tree-log.c 152.1 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
#include "tree-log.h"
24 25 26
#include "disk-io.h"
#include "locking.h"
#include "print-tree.h"
M
Mark Fasheh 已提交
27 28
#include "backref.h"
#include "hash.h"
29
#include "compression.h"
30 31 32 33 34 35 36 37 38 39

/* 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

40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82
/*
 * 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.
 */

83 84 85 86 87 88 89 90 91 92 93
/*
 * 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
94 95
#define LOG_WALK_REPLAY_DIR_INDEX 2
#define LOG_WALK_REPLAY_ALL 3
96

97
static int btrfs_log_inode(struct btrfs_trans_handle *trans,
98 99 100
			   struct btrfs_root *root, struct inode *inode,
			   int inode_only,
			   const loff_t start,
101 102
			   const loff_t end,
			   struct btrfs_log_ctx *ctx);
103 104 105
static int link_to_fixup_dir(struct btrfs_trans_handle *trans,
			     struct btrfs_root *root,
			     struct btrfs_path *path, u64 objectid);
106 107 108 109 110
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);
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

/*
 * 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,
141 142
			   struct btrfs_root *root,
			   struct btrfs_log_ctx *ctx)
143
{
144
	int ret = 0;
Y
Yan Zheng 已提交
145 146

	mutex_lock(&root->log_mutex);
147

Y
Yan Zheng 已提交
148
	if (root->log_root) {
149
		if (btrfs_need_log_full_commit(root->fs_info, trans)) {
150 151 152
			ret = -EAGAIN;
			goto out;
		}
153

154
		if (!root->log_start_pid) {
155
			clear_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state);
156
			root->log_start_pid = current->pid;
157
		} else if (root->log_start_pid != current->pid) {
158
			set_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state);
159
		}
160 161 162 163 164 165 166
	} else {
		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);
		mutex_unlock(&root->fs_info->tree_log_mutex);
		if (ret)
			goto out;
167

168
		ret = btrfs_add_log_tree(trans, root);
169
		if (ret)
170
			goto out;
171 172 173

		clear_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state);
		root->log_start_pid = current->pid;
174
	}
175

M
Miao Xie 已提交
176
	atomic_inc(&root->log_batch);
Y
Yan Zheng 已提交
177
	atomic_inc(&root->log_writers);
178
	if (ctx) {
179
		int index = root->log_transid % 2;
180
		list_add_tail(&ctx->list, &root->log_ctxs[index]);
181
		ctx->log_transid = root->log_transid;
182
	}
183

184
out:
Y
Yan Zheng 已提交
185
	mutex_unlock(&root->log_mutex);
186
	return ret;
187 188 189 190 191 192 193 194 195 196 197 198 199 200 201
}

/*
 * 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
202
	mutex_lock(&root->log_mutex);
203 204
	if (root->log_root) {
		ret = 0;
Y
Yan Zheng 已提交
205
		atomic_inc(&root->log_writers);
206
	}
Y
Yan Zheng 已提交
207
	mutex_unlock(&root->log_mutex);
208 209 210
	return ret;
}

211 212 213 214 215 216 217 218 219 220 221 222 223 224 225
/*
 * 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;
}

226 227 228 229
/*
 * indicate we're done making changes to the log tree
 * and wake up anyone waiting to do a sync
 */
230
void btrfs_end_log_trans(struct btrfs_root *root)
231
{
Y
Yan Zheng 已提交
232
	if (atomic_dec_and_test(&root->log_writers)) {
233 234 235
		/*
		 * Implicit memory barrier after atomic_dec_and_test
		 */
Y
Yan Zheng 已提交
236 237 238
		if (waitqueue_active(&root->log_writer_wait))
			wake_up(&root->log_writer_wait);
	}
239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293
}


/*
 * 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)
{
294 295
	int ret = 0;

296 297 298 299 300 301 302 303 304 305
	/*
	 * 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
306
	if (wc->pin)
307 308
		ret = btrfs_pin_extent_for_log_replay(log->fs_info->extent_root,
						      eb->start, eb->len);
309

310
	if (!ret && btrfs_buffer_uptodate(eb, gen, 0)) {
311 312
		if (wc->pin && btrfs_header_level(eb) == 0)
			ret = btrfs_exclude_logged_extents(log, eb);
313 314 315 316 317
		if (wc->write)
			btrfs_write_tree_block(eb);
		if (wc->wait)
			btrfs_wait_tree_block_writeback(eb);
	}
318
	return ret;
319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347
}

/*
 * 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;
348
	bool inode_item = key->type == BTRFS_INODE_ITEM_KEY;
349 350 351 352 353 354 355 356 357

	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);
358 359 360
	if (ret < 0)
		return ret;

361 362 363 364 365 366 367 368 369
	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) {
370
			btrfs_release_path(path);
371 372 373 374
			return 0;
		}
		dst_copy = kmalloc(item_size, GFP_NOFS);
		src_copy = kmalloc(item_size, GFP_NOFS);
375
		if (!dst_copy || !src_copy) {
376
			btrfs_release_path(path);
377 378 379 380
			kfree(dst_copy);
			kfree(src_copy);
			return -ENOMEM;
		}
381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397

		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) {
398
			btrfs_release_path(path);
399 400 401
			return 0;
		}

402 403 404 405 406 407 408
		/*
		 * 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;
409
			u32 mode;
410 411 412 413 414 415 416

			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);
417 418 419 420 421 422 423 424 425

			/*
			 * 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);
426 427 428
		}
	} else if (inode_item) {
		struct btrfs_inode_item *item;
429
		u32 mode;
430 431 432 433 434 435 436

		/*
		 * 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);
437 438 439 440 441 442 443 444 445

		/*
		 * 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);
446 447
	}
insert:
448
	btrfs_release_path(path);
449
	/* try to insert the key into the destination tree */
450
	path->skip_release_on_error = 1;
451 452
	ret = btrfs_insert_empty_item(trans, root, path,
				      key, item_size);
453
	path->skip_release_on_error = 0;
454 455

	/* make sure any existing item is the correct size */
456
	if (ret == -EEXIST || ret == -EOVERFLOW) {
457 458 459
		u32 found_size;
		found_size = btrfs_item_size_nr(path->nodes[0],
						path->slots[0]);
460
		if (found_size > item_size)
461
			btrfs_truncate_item(root, path, item_size, 1);
462
		else if (found_size < item_size)
463
			btrfs_extend_item(root, path,
464
					  item_size - found_size);
465
	} else if (ret) {
466
		return ret;
467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486
	}
	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;

487 488
		if (btrfs_inode_generation(eb, src_item) == 0) {
			struct extent_buffer *dst_eb = path->nodes[0];
489
			const u64 ino_size = btrfs_inode_size(eb, src_item);
490

491 492 493 494 495 496 497
			/*
			 * For regular files an ino_size == 0 is used only when
			 * logging that an inode exists, as part of a directory
			 * fsync, and the inode wasn't fsynced before. In this
			 * case don't set the size of the inode in the fs/subvol
			 * tree, otherwise we would be throwing valid data away.
			 */
498
			if (S_ISREG(btrfs_inode_mode(eb, src_item)) &&
499 500
			    S_ISREG(btrfs_inode_mode(dst_eb, dst_item)) &&
			    ino_size != 0) {
501 502 503 504 505 506
				struct btrfs_map_token token;

				btrfs_init_map_token(&token);
				btrfs_set_token_inode_size(dst_eb, dst_item,
							   ino_size, &token);
			}
507
			goto no_copy;
508
		}
509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538

		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]);
539
	btrfs_release_path(path);
540 541 542 543 544 545 546 547 548 549
	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)
{
550
	struct btrfs_key key;
551 552
	struct inode *inode;

553 554 555
	key.objectid = objectid;
	key.type = BTRFS_INODE_ITEM_KEY;
	key.offset = 0;
556
	inode = btrfs_iget(root->fs_info->sb, &key, root, NULL);
557 558 559
	if (IS_ERR(inode)) {
		inode = NULL;
	} else if (is_bad_inode(inode)) {
560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586
		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;
587
	u64 nbytes = 0;
588 589 590 591 592 593 594 595
	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
596
	if (found_type == BTRFS_FILE_EXTENT_REG ||
597 598 599 600 601 602 603 604 605 606 607
	    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) {
608
		size = btrfs_file_extent_inline_len(eb, slot, item);
609
		nbytes = btrfs_file_extent_ram_bytes(eb, item);
610
		extent_end = ALIGN(start + size, root->sectorsize);
611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626
	} 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
627
	ret = btrfs_lookup_file_extent(trans, root, path, btrfs_ino(inode),
628 629
				       start, 0);

Y
Yan Zheng 已提交
630 631 632
	if (ret == 0 &&
	    (found_type == BTRFS_FILE_EXTENT_REG ||
	     found_type == BTRFS_FILE_EXTENT_PREALLOC)) {
633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651
		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) {
652
			btrfs_release_path(path);
653 654 655
			goto out;
		}
	}
656
	btrfs_release_path(path);
657 658

	/* drop any overlapping extents */
659
	ret = btrfs_drop_extents(trans, root, inode, start, extent_end, 1);
660 661
	if (ret)
		goto out;
662

Y
Yan Zheng 已提交
663 664
	if (found_type == BTRFS_FILE_EXTENT_REG ||
	    found_type == BTRFS_FILE_EXTENT_PREALLOC) {
665
		u64 offset;
Y
Yan Zheng 已提交
666 667 668 669 670
		unsigned long dest_offset;
		struct btrfs_key ins;

		ret = btrfs_insert_empty_item(trans, root, path, key,
					      sizeof(*item));
671 672
		if (ret)
			goto out;
Y
Yan Zheng 已提交
673 674 675 676 677 678 679 680
		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;
681
		offset = key->offset - btrfs_file_extent_offset(eb, item);
Y
Yan Zheng 已提交
682 683 684 685 686 687 688 689 690

		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
			 */
691
			ret = btrfs_lookup_data_extent(root, ins.objectid,
Y
Yan Zheng 已提交
692 693 694 695
						ins.offset);
			if (ret == 0) {
				ret = btrfs_inc_extent_ref(trans, root,
						ins.objectid, ins.offset,
696
						0, root->root_key.objectid,
697
						key->objectid, offset);
698 699
				if (ret)
					goto out;
Y
Yan Zheng 已提交
700 701 702 703 704
			} else {
				/*
				 * insert the extent pointer in the extent
				 * allocation tree
				 */
705 706 707
				ret = btrfs_alloc_logged_file_extent(trans,
						root, root->root_key.objectid,
						key->objectid, offset, &ins);
708 709
				if (ret)
					goto out;
Y
Yan Zheng 已提交
710
			}
711
			btrfs_release_path(path);
Y
Yan Zheng 已提交
712 713 714 715 716 717 718 719 720 721 722 723 724

			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
725
						&ordered_sums, 0);
726 727
			if (ret)
				goto out;
728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776
			/*
			 * Now delete all existing cums in the csum root that
			 * cover our range. We do this because we can have an
			 * extent that is completely referenced by one file
			 * extent item and partially referenced by another
			 * file extent item (like after using the clone or
			 * extent_same ioctls). In this case if we end up doing
			 * the replay of the one that partially references the
			 * extent first, and we do not do the csum deletion
			 * below, we can get 2 csum items in the csum tree that
			 * overlap each other. For example, imagine our log has
			 * the two following file extent items:
			 *
			 * key (257 EXTENT_DATA 409600)
			 *     extent data disk byte 12845056 nr 102400
			 *     extent data offset 20480 nr 20480 ram 102400
			 *
			 * key (257 EXTENT_DATA 819200)
			 *     extent data disk byte 12845056 nr 102400
			 *     extent data offset 0 nr 102400 ram 102400
			 *
			 * Where the second one fully references the 100K extent
			 * that starts at disk byte 12845056, and the log tree
			 * has a single csum item that covers the entire range
			 * of the extent:
			 *
			 * key (EXTENT_CSUM EXTENT_CSUM 12845056) itemsize 100
			 *
			 * After the first file extent item is replayed, the
			 * csum tree gets the following csum item:
			 *
			 * key (EXTENT_CSUM EXTENT_CSUM 12865536) itemsize 20
			 *
			 * Which covers the 20K sub-range starting at offset 20K
			 * of our extent. Now when we replay the second file
			 * extent item, if we do not delete existing csum items
			 * that cover any of its blocks, we end up getting two
			 * csum items in our csum tree that overlap each other:
			 *
			 * key (EXTENT_CSUM EXTENT_CSUM 12845056) itemsize 100
			 * key (EXTENT_CSUM EXTENT_CSUM 12865536) itemsize 20
			 *
			 * Which is a problem, because after this anyone trying
			 * to lookup up for the checksum of any block of our
			 * extent starting at an offset of 40K or higher, will
			 * end up looking at the second csum item only, which
			 * does not contain the checksum for any block starting
			 * at offset 40K or higher of our extent.
			 */
Y
Yan Zheng 已提交
777 778 779 780 781
			while (!list_empty(&ordered_sums)) {
				struct btrfs_ordered_sum *sums;
				sums = list_entry(ordered_sums.next,
						struct btrfs_ordered_sum,
						list);
782 783 784 785 786
				if (!ret)
					ret = btrfs_del_csums(trans,
						      root->fs_info->csum_root,
						      sums->bytenr,
						      sums->len);
787 788
				if (!ret)
					ret = btrfs_csum_file_blocks(trans,
Y
Yan Zheng 已提交
789 790 791 792 793
						root->fs_info->csum_root,
						sums);
				list_del(&sums->list);
				kfree(sums);
			}
794 795
			if (ret)
				goto out;
Y
Yan Zheng 已提交
796
		} else {
797
			btrfs_release_path(path);
Y
Yan Zheng 已提交
798 799 800 801
		}
	} 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);
802 803
		if (ret)
			goto out;
Y
Yan Zheng 已提交
804
	}
805

806
	inode_add_bytes(inode, nbytes);
807
	ret = btrfs_update_inode(trans, root, inode);
808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839
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);
840 841 842
	if (!name)
		return -ENOMEM;

843
	read_extent_buffer(leaf, name, (unsigned long)(di + 1), name_len);
844
	btrfs_release_path(path);
845 846

	inode = read_one_inode(root, location.objectid);
847
	if (!inode) {
848 849
		ret = -EIO;
		goto out;
850
	}
851

852
	ret = link_to_fixup_dir(trans, root, path, location.objectid);
853 854
	if (ret)
		goto out;
855

856
	ret = btrfs_unlink_inode(trans, root, dir, inode, name, name_len);
857 858
	if (ret)
		goto out;
859 860
	else
		ret = btrfs_run_delayed_items(trans, root);
861
out:
862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888
	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;
889
	btrfs_release_path(path);
890 891 892 893 894 895 896 897 898 899

	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:
900
	btrfs_release_path(path);
901 902 903 904 905 906 907 908 909 910 911 912 913 914 915
	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
916
				   u64 ref_objectid,
917
				   const char *name, int namelen)
918 919 920 921 922 923 924 925 926 927 928 929
{
	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();
930 931 932
	if (!path)
		return -ENOMEM;

933 934 935 936 937
	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
938 939 940 941 942 943 944 945 946 947

	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]);
948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967
	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;
}

968
static inline int __add_inode_ref(struct btrfs_trans_handle *trans,
969 970
				  struct btrfs_root *root,
				  struct btrfs_path *path,
971 972 973
				  struct btrfs_root *log_root,
				  struct inode *dir, struct inode *inode,
				  struct extent_buffer *eb,
M
Mark Fasheh 已提交
974 975 976
				  u64 inode_objectid, u64 parent_objectid,
				  u64 ref_index, char *name, int namelen,
				  int *search_done)
977
{
L
liubo 已提交
978
	int ret;
M
Mark Fasheh 已提交
979 980 981
	char *victim_name;
	int victim_name_len;
	struct extent_buffer *leaf;
982
	struct btrfs_dir_item *di;
M
Mark Fasheh 已提交
983 984
	struct btrfs_key search_key;
	struct btrfs_inode_extref *extref;
985

M
Mark Fasheh 已提交
986 987 988 989 990 991
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);
992 993 994 995
	if (ret == 0) {
		struct btrfs_inode_ref *victim_ref;
		unsigned long ptr;
		unsigned long ptr_end;
M
Mark Fasheh 已提交
996 997

		leaf = path->nodes[0];
998 999 1000 1001

		/* 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
1002
		if (search_key.objectid == search_key.offset)
1003
			return 1;
1004 1005 1006 1007 1008 1009 1010

		/* 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
1011
		while (ptr < ptr_end) {
1012 1013 1014 1015
			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);
1016 1017
			if (!victim_name)
				return -ENOMEM;
1018 1019 1020 1021 1022

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

M
Mark Fasheh 已提交
1023 1024 1025
			if (!backref_in_log(log_root, &search_key,
					    parent_objectid,
					    victim_name,
1026
					    victim_name_len)) {
Z
Zach Brown 已提交
1027
				inc_nlink(inode);
1028
				btrfs_release_path(path);
1029

1030 1031 1032
				ret = btrfs_unlink_inode(trans, root, dir,
							 inode, victim_name,
							 victim_name_len);
M
Mark Fasheh 已提交
1033
				kfree(victim_name);
1034 1035
				if (ret)
					return ret;
1036 1037 1038
				ret = btrfs_run_delayed_items(trans, root);
				if (ret)
					return ret;
M
Mark Fasheh 已提交
1039 1040
				*search_done = 1;
				goto again;
1041 1042
			}
			kfree(victim_name);
M
Mark Fasheh 已提交
1043

1044 1045 1046
			ptr = (unsigned long)(victim_ref + 1) + victim_name_len;
		}

1047 1048
		/*
		 * NOTE: we have searched root tree and checked the
1049
		 * corresponding ref, it does not need to check again.
1050
		 */
1051
		*search_done = 1;
1052
	}
1053
	btrfs_release_path(path);
1054

M
Mark Fasheh 已提交
1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070
	/* 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) {
1071
			extref = (struct btrfs_inode_extref *)(base + cur_offset);
M
Mark Fasheh 已提交
1072 1073 1074 1075 1076 1077 1078

			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);
1079 1080
			if (!victim_name)
				return -ENOMEM;
M
Mark Fasheh 已提交
1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096
			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
1097
					inc_nlink(inode);
M
Mark Fasheh 已提交
1098 1099 1100 1101 1102 1103 1104
					btrfs_release_path(path);

					ret = btrfs_unlink_inode(trans, root,
								 victim_parent,
								 inode,
								 victim_name,
								 victim_name_len);
1105 1106 1107
					if (!ret)
						ret = btrfs_run_delayed_items(
								  trans, root);
M
Mark Fasheh 已提交
1108 1109 1110
				}
				iput(victim_parent);
				kfree(victim_name);
1111 1112
				if (ret)
					return ret;
M
Mark Fasheh 已提交
1113 1114 1115 1116
				*search_done = 1;
				goto again;
			}
			kfree(victim_name);
1117 1118
			if (ret)
				return ret;
M
Mark Fasheh 已提交
1119 1120 1121 1122 1123 1124 1125
next:
			cur_offset += victim_name_len + sizeof(*extref);
		}
		*search_done = 1;
	}
	btrfs_release_path(path);

L
liubo 已提交
1126 1127
	/* look for a conflicting sequence number */
	di = btrfs_lookup_dir_index_item(trans, root, path, btrfs_ino(dir),
M
Mark Fasheh 已提交
1128
					 ref_index, name, namelen, 0);
L
liubo 已提交
1129 1130
	if (di && !IS_ERR(di)) {
		ret = drop_one_dir_item(trans, root, path, dir, di);
1131 1132
		if (ret)
			return ret;
L
liubo 已提交
1133 1134 1135 1136 1137 1138 1139 1140
	}
	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);
1141 1142
		if (ret)
			return ret;
L
liubo 已提交
1143 1144 1145
	}
	btrfs_release_path(path);

1146 1147
	return 0;
}
1148

M
Mark Fasheh 已提交
1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190
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;
}

1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203
/*
 * 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)
{
1204 1205
	struct inode *dir = NULL;
	struct inode *inode = NULL;
1206 1207
	unsigned long ref_ptr;
	unsigned long ref_end;
1208
	char *name = NULL;
1209 1210 1211
	int namelen;
	int ret;
	int search_done = 0;
M
Mark Fasheh 已提交
1212 1213 1214
	int log_ref_ver = 0;
	u64 parent_objectid;
	u64 inode_objectid;
1215
	u64 ref_index = 0;
M
Mark Fasheh 已提交
1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232
	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;
1233

1234 1235 1236 1237 1238 1239
	/*
	 * 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
1240
	dir = read_one_inode(root, parent_objectid);
1241 1242 1243 1244
	if (!dir) {
		ret = -ENOENT;
		goto out;
	}
1245

M
Mark Fasheh 已提交
1246
	inode = read_one_inode(root, inode_objectid);
1247
	if (!inode) {
1248 1249
		ret = -EIO;
		goto out;
1250 1251 1252
	}

	while (ref_ptr < ref_end) {
M
Mark Fasheh 已提交
1253 1254 1255 1256 1257 1258 1259 1260 1261
		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);
1262 1263 1264 1265
			if (!dir) {
				ret = -ENOENT;
				goto out;
			}
M
Mark Fasheh 已提交
1266 1267 1268 1269 1270
		} else {
			ret = ref_get_fields(eb, ref_ptr, &namelen, &name,
					     &ref_index);
		}
		if (ret)
1271
			goto out;
1272 1273 1274

		/* 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
1275
				  ref_index, name, namelen)) {
1276 1277 1278 1279 1280 1281 1282 1283 1284 1285
			/*
			 * 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
1286 1287 1288 1289
						      dir, inode, eb,
						      inode_objectid,
						      parent_objectid,
						      ref_index, name, namelen,
1290
						      &search_done);
1291 1292 1293
				if (ret) {
					if (ret == 1)
						ret = 0;
1294 1295
					goto out;
				}
1296 1297 1298 1299
			}

			/* insert our name */
			ret = btrfs_add_link(trans, dir, inode, name, namelen,
M
Mark Fasheh 已提交
1300
					     0, ref_index);
1301 1302
			if (ret)
				goto out;
1303 1304 1305 1306

			btrfs_update_inode(trans, root, inode);
		}

M
Mark Fasheh 已提交
1307
		ref_ptr = (unsigned long)(ref_ptr + ref_struct_size) + namelen;
1308
		kfree(name);
1309
		name = NULL;
M
Mark Fasheh 已提交
1310 1311 1312 1313
		if (log_ref_ver) {
			iput(dir);
			dir = NULL;
		}
1314
	}
1315 1316 1317

	/* finally write the back reference in the inode */
	ret = overwrite_item(trans, root, path, eb, slot, key);
1318
out:
1319
	btrfs_release_path(path);
1320
	kfree(name);
1321 1322
	iput(dir);
	iput(inode);
1323
	return ret;
1324 1325
}

1326
static int insert_orphan_item(struct btrfs_trans_handle *trans,
1327
			      struct btrfs_root *root, u64 ino)
1328 1329
{
	int ret;
1330

1331 1332 1333
	ret = btrfs_insert_orphan_item(trans, root, ino);
	if (ret == -EEXIST)
		ret = 0;
1334

1335 1336 1337
	return ret;
}

M
Mark Fasheh 已提交
1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356
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;
1357

M
Mark Fasheh 已提交
1358 1359 1360
		leaf = path->nodes[0];
		item_size = btrfs_item_size_nr(leaf, path->slots[0]);
		ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
1361
		cur_offset = 0;
M
Mark Fasheh 已提交
1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376

		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);

1377
	if (ret < 0 && ret != -ENOENT)
M
Mark Fasheh 已提交
1378 1379 1380 1381 1382 1383
		return ret;
	return nlink;
}

static int count_inode_refs(struct btrfs_root *root,
			       struct inode *inode, struct btrfs_path *path)
1384 1385 1386
{
	int ret;
	struct btrfs_key key;
M
Mark Fasheh 已提交
1387
	unsigned int nlink = 0;
1388 1389 1390
	unsigned long ptr;
	unsigned long ptr_end;
	int name_len;
L
Li Zefan 已提交
1391
	u64 ino = btrfs_ino(inode);
1392

L
Li Zefan 已提交
1393
	key.objectid = ino;
1394 1395 1396
	key.type = BTRFS_INODE_REF_KEY;
	key.offset = (u64)-1;

C
Chris Mason 已提交
1397
	while (1) {
1398 1399 1400 1401 1402 1403 1404 1405
		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]--;
		}
1406
process_slot:
1407 1408
		btrfs_item_key_to_cpu(path->nodes[0], &key,
				      path->slots[0]);
L
Li Zefan 已提交
1409
		if (key.objectid != ino ||
1410 1411 1412 1413 1414
		    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
1415
		while (ptr < ptr_end) {
1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426
			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;
1427 1428 1429 1430
		if (path->slots[0] > 0) {
			path->slots[0]--;
			goto process_slot;
		}
1431
		key.offset--;
1432
		btrfs_release_path(path);
1433
	}
1434
	btrfs_release_path(path);
M
Mark Fasheh 已提交
1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475

	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 < 0)
		goto out;

	nlink += ret;

	ret = 0;

1476
	if (nlink != inode->i_nlink) {
M
Miklos Szeredi 已提交
1477
		set_nlink(inode, nlink);
1478 1479
		btrfs_update_inode(trans, root, inode);
	}
1480
	BTRFS_I(inode)->index_cnt = (u64)-1;
1481

1482 1483 1484
	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
1485
						 ino, 1);
1486 1487
			if (ret)
				goto out;
1488
		}
L
Li Zefan 已提交
1489
		ret = insert_orphan_item(trans, root, ino);
1490 1491
	}

M
Mark Fasheh 已提交
1492 1493 1494
out:
	btrfs_free_path(path);
	return ret;
1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507
}

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
1508
	while (1) {
1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524
		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);
1525 1526
		if (ret)
			goto out;
1527

1528
		btrfs_release_path(path);
1529
		inode = read_one_inode(root, key.offset);
1530 1531
		if (!inode)
			return -EIO;
1532 1533 1534

		ret = fixup_inode_link_count(trans, root, inode);
		iput(inode);
1535 1536
		if (ret)
			goto out;
1537

1538 1539 1540 1541 1542 1543
		/*
		 * fixup on a directory may create new entries,
		 * make sure we always look for the highset possible
		 * offset
		 */
		key.offset = (u64)-1;
1544
	}
1545 1546
	ret = 0;
out:
1547
	btrfs_release_path(path);
1548
	return ret;
1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566
}


/*
 * 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);
1567 1568
	if (!inode)
		return -EIO;
1569 1570

	key.objectid = BTRFS_TREE_LOG_FIXUP_OBJECTID;
1571
	key.type = BTRFS_ORPHAN_ITEM_KEY;
1572 1573 1574 1575
	key.offset = objectid;

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

1576
	btrfs_release_path(path);
1577
	if (ret == 0) {
1578 1579 1580
		if (!inode->i_nlink)
			set_nlink(inode, 1);
		else
Z
Zach Brown 已提交
1581
			inc_nlink(inode);
1582
		ret = btrfs_update_inode(trans, root, inode);
1583 1584 1585
	} else if (ret == -EEXIST) {
		ret = 0;
	} else {
1586
		BUG(); /* Logic Error */
1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600
	}
	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,
				    u64 dirid, u64 index,
1601
				    char *name, int name_len,
1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616
				    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;
	}
1617

1618 1619 1620 1621 1622 1623 1624 1625 1626
	ret = btrfs_add_link(trans, dir, inode, name, name_len, 1, index);

	/* FIXME, put inode into FIXUP list */

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

1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650
/*
 * Return true if an inode reference exists in the log for the given name,
 * inode and parent inode.
 */
static bool name_in_log_ref(struct btrfs_root *log_root,
			    const char *name, const int name_len,
			    const u64 dirid, const u64 ino)
{
	struct btrfs_key search_key;

	search_key.objectid = ino;
	search_key.type = BTRFS_INODE_REF_KEY;
	search_key.offset = dirid;
	if (backref_in_log(log_root, &search_key, dirid, name, name_len))
		return true;

	search_key.type = BTRFS_INODE_EXTREF_KEY;
	search_key.offset = btrfs_extref_hash(dirid, name, name_len);
	if (backref_in_log(log_root, &search_key, dirid, name, name_len))
		return true;

	return false;
}

1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662
/*
 * 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.
1663 1664 1665
 *
 * Returns < 0 on error, 0 if the name wasn't replayed (dentry points to a
 * non-existing inode) and 1 if the name was replayed.
1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680
 */
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
1681
	int exists;
1682
	int ret = 0;
1683
	bool update_size = (key->type == BTRFS_DIR_INDEX_KEY);
1684
	bool name_added = false;
1685 1686

	dir = read_one_inode(root, key->objectid);
1687 1688
	if (!dir)
		return -EIO;
1689 1690 1691

	name_len = btrfs_dir_name_len(eb, di);
	name = kmalloc(name_len, GFP_NOFS);
1692 1693 1694 1695
	if (!name) {
		ret = -ENOMEM;
		goto out;
	}
1696

1697 1698 1699 1700 1701
	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
1702 1703 1704 1705 1706
	exists = btrfs_lookup_inode(trans, root, path, &log_key, 0);
	if (exists == 0)
		exists = 1;
	else
		exists = 0;
1707
	btrfs_release_path(path);
C
Chris Mason 已提交
1708

1709 1710 1711
	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
1712
	} else if (key->type == BTRFS_DIR_INDEX_KEY) {
1713 1714 1715 1716 1717
		dst_di = btrfs_lookup_dir_index_item(trans, root, path,
						     key->objectid,
						     key->offset, name,
						     name_len, 1);
	} else {
1718 1719 1720
		/* Corruption */
		ret = -EINVAL;
		goto out;
1721
	}
1722
	if (IS_ERR_OR_NULL(dst_di)) {
1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736
		/* 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) {
1737
		update_size = false;
1738 1739 1740 1741 1742 1743 1744
		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
1745
	if (!exists)
1746 1747 1748
		goto out;

	ret = drop_one_dir_item(trans, root, path, dir, dst_di);
1749 1750
	if (ret)
		goto out;
1751 1752 1753 1754

	if (key->type == BTRFS_DIR_INDEX_KEY)
		goto insert;
out:
1755
	btrfs_release_path(path);
1756 1757 1758 1759
	if (!ret && update_size) {
		btrfs_i_size_write(dir, dir->i_size + name_len * 2);
		ret = btrfs_update_inode(trans, root, dir);
	}
1760 1761
	kfree(name);
	iput(dir);
1762 1763
	if (!ret && name_added)
		ret = 1;
1764
	return ret;
1765 1766

insert:
1767 1768 1769 1770 1771 1772 1773
	if (name_in_log_ref(root->log_root, name, name_len,
			    key->objectid, log_key.objectid)) {
		/* The dentry will be added later. */
		ret = 0;
		update_size = false;
		goto out;
	}
1774
	btrfs_release_path(path);
1775 1776
	ret = insert_one_name(trans, root, key->objectid, key->offset,
			      name, name_len, &log_key);
1777
	if (ret && ret != -ENOENT && ret != -EEXIST)
1778
		goto out;
1779 1780
	if (!ret)
		name_added = true;
1781
	update_size = false;
1782
	ret = 0;
1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797
	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)
{
1798
	int ret = 0;
1799 1800 1801 1802 1803
	u32 item_size = btrfs_item_size_nr(eb, slot);
	struct btrfs_dir_item *di;
	int name_len;
	unsigned long ptr;
	unsigned long ptr_end;
1804
	struct btrfs_path *fixup_path = NULL;
1805 1806 1807

	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
		if (verify_dir_item(root, eb, di))
			return -EIO;
1812 1813
		name_len = btrfs_dir_name_len(eb, di);
		ret = replay_one_name(trans, root, path, eb, di, key);
1814 1815
		if (ret < 0)
			break;
1816 1817
		ptr = (unsigned long)(di + 1);
		ptr += name_len;
1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863

		/*
		 * If this entry refers to a non-directory (directories can not
		 * have a link count > 1) and it was added in the transaction
		 * that was not committed, make sure we fixup the link count of
		 * the inode it the entry points to. Otherwise something like
		 * the following would result in a directory pointing to an
		 * inode with a wrong link that does not account for this dir
		 * entry:
		 *
		 * mkdir testdir
		 * touch testdir/foo
		 * touch testdir/bar
		 * sync
		 *
		 * ln testdir/bar testdir/bar_link
		 * ln testdir/foo testdir/foo_link
		 * xfs_io -c "fsync" testdir/bar
		 *
		 * <power failure>
		 *
		 * mount fs, log replay happens
		 *
		 * File foo would remain with a link count of 1 when it has two
		 * entries pointing to it in the directory testdir. This would
		 * make it impossible to ever delete the parent directory has
		 * it would result in stale dentries that can never be deleted.
		 */
		if (ret == 1 && btrfs_dir_type(eb, di) != BTRFS_FT_DIR) {
			struct btrfs_key di_key;

			if (!fixup_path) {
				fixup_path = btrfs_alloc_path();
				if (!fixup_path) {
					ret = -ENOMEM;
					break;
				}
			}

			btrfs_dir_item_key_to_cpu(eb, di, &di_key);
			ret = link_to_fixup_dir(trans, root, fixup_path,
						di_key.objectid);
			if (ret)
				break;
		}
		ret = 0;
1864
	}
1865 1866
	btrfs_free_path(fixup_path);
	return ret;
1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 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 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947
}

/*
 * 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:
1948
	btrfs_release_path(path);
1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983
	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
1984
	while (ptr < ptr_end) {
1985
		di = (struct btrfs_dir_item *)ptr;
1986 1987 1988 1989 1990
		if (verify_dir_item(root, eb, di)) {
			ret = -EIO;
			goto out;
		}

1991 1992 1993 1994 1995 1996 1997 1998 1999
		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;
2000
		if (log && dir_key->type == BTRFS_DIR_ITEM_KEY) {
2001 2002 2003
			log_di = btrfs_lookup_dir_item(trans, log, log_path,
						       dir_key->objectid,
						       name, name_len, 0);
2004
		} else if (log && dir_key->type == BTRFS_DIR_INDEX_KEY) {
2005 2006 2007 2008 2009 2010
			log_di = btrfs_lookup_dir_index_item(trans, log,
						     log_path,
						     dir_key->objectid,
						     dir_key->offset,
						     name, name_len, 0);
		}
2011
		if (!log_di || (IS_ERR(log_di) && PTR_ERR(log_di) == -ENOENT)) {
2012
			btrfs_dir_item_key_to_cpu(eb, di, &location);
2013 2014
			btrfs_release_path(path);
			btrfs_release_path(log_path);
2015
			inode = read_one_inode(root, location.objectid);
2016 2017 2018 2019
			if (!inode) {
				kfree(name);
				return -EIO;
			}
2020 2021 2022

			ret = link_to_fixup_dir(trans, root,
						path, location.objectid);
2023 2024 2025 2026 2027 2028
			if (ret) {
				kfree(name);
				iput(inode);
				goto out;
			}

Z
Zach Brown 已提交
2029
			inc_nlink(inode);
2030 2031
			ret = btrfs_unlink_inode(trans, root, dir, inode,
						 name, name_len);
2032
			if (!ret)
2033
				ret = btrfs_run_delayed_items(trans, root);
2034 2035
			kfree(name);
			iput(inode);
2036 2037
			if (ret)
				goto out;
2038 2039 2040 2041 2042 2043 2044 2045 2046 2047

			/* 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;
2048 2049 2050
		} else if (IS_ERR(log_di)) {
			kfree(name);
			return PTR_ERR(log_di);
2051
		}
2052
		btrfs_release_path(log_path);
2053 2054 2055 2056 2057 2058 2059
		kfree(name);

		ptr = (unsigned long)(di + 1);
		ptr += name_len;
	}
	ret = 0;
out:
2060 2061
	btrfs_release_path(path);
	btrfs_release_path(log_path);
2062 2063 2064
	return ret;
}

2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162
static int replay_xattr_deletes(struct btrfs_trans_handle *trans,
			      struct btrfs_root *root,
			      struct btrfs_root *log,
			      struct btrfs_path *path,
			      const u64 ino)
{
	struct btrfs_key search_key;
	struct btrfs_path *log_path;
	int i;
	int nritems;
	int ret;

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

	search_key.objectid = ino;
	search_key.type = BTRFS_XATTR_ITEM_KEY;
	search_key.offset = 0;
again:
	ret = btrfs_search_slot(NULL, root, &search_key, path, 0, 0);
	if (ret < 0)
		goto out;
process_leaf:
	nritems = btrfs_header_nritems(path->nodes[0]);
	for (i = path->slots[0]; i < nritems; i++) {
		struct btrfs_key key;
		struct btrfs_dir_item *di;
		struct btrfs_dir_item *log_di;
		u32 total_size;
		u32 cur;

		btrfs_item_key_to_cpu(path->nodes[0], &key, i);
		if (key.objectid != ino || key.type != BTRFS_XATTR_ITEM_KEY) {
			ret = 0;
			goto out;
		}

		di = btrfs_item_ptr(path->nodes[0], i, struct btrfs_dir_item);
		total_size = btrfs_item_size_nr(path->nodes[0], i);
		cur = 0;
		while (cur < total_size) {
			u16 name_len = btrfs_dir_name_len(path->nodes[0], di);
			u16 data_len = btrfs_dir_data_len(path->nodes[0], di);
			u32 this_len = sizeof(*di) + name_len + data_len;
			char *name;

			name = kmalloc(name_len, GFP_NOFS);
			if (!name) {
				ret = -ENOMEM;
				goto out;
			}
			read_extent_buffer(path->nodes[0], name,
					   (unsigned long)(di + 1), name_len);

			log_di = btrfs_lookup_xattr(NULL, log, log_path, ino,
						    name, name_len, 0);
			btrfs_release_path(log_path);
			if (!log_di) {
				/* Doesn't exist in log tree, so delete it. */
				btrfs_release_path(path);
				di = btrfs_lookup_xattr(trans, root, path, ino,
							name, name_len, -1);
				kfree(name);
				if (IS_ERR(di)) {
					ret = PTR_ERR(di);
					goto out;
				}
				ASSERT(di);
				ret = btrfs_delete_one_dir_name(trans, root,
								path, di);
				if (ret)
					goto out;
				btrfs_release_path(path);
				search_key = key;
				goto again;
			}
			kfree(name);
			if (IS_ERR(log_di)) {
				ret = PTR_ERR(log_di);
				goto out;
			}
			cur += this_len;
			di = (struct btrfs_dir_item *)((char *)di + this_len);
		}
	}
	ret = btrfs_next_leaf(root, path);
	if (ret > 0)
		ret = 0;
	else if (ret == 0)
		goto process_leaf;
out:
	btrfs_free_path(log_path);
	btrfs_release_path(path);
	return ret;
}


2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176
/*
 * 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,
2177
				       u64 dirid, int del_all)
2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205
{
	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
2206
	while (1) {
2207 2208 2209 2210 2211 2212 2213 2214
		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;
		}
2215 2216

		dir_key.offset = range_start;
C
Chris Mason 已提交
2217
		while (1) {
2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239
			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,
2240 2241
						log_path, dir,
						&found_key);
2242 2243
			if (ret)
				goto out;
2244 2245 2246 2247
			if (found_key.offset == (u64)-1)
				break;
			dir_key.offset = found_key.offset + 1;
		}
2248
		btrfs_release_path(path);
2249 2250 2251 2252 2253 2254 2255 2256 2257 2258
		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;
2259
		btrfs_release_path(path);
2260 2261 2262
		goto again;
	}
out:
2263
	btrfs_release_path(path);
2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290
	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;

2291 2292 2293
	ret = btrfs_read_buffer(eb, gen);
	if (ret)
		return ret;
2294 2295 2296 2297 2298 2299 2300

	level = btrfs_header_level(eb);

	if (level != 0)
		return 0;

	path = btrfs_alloc_path();
2301 2302
	if (!path)
		return -ENOMEM;
2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315

	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);
2316 2317 2318 2319
			ret = replay_xattr_deletes(wc->trans, root, log,
						   path, key.objectid);
			if (ret)
				break;
2320 2321 2322
			mode = btrfs_inode_mode(eb, inode_item);
			if (S_ISDIR(mode)) {
				ret = replay_dir_deletes(wc->trans,
2323
					 root, log, path, key.objectid, 0);
2324 2325
				if (ret)
					break;
2326 2327 2328
			}
			ret = overwrite_item(wc->trans, root, path,
					     eb, i, &key);
2329 2330
			if (ret)
				break;
2331

2332
			/* for regular files, make sure corresponding
2333
			 * orphan item exist. extents past the new EOF
2334
			 * will be truncated later by orphan cleanup.
2335 2336
			 */
			if (S_ISREG(mode)) {
2337 2338
				ret = insert_orphan_item(wc->trans, root,
							 key.objectid);
2339 2340
				if (ret)
					break;
2341
			}
2342

2343 2344
			ret = link_to_fixup_dir(wc->trans, root,
						path, key.objectid);
2345 2346
			if (ret)
				break;
2347
		}
2348 2349 2350 2351 2352 2353 2354 2355 2356

		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;
		}

2357 2358 2359 2360 2361 2362 2363
		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);
2364 2365
			if (ret)
				break;
2366 2367
		} else if (key.type == BTRFS_INODE_REF_KEY ||
			   key.type == BTRFS_INODE_EXTREF_KEY) {
M
Mark Fasheh 已提交
2368 2369
			ret = add_inode_ref(wc->trans, root, log, path,
					    eb, i, &key);
2370 2371 2372
			if (ret && ret != -ENOENT)
				break;
			ret = 0;
2373 2374 2375
		} else if (key.type == BTRFS_EXTENT_DATA_KEY) {
			ret = replay_one_extent(wc->trans, root, path,
						eb, i, &key);
2376 2377
			if (ret)
				break;
2378
		} else if (key.type == BTRFS_DIR_ITEM_KEY) {
2379 2380
			ret = replay_one_dir_item(wc->trans, root, path,
						  eb, i, &key);
2381 2382
			if (ret)
				break;
2383 2384 2385
		}
	}
	btrfs_free_path(path);
2386
	return ret;
2387 2388
}

C
Chris Mason 已提交
2389
static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans,
2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405
				   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
2406
	while (*level > 0) {
2407 2408 2409 2410
		WARN_ON(*level < 0);
		WARN_ON(*level >= BTRFS_MAX_LEVEL);
		cur = path->nodes[*level];

2411
		WARN_ON(btrfs_header_level(cur) != *level);
2412 2413 2414 2415 2416 2417 2418

		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]);
2419
		blocksize = root->nodesize;
2420 2421 2422 2423

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

2424
		next = btrfs_find_create_tree_block(root, bytenr);
2425 2426
		if (IS_ERR(next))
			return PTR_ERR(next);
2427 2428

		if (*level == 1) {
2429
			ret = wc->process_func(root, next, wc, ptr_gen);
2430 2431
			if (ret) {
				free_extent_buffer(next);
2432
				return ret;
2433
			}
2434

2435 2436
			path->slots[*level]++;
			if (wc->free) {
2437 2438 2439 2440 2441
				ret = btrfs_read_buffer(next, ptr_gen);
				if (ret) {
					free_extent_buffer(next);
					return ret;
				}
2442

2443 2444 2445
				if (trans) {
					btrfs_tree_lock(next);
					btrfs_set_lock_blocking(next);
2446 2447
					clean_tree_block(trans, root->fs_info,
							next);
2448 2449 2450
					btrfs_wait_tree_block_writeback(next);
					btrfs_tree_unlock(next);
				}
2451 2452 2453

				WARN_ON(root_owner !=
					BTRFS_TREE_LOG_OBJECTID);
2454
				ret = btrfs_free_and_pin_reserved_extent(root,
2455
							 bytenr, blocksize);
2456 2457 2458 2459
				if (ret) {
					free_extent_buffer(next);
					return ret;
				}
2460 2461 2462 2463
			}
			free_extent_buffer(next);
			continue;
		}
2464 2465 2466 2467 2468
		ret = btrfs_read_buffer(next, ptr_gen);
		if (ret) {
			free_extent_buffer(next);
			return ret;
		}
2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480

		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);

2481
	path->slots[*level] = btrfs_header_nritems(path->nodes[*level]);
2482 2483 2484 2485 2486

	cond_resched();
	return 0;
}

C
Chris Mason 已提交
2487
static noinline int walk_up_log_tree(struct btrfs_trans_handle *trans,
2488 2489 2490 2491 2492 2493 2494 2495 2496
				 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
2497
	for (i = *level; i < BTRFS_MAX_LEVEL - 1 && path->nodes[i]; i++) {
2498
		slot = path->slots[i];
2499
		if (slot + 1 < btrfs_header_nritems(path->nodes[i])) {
2500 2501 2502 2503 2504
			path->slots[i]++;
			*level = i;
			WARN_ON(*level == 0);
			return 0;
		} else {
Z
Zheng Yan 已提交
2505 2506 2507 2508 2509 2510 2511
			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);
2512
			ret = wc->process_func(root, path->nodes[*level], wc,
2513
				 btrfs_header_generation(path->nodes[*level]));
2514 2515 2516
			if (ret)
				return ret;

2517 2518 2519 2520 2521
			if (wc->free) {
				struct extent_buffer *next;

				next = path->nodes[*level];

2522 2523 2524
				if (trans) {
					btrfs_tree_lock(next);
					btrfs_set_lock_blocking(next);
2525 2526
					clean_tree_block(trans, root->fs_info,
							next);
2527 2528 2529
					btrfs_wait_tree_block_writeback(next);
					btrfs_tree_unlock(next);
				}
2530 2531

				WARN_ON(root_owner != BTRFS_TREE_LOG_OBJECTID);
2532
				ret = btrfs_free_and_pin_reserved_extent(root,
2533
						path->nodes[*level]->start,
2534
						path->nodes[*level]->len);
2535 2536
				if (ret)
					return ret;
2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560
			}
			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
2561 2562
	if (!path)
		return -ENOMEM;
2563 2564 2565 2566 2567 2568 2569

	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
2570
	while (1) {
2571 2572 2573
		wret = walk_down_log_tree(trans, log, path, &level, wc);
		if (wret > 0)
			break;
2574
		if (wret < 0) {
2575
			ret = wret;
2576 2577
			goto out;
		}
2578 2579 2580 2581

		wret = walk_up_log_tree(trans, log, path, &level, wc);
		if (wret > 0)
			break;
2582
		if (wret < 0) {
2583
			ret = wret;
2584 2585
			goto out;
		}
2586 2587 2588 2589
	}

	/* was the root node processed? if not, catch it here */
	if (path->nodes[orig_level]) {
2590
		ret = wc->process_func(log, path->nodes[orig_level], wc,
2591
			 btrfs_header_generation(path->nodes[orig_level]));
2592 2593
		if (ret)
			goto out;
2594 2595 2596 2597 2598
		if (wc->free) {
			struct extent_buffer *next;

			next = path->nodes[orig_level];

2599 2600 2601
			if (trans) {
				btrfs_tree_lock(next);
				btrfs_set_lock_blocking(next);
2602
				clean_tree_block(trans, log->fs_info, next);
2603 2604 2605
				btrfs_wait_tree_block_writeback(next);
				btrfs_tree_unlock(next);
			}
2606 2607 2608

			WARN_ON(log->root_key.objectid !=
				BTRFS_TREE_LOG_OBJECTID);
2609
			ret = btrfs_free_and_pin_reserved_extent(log, next->start,
2610
							 next->len);
2611 2612
			if (ret)
				goto out;
2613 2614 2615
		}
	}

2616
out:
2617 2618 2619 2620
	btrfs_free_path(path);
	return ret;
}

Y
Yan Zheng 已提交
2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640
/*
 * 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;
}

2641
static void wait_log_commit(struct btrfs_root *root, int transid)
2642 2643
{
	DEFINE_WAIT(wait);
Y
Yan Zheng 已提交
2644
	int index = transid % 2;
2645

Y
Yan Zheng 已提交
2646 2647 2648 2649 2650
	/*
	 * 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
	 */
2651
	do {
Y
Yan Zheng 已提交
2652 2653 2654
		prepare_to_wait(&root->log_commit_wait[index],
				&wait, TASK_UNINTERRUPTIBLE);
		mutex_unlock(&root->log_mutex);
2655

2656
		if (root->log_transid_committed < transid &&
Y
Yan Zheng 已提交
2657 2658
		    atomic_read(&root->log_commit[index]))
			schedule();
2659

Y
Yan Zheng 已提交
2660 2661
		finish_wait(&root->log_commit_wait[index], &wait);
		mutex_lock(&root->log_mutex);
2662
	} while (root->log_transid_committed < transid &&
Y
Yan Zheng 已提交
2663 2664 2665
		 atomic_read(&root->log_commit[index]));
}

2666
static void wait_for_writer(struct btrfs_root *root)
Y
Yan Zheng 已提交
2667 2668
{
	DEFINE_WAIT(wait);
2669 2670

	while (atomic_read(&root->log_writers)) {
Y
Yan Zheng 已提交
2671 2672 2673
		prepare_to_wait(&root->log_writer_wait,
				&wait, TASK_UNINTERRUPTIBLE);
		mutex_unlock(&root->log_mutex);
2674
		if (atomic_read(&root->log_writers))
2675
			schedule();
Y
Yan Zheng 已提交
2676
		finish_wait(&root->log_writer_wait, &wait);
2677
		mutex_lock(&root->log_mutex);
Y
Yan Zheng 已提交
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 2707 2708 2709 2710 2711
static inline void btrfs_remove_log_ctx(struct btrfs_root *root,
					struct btrfs_log_ctx *ctx)
{
	if (!ctx)
		return;

	mutex_lock(&root->log_mutex);
	list_del_init(&ctx->list);
	mutex_unlock(&root->log_mutex);
}

/* 
 * Invoked in log mutex context, or be sure there is no other task which
 * can access the list.
 */
static inline void btrfs_remove_all_log_ctxs(struct btrfs_root *root,
					     int index, int error)
{
	struct btrfs_log_ctx *ctx;

	if (!error) {
		INIT_LIST_HEAD(&root->log_ctxs[index]);
		return;
	}

	list_for_each_entry(ctx, &root->log_ctxs[index], list)
		ctx->log_ret = error;

	INIT_LIST_HEAD(&root->log_ctxs[index]);
}

2712 2713 2714
/*
 * 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,
2715 2716 2717 2718 2719 2720 2721 2722
 * 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.
2723 2724
 */
int btrfs_sync_log(struct btrfs_trans_handle *trans,
2725
		   struct btrfs_root *root, struct btrfs_log_ctx *ctx)
2726
{
Y
Yan Zheng 已提交
2727 2728
	int index1;
	int index2;
2729
	int mark;
2730 2731
	int ret;
	struct btrfs_root *log = root->log_root;
Y
Yan Zheng 已提交
2732
	struct btrfs_root *log_root_tree = root->fs_info->log_root_tree;
2733
	int log_transid = 0;
2734
	struct btrfs_log_ctx root_log_ctx;
2735
	struct blk_plug plug;
2736

Y
Yan Zheng 已提交
2737
	mutex_lock(&root->log_mutex);
2738 2739 2740 2741 2742 2743 2744
	log_transid = ctx->log_transid;
	if (root->log_transid_committed >= log_transid) {
		mutex_unlock(&root->log_mutex);
		return ctx->log_ret;
	}

	index1 = log_transid % 2;
Y
Yan Zheng 已提交
2745
	if (atomic_read(&root->log_commit[index1])) {
2746
		wait_log_commit(root, log_transid);
Y
Yan Zheng 已提交
2747
		mutex_unlock(&root->log_mutex);
2748
		return ctx->log_ret;
2749
	}
2750
	ASSERT(log_transid == root->log_transid);
Y
Yan Zheng 已提交
2751 2752 2753 2754
	atomic_set(&root->log_commit[index1], 1);

	/* wait for previous tree log sync to complete */
	if (atomic_read(&root->log_commit[(index1 + 1) % 2]))
2755
		wait_log_commit(root, log_transid - 1);
2756

2757
	while (1) {
M
Miao Xie 已提交
2758
		int batch = atomic_read(&root->log_batch);
2759
		/* when we're on an ssd, just kick the log commit out */
2760 2761
		if (!btrfs_test_opt(root, SSD) &&
		    test_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state)) {
2762 2763 2764 2765
			mutex_unlock(&root->log_mutex);
			schedule_timeout_uninterruptible(1);
			mutex_lock(&root->log_mutex);
		}
2766
		wait_for_writer(root);
M
Miao Xie 已提交
2767
		if (batch == atomic_read(&root->log_batch))
2768 2769 2770
			break;
	}

2771
	/* bail out if we need to do a full commit */
2772
	if (btrfs_need_log_full_commit(root->fs_info, trans)) {
2773
		ret = -EAGAIN;
2774
		btrfs_free_logged_extents(log, log_transid);
2775 2776 2777 2778
		mutex_unlock(&root->log_mutex);
		goto out;
	}

2779 2780 2781 2782 2783
	if (log_transid % 2 == 0)
		mark = EXTENT_DIRTY;
	else
		mark = EXTENT_NEW;

2784 2785 2786
	/* we start IO on  all the marked extents here, but we don't actually
	 * wait for them until later.
	 */
2787
	blk_start_plug(&plug);
2788
	ret = btrfs_write_marked_extents(log, &log->dirty_log_pages, mark);
2789
	if (ret) {
2790
		blk_finish_plug(&plug);
2791
		btrfs_abort_transaction(trans, root, ret);
2792
		btrfs_free_logged_extents(log, log_transid);
2793
		btrfs_set_log_full_commit(root->fs_info, trans);
2794 2795 2796
		mutex_unlock(&root->log_mutex);
		goto out;
	}
Y
Yan Zheng 已提交
2797

2798
	btrfs_set_root_node(&log->root_item, log->node);
Y
Yan Zheng 已提交
2799 2800 2801

	root->log_transid++;
	log->log_transid = root->log_transid;
2802
	root->log_start_pid = 0;
Y
Yan Zheng 已提交
2803
	/*
2804 2805 2806
	 * 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
2807 2808 2809
	 */
	mutex_unlock(&root->log_mutex);

2810 2811
	btrfs_init_log_ctx(&root_log_ctx);

Y
Yan Zheng 已提交
2812
	mutex_lock(&log_root_tree->log_mutex);
M
Miao Xie 已提交
2813
	atomic_inc(&log_root_tree->log_batch);
Y
Yan Zheng 已提交
2814
	atomic_inc(&log_root_tree->log_writers);
2815 2816 2817 2818 2819

	index2 = log_root_tree->log_transid % 2;
	list_add_tail(&root_log_ctx.list, &log_root_tree->log_ctxs[index2]);
	root_log_ctx.log_transid = log_root_tree->log_transid;

Y
Yan Zheng 已提交
2820 2821 2822 2823 2824 2825
	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)) {
2826 2827 2828
		/*
		 * Implicit memory barrier after atomic_dec_and_test
		 */
Y
Yan Zheng 已提交
2829 2830 2831 2832
		if (waitqueue_active(&log_root_tree->log_writer_wait))
			wake_up(&log_root_tree->log_writer_wait);
	}

2833
	if (ret) {
2834 2835 2836
		if (!list_empty(&root_log_ctx.list))
			list_del_init(&root_log_ctx.list);

2837
		blk_finish_plug(&plug);
2838 2839
		btrfs_set_log_full_commit(root->fs_info, trans);

2840 2841 2842 2843 2844
		if (ret != -ENOSPC) {
			btrfs_abort_transaction(trans, root, ret);
			mutex_unlock(&log_root_tree->log_mutex);
			goto out;
		}
2845
		btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark);
2846
		btrfs_free_logged_extents(log, log_transid);
2847 2848 2849 2850 2851
		mutex_unlock(&log_root_tree->log_mutex);
		ret = -EAGAIN;
		goto out;
	}

2852
	if (log_root_tree->log_transid_committed >= root_log_ctx.log_transid) {
2853
		blk_finish_plug(&plug);
2854 2855 2856 2857
		mutex_unlock(&log_root_tree->log_mutex);
		ret = root_log_ctx.log_ret;
		goto out;
	}
2858

2859
	index2 = root_log_ctx.log_transid % 2;
Y
Yan Zheng 已提交
2860
	if (atomic_read(&log_root_tree->log_commit[index2])) {
2861
		blk_finish_plug(&plug);
2862 2863
		ret = btrfs_wait_marked_extents(log, &log->dirty_log_pages,
						mark);
2864
		btrfs_wait_logged_extents(trans, log, log_transid);
2865
		wait_log_commit(log_root_tree,
2866
				root_log_ctx.log_transid);
Y
Yan Zheng 已提交
2867
		mutex_unlock(&log_root_tree->log_mutex);
2868 2869
		if (!ret)
			ret = root_log_ctx.log_ret;
Y
Yan Zheng 已提交
2870 2871
		goto out;
	}
2872
	ASSERT(root_log_ctx.log_transid == log_root_tree->log_transid);
Y
Yan Zheng 已提交
2873 2874
	atomic_set(&log_root_tree->log_commit[index2], 1);

2875
	if (atomic_read(&log_root_tree->log_commit[(index2 + 1) % 2])) {
2876
		wait_log_commit(log_root_tree,
2877
				root_log_ctx.log_transid - 1);
2878 2879
	}

2880
	wait_for_writer(log_root_tree);
Y
Yan Zheng 已提交
2881

2882 2883 2884 2885
	/*
	 * now that we've moved on to the tree of log tree roots,
	 * check the full commit flag again
	 */
2886
	if (btrfs_need_log_full_commit(root->fs_info, trans)) {
2887
		blk_finish_plug(&plug);
2888
		btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark);
2889
		btrfs_free_logged_extents(log, log_transid);
2890 2891 2892 2893
		mutex_unlock(&log_root_tree->log_mutex);
		ret = -EAGAIN;
		goto out_wake_log_root;
	}
Y
Yan Zheng 已提交
2894

2895 2896 2897 2898
	ret = btrfs_write_marked_extents(log_root_tree,
					 &log_root_tree->dirty_log_pages,
					 EXTENT_DIRTY | EXTENT_NEW);
	blk_finish_plug(&plug);
2899
	if (ret) {
2900
		btrfs_set_log_full_commit(root->fs_info, trans);
2901
		btrfs_abort_transaction(trans, root, ret);
2902
		btrfs_free_logged_extents(log, log_transid);
2903 2904 2905
		mutex_unlock(&log_root_tree->log_mutex);
		goto out_wake_log_root;
	}
2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916
	ret = btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark);
	if (!ret)
		ret = btrfs_wait_marked_extents(log_root_tree,
						&log_root_tree->dirty_log_pages,
						EXTENT_NEW | EXTENT_DIRTY);
	if (ret) {
		btrfs_set_log_full_commit(root->fs_info, trans);
		btrfs_free_logged_extents(log, log_transid);
		mutex_unlock(&log_root_tree->log_mutex);
		goto out_wake_log_root;
	}
2917
	btrfs_wait_logged_extents(trans, log, log_transid);
2918

2919
	btrfs_set_super_log_root(root->fs_info->super_for_commit,
Y
Yan Zheng 已提交
2920
				log_root_tree->node->start);
2921
	btrfs_set_super_log_root_level(root->fs_info->super_for_commit,
Y
Yan Zheng 已提交
2922
				btrfs_header_level(log_root_tree->node));
2923

Y
Yan Zheng 已提交
2924 2925 2926 2927 2928 2929 2930 2931 2932 2933
	log_root_tree->log_transid++;
	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.
	 */
2934 2935
	ret = write_ctree_super(trans, root->fs_info->tree_root, 1);
	if (ret) {
2936
		btrfs_set_log_full_commit(root->fs_info, trans);
2937 2938 2939
		btrfs_abort_transaction(trans, root, ret);
		goto out_wake_log_root;
	}
Y
Yan Zheng 已提交
2940

2941 2942 2943 2944 2945
	mutex_lock(&root->log_mutex);
	if (root->last_log_commit < log_transid)
		root->last_log_commit = log_transid;
	mutex_unlock(&root->log_mutex);

2946
out_wake_log_root:
2947 2948 2949 2950 2951 2952
	/*
	 * We needn't get log_mutex here because we are sure all
	 * the other tasks are blocked.
	 */
	btrfs_remove_all_log_ctxs(log_root_tree, index2, ret);

2953 2954
	mutex_lock(&log_root_tree->log_mutex);
	log_root_tree->log_transid_committed++;
Y
Yan Zheng 已提交
2955
	atomic_set(&log_root_tree->log_commit[index2], 0);
2956 2957
	mutex_unlock(&log_root_tree->log_mutex);

2958 2959 2960
	/*
	 * The barrier before waitqueue_active is implied by mutex_unlock
	 */
Y
Yan Zheng 已提交
2961 2962
	if (waitqueue_active(&log_root_tree->log_commit_wait[index2]))
		wake_up(&log_root_tree->log_commit_wait[index2]);
2963
out:
2964 2965 2966
	/* See above. */
	btrfs_remove_all_log_ctxs(root, index1, ret);

2967 2968
	mutex_lock(&root->log_mutex);
	root->log_transid_committed++;
Y
Yan Zheng 已提交
2969
	atomic_set(&root->log_commit[index1], 0);
2970
	mutex_unlock(&root->log_mutex);
2971

2972 2973 2974
	/*
	 * The barrier before waitqueue_active is implied by mutex_unlock
	 */
Y
Yan Zheng 已提交
2975 2976
	if (waitqueue_active(&root->log_commit_wait[index1]))
		wake_up(&root->log_commit_wait[index1]);
2977
	return ret;
2978 2979
}

2980 2981
static void free_log_tree(struct btrfs_trans_handle *trans,
			  struct btrfs_root *log)
2982 2983
{
	int ret;
2984 2985
	u64 start;
	u64 end;
2986 2987 2988 2989 2990
	struct walk_control wc = {
		.free = 1,
		.process_func = process_one_buffer
	};

2991 2992 2993 2994
	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);
2995

C
Chris Mason 已提交
2996
	while (1) {
2997
		ret = find_first_extent_bit(&log->dirty_log_pages,
2998 2999
				0, &start, &end, EXTENT_DIRTY | EXTENT_NEW,
				NULL);
3000 3001 3002
		if (ret)
			break;

3003
		clear_extent_bits(&log->dirty_log_pages, start, end,
3004
				  EXTENT_DIRTY | EXTENT_NEW);
3005 3006
	}

3007 3008 3009 3010 3011 3012 3013 3014
	/*
	 * 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
3015 3016
	free_extent_buffer(log->node);
	kfree(log);
3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038
}

/*
 * 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;
	}
3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071
	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;
3072
	int err = 0;
3073
	int bytes_del = 0;
L
Li Zefan 已提交
3074
	u64 dir_ino = btrfs_ino(dir);
3075

3076 3077 3078
	if (BTRFS_I(dir)->logged_trans < trans->transid)
		return 0;

3079 3080 3081 3082 3083 3084 3085 3086
	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();
3087 3088 3089 3090
	if (!path) {
		err = -ENOMEM;
		goto out_unlock;
	}
3091

L
Li Zefan 已提交
3092
	di = btrfs_lookup_dir_item(trans, log, path, dir_ino,
3093
				   name, name_len, -1);
3094 3095 3096 3097 3098
	if (IS_ERR(di)) {
		err = PTR_ERR(di);
		goto fail;
	}
	if (di) {
3099 3100
		ret = btrfs_delete_one_dir_name(trans, log, path, di);
		bytes_del += name_len;
3101 3102 3103 3104
		if (ret) {
			err = ret;
			goto fail;
		}
3105
	}
3106
	btrfs_release_path(path);
L
Li Zefan 已提交
3107
	di = btrfs_lookup_dir_index_item(trans, log, path, dir_ino,
3108
					 index, name, name_len, -1);
3109 3110 3111 3112 3113
	if (IS_ERR(di)) {
		err = PTR_ERR(di);
		goto fail;
	}
	if (di) {
3114 3115
		ret = btrfs_delete_one_dir_name(trans, log, path, di);
		bytes_del += name_len;
3116 3117 3118 3119
		if (ret) {
			err = ret;
			goto fail;
		}
3120 3121 3122 3123 3124 3125 3126 3127
	}

	/* 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
3128
		key.objectid = dir_ino;
3129 3130
		key.offset = 0;
		key.type = BTRFS_INODE_ITEM_KEY;
3131
		btrfs_release_path(path);
3132 3133

		ret = btrfs_search_slot(trans, log, &key, path, 0, 1);
3134 3135 3136 3137
		if (ret < 0) {
			err = ret;
			goto fail;
		}
3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152
		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;
3153
		btrfs_release_path(path);
3154
	}
3155
fail:
3156
	btrfs_free_path(path);
3157
out_unlock:
3158
	mutex_unlock(&BTRFS_I(dir)->log_mutex);
3159
	if (ret == -ENOSPC) {
3160
		btrfs_set_log_full_commit(root->fs_info, trans);
3161
		ret = 0;
3162 3163 3164
	} else if (ret < 0)
		btrfs_abort_transaction(trans, root, ret);

3165
	btrfs_end_log_trans(root);
3166

3167
	return err;
3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179
}

/* 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;

3180 3181 3182
	if (BTRFS_I(inode)->logged_trans < trans->transid)
		return 0;

3183 3184 3185 3186 3187 3188
	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
3189
	ret = btrfs_del_inode_ref(trans, log, name, name_len, btrfs_ino(inode),
3190 3191
				  dirid, &index);
	mutex_unlock(&BTRFS_I(inode)->log_mutex);
3192
	if (ret == -ENOSPC) {
3193
		btrfs_set_log_full_commit(root->fs_info, trans);
3194
		ret = 0;
3195 3196
	} else if (ret < 0 && ret != -ENOENT)
		btrfs_abort_transaction(trans, root, ret);
3197
	btrfs_end_log_trans(root);
3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223

	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));
3224 3225
	if (ret)
		return ret;
3226 3227 3228 3229 3230

	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]);
3231
	btrfs_release_path(path);
3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243
	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,
3244
			  struct btrfs_log_ctx *ctx,
3245 3246 3247 3248 3249
			  u64 min_offset, u64 *last_offset_ret)
{
	struct btrfs_key min_key;
	struct btrfs_root *log = root->log_root;
	struct extent_buffer *src;
3250
	int err = 0;
3251 3252 3253 3254 3255
	int ret;
	int i;
	int nritems;
	u64 first_offset = min_offset;
	u64 last_offset = (u64)-1;
L
Li Zefan 已提交
3256
	u64 ino = btrfs_ino(inode);
3257 3258 3259

	log = root->log_root;

L
Li Zefan 已提交
3260
	min_key.objectid = ino;
3261 3262 3263
	min_key.type = key_type;
	min_key.offset = min_offset;

3264
	ret = btrfs_search_forward(root, &min_key, path, trans->transid);
3265 3266 3267 3268 3269

	/*
	 * we didn't find anything from this transaction, see if there
	 * is anything at all
	 */
L
Li Zefan 已提交
3270 3271
	if (ret != 0 || min_key.objectid != ino || min_key.type != key_type) {
		min_key.objectid = ino;
3272 3273
		min_key.type = key_type;
		min_key.offset = (u64)-1;
3274
		btrfs_release_path(path);
3275 3276
		ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0);
		if (ret < 0) {
3277
			btrfs_release_path(path);
3278 3279
			return ret;
		}
L
Li Zefan 已提交
3280
		ret = btrfs_previous_item(root, path, ino, key_type);
3281 3282 3283 3284 3285 3286 3287 3288 3289 3290

		/* 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
3291
			if (key_type == tmp.type)
3292 3293 3294 3295 3296 3297
				first_offset = max(min_offset, tmp.offset) + 1;
		}
		goto done;
	}

	/* go backward to find any previous key */
L
Li Zefan 已提交
3298
	ret = btrfs_previous_item(root, path, ino, key_type);
3299 3300 3301 3302 3303 3304 3305 3306
	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);
3307 3308 3309 3310
			if (ret) {
				err = ret;
				goto done;
			}
3311 3312
		}
	}
3313
	btrfs_release_path(path);
3314 3315 3316

	/* find the first key from this transaction again */
	ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0);
3317
	if (WARN_ON(ret != 0))
3318 3319 3320 3321 3322 3323
		goto done;

	/*
	 * we have a block from this transaction, log every item in it
	 * from our directory
	 */
C
Chris Mason 已提交
3324
	while (1) {
3325 3326 3327 3328
		struct btrfs_key tmp;
		src = path->nodes[0];
		nritems = btrfs_header_nritems(src);
		for (i = path->slots[0]; i < nritems; i++) {
3329 3330
			struct btrfs_dir_item *di;

3331 3332
			btrfs_item_key_to_cpu(src, &min_key, i);

L
Li Zefan 已提交
3333
			if (min_key.objectid != ino || min_key.type != key_type)
3334 3335 3336
				goto done;
			ret = overwrite_item(trans, log, dst_path, src, i,
					     &min_key);
3337 3338 3339 3340
			if (ret) {
				err = ret;
				goto done;
			}
3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371

			/*
			 * We must make sure that when we log a directory entry,
			 * the corresponding inode, after log replay, has a
			 * matching link count. For example:
			 *
			 * touch foo
			 * mkdir mydir
			 * sync
			 * ln foo mydir/bar
			 * xfs_io -c "fsync" mydir
			 * <crash>
			 * <mount fs and log replay>
			 *
			 * Would result in a fsync log that when replayed, our
			 * file inode would have a link count of 1, but we get
			 * two directory entries pointing to the same inode.
			 * After removing one of the names, it would not be
			 * possible to remove the other name, which resulted
			 * always in stale file handle errors, and would not
			 * be possible to rmdir the parent directory, since
			 * its i_size could never decrement to the value
			 * BTRFS_EMPTY_DIR_SIZE, resulting in -ENOTEMPTY errors.
			 */
			di = btrfs_item_ptr(src, i, struct btrfs_dir_item);
			btrfs_dir_item_key_to_cpu(src, di, &tmp);
			if (ctx &&
			    (btrfs_dir_transid(src, di) == trans->transid ||
			     btrfs_dir_type(src, di) == BTRFS_FT_DIR) &&
			    tmp.type != BTRFS_ROOT_ITEM_KEY)
				ctx->log_new_dentries = true;
3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384
		}
		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
3385
		if (tmp.objectid != ino || tmp.type != key_type) {
3386 3387 3388 3389 3390 3391 3392
			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);
3393 3394 3395 3396
			if (ret)
				err = ret;
			else
				last_offset = tmp.offset;
3397 3398 3399 3400
			goto done;
		}
	}
done:
3401 3402
	btrfs_release_path(path);
	btrfs_release_path(dst_path);
3403

3404 3405 3406 3407 3408 3409 3410
	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
3411
					 ino, first_offset, last_offset);
3412 3413 3414 3415
		if (ret)
			err = ret;
	}
	return err;
3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432
}

/*
 * 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,
3433 3434
			  struct btrfs_path *dst_path,
			  struct btrfs_log_ctx *ctx)
3435 3436 3437 3438 3439 3440 3441 3442 3443
{
	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
3444
	while (1) {
3445
		ret = log_dir_items(trans, root, inode, path,
3446
				    dst_path, key_type, ctx, min_key,
3447
				    &max_key);
3448 3449
		if (ret)
			return ret;
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
		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;
3476
	int start_slot;
3477 3478 3479 3480 3481

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

C
Chris Mason 已提交
3482
	while (1) {
3483
		ret = btrfs_search_slot(trans, log, &key, path, -1, 1);
3484
		BUG_ON(ret == 0); /* Logic error */
3485
		if (ret < 0)
3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497
			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;

3498 3499 3500 3501 3502 3503 3504 3505 3506 3507 3508 3509
		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)
3510
			break;
3511
		btrfs_release_path(path);
3512
	}
3513
	btrfs_release_path(path);
3514 3515
	if (ret > 0)
		ret = 0;
3516
	return ret;
3517 3518
}

3519 3520 3521
static void fill_inode_item(struct btrfs_trans_handle *trans,
			    struct extent_buffer *leaf,
			    struct btrfs_inode_item *item,
3522 3523
			    struct inode *inode, int log_inode_only,
			    u64 logged_isize)
3524
{
3525 3526 3527
	struct btrfs_map_token token;

	btrfs_init_map_token(&token);
3528 3529 3530 3531 3532 3533 3534

	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'
		 */
3535
		btrfs_set_token_inode_generation(leaf, item, 0, &token);
3536
		btrfs_set_token_inode_size(leaf, item, logged_isize, &token);
3537
	} else {
3538 3539 3540 3541 3542 3543 3544 3545 3546 3547 3548
		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);

3549
	btrfs_set_token_timespec_sec(leaf, &item->atime,
3550
				     inode->i_atime.tv_sec, &token);
3551
	btrfs_set_token_timespec_nsec(leaf, &item->atime,
3552 3553
				      inode->i_atime.tv_nsec, &token);

3554
	btrfs_set_token_timespec_sec(leaf, &item->mtime,
3555
				     inode->i_mtime.tv_sec, &token);
3556
	btrfs_set_token_timespec_nsec(leaf, &item->mtime,
3557 3558
				      inode->i_mtime.tv_nsec, &token);

3559
	btrfs_set_token_timespec_sec(leaf, &item->ctime,
3560
				     inode->i_ctime.tv_sec, &token);
3561
	btrfs_set_token_timespec_nsec(leaf, &item->ctime,
3562 3563 3564 3565 3566 3567 3568 3569 3570 3571
				      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);
3572 3573
}

3574 3575 3576 3577 3578 3579 3580
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;

3581 3582
	ret = btrfs_insert_empty_item(trans, log, path,
				      &BTRFS_I(inode)->location,
3583 3584 3585 3586 3587
				      sizeof(*inode_item));
	if (ret && ret != -EEXIST)
		return ret;
	inode_item = btrfs_item_ptr(path->nodes[0], path->slots[0],
				    struct btrfs_inode_item);
3588
	fill_inode_item(trans, path->nodes[0], inode_item, inode, 0, 0);
3589 3590 3591 3592
	btrfs_release_path(path);
	return 0;
}

3593
static noinline int copy_items(struct btrfs_trans_handle *trans,
3594
			       struct inode *inode,
3595
			       struct btrfs_path *dst_path,
3596
			       struct btrfs_path *src_path, u64 *last_extent,
3597 3598
			       int start_slot, int nr, int inode_only,
			       u64 logged_isize)
3599 3600 3601
{
	unsigned long src_offset;
	unsigned long dst_offset;
3602
	struct btrfs_root *log = BTRFS_I(inode)->root->log_root;
3603 3604
	struct btrfs_file_extent_item *extent;
	struct btrfs_inode_item *inode_item;
3605 3606
	struct extent_buffer *src = src_path->nodes[0];
	struct btrfs_key first_key, last_key, key;
3607 3608 3609 3610 3611
	int ret;
	struct btrfs_key *ins_keys;
	u32 *ins_sizes;
	char *ins_data;
	int i;
3612
	struct list_head ordered_sums;
3613
	int skip_csum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
3614
	bool has_extents = false;
3615
	bool need_find_last_extent = true;
3616
	bool done = false;
3617 3618

	INIT_LIST_HEAD(&ordered_sums);
3619 3620 3621

	ins_data = kmalloc(nr * sizeof(struct btrfs_key) +
			   nr * sizeof(u32), GFP_NOFS);
3622 3623 3624
	if (!ins_data)
		return -ENOMEM;

3625 3626
	first_key.objectid = (u64)-1;

3627 3628 3629 3630 3631 3632 3633 3634 3635
	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);
3636 3637 3638 3639
	if (ret) {
		kfree(ins_data);
		return ret;
	}
3640

3641
	for (i = 0; i < nr; i++, dst_path->slots[0]++) {
3642 3643 3644 3645 3646
		dst_offset = btrfs_item_ptr_offset(dst_path->nodes[0],
						   dst_path->slots[0]);

		src_offset = btrfs_item_ptr_offset(src, start_slot + i);

3647 3648 3649
		if ((i == (nr - 1)))
			last_key = ins_keys[i];

3650
		if (ins_keys[i].type == BTRFS_INODE_ITEM_KEY) {
3651 3652 3653
			inode_item = btrfs_item_ptr(dst_path->nodes[0],
						    dst_path->slots[0],
						    struct btrfs_inode_item);
3654
			fill_inode_item(trans, dst_path->nodes[0], inode_item,
3655 3656
					inode, inode_only == LOG_INODE_EXISTS,
					logged_isize);
3657 3658 3659
		} else {
			copy_extent_buffer(dst_path->nodes[0], src, dst_offset,
					   src_offset, ins_sizes[i]);
3660
		}
3661

3662 3663 3664 3665 3666 3667 3668 3669
		/*
		 * We set need_find_last_extent here in case we know we were
		 * processing other items and then walk into the first extent in
		 * the inode.  If we don't hit an extent then nothing changes,
		 * we'll do the last search the next time around.
		 */
		if (ins_keys[i].type == BTRFS_EXTENT_DATA_KEY) {
			has_extents = true;
3670
			if (first_key.objectid == (u64)-1)
3671 3672 3673 3674 3675
				first_key = ins_keys[i];
		} else {
			need_find_last_extent = false;
		}

3676 3677 3678 3679
		/* take a reference on file data extents so that truncates
		 * or deletes of this inode don't have to relog the inode
		 * again
		 */
3680
		if (ins_keys[i].type == BTRFS_EXTENT_DATA_KEY &&
3681
		    !skip_csum) {
3682 3683 3684 3685
			int found_type;
			extent = btrfs_item_ptr(src, start_slot + i,
						struct btrfs_file_extent_item);

3686 3687 3688
			if (btrfs_file_extent_generation(src, extent) < trans->transid)
				continue;

3689
			found_type = btrfs_file_extent_type(src, extent);
3690
			if (found_type == BTRFS_FILE_EXTENT_REG) {
3691 3692 3693 3694 3695 3696 3697 3698 3699 3700 3701
				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,
3702
								extent);
3703 3704 3705 3706 3707
				if (btrfs_file_extent_compression(src,
								  extent)) {
					cs = 0;
					cl = dl;
				}
3708 3709 3710 3711

				ret = btrfs_lookup_csums_range(
						log->fs_info->csum_root,
						ds + cs, ds + cs + cl - 1,
A
Arne Jansen 已提交
3712
						&ordered_sums, 0);
3713 3714 3715 3716 3717
				if (ret) {
					btrfs_release_path(dst_path);
					kfree(ins_data);
					return ret;
				}
3718 3719 3720 3721 3722
			}
		}
	}

	btrfs_mark_buffer_dirty(dst_path->nodes[0]);
3723
	btrfs_release_path(dst_path);
3724
	kfree(ins_data);
3725 3726 3727 3728 3729

	/*
	 * we have to do this after the loop above to avoid changing the
	 * log tree while trying to change the log tree.
	 */
3730
	ret = 0;
C
Chris Mason 已提交
3731
	while (!list_empty(&ordered_sums)) {
3732 3733 3734
		struct btrfs_ordered_sum *sums = list_entry(ordered_sums.next,
						   struct btrfs_ordered_sum,
						   list);
3735 3736
		if (!ret)
			ret = btrfs_csum_file_blocks(trans, log, sums);
3737 3738 3739
		list_del(&sums->list);
		kfree(sums);
	}
3740 3741 3742 3743

	if (!has_extents)
		return ret;

3744 3745 3746 3747 3748 3749 3750 3751 3752 3753
	if (need_find_last_extent && *last_extent == first_key.offset) {
		/*
		 * We don't have any leafs between our current one and the one
		 * we processed before that can have file extent items for our
		 * inode (and have a generation number smaller than our current
		 * transaction id).
		 */
		need_find_last_extent = false;
	}

3754 3755 3756 3757 3758 3759 3760 3761 3762 3763 3764 3765 3766 3767 3768 3769 3770 3771 3772 3773 3774 3775 3776 3777 3778
	/*
	 * Because we use btrfs_search_forward we could skip leaves that were
	 * not modified and then assume *last_extent is valid when it really
	 * isn't.  So back up to the previous leaf and read the end of the last
	 * extent before we go and fill in holes.
	 */
	if (need_find_last_extent) {
		u64 len;

		ret = btrfs_prev_leaf(BTRFS_I(inode)->root, src_path);
		if (ret < 0)
			return ret;
		if (ret)
			goto fill_holes;
		if (src_path->slots[0])
			src_path->slots[0]--;
		src = src_path->nodes[0];
		btrfs_item_key_to_cpu(src, &key, src_path->slots[0]);
		if (key.objectid != btrfs_ino(inode) ||
		    key.type != BTRFS_EXTENT_DATA_KEY)
			goto fill_holes;
		extent = btrfs_item_ptr(src, src_path->slots[0],
					struct btrfs_file_extent_item);
		if (btrfs_file_extent_type(src, extent) ==
		    BTRFS_FILE_EXTENT_INLINE) {
3779 3780 3781
			len = btrfs_file_extent_inline_len(src,
							   src_path->slots[0],
							   extent);
3782 3783 3784 3785 3786 3787 3788 3789 3790 3791 3792 3793 3794 3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 3806 3807 3808 3809 3810 3811 3812 3813 3814 3815 3816 3817 3818 3819 3820 3821 3822 3823 3824 3825 3826 3827 3828 3829 3830 3831 3832 3833 3834 3835 3836 3837 3838 3839 3840 3841 3842 3843 3844
			*last_extent = ALIGN(key.offset + len,
					     log->sectorsize);
		} else {
			len = btrfs_file_extent_num_bytes(src, extent);
			*last_extent = key.offset + len;
		}
	}
fill_holes:
	/* So we did prev_leaf, now we need to move to the next leaf, but a few
	 * things could have happened
	 *
	 * 1) A merge could have happened, so we could currently be on a leaf
	 * that holds what we were copying in the first place.
	 * 2) A split could have happened, and now not all of the items we want
	 * are on the same leaf.
	 *
	 * So we need to adjust how we search for holes, we need to drop the
	 * path and re-search for the first extent key we found, and then walk
	 * forward until we hit the last one we copied.
	 */
	if (need_find_last_extent) {
		/* btrfs_prev_leaf could return 1 without releasing the path */
		btrfs_release_path(src_path);
		ret = btrfs_search_slot(NULL, BTRFS_I(inode)->root, &first_key,
					src_path, 0, 0);
		if (ret < 0)
			return ret;
		ASSERT(ret == 0);
		src = src_path->nodes[0];
		i = src_path->slots[0];
	} else {
		i = start_slot;
	}

	/*
	 * Ok so here we need to go through and fill in any holes we may have
	 * to make sure that holes are punched for those areas in case they had
	 * extents previously.
	 */
	while (!done) {
		u64 offset, len;
		u64 extent_end;

		if (i >= btrfs_header_nritems(src_path->nodes[0])) {
			ret = btrfs_next_leaf(BTRFS_I(inode)->root, src_path);
			if (ret < 0)
				return ret;
			ASSERT(ret == 0);
			src = src_path->nodes[0];
			i = 0;
		}

		btrfs_item_key_to_cpu(src, &key, i);
		if (!btrfs_comp_cpu_keys(&key, &last_key))
			done = true;
		if (key.objectid != btrfs_ino(inode) ||
		    key.type != BTRFS_EXTENT_DATA_KEY) {
			i++;
			continue;
		}
		extent = btrfs_item_ptr(src, i, struct btrfs_file_extent_item);
		if (btrfs_file_extent_type(src, extent) ==
		    BTRFS_FILE_EXTENT_INLINE) {
3845
			len = btrfs_file_extent_inline_len(src, i, extent);
3846 3847 3848 3849 3850 3851 3852 3853 3854 3855 3856 3857 3858 3859 3860 3861 3862 3863
			extent_end = ALIGN(key.offset + len, log->sectorsize);
		} else {
			len = btrfs_file_extent_num_bytes(src, extent);
			extent_end = key.offset + len;
		}
		i++;

		if (*last_extent == key.offset) {
			*last_extent = extent_end;
			continue;
		}
		offset = *last_extent;
		len = key.offset - *last_extent;
		ret = btrfs_insert_file_extent(trans, log, btrfs_ino(inode),
					       offset, 0, 0, len, 0, len, 0,
					       0, 0);
		if (ret)
			break;
3864
		*last_extent = extent_end;
3865 3866 3867 3868 3869 3870 3871
	}
	/*
	 * Need to let the callers know we dropped the path so they should
	 * re-search.
	 */
	if (!ret && need_find_last_extent)
		ret = 1;
3872
	return ret;
3873 3874
}

J
Josef Bacik 已提交
3875 3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888
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;
}

3889 3890 3891 3892 3893 3894
static int wait_ordered_extents(struct btrfs_trans_handle *trans,
				struct inode *inode,
				struct btrfs_root *root,
				const struct extent_map *em,
				const struct list_head *logged_list,
				bool *ordered_io_error)
J
Josef Bacik 已提交
3895
{
3896
	struct btrfs_ordered_extent *ordered;
3897
	struct btrfs_root *log = root->log_root;
3898 3899
	u64 mod_start = em->mod_start;
	u64 mod_len = em->mod_len;
3900
	const bool skip_csum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
3901 3902
	u64 csum_offset;
	u64 csum_len;
3903 3904
	LIST_HEAD(ordered_sums);
	int ret = 0;
3905

3906
	*ordered_io_error = false;
3907

3908 3909
	if (test_bit(EXTENT_FLAG_PREALLOC, &em->flags) ||
	    em->block_start == EXTENT_MAP_HOLE)
3910
		return 0;
J
Josef Bacik 已提交
3911

3912
	/*
3913 3914 3915
	 * Wait far any ordered extent that covers our extent map. If it
	 * finishes without an error, first check and see if our csums are on
	 * our outstanding ordered extents.
3916
	 */
3917
	list_for_each_entry(ordered, logged_list, log_list) {
3918 3919 3920 3921 3922 3923 3924 3925 3926
		struct btrfs_ordered_sum *sum;

		if (!mod_len)
			break;

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

3927 3928 3929 3930 3931 3932 3933 3934 3935 3936 3937 3938 3939 3940 3941
		if (!test_bit(BTRFS_ORDERED_IO_DONE, &ordered->flags) &&
		    !test_bit(BTRFS_ORDERED_IOERR, &ordered->flags) &&
		    !test_bit(BTRFS_ORDERED_DIRECT, &ordered->flags)) {
			const u64 start = ordered->file_offset;
			const u64 end = ordered->file_offset + ordered->len - 1;

			WARN_ON(ordered->inode != inode);
			filemap_fdatawrite_range(inode->i_mapping, start, end);
		}

		wait_event(ordered->wait,
			   (test_bit(BTRFS_ORDERED_IO_DONE, &ordered->flags) ||
			    test_bit(BTRFS_ORDERED_IOERR, &ordered->flags)));

		if (test_bit(BTRFS_ORDERED_IOERR, &ordered->flags)) {
3942 3943 3944 3945 3946 3947
			/*
			 * Clear the AS_EIO/AS_ENOSPC flags from the inode's
			 * i_mapping flags, so that the next fsync won't get
			 * an outdated io error too.
			 */
			btrfs_inode_check_errors(inode);
3948 3949 3950
			*ordered_io_error = true;
			break;
		}
3951 3952 3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975 3976 3977 3978 3979 3980 3981
		/*
		 * 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;
			}
		}

3982 3983 3984
		if (skip_csum)
			continue;

3985 3986 3987 3988 3989 3990 3991 3992 3993 3994
		/*
		 * 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;

		list_for_each_entry(sum, &ordered->list, list) {
			ret = btrfs_csum_file_blocks(trans, log, sum);
3995
			if (ret)
3996
				break;
3997 3998 3999
		}
	}

4000
	if (*ordered_io_error || !mod_len || ret || skip_csum)
4001 4002
		return ret;

4003 4004
	if (em->compress_type) {
		csum_offset = 0;
4005
		csum_len = max(em->block_len, em->orig_block_len);
4006 4007 4008 4009
	} else {
		csum_offset = mod_start - em->start;
		csum_len = mod_len;
	}
4010

4011 4012 4013 4014 4015 4016 4017
	/* 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
4018

4019 4020 4021 4022 4023 4024 4025 4026
	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
4027 4028
	}

4029
	return ret;
J
Josef Bacik 已提交
4030 4031
}

4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044 4045 4046 4047 4048 4049 4050 4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 4062 4063 4064 4065 4066 4067 4068 4069 4070 4071 4072 4073 4074 4075 4076 4077 4078 4079 4080 4081
static int log_one_extent(struct btrfs_trans_handle *trans,
			  struct inode *inode, struct btrfs_root *root,
			  const struct extent_map *em,
			  struct btrfs_path *path,
			  const struct list_head *logged_list,
			  struct btrfs_log_ctx *ctx)
{
	struct btrfs_root *log = root->log_root;
	struct btrfs_file_extent_item *fi;
	struct extent_buffer *leaf;
	struct btrfs_map_token token;
	struct btrfs_key key;
	u64 extent_offset = em->start - em->orig_start;
	u64 block_len;
	int ret;
	int extent_inserted = 0;
	bool ordered_io_err = false;

	ret = wait_ordered_extents(trans, inode, root, em, logged_list,
				   &ordered_io_err);
	if (ret)
		return ret;

	if (ordered_io_err) {
		ctx->io_err = -EIO;
		return 0;
	}

	btrfs_init_map_token(&token);

	ret = __btrfs_drop_extents(trans, log, inode, path, em->start,
				   em->start + em->len, NULL, 0, 1,
				   sizeof(*fi), &extent_inserted);
	if (ret)
		return ret;

	if (!extent_inserted) {
		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));
		if (ret)
			return ret;
	}
	leaf = path->nodes[0];
	fi = btrfs_item_ptr(leaf, path->slots[0],
			    struct btrfs_file_extent_item);

4082
	btrfs_set_token_file_extent_generation(leaf, fi, trans->transid,
4083 4084 4085 4086 4087 4088 4089 4090 4091 4092 4093 4094 4095 4096 4097 4098 4099 4100 4101 4102 4103 4104 4105 4106 4107 4108 4109 4110 4111 4112 4113 4114 4115 4116 4117 4118 4119 4120 4121 4122 4123 4124 4125
					       &token);
	if (test_bit(EXTENT_FLAG_PREALLOC, &em->flags))
		btrfs_set_token_file_extent_type(leaf, fi,
						 BTRFS_FILE_EXTENT_PREALLOC,
						 &token);
	else
		btrfs_set_token_file_extent_type(leaf, fi,
						 BTRFS_FILE_EXTENT_REG,
						 &token);

	block_len = max(em->block_len, em->orig_block_len);
	if (em->compress_type != BTRFS_COMPRESS_NONE) {
		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);
	} else if (em->block_start < EXTENT_MAP_LAST_BYTE) {
		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);
	} else {
		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, extent_offset, &token);
	btrfs_set_token_file_extent_num_bytes(leaf, fi, em->len, &token);
	btrfs_set_token_file_extent_ram_bytes(leaf, fi, em->ram_bytes, &token);
	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);
	btrfs_mark_buffer_dirty(leaf);

	btrfs_release_path(path);

	return ret;
}

J
Josef Bacik 已提交
4126 4127 4128
static int btrfs_log_changed_extents(struct btrfs_trans_handle *trans,
				     struct btrfs_root *root,
				     struct inode *inode,
4129
				     struct btrfs_path *path,
4130
				     struct list_head *logged_list,
4131 4132 4133
				     struct btrfs_log_ctx *ctx,
				     const u64 start,
				     const u64 end)
J
Josef Bacik 已提交
4134 4135 4136 4137 4138 4139
{
	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;
4140
	int num = 0;
J
Josef Bacik 已提交
4141 4142 4143

	INIT_LIST_HEAD(&extents);

4144
	down_write(&BTRFS_I(inode)->dio_sem);
J
Josef Bacik 已提交
4145 4146 4147 4148 4149
	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);
4150 4151 4152 4153 4154 4155 4156 4157 4158 4159 4160 4161 4162

		/*
		 * 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
4163 4164
		if (em->generation <= test_gen)
			continue;
4165 4166 4167
		/* 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
4168
		list_add_tail(&em->list, &extents);
4169
		num++;
J
Josef Bacik 已提交
4170 4171 4172
	}

	list_sort(NULL, &extents, extent_cmp);
4173
	btrfs_get_logged_extents(inode, logged_list, start, end);
4174
	/*
4175 4176 4177 4178 4179 4180 4181 4182
	 * Some ordered extents started by fsync might have completed
	 * before we could collect them into the list logged_list, which
	 * means they're gone, not in our logged_list nor in the inode's
	 * ordered tree. We want the application/user space to know an
	 * error happened while attempting to persist file data so that
	 * it can take proper action. If such error happened, we leave
	 * without writing to the log tree and the fsync must report the
	 * file data write error and not commit the current transaction.
4183
	 */
4184 4185 4186
	ret = btrfs_inode_check_errors(inode);
	if (ret)
		ctx->io_err = ret;
4187
process:
J
Josef Bacik 已提交
4188 4189 4190 4191 4192 4193 4194 4195 4196
	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.
		 */
4197
		if (ret) {
4198
			clear_em_logging(tree, em);
4199
			free_extent_map(em);
J
Josef Bacik 已提交
4200
			continue;
4201 4202 4203
		}

		write_unlock(&tree->lock);
J
Josef Bacik 已提交
4204

4205 4206
		ret = log_one_extent(trans, inode, root, em, path, logged_list,
				     ctx);
4207
		write_lock(&tree->lock);
4208 4209
		clear_em_logging(tree, em);
		free_extent_map(em);
J
Josef Bacik 已提交
4210
	}
4211 4212
	WARN_ON(!list_empty(&extents));
	write_unlock(&tree->lock);
4213
	up_write(&BTRFS_I(inode)->dio_sem);
J
Josef Bacik 已提交
4214 4215 4216 4217 4218

	btrfs_release_path(path);
	return ret;
}

4219 4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232
static int logged_inode_size(struct btrfs_root *log, struct inode *inode,
			     struct btrfs_path *path, u64 *size_ret)
{
	struct btrfs_key key;
	int ret;

	key.objectid = btrfs_ino(inode);
	key.type = BTRFS_INODE_ITEM_KEY;
	key.offset = 0;

	ret = btrfs_search_slot(NULL, log, &key, path, 0, 0);
	if (ret < 0) {
		return ret;
	} else if (ret > 0) {
4233
		*size_ret = 0;
4234 4235 4236 4237 4238 4239 4240 4241 4242 4243 4244 4245
	} else {
		struct btrfs_inode_item *item;

		item = btrfs_item_ptr(path->nodes[0], path->slots[0],
				      struct btrfs_inode_item);
		*size_ret = btrfs_inode_size(path->nodes[0], item);
	}

	btrfs_release_path(path);
	return 0;
}

4246 4247 4248 4249 4250 4251 4252 4253 4254 4255 4256 4257 4258 4259 4260 4261 4262 4263 4264 4265 4266 4267 4268 4269 4270 4271 4272 4273 4274 4275 4276 4277 4278 4279 4280 4281 4282 4283 4284 4285 4286 4287 4288 4289 4290 4291 4292 4293 4294 4295 4296 4297 4298 4299 4300 4301 4302 4303 4304 4305 4306 4307 4308 4309 4310 4311 4312 4313 4314 4315 4316 4317 4318 4319 4320 4321 4322 4323 4324 4325
/*
 * At the moment we always log all xattrs. This is to figure out at log replay
 * time which xattrs must have their deletion replayed. If a xattr is missing
 * in the log tree and exists in the fs/subvol tree, we delete it. This is
 * because if a xattr is deleted, the inode is fsynced and a power failure
 * happens, causing the log to be replayed the next time the fs is mounted,
 * we want the xattr to not exist anymore (same behaviour as other filesystems
 * with a journal, ext3/4, xfs, f2fs, etc).
 */
static int btrfs_log_all_xattrs(struct btrfs_trans_handle *trans,
				struct btrfs_root *root,
				struct inode *inode,
				struct btrfs_path *path,
				struct btrfs_path *dst_path)
{
	int ret;
	struct btrfs_key key;
	const u64 ino = btrfs_ino(inode);
	int ins_nr = 0;
	int start_slot = 0;

	key.objectid = ino;
	key.type = BTRFS_XATTR_ITEM_KEY;
	key.offset = 0;

	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
	if (ret < 0)
		return ret;

	while (true) {
		int slot = path->slots[0];
		struct extent_buffer *leaf = path->nodes[0];
		int nritems = btrfs_header_nritems(leaf);

		if (slot >= nritems) {
			if (ins_nr > 0) {
				u64 last_extent = 0;

				ret = copy_items(trans, inode, dst_path, path,
						 &last_extent, start_slot,
						 ins_nr, 1, 0);
				/* can't be 1, extent items aren't processed */
				ASSERT(ret <= 0);
				if (ret < 0)
					return ret;
				ins_nr = 0;
			}
			ret = btrfs_next_leaf(root, path);
			if (ret < 0)
				return ret;
			else if (ret > 0)
				break;
			continue;
		}

		btrfs_item_key_to_cpu(leaf, &key, slot);
		if (key.objectid != ino || key.type != BTRFS_XATTR_ITEM_KEY)
			break;

		if (ins_nr == 0)
			start_slot = slot;
		ins_nr++;
		path->slots[0]++;
		cond_resched();
	}
	if (ins_nr > 0) {
		u64 last_extent = 0;

		ret = copy_items(trans, inode, dst_path, path,
				 &last_extent, start_slot,
				 ins_nr, 1, 0);
		/* can't be 1, extent items aren't processed */
		ASSERT(ret <= 0);
		if (ret < 0)
			return ret;
	}

	return 0;
}

4326 4327 4328 4329 4330 4331 4332 4333 4334 4335 4336 4337 4338 4339 4340 4341 4342 4343 4344 4345 4346 4347 4348 4349 4350 4351 4352 4353 4354 4355 4356 4357 4358 4359 4360 4361 4362 4363 4364 4365 4366 4367 4368 4369 4370 4371 4372 4373 4374 4375 4376 4377 4378 4379 4380 4381 4382 4383 4384 4385 4386 4387 4388 4389 4390 4391 4392 4393 4394 4395 4396 4397 4398 4399 4400 4401 4402 4403 4404 4405 4406 4407 4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418 4419 4420 4421 4422 4423 4424 4425 4426
/*
 * If the no holes feature is enabled we need to make sure any hole between the
 * last extent and the i_size of our inode is explicitly marked in the log. This
 * is to make sure that doing something like:
 *
 *      1) create file with 128Kb of data
 *      2) truncate file to 64Kb
 *      3) truncate file to 256Kb
 *      4) fsync file
 *      5) <crash/power failure>
 *      6) mount fs and trigger log replay
 *
 * Will give us a file with a size of 256Kb, the first 64Kb of data match what
 * the file had in its first 64Kb of data at step 1 and the last 192Kb of the
 * file correspond to a hole. The presence of explicit holes in a log tree is
 * what guarantees that log replay will remove/adjust file extent items in the
 * fs/subvol tree.
 *
 * Here we do not need to care about holes between extents, that is already done
 * by copy_items(). We also only need to do this in the full sync path, where we
 * lookup for extents from the fs/subvol tree only. In the fast path case, we
 * lookup the list of modified extent maps and if any represents a hole, we
 * insert a corresponding extent representing a hole in the log tree.
 */
static int btrfs_log_trailing_hole(struct btrfs_trans_handle *trans,
				   struct btrfs_root *root,
				   struct inode *inode,
				   struct btrfs_path *path)
{
	int ret;
	struct btrfs_key key;
	u64 hole_start;
	u64 hole_size;
	struct extent_buffer *leaf;
	struct btrfs_root *log = root->log_root;
	const u64 ino = btrfs_ino(inode);
	const u64 i_size = i_size_read(inode);

	if (!btrfs_fs_incompat(root->fs_info, NO_HOLES))
		return 0;

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

	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
	ASSERT(ret != 0);
	if (ret < 0)
		return ret;

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

	if (key.objectid != ino || key.type != BTRFS_EXTENT_DATA_KEY) {
		/* inode does not have any extents */
		hole_start = 0;
		hole_size = i_size;
	} else {
		struct btrfs_file_extent_item *extent;
		u64 len;

		/*
		 * If there's an extent beyond i_size, an explicit hole was
		 * already inserted by copy_items().
		 */
		if (key.offset >= i_size)
			return 0;

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

		if (btrfs_file_extent_type(leaf, extent) ==
		    BTRFS_FILE_EXTENT_INLINE) {
			len = btrfs_file_extent_inline_len(leaf,
							   path->slots[0],
							   extent);
			ASSERT(len == i_size);
			return 0;
		}

		len = btrfs_file_extent_num_bytes(leaf, extent);
		/* Last extent goes beyond i_size, no need to log a hole. */
		if (key.offset + len > i_size)
			return 0;
		hole_start = key.offset + len;
		hole_size = i_size - hole_start;
	}
	btrfs_release_path(path);

	/* Last extent ends at i_size. */
	if (hole_size == 0)
		return 0;

	hole_size = ALIGN(hole_size, root->sectorsize);
	ret = btrfs_insert_file_extent(trans, log, ino, hole_start, 0, 0,
				       hole_size, 0, hole_size, 0, 0, 0);
	return ret;
}

4427 4428 4429 4430 4431 4432 4433 4434 4435 4436 4437 4438 4439 4440 4441 4442 4443 4444 4445 4446 4447 4448 4449 4450 4451 4452 4453 4454 4455 4456 4457 4458 4459 4460 4461 4462 4463 4464 4465 4466 4467 4468 4469 4470 4471 4472 4473 4474 4475 4476 4477 4478 4479 4480 4481 4482 4483 4484 4485 4486 4487 4488 4489 4490 4491 4492 4493 4494 4495 4496 4497 4498 4499 4500 4501 4502 4503 4504 4505 4506 4507 4508 4509 4510 4511 4512 4513 4514 4515 4516 4517 4518 4519 4520 4521 4522 4523 4524 4525 4526 4527 4528 4529 4530 4531 4532 4533 4534 4535 4536 4537 4538 4539 4540 4541 4542 4543 4544 4545 4546 4547
/*
 * When we are logging a new inode X, check if it doesn't have a reference that
 * matches the reference from some other inode Y created in a past transaction
 * and that was renamed in the current transaction. If we don't do this, then at
 * log replay time we can lose inode Y (and all its files if it's a directory):
 *
 * mkdir /mnt/x
 * echo "hello world" > /mnt/x/foobar
 * sync
 * mv /mnt/x /mnt/y
 * mkdir /mnt/x                 # or touch /mnt/x
 * xfs_io -c fsync /mnt/x
 * <power fail>
 * mount fs, trigger log replay
 *
 * After the log replay procedure, we would lose the first directory and all its
 * files (file foobar).
 * For the case where inode Y is not a directory we simply end up losing it:
 *
 * echo "123" > /mnt/foo
 * sync
 * mv /mnt/foo /mnt/bar
 * echo "abc" > /mnt/foo
 * xfs_io -c fsync /mnt/foo
 * <power fail>
 *
 * We also need this for cases where a snapshot entry is replaced by some other
 * entry (file or directory) otherwise we end up with an unreplayable log due to
 * attempts to delete the snapshot entry (entry of type BTRFS_ROOT_ITEM_KEY) as
 * if it were a regular entry:
 *
 * mkdir /mnt/x
 * btrfs subvolume snapshot /mnt /mnt/x/snap
 * btrfs subvolume delete /mnt/x/snap
 * rmdir /mnt/x
 * mkdir /mnt/x
 * fsync /mnt/x or fsync some new file inside it
 * <power fail>
 *
 * The snapshot delete, rmdir of x, mkdir of a new x and the fsync all happen in
 * the same transaction.
 */
static int btrfs_check_ref_name_override(struct extent_buffer *eb,
					 const int slot,
					 const struct btrfs_key *key,
					 struct inode *inode)
{
	int ret;
	struct btrfs_path *search_path;
	char *name = NULL;
	u32 name_len = 0;
	u32 item_size = btrfs_item_size_nr(eb, slot);
	u32 cur_offset = 0;
	unsigned long ptr = btrfs_item_ptr_offset(eb, slot);

	search_path = btrfs_alloc_path();
	if (!search_path)
		return -ENOMEM;
	search_path->search_commit_root = 1;
	search_path->skip_locking = 1;

	while (cur_offset < item_size) {
		u64 parent;
		u32 this_name_len;
		u32 this_len;
		unsigned long name_ptr;
		struct btrfs_dir_item *di;

		if (key->type == BTRFS_INODE_REF_KEY) {
			struct btrfs_inode_ref *iref;

			iref = (struct btrfs_inode_ref *)(ptr + cur_offset);
			parent = key->offset;
			this_name_len = btrfs_inode_ref_name_len(eb, iref);
			name_ptr = (unsigned long)(iref + 1);
			this_len = sizeof(*iref) + this_name_len;
		} else {
			struct btrfs_inode_extref *extref;

			extref = (struct btrfs_inode_extref *)(ptr +
							       cur_offset);
			parent = btrfs_inode_extref_parent(eb, extref);
			this_name_len = btrfs_inode_extref_name_len(eb, extref);
			name_ptr = (unsigned long)&extref->name;
			this_len = sizeof(*extref) + this_name_len;
		}

		if (this_name_len > name_len) {
			char *new_name;

			new_name = krealloc(name, this_name_len, GFP_NOFS);
			if (!new_name) {
				ret = -ENOMEM;
				goto out;
			}
			name_len = this_name_len;
			name = new_name;
		}

		read_extent_buffer(eb, name, name_ptr, this_name_len);
		di = btrfs_lookup_dir_item(NULL, BTRFS_I(inode)->root,
					   search_path, parent,
					   name, this_name_len, 0);
		if (di && !IS_ERR(di)) {
			ret = 1;
			goto out;
		} else if (IS_ERR(di)) {
			ret = PTR_ERR(di);
			goto out;
		}
		btrfs_release_path(search_path);

		cur_offset += this_len;
	}
	ret = 0;
out:
	btrfs_free_path(search_path);
	kfree(name);
	return ret;
}

4548 4549 4550 4551 4552 4553 4554 4555 4556 4557 4558 4559 4560 4561
/* 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.
 */
4562
static int btrfs_log_inode(struct btrfs_trans_handle *trans,
4563 4564 4565
			   struct btrfs_root *root, struct inode *inode,
			   int inode_only,
			   const loff_t start,
4566 4567
			   const loff_t end,
			   struct btrfs_log_ctx *ctx)
4568 4569 4570 4571 4572 4573
{
	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;
4574
	struct extent_buffer *src = NULL;
4575
	LIST_HEAD(logged_list);
4576
	u64 last_extent = 0;
4577
	int err = 0;
4578
	int ret;
4579
	int nritems;
4580 4581
	int ins_start_slot = 0;
	int ins_nr;
J
Josef Bacik 已提交
4582
	bool fast_search = false;
L
Li Zefan 已提交
4583
	u64 ino = btrfs_ino(inode);
4584
	struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
4585
	u64 logged_isize = 0;
4586
	bool need_log_inode_item = true;
4587 4588

	path = btrfs_alloc_path();
4589 4590
	if (!path)
		return -ENOMEM;
4591
	dst_path = btrfs_alloc_path();
4592 4593 4594 4595
	if (!dst_path) {
		btrfs_free_path(path);
		return -ENOMEM;
	}
4596

L
Li Zefan 已提交
4597
	min_key.objectid = ino;
4598 4599 4600
	min_key.type = BTRFS_INODE_ITEM_KEY;
	min_key.offset = 0;

L
Li Zefan 已提交
4601
	max_key.objectid = ino;
4602 4603


J
Josef Bacik 已提交
4604
	/* today the code can only do partial logging of directories */
4605 4606 4607 4608
	if (S_ISDIR(inode->i_mode) ||
	    (!test_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
		       &BTRFS_I(inode)->runtime_flags) &&
	     inode_only == LOG_INODE_EXISTS))
4609 4610 4611 4612 4613
		max_key.type = BTRFS_XATTR_ITEM_KEY;
	else
		max_key.type = (u8)-1;
	max_key.offset = (u64)-1;

4614 4615 4616 4617 4618 4619
	/*
	 * Only run delayed items if we are a dir or a new file.
	 * Otherwise commit the delayed inode only, which is needed in
	 * order for the log replay code to mark inodes for link count
	 * fixup (create temporary BTRFS_TREE_LOG_FIXUP_OBJECTID items).
	 */
4620
	if (S_ISDIR(inode->i_mode) ||
4621
	    BTRFS_I(inode)->generation > root->fs_info->last_trans_committed)
4622
		ret = btrfs_commit_inode_delayed_items(trans, inode);
4623 4624 4625 4626 4627 4628 4629
	else
		ret = btrfs_commit_inode_delayed_inode(inode);

	if (ret) {
		btrfs_free_path(path);
		btrfs_free_path(dst_path);
		return ret;
4630 4631
	}

4632 4633 4634 4635 4636 4637 4638 4639 4640
	mutex_lock(&BTRFS_I(inode)->log_mutex);

	/*
	 * 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;

4641 4642
		if (inode_only == LOG_INODE_EXISTS)
			max_key_type = BTRFS_XATTR_ITEM_KEY;
L
Li Zefan 已提交
4643
		ret = drop_objectid_items(trans, log, path, ino, max_key_type);
4644
	} else {
4645 4646 4647 4648 4649 4650 4651 4652 4653 4654 4655 4656 4657 4658 4659 4660 4661 4662 4663
		if (inode_only == LOG_INODE_EXISTS) {
			/*
			 * Make sure the new inode item we write to the log has
			 * the same isize as the current one (if it exists).
			 * This is necessary to prevent data loss after log
			 * replay, and also to prevent doing a wrong expanding
			 * truncate - for e.g. create file, write 4K into offset
			 * 0, fsync, write 4K into offset 4096, add hard link,
			 * fsync some other file (to sync log), power fail - if
			 * we use the inode's current i_size, after log replay
			 * we get a 8Kb file, with the last 4Kb extent as a hole
			 * (zeroes), as if an expanding truncate happened,
			 * instead of getting a file of 4Kb only.
			 */
			err = logged_inode_size(log, inode, path,
						&logged_isize);
			if (err)
				goto out_unlock;
		}
4664 4665 4666
		if (test_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
			     &BTRFS_I(inode)->runtime_flags)) {
			if (inode_only == LOG_INODE_EXISTS) {
4667
				max_key.type = BTRFS_XATTR_ITEM_KEY;
4668 4669 4670 4671 4672 4673 4674
				ret = drop_objectid_items(trans, log, path, ino,
							  max_key.type);
			} else {
				clear_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
					  &BTRFS_I(inode)->runtime_flags);
				clear_bit(BTRFS_INODE_COPY_EVERYTHING,
					  &BTRFS_I(inode)->runtime_flags);
4675 4676 4677 4678 4679 4680
				while(1) {
					ret = btrfs_truncate_inode_items(trans,
							 log, inode, 0, 0);
					if (ret != -EAGAIN)
						break;
				}
4681
			}
4682 4683
		} else if (test_and_clear_bit(BTRFS_INODE_COPY_EVERYTHING,
					      &BTRFS_I(inode)->runtime_flags) ||
4684
			   inode_only == LOG_INODE_EXISTS) {
4685
			if (inode_only == LOG_INODE_ALL)
4686
				fast_search = true;
4687
			max_key.type = BTRFS_XATTR_ITEM_KEY;
J
Josef Bacik 已提交
4688
			ret = drop_objectid_items(trans, log, path, ino,
4689
						  max_key.type);
4690 4691 4692 4693
		} else {
			if (inode_only == LOG_INODE_ALL)
				fast_search = true;
			goto log_extents;
J
Josef Bacik 已提交
4694
		}
4695

4696
	}
4697 4698 4699 4700
	if (ret) {
		err = ret;
		goto out_unlock;
	}
4701

C
Chris Mason 已提交
4702
	while (1) {
4703
		ins_nr = 0;
4704
		ret = btrfs_search_forward(root, &min_key,
4705
					   path, trans->transid);
4706 4707 4708 4709
		if (ret < 0) {
			err = ret;
			goto out_unlock;
		}
4710 4711
		if (ret != 0)
			break;
4712
again:
4713
		/* note, ins_nr might be > 0 here, cleanup outside the loop */
L
Li Zefan 已提交
4714
		if (min_key.objectid != ino)
4715 4716 4717
			break;
		if (min_key.type > max_key.type)
			break;
4718

4719 4720 4721
		if (min_key.type == BTRFS_INODE_ITEM_KEY)
			need_log_inode_item = false;

4722 4723 4724 4725 4726 4727 4728 4729 4730 4731 4732 4733 4734 4735 4736 4737
		if ((min_key.type == BTRFS_INODE_REF_KEY ||
		     min_key.type == BTRFS_INODE_EXTREF_KEY) &&
		    BTRFS_I(inode)->generation == trans->transid) {
			ret = btrfs_check_ref_name_override(path->nodes[0],
							    path->slots[0],
							    &min_key, inode);
			if (ret < 0) {
				err = ret;
				goto out_unlock;
			} else if (ret > 0) {
				err = 1;
				btrfs_set_log_full_commit(root->fs_info, trans);
				goto out_unlock;
			}
		}

4738 4739 4740 4741 4742 4743 4744 4745 4746 4747 4748 4749 4750 4751 4752 4753 4754 4755 4756
		/* Skip xattrs, we log them later with btrfs_log_all_xattrs() */
		if (min_key.type == BTRFS_XATTR_ITEM_KEY) {
			if (ins_nr == 0)
				goto next_slot;
			ret = copy_items(trans, inode, dst_path, path,
					 &last_extent, ins_start_slot,
					 ins_nr, inode_only, logged_isize);
			if (ret < 0) {
				err = ret;
				goto out_unlock;
			}
			ins_nr = 0;
			if (ret) {
				btrfs_release_path(path);
				continue;
			}
			goto next_slot;
		}

4757
		src = path->nodes[0];
4758 4759 4760 4761 4762 4763 4764
		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;
4765 4766
		}

4767
		ret = copy_items(trans, inode, dst_path, path, &last_extent,
4768 4769
				 ins_start_slot, ins_nr, inode_only,
				 logged_isize);
4770
		if (ret < 0) {
4771 4772
			err = ret;
			goto out_unlock;
4773 4774
		}
		if (ret) {
4775 4776 4777
			ins_nr = 0;
			btrfs_release_path(path);
			continue;
4778
		}
4779 4780 4781
		ins_nr = 1;
		ins_start_slot = path->slots[0];
next_slot:
4782

4783 4784 4785 4786 4787 4788 4789
		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;
		}
4790
		if (ins_nr) {
4791 4792
			ret = copy_items(trans, inode, dst_path, path,
					 &last_extent, ins_start_slot,
4793
					 ins_nr, inode_only, logged_isize);
4794
			if (ret < 0) {
4795 4796 4797
				err = ret;
				goto out_unlock;
			}
4798
			ret = 0;
4799 4800
			ins_nr = 0;
		}
4801
		btrfs_release_path(path);
4802

4803
		if (min_key.offset < (u64)-1) {
4804
			min_key.offset++;
4805
		} else if (min_key.type < max_key.type) {
4806
			min_key.type++;
4807 4808
			min_key.offset = 0;
		} else {
4809
			break;
4810
		}
4811
	}
4812
	if (ins_nr) {
4813
		ret = copy_items(trans, inode, dst_path, path, &last_extent,
4814 4815
				 ins_start_slot, ins_nr, inode_only,
				 logged_isize);
4816
		if (ret < 0) {
4817 4818 4819
			err = ret;
			goto out_unlock;
		}
4820
		ret = 0;
4821 4822
		ins_nr = 0;
	}
J
Josef Bacik 已提交
4823

4824 4825 4826 4827 4828
	btrfs_release_path(path);
	btrfs_release_path(dst_path);
	err = btrfs_log_all_xattrs(trans, root, inode, path, dst_path);
	if (err)
		goto out_unlock;
4829 4830 4831 4832 4833 4834 4835
	if (max_key.type >= BTRFS_EXTENT_DATA_KEY && !fast_search) {
		btrfs_release_path(path);
		btrfs_release_path(dst_path);
		err = btrfs_log_trailing_hole(trans, root, inode, path);
		if (err)
			goto out_unlock;
	}
4836
log_extents:
4837 4838
	btrfs_release_path(path);
	btrfs_release_path(dst_path);
4839 4840 4841 4842 4843
	if (need_log_inode_item) {
		err = log_inode_item(trans, log, dst_path, inode);
		if (err)
			goto out_unlock;
	}
J
Josef Bacik 已提交
4844
	if (fast_search) {
4845
		ret = btrfs_log_changed_extents(trans, root, inode, dst_path,
4846
						&logged_list, ctx, start, end);
J
Josef Bacik 已提交
4847 4848 4849 4850
		if (ret) {
			err = ret;
			goto out_unlock;
		}
4851
	} else if (inode_only == LOG_INODE_ALL) {
4852 4853
		struct extent_map *em, *n;

4854 4855 4856 4857 4858 4859 4860 4861 4862 4863 4864 4865 4866 4867 4868 4869 4870 4871 4872 4873 4874 4875 4876 4877 4878 4879 4880
		write_lock(&em_tree->lock);
		/*
		 * We can't just remove every em if we're called for a ranged
		 * fsync - that is, one that doesn't cover the whole possible
		 * file range (0 to LLONG_MAX). This is because we can have
		 * em's that fall outside the range we're logging and therefore
		 * their ordered operations haven't completed yet
		 * (btrfs_finish_ordered_io() not invoked yet). This means we
		 * didn't get their respective file extent item in the fs/subvol
		 * tree yet, and need to let the next fast fsync (one which
		 * consults the list of modified extent maps) find the em so
		 * that it logs a matching file extent item and waits for the
		 * respective ordered operation to complete (if it's still
		 * running).
		 *
		 * Removing every em outside the range we're logging would make
		 * the next fast fsync not log their matching file extent items,
		 * therefore making us lose data after a log replay.
		 */
		list_for_each_entry_safe(em, n, &em_tree->modified_extents,
					 list) {
			const u64 mod_end = em->mod_start + em->mod_len - 1;

			if (em->mod_start >= start && mod_end <= end)
				list_del_init(&em->list);
		}
		write_unlock(&em_tree->lock);
J
Josef Bacik 已提交
4881 4882
	}

4883
	if (inode_only == LOG_INODE_ALL && S_ISDIR(inode->i_mode)) {
4884 4885
		ret = log_directory_changes(trans, root, inode, path, dst_path,
					    ctx);
4886 4887 4888 4889
		if (ret) {
			err = ret;
			goto out_unlock;
		}
4890
	}
4891

4892
	spin_lock(&BTRFS_I(inode)->lock);
4893 4894
	BTRFS_I(inode)->logged_trans = trans->transid;
	BTRFS_I(inode)->last_log_commit = BTRFS_I(inode)->last_sub_trans;
4895
	spin_unlock(&BTRFS_I(inode)->lock);
4896
out_unlock:
4897 4898 4899 4900
	if (unlikely(err))
		btrfs_put_logged_extents(&logged_list);
	else
		btrfs_submit_logged_extents(&logged_list, log);
4901 4902 4903 4904
	mutex_unlock(&BTRFS_I(inode)->log_mutex);

	btrfs_free_path(path);
	btrfs_free_path(dst_path);
4905
	return err;
4906 4907
}

4908 4909 4910 4911 4912 4913 4914 4915 4916 4917 4918 4919 4920
/*
 * Check if we must fallback to a transaction commit when logging an inode.
 * This must be called after logging the inode and is used only in the context
 * when fsyncing an inode requires the need to log some other inode - in which
 * case we can't lock the i_mutex of each other inode we need to log as that
 * can lead to deadlocks with concurrent fsync against other inodes (as we can
 * log inodes up or down in the hierarchy) or rename operations for example. So
 * we take the log_mutex of the inode after we have logged it and then check for
 * its last_unlink_trans value - this is safe because any task setting
 * last_unlink_trans must take the log_mutex and it must do this before it does
 * the actual unlink operation, so if we do this check before a concurrent task
 * sets last_unlink_trans it means we've logged a consistent version/state of
 * all the inode items, otherwise we are not sure and must do a transaction
4921
 * commit (the concurrent task might have only updated last_unlink_trans before
4922 4923 4924 4925 4926 4927 4928 4929 4930 4931 4932 4933 4934 4935 4936 4937 4938 4939 4940 4941 4942 4943
 * we logged the inode or it might have also done the unlink).
 */
static bool btrfs_must_commit_transaction(struct btrfs_trans_handle *trans,
					  struct inode *inode)
{
	struct btrfs_fs_info *fs_info = BTRFS_I(inode)->root->fs_info;
	bool ret = false;

	mutex_lock(&BTRFS_I(inode)->log_mutex);
	if (BTRFS_I(inode)->last_unlink_trans > fs_info->last_trans_committed) {
		/*
		 * Make sure any commits to the log are forced to be full
		 * commits.
		 */
		btrfs_set_log_full_commit(fs_info, trans);
		ret = true;
	}
	mutex_unlock(&BTRFS_I(inode)->log_mutex);

	return ret;
}

4944 4945 4946 4947 4948 4949 4950 4951 4952 4953 4954
/*
 * 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)
4955
{
4956
	int ret = 0;
4957
	struct dentry *old_parent = NULL;
4958
	struct inode *orig_inode = inode;
4959

4960 4961 4962 4963 4964 4965 4966 4967 4968 4969 4970
	/*
	 * 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;

4971
	if (!S_ISDIR(inode->i_mode)) {
4972
		if (!parent || d_really_is_negative(parent) || sb != parent->d_sb)
4973
			goto out;
4974
		inode = d_inode(parent);
4975 4976 4977
	}

	while (1) {
4978 4979
		/*
		 * If we are logging a directory then we start with our inode,
4980
		 * not our parent's inode, so we need to skip setting the
4981 4982 4983 4984 4985
		 * 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;
4986 4987
		smp_mb();

4988
		if (btrfs_must_commit_transaction(trans, inode)) {
4989 4990 4991 4992
			ret = 1;
			break;
		}

4993
		if (!parent || d_really_is_negative(parent) || sb != parent->d_sb)
4994 4995
			break;

4996
		if (IS_ROOT(parent))
4997 4998
			break;

4999 5000 5001
		parent = dget_parent(parent);
		dput(old_parent);
		old_parent = parent;
5002
		inode = d_inode(parent);
5003 5004

	}
5005
	dput(old_parent);
5006
out:
5007 5008 5009
	return ret;
}

5010 5011 5012 5013 5014 5015 5016 5017 5018 5019 5020 5021 5022 5023 5024 5025 5026 5027 5028 5029 5030 5031 5032 5033 5034 5035 5036 5037 5038 5039 5040 5041 5042 5043 5044 5045 5046 5047 5048 5049 5050 5051 5052 5053 5054 5055 5056 5057 5058 5059 5060 5061 5062 5063 5064 5065 5066 5067 5068 5069 5070 5071 5072 5073 5074 5075 5076 5077 5078 5079 5080 5081 5082 5083 5084 5085 5086 5087 5088 5089 5090 5091 5092 5093 5094 5095 5096 5097 5098 5099 5100 5101 5102 5103 5104 5105 5106 5107 5108 5109 5110 5111 5112 5113 5114 5115 5116 5117 5118 5119 5120 5121 5122 5123 5124 5125 5126 5127 5128 5129 5130 5131 5132 5133 5134 5135 5136 5137 5138 5139 5140 5141
struct btrfs_dir_list {
	u64 ino;
	struct list_head list;
};

/*
 * Log the inodes of the new dentries of a directory. See log_dir_items() for
 * details about the why it is needed.
 * This is a recursive operation - if an existing dentry corresponds to a
 * directory, that directory's new entries are logged too (same behaviour as
 * ext3/4, xfs, f2fs, reiserfs, nilfs2). Note that when logging the inodes
 * the dentries point to we do not lock their i_mutex, otherwise lockdep
 * complains about the following circular lock dependency / possible deadlock:
 *
 *        CPU0                                        CPU1
 *        ----                                        ----
 * lock(&type->i_mutex_dir_key#3/2);
 *                                            lock(sb_internal#2);
 *                                            lock(&type->i_mutex_dir_key#3/2);
 * lock(&sb->s_type->i_mutex_key#14);
 *
 * Where sb_internal is the lock (a counter that works as a lock) acquired by
 * sb_start_intwrite() in btrfs_start_transaction().
 * Not locking i_mutex of the inodes is still safe because:
 *
 * 1) For regular files we log with a mode of LOG_INODE_EXISTS. It's possible
 *    that while logging the inode new references (names) are added or removed
 *    from the inode, leaving the logged inode item with a link count that does
 *    not match the number of logged inode reference items. This is fine because
 *    at log replay time we compute the real number of links and correct the
 *    link count in the inode item (see replay_one_buffer() and
 *    link_to_fixup_dir());
 *
 * 2) For directories we log with a mode of LOG_INODE_ALL. It's possible that
 *    while logging the inode's items new items with keys BTRFS_DIR_ITEM_KEY and
 *    BTRFS_DIR_INDEX_KEY are added to fs/subvol tree and the logged inode item
 *    has a size that doesn't match the sum of the lengths of all the logged
 *    names. This does not result in a problem because if a dir_item key is
 *    logged but its matching dir_index key is not logged, at log replay time we
 *    don't use it to replay the respective name (see replay_one_name()). On the
 *    other hand if only the dir_index key ends up being logged, the respective
 *    name is added to the fs/subvol tree with both the dir_item and dir_index
 *    keys created (see replay_one_name()).
 *    The directory's inode item with a wrong i_size is not a problem as well,
 *    since we don't use it at log replay time to set the i_size in the inode
 *    item of the fs/subvol tree (see overwrite_item()).
 */
static int log_new_dir_dentries(struct btrfs_trans_handle *trans,
				struct btrfs_root *root,
				struct inode *start_inode,
				struct btrfs_log_ctx *ctx)
{
	struct btrfs_root *log = root->log_root;
	struct btrfs_path *path;
	LIST_HEAD(dir_list);
	struct btrfs_dir_list *dir_elem;
	int ret = 0;

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

	dir_elem = kmalloc(sizeof(*dir_elem), GFP_NOFS);
	if (!dir_elem) {
		btrfs_free_path(path);
		return -ENOMEM;
	}
	dir_elem->ino = btrfs_ino(start_inode);
	list_add_tail(&dir_elem->list, &dir_list);

	while (!list_empty(&dir_list)) {
		struct extent_buffer *leaf;
		struct btrfs_key min_key;
		int nritems;
		int i;

		dir_elem = list_first_entry(&dir_list, struct btrfs_dir_list,
					    list);
		if (ret)
			goto next_dir_inode;

		min_key.objectid = dir_elem->ino;
		min_key.type = BTRFS_DIR_ITEM_KEY;
		min_key.offset = 0;
again:
		btrfs_release_path(path);
		ret = btrfs_search_forward(log, &min_key, path, trans->transid);
		if (ret < 0) {
			goto next_dir_inode;
		} else if (ret > 0) {
			ret = 0;
			goto next_dir_inode;
		}

process_leaf:
		leaf = path->nodes[0];
		nritems = btrfs_header_nritems(leaf);
		for (i = path->slots[0]; i < nritems; i++) {
			struct btrfs_dir_item *di;
			struct btrfs_key di_key;
			struct inode *di_inode;
			struct btrfs_dir_list *new_dir_elem;
			int log_mode = LOG_INODE_EXISTS;
			int type;

			btrfs_item_key_to_cpu(leaf, &min_key, i);
			if (min_key.objectid != dir_elem->ino ||
			    min_key.type != BTRFS_DIR_ITEM_KEY)
				goto next_dir_inode;

			di = btrfs_item_ptr(leaf, i, struct btrfs_dir_item);
			type = btrfs_dir_type(leaf, di);
			if (btrfs_dir_transid(leaf, di) < trans->transid &&
			    type != BTRFS_FT_DIR)
				continue;
			btrfs_dir_item_key_to_cpu(leaf, di, &di_key);
			if (di_key.type == BTRFS_ROOT_ITEM_KEY)
				continue;

			di_inode = btrfs_iget(root->fs_info->sb, &di_key,
					      root, NULL);
			if (IS_ERR(di_inode)) {
				ret = PTR_ERR(di_inode);
				goto next_dir_inode;
			}

			if (btrfs_inode_in_log(di_inode, trans->transid)) {
				iput(di_inode);
				continue;
			}

			ctx->log_new_dentries = false;
5142
			if (type == BTRFS_FT_DIR || type == BTRFS_FT_SYMLINK)
5143 5144 5145 5146
				log_mode = LOG_INODE_ALL;
			btrfs_release_path(path);
			ret = btrfs_log_inode(trans, root, di_inode,
					      log_mode, 0, LLONG_MAX, ctx);
5147 5148 5149
			if (!ret &&
			    btrfs_must_commit_transaction(trans, di_inode))
				ret = 1;
5150 5151 5152 5153 5154 5155 5156 5157 5158 5159 5160 5161 5162 5163 5164 5165 5166 5167 5168 5169 5170 5171 5172 5173 5174 5175 5176 5177 5178 5179 5180 5181 5182 5183 5184 5185 5186 5187
			iput(di_inode);
			if (ret)
				goto next_dir_inode;
			if (ctx->log_new_dentries) {
				new_dir_elem = kmalloc(sizeof(*new_dir_elem),
						       GFP_NOFS);
				if (!new_dir_elem) {
					ret = -ENOMEM;
					goto next_dir_inode;
				}
				new_dir_elem->ino = di_key.objectid;
				list_add_tail(&new_dir_elem->list, &dir_list);
			}
			break;
		}
		if (i == nritems) {
			ret = btrfs_next_leaf(log, path);
			if (ret < 0) {
				goto next_dir_inode;
			} else if (ret > 0) {
				ret = 0;
				goto next_dir_inode;
			}
			goto process_leaf;
		}
		if (min_key.offset < (u64)-1) {
			min_key.offset++;
			goto again;
		}
next_dir_inode:
		list_del(&dir_elem->list);
		kfree(dir_elem);
	}

	btrfs_free_path(path);
	return ret;
}

5188 5189 5190 5191 5192 5193 5194 5195 5196 5197 5198 5199 5200 5201 5202 5203 5204 5205 5206 5207 5208 5209 5210 5211 5212 5213 5214 5215 5216 5217 5218 5219 5220 5221 5222 5223 5224 5225 5226 5227 5228 5229 5230 5231 5232 5233 5234 5235 5236 5237 5238 5239 5240 5241 5242 5243 5244 5245 5246 5247 5248 5249 5250 5251 5252 5253 5254 5255 5256 5257 5258 5259 5260 5261
static int btrfs_log_all_parents(struct btrfs_trans_handle *trans,
				 struct inode *inode,
				 struct btrfs_log_ctx *ctx)
{
	int ret;
	struct btrfs_path *path;
	struct btrfs_key key;
	struct btrfs_root *root = BTRFS_I(inode)->root;
	const u64 ino = btrfs_ino(inode);

	path = btrfs_alloc_path();
	if (!path)
		return -ENOMEM;
	path->skip_locking = 1;
	path->search_commit_root = 1;

	key.objectid = ino;
	key.type = BTRFS_INODE_REF_KEY;
	key.offset = 0;
	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
	if (ret < 0)
		goto out;

	while (true) {
		struct extent_buffer *leaf = path->nodes[0];
		int slot = path->slots[0];
		u32 cur_offset = 0;
		u32 item_size;
		unsigned long ptr;

		if (slot >= btrfs_header_nritems(leaf)) {
			ret = btrfs_next_leaf(root, path);
			if (ret < 0)
				goto out;
			else if (ret > 0)
				break;
			continue;
		}

		btrfs_item_key_to_cpu(leaf, &key, slot);
		/* BTRFS_INODE_EXTREF_KEY is BTRFS_INODE_REF_KEY + 1 */
		if (key.objectid != ino || key.type > BTRFS_INODE_EXTREF_KEY)
			break;

		item_size = btrfs_item_size_nr(leaf, slot);
		ptr = btrfs_item_ptr_offset(leaf, slot);
		while (cur_offset < item_size) {
			struct btrfs_key inode_key;
			struct inode *dir_inode;

			inode_key.type = BTRFS_INODE_ITEM_KEY;
			inode_key.offset = 0;

			if (key.type == BTRFS_INODE_EXTREF_KEY) {
				struct btrfs_inode_extref *extref;

				extref = (struct btrfs_inode_extref *)
					(ptr + cur_offset);
				inode_key.objectid = btrfs_inode_extref_parent(
					leaf, extref);
				cur_offset += sizeof(*extref);
				cur_offset += btrfs_inode_extref_name_len(leaf,
					extref);
			} else {
				inode_key.objectid = key.offset;
				cur_offset = item_size;
			}

			dir_inode = btrfs_iget(root->fs_info->sb, &inode_key,
					       root, NULL);
			/* If parent inode was deleted, skip it. */
			if (IS_ERR(dir_inode))
				continue;

5262 5263
			if (ctx)
				ctx->log_new_dentries = false;
5264 5265
			ret = btrfs_log_inode(trans, root, dir_inode,
					      LOG_INODE_ALL, 0, LLONG_MAX, ctx);
5266 5267 5268
			if (!ret &&
			    btrfs_must_commit_transaction(trans, dir_inode))
				ret = 1;
5269 5270 5271
			if (!ret && ctx && ctx->log_new_dentries)
				ret = log_new_dir_dentries(trans, root,
							   dir_inode, ctx);
5272 5273 5274 5275 5276 5277 5278 5279 5280 5281 5282 5283
			iput(dir_inode);
			if (ret)
				goto out;
		}
		path->slots[0]++;
	}
	ret = 0;
out:
	btrfs_free_path(path);
	return ret;
}

5284 5285 5286 5287 5288 5289
/*
 * 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
 */
5290 5291
static int btrfs_log_inode_parent(struct btrfs_trans_handle *trans,
			    	  struct btrfs_root *root, struct inode *inode,
5292 5293 5294 5295
				  struct dentry *parent,
				  const loff_t start,
				  const loff_t end,
				  int exists_only,
5296
				  struct btrfs_log_ctx *ctx)
5297
{
5298
	int inode_only = exists_only ? LOG_INODE_EXISTS : LOG_INODE_ALL;
5299
	struct super_block *sb;
5300
	struct dentry *old_parent = NULL;
5301 5302
	int ret = 0;
	u64 last_committed = root->fs_info->last_trans_committed;
5303 5304
	bool log_dentries = false;
	struct inode *orig_inode = inode;
5305 5306 5307

	sb = inode->i_sb;

S
Sage Weil 已提交
5308 5309 5310 5311 5312
	if (btrfs_test_opt(root, NOTREELOG)) {
		ret = 1;
		goto end_no_trans;
	}

5313 5314 5315 5316
	/*
	 * The prev transaction commit doesn't complete, we need do
	 * full commit by ourselves.
	 */
5317 5318 5319 5320 5321 5322
	if (root->fs_info->last_trans_log_full_commit >
	    root->fs_info->last_trans_committed) {
		ret = 1;
		goto end_no_trans;
	}

5323 5324 5325 5326 5327 5328
	if (root != BTRFS_I(inode)->root ||
	    btrfs_root_refs(&root->root_item) == 0) {
		ret = 1;
		goto end_no_trans;
	}

5329 5330 5331 5332
	ret = check_parent_dirs_for_sync(trans, inode, parent,
					 sb, last_committed);
	if (ret)
		goto end_no_trans;
5333

5334
	if (btrfs_inode_in_log(inode, trans->transid)) {
5335 5336 5337 5338
		ret = BTRFS_NO_LOG_SYNC;
		goto end_no_trans;
	}

5339
	ret = start_log_trans(trans, root, ctx);
5340
	if (ret)
5341
		goto end_no_trans;
5342

5343
	ret = btrfs_log_inode(trans, root, inode, inode_only, start, end, ctx);
5344 5345
	if (ret)
		goto end_trans;
5346

5347 5348 5349 5350 5351 5352 5353 5354
	/*
	 * 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 &&
5355 5356 5357 5358
	    BTRFS_I(inode)->last_unlink_trans <= last_committed) {
		ret = 0;
		goto end_trans;
	}
5359

5360 5361 5362
	if (S_ISDIR(inode->i_mode) && ctx && ctx->log_new_dentries)
		log_dentries = true;

5363
	/*
5364
	 * On unlink we must make sure all our current and old parent directory
5365 5366 5367 5368 5369 5370 5371 5372 5373 5374 5375 5376 5377 5378 5379 5380 5381 5382 5383 5384 5385 5386 5387 5388 5389 5390 5391 5392 5393 5394 5395 5396 5397 5398 5399 5400 5401 5402 5403 5404 5405 5406 5407 5408 5409
	 * inodes are fully logged. This is to prevent leaving dangling
	 * directory index entries in directories that were our parents but are
	 * not anymore. Not doing this results in old parent directory being
	 * impossible to delete after log replay (rmdir will always fail with
	 * error -ENOTEMPTY).
	 *
	 * Example 1:
	 *
	 * mkdir testdir
	 * touch testdir/foo
	 * ln testdir/foo testdir/bar
	 * sync
	 * unlink testdir/bar
	 * xfs_io -c fsync testdir/foo
	 * <power failure>
	 * mount fs, triggers log replay
	 *
	 * If we don't log the parent directory (testdir), after log replay the
	 * directory still has an entry pointing to the file inode using the bar
	 * name, but a matching BTRFS_INODE_[REF|EXTREF]_KEY does not exist and
	 * the file inode has a link count of 1.
	 *
	 * Example 2:
	 *
	 * mkdir testdir
	 * touch foo
	 * ln foo testdir/foo2
	 * ln foo testdir/foo3
	 * sync
	 * unlink testdir/foo3
	 * xfs_io -c fsync foo
	 * <power failure>
	 * mount fs, triggers log replay
	 *
	 * Similar as the first example, after log replay the parent directory
	 * testdir still has an entry pointing to the inode file with name foo3
	 * but the file inode does not have a matching BTRFS_INODE_REF_KEY item
	 * and has a link count of 2.
	 */
	if (BTRFS_I(inode)->last_unlink_trans > last_committed) {
		ret = btrfs_log_all_parents(trans, orig_inode, ctx);
		if (ret)
			goto end_trans;
	}

5410
	while (1) {
5411
		if (!parent || d_really_is_negative(parent) || sb != parent->d_sb)
5412 5413
			break;

5414
		inode = d_inode(parent);
5415 5416 5417
		if (root != BTRFS_I(inode)->root)
			break;

5418 5419 5420
		if (BTRFS_I(inode)->generation > last_committed) {
			ret = btrfs_log_inode(trans, root, inode,
					      LOG_INODE_EXISTS,
5421
					      0, LLONG_MAX, ctx);
5422 5423
			if (ret)
				goto end_trans;
5424
		}
5425
		if (IS_ROOT(parent))
5426
			break;
5427

5428 5429 5430
		parent = dget_parent(parent);
		dput(old_parent);
		old_parent = parent;
5431
	}
5432 5433 5434 5435
	if (log_dentries)
		ret = log_new_dir_dentries(trans, root, orig_inode, ctx);
	else
		ret = 0;
5436
end_trans:
5437
	dput(old_parent);
5438
	if (ret < 0) {
5439
		btrfs_set_log_full_commit(root->fs_info, trans);
5440 5441
		ret = 1;
	}
5442 5443 5444

	if (ret)
		btrfs_remove_log_ctx(root, ctx);
5445 5446 5447
	btrfs_end_log_trans(root);
end_no_trans:
	return ret;
5448 5449 5450 5451 5452 5453 5454 5455 5456
}

/*
 * 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,
5457
			  struct btrfs_root *root, struct dentry *dentry,
5458 5459
			  const loff_t start,
			  const loff_t end,
5460
			  struct btrfs_log_ctx *ctx)
5461
{
5462 5463 5464
	struct dentry *parent = dget_parent(dentry);
	int ret;

5465
	ret = btrfs_log_inode_parent(trans, root, d_inode(dentry), parent,
5466
				     start, end, 0, ctx);
5467 5468 5469
	dput(parent);

	return ret;
5470 5471 5472 5473 5474 5475 5476 5477 5478 5479 5480 5481 5482 5483 5484 5485 5486 5487 5488 5489 5490 5491
}

/*
 * 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
5492 5493 5494 5495
	if (!path)
		return -ENOMEM;

	fs_info->log_root_recovering = 1;
5496

5497
	trans = btrfs_start_transaction(fs_info->tree_root, 0);
5498 5499 5500 5501
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto error;
	}
5502 5503 5504 5505

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

T
Tsutomu Itoh 已提交
5506
	ret = walk_log_tree(trans, log_root_tree, &wc);
5507
	if (ret) {
5508
		btrfs_handle_fs_error(fs_info, ret, "Failed to pin buffers while "
5509 5510 5511
			    "recovering log root tree.");
		goto error;
	}
5512 5513 5514 5515

again:
	key.objectid = BTRFS_TREE_LOG_OBJECTID;
	key.offset = (u64)-1;
5516
	key.type = BTRFS_ROOT_ITEM_KEY;
5517

C
Chris Mason 已提交
5518
	while (1) {
5519
		ret = btrfs_search_slot(NULL, log_root_tree, &key, path, 0, 0);
5520 5521

		if (ret < 0) {
5522
			btrfs_handle_fs_error(fs_info, ret,
5523 5524 5525
				    "Couldn't find tree log root.");
			goto error;
		}
5526 5527 5528 5529 5530 5531 5532
		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]);
5533
		btrfs_release_path(path);
5534 5535 5536
		if (found_key.objectid != BTRFS_TREE_LOG_OBJECTID)
			break;

5537
		log = btrfs_read_fs_root(log_root_tree, &found_key);
5538 5539
		if (IS_ERR(log)) {
			ret = PTR_ERR(log);
5540
			btrfs_handle_fs_error(fs_info, ret,
5541 5542 5543
				    "Couldn't read tree log root.");
			goto error;
		}
5544 5545 5546 5547 5548 5549

		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);
5550 5551
		if (IS_ERR(wc.replay_dest)) {
			ret = PTR_ERR(wc.replay_dest);
5552 5553 5554
			free_extent_buffer(log->node);
			free_extent_buffer(log->commit_root);
			kfree(log);
5555
			btrfs_handle_fs_error(fs_info, ret, "Couldn't read target root "
5556 5557 5558
				    "for tree log recovery.");
			goto error;
		}
5559

Y
Yan Zheng 已提交
5560
		wc.replay_dest->log_root = log;
5561
		btrfs_record_root_in_trans(trans, wc.replay_dest);
5562 5563
		ret = walk_log_tree(trans, log, &wc);

5564
		if (!ret && wc.stage == LOG_WALK_REPLAY_ALL) {
5565 5566 5567 5568 5569
			ret = fixup_inode_link_counts(trans, wc.replay_dest,
						      path);
		}

		key.offset = found_key.offset - 1;
Y
Yan Zheng 已提交
5570
		wc.replay_dest->log_root = NULL;
5571
		free_extent_buffer(log->node);
5572
		free_extent_buffer(log->commit_root);
5573 5574
		kfree(log);

5575 5576 5577
		if (ret)
			goto error;

5578 5579 5580
		if (found_key.offset == 0)
			break;
	}
5581
	btrfs_release_path(path);
5582 5583 5584 5585 5586 5587 5588 5589 5590 5591 5592 5593 5594 5595 5596 5597

	/* 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);

5598 5599 5600 5601 5602
	/* step 4: commit the transaction, which also unpins the blocks */
	ret = btrfs_commit_transaction(trans, fs_info->tree_root);
	if (ret)
		return ret;

5603 5604 5605 5606
	free_extent_buffer(log_root_tree->node);
	log_root_tree->log_root = NULL;
	fs_info->log_root_recovering = 0;
	kfree(log_root_tree);
5607

5608
	return 0;
5609
error:
5610 5611
	if (wc.trans)
		btrfs_end_transaction(wc.trans, fs_info->tree_root);
5612 5613
	btrfs_free_path(path);
	return ret;
5614
}
5615 5616 5617 5618 5619 5620 5621 5622

/*
 * 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.
5623 5624 5625
 *
 * Must be called before the unlink operations (updates to the subvolume tree,
 * inodes, etc) are done.
5626 5627 5628 5629 5630
 */
void btrfs_record_unlink_dir(struct btrfs_trans_handle *trans,
			     struct inode *dir, struct inode *inode,
			     int for_rename)
{
5631 5632 5633 5634 5635 5636 5637 5638 5639 5640
	/*
	 * 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.
	 */
5641 5642 5643
	mutex_lock(&BTRFS_I(inode)->log_mutex);
	BTRFS_I(inode)->last_unlink_trans = trans->transid;
	mutex_unlock(&BTRFS_I(inode)->log_mutex);
5644

5645 5646 5647 5648 5649 5650 5651 5652 5653 5654 5655 5656 5657 5658 5659 5660 5661 5662 5663 5664 5665 5666 5667 5668 5669 5670 5671 5672 5673
	/*
	 * 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:
5674
	mutex_lock(&BTRFS_I(dir)->log_mutex);
5675
	BTRFS_I(dir)->last_unlink_trans = trans->transid;
5676
	mutex_unlock(&BTRFS_I(dir)->log_mutex);
5677 5678 5679 5680 5681 5682 5683 5684 5685 5686
}

/*
 * Make sure that if someone attempts to fsync the parent directory of a deleted
 * snapshot, it ends up triggering a transaction commit. This is to guarantee
 * that after replaying the log tree of the parent directory's root we will not
 * see the snapshot anymore and at log replay time we will not see any log tree
 * corresponding to the deleted snapshot's root, which could lead to replaying
 * it after replaying the log tree of the parent directory (which would replay
 * the snapshot delete operation).
5687 5688 5689
 *
 * Must be called before the actual snapshot destroy operation (updates to the
 * parent root and tree of tree roots trees, etc) are done.
5690 5691 5692 5693
 */
void btrfs_record_snapshot_destroy(struct btrfs_trans_handle *trans,
				   struct inode *dir)
{
5694
	mutex_lock(&BTRFS_I(dir)->log_mutex);
5695
	BTRFS_I(dir)->last_unlink_trans = trans->transid;
5696
	mutex_unlock(&BTRFS_I(dir)->log_mutex);
5697 5698 5699 5700 5701 5702 5703 5704 5705 5706 5707 5708 5709 5710 5711
}

/*
 * 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;

5712 5713 5714 5715 5716 5717 5718
	/*
	 * 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;

5719 5720 5721 5722 5723 5724 5725 5726 5727 5728
	/*
	 * 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;

5729 5730
	return btrfs_log_inode_parent(trans, root, inode, parent, 0,
				      LLONG_MAX, 1, NULL);
5731 5732
}