1// SPDX-License-Identifier: GPL-2.0
   2/*
   3 * Copyright (C) 2008 Oracle.  All rights reserved.
   4 */
   5
   6#include <linux/sched.h>
   7#include <linux/slab.h>
   8#include <linux/blkdev.h>
   9#include <linux/list_sort.h>
  10#include <linux/iversion.h>
  11#include "misc.h"
  12#include "ctree.h"
  13#include "tree-log.h"
  14#include "disk-io.h"
  15#include "locking.h"
 
  16#include "backref.h"
  17#include "compression.h"
  18#include "qgroup.h"
  19#include "block-group.h"
  20#include "space-info.h"
 
  21#include "inode-item.h"
  22#include "fs.h"
  23#include "accessors.h"
  24#include "extent-tree.h"
  25#include "root-tree.h"
  26#include "dir-item.h"
  27#include "file-item.h"
  28#include "file.h"
  29#include "orphan.h"
  30#include "tree-checker.h"
  31
  32#define MAX_CONFLICT_INODES 10
  33
  34/* magic values for the inode_only field in btrfs_log_inode:
  35 *
  36 * LOG_INODE_ALL means to log everything
  37 * LOG_INODE_EXISTS means to log just enough to recreate the inode
  38 * during log replay
  39 */
  40enum {
  41	LOG_INODE_ALL,
  42	LOG_INODE_EXISTS,
  43};
  44
  45/*
  46 * directory trouble cases
  47 *
  48 * 1) on rename or unlink, if the inode being unlinked isn't in the fsync
  49 * log, we must force a full commit before doing an fsync of the directory
  50 * where the unlink was done.
  51 * ---> record transid of last unlink/rename per directory
  52 *
  53 * mkdir foo/some_dir
  54 * normal commit
  55 * rename foo/some_dir foo2/some_dir
  56 * mkdir foo/some_dir
  57 * fsync foo/some_dir/some_file
  58 *
  59 * The fsync above will unlink the original some_dir without recording
  60 * it in its new location (foo2).  After a crash, some_dir will be gone
  61 * unless the fsync of some_file forces a full commit
  62 *
  63 * 2) we must log any new names for any file or dir that is in the fsync
  64 * log. ---> check inode while renaming/linking.
  65 *
  66 * 2a) we must log any new names for any file or dir during rename
  67 * when the directory they are being removed from was logged.
  68 * ---> check inode and old parent dir during rename
  69 *
  70 *  2a is actually the more important variant.  With the extra logging
  71 *  a crash might unlink the old name without recreating the new one
  72 *
  73 * 3) after a crash, we must go through any directories with a link count
  74 * of zero and redo the rm -rf
  75 *
  76 * mkdir f1/foo
  77 * normal commit
  78 * rm -rf f1/foo
  79 * fsync(f1)
  80 *
  81 * The directory f1 was fully removed from the FS, but fsync was never
  82 * called on f1, only its parent dir.  After a crash the rm -rf must
  83 * be replayed.  This must be able to recurse down the entire
  84 * directory tree.  The inode link count fixup code takes care of the
  85 * ugly details.
  86 */
  87
  88/*
  89 * stages for the tree walking.  The first
  90 * stage (0) is to only pin down the blocks we find
  91 * the second stage (1) is to make sure that all the inodes
  92 * we find in the log are created in the subvolume.
  93 *
  94 * The last stage is to deal with directories and links and extents
  95 * and all the other fun semantics
  96 */
  97enum {
  98	LOG_WALK_PIN_ONLY,
  99	LOG_WALK_REPLAY_INODES,
 100	LOG_WALK_REPLAY_DIR_INDEX,
 101	LOG_WALK_REPLAY_ALL,
 102};
 103
 104static int btrfs_log_inode(struct btrfs_trans_handle *trans,
 105			   struct btrfs_inode *inode,
 106			   int inode_only,
 107			   struct btrfs_log_ctx *ctx);
 108static int link_to_fixup_dir(struct btrfs_trans_handle *trans,
 109			     struct btrfs_root *root,
 110			     struct btrfs_path *path, u64 objectid);
 111static noinline int replay_dir_deletes(struct btrfs_trans_handle *trans,
 112				       struct btrfs_root *root,
 113				       struct btrfs_root *log,
 114				       struct btrfs_path *path,
 115				       u64 dirid, int del_all);
 116static void wait_log_commit(struct btrfs_root *root, int transid);
 117
 118/*
 119 * tree logging is a special write ahead log used to make sure that
 120 * fsyncs and O_SYNCs can happen without doing full tree commits.
 121 *
 122 * Full tree commits are expensive because they require commonly
 123 * modified blocks to be recowed, creating many dirty pages in the
 124 * extent tree an 4x-6x higher write load than ext3.
 125 *
 126 * Instead of doing a tree commit on every fsync, we use the
 127 * key ranges and transaction ids to find items for a given file or directory
 128 * that have changed in this transaction.  Those items are copied into
 129 * a special tree (one per subvolume root), that tree is written to disk
 130 * and then the fsync is considered complete.
 131 *
 132 * After a crash, items are copied out of the log-tree back into the
 133 * subvolume tree.  Any file data extents found are recorded in the extent
 134 * allocation tree, and the log-tree freed.
 135 *
 136 * The log tree is read three times, once to pin down all the extents it is
 137 * using in ram and once, once to create all the inodes logged in the tree
 138 * and once to do all the other items.
 139 */
 140
 141/*
 142 * start a sub transaction and setup the log tree
 143 * this increments the log tree writer count to make the people
 144 * syncing the tree wait for us to finish
 145 */
 146static int start_log_trans(struct btrfs_trans_handle *trans,
 147			   struct btrfs_root *root,
 148			   struct btrfs_log_ctx *ctx)
 149{
 150	struct btrfs_fs_info *fs_info = root->fs_info;
 151	struct btrfs_root *tree_root = fs_info->tree_root;
 152	const bool zoned = btrfs_is_zoned(fs_info);
 153	int ret = 0;
 154	bool created = false;
 155
 156	/*
 157	 * First check if the log root tree was already created. If not, create
 158	 * it before locking the root's log_mutex, just to keep lockdep happy.
 159	 */
 160	if (!test_bit(BTRFS_ROOT_HAS_LOG_TREE, &tree_root->state)) {
 161		mutex_lock(&tree_root->log_mutex);
 162		if (!fs_info->log_root_tree) {
 163			ret = btrfs_init_log_root_tree(trans, fs_info);
 164			if (!ret) {
 165				set_bit(BTRFS_ROOT_HAS_LOG_TREE, &tree_root->state);
 166				created = true;
 167			}
 168		}
 169		mutex_unlock(&tree_root->log_mutex);
 170		if (ret)
 171			return ret;
 172	}
 173
 174	mutex_lock(&root->log_mutex);
 175
 176again:
 177	if (root->log_root) {
 178		int index = (root->log_transid + 1) % 2;
 179
 180		if (btrfs_need_log_full_commit(trans)) {
 181			ret = BTRFS_LOG_FORCE_COMMIT;
 182			goto out;
 183		}
 184
 185		if (zoned && atomic_read(&root->log_commit[index])) {
 186			wait_log_commit(root, root->log_transid - 1);
 187			goto again;
 188		}
 189
 190		if (!root->log_start_pid) {
 191			clear_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state);
 192			root->log_start_pid = current->pid;
 193		} else if (root->log_start_pid != current->pid) {
 194			set_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state);
 195		}
 196	} else {
 197		/*
 198		 * This means fs_info->log_root_tree was already created
 199		 * for some other FS trees. Do the full commit not to mix
 200		 * nodes from multiple log transactions to do sequential
 201		 * writing.
 202		 */
 203		if (zoned && !created) {
 204			ret = BTRFS_LOG_FORCE_COMMIT;
 205			goto out;
 206		}
 207
 208		ret = btrfs_add_log_tree(trans, root);
 209		if (ret)
 210			goto out;
 211
 212		set_bit(BTRFS_ROOT_HAS_LOG_TREE, &root->state);
 213		clear_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state);
 214		root->log_start_pid = current->pid;
 215	}
 216
 217	atomic_inc(&root->log_writers);
 218	if (!ctx->logging_new_name) {
 219		int index = root->log_transid % 2;
 220		list_add_tail(&ctx->list, &root->log_ctxs[index]);
 221		ctx->log_transid = root->log_transid;
 222	}
 223
 224out:
 225	mutex_unlock(&root->log_mutex);
 226	return ret;
 227}
 228
 229/*
 230 * returns 0 if there was a log transaction running and we were able
 231 * to join, or returns -ENOENT if there were not transactions
 232 * in progress
 233 */
 234static int join_running_log_trans(struct btrfs_root *root)
 235{
 236	const bool zoned = btrfs_is_zoned(root->fs_info);
 237	int ret = -ENOENT;
 238
 239	if (!test_bit(BTRFS_ROOT_HAS_LOG_TREE, &root->state))
 240		return ret;
 241
 242	mutex_lock(&root->log_mutex);
 243again:
 244	if (root->log_root) {
 245		int index = (root->log_transid + 1) % 2;
 246
 247		ret = 0;
 248		if (zoned && atomic_read(&root->log_commit[index])) {
 249			wait_log_commit(root, root->log_transid - 1);
 250			goto again;
 251		}
 252		atomic_inc(&root->log_writers);
 253	}
 254	mutex_unlock(&root->log_mutex);
 255	return ret;
 256}
 257
 258/*
 259 * This either makes the current running log transaction wait
 260 * until you call btrfs_end_log_trans() or it makes any future
 261 * log transactions wait until you call btrfs_end_log_trans()
 262 */
 263void btrfs_pin_log_trans(struct btrfs_root *root)
 264{
 265	atomic_inc(&root->log_writers);
 266}
 267
 268/*
 269 * indicate we're done making changes to the log tree
 270 * and wake up anyone waiting to do a sync
 271 */
 272void btrfs_end_log_trans(struct btrfs_root *root)
 273{
 274	if (atomic_dec_and_test(&root->log_writers)) {
 275		/* atomic_dec_and_test implies a barrier */
 276		cond_wake_up_nomb(&root->log_writer_wait);
 277	}
 278}
 279
 280/*
 281 * the walk control struct is used to pass state down the chain when
 282 * processing the log tree.  The stage field tells us which part
 283 * of the log tree processing we are currently doing.  The others
 284 * are state fields used for that specific part
 285 */
 286struct walk_control {
 287	/* should we free the extent on disk when done?  This is used
 288	 * at transaction commit time while freeing a log tree
 289	 */
 290	int free;
 291
 292	/* pin only walk, we record which extents on disk belong to the
 293	 * log trees
 294	 */
 295	int pin;
 296
 297	/* what stage of the replay code we're currently in */
 298	int stage;
 299
 300	/*
 301	 * Ignore any items from the inode currently being processed. Needs
 302	 * to be set every time we find a BTRFS_INODE_ITEM_KEY and we are in
 303	 * the LOG_WALK_REPLAY_INODES stage.
 304	 */
 305	bool ignore_cur_inode;
 306
 307	/* the root we are currently replaying */
 308	struct btrfs_root *replay_dest;
 309
 310	/* the trans handle for the current replay */
 311	struct btrfs_trans_handle *trans;
 312
 313	/* the function that gets used to process blocks we find in the
 314	 * tree.  Note the extent_buffer might not be up to date when it is
 315	 * passed in, and it must be checked or read if you need the data
 316	 * inside it
 317	 */
 318	int (*process_func)(struct btrfs_root *log, struct extent_buffer *eb,
 319			    struct walk_control *wc, u64 gen, int level);
 320};
 321
 322/*
 323 * process_func used to pin down extents, write them or wait on them
 324 */
 325static int process_one_buffer(struct btrfs_root *log,
 326			      struct extent_buffer *eb,
 327			      struct walk_control *wc, u64 gen, int level)
 328{
 329	struct btrfs_fs_info *fs_info = log->fs_info;
 330	int ret = 0;
 331
 332	/*
 333	 * If this fs is mixed then we need to be able to process the leaves to
 334	 * pin down any logged extents, so we have to read the block.
 335	 */
 336	if (btrfs_fs_incompat(fs_info, MIXED_GROUPS)) {
 337		struct btrfs_tree_parent_check check = {
 338			.level = level,
 339			.transid = gen
 340		};
 341
 342		ret = btrfs_read_extent_buffer(eb, &check);
 343		if (ret)
 344			return ret;
 345	}
 346
 347	if (wc->pin) {
 348		ret = btrfs_pin_extent_for_log_replay(wc->trans, eb);
 349		if (ret)
 350			return ret;
 351
 352		if (btrfs_buffer_uptodate(eb, gen, 0) &&
 353		    btrfs_header_level(eb) == 0)
 354			ret = btrfs_exclude_logged_extents(eb);
 355	}
 356	return ret;
 357}
 358
 359/*
 360 * Item overwrite used by replay and tree logging.  eb, slot and key all refer
 361 * to the src data we are copying out.
 362 *
 363 * root is the tree we are copying into, and path is a scratch
 364 * path for use in this function (it should be released on entry and
 365 * will be released on exit).
 366 *
 367 * If the key is already in the destination tree the existing item is
 368 * overwritten.  If the existing item isn't big enough, it is extended.
 369 * If it is too large, it is truncated.
 370 *
 371 * If the key isn't in the destination yet, a new item is inserted.
 372 */
 373static int overwrite_item(struct btrfs_trans_handle *trans,
 374			  struct btrfs_root *root,
 375			  struct btrfs_path *path,
 376			  struct extent_buffer *eb, int slot,
 377			  struct btrfs_key *key)
 378{
 379	int ret;
 380	u32 item_size;
 381	u64 saved_i_size = 0;
 382	int save_old_i_size = 0;
 383	unsigned long src_ptr;
 384	unsigned long dst_ptr;
 385	bool inode_item = key->type == BTRFS_INODE_ITEM_KEY;
 386
 387	/*
 388	 * This is only used during log replay, so the root is always from a
 389	 * fs/subvolume tree. In case we ever need to support a log root, then
 390	 * we'll have to clone the leaf in the path, release the path and use
 391	 * the leaf before writing into the log tree. See the comments at
 392	 * copy_items() for more details.
 393	 */
 394	ASSERT(root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID);
 395
 396	item_size = btrfs_item_size(eb, slot);
 397	src_ptr = btrfs_item_ptr_offset(eb, slot);
 398
 399	/* Look for the key in the destination tree. */
 400	ret = btrfs_search_slot(NULL, root, key, path, 0, 0);
 401	if (ret < 0)
 402		return ret;
 403
 404	if (ret == 0) {
 405		char *src_copy;
 406		char *dst_copy;
 407		u32 dst_size = btrfs_item_size(path->nodes[0],
 408						  path->slots[0]);
 409		if (dst_size != item_size)
 410			goto insert;
 411
 412		if (item_size == 0) {
 413			btrfs_release_path(path);
 414			return 0;
 415		}
 416		dst_copy = kmalloc(item_size, GFP_NOFS);
 417		src_copy = kmalloc(item_size, GFP_NOFS);
 418		if (!dst_copy || !src_copy) {
 419			btrfs_release_path(path);
 420			kfree(dst_copy);
 421			kfree(src_copy);
 422			return -ENOMEM;
 423		}
 424
 425		read_extent_buffer(eb, src_copy, src_ptr, item_size);
 426
 427		dst_ptr = btrfs_item_ptr_offset(path->nodes[0], path->slots[0]);
 428		read_extent_buffer(path->nodes[0], dst_copy, dst_ptr,
 429				   item_size);
 430		ret = memcmp(dst_copy, src_copy, item_size);
 431
 432		kfree(dst_copy);
 433		kfree(src_copy);
 434		/*
 435		 * they have the same contents, just return, this saves
 436		 * us from cowing blocks in the destination tree and doing
 437		 * extra writes that may not have been done by a previous
 438		 * sync
 439		 */
 440		if (ret == 0) {
 441			btrfs_release_path(path);
 442			return 0;
 443		}
 444
 445		/*
 446		 * We need to load the old nbytes into the inode so when we
 447		 * replay the extents we've logged we get the right nbytes.
 448		 */
 449		if (inode_item) {
 450			struct btrfs_inode_item *item;
 451			u64 nbytes;
 452			u32 mode;
 453
 454			item = btrfs_item_ptr(path->nodes[0], path->slots[0],
 455					      struct btrfs_inode_item);
 456			nbytes = btrfs_inode_nbytes(path->nodes[0], item);
 457			item = btrfs_item_ptr(eb, slot,
 458					      struct btrfs_inode_item);
 459			btrfs_set_inode_nbytes(eb, item, nbytes);
 460
 461			/*
 462			 * If this is a directory we need to reset the i_size to
 463			 * 0 so that we can set it up properly when replaying
 464			 * the rest of the items in this log.
 465			 */
 466			mode = btrfs_inode_mode(eb, item);
 467			if (S_ISDIR(mode))
 468				btrfs_set_inode_size(eb, item, 0);
 469		}
 470	} else if (inode_item) {
 471		struct btrfs_inode_item *item;
 472		u32 mode;
 473
 474		/*
 475		 * New inode, set nbytes to 0 so that the nbytes comes out
 476		 * properly when we replay the extents.
 477		 */
 478		item = btrfs_item_ptr(eb, slot, struct btrfs_inode_item);
 479		btrfs_set_inode_nbytes(eb, item, 0);
 480
 481		/*
 482		 * If this is a directory we need to reset the i_size to 0 so
 483		 * that we can set it up properly when replaying the rest of
 484		 * the items in this log.
 485		 */
 486		mode = btrfs_inode_mode(eb, item);
 487		if (S_ISDIR(mode))
 488			btrfs_set_inode_size(eb, item, 0);
 489	}
 490insert:
 491	btrfs_release_path(path);
 492	/* try to insert the key into the destination tree */
 493	path->skip_release_on_error = 1;
 494	ret = btrfs_insert_empty_item(trans, root, path,
 495				      key, item_size);
 496	path->skip_release_on_error = 0;
 497
 498	/* make sure any existing item is the correct size */
 499	if (ret == -EEXIST || ret == -EOVERFLOW) {
 500		u32 found_size;
 501		found_size = btrfs_item_size(path->nodes[0],
 502						path->slots[0]);
 503		if (found_size > item_size)
 504			btrfs_truncate_item(trans, path, item_size, 1);
 505		else if (found_size < item_size)
 506			btrfs_extend_item(trans, path, item_size - found_size);
 507	} else if (ret) {
 508		return ret;
 509	}
 510	dst_ptr = btrfs_item_ptr_offset(path->nodes[0],
 511					path->slots[0]);
 512
 513	/* don't overwrite an existing inode if the generation number
 514	 * was logged as zero.  This is done when the tree logging code
 515	 * is just logging an inode to make sure it exists after recovery.
 516	 *
 517	 * Also, don't overwrite i_size on directories during replay.
 518	 * log replay inserts and removes directory items based on the
 519	 * state of the tree found in the subvolume, and i_size is modified
 520	 * as it goes
 521	 */
 522	if (key->type == BTRFS_INODE_ITEM_KEY && ret == -EEXIST) {
 523		struct btrfs_inode_item *src_item;
 524		struct btrfs_inode_item *dst_item;
 525
 526		src_item = (struct btrfs_inode_item *)src_ptr;
 527		dst_item = (struct btrfs_inode_item *)dst_ptr;
 528
 529		if (btrfs_inode_generation(eb, src_item) == 0) {
 530			struct extent_buffer *dst_eb = path->nodes[0];
 531			const u64 ino_size = btrfs_inode_size(eb, src_item);
 532
 533			/*
 534			 * For regular files an ino_size == 0 is used only when
 535			 * logging that an inode exists, as part of a directory
 536			 * fsync, and the inode wasn't fsynced before. In this
 537			 * case don't set the size of the inode in the fs/subvol
 538			 * tree, otherwise we would be throwing valid data away.
 539			 */
 540			if (S_ISREG(btrfs_inode_mode(eb, src_item)) &&
 541			    S_ISREG(btrfs_inode_mode(dst_eb, dst_item)) &&
 542			    ino_size != 0)
 543				btrfs_set_inode_size(dst_eb, dst_item, ino_size);
 544			goto no_copy;
 545		}
 546
 547		if (S_ISDIR(btrfs_inode_mode(eb, src_item)) &&
 548		    S_ISDIR(btrfs_inode_mode(path->nodes[0], dst_item))) {
 549			save_old_i_size = 1;
 550			saved_i_size = btrfs_inode_size(path->nodes[0],
 551							dst_item);
 552		}
 553	}
 554
 555	copy_extent_buffer(path->nodes[0], eb, dst_ptr,
 556			   src_ptr, item_size);
 557
 558	if (save_old_i_size) {
 559		struct btrfs_inode_item *dst_item;
 560		dst_item = (struct btrfs_inode_item *)dst_ptr;
 561		btrfs_set_inode_size(path->nodes[0], dst_item, saved_i_size);
 562	}
 563
 564	/* make sure the generation is filled in */
 565	if (key->type == BTRFS_INODE_ITEM_KEY) {
 566		struct btrfs_inode_item *dst_item;
 567		dst_item = (struct btrfs_inode_item *)dst_ptr;
 568		if (btrfs_inode_generation(path->nodes[0], dst_item) == 0) {
 569			btrfs_set_inode_generation(path->nodes[0], dst_item,
 570						   trans->transid);
 571		}
 572	}
 573no_copy:
 574	btrfs_mark_buffer_dirty(trans, path->nodes[0]);
 575	btrfs_release_path(path);
 576	return 0;
 577}
 578
 579static int read_alloc_one_name(struct extent_buffer *eb, void *start, int len,
 580			       struct fscrypt_str *name)
 581{
 582	char *buf;
 583
 584	buf = kmalloc(len, GFP_NOFS);
 585	if (!buf)
 586		return -ENOMEM;
 587
 588	read_extent_buffer(eb, buf, (unsigned long)start, len);
 589	name->name = buf;
 590	name->len = len;
 591	return 0;
 592}
 593
 594/*
 595 * simple helper to read an inode off the disk from a given root
 596 * This can only be called for subvolume roots and not for the log
 597 */
 598static noinline struct inode *read_one_inode(struct btrfs_root *root,
 599					     u64 objectid)
 600{
 601	struct inode *inode;
 602
 603	inode = btrfs_iget(root->fs_info->sb, objectid, root);
 604	if (IS_ERR(inode))
 605		inode = NULL;
 606	return inode;
 607}
 608
 609/* replays a single extent in 'eb' at 'slot' with 'key' into the
 610 * subvolume 'root'.  path is released on entry and should be released
 611 * on exit.
 612 *
 613 * extents in the log tree have not been allocated out of the extent
 614 * tree yet.  So, this completes the allocation, taking a reference
 615 * as required if the extent already exists or creating a new extent
 616 * if it isn't in the extent allocation tree yet.
 617 *
 618 * The extent is inserted into the file, dropping any existing extents
 619 * from the file that overlap the new one.
 620 */
 621static noinline int replay_one_extent(struct btrfs_trans_handle *trans,
 622				      struct btrfs_root *root,
 623				      struct btrfs_path *path,
 624				      struct extent_buffer *eb, int slot,
 625				      struct btrfs_key *key)
 626{
 627	struct btrfs_drop_extents_args drop_args = { 0 };
 628	struct btrfs_fs_info *fs_info = root->fs_info;
 629	int found_type;
 630	u64 extent_end;
 631	u64 start = key->offset;
 632	u64 nbytes = 0;
 633	struct btrfs_file_extent_item *item;
 634	struct inode *inode = NULL;
 635	unsigned long size;
 636	int ret = 0;
 637
 638	item = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item);
 639	found_type = btrfs_file_extent_type(eb, item);
 640
 641	if (found_type == BTRFS_FILE_EXTENT_REG ||
 642	    found_type == BTRFS_FILE_EXTENT_PREALLOC) {
 643		nbytes = btrfs_file_extent_num_bytes(eb, item);
 644		extent_end = start + nbytes;
 645
 646		/*
 647		 * We don't add to the inodes nbytes if we are prealloc or a
 648		 * hole.
 649		 */
 650		if (btrfs_file_extent_disk_bytenr(eb, item) == 0)
 651			nbytes = 0;
 652	} else if (found_type == BTRFS_FILE_EXTENT_INLINE) {
 653		size = btrfs_file_extent_ram_bytes(eb, item);
 654		nbytes = btrfs_file_extent_ram_bytes(eb, item);
 655		extent_end = ALIGN(start + size,
 656				   fs_info->sectorsize);
 657	} else {
 658		ret = 0;
 659		goto out;
 660	}
 661
 662	inode = read_one_inode(root, key->objectid);
 663	if (!inode) {
 664		ret = -EIO;
 665		goto out;
 666	}
 667
 668	/*
 669	 * first check to see if we already have this extent in the
 670	 * file.  This must be done before the btrfs_drop_extents run
 671	 * so we don't try to drop this extent.
 672	 */
 673	ret = btrfs_lookup_file_extent(trans, root, path,
 674			btrfs_ino(BTRFS_I(inode)), start, 0);
 675
 676	if (ret == 0 &&
 677	    (found_type == BTRFS_FILE_EXTENT_REG ||
 678	     found_type == BTRFS_FILE_EXTENT_PREALLOC)) {
 679		struct btrfs_file_extent_item cmp1;
 680		struct btrfs_file_extent_item cmp2;
 681		struct btrfs_file_extent_item *existing;
 682		struct extent_buffer *leaf;
 683
 684		leaf = path->nodes[0];
 685		existing = btrfs_item_ptr(leaf, path->slots[0],
 686					  struct btrfs_file_extent_item);
 687
 688		read_extent_buffer(eb, &cmp1, (unsigned long)item,
 689				   sizeof(cmp1));
 690		read_extent_buffer(leaf, &cmp2, (unsigned long)existing,
 691				   sizeof(cmp2));
 692
 693		/*
 694		 * we already have a pointer to this exact extent,
 695		 * we don't have to do anything
 696		 */
 697		if (memcmp(&cmp1, &cmp2, sizeof(cmp1)) == 0) {
 698			btrfs_release_path(path);
 699			goto out;
 700		}
 701	}
 702	btrfs_release_path(path);
 703
 704	/* drop any overlapping extents */
 705	drop_args.start = start;
 706	drop_args.end = extent_end;
 707	drop_args.drop_cache = true;
 708	ret = btrfs_drop_extents(trans, root, BTRFS_I(inode), &drop_args);
 709	if (ret)
 710		goto out;
 711
 712	if (found_type == BTRFS_FILE_EXTENT_REG ||
 713	    found_type == BTRFS_FILE_EXTENT_PREALLOC) {
 714		u64 offset;
 715		unsigned long dest_offset;
 716		struct btrfs_key ins;
 717
 718		if (btrfs_file_extent_disk_bytenr(eb, item) == 0 &&
 719		    btrfs_fs_incompat(fs_info, NO_HOLES))
 720			goto update_inode;
 721
 722		ret = btrfs_insert_empty_item(trans, root, path, key,
 723					      sizeof(*item));
 724		if (ret)
 725			goto out;
 726		dest_offset = btrfs_item_ptr_offset(path->nodes[0],
 727						    path->slots[0]);
 728		copy_extent_buffer(path->nodes[0], eb, dest_offset,
 729				(unsigned long)item,  sizeof(*item));
 730
 731		ins.objectid = btrfs_file_extent_disk_bytenr(eb, item);
 732		ins.offset = btrfs_file_extent_disk_num_bytes(eb, item);
 733		ins.type = BTRFS_EXTENT_ITEM_KEY;
 734		offset = key->offset - btrfs_file_extent_offset(eb, item);
 735
 736		/*
 737		 * Manually record dirty extent, as here we did a shallow
 738		 * file extent item copy and skip normal backref update,
 739		 * but modifying extent tree all by ourselves.
 740		 * So need to manually record dirty extent for qgroup,
 741		 * as the owner of the file extent changed from log tree
 742		 * (doesn't affect qgroup) to fs/file tree(affects qgroup)
 743		 */
 744		ret = btrfs_qgroup_trace_extent(trans,
 745				btrfs_file_extent_disk_bytenr(eb, item),
 746				btrfs_file_extent_disk_num_bytes(eb, item));
 747		if (ret < 0)
 748			goto out;
 749
 750		if (ins.objectid > 0) {
 751			struct btrfs_ref ref = { 0 };
 752			u64 csum_start;
 753			u64 csum_end;
 754			LIST_HEAD(ordered_sums);
 755
 756			/*
 757			 * is this extent already allocated in the extent
 758			 * allocation tree?  If so, just add a reference
 759			 */
 760			ret = btrfs_lookup_data_extent(fs_info, ins.objectid,
 761						ins.offset);
 762			if (ret < 0) {
 763				goto out;
 764			} else if (ret == 0) {
 765				btrfs_init_generic_ref(&ref,
 766						BTRFS_ADD_DELAYED_REF,
 767						ins.objectid, ins.offset, 0,
 768						root->root_key.objectid);
 769				btrfs_init_data_ref(&ref,
 770						root->root_key.objectid,
 771						key->objectid, offset, 0, false);
 772				ret = btrfs_inc_extent_ref(trans, &ref);
 773				if (ret)
 774					goto out;
 775			} else {
 776				/*
 777				 * insert the extent pointer in the extent
 778				 * allocation tree
 779				 */
 780				ret = btrfs_alloc_logged_file_extent(trans,
 781						root->root_key.objectid,
 782						key->objectid, offset, &ins);
 783				if (ret)
 784					goto out;
 785			}
 786			btrfs_release_path(path);
 787
 788			if (btrfs_file_extent_compression(eb, item)) {
 789				csum_start = ins.objectid;
 790				csum_end = csum_start + ins.offset;
 791			} else {
 792				csum_start = ins.objectid +
 793					btrfs_file_extent_offset(eb, item);
 794				csum_end = csum_start +
 795					btrfs_file_extent_num_bytes(eb, item);
 796			}
 797
 798			ret = btrfs_lookup_csums_list(root->log_root,
 799						csum_start, csum_end - 1,
 800						&ordered_sums, 0, false);
 801			if (ret)
 802				goto out;
 803			/*
 804			 * Now delete all existing cums in the csum root that
 805			 * cover our range. We do this because we can have an
 806			 * extent that is completely referenced by one file
 807			 * extent item and partially referenced by another
 808			 * file extent item (like after using the clone or
 809			 * extent_same ioctls). In this case if we end up doing
 810			 * the replay of the one that partially references the
 811			 * extent first, and we do not do the csum deletion
 812			 * below, we can get 2 csum items in the csum tree that
 813			 * overlap each other. For example, imagine our log has
 814			 * the two following file extent items:
 815			 *
 816			 * key (257 EXTENT_DATA 409600)
 817			 *     extent data disk byte 12845056 nr 102400
 818			 *     extent data offset 20480 nr 20480 ram 102400
 819			 *
 820			 * key (257 EXTENT_DATA 819200)
 821			 *     extent data disk byte 12845056 nr 102400
 822			 *     extent data offset 0 nr 102400 ram 102400
 823			 *
 824			 * Where the second one fully references the 100K extent
 825			 * that starts at disk byte 12845056, and the log tree
 826			 * has a single csum item that covers the entire range
 827			 * of the extent:
 828			 *
 829			 * key (EXTENT_CSUM EXTENT_CSUM 12845056) itemsize 100
 830			 *
 831			 * After the first file extent item is replayed, the
 832			 * csum tree gets the following csum item:
 833			 *
 834			 * key (EXTENT_CSUM EXTENT_CSUM 12865536) itemsize 20
 835			 *
 836			 * Which covers the 20K sub-range starting at offset 20K
 837			 * of our extent. Now when we replay the second file
 838			 * extent item, if we do not delete existing csum items
 839			 * that cover any of its blocks, we end up getting two
 840			 * csum items in our csum tree that overlap each other:
 841			 *
 842			 * key (EXTENT_CSUM EXTENT_CSUM 12845056) itemsize 100
 843			 * key (EXTENT_CSUM EXTENT_CSUM 12865536) itemsize 20
 844			 *
 845			 * Which is a problem, because after this anyone trying
 846			 * to lookup up for the checksum of any block of our
 847			 * extent starting at an offset of 40K or higher, will
 848			 * end up looking at the second csum item only, which
 849			 * does not contain the checksum for any block starting
 850			 * at offset 40K or higher of our extent.
 851			 */
 852			while (!list_empty(&ordered_sums)) {
 853				struct btrfs_ordered_sum *sums;
 854				struct btrfs_root *csum_root;
 855
 856				sums = list_entry(ordered_sums.next,
 857						struct btrfs_ordered_sum,
 858						list);
 859				csum_root = btrfs_csum_root(fs_info,
 860							    sums->logical);
 861				if (!ret)
 862					ret = btrfs_del_csums(trans, csum_root,
 863							      sums->logical,
 864							      sums->len);
 865				if (!ret)
 866					ret = btrfs_csum_file_blocks(trans,
 867								     csum_root,
 868								     sums);
 869				list_del(&sums->list);
 870				kfree(sums);
 871			}
 872			if (ret)
 873				goto out;
 874		} else {
 875			btrfs_release_path(path);
 876		}
 877	} else if (found_type == BTRFS_FILE_EXTENT_INLINE) {
 878		/* inline extents are easy, we just overwrite them */
 879		ret = overwrite_item(trans, root, path, eb, slot, key);
 880		if (ret)
 881			goto out;
 882	}
 883
 884	ret = btrfs_inode_set_file_extent_range(BTRFS_I(inode), start,
 885						extent_end - start);
 886	if (ret)
 887		goto out;
 888
 889update_inode:
 890	btrfs_update_inode_bytes(BTRFS_I(inode), nbytes, drop_args.bytes_found);
 891	ret = btrfs_update_inode(trans, BTRFS_I(inode));
 892out:
 893	iput(inode);
 894	return ret;
 895}
 896
 897static int unlink_inode_for_log_replay(struct btrfs_trans_handle *trans,
 898				       struct btrfs_inode *dir,
 899				       struct btrfs_inode *inode,
 900				       const struct fscrypt_str *name)
 901{
 902	int ret;
 903
 904	ret = btrfs_unlink_inode(trans, dir, inode, name);
 905	if (ret)
 906		return ret;
 907	/*
 908	 * Whenever we need to check if a name exists or not, we check the
 909	 * fs/subvolume tree. So after an unlink we must run delayed items, so
 910	 * that future checks for a name during log replay see that the name
 911	 * does not exists anymore.
 912	 */
 913	return btrfs_run_delayed_items(trans);
 914}
 915
 916/*
 917 * when cleaning up conflicts between the directory names in the
 918 * subvolume, directory names in the log and directory names in the
 919 * inode back references, we may have to unlink inodes from directories.
 920 *
 921 * This is a helper function to do the unlink of a specific directory
 922 * item
 923 */
 924static noinline int drop_one_dir_item(struct btrfs_trans_handle *trans,
 925				      struct btrfs_path *path,
 926				      struct btrfs_inode *dir,
 927				      struct btrfs_dir_item *di)
 928{
 929	struct btrfs_root *root = dir->root;
 930	struct inode *inode;
 931	struct fscrypt_str name;
 932	struct extent_buffer *leaf;
 933	struct btrfs_key location;
 934	int ret;
 935
 936	leaf = path->nodes[0];
 937
 938	btrfs_dir_item_key_to_cpu(leaf, di, &location);
 939	ret = read_alloc_one_name(leaf, di + 1, btrfs_dir_name_len(leaf, di), &name);
 940	if (ret)
 941		return -ENOMEM;
 942
 943	btrfs_release_path(path);
 944
 945	inode = read_one_inode(root, location.objectid);
 946	if (!inode) {
 947		ret = -EIO;
 948		goto out;
 949	}
 950
 951	ret = link_to_fixup_dir(trans, root, path, location.objectid);
 952	if (ret)
 953		goto out;
 954
 955	ret = unlink_inode_for_log_replay(trans, dir, BTRFS_I(inode), &name);
 956out:
 957	kfree(name.name);
 958	iput(inode);
 959	return ret;
 960}
 961
 962/*
 963 * See if a given name and sequence number found in an inode back reference are
 964 * already in a directory and correctly point to this inode.
 965 *
 966 * Returns: < 0 on error, 0 if the directory entry does not exists and 1 if it
 967 * exists.
 968 */
 969static noinline int inode_in_dir(struct btrfs_root *root,
 970				 struct btrfs_path *path,
 971				 u64 dirid, u64 objectid, u64 index,
 972				 struct fscrypt_str *name)
 973{
 974	struct btrfs_dir_item *di;
 975	struct btrfs_key location;
 976	int ret = 0;
 977
 978	di = btrfs_lookup_dir_index_item(NULL, root, path, dirid,
 979					 index, name, 0);
 980	if (IS_ERR(di)) {
 981		ret = PTR_ERR(di);
 982		goto out;
 983	} else if (di) {
 984		btrfs_dir_item_key_to_cpu(path->nodes[0], di, &location);
 985		if (location.objectid != objectid)
 986			goto out;
 987	} else {
 988		goto out;
 989	}
 990
 991	btrfs_release_path(path);
 992	di = btrfs_lookup_dir_item(NULL, root, path, dirid, name, 0);
 993	if (IS_ERR(di)) {
 994		ret = PTR_ERR(di);
 995		goto out;
 996	} else if (di) {
 997		btrfs_dir_item_key_to_cpu(path->nodes[0], di, &location);
 998		if (location.objectid == objectid)
 999			ret = 1;
1000	}
1001out:
1002	btrfs_release_path(path);
1003	return ret;
1004}
1005
1006/*
1007 * helper function to check a log tree for a named back reference in
1008 * an inode.  This is used to decide if a back reference that is
1009 * found in the subvolume conflicts with what we find in the log.
1010 *
1011 * inode backreferences may have multiple refs in a single item,
1012 * during replay we process one reference at a time, and we don't
1013 * want to delete valid links to a file from the subvolume if that
1014 * link is also in the log.
1015 */
1016static noinline int backref_in_log(struct btrfs_root *log,
1017				   struct btrfs_key *key,
1018				   u64 ref_objectid,
1019				   const struct fscrypt_str *name)
1020{
1021	struct btrfs_path *path;
1022	int ret;
1023
1024	path = btrfs_alloc_path();
1025	if (!path)
1026		return -ENOMEM;
1027
1028	ret = btrfs_search_slot(NULL, log, key, path, 0, 0);
1029	if (ret < 0) {
1030		goto out;
1031	} else if (ret == 1) {
1032		ret = 0;
1033		goto out;
1034	}
1035
1036	if (key->type == BTRFS_INODE_EXTREF_KEY)
1037		ret = !!btrfs_find_name_in_ext_backref(path->nodes[0],
1038						       path->slots[0],
1039						       ref_objectid, name);
1040	else
1041		ret = !!btrfs_find_name_in_backref(path->nodes[0],
1042						   path->slots[0], name);
1043out:
1044	btrfs_free_path(path);
1045	return ret;
1046}
1047
1048static inline int __add_inode_ref(struct btrfs_trans_handle *trans,
1049				  struct btrfs_root *root,
1050				  struct btrfs_path *path,
1051				  struct btrfs_root *log_root,
1052				  struct btrfs_inode *dir,
1053				  struct btrfs_inode *inode,
1054				  u64 inode_objectid, u64 parent_objectid,
1055				  u64 ref_index, struct fscrypt_str *name)
1056{
1057	int ret;
1058	struct extent_buffer *leaf;
1059	struct btrfs_dir_item *di;
1060	struct btrfs_key search_key;
1061	struct btrfs_inode_extref *extref;
1062
1063again:
1064	/* Search old style refs */
1065	search_key.objectid = inode_objectid;
1066	search_key.type = BTRFS_INODE_REF_KEY;
1067	search_key.offset = parent_objectid;
1068	ret = btrfs_search_slot(NULL, root, &search_key, path, 0, 0);
1069	if (ret == 0) {
1070		struct btrfs_inode_ref *victim_ref;
1071		unsigned long ptr;
1072		unsigned long ptr_end;
1073
1074		leaf = path->nodes[0];
1075
1076		/* are we trying to overwrite a back ref for the root directory
1077		 * if so, just jump out, we're done
1078		 */
1079		if (search_key.objectid == search_key.offset)
1080			return 1;
1081
1082		/* check all the names in this back reference to see
1083		 * if they are in the log.  if so, we allow them to stay
1084		 * otherwise they must be unlinked as a conflict
1085		 */
1086		ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
1087		ptr_end = ptr + btrfs_item_size(leaf, path->slots[0]);
1088		while (ptr < ptr_end) {
1089			struct fscrypt_str victim_name;
1090
1091			victim_ref = (struct btrfs_inode_ref *)ptr;
1092			ret = read_alloc_one_name(leaf, (victim_ref + 1),
1093				 btrfs_inode_ref_name_len(leaf, victim_ref),
1094				 &victim_name);
1095			if (ret)
1096				return ret;
1097
1098			ret = backref_in_log(log_root, &search_key,
1099					     parent_objectid, &victim_name);
1100			if (ret < 0) {
1101				kfree(victim_name.name);
1102				return ret;
1103			} else if (!ret) {
1104				inc_nlink(&inode->vfs_inode);
1105				btrfs_release_path(path);
1106
1107				ret = unlink_inode_for_log_replay(trans, dir, inode,
1108						&victim_name);
1109				kfree(victim_name.name);
1110				if (ret)
1111					return ret;
1112				goto again;
1113			}
1114			kfree(victim_name.name);
1115
1116			ptr = (unsigned long)(victim_ref + 1) + victim_name.len;
1117		}
1118	}
1119	btrfs_release_path(path);
1120
1121	/* Same search but for extended refs */
1122	extref = btrfs_lookup_inode_extref(NULL, root, path, name,
1123					   inode_objectid, parent_objectid, 0,
1124					   0);
1125	if (IS_ERR(extref)) {
1126		return PTR_ERR(extref);
1127	} else if (extref) {
1128		u32 item_size;
1129		u32 cur_offset = 0;
1130		unsigned long base;
1131		struct inode *victim_parent;
1132
1133		leaf = path->nodes[0];
1134
1135		item_size = btrfs_item_size(leaf, path->slots[0]);
1136		base = btrfs_item_ptr_offset(leaf, path->slots[0]);
1137
1138		while (cur_offset < item_size) {
1139			struct fscrypt_str victim_name;
1140
1141			extref = (struct btrfs_inode_extref *)(base + cur_offset);
1142
1143			if (btrfs_inode_extref_parent(leaf, extref) != parent_objectid)
1144				goto next;
1145
1146			ret = read_alloc_one_name(leaf, &extref->name,
1147				 btrfs_inode_extref_name_len(leaf, extref),
1148				 &victim_name);
1149			if (ret)
1150				return ret;
1151
1152			search_key.objectid = inode_objectid;
1153			search_key.type = BTRFS_INODE_EXTREF_KEY;
1154			search_key.offset = btrfs_extref_hash(parent_objectid,
1155							      victim_name.name,
1156							      victim_name.len);
1157			ret = backref_in_log(log_root, &search_key,
1158					     parent_objectid, &victim_name);
1159			if (ret < 0) {
1160				kfree(victim_name.name);
1161				return ret;
1162			} else if (!ret) {
1163				ret = -ENOENT;
1164				victim_parent = read_one_inode(root,
1165						parent_objectid);
1166				if (victim_parent) {
1167					inc_nlink(&inode->vfs_inode);
1168					btrfs_release_path(path);
1169
1170					ret = unlink_inode_for_log_replay(trans,
1171							BTRFS_I(victim_parent),
1172							inode, &victim_name);
1173				}
1174				iput(victim_parent);
1175				kfree(victim_name.name);
1176				if (ret)
1177					return ret;
1178				goto again;
1179			}
1180			kfree(victim_name.name);
1181next:
1182			cur_offset += victim_name.len + sizeof(*extref);
1183		}
1184	}
1185	btrfs_release_path(path);
1186
1187	/* look for a conflicting sequence number */
1188	di = btrfs_lookup_dir_index_item(trans, root, path, btrfs_ino(dir),
1189					 ref_index, name, 0);
1190	if (IS_ERR(di)) {
1191		return PTR_ERR(di);
1192	} else if (di) {
1193		ret = drop_one_dir_item(trans, path, dir, di);
1194		if (ret)
1195			return ret;
1196	}
1197	btrfs_release_path(path);
1198
1199	/* look for a conflicting name */
1200	di = btrfs_lookup_dir_item(trans, root, path, btrfs_ino(dir), name, 0);
1201	if (IS_ERR(di)) {
1202		return PTR_ERR(di);
1203	} else if (di) {
1204		ret = drop_one_dir_item(trans, path, dir, di);
1205		if (ret)
1206			return ret;
1207	}
1208	btrfs_release_path(path);
1209
1210	return 0;
1211}
1212
1213static int extref_get_fields(struct extent_buffer *eb, unsigned long ref_ptr,
1214			     struct fscrypt_str *name, u64 *index,
1215			     u64 *parent_objectid)
1216{
1217	struct btrfs_inode_extref *extref;
1218	int ret;
1219
1220	extref = (struct btrfs_inode_extref *)ref_ptr;
1221
1222	ret = read_alloc_one_name(eb, &extref->name,
1223				  btrfs_inode_extref_name_len(eb, extref), name);
1224	if (ret)
1225		return ret;
1226
1227	if (index)
1228		*index = btrfs_inode_extref_index(eb, extref);
1229	if (parent_objectid)
1230		*parent_objectid = btrfs_inode_extref_parent(eb, extref);
1231
1232	return 0;
1233}
1234
1235static int ref_get_fields(struct extent_buffer *eb, unsigned long ref_ptr,
1236			  struct fscrypt_str *name, u64 *index)
1237{
1238	struct btrfs_inode_ref *ref;
1239	int ret;
1240
1241	ref = (struct btrfs_inode_ref *)ref_ptr;
1242
1243	ret = read_alloc_one_name(eb, ref + 1, btrfs_inode_ref_name_len(eb, ref),
1244				  name);
1245	if (ret)
1246		return ret;
1247
1248	if (index)
1249		*index = btrfs_inode_ref_index(eb, ref);
1250
1251	return 0;
1252}
1253
1254/*
1255 * Take an inode reference item from the log tree and iterate all names from the
1256 * inode reference item in the subvolume tree with the same key (if it exists).
1257 * For any name that is not in the inode reference item from the log tree, do a
1258 * proper unlink of that name (that is, remove its entry from the inode
1259 * reference item and both dir index keys).
1260 */
1261static int unlink_old_inode_refs(struct btrfs_trans_handle *trans,
1262				 struct btrfs_root *root,
1263				 struct btrfs_path *path,
1264				 struct btrfs_inode *inode,
1265				 struct extent_buffer *log_eb,
1266				 int log_slot,
1267				 struct btrfs_key *key)
1268{
1269	int ret;
1270	unsigned long ref_ptr;
1271	unsigned long ref_end;
1272	struct extent_buffer *eb;
1273
1274again:
1275	btrfs_release_path(path);
1276	ret = btrfs_search_slot(NULL, root, key, path, 0, 0);
1277	if (ret > 0) {
1278		ret = 0;
1279		goto out;
1280	}
1281	if (ret < 0)
1282		goto out;
1283
1284	eb = path->nodes[0];
1285	ref_ptr = btrfs_item_ptr_offset(eb, path->slots[0]);
1286	ref_end = ref_ptr + btrfs_item_size(eb, path->slots[0]);
1287	while (ref_ptr < ref_end) {
1288		struct fscrypt_str name;
1289		u64 parent_id;
1290
1291		if (key->type == BTRFS_INODE_EXTREF_KEY) {
1292			ret = extref_get_fields(eb, ref_ptr, &name,
1293						NULL, &parent_id);
1294		} else {
1295			parent_id = key->offset;
1296			ret = ref_get_fields(eb, ref_ptr, &name, NULL);
1297		}
1298		if (ret)
1299			goto out;
1300
1301		if (key->type == BTRFS_INODE_EXTREF_KEY)
1302			ret = !!btrfs_find_name_in_ext_backref(log_eb, log_slot,
1303							       parent_id, &name);
1304		else
1305			ret = !!btrfs_find_name_in_backref(log_eb, log_slot, &name);
1306
1307		if (!ret) {
1308			struct inode *dir;
1309
1310			btrfs_release_path(path);
1311			dir = read_one_inode(root, parent_id);
1312			if (!dir) {
1313				ret = -ENOENT;
1314				kfree(name.name);
1315				goto out;
1316			}
1317			ret = unlink_inode_for_log_replay(trans, BTRFS_I(dir),
1318						 inode, &name);
1319			kfree(name.name);
1320			iput(dir);
1321			if (ret)
1322				goto out;
1323			goto again;
1324		}
1325
1326		kfree(name.name);
1327		ref_ptr += name.len;
1328		if (key->type == BTRFS_INODE_EXTREF_KEY)
1329			ref_ptr += sizeof(struct btrfs_inode_extref);
1330		else
1331			ref_ptr += sizeof(struct btrfs_inode_ref);
1332	}
1333	ret = 0;
1334 out:
1335	btrfs_release_path(path);
1336	return ret;
1337}
1338
1339/*
1340 * replay one inode back reference item found in the log tree.
1341 * eb, slot and key refer to the buffer and key found in the log tree.
1342 * root is the destination we are replaying into, and path is for temp
1343 * use by this function.  (it should be released on return).
1344 */
1345static noinline int add_inode_ref(struct btrfs_trans_handle *trans,
1346				  struct btrfs_root *root,
1347				  struct btrfs_root *log,
1348				  struct btrfs_path *path,
1349				  struct extent_buffer *eb, int slot,
1350				  struct btrfs_key *key)
1351{
1352	struct inode *dir = NULL;
1353	struct inode *inode = NULL;
1354	unsigned long ref_ptr;
1355	unsigned long ref_end;
1356	struct fscrypt_str name;
1357	int ret;
1358	int log_ref_ver = 0;
1359	u64 parent_objectid;
1360	u64 inode_objectid;
1361	u64 ref_index = 0;
1362	int ref_struct_size;
1363
1364	ref_ptr = btrfs_item_ptr_offset(eb, slot);
1365	ref_end = ref_ptr + btrfs_item_size(eb, slot);
1366
1367	if (key->type == BTRFS_INODE_EXTREF_KEY) {
1368		struct btrfs_inode_extref *r;
1369
1370		ref_struct_size = sizeof(struct btrfs_inode_extref);
1371		log_ref_ver = 1;
1372		r = (struct btrfs_inode_extref *)ref_ptr;
1373		parent_objectid = btrfs_inode_extref_parent(eb, r);
1374	} else {
1375		ref_struct_size = sizeof(struct btrfs_inode_ref);
1376		parent_objectid = key->offset;
1377	}
1378	inode_objectid = key->objectid;
1379
1380	/*
1381	 * it is possible that we didn't log all the parent directories
1382	 * for a given inode.  If we don't find the dir, just don't
1383	 * copy the back ref in.  The link count fixup code will take
1384	 * care of the rest
1385	 */
1386	dir = read_one_inode(root, parent_objectid);
1387	if (!dir) {
1388		ret = -ENOENT;
1389		goto out;
1390	}
1391
1392	inode = read_one_inode(root, inode_objectid);
1393	if (!inode) {
1394		ret = -EIO;
1395		goto out;
1396	}
1397
1398	while (ref_ptr < ref_end) {
1399		if (log_ref_ver) {
1400			ret = extref_get_fields(eb, ref_ptr, &name,
1401						&ref_index, &parent_objectid);
1402			/*
1403			 * parent object can change from one array
1404			 * item to another.
1405			 */
1406			if (!dir)
1407				dir = read_one_inode(root, parent_objectid);
1408			if (!dir) {
1409				ret = -ENOENT;
1410				goto out;
1411			}
1412		} else {
1413			ret = ref_get_fields(eb, ref_ptr, &name, &ref_index);
1414		}
1415		if (ret)
1416			goto out;
1417
1418		ret = inode_in_dir(root, path, btrfs_ino(BTRFS_I(dir)),
1419				   btrfs_ino(BTRFS_I(inode)), ref_index, &name);
1420		if (ret < 0) {
1421			goto out;
1422		} else if (ret == 0) {
1423			/*
1424			 * look for a conflicting back reference in the
1425			 * metadata. if we find one we have to unlink that name
1426			 * of the file before we add our new link.  Later on, we
1427			 * overwrite any existing back reference, and we don't
1428			 * want to create dangling pointers in the directory.
1429			 */
1430			ret = __add_inode_ref(trans, root, path, log,
1431					      BTRFS_I(dir), BTRFS_I(inode),
1432					      inode_objectid, parent_objectid,
1433					      ref_index, &name);
1434			if (ret) {
1435				if (ret == 1)
1436					ret = 0;
1437				goto out;
1438			}
1439
1440			/* insert our name */
1441			ret = btrfs_add_link(trans, BTRFS_I(dir), BTRFS_I(inode),
1442					     &name, 0, ref_index);
1443			if (ret)
1444				goto out;
1445
1446			ret = btrfs_update_inode(trans, BTRFS_I(inode));
1447			if (ret)
1448				goto out;
1449		}
1450		/* Else, ret == 1, we already have a perfect match, we're done. */
1451
1452		ref_ptr = (unsigned long)(ref_ptr + ref_struct_size) + name.len;
1453		kfree(name.name);
1454		name.name = NULL;
1455		if (log_ref_ver) {
1456			iput(dir);
1457			dir = NULL;
1458		}
1459	}
1460
1461	/*
1462	 * Before we overwrite the inode reference item in the subvolume tree
1463	 * with the item from the log tree, we must unlink all names from the
1464	 * parent directory that are in the subvolume's tree inode reference
1465	 * item, otherwise we end up with an inconsistent subvolume tree where
1466	 * dir index entries exist for a name but there is no inode reference
1467	 * item with the same name.
1468	 */
1469	ret = unlink_old_inode_refs(trans, root, path, BTRFS_I(inode), eb, slot,
1470				    key);
1471	if (ret)
1472		goto out;
1473
1474	/* finally write the back reference in the inode */
1475	ret = overwrite_item(trans, root, path, eb, slot, key);
1476out:
1477	btrfs_release_path(path);
1478	kfree(name.name);
1479	iput(dir);
1480	iput(inode);
1481	return ret;
1482}
1483
1484static int count_inode_extrefs(struct btrfs_inode *inode, struct btrfs_path *path)
1485{
1486	int ret = 0;
1487	int name_len;
1488	unsigned int nlink = 0;
1489	u32 item_size;
1490	u32 cur_offset = 0;
1491	u64 inode_objectid = btrfs_ino(inode);
1492	u64 offset = 0;
1493	unsigned long ptr;
1494	struct btrfs_inode_extref *extref;
1495	struct extent_buffer *leaf;
1496
1497	while (1) {
1498		ret = btrfs_find_one_extref(inode->root, inode_objectid, offset,
1499					    path, &extref, &offset);
1500		if (ret)
1501			break;
1502
1503		leaf = path->nodes[0];
1504		item_size = btrfs_item_size(leaf, path->slots[0]);
1505		ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
1506		cur_offset = 0;
1507
1508		while (cur_offset < item_size) {
1509			extref = (struct btrfs_inode_extref *) (ptr + cur_offset);
1510			name_len = btrfs_inode_extref_name_len(leaf, extref);
1511
1512			nlink++;
1513
1514			cur_offset += name_len + sizeof(*extref);
1515		}
1516
1517		offset++;
1518		btrfs_release_path(path);
1519	}
1520	btrfs_release_path(path);
1521
1522	if (ret < 0 && ret != -ENOENT)
1523		return ret;
1524	return nlink;
1525}
1526
1527static int count_inode_refs(struct btrfs_inode *inode, struct btrfs_path *path)
1528{
1529	int ret;
1530	struct btrfs_key key;
1531	unsigned int nlink = 0;
1532	unsigned long ptr;
1533	unsigned long ptr_end;
1534	int name_len;
1535	u64 ino = btrfs_ino(inode);
1536
1537	key.objectid = ino;
1538	key.type = BTRFS_INODE_REF_KEY;
1539	key.offset = (u64)-1;
1540
1541	while (1) {
1542		ret = btrfs_search_slot(NULL, inode->root, &key, path, 0, 0);
1543		if (ret < 0)
1544			break;
1545		if (ret > 0) {
1546			if (path->slots[0] == 0)
1547				break;
1548			path->slots[0]--;
1549		}
1550process_slot:
1551		btrfs_item_key_to_cpu(path->nodes[0], &key,
1552				      path->slots[0]);
1553		if (key.objectid != ino ||
1554		    key.type != BTRFS_INODE_REF_KEY)
1555			break;
1556		ptr = btrfs_item_ptr_offset(path->nodes[0], path->slots[0]);
1557		ptr_end = ptr + btrfs_item_size(path->nodes[0],
1558						   path->slots[0]);
1559		while (ptr < ptr_end) {
1560			struct btrfs_inode_ref *ref;
1561
1562			ref = (struct btrfs_inode_ref *)ptr;
1563			name_len = btrfs_inode_ref_name_len(path->nodes[0],
1564							    ref);
1565			ptr = (unsigned long)(ref + 1) + name_len;
1566			nlink++;
1567		}
1568
1569		if (key.offset == 0)
1570			break;
1571		if (path->slots[0] > 0) {
1572			path->slots[0]--;
1573			goto process_slot;
1574		}
1575		key.offset--;
1576		btrfs_release_path(path);
1577	}
1578	btrfs_release_path(path);
1579
1580	return nlink;
1581}
1582
1583/*
1584 * There are a few corners where the link count of the file can't
1585 * be properly maintained during replay.  So, instead of adding
1586 * lots of complexity to the log code, we just scan the backrefs
1587 * for any file that has been through replay.
1588 *
1589 * The scan will update the link count on the inode to reflect the
1590 * number of back refs found.  If it goes down to zero, the iput
1591 * will free the inode.
1592 */
1593static noinline int fixup_inode_link_count(struct btrfs_trans_handle *trans,
1594					   struct inode *inode)
1595{
1596	struct btrfs_root *root = BTRFS_I(inode)->root;
1597	struct btrfs_path *path;
1598	int ret;
1599	u64 nlink = 0;
1600	u64 ino = btrfs_ino(BTRFS_I(inode));
1601
1602	path = btrfs_alloc_path();
1603	if (!path)
1604		return -ENOMEM;
1605
1606	ret = count_inode_refs(BTRFS_I(inode), path);
1607	if (ret < 0)
1608		goto out;
1609
1610	nlink = ret;
1611
1612	ret = count_inode_extrefs(BTRFS_I(inode), path);
1613	if (ret < 0)
1614		goto out;
1615
1616	nlink += ret;
1617
1618	ret = 0;
1619
1620	if (nlink != inode->i_nlink) {
1621		set_nlink(inode, nlink);
1622		ret = btrfs_update_inode(trans, BTRFS_I(inode));
1623		if (ret)
1624			goto out;
1625	}
1626	BTRFS_I(inode)->index_cnt = (u64)-1;
1627
1628	if (inode->i_nlink == 0) {
1629		if (S_ISDIR(inode->i_mode)) {
1630			ret = replay_dir_deletes(trans, root, NULL, path,
1631						 ino, 1);
1632			if (ret)
1633				goto out;
1634		}
1635		ret = btrfs_insert_orphan_item(trans, root, ino);
1636		if (ret == -EEXIST)
1637			ret = 0;
1638	}
1639
1640out:
1641	btrfs_free_path(path);
1642	return ret;
1643}
1644
1645static noinline int fixup_inode_link_counts(struct btrfs_trans_handle *trans,
1646					    struct btrfs_root *root,
1647					    struct btrfs_path *path)
1648{
1649	int ret;
1650	struct btrfs_key key;
1651	struct inode *inode;
1652
1653	key.objectid = BTRFS_TREE_LOG_FIXUP_OBJECTID;
1654	key.type = BTRFS_ORPHAN_ITEM_KEY;
1655	key.offset = (u64)-1;
1656	while (1) {
1657		ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
1658		if (ret < 0)
1659			break;
1660
1661		if (ret == 1) {
1662			ret = 0;
1663			if (path->slots[0] == 0)
1664				break;
1665			path->slots[0]--;
1666		}
1667
1668		btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
1669		if (key.objectid != BTRFS_TREE_LOG_FIXUP_OBJECTID ||
1670		    key.type != BTRFS_ORPHAN_ITEM_KEY)
1671			break;
1672
1673		ret = btrfs_del_item(trans, root, path);
1674		if (ret)
1675			break;
1676
1677		btrfs_release_path(path);
1678		inode = read_one_inode(root, key.offset);
1679		if (!inode) {
1680			ret = -EIO;
1681			break;
1682		}
1683
1684		ret = fixup_inode_link_count(trans, inode);
1685		iput(inode);
1686		if (ret)
1687			break;
1688
1689		/*
1690		 * fixup on a directory may create new entries,
1691		 * make sure we always look for the highset possible
1692		 * offset
1693		 */
1694		key.offset = (u64)-1;
1695	}
1696	btrfs_release_path(path);
1697	return ret;
1698}
1699
1700
1701/*
1702 * record a given inode in the fixup dir so we can check its link
1703 * count when replay is done.  The link count is incremented here
1704 * so the inode won't go away until we check it
1705 */
1706static noinline int link_to_fixup_dir(struct btrfs_trans_handle *trans,
1707				      struct btrfs_root *root,
1708				      struct btrfs_path *path,
1709				      u64 objectid)
1710{
1711	struct btrfs_key key;
1712	int ret = 0;
1713	struct inode *inode;
1714
1715	inode = read_one_inode(root, objectid);
1716	if (!inode)
1717		return -EIO;
1718
1719	key.objectid = BTRFS_TREE_LOG_FIXUP_OBJECTID;
1720	key.type = BTRFS_ORPHAN_ITEM_KEY;
1721	key.offset = objectid;
1722
1723	ret = btrfs_insert_empty_item(trans, root, path, &key, 0);
1724
1725	btrfs_release_path(path);
1726	if (ret == 0) {
1727		if (!inode->i_nlink)
1728			set_nlink(inode, 1);
1729		else
1730			inc_nlink(inode);
1731		ret = btrfs_update_inode(trans, BTRFS_I(inode));
1732	} else if (ret == -EEXIST) {
1733		ret = 0;
1734	}
1735	iput(inode);
1736
1737	return ret;
1738}
1739
1740/*
1741 * when replaying the log for a directory, we only insert names
1742 * for inodes that actually exist.  This means an fsync on a directory
1743 * does not implicitly fsync all the new files in it
1744 */
1745static noinline int insert_one_name(struct btrfs_trans_handle *trans,
1746				    struct btrfs_root *root,
1747				    u64 dirid, u64 index,
1748				    const struct fscrypt_str *name,
1749				    struct btrfs_key *location)
1750{
1751	struct inode *inode;
1752	struct inode *dir;
1753	int ret;
1754
1755	inode = read_one_inode(root, location->objectid);
1756	if (!inode)
1757		return -ENOENT;
1758
1759	dir = read_one_inode(root, dirid);
1760	if (!dir) {
1761		iput(inode);
1762		return -EIO;
1763	}
1764
1765	ret = btrfs_add_link(trans, BTRFS_I(dir), BTRFS_I(inode), name,
1766			     1, index);
1767
1768	/* FIXME, put inode into FIXUP list */
1769
1770	iput(inode);
1771	iput(dir);
1772	return ret;
1773}
1774
1775static int delete_conflicting_dir_entry(struct btrfs_trans_handle *trans,
1776					struct btrfs_inode *dir,
1777					struct btrfs_path *path,
1778					struct btrfs_dir_item *dst_di,
1779					const struct btrfs_key *log_key,
1780					u8 log_flags,
1781					bool exists)
1782{
1783	struct btrfs_key found_key;
1784
1785	btrfs_dir_item_key_to_cpu(path->nodes[0], dst_di, &found_key);
1786	/* The existing dentry points to the same inode, don't delete it. */
1787	if (found_key.objectid == log_key->objectid &&
1788	    found_key.type == log_key->type &&
1789	    found_key.offset == log_key->offset &&
1790	    btrfs_dir_flags(path->nodes[0], dst_di) == log_flags)
1791		return 1;
1792
1793	/*
1794	 * Don't drop the conflicting directory entry if the inode for the new
1795	 * entry doesn't exist.
1796	 */
1797	if (!exists)
1798		return 0;
1799
1800	return drop_one_dir_item(trans, path, dir, dst_di);
1801}
1802
1803/*
1804 * take a single entry in a log directory item and replay it into
1805 * the subvolume.
1806 *
1807 * if a conflicting item exists in the subdirectory already,
1808 * the inode it points to is unlinked and put into the link count
1809 * fix up tree.
1810 *
1811 * If a name from the log points to a file or directory that does
1812 * not exist in the FS, it is skipped.  fsyncs on directories
1813 * do not force down inodes inside that directory, just changes to the
1814 * names or unlinks in a directory.
1815 *
1816 * Returns < 0 on error, 0 if the name wasn't replayed (dentry points to a
1817 * non-existing inode) and 1 if the name was replayed.
1818 */
1819static noinline int replay_one_name(struct btrfs_trans_handle *trans,
1820				    struct btrfs_root *root,
1821				    struct btrfs_path *path,
1822				    struct extent_buffer *eb,
1823				    struct btrfs_dir_item *di,
1824				    struct btrfs_key *key)
1825{
1826	struct fscrypt_str name;
1827	struct btrfs_dir_item *dir_dst_di;
1828	struct btrfs_dir_item *index_dst_di;
1829	bool dir_dst_matches = false;
1830	bool index_dst_matches = false;
1831	struct btrfs_key log_key;
1832	struct btrfs_key search_key;
1833	struct inode *dir;
1834	u8 log_flags;
1835	bool exists;
1836	int ret;
1837	bool update_size = true;
1838	bool name_added = false;
1839
1840	dir = read_one_inode(root, key->objectid);
1841	if (!dir)
1842		return -EIO;
1843
1844	ret = read_alloc_one_name(eb, di + 1, btrfs_dir_name_len(eb, di), &name);
1845	if (ret)
1846		goto out;
1847
1848	log_flags = btrfs_dir_flags(eb, di);
1849	btrfs_dir_item_key_to_cpu(eb, di, &log_key);
1850	ret = btrfs_lookup_inode(trans, root, path, &log_key, 0);
1851	btrfs_release_path(path);
1852	if (ret < 0)
1853		goto out;
1854	exists = (ret == 0);
1855	ret = 0;
1856
1857	dir_dst_di = btrfs_lookup_dir_item(trans, root, path, key->objectid,
1858					   &name, 1);
1859	if (IS_ERR(dir_dst_di)) {
1860		ret = PTR_ERR(dir_dst_di);
1861		goto out;
1862	} else if (dir_dst_di) {
1863		ret = delete_conflicting_dir_entry(trans, BTRFS_I(dir), path,
1864						   dir_dst_di, &log_key,
1865						   log_flags, exists);
1866		if (ret < 0)
1867			goto out;
1868		dir_dst_matches = (ret == 1);
1869	}
1870
1871	btrfs_release_path(path);
1872
1873	index_dst_di = btrfs_lookup_dir_index_item(trans, root, path,
1874						   key->objectid, key->offset,
1875						   &name, 1);
1876	if (IS_ERR(index_dst_di)) {
1877		ret = PTR_ERR(index_dst_di);
1878		goto out;
1879	} else if (index_dst_di) {
1880		ret = delete_conflicting_dir_entry(trans, BTRFS_I(dir), path,
1881						   index_dst_di, &log_key,
1882						   log_flags, exists);
1883		if (ret < 0)
1884			goto out;
1885		index_dst_matches = (ret == 1);
1886	}
1887
1888	btrfs_release_path(path);
1889
1890	if (dir_dst_matches && index_dst_matches) {
1891		ret = 0;
1892		update_size = false;
1893		goto out;
1894	}
1895
1896	/*
1897	 * Check if the inode reference exists in the log for the given name,
1898	 * inode and parent inode
1899	 */
1900	search_key.objectid = log_key.objectid;
1901	search_key.type = BTRFS_INODE_REF_KEY;
1902	search_key.offset = key->objectid;
1903	ret = backref_in_log(root->log_root, &search_key, 0, &name);
1904	if (ret < 0) {
1905	        goto out;
1906	} else if (ret) {
1907	        /* The dentry will be added later. */
1908	        ret = 0;
1909	        update_size = false;
1910	        goto out;
1911	}
1912
1913	search_key.objectid = log_key.objectid;
1914	search_key.type = BTRFS_INODE_EXTREF_KEY;
1915	search_key.offset = key->objectid;
1916	ret = backref_in_log(root->log_root, &search_key, key->objectid, &name);
1917	if (ret < 0) {
1918		goto out;
1919	} else if (ret) {
1920		/* The dentry will be added later. */
1921		ret = 0;
1922		update_size = false;
1923		goto out;
1924	}
1925	btrfs_release_path(path);
1926	ret = insert_one_name(trans, root, key->objectid, key->offset,
1927			      &name, &log_key);
1928	if (ret && ret != -ENOENT && ret != -EEXIST)
1929		goto out;
1930	if (!ret)
1931		name_added = true;
1932	update_size = false;
1933	ret = 0;
1934
1935out:
1936	if (!ret && update_size) {
1937		btrfs_i_size_write(BTRFS_I(dir), dir->i_size + name.len * 2);
1938		ret = btrfs_update_inode(trans, BTRFS_I(dir));
1939	}
1940	kfree(name.name);
1941	iput(dir);
1942	if (!ret && name_added)
1943		ret = 1;
1944	return ret;
1945}
1946
1947/* Replay one dir item from a BTRFS_DIR_INDEX_KEY key. */
1948static noinline int replay_one_dir_item(struct btrfs_trans_handle *trans,
1949					struct btrfs_root *root,
1950					struct btrfs_path *path,
1951					struct extent_buffer *eb, int slot,
1952					struct btrfs_key *key)
1953{
1954	int ret;
1955	struct btrfs_dir_item *di;
1956
1957	/* We only log dir index keys, which only contain a single dir item. */
1958	ASSERT(key->type == BTRFS_DIR_INDEX_KEY);
1959
1960	di = btrfs_item_ptr(eb, slot, struct btrfs_dir_item);
1961	ret = replay_one_name(trans, root, path, eb, di, key);
1962	if (ret < 0)
1963		return ret;
1964
1965	/*
1966	 * If this entry refers to a non-directory (directories can not have a
1967	 * link count > 1) and it was added in the transaction that was not
1968	 * committed, make sure we fixup the link count of the inode the entry
1969	 * points to. Otherwise something like the following would result in a
1970	 * directory pointing to an inode with a wrong link that does not account
1971	 * for this dir entry:
1972	 *
1973	 * mkdir testdir
1974	 * touch testdir/foo
1975	 * touch testdir/bar
1976	 * sync
1977	 *
1978	 * ln testdir/bar testdir/bar_link
1979	 * ln testdir/foo testdir/foo_link
1980	 * xfs_io -c "fsync" testdir/bar
1981	 *
1982	 * <power failure>
1983	 *
1984	 * mount fs, log replay happens
1985	 *
1986	 * File foo would remain with a link count of 1 when it has two entries
1987	 * pointing to it in the directory testdir. This would make it impossible
1988	 * to ever delete the parent directory has it would result in stale
1989	 * dentries that can never be deleted.
1990	 */
1991	if (ret == 1 && btrfs_dir_ftype(eb, di) != BTRFS_FT_DIR) {
1992		struct btrfs_path *fixup_path;
1993		struct btrfs_key di_key;
1994
1995		fixup_path = btrfs_alloc_path();
1996		if (!fixup_path)
1997			return -ENOMEM;
1998
1999		btrfs_dir_item_key_to_cpu(eb, di, &di_key);
2000		ret = link_to_fixup_dir(trans, root, fixup_path, di_key.objectid);
2001		btrfs_free_path(fixup_path);
2002	}
2003
2004	return ret;
2005}
2006
2007/*
2008 * directory replay has two parts.  There are the standard directory
2009 * items in the log copied from the subvolume, and range items
2010 * created in the log while the subvolume was logged.
2011 *
2012 * The range items tell us which parts of the key space the log
2013 * is authoritative for.  During replay, if a key in the subvolume
2014 * directory is in a logged range item, but not actually in the log
2015 * that means it was deleted from the directory before the fsync
2016 * and should be removed.
2017 */
2018static noinline int find_dir_range(struct btrfs_root *root,
2019				   struct btrfs_path *path,
2020				   u64 dirid,
2021				   u64 *start_ret, u64 *end_ret)
2022{
2023	struct btrfs_key key;
2024	u64 found_end;
2025	struct btrfs_dir_log_item *item;
2026	int ret;
2027	int nritems;
2028
2029	if (*start_ret == (u64)-1)
2030		return 1;
2031
2032	key.objectid = dirid;
2033	key.type = BTRFS_DIR_LOG_INDEX_KEY;
2034	key.offset = *start_ret;
2035
2036	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
2037	if (ret < 0)
2038		goto out;
2039	if (ret > 0) {
2040		if (path->slots[0] == 0)
2041			goto out;
2042		path->slots[0]--;
2043	}
2044	if (ret != 0)
2045		btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
2046
2047	if (key.type != BTRFS_DIR_LOG_INDEX_KEY || key.objectid != dirid) {
2048		ret = 1;
2049		goto next;
2050	}
2051	item = btrfs_item_ptr(path->nodes[0], path->slots[0],
2052			      struct btrfs_dir_log_item);
2053	found_end = btrfs_dir_log_end(path->nodes[0], item);
2054
2055	if (*start_ret >= key.offset && *start_ret <= found_end) {
2056		ret = 0;
2057		*start_ret = key.offset;
2058		*end_ret = found_end;
2059		goto out;
2060	}
2061	ret = 1;
2062next:
2063	/* check the next slot in the tree to see if it is a valid item */
2064	nritems = btrfs_header_nritems(path->nodes[0]);
2065	path->slots[0]++;
2066	if (path->slots[0] >= nritems) {
2067		ret = btrfs_next_leaf(root, path);
2068		if (ret)
2069			goto out;
2070	}
2071
2072	btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
2073
2074	if (key.type != BTRFS_DIR_LOG_INDEX_KEY || key.objectid != dirid) {
2075		ret = 1;
2076		goto out;
2077	}
2078	item = btrfs_item_ptr(path->nodes[0], path->slots[0],
2079			      struct btrfs_dir_log_item);
2080	found_end = btrfs_dir_log_end(path->nodes[0], item);
2081	*start_ret = key.offset;
2082	*end_ret = found_end;
2083	ret = 0;
2084out:
2085	btrfs_release_path(path);
2086	return ret;
2087}
2088
2089/*
2090 * this looks for a given directory item in the log.  If the directory
2091 * item is not in the log, the item is removed and the inode it points
2092 * to is unlinked
2093 */
2094static noinline int check_item_in_log(struct btrfs_trans_handle *trans,
2095				      struct btrfs_root *log,
2096				      struct btrfs_path *path,
2097				      struct btrfs_path *log_path,
2098				      struct inode *dir,
2099				      struct btrfs_key *dir_key)
2100{
2101	struct btrfs_root *root = BTRFS_I(dir)->root;
2102	int ret;
2103	struct extent_buffer *eb;
2104	int slot;
2105	struct btrfs_dir_item *di;
2106	struct fscrypt_str name;
2107	struct inode *inode = NULL;
2108	struct btrfs_key location;
2109
2110	/*
2111	 * Currently we only log dir index keys. Even if we replay a log created
2112	 * by an older kernel that logged both dir index and dir item keys, all
2113	 * we need to do is process the dir index keys, we (and our caller) can
2114	 * safely ignore dir item keys (key type BTRFS_DIR_ITEM_KEY).
2115	 */
2116	ASSERT(dir_key->type == BTRFS_DIR_INDEX_KEY);
2117
2118	eb = path->nodes[0];
2119	slot = path->slots[0];
2120	di = btrfs_item_ptr(eb, slot, struct btrfs_dir_item);
2121	ret = read_alloc_one_name(eb, di + 1, btrfs_dir_name_len(eb, di), &name);
2122	if (ret)
2123		goto out;
2124
2125	if (log) {
2126		struct btrfs_dir_item *log_di;
2127
2128		log_di = btrfs_lookup_dir_index_item(trans, log, log_path,
2129						     dir_key->objectid,
2130						     dir_key->offset, &name, 0);
2131		if (IS_ERR(log_di)) {
2132			ret = PTR_ERR(log_di);
2133			goto out;
2134		} else if (log_di) {
2135			/* The dentry exists in the log, we have nothing to do. */
2136			ret = 0;
2137			goto out;
2138		}
2139	}
2140
2141	btrfs_dir_item_key_to_cpu(eb, di, &location);
2142	btrfs_release_path(path);
2143	btrfs_release_path(log_path);
2144	inode = read_one_inode(root, location.objectid);
2145	if (!inode) {
2146		ret = -EIO;
2147		goto out;
2148	}
2149
2150	ret = link_to_fixup_dir(trans, root, path, location.objectid);
2151	if (ret)
2152		goto out;
2153
2154	inc_nlink(inode);
2155	ret = unlink_inode_for_log_replay(trans, BTRFS_I(dir), BTRFS_I(inode),
2156					  &name);
2157	/*
2158	 * Unlike dir item keys, dir index keys can only have one name (entry) in
2159	 * them, as there are no key collisions since each key has a unique offset
2160	 * (an index number), so we're done.
2161	 */
2162out:
2163	btrfs_release_path(path);
2164	btrfs_release_path(log_path);
2165	kfree(name.name);
2166	iput(inode);
2167	return ret;
2168}
2169
2170static int replay_xattr_deletes(struct btrfs_trans_handle *trans,
2171			      struct btrfs_root *root,
2172			      struct btrfs_root *log,
2173			      struct btrfs_path *path,
2174			      const u64 ino)
2175{
2176	struct btrfs_key search_key;
2177	struct btrfs_path *log_path;
2178	int i;
2179	int nritems;
2180	int ret;
2181
2182	log_path = btrfs_alloc_path();
2183	if (!log_path)
2184		return -ENOMEM;
2185
2186	search_key.objectid = ino;
2187	search_key.type = BTRFS_XATTR_ITEM_KEY;
2188	search_key.offset = 0;
2189again:
2190	ret = btrfs_search_slot(NULL, root, &search_key, path, 0, 0);
2191	if (ret < 0)
2192		goto out;
2193process_leaf:
2194	nritems = btrfs_header_nritems(path->nodes[0]);
2195	for (i = path->slots[0]; i < nritems; i++) {
2196		struct btrfs_key key;
2197		struct btrfs_dir_item *di;
2198		struct btrfs_dir_item *log_di;
2199		u32 total_size;
2200		u32 cur;
2201
2202		btrfs_item_key_to_cpu(path->nodes[0], &key, i);
2203		if (key.objectid != ino || key.type != BTRFS_XATTR_ITEM_KEY) {
2204			ret = 0;
2205			goto out;
2206		}
2207
2208		di = btrfs_item_ptr(path->nodes[0], i, struct btrfs_dir_item);
2209		total_size = btrfs_item_size(path->nodes[0], i);
2210		cur = 0;
2211		while (cur < total_size) {
2212			u16 name_len = btrfs_dir_name_len(path->nodes[0], di);
2213			u16 data_len = btrfs_dir_data_len(path->nodes[0], di);
2214			u32 this_len = sizeof(*di) + name_len + data_len;
2215			char *name;
2216
2217			name = kmalloc(name_len, GFP_NOFS);
2218			if (!name) {
2219				ret = -ENOMEM;
2220				goto out;
2221			}
2222			read_extent_buffer(path->nodes[0], name,
2223					   (unsigned long)(di + 1), name_len);
2224
2225			log_di = btrfs_lookup_xattr(NULL, log, log_path, ino,
2226						    name, name_len, 0);
2227			btrfs_release_path(log_path);
2228			if (!log_di) {
2229				/* Doesn't exist in log tree, so delete it. */
2230				btrfs_release_path(path);
2231				di = btrfs_lookup_xattr(trans, root, path, ino,
2232							name, name_len, -1);
2233				kfree(name);
2234				if (IS_ERR(di)) {
2235					ret = PTR_ERR(di);
2236					goto out;
2237				}
2238				ASSERT(di);
2239				ret = btrfs_delete_one_dir_name(trans, root,
2240								path, di);
2241				if (ret)
2242					goto out;
2243				btrfs_release_path(path);
2244				search_key = key;
2245				goto again;
2246			}
2247			kfree(name);
2248			if (IS_ERR(log_di)) {
2249				ret = PTR_ERR(log_di);
2250				goto out;
2251			}
2252			cur += this_len;
2253			di = (struct btrfs_dir_item *)((char *)di + this_len);
2254		}
2255	}
2256	ret = btrfs_next_leaf(root, path);
2257	if (ret > 0)
2258		ret = 0;
2259	else if (ret == 0)
2260		goto process_leaf;
2261out:
2262	btrfs_free_path(log_path);
2263	btrfs_release_path(path);
2264	return ret;
2265}
2266
2267
2268/*
2269 * deletion replay happens before we copy any new directory items
2270 * out of the log or out of backreferences from inodes.  It
2271 * scans the log to find ranges of keys that log is authoritative for,
2272 * and then scans the directory to find items in those ranges that are
2273 * not present in the log.
2274 *
2275 * Anything we don't find in the log is unlinked and removed from the
2276 * directory.
2277 */
2278static noinline int replay_dir_deletes(struct btrfs_trans_handle *trans,
2279				       struct btrfs_root *root,
2280				       struct btrfs_root *log,
2281				       struct btrfs_path *path,
2282				       u64 dirid, int del_all)
2283{
2284	u64 range_start;
2285	u64 range_end;
2286	int ret = 0;
2287	struct btrfs_key dir_key;
2288	struct btrfs_key found_key;
2289	struct btrfs_path *log_path;
2290	struct inode *dir;
2291
2292	dir_key.objectid = dirid;
2293	dir_key.type = BTRFS_DIR_INDEX_KEY;
2294	log_path = btrfs_alloc_path();
2295	if (!log_path)
2296		return -ENOMEM;
2297
2298	dir = read_one_inode(root, dirid);
2299	/* it isn't an error if the inode isn't there, that can happen
2300	 * because we replay the deletes before we copy in the inode item
2301	 * from the log
2302	 */
2303	if (!dir) {
2304		btrfs_free_path(log_path);
2305		return 0;
2306	}
2307
2308	range_start = 0;
2309	range_end = 0;
2310	while (1) {
2311		if (del_all)
2312			range_end = (u64)-1;
2313		else {
2314			ret = find_dir_range(log, path, dirid,
2315					     &range_start, &range_end);
2316			if (ret < 0)
2317				goto out;
2318			else if (ret > 0)
2319				break;
2320		}
2321
2322		dir_key.offset = range_start;
2323		while (1) {
2324			int nritems;
2325			ret = btrfs_search_slot(NULL, root, &dir_key, path,
2326						0, 0);
2327			if (ret < 0)
2328				goto out;
2329
2330			nritems = btrfs_header_nritems(path->nodes[0]);
2331			if (path->slots[0] >= nritems) {
2332				ret = btrfs_next_leaf(root, path);
2333				if (ret == 1)
2334					break;
2335				else if (ret < 0)
2336					goto out;
2337			}
2338			btrfs_item_key_to_cpu(path->nodes[0], &found_key,
2339					      path->slots[0]);
2340			if (found_key.objectid != dirid ||
2341			    found_key.type != dir_key.type) {
2342				ret = 0;
2343				goto out;
2344			}
2345
2346			if (found_key.offset > range_end)
2347				break;
2348
2349			ret = check_item_in_log(trans, log, path,
2350						log_path, dir,
2351						&found_key);
2352			if (ret)
2353				goto out;
2354			if (found_key.offset == (u64)-1)
2355				break;
2356			dir_key.offset = found_key.offset + 1;
2357		}
2358		btrfs_release_path(path);
2359		if (range_end == (u64)-1)
2360			break;
2361		range_start = range_end + 1;
2362	}
2363	ret = 0;
2364out:
2365	btrfs_release_path(path);
2366	btrfs_free_path(log_path);
2367	iput(dir);
2368	return ret;
2369}
2370
2371/*
2372 * the process_func used to replay items from the log tree.  This
2373 * gets called in two different stages.  The first stage just looks
2374 * for inodes and makes sure they are all copied into the subvolume.
2375 *
2376 * The second stage copies all the other item types from the log into
2377 * the subvolume.  The two stage approach is slower, but gets rid of
2378 * lots of complexity around inodes referencing other inodes that exist
2379 * only in the log (references come from either directory items or inode
2380 * back refs).
2381 */
2382static int replay_one_buffer(struct btrfs_root *log, struct extent_buffer *eb,
2383			     struct walk_control *wc, u64 gen, int level)
2384{
2385	int nritems;
2386	struct btrfs_tree_parent_check check = {
2387		.transid = gen,
2388		.level = level
2389	};
2390	struct btrfs_path *path;
2391	struct btrfs_root *root = wc->replay_dest;
2392	struct btrfs_key key;
2393	int i;
2394	int ret;
2395
2396	ret = btrfs_read_extent_buffer(eb, &check);
2397	if (ret)
2398		return ret;
2399
2400	level = btrfs_header_level(eb);
2401
2402	if (level != 0)
2403		return 0;
2404
2405	path = btrfs_alloc_path();
2406	if (!path)
2407		return -ENOMEM;
2408
2409	nritems = btrfs_header_nritems(eb);
2410	for (i = 0; i < nritems; i++) {
2411		btrfs_item_key_to_cpu(eb, &key, i);
2412
2413		/* inode keys are done during the first stage */
2414		if (key.type == BTRFS_INODE_ITEM_KEY &&
2415		    wc->stage == LOG_WALK_REPLAY_INODES) {
2416			struct btrfs_inode_item *inode_item;
2417			u32 mode;
2418
2419			inode_item = btrfs_item_ptr(eb, i,
2420					    struct btrfs_inode_item);
2421			/*
2422			 * If we have a tmpfile (O_TMPFILE) that got fsync'ed
2423			 * and never got linked before the fsync, skip it, as
2424			 * replaying it is pointless since it would be deleted
2425			 * later. We skip logging tmpfiles, but it's always
2426			 * possible we are replaying a log created with a kernel
2427			 * that used to log tmpfiles.
2428			 */
2429			if (btrfs_inode_nlink(eb, inode_item) == 0) {
2430				wc->ignore_cur_inode = true;
2431				continue;
2432			} else {
2433				wc->ignore_cur_inode = false;
2434			}
2435			ret = replay_xattr_deletes(wc->trans, root, log,
2436						   path, key.objectid);
2437			if (ret)
2438				break;
2439			mode = btrfs_inode_mode(eb, inode_item);
2440			if (S_ISDIR(mode)) {
2441				ret = replay_dir_deletes(wc->trans,
2442					 root, log, path, key.objectid, 0);
2443				if (ret)
2444					break;
2445			}
2446			ret = overwrite_item(wc->trans, root, path,
2447					     eb, i, &key);
2448			if (ret)
2449				break;
2450
2451			/*
2452			 * Before replaying extents, truncate the inode to its
2453			 * size. We need to do it now and not after log replay
2454			 * because before an fsync we can have prealloc extents
2455			 * added beyond the inode's i_size. If we did it after,
2456			 * through orphan cleanup for example, we would drop
2457			 * those prealloc extents just after replaying them.
2458			 */
2459			if (S_ISREG(mode)) {
2460				struct btrfs_drop_extents_args drop_args = { 0 };
2461				struct inode *inode;
2462				u64 from;
2463
2464				inode = read_one_inode(root, key.objectid);
2465				if (!inode) {
2466					ret = -EIO;
2467					break;
2468				}
2469				from = ALIGN(i_size_read(inode),
2470					     root->fs_info->sectorsize);
2471				drop_args.start = from;
2472				drop_args.end = (u64)-1;
2473				drop_args.drop_cache = true;
2474				ret = btrfs_drop_extents(wc->trans, root,
2475							 BTRFS_I(inode),
2476							 &drop_args);
2477				if (!ret) {
2478					inode_sub_bytes(inode,
2479							drop_args.bytes_found);
2480					/* Update the inode's nbytes. */
2481					ret = btrfs_update_inode(wc->trans,
2482								 BTRFS_I(inode));
2483				}
2484				iput(inode);
2485				if (ret)
2486					break;
2487			}
2488
2489			ret = link_to_fixup_dir(wc->trans, root,
2490						path, key.objectid);
2491			if (ret)
2492				break;
2493		}
2494
2495		if (wc->ignore_cur_inode)
2496			continue;
2497
2498		if (key.type == BTRFS_DIR_INDEX_KEY &&
2499		    wc->stage == LOG_WALK_REPLAY_DIR_INDEX) {
2500			ret = replay_one_dir_item(wc->trans, root, path,
2501						  eb, i, &key);
2502			if (ret)
2503				break;
2504		}
2505
2506		if (wc->stage < LOG_WALK_REPLAY_ALL)
2507			continue;
2508
2509		/* these keys are simply copied */
2510		if (key.type == BTRFS_XATTR_ITEM_KEY) {
2511			ret = overwrite_item(wc->trans, root, path,
2512					     eb, i, &key);
2513			if (ret)
2514				break;
2515		} else if (key.type == BTRFS_INODE_REF_KEY ||
2516			   key.type == BTRFS_INODE_EXTREF_KEY) {
2517			ret = add_inode_ref(wc->trans, root, log, path,
2518					    eb, i, &key);
2519			if (ret && ret != -ENOENT)
2520				break;
2521			ret = 0;
2522		} else if (key.type == BTRFS_EXTENT_DATA_KEY) {
2523			ret = replay_one_extent(wc->trans, root, path,
2524						eb, i, &key);
2525			if (ret)
2526				break;
2527		}
2528		/*
2529		 * We don't log BTRFS_DIR_ITEM_KEY keys anymore, only the
2530		 * BTRFS_DIR_INDEX_KEY items which we use to derive the
2531		 * BTRFS_DIR_ITEM_KEY items. If we are replaying a log from an
2532		 * older kernel with such keys, ignore them.
2533		 */
2534	}
2535	btrfs_free_path(path);
2536	return ret;
2537}
2538
2539/*
2540 * Correctly adjust the reserved bytes occupied by a log tree extent buffer
2541 */
2542static void unaccount_log_buffer(struct btrfs_fs_info *fs_info, u64 start)
2543{
2544	struct btrfs_block_group *cache;
2545
2546	cache = btrfs_lookup_block_group(fs_info, start);
2547	if (!cache) {
2548		btrfs_err(fs_info, "unable to find block group for %llu", start);
2549		return;
2550	}
2551
2552	spin_lock(&cache->space_info->lock);
2553	spin_lock(&cache->lock);
2554	cache->reserved -= fs_info->nodesize;
2555	cache->space_info->bytes_reserved -= fs_info->nodesize;
2556	spin_unlock(&cache->lock);
2557	spin_unlock(&cache->space_info->lock);
2558
2559	btrfs_put_block_group(cache);
2560}
2561
2562static int clean_log_buffer(struct btrfs_trans_handle *trans,
2563			    struct extent_buffer *eb)
2564{
2565	int ret;
2566
2567	btrfs_tree_lock(eb);
2568	btrfs_clear_buffer_dirty(trans, eb);
2569	wait_on_extent_buffer_writeback(eb);
2570	btrfs_tree_unlock(eb);
2571
2572	if (trans) {
2573		ret = btrfs_pin_reserved_extent(trans, eb);
2574		if (ret)
2575			return ret;
2576	} else {
2577		unaccount_log_buffer(eb->fs_info, eb->start);
2578	}
2579
2580	return 0;
2581}
2582
2583static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans,
2584				   struct btrfs_root *root,
2585				   struct btrfs_path *path, int *level,
2586				   struct walk_control *wc)
2587{
2588	struct btrfs_fs_info *fs_info = root->fs_info;
2589	u64 bytenr;
2590	u64 ptr_gen;
2591	struct extent_buffer *next;
2592	struct extent_buffer *cur;
2593	int ret = 0;
2594
2595	while (*level > 0) {
2596		struct btrfs_tree_parent_check check = { 0 };
2597
2598		cur = path->nodes[*level];
2599
2600		WARN_ON(btrfs_header_level(cur) != *level);
2601
2602		if (path->slots[*level] >=
2603		    btrfs_header_nritems(cur))
2604			break;
2605
2606		bytenr = btrfs_node_blockptr(cur, path->slots[*level]);
2607		ptr_gen = btrfs_node_ptr_generation(cur, path->slots[*level]);
2608		check.transid = ptr_gen;
2609		check.level = *level - 1;
2610		check.has_first_key = true;
2611		btrfs_node_key_to_cpu(cur, &check.first_key, path->slots[*level]);
2612
2613		next = btrfs_find_create_tree_block(fs_info, bytenr,
2614						    btrfs_header_owner(cur),
2615						    *level - 1);
2616		if (IS_ERR(next))
2617			return PTR_ERR(next);
2618
2619		if (*level == 1) {
2620			ret = wc->process_func(root, next, wc, ptr_gen,
2621					       *level - 1);
2622			if (ret) {
2623				free_extent_buffer(next);
2624				return ret;
2625			}
2626
2627			path->slots[*level]++;
2628			if (wc->free) {
2629				ret = btrfs_read_extent_buffer(next, &check);
2630				if (ret) {
2631					free_extent_buffer(next);
2632					return ret;
2633				}
2634
2635				ret = clean_log_buffer(trans, next);
2636				if (ret) {
2637					free_extent_buffer(next);
2638					return ret;
2639				}
2640			}
2641			free_extent_buffer(next);
2642			continue;
2643		}
2644		ret = btrfs_read_extent_buffer(next, &check);
2645		if (ret) {
2646			free_extent_buffer(next);
2647			return ret;
2648		}
2649
2650		if (path->nodes[*level-1])
2651			free_extent_buffer(path->nodes[*level-1]);
2652		path->nodes[*level-1] = next;
2653		*level = btrfs_header_level(next);
2654		path->slots[*level] = 0;
2655		cond_resched();
2656	}
2657	path->slots[*level] = btrfs_header_nritems(path->nodes[*level]);
2658
2659	cond_resched();
2660	return 0;
2661}
2662
2663static noinline int walk_up_log_tree(struct btrfs_trans_handle *trans,
2664				 struct btrfs_root *root,
2665				 struct btrfs_path *path, int *level,
2666				 struct walk_control *wc)
2667{
2668	int i;
2669	int slot;
2670	int ret;
2671
2672	for (i = *level; i < BTRFS_MAX_LEVEL - 1 && path->nodes[i]; i++) {
2673		slot = path->slots[i];
2674		if (slot + 1 < btrfs_header_nritems(path->nodes[i])) {
2675			path->slots[i]++;
2676			*level = i;
2677			WARN_ON(*level == 0);
2678			return 0;
2679		} else {
2680			ret = wc->process_func(root, path->nodes[*level], wc,
2681				 btrfs_header_generation(path->nodes[*level]),
2682				 *level);
2683			if (ret)
2684				return ret;
2685
2686			if (wc->free) {
2687				ret = clean_log_buffer(trans, path->nodes[*level]);
2688				if (ret)
2689					return ret;
2690			}
2691			free_extent_buffer(path->nodes[*level]);
2692			path->nodes[*level] = NULL;
2693			*level = i + 1;
2694		}
2695	}
2696	return 1;
2697}
2698
2699/*
2700 * drop the reference count on the tree rooted at 'snap'.  This traverses
2701 * the tree freeing any blocks that have a ref count of zero after being
2702 * decremented.
2703 */
2704static int walk_log_tree(struct btrfs_trans_handle *trans,
2705			 struct btrfs_root *log, struct walk_control *wc)
2706{
2707	int ret = 0;
2708	int wret;
2709	int level;
2710	struct btrfs_path *path;
2711	int orig_level;
2712
2713	path = btrfs_alloc_path();
2714	if (!path)
2715		return -ENOMEM;
2716
2717	level = btrfs_header_level(log->node);
2718	orig_level = level;
2719	path->nodes[level] = log->node;
2720	atomic_inc(&log->node->refs);
2721	path->slots[level] = 0;
2722
2723	while (1) {
2724		wret = walk_down_log_tree(trans, log, path, &level, wc);
2725		if (wret > 0)
2726			break;
2727		if (wret < 0) {
2728			ret = wret;
2729			goto out;
2730		}
2731
2732		wret = walk_up_log_tree(trans, log, path, &level, wc);
2733		if (wret > 0)
2734			break;
2735		if (wret < 0) {
2736			ret = wret;
2737			goto out;
2738		}
2739	}
2740
2741	/* was the root node processed? if not, catch it here */
2742	if (path->nodes[orig_level]) {
2743		ret = wc->process_func(log, path->nodes[orig_level], wc,
2744			 btrfs_header_generation(path->nodes[orig_level]),
2745			 orig_level);
2746		if (ret)
2747			goto out;
2748		if (wc->free)
2749			ret = clean_log_buffer(trans, path->nodes[orig_level]);
2750	}
2751
2752out:
2753	btrfs_free_path(path);
2754	return ret;
2755}
2756
2757/*
2758 * helper function to update the item for a given subvolumes log root
2759 * in the tree of log roots
2760 */
2761static int update_log_root(struct btrfs_trans_handle *trans,
2762			   struct btrfs_root *log,
2763			   struct btrfs_root_item *root_item)
2764{
2765	struct btrfs_fs_info *fs_info = log->fs_info;
2766	int ret;
2767
2768	if (log->log_transid == 1) {
2769		/* insert root item on the first sync */
2770		ret = btrfs_insert_root(trans, fs_info->log_root_tree,
2771				&log->root_key, root_item);
2772	} else {
2773		ret = btrfs_update_root(trans, fs_info->log_root_tree,
2774				&log->root_key, root_item);
2775	}
2776	return ret;
2777}
2778
2779static void wait_log_commit(struct btrfs_root *root, int transid)
2780{
2781	DEFINE_WAIT(wait);
2782	int index = transid % 2;
2783
2784	/*
2785	 * we only allow two pending log transactions at a time,
2786	 * so we know that if ours is more than 2 older than the
2787	 * current transaction, we're done
2788	 */
2789	for (;;) {
2790		prepare_to_wait(&root->log_commit_wait[index],
2791				&wait, TASK_UNINTERRUPTIBLE);
2792
2793		if (!(root->log_transid_committed < transid &&
2794		      atomic_read(&root->log_commit[index])))
2795			break;
2796
2797		mutex_unlock(&root->log_mutex);
2798		schedule();
2799		mutex_lock(&root->log_mutex);
2800	}
2801	finish_wait(&root->log_commit_wait[index], &wait);
2802}
2803
2804static void wait_for_writer(struct btrfs_root *root)
2805{
2806	DEFINE_WAIT(wait);
2807
2808	for (;;) {
2809		prepare_to_wait(&root->log_writer_wait, &wait,
2810				TASK_UNINTERRUPTIBLE);
2811		if (!atomic_read(&root->log_writers))
2812			break;
2813
2814		mutex_unlock(&root->log_mutex);
2815		schedule();
2816		mutex_lock(&root->log_mutex);
2817	}
2818	finish_wait(&root->log_writer_wait, &wait);
2819}
2820
2821void btrfs_init_log_ctx(struct btrfs_log_ctx *ctx, struct inode *inode)
2822{
2823	ctx->log_ret = 0;
2824	ctx->log_transid = 0;
2825	ctx->log_new_dentries = false;
2826	ctx->logging_new_name = false;
2827	ctx->logging_new_delayed_dentries = false;
2828	ctx->logged_before = false;
2829	ctx->inode = inode;
2830	INIT_LIST_HEAD(&ctx->list);
2831	INIT_LIST_HEAD(&ctx->ordered_extents);
2832	INIT_LIST_HEAD(&ctx->conflict_inodes);
2833	ctx->num_conflict_inodes = 0;
2834	ctx->logging_conflict_inodes = false;
2835	ctx->scratch_eb = NULL;
2836}
2837
2838void btrfs_init_log_ctx_scratch_eb(struct btrfs_log_ctx *ctx)
2839{
2840	struct btrfs_inode *inode = BTRFS_I(ctx->inode);
2841
2842	if (!test_bit(BTRFS_INODE_NEEDS_FULL_SYNC, &inode->runtime_flags) &&
2843	    !test_bit(BTRFS_INODE_COPY_EVERYTHING, &inode->runtime_flags))
2844		return;
2845
2846	/*
2847	 * Don't care about allocation failure. This is just for optimization,
2848	 * if we fail to allocate here, we will try again later if needed.
2849	 */
2850	ctx->scratch_eb = alloc_dummy_extent_buffer(inode->root->fs_info, 0);
2851}
2852
2853void btrfs_release_log_ctx_extents(struct btrfs_log_ctx *ctx)
2854{
2855	struct btrfs_ordered_extent *ordered;
2856	struct btrfs_ordered_extent *tmp;
2857
2858	ASSERT(inode_is_locked(ctx->inode));
2859
2860	list_for_each_entry_safe(ordered, tmp, &ctx->ordered_extents, log_list) {
2861		list_del_init(&ordered->log_list);
2862		btrfs_put_ordered_extent(ordered);
2863	}
2864}
2865
2866
2867static inline void btrfs_remove_log_ctx(struct btrfs_root *root,
2868					struct btrfs_log_ctx *ctx)
2869{
2870	mutex_lock(&root->log_mutex);
2871	list_del_init(&ctx->list);
2872	mutex_unlock(&root->log_mutex);
2873}
2874
2875/* 
2876 * Invoked in log mutex context, or be sure there is no other task which
2877 * can access the list.
2878 */
2879static inline void btrfs_remove_all_log_ctxs(struct btrfs_root *root,
2880					     int index, int error)
2881{
2882	struct btrfs_log_ctx *ctx;
2883	struct btrfs_log_ctx *safe;
2884
2885	list_for_each_entry_safe(ctx, safe, &root->log_ctxs[index], list) {
2886		list_del_init(&ctx->list);
2887		ctx->log_ret = error;
2888	}
2889}
2890
2891/*
2892 * Sends a given tree log down to the disk and updates the super blocks to
2893 * record it.  When this call is done, you know that any inodes previously
2894 * logged are safely on disk only if it returns 0.
2895 *
2896 * Any other return value means you need to call btrfs_commit_transaction.
2897 * Some of the edge cases for fsyncing directories that have had unlinks
2898 * or renames done in the past mean that sometimes the only safe
2899 * fsync is to commit the whole FS.  When btrfs_sync_log returns -EAGAIN,
2900 * that has happened.
2901 */
2902int btrfs_sync_log(struct btrfs_trans_handle *trans,
2903		   struct btrfs_root *root, struct btrfs_log_ctx *ctx)
2904{
2905	int index1;
2906	int index2;
2907	int mark;
2908	int ret;
2909	struct btrfs_fs_info *fs_info = root->fs_info;
2910	struct btrfs_root *log = root->log_root;
2911	struct btrfs_root *log_root_tree = fs_info->log_root_tree;
2912	struct btrfs_root_item new_root_item;
2913	int log_transid = 0;
2914	struct btrfs_log_ctx root_log_ctx;
2915	struct blk_plug plug;
2916	u64 log_root_start;
2917	u64 log_root_level;
2918
2919	mutex_lock(&root->log_mutex);
2920	log_transid = ctx->log_transid;
2921	if (root->log_transid_committed >= log_transid) {
2922		mutex_unlock(&root->log_mutex);
2923		return ctx->log_ret;
2924	}
2925
2926	index1 = log_transid % 2;
2927	if (atomic_read(&root->log_commit[index1])) {
2928		wait_log_commit(root, log_transid);
2929		mutex_unlock(&root->log_mutex);
2930		return ctx->log_ret;
2931	}
2932	ASSERT(log_transid == root->log_transid);
2933	atomic_set(&root->log_commit[index1], 1);
2934
2935	/* wait for previous tree log sync to complete */
2936	if (atomic_read(&root->log_commit[(index1 + 1) % 2]))
2937		wait_log_commit(root, log_transid - 1);
2938
2939	while (1) {
2940		int batch = atomic_read(&root->log_batch);
2941		/* when we're on an ssd, just kick the log commit out */
2942		if (!btrfs_test_opt(fs_info, SSD) &&
2943		    test_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state)) {
2944			mutex_unlock(&root->log_mutex);
2945			schedule_timeout_uninterruptible(1);
2946			mutex_lock(&root->log_mutex);
2947		}
2948		wait_for_writer(root);
2949		if (batch == atomic_read(&root->log_batch))
2950			break;
2951	}
2952
2953	/* bail out if we need to do a full commit */
2954	if (btrfs_need_log_full_commit(trans)) {
2955		ret = BTRFS_LOG_FORCE_COMMIT;
2956		mutex_unlock(&root->log_mutex);
2957		goto out;
2958	}
2959
2960	if (log_transid % 2 == 0)
2961		mark = EXTENT_DIRTY;
2962	else
2963		mark = EXTENT_NEW;
2964
2965	/* we start IO on  all the marked extents here, but we don't actually
2966	 * wait for them until later.
2967	 */
2968	blk_start_plug(&plug);
2969	ret = btrfs_write_marked_extents(fs_info, &log->dirty_log_pages, mark);
2970	/*
2971	 * -EAGAIN happens when someone, e.g., a concurrent transaction
2972	 *  commit, writes a dirty extent in this tree-log commit. This
2973	 *  concurrent write will create a hole writing out the extents,
2974	 *  and we cannot proceed on a zoned filesystem, requiring
2975	 *  sequential writing. While we can bail out to a full commit
2976	 *  here, but we can continue hoping the concurrent writing fills
2977	 *  the hole.
2978	 */
2979	if (ret == -EAGAIN && btrfs_is_zoned(fs_info))
2980		ret = 0;
2981	if (ret) {
2982		blk_finish_plug(&plug);
2983		btrfs_set_log_full_commit(trans);
2984		mutex_unlock(&root->log_mutex);
2985		goto out;
2986	}
2987
2988	/*
2989	 * We _must_ update under the root->log_mutex in order to make sure we
2990	 * have a consistent view of the log root we are trying to commit at
2991	 * this moment.
2992	 *
2993	 * We _must_ copy this into a local copy, because we are not holding the
2994	 * log_root_tree->log_mutex yet.  This is important because when we
2995	 * commit the log_root_tree we must have a consistent view of the
2996	 * log_root_tree when we update the super block to point at the
2997	 * log_root_tree bytenr.  If we update the log_root_tree here we'll race
2998	 * with the commit and possibly point at the new block which we may not
2999	 * have written out.
3000	 */
3001	btrfs_set_root_node(&log->root_item, log->node);
3002	memcpy(&new_root_item, &log->root_item, sizeof(new_root_item));
3003
3004	btrfs_set_root_log_transid(root, root->log_transid + 1);
3005	log->log_transid = root->log_transid;
3006	root->log_start_pid = 0;
3007	/*
3008	 * IO has been started, blocks of the log tree have WRITTEN flag set
3009	 * in their headers. new modifications of the log will be written to
3010	 * new positions. so it's safe to allow log writers to go in.
3011	 */
3012	mutex_unlock(&root->log_mutex);
3013
3014	if (btrfs_is_zoned(fs_info)) {
3015		mutex_lock(&fs_info->tree_root->log_mutex);
3016		if (!log_root_tree->node) {
3017			ret = btrfs_alloc_log_tree_node(trans, log_root_tree);
3018			if (ret) {
3019				mutex_unlock(&fs_info->tree_root->log_mutex);
3020				blk_finish_plug(&plug);
3021				goto out;
3022			}
3023		}
3024		mutex_unlock(&fs_info->tree_root->log_mutex);
3025	}
3026
3027	btrfs_init_log_ctx(&root_log_ctx, NULL);
3028
3029	mutex_lock(&log_root_tree->log_mutex);
3030
3031	index2 = log_root_tree->log_transid % 2;
3032	list_add_tail(&root_log_ctx.list, &log_root_tree->log_ctxs[index2]);
3033	root_log_ctx.log_transid = log_root_tree->log_transid;
3034
3035	/*
3036	 * Now we are safe to update the log_root_tree because we're under the
3037	 * log_mutex, and we're a current writer so we're holding the commit
3038	 * open until we drop the log_mutex.
3039	 */
3040	ret = update_log_root(trans, log, &new_root_item);
3041	if (ret) {
3042		list_del_init(&root_log_ctx.list);
3043		blk_finish_plug(&plug);
3044		btrfs_set_log_full_commit(trans);
3045		if (ret != -ENOSPC)
3046			btrfs_err(fs_info,
3047				  "failed to update log for root %llu ret %d",
3048				  root->root_key.objectid, ret);
3049		btrfs_wait_tree_log_extents(log, mark);
3050		mutex_unlock(&log_root_tree->log_mutex);
3051		goto out;
3052	}
3053
3054	if (log_root_tree->log_transid_committed >= root_log_ctx.log_transid) {
3055		blk_finish_plug(&plug);
3056		list_del_init(&root_log_ctx.list);
3057		mutex_unlock(&log_root_tree->log_mutex);
3058		ret = root_log_ctx.log_ret;
3059		goto out;
3060	}
3061
3062	if (atomic_read(&log_root_tree->log_commit[index2])) {
3063		blk_finish_plug(&plug);
3064		ret = btrfs_wait_tree_log_extents(log, mark);
3065		wait_log_commit(log_root_tree,
3066				root_log_ctx.log_transid);
3067		mutex_unlock(&log_root_tree->log_mutex);
3068		if (!ret)
3069			ret = root_log_ctx.log_ret;
3070		goto out;
3071	}
3072	ASSERT(root_log_ctx.log_transid == log_root_tree->log_transid);
3073	atomic_set(&log_root_tree->log_commit[index2], 1);
3074
3075	if (atomic_read(&log_root_tree->log_commit[(index2 + 1) % 2])) {
3076		wait_log_commit(log_root_tree,
3077				root_log_ctx.log_transid - 1);
3078	}
3079
3080	/*
3081	 * now that we've moved on to the tree of log tree roots,
3082	 * check the full commit flag again
3083	 */
3084	if (btrfs_need_log_full_commit(trans)) {
3085		blk_finish_plug(&plug);
3086		btrfs_wait_tree_log_extents(log, mark);
3087		mutex_unlock(&log_root_tree->log_mutex);
3088		ret = BTRFS_LOG_FORCE_COMMIT;
3089		goto out_wake_log_root;
3090	}
3091
3092	ret = btrfs_write_marked_extents(fs_info,
3093					 &log_root_tree->dirty_log_pages,
3094					 EXTENT_DIRTY | EXTENT_NEW);
3095	blk_finish_plug(&plug);
3096	/*
3097	 * As described above, -EAGAIN indicates a hole in the extents. We
3098	 * cannot wait for these write outs since the waiting cause a
3099	 * deadlock. Bail out to the full commit instead.
3100	 */
3101	if (ret == -EAGAIN && btrfs_is_zoned(fs_info)) {
3102		btrfs_set_log_full_commit(trans);
3103		btrfs_wait_tree_log_extents(log, mark);
3104		mutex_unlock(&log_root_tree->log_mutex);
3105		goto out_wake_log_root;
3106	} else if (ret) {
3107		btrfs_set_log_full_commit(trans);
3108		mutex_unlock(&log_root_tree->log_mutex);
3109		goto out_wake_log_root;
3110	}
3111	ret = btrfs_wait_tree_log_extents(log, mark);
3112	if (!ret)
3113		ret = btrfs_wait_tree_log_extents(log_root_tree,
3114						  EXTENT_NEW | EXTENT_DIRTY);
3115	if (ret) {
3116		btrfs_set_log_full_commit(trans);
3117		mutex_unlock(&log_root_tree->log_mutex);
3118		goto out_wake_log_root;
3119	}
3120
3121	log_root_start = log_root_tree->node->start;
3122	log_root_level = btrfs_header_level(log_root_tree->node);
3123	log_root_tree->log_transid++;
3124	mutex_unlock(&log_root_tree->log_mutex);
3125
3126	/*
3127	 * Here we are guaranteed that nobody is going to write the superblock
3128	 * for the current transaction before us and that neither we do write
3129	 * our superblock before the previous transaction finishes its commit
3130	 * and writes its superblock, because:
3131	 *
3132	 * 1) We are holding a handle on the current transaction, so no body
3133	 *    can commit it until we release the handle;
3134	 *
3135	 * 2) Before writing our superblock we acquire the tree_log_mutex, so
3136	 *    if the previous transaction is still committing, and hasn't yet
3137	 *    written its superblock, we wait for it to do it, because a
3138	 *    transaction commit acquires the tree_log_mutex when the commit
3139	 *    begins and releases it only after writing its superblock.
3140	 */
3141	mutex_lock(&fs_info->tree_log_mutex);
3142
3143	/*
3144	 * The previous transaction writeout phase could have failed, and thus
3145	 * marked the fs in an error state.  We must not commit here, as we
3146	 * could have updated our generation in the super_for_commit and
3147	 * writing the super here would result in transid mismatches.  If there
3148	 * is an error here just bail.
3149	 */
3150	if (BTRFS_FS_ERROR(fs_info)) {
3151		ret = -EIO;
3152		btrfs_set_log_full_commit(trans);
3153		btrfs_abort_transaction(trans, ret);
3154		mutex_unlock(&fs_info->tree_log_mutex);
3155		goto out_wake_log_root;
3156	}
3157
3158	btrfs_set_super_log_root(fs_info->super_for_commit, log_root_start);
3159	btrfs_set_super_log_root_level(fs_info->super_for_commit, log_root_level);
3160	ret = write_all_supers(fs_info, 1);
3161	mutex_unlock(&fs_info->tree_log_mutex);
3162	if (ret) {
3163		btrfs_set_log_full_commit(trans);
3164		btrfs_abort_transaction(trans, ret);
3165		goto out_wake_log_root;
3166	}
3167
3168	/*
3169	 * We know there can only be one task here, since we have not yet set
3170	 * root->log_commit[index1] to 0 and any task attempting to sync the
3171	 * log must wait for the previous log transaction to commit if it's
3172	 * still in progress or wait for the current log transaction commit if
3173	 * someone else already started it. We use <= and not < because the
3174	 * first log transaction has an ID of 0.
3175	 */
3176	ASSERT(btrfs_get_root_last_log_commit(root) <= log_transid);
3177	btrfs_set_root_last_log_commit(root, log_transid);
3178
3179out_wake_log_root:
3180	mutex_lock(&log_root_tree->log_mutex);
3181	btrfs_remove_all_log_ctxs(log_root_tree, index2, ret);
3182
3183	log_root_tree->log_transid_committed++;
3184	atomic_set(&log_root_tree->log_commit[index2], 0);
3185	mutex_unlock(&log_root_tree->log_mutex);
3186
3187	/*
3188	 * The barrier before waitqueue_active (in cond_wake_up) is needed so
3189	 * all the updates above are seen by the woken threads. It might not be
3190	 * necessary, but proving that seems to be hard.
3191	 */
3192	cond_wake_up(&log_root_tree->log_commit_wait[index2]);
3193out:
3194	mutex_lock(&root->log_mutex);
3195	btrfs_remove_all_log_ctxs(root, index1, ret);
3196	root->log_transid_committed++;
3197	atomic_set(&root->log_commit[index1], 0);
3198	mutex_unlock(&root->log_mutex);
3199
3200	/*
3201	 * The barrier before waitqueue_active (in cond_wake_up) is needed so
3202	 * all the updates above are seen by the woken threads. It might not be
3203	 * necessary, but proving that seems to be hard.
3204	 */
3205	cond_wake_up(&root->log_commit_wait[index1]);
3206	return ret;
3207}
3208
3209static void free_log_tree(struct btrfs_trans_handle *trans,
3210			  struct btrfs_root *log)
3211{
3212	int ret;
3213	struct walk_control wc = {
3214		.free = 1,
3215		.process_func = process_one_buffer
3216	};
3217
3218	if (log->node) {
3219		ret = walk_log_tree(trans, log, &wc);
3220		if (ret) {
3221			/*
3222			 * We weren't able to traverse the entire log tree, the
3223			 * typical scenario is getting an -EIO when reading an
3224			 * extent buffer of the tree, due to a previous writeback
3225			 * failure of it.
3226			 */
3227			set_bit(BTRFS_FS_STATE_LOG_CLEANUP_ERROR,
3228				&log->fs_info->fs_state);
3229
3230			/*
3231			 * Some extent buffers of the log tree may still be dirty
3232			 * and not yet written back to storage, because we may
3233			 * have updates to a log tree without syncing a log tree,
3234			 * such as during rename and link operations. So flush
3235			 * them out and wait for their writeback to complete, so
3236			 * that we properly cleanup their state and pages.
3237			 */
3238			btrfs_write_marked_extents(log->fs_info,
3239						   &log->dirty_log_pages,
3240						   EXTENT_DIRTY | EXTENT_NEW);
3241			btrfs_wait_tree_log_extents(log,
3242						    EXTENT_DIRTY | EXTENT_NEW);
3243
3244			if (trans)
3245				btrfs_abort_transaction(trans, ret);
3246			else
3247				btrfs_handle_fs_error(log->fs_info, ret, NULL);
3248		}
3249	}
3250
3251	extent_io_tree_release(&log->dirty_log_pages);
3252	extent_io_tree_release(&log->log_csum_range);
3253
3254	btrfs_put_root(log);
3255}
3256
3257/*
3258 * free all the extents used by the tree log.  This should be called
3259 * at commit time of the full transaction
3260 */
3261int btrfs_free_log(struct btrfs_trans_handle *trans, struct btrfs_root *root)
3262{
3263	if (root->log_root) {
3264		free_log_tree(trans, root->log_root);
3265		root->log_root = NULL;
3266		clear_bit(BTRFS_ROOT_HAS_LOG_TREE, &root->state);
3267	}
3268	return 0;
3269}
3270
3271int btrfs_free_log_root_tree(struct btrfs_trans_handle *trans,
3272			     struct btrfs_fs_info *fs_info)
3273{
3274	if (fs_info->log_root_tree) {
3275		free_log_tree(trans, fs_info->log_root_tree);
3276		fs_info->log_root_tree = NULL;
3277		clear_bit(BTRFS_ROOT_HAS_LOG_TREE, &fs_info->tree_root->state);
3278	}
3279	return 0;
3280}
3281
3282/*
3283 * Check if an inode was logged in the current transaction. This correctly deals
3284 * with the case where the inode was logged but has a logged_trans of 0, which
3285 * happens if the inode is evicted and loaded again, as logged_trans is an in
3286 * memory only field (not persisted).
3287 *
3288 * Returns 1 if the inode was logged before in the transaction, 0 if it was not,
3289 * and < 0 on error.
3290 */
3291static int inode_logged(const struct btrfs_trans_handle *trans,
3292			struct btrfs_inode *inode,
3293			struct btrfs_path *path_in)
3294{
3295	struct btrfs_path *path = path_in;
3296	struct btrfs_key key;
3297	int ret;
3298
3299	if (inode->logged_trans == trans->transid)
3300		return 1;
3301
3302	/*
3303	 * If logged_trans is not 0, then we know the inode logged was not logged
3304	 * in this transaction, so we can return false right away.
3305	 */
3306	if (inode->logged_trans > 0)
3307		return 0;
3308
3309	/*
3310	 * If no log tree was created for this root in this transaction, then
3311	 * the inode can not have been logged in this transaction. In that case
3312	 * set logged_trans to anything greater than 0 and less than the current
3313	 * transaction's ID, to avoid the search below in a future call in case
3314	 * a log tree gets created after this.
3315	 */
3316	if (!test_bit(BTRFS_ROOT_HAS_LOG_TREE, &inode->root->state)) {
3317		inode->logged_trans = trans->transid - 1;
3318		return 0;
3319	}
3320
3321	/*
3322	 * We have a log tree and the inode's logged_trans is 0. We can't tell
3323	 * for sure if the inode was logged before in this transaction by looking
3324	 * only at logged_trans. We could be pessimistic and assume it was, but
3325	 * that can lead to unnecessarily logging an inode during rename and link
3326	 * operations, and then further updating the log in followup rename and
3327	 * link operations, specially if it's a directory, which adds latency
3328	 * visible to applications doing a series of rename or link operations.
3329	 *
3330	 * A logged_trans of 0 here can mean several things:
3331	 *
3332	 * 1) The inode was never logged since the filesystem was mounted, and may
3333	 *    or may have not been evicted and loaded again;
3334	 *
3335	 * 2) The inode was logged in a previous transaction, then evicted and
3336	 *    then loaded again;
3337	 *
3338	 * 3) The inode was logged in the current transaction, then evicted and
3339	 *    then loaded again.
3340	 *
3341	 * For cases 1) and 2) we don't want to return true, but we need to detect
3342	 * case 3) and return true. So we do a search in the log root for the inode
3343	 * item.
3344	 */
3345	key.objectid = btrfs_ino(inode);
3346	key.type = BTRFS_INODE_ITEM_KEY;
3347	key.offset = 0;
3348
3349	if (!path) {
3350		path = btrfs_alloc_path();
3351		if (!path)
3352			return -ENOMEM;
3353	}
3354
3355	ret = btrfs_search_slot(NULL, inode->root->log_root, &key, path, 0, 0);
3356
3357	if (path_in)
3358		btrfs_release_path(path);
3359	else
3360		btrfs_free_path(path);
3361
3362	/*
3363	 * Logging an inode always results in logging its inode item. So if we
3364	 * did not find the item we know the inode was not logged for sure.
3365	 */
3366	if (ret < 0) {
3367		return ret;
3368	} else if (ret > 0) {
3369		/*
3370		 * Set logged_trans to a value greater than 0 and less then the
3371		 * current transaction to avoid doing the search in future calls.
3372		 */
3373		inode->logged_trans = trans->transid - 1;
3374		return 0;
3375	}
3376
3377	/*
3378	 * The inode was previously logged and then evicted, set logged_trans to
3379	 * the current transacion's ID, to avoid future tree searches as long as
3380	 * the inode is not evicted again.
3381	 */
3382	inode->logged_trans = trans->transid;
3383
3384	/*
3385	 * If it's a directory, then we must set last_dir_index_offset to the
3386	 * maximum possible value, so that the next attempt to log the inode does
3387	 * not skip checking if dir index keys found in modified subvolume tree
3388	 * leaves have been logged before, otherwise it would result in attempts
3389	 * to insert duplicate dir index keys in the log tree. This must be done
3390	 * because last_dir_index_offset is an in-memory only field, not persisted
3391	 * in the inode item or any other on-disk structure, so its value is lost
3392	 * once the inode is evicted.
3393	 */
3394	if (S_ISDIR(inode->vfs_inode.i_mode))
3395		inode->last_dir_index_offset = (u64)-1;
3396
3397	return 1;
3398}
3399
3400/*
3401 * Delete a directory entry from the log if it exists.
3402 *
3403 * Returns < 0 on error
3404 *           1 if the entry does not exists
3405 *           0 if the entry existed and was successfully deleted
3406 */
3407static int del_logged_dentry(struct btrfs_trans_handle *trans,
3408			     struct btrfs_root *log,
3409			     struct btrfs_path *path,
3410			     u64 dir_ino,
3411			     const struct fscrypt_str *name,
3412			     u64 index)
3413{
3414	struct btrfs_dir_item *di;
3415
3416	/*
3417	 * We only log dir index items of a directory, so we don't need to look
3418	 * for dir item keys.
3419	 */
3420	di = btrfs_lookup_dir_index_item(trans, log, path, dir_ino,
3421					 index, name, -1);
3422	if (IS_ERR(di))
3423		return PTR_ERR(di);
3424	else if (!di)
3425		return 1;
3426
3427	/*
3428	 * We do not need to update the size field of the directory's
3429	 * inode item because on log replay we update the field to reflect
3430	 * all existing entries in the directory (see overwrite_item()).
3431	 */
3432	return btrfs_delete_one_dir_name(trans, log, path, di);
3433}
3434
3435/*
3436 * If both a file and directory are logged, and unlinks or renames are
3437 * mixed in, we have a few interesting corners:
3438 *
3439 * create file X in dir Y
3440 * link file X to X.link in dir Y
3441 * fsync file X
3442 * unlink file X but leave X.link
3443 * fsync dir Y
3444 *
3445 * After a crash we would expect only X.link to exist.  But file X
3446 * didn't get fsync'd again so the log has back refs for X and X.link.
3447 *
3448 * We solve this by removing directory entries and inode backrefs from the
3449 * log when a file that was logged in the current transaction is
3450 * unlinked.  Any later fsync will include the updated log entries, and
3451 * we'll be able to reconstruct the proper directory items from backrefs.
3452 *
3453 * This optimizations allows us to avoid relogging the entire inode
3454 * or the entire directory.
3455 */
3456void btrfs_del_dir_entries_in_log(struct btrfs_trans_handle *trans,
3457				  struct btrfs_root *root,
3458				  const struct fscrypt_str *name,
3459				  struct btrfs_inode *dir, u64 index)
3460{
3461	struct btrfs_path *path;
3462	int ret;
3463
3464	ret = inode_logged(trans, dir, NULL);
3465	if (ret == 0)
3466		return;
3467	else if (ret < 0) {
3468		btrfs_set_log_full_commit(trans);
3469		return;
3470	}
3471
3472	ret = join_running_log_trans(root);
3473	if (ret)
3474		return;
3475
3476	mutex_lock(&dir->log_mutex);
3477
3478	path = btrfs_alloc_path();
3479	if (!path) {
3480		ret = -ENOMEM;
3481		goto out_unlock;
3482	}
3483
3484	ret = del_logged_dentry(trans, root->log_root, path, btrfs_ino(dir),
3485				name, index);
3486	btrfs_free_path(path);
3487out_unlock:
3488	mutex_unlock(&dir->log_mutex);
3489	if (ret < 0)
3490		btrfs_set_log_full_commit(trans);
3491	btrfs_end_log_trans(root);
3492}
3493
3494/* see comments for btrfs_del_dir_entries_in_log */
3495void btrfs_del_inode_ref_in_log(struct btrfs_trans_handle *trans,
3496				struct btrfs_root *root,
3497				const struct fscrypt_str *name,
3498				struct btrfs_inode *inode, u64 dirid)
3499{
3500	struct btrfs_root *log;
3501	u64 index;
3502	int ret;
3503
3504	ret = inode_logged(trans, inode, NULL);
3505	if (ret == 0)
3506		return;
3507	else if (ret < 0) {
3508		btrfs_set_log_full_commit(trans);
3509		return;
3510	}
3511
3512	ret = join_running_log_trans(root);
3513	if (ret)
3514		return;
3515	log = root->log_root;
3516	mutex_lock(&inode->log_mutex);
3517
3518	ret = btrfs_del_inode_ref(trans, log, name, btrfs_ino(inode),
3519				  dirid, &index);
3520	mutex_unlock(&inode->log_mutex);
3521	if (ret < 0 && ret != -ENOENT)
3522		btrfs_set_log_full_commit(trans);
3523	btrfs_end_log_trans(root);
3524}
3525
3526/*
3527 * creates a range item in the log for 'dirid'.  first_offset and
3528 * last_offset tell us which parts of the key space the log should
3529 * be considered authoritative for.
3530 */
3531static noinline int insert_dir_log_key(struct btrfs_trans_handle *trans,
3532				       struct btrfs_root *log,
3533				       struct btrfs_path *path,
3534				       u64 dirid,
3535				       u64 first_offset, u64 last_offset)
3536{
3537	int ret;
3538	struct btrfs_key key;
3539	struct btrfs_dir_log_item *item;
3540
3541	key.objectid = dirid;
3542	key.offset = first_offset;
3543	key.type = BTRFS_DIR_LOG_INDEX_KEY;
3544	ret = btrfs_insert_empty_item(trans, log, path, &key, sizeof(*item));
3545	/*
3546	 * -EEXIST is fine and can happen sporadically when we are logging a
3547	 * directory and have concurrent insertions in the subvolume's tree for
3548	 * items from other inodes and that result in pushing off some dir items
3549	 * from one leaf to another in order to accommodate for the new items.
3550	 * This results in logging the same dir index range key.
3551	 */
3552	if (ret && ret != -EEXIST)
3553		return ret;
3554
3555	item = btrfs_item_ptr(path->nodes[0], path->slots[0],
3556			      struct btrfs_dir_log_item);
3557	if (ret == -EEXIST) {
3558		const u64 curr_end = btrfs_dir_log_end(path->nodes[0], item);
3559
3560		/*
3561		 * btrfs_del_dir_entries_in_log() might have been called during
3562		 * an unlink between the initial insertion of this key and the
3563		 * current update, or we might be logging a single entry deletion
3564		 * during a rename, so set the new last_offset to the max value.
3565		 */
3566		last_offset = max(last_offset, curr_end);
3567	}
3568	btrfs_set_dir_log_end(path->nodes[0], item, last_offset);
3569	btrfs_mark_buffer_dirty(trans, path->nodes[0]);
3570	btrfs_release_path(path);
3571	return 0;
3572}
3573
3574static int flush_dir_items_batch(struct btrfs_trans_handle *trans,
3575				 struct btrfs_inode *inode,
3576				 struct extent_buffer *src,
3577				 struct btrfs_path *dst_path,
3578				 int start_slot,
3579				 int count)
3580{
3581	struct btrfs_root *log = inode->root->log_root;
3582	char *ins_data = NULL;
3583	struct btrfs_item_batch batch;
3584	struct extent_buffer *dst;
3585	unsigned long src_offset;
3586	unsigned long dst_offset;
3587	u64 last_index;
3588	struct btrfs_key key;
3589	u32 item_size;
3590	int ret;
3591	int i;
3592
3593	ASSERT(count > 0);
3594	batch.nr = count;
3595
3596	if (count == 1) {
3597		btrfs_item_key_to_cpu(src, &key, start_slot);
3598		item_size = btrfs_item_size(src, start_slot);
3599		batch.keys = &key;
3600		batch.data_sizes = &item_size;
3601		batch.total_data_size = item_size;
3602	} else {
3603		struct btrfs_key *ins_keys;
3604		u32 *ins_sizes;
3605
3606		ins_data = kmalloc(count * sizeof(u32) +
3607				   count * sizeof(struct btrfs_key), GFP_NOFS);
3608		if (!ins_data)
3609			return -ENOMEM;
3610
3611		ins_sizes = (u32 *)ins_data;
3612		ins_keys = (struct btrfs_key *)(ins_data + count * sizeof(u32));
3613		batch.keys = ins_keys;
3614		batch.data_sizes = ins_sizes;
3615		batch.total_data_size = 0;
3616
3617		for (i = 0; i < count; i++) {
3618			const int slot = start_slot + i;
3619
3620			btrfs_item_key_to_cpu(src, &ins_keys[i], slot);
3621			ins_sizes[i] = btrfs_item_size(src, slot);
3622			batch.total_data_size += ins_sizes[i];
3623		}
3624	}
3625
3626	ret = btrfs_insert_empty_items(trans, log, dst_path, &batch);
3627	if (ret)
3628		goto out;
3629
3630	dst = dst_path->nodes[0];
3631	/*
3632	 * Copy all the items in bulk, in a single copy operation. Item data is
3633	 * organized such that it's placed at the end of a leaf and from right
3634	 * to left. For example, the data for the second item ends at an offset
3635	 * that matches the offset where the data for the first item starts, the
3636	 * data for the third item ends at an offset that matches the offset
3637	 * where the data of the second items starts, and so on.
3638	 * Therefore our source and destination start offsets for copy match the
3639	 * offsets of the last items (highest slots).
3640	 */
3641	dst_offset = btrfs_item_ptr_offset(dst, dst_path->slots[0] + count - 1);
3642	src_offset = btrfs_item_ptr_offset(src, start_slot + count - 1);
3643	copy_extent_buffer(dst, src, dst_offset, src_offset, batch.total_data_size);
3644	btrfs_release_path(dst_path);
3645
3646	last_index = batch.keys[count - 1].offset;
3647	ASSERT(last_index > inode->last_dir_index_offset);
3648
3649	/*
3650	 * If for some unexpected reason the last item's index is not greater
3651	 * than the last index we logged, warn and force a transaction commit.
3652	 */
3653	if (WARN_ON(last_index <= inode->last_dir_index_offset))
3654		ret = BTRFS_LOG_FORCE_COMMIT;
3655	else
3656		inode->last_dir_index_offset = last_index;
3657
3658	if (btrfs_get_first_dir_index_to_log(inode) == 0)
3659		btrfs_set_first_dir_index_to_log(inode, batch.keys[0].offset);
3660out:
3661	kfree(ins_data);
3662
3663	return ret;
3664}
3665
3666static int clone_leaf(struct btrfs_path *path, struct btrfs_log_ctx *ctx)
3667{
3668	const int slot = path->slots[0];
3669
3670	if (ctx->scratch_eb) {
3671		copy_extent_buffer_full(ctx->scratch_eb, path->nodes[0]);
3672	} else {
3673		ctx->scratch_eb = btrfs_clone_extent_buffer(path->nodes[0]);
3674		if (!ctx->scratch_eb)
3675			return -ENOMEM;
3676	}
3677
3678	btrfs_release_path(path);
3679	path->nodes[0] = ctx->scratch_eb;
3680	path->slots[0] = slot;
3681	/*
3682	 * Add extra ref to scratch eb so that it is not freed when callers
3683	 * release the path, so we can reuse it later if needed.
3684	 */
3685	atomic_inc(&ctx->scratch_eb->refs);
3686
3687	return 0;
3688}
3689
3690static int process_dir_items_leaf(struct btrfs_trans_handle *trans,
3691				  struct btrfs_inode *inode,
3692				  struct btrfs_path *path,
3693				  struct btrfs_path *dst_path,
3694				  struct btrfs_log_ctx *ctx,
3695				  u64 *last_old_dentry_offset)
3696{
3697	struct btrfs_root *log = inode->root->log_root;
3698	struct extent_buffer *src;
3699	const int nritems = btrfs_header_nritems(path->nodes[0]);
3700	const u64 ino = btrfs_ino(inode);
3701	bool last_found = false;
3702	int batch_start = 0;
3703	int batch_size = 0;
3704	int ret;
3705
3706	/*
3707	 * We need to clone the leaf, release the read lock on it, and use the
3708	 * clone before modifying the log tree. See the comment at copy_items()
3709	 * about why we need to do this.
3710	 */
3711	ret = clone_leaf(path, ctx);
3712	if (ret < 0)
3713		return ret;
3714
3715	src = path->nodes[0];
 
 
 
3716
3717	for (int i = path->slots[0]; i < nritems; i++) {
3718		struct btrfs_dir_item *di;
3719		struct btrfs_key key;
3720		int ret;
3721
3722		btrfs_item_key_to_cpu(src, &key, i);
3723
3724		if (key.objectid != ino || key.type != BTRFS_DIR_INDEX_KEY) {
3725			last_found = true;
3726			break;
3727		}
3728
3729		di = btrfs_item_ptr(src, i, struct btrfs_dir_item);
3730
3731		/*
3732		 * Skip ranges of items that consist only of dir item keys created
3733		 * in past transactions. However if we find a gap, we must log a
3734		 * dir index range item for that gap, so that index keys in that
3735		 * gap are deleted during log replay.
3736		 */
3737		if (btrfs_dir_transid(src, di) < trans->transid) {
3738			if (key.offset > *last_old_dentry_offset + 1) {
3739				ret = insert_dir_log_key(trans, log, dst_path,
3740						 ino, *last_old_dentry_offset + 1,
3741						 key.offset - 1);
3742				if (ret < 0)
3743					return ret;
3744			}
3745
3746			*last_old_dentry_offset = key.offset;
3747			continue;
3748		}
3749
3750		/* If we logged this dir index item before, we can skip it. */
3751		if (key.offset <= inode->last_dir_index_offset)
3752			continue;
3753
3754		/*
3755		 * We must make sure that when we log a directory entry, the
3756		 * corresponding inode, after log replay, has a matching link
3757		 * count. For example:
3758		 *
3759		 * touch foo
3760		 * mkdir mydir
3761		 * sync
3762		 * ln foo mydir/bar
3763		 * xfs_io -c "fsync" mydir
3764		 * <crash>
3765		 * <mount fs and log replay>
3766		 *
3767		 * Would result in a fsync log that when replayed, our file inode
3768		 * would have a link count of 1, but we get two directory entries
3769		 * pointing to the same inode. After removing one of the names,
3770		 * it would not be possible to remove the other name, which
3771		 * resulted always in stale file handle errors, and would not be
3772		 * possible to rmdir the parent directory, since its i_size could
3773		 * never be decremented to the value BTRFS_EMPTY_DIR_SIZE,
3774		 * resulting in -ENOTEMPTY errors.
3775		 */
3776		if (!ctx->log_new_dentries) {
3777			struct btrfs_key di_key;
3778
3779			btrfs_dir_item_key_to_cpu(src, di, &di_key);
3780			if (di_key.type != BTRFS_ROOT_ITEM_KEY)
3781				ctx->log_new_dentries = true;
3782		}
3783
3784		if (batch_size == 0)
3785			batch_start = i;
3786		batch_size++;
3787	}
3788
3789	if (batch_size > 0) {
3790		int ret;
3791
3792		ret = flush_dir_items_batch(trans, inode, src, dst_path,
3793					    batch_start, batch_size);
3794		if (ret < 0)
3795			return ret;
3796	}
3797
3798	return last_found ? 1 : 0;
3799}
3800
3801/*
3802 * log all the items included in the current transaction for a given
3803 * directory.  This also creates the range items in the log tree required
3804 * to replay anything deleted before the fsync
3805 */
3806static noinline int log_dir_items(struct btrfs_trans_handle *trans,
3807			  struct btrfs_inode *inode,
3808			  struct btrfs_path *path,
3809			  struct btrfs_path *dst_path,
3810			  struct btrfs_log_ctx *ctx,
3811			  u64 min_offset, u64 *last_offset_ret)
3812{
3813	struct btrfs_key min_key;
3814	struct btrfs_root *root = inode->root;
3815	struct btrfs_root *log = root->log_root;
3816	int ret;
3817	u64 last_old_dentry_offset = min_offset - 1;
3818	u64 last_offset = (u64)-1;
3819	u64 ino = btrfs_ino(inode);
3820
3821	min_key.objectid = ino;
3822	min_key.type = BTRFS_DIR_INDEX_KEY;
3823	min_key.offset = min_offset;
3824
3825	ret = btrfs_search_forward(root, &min_key, path, trans->transid);
3826
3827	/*
3828	 * we didn't find anything from this transaction, see if there
3829	 * is anything at all
3830	 */
3831	if (ret != 0 || min_key.objectid != ino ||
3832	    min_key.type != BTRFS_DIR_INDEX_KEY) {
3833		min_key.objectid = ino;
3834		min_key.type = BTRFS_DIR_INDEX_KEY;
3835		min_key.offset = (u64)-1;
3836		btrfs_release_path(path);
3837		ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0);
3838		if (ret < 0) {
3839			btrfs_release_path(path);
3840			return ret;
3841		}
3842		ret = btrfs_previous_item(root, path, ino, BTRFS_DIR_INDEX_KEY);
3843
3844		/* if ret == 0 there are items for this type,
3845		 * create a range to tell us the last key of this type.
3846		 * otherwise, there are no items in this directory after
3847		 * *min_offset, and we create a range to indicate that.
3848		 */
3849		if (ret == 0) {
3850			struct btrfs_key tmp;
3851
3852			btrfs_item_key_to_cpu(path->nodes[0], &tmp,
3853					      path->slots[0]);
3854			if (tmp.type == BTRFS_DIR_INDEX_KEY)
3855				last_old_dentry_offset = tmp.offset;
3856		} else if (ret > 0) {
3857			ret = 0;
3858		}
3859
3860		goto done;
3861	}
3862
3863	/* go backward to find any previous key */
3864	ret = btrfs_previous_item(root, path, ino, BTRFS_DIR_INDEX_KEY);
3865	if (ret == 0) {
3866		struct btrfs_key tmp;
3867
3868		btrfs_item_key_to_cpu(path->nodes[0], &tmp, path->slots[0]);
3869		/*
3870		 * The dir index key before the first one we found that needs to
3871		 * be logged might be in a previous leaf, and there might be a
3872		 * gap between these keys, meaning that we had deletions that
3873		 * happened. So the key range item we log (key type
3874		 * BTRFS_DIR_LOG_INDEX_KEY) must cover a range that starts at the
3875		 * previous key's offset plus 1, so that those deletes are replayed.
3876		 */
3877		if (tmp.type == BTRFS_DIR_INDEX_KEY)
3878			last_old_dentry_offset = tmp.offset;
3879	} else if (ret < 0) {
3880		goto done;
3881	}
3882
3883	btrfs_release_path(path);
3884
3885	/*
3886	 * Find the first key from this transaction again or the one we were at
3887	 * in the loop below in case we had to reschedule. We may be logging the
3888	 * directory without holding its VFS lock, which happen when logging new
3889	 * dentries (through log_new_dir_dentries()) or in some cases when we
3890	 * need to log the parent directory of an inode. This means a dir index
3891	 * key might be deleted from the inode's root, and therefore we may not
3892	 * find it anymore. If we can't find it, just move to the next key. We
3893	 * can not bail out and ignore, because if we do that we will simply
3894	 * not log dir index keys that come after the one that was just deleted
3895	 * and we can end up logging a dir index range that ends at (u64)-1
3896	 * (@last_offset is initialized to that), resulting in removing dir
3897	 * entries we should not remove at log replay time.
3898	 */
3899search:
3900	ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0);
3901	if (ret > 0) {
3902		ret = btrfs_next_item(root, path);
3903		if (ret > 0) {
3904			/* There are no more keys in the inode's root. */
3905			ret = 0;
3906			goto done;
3907		}
3908	}
3909	if (ret < 0)
3910		goto done;
3911
3912	/*
3913	 * we have a block from this transaction, log every item in it
3914	 * from our directory
3915	 */
3916	while (1) {
3917		ret = process_dir_items_leaf(trans, inode, path, dst_path, ctx,
3918					     &last_old_dentry_offset);
3919		if (ret != 0) {
3920			if (ret > 0)
3921				ret = 0;
3922			goto done;
3923		}
3924		path->slots[0] = btrfs_header_nritems(path->nodes[0]);
3925
3926		/*
3927		 * look ahead to the next item and see if it is also
3928		 * from this directory and from this transaction
3929		 */
3930		ret = btrfs_next_leaf(root, path);
3931		if (ret) {
3932			if (ret == 1) {
3933				last_offset = (u64)-1;
3934				ret = 0;
3935			}
3936			goto done;
3937		}
3938		btrfs_item_key_to_cpu(path->nodes[0], &min_key, path->slots[0]);
3939		if (min_key.objectid != ino || min_key.type != BTRFS_DIR_INDEX_KEY) {
3940			last_offset = (u64)-1;
3941			goto done;
3942		}
3943		if (btrfs_header_generation(path->nodes[0]) != trans->transid) {
3944			/*
3945			 * The next leaf was not changed in the current transaction
3946			 * and has at least one dir index key.
3947			 * We check for the next key because there might have been
3948			 * one or more deletions between the last key we logged and
3949			 * that next key. So the key range item we log (key type
3950			 * BTRFS_DIR_LOG_INDEX_KEY) must end at the next key's
3951			 * offset minus 1, so that those deletes are replayed.
3952			 */
3953			last_offset = min_key.offset - 1;
3954			goto done;
3955		}
3956		if (need_resched()) {
3957			btrfs_release_path(path);
3958			cond_resched();
3959			goto search;
3960		}
3961	}
3962done:
3963	btrfs_release_path(path);
3964	btrfs_release_path(dst_path);
3965
3966	if (ret == 0) {
3967		*last_offset_ret = last_offset;
3968		/*
3969		 * In case the leaf was changed in the current transaction but
3970		 * all its dir items are from a past transaction, the last item
3971		 * in the leaf is a dir item and there's no gap between that last
3972		 * dir item and the first one on the next leaf (which did not
3973		 * change in the current transaction), then we don't need to log
3974		 * a range, last_old_dentry_offset is == to last_offset.
3975		 */
3976		ASSERT(last_old_dentry_offset <= last_offset);
3977		if (last_old_dentry_offset < last_offset)
3978			ret = insert_dir_log_key(trans, log, path, ino,
3979						 last_old_dentry_offset + 1,
3980						 last_offset);
3981	}
3982
3983	return ret;
3984}
3985
3986/*
3987 * If the inode was logged before and it was evicted, then its
3988 * last_dir_index_offset is (u64)-1, so we don't the value of the last index
3989 * key offset. If that's the case, search for it and update the inode. This
3990 * is to avoid lookups in the log tree every time we try to insert a dir index
3991 * key from a leaf changed in the current transaction, and to allow us to always
3992 * do batch insertions of dir index keys.
3993 */
3994static int update_last_dir_index_offset(struct btrfs_inode *inode,
3995					struct btrfs_path *path,
3996					const struct btrfs_log_ctx *ctx)
3997{
3998	const u64 ino = btrfs_ino(inode);
3999	struct btrfs_key key;
4000	int ret;
4001
4002	lockdep_assert_held(&inode->log_mutex);
4003
4004	if (inode->last_dir_index_offset != (u64)-1)
4005		return 0;
4006
4007	if (!ctx->logged_before) {
4008		inode->last_dir_index_offset = BTRFS_DIR_START_INDEX - 1;
4009		return 0;
4010	}
4011
4012	key.objectid = ino;
4013	key.type = BTRFS_DIR_INDEX_KEY;
4014	key.offset = (u64)-1;
4015
4016	ret = btrfs_search_slot(NULL, inode->root->log_root, &key, path, 0, 0);
4017	/*
4018	 * An error happened or we actually have an index key with an offset
4019	 * value of (u64)-1. Bail out, we're done.
4020	 */
4021	if (ret <= 0)
4022		goto out;
4023
4024	ret = 0;
4025	inode->last_dir_index_offset = BTRFS_DIR_START_INDEX - 1;
4026
4027	/*
4028	 * No dir index items, bail out and leave last_dir_index_offset with
4029	 * the value right before the first valid index value.
4030	 */
4031	if (path->slots[0] == 0)
4032		goto out;
4033
4034	/*
4035	 * btrfs_search_slot() left us at one slot beyond the slot with the last
4036	 * index key, or beyond the last key of the directory that is not an
4037	 * index key. If we have an index key before, set last_dir_index_offset
4038	 * to its offset value, otherwise leave it with a value right before the
4039	 * first valid index value, as it means we have an empty directory.
4040	 */
4041	btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0] - 1);
4042	if (key.objectid == ino && key.type == BTRFS_DIR_INDEX_KEY)
4043		inode->last_dir_index_offset = key.offset;
4044
4045out:
4046	btrfs_release_path(path);
4047
4048	return ret;
4049}
4050
4051/*
4052 * logging directories is very similar to logging inodes, We find all the items
4053 * from the current transaction and write them to the log.
4054 *
4055 * The recovery code scans the directory in the subvolume, and if it finds a
4056 * key in the range logged that is not present in the log tree, then it means
4057 * that dir entry was unlinked during the transaction.
4058 *
4059 * In order for that scan to work, we must include one key smaller than
4060 * the smallest logged by this transaction and one key larger than the largest
4061 * key logged by this transaction.
4062 */
4063static noinline int log_directory_changes(struct btrfs_trans_handle *trans,
4064			  struct btrfs_inode *inode,
4065			  struct btrfs_path *path,
4066			  struct btrfs_path *dst_path,
4067			  struct btrfs_log_ctx *ctx)
4068{
4069	u64 min_key;
4070	u64 max_key;
4071	int ret;
4072
4073	ret = update_last_dir_index_offset(inode, path, ctx);
4074	if (ret)
4075		return ret;
4076
4077	min_key = BTRFS_DIR_START_INDEX;
4078	max_key = 0;
4079
4080	while (1) {
4081		ret = log_dir_items(trans, inode, path, dst_path,
4082				ctx, min_key, &max_key);
4083		if (ret)
4084			return ret;
4085		if (max_key == (u64)-1)
4086			break;
4087		min_key = max_key + 1;
4088	}
4089
4090	return 0;
4091}
4092
4093/*
4094 * a helper function to drop items from the log before we relog an
4095 * inode.  max_key_type indicates the highest item type to remove.
4096 * This cannot be run for file data extents because it does not
4097 * free the extents they point to.
4098 */
4099static int drop_inode_items(struct btrfs_trans_handle *trans,
4100				  struct btrfs_root *log,
4101				  struct btrfs_path *path,
4102				  struct btrfs_inode *inode,
4103				  int max_key_type)
4104{
4105	int ret;
4106	struct btrfs_key key;
4107	struct btrfs_key found_key;
4108	int start_slot;
4109
4110	key.objectid = btrfs_ino(inode);
4111	key.type = max_key_type;
4112	key.offset = (u64)-1;
4113
4114	while (1) {
4115		ret = btrfs_search_slot(trans, log, &key, path, -1, 1);
4116		if (ret < 0) {
4117			break;
4118		} else if (ret > 0) {
4119			if (path->slots[0] == 0)
4120				break;
4121			path->slots[0]--;
4122		}
4123
4124		btrfs_item_key_to_cpu(path->nodes[0], &found_key,
4125				      path->slots[0]);
4126
4127		if (found_key.objectid != key.objectid)
4128			break;
4129
4130		found_key.offset = 0;
4131		found_key.type = 0;
4132		ret = btrfs_bin_search(path->nodes[0], 0, &found_key, &start_slot);
4133		if (ret < 0)
4134			break;
4135
4136		ret = btrfs_del_items(trans, log, path, start_slot,
4137				      path->slots[0] - start_slot + 1);
4138		/*
4139		 * If start slot isn't 0 then we don't need to re-search, we've
4140		 * found the last guy with the objectid in this tree.
4141		 */
4142		if (ret || start_slot != 0)
4143			break;
4144		btrfs_release_path(path);
4145	}
4146	btrfs_release_path(path);
4147	if (ret > 0)
4148		ret = 0;
4149	return ret;
4150}
4151
4152static int truncate_inode_items(struct btrfs_trans_handle *trans,
4153				struct btrfs_root *log_root,
4154				struct btrfs_inode *inode,
4155				u64 new_size, u32 min_type)
4156{
4157	struct btrfs_truncate_control control = {
4158		.new_size = new_size,
4159		.ino = btrfs_ino(inode),
4160		.min_type = min_type,
4161		.skip_ref_updates = true,
4162	};
4163
4164	return btrfs_truncate_inode_items(trans, log_root, &control);
4165}
4166
4167static void fill_inode_item(struct btrfs_trans_handle *trans,
4168			    struct extent_buffer *leaf,
4169			    struct btrfs_inode_item *item,
4170			    struct inode *inode, int log_inode_only,
4171			    u64 logged_isize)
4172{
4173	struct btrfs_map_token token;
4174	u64 flags;
4175
4176	btrfs_init_map_token(&token, leaf);
4177
4178	if (log_inode_only) {
4179		/* set the generation to zero so the recover code
4180		 * can tell the difference between an logging
4181		 * just to say 'this inode exists' and a logging
4182		 * to say 'update this inode with these values'
4183		 */
4184		btrfs_set_token_inode_generation(&token, item, 0);
4185		btrfs_set_token_inode_size(&token, item, logged_isize);
4186	} else {
4187		btrfs_set_token_inode_generation(&token, item,
4188						 BTRFS_I(inode)->generation);
4189		btrfs_set_token_inode_size(&token, item, inode->i_size);
4190	}
4191
4192	btrfs_set_token_inode_uid(&token, item, i_uid_read(inode));
4193	btrfs_set_token_inode_gid(&token, item, i_gid_read(inode));
4194	btrfs_set_token_inode_mode(&token, item, inode->i_mode);
4195	btrfs_set_token_inode_nlink(&token, item, inode->i_nlink);
4196
4197	btrfs_set_token_timespec_sec(&token, &item->atime,
4198				     inode_get_atime_sec(inode));
4199	btrfs_set_token_timespec_nsec(&token, &item->atime,
4200				      inode_get_atime_nsec(inode));
4201
4202	btrfs_set_token_timespec_sec(&token, &item->mtime,
4203				     inode_get_mtime_sec(inode));
4204	btrfs_set_token_timespec_nsec(&token, &item->mtime,
4205				      inode_get_mtime_nsec(inode));
4206
4207	btrfs_set_token_timespec_sec(&token, &item->ctime,
4208				     inode_get_ctime_sec(inode));
4209	btrfs_set_token_timespec_nsec(&token, &item->ctime,
4210				      inode_get_ctime_nsec(inode));
4211
4212	/*
4213	 * We do not need to set the nbytes field, in fact during a fast fsync
4214	 * its value may not even be correct, since a fast fsync does not wait
4215	 * for ordered extent completion, which is where we update nbytes, it
4216	 * only waits for writeback to complete. During log replay as we find
4217	 * file extent items and replay them, we adjust the nbytes field of the
4218	 * inode item in subvolume tree as needed (see overwrite_item()).
4219	 */
4220
4221	btrfs_set_token_inode_sequence(&token, item, inode_peek_iversion(inode));
4222	btrfs_set_token_inode_transid(&token, item, trans->transid);
4223	btrfs_set_token_inode_rdev(&token, item, inode->i_rdev);
4224	flags = btrfs_inode_combine_flags(BTRFS_I(inode)->flags,
4225					  BTRFS_I(inode)->ro_flags);
4226	btrfs_set_token_inode_flags(&token, item, flags);
4227	btrfs_set_token_inode_block_group(&token, item, 0);
4228}
4229
4230static int log_inode_item(struct btrfs_trans_handle *trans,
4231			  struct btrfs_root *log, struct btrfs_path *path,
4232			  struct btrfs_inode *inode, bool inode_item_dropped)
4233{
4234	struct btrfs_inode_item *inode_item;
4235	int ret;
4236
4237	/*
4238	 * If we are doing a fast fsync and the inode was logged before in the
4239	 * current transaction, then we know the inode was previously logged and
4240	 * it exists in the log tree. For performance reasons, in this case use
4241	 * btrfs_search_slot() directly with ins_len set to 0 so that we never
4242	 * attempt a write lock on the leaf's parent, which adds unnecessary lock
4243	 * contention in case there are concurrent fsyncs for other inodes of the
4244	 * same subvolume. Using btrfs_insert_empty_item() when the inode item
4245	 * already exists can also result in unnecessarily splitting a leaf.
4246	 */
4247	if (!inode_item_dropped && inode->logged_trans == trans->transid) {
4248		ret = btrfs_search_slot(trans, log, &inode->location, path, 0, 1);
4249		ASSERT(ret <= 0);
4250		if (ret > 0)
4251			ret = -ENOENT;
4252	} else {
4253		/*
4254		 * This means it is the first fsync in the current transaction,
4255		 * so the inode item is not in the log and we need to insert it.
4256		 * We can never get -EEXIST because we are only called for a fast
4257		 * fsync and in case an inode eviction happens after the inode was
4258		 * logged before in the current transaction, when we load again
4259		 * the inode, we set BTRFS_INODE_NEEDS_FULL_SYNC on its runtime
4260		 * flags and set ->logged_trans to 0.
4261		 */
4262		ret = btrfs_insert_empty_item(trans, log, path, &inode->location,
4263					      sizeof(*inode_item));
4264		ASSERT(ret != -EEXIST);
4265	}
4266	if (ret)
4267		return ret;
4268	inode_item = btrfs_item_ptr(path->nodes[0], path->slots[0],
4269				    struct btrfs_inode_item);
4270	fill_inode_item(trans, path->nodes[0], inode_item, &inode->vfs_inode,
4271			0, 0);
4272	btrfs_release_path(path);
4273	return 0;
4274}
4275
4276static int log_csums(struct btrfs_trans_handle *trans,
4277		     struct btrfs_inode *inode,
4278		     struct btrfs_root *log_root,
4279		     struct btrfs_ordered_sum *sums)
4280{
4281	const u64 lock_end = sums->logical + sums->len - 1;
4282	struct extent_state *cached_state = NULL;
4283	int ret;
4284
4285	/*
4286	 * If this inode was not used for reflink operations in the current
4287	 * transaction with new extents, then do the fast path, no need to
4288	 * worry about logging checksum items with overlapping ranges.
4289	 */
4290	if (inode->last_reflink_trans < trans->transid)
4291		return btrfs_csum_file_blocks(trans, log_root, sums);
4292
4293	/*
4294	 * Serialize logging for checksums. This is to avoid racing with the
4295	 * same checksum being logged by another task that is logging another
4296	 * file which happens to refer to the same extent as well. Such races
4297	 * can leave checksum items in the log with overlapping ranges.
4298	 */
4299	ret = lock_extent(&log_root->log_csum_range, sums->logical, lock_end,
4300			  &cached_state);
4301	if (ret)
4302		return ret;
4303	/*
4304	 * Due to extent cloning, we might have logged a csum item that covers a
4305	 * subrange of a cloned extent, and later we can end up logging a csum
4306	 * item for a larger subrange of the same extent or the entire range.
4307	 * This would leave csum items in the log tree that cover the same range
4308	 * and break the searches for checksums in the log tree, resulting in
4309	 * some checksums missing in the fs/subvolume tree. So just delete (or
4310	 * trim and adjust) any existing csum items in the log for this range.
4311	 */
4312	ret = btrfs_del_csums(trans, log_root, sums->logical, sums->len);
4313	if (!ret)
4314		ret = btrfs_csum_file_blocks(trans, log_root, sums);
4315
4316	unlock_extent(&log_root->log_csum_range, sums->logical, lock_end,
4317		      &cached_state);
4318
4319	return ret;
4320}
4321
4322static noinline int copy_items(struct btrfs_trans_handle *trans,
4323			       struct btrfs_inode *inode,
4324			       struct btrfs_path *dst_path,
4325			       struct btrfs_path *src_path,
4326			       int start_slot, int nr, int inode_only,
4327			       u64 logged_isize, struct btrfs_log_ctx *ctx)
4328{
4329	struct btrfs_root *log = inode->root->log_root;
4330	struct btrfs_file_extent_item *extent;
4331	struct extent_buffer *src;
4332	int ret;
4333	struct btrfs_key *ins_keys;
4334	u32 *ins_sizes;
4335	struct btrfs_item_batch batch;
4336	char *ins_data;
 
4337	int dst_index;
4338	const bool skip_csum = (inode->flags & BTRFS_INODE_NODATASUM);
4339	const u64 i_size = i_size_read(&inode->vfs_inode);
4340
4341	/*
4342	 * To keep lockdep happy and avoid deadlocks, clone the source leaf and
4343	 * use the clone. This is because otherwise we would be changing the log
4344	 * tree, to insert items from the subvolume tree or insert csum items,
4345	 * while holding a read lock on a leaf from the subvolume tree, which
4346	 * creates a nasty lock dependency when COWing log tree nodes/leaves:
4347	 *
4348	 * 1) Modifying the log tree triggers an extent buffer allocation while
4349	 *    holding a write lock on a parent extent buffer from the log tree.
4350	 *    Allocating the pages for an extent buffer, or the extent buffer
4351	 *    struct, can trigger inode eviction and finally the inode eviction
4352	 *    will trigger a release/remove of a delayed node, which requires
4353	 *    taking the delayed node's mutex;
4354	 *
4355	 * 2) Allocating a metadata extent for a log tree can trigger the async
4356	 *    reclaim thread and make us wait for it to release enough space and
4357	 *    unblock our reservation ticket. The reclaim thread can start
4358	 *    flushing delayed items, and that in turn results in the need to
4359	 *    lock delayed node mutexes and in the need to write lock extent
4360	 *    buffers of a subvolume tree - all this while holding a write lock
4361	 *    on the parent extent buffer in the log tree.
4362	 *
4363	 * So one task in scenario 1) running in parallel with another task in
4364	 * scenario 2) could lead to a deadlock, one wanting to lock a delayed
4365	 * node mutex while having a read lock on a leaf from the subvolume,
4366	 * while the other is holding the delayed node's mutex and wants to
4367	 * write lock the same subvolume leaf for flushing delayed items.
4368	 */
4369	ret = clone_leaf(src_path, ctx);
4370	if (ret < 0)
4371		return ret;
4372
4373	src = src_path->nodes[0];
 
 
 
4374
4375	ins_data = kmalloc(nr * sizeof(struct btrfs_key) +
4376			   nr * sizeof(u32), GFP_NOFS);
4377	if (!ins_data)
4378		return -ENOMEM;
4379
4380	ins_sizes = (u32 *)ins_data;
4381	ins_keys = (struct btrfs_key *)(ins_data + nr * sizeof(u32));
4382	batch.keys = ins_keys;
4383	batch.data_sizes = ins_sizes;
4384	batch.total_data_size = 0;
4385	batch.nr = 0;
4386
4387	dst_index = 0;
4388	for (int i = 0; i < nr; i++) {
4389		const int src_slot = start_slot + i;
4390		struct btrfs_root *csum_root;
4391		struct btrfs_ordered_sum *sums;
4392		struct btrfs_ordered_sum *sums_next;
4393		LIST_HEAD(ordered_sums);
4394		u64 disk_bytenr;
4395		u64 disk_num_bytes;
4396		u64 extent_offset;
4397		u64 extent_num_bytes;
4398		bool is_old_extent;
4399
4400		btrfs_item_key_to_cpu(src, &ins_keys[dst_index], src_slot);
4401
4402		if (ins_keys[dst_index].type != BTRFS_EXTENT_DATA_KEY)
4403			goto add_to_batch;
4404
4405		extent = btrfs_item_ptr(src, src_slot,
4406					struct btrfs_file_extent_item);
4407
4408		is_old_extent = (btrfs_file_extent_generation(src, extent) <
4409				 trans->transid);
4410
4411		/*
4412		 * Don't copy extents from past generations. That would make us
4413		 * log a lot more metadata for common cases like doing only a
4414		 * few random writes into a file and then fsync it for the first
4415		 * time or after the full sync flag is set on the inode. We can
4416		 * get leaves full of extent items, most of which are from past
4417		 * generations, so we can skip them - as long as the inode has
4418		 * not been the target of a reflink operation in this transaction,
4419		 * as in that case it might have had file extent items with old
4420		 * generations copied into it. We also must always log prealloc
4421		 * extents that start at or beyond eof, otherwise we would lose
4422		 * them on log replay.
4423		 */
4424		if (is_old_extent &&
4425		    ins_keys[dst_index].offset < i_size &&
4426		    inode->last_reflink_trans < trans->transid)
4427			continue;
4428
4429		if (skip_csum)
4430			goto add_to_batch;
4431
4432		/* Only regular extents have checksums. */
4433		if (btrfs_file_extent_type(src, extent) != BTRFS_FILE_EXTENT_REG)
4434			goto add_to_batch;
4435
4436		/*
4437		 * If it's an extent created in a past transaction, then its
4438		 * checksums are already accessible from the committed csum tree,
4439		 * no need to log them.
4440		 */
4441		if (is_old_extent)
4442			goto add_to_batch;
4443
4444		disk_bytenr = btrfs_file_extent_disk_bytenr(src, extent);
4445		/* If it's an explicit hole, there are no checksums. */
4446		if (disk_bytenr == 0)
4447			goto add_to_batch;
4448
4449		disk_num_bytes = btrfs_file_extent_disk_num_bytes(src, extent);
4450
4451		if (btrfs_file_extent_compression(src, extent)) {
4452			extent_offset = 0;
4453			extent_num_bytes = disk_num_bytes;
4454		} else {
4455			extent_offset = btrfs_file_extent_offset(src, extent);
4456			extent_num_bytes = btrfs_file_extent_num_bytes(src, extent);
4457		}
4458
4459		csum_root = btrfs_csum_root(trans->fs_info, disk_bytenr);
4460		disk_bytenr += extent_offset;
4461		ret = btrfs_lookup_csums_list(csum_root, disk_bytenr,
4462					      disk_bytenr + extent_num_bytes - 1,
4463					      &ordered_sums, 0, false);
4464		if (ret)
4465			goto out;
4466
4467		list_for_each_entry_safe(sums, sums_next, &ordered_sums, list) {
4468			if (!ret)
4469				ret = log_csums(trans, inode, log, sums);
4470			list_del(&sums->list);
4471			kfree(sums);
4472		}
4473		if (ret)
4474			goto out;
4475
4476add_to_batch:
4477		ins_sizes[dst_index] = btrfs_item_size(src, src_slot);
4478		batch.total_data_size += ins_sizes[dst_index];
4479		batch.nr++;
4480		dst_index++;
4481	}
4482
4483	/*
4484	 * We have a leaf full of old extent items that don't need to be logged,
4485	 * so we don't need to do anything.
4486	 */
4487	if (batch.nr == 0)
4488		goto out;
4489
4490	ret = btrfs_insert_empty_items(trans, log, dst_path, &batch);
4491	if (ret)
4492		goto out;
4493
4494	dst_index = 0;
4495	for (int i = 0; i < nr; i++) {
4496		const int src_slot = start_slot + i;
4497		const int dst_slot = dst_path->slots[0] + dst_index;
4498		struct btrfs_key key;
4499		unsigned long src_offset;
4500		unsigned long dst_offset;
4501
4502		/*
4503		 * We're done, all the remaining items in the source leaf
4504		 * correspond to old file extent items.
4505		 */
4506		if (dst_index >= batch.nr)
4507			break;
4508
4509		btrfs_item_key_to_cpu(src, &key, src_slot);
4510
4511		if (key.type != BTRFS_EXTENT_DATA_KEY)
4512			goto copy_item;
4513
4514		extent = btrfs_item_ptr(src, src_slot,
4515					struct btrfs_file_extent_item);
4516
4517		/* See the comment in the previous loop, same logic. */
4518		if (btrfs_file_extent_generation(src, extent) < trans->transid &&
4519		    key.offset < i_size &&
4520		    inode->last_reflink_trans < trans->transid)
4521			continue;
4522
4523copy_item:
4524		dst_offset = btrfs_item_ptr_offset(dst_path->nodes[0], dst_slot);
4525		src_offset = btrfs_item_ptr_offset(src, src_slot);
4526
4527		if (key.type == BTRFS_INODE_ITEM_KEY) {
4528			struct btrfs_inode_item *inode_item;
4529
4530			inode_item = btrfs_item_ptr(dst_path->nodes[0], dst_slot,
4531						    struct btrfs_inode_item);
4532			fill_inode_item(trans, dst_path->nodes[0], inode_item,
4533					&inode->vfs_inode,
4534					inode_only == LOG_INODE_EXISTS,
4535					logged_isize);
4536		} else {
4537			copy_extent_buffer(dst_path->nodes[0], src, dst_offset,
4538					   src_offset, ins_sizes[dst_index]);
4539		}
4540
4541		dst_index++;
4542	}
4543
4544	btrfs_mark_buffer_dirty(trans, dst_path->nodes[0]);
4545	btrfs_release_path(dst_path);
4546out:
4547	kfree(ins_data);
4548
4549	return ret;
4550}
4551
4552static int extent_cmp(void *priv, const struct list_head *a,
4553		      const struct list_head *b)
4554{
4555	const struct extent_map *em1, *em2;
4556
4557	em1 = list_entry(a, struct extent_map, list);
4558	em2 = list_entry(b, struct extent_map, list);
4559
4560	if (em1->start < em2->start)
4561		return -1;
4562	else if (em1->start > em2->start)
4563		return 1;
4564	return 0;
4565}
4566
4567static int log_extent_csums(struct btrfs_trans_handle *trans,
4568			    struct btrfs_inode *inode,
4569			    struct btrfs_root *log_root,
4570			    const struct extent_map *em,
4571			    struct btrfs_log_ctx *ctx)
4572{
4573	struct btrfs_ordered_extent *ordered;
4574	struct btrfs_root *csum_root;
4575	u64 csum_offset;
4576	u64 csum_len;
4577	u64 mod_start = em->mod_start;
4578	u64 mod_len = em->mod_len;
4579	LIST_HEAD(ordered_sums);
4580	int ret = 0;
4581
4582	if (inode->flags & BTRFS_INODE_NODATASUM ||
4583	    (em->flags & EXTENT_FLAG_PREALLOC) ||
4584	    em->block_start == EXTENT_MAP_HOLE)
4585		return 0;
4586
4587	list_for_each_entry(ordered, &ctx->ordered_extents, log_list) {
4588		const u64 ordered_end = ordered->file_offset + ordered->num_bytes;
4589		const u64 mod_end = mod_start + mod_len;
4590		struct btrfs_ordered_sum *sums;
4591
4592		if (mod_len == 0)
4593			break;
4594
4595		if (ordered_end <= mod_start)
4596			continue;
4597		if (mod_end <= ordered->file_offset)
4598			break;
4599
4600		/*
4601		 * We are going to copy all the csums on this ordered extent, so
4602		 * go ahead and adjust mod_start and mod_len in case this ordered
4603		 * extent has already been logged.
4604		 */
4605		if (ordered->file_offset > mod_start) {
4606			if (ordered_end >= mod_end)
4607				mod_len = ordered->file_offset - mod_start;
4608			/*
4609			 * If we have this case
4610			 *
4611			 * |--------- logged extent ---------|
4612			 *       |----- ordered extent ----|
4613			 *
4614			 * Just don't mess with mod_start and mod_len, we'll
4615			 * just end up logging more csums than we need and it
4616			 * will be ok.
4617			 */
4618		} else {
4619			if (ordered_end < mod_end) {
4620				mod_len = mod_end - ordered_end;
4621				mod_start = ordered_end;
4622			} else {
4623				mod_len = 0;
4624			}
4625		}
4626
4627		/*
4628		 * To keep us from looping for the above case of an ordered
4629		 * extent that falls inside of the logged extent.
4630		 */
4631		if (test_and_set_bit(BTRFS_ORDERED_LOGGED_CSUM, &ordered->flags))
4632			continue;
4633
4634		list_for_each_entry(sums, &ordered->list, list) {
4635			ret = log_csums(trans, inode, log_root, sums);
4636			if (ret)
4637				return ret;
4638		}
4639	}
4640
4641	/* We're done, found all csums in the ordered extents. */
4642	if (mod_len == 0)
4643		return 0;
4644
4645	/* If we're compressed we have to save the entire range of csums. */
4646	if (extent_map_is_compressed(em)) {
4647		csum_offset = 0;
4648		csum_len = max(em->block_len, em->orig_block_len);
4649	} else {
4650		csum_offset = mod_start - em->start;
4651		csum_len = mod_len;
4652	}
4653
4654	/* block start is already adjusted for the file extent offset. */
4655	csum_root = btrfs_csum_root(trans->fs_info, em->block_start);
4656	ret = btrfs_lookup_csums_list(csum_root, em->block_start + csum_offset,
4657				      em->block_start + csum_offset +
4658				      csum_len - 1, &ordered_sums, 0, false);
4659	if (ret)
4660		return ret;
4661
4662	while (!list_empty(&ordered_sums)) {
4663		struct btrfs_ordered_sum *sums = list_entry(ordered_sums.next,
4664						   struct btrfs_ordered_sum,
4665						   list);
4666		if (!ret)
4667			ret = log_csums(trans, inode, log_root, sums);
4668		list_del(&sums->list);
4669		kfree(sums);
4670	}
4671
4672	return ret;
4673}
4674
4675static int log_one_extent(struct btrfs_trans_handle *trans,
4676			  struct btrfs_inode *inode,
4677			  const struct extent_map *em,
4678			  struct btrfs_path *path,
4679			  struct btrfs_log_ctx *ctx)
4680{
4681	struct btrfs_drop_extents_args drop_args = { 0 };
4682	struct btrfs_root *log = inode->root->log_root;
4683	struct btrfs_file_extent_item fi = { 0 };
4684	struct extent_buffer *leaf;
4685	struct btrfs_key key;
4686	enum btrfs_compression_type compress_type;
4687	u64 extent_offset = em->start - em->orig_start;
4688	u64 block_len;
4689	int ret;
4690
4691	btrfs_set_stack_file_extent_generation(&fi, trans->transid);
4692	if (em->flags & EXTENT_FLAG_PREALLOC)
4693		btrfs_set_stack_file_extent_type(&fi, BTRFS_FILE_EXTENT_PREALLOC);
4694	else
4695		btrfs_set_stack_file_extent_type(&fi, BTRFS_FILE_EXTENT_REG);
4696
4697	block_len = max(em->block_len, em->orig_block_len);
4698	compress_type = extent_map_compression(em);
4699	if (compress_type != BTRFS_COMPRESS_NONE) {
4700		btrfs_set_stack_file_extent_disk_bytenr(&fi, em->block_start);
4701		btrfs_set_stack_file_extent_disk_num_bytes(&fi, block_len);
4702	} else if (em->block_start < EXTENT_MAP_LAST_BYTE) {
4703		btrfs_set_stack_file_extent_disk_bytenr(&fi, em->block_start -
4704							extent_offset);
4705		btrfs_set_stack_file_extent_disk_num_bytes(&fi, block_len);
4706	}
4707
4708	btrfs_set_stack_file_extent_offset(&fi, extent_offset);
4709	btrfs_set_stack_file_extent_num_bytes(&fi, em->len);
4710	btrfs_set_stack_file_extent_ram_bytes(&fi, em->ram_bytes);
4711	btrfs_set_stack_file_extent_compression(&fi, compress_type);
4712
4713	ret = log_extent_csums(trans, inode, log, em, ctx);
4714	if (ret)
4715		return ret;
4716
4717	/*
4718	 * If this is the first time we are logging the inode in the current
4719	 * transaction, we can avoid btrfs_drop_extents(), which is expensive
4720	 * because it does a deletion search, which always acquires write locks
4721	 * for extent buffers at levels 2, 1 and 0. This not only wastes time
4722	 * but also adds significant contention in a log tree, since log trees
4723	 * are small, with a root at level 2 or 3 at most, due to their short
4724	 * life span.
4725	 */
4726	if (ctx->logged_before) {
4727		drop_args.path = path;
4728		drop_args.start = em->start;
4729		drop_args.end = em->start + em->len;
4730		drop_args.replace_extent = true;
4731		drop_args.extent_item_size = sizeof(fi);
4732		ret = btrfs_drop_extents(trans, log, inode, &drop_args);
4733		if (ret)
4734			return ret;
4735	}
4736
4737	if (!drop_args.extent_inserted) {
4738		key.objectid = btrfs_ino(inode);
4739		key.type = BTRFS_EXTENT_DATA_KEY;
4740		key.offset = em->start;
4741
4742		ret = btrfs_insert_empty_item(trans, log, path, &key,
4743					      sizeof(fi));
4744		if (ret)
4745			return ret;
4746	}
4747	leaf = path->nodes[0];
4748	write_extent_buffer(leaf, &fi,
4749			    btrfs_item_ptr_offset(leaf, path->slots[0]),
4750			    sizeof(fi));
4751	btrfs_mark_buffer_dirty(trans, leaf);
4752
4753	btrfs_release_path(path);
4754
4755	return ret;
4756}
4757
4758/*
4759 * Log all prealloc extents beyond the inode's i_size to make sure we do not
4760 * lose them after doing a full/fast fsync and replaying the log. We scan the
4761 * subvolume's root instead of iterating the inode's extent map tree because
4762 * otherwise we can log incorrect extent items based on extent map conversion.
4763 * That can happen due to the fact that extent maps are merged when they
4764 * are not in the extent map tree's list of modified extents.
4765 */
4766static int btrfs_log_prealloc_extents(struct btrfs_trans_handle *trans,
4767				      struct btrfs_inode *inode,
4768				      struct btrfs_path *path,
4769				      struct btrfs_log_ctx *ctx)
4770{
4771	struct btrfs_root *root = inode->root;
4772	struct btrfs_key key;
4773	const u64 i_size = i_size_read(&inode->vfs_inode);
4774	const u64 ino = btrfs_ino(inode);
4775	struct btrfs_path *dst_path = NULL;
4776	bool dropped_extents = false;
4777	u64 truncate_offset = i_size;
4778	struct extent_buffer *leaf;
4779	int slot;
4780	int ins_nr = 0;
4781	int start_slot = 0;
4782	int ret;
4783
4784	if (!(inode->flags & BTRFS_INODE_PREALLOC))
4785		return 0;
4786
4787	key.objectid = ino;
4788	key.type = BTRFS_EXTENT_DATA_KEY;
4789	key.offset = i_size;
4790	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
4791	if (ret < 0)
4792		goto out;
4793
4794	/*
4795	 * We must check if there is a prealloc extent that starts before the
4796	 * i_size and crosses the i_size boundary. This is to ensure later we
4797	 * truncate down to the end of that extent and not to the i_size, as
4798	 * otherwise we end up losing part of the prealloc extent after a log
4799	 * replay and with an implicit hole if there is another prealloc extent
4800	 * that starts at an offset beyond i_size.
4801	 */
4802	ret = btrfs_previous_item(root, path, ino, BTRFS_EXTENT_DATA_KEY);
4803	if (ret < 0)
4804		goto out;
4805
4806	if (ret == 0) {
4807		struct btrfs_file_extent_item *ei;
4808
4809		leaf = path->nodes[0];
4810		slot = path->slots[0];
4811		ei = btrfs_item_ptr(leaf, slot, struct btrfs_file_extent_item);
4812
4813		if (btrfs_file_extent_type(leaf, ei) ==
4814		    BTRFS_FILE_EXTENT_PREALLOC) {
4815			u64 extent_end;
4816
4817			btrfs_item_key_to_cpu(leaf, &key, slot);
4818			extent_end = key.offset +
4819				btrfs_file_extent_num_bytes(leaf, ei);
4820
4821			if (extent_end > i_size)
4822				truncate_offset = extent_end;
4823		}
4824	} else {
4825		ret = 0;
4826	}
4827
4828	while (true) {
4829		leaf = path->nodes[0];
4830		slot = path->slots[0];
4831
4832		if (slot >= btrfs_header_nritems(leaf)) {
4833			if (ins_nr > 0) {
4834				ret = copy_items(trans, inode, dst_path, path,
4835						 start_slot, ins_nr, 1, 0, ctx);
4836				if (ret < 0)
4837					goto out;
4838				ins_nr = 0;
4839			}
4840			ret = btrfs_next_leaf(root, path);
4841			if (ret < 0)
4842				goto out;
4843			if (ret > 0) {
4844				ret = 0;
4845				break;
4846			}
4847			continue;
4848		}
4849
4850		btrfs_item_key_to_cpu(leaf, &key, slot);
4851		if (key.objectid > ino)
4852			break;
4853		if (WARN_ON_ONCE(key.objectid < ino) ||
4854		    key.type < BTRFS_EXTENT_DATA_KEY ||
4855		    key.offset < i_size) {
4856			path->slots[0]++;
4857			continue;
4858		}
4859		if (!dropped_extents) {
4860			/*
4861			 * Avoid logging extent items logged in past fsync calls
4862			 * and leading to duplicate keys in the log tree.
4863			 */
4864			ret = truncate_inode_items(trans, root->log_root, inode,
4865						   truncate_offset,
4866						   BTRFS_EXTENT_DATA_KEY);
4867			if (ret)
4868				goto out;
4869			dropped_extents = true;
4870		}
4871		if (ins_nr == 0)
4872			start_slot = slot;
4873		ins_nr++;
4874		path->slots[0]++;
4875		if (!dst_path) {
4876			dst_path = btrfs_alloc_path();
4877			if (!dst_path) {
4878				ret = -ENOMEM;
4879				goto out;
4880			}
4881		}
4882	}
4883	if (ins_nr > 0)
4884		ret = copy_items(trans, inode, dst_path, path,
4885				 start_slot, ins_nr, 1, 0, ctx);
4886out:
4887	btrfs_release_path(path);
4888	btrfs_free_path(dst_path);
4889	return ret;
4890}
4891
4892static int btrfs_log_changed_extents(struct btrfs_trans_handle *trans,
4893				     struct btrfs_inode *inode,
4894				     struct btrfs_path *path,
4895				     struct btrfs_log_ctx *ctx)
4896{
4897	struct btrfs_ordered_extent *ordered;
4898	struct btrfs_ordered_extent *tmp;
4899	struct extent_map *em, *n;
4900	LIST_HEAD(extents);
4901	struct extent_map_tree *tree = &inode->extent_tree;
4902	int ret = 0;
4903	int num = 0;
4904
4905	write_lock(&tree->lock);
4906
4907	list_for_each_entry_safe(em, n, &tree->modified_extents, list) {
4908		list_del_init(&em->list);
4909		/*
4910		 * Just an arbitrary number, this can be really CPU intensive
4911		 * once we start getting a lot of extents, and really once we
4912		 * have a bunch of extents we just want to commit since it will
4913		 * be faster.
4914		 */
4915		if (++num > 32768) {
4916			list_del_init(&tree->modified_extents);
4917			ret = -EFBIG;
4918			goto process;
4919		}
4920
4921		if (em->generation < trans->transid)
4922			continue;
4923
4924		/* We log prealloc extents beyond eof later. */
4925		if ((em->flags & EXTENT_FLAG_PREALLOC) &&
4926		    em->start >= i_size_read(&inode->vfs_inode))
4927			continue;
4928
4929		/* Need a ref to keep it from getting evicted from cache */
4930		refcount_inc(&em->refs);
4931		em->flags |= EXTENT_FLAG_LOGGING;
4932		list_add_tail(&em->list, &extents);
4933		num++;
4934	}
4935
4936	list_sort(NULL, &extents, extent_cmp);
4937process:
4938	while (!list_empty(&extents)) {
4939		em = list_entry(extents.next, struct extent_map, list);
4940
4941		list_del_init(&em->list);
4942
4943		/*
4944		 * If we had an error we just need to delete everybody from our
4945		 * private list.
4946		 */
4947		if (ret) {
4948			clear_em_logging(tree, em);
4949			free_extent_map(em);
4950			continue;
4951		}
4952
4953		write_unlock(&tree->lock);
4954
4955		ret = log_one_extent(trans, inode, em, path, ctx);
4956		write_lock(&tree->lock);
4957		clear_em_logging(tree, em);
4958		free_extent_map(em);
4959	}
4960	WARN_ON(!list_empty(&extents));
4961	write_unlock(&tree->lock);
4962
4963	if (!ret)
4964		ret = btrfs_log_prealloc_extents(trans, inode, path, ctx);
4965	if (ret)
4966		return ret;
4967
4968	/*
4969	 * We have logged all extents successfully, now make sure the commit of
4970	 * the current transaction waits for the ordered extents to complete
4971	 * before it commits and wipes out the log trees, otherwise we would
4972	 * lose data if an ordered extents completes after the transaction
4973	 * commits and a power failure happens after the transaction commit.
4974	 */
4975	list_for_each_entry_safe(ordered, tmp, &ctx->ordered_extents, log_list) {
4976		list_del_init(&ordered->log_list);
4977		set_bit(BTRFS_ORDERED_LOGGED, &ordered->flags);
4978
4979		if (!test_bit(BTRFS_ORDERED_COMPLETE, &ordered->flags)) {
4980			spin_lock_irq(&inode->ordered_tree_lock);
4981			if (!test_bit(BTRFS_ORDERED_COMPLETE, &ordered->flags)) {
4982				set_bit(BTRFS_ORDERED_PENDING, &ordered->flags);
4983				atomic_inc(&trans->transaction->pending_ordered);
4984			}
4985			spin_unlock_irq(&inode->ordered_tree_lock);
4986		}
4987		btrfs_put_ordered_extent(ordered);
4988	}
4989
4990	return 0;
4991}
4992
4993static int logged_inode_size(struct btrfs_root *log, struct btrfs_inode *inode,
4994			     struct btrfs_path *path, u64 *size_ret)
4995{
4996	struct btrfs_key key;
4997	int ret;
4998
4999	key.objectid = btrfs_ino(inode);
5000	key.type = BTRFS_INODE_ITEM_KEY;
5001	key.offset = 0;
5002
5003	ret = btrfs_search_slot(NULL, log, &key, path, 0, 0);
5004	if (ret < 0) {
5005		return ret;
5006	} else if (ret > 0) {
5007		*size_ret = 0;
5008	} else {
5009		struct btrfs_inode_item *item;
5010
5011		item = btrfs_item_ptr(path->nodes[0], path->slots[0],
5012				      struct btrfs_inode_item);
5013		*size_ret = btrfs_inode_size(path->nodes[0], item);
5014		/*
5015		 * If the in-memory inode's i_size is smaller then the inode
5016		 * size stored in the btree, return the inode's i_size, so
5017		 * that we get a correct inode size after replaying the log
5018		 * when before a power failure we had a shrinking truncate
5019		 * followed by addition of a new name (rename / new hard link).
5020		 * Otherwise return the inode size from the btree, to avoid
5021		 * data loss when replaying a log due to previously doing a
5022		 * write that expands the inode's size and logging a new name
5023		 * immediately after.
5024		 */
5025		if (*size_ret > inode->vfs_inode.i_size)
5026			*size_ret = inode->vfs_inode.i_size;
5027	}
5028
5029	btrfs_release_path(path);
5030	return 0;
5031}
5032
5033/*
5034 * At the moment we always log all xattrs. This is to figure out at log replay
5035 * time which xattrs must have their deletion replayed. If a xattr is missing
5036 * in the log tree and exists in the fs/subvol tree, we delete it. This is
5037 * because if a xattr is deleted, the inode is fsynced and a power failure
5038 * happens, causing the log to be replayed the next time the fs is mounted,
5039 * we want the xattr to not exist anymore (same behaviour as other filesystems
5040 * with a journal, ext3/4, xfs, f2fs, etc).
5041 */
5042static int btrfs_log_all_xattrs(struct btrfs_trans_handle *trans,
5043				struct btrfs_inode *inode,
5044				struct btrfs_path *path,
5045				struct btrfs_path *dst_path,
5046				struct btrfs_log_ctx *ctx)
5047{
5048	struct btrfs_root *root = inode->root;
5049	int ret;
5050	struct btrfs_key key;
5051	const u64 ino = btrfs_ino(inode);
5052	int ins_nr = 0;
5053	int start_slot = 0;
5054	bool found_xattrs = false;
5055
5056	if (test_bit(BTRFS_INODE_NO_XATTRS, &inode->runtime_flags))
5057		return 0;
5058
5059	key.objectid = ino;
5060	key.type = BTRFS_XATTR_ITEM_KEY;
5061	key.offset = 0;
5062
5063	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
5064	if (ret < 0)
5065		return ret;
5066
5067	while (true) {
5068		int slot = path->slots[0];
5069		struct extent_buffer *leaf = path->nodes[0];
5070		int nritems = btrfs_header_nritems(leaf);
5071
5072		if (slot >= nritems) {
5073			if (ins_nr > 0) {
5074				ret = copy_items(trans, inode, dst_path, path,
5075						 start_slot, ins_nr, 1, 0, ctx);
5076				if (ret < 0)
5077					return ret;
5078				ins_nr = 0;
5079			}
5080			ret = btrfs_next_leaf(root, path);
5081			if (ret < 0)
5082				return ret;
5083			else if (ret > 0)
5084				break;
5085			continue;
5086		}
5087
5088		btrfs_item_key_to_cpu(leaf, &key, slot);
5089		if (key.objectid != ino || key.type != BTRFS_XATTR_ITEM_KEY)
5090			break;
5091
5092		if (ins_nr == 0)
5093			start_slot = slot;
5094		ins_nr++;
5095		path->slots[0]++;
5096		found_xattrs = true;
5097		cond_resched();
5098	}
5099	if (ins_nr > 0) {
5100		ret = copy_items(trans, inode, dst_path, path,
5101				 start_slot, ins_nr, 1, 0, ctx);
5102		if (ret < 0)
5103			return ret;
5104	}
5105
5106	if (!found_xattrs)
5107		set_bit(BTRFS_INODE_NO_XATTRS, &inode->runtime_flags);
5108
5109	return 0;
5110}
5111
5112/*
5113 * When using the NO_HOLES feature if we punched a hole that causes the
5114 * deletion of entire leafs or all the extent items of the first leaf (the one
5115 * that contains the inode item and references) we may end up not processing
5116 * any extents, because there are no leafs with a generation matching the
5117 * current transaction that have extent items for our inode. So we need to find
5118 * if any holes exist and then log them. We also need to log holes after any
5119 * truncate operation that changes the inode's size.
5120 */
5121static int btrfs_log_holes(struct btrfs_trans_handle *trans,
5122			   struct btrfs_inode *inode,
5123			   struct btrfs_path *path)
5124{
5125	struct btrfs_root *root = inode->root;
5126	struct btrfs_fs_info *fs_info = root->fs_info;
5127	struct btrfs_key key;
5128	const u64 ino = btrfs_ino(inode);
5129	const u64 i_size = i_size_read(&inode->vfs_inode);
5130	u64 prev_extent_end = 0;
5131	int ret;
5132
5133	if (!btrfs_fs_incompat(fs_info, NO_HOLES) || i_size == 0)
5134		return 0;
5135
5136	key.objectid = ino;
5137	key.type = BTRFS_EXTENT_DATA_KEY;
5138	key.offset = 0;
5139
5140	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
5141	if (ret < 0)
5142		return ret;
5143
5144	while (true) {
5145		struct extent_buffer *leaf = path->nodes[0];
5146
5147		if (path->slots[0] >= btrfs_header_nritems(path->nodes[0])) {
5148			ret = btrfs_next_leaf(root, path);
5149			if (ret < 0)
5150				return ret;
5151			if (ret > 0) {
5152				ret = 0;
5153				break;
5154			}
5155			leaf = path->nodes[0];
5156		}
5157
5158		btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
5159		if (key.objectid != ino || key.type != BTRFS_EXTENT_DATA_KEY)
5160			break;
5161
5162		/* We have a hole, log it. */
5163		if (prev_extent_end < key.offset) {
5164			const u64 hole_len = key.offset - prev_extent_end;
5165
5166			/*
5167			 * Release the path to avoid deadlocks with other code
5168			 * paths that search the root while holding locks on
5169			 * leafs from the log root.
5170			 */
5171			btrfs_release_path(path);
5172			ret = btrfs_insert_hole_extent(trans, root->log_root,
5173						       ino, prev_extent_end,
5174						       hole_len);
5175			if (ret < 0)
5176				return ret;
5177
5178			/*
5179			 * Search for the same key again in the root. Since it's
5180			 * an extent item and we are holding the inode lock, the
5181			 * key must still exist. If it doesn't just emit warning
5182			 * and return an error to fall back to a transaction
5183			 * commit.
5184			 */
5185			ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
5186			if (ret < 0)
5187				return ret;
5188			if (WARN_ON(ret > 0))
5189				return -ENOENT;
5190			leaf = path->nodes[0];
5191		}
5192
5193		prev_extent_end = btrfs_file_extent_end(path);
5194		path->slots[0]++;
5195		cond_resched();
5196	}
5197
5198	if (prev_extent_end < i_size) {
5199		u64 hole_len;
5200
5201		btrfs_release_path(path);
5202		hole_len = ALIGN(i_size - prev_extent_end, fs_info->sectorsize);
5203		ret = btrfs_insert_hole_extent(trans, root->log_root, ino,
5204					       prev_extent_end, hole_len);
5205		if (ret < 0)
5206			return ret;
5207	}
5208
5209	return 0;
5210}
5211
5212/*
5213 * When we are logging a new inode X, check if it doesn't have a reference that
5214 * matches the reference from some other inode Y created in a past transaction
5215 * and that was renamed in the current transaction. If we don't do this, then at
5216 * log replay time we can lose inode Y (and all its files if it's a directory):
5217 *
5218 * mkdir /mnt/x
5219 * echo "hello world" > /mnt/x/foobar
5220 * sync
5221 * mv /mnt/x /mnt/y
5222 * mkdir /mnt/x                 # or touch /mnt/x
5223 * xfs_io -c fsync /mnt/x
5224 * <power fail>
5225 * mount fs, trigger log replay
5226 *
5227 * After the log replay procedure, we would lose the first directory and all its
5228 * files (file foobar).
5229 * For the case where inode Y is not a directory we simply end up losing it:
5230 *
5231 * echo "123" > /mnt/foo
5232 * sync
5233 * mv /mnt/foo /mnt/bar
5234 * echo "abc" > /mnt/foo
5235 * xfs_io -c fsync /mnt/foo
5236 * <power fail>
5237 *
5238 * We also need this for cases where a snapshot entry is replaced by some other
5239 * entry (file or directory) otherwise we end up with an unreplayable log due to
5240 * attempts to delete the snapshot entry (entry of type BTRFS_ROOT_ITEM_KEY) as
5241 * if it were a regular entry:
5242 *
5243 * mkdir /mnt/x
5244 * btrfs subvolume snapshot /mnt /mnt/x/snap
5245 * btrfs subvolume delete /mnt/x/snap
5246 * rmdir /mnt/x
5247 * mkdir /mnt/x
5248 * fsync /mnt/x or fsync some new file inside it
5249 * <power fail>
5250 *
5251 * The snapshot delete, rmdir of x, mkdir of a new x and the fsync all happen in
5252 * the same transaction.
5253 */
5254static int btrfs_check_ref_name_override(struct extent_buffer *eb,
5255					 const int slot,
5256					 const struct btrfs_key *key,
5257					 struct btrfs_inode *inode,
5258					 u64 *other_ino, u64 *other_parent)
5259{
5260	int ret;
5261	struct btrfs_path *search_path;
5262	char *name = NULL;
5263	u32 name_len = 0;
5264	u32 item_size = btrfs_item_size(eb, slot);
5265	u32 cur_offset = 0;
5266	unsigned long ptr = btrfs_item_ptr_offset(eb, slot);
5267
5268	search_path = btrfs_alloc_path();
5269	if (!search_path)
5270		return -ENOMEM;
5271	search_path->search_commit_root = 1;
5272	search_path->skip_locking = 1;
5273
5274	while (cur_offset < item_size) {
5275		u64 parent;
5276		u32 this_name_len;
5277		u32 this_len;
5278		unsigned long name_ptr;
5279		struct btrfs_dir_item *di;
5280		struct fscrypt_str name_str;
5281
5282		if (key->type == BTRFS_INODE_REF_KEY) {
5283			struct btrfs_inode_ref *iref;
5284
5285			iref = (struct btrfs_inode_ref *)(ptr + cur_offset);
5286			parent = key->offset;
5287			this_name_len = btrfs_inode_ref_name_len(eb, iref);
5288			name_ptr = (unsigned long)(iref + 1);
5289			this_len = sizeof(*iref) + this_name_len;
5290		} else {
5291			struct btrfs_inode_extref *extref;
5292
5293			extref = (struct btrfs_inode_extref *)(ptr +
5294							       cur_offset);
5295			parent = btrfs_inode_extref_parent(eb, extref);
5296			this_name_len = btrfs_inode_extref_name_len(eb, extref);
5297			name_ptr = (unsigned long)&extref->name;
5298			this_len = sizeof(*extref) + this_name_len;
5299		}
5300
5301		if (this_name_len > name_len) {
5302			char *new_name;
5303
5304			new_name = krealloc(name, this_name_len, GFP_NOFS);
5305			if (!new_name) {
5306				ret = -ENOMEM;
5307				goto out;
5308			}
5309			name_len = this_name_len;
5310			name = new_name;
5311		}
5312
5313		read_extent_buffer(eb, name, name_ptr, this_name_len);
5314
5315		name_str.name = name;
5316		name_str.len = this_name_len;
5317		di = btrfs_lookup_dir_item(NULL, inode->root, search_path,
5318				parent, &name_str, 0);
5319		if (di && !IS_ERR(di)) {
5320			struct btrfs_key di_key;
5321
5322			btrfs_dir_item_key_to_cpu(search_path->nodes[0],
5323						  di, &di_key);
5324			if (di_key.type == BTRFS_INODE_ITEM_KEY) {
5325				if (di_key.objectid != key->objectid) {
5326					ret = 1;
5327					*other_ino = di_key.objectid;
5328					*other_parent = parent;
5329				} else {
5330					ret = 0;
5331				}
5332			} else {
5333				ret = -EAGAIN;
5334			}
5335			goto out;
5336		} else if (IS_ERR(di)) {
5337			ret = PTR_ERR(di);
5338			goto out;
5339		}
5340		btrfs_release_path(search_path);
5341
5342		cur_offset += this_len;
5343	}
5344	ret = 0;
5345out:
5346	btrfs_free_path(search_path);
5347	kfree(name);
5348	return ret;
5349}
5350
5351/*
5352 * Check if we need to log an inode. This is used in contexts where while
5353 * logging an inode we need to log another inode (either that it exists or in
5354 * full mode). This is used instead of btrfs_inode_in_log() because the later
5355 * requires the inode to be in the log and have the log transaction committed,
5356 * while here we do not care if the log transaction was already committed - our
5357 * caller will commit the log later - and we want to avoid logging an inode
5358 * multiple times when multiple tasks have joined the same log transaction.
5359 */
5360static bool need_log_inode(const struct btrfs_trans_handle *trans,
5361			   struct btrfs_inode *inode)
5362{
5363	/*
5364	 * If a directory was not modified, no dentries added or removed, we can
5365	 * and should avoid logging it.
5366	 */
5367	if (S_ISDIR(inode->vfs_inode.i_mode) && inode->last_trans < trans->transid)
5368		return false;
5369
5370	/*
5371	 * If this inode does not have new/updated/deleted xattrs since the last
5372	 * time it was logged and is flagged as logged in the current transaction,
5373	 * we can skip logging it. As for new/deleted names, those are updated in
5374	 * the log by link/unlink/rename operations.
5375	 * In case the inode was logged and then evicted and reloaded, its
5376	 * logged_trans will be 0, in which case we have to fully log it since
5377	 * logged_trans is a transient field, not persisted.
5378	 */
5379	if (inode_logged(trans, inode, NULL) == 1 &&
5380	    !test_bit(BTRFS_INODE_COPY_EVERYTHING, &inode->runtime_flags))
5381		return false;
5382
5383	return true;
5384}
5385
5386struct btrfs_dir_list {
5387	u64 ino;
5388	struct list_head list;
5389};
5390
5391/*
5392 * Log the inodes of the new dentries of a directory.
5393 * See process_dir_items_leaf() for details about why it is needed.
5394 * This is a recursive operation - if an existing dentry corresponds to a
5395 * directory, that directory's new entries are logged too (same behaviour as
5396 * ext3/4, xfs, f2fs, reiserfs, nilfs2). Note that when logging the inodes
5397 * the dentries point to we do not acquire their VFS lock, otherwise lockdep
5398 * complains about the following circular lock dependency / possible deadlock:
5399 *
5400 *        CPU0                                        CPU1
5401 *        ----                                        ----
5402 * lock(&type->i_mutex_dir_key#3/2);
5403 *                                            lock(sb_internal#2);
5404 *                                            lock(&type->i_mutex_dir_key#3/2);
5405 * lock(&sb->s_type->i_mutex_key#14);
5406 *
5407 * Where sb_internal is the lock (a counter that works as a lock) acquired by
5408 * sb_start_intwrite() in btrfs_start_transaction().
5409 * Not acquiring the VFS lock of the inodes is still safe because:
5410 *
5411 * 1) For regular files we log with a mode of LOG_INODE_EXISTS. It's possible
5412 *    that while logging the inode new references (names) are added or removed
5413 *    from the inode, leaving the logged inode item with a link count that does
5414 *    not match the number of logged inode reference items. This is fine because
5415 *    at log replay time we compute the real number of links and correct the
5416 *    link count in the inode item (see replay_one_buffer() and
5417 *    link_to_fixup_dir());
5418 *
5419 * 2) For directories we log with a mode of LOG_INODE_ALL. It's possible that
5420 *    while logging the inode's items new index items (key type
5421 *    BTRFS_DIR_INDEX_KEY) are added to fs/subvol tree and the logged inode item
5422 *    has a size that doesn't match the sum of the lengths of all the logged
5423 *    names - this is ok, not a problem, because at log replay time we set the
5424 *    directory's i_size to the correct value (see replay_one_name() and
5425 *    overwrite_item()).
5426 */
5427static int log_new_dir_dentries(struct btrfs_trans_handle *trans,
5428				struct btrfs_inode *start_inode,
5429				struct btrfs_log_ctx *ctx)
5430{
5431	struct btrfs_root *root = start_inode->root;
5432	struct btrfs_fs_info *fs_info = root->fs_info;
5433	struct btrfs_path *path;
5434	LIST_HEAD(dir_list);
5435	struct btrfs_dir_list *dir_elem;
5436	u64 ino = btrfs_ino(start_inode);
5437	struct btrfs_inode *curr_inode = start_inode;
5438	int ret = 0;
5439
5440	/*
5441	 * If we are logging a new name, as part of a link or rename operation,
5442	 * don't bother logging new dentries, as we just want to log the names
5443	 * of an inode and that any new parents exist.
5444	 */
5445	if (ctx->logging_new_name)
5446		return 0;
5447
5448	path = btrfs_alloc_path();
5449	if (!path)
5450		return -ENOMEM;
5451
5452	/* Pairs with btrfs_add_delayed_iput below. */
5453	ihold(&curr_inode->vfs_inode);
5454
5455	while (true) {
5456		struct inode *vfs_inode;
5457		struct btrfs_key key;
5458		struct btrfs_key found_key;
5459		u64 next_index;
5460		bool continue_curr_inode = true;
5461		int iter_ret;
5462
5463		key.objectid = ino;
5464		key.type = BTRFS_DIR_INDEX_KEY;
5465		key.offset = btrfs_get_first_dir_index_to_log(curr_inode);
5466		next_index = key.offset;
5467again:
5468		btrfs_for_each_slot(root->log_root, &key, &found_key, path, iter_ret) {
5469			struct extent_buffer *leaf = path->nodes[0];
5470			struct btrfs_dir_item *di;
5471			struct btrfs_key di_key;
5472			struct inode *di_inode;
5473			int log_mode = LOG_INODE_EXISTS;
5474			int type;
5475
5476			if (found_key.objectid != ino ||
5477			    found_key.type != BTRFS_DIR_INDEX_KEY) {
5478				continue_curr_inode = false;
5479				break;
5480			}
5481
5482			next_index = found_key.offset + 1;
5483
5484			di = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_dir_item);
5485			type = btrfs_dir_ftype(leaf, di);
5486			if (btrfs_dir_transid(leaf, di) < trans->transid)
5487				continue;
5488			btrfs_dir_item_key_to_cpu(leaf, di, &di_key);
5489			if (di_key.type == BTRFS_ROOT_ITEM_KEY)
5490				continue;
5491
5492			btrfs_release_path(path);
5493			di_inode = btrfs_iget(fs_info->sb, di_key.objectid, root);
5494			if (IS_ERR(di_inode)) {
5495				ret = PTR_ERR(di_inode);
5496				goto out;
5497			}
5498
5499			if (!need_log_inode(trans, BTRFS_I(di_inode))) {
5500				btrfs_add_delayed_iput(BTRFS_I(di_inode));
5501				break;
5502			}
5503
5504			ctx->log_new_dentries = false;
5505			if (type == BTRFS_FT_DIR)
5506				log_mode = LOG_INODE_ALL;
5507			ret = btrfs_log_inode(trans, BTRFS_I(di_inode),
5508					      log_mode, ctx);
5509			btrfs_add_delayed_iput(BTRFS_I(di_inode));
5510			if (ret)
5511				goto out;
5512			if (ctx->log_new_dentries) {
5513				dir_elem = kmalloc(sizeof(*dir_elem), GFP_NOFS);
5514				if (!dir_elem) {
5515					ret = -ENOMEM;
5516					goto out;
5517				}
5518				dir_elem->ino = di_key.objectid;
5519				list_add_tail(&dir_elem->list, &dir_list);
5520			}
5521			break;
5522		}
5523
5524		btrfs_release_path(path);
5525
5526		if (iter_ret < 0) {
5527			ret = iter_ret;
5528			goto out;
5529		} else if (iter_ret > 0) {
5530			continue_curr_inode = false;
5531		} else {
5532			key = found_key;
5533		}
5534
5535		if (continue_curr_inode && key.offset < (u64)-1) {
5536			key.offset++;
5537			goto again;
5538		}
5539
5540		btrfs_set_first_dir_index_to_log(curr_inode, next_index);
5541
5542		if (list_empty(&dir_list))
5543			break;
5544
5545		dir_elem = list_first_entry(&dir_list, struct btrfs_dir_list, list);
5546		ino = dir_elem->ino;
5547		list_del(&dir_elem->list);
5548		kfree(dir_elem);
5549
5550		btrfs_add_delayed_iput(curr_inode);
5551		curr_inode = NULL;
5552
5553		vfs_inode = btrfs_iget(fs_info->sb, ino, root);
5554		if (IS_ERR(vfs_inode)) {
5555			ret = PTR_ERR(vfs_inode);
5556			break;
5557		}
5558		curr_inode = BTRFS_I(vfs_inode);
5559	}
5560out:
5561	btrfs_free_path(path);
5562	if (curr_inode)
5563		btrfs_add_delayed_iput(curr_inode);
5564
5565	if (ret) {
5566		struct btrfs_dir_list *next;
5567
5568		list_for_each_entry_safe(dir_elem, next, &dir_list, list)
5569			kfree(dir_elem);
5570	}
5571
5572	return ret;
5573}
5574
5575struct btrfs_ino_list {
5576	u64 ino;
5577	u64 parent;
5578	struct list_head list;
5579};
5580
5581static void free_conflicting_inodes(struct btrfs_log_ctx *ctx)
5582{
5583	struct btrfs_ino_list *curr;
5584	struct btrfs_ino_list *next;
5585
5586	list_for_each_entry_safe(curr, next, &ctx->conflict_inodes, list) {
5587		list_del(&curr->list);
5588		kfree(curr);
5589	}
5590}
5591
5592static int conflicting_inode_is_dir(struct btrfs_root *root, u64 ino,
5593				    struct btrfs_path *path)
5594{
5595	struct btrfs_key key;
5596	int ret;
5597
5598	key.objectid = ino;
5599	key.type = BTRFS_INODE_ITEM_KEY;
5600	key.offset = 0;
5601
5602	path->search_commit_root = 1;
5603	path->skip_locking = 1;
5604
5605	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
5606	if (WARN_ON_ONCE(ret > 0)) {
5607		/*
5608		 * We have previously found the inode through the commit root
5609		 * so this should not happen. If it does, just error out and
5610		 * fallback to a transaction commit.
5611		 */
5612		ret = -ENOENT;
5613	} else if (ret == 0) {
5614		struct btrfs_inode_item *item;
5615
5616		item = btrfs_item_ptr(path->nodes[0], path->slots[0],
5617				      struct btrfs_inode_item);
5618		if (S_ISDIR(btrfs_inode_mode(path->nodes[0], item)))
5619			ret = 1;
5620	}
5621
5622	btrfs_release_path(path);
5623	path->search_commit_root = 0;
5624	path->skip_locking = 0;
5625
5626	return ret;
5627}
5628
5629static int add_conflicting_inode(struct btrfs_trans_handle *trans,
5630				 struct btrfs_root *root,
5631				 struct btrfs_path *path,
5632				 u64 ino, u64 parent,
5633				 struct btrfs_log_ctx *ctx)
5634{
5635	struct btrfs_ino_list *ino_elem;
5636	struct inode *inode;
5637
5638	/*
5639	 * It's rare to have a lot of conflicting inodes, in practice it is not
5640	 * common to have more than 1 or 2. We don't want to collect too many,
5641	 * as we could end up logging too many inodes (even if only in
5642	 * LOG_INODE_EXISTS mode) and slow down other fsyncs or transaction
5643	 * commits.
5644	 */
5645	if (ctx->num_conflict_inodes >= MAX_CONFLICT_INODES)
5646		return BTRFS_LOG_FORCE_COMMIT;
5647
5648	inode = btrfs_iget(root->fs_info->sb, ino, root);
5649	/*
5650	 * If the other inode that had a conflicting dir entry was deleted in
5651	 * the current transaction then we either:
5652	 *
5653	 * 1) Log the parent directory (later after adding it to the list) if
5654	 *    the inode is a directory. This is because it may be a deleted
5655	 *    subvolume/snapshot or it may be a regular directory that had
5656	 *    deleted subvolumes/snapshots (or subdirectories that had them),
5657	 *    and at the moment we can't deal with dropping subvolumes/snapshots
5658	 *    during log replay. So we just log the parent, which will result in
5659	 *    a fallback to a transaction commit if we are dealing with those
5660	 *    cases (last_unlink_trans will match the current transaction);
5661	 *
5662	 * 2) Do nothing if it's not a directory. During log replay we simply
5663	 *    unlink the conflicting dentry from the parent directory and then
5664	 *    add the dentry for our inode. Like this we can avoid logging the
5665	 *    parent directory (and maybe fallback to a transaction commit in
5666	 *    case it has a last_unlink_trans == trans->transid, due to moving
5667	 *    some inode from it to some other directory).
5668	 */
5669	if (IS_ERR(inode)) {
5670		int ret = PTR_ERR(inode);
5671
5672		if (ret != -ENOENT)
5673			return ret;
5674
5675		ret = conflicting_inode_is_dir(root, ino, path);
5676		/* Not a directory or we got an error. */
5677		if (ret <= 0)
5678			return ret;
5679
5680		/* Conflicting inode is a directory, so we'll log its parent. */
5681		ino_elem = kmalloc(sizeof(*ino_elem), GFP_NOFS);
5682		if (!ino_elem)
5683			return -ENOMEM;
5684		ino_elem->ino = ino;
5685		ino_elem->parent = parent;
5686		list_add_tail(&ino_elem->list, &ctx->conflict_inodes);
5687		ctx->num_conflict_inodes++;
5688
5689		return 0;
5690	}
5691
5692	/*
5693	 * If the inode was already logged skip it - otherwise we can hit an
5694	 * infinite loop. Example:
5695	 *
5696	 * From the commit root (previous transaction) we have the following
5697	 * inodes:
5698	 *
5699	 * inode 257 a directory
5700	 * inode 258 with references "zz" and "zz_link" on inode 257
5701	 * inode 259 with reference "a" on inode 257
5702	 *
5703	 * And in the current (uncommitted) transaction we have:
5704	 *
5705	 * inode 257 a directory, unchanged
5706	 * inode 258 with references "a" and "a2" on inode 257
5707	 * inode 259 with reference "zz_link" on inode 257
5708	 * inode 261 with reference "zz" on inode 257
5709	 *
5710	 * When logging inode 261 the following infinite loop could
5711	 * happen if we don't skip already logged inodes:
5712	 *
5713	 * - we detect inode 258 as a conflicting inode, with inode 261
5714	 *   on reference "zz", and log it;
5715	 *
5716	 * - we detect inode 259 as a conflicting inode, with inode 258
5717	 *   on reference "a", and log it;
5718	 *
5719	 * - we detect inode 258 as a conflicting inode, with inode 259
5720	 *   on reference "zz_link", and log it - again! After this we
5721	 *   repeat the above steps forever.
5722	 *
5723	 * Here we can use need_log_inode() because we only need to log the
5724	 * inode in LOG_INODE_EXISTS mode and rename operations update the log,
5725	 * so that the log ends up with the new name and without the old name.
5726	 */
5727	if (!need_log_inode(trans, BTRFS_I(inode))) {
5728		btrfs_add_delayed_iput(BTRFS_I(inode));
5729		return 0;
5730	}
5731
5732	btrfs_add_delayed_iput(BTRFS_I(inode));
5733
5734	ino_elem = kmalloc(sizeof(*ino_elem), GFP_NOFS);
5735	if (!ino_elem)
5736		return -ENOMEM;
5737	ino_elem->ino = ino;
5738	ino_elem->parent = parent;
5739	list_add_tail(&ino_elem->list, &ctx->conflict_inodes);
5740	ctx->num_conflict_inodes++;
5741
5742	return 0;
5743}
5744
5745static int log_conflicting_inodes(struct btrfs_trans_handle *trans,
5746				  struct btrfs_root *root,
5747				  struct btrfs_log_ctx *ctx)
5748{
5749	struct btrfs_fs_info *fs_info = root->fs_info;
5750	int ret = 0;
5751
5752	/*
5753	 * Conflicting inodes are logged by the first call to btrfs_log_inode(),
5754	 * otherwise we could have unbounded recursion of btrfs_log_inode()
5755	 * calls. This check guarantees we can have only 1 level of recursion.
5756	 */
5757	if (ctx->logging_conflict_inodes)
5758		return 0;
5759
5760	ctx->logging_conflict_inodes = true;
5761
5762	/*
5763	 * New conflicting inodes may be found and added to the list while we
5764	 * are logging a conflicting inode, so keep iterating while the list is
5765	 * not empty.
5766	 */
5767	while (!list_empty(&ctx->conflict_inodes)) {
5768		struct btrfs_ino_list *curr;
5769		struct inode *inode;
5770		u64 ino;
5771		u64 parent;
5772
5773		curr = list_first_entry(&ctx->conflict_inodes,
5774					struct btrfs_ino_list, list);
5775		ino = curr->ino;
5776		parent = curr->parent;
5777		list_del(&curr->list);
5778		kfree(curr);
5779
5780		inode = btrfs_iget(fs_info->sb, ino, root);
5781		/*
5782		 * If the other inode that had a conflicting dir entry was
5783		 * deleted in the current transaction, we need to log its parent
5784		 * directory. See the comment at add_conflicting_inode().
5785		 */
5786		if (IS_ERR(inode)) {
5787			ret = PTR_ERR(inode);
5788			if (ret != -ENOENT)
5789				break;
5790
5791			inode = btrfs_iget(fs_info->sb, parent, root);
5792			if (IS_ERR(inode)) {
5793				ret = PTR_ERR(inode);
5794				break;
5795			}
5796
5797			/*
5798			 * Always log the directory, we cannot make this
5799			 * conditional on need_log_inode() because the directory
5800			 * might have been logged in LOG_INODE_EXISTS mode or
5801			 * the dir index of the conflicting inode is not in a
5802			 * dir index key range logged for the directory. So we
5803			 * must make sure the deletion is recorded.
5804			 */
5805			ret = btrfs_log_inode(trans, BTRFS_I(inode),
5806					      LOG_INODE_ALL, ctx);
5807			btrfs_add_delayed_iput(BTRFS_I(inode));
5808			if (ret)
5809				break;
5810			continue;
5811		}
5812
5813		/*
5814		 * Here we can use need_log_inode() because we only need to log
5815		 * the inode in LOG_INODE_EXISTS mode and rename operations
5816		 * update the log, so that the log ends up with the new name and
5817		 * without the old name.
5818		 *
5819		 * We did this check at add_conflicting_inode(), but here we do
5820		 * it again because if some other task logged the inode after
5821		 * that, we can avoid doing it again.
5822		 */
5823		if (!need_log_inode(trans, BTRFS_I(inode))) {
5824			btrfs_add_delayed_iput(BTRFS_I(inode));
5825			continue;
5826		}
5827
5828		/*
5829		 * We are safe logging the other inode without acquiring its
5830		 * lock as long as we log with the LOG_INODE_EXISTS mode. We
5831		 * are safe against concurrent renames of the other inode as
5832		 * well because during a rename we pin the log and update the
5833		 * log with the new name before we unpin it.
5834		 */
5835		ret = btrfs_log_inode(trans, BTRFS_I(inode), LOG_INODE_EXISTS, ctx);
5836		btrfs_add_delayed_iput(BTRFS_I(inode));
5837		if (ret)
5838			break;
5839	}
5840
5841	ctx->logging_conflict_inodes = false;
5842	if (ret)
5843		free_conflicting_inodes(ctx);
5844
5845	return ret;
5846}
5847
5848static int copy_inode_items_to_log(struct btrfs_trans_handle *trans,
5849				   struct btrfs_inode *inode,
5850				   struct btrfs_key *min_key,
5851				   const struct btrfs_key *max_key,
5852				   struct btrfs_path *path,
5853				   struct btrfs_path *dst_path,
5854				   const u64 logged_isize,
5855				   const int inode_only,
5856				   struct btrfs_log_ctx *ctx,
5857				   bool *need_log_inode_item)
5858{
5859	const u64 i_size = i_size_read(&inode->vfs_inode);
5860	struct btrfs_root *root = inode->root;
5861	int ins_start_slot = 0;
5862	int ins_nr = 0;
5863	int ret;
5864
5865	while (1) {
5866		ret = btrfs_search_forward(root, min_key, path, trans->transid);
5867		if (ret < 0)
5868			return ret;
5869		if (ret > 0) {
5870			ret = 0;
5871			break;
5872		}
5873again:
5874		/* Note, ins_nr might be > 0 here, cleanup outside the loop */
5875		if (min_key->objectid != max_key->objectid)
5876			break;
5877		if (min_key->type > max_key->type)
5878			break;
5879
5880		if (min_key->type == BTRFS_INODE_ITEM_KEY) {
5881			*need_log_inode_item = false;
5882		} else if (min_key->type == BTRFS_EXTENT_DATA_KEY &&
5883			   min_key->offset >= i_size) {
5884			/*
5885			 * Extents at and beyond eof are logged with
5886			 * btrfs_log_prealloc_extents().
5887			 * Only regular files have BTRFS_EXTENT_DATA_KEY keys,
5888			 * and no keys greater than that, so bail out.
5889			 */
5890			break;
5891		} else if ((min_key->type == BTRFS_INODE_REF_KEY ||
5892			    min_key->type == BTRFS_INODE_EXTREF_KEY) &&
5893			   (inode->generation == trans->transid ||
5894			    ctx->logging_conflict_inodes)) {
5895			u64 other_ino = 0;
5896			u64 other_parent = 0;
5897
5898			ret = btrfs_check_ref_name_override(path->nodes[0],
5899					path->slots[0], min_key, inode,
5900					&other_ino, &other_parent);
5901			if (ret < 0) {
5902				return ret;
5903			} else if (ret > 0 &&
5904				   other_ino != btrfs_ino(BTRFS_I(ctx->inode))) {
5905				if (ins_nr > 0) {
5906					ins_nr++;
5907				} else {
5908					ins_nr = 1;
5909					ins_start_slot = path->slots[0];
5910				}
5911				ret = copy_items(trans, inode, dst_path, path,
5912						 ins_start_slot, ins_nr,
5913						 inode_only, logged_isize, ctx);
5914				if (ret < 0)
5915					return ret;
5916				ins_nr = 0;
5917
5918				btrfs_release_path(path);
5919				ret = add_conflicting_inode(trans, root, path,
5920							    other_ino,
5921							    other_parent, ctx);
5922				if (ret)
5923					return ret;
5924				goto next_key;
5925			}
5926		} else if (min_key->type == BTRFS_XATTR_ITEM_KEY) {
5927			/* Skip xattrs, logged later with btrfs_log_all_xattrs() */
5928			if (ins_nr == 0)
5929				goto next_slot;
5930			ret = copy_items(trans, inode, dst_path, path,
5931					 ins_start_slot,
5932					 ins_nr, inode_only, logged_isize, ctx);
5933			if (ret < 0)
5934				return ret;
5935			ins_nr = 0;
5936			goto next_slot;
5937		}
5938
5939		if (ins_nr && ins_start_slot + ins_nr == path->slots[0]) {
5940			ins_nr++;
5941			goto next_slot;
5942		} else if (!ins_nr) {
5943			ins_start_slot = path->slots[0];
5944			ins_nr = 1;
5945			goto next_slot;
5946		}
5947
5948		ret = copy_items(trans, inode, dst_path, path, ins_start_slot,
5949				 ins_nr, inode_only, logged_isize, ctx);
5950		if (ret < 0)
5951			return ret;
5952		ins_nr = 1;
5953		ins_start_slot = path->slots[0];
5954next_slot:
5955		path->slots[0]++;
5956		if (path->slots[0] < btrfs_header_nritems(path->nodes[0])) {
5957			btrfs_item_key_to_cpu(path->nodes[0], min_key,
5958					      path->slots[0]);
5959			goto again;
5960		}
5961		if (ins_nr) {
5962			ret = copy_items(trans, inode, dst_path, path,
5963					 ins_start_slot, ins_nr, inode_only,
5964					 logged_isize, ctx);
5965			if (ret < 0)
5966				return ret;
5967			ins_nr = 0;
5968		}
5969		btrfs_release_path(path);
5970next_key:
5971		if (min_key->offset < (u64)-1) {
5972			min_key->offset++;
5973		} else if (min_key->type < max_key->type) {
5974			min_key->type++;
5975			min_key->offset = 0;
5976		} else {
5977			break;
5978		}
5979
5980		/*
5981		 * We may process many leaves full of items for our inode, so
5982		 * avoid monopolizing a cpu for too long by rescheduling while
5983		 * not holding locks on any tree.
5984		 */
5985		cond_resched();
5986	}
5987	if (ins_nr) {
5988		ret = copy_items(trans, inode, dst_path, path, ins_start_slot,
5989				 ins_nr, inode_only, logged_isize, ctx);
5990		if (ret)
5991			return ret;
5992	}
5993
5994	if (inode_only == LOG_INODE_ALL && S_ISREG(inode->vfs_inode.i_mode)) {
5995		/*
5996		 * Release the path because otherwise we might attempt to double
5997		 * lock the same leaf with btrfs_log_prealloc_extents() below.
5998		 */
5999		btrfs_release_path(path);
6000		ret = btrfs_log_prealloc_extents(trans, inode, dst_path, ctx);
6001	}
6002
6003	return ret;
6004}
6005
6006static int insert_delayed_items_batch(struct btrfs_trans_handle *trans,
6007				      struct btrfs_root *log,
6008				      struct btrfs_path *path,
6009				      const struct btrfs_item_batch *batch,
6010				      const struct btrfs_delayed_item *first_item)
6011{
6012	const struct btrfs_delayed_item *curr = first_item;
6013	int ret;
6014
6015	ret = btrfs_insert_empty_items(trans, log, path, batch);
6016	if (ret)
6017		return ret;
6018
6019	for (int i = 0; i < batch->nr; i++) {
6020		char *data_ptr;
6021
6022		data_ptr = btrfs_item_ptr(path->nodes[0], path->slots[0], char);
6023		write_extent_buffer(path->nodes[0], &curr->data,
6024				    (unsigned long)data_ptr, curr->data_len);
6025		curr = list_next_entry(curr, log_list);
6026		path->slots[0]++;
6027	}
6028
6029	btrfs_release_path(path);
6030
6031	return 0;
6032}
6033
6034static int log_delayed_insertion_items(struct btrfs_trans_handle *trans,
6035				       struct btrfs_inode *inode,
6036				       struct btrfs_path *path,
6037				       const struct list_head *delayed_ins_list,
6038				       struct btrfs_log_ctx *ctx)
6039{
6040	/* 195 (4095 bytes of keys and sizes) fits in a single 4K page. */
6041	const int max_batch_size = 195;
6042	const int leaf_data_size = BTRFS_LEAF_DATA_SIZE(trans->fs_info);
6043	const u64 ino = btrfs_ino(inode);
6044	struct btrfs_root *log = inode->root->log_root;
6045	struct btrfs_item_batch batch = {
6046		.nr = 0,
6047		.total_data_size = 0,
6048	};
6049	const struct btrfs_delayed_item *first = NULL;
6050	const struct btrfs_delayed_item *curr;
6051	char *ins_data;
6052	struct btrfs_key *ins_keys;
6053	u32 *ins_sizes;
6054	u64 curr_batch_size = 0;
6055	int batch_idx = 0;
6056	int ret;
6057
6058	/* We are adding dir index items to the log tree. */
6059	lockdep_assert_held(&inode->log_mutex);
6060
6061	/*
6062	 * We collect delayed items before copying index keys from the subvolume
6063	 * to the log tree. However just after we collected them, they may have
6064	 * been flushed (all of them or just some of them), and therefore we
6065	 * could have copied them from the subvolume tree to the log tree.
6066	 * So find the first delayed item that was not yet logged (they are
6067	 * sorted by index number).
6068	 */
6069	list_for_each_entry(curr, delayed_ins_list, log_list) {
6070		if (curr->index > inode->last_dir_index_offset) {
6071			first = curr;
6072			break;
6073		}
6074	}
6075
6076	/* Empty list or all delayed items were already logged. */
6077	if (!first)
6078		return 0;
6079
6080	ins_data = kmalloc(max_batch_size * sizeof(u32) +
6081			   max_batch_size * sizeof(struct btrfs_key), GFP_NOFS);
6082	if (!ins_data)
6083		return -ENOMEM;
6084	ins_sizes = (u32 *)ins_data;
6085	batch.data_sizes = ins_sizes;
6086	ins_keys = (struct btrfs_key *)(ins_data + max_batch_size * sizeof(u32));
6087	batch.keys = ins_keys;
6088
6089	curr = first;
6090	while (!list_entry_is_head(curr, delayed_ins_list, log_list)) {
6091		const u32 curr_size = curr->data_len + sizeof(struct btrfs_item);
6092
6093		if (curr_batch_size + curr_size > leaf_data_size ||
6094		    batch.nr == max_batch_size) {
6095			ret = insert_delayed_items_batch(trans, log, path,
6096							 &batch, first);
6097			if (ret)
6098				goto out;
6099			batch_idx = 0;
6100			batch.nr = 0;
6101			batch.total_data_size = 0;
6102			curr_batch_size = 0;
6103			first = curr;
6104		}
6105
6106		ins_sizes[batch_idx] = curr->data_len;
6107		ins_keys[batch_idx].objectid = ino;
6108		ins_keys[batch_idx].type = BTRFS_DIR_INDEX_KEY;
6109		ins_keys[batch_idx].offset = curr->index;
6110		curr_batch_size += curr_size;
6111		batch.total_data_size += curr->data_len;
6112		batch.nr++;
6113		batch_idx++;
6114		curr = list_next_entry(curr, log_list);
6115	}
6116
6117	ASSERT(batch.nr >= 1);
6118	ret = insert_delayed_items_batch(trans, log, path, &batch, first);
6119
6120	curr = list_last_entry(delayed_ins_list, struct btrfs_delayed_item,
6121			       log_list);
6122	inode->last_dir_index_offset = curr->index;
6123out:
6124	kfree(ins_data);
6125
6126	return ret;
6127}
6128
6129static int log_delayed_deletions_full(struct btrfs_trans_handle *trans,
6130				      struct btrfs_inode *inode,
6131				      struct btrfs_path *path,
6132				      const struct list_head *delayed_del_list,
6133				      struct btrfs_log_ctx *ctx)
6134{
6135	const u64 ino = btrfs_ino(inode);
6136	const struct btrfs_delayed_item *curr;
6137
6138	curr = list_first_entry(delayed_del_list, struct btrfs_delayed_item,
6139				log_list);
6140
6141	while (!list_entry_is_head(curr, delayed_del_list, log_list)) {
6142		u64 first_dir_index = curr->index;
6143		u64 last_dir_index;
6144		const struct btrfs_delayed_item *next;
6145		int ret;
6146
6147		/*
6148		 * Find a range of consecutive dir index items to delete. Like
6149		 * this we log a single dir range item spanning several contiguous
6150		 * dir items instead of logging one range item per dir index item.
6151		 */
6152		next = list_next_entry(curr, log_list);
6153		while (!list_entry_is_head(next, delayed_del_list, log_list)) {
6154			if (next->index != curr->index + 1)
6155				break;
6156			curr = next;
6157			next = list_next_entry(next, log_list);
6158		}
6159
6160		last_dir_index = curr->index;
6161		ASSERT(last_dir_index >= first_dir_index);
6162
6163		ret = insert_dir_log_key(trans, inode->root->log_root, path,
6164					 ino, first_dir_index, last_dir_index);
6165		if (ret)
6166			return ret;
6167		curr = list_next_entry(curr, log_list);
6168	}
6169
6170	return 0;
6171}
6172
6173static int batch_delete_dir_index_items(struct btrfs_trans_handle *trans,
6174					struct btrfs_inode *inode,
6175					struct btrfs_path *path,
6176					struct btrfs_log_ctx *ctx,
6177					const struct list_head *delayed_del_list,
6178					const struct btrfs_delayed_item *first,
6179					const struct btrfs_delayed_item **last_ret)
6180{
6181	const struct btrfs_delayed_item *next;
6182	struct extent_buffer *leaf = path->nodes[0];
6183	const int last_slot = btrfs_header_nritems(leaf) - 1;
6184	int slot = path->slots[0] + 1;
6185	const u64 ino = btrfs_ino(inode);
6186
6187	next = list_next_entry(first, log_list);
6188
6189	while (slot < last_slot &&
6190	       !list_entry_is_head(next, delayed_del_list, log_list)) {
6191		struct btrfs_key key;
6192
6193		btrfs_item_key_to_cpu(leaf, &key, slot);
6194		if (key.objectid != ino ||
6195		    key.type != BTRFS_DIR_INDEX_KEY ||
6196		    key.offset != next->index)
6197			break;
6198
6199		slot++;
6200		*last_ret = next;
6201		next = list_next_entry(next, log_list);
6202	}
6203
6204	return btrfs_del_items(trans, inode->root->log_root, path,
6205			       path->slots[0], slot - path->slots[0]);
6206}
6207
6208static int log_delayed_deletions_incremental(struct btrfs_trans_handle *trans,
6209					     struct btrfs_inode *inode,
6210					     struct btrfs_path *path,
6211					     const struct list_head *delayed_del_list,
6212					     struct btrfs_log_ctx *ctx)
6213{
6214	struct btrfs_root *log = inode->root->log_root;
6215	const struct btrfs_delayed_item *curr;
6216	u64 last_range_start = 0;
6217	u64 last_range_end = 0;
6218	struct btrfs_key key;
6219
6220	key.objectid = btrfs_ino(inode);
6221	key.type = BTRFS_DIR_INDEX_KEY;
6222	curr = list_first_entry(delayed_del_list, struct btrfs_delayed_item,
6223				log_list);
6224
6225	while (!list_entry_is_head(curr, delayed_del_list, log_list)) {
6226		const struct btrfs_delayed_item *last = curr;
6227		u64 first_dir_index = curr->index;
6228		u64 last_dir_index;
6229		bool deleted_items = false;
6230		int ret;
6231
6232		key.offset = curr->index;
6233		ret = btrfs_search_slot(trans, log, &key, path, -1, 1);
6234		if (ret < 0) {
6235			return ret;
6236		} else if (ret == 0) {
6237			ret = batch_delete_dir_index_items(trans, inode, path, ctx,
6238							   delayed_del_list, curr,
6239							   &last);
6240			if (ret)
6241				return ret;
6242			deleted_items = true;
6243		}
6244
6245		btrfs_release_path(path);
6246
6247		/*
6248		 * If we deleted items from the leaf, it means we have a range
6249		 * item logging their range, so no need to add one or update an
6250		 * existing one. Otherwise we have to log a dir range item.
6251		 */
6252		if (deleted_items)
6253			goto next_batch;
6254
6255		last_dir_index = last->index;
6256		ASSERT(last_dir_index >= first_dir_index);
6257		/*
6258		 * If this range starts right after where the previous one ends,
6259		 * then we want to reuse the previous range item and change its
6260		 * end offset to the end of this range. This is just to minimize
6261		 * leaf space usage, by avoiding adding a new range item.
6262		 */
6263		if (last_range_end != 0 && first_dir_index == last_range_end + 1)
6264			first_dir_index = last_range_start;
6265
6266		ret = insert_dir_log_key(trans, log, path, key.objectid,
6267					 first_dir_index, last_dir_index);
6268		if (ret)
6269			return ret;
6270
6271		last_range_start = first_dir_index;
6272		last_range_end = last_dir_index;
6273next_batch:
6274		curr = list_next_entry(last, log_list);
6275	}
6276
6277	return 0;
6278}
6279
6280static int log_delayed_deletion_items(struct btrfs_trans_handle *trans,
6281				      struct btrfs_inode *inode,
6282				      struct btrfs_path *path,
6283				      const struct list_head *delayed_del_list,
6284				      struct btrfs_log_ctx *ctx)
6285{
6286	/*
6287	 * We are deleting dir index items from the log tree or adding range
6288	 * items to it.
6289	 */
6290	lockdep_assert_held(&inode->log_mutex);
6291
6292	if (list_empty(delayed_del_list))
6293		return 0;
6294
6295	if (ctx->logged_before)
6296		return log_delayed_deletions_incremental(trans, inode, path,
6297							 delayed_del_list, ctx);
6298
6299	return log_delayed_deletions_full(trans, inode, path, delayed_del_list,
6300					  ctx);
6301}
6302
6303/*
6304 * Similar logic as for log_new_dir_dentries(), but it iterates over the delayed
6305 * items instead of the subvolume tree.
6306 */
6307static int log_new_delayed_dentries(struct btrfs_trans_handle *trans,
6308				    struct btrfs_inode *inode,
6309				    const struct list_head *delayed_ins_list,
6310				    struct btrfs_log_ctx *ctx)
6311{
6312	const bool orig_log_new_dentries = ctx->log_new_dentries;
6313	struct btrfs_fs_info *fs_info = trans->fs_info;
6314	struct btrfs_delayed_item *item;
6315	int ret = 0;
6316
6317	/*
6318	 * No need for the log mutex, plus to avoid potential deadlocks or
6319	 * lockdep annotations due to nesting of delayed inode mutexes and log
6320	 * mutexes.
6321	 */
6322	lockdep_assert_not_held(&inode->log_mutex);
6323
6324	ASSERT(!ctx->logging_new_delayed_dentries);
6325	ctx->logging_new_delayed_dentries = true;
6326
6327	list_for_each_entry(item, delayed_ins_list, log_list) {
6328		struct btrfs_dir_item *dir_item;
6329		struct inode *di_inode;
6330		struct btrfs_key key;
6331		int log_mode = LOG_INODE_EXISTS;
6332
6333		dir_item = (struct btrfs_dir_item *)item->data;
6334		btrfs_disk_key_to_cpu(&key, &dir_item->location);
6335
6336		if (key.type == BTRFS_ROOT_ITEM_KEY)
6337			continue;
6338
6339		di_inode = btrfs_iget(fs_info->sb, key.objectid, inode->root);
6340		if (IS_ERR(di_inode)) {
6341			ret = PTR_ERR(di_inode);
6342			break;
6343		}
6344
6345		if (!need_log_inode(trans, BTRFS_I(di_inode))) {
6346			btrfs_add_delayed_iput(BTRFS_I(di_inode));
6347			continue;
6348		}
6349
6350		if (btrfs_stack_dir_ftype(dir_item) == BTRFS_FT_DIR)
6351			log_mode = LOG_INODE_ALL;
6352
6353		ctx->log_new_dentries = false;
6354		ret = btrfs_log_inode(trans, BTRFS_I(di_inode), log_mode, ctx);
6355
6356		if (!ret && ctx->log_new_dentries)
6357			ret = log_new_dir_dentries(trans, BTRFS_I(di_inode), ctx);
6358
6359		btrfs_add_delayed_iput(BTRFS_I(di_inode));
6360
6361		if (ret)
6362			break;
6363	}
6364
6365	ctx->log_new_dentries = orig_log_new_dentries;
6366	ctx->logging_new_delayed_dentries = false;
6367
6368	return ret;
6369}
6370
6371/* log a single inode in the tree log.
6372 * At least one parent directory for this inode must exist in the tree
6373 * or be logged already.
6374 *
6375 * Any items from this inode changed by the current transaction are copied
6376 * to the log tree.  An extra reference is taken on any extents in this
6377 * file, allowing us to avoid a whole pile of corner cases around logging
6378 * blocks that have been removed from the tree.
6379 *
6380 * See LOG_INODE_ALL and related defines for a description of what inode_only
6381 * does.
6382 *
6383 * This handles both files and directories.
6384 */
6385static int btrfs_log_inode(struct btrfs_trans_handle *trans,
6386			   struct btrfs_inode *inode,
6387			   int inode_only,
6388			   struct btrfs_log_ctx *ctx)
6389{
6390	struct btrfs_path *path;
6391	struct btrfs_path *dst_path;
6392	struct btrfs_key min_key;
6393	struct btrfs_key max_key;
6394	struct btrfs_root *log = inode->root->log_root;
6395	int ret;
6396	bool fast_search = false;
6397	u64 ino = btrfs_ino(inode);
6398	struct extent_map_tree *em_tree = &inode->extent_tree;
6399	u64 logged_isize = 0;
6400	bool need_log_inode_item = true;
6401	bool xattrs_logged = false;
6402	bool inode_item_dropped = true;
6403	bool full_dir_logging = false;
6404	LIST_HEAD(delayed_ins_list);
6405	LIST_HEAD(delayed_del_list);
6406
6407	path = btrfs_alloc_path();
6408	if (!path)
6409		return -ENOMEM;
6410	dst_path = btrfs_alloc_path();
6411	if (!dst_path) {
6412		btrfs_free_path(path);
6413		return -ENOMEM;
6414	}
6415
6416	min_key.objectid = ino;
6417	min_key.type = BTRFS_INODE_ITEM_KEY;
6418	min_key.offset = 0;
6419
6420	max_key.objectid = ino;
6421
6422
6423	/* today the code can only do partial logging of directories */
6424	if (S_ISDIR(inode->vfs_inode.i_mode) ||
6425	    (!test_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
6426		       &inode->runtime_flags) &&
6427	     inode_only >= LOG_INODE_EXISTS))
6428		max_key.type = BTRFS_XATTR_ITEM_KEY;
6429	else
6430		max_key.type = (u8)-1;
6431	max_key.offset = (u64)-1;
6432
6433	if (S_ISDIR(inode->vfs_inode.i_mode) && inode_only == LOG_INODE_ALL)
6434		full_dir_logging = true;
6435
6436	/*
6437	 * If we are logging a directory while we are logging dentries of the
6438	 * delayed items of some other inode, then we need to flush the delayed
6439	 * items of this directory and not log the delayed items directly. This
6440	 * is to prevent more than one level of recursion into btrfs_log_inode()
6441	 * by having something like this:
6442	 *
6443	 *     $ mkdir -p a/b/c/d/e/f/g/h/...
6444	 *     $ xfs_io -c "fsync" a
6445	 *
6446	 * Where all directories in the path did not exist before and are
6447	 * created in the current transaction.
6448	 * So in such a case we directly log the delayed items of the main
6449	 * directory ("a") without flushing them first, while for each of its
6450	 * subdirectories we flush their delayed items before logging them.
6451	 * This prevents a potential unbounded recursion like this:
6452	 *
6453	 * btrfs_log_inode()
6454	 *   log_new_delayed_dentries()
6455	 *      btrfs_log_inode()
6456	 *        log_new_delayed_dentries()
6457	 *          btrfs_log_inode()
6458	 *            log_new_delayed_dentries()
6459	 *              (...)
6460	 *
6461	 * We have thresholds for the maximum number of delayed items to have in
6462	 * memory, and once they are hit, the items are flushed asynchronously.
6463	 * However the limit is quite high, so lets prevent deep levels of
6464	 * recursion to happen by limiting the maximum depth to be 1.
6465	 */
6466	if (full_dir_logging && ctx->logging_new_delayed_dentries) {
6467		ret = btrfs_commit_inode_delayed_items(trans, inode);
6468		if (ret)
6469			goto out;
6470	}
6471
6472	mutex_lock(&inode->log_mutex);
6473
6474	/*
6475	 * For symlinks, we must always log their content, which is stored in an
6476	 * inline extent, otherwise we could end up with an empty symlink after
6477	 * log replay, which is invalid on linux (symlink(2) returns -ENOENT if
6478	 * one attempts to create an empty symlink).
6479	 * We don't need to worry about flushing delalloc, because when we create
6480	 * the inline extent when the symlink is created (we never have delalloc
6481	 * for symlinks).
6482	 */
6483	if (S_ISLNK(inode->vfs_inode.i_mode))
6484		inode_only = LOG_INODE_ALL;
6485
6486	/*
6487	 * Before logging the inode item, cache the value returned by
6488	 * inode_logged(), because after that we have the need to figure out if
6489	 * the inode was previously logged in this transaction.
6490	 */
6491	ret = inode_logged(trans, inode, path);
6492	if (ret < 0)
6493		goto out_unlock;
6494	ctx->logged_before = (ret == 1);
6495	ret = 0;
6496
6497	/*
6498	 * This is for cases where logging a directory could result in losing a
6499	 * a file after replaying the log. For example, if we move a file from a
6500	 * directory A to a directory B, then fsync directory A, we have no way
6501	 * to known the file was moved from A to B, so logging just A would
6502	 * result in losing the file after a log replay.
6503	 */
6504	if (full_dir_logging && inode->last_unlink_trans >= trans->transid) {
6505		ret = BTRFS_LOG_FORCE_COMMIT;
6506		goto out_unlock;
6507	}
6508
6509	/*
6510	 * a brute force approach to making sure we get the most uptodate
6511	 * copies of everything.
6512	 */
6513	if (S_ISDIR(inode->vfs_inode.i_mode)) {
6514		clear_bit(BTRFS_INODE_COPY_EVERYTHING, &inode->runtime_flags);
6515		if (ctx->logged_before)
6516			ret = drop_inode_items(trans, log, path, inode,
6517					       BTRFS_XATTR_ITEM_KEY);
6518	} else {
6519		if (inode_only == LOG_INODE_EXISTS && ctx->logged_before) {
6520			/*
6521			 * Make sure the new inode item we write to the log has
6522			 * the same isize as the current one (if it exists).
6523			 * This is necessary to prevent data loss after log
6524			 * replay, and also to prevent doing a wrong expanding
6525			 * truncate - for e.g. create file, write 4K into offset
6526			 * 0, fsync, write 4K into offset 4096, add hard link,
6527			 * fsync some other file (to sync log), power fail - if
6528			 * we use the inode's current i_size, after log replay
6529			 * we get a 8Kb file, with the last 4Kb extent as a hole
6530			 * (zeroes), as if an expanding truncate happened,
6531			 * instead of getting a file of 4Kb only.
6532			 */
6533			ret = logged_inode_size(log, inode, path, &logged_isize);
6534			if (ret)
6535				goto out_unlock;
6536		}
6537		if (test_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
6538			     &inode->runtime_flags)) {
6539			if (inode_only == LOG_INODE_EXISTS) {
6540				max_key.type = BTRFS_XATTR_ITEM_KEY;
6541				if (ctx->logged_before)
6542					ret = drop_inode_items(trans, log, path,
6543							       inode, max_key.type);
6544			} else {
6545				clear_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
6546					  &inode->runtime_flags);
6547				clear_bit(BTRFS_INODE_COPY_EVERYTHING,
6548					  &inode->runtime_flags);
6549				if (ctx->logged_before)
6550					ret = truncate_inode_items(trans, log,
6551								   inode, 0, 0);
6552			}
6553		} else if (test_and_clear_bit(BTRFS_INODE_COPY_EVERYTHING,
6554					      &inode->runtime_flags) ||
6555			   inode_only == LOG_INODE_EXISTS) {
6556			if (inode_only == LOG_INODE_ALL)
6557				fast_search = true;
6558			max_key.type = BTRFS_XATTR_ITEM_KEY;
6559			if (ctx->logged_before)
6560				ret = drop_inode_items(trans, log, path, inode,
6561						       max_key.type);
6562		} else {
6563			if (inode_only == LOG_INODE_ALL)
6564				fast_search = true;
6565			inode_item_dropped = false;
6566			goto log_extents;
6567		}
6568
6569	}
6570	if (ret)
6571		goto out_unlock;
6572
6573	/*
6574	 * If we are logging a directory in full mode, collect the delayed items
6575	 * before iterating the subvolume tree, so that we don't miss any new
6576	 * dir index items in case they get flushed while or right after we are
6577	 * iterating the subvolume tree.
6578	 */
6579	if (full_dir_logging && !ctx->logging_new_delayed_dentries)
6580		btrfs_log_get_delayed_items(inode, &delayed_ins_list,
6581					    &delayed_del_list);
6582
6583	ret = copy_inode_items_to_log(trans, inode, &min_key, &max_key,
6584				      path, dst_path, logged_isize,
6585				      inode_only, ctx,
6586				      &need_log_inode_item);
6587	if (ret)
6588		goto out_unlock;
6589
6590	btrfs_release_path(path);
6591	btrfs_release_path(dst_path);
6592	ret = btrfs_log_all_xattrs(trans, inode, path, dst_path, ctx);
6593	if (ret)
6594		goto out_unlock;
6595	xattrs_logged = true;
6596	if (max_key.type >= BTRFS_EXTENT_DATA_KEY && !fast_search) {
6597		btrfs_release_path(path);
6598		btrfs_release_path(dst_path);
6599		ret = btrfs_log_holes(trans, inode, path);
6600		if (ret)
6601			goto out_unlock;
6602	}
6603log_extents:
6604	btrfs_release_path(path);
6605	btrfs_release_path(dst_path);
6606	if (need_log_inode_item) {
6607		ret = log_inode_item(trans, log, dst_path, inode, inode_item_dropped);
6608		if (ret)
6609			goto out_unlock;
6610		/*
6611		 * If we are doing a fast fsync and the inode was logged before
6612		 * in this transaction, we don't need to log the xattrs because
6613		 * they were logged before. If xattrs were added, changed or
6614		 * deleted since the last time we logged the inode, then we have
6615		 * already logged them because the inode had the runtime flag
6616		 * BTRFS_INODE_COPY_EVERYTHING set.
6617		 */
6618		if (!xattrs_logged && inode->logged_trans < trans->transid) {
6619			ret = btrfs_log_all_xattrs(trans, inode, path, dst_path, ctx);
6620			if (ret)
6621				goto out_unlock;
6622			btrfs_release_path(path);
6623		}
6624	}
6625	if (fast_search) {
6626		ret = btrfs_log_changed_extents(trans, inode, dst_path, ctx);
6627		if (ret)
6628			goto out_unlock;
6629	} else if (inode_only == LOG_INODE_ALL) {
6630		struct extent_map *em, *n;
6631
6632		write_lock(&em_tree->lock);
6633		list_for_each_entry_safe(em, n, &em_tree->modified_extents, list)
6634			list_del_init(&em->list);
6635		write_unlock(&em_tree->lock);
6636	}
6637
6638	if (full_dir_logging) {
6639		ret = log_directory_changes(trans, inode, path, dst_path, ctx);
6640		if (ret)
6641			goto out_unlock;
6642		ret = log_delayed_insertion_items(trans, inode, path,
6643						  &delayed_ins_list, ctx);
6644		if (ret)
6645			goto out_unlock;
6646		ret = log_delayed_deletion_items(trans, inode, path,
6647						 &delayed_del_list, ctx);
6648		if (ret)
6649			goto out_unlock;
6650	}
6651
6652	spin_lock(&inode->lock);
6653	inode->logged_trans = trans->transid;
6654	/*
6655	 * Don't update last_log_commit if we logged that an inode exists.
6656	 * We do this for three reasons:
6657	 *
6658	 * 1) We might have had buffered writes to this inode that were
6659	 *    flushed and had their ordered extents completed in this
6660	 *    transaction, but we did not previously log the inode with
6661	 *    LOG_INODE_ALL. Later the inode was evicted and after that
6662	 *    it was loaded again and this LOG_INODE_EXISTS log operation
6663	 *    happened. We must make sure that if an explicit fsync against
6664	 *    the inode is performed later, it logs the new extents, an
6665	 *    updated inode item, etc, and syncs the log. The same logic
6666	 *    applies to direct IO writes instead of buffered writes.
6667	 *
6668	 * 2) When we log the inode with LOG_INODE_EXISTS, its inode item
6669	 *    is logged with an i_size of 0 or whatever value was logged
6670	 *    before. If later the i_size of the inode is increased by a
6671	 *    truncate operation, the log is synced through an fsync of
6672	 *    some other inode and then finally an explicit fsync against
6673	 *    this inode is made, we must make sure this fsync logs the
6674	 *    inode with the new i_size, the hole between old i_size and
6675	 *    the new i_size, and syncs the log.
6676	 *
6677	 * 3) If we are logging that an ancestor inode exists as part of
6678	 *    logging a new name from a link or rename operation, don't update
6679	 *    its last_log_commit - otherwise if an explicit fsync is made
6680	 *    against an ancestor, the fsync considers the inode in the log
6681	 *    and doesn't sync the log, resulting in the ancestor missing after
6682	 *    a power failure unless the log was synced as part of an fsync
6683	 *    against any other unrelated inode.
6684	 */
6685	if (inode_only != LOG_INODE_EXISTS)
6686		inode->last_log_commit = inode->last_sub_trans;
6687	spin_unlock(&inode->lock);
6688
6689	/*
6690	 * Reset the last_reflink_trans so that the next fsync does not need to
6691	 * go through the slower path when logging extents and their checksums.
6692	 */
6693	if (inode_only == LOG_INODE_ALL)
6694		inode->last_reflink_trans = 0;
6695
6696out_unlock:
6697	mutex_unlock(&inode->log_mutex);
6698out:
6699	btrfs_free_path(path);
6700	btrfs_free_path(dst_path);
6701
6702	if (ret)
6703		free_conflicting_inodes(ctx);
6704	else
6705		ret = log_conflicting_inodes(trans, inode->root, ctx);
6706
6707	if (full_dir_logging && !ctx->logging_new_delayed_dentries) {
6708		if (!ret)
6709			ret = log_new_delayed_dentries(trans, inode,
6710						       &delayed_ins_list, ctx);
6711
6712		btrfs_log_put_delayed_items(inode, &delayed_ins_list,
6713					    &delayed_del_list);
6714	}
6715
6716	return ret;
6717}
6718
6719static int btrfs_log_all_parents(struct btrfs_trans_handle *trans,
6720				 struct btrfs_inode *inode,
6721				 struct btrfs_log_ctx *ctx)
6722{
6723	struct btrfs_fs_info *fs_info = trans->fs_info;
6724	int ret;
6725	struct btrfs_path *path;
6726	struct btrfs_key key;
6727	struct btrfs_root *root = inode->root;
6728	const u64 ino = btrfs_ino(inode);
6729
6730	path = btrfs_alloc_path();
6731	if (!path)
6732		return -ENOMEM;
6733	path->skip_locking = 1;
6734	path->search_commit_root = 1;
6735
6736	key.objectid = ino;
6737	key.type = BTRFS_INODE_REF_KEY;
6738	key.offset = 0;
6739	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
6740	if (ret < 0)
6741		goto out;
6742
6743	while (true) {
6744		struct extent_buffer *leaf = path->nodes[0];
6745		int slot = path->slots[0];
6746		u32 cur_offset = 0;
6747		u32 item_size;
6748		unsigned long ptr;
6749
6750		if (slot >= btrfs_header_nritems(leaf)) {
6751			ret = btrfs_next_leaf(root, path);
6752			if (ret < 0)
6753				goto out;
6754			else if (ret > 0)
6755				break;
6756			continue;
6757		}
6758
6759		btrfs_item_key_to_cpu(leaf, &key, slot);
6760		/* BTRFS_INODE_EXTREF_KEY is BTRFS_INODE_REF_KEY + 1 */
6761		if (key.objectid != ino || key.type > BTRFS_INODE_EXTREF_KEY)
6762			break;
6763
6764		item_size = btrfs_item_size(leaf, slot);
6765		ptr = btrfs_item_ptr_offset(leaf, slot);
6766		while (cur_offset < item_size) {
6767			struct btrfs_key inode_key;
6768			struct inode *dir_inode;
6769
6770			inode_key.type = BTRFS_INODE_ITEM_KEY;
6771			inode_key.offset = 0;
6772
6773			if (key.type == BTRFS_INODE_EXTREF_KEY) {
6774				struct btrfs_inode_extref *extref;
6775
6776				extref = (struct btrfs_inode_extref *)
6777					(ptr + cur_offset);
6778				inode_key.objectid = btrfs_inode_extref_parent(
6779					leaf, extref);
6780				cur_offset += sizeof(*extref);
6781				cur_offset += btrfs_inode_extref_name_len(leaf,
6782					extref);
6783			} else {
6784				inode_key.objectid = key.offset;
6785				cur_offset = item_size;
6786			}
6787
6788			dir_inode = btrfs_iget(fs_info->sb, inode_key.objectid,
6789					       root);
6790			/*
6791			 * If the parent inode was deleted, return an error to
6792			 * fallback to a transaction commit. This is to prevent
6793			 * getting an inode that was moved from one parent A to
6794			 * a parent B, got its former parent A deleted and then
6795			 * it got fsync'ed, from existing at both parents after
6796			 * a log replay (and the old parent still existing).
6797			 * Example:
6798			 *
6799			 * mkdir /mnt/A
6800			 * mkdir /mnt/B
6801			 * touch /mnt/B/bar
6802			 * sync
6803			 * mv /mnt/B/bar /mnt/A/bar
6804			 * mv -T /mnt/A /mnt/B
6805			 * fsync /mnt/B/bar
6806			 * <power fail>
6807			 *
6808			 * If we ignore the old parent B which got deleted,
6809			 * after a log replay we would have file bar linked
6810			 * at both parents and the old parent B would still
6811			 * exist.
6812			 */
6813			if (IS_ERR(dir_inode)) {
6814				ret = PTR_ERR(dir_inode);
6815				goto out;
6816			}
6817
6818			if (!need_log_inode(trans, BTRFS_I(dir_inode))) {
6819				btrfs_add_delayed_iput(BTRFS_I(dir_inode));
6820				continue;
6821			}
6822
6823			ctx->log_new_dentries = false;
6824			ret = btrfs_log_inode(trans, BTRFS_I(dir_inode),
6825					      LOG_INODE_ALL, ctx);
6826			if (!ret && ctx->log_new_dentries)
6827				ret = log_new_dir_dentries(trans,
6828						   BTRFS_I(dir_inode), ctx);
6829			btrfs_add_delayed_iput(BTRFS_I(dir_inode));
6830			if (ret)
6831				goto out;
6832		}
6833		path->slots[0]++;
6834	}
6835	ret = 0;
6836out:
6837	btrfs_free_path(path);
6838	return ret;
6839}
6840
6841static int log_new_ancestors(struct btrfs_trans_handle *trans,
6842			     struct btrfs_root *root,
6843			     struct btrfs_path *path,
6844			     struct btrfs_log_ctx *ctx)
6845{
6846	struct btrfs_key found_key;
6847
6848	btrfs_item_key_to_cpu(path->nodes[0], &found_key, path->slots[0]);
6849
6850	while (true) {
6851		struct btrfs_fs_info *fs_info = root->fs_info;
6852		struct extent_buffer *leaf;
6853		int slot;
6854		struct btrfs_key search_key;
6855		struct inode *inode;
6856		u64 ino;
6857		int ret = 0;
6858
6859		btrfs_release_path(path);
6860
6861		ino = found_key.offset;
6862
6863		search_key.objectid = found_key.offset;
6864		search_key.type = BTRFS_INODE_ITEM_KEY;
6865		search_key.offset = 0;
6866		inode = btrfs_iget(fs_info->sb, ino, root);
6867		if (IS_ERR(inode))
6868			return PTR_ERR(inode);
6869
6870		if (BTRFS_I(inode)->generation >= trans->transid &&
6871		    need_log_inode(trans, BTRFS_I(inode)))
6872			ret = btrfs_log_inode(trans, BTRFS_I(inode),
6873					      LOG_INODE_EXISTS, ctx);
6874		btrfs_add_delayed_iput(BTRFS_I(inode));
6875		if (ret)
6876			return ret;
6877
6878		if (search_key.objectid == BTRFS_FIRST_FREE_OBJECTID)
6879			break;
6880
6881		search_key.type = BTRFS_INODE_REF_KEY;
6882		ret = btrfs_search_slot(NULL, root, &search_key, path, 0, 0);
6883		if (ret < 0)
6884			return ret;
6885
6886		leaf = path->nodes[0];
6887		slot = path->slots[0];
6888		if (slot >= btrfs_header_nritems(leaf)) {
6889			ret = btrfs_next_leaf(root, path);
6890			if (ret < 0)
6891				return ret;
6892			else if (ret > 0)
6893				return -ENOENT;
6894			leaf = path->nodes[0];
6895			slot = path->slots[0];
6896		}
6897
6898		btrfs_item_key_to_cpu(leaf, &found_key, slot);
6899		if (found_key.objectid != search_key.objectid ||
6900		    found_key.type != BTRFS_INODE_REF_KEY)
6901			return -ENOENT;
6902	}
6903	return 0;
6904}
6905
6906static int log_new_ancestors_fast(struct btrfs_trans_handle *trans,
6907				  struct btrfs_inode *inode,
6908				  struct dentry *parent,
6909				  struct btrfs_log_ctx *ctx)
6910{
6911	struct btrfs_root *root = inode->root;
6912	struct dentry *old_parent = NULL;
6913	struct super_block *sb = inode->vfs_inode.i_sb;
6914	int ret = 0;
6915
6916	while (true) {
6917		if (!parent || d_really_is_negative(parent) ||
6918		    sb != parent->d_sb)
6919			break;
6920
6921		inode = BTRFS_I(d_inode(parent));
6922		if (root != inode->root)
6923			break;
6924
6925		if (inode->generation >= trans->transid &&
6926		    need_log_inode(trans, inode)) {
6927			ret = btrfs_log_inode(trans, inode,
6928					      LOG_INODE_EXISTS, ctx);
6929			if (ret)
6930				break;
6931		}
6932		if (IS_ROOT(parent))
6933			break;
6934
6935		parent = dget_parent(parent);
6936		dput(old_parent);
6937		old_parent = parent;
6938	}
6939	dput(old_parent);
6940
6941	return ret;
6942}
6943
6944static int log_all_new_ancestors(struct btrfs_trans_handle *trans,
6945				 struct btrfs_inode *inode,
6946				 struct dentry *parent,
6947				 struct btrfs_log_ctx *ctx)
6948{
6949	struct btrfs_root *root = inode->root;
6950	const u64 ino = btrfs_ino(inode);
6951	struct btrfs_path *path;
6952	struct btrfs_key search_key;
6953	int ret;
6954
6955	/*
6956	 * For a single hard link case, go through a fast path that does not
6957	 * need to iterate the fs/subvolume tree.
6958	 */
6959	if (inode->vfs_inode.i_nlink < 2)
6960		return log_new_ancestors_fast(trans, inode, parent, ctx);
6961
6962	path = btrfs_alloc_path();
6963	if (!path)
6964		return -ENOMEM;
6965
6966	search_key.objectid = ino;
6967	search_key.type = BTRFS_INODE_REF_KEY;
6968	search_key.offset = 0;
6969again:
6970	ret = btrfs_search_slot(NULL, root, &search_key, path, 0, 0);
6971	if (ret < 0)
6972		goto out;
6973	if (ret == 0)
6974		path->slots[0]++;
6975
6976	while (true) {
6977		struct extent_buffer *leaf = path->nodes[0];
6978		int slot = path->slots[0];
6979		struct btrfs_key found_key;
6980
6981		if (slot >= btrfs_header_nritems(leaf)) {
6982			ret = btrfs_next_leaf(root, path);
6983			if (ret < 0)
6984				goto out;
6985			else if (ret > 0)
6986				break;
6987			continue;
6988		}
6989
6990		btrfs_item_key_to_cpu(leaf, &found_key, slot);
6991		if (found_key.objectid != ino ||
6992		    found_key.type > BTRFS_INODE_EXTREF_KEY)
6993			break;
6994
6995		/*
6996		 * Don't deal with extended references because they are rare
6997		 * cases and too complex to deal with (we would need to keep
6998		 * track of which subitem we are processing for each item in
6999		 * this loop, etc). So just return some error to fallback to
7000		 * a transaction commit.
7001		 */
7002		if (found_key.type == BTRFS_INODE_EXTREF_KEY) {
7003			ret = -EMLINK;
7004			goto out;
7005		}
7006
7007		/*
7008		 * Logging ancestors needs to do more searches on the fs/subvol
7009		 * tree, so it releases the path as needed to avoid deadlocks.
7010		 * Keep track of the last inode ref key and resume from that key
7011		 * after logging all new ancestors for the current hard link.
7012		 */
7013		memcpy(&search_key, &found_key, sizeof(search_key));
7014
7015		ret = log_new_ancestors(trans, root, path, ctx);
7016		if (ret)
7017			goto out;
7018		btrfs_release_path(path);
7019		goto again;
7020	}
7021	ret = 0;
7022out:
7023	btrfs_free_path(path);
7024	return ret;
7025}
7026
7027/*
7028 * helper function around btrfs_log_inode to make sure newly created
7029 * parent directories also end up in the log.  A minimal inode and backref
7030 * only logging is done of any parent directories that are older than
7031 * the last committed transaction
7032 */
7033static int btrfs_log_inode_parent(struct btrfs_trans_handle *trans,
7034				  struct btrfs_inode *inode,
7035				  struct dentry *parent,
7036				  int inode_only,
7037				  struct btrfs_log_ctx *ctx)
7038{
7039	struct btrfs_root *root = inode->root;
7040	struct btrfs_fs_info *fs_info = root->fs_info;
7041	int ret = 0;
7042	bool log_dentries = false;
7043
7044	if (btrfs_test_opt(fs_info, NOTREELOG)) {
7045		ret = BTRFS_LOG_FORCE_COMMIT;
7046		goto end_no_trans;
7047	}
7048
7049	if (btrfs_root_refs(&root->root_item) == 0) {
7050		ret = BTRFS_LOG_FORCE_COMMIT;
7051		goto end_no_trans;
7052	}
7053
7054	/*
7055	 * Skip already logged inodes or inodes corresponding to tmpfiles
7056	 * (since logging them is pointless, a link count of 0 means they
7057	 * will never be accessible).
7058	 */
7059	if ((btrfs_inode_in_log(inode, trans->transid) &&
7060	     list_empty(&ctx->ordered_extents)) ||
7061	    inode->vfs_inode.i_nlink == 0) {
7062		ret = BTRFS_NO_LOG_SYNC;
7063		goto end_no_trans;
7064	}
7065
7066	ret = start_log_trans(trans, root, ctx);
7067	if (ret)
7068		goto end_no_trans;
7069
7070	ret = btrfs_log_inode(trans, inode, inode_only, ctx);
7071	if (ret)
7072		goto end_trans;
7073
7074	/*
7075	 * for regular files, if its inode is already on disk, we don't
7076	 * have to worry about the parents at all.  This is because
7077	 * we can use the last_unlink_trans field to record renames
7078	 * and other fun in this file.
7079	 */
7080	if (S_ISREG(inode->vfs_inode.i_mode) &&
7081	    inode->generation < trans->transid &&
7082	    inode->last_unlink_trans < trans->transid) {
7083		ret = 0;
7084		goto end_trans;
7085	}
7086
7087	if (S_ISDIR(inode->vfs_inode.i_mode) && ctx->log_new_dentries)
7088		log_dentries = true;
7089
7090	/*
7091	 * On unlink we must make sure all our current and old parent directory
7092	 * inodes are fully logged. This is to prevent leaving dangling
7093	 * directory index entries in directories that were our parents but are
7094	 * not anymore. Not doing this results in old parent directory being
7095	 * impossible to delete after log replay (rmdir will always fail with
7096	 * error -ENOTEMPTY).
7097	 *
7098	 * Example 1:
7099	 *
7100	 * mkdir testdir
7101	 * touch testdir/foo
7102	 * ln testdir/foo testdir/bar
7103	 * sync
7104	 * unlink testdir/bar
7105	 * xfs_io -c fsync testdir/foo
7106	 * <power failure>
7107	 * mount fs, triggers log replay
7108	 *
7109	 * If we don't log the parent directory (testdir), after log replay the
7110	 * directory still has an entry pointing to the file inode using the bar
7111	 * name, but a matching BTRFS_INODE_[REF|EXTREF]_KEY does not exist and
7112	 * the file inode has a link count of 1.
7113	 *
7114	 * Example 2:
7115	 *
7116	 * mkdir testdir
7117	 * touch foo
7118	 * ln foo testdir/foo2
7119	 * ln foo testdir/foo3
7120	 * sync
7121	 * unlink testdir/foo3
7122	 * xfs_io -c fsync foo
7123	 * <power failure>
7124	 * mount fs, triggers log replay
7125	 *
7126	 * Similar as the first example, after log replay the parent directory
7127	 * testdir still has an entry pointing to the inode file with name foo3
7128	 * but the file inode does not have a matching BTRFS_INODE_REF_KEY item
7129	 * and has a link count of 2.
7130	 */
7131	if (inode->last_unlink_trans >= trans->transid) {
7132		ret = btrfs_log_all_parents(trans, inode, ctx);
7133		if (ret)
7134			goto end_trans;
7135	}
7136
7137	ret = log_all_new_ancestors(trans, inode, parent, ctx);
7138	if (ret)
7139		goto end_trans;
7140
7141	if (log_dentries)
7142		ret = log_new_dir_dentries(trans, inode, ctx);
7143	else
7144		ret = 0;
7145end_trans:
7146	if (ret < 0) {
7147		btrfs_set_log_full_commit(trans);
7148		ret = BTRFS_LOG_FORCE_COMMIT;
7149	}
7150
7151	if (ret)
7152		btrfs_remove_log_ctx(root, ctx);
7153	btrfs_end_log_trans(root);
7154end_no_trans:
7155	return ret;
7156}
7157
7158/*
7159 * it is not safe to log dentry if the chunk root has added new
7160 * chunks.  This returns 0 if the dentry was logged, and 1 otherwise.
7161 * If this returns 1, you must commit the transaction to safely get your
7162 * data on disk.
7163 */
7164int btrfs_log_dentry_safe(struct btrfs_trans_handle *trans,
7165			  struct dentry *dentry,
7166			  struct btrfs_log_ctx *ctx)
7167{
7168	struct dentry *parent = dget_parent(dentry);
7169	int ret;
7170
7171	ret = btrfs_log_inode_parent(trans, BTRFS_I(d_inode(dentry)), parent,
7172				     LOG_INODE_ALL, ctx);
7173	dput(parent);
7174
7175	return ret;
7176}
7177
7178/*
7179 * should be called during mount to recover any replay any log trees
7180 * from the FS
7181 */
7182int btrfs_recover_log_trees(struct btrfs_root *log_root_tree)
7183{
7184	int ret;
7185	struct btrfs_path *path;
7186	struct btrfs_trans_handle *trans;
7187	struct btrfs_key key;
7188	struct btrfs_key found_key;
7189	struct btrfs_root *log;
7190	struct btrfs_fs_info *fs_info = log_root_tree->fs_info;
7191	struct walk_control wc = {
7192		.process_func = process_one_buffer,
7193		.stage = LOG_WALK_PIN_ONLY,
7194	};
7195
7196	path = btrfs_alloc_path();
7197	if (!path)
7198		return -ENOMEM;
7199
7200	set_bit(BTRFS_FS_LOG_RECOVERING, &fs_info->flags);
7201
7202	trans = btrfs_start_transaction(fs_info->tree_root, 0);
7203	if (IS_ERR(trans)) {
7204		ret = PTR_ERR(trans);
7205		goto error;
7206	}
7207
7208	wc.trans = trans;
7209	wc.pin = 1;
7210
7211	ret = walk_log_tree(trans, log_root_tree, &wc);
7212	if (ret) {
7213		btrfs_abort_transaction(trans, ret);
7214		goto error;
7215	}
7216
7217again:
7218	key.objectid = BTRFS_TREE_LOG_OBJECTID;
7219	key.offset = (u64)-1;
7220	key.type = BTRFS_ROOT_ITEM_KEY;
7221
7222	while (1) {
7223		ret = btrfs_search_slot(NULL, log_root_tree, &key, path, 0, 0);
7224
7225		if (ret < 0) {
7226			btrfs_abort_transaction(trans, ret);
7227			goto error;
7228		}
7229		if (ret > 0) {
7230			if (path->slots[0] == 0)
7231				break;
7232			path->slots[0]--;
7233		}
7234		btrfs_item_key_to_cpu(path->nodes[0], &found_key,
7235				      path->slots[0]);
7236		btrfs_release_path(path);
7237		if (found_key.objectid != BTRFS_TREE_LOG_OBJECTID)
7238			break;
7239
7240		log = btrfs_read_tree_root(log_root_tree, &found_key);
7241		if (IS_ERR(log)) {
7242			ret = PTR_ERR(log);
7243			btrfs_abort_transaction(trans, ret);
7244			goto error;
7245		}
7246
7247		wc.replay_dest = btrfs_get_fs_root(fs_info, found_key.offset,
7248						   true);
7249		if (IS_ERR(wc.replay_dest)) {
7250			ret = PTR_ERR(wc.replay_dest);
7251
7252			/*
7253			 * We didn't find the subvol, likely because it was
7254			 * deleted.  This is ok, simply skip this log and go to
7255			 * the next one.
7256			 *
7257			 * We need to exclude the root because we can't have
7258			 * other log replays overwriting this log as we'll read
7259			 * it back in a few more times.  This will keep our
7260			 * block from being modified, and we'll just bail for
7261			 * each subsequent pass.
7262			 */
7263			if (ret == -ENOENT)
7264				ret = btrfs_pin_extent_for_log_replay(trans, log->node);
7265			btrfs_put_root(log);
7266
7267			if (!ret)
7268				goto next;
7269			btrfs_abort_transaction(trans, ret);
7270			goto error;
7271		}
7272
7273		wc.replay_dest->log_root = log;
7274		ret = btrfs_record_root_in_trans(trans, wc.replay_dest);
7275		if (ret)
7276			/* The loop needs to continue due to the root refs */
7277			btrfs_abort_transaction(trans, ret);
7278		else
7279			ret = walk_log_tree(trans, log, &wc);
7280
7281		if (!ret && wc.stage == LOG_WALK_REPLAY_ALL) {
7282			ret = fixup_inode_link_counts(trans, wc.replay_dest,
7283						      path);
7284			if (ret)
7285				btrfs_abort_transaction(trans, ret);
7286		}
7287
7288		if (!ret && wc.stage == LOG_WALK_REPLAY_ALL) {
7289			struct btrfs_root *root = wc.replay_dest;
7290
7291			btrfs_release_path(path);
7292
7293			/*
7294			 * We have just replayed everything, and the highest
7295			 * objectid of fs roots probably has changed in case
7296			 * some inode_item's got replayed.
7297			 *
7298			 * root->objectid_mutex is not acquired as log replay
7299			 * could only happen during mount.
7300			 */
7301			ret = btrfs_init_root_free_objectid(root);
7302			if (ret)
7303				btrfs_abort_transaction(trans, ret);
7304		}
7305
7306		wc.replay_dest->log_root = NULL;
7307		btrfs_put_root(wc.replay_dest);
7308		btrfs_put_root(log);
7309
7310		if (ret)
7311			goto error;
7312next:
7313		if (found_key.offset == 0)
7314			break;
7315		key.offset = found_key.offset - 1;
7316	}
7317	btrfs_release_path(path);
7318
7319	/* step one is to pin it all, step two is to replay just inodes */
7320	if (wc.pin) {
7321		wc.pin = 0;
7322		wc.process_func = replay_one_buffer;
7323		wc.stage = LOG_WALK_REPLAY_INODES;
7324		goto again;
7325	}
7326	/* step three is to replay everything */
7327	if (wc.stage < LOG_WALK_REPLAY_ALL) {
7328		wc.stage++;
7329		goto again;
7330	}
7331
7332	btrfs_free_path(path);
7333
7334	/* step 4: commit the transaction, which also unpins the blocks */
7335	ret = btrfs_commit_transaction(trans);
7336	if (ret)
7337		return ret;
7338
7339	log_root_tree->log_root = NULL;
7340	clear_bit(BTRFS_FS_LOG_RECOVERING, &fs_info->flags);
7341	btrfs_put_root(log_root_tree);
7342
7343	return 0;
7344error:
7345	if (wc.trans)
7346		btrfs_end_transaction(wc.trans);
7347	clear_bit(BTRFS_FS_LOG_RECOVERING, &fs_info->flags);
7348	btrfs_free_path(path);
7349	return ret;
7350}
7351
7352/*
7353 * there are some corner cases where we want to force a full
7354 * commit instead of allowing a directory to be logged.
7355 *
7356 * They revolve around files there were unlinked from the directory, and
7357 * this function updates the parent directory so that a full commit is
7358 * properly done if it is fsync'd later after the unlinks are done.
7359 *
7360 * Must be called before the unlink operations (updates to the subvolume tree,
7361 * inodes, etc) are done.
7362 */
7363void btrfs_record_unlink_dir(struct btrfs_trans_handle *trans,
7364			     struct btrfs_inode *dir, struct btrfs_inode *inode,
7365			     bool for_rename)
7366{
7367	/*
7368	 * when we're logging a file, if it hasn't been renamed
7369	 * or unlinked, and its inode is fully committed on disk,
7370	 * we don't have to worry about walking up the directory chain
7371	 * to log its parents.
7372	 *
7373	 * So, we use the last_unlink_trans field to put this transid
7374	 * into the file.  When the file is logged we check it and
7375	 * don't log the parents if the file is fully on disk.
7376	 */
7377	mutex_lock(&inode->log_mutex);
7378	inode->last_unlink_trans = trans->transid;
7379	mutex_unlock(&inode->log_mutex);
7380
7381	if (!for_rename)
7382		return;
7383
7384	/*
7385	 * If this directory was already logged, any new names will be logged
7386	 * with btrfs_log_new_name() and old names will be deleted from the log
7387	 * tree with btrfs_del_dir_entries_in_log() or with
7388	 * btrfs_del_inode_ref_in_log().
7389	 */
7390	if (inode_logged(trans, dir, NULL) == 1)
7391		return;
7392
7393	/*
7394	 * If the inode we're about to unlink was logged before, the log will be
7395	 * properly updated with the new name with btrfs_log_new_name() and the
7396	 * old name removed with btrfs_del_dir_entries_in_log() or with
7397	 * btrfs_del_inode_ref_in_log().
7398	 */
7399	if (inode_logged(trans, inode, NULL) == 1)
7400		return;
7401
7402	/*
7403	 * when renaming files across directories, if the directory
7404	 * there we're unlinking from gets fsync'd later on, there's
7405	 * no way to find the destination directory later and fsync it
7406	 * properly.  So, we have to be conservative and force commits
7407	 * so the new name gets discovered.
7408	 */
7409	mutex_lock(&dir->log_mutex);
7410	dir->last_unlink_trans = trans->transid;
7411	mutex_unlock(&dir->log_mutex);
7412}
7413
7414/*
7415 * Make sure that if someone attempts to fsync the parent directory of a deleted
7416 * snapshot, it ends up triggering a transaction commit. This is to guarantee
7417 * that after replaying the log tree of the parent directory's root we will not
7418 * see the snapshot anymore and at log replay time we will not see any log tree
7419 * corresponding to the deleted snapshot's root, which could lead to replaying
7420 * it after replaying the log tree of the parent directory (which would replay
7421 * the snapshot delete operation).
7422 *
7423 * Must be called before the actual snapshot destroy operation (updates to the
7424 * parent root and tree of tree roots trees, etc) are done.
7425 */
7426void btrfs_record_snapshot_destroy(struct btrfs_trans_handle *trans,
7427				   struct btrfs_inode *dir)
7428{
7429	mutex_lock(&dir->log_mutex);
7430	dir->last_unlink_trans = trans->transid;
7431	mutex_unlock(&dir->log_mutex);
7432}
7433
7434/*
7435 * Update the log after adding a new name for an inode.
7436 *
7437 * @trans:              Transaction handle.
7438 * @old_dentry:         The dentry associated with the old name and the old
7439 *                      parent directory.
7440 * @old_dir:            The inode of the previous parent directory for the case
7441 *                      of a rename. For a link operation, it must be NULL.
7442 * @old_dir_index:      The index number associated with the old name, meaningful
7443 *                      only for rename operations (when @old_dir is not NULL).
7444 *                      Ignored for link operations.
7445 * @parent:             The dentry associated with the directory under which the
7446 *                      new name is located.
7447 *
7448 * Call this after adding a new name for an inode, as a result of a link or
7449 * rename operation, and it will properly update the log to reflect the new name.
7450 */
7451void btrfs_log_new_name(struct btrfs_trans_handle *trans,
7452			struct dentry *old_dentry, struct btrfs_inode *old_dir,
7453			u64 old_dir_index, struct dentry *parent)
7454{
7455	struct btrfs_inode *inode = BTRFS_I(d_inode(old_dentry));
7456	struct btrfs_root *root = inode->root;
7457	struct btrfs_log_ctx ctx;
7458	bool log_pinned = false;
7459	int ret;
7460
7461	/*
7462	 * this will force the logging code to walk the dentry chain
7463	 * up for the file
7464	 */
7465	if (!S_ISDIR(inode->vfs_inode.i_mode))
7466		inode->last_unlink_trans = trans->transid;
7467
7468	/*
7469	 * if this inode hasn't been logged and directory we're renaming it
7470	 * from hasn't been logged, we don't need to log it
7471	 */
7472	ret = inode_logged(trans, inode, NULL);
7473	if (ret < 0) {
7474		goto out;
7475	} else if (ret == 0) {
7476		if (!old_dir)
7477			return;
7478		/*
7479		 * If the inode was not logged and we are doing a rename (old_dir is not
7480		 * NULL), check if old_dir was logged - if it was not we can return and
7481		 * do nothing.
7482		 */
7483		ret = inode_logged(trans, old_dir, NULL);
7484		if (ret < 0)
7485			goto out;
7486		else if (ret == 0)
7487			return;
7488	}
7489	ret = 0;
7490
7491	/*
7492	 * If we are doing a rename (old_dir is not NULL) from a directory that
7493	 * was previously logged, make sure that on log replay we get the old
7494	 * dir entry deleted. This is needed because we will also log the new
7495	 * name of the renamed inode, so we need to make sure that after log
7496	 * replay we don't end up with both the new and old dir entries existing.
7497	 */
7498	if (old_dir && old_dir->logged_trans == trans->transid) {
7499		struct btrfs_root *log = old_dir->root->log_root;
7500		struct btrfs_path *path;
7501		struct fscrypt_name fname;
7502
7503		ASSERT(old_dir_index >= BTRFS_DIR_START_INDEX);
7504
7505		ret = fscrypt_setup_filename(&old_dir->vfs_inode,
7506					     &old_dentry->d_name, 0, &fname);
7507		if (ret)
7508			goto out;
7509		/*
7510		 * We have two inodes to update in the log, the old directory and
7511		 * the inode that got renamed, so we must pin the log to prevent
7512		 * anyone from syncing the log until we have updated both inodes
7513		 * in the log.
7514		 */
7515		ret = join_running_log_trans(root);
7516		/*
7517		 * At least one of the inodes was logged before, so this should
7518		 * not fail, but if it does, it's not serious, just bail out and
7519		 * mark the log for a full commit.
7520		 */
7521		if (WARN_ON_ONCE(ret < 0)) {
7522			fscrypt_free_filename(&fname);
7523			goto out;
7524		}
7525
7526		log_pinned = true;
7527
7528		path = btrfs_alloc_path();
7529		if (!path) {
7530			ret = -ENOMEM;
7531			fscrypt_free_filename(&fname);
7532			goto out;
7533		}
7534
7535		/*
7536		 * Other concurrent task might be logging the old directory,
7537		 * as it can be triggered when logging other inode that had or
7538		 * still has a dentry in the old directory. We lock the old
7539		 * directory's log_mutex to ensure the deletion of the old
7540		 * name is persisted, because during directory logging we
7541		 * delete all BTRFS_DIR_LOG_INDEX_KEY keys and the deletion of
7542		 * the old name's dir index item is in the delayed items, so
7543		 * it could be missed by an in progress directory logging.
7544		 */
7545		mutex_lock(&old_dir->log_mutex);
7546		ret = del_logged_dentry(trans, log, path, btrfs_ino(old_dir),
7547					&fname.disk_name, old_dir_index);
7548		if (ret > 0) {
7549			/*
7550			 * The dentry does not exist in the log, so record its
7551			 * deletion.
7552			 */
7553			btrfs_release_path(path);
7554			ret = insert_dir_log_key(trans, log, path,
7555						 btrfs_ino(old_dir),
7556						 old_dir_index, old_dir_index);
7557		}
7558		mutex_unlock(&old_dir->log_mutex);
7559
7560		btrfs_free_path(path);
7561		fscrypt_free_filename(&fname);
7562		if (ret < 0)
7563			goto out;
7564	}
7565
7566	btrfs_init_log_ctx(&ctx, &inode->vfs_inode);
7567	ctx.logging_new_name = true;
7568	btrfs_init_log_ctx_scratch_eb(&ctx);
7569	/*
7570	 * We don't care about the return value. If we fail to log the new name
7571	 * then we know the next attempt to sync the log will fallback to a full
7572	 * transaction commit (due to a call to btrfs_set_log_full_commit()), so
7573	 * we don't need to worry about getting a log committed that has an
7574	 * inconsistent state after a rename operation.
7575	 */
7576	btrfs_log_inode_parent(trans, inode, parent, LOG_INODE_EXISTS, &ctx);
7577	free_extent_buffer(ctx.scratch_eb);
7578	ASSERT(list_empty(&ctx.conflict_inodes));
7579out:
7580	/*
7581	 * If an error happened mark the log for a full commit because it's not
7582	 * consistent and up to date or we couldn't find out if one of the
7583	 * inodes was logged before in this transaction. Do it before unpinning
7584	 * the log, to avoid any races with someone else trying to commit it.
7585	 */
7586	if (ret < 0)
7587		btrfs_set_log_full_commit(trans);
7588	if (log_pinned)
7589		btrfs_end_log_trans(root);
7590}
7591