send.c 132.4 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
/*
 * Copyright (C) 2012 Alexander Block.  All rights reserved.
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public
 * License v2 as published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * General Public License for more details.
 *
 * You should have received a copy of the GNU General Public
 * License along with this program; if not, write to the
 * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
 * Boston, MA 021110-1307, USA.
 */

#include <linux/bsearch.h>
#include <linux/fs.h>
#include <linux/file.h>
#include <linux/sort.h>
#include <linux/mount.h>
#include <linux/xattr.h>
#include <linux/posix_acl_xattr.h>
#include <linux/radix-tree.h>
27
#include <linux/vmalloc.h>
A
Andy Shevchenko 已提交
28
#include <linux/string.h>
29 30 31

#include "send.h"
#include "backref.h"
32
#include "hash.h"
33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55
#include "locking.h"
#include "disk-io.h"
#include "btrfs_inode.h"
#include "transaction.h"

static int g_verbose = 0;

#define verbose_printk(...) if (g_verbose) printk(__VA_ARGS__)

/*
 * A fs_path is a helper to dynamically build path names with unknown size.
 * It reallocates the internal buffer on demand.
 * It allows fast adding of path elements on the right side (normal path) and
 * fast adding to the left side (reversed path). A reversed path can also be
 * unreversed if needed.
 */
struct fs_path {
	union {
		struct {
			char *start;
			char *end;

			char *buf;
56 57
			unsigned short buf_len:15;
			unsigned short reversed:1;
58 59
			char inline_buf[];
		};
60 61 62 63 64 65
		/*
		 * Average path length does not exceed 200 bytes, we'll have
		 * better packing in the slab and higher chance to satisfy
		 * a allocation later during send.
		 */
		char pad[256];
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91
	};
};
#define FS_PATH_INLINE_SIZE \
	(sizeof(struct fs_path) - offsetof(struct fs_path, inline_buf))


/* reused for each extent */
struct clone_root {
	struct btrfs_root *root;
	u64 ino;
	u64 offset;

	u64 found_refs;
};

#define SEND_CTX_MAX_NAME_CACHE_SIZE 128
#define SEND_CTX_NAME_CACHE_CLEAN_SIZE (SEND_CTX_MAX_NAME_CACHE_SIZE * 2)

struct send_ctx {
	struct file *send_filp;
	loff_t send_off;
	char *send_buf;
	u32 send_size;
	u32 send_max_size;
	u64 total_send_size;
	u64 cmd_send_size[BTRFS_SEND_C_MAX + 1];
92
	u64 flags;	/* 'flags' member of btrfs_ioctl_send_args is u64 */
93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114

	struct btrfs_root *send_root;
	struct btrfs_root *parent_root;
	struct clone_root *clone_roots;
	int clone_roots_cnt;

	/* current state of the compare_tree call */
	struct btrfs_path *left_path;
	struct btrfs_path *right_path;
	struct btrfs_key *cmp_key;

	/*
	 * infos of the currently processed inode. In case of deleted inodes,
	 * these are the values from the deleted inode.
	 */
	u64 cur_ino;
	u64 cur_inode_gen;
	int cur_inode_new;
	int cur_inode_new_gen;
	int cur_inode_deleted;
	u64 cur_inode_size;
	u64 cur_inode_mode;
L
Liu Bo 已提交
115
	u64 cur_inode_rdev;
116
	u64 cur_inode_last_extent;
117 118 119 120 121 122 123 124 125 126

	u64 send_progress;

	struct list_head new_refs;
	struct list_head deleted_refs;

	struct radix_tree_root name_cache;
	struct list_head name_cache_list;
	int name_cache_size;

L
Liu Bo 已提交
127 128
	struct file_ra_state ra;

129
	char *read_buf;
130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183

	/*
	 * We process inodes by their increasing order, so if before an
	 * incremental send we reverse the parent/child relationship of
	 * directories such that a directory with a lower inode number was
	 * the parent of a directory with a higher inode number, and the one
	 * becoming the new parent got renamed too, we can't rename/move the
	 * directory with lower inode number when we finish processing it - we
	 * must process the directory with higher inode number first, then
	 * rename/move it and then rename/move the directory with lower inode
	 * number. Example follows.
	 *
	 * Tree state when the first send was performed:
	 *
	 * .
	 * |-- a                   (ino 257)
	 *     |-- b               (ino 258)
	 *         |
	 *         |
	 *         |-- c           (ino 259)
	 *         |   |-- d       (ino 260)
	 *         |
	 *         |-- c2          (ino 261)
	 *
	 * Tree state when the second (incremental) send is performed:
	 *
	 * .
	 * |-- a                   (ino 257)
	 *     |-- b               (ino 258)
	 *         |-- c2          (ino 261)
	 *             |-- d2      (ino 260)
	 *                 |-- cc  (ino 259)
	 *
	 * The sequence of steps that lead to the second state was:
	 *
	 * mv /a/b/c/d /a/b/c2/d2
	 * mv /a/b/c /a/b/c2/d2/cc
	 *
	 * "c" has lower inode number, but we can't move it (2nd mv operation)
	 * before we move "d", which has higher inode number.
	 *
	 * So we just memorize which move/rename operations must be performed
	 * later when their respective parent is processed and moved/renamed.
	 */

	/* Indexed by parent directory inode number. */
	struct rb_root pending_dir_moves;

	/*
	 * Reverse index, indexed by the inode number of a directory that
	 * is waiting for the move/rename of its immediate parent before its
	 * own move/rename can be performed.
	 */
	struct rb_root waiting_dir_moves;
184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224

	/*
	 * A directory that is going to be rm'ed might have a child directory
	 * which is in the pending directory moves index above. In this case,
	 * the directory can only be removed after the move/rename of its child
	 * is performed. Example:
	 *
	 * Parent snapshot:
	 *
	 * .                        (ino 256)
	 * |-- a/                   (ino 257)
	 *     |-- b/               (ino 258)
	 *         |-- c/           (ino 259)
	 *         |   |-- x/       (ino 260)
	 *         |
	 *         |-- y/           (ino 261)
	 *
	 * Send snapshot:
	 *
	 * .                        (ino 256)
	 * |-- a/                   (ino 257)
	 *     |-- b/               (ino 258)
	 *         |-- YY/          (ino 261)
	 *              |-- x/      (ino 260)
	 *
	 * Sequence of steps that lead to the send snapshot:
	 * rm -f /a/b/c/foo.txt
	 * mv /a/b/y /a/b/YY
	 * mv /a/b/c/x /a/b/YY
	 * rmdir /a/b/c
	 *
	 * When the child is processed, its move/rename is delayed until its
	 * parent is processed (as explained above), but all other operations
	 * like update utimes, chown, chgrp, etc, are performed and the paths
	 * that it uses for those operations must use the orphanized name of
	 * its parent (the directory we're going to rm later), so we need to
	 * memorize that name.
	 *
	 * Indexed by the inode number of the directory to be deleted.
	 */
	struct rb_root orphan_dirs;
225 226 227 228 229 230 231 232 233 234 235 236 237 238
};

struct pending_dir_move {
	struct rb_node node;
	struct list_head list;
	u64 parent_ino;
	u64 ino;
	u64 gen;
	struct list_head update_refs;
};

struct waiting_dir_move {
	struct rb_node node;
	u64 ino;
239 240 241 242 243 244 245 246 247 248 249 250
	/*
	 * There might be some directory that could not be removed because it
	 * was waiting for this directory inode to be moved first. Therefore
	 * after this directory is moved, we can try to rmdir the ino rmdir_ino.
	 */
	u64 rmdir_ino;
};

struct orphan_dir_info {
	struct rb_node node;
	u64 ino;
	u64 gen;
251 252 253 254
};

struct name_cache_entry {
	struct list_head list;
255 256 257 258 259 260 261 262 263
	/*
	 * radix_tree has only 32bit entries but we need to handle 64bit inums.
	 * We use the lower 32bit of the 64bit inum to store it in the tree. If
	 * more then one inum would fall into the same entry, we use radix_list
	 * to store the additional entries. radix_list is also used to store
	 * entries where two entries have the same inum but different
	 * generations.
	 */
	struct list_head radix_list;
264 265 266 267 268 269 270 271 272 273
	u64 ino;
	u64 gen;
	u64 parent_ino;
	u64 parent_gen;
	int ret;
	int need_later_update;
	int name_len;
	char name[];
};

274 275
static int is_waiting_for_move(struct send_ctx *sctx, u64 ino);

276 277 278 279 280
static struct waiting_dir_move *
get_waiting_dir_move(struct send_ctx *sctx, u64 ino);

static int is_waiting_for_rm(struct send_ctx *sctx, u64 dir_ino);

281 282 283 284 285 286 287
static int need_send_hole(struct send_ctx *sctx)
{
	return (sctx->parent_root && !sctx->cur_inode_new &&
		!sctx->cur_inode_new_gen && !sctx->cur_inode_deleted &&
		S_ISREG(sctx->cur_inode_mode));
}

288 289 290 291 292 293 294 295 296 297 298 299 300
static void fs_path_reset(struct fs_path *p)
{
	if (p->reversed) {
		p->start = p->buf + p->buf_len - 1;
		p->end = p->start;
		*p->start = 0;
	} else {
		p->start = p->buf;
		p->end = p->start;
		*p->start = 0;
	}
}

301
static struct fs_path *fs_path_alloc(void)
302 303 304 305 306 307 308 309 310 311 312 313 314
{
	struct fs_path *p;

	p = kmalloc(sizeof(*p), GFP_NOFS);
	if (!p)
		return NULL;
	p->reversed = 0;
	p->buf = p->inline_buf;
	p->buf_len = FS_PATH_INLINE_SIZE;
	fs_path_reset(p);
	return p;
}

315
static struct fs_path *fs_path_alloc_reversed(void)
316 317 318
{
	struct fs_path *p;

319
	p = fs_path_alloc();
320 321 322 323 324 325 326
	if (!p)
		return NULL;
	p->reversed = 1;
	fs_path_reset(p);
	return p;
}

327
static void fs_path_free(struct fs_path *p)
328 329 330
{
	if (!p)
		return;
331 332
	if (p->buf != p->inline_buf)
		kfree(p->buf);
333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351
	kfree(p);
}

static int fs_path_len(struct fs_path *p)
{
	return p->end - p->start;
}

static int fs_path_ensure_buf(struct fs_path *p, int len)
{
	char *tmp_buf;
	int path_len;
	int old_buf_len;

	len++;

	if (p->buf_len >= len)
		return 0;

352 353 354 355 356
	if (len > PATH_MAX) {
		WARN_ON(1);
		return -ENOMEM;
	}

357 358 359
	path_len = p->end - p->start;
	old_buf_len = p->buf_len;

360 361 362
	/*
	 * First time the inline_buf does not suffice
	 */
363
	if (p->buf == p->inline_buf) {
364
		tmp_buf = kmalloc(len, GFP_NOFS);
365 366 367
		if (tmp_buf)
			memcpy(tmp_buf, p->buf, old_buf_len);
	} else {
368
		tmp_buf = krealloc(p->buf, len, GFP_NOFS);
369
	}
370 371 372 373 374 375 376 377
	if (!tmp_buf)
		return -ENOMEM;
	p->buf = tmp_buf;
	/*
	 * The real size of the buffer is bigger, this will let the fast path
	 * happen most of the time
	 */
	p->buf_len = ksize(p->buf);
378

379 380 381 382 383 384 385 386 387 388 389 390
	if (p->reversed) {
		tmp_buf = p->buf + old_buf_len - path_len - 1;
		p->end = p->buf + p->buf_len - 1;
		p->start = p->end - path_len;
		memmove(p->start, tmp_buf, path_len + 1);
	} else {
		p->start = p->buf;
		p->end = p->start + path_len;
	}
	return 0;
}

391 392
static int fs_path_prepare_for_add(struct fs_path *p, int name_len,
				   char **prepared)
393 394 395 396 397 398 399 400 401 402 403 404 405 406 407
{
	int ret;
	int new_len;

	new_len = p->end - p->start + name_len;
	if (p->start != p->end)
		new_len++;
	ret = fs_path_ensure_buf(p, new_len);
	if (ret < 0)
		goto out;

	if (p->reversed) {
		if (p->start != p->end)
			*--p->start = '/';
		p->start -= name_len;
408
		*prepared = p->start;
409 410 411
	} else {
		if (p->start != p->end)
			*p->end++ = '/';
412
		*prepared = p->end;
413 414 415 416 417 418 419 420 421 422 423
		p->end += name_len;
		*p->end = 0;
	}

out:
	return ret;
}

static int fs_path_add(struct fs_path *p, const char *name, int name_len)
{
	int ret;
424
	char *prepared;
425

426
	ret = fs_path_prepare_for_add(p, name_len, &prepared);
427 428
	if (ret < 0)
		goto out;
429
	memcpy(prepared, name, name_len);
430 431 432 433 434 435 436 437

out:
	return ret;
}

static int fs_path_add_path(struct fs_path *p, struct fs_path *p2)
{
	int ret;
438
	char *prepared;
439

440
	ret = fs_path_prepare_for_add(p, p2->end - p2->start, &prepared);
441 442
	if (ret < 0)
		goto out;
443
	memcpy(prepared, p2->start, p2->end - p2->start);
444 445 446 447 448 449 450 451 452 453

out:
	return ret;
}

static int fs_path_add_from_extent_buffer(struct fs_path *p,
					  struct extent_buffer *eb,
					  unsigned long off, int len)
{
	int ret;
454
	char *prepared;
455

456
	ret = fs_path_prepare_for_add(p, len, &prepared);
457 458 459
	if (ret < 0)
		goto out;

460
	read_extent_buffer(eb, prepared, off, len);
461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503

out:
	return ret;
}

static int fs_path_copy(struct fs_path *p, struct fs_path *from)
{
	int ret;

	p->reversed = from->reversed;
	fs_path_reset(p);

	ret = fs_path_add_path(p, from);

	return ret;
}


static void fs_path_unreverse(struct fs_path *p)
{
	char *tmp;
	int len;

	if (!p->reversed)
		return;

	tmp = p->start;
	len = p->end - p->start;
	p->start = p->buf;
	p->end = p->start + len;
	memmove(p->start, tmp, len + 1);
	p->reversed = 0;
}

static struct btrfs_path *alloc_path_for_send(void)
{
	struct btrfs_path *path;

	path = btrfs_alloc_path();
	if (!path)
		return NULL;
	path->search_commit_root = 1;
	path->skip_locking = 1;
504
	path->need_commit_sem = 1;
505 506 507
	return path;
}

508
static int write_buf(struct file *filp, const void *buf, u32 len, loff_t *off)
509 510 511 512 513 514 515 516 517
{
	int ret;
	mm_segment_t old_fs;
	u32 pos = 0;

	old_fs = get_fs();
	set_fs(KERNEL_DS);

	while (pos < len) {
518
		ret = vfs_write(filp, (char *)buf + pos, len - pos, off);
519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556
		/* TODO handle that correctly */
		/*if (ret == -ERESTARTSYS) {
			continue;
		}*/
		if (ret < 0)
			goto out;
		if (ret == 0) {
			ret = -EIO;
			goto out;
		}
		pos += ret;
	}

	ret = 0;

out:
	set_fs(old_fs);
	return ret;
}

static int tlv_put(struct send_ctx *sctx, u16 attr, const void *data, int len)
{
	struct btrfs_tlv_header *hdr;
	int total_len = sizeof(*hdr) + len;
	int left = sctx->send_max_size - sctx->send_size;

	if (unlikely(left < total_len))
		return -EOVERFLOW;

	hdr = (struct btrfs_tlv_header *) (sctx->send_buf + sctx->send_size);
	hdr->tlv_type = cpu_to_le16(attr);
	hdr->tlv_len = cpu_to_le16(len);
	memcpy(hdr + 1, data, len);
	sctx->send_size += total_len;

	return 0;
}

D
David Sterba 已提交
557 558 559 560 561 562 563
#define TLV_PUT_DEFINE_INT(bits) \
	static int tlv_put_u##bits(struct send_ctx *sctx,	 	\
			u##bits attr, u##bits value)			\
	{								\
		__le##bits __tmp = cpu_to_le##bits(value);		\
		return tlv_put(sctx, attr, &__tmp, sizeof(__tmp));	\
	}
564

D
David Sterba 已提交
565
TLV_PUT_DEFINE_INT(64)
566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641

static int tlv_put_string(struct send_ctx *sctx, u16 attr,
			  const char *str, int len)
{
	if (len == -1)
		len = strlen(str);
	return tlv_put(sctx, attr, str, len);
}

static int tlv_put_uuid(struct send_ctx *sctx, u16 attr,
			const u8 *uuid)
{
	return tlv_put(sctx, attr, uuid, BTRFS_UUID_SIZE);
}

static int tlv_put_btrfs_timespec(struct send_ctx *sctx, u16 attr,
				  struct extent_buffer *eb,
				  struct btrfs_timespec *ts)
{
	struct btrfs_timespec bts;
	read_extent_buffer(eb, &bts, (unsigned long)ts, sizeof(bts));
	return tlv_put(sctx, attr, &bts, sizeof(bts));
}


#define TLV_PUT(sctx, attrtype, attrlen, data) \
	do { \
		ret = tlv_put(sctx, attrtype, attrlen, data); \
		if (ret < 0) \
			goto tlv_put_failure; \
	} while (0)

#define TLV_PUT_INT(sctx, attrtype, bits, value) \
	do { \
		ret = tlv_put_u##bits(sctx, attrtype, value); \
		if (ret < 0) \
			goto tlv_put_failure; \
	} while (0)

#define TLV_PUT_U8(sctx, attrtype, data) TLV_PUT_INT(sctx, attrtype, 8, data)
#define TLV_PUT_U16(sctx, attrtype, data) TLV_PUT_INT(sctx, attrtype, 16, data)
#define TLV_PUT_U32(sctx, attrtype, data) TLV_PUT_INT(sctx, attrtype, 32, data)
#define TLV_PUT_U64(sctx, attrtype, data) TLV_PUT_INT(sctx, attrtype, 64, data)
#define TLV_PUT_STRING(sctx, attrtype, str, len) \
	do { \
		ret = tlv_put_string(sctx, attrtype, str, len); \
		if (ret < 0) \
			goto tlv_put_failure; \
	} while (0)
#define TLV_PUT_PATH(sctx, attrtype, p) \
	do { \
		ret = tlv_put_string(sctx, attrtype, p->start, \
			p->end - p->start); \
		if (ret < 0) \
			goto tlv_put_failure; \
	} while(0)
#define TLV_PUT_UUID(sctx, attrtype, uuid) \
	do { \
		ret = tlv_put_uuid(sctx, attrtype, uuid); \
		if (ret < 0) \
			goto tlv_put_failure; \
	} while (0)
#define TLV_PUT_BTRFS_TIMESPEC(sctx, attrtype, eb, ts) \
	do { \
		ret = tlv_put_btrfs_timespec(sctx, attrtype, eb, ts); \
		if (ret < 0) \
			goto tlv_put_failure; \
	} while (0)

static int send_header(struct send_ctx *sctx)
{
	struct btrfs_stream_header hdr;

	strcpy(hdr.magic, BTRFS_SEND_STREAM_MAGIC);
	hdr.version = cpu_to_le32(BTRFS_SEND_STREAM_VERSION);

642 643
	return write_buf(sctx->send_filp, &hdr, sizeof(hdr),
					&sctx->send_off);
644 645 646 647 648 649 650 651 652
}

/*
 * For each command/item we want to send to userspace, we call this function.
 */
static int begin_cmd(struct send_ctx *sctx, int cmd)
{
	struct btrfs_cmd_header *hdr;

653
	if (WARN_ON(!sctx->send_buf))
654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674
		return -EINVAL;

	BUG_ON(sctx->send_size);

	sctx->send_size += sizeof(*hdr);
	hdr = (struct btrfs_cmd_header *)sctx->send_buf;
	hdr->cmd = cpu_to_le16(cmd);

	return 0;
}

static int send_cmd(struct send_ctx *sctx)
{
	int ret;
	struct btrfs_cmd_header *hdr;
	u32 crc;

	hdr = (struct btrfs_cmd_header *)sctx->send_buf;
	hdr->len = cpu_to_le32(sctx->send_size - sizeof(*hdr));
	hdr->crc = 0;

675
	crc = btrfs_crc32c(0, (unsigned char *)sctx->send_buf, sctx->send_size);
676 677
	hdr->crc = cpu_to_le32(crc);

678 679
	ret = write_buf(sctx->send_filp, sctx->send_buf, sctx->send_size,
					&sctx->send_off);
680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782

	sctx->total_send_size += sctx->send_size;
	sctx->cmd_send_size[le16_to_cpu(hdr->cmd)] += sctx->send_size;
	sctx->send_size = 0;

	return ret;
}

/*
 * Sends a move instruction to user space
 */
static int send_rename(struct send_ctx *sctx,
		     struct fs_path *from, struct fs_path *to)
{
	int ret;

verbose_printk("btrfs: send_rename %s -> %s\n", from->start, to->start);

	ret = begin_cmd(sctx, BTRFS_SEND_C_RENAME);
	if (ret < 0)
		goto out;

	TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, from);
	TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH_TO, to);

	ret = send_cmd(sctx);

tlv_put_failure:
out:
	return ret;
}

/*
 * Sends a link instruction to user space
 */
static int send_link(struct send_ctx *sctx,
		     struct fs_path *path, struct fs_path *lnk)
{
	int ret;

verbose_printk("btrfs: send_link %s -> %s\n", path->start, lnk->start);

	ret = begin_cmd(sctx, BTRFS_SEND_C_LINK);
	if (ret < 0)
		goto out;

	TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, path);
	TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH_LINK, lnk);

	ret = send_cmd(sctx);

tlv_put_failure:
out:
	return ret;
}

/*
 * Sends an unlink instruction to user space
 */
static int send_unlink(struct send_ctx *sctx, struct fs_path *path)
{
	int ret;

verbose_printk("btrfs: send_unlink %s\n", path->start);

	ret = begin_cmd(sctx, BTRFS_SEND_C_UNLINK);
	if (ret < 0)
		goto out;

	TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, path);

	ret = send_cmd(sctx);

tlv_put_failure:
out:
	return ret;
}

/*
 * Sends a rmdir instruction to user space
 */
static int send_rmdir(struct send_ctx *sctx, struct fs_path *path)
{
	int ret;

verbose_printk("btrfs: send_rmdir %s\n", path->start);

	ret = begin_cmd(sctx, BTRFS_SEND_C_RMDIR);
	if (ret < 0)
		goto out;

	TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, path);

	ret = send_cmd(sctx);

tlv_put_failure:
out:
	return ret;
}

/*
 * Helper function to retrieve some fields from an inode item.
 */
783 784 785
static int __get_inode_info(struct btrfs_root *root, struct btrfs_path *path,
			  u64 ino, u64 *size, u64 *gen, u64 *mode, u64 *uid,
			  u64 *gid, u64 *rdev)
786 787 788 789 790 791 792 793 794 795
{
	int ret;
	struct btrfs_inode_item *ii;
	struct btrfs_key key;

	key.objectid = ino;
	key.type = BTRFS_INODE_ITEM_KEY;
	key.offset = 0;
	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
	if (ret) {
796 797 798
		if (ret > 0)
			ret = -ENOENT;
		return ret;
799 800 801 802 803 804 805 806 807 808 809 810 811 812
	}

	ii = btrfs_item_ptr(path->nodes[0], path->slots[0],
			struct btrfs_inode_item);
	if (size)
		*size = btrfs_inode_size(path->nodes[0], ii);
	if (gen)
		*gen = btrfs_inode_generation(path->nodes[0], ii);
	if (mode)
		*mode = btrfs_inode_mode(path->nodes[0], ii);
	if (uid)
		*uid = btrfs_inode_uid(path->nodes[0], ii);
	if (gid)
		*gid = btrfs_inode_gid(path->nodes[0], ii);
813 814
	if (rdev)
		*rdev = btrfs_inode_rdev(path->nodes[0], ii);
815

816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831
	return ret;
}

static int get_inode_info(struct btrfs_root *root,
			  u64 ino, u64 *size, u64 *gen,
			  u64 *mode, u64 *uid, u64 *gid,
			  u64 *rdev)
{
	struct btrfs_path *path;
	int ret;

	path = alloc_path_for_send();
	if (!path)
		return -ENOMEM;
	ret = __get_inode_info(root, path, ino, size, gen, mode, uid, gid,
			       rdev);
832 833 834 835 836 837 838 839 840
	btrfs_free_path(path);
	return ret;
}

typedef int (*iterate_inode_ref_t)(int num, u64 dir, int index,
				   struct fs_path *p,
				   void *ctx);

/*
841 842
 * Helper function to iterate the entries in ONE btrfs_inode_ref or
 * btrfs_inode_extref.
843 844 845
 * The iterate callback may return a non zero value to stop iteration. This can
 * be a negative value for error codes or 1 to simply stop it.
 *
846
 * path must point to the INODE_REF or INODE_EXTREF when called.
847
 */
848
static int iterate_inode_ref(struct btrfs_root *root, struct btrfs_path *path,
849 850 851
			     struct btrfs_key *found_key, int resolve,
			     iterate_inode_ref_t iterate, void *ctx)
{
852
	struct extent_buffer *eb = path->nodes[0];
853 854
	struct btrfs_item *item;
	struct btrfs_inode_ref *iref;
855
	struct btrfs_inode_extref *extref;
856 857
	struct btrfs_path *tmp_path;
	struct fs_path *p;
858
	u32 cur = 0;
859
	u32 total;
860
	int slot = path->slots[0];
861 862 863
	u32 name_len;
	char *start;
	int ret = 0;
864
	int num = 0;
865
	int index;
866 867 868 869
	u64 dir;
	unsigned long name_off;
	unsigned long elem_size;
	unsigned long ptr;
870

871
	p = fs_path_alloc_reversed();
872 873 874 875 876
	if (!p)
		return -ENOMEM;

	tmp_path = alloc_path_for_send();
	if (!tmp_path) {
877
		fs_path_free(p);
878 879 880 881
		return -ENOMEM;
	}


882 883 884
	if (found_key->type == BTRFS_INODE_REF_KEY) {
		ptr = (unsigned long)btrfs_item_ptr(eb, slot,
						    struct btrfs_inode_ref);
885
		item = btrfs_item_nr(slot);
886 887 888 889 890 891 892 893
		total = btrfs_item_size(eb, item);
		elem_size = sizeof(*iref);
	} else {
		ptr = btrfs_item_ptr_offset(eb, slot);
		total = btrfs_item_size_nr(eb, slot);
		elem_size = sizeof(*extref);
	}

894 895 896
	while (cur < total) {
		fs_path_reset(p);

897 898 899 900 901 902 903 904 905 906 907 908 909 910
		if (found_key->type == BTRFS_INODE_REF_KEY) {
			iref = (struct btrfs_inode_ref *)(ptr + cur);
			name_len = btrfs_inode_ref_name_len(eb, iref);
			name_off = (unsigned long)(iref + 1);
			index = btrfs_inode_ref_index(eb, iref);
			dir = found_key->offset;
		} else {
			extref = (struct btrfs_inode_extref *)(ptr + cur);
			name_len = btrfs_inode_extref_name_len(eb, extref);
			name_off = (unsigned long)&extref->name;
			index = btrfs_inode_extref_index(eb, extref);
			dir = btrfs_inode_extref_parent(eb, extref);
		}

911
		if (resolve) {
912 913 914
			start = btrfs_ref_to_path(root, tmp_path, name_len,
						  name_off, eb, dir,
						  p->buf, p->buf_len);
915 916 917 918 919 920 921 922 923 924
			if (IS_ERR(start)) {
				ret = PTR_ERR(start);
				goto out;
			}
			if (start < p->buf) {
				/* overflow , try again with larger buffer */
				ret = fs_path_ensure_buf(p,
						p->buf_len + p->buf - start);
				if (ret < 0)
					goto out;
925 926 927 928
				start = btrfs_ref_to_path(root, tmp_path,
							  name_len, name_off,
							  eb, dir,
							  p->buf, p->buf_len);
929 930 931 932 933 934 935 936
				if (IS_ERR(start)) {
					ret = PTR_ERR(start);
					goto out;
				}
				BUG_ON(start < p->buf);
			}
			p->start = start;
		} else {
937 938
			ret = fs_path_add_from_extent_buffer(p, eb, name_off,
							     name_len);
939 940 941 942
			if (ret < 0)
				goto out;
		}

943 944
		cur += elem_size + name_len;
		ret = iterate(num, dir, index, p, ctx);
945 946 947 948 949 950 951
		if (ret)
			goto out;
		num++;
	}

out:
	btrfs_free_path(tmp_path);
952
	fs_path_free(p);
953 954 955 956 957 958 959 960 961 962 963 964 965 966 967
	return ret;
}

typedef int (*iterate_dir_item_t)(int num, struct btrfs_key *di_key,
				  const char *name, int name_len,
				  const char *data, int data_len,
				  u8 type, void *ctx);

/*
 * Helper function to iterate the entries in ONE btrfs_dir_item.
 * The iterate callback may return a non zero value to stop iteration. This can
 * be a negative value for error codes or 1 to simply stop it.
 *
 * path must point to the dir item when called.
 */
968
static int iterate_dir_item(struct btrfs_root *root, struct btrfs_path *path,
969 970 971 972 973 974 975 976 977
			    struct btrfs_key *found_key,
			    iterate_dir_item_t iterate, void *ctx)
{
	int ret = 0;
	struct extent_buffer *eb;
	struct btrfs_item *item;
	struct btrfs_dir_item *di;
	struct btrfs_key di_key;
	char *buf = NULL;
978
	const int buf_len = PATH_MAX;
979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995
	u32 name_len;
	u32 data_len;
	u32 cur;
	u32 len;
	u32 total;
	int slot;
	int num;
	u8 type;

	buf = kmalloc(buf_len, GFP_NOFS);
	if (!buf) {
		ret = -ENOMEM;
		goto out;
	}

	eb = path->nodes[0];
	slot = path->slots[0];
996
	item = btrfs_item_nr(slot);
997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008
	di = btrfs_item_ptr(eb, slot, struct btrfs_dir_item);
	cur = 0;
	len = 0;
	total = btrfs_item_size(eb, item);

	num = 0;
	while (cur < total) {
		name_len = btrfs_dir_name_len(eb, di);
		data_len = btrfs_dir_data_len(eb, di);
		type = btrfs_dir_type(eb, di);
		btrfs_dir_item_key_to_cpu(eb, di, &di_key);

1009 1010 1011
		/*
		 * Path too long
		 */
1012
		if (name_len + data_len > buf_len) {
1013 1014
			ret = -ENAMETOOLONG;
			goto out;
1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036
		}

		read_extent_buffer(eb, buf, (unsigned long)(di + 1),
				name_len + data_len);

		len = sizeof(*di) + name_len + data_len;
		di = (struct btrfs_dir_item *)((char *)di + len);
		cur += len;

		ret = iterate(num, &di_key, buf, name_len, buf + name_len,
				data_len, type, ctx);
		if (ret < 0)
			goto out;
		if (ret) {
			ret = 0;
			goto out;
		}

		num++;
	}

out:
1037
	kfree(buf);
1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058
	return ret;
}

static int __copy_first_ref(int num, u64 dir, int index,
			    struct fs_path *p, void *ctx)
{
	int ret;
	struct fs_path *pt = ctx;

	ret = fs_path_copy(pt, p);
	if (ret < 0)
		return ret;

	/* we want the first only */
	return 1;
}

/*
 * Retrieve the first path of an inode. If an inode has more then one
 * ref/hardlink, this is ignored.
 */
1059
static int get_inode_path(struct btrfs_root *root,
1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084
			  u64 ino, struct fs_path *path)
{
	int ret;
	struct btrfs_key key, found_key;
	struct btrfs_path *p;

	p = alloc_path_for_send();
	if (!p)
		return -ENOMEM;

	fs_path_reset(path);

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

	ret = btrfs_search_slot_for_read(root, &key, p, 1, 0);
	if (ret < 0)
		goto out;
	if (ret) {
		ret = 1;
		goto out;
	}
	btrfs_item_key_to_cpu(p->nodes[0], &found_key, p->slots[0]);
	if (found_key.objectid != ino ||
1085 1086
	    (found_key.type != BTRFS_INODE_REF_KEY &&
	     found_key.type != BTRFS_INODE_EXTREF_KEY)) {
1087 1088 1089 1090
		ret = -ENOENT;
		goto out;
	}

1091 1092
	ret = iterate_inode_ref(root, p, &found_key, 1,
				__copy_first_ref, path);
1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104
	if (ret < 0)
		goto out;
	ret = 0;

out:
	btrfs_free_path(p);
	return ret;
}

struct backref_ctx {
	struct send_ctx *sctx;

1105
	struct btrfs_path *path;
1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119
	/* number of total found references */
	u64 found;

	/*
	 * used for clones found in send_root. clones found behind cur_objectid
	 * and cur_offset are not considered as allowed clones.
	 */
	u64 cur_objectid;
	u64 cur_offset;

	/* may be truncated in case it's the last extent in a file */
	u64 extent_len;

	/* Just to check for bugs in backref resolving */
1120
	int found_itself;
1121 1122 1123 1124
};

static int __clone_root_cmp_bsearch(const void *key, const void *elt)
{
1125
	u64 root = (u64)(uintptr_t)key;
1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148
	struct clone_root *cr = (struct clone_root *)elt;

	if (root < cr->root->objectid)
		return -1;
	if (root > cr->root->objectid)
		return 1;
	return 0;
}

static int __clone_root_cmp_sort(const void *e1, const void *e2)
{
	struct clone_root *cr1 = (struct clone_root *)e1;
	struct clone_root *cr2 = (struct clone_root *)e2;

	if (cr1->root->objectid < cr2->root->objectid)
		return -1;
	if (cr1->root->objectid > cr2->root->objectid)
		return 1;
	return 0;
}

/*
 * Called for every backref that is found for the current extent.
1149
 * Results are collected in sctx->clone_roots->ino/offset/found_refs
1150 1151 1152 1153 1154 1155 1156 1157 1158
 */
static int __iterate_backrefs(u64 ino, u64 offset, u64 root, void *ctx_)
{
	struct backref_ctx *bctx = ctx_;
	struct clone_root *found;
	int ret;
	u64 i_size;

	/* First check if the root is in the list of accepted clone sources */
1159
	found = bsearch((void *)(uintptr_t)root, bctx->sctx->clone_roots,
1160 1161 1162 1163 1164 1165 1166 1167 1168
			bctx->sctx->clone_roots_cnt,
			sizeof(struct clone_root),
			__clone_root_cmp_bsearch);
	if (!found)
		return 0;

	if (found->root == bctx->sctx->send_root &&
	    ino == bctx->cur_objectid &&
	    offset == bctx->cur_offset) {
1169
		bctx->found_itself = 1;
1170 1171 1172
	}

	/*
1173
	 * There are inodes that have extents that lie behind its i_size. Don't
1174 1175
	 * accept clones from these extents.
	 */
1176 1177 1178
	ret = __get_inode_info(found->root, bctx->path, ino, &i_size, NULL, NULL,
			       NULL, NULL, NULL);
	btrfs_release_path(bctx->path);
1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197
	if (ret < 0)
		return ret;

	if (offset + bctx->extent_len > i_size)
		return 0;

	/*
	 * Make sure we don't consider clones from send_root that are
	 * behind the current inode/offset.
	 */
	if (found->root == bctx->sctx->send_root) {
		/*
		 * TODO for the moment we don't accept clones from the inode
		 * that is currently send. We may change this when
		 * BTRFS_IOC_CLONE_RANGE supports cloning from and to the same
		 * file.
		 */
		if (ino >= bctx->cur_objectid)
			return 0;
1198 1199 1200 1201
#if 0
		if (ino > bctx->cur_objectid)
			return 0;
		if (offset + bctx->extent_len > bctx->cur_offset)
1202
			return 0;
1203
#endif
1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222
	}

	bctx->found++;
	found->found_refs++;
	if (ino < found->ino) {
		found->ino = ino;
		found->offset = offset;
	} else if (found->ino == ino) {
		/*
		 * same extent found more then once in the same file.
		 */
		if (found->offset > offset + bctx->extent_len)
			found->offset = offset;
	}

	return 0;
}

/*
1223 1224 1225 1226 1227 1228
 * Given an inode, offset and extent item, it finds a good clone for a clone
 * instruction. Returns -ENOENT when none could be found. The function makes
 * sure that the returned clone is usable at the point where sending is at the
 * moment. This means, that no clones are accepted which lie behind the current
 * inode+offset.
 *
1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239
 * path must point to the extent item when called.
 */
static int find_extent_clone(struct send_ctx *sctx,
			     struct btrfs_path *path,
			     u64 ino, u64 data_offset,
			     u64 ino_size,
			     struct clone_root **found)
{
	int ret;
	int extent_type;
	u64 logical;
1240
	u64 disk_byte;
1241 1242
	u64 num_bytes;
	u64 extent_item_pos;
1243
	u64 flags = 0;
1244 1245
	struct btrfs_file_extent_item *fi;
	struct extent_buffer *eb = path->nodes[0];
1246
	struct backref_ctx *backref_ctx = NULL;
1247 1248 1249
	struct clone_root *cur_clone_root;
	struct btrfs_key found_key;
	struct btrfs_path *tmp_path;
1250
	int compressed;
1251 1252 1253 1254 1255 1256
	u32 i;

	tmp_path = alloc_path_for_send();
	if (!tmp_path)
		return -ENOMEM;

1257 1258 1259
	/* We only use this path under the commit sem */
	tmp_path->need_commit_sem = 0;

1260 1261 1262 1263 1264 1265
	backref_ctx = kmalloc(sizeof(*backref_ctx), GFP_NOFS);
	if (!backref_ctx) {
		ret = -ENOMEM;
		goto out;
	}

1266 1267
	backref_ctx->path = tmp_path;

1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284
	if (data_offset >= ino_size) {
		/*
		 * There may be extents that lie behind the file's size.
		 * I at least had this in combination with snapshotting while
		 * writing large files.
		 */
		ret = 0;
		goto out;
	}

	fi = btrfs_item_ptr(eb, path->slots[0],
			struct btrfs_file_extent_item);
	extent_type = btrfs_file_extent_type(eb, fi);
	if (extent_type == BTRFS_FILE_EXTENT_INLINE) {
		ret = -ENOENT;
		goto out;
	}
1285
	compressed = btrfs_file_extent_compression(eb, fi);
1286 1287

	num_bytes = btrfs_file_extent_num_bytes(eb, fi);
1288 1289
	disk_byte = btrfs_file_extent_disk_bytenr(eb, fi);
	if (disk_byte == 0) {
1290 1291 1292
		ret = -ENOENT;
		goto out;
	}
1293
	logical = disk_byte + btrfs_file_extent_offset(eb, fi);
1294

1295
	down_read(&sctx->send_root->fs_info->commit_root_sem);
1296 1297
	ret = extent_from_logical(sctx->send_root->fs_info, disk_byte, tmp_path,
				  &found_key, &flags);
1298
	up_read(&sctx->send_root->fs_info->commit_root_sem);
1299 1300 1301 1302
	btrfs_release_path(tmp_path);

	if (ret < 0)
		goto out;
1303
	if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) {
1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317
		ret = -EIO;
		goto out;
	}

	/*
	 * Setup the clone roots.
	 */
	for (i = 0; i < sctx->clone_roots_cnt; i++) {
		cur_clone_root = sctx->clone_roots + i;
		cur_clone_root->ino = (u64)-1;
		cur_clone_root->offset = 0;
		cur_clone_root->found_refs = 0;
	}

1318 1319 1320 1321 1322 1323
	backref_ctx->sctx = sctx;
	backref_ctx->found = 0;
	backref_ctx->cur_objectid = ino;
	backref_ctx->cur_offset = data_offset;
	backref_ctx->found_itself = 0;
	backref_ctx->extent_len = num_bytes;
1324 1325 1326 1327 1328 1329 1330

	/*
	 * The last extent of a file may be too large due to page alignment.
	 * We need to adjust extent_len in this case so that the checks in
	 * __iterate_backrefs work.
	 */
	if (data_offset + num_bytes >= ino_size)
1331
		backref_ctx->extent_len = ino_size - data_offset;
1332 1333 1334 1335

	/*
	 * Now collect all backrefs.
	 */
1336 1337 1338 1339
	if (compressed == BTRFS_COMPRESS_NONE)
		extent_item_pos = logical - found_key.objectid;
	else
		extent_item_pos = 0;
1340 1341
	ret = iterate_extent_inodes(sctx->send_root->fs_info,
					found_key.objectid, extent_item_pos, 1,
1342
					__iterate_backrefs, backref_ctx);
1343

1344 1345 1346
	if (ret < 0)
		goto out;

1347
	if (!backref_ctx->found_itself) {
1348 1349
		/* found a bug in backref code? */
		ret = -EIO;
1350
		btrfs_err(sctx->send_root->fs_info, "did not find backref in "
1351
				"send_root. inode=%llu, offset=%llu, "
1352 1353
				"disk_byte=%llu found extent=%llu\n",
				ino, data_offset, disk_byte, found_key.objectid);
1354 1355 1356 1357 1358 1359 1360 1361
		goto out;
	}

verbose_printk(KERN_DEBUG "btrfs: find_extent_clone: data_offset=%llu, "
		"ino=%llu, "
		"num_bytes=%llu, logical=%llu\n",
		data_offset, ino, num_bytes, logical);

1362
	if (!backref_ctx->found)
1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377
		verbose_printk("btrfs:    no clones found\n");

	cur_clone_root = NULL;
	for (i = 0; i < sctx->clone_roots_cnt; i++) {
		if (sctx->clone_roots[i].found_refs) {
			if (!cur_clone_root)
				cur_clone_root = sctx->clone_roots + i;
			else if (sctx->clone_roots[i].root == sctx->send_root)
				/* prefer clones from send_root over others */
				cur_clone_root = sctx->clone_roots + i;
		}

	}

	if (cur_clone_root) {
1378 1379 1380 1381 1382 1383 1384 1385 1386 1387
		if (compressed != BTRFS_COMPRESS_NONE) {
			/*
			 * Offsets given by iterate_extent_inodes() are relative
			 * to the start of the extent, we need to add logical
			 * offset from the file extent item.
			 * (See why at backref.c:check_extent_in_eb())
			 */
			cur_clone_root->offset += btrfs_file_extent_offset(eb,
									   fi);
		}
1388 1389 1390 1391 1392 1393 1394 1395
		*found = cur_clone_root;
		ret = 0;
	} else {
		ret = -ENOENT;
	}

out:
	btrfs_free_path(tmp_path);
1396
	kfree(backref_ctx);
1397 1398 1399
	return ret;
}

1400
static int read_symlink(struct btrfs_root *root,
1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432
			u64 ino,
			struct fs_path *dest)
{
	int ret;
	struct btrfs_path *path;
	struct btrfs_key key;
	struct btrfs_file_extent_item *ei;
	u8 type;
	u8 compression;
	unsigned long off;
	int len;

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

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

	ei = btrfs_item_ptr(path->nodes[0], path->slots[0],
			struct btrfs_file_extent_item);
	type = btrfs_file_extent_type(path->nodes[0], ei);
	compression = btrfs_file_extent_compression(path->nodes[0], ei);
	BUG_ON(type != BTRFS_FILE_EXTENT_INLINE);
	BUG_ON(compression);

	off = btrfs_file_extent_inline_start(ei);
1433
	len = btrfs_file_extent_inline_len(path->nodes[0], path->slots[0], ei);
1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461

	ret = fs_path_add_from_extent_buffer(dest, path->nodes[0], off, len);

out:
	btrfs_free_path(path);
	return ret;
}

/*
 * Helper function to generate a file name that is unique in the root of
 * send_root and parent_root. This is used to generate names for orphan inodes.
 */
static int gen_unique_name(struct send_ctx *sctx,
			   u64 ino, u64 gen,
			   struct fs_path *dest)
{
	int ret = 0;
	struct btrfs_path *path;
	struct btrfs_dir_item *di;
	char tmp[64];
	int len;
	u64 idx = 0;

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

	while (1) {
1462
		len = snprintf(tmp, sizeof(tmp), "o%llu-%llu-%llu",
1463
				ino, gen, idx);
1464
		ASSERT(len < sizeof(tmp));
1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526

		di = btrfs_lookup_dir_item(NULL, sctx->send_root,
				path, BTRFS_FIRST_FREE_OBJECTID,
				tmp, strlen(tmp), 0);
		btrfs_release_path(path);
		if (IS_ERR(di)) {
			ret = PTR_ERR(di);
			goto out;
		}
		if (di) {
			/* not unique, try again */
			idx++;
			continue;
		}

		if (!sctx->parent_root) {
			/* unique */
			ret = 0;
			break;
		}

		di = btrfs_lookup_dir_item(NULL, sctx->parent_root,
				path, BTRFS_FIRST_FREE_OBJECTID,
				tmp, strlen(tmp), 0);
		btrfs_release_path(path);
		if (IS_ERR(di)) {
			ret = PTR_ERR(di);
			goto out;
		}
		if (di) {
			/* not unique, try again */
			idx++;
			continue;
		}
		/* unique */
		break;
	}

	ret = fs_path_add(dest, tmp, strlen(tmp));

out:
	btrfs_free_path(path);
	return ret;
}

enum inode_state {
	inode_state_no_change,
	inode_state_will_create,
	inode_state_did_create,
	inode_state_will_delete,
	inode_state_did_delete,
};

static int get_cur_inode_state(struct send_ctx *sctx, u64 ino, u64 gen)
{
	int ret;
	int left_ret;
	int right_ret;
	u64 left_gen;
	u64 right_gen;

	ret = get_inode_info(sctx->send_root, ino, NULL, &left_gen, NULL, NULL,
1527
			NULL, NULL);
1528 1529 1530 1531 1532 1533 1534 1535
	if (ret < 0 && ret != -ENOENT)
		goto out;
	left_ret = ret;

	if (!sctx->parent_root) {
		right_ret = -ENOENT;
	} else {
		ret = get_inode_info(sctx->parent_root, ino, NULL, &right_gen,
1536
				NULL, NULL, NULL, NULL);
1537 1538 1539 1540 1541 1542
		if (ret < 0 && ret != -ENOENT)
			goto out;
		right_ret = ret;
	}

	if (!left_ret && !right_ret) {
1543
		if (left_gen == gen && right_gen == gen) {
1544
			ret = inode_state_no_change;
1545
		} else if (left_gen == gen) {
1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638
			if (ino < sctx->send_progress)
				ret = inode_state_did_create;
			else
				ret = inode_state_will_create;
		} else if (right_gen == gen) {
			if (ino < sctx->send_progress)
				ret = inode_state_did_delete;
			else
				ret = inode_state_will_delete;
		} else  {
			ret = -ENOENT;
		}
	} else if (!left_ret) {
		if (left_gen == gen) {
			if (ino < sctx->send_progress)
				ret = inode_state_did_create;
			else
				ret = inode_state_will_create;
		} else {
			ret = -ENOENT;
		}
	} else if (!right_ret) {
		if (right_gen == gen) {
			if (ino < sctx->send_progress)
				ret = inode_state_did_delete;
			else
				ret = inode_state_will_delete;
		} else {
			ret = -ENOENT;
		}
	} else {
		ret = -ENOENT;
	}

out:
	return ret;
}

static int is_inode_existent(struct send_ctx *sctx, u64 ino, u64 gen)
{
	int ret;

	ret = get_cur_inode_state(sctx, ino, gen);
	if (ret < 0)
		goto out;

	if (ret == inode_state_no_change ||
	    ret == inode_state_did_create ||
	    ret == inode_state_will_delete)
		ret = 1;
	else
		ret = 0;

out:
	return ret;
}

/*
 * Helper function to lookup a dir item in a dir.
 */
static int lookup_dir_item_inode(struct btrfs_root *root,
				 u64 dir, const char *name, int name_len,
				 u64 *found_inode,
				 u8 *found_type)
{
	int ret = 0;
	struct btrfs_dir_item *di;
	struct btrfs_key key;
	struct btrfs_path *path;

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

	di = btrfs_lookup_dir_item(NULL, root, path,
			dir, name, name_len, 0);
	if (!di) {
		ret = -ENOENT;
		goto out;
	}
	if (IS_ERR(di)) {
		ret = PTR_ERR(di);
		goto out;
	}
	btrfs_dir_item_key_to_cpu(path->nodes[0], di, &key);
	*found_inode = key.objectid;
	*found_type = btrfs_dir_type(path->nodes[0], di);

out:
	btrfs_free_path(path);
	return ret;
}

1639 1640 1641 1642
/*
 * Looks up the first btrfs_inode_ref of a given ino. It returns the parent dir,
 * generation of the parent dir and the name of the dir entry.
 */
1643
static int get_first_ref(struct btrfs_root *root, u64 ino,
1644 1645 1646 1647 1648 1649 1650
			 u64 *dir, u64 *dir_gen, struct fs_path *name)
{
	int ret;
	struct btrfs_key key;
	struct btrfs_key found_key;
	struct btrfs_path *path;
	int len;
1651
	u64 parent_dir;
1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666

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

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

	ret = btrfs_search_slot_for_read(root, &key, path, 1, 0);
	if (ret < 0)
		goto out;
	if (!ret)
		btrfs_item_key_to_cpu(path->nodes[0], &found_key,
				path->slots[0]);
1667 1668 1669
	if (ret || found_key.objectid != ino ||
	    (found_key.type != BTRFS_INODE_REF_KEY &&
	     found_key.type != BTRFS_INODE_EXTREF_KEY)) {
1670 1671 1672 1673
		ret = -ENOENT;
		goto out;
	}

1674
	if (found_key.type == BTRFS_INODE_REF_KEY) {
1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691
		struct btrfs_inode_ref *iref;
		iref = btrfs_item_ptr(path->nodes[0], path->slots[0],
				      struct btrfs_inode_ref);
		len = btrfs_inode_ref_name_len(path->nodes[0], iref);
		ret = fs_path_add_from_extent_buffer(name, path->nodes[0],
						     (unsigned long)(iref + 1),
						     len);
		parent_dir = found_key.offset;
	} else {
		struct btrfs_inode_extref *extref;
		extref = btrfs_item_ptr(path->nodes[0], path->slots[0],
					struct btrfs_inode_extref);
		len = btrfs_inode_extref_name_len(path->nodes[0], extref);
		ret = fs_path_add_from_extent_buffer(name, path->nodes[0],
					(unsigned long)&extref->name, len);
		parent_dir = btrfs_inode_extref_parent(path->nodes[0], extref);
	}
1692 1693 1694 1695
	if (ret < 0)
		goto out;
	btrfs_release_path(path);

1696 1697 1698 1699 1700 1701
	if (dir_gen) {
		ret = get_inode_info(root, parent_dir, NULL, dir_gen, NULL,
				     NULL, NULL, NULL);
		if (ret < 0)
			goto out;
	}
1702

1703
	*dir = parent_dir;
1704 1705 1706 1707 1708 1709

out:
	btrfs_free_path(path);
	return ret;
}

1710
static int is_first_ref(struct btrfs_root *root,
1711 1712 1713 1714 1715 1716 1717
			u64 ino, u64 dir,
			const char *name, int name_len)
{
	int ret;
	struct fs_path *tmp_name;
	u64 tmp_dir;

1718
	tmp_name = fs_path_alloc();
1719 1720 1721
	if (!tmp_name)
		return -ENOMEM;

1722
	ret = get_first_ref(root, ino, &tmp_dir, NULL, tmp_name);
1723 1724 1725
	if (ret < 0)
		goto out;

1726
	if (dir != tmp_dir || name_len != fs_path_len(tmp_name)) {
1727 1728 1729 1730
		ret = 0;
		goto out;
	}

1731
	ret = !memcmp(tmp_name->start, name, name_len);
1732 1733

out:
1734
	fs_path_free(tmp_name);
1735 1736 1737
	return ret;
}

1738 1739 1740 1741 1742 1743 1744 1745 1746 1747
/*
 * Used by process_recorded_refs to determine if a new ref would overwrite an
 * already existing ref. In case it detects an overwrite, it returns the
 * inode/gen in who_ino/who_gen.
 * When an overwrite is detected, process_recorded_refs does proper orphanizing
 * to make sure later references to the overwritten inode are possible.
 * Orphanizing is however only required for the first ref of an inode.
 * process_recorded_refs does an additional is_first_ref check to see if
 * orphanizing is really required.
 */
1748 1749 1750 1751 1752
static int will_overwrite_ref(struct send_ctx *sctx, u64 dir, u64 dir_gen,
			      const char *name, int name_len,
			      u64 *who_ino, u64 *who_gen)
{
	int ret = 0;
1753
	u64 gen;
1754 1755 1756 1757 1758 1759 1760 1761 1762 1763
	u64 other_inode = 0;
	u8 other_type = 0;

	if (!sctx->parent_root)
		goto out;

	ret = is_inode_existent(sctx, dir, dir_gen);
	if (ret <= 0)
		goto out;

1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781
	/*
	 * If we have a parent root we need to verify that the parent dir was
	 * not delted and then re-created, if it was then we have no overwrite
	 * and we can just unlink this entry.
	 */
	if (sctx->parent_root) {
		ret = get_inode_info(sctx->parent_root, dir, NULL, &gen, NULL,
				     NULL, NULL, NULL);
		if (ret < 0 && ret != -ENOENT)
			goto out;
		if (ret) {
			ret = 0;
			goto out;
		}
		if (gen != dir_gen)
			goto out;
	}

1782 1783 1784 1785 1786 1787 1788 1789 1790
	ret = lookup_dir_item_inode(sctx->parent_root, dir, name, name_len,
			&other_inode, &other_type);
	if (ret < 0 && ret != -ENOENT)
		goto out;
	if (ret) {
		ret = 0;
		goto out;
	}

1791 1792 1793 1794 1795
	/*
	 * Check if the overwritten ref was already processed. If yes, the ref
	 * was already unlinked/moved, so we can safely assume that we will not
	 * overwrite anything at this point in time.
	 */
1796 1797
	if (other_inode > sctx->send_progress) {
		ret = get_inode_info(sctx->parent_root, other_inode, NULL,
1798
				who_gen, NULL, NULL, NULL, NULL);
1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811
		if (ret < 0)
			goto out;

		ret = 1;
		*who_ino = other_inode;
	} else {
		ret = 0;
	}

out:
	return ret;
}

1812 1813 1814 1815 1816 1817 1818
/*
 * Checks if the ref was overwritten by an already processed inode. This is
 * used by __get_cur_name_and_parent to find out if the ref was orphanized and
 * thus the orphan name needs be used.
 * process_recorded_refs also uses it to avoid unlinking of refs that were
 * overwritten.
 */
1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847
static int did_overwrite_ref(struct send_ctx *sctx,
			    u64 dir, u64 dir_gen,
			    u64 ino, u64 ino_gen,
			    const char *name, int name_len)
{
	int ret = 0;
	u64 gen;
	u64 ow_inode;
	u8 other_type;

	if (!sctx->parent_root)
		goto out;

	ret = is_inode_existent(sctx, dir, dir_gen);
	if (ret <= 0)
		goto out;

	/* check if the ref was overwritten by another ref */
	ret = lookup_dir_item_inode(sctx->send_root, dir, name, name_len,
			&ow_inode, &other_type);
	if (ret < 0 && ret != -ENOENT)
		goto out;
	if (ret) {
		/* was never and will never be overwritten */
		ret = 0;
		goto out;
	}

	ret = get_inode_info(sctx->send_root, ow_inode, NULL, &gen, NULL, NULL,
1848
			NULL, NULL);
1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866
	if (ret < 0)
		goto out;

	if (ow_inode == ino && gen == ino_gen) {
		ret = 0;
		goto out;
	}

	/* we know that it is or will be overwritten. check this now */
	if (ow_inode < sctx->send_progress)
		ret = 1;
	else
		ret = 0;

out:
	return ret;
}

1867 1868 1869 1870 1871
/*
 * Same as did_overwrite_ref, but also checks if it is the first ref of an inode
 * that got overwritten. This is used by process_recorded_refs to determine
 * if it has to use the path as returned by get_cur_path or the orphan name.
 */
1872 1873 1874 1875 1876 1877 1878 1879 1880 1881
static int did_overwrite_first_ref(struct send_ctx *sctx, u64 ino, u64 gen)
{
	int ret = 0;
	struct fs_path *name = NULL;
	u64 dir;
	u64 dir_gen;

	if (!sctx->parent_root)
		goto out;

1882
	name = fs_path_alloc();
1883 1884 1885
	if (!name)
		return -ENOMEM;

1886
	ret = get_first_ref(sctx->parent_root, ino, &dir, &dir_gen, name);
1887 1888 1889 1890 1891 1892 1893
	if (ret < 0)
		goto out;

	ret = did_overwrite_ref(sctx, dir, dir_gen, ino, gen,
			name->start, fs_path_len(name));

out:
1894
	fs_path_free(name);
1895 1896 1897
	return ret;
}

1898 1899 1900 1901
/*
 * Insert a name cache entry. On 32bit kernels the radix tree index is 32bit,
 * so we need to do some special handling in case we have clashes. This function
 * takes care of this with the help of name_cache_entry::radix_list.
1902
 * In case of error, nce is kfreed.
1903
 */
1904 1905 1906 1907
static int name_cache_insert(struct send_ctx *sctx,
			     struct name_cache_entry *nce)
{
	int ret = 0;
1908 1909 1910 1911 1912 1913
	struct list_head *nce_head;

	nce_head = radix_tree_lookup(&sctx->name_cache,
			(unsigned long)nce->ino);
	if (!nce_head) {
		nce_head = kmalloc(sizeof(*nce_head), GFP_NOFS);
1914 1915
		if (!nce_head) {
			kfree(nce);
1916
			return -ENOMEM;
1917
		}
1918
		INIT_LIST_HEAD(nce_head);
1919

1920
		ret = radix_tree_insert(&sctx->name_cache, nce->ino, nce_head);
1921 1922 1923
		if (ret < 0) {
			kfree(nce_head);
			kfree(nce);
1924
			return ret;
1925
		}
1926
	}
1927
	list_add_tail(&nce->radix_list, nce_head);
1928 1929 1930 1931 1932 1933 1934 1935 1936
	list_add_tail(&nce->list, &sctx->name_cache_list);
	sctx->name_cache_size++;

	return ret;
}

static void name_cache_delete(struct send_ctx *sctx,
			      struct name_cache_entry *nce)
{
1937
	struct list_head *nce_head;
1938

1939 1940
	nce_head = radix_tree_lookup(&sctx->name_cache,
			(unsigned long)nce->ino);
1941 1942 1943 1944 1945
	if (!nce_head) {
		btrfs_err(sctx->send_root->fs_info,
	      "name_cache_delete lookup failed ino %llu cache size %d, leaking memory",
			nce->ino, sctx->name_cache_size);
	}
1946

1947
	list_del(&nce->radix_list);
1948 1949
	list_del(&nce->list);
	sctx->name_cache_size--;
1950

1951 1952 1953 1954
	/*
	 * We may not get to the final release of nce_head if the lookup fails
	 */
	if (nce_head && list_empty(nce_head)) {
1955 1956 1957
		radix_tree_delete(&sctx->name_cache, (unsigned long)nce->ino);
		kfree(nce_head);
	}
1958 1959 1960 1961 1962
}

static struct name_cache_entry *name_cache_search(struct send_ctx *sctx,
						    u64 ino, u64 gen)
{
1963 1964
	struct list_head *nce_head;
	struct name_cache_entry *cur;
1965

1966 1967
	nce_head = radix_tree_lookup(&sctx->name_cache, (unsigned long)ino);
	if (!nce_head)
1968 1969
		return NULL;

1970 1971 1972 1973
	list_for_each_entry(cur, nce_head, radix_list) {
		if (cur->ino == ino && cur->gen == gen)
			return cur;
	}
1974 1975 1976
	return NULL;
}

1977 1978 1979 1980
/*
 * Removes the entry from the list and adds it back to the end. This marks the
 * entry as recently used so that name_cache_clean_unused does not remove it.
 */
1981 1982 1983 1984 1985 1986
static void name_cache_used(struct send_ctx *sctx, struct name_cache_entry *nce)
{
	list_del(&nce->list);
	list_add_tail(&nce->list, &sctx->name_cache_list);
}

1987 1988 1989
/*
 * Remove some entries from the beginning of name_cache_list.
 */
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
static void name_cache_clean_unused(struct send_ctx *sctx)
{
	struct name_cache_entry *nce;

	if (sctx->name_cache_size < SEND_CTX_NAME_CACHE_CLEAN_SIZE)
		return;

	while (sctx->name_cache_size > SEND_CTX_MAX_NAME_CACHE_SIZE) {
		nce = list_entry(sctx->name_cache_list.next,
				struct name_cache_entry, list);
		name_cache_delete(sctx, nce);
		kfree(nce);
	}
}

static void name_cache_free(struct send_ctx *sctx)
{
	struct name_cache_entry *nce;

2009 2010 2011
	while (!list_empty(&sctx->name_cache_list)) {
		nce = list_entry(sctx->name_cache_list.next,
				struct name_cache_entry, list);
2012
		name_cache_delete(sctx, nce);
2013
		kfree(nce);
2014 2015 2016
	}
}

2017 2018 2019 2020 2021 2022 2023 2024
/*
 * Used by get_cur_path for each ref up to the root.
 * Returns 0 if it succeeded.
 * Returns 1 if the inode is not existent or got overwritten. In that case, the
 * name is an orphan name. This instructs get_cur_path to stop iterating. If 1
 * is returned, parent_ino/parent_gen are not guaranteed to be valid.
 * Returns <0 in case of error.
 */
2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035
static int __get_cur_name_and_parent(struct send_ctx *sctx,
				     u64 ino, u64 gen,
				     u64 *parent_ino,
				     u64 *parent_gen,
				     struct fs_path *dest)
{
	int ret;
	int nce_ret;
	struct btrfs_path *path = NULL;
	struct name_cache_entry *nce = NULL;

2036 2037 2038 2039 2040
	/*
	 * First check if we already did a call to this function with the same
	 * ino/gen. If yes, check if the cache entry is still up-to-date. If yes
	 * return the cached result.
	 */
2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062
	nce = name_cache_search(sctx, ino, gen);
	if (nce) {
		if (ino < sctx->send_progress && nce->need_later_update) {
			name_cache_delete(sctx, nce);
			kfree(nce);
			nce = NULL;
		} else {
			name_cache_used(sctx, nce);
			*parent_ino = nce->parent_ino;
			*parent_gen = nce->parent_gen;
			ret = fs_path_add(dest, nce->name, nce->name_len);
			if (ret < 0)
				goto out;
			ret = nce->ret;
			goto out;
		}
	}

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

2063 2064 2065 2066 2067
	/*
	 * If the inode is not existent yet, add the orphan name and return 1.
	 * This should only happen for the parent dir that we determine in
	 * __record_new_ref
	 */
2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079
	ret = is_inode_existent(sctx, ino, gen);
	if (ret < 0)
		goto out;

	if (!ret) {
		ret = gen_unique_name(sctx, ino, gen, dest);
		if (ret < 0)
			goto out;
		ret = 1;
		goto out_cache;
	}

2080 2081 2082 2083
	/*
	 * Depending on whether the inode was already processed or not, use
	 * send_root or parent_root for ref lookup.
	 */
2084
	if (ino < sctx->send_progress)
2085 2086
		ret = get_first_ref(sctx->send_root, ino,
				    parent_ino, parent_gen, dest);
2087
	else
2088 2089
		ret = get_first_ref(sctx->parent_root, ino,
				    parent_ino, parent_gen, dest);
2090 2091 2092
	if (ret < 0)
		goto out;

2093 2094 2095 2096
	/*
	 * Check if the ref was overwritten by an inode's ref that was processed
	 * earlier. If yes, treat as orphan and return 1.
	 */
2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109
	ret = did_overwrite_ref(sctx, *parent_ino, *parent_gen, ino, gen,
			dest->start, dest->end - dest->start);
	if (ret < 0)
		goto out;
	if (ret) {
		fs_path_reset(dest);
		ret = gen_unique_name(sctx, ino, gen, dest);
		if (ret < 0)
			goto out;
		ret = 1;
	}

out_cache:
2110 2111 2112
	/*
	 * Store the result of the lookup in the name cache.
	 */
2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175
	nce = kmalloc(sizeof(*nce) + fs_path_len(dest) + 1, GFP_NOFS);
	if (!nce) {
		ret = -ENOMEM;
		goto out;
	}

	nce->ino = ino;
	nce->gen = gen;
	nce->parent_ino = *parent_ino;
	nce->parent_gen = *parent_gen;
	nce->name_len = fs_path_len(dest);
	nce->ret = ret;
	strcpy(nce->name, dest->start);

	if (ino < sctx->send_progress)
		nce->need_later_update = 0;
	else
		nce->need_later_update = 1;

	nce_ret = name_cache_insert(sctx, nce);
	if (nce_ret < 0)
		ret = nce_ret;
	name_cache_clean_unused(sctx);

out:
	btrfs_free_path(path);
	return ret;
}

/*
 * Magic happens here. This function returns the first ref to an inode as it
 * would look like while receiving the stream at this point in time.
 * We walk the path up to the root. For every inode in between, we check if it
 * was already processed/sent. If yes, we continue with the parent as found
 * in send_root. If not, we continue with the parent as found in parent_root.
 * If we encounter an inode that was deleted at this point in time, we use the
 * inodes "orphan" name instead of the real name and stop. Same with new inodes
 * that were not created yet and overwritten inodes/refs.
 *
 * When do we have have orphan inodes:
 * 1. When an inode is freshly created and thus no valid refs are available yet
 * 2. When a directory lost all it's refs (deleted) but still has dir items
 *    inside which were not processed yet (pending for move/delete). If anyone
 *    tried to get the path to the dir items, it would get a path inside that
 *    orphan directory.
 * 3. When an inode is moved around or gets new links, it may overwrite the ref
 *    of an unprocessed inode. If in that case the first ref would be
 *    overwritten, the overwritten inode gets "orphanized". Later when we
 *    process this overwritten inode, it is restored at a new place by moving
 *    the orphan inode.
 *
 * sctx->send_progress tells this function at which point in time receiving
 * would be.
 */
static int get_cur_path(struct send_ctx *sctx, u64 ino, u64 gen,
			struct fs_path *dest)
{
	int ret = 0;
	struct fs_path *name = NULL;
	u64 parent_inode = 0;
	u64 parent_gen = 0;
	int stop = 0;

2176
	name = fs_path_alloc();
2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187
	if (!name) {
		ret = -ENOMEM;
		goto out;
	}

	dest->reversed = 1;
	fs_path_reset(dest);

	while (!stop && ino != BTRFS_FIRST_FREE_OBJECTID) {
		fs_path_reset(name);

2188 2189 2190 2191 2192 2193 2194 2195
		if (is_waiting_for_rm(sctx, ino)) {
			ret = gen_unique_name(sctx, ino, gen, name);
			if (ret < 0)
				goto out;
			ret = fs_path_add_path(dest, name);
			break;
		}

2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206
		if (is_waiting_for_move(sctx, ino)) {
			ret = get_first_ref(sctx->parent_root, ino,
					    &parent_inode, &parent_gen, name);
		} else {
			ret = __get_cur_name_and_parent(sctx, ino, gen,
							&parent_inode,
							&parent_gen, name);
			if (ret)
				stop = 1;
		}

2207 2208
		if (ret < 0)
			goto out;
2209

2210 2211 2212 2213 2214 2215 2216 2217 2218
		ret = fs_path_add_path(dest, name);
		if (ret < 0)
			goto out;

		ino = parent_inode;
		gen = parent_gen;
	}

out:
2219
	fs_path_free(name);
2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239
	if (!ret)
		fs_path_unreverse(dest);
	return ret;
}

/*
 * Sends a BTRFS_SEND_C_SUBVOL command/item to userspace
 */
static int send_subvol_begin(struct send_ctx *sctx)
{
	int ret;
	struct btrfs_root *send_root = sctx->send_root;
	struct btrfs_root *parent_root = sctx->parent_root;
	struct btrfs_path *path;
	struct btrfs_key key;
	struct btrfs_root_ref *ref;
	struct extent_buffer *leaf;
	char *name = NULL;
	int namelen;

2240
	path = btrfs_alloc_path();
2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288
	if (!path)
		return -ENOMEM;

	name = kmalloc(BTRFS_PATH_NAME_MAX, GFP_NOFS);
	if (!name) {
		btrfs_free_path(path);
		return -ENOMEM;
	}

	key.objectid = send_root->objectid;
	key.type = BTRFS_ROOT_BACKREF_KEY;
	key.offset = 0;

	ret = btrfs_search_slot_for_read(send_root->fs_info->tree_root,
				&key, path, 1, 0);
	if (ret < 0)
		goto out;
	if (ret) {
		ret = -ENOENT;
		goto out;
	}

	leaf = path->nodes[0];
	btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
	if (key.type != BTRFS_ROOT_BACKREF_KEY ||
	    key.objectid != send_root->objectid) {
		ret = -ENOENT;
		goto out;
	}
	ref = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_root_ref);
	namelen = btrfs_root_ref_name_len(leaf, ref);
	read_extent_buffer(leaf, name, (unsigned long)(ref + 1), namelen);
	btrfs_release_path(path);

	if (parent_root) {
		ret = begin_cmd(sctx, BTRFS_SEND_C_SNAPSHOT);
		if (ret < 0)
			goto out;
	} else {
		ret = begin_cmd(sctx, BTRFS_SEND_C_SUBVOL);
		if (ret < 0)
			goto out;
	}

	TLV_PUT_STRING(sctx, BTRFS_SEND_A_PATH, name, namelen);
	TLV_PUT_UUID(sctx, BTRFS_SEND_A_UUID,
			sctx->send_root->root_item.uuid);
	TLV_PUT_U64(sctx, BTRFS_SEND_A_CTRANSID,
2289
		    le64_to_cpu(sctx->send_root->root_item.ctransid));
2290 2291 2292 2293
	if (parent_root) {
		TLV_PUT_UUID(sctx, BTRFS_SEND_A_CLONE_UUID,
				sctx->parent_root->root_item.uuid);
		TLV_PUT_U64(sctx, BTRFS_SEND_A_CLONE_CTRANSID,
2294
			    le64_to_cpu(sctx->parent_root->root_item.ctransid));
2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312
	}

	ret = send_cmd(sctx);

tlv_put_failure:
out:
	btrfs_free_path(path);
	kfree(name);
	return ret;
}

static int send_truncate(struct send_ctx *sctx, u64 ino, u64 gen, u64 size)
{
	int ret = 0;
	struct fs_path *p;

verbose_printk("btrfs: send_truncate %llu size=%llu\n", ino, size);

2313
	p = fs_path_alloc();
2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330
	if (!p)
		return -ENOMEM;

	ret = begin_cmd(sctx, BTRFS_SEND_C_TRUNCATE);
	if (ret < 0)
		goto out;

	ret = get_cur_path(sctx, ino, gen, p);
	if (ret < 0)
		goto out;
	TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p);
	TLV_PUT_U64(sctx, BTRFS_SEND_A_SIZE, size);

	ret = send_cmd(sctx);

tlv_put_failure:
out:
2331
	fs_path_free(p);
2332 2333 2334 2335 2336 2337 2338 2339 2340 2341
	return ret;
}

static int send_chmod(struct send_ctx *sctx, u64 ino, u64 gen, u64 mode)
{
	int ret = 0;
	struct fs_path *p;

verbose_printk("btrfs: send_chmod %llu mode=%llu\n", ino, mode);

2342
	p = fs_path_alloc();
2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359
	if (!p)
		return -ENOMEM;

	ret = begin_cmd(sctx, BTRFS_SEND_C_CHMOD);
	if (ret < 0)
		goto out;

	ret = get_cur_path(sctx, ino, gen, p);
	if (ret < 0)
		goto out;
	TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p);
	TLV_PUT_U64(sctx, BTRFS_SEND_A_MODE, mode & 07777);

	ret = send_cmd(sctx);

tlv_put_failure:
out:
2360
	fs_path_free(p);
2361 2362 2363 2364 2365 2366 2367 2368 2369 2370
	return ret;
}

static int send_chown(struct send_ctx *sctx, u64 ino, u64 gen, u64 uid, u64 gid)
{
	int ret = 0;
	struct fs_path *p;

verbose_printk("btrfs: send_chown %llu uid=%llu, gid=%llu\n", ino, uid, gid);

2371
	p = fs_path_alloc();
2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389
	if (!p)
		return -ENOMEM;

	ret = begin_cmd(sctx, BTRFS_SEND_C_CHOWN);
	if (ret < 0)
		goto out;

	ret = get_cur_path(sctx, ino, gen, p);
	if (ret < 0)
		goto out;
	TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p);
	TLV_PUT_U64(sctx, BTRFS_SEND_A_UID, uid);
	TLV_PUT_U64(sctx, BTRFS_SEND_A_GID, gid);

	ret = send_cmd(sctx);

tlv_put_failure:
out:
2390
	fs_path_free(p);
2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405
	return ret;
}

static int send_utimes(struct send_ctx *sctx, u64 ino, u64 gen)
{
	int ret = 0;
	struct fs_path *p = NULL;
	struct btrfs_inode_item *ii;
	struct btrfs_path *path = NULL;
	struct extent_buffer *eb;
	struct btrfs_key key;
	int slot;

verbose_printk("btrfs: send_utimes %llu\n", ino);

2406
	p = fs_path_alloc();
2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440
	if (!p)
		return -ENOMEM;

	path = alloc_path_for_send();
	if (!path) {
		ret = -ENOMEM;
		goto out;
	}

	key.objectid = ino;
	key.type = BTRFS_INODE_ITEM_KEY;
	key.offset = 0;
	ret = btrfs_search_slot(NULL, sctx->send_root, &key, path, 0, 0);
	if (ret < 0)
		goto out;

	eb = path->nodes[0];
	slot = path->slots[0];
	ii = btrfs_item_ptr(eb, slot, struct btrfs_inode_item);

	ret = begin_cmd(sctx, BTRFS_SEND_C_UTIMES);
	if (ret < 0)
		goto out;

	ret = get_cur_path(sctx, ino, gen, p);
	if (ret < 0)
		goto out;
	TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p);
	TLV_PUT_BTRFS_TIMESPEC(sctx, BTRFS_SEND_A_ATIME, eb,
			btrfs_inode_atime(ii));
	TLV_PUT_BTRFS_TIMESPEC(sctx, BTRFS_SEND_A_MTIME, eb,
			btrfs_inode_mtime(ii));
	TLV_PUT_BTRFS_TIMESPEC(sctx, BTRFS_SEND_A_CTIME, eb,
			btrfs_inode_ctime(ii));
2441
	/* TODO Add otime support when the otime patches get into upstream */
2442 2443 2444 2445 2446

	ret = send_cmd(sctx);

tlv_put_failure:
out:
2447
	fs_path_free(p);
2448 2449 2450 2451 2452 2453 2454 2455 2456
	btrfs_free_path(path);
	return ret;
}

/*
 * Sends a BTRFS_SEND_C_MKXXX or SYMLINK command to user space. We don't have
 * a valid path yet because we did not process the refs yet. So, the inode
 * is created as orphan.
 */
2457
static int send_create_inode(struct send_ctx *sctx, u64 ino)
2458 2459 2460 2461
{
	int ret = 0;
	struct fs_path *p;
	int cmd;
2462
	u64 gen;
2463
	u64 mode;
2464
	u64 rdev;
2465

2466
verbose_printk("btrfs: send_create_inode %llu\n", ino);
2467

2468
	p = fs_path_alloc();
2469 2470 2471
	if (!p)
		return -ENOMEM;

L
Liu Bo 已提交
2472 2473 2474 2475 2476 2477 2478 2479 2480 2481
	if (ino != sctx->cur_ino) {
		ret = get_inode_info(sctx->send_root, ino, NULL, &gen, &mode,
				     NULL, NULL, &rdev);
		if (ret < 0)
			goto out;
	} else {
		gen = sctx->cur_inode_gen;
		mode = sctx->cur_inode_mode;
		rdev = sctx->cur_inode_rdev;
	}
2482

2483
	if (S_ISREG(mode)) {
2484
		cmd = BTRFS_SEND_C_MKFILE;
2485
	} else if (S_ISDIR(mode)) {
2486
		cmd = BTRFS_SEND_C_MKDIR;
2487
	} else if (S_ISLNK(mode)) {
2488
		cmd = BTRFS_SEND_C_SYMLINK;
2489
	} else if (S_ISCHR(mode) || S_ISBLK(mode)) {
2490
		cmd = BTRFS_SEND_C_MKNOD;
2491
	} else if (S_ISFIFO(mode)) {
2492
		cmd = BTRFS_SEND_C_MKFIFO;
2493
	} else if (S_ISSOCK(mode)) {
2494
		cmd = BTRFS_SEND_C_MKSOCK;
2495
	} else {
2496 2497 2498 2499 2500 2501 2502 2503 2504 2505
		printk(KERN_WARNING "btrfs: unexpected inode type %o",
				(int)(mode & S_IFMT));
		ret = -ENOTSUPP;
		goto out;
	}

	ret = begin_cmd(sctx, cmd);
	if (ret < 0)
		goto out;

2506
	ret = gen_unique_name(sctx, ino, gen, p);
2507 2508 2509 2510
	if (ret < 0)
		goto out;

	TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p);
2511
	TLV_PUT_U64(sctx, BTRFS_SEND_A_INO, ino);
2512 2513 2514

	if (S_ISLNK(mode)) {
		fs_path_reset(p);
2515
		ret = read_symlink(sctx->send_root, ino, p);
2516 2517 2518 2519 2520
		if (ret < 0)
			goto out;
		TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH_LINK, p);
	} else if (S_ISCHR(mode) || S_ISBLK(mode) ||
		   S_ISFIFO(mode) || S_ISSOCK(mode)) {
2521 2522
		TLV_PUT_U64(sctx, BTRFS_SEND_A_RDEV, new_encode_dev(rdev));
		TLV_PUT_U64(sctx, BTRFS_SEND_A_MODE, mode);
2523 2524 2525 2526 2527 2528 2529 2530 2531
	}

	ret = send_cmd(sctx);
	if (ret < 0)
		goto out;


tlv_put_failure:
out:
2532
	fs_path_free(p);
2533 2534 2535
	return ret;
}

2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560
/*
 * We need some special handling for inodes that get processed before the parent
 * directory got created. See process_recorded_refs for details.
 * This function does the check if we already created the dir out of order.
 */
static int did_create_dir(struct send_ctx *sctx, u64 dir)
{
	int ret = 0;
	struct btrfs_path *path = NULL;
	struct btrfs_key key;
	struct btrfs_key found_key;
	struct btrfs_key di_key;
	struct extent_buffer *eb;
	struct btrfs_dir_item *di;
	int slot;

	path = alloc_path_for_send();
	if (!path) {
		ret = -ENOMEM;
		goto out;
	}

	key.objectid = dir;
	key.type = BTRFS_DIR_INDEX_KEY;
	key.offset = 0;
2561 2562 2563 2564
	ret = btrfs_search_slot(NULL, sctx->send_root, &key, path, 0, 0);
	if (ret < 0)
		goto out;

2565
	while (1) {
2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576
		eb = path->nodes[0];
		slot = path->slots[0];
		if (slot >= btrfs_header_nritems(eb)) {
			ret = btrfs_next_leaf(sctx->send_root, path);
			if (ret < 0) {
				goto out;
			} else if (ret > 0) {
				ret = 0;
				break;
			}
			continue;
2577
		}
2578 2579 2580

		btrfs_item_key_to_cpu(eb, &found_key, slot);
		if (found_key.objectid != key.objectid ||
2581 2582 2583 2584 2585 2586 2587 2588
		    found_key.type != key.type) {
			ret = 0;
			goto out;
		}

		di = btrfs_item_ptr(eb, slot, struct btrfs_dir_item);
		btrfs_dir_item_key_to_cpu(eb, di, &di_key);

2589 2590
		if (di_key.type != BTRFS_ROOT_ITEM_KEY &&
		    di_key.objectid < sctx->send_progress) {
2591 2592 2593 2594
			ret = 1;
			goto out;
		}

2595
		path->slots[0]++;
2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630
	}

out:
	btrfs_free_path(path);
	return ret;
}

/*
 * Only creates the inode if it is:
 * 1. Not a directory
 * 2. Or a directory which was not created already due to out of order
 *    directories. See did_create_dir and process_recorded_refs for details.
 */
static int send_create_inode_if_needed(struct send_ctx *sctx)
{
	int ret;

	if (S_ISDIR(sctx->cur_inode_mode)) {
		ret = did_create_dir(sctx, sctx->cur_ino);
		if (ret < 0)
			goto out;
		if (ret) {
			ret = 0;
			goto out;
		}
	}

	ret = send_create_inode(sctx, sctx->cur_ino);
	if (ret < 0)
		goto out;

out:
	return ret;
}

2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646
struct recorded_ref {
	struct list_head list;
	char *dir_path;
	char *name;
	struct fs_path *full_path;
	u64 dir;
	u64 dir_gen;
	int dir_path_len;
	int name_len;
};

/*
 * We need to process new refs before deleted refs, but compare_tree gives us
 * everything mixed. So we first record all refs and later process them.
 * This function is a helper to record one ref.
 */
2647
static int __record_ref(struct list_head *head, u64 dir,
2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659
		      u64 dir_gen, struct fs_path *path)
{
	struct recorded_ref *ref;

	ref = kmalloc(sizeof(*ref), GFP_NOFS);
	if (!ref)
		return -ENOMEM;

	ref->dir = dir;
	ref->dir_gen = dir_gen;
	ref->full_path = path;

A
Andy Shevchenko 已提交
2660 2661 2662 2663
	ref->name = (char *)kbasename(ref->full_path->start);
	ref->name_len = ref->full_path->end - ref->name;
	ref->dir_path = ref->full_path->start;
	if (ref->name == ref->full_path->start)
2664
		ref->dir_path_len = 0;
A
Andy Shevchenko 已提交
2665
	else
2666 2667 2668 2669 2670 2671 2672
		ref->dir_path_len = ref->full_path->end -
				ref->full_path->start - 1 - ref->name_len;

	list_add_tail(&ref->list, head);
	return 0;
}

2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688
static int dup_ref(struct recorded_ref *ref, struct list_head *list)
{
	struct recorded_ref *new;

	new = kmalloc(sizeof(*ref), GFP_NOFS);
	if (!new)
		return -ENOMEM;

	new->dir = ref->dir;
	new->dir_gen = ref->dir_gen;
	new->full_path = NULL;
	INIT_LIST_HEAD(&new->list);
	list_add_tail(&new->list, list);
	return 0;
}

2689
static void __free_recorded_refs(struct list_head *head)
2690 2691 2692
{
	struct recorded_ref *cur;

2693 2694
	while (!list_empty(head)) {
		cur = list_entry(head->next, struct recorded_ref, list);
2695
		fs_path_free(cur->full_path);
2696
		list_del(&cur->list);
2697 2698 2699 2700 2701 2702
		kfree(cur);
	}
}

static void free_recorded_refs(struct send_ctx *sctx)
{
2703 2704
	__free_recorded_refs(&sctx->new_refs);
	__free_recorded_refs(&sctx->deleted_refs);
2705 2706 2707
}

/*
2708
 * Renames/moves a file/dir to its orphan name. Used when the first
2709 2710 2711 2712 2713 2714 2715 2716 2717
 * ref of an unprocessed inode gets overwritten and for all non empty
 * directories.
 */
static int orphanize_inode(struct send_ctx *sctx, u64 ino, u64 gen,
			  struct fs_path *path)
{
	int ret;
	struct fs_path *orphan;

2718
	orphan = fs_path_alloc();
2719 2720 2721 2722 2723 2724 2725 2726 2727 2728
	if (!orphan)
		return -ENOMEM;

	ret = gen_unique_name(sctx, ino, gen, orphan);
	if (ret < 0)
		goto out;

	ret = send_rename(sctx, path, orphan);

out:
2729
	fs_path_free(orphan);
2730 2731 2732
	return ret;
}

2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797
static struct orphan_dir_info *
add_orphan_dir_info(struct send_ctx *sctx, u64 dir_ino)
{
	struct rb_node **p = &sctx->orphan_dirs.rb_node;
	struct rb_node *parent = NULL;
	struct orphan_dir_info *entry, *odi;

	odi = kmalloc(sizeof(*odi), GFP_NOFS);
	if (!odi)
		return ERR_PTR(-ENOMEM);
	odi->ino = dir_ino;
	odi->gen = 0;

	while (*p) {
		parent = *p;
		entry = rb_entry(parent, struct orphan_dir_info, node);
		if (dir_ino < entry->ino) {
			p = &(*p)->rb_left;
		} else if (dir_ino > entry->ino) {
			p = &(*p)->rb_right;
		} else {
			kfree(odi);
			return entry;
		}
	}

	rb_link_node(&odi->node, parent, p);
	rb_insert_color(&odi->node, &sctx->orphan_dirs);
	return odi;
}

static struct orphan_dir_info *
get_orphan_dir_info(struct send_ctx *sctx, u64 dir_ino)
{
	struct rb_node *n = sctx->orphan_dirs.rb_node;
	struct orphan_dir_info *entry;

	while (n) {
		entry = rb_entry(n, struct orphan_dir_info, node);
		if (dir_ino < entry->ino)
			n = n->rb_left;
		else if (dir_ino > entry->ino)
			n = n->rb_right;
		else
			return entry;
	}
	return NULL;
}

static int is_waiting_for_rm(struct send_ctx *sctx, u64 dir_ino)
{
	struct orphan_dir_info *odi = get_orphan_dir_info(sctx, dir_ino);

	return odi != NULL;
}

static void free_orphan_dir_info(struct send_ctx *sctx,
				 struct orphan_dir_info *odi)
{
	if (!odi)
		return;
	rb_erase(&odi->node, &sctx->orphan_dirs);
	kfree(odi);
}

2798 2799 2800 2801 2802
/*
 * Returns 1 if a directory can be removed at this point in time.
 * We check this by iterating all dir items and checking if the inode behind
 * the dir item was already processed.
 */
2803 2804
static int can_rmdir(struct send_ctx *sctx, u64 dir, u64 dir_gen,
		     u64 send_progress)
2805 2806 2807 2808 2809 2810 2811 2812 2813
{
	int ret = 0;
	struct btrfs_root *root = sctx->parent_root;
	struct btrfs_path *path;
	struct btrfs_key key;
	struct btrfs_key found_key;
	struct btrfs_key loc;
	struct btrfs_dir_item *di;

2814 2815 2816 2817 2818 2819
	/*
	 * Don't try to rmdir the top/root subvolume dir.
	 */
	if (dir == BTRFS_FIRST_FREE_OBJECTID)
		return 0;

2820 2821 2822 2823 2824 2825 2826
	path = alloc_path_for_send();
	if (!path)
		return -ENOMEM;

	key.objectid = dir;
	key.type = BTRFS_DIR_INDEX_KEY;
	key.offset = 0;
2827 2828 2829
	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
	if (ret < 0)
		goto out;
2830 2831

	while (1) {
2832 2833
		struct waiting_dir_move *dm;

2834 2835 2836 2837 2838 2839 2840
		if (path->slots[0] >= btrfs_header_nritems(path->nodes[0])) {
			ret = btrfs_next_leaf(root, path);
			if (ret < 0)
				goto out;
			else if (ret > 0)
				break;
			continue;
2841
		}
2842 2843 2844 2845
		btrfs_item_key_to_cpu(path->nodes[0], &found_key,
				      path->slots[0]);
		if (found_key.objectid != key.objectid ||
		    found_key.type != key.type)
2846 2847 2848 2849 2850 2851
			break;

		di = btrfs_item_ptr(path->nodes[0], path->slots[0],
				struct btrfs_dir_item);
		btrfs_dir_item_key_to_cpu(path->nodes[0], di, &loc);

2852 2853 2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864 2865 2866
		dm = get_waiting_dir_move(sctx, loc.objectid);
		if (dm) {
			struct orphan_dir_info *odi;

			odi = add_orphan_dir_info(sctx, dir);
			if (IS_ERR(odi)) {
				ret = PTR_ERR(odi);
				goto out;
			}
			odi->gen = dir_gen;
			dm->rmdir_ino = dir;
			ret = 0;
			goto out;
		}

2867 2868 2869 2870 2871
		if (loc.objectid > send_progress) {
			ret = 0;
			goto out;
		}

2872
		path->slots[0]++;
2873 2874 2875 2876 2877 2878 2879 2880 2881
	}

	ret = 1;

out:
	btrfs_free_path(path);
	return ret;
}

2882 2883
static int is_waiting_for_move(struct send_ctx *sctx, u64 ino)
{
2884
	struct waiting_dir_move *entry = get_waiting_dir_move(sctx, ino);
2885

2886
	return entry != NULL;
2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898
}

static int add_waiting_dir_move(struct send_ctx *sctx, u64 ino)
{
	struct rb_node **p = &sctx->waiting_dir_moves.rb_node;
	struct rb_node *parent = NULL;
	struct waiting_dir_move *entry, *dm;

	dm = kmalloc(sizeof(*dm), GFP_NOFS);
	if (!dm)
		return -ENOMEM;
	dm->ino = ino;
2899
	dm->rmdir_ino = 0;
2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918

	while (*p) {
		parent = *p;
		entry = rb_entry(parent, struct waiting_dir_move, node);
		if (ino < entry->ino) {
			p = &(*p)->rb_left;
		} else if (ino > entry->ino) {
			p = &(*p)->rb_right;
		} else {
			kfree(dm);
			return -EEXIST;
		}
	}

	rb_link_node(&dm->node, parent, p);
	rb_insert_color(&dm->node, &sctx->waiting_dir_moves);
	return 0;
}

2919 2920
static struct waiting_dir_move *
get_waiting_dir_move(struct send_ctx *sctx, u64 ino)
2921 2922 2923 2924 2925 2926
{
	struct rb_node *n = sctx->waiting_dir_moves.rb_node;
	struct waiting_dir_move *entry;

	while (n) {
		entry = rb_entry(n, struct waiting_dir_move, node);
2927
		if (ino < entry->ino)
2928
			n = n->rb_left;
2929
		else if (ino > entry->ino)
2930
			n = n->rb_right;
2931 2932
		else
			return entry;
2933
	}
2934 2935 2936 2937 2938 2939 2940 2941 2942 2943
	return NULL;
}

static void free_waiting_dir_move(struct send_ctx *sctx,
				  struct waiting_dir_move *dm)
{
	if (!dm)
		return;
	rb_erase(&dm->node, &sctx->waiting_dir_moves);
	kfree(dm);
2944 2945
}

2946 2947 2948 2949
static int add_pending_dir_move(struct send_ctx *sctx,
				u64 ino,
				u64 ino_gen,
				u64 parent_ino)
2950 2951 2952
{
	struct rb_node **p = &sctx->pending_dir_moves.rb_node;
	struct rb_node *parent = NULL;
C
Chris Mason 已提交
2953
	struct pending_dir_move *entry = NULL, *pm;
2954 2955 2956 2957 2958 2959 2960 2961
	struct recorded_ref *cur;
	int exists = 0;
	int ret;

	pm = kmalloc(sizeof(*pm), GFP_NOFS);
	if (!pm)
		return -ENOMEM;
	pm->parent_ino = parent_ino;
2962 2963
	pm->ino = ino;
	pm->gen = ino_gen;
2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032
	INIT_LIST_HEAD(&pm->list);
	INIT_LIST_HEAD(&pm->update_refs);
	RB_CLEAR_NODE(&pm->node);

	while (*p) {
		parent = *p;
		entry = rb_entry(parent, struct pending_dir_move, node);
		if (parent_ino < entry->parent_ino) {
			p = &(*p)->rb_left;
		} else if (parent_ino > entry->parent_ino) {
			p = &(*p)->rb_right;
		} else {
			exists = 1;
			break;
		}
	}

	list_for_each_entry(cur, &sctx->deleted_refs, list) {
		ret = dup_ref(cur, &pm->update_refs);
		if (ret < 0)
			goto out;
	}
	list_for_each_entry(cur, &sctx->new_refs, list) {
		ret = dup_ref(cur, &pm->update_refs);
		if (ret < 0)
			goto out;
	}

	ret = add_waiting_dir_move(sctx, pm->ino);
	if (ret)
		goto out;

	if (exists) {
		list_add_tail(&pm->list, &entry->list);
	} else {
		rb_link_node(&pm->node, parent, p);
		rb_insert_color(&pm->node, &sctx->pending_dir_moves);
	}
	ret = 0;
out:
	if (ret) {
		__free_recorded_refs(&pm->update_refs);
		kfree(pm);
	}
	return ret;
}

static struct pending_dir_move *get_pending_dir_moves(struct send_ctx *sctx,
						      u64 parent_ino)
{
	struct rb_node *n = sctx->pending_dir_moves.rb_node;
	struct pending_dir_move *entry;

	while (n) {
		entry = rb_entry(n, struct pending_dir_move, node);
		if (parent_ino < entry->parent_ino)
			n = n->rb_left;
		else if (parent_ino > entry->parent_ino)
			n = n->rb_right;
		else
			return entry;
	}
	return NULL;
}

static int apply_dir_move(struct send_ctx *sctx, struct pending_dir_move *pm)
{
	struct fs_path *from_path = NULL;
	struct fs_path *to_path = NULL;
3033
	struct fs_path *name = NULL;
3034 3035
	u64 orig_progress = sctx->send_progress;
	struct recorded_ref *cur;
3036
	u64 parent_ino, parent_gen;
3037 3038
	struct waiting_dir_move *dm = NULL;
	u64 rmdir_ino = 0;
3039 3040
	int ret;

3041
	name = fs_path_alloc();
3042
	from_path = fs_path_alloc();
3043 3044 3045 3046
	if (!name || !from_path) {
		ret = -ENOMEM;
		goto out;
	}
3047

3048 3049 3050 3051
	dm = get_waiting_dir_move(sctx, pm->ino);
	ASSERT(dm);
	rmdir_ino = dm->rmdir_ino;
	free_waiting_dir_move(sctx, dm);
3052 3053 3054

	ret = get_first_ref(sctx->parent_root, pm->ino,
			    &parent_ino, &parent_gen, name);
3055 3056 3057
	if (ret < 0)
		goto out;

3058 3059 3060 3061 3062 3063 3064
	ret = get_cur_path(sctx, parent_ino, parent_gen,
			   from_path);
	if (ret < 0)
		goto out;
	ret = fs_path_add_path(from_path, name);
	if (ret < 0)
		goto out;
3065

3066 3067
	fs_path_reset(name);
	to_path = name;
3068 3069
	name = NULL;

3070 3071 3072 3073 3074 3075 3076 3077 3078
	sctx->send_progress = sctx->cur_ino + 1;
	ret = get_cur_path(sctx, pm->ino, pm->gen, to_path);
	if (ret < 0)
		goto out;

	ret = send_rename(sctx, from_path, to_path);
	if (ret < 0)
		goto out;

3079 3080 3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107
	if (rmdir_ino) {
		struct orphan_dir_info *odi;

		odi = get_orphan_dir_info(sctx, rmdir_ino);
		if (!odi) {
			/* already deleted */
			goto finish;
		}
		ret = can_rmdir(sctx, rmdir_ino, odi->gen, sctx->cur_ino + 1);
		if (ret < 0)
			goto out;
		if (!ret)
			goto finish;

		name = fs_path_alloc();
		if (!name) {
			ret = -ENOMEM;
			goto out;
		}
		ret = get_cur_path(sctx, rmdir_ino, odi->gen, name);
		if (ret < 0)
			goto out;
		ret = send_rmdir(sctx, name);
		if (ret < 0)
			goto out;
		free_orphan_dir_info(sctx, odi);
	}

finish:
3108 3109 3110 3111 3112 3113 3114 3115 3116
	ret = send_utimes(sctx, pm->ino, pm->gen);
	if (ret < 0)
		goto out;

	/*
	 * After rename/move, need to update the utimes of both new parent(s)
	 * and old parent(s).
	 */
	list_for_each_entry(cur, &pm->update_refs, list) {
3117 3118
		if (cur->dir == rmdir_ino)
			continue;
3119 3120 3121 3122 3123 3124
		ret = send_utimes(sctx, cur->dir, cur->dir_gen);
		if (ret < 0)
			goto out;
	}

out:
3125
	fs_path_free(name);
3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196
	fs_path_free(from_path);
	fs_path_free(to_path);
	sctx->send_progress = orig_progress;

	return ret;
}

static void free_pending_move(struct send_ctx *sctx, struct pending_dir_move *m)
{
	if (!list_empty(&m->list))
		list_del(&m->list);
	if (!RB_EMPTY_NODE(&m->node))
		rb_erase(&m->node, &sctx->pending_dir_moves);
	__free_recorded_refs(&m->update_refs);
	kfree(m);
}

static void tail_append_pending_moves(struct pending_dir_move *moves,
				      struct list_head *stack)
{
	if (list_empty(&moves->list)) {
		list_add_tail(&moves->list, stack);
	} else {
		LIST_HEAD(list);
		list_splice_init(&moves->list, &list);
		list_add_tail(&moves->list, stack);
		list_splice_tail(&list, stack);
	}
}

static int apply_children_dir_moves(struct send_ctx *sctx)
{
	struct pending_dir_move *pm;
	struct list_head stack;
	u64 parent_ino = sctx->cur_ino;
	int ret = 0;

	pm = get_pending_dir_moves(sctx, parent_ino);
	if (!pm)
		return 0;

	INIT_LIST_HEAD(&stack);
	tail_append_pending_moves(pm, &stack);

	while (!list_empty(&stack)) {
		pm = list_first_entry(&stack, struct pending_dir_move, list);
		parent_ino = pm->ino;
		ret = apply_dir_move(sctx, pm);
		free_pending_move(sctx, pm);
		if (ret)
			goto out;
		pm = get_pending_dir_moves(sctx, parent_ino);
		if (pm)
			tail_append_pending_moves(pm, &stack);
	}
	return 0;

out:
	while (!list_empty(&stack)) {
		pm = list_first_entry(&stack, struct pending_dir_move, list);
		free_pending_move(sctx, pm);
	}
	return ret;
}

static int wait_for_parent_move(struct send_ctx *sctx,
				struct recorded_ref *parent_ref)
{
	int ret;
	u64 ino = parent_ref->dir;
	u64 parent_ino_before, parent_ino_after;
3197
	u64 old_gen;
3198 3199 3200
	struct fs_path *path_before = NULL;
	struct fs_path *path_after = NULL;
	int len1, len2;
3201 3202
	int register_upper_dirs;
	u64 gen;
3203 3204 3205 3206

	if (is_waiting_for_move(sctx, ino))
		return 1;

3207 3208 3209
	if (parent_ref->dir <= sctx->cur_ino)
		return 0;

3210 3211 3212 3213 3214 3215 3216
	ret = get_inode_info(sctx->parent_root, ino, NULL, &old_gen,
			     NULL, NULL, NULL, NULL);
	if (ret == -ENOENT)
		return 0;
	else if (ret < 0)
		return ret;

3217
	if (parent_ref->dir_gen != old_gen)
3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239
		return 0;

	path_before = fs_path_alloc();
	if (!path_before)
		return -ENOMEM;

	ret = get_first_ref(sctx->parent_root, ino, &parent_ino_before,
			    NULL, path_before);
	if (ret == -ENOENT) {
		ret = 0;
		goto out;
	} else if (ret < 0) {
		goto out;
	}

	path_after = fs_path_alloc();
	if (!path_after) {
		ret = -ENOMEM;
		goto out;
	}

	ret = get_first_ref(sctx->send_root, ino, &parent_ino_after,
3240
			    &gen, path_after);
3241 3242 3243 3244 3245 3246 3247 3248 3249
	if (ret == -ENOENT) {
		ret = 0;
		goto out;
	} else if (ret < 0) {
		goto out;
	}

	len1 = fs_path_len(path_before);
	len2 = fs_path_len(path_after);
3250 3251
	if (parent_ino_before != parent_ino_after || len1 != len2 ||
	     memcmp(path_before->start, path_after->start, len1)) {
3252 3253 3254 3255 3256
		ret = 1;
		goto out;
	}
	ret = 0;

3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310
	/*
	 * Ok, our new most direct ancestor has a higher inode number but
	 * wasn't moved/renamed. So maybe some of the new ancestors higher in
	 * the hierarchy have an higher inode number too *and* were renamed
	 * or moved - in this case we need to wait for the ancestor's rename
	 * or move operation before we can do the move/rename for the current
	 * inode.
	 */
	register_upper_dirs = 0;
	ino = parent_ino_after;
again:
	while ((ret == 0 || register_upper_dirs) && ino > sctx->cur_ino) {
		u64 parent_gen;

		fs_path_reset(path_before);
		fs_path_reset(path_after);

		ret = get_first_ref(sctx->send_root, ino, &parent_ino_after,
				    &parent_gen, path_after);
		if (ret < 0)
			goto out;
		ret = get_first_ref(sctx->parent_root, ino, &parent_ino_before,
				    NULL, path_before);
		if (ret == -ENOENT) {
			ret = 0;
			break;
		} else if (ret < 0) {
			goto out;
		}

		len1 = fs_path_len(path_before);
		len2 = fs_path_len(path_after);
		if (parent_ino_before != parent_ino_after || len1 != len2 ||
		    memcmp(path_before->start, path_after->start, len1)) {
			ret = 1;
			if (register_upper_dirs) {
				break;
			} else {
				register_upper_dirs = 1;
				ino = parent_ref->dir;
				gen = parent_ref->dir_gen;
				goto again;
			}
		} else if (register_upper_dirs) {
			ret = add_pending_dir_move(sctx, ino, gen,
						   parent_ino_after);
			if (ret < 0 && ret != -EEXIST)
				goto out;
		}

		ino = parent_ino_after;
		gen = parent_gen;
	}

3311 3312 3313 3314 3315 3316 3317
out:
	fs_path_free(path_before);
	fs_path_free(path_after);

	return ret;
}

3318 3319 3320
/*
 * This does all the move/link/unlink/rmdir magic.
 */
3321
static int process_recorded_refs(struct send_ctx *sctx, int *pending_move)
3322 3323 3324
{
	int ret = 0;
	struct recorded_ref *cur;
3325
	struct recorded_ref *cur2;
3326
	struct list_head check_dirs;
3327
	struct fs_path *valid_path = NULL;
3328
	u64 ow_inode = 0;
3329 3330 3331
	u64 ow_gen;
	int did_overwrite = 0;
	int is_orphan = 0;
3332
	u64 last_dir_ino_rm = 0;
3333 3334 3335

verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino);

3336 3337 3338 3339 3340
	/*
	 * This should never happen as the root dir always has the same ref
	 * which is always '..'
	 */
	BUG_ON(sctx->cur_ino <= BTRFS_FIRST_FREE_OBJECTID);
3341
	INIT_LIST_HEAD(&check_dirs);
3342

3343
	valid_path = fs_path_alloc();
3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381
	if (!valid_path) {
		ret = -ENOMEM;
		goto out;
	}

	/*
	 * First, check if the first ref of the current inode was overwritten
	 * before. If yes, we know that the current inode was already orphanized
	 * and thus use the orphan name. If not, we can use get_cur_path to
	 * get the path of the first ref as it would like while receiving at
	 * this point in time.
	 * New inodes are always orphan at the beginning, so force to use the
	 * orphan name in this case.
	 * The first ref is stored in valid_path and will be updated if it
	 * gets moved around.
	 */
	if (!sctx->cur_inode_new) {
		ret = did_overwrite_first_ref(sctx, sctx->cur_ino,
				sctx->cur_inode_gen);
		if (ret < 0)
			goto out;
		if (ret)
			did_overwrite = 1;
	}
	if (sctx->cur_inode_new || did_overwrite) {
		ret = gen_unique_name(sctx, sctx->cur_ino,
				sctx->cur_inode_gen, valid_path);
		if (ret < 0)
			goto out;
		is_orphan = 1;
	} else {
		ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen,
				valid_path);
		if (ret < 0)
			goto out;
	}

	list_for_each_entry(cur, &sctx->new_refs, list) {
3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421
		/*
		 * We may have refs where the parent directory does not exist
		 * yet. This happens if the parent directories inum is higher
		 * the the current inum. To handle this case, we create the
		 * parent directory out of order. But we need to check if this
		 * did already happen before due to other refs in the same dir.
		 */
		ret = get_cur_inode_state(sctx, cur->dir, cur->dir_gen);
		if (ret < 0)
			goto out;
		if (ret == inode_state_will_create) {
			ret = 0;
			/*
			 * First check if any of the current inodes refs did
			 * already create the dir.
			 */
			list_for_each_entry(cur2, &sctx->new_refs, list) {
				if (cur == cur2)
					break;
				if (cur2->dir == cur->dir) {
					ret = 1;
					break;
				}
			}

			/*
			 * If that did not happen, check if a previous inode
			 * did already create the dir.
			 */
			if (!ret)
				ret = did_create_dir(sctx, cur->dir);
			if (ret < 0)
				goto out;
			if (!ret) {
				ret = send_create_inode(sctx, cur->dir);
				if (ret < 0)
					goto out;
			}
		}

3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433
		/*
		 * Check if this new ref would overwrite the first ref of
		 * another unprocessed inode. If yes, orphanize the
		 * overwritten inode. If we find an overwritten ref that is
		 * not the first ref, simply unlink it.
		 */
		ret = will_overwrite_ref(sctx, cur->dir, cur->dir_gen,
				cur->name, cur->name_len,
				&ow_inode, &ow_gen);
		if (ret < 0)
			goto out;
		if (ret) {
3434 3435 3436
			ret = is_first_ref(sctx->parent_root,
					   ow_inode, cur->dir, cur->name,
					   cur->name_len);
3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455
			if (ret < 0)
				goto out;
			if (ret) {
				ret = orphanize_inode(sctx, ow_inode, ow_gen,
						cur->full_path);
				if (ret < 0)
					goto out;
			} else {
				ret = send_unlink(sctx, cur->full_path);
				if (ret < 0)
					goto out;
			}
		}

		/*
		 * link/move the ref to the new place. If we have an orphan
		 * inode, move it and update valid_path. If not, link or move
		 * it depending on the inode mode.
		 */
3456
		if (is_orphan) {
3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470
			ret = send_rename(sctx, valid_path, cur->full_path);
			if (ret < 0)
				goto out;
			is_orphan = 0;
			ret = fs_path_copy(valid_path, cur->full_path);
			if (ret < 0)
				goto out;
		} else {
			if (S_ISDIR(sctx->cur_inode_mode)) {
				/*
				 * Dirs can't be linked, so move it. For moved
				 * dirs, we always have one new and one deleted
				 * ref. The deleted ref is ignored later.
				 */
3471 3472 3473 3474
				ret = wait_for_parent_move(sctx, cur);
				if (ret < 0)
					goto out;
				if (ret) {
3475
					ret = add_pending_dir_move(sctx,
3476 3477 3478
							   sctx->cur_ino,
							   sctx->cur_inode_gen,
							   cur->dir);
3479 3480 3481 3482 3483 3484 3485 3486
					*pending_move = 1;
				} else {
					ret = send_rename(sctx, valid_path,
							  cur->full_path);
					if (!ret)
						ret = fs_path_copy(valid_path,
							       cur->full_path);
				}
3487 3488 3489 3490 3491 3492 3493 3494 3495
				if (ret < 0)
					goto out;
			} else {
				ret = send_link(sctx, cur->full_path,
						valid_path);
				if (ret < 0)
					goto out;
			}
		}
3496
		ret = dup_ref(cur, &check_dirs);
3497 3498 3499 3500 3501 3502 3503 3504 3505 3506 3507
		if (ret < 0)
			goto out;
	}

	if (S_ISDIR(sctx->cur_inode_mode) && sctx->cur_inode_deleted) {
		/*
		 * Check if we can already rmdir the directory. If not,
		 * orphanize it. For every dir item inside that gets deleted
		 * later, we do this check again and rmdir it then if possible.
		 * See the use of check_dirs for more details.
		 */
3508 3509
		ret = can_rmdir(sctx, sctx->cur_ino, sctx->cur_inode_gen,
				sctx->cur_ino);
3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524
		if (ret < 0)
			goto out;
		if (ret) {
			ret = send_rmdir(sctx, valid_path);
			if (ret < 0)
				goto out;
		} else if (!is_orphan) {
			ret = orphanize_inode(sctx, sctx->cur_ino,
					sctx->cur_inode_gen, valid_path);
			if (ret < 0)
				goto out;
			is_orphan = 1;
		}

		list_for_each_entry(cur, &sctx->deleted_refs, list) {
3525
			ret = dup_ref(cur, &check_dirs);
3526 3527 3528
			if (ret < 0)
				goto out;
		}
3529 3530 3531 3532 3533 3534 3535
	} else if (S_ISDIR(sctx->cur_inode_mode) &&
		   !list_empty(&sctx->deleted_refs)) {
		/*
		 * We have a moved dir. Add the old parent to check_dirs
		 */
		cur = list_entry(sctx->deleted_refs.next, struct recorded_ref,
				list);
3536
		ret = dup_ref(cur, &check_dirs);
3537 3538
		if (ret < 0)
			goto out;
3539 3540 3541 3542 3543 3544 3545 3546 3547 3548 3549 3550 3551
	} else if (!S_ISDIR(sctx->cur_inode_mode)) {
		/*
		 * We have a non dir inode. Go through all deleted refs and
		 * unlink them if they were not already overwritten by other
		 * inodes.
		 */
		list_for_each_entry(cur, &sctx->deleted_refs, list) {
			ret = did_overwrite_ref(sctx, cur->dir, cur->dir_gen,
					sctx->cur_ino, sctx->cur_inode_gen,
					cur->name, cur->name_len);
			if (ret < 0)
				goto out;
			if (!ret) {
3552 3553 3554
				ret = send_unlink(sctx, cur->full_path);
				if (ret < 0)
					goto out;
3555
			}
3556
			ret = dup_ref(cur, &check_dirs);
3557 3558 3559 3560 3561 3562 3563
			if (ret < 0)
				goto out;
		}
		/*
		 * If the inode is still orphan, unlink the orphan. This may
		 * happen when a previous inode did overwrite the first ref
		 * of this inode and no new refs were added for the current
3564 3565 3566
		 * inode. Unlinking does not mean that the inode is deleted in
		 * all cases. There may still be links to this inode in other
		 * places.
3567
		 */
3568
		if (is_orphan) {
3569 3570 3571 3572 3573 3574 3575 3576 3577 3578 3579 3580
			ret = send_unlink(sctx, valid_path);
			if (ret < 0)
				goto out;
		}
	}

	/*
	 * We did collect all parent dirs where cur_inode was once located. We
	 * now go through all these dirs and check if they are pending for
	 * deletion and if it's finally possible to perform the rmdir now.
	 * We also update the inode stats of the parent dirs here.
	 */
3581
	list_for_each_entry(cur, &check_dirs, list) {
3582 3583 3584 3585 3586
		/*
		 * In case we had refs into dirs that were not processed yet,
		 * we don't need to do the utime and rmdir logic for these dirs.
		 * The dir will be processed later.
		 */
3587
		if (cur->dir > sctx->cur_ino)
3588 3589
			continue;

3590
		ret = get_cur_inode_state(sctx, cur->dir, cur->dir_gen);
3591 3592 3593 3594 3595 3596
		if (ret < 0)
			goto out;

		if (ret == inode_state_did_create ||
		    ret == inode_state_no_change) {
			/* TODO delayed utimes */
3597
			ret = send_utimes(sctx, cur->dir, cur->dir_gen);
3598 3599
			if (ret < 0)
				goto out;
3600 3601
		} else if (ret == inode_state_did_delete &&
			   cur->dir != last_dir_ino_rm) {
3602 3603
			ret = can_rmdir(sctx, cur->dir, cur->dir_gen,
					sctx->cur_ino);
3604 3605 3606
			if (ret < 0)
				goto out;
			if (ret) {
3607 3608
				ret = get_cur_path(sctx, cur->dir,
						   cur->dir_gen, valid_path);
3609 3610 3611 3612 3613
				if (ret < 0)
					goto out;
				ret = send_rmdir(sctx, valid_path);
				if (ret < 0)
					goto out;
3614
				last_dir_ino_rm = cur->dir;
3615 3616 3617 3618 3619 3620 3621
			}
		}
	}

	ret = 0;

out:
3622
	__free_recorded_refs(&check_dirs);
3623
	free_recorded_refs(sctx);
3624
	fs_path_free(valid_path);
3625 3626 3627
	return ret;
}

3628 3629
static int record_ref(struct btrfs_root *root, int num, u64 dir, int index,
		      struct fs_path *name, void *ctx, struct list_head *refs)
3630 3631 3632 3633 3634 3635
{
	int ret = 0;
	struct send_ctx *sctx = ctx;
	struct fs_path *p;
	u64 gen;

3636
	p = fs_path_alloc();
3637 3638 3639
	if (!p)
		return -ENOMEM;

3640
	ret = get_inode_info(root, dir, NULL, &gen, NULL, NULL,
3641
			NULL, NULL);
3642 3643 3644 3645 3646 3647 3648 3649 3650 3651
	if (ret < 0)
		goto out;

	ret = get_cur_path(sctx, dir, gen, p);
	if (ret < 0)
		goto out;
	ret = fs_path_add_path(p, name);
	if (ret < 0)
		goto out;

3652
	ret = __record_ref(refs, dir, gen, p);
3653 3654 3655

out:
	if (ret)
3656
		fs_path_free(p);
3657 3658 3659
	return ret;
}

3660 3661 3662 3663 3664 3665 3666 3667 3668 3669
static int __record_new_ref(int num, u64 dir, int index,
			    struct fs_path *name,
			    void *ctx)
{
	struct send_ctx *sctx = ctx;
	return record_ref(sctx->send_root, num, dir, index, name,
			  ctx, &sctx->new_refs);
}


3670 3671 3672 3673 3674
static int __record_deleted_ref(int num, u64 dir, int index,
				struct fs_path *name,
				void *ctx)
{
	struct send_ctx *sctx = ctx;
3675 3676
	return record_ref(sctx->parent_root, num, dir, index, name,
			  ctx, &sctx->deleted_refs);
3677 3678 3679 3680 3681 3682
}

static int record_new_ref(struct send_ctx *sctx)
{
	int ret;

3683 3684
	ret = iterate_inode_ref(sctx->send_root, sctx->left_path,
				sctx->cmp_key, 0, __record_new_ref, sctx);
3685 3686 3687 3688 3689 3690 3691 3692 3693 3694 3695 3696
	if (ret < 0)
		goto out;
	ret = 0;

out:
	return ret;
}

static int record_deleted_ref(struct send_ctx *sctx)
{
	int ret;

3697 3698
	ret = iterate_inode_ref(sctx->parent_root, sctx->right_path,
				sctx->cmp_key, 0, __record_deleted_ref, sctx);
3699 3700 3701 3702 3703 3704 3705 3706 3707 3708
	if (ret < 0)
		goto out;
	ret = 0;

out:
	return ret;
}

struct find_ref_ctx {
	u64 dir;
3709 3710
	u64 dir_gen;
	struct btrfs_root *root;
3711 3712 3713 3714 3715 3716 3717 3718 3719
	struct fs_path *name;
	int found_idx;
};

static int __find_iref(int num, u64 dir, int index,
		       struct fs_path *name,
		       void *ctx_)
{
	struct find_ref_ctx *ctx = ctx_;
3720 3721
	u64 dir_gen;
	int ret;
3722 3723 3724

	if (dir == ctx->dir && fs_path_len(name) == fs_path_len(ctx->name) &&
	    strncmp(name->start, ctx->name->start, fs_path_len(name)) == 0) {
3725 3726 3727 3728 3729 3730 3731 3732 3733 3734
		/*
		 * To avoid doing extra lookups we'll only do this if everything
		 * else matches.
		 */
		ret = get_inode_info(ctx->root, dir, NULL, &dir_gen, NULL,
				     NULL, NULL, NULL);
		if (ret)
			return ret;
		if (dir_gen != ctx->dir_gen)
			return 0;
3735 3736 3737 3738 3739 3740
		ctx->found_idx = num;
		return 1;
	}
	return 0;
}

3741
static int find_iref(struct btrfs_root *root,
3742 3743
		     struct btrfs_path *path,
		     struct btrfs_key *key,
3744
		     u64 dir, u64 dir_gen, struct fs_path *name)
3745 3746 3747 3748 3749 3750
{
	int ret;
	struct find_ref_ctx ctx;

	ctx.dir = dir;
	ctx.name = name;
3751
	ctx.dir_gen = dir_gen;
3752
	ctx.found_idx = -1;
3753
	ctx.root = root;
3754

3755
	ret = iterate_inode_ref(root, path, key, 0, __find_iref, &ctx);
3756 3757 3758 3759 3760 3761 3762 3763 3764 3765 3766 3767 3768
	if (ret < 0)
		return ret;

	if (ctx.found_idx == -1)
		return -ENOENT;

	return ctx.found_idx;
}

static int __record_changed_new_ref(int num, u64 dir, int index,
				    struct fs_path *name,
				    void *ctx)
{
3769
	u64 dir_gen;
3770 3771 3772
	int ret;
	struct send_ctx *sctx = ctx;

3773 3774 3775 3776 3777
	ret = get_inode_info(sctx->send_root, dir, NULL, &dir_gen, NULL,
			     NULL, NULL, NULL);
	if (ret)
		return ret;

3778
	ret = find_iref(sctx->parent_root, sctx->right_path,
3779
			sctx->cmp_key, dir, dir_gen, name);
3780 3781 3782 3783 3784 3785 3786 3787 3788 3789 3790 3791
	if (ret == -ENOENT)
		ret = __record_new_ref(num, dir, index, name, sctx);
	else if (ret > 0)
		ret = 0;

	return ret;
}

static int __record_changed_deleted_ref(int num, u64 dir, int index,
					struct fs_path *name,
					void *ctx)
{
3792
	u64 dir_gen;
3793 3794 3795
	int ret;
	struct send_ctx *sctx = ctx;

3796 3797 3798 3799 3800
	ret = get_inode_info(sctx->parent_root, dir, NULL, &dir_gen, NULL,
			     NULL, NULL, NULL);
	if (ret)
		return ret;

3801
	ret = find_iref(sctx->send_root, sctx->left_path, sctx->cmp_key,
3802
			dir, dir_gen, name);
3803 3804 3805 3806 3807 3808 3809 3810 3811 3812 3813 3814
	if (ret == -ENOENT)
		ret = __record_deleted_ref(num, dir, index, name, sctx);
	else if (ret > 0)
		ret = 0;

	return ret;
}

static int record_changed_ref(struct send_ctx *sctx)
{
	int ret = 0;

3815
	ret = iterate_inode_ref(sctx->send_root, sctx->left_path,
3816 3817 3818
			sctx->cmp_key, 0, __record_changed_new_ref, sctx);
	if (ret < 0)
		goto out;
3819
	ret = iterate_inode_ref(sctx->parent_root, sctx->right_path,
3820 3821 3822 3823 3824 3825 3826 3827 3828 3829 3830 3831 3832 3833 3834 3835 3836 3837 3838 3839 3840 3841 3842 3843
			sctx->cmp_key, 0, __record_changed_deleted_ref, sctx);
	if (ret < 0)
		goto out;
	ret = 0;

out:
	return ret;
}

/*
 * Record and process all refs at once. Needed when an inode changes the
 * generation number, which means that it was deleted and recreated.
 */
static int process_all_refs(struct send_ctx *sctx,
			    enum btrfs_compare_tree_result cmd)
{
	int ret;
	struct btrfs_root *root;
	struct btrfs_path *path;
	struct btrfs_key key;
	struct btrfs_key found_key;
	struct extent_buffer *eb;
	int slot;
	iterate_inode_ref_t cb;
3844
	int pending_move = 0;
3845 3846 3847 3848 3849 3850 3851 3852 3853 3854 3855 3856

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

	if (cmd == BTRFS_COMPARE_TREE_NEW) {
		root = sctx->send_root;
		cb = __record_new_ref;
	} else if (cmd == BTRFS_COMPARE_TREE_DELETED) {
		root = sctx->parent_root;
		cb = __record_deleted_ref;
	} else {
3857 3858 3859 3860
		btrfs_err(sctx->send_root->fs_info,
				"Wrong command %d in process_all_refs", cmd);
		ret = -EINVAL;
		goto out;
3861 3862 3863 3864 3865
	}

	key.objectid = sctx->cmp_key->objectid;
	key.type = BTRFS_INODE_REF_KEY;
	key.offset = 0;
3866 3867 3868
	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
	if (ret < 0)
		goto out;
3869

3870
	while (1) {
3871 3872
		eb = path->nodes[0];
		slot = path->slots[0];
3873 3874 3875 3876 3877 3878 3879 3880 3881
		if (slot >= btrfs_header_nritems(eb)) {
			ret = btrfs_next_leaf(root, path);
			if (ret < 0)
				goto out;
			else if (ret > 0)
				break;
			continue;
		}

3882 3883 3884
		btrfs_item_key_to_cpu(eb, &found_key, slot);

		if (found_key.objectid != key.objectid ||
3885 3886
		    (found_key.type != BTRFS_INODE_REF_KEY &&
		     found_key.type != BTRFS_INODE_EXTREF_KEY))
3887 3888
			break;

3889
		ret = iterate_inode_ref(root, path, &found_key, 0, cb, sctx);
3890 3891 3892
		if (ret < 0)
			goto out;

3893
		path->slots[0]++;
3894
	}
3895
	btrfs_release_path(path);
3896

3897 3898 3899
	ret = process_recorded_refs(sctx, &pending_move);
	/* Only applicable to an incremental send. */
	ASSERT(pending_move == 0);
3900 3901 3902 3903 3904 3905 3906 3907 3908 3909 3910 3911 3912 3913 3914 3915 3916 3917 3918 3919 3920 3921 3922 3923 3924 3925 3926 3927 3928 3929 3930 3931 3932 3933 3934 3935 3936 3937 3938 3939 3940 3941 3942 3943 3944 3945 3946 3947 3948 3949 3950 3951 3952 3953 3954 3955 3956 3957

out:
	btrfs_free_path(path);
	return ret;
}

static int send_set_xattr(struct send_ctx *sctx,
			  struct fs_path *path,
			  const char *name, int name_len,
			  const char *data, int data_len)
{
	int ret = 0;

	ret = begin_cmd(sctx, BTRFS_SEND_C_SET_XATTR);
	if (ret < 0)
		goto out;

	TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, path);
	TLV_PUT_STRING(sctx, BTRFS_SEND_A_XATTR_NAME, name, name_len);
	TLV_PUT(sctx, BTRFS_SEND_A_XATTR_DATA, data, data_len);

	ret = send_cmd(sctx);

tlv_put_failure:
out:
	return ret;
}

static int send_remove_xattr(struct send_ctx *sctx,
			  struct fs_path *path,
			  const char *name, int name_len)
{
	int ret = 0;

	ret = begin_cmd(sctx, BTRFS_SEND_C_REMOVE_XATTR);
	if (ret < 0)
		goto out;

	TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, path);
	TLV_PUT_STRING(sctx, BTRFS_SEND_A_XATTR_NAME, name, name_len);

	ret = send_cmd(sctx);

tlv_put_failure:
out:
	return ret;
}

static int __process_new_xattr(int num, struct btrfs_key *di_key,
			       const char *name, int name_len,
			       const char *data, int data_len,
			       u8 type, void *ctx)
{
	int ret;
	struct send_ctx *sctx = ctx;
	struct fs_path *p;
	posix_acl_xattr_header dummy_acl;

3958
	p = fs_path_alloc();
3959 3960 3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975 3976 3977 3978 3979 3980 3981 3982 3983 3984
	if (!p)
		return -ENOMEM;

	/*
	 * This hack is needed because empty acl's are stored as zero byte
	 * data in xattrs. Problem with that is, that receiving these zero byte
	 * acl's will fail later. To fix this, we send a dummy acl list that
	 * only contains the version number and no entries.
	 */
	if (!strncmp(name, XATTR_NAME_POSIX_ACL_ACCESS, name_len) ||
	    !strncmp(name, XATTR_NAME_POSIX_ACL_DEFAULT, name_len)) {
		if (data_len == 0) {
			dummy_acl.a_version =
					cpu_to_le32(POSIX_ACL_XATTR_VERSION);
			data = (char *)&dummy_acl;
			data_len = sizeof(dummy_acl);
		}
	}

	ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen, p);
	if (ret < 0)
		goto out;

	ret = send_set_xattr(sctx, p, name, name_len, data, data_len);

out:
3985
	fs_path_free(p);
3986 3987 3988 3989 3990 3991 3992 3993 3994 3995 3996 3997
	return ret;
}

static int __process_deleted_xattr(int num, struct btrfs_key *di_key,
				   const char *name, int name_len,
				   const char *data, int data_len,
				   u8 type, void *ctx)
{
	int ret;
	struct send_ctx *sctx = ctx;
	struct fs_path *p;

3998
	p = fs_path_alloc();
3999 4000 4001 4002 4003 4004 4005 4006 4007 4008
	if (!p)
		return -ENOMEM;

	ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen, p);
	if (ret < 0)
		goto out;

	ret = send_remove_xattr(sctx, p, name, name_len);

out:
4009
	fs_path_free(p);
4010 4011 4012 4013 4014 4015 4016
	return ret;
}

static int process_new_xattr(struct send_ctx *sctx)
{
	int ret = 0;

4017 4018
	ret = iterate_dir_item(sctx->send_root, sctx->left_path,
			       sctx->cmp_key, __process_new_xattr, sctx);
4019 4020 4021 4022 4023 4024 4025 4026

	return ret;
}

static int process_deleted_xattr(struct send_ctx *sctx)
{
	int ret;

4027 4028
	ret = iterate_dir_item(sctx->parent_root, sctx->right_path,
			       sctx->cmp_key, __process_deleted_xattr, sctx);
4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044 4045 4046 4047 4048 4049 4050 4051

	return ret;
}

struct find_xattr_ctx {
	const char *name;
	int name_len;
	int found_idx;
	char *found_data;
	int found_data_len;
};

static int __find_xattr(int num, struct btrfs_key *di_key,
			const char *name, int name_len,
			const char *data, int data_len,
			u8 type, void *vctx)
{
	struct find_xattr_ctx *ctx = vctx;

	if (name_len == ctx->name_len &&
	    strncmp(name, ctx->name, name_len) == 0) {
		ctx->found_idx = num;
		ctx->found_data_len = data_len;
4052
		ctx->found_data = kmemdup(data, data_len, GFP_NOFS);
4053 4054 4055 4056 4057 4058 4059
		if (!ctx->found_data)
			return -ENOMEM;
		return 1;
	}
	return 0;
}

4060
static int find_xattr(struct btrfs_root *root,
4061 4062 4063 4064 4065 4066 4067 4068 4069 4070 4071 4072 4073 4074
		      struct btrfs_path *path,
		      struct btrfs_key *key,
		      const char *name, int name_len,
		      char **data, int *data_len)
{
	int ret;
	struct find_xattr_ctx ctx;

	ctx.name = name;
	ctx.name_len = name_len;
	ctx.found_idx = -1;
	ctx.found_data = NULL;
	ctx.found_data_len = 0;

4075
	ret = iterate_dir_item(root, path, key, __find_xattr, &ctx);
4076 4077 4078 4079 4080 4081 4082 4083 4084 4085 4086 4087 4088 4089 4090 4091 4092 4093 4094 4095 4096 4097 4098 4099 4100
	if (ret < 0)
		return ret;

	if (ctx.found_idx == -1)
		return -ENOENT;
	if (data) {
		*data = ctx.found_data;
		*data_len = ctx.found_data_len;
	} else {
		kfree(ctx.found_data);
	}
	return ctx.found_idx;
}


static int __process_changed_new_xattr(int num, struct btrfs_key *di_key,
				       const char *name, int name_len,
				       const char *data, int data_len,
				       u8 type, void *ctx)
{
	int ret;
	struct send_ctx *sctx = ctx;
	char *found_data = NULL;
	int found_data_len  = 0;

4101 4102 4103
	ret = find_xattr(sctx->parent_root, sctx->right_path,
			 sctx->cmp_key, name, name_len, &found_data,
			 &found_data_len);
4104 4105 4106 4107 4108 4109 4110 4111 4112 4113 4114 4115 4116 4117 4118 4119 4120 4121 4122 4123 4124 4125 4126 4127 4128
	if (ret == -ENOENT) {
		ret = __process_new_xattr(num, di_key, name, name_len, data,
				data_len, type, ctx);
	} else if (ret >= 0) {
		if (data_len != found_data_len ||
		    memcmp(data, found_data, data_len)) {
			ret = __process_new_xattr(num, di_key, name, name_len,
					data, data_len, type, ctx);
		} else {
			ret = 0;
		}
	}

	kfree(found_data);
	return ret;
}

static int __process_changed_deleted_xattr(int num, struct btrfs_key *di_key,
					   const char *name, int name_len,
					   const char *data, int data_len,
					   u8 type, void *ctx)
{
	int ret;
	struct send_ctx *sctx = ctx;

4129 4130
	ret = find_xattr(sctx->send_root, sctx->left_path, sctx->cmp_key,
			 name, name_len, NULL, NULL);
4131 4132 4133 4134 4135 4136 4137 4138 4139 4140 4141 4142 4143
	if (ret == -ENOENT)
		ret = __process_deleted_xattr(num, di_key, name, name_len, data,
				data_len, type, ctx);
	else if (ret >= 0)
		ret = 0;

	return ret;
}

static int process_changed_xattr(struct send_ctx *sctx)
{
	int ret = 0;

4144
	ret = iterate_dir_item(sctx->send_root, sctx->left_path,
4145 4146 4147
			sctx->cmp_key, __process_changed_new_xattr, sctx);
	if (ret < 0)
		goto out;
4148
	ret = iterate_dir_item(sctx->parent_root, sctx->right_path,
4149 4150 4151 4152 4153 4154 4155 4156 4157 4158 4159 4160 4161 4162 4163 4164 4165 4166 4167 4168 4169 4170 4171 4172 4173
			sctx->cmp_key, __process_changed_deleted_xattr, sctx);

out:
	return ret;
}

static int process_all_new_xattrs(struct send_ctx *sctx)
{
	int ret;
	struct btrfs_root *root;
	struct btrfs_path *path;
	struct btrfs_key key;
	struct btrfs_key found_key;
	struct extent_buffer *eb;
	int slot;

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

	root = sctx->send_root;

	key.objectid = sctx->cmp_key->objectid;
	key.type = BTRFS_XATTR_ITEM_KEY;
	key.offset = 0;
4174 4175 4176
	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
	if (ret < 0)
		goto out;
4177

4178
	while (1) {
4179 4180
		eb = path->nodes[0];
		slot = path->slots[0];
4181 4182 4183 4184 4185 4186 4187 4188 4189 4190
		if (slot >= btrfs_header_nritems(eb)) {
			ret = btrfs_next_leaf(root, path);
			if (ret < 0) {
				goto out;
			} else if (ret > 0) {
				ret = 0;
				break;
			}
			continue;
		}
4191

4192
		btrfs_item_key_to_cpu(eb, &found_key, slot);
4193 4194 4195 4196 4197 4198
		if (found_key.objectid != key.objectid ||
		    found_key.type != key.type) {
			ret = 0;
			goto out;
		}

4199 4200
		ret = iterate_dir_item(root, path, &found_key,
				       __process_new_xattr, sctx);
4201 4202 4203
		if (ret < 0)
			goto out;

4204
		path->slots[0]++;
4205 4206 4207 4208 4209 4210 4211
	}

out:
	btrfs_free_path(path);
	return ret;
}

J
Josef Bacik 已提交
4212 4213 4214 4215 4216 4217 4218 4219 4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232 4233 4234 4235 4236 4237 4238 4239 4240 4241 4242
static ssize_t fill_read_buf(struct send_ctx *sctx, u64 offset, u32 len)
{
	struct btrfs_root *root = sctx->send_root;
	struct btrfs_fs_info *fs_info = root->fs_info;
	struct inode *inode;
	struct page *page;
	char *addr;
	struct btrfs_key key;
	pgoff_t index = offset >> PAGE_CACHE_SHIFT;
	pgoff_t last_index;
	unsigned pg_offset = offset & ~PAGE_CACHE_MASK;
	ssize_t ret = 0;

	key.objectid = sctx->cur_ino;
	key.type = BTRFS_INODE_ITEM_KEY;
	key.offset = 0;

	inode = btrfs_iget(fs_info->sb, &key, root, NULL);
	if (IS_ERR(inode))
		return PTR_ERR(inode);

	if (offset + len > i_size_read(inode)) {
		if (offset > i_size_read(inode))
			len = 0;
		else
			len = offset - i_size_read(inode);
	}
	if (len == 0)
		goto out;

	last_index = (offset + len - 1) >> PAGE_CACHE_SHIFT;
L
Liu Bo 已提交
4243 4244 4245 4246 4247 4248 4249

	/* initial readahead */
	memset(&sctx->ra, 0, sizeof(struct file_ra_state));
	file_ra_state_init(&sctx->ra, inode->i_mapping);
	btrfs_force_ra(inode->i_mapping, &sctx->ra, NULL, index,
		       last_index - index + 1);

J
Josef Bacik 已提交
4250 4251 4252 4253 4254 4255 4256 4257 4258 4259 4260 4261 4262 4263 4264 4265 4266 4267 4268 4269 4270 4271 4272 4273 4274 4275 4276 4277 4278 4279 4280 4281 4282 4283 4284
	while (index <= last_index) {
		unsigned cur_len = min_t(unsigned, len,
					 PAGE_CACHE_SIZE - pg_offset);
		page = find_or_create_page(inode->i_mapping, index, GFP_NOFS);
		if (!page) {
			ret = -ENOMEM;
			break;
		}

		if (!PageUptodate(page)) {
			btrfs_readpage(NULL, page);
			lock_page(page);
			if (!PageUptodate(page)) {
				unlock_page(page);
				page_cache_release(page);
				ret = -EIO;
				break;
			}
		}

		addr = kmap(page);
		memcpy(sctx->read_buf + ret, addr + pg_offset, cur_len);
		kunmap(page);
		unlock_page(page);
		page_cache_release(page);
		index++;
		pg_offset = 0;
		len -= cur_len;
		ret += cur_len;
	}
out:
	iput(inode);
	return ret;
}

4285 4286 4287 4288 4289 4290 4291 4292
/*
 * Read some bytes from the current inode/file and send a write command to
 * user space.
 */
static int send_write(struct send_ctx *sctx, u64 offset, u32 len)
{
	int ret = 0;
	struct fs_path *p;
J
Josef Bacik 已提交
4293
	ssize_t num_read = 0;
4294

4295
	p = fs_path_alloc();
4296 4297 4298 4299 4300
	if (!p)
		return -ENOMEM;

verbose_printk("btrfs: send_write offset=%llu, len=%d\n", offset, len);

J
Josef Bacik 已提交
4301 4302 4303 4304
	num_read = fill_read_buf(sctx, offset, len);
	if (num_read <= 0) {
		if (num_read < 0)
			ret = num_read;
4305
		goto out;
J
Josef Bacik 已提交
4306
	}
4307 4308 4309 4310 4311 4312 4313 4314 4315 4316 4317

	ret = begin_cmd(sctx, BTRFS_SEND_C_WRITE);
	if (ret < 0)
		goto out;

	ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen, p);
	if (ret < 0)
		goto out;

	TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p);
	TLV_PUT_U64(sctx, BTRFS_SEND_A_FILE_OFFSET, offset);
4318
	TLV_PUT(sctx, BTRFS_SEND_A_DATA, sctx->read_buf, num_read);
4319 4320 4321 4322 4323

	ret = send_cmd(sctx);

tlv_put_failure:
out:
4324
	fs_path_free(p);
4325 4326
	if (ret < 0)
		return ret;
4327
	return num_read;
4328 4329 4330 4331 4332 4333 4334 4335 4336 4337 4338 4339 4340 4341 4342 4343 4344 4345
}

/*
 * Send a clone command to user space.
 */
static int send_clone(struct send_ctx *sctx,
		      u64 offset, u32 len,
		      struct clone_root *clone_root)
{
	int ret = 0;
	struct fs_path *p;
	u64 gen;

verbose_printk("btrfs: send_clone offset=%llu, len=%d, clone_root=%llu, "
	       "clone_inode=%llu, clone_offset=%llu\n", offset, len,
		clone_root->root->objectid, clone_root->ino,
		clone_root->offset);

4346
	p = fs_path_alloc();
4347 4348 4349 4350 4351 4352 4353 4354 4355 4356 4357 4358 4359 4360 4361
	if (!p)
		return -ENOMEM;

	ret = begin_cmd(sctx, BTRFS_SEND_C_CLONE);
	if (ret < 0)
		goto out;

	ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen, p);
	if (ret < 0)
		goto out;

	TLV_PUT_U64(sctx, BTRFS_SEND_A_FILE_OFFSET, offset);
	TLV_PUT_U64(sctx, BTRFS_SEND_A_CLONE_LEN, len);
	TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p);

4362
	if (clone_root->root == sctx->send_root) {
4363
		ret = get_inode_info(sctx->send_root, clone_root->ino, NULL,
4364
				&gen, NULL, NULL, NULL, NULL);
4365 4366 4367 4368
		if (ret < 0)
			goto out;
		ret = get_cur_path(sctx, clone_root->ino, gen, p);
	} else {
4369
		ret = get_inode_path(clone_root->root, clone_root->ino, p);
4370 4371 4372 4373 4374
	}
	if (ret < 0)
		goto out;

	TLV_PUT_UUID(sctx, BTRFS_SEND_A_CLONE_UUID,
4375
			clone_root->root->root_item.uuid);
4376
	TLV_PUT_U64(sctx, BTRFS_SEND_A_CLONE_CTRANSID,
4377
		    le64_to_cpu(clone_root->root->root_item.ctransid));
4378 4379 4380 4381 4382 4383 4384 4385
	TLV_PUT_PATH(sctx, BTRFS_SEND_A_CLONE_PATH, p);
	TLV_PUT_U64(sctx, BTRFS_SEND_A_CLONE_OFFSET,
			clone_root->offset);

	ret = send_cmd(sctx);

tlv_put_failure:
out:
4386
	fs_path_free(p);
4387 4388 4389
	return ret;
}

4390 4391 4392 4393 4394 4395 4396 4397 4398
/*
 * Send an update extent command to user space.
 */
static int send_update_extent(struct send_ctx *sctx,
			      u64 offset, u32 len)
{
	int ret = 0;
	struct fs_path *p;

4399
	p = fs_path_alloc();
4400 4401 4402 4403 4404 4405 4406 4407 4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418
	if (!p)
		return -ENOMEM;

	ret = begin_cmd(sctx, BTRFS_SEND_C_UPDATE_EXTENT);
	if (ret < 0)
		goto out;

	ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen, p);
	if (ret < 0)
		goto out;

	TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p);
	TLV_PUT_U64(sctx, BTRFS_SEND_A_FILE_OFFSET, offset);
	TLV_PUT_U64(sctx, BTRFS_SEND_A_SIZE, len);

	ret = send_cmd(sctx);

tlv_put_failure:
out:
4419
	fs_path_free(p);
4420 4421 4422
	return ret;
}

4423 4424 4425 4426 4427 4428 4429 4430 4431 4432
static int send_hole(struct send_ctx *sctx, u64 end)
{
	struct fs_path *p = NULL;
	u64 offset = sctx->cur_inode_last_extent;
	u64 len;
	int ret = 0;

	p = fs_path_alloc();
	if (!p)
		return -ENOMEM;
4433 4434 4435
	ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen, p);
	if (ret < 0)
		goto tlv_put_failure;
4436 4437 4438 4439 4440 4441 4442 4443 4444 4445 4446 4447 4448 4449 4450 4451 4452 4453 4454 4455
	memset(sctx->read_buf, 0, BTRFS_SEND_READ_SIZE);
	while (offset < end) {
		len = min_t(u64, end - offset, BTRFS_SEND_READ_SIZE);

		ret = begin_cmd(sctx, BTRFS_SEND_C_WRITE);
		if (ret < 0)
			break;
		TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p);
		TLV_PUT_U64(sctx, BTRFS_SEND_A_FILE_OFFSET, offset);
		TLV_PUT(sctx, BTRFS_SEND_A_DATA, sctx->read_buf, len);
		ret = send_cmd(sctx);
		if (ret < 0)
			break;
		offset += len;
	}
tlv_put_failure:
	fs_path_free(p);
	return ret;
}

4456 4457 4458 4459 4460 4461 4462 4463 4464 4465 4466 4467
static int send_write_or_clone(struct send_ctx *sctx,
			       struct btrfs_path *path,
			       struct btrfs_key *key,
			       struct clone_root *clone_root)
{
	int ret = 0;
	struct btrfs_file_extent_item *ei;
	u64 offset = key->offset;
	u64 pos = 0;
	u64 len;
	u32 l;
	u8 type;
4468
	u64 bs = sctx->send_root->fs_info->sb->s_blocksize;
4469 4470 4471 4472

	ei = btrfs_item_ptr(path->nodes[0], path->slots[0],
			struct btrfs_file_extent_item);
	type = btrfs_file_extent_type(path->nodes[0], ei);
4473
	if (type == BTRFS_FILE_EXTENT_INLINE) {
4474 4475
		len = btrfs_file_extent_inline_len(path->nodes[0],
						   path->slots[0], ei);
4476 4477 4478 4479 4480 4481 4482
		/*
		 * it is possible the inline item won't cover the whole page,
		 * but there may be items after this page.  Make
		 * sure to send the whole thing
		 */
		len = PAGE_CACHE_ALIGN(len);
	} else {
4483
		len = btrfs_file_extent_num_bytes(path->nodes[0], ei);
4484
	}
4485 4486 4487 4488 4489 4490 4491 4492

	if (offset + len > sctx->cur_inode_size)
		len = sctx->cur_inode_size - offset;
	if (len == 0) {
		ret = 0;
		goto out;
	}

4493
	if (clone_root && IS_ALIGNED(offset + len, bs)) {
4494 4495 4496 4497
		ret = send_clone(sctx, offset, len, clone_root);
	} else if (sctx->flags & BTRFS_SEND_FLAG_NO_FILE_DATA) {
		ret = send_update_extent(sctx, offset, len);
	} else {
4498 4499 4500 4501 4502 4503 4504 4505 4506 4507 4508 4509 4510 4511 4512 4513 4514 4515 4516 4517 4518 4519 4520 4521 4522 4523 4524 4525 4526 4527 4528 4529 4530 4531 4532
		while (pos < len) {
			l = len - pos;
			if (l > BTRFS_SEND_READ_SIZE)
				l = BTRFS_SEND_READ_SIZE;
			ret = send_write(sctx, pos + offset, l);
			if (ret < 0)
				goto out;
			if (!ret)
				break;
			pos += ret;
		}
		ret = 0;
	}
out:
	return ret;
}

static int is_extent_unchanged(struct send_ctx *sctx,
			       struct btrfs_path *left_path,
			       struct btrfs_key *ekey)
{
	int ret = 0;
	struct btrfs_key key;
	struct btrfs_path *path = NULL;
	struct extent_buffer *eb;
	int slot;
	struct btrfs_key found_key;
	struct btrfs_file_extent_item *ei;
	u64 left_disknr;
	u64 right_disknr;
	u64 left_offset;
	u64 right_offset;
	u64 left_offset_fixed;
	u64 left_len;
	u64 right_len;
4533 4534
	u64 left_gen;
	u64 right_gen;
4535 4536 4537 4538 4539 4540 4541 4542 4543 4544 4545 4546 4547 4548 4549 4550
	u8 left_type;
	u8 right_type;

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

	eb = left_path->nodes[0];
	slot = left_path->slots[0];
	ei = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item);
	left_type = btrfs_file_extent_type(eb, ei);

	if (left_type != BTRFS_FILE_EXTENT_REG) {
		ret = 0;
		goto out;
	}
4551 4552 4553 4554
	left_disknr = btrfs_file_extent_disk_bytenr(eb, ei);
	left_len = btrfs_file_extent_num_bytes(eb, ei);
	left_offset = btrfs_file_extent_offset(eb, ei);
	left_gen = btrfs_file_extent_generation(eb, ei);
4555 4556 4557 4558 4559 4560 4561 4562 4563 4564 4565 4566 4567 4568 4569 4570 4571 4572 4573 4574 4575 4576 4577 4578 4579 4580 4581 4582 4583 4584 4585 4586 4587 4588 4589 4590 4591 4592 4593 4594 4595

	/*
	 * Following comments will refer to these graphics. L is the left
	 * extents which we are checking at the moment. 1-8 are the right
	 * extents that we iterate.
	 *
	 *       |-----L-----|
	 * |-1-|-2a-|-3-|-4-|-5-|-6-|
	 *
	 *       |-----L-----|
	 * |--1--|-2b-|...(same as above)
	 *
	 * Alternative situation. Happens on files where extents got split.
	 *       |-----L-----|
	 * |-----------7-----------|-6-|
	 *
	 * Alternative situation. Happens on files which got larger.
	 *       |-----L-----|
	 * |-8-|
	 * Nothing follows after 8.
	 */

	key.objectid = ekey->objectid;
	key.type = BTRFS_EXTENT_DATA_KEY;
	key.offset = ekey->offset;
	ret = btrfs_search_slot_for_read(sctx->parent_root, &key, path, 0, 0);
	if (ret < 0)
		goto out;
	if (ret) {
		ret = 0;
		goto out;
	}

	/*
	 * Handle special case where the right side has no extents at all.
	 */
	eb = path->nodes[0];
	slot = path->slots[0];
	btrfs_item_key_to_cpu(eb, &found_key, slot);
	if (found_key.objectid != key.objectid ||
	    found_key.type != key.type) {
4596 4597
		/* If we're a hole then just pretend nothing changed */
		ret = (left_disknr) ? 0 : 1;
4598 4599 4600 4601 4602 4603 4604 4605 4606 4607 4608 4609 4610 4611 4612
		goto out;
	}

	/*
	 * We're now on 2a, 2b or 7.
	 */
	key = found_key;
	while (key.offset < ekey->offset + left_len) {
		ei = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item);
		right_type = btrfs_file_extent_type(eb, ei);
		if (right_type != BTRFS_FILE_EXTENT_REG) {
			ret = 0;
			goto out;
		}

4613 4614 4615 4616 4617
		right_disknr = btrfs_file_extent_disk_bytenr(eb, ei);
		right_len = btrfs_file_extent_num_bytes(eb, ei);
		right_offset = btrfs_file_extent_offset(eb, ei);
		right_gen = btrfs_file_extent_generation(eb, ei);

4618 4619 4620 4621
		/*
		 * Are we at extent 8? If yes, we know the extent is changed.
		 * This may only happen on the first iteration.
		 */
4622
		if (found_key.offset + right_len <= ekey->offset) {
4623 4624
			/* If we're a hole just pretend nothing changed */
			ret = (left_disknr) ? 0 : 1;
4625 4626 4627 4628 4629 4630 4631 4632 4633 4634 4635 4636 4637 4638 4639
			goto out;
		}

		left_offset_fixed = left_offset;
		if (key.offset < ekey->offset) {
			/* Fix the right offset for 2a and 7. */
			right_offset += ekey->offset - key.offset;
		} else {
			/* Fix the left offset for all behind 2a and 2b */
			left_offset_fixed += key.offset - ekey->offset;
		}

		/*
		 * Check if we have the same extent.
		 */
4640
		if (left_disknr != right_disknr ||
4641 4642
		    left_offset_fixed != right_offset ||
		    left_gen != right_gen) {
4643 4644 4645 4646 4647 4648 4649 4650 4651 4652 4653 4654 4655 4656 4657 4658 4659 4660 4661
			ret = 0;
			goto out;
		}

		/*
		 * Go to the next extent.
		 */
		ret = btrfs_next_item(sctx->parent_root, path);
		if (ret < 0)
			goto out;
		if (!ret) {
			eb = path->nodes[0];
			slot = path->slots[0];
			btrfs_item_key_to_cpu(eb, &found_key, slot);
		}
		if (ret || found_key.objectid != key.objectid ||
		    found_key.type != key.type) {
			key.offset += right_len;
			break;
4662 4663 4664 4665
		}
		if (found_key.offset != key.offset + right_len) {
			ret = 0;
			goto out;
4666 4667 4668 4669 4670 4671 4672 4673 4674 4675 4676 4677 4678 4679 4680 4681 4682 4683 4684
		}
		key = found_key;
	}

	/*
	 * We're now behind the left extent (treat as unchanged) or at the end
	 * of the right side (treat as changed).
	 */
	if (key.offset >= ekey->offset + left_len)
		ret = 1;
	else
		ret = 0;


out:
	btrfs_free_path(path);
	return ret;
}

4685 4686 4687 4688 4689 4690 4691 4692 4693 4694 4695 4696 4697 4698 4699 4700 4701 4702 4703 4704 4705 4706 4707 4708 4709 4710 4711 4712 4713 4714 4715
static int get_last_extent(struct send_ctx *sctx, u64 offset)
{
	struct btrfs_path *path;
	struct btrfs_root *root = sctx->send_root;
	struct btrfs_file_extent_item *fi;
	struct btrfs_key key;
	u64 extent_end;
	u8 type;
	int ret;

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

	sctx->cur_inode_last_extent = 0;

	key.objectid = sctx->cur_ino;
	key.type = BTRFS_EXTENT_DATA_KEY;
	key.offset = offset;
	ret = btrfs_search_slot_for_read(root, &key, path, 0, 1);
	if (ret < 0)
		goto out;
	ret = 0;
	btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
	if (key.objectid != sctx->cur_ino || key.type != BTRFS_EXTENT_DATA_KEY)
		goto out;

	fi = btrfs_item_ptr(path->nodes[0], path->slots[0],
			    struct btrfs_file_extent_item);
	type = btrfs_file_extent_type(path->nodes[0], fi);
	if (type == BTRFS_FILE_EXTENT_INLINE) {
4716 4717
		u64 size = btrfs_file_extent_inline_len(path->nodes[0],
							path->slots[0], fi);
4718 4719 4720 4721 4722 4723 4724 4725 4726 4727 4728 4729 4730 4731 4732 4733 4734 4735 4736 4737 4738 4739 4740 4741 4742 4743 4744 4745 4746 4747 4748 4749 4750
		extent_end = ALIGN(key.offset + size,
				   sctx->send_root->sectorsize);
	} else {
		extent_end = key.offset +
			btrfs_file_extent_num_bytes(path->nodes[0], fi);
	}
	sctx->cur_inode_last_extent = extent_end;
out:
	btrfs_free_path(path);
	return ret;
}

static int maybe_send_hole(struct send_ctx *sctx, struct btrfs_path *path,
			   struct btrfs_key *key)
{
	struct btrfs_file_extent_item *fi;
	u64 extent_end;
	u8 type;
	int ret = 0;

	if (sctx->cur_ino != key->objectid || !need_send_hole(sctx))
		return 0;

	if (sctx->cur_inode_last_extent == (u64)-1) {
		ret = get_last_extent(sctx, key->offset - 1);
		if (ret)
			return ret;
	}

	fi = btrfs_item_ptr(path->nodes[0], path->slots[0],
			    struct btrfs_file_extent_item);
	type = btrfs_file_extent_type(path->nodes[0], fi);
	if (type == BTRFS_FILE_EXTENT_INLINE) {
4751 4752
		u64 size = btrfs_file_extent_inline_len(path->nodes[0],
							path->slots[0], fi);
4753 4754 4755 4756 4757 4758
		extent_end = ALIGN(key->offset + size,
				   sctx->send_root->sectorsize);
	} else {
		extent_end = key->offset +
			btrfs_file_extent_num_bytes(path->nodes[0], fi);
	}
4759 4760 4761 4762 4763 4764 4765 4766 4767 4768 4769 4770 4771 4772 4773

	if (path->slots[0] == 0 &&
	    sctx->cur_inode_last_extent < key->offset) {
		/*
		 * We might have skipped entire leafs that contained only
		 * file extent items for our current inode. These leafs have
		 * a generation number smaller (older) than the one in the
		 * current leaf and the leaf our last extent came from, and
		 * are located between these 2 leafs.
		 */
		ret = get_last_extent(sctx, key->offset - 1);
		if (ret)
			return ret;
	}

4774 4775 4776 4777 4778 4779
	if (sctx->cur_inode_last_extent < key->offset)
		ret = send_hole(sctx, key->offset);
	sctx->cur_inode_last_extent = extent_end;
	return ret;
}

4780 4781 4782 4783 4784
static int process_extent(struct send_ctx *sctx,
			  struct btrfs_path *path,
			  struct btrfs_key *key)
{
	struct clone_root *found_clone = NULL;
4785
	int ret = 0;
4786 4787 4788 4789 4790 4791 4792 4793 4794 4795

	if (S_ISLNK(sctx->cur_inode_mode))
		return 0;

	if (sctx->parent_root && !sctx->cur_inode_new) {
		ret = is_extent_unchanged(sctx, path, key);
		if (ret < 0)
			goto out;
		if (ret) {
			ret = 0;
4796
			goto out_hole;
4797
		}
4798 4799 4800 4801 4802 4803 4804 4805 4806 4807 4808 4809 4810 4811 4812 4813 4814 4815 4816 4817 4818 4819 4820 4821 4822 4823
	} else {
		struct btrfs_file_extent_item *ei;
		u8 type;

		ei = btrfs_item_ptr(path->nodes[0], path->slots[0],
				    struct btrfs_file_extent_item);
		type = btrfs_file_extent_type(path->nodes[0], ei);
		if (type == BTRFS_FILE_EXTENT_PREALLOC ||
		    type == BTRFS_FILE_EXTENT_REG) {
			/*
			 * The send spec does not have a prealloc command yet,
			 * so just leave a hole for prealloc'ed extents until
			 * we have enough commands queued up to justify rev'ing
			 * the send spec.
			 */
			if (type == BTRFS_FILE_EXTENT_PREALLOC) {
				ret = 0;
				goto out;
			}

			/* Have a hole, just skip it. */
			if (btrfs_file_extent_disk_bytenr(path->nodes[0], ei) == 0) {
				ret = 0;
				goto out;
			}
		}
4824 4825 4826 4827 4828 4829 4830 4831
	}

	ret = find_extent_clone(sctx, path, key->objectid, key->offset,
			sctx->cur_inode_size, &found_clone);
	if (ret != -ENOENT && ret < 0)
		goto out;

	ret = send_write_or_clone(sctx, path, key, found_clone);
4832 4833 4834 4835
	if (ret)
		goto out;
out_hole:
	ret = maybe_send_hole(sctx, path, key);
4836 4837 4838 4839 4840 4841 4842 4843 4844 4845 4846 4847 4848 4849 4850 4851 4852 4853 4854 4855 4856 4857
out:
	return ret;
}

static int process_all_extents(struct send_ctx *sctx)
{
	int ret;
	struct btrfs_root *root;
	struct btrfs_path *path;
	struct btrfs_key key;
	struct btrfs_key found_key;
	struct extent_buffer *eb;
	int slot;

	root = sctx->send_root;
	path = alloc_path_for_send();
	if (!path)
		return -ENOMEM;

	key.objectid = sctx->cmp_key->objectid;
	key.type = BTRFS_EXTENT_DATA_KEY;
	key.offset = 0;
4858 4859 4860
	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
	if (ret < 0)
		goto out;
4861

4862
	while (1) {
4863 4864
		eb = path->nodes[0];
		slot = path->slots[0];
4865 4866 4867 4868 4869 4870 4871 4872 4873 4874 4875 4876

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

4877 4878 4879 4880 4881 4882 4883 4884 4885 4886 4887 4888
		btrfs_item_key_to_cpu(eb, &found_key, slot);

		if (found_key.objectid != key.objectid ||
		    found_key.type != key.type) {
			ret = 0;
			goto out;
		}

		ret = process_extent(sctx, path, &found_key);
		if (ret < 0)
			goto out;

4889
		path->slots[0]++;
4890 4891 4892 4893 4894 4895 4896
	}

out:
	btrfs_free_path(path);
	return ret;
}

4897 4898 4899
static int process_recorded_refs_if_needed(struct send_ctx *sctx, int at_end,
					   int *pending_move,
					   int *refs_processed)
4900 4901 4902 4903 4904 4905
{
	int ret = 0;

	if (sctx->cur_ino == 0)
		goto out;
	if (!at_end && sctx->cur_ino == sctx->cmp_key->objectid &&
4906
	    sctx->cmp_key->type <= BTRFS_INODE_EXTREF_KEY)
4907 4908 4909 4910
		goto out;
	if (list_empty(&sctx->new_refs) && list_empty(&sctx->deleted_refs))
		goto out;

4911
	ret = process_recorded_refs(sctx, pending_move);
4912 4913 4914
	if (ret < 0)
		goto out;

4915
	*refs_processed = 1;
4916 4917 4918 4919 4920 4921 4922 4923 4924 4925 4926 4927 4928 4929 4930
out:
	return ret;
}

static int finish_inode_if_needed(struct send_ctx *sctx, int at_end)
{
	int ret = 0;
	u64 left_mode;
	u64 left_uid;
	u64 left_gid;
	u64 right_mode;
	u64 right_uid;
	u64 right_gid;
	int need_chmod = 0;
	int need_chown = 0;
4931 4932
	int pending_move = 0;
	int refs_processed = 0;
4933

4934 4935
	ret = process_recorded_refs_if_needed(sctx, at_end, &pending_move,
					      &refs_processed);
4936 4937 4938
	if (ret < 0)
		goto out;

4939 4940 4941 4942 4943 4944 4945 4946 4947 4948 4949 4950 4951 4952 4953
	/*
	 * We have processed the refs and thus need to advance send_progress.
	 * Now, calls to get_cur_xxx will take the updated refs of the current
	 * inode into account.
	 *
	 * On the other hand, if our current inode is a directory and couldn't
	 * be moved/renamed because its parent was renamed/moved too and it has
	 * a higher inode number, we can only move/rename our current inode
	 * after we moved/renamed its parent. Therefore in this case operate on
	 * the old path (pre move/rename) of our current inode, and the
	 * move/rename will be performed later.
	 */
	if (refs_processed && !pending_move)
		sctx->send_progress = sctx->cur_ino + 1;

4954 4955 4956 4957 4958 4959
	if (sctx->cur_ino == 0 || sctx->cur_inode_deleted)
		goto out;
	if (!at_end && sctx->cmp_key->objectid == sctx->cur_ino)
		goto out;

	ret = get_inode_info(sctx->send_root, sctx->cur_ino, NULL, NULL,
4960
			&left_mode, &left_uid, &left_gid, NULL);
4961 4962 4963
	if (ret < 0)
		goto out;

4964 4965 4966
	if (!sctx->parent_root || sctx->cur_inode_new) {
		need_chown = 1;
		if (!S_ISLNK(sctx->cur_inode_mode))
4967
			need_chmod = 1;
4968 4969 4970 4971 4972 4973
	} else {
		ret = get_inode_info(sctx->parent_root, sctx->cur_ino,
				NULL, NULL, &right_mode, &right_uid,
				&right_gid, NULL);
		if (ret < 0)
			goto out;
4974

4975 4976 4977 4978
		if (left_uid != right_uid || left_gid != right_gid)
			need_chown = 1;
		if (!S_ISLNK(sctx->cur_inode_mode) && left_mode != right_mode)
			need_chmod = 1;
4979 4980 4981
	}

	if (S_ISREG(sctx->cur_inode_mode)) {
4982
		if (need_send_hole(sctx)) {
4983 4984 4985
			if (sctx->cur_inode_last_extent == (u64)-1 ||
			    sctx->cur_inode_last_extent <
			    sctx->cur_inode_size) {
4986 4987 4988 4989 4990 4991 4992 4993 4994 4995 4996
				ret = get_last_extent(sctx, (u64)-1);
				if (ret)
					goto out;
			}
			if (sctx->cur_inode_last_extent <
			    sctx->cur_inode_size) {
				ret = send_hole(sctx, sctx->cur_inode_size);
				if (ret)
					goto out;
			}
		}
4997 4998 4999 5000 5001 5002 5003 5004 5005 5006 5007 5008 5009 5010 5011 5012 5013 5014 5015 5016
		ret = send_truncate(sctx, sctx->cur_ino, sctx->cur_inode_gen,
				sctx->cur_inode_size);
		if (ret < 0)
			goto out;
	}

	if (need_chown) {
		ret = send_chown(sctx, sctx->cur_ino, sctx->cur_inode_gen,
				left_uid, left_gid);
		if (ret < 0)
			goto out;
	}
	if (need_chmod) {
		ret = send_chmod(sctx, sctx->cur_ino, sctx->cur_inode_gen,
				left_mode);
		if (ret < 0)
			goto out;
	}

	/*
5017 5018
	 * If other directory inodes depended on our current directory
	 * inode's move/rename, now do their move/rename operations.
5019
	 */
5020 5021 5022 5023
	if (!is_waiting_for_move(sctx, sctx->cur_ino)) {
		ret = apply_children_dir_moves(sctx);
		if (ret)
			goto out;
5024 5025 5026 5027 5028 5029 5030 5031 5032 5033 5034
		/*
		 * Need to send that every time, no matter if it actually
		 * changed between the two trees as we have done changes to
		 * the inode before. If our inode is a directory and it's
		 * waiting to be moved/renamed, we will send its utimes when
		 * it's moved/renamed, therefore we don't need to do it here.
		 */
		sctx->send_progress = sctx->cur_ino + 1;
		ret = send_utimes(sctx, sctx->cur_ino, sctx->cur_inode_gen);
		if (ret < 0)
			goto out;
5035 5036
	}

5037 5038 5039 5040 5041 5042 5043 5044 5045 5046 5047 5048 5049 5050 5051 5052
out:
	return ret;
}

static int changed_inode(struct send_ctx *sctx,
			 enum btrfs_compare_tree_result result)
{
	int ret = 0;
	struct btrfs_key *key = sctx->cmp_key;
	struct btrfs_inode_item *left_ii = NULL;
	struct btrfs_inode_item *right_ii = NULL;
	u64 left_gen = 0;
	u64 right_gen = 0;

	sctx->cur_ino = key->objectid;
	sctx->cur_inode_new_gen = 0;
5053
	sctx->cur_inode_last_extent = (u64)-1;
5054 5055 5056 5057 5058 5059

	/*
	 * Set send_progress to current inode. This will tell all get_cur_xxx
	 * functions that the current inode's refs are not updated yet. Later,
	 * when process_recorded_refs is finished, it is set to cur_ino + 1.
	 */
5060 5061 5062 5063 5064 5065 5066 5067 5068 5069 5070 5071 5072 5073 5074 5075 5076 5077 5078 5079 5080 5081 5082
	sctx->send_progress = sctx->cur_ino;

	if (result == BTRFS_COMPARE_TREE_NEW ||
	    result == BTRFS_COMPARE_TREE_CHANGED) {
		left_ii = btrfs_item_ptr(sctx->left_path->nodes[0],
				sctx->left_path->slots[0],
				struct btrfs_inode_item);
		left_gen = btrfs_inode_generation(sctx->left_path->nodes[0],
				left_ii);
	} else {
		right_ii = btrfs_item_ptr(sctx->right_path->nodes[0],
				sctx->right_path->slots[0],
				struct btrfs_inode_item);
		right_gen = btrfs_inode_generation(sctx->right_path->nodes[0],
				right_ii);
	}
	if (result == BTRFS_COMPARE_TREE_CHANGED) {
		right_ii = btrfs_item_ptr(sctx->right_path->nodes[0],
				sctx->right_path->slots[0],
				struct btrfs_inode_item);

		right_gen = btrfs_inode_generation(sctx->right_path->nodes[0],
				right_ii);
5083 5084 5085 5086 5087 5088 5089 5090

		/*
		 * The cur_ino = root dir case is special here. We can't treat
		 * the inode as deleted+reused because it would generate a
		 * stream that tries to delete/mkdir the root dir.
		 */
		if (left_gen != right_gen &&
		    sctx->cur_ino != BTRFS_FIRST_FREE_OBJECTID)
5091 5092 5093 5094 5095 5096 5097 5098 5099 5100 5101
			sctx->cur_inode_new_gen = 1;
	}

	if (result == BTRFS_COMPARE_TREE_NEW) {
		sctx->cur_inode_gen = left_gen;
		sctx->cur_inode_new = 1;
		sctx->cur_inode_deleted = 0;
		sctx->cur_inode_size = btrfs_inode_size(
				sctx->left_path->nodes[0], left_ii);
		sctx->cur_inode_mode = btrfs_inode_mode(
				sctx->left_path->nodes[0], left_ii);
L
Liu Bo 已提交
5102 5103
		sctx->cur_inode_rdev = btrfs_inode_rdev(
				sctx->left_path->nodes[0], left_ii);
5104
		if (sctx->cur_ino != BTRFS_FIRST_FREE_OBJECTID)
5105
			ret = send_create_inode_if_needed(sctx);
5106 5107 5108 5109 5110 5111 5112 5113 5114
	} else if (result == BTRFS_COMPARE_TREE_DELETED) {
		sctx->cur_inode_gen = right_gen;
		sctx->cur_inode_new = 0;
		sctx->cur_inode_deleted = 1;
		sctx->cur_inode_size = btrfs_inode_size(
				sctx->right_path->nodes[0], right_ii);
		sctx->cur_inode_mode = btrfs_inode_mode(
				sctx->right_path->nodes[0], right_ii);
	} else if (result == BTRFS_COMPARE_TREE_CHANGED) {
5115 5116 5117 5118 5119 5120 5121
		/*
		 * We need to do some special handling in case the inode was
		 * reported as changed with a changed generation number. This
		 * means that the original inode was deleted and new inode
		 * reused the same inum. So we have to treat the old inode as
		 * deleted and the new one as new.
		 */
5122
		if (sctx->cur_inode_new_gen) {
5123 5124 5125
			/*
			 * First, process the inode as if it was deleted.
			 */
5126 5127 5128 5129 5130 5131 5132 5133 5134 5135 5136 5137
			sctx->cur_inode_gen = right_gen;
			sctx->cur_inode_new = 0;
			sctx->cur_inode_deleted = 1;
			sctx->cur_inode_size = btrfs_inode_size(
					sctx->right_path->nodes[0], right_ii);
			sctx->cur_inode_mode = btrfs_inode_mode(
					sctx->right_path->nodes[0], right_ii);
			ret = process_all_refs(sctx,
					BTRFS_COMPARE_TREE_DELETED);
			if (ret < 0)
				goto out;

5138 5139 5140
			/*
			 * Now process the inode as if it was new.
			 */
5141 5142 5143 5144 5145 5146 5147
			sctx->cur_inode_gen = left_gen;
			sctx->cur_inode_new = 1;
			sctx->cur_inode_deleted = 0;
			sctx->cur_inode_size = btrfs_inode_size(
					sctx->left_path->nodes[0], left_ii);
			sctx->cur_inode_mode = btrfs_inode_mode(
					sctx->left_path->nodes[0], left_ii);
L
Liu Bo 已提交
5148 5149
			sctx->cur_inode_rdev = btrfs_inode_rdev(
					sctx->left_path->nodes[0], left_ii);
5150
			ret = send_create_inode_if_needed(sctx);
5151 5152 5153 5154 5155 5156
			if (ret < 0)
				goto out;

			ret = process_all_refs(sctx, BTRFS_COMPARE_TREE_NEW);
			if (ret < 0)
				goto out;
5157 5158 5159 5160 5161
			/*
			 * Advance send_progress now as we did not get into
			 * process_recorded_refs_if_needed in the new_gen case.
			 */
			sctx->send_progress = sctx->cur_ino + 1;
5162 5163 5164 5165 5166

			/*
			 * Now process all extents and xattrs of the inode as if
			 * they were all new.
			 */
5167 5168 5169 5170 5171 5172 5173 5174 5175 5176 5177 5178 5179 5180 5181 5182 5183 5184 5185 5186 5187 5188
			ret = process_all_extents(sctx);
			if (ret < 0)
				goto out;
			ret = process_all_new_xattrs(sctx);
			if (ret < 0)
				goto out;
		} else {
			sctx->cur_inode_gen = left_gen;
			sctx->cur_inode_new = 0;
			sctx->cur_inode_new_gen = 0;
			sctx->cur_inode_deleted = 0;
			sctx->cur_inode_size = btrfs_inode_size(
					sctx->left_path->nodes[0], left_ii);
			sctx->cur_inode_mode = btrfs_inode_mode(
					sctx->left_path->nodes[0], left_ii);
		}
	}

out:
	return ret;
}

5189 5190 5191 5192 5193 5194 5195 5196 5197 5198
/*
 * We have to process new refs before deleted refs, but compare_trees gives us
 * the new and deleted refs mixed. To fix this, we record the new/deleted refs
 * first and later process them in process_recorded_refs.
 * For the cur_inode_new_gen case, we skip recording completely because
 * changed_inode did already initiate processing of refs. The reason for this is
 * that in this case, compare_tree actually compares the refs of 2 different
 * inodes. To fix this, process_all_refs is used in changed_inode to handle all
 * refs of the right tree as deleted and all refs of the left tree as new.
 */
5199 5200 5201 5202 5203 5204 5205 5206 5207 5208 5209 5210 5211 5212 5213 5214 5215 5216 5217 5218
static int changed_ref(struct send_ctx *sctx,
		       enum btrfs_compare_tree_result result)
{
	int ret = 0;

	BUG_ON(sctx->cur_ino != sctx->cmp_key->objectid);

	if (!sctx->cur_inode_new_gen &&
	    sctx->cur_ino != BTRFS_FIRST_FREE_OBJECTID) {
		if (result == BTRFS_COMPARE_TREE_NEW)
			ret = record_new_ref(sctx);
		else if (result == BTRFS_COMPARE_TREE_DELETED)
			ret = record_deleted_ref(sctx);
		else if (result == BTRFS_COMPARE_TREE_CHANGED)
			ret = record_changed_ref(sctx);
	}

	return ret;
}

5219 5220 5221 5222 5223
/*
 * Process new/deleted/changed xattrs. We skip processing in the
 * cur_inode_new_gen case because changed_inode did already initiate processing
 * of xattrs. The reason is the same as in changed_ref
 */
5224 5225 5226 5227 5228 5229 5230 5231 5232 5233 5234 5235 5236 5237 5238 5239 5240 5241 5242
static int changed_xattr(struct send_ctx *sctx,
			 enum btrfs_compare_tree_result result)
{
	int ret = 0;

	BUG_ON(sctx->cur_ino != sctx->cmp_key->objectid);

	if (!sctx->cur_inode_new_gen && !sctx->cur_inode_deleted) {
		if (result == BTRFS_COMPARE_TREE_NEW)
			ret = process_new_xattr(sctx);
		else if (result == BTRFS_COMPARE_TREE_DELETED)
			ret = process_deleted_xattr(sctx);
		else if (result == BTRFS_COMPARE_TREE_CHANGED)
			ret = process_changed_xattr(sctx);
	}

	return ret;
}

5243 5244 5245 5246 5247
/*
 * Process new/deleted/changed extents. We skip processing in the
 * cur_inode_new_gen case because changed_inode did already initiate processing
 * of extents. The reason is the same as in changed_ref
 */
5248 5249 5250 5251 5252 5253 5254 5255 5256 5257 5258 5259 5260 5261 5262 5263
static int changed_extent(struct send_ctx *sctx,
			  enum btrfs_compare_tree_result result)
{
	int ret = 0;

	BUG_ON(sctx->cur_ino != sctx->cmp_key->objectid);

	if (!sctx->cur_inode_new_gen && !sctx->cur_inode_deleted) {
		if (result != BTRFS_COMPARE_TREE_DELETED)
			ret = process_extent(sctx, sctx->left_path,
					sctx->cmp_key);
	}

	return ret;
}

5264 5265 5266 5267 5268 5269 5270 5271 5272 5273 5274 5275 5276 5277 5278 5279 5280 5281 5282 5283 5284 5285 5286 5287 5288 5289 5290 5291 5292 5293 5294 5295 5296 5297 5298 5299 5300 5301 5302 5303 5304 5305 5306 5307 5308 5309 5310 5311 5312 5313 5314 5315 5316 5317 5318 5319 5320 5321
static int dir_changed(struct send_ctx *sctx, u64 dir)
{
	u64 orig_gen, new_gen;
	int ret;

	ret = get_inode_info(sctx->send_root, dir, NULL, &new_gen, NULL, NULL,
			     NULL, NULL);
	if (ret)
		return ret;

	ret = get_inode_info(sctx->parent_root, dir, NULL, &orig_gen, NULL,
			     NULL, NULL, NULL);
	if (ret)
		return ret;

	return (orig_gen != new_gen) ? 1 : 0;
}

static int compare_refs(struct send_ctx *sctx, struct btrfs_path *path,
			struct btrfs_key *key)
{
	struct btrfs_inode_extref *extref;
	struct extent_buffer *leaf;
	u64 dirid = 0, last_dirid = 0;
	unsigned long ptr;
	u32 item_size;
	u32 cur_offset = 0;
	int ref_name_len;
	int ret = 0;

	/* Easy case, just check this one dirid */
	if (key->type == BTRFS_INODE_REF_KEY) {
		dirid = key->offset;

		ret = dir_changed(sctx, dirid);
		goto out;
	}

	leaf = path->nodes[0];
	item_size = btrfs_item_size_nr(leaf, path->slots[0]);
	ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
	while (cur_offset < item_size) {
		extref = (struct btrfs_inode_extref *)(ptr +
						       cur_offset);
		dirid = btrfs_inode_extref_parent(leaf, extref);
		ref_name_len = btrfs_inode_extref_name_len(leaf, extref);
		cur_offset += ref_name_len + sizeof(*extref);
		if (dirid == last_dirid)
			continue;
		ret = dir_changed(sctx, dirid);
		if (ret)
			break;
		last_dirid = dirid;
	}
out:
	return ret;
}

5322 5323 5324 5325
/*
 * Updates compare related fields in sctx and simply forwards to the actual
 * changed_xxx functions.
 */
5326 5327 5328 5329 5330 5331 5332 5333 5334 5335 5336
static int changed_cb(struct btrfs_root *left_root,
		      struct btrfs_root *right_root,
		      struct btrfs_path *left_path,
		      struct btrfs_path *right_path,
		      struct btrfs_key *key,
		      enum btrfs_compare_tree_result result,
		      void *ctx)
{
	int ret = 0;
	struct send_ctx *sctx = ctx;

5337
	if (result == BTRFS_COMPARE_TREE_SAME) {
5338 5339 5340 5341 5342 5343 5344 5345 5346 5347
		if (key->type == BTRFS_INODE_REF_KEY ||
		    key->type == BTRFS_INODE_EXTREF_KEY) {
			ret = compare_refs(sctx, left_path, key);
			if (!ret)
				return 0;
			if (ret < 0)
				return ret;
		} else if (key->type == BTRFS_EXTENT_DATA_KEY) {
			return maybe_send_hole(sctx, left_path, key);
		} else {
5348
			return 0;
5349
		}
5350 5351 5352 5353
		result = BTRFS_COMPARE_TREE_CHANGED;
		ret = 0;
	}

5354 5355 5356 5357 5358 5359 5360 5361
	sctx->left_path = left_path;
	sctx->right_path = right_path;
	sctx->cmp_key = key;

	ret = finish_inode_if_needed(sctx, 0);
	if (ret < 0)
		goto out;

5362 5363 5364 5365 5366
	/* Ignore non-FS objects */
	if (key->objectid == BTRFS_FREE_INO_OBJECTID ||
	    key->objectid == BTRFS_FREE_SPACE_OBJECTID)
		goto out;

5367 5368
	if (key->type == BTRFS_INODE_ITEM_KEY)
		ret = changed_inode(sctx, result);
5369 5370
	else if (key->type == BTRFS_INODE_REF_KEY ||
		 key->type == BTRFS_INODE_EXTREF_KEY)
5371 5372 5373 5374 5375 5376 5377 5378 5379 5380 5381 5382 5383 5384 5385 5386 5387 5388 5389 5390 5391 5392 5393 5394 5395 5396 5397 5398 5399 5400 5401 5402 5403 5404 5405 5406 5407 5408 5409 5410 5411 5412 5413 5414 5415 5416 5417 5418 5419 5420 5421 5422 5423 5424 5425 5426 5427 5428 5429 5430 5431 5432 5433 5434 5435 5436 5437 5438 5439
		ret = changed_ref(sctx, result);
	else if (key->type == BTRFS_XATTR_ITEM_KEY)
		ret = changed_xattr(sctx, result);
	else if (key->type == BTRFS_EXTENT_DATA_KEY)
		ret = changed_extent(sctx, result);

out:
	return ret;
}

static int full_send_tree(struct send_ctx *sctx)
{
	int ret;
	struct btrfs_root *send_root = sctx->send_root;
	struct btrfs_key key;
	struct btrfs_key found_key;
	struct btrfs_path *path;
	struct extent_buffer *eb;
	int slot;

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

	key.objectid = BTRFS_FIRST_FREE_OBJECTID;
	key.type = BTRFS_INODE_ITEM_KEY;
	key.offset = 0;

	ret = btrfs_search_slot_for_read(send_root, &key, path, 1, 0);
	if (ret < 0)
		goto out;
	if (ret)
		goto out_finish;

	while (1) {
		eb = path->nodes[0];
		slot = path->slots[0];
		btrfs_item_key_to_cpu(eb, &found_key, slot);

		ret = changed_cb(send_root, NULL, path, NULL,
				&found_key, BTRFS_COMPARE_TREE_NEW, sctx);
		if (ret < 0)
			goto out;

		key.objectid = found_key.objectid;
		key.type = found_key.type;
		key.offset = found_key.offset + 1;

		ret = btrfs_next_item(send_root, path);
		if (ret < 0)
			goto out;
		if (ret) {
			ret  = 0;
			break;
		}
	}

out_finish:
	ret = finish_inode_if_needed(sctx, 1);

out:
	btrfs_free_path(path);
	return ret;
}

static int send_subvol(struct send_ctx *sctx)
{
	int ret;

5440 5441 5442 5443 5444
	if (!(sctx->flags & BTRFS_SEND_FLAG_OMIT_STREAM_HEADER)) {
		ret = send_header(sctx);
		if (ret < 0)
			goto out;
	}
5445 5446 5447 5448 5449 5450 5451 5452 5453 5454 5455 5456 5457 5458 5459 5460 5461 5462 5463 5464 5465 5466 5467 5468

	ret = send_subvol_begin(sctx);
	if (ret < 0)
		goto out;

	if (sctx->parent_root) {
		ret = btrfs_compare_trees(sctx->send_root, sctx->parent_root,
				changed_cb, sctx);
		if (ret < 0)
			goto out;
		ret = finish_inode_if_needed(sctx, 1);
		if (ret < 0)
			goto out;
	} else {
		ret = full_send_tree(sctx);
		if (ret < 0)
			goto out;
	}

out:
	free_recorded_refs(sctx);
	return ret;
}

5469 5470 5471 5472 5473 5474 5475 5476 5477 5478 5479 5480 5481 5482 5483
static void btrfs_root_dec_send_in_progress(struct btrfs_root* root)
{
	spin_lock(&root->root_item_lock);
	root->send_in_progress--;
	/*
	 * Not much left to do, we don't know why it's unbalanced and
	 * can't blindly reset it to 0.
	 */
	if (root->send_in_progress < 0)
		btrfs_err(root->fs_info,
			"send_in_progres unbalanced %d root %llu\n",
			root->send_in_progress, root->root_key.objectid);
	spin_unlock(&root->root_item_lock);
}

5484 5485 5486 5487 5488 5489 5490 5491 5492 5493 5494
long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_)
{
	int ret = 0;
	struct btrfs_root *send_root;
	struct btrfs_root *clone_root;
	struct btrfs_fs_info *fs_info;
	struct btrfs_ioctl_send_args *arg = NULL;
	struct btrfs_key key;
	struct send_ctx *sctx = NULL;
	u32 i;
	u64 *clone_sources_tmp = NULL;
5495
	int clone_sources_to_rollback = 0;
5496
	int sort_clone_roots = 0;
5497
	int index;
5498 5499 5500 5501

	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

A
Al Viro 已提交
5502
	send_root = BTRFS_I(file_inode(mnt_file))->root;
5503 5504
	fs_info = send_root->fs_info;

5505 5506
	/*
	 * The subvolume must remain read-only during send, protect against
5507
	 * making it RW. This also protects against deletion.
5508 5509 5510 5511 5512
	 */
	spin_lock(&send_root->root_item_lock);
	send_root->send_in_progress++;
	spin_unlock(&send_root->root_item_lock);

J
Josef Bacik 已提交
5513 5514 5515 5516 5517 5518
	/*
	 * This is done when we lookup the root, it should already be complete
	 * by the time we get here.
	 */
	WARN_ON(send_root->orphan_cleanup_state != ORPHAN_CLEANUP_DONE);

5519 5520 5521 5522 5523 5524 5525 5526 5527
	/*
	 * Userspace tools do the checks and warn the user if it's
	 * not RO.
	 */
	if (!btrfs_root_readonly(send_root)) {
		ret = -EPERM;
		goto out;
	}

5528 5529 5530 5531 5532 5533 5534 5535
	arg = memdup_user(arg_, sizeof(*arg));
	if (IS_ERR(arg)) {
		ret = PTR_ERR(arg);
		arg = NULL;
		goto out;
	}

	if (!access_ok(VERIFY_READ, arg->clone_sources,
5536 5537
			sizeof(*arg->clone_sources) *
			arg->clone_sources_count)) {
5538 5539 5540 5541
		ret = -EFAULT;
		goto out;
	}

5542
	if (arg->flags & ~BTRFS_SEND_FLAG_MASK) {
5543 5544 5545 5546
		ret = -EINVAL;
		goto out;
	}

5547 5548 5549 5550 5551 5552 5553 5554 5555 5556 5557
	sctx = kzalloc(sizeof(struct send_ctx), GFP_NOFS);
	if (!sctx) {
		ret = -ENOMEM;
		goto out;
	}

	INIT_LIST_HEAD(&sctx->new_refs);
	INIT_LIST_HEAD(&sctx->deleted_refs);
	INIT_RADIX_TREE(&sctx->name_cache, GFP_NOFS);
	INIT_LIST_HEAD(&sctx->name_cache_list);

5558 5559
	sctx->flags = arg->flags;

5560
	sctx->send_filp = fget(arg->send_fd);
5561 5562
	if (!sctx->send_filp) {
		ret = -EBADF;
5563 5564 5565 5566
		goto out;
	}

	sctx->send_root = send_root;
5567 5568 5569 5570 5571 5572 5573 5574 5575
	/*
	 * Unlikely but possible, if the subvolume is marked for deletion but
	 * is slow to remove the directory entry, send can still be started
	 */
	if (btrfs_root_dead(sctx->send_root)) {
		ret = -EPERM;
		goto out;
	}

5576 5577 5578 5579 5580 5581 5582 5583 5584 5585 5586 5587 5588 5589 5590
	sctx->clone_roots_cnt = arg->clone_sources_count;

	sctx->send_max_size = BTRFS_SEND_BUF_SIZE;
	sctx->send_buf = vmalloc(sctx->send_max_size);
	if (!sctx->send_buf) {
		ret = -ENOMEM;
		goto out;
	}

	sctx->read_buf = vmalloc(BTRFS_SEND_READ_SIZE);
	if (!sctx->read_buf) {
		ret = -ENOMEM;
		goto out;
	}

5591 5592
	sctx->pending_dir_moves = RB_ROOT;
	sctx->waiting_dir_moves = RB_ROOT;
5593
	sctx->orphan_dirs = RB_ROOT;
5594

5595 5596 5597 5598 5599 5600 5601 5602 5603 5604 5605 5606 5607 5608 5609 5610 5611 5612 5613 5614 5615 5616 5617 5618 5619 5620 5621
	sctx->clone_roots = vzalloc(sizeof(struct clone_root) *
			(arg->clone_sources_count + 1));
	if (!sctx->clone_roots) {
		ret = -ENOMEM;
		goto out;
	}

	if (arg->clone_sources_count) {
		clone_sources_tmp = vmalloc(arg->clone_sources_count *
				sizeof(*arg->clone_sources));
		if (!clone_sources_tmp) {
			ret = -ENOMEM;
			goto out;
		}

		ret = copy_from_user(clone_sources_tmp, arg->clone_sources,
				arg->clone_sources_count *
				sizeof(*arg->clone_sources));
		if (ret) {
			ret = -EFAULT;
			goto out;
		}

		for (i = 0; i < arg->clone_sources_count; i++) {
			key.objectid = clone_sources_tmp[i];
			key.type = BTRFS_ROOT_ITEM_KEY;
			key.offset = (u64)-1;
5622 5623 5624

			index = srcu_read_lock(&fs_info->subvol_srcu);

5625 5626
			clone_root = btrfs_read_fs_root_no_name(fs_info, &key);
			if (IS_ERR(clone_root)) {
5627
				srcu_read_unlock(&fs_info->subvol_srcu, index);
5628 5629 5630
				ret = PTR_ERR(clone_root);
				goto out;
			}
5631 5632 5633 5634 5635
			clone_sources_to_rollback = i + 1;
			spin_lock(&clone_root->root_item_lock);
			clone_root->send_in_progress++;
			if (!btrfs_root_readonly(clone_root)) {
				spin_unlock(&clone_root->root_item_lock);
5636
				srcu_read_unlock(&fs_info->subvol_srcu, index);
5637 5638 5639 5640
				ret = -EPERM;
				goto out;
			}
			spin_unlock(&clone_root->root_item_lock);
5641 5642
			srcu_read_unlock(&fs_info->subvol_srcu, index);

5643 5644 5645 5646 5647 5648 5649 5650 5651 5652
			sctx->clone_roots[i].root = clone_root;
		}
		vfree(clone_sources_tmp);
		clone_sources_tmp = NULL;
	}

	if (arg->parent_root) {
		key.objectid = arg->parent_root;
		key.type = BTRFS_ROOT_ITEM_KEY;
		key.offset = (u64)-1;
5653 5654 5655

		index = srcu_read_lock(&fs_info->subvol_srcu);

5656
		sctx->parent_root = btrfs_read_fs_root_no_name(fs_info, &key);
5657
		if (IS_ERR(sctx->parent_root)) {
5658
			srcu_read_unlock(&fs_info->subvol_srcu, index);
5659
			ret = PTR_ERR(sctx->parent_root);
5660 5661
			goto out;
		}
5662

5663 5664
		spin_lock(&sctx->parent_root->root_item_lock);
		sctx->parent_root->send_in_progress++;
5665 5666
		if (!btrfs_root_readonly(sctx->parent_root) ||
				btrfs_root_dead(sctx->parent_root)) {
5667
			spin_unlock(&sctx->parent_root->root_item_lock);
5668
			srcu_read_unlock(&fs_info->subvol_srcu, index);
5669 5670 5671 5672
			ret = -EPERM;
			goto out;
		}
		spin_unlock(&sctx->parent_root->root_item_lock);
5673 5674

		srcu_read_unlock(&fs_info->subvol_srcu, index);
5675 5676 5677 5678 5679 5680 5681 5682 5683 5684 5685 5686 5687
	}

	/*
	 * Clones from send_root are allowed, but only if the clone source
	 * is behind the current send position. This is checked while searching
	 * for possible clone sources.
	 */
	sctx->clone_roots[sctx->clone_roots_cnt++].root = sctx->send_root;

	/* We do a bsearch later */
	sort(sctx->clone_roots, sctx->clone_roots_cnt,
			sizeof(*sctx->clone_roots), __clone_root_cmp_sort,
			NULL);
5688
	sort_clone_roots = 1;
5689

5690
	current->journal_info = (void *)BTRFS_SEND_TRANS_STUB;
5691
	ret = send_subvol(sctx);
5692
	current->journal_info = NULL;
5693 5694 5695
	if (ret < 0)
		goto out;

5696 5697 5698 5699 5700 5701 5702 5703
	if (!(sctx->flags & BTRFS_SEND_FLAG_OMIT_END_CMD)) {
		ret = begin_cmd(sctx, BTRFS_SEND_C_END);
		if (ret < 0)
			goto out;
		ret = send_cmd(sctx);
		if (ret < 0)
			goto out;
	}
5704 5705

out:
5706 5707 5708 5709 5710 5711 5712 5713 5714 5715 5716 5717 5718 5719 5720 5721 5722 5723 5724 5725 5726 5727 5728 5729 5730 5731 5732 5733
	WARN_ON(sctx && !ret && !RB_EMPTY_ROOT(&sctx->pending_dir_moves));
	while (sctx && !RB_EMPTY_ROOT(&sctx->pending_dir_moves)) {
		struct rb_node *n;
		struct pending_dir_move *pm;

		n = rb_first(&sctx->pending_dir_moves);
		pm = rb_entry(n, struct pending_dir_move, node);
		while (!list_empty(&pm->list)) {
			struct pending_dir_move *pm2;

			pm2 = list_first_entry(&pm->list,
					       struct pending_dir_move, list);
			free_pending_move(sctx, pm2);
		}
		free_pending_move(sctx, pm);
	}

	WARN_ON(sctx && !ret && !RB_EMPTY_ROOT(&sctx->waiting_dir_moves));
	while (sctx && !RB_EMPTY_ROOT(&sctx->waiting_dir_moves)) {
		struct rb_node *n;
		struct waiting_dir_move *dm;

		n = rb_first(&sctx->waiting_dir_moves);
		dm = rb_entry(n, struct waiting_dir_move, node);
		rb_erase(&dm->node, &sctx->waiting_dir_moves);
		kfree(dm);
	}

5734 5735 5736 5737 5738 5739 5740 5741 5742 5743
	WARN_ON(sctx && !ret && !RB_EMPTY_ROOT(&sctx->orphan_dirs));
	while (sctx && !RB_EMPTY_ROOT(&sctx->orphan_dirs)) {
		struct rb_node *n;
		struct orphan_dir_info *odi;

		n = rb_first(&sctx->orphan_dirs);
		odi = rb_entry(n, struct orphan_dir_info, node);
		free_orphan_dir_info(sctx, odi);
	}

5744 5745 5746 5747 5748 5749 5750 5751 5752 5753 5754
	if (sort_clone_roots) {
		for (i = 0; i < sctx->clone_roots_cnt; i++)
			btrfs_root_dec_send_in_progress(
					sctx->clone_roots[i].root);
	} else {
		for (i = 0; sctx && i < clone_sources_to_rollback; i++)
			btrfs_root_dec_send_in_progress(
					sctx->clone_roots[i].root);

		btrfs_root_dec_send_in_progress(send_root);
	}
5755 5756
	if (sctx && !IS_ERR_OR_NULL(sctx->parent_root))
		btrfs_root_dec_send_in_progress(sctx->parent_root);
5757

5758 5759 5760 5761 5762 5763 5764 5765 5766 5767 5768 5769 5770 5771 5772 5773 5774 5775
	kfree(arg);
	vfree(clone_sources_tmp);

	if (sctx) {
		if (sctx->send_filp)
			fput(sctx->send_filp);

		vfree(sctx->clone_roots);
		vfree(sctx->send_buf);
		vfree(sctx->read_buf);

		name_cache_free(sctx);

		kfree(sctx);
	}

	return ret;
}