ioctl.c 128.4 KB
Newer Older
C
Christoph Hellwig 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
/*
 * Copyright (C) 2007 Oracle.  All rights reserved.
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public
 * License v2 as published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * General Public License for more details.
 *
 * You should have received a copy of the GNU General Public
 * License along with this program; if not, write to the
 * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
 * Boston, MA 021110-1307, USA.
 */

#include <linux/kernel.h>
#include <linux/bio.h>
#include <linux/buffer_head.h>
#include <linux/file.h>
#include <linux/fs.h>
24
#include <linux/fsnotify.h>
C
Christoph Hellwig 已提交
25 26 27 28 29 30
#include <linux/pagemap.h>
#include <linux/highmem.h>
#include <linux/time.h>
#include <linux/init.h>
#include <linux/string.h>
#include <linux/backing-dev.h>
31
#include <linux/mount.h>
C
Christoph Hellwig 已提交
32
#include <linux/mpage.h>
33
#include <linux/namei.h>
C
Christoph Hellwig 已提交
34 35 36 37 38
#include <linux/swap.h>
#include <linux/writeback.h>
#include <linux/statfs.h>
#include <linux/compat.h>
#include <linux/bit_spinlock.h>
39
#include <linux/security.h>
C
Christoph Hellwig 已提交
40
#include <linux/xattr.h>
41
#include <linux/vmalloc.h>
42
#include <linux/slab.h>
43
#include <linux/blkdev.h>
44
#include <linux/uuid.h>
45
#include <linux/btrfs.h>
M
Mark Fasheh 已提交
46
#include <linux/uaccess.h>
C
Christoph Hellwig 已提交
47 48 49 50 51 52
#include "ctree.h"
#include "disk-io.h"
#include "transaction.h"
#include "btrfs_inode.h"
#include "print-tree.h"
#include "volumes.h"
53
#include "locking.h"
54
#include "inode-map.h"
55
#include "backref.h"
56
#include "rcu-string.h"
57
#include "send.h"
58
#include "dev-replace.h"
59
#include "props.h"
60
#include "sysfs.h"
J
Josef Bacik 已提交
61
#include "qgroup.h"
C
Christoph Hellwig 已提交
62

63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88
#ifdef CONFIG_64BIT
/* If we have a 32-bit userspace and 64-bit kernel, then the UAPI
 * structures are incorrect, as the timespec structure from userspace
 * is 4 bytes too small. We define these alternatives here to teach
 * the kernel about the 32-bit struct packing.
 */
struct btrfs_ioctl_timespec_32 {
	__u64 sec;
	__u32 nsec;
} __attribute__ ((__packed__));

struct btrfs_ioctl_received_subvol_args_32 {
	char	uuid[BTRFS_UUID_SIZE];	/* in */
	__u64	stransid;		/* in */
	__u64	rtransid;		/* out */
	struct btrfs_ioctl_timespec_32 stime; /* in */
	struct btrfs_ioctl_timespec_32 rtime; /* out */
	__u64	flags;			/* in */
	__u64	reserved[16];		/* in */
} __attribute__ ((__packed__));

#define BTRFS_IOC_SET_RECEIVED_SUBVOL_32 _IOWR(BTRFS_IOCTL_MAGIC, 37, \
				struct btrfs_ioctl_received_subvol_args_32)
#endif


M
Mark Fasheh 已提交
89 90 91
static int btrfs_clone(struct inode *src, struct inode *inode,
		       u64 off, u64 olen, u64 olen_aligned, u64 destoff);

92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121
/* Mask out flags that are inappropriate for the given type of inode. */
static inline __u32 btrfs_mask_flags(umode_t mode, __u32 flags)
{
	if (S_ISDIR(mode))
		return flags;
	else if (S_ISREG(mode))
		return flags & ~FS_DIRSYNC_FL;
	else
		return flags & (FS_NODUMP_FL | FS_NOATIME_FL);
}

/*
 * Export inode flags to the format expected by the FS_IOC_GETFLAGS ioctl.
 */
static unsigned int btrfs_flags_to_ioctl(unsigned int flags)
{
	unsigned int iflags = 0;

	if (flags & BTRFS_INODE_SYNC)
		iflags |= FS_SYNC_FL;
	if (flags & BTRFS_INODE_IMMUTABLE)
		iflags |= FS_IMMUTABLE_FL;
	if (flags & BTRFS_INODE_APPEND)
		iflags |= FS_APPEND_FL;
	if (flags & BTRFS_INODE_NODUMP)
		iflags |= FS_NODUMP_FL;
	if (flags & BTRFS_INODE_NOATIME)
		iflags |= FS_NOATIME_FL;
	if (flags & BTRFS_INODE_DIRSYNC)
		iflags |= FS_DIRSYNC_FL;
L
Li Zefan 已提交
122 123 124 125 126 127 128
	if (flags & BTRFS_INODE_NODATACOW)
		iflags |= FS_NOCOW_FL;

	if ((flags & BTRFS_INODE_COMPRESS) && !(flags & BTRFS_INODE_NOCOMPRESS))
		iflags |= FS_COMPR_FL;
	else if (flags & BTRFS_INODE_NOCOMPRESS)
		iflags |= FS_NOCOMP_FL;
129 130 131 132 133 134 135 136 137 138

	return iflags;
}

/*
 * Update inode->i_flags based on the btrfs internal flags.
 */
void btrfs_update_iflags(struct inode *inode)
{
	struct btrfs_inode *ip = BTRFS_I(inode);
139
	unsigned int new_fl = 0;
140 141

	if (ip->flags & BTRFS_INODE_SYNC)
142
		new_fl |= S_SYNC;
143
	if (ip->flags & BTRFS_INODE_IMMUTABLE)
144
		new_fl |= S_IMMUTABLE;
145
	if (ip->flags & BTRFS_INODE_APPEND)
146
		new_fl |= S_APPEND;
147
	if (ip->flags & BTRFS_INODE_NOATIME)
148
		new_fl |= S_NOATIME;
149
	if (ip->flags & BTRFS_INODE_DIRSYNC)
150 151 152 153 154
		new_fl |= S_DIRSYNC;

	set_mask_bits(&inode->i_flags,
		      S_SYNC | S_APPEND | S_IMMUTABLE | S_NOATIME | S_DIRSYNC,
		      new_fl);
155 156 157 158 159
}

/*
 * Inherit flags from the parent inode.
 *
160
 * Currently only the compression flags and the cow flags are inherited.
161 162 163
 */
void btrfs_inherit_iflags(struct inode *inode, struct inode *dir)
{
164 165 166 167 168 169
	unsigned int flags;

	if (!dir)
		return;

	flags = BTRFS_I(dir)->flags;
170

171 172 173 174 175 176 177 178
	if (flags & BTRFS_INODE_NOCOMPRESS) {
		BTRFS_I(inode)->flags &= ~BTRFS_INODE_COMPRESS;
		BTRFS_I(inode)->flags |= BTRFS_INODE_NOCOMPRESS;
	} else if (flags & BTRFS_INODE_COMPRESS) {
		BTRFS_I(inode)->flags &= ~BTRFS_INODE_NOCOMPRESS;
		BTRFS_I(inode)->flags |= BTRFS_INODE_COMPRESS;
	}

L
Liu Bo 已提交
179
	if (flags & BTRFS_INODE_NODATACOW) {
180
		BTRFS_I(inode)->flags |= BTRFS_INODE_NODATACOW;
L
Liu Bo 已提交
181 182 183
		if (S_ISREG(inode->i_mode))
			BTRFS_I(inode)->flags |= BTRFS_INODE_NODATASUM;
	}
184 185 186 187 188 189

	btrfs_update_iflags(inode);
}

static int btrfs_ioctl_getflags(struct file *file, void __user *arg)
{
A
Al Viro 已提交
190
	struct btrfs_inode *ip = BTRFS_I(file_inode(file));
191 192 193 194 195 196 197
	unsigned int flags = btrfs_flags_to_ioctl(ip->flags);

	if (copy_to_user(arg, &flags, sizeof(flags)))
		return -EFAULT;
	return 0;
}

198 199 200 201 202
static int check_flags(unsigned int flags)
{
	if (flags & ~(FS_IMMUTABLE_FL | FS_APPEND_FL | \
		      FS_NOATIME_FL | FS_NODUMP_FL | \
		      FS_SYNC_FL | FS_DIRSYNC_FL | \
L
Li Zefan 已提交
203 204
		      FS_NOCOMP_FL | FS_COMPR_FL |
		      FS_NOCOW_FL))
205 206 207 208 209 210 211 212
		return -EOPNOTSUPP;

	if ((flags & FS_NOCOMP_FL) && (flags & FS_COMPR_FL))
		return -EINVAL;

	return 0;
}

213 214
static int btrfs_ioctl_setflags(struct file *file, void __user *arg)
{
A
Al Viro 已提交
215
	struct inode *inode = file_inode(file);
216 217 218 219 220
	struct btrfs_inode *ip = BTRFS_I(inode);
	struct btrfs_root *root = ip->root;
	struct btrfs_trans_handle *trans;
	unsigned int flags, oldflags;
	int ret;
221 222
	u64 ip_oldflags;
	unsigned int i_oldflags;
223
	umode_t mode;
224

225 226 227
	if (!inode_owner_or_capable(inode))
		return -EPERM;

L
Li Zefan 已提交
228 229 230
	if (btrfs_root_readonly(root))
		return -EROFS;

231 232 233
	if (copy_from_user(&flags, arg, sizeof(flags)))
		return -EFAULT;

234 235 236
	ret = check_flags(flags);
	if (ret)
		return ret;
C
Christoph Hellwig 已提交
237

238 239 240 241
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;

242 243
	mutex_lock(&inode->i_mutex);

244 245
	ip_oldflags = ip->flags;
	i_oldflags = inode->i_flags;
246
	mode = inode->i_mode;
247

248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280
	flags = btrfs_mask_flags(inode->i_mode, flags);
	oldflags = btrfs_flags_to_ioctl(ip->flags);
	if ((flags ^ oldflags) & (FS_APPEND_FL | FS_IMMUTABLE_FL)) {
		if (!capable(CAP_LINUX_IMMUTABLE)) {
			ret = -EPERM;
			goto out_unlock;
		}
	}

	if (flags & FS_SYNC_FL)
		ip->flags |= BTRFS_INODE_SYNC;
	else
		ip->flags &= ~BTRFS_INODE_SYNC;
	if (flags & FS_IMMUTABLE_FL)
		ip->flags |= BTRFS_INODE_IMMUTABLE;
	else
		ip->flags &= ~BTRFS_INODE_IMMUTABLE;
	if (flags & FS_APPEND_FL)
		ip->flags |= BTRFS_INODE_APPEND;
	else
		ip->flags &= ~BTRFS_INODE_APPEND;
	if (flags & FS_NODUMP_FL)
		ip->flags |= BTRFS_INODE_NODUMP;
	else
		ip->flags &= ~BTRFS_INODE_NODUMP;
	if (flags & FS_NOATIME_FL)
		ip->flags |= BTRFS_INODE_NOATIME;
	else
		ip->flags &= ~BTRFS_INODE_NOATIME;
	if (flags & FS_DIRSYNC_FL)
		ip->flags |= BTRFS_INODE_DIRSYNC;
	else
		ip->flags &= ~BTRFS_INODE_DIRSYNC;
281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305
	if (flags & FS_NOCOW_FL) {
		if (S_ISREG(mode)) {
			/*
			 * It's safe to turn csums off here, no extents exist.
			 * Otherwise we want the flag to reflect the real COW
			 * status of the file and will not set it.
			 */
			if (inode->i_size == 0)
				ip->flags |= BTRFS_INODE_NODATACOW
					   | BTRFS_INODE_NODATASUM;
		} else {
			ip->flags |= BTRFS_INODE_NODATACOW;
		}
	} else {
		/*
		 * Revert back under same assuptions as above
		 */
		if (S_ISREG(mode)) {
			if (inode->i_size == 0)
				ip->flags &= ~(BTRFS_INODE_NODATACOW
				             | BTRFS_INODE_NODATASUM);
		} else {
			ip->flags &= ~BTRFS_INODE_NODATACOW;
		}
	}
306

307 308 309 310 311 312 313 314
	/*
	 * The COMPRESS flag can only be changed by users, while the NOCOMPRESS
	 * flag may be changed automatically if compression code won't make
	 * things smaller.
	 */
	if (flags & FS_NOCOMP_FL) {
		ip->flags &= ~BTRFS_INODE_COMPRESS;
		ip->flags |= BTRFS_INODE_NOCOMPRESS;
315 316 317 318

		ret = btrfs_set_prop(inode, "btrfs.compression", NULL, 0, 0);
		if (ret && ret != -ENODATA)
			goto out_drop;
319
	} else if (flags & FS_COMPR_FL) {
320 321
		const char *comp;

322 323
		ip->flags |= BTRFS_INODE_COMPRESS;
		ip->flags &= ~BTRFS_INODE_NOCOMPRESS;
324 325 326 327 328 329 330 331 332 333

		if (root->fs_info->compress_type == BTRFS_COMPRESS_LZO)
			comp = "lzo";
		else
			comp = "zlib";
		ret = btrfs_set_prop(inode, "btrfs.compression",
				     comp, strlen(comp), 0);
		if (ret)
			goto out_drop;

L
Li Zefan 已提交
334
	} else {
335 336 337
		ret = btrfs_set_prop(inode, "btrfs.compression", NULL, 0, 0);
		if (ret && ret != -ENODATA)
			goto out_drop;
L
Li Zefan 已提交
338
		ip->flags &= ~(BTRFS_INODE_COMPRESS | BTRFS_INODE_NOCOMPRESS);
339
	}
340

341
	trans = btrfs_start_transaction(root, 1);
342 343 344 345
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto out_drop;
	}
346

347
	btrfs_update_iflags(inode);
348
	inode_inc_iversion(inode);
349
	inode->i_ctime = CURRENT_TIME;
350 351 352
	ret = btrfs_update_inode(trans, root, inode);

	btrfs_end_transaction(trans, root);
353 354 355 356 357
 out_drop:
	if (ret) {
		ip->flags = ip_oldflags;
		inode->i_flags = i_oldflags;
	}
358 359 360

 out_unlock:
	mutex_unlock(&inode->i_mutex);
361
	mnt_drop_write_file(file);
362
	return ret;
363 364 365 366
}

static int btrfs_ioctl_getversion(struct file *file, int __user *arg)
{
A
Al Viro 已提交
367
	struct inode *inode = file_inode(file);
368 369 370

	return put_user(inode->i_generation, arg);
}
C
Christoph Hellwig 已提交
371

372 373
static noinline int btrfs_ioctl_fitrim(struct file *file, void __user *arg)
{
A
Al Viro 已提交
374
	struct btrfs_fs_info *fs_info = btrfs_sb(file_inode(file)->i_sb);
375 376 377 378 379
	struct btrfs_device *device;
	struct request_queue *q;
	struct fstrim_range range;
	u64 minlen = ULLONG_MAX;
	u64 num_devices = 0;
380
	u64 total_bytes = btrfs_super_total_bytes(fs_info->super_copy);
381 382 383 384 385
	int ret;

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

386 387 388
	rcu_read_lock();
	list_for_each_entry_rcu(device, &fs_info->fs_devices->devices,
				dev_list) {
389 390 391 392 393 394 395 396 397
		if (!device->bdev)
			continue;
		q = bdev_get_queue(device->bdev);
		if (blk_queue_discard(q)) {
			num_devices++;
			minlen = min((u64)q->limits.discard_granularity,
				     minlen);
		}
	}
398
	rcu_read_unlock();
399

400 401 402 403
	if (!num_devices)
		return -EOPNOTSUPP;
	if (copy_from_user(&range, arg, sizeof(range)))
		return -EFAULT;
404 405
	if (range.start > total_bytes ||
	    range.len < fs_info->sb->s_blocksize)
406
		return -EINVAL;
407

408
	range.len = min(range.len, total_bytes - range.start);
409
	range.minlen = max(range.minlen, minlen);
410
	ret = btrfs_trim_fs(fs_info->tree_root, &range);
411 412 413 414 415 416 417 418 419
	if (ret < 0)
		return ret;

	if (copy_to_user(arg, &range, sizeof(range)))
		return -EFAULT;

	return 0;
}

420 421
int btrfs_is_empty_uuid(u8 *uuid)
{
C
Chris Mason 已提交
422 423 424 425 426 427 428
	int i;

	for (i = 0; i < BTRFS_UUID_SIZE; i++) {
		if (uuid[i])
			return 0;
	}
	return 1;
429 430
}

431
static noinline int create_subvol(struct inode *dir,
432
				  struct dentry *dentry,
S
Sage Weil 已提交
433
				  char *name, int namelen,
A
Arne Jansen 已提交
434
				  u64 *async_transid,
435
				  struct btrfs_qgroup_inherit *inherit)
C
Christoph Hellwig 已提交
436 437 438 439 440 441
{
	struct btrfs_trans_handle *trans;
	struct btrfs_key key;
	struct btrfs_root_item root_item;
	struct btrfs_inode_item *inode_item;
	struct extent_buffer *leaf;
442
	struct btrfs_root *root = BTRFS_I(dir)->root;
443
	struct btrfs_root *new_root;
444
	struct btrfs_block_rsv block_rsv;
445
	struct timespec cur_time = CURRENT_TIME;
446
	struct inode *inode;
C
Christoph Hellwig 已提交
447 448 449 450
	int ret;
	int err;
	u64 objectid;
	u64 new_dirid = BTRFS_FIRST_FREE_OBJECTID;
451
	u64 index = 0;
452
	u64 qgroup_reserved;
453
	uuid_le new_uuid;
C
Christoph Hellwig 已提交
454

455
	ret = btrfs_find_free_objectid(root->fs_info->tree_root, &objectid);
456
	if (ret)
457
		return ret;
458

459
	btrfs_init_block_rsv(&block_rsv, BTRFS_BLOCK_RSV_TEMP);
J
Josef Bacik 已提交
460
	/*
461 462
	 * The same as the snapshot creation, please see the comment
	 * of create_snapshot().
J
Josef Bacik 已提交
463
	 */
464
	ret = btrfs_subvolume_reserve_metadata(root, &block_rsv,
465
					       8, &qgroup_reserved, false);
466 467 468 469 470 471
	if (ret)
		return ret;

	trans = btrfs_start_transaction(root, 0);
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
472 473 474
		btrfs_subvolume_release_metadata(root, &block_rsv,
						 qgroup_reserved);
		return ret;
475 476 477
	}
	trans->block_rsv = &block_rsv;
	trans->bytes_reserved = block_rsv.size;
C
Christoph Hellwig 已提交
478

479
	ret = btrfs_qgroup_inherit(trans, root->fs_info, 0, objectid, inherit);
A
Arne Jansen 已提交
480 481 482
	if (ret)
		goto fail;

483
	leaf = btrfs_alloc_tree_block(trans, root, 0, objectid, NULL, 0, 0, 0);
484 485 486 487
	if (IS_ERR(leaf)) {
		ret = PTR_ERR(leaf);
		goto fail;
	}
C
Christoph Hellwig 已提交
488

489
	memset_extent_buffer(leaf, 0, 0, sizeof(struct btrfs_header));
C
Christoph Hellwig 已提交
490 491
	btrfs_set_header_bytenr(leaf, leaf->start);
	btrfs_set_header_generation(leaf, trans->transid);
492
	btrfs_set_header_backref_rev(leaf, BTRFS_MIXED_BACKREF_REV);
C
Christoph Hellwig 已提交
493 494
	btrfs_set_header_owner(leaf, objectid);

495
	write_extent_buffer(leaf, root->fs_info->fsid, btrfs_header_fsid(),
C
Christoph Hellwig 已提交
496
			    BTRFS_FSID_SIZE);
497
	write_extent_buffer(leaf, root->fs_info->chunk_tree_uuid,
498
			    btrfs_header_chunk_tree_uuid(leaf),
499
			    BTRFS_UUID_SIZE);
C
Christoph Hellwig 已提交
500 501
	btrfs_mark_buffer_dirty(leaf);

502 503
	memset(&root_item, 0, sizeof(root_item));

C
Christoph Hellwig 已提交
504
	inode_item = &root_item.inode;
505 506 507
	btrfs_set_stack_inode_generation(inode_item, 1);
	btrfs_set_stack_inode_size(inode_item, 3);
	btrfs_set_stack_inode_nlink(inode_item, 1);
508
	btrfs_set_stack_inode_nbytes(inode_item, root->nodesize);
509
	btrfs_set_stack_inode_mode(inode_item, S_IFDIR | 0755);
C
Christoph Hellwig 已提交
510

511 512 513
	btrfs_set_root_flags(&root_item, 0);
	btrfs_set_root_limit(&root_item, 0);
	btrfs_set_stack_inode_flags(inode_item, BTRFS_INODE_ROOT_ITEM_INIT);
514

C
Christoph Hellwig 已提交
515
	btrfs_set_root_bytenr(&root_item, leaf->start);
516
	btrfs_set_root_generation(&root_item, trans->transid);
C
Christoph Hellwig 已提交
517 518
	btrfs_set_root_level(&root_item, 0);
	btrfs_set_root_refs(&root_item, 1);
519
	btrfs_set_root_used(&root_item, leaf->len);
Y
Yan Zheng 已提交
520
	btrfs_set_root_last_snapshot(&root_item, 0);
C
Christoph Hellwig 已提交
521

522 523 524 525
	btrfs_set_root_generation_v2(&root_item,
			btrfs_root_generation(&root_item));
	uuid_le_gen(&new_uuid);
	memcpy(root_item.uuid, new_uuid.b, BTRFS_UUID_SIZE);
526 527
	btrfs_set_stack_timespec_sec(&root_item.otime, cur_time.tv_sec);
	btrfs_set_stack_timespec_nsec(&root_item.otime, cur_time.tv_nsec);
528 529 530
	root_item.ctime = root_item.otime;
	btrfs_set_root_ctransid(&root_item, trans->transid);
	btrfs_set_root_otransid(&root_item, trans->transid);
C
Christoph Hellwig 已提交
531

532
	btrfs_tree_unlock(leaf);
C
Christoph Hellwig 已提交
533 534 535 536 537 538
	free_extent_buffer(leaf);
	leaf = NULL;

	btrfs_set_root_dirid(&root_item, new_dirid);

	key.objectid = objectid;
539
	key.offset = 0;
540
	key.type = BTRFS_ROOT_ITEM_KEY;
C
Christoph Hellwig 已提交
541 542 543 544 545
	ret = btrfs_insert_root(trans, root->fs_info->tree_root, &key,
				&root_item);
	if (ret)
		goto fail;

546 547
	key.offset = (u64)-1;
	new_root = btrfs_read_fs_root_no_name(root->fs_info, &key);
548 549 550 551 552
	if (IS_ERR(new_root)) {
		btrfs_abort_transaction(trans, root, PTR_ERR(new_root));
		ret = PTR_ERR(new_root);
		goto fail;
	}
553 554 555

	btrfs_record_root_in_trans(trans, new_root);

556
	ret = btrfs_create_subvol_root(trans, new_root, root, new_dirid);
557 558
	if (ret) {
		/* We potentially lose an unused inode item here */
559
		btrfs_abort_transaction(trans, root, ret);
560 561 562
		goto fail;
	}

C
Christoph Hellwig 已提交
563 564 565
	/*
	 * insert the directory item
	 */
566
	ret = btrfs_set_inode_index(dir, &index);
567 568 569 570
	if (ret) {
		btrfs_abort_transaction(trans, root, ret);
		goto fail;
	}
571 572

	ret = btrfs_insert_dir_item(trans, root,
573
				    name, namelen, dir, &key,
574
				    BTRFS_FT_DIR, index);
575 576
	if (ret) {
		btrfs_abort_transaction(trans, root, ret);
C
Christoph Hellwig 已提交
577
		goto fail;
578
	}
579

580 581 582 583
	btrfs_i_size_write(dir, dir->i_size + namelen * 2);
	ret = btrfs_update_inode(trans, root, dir);
	BUG_ON(ret);

584
	ret = btrfs_add_root_ref(trans, root->fs_info->tree_root,
585
				 objectid, root->root_key.objectid,
L
Li Zefan 已提交
586
				 btrfs_ino(dir), index, name, namelen);
587
	BUG_ON(ret);
C
Christoph Hellwig 已提交
588

589 590 591 592 593 594
	ret = btrfs_uuid_tree_add(trans, root->fs_info->uuid_root,
				  root_item.uuid, BTRFS_UUID_KEY_SUBVOL,
				  objectid);
	if (ret)
		btrfs_abort_transaction(trans, root, ret);

C
Christoph Hellwig 已提交
595
fail:
596 597
	trans->block_rsv = NULL;
	trans->bytes_reserved = 0;
598 599
	btrfs_subvolume_release_metadata(root, &block_rsv, qgroup_reserved);

S
Sage Weil 已提交
600 601 602
	if (async_transid) {
		*async_transid = trans->transid;
		err = btrfs_commit_transaction_async(trans, root, 1);
603 604
		if (err)
			err = btrfs_commit_transaction(trans, root);
S
Sage Weil 已提交
605 606 607
	} else {
		err = btrfs_commit_transaction(trans, root);
	}
C
Christoph Hellwig 已提交
608 609
	if (err && !ret)
		ret = err;
610

611 612
	if (!ret) {
		inode = btrfs_lookup_dentry(dir, dentry);
613 614
		if (IS_ERR(inode))
			return PTR_ERR(inode);
615 616
		d_instantiate(dentry, inode);
	}
C
Christoph Hellwig 已提交
617 618 619
	return ret;
}

620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636
static void btrfs_wait_nocow_write(struct btrfs_root *root)
{
	s64 writers;
	DEFINE_WAIT(wait);

	do {
		prepare_to_wait(&root->subv_writers->wait, &wait,
				TASK_UNINTERRUPTIBLE);

		writers = percpu_counter_sum(&root->subv_writers->counter);
		if (writers)
			schedule();

		finish_wait(&root->subv_writers->wait, &wait);
	} while (writers);
}

637 638 639 640
static int create_snapshot(struct btrfs_root *root, struct inode *dir,
			   struct dentry *dentry, char *name, int namelen,
			   u64 *async_transid, bool readonly,
			   struct btrfs_qgroup_inherit *inherit)
C
Christoph Hellwig 已提交
641
{
642
	struct inode *inode;
C
Christoph Hellwig 已提交
643 644
	struct btrfs_pending_snapshot *pending_snapshot;
	struct btrfs_trans_handle *trans;
645
	int ret;
C
Christoph Hellwig 已提交
646

647
	if (!test_bit(BTRFS_ROOT_REF_COWS, &root->state))
C
Christoph Hellwig 已提交
648 649
		return -EINVAL;

650
	atomic_inc(&root->will_be_snapshoted);
651
	smp_mb__after_atomic();
652 653
	btrfs_wait_nocow_write(root);

654 655
	ret = btrfs_start_delalloc_inodes(root, 0);
	if (ret)
656
		goto out;
657

658
	btrfs_wait_ordered_extents(root, -1);
659

660
	pending_snapshot = kzalloc(sizeof(*pending_snapshot), GFP_NOFS);
661 662 663 664
	if (!pending_snapshot) {
		ret = -ENOMEM;
		goto out;
	}
665

666 667
	btrfs_init_block_rsv(&pending_snapshot->block_rsv,
			     BTRFS_BLOCK_RSV_TEMP);
668 669 670 671 672 673
	/*
	 * 1 - parent dir inode
	 * 2 - dir entries
	 * 1 - root item
	 * 2 - root ref/backref
	 * 1 - root of snapshot
674
	 * 1 - UUID item
675 676
	 */
	ret = btrfs_subvolume_reserve_metadata(BTRFS_I(dir)->root,
677
					&pending_snapshot->block_rsv, 8,
678 679
					&pending_snapshot->qgroup_reserved,
					false);
680
	if (ret)
681
		goto free;
682

683
	pending_snapshot->dentry = dentry;
C
Christoph Hellwig 已提交
684
	pending_snapshot->root = root;
L
Li Zefan 已提交
685
	pending_snapshot->readonly = readonly;
686
	pending_snapshot->dir = dir;
687
	pending_snapshot->inherit = inherit;
688

689
	trans = btrfs_start_transaction(root, 0);
690 691 692 693 694
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto fail;
	}

695
	spin_lock(&root->fs_info->trans_lock);
C
Christoph Hellwig 已提交
696 697
	list_add(&pending_snapshot->list,
		 &trans->transaction->pending_snapshots);
698
	spin_unlock(&root->fs_info->trans_lock);
S
Sage Weil 已提交
699 700 701 702
	if (async_transid) {
		*async_transid = trans->transid;
		ret = btrfs_commit_transaction_async(trans,
				     root->fs_info->extent_root, 1);
703 704
		if (ret)
			ret = btrfs_commit_transaction(trans, root);
S
Sage Weil 已提交
705 706 707 708
	} else {
		ret = btrfs_commit_transaction(trans,
					       root->fs_info->extent_root);
	}
709
	if (ret)
710
		goto fail;
711 712 713 714 715

	ret = pending_snapshot->error;
	if (ret)
		goto fail;

716
	inode = btrfs_lookup_dentry(dentry->d_parent->d_inode, dentry);
717 718 719 720
	if (IS_ERR(inode)) {
		ret = PTR_ERR(inode);
		goto fail;
	}
721

722 723 724
	d_instantiate(dentry, inode);
	ret = 0;
fail:
725 726 727
	btrfs_subvolume_release_metadata(BTRFS_I(dir)->root,
					 &pending_snapshot->block_rsv,
					 pending_snapshot->qgroup_reserved);
728
free:
729
	kfree(pending_snapshot);
730 731
out:
	atomic_dec(&root->will_be_snapshoted);
C
Christoph Hellwig 已提交
732 733 734
	return ret;
}

735 736 737 738 739 740
/*  copy of check_sticky in fs/namei.c()
* It's inline, so penalty for filesystems that don't use sticky bit is
* minimal.
*/
static inline int btrfs_check_sticky(struct inode *dir, struct inode *inode)
{
741
	kuid_t fsuid = current_fsuid();
742 743 744

	if (!(dir->i_mode & S_ISVTX))
		return 0;
745
	if (uid_eq(inode->i_uid, fsuid))
746
		return 0;
747
	if (uid_eq(dir->i_uid, fsuid))
748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771
		return 0;
	return !capable(CAP_FOWNER);
}

/*  copy of may_delete in fs/namei.c()
 *	Check whether we can remove a link victim from directory dir, check
 *  whether the type of victim is right.
 *  1. We can't do it if dir is read-only (done in permission())
 *  2. We should have write and exec permissions on dir
 *  3. We can't remove anything from append-only dir
 *  4. We can't do anything with immutable dir (done in permission())
 *  5. If the sticky bit on dir is set we should either
 *	a. be owner of dir, or
 *	b. be owner of victim, or
 *	c. have CAP_FOWNER capability
 *  6. If the victim is append-only or immutable we can't do antyhing with
 *     links pointing to it.
 *  7. If we were asked to remove a directory and victim isn't one - ENOTDIR.
 *  8. If we were asked to remove a non-directory and victim isn't one - EISDIR.
 *  9. We can't remove a root or mountpoint.
 * 10. We don't allow removal of NFS sillyrenamed files; it's handled by
 *     nfs_async_unlink().
 */

772
static int btrfs_may_delete(struct inode *dir, struct dentry *victim, int isdir)
773 774 775 776 777 778 779
{
	int error;

	if (!victim->d_inode)
		return -ENOENT;

	BUG_ON(victim->d_parent->d_inode != dir);
780
	audit_inode_child(dir, victim, AUDIT_TYPE_CHILD_DELETE);
781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804

	error = inode_permission(dir, MAY_WRITE | MAY_EXEC);
	if (error)
		return error;
	if (IS_APPEND(dir))
		return -EPERM;
	if (btrfs_check_sticky(dir, victim->d_inode)||
		IS_APPEND(victim->d_inode)||
	    IS_IMMUTABLE(victim->d_inode) || IS_SWAPFILE(victim->d_inode))
		return -EPERM;
	if (isdir) {
		if (!S_ISDIR(victim->d_inode->i_mode))
			return -ENOTDIR;
		if (IS_ROOT(victim))
			return -EBUSY;
	} else if (S_ISDIR(victim->d_inode->i_mode))
		return -EISDIR;
	if (IS_DEADDIR(dir))
		return -ENOENT;
	if (victim->d_flags & DCACHE_NFSFS_RENAMED)
		return -EBUSY;
	return 0;
}

805 806 807 808 809 810 811 812 813 814 815 816 817 818 819
/* copy of may_create in fs/namei.c() */
static inline int btrfs_may_create(struct inode *dir, struct dentry *child)
{
	if (child->d_inode)
		return -EEXIST;
	if (IS_DEADDIR(dir))
		return -ENOENT;
	return inode_permission(dir, MAY_WRITE | MAY_EXEC);
}

/*
 * Create a new subvolume below @parent.  This is largely modeled after
 * sys_mkdirat and vfs_mkdir, but we only do a single component lookup
 * inside this filesystem so it's quite a bit simpler.
 */
820 821
static noinline int btrfs_mksubvol(struct path *parent,
				   char *name, int namelen,
S
Sage Weil 已提交
822
				   struct btrfs_root *snap_src,
A
Arne Jansen 已提交
823
				   u64 *async_transid, bool readonly,
824
				   struct btrfs_qgroup_inherit *inherit)
825
{
826
	struct inode *dir  = parent->dentry->d_inode;
827 828 829
	struct dentry *dentry;
	int error;

830 831 832
	error = mutex_lock_killable_nested(&dir->i_mutex, I_MUTEX_PARENT);
	if (error == -EINTR)
		return error;
833 834 835 836 837 838 839 840 841 842

	dentry = lookup_one_len(name, parent->dentry, namelen);
	error = PTR_ERR(dentry);
	if (IS_ERR(dentry))
		goto out_unlock;

	error = -EEXIST;
	if (dentry->d_inode)
		goto out_dput;

843
	error = btrfs_may_create(dir, dentry);
844
	if (error)
845
		goto out_dput;
846

C
Chris Mason 已提交
847 848 849 850 851 852 853 854 855 856
	/*
	 * even if this name doesn't exist, we may get hash collisions.
	 * check for them now when we can safely fail
	 */
	error = btrfs_check_dir_item_collision(BTRFS_I(dir)->root,
					       dir->i_ino, name,
					       namelen);
	if (error)
		goto out_dput;

857 858 859 860 861
	down_read(&BTRFS_I(dir)->root->fs_info->subvol_sem);

	if (btrfs_root_refs(&BTRFS_I(dir)->root->root_item) == 0)
		goto out_up_read;

862
	if (snap_src) {
863
		error = create_snapshot(snap_src, dir, dentry, name, namelen,
A
Arne Jansen 已提交
864
					async_transid, readonly, inherit);
865
	} else {
866 867
		error = create_subvol(dir, dentry, name, namelen,
				      async_transid, inherit);
868
	}
869 870 871 872
	if (!error)
		fsnotify_mkdir(dir, dentry);
out_up_read:
	up_read(&BTRFS_I(dir)->root->fs_info->subvol_sem);
873 874 875
out_dput:
	dput(dentry);
out_unlock:
876
	mutex_unlock(&dir->i_mutex);
877 878 879
	return error;
}

C
Chris Mason 已提交
880 881 882 883 884 885 886
/*
 * When we're defragging a range, we don't want to kick it off again
 * if it is really just waiting for delalloc to send it down.
 * If we find a nice big extent or delalloc range for the bytes in the
 * file you want to defrag, we return 0 to let you know to skip this
 * part of the file
 */
887
static int check_defrag_in_cache(struct inode *inode, u64 offset, u32 thresh)
C
Chris Mason 已提交
888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921
{
	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
	struct extent_map *em = NULL;
	struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
	u64 end;

	read_lock(&em_tree->lock);
	em = lookup_extent_mapping(em_tree, offset, PAGE_CACHE_SIZE);
	read_unlock(&em_tree->lock);

	if (em) {
		end = extent_map_end(em);
		free_extent_map(em);
		if (end - offset > thresh)
			return 0;
	}
	/* if we already have a nice delalloc here, just stop */
	thresh /= 2;
	end = count_range_bits(io_tree, &offset, offset + thresh,
			       thresh, EXTENT_DELALLOC, 1);
	if (end >= thresh)
		return 0;
	return 1;
}

/*
 * helper function to walk through a file and find extents
 * newer than a specific transid, and smaller than thresh.
 *
 * This is used by the defragging code to find new and small
 * extents
 */
static int find_new_extents(struct btrfs_root *root,
			    struct inode *inode, u64 newer_than,
922
			    u64 *off, u32 thresh)
C
Chris Mason 已提交
923 924 925 926 927 928 929
{
	struct btrfs_path *path;
	struct btrfs_key min_key;
	struct extent_buffer *leaf;
	struct btrfs_file_extent_item *extent;
	int type;
	int ret;
930
	u64 ino = btrfs_ino(inode);
C
Chris Mason 已提交
931 932 933 934 935

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

936
	min_key.objectid = ino;
C
Chris Mason 已提交
937 938 939
	min_key.type = BTRFS_EXTENT_DATA_KEY;
	min_key.offset = *off;

940
	while (1) {
941
		ret = btrfs_search_forward(root, &min_key, path, newer_than);
C
Chris Mason 已提交
942 943
		if (ret != 0)
			goto none;
944
process_slot:
945
		if (min_key.objectid != ino)
C
Chris Mason 已提交
946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962
			goto none;
		if (min_key.type != BTRFS_EXTENT_DATA_KEY)
			goto none;

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

		type = btrfs_file_extent_type(leaf, extent);
		if (type == BTRFS_FILE_EXTENT_REG &&
		    btrfs_file_extent_num_bytes(leaf, extent) < thresh &&
		    check_defrag_in_cache(inode, min_key.offset, thresh)) {
			*off = min_key.offset;
			btrfs_free_path(path);
			return 0;
		}

963 964 965 966 967 968
		path->slots[0]++;
		if (path->slots[0] < btrfs_header_nritems(leaf)) {
			btrfs_item_key_to_cpu(leaf, &min_key, path->slots[0]);
			goto process_slot;
		}

C
Chris Mason 已提交
969 970 971 972 973 974 975 976 977 978 979
		if (min_key.offset == (u64)-1)
			goto none;

		min_key.offset++;
		btrfs_release_path(path);
	}
none:
	btrfs_free_path(path);
	return -ENOENT;
}

L
Li Zefan 已提交
980
static struct extent_map *defrag_lookup_extent(struct inode *inode, u64 start)
981 982
{
	struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
L
Li Zefan 已提交
983 984 985
	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
	struct extent_map *em;
	u64 len = PAGE_CACHE_SIZE;
986

L
Li Zefan 已提交
987 988 989 990
	/*
	 * hopefully we have this extent in the tree already, try without
	 * the full extent lock
	 */
991
	read_lock(&em_tree->lock);
L
Li Zefan 已提交
992
	em = lookup_extent_mapping(em_tree, start, len);
993 994
	read_unlock(&em_tree->lock);

L
Li Zefan 已提交
995
	if (!em) {
996 997 998
		struct extent_state *cached = NULL;
		u64 end = start + len - 1;

L
Li Zefan 已提交
999
		/* get the big lock and read metadata off disk */
1000
		lock_extent_bits(io_tree, start, end, 0, &cached);
L
Li Zefan 已提交
1001
		em = btrfs_get_extent(inode, NULL, 0, start, len, 0);
1002
		unlock_extent_cached(io_tree, start, end, &cached, GFP_NOFS);
L
Li Zefan 已提交
1003 1004 1005 1006 1007 1008 1009

		if (IS_ERR(em))
			return NULL;
	}

	return em;
}
1010

L
Li Zefan 已提交
1011 1012 1013 1014 1015 1016 1017 1018 1019 1020
static bool defrag_check_next_extent(struct inode *inode, struct extent_map *em)
{
	struct extent_map *next;
	bool ret = true;

	/* this is the last extent */
	if (em->start + em->len >= i_size_read(inode))
		return false;

	next = defrag_lookup_extent(inode, em->start + em->len);
1021 1022 1023 1024
	if (!next || next->block_start >= EXTENT_MAP_LAST_BYTE)
		ret = false;
	else if ((em->block_start + em->block_len == next->block_start) &&
		 (em->block_len > 128 * 1024 && next->block_len > 128 * 1024))
L
Li Zefan 已提交
1025 1026 1027
		ret = false;

	free_extent_map(next);
1028 1029 1030
	return ret;
}

1031
static int should_defrag_range(struct inode *inode, u64 start, u32 thresh,
1032 1033
			       u64 *last_len, u64 *skip, u64 *defrag_end,
			       int compress)
1034
{
L
Li Zefan 已提交
1035
	struct extent_map *em;
1036
	int ret = 1;
L
Li Zefan 已提交
1037
	bool next_mergeable = true;
1038 1039

	/*
1040
	 * make sure that once we start defragging an extent, we keep on
1041 1042 1043 1044 1045 1046 1047
	 * defragging it
	 */
	if (start < *defrag_end)
		return 1;

	*skip = 0;

L
Li Zefan 已提交
1048 1049 1050
	em = defrag_lookup_extent(inode, start);
	if (!em)
		return 0;
1051 1052

	/* this will cover holes, and inline extents */
1053
	if (em->block_start >= EXTENT_MAP_LAST_BYTE) {
1054
		ret = 0;
1055 1056 1057
		goto out;
	}

L
Li Zefan 已提交
1058
	next_mergeable = defrag_check_next_extent(inode, em);
1059
	/*
L
Li Zefan 已提交
1060 1061
	 * we hit a real extent, if it is big or the next extent is not a
	 * real extent, don't bother defragging it
1062
	 */
1063
	if (!compress && (*last_len == 0 || *last_len >= thresh) &&
L
Li Zefan 已提交
1064
	    (em->len >= thresh || !next_mergeable))
1065
		ret = 0;
1066
out:
1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086
	/*
	 * last_len ends up being a counter of how many bytes we've defragged.
	 * every time we choose not to defrag an extent, we reset *last_len
	 * so that the next tiny extent will force a defrag.
	 *
	 * The end result of this is that tiny extents before a single big
	 * extent will force at least part of that big extent to be defragged.
	 */
	if (ret) {
		*defrag_end = extent_map_end(em);
	} else {
		*last_len = 0;
		*skip = extent_map_end(em);
		*defrag_end = 0;
	}

	free_extent_map(em);
	return ret;
}

C
Chris Mason 已提交
1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101
/*
 * it doesn't do much good to defrag one or two pages
 * at a time.  This pulls in a nice chunk of pages
 * to COW and defrag.
 *
 * It also makes sure the delalloc code has enough
 * dirty data to avoid making new small extents as part
 * of the defrag
 *
 * It's a good idea to start RA on this range
 * before calling this.
 */
static int cluster_pages_for_defrag(struct inode *inode,
				    struct page **pages,
				    unsigned long start_index,
1102
				    unsigned long num_pages)
C
Christoph Hellwig 已提交
1103
{
C
Chris Mason 已提交
1104 1105 1106 1107
	unsigned long file_end;
	u64 isize = i_size_read(inode);
	u64 page_start;
	u64 page_end;
1108
	u64 page_cnt;
C
Chris Mason 已提交
1109 1110 1111
	int ret;
	int i;
	int i_done;
1112
	struct btrfs_ordered_extent *ordered;
C
Chris Mason 已提交
1113
	struct extent_state *cached_state = NULL;
1114
	struct extent_io_tree *tree;
1115
	gfp_t mask = btrfs_alloc_write_mask(inode->i_mapping);
C
Chris Mason 已提交
1116 1117

	file_end = (isize - 1) >> PAGE_CACHE_SHIFT;
1118 1119 1120 1121
	if (!isize || start_index > file_end)
		return 0;

	page_cnt = min_t(u64, (u64)num_pages, (u64)file_end - start_index + 1);
C
Chris Mason 已提交
1122 1123

	ret = btrfs_delalloc_reserve_space(inode,
1124
					   page_cnt << PAGE_CACHE_SHIFT);
C
Chris Mason 已提交
1125 1126 1127
	if (ret)
		return ret;
	i_done = 0;
1128
	tree = &BTRFS_I(inode)->io_tree;
C
Chris Mason 已提交
1129 1130

	/* step one, lock all the pages */
1131
	for (i = 0; i < page_cnt; i++) {
C
Chris Mason 已提交
1132
		struct page *page;
1133
again:
1134
		page = find_or_create_page(inode->i_mapping,
1135
					   start_index + i, mask);
C
Chris Mason 已提交
1136 1137 1138
		if (!page)
			break;

1139 1140 1141
		page_start = page_offset(page);
		page_end = page_start + PAGE_CACHE_SIZE - 1;
		while (1) {
1142 1143
			lock_extent_bits(tree, page_start, page_end,
					 0, &cached_state);
1144 1145
			ordered = btrfs_lookup_ordered_extent(inode,
							      page_start);
1146 1147
			unlock_extent_cached(tree, page_start, page_end,
					     &cached_state, GFP_NOFS);
1148 1149 1150 1151 1152 1153 1154
			if (!ordered)
				break;

			unlock_page(page);
			btrfs_start_ordered_extent(inode, ordered, 1);
			btrfs_put_ordered_extent(ordered);
			lock_page(page);
1155 1156 1157 1158 1159 1160 1161 1162 1163
			/*
			 * we unlocked the page above, so we need check if
			 * it was released or not.
			 */
			if (page->mapping != inode->i_mapping) {
				unlock_page(page);
				page_cache_release(page);
				goto again;
			}
1164 1165
		}

C
Chris Mason 已提交
1166 1167 1168 1169 1170 1171 1172 1173 1174 1175
		if (!PageUptodate(page)) {
			btrfs_readpage(NULL, page);
			lock_page(page);
			if (!PageUptodate(page)) {
				unlock_page(page);
				page_cache_release(page);
				ret = -EIO;
				break;
			}
		}
1176 1177 1178 1179 1180 1181 1182

		if (page->mapping != inode->i_mapping) {
			unlock_page(page);
			page_cache_release(page);
			goto again;
		}

C
Chris Mason 已提交
1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202
		pages[i] = page;
		i_done++;
	}
	if (!i_done || ret)
		goto out;

	if (!(inode->i_sb->s_flags & MS_ACTIVE))
		goto out;

	/*
	 * so now we have a nice long stream of locked
	 * and up to date pages, lets wait on them
	 */
	for (i = 0; i < i_done; i++)
		wait_on_page_writeback(pages[i]);

	page_start = page_offset(pages[0]);
	page_end = page_offset(pages[i_done - 1]) + PAGE_CACHE_SIZE;

	lock_extent_bits(&BTRFS_I(inode)->io_tree,
1203
			 page_start, page_end - 1, 0, &cached_state);
C
Chris Mason 已提交
1204 1205
	clear_extent_bit(&BTRFS_I(inode)->io_tree, page_start,
			  page_end - 1, EXTENT_DIRTY | EXTENT_DELALLOC |
1206 1207
			  EXTENT_DO_ACCOUNTING | EXTENT_DEFRAG, 0, 0,
			  &cached_state, GFP_NOFS);
C
Chris Mason 已提交
1208

1209
	if (i_done != page_cnt) {
1210 1211 1212
		spin_lock(&BTRFS_I(inode)->lock);
		BTRFS_I(inode)->outstanding_extents++;
		spin_unlock(&BTRFS_I(inode)->lock);
C
Chris Mason 已提交
1213
		btrfs_delalloc_release_space(inode,
1214
				     (page_cnt - i_done) << PAGE_CACHE_SHIFT);
C
Chris Mason 已提交
1215 1216 1217
	}


1218 1219
	set_extent_defrag(&BTRFS_I(inode)->io_tree, page_start, page_end - 1,
			  &cached_state, GFP_NOFS);
C
Chris Mason 已提交
1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238

	unlock_extent_cached(&BTRFS_I(inode)->io_tree,
			     page_start, page_end - 1, &cached_state,
			     GFP_NOFS);

	for (i = 0; i < i_done; i++) {
		clear_page_dirty_for_io(pages[i]);
		ClearPageChecked(pages[i]);
		set_page_extent_mapped(pages[i]);
		set_page_dirty(pages[i]);
		unlock_page(pages[i]);
		page_cache_release(pages[i]);
	}
	return i_done;
out:
	for (i = 0; i < i_done; i++) {
		unlock_page(pages[i]);
		page_cache_release(pages[i]);
	}
1239
	btrfs_delalloc_release_space(inode, page_cnt << PAGE_CACHE_SHIFT);
C
Chris Mason 已提交
1240 1241 1242 1243 1244 1245 1246 1247 1248 1249
	return ret;

}

int btrfs_defrag_file(struct inode *inode, struct file *file,
		      struct btrfs_ioctl_defrag_range_args *range,
		      u64 newer_than, unsigned long max_to_defrag)
{
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct file_ra_state *ra = NULL;
C
Christoph Hellwig 已提交
1250
	unsigned long last_index;
1251
	u64 isize = i_size_read(inode);
1252 1253 1254
	u64 last_len = 0;
	u64 skip = 0;
	u64 defrag_end = 0;
C
Chris Mason 已提交
1255
	u64 newer_off = range->start;
C
Christoph Hellwig 已提交
1256
	unsigned long i;
1257
	unsigned long ra_index = 0;
C
Christoph Hellwig 已提交
1258
	int ret;
C
Chris Mason 已提交
1259
	int defrag_count = 0;
1260
	int compress_type = BTRFS_COMPRESS_ZLIB;
1261
	u32 extent_thresh = range->extent_thresh;
1262 1263
	unsigned long max_cluster = (256 * 1024) >> PAGE_CACHE_SHIFT;
	unsigned long cluster = max_cluster;
C
Chris Mason 已提交
1264 1265 1266
	u64 new_align = ~((u64)128 * 1024 - 1);
	struct page **pages = NULL;

1267 1268 1269 1270 1271
	if (isize == 0)
		return 0;

	if (range->start >= isize)
		return -EINVAL;
1272 1273 1274 1275 1276 1277 1278

	if (range->flags & BTRFS_DEFRAG_RANGE_COMPRESS) {
		if (range->compress_type > BTRFS_COMPRESS_TYPES)
			return -EINVAL;
		if (range->compress_type)
			compress_type = range->compress_type;
	}
C
Christoph Hellwig 已提交
1279

1280 1281
	if (extent_thresh == 0)
		extent_thresh = 256 * 1024;
1282

C
Chris Mason 已提交
1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295
	/*
	 * if we were not given a file, allocate a readahead
	 * context
	 */
	if (!file) {
		ra = kzalloc(sizeof(*ra), GFP_NOFS);
		if (!ra)
			return -ENOMEM;
		file_ra_state_init(ra, inode->i_mapping);
	} else {
		ra = &file->f_ra;
	}

1296
	pages = kmalloc_array(max_cluster, sizeof(struct page *),
C
Chris Mason 已提交
1297 1298 1299 1300 1301 1302 1303
			GFP_NOFS);
	if (!pages) {
		ret = -ENOMEM;
		goto out_ra;
	}

	/* find the last page to defrag */
C
Chris Mason 已提交
1304
	if (range->start + range->len > range->start) {
1305
		last_index = min_t(u64, isize - 1,
C
Chris Mason 已提交
1306 1307
			 range->start + range->len - 1) >> PAGE_CACHE_SHIFT;
	} else {
1308
		last_index = (isize - 1) >> PAGE_CACHE_SHIFT;
C
Chris Mason 已提交
1309 1310
	}

C
Chris Mason 已提交
1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326
	if (newer_than) {
		ret = find_new_extents(root, inode, newer_than,
				       &newer_off, 64 * 1024);
		if (!ret) {
			range->start = newer_off;
			/*
			 * we always align our defrag to help keep
			 * the extents in the file evenly spaced
			 */
			i = (newer_off & new_align) >> PAGE_CACHE_SHIFT;
		} else
			goto out_ra;
	} else {
		i = range->start >> PAGE_CACHE_SHIFT;
	}
	if (!max_to_defrag)
L
Liu Bo 已提交
1327
		max_to_defrag = last_index + 1;
C
Chris Mason 已提交
1328

L
Li Zefan 已提交
1329 1330 1331 1332 1333 1334 1335
	/*
	 * make writeback starts from i, so the defrag range can be
	 * written sequentially.
	 */
	if (i < inode->i_mapping->writeback_index)
		inode->i_mapping->writeback_index = i;

1336
	while (i <= last_index && defrag_count < max_to_defrag &&
1337
	       (i < DIV_ROUND_UP(i_size_read(inode), PAGE_CACHE_SIZE))) {
C
Chris Mason 已提交
1338 1339 1340 1341 1342 1343 1344
		/*
		 * make sure we stop running if someone unmounts
		 * the FS
		 */
		if (!(inode->i_sb->s_flags & MS_ACTIVE))
			break;

1345
		if (btrfs_defrag_cancelled(root->fs_info)) {
1346
			printk(KERN_DEBUG "BTRFS: defrag_file cancelled\n");
1347 1348 1349 1350
			ret = -EAGAIN;
			break;
		}

1351
		if (!should_defrag_range(inode, (u64)i << PAGE_CACHE_SHIFT,
L
Li Zefan 已提交
1352
					 extent_thresh, &last_len, &skip,
1353 1354
					 &defrag_end, range->flags &
					 BTRFS_DEFRAG_RANGE_COMPRESS)) {
1355 1356 1357 1358 1359
			unsigned long next;
			/*
			 * the should_defrag function tells us how much to skip
			 * bump our counter by the suggested amount
			 */
1360
			next = DIV_ROUND_UP(skip, PAGE_CACHE_SIZE);
1361 1362 1363
			i = max(i + 1, next);
			continue;
		}
1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378

		if (!newer_than) {
			cluster = (PAGE_CACHE_ALIGN(defrag_end) >>
				   PAGE_CACHE_SHIFT) - i;
			cluster = min(cluster, max_cluster);
		} else {
			cluster = max_cluster;
		}

		if (i + cluster > ra_index) {
			ra_index = max(i, ra_index);
			btrfs_force_ra(inode->i_mapping, ra, file, ra_index,
				       cluster);
			ra_index += max_cluster;
		}
1379

1380
		mutex_lock(&inode->i_mutex);
1381 1382
		if (range->flags & BTRFS_DEFRAG_RANGE_COMPRESS)
			BTRFS_I(inode)->force_compress = compress_type;
1383
		ret = cluster_pages_for_defrag(inode, pages, i, cluster);
1384 1385
		if (ret < 0) {
			mutex_unlock(&inode->i_mutex);
C
Chris Mason 已提交
1386
			goto out_ra;
1387
		}
C
Chris Mason 已提交
1388 1389

		defrag_count += ret;
1390
		balance_dirty_pages_ratelimited(inode->i_mapping);
1391
		mutex_unlock(&inode->i_mutex);
C
Chris Mason 已提交
1392 1393 1394 1395 1396

		if (newer_than) {
			if (newer_off == (u64)-1)
				break;

1397 1398 1399
			if (ret > 0)
				i += ret;

C
Chris Mason 已提交
1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410
			newer_off = max(newer_off + 1,
					(u64)i << PAGE_CACHE_SHIFT);

			ret = find_new_extents(root, inode,
					       newer_than, &newer_off,
					       64 * 1024);
			if (!ret) {
				range->start = newer_off;
				i = (newer_off & new_align) >> PAGE_CACHE_SHIFT;
			} else {
				break;
C
Christoph Hellwig 已提交
1411
			}
C
Chris Mason 已提交
1412
		} else {
1413
			if (ret > 0) {
L
Li Zefan 已提交
1414
				i += ret;
1415 1416
				last_len += ret << PAGE_CACHE_SHIFT;
			} else {
L
Li Zefan 已提交
1417
				i++;
1418 1419
				last_len = 0;
			}
C
Christoph Hellwig 已提交
1420 1421 1422
		}
	}

1423
	if ((range->flags & BTRFS_DEFRAG_RANGE_START_IO)) {
C
Chris Mason 已提交
1424
		filemap_flush(inode->i_mapping);
1425 1426 1427 1428
		if (test_bit(BTRFS_INODE_HAS_ASYNC_EXTENT,
			     &BTRFS_I(inode)->runtime_flags))
			filemap_flush(inode->i_mapping);
	}
C
Chris Mason 已提交
1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444

	if ((range->flags & BTRFS_DEFRAG_RANGE_COMPRESS)) {
		/* the filemap_flush will queue IO into the worker threads, but
		 * we have to make sure the IO is actually started and that
		 * ordered extents get created before we return
		 */
		atomic_inc(&root->fs_info->async_submit_draining);
		while (atomic_read(&root->fs_info->nr_async_submits) ||
		      atomic_read(&root->fs_info->async_delalloc_pages)) {
			wait_event(root->fs_info->async_submit_wait,
			   (atomic_read(&root->fs_info->nr_async_submits) == 0 &&
			    atomic_read(&root->fs_info->async_delalloc_pages) == 0));
		}
		atomic_dec(&root->fs_info->async_submit_draining);
	}

1445
	if (range->compress_type == BTRFS_COMPRESS_LZO) {
1446
		btrfs_set_fs_incompat(root->fs_info, COMPRESS_LZO);
1447 1448
	}

1449
	ret = defrag_count;
1450

C
Chris Mason 已提交
1451
out_ra:
1452 1453 1454 1455 1456
	if (range->flags & BTRFS_DEFRAG_RANGE_COMPRESS) {
		mutex_lock(&inode->i_mutex);
		BTRFS_I(inode)->force_compress = BTRFS_COMPRESS_NONE;
		mutex_unlock(&inode->i_mutex);
	}
C
Chris Mason 已提交
1457 1458 1459
	if (!file)
		kfree(ra);
	kfree(pages);
1460
	return ret;
C
Christoph Hellwig 已提交
1461 1462
}

1463
static noinline int btrfs_ioctl_resize(struct file *file,
1464
					void __user *arg)
C
Christoph Hellwig 已提交
1465 1466 1467 1468
{
	u64 new_size;
	u64 old_size;
	u64 devid = 1;
A
Al Viro 已提交
1469
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
C
Christoph Hellwig 已提交
1470 1471 1472 1473
	struct btrfs_ioctl_vol_args *vol_args;
	struct btrfs_trans_handle *trans;
	struct btrfs_device *device = NULL;
	char *sizestr;
1474
	char *retptr;
C
Christoph Hellwig 已提交
1475 1476 1477 1478
	char *devstr = NULL;
	int ret = 0;
	int mod = 0;

1479 1480 1481
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

1482 1483 1484 1485
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;

1486 1487
	if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
			1)) {
1488
		mnt_drop_write_file(file);
1489
		return BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
1490 1491
	}

1492
	mutex_lock(&root->fs_info->volume_mutex);
L
Li Zefan 已提交
1493
	vol_args = memdup_user(arg, sizeof(*vol_args));
1494 1495 1496 1497
	if (IS_ERR(vol_args)) {
		ret = PTR_ERR(vol_args);
		goto out;
	}
1498 1499

	vol_args->name[BTRFS_PATH_NAME_MAX] = '\0';
C
Christoph Hellwig 已提交
1500 1501 1502 1503 1504 1505 1506

	sizestr = vol_args->name;
	devstr = strchr(sizestr, ':');
	if (devstr) {
		sizestr = devstr + 1;
		*devstr = '\0';
		devstr = vol_args->name;
1507 1508 1509
		ret = kstrtoull(devstr, 10, &devid);
		if (ret)
			goto out_free;
1510 1511 1512 1513
		if (!devid) {
			ret = -EINVAL;
			goto out_free;
		}
1514
		btrfs_info(root->fs_info, "resizing devid %llu", devid);
C
Christoph Hellwig 已提交
1515
	}
M
Miao Xie 已提交
1516

1517
	device = btrfs_find_device(root->fs_info, devid, NULL, NULL);
C
Christoph Hellwig 已提交
1518
	if (!device) {
1519
		btrfs_info(root->fs_info, "resizer unable to find device %llu",
1520
		       devid);
1521
		ret = -ENODEV;
1522
		goto out_free;
C
Christoph Hellwig 已提交
1523
	}
M
Miao Xie 已提交
1524 1525

	if (!device->writeable) {
1526 1527
		btrfs_info(root->fs_info,
			   "resizer unable to apply on readonly device %llu",
1528
		       devid);
1529
		ret = -EPERM;
L
Liu Bo 已提交
1530 1531 1532
		goto out_free;
	}

C
Christoph Hellwig 已提交
1533 1534 1535 1536 1537 1538 1539 1540 1541 1542
	if (!strcmp(sizestr, "max"))
		new_size = device->bdev->bd_inode->i_size;
	else {
		if (sizestr[0] == '-') {
			mod = -1;
			sizestr++;
		} else if (sizestr[0] == '+') {
			mod = 1;
			sizestr++;
		}
1543 1544
		new_size = memparse(sizestr, &retptr);
		if (*retptr != '\0' || new_size == 0) {
C
Christoph Hellwig 已提交
1545
			ret = -EINVAL;
1546
			goto out_free;
C
Christoph Hellwig 已提交
1547 1548 1549
		}
	}

1550
	if (device->is_tgtdev_for_dev_replace) {
1551
		ret = -EPERM;
1552 1553 1554
		goto out_free;
	}

1555
	old_size = btrfs_device_get_total_bytes(device);
C
Christoph Hellwig 已提交
1556 1557 1558 1559

	if (mod < 0) {
		if (new_size > old_size) {
			ret = -EINVAL;
1560
			goto out_free;
C
Christoph Hellwig 已提交
1561 1562 1563
		}
		new_size = old_size - new_size;
	} else if (mod > 0) {
1564
		if (new_size > ULLONG_MAX - old_size) {
1565
			ret = -ERANGE;
1566 1567
			goto out_free;
		}
C
Christoph Hellwig 已提交
1568 1569 1570 1571 1572
		new_size = old_size + new_size;
	}

	if (new_size < 256 * 1024 * 1024) {
		ret = -EINVAL;
1573
		goto out_free;
C
Christoph Hellwig 已提交
1574 1575 1576
	}
	if (new_size > device->bdev->bd_inode->i_size) {
		ret = -EFBIG;
1577
		goto out_free;
C
Christoph Hellwig 已提交
1578 1579 1580 1581 1582
	}

	do_div(new_size, root->sectorsize);
	new_size *= root->sectorsize;

1583
	printk_in_rcu(KERN_INFO "BTRFS: new size for %s is %llu\n",
1584
		      rcu_str_deref(device->name), new_size);
C
Christoph Hellwig 已提交
1585 1586

	if (new_size > old_size) {
1587
		trans = btrfs_start_transaction(root, 0);
1588 1589
		if (IS_ERR(trans)) {
			ret = PTR_ERR(trans);
1590
			goto out_free;
1591
		}
C
Christoph Hellwig 已提交
1592 1593
		ret = btrfs_grow_device(trans, device, new_size);
		btrfs_commit_transaction(trans, root);
1594
	} else if (new_size < old_size) {
C
Christoph Hellwig 已提交
1595
		ret = btrfs_shrink_device(device, new_size);
1596
	} /* equal, nothing need to do */
C
Christoph Hellwig 已提交
1597

1598
out_free:
C
Christoph Hellwig 已提交
1599
	kfree(vol_args);
1600 1601
out:
	mutex_unlock(&root->fs_info->volume_mutex);
1602
	atomic_set(&root->fs_info->mutually_exclusive_operation_running, 0);
1603
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
1604 1605 1606
	return ret;
}

S
Sage Weil 已提交
1607
static noinline int btrfs_ioctl_snap_create_transid(struct file *file,
A
Arne Jansen 已提交
1608 1609
				char *name, unsigned long fd, int subvol,
				u64 *transid, bool readonly,
1610
				struct btrfs_qgroup_inherit *inherit)
C
Christoph Hellwig 已提交
1611 1612
{
	int namelen;
1613
	int ret = 0;
C
Christoph Hellwig 已提交
1614

1615 1616 1617 1618
	ret = mnt_want_write_file(file);
	if (ret)
		goto out;

S
Sage Weil 已提交
1619 1620
	namelen = strlen(name);
	if (strchr(name, '/')) {
C
Christoph Hellwig 已提交
1621
		ret = -EINVAL;
1622
		goto out_drop_write;
C
Christoph Hellwig 已提交
1623 1624
	}

1625 1626 1627
	if (name[0] == '.' &&
	   (namelen == 1 || (name[1] == '.' && namelen == 2))) {
		ret = -EEXIST;
1628
		goto out_drop_write;
1629 1630
	}

1631
	if (subvol) {
S
Sage Weil 已提交
1632
		ret = btrfs_mksubvol(&file->f_path, name, namelen,
A
Arne Jansen 已提交
1633
				     NULL, transid, readonly, inherit);
1634
	} else {
1635
		struct fd src = fdget(fd);
1636
		struct inode *src_inode;
1637
		if (!src.file) {
1638
			ret = -EINVAL;
1639
			goto out_drop_write;
1640 1641
		}

A
Al Viro 已提交
1642 1643
		src_inode = file_inode(src.file);
		if (src_inode->i_sb != file_inode(file)->i_sb) {
1644 1645
			btrfs_info(BTRFS_I(src_inode)->root->fs_info,
				   "Snapshot src from another FS");
1646
			ret = -EXDEV;
1647 1648 1649 1650 1651 1652
		} else if (!inode_owner_or_capable(src_inode)) {
			/*
			 * Subvolume creation is not restricted, but snapshots
			 * are limited to own subvolumes only
			 */
			ret = -EPERM;
1653 1654 1655 1656
		} else {
			ret = btrfs_mksubvol(&file->f_path, name, namelen,
					     BTRFS_I(src_inode)->root,
					     transid, readonly, inherit);
1657
		}
1658
		fdput(src);
1659
	}
1660 1661
out_drop_write:
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
1662
out:
S
Sage Weil 已提交
1663 1664 1665 1666
	return ret;
}

static noinline int btrfs_ioctl_snap_create(struct file *file,
1667
					    void __user *arg, int subvol)
S
Sage Weil 已提交
1668
{
1669
	struct btrfs_ioctl_vol_args *vol_args;
S
Sage Weil 已提交
1670 1671
	int ret;

1672 1673 1674 1675
	vol_args = memdup_user(arg, sizeof(*vol_args));
	if (IS_ERR(vol_args))
		return PTR_ERR(vol_args);
	vol_args->name[BTRFS_PATH_NAME_MAX] = '\0';
S
Sage Weil 已提交
1676

1677
	ret = btrfs_ioctl_snap_create_transid(file, vol_args->name,
L
Li Zefan 已提交
1678
					      vol_args->fd, subvol,
A
Arne Jansen 已提交
1679
					      NULL, false, NULL);
1680

1681 1682 1683
	kfree(vol_args);
	return ret;
}
1684

1685 1686 1687 1688 1689 1690 1691
static noinline int btrfs_ioctl_snap_create_v2(struct file *file,
					       void __user *arg, int subvol)
{
	struct btrfs_ioctl_vol_args_v2 *vol_args;
	int ret;
	u64 transid = 0;
	u64 *ptr = NULL;
L
Li Zefan 已提交
1692
	bool readonly = false;
A
Arne Jansen 已提交
1693
	struct btrfs_qgroup_inherit *inherit = NULL;
1694

1695 1696 1697 1698
	vol_args = memdup_user(arg, sizeof(*vol_args));
	if (IS_ERR(vol_args))
		return PTR_ERR(vol_args);
	vol_args->name[BTRFS_SUBVOL_NAME_MAX] = '\0';
1699

L
Li Zefan 已提交
1700
	if (vol_args->flags &
A
Arne Jansen 已提交
1701 1702
	    ~(BTRFS_SUBVOL_CREATE_ASYNC | BTRFS_SUBVOL_RDONLY |
	      BTRFS_SUBVOL_QGROUP_INHERIT)) {
L
Li Zefan 已提交
1703
		ret = -EOPNOTSUPP;
D
Dan Carpenter 已提交
1704
		goto free_args;
S
Sage Weil 已提交
1705
	}
1706 1707 1708

	if (vol_args->flags & BTRFS_SUBVOL_CREATE_ASYNC)
		ptr = &transid;
L
Li Zefan 已提交
1709 1710
	if (vol_args->flags & BTRFS_SUBVOL_RDONLY)
		readonly = true;
A
Arne Jansen 已提交
1711 1712 1713
	if (vol_args->flags & BTRFS_SUBVOL_QGROUP_INHERIT) {
		if (vol_args->size > PAGE_CACHE_SIZE) {
			ret = -EINVAL;
D
Dan Carpenter 已提交
1714
			goto free_args;
A
Arne Jansen 已提交
1715 1716 1717 1718
		}
		inherit = memdup_user(vol_args->qgroup_inherit, vol_args->size);
		if (IS_ERR(inherit)) {
			ret = PTR_ERR(inherit);
D
Dan Carpenter 已提交
1719
			goto free_args;
A
Arne Jansen 已提交
1720 1721
		}
	}
1722 1723

	ret = btrfs_ioctl_snap_create_transid(file, vol_args->name,
A
Arne Jansen 已提交
1724
					      vol_args->fd, subvol, ptr,
1725
					      readonly, inherit);
D
Dan Carpenter 已提交
1726 1727
	if (ret)
		goto free_inherit;
1728

D
Dan Carpenter 已提交
1729 1730 1731 1732
	if (ptr && copy_to_user(arg +
				offsetof(struct btrfs_ioctl_vol_args_v2,
					transid),
				ptr, sizeof(*ptr)))
1733
		ret = -EFAULT;
D
Dan Carpenter 已提交
1734 1735

free_inherit:
A
Arne Jansen 已提交
1736
	kfree(inherit);
D
Dan Carpenter 已提交
1737 1738
free_args:
	kfree(vol_args);
C
Christoph Hellwig 已提交
1739 1740 1741
	return ret;
}

1742 1743 1744
static noinline int btrfs_ioctl_subvol_getflags(struct file *file,
						void __user *arg)
{
A
Al Viro 已提交
1745
	struct inode *inode = file_inode(file);
1746 1747 1748 1749
	struct btrfs_root *root = BTRFS_I(inode)->root;
	int ret = 0;
	u64 flags = 0;

L
Li Zefan 已提交
1750
	if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID)
1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766
		return -EINVAL;

	down_read(&root->fs_info->subvol_sem);
	if (btrfs_root_readonly(root))
		flags |= BTRFS_SUBVOL_RDONLY;
	up_read(&root->fs_info->subvol_sem);

	if (copy_to_user(arg, &flags, sizeof(flags)))
		ret = -EFAULT;

	return ret;
}

static noinline int btrfs_ioctl_subvol_setflags(struct file *file,
					      void __user *arg)
{
A
Al Viro 已提交
1767
	struct inode *inode = file_inode(file);
1768 1769 1770 1771 1772 1773
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;
	u64 root_flags;
	u64 flags;
	int ret = 0;

1774 1775 1776
	if (!inode_owner_or_capable(inode))
		return -EPERM;

1777 1778 1779
	ret = mnt_want_write_file(file);
	if (ret)
		goto out;
1780

1781 1782 1783 1784
	if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID) {
		ret = -EINVAL;
		goto out_drop_write;
	}
1785

1786 1787 1788 1789
	if (copy_from_user(&flags, arg, sizeof(flags))) {
		ret = -EFAULT;
		goto out_drop_write;
	}
1790

1791 1792 1793 1794
	if (flags & BTRFS_SUBVOL_CREATE_ASYNC) {
		ret = -EINVAL;
		goto out_drop_write;
	}
1795

1796 1797 1798 1799
	if (flags & ~BTRFS_SUBVOL_RDONLY) {
		ret = -EOPNOTSUPP;
		goto out_drop_write;
	}
1800 1801 1802 1803 1804

	down_write(&root->fs_info->subvol_sem);

	/* nothing to do */
	if (!!(flags & BTRFS_SUBVOL_RDONLY) == btrfs_root_readonly(root))
1805
		goto out_drop_sem;
1806 1807

	root_flags = btrfs_root_flags(&root->root_item);
1808
	if (flags & BTRFS_SUBVOL_RDONLY) {
1809 1810
		btrfs_set_root_flags(&root->root_item,
				     root_flags | BTRFS_ROOT_SUBVOL_RDONLY);
1811 1812 1813 1814 1815 1816 1817 1818
	} else {
		/*
		 * Block RO -> RW transition if this subvolume is involved in
		 * send
		 */
		spin_lock(&root->root_item_lock);
		if (root->send_in_progress == 0) {
			btrfs_set_root_flags(&root->root_item,
1819
				     root_flags & ~BTRFS_ROOT_SUBVOL_RDONLY);
1820 1821 1822 1823 1824 1825 1826 1827 1828 1829
			spin_unlock(&root->root_item_lock);
		} else {
			spin_unlock(&root->root_item_lock);
			btrfs_warn(root->fs_info,
			"Attempt to set subvolume %llu read-write during send",
					root->root_key.objectid);
			ret = -EPERM;
			goto out_drop_sem;
		}
	}
1830 1831 1832 1833 1834 1835 1836

	trans = btrfs_start_transaction(root, 1);
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto out_reset;
	}

1837
	ret = btrfs_update_root(trans, root->fs_info->tree_root,
1838 1839 1840 1841 1842 1843
				&root->root_key, &root->root_item);

	btrfs_commit_transaction(trans, root);
out_reset:
	if (ret)
		btrfs_set_root_flags(&root->root_item, root_flags);
1844
out_drop_sem:
1845
	up_write(&root->fs_info->subvol_sem);
1846 1847 1848
out_drop_write:
	mnt_drop_write_file(file);
out:
1849 1850 1851
	return ret;
}

1852 1853 1854 1855 1856 1857
/*
 * helper to check if the subvolume references other subvolumes
 */
static noinline int may_destroy_subvol(struct btrfs_root *root)
{
	struct btrfs_path *path;
1858
	struct btrfs_dir_item *di;
1859
	struct btrfs_key key;
1860
	u64 dir_id;
1861 1862 1863 1864 1865 1866
	int ret;

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

1867 1868 1869 1870 1871 1872 1873
	/* Make sure this root isn't set as the default subvol */
	dir_id = btrfs_super_root_dir(root->fs_info->super_copy);
	di = btrfs_lookup_dir_item(NULL, root->fs_info->tree_root, path,
				   dir_id, "default", 7, 0);
	if (di && !IS_ERR(di)) {
		btrfs_dir_item_key_to_cpu(path->nodes[0], di, &key);
		if (key.objectid == root->root_key.objectid) {
1874 1875 1876
			ret = -EPERM;
			btrfs_err(root->fs_info, "deleting default subvolume "
				  "%llu is not allowed", key.objectid);
1877 1878 1879 1880 1881
			goto out;
		}
		btrfs_release_path(path);
	}

1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904
	key.objectid = root->root_key.objectid;
	key.type = BTRFS_ROOT_REF_KEY;
	key.offset = (u64)-1;

	ret = btrfs_search_slot(NULL, root->fs_info->tree_root,
				&key, path, 0, 0);
	if (ret < 0)
		goto out;
	BUG_ON(ret == 0);

	ret = 0;
	if (path->slots[0] > 0) {
		path->slots[0]--;
		btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]);
		if (key.objectid == root->root_key.objectid &&
		    key.type == BTRFS_ROOT_REF_KEY)
			ret = -ENOTEMPTY;
	}
out:
	btrfs_free_path(path);
	return ret;
}

1905 1906 1907
static noinline int key_in_sk(struct btrfs_key *key,
			      struct btrfs_ioctl_search_key *sk)
{
1908 1909 1910 1911 1912 1913 1914 1915 1916
	struct btrfs_key test;
	int ret;

	test.objectid = sk->min_objectid;
	test.type = sk->min_type;
	test.offset = sk->min_offset;

	ret = btrfs_comp_cpu_keys(key, &test);
	if (ret < 0)
1917
		return 0;
1918 1919 1920 1921 1922 1923 1924

	test.objectid = sk->max_objectid;
	test.type = sk->max_type;
	test.offset = sk->max_offset;

	ret = btrfs_comp_cpu_keys(key, &test);
	if (ret > 0)
1925 1926 1927 1928 1929 1930 1931 1932
		return 0;
	return 1;
}

static noinline int copy_to_sk(struct btrfs_root *root,
			       struct btrfs_path *path,
			       struct btrfs_key *key,
			       struct btrfs_ioctl_search_key *sk,
1933
			       size_t *buf_size,
1934
			       char __user *ubuf,
1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961
			       unsigned long *sk_offset,
			       int *num_found)
{
	u64 found_transid;
	struct extent_buffer *leaf;
	struct btrfs_ioctl_search_header sh;
	unsigned long item_off;
	unsigned long item_len;
	int nritems;
	int i;
	int slot;
	int ret = 0;

	leaf = path->nodes[0];
	slot = path->slots[0];
	nritems = btrfs_header_nritems(leaf);

	if (btrfs_header_generation(leaf) > sk->max_transid) {
		i = nritems;
		goto advance_key;
	}
	found_transid = btrfs_header_generation(leaf);

	for (i = slot; i < nritems; i++) {
		item_off = btrfs_item_ptr_offset(leaf, i);
		item_len = btrfs_item_size_nr(leaf, i);

1962 1963 1964 1965
		btrfs_item_key_to_cpu(leaf, key, i);
		if (!key_in_sk(key, sk))
			continue;

1966
		if (sizeof(sh) + item_len > *buf_size) {
1967 1968 1969 1970 1971 1972 1973 1974 1975 1976
			if (*num_found) {
				ret = 1;
				goto out;
			}

			/*
			 * return one empty item back for v1, which does not
			 * handle -EOVERFLOW
			 */

1977
			*buf_size = sizeof(sh) + item_len;
1978
			item_len = 0;
1979 1980
			ret = -EOVERFLOW;
		}
1981

1982
		if (sizeof(sh) + item_len + *sk_offset > *buf_size) {
1983
			ret = 1;
1984
			goto out;
1985 1986 1987 1988 1989 1990 1991 1992 1993
		}

		sh.objectid = key->objectid;
		sh.offset = key->offset;
		sh.type = key->type;
		sh.len = item_len;
		sh.transid = found_transid;

		/* copy search result header */
1994 1995 1996 1997 1998
		if (copy_to_user(ubuf + *sk_offset, &sh, sizeof(sh))) {
			ret = -EFAULT;
			goto out;
		}

1999 2000 2001
		*sk_offset += sizeof(sh);

		if (item_len) {
2002
			char __user *up = ubuf + *sk_offset;
2003
			/* copy the item */
2004 2005 2006 2007 2008 2009
			if (read_extent_buffer_to_user(leaf, up,
						       item_off, item_len)) {
				ret = -EFAULT;
				goto out;
			}

2010 2011
			*sk_offset += item_len;
		}
2012
		(*num_found)++;
2013

2014 2015 2016
		if (ret) /* -EOVERFLOW from above */
			goto out;

2017 2018 2019 2020
		if (*num_found >= sk->nr_items) {
			ret = 1;
			goto out;
		}
2021 2022
	}
advance_key:
2023 2024
	ret = 0;
	if (key->offset < (u64)-1 && key->offset < sk->max_offset)
2025
		key->offset++;
2026 2027
	else if (key->type < (u8)-1 && key->type < sk->max_type) {
		key->offset = 0;
2028
		key->type++;
2029 2030 2031
	} else if (key->objectid < (u64)-1 && key->objectid < sk->max_objectid) {
		key->offset = 0;
		key->type = 0;
2032
		key->objectid++;
2033 2034
	} else
		ret = 1;
2035
out:
2036 2037 2038 2039 2040 2041 2042 2043 2044
	/*
	 *  0: all items from this leaf copied, continue with next
	 *  1: * more items can be copied, but unused buffer is too small
	 *     * all items were found
	 *     Either way, it will stops the loop which iterates to the next
	 *     leaf
	 *  -EOVERFLOW: item was to large for buffer
	 *  -EFAULT: could not copy extent buffer back to userspace
	 */
2045 2046 2047 2048
	return ret;
}

static noinline int search_ioctl(struct inode *inode,
2049
				 struct btrfs_ioctl_search_key *sk,
2050
				 size_t *buf_size,
2051
				 char __user *ubuf)
2052 2053 2054 2055 2056 2057 2058 2059 2060
{
	struct btrfs_root *root;
	struct btrfs_key key;
	struct btrfs_path *path;
	struct btrfs_fs_info *info = BTRFS_I(inode)->root->fs_info;
	int ret;
	int num_found = 0;
	unsigned long sk_offset = 0;

2061 2062
	if (*buf_size < sizeof(struct btrfs_ioctl_search_header)) {
		*buf_size = sizeof(struct btrfs_ioctl_search_header);
2063
		return -EOVERFLOW;
2064
	}
2065

2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078
	path = btrfs_alloc_path();
	if (!path)
		return -ENOMEM;

	if (sk->tree_id == 0) {
		/* search the root of the inode that was passed */
		root = BTRFS_I(inode)->root;
	} else {
		key.objectid = sk->tree_id;
		key.type = BTRFS_ROOT_ITEM_KEY;
		key.offset = (u64)-1;
		root = btrfs_read_fs_root_no_name(info, &key);
		if (IS_ERR(root)) {
2079
			printk(KERN_ERR "BTRFS: could not find root %llu\n",
2080 2081 2082 2083 2084 2085 2086 2087 2088 2089
			       sk->tree_id);
			btrfs_free_path(path);
			return -ENOENT;
		}
	}

	key.objectid = sk->min_objectid;
	key.type = sk->min_type;
	key.offset = sk->min_offset;

2090
	while (1) {
2091
		ret = btrfs_search_forward(root, &key, path, sk->min_transid);
2092 2093 2094 2095 2096
		if (ret != 0) {
			if (ret > 0)
				ret = 0;
			goto err;
		}
2097
		ret = copy_to_sk(root, path, &key, sk, buf_size, ubuf,
2098
				 &sk_offset, &num_found);
2099
		btrfs_release_path(path);
2100
		if (ret)
2101 2102 2103
			break;

	}
2104 2105
	if (ret > 0)
		ret = 0;
2106 2107 2108 2109 2110 2111 2112 2113 2114
err:
	sk->nr_items = num_found;
	btrfs_free_path(path);
	return ret;
}

static noinline int btrfs_ioctl_tree_search(struct file *file,
					   void __user *argp)
{
2115 2116
	struct btrfs_ioctl_search_args __user *uargs;
	struct btrfs_ioctl_search_key sk;
2117 2118 2119
	struct inode *inode;
	int ret;
	size_t buf_size;
2120 2121 2122 2123

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

2124 2125 2126 2127
	uargs = (struct btrfs_ioctl_search_args __user *)argp;

	if (copy_from_user(&sk, &uargs->key, sizeof(sk)))
		return -EFAULT;
2128

2129
	buf_size = sizeof(uargs->buf);
2130

A
Al Viro 已提交
2131
	inode = file_inode(file);
2132
	ret = search_ioctl(inode, &sk, &buf_size, uargs->buf);
2133 2134 2135 2136 2137 2138 2139 2140

	/*
	 * In the origin implementation an overflow is handled by returning a
	 * search header with a len of zero, so reset ret.
	 */
	if (ret == -EOVERFLOW)
		ret = 0;

2141
	if (ret == 0 && copy_to_user(&uargs->key, &sk, sizeof(sk)))
2142 2143 2144 2145
		ret = -EFAULT;
	return ret;
}

G
Gerhard Heift 已提交
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 2176 2177 2178 2179 2180 2181
static noinline int btrfs_ioctl_tree_search_v2(struct file *file,
					       void __user *argp)
{
	struct btrfs_ioctl_search_args_v2 __user *uarg;
	struct btrfs_ioctl_search_args_v2 args;
	struct inode *inode;
	int ret;
	size_t buf_size;
	const size_t buf_limit = 16 * 1024 * 1024;

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

	/* copy search header and buffer size */
	uarg = (struct btrfs_ioctl_search_args_v2 __user *)argp;
	if (copy_from_user(&args, uarg, sizeof(args)))
		return -EFAULT;

	buf_size = args.buf_size;

	if (buf_size < sizeof(struct btrfs_ioctl_search_header))
		return -EOVERFLOW;

	/* limit result size to 16MB */
	if (buf_size > buf_limit)
		buf_size = buf_limit;

	inode = file_inode(file);
	ret = search_ioctl(inode, &args.key, &buf_size,
			   (char *)(&uarg->buf[0]));
	if (ret == 0 && copy_to_user(&uarg->key, &args.key, sizeof(args.key)))
		ret = -EFAULT;
	else if (ret == -EOVERFLOW &&
		copy_to_user(&uarg->buf_size, &buf_size, sizeof(buf_size)))
		ret = -EFAULT;

2182 2183 2184
	return ret;
}

2185
/*
2186 2187 2188
 * Search INODE_REFs to identify path name of 'dirid' directory
 * in a 'tree_id' tree. and sets path name to 'name'.
 */
2189 2190 2191 2192 2193
static noinline int btrfs_search_path_in_tree(struct btrfs_fs_info *info,
				u64 tree_id, u64 dirid, char *name)
{
	struct btrfs_root *root;
	struct btrfs_key key;
2194
	char *ptr;
2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211
	int ret = -1;
	int slot;
	int len;
	int total_len = 0;
	struct btrfs_inode_ref *iref;
	struct extent_buffer *l;
	struct btrfs_path *path;

	if (dirid == BTRFS_FIRST_FREE_OBJECTID) {
		name[0]='\0';
		return 0;
	}

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

2212
	ptr = &name[BTRFS_INO_LOOKUP_PATH_MAX];
2213 2214 2215 2216 2217 2218

	key.objectid = tree_id;
	key.type = BTRFS_ROOT_ITEM_KEY;
	key.offset = (u64)-1;
	root = btrfs_read_fs_root_no_name(info, &key);
	if (IS_ERR(root)) {
2219
		printk(KERN_ERR "BTRFS: could not find root %llu\n", tree_id);
2220 2221
		ret = -ENOENT;
		goto out;
2222 2223 2224 2225
	}

	key.objectid = dirid;
	key.type = BTRFS_INODE_REF_KEY;
2226
	key.offset = (u64)-1;
2227

2228
	while (1) {
2229 2230 2231
		ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
		if (ret < 0)
			goto out;
2232 2233 2234 2235 2236 2237 2238 2239 2240 2241
		else if (ret > 0) {
			ret = btrfs_previous_item(root, path, dirid,
						  BTRFS_INODE_REF_KEY);
			if (ret < 0)
				goto out;
			else if (ret > 0) {
				ret = -ENOENT;
				goto out;
			}
		}
2242 2243 2244 2245 2246 2247 2248 2249 2250

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

		iref = btrfs_item_ptr(l, slot, struct btrfs_inode_ref);
		len = btrfs_inode_ref_name_len(l, iref);
		ptr -= len + 1;
		total_len += len + 1;
2251 2252
		if (ptr < name) {
			ret = -ENAMETOOLONG;
2253
			goto out;
2254
		}
2255 2256

		*(ptr + len) = '/';
2257
		read_extent_buffer(l, ptr, (unsigned long)(iref + 1), len);
2258 2259 2260 2261

		if (key.offset == BTRFS_FIRST_FREE_OBJECTID)
			break;

2262
		btrfs_release_path(path);
2263
		key.objectid = key.offset;
2264
		key.offset = (u64)-1;
2265 2266
		dirid = key.objectid;
	}
2267
	memmove(name, ptr, total_len);
2268
	name[total_len] = '\0';
2269 2270 2271
	ret = 0;
out:
	btrfs_free_path(path);
2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284
	return ret;
}

static noinline int btrfs_ioctl_ino_lookup(struct file *file,
					   void __user *argp)
{
	 struct btrfs_ioctl_ino_lookup_args *args;
	 struct inode *inode;
	 int ret;

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

J
Julia Lawall 已提交
2285 2286 2287
	args = memdup_user(argp, sizeof(*args));
	if (IS_ERR(args))
		return PTR_ERR(args);
2288

A
Al Viro 已提交
2289
	inode = file_inode(file);
2290

2291 2292 2293
	if (args->treeid == 0)
		args->treeid = BTRFS_I(inode)->root->root_key.objectid;

2294 2295 2296 2297 2298 2299 2300 2301
	ret = btrfs_search_path_in_tree(BTRFS_I(inode)->root->fs_info,
					args->treeid, args->objectid,
					args->name);

	if (ret == 0 && copy_to_user(argp, args, sizeof(*args)))
		ret = -EFAULT;

	kfree(args);
2302 2303 2304
	return ret;
}

2305 2306 2307
static noinline int btrfs_ioctl_snap_destroy(struct file *file,
					     void __user *arg)
{
A
Al Viro 已提交
2308
	struct dentry *parent = file->f_path.dentry;
2309 2310 2311 2312 2313 2314 2315
	struct dentry *dentry;
	struct inode *dir = parent->d_inode;
	struct inode *inode;
	struct btrfs_root *root = BTRFS_I(dir)->root;
	struct btrfs_root *dest = NULL;
	struct btrfs_ioctl_vol_args *vol_args;
	struct btrfs_trans_handle *trans;
2316
	struct btrfs_block_rsv block_rsv;
2317
	u64 root_flags;
2318
	u64 qgroup_reserved;
2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334
	int namelen;
	int ret;
	int err = 0;

	vol_args = memdup_user(arg, sizeof(*vol_args));
	if (IS_ERR(vol_args))
		return PTR_ERR(vol_args);

	vol_args->name[BTRFS_PATH_NAME_MAX] = '\0';
	namelen = strlen(vol_args->name);
	if (strchr(vol_args->name, '/') ||
	    strncmp(vol_args->name, "..", namelen) == 0) {
		err = -EINVAL;
		goto out;
	}

2335
	err = mnt_want_write_file(file);
2336 2337 2338
	if (err)
		goto out;

2339

2340 2341
	err = mutex_lock_killable_nested(&dir->i_mutex, I_MUTEX_PARENT);
	if (err == -EINTR)
2342
		goto out_drop_write;
2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354
	dentry = lookup_one_len(vol_args->name, parent, namelen);
	if (IS_ERR(dentry)) {
		err = PTR_ERR(dentry);
		goto out_unlock_dir;
	}

	if (!dentry->d_inode) {
		err = -ENOENT;
		goto out_dput;
	}

	inode = dentry->d_inode;
2355
	dest = BTRFS_I(inode)->root;
2356
	if (!capable(CAP_SYS_ADMIN)) {
2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389
		/*
		 * Regular user.  Only allow this with a special mount
		 * option, when the user has write+exec access to the
		 * subvol root, and when rmdir(2) would have been
		 * allowed.
		 *
		 * Note that this is _not_ check that the subvol is
		 * empty or doesn't contain data that we wouldn't
		 * otherwise be able to delete.
		 *
		 * Users who want to delete empty subvols should try
		 * rmdir(2).
		 */
		err = -EPERM;
		if (!btrfs_test_opt(root, USER_SUBVOL_RM_ALLOWED))
			goto out_dput;

		/*
		 * Do not allow deletion if the parent dir is the same
		 * as the dir to be deleted.  That means the ioctl
		 * must be called on the dentry referencing the root
		 * of the subvol, not a random directory contained
		 * within it.
		 */
		err = -EINVAL;
		if (root == dest)
			goto out_dput;

		err = inode_permission(inode, MAY_WRITE | MAY_EXEC);
		if (err)
			goto out_dput;
	}

2390 2391 2392 2393 2394
	/* check if subvolume may be deleted by a user */
	err = btrfs_may_delete(dir, dentry, 1);
	if (err)
		goto out_dput;

L
Li Zefan 已提交
2395
	if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID) {
2396 2397 2398 2399 2400
		err = -EINVAL;
		goto out_dput;
	}

	mutex_lock(&inode->i_mutex);
2401 2402 2403 2404 2405 2406 2407

	/*
	 * Don't allow to delete a subvolume with send in progress. This is
	 * inside the i_mutex so the error handling that has to drop the bit
	 * again is not run concurrently.
	 */
	spin_lock(&dest->root_item_lock);
2408 2409 2410
	root_flags = btrfs_root_flags(&dest->root_item);
	if (dest->send_in_progress == 0) {
		btrfs_set_root_flags(&dest->root_item,
2411 2412 2413 2414 2415 2416
				root_flags | BTRFS_ROOT_SUBVOL_DEAD);
		spin_unlock(&dest->root_item_lock);
	} else {
		spin_unlock(&dest->root_item_lock);
		btrfs_warn(root->fs_info,
			"Attempt to delete subvolume %llu during send",
2417
			dest->root_key.objectid);
2418 2419 2420 2421
		err = -EPERM;
		goto out_dput;
	}

2422 2423 2424 2425 2426 2427 2428 2429 2430 2431
	err = d_invalidate(dentry);
	if (err)
		goto out_unlock;

	down_write(&root->fs_info->subvol_sem);

	err = may_destroy_subvol(dest);
	if (err)
		goto out_up_write;

2432 2433 2434 2435 2436 2437
	btrfs_init_block_rsv(&block_rsv, BTRFS_BLOCK_RSV_TEMP);
	/*
	 * One for dir inode, two for dir entries, two for root
	 * ref/backref.
	 */
	err = btrfs_subvolume_reserve_metadata(root, &block_rsv,
2438
					       5, &qgroup_reserved, true);
2439 2440 2441
	if (err)
		goto out_up_write;

2442 2443 2444
	trans = btrfs_start_transaction(root, 0);
	if (IS_ERR(trans)) {
		err = PTR_ERR(trans);
2445
		goto out_release;
2446
	}
2447 2448
	trans->block_rsv = &block_rsv;
	trans->bytes_reserved = block_rsv.size;
2449

2450 2451 2452 2453
	ret = btrfs_unlink_subvol(trans, root, dir,
				dest->root_key.objectid,
				dentry->d_name.name,
				dentry->d_name.len);
2454 2455 2456 2457 2458
	if (ret) {
		err = ret;
		btrfs_abort_transaction(trans, root, ret);
		goto out_end_trans;
	}
2459 2460 2461 2462 2463 2464 2465 2466

	btrfs_record_root_in_trans(trans, dest);

	memset(&dest->root_item.drop_progress, 0,
		sizeof(dest->root_item.drop_progress));
	dest->root_item.drop_level = 0;
	btrfs_set_root_refs(&dest->root_item, 0);

2467
	if (!test_and_set_bit(BTRFS_ROOT_ORPHAN_ITEM_INSERTED, &dest->state)) {
2468 2469 2470
		ret = btrfs_insert_orphan_item(trans,
					root->fs_info->tree_root,
					dest->root_key.objectid);
2471 2472 2473 2474 2475
		if (ret) {
			btrfs_abort_transaction(trans, root, ret);
			err = ret;
			goto out_end_trans;
		}
2476
	}
2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497

	ret = btrfs_uuid_tree_rem(trans, root->fs_info->uuid_root,
				  dest->root_item.uuid, BTRFS_UUID_KEY_SUBVOL,
				  dest->root_key.objectid);
	if (ret && ret != -ENOENT) {
		btrfs_abort_transaction(trans, root, ret);
		err = ret;
		goto out_end_trans;
	}
	if (!btrfs_is_empty_uuid(dest->root_item.received_uuid)) {
		ret = btrfs_uuid_tree_rem(trans, root->fs_info->uuid_root,
					  dest->root_item.received_uuid,
					  BTRFS_UUID_KEY_RECEIVED_SUBVOL,
					  dest->root_key.objectid);
		if (ret && ret != -ENOENT) {
			btrfs_abort_transaction(trans, root, ret);
			err = ret;
			goto out_end_trans;
		}
	}

2498
out_end_trans:
2499 2500
	trans->block_rsv = NULL;
	trans->bytes_reserved = 0;
S
Sage Weil 已提交
2501
	ret = btrfs_end_transaction(trans, root);
2502 2503
	if (ret && !err)
		err = ret;
2504
	inode->i_flags |= S_DEAD;
2505 2506
out_release:
	btrfs_subvolume_release_metadata(root, &block_rsv, qgroup_reserved);
2507 2508 2509
out_up_write:
	up_write(&root->fs_info->subvol_sem);
out_unlock:
2510 2511
	if (err) {
		spin_lock(&dest->root_item_lock);
2512 2513
		root_flags = btrfs_root_flags(&dest->root_item);
		btrfs_set_root_flags(&dest->root_item,
2514 2515 2516
				root_flags & ~BTRFS_ROOT_SUBVOL_DEAD);
		spin_unlock(&dest->root_item_lock);
	}
2517 2518
	mutex_unlock(&inode->i_mutex);
	if (!err) {
2519
		shrink_dcache_sb(root->fs_info->sb);
2520 2521
		btrfs_invalidate_inodes(dest);
		d_delete(dentry);
2522
		ASSERT(dest->send_in_progress == 0);
2523 2524

		/* the last ref */
2525 2526 2527
		if (dest->ino_cache_inode) {
			iput(dest->ino_cache_inode);
			dest->ino_cache_inode = NULL;
2528
		}
2529 2530 2531 2532 2533
	}
out_dput:
	dput(dentry);
out_unlock_dir:
	mutex_unlock(&dir->i_mutex);
2534
out_drop_write:
A
Al Viro 已提交
2535
	mnt_drop_write_file(file);
2536 2537 2538 2539 2540
out:
	kfree(vol_args);
	return err;
}

C
Chris Mason 已提交
2541
static int btrfs_ioctl_defrag(struct file *file, void __user *argp)
C
Christoph Hellwig 已提交
2542
{
A
Al Viro 已提交
2543
	struct inode *inode = file_inode(file);
C
Christoph Hellwig 已提交
2544
	struct btrfs_root *root = BTRFS_I(inode)->root;
C
Chris Mason 已提交
2545
	struct btrfs_ioctl_defrag_range_args *range;
Y
Yan Zheng 已提交
2546 2547
	int ret;

2548 2549 2550
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
L
Li Zefan 已提交
2551

2552 2553 2554
	if (btrfs_root_readonly(root)) {
		ret = -EROFS;
		goto out;
2555
	}
C
Christoph Hellwig 已提交
2556 2557 2558

	switch (inode->i_mode & S_IFMT) {
	case S_IFDIR:
2559 2560 2561 2562
		if (!capable(CAP_SYS_ADMIN)) {
			ret = -EPERM;
			goto out;
		}
2563
		ret = btrfs_defrag_root(root);
2564 2565
		if (ret)
			goto out;
2566
		ret = btrfs_defrag_root(root->fs_info->extent_root);
C
Christoph Hellwig 已提交
2567 2568
		break;
	case S_IFREG:
2569 2570 2571 2572
		if (!(file->f_mode & FMODE_WRITE)) {
			ret = -EINVAL;
			goto out;
		}
C
Chris Mason 已提交
2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584

		range = kzalloc(sizeof(*range), GFP_KERNEL);
		if (!range) {
			ret = -ENOMEM;
			goto out;
		}

		if (argp) {
			if (copy_from_user(range, argp,
					   sizeof(*range))) {
				ret = -EFAULT;
				kfree(range);
2585
				goto out;
C
Chris Mason 已提交
2586 2587 2588 2589 2590 2591 2592 2593 2594 2595
			}
			/* compression requires us to start the IO */
			if ((range->flags & BTRFS_DEFRAG_RANGE_COMPRESS)) {
				range->flags |= BTRFS_DEFRAG_RANGE_START_IO;
				range->extent_thresh = (u32)-1;
			}
		} else {
			/* the rest are all set to zero by kzalloc */
			range->len = (u64)-1;
		}
A
Al Viro 已提交
2596
		ret = btrfs_defrag_file(file_inode(file), file,
C
Chris Mason 已提交
2597 2598 2599
					range, 0, 0);
		if (ret > 0)
			ret = 0;
C
Chris Mason 已提交
2600
		kfree(range);
C
Christoph Hellwig 已提交
2601
		break;
2602 2603
	default:
		ret = -EINVAL;
C
Christoph Hellwig 已提交
2604
	}
2605
out:
2606
	mnt_drop_write_file(file);
2607
	return ret;
C
Christoph Hellwig 已提交
2608 2609
}

2610
static long btrfs_ioctl_add_dev(struct btrfs_root *root, void __user *arg)
C
Christoph Hellwig 已提交
2611 2612 2613 2614
{
	struct btrfs_ioctl_vol_args *vol_args;
	int ret;

2615 2616 2617
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

2618 2619
	if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
			1)) {
2620
		return BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
2621 2622
	}

2623
	mutex_lock(&root->fs_info->volume_mutex);
L
Li Zefan 已提交
2624
	vol_args = memdup_user(arg, sizeof(*vol_args));
2625 2626 2627 2628
	if (IS_ERR(vol_args)) {
		ret = PTR_ERR(vol_args);
		goto out;
	}
C
Christoph Hellwig 已提交
2629

2630
	vol_args->name[BTRFS_PATH_NAME_MAX] = '\0';
C
Christoph Hellwig 已提交
2631 2632
	ret = btrfs_init_new_device(root, vol_args->name);

A
Anand Jain 已提交
2633 2634 2635
	if (!ret)
		btrfs_info(root->fs_info, "disk added %s",vol_args->name);

C
Christoph Hellwig 已提交
2636
	kfree(vol_args);
2637 2638
out:
	mutex_unlock(&root->fs_info->volume_mutex);
2639
	atomic_set(&root->fs_info->mutually_exclusive_operation_running, 0);
C
Christoph Hellwig 已提交
2640 2641 2642
	return ret;
}

2643
static long btrfs_ioctl_rm_dev(struct file *file, void __user *arg)
C
Christoph Hellwig 已提交
2644
{
A
Al Viro 已提交
2645
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
C
Christoph Hellwig 已提交
2646 2647 2648
	struct btrfs_ioctl_vol_args *vol_args;
	int ret;

2649 2650 2651
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

2652 2653 2654
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
Y
Yan Zheng 已提交
2655

L
Li Zefan 已提交
2656
	vol_args = memdup_user(arg, sizeof(*vol_args));
2657 2658
	if (IS_ERR(vol_args)) {
		ret = PTR_ERR(vol_args);
D
Dan Carpenter 已提交
2659
		goto err_drop;
2660
	}
C
Christoph Hellwig 已提交
2661

2662
	vol_args->name[BTRFS_PATH_NAME_MAX] = '\0';
C
Christoph Hellwig 已提交
2663

2664 2665 2666 2667 2668 2669 2670 2671
	if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
			1)) {
		ret = BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
		goto out;
	}

	mutex_lock(&root->fs_info->volume_mutex);
	ret = btrfs_rm_device(root, vol_args->name);
2672
	mutex_unlock(&root->fs_info->volume_mutex);
2673
	atomic_set(&root->fs_info->mutually_exclusive_operation_running, 0);
2674

2675 2676 2677
	if (!ret)
		btrfs_info(root->fs_info, "disk deleted %s",vol_args->name);

2678 2679
out:
	kfree(vol_args);
D
Dan Carpenter 已提交
2680
err_drop:
2681
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
2682 2683 2684
	return ret;
}

J
Jan Schmidt 已提交
2685 2686
static long btrfs_ioctl_fs_info(struct btrfs_root *root, void __user *arg)
{
2687
	struct btrfs_ioctl_fs_info_args *fi_args;
J
Jan Schmidt 已提交
2688 2689 2690
	struct btrfs_device *device;
	struct btrfs_device *next;
	struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
2691
	int ret = 0;
J
Jan Schmidt 已提交
2692

2693 2694 2695 2696
	fi_args = kzalloc(sizeof(*fi_args), GFP_KERNEL);
	if (!fi_args)
		return -ENOMEM;

2697
	mutex_lock(&fs_devices->device_list_mutex);
2698 2699
	fi_args->num_devices = fs_devices->num_devices;
	memcpy(&fi_args->fsid, root->fs_info->fsid, sizeof(fi_args->fsid));
J
Jan Schmidt 已提交
2700 2701

	list_for_each_entry_safe(device, next, &fs_devices->devices, dev_list) {
2702 2703
		if (device->devid > fi_args->max_id)
			fi_args->max_id = device->devid;
J
Jan Schmidt 已提交
2704 2705 2706
	}
	mutex_unlock(&fs_devices->device_list_mutex);

2707 2708 2709 2710
	fi_args->nodesize = root->fs_info->super_copy->nodesize;
	fi_args->sectorsize = root->fs_info->super_copy->sectorsize;
	fi_args->clone_alignment = root->fs_info->super_copy->sectorsize;

2711 2712
	if (copy_to_user(arg, fi_args, sizeof(*fi_args)))
		ret = -EFAULT;
J
Jan Schmidt 已提交
2713

2714 2715
	kfree(fi_args);
	return ret;
J
Jan Schmidt 已提交
2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729
}

static long btrfs_ioctl_dev_info(struct btrfs_root *root, void __user *arg)
{
	struct btrfs_ioctl_dev_info_args *di_args;
	struct btrfs_device *dev;
	struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
	int ret = 0;
	char *s_uuid = NULL;

	di_args = memdup_user(arg, sizeof(*di_args));
	if (IS_ERR(di_args))
		return PTR_ERR(di_args);

2730
	if (!btrfs_is_empty_uuid(di_args->uuid))
J
Jan Schmidt 已提交
2731 2732 2733
		s_uuid = di_args->uuid;

	mutex_lock(&fs_devices->device_list_mutex);
2734
	dev = btrfs_find_device(root->fs_info, di_args->devid, s_uuid, NULL);
J
Jan Schmidt 已提交
2735 2736 2737 2738 2739 2740 2741

	if (!dev) {
		ret = -ENODEV;
		goto out;
	}

	di_args->devid = dev->devid;
2742 2743
	di_args->bytes_used = btrfs_device_get_bytes_used(dev);
	di_args->total_bytes = btrfs_device_get_total_bytes(dev);
J
Jan Schmidt 已提交
2744
	memcpy(di_args->uuid, dev->uuid, sizeof(di_args->uuid));
2745
	if (dev->name) {
2746 2747 2748 2749 2750 2751
		struct rcu_string *name;

		rcu_read_lock();
		name = rcu_dereference(dev->name);
		strncpy(di_args->path, name->str, sizeof(di_args->path));
		rcu_read_unlock();
2752 2753
		di_args->path[sizeof(di_args->path) - 1] = 0;
	} else {
2754
		di_args->path[0] = '\0';
2755
	}
J
Jan Schmidt 已提交
2756 2757

out:
2758
	mutex_unlock(&fs_devices->device_list_mutex);
J
Jan Schmidt 已提交
2759 2760 2761 2762 2763 2764 2765
	if (ret == 0 && copy_to_user(arg, di_args, sizeof(*di_args)))
		ret = -EFAULT;

	kfree(di_args);
	return ret;
}

M
Mark Fasheh 已提交
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
static struct page *extent_same_get_page(struct inode *inode, u64 off)
{
	struct page *page;
	pgoff_t index;
	struct extent_io_tree *tree = &BTRFS_I(inode)->io_tree;

	index = off >> PAGE_CACHE_SHIFT;

	page = grab_cache_page(inode->i_mapping, index);
	if (!page)
		return NULL;

	if (!PageUptodate(page)) {
		if (extent_read_full_page_nolock(tree, page, btrfs_get_extent,
						 0))
			return NULL;
		lock_page(page);
		if (!PageUptodate(page)) {
			unlock_page(page);
			page_cache_release(page);
			return NULL;
		}
	}
	unlock_page(page);

	return page;
}

2794 2795 2796 2797 2798 2799 2800 2801 2802
static inline void lock_extent_range(struct inode *inode, u64 off, u64 len)
{
	/* do any pending delalloc/csum calc on src, one way or
	   another, and lock file content */
	while (1) {
		struct btrfs_ordered_extent *ordered;
		lock_extent(&BTRFS_I(inode)->io_tree, off, off + len - 1);
		ordered = btrfs_lookup_first_ordered_extent(inode,
							    off + len - 1);
2803 2804 2805
		if ((!ordered ||
		     ordered->file_offset + ordered->len <= off ||
		     ordered->file_offset >= off + len) &&
2806
		    !test_range_bit(&BTRFS_I(inode)->io_tree, off,
2807 2808 2809
				    off + len - 1, EXTENT_DELALLOC, 0, NULL)) {
			if (ordered)
				btrfs_put_ordered_extent(ordered);
2810
			break;
2811
		}
2812 2813 2814 2815 2816 2817 2818
		unlock_extent(&BTRFS_I(inode)->io_tree, off, off + len - 1);
		if (ordered)
			btrfs_put_ordered_extent(ordered);
		btrfs_wait_ordered_range(inode, off, len);
	}
}

M
Mark Fasheh 已提交
2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945
static void btrfs_double_unlock(struct inode *inode1, u64 loff1,
				struct inode *inode2, u64 loff2, u64 len)
{
	unlock_extent(&BTRFS_I(inode1)->io_tree, loff1, loff1 + len - 1);
	unlock_extent(&BTRFS_I(inode2)->io_tree, loff2, loff2 + len - 1);

	mutex_unlock(&inode1->i_mutex);
	mutex_unlock(&inode2->i_mutex);
}

static void btrfs_double_lock(struct inode *inode1, u64 loff1,
			      struct inode *inode2, u64 loff2, u64 len)
{
	if (inode1 < inode2) {
		swap(inode1, inode2);
		swap(loff1, loff2);
	}

	mutex_lock_nested(&inode1->i_mutex, I_MUTEX_PARENT);
	lock_extent_range(inode1, loff1, len);
	if (inode1 != inode2) {
		mutex_lock_nested(&inode2->i_mutex, I_MUTEX_CHILD);
		lock_extent_range(inode2, loff2, len);
	}
}

static int btrfs_cmp_data(struct inode *src, u64 loff, struct inode *dst,
			  u64 dst_loff, u64 len)
{
	int ret = 0;
	struct page *src_page, *dst_page;
	unsigned int cmp_len = PAGE_CACHE_SIZE;
	void *addr, *dst_addr;

	while (len) {
		if (len < PAGE_CACHE_SIZE)
			cmp_len = len;

		src_page = extent_same_get_page(src, loff);
		if (!src_page)
			return -EINVAL;
		dst_page = extent_same_get_page(dst, dst_loff);
		if (!dst_page) {
			page_cache_release(src_page);
			return -EINVAL;
		}
		addr = kmap_atomic(src_page);
		dst_addr = kmap_atomic(dst_page);

		flush_dcache_page(src_page);
		flush_dcache_page(dst_page);

		if (memcmp(addr, dst_addr, cmp_len))
			ret = BTRFS_SAME_DATA_DIFFERS;

		kunmap_atomic(addr);
		kunmap_atomic(dst_addr);
		page_cache_release(src_page);
		page_cache_release(dst_page);

		if (ret)
			break;

		loff += cmp_len;
		dst_loff += cmp_len;
		len -= cmp_len;
	}

	return ret;
}

static int extent_same_check_offsets(struct inode *inode, u64 off, u64 len)
{
	u64 bs = BTRFS_I(inode)->root->fs_info->sb->s_blocksize;

	if (off + len > inode->i_size || off + len < off)
		return -EINVAL;
	/* Check that we are block aligned - btrfs_clone() requires this */
	if (!IS_ALIGNED(off, bs) || !IS_ALIGNED(off + len, bs))
		return -EINVAL;

	return 0;
}

static int btrfs_extent_same(struct inode *src, u64 loff, u64 len,
			     struct inode *dst, u64 dst_loff)
{
	int ret;

	/*
	 * btrfs_clone() can't handle extents in the same file
	 * yet. Once that works, we can drop this check and replace it
	 * with a check for the same inode, but overlapping extents.
	 */
	if (src == dst)
		return -EINVAL;

	btrfs_double_lock(src, loff, dst, dst_loff, len);

	ret = extent_same_check_offsets(src, loff, len);
	if (ret)
		goto out_unlock;

	ret = extent_same_check_offsets(dst, dst_loff, len);
	if (ret)
		goto out_unlock;

	/* don't make the dst file partly checksummed */
	if ((BTRFS_I(src)->flags & BTRFS_INODE_NODATASUM) !=
	    (BTRFS_I(dst)->flags & BTRFS_INODE_NODATASUM)) {
		ret = -EINVAL;
		goto out_unlock;
	}

	ret = btrfs_cmp_data(src, loff, dst, dst_loff, len);
	if (ret == 0)
		ret = btrfs_clone(src, dst, loff, len, len, dst_loff);

out_unlock:
	btrfs_double_unlock(src, loff, dst, dst_loff, len);

	return ret;
}

#define BTRFS_MAX_DEDUPE_LEN	(16 * 1024 * 1024)

static long btrfs_ioctl_file_extent_same(struct file *file,
2946
			struct btrfs_ioctl_same_args __user *argp)
M
Mark Fasheh 已提交
2947
{
2948 2949
	struct btrfs_ioctl_same_args *same;
	struct btrfs_ioctl_same_extent_info *info;
2950
	struct inode *src = file_inode(file);
M
Mark Fasheh 已提交
2951 2952 2953 2954
	u64 off;
	u64 len;
	int i;
	int ret;
2955
	unsigned long size;
M
Mark Fasheh 已提交
2956 2957
	u64 bs = BTRFS_I(src)->root->fs_info->sb->s_blocksize;
	bool is_admin = capable(CAP_SYS_ADMIN);
2958
	u16 count;
M
Mark Fasheh 已提交
2959 2960 2961 2962 2963 2964 2965 2966

	if (!(file->f_mode & FMODE_READ))
		return -EINVAL;

	ret = mnt_want_write_file(file);
	if (ret)
		return ret;

2967
	if (get_user(count, &argp->dest_count)) {
M
Mark Fasheh 已提交
2968 2969 2970 2971
		ret = -EFAULT;
		goto out;
	}

2972
	size = offsetof(struct btrfs_ioctl_same_args __user, info[count]);
2973

2974
	same = memdup_user(argp, size);
2975

2976 2977
	if (IS_ERR(same)) {
		ret = PTR_ERR(same);
2978 2979 2980 2981 2982
		goto out;
	}

	off = same->logical_offset;
	len = same->length;
M
Mark Fasheh 已提交
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

	/*
	 * Limit the total length we will dedupe for each operation.
	 * This is intended to bound the total time spent in this
	 * ioctl to something sane.
	 */
	if (len > BTRFS_MAX_DEDUPE_LEN)
		len = BTRFS_MAX_DEDUPE_LEN;

	if (WARN_ON_ONCE(bs < PAGE_CACHE_SIZE)) {
		/*
		 * Btrfs does not support blocksize < page_size. As a
		 * result, btrfs_cmp_data() won't correctly handle
		 * this situation without an update.
		 */
		ret = -EINVAL;
		goto out;
	}

	ret = -EISDIR;
	if (S_ISDIR(src->i_mode))
		goto out;

	ret = -EACCES;
	if (!S_ISREG(src->i_mode))
		goto out;

3010
	/* pre-format output fields to sane values */
3011
	for (i = 0; i < count; i++) {
3012 3013 3014
		same->info[i].bytes_deduped = 0ULL;
		same->info[i].status = 0;
	}
M
Mark Fasheh 已提交
3015

3016 3017 3018 3019
	for (i = 0, info = same->info; i < count; i++, info++) {
		struct inode *dst;
		struct fd dst_file = fdget(info->fd);
		if (!dst_file.file) {
3020
			info->status = -EBADF;
3021
			continue;
M
Mark Fasheh 已提交
3022
		}
3023
		dst = file_inode(dst_file.file);
M
Mark Fasheh 已提交
3024

3025
		if (!(is_admin || (dst_file.file->f_mode & FMODE_WRITE))) {
3026
			info->status = -EINVAL;
3027 3028 3029
		} else if (file->f_path.mnt != dst_file.file->f_path.mnt) {
			info->status = -EXDEV;
		} else if (S_ISDIR(dst->i_mode)) {
3030
			info->status = -EISDIR;
3031
		} else if (!S_ISREG(dst->i_mode)) {
3032
			info->status = -EACCES;
3033 3034 3035 3036 3037
		} else {
			info->status = btrfs_extent_same(src, off, len, dst,
							info->logical_offset);
			if (info->status == 0)
				info->bytes_deduped += len;
M
Mark Fasheh 已提交
3038
		}
3039
		fdput(dst_file);
M
Mark Fasheh 已提交
3040 3041
	}

3042 3043 3044 3045
	ret = copy_to_user(argp, same, size);
	if (ret)
		ret = -EFAULT;

M
Mark Fasheh 已提交
3046 3047 3048 3049 3050
out:
	mnt_drop_write_file(file);
	return ret;
}

J
Josef Bacik 已提交
3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083 3084 3085
/* Helper to check and see if this root currently has a ref on the given disk
 * bytenr.  If it does then we need to update the quota for this root.  This
 * doesn't do anything if quotas aren't enabled.
 */
static int check_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root,
		     u64 disko)
{
	struct seq_list tree_mod_seq_elem = {};
	struct ulist *roots;
	struct ulist_iterator uiter;
	struct ulist_node *root_node = NULL;
	int ret;

	if (!root->fs_info->quota_enabled)
		return 1;

	btrfs_get_tree_mod_seq(root->fs_info, &tree_mod_seq_elem);
	ret = btrfs_find_all_roots(trans, root->fs_info, disko,
				   tree_mod_seq_elem.seq, &roots);
	if (ret < 0)
		goto out;
	ret = 0;
	ULIST_ITER_INIT(&uiter);
	while ((root_node = ulist_next(roots, &uiter))) {
		if (root_node->val == root->objectid) {
			ret = 1;
			break;
		}
	}
	ulist_free(roots);
out:
	btrfs_put_tree_mod_seq(root->fs_info, &tree_mod_seq_elem);
	return ret;
}

3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112 3113 3114 3115 3116
static int clone_finish_inode_update(struct btrfs_trans_handle *trans,
				     struct inode *inode,
				     u64 endoff,
				     const u64 destoff,
				     const u64 olen)
{
	struct btrfs_root *root = BTRFS_I(inode)->root;
	int ret;

	inode_inc_iversion(inode);
	inode->i_mtime = inode->i_ctime = CURRENT_TIME;
	/*
	 * We round up to the block size at eof when determining which
	 * extents to clone above, but shouldn't round up the file size.
	 */
	if (endoff > destoff + olen)
		endoff = destoff + olen;
	if (endoff > inode->i_size)
		btrfs_i_size_write(inode, endoff);

	ret = btrfs_update_inode(trans, root, inode);
	if (ret) {
		btrfs_abort_transaction(trans, root, ret);
		btrfs_end_transaction(trans, root);
		goto out;
	}
	ret = btrfs_end_transaction(trans, root);
out:
	return ret;
}

3117 3118 3119 3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133
static void clone_update_extent_map(struct inode *inode,
				    const struct btrfs_trans_handle *trans,
				    const struct btrfs_path *path,
				    const u64 hole_offset,
				    const u64 hole_len)
{
	struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
	struct extent_map *em;
	int ret;

	em = alloc_extent_map();
	if (!em) {
		set_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
			&BTRFS_I(inode)->runtime_flags);
		return;
	}

3134 3135 3136 3137 3138
	if (path) {
		struct btrfs_file_extent_item *fi;

		fi = btrfs_item_ptr(path->nodes[0], path->slots[0],
				    struct btrfs_file_extent_item);
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
		btrfs_extent_item_to_extent_map(inode, path, fi, false, em);
		em->generation = -1;
		if (btrfs_file_extent_type(path->nodes[0], fi) ==
		    BTRFS_FILE_EXTENT_INLINE)
			set_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
				&BTRFS_I(inode)->runtime_flags);
	} else {
		em->start = hole_offset;
		em->len = hole_len;
		em->ram_bytes = em->len;
		em->orig_start = hole_offset;
		em->block_start = EXTENT_MAP_HOLE;
		em->block_len = 0;
		em->orig_block_len = 0;
		em->compress_type = BTRFS_COMPRESS_NONE;
		em->generation = trans->transid;
	}

	while (1) {
		write_lock(&em_tree->lock);
		ret = add_extent_mapping(em_tree, em, 1);
		write_unlock(&em_tree->lock);
		if (ret != -EEXIST) {
			free_extent_map(em);
			break;
		}
		btrfs_drop_extent_cache(inode, em->start,
					em->start + em->len - 1, 0);
	}

3169
	if (ret)
3170 3171 3172 3173
		set_bit(BTRFS_INODE_NEEDS_FULL_SYNC,
			&BTRFS_I(inode)->runtime_flags);
}

3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185
/**
 * btrfs_clone() - clone a range from inode file to another
 *
 * @src: Inode to clone from
 * @inode: Inode to clone to
 * @off: Offset within source to start clone from
 * @olen: Original length, passed by user, of range to clone
 * @olen_aligned: Block-aligned value of olen, extent_same uses
 *               identical values here
 * @destoff: Offset within @inode to start clone
 */
static int btrfs_clone(struct inode *src, struct inode *inode,
3186 3187
		       const u64 off, const u64 olen, const u64 olen_aligned,
		       const u64 destoff)
C
Christoph Hellwig 已提交
3188 3189
{
	struct btrfs_root *root = BTRFS_I(inode)->root;
3190
	struct btrfs_path *path = NULL;
C
Christoph Hellwig 已提交
3191
	struct extent_buffer *leaf;
3192 3193
	struct btrfs_trans_handle *trans;
	char *buf = NULL;
Y
Yan Zheng 已提交
3194
	struct btrfs_key key;
C
Christoph Hellwig 已提交
3195 3196
	u32 nritems;
	int slot;
Y
Yan Zheng 已提交
3197
	int ret;
J
Josef Bacik 已提交
3198
	int no_quota;
3199
	const u64 len = olen_aligned;
J
Josef Bacik 已提交
3200
	u64 last_disko = 0;
3201
	u64 last_dest_end = destoff;
Y
Yan Zheng 已提交
3202 3203

	ret = -ENOMEM;
3204
	buf = vmalloc(root->nodesize);
Y
Yan Zheng 已提交
3205
	if (!buf)
3206
		return ret;
Y
Yan Zheng 已提交
3207 3208 3209 3210

	path = btrfs_alloc_path();
	if (!path) {
		vfree(buf);
3211
		return ret;
3212 3213
	}

3214
	path->reada = 2;
3215
	/* clone data */
L
Li Zefan 已提交
3216
	key.objectid = btrfs_ino(src);
Y
Yan Zheng 已提交
3217
	key.type = BTRFS_EXTENT_DATA_KEY;
3218
	key.offset = off;
C
Christoph Hellwig 已提交
3219 3220 3221 3222 3223 3224

	while (1) {
		/*
		 * note the key will change type as we walk through the
		 * tree.
		 */
3225
		path->leave_spinning = 1;
3226 3227
		ret = btrfs_search_slot(NULL, BTRFS_I(src)->root, &key, path,
				0, 0);
C
Christoph Hellwig 已提交
3228 3229
		if (ret < 0)
			goto out;
3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240
		/*
		 * First search, if no extent item that starts at offset off was
		 * found but the previous item is an extent item, it's possible
		 * it might overlap our target range, therefore process it.
		 */
		if (key.offset == off && ret > 0 && path->slots[0] > 0) {
			btrfs_item_key_to_cpu(path->nodes[0], &key,
					      path->slots[0] - 1);
			if (key.type == BTRFS_EXTENT_DATA_KEY)
				path->slots[0]--;
		}
C
Christoph Hellwig 已提交
3241

Y
Yan Zheng 已提交
3242
		nritems = btrfs_header_nritems(path->nodes[0]);
3243
process_slot:
J
Josef Bacik 已提交
3244
		no_quota = 1;
Y
Yan Zheng 已提交
3245
		if (path->slots[0] >= nritems) {
3246
			ret = btrfs_next_leaf(BTRFS_I(src)->root, path);
C
Christoph Hellwig 已提交
3247 3248 3249 3250
			if (ret < 0)
				goto out;
			if (ret > 0)
				break;
Y
Yan Zheng 已提交
3251
			nritems = btrfs_header_nritems(path->nodes[0]);
C
Christoph Hellwig 已提交
3252 3253 3254 3255
		}
		leaf = path->nodes[0];
		slot = path->slots[0];

Y
Yan Zheng 已提交
3256
		btrfs_item_key_to_cpu(leaf, &key, slot);
3257
		if (key.type > BTRFS_EXTENT_DATA_KEY ||
L
Li Zefan 已提交
3258
		    key.objectid != btrfs_ino(src))
C
Christoph Hellwig 已提交
3259 3260
			break;

3261
		if (key.type == BTRFS_EXTENT_DATA_KEY) {
3262 3263
			struct btrfs_file_extent_item *extent;
			int type;
Z
Zheng Yan 已提交
3264 3265
			u32 size;
			struct btrfs_key new_key;
3266 3267 3268
			u64 disko = 0, diskl = 0;
			u64 datao = 0, datal = 0;
			u8 comp;
3269
			u64 drop_start;
Z
Zheng Yan 已提交
3270

3271 3272 3273 3274
			extent = btrfs_item_ptr(leaf, slot,
						struct btrfs_file_extent_item);
			comp = btrfs_file_extent_compression(leaf, extent);
			type = btrfs_file_extent_type(leaf, extent);
3275 3276
			if (type == BTRFS_FILE_EXTENT_REG ||
			    type == BTRFS_FILE_EXTENT_PREALLOC) {
C
Chris Mason 已提交
3277 3278 3279 3280
				disko = btrfs_file_extent_disk_bytenr(leaf,
								      extent);
				diskl = btrfs_file_extent_disk_num_bytes(leaf,
								 extent);
3281
				datao = btrfs_file_extent_offset(leaf, extent);
C
Chris Mason 已提交
3282 3283
				datal = btrfs_file_extent_num_bytes(leaf,
								    extent);
3284 3285 3286 3287 3288
			} else if (type == BTRFS_FILE_EXTENT_INLINE) {
				/* take upper bound, may be compressed */
				datal = btrfs_file_extent_ram_bytes(leaf,
								    extent);
			}
Z
Zheng Yan 已提交
3289

3290 3291 3292 3293 3294 3295
			/*
			 * The first search might have left us at an extent
			 * item that ends before our target range's start, can
			 * happen if we have holes and NO_HOLES feature enabled.
			 */
			if (key.offset + datal <= off) {
3296 3297
				path->slots[0]++;
				goto process_slot;
3298 3299
			} else if (key.offset >= off + len) {
				break;
3300 3301 3302 3303 3304 3305 3306 3307 3308
			}

			size = btrfs_item_size_nr(leaf, slot);
			read_extent_buffer(leaf, buf,
					   btrfs_item_ptr_offset(leaf, slot),
					   size);

			btrfs_release_path(path);
			path->leave_spinning = 0;
3309

Z
Zheng Yan 已提交
3310
			memcpy(&new_key, &key, sizeof(new_key));
L
Li Zefan 已提交
3311
			new_key.objectid = btrfs_ino(inode);
3312 3313 3314 3315
			if (off <= key.offset)
				new_key.offset = key.offset + destoff - off;
			else
				new_key.offset = destoff;
Z
Zheng Yan 已提交
3316

3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328
			/*
			 * Deal with a hole that doesn't have an extent item
			 * that represents it (NO_HOLES feature enabled).
			 * This hole is either in the middle of the cloning
			 * range or at the beginning (fully overlaps it or
			 * partially overlaps it).
			 */
			if (new_key.offset != last_dest_end)
				drop_start = last_dest_end;
			else
				drop_start = new_key.offset;

3329 3330 3331 3332 3333 3334
			/*
			 * 1 - adjusting old extent (we may have to split it)
			 * 1 - add new extent
			 * 1 - inode update
			 */
			trans = btrfs_start_transaction(root, 3);
3335 3336 3337 3338 3339
			if (IS_ERR(trans)) {
				ret = PTR_ERR(trans);
				goto out;
			}

3340 3341
			if (type == BTRFS_FILE_EXTENT_REG ||
			    type == BTRFS_FILE_EXTENT_PREALLOC) {
3342 3343 3344 3345 3346
				/*
				 *    a  | --- range to clone ---|  b
				 * | ------------- extent ------------- |
				 */

3347
				/* subtract range b */
3348 3349 3350
				if (key.offset + datal > off + len)
					datal = off + len - key.offset;

3351
				/* subtract range a */
3352 3353 3354 3355 3356
				if (off > key.offset) {
					datao += off - key.offset;
					datal -= off - key.offset;
				}

J
Josef Bacik 已提交
3357
				ret = btrfs_drop_extents(trans, root, inode,
3358
							 drop_start,
3359
							 new_key.offset + datal,
3360
							 1);
3361
				if (ret) {
3362
					if (ret != -EOPNOTSUPP)
3363 3364
						btrfs_abort_transaction(trans,
								root, ret);
3365 3366 3367
					btrfs_end_transaction(trans, root);
					goto out;
				}
3368

3369 3370
				ret = btrfs_insert_empty_item(trans, root, path,
							      &new_key, size);
3371 3372 3373 3374 3375 3376
				if (ret) {
					btrfs_abort_transaction(trans, root,
								ret);
					btrfs_end_transaction(trans, root);
					goto out;
				}
3377 3378 3379 3380

				leaf = path->nodes[0];
				slot = path->slots[0];
				write_extent_buffer(leaf, buf,
Z
Zheng Yan 已提交
3381 3382
					    btrfs_item_ptr_offset(leaf, slot),
					    size);
Y
Yan Zheng 已提交
3383

3384
				extent = btrfs_item_ptr(leaf, slot,
C
Christoph Hellwig 已提交
3385
						struct btrfs_file_extent_item);
3386 3387 3388 3389 3390 3391 3392 3393 3394

				/* disko == 0 means it's a hole */
				if (!disko)
					datao = 0;

				btrfs_set_file_extent_offset(leaf, extent,
							     datao);
				btrfs_set_file_extent_num_bytes(leaf, extent,
								datal);
J
Josef Bacik 已提交
3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416

				/*
				 * We need to look up the roots that point at
				 * this bytenr and see if the new root does.  If
				 * it does not we need to make sure we update
				 * quotas appropriately.
				 */
				if (disko && root != BTRFS_I(src)->root &&
				    disko != last_disko) {
					no_quota = check_ref(trans, root,
							     disko);
					if (no_quota < 0) {
						btrfs_abort_transaction(trans,
									root,
									ret);
						btrfs_end_transaction(trans,
								      root);
						ret = no_quota;
						goto out;
					}
				}

3417 3418
				if (disko) {
					inode_add_bytes(inode, datal);
Y
Yan Zheng 已提交
3419
					ret = btrfs_inc_extent_ref(trans, root,
3420 3421
							disko, diskl, 0,
							root->root_key.objectid,
L
Li Zefan 已提交
3422
							btrfs_ino(inode),
A
Arne Jansen 已提交
3423
							new_key.offset - datao,
J
Josef Bacik 已提交
3424
							no_quota);
3425 3426 3427 3428 3429 3430 3431 3432 3433
					if (ret) {
						btrfs_abort_transaction(trans,
									root,
									ret);
						btrfs_end_transaction(trans,
								      root);
						goto out;

					}
C
Christoph Hellwig 已提交
3434
				}
3435 3436 3437
			} else if (type == BTRFS_FILE_EXTENT_INLINE) {
				u64 skip = 0;
				u64 trim = 0;
3438 3439
				u64 aligned_end = 0;

3440 3441 3442 3443
				if (off > key.offset) {
					skip = off - key.offset;
					new_key.offset += skip;
				}
C
Chris Mason 已提交
3444

L
Liu Bo 已提交
3445 3446
				if (key.offset + datal > off + len)
					trim = key.offset + datal - (off + len);
C
Chris Mason 已提交
3447

3448 3449
				if (comp && (skip || trim)) {
					ret = -EINVAL;
3450
					btrfs_end_transaction(trans, root);
3451 3452 3453 3454
					goto out;
				}
				size -= skip + trim;
				datal -= skip + trim;
3455

3456 3457
				aligned_end = ALIGN(new_key.offset + datal,
						    root->sectorsize);
J
Josef Bacik 已提交
3458
				ret = btrfs_drop_extents(trans, root, inode,
3459
							 drop_start,
3460
							 aligned_end,
3461
							 1);
3462
				if (ret) {
3463
					if (ret != -EOPNOTSUPP)
3464 3465
						btrfs_abort_transaction(trans,
							root, ret);
3466 3467 3468
					btrfs_end_transaction(trans, root);
					goto out;
				}
3469

3470 3471
				ret = btrfs_insert_empty_item(trans, root, path,
							      &new_key, size);
3472 3473 3474 3475 3476 3477
				if (ret) {
					btrfs_abort_transaction(trans, root,
								ret);
					btrfs_end_transaction(trans, root);
					goto out;
				}
3478 3479

				if (skip) {
C
Chris Mason 已提交
3480 3481
					u32 start =
					  btrfs_file_extent_calc_inline_size(0);
3482 3483 3484 3485 3486 3487 3488 3489 3490 3491
					memmove(buf+start, buf+start+skip,
						datal);
				}

				leaf = path->nodes[0];
				slot = path->slots[0];
				write_extent_buffer(leaf, buf,
					    btrfs_item_ptr_offset(leaf, slot),
					    size);
				inode_add_bytes(inode, datal);
C
Christoph Hellwig 已提交
3492
			}
3493

3494 3495 3496
			/* If we have an implicit hole (NO_HOLES feature). */
			if (drop_start < new_key.offset)
				clone_update_extent_map(inode, trans,
3497
						NULL, drop_start,
3498 3499
						new_key.offset - drop_start);

3500
			clone_update_extent_map(inode, trans, path, 0, 0);
3501

3502
			btrfs_mark_buffer_dirty(leaf);
3503
			btrfs_release_path(path);
3504

3505 3506
			last_dest_end = ALIGN(new_key.offset + datal,
					      root->sectorsize);
3507 3508 3509 3510
			ret = clone_finish_inode_update(trans, inode,
							last_dest_end,
							destoff, olen);
			if (ret)
3511
				goto out;
3512 3513
			if (new_key.offset + datal >= destoff + len)
				break;
3514
		}
3515
		btrfs_release_path(path);
C
Christoph Hellwig 已提交
3516 3517 3518
		key.offset++;
	}
	ret = 0;
3519

3520 3521 3522 3523 3524 3525 3526 3527 3528 3529 3530 3531 3532 3533 3534 3535 3536 3537 3538 3539 3540 3541 3542 3543
	if (last_dest_end < destoff + len) {
		/*
		 * We have an implicit hole (NO_HOLES feature is enabled) that
		 * fully or partially overlaps our cloning range at its end.
		 */
		btrfs_release_path(path);

		/*
		 * 1 - remove extent(s)
		 * 1 - inode update
		 */
		trans = btrfs_start_transaction(root, 2);
		if (IS_ERR(trans)) {
			ret = PTR_ERR(trans);
			goto out;
		}
		ret = btrfs_drop_extents(trans, root, inode,
					 last_dest_end, destoff + len, 1);
		if (ret) {
			if (ret != -EOPNOTSUPP)
				btrfs_abort_transaction(trans, root, ret);
			btrfs_end_transaction(trans, root);
			goto out;
		}
3544 3545
		clone_update_extent_map(inode, trans, NULL, last_dest_end,
					destoff + len - last_dest_end);
3546 3547 3548 3549
		ret = clone_finish_inode_update(trans, inode, destoff + len,
						destoff, olen);
	}

C
Christoph Hellwig 已提交
3550
out:
3551 3552 3553 3554 3555 3556 3557 3558
	btrfs_free_path(path);
	vfree(buf);
	return ret;
}

static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd,
				       u64 off, u64 olen, u64 destoff)
{
A
Al Viro 已提交
3559
	struct inode *inode = file_inode(file);
3560 3561 3562 3563 3564 3565 3566 3567 3568 3569 3570 3571 3572 3573
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct fd src_file;
	struct inode *src;
	int ret;
	u64 len = olen;
	u64 bs = root->fs_info->sb->s_blocksize;
	int same_inode = 0;

	/*
	 * TODO:
	 * - split compressed inline extents.  annoying: we need to
	 *   decompress into destination's address_space (the file offset
	 *   may change, so source mapping won't do), then recompress (or
	 *   otherwise reinsert) a subrange.
3574 3575 3576
	 *
	 * - split destination inode's inline extents.  The inline extents can
	 *   be either compressed or non-compressed.
3577 3578 3579 3580 3581 3582 3583 3584 3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 3638 3639 3640 3641 3642 3643 3644 3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661
	 */

	/* the destination must be opened for writing */
	if (!(file->f_mode & FMODE_WRITE) || (file->f_flags & O_APPEND))
		return -EINVAL;

	if (btrfs_root_readonly(root))
		return -EROFS;

	ret = mnt_want_write_file(file);
	if (ret)
		return ret;

	src_file = fdget(srcfd);
	if (!src_file.file) {
		ret = -EBADF;
		goto out_drop_write;
	}

	ret = -EXDEV;
	if (src_file.file->f_path.mnt != file->f_path.mnt)
		goto out_fput;

	src = file_inode(src_file.file);

	ret = -EINVAL;
	if (src == inode)
		same_inode = 1;

	/* the src must be open for reading */
	if (!(src_file.file->f_mode & FMODE_READ))
		goto out_fput;

	/* don't make the dst file partly checksummed */
	if ((BTRFS_I(src)->flags & BTRFS_INODE_NODATASUM) !=
	    (BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM))
		goto out_fput;

	ret = -EISDIR;
	if (S_ISDIR(src->i_mode) || S_ISDIR(inode->i_mode))
		goto out_fput;

	ret = -EXDEV;
	if (src->i_sb != inode->i_sb)
		goto out_fput;

	if (!same_inode) {
		if (inode < src) {
			mutex_lock_nested(&inode->i_mutex, I_MUTEX_PARENT);
			mutex_lock_nested(&src->i_mutex, I_MUTEX_CHILD);
		} else {
			mutex_lock_nested(&src->i_mutex, I_MUTEX_PARENT);
			mutex_lock_nested(&inode->i_mutex, I_MUTEX_CHILD);
		}
	} else {
		mutex_lock(&src->i_mutex);
	}

	/* determine range to clone */
	ret = -EINVAL;
	if (off + len > src->i_size || off + len < off)
		goto out_unlock;
	if (len == 0)
		olen = len = src->i_size - off;
	/* if we extend to eof, continue to block boundary */
	if (off + len == src->i_size)
		len = ALIGN(src->i_size, bs) - off;

	/* verify the end result is block aligned */
	if (!IS_ALIGNED(off, bs) || !IS_ALIGNED(off + len, bs) ||
	    !IS_ALIGNED(destoff, bs))
		goto out_unlock;

	/* verify if ranges are overlapped within the same file */
	if (same_inode) {
		if (destoff + len > off && destoff < off + len)
			goto out_unlock;
	}

	if (destoff > inode->i_size) {
		ret = btrfs_cont_expand(inode, inode->i_size, destoff);
		if (ret)
			goto out_unlock;
	}

3662 3663 3664 3665 3666 3667 3668 3669 3670 3671
	/*
	 * Lock the target range too. Right after we replace the file extent
	 * items in the fs tree (which now point to the cloned data), we might
	 * have a worker replace them with extent items relative to a write
	 * operation that was issued before this clone operation (i.e. confront
	 * with inode.c:btrfs_finish_ordered_io).
	 */
	if (same_inode) {
		u64 lock_start = min_t(u64, off, destoff);
		u64 lock_len = max_t(u64, off, destoff) + len - lock_start;
3672

3673 3674 3675 3676 3677
		lock_extent_range(src, lock_start, lock_len);
	} else {
		lock_extent_range(src, off, len);
		lock_extent_range(inode, destoff, len);
	}
3678 3679 3680

	ret = btrfs_clone(src, inode, off, olen, len, destoff);

3681 3682 3683 3684 3685 3686 3687 3688 3689 3690 3691 3692 3693 3694 3695 3696
	if (same_inode) {
		u64 lock_start = min_t(u64, off, destoff);
		u64 lock_end = max_t(u64, off, destoff) + len - 1;

		unlock_extent(&BTRFS_I(src)->io_tree, lock_start, lock_end);
	} else {
		unlock_extent(&BTRFS_I(src)->io_tree, off, off + len - 1);
		unlock_extent(&BTRFS_I(inode)->io_tree, destoff,
			      destoff + len - 1);
	}
	/*
	 * Truncate page cache pages so that future reads will see the cloned
	 * data immediately and not the previous data.
	 */
	truncate_inode_pages_range(&inode->i_data, destoff,
				   PAGE_CACHE_ALIGN(destoff + len) - 1);
C
Christoph Hellwig 已提交
3697
out_unlock:
3698 3699 3700 3701 3702 3703 3704 3705 3706 3707 3708
	if (!same_inode) {
		if (inode < src) {
			mutex_unlock(&src->i_mutex);
			mutex_unlock(&inode->i_mutex);
		} else {
			mutex_unlock(&inode->i_mutex);
			mutex_unlock(&src->i_mutex);
		}
	} else {
		mutex_unlock(&src->i_mutex);
	}
C
Christoph Hellwig 已提交
3709
out_fput:
3710
	fdput(src_file);
3711
out_drop_write:
A
Al Viro 已提交
3712
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
3713 3714 3715
	return ret;
}

3716
static long btrfs_ioctl_clone_range(struct file *file, void __user *argp)
3717 3718 3719
{
	struct btrfs_ioctl_clone_range_args args;

3720
	if (copy_from_user(&args, argp, sizeof(args)))
3721 3722 3723 3724 3725
		return -EFAULT;
	return btrfs_ioctl_clone(file, args.src_fd, args.src_offset,
				 args.src_length, args.dest_offset);
}

C
Christoph Hellwig 已提交
3726 3727 3728 3729 3730 3731
/*
 * there are many ways the trans_start and trans_end ioctls can lead
 * to deadlocks.  They should only be used by applications that
 * basically own the machine, and have a very in depth understanding
 * of all the possible deadlocks and enospc problems.
 */
3732
static long btrfs_ioctl_trans_start(struct file *file)
C
Christoph Hellwig 已提交
3733
{
A
Al Viro 已提交
3734
	struct inode *inode = file_inode(file);
C
Christoph Hellwig 已提交
3735 3736
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;
3737
	int ret;
C
Christoph Hellwig 已提交
3738

3739
	ret = -EPERM;
3740
	if (!capable(CAP_SYS_ADMIN))
3741
		goto out;
3742

3743 3744
	ret = -EINPROGRESS;
	if (file->private_data)
C
Christoph Hellwig 已提交
3745
		goto out;
3746

L
Li Zefan 已提交
3747 3748 3749 3750
	ret = -EROFS;
	if (btrfs_root_readonly(root))
		goto out;

3751
	ret = mnt_want_write_file(file);
Y
Yan Zheng 已提交
3752 3753 3754
	if (ret)
		goto out;

J
Josef Bacik 已提交
3755
	atomic_inc(&root->fs_info->open_ioctl_trans);
3756

3757
	ret = -ENOMEM;
3758
	trans = btrfs_start_ioctl_transaction(root);
3759
	if (IS_ERR(trans))
3760 3761 3762 3763 3764 3765
		goto out_drop;

	file->private_data = trans;
	return 0;

out_drop:
J
Josef Bacik 已提交
3766
	atomic_dec(&root->fs_info->open_ioctl_trans);
A
Al Viro 已提交
3767
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
3768 3769 3770 3771
out:
	return ret;
}

3772 3773
static long btrfs_ioctl_default_subvol(struct file *file, void __user *argp)
{
A
Al Viro 已提交
3774
	struct inode *inode = file_inode(file);
3775 3776 3777 3778 3779 3780 3781 3782 3783
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_root *new_root;
	struct btrfs_dir_item *di;
	struct btrfs_trans_handle *trans;
	struct btrfs_path *path;
	struct btrfs_key location;
	struct btrfs_disk_key disk_key;
	u64 objectid = 0;
	u64 dir_id;
3784
	int ret;
3785 3786 3787 3788

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

3789 3790 3791 3792 3793 3794 3795 3796
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;

	if (copy_from_user(&objectid, argp, sizeof(objectid))) {
		ret = -EFAULT;
		goto out;
	}
3797 3798

	if (!objectid)
3799
		objectid = BTRFS_FS_TREE_OBJECTID;
3800 3801 3802 3803 3804 3805

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

	new_root = btrfs_read_fs_root_no_name(root->fs_info, &location);
3806 3807 3808 3809
	if (IS_ERR(new_root)) {
		ret = PTR_ERR(new_root);
		goto out;
	}
3810 3811

	path = btrfs_alloc_path();
3812 3813 3814 3815
	if (!path) {
		ret = -ENOMEM;
		goto out;
	}
3816 3817 3818
	path->leave_spinning = 1;

	trans = btrfs_start_transaction(root, 1);
3819
	if (IS_ERR(trans)) {
3820
		btrfs_free_path(path);
3821 3822
		ret = PTR_ERR(trans);
		goto out;
3823 3824
	}

3825
	dir_id = btrfs_super_root_dir(root->fs_info->super_copy);
3826 3827
	di = btrfs_lookup_dir_item(trans, root->fs_info->tree_root, path,
				   dir_id, "default", 7, 1);
3828
	if (IS_ERR_OR_NULL(di)) {
3829 3830
		btrfs_free_path(path);
		btrfs_end_transaction(trans, root);
3831 3832
		btrfs_err(new_root->fs_info, "Umm, you don't have the default dir"
			   "item, this isn't going to work");
3833 3834
		ret = -ENOENT;
		goto out;
3835 3836 3837 3838 3839 3840 3841
	}

	btrfs_cpu_key_to_disk(&disk_key, &new_root->root_key);
	btrfs_set_dir_item_key(path->nodes[0], di, &disk_key);
	btrfs_mark_buffer_dirty(path->nodes[0]);
	btrfs_free_path(path);

3842
	btrfs_set_fs_incompat(root->fs_info, DEFAULT_SUBVOL);
3843
	btrfs_end_transaction(trans, root);
3844 3845 3846
out:
	mnt_drop_write_file(file);
	return ret;
3847 3848
}

3849 3850
void btrfs_get_block_group_info(struct list_head *groups_list,
				struct btrfs_ioctl_space_info *space)
3851 3852 3853 3854 3855 3856 3857 3858 3859 3860 3861 3862 3863 3864
{
	struct btrfs_block_group_cache *block_group;

	space->total_bytes = 0;
	space->used_bytes = 0;
	space->flags = 0;
	list_for_each_entry(block_group, groups_list, list) {
		space->flags = block_group->flags;
		space->total_bytes += block_group->key.offset;
		space->used_bytes +=
			btrfs_block_group_used(&block_group->item);
	}
}

3865
static long btrfs_ioctl_space_info(struct btrfs_root *root, void __user *arg)
J
Josef Bacik 已提交
3866 3867 3868 3869
{
	struct btrfs_ioctl_space_args space_args;
	struct btrfs_ioctl_space_info space;
	struct btrfs_ioctl_space_info *dest;
3870
	struct btrfs_ioctl_space_info *dest_orig;
3871
	struct btrfs_ioctl_space_info __user *user_dest;
J
Josef Bacik 已提交
3872
	struct btrfs_space_info *info;
3873 3874 3875 3876 3877
	u64 types[] = {BTRFS_BLOCK_GROUP_DATA,
		       BTRFS_BLOCK_GROUP_SYSTEM,
		       BTRFS_BLOCK_GROUP_METADATA,
		       BTRFS_BLOCK_GROUP_DATA | BTRFS_BLOCK_GROUP_METADATA};
	int num_types = 4;
3878
	int alloc_size;
J
Josef Bacik 已提交
3879
	int ret = 0;
3880
	u64 slot_count = 0;
3881
	int i, c;
J
Josef Bacik 已提交
3882 3883 3884 3885 3886 3887

	if (copy_from_user(&space_args,
			   (struct btrfs_ioctl_space_args __user *)arg,
			   sizeof(space_args)))
		return -EFAULT;

3888 3889 3890 3891 3892 3893 3894 3895 3896 3897 3898 3899 3900 3901 3902 3903 3904 3905 3906 3907 3908 3909 3910 3911
	for (i = 0; i < num_types; i++) {
		struct btrfs_space_info *tmp;

		info = NULL;
		rcu_read_lock();
		list_for_each_entry_rcu(tmp, &root->fs_info->space_info,
					list) {
			if (tmp->flags == types[i]) {
				info = tmp;
				break;
			}
		}
		rcu_read_unlock();

		if (!info)
			continue;

		down_read(&info->groups_sem);
		for (c = 0; c < BTRFS_NR_RAID_TYPES; c++) {
			if (!list_empty(&info->block_groups[c]))
				slot_count++;
		}
		up_read(&info->groups_sem);
	}
3912

3913 3914 3915 3916 3917
	/*
	 * Global block reserve, exported as a space_info
	 */
	slot_count++;

3918 3919 3920 3921 3922
	/* space_slots == 0 means they are asking for a count */
	if (space_args.space_slots == 0) {
		space_args.total_spaces = slot_count;
		goto out;
	}
3923

3924
	slot_count = min_t(u64, space_args.space_slots, slot_count);
3925

3926
	alloc_size = sizeof(*dest) * slot_count;
3927

3928 3929 3930 3931 3932 3933
	/* we generally have at most 6 or so space infos, one for each raid
	 * level.  So, a whole page should be more than enough for everyone
	 */
	if (alloc_size > PAGE_CACHE_SIZE)
		return -ENOMEM;

J
Josef Bacik 已提交
3934
	space_args.total_spaces = 0;
3935 3936 3937 3938
	dest = kmalloc(alloc_size, GFP_NOFS);
	if (!dest)
		return -ENOMEM;
	dest_orig = dest;
J
Josef Bacik 已提交
3939

3940
	/* now we have a buffer to copy into */
3941 3942 3943
	for (i = 0; i < num_types; i++) {
		struct btrfs_space_info *tmp;

3944 3945 3946
		if (!slot_count)
			break;

3947 3948 3949 3950 3951 3952 3953 3954 3955 3956
		info = NULL;
		rcu_read_lock();
		list_for_each_entry_rcu(tmp, &root->fs_info->space_info,
					list) {
			if (tmp->flags == types[i]) {
				info = tmp;
				break;
			}
		}
		rcu_read_unlock();
3957

3958 3959 3960 3961 3962
		if (!info)
			continue;
		down_read(&info->groups_sem);
		for (c = 0; c < BTRFS_NR_RAID_TYPES; c++) {
			if (!list_empty(&info->block_groups[c])) {
3963 3964
				btrfs_get_block_group_info(
					&info->block_groups[c], &space);
3965 3966 3967
				memcpy(dest, &space, sizeof(space));
				dest++;
				space_args.total_spaces++;
3968
				slot_count--;
3969
			}
3970 3971
			if (!slot_count)
				break;
3972 3973
		}
		up_read(&info->groups_sem);
J
Josef Bacik 已提交
3974 3975
	}

3976 3977 3978 3979 3980 3981 3982 3983 3984 3985 3986 3987 3988 3989 3990
	/*
	 * Add global block reserve
	 */
	if (slot_count) {
		struct btrfs_block_rsv *block_rsv = &root->fs_info->global_block_rsv;

		spin_lock(&block_rsv->lock);
		space.total_bytes = block_rsv->size;
		space.used_bytes = block_rsv->size - block_rsv->reserved;
		spin_unlock(&block_rsv->lock);
		space.flags = BTRFS_SPACE_INFO_GLOBAL_RSV;
		memcpy(dest, &space, sizeof(space));
		space_args.total_spaces++;
	}

D
Daniel J Blueman 已提交
3991
	user_dest = (struct btrfs_ioctl_space_info __user *)
3992 3993 3994 3995 3996 3997 3998 3999
		(arg + sizeof(struct btrfs_ioctl_space_args));

	if (copy_to_user(user_dest, dest_orig, alloc_size))
		ret = -EFAULT;

	kfree(dest_orig);
out:
	if (ret == 0 && copy_to_user(arg, &space_args, sizeof(space_args)))
J
Josef Bacik 已提交
4000 4001 4002 4003 4004
		ret = -EFAULT;

	return ret;
}

C
Christoph Hellwig 已提交
4005 4006 4007 4008 4009 4010 4011 4012
/*
 * there are many ways the trans_start and trans_end ioctls can lead
 * to deadlocks.  They should only be used by applications that
 * basically own the machine, and have a very in depth understanding
 * of all the possible deadlocks and enospc problems.
 */
long btrfs_ioctl_trans_end(struct file *file)
{
A
Al Viro 已提交
4013
	struct inode *inode = file_inode(file);
C
Christoph Hellwig 已提交
4014 4015 4016 4017
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;

	trans = file->private_data;
4018 4019
	if (!trans)
		return -EINVAL;
C
Christoph Hellwig 已提交
4020
	file->private_data = NULL;
4021

4022 4023
	btrfs_end_transaction(trans, root);

J
Josef Bacik 已提交
4024
	atomic_dec(&root->fs_info->open_ioctl_trans);
4025

A
Al Viro 已提交
4026
	mnt_drop_write_file(file);
4027
	return 0;
C
Christoph Hellwig 已提交
4028 4029
}

4030 4031
static noinline long btrfs_ioctl_start_sync(struct btrfs_root *root,
					    void __user *argp)
4032 4033 4034
{
	struct btrfs_trans_handle *trans;
	u64 transid;
T
Tsutomu Itoh 已提交
4035
	int ret;
4036

M
Miao Xie 已提交
4037
	trans = btrfs_attach_transaction_barrier(root);
4038 4039 4040 4041 4042 4043 4044 4045
	if (IS_ERR(trans)) {
		if (PTR_ERR(trans) != -ENOENT)
			return PTR_ERR(trans);

		/* No running transaction, don't bother */
		transid = root->fs_info->last_trans_committed;
		goto out;
	}
4046
	transid = trans->transid;
T
Tsutomu Itoh 已提交
4047
	ret = btrfs_commit_transaction_async(trans, root, 0);
4048 4049
	if (ret) {
		btrfs_end_transaction(trans, root);
T
Tsutomu Itoh 已提交
4050
		return ret;
4051
	}
4052
out:
4053 4054 4055 4056 4057 4058
	if (argp)
		if (copy_to_user(argp, &transid, sizeof(transid)))
			return -EFAULT;
	return 0;
}

4059 4060
static noinline long btrfs_ioctl_wait_sync(struct btrfs_root *root,
					   void __user *argp)
4061 4062 4063 4064 4065 4066 4067 4068 4069 4070 4071 4072
{
	u64 transid;

	if (argp) {
		if (copy_from_user(&transid, argp, sizeof(transid)))
			return -EFAULT;
	} else {
		transid = 0;  /* current trans */
	}
	return btrfs_wait_for_commit(root, transid);
}

M
Miao Xie 已提交
4073
static long btrfs_ioctl_scrub(struct file *file, void __user *arg)
J
Jan Schmidt 已提交
4074
{
A
Al Viro 已提交
4075
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
J
Jan Schmidt 已提交
4076
	struct btrfs_ioctl_scrub_args *sa;
M
Miao Xie 已提交
4077
	int ret;
J
Jan Schmidt 已提交
4078 4079 4080 4081 4082 4083 4084 4085

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

	sa = memdup_user(arg, sizeof(*sa));
	if (IS_ERR(sa))
		return PTR_ERR(sa);

M
Miao Xie 已提交
4086 4087 4088 4089 4090 4091
	if (!(sa->flags & BTRFS_SCRUB_READONLY)) {
		ret = mnt_want_write_file(file);
		if (ret)
			goto out;
	}

4092
	ret = btrfs_scrub_dev(root->fs_info, sa->devid, sa->start, sa->end,
4093 4094
			      &sa->progress, sa->flags & BTRFS_SCRUB_READONLY,
			      0);
J
Jan Schmidt 已提交
4095 4096 4097 4098

	if (copy_to_user(arg, sa, sizeof(*sa)))
		ret = -EFAULT;

M
Miao Xie 已提交
4099 4100 4101
	if (!(sa->flags & BTRFS_SCRUB_READONLY))
		mnt_drop_write_file(file);
out:
J
Jan Schmidt 已提交
4102 4103 4104 4105 4106 4107 4108 4109 4110
	kfree(sa);
	return ret;
}

static long btrfs_ioctl_scrub_cancel(struct btrfs_root *root, void __user *arg)
{
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

4111
	return btrfs_scrub_cancel(root->fs_info);
J
Jan Schmidt 已提交
4112 4113 4114 4115 4116 4117 4118 4119 4120 4121 4122 4123 4124 4125 4126 4127 4128 4129 4130 4131 4132 4133 4134 4135
}

static long btrfs_ioctl_scrub_progress(struct btrfs_root *root,
				       void __user *arg)
{
	struct btrfs_ioctl_scrub_args *sa;
	int ret;

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

	sa = memdup_user(arg, sizeof(*sa));
	if (IS_ERR(sa))
		return PTR_ERR(sa);

	ret = btrfs_scrub_progress(root, sa->devid, &sa->progress);

	if (copy_to_user(arg, sa, sizeof(*sa)))
		ret = -EFAULT;

	kfree(sa);
	return ret;
}

4136
static long btrfs_ioctl_get_dev_stats(struct btrfs_root *root,
4137
				      void __user *arg)
4138 4139 4140 4141 4142 4143 4144 4145
{
	struct btrfs_ioctl_get_dev_stats *sa;
	int ret;

	sa = memdup_user(arg, sizeof(*sa));
	if (IS_ERR(sa))
		return PTR_ERR(sa);

4146 4147 4148 4149 4150 4151
	if ((sa->flags & BTRFS_DEV_STATS_RESET) && !capable(CAP_SYS_ADMIN)) {
		kfree(sa);
		return -EPERM;
	}

	ret = btrfs_get_dev_stats(root, sa);
4152 4153 4154 4155 4156 4157 4158 4159

	if (copy_to_user(arg, sa, sizeof(*sa)))
		ret = -EFAULT;

	kfree(sa);
	return ret;
}

4160 4161 4162 4163 4164 4165 4166 4167 4168 4169 4170 4171 4172 4173
static long btrfs_ioctl_dev_replace(struct btrfs_root *root, void __user *arg)
{
	struct btrfs_ioctl_dev_replace_args *p;
	int ret;

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

	p = memdup_user(arg, sizeof(*p));
	if (IS_ERR(p))
		return PTR_ERR(p);

	switch (p->cmd) {
	case BTRFS_IOCTL_DEV_REPLACE_CMD_START:
4174 4175 4176 4177
		if (root->fs_info->sb->s_flags & MS_RDONLY) {
			ret = -EROFS;
			goto out;
		}
4178 4179 4180
		if (atomic_xchg(
			&root->fs_info->mutually_exclusive_operation_running,
			1)) {
4181
			ret = BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
4182 4183 4184 4185 4186 4187 4188 4189 4190 4191 4192 4193 4194 4195 4196 4197 4198 4199 4200 4201 4202
		} else {
			ret = btrfs_dev_replace_start(root, p);
			atomic_set(
			 &root->fs_info->mutually_exclusive_operation_running,
			 0);
		}
		break;
	case BTRFS_IOCTL_DEV_REPLACE_CMD_STATUS:
		btrfs_dev_replace_status(root->fs_info, p);
		ret = 0;
		break;
	case BTRFS_IOCTL_DEV_REPLACE_CMD_CANCEL:
		ret = btrfs_dev_replace_cancel(root->fs_info, p);
		break;
	default:
		ret = -EINVAL;
		break;
	}

	if (copy_to_user(arg, p, sizeof(*p)))
		ret = -EFAULT;
4203
out:
4204 4205 4206 4207
	kfree(p);
	return ret;
}

4208 4209 4210 4211
static long btrfs_ioctl_ino_to_path(struct btrfs_root *root, void __user *arg)
{
	int ret = 0;
	int i;
4212
	u64 rel_ptr;
4213
	int size;
4214
	struct btrfs_ioctl_ino_path_args *ipa = NULL;
4215 4216 4217
	struct inode_fs_paths *ipath = NULL;
	struct btrfs_path *path;

4218
	if (!capable(CAP_DAC_READ_SEARCH))
4219 4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232 4233 4234 4235 4236 4237 4238 4239 4240 4241 4242 4243 4244 4245 4246
		return -EPERM;

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

	ipa = memdup_user(arg, sizeof(*ipa));
	if (IS_ERR(ipa)) {
		ret = PTR_ERR(ipa);
		ipa = NULL;
		goto out;
	}

	size = min_t(u32, ipa->size, 4096);
	ipath = init_ipath(size, root, path);
	if (IS_ERR(ipath)) {
		ret = PTR_ERR(ipath);
		ipath = NULL;
		goto out;
	}

	ret = paths_from_inode(ipa->inum, ipath);
	if (ret < 0)
		goto out;

	for (i = 0; i < ipath->fspath->elem_cnt; ++i) {
4247 4248
		rel_ptr = ipath->fspath->val[i] -
			  (u64)(unsigned long)ipath->fspath->val;
4249
		ipath->fspath->val[i] = rel_ptr;
4250 4251
	}

4252 4253
	ret = copy_to_user((void *)(unsigned long)ipa->fspath,
			   (void *)(unsigned long)ipath->fspath, size);
4254 4255 4256 4257 4258 4259 4260 4261 4262 4263 4264 4265 4266 4267 4268 4269 4270 4271 4272 4273 4274 4275 4276 4277 4278 4279 4280 4281 4282 4283 4284 4285 4286 4287 4288 4289 4290 4291 4292 4293 4294 4295 4296 4297 4298 4299 4300 4301 4302 4303 4304 4305 4306 4307 4308 4309 4310 4311
	if (ret) {
		ret = -EFAULT;
		goto out;
	}

out:
	btrfs_free_path(path);
	free_ipath(ipath);
	kfree(ipa);

	return ret;
}

static int build_ino_list(u64 inum, u64 offset, u64 root, void *ctx)
{
	struct btrfs_data_container *inodes = ctx;
	const size_t c = 3 * sizeof(u64);

	if (inodes->bytes_left >= c) {
		inodes->bytes_left -= c;
		inodes->val[inodes->elem_cnt] = inum;
		inodes->val[inodes->elem_cnt + 1] = offset;
		inodes->val[inodes->elem_cnt + 2] = root;
		inodes->elem_cnt += 3;
	} else {
		inodes->bytes_missing += c - inodes->bytes_left;
		inodes->bytes_left = 0;
		inodes->elem_missed += 3;
	}

	return 0;
}

static long btrfs_ioctl_logical_to_ino(struct btrfs_root *root,
					void __user *arg)
{
	int ret = 0;
	int size;
	struct btrfs_ioctl_logical_ino_args *loi;
	struct btrfs_data_container *inodes = NULL;
	struct btrfs_path *path = NULL;

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

	loi = memdup_user(arg, sizeof(*loi));
	if (IS_ERR(loi)) {
		ret = PTR_ERR(loi);
		loi = NULL;
		goto out;
	}

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

4312
	size = min_t(u32, loi->size, 64 * 1024);
4313 4314 4315 4316 4317 4318 4319
	inodes = init_data_container(size);
	if (IS_ERR(inodes)) {
		ret = PTR_ERR(inodes);
		inodes = NULL;
		goto out;
	}

L
Liu Bo 已提交
4320 4321 4322
	ret = iterate_inodes_from_logical(loi->logical, root->fs_info, path,
					  build_ino_list, inodes);
	if (ret == -EINVAL)
4323 4324 4325 4326
		ret = -ENOENT;
	if (ret < 0)
		goto out;

4327 4328
	ret = copy_to_user((void *)(unsigned long)loi->inodes,
			   (void *)(unsigned long)inodes, size);
4329 4330 4331 4332 4333
	if (ret)
		ret = -EFAULT;

out:
	btrfs_free_path(path);
4334
	vfree(inodes);
4335 4336 4337 4338 4339
	kfree(loi);

	return ret;
}

4340
void update_ioctl_balance_args(struct btrfs_fs_info *fs_info, int lock,
4341 4342 4343 4344 4345 4346
			       struct btrfs_ioctl_balance_args *bargs)
{
	struct btrfs_balance_control *bctl = fs_info->balance_ctl;

	bargs->flags = bctl->flags;

4347 4348 4349 4350
	if (atomic_read(&fs_info->balance_running))
		bargs->state |= BTRFS_BALANCE_STATE_RUNNING;
	if (atomic_read(&fs_info->balance_pause_req))
		bargs->state |= BTRFS_BALANCE_STATE_PAUSE_REQ;
4351 4352
	if (atomic_read(&fs_info->balance_cancel_req))
		bargs->state |= BTRFS_BALANCE_STATE_CANCEL_REQ;
4353

4354 4355 4356
	memcpy(&bargs->data, &bctl->data, sizeof(bargs->data));
	memcpy(&bargs->meta, &bctl->meta, sizeof(bargs->meta));
	memcpy(&bargs->sys, &bctl->sys, sizeof(bargs->sys));
4357 4358 4359 4360 4361 4362 4363 4364

	if (lock) {
		spin_lock(&fs_info->balance_lock);
		memcpy(&bargs->stat, &bctl->stat, sizeof(bargs->stat));
		spin_unlock(&fs_info->balance_lock);
	} else {
		memcpy(&bargs->stat, &bctl->stat, sizeof(bargs->stat));
	}
4365 4366
}

4367
static long btrfs_ioctl_balance(struct file *file, void __user *arg)
4368
{
A
Al Viro 已提交
4369
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
4370 4371 4372
	struct btrfs_fs_info *fs_info = root->fs_info;
	struct btrfs_ioctl_balance_args *bargs;
	struct btrfs_balance_control *bctl;
4373
	bool need_unlock; /* for mut. excl. ops lock */
4374 4375 4376 4377 4378
	int ret;

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

4379
	ret = mnt_want_write_file(file);
4380 4381 4382
	if (ret)
		return ret;

4383 4384 4385 4386 4387 4388 4389 4390 4391 4392 4393 4394 4395 4396
again:
	if (!atomic_xchg(&fs_info->mutually_exclusive_operation_running, 1)) {
		mutex_lock(&fs_info->volume_mutex);
		mutex_lock(&fs_info->balance_mutex);
		need_unlock = true;
		goto locked;
	}

	/*
	 * mut. excl. ops lock is locked.  Three possibilites:
	 *   (1) some other op is running
	 *   (2) balance is running
	 *   (3) balance is paused -- special case (think resume)
	 */
4397
	mutex_lock(&fs_info->balance_mutex);
4398 4399 4400 4401 4402 4403 4404 4405 4406 4407 4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418 4419 4420 4421 4422 4423 4424
	if (fs_info->balance_ctl) {
		/* this is either (2) or (3) */
		if (!atomic_read(&fs_info->balance_running)) {
			mutex_unlock(&fs_info->balance_mutex);
			if (!mutex_trylock(&fs_info->volume_mutex))
				goto again;
			mutex_lock(&fs_info->balance_mutex);

			if (fs_info->balance_ctl &&
			    !atomic_read(&fs_info->balance_running)) {
				/* this is (3) */
				need_unlock = false;
				goto locked;
			}

			mutex_unlock(&fs_info->balance_mutex);
			mutex_unlock(&fs_info->volume_mutex);
			goto again;
		} else {
			/* this is (2) */
			mutex_unlock(&fs_info->balance_mutex);
			ret = -EINPROGRESS;
			goto out;
		}
	} else {
		/* this is (1) */
		mutex_unlock(&fs_info->balance_mutex);
4425
		ret = BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
4426 4427 4428 4429 4430
		goto out;
	}

locked:
	BUG_ON(!atomic_read(&fs_info->mutually_exclusive_operation_running));
4431 4432 4433 4434 4435

	if (arg) {
		bargs = memdup_user(arg, sizeof(*bargs));
		if (IS_ERR(bargs)) {
			ret = PTR_ERR(bargs);
4436
			goto out_unlock;
4437
		}
4438 4439 4440 4441 4442 4443 4444 4445 4446 4447 4448 4449 4450 4451

		if (bargs->flags & BTRFS_BALANCE_RESUME) {
			if (!fs_info->balance_ctl) {
				ret = -ENOTCONN;
				goto out_bargs;
			}

			bctl = fs_info->balance_ctl;
			spin_lock(&fs_info->balance_lock);
			bctl->flags |= BTRFS_BALANCE_RESUME;
			spin_unlock(&fs_info->balance_lock);

			goto do_balance;
		}
4452 4453 4454 4455
	} else {
		bargs = NULL;
	}

4456
	if (fs_info->balance_ctl) {
4457 4458 4459 4460
		ret = -EINPROGRESS;
		goto out_bargs;
	}

4461 4462 4463 4464 4465 4466 4467 4468 4469 4470 4471 4472 4473
	bctl = kzalloc(sizeof(*bctl), GFP_NOFS);
	if (!bctl) {
		ret = -ENOMEM;
		goto out_bargs;
	}

	bctl->fs_info = fs_info;
	if (arg) {
		memcpy(&bctl->data, &bargs->data, sizeof(bctl->data));
		memcpy(&bctl->meta, &bargs->meta, sizeof(bctl->meta));
		memcpy(&bctl->sys, &bargs->sys, sizeof(bctl->sys));

		bctl->flags = bargs->flags;
4474 4475 4476
	} else {
		/* balance everything - no filters */
		bctl->flags |= BTRFS_BALANCE_TYPE_MASK;
4477 4478
	}

4479
do_balance:
4480
	/*
4481 4482 4483 4484 4485
	 * Ownership of bctl and mutually_exclusive_operation_running
	 * goes to to btrfs_balance.  bctl is freed in __cancel_balance,
	 * or, if restriper was paused all the way until unmount, in
	 * free_fs_info.  mutually_exclusive_operation_running is
	 * cleared in __cancel_balance.
4486
	 */
4487 4488 4489 4490
	need_unlock = false;

	ret = btrfs_balance(bctl, bargs);

4491 4492 4493 4494 4495 4496 4497
	if (arg) {
		if (copy_to_user(arg, bargs, sizeof(*bargs)))
			ret = -EFAULT;
	}

out_bargs:
	kfree(bargs);
4498
out_unlock:
4499 4500
	mutex_unlock(&fs_info->balance_mutex);
	mutex_unlock(&fs_info->volume_mutex);
4501 4502 4503
	if (need_unlock)
		atomic_set(&fs_info->mutually_exclusive_operation_running, 0);
out:
4504
	mnt_drop_write_file(file);
4505 4506 4507
	return ret;
}

4508 4509 4510 4511 4512 4513 4514 4515
static long btrfs_ioctl_balance_ctl(struct btrfs_root *root, int cmd)
{
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

	switch (cmd) {
	case BTRFS_BALANCE_CTL_PAUSE:
		return btrfs_pause_balance(root->fs_info);
4516 4517
	case BTRFS_BALANCE_CTL_CANCEL:
		return btrfs_cancel_balance(root->fs_info);
4518 4519 4520 4521 4522
	}

	return -EINVAL;
}

4523 4524 4525 4526 4527 4528 4529 4530 4531 4532 4533 4534 4535 4536 4537 4538 4539 4540 4541 4542 4543 4544 4545 4546 4547 4548 4549 4550 4551 4552 4553 4554 4555
static long btrfs_ioctl_balance_progress(struct btrfs_root *root,
					 void __user *arg)
{
	struct btrfs_fs_info *fs_info = root->fs_info;
	struct btrfs_ioctl_balance_args *bargs;
	int ret = 0;

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

	mutex_lock(&fs_info->balance_mutex);
	if (!fs_info->balance_ctl) {
		ret = -ENOTCONN;
		goto out;
	}

	bargs = kzalloc(sizeof(*bargs), GFP_NOFS);
	if (!bargs) {
		ret = -ENOMEM;
		goto out;
	}

	update_ioctl_balance_args(fs_info, 1, bargs);

	if (copy_to_user(arg, bargs, sizeof(*bargs)))
		ret = -EFAULT;

	kfree(bargs);
out:
	mutex_unlock(&fs_info->balance_mutex);
	return ret;
}

4556
static long btrfs_ioctl_quota_ctl(struct file *file, void __user *arg)
A
Arne Jansen 已提交
4557
{
A
Al Viro 已提交
4558
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
A
Arne Jansen 已提交
4559 4560 4561 4562 4563 4564 4565 4566
	struct btrfs_ioctl_quota_ctl_args *sa;
	struct btrfs_trans_handle *trans = NULL;
	int ret;
	int err;

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

4567 4568 4569
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
4570 4571

	sa = memdup_user(arg, sizeof(*sa));
4572 4573 4574 4575
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
4576

4577
	down_write(&root->fs_info->subvol_sem);
J
Jan Schmidt 已提交
4578 4579 4580 4581
	trans = btrfs_start_transaction(root->fs_info->tree_root, 2);
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto out;
A
Arne Jansen 已提交
4582 4583 4584 4585 4586 4587 4588 4589 4590 4591 4592 4593 4594 4595
	}

	switch (sa->cmd) {
	case BTRFS_QUOTA_CTL_ENABLE:
		ret = btrfs_quota_enable(trans, root->fs_info);
		break;
	case BTRFS_QUOTA_CTL_DISABLE:
		ret = btrfs_quota_disable(trans, root->fs_info);
		break;
	default:
		ret = -EINVAL;
		break;
	}

J
Jan Schmidt 已提交
4596 4597 4598
	err = btrfs_commit_transaction(trans, root->fs_info->tree_root);
	if (err && !ret)
		ret = err;
A
Arne Jansen 已提交
4599 4600
out:
	kfree(sa);
4601
	up_write(&root->fs_info->subvol_sem);
4602 4603
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
4604 4605 4606
	return ret;
}

4607
static long btrfs_ioctl_qgroup_assign(struct file *file, void __user *arg)
A
Arne Jansen 已提交
4608
{
A
Al Viro 已提交
4609
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
A
Arne Jansen 已提交
4610 4611 4612 4613 4614 4615 4616 4617
	struct btrfs_ioctl_qgroup_assign_args *sa;
	struct btrfs_trans_handle *trans;
	int ret;
	int err;

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

4618 4619 4620
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
4621 4622

	sa = memdup_user(arg, sizeof(*sa));
4623 4624 4625 4626
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
4627 4628 4629 4630 4631 4632 4633 4634 4635 4636 4637 4638 4639 4640 4641 4642 4643 4644 4645 4646 4647 4648

	trans = btrfs_join_transaction(root);
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto out;
	}

	/* FIXME: check if the IDs really exist */
	if (sa->assign) {
		ret = btrfs_add_qgroup_relation(trans, root->fs_info,
						sa->src, sa->dst);
	} else {
		ret = btrfs_del_qgroup_relation(trans, root->fs_info,
						sa->src, sa->dst);
	}

	err = btrfs_end_transaction(trans, root);
	if (err && !ret)
		ret = err;

out:
	kfree(sa);
4649 4650
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
4651 4652 4653
	return ret;
}

4654
static long btrfs_ioctl_qgroup_create(struct file *file, void __user *arg)
A
Arne Jansen 已提交
4655
{
A
Al Viro 已提交
4656
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
A
Arne Jansen 已提交
4657 4658 4659 4660 4661 4662 4663 4664
	struct btrfs_ioctl_qgroup_create_args *sa;
	struct btrfs_trans_handle *trans;
	int ret;
	int err;

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

4665 4666 4667
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
4668 4669

	sa = memdup_user(arg, sizeof(*sa));
4670 4671 4672 4673
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
4674

M
Miao Xie 已提交
4675 4676 4677 4678 4679
	if (!sa->qgroupid) {
		ret = -EINVAL;
		goto out;
	}

A
Arne Jansen 已提交
4680 4681 4682 4683 4684 4685 4686 4687 4688 4689 4690 4691 4692 4693 4694 4695 4696 4697 4698 4699
	trans = btrfs_join_transaction(root);
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto out;
	}

	/* FIXME: check if the IDs really exist */
	if (sa->create) {
		ret = btrfs_create_qgroup(trans, root->fs_info, sa->qgroupid,
					  NULL);
	} else {
		ret = btrfs_remove_qgroup(trans, root->fs_info, sa->qgroupid);
	}

	err = btrfs_end_transaction(trans, root);
	if (err && !ret)
		ret = err;

out:
	kfree(sa);
4700 4701
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
4702 4703 4704
	return ret;
}

4705
static long btrfs_ioctl_qgroup_limit(struct file *file, void __user *arg)
A
Arne Jansen 已提交
4706
{
A
Al Viro 已提交
4707
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
A
Arne Jansen 已提交
4708 4709 4710 4711 4712 4713 4714 4715 4716
	struct btrfs_ioctl_qgroup_limit_args *sa;
	struct btrfs_trans_handle *trans;
	int ret;
	int err;
	u64 qgroupid;

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

4717 4718 4719
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
4720 4721

	sa = memdup_user(arg, sizeof(*sa));
4722 4723 4724 4725
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
4726 4727 4728 4729 4730 4731 4732 4733 4734 4735 4736 4737 4738 4739 4740 4741 4742 4743 4744 4745 4746 4747

	trans = btrfs_join_transaction(root);
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto out;
	}

	qgroupid = sa->qgroupid;
	if (!qgroupid) {
		/* take the current subvol as qgroup */
		qgroupid = root->root_key.objectid;
	}

	/* FIXME: check if the IDs really exist */
	ret = btrfs_limit_qgroup(trans, root->fs_info, qgroupid, &sa->lim);

	err = btrfs_end_transaction(trans, root);
	if (err && !ret)
		ret = err;

out:
	kfree(sa);
4748 4749
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
4750 4751 4752
	return ret;
}

J
Jan Schmidt 已提交
4753 4754
static long btrfs_ioctl_quota_rescan(struct file *file, void __user *arg)
{
A
Al Viro 已提交
4755
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
J
Jan Schmidt 已提交
4756 4757 4758 4759 4760 4761 4762 4763 4764 4765 4766 4767 4768 4769 4770 4771 4772 4773 4774 4775 4776 4777 4778 4779 4780 4781 4782 4783 4784 4785 4786 4787
	struct btrfs_ioctl_quota_rescan_args *qsa;
	int ret;

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

	ret = mnt_want_write_file(file);
	if (ret)
		return ret;

	qsa = memdup_user(arg, sizeof(*qsa));
	if (IS_ERR(qsa)) {
		ret = PTR_ERR(qsa);
		goto drop_write;
	}

	if (qsa->flags) {
		ret = -EINVAL;
		goto out;
	}

	ret = btrfs_qgroup_rescan(root->fs_info);

out:
	kfree(qsa);
drop_write:
	mnt_drop_write_file(file);
	return ret;
}

static long btrfs_ioctl_quota_rescan_status(struct file *file, void __user *arg)
{
A
Al Viro 已提交
4788
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
J
Jan Schmidt 已提交
4789 4790 4791 4792 4793 4794 4795 4796 4797 4798 4799 4800 4801 4802 4803 4804 4805 4806 4807 4808 4809 4810
	struct btrfs_ioctl_quota_rescan_args *qsa;
	int ret = 0;

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

	qsa = kzalloc(sizeof(*qsa), GFP_NOFS);
	if (!qsa)
		return -ENOMEM;

	if (root->fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_RESCAN) {
		qsa->flags = 1;
		qsa->progress = root->fs_info->qgroup_rescan_progress.objectid;
	}

	if (copy_to_user(arg, qsa, sizeof(*qsa)))
		ret = -EFAULT;

	kfree(qsa);
	return ret;
}

4811 4812
static long btrfs_ioctl_quota_rescan_wait(struct file *file, void __user *arg)
{
A
Al Viro 已提交
4813
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
4814 4815 4816 4817 4818 4819 4820

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

	return btrfs_qgroup_wait_for_completion(root->fs_info);
}

4821 4822
static long _btrfs_ioctl_set_received_subvol(struct file *file,
					    struct btrfs_ioctl_received_subvol_args *sa)
4823
{
A
Al Viro 已提交
4824
	struct inode *inode = file_inode(file);
4825 4826 4827 4828 4829
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_root_item *root_item = &root->root_item;
	struct btrfs_trans_handle *trans;
	struct timespec ct = CURRENT_TIME;
	int ret = 0;
4830
	int received_uuid_changed;
4831

4832 4833 4834
	if (!inode_owner_or_capable(inode))
		return -EPERM;

4835 4836 4837 4838 4839 4840 4841 4842 4843 4844 4845 4846 4847 4848 4849 4850
	ret = mnt_want_write_file(file);
	if (ret < 0)
		return ret;

	down_write(&root->fs_info->subvol_sem);

	if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID) {
		ret = -EINVAL;
		goto out;
	}

	if (btrfs_root_readonly(root)) {
		ret = -EROFS;
		goto out;
	}

4851 4852 4853 4854 4855
	/*
	 * 1 - root item
	 * 2 - uuid items (received uuid + subvol uuid)
	 */
	trans = btrfs_start_transaction(root, 3);
4856 4857 4858 4859 4860 4861 4862 4863 4864 4865
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		trans = NULL;
		goto out;
	}

	sa->rtransid = trans->transid;
	sa->rtime.sec = ct.tv_sec;
	sa->rtime.nsec = ct.tv_nsec;

4866 4867 4868 4869 4870 4871 4872 4873
	received_uuid_changed = memcmp(root_item->received_uuid, sa->uuid,
				       BTRFS_UUID_SIZE);
	if (received_uuid_changed &&
	    !btrfs_is_empty_uuid(root_item->received_uuid))
		btrfs_uuid_tree_rem(trans, root->fs_info->uuid_root,
				    root_item->received_uuid,
				    BTRFS_UUID_KEY_RECEIVED_SUBVOL,
				    root->root_key.objectid);
4874 4875 4876
	memcpy(root_item->received_uuid, sa->uuid, BTRFS_UUID_SIZE);
	btrfs_set_root_stransid(root_item, sa->stransid);
	btrfs_set_root_rtransid(root_item, sa->rtransid);
4877 4878 4879 4880
	btrfs_set_stack_timespec_sec(&root_item->stime, sa->stime.sec);
	btrfs_set_stack_timespec_nsec(&root_item->stime, sa->stime.nsec);
	btrfs_set_stack_timespec_sec(&root_item->rtime, sa->rtime.sec);
	btrfs_set_stack_timespec_nsec(&root_item->rtime, sa->rtime.nsec);
4881 4882 4883 4884 4885 4886

	ret = btrfs_update_root(trans, root->fs_info->tree_root,
				&root->root_key, &root->root_item);
	if (ret < 0) {
		btrfs_end_transaction(trans, root);
		goto out;
4887 4888 4889 4890 4891 4892 4893 4894
	}
	if (received_uuid_changed && !btrfs_is_empty_uuid(sa->uuid)) {
		ret = btrfs_uuid_tree_add(trans, root->fs_info->uuid_root,
					  sa->uuid,
					  BTRFS_UUID_KEY_RECEIVED_SUBVOL,
					  root->root_key.objectid);
		if (ret < 0 && ret != -EEXIST) {
			btrfs_abort_transaction(trans, root, ret);
4895
			goto out;
4896 4897 4898 4899 4900 4901
		}
	}
	ret = btrfs_commit_transaction(trans, root);
	if (ret < 0) {
		btrfs_abort_transaction(trans, root, ret);
		goto out;
4902 4903
	}

4904 4905 4906 4907 4908 4909 4910 4911 4912 4913 4914 4915 4916 4917 4918 4919 4920 4921 4922 4923 4924 4925
out:
	up_write(&root->fs_info->subvol_sem);
	mnt_drop_write_file(file);
	return ret;
}

#ifdef CONFIG_64BIT
static long btrfs_ioctl_set_received_subvol_32(struct file *file,
						void __user *arg)
{
	struct btrfs_ioctl_received_subvol_args_32 *args32 = NULL;
	struct btrfs_ioctl_received_subvol_args *args64 = NULL;
	int ret = 0;

	args32 = memdup_user(arg, sizeof(*args32));
	if (IS_ERR(args32)) {
		ret = PTR_ERR(args32);
		args32 = NULL;
		goto out;
	}

	args64 = kmalloc(sizeof(*args64), GFP_NOFS);
4926 4927
	if (!args64) {
		ret = -ENOMEM;
4928 4929 4930 4931 4932 4933 4934 4935 4936 4937 4938 4939 4940 4941 4942 4943 4944 4945 4946 4947 4948 4949 4950 4951 4952 4953 4954 4955 4956 4957 4958 4959 4960 4961 4962 4963 4964 4965 4966 4967 4968 4969 4970 4971 4972 4973 4974 4975 4976 4977 4978 4979 4980 4981
		goto out;
	}

	memcpy(args64->uuid, args32->uuid, BTRFS_UUID_SIZE);
	args64->stransid = args32->stransid;
	args64->rtransid = args32->rtransid;
	args64->stime.sec = args32->stime.sec;
	args64->stime.nsec = args32->stime.nsec;
	args64->rtime.sec = args32->rtime.sec;
	args64->rtime.nsec = args32->rtime.nsec;
	args64->flags = args32->flags;

	ret = _btrfs_ioctl_set_received_subvol(file, args64);
	if (ret)
		goto out;

	memcpy(args32->uuid, args64->uuid, BTRFS_UUID_SIZE);
	args32->stransid = args64->stransid;
	args32->rtransid = args64->rtransid;
	args32->stime.sec = args64->stime.sec;
	args32->stime.nsec = args64->stime.nsec;
	args32->rtime.sec = args64->rtime.sec;
	args32->rtime.nsec = args64->rtime.nsec;
	args32->flags = args64->flags;

	ret = copy_to_user(arg, args32, sizeof(*args32));
	if (ret)
		ret = -EFAULT;

out:
	kfree(args32);
	kfree(args64);
	return ret;
}
#endif

static long btrfs_ioctl_set_received_subvol(struct file *file,
					    void __user *arg)
{
	struct btrfs_ioctl_received_subvol_args *sa = NULL;
	int ret = 0;

	sa = memdup_user(arg, sizeof(*sa));
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		sa = NULL;
		goto out;
	}

	ret = _btrfs_ioctl_set_received_subvol(file, sa);

	if (ret)
		goto out;

4982 4983 4984 4985 4986 4987 4988 4989 4990
	ret = copy_to_user(arg, sa, sizeof(*sa));
	if (ret)
		ret = -EFAULT;

out:
	kfree(sa);
	return ret;
}

4991 4992
static int btrfs_ioctl_get_fslabel(struct file *file, void __user *arg)
{
A
Al Viro 已提交
4993
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
4994
	size_t len;
4995
	int ret;
4996 4997 4998 4999 5000 5001 5002
	char label[BTRFS_LABEL_SIZE];

	spin_lock(&root->fs_info->super_lock);
	memcpy(label, root->fs_info->super_copy->label, BTRFS_LABEL_SIZE);
	spin_unlock(&root->fs_info->super_lock);

	len = strnlen(label, BTRFS_LABEL_SIZE);
5003 5004

	if (len == BTRFS_LABEL_SIZE) {
5005 5006
		btrfs_warn(root->fs_info,
			"label is too long, return the first %zu bytes", --len);
5007 5008 5009 5010 5011 5012 5013
	}

	ret = copy_to_user(arg, label, len);

	return ret ? -EFAULT : 0;
}

5014 5015
static int btrfs_ioctl_set_fslabel(struct file *file, void __user *arg)
{
A
Al Viro 已提交
5016
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
5017 5018 5019 5020 5021 5022 5023 5024 5025 5026 5027 5028
	struct btrfs_super_block *super_block = root->fs_info->super_copy;
	struct btrfs_trans_handle *trans;
	char label[BTRFS_LABEL_SIZE];
	int ret;

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

	if (copy_from_user(label, arg, sizeof(label)))
		return -EFAULT;

	if (strnlen(label, BTRFS_LABEL_SIZE) == BTRFS_LABEL_SIZE) {
5029
		btrfs_err(root->fs_info, "unable to set label with more than %d bytes",
5030 5031 5032 5033 5034 5035 5036 5037 5038 5039 5040 5041 5042 5043
		       BTRFS_LABEL_SIZE - 1);
		return -EINVAL;
	}

	ret = mnt_want_write_file(file);
	if (ret)
		return ret;

	trans = btrfs_start_transaction(root, 0);
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto out_unlock;
	}

5044
	spin_lock(&root->fs_info->super_lock);
5045
	strcpy(super_block->label, label);
5046
	spin_unlock(&root->fs_info->super_lock);
5047
	ret = btrfs_commit_transaction(trans, root);
5048 5049 5050 5051 5052 5053

out_unlock:
	mnt_drop_write_file(file);
	return ret;
}

5054 5055 5056 5057 5058 5059 5060 5061 5062 5063 5064 5065 5066 5067 5068 5069 5070 5071 5072 5073 5074 5075 5076 5077 5078 5079 5080 5081 5082 5083 5084 5085 5086 5087 5088 5089
#define INIT_FEATURE_FLAGS(suffix) \
	{ .compat_flags = BTRFS_FEATURE_COMPAT_##suffix, \
	  .compat_ro_flags = BTRFS_FEATURE_COMPAT_RO_##suffix, \
	  .incompat_flags = BTRFS_FEATURE_INCOMPAT_##suffix }

static int btrfs_ioctl_get_supported_features(struct file *file,
					      void __user *arg)
{
	static struct btrfs_ioctl_feature_flags features[3] = {
		INIT_FEATURE_FLAGS(SUPP),
		INIT_FEATURE_FLAGS(SAFE_SET),
		INIT_FEATURE_FLAGS(SAFE_CLEAR)
	};

	if (copy_to_user(arg, &features, sizeof(features)))
		return -EFAULT;

	return 0;
}

static int btrfs_ioctl_get_features(struct file *file, void __user *arg)
{
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
	struct btrfs_super_block *super_block = root->fs_info->super_copy;
	struct btrfs_ioctl_feature_flags features;

	features.compat_flags = btrfs_super_compat_flags(super_block);
	features.compat_ro_flags = btrfs_super_compat_ro_flags(super_block);
	features.incompat_flags = btrfs_super_incompat_flags(super_block);

	if (copy_to_user(arg, &features, sizeof(features)))
		return -EFAULT;

	return 0;
}

5090 5091
static int check_feature_bits(struct btrfs_root *root,
			      enum btrfs_feature_set set,
5092 5093 5094
			      u64 change_mask, u64 flags, u64 supported_flags,
			      u64 safe_set, u64 safe_clear)
{
5095 5096
	const char *type = btrfs_feature_set_names[set];
	char *names;
5097 5098 5099 5100 5101 5102
	u64 disallowed, unsupported;
	u64 set_mask = flags & change_mask;
	u64 clear_mask = ~flags & change_mask;

	unsupported = set_mask & ~supported_flags;
	if (unsupported) {
5103 5104 5105 5106 5107 5108 5109 5110
		names = btrfs_printable_features(set, unsupported);
		if (names) {
			btrfs_warn(root->fs_info,
			   "this kernel does not support the %s feature bit%s",
			   names, strchr(names, ',') ? "s" : "");
			kfree(names);
		} else
			btrfs_warn(root->fs_info,
5111 5112 5113 5114 5115 5116 5117
			   "this kernel does not support %s bits 0x%llx",
			   type, unsupported);
		return -EOPNOTSUPP;
	}

	disallowed = set_mask & ~safe_set;
	if (disallowed) {
5118 5119 5120 5121 5122 5123 5124 5125
		names = btrfs_printable_features(set, disallowed);
		if (names) {
			btrfs_warn(root->fs_info,
			   "can't set the %s feature bit%s while mounted",
			   names, strchr(names, ',') ? "s" : "");
			kfree(names);
		} else
			btrfs_warn(root->fs_info,
5126 5127 5128 5129 5130 5131 5132
			   "can't set %s bits 0x%llx while mounted",
			   type, disallowed);
		return -EPERM;
	}

	disallowed = clear_mask & ~safe_clear;
	if (disallowed) {
5133 5134 5135 5136 5137 5138 5139 5140
		names = btrfs_printable_features(set, disallowed);
		if (names) {
			btrfs_warn(root->fs_info,
			   "can't clear the %s feature bit%s while mounted",
			   names, strchr(names, ',') ? "s" : "");
			kfree(names);
		} else
			btrfs_warn(root->fs_info,
5141 5142 5143 5144 5145 5146 5147 5148 5149
			   "can't clear %s bits 0x%llx while mounted",
			   type, disallowed);
		return -EPERM;
	}

	return 0;
}

#define check_feature(root, change_mask, flags, mask_base)	\
5150
check_feature_bits(root, FEAT_##mask_base, change_mask, flags,	\
5151 5152 5153 5154 5155 5156 5157 5158 5159 5160 5161 5162 5163 5164 5165 5166 5167 5168 5169 5170 5171 5172 5173 5174 5175 5176 5177 5178 5179 5180 5181 5182 5183 5184 5185 5186 5187 5188 5189
		   BTRFS_FEATURE_ ## mask_base ## _SUPP,	\
		   BTRFS_FEATURE_ ## mask_base ## _SAFE_SET,	\
		   BTRFS_FEATURE_ ## mask_base ## _SAFE_CLEAR)

static int btrfs_ioctl_set_features(struct file *file, void __user *arg)
{
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
	struct btrfs_super_block *super_block = root->fs_info->super_copy;
	struct btrfs_ioctl_feature_flags flags[2];
	struct btrfs_trans_handle *trans;
	u64 newflags;
	int ret;

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

	if (copy_from_user(flags, arg, sizeof(flags)))
		return -EFAULT;

	/* Nothing to do */
	if (!flags[0].compat_flags && !flags[0].compat_ro_flags &&
	    !flags[0].incompat_flags)
		return 0;

	ret = check_feature(root, flags[0].compat_flags,
			    flags[1].compat_flags, COMPAT);
	if (ret)
		return ret;

	ret = check_feature(root, flags[0].compat_ro_flags,
			    flags[1].compat_ro_flags, COMPAT_RO);
	if (ret)
		return ret;

	ret = check_feature(root, flags[0].incompat_flags,
			    flags[1].incompat_flags, INCOMPAT);
	if (ret)
		return ret;

5190
	trans = btrfs_start_transaction(root, 0);
5191 5192 5193 5194 5195 5196 5197 5198 5199 5200 5201 5202 5203 5204 5205 5206 5207 5208 5209 5210
	if (IS_ERR(trans))
		return PTR_ERR(trans);

	spin_lock(&root->fs_info->super_lock);
	newflags = btrfs_super_compat_flags(super_block);
	newflags |= flags[0].compat_flags & flags[1].compat_flags;
	newflags &= ~(flags[0].compat_flags & ~flags[1].compat_flags);
	btrfs_set_super_compat_flags(super_block, newflags);

	newflags = btrfs_super_compat_ro_flags(super_block);
	newflags |= flags[0].compat_ro_flags & flags[1].compat_ro_flags;
	newflags &= ~(flags[0].compat_ro_flags & ~flags[1].compat_ro_flags);
	btrfs_set_super_compat_ro_flags(super_block, newflags);

	newflags = btrfs_super_incompat_flags(super_block);
	newflags |= flags[0].incompat_flags & flags[1].incompat_flags;
	newflags &= ~(flags[0].incompat_flags & ~flags[1].incompat_flags);
	btrfs_set_super_incompat_flags(super_block, newflags);
	spin_unlock(&root->fs_info->super_lock);

5211
	return btrfs_commit_transaction(trans, root);
5212 5213
}

C
Christoph Hellwig 已提交
5214 5215 5216
long btrfs_ioctl(struct file *file, unsigned int
		cmd, unsigned long arg)
{
A
Al Viro 已提交
5217
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
5218
	void __user *argp = (void __user *)arg;
C
Christoph Hellwig 已提交
5219 5220

	switch (cmd) {
5221 5222 5223 5224 5225 5226
	case FS_IOC_GETFLAGS:
		return btrfs_ioctl_getflags(file, argp);
	case FS_IOC_SETFLAGS:
		return btrfs_ioctl_setflags(file, argp);
	case FS_IOC_GETVERSION:
		return btrfs_ioctl_getversion(file, argp);
5227 5228
	case FITRIM:
		return btrfs_ioctl_fitrim(file, argp);
C
Christoph Hellwig 已提交
5229
	case BTRFS_IOC_SNAP_CREATE:
5230
		return btrfs_ioctl_snap_create(file, argp, 0);
5231
	case BTRFS_IOC_SNAP_CREATE_V2:
5232
		return btrfs_ioctl_snap_create_v2(file, argp, 0);
5233
	case BTRFS_IOC_SUBVOL_CREATE:
5234
		return btrfs_ioctl_snap_create(file, argp, 1);
A
Arne Jansen 已提交
5235 5236
	case BTRFS_IOC_SUBVOL_CREATE_V2:
		return btrfs_ioctl_snap_create_v2(file, argp, 1);
5237 5238
	case BTRFS_IOC_SNAP_DESTROY:
		return btrfs_ioctl_snap_destroy(file, argp);
5239 5240 5241 5242
	case BTRFS_IOC_SUBVOL_GETFLAGS:
		return btrfs_ioctl_subvol_getflags(file, argp);
	case BTRFS_IOC_SUBVOL_SETFLAGS:
		return btrfs_ioctl_subvol_setflags(file, argp);
5243 5244
	case BTRFS_IOC_DEFAULT_SUBVOL:
		return btrfs_ioctl_default_subvol(file, argp);
C
Christoph Hellwig 已提交
5245
	case BTRFS_IOC_DEFRAG:
C
Chris Mason 已提交
5246 5247 5248
		return btrfs_ioctl_defrag(file, NULL);
	case BTRFS_IOC_DEFRAG_RANGE:
		return btrfs_ioctl_defrag(file, argp);
C
Christoph Hellwig 已提交
5249
	case BTRFS_IOC_RESIZE:
5250
		return btrfs_ioctl_resize(file, argp);
C
Christoph Hellwig 已提交
5251
	case BTRFS_IOC_ADD_DEV:
5252
		return btrfs_ioctl_add_dev(root, argp);
C
Christoph Hellwig 已提交
5253
	case BTRFS_IOC_RM_DEV:
5254
		return btrfs_ioctl_rm_dev(file, argp);
J
Jan Schmidt 已提交
5255 5256 5257 5258
	case BTRFS_IOC_FS_INFO:
		return btrfs_ioctl_fs_info(root, argp);
	case BTRFS_IOC_DEV_INFO:
		return btrfs_ioctl_dev_info(root, argp);
C
Christoph Hellwig 已提交
5259
	case BTRFS_IOC_BALANCE:
5260
		return btrfs_ioctl_balance(file, NULL);
C
Christoph Hellwig 已提交
5261
	case BTRFS_IOC_CLONE:
5262 5263
		return btrfs_ioctl_clone(file, arg, 0, 0, 0);
	case BTRFS_IOC_CLONE_RANGE:
5264
		return btrfs_ioctl_clone_range(file, argp);
C
Christoph Hellwig 已提交
5265 5266 5267 5268
	case BTRFS_IOC_TRANS_START:
		return btrfs_ioctl_trans_start(file);
	case BTRFS_IOC_TRANS_END:
		return btrfs_ioctl_trans_end(file);
5269 5270
	case BTRFS_IOC_TREE_SEARCH:
		return btrfs_ioctl_tree_search(file, argp);
G
Gerhard Heift 已提交
5271 5272
	case BTRFS_IOC_TREE_SEARCH_V2:
		return btrfs_ioctl_tree_search_v2(file, argp);
5273 5274
	case BTRFS_IOC_INO_LOOKUP:
		return btrfs_ioctl_ino_lookup(file, argp);
5275 5276 5277 5278
	case BTRFS_IOC_INO_PATHS:
		return btrfs_ioctl_ino_to_path(root, argp);
	case BTRFS_IOC_LOGICAL_INO:
		return btrfs_ioctl_logical_to_ino(root, argp);
J
Josef Bacik 已提交
5279 5280
	case BTRFS_IOC_SPACE_INFO:
		return btrfs_ioctl_space_info(root, argp);
5281 5282 5283
	case BTRFS_IOC_SYNC: {
		int ret;

5284
		ret = btrfs_start_delalloc_roots(root->fs_info, 0, -1);
5285 5286 5287
		if (ret)
			return ret;
		ret = btrfs_sync_fs(file->f_dentry->d_sb, 1);
5288 5289 5290 5291 5292 5293
		/*
		 * The transaction thread may want to do more work,
		 * namely it pokes the cleaner ktread that will start
		 * processing uncleaned subvols.
		 */
		wake_up_process(root->fs_info->transaction_kthread);
5294 5295
		return ret;
	}
5296
	case BTRFS_IOC_START_SYNC:
5297
		return btrfs_ioctl_start_sync(root, argp);
5298
	case BTRFS_IOC_WAIT_SYNC:
5299
		return btrfs_ioctl_wait_sync(root, argp);
J
Jan Schmidt 已提交
5300
	case BTRFS_IOC_SCRUB:
M
Miao Xie 已提交
5301
		return btrfs_ioctl_scrub(file, argp);
J
Jan Schmidt 已提交
5302 5303 5304 5305
	case BTRFS_IOC_SCRUB_CANCEL:
		return btrfs_ioctl_scrub_cancel(root, argp);
	case BTRFS_IOC_SCRUB_PROGRESS:
		return btrfs_ioctl_scrub_progress(root, argp);
5306
	case BTRFS_IOC_BALANCE_V2:
5307
		return btrfs_ioctl_balance(file, argp);
5308 5309
	case BTRFS_IOC_BALANCE_CTL:
		return btrfs_ioctl_balance_ctl(root, arg);
5310 5311
	case BTRFS_IOC_BALANCE_PROGRESS:
		return btrfs_ioctl_balance_progress(root, argp);
5312 5313
	case BTRFS_IOC_SET_RECEIVED_SUBVOL:
		return btrfs_ioctl_set_received_subvol(file, argp);
5314 5315 5316 5317
#ifdef CONFIG_64BIT
	case BTRFS_IOC_SET_RECEIVED_SUBVOL_32:
		return btrfs_ioctl_set_received_subvol_32(file, argp);
#endif
5318 5319
	case BTRFS_IOC_SEND:
		return btrfs_ioctl_send(file, argp);
5320
	case BTRFS_IOC_GET_DEV_STATS:
5321
		return btrfs_ioctl_get_dev_stats(root, argp);
A
Arne Jansen 已提交
5322
	case BTRFS_IOC_QUOTA_CTL:
5323
		return btrfs_ioctl_quota_ctl(file, argp);
A
Arne Jansen 已提交
5324
	case BTRFS_IOC_QGROUP_ASSIGN:
5325
		return btrfs_ioctl_qgroup_assign(file, argp);
A
Arne Jansen 已提交
5326
	case BTRFS_IOC_QGROUP_CREATE:
5327
		return btrfs_ioctl_qgroup_create(file, argp);
A
Arne Jansen 已提交
5328
	case BTRFS_IOC_QGROUP_LIMIT:
5329
		return btrfs_ioctl_qgroup_limit(file, argp);
J
Jan Schmidt 已提交
5330 5331 5332 5333
	case BTRFS_IOC_QUOTA_RESCAN:
		return btrfs_ioctl_quota_rescan(file, argp);
	case BTRFS_IOC_QUOTA_RESCAN_STATUS:
		return btrfs_ioctl_quota_rescan_status(file, argp);
5334 5335
	case BTRFS_IOC_QUOTA_RESCAN_WAIT:
		return btrfs_ioctl_quota_rescan_wait(file, argp);
5336 5337
	case BTRFS_IOC_DEV_REPLACE:
		return btrfs_ioctl_dev_replace(root, argp);
5338 5339
	case BTRFS_IOC_GET_FSLABEL:
		return btrfs_ioctl_get_fslabel(file, argp);
5340 5341
	case BTRFS_IOC_SET_FSLABEL:
		return btrfs_ioctl_set_fslabel(file, argp);
M
Mark Fasheh 已提交
5342 5343
	case BTRFS_IOC_FILE_EXTENT_SAME:
		return btrfs_ioctl_file_extent_same(file, argp);
5344 5345 5346 5347 5348 5349
	case BTRFS_IOC_GET_SUPPORTED_FEATURES:
		return btrfs_ioctl_get_supported_features(file, argp);
	case BTRFS_IOC_GET_FEATURES:
		return btrfs_ioctl_get_features(file, argp);
	case BTRFS_IOC_SET_FEATURES:
		return btrfs_ioctl_set_features(file, argp);
C
Christoph Hellwig 已提交
5350 5351 5352 5353
	}

	return -ENOTTY;
}