ioctl.c 107.8 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"
C
Christoph Hellwig 已提交
59

M
Mark Fasheh 已提交
60 61 62
static int btrfs_clone(struct inode *src, struct inode *inode,
		       u64 off, u64 olen, u64 olen_aligned, u64 destoff);

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 89 90 91 92
/* 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
93 94 95 96 97 98 99
	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;
100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127

	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);

	inode->i_flags &= ~(S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC);

	if (ip->flags & BTRFS_INODE_SYNC)
		inode->i_flags |= S_SYNC;
	if (ip->flags & BTRFS_INODE_IMMUTABLE)
		inode->i_flags |= S_IMMUTABLE;
	if (ip->flags & BTRFS_INODE_APPEND)
		inode->i_flags |= S_APPEND;
	if (ip->flags & BTRFS_INODE_NOATIME)
		inode->i_flags |= S_NOATIME;
	if (ip->flags & BTRFS_INODE_DIRSYNC)
		inode->i_flags |= S_DIRSYNC;
}

/*
 * Inherit flags from the parent inode.
 *
128
 * Currently only the compression flags and the cow flags are inherited.
129 130 131
 */
void btrfs_inherit_iflags(struct inode *inode, struct inode *dir)
{
132 133 134 135 136 137
	unsigned int flags;

	if (!dir)
		return;

	flags = BTRFS_I(dir)->flags;
138

139 140 141 142 143 144 145 146
	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
147
	if (flags & BTRFS_INODE_NODATACOW) {
148
		BTRFS_I(inode)->flags |= BTRFS_INODE_NODATACOW;
L
Liu Bo 已提交
149 150 151
		if (S_ISREG(inode->i_mode))
			BTRFS_I(inode)->flags |= BTRFS_INODE_NODATASUM;
	}
152 153 154 155 156 157

	btrfs_update_iflags(inode);
}

static int btrfs_ioctl_getflags(struct file *file, void __user *arg)
{
A
Al Viro 已提交
158
	struct btrfs_inode *ip = BTRFS_I(file_inode(file));
159 160 161 162 163 164 165
	unsigned int flags = btrfs_flags_to_ioctl(ip->flags);

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

166 167 168 169 170
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
171 172
		      FS_NOCOMP_FL | FS_COMPR_FL |
		      FS_NOCOW_FL))
173 174 175 176 177 178 179 180
		return -EOPNOTSUPP;

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

	return 0;
}

181 182
static int btrfs_ioctl_setflags(struct file *file, void __user *arg)
{
A
Al Viro 已提交
183
	struct inode *inode = file_inode(file);
184 185 186 187 188
	struct btrfs_inode *ip = BTRFS_I(inode);
	struct btrfs_root *root = ip->root;
	struct btrfs_trans_handle *trans;
	unsigned int flags, oldflags;
	int ret;
189 190
	u64 ip_oldflags;
	unsigned int i_oldflags;
191
	umode_t mode;
192

L
Li Zefan 已提交
193 194 195
	if (btrfs_root_readonly(root))
		return -EROFS;

196 197 198
	if (copy_from_user(&flags, arg, sizeof(flags)))
		return -EFAULT;

199 200 201
	ret = check_flags(flags);
	if (ret)
		return ret;
C
Christoph Hellwig 已提交
202

203
	if (!inode_owner_or_capable(inode))
204 205
		return -EACCES;

206 207 208 209
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;

210 211
	mutex_lock(&inode->i_mutex);

212 213
	ip_oldflags = ip->flags;
	i_oldflags = inode->i_flags;
214
	mode = inode->i_mode;
215

216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248
	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;
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
	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;
		}
	}
274

275 276 277 278 279 280 281 282 283 284 285
	/*
	 * 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;
	} else if (flags & FS_COMPR_FL) {
		ip->flags |= BTRFS_INODE_COMPRESS;
		ip->flags &= ~BTRFS_INODE_NOCOMPRESS;
L
Li Zefan 已提交
286 287
	} else {
		ip->flags &= ~(BTRFS_INODE_COMPRESS | BTRFS_INODE_NOCOMPRESS);
288
	}
289

290
	trans = btrfs_start_transaction(root, 1);
291 292 293 294
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto out_drop;
	}
295

296
	btrfs_update_iflags(inode);
297
	inode_inc_iversion(inode);
298
	inode->i_ctime = CURRENT_TIME;
299 300 301
	ret = btrfs_update_inode(trans, root, inode);

	btrfs_end_transaction(trans, root);
302 303 304 305 306
 out_drop:
	if (ret) {
		ip->flags = ip_oldflags;
		inode->i_flags = i_oldflags;
	}
307 308 309

 out_unlock:
	mutex_unlock(&inode->i_mutex);
310
	mnt_drop_write_file(file);
311
	return ret;
312 313 314 315
}

static int btrfs_ioctl_getversion(struct file *file, int __user *arg)
{
A
Al Viro 已提交
316
	struct inode *inode = file_inode(file);
317 318 319

	return put_user(inode->i_generation, arg);
}
C
Christoph Hellwig 已提交
320

321 322
static noinline int btrfs_ioctl_fitrim(struct file *file, void __user *arg)
{
323
	struct btrfs_fs_info *fs_info = btrfs_sb(fdentry(file)->d_sb);
324 325 326 327 328
	struct btrfs_device *device;
	struct request_queue *q;
	struct fstrim_range range;
	u64 minlen = ULLONG_MAX;
	u64 num_devices = 0;
329
	u64 total_bytes = btrfs_super_total_bytes(fs_info->super_copy);
330 331 332 333 334
	int ret;

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

335 336 337
	rcu_read_lock();
	list_for_each_entry_rcu(device, &fs_info->fs_devices->devices,
				dev_list) {
338 339 340 341 342 343 344 345 346
		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);
		}
	}
347
	rcu_read_unlock();
348

349 350 351 352
	if (!num_devices)
		return -EOPNOTSUPP;
	if (copy_from_user(&range, arg, sizeof(range)))
		return -EFAULT;
353 354
	if (range.start > total_bytes ||
	    range.len < fs_info->sb->s_blocksize)
355
		return -EINVAL;
356

357
	range.len = min(range.len, total_bytes - range.start);
358
	range.minlen = max(range.minlen, minlen);
359
	ret = btrfs_trim_fs(fs_info->tree_root, &range);
360 361 362 363 364 365 366 367 368
	if (ret < 0)
		return ret;

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

	return 0;
}

369 370
int btrfs_is_empty_uuid(u8 *uuid)
{
C
Chris Mason 已提交
371 372 373 374 375 376 377
	int i;

	for (i = 0; i < BTRFS_UUID_SIZE; i++) {
		if (uuid[i])
			return 0;
	}
	return 1;
378 379
}

380
static noinline int create_subvol(struct inode *dir,
381
				  struct dentry *dentry,
S
Sage Weil 已提交
382
				  char *name, int namelen,
A
Arne Jansen 已提交
383
				  u64 *async_transid,
384
				  struct btrfs_qgroup_inherit *inherit)
C
Christoph Hellwig 已提交
385 386 387 388 389 390
{
	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;
391
	struct btrfs_root *root = BTRFS_I(dir)->root;
392
	struct btrfs_root *new_root;
393
	struct btrfs_block_rsv block_rsv;
394
	struct timespec cur_time = CURRENT_TIME;
C
Christoph Hellwig 已提交
395 396 397 398
	int ret;
	int err;
	u64 objectid;
	u64 new_dirid = BTRFS_FIRST_FREE_OBJECTID;
399
	u64 index = 0;
400
	u64 qgroup_reserved;
401
	uuid_le new_uuid;
C
Christoph Hellwig 已提交
402

403
	ret = btrfs_find_free_objectid(root->fs_info->tree_root, &objectid);
404
	if (ret)
405
		return ret;
406

407
	btrfs_init_block_rsv(&block_rsv, BTRFS_BLOCK_RSV_TEMP);
J
Josef Bacik 已提交
408
	/*
409 410
	 * The same as the snapshot creation, please see the comment
	 * of create_snapshot().
J
Josef Bacik 已提交
411
	 */
412
	ret = btrfs_subvolume_reserve_metadata(root, &block_rsv,
413
					       8, &qgroup_reserved, false);
414 415 416 417 418 419 420 421 422 423
	if (ret)
		return ret;

	trans = btrfs_start_transaction(root, 0);
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto out;
	}
	trans->block_rsv = &block_rsv;
	trans->bytes_reserved = block_rsv.size;
C
Christoph Hellwig 已提交
424

425
	ret = btrfs_qgroup_inherit(trans, root->fs_info, 0, objectid, inherit);
A
Arne Jansen 已提交
426 427 428
	if (ret)
		goto fail;

429
	leaf = btrfs_alloc_free_block(trans, root, root->leafsize,
430
				      0, objectid, NULL, 0, 0, 0);
431 432 433 434
	if (IS_ERR(leaf)) {
		ret = PTR_ERR(leaf);
		goto fail;
	}
C
Christoph Hellwig 已提交
435

436
	memset_extent_buffer(leaf, 0, 0, sizeof(struct btrfs_header));
C
Christoph Hellwig 已提交
437 438
	btrfs_set_header_bytenr(leaf, leaf->start);
	btrfs_set_header_generation(leaf, trans->transid);
439
	btrfs_set_header_backref_rev(leaf, BTRFS_MIXED_BACKREF_REV);
C
Christoph Hellwig 已提交
440 441
	btrfs_set_header_owner(leaf, objectid);

442
	write_extent_buffer(leaf, root->fs_info->fsid, btrfs_header_fsid(),
C
Christoph Hellwig 已提交
443
			    BTRFS_FSID_SIZE);
444
	write_extent_buffer(leaf, root->fs_info->chunk_tree_uuid,
445
			    btrfs_header_chunk_tree_uuid(leaf),
446
			    BTRFS_UUID_SIZE);
C
Christoph Hellwig 已提交
447 448
	btrfs_mark_buffer_dirty(leaf);

449 450
	memset(&root_item, 0, sizeof(root_item));

C
Christoph Hellwig 已提交
451
	inode_item = &root_item.inode;
452 453 454 455 456
	btrfs_set_stack_inode_generation(inode_item, 1);
	btrfs_set_stack_inode_size(inode_item, 3);
	btrfs_set_stack_inode_nlink(inode_item, 1);
	btrfs_set_stack_inode_nbytes(inode_item, root->leafsize);
	btrfs_set_stack_inode_mode(inode_item, S_IFDIR | 0755);
C
Christoph Hellwig 已提交
457

458 459 460
	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);
461

C
Christoph Hellwig 已提交
462
	btrfs_set_root_bytenr(&root_item, leaf->start);
463
	btrfs_set_root_generation(&root_item, trans->transid);
C
Christoph Hellwig 已提交
464 465
	btrfs_set_root_level(&root_item, 0);
	btrfs_set_root_refs(&root_item, 1);
466
	btrfs_set_root_used(&root_item, leaf->len);
Y
Yan Zheng 已提交
467
	btrfs_set_root_last_snapshot(&root_item, 0);
C
Christoph Hellwig 已提交
468

469 470 471 472
	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);
473 474
	btrfs_set_stack_timespec_sec(&root_item.otime, cur_time.tv_sec);
	btrfs_set_stack_timespec_nsec(&root_item.otime, cur_time.tv_nsec);
475 476 477
	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
478

479
	btrfs_tree_unlock(leaf);
C
Christoph Hellwig 已提交
480 481 482 483 484 485
	free_extent_buffer(leaf);
	leaf = NULL;

	btrfs_set_root_dirid(&root_item, new_dirid);

	key.objectid = objectid;
486
	key.offset = 0;
C
Christoph Hellwig 已提交
487 488 489 490 491 492
	btrfs_set_key_type(&key, BTRFS_ROOT_ITEM_KEY);
	ret = btrfs_insert_root(trans, root->fs_info->tree_root, &key,
				&root_item);
	if (ret)
		goto fail;

493 494
	key.offset = (u64)-1;
	new_root = btrfs_read_fs_root_no_name(root->fs_info, &key);
495 496 497 498 499
	if (IS_ERR(new_root)) {
		btrfs_abort_transaction(trans, root, PTR_ERR(new_root));
		ret = PTR_ERR(new_root);
		goto fail;
	}
500 501 502

	btrfs_record_root_in_trans(trans, new_root);

503
	ret = btrfs_create_subvol_root(trans, new_root, new_dirid);
504 505
	if (ret) {
		/* We potentially lose an unused inode item here */
506
		btrfs_abort_transaction(trans, root, ret);
507 508 509
		goto fail;
	}

C
Christoph Hellwig 已提交
510 511 512
	/*
	 * insert the directory item
	 */
513
	ret = btrfs_set_inode_index(dir, &index);
514 515 516 517
	if (ret) {
		btrfs_abort_transaction(trans, root, ret);
		goto fail;
	}
518 519

	ret = btrfs_insert_dir_item(trans, root,
520
				    name, namelen, dir, &key,
521
				    BTRFS_FT_DIR, index);
522 523
	if (ret) {
		btrfs_abort_transaction(trans, root, ret);
C
Christoph Hellwig 已提交
524
		goto fail;
525
	}
526

527 528 529 530
	btrfs_i_size_write(dir, dir->i_size + namelen * 2);
	ret = btrfs_update_inode(trans, root, dir);
	BUG_ON(ret);

531
	ret = btrfs_add_root_ref(trans, root->fs_info->tree_root,
532
				 objectid, root->root_key.objectid,
L
Li Zefan 已提交
533
				 btrfs_ino(dir), index, name, namelen);
534
	BUG_ON(ret);
C
Christoph Hellwig 已提交
535

536 537 538 539 540 541
	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
542
fail:
543 544
	trans->block_rsv = NULL;
	trans->bytes_reserved = 0;
S
Sage Weil 已提交
545 546 547
	if (async_transid) {
		*async_transid = trans->transid;
		err = btrfs_commit_transaction_async(trans, root, 1);
548 549
		if (err)
			err = btrfs_commit_transaction(trans, root);
S
Sage Weil 已提交
550 551 552
	} else {
		err = btrfs_commit_transaction(trans, root);
	}
C
Christoph Hellwig 已提交
553 554
	if (err && !ret)
		ret = err;
555 556 557

	if (!ret)
		d_instantiate(dentry, btrfs_lookup_dentry(dir, dentry));
558 559
out:
	btrfs_subvolume_release_metadata(root, &block_rsv, qgroup_reserved);
C
Christoph Hellwig 已提交
560 561 562
	return ret;
}

563 564 565 566
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
567
{
568
	struct inode *inode;
C
Christoph Hellwig 已提交
569 570
	struct btrfs_pending_snapshot *pending_snapshot;
	struct btrfs_trans_handle *trans;
571
	int ret;
C
Christoph Hellwig 已提交
572 573 574 575

	if (!root->ref_cows)
		return -EINVAL;

576 577 578 579
	ret = btrfs_start_delalloc_inodes(root, 0);
	if (ret)
		return ret;

580
	btrfs_wait_ordered_extents(root, -1);
581

582
	pending_snapshot = kzalloc(sizeof(*pending_snapshot), GFP_NOFS);
583 584 585
	if (!pending_snapshot)
		return -ENOMEM;

586 587
	btrfs_init_block_rsv(&pending_snapshot->block_rsv,
			     BTRFS_BLOCK_RSV_TEMP);
588 589 590 591 592 593
	/*
	 * 1 - parent dir inode
	 * 2 - dir entries
	 * 1 - root item
	 * 2 - root ref/backref
	 * 1 - root of snapshot
594
	 * 1 - UUID item
595 596
	 */
	ret = btrfs_subvolume_reserve_metadata(BTRFS_I(dir)->root,
597
					&pending_snapshot->block_rsv, 8,
598 599
					&pending_snapshot->qgroup_reserved,
					false);
600 601 602
	if (ret)
		goto out;

603
	pending_snapshot->dentry = dentry;
C
Christoph Hellwig 已提交
604
	pending_snapshot->root = root;
L
Li Zefan 已提交
605
	pending_snapshot->readonly = readonly;
606
	pending_snapshot->dir = dir;
607
	pending_snapshot->inherit = inherit;
608

609
	trans = btrfs_start_transaction(root, 0);
610 611 612 613 614
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto fail;
	}

615
	spin_lock(&root->fs_info->trans_lock);
C
Christoph Hellwig 已提交
616 617
	list_add(&pending_snapshot->list,
		 &trans->transaction->pending_snapshots);
618
	spin_unlock(&root->fs_info->trans_lock);
S
Sage Weil 已提交
619 620 621 622
	if (async_transid) {
		*async_transid = trans->transid;
		ret = btrfs_commit_transaction_async(trans,
				     root->fs_info->extent_root, 1);
623 624
		if (ret)
			ret = btrfs_commit_transaction(trans, root);
S
Sage Weil 已提交
625 626 627 628
	} else {
		ret = btrfs_commit_transaction(trans,
					       root->fs_info->extent_root);
	}
629
	if (ret)
630
		goto fail;
631 632 633 634 635

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

636 637 638
	ret = btrfs_orphan_cleanup(pending_snapshot->snap);
	if (ret)
		goto fail;
C
Christoph Hellwig 已提交
639

640
	inode = btrfs_lookup_dentry(dentry->d_parent->d_inode, dentry);
641 642 643 644 645 646 647 648
	if (IS_ERR(inode)) {
		ret = PTR_ERR(inode);
		goto fail;
	}
	BUG_ON(!inode);
	d_instantiate(dentry, inode);
	ret = 0;
fail:
649 650 651 652
	btrfs_subvolume_release_metadata(BTRFS_I(dir)->root,
					 &pending_snapshot->block_rsv,
					 pending_snapshot->qgroup_reserved);
out:
653
	kfree(pending_snapshot);
C
Christoph Hellwig 已提交
654 655 656
	return ret;
}

657 658 659 660 661 662
/*  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)
{
663
	kuid_t fsuid = current_fsuid();
664 665 666

	if (!(dir->i_mode & S_ISVTX))
		return 0;
667
	if (uid_eq(inode->i_uid, fsuid))
668
		return 0;
669
	if (uid_eq(dir->i_uid, fsuid))
670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693
		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().
 */

694
static int btrfs_may_delete(struct inode *dir, struct dentry *victim, int isdir)
695 696 697 698 699 700 701
{
	int error;

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

	BUG_ON(victim->d_parent->d_inode != dir);
702
	audit_inode_child(dir, victim, AUDIT_TYPE_CHILD_DELETE);
703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726

	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;
}

727 728 729 730 731 732 733 734 735 736 737 738 739 740 741
/* 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.
 */
742 743
static noinline int btrfs_mksubvol(struct path *parent,
				   char *name, int namelen,
S
Sage Weil 已提交
744
				   struct btrfs_root *snap_src,
A
Arne Jansen 已提交
745
				   u64 *async_transid, bool readonly,
746
				   struct btrfs_qgroup_inherit *inherit)
747
{
748
	struct inode *dir  = parent->dentry->d_inode;
749 750 751
	struct dentry *dentry;
	int error;

752 753 754
	error = mutex_lock_killable_nested(&dir->i_mutex, I_MUTEX_PARENT);
	if (error == -EINTR)
		return error;
755 756 757 758 759 760 761 762 763 764

	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;

765
	error = btrfs_may_create(dir, dentry);
766
	if (error)
767
		goto out_dput;
768

C
Chris Mason 已提交
769 770 771 772 773 774 775 776 777 778
	/*
	 * 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;

779 780 781 782 783
	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;

784
	if (snap_src) {
785
		error = create_snapshot(snap_src, dir, dentry, name, namelen,
A
Arne Jansen 已提交
786
					async_transid, readonly, inherit);
787
	} else {
788 789
		error = create_subvol(dir, dentry, name, namelen,
				      async_transid, inherit);
790
	}
791 792 793 794
	if (!error)
		fsnotify_mkdir(dir, dentry);
out_up_read:
	up_read(&BTRFS_I(dir)->root->fs_info->subvol_sem);
795 796 797
out_dput:
	dput(dentry);
out_unlock:
798
	mutex_unlock(&dir->i_mutex);
799 800 801
	return error;
}

C
Chris Mason 已提交
802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851
/*
 * 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
 */
static int check_defrag_in_cache(struct inode *inode, u64 offset, int thresh)
{
	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,
			    u64 *off, int thresh)
{
	struct btrfs_path *path;
	struct btrfs_key min_key;
	struct extent_buffer *leaf;
	struct btrfs_file_extent_item *extent;
	int type;
	int ret;
852
	u64 ino = btrfs_ino(inode);
C
Chris Mason 已提交
853 854 855 856 857

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

858
	min_key.objectid = ino;
C
Chris Mason 已提交
859 860 861 862 863
	min_key.type = BTRFS_EXTENT_DATA_KEY;
	min_key.offset = *off;

	path->keep_locks = 1;

864
	while (1) {
865
		ret = btrfs_search_forward(root, &min_key, path, newer_than);
C
Chris Mason 已提交
866 867
		if (ret != 0)
			goto none;
868
		if (min_key.objectid != ino)
C
Chris Mason 已提交
869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896
			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;
		}

		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
897
static struct extent_map *defrag_lookup_extent(struct inode *inode, u64 start)
898 899
{
	struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
L
Li Zefan 已提交
900 901 902
	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
	struct extent_map *em;
	u64 len = PAGE_CACHE_SIZE;
903

L
Li Zefan 已提交
904 905 906 907
	/*
	 * hopefully we have this extent in the tree already, try without
	 * the full extent lock
	 */
908
	read_lock(&em_tree->lock);
L
Li Zefan 已提交
909
	em = lookup_extent_mapping(em_tree, start, len);
910 911
	read_unlock(&em_tree->lock);

L
Li Zefan 已提交
912 913 914 915 916 917 918 919 920 921 922 923
	if (!em) {
		/* get the big lock and read metadata off disk */
		lock_extent(io_tree, start, start + len - 1);
		em = btrfs_get_extent(inode, NULL, 0, start, len, 0);
		unlock_extent(io_tree, start, start + len - 1);

		if (IS_ERR(em))
			return NULL;
	}

	return em;
}
924

L
Li Zefan 已提交
925 926 927 928 929 930 931 932 933 934 935 936 937 938
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);
	if (!next || next->block_start >= EXTENT_MAP_LAST_BYTE)
		ret = false;

	free_extent_map(next);
939 940 941
	return ret;
}

L
Li Zefan 已提交
942
static int should_defrag_range(struct inode *inode, u64 start, int thresh,
943 944
			       u64 *last_len, u64 *skip, u64 *defrag_end,
			       int compress)
945
{
L
Li Zefan 已提交
946
	struct extent_map *em;
947
	int ret = 1;
L
Li Zefan 已提交
948
	bool next_mergeable = true;
949 950

	/*
951
	 * make sure that once we start defragging an extent, we keep on
952 953 954 955 956 957 958
	 * defragging it
	 */
	if (start < *defrag_end)
		return 1;

	*skip = 0;

L
Li Zefan 已提交
959 960 961
	em = defrag_lookup_extent(inode, start);
	if (!em)
		return 0;
962 963

	/* this will cover holes, and inline extents */
964
	if (em->block_start >= EXTENT_MAP_LAST_BYTE) {
965
		ret = 0;
966 967 968
		goto out;
	}

L
Li Zefan 已提交
969
	next_mergeable = defrag_check_next_extent(inode, em);
970 971

	/*
L
Li Zefan 已提交
972 973
	 * we hit a real extent, if it is big or the next extent is not a
	 * real extent, don't bother defragging it
974
	 */
975
	if (!compress && (*last_len == 0 || *last_len >= thresh) &&
L
Li Zefan 已提交
976
	    (em->len >= thresh || !next_mergeable))
977
		ret = 0;
978
out:
979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998
	/*
	 * 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
999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014
/*
 * 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,
				    int num_pages)
C
Christoph Hellwig 已提交
1015
{
C
Chris Mason 已提交
1016 1017 1018 1019
	unsigned long file_end;
	u64 isize = i_size_read(inode);
	u64 page_start;
	u64 page_end;
1020
	u64 page_cnt;
C
Chris Mason 已提交
1021 1022 1023
	int ret;
	int i;
	int i_done;
1024
	struct btrfs_ordered_extent *ordered;
C
Chris Mason 已提交
1025
	struct extent_state *cached_state = NULL;
1026
	struct extent_io_tree *tree;
1027
	gfp_t mask = btrfs_alloc_write_mask(inode->i_mapping);
C
Chris Mason 已提交
1028 1029

	file_end = (isize - 1) >> PAGE_CACHE_SHIFT;
1030 1031 1032 1033
	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
1034 1035

	ret = btrfs_delalloc_reserve_space(inode,
1036
					   page_cnt << PAGE_CACHE_SHIFT);
C
Chris Mason 已提交
1037 1038 1039
	if (ret)
		return ret;
	i_done = 0;
1040
	tree = &BTRFS_I(inode)->io_tree;
C
Chris Mason 已提交
1041 1042

	/* step one, lock all the pages */
1043
	for (i = 0; i < page_cnt; i++) {
C
Chris Mason 已提交
1044
		struct page *page;
1045
again:
1046
		page = find_or_create_page(inode->i_mapping,
1047
					   start_index + i, mask);
C
Chris Mason 已提交
1048 1049 1050
		if (!page)
			break;

1051 1052 1053
		page_start = page_offset(page);
		page_end = page_start + PAGE_CACHE_SIZE - 1;
		while (1) {
1054
			lock_extent(tree, page_start, page_end);
1055 1056
			ordered = btrfs_lookup_ordered_extent(inode,
							      page_start);
1057
			unlock_extent(tree, page_start, page_end);
1058 1059 1060 1061 1062 1063 1064
			if (!ordered)
				break;

			unlock_page(page);
			btrfs_start_ordered_extent(inode, ordered, 1);
			btrfs_put_ordered_extent(ordered);
			lock_page(page);
1065 1066 1067 1068 1069 1070 1071 1072 1073
			/*
			 * 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;
			}
1074 1075
		}

C
Chris Mason 已提交
1076 1077 1078 1079 1080 1081 1082 1083 1084 1085
		if (!PageUptodate(page)) {
			btrfs_readpage(NULL, page);
			lock_page(page);
			if (!PageUptodate(page)) {
				unlock_page(page);
				page_cache_release(page);
				ret = -EIO;
				break;
			}
		}
1086 1087 1088 1089 1090 1091 1092

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

C
Chris Mason 已提交
1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112
		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,
1113
			 page_start, page_end - 1, 0, &cached_state);
C
Chris Mason 已提交
1114 1115
	clear_extent_bit(&BTRFS_I(inode)->io_tree, page_start,
			  page_end - 1, EXTENT_DIRTY | EXTENT_DELALLOC |
1116 1117
			  EXTENT_DO_ACCOUNTING | EXTENT_DEFRAG, 0, 0,
			  &cached_state, GFP_NOFS);
C
Chris Mason 已提交
1118

1119
	if (i_done != page_cnt) {
1120 1121 1122
		spin_lock(&BTRFS_I(inode)->lock);
		BTRFS_I(inode)->outstanding_extents++;
		spin_unlock(&BTRFS_I(inode)->lock);
C
Chris Mason 已提交
1123
		btrfs_delalloc_release_space(inode,
1124
				     (page_cnt - i_done) << PAGE_CACHE_SHIFT);
C
Chris Mason 已提交
1125 1126 1127
	}


1128 1129
	set_extent_defrag(&BTRFS_I(inode)->io_tree, page_start, page_end - 1,
			  &cached_state, GFP_NOFS);
C
Chris Mason 已提交
1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148

	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]);
	}
1149
	btrfs_delalloc_release_space(inode, page_cnt << PAGE_CACHE_SHIFT);
C
Chris Mason 已提交
1150 1151 1152 1153 1154 1155 1156 1157 1158 1159
	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
1160
	unsigned long last_index;
1161
	u64 isize = i_size_read(inode);
1162 1163 1164
	u64 last_len = 0;
	u64 skip = 0;
	u64 defrag_end = 0;
C
Chris Mason 已提交
1165
	u64 newer_off = range->start;
C
Christoph Hellwig 已提交
1166
	unsigned long i;
1167
	unsigned long ra_index = 0;
C
Christoph Hellwig 已提交
1168
	int ret;
C
Chris Mason 已提交
1169
	int defrag_count = 0;
1170
	int compress_type = BTRFS_COMPRESS_ZLIB;
C
Chris Mason 已提交
1171
	int extent_thresh = range->extent_thresh;
1172 1173
	int max_cluster = (256 * 1024) >> PAGE_CACHE_SHIFT;
	int cluster = max_cluster;
C
Chris Mason 已提交
1174 1175 1176
	u64 new_align = ~((u64)128 * 1024 - 1);
	struct page **pages = NULL;

1177 1178 1179 1180 1181
	if (isize == 0)
		return 0;

	if (range->start >= isize)
		return -EINVAL;
1182 1183 1184 1185 1186 1187 1188

	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
1189

1190 1191
	if (extent_thresh == 0)
		extent_thresh = 256 * 1024;
1192

C
Chris Mason 已提交
1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205
	/*
	 * 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;
	}

1206
	pages = kmalloc_array(max_cluster, sizeof(struct page *),
C
Chris Mason 已提交
1207 1208 1209 1210 1211 1212 1213
			GFP_NOFS);
	if (!pages) {
		ret = -ENOMEM;
		goto out_ra;
	}

	/* find the last page to defrag */
C
Chris Mason 已提交
1214
	if (range->start + range->len > range->start) {
1215
		last_index = min_t(u64, isize - 1,
C
Chris Mason 已提交
1216 1217
			 range->start + range->len - 1) >> PAGE_CACHE_SHIFT;
	} else {
1218
		last_index = (isize - 1) >> PAGE_CACHE_SHIFT;
C
Chris Mason 已提交
1219 1220
	}

C
Chris Mason 已提交
1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236
	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
1237
		max_to_defrag = last_index + 1;
C
Chris Mason 已提交
1238

L
Li Zefan 已提交
1239 1240 1241 1242 1243 1244 1245
	/*
	 * 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;

1246 1247 1248
	while (i <= last_index && defrag_count < max_to_defrag &&
	       (i < (i_size_read(inode) + PAGE_CACHE_SIZE - 1) >>
		PAGE_CACHE_SHIFT)) {
C
Chris Mason 已提交
1249 1250 1251 1252 1253 1254 1255
		/*
		 * make sure we stop running if someone unmounts
		 * the FS
		 */
		if (!(inode->i_sb->s_flags & MS_ACTIVE))
			break;

1256 1257 1258 1259 1260 1261
		if (btrfs_defrag_cancelled(root->fs_info)) {
			printk(KERN_DEBUG "btrfs: defrag_file cancelled\n");
			ret = -EAGAIN;
			break;
		}

1262
		if (!should_defrag_range(inode, (u64)i << PAGE_CACHE_SHIFT,
L
Li Zefan 已提交
1263
					 extent_thresh, &last_len, &skip,
1264 1265
					 &defrag_end, range->flags &
					 BTRFS_DEFRAG_RANGE_COMPRESS)) {
1266 1267 1268 1269 1270 1271 1272 1273 1274
			unsigned long next;
			/*
			 * the should_defrag function tells us how much to skip
			 * bump our counter by the suggested amount
			 */
			next = (skip + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
			i = max(i + 1, next);
			continue;
		}
1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289

		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;
		}
1290

1291
		mutex_lock(&inode->i_mutex);
1292 1293
		if (range->flags & BTRFS_DEFRAG_RANGE_COMPRESS)
			BTRFS_I(inode)->force_compress = compress_type;
1294
		ret = cluster_pages_for_defrag(inode, pages, i, cluster);
1295 1296
		if (ret < 0) {
			mutex_unlock(&inode->i_mutex);
C
Chris Mason 已提交
1297
			goto out_ra;
1298
		}
C
Chris Mason 已提交
1299 1300

		defrag_count += ret;
1301
		balance_dirty_pages_ratelimited(inode->i_mapping);
1302
		mutex_unlock(&inode->i_mutex);
C
Chris Mason 已提交
1303 1304 1305 1306 1307

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

1308 1309 1310
			if (ret > 0)
				i += ret;

C
Chris Mason 已提交
1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321
			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
1322
			}
C
Chris Mason 已提交
1323
		} else {
1324
			if (ret > 0) {
L
Li Zefan 已提交
1325
				i += ret;
1326 1327
				last_len += ret << PAGE_CACHE_SHIFT;
			} else {
L
Li Zefan 已提交
1328
				i++;
1329 1330
				last_len = 0;
			}
C
Christoph Hellwig 已提交
1331 1332 1333
		}
	}

C
Chris Mason 已提交
1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351
	if ((range->flags & BTRFS_DEFRAG_RANGE_START_IO))
		filemap_flush(inode->i_mapping);

	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);
	}

1352
	if (range->compress_type == BTRFS_COMPRESS_LZO) {
1353
		btrfs_set_fs_incompat(root->fs_info, COMPRESS_LZO);
1354 1355
	}

1356
	ret = defrag_count;
1357

C
Chris Mason 已提交
1358
out_ra:
1359 1360 1361 1362 1363
	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
1364 1365 1366
	if (!file)
		kfree(ra);
	kfree(pages);
1367
	return ret;
C
Christoph Hellwig 已提交
1368 1369
}

1370
static noinline int btrfs_ioctl_resize(struct file *file,
1371
					void __user *arg)
C
Christoph Hellwig 已提交
1372 1373 1374 1375
{
	u64 new_size;
	u64 old_size;
	u64 devid = 1;
A
Al Viro 已提交
1376
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
C
Christoph Hellwig 已提交
1377 1378 1379 1380 1381 1382 1383 1384
	struct btrfs_ioctl_vol_args *vol_args;
	struct btrfs_trans_handle *trans;
	struct btrfs_device *device = NULL;
	char *sizestr;
	char *devstr = NULL;
	int ret = 0;
	int mod = 0;

1385 1386 1387
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

1388 1389 1390 1391
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;

1392 1393
	if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
			1)) {
1394
		mnt_drop_write_file(file);
1395
		return BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
1396 1397
	}

1398
	mutex_lock(&root->fs_info->volume_mutex);
L
Li Zefan 已提交
1399
	vol_args = memdup_user(arg, sizeof(*vol_args));
1400 1401 1402 1403
	if (IS_ERR(vol_args)) {
		ret = PTR_ERR(vol_args);
		goto out;
	}
1404 1405

	vol_args->name[BTRFS_PATH_NAME_MAX] = '\0';
C
Christoph Hellwig 已提交
1406 1407 1408 1409 1410 1411 1412 1413 1414

	sizestr = vol_args->name;
	devstr = strchr(sizestr, ':');
	if (devstr) {
		char *end;
		sizestr = devstr + 1;
		*devstr = '\0';
		devstr = vol_args->name;
		devid = simple_strtoull(devstr, &end, 10);
1415 1416 1417 1418
		if (!devid) {
			ret = -EINVAL;
			goto out_free;
		}
1419
		printk(KERN_INFO "btrfs: resizing devid %llu\n", devid);
C
Christoph Hellwig 已提交
1420
	}
M
Miao Xie 已提交
1421

1422
	device = btrfs_find_device(root->fs_info, devid, NULL, NULL);
C
Christoph Hellwig 已提交
1423
	if (!device) {
1424
		printk(KERN_INFO "btrfs: resizer unable to find device %llu\n",
1425
		       devid);
1426
		ret = -ENODEV;
1427
		goto out_free;
C
Christoph Hellwig 已提交
1428
	}
M
Miao Xie 已提交
1429 1430

	if (!device->writeable) {
L
Liu Bo 已提交
1431
		printk(KERN_INFO "btrfs: resizer unable to apply on "
M
Miao Xie 已提交
1432
		       "readonly device %llu\n",
1433
		       devid);
1434
		ret = -EPERM;
L
Liu Bo 已提交
1435 1436 1437
		goto out_free;
	}

C
Christoph Hellwig 已提交
1438 1439 1440 1441 1442 1443 1444 1445 1446 1447
	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++;
		}
A
Akinobu Mita 已提交
1448
		new_size = memparse(sizestr, NULL);
C
Christoph Hellwig 已提交
1449 1450
		if (new_size == 0) {
			ret = -EINVAL;
1451
			goto out_free;
C
Christoph Hellwig 已提交
1452 1453 1454
		}
	}

1455
	if (device->is_tgtdev_for_dev_replace) {
1456
		ret = -EPERM;
1457 1458 1459
		goto out_free;
	}

C
Christoph Hellwig 已提交
1460 1461 1462 1463 1464
	old_size = device->total_bytes;

	if (mod < 0) {
		if (new_size > old_size) {
			ret = -EINVAL;
1465
			goto out_free;
C
Christoph Hellwig 已提交
1466 1467 1468 1469 1470 1471 1472 1473
		}
		new_size = old_size - new_size;
	} else if (mod > 0) {
		new_size = old_size + new_size;
	}

	if (new_size < 256 * 1024 * 1024) {
		ret = -EINVAL;
1474
		goto out_free;
C
Christoph Hellwig 已提交
1475 1476 1477
	}
	if (new_size > device->bdev->bd_inode->i_size) {
		ret = -EFBIG;
1478
		goto out_free;
C
Christoph Hellwig 已提交
1479 1480 1481 1482 1483
	}

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

1484
	printk_in_rcu(KERN_INFO "btrfs: new size for %s is %llu\n",
1485
		      rcu_str_deref(device->name), new_size);
C
Christoph Hellwig 已提交
1486 1487

	if (new_size > old_size) {
1488
		trans = btrfs_start_transaction(root, 0);
1489 1490
		if (IS_ERR(trans)) {
			ret = PTR_ERR(trans);
1491
			goto out_free;
1492
		}
C
Christoph Hellwig 已提交
1493 1494
		ret = btrfs_grow_device(trans, device, new_size);
		btrfs_commit_transaction(trans, root);
1495
	} else if (new_size < old_size) {
C
Christoph Hellwig 已提交
1496
		ret = btrfs_shrink_device(device, new_size);
1497
	} /* equal, nothing need to do */
C
Christoph Hellwig 已提交
1498

1499
out_free:
C
Christoph Hellwig 已提交
1500
	kfree(vol_args);
1501 1502
out:
	mutex_unlock(&root->fs_info->volume_mutex);
1503
	atomic_set(&root->fs_info->mutually_exclusive_operation_running, 0);
1504
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
1505 1506 1507
	return ret;
}

S
Sage Weil 已提交
1508
static noinline int btrfs_ioctl_snap_create_transid(struct file *file,
A
Arne Jansen 已提交
1509 1510
				char *name, unsigned long fd, int subvol,
				u64 *transid, bool readonly,
1511
				struct btrfs_qgroup_inherit *inherit)
C
Christoph Hellwig 已提交
1512 1513
{
	int namelen;
1514
	int ret = 0;
C
Christoph Hellwig 已提交
1515

1516 1517 1518 1519
	ret = mnt_want_write_file(file);
	if (ret)
		goto out;

S
Sage Weil 已提交
1520 1521
	namelen = strlen(name);
	if (strchr(name, '/')) {
C
Christoph Hellwig 已提交
1522
		ret = -EINVAL;
1523
		goto out_drop_write;
C
Christoph Hellwig 已提交
1524 1525
	}

1526 1527 1528
	if (name[0] == '.' &&
	   (namelen == 1 || (name[1] == '.' && namelen == 2))) {
		ret = -EEXIST;
1529
		goto out_drop_write;
1530 1531
	}

1532
	if (subvol) {
S
Sage Weil 已提交
1533
		ret = btrfs_mksubvol(&file->f_path, name, namelen,
A
Arne Jansen 已提交
1534
				     NULL, transid, readonly, inherit);
1535
	} else {
1536
		struct fd src = fdget(fd);
1537
		struct inode *src_inode;
1538
		if (!src.file) {
1539
			ret = -EINVAL;
1540
			goto out_drop_write;
1541 1542
		}

A
Al Viro 已提交
1543 1544
		src_inode = file_inode(src.file);
		if (src_inode->i_sb != file_inode(file)->i_sb) {
C
Chris Mason 已提交
1545 1546
			printk(KERN_INFO "btrfs: Snapshot src from "
			       "another FS\n");
1547
			ret = -EINVAL;
1548 1549 1550 1551
		} else {
			ret = btrfs_mksubvol(&file->f_path, name, namelen,
					     BTRFS_I(src_inode)->root,
					     transid, readonly, inherit);
1552
		}
1553
		fdput(src);
1554
	}
1555 1556
out_drop_write:
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
1557
out:
S
Sage Weil 已提交
1558 1559 1560 1561
	return ret;
}

static noinline int btrfs_ioctl_snap_create(struct file *file,
1562
					    void __user *arg, int subvol)
S
Sage Weil 已提交
1563
{
1564
	struct btrfs_ioctl_vol_args *vol_args;
S
Sage Weil 已提交
1565 1566
	int ret;

1567 1568 1569 1570
	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
1571

1572
	ret = btrfs_ioctl_snap_create_transid(file, vol_args->name,
L
Li Zefan 已提交
1573
					      vol_args->fd, subvol,
A
Arne Jansen 已提交
1574
					      NULL, false, NULL);
1575

1576 1577 1578
	kfree(vol_args);
	return ret;
}
1579

1580 1581 1582 1583 1584 1585 1586
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
1587
	bool readonly = false;
A
Arne Jansen 已提交
1588
	struct btrfs_qgroup_inherit *inherit = NULL;
1589

1590 1591 1592 1593
	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';
1594

L
Li Zefan 已提交
1595
	if (vol_args->flags &
A
Arne Jansen 已提交
1596 1597
	    ~(BTRFS_SUBVOL_CREATE_ASYNC | BTRFS_SUBVOL_RDONLY |
	      BTRFS_SUBVOL_QGROUP_INHERIT)) {
L
Li Zefan 已提交
1598
		ret = -EOPNOTSUPP;
1599
		goto out;
S
Sage Weil 已提交
1600
	}
1601 1602 1603

	if (vol_args->flags & BTRFS_SUBVOL_CREATE_ASYNC)
		ptr = &transid;
L
Li Zefan 已提交
1604 1605
	if (vol_args->flags & BTRFS_SUBVOL_RDONLY)
		readonly = true;
A
Arne Jansen 已提交
1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616
	if (vol_args->flags & BTRFS_SUBVOL_QGROUP_INHERIT) {
		if (vol_args->size > PAGE_CACHE_SIZE) {
			ret = -EINVAL;
			goto out;
		}
		inherit = memdup_user(vol_args->qgroup_inherit, vol_args->size);
		if (IS_ERR(inherit)) {
			ret = PTR_ERR(inherit);
			goto out;
		}
	}
1617 1618

	ret = btrfs_ioctl_snap_create_transid(file, vol_args->name,
A
Arne Jansen 已提交
1619
					      vol_args->fd, subvol, ptr,
1620
					      readonly, inherit);
1621 1622 1623 1624 1625 1626

	if (ret == 0 && ptr &&
	    copy_to_user(arg +
			 offsetof(struct btrfs_ioctl_vol_args_v2,
				  transid), ptr, sizeof(*ptr)))
		ret = -EFAULT;
1627
out:
C
Christoph Hellwig 已提交
1628
	kfree(vol_args);
A
Arne Jansen 已提交
1629
	kfree(inherit);
C
Christoph Hellwig 已提交
1630 1631 1632
	return ret;
}

1633 1634 1635
static noinline int btrfs_ioctl_subvol_getflags(struct file *file,
						void __user *arg)
{
A
Al Viro 已提交
1636
	struct inode *inode = file_inode(file);
1637 1638 1639 1640
	struct btrfs_root *root = BTRFS_I(inode)->root;
	int ret = 0;
	u64 flags = 0;

L
Li Zefan 已提交
1641
	if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID)
1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657
		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
1658
	struct inode *inode = file_inode(file);
1659 1660 1661 1662 1663 1664
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;
	u64 root_flags;
	u64 flags;
	int ret = 0;

1665 1666 1667
	ret = mnt_want_write_file(file);
	if (ret)
		goto out;
1668

1669 1670 1671 1672
	if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID) {
		ret = -EINVAL;
		goto out_drop_write;
	}
1673

1674 1675 1676 1677
	if (copy_from_user(&flags, arg, sizeof(flags))) {
		ret = -EFAULT;
		goto out_drop_write;
	}
1678

1679 1680 1681 1682
	if (flags & BTRFS_SUBVOL_CREATE_ASYNC) {
		ret = -EINVAL;
		goto out_drop_write;
	}
1683

1684 1685 1686 1687
	if (flags & ~BTRFS_SUBVOL_RDONLY) {
		ret = -EOPNOTSUPP;
		goto out_drop_write;
	}
1688

1689 1690 1691 1692
	if (!inode_owner_or_capable(inode)) {
		ret = -EACCES;
		goto out_drop_write;
	}
1693

1694 1695 1696 1697
	down_write(&root->fs_info->subvol_sem);

	/* nothing to do */
	if (!!(flags & BTRFS_SUBVOL_RDONLY) == btrfs_root_readonly(root))
1698
		goto out_drop_sem;
1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713

	root_flags = btrfs_root_flags(&root->root_item);
	if (flags & BTRFS_SUBVOL_RDONLY)
		btrfs_set_root_flags(&root->root_item,
				     root_flags | BTRFS_ROOT_SUBVOL_RDONLY);
	else
		btrfs_set_root_flags(&root->root_item,
				     root_flags & ~BTRFS_ROOT_SUBVOL_RDONLY);

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

1714
	ret = btrfs_update_root(trans, root->fs_info->tree_root,
1715 1716 1717 1718 1719 1720
				&root->root_key, &root->root_item);

	btrfs_commit_transaction(trans, root);
out_reset:
	if (ret)
		btrfs_set_root_flags(&root->root_item, root_flags);
1721
out_drop_sem:
1722
	up_write(&root->fs_info->subvol_sem);
1723 1724 1725
out_drop_write:
	mnt_drop_write_file(file);
out:
1726 1727 1728
	return ret;
}

1729 1730 1731 1732 1733 1734
/*
 * helper to check if the subvolume references other subvolumes
 */
static noinline int may_destroy_subvol(struct btrfs_root *root)
{
	struct btrfs_path *path;
1735
	struct btrfs_dir_item *di;
1736
	struct btrfs_key key;
1737
	u64 dir_id;
1738 1739 1740 1741 1742 1743
	int ret;

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

1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756
	/* 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) {
			ret = -ENOTEMPTY;
			goto out;
		}
		btrfs_release_path(path);
	}

1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779
	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;
}

1780 1781 1782
static noinline int key_in_sk(struct btrfs_key *key,
			      struct btrfs_ioctl_search_key *sk)
{
1783 1784 1785 1786 1787 1788 1789 1790 1791
	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)
1792
		return 0;
1793 1794 1795 1796 1797 1798 1799

	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)
1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835
		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,
			       char *buf,
			       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);

1836 1837 1838 1839 1840
		btrfs_item_key_to_cpu(leaf, key, i);
		if (!key_in_sk(key, sk))
			continue;

		if (sizeof(sh) + item_len > BTRFS_SEARCH_ARGS_BUFSIZE)
1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865
			item_len = 0;

		if (sizeof(sh) + item_len + *sk_offset >
		    BTRFS_SEARCH_ARGS_BUFSIZE) {
			ret = 1;
			goto overflow;
		}

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

		/* copy search result header */
		memcpy(buf + *sk_offset, &sh, sizeof(sh));
		*sk_offset += sizeof(sh);

		if (item_len) {
			char *p = buf + *sk_offset;
			/* copy the item */
			read_extent_buffer(leaf, p,
					   item_off, item_len);
			*sk_offset += item_len;
		}
1866
		(*num_found)++;
1867 1868 1869 1870 1871

		if (*num_found >= sk->nr_items)
			break;
	}
advance_key:
1872 1873
	ret = 0;
	if (key->offset < (u64)-1 && key->offset < sk->max_offset)
1874
		key->offset++;
1875 1876
	else if (key->type < (u8)-1 && key->type < sk->max_type) {
		key->offset = 0;
1877
		key->type++;
1878 1879 1880
	} else if (key->objectid < (u64)-1 && key->objectid < sk->max_objectid) {
		key->offset = 0;
		key->type = 0;
1881
		key->objectid++;
1882 1883
	} else
		ret = 1;
1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925
overflow:
	return ret;
}

static noinline int search_ioctl(struct inode *inode,
				 struct btrfs_ioctl_search_args *args)
{
	struct btrfs_root *root;
	struct btrfs_key key;
	struct btrfs_path *path;
	struct btrfs_ioctl_search_key *sk = &args->key;
	struct btrfs_fs_info *info = BTRFS_I(inode)->root->fs_info;
	int ret;
	int num_found = 0;
	unsigned long sk_offset = 0;

	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)) {
			printk(KERN_ERR "could not find root %llu\n",
			       sk->tree_id);
			btrfs_free_path(path);
			return -ENOENT;
		}
	}

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

	path->keep_locks = 1;

1926
	while (1) {
1927
		ret = btrfs_search_forward(root, &key, path, sk->min_transid);
1928 1929 1930 1931 1932 1933 1934
		if (ret != 0) {
			if (ret > 0)
				ret = 0;
			goto err;
		}
		ret = copy_to_sk(root, path, &key, sk, args->buf,
				 &sk_offset, &num_found);
1935
		btrfs_release_path(path);
1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956
		if (ret || num_found >= sk->nr_items)
			break;

	}
	ret = 0;
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)
{
	 struct btrfs_ioctl_search_args *args;
	 struct inode *inode;
	 int ret;

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

J
Julia Lawall 已提交
1957 1958 1959
	args = memdup_user(argp, sizeof(*args));
	if (IS_ERR(args))
		return PTR_ERR(args);
1960

A
Al Viro 已提交
1961
	inode = file_inode(file);
1962 1963 1964 1965 1966 1967 1968
	ret = search_ioctl(inode, args);
	if (ret == 0 && copy_to_user(argp, args, sizeof(*args)))
		ret = -EFAULT;
	kfree(args);
	return ret;
}

1969
/*
1970 1971 1972
 * Search INODE_REFs to identify path name of 'dirid' directory
 * in a 'tree_id' tree. and sets path name to 'name'.
 */
1973 1974 1975 1976 1977
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;
1978
	char *ptr;
1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995
	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;

1996
	ptr = &name[BTRFS_INO_LOOKUP_PATH_MAX];
1997 1998 1999 2000 2001 2002 2003

	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)) {
		printk(KERN_ERR "could not find root %llu\n", tree_id);
2004 2005
		ret = -ENOENT;
		goto out;
2006 2007 2008 2009
	}

	key.objectid = dirid;
	key.type = BTRFS_INODE_REF_KEY;
2010
	key.offset = (u64)-1;
2011

2012
	while (1) {
2013 2014 2015
		ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
		if (ret < 0)
			goto out;
2016 2017 2018 2019 2020 2021 2022 2023 2024 2025
		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;
			}
		}
2026 2027 2028 2029 2030 2031 2032 2033 2034

		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;
2035 2036
		if (ptr < name) {
			ret = -ENAMETOOLONG;
2037
			goto out;
2038
		}
2039 2040

		*(ptr + len) = '/';
2041
		read_extent_buffer(l, ptr, (unsigned long)(iref + 1), len);
2042 2043 2044 2045

		if (key.offset == BTRFS_FIRST_FREE_OBJECTID)
			break;

2046
		btrfs_release_path(path);
2047
		key.objectid = key.offset;
2048
		key.offset = (u64)-1;
2049 2050
		dirid = key.objectid;
	}
2051
	memmove(name, ptr, total_len);
2052
	name[total_len] = '\0';
2053 2054 2055
	ret = 0;
out:
	btrfs_free_path(path);
2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068
	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
2069 2070 2071
	args = memdup_user(argp, sizeof(*args));
	if (IS_ERR(args))
		return PTR_ERR(args);
2072

A
Al Viro 已提交
2073
	inode = file_inode(file);
2074

2075 2076 2077
	if (args->treeid == 0)
		args->treeid = BTRFS_I(inode)->root->root_key.objectid;

2078 2079 2080 2081 2082 2083 2084 2085
	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);
2086 2087 2088
	return ret;
}

2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099
static noinline int btrfs_ioctl_snap_destroy(struct file *file,
					     void __user *arg)
{
	struct dentry *parent = fdentry(file);
	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;
2100 2101
	struct btrfs_block_rsv block_rsv;
	u64 qgroup_reserved;
2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117
	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;
	}

2118
	err = mnt_want_write_file(file);
2119 2120 2121
	if (err)
		goto out;

2122 2123 2124
	err = mutex_lock_killable_nested(&dir->i_mutex, I_MUTEX_PARENT);
	if (err == -EINTR)
		goto out;
2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136
	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;
2137
	dest = BTRFS_I(inode)->root;
2138
	if (!capable(CAP_SYS_ADMIN)) {
2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171
		/*
		 * 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;
	}

2172 2173 2174 2175 2176
	/* 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
2177
	if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID) {
2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192
		err = -EINVAL;
		goto out_dput;
	}

	mutex_lock(&inode->i_mutex);
	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;

2193 2194 2195 2196 2197 2198
	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,
2199
					       5, &qgroup_reserved, true);
2200 2201 2202
	if (err)
		goto out_up_write;

2203 2204 2205
	trans = btrfs_start_transaction(root, 0);
	if (IS_ERR(trans)) {
		err = PTR_ERR(trans);
2206
		goto out_release;
2207
	}
2208 2209
	trans->block_rsv = &block_rsv;
	trans->bytes_reserved = block_rsv.size;
2210

2211 2212 2213 2214
	ret = btrfs_unlink_subvol(trans, root, dir,
				dest->root_key.objectid,
				dentry->d_name.name,
				dentry->d_name.len);
2215 2216 2217 2218 2219
	if (ret) {
		err = ret;
		btrfs_abort_transaction(trans, root, ret);
		goto out_end_trans;
	}
2220 2221 2222 2223 2224 2225 2226 2227

	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);

2228 2229 2230 2231
	if (!xchg(&dest->orphan_item_inserted, 1)) {
		ret = btrfs_insert_orphan_item(trans,
					root->fs_info->tree_root,
					dest->root_key.objectid);
2232 2233 2234 2235 2236
		if (ret) {
			btrfs_abort_transaction(trans, root, ret);
			err = ret;
			goto out_end_trans;
		}
2237
	}
2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258

	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;
		}
	}

2259
out_end_trans:
2260 2261
	trans->block_rsv = NULL;
	trans->bytes_reserved = 0;
S
Sage Weil 已提交
2262
	ret = btrfs_end_transaction(trans, root);
2263 2264
	if (ret && !err)
		err = ret;
2265
	inode->i_flags |= S_DEAD;
2266 2267
out_release:
	btrfs_subvolume_release_metadata(root, &block_rsv, qgroup_reserved);
2268 2269 2270 2271 2272
out_up_write:
	up_write(&root->fs_info->subvol_sem);
out_unlock:
	mutex_unlock(&inode->i_mutex);
	if (!err) {
2273
		shrink_dcache_sb(root->fs_info->sb);
2274 2275
		btrfs_invalidate_inodes(dest);
		d_delete(dentry);
2276 2277 2278 2279 2280 2281

		/* the last ref */
		if (dest->cache_inode) {
			iput(dest->cache_inode);
			dest->cache_inode = NULL;
		}
2282 2283 2284 2285 2286
	}
out_dput:
	dput(dentry);
out_unlock_dir:
	mutex_unlock(&dir->i_mutex);
A
Al Viro 已提交
2287
	mnt_drop_write_file(file);
2288 2289 2290 2291 2292
out:
	kfree(vol_args);
	return err;
}

C
Chris Mason 已提交
2293
static int btrfs_ioctl_defrag(struct file *file, void __user *argp)
C
Christoph Hellwig 已提交
2294
{
A
Al Viro 已提交
2295
	struct inode *inode = file_inode(file);
C
Christoph Hellwig 已提交
2296
	struct btrfs_root *root = BTRFS_I(inode)->root;
C
Chris Mason 已提交
2297
	struct btrfs_ioctl_defrag_range_args *range;
Y
Yan Zheng 已提交
2298 2299
	int ret;

2300 2301 2302
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
L
Li Zefan 已提交
2303

2304 2305 2306
	if (btrfs_root_readonly(root)) {
		ret = -EROFS;
		goto out;
2307
	}
C
Christoph Hellwig 已提交
2308 2309 2310

	switch (inode->i_mode & S_IFMT) {
	case S_IFDIR:
2311 2312 2313 2314
		if (!capable(CAP_SYS_ADMIN)) {
			ret = -EPERM;
			goto out;
		}
2315
		ret = btrfs_defrag_root(root);
2316 2317
		if (ret)
			goto out;
2318
		ret = btrfs_defrag_root(root->fs_info->extent_root);
C
Christoph Hellwig 已提交
2319 2320
		break;
	case S_IFREG:
2321 2322 2323 2324
		if (!(file->f_mode & FMODE_WRITE)) {
			ret = -EINVAL;
			goto out;
		}
C
Chris Mason 已提交
2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336

		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);
2337
				goto out;
C
Chris Mason 已提交
2338 2339 2340 2341 2342 2343 2344 2345 2346 2347
			}
			/* 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
2348
		ret = btrfs_defrag_file(file_inode(file), file,
C
Chris Mason 已提交
2349 2350 2351
					range, 0, 0);
		if (ret > 0)
			ret = 0;
C
Chris Mason 已提交
2352
		kfree(range);
C
Christoph Hellwig 已提交
2353
		break;
2354 2355
	default:
		ret = -EINVAL;
C
Christoph Hellwig 已提交
2356
	}
2357
out:
2358
	mnt_drop_write_file(file);
2359
	return ret;
C
Christoph Hellwig 已提交
2360 2361
}

2362
static long btrfs_ioctl_add_dev(struct btrfs_root *root, void __user *arg)
C
Christoph Hellwig 已提交
2363 2364 2365 2366
{
	struct btrfs_ioctl_vol_args *vol_args;
	int ret;

2367 2368 2369
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

2370 2371
	if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
			1)) {
2372
		return BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
2373 2374
	}

2375
	mutex_lock(&root->fs_info->volume_mutex);
L
Li Zefan 已提交
2376
	vol_args = memdup_user(arg, sizeof(*vol_args));
2377 2378 2379 2380
	if (IS_ERR(vol_args)) {
		ret = PTR_ERR(vol_args);
		goto out;
	}
C
Christoph Hellwig 已提交
2381

2382
	vol_args->name[BTRFS_PATH_NAME_MAX] = '\0';
C
Christoph Hellwig 已提交
2383 2384 2385
	ret = btrfs_init_new_device(root, vol_args->name);

	kfree(vol_args);
2386 2387
out:
	mutex_unlock(&root->fs_info->volume_mutex);
2388
	atomic_set(&root->fs_info->mutually_exclusive_operation_running, 0);
C
Christoph Hellwig 已提交
2389 2390 2391
	return ret;
}

2392
static long btrfs_ioctl_rm_dev(struct file *file, void __user *arg)
C
Christoph Hellwig 已提交
2393
{
A
Al Viro 已提交
2394
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
C
Christoph Hellwig 已提交
2395 2396 2397
	struct btrfs_ioctl_vol_args *vol_args;
	int ret;

2398 2399 2400
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

2401 2402 2403
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
Y
Yan Zheng 已提交
2404

L
Li Zefan 已提交
2405
	vol_args = memdup_user(arg, sizeof(*vol_args));
2406 2407 2408 2409
	if (IS_ERR(vol_args)) {
		ret = PTR_ERR(vol_args);
		goto out;
	}
C
Christoph Hellwig 已提交
2410

2411
	vol_args->name[BTRFS_PATH_NAME_MAX] = '\0';
C
Christoph Hellwig 已提交
2412

2413 2414 2415 2416 2417 2418 2419 2420
	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);
2421
	mutex_unlock(&root->fs_info->volume_mutex);
2422
	atomic_set(&root->fs_info->mutually_exclusive_operation_running, 0);
2423 2424 2425

out:
	kfree(vol_args);
2426
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
2427 2428 2429
	return ret;
}

J
Jan Schmidt 已提交
2430 2431
static long btrfs_ioctl_fs_info(struct btrfs_root *root, void __user *arg)
{
2432
	struct btrfs_ioctl_fs_info_args *fi_args;
J
Jan Schmidt 已提交
2433 2434 2435
	struct btrfs_device *device;
	struct btrfs_device *next;
	struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
2436
	int ret = 0;
J
Jan Schmidt 已提交
2437 2438 2439 2440

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

2441 2442 2443 2444
	fi_args = kzalloc(sizeof(*fi_args), GFP_KERNEL);
	if (!fi_args)
		return -ENOMEM;

2445
	mutex_lock(&fs_devices->device_list_mutex);
2446 2447
	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
2448 2449

	list_for_each_entry_safe(device, next, &fs_devices->devices, dev_list) {
2450 2451
		if (device->devid > fi_args->max_id)
			fi_args->max_id = device->devid;
J
Jan Schmidt 已提交
2452 2453 2454
	}
	mutex_unlock(&fs_devices->device_list_mutex);

2455 2456
	if (copy_to_user(arg, fi_args, sizeof(*fi_args)))
		ret = -EFAULT;
J
Jan Schmidt 已提交
2457

2458 2459
	kfree(fi_args);
	return ret;
J
Jan Schmidt 已提交
2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476
}

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;

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

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

2477
	if (!btrfs_is_empty_uuid(di_args->uuid))
J
Jan Schmidt 已提交
2478 2479 2480
		s_uuid = di_args->uuid;

	mutex_lock(&fs_devices->device_list_mutex);
2481
	dev = btrfs_find_device(root->fs_info, di_args->devid, s_uuid, NULL);
J
Jan Schmidt 已提交
2482 2483 2484 2485 2486 2487 2488 2489 2490 2491

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

	di_args->devid = dev->devid;
	di_args->bytes_used = dev->bytes_used;
	di_args->total_bytes = dev->total_bytes;
	memcpy(di_args->uuid, dev->uuid, sizeof(di_args->uuid));
2492
	if (dev->name) {
2493 2494 2495 2496 2497 2498
		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();
2499 2500
		di_args->path[sizeof(di_args->path) - 1] = 0;
	} else {
2501
		di_args->path[0] = '\0';
2502
	}
J
Jan Schmidt 已提交
2503 2504

out:
2505
	mutex_unlock(&fs_devices->device_list_mutex);
J
Jan Schmidt 已提交
2506 2507 2508 2509 2510 2511 2512
	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
2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540
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;
}

2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560
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);
		if (!ordered &&
		    !test_range_bit(&BTRFS_I(inode)->io_tree, off,
				    off + len - 1, EXTENT_DELALLOC, 0, NULL))
			break;
		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
2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689
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,
					 void __user *argp)
{
2690 2691 2692
	struct btrfs_ioctl_same_args tmp;
	struct btrfs_ioctl_same_args *same;
	struct btrfs_ioctl_same_extent_info *info;
M
Mark Fasheh 已提交
2693 2694 2695 2696 2697 2698 2699
	struct inode *src = file->f_dentry->d_inode;
	struct file *dst_file = NULL;
	struct inode *dst;
	u64 off;
	u64 len;
	int i;
	int ret;
2700
	unsigned long size;
M
Mark Fasheh 已提交
2701 2702 2703 2704 2705 2706 2707 2708 2709 2710
	u64 bs = BTRFS_I(src)->root->fs_info->sb->s_blocksize;
	bool is_admin = capable(CAP_SYS_ADMIN);

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

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

2711
	if (copy_from_user(&tmp,
M
Mark Fasheh 已提交
2712
			   (struct btrfs_ioctl_same_args __user *)argp,
2713
			   sizeof(tmp))) {
M
Mark Fasheh 已提交
2714 2715 2716 2717
		ret = -EFAULT;
		goto out;
	}

2718 2719 2720
	size = sizeof(tmp) +
		tmp.dest_count * sizeof(struct btrfs_ioctl_same_extent_info);

2721
	same = memdup_user((struct btrfs_ioctl_same_args __user *)argp, size);
2722

2723 2724
	if (IS_ERR(same)) {
		ret = PTR_ERR(same);
2725 2726 2727 2728 2729
		goto out;
	}

	off = same->logical_offset;
	len = same->length;
M
Mark Fasheh 已提交
2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756

	/*
	 * 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;

2757 2758 2759 2760 2761
	/* pre-format output fields to sane values */
	for (i = 0; i < same->dest_count; i++) {
		same->info[i].bytes_deduped = 0ULL;
		same->info[i].status = 0;
	}
M
Mark Fasheh 已提交
2762

2763 2764 2765
	ret = 0;
	for (i = 0; i < same->dest_count; i++) {
		info = &same->info[i];
M
Mark Fasheh 已提交
2766

2767
		dst_file = fget(info->fd);
M
Mark Fasheh 已提交
2768
		if (!dst_file) {
2769
			info->status = -EBADF;
M
Mark Fasheh 已提交
2770 2771 2772 2773
			goto next;
		}

		if (!(is_admin || (dst_file->f_mode & FMODE_WRITE))) {
2774
			info->status = -EINVAL;
M
Mark Fasheh 已提交
2775 2776 2777
			goto next;
		}

2778
		info->status = -EXDEV;
M
Mark Fasheh 已提交
2779 2780 2781 2782 2783 2784 2785 2786
		if (file->f_path.mnt != dst_file->f_path.mnt)
			goto next;

		dst = dst_file->f_dentry->d_inode;
		if (src->i_sb != dst->i_sb)
			goto next;

		if (S_ISDIR(dst->i_mode)) {
2787
			info->status = -EISDIR;
M
Mark Fasheh 已提交
2788 2789 2790 2791
			goto next;
		}

		if (!S_ISREG(dst->i_mode)) {
2792
			info->status = -EACCES;
M
Mark Fasheh 已提交
2793 2794 2795
			goto next;
		}

2796 2797 2798 2799
		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
2800 2801 2802 2803 2804 2805

next:
		if (dst_file)
			fput(dst_file);
	}

2806 2807 2808 2809
	ret = copy_to_user(argp, same, size);
	if (ret)
		ret = -EFAULT;

M
Mark Fasheh 已提交
2810 2811 2812 2813 2814
out:
	mnt_drop_write_file(file);
	return ret;
}

2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827
/**
 * 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,
		       u64 off, u64 olen, u64 olen_aligned, u64 destoff)
C
Christoph Hellwig 已提交
2828 2829
{
	struct btrfs_root *root = BTRFS_I(inode)->root;
2830
	struct btrfs_path *path = NULL;
C
Christoph Hellwig 已提交
2831
	struct extent_buffer *leaf;
2832 2833
	struct btrfs_trans_handle *trans;
	char *buf = NULL;
Y
Yan Zheng 已提交
2834
	struct btrfs_key key;
C
Christoph Hellwig 已提交
2835 2836
	u32 nritems;
	int slot;
Y
Yan Zheng 已提交
2837
	int ret;
2838
	u64 len = olen_aligned;
Y
Yan Zheng 已提交
2839 2840 2841 2842

	ret = -ENOMEM;
	buf = vmalloc(btrfs_level_size(root, 0));
	if (!buf)
2843
		return ret;
Y
Yan Zheng 已提交
2844 2845 2846 2847

	path = btrfs_alloc_path();
	if (!path) {
		vfree(buf);
2848
		return ret;
2849 2850
	}

2851
	path->reada = 2;
2852
	/* clone data */
L
Li Zefan 已提交
2853
	key.objectid = btrfs_ino(src);
Y
Yan Zheng 已提交
2854 2855
	key.type = BTRFS_EXTENT_DATA_KEY;
	key.offset = 0;
C
Christoph Hellwig 已提交
2856 2857 2858 2859 2860 2861

	while (1) {
		/*
		 * note the key will change type as we walk through the
		 * tree.
		 */
2862 2863
		ret = btrfs_search_slot(NULL, BTRFS_I(src)->root, &key, path,
				0, 0);
C
Christoph Hellwig 已提交
2864 2865 2866
		if (ret < 0)
			goto out;

Y
Yan Zheng 已提交
2867 2868
		nritems = btrfs_header_nritems(path->nodes[0]);
		if (path->slots[0] >= nritems) {
2869
			ret = btrfs_next_leaf(BTRFS_I(src)->root, path);
C
Christoph Hellwig 已提交
2870 2871 2872 2873
			if (ret < 0)
				goto out;
			if (ret > 0)
				break;
Y
Yan Zheng 已提交
2874
			nritems = btrfs_header_nritems(path->nodes[0]);
C
Christoph Hellwig 已提交
2875 2876 2877 2878
		}
		leaf = path->nodes[0];
		slot = path->slots[0];

Y
Yan Zheng 已提交
2879
		btrfs_item_key_to_cpu(leaf, &key, slot);
2880
		if (btrfs_key_type(&key) > BTRFS_EXTENT_DATA_KEY ||
L
Li Zefan 已提交
2881
		    key.objectid != btrfs_ino(src))
C
Christoph Hellwig 已提交
2882 2883
			break;

2884 2885 2886
		if (btrfs_key_type(&key) == BTRFS_EXTENT_DATA_KEY) {
			struct btrfs_file_extent_item *extent;
			int type;
Z
Zheng Yan 已提交
2887 2888
			u32 size;
			struct btrfs_key new_key;
2889 2890 2891
			u64 disko = 0, diskl = 0;
			u64 datao = 0, datal = 0;
			u8 comp;
2892
			u64 endoff;
Z
Zheng Yan 已提交
2893 2894 2895 2896 2897

			size = btrfs_item_size_nr(leaf, slot);
			read_extent_buffer(leaf, buf,
					   btrfs_item_ptr_offset(leaf, slot),
					   size);
2898 2899 2900 2901 2902

			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);
2903 2904
			if (type == BTRFS_FILE_EXTENT_REG ||
			    type == BTRFS_FILE_EXTENT_PREALLOC) {
C
Chris Mason 已提交
2905 2906 2907 2908
				disko = btrfs_file_extent_disk_bytenr(leaf,
								      extent);
				diskl = btrfs_file_extent_disk_num_bytes(leaf,
								 extent);
2909
				datao = btrfs_file_extent_offset(leaf, extent);
C
Chris Mason 已提交
2910 2911
				datal = btrfs_file_extent_num_bytes(leaf,
								    extent);
2912 2913 2914 2915 2916
			} else if (type == BTRFS_FILE_EXTENT_INLINE) {
				/* take upper bound, may be compressed */
				datal = btrfs_file_extent_ram_bytes(leaf,
								    extent);
			}
2917
			btrfs_release_path(path);
Z
Zheng Yan 已提交
2918

2919
			if (key.offset + datal <= off ||
L
Liu Bo 已提交
2920
			    key.offset >= off + len - 1)
2921 2922
				goto next;

Z
Zheng Yan 已提交
2923
			memcpy(&new_key, &key, sizeof(new_key));
L
Li Zefan 已提交
2924
			new_key.objectid = btrfs_ino(inode);
2925 2926 2927 2928
			if (off <= key.offset)
				new_key.offset = key.offset + destoff - off;
			else
				new_key.offset = destoff;
Z
Zheng Yan 已提交
2929

2930 2931 2932 2933 2934 2935
			/*
			 * 1 - adjusting old extent (we may have to split it)
			 * 1 - add new extent
			 * 1 - inode update
			 */
			trans = btrfs_start_transaction(root, 3);
2936 2937 2938 2939 2940
			if (IS_ERR(trans)) {
				ret = PTR_ERR(trans);
				goto out;
			}

2941 2942
			if (type == BTRFS_FILE_EXTENT_REG ||
			    type == BTRFS_FILE_EXTENT_PREALLOC) {
2943 2944 2945 2946 2947 2948 2949 2950 2951 2952
				/*
				 *    a  | --- range to clone ---|  b
				 * | ------------- extent ------------- |
				 */

				/* substract range b */
				if (key.offset + datal > off + len)
					datal = off + len - key.offset;

				/* substract range a */
2953 2954 2955 2956 2957
				if (off > key.offset) {
					datao += off - key.offset;
					datal -= off - key.offset;
				}

J
Josef Bacik 已提交
2958
				ret = btrfs_drop_extents(trans, root, inode,
2959 2960
							 new_key.offset,
							 new_key.offset + datal,
2961
							 1);
2962 2963 2964 2965 2966 2967
				if (ret) {
					btrfs_abort_transaction(trans, root,
								ret);
					btrfs_end_transaction(trans, root);
					goto out;
				}
2968

2969 2970
				ret = btrfs_insert_empty_item(trans, root, path,
							      &new_key, size);
2971 2972 2973 2974 2975 2976
				if (ret) {
					btrfs_abort_transaction(trans, root,
								ret);
					btrfs_end_transaction(trans, root);
					goto out;
				}
2977 2978 2979 2980

				leaf = path->nodes[0];
				slot = path->slots[0];
				write_extent_buffer(leaf, buf,
Z
Zheng Yan 已提交
2981 2982
					    btrfs_item_ptr_offset(leaf, slot),
					    size);
Y
Yan Zheng 已提交
2983

2984
				extent = btrfs_item_ptr(leaf, slot,
C
Christoph Hellwig 已提交
2985
						struct btrfs_file_extent_item);
2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996

				/* 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);
				if (disko) {
					inode_add_bytes(inode, datal);
Y
Yan Zheng 已提交
2997
					ret = btrfs_inc_extent_ref(trans, root,
2998 2999
							disko, diskl, 0,
							root->root_key.objectid,
L
Li Zefan 已提交
3000
							btrfs_ino(inode),
A
Arne Jansen 已提交
3001 3002
							new_key.offset - datao,
							0);
3003 3004 3005 3006 3007 3008 3009 3010 3011
					if (ret) {
						btrfs_abort_transaction(trans,
									root,
									ret);
						btrfs_end_transaction(trans,
								      root);
						goto out;

					}
C
Christoph Hellwig 已提交
3012
				}
3013 3014 3015 3016 3017 3018 3019
			} else if (type == BTRFS_FILE_EXTENT_INLINE) {
				u64 skip = 0;
				u64 trim = 0;
				if (off > key.offset) {
					skip = off - key.offset;
					new_key.offset += skip;
				}
C
Chris Mason 已提交
3020

L
Liu Bo 已提交
3021 3022
				if (key.offset + datal > off + len)
					trim = key.offset + datal - (off + len);
C
Chris Mason 已提交
3023

3024 3025
				if (comp && (skip || trim)) {
					ret = -EINVAL;
3026
					btrfs_end_transaction(trans, root);
3027 3028 3029 3030
					goto out;
				}
				size -= skip + trim;
				datal -= skip + trim;
3031

J
Josef Bacik 已提交
3032
				ret = btrfs_drop_extents(trans, root, inode,
3033 3034
							 new_key.offset,
							 new_key.offset + datal,
3035
							 1);
3036 3037 3038 3039 3040 3041
				if (ret) {
					btrfs_abort_transaction(trans, root,
								ret);
					btrfs_end_transaction(trans, root);
					goto out;
				}
3042

3043 3044
				ret = btrfs_insert_empty_item(trans, root, path,
							      &new_key, size);
3045 3046 3047 3048 3049 3050
				if (ret) {
					btrfs_abort_transaction(trans, root,
								ret);
					btrfs_end_transaction(trans, root);
					goto out;
				}
3051 3052

				if (skip) {
C
Chris Mason 已提交
3053 3054
					u32 start =
					  btrfs_file_extent_calc_inline_size(0);
3055 3056 3057 3058 3059 3060 3061 3062 3063 3064
					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
3065
			}
3066 3067

			btrfs_mark_buffer_dirty(leaf);
3068
			btrfs_release_path(path);
3069

3070
			inode_inc_iversion(inode);
3071
			inode->i_mtime = inode->i_ctime = CURRENT_TIME;
3072 3073 3074 3075 3076 3077 3078

			/*
			 * we round up to the block size at eof when
			 * determining which extents to clone above,
			 * but shouldn't round up the file size
			 */
			endoff = new_key.offset + datal;
3079 3080
			if (endoff > destoff+olen)
				endoff = destoff+olen;
3081 3082 3083
			if (endoff > inode->i_size)
				btrfs_i_size_write(inode, endoff);

3084
			ret = btrfs_update_inode(trans, root, inode);
3085 3086 3087 3088 3089 3090
			if (ret) {
				btrfs_abort_transaction(trans, root, ret);
				btrfs_end_transaction(trans, root);
				goto out;
			}
			ret = btrfs_end_transaction(trans, root);
3091
		}
C
Chris Mason 已提交
3092
next:
3093
		btrfs_release_path(path);
C
Christoph Hellwig 已提交
3094 3095 3096
		key.offset++;
	}
	ret = 0;
3097

C
Christoph Hellwig 已提交
3098
out:
3099
	btrfs_release_path(path);
3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112 3113 3114 3115 3116 3117 3118 3119 3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217
	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)
{
	struct inode *inode = fdentry(file)->d_inode;
	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.
	 * - allow ranges within the same file to be cloned (provided
	 *   they don't overlap)?
	 */

	/* 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;
	}

	/* truncate page cache pages from target inode range */
	truncate_inode_pages_range(&inode->i_data, destoff,
				   PAGE_CACHE_ALIGN(destoff + len) - 1);

	lock_extent_range(src, off, len);

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

L
Liu Bo 已提交
3218
	unlock_extent(&BTRFS_I(src)->io_tree, off, off + len - 1);
C
Christoph Hellwig 已提交
3219 3220
out_unlock:
	mutex_unlock(&src->i_mutex);
3221 3222
	if (!same_inode)
		mutex_unlock(&inode->i_mutex);
C
Christoph Hellwig 已提交
3223
out_fput:
3224
	fdput(src_file);
3225
out_drop_write:
A
Al Viro 已提交
3226
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
3227 3228 3229
	return ret;
}

3230
static long btrfs_ioctl_clone_range(struct file *file, void __user *argp)
3231 3232 3233
{
	struct btrfs_ioctl_clone_range_args args;

3234
	if (copy_from_user(&args, argp, sizeof(args)))
3235 3236 3237 3238 3239
		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
3240 3241 3242 3243 3244 3245
/*
 * 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.
 */
3246
static long btrfs_ioctl_trans_start(struct file *file)
C
Christoph Hellwig 已提交
3247
{
A
Al Viro 已提交
3248
	struct inode *inode = file_inode(file);
C
Christoph Hellwig 已提交
3249 3250
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;
3251
	int ret;
C
Christoph Hellwig 已提交
3252

3253
	ret = -EPERM;
3254
	if (!capable(CAP_SYS_ADMIN))
3255
		goto out;
3256

3257 3258
	ret = -EINPROGRESS;
	if (file->private_data)
C
Christoph Hellwig 已提交
3259
		goto out;
3260

L
Li Zefan 已提交
3261 3262 3263 3264
	ret = -EROFS;
	if (btrfs_root_readonly(root))
		goto out;

3265
	ret = mnt_want_write_file(file);
Y
Yan Zheng 已提交
3266 3267 3268
	if (ret)
		goto out;

J
Josef Bacik 已提交
3269
	atomic_inc(&root->fs_info->open_ioctl_trans);
3270

3271
	ret = -ENOMEM;
3272
	trans = btrfs_start_ioctl_transaction(root);
3273
	if (IS_ERR(trans))
3274 3275 3276 3277 3278 3279
		goto out_drop;

	file->private_data = trans;
	return 0;

out_drop:
J
Josef Bacik 已提交
3280
	atomic_dec(&root->fs_info->open_ioctl_trans);
A
Al Viro 已提交
3281
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
3282 3283 3284 3285
out:
	return ret;
}

3286 3287
static long btrfs_ioctl_default_subvol(struct file *file, void __user *argp)
{
A
Al Viro 已提交
3288
	struct inode *inode = file_inode(file);
3289 3290 3291 3292 3293 3294 3295 3296 3297
	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;
3298
	int ret;
3299 3300 3301 3302

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

3303 3304 3305 3306 3307 3308 3309 3310
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;

	if (copy_from_user(&objectid, argp, sizeof(objectid))) {
		ret = -EFAULT;
		goto out;
	}
3311 3312

	if (!objectid)
3313
		objectid = BTRFS_FS_TREE_OBJECTID;
3314 3315 3316 3317 3318 3319

	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);
3320 3321 3322 3323
	if (IS_ERR(new_root)) {
		ret = PTR_ERR(new_root);
		goto out;
	}
3324 3325

	path = btrfs_alloc_path();
3326 3327 3328 3329
	if (!path) {
		ret = -ENOMEM;
		goto out;
	}
3330 3331 3332
	path->leave_spinning = 1;

	trans = btrfs_start_transaction(root, 1);
3333
	if (IS_ERR(trans)) {
3334
		btrfs_free_path(path);
3335 3336
		ret = PTR_ERR(trans);
		goto out;
3337 3338
	}

3339
	dir_id = btrfs_super_root_dir(root->fs_info->super_copy);
3340 3341
	di = btrfs_lookup_dir_item(trans, root->fs_info->tree_root, path,
				   dir_id, "default", 7, 1);
3342
	if (IS_ERR_OR_NULL(di)) {
3343 3344 3345 3346
		btrfs_free_path(path);
		btrfs_end_transaction(trans, root);
		printk(KERN_ERR "Umm, you don't have the default dir item, "
		       "this isn't going to work\n");
3347 3348
		ret = -ENOENT;
		goto out;
3349 3350 3351 3352 3353 3354 3355
	}

	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);

3356
	btrfs_set_fs_incompat(root->fs_info, DEFAULT_SUBVOL);
3357
	btrfs_end_transaction(trans, root);
3358 3359 3360
out:
	mnt_drop_write_file(file);
	return ret;
3361 3362
}

3363 3364
void btrfs_get_block_group_info(struct list_head *groups_list,
				struct btrfs_ioctl_space_info *space)
3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378
{
	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);
	}
}

3379
static long btrfs_ioctl_space_info(struct btrfs_root *root, void __user *arg)
J
Josef Bacik 已提交
3380 3381 3382 3383
{
	struct btrfs_ioctl_space_args space_args;
	struct btrfs_ioctl_space_info space;
	struct btrfs_ioctl_space_info *dest;
3384
	struct btrfs_ioctl_space_info *dest_orig;
3385
	struct btrfs_ioctl_space_info __user *user_dest;
J
Josef Bacik 已提交
3386
	struct btrfs_space_info *info;
3387 3388 3389 3390 3391
	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;
3392
	int alloc_size;
J
Josef Bacik 已提交
3393
	int ret = 0;
3394
	u64 slot_count = 0;
3395
	int i, c;
J
Josef Bacik 已提交
3396 3397 3398 3399 3400 3401

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

3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425
	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);
	}
3426 3427 3428 3429 3430 3431

	/* space_slots == 0 means they are asking for a count */
	if (space_args.space_slots == 0) {
		space_args.total_spaces = slot_count;
		goto out;
	}
3432

3433
	slot_count = min_t(u64, space_args.space_slots, slot_count);
3434

3435
	alloc_size = sizeof(*dest) * slot_count;
3436

3437 3438 3439 3440 3441 3442
	/* 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
3443
	space_args.total_spaces = 0;
3444 3445 3446 3447
	dest = kmalloc(alloc_size, GFP_NOFS);
	if (!dest)
		return -ENOMEM;
	dest_orig = dest;
J
Josef Bacik 已提交
3448

3449
	/* now we have a buffer to copy into */
3450 3451 3452
	for (i = 0; i < num_types; i++) {
		struct btrfs_space_info *tmp;

3453 3454 3455
		if (!slot_count)
			break;

3456 3457 3458 3459 3460 3461 3462 3463 3464 3465
		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();
3466

3467 3468 3469 3470 3471
		if (!info)
			continue;
		down_read(&info->groups_sem);
		for (c = 0; c < BTRFS_NR_RAID_TYPES; c++) {
			if (!list_empty(&info->block_groups[c])) {
3472 3473
				btrfs_get_block_group_info(
					&info->block_groups[c], &space);
3474 3475 3476
				memcpy(dest, &space, sizeof(space));
				dest++;
				space_args.total_spaces++;
3477
				slot_count--;
3478
			}
3479 3480
			if (!slot_count)
				break;
3481 3482
		}
		up_read(&info->groups_sem);
J
Josef Bacik 已提交
3483 3484
	}

D
Daniel J Blueman 已提交
3485
	user_dest = (struct btrfs_ioctl_space_info __user *)
3486 3487 3488 3489 3490 3491 3492 3493
		(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
3494 3495 3496 3497 3498
		ret = -EFAULT;

	return ret;
}

C
Christoph Hellwig 已提交
3499 3500 3501 3502 3503 3504 3505 3506
/*
 * 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
3507
	struct inode *inode = file_inode(file);
C
Christoph Hellwig 已提交
3508 3509 3510 3511
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;

	trans = file->private_data;
3512 3513
	if (!trans)
		return -EINVAL;
C
Christoph Hellwig 已提交
3514
	file->private_data = NULL;
3515

3516 3517
	btrfs_end_transaction(trans, root);

J
Josef Bacik 已提交
3518
	atomic_dec(&root->fs_info->open_ioctl_trans);
3519

A
Al Viro 已提交
3520
	mnt_drop_write_file(file);
3521
	return 0;
C
Christoph Hellwig 已提交
3522 3523
}

3524 3525
static noinline long btrfs_ioctl_start_sync(struct btrfs_root *root,
					    void __user *argp)
3526 3527 3528
{
	struct btrfs_trans_handle *trans;
	u64 transid;
T
Tsutomu Itoh 已提交
3529
	int ret;
3530

M
Miao Xie 已提交
3531
	trans = btrfs_attach_transaction_barrier(root);
3532 3533 3534 3535 3536 3537 3538 3539
	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;
	}
3540
	transid = trans->transid;
T
Tsutomu Itoh 已提交
3541
	ret = btrfs_commit_transaction_async(trans, root, 0);
3542 3543
	if (ret) {
		btrfs_end_transaction(trans, root);
T
Tsutomu Itoh 已提交
3544
		return ret;
3545
	}
3546
out:
3547 3548 3549 3550 3551 3552
	if (argp)
		if (copy_to_user(argp, &transid, sizeof(transid)))
			return -EFAULT;
	return 0;
}

3553 3554
static noinline long btrfs_ioctl_wait_sync(struct btrfs_root *root,
					   void __user *argp)
3555 3556 3557 3558 3559 3560 3561 3562 3563 3564 3565 3566
{
	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
3567
static long btrfs_ioctl_scrub(struct file *file, void __user *arg)
J
Jan Schmidt 已提交
3568
{
A
Al Viro 已提交
3569
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
J
Jan Schmidt 已提交
3570
	struct btrfs_ioctl_scrub_args *sa;
M
Miao Xie 已提交
3571
	int ret;
J
Jan Schmidt 已提交
3572 3573 3574 3575 3576 3577 3578 3579

	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
3580 3581 3582 3583 3584 3585
	if (!(sa->flags & BTRFS_SCRUB_READONLY)) {
		ret = mnt_want_write_file(file);
		if (ret)
			goto out;
	}

3586
	ret = btrfs_scrub_dev(root->fs_info, sa->devid, sa->start, sa->end,
3587 3588
			      &sa->progress, sa->flags & BTRFS_SCRUB_READONLY,
			      0);
J
Jan Schmidt 已提交
3589 3590 3591 3592

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

M
Miao Xie 已提交
3593 3594 3595
	if (!(sa->flags & BTRFS_SCRUB_READONLY))
		mnt_drop_write_file(file);
out:
J
Jan Schmidt 已提交
3596 3597 3598 3599 3600 3601 3602 3603 3604
	kfree(sa);
	return ret;
}

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

3605
	return btrfs_scrub_cancel(root->fs_info);
J
Jan Schmidt 已提交
3606 3607 3608 3609 3610 3611 3612 3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 3629
}

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;
}

3630
static long btrfs_ioctl_get_dev_stats(struct btrfs_root *root,
3631
				      void __user *arg)
3632 3633 3634 3635 3636 3637 3638 3639
{
	struct btrfs_ioctl_get_dev_stats *sa;
	int ret;

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

3640 3641 3642 3643 3644 3645
	if ((sa->flags & BTRFS_DEV_STATS_RESET) && !capable(CAP_SYS_ADMIN)) {
		kfree(sa);
		return -EPERM;
	}

	ret = btrfs_get_dev_stats(root, sa);
3646 3647 3648 3649 3650 3651 3652 3653

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

	kfree(sa);
	return ret;
}

3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 3667
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:
3668 3669 3670 3671
		if (root->fs_info->sb->s_flags & MS_RDONLY) {
			ret = -EROFS;
			goto out;
		}
3672 3673 3674
		if (atomic_xchg(
			&root->fs_info->mutually_exclusive_operation_running,
			1)) {
3675
			ret = BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
3676 3677 3678 3679 3680 3681 3682 3683 3684 3685 3686 3687 3688 3689 3690 3691 3692 3693 3694 3695 3696
		} 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;
3697
out:
3698 3699 3700 3701
	kfree(p);
	return ret;
}

3702 3703 3704 3705
static long btrfs_ioctl_ino_to_path(struct btrfs_root *root, void __user *arg)
{
	int ret = 0;
	int i;
3706
	u64 rel_ptr;
3707
	int size;
3708
	struct btrfs_ioctl_ino_path_args *ipa = NULL;
3709 3710 3711
	struct inode_fs_paths *ipath = NULL;
	struct btrfs_path *path;

3712
	if (!capable(CAP_DAC_READ_SEARCH))
3713 3714 3715 3716 3717 3718 3719 3720 3721 3722 3723 3724 3725 3726 3727 3728 3729 3730 3731 3732 3733 3734 3735 3736 3737 3738 3739 3740
		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) {
3741 3742
		rel_ptr = ipath->fspath->val[i] -
			  (u64)(unsigned long)ipath->fspath->val;
3743
		ipath->fspath->val[i] = rel_ptr;
3744 3745
	}

3746 3747
	ret = copy_to_user((void *)(unsigned long)ipa->fspath,
			   (void *)(unsigned long)ipath->fspath, size);
3748 3749 3750 3751 3752 3753 3754 3755 3756 3757 3758 3759 3760 3761 3762 3763 3764 3765 3766 3767 3768 3769 3770 3771 3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786 3787 3788 3789 3790 3791 3792 3793 3794 3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805
	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;
	}

3806
	size = min_t(u32, loi->size, 64 * 1024);
3807 3808 3809 3810 3811 3812 3813
	inodes = init_data_container(size);
	if (IS_ERR(inodes)) {
		ret = PTR_ERR(inodes);
		inodes = NULL;
		goto out;
	}

L
Liu Bo 已提交
3814 3815 3816
	ret = iterate_inodes_from_logical(loi->logical, root->fs_info, path,
					  build_ino_list, inodes);
	if (ret == -EINVAL)
3817 3818 3819 3820
		ret = -ENOENT;
	if (ret < 0)
		goto out;

3821 3822
	ret = copy_to_user((void *)(unsigned long)loi->inodes,
			   (void *)(unsigned long)inodes, size);
3823 3824 3825 3826 3827
	if (ret)
		ret = -EFAULT;

out:
	btrfs_free_path(path);
3828
	vfree(inodes);
3829 3830 3831 3832 3833
	kfree(loi);

	return ret;
}

3834
void update_ioctl_balance_args(struct btrfs_fs_info *fs_info, int lock,
3835 3836 3837 3838 3839 3840
			       struct btrfs_ioctl_balance_args *bargs)
{
	struct btrfs_balance_control *bctl = fs_info->balance_ctl;

	bargs->flags = bctl->flags;

3841 3842 3843 3844
	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;
3845 3846
	if (atomic_read(&fs_info->balance_cancel_req))
		bargs->state |= BTRFS_BALANCE_STATE_CANCEL_REQ;
3847

3848 3849 3850
	memcpy(&bargs->data, &bctl->data, sizeof(bargs->data));
	memcpy(&bargs->meta, &bctl->meta, sizeof(bargs->meta));
	memcpy(&bargs->sys, &bctl->sys, sizeof(bargs->sys));
3851 3852 3853 3854 3855 3856 3857 3858

	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));
	}
3859 3860
}

3861
static long btrfs_ioctl_balance(struct file *file, void __user *arg)
3862
{
A
Al Viro 已提交
3863
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
3864 3865 3866
	struct btrfs_fs_info *fs_info = root->fs_info;
	struct btrfs_ioctl_balance_args *bargs;
	struct btrfs_balance_control *bctl;
3867
	bool need_unlock; /* for mut. excl. ops lock */
3868 3869 3870 3871 3872
	int ret;

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

3873
	ret = mnt_want_write_file(file);
3874 3875 3876
	if (ret)
		return ret;

3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890
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)
	 */
3891
	mutex_lock(&fs_info->balance_mutex);
3892 3893 3894 3895 3896 3897 3898 3899 3900 3901 3902 3903 3904 3905 3906 3907 3908 3909 3910 3911 3912 3913 3914 3915 3916 3917 3918
	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);
3919
		ret = BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
3920 3921 3922 3923 3924
		goto out;
	}

locked:
	BUG_ON(!atomic_read(&fs_info->mutually_exclusive_operation_running));
3925 3926 3927 3928 3929

	if (arg) {
		bargs = memdup_user(arg, sizeof(*bargs));
		if (IS_ERR(bargs)) {
			ret = PTR_ERR(bargs);
3930
			goto out_unlock;
3931
		}
3932 3933 3934 3935 3936 3937 3938 3939 3940 3941 3942 3943 3944 3945

		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;
		}
3946 3947 3948 3949
	} else {
		bargs = NULL;
	}

3950
	if (fs_info->balance_ctl) {
3951 3952 3953 3954
		ret = -EINPROGRESS;
		goto out_bargs;
	}

3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966 3967
	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;
3968 3969 3970
	} else {
		/* balance everything - no filters */
		bctl->flags |= BTRFS_BALANCE_TYPE_MASK;
3971 3972
	}

3973
do_balance:
3974
	/*
3975 3976 3977 3978 3979
	 * 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.
3980
	 */
3981 3982 3983 3984
	need_unlock = false;

	ret = btrfs_balance(bctl, bargs);

3985 3986 3987 3988 3989 3990 3991
	if (arg) {
		if (copy_to_user(arg, bargs, sizeof(*bargs)))
			ret = -EFAULT;
	}

out_bargs:
	kfree(bargs);
3992
out_unlock:
3993 3994
	mutex_unlock(&fs_info->balance_mutex);
	mutex_unlock(&fs_info->volume_mutex);
3995 3996 3997
	if (need_unlock)
		atomic_set(&fs_info->mutually_exclusive_operation_running, 0);
out:
3998
	mnt_drop_write_file(file);
3999 4000 4001
	return ret;
}

4002 4003 4004 4005 4006 4007 4008 4009
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);
4010 4011
	case BTRFS_BALANCE_CTL_CANCEL:
		return btrfs_cancel_balance(root->fs_info);
4012 4013 4014 4015 4016
	}

	return -EINVAL;
}

4017 4018 4019 4020 4021 4022 4023 4024 4025 4026 4027 4028 4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044 4045 4046 4047 4048 4049
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;
}

4050
static long btrfs_ioctl_quota_ctl(struct file *file, void __user *arg)
A
Arne Jansen 已提交
4051
{
A
Al Viro 已提交
4052
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
A
Arne Jansen 已提交
4053 4054 4055 4056 4057 4058 4059 4060
	struct btrfs_ioctl_quota_ctl_args *sa;
	struct btrfs_trans_handle *trans = NULL;
	int ret;
	int err;

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

4061 4062 4063
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
4064 4065

	sa = memdup_user(arg, sizeof(*sa));
4066 4067 4068 4069
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
4070

4071
	down_write(&root->fs_info->subvol_sem);
J
Jan Schmidt 已提交
4072 4073 4074 4075
	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
4076 4077 4078 4079 4080 4081 4082 4083 4084 4085 4086 4087 4088 4089
	}

	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
4090 4091 4092
	err = btrfs_commit_transaction(trans, root->fs_info->tree_root);
	if (err && !ret)
		ret = err;
A
Arne Jansen 已提交
4093 4094
out:
	kfree(sa);
4095
	up_write(&root->fs_info->subvol_sem);
4096 4097
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
4098 4099 4100
	return ret;
}

4101
static long btrfs_ioctl_qgroup_assign(struct file *file, void __user *arg)
A
Arne Jansen 已提交
4102
{
A
Al Viro 已提交
4103
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
A
Arne Jansen 已提交
4104 4105 4106 4107 4108 4109 4110 4111
	struct btrfs_ioctl_qgroup_assign_args *sa;
	struct btrfs_trans_handle *trans;
	int ret;
	int err;

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

4112 4113 4114
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
4115 4116

	sa = memdup_user(arg, sizeof(*sa));
4117 4118 4119 4120
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
4121 4122 4123 4124 4125 4126 4127 4128 4129 4130 4131 4132 4133 4134 4135 4136 4137 4138 4139 4140 4141 4142

	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);
4143 4144
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
4145 4146 4147
	return ret;
}

4148
static long btrfs_ioctl_qgroup_create(struct file *file, void __user *arg)
A
Arne Jansen 已提交
4149
{
A
Al Viro 已提交
4150
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
A
Arne Jansen 已提交
4151 4152 4153 4154 4155 4156 4157 4158
	struct btrfs_ioctl_qgroup_create_args *sa;
	struct btrfs_trans_handle *trans;
	int ret;
	int err;

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

4159 4160 4161
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
4162 4163

	sa = memdup_user(arg, sizeof(*sa));
4164 4165 4166 4167
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
4168

M
Miao Xie 已提交
4169 4170 4171 4172 4173
	if (!sa->qgroupid) {
		ret = -EINVAL;
		goto out;
	}

A
Arne Jansen 已提交
4174 4175 4176 4177 4178 4179 4180 4181 4182 4183 4184 4185 4186 4187 4188 4189 4190 4191 4192 4193
	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);
4194 4195
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
4196 4197 4198
	return ret;
}

4199
static long btrfs_ioctl_qgroup_limit(struct file *file, void __user *arg)
A
Arne Jansen 已提交
4200
{
A
Al Viro 已提交
4201
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
A
Arne Jansen 已提交
4202 4203 4204 4205 4206 4207 4208 4209 4210
	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;

4211 4212 4213
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
4214 4215

	sa = memdup_user(arg, sizeof(*sa));
4216 4217 4218 4219
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232 4233 4234 4235 4236 4237 4238 4239 4240 4241

	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);
4242 4243
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
4244 4245 4246
	return ret;
}

J
Jan Schmidt 已提交
4247 4248
static long btrfs_ioctl_quota_rescan(struct file *file, void __user *arg)
{
A
Al Viro 已提交
4249
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
J
Jan Schmidt 已提交
4250 4251 4252 4253 4254 4255 4256 4257 4258 4259 4260 4261 4262 4263 4264 4265 4266 4267 4268 4269 4270 4271 4272 4273 4274 4275 4276 4277 4278 4279 4280 4281
	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
4282
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
J
Jan Schmidt 已提交
4283 4284 4285 4286 4287 4288 4289 4290 4291 4292 4293 4294 4295 4296 4297 4298 4299 4300 4301 4302 4303 4304
	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;
}

4305 4306 4307 4308 4309 4310 4311 4312 4313 4314
static long btrfs_ioctl_quota_rescan_wait(struct file *file, void __user *arg)
{
	struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;

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

	return btrfs_qgroup_wait_for_completion(root->fs_info);
}

4315 4316 4317 4318
static long btrfs_ioctl_set_received_subvol(struct file *file,
					    void __user *arg)
{
	struct btrfs_ioctl_received_subvol_args *sa = NULL;
A
Al Viro 已提交
4319
	struct inode *inode = file_inode(file);
4320 4321 4322 4323 4324
	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;
4325
	int received_uuid_changed;
4326 4327 4328 4329 4330 4331 4332 4333 4334 4335 4336 4337 4338 4339 4340 4341 4342 4343 4344 4345 4346 4347 4348 4349 4350 4351 4352 4353 4354

	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;
	}

	if (!inode_owner_or_capable(inode)) {
		ret = -EACCES;
		goto out;
	}

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

4355 4356 4357 4358 4359
	/*
	 * 1 - root item
	 * 2 - uuid items (received uuid + subvol uuid)
	 */
	trans = btrfs_start_transaction(root, 3);
4360 4361 4362 4363 4364 4365 4366 4367 4368 4369
	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;

4370 4371 4372 4373 4374 4375 4376 4377
	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);
4378 4379 4380
	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);
4381 4382 4383 4384
	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);
4385 4386 4387 4388 4389 4390

	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;
4391 4392 4393 4394 4395 4396 4397 4398
	}
	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);
4399
			goto out;
4400 4401 4402 4403 4404 4405
		}
	}
	ret = btrfs_commit_transaction(trans, root);
	if (ret < 0) {
		btrfs_abort_transaction(trans, root, ret);
		goto out;
4406 4407 4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418
	}

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

out:
	kfree(sa);
	up_write(&root->fs_info->subvol_sem);
	mnt_drop_write_file(file);
	return ret;
}

4419 4420
static int btrfs_ioctl_get_fslabel(struct file *file, void __user *arg)
{
A
Al Viro 已提交
4421
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
4422
	size_t len;
4423
	int ret;
4424 4425 4426 4427 4428 4429 4430
	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);
4431 4432 4433 4434 4435 4436 4437 4438 4439 4440 4441

	if (len == BTRFS_LABEL_SIZE) {
		pr_warn("btrfs: label is too long, return the first %zu bytes\n",
			--len);
	}

	ret = copy_to_user(arg, label, len);

	return ret ? -EFAULT : 0;
}

4442 4443
static int btrfs_ioctl_set_fslabel(struct file *file, void __user *arg)
{
A
Al Viro 已提交
4444
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
4445 4446 4447 4448 4449 4450 4451 4452 4453 4454 4455 4456 4457 4458 4459 4460 4461 4462 4463 4464 4465 4466 4467 4468 4469 4470 4471
	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) {
		pr_err("btrfs: unable to set label with more than %d bytes\n",
		       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;
	}

4472
	spin_lock(&root->fs_info->super_lock);
4473
	strcpy(super_block->label, label);
4474
	spin_unlock(&root->fs_info->super_lock);
4475 4476 4477 4478 4479 4480 4481
	ret = btrfs_end_transaction(trans, root);

out_unlock:
	mnt_drop_write_file(file);
	return ret;
}

C
Christoph Hellwig 已提交
4482 4483 4484
long btrfs_ioctl(struct file *file, unsigned int
		cmd, unsigned long arg)
{
A
Al Viro 已提交
4485
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
4486
	void __user *argp = (void __user *)arg;
C
Christoph Hellwig 已提交
4487 4488

	switch (cmd) {
4489 4490 4491 4492 4493 4494
	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);
4495 4496
	case FITRIM:
		return btrfs_ioctl_fitrim(file, argp);
C
Christoph Hellwig 已提交
4497
	case BTRFS_IOC_SNAP_CREATE:
4498
		return btrfs_ioctl_snap_create(file, argp, 0);
4499
	case BTRFS_IOC_SNAP_CREATE_V2:
4500
		return btrfs_ioctl_snap_create_v2(file, argp, 0);
4501
	case BTRFS_IOC_SUBVOL_CREATE:
4502
		return btrfs_ioctl_snap_create(file, argp, 1);
A
Arne Jansen 已提交
4503 4504
	case BTRFS_IOC_SUBVOL_CREATE_V2:
		return btrfs_ioctl_snap_create_v2(file, argp, 1);
4505 4506
	case BTRFS_IOC_SNAP_DESTROY:
		return btrfs_ioctl_snap_destroy(file, argp);
4507 4508 4509 4510
	case BTRFS_IOC_SUBVOL_GETFLAGS:
		return btrfs_ioctl_subvol_getflags(file, argp);
	case BTRFS_IOC_SUBVOL_SETFLAGS:
		return btrfs_ioctl_subvol_setflags(file, argp);
4511 4512
	case BTRFS_IOC_DEFAULT_SUBVOL:
		return btrfs_ioctl_default_subvol(file, argp);
C
Christoph Hellwig 已提交
4513
	case BTRFS_IOC_DEFRAG:
C
Chris Mason 已提交
4514 4515 4516
		return btrfs_ioctl_defrag(file, NULL);
	case BTRFS_IOC_DEFRAG_RANGE:
		return btrfs_ioctl_defrag(file, argp);
C
Christoph Hellwig 已提交
4517
	case BTRFS_IOC_RESIZE:
4518
		return btrfs_ioctl_resize(file, argp);
C
Christoph Hellwig 已提交
4519
	case BTRFS_IOC_ADD_DEV:
4520
		return btrfs_ioctl_add_dev(root, argp);
C
Christoph Hellwig 已提交
4521
	case BTRFS_IOC_RM_DEV:
4522
		return btrfs_ioctl_rm_dev(file, argp);
J
Jan Schmidt 已提交
4523 4524 4525 4526
	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
4527
	case BTRFS_IOC_BALANCE:
4528
		return btrfs_ioctl_balance(file, NULL);
C
Christoph Hellwig 已提交
4529
	case BTRFS_IOC_CLONE:
4530 4531
		return btrfs_ioctl_clone(file, arg, 0, 0, 0);
	case BTRFS_IOC_CLONE_RANGE:
4532
		return btrfs_ioctl_clone_range(file, argp);
C
Christoph Hellwig 已提交
4533 4534 4535 4536
	case BTRFS_IOC_TRANS_START:
		return btrfs_ioctl_trans_start(file);
	case BTRFS_IOC_TRANS_END:
		return btrfs_ioctl_trans_end(file);
4537 4538 4539 4540
	case BTRFS_IOC_TREE_SEARCH:
		return btrfs_ioctl_tree_search(file, argp);
	case BTRFS_IOC_INO_LOOKUP:
		return btrfs_ioctl_ino_lookup(file, argp);
4541 4542 4543 4544
	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
4545 4546
	case BTRFS_IOC_SPACE_INFO:
		return btrfs_ioctl_space_info(root, argp);
4547 4548 4549
	case BTRFS_IOC_SYNC: {
		int ret;

4550
		ret = btrfs_start_delalloc_roots(root->fs_info, 0);
4551 4552 4553 4554 4555
		if (ret)
			return ret;
		ret = btrfs_sync_fs(file->f_dentry->d_sb, 1);
		return ret;
	}
4556
	case BTRFS_IOC_START_SYNC:
4557
		return btrfs_ioctl_start_sync(root, argp);
4558
	case BTRFS_IOC_WAIT_SYNC:
4559
		return btrfs_ioctl_wait_sync(root, argp);
J
Jan Schmidt 已提交
4560
	case BTRFS_IOC_SCRUB:
M
Miao Xie 已提交
4561
		return btrfs_ioctl_scrub(file, argp);
J
Jan Schmidt 已提交
4562 4563 4564 4565
	case BTRFS_IOC_SCRUB_CANCEL:
		return btrfs_ioctl_scrub_cancel(root, argp);
	case BTRFS_IOC_SCRUB_PROGRESS:
		return btrfs_ioctl_scrub_progress(root, argp);
4566
	case BTRFS_IOC_BALANCE_V2:
4567
		return btrfs_ioctl_balance(file, argp);
4568 4569
	case BTRFS_IOC_BALANCE_CTL:
		return btrfs_ioctl_balance_ctl(root, arg);
4570 4571
	case BTRFS_IOC_BALANCE_PROGRESS:
		return btrfs_ioctl_balance_progress(root, argp);
4572 4573
	case BTRFS_IOC_SET_RECEIVED_SUBVOL:
		return btrfs_ioctl_set_received_subvol(file, argp);
4574 4575
	case BTRFS_IOC_SEND:
		return btrfs_ioctl_send(file, argp);
4576
	case BTRFS_IOC_GET_DEV_STATS:
4577
		return btrfs_ioctl_get_dev_stats(root, argp);
A
Arne Jansen 已提交
4578
	case BTRFS_IOC_QUOTA_CTL:
4579
		return btrfs_ioctl_quota_ctl(file, argp);
A
Arne Jansen 已提交
4580
	case BTRFS_IOC_QGROUP_ASSIGN:
4581
		return btrfs_ioctl_qgroup_assign(file, argp);
A
Arne Jansen 已提交
4582
	case BTRFS_IOC_QGROUP_CREATE:
4583
		return btrfs_ioctl_qgroup_create(file, argp);
A
Arne Jansen 已提交
4584
	case BTRFS_IOC_QGROUP_LIMIT:
4585
		return btrfs_ioctl_qgroup_limit(file, argp);
J
Jan Schmidt 已提交
4586 4587 4588 4589
	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);
4590 4591
	case BTRFS_IOC_QUOTA_RESCAN_WAIT:
		return btrfs_ioctl_quota_rescan_wait(file, argp);
4592 4593
	case BTRFS_IOC_DEV_REPLACE:
		return btrfs_ioctl_dev_replace(root, argp);
4594 4595
	case BTRFS_IOC_GET_FSLABEL:
		return btrfs_ioctl_get_fslabel(file, argp);
4596 4597
	case BTRFS_IOC_SET_FSLABEL:
		return btrfs_ioctl_set_fslabel(file, argp);
M
Mark Fasheh 已提交
4598 4599
	case BTRFS_IOC_FILE_EXTENT_SAME:
		return btrfs_ioctl_file_extent_same(file, argp);
C
Christoph Hellwig 已提交
4600 4601 4602 4603
	}

	return -ENOTTY;
}