ioctl.c 97.1 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>
C
Chris Mason 已提交
46
#include "compat.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

60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89
/* 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
90 91 92 93 94 95 96
	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;
97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124

	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.
 *
125
 * Currently only the compression flags and the cow flags are inherited.
126 127 128
 */
void btrfs_inherit_iflags(struct inode *inode, struct inode *dir)
{
129 130 131 132 133 134
	unsigned int flags;

	if (!dir)
		return;

	flags = BTRFS_I(dir)->flags;
135

136 137 138 139 140 141 142 143
	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
144
	if (flags & BTRFS_INODE_NODATACOW) {
145
		BTRFS_I(inode)->flags |= BTRFS_INODE_NODATACOW;
L
Liu Bo 已提交
146 147 148
		if (S_ISREG(inode->i_mode))
			BTRFS_I(inode)->flags |= BTRFS_INODE_NODATASUM;
	}
149 150 151 152 153 154

	btrfs_update_iflags(inode);
}

static int btrfs_ioctl_getflags(struct file *file, void __user *arg)
{
A
Al Viro 已提交
155
	struct btrfs_inode *ip = BTRFS_I(file_inode(file));
156 157 158 159 160 161 162
	unsigned int flags = btrfs_flags_to_ioctl(ip->flags);

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

163 164 165 166 167
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
168 169
		      FS_NOCOMP_FL | FS_COMPR_FL |
		      FS_NOCOW_FL))
170 171 172 173 174 175 176 177
		return -EOPNOTSUPP;

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

	return 0;
}

178 179
static int btrfs_ioctl_setflags(struct file *file, void __user *arg)
{
A
Al Viro 已提交
180
	struct inode *inode = file_inode(file);
181 182 183 184 185
	struct btrfs_inode *ip = BTRFS_I(inode);
	struct btrfs_root *root = ip->root;
	struct btrfs_trans_handle *trans;
	unsigned int flags, oldflags;
	int ret;
186 187
	u64 ip_oldflags;
	unsigned int i_oldflags;
188
	umode_t mode;
189

L
Li Zefan 已提交
190 191 192
	if (btrfs_root_readonly(root))
		return -EROFS;

193 194 195
	if (copy_from_user(&flags, arg, sizeof(flags)))
		return -EFAULT;

196 197 198
	ret = check_flags(flags);
	if (ret)
		return ret;
C
Christoph Hellwig 已提交
199

200
	if (!inode_owner_or_capable(inode))
201 202
		return -EACCES;

203 204 205 206
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;

207 208
	mutex_lock(&inode->i_mutex);

209 210
	ip_oldflags = ip->flags;
	i_oldflags = inode->i_flags;
211
	mode = inode->i_mode;
212

213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245
	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;
246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270
	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;
		}
	}
271

272 273 274 275 276 277 278 279 280 281 282
	/*
	 * 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
283 284
	} else {
		ip->flags &= ~(BTRFS_INODE_COMPRESS | BTRFS_INODE_NOCOMPRESS);
285
	}
286

287
	trans = btrfs_start_transaction(root, 1);
288 289 290 291
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto out_drop;
	}
292

293
	btrfs_update_iflags(inode);
294
	inode_inc_iversion(inode);
295
	inode->i_ctime = CURRENT_TIME;
296 297 298
	ret = btrfs_update_inode(trans, root, inode);

	btrfs_end_transaction(trans, root);
299 300 301 302 303
 out_drop:
	if (ret) {
		ip->flags = ip_oldflags;
		inode->i_flags = i_oldflags;
	}
304 305 306

 out_unlock:
	mutex_unlock(&inode->i_mutex);
307
	mnt_drop_write_file(file);
308
	return ret;
309 310 311 312
}

static int btrfs_ioctl_getversion(struct file *file, int __user *arg)
{
A
Al Viro 已提交
313
	struct inode *inode = file_inode(file);
314 315 316

	return put_user(inode->i_generation, arg);
}
C
Christoph Hellwig 已提交
317

318 319
static noinline int btrfs_ioctl_fitrim(struct file *file, void __user *arg)
{
320
	struct btrfs_fs_info *fs_info = btrfs_sb(fdentry(file)->d_sb);
321 322 323 324 325
	struct btrfs_device *device;
	struct request_queue *q;
	struct fstrim_range range;
	u64 minlen = ULLONG_MAX;
	u64 num_devices = 0;
326
	u64 total_bytes = btrfs_super_total_bytes(fs_info->super_copy);
327 328 329 330 331
	int ret;

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

332 333 334
	rcu_read_lock();
	list_for_each_entry_rcu(device, &fs_info->fs_devices->devices,
				dev_list) {
335 336 337 338 339 340 341 342 343
		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);
		}
	}
344
	rcu_read_unlock();
345

346 347 348 349
	if (!num_devices)
		return -EOPNOTSUPP;
	if (copy_from_user(&range, arg, sizeof(range)))
		return -EFAULT;
350 351
	if (range.start > total_bytes ||
	    range.len < fs_info->sb->s_blocksize)
352
		return -EINVAL;
353

354
	range.len = min(range.len, total_bytes - range.start);
355
	range.minlen = max(range.minlen, minlen);
356
	ret = btrfs_trim_fs(fs_info->tree_root, &range);
357 358 359 360 361 362 363 364 365
	if (ret < 0)
		return ret;

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

	return 0;
}

366
static noinline int create_subvol(struct inode *dir,
367
				  struct dentry *dentry,
S
Sage Weil 已提交
368
				  char *name, int namelen,
A
Arne Jansen 已提交
369
				  u64 *async_transid,
370
				  struct btrfs_qgroup_inherit *inherit)
C
Christoph Hellwig 已提交
371 372 373 374 375 376
{
	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;
377
	struct btrfs_root *root = BTRFS_I(dir)->root;
378
	struct btrfs_root *new_root;
379
	struct btrfs_block_rsv block_rsv;
380
	struct timespec cur_time = CURRENT_TIME;
C
Christoph Hellwig 已提交
381 382 383 384
	int ret;
	int err;
	u64 objectid;
	u64 new_dirid = BTRFS_FIRST_FREE_OBJECTID;
385
	u64 index = 0;
386
	u64 qgroup_reserved;
387
	uuid_le new_uuid;
C
Christoph Hellwig 已提交
388

389
	ret = btrfs_find_free_objectid(root->fs_info->tree_root, &objectid);
390
	if (ret)
391
		return ret;
392

393
	btrfs_init_block_rsv(&block_rsv, BTRFS_BLOCK_RSV_TEMP);
J
Josef Bacik 已提交
394
	/*
395 396
	 * The same as the snapshot creation, please see the comment
	 * of create_snapshot().
J
Josef Bacik 已提交
397
	 */
398 399 400 401 402 403 404 405 406 407 408 409
	ret = btrfs_subvolume_reserve_metadata(root, &block_rsv,
					       7, &qgroup_reserved);
	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
410

411
	ret = btrfs_qgroup_inherit(trans, root->fs_info, 0, objectid, inherit);
A
Arne Jansen 已提交
412 413 414
	if (ret)
		goto fail;

415
	leaf = btrfs_alloc_free_block(trans, root, root->leafsize,
416
				      0, objectid, NULL, 0, 0, 0);
417 418 419 420
	if (IS_ERR(leaf)) {
		ret = PTR_ERR(leaf);
		goto fail;
	}
C
Christoph Hellwig 已提交
421

422
	memset_extent_buffer(leaf, 0, 0, sizeof(struct btrfs_header));
C
Christoph Hellwig 已提交
423 424
	btrfs_set_header_bytenr(leaf, leaf->start);
	btrfs_set_header_generation(leaf, trans->transid);
425
	btrfs_set_header_backref_rev(leaf, BTRFS_MIXED_BACKREF_REV);
C
Christoph Hellwig 已提交
426 427 428 429 430
	btrfs_set_header_owner(leaf, objectid);

	write_extent_buffer(leaf, root->fs_info->fsid,
			    (unsigned long)btrfs_header_fsid(leaf),
			    BTRFS_FSID_SIZE);
431 432 433
	write_extent_buffer(leaf, root->fs_info->chunk_tree_uuid,
			    (unsigned long)btrfs_header_chunk_tree_uuid(leaf),
			    BTRFS_UUID_SIZE);
C
Christoph Hellwig 已提交
434 435
	btrfs_mark_buffer_dirty(leaf);

436 437
	memset(&root_item, 0, sizeof(root_item));

C
Christoph Hellwig 已提交
438 439 440 441
	inode_item = &root_item.inode;
	inode_item->generation = cpu_to_le64(1);
	inode_item->size = cpu_to_le64(3);
	inode_item->nlink = cpu_to_le32(1);
442
	inode_item->nbytes = cpu_to_le64(root->leafsize);
C
Christoph Hellwig 已提交
443 444
	inode_item->mode = cpu_to_le32(S_IFDIR | 0755);

445 446 447 448
	root_item.flags = 0;
	root_item.byte_limit = 0;
	inode_item->flags = cpu_to_le64(BTRFS_INODE_ROOT_ITEM_INIT);

C
Christoph Hellwig 已提交
449
	btrfs_set_root_bytenr(&root_item, leaf->start);
450
	btrfs_set_root_generation(&root_item, trans->transid);
C
Christoph Hellwig 已提交
451 452
	btrfs_set_root_level(&root_item, 0);
	btrfs_set_root_refs(&root_item, 1);
453
	btrfs_set_root_used(&root_item, leaf->len);
Y
Yan Zheng 已提交
454
	btrfs_set_root_last_snapshot(&root_item, 0);
C
Christoph Hellwig 已提交
455

456 457 458 459 460
	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);
	root_item.otime.sec = cpu_to_le64(cur_time.tv_sec);
461
	root_item.otime.nsec = cpu_to_le32(cur_time.tv_nsec);
462 463 464
	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
465

466
	btrfs_tree_unlock(leaf);
C
Christoph Hellwig 已提交
467 468 469 470 471 472
	free_extent_buffer(leaf);
	leaf = NULL;

	btrfs_set_root_dirid(&root_item, new_dirid);

	key.objectid = objectid;
473
	key.offset = 0;
C
Christoph Hellwig 已提交
474 475 476 477 478 479
	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;

480 481
	key.offset = (u64)-1;
	new_root = btrfs_read_fs_root_no_name(root->fs_info, &key);
482 483 484 485 486
	if (IS_ERR(new_root)) {
		btrfs_abort_transaction(trans, root, PTR_ERR(new_root));
		ret = PTR_ERR(new_root);
		goto fail;
	}
487 488 489

	btrfs_record_root_in_trans(trans, new_root);

490
	ret = btrfs_create_subvol_root(trans, new_root, new_dirid);
491 492
	if (ret) {
		/* We potentially lose an unused inode item here */
493
		btrfs_abort_transaction(trans, root, ret);
494 495 496
		goto fail;
	}

C
Christoph Hellwig 已提交
497 498 499
	/*
	 * insert the directory item
	 */
500
	ret = btrfs_set_inode_index(dir, &index);
501 502 503 504
	if (ret) {
		btrfs_abort_transaction(trans, root, ret);
		goto fail;
	}
505 506

	ret = btrfs_insert_dir_item(trans, root,
507
				    name, namelen, dir, &key,
508
				    BTRFS_FT_DIR, index);
509 510
	if (ret) {
		btrfs_abort_transaction(trans, root, ret);
C
Christoph Hellwig 已提交
511
		goto fail;
512
	}
513

514 515 516 517
	btrfs_i_size_write(dir, dir->i_size + namelen * 2);
	ret = btrfs_update_inode(trans, root, dir);
	BUG_ON(ret);

518
	ret = btrfs_add_root_ref(trans, root->fs_info->tree_root,
519
				 objectid, root->root_key.objectid,
L
Li Zefan 已提交
520
				 btrfs_ino(dir), index, name, namelen);
521

522
	BUG_ON(ret);
C
Christoph Hellwig 已提交
523 524

fail:
525 526
	trans->block_rsv = NULL;
	trans->bytes_reserved = 0;
S
Sage Weil 已提交
527 528 529
	if (async_transid) {
		*async_transid = trans->transid;
		err = btrfs_commit_transaction_async(trans, root, 1);
530 531
		if (err)
			err = btrfs_commit_transaction(trans, root);
S
Sage Weil 已提交
532 533 534
	} else {
		err = btrfs_commit_transaction(trans, root);
	}
C
Christoph Hellwig 已提交
535 536
	if (err && !ret)
		ret = err;
537 538 539

	if (!ret)
		d_instantiate(dentry, btrfs_lookup_dentry(dir, dentry));
540 541
out:
	btrfs_subvolume_release_metadata(root, &block_rsv, qgroup_reserved);
C
Christoph Hellwig 已提交
542 543 544
	return ret;
}

545 546 547 548
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
549
{
550
	struct inode *inode;
C
Christoph Hellwig 已提交
551 552
	struct btrfs_pending_snapshot *pending_snapshot;
	struct btrfs_trans_handle *trans;
553
	int ret;
C
Christoph Hellwig 已提交
554 555 556 557

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

558
	pending_snapshot = kzalloc(sizeof(*pending_snapshot), GFP_NOFS);
559 560 561
	if (!pending_snapshot)
		return -ENOMEM;

562 563
	btrfs_init_block_rsv(&pending_snapshot->block_rsv,
			     BTRFS_BLOCK_RSV_TEMP);
564 565 566 567 568 569 570 571 572 573 574 575 576
	/*
	 * 1 - parent dir inode
	 * 2 - dir entries
	 * 1 - root item
	 * 2 - root ref/backref
	 * 1 - root of snapshot
	 */
	ret = btrfs_subvolume_reserve_metadata(BTRFS_I(dir)->root,
					&pending_snapshot->block_rsv, 7,
					&pending_snapshot->qgroup_reserved);
	if (ret)
		goto out;

577
	pending_snapshot->dentry = dentry;
C
Christoph Hellwig 已提交
578
	pending_snapshot->root = root;
L
Li Zefan 已提交
579
	pending_snapshot->readonly = readonly;
580
	pending_snapshot->dir = dir;
581
	pending_snapshot->inherit = inherit;
582

583
	trans = btrfs_start_transaction(root, 0);
584 585 586 587 588
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto fail;
	}

589
	spin_lock(&root->fs_info->trans_lock);
C
Christoph Hellwig 已提交
590 591
	list_add(&pending_snapshot->list,
		 &trans->transaction->pending_snapshots);
592
	spin_unlock(&root->fs_info->trans_lock);
S
Sage Weil 已提交
593 594 595 596
	if (async_transid) {
		*async_transid = trans->transid;
		ret = btrfs_commit_transaction_async(trans,
				     root->fs_info->extent_root, 1);
597 598
		if (ret)
			ret = btrfs_commit_transaction(trans, root);
S
Sage Weil 已提交
599 600 601 602
	} else {
		ret = btrfs_commit_transaction(trans,
					       root->fs_info->extent_root);
	}
603
	if (ret)
604
		goto fail;
605 606 607 608 609

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

610 611 612
	ret = btrfs_orphan_cleanup(pending_snapshot->snap);
	if (ret)
		goto fail;
C
Christoph Hellwig 已提交
613

614
	inode = btrfs_lookup_dentry(dentry->d_parent->d_inode, dentry);
615 616 617 618 619 620 621 622
	if (IS_ERR(inode)) {
		ret = PTR_ERR(inode);
		goto fail;
	}
	BUG_ON(!inode);
	d_instantiate(dentry, inode);
	ret = 0;
fail:
623 624 625 626
	btrfs_subvolume_release_metadata(BTRFS_I(dir)->root,
					 &pending_snapshot->block_rsv,
					 pending_snapshot->qgroup_reserved);
out:
627
	kfree(pending_snapshot);
C
Christoph Hellwig 已提交
628 629 630
	return ret;
}

631 632 633 634 635 636
/*  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)
{
637
	kuid_t fsuid = current_fsuid();
638 639 640

	if (!(dir->i_mode & S_ISVTX))
		return 0;
641
	if (uid_eq(inode->i_uid, fsuid))
642
		return 0;
643
	if (uid_eq(dir->i_uid, fsuid))
644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675
		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().
 */

static int btrfs_may_delete(struct inode *dir,struct dentry *victim,int isdir)
{
	int error;

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

	BUG_ON(victim->d_parent->d_inode != dir);
676
	audit_inode_child(dir, victim, AUDIT_TYPE_CHILD_DELETE);
677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700

	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;
}

701 702 703 704 705 706 707 708 709 710 711 712 713 714 715
/* 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.
 */
716 717
static noinline int btrfs_mksubvol(struct path *parent,
				   char *name, int namelen,
S
Sage Weil 已提交
718
				   struct btrfs_root *snap_src,
A
Arne Jansen 已提交
719
				   u64 *async_transid, bool readonly,
720
				   struct btrfs_qgroup_inherit *inherit)
721
{
722
	struct inode *dir  = parent->dentry->d_inode;
723 724 725
	struct dentry *dentry;
	int error;

726 727 728
	error = mutex_lock_killable_nested(&dir->i_mutex, I_MUTEX_PARENT);
	if (error == -EINTR)
		return error;
729 730 731 732 733 734 735 736 737 738

	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;

739
	error = btrfs_may_create(dir, dentry);
740
	if (error)
741
		goto out_dput;
742

C
Chris Mason 已提交
743 744 745 746 747 748 749 750 751 752
	/*
	 * 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;

753 754 755 756 757
	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;

758
	if (snap_src) {
759
		error = create_snapshot(snap_src, dir, dentry, name, namelen,
A
Arne Jansen 已提交
760
					async_transid, readonly, inherit);
761
	} else {
762 763
		error = create_subvol(dir, dentry, name, namelen,
				      async_transid, inherit);
764
	}
765 766 767 768
	if (!error)
		fsnotify_mkdir(dir, dentry);
out_up_read:
	up_read(&BTRFS_I(dir)->root->fs_info->subvol_sem);
769 770 771
out_dput:
	dput(dentry);
out_unlock:
772
	mutex_unlock(&dir->i_mutex);
773 774 775
	return error;
}

C
Chris Mason 已提交
776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826
/*
 * 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 btrfs_key max_key;
	struct extent_buffer *leaf;
	struct btrfs_file_extent_item *extent;
	int type;
	int ret;
827
	u64 ino = btrfs_ino(inode);
C
Chris Mason 已提交
828 829 830 831 832

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

833
	min_key.objectid = ino;
C
Chris Mason 已提交
834 835 836
	min_key.type = BTRFS_EXTENT_DATA_KEY;
	min_key.offset = *off;

837
	max_key.objectid = ino;
C
Chris Mason 已提交
838 839 840 841 842 843 844
	max_key.type = (u8)-1;
	max_key.offset = (u64)-1;

	path->keep_locks = 1;

	while(1) {
		ret = btrfs_search_forward(root, &min_key, &max_key,
845
					   path, newer_than);
C
Chris Mason 已提交
846 847
		if (ret != 0)
			goto none;
848
		if (min_key.objectid != ino)
C
Chris Mason 已提交
849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876
			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
877
static struct extent_map *defrag_lookup_extent(struct inode *inode, u64 start)
878 879
{
	struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
L
Li Zefan 已提交
880 881 882
	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
	struct extent_map *em;
	u64 len = PAGE_CACHE_SIZE;
883

L
Li Zefan 已提交
884 885 886 887
	/*
	 * hopefully we have this extent in the tree already, try without
	 * the full extent lock
	 */
888
	read_lock(&em_tree->lock);
L
Li Zefan 已提交
889
	em = lookup_extent_mapping(em_tree, start, len);
890 891
	read_unlock(&em_tree->lock);

L
Li Zefan 已提交
892 893 894 895 896 897 898 899 900 901 902 903
	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;
}
904

L
Li Zefan 已提交
905 906 907 908 909 910 911 912 913 914 915 916 917 918
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);
919 920 921
	return ret;
}

L
Li Zefan 已提交
922
static int should_defrag_range(struct inode *inode, u64 start, int thresh,
923 924
			       u64 *last_len, u64 *skip, u64 *defrag_end,
			       int compress)
925
{
L
Li Zefan 已提交
926
	struct extent_map *em;
927
	int ret = 1;
L
Li Zefan 已提交
928
	bool next_mergeable = true;
929 930

	/*
931
	 * make sure that once we start defragging an extent, we keep on
932 933 934 935 936 937 938
	 * defragging it
	 */
	if (start < *defrag_end)
		return 1;

	*skip = 0;

L
Li Zefan 已提交
939 940 941
	em = defrag_lookup_extent(inode, start);
	if (!em)
		return 0;
942 943

	/* this will cover holes, and inline extents */
944
	if (em->block_start >= EXTENT_MAP_LAST_BYTE) {
945
		ret = 0;
946 947 948
		goto out;
	}

L
Li Zefan 已提交
949
	next_mergeable = defrag_check_next_extent(inode, em);
950 951

	/*
L
Li Zefan 已提交
952 953
	 * we hit a real extent, if it is big or the next extent is not a
	 * real extent, don't bother defragging it
954
	 */
955
	if (!compress && (*last_len == 0 || *last_len >= thresh) &&
L
Li Zefan 已提交
956
	    (em->len >= thresh || !next_mergeable))
957
		ret = 0;
958
out:
959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978
	/*
	 * 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
979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994
/*
 * 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
995
{
C
Chris Mason 已提交
996 997 998 999
	unsigned long file_end;
	u64 isize = i_size_read(inode);
	u64 page_start;
	u64 page_end;
1000
	u64 page_cnt;
C
Chris Mason 已提交
1001 1002 1003
	int ret;
	int i;
	int i_done;
1004
	struct btrfs_ordered_extent *ordered;
C
Chris Mason 已提交
1005
	struct extent_state *cached_state = NULL;
1006
	struct extent_io_tree *tree;
1007
	gfp_t mask = btrfs_alloc_write_mask(inode->i_mapping);
C
Chris Mason 已提交
1008 1009

	file_end = (isize - 1) >> PAGE_CACHE_SHIFT;
1010 1011 1012 1013
	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
1014 1015

	ret = btrfs_delalloc_reserve_space(inode,
1016
					   page_cnt << PAGE_CACHE_SHIFT);
C
Chris Mason 已提交
1017 1018 1019
	if (ret)
		return ret;
	i_done = 0;
1020
	tree = &BTRFS_I(inode)->io_tree;
C
Chris Mason 已提交
1021 1022

	/* step one, lock all the pages */
1023
	for (i = 0; i < page_cnt; i++) {
C
Chris Mason 已提交
1024
		struct page *page;
1025
again:
1026
		page = find_or_create_page(inode->i_mapping,
1027
					   start_index + i, mask);
C
Chris Mason 已提交
1028 1029 1030
		if (!page)
			break;

1031 1032 1033
		page_start = page_offset(page);
		page_end = page_start + PAGE_CACHE_SIZE - 1;
		while (1) {
1034
			lock_extent(tree, page_start, page_end);
1035 1036
			ordered = btrfs_lookup_ordered_extent(inode,
							      page_start);
1037
			unlock_extent(tree, page_start, page_end);
1038 1039 1040 1041 1042 1043 1044
			if (!ordered)
				break;

			unlock_page(page);
			btrfs_start_ordered_extent(inode, ordered, 1);
			btrfs_put_ordered_extent(ordered);
			lock_page(page);
1045 1046 1047 1048 1049 1050 1051 1052 1053
			/*
			 * 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;
			}
1054 1055
		}

C
Chris Mason 已提交
1056 1057 1058 1059 1060 1061 1062 1063 1064 1065
		if (!PageUptodate(page)) {
			btrfs_readpage(NULL, page);
			lock_page(page);
			if (!PageUptodate(page)) {
				unlock_page(page);
				page_cache_release(page);
				ret = -EIO;
				break;
			}
		}
1066 1067 1068 1069 1070 1071 1072

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

C
Chris Mason 已提交
1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092
		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,
1093
			 page_start, page_end - 1, 0, &cached_state);
C
Chris Mason 已提交
1094 1095
	clear_extent_bit(&BTRFS_I(inode)->io_tree, page_start,
			  page_end - 1, EXTENT_DIRTY | EXTENT_DELALLOC |
1096 1097
			  EXTENT_DO_ACCOUNTING | EXTENT_DEFRAG, 0, 0,
			  &cached_state, GFP_NOFS);
C
Chris Mason 已提交
1098

1099
	if (i_done != page_cnt) {
1100 1101 1102
		spin_lock(&BTRFS_I(inode)->lock);
		BTRFS_I(inode)->outstanding_extents++;
		spin_unlock(&BTRFS_I(inode)->lock);
C
Chris Mason 已提交
1103
		btrfs_delalloc_release_space(inode,
1104
				     (page_cnt - i_done) << PAGE_CACHE_SHIFT);
C
Chris Mason 已提交
1105 1106 1107
	}


1108 1109
	set_extent_defrag(&BTRFS_I(inode)->io_tree, page_start, page_end - 1,
			  &cached_state, GFP_NOFS);
C
Chris Mason 已提交
1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128

	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]);
	}
1129
	btrfs_delalloc_release_space(inode, page_cnt << PAGE_CACHE_SHIFT);
C
Chris Mason 已提交
1130 1131 1132 1133 1134 1135 1136 1137 1138 1139
	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
1140
	unsigned long last_index;
1141
	u64 isize = i_size_read(inode);
1142 1143 1144
	u64 last_len = 0;
	u64 skip = 0;
	u64 defrag_end = 0;
C
Chris Mason 已提交
1145
	u64 newer_off = range->start;
C
Christoph Hellwig 已提交
1146
	unsigned long i;
1147
	unsigned long ra_index = 0;
C
Christoph Hellwig 已提交
1148
	int ret;
C
Chris Mason 已提交
1149
	int defrag_count = 0;
1150
	int compress_type = BTRFS_COMPRESS_ZLIB;
C
Chris Mason 已提交
1151
	int extent_thresh = range->extent_thresh;
1152 1153
	int max_cluster = (256 * 1024) >> PAGE_CACHE_SHIFT;
	int cluster = max_cluster;
C
Chris Mason 已提交
1154 1155 1156 1157 1158
	u64 new_align = ~((u64)128 * 1024 - 1);
	struct page **pages = NULL;

	if (extent_thresh == 0)
		extent_thresh = 256 * 1024;
1159 1160 1161 1162 1163 1164 1165

	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
1166

1167
	if (isize == 0)
1168 1169
		return 0;

C
Chris Mason 已提交
1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182
	/*
	 * 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;
	}

1183
	pages = kmalloc(sizeof(struct page *) * max_cluster,
C
Chris Mason 已提交
1184 1185 1186 1187 1188 1189 1190
			GFP_NOFS);
	if (!pages) {
		ret = -ENOMEM;
		goto out_ra;
	}

	/* find the last page to defrag */
C
Chris Mason 已提交
1191
	if (range->start + range->len > range->start) {
1192
		last_index = min_t(u64, isize - 1,
C
Chris Mason 已提交
1193 1194
			 range->start + range->len - 1) >> PAGE_CACHE_SHIFT;
	} else {
1195
		last_index = (isize - 1) >> PAGE_CACHE_SHIFT;
C
Chris Mason 已提交
1196 1197
	}

C
Chris Mason 已提交
1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213
	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
1214
		max_to_defrag = last_index + 1;
C
Chris Mason 已提交
1215

L
Li Zefan 已提交
1216 1217 1218 1219 1220 1221 1222
	/*
	 * 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;

1223 1224 1225
	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
1226 1227 1228 1229 1230 1231 1232
		/*
		 * make sure we stop running if someone unmounts
		 * the FS
		 */
		if (!(inode->i_sb->s_flags & MS_ACTIVE))
			break;

1233 1234 1235 1236 1237 1238
		if (btrfs_defrag_cancelled(root->fs_info)) {
			printk(KERN_DEBUG "btrfs: defrag_file cancelled\n");
			ret = -EAGAIN;
			break;
		}

1239
		if (!should_defrag_range(inode, (u64)i << PAGE_CACHE_SHIFT,
L
Li Zefan 已提交
1240
					 extent_thresh, &last_len, &skip,
1241 1242
					 &defrag_end, range->flags &
					 BTRFS_DEFRAG_RANGE_COMPRESS)) {
1243 1244 1245 1246 1247 1248 1249 1250 1251
			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;
		}
1252 1253 1254 1255 1256 1257 1258 1259 1260

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

C
Chris Mason 已提交
1261
		if (range->flags & BTRFS_DEFRAG_RANGE_COMPRESS)
1262
			BTRFS_I(inode)->force_compress = compress_type;
1263

1264 1265 1266 1267 1268 1269
		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;
		}
1270

1271
		mutex_lock(&inode->i_mutex);
1272
		ret = cluster_pages_for_defrag(inode, pages, i, cluster);
1273 1274
		if (ret < 0) {
			mutex_unlock(&inode->i_mutex);
C
Chris Mason 已提交
1275
			goto out_ra;
1276
		}
C
Chris Mason 已提交
1277 1278

		defrag_count += ret;
1279
		balance_dirty_pages_ratelimited(inode->i_mapping);
1280
		mutex_unlock(&inode->i_mutex);
C
Chris Mason 已提交
1281 1282 1283 1284 1285

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

1286 1287 1288
			if (ret > 0)
				i += ret;

C
Chris Mason 已提交
1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299
			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
1300
			}
C
Chris Mason 已提交
1301
		} else {
1302
			if (ret > 0) {
L
Li Zefan 已提交
1303
				i += ret;
1304 1305
				last_len += ret << PAGE_CACHE_SHIFT;
			} else {
L
Li Zefan 已提交
1306
				i++;
1307 1308
				last_len = 0;
			}
C
Christoph Hellwig 已提交
1309 1310 1311
		}
	}

C
Chris Mason 已提交
1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329
	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);

		mutex_lock(&inode->i_mutex);
1330
		BTRFS_I(inode)->force_compress = BTRFS_COMPRESS_NONE;
C
Chris Mason 已提交
1331 1332 1333
		mutex_unlock(&inode->i_mutex);
	}

1334
	if (range->compress_type == BTRFS_COMPRESS_LZO) {
1335
		btrfs_set_fs_incompat(root->fs_info, COMPRESS_LZO);
1336 1337
	}

1338
	ret = defrag_count;
1339

C
Chris Mason 已提交
1340 1341 1342 1343
out_ra:
	if (!file)
		kfree(ra);
	kfree(pages);
1344
	return ret;
C
Christoph Hellwig 已提交
1345 1346
}

1347
static noinline int btrfs_ioctl_resize(struct file *file,
1348
					void __user *arg)
C
Christoph Hellwig 已提交
1349 1350 1351 1352
{
	u64 new_size;
	u64 old_size;
	u64 devid = 1;
A
Al Viro 已提交
1353
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
C
Christoph Hellwig 已提交
1354 1355 1356 1357 1358 1359 1360 1361
	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;

1362 1363 1364
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

1365 1366 1367 1368
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;

1369 1370 1371
	if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
			1)) {
		pr_info("btrfs: dev add/delete/balance/replace/resize operation in progress\n");
1372
		mnt_drop_write_file(file);
1373
		return -EINVAL;
1374 1375
	}

1376
	mutex_lock(&root->fs_info->volume_mutex);
L
Li Zefan 已提交
1377
	vol_args = memdup_user(arg, sizeof(*vol_args));
1378 1379 1380 1381
	if (IS_ERR(vol_args)) {
		ret = PTR_ERR(vol_args);
		goto out;
	}
1382 1383

	vol_args->name[BTRFS_PATH_NAME_MAX] = '\0';
C
Christoph Hellwig 已提交
1384 1385 1386 1387 1388 1389 1390 1391 1392

	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);
1393 1394 1395 1396
		if (!devid) {
			ret = -EINVAL;
			goto out_free;
		}
1397
		printk(KERN_INFO "btrfs: resizing devid %llu\n",
1398
		       (unsigned long long)devid);
C
Christoph Hellwig 已提交
1399
	}
M
Miao Xie 已提交
1400

1401
	device = btrfs_find_device(root->fs_info, devid, NULL, NULL);
C
Christoph Hellwig 已提交
1402
	if (!device) {
1403
		printk(KERN_INFO "btrfs: resizer unable to find device %llu\n",
1404
		       (unsigned long long)devid);
1405
		ret = -ENODEV;
1406
		goto out_free;
C
Christoph Hellwig 已提交
1407
	}
M
Miao Xie 已提交
1408 1409

	if (!device->writeable) {
L
Liu Bo 已提交
1410
		printk(KERN_INFO "btrfs: resizer unable to apply on "
M
Miao Xie 已提交
1411
		       "readonly device %llu\n",
1412
		       (unsigned long long)devid);
1413
		ret = -EPERM;
L
Liu Bo 已提交
1414 1415 1416
		goto out_free;
	}

C
Christoph Hellwig 已提交
1417 1418 1419 1420 1421 1422 1423 1424 1425 1426
	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
1427
		new_size = memparse(sizestr, NULL);
C
Christoph Hellwig 已提交
1428 1429
		if (new_size == 0) {
			ret = -EINVAL;
1430
			goto out_free;
C
Christoph Hellwig 已提交
1431 1432 1433
		}
	}

1434
	if (device->is_tgtdev_for_dev_replace) {
1435
		ret = -EPERM;
1436 1437 1438
		goto out_free;
	}

C
Christoph Hellwig 已提交
1439 1440 1441 1442 1443
	old_size = device->total_bytes;

	if (mod < 0) {
		if (new_size > old_size) {
			ret = -EINVAL;
1444
			goto out_free;
C
Christoph Hellwig 已提交
1445 1446 1447 1448 1449 1450 1451 1452
		}
		new_size = old_size - new_size;
	} else if (mod > 0) {
		new_size = old_size + new_size;
	}

	if (new_size < 256 * 1024 * 1024) {
		ret = -EINVAL;
1453
		goto out_free;
C
Christoph Hellwig 已提交
1454 1455 1456
	}
	if (new_size > device->bdev->bd_inode->i_size) {
		ret = -EFBIG;
1457
		goto out_free;
C
Christoph Hellwig 已提交
1458 1459 1460 1461 1462
	}

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

1463 1464 1465
	printk_in_rcu(KERN_INFO "btrfs: new size for %s is %llu\n",
		      rcu_str_deref(device->name),
		      (unsigned long long)new_size);
C
Christoph Hellwig 已提交
1466 1467

	if (new_size > old_size) {
1468
		trans = btrfs_start_transaction(root, 0);
1469 1470
		if (IS_ERR(trans)) {
			ret = PTR_ERR(trans);
1471
			goto out_free;
1472
		}
C
Christoph Hellwig 已提交
1473 1474
		ret = btrfs_grow_device(trans, device, new_size);
		btrfs_commit_transaction(trans, root);
1475
	} else if (new_size < old_size) {
C
Christoph Hellwig 已提交
1476
		ret = btrfs_shrink_device(device, new_size);
1477
	} /* equal, nothing need to do */
C
Christoph Hellwig 已提交
1478

1479
out_free:
C
Christoph Hellwig 已提交
1480
	kfree(vol_args);
1481 1482
out:
	mutex_unlock(&root->fs_info->volume_mutex);
1483
	atomic_set(&root->fs_info->mutually_exclusive_operation_running, 0);
1484
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
1485 1486 1487
	return ret;
}

S
Sage Weil 已提交
1488
static noinline int btrfs_ioctl_snap_create_transid(struct file *file,
A
Arne Jansen 已提交
1489 1490
				char *name, unsigned long fd, int subvol,
				u64 *transid, bool readonly,
1491
				struct btrfs_qgroup_inherit *inherit)
C
Christoph Hellwig 已提交
1492 1493
{
	int namelen;
1494
	int ret = 0;
C
Christoph Hellwig 已提交
1495

1496 1497 1498 1499
	ret = mnt_want_write_file(file);
	if (ret)
		goto out;

S
Sage Weil 已提交
1500 1501
	namelen = strlen(name);
	if (strchr(name, '/')) {
C
Christoph Hellwig 已提交
1502
		ret = -EINVAL;
1503
		goto out_drop_write;
C
Christoph Hellwig 已提交
1504 1505
	}

1506 1507 1508
	if (name[0] == '.' &&
	   (namelen == 1 || (name[1] == '.' && namelen == 2))) {
		ret = -EEXIST;
1509
		goto out_drop_write;
1510 1511
	}

1512
	if (subvol) {
S
Sage Weil 已提交
1513
		ret = btrfs_mksubvol(&file->f_path, name, namelen,
A
Arne Jansen 已提交
1514
				     NULL, transid, readonly, inherit);
1515
	} else {
1516
		struct fd src = fdget(fd);
1517
		struct inode *src_inode;
1518
		if (!src.file) {
1519
			ret = -EINVAL;
1520
			goto out_drop_write;
1521 1522
		}

A
Al Viro 已提交
1523 1524
		src_inode = file_inode(src.file);
		if (src_inode->i_sb != file_inode(file)->i_sb) {
C
Chris Mason 已提交
1525 1526
			printk(KERN_INFO "btrfs: Snapshot src from "
			       "another FS\n");
1527
			ret = -EINVAL;
1528 1529 1530 1531
		} else {
			ret = btrfs_mksubvol(&file->f_path, name, namelen,
					     BTRFS_I(src_inode)->root,
					     transid, readonly, inherit);
1532
		}
1533
		fdput(src);
1534
	}
1535 1536
out_drop_write:
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
1537
out:
S
Sage Weil 已提交
1538 1539 1540 1541
	return ret;
}

static noinline int btrfs_ioctl_snap_create(struct file *file,
1542
					    void __user *arg, int subvol)
S
Sage Weil 已提交
1543
{
1544
	struct btrfs_ioctl_vol_args *vol_args;
S
Sage Weil 已提交
1545 1546
	int ret;

1547 1548 1549 1550
	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
1551

1552
	ret = btrfs_ioctl_snap_create_transid(file, vol_args->name,
L
Li Zefan 已提交
1553
					      vol_args->fd, subvol,
A
Arne Jansen 已提交
1554
					      NULL, false, NULL);
1555

1556 1557 1558
	kfree(vol_args);
	return ret;
}
1559

1560 1561 1562 1563 1564 1565 1566
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
1567
	bool readonly = false;
A
Arne Jansen 已提交
1568
	struct btrfs_qgroup_inherit *inherit = NULL;
1569

1570 1571 1572 1573
	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';
1574

L
Li Zefan 已提交
1575
	if (vol_args->flags &
A
Arne Jansen 已提交
1576 1577
	    ~(BTRFS_SUBVOL_CREATE_ASYNC | BTRFS_SUBVOL_RDONLY |
	      BTRFS_SUBVOL_QGROUP_INHERIT)) {
L
Li Zefan 已提交
1578
		ret = -EOPNOTSUPP;
1579
		goto out;
S
Sage Weil 已提交
1580
	}
1581 1582 1583

	if (vol_args->flags & BTRFS_SUBVOL_CREATE_ASYNC)
		ptr = &transid;
L
Li Zefan 已提交
1584 1585
	if (vol_args->flags & BTRFS_SUBVOL_RDONLY)
		readonly = true;
A
Arne Jansen 已提交
1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596
	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;
		}
	}
1597 1598

	ret = btrfs_ioctl_snap_create_transid(file, vol_args->name,
A
Arne Jansen 已提交
1599
					      vol_args->fd, subvol, ptr,
1600
					      readonly, inherit);
1601 1602 1603 1604 1605 1606

	if (ret == 0 && ptr &&
	    copy_to_user(arg +
			 offsetof(struct btrfs_ioctl_vol_args_v2,
				  transid), ptr, sizeof(*ptr)))
		ret = -EFAULT;
1607
out:
C
Christoph Hellwig 已提交
1608
	kfree(vol_args);
A
Arne Jansen 已提交
1609
	kfree(inherit);
C
Christoph Hellwig 已提交
1610 1611 1612
	return ret;
}

1613 1614 1615
static noinline int btrfs_ioctl_subvol_getflags(struct file *file,
						void __user *arg)
{
A
Al Viro 已提交
1616
	struct inode *inode = file_inode(file);
1617 1618 1619 1620
	struct btrfs_root *root = BTRFS_I(inode)->root;
	int ret = 0;
	u64 flags = 0;

L
Li Zefan 已提交
1621
	if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID)
1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637
		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
1638
	struct inode *inode = file_inode(file);
1639 1640 1641 1642 1643 1644
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;
	u64 root_flags;
	u64 flags;
	int ret = 0;

1645 1646 1647
	ret = mnt_want_write_file(file);
	if (ret)
		goto out;
1648

1649 1650 1651 1652
	if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID) {
		ret = -EINVAL;
		goto out_drop_write;
	}
1653

1654 1655 1656 1657
	if (copy_from_user(&flags, arg, sizeof(flags))) {
		ret = -EFAULT;
		goto out_drop_write;
	}
1658

1659 1660 1661 1662
	if (flags & BTRFS_SUBVOL_CREATE_ASYNC) {
		ret = -EINVAL;
		goto out_drop_write;
	}
1663

1664 1665 1666 1667
	if (flags & ~BTRFS_SUBVOL_RDONLY) {
		ret = -EOPNOTSUPP;
		goto out_drop_write;
	}
1668

1669 1670 1671 1672
	if (!inode_owner_or_capable(inode)) {
		ret = -EACCES;
		goto out_drop_write;
	}
1673

1674 1675 1676 1677
	down_write(&root->fs_info->subvol_sem);

	/* nothing to do */
	if (!!(flags & BTRFS_SUBVOL_RDONLY) == btrfs_root_readonly(root))
1678
		goto out_drop_sem;
1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693

	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;
	}

1694
	ret = btrfs_update_root(trans, root->fs_info->tree_root,
1695 1696 1697 1698 1699 1700
				&root->root_key, &root->root_item);

	btrfs_commit_transaction(trans, root);
out_reset:
	if (ret)
		btrfs_set_root_flags(&root->root_item, root_flags);
1701
out_drop_sem:
1702
	up_write(&root->fs_info->subvol_sem);
1703 1704 1705
out_drop_write:
	mnt_drop_write_file(file);
out:
1706 1707 1708
	return ret;
}

1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744
/*
 * helper to check if the subvolume references other subvolumes
 */
static noinline int may_destroy_subvol(struct btrfs_root *root)
{
	struct btrfs_path *path;
	struct btrfs_key key;
	int ret;

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

	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;
}

1745 1746 1747
static noinline int key_in_sk(struct btrfs_key *key,
			      struct btrfs_ioctl_search_key *sk)
{
1748 1749 1750 1751 1752 1753 1754 1755 1756
	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)
1757
		return 0;
1758 1759 1760 1761 1762 1763 1764

	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)
1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 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
		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);

		if (item_len > BTRFS_SEARCH_ARGS_BUFSIZE)
			item_len = 0;

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

		btrfs_item_key_to_cpu(leaf, key, i);
		if (!key_in_sk(key, sk))
			continue;

		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;
		}
1831
		(*num_found)++;
1832 1833 1834 1835 1836

		if (*num_found >= sk->nr_items)
			break;
	}
advance_key:
1837 1838
	ret = 0;
	if (key->offset < (u64)-1 && key->offset < sk->max_offset)
1839
		key->offset++;
1840 1841
	else if (key->type < (u8)-1 && key->type < sk->max_type) {
		key->offset = 0;
1842
		key->type++;
1843 1844 1845
	} else if (key->objectid < (u64)-1 && key->objectid < sk->max_objectid) {
		key->offset = 0;
		key->type = 0;
1846
		key->objectid++;
1847 1848
	} else
		ret = 1;
1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896
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_key max_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;

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

	path->keep_locks = 1;

	while(1) {
1897
		ret = btrfs_search_forward(root, &key, &max_key, path,
1898 1899 1900 1901 1902 1903 1904 1905
					   sk->min_transid);
		if (ret != 0) {
			if (ret > 0)
				ret = 0;
			goto err;
		}
		ret = copy_to_sk(root, path, &key, sk, args->buf,
				 &sk_offset, &num_found);
1906
		btrfs_release_path(path);
1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927
		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
1928 1929 1930
	args = memdup_user(argp, sizeof(*args));
	if (IS_ERR(args))
		return PTR_ERR(args);
1931

A
Al Viro 已提交
1932
	inode = file_inode(file);
1933 1934 1935 1936 1937 1938 1939
	ret = search_ioctl(inode, args);
	if (ret == 0 && copy_to_user(argp, args, sizeof(*args)))
		ret = -EFAULT;
	kfree(args);
	return ret;
}

1940
/*
1941 1942 1943
 * Search INODE_REFs to identify path name of 'dirid' directory
 * in a 'tree_id' tree. and sets path name to 'name'.
 */
1944 1945 1946 1947 1948
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;
1949
	char *ptr;
1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966
	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;

1967
	ptr = &name[BTRFS_INO_LOOKUP_PATH_MAX];
1968 1969 1970 1971 1972 1973 1974

	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);
1975 1976
		ret = -ENOENT;
		goto out;
1977 1978 1979 1980
	}

	key.objectid = dirid;
	key.type = BTRFS_INODE_REF_KEY;
1981
	key.offset = (u64)-1;
1982 1983 1984 1985 1986 1987 1988 1989

	while(1) {
		ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
		if (ret < 0)
			goto out;

		l = path->nodes[0];
		slot = path->slots[0];
1990 1991
		if (ret > 0 && slot > 0)
			slot--;
1992 1993 1994
		btrfs_item_key_to_cpu(l, &key, slot);

		if (ret > 0 && (key.objectid != dirid ||
1995 1996
				key.type != BTRFS_INODE_REF_KEY)) {
			ret = -ENOENT;
1997
			goto out;
1998
		}
1999 2000 2001 2002 2003

		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;
2004
		if (ptr < name)
2005 2006 2007 2008 2009 2010 2011 2012
			goto out;

		*(ptr + len) = '/';
		read_extent_buffer(l, ptr,(unsigned long)(iref + 1), len);

		if (key.offset == BTRFS_FIRST_FREE_OBJECTID)
			break;

2013
		btrfs_release_path(path);
2014
		key.objectid = key.offset;
2015
		key.offset = (u64)-1;
2016 2017
		dirid = key.objectid;
	}
2018
	if (ptr < name)
2019
		goto out;
2020
	memmove(name, ptr, total_len);
2021 2022 2023 2024
	name[total_len]='\0';
	ret = 0;
out:
	btrfs_free_path(path);
2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037
	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
2038 2039 2040
	args = memdup_user(argp, sizeof(*args));
	if (IS_ERR(args))
		return PTR_ERR(args);
2041

A
Al Viro 已提交
2042
	inode = file_inode(file);
2043

2044 2045 2046
	if (args->treeid == 0)
		args->treeid = BTRFS_I(inode)->root->root_key.objectid;

2047 2048 2049 2050 2051 2052 2053 2054
	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);
2055 2056 2057
	return ret;
}

2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068
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;
2069 2070
	struct btrfs_block_rsv block_rsv;
	u64 qgroup_reserved;
2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086
	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;
	}

2087
	err = mnt_want_write_file(file);
2088 2089 2090
	if (err)
		goto out;

2091 2092 2093
	err = mutex_lock_killable_nested(&dir->i_mutex, I_MUTEX_PARENT);
	if (err == -EINTR)
		goto out;
2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105
	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;
2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140
	dest = BTRFS_I(inode)->root;
	if (!capable(CAP_SYS_ADMIN)){
		/*
		 * 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;
	}

2141 2142 2143 2144 2145
	/* 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
2146
	if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID) {
2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161
		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;

2162 2163 2164 2165 2166 2167 2168 2169 2170 2171
	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,
					       5, &qgroup_reserved);
	if (err)
		goto out_up_write;

2172 2173 2174
	trans = btrfs_start_transaction(root, 0);
	if (IS_ERR(trans)) {
		err = PTR_ERR(trans);
2175
		goto out_release;
2176
	}
2177 2178
	trans->block_rsv = &block_rsv;
	trans->bytes_reserved = block_rsv.size;
2179

2180 2181 2182 2183
	ret = btrfs_unlink_subvol(trans, root, dir,
				dest->root_key.objectid,
				dentry->d_name.name,
				dentry->d_name.len);
2184 2185 2186 2187 2188
	if (ret) {
		err = ret;
		btrfs_abort_transaction(trans, root, ret);
		goto out_end_trans;
	}
2189 2190 2191 2192 2193 2194 2195 2196

	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);

2197 2198 2199 2200
	if (!xchg(&dest->orphan_item_inserted, 1)) {
		ret = btrfs_insert_orphan_item(trans,
					root->fs_info->tree_root,
					dest->root_key.objectid);
2201 2202 2203 2204 2205
		if (ret) {
			btrfs_abort_transaction(trans, root, ret);
			err = ret;
			goto out_end_trans;
		}
2206
	}
2207
out_end_trans:
2208 2209
	trans->block_rsv = NULL;
	trans->bytes_reserved = 0;
S
Sage Weil 已提交
2210
	ret = btrfs_end_transaction(trans, root);
2211 2212
	if (ret && !err)
		err = ret;
2213
	inode->i_flags |= S_DEAD;
2214 2215
out_release:
	btrfs_subvolume_release_metadata(root, &block_rsv, qgroup_reserved);
2216 2217 2218 2219 2220
out_up_write:
	up_write(&root->fs_info->subvol_sem);
out_unlock:
	mutex_unlock(&inode->i_mutex);
	if (!err) {
2221
		shrink_dcache_sb(root->fs_info->sb);
2222 2223
		btrfs_invalidate_inodes(dest);
		d_delete(dentry);
2224 2225 2226 2227 2228 2229

		/* the last ref */
		if (dest->cache_inode) {
			iput(dest->cache_inode);
			dest->cache_inode = NULL;
		}
2230 2231 2232 2233 2234
	}
out_dput:
	dput(dentry);
out_unlock_dir:
	mutex_unlock(&dir->i_mutex);
A
Al Viro 已提交
2235
	mnt_drop_write_file(file);
2236 2237 2238 2239 2240
out:
	kfree(vol_args);
	return err;
}

C
Chris Mason 已提交
2241
static int btrfs_ioctl_defrag(struct file *file, void __user *argp)
C
Christoph Hellwig 已提交
2242
{
A
Al Viro 已提交
2243
	struct inode *inode = file_inode(file);
C
Christoph Hellwig 已提交
2244
	struct btrfs_root *root = BTRFS_I(inode)->root;
C
Chris Mason 已提交
2245
	struct btrfs_ioctl_defrag_range_args *range;
Y
Yan Zheng 已提交
2246 2247
	int ret;

2248 2249 2250
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
L
Li Zefan 已提交
2251

2252 2253 2254
	if (btrfs_root_readonly(root)) {
		ret = -EROFS;
		goto out;
2255
	}
C
Christoph Hellwig 已提交
2256 2257 2258

	switch (inode->i_mode & S_IFMT) {
	case S_IFDIR:
2259 2260 2261 2262
		if (!capable(CAP_SYS_ADMIN)) {
			ret = -EPERM;
			goto out;
		}
2263
		ret = btrfs_defrag_root(root);
2264 2265
		if (ret)
			goto out;
2266
		ret = btrfs_defrag_root(root->fs_info->extent_root);
C
Christoph Hellwig 已提交
2267 2268
		break;
	case S_IFREG:
2269 2270 2271 2272
		if (!(file->f_mode & FMODE_WRITE)) {
			ret = -EINVAL;
			goto out;
		}
C
Chris Mason 已提交
2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284

		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);
2285
				goto out;
C
Chris Mason 已提交
2286 2287 2288 2289 2290 2291 2292 2293 2294 2295
			}
			/* 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
2296
		ret = btrfs_defrag_file(file_inode(file), file,
C
Chris Mason 已提交
2297 2298 2299
					range, 0, 0);
		if (ret > 0)
			ret = 0;
C
Chris Mason 已提交
2300
		kfree(range);
C
Christoph Hellwig 已提交
2301
		break;
2302 2303
	default:
		ret = -EINVAL;
C
Christoph Hellwig 已提交
2304
	}
2305
out:
2306
	mnt_drop_write_file(file);
2307
	return ret;
C
Christoph Hellwig 已提交
2308 2309
}

2310
static long btrfs_ioctl_add_dev(struct btrfs_root *root, void __user *arg)
C
Christoph Hellwig 已提交
2311 2312 2313 2314
{
	struct btrfs_ioctl_vol_args *vol_args;
	int ret;

2315 2316 2317
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

2318 2319 2320
	if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
			1)) {
		pr_info("btrfs: dev add/delete/balance/replace/resize operation in progress\n");
2321
		return -EINVAL;
2322 2323
	}

2324
	mutex_lock(&root->fs_info->volume_mutex);
L
Li Zefan 已提交
2325
	vol_args = memdup_user(arg, sizeof(*vol_args));
2326 2327 2328 2329
	if (IS_ERR(vol_args)) {
		ret = PTR_ERR(vol_args);
		goto out;
	}
C
Christoph Hellwig 已提交
2330

2331
	vol_args->name[BTRFS_PATH_NAME_MAX] = '\0';
C
Christoph Hellwig 已提交
2332 2333 2334
	ret = btrfs_init_new_device(root, vol_args->name);

	kfree(vol_args);
2335 2336
out:
	mutex_unlock(&root->fs_info->volume_mutex);
2337
	atomic_set(&root->fs_info->mutually_exclusive_operation_running, 0);
C
Christoph Hellwig 已提交
2338 2339 2340
	return ret;
}

2341
static long btrfs_ioctl_rm_dev(struct file *file, void __user *arg)
C
Christoph Hellwig 已提交
2342
{
A
Al Viro 已提交
2343
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
C
Christoph Hellwig 已提交
2344 2345 2346
	struct btrfs_ioctl_vol_args *vol_args;
	int ret;

2347 2348 2349
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

2350 2351 2352
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
Y
Yan Zheng 已提交
2353

2354 2355 2356
	if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
			1)) {
		pr_info("btrfs: dev add/delete/balance/replace/resize operation in progress\n");
2357
		mnt_drop_write_file(file);
2358
		return -EINVAL;
2359 2360
	}

2361
	mutex_lock(&root->fs_info->volume_mutex);
L
Li Zefan 已提交
2362
	vol_args = memdup_user(arg, sizeof(*vol_args));
2363 2364 2365 2366
	if (IS_ERR(vol_args)) {
		ret = PTR_ERR(vol_args);
		goto out;
	}
C
Christoph Hellwig 已提交
2367

2368
	vol_args->name[BTRFS_PATH_NAME_MAX] = '\0';
C
Christoph Hellwig 已提交
2369 2370 2371
	ret = btrfs_rm_device(root, vol_args->name);

	kfree(vol_args);
2372 2373
out:
	mutex_unlock(&root->fs_info->volume_mutex);
2374
	atomic_set(&root->fs_info->mutually_exclusive_operation_running, 0);
2375
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
2376 2377 2378
	return ret;
}

J
Jan Schmidt 已提交
2379 2380
static long btrfs_ioctl_fs_info(struct btrfs_root *root, void __user *arg)
{
2381
	struct btrfs_ioctl_fs_info_args *fi_args;
J
Jan Schmidt 已提交
2382 2383 2384
	struct btrfs_device *device;
	struct btrfs_device *next;
	struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
2385
	int ret = 0;
J
Jan Schmidt 已提交
2386 2387 2388 2389

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

2390 2391 2392 2393 2394 2395
	fi_args = kzalloc(sizeof(*fi_args), GFP_KERNEL);
	if (!fi_args)
		return -ENOMEM;

	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
2396 2397 2398

	mutex_lock(&fs_devices->device_list_mutex);
	list_for_each_entry_safe(device, next, &fs_devices->devices, dev_list) {
2399 2400
		if (device->devid > fi_args->max_id)
			fi_args->max_id = device->devid;
J
Jan Schmidt 已提交
2401 2402 2403
	}
	mutex_unlock(&fs_devices->device_list_mutex);

2404 2405
	if (copy_to_user(arg, fi_args, sizeof(*fi_args)))
		ret = -EFAULT;
J
Jan Schmidt 已提交
2406

2407 2408
	kfree(fi_args);
	return ret;
J
Jan Schmidt 已提交
2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430
}

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;
	char empty_uuid[BTRFS_UUID_SIZE] = {0};

	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);

	if (memcmp(empty_uuid, di_args->uuid, BTRFS_UUID_SIZE) != 0)
		s_uuid = di_args->uuid;

	mutex_lock(&fs_devices->device_list_mutex);
2431
	dev = btrfs_find_device(root->fs_info, di_args->devid, s_uuid, NULL);
J
Jan Schmidt 已提交
2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442
	mutex_unlock(&fs_devices->device_list_mutex);

	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));
2443
	if (dev->name) {
2444 2445 2446 2447 2448 2449
		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();
2450 2451
		di_args->path[sizeof(di_args->path) - 1] = 0;
	} else {
2452
		di_args->path[0] = '\0';
2453
	}
J
Jan Schmidt 已提交
2454 2455 2456 2457 2458 2459 2460 2461 2462

out:
	if (ret == 0 && copy_to_user(arg, di_args, sizeof(*di_args)))
		ret = -EFAULT;

	kfree(di_args);
	return ret;
}

2463 2464
static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd,
				       u64 off, u64 olen, u64 destoff)
C
Christoph Hellwig 已提交
2465
{
A
Al Viro 已提交
2466
	struct inode *inode = file_inode(file);
C
Christoph Hellwig 已提交
2467
	struct btrfs_root *root = BTRFS_I(inode)->root;
2468
	struct fd src_file;
C
Christoph Hellwig 已提交
2469 2470 2471 2472
	struct inode *src;
	struct btrfs_trans_handle *trans;
	struct btrfs_path *path;
	struct extent_buffer *leaf;
Y
Yan Zheng 已提交
2473 2474
	char *buf;
	struct btrfs_key key;
C
Christoph Hellwig 已提交
2475 2476
	u32 nritems;
	int slot;
Y
Yan Zheng 已提交
2477
	int ret;
2478 2479
	u64 len = olen;
	u64 bs = root->fs_info->sb->s_blocksize;
2480

2481 2482 2483 2484 2485 2486 2487 2488 2489 2490
	/*
	 * 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)?
	 */

2491
	/* the destination must be opened for writing */
2492
	if (!(file->f_mode & FMODE_WRITE) || (file->f_flags & O_APPEND))
2493 2494
		return -EINVAL;

L
Li Zefan 已提交
2495 2496 2497
	if (btrfs_root_readonly(root))
		return -EROFS;

2498
	ret = mnt_want_write_file(file);
Y
Yan Zheng 已提交
2499 2500 2501
	if (ret)
		return ret;

2502 2503
	src_file = fdget(srcfd);
	if (!src_file.file) {
2504 2505 2506
		ret = -EBADF;
		goto out_drop_write;
	}
2507

2508
	ret = -EXDEV;
2509
	if (src_file.file->f_path.mnt != file->f_path.mnt)
2510 2511
		goto out_fput;

A
Al Viro 已提交
2512
	src = file_inode(src_file.file);
C
Christoph Hellwig 已提交
2513

2514 2515 2516 2517
	ret = -EINVAL;
	if (src == inode)
		goto out_fput;

2518
	/* the src must be open for reading */
2519
	if (!(src_file.file->f_mode & FMODE_READ))
2520 2521
		goto out_fput;

2522 2523 2524 2525 2526
	/* 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;

Y
Yan Zheng 已提交
2527 2528 2529 2530
	ret = -EISDIR;
	if (S_ISDIR(src->i_mode) || S_ISDIR(inode->i_mode))
		goto out_fput;

C
Christoph Hellwig 已提交
2531
	ret = -EXDEV;
2532
	if (src->i_sb != inode->i_sb)
Y
Yan Zheng 已提交
2533 2534 2535 2536 2537 2538 2539 2540 2541 2542
		goto out_fput;

	ret = -ENOMEM;
	buf = vmalloc(btrfs_level_size(root, 0));
	if (!buf)
		goto out_fput;

	path = btrfs_alloc_path();
	if (!path) {
		vfree(buf);
C
Christoph Hellwig 已提交
2543
		goto out_fput;
Y
Yan Zheng 已提交
2544 2545
	}
	path->reada = 2;
C
Christoph Hellwig 已提交
2546 2547

	if (inode < src) {
2548 2549
		mutex_lock_nested(&inode->i_mutex, I_MUTEX_PARENT);
		mutex_lock_nested(&src->i_mutex, I_MUTEX_CHILD);
C
Christoph Hellwig 已提交
2550
	} else {
2551 2552
		mutex_lock_nested(&src->i_mutex, I_MUTEX_PARENT);
		mutex_lock_nested(&inode->i_mutex, I_MUTEX_CHILD);
C
Christoph Hellwig 已提交
2553 2554
	}

2555 2556
	/* determine range to clone */
	ret = -EINVAL;
2557
	if (off + len > src->i_size || off + len < off)
C
Christoph Hellwig 已提交
2558
		goto out_unlock;
2559 2560 2561 2562
	if (len == 0)
		olen = len = src->i_size - off;
	/* if we extend to eof, continue to block boundary */
	if (off + len == src->i_size)
2563
		len = ALIGN(src->i_size, bs) - off;
2564 2565

	/* verify the end result is block aligned */
2566 2567
	if (!IS_ALIGNED(off, bs) || !IS_ALIGNED(off + len, bs) ||
	    !IS_ALIGNED(destoff, bs))
2568 2569
		goto out_unlock;

2570 2571 2572 2573 2574 2575
	if (destoff > inode->i_size) {
		ret = btrfs_cont_expand(inode, inode->i_size, destoff);
		if (ret)
			goto out_unlock;
	}

2576 2577 2578 2579
	/* truncate page cache pages from target inode range */
	truncate_inode_pages_range(&inode->i_data, destoff,
				   PAGE_CACHE_ALIGN(destoff + len) - 1);

C
Christoph Hellwig 已提交
2580 2581 2582
	/* do any pending delalloc/csum calc on src, one way or
	   another, and lock file content */
	while (1) {
Z
Zheng Yan 已提交
2583
		struct btrfs_ordered_extent *ordered;
L
Liu Bo 已提交
2584 2585
		lock_extent(&BTRFS_I(src)->io_tree, off, off + len - 1);
		ordered = btrfs_lookup_first_ordered_extent(src, off + len - 1);
2586
		if (!ordered &&
L
Liu Bo 已提交
2587 2588
		    !test_range_bit(&BTRFS_I(src)->io_tree, off, off + len - 1,
				    EXTENT_DELALLOC, 0, NULL))
C
Christoph Hellwig 已提交
2589
			break;
L
Liu Bo 已提交
2590
		unlock_extent(&BTRFS_I(src)->io_tree, off, off + len - 1);
Y
Yan Zheng 已提交
2591 2592
		if (ordered)
			btrfs_put_ordered_extent(ordered);
2593
		btrfs_wait_ordered_range(src, off, len);
C
Christoph Hellwig 已提交
2594 2595
	}

2596
	/* clone data */
L
Li Zefan 已提交
2597
	key.objectid = btrfs_ino(src);
Y
Yan Zheng 已提交
2598 2599
	key.type = BTRFS_EXTENT_DATA_KEY;
	key.offset = 0;
C
Christoph Hellwig 已提交
2600 2601 2602 2603 2604 2605

	while (1) {
		/*
		 * note the key will change type as we walk through the
		 * tree.
		 */
2606 2607
		ret = btrfs_search_slot(NULL, BTRFS_I(src)->root, &key, path,
				0, 0);
C
Christoph Hellwig 已提交
2608 2609 2610
		if (ret < 0)
			goto out;

Y
Yan Zheng 已提交
2611 2612
		nritems = btrfs_header_nritems(path->nodes[0]);
		if (path->slots[0] >= nritems) {
2613
			ret = btrfs_next_leaf(BTRFS_I(src)->root, path);
C
Christoph Hellwig 已提交
2614 2615 2616 2617
			if (ret < 0)
				goto out;
			if (ret > 0)
				break;
Y
Yan Zheng 已提交
2618
			nritems = btrfs_header_nritems(path->nodes[0]);
C
Christoph Hellwig 已提交
2619 2620 2621 2622
		}
		leaf = path->nodes[0];
		slot = path->slots[0];

Y
Yan Zheng 已提交
2623
		btrfs_item_key_to_cpu(leaf, &key, slot);
2624
		if (btrfs_key_type(&key) > BTRFS_EXTENT_DATA_KEY ||
L
Li Zefan 已提交
2625
		    key.objectid != btrfs_ino(src))
C
Christoph Hellwig 已提交
2626 2627
			break;

2628 2629 2630
		if (btrfs_key_type(&key) == BTRFS_EXTENT_DATA_KEY) {
			struct btrfs_file_extent_item *extent;
			int type;
Z
Zheng Yan 已提交
2631 2632
			u32 size;
			struct btrfs_key new_key;
2633 2634 2635
			u64 disko = 0, diskl = 0;
			u64 datao = 0, datal = 0;
			u8 comp;
2636
			u64 endoff;
Z
Zheng Yan 已提交
2637 2638 2639 2640 2641

			size = btrfs_item_size_nr(leaf, slot);
			read_extent_buffer(leaf, buf,
					   btrfs_item_ptr_offset(leaf, slot),
					   size);
2642 2643 2644 2645 2646

			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);
2647 2648
			if (type == BTRFS_FILE_EXTENT_REG ||
			    type == BTRFS_FILE_EXTENT_PREALLOC) {
C
Chris Mason 已提交
2649 2650 2651 2652
				disko = btrfs_file_extent_disk_bytenr(leaf,
								      extent);
				diskl = btrfs_file_extent_disk_num_bytes(leaf,
								 extent);
2653
				datao = btrfs_file_extent_offset(leaf, extent);
C
Chris Mason 已提交
2654 2655
				datal = btrfs_file_extent_num_bytes(leaf,
								    extent);
2656 2657 2658 2659 2660
			} else if (type == BTRFS_FILE_EXTENT_INLINE) {
				/* take upper bound, may be compressed */
				datal = btrfs_file_extent_ram_bytes(leaf,
								    extent);
			}
2661
			btrfs_release_path(path);
Z
Zheng Yan 已提交
2662

2663
			if (key.offset + datal <= off ||
L
Liu Bo 已提交
2664
			    key.offset >= off + len - 1)
2665 2666
				goto next;

Z
Zheng Yan 已提交
2667
			memcpy(&new_key, &key, sizeof(new_key));
L
Li Zefan 已提交
2668
			new_key.objectid = btrfs_ino(inode);
2669 2670 2671 2672
			if (off <= key.offset)
				new_key.offset = key.offset + destoff - off;
			else
				new_key.offset = destoff;
Z
Zheng Yan 已提交
2673

2674 2675 2676 2677 2678 2679
			/*
			 * 1 - adjusting old extent (we may have to split it)
			 * 1 - add new extent
			 * 1 - inode update
			 */
			trans = btrfs_start_transaction(root, 3);
2680 2681 2682 2683 2684
			if (IS_ERR(trans)) {
				ret = PTR_ERR(trans);
				goto out;
			}

2685 2686
			if (type == BTRFS_FILE_EXTENT_REG ||
			    type == BTRFS_FILE_EXTENT_PREALLOC) {
2687 2688 2689 2690 2691 2692 2693 2694 2695 2696
				/*
				 *    a  | --- range to clone ---|  b
				 * | ------------- extent ------------- |
				 */

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

				/* substract range a */
2697 2698 2699 2700 2701
				if (off > key.offset) {
					datao += off - key.offset;
					datal -= off - key.offset;
				}

J
Josef Bacik 已提交
2702
				ret = btrfs_drop_extents(trans, root, inode,
2703 2704
							 new_key.offset,
							 new_key.offset + datal,
2705
							 1);
2706 2707 2708 2709 2710 2711
				if (ret) {
					btrfs_abort_transaction(trans, root,
								ret);
					btrfs_end_transaction(trans, root);
					goto out;
				}
2712

2713 2714
				ret = btrfs_insert_empty_item(trans, root, path,
							      &new_key, size);
2715 2716 2717 2718 2719 2720
				if (ret) {
					btrfs_abort_transaction(trans, root,
								ret);
					btrfs_end_transaction(trans, root);
					goto out;
				}
2721 2722 2723 2724

				leaf = path->nodes[0];
				slot = path->slots[0];
				write_extent_buffer(leaf, buf,
Z
Zheng Yan 已提交
2725 2726
					    btrfs_item_ptr_offset(leaf, slot),
					    size);
Y
Yan Zheng 已提交
2727

2728
				extent = btrfs_item_ptr(leaf, slot,
C
Christoph Hellwig 已提交
2729
						struct btrfs_file_extent_item);
2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740

				/* 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
2741
					ret = btrfs_inc_extent_ref(trans, root,
2742 2743
							disko, diskl, 0,
							root->root_key.objectid,
L
Li Zefan 已提交
2744
							btrfs_ino(inode),
A
Arne Jansen 已提交
2745 2746
							new_key.offset - datao,
							0);
2747 2748 2749 2750 2751 2752 2753 2754 2755
					if (ret) {
						btrfs_abort_transaction(trans,
									root,
									ret);
						btrfs_end_transaction(trans,
								      root);
						goto out;

					}
C
Christoph Hellwig 已提交
2756
				}
2757 2758 2759 2760 2761 2762 2763
			} 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
2764

L
Liu Bo 已提交
2765 2766
				if (key.offset + datal > off + len)
					trim = key.offset + datal - (off + len);
C
Chris Mason 已提交
2767

2768 2769
				if (comp && (skip || trim)) {
					ret = -EINVAL;
2770
					btrfs_end_transaction(trans, root);
2771 2772 2773 2774
					goto out;
				}
				size -= skip + trim;
				datal -= skip + trim;
2775

J
Josef Bacik 已提交
2776
				ret = btrfs_drop_extents(trans, root, inode,
2777 2778
							 new_key.offset,
							 new_key.offset + datal,
2779
							 1);
2780 2781 2782 2783 2784 2785
				if (ret) {
					btrfs_abort_transaction(trans, root,
								ret);
					btrfs_end_transaction(trans, root);
					goto out;
				}
2786

2787 2788
				ret = btrfs_insert_empty_item(trans, root, path,
							      &new_key, size);
2789 2790 2791 2792 2793 2794
				if (ret) {
					btrfs_abort_transaction(trans, root,
								ret);
					btrfs_end_transaction(trans, root);
					goto out;
				}
2795 2796

				if (skip) {
C
Chris Mason 已提交
2797 2798
					u32 start =
					  btrfs_file_extent_calc_inline_size(0);
2799 2800 2801 2802 2803 2804 2805 2806 2807 2808
					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
2809
			}
2810 2811

			btrfs_mark_buffer_dirty(leaf);
2812
			btrfs_release_path(path);
2813

2814
			inode_inc_iversion(inode);
2815
			inode->i_mtime = inode->i_ctime = CURRENT_TIME;
2816 2817 2818 2819 2820 2821 2822

			/*
			 * 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;
2823 2824
			if (endoff > destoff+olen)
				endoff = destoff+olen;
2825 2826 2827
			if (endoff > inode->i_size)
				btrfs_i_size_write(inode, endoff);

2828
			ret = btrfs_update_inode(trans, root, inode);
2829 2830 2831 2832 2833 2834
			if (ret) {
				btrfs_abort_transaction(trans, root, ret);
				btrfs_end_transaction(trans, root);
				goto out;
			}
			ret = btrfs_end_transaction(trans, root);
2835
		}
C
Chris Mason 已提交
2836
next:
2837
		btrfs_release_path(path);
C
Christoph Hellwig 已提交
2838 2839 2840 2841
		key.offset++;
	}
	ret = 0;
out:
2842
	btrfs_release_path(path);
L
Liu Bo 已提交
2843
	unlock_extent(&BTRFS_I(src)->io_tree, off, off + len - 1);
C
Christoph Hellwig 已提交
2844 2845 2846
out_unlock:
	mutex_unlock(&src->i_mutex);
	mutex_unlock(&inode->i_mutex);
Y
Yan Zheng 已提交
2847 2848
	vfree(buf);
	btrfs_free_path(path);
C
Christoph Hellwig 已提交
2849
out_fput:
2850
	fdput(src_file);
2851
out_drop_write:
A
Al Viro 已提交
2852
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
2853 2854 2855
	return ret;
}

2856
static long btrfs_ioctl_clone_range(struct file *file, void __user *argp)
2857 2858 2859
{
	struct btrfs_ioctl_clone_range_args args;

2860
	if (copy_from_user(&args, argp, sizeof(args)))
2861 2862 2863 2864 2865
		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
2866 2867 2868 2869 2870 2871
/*
 * 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.
 */
2872
static long btrfs_ioctl_trans_start(struct file *file)
C
Christoph Hellwig 已提交
2873
{
A
Al Viro 已提交
2874
	struct inode *inode = file_inode(file);
C
Christoph Hellwig 已提交
2875 2876
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;
2877
	int ret;
C
Christoph Hellwig 已提交
2878

2879
	ret = -EPERM;
2880
	if (!capable(CAP_SYS_ADMIN))
2881
		goto out;
2882

2883 2884
	ret = -EINPROGRESS;
	if (file->private_data)
C
Christoph Hellwig 已提交
2885
		goto out;
2886

L
Li Zefan 已提交
2887 2888 2889 2890
	ret = -EROFS;
	if (btrfs_root_readonly(root))
		goto out;

2891
	ret = mnt_want_write_file(file);
Y
Yan Zheng 已提交
2892 2893 2894
	if (ret)
		goto out;

J
Josef Bacik 已提交
2895
	atomic_inc(&root->fs_info->open_ioctl_trans);
2896

2897
	ret = -ENOMEM;
2898
	trans = btrfs_start_ioctl_transaction(root);
2899
	if (IS_ERR(trans))
2900 2901 2902 2903 2904 2905
		goto out_drop;

	file->private_data = trans;
	return 0;

out_drop:
J
Josef Bacik 已提交
2906
	atomic_dec(&root->fs_info->open_ioctl_trans);
A
Al Viro 已提交
2907
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
2908 2909 2910 2911
out:
	return ret;
}

2912 2913
static long btrfs_ioctl_default_subvol(struct file *file, void __user *argp)
{
A
Al Viro 已提交
2914
	struct inode *inode = file_inode(file);
2915 2916 2917 2918 2919 2920 2921 2922 2923
	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;
2924
	int ret;
2925 2926 2927 2928

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

2929 2930 2931 2932 2933 2934 2935 2936
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;

	if (copy_from_user(&objectid, argp, sizeof(objectid))) {
		ret = -EFAULT;
		goto out;
	}
2937 2938 2939 2940 2941 2942 2943 2944 2945

	if (!objectid)
		objectid = root->root_key.objectid;

	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);
2946 2947 2948 2949
	if (IS_ERR(new_root)) {
		ret = PTR_ERR(new_root);
		goto out;
	}
2950

2951 2952 2953 2954
	if (btrfs_root_refs(&new_root->root_item) == 0) {
		ret = -ENOENT;
		goto out;
	}
2955 2956

	path = btrfs_alloc_path();
2957 2958 2959 2960
	if (!path) {
		ret = -ENOMEM;
		goto out;
	}
2961 2962 2963
	path->leave_spinning = 1;

	trans = btrfs_start_transaction(root, 1);
2964
	if (IS_ERR(trans)) {
2965
		btrfs_free_path(path);
2966 2967
		ret = PTR_ERR(trans);
		goto out;
2968 2969
	}

2970
	dir_id = btrfs_super_root_dir(root->fs_info->super_copy);
2971 2972
	di = btrfs_lookup_dir_item(trans, root->fs_info->tree_root, path,
				   dir_id, "default", 7, 1);
2973
	if (IS_ERR_OR_NULL(di)) {
2974 2975 2976 2977
		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");
2978 2979
		ret = -ENOENT;
		goto out;
2980 2981 2982 2983 2984 2985 2986
	}

	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);

2987
	btrfs_set_fs_incompat(root->fs_info, DEFAULT_SUBVOL);
2988
	btrfs_end_transaction(trans, root);
2989 2990 2991
out:
	mnt_drop_write_file(file);
	return ret;
2992 2993
}

2994 2995
void btrfs_get_block_group_info(struct list_head *groups_list,
				struct btrfs_ioctl_space_info *space)
2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009
{
	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);
	}
}

J
Josef Bacik 已提交
3010 3011 3012 3013 3014
long btrfs_ioctl_space_info(struct btrfs_root *root, void __user *arg)
{
	struct btrfs_ioctl_space_args space_args;
	struct btrfs_ioctl_space_info space;
	struct btrfs_ioctl_space_info *dest;
3015
	struct btrfs_ioctl_space_info *dest_orig;
3016
	struct btrfs_ioctl_space_info __user *user_dest;
J
Josef Bacik 已提交
3017
	struct btrfs_space_info *info;
3018 3019 3020 3021 3022
	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;
3023
	int alloc_size;
J
Josef Bacik 已提交
3024
	int ret = 0;
3025
	u64 slot_count = 0;
3026
	int i, c;
J
Josef Bacik 已提交
3027 3028 3029 3030 3031 3032

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

3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056
	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);
	}
3057 3058 3059 3060 3061 3062

	/* space_slots == 0 means they are asking for a count */
	if (space_args.space_slots == 0) {
		space_args.total_spaces = slot_count;
		goto out;
	}
3063

3064
	slot_count = min_t(u64, space_args.space_slots, slot_count);
3065

3066
	alloc_size = sizeof(*dest) * slot_count;
3067

3068 3069 3070 3071 3072 3073
	/* 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
3074
	space_args.total_spaces = 0;
3075 3076 3077 3078
	dest = kmalloc(alloc_size, GFP_NOFS);
	if (!dest)
		return -ENOMEM;
	dest_orig = dest;
J
Josef Bacik 已提交
3079

3080
	/* now we have a buffer to copy into */
3081 3082 3083
	for (i = 0; i < num_types; i++) {
		struct btrfs_space_info *tmp;

3084 3085 3086
		if (!slot_count)
			break;

3087 3088 3089 3090 3091 3092 3093 3094 3095 3096
		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();
3097

3098 3099 3100 3101 3102
		if (!info)
			continue;
		down_read(&info->groups_sem);
		for (c = 0; c < BTRFS_NR_RAID_TYPES; c++) {
			if (!list_empty(&info->block_groups[c])) {
3103 3104
				btrfs_get_block_group_info(
					&info->block_groups[c], &space);
3105 3106 3107
				memcpy(dest, &space, sizeof(space));
				dest++;
				space_args.total_spaces++;
3108
				slot_count--;
3109
			}
3110 3111
			if (!slot_count)
				break;
3112 3113
		}
		up_read(&info->groups_sem);
J
Josef Bacik 已提交
3114 3115
	}

D
Daniel J Blueman 已提交
3116
	user_dest = (struct btrfs_ioctl_space_info __user *)
3117 3118 3119 3120 3121 3122 3123 3124
		(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
3125 3126 3127 3128 3129
		ret = -EFAULT;

	return ret;
}

C
Christoph Hellwig 已提交
3130 3131 3132 3133 3134 3135 3136 3137
/*
 * 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
3138
	struct inode *inode = file_inode(file);
C
Christoph Hellwig 已提交
3139 3140 3141 3142
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;

	trans = file->private_data;
3143 3144
	if (!trans)
		return -EINVAL;
C
Christoph Hellwig 已提交
3145
	file->private_data = NULL;
3146

3147 3148
	btrfs_end_transaction(trans, root);

J
Josef Bacik 已提交
3149
	atomic_dec(&root->fs_info->open_ioctl_trans);
3150

A
Al Viro 已提交
3151
	mnt_drop_write_file(file);
3152
	return 0;
C
Christoph Hellwig 已提交
3153 3154
}

3155 3156
static noinline long btrfs_ioctl_start_sync(struct btrfs_root *root,
					    void __user *argp)
3157 3158 3159
{
	struct btrfs_trans_handle *trans;
	u64 transid;
T
Tsutomu Itoh 已提交
3160
	int ret;
3161

M
Miao Xie 已提交
3162
	trans = btrfs_attach_transaction_barrier(root);
3163 3164 3165 3166 3167 3168 3169 3170
	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;
	}
3171
	transid = trans->transid;
T
Tsutomu Itoh 已提交
3172
	ret = btrfs_commit_transaction_async(trans, root, 0);
3173 3174
	if (ret) {
		btrfs_end_transaction(trans, root);
T
Tsutomu Itoh 已提交
3175
		return ret;
3176
	}
3177
out:
3178 3179 3180 3181 3182 3183
	if (argp)
		if (copy_to_user(argp, &transid, sizeof(transid)))
			return -EFAULT;
	return 0;
}

3184 3185
static noinline long btrfs_ioctl_wait_sync(struct btrfs_root *root,
					   void __user *argp)
3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197
{
	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
3198
static long btrfs_ioctl_scrub(struct file *file, void __user *arg)
J
Jan Schmidt 已提交
3199
{
A
Al Viro 已提交
3200
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
J
Jan Schmidt 已提交
3201
	struct btrfs_ioctl_scrub_args *sa;
M
Miao Xie 已提交
3202
	int ret;
J
Jan Schmidt 已提交
3203 3204 3205 3206 3207 3208 3209 3210

	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
3211 3212 3213 3214 3215 3216
	if (!(sa->flags & BTRFS_SCRUB_READONLY)) {
		ret = mnt_want_write_file(file);
		if (ret)
			goto out;
	}

3217
	ret = btrfs_scrub_dev(root->fs_info, sa->devid, sa->start, sa->end,
3218 3219
			      &sa->progress, sa->flags & BTRFS_SCRUB_READONLY,
			      0);
J
Jan Schmidt 已提交
3220 3221 3222 3223

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

M
Miao Xie 已提交
3224 3225 3226
	if (!(sa->flags & BTRFS_SCRUB_READONLY))
		mnt_drop_write_file(file);
out:
J
Jan Schmidt 已提交
3227 3228 3229 3230 3231 3232 3233 3234 3235
	kfree(sa);
	return ret;
}

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

3236
	return btrfs_scrub_cancel(root->fs_info);
J
Jan Schmidt 已提交
3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260
}

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;
}

3261
static long btrfs_ioctl_get_dev_stats(struct btrfs_root *root,
3262
				      void __user *arg)
3263 3264 3265 3266 3267 3268 3269 3270
{
	struct btrfs_ioctl_get_dev_stats *sa;
	int ret;

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

3271 3272 3273 3274 3275 3276
	if ((sa->flags & BTRFS_DEV_STATS_RESET) && !capable(CAP_SYS_ADMIN)) {
		kfree(sa);
		return -EPERM;
	}

	ret = btrfs_get_dev_stats(root, sa);
3277 3278 3279 3280 3281 3282 3283 3284

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

	kfree(sa);
	return ret;
}

3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329
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:
		if (atomic_xchg(
			&root->fs_info->mutually_exclusive_operation_running,
			1)) {
			pr_info("btrfs: dev add/delete/balance/replace/resize operation in progress\n");
			ret = -EINPROGRESS;
		} 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;

	kfree(p);
	return ret;
}

3330 3331 3332 3333
static long btrfs_ioctl_ino_to_path(struct btrfs_root *root, void __user *arg)
{
	int ret = 0;
	int i;
3334
	u64 rel_ptr;
3335
	int size;
3336
	struct btrfs_ioctl_ino_path_args *ipa = NULL;
3337 3338 3339
	struct inode_fs_paths *ipath = NULL;
	struct btrfs_path *path;

3340
	if (!capable(CAP_DAC_READ_SEARCH))
3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368
		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) {
3369 3370
		rel_ptr = ipath->fspath->val[i] -
			  (u64)(unsigned long)ipath->fspath->val;
3371
		ipath->fspath->val[i] = rel_ptr;
3372 3373
	}

3374 3375
	ret = copy_to_user((void *)(unsigned long)ipa->fspath,
			   (void *)(unsigned long)ipath->fspath, size);
3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433
	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;
	}

3434
	size = min_t(u32, loi->size, 64 * 1024);
3435 3436 3437 3438 3439 3440 3441
	inodes = init_data_container(size);
	if (IS_ERR(inodes)) {
		ret = PTR_ERR(inodes);
		inodes = NULL;
		goto out;
	}

L
Liu Bo 已提交
3442 3443 3444
	ret = iterate_inodes_from_logical(loi->logical, root->fs_info, path,
					  build_ino_list, inodes);
	if (ret == -EINVAL)
3445 3446 3447 3448
		ret = -ENOENT;
	if (ret < 0)
		goto out;

3449 3450
	ret = copy_to_user((void *)(unsigned long)loi->inodes,
			   (void *)(unsigned long)inodes, size);
3451 3452 3453 3454 3455
	if (ret)
		ret = -EFAULT;

out:
	btrfs_free_path(path);
3456
	vfree(inodes);
3457 3458 3459 3460 3461
	kfree(loi);

	return ret;
}

3462
void update_ioctl_balance_args(struct btrfs_fs_info *fs_info, int lock,
3463 3464 3465 3466 3467 3468
			       struct btrfs_ioctl_balance_args *bargs)
{
	struct btrfs_balance_control *bctl = fs_info->balance_ctl;

	bargs->flags = bctl->flags;

3469 3470 3471 3472
	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;
3473 3474
	if (atomic_read(&fs_info->balance_cancel_req))
		bargs->state |= BTRFS_BALANCE_STATE_CANCEL_REQ;
3475

3476 3477 3478
	memcpy(&bargs->data, &bctl->data, sizeof(bargs->data));
	memcpy(&bargs->meta, &bctl->meta, sizeof(bargs->meta));
	memcpy(&bargs->sys, &bctl->sys, sizeof(bargs->sys));
3479 3480 3481 3482 3483 3484 3485 3486

	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));
	}
3487 3488
}

3489
static long btrfs_ioctl_balance(struct file *file, void __user *arg)
3490
{
A
Al Viro 已提交
3491
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
3492 3493 3494
	struct btrfs_fs_info *fs_info = root->fs_info;
	struct btrfs_ioctl_balance_args *bargs;
	struct btrfs_balance_control *bctl;
3495
	bool need_unlock; /* for mut. excl. ops lock */
3496 3497 3498 3499 3500
	int ret;

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

3501
	ret = mnt_want_write_file(file);
3502 3503 3504
	if (ret)
		return ret;

3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 3518
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)
	 */
3519
	mutex_lock(&fs_info->balance_mutex);
3520 3521 3522 3523 3524 3525 3526 3527 3528 3529 3530 3531 3532 3533 3534 3535 3536 3537 3538 3539 3540 3541 3542 3543 3544 3545 3546 3547 3548 3549 3550 3551 3552 3553
	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);
		pr_info("btrfs: dev add/delete/balance/replace/resize operation in progress\n");
		ret = -EINVAL;
		goto out;
	}

locked:
	BUG_ON(!atomic_read(&fs_info->mutually_exclusive_operation_running));
3554 3555 3556 3557 3558

	if (arg) {
		bargs = memdup_user(arg, sizeof(*bargs));
		if (IS_ERR(bargs)) {
			ret = PTR_ERR(bargs);
3559
			goto out_unlock;
3560
		}
3561 3562 3563 3564 3565 3566 3567 3568 3569 3570 3571 3572 3573 3574

		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;
		}
3575 3576 3577 3578
	} else {
		bargs = NULL;
	}

3579
	if (fs_info->balance_ctl) {
3580 3581 3582 3583
		ret = -EINPROGRESS;
		goto out_bargs;
	}

3584 3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596
	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;
3597 3598 3599
	} else {
		/* balance everything - no filters */
		bctl->flags |= BTRFS_BALANCE_TYPE_MASK;
3600 3601
	}

3602
do_balance:
3603
	/*
3604 3605 3606 3607 3608
	 * 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.
3609
	 */
3610 3611 3612 3613
	need_unlock = false;

	ret = btrfs_balance(bctl, bargs);

3614 3615 3616 3617 3618 3619 3620
	if (arg) {
		if (copy_to_user(arg, bargs, sizeof(*bargs)))
			ret = -EFAULT;
	}

out_bargs:
	kfree(bargs);
3621
out_unlock:
3622 3623
	mutex_unlock(&fs_info->balance_mutex);
	mutex_unlock(&fs_info->volume_mutex);
3624 3625 3626
	if (need_unlock)
		atomic_set(&fs_info->mutually_exclusive_operation_running, 0);
out:
3627
	mnt_drop_write_file(file);
3628 3629 3630
	return ret;
}

3631 3632 3633 3634 3635 3636 3637 3638
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);
3639 3640
	case BTRFS_BALANCE_CTL_CANCEL:
		return btrfs_cancel_balance(root->fs_info);
3641 3642 3643 3644 3645
	}

	return -EINVAL;
}

3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 3673 3674 3675 3676 3677 3678
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;
}

3679
static long btrfs_ioctl_quota_ctl(struct file *file, void __user *arg)
A
Arne Jansen 已提交
3680
{
A
Al Viro 已提交
3681
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
A
Arne Jansen 已提交
3682 3683 3684 3685 3686 3687 3688 3689
	struct btrfs_ioctl_quota_ctl_args *sa;
	struct btrfs_trans_handle *trans = NULL;
	int ret;
	int err;

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

3690 3691 3692
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
3693 3694

	sa = memdup_user(arg, sizeof(*sa));
3695 3696 3697 3698
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
3699

3700
	down_write(&root->fs_info->subvol_sem);
A
Arne Jansen 已提交
3701
	if (sa->cmd != BTRFS_QUOTA_CTL_RESCAN) {
3702
		trans = btrfs_start_transaction(root->fs_info->tree_root, 2);
A
Arne Jansen 已提交
3703 3704 3705 3706 3707 3708 3709 3710 3711 3712 3713 3714 3715 3716 3717 3718 3719 3720 3721 3722 3723 3724 3725 3726 3727
		if (IS_ERR(trans)) {
			ret = PTR_ERR(trans);
			goto out;
		}
	}

	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;
	case BTRFS_QUOTA_CTL_RESCAN:
		ret = btrfs_quota_rescan(root->fs_info);
		break;
	default:
		ret = -EINVAL;
		break;
	}

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

	if (trans) {
3728
		err = btrfs_commit_transaction(trans, root->fs_info->tree_root);
A
Arne Jansen 已提交
3729 3730 3731 3732 3733
		if (err && !ret)
			ret = err;
	}
out:
	kfree(sa);
3734
	up_write(&root->fs_info->subvol_sem);
3735 3736
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
3737 3738 3739
	return ret;
}

3740
static long btrfs_ioctl_qgroup_assign(struct file *file, void __user *arg)
A
Arne Jansen 已提交
3741
{
A
Al Viro 已提交
3742
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
A
Arne Jansen 已提交
3743 3744 3745 3746 3747 3748 3749 3750
	struct btrfs_ioctl_qgroup_assign_args *sa;
	struct btrfs_trans_handle *trans;
	int ret;
	int err;

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

3751 3752 3753
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
3754 3755

	sa = memdup_user(arg, sizeof(*sa));
3756 3757 3758 3759
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
3760 3761 3762 3763 3764 3765 3766 3767 3768 3769 3770 3771 3772 3773 3774 3775 3776 3777 3778 3779 3780 3781

	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);
3782 3783
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
3784 3785 3786
	return ret;
}

3787
static long btrfs_ioctl_qgroup_create(struct file *file, void __user *arg)
A
Arne Jansen 已提交
3788
{
A
Al Viro 已提交
3789
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
A
Arne Jansen 已提交
3790 3791 3792 3793 3794 3795 3796 3797
	struct btrfs_ioctl_qgroup_create_args *sa;
	struct btrfs_trans_handle *trans;
	int ret;
	int err;

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

3798 3799 3800
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
3801 3802

	sa = memdup_user(arg, sizeof(*sa));
3803 3804 3805 3806
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
3807

M
Miao Xie 已提交
3808 3809 3810 3811 3812
	if (!sa->qgroupid) {
		ret = -EINVAL;
		goto out;
	}

A
Arne Jansen 已提交
3813 3814 3815 3816 3817 3818 3819 3820 3821 3822 3823 3824 3825 3826 3827 3828 3829 3830 3831 3832
	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);
3833 3834
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
3835 3836 3837
	return ret;
}

3838
static long btrfs_ioctl_qgroup_limit(struct file *file, void __user *arg)
A
Arne Jansen 已提交
3839
{
A
Al Viro 已提交
3840
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
A
Arne Jansen 已提交
3841 3842 3843 3844 3845 3846 3847 3848 3849
	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;

3850 3851 3852
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
3853 3854

	sa = memdup_user(arg, sizeof(*sa));
3855 3856 3857 3858
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
3859 3860 3861 3862 3863 3864 3865 3866 3867 3868 3869 3870 3871 3872 3873 3874 3875 3876 3877 3878 3879 3880

	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);
3881 3882
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
3883 3884 3885
	return ret;
}

3886 3887 3888 3889
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
3890
	struct inode *inode = file_inode(file);
3891 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 3919 3920 3921 3922 3923 3924 3925 3926 3927 3928 3929 3930 3931 3932 3933 3934 3935 3936 3937 3938 3939 3940 3941 3942 3943 3944 3945 3946 3947 3948 3949 3950 3951 3952 3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966
	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;

	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;
	}

	trans = btrfs_start_transaction(root, 1);
	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;

	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);
	root_item->stime.sec = cpu_to_le64(sa->stime.sec);
	root_item->stime.nsec = cpu_to_le32(sa->stime.nsec);
	root_item->rtime.sec = cpu_to_le64(sa->rtime.sec);
	root_item->rtime.nsec = cpu_to_le32(sa->rtime.nsec);

	ret = btrfs_update_root(trans, root->fs_info->tree_root,
				&root->root_key, &root->root_item);
	if (ret < 0) {
		btrfs_end_transaction(trans, root);
		trans = NULL;
		goto out;
	} else {
		ret = btrfs_commit_transaction(trans, root);
		if (ret < 0)
			goto out;
	}

	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;
}

3967 3968 3969 3970 3971 3972 3973 3974 3975 3976 3977 3978 3979 3980 3981 3982 3983 3984 3985
static int btrfs_ioctl_get_fslabel(struct file *file, void __user *arg)
{
	struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;
	const char *label = root->fs_info->super_copy->label;
	size_t len = strnlen(label, BTRFS_LABEL_SIZE);
	int ret;

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

	mutex_lock(&root->fs_info->volume_mutex);
	ret = copy_to_user(arg, label, len);
	mutex_unlock(&root->fs_info->volume_mutex);

	return ret ? -EFAULT : 0;
}

3986 3987 3988 3989 3990 3991 3992 3993 3994 3995 3996 3997 3998 3999 4000 4001 4002 4003 4004 4005 4006 4007 4008 4009 4010 4011 4012 4013 4014 4015 4016 4017 4018 4019 4020 4021 4022 4023 4024 4025
static int btrfs_ioctl_set_fslabel(struct file *file, void __user *arg)
{
	struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;
	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;

	mutex_lock(&root->fs_info->volume_mutex);
	trans = btrfs_start_transaction(root, 0);
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto out_unlock;
	}

	strcpy(super_block->label, label);
	ret = btrfs_end_transaction(trans, root);

out_unlock:
	mutex_unlock(&root->fs_info->volume_mutex);
	mnt_drop_write_file(file);
	return ret;
}

C
Christoph Hellwig 已提交
4026 4027 4028
long btrfs_ioctl(struct file *file, unsigned int
		cmd, unsigned long arg)
{
A
Al Viro 已提交
4029
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
4030
	void __user *argp = (void __user *)arg;
C
Christoph Hellwig 已提交
4031 4032

	switch (cmd) {
4033 4034 4035 4036 4037 4038
	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);
4039 4040
	case FITRIM:
		return btrfs_ioctl_fitrim(file, argp);
C
Christoph Hellwig 已提交
4041
	case BTRFS_IOC_SNAP_CREATE:
4042
		return btrfs_ioctl_snap_create(file, argp, 0);
4043
	case BTRFS_IOC_SNAP_CREATE_V2:
4044
		return btrfs_ioctl_snap_create_v2(file, argp, 0);
4045
	case BTRFS_IOC_SUBVOL_CREATE:
4046
		return btrfs_ioctl_snap_create(file, argp, 1);
A
Arne Jansen 已提交
4047 4048
	case BTRFS_IOC_SUBVOL_CREATE_V2:
		return btrfs_ioctl_snap_create_v2(file, argp, 1);
4049 4050
	case BTRFS_IOC_SNAP_DESTROY:
		return btrfs_ioctl_snap_destroy(file, argp);
4051 4052 4053 4054
	case BTRFS_IOC_SUBVOL_GETFLAGS:
		return btrfs_ioctl_subvol_getflags(file, argp);
	case BTRFS_IOC_SUBVOL_SETFLAGS:
		return btrfs_ioctl_subvol_setflags(file, argp);
4055 4056
	case BTRFS_IOC_DEFAULT_SUBVOL:
		return btrfs_ioctl_default_subvol(file, argp);
C
Christoph Hellwig 已提交
4057
	case BTRFS_IOC_DEFRAG:
C
Chris Mason 已提交
4058 4059 4060
		return btrfs_ioctl_defrag(file, NULL);
	case BTRFS_IOC_DEFRAG_RANGE:
		return btrfs_ioctl_defrag(file, argp);
C
Christoph Hellwig 已提交
4061
	case BTRFS_IOC_RESIZE:
4062
		return btrfs_ioctl_resize(file, argp);
C
Christoph Hellwig 已提交
4063
	case BTRFS_IOC_ADD_DEV:
4064
		return btrfs_ioctl_add_dev(root, argp);
C
Christoph Hellwig 已提交
4065
	case BTRFS_IOC_RM_DEV:
4066
		return btrfs_ioctl_rm_dev(file, argp);
J
Jan Schmidt 已提交
4067 4068 4069 4070
	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
4071
	case BTRFS_IOC_BALANCE:
4072
		return btrfs_ioctl_balance(file, NULL);
C
Christoph Hellwig 已提交
4073
	case BTRFS_IOC_CLONE:
4074 4075
		return btrfs_ioctl_clone(file, arg, 0, 0, 0);
	case BTRFS_IOC_CLONE_RANGE:
4076
		return btrfs_ioctl_clone_range(file, argp);
C
Christoph Hellwig 已提交
4077 4078 4079 4080
	case BTRFS_IOC_TRANS_START:
		return btrfs_ioctl_trans_start(file);
	case BTRFS_IOC_TRANS_END:
		return btrfs_ioctl_trans_end(file);
4081 4082 4083 4084
	case BTRFS_IOC_TREE_SEARCH:
		return btrfs_ioctl_tree_search(file, argp);
	case BTRFS_IOC_INO_LOOKUP:
		return btrfs_ioctl_ino_lookup(file, argp);
4085 4086 4087 4088
	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
4089 4090
	case BTRFS_IOC_SPACE_INFO:
		return btrfs_ioctl_space_info(root, argp);
C
Christoph Hellwig 已提交
4091 4092 4093
	case BTRFS_IOC_SYNC:
		btrfs_sync_fs(file->f_dentry->d_sb, 1);
		return 0;
4094
	case BTRFS_IOC_START_SYNC:
4095
		return btrfs_ioctl_start_sync(root, argp);
4096
	case BTRFS_IOC_WAIT_SYNC:
4097
		return btrfs_ioctl_wait_sync(root, argp);
J
Jan Schmidt 已提交
4098
	case BTRFS_IOC_SCRUB:
M
Miao Xie 已提交
4099
		return btrfs_ioctl_scrub(file, argp);
J
Jan Schmidt 已提交
4100 4101 4102 4103
	case BTRFS_IOC_SCRUB_CANCEL:
		return btrfs_ioctl_scrub_cancel(root, argp);
	case BTRFS_IOC_SCRUB_PROGRESS:
		return btrfs_ioctl_scrub_progress(root, argp);
4104
	case BTRFS_IOC_BALANCE_V2:
4105
		return btrfs_ioctl_balance(file, argp);
4106 4107
	case BTRFS_IOC_BALANCE_CTL:
		return btrfs_ioctl_balance_ctl(root, arg);
4108 4109
	case BTRFS_IOC_BALANCE_PROGRESS:
		return btrfs_ioctl_balance_progress(root, argp);
4110 4111
	case BTRFS_IOC_SET_RECEIVED_SUBVOL:
		return btrfs_ioctl_set_received_subvol(file, argp);
4112 4113
	case BTRFS_IOC_SEND:
		return btrfs_ioctl_send(file, argp);
4114
	case BTRFS_IOC_GET_DEV_STATS:
4115
		return btrfs_ioctl_get_dev_stats(root, argp);
A
Arne Jansen 已提交
4116
	case BTRFS_IOC_QUOTA_CTL:
4117
		return btrfs_ioctl_quota_ctl(file, argp);
A
Arne Jansen 已提交
4118
	case BTRFS_IOC_QGROUP_ASSIGN:
4119
		return btrfs_ioctl_qgroup_assign(file, argp);
A
Arne Jansen 已提交
4120
	case BTRFS_IOC_QGROUP_CREATE:
4121
		return btrfs_ioctl_qgroup_create(file, argp);
A
Arne Jansen 已提交
4122
	case BTRFS_IOC_QGROUP_LIMIT:
4123
		return btrfs_ioctl_qgroup_limit(file, argp);
4124 4125
	case BTRFS_IOC_DEV_REPLACE:
		return btrfs_ioctl_dev_replace(root, argp);
4126 4127
	case BTRFS_IOC_GET_FSLABEL:
		return btrfs_ioctl_get_fslabel(file, argp);
4128 4129
	case BTRFS_IOC_SET_FSLABEL:
		return btrfs_ioctl_set_fslabel(file, argp);
C
Christoph Hellwig 已提交
4130 4131 4132 4133
	}

	return -ENOTTY;
}