ioctl.c 98.3 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
	u64 new_align = ~((u64)128 * 1024 - 1);
	struct page **pages = NULL;

1157 1158 1159 1160 1161
	if (isize == 0)
		return 0;

	if (range->start >= isize)
		return -EINVAL;
1162 1163 1164 1165 1166 1167 1168

	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
1169

1170 1171
	if (extent_thresh == 0)
		extent_thresh = 256 * 1024;
1172

C
Chris Mason 已提交
1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185
	/*
	 * 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;
	}

1186
	pages = kmalloc(sizeof(struct page *) * max_cluster,
C
Chris Mason 已提交
1187 1188 1189 1190 1191 1192 1193
			GFP_NOFS);
	if (!pages) {
		ret = -ENOMEM;
		goto out_ra;
	}

	/* find the last page to defrag */
C
Chris Mason 已提交
1194
	if (range->start + range->len > range->start) {
1195
		last_index = min_t(u64, isize - 1,
C
Chris Mason 已提交
1196 1197
			 range->start + range->len - 1) >> PAGE_CACHE_SHIFT;
	} else {
1198
		last_index = (isize - 1) >> PAGE_CACHE_SHIFT;
C
Chris Mason 已提交
1199 1200
	}

C
Chris Mason 已提交
1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216
	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
1217
		max_to_defrag = last_index + 1;
C
Chris Mason 已提交
1218

L
Li Zefan 已提交
1219 1220 1221 1222 1223 1224 1225
	/*
	 * 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;

1226 1227 1228
	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
1229 1230 1231 1232 1233 1234 1235
		/*
		 * make sure we stop running if someone unmounts
		 * the FS
		 */
		if (!(inode->i_sb->s_flags & MS_ACTIVE))
			break;

1236 1237 1238 1239 1240 1241
		if (btrfs_defrag_cancelled(root->fs_info)) {
			printk(KERN_DEBUG "btrfs: defrag_file cancelled\n");
			ret = -EAGAIN;
			break;
		}

1242
		if (!should_defrag_range(inode, (u64)i << PAGE_CACHE_SHIFT,
L
Li Zefan 已提交
1243
					 extent_thresh, &last_len, &skip,
1244 1245
					 &defrag_end, range->flags &
					 BTRFS_DEFRAG_RANGE_COMPRESS)) {
1246 1247 1248 1249 1250 1251 1252 1253 1254
			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;
		}
1255 1256 1257 1258 1259 1260 1261 1262 1263

		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
1264
		if (range->flags & BTRFS_DEFRAG_RANGE_COMPRESS)
1265
			BTRFS_I(inode)->force_compress = compress_type;
1266

1267 1268 1269 1270 1271 1272
		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;
		}
1273

1274
		mutex_lock(&inode->i_mutex);
1275
		ret = cluster_pages_for_defrag(inode, pages, i, cluster);
1276 1277
		if (ret < 0) {
			mutex_unlock(&inode->i_mutex);
C
Chris Mason 已提交
1278
			goto out_ra;
1279
		}
C
Chris Mason 已提交
1280 1281

		defrag_count += ret;
1282
		balance_dirty_pages_ratelimited(inode->i_mapping);
1283
		mutex_unlock(&inode->i_mutex);
C
Chris Mason 已提交
1284 1285 1286 1287 1288

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

1289 1290 1291
			if (ret > 0)
				i += ret;

C
Chris Mason 已提交
1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302
			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
1303
			}
C
Chris Mason 已提交
1304
		} else {
1305
			if (ret > 0) {
L
Li Zefan 已提交
1306
				i += ret;
1307 1308
				last_len += ret << PAGE_CACHE_SHIFT;
			} else {
L
Li Zefan 已提交
1309
				i++;
1310 1311
				last_len = 0;
			}
C
Christoph Hellwig 已提交
1312 1313 1314
		}
	}

C
Chris Mason 已提交
1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332
	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);
1333
		BTRFS_I(inode)->force_compress = BTRFS_COMPRESS_NONE;
C
Chris Mason 已提交
1334 1335 1336
		mutex_unlock(&inode->i_mutex);
	}

1337
	if (range->compress_type == BTRFS_COMPRESS_LZO) {
1338
		btrfs_set_fs_incompat(root->fs_info, COMPRESS_LZO);
1339 1340
	}

1341
	ret = defrag_count;
1342

C
Chris Mason 已提交
1343 1344 1345 1346
out_ra:
	if (!file)
		kfree(ra);
	kfree(pages);
1347
	return ret;
C
Christoph Hellwig 已提交
1348 1349
}

1350
static noinline int btrfs_ioctl_resize(struct file *file,
1351
					void __user *arg)
C
Christoph Hellwig 已提交
1352 1353 1354 1355
{
	u64 new_size;
	u64 old_size;
	u64 devid = 1;
A
Al Viro 已提交
1356
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
C
Christoph Hellwig 已提交
1357 1358 1359 1360 1361 1362 1363 1364
	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;

1365 1366 1367
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

1368 1369 1370 1371
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;

1372 1373 1374
	if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
			1)) {
		pr_info("btrfs: dev add/delete/balance/replace/resize operation in progress\n");
1375
		mnt_drop_write_file(file);
1376
		return -EINVAL;
1377 1378
	}

1379
	mutex_lock(&root->fs_info->volume_mutex);
L
Li Zefan 已提交
1380
	vol_args = memdup_user(arg, sizeof(*vol_args));
1381 1382 1383 1384
	if (IS_ERR(vol_args)) {
		ret = PTR_ERR(vol_args);
		goto out;
	}
1385 1386

	vol_args->name[BTRFS_PATH_NAME_MAX] = '\0';
C
Christoph Hellwig 已提交
1387 1388 1389 1390 1391 1392 1393 1394 1395

	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);
1396 1397 1398 1399
		if (!devid) {
			ret = -EINVAL;
			goto out_free;
		}
1400
		printk(KERN_INFO "btrfs: resizing devid %llu\n",
1401
		       (unsigned long long)devid);
C
Christoph Hellwig 已提交
1402
	}
M
Miao Xie 已提交
1403

1404
	device = btrfs_find_device(root->fs_info, devid, NULL, NULL);
C
Christoph Hellwig 已提交
1405
	if (!device) {
1406
		printk(KERN_INFO "btrfs: resizer unable to find device %llu\n",
1407
		       (unsigned long long)devid);
1408
		ret = -ENODEV;
1409
		goto out_free;
C
Christoph Hellwig 已提交
1410
	}
M
Miao Xie 已提交
1411 1412

	if (!device->writeable) {
L
Liu Bo 已提交
1413
		printk(KERN_INFO "btrfs: resizer unable to apply on "
M
Miao Xie 已提交
1414
		       "readonly device %llu\n",
1415
		       (unsigned long long)devid);
1416
		ret = -EPERM;
L
Liu Bo 已提交
1417 1418 1419
		goto out_free;
	}

C
Christoph Hellwig 已提交
1420 1421 1422 1423 1424 1425 1426 1427 1428 1429
	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
1430
		new_size = memparse(sizestr, NULL);
C
Christoph Hellwig 已提交
1431 1432
		if (new_size == 0) {
			ret = -EINVAL;
1433
			goto out_free;
C
Christoph Hellwig 已提交
1434 1435 1436
		}
	}

1437
	if (device->is_tgtdev_for_dev_replace) {
1438
		ret = -EPERM;
1439 1440 1441
		goto out_free;
	}

C
Christoph Hellwig 已提交
1442 1443 1444 1445 1446
	old_size = device->total_bytes;

	if (mod < 0) {
		if (new_size > old_size) {
			ret = -EINVAL;
1447
			goto out_free;
C
Christoph Hellwig 已提交
1448 1449 1450 1451 1452 1453 1454 1455
		}
		new_size = old_size - new_size;
	} else if (mod > 0) {
		new_size = old_size + new_size;
	}

	if (new_size < 256 * 1024 * 1024) {
		ret = -EINVAL;
1456
		goto out_free;
C
Christoph Hellwig 已提交
1457 1458 1459
	}
	if (new_size > device->bdev->bd_inode->i_size) {
		ret = -EFBIG;
1460
		goto out_free;
C
Christoph Hellwig 已提交
1461 1462 1463 1464 1465
	}

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

1466 1467 1468
	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
1469 1470

	if (new_size > old_size) {
1471
		trans = btrfs_start_transaction(root, 0);
1472 1473
		if (IS_ERR(trans)) {
			ret = PTR_ERR(trans);
1474
			goto out_free;
1475
		}
C
Christoph Hellwig 已提交
1476 1477
		ret = btrfs_grow_device(trans, device, new_size);
		btrfs_commit_transaction(trans, root);
1478
	} else if (new_size < old_size) {
C
Christoph Hellwig 已提交
1479
		ret = btrfs_shrink_device(device, new_size);
1480
	} /* equal, nothing need to do */
C
Christoph Hellwig 已提交
1481

1482
out_free:
C
Christoph Hellwig 已提交
1483
	kfree(vol_args);
1484 1485
out:
	mutex_unlock(&root->fs_info->volume_mutex);
1486
	atomic_set(&root->fs_info->mutually_exclusive_operation_running, 0);
1487
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
1488 1489 1490
	return ret;
}

S
Sage Weil 已提交
1491
static noinline int btrfs_ioctl_snap_create_transid(struct file *file,
A
Arne Jansen 已提交
1492 1493
				char *name, unsigned long fd, int subvol,
				u64 *transid, bool readonly,
1494
				struct btrfs_qgroup_inherit *inherit)
C
Christoph Hellwig 已提交
1495 1496
{
	int namelen;
1497
	int ret = 0;
C
Christoph Hellwig 已提交
1498

1499 1500 1501 1502
	ret = mnt_want_write_file(file);
	if (ret)
		goto out;

S
Sage Weil 已提交
1503 1504
	namelen = strlen(name);
	if (strchr(name, '/')) {
C
Christoph Hellwig 已提交
1505
		ret = -EINVAL;
1506
		goto out_drop_write;
C
Christoph Hellwig 已提交
1507 1508
	}

1509 1510 1511
	if (name[0] == '.' &&
	   (namelen == 1 || (name[1] == '.' && namelen == 2))) {
		ret = -EEXIST;
1512
		goto out_drop_write;
1513 1514
	}

1515
	if (subvol) {
S
Sage Weil 已提交
1516
		ret = btrfs_mksubvol(&file->f_path, name, namelen,
A
Arne Jansen 已提交
1517
				     NULL, transid, readonly, inherit);
1518
	} else {
1519
		struct fd src = fdget(fd);
1520
		struct inode *src_inode;
1521
		if (!src.file) {
1522
			ret = -EINVAL;
1523
			goto out_drop_write;
1524 1525
		}

A
Al Viro 已提交
1526 1527
		src_inode = file_inode(src.file);
		if (src_inode->i_sb != file_inode(file)->i_sb) {
C
Chris Mason 已提交
1528 1529
			printk(KERN_INFO "btrfs: Snapshot src from "
			       "another FS\n");
1530
			ret = -EINVAL;
1531 1532 1533 1534
		} else {
			ret = btrfs_mksubvol(&file->f_path, name, namelen,
					     BTRFS_I(src_inode)->root,
					     transid, readonly, inherit);
1535
		}
1536
		fdput(src);
1537
	}
1538 1539
out_drop_write:
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
1540
out:
S
Sage Weil 已提交
1541 1542 1543 1544
	return ret;
}

static noinline int btrfs_ioctl_snap_create(struct file *file,
1545
					    void __user *arg, int subvol)
S
Sage Weil 已提交
1546
{
1547
	struct btrfs_ioctl_vol_args *vol_args;
S
Sage Weil 已提交
1548 1549
	int ret;

1550 1551 1552 1553
	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
1554

1555
	ret = btrfs_ioctl_snap_create_transid(file, vol_args->name,
L
Li Zefan 已提交
1556
					      vol_args->fd, subvol,
A
Arne Jansen 已提交
1557
					      NULL, false, NULL);
1558

1559 1560 1561
	kfree(vol_args);
	return ret;
}
1562

1563 1564 1565 1566 1567 1568 1569
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
1570
	bool readonly = false;
A
Arne Jansen 已提交
1571
	struct btrfs_qgroup_inherit *inherit = NULL;
1572

1573 1574 1575 1576
	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';
1577

L
Li Zefan 已提交
1578
	if (vol_args->flags &
A
Arne Jansen 已提交
1579 1580
	    ~(BTRFS_SUBVOL_CREATE_ASYNC | BTRFS_SUBVOL_RDONLY |
	      BTRFS_SUBVOL_QGROUP_INHERIT)) {
L
Li Zefan 已提交
1581
		ret = -EOPNOTSUPP;
1582
		goto out;
S
Sage Weil 已提交
1583
	}
1584 1585 1586

	if (vol_args->flags & BTRFS_SUBVOL_CREATE_ASYNC)
		ptr = &transid;
L
Li Zefan 已提交
1587 1588
	if (vol_args->flags & BTRFS_SUBVOL_RDONLY)
		readonly = true;
A
Arne Jansen 已提交
1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599
	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;
		}
	}
1600 1601

	ret = btrfs_ioctl_snap_create_transid(file, vol_args->name,
A
Arne Jansen 已提交
1602
					      vol_args->fd, subvol, ptr,
1603
					      readonly, inherit);
1604 1605 1606 1607 1608 1609

	if (ret == 0 && ptr &&
	    copy_to_user(arg +
			 offsetof(struct btrfs_ioctl_vol_args_v2,
				  transid), ptr, sizeof(*ptr)))
		ret = -EFAULT;
1610
out:
C
Christoph Hellwig 已提交
1611
	kfree(vol_args);
A
Arne Jansen 已提交
1612
	kfree(inherit);
C
Christoph Hellwig 已提交
1613 1614 1615
	return ret;
}

1616 1617 1618
static noinline int btrfs_ioctl_subvol_getflags(struct file *file,
						void __user *arg)
{
A
Al Viro 已提交
1619
	struct inode *inode = file_inode(file);
1620 1621 1622 1623
	struct btrfs_root *root = BTRFS_I(inode)->root;
	int ret = 0;
	u64 flags = 0;

L
Li Zefan 已提交
1624
	if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID)
1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640
		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
1641
	struct inode *inode = file_inode(file);
1642 1643 1644 1645 1646 1647
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;
	u64 root_flags;
	u64 flags;
	int ret = 0;

1648 1649 1650
	ret = mnt_want_write_file(file);
	if (ret)
		goto out;
1651

1652 1653 1654 1655
	if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID) {
		ret = -EINVAL;
		goto out_drop_write;
	}
1656

1657 1658 1659 1660
	if (copy_from_user(&flags, arg, sizeof(flags))) {
		ret = -EFAULT;
		goto out_drop_write;
	}
1661

1662 1663 1664 1665
	if (flags & BTRFS_SUBVOL_CREATE_ASYNC) {
		ret = -EINVAL;
		goto out_drop_write;
	}
1666

1667 1668 1669 1670
	if (flags & ~BTRFS_SUBVOL_RDONLY) {
		ret = -EOPNOTSUPP;
		goto out_drop_write;
	}
1671

1672 1673 1674 1675
	if (!inode_owner_or_capable(inode)) {
		ret = -EACCES;
		goto out_drop_write;
	}
1676

1677 1678 1679 1680
	down_write(&root->fs_info->subvol_sem);

	/* nothing to do */
	if (!!(flags & BTRFS_SUBVOL_RDONLY) == btrfs_root_readonly(root))
1681
		goto out_drop_sem;
1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696

	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;
	}

1697
	ret = btrfs_update_root(trans, root->fs_info->tree_root,
1698 1699 1700 1701 1702 1703
				&root->root_key, &root->root_item);

	btrfs_commit_transaction(trans, root);
out_reset:
	if (ret)
		btrfs_set_root_flags(&root->root_item, root_flags);
1704
out_drop_sem:
1705
	up_write(&root->fs_info->subvol_sem);
1706 1707 1708
out_drop_write:
	mnt_drop_write_file(file);
out:
1709 1710 1711
	return ret;
}

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 1745 1746 1747
/*
 * 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;
}

1748 1749 1750
static noinline int key_in_sk(struct btrfs_key *key,
			      struct btrfs_ioctl_search_key *sk)
{
1751 1752 1753 1754 1755 1756 1757 1758 1759
	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)
1760
		return 0;
1761 1762 1763 1764 1765 1766 1767

	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)
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 1831 1832 1833
		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;
		}
1834
		(*num_found)++;
1835 1836 1837 1838 1839

		if (*num_found >= sk->nr_items)
			break;
	}
advance_key:
1840 1841
	ret = 0;
	if (key->offset < (u64)-1 && key->offset < sk->max_offset)
1842
		key->offset++;
1843 1844
	else if (key->type < (u8)-1 && key->type < sk->max_type) {
		key->offset = 0;
1845
		key->type++;
1846 1847 1848
	} else if (key->objectid < (u64)-1 && key->objectid < sk->max_objectid) {
		key->offset = 0;
		key->type = 0;
1849
		key->objectid++;
1850 1851
	} else
		ret = 1;
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 1897 1898 1899
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) {
1900
		ret = btrfs_search_forward(root, &key, &max_key, path,
1901 1902 1903 1904 1905 1906 1907 1908
					   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);
1909
		btrfs_release_path(path);
1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930
		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
1931 1932 1933
	args = memdup_user(argp, sizeof(*args));
	if (IS_ERR(args))
		return PTR_ERR(args);
1934

A
Al Viro 已提交
1935
	inode = file_inode(file);
1936 1937 1938 1939 1940 1941 1942
	ret = search_ioctl(inode, args);
	if (ret == 0 && copy_to_user(argp, args, sizeof(*args)))
		ret = -EFAULT;
	kfree(args);
	return ret;
}

1943
/*
1944 1945 1946
 * Search INODE_REFs to identify path name of 'dirid' directory
 * in a 'tree_id' tree. and sets path name to 'name'.
 */
1947 1948 1949 1950 1951
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;
1952
	char *ptr;
1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969
	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;

1970
	ptr = &name[BTRFS_INO_LOOKUP_PATH_MAX];
1971 1972 1973 1974 1975 1976 1977

	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);
1978 1979
		ret = -ENOENT;
		goto out;
1980 1981 1982 1983
	}

	key.objectid = dirid;
	key.type = BTRFS_INODE_REF_KEY;
1984
	key.offset = (u64)-1;
1985 1986 1987 1988 1989 1990 1991 1992

	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];
1993 1994
		if (ret > 0 && slot > 0)
			slot--;
1995 1996 1997
		btrfs_item_key_to_cpu(l, &key, slot);

		if (ret > 0 && (key.objectid != dirid ||
1998 1999
				key.type != BTRFS_INODE_REF_KEY)) {
			ret = -ENOENT;
2000
			goto out;
2001
		}
2002 2003 2004 2005 2006

		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;
2007
		if (ptr < name)
2008 2009 2010 2011 2012 2013 2014 2015
			goto out;

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

		if (key.offset == BTRFS_FIRST_FREE_OBJECTID)
			break;

2016
		btrfs_release_path(path);
2017
		key.objectid = key.offset;
2018
		key.offset = (u64)-1;
2019 2020
		dirid = key.objectid;
	}
2021
	if (ptr < name)
2022
		goto out;
2023
	memmove(name, ptr, total_len);
2024 2025 2026 2027
	name[total_len]='\0';
	ret = 0;
out:
	btrfs_free_path(path);
2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040
	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
2041 2042 2043
	args = memdup_user(argp, sizeof(*args));
	if (IS_ERR(args))
		return PTR_ERR(args);
2044

A
Al Viro 已提交
2045
	inode = file_inode(file);
2046

2047 2048 2049
	if (args->treeid == 0)
		args->treeid = BTRFS_I(inode)->root->root_key.objectid;

2050 2051 2052 2053 2054 2055 2056 2057
	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);
2058 2059 2060
	return ret;
}

2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071
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;
2072 2073
	struct btrfs_block_rsv block_rsv;
	u64 qgroup_reserved;
2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089
	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;
	}

2090
	err = mnt_want_write_file(file);
2091 2092 2093
	if (err)
		goto out;

2094 2095 2096
	err = mutex_lock_killable_nested(&dir->i_mutex, I_MUTEX_PARENT);
	if (err == -EINTR)
		goto out;
2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108
	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;
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 2141 2142 2143
	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;
	}

2144 2145 2146 2147 2148
	/* 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
2149
	if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID) {
2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164
		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;

2165 2166 2167 2168 2169 2170 2171 2172 2173 2174
	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;

2175 2176 2177
	trans = btrfs_start_transaction(root, 0);
	if (IS_ERR(trans)) {
		err = PTR_ERR(trans);
2178
		goto out_release;
2179
	}
2180 2181
	trans->block_rsv = &block_rsv;
	trans->bytes_reserved = block_rsv.size;
2182

2183 2184 2185 2186
	ret = btrfs_unlink_subvol(trans, root, dir,
				dest->root_key.objectid,
				dentry->d_name.name,
				dentry->d_name.len);
2187 2188 2189 2190 2191
	if (ret) {
		err = ret;
		btrfs_abort_transaction(trans, root, ret);
		goto out_end_trans;
	}
2192 2193 2194 2195 2196 2197 2198 2199

	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);

2200 2201 2202 2203
	if (!xchg(&dest->orphan_item_inserted, 1)) {
		ret = btrfs_insert_orphan_item(trans,
					root->fs_info->tree_root,
					dest->root_key.objectid);
2204 2205 2206 2207 2208
		if (ret) {
			btrfs_abort_transaction(trans, root, ret);
			err = ret;
			goto out_end_trans;
		}
2209
	}
2210
out_end_trans:
2211 2212
	trans->block_rsv = NULL;
	trans->bytes_reserved = 0;
S
Sage Weil 已提交
2213
	ret = btrfs_end_transaction(trans, root);
2214 2215
	if (ret && !err)
		err = ret;
2216
	inode->i_flags |= S_DEAD;
2217 2218
out_release:
	btrfs_subvolume_release_metadata(root, &block_rsv, qgroup_reserved);
2219 2220 2221 2222 2223
out_up_write:
	up_write(&root->fs_info->subvol_sem);
out_unlock:
	mutex_unlock(&inode->i_mutex);
	if (!err) {
2224
		shrink_dcache_sb(root->fs_info->sb);
2225 2226
		btrfs_invalidate_inodes(dest);
		d_delete(dentry);
2227 2228 2229 2230 2231 2232

		/* the last ref */
		if (dest->cache_inode) {
			iput(dest->cache_inode);
			dest->cache_inode = NULL;
		}
2233 2234 2235 2236 2237
	}
out_dput:
	dput(dentry);
out_unlock_dir:
	mutex_unlock(&dir->i_mutex);
A
Al Viro 已提交
2238
	mnt_drop_write_file(file);
2239 2240 2241 2242 2243
out:
	kfree(vol_args);
	return err;
}

C
Chris Mason 已提交
2244
static int btrfs_ioctl_defrag(struct file *file, void __user *argp)
C
Christoph Hellwig 已提交
2245
{
A
Al Viro 已提交
2246
	struct inode *inode = file_inode(file);
C
Christoph Hellwig 已提交
2247
	struct btrfs_root *root = BTRFS_I(inode)->root;
C
Chris Mason 已提交
2248
	struct btrfs_ioctl_defrag_range_args *range;
Y
Yan Zheng 已提交
2249 2250
	int ret;

2251 2252 2253
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
L
Li Zefan 已提交
2254

2255 2256 2257
	if (btrfs_root_readonly(root)) {
		ret = -EROFS;
		goto out;
2258
	}
C
Christoph Hellwig 已提交
2259 2260 2261

	switch (inode->i_mode & S_IFMT) {
	case S_IFDIR:
2262 2263 2264 2265
		if (!capable(CAP_SYS_ADMIN)) {
			ret = -EPERM;
			goto out;
		}
2266
		ret = btrfs_defrag_root(root);
2267 2268
		if (ret)
			goto out;
2269
		ret = btrfs_defrag_root(root->fs_info->extent_root);
C
Christoph Hellwig 已提交
2270 2271
		break;
	case S_IFREG:
2272 2273 2274 2275
		if (!(file->f_mode & FMODE_WRITE)) {
			ret = -EINVAL;
			goto out;
		}
C
Chris Mason 已提交
2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287

		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);
2288
				goto out;
C
Chris Mason 已提交
2289 2290 2291 2292 2293 2294 2295 2296 2297 2298
			}
			/* 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
2299
		ret = btrfs_defrag_file(file_inode(file), file,
C
Chris Mason 已提交
2300 2301 2302
					range, 0, 0);
		if (ret > 0)
			ret = 0;
C
Chris Mason 已提交
2303
		kfree(range);
C
Christoph Hellwig 已提交
2304
		break;
2305 2306
	default:
		ret = -EINVAL;
C
Christoph Hellwig 已提交
2307
	}
2308
out:
2309
	mnt_drop_write_file(file);
2310
	return ret;
C
Christoph Hellwig 已提交
2311 2312
}

2313
static long btrfs_ioctl_add_dev(struct btrfs_root *root, void __user *arg)
C
Christoph Hellwig 已提交
2314 2315 2316 2317
{
	struct btrfs_ioctl_vol_args *vol_args;
	int ret;

2318 2319 2320
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

2321 2322 2323
	if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
			1)) {
		pr_info("btrfs: dev add/delete/balance/replace/resize operation in progress\n");
2324
		return -EINVAL;
2325 2326
	}

2327
	mutex_lock(&root->fs_info->volume_mutex);
L
Li Zefan 已提交
2328
	vol_args = memdup_user(arg, sizeof(*vol_args));
2329 2330 2331 2332
	if (IS_ERR(vol_args)) {
		ret = PTR_ERR(vol_args);
		goto out;
	}
C
Christoph Hellwig 已提交
2333

2334
	vol_args->name[BTRFS_PATH_NAME_MAX] = '\0';
C
Christoph Hellwig 已提交
2335 2336 2337
	ret = btrfs_init_new_device(root, vol_args->name);

	kfree(vol_args);
2338 2339
out:
	mutex_unlock(&root->fs_info->volume_mutex);
2340
	atomic_set(&root->fs_info->mutually_exclusive_operation_running, 0);
C
Christoph Hellwig 已提交
2341 2342 2343
	return ret;
}

2344
static long btrfs_ioctl_rm_dev(struct file *file, void __user *arg)
C
Christoph Hellwig 已提交
2345
{
A
Al Viro 已提交
2346
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
C
Christoph Hellwig 已提交
2347 2348 2349
	struct btrfs_ioctl_vol_args *vol_args;
	int ret;

2350 2351 2352
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

2353 2354 2355
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
Y
Yan Zheng 已提交
2356

2357 2358 2359
	if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
			1)) {
		pr_info("btrfs: dev add/delete/balance/replace/resize operation in progress\n");
2360
		mnt_drop_write_file(file);
2361
		return -EINVAL;
2362 2363
	}

2364
	mutex_lock(&root->fs_info->volume_mutex);
L
Li Zefan 已提交
2365
	vol_args = memdup_user(arg, sizeof(*vol_args));
2366 2367 2368 2369
	if (IS_ERR(vol_args)) {
		ret = PTR_ERR(vol_args);
		goto out;
	}
C
Christoph Hellwig 已提交
2370

2371
	vol_args->name[BTRFS_PATH_NAME_MAX] = '\0';
C
Christoph Hellwig 已提交
2372 2373 2374
	ret = btrfs_rm_device(root, vol_args->name);

	kfree(vol_args);
2375 2376
out:
	mutex_unlock(&root->fs_info->volume_mutex);
2377
	atomic_set(&root->fs_info->mutually_exclusive_operation_running, 0);
2378
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
2379 2380 2381
	return ret;
}

J
Jan Schmidt 已提交
2382 2383
static long btrfs_ioctl_fs_info(struct btrfs_root *root, void __user *arg)
{
2384
	struct btrfs_ioctl_fs_info_args *fi_args;
J
Jan Schmidt 已提交
2385 2386 2387
	struct btrfs_device *device;
	struct btrfs_device *next;
	struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
2388
	int ret = 0;
J
Jan Schmidt 已提交
2389 2390 2391 2392

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

2393 2394 2395 2396 2397 2398
	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
2399 2400 2401

	mutex_lock(&fs_devices->device_list_mutex);
	list_for_each_entry_safe(device, next, &fs_devices->devices, dev_list) {
2402 2403
		if (device->devid > fi_args->max_id)
			fi_args->max_id = device->devid;
J
Jan Schmidt 已提交
2404 2405 2406
	}
	mutex_unlock(&fs_devices->device_list_mutex);

2407 2408
	if (copy_to_user(arg, fi_args, sizeof(*fi_args)))
		ret = -EFAULT;
J
Jan Schmidt 已提交
2409

2410 2411
	kfree(fi_args);
	return ret;
J
Jan Schmidt 已提交
2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433
}

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);
2434
	dev = btrfs_find_device(root->fs_info, di_args->devid, s_uuid, NULL);
J
Jan Schmidt 已提交
2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445
	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));
2446
	if (dev->name) {
2447 2448 2449 2450 2451 2452
		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();
2453 2454
		di_args->path[sizeof(di_args->path) - 1] = 0;
	} else {
2455
		di_args->path[0] = '\0';
2456
	}
J
Jan Schmidt 已提交
2457 2458 2459 2460 2461 2462 2463 2464 2465

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

	kfree(di_args);
	return ret;
}

2466 2467
static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd,
				       u64 off, u64 olen, u64 destoff)
C
Christoph Hellwig 已提交
2468
{
A
Al Viro 已提交
2469
	struct inode *inode = file_inode(file);
C
Christoph Hellwig 已提交
2470
	struct btrfs_root *root = BTRFS_I(inode)->root;
2471
	struct fd src_file;
C
Christoph Hellwig 已提交
2472 2473 2474 2475
	struct inode *src;
	struct btrfs_trans_handle *trans;
	struct btrfs_path *path;
	struct extent_buffer *leaf;
Y
Yan Zheng 已提交
2476 2477
	char *buf;
	struct btrfs_key key;
C
Christoph Hellwig 已提交
2478 2479
	u32 nritems;
	int slot;
Y
Yan Zheng 已提交
2480
	int ret;
2481 2482
	u64 len = olen;
	u64 bs = root->fs_info->sb->s_blocksize;
2483

2484 2485 2486 2487 2488 2489 2490 2491 2492 2493
	/*
	 * 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)?
	 */

2494
	/* the destination must be opened for writing */
2495
	if (!(file->f_mode & FMODE_WRITE) || (file->f_flags & O_APPEND))
2496 2497
		return -EINVAL;

L
Li Zefan 已提交
2498 2499 2500
	if (btrfs_root_readonly(root))
		return -EROFS;

2501
	ret = mnt_want_write_file(file);
Y
Yan Zheng 已提交
2502 2503 2504
	if (ret)
		return ret;

2505 2506
	src_file = fdget(srcfd);
	if (!src_file.file) {
2507 2508 2509
		ret = -EBADF;
		goto out_drop_write;
	}
2510

2511
	ret = -EXDEV;
2512
	if (src_file.file->f_path.mnt != file->f_path.mnt)
2513 2514
		goto out_fput;

A
Al Viro 已提交
2515
	src = file_inode(src_file.file);
C
Christoph Hellwig 已提交
2516

2517 2518 2519 2520
	ret = -EINVAL;
	if (src == inode)
		goto out_fput;

2521
	/* the src must be open for reading */
2522
	if (!(src_file.file->f_mode & FMODE_READ))
2523 2524
		goto out_fput;

2525 2526 2527 2528 2529
	/* 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
2530 2531 2532 2533
	ret = -EISDIR;
	if (S_ISDIR(src->i_mode) || S_ISDIR(inode->i_mode))
		goto out_fput;

C
Christoph Hellwig 已提交
2534
	ret = -EXDEV;
2535
	if (src->i_sb != inode->i_sb)
Y
Yan Zheng 已提交
2536 2537 2538 2539 2540 2541 2542 2543 2544 2545
		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
2546
		goto out_fput;
Y
Yan Zheng 已提交
2547 2548
	}
	path->reada = 2;
C
Christoph Hellwig 已提交
2549 2550

	if (inode < src) {
2551 2552
		mutex_lock_nested(&inode->i_mutex, I_MUTEX_PARENT);
		mutex_lock_nested(&src->i_mutex, I_MUTEX_CHILD);
C
Christoph Hellwig 已提交
2553
	} else {
2554 2555
		mutex_lock_nested(&src->i_mutex, I_MUTEX_PARENT);
		mutex_lock_nested(&inode->i_mutex, I_MUTEX_CHILD);
C
Christoph Hellwig 已提交
2556 2557
	}

2558 2559
	/* determine range to clone */
	ret = -EINVAL;
2560
	if (off + len > src->i_size || off + len < off)
C
Christoph Hellwig 已提交
2561
		goto out_unlock;
2562 2563 2564 2565
	if (len == 0)
		olen = len = src->i_size - off;
	/* if we extend to eof, continue to block boundary */
	if (off + len == src->i_size)
2566
		len = ALIGN(src->i_size, bs) - off;
2567 2568

	/* verify the end result is block aligned */
2569 2570
	if (!IS_ALIGNED(off, bs) || !IS_ALIGNED(off + len, bs) ||
	    !IS_ALIGNED(destoff, bs))
2571 2572
		goto out_unlock;

2573 2574 2575 2576 2577 2578
	if (destoff > inode->i_size) {
		ret = btrfs_cont_expand(inode, inode->i_size, destoff);
		if (ret)
			goto out_unlock;
	}

2579 2580 2581 2582
	/* 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
2583 2584 2585
	/* do any pending delalloc/csum calc on src, one way or
	   another, and lock file content */
	while (1) {
Z
Zheng Yan 已提交
2586
		struct btrfs_ordered_extent *ordered;
L
Liu Bo 已提交
2587 2588
		lock_extent(&BTRFS_I(src)->io_tree, off, off + len - 1);
		ordered = btrfs_lookup_first_ordered_extent(src, off + len - 1);
2589
		if (!ordered &&
L
Liu Bo 已提交
2590 2591
		    !test_range_bit(&BTRFS_I(src)->io_tree, off, off + len - 1,
				    EXTENT_DELALLOC, 0, NULL))
C
Christoph Hellwig 已提交
2592
			break;
L
Liu Bo 已提交
2593
		unlock_extent(&BTRFS_I(src)->io_tree, off, off + len - 1);
Y
Yan Zheng 已提交
2594 2595
		if (ordered)
			btrfs_put_ordered_extent(ordered);
2596
		btrfs_wait_ordered_range(src, off, len);
C
Christoph Hellwig 已提交
2597 2598
	}

2599
	/* clone data */
L
Li Zefan 已提交
2600
	key.objectid = btrfs_ino(src);
Y
Yan Zheng 已提交
2601 2602
	key.type = BTRFS_EXTENT_DATA_KEY;
	key.offset = 0;
C
Christoph Hellwig 已提交
2603 2604 2605 2606 2607 2608

	while (1) {
		/*
		 * note the key will change type as we walk through the
		 * tree.
		 */
2609 2610
		ret = btrfs_search_slot(NULL, BTRFS_I(src)->root, &key, path,
				0, 0);
C
Christoph Hellwig 已提交
2611 2612 2613
		if (ret < 0)
			goto out;

Y
Yan Zheng 已提交
2614 2615
		nritems = btrfs_header_nritems(path->nodes[0]);
		if (path->slots[0] >= nritems) {
2616
			ret = btrfs_next_leaf(BTRFS_I(src)->root, path);
C
Christoph Hellwig 已提交
2617 2618 2619 2620
			if (ret < 0)
				goto out;
			if (ret > 0)
				break;
Y
Yan Zheng 已提交
2621
			nritems = btrfs_header_nritems(path->nodes[0]);
C
Christoph Hellwig 已提交
2622 2623 2624 2625
		}
		leaf = path->nodes[0];
		slot = path->slots[0];

Y
Yan Zheng 已提交
2626
		btrfs_item_key_to_cpu(leaf, &key, slot);
2627
		if (btrfs_key_type(&key) > BTRFS_EXTENT_DATA_KEY ||
L
Li Zefan 已提交
2628
		    key.objectid != btrfs_ino(src))
C
Christoph Hellwig 已提交
2629 2630
			break;

2631 2632 2633
		if (btrfs_key_type(&key) == BTRFS_EXTENT_DATA_KEY) {
			struct btrfs_file_extent_item *extent;
			int type;
Z
Zheng Yan 已提交
2634 2635
			u32 size;
			struct btrfs_key new_key;
2636 2637 2638
			u64 disko = 0, diskl = 0;
			u64 datao = 0, datal = 0;
			u8 comp;
2639
			u64 endoff;
Z
Zheng Yan 已提交
2640 2641 2642 2643 2644

			size = btrfs_item_size_nr(leaf, slot);
			read_extent_buffer(leaf, buf,
					   btrfs_item_ptr_offset(leaf, slot),
					   size);
2645 2646 2647 2648 2649

			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);
2650 2651
			if (type == BTRFS_FILE_EXTENT_REG ||
			    type == BTRFS_FILE_EXTENT_PREALLOC) {
C
Chris Mason 已提交
2652 2653 2654 2655
				disko = btrfs_file_extent_disk_bytenr(leaf,
								      extent);
				diskl = btrfs_file_extent_disk_num_bytes(leaf,
								 extent);
2656
				datao = btrfs_file_extent_offset(leaf, extent);
C
Chris Mason 已提交
2657 2658
				datal = btrfs_file_extent_num_bytes(leaf,
								    extent);
2659 2660 2661 2662 2663
			} else if (type == BTRFS_FILE_EXTENT_INLINE) {
				/* take upper bound, may be compressed */
				datal = btrfs_file_extent_ram_bytes(leaf,
								    extent);
			}
2664
			btrfs_release_path(path);
Z
Zheng Yan 已提交
2665

2666
			if (key.offset + datal <= off ||
L
Liu Bo 已提交
2667
			    key.offset >= off + len - 1)
2668 2669
				goto next;

Z
Zheng Yan 已提交
2670
			memcpy(&new_key, &key, sizeof(new_key));
L
Li Zefan 已提交
2671
			new_key.objectid = btrfs_ino(inode);
2672 2673 2674 2675
			if (off <= key.offset)
				new_key.offset = key.offset + destoff - off;
			else
				new_key.offset = destoff;
Z
Zheng Yan 已提交
2676

2677 2678 2679 2680 2681 2682
			/*
			 * 1 - adjusting old extent (we may have to split it)
			 * 1 - add new extent
			 * 1 - inode update
			 */
			trans = btrfs_start_transaction(root, 3);
2683 2684 2685 2686 2687
			if (IS_ERR(trans)) {
				ret = PTR_ERR(trans);
				goto out;
			}

2688 2689
			if (type == BTRFS_FILE_EXTENT_REG ||
			    type == BTRFS_FILE_EXTENT_PREALLOC) {
2690 2691 2692 2693 2694 2695 2696 2697 2698 2699
				/*
				 *    a  | --- range to clone ---|  b
				 * | ------------- extent ------------- |
				 */

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

				/* substract range a */
2700 2701 2702 2703 2704
				if (off > key.offset) {
					datao += off - key.offset;
					datal -= off - key.offset;
				}

J
Josef Bacik 已提交
2705
				ret = btrfs_drop_extents(trans, root, inode,
2706 2707
							 new_key.offset,
							 new_key.offset + datal,
2708
							 1);
2709 2710 2711 2712 2713 2714
				if (ret) {
					btrfs_abort_transaction(trans, root,
								ret);
					btrfs_end_transaction(trans, root);
					goto out;
				}
2715

2716 2717
				ret = btrfs_insert_empty_item(trans, root, path,
							      &new_key, size);
2718 2719 2720 2721 2722 2723
				if (ret) {
					btrfs_abort_transaction(trans, root,
								ret);
					btrfs_end_transaction(trans, root);
					goto out;
				}
2724 2725 2726 2727

				leaf = path->nodes[0];
				slot = path->slots[0];
				write_extent_buffer(leaf, buf,
Z
Zheng Yan 已提交
2728 2729
					    btrfs_item_ptr_offset(leaf, slot),
					    size);
Y
Yan Zheng 已提交
2730

2731
				extent = btrfs_item_ptr(leaf, slot,
C
Christoph Hellwig 已提交
2732
						struct btrfs_file_extent_item);
2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743

				/* 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
2744
					ret = btrfs_inc_extent_ref(trans, root,
2745 2746
							disko, diskl, 0,
							root->root_key.objectid,
L
Li Zefan 已提交
2747
							btrfs_ino(inode),
A
Arne Jansen 已提交
2748 2749
							new_key.offset - datao,
							0);
2750 2751 2752 2753 2754 2755 2756 2757 2758
					if (ret) {
						btrfs_abort_transaction(trans,
									root,
									ret);
						btrfs_end_transaction(trans,
								      root);
						goto out;

					}
C
Christoph Hellwig 已提交
2759
				}
2760 2761 2762 2763 2764 2765 2766
			} 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
2767

L
Liu Bo 已提交
2768 2769
				if (key.offset + datal > off + len)
					trim = key.offset + datal - (off + len);
C
Chris Mason 已提交
2770

2771 2772
				if (comp && (skip || trim)) {
					ret = -EINVAL;
2773
					btrfs_end_transaction(trans, root);
2774 2775 2776 2777
					goto out;
				}
				size -= skip + trim;
				datal -= skip + trim;
2778

J
Josef Bacik 已提交
2779
				ret = btrfs_drop_extents(trans, root, inode,
2780 2781
							 new_key.offset,
							 new_key.offset + datal,
2782
							 1);
2783 2784 2785 2786 2787 2788
				if (ret) {
					btrfs_abort_transaction(trans, root,
								ret);
					btrfs_end_transaction(trans, root);
					goto out;
				}
2789

2790 2791
				ret = btrfs_insert_empty_item(trans, root, path,
							      &new_key, size);
2792 2793 2794 2795 2796 2797
				if (ret) {
					btrfs_abort_transaction(trans, root,
								ret);
					btrfs_end_transaction(trans, root);
					goto out;
				}
2798 2799

				if (skip) {
C
Chris Mason 已提交
2800 2801
					u32 start =
					  btrfs_file_extent_calc_inline_size(0);
2802 2803 2804 2805 2806 2807 2808 2809 2810 2811
					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
2812
			}
2813 2814

			btrfs_mark_buffer_dirty(leaf);
2815
			btrfs_release_path(path);
2816

2817
			inode_inc_iversion(inode);
2818
			inode->i_mtime = inode->i_ctime = CURRENT_TIME;
2819 2820 2821 2822 2823 2824 2825

			/*
			 * 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;
2826 2827
			if (endoff > destoff+olen)
				endoff = destoff+olen;
2828 2829 2830
			if (endoff > inode->i_size)
				btrfs_i_size_write(inode, endoff);

2831
			ret = btrfs_update_inode(trans, root, inode);
2832 2833 2834 2835 2836 2837
			if (ret) {
				btrfs_abort_transaction(trans, root, ret);
				btrfs_end_transaction(trans, root);
				goto out;
			}
			ret = btrfs_end_transaction(trans, root);
2838
		}
C
Chris Mason 已提交
2839
next:
2840
		btrfs_release_path(path);
C
Christoph Hellwig 已提交
2841 2842 2843 2844
		key.offset++;
	}
	ret = 0;
out:
2845
	btrfs_release_path(path);
L
Liu Bo 已提交
2846
	unlock_extent(&BTRFS_I(src)->io_tree, off, off + len - 1);
C
Christoph Hellwig 已提交
2847 2848 2849
out_unlock:
	mutex_unlock(&src->i_mutex);
	mutex_unlock(&inode->i_mutex);
Y
Yan Zheng 已提交
2850 2851
	vfree(buf);
	btrfs_free_path(path);
C
Christoph Hellwig 已提交
2852
out_fput:
2853
	fdput(src_file);
2854
out_drop_write:
A
Al Viro 已提交
2855
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
2856 2857 2858
	return ret;
}

2859
static long btrfs_ioctl_clone_range(struct file *file, void __user *argp)
2860 2861 2862
{
	struct btrfs_ioctl_clone_range_args args;

2863
	if (copy_from_user(&args, argp, sizeof(args)))
2864 2865 2866 2867 2868
		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
2869 2870 2871 2872 2873 2874
/*
 * 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.
 */
2875
static long btrfs_ioctl_trans_start(struct file *file)
C
Christoph Hellwig 已提交
2876
{
A
Al Viro 已提交
2877
	struct inode *inode = file_inode(file);
C
Christoph Hellwig 已提交
2878 2879
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;
2880
	int ret;
C
Christoph Hellwig 已提交
2881

2882
	ret = -EPERM;
2883
	if (!capable(CAP_SYS_ADMIN))
2884
		goto out;
2885

2886 2887
	ret = -EINPROGRESS;
	if (file->private_data)
C
Christoph Hellwig 已提交
2888
		goto out;
2889

L
Li Zefan 已提交
2890 2891 2892 2893
	ret = -EROFS;
	if (btrfs_root_readonly(root))
		goto out;

2894
	ret = mnt_want_write_file(file);
Y
Yan Zheng 已提交
2895 2896 2897
	if (ret)
		goto out;

J
Josef Bacik 已提交
2898
	atomic_inc(&root->fs_info->open_ioctl_trans);
2899

2900
	ret = -ENOMEM;
2901
	trans = btrfs_start_ioctl_transaction(root);
2902
	if (IS_ERR(trans))
2903 2904 2905 2906 2907 2908
		goto out_drop;

	file->private_data = trans;
	return 0;

out_drop:
J
Josef Bacik 已提交
2909
	atomic_dec(&root->fs_info->open_ioctl_trans);
A
Al Viro 已提交
2910
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
2911 2912 2913 2914
out:
	return ret;
}

2915 2916
static long btrfs_ioctl_default_subvol(struct file *file, void __user *argp)
{
A
Al Viro 已提交
2917
	struct inode *inode = file_inode(file);
2918 2919 2920 2921 2922 2923 2924 2925 2926
	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;
2927
	int ret;
2928 2929 2930 2931

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

2932 2933 2934 2935 2936 2937 2938 2939
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;

	if (copy_from_user(&objectid, argp, sizeof(objectid))) {
		ret = -EFAULT;
		goto out;
	}
2940 2941 2942 2943 2944 2945 2946 2947 2948

	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);
2949 2950 2951 2952
	if (IS_ERR(new_root)) {
		ret = PTR_ERR(new_root);
		goto out;
	}
2953

2954 2955 2956 2957
	if (btrfs_root_refs(&new_root->root_item) == 0) {
		ret = -ENOENT;
		goto out;
	}
2958 2959

	path = btrfs_alloc_path();
2960 2961 2962 2963
	if (!path) {
		ret = -ENOMEM;
		goto out;
	}
2964 2965 2966
	path->leave_spinning = 1;

	trans = btrfs_start_transaction(root, 1);
2967
	if (IS_ERR(trans)) {
2968
		btrfs_free_path(path);
2969 2970
		ret = PTR_ERR(trans);
		goto out;
2971 2972
	}

2973
	dir_id = btrfs_super_root_dir(root->fs_info->super_copy);
2974 2975
	di = btrfs_lookup_dir_item(trans, root->fs_info->tree_root, path,
				   dir_id, "default", 7, 1);
2976
	if (IS_ERR_OR_NULL(di)) {
2977 2978 2979 2980
		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");
2981 2982
		ret = -ENOENT;
		goto out;
2983 2984 2985 2986 2987 2988 2989
	}

	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);

2990
	btrfs_set_fs_incompat(root->fs_info, DEFAULT_SUBVOL);
2991
	btrfs_end_transaction(trans, root);
2992 2993 2994
out:
	mnt_drop_write_file(file);
	return ret;
2995 2996
}

2997 2998
void btrfs_get_block_group_info(struct list_head *groups_list,
				struct btrfs_ioctl_space_info *space)
2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012
{
	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);
	}
}

3013
static long btrfs_ioctl_space_info(struct btrfs_root *root, void __user *arg)
J
Josef Bacik 已提交
3014 3015 3016 3017
{
	struct btrfs_ioctl_space_args space_args;
	struct btrfs_ioctl_space_info space;
	struct btrfs_ioctl_space_info *dest;
3018
	struct btrfs_ioctl_space_info *dest_orig;
3019
	struct btrfs_ioctl_space_info __user *user_dest;
J
Josef Bacik 已提交
3020
	struct btrfs_space_info *info;
3021 3022 3023 3024 3025
	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;
3026
	int alloc_size;
J
Josef Bacik 已提交
3027
	int ret = 0;
3028
	u64 slot_count = 0;
3029
	int i, c;
J
Josef Bacik 已提交
3030 3031 3032 3033 3034 3035

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

3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058 3059
	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);
	}
3060 3061 3062 3063 3064 3065

	/* space_slots == 0 means they are asking for a count */
	if (space_args.space_slots == 0) {
		space_args.total_spaces = slot_count;
		goto out;
	}
3066

3067
	slot_count = min_t(u64, space_args.space_slots, slot_count);
3068

3069
	alloc_size = sizeof(*dest) * slot_count;
3070

3071 3072 3073 3074 3075 3076
	/* 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
3077
	space_args.total_spaces = 0;
3078 3079 3080 3081
	dest = kmalloc(alloc_size, GFP_NOFS);
	if (!dest)
		return -ENOMEM;
	dest_orig = dest;
J
Josef Bacik 已提交
3082

3083
	/* now we have a buffer to copy into */
3084 3085 3086
	for (i = 0; i < num_types; i++) {
		struct btrfs_space_info *tmp;

3087 3088 3089
		if (!slot_count)
			break;

3090 3091 3092 3093 3094 3095 3096 3097 3098 3099
		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();
3100

3101 3102 3103 3104 3105
		if (!info)
			continue;
		down_read(&info->groups_sem);
		for (c = 0; c < BTRFS_NR_RAID_TYPES; c++) {
			if (!list_empty(&info->block_groups[c])) {
3106 3107
				btrfs_get_block_group_info(
					&info->block_groups[c], &space);
3108 3109 3110
				memcpy(dest, &space, sizeof(space));
				dest++;
				space_args.total_spaces++;
3111
				slot_count--;
3112
			}
3113 3114
			if (!slot_count)
				break;
3115 3116
		}
		up_read(&info->groups_sem);
J
Josef Bacik 已提交
3117 3118
	}

D
Daniel J Blueman 已提交
3119
	user_dest = (struct btrfs_ioctl_space_info __user *)
3120 3121 3122 3123 3124 3125 3126 3127
		(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
3128 3129 3130 3131 3132
		ret = -EFAULT;

	return ret;
}

C
Christoph Hellwig 已提交
3133 3134 3135 3136 3137 3138 3139 3140
/*
 * 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
3141
	struct inode *inode = file_inode(file);
C
Christoph Hellwig 已提交
3142 3143 3144 3145
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;

	trans = file->private_data;
3146 3147
	if (!trans)
		return -EINVAL;
C
Christoph Hellwig 已提交
3148
	file->private_data = NULL;
3149

3150 3151
	btrfs_end_transaction(trans, root);

J
Josef Bacik 已提交
3152
	atomic_dec(&root->fs_info->open_ioctl_trans);
3153

A
Al Viro 已提交
3154
	mnt_drop_write_file(file);
3155
	return 0;
C
Christoph Hellwig 已提交
3156 3157
}

3158 3159
static noinline long btrfs_ioctl_start_sync(struct btrfs_root *root,
					    void __user *argp)
3160 3161 3162
{
	struct btrfs_trans_handle *trans;
	u64 transid;
T
Tsutomu Itoh 已提交
3163
	int ret;
3164

M
Miao Xie 已提交
3165
	trans = btrfs_attach_transaction_barrier(root);
3166 3167 3168 3169 3170 3171 3172 3173
	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;
	}
3174
	transid = trans->transid;
T
Tsutomu Itoh 已提交
3175
	ret = btrfs_commit_transaction_async(trans, root, 0);
3176 3177
	if (ret) {
		btrfs_end_transaction(trans, root);
T
Tsutomu Itoh 已提交
3178
		return ret;
3179
	}
3180
out:
3181 3182 3183 3184 3185 3186
	if (argp)
		if (copy_to_user(argp, &transid, sizeof(transid)))
			return -EFAULT;
	return 0;
}

3187 3188
static noinline long btrfs_ioctl_wait_sync(struct btrfs_root *root,
					   void __user *argp)
3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200
{
	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
3201
static long btrfs_ioctl_scrub(struct file *file, void __user *arg)
J
Jan Schmidt 已提交
3202
{
A
Al Viro 已提交
3203
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
J
Jan Schmidt 已提交
3204
	struct btrfs_ioctl_scrub_args *sa;
M
Miao Xie 已提交
3205
	int ret;
J
Jan Schmidt 已提交
3206 3207 3208 3209 3210 3211 3212 3213

	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
3214 3215 3216 3217 3218 3219
	if (!(sa->flags & BTRFS_SCRUB_READONLY)) {
		ret = mnt_want_write_file(file);
		if (ret)
			goto out;
	}

3220
	ret = btrfs_scrub_dev(root->fs_info, sa->devid, sa->start, sa->end,
3221 3222
			      &sa->progress, sa->flags & BTRFS_SCRUB_READONLY,
			      0);
J
Jan Schmidt 已提交
3223 3224 3225 3226

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

M
Miao Xie 已提交
3227 3228 3229
	if (!(sa->flags & BTRFS_SCRUB_READONLY))
		mnt_drop_write_file(file);
out:
J
Jan Schmidt 已提交
3230 3231 3232 3233 3234 3235 3236 3237 3238
	kfree(sa);
	return ret;
}

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

3239
	return btrfs_scrub_cancel(root->fs_info);
J
Jan Schmidt 已提交
3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263
}

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;
}

3264
static long btrfs_ioctl_get_dev_stats(struct btrfs_root *root,
3265
				      void __user *arg)
3266 3267 3268 3269 3270 3271 3272 3273
{
	struct btrfs_ioctl_get_dev_stats *sa;
	int ret;

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

3274 3275 3276 3277 3278 3279
	if ((sa->flags & BTRFS_DEV_STATS_RESET) && !capable(CAP_SYS_ADMIN)) {
		kfree(sa);
		return -EPERM;
	}

	ret = btrfs_get_dev_stats(root, sa);
3280 3281 3282 3283 3284 3285 3286 3287

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

	kfree(sa);
	return ret;
}

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 3330 3331 3332
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;
}

3333 3334 3335 3336
static long btrfs_ioctl_ino_to_path(struct btrfs_root *root, void __user *arg)
{
	int ret = 0;
	int i;
3337
	u64 rel_ptr;
3338
	int size;
3339
	struct btrfs_ioctl_ino_path_args *ipa = NULL;
3340 3341 3342
	struct inode_fs_paths *ipath = NULL;
	struct btrfs_path *path;

3343
	if (!capable(CAP_DAC_READ_SEARCH))
3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371
		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) {
3372 3373
		rel_ptr = ipath->fspath->val[i] -
			  (u64)(unsigned long)ipath->fspath->val;
3374
		ipath->fspath->val[i] = rel_ptr;
3375 3376
	}

3377 3378
	ret = copy_to_user((void *)(unsigned long)ipa->fspath,
			   (void *)(unsigned long)ipath->fspath, size);
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 3434 3435 3436
	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;
	}

3437
	size = min_t(u32, loi->size, 64 * 1024);
3438 3439 3440 3441 3442 3443 3444
	inodes = init_data_container(size);
	if (IS_ERR(inodes)) {
		ret = PTR_ERR(inodes);
		inodes = NULL;
		goto out;
	}

L
Liu Bo 已提交
3445 3446 3447
	ret = iterate_inodes_from_logical(loi->logical, root->fs_info, path,
					  build_ino_list, inodes);
	if (ret == -EINVAL)
3448 3449 3450 3451
		ret = -ENOENT;
	if (ret < 0)
		goto out;

3452 3453
	ret = copy_to_user((void *)(unsigned long)loi->inodes,
			   (void *)(unsigned long)inodes, size);
3454 3455 3456 3457 3458
	if (ret)
		ret = -EFAULT;

out:
	btrfs_free_path(path);
3459
	vfree(inodes);
3460 3461 3462 3463 3464
	kfree(loi);

	return ret;
}

3465
void update_ioctl_balance_args(struct btrfs_fs_info *fs_info, int lock,
3466 3467 3468 3469 3470 3471
			       struct btrfs_ioctl_balance_args *bargs)
{
	struct btrfs_balance_control *bctl = fs_info->balance_ctl;

	bargs->flags = bctl->flags;

3472 3473 3474 3475
	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;
3476 3477
	if (atomic_read(&fs_info->balance_cancel_req))
		bargs->state |= BTRFS_BALANCE_STATE_CANCEL_REQ;
3478

3479 3480 3481
	memcpy(&bargs->data, &bctl->data, sizeof(bargs->data));
	memcpy(&bargs->meta, &bctl->meta, sizeof(bargs->meta));
	memcpy(&bargs->sys, &bctl->sys, sizeof(bargs->sys));
3482 3483 3484 3485 3486 3487 3488 3489

	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));
	}
3490 3491
}

3492
static long btrfs_ioctl_balance(struct file *file, void __user *arg)
3493
{
A
Al Viro 已提交
3494
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
3495 3496 3497
	struct btrfs_fs_info *fs_info = root->fs_info;
	struct btrfs_ioctl_balance_args *bargs;
	struct btrfs_balance_control *bctl;
3498
	bool need_unlock; /* for mut. excl. ops lock */
3499 3500 3501 3502 3503
	int ret;

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

3504
	ret = mnt_want_write_file(file);
3505 3506 3507
	if (ret)
		return ret;

3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521
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)
	 */
3522
	mutex_lock(&fs_info->balance_mutex);
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 3554 3555 3556
	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));
3557 3558 3559 3560 3561

	if (arg) {
		bargs = memdup_user(arg, sizeof(*bargs));
		if (IS_ERR(bargs)) {
			ret = PTR_ERR(bargs);
3562
			goto out_unlock;
3563
		}
3564 3565 3566 3567 3568 3569 3570 3571 3572 3573 3574 3575 3576 3577

		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;
		}
3578 3579 3580 3581
	} else {
		bargs = NULL;
	}

3582
	if (fs_info->balance_ctl) {
3583 3584 3585 3586
		ret = -EINPROGRESS;
		goto out_bargs;
	}

3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 3598 3599
	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;
3600 3601 3602
	} else {
		/* balance everything - no filters */
		bctl->flags |= BTRFS_BALANCE_TYPE_MASK;
3603 3604
	}

3605
do_balance:
3606
	/*
3607 3608 3609 3610 3611
	 * 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.
3612
	 */
3613 3614 3615 3616
	need_unlock = false;

	ret = btrfs_balance(bctl, bargs);

3617 3618 3619 3620 3621 3622 3623
	if (arg) {
		if (copy_to_user(arg, bargs, sizeof(*bargs)))
			ret = -EFAULT;
	}

out_bargs:
	kfree(bargs);
3624
out_unlock:
3625 3626
	mutex_unlock(&fs_info->balance_mutex);
	mutex_unlock(&fs_info->volume_mutex);
3627 3628 3629
	if (need_unlock)
		atomic_set(&fs_info->mutually_exclusive_operation_running, 0);
out:
3630
	mnt_drop_write_file(file);
3631 3632 3633
	return ret;
}

3634 3635 3636 3637 3638 3639 3640 3641
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);
3642 3643
	case BTRFS_BALANCE_CTL_CANCEL:
		return btrfs_cancel_balance(root->fs_info);
3644 3645 3646 3647 3648
	}

	return -EINVAL;
}

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 3679 3680 3681
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;
}

3682
static long btrfs_ioctl_quota_ctl(struct file *file, void __user *arg)
A
Arne Jansen 已提交
3683
{
A
Al Viro 已提交
3684
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
A
Arne Jansen 已提交
3685 3686 3687 3688 3689 3690 3691 3692
	struct btrfs_ioctl_quota_ctl_args *sa;
	struct btrfs_trans_handle *trans = NULL;
	int ret;
	int err;

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

3693 3694 3695
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
3696 3697

	sa = memdup_user(arg, sizeof(*sa));
3698 3699 3700 3701
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
3702

3703
	down_write(&root->fs_info->subvol_sem);
J
Jan Schmidt 已提交
3704 3705 3706 3707
	trans = btrfs_start_transaction(root->fs_info->tree_root, 2);
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto out;
A
Arne Jansen 已提交
3708 3709 3710 3711 3712 3713 3714 3715 3716 3717 3718 3719 3720 3721 3722 3723 3724
	}

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

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

J
Jan Schmidt 已提交
3725 3726 3727
	err = btrfs_commit_transaction(trans, root->fs_info->tree_root);
	if (err && !ret)
		ret = err;
A
Arne Jansen 已提交
3728 3729
out:
	kfree(sa);
3730
	up_write(&root->fs_info->subvol_sem);
3731 3732
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
3733 3734 3735
	return ret;
}

3736
static long btrfs_ioctl_qgroup_assign(struct file *file, void __user *arg)
A
Arne Jansen 已提交
3737
{
A
Al Viro 已提交
3738
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
A
Arne Jansen 已提交
3739 3740 3741 3742 3743 3744 3745 3746
	struct btrfs_ioctl_qgroup_assign_args *sa;
	struct btrfs_trans_handle *trans;
	int ret;
	int err;

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

3747 3748 3749
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
3750 3751

	sa = memdup_user(arg, sizeof(*sa));
3752 3753 3754 3755
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
3756 3757 3758 3759 3760 3761 3762 3763 3764 3765 3766 3767 3768 3769 3770 3771 3772 3773 3774 3775 3776 3777

	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);
3778 3779
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
3780 3781 3782
	return ret;
}

3783
static long btrfs_ioctl_qgroup_create(struct file *file, void __user *arg)
A
Arne Jansen 已提交
3784
{
A
Al Viro 已提交
3785
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
A
Arne Jansen 已提交
3786 3787 3788 3789 3790 3791 3792 3793
	struct btrfs_ioctl_qgroup_create_args *sa;
	struct btrfs_trans_handle *trans;
	int ret;
	int err;

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

3794 3795 3796
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
3797 3798

	sa = memdup_user(arg, sizeof(*sa));
3799 3800 3801 3802
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
3803

M
Miao Xie 已提交
3804 3805 3806 3807 3808
	if (!sa->qgroupid) {
		ret = -EINVAL;
		goto out;
	}

A
Arne Jansen 已提交
3809 3810 3811 3812 3813 3814 3815 3816 3817 3818 3819 3820 3821 3822 3823 3824 3825 3826 3827 3828
	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);
3829 3830
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
3831 3832 3833
	return ret;
}

3834
static long btrfs_ioctl_qgroup_limit(struct file *file, void __user *arg)
A
Arne Jansen 已提交
3835
{
A
Al Viro 已提交
3836
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
A
Arne Jansen 已提交
3837 3838 3839 3840 3841 3842 3843 3844 3845
	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;

3846 3847 3848
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
3849 3850

	sa = memdup_user(arg, sizeof(*sa));
3851 3852 3853 3854
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 3865 3866 3867 3868 3869 3870 3871 3872 3873 3874 3875 3876

	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);
3877 3878
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
3879 3880 3881
	return ret;
}

J
Jan Schmidt 已提交
3882 3883 3884 3885 3886 3887 3888 3889 3890 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
static long btrfs_ioctl_quota_rescan(struct file *file, void __user *arg)
{
	struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;
	struct btrfs_ioctl_quota_rescan_args *qsa;
	int ret;

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

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

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

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

	ret = btrfs_qgroup_rescan(root->fs_info);

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

static long btrfs_ioctl_quota_rescan_status(struct file *file, void __user *arg)
{
	struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;
	struct btrfs_ioctl_quota_rescan_args *qsa;
	int ret = 0;

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

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

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

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

	kfree(qsa);
	return ret;
}

3940 3941 3942 3943
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
3944
	struct inode *inode = file_inode(file);
3945 3946 3947 3948 3949 3950 3951 3952 3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975 3976 3977 3978 3979 3980 3981 3982 3983 3984 3985 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
	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;
}

4021 4022 4023 4024 4025 4026 4027 4028 4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039
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;
}

4040 4041 4042 4043 4044 4045 4046 4047 4048 4049 4050 4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 4062 4063 4064 4065 4066 4067 4068 4069 4070 4071 4072 4073 4074 4075 4076 4077 4078 4079
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
4080 4081 4082
long btrfs_ioctl(struct file *file, unsigned int
		cmd, unsigned long arg)
{
A
Al Viro 已提交
4083
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
4084
	void __user *argp = (void __user *)arg;
C
Christoph Hellwig 已提交
4085 4086

	switch (cmd) {
4087 4088 4089 4090 4091 4092
	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);
4093 4094
	case FITRIM:
		return btrfs_ioctl_fitrim(file, argp);
C
Christoph Hellwig 已提交
4095
	case BTRFS_IOC_SNAP_CREATE:
4096
		return btrfs_ioctl_snap_create(file, argp, 0);
4097
	case BTRFS_IOC_SNAP_CREATE_V2:
4098
		return btrfs_ioctl_snap_create_v2(file, argp, 0);
4099
	case BTRFS_IOC_SUBVOL_CREATE:
4100
		return btrfs_ioctl_snap_create(file, argp, 1);
A
Arne Jansen 已提交
4101 4102
	case BTRFS_IOC_SUBVOL_CREATE_V2:
		return btrfs_ioctl_snap_create_v2(file, argp, 1);
4103 4104
	case BTRFS_IOC_SNAP_DESTROY:
		return btrfs_ioctl_snap_destroy(file, argp);
4105 4106 4107 4108
	case BTRFS_IOC_SUBVOL_GETFLAGS:
		return btrfs_ioctl_subvol_getflags(file, argp);
	case BTRFS_IOC_SUBVOL_SETFLAGS:
		return btrfs_ioctl_subvol_setflags(file, argp);
4109 4110
	case BTRFS_IOC_DEFAULT_SUBVOL:
		return btrfs_ioctl_default_subvol(file, argp);
C
Christoph Hellwig 已提交
4111
	case BTRFS_IOC_DEFRAG:
C
Chris Mason 已提交
4112 4113 4114
		return btrfs_ioctl_defrag(file, NULL);
	case BTRFS_IOC_DEFRAG_RANGE:
		return btrfs_ioctl_defrag(file, argp);
C
Christoph Hellwig 已提交
4115
	case BTRFS_IOC_RESIZE:
4116
		return btrfs_ioctl_resize(file, argp);
C
Christoph Hellwig 已提交
4117
	case BTRFS_IOC_ADD_DEV:
4118
		return btrfs_ioctl_add_dev(root, argp);
C
Christoph Hellwig 已提交
4119
	case BTRFS_IOC_RM_DEV:
4120
		return btrfs_ioctl_rm_dev(file, argp);
J
Jan Schmidt 已提交
4121 4122 4123 4124
	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
4125
	case BTRFS_IOC_BALANCE:
4126
		return btrfs_ioctl_balance(file, NULL);
C
Christoph Hellwig 已提交
4127
	case BTRFS_IOC_CLONE:
4128 4129
		return btrfs_ioctl_clone(file, arg, 0, 0, 0);
	case BTRFS_IOC_CLONE_RANGE:
4130
		return btrfs_ioctl_clone_range(file, argp);
C
Christoph Hellwig 已提交
4131 4132 4133 4134
	case BTRFS_IOC_TRANS_START:
		return btrfs_ioctl_trans_start(file);
	case BTRFS_IOC_TRANS_END:
		return btrfs_ioctl_trans_end(file);
4135 4136 4137 4138
	case BTRFS_IOC_TREE_SEARCH:
		return btrfs_ioctl_tree_search(file, argp);
	case BTRFS_IOC_INO_LOOKUP:
		return btrfs_ioctl_ino_lookup(file, argp);
4139 4140 4141 4142
	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
4143 4144
	case BTRFS_IOC_SPACE_INFO:
		return btrfs_ioctl_space_info(root, argp);
C
Christoph Hellwig 已提交
4145 4146 4147
	case BTRFS_IOC_SYNC:
		btrfs_sync_fs(file->f_dentry->d_sb, 1);
		return 0;
4148
	case BTRFS_IOC_START_SYNC:
4149
		return btrfs_ioctl_start_sync(root, argp);
4150
	case BTRFS_IOC_WAIT_SYNC:
4151
		return btrfs_ioctl_wait_sync(root, argp);
J
Jan Schmidt 已提交
4152
	case BTRFS_IOC_SCRUB:
M
Miao Xie 已提交
4153
		return btrfs_ioctl_scrub(file, argp);
J
Jan Schmidt 已提交
4154 4155 4156 4157
	case BTRFS_IOC_SCRUB_CANCEL:
		return btrfs_ioctl_scrub_cancel(root, argp);
	case BTRFS_IOC_SCRUB_PROGRESS:
		return btrfs_ioctl_scrub_progress(root, argp);
4158
	case BTRFS_IOC_BALANCE_V2:
4159
		return btrfs_ioctl_balance(file, argp);
4160 4161
	case BTRFS_IOC_BALANCE_CTL:
		return btrfs_ioctl_balance_ctl(root, arg);
4162 4163
	case BTRFS_IOC_BALANCE_PROGRESS:
		return btrfs_ioctl_balance_progress(root, argp);
4164 4165
	case BTRFS_IOC_SET_RECEIVED_SUBVOL:
		return btrfs_ioctl_set_received_subvol(file, argp);
4166 4167
	case BTRFS_IOC_SEND:
		return btrfs_ioctl_send(file, argp);
4168
	case BTRFS_IOC_GET_DEV_STATS:
4169
		return btrfs_ioctl_get_dev_stats(root, argp);
A
Arne Jansen 已提交
4170
	case BTRFS_IOC_QUOTA_CTL:
4171
		return btrfs_ioctl_quota_ctl(file, argp);
A
Arne Jansen 已提交
4172
	case BTRFS_IOC_QGROUP_ASSIGN:
4173
		return btrfs_ioctl_qgroup_assign(file, argp);
A
Arne Jansen 已提交
4174
	case BTRFS_IOC_QGROUP_CREATE:
4175
		return btrfs_ioctl_qgroup_create(file, argp);
A
Arne Jansen 已提交
4176
	case BTRFS_IOC_QGROUP_LIMIT:
4177
		return btrfs_ioctl_qgroup_limit(file, argp);
J
Jan Schmidt 已提交
4178 4179 4180 4181
	case BTRFS_IOC_QUOTA_RESCAN:
		return btrfs_ioctl_quota_rescan(file, argp);
	case BTRFS_IOC_QUOTA_RESCAN_STATUS:
		return btrfs_ioctl_quota_rescan_status(file, argp);
4182 4183
	case BTRFS_IOC_DEV_REPLACE:
		return btrfs_ioctl_dev_replace(root, argp);
4184 4185
	case BTRFS_IOC_GET_FSLABEL:
		return btrfs_ioctl_get_fslabel(file, argp);
4186 4187
	case BTRFS_IOC_SET_FSLABEL:
		return btrfs_ioctl_set_fslabel(file, argp);
C
Christoph Hellwig 已提交
4188 4189 4190 4191
	}

	return -ENOTTY;
}