ioctl.c 93.0 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>
C
Chris Mason 已提交
45
#include "compat.h"
C
Christoph Hellwig 已提交
46 47 48 49 50 51 52
#include "ctree.h"
#include "disk-io.h"
#include "transaction.h"
#include "btrfs_inode.h"
#include "ioctl.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

136 137 138 139 140 141 142 143 144 145
	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;
	}

	if (flags & BTRFS_INODE_NODATACOW)
		BTRFS_I(inode)->flags |= BTRFS_INODE_NODATACOW;
146 147 148 149 150 151 152 153 154 155 156 157 158 159

	btrfs_update_iflags(inode);
}

static int btrfs_ioctl_getflags(struct file *file, void __user *arg)
{
	struct btrfs_inode *ip = BTRFS_I(file->f_path.dentry->d_inode);
	unsigned int flags = btrfs_flags_to_ioctl(ip->flags);

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

160 161 162 163 164
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
165 166
		      FS_NOCOMP_FL | FS_COMPR_FL |
		      FS_NOCOW_FL))
167 168 169 170 171 172 173 174
		return -EOPNOTSUPP;

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

	return 0;
}

175 176 177 178 179 180 181 182
static int btrfs_ioctl_setflags(struct file *file, void __user *arg)
{
	struct inode *inode = file->f_path.dentry->d_inode;
	struct btrfs_inode *ip = BTRFS_I(inode);
	struct btrfs_root *root = ip->root;
	struct btrfs_trans_handle *trans;
	unsigned int flags, oldflags;
	int ret;
183 184
	u64 ip_oldflags;
	unsigned int i_oldflags;
185
	umode_t mode;
186

L
Li Zefan 已提交
187 188 189
	if (btrfs_root_readonly(root))
		return -EROFS;

190 191 192
	if (copy_from_user(&flags, arg, sizeof(flags)))
		return -EFAULT;

193 194 195
	ret = check_flags(flags);
	if (ret)
		return ret;
C
Christoph Hellwig 已提交
196

197
	if (!inode_owner_or_capable(inode))
198 199
		return -EACCES;

200 201 202 203
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;

204 205
	mutex_lock(&inode->i_mutex);

206 207
	ip_oldflags = ip->flags;
	i_oldflags = inode->i_flags;
208
	mode = inode->i_mode;
209

210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242
	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;
243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267
	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;
		}
	}
268

269 270 271 272 273 274 275 276 277 278 279
	/*
	 * 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
280 281
	} else {
		ip->flags &= ~(BTRFS_INODE_COMPRESS | BTRFS_INODE_NOCOMPRESS);
282
	}
283

284
	trans = btrfs_start_transaction(root, 1);
285 286 287 288
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto out_drop;
	}
289

290
	btrfs_update_iflags(inode);
291
	inode_inc_iversion(inode);
292
	inode->i_ctime = CURRENT_TIME;
293 294 295
	ret = btrfs_update_inode(trans, root, inode);

	btrfs_end_transaction(trans, root);
296 297 298 299 300
 out_drop:
	if (ret) {
		ip->flags = ip_oldflags;
		inode->i_flags = i_oldflags;
	}
301 302 303

 out_unlock:
	mutex_unlock(&inode->i_mutex);
304
	mnt_drop_write_file(file);
305
	return ret;
306 307 308 309 310 311 312 313
}

static int btrfs_ioctl_getversion(struct file *file, int __user *arg)
{
	struct inode *inode = file->f_path.dentry->d_inode;

	return put_user(inode->i_generation, arg);
}
C
Christoph Hellwig 已提交
314

315 316
static noinline int btrfs_ioctl_fitrim(struct file *file, void __user *arg)
{
317
	struct btrfs_fs_info *fs_info = btrfs_sb(fdentry(file)->d_sb);
318 319 320 321 322
	struct btrfs_device *device;
	struct request_queue *q;
	struct fstrim_range range;
	u64 minlen = ULLONG_MAX;
	u64 num_devices = 0;
323
	u64 total_bytes = btrfs_super_total_bytes(fs_info->super_copy);
324 325 326 327 328
	int ret;

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

329 330 331
	rcu_read_lock();
	list_for_each_entry_rcu(device, &fs_info->fs_devices->devices,
				dev_list) {
332 333 334 335 336 337 338 339 340
		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);
		}
	}
341
	rcu_read_unlock();
342

343 344 345 346
	if (!num_devices)
		return -EOPNOTSUPP;
	if (copy_from_user(&range, arg, sizeof(range)))
		return -EFAULT;
347 348
	if (range.start > total_bytes ||
	    range.len < fs_info->sb->s_blocksize)
349
		return -EINVAL;
350

351
	range.len = min(range.len, total_bytes - range.start);
352
	range.minlen = max(range.minlen, minlen);
353
	ret = btrfs_trim_fs(fs_info->tree_root, &range);
354 355 356 357 358 359 360 361 362
	if (ret < 0)
		return ret;

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

	return 0;
}

363 364
static noinline int create_subvol(struct btrfs_root *root,
				  struct dentry *dentry,
S
Sage Weil 已提交
365
				  char *name, int namelen,
A
Arne Jansen 已提交
366 367
				  u64 *async_transid,
				  struct btrfs_qgroup_inherit **inherit)
C
Christoph Hellwig 已提交
368 369 370 371 372 373
{
	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;
374
	struct btrfs_root *new_root;
375
	struct dentry *parent = dentry->d_parent;
376
	struct inode *dir;
377
	struct timespec cur_time = CURRENT_TIME;
C
Christoph Hellwig 已提交
378 379 380 381
	int ret;
	int err;
	u64 objectid;
	u64 new_dirid = BTRFS_FIRST_FREE_OBJECTID;
382
	u64 index = 0;
383
	uuid_le new_uuid;
C
Christoph Hellwig 已提交
384

385
	ret = btrfs_find_free_objectid(root->fs_info->tree_root, &objectid);
386
	if (ret)
387
		return ret;
388 389 390

	dir = parent->d_inode;

J
Josef Bacik 已提交
391 392 393 394 395 396
	/*
	 * 1 - inode item
	 * 2 - refs
	 * 1 - root item
	 * 2 - dir items
	 */
397
	trans = btrfs_start_transaction(root, 6);
398
	if (IS_ERR(trans))
399
		return PTR_ERR(trans);
C
Christoph Hellwig 已提交
400

A
Arne Jansen 已提交
401 402 403 404 405
	ret = btrfs_qgroup_inherit(trans, root->fs_info, 0, objectid,
				   inherit ? *inherit : NULL);
	if (ret)
		goto fail;

406
	leaf = btrfs_alloc_free_block(trans, root, root->leafsize,
407
				      0, objectid, NULL, 0, 0, 0);
408 409 410 411
	if (IS_ERR(leaf)) {
		ret = PTR_ERR(leaf);
		goto fail;
	}
C
Christoph Hellwig 已提交
412

413
	memset_extent_buffer(leaf, 0, 0, sizeof(struct btrfs_header));
C
Christoph Hellwig 已提交
414 415
	btrfs_set_header_bytenr(leaf, leaf->start);
	btrfs_set_header_generation(leaf, trans->transid);
416
	btrfs_set_header_backref_rev(leaf, BTRFS_MIXED_BACKREF_REV);
C
Christoph Hellwig 已提交
417 418 419 420 421
	btrfs_set_header_owner(leaf, objectid);

	write_extent_buffer(leaf, root->fs_info->fsid,
			    (unsigned long)btrfs_header_fsid(leaf),
			    BTRFS_FSID_SIZE);
422 423 424
	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
425 426
	btrfs_mark_buffer_dirty(leaf);

427 428
	memset(&root_item, 0, sizeof(root_item));

C
Christoph Hellwig 已提交
429 430 431 432
	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);
433
	inode_item->nbytes = cpu_to_le64(root->leafsize);
C
Christoph Hellwig 已提交
434 435
	inode_item->mode = cpu_to_le32(S_IFDIR | 0755);

436 437 438 439
	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
440
	btrfs_set_root_bytenr(&root_item, leaf->start);
441
	btrfs_set_root_generation(&root_item, trans->transid);
C
Christoph Hellwig 已提交
442 443
	btrfs_set_root_level(&root_item, 0);
	btrfs_set_root_refs(&root_item, 1);
444
	btrfs_set_root_used(&root_item, leaf->len);
Y
Yan Zheng 已提交
445
	btrfs_set_root_last_snapshot(&root_item, 0);
C
Christoph Hellwig 已提交
446

447 448 449 450 451
	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);
452
	root_item.otime.nsec = cpu_to_le32(cur_time.tv_nsec);
453 454 455
	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
456

457
	btrfs_tree_unlock(leaf);
C
Christoph Hellwig 已提交
458 459 460 461 462 463
	free_extent_buffer(leaf);
	leaf = NULL;

	btrfs_set_root_dirid(&root_item, new_dirid);

	key.objectid = objectid;
464
	key.offset = 0;
C
Christoph Hellwig 已提交
465 466 467 468 469 470
	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;

471 472
	key.offset = (u64)-1;
	new_root = btrfs_read_fs_root_no_name(root->fs_info, &key);
473 474 475 476 477
	if (IS_ERR(new_root)) {
		btrfs_abort_transaction(trans, root, PTR_ERR(new_root));
		ret = PTR_ERR(new_root);
		goto fail;
	}
478 479 480

	btrfs_record_root_in_trans(trans, new_root);

481
	ret = btrfs_create_subvol_root(trans, new_root, new_dirid);
482 483
	if (ret) {
		/* We potentially lose an unused inode item here */
484
		btrfs_abort_transaction(trans, root, ret);
485 486 487
		goto fail;
	}

C
Christoph Hellwig 已提交
488 489 490
	/*
	 * insert the directory item
	 */
491
	ret = btrfs_set_inode_index(dir, &index);
492 493 494 495
	if (ret) {
		btrfs_abort_transaction(trans, root, ret);
		goto fail;
	}
496 497

	ret = btrfs_insert_dir_item(trans, root,
498
				    name, namelen, dir, &key,
499
				    BTRFS_FT_DIR, index);
500 501
	if (ret) {
		btrfs_abort_transaction(trans, root, ret);
C
Christoph Hellwig 已提交
502
		goto fail;
503
	}
504

505 506 507 508
	btrfs_i_size_write(dir, dir->i_size + namelen * 2);
	ret = btrfs_update_inode(trans, root, dir);
	BUG_ON(ret);

509
	ret = btrfs_add_root_ref(trans, root->fs_info->tree_root,
510
				 objectid, root->root_key.objectid,
L
Li Zefan 已提交
511
				 btrfs_ino(dir), index, name, namelen);
512

513
	BUG_ON(ret);
C
Christoph Hellwig 已提交
514

515
	d_instantiate(dentry, btrfs_lookup_dentry(dir, dentry));
C
Christoph Hellwig 已提交
516
fail:
S
Sage Weil 已提交
517 518 519 520 521 522
	if (async_transid) {
		*async_transid = trans->transid;
		err = btrfs_commit_transaction_async(trans, root, 1);
	} else {
		err = btrfs_commit_transaction(trans, root);
	}
C
Christoph Hellwig 已提交
523 524 525 526 527
	if (err && !ret)
		ret = err;
	return ret;
}

S
Sage Weil 已提交
528
static int create_snapshot(struct btrfs_root *root, struct dentry *dentry,
L
Li Zefan 已提交
529
			   char *name, int namelen, u64 *async_transid,
A
Arne Jansen 已提交
530
			   bool readonly, struct btrfs_qgroup_inherit **inherit)
C
Christoph Hellwig 已提交
531
{
532
	struct inode *inode;
C
Christoph Hellwig 已提交
533 534
	struct btrfs_pending_snapshot *pending_snapshot;
	struct btrfs_trans_handle *trans;
535
	int ret;
C
Christoph Hellwig 已提交
536 537 538 539

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

540
	pending_snapshot = kzalloc(sizeof(*pending_snapshot), GFP_NOFS);
541 542 543
	if (!pending_snapshot)
		return -ENOMEM;

544 545
	btrfs_init_block_rsv(&pending_snapshot->block_rsv,
			     BTRFS_BLOCK_RSV_TEMP);
546
	pending_snapshot->dentry = dentry;
C
Christoph Hellwig 已提交
547
	pending_snapshot->root = root;
L
Li Zefan 已提交
548
	pending_snapshot->readonly = readonly;
A
Arne Jansen 已提交
549 550 551 552
	if (inherit) {
		pending_snapshot->inherit = *inherit;
		*inherit = NULL;	/* take responsibility to free it */
	}
553

554
	trans = btrfs_start_transaction(root->fs_info->extent_root, 6);
555 556 557 558 559 560 561 562
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto fail;
	}

	ret = btrfs_snap_reserve_metadata(trans, pending_snapshot);
	BUG_ON(ret);

563
	spin_lock(&root->fs_info->trans_lock);
C
Christoph Hellwig 已提交
564 565
	list_add(&pending_snapshot->list,
		 &trans->transaction->pending_snapshots);
566
	spin_unlock(&root->fs_info->trans_lock);
S
Sage Weil 已提交
567 568 569 570 571 572 573 574
	if (async_transid) {
		*async_transid = trans->transid;
		ret = btrfs_commit_transaction_async(trans,
				     root->fs_info->extent_root, 1);
	} else {
		ret = btrfs_commit_transaction(trans,
					       root->fs_info->extent_root);
	}
575 576 577 578
	if (ret) {
		/* cleanup_transaction has freed this for us */
		if (trans->aborted)
			pending_snapshot = NULL;
579
		goto fail;
580
	}
581 582 583 584 585

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

586 587 588
	ret = btrfs_orphan_cleanup(pending_snapshot->snap);
	if (ret)
		goto fail;
C
Christoph Hellwig 已提交
589

590
	inode = btrfs_lookup_dentry(dentry->d_parent->d_inode, dentry);
591 592 593 594 595 596 597 598
	if (IS_ERR(inode)) {
		ret = PTR_ERR(inode);
		goto fail;
	}
	BUG_ON(!inode);
	d_instantiate(dentry, inode);
	ret = 0;
fail:
599
	kfree(pending_snapshot);
C
Christoph Hellwig 已提交
600 601 602
	return ret;
}

603 604 605 606 607 608
/*  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)
{
609
	kuid_t fsuid = current_fsuid();
610 611 612

	if (!(dir->i_mode & S_ISVTX))
		return 0;
613
	if (uid_eq(inode->i_uid, fsuid))
614
		return 0;
615
	if (uid_eq(dir->i_uid, fsuid))
616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647
		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);
648
	audit_inode_child(dir, victim, AUDIT_TYPE_CHILD_DELETE);
649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672

	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;
}

673 674 675 676 677 678 679 680 681 682 683 684 685 686 687
/* 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.
 */
688 689
static noinline int btrfs_mksubvol(struct path *parent,
				   char *name, int namelen,
S
Sage Weil 已提交
690
				   struct btrfs_root *snap_src,
A
Arne Jansen 已提交
691 692
				   u64 *async_transid, bool readonly,
				   struct btrfs_qgroup_inherit **inherit)
693
{
694
	struct inode *dir  = parent->dentry->d_inode;
695 696 697
	struct dentry *dentry;
	int error;

698
	mutex_lock_nested(&dir->i_mutex, I_MUTEX_PARENT);
699 700 701 702 703 704 705 706 707 708

	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;

709
	error = btrfs_may_create(dir, dentry);
710
	if (error)
711
		goto out_dput;
712

C
Chris Mason 已提交
713 714 715 716 717 718 719 720 721 722
	/*
	 * 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;

723 724 725 726 727
	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;

728
	if (snap_src) {
A
Arne Jansen 已提交
729 730
		error = create_snapshot(snap_src, dentry, name, namelen,
					async_transid, readonly, inherit);
731
	} else {
732
		error = create_subvol(BTRFS_I(dir)->root, dentry,
A
Arne Jansen 已提交
733
				      name, namelen, async_transid, inherit);
734
	}
735 736 737 738
	if (!error)
		fsnotify_mkdir(dir, dentry);
out_up_read:
	up_read(&BTRFS_I(dir)->root->fs_info->subvol_sem);
739 740 741
out_dput:
	dput(dentry);
out_unlock:
742
	mutex_unlock(&dir->i_mutex);
743 744 745
	return error;
}

C
Chris Mason 已提交
746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796
/*
 * 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;
797
	u64 ino = btrfs_ino(inode);
C
Chris Mason 已提交
798 799 800 801 802

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

803
	min_key.objectid = ino;
C
Chris Mason 已提交
804 805 806
	min_key.type = BTRFS_EXTENT_DATA_KEY;
	min_key.offset = *off;

807
	max_key.objectid = ino;
C
Chris Mason 已提交
808 809 810 811 812 813 814 815 816 817
	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,
					   path, 0, newer_than);
		if (ret != 0)
			goto none;
818
		if (min_key.objectid != ino)
C
Chris Mason 已提交
819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846
			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
847
static struct extent_map *defrag_lookup_extent(struct inode *inode, u64 start)
848 849
{
	struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
L
Li Zefan 已提交
850 851 852
	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
	struct extent_map *em;
	u64 len = PAGE_CACHE_SIZE;
853

L
Li Zefan 已提交
854 855 856 857
	/*
	 * hopefully we have this extent in the tree already, try without
	 * the full extent lock
	 */
858
	read_lock(&em_tree->lock);
L
Li Zefan 已提交
859
	em = lookup_extent_mapping(em_tree, start, len);
860 861
	read_unlock(&em_tree->lock);

L
Li Zefan 已提交
862 863 864 865 866 867 868 869 870 871 872 873
	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;
}
874

L
Li Zefan 已提交
875 876 877 878 879 880 881 882 883 884 885 886 887 888
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);
889 890 891
	return ret;
}

L
Li Zefan 已提交
892
static int should_defrag_range(struct inode *inode, u64 start, int thresh,
893 894
			       u64 *last_len, u64 *skip, u64 *defrag_end,
			       int compress)
895
{
L
Li Zefan 已提交
896
	struct extent_map *em;
897
	int ret = 1;
L
Li Zefan 已提交
898
	bool next_mergeable = true;
899 900

	/*
901
	 * make sure that once we start defragging an extent, we keep on
902 903 904 905 906 907 908
	 * defragging it
	 */
	if (start < *defrag_end)
		return 1;

	*skip = 0;

L
Li Zefan 已提交
909 910 911
	em = defrag_lookup_extent(inode, start);
	if (!em)
		return 0;
912 913

	/* this will cover holes, and inline extents */
914
	if (em->block_start >= EXTENT_MAP_LAST_BYTE) {
915
		ret = 0;
916 917 918
		goto out;
	}

L
Li Zefan 已提交
919
	next_mergeable = defrag_check_next_extent(inode, em);
920 921

	/*
L
Li Zefan 已提交
922 923
	 * we hit a real extent, if it is big or the next extent is not a
	 * real extent, don't bother defragging it
924
	 */
925
	if (!compress && (*last_len == 0 || *last_len >= thresh) &&
L
Li Zefan 已提交
926
	    (em->len >= thresh || !next_mergeable))
927
		ret = 0;
928
out:
929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948
	/*
	 * 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
949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964
/*
 * 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
965
{
C
Chris Mason 已提交
966 967 968 969
	unsigned long file_end;
	u64 isize = i_size_read(inode);
	u64 page_start;
	u64 page_end;
970
	u64 page_cnt;
C
Chris Mason 已提交
971 972 973
	int ret;
	int i;
	int i_done;
974
	struct btrfs_ordered_extent *ordered;
C
Chris Mason 已提交
975
	struct extent_state *cached_state = NULL;
976
	struct extent_io_tree *tree;
977
	gfp_t mask = btrfs_alloc_write_mask(inode->i_mapping);
C
Chris Mason 已提交
978 979

	file_end = (isize - 1) >> PAGE_CACHE_SHIFT;
980 981 982 983
	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
984 985

	ret = btrfs_delalloc_reserve_space(inode,
986
					   page_cnt << PAGE_CACHE_SHIFT);
C
Chris Mason 已提交
987 988 989
	if (ret)
		return ret;
	i_done = 0;
990
	tree = &BTRFS_I(inode)->io_tree;
C
Chris Mason 已提交
991 992

	/* step one, lock all the pages */
993
	for (i = 0; i < page_cnt; i++) {
C
Chris Mason 已提交
994
		struct page *page;
995
again:
996
		page = find_or_create_page(inode->i_mapping,
997
					   start_index + i, mask);
C
Chris Mason 已提交
998 999 1000
		if (!page)
			break;

1001 1002 1003
		page_start = page_offset(page);
		page_end = page_start + PAGE_CACHE_SIZE - 1;
		while (1) {
1004
			lock_extent(tree, page_start, page_end);
1005 1006
			ordered = btrfs_lookup_ordered_extent(inode,
							      page_start);
1007
			unlock_extent(tree, page_start, page_end);
1008 1009 1010 1011 1012 1013 1014
			if (!ordered)
				break;

			unlock_page(page);
			btrfs_start_ordered_extent(inode, ordered, 1);
			btrfs_put_ordered_extent(ordered);
			lock_page(page);
1015 1016 1017 1018 1019 1020 1021 1022 1023
			/*
			 * 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;
			}
1024 1025
		}

C
Chris Mason 已提交
1026 1027 1028 1029 1030 1031 1032 1033 1034 1035
		if (!PageUptodate(page)) {
			btrfs_readpage(NULL, page);
			lock_page(page);
			if (!PageUptodate(page)) {
				unlock_page(page);
				page_cache_release(page);
				ret = -EIO;
				break;
			}
		}
1036 1037 1038 1039 1040 1041 1042

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

C
Chris Mason 已提交
1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062
		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,
1063
			 page_start, page_end - 1, 0, &cached_state);
C
Chris Mason 已提交
1064 1065
	clear_extent_bit(&BTRFS_I(inode)->io_tree, page_start,
			  page_end - 1, EXTENT_DIRTY | EXTENT_DELALLOC |
1066 1067
			  EXTENT_DO_ACCOUNTING | EXTENT_DEFRAG, 0, 0,
			  &cached_state, GFP_NOFS);
C
Chris Mason 已提交
1068

1069
	if (i_done != page_cnt) {
1070 1071 1072
		spin_lock(&BTRFS_I(inode)->lock);
		BTRFS_I(inode)->outstanding_extents++;
		spin_unlock(&BTRFS_I(inode)->lock);
C
Chris Mason 已提交
1073
		btrfs_delalloc_release_space(inode,
1074
				     (page_cnt - i_done) << PAGE_CACHE_SHIFT);
C
Chris Mason 已提交
1075 1076 1077
	}


1078 1079
	set_extent_defrag(&BTRFS_I(inode)->io_tree, page_start, page_end - 1,
			  &cached_state, GFP_NOFS);
C
Chris Mason 已提交
1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098

	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]);
	}
1099
	btrfs_delalloc_release_space(inode, page_cnt << PAGE_CACHE_SHIFT);
C
Chris Mason 已提交
1100 1101 1102 1103 1104 1105 1106 1107 1108 1109
	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
1110
	unsigned long last_index;
1111
	u64 isize = i_size_read(inode);
1112 1113 1114
	u64 last_len = 0;
	u64 skip = 0;
	u64 defrag_end = 0;
C
Chris Mason 已提交
1115
	u64 newer_off = range->start;
C
Christoph Hellwig 已提交
1116
	unsigned long i;
1117
	unsigned long ra_index = 0;
C
Christoph Hellwig 已提交
1118
	int ret;
C
Chris Mason 已提交
1119
	int defrag_count = 0;
1120
	int compress_type = BTRFS_COMPRESS_ZLIB;
C
Chris Mason 已提交
1121
	int extent_thresh = range->extent_thresh;
1122 1123
	int max_cluster = (256 * 1024) >> PAGE_CACHE_SHIFT;
	int cluster = max_cluster;
C
Chris Mason 已提交
1124 1125 1126 1127 1128
	u64 new_align = ~((u64)128 * 1024 - 1);
	struct page **pages = NULL;

	if (extent_thresh == 0)
		extent_thresh = 256 * 1024;
1129 1130 1131 1132 1133 1134 1135

	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
1136

1137
	if (isize == 0)
1138 1139
		return 0;

C
Chris Mason 已提交
1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152
	/*
	 * 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;
	}

1153
	pages = kmalloc(sizeof(struct page *) * max_cluster,
C
Chris Mason 已提交
1154 1155 1156 1157 1158 1159 1160
			GFP_NOFS);
	if (!pages) {
		ret = -ENOMEM;
		goto out_ra;
	}

	/* find the last page to defrag */
C
Chris Mason 已提交
1161
	if (range->start + range->len > range->start) {
1162
		last_index = min_t(u64, isize - 1,
C
Chris Mason 已提交
1163 1164
			 range->start + range->len - 1) >> PAGE_CACHE_SHIFT;
	} else {
1165
		last_index = (isize - 1) >> PAGE_CACHE_SHIFT;
C
Chris Mason 已提交
1166 1167
	}

C
Chris Mason 已提交
1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183
	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
1184
		max_to_defrag = last_index + 1;
C
Chris Mason 已提交
1185

L
Li Zefan 已提交
1186 1187 1188 1189 1190 1191 1192
	/*
	 * 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;

1193 1194 1195
	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
1196 1197 1198 1199 1200 1201 1202
		/*
		 * make sure we stop running if someone unmounts
		 * the FS
		 */
		if (!(inode->i_sb->s_flags & MS_ACTIVE))
			break;

1203
		if (!should_defrag_range(inode, (u64)i << PAGE_CACHE_SHIFT,
L
Li Zefan 已提交
1204
					 extent_thresh, &last_len, &skip,
1205 1206
					 &defrag_end, range->flags &
					 BTRFS_DEFRAG_RANGE_COMPRESS)) {
1207 1208 1209 1210 1211 1212 1213 1214 1215
			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;
		}
1216 1217 1218 1219 1220 1221 1222 1223 1224

		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
1225
		if (range->flags & BTRFS_DEFRAG_RANGE_COMPRESS)
1226
			BTRFS_I(inode)->force_compress = compress_type;
1227

1228 1229 1230 1231 1232 1233
		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;
		}
1234

1235
		mutex_lock(&inode->i_mutex);
1236
		ret = cluster_pages_for_defrag(inode, pages, i, cluster);
1237 1238
		if (ret < 0) {
			mutex_unlock(&inode->i_mutex);
C
Chris Mason 已提交
1239
			goto out_ra;
1240
		}
C
Chris Mason 已提交
1241 1242 1243

		defrag_count += ret;
		balance_dirty_pages_ratelimited_nr(inode->i_mapping, ret);
1244
		mutex_unlock(&inode->i_mutex);
C
Chris Mason 已提交
1245 1246 1247 1248 1249

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

1250 1251 1252
			if (ret > 0)
				i += ret;

C
Chris Mason 已提交
1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263
			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
1264
			}
C
Chris Mason 已提交
1265
		} else {
1266
			if (ret > 0) {
L
Li Zefan 已提交
1267
				i += ret;
1268 1269
				last_len += ret << PAGE_CACHE_SHIFT;
			} else {
L
Li Zefan 已提交
1270
				i++;
1271 1272
				last_len = 0;
			}
C
Christoph Hellwig 已提交
1273 1274 1275
		}
	}

C
Chris Mason 已提交
1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293
	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);
1294
		BTRFS_I(inode)->force_compress = BTRFS_COMPRESS_NONE;
C
Chris Mason 已提交
1295 1296 1297
		mutex_unlock(&inode->i_mutex);
	}

1298
	if (range->compress_type == BTRFS_COMPRESS_LZO) {
1299
		btrfs_set_fs_incompat(root->fs_info, COMPRESS_LZO);
1300 1301
	}

1302
	ret = defrag_count;
1303

C
Chris Mason 已提交
1304 1305 1306 1307
out_ra:
	if (!file)
		kfree(ra);
	kfree(pages);
1308
	return ret;
C
Christoph Hellwig 已提交
1309 1310
}

1311
static noinline int btrfs_ioctl_resize(struct file *file,
1312
					void __user *arg)
C
Christoph Hellwig 已提交
1313 1314 1315 1316
{
	u64 new_size;
	u64 old_size;
	u64 devid = 1;
1317
	struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;
C
Christoph Hellwig 已提交
1318 1319 1320 1321 1322 1323 1324 1325
	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;

Y
Yan Zheng 已提交
1326 1327 1328
	if (root->fs_info->sb->s_flags & MS_RDONLY)
		return -EROFS;

1329 1330 1331
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

1332 1333 1334 1335
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;

1336 1337 1338 1339
	if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
			1)) {
		pr_info("btrfs: dev add/delete/balance/replace/resize operation in progress\n");
		return -EINPROGRESS;
1340 1341
	}

1342
	mutex_lock(&root->fs_info->volume_mutex);
L
Li Zefan 已提交
1343
	vol_args = memdup_user(arg, sizeof(*vol_args));
1344 1345 1346 1347
	if (IS_ERR(vol_args)) {
		ret = PTR_ERR(vol_args);
		goto out;
	}
1348 1349

	vol_args->name[BTRFS_PATH_NAME_MAX] = '\0';
C
Christoph Hellwig 已提交
1350 1351 1352 1353 1354 1355 1356 1357 1358

	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);
1359
		printk(KERN_INFO "btrfs: resizing devid %llu\n",
1360
		       (unsigned long long)devid);
C
Christoph Hellwig 已提交
1361
	}
1362
	device = btrfs_find_device(root->fs_info, devid, NULL, NULL);
C
Christoph Hellwig 已提交
1363
	if (!device) {
1364
		printk(KERN_INFO "btrfs: resizer unable to find device %llu\n",
1365
		       (unsigned long long)devid);
C
Christoph Hellwig 已提交
1366
		ret = -EINVAL;
1367
		goto out_free;
C
Christoph Hellwig 已提交
1368
	}
L
Liu Bo 已提交
1369 1370
	if (device->fs_devices && device->fs_devices->seeding) {
		printk(KERN_INFO "btrfs: resizer unable to apply on "
1371 1372
		       "seeding device %llu\n",
		       (unsigned long long)devid);
L
Liu Bo 已提交
1373 1374 1375 1376
		ret = -EINVAL;
		goto out_free;
	}

C
Christoph Hellwig 已提交
1377 1378 1379 1380 1381 1382 1383 1384 1385 1386
	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
1387
		new_size = memparse(sizestr, NULL);
C
Christoph Hellwig 已提交
1388 1389
		if (new_size == 0) {
			ret = -EINVAL;
1390
			goto out_free;
C
Christoph Hellwig 已提交
1391 1392 1393
		}
	}

1394 1395 1396 1397 1398
	if (device->is_tgtdev_for_dev_replace) {
		ret = -EINVAL;
		goto out_free;
	}

C
Christoph Hellwig 已提交
1399 1400 1401 1402 1403
	old_size = device->total_bytes;

	if (mod < 0) {
		if (new_size > old_size) {
			ret = -EINVAL;
1404
			goto out_free;
C
Christoph Hellwig 已提交
1405 1406 1407 1408 1409 1410 1411 1412
		}
		new_size = old_size - new_size;
	} else if (mod > 0) {
		new_size = old_size + new_size;
	}

	if (new_size < 256 * 1024 * 1024) {
		ret = -EINVAL;
1413
		goto out_free;
C
Christoph Hellwig 已提交
1414 1415 1416
	}
	if (new_size > device->bdev->bd_inode->i_size) {
		ret = -EFBIG;
1417
		goto out_free;
C
Christoph Hellwig 已提交
1418 1419 1420 1421 1422
	}

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

1423 1424 1425
	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
1426 1427

	if (new_size > old_size) {
1428
		trans = btrfs_start_transaction(root, 0);
1429 1430
		if (IS_ERR(trans)) {
			ret = PTR_ERR(trans);
1431
			goto out_free;
1432
		}
C
Christoph Hellwig 已提交
1433 1434
		ret = btrfs_grow_device(trans, device, new_size);
		btrfs_commit_transaction(trans, root);
1435
	} else if (new_size < old_size) {
C
Christoph Hellwig 已提交
1436
		ret = btrfs_shrink_device(device, new_size);
1437
	} /* equal, nothing need to do */
C
Christoph Hellwig 已提交
1438

1439
out_free:
C
Christoph Hellwig 已提交
1440
	kfree(vol_args);
1441 1442
out:
	mutex_unlock(&root->fs_info->volume_mutex);
1443
	mnt_drop_write_file(file);
1444
	atomic_set(&root->fs_info->mutually_exclusive_operation_running, 0);
C
Christoph Hellwig 已提交
1445 1446 1447
	return ret;
}

S
Sage Weil 已提交
1448
static noinline int btrfs_ioctl_snap_create_transid(struct file *file,
A
Arne Jansen 已提交
1449 1450 1451
				char *name, unsigned long fd, int subvol,
				u64 *transid, bool readonly,
				struct btrfs_qgroup_inherit **inherit)
C
Christoph Hellwig 已提交
1452 1453
{
	int namelen;
1454
	int ret = 0;
C
Christoph Hellwig 已提交
1455

1456 1457 1458 1459
	ret = mnt_want_write_file(file);
	if (ret)
		goto out;

S
Sage Weil 已提交
1460 1461
	namelen = strlen(name);
	if (strchr(name, '/')) {
C
Christoph Hellwig 已提交
1462
		ret = -EINVAL;
1463
		goto out_drop_write;
C
Christoph Hellwig 已提交
1464 1465
	}

1466 1467 1468
	if (name[0] == '.' &&
	   (namelen == 1 || (name[1] == '.' && namelen == 2))) {
		ret = -EEXIST;
1469
		goto out_drop_write;
1470 1471
	}

1472
	if (subvol) {
S
Sage Weil 已提交
1473
		ret = btrfs_mksubvol(&file->f_path, name, namelen,
A
Arne Jansen 已提交
1474
				     NULL, transid, readonly, inherit);
1475
	} else {
1476
		struct fd src = fdget(fd);
1477
		struct inode *src_inode;
1478
		if (!src.file) {
1479
			ret = -EINVAL;
1480
			goto out_drop_write;
1481 1482
		}

1483
		src_inode = src.file->f_path.dentry->d_inode;
1484
		if (src_inode->i_sb != file->f_path.dentry->d_inode->i_sb) {
C
Chris Mason 已提交
1485 1486
			printk(KERN_INFO "btrfs: Snapshot src from "
			       "another FS\n");
1487
			ret = -EINVAL;
1488 1489 1490 1491
		} else {
			ret = btrfs_mksubvol(&file->f_path, name, namelen,
					     BTRFS_I(src_inode)->root,
					     transid, readonly, inherit);
1492
		}
1493
		fdput(src);
1494
	}
1495 1496
out_drop_write:
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
1497
out:
S
Sage Weil 已提交
1498 1499 1500 1501
	return ret;
}

static noinline int btrfs_ioctl_snap_create(struct file *file,
1502
					    void __user *arg, int subvol)
S
Sage Weil 已提交
1503
{
1504
	struct btrfs_ioctl_vol_args *vol_args;
S
Sage Weil 已提交
1505 1506
	int ret;

1507 1508 1509 1510
	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
1511

1512
	ret = btrfs_ioctl_snap_create_transid(file, vol_args->name,
L
Li Zefan 已提交
1513
					      vol_args->fd, subvol,
A
Arne Jansen 已提交
1514
					      NULL, false, NULL);
1515

1516 1517 1518
	kfree(vol_args);
	return ret;
}
1519

1520 1521 1522 1523 1524 1525 1526
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
1527
	bool readonly = false;
A
Arne Jansen 已提交
1528
	struct btrfs_qgroup_inherit *inherit = NULL;
1529

1530 1531 1532 1533
	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';
1534

L
Li Zefan 已提交
1535
	if (vol_args->flags &
A
Arne Jansen 已提交
1536 1537
	    ~(BTRFS_SUBVOL_CREATE_ASYNC | BTRFS_SUBVOL_RDONLY |
	      BTRFS_SUBVOL_QGROUP_INHERIT)) {
L
Li Zefan 已提交
1538
		ret = -EOPNOTSUPP;
1539
		goto out;
S
Sage Weil 已提交
1540
	}
1541 1542 1543

	if (vol_args->flags & BTRFS_SUBVOL_CREATE_ASYNC)
		ptr = &transid;
L
Li Zefan 已提交
1544 1545
	if (vol_args->flags & BTRFS_SUBVOL_RDONLY)
		readonly = true;
A
Arne Jansen 已提交
1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556
	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;
		}
	}
1557 1558

	ret = btrfs_ioctl_snap_create_transid(file, vol_args->name,
A
Arne Jansen 已提交
1559 1560
					      vol_args->fd, subvol, ptr,
					      readonly, &inherit);
1561 1562 1563 1564 1565 1566

	if (ret == 0 && ptr &&
	    copy_to_user(arg +
			 offsetof(struct btrfs_ioctl_vol_args_v2,
				  transid), ptr, sizeof(*ptr)))
		ret = -EFAULT;
1567
out:
C
Christoph Hellwig 已提交
1568
	kfree(vol_args);
A
Arne Jansen 已提交
1569
	kfree(inherit);
C
Christoph Hellwig 已提交
1570 1571 1572
	return ret;
}

1573 1574 1575 1576 1577 1578 1579 1580
static noinline int btrfs_ioctl_subvol_getflags(struct file *file,
						void __user *arg)
{
	struct inode *inode = fdentry(file)->d_inode;
	struct btrfs_root *root = BTRFS_I(inode)->root;
	int ret = 0;
	u64 flags = 0;

L
Li Zefan 已提交
1581
	if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID)
1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604
		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)
{
	struct inode *inode = fdentry(file)->d_inode;
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;
	u64 root_flags;
	u64 flags;
	int ret = 0;

1605 1606 1607
	ret = mnt_want_write_file(file);
	if (ret)
		goto out;
1608

1609 1610 1611 1612
	if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID) {
		ret = -EINVAL;
		goto out_drop_write;
	}
1613

1614 1615 1616 1617
	if (copy_from_user(&flags, arg, sizeof(flags))) {
		ret = -EFAULT;
		goto out_drop_write;
	}
1618

1619 1620 1621 1622
	if (flags & BTRFS_SUBVOL_CREATE_ASYNC) {
		ret = -EINVAL;
		goto out_drop_write;
	}
1623

1624 1625 1626 1627
	if (flags & ~BTRFS_SUBVOL_RDONLY) {
		ret = -EOPNOTSUPP;
		goto out_drop_write;
	}
1628

1629 1630 1631 1632
	if (!inode_owner_or_capable(inode)) {
		ret = -EACCES;
		goto out_drop_write;
	}
1633

1634 1635 1636 1637
	down_write(&root->fs_info->subvol_sem);

	/* nothing to do */
	if (!!(flags & BTRFS_SUBVOL_RDONLY) == btrfs_root_readonly(root))
1638
		goto out_drop_sem;
1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653

	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;
	}

1654
	ret = btrfs_update_root(trans, root->fs_info->tree_root,
1655 1656 1657 1658 1659 1660
				&root->root_key, &root->root_item);

	btrfs_commit_transaction(trans, root);
out_reset:
	if (ret)
		btrfs_set_root_flags(&root->root_item, root_flags);
1661
out_drop_sem:
1662
	up_write(&root->fs_info->subvol_sem);
1663 1664 1665
out_drop_write:
	mnt_drop_write_file(file);
out:
1666 1667 1668
	return ret;
}

1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704
/*
 * 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;
}

1705 1706 1707
static noinline int key_in_sk(struct btrfs_key *key,
			      struct btrfs_ioctl_search_key *sk)
{
1708 1709 1710 1711 1712 1713 1714 1715 1716
	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)
1717
		return 0;
1718 1719 1720 1721 1722 1723 1724

	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)
1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790
		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;
		}
1791
		(*num_found)++;
1792 1793 1794 1795 1796

		if (*num_found >= sk->nr_items)
			break;
	}
advance_key:
1797 1798
	ret = 0;
	if (key->offset < (u64)-1 && key->offset < sk->max_offset)
1799
		key->offset++;
1800 1801
	else if (key->type < (u8)-1 && key->type < sk->max_type) {
		key->offset = 0;
1802
		key->type++;
1803 1804 1805
	} else if (key->objectid < (u64)-1 && key->objectid < sk->max_objectid) {
		key->offset = 0;
		key->type = 0;
1806
		key->objectid++;
1807 1808
	} else
		ret = 1;
1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865
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) {
		ret = btrfs_search_forward(root, &key, &max_key, path, 0,
					   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);
1866
		btrfs_release_path(path);
1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887
		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
1888 1889 1890
	args = memdup_user(argp, sizeof(*args));
	if (IS_ERR(args))
		return PTR_ERR(args);
1891 1892 1893 1894 1895 1896 1897 1898 1899

	inode = fdentry(file)->d_inode;
	ret = search_ioctl(inode, args);
	if (ret == 0 && copy_to_user(argp, args, sizeof(*args)))
		ret = -EFAULT;
	kfree(args);
	return ret;
}

1900
/*
1901 1902 1903
 * Search INODE_REFs to identify path name of 'dirid' directory
 * in a 'tree_id' tree. and sets path name to 'name'.
 */
1904 1905 1906 1907 1908
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;
1909
	char *ptr;
1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926
	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;

1927
	ptr = &name[BTRFS_INO_LOOKUP_PATH_MAX];
1928 1929 1930 1931 1932 1933 1934

	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);
1935 1936
		ret = -ENOENT;
		goto out;
1937 1938 1939 1940
	}

	key.objectid = dirid;
	key.type = BTRFS_INODE_REF_KEY;
1941
	key.offset = (u64)-1;
1942 1943 1944 1945 1946 1947 1948 1949

	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];
1950 1951
		if (ret > 0 && slot > 0)
			slot--;
1952 1953 1954
		btrfs_item_key_to_cpu(l, &key, slot);

		if (ret > 0 && (key.objectid != dirid ||
1955 1956
				key.type != BTRFS_INODE_REF_KEY)) {
			ret = -ENOENT;
1957
			goto out;
1958
		}
1959 1960 1961 1962 1963

		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;
1964
		if (ptr < name)
1965 1966 1967 1968 1969 1970 1971 1972
			goto out;

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

		if (key.offset == BTRFS_FIRST_FREE_OBJECTID)
			break;

1973
		btrfs_release_path(path);
1974
		key.objectid = key.offset;
1975
		key.offset = (u64)-1;
1976 1977
		dirid = key.objectid;
	}
1978
	if (ptr < name)
1979
		goto out;
1980
	memmove(name, ptr, total_len);
1981 1982 1983 1984
	name[total_len]='\0';
	ret = 0;
out:
	btrfs_free_path(path);
1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997
	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
1998 1999 2000
	args = memdup_user(argp, sizeof(*args));
	if (IS_ERR(args))
		return PTR_ERR(args);
2001

2002 2003
	inode = fdentry(file)->d_inode;

2004 2005 2006
	if (args->treeid == 0)
		args->treeid = BTRFS_I(inode)->root->root_key.objectid;

2007 2008 2009 2010 2011 2012 2013 2014
	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);
2015 2016 2017
	return ret;
}

2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044
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;
	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;
	}

2045
	err = mnt_want_write_file(file);
2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061
	if (err)
		goto out;

	mutex_lock_nested(&dir->i_mutex, I_MUTEX_PARENT);
	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;
2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101
	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;

		/* check if subvolume may be deleted by a non-root user */
		err = btrfs_may_delete(dir, dentry, 1);
		if (err)
			goto out_dput;
	}

L
Li Zefan 已提交
2102
	if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID) {
2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117
		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;

2118 2119 2120
	trans = btrfs_start_transaction(root, 0);
	if (IS_ERR(trans)) {
		err = PTR_ERR(trans);
2121
		goto out_up_write;
2122 2123 2124
	}
	trans->block_rsv = &root->fs_info->global_block_rsv;

2125 2126 2127 2128
	ret = btrfs_unlink_subvol(trans, root, dir,
				dest->root_key.objectid,
				dentry->d_name.name,
				dentry->d_name.len);
2129 2130 2131 2132 2133
	if (ret) {
		err = ret;
		btrfs_abort_transaction(trans, root, ret);
		goto out_end_trans;
	}
2134 2135 2136 2137 2138 2139 2140 2141

	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);

2142 2143 2144 2145
	if (!xchg(&dest->orphan_item_inserted, 1)) {
		ret = btrfs_insert_orphan_item(trans,
					root->fs_info->tree_root,
					dest->root_key.objectid);
2146 2147 2148 2149 2150
		if (ret) {
			btrfs_abort_transaction(trans, root, ret);
			err = ret;
			goto out_end_trans;
		}
2151
	}
2152
out_end_trans:
S
Sage Weil 已提交
2153
	ret = btrfs_end_transaction(trans, root);
2154 2155
	if (ret && !err)
		err = ret;
2156 2157 2158 2159 2160 2161
	inode->i_flags |= S_DEAD;
out_up_write:
	up_write(&root->fs_info->subvol_sem);
out_unlock:
	mutex_unlock(&inode->i_mutex);
	if (!err) {
2162
		shrink_dcache_sb(root->fs_info->sb);
2163 2164 2165 2166 2167 2168 2169
		btrfs_invalidate_inodes(dest);
		d_delete(dentry);
	}
out_dput:
	dput(dentry);
out_unlock_dir:
	mutex_unlock(&dir->i_mutex);
A
Al Viro 已提交
2170
	mnt_drop_write_file(file);
2171 2172 2173 2174 2175
out:
	kfree(vol_args);
	return err;
}

C
Chris Mason 已提交
2176
static int btrfs_ioctl_defrag(struct file *file, void __user *argp)
C
Christoph Hellwig 已提交
2177 2178 2179
{
	struct inode *inode = fdentry(file)->d_inode;
	struct btrfs_root *root = BTRFS_I(inode)->root;
C
Chris Mason 已提交
2180
	struct btrfs_ioctl_defrag_range_args *range;
Y
Yan Zheng 已提交
2181 2182
	int ret;

L
Li Zefan 已提交
2183 2184 2185
	if (btrfs_root_readonly(root))
		return -EROFS;

2186 2187 2188 2189 2190
	if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
			1)) {
		pr_info("btrfs: dev add/delete/balance/replace/resize operation in progress\n");
		return -EINPROGRESS;
	}
2191
	ret = mnt_want_write_file(file);
2192 2193 2194
	if (ret) {
		atomic_set(&root->fs_info->mutually_exclusive_operation_running,
			   0);
Y
Yan Zheng 已提交
2195
		return ret;
2196
	}
C
Christoph Hellwig 已提交
2197 2198 2199

	switch (inode->i_mode & S_IFMT) {
	case S_IFDIR:
2200 2201 2202 2203
		if (!capable(CAP_SYS_ADMIN)) {
			ret = -EPERM;
			goto out;
		}
2204 2205 2206 2207
		ret = btrfs_defrag_root(root, 0);
		if (ret)
			goto out;
		ret = btrfs_defrag_root(root->fs_info->extent_root, 0);
C
Christoph Hellwig 已提交
2208 2209
		break;
	case S_IFREG:
2210 2211 2212 2213
		if (!(file->f_mode & FMODE_WRITE)) {
			ret = -EINVAL;
			goto out;
		}
C
Chris Mason 已提交
2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225

		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);
2226
				goto out;
C
Chris Mason 已提交
2227 2228 2229 2230 2231 2232 2233 2234 2235 2236
			}
			/* 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;
		}
C
Chris Mason 已提交
2237 2238 2239 2240
		ret = btrfs_defrag_file(fdentry(file)->d_inode, file,
					range, 0, 0);
		if (ret > 0)
			ret = 0;
C
Chris Mason 已提交
2241
		kfree(range);
C
Christoph Hellwig 已提交
2242
		break;
2243 2244
	default:
		ret = -EINVAL;
C
Christoph Hellwig 已提交
2245
	}
2246
out:
A
Al Viro 已提交
2247
	mnt_drop_write_file(file);
2248
	atomic_set(&root->fs_info->mutually_exclusive_operation_running, 0);
2249
	return ret;
C
Christoph Hellwig 已提交
2250 2251
}

2252
static long btrfs_ioctl_add_dev(struct btrfs_root *root, void __user *arg)
C
Christoph Hellwig 已提交
2253 2254 2255 2256
{
	struct btrfs_ioctl_vol_args *vol_args;
	int ret;

2257 2258 2259
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

2260 2261 2262 2263
	if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
			1)) {
		pr_info("btrfs: dev add/delete/balance/replace/resize operation in progress\n");
		return -EINPROGRESS;
2264 2265
	}

2266
	mutex_lock(&root->fs_info->volume_mutex);
L
Li Zefan 已提交
2267
	vol_args = memdup_user(arg, sizeof(*vol_args));
2268 2269 2270 2271
	if (IS_ERR(vol_args)) {
		ret = PTR_ERR(vol_args);
		goto out;
	}
C
Christoph Hellwig 已提交
2272

2273
	vol_args->name[BTRFS_PATH_NAME_MAX] = '\0';
C
Christoph Hellwig 已提交
2274 2275 2276
	ret = btrfs_init_new_device(root, vol_args->name);

	kfree(vol_args);
2277 2278
out:
	mutex_unlock(&root->fs_info->volume_mutex);
2279
	atomic_set(&root->fs_info->mutually_exclusive_operation_running, 0);
C
Christoph Hellwig 已提交
2280 2281 2282
	return ret;
}

2283
static long btrfs_ioctl_rm_dev(struct file *file, void __user *arg)
C
Christoph Hellwig 已提交
2284
{
2285
	struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;
C
Christoph Hellwig 已提交
2286 2287 2288
	struct btrfs_ioctl_vol_args *vol_args;
	int ret;

2289 2290 2291
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

2292 2293 2294
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
Y
Yan Zheng 已提交
2295

2296 2297 2298
	if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
			1)) {
		pr_info("btrfs: dev add/delete/balance/replace/resize operation in progress\n");
2299
		mnt_drop_write_file(file);
2300
		return -EINPROGRESS;
2301 2302
	}

2303
	mutex_lock(&root->fs_info->volume_mutex);
L
Li Zefan 已提交
2304
	vol_args = memdup_user(arg, sizeof(*vol_args));
2305 2306 2307 2308
	if (IS_ERR(vol_args)) {
		ret = PTR_ERR(vol_args);
		goto out;
	}
C
Christoph Hellwig 已提交
2309

2310
	vol_args->name[BTRFS_PATH_NAME_MAX] = '\0';
C
Christoph Hellwig 已提交
2311 2312 2313
	ret = btrfs_rm_device(root, vol_args->name);

	kfree(vol_args);
2314 2315
out:
	mutex_unlock(&root->fs_info->volume_mutex);
2316
	mnt_drop_write_file(file);
2317
	atomic_set(&root->fs_info->mutually_exclusive_operation_running, 0);
C
Christoph Hellwig 已提交
2318 2319 2320
	return ret;
}

J
Jan Schmidt 已提交
2321 2322
static long btrfs_ioctl_fs_info(struct btrfs_root *root, void __user *arg)
{
2323
	struct btrfs_ioctl_fs_info_args *fi_args;
J
Jan Schmidt 已提交
2324 2325 2326
	struct btrfs_device *device;
	struct btrfs_device *next;
	struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
2327
	int ret = 0;
J
Jan Schmidt 已提交
2328 2329 2330 2331

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

2332 2333 2334 2335 2336 2337
	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
2338 2339 2340

	mutex_lock(&fs_devices->device_list_mutex);
	list_for_each_entry_safe(device, next, &fs_devices->devices, dev_list) {
2341 2342
		if (device->devid > fi_args->max_id)
			fi_args->max_id = device->devid;
J
Jan Schmidt 已提交
2343 2344 2345
	}
	mutex_unlock(&fs_devices->device_list_mutex);

2346 2347
	if (copy_to_user(arg, fi_args, sizeof(*fi_args)))
		ret = -EFAULT;
J
Jan Schmidt 已提交
2348

2349 2350
	kfree(fi_args);
	return ret;
J
Jan Schmidt 已提交
2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372
}

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);
2373
	dev = btrfs_find_device(root->fs_info, di_args->devid, s_uuid, NULL);
J
Jan Schmidt 已提交
2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384
	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));
2385
	if (dev->name) {
2386 2387 2388 2389 2390 2391
		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();
2392 2393
		di_args->path[sizeof(di_args->path) - 1] = 0;
	} else {
2394
		di_args->path[0] = '\0';
2395
	}
J
Jan Schmidt 已提交
2396 2397 2398 2399 2400 2401 2402 2403 2404

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

	kfree(di_args);
	return ret;
}

2405 2406
static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd,
				       u64 off, u64 olen, u64 destoff)
C
Christoph Hellwig 已提交
2407 2408 2409
{
	struct inode *inode = fdentry(file)->d_inode;
	struct btrfs_root *root = BTRFS_I(inode)->root;
2410
	struct fd src_file;
C
Christoph Hellwig 已提交
2411 2412 2413 2414
	struct inode *src;
	struct btrfs_trans_handle *trans;
	struct btrfs_path *path;
	struct extent_buffer *leaf;
Y
Yan Zheng 已提交
2415 2416
	char *buf;
	struct btrfs_key key;
C
Christoph Hellwig 已提交
2417 2418
	u32 nritems;
	int slot;
Y
Yan Zheng 已提交
2419
	int ret;
2420 2421
	u64 len = olen;
	u64 bs = root->fs_info->sb->s_blocksize;
2422

2423 2424 2425 2426 2427 2428 2429 2430 2431 2432
	/*
	 * 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)?
	 */

2433
	/* the destination must be opened for writing */
2434
	if (!(file->f_mode & FMODE_WRITE) || (file->f_flags & O_APPEND))
2435 2436
		return -EINVAL;

L
Li Zefan 已提交
2437 2438 2439
	if (btrfs_root_readonly(root))
		return -EROFS;

2440
	ret = mnt_want_write_file(file);
Y
Yan Zheng 已提交
2441 2442 2443
	if (ret)
		return ret;

2444 2445
	src_file = fdget(srcfd);
	if (!src_file.file) {
2446 2447 2448
		ret = -EBADF;
		goto out_drop_write;
	}
2449

2450
	ret = -EXDEV;
2451
	if (src_file.file->f_path.mnt != file->f_path.mnt)
2452 2453
		goto out_fput;

2454
	src = src_file.file->f_dentry->d_inode;
C
Christoph Hellwig 已提交
2455

2456 2457 2458 2459
	ret = -EINVAL;
	if (src == inode)
		goto out_fput;

2460
	/* the src must be open for reading */
2461
	if (!(src_file.file->f_mode & FMODE_READ))
2462 2463
		goto out_fput;

2464 2465 2466 2467 2468
	/* 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
2469 2470 2471 2472
	ret = -EISDIR;
	if (S_ISDIR(src->i_mode) || S_ISDIR(inode->i_mode))
		goto out_fput;

C
Christoph Hellwig 已提交
2473
	ret = -EXDEV;
2474
	if (src->i_sb != inode->i_sb)
Y
Yan Zheng 已提交
2475 2476 2477 2478 2479 2480 2481 2482 2483 2484
		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
2485
		goto out_fput;
Y
Yan Zheng 已提交
2486 2487
	}
	path->reada = 2;
C
Christoph Hellwig 已提交
2488 2489

	if (inode < src) {
2490 2491
		mutex_lock_nested(&inode->i_mutex, I_MUTEX_PARENT);
		mutex_lock_nested(&src->i_mutex, I_MUTEX_CHILD);
C
Christoph Hellwig 已提交
2492
	} else {
2493 2494
		mutex_lock_nested(&src->i_mutex, I_MUTEX_PARENT);
		mutex_lock_nested(&inode->i_mutex, I_MUTEX_CHILD);
C
Christoph Hellwig 已提交
2495 2496
	}

2497 2498
	/* determine range to clone */
	ret = -EINVAL;
2499
	if (off + len > src->i_size || off + len < off)
C
Christoph Hellwig 已提交
2500
		goto out_unlock;
2501 2502 2503 2504
	if (len == 0)
		olen = len = src->i_size - off;
	/* if we extend to eof, continue to block boundary */
	if (off + len == src->i_size)
2505
		len = ALIGN(src->i_size, bs) - off;
2506 2507

	/* verify the end result is block aligned */
2508 2509
	if (!IS_ALIGNED(off, bs) || !IS_ALIGNED(off + len, bs) ||
	    !IS_ALIGNED(destoff, bs))
2510 2511
		goto out_unlock;

2512 2513 2514 2515 2516 2517
	if (destoff > inode->i_size) {
		ret = btrfs_cont_expand(inode, inode->i_size, destoff);
		if (ret)
			goto out_unlock;
	}

2518 2519 2520 2521
	/* 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
2522 2523 2524
	/* do any pending delalloc/csum calc on src, one way or
	   another, and lock file content */
	while (1) {
Z
Zheng Yan 已提交
2525
		struct btrfs_ordered_extent *ordered;
L
Liu Bo 已提交
2526 2527
		lock_extent(&BTRFS_I(src)->io_tree, off, off + len - 1);
		ordered = btrfs_lookup_first_ordered_extent(src, off + len - 1);
2528
		if (!ordered &&
L
Liu Bo 已提交
2529 2530
		    !test_range_bit(&BTRFS_I(src)->io_tree, off, off + len - 1,
				    EXTENT_DELALLOC, 0, NULL))
C
Christoph Hellwig 已提交
2531
			break;
L
Liu Bo 已提交
2532
		unlock_extent(&BTRFS_I(src)->io_tree, off, off + len - 1);
Y
Yan Zheng 已提交
2533 2534
		if (ordered)
			btrfs_put_ordered_extent(ordered);
2535
		btrfs_wait_ordered_range(src, off, len);
C
Christoph Hellwig 已提交
2536 2537
	}

2538
	/* clone data */
L
Li Zefan 已提交
2539
	key.objectid = btrfs_ino(src);
Y
Yan Zheng 已提交
2540 2541
	key.type = BTRFS_EXTENT_DATA_KEY;
	key.offset = 0;
C
Christoph Hellwig 已提交
2542 2543 2544 2545 2546 2547

	while (1) {
		/*
		 * note the key will change type as we walk through the
		 * tree.
		 */
2548 2549
		ret = btrfs_search_slot(NULL, BTRFS_I(src)->root, &key, path,
				0, 0);
C
Christoph Hellwig 已提交
2550 2551 2552
		if (ret < 0)
			goto out;

Y
Yan Zheng 已提交
2553 2554
		nritems = btrfs_header_nritems(path->nodes[0]);
		if (path->slots[0] >= nritems) {
2555
			ret = btrfs_next_leaf(BTRFS_I(src)->root, path);
C
Christoph Hellwig 已提交
2556 2557 2558 2559
			if (ret < 0)
				goto out;
			if (ret > 0)
				break;
Y
Yan Zheng 已提交
2560
			nritems = btrfs_header_nritems(path->nodes[0]);
C
Christoph Hellwig 已提交
2561 2562 2563 2564
		}
		leaf = path->nodes[0];
		slot = path->slots[0];

Y
Yan Zheng 已提交
2565
		btrfs_item_key_to_cpu(leaf, &key, slot);
2566
		if (btrfs_key_type(&key) > BTRFS_EXTENT_DATA_KEY ||
L
Li Zefan 已提交
2567
		    key.objectid != btrfs_ino(src))
C
Christoph Hellwig 已提交
2568 2569
			break;

2570 2571 2572
		if (btrfs_key_type(&key) == BTRFS_EXTENT_DATA_KEY) {
			struct btrfs_file_extent_item *extent;
			int type;
Z
Zheng Yan 已提交
2573 2574
			u32 size;
			struct btrfs_key new_key;
2575 2576 2577
			u64 disko = 0, diskl = 0;
			u64 datao = 0, datal = 0;
			u8 comp;
2578
			u64 endoff;
Z
Zheng Yan 已提交
2579 2580 2581 2582 2583

			size = btrfs_item_size_nr(leaf, slot);
			read_extent_buffer(leaf, buf,
					   btrfs_item_ptr_offset(leaf, slot),
					   size);
2584 2585 2586 2587 2588

			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);
2589 2590
			if (type == BTRFS_FILE_EXTENT_REG ||
			    type == BTRFS_FILE_EXTENT_PREALLOC) {
C
Chris Mason 已提交
2591 2592 2593 2594
				disko = btrfs_file_extent_disk_bytenr(leaf,
								      extent);
				diskl = btrfs_file_extent_disk_num_bytes(leaf,
								 extent);
2595
				datao = btrfs_file_extent_offset(leaf, extent);
C
Chris Mason 已提交
2596 2597
				datal = btrfs_file_extent_num_bytes(leaf,
								    extent);
2598 2599 2600 2601 2602
			} else if (type == BTRFS_FILE_EXTENT_INLINE) {
				/* take upper bound, may be compressed */
				datal = btrfs_file_extent_ram_bytes(leaf,
								    extent);
			}
2603
			btrfs_release_path(path);
Z
Zheng Yan 已提交
2604

2605
			if (key.offset + datal <= off ||
L
Liu Bo 已提交
2606
			    key.offset >= off + len - 1)
2607 2608
				goto next;

Z
Zheng Yan 已提交
2609
			memcpy(&new_key, &key, sizeof(new_key));
L
Li Zefan 已提交
2610
			new_key.objectid = btrfs_ino(inode);
2611 2612 2613 2614
			if (off <= key.offset)
				new_key.offset = key.offset + destoff - off;
			else
				new_key.offset = destoff;
Z
Zheng Yan 已提交
2615

2616 2617 2618 2619 2620 2621
			/*
			 * 1 - adjusting old extent (we may have to split it)
			 * 1 - add new extent
			 * 1 - inode update
			 */
			trans = btrfs_start_transaction(root, 3);
2622 2623 2624 2625 2626
			if (IS_ERR(trans)) {
				ret = PTR_ERR(trans);
				goto out;
			}

2627 2628
			if (type == BTRFS_FILE_EXTENT_REG ||
			    type == BTRFS_FILE_EXTENT_PREALLOC) {
2629 2630 2631 2632 2633 2634 2635 2636 2637 2638
				/*
				 *    a  | --- range to clone ---|  b
				 * | ------------- extent ------------- |
				 */

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

				/* substract range a */
2639 2640 2641 2642 2643
				if (off > key.offset) {
					datao += off - key.offset;
					datal -= off - key.offset;
				}

J
Josef Bacik 已提交
2644
				ret = btrfs_drop_extents(trans, root, inode,
2645 2646
							 new_key.offset,
							 new_key.offset + datal,
2647
							 1);
2648 2649 2650 2651 2652 2653
				if (ret) {
					btrfs_abort_transaction(trans, root,
								ret);
					btrfs_end_transaction(trans, root);
					goto out;
				}
2654

2655 2656
				ret = btrfs_insert_empty_item(trans, root, path,
							      &new_key, size);
2657 2658 2659 2660 2661 2662
				if (ret) {
					btrfs_abort_transaction(trans, root,
								ret);
					btrfs_end_transaction(trans, root);
					goto out;
				}
2663 2664 2665 2666

				leaf = path->nodes[0];
				slot = path->slots[0];
				write_extent_buffer(leaf, buf,
Z
Zheng Yan 已提交
2667 2668
					    btrfs_item_ptr_offset(leaf, slot),
					    size);
Y
Yan Zheng 已提交
2669

2670
				extent = btrfs_item_ptr(leaf, slot,
C
Christoph Hellwig 已提交
2671
						struct btrfs_file_extent_item);
2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682

				/* 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
2683
					ret = btrfs_inc_extent_ref(trans, root,
2684 2685
							disko, diskl, 0,
							root->root_key.objectid,
L
Li Zefan 已提交
2686
							btrfs_ino(inode),
A
Arne Jansen 已提交
2687 2688
							new_key.offset - datao,
							0);
2689 2690 2691 2692 2693 2694 2695 2696 2697
					if (ret) {
						btrfs_abort_transaction(trans,
									root,
									ret);
						btrfs_end_transaction(trans,
								      root);
						goto out;

					}
C
Christoph Hellwig 已提交
2698
				}
2699 2700 2701 2702 2703 2704 2705
			} 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
2706

L
Liu Bo 已提交
2707 2708
				if (key.offset + datal > off + len)
					trim = key.offset + datal - (off + len);
C
Chris Mason 已提交
2709

2710 2711
				if (comp && (skip || trim)) {
					ret = -EINVAL;
2712
					btrfs_end_transaction(trans, root);
2713 2714 2715 2716
					goto out;
				}
				size -= skip + trim;
				datal -= skip + trim;
2717

J
Josef Bacik 已提交
2718
				ret = btrfs_drop_extents(trans, root, inode,
2719 2720
							 new_key.offset,
							 new_key.offset + datal,
2721
							 1);
2722 2723 2724 2725 2726 2727
				if (ret) {
					btrfs_abort_transaction(trans, root,
								ret);
					btrfs_end_transaction(trans, root);
					goto out;
				}
2728

2729 2730
				ret = btrfs_insert_empty_item(trans, root, path,
							      &new_key, size);
2731 2732 2733 2734 2735 2736
				if (ret) {
					btrfs_abort_transaction(trans, root,
								ret);
					btrfs_end_transaction(trans, root);
					goto out;
				}
2737 2738

				if (skip) {
C
Chris Mason 已提交
2739 2740
					u32 start =
					  btrfs_file_extent_calc_inline_size(0);
2741 2742 2743 2744 2745 2746 2747 2748 2749 2750
					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
2751
			}
2752 2753

			btrfs_mark_buffer_dirty(leaf);
2754
			btrfs_release_path(path);
2755

2756
			inode_inc_iversion(inode);
2757
			inode->i_mtime = inode->i_ctime = CURRENT_TIME;
2758 2759 2760 2761 2762 2763 2764

			/*
			 * 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;
2765 2766
			if (endoff > destoff+olen)
				endoff = destoff+olen;
2767 2768 2769
			if (endoff > inode->i_size)
				btrfs_i_size_write(inode, endoff);

2770
			ret = btrfs_update_inode(trans, root, inode);
2771 2772 2773 2774 2775 2776
			if (ret) {
				btrfs_abort_transaction(trans, root, ret);
				btrfs_end_transaction(trans, root);
				goto out;
			}
			ret = btrfs_end_transaction(trans, root);
2777
		}
C
Chris Mason 已提交
2778
next:
2779
		btrfs_release_path(path);
C
Christoph Hellwig 已提交
2780 2781 2782 2783
		key.offset++;
	}
	ret = 0;
out:
2784
	btrfs_release_path(path);
L
Liu Bo 已提交
2785
	unlock_extent(&BTRFS_I(src)->io_tree, off, off + len - 1);
C
Christoph Hellwig 已提交
2786 2787 2788
out_unlock:
	mutex_unlock(&src->i_mutex);
	mutex_unlock(&inode->i_mutex);
Y
Yan Zheng 已提交
2789 2790
	vfree(buf);
	btrfs_free_path(path);
C
Christoph Hellwig 已提交
2791
out_fput:
2792
	fdput(src_file);
2793
out_drop_write:
A
Al Viro 已提交
2794
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
2795 2796 2797
	return ret;
}

2798
static long btrfs_ioctl_clone_range(struct file *file, void __user *argp)
2799 2800 2801
{
	struct btrfs_ioctl_clone_range_args args;

2802
	if (copy_from_user(&args, argp, sizeof(args)))
2803 2804 2805 2806 2807
		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
2808 2809 2810 2811 2812 2813
/*
 * 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.
 */
2814
static long btrfs_ioctl_trans_start(struct file *file)
C
Christoph Hellwig 已提交
2815 2816 2817 2818
{
	struct inode *inode = fdentry(file)->d_inode;
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;
2819
	int ret;
C
Christoph Hellwig 已提交
2820

2821
	ret = -EPERM;
2822
	if (!capable(CAP_SYS_ADMIN))
2823
		goto out;
2824

2825 2826
	ret = -EINPROGRESS;
	if (file->private_data)
C
Christoph Hellwig 已提交
2827
		goto out;
2828

L
Li Zefan 已提交
2829 2830 2831 2832
	ret = -EROFS;
	if (btrfs_root_readonly(root))
		goto out;

2833
	ret = mnt_want_write_file(file);
Y
Yan Zheng 已提交
2834 2835 2836
	if (ret)
		goto out;

J
Josef Bacik 已提交
2837
	atomic_inc(&root->fs_info->open_ioctl_trans);
2838

2839
	ret = -ENOMEM;
2840
	trans = btrfs_start_ioctl_transaction(root);
2841
	if (IS_ERR(trans))
2842 2843 2844 2845 2846 2847
		goto out_drop;

	file->private_data = trans;
	return 0;

out_drop:
J
Josef Bacik 已提交
2848
	atomic_dec(&root->fs_info->open_ioctl_trans);
A
Al Viro 已提交
2849
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
2850 2851 2852 2853
out:
	return ret;
}

2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864 2865
static long btrfs_ioctl_default_subvol(struct file *file, void __user *argp)
{
	struct inode *inode = fdentry(file)->d_inode;
	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;
2866
	int ret;
2867 2868 2869 2870

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

2871 2872 2873 2874 2875 2876 2877 2878
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;

	if (copy_from_user(&objectid, argp, sizeof(objectid))) {
		ret = -EFAULT;
		goto out;
	}
2879 2880 2881 2882 2883 2884 2885 2886 2887

	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);
2888 2889 2890 2891
	if (IS_ERR(new_root)) {
		ret = PTR_ERR(new_root);
		goto out;
	}
2892

2893 2894 2895 2896
	if (btrfs_root_refs(&new_root->root_item) == 0) {
		ret = -ENOENT;
		goto out;
	}
2897 2898

	path = btrfs_alloc_path();
2899 2900 2901 2902
	if (!path) {
		ret = -ENOMEM;
		goto out;
	}
2903 2904 2905
	path->leave_spinning = 1;

	trans = btrfs_start_transaction(root, 1);
2906
	if (IS_ERR(trans)) {
2907
		btrfs_free_path(path);
2908 2909
		ret = PTR_ERR(trans);
		goto out;
2910 2911
	}

2912
	dir_id = btrfs_super_root_dir(root->fs_info->super_copy);
2913 2914
	di = btrfs_lookup_dir_item(trans, root->fs_info->tree_root, path,
				   dir_id, "default", 7, 1);
2915
	if (IS_ERR_OR_NULL(di)) {
2916 2917 2918 2919
		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");
2920 2921
		ret = -ENOENT;
		goto out;
2922 2923 2924 2925 2926 2927 2928
	}

	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);

2929
	btrfs_set_fs_incompat(root->fs_info, DEFAULT_SUBVOL);
2930
	btrfs_end_transaction(trans, root);
2931 2932 2933
out:
	mnt_drop_write_file(file);
	return ret;
2934 2935
}

2936 2937
void btrfs_get_block_group_info(struct list_head *groups_list,
				struct btrfs_ioctl_space_info *space)
2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951
{
	struct btrfs_block_group_cache *block_group;

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

J
Josef Bacik 已提交
2952 2953 2954 2955 2956
long btrfs_ioctl_space_info(struct btrfs_root *root, void __user *arg)
{
	struct btrfs_ioctl_space_args space_args;
	struct btrfs_ioctl_space_info space;
	struct btrfs_ioctl_space_info *dest;
2957
	struct btrfs_ioctl_space_info *dest_orig;
2958
	struct btrfs_ioctl_space_info __user *user_dest;
J
Josef Bacik 已提交
2959
	struct btrfs_space_info *info;
2960 2961 2962 2963 2964
	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;
2965
	int alloc_size;
J
Josef Bacik 已提交
2966
	int ret = 0;
2967
	u64 slot_count = 0;
2968
	int i, c;
J
Josef Bacik 已提交
2969 2970 2971 2972 2973 2974

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

2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998
	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);
	}
2999 3000 3001 3002 3003 3004

	/* space_slots == 0 means they are asking for a count */
	if (space_args.space_slots == 0) {
		space_args.total_spaces = slot_count;
		goto out;
	}
3005

3006
	slot_count = min_t(u64, space_args.space_slots, slot_count);
3007

3008
	alloc_size = sizeof(*dest) * slot_count;
3009

3010 3011 3012 3013 3014 3015
	/* 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
3016
	space_args.total_spaces = 0;
3017 3018 3019 3020
	dest = kmalloc(alloc_size, GFP_NOFS);
	if (!dest)
		return -ENOMEM;
	dest_orig = dest;
J
Josef Bacik 已提交
3021

3022
	/* now we have a buffer to copy into */
3023 3024 3025
	for (i = 0; i < num_types; i++) {
		struct btrfs_space_info *tmp;

3026 3027 3028
		if (!slot_count)
			break;

3029 3030 3031 3032 3033 3034 3035 3036 3037 3038
		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();
3039

3040 3041 3042 3043 3044
		if (!info)
			continue;
		down_read(&info->groups_sem);
		for (c = 0; c < BTRFS_NR_RAID_TYPES; c++) {
			if (!list_empty(&info->block_groups[c])) {
3045 3046
				btrfs_get_block_group_info(
					&info->block_groups[c], &space);
3047 3048 3049
				memcpy(dest, &space, sizeof(space));
				dest++;
				space_args.total_spaces++;
3050
				slot_count--;
3051
			}
3052 3053
			if (!slot_count)
				break;
3054 3055
		}
		up_read(&info->groups_sem);
J
Josef Bacik 已提交
3056 3057
	}

D
Daniel J Blueman 已提交
3058
	user_dest = (struct btrfs_ioctl_space_info __user *)
3059 3060 3061 3062 3063 3064 3065 3066
		(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
3067 3068 3069 3070 3071
		ret = -EFAULT;

	return ret;
}

C
Christoph Hellwig 已提交
3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083 3084
/*
 * 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)
{
	struct inode *inode = fdentry(file)->d_inode;
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;

	trans = file->private_data;
3085 3086
	if (!trans)
		return -EINVAL;
C
Christoph Hellwig 已提交
3087
	file->private_data = NULL;
3088

3089 3090
	btrfs_end_transaction(trans, root);

J
Josef Bacik 已提交
3091
	atomic_dec(&root->fs_info->open_ioctl_trans);
3092

A
Al Viro 已提交
3093
	mnt_drop_write_file(file);
3094
	return 0;
C
Christoph Hellwig 已提交
3095 3096
}

3097 3098
static noinline long btrfs_ioctl_start_sync(struct btrfs_root *root,
					    void __user *argp)
3099 3100 3101
{
	struct btrfs_trans_handle *trans;
	u64 transid;
T
Tsutomu Itoh 已提交
3102
	int ret;
3103

3104 3105 3106 3107 3108 3109 3110 3111 3112
	trans = btrfs_attach_transaction(root);
	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;
	}
3113
	transid = trans->transid;
T
Tsutomu Itoh 已提交
3114
	ret = btrfs_commit_transaction_async(trans, root, 0);
3115 3116
	if (ret) {
		btrfs_end_transaction(trans, root);
T
Tsutomu Itoh 已提交
3117
		return ret;
3118
	}
3119
out:
3120 3121 3122 3123 3124 3125
	if (argp)
		if (copy_to_user(argp, &transid, sizeof(transid)))
			return -EFAULT;
	return 0;
}

3126 3127
static noinline long btrfs_ioctl_wait_sync(struct btrfs_root *root,
					   void __user *argp)
3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139
{
	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
3140
static long btrfs_ioctl_scrub(struct file *file, void __user *arg)
J
Jan Schmidt 已提交
3141
{
M
Miao Xie 已提交
3142
	struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;
J
Jan Schmidt 已提交
3143
	struct btrfs_ioctl_scrub_args *sa;
M
Miao Xie 已提交
3144
	int ret;
J
Jan Schmidt 已提交
3145 3146 3147 3148 3149 3150 3151 3152

	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
3153 3154 3155 3156 3157 3158
	if (!(sa->flags & BTRFS_SCRUB_READONLY)) {
		ret = mnt_want_write_file(file);
		if (ret)
			goto out;
	}

3159
	ret = btrfs_scrub_dev(root->fs_info, sa->devid, sa->start, sa->end,
3160 3161
			      &sa->progress, sa->flags & BTRFS_SCRUB_READONLY,
			      0);
J
Jan Schmidt 已提交
3162 3163 3164 3165

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

M
Miao Xie 已提交
3166 3167 3168
	if (!(sa->flags & BTRFS_SCRUB_READONLY))
		mnt_drop_write_file(file);
out:
J
Jan Schmidt 已提交
3169 3170 3171 3172 3173 3174 3175 3176 3177
	kfree(sa);
	return ret;
}

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

3178
	return btrfs_scrub_cancel(root->fs_info);
J
Jan Schmidt 已提交
3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202
}

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;
}

3203
static long btrfs_ioctl_get_dev_stats(struct btrfs_root *root,
3204
				      void __user *arg)
3205 3206 3207 3208 3209 3210 3211 3212
{
	struct btrfs_ioctl_get_dev_stats *sa;
	int ret;

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

3213 3214 3215 3216 3217 3218
	if ((sa->flags & BTRFS_DEV_STATS_RESET) && !capable(CAP_SYS_ADMIN)) {
		kfree(sa);
		return -EPERM;
	}

	ret = btrfs_get_dev_stats(root, sa);
3219 3220 3221 3222 3223 3224 3225 3226

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

	kfree(sa);
	return ret;
}

3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271
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;
}

3272 3273 3274 3275
static long btrfs_ioctl_ino_to_path(struct btrfs_root *root, void __user *arg)
{
	int ret = 0;
	int i;
3276
	u64 rel_ptr;
3277
	int size;
3278
	struct btrfs_ioctl_ino_path_args *ipa = NULL;
3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310
	struct inode_fs_paths *ipath = NULL;
	struct btrfs_path *path;

	if (!capable(CAP_SYS_ADMIN))
		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) {
3311 3312
		rel_ptr = ipath->fspath->val[i] -
			  (u64)(unsigned long)ipath->fspath->val;
3313
		ipath->fspath->val[i] = rel_ptr;
3314 3315
	}

3316 3317
	ret = copy_to_user((void *)(unsigned long)ipa->fspath,
			   (void *)(unsigned long)ipath->fspath, size);
3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375
	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;
	}

3376
	size = min_t(u32, loi->size, 64 * 1024);
3377 3378 3379 3380 3381 3382 3383
	inodes = init_data_container(size);
	if (IS_ERR(inodes)) {
		ret = PTR_ERR(inodes);
		inodes = NULL;
		goto out;
	}

L
Liu Bo 已提交
3384 3385 3386
	ret = iterate_inodes_from_logical(loi->logical, root->fs_info, path,
					  build_ino_list, inodes);
	if (ret == -EINVAL)
3387 3388 3389 3390
		ret = -ENOENT;
	if (ret < 0)
		goto out;

3391 3392
	ret = copy_to_user((void *)(unsigned long)loi->inodes,
			   (void *)(unsigned long)inodes, size);
3393 3394 3395 3396 3397
	if (ret)
		ret = -EFAULT;

out:
	btrfs_free_path(path);
3398
	vfree(inodes);
3399 3400 3401 3402 3403
	kfree(loi);

	return ret;
}

3404
void update_ioctl_balance_args(struct btrfs_fs_info *fs_info, int lock,
3405 3406 3407 3408 3409 3410
			       struct btrfs_ioctl_balance_args *bargs)
{
	struct btrfs_balance_control *bctl = fs_info->balance_ctl;

	bargs->flags = bctl->flags;

3411 3412 3413 3414
	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;
3415 3416
	if (atomic_read(&fs_info->balance_cancel_req))
		bargs->state |= BTRFS_BALANCE_STATE_CANCEL_REQ;
3417

3418 3419 3420
	memcpy(&bargs->data, &bctl->data, sizeof(bargs->data));
	memcpy(&bargs->meta, &bctl->meta, sizeof(bargs->meta));
	memcpy(&bargs->sys, &bctl->sys, sizeof(bargs->sys));
3421 3422 3423 3424 3425 3426 3427 3428

	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));
	}
3429 3430
}

3431
static long btrfs_ioctl_balance(struct file *file, void __user *arg)
3432
{
3433
	struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;
3434 3435 3436 3437
	struct btrfs_fs_info *fs_info = root->fs_info;
	struct btrfs_ioctl_balance_args *bargs;
	struct btrfs_balance_control *bctl;
	int ret;
3438
	int need_to_clear_lock = 0;
3439 3440 3441 3442

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

3443
	ret = mnt_want_write_file(file);
3444 3445 3446
	if (ret)
		return ret;

3447 3448 3449 3450 3451 3452 3453 3454 3455
	mutex_lock(&fs_info->volume_mutex);
	mutex_lock(&fs_info->balance_mutex);

	if (arg) {
		bargs = memdup_user(arg, sizeof(*bargs));
		if (IS_ERR(bargs)) {
			ret = PTR_ERR(bargs);
			goto out;
		}
3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469

		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;
		}
3470 3471 3472 3473
	} else {
		bargs = NULL;
	}

3474 3475 3476
	if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
			1)) {
		pr_info("btrfs: dev add/delete/balance/replace/resize operation in progress\n");
3477 3478 3479
		ret = -EINPROGRESS;
		goto out_bargs;
	}
3480
	need_to_clear_lock = 1;
3481

3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494
	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;
3495 3496 3497
	} else {
		/* balance everything - no filters */
		bctl->flags |= BTRFS_BALANCE_TYPE_MASK;
3498 3499
	}

3500
do_balance:
3501 3502
	ret = btrfs_balance(bctl, bargs);
	/*
3503 3504
	 * bctl is freed in __cancel_balance or in free_fs_info if
	 * restriper was paused all the way until unmount
3505 3506 3507 3508 3509 3510 3511 3512 3513
	 */
	if (arg) {
		if (copy_to_user(arg, bargs, sizeof(*bargs)))
			ret = -EFAULT;
	}

out_bargs:
	kfree(bargs);
out:
3514 3515 3516
	if (need_to_clear_lock)
		atomic_set(&root->fs_info->mutually_exclusive_operation_running,
			   0);
3517 3518
	mutex_unlock(&fs_info->balance_mutex);
	mutex_unlock(&fs_info->volume_mutex);
3519
	mnt_drop_write_file(file);
3520 3521 3522
	return ret;
}

3523 3524 3525 3526 3527 3528 3529 3530
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);
3531 3532
	case BTRFS_BALANCE_CTL_CANCEL:
		return btrfs_cancel_balance(root->fs_info);
3533 3534 3535 3536 3537
	}

	return -EINVAL;
}

3538 3539 3540 3541 3542 3543 3544 3545 3546 3547 3548 3549 3550 3551 3552 3553 3554 3555 3556 3557 3558 3559 3560 3561 3562 3563 3564 3565 3566 3567 3568 3569 3570
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;
}

3571
static long btrfs_ioctl_quota_ctl(struct file *file, void __user *arg)
A
Arne Jansen 已提交
3572
{
3573
	struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;
A
Arne Jansen 已提交
3574 3575 3576 3577 3578 3579 3580 3581
	struct btrfs_ioctl_quota_ctl_args *sa;
	struct btrfs_trans_handle *trans = NULL;
	int ret;
	int err;

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

3582 3583 3584
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
3585 3586

	sa = memdup_user(arg, sizeof(*sa));
3587 3588 3589 3590
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
3591 3592 3593 3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624

	if (sa->cmd != BTRFS_QUOTA_CTL_RESCAN) {
		trans = btrfs_start_transaction(root, 2);
		if (IS_ERR(trans)) {
			ret = PTR_ERR(trans);
			goto out;
		}
	}

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

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

	if (trans) {
		err = btrfs_commit_transaction(trans, root);
		if (err && !ret)
			ret = err;
	}
out:
	kfree(sa);
3625 3626
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
3627 3628 3629
	return ret;
}

3630
static long btrfs_ioctl_qgroup_assign(struct file *file, void __user *arg)
A
Arne Jansen 已提交
3631
{
3632
	struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;
A
Arne Jansen 已提交
3633 3634 3635 3636 3637 3638 3639 3640
	struct btrfs_ioctl_qgroup_assign_args *sa;
	struct btrfs_trans_handle *trans;
	int ret;
	int err;

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

3641 3642 3643
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
3644 3645

	sa = memdup_user(arg, sizeof(*sa));
3646 3647 3648 3649
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 3667 3668 3669 3670 3671

	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);
3672 3673
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
3674 3675 3676
	return ret;
}

3677
static long btrfs_ioctl_qgroup_create(struct file *file, void __user *arg)
A
Arne Jansen 已提交
3678
{
3679
	struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;
A
Arne Jansen 已提交
3680 3681 3682 3683 3684 3685 3686 3687
	struct btrfs_ioctl_qgroup_create_args *sa;
	struct btrfs_trans_handle *trans;
	int ret;
	int err;

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

3688 3689 3690
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
3691 3692

	sa = memdup_user(arg, sizeof(*sa));
3693 3694 3695 3696
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
3697 3698 3699 3700 3701 3702 3703 3704 3705 3706 3707 3708 3709 3710 3711 3712 3713 3714 3715 3716 3717

	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);
3718 3719
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
3720 3721 3722
	return ret;
}

3723
static long btrfs_ioctl_qgroup_limit(struct file *file, void __user *arg)
A
Arne Jansen 已提交
3724
{
3725
	struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;
A
Arne Jansen 已提交
3726 3727 3728 3729 3730 3731 3732 3733 3734
	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;

3735 3736 3737
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
3738 3739

	sa = memdup_user(arg, sizeof(*sa));
3740 3741 3742 3743
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
3744 3745 3746 3747 3748 3749 3750 3751 3752 3753 3754 3755 3756 3757 3758 3759 3760 3761 3762 3763 3764 3765

	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);
3766 3767
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
3768 3769 3770
	return ret;
}

3771 3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786 3787 3788 3789 3790 3791 3792 3793 3794 3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 3806 3807 3808 3809 3810 3811 3812 3813 3814 3815 3816 3817 3818 3819 3820 3821 3822 3823 3824 3825 3826 3827 3828 3829 3830 3831 3832 3833 3834 3835 3836 3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847 3848 3849 3850 3851
static long btrfs_ioctl_set_received_subvol(struct file *file,
					    void __user *arg)
{
	struct btrfs_ioctl_received_subvol_args *sa = NULL;
	struct inode *inode = fdentry(file)->d_inode;
	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;
}

C
Christoph Hellwig 已提交
3852 3853 3854 3855
long btrfs_ioctl(struct file *file, unsigned int
		cmd, unsigned long arg)
{
	struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;
3856
	void __user *argp = (void __user *)arg;
C
Christoph Hellwig 已提交
3857 3858

	switch (cmd) {
3859 3860 3861 3862 3863 3864
	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);
3865 3866
	case FITRIM:
		return btrfs_ioctl_fitrim(file, argp);
C
Christoph Hellwig 已提交
3867
	case BTRFS_IOC_SNAP_CREATE:
3868
		return btrfs_ioctl_snap_create(file, argp, 0);
3869
	case BTRFS_IOC_SNAP_CREATE_V2:
3870
		return btrfs_ioctl_snap_create_v2(file, argp, 0);
3871
	case BTRFS_IOC_SUBVOL_CREATE:
3872
		return btrfs_ioctl_snap_create(file, argp, 1);
A
Arne Jansen 已提交
3873 3874
	case BTRFS_IOC_SUBVOL_CREATE_V2:
		return btrfs_ioctl_snap_create_v2(file, argp, 1);
3875 3876
	case BTRFS_IOC_SNAP_DESTROY:
		return btrfs_ioctl_snap_destroy(file, argp);
3877 3878 3879 3880
	case BTRFS_IOC_SUBVOL_GETFLAGS:
		return btrfs_ioctl_subvol_getflags(file, argp);
	case BTRFS_IOC_SUBVOL_SETFLAGS:
		return btrfs_ioctl_subvol_setflags(file, argp);
3881 3882
	case BTRFS_IOC_DEFAULT_SUBVOL:
		return btrfs_ioctl_default_subvol(file, argp);
C
Christoph Hellwig 已提交
3883
	case BTRFS_IOC_DEFRAG:
C
Chris Mason 已提交
3884 3885 3886
		return btrfs_ioctl_defrag(file, NULL);
	case BTRFS_IOC_DEFRAG_RANGE:
		return btrfs_ioctl_defrag(file, argp);
C
Christoph Hellwig 已提交
3887
	case BTRFS_IOC_RESIZE:
3888
		return btrfs_ioctl_resize(file, argp);
C
Christoph Hellwig 已提交
3889
	case BTRFS_IOC_ADD_DEV:
3890
		return btrfs_ioctl_add_dev(root, argp);
C
Christoph Hellwig 已提交
3891
	case BTRFS_IOC_RM_DEV:
3892
		return btrfs_ioctl_rm_dev(file, argp);
J
Jan Schmidt 已提交
3893 3894 3895 3896
	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
3897
	case BTRFS_IOC_BALANCE:
3898
		return btrfs_ioctl_balance(file, NULL);
C
Christoph Hellwig 已提交
3899
	case BTRFS_IOC_CLONE:
3900 3901
		return btrfs_ioctl_clone(file, arg, 0, 0, 0);
	case BTRFS_IOC_CLONE_RANGE:
3902
		return btrfs_ioctl_clone_range(file, argp);
C
Christoph Hellwig 已提交
3903 3904 3905 3906
	case BTRFS_IOC_TRANS_START:
		return btrfs_ioctl_trans_start(file);
	case BTRFS_IOC_TRANS_END:
		return btrfs_ioctl_trans_end(file);
3907 3908 3909 3910
	case BTRFS_IOC_TREE_SEARCH:
		return btrfs_ioctl_tree_search(file, argp);
	case BTRFS_IOC_INO_LOOKUP:
		return btrfs_ioctl_ino_lookup(file, argp);
3911 3912 3913 3914
	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
3915 3916
	case BTRFS_IOC_SPACE_INFO:
		return btrfs_ioctl_space_info(root, argp);
C
Christoph Hellwig 已提交
3917 3918 3919
	case BTRFS_IOC_SYNC:
		btrfs_sync_fs(file->f_dentry->d_sb, 1);
		return 0;
3920
	case BTRFS_IOC_START_SYNC:
3921
		return btrfs_ioctl_start_sync(root, argp);
3922
	case BTRFS_IOC_WAIT_SYNC:
3923
		return btrfs_ioctl_wait_sync(root, argp);
J
Jan Schmidt 已提交
3924
	case BTRFS_IOC_SCRUB:
M
Miao Xie 已提交
3925
		return btrfs_ioctl_scrub(file, argp);
J
Jan Schmidt 已提交
3926 3927 3928 3929
	case BTRFS_IOC_SCRUB_CANCEL:
		return btrfs_ioctl_scrub_cancel(root, argp);
	case BTRFS_IOC_SCRUB_PROGRESS:
		return btrfs_ioctl_scrub_progress(root, argp);
3930
	case BTRFS_IOC_BALANCE_V2:
3931
		return btrfs_ioctl_balance(file, argp);
3932 3933
	case BTRFS_IOC_BALANCE_CTL:
		return btrfs_ioctl_balance_ctl(root, arg);
3934 3935
	case BTRFS_IOC_BALANCE_PROGRESS:
		return btrfs_ioctl_balance_progress(root, argp);
3936 3937
	case BTRFS_IOC_SET_RECEIVED_SUBVOL:
		return btrfs_ioctl_set_received_subvol(file, argp);
3938 3939
	case BTRFS_IOC_SEND:
		return btrfs_ioctl_send(file, argp);
3940
	case BTRFS_IOC_GET_DEV_STATS:
3941
		return btrfs_ioctl_get_dev_stats(root, argp);
A
Arne Jansen 已提交
3942
	case BTRFS_IOC_QUOTA_CTL:
3943
		return btrfs_ioctl_quota_ctl(file, argp);
A
Arne Jansen 已提交
3944
	case BTRFS_IOC_QGROUP_ASSIGN:
3945
		return btrfs_ioctl_qgroup_assign(file, argp);
A
Arne Jansen 已提交
3946
	case BTRFS_IOC_QGROUP_CREATE:
3947
		return btrfs_ioctl_qgroup_create(file, argp);
A
Arne Jansen 已提交
3948
	case BTRFS_IOC_QGROUP_LIMIT:
3949
		return btrfs_ioctl_qgroup_limit(file, argp);
3950 3951
	case BTRFS_IOC_DEV_REPLACE:
		return btrfs_ioctl_dev_replace(root, argp);
C
Christoph Hellwig 已提交
3952 3953 3954 3955
	}

	return -ENOTTY;
}