ioctl.c 121.8 KB
Newer Older
C
Christoph Hellwig 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
/*
 * Copyright (C) 2007 Oracle.  All rights reserved.
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public
 * License v2 as published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * General Public License for more details.
 *
 * You should have received a copy of the GNU General Public
 * License along with this program; if not, write to the
 * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
 * Boston, MA 021110-1307, USA.
 */

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

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

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

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


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

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

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

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

	if ((flags & BTRFS_INODE_COMPRESS) && !(flags & BTRFS_INODE_NOCOMPRESS))
		iflags |= FS_COMPR_FL;
	else if (flags & BTRFS_INODE_NOCOMPRESS)
		iflags |= FS_NOCOMP_FL;
129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156

	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.
 *
157
 * Currently only the compression flags and the cow flags are inherited.
158 159 160
 */
void btrfs_inherit_iflags(struct inode *inode, struct inode *dir)
{
161 162 163 164 165 166
	unsigned int flags;

	if (!dir)
		return;

	flags = BTRFS_I(dir)->flags;
167

168 169 170 171 172 173 174 175
	if (flags & BTRFS_INODE_NOCOMPRESS) {
		BTRFS_I(inode)->flags &= ~BTRFS_INODE_COMPRESS;
		BTRFS_I(inode)->flags |= BTRFS_INODE_NOCOMPRESS;
	} else if (flags & BTRFS_INODE_COMPRESS) {
		BTRFS_I(inode)->flags &= ~BTRFS_INODE_NOCOMPRESS;
		BTRFS_I(inode)->flags |= BTRFS_INODE_COMPRESS;
	}

L
Liu Bo 已提交
176
	if (flags & BTRFS_INODE_NODATACOW) {
177
		BTRFS_I(inode)->flags |= BTRFS_INODE_NODATACOW;
L
Liu Bo 已提交
178 179 180
		if (S_ISREG(inode->i_mode))
			BTRFS_I(inode)->flags |= BTRFS_INODE_NODATASUM;
	}
181 182 183 184 185 186

	btrfs_update_iflags(inode);
}

static int btrfs_ioctl_getflags(struct file *file, void __user *arg)
{
A
Al Viro 已提交
187
	struct btrfs_inode *ip = BTRFS_I(file_inode(file));
188 189 190 191 192 193 194
	unsigned int flags = btrfs_flags_to_ioctl(ip->flags);

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

195 196 197 198 199
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
200 201
		      FS_NOCOMP_FL | FS_COMPR_FL |
		      FS_NOCOW_FL))
202 203 204 205 206 207 208 209
		return -EOPNOTSUPP;

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

	return 0;
}

210 211
static int btrfs_ioctl_setflags(struct file *file, void __user *arg)
{
A
Al Viro 已提交
212
	struct inode *inode = file_inode(file);
213 214 215 216 217
	struct btrfs_inode *ip = BTRFS_I(inode);
	struct btrfs_root *root = ip->root;
	struct btrfs_trans_handle *trans;
	unsigned int flags, oldflags;
	int ret;
218 219
	u64 ip_oldflags;
	unsigned int i_oldflags;
220
	umode_t mode;
221

222 223 224
	if (!inode_owner_or_capable(inode))
		return -EPERM;

L
Li Zefan 已提交
225 226 227
	if (btrfs_root_readonly(root))
		return -EROFS;

228 229 230
	if (copy_from_user(&flags, arg, sizeof(flags)))
		return -EFAULT;

231 232 233
	ret = check_flags(flags);
	if (ret)
		return ret;
C
Christoph Hellwig 已提交
234

235 236 237 238
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;

239 240
	mutex_lock(&inode->i_mutex);

241 242
	ip_oldflags = ip->flags;
	i_oldflags = inode->i_flags;
243
	mode = inode->i_mode;
244

245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277
	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;
278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302
	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;
		}
	}
303

304 305 306 307 308 309 310 311
	/*
	 * 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;
312 313 314 315

		ret = btrfs_set_prop(inode, "btrfs.compression", NULL, 0, 0);
		if (ret && ret != -ENODATA)
			goto out_drop;
316
	} else if (flags & FS_COMPR_FL) {
317 318
		const char *comp;

319 320
		ip->flags |= BTRFS_INODE_COMPRESS;
		ip->flags &= ~BTRFS_INODE_NOCOMPRESS;
321 322 323 324 325 326 327 328 329 330

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

L
Li Zefan 已提交
331 332
	} else {
		ip->flags &= ~(BTRFS_INODE_COMPRESS | BTRFS_INODE_NOCOMPRESS);
333
	}
334

335
	trans = btrfs_start_transaction(root, 1);
336 337 338 339
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto out_drop;
	}
340

341
	btrfs_update_iflags(inode);
342
	inode_inc_iversion(inode);
343
	inode->i_ctime = CURRENT_TIME;
344 345 346
	ret = btrfs_update_inode(trans, root, inode);

	btrfs_end_transaction(trans, root);
347 348 349 350 351
 out_drop:
	if (ret) {
		ip->flags = ip_oldflags;
		inode->i_flags = i_oldflags;
	}
352 353 354

 out_unlock:
	mutex_unlock(&inode->i_mutex);
355
	mnt_drop_write_file(file);
356
	return ret;
357 358 359 360
}

static int btrfs_ioctl_getversion(struct file *file, int __user *arg)
{
A
Al Viro 已提交
361
	struct inode *inode = file_inode(file);
362 363 364

	return put_user(inode->i_generation, arg);
}
C
Christoph Hellwig 已提交
365

366 367
static noinline int btrfs_ioctl_fitrim(struct file *file, void __user *arg)
{
A
Al Viro 已提交
368
	struct btrfs_fs_info *fs_info = btrfs_sb(file_inode(file)->i_sb);
369 370 371 372 373
	struct btrfs_device *device;
	struct request_queue *q;
	struct fstrim_range range;
	u64 minlen = ULLONG_MAX;
	u64 num_devices = 0;
374
	u64 total_bytes = btrfs_super_total_bytes(fs_info->super_copy);
375 376 377 378 379
	int ret;

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

380 381 382
	rcu_read_lock();
	list_for_each_entry_rcu(device, &fs_info->fs_devices->devices,
				dev_list) {
383 384 385 386 387 388 389 390 391
		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);
		}
	}
392
	rcu_read_unlock();
393

394 395 396 397
	if (!num_devices)
		return -EOPNOTSUPP;
	if (copy_from_user(&range, arg, sizeof(range)))
		return -EFAULT;
398 399
	if (range.start > total_bytes ||
	    range.len < fs_info->sb->s_blocksize)
400
		return -EINVAL;
401

402
	range.len = min(range.len, total_bytes - range.start);
403
	range.minlen = max(range.minlen, minlen);
404
	ret = btrfs_trim_fs(fs_info->tree_root, &range);
405 406 407 408 409 410 411 412 413
	if (ret < 0)
		return ret;

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

	return 0;
}

414 415
int btrfs_is_empty_uuid(u8 *uuid)
{
C
Chris Mason 已提交
416 417 418 419 420 421 422
	int i;

	for (i = 0; i < BTRFS_UUID_SIZE; i++) {
		if (uuid[i])
			return 0;
	}
	return 1;
423 424
}

425
static noinline int create_subvol(struct inode *dir,
426
				  struct dentry *dentry,
S
Sage Weil 已提交
427
				  char *name, int namelen,
A
Arne Jansen 已提交
428
				  u64 *async_transid,
429
				  struct btrfs_qgroup_inherit *inherit)
C
Christoph Hellwig 已提交
430 431 432 433 434 435
{
	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;
436
	struct btrfs_root *root = BTRFS_I(dir)->root;
437
	struct btrfs_root *new_root;
438
	struct btrfs_block_rsv block_rsv;
439
	struct timespec cur_time = CURRENT_TIME;
440
	struct inode *inode;
C
Christoph Hellwig 已提交
441 442 443 444
	int ret;
	int err;
	u64 objectid;
	u64 new_dirid = BTRFS_FIRST_FREE_OBJECTID;
445
	u64 index = 0;
446
	u64 qgroup_reserved;
447
	uuid_le new_uuid;
C
Christoph Hellwig 已提交
448

449
	ret = btrfs_find_free_objectid(root->fs_info->tree_root, &objectid);
450
	if (ret)
451
		return ret;
452

453
	btrfs_init_block_rsv(&block_rsv, BTRFS_BLOCK_RSV_TEMP);
J
Josef Bacik 已提交
454
	/*
455 456
	 * The same as the snapshot creation, please see the comment
	 * of create_snapshot().
J
Josef Bacik 已提交
457
	 */
458
	ret = btrfs_subvolume_reserve_metadata(root, &block_rsv,
459
					       8, &qgroup_reserved, false);
460 461 462 463 464 465
	if (ret)
		return ret;

	trans = btrfs_start_transaction(root, 0);
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
466 467 468
		btrfs_subvolume_release_metadata(root, &block_rsv,
						 qgroup_reserved);
		return ret;
469 470 471
	}
	trans->block_rsv = &block_rsv;
	trans->bytes_reserved = block_rsv.size;
C
Christoph Hellwig 已提交
472

473
	ret = btrfs_qgroup_inherit(trans, root->fs_info, 0, objectid, inherit);
A
Arne Jansen 已提交
474 475 476
	if (ret)
		goto fail;

477
	leaf = btrfs_alloc_free_block(trans, root, root->leafsize,
478
				      0, objectid, NULL, 0, 0, 0);
479 480 481 482
	if (IS_ERR(leaf)) {
		ret = PTR_ERR(leaf);
		goto fail;
	}
C
Christoph Hellwig 已提交
483

484
	memset_extent_buffer(leaf, 0, 0, sizeof(struct btrfs_header));
C
Christoph Hellwig 已提交
485 486
	btrfs_set_header_bytenr(leaf, leaf->start);
	btrfs_set_header_generation(leaf, trans->transid);
487
	btrfs_set_header_backref_rev(leaf, BTRFS_MIXED_BACKREF_REV);
C
Christoph Hellwig 已提交
488 489
	btrfs_set_header_owner(leaf, objectid);

490
	write_extent_buffer(leaf, root->fs_info->fsid, btrfs_header_fsid(),
C
Christoph Hellwig 已提交
491
			    BTRFS_FSID_SIZE);
492
	write_extent_buffer(leaf, root->fs_info->chunk_tree_uuid,
493
			    btrfs_header_chunk_tree_uuid(leaf),
494
			    BTRFS_UUID_SIZE);
C
Christoph Hellwig 已提交
495 496
	btrfs_mark_buffer_dirty(leaf);

497 498
	memset(&root_item, 0, sizeof(root_item));

C
Christoph Hellwig 已提交
499
	inode_item = &root_item.inode;
500 501 502 503 504
	btrfs_set_stack_inode_generation(inode_item, 1);
	btrfs_set_stack_inode_size(inode_item, 3);
	btrfs_set_stack_inode_nlink(inode_item, 1);
	btrfs_set_stack_inode_nbytes(inode_item, root->leafsize);
	btrfs_set_stack_inode_mode(inode_item, S_IFDIR | 0755);
C
Christoph Hellwig 已提交
505

506 507 508
	btrfs_set_root_flags(&root_item, 0);
	btrfs_set_root_limit(&root_item, 0);
	btrfs_set_stack_inode_flags(inode_item, BTRFS_INODE_ROOT_ITEM_INIT);
509

C
Christoph Hellwig 已提交
510
	btrfs_set_root_bytenr(&root_item, leaf->start);
511
	btrfs_set_root_generation(&root_item, trans->transid);
C
Christoph Hellwig 已提交
512 513
	btrfs_set_root_level(&root_item, 0);
	btrfs_set_root_refs(&root_item, 1);
514
	btrfs_set_root_used(&root_item, leaf->len);
Y
Yan Zheng 已提交
515
	btrfs_set_root_last_snapshot(&root_item, 0);
C
Christoph Hellwig 已提交
516

517 518 519 520
	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);
521 522
	btrfs_set_stack_timespec_sec(&root_item.otime, cur_time.tv_sec);
	btrfs_set_stack_timespec_nsec(&root_item.otime, cur_time.tv_nsec);
523 524 525
	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
526

527
	btrfs_tree_unlock(leaf);
C
Christoph Hellwig 已提交
528 529 530 531 532 533
	free_extent_buffer(leaf);
	leaf = NULL;

	btrfs_set_root_dirid(&root_item, new_dirid);

	key.objectid = objectid;
534
	key.offset = 0;
C
Christoph Hellwig 已提交
535 536 537 538 539 540
	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;

541 542
	key.offset = (u64)-1;
	new_root = btrfs_read_fs_root_no_name(root->fs_info, &key);
543 544 545 546 547
	if (IS_ERR(new_root)) {
		btrfs_abort_transaction(trans, root, PTR_ERR(new_root));
		ret = PTR_ERR(new_root);
		goto fail;
	}
548 549 550

	btrfs_record_root_in_trans(trans, new_root);

551
	ret = btrfs_create_subvol_root(trans, new_root, root, new_dirid);
552 553
	if (ret) {
		/* We potentially lose an unused inode item here */
554
		btrfs_abort_transaction(trans, root, ret);
555 556 557
		goto fail;
	}

C
Christoph Hellwig 已提交
558 559 560
	/*
	 * insert the directory item
	 */
561
	ret = btrfs_set_inode_index(dir, &index);
562 563 564 565
	if (ret) {
		btrfs_abort_transaction(trans, root, ret);
		goto fail;
	}
566 567

	ret = btrfs_insert_dir_item(trans, root,
568
				    name, namelen, dir, &key,
569
				    BTRFS_FT_DIR, index);
570 571
	if (ret) {
		btrfs_abort_transaction(trans, root, ret);
C
Christoph Hellwig 已提交
572
		goto fail;
573
	}
574

575 576 577 578
	btrfs_i_size_write(dir, dir->i_size + namelen * 2);
	ret = btrfs_update_inode(trans, root, dir);
	BUG_ON(ret);

579
	ret = btrfs_add_root_ref(trans, root->fs_info->tree_root,
580
				 objectid, root->root_key.objectid,
L
Li Zefan 已提交
581
				 btrfs_ino(dir), index, name, namelen);
582
	BUG_ON(ret);
C
Christoph Hellwig 已提交
583

584 585 586 587 588 589
	ret = btrfs_uuid_tree_add(trans, root->fs_info->uuid_root,
				  root_item.uuid, BTRFS_UUID_KEY_SUBVOL,
				  objectid);
	if (ret)
		btrfs_abort_transaction(trans, root, ret);

C
Christoph Hellwig 已提交
590
fail:
591 592
	trans->block_rsv = NULL;
	trans->bytes_reserved = 0;
593 594
	btrfs_subvolume_release_metadata(root, &block_rsv, qgroup_reserved);

S
Sage Weil 已提交
595 596 597
	if (async_transid) {
		*async_transid = trans->transid;
		err = btrfs_commit_transaction_async(trans, root, 1);
598 599
		if (err)
			err = btrfs_commit_transaction(trans, root);
S
Sage Weil 已提交
600 601 602
	} else {
		err = btrfs_commit_transaction(trans, root);
	}
C
Christoph Hellwig 已提交
603 604
	if (err && !ret)
		ret = err;
605

606 607
	if (!ret) {
		inode = btrfs_lookup_dentry(dir, dentry);
608 609
		if (IS_ERR(inode))
			return PTR_ERR(inode);
610 611
		d_instantiate(dentry, inode);
	}
C
Christoph Hellwig 已提交
612 613 614
	return ret;
}

615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631
static void btrfs_wait_nocow_write(struct btrfs_root *root)
{
	s64 writers;
	DEFINE_WAIT(wait);

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

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

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

632 633 634 635
static int create_snapshot(struct btrfs_root *root, struct inode *dir,
			   struct dentry *dentry, char *name, int namelen,
			   u64 *async_transid, bool readonly,
			   struct btrfs_qgroup_inherit *inherit)
C
Christoph Hellwig 已提交
636
{
637
	struct inode *inode;
C
Christoph Hellwig 已提交
638 639
	struct btrfs_pending_snapshot *pending_snapshot;
	struct btrfs_trans_handle *trans;
640
	int ret;
C
Christoph Hellwig 已提交
641

642
	if (!test_bit(BTRFS_ROOT_REF_COWS, &root->state))
C
Christoph Hellwig 已提交
643 644
		return -EINVAL;

645 646 647 648
	atomic_inc(&root->will_be_snapshoted);
	smp_mb__after_atomic_inc();
	btrfs_wait_nocow_write(root);

649 650
	ret = btrfs_start_delalloc_inodes(root, 0);
	if (ret)
651
		goto out;
652

653
	btrfs_wait_ordered_extents(root, -1);
654

655
	pending_snapshot = kzalloc(sizeof(*pending_snapshot), GFP_NOFS);
656 657 658 659
	if (!pending_snapshot) {
		ret = -ENOMEM;
		goto out;
	}
660

661 662
	btrfs_init_block_rsv(&pending_snapshot->block_rsv,
			     BTRFS_BLOCK_RSV_TEMP);
663 664 665 666 667 668
	/*
	 * 1 - parent dir inode
	 * 2 - dir entries
	 * 1 - root item
	 * 2 - root ref/backref
	 * 1 - root of snapshot
669
	 * 1 - UUID item
670 671
	 */
	ret = btrfs_subvolume_reserve_metadata(BTRFS_I(dir)->root,
672
					&pending_snapshot->block_rsv, 8,
673 674
					&pending_snapshot->qgroup_reserved,
					false);
675
	if (ret)
676
		goto free;
677

678
	pending_snapshot->dentry = dentry;
C
Christoph Hellwig 已提交
679
	pending_snapshot->root = root;
L
Li Zefan 已提交
680
	pending_snapshot->readonly = readonly;
681
	pending_snapshot->dir = dir;
682
	pending_snapshot->inherit = inherit;
683

684
	trans = btrfs_start_transaction(root, 0);
685 686 687 688 689
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto fail;
	}

690
	spin_lock(&root->fs_info->trans_lock);
C
Christoph Hellwig 已提交
691 692
	list_add(&pending_snapshot->list,
		 &trans->transaction->pending_snapshots);
693
	spin_unlock(&root->fs_info->trans_lock);
S
Sage Weil 已提交
694 695 696 697
	if (async_transid) {
		*async_transid = trans->transid;
		ret = btrfs_commit_transaction_async(trans,
				     root->fs_info->extent_root, 1);
698 699
		if (ret)
			ret = btrfs_commit_transaction(trans, root);
S
Sage Weil 已提交
700 701 702 703
	} else {
		ret = btrfs_commit_transaction(trans,
					       root->fs_info->extent_root);
	}
704
	if (ret)
705
		goto fail;
706 707 708 709 710

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

711 712 713
	ret = btrfs_orphan_cleanup(pending_snapshot->snap);
	if (ret)
		goto fail;
C
Christoph Hellwig 已提交
714

715
	inode = btrfs_lookup_dentry(dentry->d_parent->d_inode, dentry);
716 717 718 719
	if (IS_ERR(inode)) {
		ret = PTR_ERR(inode);
		goto fail;
	}
720

721 722 723
	d_instantiate(dentry, inode);
	ret = 0;
fail:
724 725 726
	btrfs_subvolume_release_metadata(BTRFS_I(dir)->root,
					 &pending_snapshot->block_rsv,
					 pending_snapshot->qgroup_reserved);
727
free:
728
	kfree(pending_snapshot);
729 730
out:
	atomic_dec(&root->will_be_snapshoted);
C
Christoph Hellwig 已提交
731 732 733
	return ret;
}

734 735 736 737 738 739
/*  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)
{
740
	kuid_t fsuid = current_fsuid();
741 742 743

	if (!(dir->i_mode & S_ISVTX))
		return 0;
744
	if (uid_eq(inode->i_uid, fsuid))
745
		return 0;
746
	if (uid_eq(dir->i_uid, fsuid))
747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770
		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().
 */

771
static int btrfs_may_delete(struct inode *dir, struct dentry *victim, int isdir)
772 773 774 775 776 777 778
{
	int error;

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

	BUG_ON(victim->d_parent->d_inode != dir);
779
	audit_inode_child(dir, victim, AUDIT_TYPE_CHILD_DELETE);
780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803

	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;
}

804 805 806 807 808 809 810 811 812 813 814 815 816 817 818
/* 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.
 */
819 820
static noinline int btrfs_mksubvol(struct path *parent,
				   char *name, int namelen,
S
Sage Weil 已提交
821
				   struct btrfs_root *snap_src,
A
Arne Jansen 已提交
822
				   u64 *async_transid, bool readonly,
823
				   struct btrfs_qgroup_inherit *inherit)
824
{
825
	struct inode *dir  = parent->dentry->d_inode;
826 827 828
	struct dentry *dentry;
	int error;

829 830 831
	error = mutex_lock_killable_nested(&dir->i_mutex, I_MUTEX_PARENT);
	if (error == -EINTR)
		return error;
832 833 834 835 836 837 838 839 840 841

	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;

842
	error = btrfs_may_create(dir, dentry);
843
	if (error)
844
		goto out_dput;
845

C
Chris Mason 已提交
846 847 848 849 850 851 852 853 854 855
	/*
	 * 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;

856 857 858 859 860
	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;

861
	if (snap_src) {
862
		error = create_snapshot(snap_src, dir, dentry, name, namelen,
A
Arne Jansen 已提交
863
					async_transid, readonly, inherit);
864
	} else {
865 866
		error = create_subvol(dir, dentry, name, namelen,
				      async_transid, inherit);
867
	}
868 869 870 871
	if (!error)
		fsnotify_mkdir(dir, dentry);
out_up_read:
	up_read(&BTRFS_I(dir)->root->fs_info->subvol_sem);
872 873 874
out_dput:
	dput(dentry);
out_unlock:
875
	mutex_unlock(&dir->i_mutex);
876 877 878
	return error;
}

C
Chris Mason 已提交
879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928
/*
 * When we're defragging a range, we don't want to kick it off again
 * if it is really just waiting for delalloc to send it down.
 * If we find a nice big extent or delalloc range for the bytes in the
 * file you want to defrag, we return 0 to let you know to skip this
 * part of the file
 */
static int check_defrag_in_cache(struct inode *inode, u64 offset, int thresh)
{
	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
	struct extent_map *em = NULL;
	struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
	u64 end;

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

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

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

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

935
	min_key.objectid = ino;
C
Chris Mason 已提交
936 937 938
	min_key.type = BTRFS_EXTENT_DATA_KEY;
	min_key.offset = *off;

939
	while (1) {
940
		path->keep_locks = 1;
941
		ret = btrfs_search_forward(root, &min_key, path, newer_than);
C
Chris Mason 已提交
942 943
		if (ret != 0)
			goto none;
944 945 946
		path->keep_locks = 0;
		btrfs_unlock_up_safe(path, 1);
process_slot:
947
		if (min_key.objectid != ino)
C
Chris Mason 已提交
948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964
			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;
		}

965 966 967 968 969 970
		path->slots[0]++;
		if (path->slots[0] < btrfs_header_nritems(leaf)) {
			btrfs_item_key_to_cpu(leaf, &min_key, path->slots[0]);
			goto process_slot;
		}

C
Chris Mason 已提交
971 972 973 974 975 976 977 978 979 980 981
		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
982
static struct extent_map *defrag_lookup_extent(struct inode *inode, u64 start)
983 984
{
	struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
L
Li Zefan 已提交
985 986 987
	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
	struct extent_map *em;
	u64 len = PAGE_CACHE_SIZE;
988

L
Li Zefan 已提交
989 990 991 992
	/*
	 * hopefully we have this extent in the tree already, try without
	 * the full extent lock
	 */
993
	read_lock(&em_tree->lock);
L
Li Zefan 已提交
994
	em = lookup_extent_mapping(em_tree, start, len);
995 996
	read_unlock(&em_tree->lock);

L
Li Zefan 已提交
997
	if (!em) {
998 999 1000
		struct extent_state *cached = NULL;
		u64 end = start + len - 1;

L
Li Zefan 已提交
1001
		/* get the big lock and read metadata off disk */
1002
		lock_extent_bits(io_tree, start, end, 0, &cached);
L
Li Zefan 已提交
1003
		em = btrfs_get_extent(inode, NULL, 0, start, len, 0);
1004
		unlock_extent_cached(io_tree, start, end, &cached, GFP_NOFS);
L
Li Zefan 已提交
1005 1006 1007 1008 1009 1010 1011

		if (IS_ERR(em))
			return NULL;
	}

	return em;
}
1012

L
Li Zefan 已提交
1013 1014 1015 1016 1017 1018 1019 1020 1021 1022
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);
1023 1024
	if (!next || next->block_start >= EXTENT_MAP_LAST_BYTE ||
	    (em->block_start + em->block_len == next->block_start))
L
Li Zefan 已提交
1025 1026 1027
		ret = false;

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

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

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

	*skip = 0;

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

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

L
Li Zefan 已提交
1058
	next_mergeable = defrag_check_next_extent(inode, em);
1059 1060

	/*
L
Li Zefan 已提交
1061 1062
	 * we hit a real extent, if it is big or the next extent is not a
	 * real extent, don't bother defragging it
1063
	 */
1064
	if (!compress && (*last_len == 0 || *last_len >= thresh) &&
L
Li Zefan 已提交
1065
	    (em->len >= thresh || !next_mergeable))
1066
		ret = 0;
1067
out:
1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087
	/*
	 * 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
1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102
/*
 * 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,
1103
				    unsigned long num_pages)
C
Christoph Hellwig 已提交
1104
{
C
Chris Mason 已提交
1105 1106 1107 1108
	unsigned long file_end;
	u64 isize = i_size_read(inode);
	u64 page_start;
	u64 page_end;
1109
	u64 page_cnt;
C
Chris Mason 已提交
1110 1111 1112
	int ret;
	int i;
	int i_done;
1113
	struct btrfs_ordered_extent *ordered;
C
Chris Mason 已提交
1114
	struct extent_state *cached_state = NULL;
1115
	struct extent_io_tree *tree;
1116
	gfp_t mask = btrfs_alloc_write_mask(inode->i_mapping);
C
Chris Mason 已提交
1117 1118

	file_end = (isize - 1) >> PAGE_CACHE_SHIFT;
1119 1120 1121 1122
	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
1123 1124

	ret = btrfs_delalloc_reserve_space(inode,
1125
					   page_cnt << PAGE_CACHE_SHIFT);
C
Chris Mason 已提交
1126 1127 1128
	if (ret)
		return ret;
	i_done = 0;
1129
	tree = &BTRFS_I(inode)->io_tree;
C
Chris Mason 已提交
1130 1131

	/* step one, lock all the pages */
1132
	for (i = 0; i < page_cnt; i++) {
C
Chris Mason 已提交
1133
		struct page *page;
1134
again:
1135
		page = find_or_create_page(inode->i_mapping,
1136
					   start_index + i, mask);
C
Chris Mason 已提交
1137 1138 1139
		if (!page)
			break;

1140 1141 1142
		page_start = page_offset(page);
		page_end = page_start + PAGE_CACHE_SIZE - 1;
		while (1) {
1143 1144
			lock_extent_bits(tree, page_start, page_end,
					 0, &cached_state);
1145 1146
			ordered = btrfs_lookup_ordered_extent(inode,
							      page_start);
1147 1148
			unlock_extent_cached(tree, page_start, page_end,
					     &cached_state, GFP_NOFS);
1149 1150 1151 1152 1153 1154 1155
			if (!ordered)
				break;

			unlock_page(page);
			btrfs_start_ordered_extent(inode, ordered, 1);
			btrfs_put_ordered_extent(ordered);
			lock_page(page);
1156 1157 1158 1159 1160 1161 1162 1163 1164
			/*
			 * 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;
			}
1165 1166
		}

C
Chris Mason 已提交
1167 1168 1169 1170 1171 1172 1173 1174 1175 1176
		if (!PageUptodate(page)) {
			btrfs_readpage(NULL, page);
			lock_page(page);
			if (!PageUptodate(page)) {
				unlock_page(page);
				page_cache_release(page);
				ret = -EIO;
				break;
			}
		}
1177 1178 1179 1180 1181 1182 1183

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

C
Chris Mason 已提交
1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203
		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,
1204
			 page_start, page_end - 1, 0, &cached_state);
C
Chris Mason 已提交
1205 1206
	clear_extent_bit(&BTRFS_I(inode)->io_tree, page_start,
			  page_end - 1, EXTENT_DIRTY | EXTENT_DELALLOC |
1207 1208
			  EXTENT_DO_ACCOUNTING | EXTENT_DEFRAG, 0, 0,
			  &cached_state, GFP_NOFS);
C
Chris Mason 已提交
1209

1210
	if (i_done != page_cnt) {
1211 1212 1213
		spin_lock(&BTRFS_I(inode)->lock);
		BTRFS_I(inode)->outstanding_extents++;
		spin_unlock(&BTRFS_I(inode)->lock);
C
Chris Mason 已提交
1214
		btrfs_delalloc_release_space(inode,
1215
				     (page_cnt - i_done) << PAGE_CACHE_SHIFT);
C
Chris Mason 已提交
1216 1217 1218
	}


1219 1220
	set_extent_defrag(&BTRFS_I(inode)->io_tree, page_start, page_end - 1,
			  &cached_state, GFP_NOFS);
C
Chris Mason 已提交
1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239

	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]);
	}
1240
	btrfs_delalloc_release_space(inode, page_cnt << PAGE_CACHE_SHIFT);
C
Chris Mason 已提交
1241 1242 1243 1244 1245 1246 1247 1248 1249 1250
	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
1251
	unsigned long last_index;
1252
	u64 isize = i_size_read(inode);
1253 1254 1255
	u64 last_len = 0;
	u64 skip = 0;
	u64 defrag_end = 0;
C
Chris Mason 已提交
1256
	u64 newer_off = range->start;
C
Christoph Hellwig 已提交
1257
	unsigned long i;
1258
	unsigned long ra_index = 0;
C
Christoph Hellwig 已提交
1259
	int ret;
C
Chris Mason 已提交
1260
	int defrag_count = 0;
1261
	int compress_type = BTRFS_COMPRESS_ZLIB;
C
Chris Mason 已提交
1262
	int extent_thresh = range->extent_thresh;
1263 1264
	unsigned long max_cluster = (256 * 1024) >> PAGE_CACHE_SHIFT;
	unsigned long cluster = max_cluster;
C
Chris Mason 已提交
1265 1266 1267
	u64 new_align = ~((u64)128 * 1024 - 1);
	struct page **pages = NULL;

1268 1269 1270 1271 1272
	if (isize == 0)
		return 0;

	if (range->start >= isize)
		return -EINVAL;
1273 1274 1275 1276 1277 1278 1279

	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
1280

1281 1282
	if (extent_thresh == 0)
		extent_thresh = 256 * 1024;
1283

C
Chris Mason 已提交
1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296
	/*
	 * 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;
	}

1297
	pages = kmalloc_array(max_cluster, sizeof(struct page *),
C
Chris Mason 已提交
1298 1299 1300 1301 1302 1303 1304
			GFP_NOFS);
	if (!pages) {
		ret = -ENOMEM;
		goto out_ra;
	}

	/* find the last page to defrag */
C
Chris Mason 已提交
1305
	if (range->start + range->len > range->start) {
1306
		last_index = min_t(u64, isize - 1,
C
Chris Mason 已提交
1307 1308
			 range->start + range->len - 1) >> PAGE_CACHE_SHIFT;
	} else {
1309
		last_index = (isize - 1) >> PAGE_CACHE_SHIFT;
C
Chris Mason 已提交
1310 1311
	}

C
Chris Mason 已提交
1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327
	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
1328
		max_to_defrag = last_index + 1;
C
Chris Mason 已提交
1329

L
Li Zefan 已提交
1330 1331 1332 1333 1334 1335 1336
	/*
	 * 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;

1337 1338 1339
	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
1340 1341 1342 1343 1344 1345 1346
		/*
		 * make sure we stop running if someone unmounts
		 * the FS
		 */
		if (!(inode->i_sb->s_flags & MS_ACTIVE))
			break;

1347
		if (btrfs_defrag_cancelled(root->fs_info)) {
1348
			printk(KERN_DEBUG "BTRFS: defrag_file cancelled\n");
1349 1350 1351 1352
			ret = -EAGAIN;
			break;
		}

1353
		if (!should_defrag_range(inode, (u64)i << PAGE_CACHE_SHIFT,
L
Li Zefan 已提交
1354
					 extent_thresh, &last_len, &skip,
1355 1356
					 &defrag_end, range->flags &
					 BTRFS_DEFRAG_RANGE_COMPRESS)) {
1357 1358 1359 1360 1361 1362 1363 1364 1365
			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;
		}
1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380

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

		if (i + cluster > ra_index) {
			ra_index = max(i, ra_index);
			btrfs_force_ra(inode->i_mapping, ra, file, ra_index,
				       cluster);
			ra_index += max_cluster;
		}
1381

1382
		mutex_lock(&inode->i_mutex);
1383 1384
		if (range->flags & BTRFS_DEFRAG_RANGE_COMPRESS)
			BTRFS_I(inode)->force_compress = compress_type;
1385
		ret = cluster_pages_for_defrag(inode, pages, i, cluster);
1386 1387
		if (ret < 0) {
			mutex_unlock(&inode->i_mutex);
C
Chris Mason 已提交
1388
			goto out_ra;
1389
		}
C
Chris Mason 已提交
1390 1391

		defrag_count += ret;
1392
		balance_dirty_pages_ratelimited(inode->i_mapping);
1393
		mutex_unlock(&inode->i_mutex);
C
Chris Mason 已提交
1394 1395 1396 1397 1398

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

1399 1400 1401
			if (ret > 0)
				i += ret;

C
Chris Mason 已提交
1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412
			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
1413
			}
C
Chris Mason 已提交
1414
		} else {
1415
			if (ret > 0) {
L
Li Zefan 已提交
1416
				i += ret;
1417 1418
				last_len += ret << PAGE_CACHE_SHIFT;
			} else {
L
Li Zefan 已提交
1419
				i++;
1420 1421
				last_len = 0;
			}
C
Christoph Hellwig 已提交
1422 1423 1424
		}
	}

1425
	if ((range->flags & BTRFS_DEFRAG_RANGE_START_IO)) {
C
Chris Mason 已提交
1426
		filemap_flush(inode->i_mapping);
1427 1428 1429 1430
		if (test_bit(BTRFS_INODE_HAS_ASYNC_EXTENT,
			     &BTRFS_I(inode)->runtime_flags))
			filemap_flush(inode->i_mapping);
	}
C
Chris Mason 已提交
1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446

	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);
	}

1447
	if (range->compress_type == BTRFS_COMPRESS_LZO) {
1448
		btrfs_set_fs_incompat(root->fs_info, COMPRESS_LZO);
1449 1450
	}

1451
	ret = defrag_count;
1452

C
Chris Mason 已提交
1453
out_ra:
1454 1455 1456 1457 1458
	if (range->flags & BTRFS_DEFRAG_RANGE_COMPRESS) {
		mutex_lock(&inode->i_mutex);
		BTRFS_I(inode)->force_compress = BTRFS_COMPRESS_NONE;
		mutex_unlock(&inode->i_mutex);
	}
C
Chris Mason 已提交
1459 1460 1461
	if (!file)
		kfree(ra);
	kfree(pages);
1462
	return ret;
C
Christoph Hellwig 已提交
1463 1464
}

1465
static noinline int btrfs_ioctl_resize(struct file *file,
1466
					void __user *arg)
C
Christoph Hellwig 已提交
1467 1468 1469 1470
{
	u64 new_size;
	u64 old_size;
	u64 devid = 1;
A
Al Viro 已提交
1471
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
C
Christoph Hellwig 已提交
1472 1473 1474 1475
	struct btrfs_ioctl_vol_args *vol_args;
	struct btrfs_trans_handle *trans;
	struct btrfs_device *device = NULL;
	char *sizestr;
1476
	char *retptr;
C
Christoph Hellwig 已提交
1477 1478 1479 1480
	char *devstr = NULL;
	int ret = 0;
	int mod = 0;

1481 1482 1483
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

1484 1485 1486 1487
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;

1488 1489
	if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
			1)) {
1490
		mnt_drop_write_file(file);
1491
		return BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
1492 1493
	}

1494
	mutex_lock(&root->fs_info->volume_mutex);
L
Li Zefan 已提交
1495
	vol_args = memdup_user(arg, sizeof(*vol_args));
1496 1497 1498 1499
	if (IS_ERR(vol_args)) {
		ret = PTR_ERR(vol_args);
		goto out;
	}
1500 1501

	vol_args->name[BTRFS_PATH_NAME_MAX] = '\0';
C
Christoph Hellwig 已提交
1502 1503 1504 1505 1506 1507 1508

	sizestr = vol_args->name;
	devstr = strchr(sizestr, ':');
	if (devstr) {
		sizestr = devstr + 1;
		*devstr = '\0';
		devstr = vol_args->name;
1509 1510 1511
		ret = kstrtoull(devstr, 10, &devid);
		if (ret)
			goto out_free;
1512 1513 1514 1515
		if (!devid) {
			ret = -EINVAL;
			goto out_free;
		}
1516
		btrfs_info(root->fs_info, "resizing devid %llu", devid);
C
Christoph Hellwig 已提交
1517
	}
M
Miao Xie 已提交
1518

1519
	device = btrfs_find_device(root->fs_info, devid, NULL, NULL);
C
Christoph Hellwig 已提交
1520
	if (!device) {
1521
		btrfs_info(root->fs_info, "resizer unable to find device %llu",
1522
		       devid);
1523
		ret = -ENODEV;
1524
		goto out_free;
C
Christoph Hellwig 已提交
1525
	}
M
Miao Xie 已提交
1526 1527

	if (!device->writeable) {
1528 1529
		btrfs_info(root->fs_info,
			   "resizer unable to apply on readonly device %llu",
1530
		       devid);
1531
		ret = -EPERM;
L
Liu Bo 已提交
1532 1533 1534
		goto out_free;
	}

C
Christoph Hellwig 已提交
1535 1536 1537 1538 1539 1540 1541 1542 1543 1544
	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++;
		}
1545 1546
		new_size = memparse(sizestr, &retptr);
		if (*retptr != '\0' || new_size == 0) {
C
Christoph Hellwig 已提交
1547
			ret = -EINVAL;
1548
			goto out_free;
C
Christoph Hellwig 已提交
1549 1550 1551
		}
	}

1552
	if (device->is_tgtdev_for_dev_replace) {
1553
		ret = -EPERM;
1554 1555 1556
		goto out_free;
	}

C
Christoph Hellwig 已提交
1557 1558 1559 1560 1561
	old_size = device->total_bytes;

	if (mod < 0) {
		if (new_size > old_size) {
			ret = -EINVAL;
1562
			goto out_free;
C
Christoph Hellwig 已提交
1563 1564 1565
		}
		new_size = old_size - new_size;
	} else if (mod > 0) {
1566 1567 1568 1569
		if (new_size > ULLONG_MAX - old_size) {
			ret = -EINVAL;
			goto out_free;
		}
C
Christoph Hellwig 已提交
1570 1571 1572 1573 1574
		new_size = old_size + new_size;
	}

	if (new_size < 256 * 1024 * 1024) {
		ret = -EINVAL;
1575
		goto out_free;
C
Christoph Hellwig 已提交
1576 1577 1578
	}
	if (new_size > device->bdev->bd_inode->i_size) {
		ret = -EFBIG;
1579
		goto out_free;
C
Christoph Hellwig 已提交
1580 1581 1582 1583 1584
	}

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

1585
	printk_in_rcu(KERN_INFO "BTRFS: new size for %s is %llu\n",
1586
		      rcu_str_deref(device->name), new_size);
C
Christoph Hellwig 已提交
1587 1588

	if (new_size > old_size) {
1589
		trans = btrfs_start_transaction(root, 0);
1590 1591
		if (IS_ERR(trans)) {
			ret = PTR_ERR(trans);
1592
			goto out_free;
1593
		}
C
Christoph Hellwig 已提交
1594 1595
		ret = btrfs_grow_device(trans, device, new_size);
		btrfs_commit_transaction(trans, root);
1596
	} else if (new_size < old_size) {
C
Christoph Hellwig 已提交
1597
		ret = btrfs_shrink_device(device, new_size);
1598
	} /* equal, nothing need to do */
C
Christoph Hellwig 已提交
1599

1600
out_free:
C
Christoph Hellwig 已提交
1601
	kfree(vol_args);
1602 1603
out:
	mutex_unlock(&root->fs_info->volume_mutex);
1604
	atomic_set(&root->fs_info->mutually_exclusive_operation_running, 0);
1605
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
1606 1607 1608
	return ret;
}

S
Sage Weil 已提交
1609
static noinline int btrfs_ioctl_snap_create_transid(struct file *file,
A
Arne Jansen 已提交
1610 1611
				char *name, unsigned long fd, int subvol,
				u64 *transid, bool readonly,
1612
				struct btrfs_qgroup_inherit *inherit)
C
Christoph Hellwig 已提交
1613 1614
{
	int namelen;
1615
	int ret = 0;
C
Christoph Hellwig 已提交
1616

1617 1618 1619 1620
	ret = mnt_want_write_file(file);
	if (ret)
		goto out;

S
Sage Weil 已提交
1621 1622
	namelen = strlen(name);
	if (strchr(name, '/')) {
C
Christoph Hellwig 已提交
1623
		ret = -EINVAL;
1624
		goto out_drop_write;
C
Christoph Hellwig 已提交
1625 1626
	}

1627 1628 1629
	if (name[0] == '.' &&
	   (namelen == 1 || (name[1] == '.' && namelen == 2))) {
		ret = -EEXIST;
1630
		goto out_drop_write;
1631 1632
	}

1633
	if (subvol) {
S
Sage Weil 已提交
1634
		ret = btrfs_mksubvol(&file->f_path, name, namelen,
A
Arne Jansen 已提交
1635
				     NULL, transid, readonly, inherit);
1636
	} else {
1637
		struct fd src = fdget(fd);
1638
		struct inode *src_inode;
1639
		if (!src.file) {
1640
			ret = -EINVAL;
1641
			goto out_drop_write;
1642 1643
		}

A
Al Viro 已提交
1644 1645
		src_inode = file_inode(src.file);
		if (src_inode->i_sb != file_inode(file)->i_sb) {
1646 1647
			btrfs_info(BTRFS_I(src_inode)->root->fs_info,
				   "Snapshot src from another FS");
1648
			ret = -EXDEV;
1649 1650 1651 1652 1653 1654
		} else if (!inode_owner_or_capable(src_inode)) {
			/*
			 * Subvolume creation is not restricted, but snapshots
			 * are limited to own subvolumes only
			 */
			ret = -EPERM;
1655 1656 1657 1658
		} else {
			ret = btrfs_mksubvol(&file->f_path, name, namelen,
					     BTRFS_I(src_inode)->root,
					     transid, readonly, inherit);
1659
		}
1660
		fdput(src);
1661
	}
1662 1663
out_drop_write:
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
1664
out:
S
Sage Weil 已提交
1665 1666 1667 1668
	return ret;
}

static noinline int btrfs_ioctl_snap_create(struct file *file,
1669
					    void __user *arg, int subvol)
S
Sage Weil 已提交
1670
{
1671
	struct btrfs_ioctl_vol_args *vol_args;
S
Sage Weil 已提交
1672 1673
	int ret;

1674 1675 1676 1677
	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
1678

1679
	ret = btrfs_ioctl_snap_create_transid(file, vol_args->name,
L
Li Zefan 已提交
1680
					      vol_args->fd, subvol,
A
Arne Jansen 已提交
1681
					      NULL, false, NULL);
1682

1683 1684 1685
	kfree(vol_args);
	return ret;
}
1686

1687 1688 1689 1690 1691 1692 1693
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
1694
	bool readonly = false;
A
Arne Jansen 已提交
1695
	struct btrfs_qgroup_inherit *inherit = NULL;
1696

1697 1698 1699 1700
	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';
1701

L
Li Zefan 已提交
1702
	if (vol_args->flags &
A
Arne Jansen 已提交
1703 1704
	    ~(BTRFS_SUBVOL_CREATE_ASYNC | BTRFS_SUBVOL_RDONLY |
	      BTRFS_SUBVOL_QGROUP_INHERIT)) {
L
Li Zefan 已提交
1705
		ret = -EOPNOTSUPP;
1706
		goto out;
S
Sage Weil 已提交
1707
	}
1708 1709 1710

	if (vol_args->flags & BTRFS_SUBVOL_CREATE_ASYNC)
		ptr = &transid;
L
Li Zefan 已提交
1711 1712
	if (vol_args->flags & BTRFS_SUBVOL_RDONLY)
		readonly = true;
A
Arne Jansen 已提交
1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723
	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;
		}
	}
1724 1725

	ret = btrfs_ioctl_snap_create_transid(file, vol_args->name,
A
Arne Jansen 已提交
1726
					      vol_args->fd, subvol, ptr,
1727
					      readonly, inherit);
1728 1729 1730 1731 1732 1733

	if (ret == 0 && ptr &&
	    copy_to_user(arg +
			 offsetof(struct btrfs_ioctl_vol_args_v2,
				  transid), ptr, sizeof(*ptr)))
		ret = -EFAULT;
1734
out:
C
Christoph Hellwig 已提交
1735
	kfree(vol_args);
A
Arne Jansen 已提交
1736
	kfree(inherit);
C
Christoph Hellwig 已提交
1737 1738 1739
	return ret;
}

1740 1741 1742
static noinline int btrfs_ioctl_subvol_getflags(struct file *file,
						void __user *arg)
{
A
Al Viro 已提交
1743
	struct inode *inode = file_inode(file);
1744 1745 1746 1747
	struct btrfs_root *root = BTRFS_I(inode)->root;
	int ret = 0;
	u64 flags = 0;

L
Li Zefan 已提交
1748
	if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID)
1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764
		return -EINVAL;

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

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

	return ret;
}

static noinline int btrfs_ioctl_subvol_setflags(struct file *file,
					      void __user *arg)
{
A
Al Viro 已提交
1765
	struct inode *inode = file_inode(file);
1766 1767 1768 1769 1770 1771
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;
	u64 root_flags;
	u64 flags;
	int ret = 0;

1772 1773 1774
	if (!inode_owner_or_capable(inode))
		return -EPERM;

1775 1776 1777
	ret = mnt_want_write_file(file);
	if (ret)
		goto out;
1778

1779 1780 1781 1782
	if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID) {
		ret = -EINVAL;
		goto out_drop_write;
	}
1783

1784 1785 1786 1787
	if (copy_from_user(&flags, arg, sizeof(flags))) {
		ret = -EFAULT;
		goto out_drop_write;
	}
1788

1789 1790 1791 1792
	if (flags & BTRFS_SUBVOL_CREATE_ASYNC) {
		ret = -EINVAL;
		goto out_drop_write;
	}
1793

1794 1795 1796 1797
	if (flags & ~BTRFS_SUBVOL_RDONLY) {
		ret = -EOPNOTSUPP;
		goto out_drop_write;
	}
1798 1799 1800 1801 1802

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

	/* nothing to do */
	if (!!(flags & BTRFS_SUBVOL_RDONLY) == btrfs_root_readonly(root))
1803
		goto out_drop_sem;
1804 1805

	root_flags = btrfs_root_flags(&root->root_item);
1806
	if (flags & BTRFS_SUBVOL_RDONLY) {
1807 1808
		btrfs_set_root_flags(&root->root_item,
				     root_flags | BTRFS_ROOT_SUBVOL_RDONLY);
1809 1810 1811 1812 1813 1814 1815 1816
	} else {
		/*
		 * Block RO -> RW transition if this subvolume is involved in
		 * send
		 */
		spin_lock(&root->root_item_lock);
		if (root->send_in_progress == 0) {
			btrfs_set_root_flags(&root->root_item,
1817
				     root_flags & ~BTRFS_ROOT_SUBVOL_RDONLY);
1818 1819 1820 1821 1822 1823 1824 1825 1826 1827
			spin_unlock(&root->root_item_lock);
		} else {
			spin_unlock(&root->root_item_lock);
			btrfs_warn(root->fs_info,
			"Attempt to set subvolume %llu read-write during send",
					root->root_key.objectid);
			ret = -EPERM;
			goto out_drop_sem;
		}
	}
1828 1829 1830 1831 1832 1833 1834

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

1835
	ret = btrfs_update_root(trans, root->fs_info->tree_root,
1836 1837 1838 1839 1840 1841
				&root->root_key, &root->root_item);

	btrfs_commit_transaction(trans, root);
out_reset:
	if (ret)
		btrfs_set_root_flags(&root->root_item, root_flags);
1842
out_drop_sem:
1843
	up_write(&root->fs_info->subvol_sem);
1844 1845 1846
out_drop_write:
	mnt_drop_write_file(file);
out:
1847 1848 1849
	return ret;
}

1850 1851 1852 1853 1854 1855
/*
 * helper to check if the subvolume references other subvolumes
 */
static noinline int may_destroy_subvol(struct btrfs_root *root)
{
	struct btrfs_path *path;
1856
	struct btrfs_dir_item *di;
1857
	struct btrfs_key key;
1858
	u64 dir_id;
1859 1860 1861 1862 1863 1864
	int ret;

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

1865 1866 1867 1868 1869 1870 1871
	/* Make sure this root isn't set as the default subvol */
	dir_id = btrfs_super_root_dir(root->fs_info->super_copy);
	di = btrfs_lookup_dir_item(NULL, root->fs_info->tree_root, path,
				   dir_id, "default", 7, 0);
	if (di && !IS_ERR(di)) {
		btrfs_dir_item_key_to_cpu(path->nodes[0], di, &key);
		if (key.objectid == root->root_key.objectid) {
1872 1873 1874
			ret = -EPERM;
			btrfs_err(root->fs_info, "deleting default subvolume "
				  "%llu is not allowed", key.objectid);
1875 1876 1877 1878 1879
			goto out;
		}
		btrfs_release_path(path);
	}

1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902
	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;
}

1903 1904 1905
static noinline int key_in_sk(struct btrfs_key *key,
			      struct btrfs_ioctl_search_key *sk)
{
1906 1907 1908 1909 1910 1911 1912 1913 1914
	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)
1915
		return 0;
1916 1917 1918 1919 1920 1921 1922

	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)
1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958
		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);

1959 1960 1961 1962 1963
		btrfs_item_key_to_cpu(leaf, key, i);
		if (!key_in_sk(key, sk))
			continue;

		if (sizeof(sh) + item_len > BTRFS_SEARCH_ARGS_BUFSIZE)
1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988
			item_len = 0;

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

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

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

		if (item_len) {
			char *p = buf + *sk_offset;
			/* copy the item */
			read_extent_buffer(leaf, p,
					   item_off, item_len);
			*sk_offset += item_len;
		}
1989
		(*num_found)++;
1990 1991 1992 1993 1994

		if (*num_found >= sk->nr_items)
			break;
	}
advance_key:
1995 1996
	ret = 0;
	if (key->offset < (u64)-1 && key->offset < sk->max_offset)
1997
		key->offset++;
1998 1999
	else if (key->type < (u8)-1 && key->type < sk->max_type) {
		key->offset = 0;
2000
		key->type++;
2001 2002 2003
	} else if (key->objectid < (u64)-1 && key->objectid < sk->max_objectid) {
		key->offset = 0;
		key->type = 0;
2004
		key->objectid++;
2005 2006
	} else
		ret = 1;
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035
overflow:
	return ret;
}

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

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

	if (sk->tree_id == 0) {
		/* search the root of the inode that was passed */
		root = BTRFS_I(inode)->root;
	} else {
		key.objectid = sk->tree_id;
		key.type = BTRFS_ROOT_ITEM_KEY;
		key.offset = (u64)-1;
		root = btrfs_read_fs_root_no_name(info, &key);
		if (IS_ERR(root)) {
2036
			printk(KERN_ERR "BTRFS: could not find root %llu\n",
2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048
			       sk->tree_id);
			btrfs_free_path(path);
			return -ENOENT;
		}
	}

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

	path->keep_locks = 1;

2049
	while (1) {
2050
		ret = btrfs_search_forward(root, &key, path, sk->min_transid);
2051 2052 2053 2054 2055 2056 2057
		if (ret != 0) {
			if (ret > 0)
				ret = 0;
			goto err;
		}
		ret = copy_to_sk(root, path, &key, sk, args->buf,
				 &sk_offset, &num_found);
2058
		btrfs_release_path(path);
2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079
		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
2080 2081 2082
	args = memdup_user(argp, sizeof(*args));
	if (IS_ERR(args))
		return PTR_ERR(args);
2083

A
Al Viro 已提交
2084
	inode = file_inode(file);
2085 2086 2087 2088 2089 2090 2091
	ret = search_ioctl(inode, args);
	if (ret == 0 && copy_to_user(argp, args, sizeof(*args)))
		ret = -EFAULT;
	kfree(args);
	return ret;
}

2092
/*
2093 2094 2095
 * Search INODE_REFs to identify path name of 'dirid' directory
 * in a 'tree_id' tree. and sets path name to 'name'.
 */
2096 2097 2098 2099 2100
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;
2101
	char *ptr;
2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118
	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;

2119
	ptr = &name[BTRFS_INO_LOOKUP_PATH_MAX];
2120 2121 2122 2123 2124 2125

	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)) {
2126
		printk(KERN_ERR "BTRFS: could not find root %llu\n", tree_id);
2127 2128
		ret = -ENOENT;
		goto out;
2129 2130 2131 2132
	}

	key.objectid = dirid;
	key.type = BTRFS_INODE_REF_KEY;
2133
	key.offset = (u64)-1;
2134

2135
	while (1) {
2136 2137 2138
		ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
		if (ret < 0)
			goto out;
2139 2140 2141 2142 2143 2144 2145 2146 2147 2148
		else if (ret > 0) {
			ret = btrfs_previous_item(root, path, dirid,
						  BTRFS_INODE_REF_KEY);
			if (ret < 0)
				goto out;
			else if (ret > 0) {
				ret = -ENOENT;
				goto out;
			}
		}
2149 2150 2151 2152 2153 2154 2155 2156 2157

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

		iref = btrfs_item_ptr(l, slot, struct btrfs_inode_ref);
		len = btrfs_inode_ref_name_len(l, iref);
		ptr -= len + 1;
		total_len += len + 1;
2158 2159
		if (ptr < name) {
			ret = -ENAMETOOLONG;
2160
			goto out;
2161
		}
2162 2163

		*(ptr + len) = '/';
2164
		read_extent_buffer(l, ptr, (unsigned long)(iref + 1), len);
2165 2166 2167 2168

		if (key.offset == BTRFS_FIRST_FREE_OBJECTID)
			break;

2169
		btrfs_release_path(path);
2170
		key.objectid = key.offset;
2171
		key.offset = (u64)-1;
2172 2173
		dirid = key.objectid;
	}
2174
	memmove(name, ptr, total_len);
2175
	name[total_len] = '\0';
2176 2177 2178
	ret = 0;
out:
	btrfs_free_path(path);
2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191
	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
2192 2193 2194
	args = memdup_user(argp, sizeof(*args));
	if (IS_ERR(args))
		return PTR_ERR(args);
2195

A
Al Viro 已提交
2196
	inode = file_inode(file);
2197

2198 2199 2200
	if (args->treeid == 0)
		args->treeid = BTRFS_I(inode)->root->root_key.objectid;

2201 2202 2203 2204 2205 2206 2207 2208
	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);
2209 2210 2211
	return ret;
}

2212 2213 2214
static noinline int btrfs_ioctl_snap_destroy(struct file *file,
					     void __user *arg)
{
A
Al Viro 已提交
2215
	struct dentry *parent = file->f_path.dentry;
2216 2217 2218 2219 2220 2221 2222
	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;
2223
	struct btrfs_block_rsv block_rsv;
2224
	u64 root_flags;
2225
	u64 qgroup_reserved;
2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241
	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;
	}

2242
	err = mnt_want_write_file(file);
2243 2244 2245
	if (err)
		goto out;

2246

2247 2248
	err = mutex_lock_killable_nested(&dir->i_mutex, I_MUTEX_PARENT);
	if (err == -EINTR)
2249
		goto out_drop_write;
2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261
	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;
2262
	dest = BTRFS_I(inode)->root;
2263
	if (!capable(CAP_SYS_ADMIN)) {
2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296
		/*
		 * 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;
	}

2297 2298 2299 2300 2301
	/* check if subvolume may be deleted by a user */
	err = btrfs_may_delete(dir, dentry, 1);
	if (err)
		goto out_dput;

L
Li Zefan 已提交
2302
	if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID) {
2303 2304 2305 2306 2307
		err = -EINVAL;
		goto out_dput;
	}

	mutex_lock(&inode->i_mutex);
2308 2309 2310 2311 2312 2313 2314

	/*
	 * Don't allow to delete a subvolume with send in progress. This is
	 * inside the i_mutex so the error handling that has to drop the bit
	 * again is not run concurrently.
	 */
	spin_lock(&dest->root_item_lock);
2315 2316 2317
	root_flags = btrfs_root_flags(&dest->root_item);
	if (dest->send_in_progress == 0) {
		btrfs_set_root_flags(&dest->root_item,
2318 2319 2320 2321 2322 2323
				root_flags | BTRFS_ROOT_SUBVOL_DEAD);
		spin_unlock(&dest->root_item_lock);
	} else {
		spin_unlock(&dest->root_item_lock);
		btrfs_warn(root->fs_info,
			"Attempt to delete subvolume %llu during send",
2324
			dest->root_key.objectid);
2325 2326 2327 2328
		err = -EPERM;
		goto out_dput;
	}

2329 2330 2331 2332 2333 2334 2335 2336 2337 2338
	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;

2339 2340 2341 2342 2343 2344
	btrfs_init_block_rsv(&block_rsv, BTRFS_BLOCK_RSV_TEMP);
	/*
	 * One for dir inode, two for dir entries, two for root
	 * ref/backref.
	 */
	err = btrfs_subvolume_reserve_metadata(root, &block_rsv,
2345
					       5, &qgroup_reserved, true);
2346 2347 2348
	if (err)
		goto out_up_write;

2349 2350 2351
	trans = btrfs_start_transaction(root, 0);
	if (IS_ERR(trans)) {
		err = PTR_ERR(trans);
2352
		goto out_release;
2353
	}
2354 2355
	trans->block_rsv = &block_rsv;
	trans->bytes_reserved = block_rsv.size;
2356

2357 2358 2359 2360
	ret = btrfs_unlink_subvol(trans, root, dir,
				dest->root_key.objectid,
				dentry->d_name.name,
				dentry->d_name.len);
2361 2362 2363 2364 2365
	if (ret) {
		err = ret;
		btrfs_abort_transaction(trans, root, ret);
		goto out_end_trans;
	}
2366 2367 2368 2369 2370 2371 2372 2373

	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);

2374
	if (!test_and_set_bit(BTRFS_ROOT_ORPHAN_ITEM_INSERTED, &dest->state)) {
2375 2376 2377
		ret = btrfs_insert_orphan_item(trans,
					root->fs_info->tree_root,
					dest->root_key.objectid);
2378 2379 2380 2381 2382
		if (ret) {
			btrfs_abort_transaction(trans, root, ret);
			err = ret;
			goto out_end_trans;
		}
2383
	}
2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404

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

2405
out_end_trans:
2406 2407
	trans->block_rsv = NULL;
	trans->bytes_reserved = 0;
S
Sage Weil 已提交
2408
	ret = btrfs_end_transaction(trans, root);
2409 2410
	if (ret && !err)
		err = ret;
2411
	inode->i_flags |= S_DEAD;
2412 2413
out_release:
	btrfs_subvolume_release_metadata(root, &block_rsv, qgroup_reserved);
2414 2415 2416
out_up_write:
	up_write(&root->fs_info->subvol_sem);
out_unlock:
2417 2418
	if (err) {
		spin_lock(&dest->root_item_lock);
2419 2420
		root_flags = btrfs_root_flags(&dest->root_item);
		btrfs_set_root_flags(&dest->root_item,
2421 2422 2423
				root_flags & ~BTRFS_ROOT_SUBVOL_DEAD);
		spin_unlock(&dest->root_item_lock);
	}
2424 2425
	mutex_unlock(&inode->i_mutex);
	if (!err) {
2426
		shrink_dcache_sb(root->fs_info->sb);
2427 2428
		btrfs_invalidate_inodes(dest);
		d_delete(dentry);
2429
		ASSERT(dest->send_in_progress == 0);
2430 2431 2432 2433 2434 2435

		/* the last ref */
		if (dest->cache_inode) {
			iput(dest->cache_inode);
			dest->cache_inode = NULL;
		}
2436 2437 2438 2439 2440
	}
out_dput:
	dput(dentry);
out_unlock_dir:
	mutex_unlock(&dir->i_mutex);
2441
out_drop_write:
A
Al Viro 已提交
2442
	mnt_drop_write_file(file);
2443 2444 2445 2446 2447
out:
	kfree(vol_args);
	return err;
}

C
Chris Mason 已提交
2448
static int btrfs_ioctl_defrag(struct file *file, void __user *argp)
C
Christoph Hellwig 已提交
2449
{
A
Al Viro 已提交
2450
	struct inode *inode = file_inode(file);
C
Christoph Hellwig 已提交
2451
	struct btrfs_root *root = BTRFS_I(inode)->root;
C
Chris Mason 已提交
2452
	struct btrfs_ioctl_defrag_range_args *range;
Y
Yan Zheng 已提交
2453 2454
	int ret;

2455 2456 2457
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
L
Li Zefan 已提交
2458

2459 2460 2461
	if (btrfs_root_readonly(root)) {
		ret = -EROFS;
		goto out;
2462
	}
C
Christoph Hellwig 已提交
2463 2464 2465

	switch (inode->i_mode & S_IFMT) {
	case S_IFDIR:
2466 2467 2468 2469
		if (!capable(CAP_SYS_ADMIN)) {
			ret = -EPERM;
			goto out;
		}
2470
		ret = btrfs_defrag_root(root);
2471 2472
		if (ret)
			goto out;
2473
		ret = btrfs_defrag_root(root->fs_info->extent_root);
C
Christoph Hellwig 已提交
2474 2475
		break;
	case S_IFREG:
2476 2477 2478 2479
		if (!(file->f_mode & FMODE_WRITE)) {
			ret = -EINVAL;
			goto out;
		}
C
Chris Mason 已提交
2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491

		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);
2492
				goto out;
C
Chris Mason 已提交
2493 2494 2495 2496 2497 2498 2499 2500 2501 2502
			}
			/* compression requires us to start the IO */
			if ((range->flags & BTRFS_DEFRAG_RANGE_COMPRESS)) {
				range->flags |= BTRFS_DEFRAG_RANGE_START_IO;
				range->extent_thresh = (u32)-1;
			}
		} else {
			/* the rest are all set to zero by kzalloc */
			range->len = (u64)-1;
		}
A
Al Viro 已提交
2503
		ret = btrfs_defrag_file(file_inode(file), file,
C
Chris Mason 已提交
2504 2505 2506
					range, 0, 0);
		if (ret > 0)
			ret = 0;
C
Chris Mason 已提交
2507
		kfree(range);
C
Christoph Hellwig 已提交
2508
		break;
2509 2510
	default:
		ret = -EINVAL;
C
Christoph Hellwig 已提交
2511
	}
2512
out:
2513
	mnt_drop_write_file(file);
2514
	return ret;
C
Christoph Hellwig 已提交
2515 2516
}

2517
static long btrfs_ioctl_add_dev(struct btrfs_root *root, void __user *arg)
C
Christoph Hellwig 已提交
2518 2519 2520 2521
{
	struct btrfs_ioctl_vol_args *vol_args;
	int ret;

2522 2523 2524
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

2525 2526
	if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
			1)) {
2527
		return BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
2528 2529
	}

2530
	mutex_lock(&root->fs_info->volume_mutex);
L
Li Zefan 已提交
2531
	vol_args = memdup_user(arg, sizeof(*vol_args));
2532 2533 2534 2535
	if (IS_ERR(vol_args)) {
		ret = PTR_ERR(vol_args);
		goto out;
	}
C
Christoph Hellwig 已提交
2536

2537
	vol_args->name[BTRFS_PATH_NAME_MAX] = '\0';
C
Christoph Hellwig 已提交
2538 2539 2540
	ret = btrfs_init_new_device(root, vol_args->name);

	kfree(vol_args);
2541 2542
out:
	mutex_unlock(&root->fs_info->volume_mutex);
2543
	atomic_set(&root->fs_info->mutually_exclusive_operation_running, 0);
C
Christoph Hellwig 已提交
2544 2545 2546
	return ret;
}

2547
static long btrfs_ioctl_rm_dev(struct file *file, void __user *arg)
C
Christoph Hellwig 已提交
2548
{
A
Al Viro 已提交
2549
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
C
Christoph Hellwig 已提交
2550 2551 2552
	struct btrfs_ioctl_vol_args *vol_args;
	int ret;

2553 2554 2555
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

2556 2557 2558
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
Y
Yan Zheng 已提交
2559

L
Li Zefan 已提交
2560
	vol_args = memdup_user(arg, sizeof(*vol_args));
2561 2562 2563 2564
	if (IS_ERR(vol_args)) {
		ret = PTR_ERR(vol_args);
		goto out;
	}
C
Christoph Hellwig 已提交
2565

2566
	vol_args->name[BTRFS_PATH_NAME_MAX] = '\0';
C
Christoph Hellwig 已提交
2567

2568 2569 2570 2571 2572 2573 2574 2575
	if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running,
			1)) {
		ret = BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
		goto out;
	}

	mutex_lock(&root->fs_info->volume_mutex);
	ret = btrfs_rm_device(root, vol_args->name);
2576
	mutex_unlock(&root->fs_info->volume_mutex);
2577
	atomic_set(&root->fs_info->mutually_exclusive_operation_running, 0);
2578 2579 2580

out:
	kfree(vol_args);
2581
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
2582 2583 2584
	return ret;
}

J
Jan Schmidt 已提交
2585 2586
static long btrfs_ioctl_fs_info(struct btrfs_root *root, void __user *arg)
{
2587
	struct btrfs_ioctl_fs_info_args *fi_args;
J
Jan Schmidt 已提交
2588 2589 2590
	struct btrfs_device *device;
	struct btrfs_device *next;
	struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
2591
	int ret = 0;
J
Jan Schmidt 已提交
2592

2593 2594 2595 2596
	fi_args = kzalloc(sizeof(*fi_args), GFP_KERNEL);
	if (!fi_args)
		return -ENOMEM;

2597
	mutex_lock(&fs_devices->device_list_mutex);
2598 2599
	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
2600 2601

	list_for_each_entry_safe(device, next, &fs_devices->devices, dev_list) {
2602 2603
		if (device->devid > fi_args->max_id)
			fi_args->max_id = device->devid;
J
Jan Schmidt 已提交
2604 2605 2606
	}
	mutex_unlock(&fs_devices->device_list_mutex);

2607 2608 2609 2610
	fi_args->nodesize = root->fs_info->super_copy->nodesize;
	fi_args->sectorsize = root->fs_info->super_copy->sectorsize;
	fi_args->clone_alignment = root->fs_info->super_copy->sectorsize;

2611 2612
	if (copy_to_user(arg, fi_args, sizeof(*fi_args)))
		ret = -EFAULT;
J
Jan Schmidt 已提交
2613

2614 2615
	kfree(fi_args);
	return ret;
J
Jan Schmidt 已提交
2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629
}

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

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

2630
	if (!btrfs_is_empty_uuid(di_args->uuid))
J
Jan Schmidt 已提交
2631 2632 2633
		s_uuid = di_args->uuid;

	mutex_lock(&fs_devices->device_list_mutex);
2634
	dev = btrfs_find_device(root->fs_info, di_args->devid, s_uuid, NULL);
J
Jan Schmidt 已提交
2635 2636 2637 2638 2639 2640 2641 2642 2643 2644

	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));
2645
	if (dev->name) {
2646 2647 2648 2649 2650 2651
		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();
2652 2653
		di_args->path[sizeof(di_args->path) - 1] = 0;
	} else {
2654
		di_args->path[0] = '\0';
2655
	}
J
Jan Schmidt 已提交
2656 2657

out:
2658
	mutex_unlock(&fs_devices->device_list_mutex);
J
Jan Schmidt 已提交
2659 2660 2661 2662 2663 2664 2665
	if (ret == 0 && copy_to_user(arg, di_args, sizeof(*di_args)))
		ret = -EFAULT;

	kfree(di_args);
	return ret;
}

M
Mark Fasheh 已提交
2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693
static struct page *extent_same_get_page(struct inode *inode, u64 off)
{
	struct page *page;
	pgoff_t index;
	struct extent_io_tree *tree = &BTRFS_I(inode)->io_tree;

	index = off >> PAGE_CACHE_SHIFT;

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

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

	return page;
}

2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713
static inline void lock_extent_range(struct inode *inode, u64 off, u64 len)
{
	/* do any pending delalloc/csum calc on src, one way or
	   another, and lock file content */
	while (1) {
		struct btrfs_ordered_extent *ordered;
		lock_extent(&BTRFS_I(inode)->io_tree, off, off + len - 1);
		ordered = btrfs_lookup_first_ordered_extent(inode,
							    off + len - 1);
		if (!ordered &&
		    !test_range_bit(&BTRFS_I(inode)->io_tree, off,
				    off + len - 1, EXTENT_DELALLOC, 0, NULL))
			break;
		unlock_extent(&BTRFS_I(inode)->io_tree, off, off + len - 1);
		if (ordered)
			btrfs_put_ordered_extent(ordered);
		btrfs_wait_ordered_range(inode, off, len);
	}
}

M
Mark Fasheh 已提交
2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840
static void btrfs_double_unlock(struct inode *inode1, u64 loff1,
				struct inode *inode2, u64 loff2, u64 len)
{
	unlock_extent(&BTRFS_I(inode1)->io_tree, loff1, loff1 + len - 1);
	unlock_extent(&BTRFS_I(inode2)->io_tree, loff2, loff2 + len - 1);

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

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

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

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

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

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

		flush_dcache_page(src_page);
		flush_dcache_page(dst_page);

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

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

		if (ret)
			break;

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

	return ret;
}

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

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

	return 0;
}

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

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

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

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

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

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

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

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

	return ret;
}

#define BTRFS_MAX_DEDUPE_LEN	(16 * 1024 * 1024)

static long btrfs_ioctl_file_extent_same(struct file *file,
2841
			struct btrfs_ioctl_same_args __user *argp)
M
Mark Fasheh 已提交
2842
{
2843 2844
	struct btrfs_ioctl_same_args *same;
	struct btrfs_ioctl_same_extent_info *info;
2845
	struct inode *src = file_inode(file);
M
Mark Fasheh 已提交
2846 2847 2848 2849
	u64 off;
	u64 len;
	int i;
	int ret;
2850
	unsigned long size;
M
Mark Fasheh 已提交
2851 2852
	u64 bs = BTRFS_I(src)->root->fs_info->sb->s_blocksize;
	bool is_admin = capable(CAP_SYS_ADMIN);
2853
	u16 count;
M
Mark Fasheh 已提交
2854 2855 2856 2857 2858 2859 2860 2861

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

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

2862
	if (get_user(count, &argp->dest_count)) {
M
Mark Fasheh 已提交
2863 2864 2865 2866
		ret = -EFAULT;
		goto out;
	}

2867
	size = offsetof(struct btrfs_ioctl_same_args __user, info[count]);
2868

2869
	same = memdup_user(argp, size);
2870

2871 2872
	if (IS_ERR(same)) {
		ret = PTR_ERR(same);
2873 2874 2875 2876 2877
		goto out;
	}

	off = same->logical_offset;
	len = same->length;
M
Mark Fasheh 已提交
2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904

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

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

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

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

2905
	/* pre-format output fields to sane values */
2906
	for (i = 0; i < count; i++) {
2907 2908 2909
		same->info[i].bytes_deduped = 0ULL;
		same->info[i].status = 0;
	}
M
Mark Fasheh 已提交
2910

2911 2912 2913 2914
	for (i = 0, info = same->info; i < count; i++, info++) {
		struct inode *dst;
		struct fd dst_file = fdget(info->fd);
		if (!dst_file.file) {
2915
			info->status = -EBADF;
2916
			continue;
M
Mark Fasheh 已提交
2917
		}
2918
		dst = file_inode(dst_file.file);
M
Mark Fasheh 已提交
2919

2920
		if (!(is_admin || (dst_file.file->f_mode & FMODE_WRITE))) {
2921
			info->status = -EINVAL;
2922 2923 2924
		} else if (file->f_path.mnt != dst_file.file->f_path.mnt) {
			info->status = -EXDEV;
		} else if (S_ISDIR(dst->i_mode)) {
2925
			info->status = -EISDIR;
2926
		} else if (!S_ISREG(dst->i_mode)) {
2927
			info->status = -EACCES;
2928 2929 2930 2931 2932
		} else {
			info->status = btrfs_extent_same(src, off, len, dst,
							info->logical_offset);
			if (info->status == 0)
				info->bytes_deduped += len;
M
Mark Fasheh 已提交
2933
		}
2934
		fdput(dst_file);
M
Mark Fasheh 已提交
2935 2936
	}

2937 2938 2939 2940
	ret = copy_to_user(argp, same, size);
	if (ret)
		ret = -EFAULT;

M
Mark Fasheh 已提交
2941 2942 2943 2944 2945
out:
	mnt_drop_write_file(file);
	return ret;
}

J
Josef Bacik 已提交
2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980
/* Helper to check and see if this root currently has a ref on the given disk
 * bytenr.  If it does then we need to update the quota for this root.  This
 * doesn't do anything if quotas aren't enabled.
 */
static int check_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root,
		     u64 disko)
{
	struct seq_list tree_mod_seq_elem = {};
	struct ulist *roots;
	struct ulist_iterator uiter;
	struct ulist_node *root_node = NULL;
	int ret;

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

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

2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993
/**
 * btrfs_clone() - clone a range from inode file to another
 *
 * @src: Inode to clone from
 * @inode: Inode to clone to
 * @off: Offset within source to start clone from
 * @olen: Original length, passed by user, of range to clone
 * @olen_aligned: Block-aligned value of olen, extent_same uses
 *               identical values here
 * @destoff: Offset within @inode to start clone
 */
static int btrfs_clone(struct inode *src, struct inode *inode,
		       u64 off, u64 olen, u64 olen_aligned, u64 destoff)
C
Christoph Hellwig 已提交
2994 2995
{
	struct btrfs_root *root = BTRFS_I(inode)->root;
2996
	struct btrfs_path *path = NULL;
C
Christoph Hellwig 已提交
2997
	struct extent_buffer *leaf;
2998 2999
	struct btrfs_trans_handle *trans;
	char *buf = NULL;
Y
Yan Zheng 已提交
3000
	struct btrfs_key key;
C
Christoph Hellwig 已提交
3001 3002
	u32 nritems;
	int slot;
Y
Yan Zheng 已提交
3003
	int ret;
J
Josef Bacik 已提交
3004
	int no_quota;
3005
	u64 len = olen_aligned;
J
Josef Bacik 已提交
3006
	u64 last_disko = 0;
Y
Yan Zheng 已提交
3007 3008 3009 3010

	ret = -ENOMEM;
	buf = vmalloc(btrfs_level_size(root, 0));
	if (!buf)
3011
		return ret;
Y
Yan Zheng 已提交
3012 3013 3014 3015

	path = btrfs_alloc_path();
	if (!path) {
		vfree(buf);
3016
		return ret;
3017 3018
	}

3019
	path->reada = 2;
3020
	/* clone data */
L
Li Zefan 已提交
3021
	key.objectid = btrfs_ino(src);
Y
Yan Zheng 已提交
3022 3023
	key.type = BTRFS_EXTENT_DATA_KEY;
	key.offset = 0;
C
Christoph Hellwig 已提交
3024 3025 3026 3027 3028 3029

	while (1) {
		/*
		 * note the key will change type as we walk through the
		 * tree.
		 */
3030
		path->leave_spinning = 1;
3031 3032
		ret = btrfs_search_slot(NULL, BTRFS_I(src)->root, &key, path,
				0, 0);
C
Christoph Hellwig 已提交
3033 3034 3035
		if (ret < 0)
			goto out;

Y
Yan Zheng 已提交
3036
		nritems = btrfs_header_nritems(path->nodes[0]);
3037
process_slot:
J
Josef Bacik 已提交
3038
		no_quota = 1;
Y
Yan Zheng 已提交
3039
		if (path->slots[0] >= nritems) {
3040
			ret = btrfs_next_leaf(BTRFS_I(src)->root, path);
C
Christoph Hellwig 已提交
3041 3042 3043 3044
			if (ret < 0)
				goto out;
			if (ret > 0)
				break;
Y
Yan Zheng 已提交
3045
			nritems = btrfs_header_nritems(path->nodes[0]);
C
Christoph Hellwig 已提交
3046 3047 3048 3049
		}
		leaf = path->nodes[0];
		slot = path->slots[0];

Y
Yan Zheng 已提交
3050
		btrfs_item_key_to_cpu(leaf, &key, slot);
3051
		if (btrfs_key_type(&key) > BTRFS_EXTENT_DATA_KEY ||
L
Li Zefan 已提交
3052
		    key.objectid != btrfs_ino(src))
C
Christoph Hellwig 已提交
3053 3054
			break;

3055 3056 3057
		if (btrfs_key_type(&key) == BTRFS_EXTENT_DATA_KEY) {
			struct btrfs_file_extent_item *extent;
			int type;
Z
Zheng Yan 已提交
3058 3059
			u32 size;
			struct btrfs_key new_key;
3060 3061 3062
			u64 disko = 0, diskl = 0;
			u64 datao = 0, datal = 0;
			u8 comp;
3063
			u64 endoff;
Z
Zheng Yan 已提交
3064

3065 3066 3067 3068
			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);
3069 3070
			if (type == BTRFS_FILE_EXTENT_REG ||
			    type == BTRFS_FILE_EXTENT_PREALLOC) {
C
Chris Mason 已提交
3071 3072 3073 3074
				disko = btrfs_file_extent_disk_bytenr(leaf,
								      extent);
				diskl = btrfs_file_extent_disk_num_bytes(leaf,
								 extent);
3075
				datao = btrfs_file_extent_offset(leaf, extent);
C
Chris Mason 已提交
3076 3077
				datal = btrfs_file_extent_num_bytes(leaf,
								    extent);
3078 3079 3080 3081 3082
			} else if (type == BTRFS_FILE_EXTENT_INLINE) {
				/* take upper bound, may be compressed */
				datal = btrfs_file_extent_ram_bytes(leaf,
								    extent);
			}
Z
Zheng Yan 已提交
3083

3084
			if (key.offset + datal <= off ||
3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096
			    key.offset >= off + len - 1) {
				path->slots[0]++;
				goto process_slot;
			}

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

			btrfs_release_path(path);
			path->leave_spinning = 0;
3097

Z
Zheng Yan 已提交
3098
			memcpy(&new_key, &key, sizeof(new_key));
L
Li Zefan 已提交
3099
			new_key.objectid = btrfs_ino(inode);
3100 3101 3102 3103
			if (off <= key.offset)
				new_key.offset = key.offset + destoff - off;
			else
				new_key.offset = destoff;
Z
Zheng Yan 已提交
3104

3105 3106 3107 3108 3109 3110
			/*
			 * 1 - adjusting old extent (we may have to split it)
			 * 1 - add new extent
			 * 1 - inode update
			 */
			trans = btrfs_start_transaction(root, 3);
3111 3112 3113 3114 3115
			if (IS_ERR(trans)) {
				ret = PTR_ERR(trans);
				goto out;
			}

3116 3117
			if (type == BTRFS_FILE_EXTENT_REG ||
			    type == BTRFS_FILE_EXTENT_PREALLOC) {
3118 3119 3120 3121 3122 3123 3124 3125 3126 3127
				/*
				 *    a  | --- range to clone ---|  b
				 * | ------------- extent ------------- |
				 */

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

				/* substract range a */
3128 3129 3130 3131 3132
				if (off > key.offset) {
					datao += off - key.offset;
					datal -= off - key.offset;
				}

J
Josef Bacik 已提交
3133
				ret = btrfs_drop_extents(trans, root, inode,
3134 3135
							 new_key.offset,
							 new_key.offset + datal,
3136
							 1);
3137
				if (ret) {
3138
					if (ret != -EOPNOTSUPP)
3139 3140
						btrfs_abort_transaction(trans,
								root, ret);
3141 3142 3143
					btrfs_end_transaction(trans, root);
					goto out;
				}
3144

3145 3146
				ret = btrfs_insert_empty_item(trans, root, path,
							      &new_key, size);
3147 3148 3149 3150 3151 3152
				if (ret) {
					btrfs_abort_transaction(trans, root,
								ret);
					btrfs_end_transaction(trans, root);
					goto out;
				}
3153 3154 3155 3156

				leaf = path->nodes[0];
				slot = path->slots[0];
				write_extent_buffer(leaf, buf,
Z
Zheng Yan 已提交
3157 3158
					    btrfs_item_ptr_offset(leaf, slot),
					    size);
Y
Yan Zheng 已提交
3159

3160
				extent = btrfs_item_ptr(leaf, slot,
C
Christoph Hellwig 已提交
3161
						struct btrfs_file_extent_item);
3162 3163 3164 3165 3166 3167 3168 3169 3170

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

				btrfs_set_file_extent_offset(leaf, extent,
							     datao);
				btrfs_set_file_extent_num_bytes(leaf, extent,
								datal);
J
Josef Bacik 已提交
3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192

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

3193 3194
				if (disko) {
					inode_add_bytes(inode, datal);
Y
Yan Zheng 已提交
3195
					ret = btrfs_inc_extent_ref(trans, root,
3196 3197
							disko, diskl, 0,
							root->root_key.objectid,
L
Li Zefan 已提交
3198
							btrfs_ino(inode),
A
Arne Jansen 已提交
3199
							new_key.offset - datao,
J
Josef Bacik 已提交
3200
							no_quota);
3201 3202 3203 3204 3205 3206 3207 3208 3209
					if (ret) {
						btrfs_abort_transaction(trans,
									root,
									ret);
						btrfs_end_transaction(trans,
								      root);
						goto out;

					}
C
Christoph Hellwig 已提交
3210
				}
3211 3212 3213
			} else if (type == BTRFS_FILE_EXTENT_INLINE) {
				u64 skip = 0;
				u64 trim = 0;
3214 3215
				u64 aligned_end = 0;

3216 3217 3218 3219
				if (off > key.offset) {
					skip = off - key.offset;
					new_key.offset += skip;
				}
C
Chris Mason 已提交
3220

L
Liu Bo 已提交
3221 3222
				if (key.offset + datal > off + len)
					trim = key.offset + datal - (off + len);
C
Chris Mason 已提交
3223

3224 3225
				if (comp && (skip || trim)) {
					ret = -EINVAL;
3226
					btrfs_end_transaction(trans, root);
3227 3228 3229 3230
					goto out;
				}
				size -= skip + trim;
				datal -= skip + trim;
3231

3232 3233
				aligned_end = ALIGN(new_key.offset + datal,
						    root->sectorsize);
J
Josef Bacik 已提交
3234
				ret = btrfs_drop_extents(trans, root, inode,
3235
							 new_key.offset,
3236
							 aligned_end,
3237
							 1);
3238
				if (ret) {
3239
					if (ret != -EOPNOTSUPP)
3240 3241
						btrfs_abort_transaction(trans,
							root, ret);
3242 3243 3244
					btrfs_end_transaction(trans, root);
					goto out;
				}
3245

3246 3247
				ret = btrfs_insert_empty_item(trans, root, path,
							      &new_key, size);
3248 3249 3250 3251 3252 3253
				if (ret) {
					btrfs_abort_transaction(trans, root,
								ret);
					btrfs_end_transaction(trans, root);
					goto out;
				}
3254 3255

				if (skip) {
C
Chris Mason 已提交
3256 3257
					u32 start =
					  btrfs_file_extent_calc_inline_size(0);
3258 3259 3260 3261 3262 3263 3264 3265 3266 3267
					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
3268
			}
3269 3270

			btrfs_mark_buffer_dirty(leaf);
3271
			btrfs_release_path(path);
3272

3273
			inode_inc_iversion(inode);
3274
			inode->i_mtime = inode->i_ctime = CURRENT_TIME;
3275 3276 3277 3278 3279 3280 3281

			/*
			 * 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;
3282 3283
			if (endoff > destoff+olen)
				endoff = destoff+olen;
3284 3285 3286
			if (endoff > inode->i_size)
				btrfs_i_size_write(inode, endoff);

3287
			ret = btrfs_update_inode(trans, root, inode);
3288 3289 3290 3291 3292 3293
			if (ret) {
				btrfs_abort_transaction(trans, root, ret);
				btrfs_end_transaction(trans, root);
				goto out;
			}
			ret = btrfs_end_transaction(trans, root);
3294
		}
3295
		btrfs_release_path(path);
C
Christoph Hellwig 已提交
3296 3297 3298
		key.offset++;
	}
	ret = 0;
3299

C
Christoph Hellwig 已提交
3300
out:
3301
	btrfs_release_path(path);
3302 3303 3304 3305 3306 3307 3308 3309
	btrfs_free_path(path);
	vfree(buf);
	return ret;
}

static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd,
				       u64 off, u64 olen, u64 destoff)
{
A
Al Viro 已提交
3310
	struct inode *inode = file_inode(file);
3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct fd src_file;
	struct inode *src;
	int ret;
	u64 len = olen;
	u64 bs = root->fs_info->sb->s_blocksize;
	int same_inode = 0;

	/*
	 * TODO:
	 * - split compressed inline extents.  annoying: we need to
	 *   decompress into destination's address_space (the file offset
	 *   may change, so source mapping won't do), then recompress (or
	 *   otherwise reinsert) a subrange.
3325 3326 3327
	 *
	 * - split destination inode's inline extents.  The inline extents can
	 *   be either compressed or non-compressed.
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 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412
	 */

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

	if (btrfs_root_readonly(root))
		return -EROFS;

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

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

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

	src = file_inode(src_file.file);

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

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

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

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

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

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

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

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

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

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

3413 3414 3415 3416 3417 3418 3419 3420 3421 3422
	/*
	 * Lock the target range too. Right after we replace the file extent
	 * items in the fs tree (which now point to the cloned data), we might
	 * have a worker replace them with extent items relative to a write
	 * operation that was issued before this clone operation (i.e. confront
	 * with inode.c:btrfs_finish_ordered_io).
	 */
	if (same_inode) {
		u64 lock_start = min_t(u64, off, destoff);
		u64 lock_len = max_t(u64, off, destoff) + len - lock_start;
3423

3424 3425 3426 3427 3428
		lock_extent_range(src, lock_start, lock_len);
	} else {
		lock_extent_range(src, off, len);
		lock_extent_range(inode, destoff, len);
	}
3429 3430 3431

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

3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447
	if (same_inode) {
		u64 lock_start = min_t(u64, off, destoff);
		u64 lock_end = max_t(u64, off, destoff) + len - 1;

		unlock_extent(&BTRFS_I(src)->io_tree, lock_start, lock_end);
	} else {
		unlock_extent(&BTRFS_I(src)->io_tree, off, off + len - 1);
		unlock_extent(&BTRFS_I(inode)->io_tree, destoff,
			      destoff + len - 1);
	}
	/*
	 * Truncate page cache pages so that future reads will see the cloned
	 * data immediately and not the previous data.
	 */
	truncate_inode_pages_range(&inode->i_data, destoff,
				   PAGE_CACHE_ALIGN(destoff + len) - 1);
C
Christoph Hellwig 已提交
3448
out_unlock:
3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459
	if (!same_inode) {
		if (inode < src) {
			mutex_unlock(&src->i_mutex);
			mutex_unlock(&inode->i_mutex);
		} else {
			mutex_unlock(&inode->i_mutex);
			mutex_unlock(&src->i_mutex);
		}
	} else {
		mutex_unlock(&src->i_mutex);
	}
C
Christoph Hellwig 已提交
3460
out_fput:
3461
	fdput(src_file);
3462
out_drop_write:
A
Al Viro 已提交
3463
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
3464 3465 3466
	return ret;
}

3467
static long btrfs_ioctl_clone_range(struct file *file, void __user *argp)
3468 3469 3470
{
	struct btrfs_ioctl_clone_range_args args;

3471
	if (copy_from_user(&args, argp, sizeof(args)))
3472 3473 3474 3475 3476
		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
3477 3478 3479 3480 3481 3482
/*
 * 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.
 */
3483
static long btrfs_ioctl_trans_start(struct file *file)
C
Christoph Hellwig 已提交
3484
{
A
Al Viro 已提交
3485
	struct inode *inode = file_inode(file);
C
Christoph Hellwig 已提交
3486 3487
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;
3488
	int ret;
C
Christoph Hellwig 已提交
3489

3490
	ret = -EPERM;
3491
	if (!capable(CAP_SYS_ADMIN))
3492
		goto out;
3493

3494 3495
	ret = -EINPROGRESS;
	if (file->private_data)
C
Christoph Hellwig 已提交
3496
		goto out;
3497

L
Li Zefan 已提交
3498 3499 3500 3501
	ret = -EROFS;
	if (btrfs_root_readonly(root))
		goto out;

3502
	ret = mnt_want_write_file(file);
Y
Yan Zheng 已提交
3503 3504 3505
	if (ret)
		goto out;

J
Josef Bacik 已提交
3506
	atomic_inc(&root->fs_info->open_ioctl_trans);
3507

3508
	ret = -ENOMEM;
3509
	trans = btrfs_start_ioctl_transaction(root);
3510
	if (IS_ERR(trans))
3511 3512 3513 3514 3515 3516
		goto out_drop;

	file->private_data = trans;
	return 0;

out_drop:
J
Josef Bacik 已提交
3517
	atomic_dec(&root->fs_info->open_ioctl_trans);
A
Al Viro 已提交
3518
	mnt_drop_write_file(file);
C
Christoph Hellwig 已提交
3519 3520 3521 3522
out:
	return ret;
}

3523 3524
static long btrfs_ioctl_default_subvol(struct file *file, void __user *argp)
{
A
Al Viro 已提交
3525
	struct inode *inode = file_inode(file);
3526 3527 3528 3529 3530 3531 3532 3533 3534
	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;
3535
	int ret;
3536 3537 3538 3539

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

3540 3541 3542 3543 3544 3545 3546 3547
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;

	if (copy_from_user(&objectid, argp, sizeof(objectid))) {
		ret = -EFAULT;
		goto out;
	}
3548 3549

	if (!objectid)
3550
		objectid = BTRFS_FS_TREE_OBJECTID;
3551 3552 3553 3554 3555 3556

	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);
3557 3558 3559 3560
	if (IS_ERR(new_root)) {
		ret = PTR_ERR(new_root);
		goto out;
	}
3561 3562

	path = btrfs_alloc_path();
3563 3564 3565 3566
	if (!path) {
		ret = -ENOMEM;
		goto out;
	}
3567 3568 3569
	path->leave_spinning = 1;

	trans = btrfs_start_transaction(root, 1);
3570
	if (IS_ERR(trans)) {
3571
		btrfs_free_path(path);
3572 3573
		ret = PTR_ERR(trans);
		goto out;
3574 3575
	}

3576
	dir_id = btrfs_super_root_dir(root->fs_info->super_copy);
3577 3578
	di = btrfs_lookup_dir_item(trans, root->fs_info->tree_root, path,
				   dir_id, "default", 7, 1);
3579
	if (IS_ERR_OR_NULL(di)) {
3580 3581
		btrfs_free_path(path);
		btrfs_end_transaction(trans, root);
3582 3583
		btrfs_err(new_root->fs_info, "Umm, you don't have the default dir"
			   "item, this isn't going to work");
3584 3585
		ret = -ENOENT;
		goto out;
3586 3587 3588 3589 3590 3591 3592
	}

	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);

3593
	btrfs_set_fs_incompat(root->fs_info, DEFAULT_SUBVOL);
3594
	btrfs_end_transaction(trans, root);
3595 3596 3597
out:
	mnt_drop_write_file(file);
	return ret;
3598 3599
}

3600 3601
void btrfs_get_block_group_info(struct list_head *groups_list,
				struct btrfs_ioctl_space_info *space)
3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 3613 3614 3615
{
	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);
	}
}

3616
static long btrfs_ioctl_space_info(struct btrfs_root *root, void __user *arg)
J
Josef Bacik 已提交
3617 3618 3619 3620
{
	struct btrfs_ioctl_space_args space_args;
	struct btrfs_ioctl_space_info space;
	struct btrfs_ioctl_space_info *dest;
3621
	struct btrfs_ioctl_space_info *dest_orig;
3622
	struct btrfs_ioctl_space_info __user *user_dest;
J
Josef Bacik 已提交
3623
	struct btrfs_space_info *info;
3624 3625 3626 3627 3628
	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;
3629
	int alloc_size;
J
Josef Bacik 已提交
3630
	int ret = 0;
3631
	u64 slot_count = 0;
3632
	int i, c;
J
Josef Bacik 已提交
3633 3634 3635 3636 3637 3638

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

3639 3640 3641 3642 3643 3644 3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662
	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);
	}
3663

3664 3665 3666 3667 3668
	/*
	 * Global block reserve, exported as a space_info
	 */
	slot_count++;

3669 3670 3671 3672 3673
	/* space_slots == 0 means they are asking for a count */
	if (space_args.space_slots == 0) {
		space_args.total_spaces = slot_count;
		goto out;
	}
3674

3675
	slot_count = min_t(u64, space_args.space_slots, slot_count);
3676

3677
	alloc_size = sizeof(*dest) * slot_count;
3678

3679 3680 3681 3682 3683 3684
	/* 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
3685
	space_args.total_spaces = 0;
3686 3687 3688 3689
	dest = kmalloc(alloc_size, GFP_NOFS);
	if (!dest)
		return -ENOMEM;
	dest_orig = dest;
J
Josef Bacik 已提交
3690

3691
	/* now we have a buffer to copy into */
3692 3693 3694
	for (i = 0; i < num_types; i++) {
		struct btrfs_space_info *tmp;

3695 3696 3697
		if (!slot_count)
			break;

3698 3699 3700 3701 3702 3703 3704 3705 3706 3707
		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();
3708

3709 3710 3711 3712 3713
		if (!info)
			continue;
		down_read(&info->groups_sem);
		for (c = 0; c < BTRFS_NR_RAID_TYPES; c++) {
			if (!list_empty(&info->block_groups[c])) {
3714 3715
				btrfs_get_block_group_info(
					&info->block_groups[c], &space);
3716 3717 3718
				memcpy(dest, &space, sizeof(space));
				dest++;
				space_args.total_spaces++;
3719
				slot_count--;
3720
			}
3721 3722
			if (!slot_count)
				break;
3723 3724
		}
		up_read(&info->groups_sem);
J
Josef Bacik 已提交
3725 3726
	}

3727 3728 3729 3730 3731 3732 3733 3734 3735 3736 3737 3738 3739 3740 3741
	/*
	 * Add global block reserve
	 */
	if (slot_count) {
		struct btrfs_block_rsv *block_rsv = &root->fs_info->global_block_rsv;

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

D
Daniel J Blueman 已提交
3742
	user_dest = (struct btrfs_ioctl_space_info __user *)
3743 3744 3745 3746 3747 3748 3749 3750
		(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
3751 3752 3753 3754 3755
		ret = -EFAULT;

	return ret;
}

C
Christoph Hellwig 已提交
3756 3757 3758 3759 3760 3761 3762 3763
/*
 * there are many ways the trans_start and trans_end ioctls can lead
 * to deadlocks.  They should only be used by applications that
 * basically own the machine, and have a very in depth understanding
 * of all the possible deadlocks and enospc problems.
 */
long btrfs_ioctl_trans_end(struct file *file)
{
A
Al Viro 已提交
3764
	struct inode *inode = file_inode(file);
C
Christoph Hellwig 已提交
3765 3766 3767 3768
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;

	trans = file->private_data;
3769 3770
	if (!trans)
		return -EINVAL;
C
Christoph Hellwig 已提交
3771
	file->private_data = NULL;
3772

3773 3774
	btrfs_end_transaction(trans, root);

J
Josef Bacik 已提交
3775
	atomic_dec(&root->fs_info->open_ioctl_trans);
3776

A
Al Viro 已提交
3777
	mnt_drop_write_file(file);
3778
	return 0;
C
Christoph Hellwig 已提交
3779 3780
}

3781 3782
static noinline long btrfs_ioctl_start_sync(struct btrfs_root *root,
					    void __user *argp)
3783 3784 3785
{
	struct btrfs_trans_handle *trans;
	u64 transid;
T
Tsutomu Itoh 已提交
3786
	int ret;
3787

M
Miao Xie 已提交
3788
	trans = btrfs_attach_transaction_barrier(root);
3789 3790 3791 3792 3793 3794 3795 3796
	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;
	}
3797
	transid = trans->transid;
T
Tsutomu Itoh 已提交
3798
	ret = btrfs_commit_transaction_async(trans, root, 0);
3799 3800
	if (ret) {
		btrfs_end_transaction(trans, root);
T
Tsutomu Itoh 已提交
3801
		return ret;
3802
	}
3803
out:
3804 3805 3806 3807 3808 3809
	if (argp)
		if (copy_to_user(argp, &transid, sizeof(transid)))
			return -EFAULT;
	return 0;
}

3810 3811
static noinline long btrfs_ioctl_wait_sync(struct btrfs_root *root,
					   void __user *argp)
3812 3813 3814 3815 3816 3817 3818 3819 3820 3821 3822 3823
{
	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
3824
static long btrfs_ioctl_scrub(struct file *file, void __user *arg)
J
Jan Schmidt 已提交
3825
{
A
Al Viro 已提交
3826
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
J
Jan Schmidt 已提交
3827
	struct btrfs_ioctl_scrub_args *sa;
M
Miao Xie 已提交
3828
	int ret;
J
Jan Schmidt 已提交
3829 3830 3831 3832 3833 3834 3835 3836

	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
3837 3838 3839 3840 3841 3842
	if (!(sa->flags & BTRFS_SCRUB_READONLY)) {
		ret = mnt_want_write_file(file);
		if (ret)
			goto out;
	}

3843
	ret = btrfs_scrub_dev(root->fs_info, sa->devid, sa->start, sa->end,
3844 3845
			      &sa->progress, sa->flags & BTRFS_SCRUB_READONLY,
			      0);
J
Jan Schmidt 已提交
3846 3847 3848 3849

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

M
Miao Xie 已提交
3850 3851 3852
	if (!(sa->flags & BTRFS_SCRUB_READONLY))
		mnt_drop_write_file(file);
out:
J
Jan Schmidt 已提交
3853 3854 3855 3856 3857 3858 3859 3860 3861
	kfree(sa);
	return ret;
}

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

3862
	return btrfs_scrub_cancel(root->fs_info);
J
Jan Schmidt 已提交
3863 3864 3865 3866 3867 3868 3869 3870 3871 3872 3873 3874 3875 3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886
}

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;
}

3887
static long btrfs_ioctl_get_dev_stats(struct btrfs_root *root,
3888
				      void __user *arg)
3889 3890 3891 3892 3893 3894 3895 3896
{
	struct btrfs_ioctl_get_dev_stats *sa;
	int ret;

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

3897 3898 3899 3900 3901 3902
	if ((sa->flags & BTRFS_DEV_STATS_RESET) && !capable(CAP_SYS_ADMIN)) {
		kfree(sa);
		return -EPERM;
	}

	ret = btrfs_get_dev_stats(root, sa);
3903 3904 3905 3906 3907 3908 3909 3910

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

	kfree(sa);
	return ret;
}

3911 3912 3913 3914 3915 3916 3917 3918 3919 3920 3921 3922 3923 3924
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:
3925 3926 3927 3928
		if (root->fs_info->sb->s_flags & MS_RDONLY) {
			ret = -EROFS;
			goto out;
		}
3929 3930 3931
		if (atomic_xchg(
			&root->fs_info->mutually_exclusive_operation_running,
			1)) {
3932
			ret = BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
3933 3934 3935 3936 3937 3938 3939 3940 3941 3942 3943 3944 3945 3946 3947 3948 3949 3950 3951 3952 3953
		} 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;
3954
out:
3955 3956 3957 3958
	kfree(p);
	return ret;
}

3959 3960 3961 3962
static long btrfs_ioctl_ino_to_path(struct btrfs_root *root, void __user *arg)
{
	int ret = 0;
	int i;
3963
	u64 rel_ptr;
3964
	int size;
3965
	struct btrfs_ioctl_ino_path_args *ipa = NULL;
3966 3967 3968
	struct inode_fs_paths *ipath = NULL;
	struct btrfs_path *path;

3969
	if (!capable(CAP_DAC_READ_SEARCH))
3970 3971 3972 3973 3974 3975 3976 3977 3978 3979 3980 3981 3982 3983 3984 3985 3986 3987 3988 3989 3990 3991 3992 3993 3994 3995 3996 3997
		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) {
3998 3999
		rel_ptr = ipath->fspath->val[i] -
			  (u64)(unsigned long)ipath->fspath->val;
4000
		ipath->fspath->val[i] = rel_ptr;
4001 4002
	}

4003 4004
	ret = copy_to_user((void *)(unsigned long)ipa->fspath,
			   (void *)(unsigned long)ipath->fspath, size);
4005 4006 4007 4008 4009 4010 4011 4012 4013 4014 4015 4016 4017 4018 4019 4020 4021 4022 4023 4024 4025 4026 4027 4028 4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044 4045 4046 4047 4048 4049 4050 4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 4062
	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;
	}

4063
	size = min_t(u32, loi->size, 64 * 1024);
4064 4065 4066 4067 4068 4069 4070
	inodes = init_data_container(size);
	if (IS_ERR(inodes)) {
		ret = PTR_ERR(inodes);
		inodes = NULL;
		goto out;
	}

L
Liu Bo 已提交
4071 4072 4073
	ret = iterate_inodes_from_logical(loi->logical, root->fs_info, path,
					  build_ino_list, inodes);
	if (ret == -EINVAL)
4074 4075 4076 4077
		ret = -ENOENT;
	if (ret < 0)
		goto out;

4078 4079
	ret = copy_to_user((void *)(unsigned long)loi->inodes,
			   (void *)(unsigned long)inodes, size);
4080 4081 4082 4083 4084
	if (ret)
		ret = -EFAULT;

out:
	btrfs_free_path(path);
4085
	vfree(inodes);
4086 4087 4088 4089 4090
	kfree(loi);

	return ret;
}

4091
void update_ioctl_balance_args(struct btrfs_fs_info *fs_info, int lock,
4092 4093 4094 4095 4096 4097
			       struct btrfs_ioctl_balance_args *bargs)
{
	struct btrfs_balance_control *bctl = fs_info->balance_ctl;

	bargs->flags = bctl->flags;

4098 4099 4100 4101
	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;
4102 4103
	if (atomic_read(&fs_info->balance_cancel_req))
		bargs->state |= BTRFS_BALANCE_STATE_CANCEL_REQ;
4104

4105 4106 4107
	memcpy(&bargs->data, &bctl->data, sizeof(bargs->data));
	memcpy(&bargs->meta, &bctl->meta, sizeof(bargs->meta));
	memcpy(&bargs->sys, &bctl->sys, sizeof(bargs->sys));
4108 4109 4110 4111 4112 4113 4114 4115

	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));
	}
4116 4117
}

4118
static long btrfs_ioctl_balance(struct file *file, void __user *arg)
4119
{
A
Al Viro 已提交
4120
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
4121 4122 4123
	struct btrfs_fs_info *fs_info = root->fs_info;
	struct btrfs_ioctl_balance_args *bargs;
	struct btrfs_balance_control *bctl;
4124
	bool need_unlock; /* for mut. excl. ops lock */
4125 4126 4127 4128 4129
	int ret;

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

4130
	ret = mnt_want_write_file(file);
4131 4132 4133
	if (ret)
		return ret;

4134 4135 4136 4137 4138 4139 4140 4141 4142 4143 4144 4145 4146 4147
again:
	if (!atomic_xchg(&fs_info->mutually_exclusive_operation_running, 1)) {
		mutex_lock(&fs_info->volume_mutex);
		mutex_lock(&fs_info->balance_mutex);
		need_unlock = true;
		goto locked;
	}

	/*
	 * mut. excl. ops lock is locked.  Three possibilites:
	 *   (1) some other op is running
	 *   (2) balance is running
	 *   (3) balance is paused -- special case (think resume)
	 */
4148
	mutex_lock(&fs_info->balance_mutex);
4149 4150 4151 4152 4153 4154 4155 4156 4157 4158 4159 4160 4161 4162 4163 4164 4165 4166 4167 4168 4169 4170 4171 4172 4173 4174 4175
	if (fs_info->balance_ctl) {
		/* this is either (2) or (3) */
		if (!atomic_read(&fs_info->balance_running)) {
			mutex_unlock(&fs_info->balance_mutex);
			if (!mutex_trylock(&fs_info->volume_mutex))
				goto again;
			mutex_lock(&fs_info->balance_mutex);

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

			mutex_unlock(&fs_info->balance_mutex);
			mutex_unlock(&fs_info->volume_mutex);
			goto again;
		} else {
			/* this is (2) */
			mutex_unlock(&fs_info->balance_mutex);
			ret = -EINPROGRESS;
			goto out;
		}
	} else {
		/* this is (1) */
		mutex_unlock(&fs_info->balance_mutex);
4176
		ret = BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
4177 4178 4179 4180 4181
		goto out;
	}

locked:
	BUG_ON(!atomic_read(&fs_info->mutually_exclusive_operation_running));
4182 4183 4184 4185 4186

	if (arg) {
		bargs = memdup_user(arg, sizeof(*bargs));
		if (IS_ERR(bargs)) {
			ret = PTR_ERR(bargs);
4187
			goto out_unlock;
4188
		}
4189 4190 4191 4192 4193 4194 4195 4196 4197 4198 4199 4200 4201 4202

		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;
		}
4203 4204 4205 4206
	} else {
		bargs = NULL;
	}

4207
	if (fs_info->balance_ctl) {
4208 4209 4210 4211
		ret = -EINPROGRESS;
		goto out_bargs;
	}

4212 4213 4214 4215 4216 4217 4218 4219 4220 4221 4222 4223 4224
	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;
4225 4226 4227
	} else {
		/* balance everything - no filters */
		bctl->flags |= BTRFS_BALANCE_TYPE_MASK;
4228 4229
	}

4230
do_balance:
4231
	/*
4232 4233 4234 4235 4236
	 * Ownership of bctl and mutually_exclusive_operation_running
	 * goes to to btrfs_balance.  bctl is freed in __cancel_balance,
	 * or, if restriper was paused all the way until unmount, in
	 * free_fs_info.  mutually_exclusive_operation_running is
	 * cleared in __cancel_balance.
4237
	 */
4238 4239 4240 4241
	need_unlock = false;

	ret = btrfs_balance(bctl, bargs);

4242 4243 4244 4245 4246 4247 4248
	if (arg) {
		if (copy_to_user(arg, bargs, sizeof(*bargs)))
			ret = -EFAULT;
	}

out_bargs:
	kfree(bargs);
4249
out_unlock:
4250 4251
	mutex_unlock(&fs_info->balance_mutex);
	mutex_unlock(&fs_info->volume_mutex);
4252 4253 4254
	if (need_unlock)
		atomic_set(&fs_info->mutually_exclusive_operation_running, 0);
out:
4255
	mnt_drop_write_file(file);
4256 4257 4258
	return ret;
}

4259 4260 4261 4262 4263 4264 4265 4266
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);
4267 4268
	case BTRFS_BALANCE_CTL_CANCEL:
		return btrfs_cancel_balance(root->fs_info);
4269 4270 4271 4272 4273
	}

	return -EINVAL;
}

4274 4275 4276 4277 4278 4279 4280 4281 4282 4283 4284 4285 4286 4287 4288 4289 4290 4291 4292 4293 4294 4295 4296 4297 4298 4299 4300 4301 4302 4303 4304 4305 4306
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;
}

4307
static long btrfs_ioctl_quota_ctl(struct file *file, void __user *arg)
A
Arne Jansen 已提交
4308
{
A
Al Viro 已提交
4309
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
A
Arne Jansen 已提交
4310 4311 4312 4313 4314 4315 4316 4317
	struct btrfs_ioctl_quota_ctl_args *sa;
	struct btrfs_trans_handle *trans = NULL;
	int ret;
	int err;

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

4318 4319 4320
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
4321 4322

	sa = memdup_user(arg, sizeof(*sa));
4323 4324 4325 4326
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
4327

4328
	down_write(&root->fs_info->subvol_sem);
J
Jan Schmidt 已提交
4329 4330 4331 4332
	trans = btrfs_start_transaction(root->fs_info->tree_root, 2);
	if (IS_ERR(trans)) {
		ret = PTR_ERR(trans);
		goto out;
A
Arne Jansen 已提交
4333 4334 4335 4336 4337 4338 4339 4340 4341 4342 4343 4344 4345 4346
	}

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

J
Jan Schmidt 已提交
4347 4348 4349
	err = btrfs_commit_transaction(trans, root->fs_info->tree_root);
	if (err && !ret)
		ret = err;
A
Arne Jansen 已提交
4350 4351
out:
	kfree(sa);
4352
	up_write(&root->fs_info->subvol_sem);
4353 4354
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
4355 4356 4357
	return ret;
}

4358
static long btrfs_ioctl_qgroup_assign(struct file *file, void __user *arg)
A
Arne Jansen 已提交
4359
{
A
Al Viro 已提交
4360
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
A
Arne Jansen 已提交
4361 4362 4363 4364 4365 4366 4367 4368
	struct btrfs_ioctl_qgroup_assign_args *sa;
	struct btrfs_trans_handle *trans;
	int ret;
	int err;

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

4369 4370 4371
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
4372 4373

	sa = memdup_user(arg, sizeof(*sa));
4374 4375 4376 4377
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
4378 4379 4380 4381 4382 4383 4384 4385 4386 4387 4388 4389 4390 4391 4392 4393 4394 4395 4396 4397 4398 4399

	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);
4400 4401
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
4402 4403 4404
	return ret;
}

4405
static long btrfs_ioctl_qgroup_create(struct file *file, void __user *arg)
A
Arne Jansen 已提交
4406
{
A
Al Viro 已提交
4407
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
A
Arne Jansen 已提交
4408 4409 4410 4411 4412 4413 4414 4415
	struct btrfs_ioctl_qgroup_create_args *sa;
	struct btrfs_trans_handle *trans;
	int ret;
	int err;

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

4416 4417 4418
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
4419 4420

	sa = memdup_user(arg, sizeof(*sa));
4421 4422 4423 4424
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
4425

M
Miao Xie 已提交
4426 4427 4428 4429 4430
	if (!sa->qgroupid) {
		ret = -EINVAL;
		goto out;
	}

A
Arne Jansen 已提交
4431 4432 4433 4434 4435 4436 4437 4438 4439 4440 4441 4442 4443 4444 4445 4446 4447 4448 4449 4450
	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);
4451 4452
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
4453 4454 4455
	return ret;
}

4456
static long btrfs_ioctl_qgroup_limit(struct file *file, void __user *arg)
A
Arne Jansen 已提交
4457
{
A
Al Viro 已提交
4458
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
A
Arne Jansen 已提交
4459 4460 4461 4462 4463 4464 4465 4466 4467
	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;

4468 4469 4470
	ret = mnt_want_write_file(file);
	if (ret)
		return ret;
A
Arne Jansen 已提交
4471 4472

	sa = memdup_user(arg, sizeof(*sa));
4473 4474 4475 4476
	if (IS_ERR(sa)) {
		ret = PTR_ERR(sa);
		goto drop_write;
	}
A
Arne Jansen 已提交
4477 4478 4479 4480 4481 4482 4483 4484 4485 4486 4487 4488 4489 4490 4491 4492 4493 4494 4495 4496 4497 4498

	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);
4499 4500
drop_write:
	mnt_drop_write_file(file);
A
Arne Jansen 已提交
4501 4502 4503
	return ret;
}

J
Jan Schmidt 已提交
4504 4505
static long btrfs_ioctl_quota_rescan(struct file *file, void __user *arg)
{
A
Al Viro 已提交
4506
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
J
Jan Schmidt 已提交
4507 4508 4509 4510 4511 4512 4513 4514 4515 4516 4517 4518 4519 4520 4521 4522 4523 4524 4525 4526 4527 4528 4529 4530 4531 4532 4533 4534 4535 4536 4537 4538
	struct btrfs_ioctl_quota_rescan_args *qsa;
	int ret;

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

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

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

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

	ret = btrfs_qgroup_rescan(root->fs_info);

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

static long btrfs_ioctl_quota_rescan_status(struct file *file, void __user *arg)
{
A
Al Viro 已提交
4539
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
J
Jan Schmidt 已提交
4540 4541 4542 4543 4544 4545 4546 4547 4548 4549 4550 4551 4552 4553 4554 4555 4556 4557 4558 4559 4560 4561
	struct btrfs_ioctl_quota_rescan_args *qsa;
	int ret = 0;

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

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

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

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

	kfree(qsa);
	return ret;
}

4562 4563
static long btrfs_ioctl_quota_rescan_wait(struct file *file, void __user *arg)
{
A
Al Viro 已提交
4564
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
4565 4566 4567 4568 4569 4570 4571

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

	return btrfs_qgroup_wait_for_completion(root->fs_info);
}

4572 4573
static long _btrfs_ioctl_set_received_subvol(struct file *file,
					    struct btrfs_ioctl_received_subvol_args *sa)
4574
{
A
Al Viro 已提交
4575
	struct inode *inode = file_inode(file);
4576 4577 4578 4579 4580
	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;
4581
	int received_uuid_changed;
4582

4583 4584 4585
	if (!inode_owner_or_capable(inode))
		return -EPERM;

4586 4587 4588 4589 4590 4591 4592 4593 4594 4595 4596 4597 4598 4599 4600 4601
	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;
	}

4602 4603 4604 4605 4606
	/*
	 * 1 - root item
	 * 2 - uuid items (received uuid + subvol uuid)
	 */
	trans = btrfs_start_transaction(root, 3);
4607 4608 4609 4610 4611 4612 4613 4614 4615 4616
	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;

4617 4618 4619 4620 4621 4622 4623 4624
	received_uuid_changed = memcmp(root_item->received_uuid, sa->uuid,
				       BTRFS_UUID_SIZE);
	if (received_uuid_changed &&
	    !btrfs_is_empty_uuid(root_item->received_uuid))
		btrfs_uuid_tree_rem(trans, root->fs_info->uuid_root,
				    root_item->received_uuid,
				    BTRFS_UUID_KEY_RECEIVED_SUBVOL,
				    root->root_key.objectid);
4625 4626 4627
	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);
4628 4629 4630 4631
	btrfs_set_stack_timespec_sec(&root_item->stime, sa->stime.sec);
	btrfs_set_stack_timespec_nsec(&root_item->stime, sa->stime.nsec);
	btrfs_set_stack_timespec_sec(&root_item->rtime, sa->rtime.sec);
	btrfs_set_stack_timespec_nsec(&root_item->rtime, sa->rtime.nsec);
4632 4633 4634 4635 4636 4637

	ret = btrfs_update_root(trans, root->fs_info->tree_root,
				&root->root_key, &root->root_item);
	if (ret < 0) {
		btrfs_end_transaction(trans, root);
		goto out;
4638 4639 4640 4641 4642 4643 4644 4645
	}
	if (received_uuid_changed && !btrfs_is_empty_uuid(sa->uuid)) {
		ret = btrfs_uuid_tree_add(trans, root->fs_info->uuid_root,
					  sa->uuid,
					  BTRFS_UUID_KEY_RECEIVED_SUBVOL,
					  root->root_key.objectid);
		if (ret < 0 && ret != -EEXIST) {
			btrfs_abort_transaction(trans, root, ret);
4646
			goto out;
4647 4648 4649 4650 4651 4652
		}
	}
	ret = btrfs_commit_transaction(trans, root);
	if (ret < 0) {
		btrfs_abort_transaction(trans, root, ret);
		goto out;
4653 4654
	}

4655 4656 4657 4658 4659 4660 4661 4662 4663 4664 4665 4666 4667 4668 4669 4670 4671 4672 4673 4674 4675 4676
out:
	up_write(&root->fs_info->subvol_sem);
	mnt_drop_write_file(file);
	return ret;
}

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

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

	args64 = kmalloc(sizeof(*args64), GFP_NOFS);
4677 4678
	if (!args64) {
		ret = -ENOMEM;
4679 4680 4681 4682 4683 4684 4685 4686 4687 4688 4689 4690 4691 4692 4693 4694 4695 4696 4697 4698 4699 4700 4701 4702 4703 4704 4705 4706 4707 4708 4709 4710 4711 4712 4713 4714 4715 4716 4717 4718 4719 4720 4721 4722 4723 4724 4725 4726 4727 4728 4729 4730 4731 4732
		goto out;
	}

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

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

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

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

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

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

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

	ret = _btrfs_ioctl_set_received_subvol(file, sa);

	if (ret)
		goto out;

4733 4734 4735 4736 4737 4738 4739 4740 4741
	ret = copy_to_user(arg, sa, sizeof(*sa));
	if (ret)
		ret = -EFAULT;

out:
	kfree(sa);
	return ret;
}

4742 4743
static int btrfs_ioctl_get_fslabel(struct file *file, void __user *arg)
{
A
Al Viro 已提交
4744
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
4745
	size_t len;
4746
	int ret;
4747 4748 4749 4750 4751 4752 4753
	char label[BTRFS_LABEL_SIZE];

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

	len = strnlen(label, BTRFS_LABEL_SIZE);
4754 4755

	if (len == BTRFS_LABEL_SIZE) {
4756 4757
		btrfs_warn(root->fs_info,
			"label is too long, return the first %zu bytes", --len);
4758 4759 4760 4761 4762 4763 4764
	}

	ret = copy_to_user(arg, label, len);

	return ret ? -EFAULT : 0;
}

4765 4766
static int btrfs_ioctl_set_fslabel(struct file *file, void __user *arg)
{
A
Al Viro 已提交
4767
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
4768 4769 4770 4771 4772 4773 4774 4775 4776 4777 4778 4779
	struct btrfs_super_block *super_block = root->fs_info->super_copy;
	struct btrfs_trans_handle *trans;
	char label[BTRFS_LABEL_SIZE];
	int ret;

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

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

	if (strnlen(label, BTRFS_LABEL_SIZE) == BTRFS_LABEL_SIZE) {
4780
		btrfs_err(root->fs_info, "unable to set label with more than %d bytes",
4781 4782 4783 4784 4785 4786 4787 4788 4789 4790 4791 4792 4793 4794
		       BTRFS_LABEL_SIZE - 1);
		return -EINVAL;
	}

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

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

4795
	spin_lock(&root->fs_info->super_lock);
4796
	strcpy(super_block->label, label);
4797
	spin_unlock(&root->fs_info->super_lock);
4798
	ret = btrfs_commit_transaction(trans, root);
4799 4800 4801 4802 4803 4804

out_unlock:
	mnt_drop_write_file(file);
	return ret;
}

4805 4806 4807 4808 4809 4810 4811 4812 4813 4814 4815 4816 4817 4818 4819 4820 4821 4822 4823 4824 4825 4826 4827 4828 4829 4830 4831 4832 4833 4834 4835 4836 4837 4838 4839 4840
#define INIT_FEATURE_FLAGS(suffix) \
	{ .compat_flags = BTRFS_FEATURE_COMPAT_##suffix, \
	  .compat_ro_flags = BTRFS_FEATURE_COMPAT_RO_##suffix, \
	  .incompat_flags = BTRFS_FEATURE_INCOMPAT_##suffix }

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

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

	return 0;
}

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

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

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

	return 0;
}

4841 4842
static int check_feature_bits(struct btrfs_root *root,
			      enum btrfs_feature_set set,
4843 4844 4845
			      u64 change_mask, u64 flags, u64 supported_flags,
			      u64 safe_set, u64 safe_clear)
{
4846 4847
	const char *type = btrfs_feature_set_names[set];
	char *names;
4848 4849 4850 4851 4852 4853
	u64 disallowed, unsupported;
	u64 set_mask = flags & change_mask;
	u64 clear_mask = ~flags & change_mask;

	unsupported = set_mask & ~supported_flags;
	if (unsupported) {
4854 4855 4856 4857 4858 4859 4860 4861
		names = btrfs_printable_features(set, unsupported);
		if (names) {
			btrfs_warn(root->fs_info,
			   "this kernel does not support the %s feature bit%s",
			   names, strchr(names, ',') ? "s" : "");
			kfree(names);
		} else
			btrfs_warn(root->fs_info,
4862 4863 4864 4865 4866 4867 4868
			   "this kernel does not support %s bits 0x%llx",
			   type, unsupported);
		return -EOPNOTSUPP;
	}

	disallowed = set_mask & ~safe_set;
	if (disallowed) {
4869 4870 4871 4872 4873 4874 4875 4876
		names = btrfs_printable_features(set, disallowed);
		if (names) {
			btrfs_warn(root->fs_info,
			   "can't set the %s feature bit%s while mounted",
			   names, strchr(names, ',') ? "s" : "");
			kfree(names);
		} else
			btrfs_warn(root->fs_info,
4877 4878 4879 4880 4881 4882 4883
			   "can't set %s bits 0x%llx while mounted",
			   type, disallowed);
		return -EPERM;
	}

	disallowed = clear_mask & ~safe_clear;
	if (disallowed) {
4884 4885 4886 4887 4888 4889 4890 4891
		names = btrfs_printable_features(set, disallowed);
		if (names) {
			btrfs_warn(root->fs_info,
			   "can't clear the %s feature bit%s while mounted",
			   names, strchr(names, ',') ? "s" : "");
			kfree(names);
		} else
			btrfs_warn(root->fs_info,
4892 4893 4894 4895 4896 4897 4898 4899 4900
			   "can't clear %s bits 0x%llx while mounted",
			   type, disallowed);
		return -EPERM;
	}

	return 0;
}

#define check_feature(root, change_mask, flags, mask_base)	\
4901
check_feature_bits(root, FEAT_##mask_base, change_mask, flags,	\
4902 4903 4904 4905 4906 4907 4908 4909 4910 4911 4912 4913 4914 4915 4916 4917 4918 4919 4920 4921 4922 4923 4924 4925 4926 4927 4928 4929 4930 4931 4932 4933 4934 4935 4936 4937 4938 4939 4940
		   BTRFS_FEATURE_ ## mask_base ## _SUPP,	\
		   BTRFS_FEATURE_ ## mask_base ## _SAFE_SET,	\
		   BTRFS_FEATURE_ ## mask_base ## _SAFE_CLEAR)

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

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

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

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

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

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

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

4941
	trans = btrfs_start_transaction(root, 0);
4942 4943 4944 4945 4946 4947 4948 4949 4950 4951 4952 4953 4954 4955 4956 4957 4958 4959 4960 4961
	if (IS_ERR(trans))
		return PTR_ERR(trans);

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

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

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

4962
	return btrfs_commit_transaction(trans, root);
4963 4964
}

C
Christoph Hellwig 已提交
4965 4966 4967
long btrfs_ioctl(struct file *file, unsigned int
		cmd, unsigned long arg)
{
A
Al Viro 已提交
4968
	struct btrfs_root *root = BTRFS_I(file_inode(file))->root;
4969
	void __user *argp = (void __user *)arg;
C
Christoph Hellwig 已提交
4970 4971

	switch (cmd) {
4972 4973 4974 4975 4976 4977
	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);
4978 4979
	case FITRIM:
		return btrfs_ioctl_fitrim(file, argp);
C
Christoph Hellwig 已提交
4980
	case BTRFS_IOC_SNAP_CREATE:
4981
		return btrfs_ioctl_snap_create(file, argp, 0);
4982
	case BTRFS_IOC_SNAP_CREATE_V2:
4983
		return btrfs_ioctl_snap_create_v2(file, argp, 0);
4984
	case BTRFS_IOC_SUBVOL_CREATE:
4985
		return btrfs_ioctl_snap_create(file, argp, 1);
A
Arne Jansen 已提交
4986 4987
	case BTRFS_IOC_SUBVOL_CREATE_V2:
		return btrfs_ioctl_snap_create_v2(file, argp, 1);
4988 4989
	case BTRFS_IOC_SNAP_DESTROY:
		return btrfs_ioctl_snap_destroy(file, argp);
4990 4991 4992 4993
	case BTRFS_IOC_SUBVOL_GETFLAGS:
		return btrfs_ioctl_subvol_getflags(file, argp);
	case BTRFS_IOC_SUBVOL_SETFLAGS:
		return btrfs_ioctl_subvol_setflags(file, argp);
4994 4995
	case BTRFS_IOC_DEFAULT_SUBVOL:
		return btrfs_ioctl_default_subvol(file, argp);
C
Christoph Hellwig 已提交
4996
	case BTRFS_IOC_DEFRAG:
C
Chris Mason 已提交
4997 4998 4999
		return btrfs_ioctl_defrag(file, NULL);
	case BTRFS_IOC_DEFRAG_RANGE:
		return btrfs_ioctl_defrag(file, argp);
C
Christoph Hellwig 已提交
5000
	case BTRFS_IOC_RESIZE:
5001
		return btrfs_ioctl_resize(file, argp);
C
Christoph Hellwig 已提交
5002
	case BTRFS_IOC_ADD_DEV:
5003
		return btrfs_ioctl_add_dev(root, argp);
C
Christoph Hellwig 已提交
5004
	case BTRFS_IOC_RM_DEV:
5005
		return btrfs_ioctl_rm_dev(file, argp);
J
Jan Schmidt 已提交
5006 5007 5008 5009
	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
5010
	case BTRFS_IOC_BALANCE:
5011
		return btrfs_ioctl_balance(file, NULL);
C
Christoph Hellwig 已提交
5012
	case BTRFS_IOC_CLONE:
5013 5014
		return btrfs_ioctl_clone(file, arg, 0, 0, 0);
	case BTRFS_IOC_CLONE_RANGE:
5015
		return btrfs_ioctl_clone_range(file, argp);
C
Christoph Hellwig 已提交
5016 5017 5018 5019
	case BTRFS_IOC_TRANS_START:
		return btrfs_ioctl_trans_start(file);
	case BTRFS_IOC_TRANS_END:
		return btrfs_ioctl_trans_end(file);
5020 5021 5022 5023
	case BTRFS_IOC_TREE_SEARCH:
		return btrfs_ioctl_tree_search(file, argp);
	case BTRFS_IOC_INO_LOOKUP:
		return btrfs_ioctl_ino_lookup(file, argp);
5024 5025 5026 5027
	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
5028 5029
	case BTRFS_IOC_SPACE_INFO:
		return btrfs_ioctl_space_info(root, argp);
5030 5031 5032
	case BTRFS_IOC_SYNC: {
		int ret;

5033
		ret = btrfs_start_delalloc_roots(root->fs_info, 0, -1);
5034 5035 5036 5037 5038
		if (ret)
			return ret;
		ret = btrfs_sync_fs(file->f_dentry->d_sb, 1);
		return ret;
	}
5039
	case BTRFS_IOC_START_SYNC:
5040
		return btrfs_ioctl_start_sync(root, argp);
5041
	case BTRFS_IOC_WAIT_SYNC:
5042
		return btrfs_ioctl_wait_sync(root, argp);
J
Jan Schmidt 已提交
5043
	case BTRFS_IOC_SCRUB:
M
Miao Xie 已提交
5044
		return btrfs_ioctl_scrub(file, argp);
J
Jan Schmidt 已提交
5045 5046 5047 5048
	case BTRFS_IOC_SCRUB_CANCEL:
		return btrfs_ioctl_scrub_cancel(root, argp);
	case BTRFS_IOC_SCRUB_PROGRESS:
		return btrfs_ioctl_scrub_progress(root, argp);
5049
	case BTRFS_IOC_BALANCE_V2:
5050
		return btrfs_ioctl_balance(file, argp);
5051 5052
	case BTRFS_IOC_BALANCE_CTL:
		return btrfs_ioctl_balance_ctl(root, arg);
5053 5054
	case BTRFS_IOC_BALANCE_PROGRESS:
		return btrfs_ioctl_balance_progress(root, argp);
5055 5056
	case BTRFS_IOC_SET_RECEIVED_SUBVOL:
		return btrfs_ioctl_set_received_subvol(file, argp);
5057 5058 5059 5060
#ifdef CONFIG_64BIT
	case BTRFS_IOC_SET_RECEIVED_SUBVOL_32:
		return btrfs_ioctl_set_received_subvol_32(file, argp);
#endif
5061 5062
	case BTRFS_IOC_SEND:
		return btrfs_ioctl_send(file, argp);
5063
	case BTRFS_IOC_GET_DEV_STATS:
5064
		return btrfs_ioctl_get_dev_stats(root, argp);
A
Arne Jansen 已提交
5065
	case BTRFS_IOC_QUOTA_CTL:
5066
		return btrfs_ioctl_quota_ctl(file, argp);
A
Arne Jansen 已提交
5067
	case BTRFS_IOC_QGROUP_ASSIGN:
5068
		return btrfs_ioctl_qgroup_assign(file, argp);
A
Arne Jansen 已提交
5069
	case BTRFS_IOC_QGROUP_CREATE:
5070
		return btrfs_ioctl_qgroup_create(file, argp);
A
Arne Jansen 已提交
5071
	case BTRFS_IOC_QGROUP_LIMIT:
5072
		return btrfs_ioctl_qgroup_limit(file, argp);
J
Jan Schmidt 已提交
5073 5074 5075 5076
	case BTRFS_IOC_QUOTA_RESCAN:
		return btrfs_ioctl_quota_rescan(file, argp);
	case BTRFS_IOC_QUOTA_RESCAN_STATUS:
		return btrfs_ioctl_quota_rescan_status(file, argp);
5077 5078
	case BTRFS_IOC_QUOTA_RESCAN_WAIT:
		return btrfs_ioctl_quota_rescan_wait(file, argp);
5079 5080
	case BTRFS_IOC_DEV_REPLACE:
		return btrfs_ioctl_dev_replace(root, argp);
5081 5082
	case BTRFS_IOC_GET_FSLABEL:
		return btrfs_ioctl_get_fslabel(file, argp);
5083 5084
	case BTRFS_IOC_SET_FSLABEL:
		return btrfs_ioctl_set_fslabel(file, argp);
M
Mark Fasheh 已提交
5085 5086
	case BTRFS_IOC_FILE_EXTENT_SAME:
		return btrfs_ioctl_file_extent_same(file, argp);
5087 5088 5089 5090 5091 5092
	case BTRFS_IOC_GET_SUPPORTED_FEATURES:
		return btrfs_ioctl_get_supported_features(file, argp);
	case BTRFS_IOC_GET_FEATURES:
		return btrfs_ioctl_get_features(file, argp);
	case BTRFS_IOC_SET_FEATURES:
		return btrfs_ioctl_set_features(file, argp);
C
Christoph Hellwig 已提交
5093 5094 5095 5096
	}

	return -ENOTTY;
}