super.c 34.3 KB
Newer Older
C
Chris Mason 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
/*
 * 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.
 */

Y
Yan 已提交
19
#include <linux/blkdev.h>
20
#include <linux/module.h>
C
Chris Mason 已提交
21
#include <linux/buffer_head.h>
22 23 24 25 26
#include <linux/fs.h>
#include <linux/pagemap.h>
#include <linux/highmem.h>
#include <linux/time.h>
#include <linux/init.h>
E
Eric Paris 已提交
27
#include <linux/seq_file.h>
28 29
#include <linux/string.h>
#include <linux/backing-dev.h>
Y
Yan 已提交
30
#include <linux/mount.h>
C
Chris Mason 已提交
31
#include <linux/mpage.h>
C
Chris Mason 已提交
32 33
#include <linux/swap.h>
#include <linux/writeback.h>
C
Chris Mason 已提交
34
#include <linux/statfs.h>
C
Chris Mason 已提交
35
#include <linux/compat.h>
36
#include <linux/parser.h>
37
#include <linux/ctype.h>
38
#include <linux/namei.h>
39
#include <linux/miscdevice.h>
40
#include <linux/magic.h>
41
#include <linux/slab.h>
D
Dan Magenheimer 已提交
42
#include <linux/cleancache.h>
43
#include <linux/mnt_namespace.h>
C
Chris Mason 已提交
44
#include "compat.h"
45
#include "delayed-inode.h"
46
#include "ctree.h"
C
Chris Mason 已提交
47
#include "disk-io.h"
48
#include "transaction.h"
C
Chris Mason 已提交
49
#include "btrfs_inode.h"
C
Chris Mason 已提交
50
#include "ioctl.h"
C
Chris Mason 已提交
51
#include "print-tree.h"
J
Josef Bacik 已提交
52
#include "xattr.h"
53
#include "volumes.h"
54
#include "version.h"
B
Balaji Rao 已提交
55
#include "export.h"
C
Chris Mason 已提交
56
#include "compression.h"
57

58 59 60
#define CREATE_TRACE_POINTS
#include <trace/events/btrfs.h>

61
static const struct super_operations btrfs_super_ops;
62
static struct file_system_type btrfs_fs_type;
C
Chris Mason 已提交
63

L
liubo 已提交
64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147
static const char *btrfs_decode_error(struct btrfs_fs_info *fs_info, int errno,
				      char nbuf[16])
{
	char *errstr = NULL;

	switch (errno) {
	case -EIO:
		errstr = "IO failure";
		break;
	case -ENOMEM:
		errstr = "Out of memory";
		break;
	case -EROFS:
		errstr = "Readonly filesystem";
		break;
	default:
		if (nbuf) {
			if (snprintf(nbuf, 16, "error %d", -errno) >= 0)
				errstr = nbuf;
		}
		break;
	}

	return errstr;
}

static void __save_error_info(struct btrfs_fs_info *fs_info)
{
	/*
	 * today we only save the error info into ram.  Long term we'll
	 * also send it down to the disk
	 */
	fs_info->fs_state = BTRFS_SUPER_FLAG_ERROR;
}

/* NOTE:
 *	We move write_super stuff at umount in order to avoid deadlock
 *	for umount hold all lock.
 */
static void save_error_info(struct btrfs_fs_info *fs_info)
{
	__save_error_info(fs_info);
}

/* btrfs handle error by forcing the filesystem readonly */
static void btrfs_handle_error(struct btrfs_fs_info *fs_info)
{
	struct super_block *sb = fs_info->sb;

	if (sb->s_flags & MS_RDONLY)
		return;

	if (fs_info->fs_state & BTRFS_SUPER_FLAG_ERROR) {
		sb->s_flags |= MS_RDONLY;
		printk(KERN_INFO "btrfs is forced readonly\n");
	}
}

/*
 * __btrfs_std_error decodes expected errors from the caller and
 * invokes the approciate error response.
 */
void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function,
		     unsigned int line, int errno)
{
	struct super_block *sb = fs_info->sb;
	char nbuf[16];
	const char *errstr;

	/*
	 * Special case: if the error is EROFS, and we're already
	 * under MS_RDONLY, then it is safe here.
	 */
	if (errno == -EROFS && (sb->s_flags & MS_RDONLY))
		return;

	errstr = btrfs_decode_error(fs_info, errno, nbuf);
	printk(KERN_CRIT "BTRFS error (device %s) in %s:%d: %s\n",
		sb->s_id, function, line, errstr);
	save_error_info(fs_info);

	btrfs_handle_error(fs_info);
}

C
Chris Mason 已提交
148
static void btrfs_put_super(struct super_block *sb)
C
Chris Mason 已提交
149
{
C
Chris Mason 已提交
150
	struct btrfs_root *root = btrfs_sb(sb);
C
Chris Mason 已提交
151 152
	int ret;

C
Chris Mason 已提交
153 154
	ret = close_ctree(root);
	sb->s_fs_info = NULL;
155 156

	(void)ret; /* FIXME: need to fix VFS to return error? */
C
Chris Mason 已提交
157 158
}

159
enum {
160
	Opt_degraded, Opt_subvol, Opt_subvolid, Opt_device, Opt_nodatasum,
161 162
	Opt_nodatacow, Opt_max_inline, Opt_alloc_start, Opt_nobarrier, Opt_ssd,
	Opt_nossd, Opt_ssd_spread, Opt_thread_pool, Opt_noacl, Opt_compress,
163 164
	Opt_compress_type, Opt_compress_force, Opt_compress_force_type,
	Opt_notreelog, Opt_ratio, Opt_flushoncommit, Opt_discard,
165
	Opt_space_cache, Opt_clear_cache, Opt_user_subvol_rm_allowed,
C
Chris Mason 已提交
166
	Opt_enospc_debug, Opt_subvolrootid, Opt_defrag,
167
	Opt_inode_cache, Opt_no_space_cache, Opt_err,
168 169 170
};

static match_table_t tokens = {
171
	{Opt_degraded, "degraded"},
172
	{Opt_subvol, "subvol=%s"},
173
	{Opt_subvolid, "subvolid=%d"},
174
	{Opt_device, "device=%s"},
175
	{Opt_nodatasum, "nodatasum"},
176
	{Opt_nodatacow, "nodatacow"},
177
	{Opt_nobarrier, "nobarrier"},
178
	{Opt_max_inline, "max_inline=%s"},
179
	{Opt_alloc_start, "alloc_start=%s"},
180
	{Opt_thread_pool, "thread_pool=%d"},
C
Chris Mason 已提交
181
	{Opt_compress, "compress"},
182
	{Opt_compress_type, "compress=%s"},
C
Chris Mason 已提交
183
	{Opt_compress_force, "compress-force"},
184
	{Opt_compress_force_type, "compress-force=%s"},
185
	{Opt_ssd, "ssd"},
186
	{Opt_ssd_spread, "ssd_spread"},
C
Chris Mason 已提交
187
	{Opt_nossd, "nossd"},
J
Josef Bacik 已提交
188
	{Opt_noacl, "noacl"},
S
Sage Weil 已提交
189
	{Opt_notreelog, "notreelog"},
190
	{Opt_flushoncommit, "flushoncommit"},
191
	{Opt_ratio, "metadata_ratio=%d"},
C
Christoph Hellwig 已提交
192
	{Opt_discard, "discard"},
193
	{Opt_space_cache, "space_cache"},
194
	{Opt_clear_cache, "clear_cache"},
195
	{Opt_user_subvol_rm_allowed, "user_subvol_rm_allowed"},
196
	{Opt_enospc_debug, "enospc_debug"},
197
	{Opt_subvolrootid, "subvolrootid=%d"},
C
Chris Mason 已提交
198
	{Opt_defrag, "autodefrag"},
C
Chris Mason 已提交
199
	{Opt_inode_cache, "inode_cache"},
200
	{Opt_no_space_cache, "no_space_cache"},
J
Josef Bacik 已提交
201
	{Opt_err, NULL},
202 203
};

204 205 206 207 208
/*
 * Regular mount options parser.  Everything that is needed only when
 * reading in a new superblock is parsed here.
 */
int btrfs_parse_options(struct btrfs_root *root, char *options)
209
{
210
	struct btrfs_fs_info *info = root->fs_info;
211
	substring_t args[MAX_OPT_ARGS];
212 213
	char *p, *num, *orig = NULL;
	u64 cache_gen;
214
	int intarg;
S
Sage Weil 已提交
215
	int ret = 0;
216 217
	char *compress_type;
	bool compress_force = false;
218

219 220 221 222
	cache_gen = btrfs_super_cache_generation(&root->fs_info->super_copy);
	if (cache_gen)
		btrfs_set_opt(info->mount_opt, SPACE_CACHE);

223
	if (!options)
224
		goto out;
225

226 227 228 229 230 231 232 233
	/*
	 * strsep changes the string, duplicate it because parse_options
	 * gets called twice
	 */
	options = kstrdup(options, GFP_NOFS);
	if (!options)
		return -ENOMEM;

234
	orig = options;
235

236
	while ((p = strsep(&options, ",")) != NULL) {
237 238 239 240 241 242
		int token;
		if (!*p)
			continue;

		token = match_token(p, tokens, args);
		switch (token) {
243
		case Opt_degraded:
244 245
			printk(KERN_INFO "btrfs: allowing degraded mounts\n");
			btrfs_set_opt(info->mount_opt, DEGRADED);
246
			break;
247
		case Opt_subvol:
248
		case Opt_subvolid:
249
		case Opt_subvolrootid:
250
		case Opt_device:
251
			/*
252
			 * These are parsed by btrfs_parse_early_options
253 254
			 * and can be happily ignored here.
			 */
255 256
			break;
		case Opt_nodatasum:
257
			printk(KERN_INFO "btrfs: setting nodatasum\n");
258
			btrfs_set_opt(info->mount_opt, NODATASUM);
259 260
			break;
		case Opt_nodatacow:
261 262 263
			printk(KERN_INFO "btrfs: setting nodatacow\n");
			btrfs_set_opt(info->mount_opt, NODATACOW);
			btrfs_set_opt(info->mount_opt, NODATASUM);
264
			break;
C
Chris Mason 已提交
265
		case Opt_compress_force:
266 267 268 269 270 271 272 273 274
		case Opt_compress_force_type:
			compress_force = true;
		case Opt_compress:
		case Opt_compress_type:
			if (token == Opt_compress ||
			    token == Opt_compress_force ||
			    strcmp(args[0].from, "zlib") == 0) {
				compress_type = "zlib";
				info->compress_type = BTRFS_COMPRESS_ZLIB;
L
Li Zefan 已提交
275 276 277
			} else if (strcmp(args[0].from, "lzo") == 0) {
				compress_type = "lzo";
				info->compress_type = BTRFS_COMPRESS_LZO;
278 279 280 281 282
			} else {
				ret = -EINVAL;
				goto out;
			}

C
Chris Mason 已提交
283
			btrfs_set_opt(info->mount_opt, COMPRESS);
284 285 286 287 288 289 290
			if (compress_force) {
				btrfs_set_opt(info->mount_opt, FORCE_COMPRESS);
				pr_info("btrfs: force %s compression\n",
					compress_type);
			} else
				pr_info("btrfs: use %s compression\n",
					compress_type);
C
Chris Mason 已提交
291
			break;
292
		case Opt_ssd:
293 294
			printk(KERN_INFO "btrfs: use ssd allocation scheme\n");
			btrfs_set_opt(info->mount_opt, SSD);
295
			break;
296 297 298 299 300 301
		case Opt_ssd_spread:
			printk(KERN_INFO "btrfs: use spread ssd "
			       "allocation scheme\n");
			btrfs_set_opt(info->mount_opt, SSD);
			btrfs_set_opt(info->mount_opt, SSD_SPREAD);
			break;
C
Chris Mason 已提交
302
		case Opt_nossd:
303 304
			printk(KERN_INFO "btrfs: not using ssd allocation "
			       "scheme\n");
C
Chris Mason 已提交
305
			btrfs_set_opt(info->mount_opt, NOSSD);
C
Chris Mason 已提交
306
			btrfs_clear_opt(info->mount_opt, SSD);
307
			btrfs_clear_opt(info->mount_opt, SSD_SPREAD);
C
Chris Mason 已提交
308
			break;
309
		case Opt_nobarrier:
310 311
			printk(KERN_INFO "btrfs: turning off barriers\n");
			btrfs_set_opt(info->mount_opt, NOBARRIER);
312
			break;
313 314 315 316 317 318 319 320 321
		case Opt_thread_pool:
			intarg = 0;
			match_int(&args[0], &intarg);
			if (intarg) {
				info->thread_pool_size = intarg;
				printk(KERN_INFO "btrfs: thread pool %d\n",
				       info->thread_pool_size);
			}
			break;
322
		case Opt_max_inline:
323 324
			num = match_strdup(&args[0]);
			if (num) {
A
Akinobu Mita 已提交
325
				info->max_inline = memparse(num, NULL);
326 327
				kfree(num);

C
Chris Mason 已提交
328 329 330 331 332
				if (info->max_inline) {
					info->max_inline = max_t(u64,
						info->max_inline,
						root->sectorsize);
				}
333
				printk(KERN_INFO "btrfs: max_inline at %llu\n",
334
					(unsigned long long)info->max_inline);
335 336
			}
			break;
337
		case Opt_alloc_start:
338 339
			num = match_strdup(&args[0]);
			if (num) {
A
Akinobu Mita 已提交
340
				info->alloc_start = memparse(num, NULL);
341 342 343
				kfree(num);
				printk(KERN_INFO
					"btrfs: allocations start at %llu\n",
344
					(unsigned long long)info->alloc_start);
345 346
			}
			break;
J
Josef Bacik 已提交
347 348 349
		case Opt_noacl:
			root->fs_info->sb->s_flags &= ~MS_POSIXACL;
			break;
S
Sage Weil 已提交
350 351 352 353
		case Opt_notreelog:
			printk(KERN_INFO "btrfs: disabling tree log\n");
			btrfs_set_opt(info->mount_opt, NOTREELOG);
			break;
354 355 356 357
		case Opt_flushoncommit:
			printk(KERN_INFO "btrfs: turning on flush-on-commit\n");
			btrfs_set_opt(info->mount_opt, FLUSHONCOMMIT);
			break;
358 359 360 361 362 363 364 365 366
		case Opt_ratio:
			intarg = 0;
			match_int(&args[0], &intarg);
			if (intarg) {
				info->metadata_ratio = intarg;
				printk(KERN_INFO "btrfs: metadata ratio %d\n",
				       info->metadata_ratio);
			}
			break;
C
Christoph Hellwig 已提交
367 368 369
		case Opt_discard:
			btrfs_set_opt(info->mount_opt, DISCARD);
			break;
370 371
		case Opt_space_cache:
			btrfs_set_opt(info->mount_opt, SPACE_CACHE);
372
			break;
373 374 375 376
		case Opt_no_space_cache:
			printk(KERN_INFO "btrfs: disabling disk space caching\n");
			btrfs_clear_opt(info->mount_opt, SPACE_CACHE);
			break;
C
Chris Mason 已提交
377 378 379 380
		case Opt_inode_cache:
			printk(KERN_INFO "btrfs: enabling inode map caching\n");
			btrfs_set_opt(info->mount_opt, INODE_MAP_CACHE);
			break;
381 382 383
		case Opt_clear_cache:
			printk(KERN_INFO "btrfs: force clearing of disk cache\n");
			btrfs_set_opt(info->mount_opt, CLEAR_CACHE);
384
			break;
385 386 387
		case Opt_user_subvol_rm_allowed:
			btrfs_set_opt(info->mount_opt, USER_SUBVOL_RM_ALLOWED);
			break;
388 389 390
		case Opt_enospc_debug:
			btrfs_set_opt(info->mount_opt, ENOSPC_DEBUG);
			break;
C
Chris Mason 已提交
391 392 393 394
		case Opt_defrag:
			printk(KERN_INFO "btrfs: enabling auto defrag");
			btrfs_set_opt(info->mount_opt, AUTO_DEFRAG);
			break;
S
Sage Weil 已提交
395 396 397 398 399
		case Opt_err:
			printk(KERN_INFO "btrfs: unrecognized mount option "
			       "'%s'\n", p);
			ret = -EINVAL;
			goto out;
400
		default:
401
			break;
402 403
		}
	}
S
Sage Weil 已提交
404
out:
405 406
	if (!ret && btrfs_test_opt(root, SPACE_CACHE))
		printk(KERN_INFO "btrfs: disk space caching is enabled\n");
407
	kfree(orig);
S
Sage Weil 已提交
408
	return ret;
409 410 411 412 413 414 415 416
}

/*
 * Parse mount options that are required early in the mount process.
 *
 * All other options will be parsed on much later in the mount process and
 * only when we need to allocate a new super block.
 */
417
static int btrfs_parse_early_options(const char *options, fmode_t flags,
418
		void *holder, char **subvol_name, u64 *subvol_objectid,
419
		u64 *subvol_rootid, struct btrfs_fs_devices **fs_devices)
420 421
{
	substring_t args[MAX_OPT_ARGS];
422
	char *device_name, *opts, *orig, *p;
423
	int error = 0;
424
	int intarg;
425 426

	if (!options)
427
		return 0;
428 429 430 431 432 433 434 435

	/*
	 * strsep changes the string, duplicate it because parse_options
	 * gets called twice
	 */
	opts = kstrdup(options, GFP_KERNEL);
	if (!opts)
		return -ENOMEM;
436
	orig = opts;
437 438 439 440 441 442 443 444 445 446 447

	while ((p = strsep(&opts, ",")) != NULL) {
		int token;
		if (!*p)
			continue;

		token = match_token(p, tokens, args);
		switch (token) {
		case Opt_subvol:
			*subvol_name = match_strdup(&args[0]);
			break;
448 449
		case Opt_subvolid:
			intarg = 0;
450 451 452 453 454 455 456 457 458
			error = match_int(&args[0], &intarg);
			if (!error) {
				/* we want the original fs_tree */
				if (!intarg)
					*subvol_objectid =
						BTRFS_FS_TREE_OBJECTID;
				else
					*subvol_objectid = intarg;
			}
459
			break;
460 461 462 463 464 465 466 467 468 469 470 471
		case Opt_subvolrootid:
			intarg = 0;
			error = match_int(&args[0], &intarg);
			if (!error) {
				/* we want the original fs_tree */
				if (!intarg)
					*subvol_rootid =
						BTRFS_FS_TREE_OBJECTID;
				else
					*subvol_rootid = intarg;
			}
			break;
472
		case Opt_device:
473 474 475 476 477 478
			device_name = match_strdup(&args[0]);
			if (!device_name) {
				error = -ENOMEM;
				goto out;
			}
			error = btrfs_scan_one_device(device_name,
479
					flags, holder, fs_devices);
480
			kfree(device_name);
481
			if (error)
482
				goto out;
483
			break;
484 485 486 487 488
		default:
			break;
		}
	}

489
out:
490
	kfree(orig);
491
	return error;
492 493
}

494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528
static struct dentry *get_default_root(struct super_block *sb,
				       u64 subvol_objectid)
{
	struct btrfs_root *root = sb->s_fs_info;
	struct btrfs_root *new_root;
	struct btrfs_dir_item *di;
	struct btrfs_path *path;
	struct btrfs_key location;
	struct inode *inode;
	u64 dir_id;
	int new = 0;

	/*
	 * We have a specific subvol we want to mount, just setup location and
	 * go look up the root.
	 */
	if (subvol_objectid) {
		location.objectid = subvol_objectid;
		location.type = BTRFS_ROOT_ITEM_KEY;
		location.offset = (u64)-1;
		goto find_root;
	}

	path = btrfs_alloc_path();
	if (!path)
		return ERR_PTR(-ENOMEM);
	path->leave_spinning = 1;

	/*
	 * Find the "default" dir item which points to the root item that we
	 * will mount by default if we haven't been given a specific subvolume
	 * to mount.
	 */
	dir_id = btrfs_super_root_dir(&root->fs_info->super_copy);
	di = btrfs_lookup_dir_item(NULL, root, path, dir_id, "default", 7, 0);
529 530
	if (IS_ERR(di)) {
		btrfs_free_path(path);
531
		return ERR_CAST(di);
532
	}
533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550
	if (!di) {
		/*
		 * Ok the default dir item isn't there.  This is weird since
		 * it's always been there, but don't freak out, just try and
		 * mount to root most subvolume.
		 */
		btrfs_free_path(path);
		dir_id = BTRFS_FIRST_FREE_OBJECTID;
		new_root = root->fs_info->fs_root;
		goto setup_root;
	}

	btrfs_dir_item_key_to_cpu(path->nodes[0], di, &location);
	btrfs_free_path(path);

find_root:
	new_root = btrfs_read_fs_root_no_name(root->fs_info, &location);
	if (IS_ERR(new_root))
J
Julia Lawall 已提交
551
		return ERR_CAST(new_root);
552 553 554 555 556 557 558 559 560 561 562

	if (btrfs_root_refs(&new_root->root_item) == 0)
		return ERR_PTR(-ENOENT);

	dir_id = btrfs_root_dirid(&new_root->root_item);
setup_root:
	location.objectid = dir_id;
	location.type = BTRFS_INODE_ITEM_KEY;
	location.offset = 0;

	inode = btrfs_iget(sb, &location, new_root, &new);
563 564
	if (IS_ERR(inode))
		return ERR_CAST(inode);
565 566 567 568 569 570 571 572 573 574 575

	/*
	 * If we're just mounting the root most subvol put the inode and return
	 * a reference to the dentry.  We will have already gotten a reference
	 * to the inode in btrfs_fill_super so we're good to go.
	 */
	if (!new && sb->s_root->d_inode == inode) {
		iput(inode);
		return dget(sb->s_root);
	}

576
	return d_obtain_alias(inode);
577 578
}

C
Chris Mason 已提交
579
static int btrfs_fill_super(struct super_block *sb,
580
			    struct btrfs_fs_devices *fs_devices,
C
Chris Mason 已提交
581
			    void *data, int silent)
C
Chris Mason 已提交
582
{
C
Chris Mason 已提交
583 584
	struct inode *inode;
	struct dentry *root_dentry;
C
Chris Mason 已提交
585
	struct btrfs_root *tree_root;
586
	struct btrfs_key key;
C
Chris Mason 已提交
587
	int err;
588

C
Chris Mason 已提交
589 590 591
	sb->s_maxbytes = MAX_LFS_FILESIZE;
	sb->s_magic = BTRFS_SUPER_MAGIC;
	sb->s_op = &btrfs_super_ops;
A
Al Viro 已提交
592
	sb->s_d_op = &btrfs_dentry_operations;
B
Balaji Rao 已提交
593
	sb->s_export_op = &btrfs_export_ops;
J
Josef Bacik 已提交
594
	sb->s_xattr = btrfs_xattr_handlers;
C
Chris Mason 已提交
595
	sb->s_time_gran = 1;
C
Chris Mason 已提交
596
#ifdef CONFIG_BTRFS_FS_POSIX_ACL
J
Josef Bacik 已提交
597
	sb->s_flags |= MS_POSIXACL;
598
#endif
599

600
	tree_root = open_ctree(sb, fs_devices, (char *)data);
601

602
	if (IS_ERR(tree_root)) {
C
Chris Mason 已提交
603
		printk("btrfs: open_ctree failed\n");
604
		return PTR_ERR(tree_root);
605
	}
C
Chris Mason 已提交
606
	sb->s_fs_info = tree_root;
607

608 609 610
	key.objectid = BTRFS_FIRST_FREE_OBJECTID;
	key.type = BTRFS_INODE_ITEM_KEY;
	key.offset = 0;
611
	inode = btrfs_iget(sb, &key, tree_root->fs_info->fs_root, NULL);
612 613
	if (IS_ERR(inode)) {
		err = PTR_ERR(inode);
C
Chris Mason 已提交
614
		goto fail_close;
C
Chris Mason 已提交
615 616
	}

C
Chris Mason 已提交
617 618 619 620 621
	root_dentry = d_alloc_root(inode);
	if (!root_dentry) {
		iput(inode);
		err = -ENOMEM;
		goto fail_close;
C
Chris Mason 已提交
622
	}
623

C
Chris Mason 已提交
624
	sb->s_root = root_dentry;
C
Chris Mason 已提交
625 626

	save_mount_options(sb, data);
D
Dan Magenheimer 已提交
627
	cleancache_init_fs(sb);
C
Chris Mason 已提交
628
	return 0;
C
Chris Mason 已提交
629 630 631 632

fail_close:
	close_ctree(tree_root);
	return err;
C
Chris Mason 已提交
633 634
}

S
Sage Weil 已提交
635
int btrfs_sync_fs(struct super_block *sb, int wait)
C
Chris Mason 已提交
636 637
{
	struct btrfs_trans_handle *trans;
638
	struct btrfs_root *root = btrfs_sb(sb);
C
Chris Mason 已提交
639
	int ret;
C
Chris Mason 已提交
640

641 642
	trace_btrfs_sync_fs(wait);

C
Chris Mason 已提交
643 644 645 646
	if (!wait) {
		filemap_flush(root->fs_info->btree_inode->i_mapping);
		return 0;
	}
647

Y
Yan, Zheng 已提交
648 649
	btrfs_start_delalloc_inodes(root, 0);
	btrfs_wait_ordered_extents(root, 0, 0);
650

651
	trans = btrfs_start_transaction(root, 0);
652 653
	if (IS_ERR(trans))
		return PTR_ERR(trans);
C
Chris Mason 已提交
654
	ret = btrfs_commit_transaction(trans, root);
655
	return ret;
C
Chris Mason 已提交
656 657
}

E
Eric Paris 已提交
658 659 660 661
static int btrfs_show_options(struct seq_file *seq, struct vfsmount *vfs)
{
	struct btrfs_root *root = btrfs_sb(vfs->mnt_sb);
	struct btrfs_fs_info *info = root->fs_info;
T
Tsutomu Itoh 已提交
662
	char *compress_type;
E
Eric Paris 已提交
663 664 665 666 667 668 669 670 671 672

	if (btrfs_test_opt(root, DEGRADED))
		seq_puts(seq, ",degraded");
	if (btrfs_test_opt(root, NODATASUM))
		seq_puts(seq, ",nodatasum");
	if (btrfs_test_opt(root, NODATACOW))
		seq_puts(seq, ",nodatacow");
	if (btrfs_test_opt(root, NOBARRIER))
		seq_puts(seq, ",nobarrier");
	if (info->max_inline != 8192 * 1024)
673 674
		seq_printf(seq, ",max_inline=%llu",
			   (unsigned long long)info->max_inline);
E
Eric Paris 已提交
675
	if (info->alloc_start != 0)
676 677
		seq_printf(seq, ",alloc_start=%llu",
			   (unsigned long long)info->alloc_start);
E
Eric Paris 已提交
678 679 680
	if (info->thread_pool_size !=  min_t(unsigned long,
					     num_online_cpus() + 2, 8))
		seq_printf(seq, ",thread_pool=%d", info->thread_pool_size);
T
Tsutomu Itoh 已提交
681 682 683 684 685 686 687 688 689 690
	if (btrfs_test_opt(root, COMPRESS)) {
		if (info->compress_type == BTRFS_COMPRESS_ZLIB)
			compress_type = "zlib";
		else
			compress_type = "lzo";
		if (btrfs_test_opt(root, FORCE_COMPRESS))
			seq_printf(seq, ",compress-force=%s", compress_type);
		else
			seq_printf(seq, ",compress=%s", compress_type);
	}
C
Chris Mason 已提交
691 692
	if (btrfs_test_opt(root, NOSSD))
		seq_puts(seq, ",nossd");
693 694 695
	if (btrfs_test_opt(root, SSD_SPREAD))
		seq_puts(seq, ",ssd_spread");
	else if (btrfs_test_opt(root, SSD))
E
Eric Paris 已提交
696
		seq_puts(seq, ",ssd");
S
Sage Weil 已提交
697
	if (btrfs_test_opt(root, NOTREELOG))
698
		seq_puts(seq, ",notreelog");
699
	if (btrfs_test_opt(root, FLUSHONCOMMIT))
700
		seq_puts(seq, ",flushoncommit");
701 702
	if (btrfs_test_opt(root, DISCARD))
		seq_puts(seq, ",discard");
E
Eric Paris 已提交
703 704
	if (!(root->fs_info->sb->s_flags & MS_POSIXACL))
		seq_puts(seq, ",noacl");
T
Tsutomu Itoh 已提交
705 706
	if (btrfs_test_opt(root, SPACE_CACHE))
		seq_puts(seq, ",space_cache");
707 708
	else
		seq_puts(seq, ",no_space_cache");
T
Tsutomu Itoh 已提交
709 710 711 712
	if (btrfs_test_opt(root, CLEAR_CACHE))
		seq_puts(seq, ",clear_cache");
	if (btrfs_test_opt(root, USER_SUBVOL_RM_ALLOWED))
		seq_puts(seq, ",user_subvol_rm_allowed");
713 714 715 716 717 718
	if (btrfs_test_opt(root, ENOSPC_DEBUG))
		seq_puts(seq, ",enospc_debug");
	if (btrfs_test_opt(root, AUTO_DEFRAG))
		seq_puts(seq, ",autodefrag");
	if (btrfs_test_opt(root, INODE_MAP_CACHE))
		seq_puts(seq, ",inode_cache");
E
Eric Paris 已提交
719 720 721
	return 0;
}

722
static int btrfs_test_super(struct super_block *s, void *data)
Y
Yan 已提交
723
{
724
	struct btrfs_root *test_root = data;
725
	struct btrfs_root *root = btrfs_sb(s);
Y
Yan 已提交
726

727 728 729 730 731 732
	/*
	 * If this super block is going away, return false as it
	 * can't match as an existing super block.
	 */
	if (!atomic_read(&s->s_active))
		return 0;
733
	return root->fs_info->fs_devices == test_root->fs_info->fs_devices;
Y
Yan 已提交
734 735
}

736 737 738 739 740
static int btrfs_set_super(struct super_block *s, void *data)
{
	s->s_fs_info = data;

	return set_anon_super(s, data);
Y
Yan 已提交
741 742
}

743 744 745 746 747 748 749 750 751 752
/*
 * subvolumes are identified by ino 256
 */
static inline int is_subvolume_inode(struct inode *inode)
{
	if (inode && inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)
		return 1;
	return 0;
}

753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855
/*
 * This will strip out the subvol=%s argument for an argument string and add
 * subvolid=0 to make sure we get the actual tree root for path walking to the
 * subvol we want.
 */
static char *setup_root_args(char *args)
{
	unsigned copied = 0;
	unsigned len = strlen(args) + 2;
	char *pos;
	char *ret;

	/*
	 * We need the same args as before, but minus
	 *
	 * subvol=a
	 *
	 * and add
	 *
	 * subvolid=0
	 *
	 * which is a difference of 2 characters, so we allocate strlen(args) +
	 * 2 characters.
	 */
	ret = kzalloc(len * sizeof(char), GFP_NOFS);
	if (!ret)
		return NULL;
	pos = strstr(args, "subvol=");

	/* This shouldn't happen, but just in case.. */
	if (!pos) {
		kfree(ret);
		return NULL;
	}

	/*
	 * The subvol=<> arg is not at the front of the string, copy everybody
	 * up to that into ret.
	 */
	if (pos != args) {
		*pos = '\0';
		strcpy(ret, args);
		copied += strlen(args);
		pos++;
	}

	strncpy(ret + copied, "subvolid=0", len - copied);

	/* Length of subvolid=0 */
	copied += 10;

	/*
	 * If there is no , after the subvol= option then we know there's no
	 * other options and we can just return.
	 */
	pos = strchr(pos, ',');
	if (!pos)
		return ret;

	/* Copy the rest of the arguments into our buffer */
	strncpy(ret + copied, pos, len - copied);
	copied += strlen(pos);

	return ret;
}

static struct dentry *mount_subvol(const char *subvol_name, int flags,
				   const char *device_name, char *data)
{
	struct super_block *s;
	struct dentry *root;
	struct vfsmount *mnt;
	struct mnt_namespace *ns_private;
	char *newargs;
	struct path path;
	int error;

	newargs = setup_root_args(data);
	if (!newargs)
		return ERR_PTR(-ENOMEM);
	mnt = vfs_kern_mount(&btrfs_fs_type, flags, device_name,
			     newargs);
	kfree(newargs);
	if (IS_ERR(mnt))
		return ERR_CAST(mnt);

	ns_private = create_mnt_ns(mnt);
	if (IS_ERR(ns_private)) {
		mntput(mnt);
		return ERR_CAST(ns_private);
	}

	/*
	 * This will trigger the automount of the subvol so we can just
	 * drop the mnt we have here and return the dentry that we
	 * found.
	 */
	error = vfs_path_lookup(mnt->mnt_root, mnt, subvol_name,
				LOOKUP_FOLLOW, &path);
	put_mnt_ns(ns_private);
	if (error)
		return ERR_PTR(error);

856 857 858 859 860 861 862 863 864
	if (!is_subvolume_inode(path.dentry->d_inode)) {
		path_put(&path);
		mntput(mnt);
		error = -EINVAL;
		printk(KERN_ERR "btrfs: '%s' is not a valid subvolume\n",
				subvol_name);
		return ERR_PTR(-EINVAL);
	}

865 866 867 868 869 870 871 872 873
	/* Get a ref to the sb and the dentry we found and return it */
	s = path.mnt->mnt_sb;
	atomic_inc(&s->s_active);
	root = dget(path.dentry);
	path_put(&path);
	down_write(&s->s_umount);

	return root;
}
874

875 876 877 878 879 880
/*
 * Find a superblock for the given device / mount point.
 *
 * Note:  This is based on get_sb_bdev from fs/super.c with a few additions
 *	  for multiple device setup.  Make sure to keep it in sync.
 */
A
Al Viro 已提交
881
static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags,
882
		const char *device_name, void *data)
Y
Yan 已提交
883 884 885 886
{
	struct block_device *bdev = NULL;
	struct super_block *s;
	struct dentry *root;
887
	struct btrfs_fs_devices *fs_devices = NULL;
888 889
	struct btrfs_root *tree_root = NULL;
	struct btrfs_fs_info *fs_info = NULL;
890
	fmode_t mode = FMODE_READ;
891 892
	char *subvol_name = NULL;
	u64 subvol_objectid = 0;
893
	u64 subvol_rootid = 0;
Y
Yan 已提交
894 895
	int error = 0;

896 897 898 899
	if (!(flags & MS_RDONLY))
		mode |= FMODE_WRITE;

	error = btrfs_parse_early_options(data, mode, fs_type,
900
					  &subvol_name, &subvol_objectid,
901
					  &subvol_rootid, &fs_devices);
902
	if (error)
A
Al Viro 已提交
903
		return ERR_PTR(error);
904

905 906 907 908 909 910
	if (subvol_name) {
		root = mount_subvol(subvol_name, flags, device_name, data);
		kfree(subvol_name);
		return root;
	}

911
	error = btrfs_scan_one_device(device_name, mode, fs_type, &fs_devices);
912
	if (error)
913
		return ERR_PTR(error);
Y
Yan 已提交
914

915
	error = btrfs_open_devices(fs_devices, mode, fs_type);
916
	if (error)
917
		return ERR_PTR(error);
918

Y
Yan Zheng 已提交
919 920 921 922 923
	if (!(flags & MS_RDONLY) && fs_devices->rw_devices == 0) {
		error = -EACCES;
		goto error_close_devices;
	}

924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939
	/*
	 * Setup a dummy root and fs_info for test/set super.  This is because
	 * we don't actually fill this stuff out until open_ctree, but we need
	 * it for searching for existing supers, so this lets us do that and
	 * then open_ctree will properly initialize everything later.
	 */
	fs_info = kzalloc(sizeof(struct btrfs_fs_info), GFP_NOFS);
	tree_root = kzalloc(sizeof(struct btrfs_root), GFP_NOFS);
	if (!fs_info || !tree_root) {
		error = -ENOMEM;
		goto error_close_devices;
	}
	fs_info->tree_root = tree_root;
	fs_info->fs_devices = fs_devices;
	tree_root->fs_info = fs_info;

940
	bdev = fs_devices->latest_bdev;
941
	s = sget(fs_type, btrfs_test_super, btrfs_set_super, tree_root);
942 943 944 945
	if (IS_ERR(s)) {
		error = PTR_ERR(s);
		goto error_close_devices;
	}
Y
Yan 已提交
946 947 948

	if (s->s_root) {
		if ((flags ^ s->s_flags) & MS_RDONLY) {
949
			deactivate_locked_super(s);
950
			return ERR_PTR(-EBUSY);
Y
Yan 已提交
951 952
		}

Y
Yan Zheng 已提交
953
		btrfs_close_devices(fs_devices);
954 955
		kfree(fs_info);
		kfree(tree_root);
Y
Yan 已提交
956 957 958
	} else {
		char b[BDEVNAME_SIZE];

A
Al Viro 已提交
959
		s->s_flags = flags | MS_NOSEC;
Y
Yan 已提交
960
		strlcpy(s->s_id, bdevname(bdev, b), sizeof(s->s_id));
961
		btrfs_sb(s)->fs_info->bdev_holder = fs_type;
962 963
		error = btrfs_fill_super(s, fs_devices, data,
					 flags & MS_SILENT ? 1 : 0);
Y
Yan 已提交
964
		if (error) {
965
			deactivate_locked_super(s);
966
			return ERR_PTR(error);
Y
Yan 已提交
967 968 969 970 971
		}

		s->s_flags |= MS_ACTIVE;
	}

972 973 974 975
	root = get_default_root(s, subvol_objectid);
	if (IS_ERR(root)) {
		deactivate_locked_super(s);
		return root;
Y
Yan 已提交
976 977
	}

A
Al Viro 已提交
978
	return root;
Y
Yan 已提交
979

Y
Yan Zheng 已提交
980
error_close_devices:
981
	btrfs_close_devices(fs_devices);
982 983
	kfree(fs_info);
	kfree(tree_root);
A
Al Viro 已提交
984
	return ERR_PTR(error);
Y
Yan 已提交
985
}
986

Y
Yan Zheng 已提交
987 988 989 990 991
static int btrfs_remount(struct super_block *sb, int *flags, char *data)
{
	struct btrfs_root *root = btrfs_sb(sb);
	int ret;

992 993 994 995
	ret = btrfs_parse_options(root, data);
	if (ret)
		return -EINVAL;

Y
Yan Zheng 已提交
996 997 998 999 1000 1001 1002 1003 1004
	if ((*flags & MS_RDONLY) == (sb->s_flags & MS_RDONLY))
		return 0;

	if (*flags & MS_RDONLY) {
		sb->s_flags |= MS_RDONLY;

		ret =  btrfs_commit_super(root);
		WARN_ON(ret);
	} else {
Y
Yan Zheng 已提交
1005 1006 1007
		if (root->fs_info->fs_devices->rw_devices == 0)
			return -EACCES;

Y
Yan Zheng 已提交
1008 1009 1010
		if (btrfs_super_log_root(&root->fs_info->super_copy) != 0)
			return -EINVAL;

1011
		ret = btrfs_cleanup_fs_roots(root->fs_info);
Y
Yan Zheng 已提交
1012 1013
		WARN_ON(ret);

1014 1015
		/* recover relocation */
		ret = btrfs_recover_relocation(root);
Y
Yan Zheng 已提交
1016 1017 1018 1019 1020 1021 1022 1023
		WARN_ON(ret);

		sb->s_flags &= ~MS_RDONLY;
	}

	return 0;
}

1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049
/* Used to sort the devices by max_avail(descending sort) */
static int btrfs_cmp_device_free_bytes(const void *dev_info1,
				       const void *dev_info2)
{
	if (((struct btrfs_device_info *)dev_info1)->max_avail >
	    ((struct btrfs_device_info *)dev_info2)->max_avail)
		return -1;
	else if (((struct btrfs_device_info *)dev_info1)->max_avail <
		 ((struct btrfs_device_info *)dev_info2)->max_avail)
		return 1;
	else
	return 0;
}

/*
 * sort the devices by max_avail, in which max free extent size of each device
 * is stored.(Descending Sort)
 */
static inline void btrfs_descending_sort_devices(
					struct btrfs_device_info *devices,
					size_t nr_devices)
{
	sort(devices, nr_devices, sizeof(struct btrfs_device_info),
	     btrfs_cmp_device_free_bytes, NULL);
}

1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170
/*
 * The helper to calc the free space on the devices that can be used to store
 * file data.
 */
static int btrfs_calc_avail_data_space(struct btrfs_root *root, u64 *free_bytes)
{
	struct btrfs_fs_info *fs_info = root->fs_info;
	struct btrfs_device_info *devices_info;
	struct btrfs_fs_devices *fs_devices = fs_info->fs_devices;
	struct btrfs_device *device;
	u64 skip_space;
	u64 type;
	u64 avail_space;
	u64 used_space;
	u64 min_stripe_size;
	int min_stripes = 1;
	int i = 0, nr_devices;
	int ret;

	nr_devices = fs_info->fs_devices->rw_devices;
	BUG_ON(!nr_devices);

	devices_info = kmalloc(sizeof(*devices_info) * nr_devices,
			       GFP_NOFS);
	if (!devices_info)
		return -ENOMEM;

	/* calc min stripe number for data space alloction */
	type = btrfs_get_alloc_profile(root, 1);
	if (type & BTRFS_BLOCK_GROUP_RAID0)
		min_stripes = 2;
	else if (type & BTRFS_BLOCK_GROUP_RAID1)
		min_stripes = 2;
	else if (type & BTRFS_BLOCK_GROUP_RAID10)
		min_stripes = 4;

	if (type & BTRFS_BLOCK_GROUP_DUP)
		min_stripe_size = 2 * BTRFS_STRIPE_LEN;
	else
		min_stripe_size = BTRFS_STRIPE_LEN;

	list_for_each_entry(device, &fs_devices->alloc_list, dev_alloc_list) {
		if (!device->in_fs_metadata)
			continue;

		avail_space = device->total_bytes - device->bytes_used;

		/* align with stripe_len */
		do_div(avail_space, BTRFS_STRIPE_LEN);
		avail_space *= BTRFS_STRIPE_LEN;

		/*
		 * In order to avoid overwritting the superblock on the drive,
		 * btrfs starts at an offset of at least 1MB when doing chunk
		 * allocation.
		 */
		skip_space = 1024 * 1024;

		/* user can set the offset in fs_info->alloc_start. */
		if (fs_info->alloc_start + BTRFS_STRIPE_LEN <=
		    device->total_bytes)
			skip_space = max(fs_info->alloc_start, skip_space);

		/*
		 * btrfs can not use the free space in [0, skip_space - 1],
		 * we must subtract it from the total. In order to implement
		 * it, we account the used space in this range first.
		 */
		ret = btrfs_account_dev_extents_size(device, 0, skip_space - 1,
						     &used_space);
		if (ret) {
			kfree(devices_info);
			return ret;
		}

		/* calc the free space in [0, skip_space - 1] */
		skip_space -= used_space;

		/*
		 * we can use the free space in [0, skip_space - 1], subtract
		 * it from the total.
		 */
		if (avail_space && avail_space >= skip_space)
			avail_space -= skip_space;
		else
			avail_space = 0;

		if (avail_space < min_stripe_size)
			continue;

		devices_info[i].dev = device;
		devices_info[i].max_avail = avail_space;

		i++;
	}

	nr_devices = i;

	btrfs_descending_sort_devices(devices_info, nr_devices);

	i = nr_devices - 1;
	avail_space = 0;
	while (nr_devices >= min_stripes) {
		if (devices_info[i].max_avail >= min_stripe_size) {
			int j;
			u64 alloc_size;

			avail_space += devices_info[i].max_avail * min_stripes;
			alloc_size = devices_info[i].max_avail;
			for (j = i + 1 - min_stripes; j <= i; j++)
				devices_info[j].max_avail -= alloc_size;
		}
		i--;
		nr_devices--;
	}

	kfree(devices_info);
	*free_bytes = avail_space;
	return 0;
}

C
Chris Mason 已提交
1171 1172 1173
static int btrfs_statfs(struct dentry *dentry, struct kstatfs *buf)
{
	struct btrfs_root *root = btrfs_sb(dentry->d_sb);
1174
	struct btrfs_super_block *disk_super = &root->fs_info->super_copy;
1175 1176 1177
	struct list_head *head = &root->fs_info->space_info;
	struct btrfs_space_info *found;
	u64 total_used = 0;
1178
	u64 total_free_data = 0;
1179
	int bits = dentry->d_sb->s_blocksize_bits;
1180
	__be32 *fsid = (__be32 *)root->fs_info->fsid;
1181
	int ret;
C
Chris Mason 已提交
1182

1183 1184
	/* holding chunk_muext to avoid allocating new chunks */
	mutex_lock(&root->fs_info->chunk_mutex);
1185
	rcu_read_lock();
J
Josef Bacik 已提交
1186
	list_for_each_entry_rcu(found, head, list) {
1187 1188 1189 1190 1191 1192
		if (found->flags & BTRFS_BLOCK_GROUP_DATA) {
			total_free_data += found->disk_total - found->disk_used;
			total_free_data -=
				btrfs_account_ro_block_groups_free_space(found);
		}

1193
		total_used += found->disk_used;
J
Josef Bacik 已提交
1194
	}
1195 1196
	rcu_read_unlock();

C
Chris Mason 已提交
1197
	buf->f_namelen = BTRFS_NAME_LEN;
1198
	buf->f_blocks = btrfs_super_total_bytes(disk_super) >> bits;
1199
	buf->f_bfree = buf->f_blocks - (total_used >> bits);
C
Chris Mason 已提交
1200 1201
	buf->f_bsize = dentry->d_sb->s_blocksize;
	buf->f_type = BTRFS_SUPER_MAGIC;
1202 1203 1204 1205 1206 1207 1208 1209 1210
	buf->f_bavail = total_free_data;
	ret = btrfs_calc_avail_data_space(root, &total_free_data);
	if (ret) {
		mutex_unlock(&root->fs_info->chunk_mutex);
		return ret;
	}
	buf->f_bavail += total_free_data;
	buf->f_bavail = buf->f_bavail >> bits;
	mutex_unlock(&root->fs_info->chunk_mutex);
C
Chris Mason 已提交
1211

1212
	/* We treat it as constant endianness (it doesn't matter _which_)
C
Chris Mason 已提交
1213
	   because we want the fsid to come out the same whether mounted
1214 1215 1216
	   on a big-endian or little-endian host */
	buf->f_fsid.val[0] = be32_to_cpu(fsid[0]) ^ be32_to_cpu(fsid[2]);
	buf->f_fsid.val[1] = be32_to_cpu(fsid[1]) ^ be32_to_cpu(fsid[3]);
1217 1218 1219 1220
	/* Mask in the root object ID too, to disambiguate subvols */
	buf->f_fsid.val[0] ^= BTRFS_I(dentry->d_inode)->root->objectid >> 32;
	buf->f_fsid.val[1] ^= BTRFS_I(dentry->d_inode)->root->objectid;

C
Chris Mason 已提交
1221 1222
	return 0;
}
C
Chris Mason 已提交
1223

1224 1225 1226
static struct file_system_type btrfs_fs_type = {
	.owner		= THIS_MODULE,
	.name		= "btrfs",
A
Al Viro 已提交
1227
	.mount		= btrfs_mount,
1228
	.kill_sb	= kill_anon_super,
1229 1230
	.fs_flags	= FS_REQUIRES_DEV,
};
1231

C
Chris Mason 已提交
1232 1233 1234
/*
 * used by btrfsctl to scan devices when no FS is mounted
 */
1235 1236 1237 1238 1239
static long btrfs_control_ioctl(struct file *file, unsigned int cmd,
				unsigned long arg)
{
	struct btrfs_ioctl_vol_args *vol;
	struct btrfs_fs_devices *fs_devices;
1240
	int ret = -ENOTTY;
1241

1242 1243 1244
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

L
Li Zefan 已提交
1245 1246 1247
	vol = memdup_user((void __user *)arg, sizeof(*vol));
	if (IS_ERR(vol))
		return PTR_ERR(vol);
1248

1249 1250
	switch (cmd) {
	case BTRFS_IOC_SCAN_DEV:
1251
		ret = btrfs_scan_one_device(vol->name, FMODE_READ,
1252 1253 1254
					    &btrfs_fs_type, &fs_devices);
		break;
	}
L
Li Zefan 已提交
1255

1256
	kfree(vol);
L
Linda Knippers 已提交
1257
	return ret;
1258 1259
}

1260
static int btrfs_freeze(struct super_block *sb)
Y
Yan 已提交
1261 1262
{
	struct btrfs_root *root = btrfs_sb(sb);
1263 1264
	mutex_lock(&root->fs_info->transaction_kthread_mutex);
	mutex_lock(&root->fs_info->cleaner_mutex);
1265
	return 0;
Y
Yan 已提交
1266 1267
}

1268
static int btrfs_unfreeze(struct super_block *sb)
Y
Yan 已提交
1269 1270
{
	struct btrfs_root *root = btrfs_sb(sb);
1271 1272
	mutex_unlock(&root->fs_info->cleaner_mutex);
	mutex_unlock(&root->fs_info->transaction_kthread_mutex);
1273
	return 0;
Y
Yan 已提交
1274
}
1275

1276
static const struct super_operations btrfs_super_ops = {
1277
	.drop_inode	= btrfs_drop_inode,
A
Al Viro 已提交
1278
	.evict_inode	= btrfs_evict_inode,
C
Chris Mason 已提交
1279
	.put_super	= btrfs_put_super,
1280
	.sync_fs	= btrfs_sync_fs,
E
Eric Paris 已提交
1281
	.show_options	= btrfs_show_options,
C
Chris Mason 已提交
1282
	.write_inode	= btrfs_write_inode,
C
Chris Mason 已提交
1283
	.dirty_inode	= btrfs_dirty_inode,
C
Chris Mason 已提交
1284 1285
	.alloc_inode	= btrfs_alloc_inode,
	.destroy_inode	= btrfs_destroy_inode,
C
Chris Mason 已提交
1286
	.statfs		= btrfs_statfs,
Y
Yan Zheng 已提交
1287
	.remount_fs	= btrfs_remount,
1288 1289
	.freeze_fs	= btrfs_freeze,
	.unfreeze_fs	= btrfs_unfreeze,
C
Chris Mason 已提交
1290
};
1291 1292 1293 1294 1295

static const struct file_operations btrfs_ctl_fops = {
	.unlocked_ioctl	 = btrfs_control_ioctl,
	.compat_ioctl = btrfs_control_ioctl,
	.owner	 = THIS_MODULE,
1296
	.llseek = noop_llseek,
1297 1298 1299
};

static struct miscdevice btrfs_misc = {
1300
	.minor		= BTRFS_MINOR,
1301 1302 1303 1304
	.name		= "btrfs-control",
	.fops		= &btrfs_ctl_fops
};

1305 1306 1307
MODULE_ALIAS_MISCDEV(BTRFS_MINOR);
MODULE_ALIAS("devname:btrfs-control");

1308 1309 1310 1311 1312
static int btrfs_interface_init(void)
{
	return misc_register(&btrfs_misc);
}

1313
static void btrfs_interface_exit(void)
1314 1315
{
	if (misc_deregister(&btrfs_misc) < 0)
C
Chris Mason 已提交
1316
		printk(KERN_INFO "misc_deregister failed for control device");
1317 1318
}

1319 1320
static int __init init_btrfs_fs(void)
{
C
Chris Mason 已提交
1321
	int err;
1322 1323 1324 1325 1326

	err = btrfs_init_sysfs();
	if (err)
		return err;

1327
	err = btrfs_init_compress();
C
Chris Mason 已提交
1328
	if (err)
1329
		goto free_sysfs;
1330

1331 1332 1333 1334
	err = btrfs_init_cachep();
	if (err)
		goto free_compress;

1335
	err = extent_io_init();
1336 1337 1338
	if (err)
		goto free_cachep;

1339 1340 1341 1342
	err = extent_map_init();
	if (err)
		goto free_extent_io;

1343
	err = btrfs_delayed_inode_init();
1344 1345
	if (err)
		goto free_extent_map;
C
Chris Mason 已提交
1346

1347 1348 1349 1350
	err = btrfs_interface_init();
	if (err)
		goto free_delayed_inode;

1351 1352 1353
	err = register_filesystem(&btrfs_fs_type);
	if (err)
		goto unregister_ioctl;
1354 1355

	printk(KERN_INFO "%s loaded\n", BTRFS_BUILD_VERSION);
1356 1357
	return 0;

1358 1359
unregister_ioctl:
	btrfs_interface_exit();
1360 1361
free_delayed_inode:
	btrfs_delayed_inode_exit();
1362 1363
free_extent_map:
	extent_map_exit();
1364 1365
free_extent_io:
	extent_io_exit();
1366 1367
free_cachep:
	btrfs_destroy_cachep();
1368 1369
free_compress:
	btrfs_exit_compress();
1370
free_sysfs:
1371 1372
	btrfs_exit_sysfs();
	return err;
1373 1374 1375 1376
}

static void __exit exit_btrfs_fs(void)
{
C
Chris Mason 已提交
1377
	btrfs_destroy_cachep();
1378
	btrfs_delayed_inode_exit();
1379
	extent_map_exit();
1380
	extent_io_exit();
1381
	btrfs_interface_exit();
1382
	unregister_filesystem(&btrfs_fs_type);
1383
	btrfs_exit_sysfs();
1384
	btrfs_cleanup_fs_uuids();
1385
	btrfs_exit_compress();
1386 1387 1388 1389 1390 1391
}

module_init(init_btrfs_fs)
module_exit(exit_btrfs_fs)

MODULE_LICENSE("GPL");