super.c 36.0 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>
44
#include <linux/ratelimit.h>
C
Chris Mason 已提交
45
#include "compat.h"
46
#include "delayed-inode.h"
47
#include "ctree.h"
C
Chris Mason 已提交
48
#include "disk-io.h"
49
#include "transaction.h"
C
Chris Mason 已提交
50
#include "btrfs_inode.h"
C
Chris Mason 已提交
51
#include "ioctl.h"
C
Chris Mason 已提交
52
#include "print-tree.h"
J
Josef Bacik 已提交
53
#include "xattr.h"
54
#include "volumes.h"
55
#include "version.h"
B
Balaji Rao 已提交
56
#include "export.h"
C
Chris Mason 已提交
57
#include "compression.h"
58

59 60 61
#define CREATE_TRACE_POINTS
#include <trace/events/btrfs.h>

62
static const struct super_operations btrfs_super_ops;
63
static struct file_system_type btrfs_fs_type;
C
Chris Mason 已提交
64

L
liubo 已提交
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148
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
149
static void btrfs_put_super(struct super_block *sb)
C
Chris Mason 已提交
150
{
C
Chris Mason 已提交
151
	struct btrfs_root *root = btrfs_sb(sb);
C
Chris Mason 已提交
152 153
	int ret;

C
Chris Mason 已提交
154 155
	ret = close_ctree(root);
	sb->s_fs_info = NULL;
156 157

	(void)ret; /* FIXME: need to fix VFS to return error? */
C
Chris Mason 已提交
158 159
}

160
enum {
161
	Opt_degraded, Opt_subvol, Opt_subvolid, Opt_device, Opt_nodatasum,
162 163
	Opt_nodatacow, Opt_max_inline, Opt_alloc_start, Opt_nobarrier, Opt_ssd,
	Opt_nossd, Opt_ssd_spread, Opt_thread_pool, Opt_noacl, Opt_compress,
164 165
	Opt_compress_type, Opt_compress_force, Opt_compress_force_type,
	Opt_notreelog, Opt_ratio, Opt_flushoncommit, Opt_discard,
166
	Opt_space_cache, Opt_clear_cache, Opt_user_subvol_rm_allowed,
167 168
	Opt_enospc_debug, Opt_subvolrootid, Opt_defrag, Opt_inode_cache,
	Opt_no_space_cache, Opt_recovery, Opt_skip_balance,
169 170
	Opt_check_integrity, Opt_check_integrity_including_extent_data,
	Opt_check_integrity_print_mask,
171
	Opt_err,
172 173 174
};

static match_table_t tokens = {
175
	{Opt_degraded, "degraded"},
176
	{Opt_subvol, "subvol=%s"},
177
	{Opt_subvolid, "subvolid=%d"},
178
	{Opt_device, "device=%s"},
179
	{Opt_nodatasum, "nodatasum"},
180
	{Opt_nodatacow, "nodatacow"},
181
	{Opt_nobarrier, "nobarrier"},
182
	{Opt_max_inline, "max_inline=%s"},
183
	{Opt_alloc_start, "alloc_start=%s"},
184
	{Opt_thread_pool, "thread_pool=%d"},
C
Chris Mason 已提交
185
	{Opt_compress, "compress"},
186
	{Opt_compress_type, "compress=%s"},
C
Chris Mason 已提交
187
	{Opt_compress_force, "compress-force"},
188
	{Opt_compress_force_type, "compress-force=%s"},
189
	{Opt_ssd, "ssd"},
190
	{Opt_ssd_spread, "ssd_spread"},
C
Chris Mason 已提交
191
	{Opt_nossd, "nossd"},
J
Josef Bacik 已提交
192
	{Opt_noacl, "noacl"},
S
Sage Weil 已提交
193
	{Opt_notreelog, "notreelog"},
194
	{Opt_flushoncommit, "flushoncommit"},
195
	{Opt_ratio, "metadata_ratio=%d"},
C
Christoph Hellwig 已提交
196
	{Opt_discard, "discard"},
197
	{Opt_space_cache, "space_cache"},
198
	{Opt_clear_cache, "clear_cache"},
199
	{Opt_user_subvol_rm_allowed, "user_subvol_rm_allowed"},
200
	{Opt_enospc_debug, "enospc_debug"},
201
	{Opt_subvolrootid, "subvolrootid=%d"},
C
Chris Mason 已提交
202
	{Opt_defrag, "autodefrag"},
C
Chris Mason 已提交
203
	{Opt_inode_cache, "inode_cache"},
204
	{Opt_no_space_cache, "nospace_cache"},
C
Chris Mason 已提交
205
	{Opt_recovery, "recovery"},
206
	{Opt_skip_balance, "skip_balance"},
207 208 209
	{Opt_check_integrity, "check_int"},
	{Opt_check_integrity_including_extent_data, "check_int_data"},
	{Opt_check_integrity_print_mask, "check_int_print_mask=%d"},
J
Josef Bacik 已提交
210
	{Opt_err, NULL},
211 212
};

213 214 215 216 217
/*
 * 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)
218
{
219
	struct btrfs_fs_info *info = root->fs_info;
220
	substring_t args[MAX_OPT_ARGS];
221 222
	char *p, *num, *orig = NULL;
	u64 cache_gen;
223
	int intarg;
S
Sage Weil 已提交
224
	int ret = 0;
225 226
	char *compress_type;
	bool compress_force = false;
227

228
	cache_gen = btrfs_super_cache_generation(root->fs_info->super_copy);
229 230 231
	if (cache_gen)
		btrfs_set_opt(info->mount_opt, SPACE_CACHE);

232
	if (!options)
233
		goto out;
234

235 236 237 238 239 240 241 242
	/*
	 * strsep changes the string, duplicate it because parse_options
	 * gets called twice
	 */
	options = kstrdup(options, GFP_NOFS);
	if (!options)
		return -ENOMEM;

243
	orig = options;
244

245
	while ((p = strsep(&options, ",")) != NULL) {
246 247 248 249 250 251
		int token;
		if (!*p)
			continue;

		token = match_token(p, tokens, args);
		switch (token) {
252
		case Opt_degraded:
253 254
			printk(KERN_INFO "btrfs: allowing degraded mounts\n");
			btrfs_set_opt(info->mount_opt, DEGRADED);
255
			break;
256
		case Opt_subvol:
257
		case Opt_subvolid:
258
		case Opt_subvolrootid:
259
		case Opt_device:
260
			/*
261
			 * These are parsed by btrfs_parse_early_options
262 263
			 * and can be happily ignored here.
			 */
264 265
			break;
		case Opt_nodatasum:
266
			printk(KERN_INFO "btrfs: setting nodatasum\n");
267
			btrfs_set_opt(info->mount_opt, NODATASUM);
268 269
			break;
		case Opt_nodatacow:
270 271 272
			printk(KERN_INFO "btrfs: setting nodatacow\n");
			btrfs_set_opt(info->mount_opt, NODATACOW);
			btrfs_set_opt(info->mount_opt, NODATASUM);
273
			break;
C
Chris Mason 已提交
274
		case Opt_compress_force:
275 276 277 278 279 280 281 282 283
		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
284 285 286
			} else if (strcmp(args[0].from, "lzo") == 0) {
				compress_type = "lzo";
				info->compress_type = BTRFS_COMPRESS_LZO;
287 288 289 290 291
			} else {
				ret = -EINVAL;
				goto out;
			}

C
Chris Mason 已提交
292
			btrfs_set_opt(info->mount_opt, COMPRESS);
293 294 295 296 297 298 299
			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
300
			break;
301
		case Opt_ssd:
302 303
			printk(KERN_INFO "btrfs: use ssd allocation scheme\n");
			btrfs_set_opt(info->mount_opt, SSD);
304
			break;
305 306 307 308 309 310
		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
311
		case Opt_nossd:
312 313
			printk(KERN_INFO "btrfs: not using ssd allocation "
			       "scheme\n");
C
Chris Mason 已提交
314
			btrfs_set_opt(info->mount_opt, NOSSD);
C
Chris Mason 已提交
315
			btrfs_clear_opt(info->mount_opt, SSD);
316
			btrfs_clear_opt(info->mount_opt, SSD_SPREAD);
C
Chris Mason 已提交
317
			break;
318
		case Opt_nobarrier:
319 320
			printk(KERN_INFO "btrfs: turning off barriers\n");
			btrfs_set_opt(info->mount_opt, NOBARRIER);
321
			break;
322 323 324 325 326 327 328 329 330
		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;
331
		case Opt_max_inline:
332 333
			num = match_strdup(&args[0]);
			if (num) {
A
Akinobu Mita 已提交
334
				info->max_inline = memparse(num, NULL);
335 336
				kfree(num);

C
Chris Mason 已提交
337 338 339 340 341
				if (info->max_inline) {
					info->max_inline = max_t(u64,
						info->max_inline,
						root->sectorsize);
				}
342
				printk(KERN_INFO "btrfs: max_inline at %llu\n",
343
					(unsigned long long)info->max_inline);
344 345
			}
			break;
346
		case Opt_alloc_start:
347 348
			num = match_strdup(&args[0]);
			if (num) {
A
Akinobu Mita 已提交
349
				info->alloc_start = memparse(num, NULL);
350 351 352
				kfree(num);
				printk(KERN_INFO
					"btrfs: allocations start at %llu\n",
353
					(unsigned long long)info->alloc_start);
354 355
			}
			break;
J
Josef Bacik 已提交
356 357 358
		case Opt_noacl:
			root->fs_info->sb->s_flags &= ~MS_POSIXACL;
			break;
S
Sage Weil 已提交
359 360 361 362
		case Opt_notreelog:
			printk(KERN_INFO "btrfs: disabling tree log\n");
			btrfs_set_opt(info->mount_opt, NOTREELOG);
			break;
363 364 365 366
		case Opt_flushoncommit:
			printk(KERN_INFO "btrfs: turning on flush-on-commit\n");
			btrfs_set_opt(info->mount_opt, FLUSHONCOMMIT);
			break;
367 368 369 370 371 372 373 374 375
		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
376 377 378
		case Opt_discard:
			btrfs_set_opt(info->mount_opt, DISCARD);
			break;
379 380
		case Opt_space_cache:
			btrfs_set_opt(info->mount_opt, SPACE_CACHE);
381
			break;
382 383 384 385
		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
386 387 388 389
		case Opt_inode_cache:
			printk(KERN_INFO "btrfs: enabling inode map caching\n");
			btrfs_set_opt(info->mount_opt, INODE_MAP_CACHE);
			break;
390 391 392
		case Opt_clear_cache:
			printk(KERN_INFO "btrfs: force clearing of disk cache\n");
			btrfs_set_opt(info->mount_opt, CLEAR_CACHE);
393
			break;
394 395 396
		case Opt_user_subvol_rm_allowed:
			btrfs_set_opt(info->mount_opt, USER_SUBVOL_RM_ALLOWED);
			break;
397 398 399
		case Opt_enospc_debug:
			btrfs_set_opt(info->mount_opt, ENOSPC_DEBUG);
			break;
C
Chris Mason 已提交
400 401 402 403
		case Opt_defrag:
			printk(KERN_INFO "btrfs: enabling auto defrag");
			btrfs_set_opt(info->mount_opt, AUTO_DEFRAG);
			break;
C
Chris Mason 已提交
404 405 406 407
		case Opt_recovery:
			printk(KERN_INFO "btrfs: enabling auto recovery");
			btrfs_set_opt(info->mount_opt, RECOVERY);
			break;
408 409 410
		case Opt_skip_balance:
			btrfs_set_opt(info->mount_opt, SKIP_BALANCE);
			break;
411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441
#ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY
		case Opt_check_integrity_including_extent_data:
			printk(KERN_INFO "btrfs: enabling check integrity"
			       " including extent data\n");
			btrfs_set_opt(info->mount_opt,
				      CHECK_INTEGRITY_INCLUDING_EXTENT_DATA);
			btrfs_set_opt(info->mount_opt, CHECK_INTEGRITY);
			break;
		case Opt_check_integrity:
			printk(KERN_INFO "btrfs: enabling check integrity\n");
			btrfs_set_opt(info->mount_opt, CHECK_INTEGRITY);
			break;
		case Opt_check_integrity_print_mask:
			intarg = 0;
			match_int(&args[0], &intarg);
			if (intarg) {
				info->check_integrity_print_mask = intarg;
				printk(KERN_INFO "btrfs:"
				       " check_integrity_print_mask 0x%x\n",
				       info->check_integrity_print_mask);
			}
			break;
#else
		case Opt_check_integrity_including_extent_data:
		case Opt_check_integrity:
		case Opt_check_integrity_print_mask:
			printk(KERN_ERR "btrfs: support for check_integrity*"
			       " not compiled in!\n");
			ret = -EINVAL;
			goto out;
#endif
S
Sage Weil 已提交
442 443 444 445 446
		case Opt_err:
			printk(KERN_INFO "btrfs: unrecognized mount option "
			       "'%s'\n", p);
			ret = -EINVAL;
			goto out;
447
		default:
448
			break;
449 450
		}
	}
S
Sage Weil 已提交
451
out:
452 453
	if (!ret && btrfs_test_opt(root, SPACE_CACHE))
		printk(KERN_INFO "btrfs: disk space caching is enabled\n");
454
	kfree(orig);
S
Sage Weil 已提交
455
	return ret;
456 457 458 459 460 461 462 463
}

/*
 * 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.
 */
464
static int btrfs_parse_early_options(const char *options, fmode_t flags,
465
		void *holder, char **subvol_name, u64 *subvol_objectid,
466
		u64 *subvol_rootid, struct btrfs_fs_devices **fs_devices)
467 468
{
	substring_t args[MAX_OPT_ARGS];
469
	char *device_name, *opts, *orig, *p;
470
	int error = 0;
471
	int intarg;
472 473

	if (!options)
474
		return 0;
475 476 477 478 479 480 481 482

	/*
	 * strsep changes the string, duplicate it because parse_options
	 * gets called twice
	 */
	opts = kstrdup(options, GFP_KERNEL);
	if (!opts)
		return -ENOMEM;
483
	orig = opts;
484 485 486 487 488 489 490 491 492

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

		token = match_token(p, tokens, args);
		switch (token) {
		case Opt_subvol:
493
			kfree(*subvol_name);
494 495
			*subvol_name = match_strdup(&args[0]);
			break;
496 497
		case Opt_subvolid:
			intarg = 0;
498 499 500 501 502 503 504 505 506
			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;
			}
507
			break;
508 509 510 511 512 513 514 515 516 517 518 519
		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;
520
		case Opt_device:
521 522 523 524 525 526
			device_name = match_strdup(&args[0]);
			if (!device_name) {
				error = -ENOMEM;
				goto out;
			}
			error = btrfs_scan_one_device(device_name,
527
					flags, holder, fs_devices);
528
			kfree(device_name);
529
			if (error)
530
				goto out;
531
			break;
532 533 534 535 536
		default:
			break;
		}
	}

537
out:
538
	kfree(orig);
539
	return error;
540 541
}

542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574
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.
	 */
575
	dir_id = btrfs_super_root_dir(root->fs_info->super_copy);
576
	di = btrfs_lookup_dir_item(NULL, root, path, dir_id, "default", 7, 0);
577 578
	if (IS_ERR(di)) {
		btrfs_free_path(path);
579
		return ERR_CAST(di);
580
	}
581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598
	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
599
		return ERR_CAST(new_root);
600 601 602 603 604 605 606 607 608 609 610

	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);
611 612
	if (IS_ERR(inode))
		return ERR_CAST(inode);
613 614 615 616 617 618 619 620 621 622 623

	/*
	 * 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);
	}

624
	return d_obtain_alias(inode);
625 626
}

C
Chris Mason 已提交
627
static int btrfs_fill_super(struct super_block *sb,
628
			    struct btrfs_fs_devices *fs_devices,
C
Chris Mason 已提交
629
			    void *data, int silent)
C
Chris Mason 已提交
630
{
C
Chris Mason 已提交
631 632
	struct inode *inode;
	struct dentry *root_dentry;
C
Chris Mason 已提交
633
	struct btrfs_root *tree_root;
634
	struct btrfs_key key;
C
Chris Mason 已提交
635
	int err;
636

C
Chris Mason 已提交
637 638 639
	sb->s_maxbytes = MAX_LFS_FILESIZE;
	sb->s_magic = BTRFS_SUPER_MAGIC;
	sb->s_op = &btrfs_super_ops;
A
Al Viro 已提交
640
	sb->s_d_op = &btrfs_dentry_operations;
B
Balaji Rao 已提交
641
	sb->s_export_op = &btrfs_export_ops;
J
Josef Bacik 已提交
642
	sb->s_xattr = btrfs_xattr_handlers;
C
Chris Mason 已提交
643
	sb->s_time_gran = 1;
C
Chris Mason 已提交
644
#ifdef CONFIG_BTRFS_FS_POSIX_ACL
J
Josef Bacik 已提交
645
	sb->s_flags |= MS_POSIXACL;
646
#endif
647

648
	tree_root = open_ctree(sb, fs_devices, (char *)data);
649

650
	if (IS_ERR(tree_root)) {
C
Chris Mason 已提交
651
		printk("btrfs: open_ctree failed\n");
652
		return PTR_ERR(tree_root);
653
	}
C
Chris Mason 已提交
654
	sb->s_fs_info = tree_root;
655

656 657 658
	key.objectid = BTRFS_FIRST_FREE_OBJECTID;
	key.type = BTRFS_INODE_ITEM_KEY;
	key.offset = 0;
659
	inode = btrfs_iget(sb, &key, tree_root->fs_info->fs_root, NULL);
660 661
	if (IS_ERR(inode)) {
		err = PTR_ERR(inode);
C
Chris Mason 已提交
662
		goto fail_close;
C
Chris Mason 已提交
663 664
	}

C
Chris Mason 已提交
665 666 667 668 669
	root_dentry = d_alloc_root(inode);
	if (!root_dentry) {
		iput(inode);
		err = -ENOMEM;
		goto fail_close;
C
Chris Mason 已提交
670
	}
671

C
Chris Mason 已提交
672
	sb->s_root = root_dentry;
C
Chris Mason 已提交
673 674

	save_mount_options(sb, data);
D
Dan Magenheimer 已提交
675
	cleancache_init_fs(sb);
C
Chris Mason 已提交
676
	return 0;
C
Chris Mason 已提交
677 678 679 680

fail_close:
	close_ctree(tree_root);
	return err;
C
Chris Mason 已提交
681 682
}

S
Sage Weil 已提交
683
int btrfs_sync_fs(struct super_block *sb, int wait)
C
Chris Mason 已提交
684 685
{
	struct btrfs_trans_handle *trans;
686
	struct btrfs_root *root = btrfs_sb(sb);
C
Chris Mason 已提交
687
	int ret;
C
Chris Mason 已提交
688

689 690
	trace_btrfs_sync_fs(wait);

C
Chris Mason 已提交
691 692 693 694
	if (!wait) {
		filemap_flush(root->fs_info->btree_inode->i_mapping);
		return 0;
	}
695

Y
Yan, Zheng 已提交
696 697
	btrfs_start_delalloc_inodes(root, 0);
	btrfs_wait_ordered_extents(root, 0, 0);
698

699
	trans = btrfs_start_transaction(root, 0);
700 701
	if (IS_ERR(trans))
		return PTR_ERR(trans);
C
Chris Mason 已提交
702
	ret = btrfs_commit_transaction(trans, root);
703
	return ret;
C
Chris Mason 已提交
704 705
}

E
Eric Paris 已提交
706 707 708 709
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
710
	char *compress_type;
E
Eric Paris 已提交
711 712 713 714 715 716 717 718 719 720

	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)
721 722
		seq_printf(seq, ",max_inline=%llu",
			   (unsigned long long)info->max_inline);
E
Eric Paris 已提交
723
	if (info->alloc_start != 0)
724 725
		seq_printf(seq, ",alloc_start=%llu",
			   (unsigned long long)info->alloc_start);
E
Eric Paris 已提交
726 727 728
	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
729 730 731 732 733 734 735 736 737 738
	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
739 740
	if (btrfs_test_opt(root, NOSSD))
		seq_puts(seq, ",nossd");
741 742 743
	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
744
		seq_puts(seq, ",ssd");
S
Sage Weil 已提交
745
	if (btrfs_test_opt(root, NOTREELOG))
746
		seq_puts(seq, ",notreelog");
747
	if (btrfs_test_opt(root, FLUSHONCOMMIT))
748
		seq_puts(seq, ",flushoncommit");
749 750
	if (btrfs_test_opt(root, DISCARD))
		seq_puts(seq, ",discard");
E
Eric Paris 已提交
751 752
	if (!(root->fs_info->sb->s_flags & MS_POSIXACL))
		seq_puts(seq, ",noacl");
T
Tsutomu Itoh 已提交
753 754
	if (btrfs_test_opt(root, SPACE_CACHE))
		seq_puts(seq, ",space_cache");
755
	else
756
		seq_puts(seq, ",nospace_cache");
T
Tsutomu Itoh 已提交
757 758 759 760
	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");
761 762 763 764 765 766
	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");
767 768
	if (btrfs_test_opt(root, SKIP_BALANCE))
		seq_puts(seq, ",skip_balance");
E
Eric Paris 已提交
769 770 771
	return 0;
}

772
static int btrfs_test_super(struct super_block *s, void *data)
Y
Yan 已提交
773
{
774
	struct btrfs_root *test_root = data;
775
	struct btrfs_root *root = btrfs_sb(s);
Y
Yan 已提交
776

777 778 779 780 781 782
	/*
	 * 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;
783
	return root->fs_info->fs_devices == test_root->fs_info->fs_devices;
Y
Yan 已提交
784 785
}

786 787 788 789 790
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
791 792
}

793 794 795 796 797 798 799 800 801 802
/*
 * 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;
}

803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884
/*
 * 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 dentry *root;
	struct vfsmount *mnt;
	char *newargs;

	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);

A
Al Viro 已提交
885
	root = mount_subtree(mnt, subvol_name);
886

A
Al Viro 已提交
887 888 889 890 891
	if (!IS_ERR(root) && !is_subvolume_inode(root->d_inode)) {
		struct super_block *s = root->d_sb;
		dput(root);
		root = ERR_PTR(-EINVAL);
		deactivate_locked_super(s);
892 893 894 895
		printk(KERN_ERR "btrfs: '%s' is not a valid subvolume\n",
				subvol_name);
	}

896 897
	return root;
}
898

899 900 901 902 903 904
/*
 * 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
905
static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags,
906
		const char *device_name, void *data)
Y
Yan 已提交
907 908 909 910
{
	struct block_device *bdev = NULL;
	struct super_block *s;
	struct dentry *root;
911
	struct btrfs_fs_devices *fs_devices = NULL;
912
	struct btrfs_fs_info *fs_info = NULL;
913
	fmode_t mode = FMODE_READ;
914 915
	char *subvol_name = NULL;
	u64 subvol_objectid = 0;
916
	u64 subvol_rootid = 0;
Y
Yan 已提交
917 918
	int error = 0;

919 920 921 922
	if (!(flags & MS_RDONLY))
		mode |= FMODE_WRITE;

	error = btrfs_parse_early_options(data, mode, fs_type,
923
					  &subvol_name, &subvol_objectid,
924
					  &subvol_rootid, &fs_devices);
925 926
	if (error) {
		kfree(subvol_name);
A
Al Viro 已提交
927
		return ERR_PTR(error);
928
	}
929

930 931 932 933 934 935
	if (subvol_name) {
		root = mount_subvol(subvol_name, flags, device_name, data);
		kfree(subvol_name);
		return root;
	}

936
	error = btrfs_scan_one_device(device_name, mode, fs_type, &fs_devices);
937
	if (error)
938
		return ERR_PTR(error);
Y
Yan 已提交
939

940 941 942 943 944 945 946
	/*
	 * 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);
947 948 949 950 951
	if (!fs_info)
		return ERR_PTR(-ENOMEM);

	fs_info->tree_root = kzalloc(sizeof(struct btrfs_root), GFP_NOFS);
	if (!fs_info->tree_root) {
952
		error = -ENOMEM;
953
		goto error_fs_info;
954
	}
955
	fs_info->tree_root->fs_info = fs_info;
956 957
	fs_info->fs_devices = fs_devices;

958 959 960 961
	fs_info->super_copy = kzalloc(BTRFS_SUPER_INFO_SIZE, GFP_NOFS);
	fs_info->super_for_commit = kzalloc(BTRFS_SUPER_INFO_SIZE, GFP_NOFS);
	if (!fs_info->super_copy || !fs_info->super_for_commit) {
		error = -ENOMEM;
962 963 964 965 966 967 968 969 970
		goto error_fs_info;
	}

	error = btrfs_open_devices(fs_devices, mode, fs_type);
	if (error)
		goto error_fs_info;

	if (!(flags & MS_RDONLY) && fs_devices->rw_devices == 0) {
		error = -EACCES;
971 972 973
		goto error_close_devices;
	}

974
	bdev = fs_devices->latest_bdev;
975 976
	s = sget(fs_type, btrfs_test_super, btrfs_set_super,
		 fs_info->tree_root);
977 978 979 980
	if (IS_ERR(s)) {
		error = PTR_ERR(s);
		goto error_close_devices;
	}
Y
Yan 已提交
981 982 983

	if (s->s_root) {
		if ((flags ^ s->s_flags) & MS_RDONLY) {
984
			deactivate_locked_super(s);
985 986
			error = -EBUSY;
			goto error_close_devices;
Y
Yan 已提交
987 988
		}

Y
Yan Zheng 已提交
989
		btrfs_close_devices(fs_devices);
990
		free_fs_info(fs_info);
Y
Yan 已提交
991 992 993
	} else {
		char b[BDEVNAME_SIZE];

A
Al Viro 已提交
994
		s->s_flags = flags | MS_NOSEC;
Y
Yan 已提交
995
		strlcpy(s->s_id, bdevname(bdev, b), sizeof(s->s_id));
996
		btrfs_sb(s)->fs_info->bdev_holder = fs_type;
997 998
		error = btrfs_fill_super(s, fs_devices, data,
					 flags & MS_SILENT ? 1 : 0);
Y
Yan 已提交
999
		if (error) {
1000
			deactivate_locked_super(s);
1001
			return ERR_PTR(error);
Y
Yan 已提交
1002 1003 1004 1005 1006
		}

		s->s_flags |= MS_ACTIVE;
	}

1007 1008 1009 1010
	root = get_default_root(s, subvol_objectid);
	if (IS_ERR(root)) {
		deactivate_locked_super(s);
		return root;
Y
Yan 已提交
1011 1012
	}

A
Al Viro 已提交
1013
	return root;
Y
Yan 已提交
1014

Y
Yan Zheng 已提交
1015
error_close_devices:
1016
	btrfs_close_devices(fs_devices);
1017
error_fs_info:
1018
	free_fs_info(fs_info);
A
Al Viro 已提交
1019
	return ERR_PTR(error);
Y
Yan 已提交
1020
}
1021

Y
Yan Zheng 已提交
1022 1023 1024 1025 1026
static int btrfs_remount(struct super_block *sb, int *flags, char *data)
{
	struct btrfs_root *root = btrfs_sb(sb);
	int ret;

1027 1028 1029 1030
	ret = btrfs_parse_options(root, data);
	if (ret)
		return -EINVAL;

Y
Yan Zheng 已提交
1031 1032 1033 1034 1035 1036 1037 1038 1039
	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
1040 1041 1042
		if (root->fs_info->fs_devices->rw_devices == 0)
			return -EACCES;

1043
		if (btrfs_super_log_root(root->fs_info->super_copy) != 0)
Y
Yan Zheng 已提交
1044 1045
			return -EINVAL;

1046
		ret = btrfs_cleanup_fs_roots(root->fs_info);
Y
Yan Zheng 已提交
1047 1048
		WARN_ON(ret);

1049 1050
		/* recover relocation */
		ret = btrfs_recover_relocation(root);
Y
Yan Zheng 已提交
1051 1052 1053 1054 1055 1056 1057 1058
		WARN_ON(ret);

		sb->s_flags &= ~MS_RDONLY;
	}

	return 0;
}

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
/* 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);
}

1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099
/*
 * 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;
1100
	int min_stripes = 1, num_stripes = 1;
1101 1102 1103
	int i = 0, nr_devices;
	int ret;

1104
	nr_devices = fs_info->fs_devices->open_devices;
1105 1106 1107 1108 1109 1110 1111 1112 1113
	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);
1114
	if (type & BTRFS_BLOCK_GROUP_RAID0) {
1115
		min_stripes = 2;
1116 1117
		num_stripes = nr_devices;
	} else if (type & BTRFS_BLOCK_GROUP_RAID1) {
1118
		min_stripes = 2;
1119 1120
		num_stripes = 2;
	} else if (type & BTRFS_BLOCK_GROUP_RAID10) {
1121
		min_stripes = 4;
1122 1123
		num_stripes = 4;
	}
1124 1125 1126 1127 1128 1129

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

1130 1131
	list_for_each_entry(device, &fs_devices->devices, dev_list) {
		if (!device->in_fs_metadata || !device->bdev)
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 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191
			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) {
1192 1193 1194
		if (num_stripes > nr_devices)
			num_stripes = nr_devices;

1195 1196 1197 1198
		if (devices_info[i].max_avail >= min_stripe_size) {
			int j;
			u64 alloc_size;

1199
			avail_space += devices_info[i].max_avail * num_stripes;
1200
			alloc_size = devices_info[i].max_avail;
1201
			for (j = i + 1 - num_stripes; j <= i; j++)
1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212
				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
1213 1214 1215
static int btrfs_statfs(struct dentry *dentry, struct kstatfs *buf)
{
	struct btrfs_root *root = btrfs_sb(dentry->d_sb);
1216
	struct btrfs_super_block *disk_super = root->fs_info->super_copy;
1217 1218 1219
	struct list_head *head = &root->fs_info->space_info;
	struct btrfs_space_info *found;
	u64 total_used = 0;
1220
	u64 total_free_data = 0;
1221
	int bits = dentry->d_sb->s_blocksize_bits;
1222
	__be32 *fsid = (__be32 *)root->fs_info->fsid;
1223
	int ret;
C
Chris Mason 已提交
1224

1225 1226
	/* holding chunk_muext to avoid allocating new chunks */
	mutex_lock(&root->fs_info->chunk_mutex);
1227
	rcu_read_lock();
J
Josef Bacik 已提交
1228
	list_for_each_entry_rcu(found, head, list) {
1229 1230 1231 1232 1233 1234
		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);
		}

1235
		total_used += found->disk_used;
J
Josef Bacik 已提交
1236
	}
1237 1238
	rcu_read_unlock();

C
Chris Mason 已提交
1239
	buf->f_namelen = BTRFS_NAME_LEN;
1240
	buf->f_blocks = btrfs_super_total_bytes(disk_super) >> bits;
1241
	buf->f_bfree = buf->f_blocks - (total_used >> bits);
C
Chris Mason 已提交
1242 1243
	buf->f_bsize = dentry->d_sb->s_blocksize;
	buf->f_type = BTRFS_SUPER_MAGIC;
1244 1245 1246 1247 1248 1249 1250 1251 1252
	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
1253

1254
	/* We treat it as constant endianness (it doesn't matter _which_)
C
Chris Mason 已提交
1255
	   because we want the fsid to come out the same whether mounted
1256 1257 1258
	   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]);
1259 1260 1261 1262
	/* 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
1263 1264
	return 0;
}
C
Chris Mason 已提交
1265

1266 1267 1268
static struct file_system_type btrfs_fs_type = {
	.owner		= THIS_MODULE,
	.name		= "btrfs",
A
Al Viro 已提交
1269
	.mount		= btrfs_mount,
1270
	.kill_sb	= kill_anon_super,
1271 1272
	.fs_flags	= FS_REQUIRES_DEV,
};
1273

C
Chris Mason 已提交
1274 1275 1276
/*
 * used by btrfsctl to scan devices when no FS is mounted
 */
1277 1278 1279 1280 1281
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;
1282
	int ret = -ENOTTY;
1283

1284 1285 1286
	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

L
Li Zefan 已提交
1287 1288 1289
	vol = memdup_user((void __user *)arg, sizeof(*vol));
	if (IS_ERR(vol))
		return PTR_ERR(vol);
1290

1291 1292
	switch (cmd) {
	case BTRFS_IOC_SCAN_DEV:
1293
		ret = btrfs_scan_one_device(vol->name, FMODE_READ,
1294 1295 1296
					    &btrfs_fs_type, &fs_devices);
		break;
	}
L
Li Zefan 已提交
1297

1298
	kfree(vol);
L
Linda Knippers 已提交
1299
	return ret;
1300 1301
}

1302
static int btrfs_freeze(struct super_block *sb)
Y
Yan 已提交
1303 1304
{
	struct btrfs_root *root = btrfs_sb(sb);
1305 1306
	mutex_lock(&root->fs_info->transaction_kthread_mutex);
	mutex_lock(&root->fs_info->cleaner_mutex);
1307
	return 0;
Y
Yan 已提交
1308 1309
}

1310
static int btrfs_unfreeze(struct super_block *sb)
Y
Yan 已提交
1311 1312
{
	struct btrfs_root *root = btrfs_sb(sb);
1313 1314
	mutex_unlock(&root->fs_info->cleaner_mutex);
	mutex_unlock(&root->fs_info->transaction_kthread_mutex);
1315
	return 0;
Y
Yan 已提交
1316
}
1317

1318 1319 1320 1321 1322 1323 1324 1325 1326 1327
static void btrfs_fs_dirty_inode(struct inode *inode, int flags)
{
	int ret;

	ret = btrfs_dirty_inode(inode);
	if (ret)
		printk_ratelimited(KERN_ERR "btrfs: fail to dirty inode %Lu "
				   "error %d\n", btrfs_ino(inode), ret);
}

1328
static const struct super_operations btrfs_super_ops = {
1329
	.drop_inode	= btrfs_drop_inode,
A
Al Viro 已提交
1330
	.evict_inode	= btrfs_evict_inode,
C
Chris Mason 已提交
1331
	.put_super	= btrfs_put_super,
1332
	.sync_fs	= btrfs_sync_fs,
E
Eric Paris 已提交
1333
	.show_options	= btrfs_show_options,
C
Chris Mason 已提交
1334
	.write_inode	= btrfs_write_inode,
1335
	.dirty_inode	= btrfs_fs_dirty_inode,
C
Chris Mason 已提交
1336 1337
	.alloc_inode	= btrfs_alloc_inode,
	.destroy_inode	= btrfs_destroy_inode,
C
Chris Mason 已提交
1338
	.statfs		= btrfs_statfs,
Y
Yan Zheng 已提交
1339
	.remount_fs	= btrfs_remount,
1340 1341
	.freeze_fs	= btrfs_freeze,
	.unfreeze_fs	= btrfs_unfreeze,
C
Chris Mason 已提交
1342
};
1343 1344 1345 1346 1347

static const struct file_operations btrfs_ctl_fops = {
	.unlocked_ioctl	 = btrfs_control_ioctl,
	.compat_ioctl = btrfs_control_ioctl,
	.owner	 = THIS_MODULE,
1348
	.llseek = noop_llseek,
1349 1350 1351
};

static struct miscdevice btrfs_misc = {
1352
	.minor		= BTRFS_MINOR,
1353 1354 1355 1356
	.name		= "btrfs-control",
	.fops		= &btrfs_ctl_fops
};

1357 1358 1359
MODULE_ALIAS_MISCDEV(BTRFS_MINOR);
MODULE_ALIAS("devname:btrfs-control");

1360 1361 1362 1363 1364
static int btrfs_interface_init(void)
{
	return misc_register(&btrfs_misc);
}

1365
static void btrfs_interface_exit(void)
1366 1367
{
	if (misc_deregister(&btrfs_misc) < 0)
C
Chris Mason 已提交
1368
		printk(KERN_INFO "misc_deregister failed for control device");
1369 1370
}

1371 1372
static int __init init_btrfs_fs(void)
{
C
Chris Mason 已提交
1373
	int err;
1374 1375 1376 1377 1378

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

1379
	err = btrfs_init_compress();
C
Chris Mason 已提交
1380
	if (err)
1381
		goto free_sysfs;
1382

1383 1384 1385 1386
	err = btrfs_init_cachep();
	if (err)
		goto free_compress;

1387
	err = extent_io_init();
1388 1389 1390
	if (err)
		goto free_cachep;

1391 1392 1393 1394
	err = extent_map_init();
	if (err)
		goto free_extent_io;

1395
	err = btrfs_delayed_inode_init();
1396 1397
	if (err)
		goto free_extent_map;
C
Chris Mason 已提交
1398

1399 1400 1401 1402
	err = btrfs_interface_init();
	if (err)
		goto free_delayed_inode;

1403 1404 1405
	err = register_filesystem(&btrfs_fs_type);
	if (err)
		goto unregister_ioctl;
1406 1407

	printk(KERN_INFO "%s loaded\n", BTRFS_BUILD_VERSION);
1408 1409
	return 0;

1410 1411
unregister_ioctl:
	btrfs_interface_exit();
1412 1413
free_delayed_inode:
	btrfs_delayed_inode_exit();
1414 1415
free_extent_map:
	extent_map_exit();
1416 1417
free_extent_io:
	extent_io_exit();
1418 1419
free_cachep:
	btrfs_destroy_cachep();
1420 1421
free_compress:
	btrfs_exit_compress();
1422
free_sysfs:
1423 1424
	btrfs_exit_sysfs();
	return err;
1425 1426 1427 1428
}

static void __exit exit_btrfs_fs(void)
{
C
Chris Mason 已提交
1429
	btrfs_destroy_cachep();
1430
	btrfs_delayed_inode_exit();
1431
	extent_map_exit();
1432
	extent_io_exit();
1433
	btrfs_interface_exit();
1434
	unregister_filesystem(&btrfs_fs_type);
1435
	btrfs_exit_sysfs();
1436
	btrfs_cleanup_fs_uuids();
1437
	btrfs_exit_compress();
1438 1439 1440 1441 1442 1443
}

module_init(init_btrfs_fs)
module_exit(exit_btrfs_fs)

MODULE_LICENSE("GPL");