super.c 66.2 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
/* -*- mode: c; c-basic-offset: 8; -*-
 * vim: noexpandtab sw=8 ts=8 sts=0:
 *
 * super.c
 *
 * load/unload driver, mount/dismount volumes
 *
 * Copyright (C) 2002, 2004 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 as published by the Free Software Foundation; either
 * version 2 of the License, or (at your option) any later version.
 *
 * 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/module.h>
#include <linux/fs.h>
#include <linux/types.h>
#include <linux/slab.h>
#include <linux/highmem.h>
#include <linux/init.h>
#include <linux/random.h>
#include <linux/statfs.h>
#include <linux/moduleparam.h>
#include <linux/blkdev.h>
#include <linux/socket.h>
#include <linux/inet.h>
#include <linux/parser.h>
#include <linux/crc32.h>
#include <linux/debugfs.h>
S
Sunil Mushran 已提交
41
#include <linux/mount.h>
42
#include <linux/seq_file.h>
43
#include <linux/quotaops.h>
44
#include <linux/smp_lock.h>
45 46 47 48 49 50 51 52 53 54

#define MLOG_MASK_PREFIX ML_SUPER
#include <cluster/masklog.h>

#include "ocfs2.h"

/* this should be the only file to include a version 1 header */
#include "ocfs1_fs_compat.h"

#include "alloc.h"
55
#include "blockcheck.h"
56 57 58 59 60 61 62 63 64 65 66 67 68
#include "dlmglue.h"
#include "export.h"
#include "extent_map.h"
#include "heartbeat.h"
#include "inode.h"
#include "journal.h"
#include "localalloc.h"
#include "namei.h"
#include "slot_map.h"
#include "super.h"
#include "sysfile.h"
#include "uptodate.h"
#include "ver.h"
T
Tiger Yang 已提交
69
#include "xattr.h"
70
#include "quota.h"
71
#include "refcounttree.h"
72
#include "suballoc.h"
73 74 75

#include "buffer_head_io.h"

76
static struct kmem_cache *ocfs2_inode_cachep = NULL;
77 78
struct kmem_cache *ocfs2_dquot_cachep;
struct kmem_cache *ocfs2_qf_chunk_cachep;
79 80 81 82 83 84 85 86 87 88 89 90

/* OCFS2 needs to schedule several differnt types of work which
 * require cluster locking, disk I/O, recovery waits, etc. Since these
 * types of work tend to be heavy we avoid using the kernel events
 * workqueue and schedule on our own. */
struct workqueue_struct *ocfs2_wq = NULL;

static struct dentry *ocfs2_debugfs_root = NULL;

MODULE_AUTHOR("Oracle");
MODULE_LICENSE("GPL");

91 92
struct mount_options
{
93
	unsigned long	commit_interval;
94 95 96
	unsigned long	mount_opt;
	unsigned int	atime_quantum;
	signed short	slot;
97
	int		localalloc_opt;
M
Mark Fasheh 已提交
98
	unsigned int	resv_level;
99
	int		dir_resv_level;
100
	char		cluster_stack[OCFS2_STACK_LABEL_LEN + 1];
101 102
};

103
static int ocfs2_parse_options(struct super_block *sb, char *options,
104
			       struct mount_options *mopt,
105
			       int is_remount);
J
Jan Kara 已提交
106 107
static int ocfs2_check_set_options(struct super_block *sb,
				   struct mount_options *options);
S
Sunil Mushran 已提交
108
static int ocfs2_show_options(struct seq_file *s, struct vfsmount *mnt);
109 110 111 112 113 114 115 116
static void ocfs2_put_super(struct super_block *sb);
static int ocfs2_mount_volume(struct super_block *sb);
static int ocfs2_remount(struct super_block *sb, int *flags, char *data);
static void ocfs2_dismount_volume(struct super_block *sb, int mnt_err);
static int ocfs2_initialize_mem_caches(void);
static void ocfs2_free_mem_caches(void);
static void ocfs2_delete_osb(struct ocfs2_super *osb);

117
static int ocfs2_statfs(struct dentry *dentry, struct kstatfs *buf);
118 119 120 121 122

static int ocfs2_sync_fs(struct super_block *sb, int wait);

static int ocfs2_init_global_system_inodes(struct ocfs2_super *osb);
static int ocfs2_init_local_system_inodes(struct ocfs2_super *osb);
123
static void ocfs2_release_system_inodes(struct ocfs2_super *osb);
124 125 126
static int ocfs2_check_volume(struct ocfs2_super *osb);
static int ocfs2_verify_volume(struct ocfs2_dinode *di,
			       struct buffer_head *bh,
127 128
			       u32 sectsize,
			       struct ocfs2_blockcheck_stats *stats);
129 130
static int ocfs2_initialize_super(struct super_block *sb,
				  struct buffer_head *bh,
131 132
				  int sector_size,
				  struct ocfs2_blockcheck_stats *stats);
133 134 135 136 137 138
static int ocfs2_get_sector(struct super_block *sb,
			    struct buffer_head **bh,
			    int block,
			    int sect_size);
static struct inode *ocfs2_alloc_inode(struct super_block *sb);
static void ocfs2_destroy_inode(struct inode *inode);
139 140 141
static int ocfs2_susp_quotas(struct ocfs2_super *osb, int unsuspend);
static int ocfs2_enable_quotas(struct ocfs2_super *osb);
static void ocfs2_disable_quotas(struct ocfs2_super *osb);
142

143
static const struct super_operations ocfs2_sops = {
144 145 146 147 148 149 150 151 152
	.statfs		= ocfs2_statfs,
	.alloc_inode	= ocfs2_alloc_inode,
	.destroy_inode	= ocfs2_destroy_inode,
	.drop_inode	= ocfs2_drop_inode,
	.clear_inode	= ocfs2_clear_inode,
	.delete_inode	= ocfs2_delete_inode,
	.sync_fs	= ocfs2_sync_fs,
	.put_super	= ocfs2_put_super,
	.remount_fs	= ocfs2_remount,
S
Sunil Mushran 已提交
153
	.show_options   = ocfs2_show_options,
154 155
	.quota_read	= ocfs2_quota_read,
	.quota_write	= ocfs2_quota_write,
156 157 158 159 160 161 162 163 164 165 166 167
};

enum {
	Opt_barrier,
	Opt_err_panic,
	Opt_err_ro,
	Opt_intr,
	Opt_nointr,
	Opt_hb_none,
	Opt_hb_local,
	Opt_data_ordered,
	Opt_data_writeback,
T
Tiger Yang 已提交
168
	Opt_atime_quantum,
169
	Opt_slot,
170
	Opt_commit,
171
	Opt_localalloc,
172
	Opt_localflocks,
173
	Opt_stack,
T
Tiger Yang 已提交
174 175
	Opt_user_xattr,
	Opt_nouser_xattr,
176
	Opt_inode64,
177 178
	Opt_acl,
	Opt_noacl,
179 180
	Opt_usrquota,
	Opt_grpquota,
M
Mark Fasheh 已提交
181
	Opt_resv_level,
182
	Opt_dir_resv_level,
183 184 185
	Opt_err,
};

186
static const match_table_t tokens = {
187 188 189 190 191 192 193 194 195
	{Opt_barrier, "barrier=%u"},
	{Opt_err_panic, "errors=panic"},
	{Opt_err_ro, "errors=remount-ro"},
	{Opt_intr, "intr"},
	{Opt_nointr, "nointr"},
	{Opt_hb_none, OCFS2_HB_NONE},
	{Opt_hb_local, OCFS2_HB_LOCAL},
	{Opt_data_ordered, "data=ordered"},
	{Opt_data_writeback, "data=writeback"},
T
Tiger Yang 已提交
196
	{Opt_atime_quantum, "atime_quantum=%u"},
197
	{Opt_slot, "preferred_slot=%u"},
198
	{Opt_commit, "commit=%u"},
199
	{Opt_localalloc, "localalloc=%d"},
200
	{Opt_localflocks, "localflocks"},
201
	{Opt_stack, "cluster_stack=%s"},
T
Tiger Yang 已提交
202 203
	{Opt_user_xattr, "user_xattr"},
	{Opt_nouser_xattr, "nouser_xattr"},
204
	{Opt_inode64, "inode64"},
205 206
	{Opt_acl, "acl"},
	{Opt_noacl, "noacl"},
207 208
	{Opt_usrquota, "usrquota"},
	{Opt_grpquota, "grpquota"},
M
Mark Fasheh 已提交
209
	{Opt_resv_level, "resv_level=%u"},
210
	{Opt_dir_resv_level, "dir_resv_level=%u"},
211 212 213
	{Opt_err, NULL}
};

214 215 216 217 218
#ifdef CONFIG_DEBUG_FS
static int ocfs2_osb_dump(struct ocfs2_super *osb, char *buf, int len)
{
	struct ocfs2_cluster_connection *cconn = osb->cconn;
	struct ocfs2_recovery_map *rm = osb->recovery_map;
219 220
	struct ocfs2_orphan_scan *os = &osb->osb_orphan_scan;
	int i, out = 0;
221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244

	out += snprintf(buf + out, len - out,
			"%10s => Id: %-s  Uuid: %-s  Gen: 0x%X  Label: %-s\n",
			"Device", osb->dev_str, osb->uuid_str,
			osb->fs_generation, osb->vol_label);

	out += snprintf(buf + out, len - out,
			"%10s => State: %d  Flags: 0x%lX\n", "Volume",
			atomic_read(&osb->vol_state), osb->osb_flags);

	out += snprintf(buf + out, len - out,
			"%10s => Block: %lu  Cluster: %d\n", "Sizes",
			osb->sb->s_blocksize, osb->s_clustersize);

	out += snprintf(buf + out, len - out,
			"%10s => Compat: 0x%X  Incompat: 0x%X  "
			"ROcompat: 0x%X\n",
			"Features", osb->s_feature_compat,
			osb->s_feature_incompat, osb->s_feature_ro_compat);

	out += snprintf(buf + out, len - out,
			"%10s => Opts: 0x%lX  AtimeQuanta: %u\n", "Mount",
			osb->s_mount_opt, osb->s_atime_quantum);

S
Sunil Mushran 已提交
245 246 247 248 249 250 251 252 253 254
	if (cconn) {
		out += snprintf(buf + out, len - out,
				"%10s => Stack: %s  Name: %*s  "
				"Version: %d.%d\n", "Cluster",
				(*osb->osb_cluster_stack == '\0' ?
				 "o2cb" : osb->osb_cluster_stack),
				cconn->cc_namelen, cconn->cc_name,
				cconn->cc_version.pv_major,
				cconn->cc_version.pv_minor);
	}
255 256 257 258 259

	spin_lock(&osb->dc_task_lock);
	out += snprintf(buf + out, len - out,
			"%10s => Pid: %d  Count: %lu  WakeSeq: %lu  "
			"WorkSeq: %lu\n", "DownCnvt",
S
Sunil Mushran 已提交
260 261 262
			(osb->dc_task ?  task_pid_nr(osb->dc_task) : -1),
			osb->blocked_lock_count, osb->dc_wake_sequence,
			osb->dc_work_sequence);
263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281
	spin_unlock(&osb->dc_task_lock);

	spin_lock(&osb->osb_lock);
	out += snprintf(buf + out, len - out, "%10s => Pid: %d  Nodes:",
			"Recovery",
			(osb->recovery_thread_task ?
			 task_pid_nr(osb->recovery_thread_task) : -1));
	if (rm->rm_used == 0)
		out += snprintf(buf + out, len - out, " None\n");
	else {
		for (i = 0; i < rm->rm_used; i++)
			out += snprintf(buf + out, len - out, " %d",
					rm->rm_entries[i]);
		out += snprintf(buf + out, len - out, "\n");
	}
	spin_unlock(&osb->osb_lock);

	out += snprintf(buf + out, len - out,
			"%10s => Pid: %d  Interval: %lu  Needs: %d\n", "Commit",
S
Sunil Mushran 已提交
282 283
			(osb->commit_task ? task_pid_nr(osb->commit_task) : -1),
			osb->osb_commit_interval,
284 285 286
			atomic_read(&osb->needs_checkpoint));

	out += snprintf(buf + out, len - out,
S
Sunil Mushran 已提交
287
			"%10s => State: %d  TxnId: %lu  NumTxns: %d\n",
288
			"Journal", osb->journal->j_state,
S
Sunil Mushran 已提交
289 290
			osb->journal->j_trans_id,
			atomic_read(&osb->journal->j_num_trans));
291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310

	out += snprintf(buf + out, len - out,
			"%10s => GlobalAllocs: %d  LocalAllocs: %d  "
			"SubAllocs: %d  LAWinMoves: %d  SAExtends: %d\n",
			"Stats",
			atomic_read(&osb->alloc_stats.bitmap_data),
			atomic_read(&osb->alloc_stats.local_data),
			atomic_read(&osb->alloc_stats.bg_allocs),
			atomic_read(&osb->alloc_stats.moves),
			atomic_read(&osb->alloc_stats.bg_extends));

	out += snprintf(buf + out, len - out,
			"%10s => State: %u  Descriptor: %llu  Size: %u bits  "
			"Default: %u bits\n",
			"LocalAlloc", osb->local_alloc_state,
			(unsigned long long)osb->la_last_gd,
			osb->local_alloc_bits, osb->local_alloc_default_bits);

	spin_lock(&osb->osb_lock);
	out += snprintf(buf + out, len - out,
311 312
			"%10s => InodeSlot: %d  StolenInodes: %d, "
			"MetaSlot: %d  StolenMeta: %d\n", "Steal",
313
			osb->s_inode_steal_slot,
314 315 316
			atomic_read(&osb->s_num_inodes_stolen),
			osb->s_meta_steal_slot,
			atomic_read(&osb->s_num_meta_stolen));
317 318
	spin_unlock(&osb->osb_lock);

319 320 321 322 323 324 325 326 327 328
	out += snprintf(buf + out, len - out, "OrphanScan => ");
	out += snprintf(buf + out, len - out, "Local: %u  Global: %u ",
			os->os_count, os->os_seqno);
	out += snprintf(buf + out, len - out, " Last Scan: ");
	if (atomic_read(&os->os_state) == ORPHAN_SCAN_INACTIVE)
		out += snprintf(buf + out, len - out, "Disabled\n");
	else
		out += snprintf(buf + out, len - out, "%lu seconds ago\n",
				(get_seconds() - os->os_scantime.tv_sec));

329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387
	out += snprintf(buf + out, len - out, "%10s => %3s  %10s\n",
			"Slots", "Num", "RecoGen");
	for (i = 0; i < osb->max_slots; ++i) {
		out += snprintf(buf + out, len - out,
				"%10s  %c %3d  %10d\n",
				" ",
				(i == osb->slot_num ? '*' : ' '),
				i, osb->slot_recovery_generations[i]);
	}

	return out;
}

static int ocfs2_osb_debug_open(struct inode *inode, struct file *file)
{
	struct ocfs2_super *osb = inode->i_private;
	char *buf = NULL;

	buf = kmalloc(PAGE_SIZE, GFP_KERNEL);
	if (!buf)
		goto bail;

	i_size_write(inode, ocfs2_osb_dump(osb, buf, PAGE_SIZE));

	file->private_data = buf;

	return 0;
bail:
	return -ENOMEM;
}

static int ocfs2_debug_release(struct inode *inode, struct file *file)
{
	kfree(file->private_data);
	return 0;
}

static ssize_t ocfs2_debug_read(struct file *file, char __user *buf,
				size_t nbytes, loff_t *ppos)
{
	return simple_read_from_buffer(buf, nbytes, ppos, file->private_data,
				       i_size_read(file->f_mapping->host));
}
#else
static int ocfs2_osb_debug_open(struct inode *inode, struct file *file)
{
	return 0;
}
static int ocfs2_debug_release(struct inode *inode, struct file *file)
{
	return 0;
}
static ssize_t ocfs2_debug_read(struct file *file, char __user *buf,
				size_t nbytes, loff_t *ppos)
{
	return 0;
}
#endif	/* CONFIG_DEBUG_FS */

388
static const struct file_operations ocfs2_osb_debug_fops = {
389 390 391 392 393 394
	.open =		ocfs2_osb_debug_open,
	.release =	ocfs2_debug_release,
	.read =		ocfs2_debug_read,
	.llseek =	generic_file_llseek,
};

395 396
static int ocfs2_sync_fs(struct super_block *sb, int wait)
{
397
	int status;
398 399 400 401 402 403 404 405 406 407 408 409 410 411
	tid_t target;
	struct ocfs2_super *osb = OCFS2_SB(sb);

	if (ocfs2_is_hard_readonly(osb))
		return -EROFS;

	if (wait) {
		status = ocfs2_flush_truncate_log(osb);
		if (status < 0)
			mlog_errno(status);
	} else {
		ocfs2_schedule_truncate_log_flush(osb, 0);
	}

J
Joel Becker 已提交
412 413
	if (jbd2_journal_start_commit(OCFS2_SB(sb)->journal->j_journal,
				      &target)) {
414
		if (wait)
J
Joel Becker 已提交
415 416
			jbd2_log_wait_commit(OCFS2_SB(sb)->journal->j_journal,
					     target);
417 418 419 420
	}
	return 0;
}

421 422 423 424 425 426 427 428 429 430 431 432 433
static int ocfs2_need_system_inode(struct ocfs2_super *osb, int ino)
{
	if (!OCFS2_HAS_RO_COMPAT_FEATURE(osb->sb, OCFS2_FEATURE_RO_COMPAT_USRQUOTA)
	    && (ino == USER_QUOTA_SYSTEM_INODE
		|| ino == LOCAL_USER_QUOTA_SYSTEM_INODE))
		return 0;
	if (!OCFS2_HAS_RO_COMPAT_FEATURE(osb->sb, OCFS2_FEATURE_RO_COMPAT_GRPQUOTA)
	    && (ino == GROUP_QUOTA_SYSTEM_INODE
		|| ino == LOCAL_GROUP_QUOTA_SYSTEM_INODE))
		return 0;
	return 1;
}

434 435 436 437 438 439 440 441
static int ocfs2_init_global_system_inodes(struct ocfs2_super *osb)
{
	struct inode *new = NULL;
	int status = 0;
	int i;

	mlog_entry_void();

J
Jan Kara 已提交
442
	new = ocfs2_iget(osb, osb->root_blkno, OCFS2_FI_FLAG_SYSFILE, 0);
443 444 445 446 447 448 449
	if (IS_ERR(new)) {
		status = PTR_ERR(new);
		mlog_errno(status);
		goto bail;
	}
	osb->root_inode = new;

J
Jan Kara 已提交
450
	new = ocfs2_iget(osb, osb->system_dir_blkno, OCFS2_FI_FLAG_SYSFILE, 0);
451 452 453 454 455 456 457 458 459
	if (IS_ERR(new)) {
		status = PTR_ERR(new);
		mlog_errno(status);
		goto bail;
	}
	osb->sys_root_inode = new;

	for (i = OCFS2_FIRST_ONLINE_SYSTEM_INODE;
	     i <= OCFS2_LAST_GLOBAL_SYSTEM_INODE; i++) {
460 461
		if (!ocfs2_need_system_inode(osb, i))
			continue;
462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491
		new = ocfs2_get_system_file_inode(osb, i, osb->slot_num);
		if (!new) {
			ocfs2_release_system_inodes(osb);
			status = -EINVAL;
			mlog_errno(status);
			/* FIXME: Should ERROR_RO_FS */
			mlog(ML_ERROR, "Unable to load system inode %d, "
			     "possibly corrupt fs?", i);
			goto bail;
		}
		// the array now has one ref, so drop this one
		iput(new);
	}

bail:
	mlog_exit(status);
	return status;
}

static int ocfs2_init_local_system_inodes(struct ocfs2_super *osb)
{
	struct inode *new = NULL;
	int status = 0;
	int i;

	mlog_entry_void();

	for (i = OCFS2_LAST_GLOBAL_SYSTEM_INODE + 1;
	     i < NUM_SYSTEM_INODES;
	     i++) {
492 493
		if (!ocfs2_need_system_inode(osb, i))
			continue;
494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510
		new = ocfs2_get_system_file_inode(osb, i, osb->slot_num);
		if (!new) {
			ocfs2_release_system_inodes(osb);
			status = -EINVAL;
			mlog(ML_ERROR, "status=%d, sysfile=%d, slot=%d\n",
			     status, i, osb->slot_num);
			goto bail;
		}
		/* the array now has one ref, so drop this one */
		iput(new);
	}

bail:
	mlog_exit(status);
	return status;
}

511
static void ocfs2_release_system_inodes(struct ocfs2_super *osb)
512
{
513
	int i;
514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537
	struct inode *inode;

	mlog_entry_void();

	for (i = 0; i < NUM_SYSTEM_INODES; i++) {
		inode = osb->system_inodes[i];
		if (inode) {
			iput(inode);
			osb->system_inodes[i] = NULL;
		}
	}

	inode = osb->sys_root_inode;
	if (inode) {
		iput(inode);
		osb->sys_root_inode = NULL;
	}

	inode = osb->root_inode;
	if (inode) {
		iput(inode);
		osb->root_inode = NULL;
	}

538
	mlog_exit(0);
539 540 541 542 543 544 545
}

/* We're allocating fs objects, use GFP_NOFS */
static struct inode *ocfs2_alloc_inode(struct super_block *sb)
{
	struct ocfs2_inode_info *oi;

C
Christoph Lameter 已提交
546
	oi = kmem_cache_alloc(ocfs2_inode_cachep, GFP_NOFS);
547 548 549
	if (!oi)
		return NULL;

J
Joel Becker 已提交
550
	jbd2_journal_init_jbd_inode(&oi->ip_jinode, &oi->vfs_inode);
551 552 553 554 555 556 557 558
	return &oi->vfs_inode;
}

static void ocfs2_destroy_inode(struct inode *inode)
{
	kmem_cache_free(ocfs2_inode_cachep, OCFS2_I(inode));
}

M
Mark Fasheh 已提交
559 560
static unsigned long long ocfs2_max_file_offset(unsigned int bbits,
						unsigned int cbits)
561
{
M
Mark Fasheh 已提交
562 563 564 565 566 567 568 569
	unsigned int bytes = 1 << cbits;
	unsigned int trim = bytes;
	unsigned int bitshift = 32;

	/*
	 * i_size and all block offsets in ocfs2 are always 64 bits
	 * wide. i_clusters is 32 bits, in cluster-sized units. So on
	 * 64 bit platforms, cluster size will be the limiting factor.
570 571 572
	 */

#if BITS_PER_LONG == 32
573
# if defined(CONFIG_LBDAF)
A
Alexey Dobriyan 已提交
574
	BUILD_BUG_ON(sizeof(sector_t) != 8);
M
Mark Fasheh 已提交
575 576 577 578 579 580 581 582 583 584 585 586
	/*
	 * We might be limited by page cache size.
	 */
	if (bytes > PAGE_CACHE_SIZE) {
		bytes = PAGE_CACHE_SIZE;
		trim = 1;
		/*
		 * Shift by 31 here so that we don't get larger than
		 * MAX_LFS_FILESIZE
		 */
		bitshift = 31;
	}
587
# else
M
Mark Fasheh 已提交
588 589 590 591 592 593 594
	/*
	 * We are limited by the size of sector_t. Use block size, as
	 * that's what we expose to the VFS.
	 */
	bytes = 1 << bbits;
	trim = 1;
	bitshift = 31;
595 596 597
# endif
#endif

M
Mark Fasheh 已提交
598 599 600 601 602 603
	/*
	 * Trim by a whole cluster when we can actually approach the
	 * on-disk limits. Otherwise we can overflow i_clusters when
	 * an extent start is at the max offset.
	 */
	return (((unsigned long long)bytes) << bitshift) - trim;
604 605 606 607 608 609
}

static int ocfs2_remount(struct super_block *sb, int *flags, char *data)
{
	int incompat_features;
	int ret = 0;
610
	struct mount_options parsed_options;
611 612
	struct ocfs2_super *osb = OCFS2_SB(sb);

613 614
	lock_kernel();

J
Jan Kara 已提交
615 616
	if (!ocfs2_parse_options(sb, data, &parsed_options, 1) ||
	    !ocfs2_check_set_options(sb, &parsed_options)) {
617 618 619 620 621
		ret = -EINVAL;
		goto out;
	}

	if ((osb->s_mount_opt & OCFS2_MOUNT_HB_LOCAL) !=
622
	    (parsed_options.mount_opt & OCFS2_MOUNT_HB_LOCAL)) {
623 624 625 626 627 628
		ret = -EINVAL;
		mlog(ML_ERROR, "Cannot change heartbeat mode on remount\n");
		goto out;
	}

	if ((osb->s_mount_opt & OCFS2_MOUNT_DATA_WRITEBACK) !=
629
	    (parsed_options.mount_opt & OCFS2_MOUNT_DATA_WRITEBACK)) {
630 631 632 633 634
		ret = -EINVAL;
		mlog(ML_ERROR, "Cannot change data mode on remount\n");
		goto out;
	}

635 636 637 638 639 640 641 642 643
	/* Probably don't want this on remount; it might
	 * mess with other nodes */
	if (!(osb->s_mount_opt & OCFS2_MOUNT_INODE64) &&
	    (parsed_options.mount_opt & OCFS2_MOUNT_INODE64)) {
		ret = -EINVAL;
		mlog(ML_ERROR, "Cannot enable inode64 on remount\n");
		goto out;
	}

644 645
	/* We're going to/from readonly mode. */
	if ((*flags & MS_RDONLY) != (sb->s_flags & MS_RDONLY)) {
646 647 648 649 650 651
		/* Disable quota accounting before remounting RO */
		if (*flags & MS_RDONLY) {
			ret = ocfs2_susp_quotas(osb, 0);
			if (ret < 0)
				goto out;
		}
652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686
		/* Lock here so the check of HARD_RO and the potential
		 * setting of SOFT_RO is atomic. */
		spin_lock(&osb->osb_lock);
		if (osb->osb_flags & OCFS2_OSB_HARD_RO) {
			mlog(ML_ERROR, "Remount on readonly device is forbidden.\n");
			ret = -EROFS;
			goto unlock_osb;
		}

		if (*flags & MS_RDONLY) {
			mlog(0, "Going to ro mode.\n");
			sb->s_flags |= MS_RDONLY;
			osb->osb_flags |= OCFS2_OSB_SOFT_RO;
		} else {
			mlog(0, "Making ro filesystem writeable.\n");

			if (osb->osb_flags & OCFS2_OSB_ERROR_FS) {
				mlog(ML_ERROR, "Cannot remount RDWR "
				     "filesystem due to previous errors.\n");
				ret = -EROFS;
				goto unlock_osb;
			}
			incompat_features = OCFS2_HAS_RO_COMPAT_FEATURE(sb, ~OCFS2_FEATURE_RO_COMPAT_SUPP);
			if (incompat_features) {
				mlog(ML_ERROR, "Cannot remount RDWR because "
				     "of unsupported optional features "
				     "(%x).\n", incompat_features);
				ret = -EINVAL;
				goto unlock_osb;
			}
			sb->s_flags &= ~MS_RDONLY;
			osb->osb_flags &= ~OCFS2_OSB_SOFT_RO;
		}
unlock_osb:
		spin_unlock(&osb->osb_lock);
687 688 689 690 691 692 693 694 695 696 697 698 699 700 701
		/* Enable quota accounting after remounting RW */
		if (!ret && !(*flags & MS_RDONLY)) {
			if (sb_any_quota_suspended(sb))
				ret = ocfs2_susp_quotas(osb, 1);
			else
				ret = ocfs2_enable_quotas(osb);
			if (ret < 0) {
				/* Return back changes... */
				spin_lock(&osb->osb_lock);
				sb->s_flags |= MS_RDONLY;
				osb->osb_flags |= OCFS2_OSB_SOFT_RO;
				spin_unlock(&osb->osb_lock);
				goto out;
			}
		}
702 703 704 705 706
	}

	if (!ret) {
		/* Only save off the new mount options in case of a successful
		 * remount. */
707 708 709
		osb->s_mount_opt = parsed_options.mount_opt;
		osb->s_atime_quantum = parsed_options.atime_quantum;
		osb->preferred_slot = parsed_options.slot;
710 711
		if (parsed_options.commit_interval)
			osb->osb_commit_interval = parsed_options.commit_interval;
712 713 714

		if (!ocfs2_is_hard_readonly(osb))
			ocfs2_set_journal_params(osb);
J
Jan Kara 已提交
715 716 717 718

		sb->s_flags = (sb->s_flags & ~MS_POSIXACL) |
			((osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL) ?
							MS_POSIXACL : 0);
719 720
	}
out:
721
	unlock_kernel();
722 723 724 725 726
	return ret;
}

static int ocfs2_sb_probe(struct super_block *sb,
			  struct buffer_head **bh,
727 728
			  int *sector_size,
			  struct ocfs2_blockcheck_stats *stats)
729
{
730
	int status, tmpstat;
731 732 733 734 735 736 737
	struct ocfs1_vol_disk_hdr *hdr;
	struct ocfs2_dinode *di;
	int blksize;

	*bh = NULL;

	/* may be > 512 */
738
	*sector_size = bdev_logical_block_size(sb->s_bdev);
739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790
	if (*sector_size > OCFS2_MAX_BLOCKSIZE) {
		mlog(ML_ERROR, "Hardware sector size too large: %d (max=%d)\n",
		     *sector_size, OCFS2_MAX_BLOCKSIZE);
		status = -EINVAL;
		goto bail;
	}

	/* Can this really happen? */
	if (*sector_size < OCFS2_MIN_BLOCKSIZE)
		*sector_size = OCFS2_MIN_BLOCKSIZE;

	/* check block zero for old format */
	status = ocfs2_get_sector(sb, bh, 0, *sector_size);
	if (status < 0) {
		mlog_errno(status);
		goto bail;
	}
	hdr = (struct ocfs1_vol_disk_hdr *) (*bh)->b_data;
	if (hdr->major_version == OCFS1_MAJOR_VERSION) {
		mlog(ML_ERROR, "incompatible version: %u.%u\n",
		     hdr->major_version, hdr->minor_version);
		status = -EINVAL;
	}
	if (memcmp(hdr->signature, OCFS1_VOLUME_SIGNATURE,
		   strlen(OCFS1_VOLUME_SIGNATURE)) == 0) {
		mlog(ML_ERROR, "incompatible volume signature: %8s\n",
		     hdr->signature);
		status = -EINVAL;
	}
	brelse(*bh);
	*bh = NULL;
	if (status < 0) {
		mlog(ML_ERROR, "This is an ocfs v1 filesystem which must be "
		     "upgraded before mounting with ocfs v2\n");
		goto bail;
	}

	/*
	 * Now check at magic offset for 512, 1024, 2048, 4096
	 * blocksizes.  4096 is the maximum blocksize because it is
	 * the minimum clustersize.
	 */
	status = -EINVAL;
	for (blksize = *sector_size;
	     blksize <= OCFS2_MAX_BLOCKSIZE;
	     blksize <<= 1) {
		tmpstat = ocfs2_get_sector(sb, bh,
					   OCFS2_SUPER_BLOCK_BLKNO,
					   blksize);
		if (tmpstat < 0) {
			status = tmpstat;
			mlog_errno(status);
791
			break;
792 793
		}
		di = (struct ocfs2_dinode *) (*bh)->b_data;
794
		memset(stats, 0, sizeof(struct ocfs2_blockcheck_stats));
795
		spin_lock_init(&stats->b_lock);
796 797 798 799 800 801 802
		tmpstat = ocfs2_verify_volume(di, *bh, blksize, stats);
		if (tmpstat < 0) {
			brelse(*bh);
			*bh = NULL;
		}
		if (tmpstat != -EAGAIN) {
			status = tmpstat;
803
			break;
804
		}
805 806 807 808 809 810
	}

bail:
	return status;
}

S
Sunil Mushran 已提交
811 812 813 814 815 816 817 818 819 820
static int ocfs2_verify_heartbeat(struct ocfs2_super *osb)
{
	if (ocfs2_mount_local(osb)) {
		if (osb->s_mount_opt & OCFS2_MOUNT_HB_LOCAL) {
			mlog(ML_ERROR, "Cannot heartbeat on a locally "
			     "mounted device.\n");
			return -EINVAL;
		}
	}

821 822 823 824 825 826 827 828
	if (ocfs2_userspace_stack(osb)) {
		if (osb->s_mount_opt & OCFS2_MOUNT_HB_LOCAL) {
			mlog(ML_ERROR, "Userspace stack expected, but "
			     "o2cb heartbeat arguments passed to mount\n");
			return -EINVAL;
		}
	}

S
Sunil Mushran 已提交
829
	if (!(osb->s_mount_opt & OCFS2_MOUNT_HB_LOCAL)) {
830 831
		if (!ocfs2_mount_local(osb) && !ocfs2_is_hard_readonly(osb) &&
		    !ocfs2_userspace_stack(osb)) {
S
Sunil Mushran 已提交
832 833 834 835 836 837 838 839 840
			mlog(ML_ERROR, "Heartbeat has to be started to mount "
			     "a read-write clustered device.\n");
			return -EINVAL;
		}
	}

	return 0;
}

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
/*
 * If we're using a userspace stack, mount should have passed
 * a name that matches the disk.  If not, mount should not
 * have passed a stack.
 */
static int ocfs2_verify_userspace_stack(struct ocfs2_super *osb,
					struct mount_options *mopt)
{
	if (!ocfs2_userspace_stack(osb) && mopt->cluster_stack[0]) {
		mlog(ML_ERROR,
		     "cluster stack passed to mount, but this filesystem "
		     "does not support it\n");
		return -EINVAL;
	}

	if (ocfs2_userspace_stack(osb) &&
	    strncmp(osb->osb_cluster_stack, mopt->cluster_stack,
		    OCFS2_STACK_LABEL_LEN)) {
		mlog(ML_ERROR,
		     "cluster stack passed to mount (\"%s\") does not "
		     "match the filesystem (\"%s\")\n",
		     mopt->cluster_stack,
		     osb->osb_cluster_stack);
		return -EINVAL;
	}

	return 0;
}

870 871 872 873 874 875 876 877 878 879 880 881
static int ocfs2_susp_quotas(struct ocfs2_super *osb, int unsuspend)
{
	int type;
	struct super_block *sb = osb->sb;
	unsigned int feature[MAXQUOTAS] = { OCFS2_FEATURE_RO_COMPAT_USRQUOTA,
					     OCFS2_FEATURE_RO_COMPAT_GRPQUOTA};
	int status = 0;

	for (type = 0; type < MAXQUOTAS; type++) {
		if (!OCFS2_HAS_RO_COMPAT_FEATURE(sb, feature[type]))
			continue;
		if (unsuspend)
882
			status = dquot_resume(sb, type);
883 884 885 886 887 888
		else {
			struct ocfs2_mem_dqinfo *oinfo;

			/* Cancel periodic syncing before suspending */
			oinfo = sb_dqinfo(sb, type)->dqi_priv;
			cancel_delayed_work_sync(&oinfo->dqi_sync_work);
889
			status = dquot_suspend(sb, type);
890
		}
891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942
		if (status < 0)
			break;
	}
	if (status < 0)
		mlog(ML_ERROR, "Failed to suspend/unsuspend quotas on "
		     "remount (error = %d).\n", status);
	return status;
}

static int ocfs2_enable_quotas(struct ocfs2_super *osb)
{
	struct inode *inode[MAXQUOTAS] = { NULL, NULL };
	struct super_block *sb = osb->sb;
	unsigned int feature[MAXQUOTAS] = { OCFS2_FEATURE_RO_COMPAT_USRQUOTA,
					     OCFS2_FEATURE_RO_COMPAT_GRPQUOTA};
	unsigned int ino[MAXQUOTAS] = { LOCAL_USER_QUOTA_SYSTEM_INODE,
					LOCAL_GROUP_QUOTA_SYSTEM_INODE };
	int status;
	int type;

	sb_dqopt(sb)->flags |= DQUOT_QUOTA_SYS_FILE | DQUOT_NEGATIVE_USAGE;
	for (type = 0; type < MAXQUOTAS; type++) {
		if (!OCFS2_HAS_RO_COMPAT_FEATURE(sb, feature[type]))
			continue;
		inode[type] = ocfs2_get_system_file_inode(osb, ino[type],
							osb->slot_num);
		if (!inode[type]) {
			status = -ENOENT;
			goto out_quota_off;
		}
		status = vfs_quota_enable(inode[type], type, QFMT_OCFS2,
						DQUOT_USAGE_ENABLED);
		if (status < 0)
			goto out_quota_off;
	}

	for (type = 0; type < MAXQUOTAS; type++)
		iput(inode[type]);
	return 0;
out_quota_off:
	ocfs2_disable_quotas(osb);
	for (type = 0; type < MAXQUOTAS; type++)
		iput(inode[type]);
	mlog_errno(status);
	return status;
}

static void ocfs2_disable_quotas(struct ocfs2_super *osb)
{
	int type;
	struct inode *inode;
	struct super_block *sb = osb->sb;
943
	struct ocfs2_mem_dqinfo *oinfo;
944 945 946 947 948 949

	/* We mostly ignore errors in this function because there's not much
	 * we can do when we see them */
	for (type = 0; type < MAXQUOTAS; type++) {
		if (!sb_has_quota_loaded(sb, type))
			continue;
950 951 952
		/* Cancel periodic syncing before we grab dqonoff_mutex */
		oinfo = sb_dqinfo(sb, type)->dqi_priv;
		cancel_delayed_work_sync(&oinfo->dqi_sync_work);
953 954 955 956
		inode = igrab(sb->s_dquot.files[type]);
		/* Turn off quotas. This will remove all dquot structures from
		 * memory and so they will be automatically synced to global
		 * quota files */
957 958
		dquot_disable(sb, type, DQUOT_USAGE_ENABLED |
					DQUOT_LIMITS_ENABLED);
959 960 961 962 963 964 965 966
		if (!inode)
			continue;
		iput(inode);
	}
}

/* Handle quota on quotactl */
static int ocfs2_quota_on(struct super_block *sb, int type, int format_id,
967
			  char *path)
968 969 970 971 972 973 974 975 976 977 978 979
{
	unsigned int feature[MAXQUOTAS] = { OCFS2_FEATURE_RO_COMPAT_USRQUOTA,
					     OCFS2_FEATURE_RO_COMPAT_GRPQUOTA};

	if (!OCFS2_HAS_RO_COMPAT_FEATURE(sb, feature[type]))
		return -EINVAL;

	return vfs_quota_enable(sb_dqopt(sb)->files[type], type,
				    format_id, DQUOT_LIMITS_ENABLED);
}

/* Handle quota off quotactl */
980
static int ocfs2_quota_off(struct super_block *sb, int type)
981
{
982
	return dquot_disable(sb, type, DQUOT_LIMITS_ENABLED);
983 984
}

985
static const struct quotactl_ops ocfs2_quotactl_ops = {
986 987 988 989 990 991 992 993 994
	.quota_on	= ocfs2_quota_on,
	.quota_off	= ocfs2_quota_off,
	.quota_sync	= vfs_quota_sync,
	.get_info	= vfs_get_dqinfo,
	.set_info	= vfs_set_dqinfo,
	.get_dqblk	= vfs_get_dqblk,
	.set_dqblk	= vfs_set_dqblk,
};

995 996 997 998
static int ocfs2_fill_super(struct super_block *sb, void *data, int silent)
{
	struct dentry *root;
	int status, sector_size;
999
	struct mount_options parsed_options;
1000 1001 1002
	struct inode *inode = NULL;
	struct ocfs2_super *osb = NULL;
	struct buffer_head *bh = NULL;
S
Sunil Mushran 已提交
1003
	char nodestr[8];
1004
	struct ocfs2_blockcheck_stats stats;
1005 1006 1007

	mlog_entry("%p, %p, %i", sb, data, silent);

1008
	if (!ocfs2_parse_options(sb, data, &parsed_options, 0)) {
1009 1010 1011 1012 1013
		status = -EINVAL;
		goto read_super_error;
	}

	/* probe for superblock */
1014
	status = ocfs2_sb_probe(sb, &bh, &sector_size, &stats);
1015 1016 1017 1018 1019
	if (status < 0) {
		mlog(ML_ERROR, "superblock probe failed!\n");
		goto read_super_error;
	}

1020
	status = ocfs2_initialize_super(sb, bh, sector_size, &stats);
1021 1022 1023 1024 1025 1026 1027
	osb = OCFS2_SB(sb);
	if (status < 0) {
		mlog_errno(status);
		goto read_super_error;
	}
	brelse(bh);
	bh = NULL;
1028

J
Jan Kara 已提交
1029 1030 1031 1032
	if (!ocfs2_check_set_options(sb, &parsed_options)) {
		status = -EINVAL;
		goto read_super_error;
	}
1033 1034 1035
	osb->s_mount_opt = parsed_options.mount_opt;
	osb->s_atime_quantum = parsed_options.atime_quantum;
	osb->preferred_slot = parsed_options.slot;
1036
	osb->osb_commit_interval = parsed_options.commit_interval;
1037 1038

	ocfs2_la_set_sizes(osb, parsed_options.localalloc_opt);
M
Mark Fasheh 已提交
1039
	osb->osb_resv_level = parsed_options.resv_level;
1040 1041 1042 1043 1044
	osb->osb_dir_resv_level = parsed_options.resv_level;
	if (parsed_options.dir_resv_level == -1)
		osb->osb_dir_resv_level = parsed_options.resv_level;
	else
		osb->osb_dir_resv_level = parsed_options.dir_resv_level;
1045

1046 1047 1048 1049
	status = ocfs2_verify_userspace_stack(osb, &parsed_options);
	if (status)
		goto read_super_error;

1050 1051
	sb->s_magic = OCFS2_SUPER_MAGIC;

1052 1053 1054
	sb->s_flags = (sb->s_flags & ~MS_POSIXACL) |
		((osb->s_mount_opt & OCFS2_MOUNT_POSIX_ACL) ? MS_POSIXACL : 0);

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
	/* Hard readonly mode only if: bdev_read_only, MS_RDONLY,
	 * heartbeat=none */
	if (bdev_read_only(sb->s_bdev)) {
		if (!(sb->s_flags & MS_RDONLY)) {
			status = -EACCES;
			mlog(ML_ERROR, "Readonly device detected but readonly "
			     "mount was not specified.\n");
			goto read_super_error;
		}

		/* You should not be able to start a local heartbeat
		 * on a readonly device. */
		if (osb->s_mount_opt & OCFS2_MOUNT_HB_LOCAL) {
			status = -EROFS;
			mlog(ML_ERROR, "Local heartbeat specified on readonly "
			     "device.\n");
			goto read_super_error;
		}

		status = ocfs2_check_journals_nolocks(osb);
		if (status < 0) {
			if (status == -EROFS)
				mlog(ML_ERROR, "Recovery required on readonly "
				     "file system, but write access is "
				     "unavailable.\n");
			else
1081
				mlog_errno(status);
1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096
			goto read_super_error;
		}

		ocfs2_set_ro_flag(osb, 1);

		printk(KERN_NOTICE "Readonly device detected. No cluster "
		       "services will be utilized for this mount. Recovery "
		       "will be skipped.\n");
	}

	if (!ocfs2_is_hard_readonly(osb)) {
		if (sb->s_flags & MS_RDONLY)
			ocfs2_set_ro_flag(osb, 0);
	}

S
Sunil Mushran 已提交
1097 1098 1099 1100 1101 1102
	status = ocfs2_verify_heartbeat(osb);
	if (status < 0) {
		mlog_errno(status);
		goto read_super_error;
	}

1103 1104 1105 1106 1107 1108 1109 1110
	osb->osb_debug_root = debugfs_create_dir(osb->uuid_str,
						 ocfs2_debugfs_root);
	if (!osb->osb_debug_root) {
		status = -EINVAL;
		mlog(ML_ERROR, "Unable to create per-mount debugfs root.\n");
		goto read_super_error;
	}

1111 1112 1113 1114 1115 1116 1117 1118 1119 1120
	osb->osb_ctxt = debugfs_create_file("fs_state", S_IFREG|S_IRUSR,
					    osb->osb_debug_root,
					    osb,
					    &ocfs2_osb_debug_fops);
	if (!osb->osb_ctxt) {
		status = -EINVAL;
		mlog_errno(status);
		goto read_super_error;
	}

1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132
	if (ocfs2_meta_ecc(osb)) {
		status = ocfs2_blockcheck_stats_debugfs_install(
						&osb->osb_ecc_stats,
						osb->osb_debug_root);
		if (status) {
			mlog(ML_ERROR,
			     "Unable to create blockcheck statistics "
			     "files\n");
			goto read_super_error;
		}
	}

1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156
	status = ocfs2_mount_volume(sb);
	if (osb->root_inode)
		inode = igrab(osb->root_inode);

	if (status < 0)
		goto read_super_error;

	if (!inode) {
		status = -EIO;
		mlog_errno(status);
		goto read_super_error;
	}

	root = d_alloc_root(inode);
	if (!root) {
		status = -ENOMEM;
		mlog_errno(status);
		goto read_super_error;
	}

	sb->s_root = root;

	ocfs2_complete_mount_recovery(osb);

S
Sunil Mushran 已提交
1157 1158 1159
	if (ocfs2_mount_local(osb))
		snprintf(nodestr, sizeof(nodestr), "local");
	else
1160
		snprintf(nodestr, sizeof(nodestr), "%u", osb->node_num);
S
Sunil Mushran 已提交
1161 1162

	printk(KERN_INFO "ocfs2: Mounting device (%s) on (node %s, slot %d) "
S
Sunil Mushran 已提交
1163
	       "with %s data mode.\n",
S
Sunil Mushran 已提交
1164
	       osb->dev_str, nodestr, osb->slot_num,
1165 1166 1167 1168 1169 1170
	       osb->s_mount_opt & OCFS2_MOUNT_DATA_WRITEBACK ? "writeback" :
	       "ordered");

	atomic_set(&osb->vol_state, VOLUME_MOUNTED);
	wake_up(&osb->osb_mount_event);

1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192
	/* Now we can initialize quotas because we can afford to wait
	 * for cluster locks recovery now. That also means that truncation
	 * log recovery can happen but that waits for proper quota setup */
	if (!(sb->s_flags & MS_RDONLY)) {
		status = ocfs2_enable_quotas(osb);
		if (status < 0) {
			/* We have to err-out specially here because
			 * s_root is already set */
			mlog_errno(status);
			atomic_set(&osb->vol_state, VOLUME_DISABLED);
			wake_up(&osb->osb_mount_event);
			mlog_exit(status);
			return status;
		}
	}

	ocfs2_complete_quota_recovery(osb);

	/* Now we wake up again for processes waiting for quotas */
	atomic_set(&osb->vol_state, VOLUME_MOUNTED_QUOTAS);
	wake_up(&osb->osb_mount_event);

1193
	/* Start this when the mount is almost sure of being successful */
1194
	ocfs2_orphan_scan_start(osb);
1195

1196 1197 1198 1199
	mlog_exit(status);
	return status;

read_super_error:
1200
	brelse(bh);
1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214

	if (inode)
		iput(inode);

	if (osb) {
		atomic_set(&osb->vol_state, VOLUME_DISABLED);
		wake_up(&osb->osb_mount_event);
		ocfs2_dismount_volume(sb, 1);
	}

	mlog_exit(status);
	return status;
}

1215 1216 1217 1218 1219
static int ocfs2_get_sb(struct file_system_type *fs_type,
			int flags,
			const char *dev_name,
			void *data,
			struct vfsmount *mnt)
1220
{
1221 1222
	return get_sb_bdev(fs_type, flags, dev_name, data, ocfs2_fill_super,
			   mnt);
1223 1224
}

J
Jan Kara 已提交
1225 1226 1227 1228
static void ocfs2_kill_sb(struct super_block *sb)
{
	struct ocfs2_super *osb = OCFS2_SB(sb);

1229 1230 1231 1232
	/* Failed mount? */
	if (!osb || atomic_read(&osb->vol_state) == VOLUME_DISABLED)
		goto out;

J
Jan Kara 已提交
1233 1234 1235 1236 1237 1238
	/* Prevent further queueing of inode drop events */
	spin_lock(&dentry_list_lock);
	ocfs2_set_osb_flag(osb, OCFS2_OSB_DROP_DENTRY_LOCK_IMMED);
	spin_unlock(&dentry_list_lock);
	/* Wait for work to finish and/or remove it */
	cancel_work_sync(&osb->dentry_lock_work);
1239
out:
J
Jan Kara 已提交
1240 1241 1242
	kill_block_super(sb);
}

1243 1244 1245 1246 1247
static struct file_system_type ocfs2_fs_type = {
	.owner          = THIS_MODULE,
	.name           = "ocfs2",
	.get_sb         = ocfs2_get_sb, /* is this called when we mount
					* the fs? */
J
Jan Kara 已提交
1248 1249
	.kill_sb        = ocfs2_kill_sb,

1250
	.fs_flags       = FS_REQUIRES_DEV|FS_RENAME_DOES_D_MOVE,
1251 1252 1253
	.next           = NULL
};

J
Jan Kara 已提交
1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287
static int ocfs2_check_set_options(struct super_block *sb,
				   struct mount_options *options)
{
	if (options->mount_opt & OCFS2_MOUNT_USRQUOTA &&
	    !OCFS2_HAS_RO_COMPAT_FEATURE(sb,
					 OCFS2_FEATURE_RO_COMPAT_USRQUOTA)) {
		mlog(ML_ERROR, "User quotas were requested, but this "
		     "filesystem does not have the feature enabled.\n");
		return 0;
	}
	if (options->mount_opt & OCFS2_MOUNT_GRPQUOTA &&
	    !OCFS2_HAS_RO_COMPAT_FEATURE(sb,
					 OCFS2_FEATURE_RO_COMPAT_GRPQUOTA)) {
		mlog(ML_ERROR, "Group quotas were requested, but this "
		     "filesystem does not have the feature enabled.\n");
		return 0;
	}
	if (options->mount_opt & OCFS2_MOUNT_POSIX_ACL &&
	    !OCFS2_HAS_INCOMPAT_FEATURE(sb, OCFS2_FEATURE_INCOMPAT_XATTR)) {
		mlog(ML_ERROR, "ACL support requested but extended attributes "
		     "feature is not enabled\n");
		return 0;
	}
	/* No ACL setting specified? Use XATTR feature... */
	if (!(options->mount_opt & (OCFS2_MOUNT_POSIX_ACL |
				    OCFS2_MOUNT_NO_POSIX_ACL))) {
		if (OCFS2_HAS_INCOMPAT_FEATURE(sb, OCFS2_FEATURE_INCOMPAT_XATTR))
			options->mount_opt |= OCFS2_MOUNT_POSIX_ACL;
		else
			options->mount_opt |= OCFS2_MOUNT_NO_POSIX_ACL;
	}
	return 1;
}

1288 1289
static int ocfs2_parse_options(struct super_block *sb,
			       char *options,
1290
			       struct mount_options *mopt,
1291 1292 1293 1294 1295 1296 1297 1298
			       int is_remount)
{
	int status;
	char *p;

	mlog_entry("remount: %d, options: \"%s\"\n", is_remount,
		   options ? options : "(none)");

1299
	mopt->commit_interval = 0;
1300
	mopt->mount_opt = OCFS2_MOUNT_NOINTR;
1301 1302
	mopt->atime_quantum = OCFS2_DEFAULT_ATIME_QUANTUM;
	mopt->slot = OCFS2_INVALID_SLOT;
1303
	mopt->localalloc_opt = -1;
1304
	mopt->cluster_stack[0] = '\0';
M
Mark Fasheh 已提交
1305
	mopt->resv_level = OCFS2_DEFAULT_RESV_LEVEL;
1306
	mopt->dir_resv_level = -1;
1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322

	if (!options) {
		status = 1;
		goto bail;
	}

	while ((p = strsep(&options, ",")) != NULL) {
		int token, option;
		substring_t args[MAX_OPT_ARGS];

		if (!*p)
			continue;

		token = match_token(p, tokens, args);
		switch (token) {
		case Opt_hb_local:
1323
			mopt->mount_opt |= OCFS2_MOUNT_HB_LOCAL;
1324 1325
			break;
		case Opt_hb_none:
1326
			mopt->mount_opt &= ~OCFS2_MOUNT_HB_LOCAL;
1327 1328 1329 1330 1331 1332 1333
			break;
		case Opt_barrier:
			if (match_int(&args[0], &option)) {
				status = 0;
				goto bail;
			}
			if (option)
1334
				mopt->mount_opt |= OCFS2_MOUNT_BARRIER;
1335
			else
1336
				mopt->mount_opt &= ~OCFS2_MOUNT_BARRIER;
1337 1338
			break;
		case Opt_intr:
1339
			mopt->mount_opt &= ~OCFS2_MOUNT_NOINTR;
1340 1341
			break;
		case Opt_nointr:
1342
			mopt->mount_opt |= OCFS2_MOUNT_NOINTR;
1343 1344
			break;
		case Opt_err_panic:
1345
			mopt->mount_opt |= OCFS2_MOUNT_ERRORS_PANIC;
1346 1347
			break;
		case Opt_err_ro:
1348
			mopt->mount_opt &= ~OCFS2_MOUNT_ERRORS_PANIC;
1349 1350
			break;
		case Opt_data_ordered:
1351
			mopt->mount_opt &= ~OCFS2_MOUNT_DATA_WRITEBACK;
1352 1353
			break;
		case Opt_data_writeback:
1354
			mopt->mount_opt |= OCFS2_MOUNT_DATA_WRITEBACK;
1355
			break;
T
Tiger Yang 已提交
1356 1357 1358 1359 1360 1361
		case Opt_user_xattr:
			mopt->mount_opt &= ~OCFS2_MOUNT_NOUSERXATTR;
			break;
		case Opt_nouser_xattr:
			mopt->mount_opt |= OCFS2_MOUNT_NOUSERXATTR;
			break;
T
Tiger Yang 已提交
1362 1363 1364 1365 1366 1367
		case Opt_atime_quantum:
			if (match_int(&args[0], &option)) {
				status = 0;
				goto bail;
			}
			if (option >= 0)
1368
				mopt->atime_quantum = option;
T
Tiger Yang 已提交
1369
			break;
1370 1371 1372 1373 1374 1375 1376
		case Opt_slot:
			option = 0;
			if (match_int(&args[0], &option)) {
				status = 0;
				goto bail;
			}
			if (option)
1377
				mopt->slot = (s16)option;
1378
			break;
1379 1380 1381 1382 1383 1384 1385 1386 1387
		case Opt_commit:
			option = 0;
			if (match_int(&args[0], &option)) {
				status = 0;
				goto bail;
			}
			if (option < 0)
				return 0;
			if (option == 0)
J
Joel Becker 已提交
1388
				option = JBD2_DEFAULT_MAX_COMMIT_AGE;
1389 1390
			mopt->commit_interval = HZ * option;
			break;
1391 1392 1393 1394 1395 1396
		case Opt_localalloc:
			option = 0;
			if (match_int(&args[0], &option)) {
				status = 0;
				goto bail;
			}
1397
			if (option >= 0)
1398 1399
				mopt->localalloc_opt = option;
			break;
1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413
		case Opt_localflocks:
			/*
			 * Changing this during remount could race
			 * flock() requests, or "unbalance" existing
			 * ones (e.g., a lock is taken in one mode but
			 * dropped in the other). If users care enough
			 * to flip locking modes during remount, we
			 * could add a "local" flag to individual
			 * flock structures for proper tracking of
			 * state.
			 */
			if (!is_remount)
				mopt->mount_opt |= OCFS2_MOUNT_LOCALFLOCKS;
			break;
1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432
		case Opt_stack:
			/* Check both that the option we were passed
			 * is of the right length and that it is a proper
			 * string of the right length.
			 */
			if (((args[0].to - args[0].from) !=
			     OCFS2_STACK_LABEL_LEN) ||
			    (strnlen(args[0].from,
				     OCFS2_STACK_LABEL_LEN) !=
			     OCFS2_STACK_LABEL_LEN)) {
				mlog(ML_ERROR,
				     "Invalid cluster_stack option\n");
				status = 0;
				goto bail;
			}
			memcpy(mopt->cluster_stack, args[0].from,
			       OCFS2_STACK_LABEL_LEN);
			mopt->cluster_stack[OCFS2_STACK_LABEL_LEN] = '\0';
			break;
1433 1434 1435
		case Opt_inode64:
			mopt->mount_opt |= OCFS2_MOUNT_INODE64;
			break;
1436 1437 1438 1439 1440 1441
		case Opt_usrquota:
			mopt->mount_opt |= OCFS2_MOUNT_USRQUOTA;
			break;
		case Opt_grpquota:
			mopt->mount_opt |= OCFS2_MOUNT_GRPQUOTA;
			break;
1442 1443
		case Opt_acl:
			mopt->mount_opt |= OCFS2_MOUNT_POSIX_ACL;
J
Jan Kara 已提交
1444
			mopt->mount_opt &= ~OCFS2_MOUNT_NO_POSIX_ACL;
1445 1446
			break;
		case Opt_noacl:
J
Jan Kara 已提交
1447
			mopt->mount_opt |= OCFS2_MOUNT_NO_POSIX_ACL;
1448 1449
			mopt->mount_opt &= ~OCFS2_MOUNT_POSIX_ACL;
			break;
M
Mark Fasheh 已提交
1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460
		case Opt_resv_level:
			if (is_remount)
				break;
			if (match_int(&args[0], &option)) {
				status = 0;
				goto bail;
			}
			if (option >= OCFS2_MIN_RESV_LEVEL &&
			    option < OCFS2_MAX_RESV_LEVEL)
				mopt->resv_level = option;
			break;
1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471
		case Opt_dir_resv_level:
			if (is_remount)
				break;
			if (match_int(&args[0], &option)) {
				status = 0;
				goto bail;
			}
			if (option >= OCFS2_MIN_RESV_LEVEL &&
			    option < OCFS2_MAX_RESV_LEVEL)
				mopt->dir_resv_level = option;
			break;
1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487
		default:
			mlog(ML_ERROR,
			     "Unrecognized mount option \"%s\" "
			     "or missing value\n", p);
			status = 0;
			goto bail;
		}
	}

	status = 1;

bail:
	mlog_exit(status);
	return status;
}

S
Sunil Mushran 已提交
1488 1489 1490 1491
static int ocfs2_show_options(struct seq_file *s, struct vfsmount *mnt)
{
	struct ocfs2_super *osb = OCFS2_SB(mnt->mnt_sb);
	unsigned long opts = osb->s_mount_opt;
1492
	unsigned int local_alloc_megs;
S
Sunil Mushran 已提交
1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520

	if (opts & OCFS2_MOUNT_HB_LOCAL)
		seq_printf(s, ",_netdev,heartbeat=local");
	else
		seq_printf(s, ",heartbeat=none");

	if (opts & OCFS2_MOUNT_NOINTR)
		seq_printf(s, ",nointr");

	if (opts & OCFS2_MOUNT_DATA_WRITEBACK)
		seq_printf(s, ",data=writeback");
	else
		seq_printf(s, ",data=ordered");

	if (opts & OCFS2_MOUNT_BARRIER)
		seq_printf(s, ",barrier=1");

	if (opts & OCFS2_MOUNT_ERRORS_PANIC)
		seq_printf(s, ",errors=panic");
	else
		seq_printf(s, ",errors=remount-ro");

	if (osb->preferred_slot != OCFS2_INVALID_SLOT)
		seq_printf(s, ",preferred_slot=%d", osb->preferred_slot);

	if (osb->s_atime_quantum != OCFS2_DEFAULT_ATIME_QUANTUM)
		seq_printf(s, ",atime_quantum=%u", osb->s_atime_quantum);

1521 1522 1523 1524
	if (osb->osb_commit_interval)
		seq_printf(s, ",commit=%u",
			   (unsigned) (osb->osb_commit_interval / HZ));

1525
	local_alloc_megs = osb->local_alloc_bits >> (20 - osb->s_clustersize_bits);
1526
	if (local_alloc_megs != ocfs2_la_default_mb(osb))
1527
		seq_printf(s, ",localalloc=%d", local_alloc_megs);
1528

1529 1530 1531
	if (opts & OCFS2_MOUNT_LOCALFLOCKS)
		seq_printf(s, ",localflocks,");

1532 1533 1534
	if (osb->osb_cluster_stack[0])
		seq_printf(s, ",cluster_stack=%.*s", OCFS2_STACK_LABEL_LEN,
			   osb->osb_cluster_stack);
1535 1536 1537 1538
	if (opts & OCFS2_MOUNT_USRQUOTA)
		seq_printf(s, ",usrquota");
	if (opts & OCFS2_MOUNT_GRPQUOTA)
		seq_printf(s, ",grpquota");
1539

1540 1541 1542 1543 1544
	if (opts & OCFS2_MOUNT_NOUSERXATTR)
		seq_printf(s, ",nouser_xattr");
	else
		seq_printf(s, ",user_xattr");

1545 1546 1547
	if (opts & OCFS2_MOUNT_INODE64)
		seq_printf(s, ",inode64");

1548 1549 1550 1551 1552
	if (opts & OCFS2_MOUNT_POSIX_ACL)
		seq_printf(s, ",acl");
	else
		seq_printf(s, ",noacl");

M
Mark Fasheh 已提交
1553 1554 1555
	if (osb->osb_resv_level != OCFS2_DEFAULT_RESV_LEVEL)
		seq_printf(s, ",resv_level=%d", osb->osb_resv_level);

1556 1557 1558
	if (osb->osb_dir_resv_level != osb->osb_resv_level)
		seq_printf(s, ",dir_resv_level=%d", osb->osb_resv_level);

S
Sunil Mushran 已提交
1559 1560 1561
	return 0;
}

1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593
static int __init ocfs2_init(void)
{
	int status;

	mlog_entry_void();

	ocfs2_print_version();

	status = init_ocfs2_uptodate_cache();
	if (status < 0) {
		mlog_errno(status);
		goto leave;
	}

	status = ocfs2_initialize_mem_caches();
	if (status < 0) {
		mlog_errno(status);
		goto leave;
	}

	ocfs2_wq = create_singlethread_workqueue("ocfs2_wq");
	if (!ocfs2_wq) {
		status = -ENOMEM;
		goto leave;
	}

	ocfs2_debugfs_root = debugfs_create_dir("ocfs2", NULL);
	if (!ocfs2_debugfs_root) {
		status = -EFAULT;
		mlog(ML_ERROR, "Unable to create ocfs2 debugfs root.\n");
	}

M
Mark Fasheh 已提交
1594 1595 1596 1597
	status = ocfs2_quota_setup();
	if (status)
		goto leave;

1598 1599
	ocfs2_set_locking_protocol();

1600
	status = register_quota_format(&ocfs2_quota_format);
1601 1602
leave:
	if (status < 0) {
M
Mark Fasheh 已提交
1603
		ocfs2_quota_shutdown();
1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619
		ocfs2_free_mem_caches();
		exit_ocfs2_uptodate_cache();
	}

	mlog_exit(status);

	if (status >= 0) {
		return register_filesystem(&ocfs2_fs_type);
	} else
		return -1;
}

static void __exit ocfs2_exit(void)
{
	mlog_entry_void();

M
Mark Fasheh 已提交
1620 1621
	ocfs2_quota_shutdown();

1622 1623 1624 1625 1626
	if (ocfs2_wq) {
		flush_workqueue(ocfs2_wq);
		destroy_workqueue(ocfs2_wq);
	}

1627 1628
	unregister_quota_format(&ocfs2_quota_format);

1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643
	debugfs_remove(ocfs2_debugfs_root);

	ocfs2_free_mem_caches();

	unregister_filesystem(&ocfs2_fs_type);

	exit_ocfs2_uptodate_cache();

	mlog_exit_void();
}

static void ocfs2_put_super(struct super_block *sb)
{
	mlog_entry("(0x%p)\n", sb);

1644 1645
	lock_kernel();

1646 1647 1648
	ocfs2_sync_blockdev(sb);
	ocfs2_dismount_volume(sb, 0);

1649 1650
	unlock_kernel();

1651 1652 1653
	mlog_exit_void();
}

1654
static int ocfs2_statfs(struct dentry *dentry, struct kstatfs *buf)
1655 1656 1657 1658 1659 1660 1661 1662
{
	struct ocfs2_super *osb;
	u32 numbits, freebits;
	int status;
	struct ocfs2_dinode *bm_lock;
	struct buffer_head *bh = NULL;
	struct inode *inode = NULL;

1663
	mlog_entry("(%p, %p)\n", dentry->d_sb, buf);
1664

1665
	osb = OCFS2_SB(dentry->d_sb);
1666 1667 1668 1669 1670 1671 1672 1673 1674 1675

	inode = ocfs2_get_system_file_inode(osb,
					    GLOBAL_BITMAP_SYSTEM_INODE,
					    OCFS2_INVALID_SLOT);
	if (!inode) {
		mlog(ML_ERROR, "failed to get bitmap inode\n");
		status = -EIO;
		goto bail;
	}

M
Mark Fasheh 已提交
1676
	status = ocfs2_inode_lock(inode, &bh, 0);
1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687
	if (status < 0) {
		mlog_errno(status);
		goto bail;
	}

	bm_lock = (struct ocfs2_dinode *) bh->b_data;

	numbits = le32_to_cpu(bm_lock->id1.bitmap1.i_total);
	freebits = numbits - le32_to_cpu(bm_lock->id1.bitmap1.i_used);

	buf->f_type = OCFS2_SUPER_MAGIC;
1688
	buf->f_bsize = dentry->d_sb->s_blocksize;
1689 1690 1691 1692 1693 1694 1695 1696
	buf->f_namelen = OCFS2_MAX_FILENAME_LEN;
	buf->f_blocks = ((sector_t) numbits) *
			(osb->s_clustersize >> osb->sb->s_blocksize_bits);
	buf->f_bfree = ((sector_t) freebits) *
		       (osb->s_clustersize >> osb->sb->s_blocksize_bits);
	buf->f_bavail = buf->f_bfree;
	buf->f_files = numbits;
	buf->f_ffree = freebits;
1697 1698 1699 1700
	buf->f_fsid.val[0] = crc32_le(0, osb->uuid_str, OCFS2_VOL_UUID_LEN)
				& 0xFFFFFFFFUL;
	buf->f_fsid.val[1] = crc32_le(0, osb->uuid_str + OCFS2_VOL_UUID_LEN,
				OCFS2_VOL_UUID_LEN) & 0xFFFFFFFFUL;
1701 1702 1703

	brelse(bh);

M
Mark Fasheh 已提交
1704
	ocfs2_inode_unlock(inode, 0);
1705 1706 1707 1708 1709 1710 1711 1712 1713 1714
	status = 0;
bail:
	if (inode)
		iput(inode);

	mlog_exit(status);

	return status;
}

1715
static void ocfs2_inode_init_once(void *data)
1716 1717 1718
{
	struct ocfs2_inode_info *oi = data;

C
Christoph Lameter 已提交
1719 1720 1721 1722 1723 1724
	oi->ip_flags = 0;
	oi->ip_open_count = 0;
	spin_lock_init(&oi->ip_lock);
	ocfs2_extent_map_init(&oi->vfs_inode);
	INIT_LIST_HEAD(&oi->ip_io_markers);
	oi->ip_dir_start_lookup = 0;
1725

C
Christoph Lameter 已提交
1726
	init_rwsem(&oi->ip_alloc_sem);
T
Tiger Yang 已提交
1727
	init_rwsem(&oi->ip_xattr_sem);
C
Christoph Lameter 已提交
1728
	mutex_init(&oi->ip_io_mutex);
1729

C
Christoph Lameter 已提交
1730 1731
	oi->ip_blkno = 0ULL;
	oi->ip_clusters = 0;
1732

1733 1734
	ocfs2_resv_init_once(&oi->ip_la_data_resv);

C
Christoph Lameter 已提交
1735
	ocfs2_lock_res_init_once(&oi->ip_rw_lockres);
M
Mark Fasheh 已提交
1736
	ocfs2_lock_res_init_once(&oi->ip_inode_lockres);
C
Christoph Lameter 已提交
1737
	ocfs2_lock_res_init_once(&oi->ip_open_lockres);
1738

1739
	ocfs2_metadata_cache_init(INODE_CACHE(&oi->vfs_inode),
1740
				  &ocfs2_inode_caching_ops);
1741

C
Christoph Lameter 已提交
1742
	inode_init_once(&oi->vfs_inode);
1743 1744 1745 1746 1747
}

static int ocfs2_initialize_mem_caches(void)
{
	ocfs2_inode_cachep = kmem_cache_create("ocfs2_inode_cache",
1748 1749 1750 1751
				       sizeof(struct ocfs2_inode_info),
				       0,
				       (SLAB_HWCACHE_ALIGN|SLAB_RECLAIM_ACCOUNT|
						SLAB_MEM_SPREAD),
1752
				       ocfs2_inode_init_once);
1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771
	ocfs2_dquot_cachep = kmem_cache_create("ocfs2_dquot_cache",
					sizeof(struct ocfs2_dquot),
					0,
					(SLAB_HWCACHE_ALIGN|SLAB_RECLAIM_ACCOUNT|
						SLAB_MEM_SPREAD),
					NULL);
	ocfs2_qf_chunk_cachep = kmem_cache_create("ocfs2_qf_chunk_cache",
					sizeof(struct ocfs2_quota_chunk),
					0,
					(SLAB_RECLAIM_ACCOUNT|SLAB_MEM_SPREAD),
					NULL);
	if (!ocfs2_inode_cachep || !ocfs2_dquot_cachep ||
	    !ocfs2_qf_chunk_cachep) {
		if (ocfs2_inode_cachep)
			kmem_cache_destroy(ocfs2_inode_cachep);
		if (ocfs2_dquot_cachep)
			kmem_cache_destroy(ocfs2_dquot_cachep);
		if (ocfs2_qf_chunk_cachep)
			kmem_cache_destroy(ocfs2_qf_chunk_cachep);
1772
		return -ENOMEM;
1773
	}
1774 1775 1776 1777 1778 1779 1780 1781 1782

	return 0;
}

static void ocfs2_free_mem_caches(void)
{
	if (ocfs2_inode_cachep)
		kmem_cache_destroy(ocfs2_inode_cachep);
	ocfs2_inode_cachep = NULL;
1783 1784 1785 1786 1787 1788 1789 1790

	if (ocfs2_dquot_cachep)
		kmem_cache_destroy(ocfs2_dquot_cachep);
	ocfs2_dquot_cachep = NULL;

	if (ocfs2_qf_chunk_cachep)
		kmem_cache_destroy(ocfs2_qf_chunk_cachep);
	ocfs2_qf_chunk_cachep = NULL;
1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813
}

static int ocfs2_get_sector(struct super_block *sb,
			    struct buffer_head **bh,
			    int block,
			    int sect_size)
{
	if (!sb_set_blocksize(sb, sect_size)) {
		mlog(ML_ERROR, "unable to set blocksize\n");
		return -EIO;
	}

	*bh = sb_getblk(sb, block);
	if (!*bh) {
		mlog_errno(-EIO);
		return -EIO;
	}
	lock_buffer(*bh);
	if (!buffer_dirty(*bh))
		clear_buffer_uptodate(*bh);
	unlock_buffer(*bh);
	ll_rw_block(READ, 1, bh);
	wait_on_buffer(*bh);
1814 1815 1816 1817 1818 1819 1820
	if (!buffer_uptodate(*bh)) {
		mlog_errno(-EIO);
		brelse(*bh);
		*bh = NULL;
		return -EIO;
	}

1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868
	return 0;
}

static int ocfs2_mount_volume(struct super_block *sb)
{
	int status = 0;
	int unlock_super = 0;
	struct ocfs2_super *osb = OCFS2_SB(sb);

	mlog_entry_void();

	if (ocfs2_is_hard_readonly(osb))
		goto leave;

	status = ocfs2_dlm_init(osb);
	if (status < 0) {
		mlog_errno(status);
		goto leave;
	}

	status = ocfs2_super_lock(osb, 1);
	if (status < 0) {
		mlog_errno(status);
		goto leave;
	}
	unlock_super = 1;

	/* This will load up the node map and add ourselves to it. */
	status = ocfs2_find_slot(osb);
	if (status < 0) {
		mlog_errno(status);
		goto leave;
	}

	/* load all node-local system inodes */
	status = ocfs2_init_local_system_inodes(osb);
	if (status < 0) {
		mlog_errno(status);
		goto leave;
	}

	status = ocfs2_check_volume(osb);
	if (status < 0) {
		mlog_errno(status);
		goto leave;
	}

	status = ocfs2_truncate_log_init(osb);
1869
	if (status < 0)
1870
		mlog_errno(status);
S
Sunil Mushran 已提交
1871

1872 1873 1874 1875 1876 1877 1878 1879 1880 1881
leave:
	if (unlock_super)
		ocfs2_super_unlock(osb, 1);

	mlog_exit(status);
	return status;
}

static void ocfs2_dismount_volume(struct super_block *sb, int mnt_err)
{
1882
	int tmp, hangup_needed = 0;
1883
	struct ocfs2_super *osb = NULL;
S
Sunil Mushran 已提交
1884
	char nodestr[8];
1885 1886 1887 1888 1889 1890 1891

	mlog_entry("(0x%p)\n", sb);

	BUG_ON(!sb);
	osb = OCFS2_SB(sb);
	BUG_ON(!osb);

1892 1893
	debugfs_remove(osb->osb_ctxt);

J
Jan Kara 已提交
1894 1895 1896 1897 1898 1899
	/*
	 * Flush inode dropping work queue so that deletes are
	 * performed while the filesystem is still working
	 */
	ocfs2_drop_all_dl_inodes(osb);

1900 1901 1902
	/* Orphan scan should be stopped as early as possible */
	ocfs2_orphan_scan_stop(osb);

1903 1904
	ocfs2_disable_quotas(osb);

1905 1906 1907 1908
	ocfs2_shutdown_local_alloc(osb);

	ocfs2_truncate_log_shutdown(osb);

1909 1910
	/* This will disable recovery and flush any recovery work. */
	ocfs2_recovery_exit(osb);
1911 1912 1913 1914 1915

	ocfs2_journal_shutdown(osb);

	ocfs2_sync_blockdev(sb);

1916 1917
	ocfs2_purge_refcount_trees(osb);

1918 1919 1920
	/* No cluster connection means we've failed during mount, so skip
	 * all the steps which depended on that to complete. */
	if (osb->cconn) {
1921 1922 1923 1924 1925
		tmp = ocfs2_super_lock(osb, 1);
		if (tmp < 0) {
			mlog_errno(tmp);
			return;
		}
1926
	}
1927

1928 1929
	if (osb->slot_num != OCFS2_INVALID_SLOT)
		ocfs2_put_slot(osb);
1930

1931
	if (osb->cconn)
1932 1933 1934 1935
		ocfs2_super_unlock(osb, 1);

	ocfs2_release_system_inodes(osb);

1936 1937 1938 1939 1940 1941 1942
	/*
	 * If we're dismounting due to mount error, mount.ocfs2 will clean
	 * up heartbeat.  If we're a local mount, there is no heartbeat.
	 * If we failed before we got a uuid_str yet, we can't stop
	 * heartbeat.  Otherwise, do it.
	 */
	if (!mnt_err && !ocfs2_mount_local(osb) && osb->uuid_str)
1943 1944 1945 1946 1947
		hangup_needed = 1;

	if (osb->cconn)
		ocfs2_dlm_shutdown(osb, hangup_needed);

1948
	ocfs2_blockcheck_stats_debugfs_remove(&osb->osb_ecc_stats);
1949 1950 1951
	debugfs_remove(osb->osb_debug_root);

	if (hangup_needed)
1952
		ocfs2_cluster_hangup(osb->uuid_str, strlen(osb->uuid_str));
1953 1954 1955

	atomic_set(&osb->vol_state, VOLUME_DISMOUNTED);

S
Sunil Mushran 已提交
1956 1957 1958
	if (ocfs2_mount_local(osb))
		snprintf(nodestr, sizeof(nodestr), "local");
	else
1959
		snprintf(nodestr, sizeof(nodestr), "%u", osb->node_num);
S
Sunil Mushran 已提交
1960 1961 1962

	printk(KERN_INFO "ocfs2: Unmounting device (%s) on (node %s)\n",
	       osb->dev_str, nodestr);
1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977

	ocfs2_delete_osb(osb);
	kfree(osb);
	sb->s_dev = 0;
	sb->s_fs_info = NULL;
}

static int ocfs2_setup_osb_uuid(struct ocfs2_super *osb, const unsigned char *uuid,
				unsigned uuid_bytes)
{
	int i, ret;
	char *ptr;

	BUG_ON(uuid_bytes != OCFS2_VOL_UUID_LEN);

1978
	osb->uuid_str = kzalloc(OCFS2_VOL_UUID_LEN * 2 + 1, GFP_KERNEL);
1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995
	if (osb->uuid_str == NULL)
		return -ENOMEM;

	for (i = 0, ptr = osb->uuid_str; i < OCFS2_VOL_UUID_LEN; i++) {
		/* print with null */
		ret = snprintf(ptr, 3, "%02X", uuid[i]);
		if (ret != 2) /* drop super cleans up */
			return -EINVAL;
		/* then only advance past the last char */
		ptr += 2;
	}

	return 0;
}

static int ocfs2_initialize_super(struct super_block *sb,
				  struct buffer_head *bh,
1996 1997
				  int sector_size,
				  struct ocfs2_blockcheck_stats *stats)
1998
{
1999
	int status;
M
Mark Fasheh 已提交
2000 2001
	int i, cbits, bbits;
	struct ocfs2_dinode *di = (struct ocfs2_dinode *)bh->b_data;
2002 2003 2004 2005 2006 2007 2008
	struct inode *inode = NULL;
	struct ocfs2_journal *journal;
	__le32 uuid_net_key;
	struct ocfs2_super *osb;

	mlog_entry_void();

2009
	osb = kzalloc(sizeof(struct ocfs2_super), GFP_KERNEL);
2010 2011 2012 2013 2014 2015 2016 2017 2018
	if (!osb) {
		status = -ENOMEM;
		mlog_errno(status);
		goto bail;
	}

	sb->s_fs_info = osb;
	sb->s_op = &ocfs2_sops;
	sb->s_export_op = &ocfs2_export_ops;
2019 2020
	sb->s_qcop = &ocfs2_quotactl_ops;
	sb->dq_op = &ocfs2_quota_operations;
T
Tiger Yang 已提交
2021
	sb->s_xattr = ocfs2_xattr_handlers;
2022
	sb->s_time_gran = 1;
2023 2024
	sb->s_flags |= MS_NOATIME;
	/* this is needed to support O_LARGEFILE */
M
Mark Fasheh 已提交
2025 2026 2027
	cbits = le32_to_cpu(di->id2.i_super.s_clustersize_bits);
	bbits = le32_to_cpu(di->id2.i_super.s_blocksize_bits);
	sb->s_maxbytes = ocfs2_max_file_offset(bbits, cbits);
2028

2029 2030 2031 2032 2033 2034
	osb->osb_dx_mask = (1 << (cbits - bbits)) - 1;

	for (i = 0; i < 3; i++)
		osb->osb_dx_seed[i] = le32_to_cpu(di->id2.i_super.s_dx_seed[i]);
	osb->osb_dx_seed[3] = le32_to_cpu(di->id2.i_super.s_uuid_hash);

2035 2036 2037
	osb->sb = sb;
	/* Save off for ocfs2_rw_direct */
	osb->s_sectsize_bits = blksize_bits(sector_size);
2038
	BUG_ON(!osb->s_sectsize_bits);
2039

M
Mark Fasheh 已提交
2040 2041 2042 2043
	spin_lock_init(&osb->dc_task_lock);
	init_waitqueue_head(&osb->dc_event);
	osb->dc_work_sequence = 0;
	osb->dc_wake_sequence = 0;
2044 2045 2046
	INIT_LIST_HEAD(&osb->blocked_lock_list);
	osb->blocked_lock_count = 0;
	spin_lock_init(&osb->osb_lock);
2047
	spin_lock_init(&osb->osb_xattr_lock);
2048
	ocfs2_init_steal_slots(osb);
2049 2050 2051 2052 2053 2054 2055

	atomic_set(&osb->alloc_stats.moves, 0);
	atomic_set(&osb->alloc_stats.local_data, 0);
	atomic_set(&osb->alloc_stats.bitmap_data, 0);
	atomic_set(&osb->alloc_stats.bg_allocs, 0);
	atomic_set(&osb->alloc_stats.bg_extends, 0);

2056 2057 2058
	/* Copy the blockcheck stats from the superblock probe */
	osb->osb_ecc_stats = *stats;

2059 2060 2061 2062 2063
	ocfs2_init_node_maps(osb);

	snprintf(osb->dev_str, sizeof(osb->dev_str), "%u,%u",
		 MAJOR(osb->sb->s_dev), MINOR(osb->sb->s_dev));

2064 2065
	ocfs2_orphan_scan_init(osb);

2066 2067 2068 2069 2070 2071
	status = ocfs2_recovery_init(osb);
	if (status) {
		mlog(ML_ERROR, "Unable to initialize recovery state\n");
		mlog_errno(status);
		goto bail;
	}
2072 2073 2074 2075

	init_waitqueue_head(&osb->checkpoint_event);
	atomic_set(&osb->needs_checkpoint, 0);

T
Tiger Yang 已提交
2076 2077
	osb->s_atime_quantum = OCFS2_DEFAULT_ATIME_QUANTUM;

2078 2079
	osb->slot_num = OCFS2_INVALID_SLOT;

2080 2081
	osb->s_xattr_inline_size = le16_to_cpu(
					di->id2.i_super.s_xattr_inline_size);
2082

2083 2084
	osb->local_alloc_state = OCFS2_LA_UNUSED;
	osb->local_alloc_bh = NULL;
2085
	INIT_DELAYED_WORK(&osb->la_enable_wq, ocfs2_la_enable_worker);
2086 2087 2088

	init_waitqueue_head(&osb->osb_mount_event);

M
Mark Fasheh 已提交
2089 2090 2091 2092 2093 2094
	status = ocfs2_resmap_init(osb, &osb->osb_la_resmap);
	if (status) {
		mlog_errno(status);
		goto bail;
	}

2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108
	osb->vol_label = kmalloc(OCFS2_MAX_VOL_LABEL_LEN, GFP_KERNEL);
	if (!osb->vol_label) {
		mlog(ML_ERROR, "unable to alloc vol label\n");
		status = -ENOMEM;
		goto bail;
	}

	osb->max_slots = le16_to_cpu(di->id2.i_super.s_max_slots);
	if (osb->max_slots > OCFS2_MAX_SLOTS || osb->max_slots == 0) {
		mlog(ML_ERROR, "Invalid number of node slots (%u)\n",
		     osb->max_slots);
		status = -EINVAL;
		goto bail;
	}
S
Sunil Mushran 已提交
2109
	mlog(0, "max_slots for this device: %u\n", osb->max_slots);
2110

2111 2112 2113 2114 2115 2116 2117 2118 2119
	osb->slot_recovery_generations =
		kcalloc(osb->max_slots, sizeof(*osb->slot_recovery_generations),
			GFP_KERNEL);
	if (!osb->slot_recovery_generations) {
		status = -ENOMEM;
		mlog_errno(status);
		goto bail;
	}

2120 2121 2122 2123 2124 2125 2126 2127 2128 2129
	init_waitqueue_head(&osb->osb_wipe_event);
	osb->osb_orphan_wipes = kcalloc(osb->max_slots,
					sizeof(*osb->osb_orphan_wipes),
					GFP_KERNEL);
	if (!osb->osb_orphan_wipes) {
		status = -ENOMEM;
		mlog_errno(status);
		goto bail;
	}

2130 2131
	osb->osb_rf_lock_tree = RB_ROOT;

2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152
	osb->s_feature_compat =
		le32_to_cpu(OCFS2_RAW_SB(di)->s_feature_compat);
	osb->s_feature_ro_compat =
		le32_to_cpu(OCFS2_RAW_SB(di)->s_feature_ro_compat);
	osb->s_feature_incompat =
		le32_to_cpu(OCFS2_RAW_SB(di)->s_feature_incompat);

	if ((i = OCFS2_HAS_INCOMPAT_FEATURE(osb->sb, ~OCFS2_FEATURE_INCOMPAT_SUPP))) {
		mlog(ML_ERROR, "couldn't mount because of unsupported "
		     "optional features (%x).\n", i);
		status = -EINVAL;
		goto bail;
	}
	if (!(osb->sb->s_flags & MS_RDONLY) &&
	    (i = OCFS2_HAS_RO_COMPAT_FEATURE(osb->sb, ~OCFS2_FEATURE_RO_COMPAT_SUPP))) {
		mlog(ML_ERROR, "couldn't mount RDWR because of "
		     "unsupported optional features (%x).\n", i);
		status = -EINVAL;
		goto bail;
	}

2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171
	if (ocfs2_userspace_stack(osb)) {
		memcpy(osb->osb_cluster_stack,
		       OCFS2_RAW_SB(di)->s_cluster_info.ci_stack,
		       OCFS2_STACK_LABEL_LEN);
		osb->osb_cluster_stack[OCFS2_STACK_LABEL_LEN] = '\0';
		if (strlen(osb->osb_cluster_stack) != OCFS2_STACK_LABEL_LEN) {
			mlog(ML_ERROR,
			     "couldn't mount because of an invalid "
			     "cluster stack label (%s) \n",
			     osb->osb_cluster_stack);
			status = -EINVAL;
			goto bail;
		}
	} else {
		/* The empty string is identical with classic tools that
		 * don't know about s_cluster_info. */
		osb->osb_cluster_stack[0] = '\0';
	}

2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182
	get_random_bytes(&osb->s_next_generation, sizeof(u32));

	/* FIXME
	 * This should be done in ocfs2_journal_init(), but unknown
	 * ordering issues will cause the filesystem to crash.
	 * If anyone wants to figure out what part of the code
	 * refers to osb->journal before ocfs2_journal_init() is run,
	 * be my guest.
	 */
	/* initialize our journal structure */

2183
	journal = kzalloc(sizeof(struct ocfs2_journal), GFP_KERNEL);
2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197
	if (!journal) {
		mlog(ML_ERROR, "unable to alloc journal\n");
		status = -ENOMEM;
		goto bail;
	}
	osb->journal = journal;
	journal->j_osb = osb;

	atomic_set(&journal->j_num_trans, 0);
	init_rwsem(&journal->j_trans_barrier);
	init_waitqueue_head(&journal->j_checkpointed);
	spin_lock_init(&journal->j_lock);
	journal->j_trans_id = (unsigned long) 1;
	INIT_LIST_HEAD(&journal->j_la_cleanups);
D
David Howells 已提交
2198
	INIT_WORK(&journal->j_recovery_work, ocfs2_complete_recovery);
2199 2200
	journal->j_state = OCFS2_JOURNAL_FREE;

2201 2202 2203
	INIT_WORK(&osb->dentry_lock_work, ocfs2_drop_dl_inodes);
	osb->dentry_lock_list = NULL;

2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232
	/* get some pseudo constants for clustersize bits */
	osb->s_clustersize_bits =
		le32_to_cpu(di->id2.i_super.s_clustersize_bits);
	osb->s_clustersize = 1 << osb->s_clustersize_bits;
	mlog(0, "clusterbits=%d\n", osb->s_clustersize_bits);

	if (osb->s_clustersize < OCFS2_MIN_CLUSTERSIZE ||
	    osb->s_clustersize > OCFS2_MAX_CLUSTERSIZE) {
		mlog(ML_ERROR, "Volume has invalid cluster size (%d)\n",
		     osb->s_clustersize);
		status = -EINVAL;
		goto bail;
	}

	if (ocfs2_clusters_to_blocks(osb->sb, le32_to_cpu(di->i_clusters) - 1)
	    > (u32)~0UL) {
		mlog(ML_ERROR, "Volume might try to write to blocks beyond "
		     "what jbd can address in 32 bits.\n");
		status = -EINVAL;
		goto bail;
	}

	if (ocfs2_setup_osb_uuid(osb, di->id2.i_super.s_uuid,
				 sizeof(di->id2.i_super.s_uuid))) {
		mlog(ML_ERROR, "Out of memory trying to setup our uuid.\n");
		status = -ENOMEM;
		goto bail;
	}

M
Mark Fasheh 已提交
2233
	memcpy(&uuid_net_key, di->id2.i_super.s_uuid, sizeof(uuid_net_key));
2234 2235 2236 2237 2238 2239 2240 2241

	strncpy(osb->vol_label, di->id2.i_super.s_label, 63);
	osb->vol_label[63] = '\0';
	osb->root_blkno = le64_to_cpu(di->id2.i_super.s_root_blkno);
	osb->system_dir_blkno = le64_to_cpu(di->id2.i_super.s_system_dir_blkno);
	osb->first_cluster_group_blkno =
		le64_to_cpu(di->id2.i_super.s_first_cluster_group);
	osb->fs_generation = le32_to_cpu(di->i_fs_generation);
T
Tiger Yang 已提交
2242
	osb->uuid_hash = le32_to_cpu(di->id2.i_super.s_uuid_hash);
2243 2244
	mlog(0, "vol_label: %s\n", osb->vol_label);
	mlog(0, "uuid: %s\n", osb->uuid_str);
2245 2246 2247
	mlog(0, "root_blkno=%llu, system_dir_blkno=%llu\n",
	     (unsigned long long)osb->root_blkno,
	     (unsigned long long)osb->system_dir_blkno);
2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276

	osb->osb_dlm_debug = ocfs2_new_dlm_debug();
	if (!osb->osb_dlm_debug) {
		status = -ENOMEM;
		mlog_errno(status);
		goto bail;
	}

	atomic_set(&osb->vol_state, VOLUME_INIT);

	/* load root, system_dir, and all global system inodes */
	status = ocfs2_init_global_system_inodes(osb);
	if (status < 0) {
		mlog_errno(status);
		goto bail;
	}

	/*
	 * global bitmap
	 */
	inode = ocfs2_get_system_file_inode(osb, GLOBAL_BITMAP_SYSTEM_INODE,
					    OCFS2_INVALID_SLOT);
	if (!inode) {
		status = -EINVAL;
		mlog_errno(status);
		goto bail;
	}

	osb->bitmap_blkno = OCFS2_I(inode)->ip_blkno;
2277
	osb->osb_clusters_at_boot = OCFS2_I(inode)->ip_clusters;
2278 2279
	iput(inode);

2280 2281
	osb->bitmap_cpg = ocfs2_group_bitmap_size(sb, 0,
				 osb->s_feature_incompat) * 8;
2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300

	status = ocfs2_init_slot_info(osb);
	if (status < 0) {
		mlog_errno(status);
		goto bail;
	}

bail:
	mlog_exit(status);
	return status;
}

/*
 * will return: -EAGAIN if it is ok to keep searching for superblocks
 *              -EINVAL if there is a bad superblock
 *              0 on success
 */
static int ocfs2_verify_volume(struct ocfs2_dinode *di,
			       struct buffer_head *bh,
2301 2302
			       u32 blksz,
			       struct ocfs2_blockcheck_stats *stats)
2303 2304 2305 2306 2307 2308 2309
{
	int status = -EAGAIN;

	mlog_entry_void();

	if (memcmp(di->i_signature, OCFS2_SUPER_BLOCK_SIGNATURE,
		   strlen(OCFS2_SUPER_BLOCK_SIGNATURE)) == 0) {
2310 2311 2312 2313 2314
		/* We have to do a raw check of the feature here */
		if (le32_to_cpu(di->id2.i_super.s_feature_incompat) &
		    OCFS2_FEATURE_INCOMPAT_META_ECC) {
			status = ocfs2_block_check_validate(bh->b_data,
							    bh->b_size,
2315 2316
							    &di->i_check,
							    stats);
2317 2318 2319
			if (status)
				goto out;
		}
2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337
		status = -EINVAL;
		if ((1 << le32_to_cpu(di->id2.i_super.s_blocksize_bits)) != blksz) {
			mlog(ML_ERROR, "found superblock with incorrect block "
			     "size: found %u, should be %u\n",
			     1 << le32_to_cpu(di->id2.i_super.s_blocksize_bits),
			       blksz);
		} else if (le16_to_cpu(di->id2.i_super.s_major_rev_level) !=
			   OCFS2_MAJOR_REV_LEVEL ||
			   le16_to_cpu(di->id2.i_super.s_minor_rev_level) !=
			   OCFS2_MINOR_REV_LEVEL) {
			mlog(ML_ERROR, "found superblock with bad version: "
			     "found %u.%u, should be %u.%u\n",
			     le16_to_cpu(di->id2.i_super.s_major_rev_level),
			     le16_to_cpu(di->id2.i_super.s_minor_rev_level),
			     OCFS2_MAJOR_REV_LEVEL,
			     OCFS2_MINOR_REV_LEVEL);
		} else if (bh->b_blocknr != le64_to_cpu(di->i_blkno)) {
			mlog(ML_ERROR, "bad block number on superblock: "
2338
			     "found %llu, should be %llu\n",
2339
			     (unsigned long long)le64_to_cpu(di->i_blkno),
2340
			     (unsigned long long)bh->b_blocknr);
2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360
		} else if (le32_to_cpu(di->id2.i_super.s_clustersize_bits) < 12 ||
			    le32_to_cpu(di->id2.i_super.s_clustersize_bits) > 20) {
			mlog(ML_ERROR, "bad cluster size found: %u\n",
			     1 << le32_to_cpu(di->id2.i_super.s_clustersize_bits));
		} else if (!le64_to_cpu(di->id2.i_super.s_root_blkno)) {
			mlog(ML_ERROR, "bad root_blkno: 0\n");
		} else if (!le64_to_cpu(di->id2.i_super.s_system_dir_blkno)) {
			mlog(ML_ERROR, "bad system_dir_blkno: 0\n");
		} else if (le16_to_cpu(di->id2.i_super.s_max_slots) > OCFS2_MAX_SLOTS) {
			mlog(ML_ERROR,
			     "Superblock slots found greater than file system "
			     "maximum: found %u, max %u\n",
			     le16_to_cpu(di->id2.i_super.s_max_slots),
			     OCFS2_MAX_SLOTS);
		} else {
			/* found it! */
			status = 0;
		}
	}

2361
out:
2362 2363 2364 2365 2366 2367
	mlog_exit(status);
	return status;
}

static int ocfs2_check_volume(struct ocfs2_super *osb)
{
2368
	int status;
2369
	int dirty;
S
Sunil Mushran 已提交
2370
	int local;
2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397
	struct ocfs2_dinode *local_alloc = NULL; /* only used if we
						  * recover
						  * ourselves. */

	mlog_entry_void();

	/* Init our journal object. */
	status = ocfs2_journal_init(osb->journal, &dirty);
	if (status < 0) {
		mlog(ML_ERROR, "Could not initialize journal!\n");
		goto finally;
	}

	/* If the journal was unmounted cleanly then we don't want to
	 * recover anything. Otherwise, journal_load will do that
	 * dirty work for us :) */
	if (!dirty) {
		status = ocfs2_journal_wipe(osb->journal, 0);
		if (status < 0) {
			mlog_errno(status);
			goto finally;
		}
	} else {
		mlog(ML_NOTICE, "File system was not unmounted cleanly, "
		     "recovering volume.\n");
	}

S
Sunil Mushran 已提交
2398 2399
	local = ocfs2_mount_local(osb);

2400
	/* will play back anything left in the journal. */
2401
	status = ocfs2_journal_load(osb->journal, local, dirty);
2402 2403 2404 2405
	if (status < 0) {
		mlog(ML_ERROR, "ocfs2 journal load failed! %d\n", status);
		goto finally;
	}
2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439

	if (dirty) {
		/* recover my local alloc if we didn't unmount cleanly. */
		status = ocfs2_begin_local_alloc_recovery(osb,
							  osb->slot_num,
							  &local_alloc);
		if (status < 0) {
			mlog_errno(status);
			goto finally;
		}
		/* we complete the recovery process after we've marked
		 * ourselves as mounted. */
	}

	mlog(0, "Journal loaded.\n");

	status = ocfs2_load_local_alloc(osb);
	if (status < 0) {
		mlog_errno(status);
		goto finally;
	}

	if (dirty) {
		/* Recovery will be completed after we've mounted the
		 * rest of the volume. */
		osb->dirty = 1;
		osb->local_alloc_copy = local_alloc;
		local_alloc = NULL;
	}

	/* go through each journal, trylock it and if you get the
	 * lock, and it's marked as dirty, set the bit in the recover
	 * map and launch a recovery thread for it. */
	status = ocfs2_mark_dead_nodes(osb);
2440 2441 2442 2443 2444 2445
	if (status < 0) {
		mlog_errno(status);
		goto finally;
	}

	status = ocfs2_compute_replay_slots(osb);
2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468
	if (status < 0)
		mlog_errno(status);

finally:
	if (local_alloc)
		kfree(local_alloc);

	mlog_exit(status);
	return status;
}

/*
 * The routine gets called from dismount or close whenever a dismount on
 * volume is requested and the osb open count becomes 1.
 * It will remove the osb from the global list and also free up all the
 * initialized resources and fileobject.
 */
static void ocfs2_delete_osb(struct ocfs2_super *osb)
{
	mlog_entry_void();

	/* This function assumes that the caller has the main osb resource */

2469
	ocfs2_free_slot_info(osb);
2470

2471
	kfree(osb->osb_orphan_wipes);
2472
	kfree(osb->slot_recovery_generations);
2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520
	/* FIXME
	 * This belongs in journal shutdown, but because we have to
	 * allocate osb->journal at the start of ocfs2_initalize_osb(),
	 * we free it here.
	 */
	kfree(osb->journal);
	if (osb->local_alloc_copy)
		kfree(osb->local_alloc_copy);
	kfree(osb->uuid_str);
	ocfs2_put_dlm_debug(osb->osb_dlm_debug);
	memset(osb, 0, sizeof(struct ocfs2_super));

	mlog_exit_void();
}

/* Put OCFS2 into a readonly state, or (if the user specifies it),
 * panic(). We do not support continue-on-error operation. */
static void ocfs2_handle_error(struct super_block *sb)
{
	struct ocfs2_super *osb = OCFS2_SB(sb);

	if (osb->s_mount_opt & OCFS2_MOUNT_ERRORS_PANIC)
		panic("OCFS2: (device %s): panic forced after error\n",
		      sb->s_id);

	ocfs2_set_osb_flag(osb, OCFS2_OSB_ERROR_FS);

	if (sb->s_flags & MS_RDONLY &&
	    (ocfs2_is_soft_readonly(osb) ||
	     ocfs2_is_hard_readonly(osb)))
		return;

	printk(KERN_CRIT "File system is now read-only due to the potential "
	       "of on-disk corruption. Please run fsck.ocfs2 once the file "
	       "system is unmounted.\n");
	sb->s_flags |= MS_RDONLY;
	ocfs2_set_ro_flag(osb, 0);
}

static char error_buf[1024];

void __ocfs2_error(struct super_block *sb,
		   const char *function,
		   const char *fmt, ...)
{
	va_list args;

	va_start(args, fmt);
2521
	vsnprintf(error_buf, sizeof(error_buf), fmt, args);
2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541
	va_end(args);

	/* Not using mlog here because we want to show the actual
	 * function the error came from. */
	printk(KERN_CRIT "OCFS2: ERROR (device %s): %s: %s\n",
	       sb->s_id, function, error_buf);

	ocfs2_handle_error(sb);
}

/* Handle critical errors. This is intentionally more drastic than
 * ocfs2_handle_error, so we only use for things like journal errors,
 * etc. */
void __ocfs2_abort(struct super_block* sb,
		   const char *function,
		   const char *fmt, ...)
{
	va_list args;

	va_start(args, fmt);
2542
	vsnprintf(error_buf, sizeof(error_buf), fmt, args);
2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558
	va_end(args);

	printk(KERN_CRIT "OCFS2: abort (device %s): %s: %s\n",
	       sb->s_id, function, error_buf);

	/* We don't have the cluster support yet to go straight to
	 * hard readonly in here. Until then, we want to keep
	 * ocfs2_abort() so that we can at least mark critical
	 * errors.
	 *
	 * TODO: This should abort the journal and alert other nodes
	 * that our slot needs recovery. */

	/* Force a panic(). This stinks, but it's better than letting
	 * things continue without having a proper hard readonly
	 * here. */
2559 2560
	if (!ocfs2_mount_local(OCFS2_SB(sb)))
		OCFS2_SB(sb)->s_mount_opt |= OCFS2_MOUNT_ERRORS_PANIC;
2561 2562 2563
	ocfs2_handle_error(sb);
}

2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583
/*
 * Void signal blockers, because in-kernel sigprocmask() only fails
 * when SIG_* is wrong.
 */
void ocfs2_block_signals(sigset_t *oldset)
{
	int rc;
	sigset_t blocked;

	sigfillset(&blocked);
	rc = sigprocmask(SIG_BLOCK, &blocked, oldset);
	BUG_ON(rc);
}

void ocfs2_unblock_signals(sigset_t *oldset)
{
	int rc = sigprocmask(SIG_SETMASK, oldset, NULL);
	BUG_ON(rc);
}

2584 2585
module_init(ocfs2_init);
module_exit(ocfs2_exit);