xfs_super.c 52.4 KB
Newer Older
L
Linus Torvalds 已提交
1
/*
2
 * Copyright (c) 2000-2006 Silicon Graphics, Inc.
3
 * All Rights Reserved.
L
Linus Torvalds 已提交
4
 *
5 6
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License as
L
Linus Torvalds 已提交
7 8
 * published by the Free Software Foundation.
 *
9 10 11 12
 * This program is distributed in the hope that it would 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.
L
Linus Torvalds 已提交
13
 *
14 15 16
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write the Free Software Foundation,
 * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
L
Linus Torvalds 已提交
17 18
 */
#include "xfs.h"
19
#include "xfs_bit.h"
L
Linus Torvalds 已提交
20 21
#include "xfs_log.h"
#include "xfs_clnt.h"
22
#include "xfs_inum.h"
L
Linus Torvalds 已提交
23 24
#include "xfs_trans.h"
#include "xfs_sb.h"
25
#include "xfs_ag.h"
L
Linus Torvalds 已提交
26 27 28 29 30 31
#include "xfs_dir2.h"
#include "xfs_alloc.h"
#include "xfs_dmapi.h"
#include "xfs_quota.h"
#include "xfs_mount.h"
#include "xfs_bmap_btree.h"
32
#include "xfs_alloc_btree.h"
L
Linus Torvalds 已提交
33 34
#include "xfs_ialloc_btree.h"
#include "xfs_dir2_sf.h"
35
#include "xfs_attr_sf.h"
L
Linus Torvalds 已提交
36 37
#include "xfs_dinode.h"
#include "xfs_inode.h"
38
#include "xfs_btree.h"
39
#include "xfs_btree_trace.h"
40
#include "xfs_ialloc.h"
L
Linus Torvalds 已提交
41 42 43 44
#include "xfs_bmap.h"
#include "xfs_rtalloc.h"
#include "xfs_error.h"
#include "xfs_itable.h"
C
Christoph Hellwig 已提交
45
#include "xfs_fsops.h"
L
Linus Torvalds 已提交
46 47 48 49 50
#include "xfs_rw.h"
#include "xfs_acl.h"
#include "xfs_attr.h"
#include "xfs_buf_item.h"
#include "xfs_utils.h"
51
#include "xfs_vnodeops.h"
52
#include "xfs_vfsops.h"
L
Linus Torvalds 已提交
53
#include "xfs_version.h"
54
#include "xfs_log_priv.h"
55
#include "xfs_trans_priv.h"
56
#include "xfs_filestream.h"
57 58 59 60 61
#include "xfs_da_btree.h"
#include "xfs_dir2_trace.h"
#include "xfs_extfree_item.h"
#include "xfs_mru_cache.h"
#include "xfs_inode_item.h"
62
#include "xfs_sync.h"
L
Linus Torvalds 已提交
63 64 65 66

#include <linux/namei.h>
#include <linux/init.h>
#include <linux/mount.h>
67
#include <linux/mempool.h>
L
Linus Torvalds 已提交
68
#include <linux/writeback.h>
69
#include <linux/kthread.h>
70
#include <linux/freezer.h>
71
#include <linux/parser.h>
L
Linus Torvalds 已提交
72

73 74 75 76
static struct quotactl_ops xfs_quotactl_operations;
static struct super_operations xfs_super_operations;
static kmem_zone_t *xfs_vnode_zone;
static kmem_zone_t *xfs_ioend_zone;
77
mempool_t *xfs_ioend_pool;
L
Linus Torvalds 已提交
78 79 80

STATIC struct xfs_mount_args *
xfs_args_allocate(
81 82
	struct super_block	*sb,
	int			silent)
L
Linus Torvalds 已提交
83 84 85
{
	struct xfs_mount_args	*args;

86 87 88 89
	args = kzalloc(sizeof(struct xfs_mount_args), GFP_KERNEL);
	if (!args)
		return NULL;

L
Linus Torvalds 已提交
90 91 92 93 94 95 96 97
	args->logbufs = args->logbufsize = -1;
	strncpy(args->fsname, sb->s_id, MAXNAMELEN);

	/* Copy the already-parsed mount(2) flags we're interested in */
	if (sb->s_flags & MS_DIRSYNC)
		args->flags |= XFSMNT_DIRSYNC;
	if (sb->s_flags & MS_SYNCHRONOUS)
		args->flags |= XFSMNT_WSYNC;
98 99
	if (silent)
		args->flags |= XFSMNT_QUIET;
L
Linus Torvalds 已提交
100 101 102 103 104
	args->flags |= XFSMNT_32BITINODES;

	return args;
}

105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152
#define MNTOPT_LOGBUFS	"logbufs"	/* number of XFS log buffers */
#define MNTOPT_LOGBSIZE	"logbsize"	/* size of XFS log buffers */
#define MNTOPT_LOGDEV	"logdev"	/* log device */
#define MNTOPT_RTDEV	"rtdev"		/* realtime I/O device */
#define MNTOPT_BIOSIZE	"biosize"	/* log2 of preferred buffered io size */
#define MNTOPT_WSYNC	"wsync"		/* safe-mode nfs compatible mount */
#define MNTOPT_INO64	"ino64"		/* force inodes into 64-bit range */
#define MNTOPT_NOALIGN	"noalign"	/* turn off stripe alignment */
#define MNTOPT_SWALLOC	"swalloc"	/* turn on stripe width allocation */
#define MNTOPT_SUNIT	"sunit"		/* data volume stripe unit */
#define MNTOPT_SWIDTH	"swidth"	/* data volume stripe width */
#define MNTOPT_NOUUID	"nouuid"	/* ignore filesystem UUID */
#define MNTOPT_MTPT	"mtpt"		/* filesystem mount point */
#define MNTOPT_GRPID	"grpid"		/* group-ID from parent directory */
#define MNTOPT_NOGRPID	"nogrpid"	/* group-ID from current process */
#define MNTOPT_BSDGROUPS    "bsdgroups"    /* group-ID from parent directory */
#define MNTOPT_SYSVGROUPS   "sysvgroups"   /* group-ID from current process */
#define MNTOPT_ALLOCSIZE    "allocsize"    /* preferred allocation size */
#define MNTOPT_NORECOVERY   "norecovery"   /* don't run XFS recovery */
#define MNTOPT_BARRIER	"barrier"	/* use writer barriers for log write and
					 * unwritten extent conversion */
#define MNTOPT_NOBARRIER "nobarrier"	/* .. disable */
#define MNTOPT_OSYNCISOSYNC "osyncisosync" /* o_sync is REALLY o_sync */
#define MNTOPT_64BITINODE   "inode64"	/* inodes can be allocated anywhere */
#define MNTOPT_IKEEP	"ikeep"		/* do not free empty inode clusters */
#define MNTOPT_NOIKEEP	"noikeep"	/* free empty inode clusters */
#define MNTOPT_LARGEIO	   "largeio"	/* report large I/O sizes in stat() */
#define MNTOPT_NOLARGEIO   "nolargeio"	/* do not report large I/O sizes
					 * in stat(). */
#define MNTOPT_ATTR2	"attr2"		/* do use attr2 attribute format */
#define MNTOPT_NOATTR2	"noattr2"	/* do not use attr2 attribute format */
#define MNTOPT_FILESTREAM  "filestreams" /* use filestreams allocator */
#define MNTOPT_QUOTA	"quota"		/* disk quotas (user) */
#define MNTOPT_NOQUOTA	"noquota"	/* no quotas */
#define MNTOPT_USRQUOTA	"usrquota"	/* user quota enabled */
#define MNTOPT_GRPQUOTA	"grpquota"	/* group quota enabled */
#define MNTOPT_PRJQUOTA	"prjquota"	/* project quota enabled */
#define MNTOPT_UQUOTA	"uquota"	/* user quota (IRIX variant) */
#define MNTOPT_GQUOTA	"gquota"	/* group quota (IRIX variant) */
#define MNTOPT_PQUOTA	"pquota"	/* project quota (IRIX variant) */
#define MNTOPT_UQUOTANOENF "uqnoenforce"/* user quota limit enforcement */
#define MNTOPT_GQUOTANOENF "gqnoenforce"/* group quota limit enforcement */
#define MNTOPT_PQUOTANOENF "pqnoenforce"/* project quota limit enforcement */
#define MNTOPT_QUOTANOENF  "qnoenforce"	/* same as uqnoenforce */
#define MNTOPT_DMAPI	"dmapi"		/* DMI enabled (DMAPI / XDSM) */
#define MNTOPT_XDSM	"xdsm"		/* DMI enabled (DMAPI / XDSM) */
#define MNTOPT_DMI	"dmi"		/* DMI enabled (DMAPI / XDSM) */

153 154 155 156 157 158 159 160 161 162
/*
 * Table driven mount option parser.
 *
 * Currently only used for remount, but it will be used for mount
 * in the future, too.
 */
enum {
	Opt_barrier, Opt_nobarrier, Opt_err
};

163
static const match_table_t tokens = {
164 165 166 167 168 169
	{Opt_barrier, "barrier"},
	{Opt_nobarrier, "nobarrier"},
	{Opt_err, NULL}
};


170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202
STATIC unsigned long
suffix_strtoul(char *s, char **endp, unsigned int base)
{
	int	last, shift_left_factor = 0;
	char	*value = s;

	last = strlen(value) - 1;
	if (value[last] == 'K' || value[last] == 'k') {
		shift_left_factor = 10;
		value[last] = '\0';
	}
	if (value[last] == 'M' || value[last] == 'm') {
		shift_left_factor = 20;
		value[last] = '\0';
	}
	if (value[last] == 'G' || value[last] == 'g') {
		shift_left_factor = 30;
		value[last] = '\0';
	}

	return simple_strtoul((const char *)s, endp, base) << shift_left_factor;
}

STATIC int
xfs_parseargs(
	struct xfs_mount	*mp,
	char			*options,
	struct xfs_mount_args	*args,
	int			update)
{
	char			*this_char, *value, *eov;
	int			dsunit, dswidth, vol_dsunit, vol_dswidth;
	int			iosize;
203
	int			dmapi_implies_ikeep = 1;
204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333

	args->flags |= XFSMNT_BARRIER;
	args->flags2 |= XFSMNT2_COMPAT_IOSIZE;

	if (!options)
		goto done;

	iosize = dsunit = dswidth = vol_dsunit = vol_dswidth = 0;

	while ((this_char = strsep(&options, ",")) != NULL) {
		if (!*this_char)
			continue;
		if ((value = strchr(this_char, '=')) != NULL)
			*value++ = 0;

		if (!strcmp(this_char, MNTOPT_LOGBUFS)) {
			if (!value || !*value) {
				cmn_err(CE_WARN,
					"XFS: %s option requires an argument",
					this_char);
				return EINVAL;
			}
			args->logbufs = simple_strtoul(value, &eov, 10);
		} else if (!strcmp(this_char, MNTOPT_LOGBSIZE)) {
			if (!value || !*value) {
				cmn_err(CE_WARN,
					"XFS: %s option requires an argument",
					this_char);
				return EINVAL;
			}
			args->logbufsize = suffix_strtoul(value, &eov, 10);
		} else if (!strcmp(this_char, MNTOPT_LOGDEV)) {
			if (!value || !*value) {
				cmn_err(CE_WARN,
					"XFS: %s option requires an argument",
					this_char);
				return EINVAL;
			}
			strncpy(args->logname, value, MAXNAMELEN);
		} else if (!strcmp(this_char, MNTOPT_MTPT)) {
			if (!value || !*value) {
				cmn_err(CE_WARN,
					"XFS: %s option requires an argument",
					this_char);
				return EINVAL;
			}
			strncpy(args->mtpt, value, MAXNAMELEN);
		} else if (!strcmp(this_char, MNTOPT_RTDEV)) {
			if (!value || !*value) {
				cmn_err(CE_WARN,
					"XFS: %s option requires an argument",
					this_char);
				return EINVAL;
			}
			strncpy(args->rtname, value, MAXNAMELEN);
		} else if (!strcmp(this_char, MNTOPT_BIOSIZE)) {
			if (!value || !*value) {
				cmn_err(CE_WARN,
					"XFS: %s option requires an argument",
					this_char);
				return EINVAL;
			}
			iosize = simple_strtoul(value, &eov, 10);
			args->flags |= XFSMNT_IOSIZE;
			args->iosizelog = (uint8_t) iosize;
		} else if (!strcmp(this_char, MNTOPT_ALLOCSIZE)) {
			if (!value || !*value) {
				cmn_err(CE_WARN,
					"XFS: %s option requires an argument",
					this_char);
				return EINVAL;
			}
			iosize = suffix_strtoul(value, &eov, 10);
			args->flags |= XFSMNT_IOSIZE;
			args->iosizelog = ffs(iosize) - 1;
		} else if (!strcmp(this_char, MNTOPT_GRPID) ||
			   !strcmp(this_char, MNTOPT_BSDGROUPS)) {
			mp->m_flags |= XFS_MOUNT_GRPID;
		} else if (!strcmp(this_char, MNTOPT_NOGRPID) ||
			   !strcmp(this_char, MNTOPT_SYSVGROUPS)) {
			mp->m_flags &= ~XFS_MOUNT_GRPID;
		} else if (!strcmp(this_char, MNTOPT_WSYNC)) {
			args->flags |= XFSMNT_WSYNC;
		} else if (!strcmp(this_char, MNTOPT_OSYNCISOSYNC)) {
			args->flags |= XFSMNT_OSYNCISOSYNC;
		} else if (!strcmp(this_char, MNTOPT_NORECOVERY)) {
			args->flags |= XFSMNT_NORECOVERY;
		} else if (!strcmp(this_char, MNTOPT_INO64)) {
			args->flags |= XFSMNT_INO64;
#if !XFS_BIG_INUMS
			cmn_err(CE_WARN,
				"XFS: %s option not allowed on this system",
				this_char);
			return EINVAL;
#endif
		} else if (!strcmp(this_char, MNTOPT_NOALIGN)) {
			args->flags |= XFSMNT_NOALIGN;
		} else if (!strcmp(this_char, MNTOPT_SWALLOC)) {
			args->flags |= XFSMNT_SWALLOC;
		} else if (!strcmp(this_char, MNTOPT_SUNIT)) {
			if (!value || !*value) {
				cmn_err(CE_WARN,
					"XFS: %s option requires an argument",
					this_char);
				return EINVAL;
			}
			dsunit = simple_strtoul(value, &eov, 10);
		} else if (!strcmp(this_char, MNTOPT_SWIDTH)) {
			if (!value || !*value) {
				cmn_err(CE_WARN,
					"XFS: %s option requires an argument",
					this_char);
				return EINVAL;
			}
			dswidth = simple_strtoul(value, &eov, 10);
		} else if (!strcmp(this_char, MNTOPT_64BITINODE)) {
			args->flags &= ~XFSMNT_32BITINODES;
#if !XFS_BIG_INUMS
			cmn_err(CE_WARN,
				"XFS: %s option not allowed on this system",
				this_char);
			return EINVAL;
#endif
		} else if (!strcmp(this_char, MNTOPT_NOUUID)) {
			args->flags |= XFSMNT_NOUUID;
		} else if (!strcmp(this_char, MNTOPT_BARRIER)) {
			args->flags |= XFSMNT_BARRIER;
		} else if (!strcmp(this_char, MNTOPT_NOBARRIER)) {
			args->flags &= ~XFSMNT_BARRIER;
		} else if (!strcmp(this_char, MNTOPT_IKEEP)) {
334
			args->flags |= XFSMNT_IKEEP;
335
		} else if (!strcmp(this_char, MNTOPT_NOIKEEP)) {
336 337
			dmapi_implies_ikeep = 0;
			args->flags &= ~XFSMNT_IKEEP;
338 339 340 341 342 343 344 345
		} else if (!strcmp(this_char, MNTOPT_LARGEIO)) {
			args->flags2 &= ~XFSMNT2_COMPAT_IOSIZE;
		} else if (!strcmp(this_char, MNTOPT_NOLARGEIO)) {
			args->flags2 |= XFSMNT2_COMPAT_IOSIZE;
		} else if (!strcmp(this_char, MNTOPT_ATTR2)) {
			args->flags |= XFSMNT_ATTR2;
		} else if (!strcmp(this_char, MNTOPT_NOATTR2)) {
			args->flags &= ~XFSMNT_ATTR2;
346
			args->flags |= XFSMNT_NOATTR2;
347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442
		} else if (!strcmp(this_char, MNTOPT_FILESTREAM)) {
			args->flags2 |= XFSMNT2_FILESTREAMS;
		} else if (!strcmp(this_char, MNTOPT_NOQUOTA)) {
			args->flags &= ~(XFSMNT_UQUOTAENF|XFSMNT_UQUOTA);
			args->flags &= ~(XFSMNT_GQUOTAENF|XFSMNT_GQUOTA);
		} else if (!strcmp(this_char, MNTOPT_QUOTA) ||
			   !strcmp(this_char, MNTOPT_UQUOTA) ||
			   !strcmp(this_char, MNTOPT_USRQUOTA)) {
			args->flags |= XFSMNT_UQUOTA | XFSMNT_UQUOTAENF;
		} else if (!strcmp(this_char, MNTOPT_QUOTANOENF) ||
			   !strcmp(this_char, MNTOPT_UQUOTANOENF)) {
			args->flags |= XFSMNT_UQUOTA;
			args->flags &= ~XFSMNT_UQUOTAENF;
		} else if (!strcmp(this_char, MNTOPT_PQUOTA) ||
			   !strcmp(this_char, MNTOPT_PRJQUOTA)) {
			args->flags |= XFSMNT_PQUOTA | XFSMNT_PQUOTAENF;
		} else if (!strcmp(this_char, MNTOPT_PQUOTANOENF)) {
			args->flags |= XFSMNT_PQUOTA;
			args->flags &= ~XFSMNT_PQUOTAENF;
		} else if (!strcmp(this_char, MNTOPT_GQUOTA) ||
			   !strcmp(this_char, MNTOPT_GRPQUOTA)) {
			args->flags |= XFSMNT_GQUOTA | XFSMNT_GQUOTAENF;
		} else if (!strcmp(this_char, MNTOPT_GQUOTANOENF)) {
			args->flags |= XFSMNT_GQUOTA;
			args->flags &= ~XFSMNT_GQUOTAENF;
		} else if (!strcmp(this_char, MNTOPT_DMAPI)) {
			args->flags |= XFSMNT_DMAPI;
		} else if (!strcmp(this_char, MNTOPT_XDSM)) {
			args->flags |= XFSMNT_DMAPI;
		} else if (!strcmp(this_char, MNTOPT_DMI)) {
			args->flags |= XFSMNT_DMAPI;
		} else if (!strcmp(this_char, "ihashsize")) {
			cmn_err(CE_WARN,
	"XFS: ihashsize no longer used, option is deprecated.");
		} else if (!strcmp(this_char, "osyncisdsync")) {
			/* no-op, this is now the default */
			cmn_err(CE_WARN,
	"XFS: osyncisdsync is now the default, option is deprecated.");
		} else if (!strcmp(this_char, "irixsgid")) {
			cmn_err(CE_WARN,
	"XFS: irixsgid is now a sysctl(2) variable, option is deprecated.");
		} else {
			cmn_err(CE_WARN,
				"XFS: unknown mount option [%s].", this_char);
			return EINVAL;
		}
	}

	if (args->flags & XFSMNT_NORECOVERY) {
		if ((mp->m_flags & XFS_MOUNT_RDONLY) == 0) {
			cmn_err(CE_WARN,
				"XFS: no-recovery mounts must be read-only.");
			return EINVAL;
		}
	}

	if ((args->flags & XFSMNT_NOALIGN) && (dsunit || dswidth)) {
		cmn_err(CE_WARN,
	"XFS: sunit and swidth options incompatible with the noalign option");
		return EINVAL;
	}

	if ((args->flags & XFSMNT_GQUOTA) && (args->flags & XFSMNT_PQUOTA)) {
		cmn_err(CE_WARN,
			"XFS: cannot mount with both project and group quota");
		return EINVAL;
	}

	if ((args->flags & XFSMNT_DMAPI) && *args->mtpt == '\0') {
		printk("XFS: %s option needs the mount point option as well\n",
			MNTOPT_DMAPI);
		return EINVAL;
	}

	if ((dsunit && !dswidth) || (!dsunit && dswidth)) {
		cmn_err(CE_WARN,
			"XFS: sunit and swidth must be specified together");
		return EINVAL;
	}

	if (dsunit && (dswidth % dsunit != 0)) {
		cmn_err(CE_WARN,
	"XFS: stripe width (%d) must be a multiple of the stripe unit (%d)",
			dswidth, dsunit);
		return EINVAL;
	}

	/*
	 * Applications using DMI filesystems often expect the
	 * inode generation number to be monotonically increasing.
	 * If we delete inode chunks we break this assumption, so
	 * keep unused inode chunks on disk for DMI filesystems
	 * until we come up with a better solution.
	 * Note that if "ikeep" or "noikeep" mount options are
	 * supplied, then they are honored.
	 */
443 444
	if ((args->flags & XFSMNT_DMAPI) && dmapi_implies_ikeep)
		args->flags |= XFSMNT_IKEEP;
445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478

	if ((args->flags & XFSMNT_NOALIGN) != XFSMNT_NOALIGN) {
		if (dsunit) {
			args->sunit = dsunit;
			args->flags |= XFSMNT_RETERR;
		} else {
			args->sunit = vol_dsunit;
		}
		dswidth ? (args->swidth = dswidth) :
			  (args->swidth = vol_dswidth);
	} else {
		args->sunit = args->swidth = 0;
	}

done:
	if (args->flags & XFSMNT_32BITINODES)
		mp->m_flags |= XFS_MOUNT_SMALL_INUMS;
	if (args->flags2)
		args->flags |= XFSMNT_FLAGS2;
	return 0;
}

struct proc_xfs_info {
	int	flag;
	char	*str;
};

STATIC int
xfs_showargs(
	struct xfs_mount	*mp,
	struct seq_file		*m)
{
	static struct proc_xfs_info xfs_info_set[] = {
		/* the few simple ones we can get from the mount struct */
479
		{ XFS_MOUNT_IKEEP,		"," MNTOPT_IKEEP },
480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551
		{ XFS_MOUNT_WSYNC,		"," MNTOPT_WSYNC },
		{ XFS_MOUNT_INO64,		"," MNTOPT_INO64 },
		{ XFS_MOUNT_NOALIGN,		"," MNTOPT_NOALIGN },
		{ XFS_MOUNT_SWALLOC,		"," MNTOPT_SWALLOC },
		{ XFS_MOUNT_NOUUID,		"," MNTOPT_NOUUID },
		{ XFS_MOUNT_NORECOVERY,		"," MNTOPT_NORECOVERY },
		{ XFS_MOUNT_OSYNCISOSYNC,	"," MNTOPT_OSYNCISOSYNC },
		{ XFS_MOUNT_ATTR2,		"," MNTOPT_ATTR2 },
		{ XFS_MOUNT_FILESTREAMS,	"," MNTOPT_FILESTREAM },
		{ XFS_MOUNT_DMAPI,		"," MNTOPT_DMAPI },
		{ XFS_MOUNT_GRPID,		"," MNTOPT_GRPID },
		{ 0, NULL }
	};
	static struct proc_xfs_info xfs_info_unset[] = {
		/* the few simple ones we can get from the mount struct */
		{ XFS_MOUNT_COMPAT_IOSIZE,	"," MNTOPT_LARGEIO },
		{ XFS_MOUNT_BARRIER,		"," MNTOPT_NOBARRIER },
		{ XFS_MOUNT_SMALL_INUMS,	"," MNTOPT_64BITINODE },
		{ 0, NULL }
	};
	struct proc_xfs_info	*xfs_infop;

	for (xfs_infop = xfs_info_set; xfs_infop->flag; xfs_infop++) {
		if (mp->m_flags & xfs_infop->flag)
			seq_puts(m, xfs_infop->str);
	}
	for (xfs_infop = xfs_info_unset; xfs_infop->flag; xfs_infop++) {
		if (!(mp->m_flags & xfs_infop->flag))
			seq_puts(m, xfs_infop->str);
	}

	if (mp->m_flags & XFS_MOUNT_DFLT_IOSIZE)
		seq_printf(m, "," MNTOPT_ALLOCSIZE "=%dk",
				(int)(1 << mp->m_writeio_log) >> 10);

	if (mp->m_logbufs > 0)
		seq_printf(m, "," MNTOPT_LOGBUFS "=%d", mp->m_logbufs);
	if (mp->m_logbsize > 0)
		seq_printf(m, "," MNTOPT_LOGBSIZE "=%dk", mp->m_logbsize >> 10);

	if (mp->m_logname)
		seq_printf(m, "," MNTOPT_LOGDEV "=%s", mp->m_logname);
	if (mp->m_rtname)
		seq_printf(m, "," MNTOPT_RTDEV "=%s", mp->m_rtname);

	if (mp->m_dalign > 0)
		seq_printf(m, "," MNTOPT_SUNIT "=%d",
				(int)XFS_FSB_TO_BB(mp, mp->m_dalign));
	if (mp->m_swidth > 0)
		seq_printf(m, "," MNTOPT_SWIDTH "=%d",
				(int)XFS_FSB_TO_BB(mp, mp->m_swidth));

	if (mp->m_qflags & (XFS_UQUOTA_ACCT|XFS_UQUOTA_ENFD))
		seq_puts(m, "," MNTOPT_USRQUOTA);
	else if (mp->m_qflags & XFS_UQUOTA_ACCT)
		seq_puts(m, "," MNTOPT_UQUOTANOENF);

	if (mp->m_qflags & (XFS_PQUOTA_ACCT|XFS_OQUOTA_ENFD))
		seq_puts(m, "," MNTOPT_PRJQUOTA);
	else if (mp->m_qflags & XFS_PQUOTA_ACCT)
		seq_puts(m, "," MNTOPT_PQUOTANOENF);

	if (mp->m_qflags & (XFS_GQUOTA_ACCT|XFS_OQUOTA_ENFD))
		seq_puts(m, "," MNTOPT_GRPQUOTA);
	else if (mp->m_qflags & XFS_GQUOTA_ACCT)
		seq_puts(m, "," MNTOPT_GQUOTANOENF);

	if (!(mp->m_qflags & XFS_ALL_QUOTA_ACCT))
		seq_puts(m, "," MNTOPT_NOQUOTA);

	return 0;
}
L
Linus Torvalds 已提交
552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593
__uint64_t
xfs_max_file_offset(
	unsigned int		blockshift)
{
	unsigned int		pagefactor = 1;
	unsigned int		bitshift = BITS_PER_LONG - 1;

	/* Figure out maximum filesize, on Linux this can depend on
	 * the filesystem blocksize (on 32 bit platforms).
	 * __block_prepare_write does this in an [unsigned] long...
	 *      page->index << (PAGE_CACHE_SHIFT - bbits)
	 * So, for page sized blocks (4K on 32 bit platforms),
	 * this wraps at around 8Tb (hence MAX_LFS_FILESIZE which is
	 *      (((u64)PAGE_CACHE_SIZE << (BITS_PER_LONG-1))-1)
	 * but for smaller blocksizes it is less (bbits = log2 bsize).
	 * Note1: get_block_t takes a long (implicit cast from above)
	 * Note2: The Large Block Device (LBD and HAVE_SECTOR_T) patch
	 * can optionally convert the [unsigned] long from above into
	 * an [unsigned] long long.
	 */

#if BITS_PER_LONG == 32
# if defined(CONFIG_LBD)
	ASSERT(sizeof(sector_t) == 8);
	pagefactor = PAGE_CACHE_SIZE;
	bitshift = BITS_PER_LONG;
# else
	pagefactor = PAGE_CACHE_SIZE >> (PAGE_CACHE_SHIFT - blockshift);
# endif
#endif

	return (((__uint64_t)pagefactor) << bitshift) - 1;
}

int
xfs_blkdev_get(
	xfs_mount_t		*mp,
	const char		*name,
	struct block_device	**bdevp)
{
	int			error = 0;

594
	*bdevp = open_bdev_exclusive(name, FMODE_READ|FMODE_WRITE, mp);
L
Linus Torvalds 已提交
595 596 597 598 599 600 601 602 603 604 605 606 607
	if (IS_ERR(*bdevp)) {
		error = PTR_ERR(*bdevp);
		printk("XFS: Invalid device [%s], error=%d\n", name, error);
	}

	return -error;
}

void
xfs_blkdev_put(
	struct block_device	*bdev)
{
	if (bdev)
608
		close_bdev_exclusive(bdev, FMODE_READ|FMODE_WRITE);
L
Linus Torvalds 已提交
609 610
}

611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653
/*
 * Try to write out the superblock using barriers.
 */
STATIC int
xfs_barrier_test(
	xfs_mount_t	*mp)
{
	xfs_buf_t	*sbp = xfs_getsb(mp, 0);
	int		error;

	XFS_BUF_UNDONE(sbp);
	XFS_BUF_UNREAD(sbp);
	XFS_BUF_UNDELAYWRITE(sbp);
	XFS_BUF_WRITE(sbp);
	XFS_BUF_UNASYNC(sbp);
	XFS_BUF_ORDERED(sbp);

	xfsbdstrat(mp, sbp);
	error = xfs_iowait(sbp);

	/*
	 * Clear all the flags we set and possible error state in the
	 * buffer.  We only did the write to try out whether barriers
	 * worked and shouldn't leave any traces in the superblock
	 * buffer.
	 */
	XFS_BUF_DONE(sbp);
	XFS_BUF_ERROR(sbp, 0);
	XFS_BUF_UNORDERED(sbp);

	xfs_buf_relse(sbp);
	return error;
}

void
xfs_mountfs_check_barriers(xfs_mount_t *mp)
{
	int error;

	if (mp->m_logdev_targp != mp->m_ddev_targp) {
		xfs_fs_cmn_err(CE_NOTE, mp,
		  "Disabling barriers, not supported with external log device");
		mp->m_flags &= ~XFS_MOUNT_BARRIER;
654
		return;
655 656
	}

657 658 659 660 661 662 663
	if (xfs_readonly_buftarg(mp->m_ddev_targp)) {
		xfs_fs_cmn_err(CE_NOTE, mp,
		  "Disabling barriers, underlying device is readonly");
		mp->m_flags &= ~XFS_MOUNT_BARRIER;
		return;
	}

664 665 666 667 668
	error = xfs_barrier_test(mp);
	if (error) {
		xfs_fs_cmn_err(CE_NOTE, mp,
		  "Disabling barriers, trial barrier write failed");
		mp->m_flags &= ~XFS_MOUNT_BARRIER;
669
		return;
670 671 672 673 674 675 676
	}
}

void
xfs_blkdev_issue_flush(
	xfs_buftarg_t		*buftarg)
{
677
	blkdev_issue_flush(buftarg->bt_bdev, NULL);
678
}
L
Linus Torvalds 已提交
679

680 681 682 683 684
STATIC void
xfs_close_devices(
	struct xfs_mount	*mp)
{
	if (mp->m_logdev_targp && mp->m_logdev_targp != mp->m_ddev_targp) {
685
		struct block_device *logdev = mp->m_logdev_targp->bt_bdev;
686
		xfs_free_buftarg(mp->m_logdev_targp);
687
		xfs_blkdev_put(logdev);
688 689
	}
	if (mp->m_rtdev_targp) {
690
		struct block_device *rtdev = mp->m_rtdev_targp->bt_bdev;
691
		xfs_free_buftarg(mp->m_rtdev_targp);
692
		xfs_blkdev_put(rtdev);
693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776
	}
	xfs_free_buftarg(mp->m_ddev_targp);
}

/*
 * The file system configurations are:
 *	(1) device (partition) with data and internal log
 *	(2) logical volume with data and log subvolumes.
 *	(3) logical volume with data, log, and realtime subvolumes.
 *
 * We only have to handle opening the log and realtime volumes here if
 * they are present.  The data subvolume has already been opened by
 * get_sb_bdev() and is stored in sb->s_bdev.
 */
STATIC int
xfs_open_devices(
	struct xfs_mount	*mp,
	struct xfs_mount_args	*args)
{
	struct block_device	*ddev = mp->m_super->s_bdev;
	struct block_device	*logdev = NULL, *rtdev = NULL;
	int			error;

	/*
	 * Open real time and log devices - order is important.
	 */
	if (args->logname[0]) {
		error = xfs_blkdev_get(mp, args->logname, &logdev);
		if (error)
			goto out;
	}

	if (args->rtname[0]) {
		error = xfs_blkdev_get(mp, args->rtname, &rtdev);
		if (error)
			goto out_close_logdev;

		if (rtdev == ddev || rtdev == logdev) {
			cmn_err(CE_WARN,
	"XFS: Cannot mount filesystem with identical rtdev and ddev/logdev.");
			error = EINVAL;
			goto out_close_rtdev;
		}
	}

	/*
	 * Setup xfs_mount buffer target pointers
	 */
	error = ENOMEM;
	mp->m_ddev_targp = xfs_alloc_buftarg(ddev, 0);
	if (!mp->m_ddev_targp)
		goto out_close_rtdev;

	if (rtdev) {
		mp->m_rtdev_targp = xfs_alloc_buftarg(rtdev, 1);
		if (!mp->m_rtdev_targp)
			goto out_free_ddev_targ;
	}

	if (logdev && logdev != ddev) {
		mp->m_logdev_targp = xfs_alloc_buftarg(logdev, 1);
		if (!mp->m_logdev_targp)
			goto out_free_rtdev_targ;
	} else {
		mp->m_logdev_targp = mp->m_ddev_targp;
	}

	return 0;

 out_free_rtdev_targ:
	if (mp->m_rtdev_targp)
		xfs_free_buftarg(mp->m_rtdev_targp);
 out_free_ddev_targ:
	xfs_free_buftarg(mp->m_ddev_targp);
 out_close_rtdev:
	if (rtdev)
		xfs_blkdev_put(rtdev);
 out_close_logdev:
	if (logdev && logdev != ddev)
		xfs_blkdev_put(logdev);
 out:
	return error;
}

777 778 779 780 781 782 783 784
/*
 * Setup xfs_mount buffer target pointers based on superblock
 */
STATIC int
xfs_setup_devices(
	struct xfs_mount	*mp)
{
	int			error;
785

786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811
	error = xfs_setsize_buftarg(mp->m_ddev_targp, mp->m_sb.sb_blocksize,
				    mp->m_sb.sb_sectsize);
	if (error)
		return error;

	if (mp->m_logdev_targp && mp->m_logdev_targp != mp->m_ddev_targp) {
		unsigned int	log_sector_size = BBSIZE;

		if (xfs_sb_version_hassector(&mp->m_sb))
			log_sector_size = mp->m_sb.sb_logsectsize;
		error = xfs_setsize_buftarg(mp->m_logdev_targp,
					    mp->m_sb.sb_blocksize,
					    log_sector_size);
		if (error)
			return error;
	}
	if (mp->m_rtdev_targp) {
		error = xfs_setsize_buftarg(mp->m_rtdev_targp,
					    mp->m_sb.sb_blocksize,
					    mp->m_sb.sb_sectsize);
		if (error)
			return error;
	}

	return 0;
}
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
/*
 * XFS AIL push thread support
 */
void
xfsaild_wakeup(
	xfs_mount_t		*mp,
	xfs_lsn_t		threshold_lsn)
{
	mp->m_ail.xa_target = threshold_lsn;
	wake_up_process(mp->m_ail.xa_task);
}

int
xfsaild(
	void	*data)
{
	xfs_mount_t	*mp = (xfs_mount_t *)data;
	xfs_lsn_t	last_pushed_lsn = 0;
	long		tout = 0;

	while (!kthread_should_stop()) {
		if (tout)
			schedule_timeout_interruptible(msecs_to_jiffies(tout));
		tout = 1000;

		/* swsusp */
		try_to_freeze();

		ASSERT(mp->m_log);
		if (XFS_FORCED_SHUTDOWN(mp))
			continue;

		tout = xfsaild_push(mp, &last_pushed_lsn);
	}

	return 0;
}	/* xfsaild */

int
xfsaild_start(
	xfs_mount_t	*mp)
{
	mp->m_ail.xa_target = 0;
	mp->m_ail.xa_task = kthread_run(xfsaild, mp, "xfsaild");
	if (IS_ERR(mp->m_ail.xa_task))
		return -PTR_ERR(mp->m_ail.xa_task);
	return 0;
}

void
xfsaild_stop(
	xfs_mount_t	*mp)
{
	kthread_stop(mp->m_ail.xa_task);
}



L
Linus Torvalds 已提交
871
STATIC struct inode *
872
xfs_fs_alloc_inode(
L
Linus Torvalds 已提交
873 874
	struct super_block	*sb)
{
875
	return kmem_zone_alloc(xfs_vnode_zone, KM_SLEEP);
L
Linus Torvalds 已提交
876 877 878
}

STATIC void
879
xfs_fs_destroy_inode(
L
Linus Torvalds 已提交
880 881
	struct inode		*inode)
{
C
Christoph Hellwig 已提交
882
	kmem_zone_free(xfs_vnode_zone, inode);
L
Linus Torvalds 已提交
883 884 885
}

STATIC void
886
xfs_fs_inode_init_once(
887
	void			*vnode)
L
Linus Torvalds 已提交
888
{
C
Christoph Hellwig 已提交
889
	inode_init_once((struct inode *)vnode);
L
Linus Torvalds 已提交
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

/*
 * Slab object creation initialisation for the XFS inode.
 * This covers only the idempotent fields in the XFS inode;
 * all other fields need to be initialised on allocation
 * from the slab. This avoids the need to repeatedly intialise
 * fields in the xfs inode that left in the initialise state
 * when freeing the inode.
 */
void
xfs_inode_init_once(
	void			*inode)
{
	struct xfs_inode	*ip = inode;

	memset(ip, 0, sizeof(struct xfs_inode));
	atomic_set(&ip->i_iocount, 0);
	atomic_set(&ip->i_pincount, 0);
	spin_lock_init(&ip->i_flags_lock);
	INIT_LIST_HEAD(&ip->i_reclaim);
	init_waitqueue_head(&ip->i_ipin_wait);
	/*
	 * Because we want to use a counting completion, complete
	 * the flush completion once to allow a single access to
	 * the flush completion without blocking.
	 */
	init_completion(&ip->i_flush);
	complete(&ip->i_flush);

	mrlock_init(&ip->i_lock, MRLOCK_ALLOW_EQUAL_PRI|MRLOCK_BARRIER,
		     "xfsino", ip->i_ino);
	mrlock_init(&ip->i_iolock, MRLOCK_BARRIER, "xfsio", ip->i_ino);
}

L
Linus Torvalds 已提交
926 927 928 929
/*
 * Attempt to flush the inode, this will actually fail
 * if the inode is pinned, but we dirty the inode again
 * at the point when it is unpinned after a log write,
930
 * since this is when the inode itself becomes flushable.
L
Linus Torvalds 已提交
931 932
 */
STATIC int
933
xfs_fs_write_inode(
L
Linus Torvalds 已提交
934 935 936
	struct inode		*inode,
	int			sync)
{
937 938
	int			error = 0;
	int			flags = 0;
L
Linus Torvalds 已提交
939

940
	xfs_itrace_entry(XFS_I(inode));
941 942 943
	if (sync) {
		filemap_fdatawait(inode->i_mapping);
		flags |= FLUSH_SYNC;
L
Linus Torvalds 已提交
944
	}
945
	error = xfs_inode_flush(XFS_I(inode), flags);
946 947 948 949 950 951
	/*
	 * if we failed to write out the inode then mark
	 * it dirty again so we'll try again later.
	 */
	if (error)
		mark_inode_dirty_sync(inode);
952

L
Linus Torvalds 已提交
953 954 955 956
	return -error;
}

STATIC void
957
xfs_fs_clear_inode(
L
Linus Torvalds 已提交
958 959
	struct inode		*inode)
{
960
	xfs_inode_t		*ip = XFS_I(inode);
961

962
	/*
963
	 * ip can be null when xfs_iget_core calls xfs_idestroy if we
964 965
	 * find an inode with di_mode == 0 but without IGET_CREATE set.
	 */
966
	if (ip) {
967
		xfs_itrace_entry(ip);
968 969 970 971 972 973 974 975
		XFS_STATS_INC(vn_rele);
		XFS_STATS_INC(vn_remove);
		XFS_STATS_INC(vn_reclaim);
		XFS_STATS_DEC(vn_active);

		xfs_inactive(ip);
		xfs_iflags_clear(ip, XFS_IMODIFIED);
		if (xfs_reclaim(ip))
976
			panic("%s: cannot reclaim 0x%p\n", __func__, inode);
977
	}
978

979
	ASSERT(XFS_I(inode) == NULL);
980
}
L
Linus Torvalds 已提交
981

982 983 984 985 986 987 988 989 990
STATIC void
xfs_free_fsname(
	struct xfs_mount	*mp)
{
	kfree(mp->m_fsname);
	kfree(mp->m_rtname);
	kfree(mp->m_logname);
}

L
Linus Torvalds 已提交
991
STATIC void
992
xfs_fs_put_super(
L
Linus Torvalds 已提交
993 994
	struct super_block	*sb)
{
995
	struct xfs_mount	*mp = XFS_M(sb);
996 997
	struct xfs_inode	*rip = mp->m_rootip;
	int			unmount_event_flags = 0;
L
Linus Torvalds 已提交
998 999
	int			error;

1000
	xfs_syncd_stop(mp);
1001
	xfs_sync_inodes(mp, SYNC_ATTR|SYNC_DELWRI);
1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033

#ifdef HAVE_DMAPI
	if (mp->m_flags & XFS_MOUNT_DMAPI) {
		unmount_event_flags =
			(mp->m_dmevmask & (1 << DM_EVENT_UNMOUNT)) ?
				0 : DM_FLAGS_UNWANTED;
		/*
		 * Ignore error from dmapi here, first unmount is not allowed
		 * to fail anyway, and second we wouldn't want to fail a
		 * unmount because of dmapi.
		 */
		XFS_SEND_PREUNMOUNT(mp, rip, DM_RIGHT_NULL, rip, DM_RIGHT_NULL,
				NULL, NULL, 0, 0, unmount_event_flags);
	}
#endif

	/*
	 * Blow away any referenced inode in the filestreams cache.
	 * This can and will cause log traffic as inodes go inactive
	 * here.
	 */
	xfs_filestream_unmount(mp);

	XFS_bflush(mp->m_ddev_targp);
	error = xfs_unmount_flush(mp, 0);
	WARN_ON(error);

	if (mp->m_flags & XFS_MOUNT_DMAPI) {
		XFS_SEND_UNMOUNT(mp, rip, DM_RIGHT_NULL, 0, 0,
				unmount_event_flags);
	}

1034
	xfs_unmountfs(mp);
1035
	xfs_freesb(mp);
C
Christoph Hellwig 已提交
1036
	xfs_icsb_destroy_counters(mp);
1037
	xfs_close_devices(mp);
1038 1039
	xfs_qmops_put(mp);
	xfs_dmops_put(mp);
1040
	xfs_free_fsname(mp);
C
Christoph Hellwig 已提交
1041
	kfree(mp);
L
Linus Torvalds 已提交
1042 1043 1044
}

STATIC void
1045
xfs_fs_write_super(
L
Linus Torvalds 已提交
1046 1047
	struct super_block	*sb)
{
1048
	if (!(sb->s_flags & MS_RDONLY))
1049
		xfs_sync_fsdata(XFS_M(sb), 0);
L
Linus Torvalds 已提交
1050 1051 1052 1053
	sb->s_dirt = 0;
}

STATIC int
1054
xfs_fs_sync_super(
L
Linus Torvalds 已提交
1055 1056 1057
	struct super_block	*sb,
	int			wait)
{
1058
	struct xfs_mount	*mp = XFS_M(sb);
1059
	int			error;
L
Linus Torvalds 已提交
1060

1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072
	/*
	 * Treat a sync operation like a freeze.  This is to work
	 * around a race in sync_inodes() which works in two phases
	 * - an asynchronous flush, which can write out an inode
	 * without waiting for file size updates to complete, and a
	 * synchronous flush, which wont do anything because the
	 * async flush removed the inode's dirty flag.  Also
	 * sync_inodes() will not see any files that just have
	 * outstanding transactions to be flushed because we don't
	 * dirty the Linux inode until after the transaction I/O
	 * completes.
	 */
1073 1074 1075 1076
	if (wait || unlikely(sb->s_frozen == SB_FREEZE_WRITE))
		error = xfs_quiesce_data(mp);
	else
		error = xfs_sync_fsdata(mp, 0);
L
Linus Torvalds 已提交
1077 1078 1079
	sb->s_dirt = 0;

	if (unlikely(laptop_mode)) {
1080
		int	prev_sync_seq = mp->m_sync_seq;
L
Linus Torvalds 已提交
1081 1082 1083 1084 1085 1086

		/*
		 * The disk must be active because we're syncing.
		 * We schedule xfssyncd now (now that the disk is
		 * active) instead of later (when it might not be).
		 */
1087
		wake_up_process(mp->m_sync_task);
L
Linus Torvalds 已提交
1088 1089 1090 1091 1092 1093
		/*
		 * We have to wait for the sync iteration to complete.
		 * If we don't, the disk activity caused by the sync
		 * will come after the sync is completed, and that
		 * triggers another sync from laptop mode.
		 */
1094 1095
		wait_event(mp->m_wait_single_sync_task,
				mp->m_sync_seq != prev_sync_seq);
L
Linus Torvalds 已提交
1096 1097 1098 1099 1100 1101
	}

	return -error;
}

STATIC int
1102
xfs_fs_statfs(
1103
	struct dentry		*dentry,
L
Linus Torvalds 已提交
1104 1105
	struct kstatfs		*statp)
{
C
Christoph Hellwig 已提交
1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117
	struct xfs_mount	*mp = XFS_M(dentry->d_sb);
	xfs_sb_t		*sbp = &mp->m_sb;
	__uint64_t		fakeinos, id;
	xfs_extlen_t		lsize;

	statp->f_type = XFS_SB_MAGIC;
	statp->f_namelen = MAXNAMELEN - 1;

	id = huge_encode_dev(mp->m_ddev_targp->bt_dev);
	statp->f_fsid.val[0] = (u32)id;
	statp->f_fsid.val[1] = (u32)(id >> 32);

1118
	xfs_icsb_sync_counters(mp, XFS_ICSB_LAZY_COUNT);
C
Christoph Hellwig 已提交
1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143

	spin_lock(&mp->m_sb_lock);
	statp->f_bsize = sbp->sb_blocksize;
	lsize = sbp->sb_logstart ? sbp->sb_logblocks : 0;
	statp->f_blocks = sbp->sb_dblocks - lsize;
	statp->f_bfree = statp->f_bavail =
				sbp->sb_fdblocks - XFS_ALLOC_SET_ASIDE(mp);
	fakeinos = statp->f_bfree << sbp->sb_inopblog;
#if XFS_BIG_INUMS
	fakeinos += mp->m_inoadd;
#endif
	statp->f_files =
	    MIN(sbp->sb_icount + fakeinos, (__uint64_t)XFS_MAXINUMBER);
	if (mp->m_maxicount)
#if XFS_BIG_INUMS
		if (!mp->m_inoadd)
#endif
			statp->f_files = min_t(typeof(statp->f_files),
						statp->f_files,
						mp->m_maxicount);
	statp->f_ffree = statp->f_files - (sbp->sb_icount - sbp->sb_ifree);
	spin_unlock(&mp->m_sb_lock);

	XFS_QM_DQSTATVFS(XFS_I(dentry->d_inode), statp);
	return 0;
L
Linus Torvalds 已提交
1144 1145 1146
}

STATIC int
1147
xfs_fs_remount(
L
Linus Torvalds 已提交
1148 1149 1150 1151
	struct super_block	*sb,
	int			*flags,
	char			*options)
{
1152
	struct xfs_mount	*mp = XFS_M(sb);
1153 1154
	substring_t		args[MAX_OPT_ARGS];
	char			*p;
L
Linus Torvalds 已提交
1155

1156 1157
	while ((p = strsep(&options, ",")) != NULL) {
		int token;
1158

1159 1160
		if (!*p)
			continue;
1161

1162 1163 1164
		token = match_token(p, tokens, args);
		switch (token) {
		case Opt_barrier:
1165
			mp->m_flags |= XFS_MOUNT_BARRIER;
1166 1167 1168 1169 1170 1171 1172 1173 1174 1175

			/*
			 * Test if barriers are actually working if we can,
			 * else delay this check until the filesystem is
			 * marked writeable.
			 */
			if (!(mp->m_flags & XFS_MOUNT_RDONLY))
				xfs_mountfs_check_barriers(mp);
			break;
		case Opt_nobarrier:
1176
			mp->m_flags &= ~XFS_MOUNT_BARRIER;
1177 1178
			break;
		default:
1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195
			/*
			 * Logically we would return an error here to prevent
			 * users from believing they might have changed
			 * mount options using remount which can't be changed.
			 *
			 * But unfortunately mount(8) adds all options from
			 * mtab and fstab to the mount arguments in some cases
			 * so we can't blindly reject options, but have to
			 * check for each specified option if it actually
			 * differs from the currently set option and only
			 * reject it if that's the case.
			 *
			 * Until that is implemented we return success for
			 * every remount request, and silently ignore all
			 * options that we can't actually change.
			 */
#if 0
1196 1197 1198
			printk(KERN_INFO
	"XFS: mount option \"%s\" not supported for remount\n", p);
			return -EINVAL;
1199
#else
1200
			break;
1201
#endif
1202
		}
1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213
	}

	/* rw/ro -> rw */
	if ((mp->m_flags & XFS_MOUNT_RDONLY) && !(*flags & MS_RDONLY)) {
		mp->m_flags &= ~XFS_MOUNT_RDONLY;
		if (mp->m_flags & XFS_MOUNT_BARRIER)
			xfs_mountfs_check_barriers(mp);
	}

	/* rw -> ro */
	if (!(mp->m_flags & XFS_MOUNT_RDONLY) && (*flags & MS_RDONLY)) {
1214
		xfs_quiesce_data(mp);
D
David Chinner 已提交
1215
		xfs_quiesce_attr(mp);
1216 1217 1218
		mp->m_flags |= XFS_MOUNT_RDONLY;
	}

1219
	return 0;
L
Linus Torvalds 已提交
1220 1221
}

C
Christoph Hellwig 已提交
1222 1223
/*
 * Second stage of a freeze. The data is already frozen so we only
D
David Chinner 已提交
1224
 * need to take care of the metadata. Once that's done write a dummy
C
Christoph Hellwig 已提交
1225 1226
 * record to dirty the log in case of a crash while frozen.
 */
L
Linus Torvalds 已提交
1227
STATIC void
1228
xfs_fs_lockfs(
L
Linus Torvalds 已提交
1229 1230
	struct super_block	*sb)
{
C
Christoph Hellwig 已提交
1231 1232
	struct xfs_mount	*mp = XFS_M(sb);

D
David Chinner 已提交
1233
	xfs_quiesce_attr(mp);
C
Christoph Hellwig 已提交
1234
	xfs_fs_log_dummy(mp);
L
Linus Torvalds 已提交
1235 1236 1237
}

STATIC int
1238
xfs_fs_show_options(
L
Linus Torvalds 已提交
1239 1240 1241
	struct seq_file		*m,
	struct vfsmount		*mnt)
{
1242
	return -xfs_showargs(XFS_M(mnt->mnt_sb), m);
L
Linus Torvalds 已提交
1243 1244
}

1245
STATIC int
1246
xfs_fs_quotasync(
1247 1248 1249
	struct super_block	*sb,
	int			type)
{
1250
	return -XFS_QM_QUOTACTL(XFS_M(sb), Q_XQUOTASYNC, 0, NULL);
1251 1252
}

L
Linus Torvalds 已提交
1253
STATIC int
1254
xfs_fs_getxstate(
L
Linus Torvalds 已提交
1255 1256 1257
	struct super_block	*sb,
	struct fs_quota_stat	*fqs)
{
1258
	return -XFS_QM_QUOTACTL(XFS_M(sb), Q_XGETQSTAT, 0, (caddr_t)fqs);
L
Linus Torvalds 已提交
1259 1260 1261
}

STATIC int
1262
xfs_fs_setxstate(
L
Linus Torvalds 已提交
1263 1264 1265 1266
	struct super_block	*sb,
	unsigned int		flags,
	int			op)
{
1267
	return -XFS_QM_QUOTACTL(XFS_M(sb), op, 0, (caddr_t)&flags);
L
Linus Torvalds 已提交
1268 1269 1270
}

STATIC int
1271
xfs_fs_getxquota(
L
Linus Torvalds 已提交
1272 1273 1274 1275 1276
	struct super_block	*sb,
	int			type,
	qid_t			id,
	struct fs_disk_quota	*fdq)
{
1277
	return -XFS_QM_QUOTACTL(XFS_M(sb),
1278 1279 1280
				 (type == USRQUOTA) ? Q_XGETQUOTA :
				  ((type == GRPQUOTA) ? Q_XGETGQUOTA :
				   Q_XGETPQUOTA), id, (caddr_t)fdq);
L
Linus Torvalds 已提交
1281 1282 1283
}

STATIC int
1284
xfs_fs_setxquota(
L
Linus Torvalds 已提交
1285 1286 1287 1288 1289
	struct super_block	*sb,
	int			type,
	qid_t			id,
	struct fs_disk_quota	*fdq)
{
1290
	return -XFS_QM_QUOTACTL(XFS_M(sb),
1291 1292 1293
				 (type == USRQUOTA) ? Q_XSETQLIM :
				  ((type == GRPQUOTA) ? Q_XSETGQLIM :
				   Q_XSETPQLIM), id, (caddr_t)fdq);
L
Linus Torvalds 已提交
1294 1295
}

1296 1297 1298 1299 1300 1301 1302 1303 1304
/*
 * This function fills in xfs_mount_t fields based on mount args.
 * Note: the superblock has _not_ yet been read in.
 */
STATIC int
xfs_start_flags(
	struct xfs_mount_args	*ap,
	struct xfs_mount	*mp)
{
1305 1306
	int			error;

1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338
	/* Values are in BBs */
	if ((ap->flags & XFSMNT_NOALIGN) != XFSMNT_NOALIGN) {
		/*
		 * At this point the superblock has not been read
		 * in, therefore we do not know the block size.
		 * Before the mount call ends we will convert
		 * these to FSBs.
		 */
		mp->m_dalign = ap->sunit;
		mp->m_swidth = ap->swidth;
	}

	if (ap->logbufs != -1 &&
	    ap->logbufs != 0 &&
	    (ap->logbufs < XLOG_MIN_ICLOGS ||
	     ap->logbufs > XLOG_MAX_ICLOGS)) {
		cmn_err(CE_WARN,
			"XFS: invalid logbufs value: %d [not %d-%d]",
			ap->logbufs, XLOG_MIN_ICLOGS, XLOG_MAX_ICLOGS);
		return XFS_ERROR(EINVAL);
	}
	mp->m_logbufs = ap->logbufs;
	if (ap->logbufsize != -1 &&
	    ap->logbufsize !=  0 &&
	    (ap->logbufsize < XLOG_MIN_RECORD_BSIZE ||
	     ap->logbufsize > XLOG_MAX_RECORD_BSIZE ||
	     !is_power_of_2(ap->logbufsize))) {
		cmn_err(CE_WARN,
	"XFS: invalid logbufsize: %d [not 16k,32k,64k,128k or 256k]",
			ap->logbufsize);
		return XFS_ERROR(EINVAL);
	}
1339 1340 1341

	error = ENOMEM;

1342 1343
	mp->m_logbsize = ap->logbufsize;
	mp->m_fsname_len = strlen(ap->fsname) + 1;
1344 1345 1346 1347 1348

	mp->m_fsname = kstrdup(ap->fsname, GFP_KERNEL);
	if (!mp->m_fsname)
		goto out;

1349
	if (ap->rtname[0]) {
1350 1351 1352 1353
		mp->m_rtname = kstrdup(ap->rtname, GFP_KERNEL);
		if (!mp->m_rtname)
			goto out_free_fsname;

1354
	}
1355

1356
	if (ap->logname[0]) {
1357 1358 1359
		mp->m_logname = kstrdup(ap->logname, GFP_KERNEL);
		if (!mp->m_logname)
			goto out_free_rtname;
1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431
	}

	if (ap->flags & XFSMNT_WSYNC)
		mp->m_flags |= XFS_MOUNT_WSYNC;
#if XFS_BIG_INUMS
	if (ap->flags & XFSMNT_INO64) {
		mp->m_flags |= XFS_MOUNT_INO64;
		mp->m_inoadd = XFS_INO64_OFFSET;
	}
#endif
	if (ap->flags & XFSMNT_RETERR)
		mp->m_flags |= XFS_MOUNT_RETERR;
	if (ap->flags & XFSMNT_NOALIGN)
		mp->m_flags |= XFS_MOUNT_NOALIGN;
	if (ap->flags & XFSMNT_SWALLOC)
		mp->m_flags |= XFS_MOUNT_SWALLOC;
	if (ap->flags & XFSMNT_OSYNCISOSYNC)
		mp->m_flags |= XFS_MOUNT_OSYNCISOSYNC;
	if (ap->flags & XFSMNT_32BITINODES)
		mp->m_flags |= XFS_MOUNT_32BITINODES;

	if (ap->flags & XFSMNT_IOSIZE) {
		if (ap->iosizelog > XFS_MAX_IO_LOG ||
		    ap->iosizelog < XFS_MIN_IO_LOG) {
			cmn_err(CE_WARN,
		"XFS: invalid log iosize: %d [not %d-%d]",
				ap->iosizelog, XFS_MIN_IO_LOG,
				XFS_MAX_IO_LOG);
			return XFS_ERROR(EINVAL);
		}

		mp->m_flags |= XFS_MOUNT_DFLT_IOSIZE;
		mp->m_readio_log = mp->m_writeio_log = ap->iosizelog;
	}

	if (ap->flags & XFSMNT_IKEEP)
		mp->m_flags |= XFS_MOUNT_IKEEP;
	if (ap->flags & XFSMNT_DIRSYNC)
		mp->m_flags |= XFS_MOUNT_DIRSYNC;
	if (ap->flags & XFSMNT_ATTR2)
		mp->m_flags |= XFS_MOUNT_ATTR2;
	if (ap->flags & XFSMNT_NOATTR2)
		mp->m_flags |= XFS_MOUNT_NOATTR2;

	if (ap->flags2 & XFSMNT2_COMPAT_IOSIZE)
		mp->m_flags |= XFS_MOUNT_COMPAT_IOSIZE;

	/*
	 * no recovery flag requires a read-only mount
	 */
	if (ap->flags & XFSMNT_NORECOVERY) {
		if (!(mp->m_flags & XFS_MOUNT_RDONLY)) {
			cmn_err(CE_WARN,
	"XFS: tried to mount a FS read-write without recovery!");
			return XFS_ERROR(EINVAL);
		}
		mp->m_flags |= XFS_MOUNT_NORECOVERY;
	}

	if (ap->flags & XFSMNT_NOUUID)
		mp->m_flags |= XFS_MOUNT_NOUUID;
	if (ap->flags & XFSMNT_BARRIER)
		mp->m_flags |= XFS_MOUNT_BARRIER;
	else
		mp->m_flags &= ~XFS_MOUNT_BARRIER;

	if (ap->flags2 & XFSMNT2_FILESTREAMS)
		mp->m_flags |= XFS_MOUNT_FILESTREAMS;

	if (ap->flags & XFSMNT_DMAPI)
		mp->m_flags |= XFS_MOUNT_DMAPI;
	return 0;
1432 1433 1434 1435 1436 1437 1438 1439


 out_free_rtname:
	kfree(mp->m_rtname);
 out_free_fsname:
	kfree(mp->m_fsname);
 out:
	return error;
1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 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 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534
}

/*
 * This function fills in xfs_mount_t fields based on mount args.
 * Note: the superblock _has_ now been read in.
 */
STATIC int
xfs_finish_flags(
	struct xfs_mount_args	*ap,
	struct xfs_mount	*mp)
{
	int			ronly = (mp->m_flags & XFS_MOUNT_RDONLY);

	/* Fail a mount where the logbuf is smaller then the log stripe */
	if (xfs_sb_version_haslogv2(&mp->m_sb)) {
		if ((ap->logbufsize <= 0) &&
		    (mp->m_sb.sb_logsunit > XLOG_BIG_RECORD_BSIZE)) {
			mp->m_logbsize = mp->m_sb.sb_logsunit;
		} else if (ap->logbufsize > 0 &&
			   ap->logbufsize < mp->m_sb.sb_logsunit) {
			cmn_err(CE_WARN,
	"XFS: logbuf size must be greater than or equal to log stripe size");
			return XFS_ERROR(EINVAL);
		}
	} else {
		/* Fail a mount if the logbuf is larger than 32K */
		if (ap->logbufsize > XLOG_BIG_RECORD_BSIZE) {
			cmn_err(CE_WARN,
	"XFS: logbuf size for version 1 logs must be 16K or 32K");
			return XFS_ERROR(EINVAL);
		}
	}

	/*
	 * mkfs'ed attr2 will turn on attr2 mount unless explicitly
	 * told by noattr2 to turn it off
	 */
	if (xfs_sb_version_hasattr2(&mp->m_sb) &&
	    !(ap->flags & XFSMNT_NOATTR2))
		mp->m_flags |= XFS_MOUNT_ATTR2;

	/*
	 * prohibit r/w mounts of read-only filesystems
	 */
	if ((mp->m_sb.sb_flags & XFS_SBF_READONLY) && !ronly) {
		cmn_err(CE_WARN,
	"XFS: cannot mount a read-only filesystem as read-write");
		return XFS_ERROR(EROFS);
	}

	/*
	 * check for shared mount.
	 */
	if (ap->flags & XFSMNT_SHARED) {
		if (!xfs_sb_version_hasshared(&mp->m_sb))
			return XFS_ERROR(EINVAL);

		/*
		 * For IRIX 6.5, shared mounts must have the shared
		 * version bit set, have the persistent readonly
		 * field set, must be version 0 and can only be mounted
		 * read-only.
		 */
		if (!ronly || !(mp->m_sb.sb_flags & XFS_SBF_READONLY) ||
		     (mp->m_sb.sb_shared_vn != 0))
			return XFS_ERROR(EINVAL);

		mp->m_flags |= XFS_MOUNT_SHARED;

		/*
		 * Shared XFS V0 can't deal with DMI.  Return EINVAL.
		 */
		if (mp->m_sb.sb_shared_vn == 0 && (ap->flags & XFSMNT_DMAPI))
			return XFS_ERROR(EINVAL);
	}

	if (ap->flags & XFSMNT_UQUOTA) {
		mp->m_qflags |= (XFS_UQUOTA_ACCT | XFS_UQUOTA_ACTIVE);
		if (ap->flags & XFSMNT_UQUOTAENF)
			mp->m_qflags |= XFS_UQUOTA_ENFD;
	}

	if (ap->flags & XFSMNT_GQUOTA) {
		mp->m_qflags |= (XFS_GQUOTA_ACCT | XFS_GQUOTA_ACTIVE);
		if (ap->flags & XFSMNT_GQUOTAENF)
			mp->m_qflags |= XFS_OQUOTA_ENFD;
	} else if (ap->flags & XFSMNT_PQUOTA) {
		mp->m_qflags |= (XFS_PQUOTA_ACCT | XFS_PQUOTA_ACTIVE);
		if (ap->flags & XFSMNT_PQUOTAENF)
			mp->m_qflags |= XFS_OQUOTA_ENFD;
	}

	return 0;
}

L
Linus Torvalds 已提交
1535
STATIC int
1536
xfs_fs_fill_super(
L
Linus Torvalds 已提交
1537 1538 1539 1540
	struct super_block	*sb,
	void			*data,
	int			silent)
{
1541
	struct inode		*root;
1542
	struct xfs_mount	*mp = NULL;
1543
	struct xfs_mount_args	*args;
C
Christoph Hellwig 已提交
1544
	int			flags = 0, error = ENOMEM;
L
Linus Torvalds 已提交
1545

1546 1547 1548 1549
	args = xfs_args_allocate(sb, silent);
	if (!args)
		return -ENOMEM;

C
Christoph Hellwig 已提交
1550 1551 1552
	mp = kzalloc(sizeof(struct xfs_mount), GFP_KERNEL);
	if (!mp)
		goto out_free_args;
L
Linus Torvalds 已提交
1553

C
Christoph Hellwig 已提交
1554 1555 1556 1557
	spin_lock_init(&mp->m_sb_lock);
	mutex_init(&mp->m_ilock);
	mutex_init(&mp->m_growlock);
	atomic_set(&mp->m_active_trans, 0);
1558 1559 1560 1561
	INIT_LIST_HEAD(&mp->m_sync_list);
	spin_lock_init(&mp->m_sync_lock);
	init_waitqueue_head(&mp->m_wait_single_sync_task);

C
Christoph Hellwig 已提交
1562 1563
	mp->m_super = sb;
	sb->s_fs_info = mp;
L
Linus Torvalds 已提交
1564

1565 1566 1567
	if (sb->s_flags & MS_RDONLY)
		mp->m_flags |= XFS_MOUNT_RDONLY;

1568 1569
	error = xfs_parseargs(mp, (char *)data, args, 0);
	if (error)
C
Christoph Hellwig 已提交
1570
		goto out_free_mp;
L
Linus Torvalds 已提交
1571 1572

	sb_min_blocksize(sb, BBSIZE);
1573
	sb->s_xattr = xfs_xattr_handlers;
1574 1575 1576
	sb->s_export_op = &xfs_export_operations;
	sb->s_qcop = &xfs_quotactl_operations;
	sb->s_op = &xfs_super_operations;
L
Linus Torvalds 已提交
1577

1578 1579
	error = xfs_dmops_get(mp, args);
	if (error)
C
Christoph Hellwig 已提交
1580
		goto out_free_mp;
1581
	error = xfs_qmops_get(mp, args);
1582
	if (error)
1583
		goto out_put_dmops;
L
Linus Torvalds 已提交
1584

1585 1586 1587
	if (args->flags & XFSMNT_QUIET)
		flags |= XFS_MFSI_QUIET;

1588 1589 1590
	error = xfs_open_devices(mp, args);
	if (error)
		goto out_put_qmops;
1591

C
Christoph Hellwig 已提交
1592 1593 1594
	if (xfs_icsb_init_counters(mp))
		mp->m_flags |= XFS_MOUNT_NO_PERCPU_SB;

1595 1596 1597 1598 1599
	/*
	 * Setup flags based on mount(2) options and then the superblock
	 */
	error = xfs_start_flags(args, mp);
	if (error)
1600
		goto out_free_fsname;
1601 1602
	error = xfs_readsb(mp, flags);
	if (error)
1603
		goto out_free_fsname;
1604 1605
	error = xfs_finish_flags(args, mp);
	if (error)
1606
		goto out_free_sb;
1607

1608
	error = xfs_setup_devices(mp);
1609
	if (error)
1610
		goto out_free_sb;
1611 1612 1613 1614 1615 1616

	if (mp->m_flags & XFS_MOUNT_BARRIER)
		xfs_mountfs_check_barriers(mp);

	error = xfs_filestream_mount(mp);
	if (error)
1617
		goto out_free_sb;
1618

C
Christoph Hellwig 已提交
1619
	error = xfs_mountfs(mp);
1620
	if (error)
1621
		goto out_filestream_unmount;
1622 1623 1624

	XFS_SEND_MOUNT(mp, DM_RIGHT_NULL, args->mtpt, args->fsname);

L
Linus Torvalds 已提交
1625
	sb->s_dirt = 1;
C
Christoph Hellwig 已提交
1626 1627 1628
	sb->s_magic = XFS_SB_MAGIC;
	sb->s_blocksize = mp->m_sb.sb_blocksize;
	sb->s_blocksize_bits = ffs(sb->s_blocksize) - 1;
L
Linus Torvalds 已提交
1629 1630 1631 1632
	sb->s_maxbytes = xfs_max_file_offset(sb->s_blocksize_bits);
	sb->s_time_gran = 1;
	set_posix_acl_flag(sb);

1633
	root = igrab(VFS_I(mp->m_rootip));
1634
	if (!root) {
C
Christoph Hellwig 已提交
1635
		error = ENOENT;
L
Linus Torvalds 已提交
1636
		goto fail_unmount;
C
Christoph Hellwig 已提交
1637
	}
1638 1639
	if (is_bad_inode(root)) {
		error = EINVAL;
L
Linus Torvalds 已提交
1640 1641
		goto fail_vnrele;
	}
1642 1643 1644
	sb->s_root = d_alloc_root(root);
	if (!sb->s_root) {
		error = ENOMEM;
L
Linus Torvalds 已提交
1645 1646
		goto fail_vnrele;
	}
1647

1648 1649
	error = xfs_syncd_init(mp);
	if (error)
L
Linus Torvalds 已提交
1650
		goto fail_vnrele;
1651

1652
	xfs_itrace_exit(XFS_I(sb->s_root->d_inode));
L
Linus Torvalds 已提交
1653

1654
	kfree(args);
L
Linus Torvalds 已提交
1655 1656
	return 0;

1657 1658
 out_filestream_unmount:
	xfs_filestream_unmount(mp);
1659 1660
 out_free_sb:
	xfs_freesb(mp);
1661 1662
 out_free_fsname:
	xfs_free_fsname(mp);
C
Christoph Hellwig 已提交
1663
	xfs_icsb_destroy_counters(mp);
1664 1665
	xfs_close_devices(mp);
 out_put_qmops:
1666
	xfs_qmops_put(mp);
1667
 out_put_dmops:
1668
	xfs_dmops_put(mp);
C
Christoph Hellwig 已提交
1669 1670 1671 1672 1673
 out_free_mp:
	kfree(mp);
 out_free_args:
	kfree(args);
	return -error;
1674 1675

 fail_vnrele:
L
Linus Torvalds 已提交
1676 1677 1678 1679
	if (sb->s_root) {
		dput(sb->s_root);
		sb->s_root = NULL;
	} else {
1680
		iput(root);
L
Linus Torvalds 已提交
1681 1682
	}

1683
 fail_unmount:
1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694
	/*
	 * Blow away any referenced inode in the filestreams cache.
	 * This can and will cause log traffic as inodes go inactive
	 * here.
	 */
	xfs_filestream_unmount(mp);

	XFS_bflush(mp->m_ddev_targp);
	error = xfs_unmount_flush(mp, 0);
	WARN_ON(error);

1695
	xfs_unmountfs(mp);
1696
	goto out_free_sb;
L
Linus Torvalds 已提交
1697 1698
}

1699
STATIC int
1700
xfs_fs_get_sb(
L
Linus Torvalds 已提交
1701 1702 1703
	struct file_system_type	*fs_type,
	int			flags,
	const char		*dev_name,
1704 1705
	void			*data,
	struct vfsmount		*mnt)
L
Linus Torvalds 已提交
1706
{
1707 1708
	return get_sb_bdev(fs_type, flags, dev_name, data, xfs_fs_fill_super,
			   mnt);
1709 1710
}

1711
static struct super_operations xfs_super_operations = {
1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722
	.alloc_inode		= xfs_fs_alloc_inode,
	.destroy_inode		= xfs_fs_destroy_inode,
	.write_inode		= xfs_fs_write_inode,
	.clear_inode		= xfs_fs_clear_inode,
	.put_super		= xfs_fs_put_super,
	.write_super		= xfs_fs_write_super,
	.sync_fs		= xfs_fs_sync_super,
	.write_super_lockfs	= xfs_fs_lockfs,
	.statfs			= xfs_fs_statfs,
	.remount_fs		= xfs_fs_remount,
	.show_options		= xfs_fs_show_options,
L
Linus Torvalds 已提交
1723 1724
};

1725
static struct quotactl_ops xfs_quotactl_operations = {
1726 1727 1728 1729 1730
	.quota_sync		= xfs_fs_quotasync,
	.get_xstate		= xfs_fs_getxstate,
	.set_xstate		= xfs_fs_setxstate,
	.get_xquota		= xfs_fs_getxquota,
	.set_xquota		= xfs_fs_setxquota,
L
Linus Torvalds 已提交
1731 1732
};

A
Andrew Morton 已提交
1733
static struct file_system_type xfs_fs_type = {
L
Linus Torvalds 已提交
1734 1735
	.owner			= THIS_MODULE,
	.name			= "xfs",
1736
	.get_sb			= xfs_fs_get_sb,
L
Linus Torvalds 已提交
1737 1738 1739 1740
	.kill_sb		= kill_block_super,
	.fs_flags		= FS_REQUIRES_DEV,
};

1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753
STATIC int __init
xfs_alloc_trace_bufs(void)
{
#ifdef XFS_ALLOC_TRACE
	xfs_alloc_trace_buf = ktrace_alloc(XFS_ALLOC_TRACE_SIZE, KM_MAYFAIL);
	if (!xfs_alloc_trace_buf)
		goto out;
#endif
#ifdef XFS_BMAP_TRACE
	xfs_bmap_trace_buf = ktrace_alloc(XFS_BMAP_TRACE_SIZE, KM_MAYFAIL);
	if (!xfs_bmap_trace_buf)
		goto out_free_alloc_trace;
#endif
1754 1755 1756 1757 1758 1759 1760 1761 1762 1763
#ifdef XFS_BTREE_TRACE
	xfs_allocbt_trace_buf = ktrace_alloc(XFS_ALLOCBT_TRACE_SIZE,
					     KM_MAYFAIL);
	if (!xfs_allocbt_trace_buf)
		goto out_free_bmap_trace;

	xfs_inobt_trace_buf = ktrace_alloc(XFS_INOBT_TRACE_SIZE, KM_MAYFAIL);
	if (!xfs_inobt_trace_buf)
		goto out_free_allocbt_trace;

1764 1765
	xfs_bmbt_trace_buf = ktrace_alloc(XFS_BMBT_TRACE_SIZE, KM_MAYFAIL);
	if (!xfs_bmbt_trace_buf)
1766
		goto out_free_inobt_trace;
1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787
#endif
#ifdef XFS_ATTR_TRACE
	xfs_attr_trace_buf = ktrace_alloc(XFS_ATTR_TRACE_SIZE, KM_MAYFAIL);
	if (!xfs_attr_trace_buf)
		goto out_free_bmbt_trace;
#endif
#ifdef XFS_DIR2_TRACE
	xfs_dir2_trace_buf = ktrace_alloc(XFS_DIR2_GTRACE_SIZE, KM_MAYFAIL);
	if (!xfs_dir2_trace_buf)
		goto out_free_attr_trace;
#endif

	return 0;

#ifdef XFS_DIR2_TRACE
 out_free_attr_trace:
#endif
#ifdef XFS_ATTR_TRACE
	ktrace_free(xfs_attr_trace_buf);
 out_free_bmbt_trace:
#endif
1788
#ifdef XFS_BTREE_TRACE
1789
	ktrace_free(xfs_bmbt_trace_buf);
1790 1791 1792 1793
 out_free_inobt_trace:
	ktrace_free(xfs_inobt_trace_buf);
 out_free_allocbt_trace:
	ktrace_free(xfs_allocbt_trace_buf);
1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815
 out_free_bmap_trace:
#endif
#ifdef XFS_BMAP_TRACE
	ktrace_free(xfs_bmap_trace_buf);
 out_free_alloc_trace:
#endif
#ifdef XFS_ALLOC_TRACE
	ktrace_free(xfs_alloc_trace_buf);
 out:
#endif
	return -ENOMEM;
}

STATIC void
xfs_free_trace_bufs(void)
{
#ifdef XFS_DIR2_TRACE
	ktrace_free(xfs_dir2_trace_buf);
#endif
#ifdef XFS_ATTR_TRACE
	ktrace_free(xfs_attr_trace_buf);
#endif
1816
#ifdef XFS_BTREE_TRACE
1817
	ktrace_free(xfs_bmbt_trace_buf);
1818 1819
	ktrace_free(xfs_inobt_trace_buf);
	ktrace_free(xfs_allocbt_trace_buf);
1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831
#endif
#ifdef XFS_BMAP_TRACE
	ktrace_free(xfs_bmap_trace_buf);
#endif
#ifdef XFS_ALLOC_TRACE
	ktrace_free(xfs_alloc_trace_buf);
#endif
}

STATIC int __init
xfs_init_zones(void)
{
C
Christoph Hellwig 已提交
1832
	xfs_vnode_zone = kmem_zone_init_flags(sizeof(struct inode), "xfs_vnode",
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 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904
					KM_ZONE_HWALIGN | KM_ZONE_RECLAIM |
					KM_ZONE_SPREAD,
					xfs_fs_inode_init_once);
	if (!xfs_vnode_zone)
		goto out;

	xfs_ioend_zone = kmem_zone_init(sizeof(xfs_ioend_t), "xfs_ioend");
	if (!xfs_ioend_zone)
		goto out_destroy_vnode_zone;

	xfs_ioend_pool = mempool_create_slab_pool(4 * MAX_BUF_PER_PAGE,
						  xfs_ioend_zone);
	if (!xfs_ioend_pool)
		goto out_destroy_ioend_zone;

	xfs_log_ticket_zone = kmem_zone_init(sizeof(xlog_ticket_t),
						"xfs_log_ticket");
	if (!xfs_log_ticket_zone)
		goto out_destroy_ioend_pool;

	xfs_bmap_free_item_zone = kmem_zone_init(sizeof(xfs_bmap_free_item_t),
						"xfs_bmap_free_item");
	if (!xfs_bmap_free_item_zone)
		goto out_destroy_log_ticket_zone;
	xfs_btree_cur_zone = kmem_zone_init(sizeof(xfs_btree_cur_t),
						"xfs_btree_cur");
	if (!xfs_btree_cur_zone)
		goto out_destroy_bmap_free_item_zone;

	xfs_da_state_zone = kmem_zone_init(sizeof(xfs_da_state_t),
						"xfs_da_state");
	if (!xfs_da_state_zone)
		goto out_destroy_btree_cur_zone;

	xfs_dabuf_zone = kmem_zone_init(sizeof(xfs_dabuf_t), "xfs_dabuf");
	if (!xfs_dabuf_zone)
		goto out_destroy_da_state_zone;

	xfs_ifork_zone = kmem_zone_init(sizeof(xfs_ifork_t), "xfs_ifork");
	if (!xfs_ifork_zone)
		goto out_destroy_dabuf_zone;

	xfs_trans_zone = kmem_zone_init(sizeof(xfs_trans_t), "xfs_trans");
	if (!xfs_trans_zone)
		goto out_destroy_ifork_zone;

	/*
	 * The size of the zone allocated buf log item is the maximum
	 * size possible under XFS.  This wastes a little bit of memory,
	 * but it is much faster.
	 */
	xfs_buf_item_zone = kmem_zone_init((sizeof(xfs_buf_log_item_t) +
				(((XFS_MAX_BLOCKSIZE / XFS_BLI_CHUNK) /
				  NBWORD) * sizeof(int))), "xfs_buf_item");
	if (!xfs_buf_item_zone)
		goto out_destroy_trans_zone;

	xfs_efd_zone = kmem_zone_init((sizeof(xfs_efd_log_item_t) +
			((XFS_EFD_MAX_FAST_EXTENTS - 1) *
				 sizeof(xfs_extent_t))), "xfs_efd_item");
	if (!xfs_efd_zone)
		goto out_destroy_buf_item_zone;

	xfs_efi_zone = kmem_zone_init((sizeof(xfs_efi_log_item_t) +
			((XFS_EFI_MAX_FAST_EXTENTS - 1) *
				sizeof(xfs_extent_t))), "xfs_efi_item");
	if (!xfs_efi_zone)
		goto out_destroy_efd_zone;

	xfs_inode_zone =
		kmem_zone_init_flags(sizeof(xfs_inode_t), "xfs_inode",
					KM_ZONE_HWALIGN | KM_ZONE_RECLAIM |
1905
					KM_ZONE_SPREAD, xfs_inode_init_once);
1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981
	if (!xfs_inode_zone)
		goto out_destroy_efi_zone;

	xfs_ili_zone =
		kmem_zone_init_flags(sizeof(xfs_inode_log_item_t), "xfs_ili",
					KM_ZONE_SPREAD, NULL);
	if (!xfs_ili_zone)
		goto out_destroy_inode_zone;

#ifdef CONFIG_XFS_POSIX_ACL
	xfs_acl_zone = kmem_zone_init(sizeof(xfs_acl_t), "xfs_acl");
	if (!xfs_acl_zone)
		goto out_destroy_ili_zone;
#endif

	return 0;

#ifdef CONFIG_XFS_POSIX_ACL
 out_destroy_ili_zone:
#endif
	kmem_zone_destroy(xfs_ili_zone);
 out_destroy_inode_zone:
	kmem_zone_destroy(xfs_inode_zone);
 out_destroy_efi_zone:
	kmem_zone_destroy(xfs_efi_zone);
 out_destroy_efd_zone:
	kmem_zone_destroy(xfs_efd_zone);
 out_destroy_buf_item_zone:
	kmem_zone_destroy(xfs_buf_item_zone);
 out_destroy_trans_zone:
	kmem_zone_destroy(xfs_trans_zone);
 out_destroy_ifork_zone:
	kmem_zone_destroy(xfs_ifork_zone);
 out_destroy_dabuf_zone:
	kmem_zone_destroy(xfs_dabuf_zone);
 out_destroy_da_state_zone:
	kmem_zone_destroy(xfs_da_state_zone);
 out_destroy_btree_cur_zone:
	kmem_zone_destroy(xfs_btree_cur_zone);
 out_destroy_bmap_free_item_zone:
	kmem_zone_destroy(xfs_bmap_free_item_zone);
 out_destroy_log_ticket_zone:
	kmem_zone_destroy(xfs_log_ticket_zone);
 out_destroy_ioend_pool:
	mempool_destroy(xfs_ioend_pool);
 out_destroy_ioend_zone:
	kmem_zone_destroy(xfs_ioend_zone);
 out_destroy_vnode_zone:
	kmem_zone_destroy(xfs_vnode_zone);
 out:
	return -ENOMEM;
}

STATIC void
xfs_destroy_zones(void)
{
#ifdef CONFIG_XFS_POSIX_ACL
	kmem_zone_destroy(xfs_acl_zone);
#endif
	kmem_zone_destroy(xfs_ili_zone);
	kmem_zone_destroy(xfs_inode_zone);
	kmem_zone_destroy(xfs_efi_zone);
	kmem_zone_destroy(xfs_efd_zone);
	kmem_zone_destroy(xfs_buf_item_zone);
	kmem_zone_destroy(xfs_trans_zone);
	kmem_zone_destroy(xfs_ifork_zone);
	kmem_zone_destroy(xfs_dabuf_zone);
	kmem_zone_destroy(xfs_da_state_zone);
	kmem_zone_destroy(xfs_btree_cur_zone);
	kmem_zone_destroy(xfs_bmap_free_item_zone);
	kmem_zone_destroy(xfs_log_ticket_zone);
	mempool_destroy(xfs_ioend_pool);
	kmem_zone_destroy(xfs_ioend_zone);
	kmem_zone_destroy(xfs_vnode_zone);

}
L
Linus Torvalds 已提交
1982 1983

STATIC int __init
1984
init_xfs_fs(void)
L
Linus Torvalds 已提交
1985 1986 1987 1988 1989 1990 1991 1992
{
	int			error;
	static char		message[] __initdata = KERN_INFO \
		XFS_VERSION_STRING " with " XFS_BUILD_OPTIONS " enabled\n";

	printk(message);

	ktrace_init(64);
1993 1994
	vn_init();
	xfs_dir_startup();
L
Linus Torvalds 已提交
1995

1996
	error = xfs_init_zones();
1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
	if (error)
		goto out;

	error = xfs_alloc_trace_bufs();
	if (error)
		goto out_destroy_zones;

	error = xfs_mru_cache_init();
	if (error)
		goto out_free_trace_buffers;

	error = xfs_filestream_init();
	if (error)
		goto out_mru_cache_uninit;
L
Linus Torvalds 已提交
2011

2012
	error = xfs_buf_init();
2013 2014 2015 2016 2017 2018 2019 2020 2021 2022
	if (error)
		goto out_filestream_uninit;

	error = xfs_init_procfs();
	if (error)
		goto out_buf_terminate;

	error = xfs_sysctl_register();
	if (error)
		goto out_cleanup_procfs;
L
Linus Torvalds 已提交
2023 2024 2025 2026 2027

	vfs_initquota();

	error = register_filesystem(&xfs_fs_type);
	if (error)
2028
		goto out_sysctl_unregister;
L
Linus Torvalds 已提交
2029 2030
	return 0;

2031 2032 2033 2034 2035
 out_sysctl_unregister:
	xfs_sysctl_unregister();
 out_cleanup_procfs:
	xfs_cleanup_procfs();
 out_buf_terminate:
2036
	xfs_buf_terminate();
2037 2038 2039 2040 2041 2042 2043
 out_filestream_uninit:
	xfs_filestream_uninit();
 out_mru_cache_uninit:
	xfs_mru_cache_uninit();
 out_free_trace_buffers:
	xfs_free_trace_bufs();
 out_destroy_zones:
2044
	xfs_destroy_zones();
2045
 out:
L
Linus Torvalds 已提交
2046 2047 2048 2049
	return error;
}

STATIC void __exit
2050
exit_xfs_fs(void)
L
Linus Torvalds 已提交
2051 2052 2053
{
	vfs_exitquota();
	unregister_filesystem(&xfs_fs_type);
2054 2055
	xfs_sysctl_unregister();
	xfs_cleanup_procfs();
2056
	xfs_buf_terminate();
2057 2058 2059
	xfs_filestream_uninit();
	xfs_mru_cache_uninit();
	xfs_free_trace_bufs();
2060
	xfs_destroy_zones();
L
Linus Torvalds 已提交
2061 2062 2063 2064 2065 2066 2067 2068 2069
	ktrace_uninit();
}

module_init(init_xfs_fs);
module_exit(exit_xfs_fs);

MODULE_AUTHOR("Silicon Graphics, Inc.");
MODULE_DESCRIPTION(XFS_VERSION_STRING " with " XFS_BUILD_OPTIONS " enabled");
MODULE_LICENSE("GPL");