xfs_super.c 54.3 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 39
#include "xfs_btree.h"
#include "xfs_ialloc.h"
L
Linus Torvalds 已提交
40 41 42 43
#include "xfs_bmap.h"
#include "xfs_rtalloc.h"
#include "xfs_error.h"
#include "xfs_itable.h"
C
Christoph Hellwig 已提交
44
#include "xfs_fsops.h"
L
Linus Torvalds 已提交
45 46 47 48 49
#include "xfs_rw.h"
#include "xfs_acl.h"
#include "xfs_attr.h"
#include "xfs_buf_item.h"
#include "xfs_utils.h"
50
#include "xfs_vnodeops.h"
51
#include "xfs_vfsops.h"
L
Linus Torvalds 已提交
52
#include "xfs_version.h"
53
#include "xfs_log_priv.h"
54
#include "xfs_trans_priv.h"
55
#include "xfs_filestream.h"
56 57 58 59 60
#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"
L
Linus Torvalds 已提交
61 62 63 64

#include <linux/namei.h>
#include <linux/init.h>
#include <linux/mount.h>
65
#include <linux/mempool.h>
L
Linus Torvalds 已提交
66
#include <linux/writeback.h>
67
#include <linux/kthread.h>
68
#include <linux/freezer.h>
69
#include <linux/parser.h>
L
Linus Torvalds 已提交
70

71 72 73 74
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;
75
mempool_t *xfs_ioend_pool;
L
Linus Torvalds 已提交
76 77 78

STATIC struct xfs_mount_args *
xfs_args_allocate(
79 80
	struct super_block	*sb,
	int			silent)
L
Linus Torvalds 已提交
81 82 83
{
	struct xfs_mount_args	*args;

84 85 86 87
	args = kzalloc(sizeof(struct xfs_mount_args), GFP_KERNEL);
	if (!args)
		return NULL;

L
Linus Torvalds 已提交
88 89 90 91 92 93 94 95
	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;
96 97
	if (silent)
		args->flags |= XFSMNT_QUIET;
L
Linus Torvalds 已提交
98 99 100 101 102
	args->flags |= XFSMNT_32BITINODES;

	return args;
}

103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150
#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) */

151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167
/*
 * 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
};

static match_table_t tokens = {
	{Opt_barrier, "barrier"},
	{Opt_nobarrier, "nobarrier"},
	{Opt_err, NULL}
};


168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200
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;
201
	int			dmapi_implies_ikeep = 1;
202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331

	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)) {
332
			args->flags |= XFSMNT_IKEEP;
333
		} else if (!strcmp(this_char, MNTOPT_NOIKEEP)) {
334 335
			dmapi_implies_ikeep = 0;
			args->flags &= ~XFSMNT_IKEEP;
336 337 338 339 340 341 342 343
		} 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;
344
			args->flags |= XFSMNT_NOATTR2;
345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440
		} 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.
	 */
441 442
	if ((args->flags & XFSMNT_DMAPI) && dmapi_implies_ikeep)
		args->flags |= XFSMNT_IKEEP;
443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476

	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 */
477
		{ XFS_MOUNT_IKEEP,		"," MNTOPT_IKEEP },
478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549
		{ 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
550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608
__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;

	*bdevp = open_bdev_excl(name, 0, mp);
	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)
		close_bdev_excl(bdev);
}

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
/*
 * 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;
652
		return;
653 654
	}

655 656 657 658 659 660 661
	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;
	}

662 663 664 665 666
	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;
667
		return;
668 669 670 671 672 673 674
	}
}

void
xfs_blkdev_issue_flush(
	xfs_buftarg_t		*buftarg)
{
675
	blkdev_issue_flush(buftarg->bt_bdev, NULL);
676
}
L
Linus Torvalds 已提交
677

678 679 680 681 682
STATIC void
xfs_close_devices(
	struct xfs_mount	*mp)
{
	if (mp->m_logdev_targp && mp->m_logdev_targp != mp->m_ddev_targp) {
683
		struct block_device *logdev = mp->m_logdev_targp->bt_bdev;
684
		xfs_free_buftarg(mp->m_logdev_targp);
685
		xfs_blkdev_put(logdev);
686 687
	}
	if (mp->m_rtdev_targp) {
688
		struct block_device *rtdev = mp->m_rtdev_targp->bt_bdev;
689
		xfs_free_buftarg(mp->m_rtdev_targp);
690
		xfs_blkdev_put(rtdev);
691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774
	}
	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;
}

775 776 777 778 779 780 781 782
/*
 * Setup xfs_mount buffer target pointers based on superblock
 */
STATIC int
xfs_setup_devices(
	struct xfs_mount	*mp)
{
	int			error;
783

784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809
	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;
}
810

811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868
/*
 * 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
869
STATIC struct inode *
870
xfs_fs_alloc_inode(
L
Linus Torvalds 已提交
871 872
	struct super_block	*sb)
{
873
	return kmem_zone_alloc(xfs_vnode_zone, KM_SLEEP);
L
Linus Torvalds 已提交
874 875 876
}

STATIC void
877
xfs_fs_destroy_inode(
L
Linus Torvalds 已提交
878 879
	struct inode		*inode)
{
C
Christoph Hellwig 已提交
880
	kmem_zone_free(xfs_vnode_zone, inode);
L
Linus Torvalds 已提交
881 882 883
}

STATIC void
884
xfs_fs_inode_init_once(
885
	void			*vnode)
L
Linus Torvalds 已提交
886
{
C
Christoph Hellwig 已提交
887
	inode_init_once((struct inode *)vnode);
L
Linus Torvalds 已提交
888 889 890 891 892 893
}

/*
 * 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,
894
 * since this is when the inode itself becomes flushable.
L
Linus Torvalds 已提交
895 896
 */
STATIC int
897
xfs_fs_write_inode(
L
Linus Torvalds 已提交
898 899 900
	struct inode		*inode,
	int			sync)
{
901 902
	int			error = 0;
	int			flags = 0;
L
Linus Torvalds 已提交
903

904
	xfs_itrace_entry(XFS_I(inode));
905 906 907
	if (sync) {
		filemap_fdatawait(inode->i_mapping);
		flags |= FLUSH_SYNC;
L
Linus Torvalds 已提交
908
	}
909
	error = xfs_inode_flush(XFS_I(inode), flags);
910 911 912 913 914 915
	/*
	 * 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);
916

L
Linus Torvalds 已提交
917 918 919 920
	return -error;
}

STATIC void
921
xfs_fs_clear_inode(
L
Linus Torvalds 已提交
922 923
	struct inode		*inode)
{
924
	xfs_inode_t		*ip = XFS_I(inode);
925

926
	/*
927
	 * ip can be null when xfs_iget_core calls xfs_idestroy if we
928 929
	 * find an inode with di_mode == 0 but without IGET_CREATE set.
	 */
930
	if (ip) {
931
		xfs_itrace_entry(ip);
932 933 934 935 936 937 938 939
		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))
940
			panic("%s: cannot reclaim 0x%p\n", __func__, inode);
941
	}
942

943
	ASSERT(XFS_I(inode) == NULL);
944
}
L
Linus Torvalds 已提交
945 946 947 948 949 950 951 952 953 954

/*
 * Enqueue a work item to be picked up by the vfs xfssyncd thread.
 * Doing this has two advantages:
 * - It saves on stack space, which is tight in certain situations
 * - It can be used (with care) as a mechanism to avoid deadlocks.
 * Flushing while allocating in a full filesystem requires both.
 */
STATIC void
xfs_syncd_queue_work(
955
	struct xfs_mount *mp,
L
Linus Torvalds 已提交
956
	void		*data,
957
	void		(*syncer)(struct xfs_mount *, void *))
L
Linus Torvalds 已提交
958
{
959
	struct bhv_vfs_sync_work *work;
L
Linus Torvalds 已提交
960

961
	work = kmem_alloc(sizeof(struct bhv_vfs_sync_work), KM_SLEEP);
L
Linus Torvalds 已提交
962 963 964
	INIT_LIST_HEAD(&work->w_list);
	work->w_syncer = syncer;
	work->w_data = data;
965 966 967 968 969
	work->w_mount = mp;
	spin_lock(&mp->m_sync_lock);
	list_add_tail(&work->w_list, &mp->m_sync_list);
	spin_unlock(&mp->m_sync_lock);
	wake_up_process(mp->m_sync_task);
L
Linus Torvalds 已提交
970 971 972 973 974 975 976 977 978 979
}

/*
 * Flush delayed allocate data, attempting to free up reserved space
 * from existing allocations.  At this point a new allocation attempt
 * has failed with ENOSPC and we are in the process of scratching our
 * heads, looking about for more room...
 */
STATIC void
xfs_flush_inode_work(
980 981
	struct xfs_mount *mp,
	void		*arg)
L
Linus Torvalds 已提交
982
{
983 984 985
	struct inode	*inode = arg;
	filemap_flush(inode->i_mapping);
	iput(inode);
L
Linus Torvalds 已提交
986 987 988 989 990 991
}

void
xfs_flush_inode(
	xfs_inode_t	*ip)
{
992
	struct inode	*inode = VFS_I(ip);
L
Linus Torvalds 已提交
993 994

	igrab(inode);
995
	xfs_syncd_queue_work(ip->i_mount, inode, xfs_flush_inode_work);
996
	delay(msecs_to_jiffies(500));
L
Linus Torvalds 已提交
997 998 999 1000 1001 1002 1003 1004
}

/*
 * This is the "bigger hammer" version of xfs_flush_inode_work...
 * (IOW, "If at first you don't succeed, use a Bigger Hammer").
 */
STATIC void
xfs_flush_device_work(
1005 1006
	struct xfs_mount *mp,
	void		*arg)
L
Linus Torvalds 已提交
1007
{
1008
	struct inode	*inode = arg;
C
Christoph Hellwig 已提交
1009
	sync_blockdev(mp->m_super->s_bdev);
1010
	iput(inode);
L
Linus Torvalds 已提交
1011 1012 1013 1014 1015 1016
}

void
xfs_flush_device(
	xfs_inode_t	*ip)
{
1017
	struct inode	*inode = VFS_I(ip);
L
Linus Torvalds 已提交
1018 1019

	igrab(inode);
1020
	xfs_syncd_queue_work(ip->i_mount, inode, xfs_flush_device_work);
1021
	delay(msecs_to_jiffies(500));
L
Linus Torvalds 已提交
1022 1023 1024 1025
	xfs_log_force(ip->i_mount, (xfs_lsn_t)0, XFS_LOG_FORCE|XFS_LOG_SYNC);
}

STATIC void
1026 1027
xfs_sync_worker(
	struct xfs_mount *mp,
L
Linus Torvalds 已提交
1028 1029 1030 1031
	void		*unused)
{
	int		error;

1032
	if (!(mp->m_flags & XFS_MOUNT_RDONLY))
1033
		error = xfs_sync(mp, SYNC_FSDATA | SYNC_BDFLUSH | SYNC_ATTR);
1034 1035
	mp->m_sync_seq++;
	wake_up(&mp->m_wait_single_sync_task);
L
Linus Torvalds 已提交
1036 1037 1038 1039 1040 1041
}

STATIC int
xfssyncd(
	void			*arg)
{
1042
	struct xfs_mount	*mp = arg;
L
Linus Torvalds 已提交
1043
	long			timeleft;
1044
	bhv_vfs_sync_work_t	*work, *n;
1045
	LIST_HEAD		(tmp);
L
Linus Torvalds 已提交
1046

1047
	set_freezable();
1048
	timeleft = xfs_syncd_centisecs * msecs_to_jiffies(10);
L
Linus Torvalds 已提交
1049
	for (;;) {
1050
		timeleft = schedule_timeout_interruptible(timeleft);
L
Linus Torvalds 已提交
1051
		/* swsusp */
1052
		try_to_freeze();
1053
		if (kthread_should_stop() && list_empty(&mp->m_sync_list))
L
Linus Torvalds 已提交
1054 1055
			break;

1056
		spin_lock(&mp->m_sync_lock);
L
Linus Torvalds 已提交
1057 1058 1059 1060 1061
		/*
		 * We can get woken by laptop mode, to do a sync -
		 * that's the (only!) case where the list would be
		 * empty with time remaining.
		 */
1062
		if (!timeleft || list_empty(&mp->m_sync_list)) {
L
Linus Torvalds 已提交
1063
			if (!timeleft)
1064 1065
				timeleft = xfs_syncd_centisecs *
							msecs_to_jiffies(10);
1066 1067 1068
			INIT_LIST_HEAD(&mp->m_sync_work.w_list);
			list_add_tail(&mp->m_sync_work.w_list,
					&mp->m_sync_list);
L
Linus Torvalds 已提交
1069
		}
1070
		list_for_each_entry_safe(work, n, &mp->m_sync_list, w_list)
L
Linus Torvalds 已提交
1071
			list_move(&work->w_list, &tmp);
1072
		spin_unlock(&mp->m_sync_lock);
L
Linus Torvalds 已提交
1073 1074

		list_for_each_entry_safe(work, n, &tmp, w_list) {
1075
			(*work->w_syncer)(mp, work->w_data);
L
Linus Torvalds 已提交
1076
			list_del(&work->w_list);
1077
			if (work == &mp->m_sync_work)
L
Linus Torvalds 已提交
1078
				continue;
1079
			kmem_free(work);
L
Linus Torvalds 已提交
1080 1081 1082 1083 1084 1085
		}
	}

	return 0;
}

1086 1087 1088 1089 1090 1091 1092 1093 1094
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
1095
STATIC void
1096
xfs_fs_put_super(
L
Linus Torvalds 已提交
1097 1098
	struct super_block	*sb)
{
1099
	struct xfs_mount	*mp = XFS_M(sb);
1100 1101
	struct xfs_inode	*rip = mp->m_rootip;
	int			unmount_event_flags = 0;
L
Linus Torvalds 已提交
1102 1103
	int			error;

1104 1105
	kthread_stop(mp->m_sync_task);

1106
	xfs_sync(mp, SYNC_ATTR | SYNC_DELWRI);
1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148

#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 we're forcing a shutdown, typically because of a media error,
	 * we want to make sure we invalidate dirty pages that belong to
	 * referenced vnodes as well.
	 */
	if (XFS_FORCED_SHUTDOWN(mp)) {
		error = xfs_sync(mp, SYNC_WAIT | SYNC_CLOSE);
		ASSERT(error != EFSCORRUPTED);
	}

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

1149
	xfs_unmountfs(mp);
1150
	xfs_freesb(mp);
C
Christoph Hellwig 已提交
1151
	xfs_icsb_destroy_counters(mp);
1152
	xfs_close_devices(mp);
1153 1154
	xfs_qmops_put(mp);
	xfs_dmops_put(mp);
1155
	xfs_free_fsname(mp);
C
Christoph Hellwig 已提交
1156
	kfree(mp);
L
Linus Torvalds 已提交
1157 1158 1159
}

STATIC void
1160
xfs_fs_write_super(
L
Linus Torvalds 已提交
1161 1162
	struct super_block	*sb)
{
1163
	if (!(sb->s_flags & MS_RDONLY))
1164
		xfs_sync(XFS_M(sb), SYNC_FSDATA);
L
Linus Torvalds 已提交
1165 1166 1167 1168
	sb->s_dirt = 0;
}

STATIC int
1169
xfs_fs_sync_super(
L
Linus Torvalds 已提交
1170 1171 1172
	struct super_block	*sb,
	int			wait)
{
1173
	struct xfs_mount	*mp = XFS_M(sb);
1174 1175
	int			error;
	int			flags;
L
Linus Torvalds 已提交
1176

1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189
	/*
	 * 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.
	 */
	if (wait || unlikely(sb->s_frozen == SB_FREEZE_WRITE)) {
1190 1191 1192 1193 1194 1195 1196 1197
		/*
		 * First stage of freeze - no more writers will make progress
		 * now we are here, so we flush delwri and delalloc buffers
		 * here, then wait for all I/O to complete.  Data is frozen at
		 * that point. Metadata is not frozen, transactions can still
		 * occur here so don't bother flushing the buftarg (i.e
		 * SYNC_QUIESCE) because it'll just get dirty again.
		 */
1198
		flags = SYNC_DATA_QUIESCE;
1199
	} else
1200
		flags = SYNC_FSDATA;
L
Linus Torvalds 已提交
1201

1202
	error = xfs_sync(mp, flags);
L
Linus Torvalds 已提交
1203 1204 1205
	sb->s_dirt = 0;

	if (unlikely(laptop_mode)) {
1206
		int	prev_sync_seq = mp->m_sync_seq;
L
Linus Torvalds 已提交
1207 1208 1209 1210 1211 1212

		/*
		 * 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).
		 */
1213
		wake_up_process(mp->m_sync_task);
L
Linus Torvalds 已提交
1214 1215 1216 1217 1218 1219
		/*
		 * 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.
		 */
1220 1221
		wait_event(mp->m_wait_single_sync_task,
				mp->m_sync_seq != prev_sync_seq);
L
Linus Torvalds 已提交
1222 1223 1224 1225 1226 1227
	}

	return -error;
}

STATIC int
1228
xfs_fs_statfs(
1229
	struct dentry		*dentry,
L
Linus Torvalds 已提交
1230 1231
	struct kstatfs		*statp)
{
C
Christoph Hellwig 已提交
1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243
	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);

1244
	xfs_icsb_sync_counters(mp, XFS_ICSB_LAZY_COUNT);
C
Christoph Hellwig 已提交
1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269

	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
1270 1271 1272
}

STATIC int
1273
xfs_fs_remount(
L
Linus Torvalds 已提交
1274 1275 1276 1277
	struct super_block	*sb,
	int			*flags,
	char			*options)
{
1278
	struct xfs_mount	*mp = XFS_M(sb);
1279 1280
	substring_t		args[MAX_OPT_ARGS];
	char			*p;
L
Linus Torvalds 已提交
1281

1282 1283
	while ((p = strsep(&options, ",")) != NULL) {
		int token;
1284

1285 1286
		if (!*p)
			continue;
1287

1288 1289 1290
		token = match_token(p, tokens, args);
		switch (token) {
		case Opt_barrier:
1291
			mp->m_flags |= XFS_MOUNT_BARRIER;
1292 1293 1294 1295 1296 1297 1298 1299 1300 1301

			/*
			 * 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:
1302
			mp->m_flags &= ~XFS_MOUNT_BARRIER;
1303 1304 1305 1306 1307
			break;
		default:
			printk(KERN_INFO
	"XFS: mount option \"%s\" not supported for remount\n", p);
			return -EINVAL;
1308
		}
1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319
	}

	/* 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)) {
1320 1321 1322 1323 1324 1325
		xfs_filestream_flush(mp);
		xfs_sync(mp, SYNC_DATA_QUIESCE);
		xfs_attr_quiesce(mp);
		mp->m_flags |= XFS_MOUNT_RDONLY;
	}

1326
	return 0;
L
Linus Torvalds 已提交
1327 1328
}

C
Christoph Hellwig 已提交
1329 1330 1331 1332 1333
/*
 * Second stage of a freeze. The data is already frozen so we only
 * need to take care of themetadata. Once that's done write a dummy
 * record to dirty the log in case of a crash while frozen.
 */
L
Linus Torvalds 已提交
1334
STATIC void
1335
xfs_fs_lockfs(
L
Linus Torvalds 已提交
1336 1337
	struct super_block	*sb)
{
C
Christoph Hellwig 已提交
1338 1339 1340 1341
	struct xfs_mount	*mp = XFS_M(sb);

	xfs_attr_quiesce(mp);
	xfs_fs_log_dummy(mp);
L
Linus Torvalds 已提交
1342 1343 1344
}

STATIC int
1345
xfs_fs_show_options(
L
Linus Torvalds 已提交
1346 1347 1348
	struct seq_file		*m,
	struct vfsmount		*mnt)
{
1349
	return -xfs_showargs(XFS_M(mnt->mnt_sb), m);
L
Linus Torvalds 已提交
1350 1351
}

1352
STATIC int
1353
xfs_fs_quotasync(
1354 1355 1356
	struct super_block	*sb,
	int			type)
{
1357
	return -XFS_QM_QUOTACTL(XFS_M(sb), Q_XQUOTASYNC, 0, NULL);
1358 1359
}

L
Linus Torvalds 已提交
1360
STATIC int
1361
xfs_fs_getxstate(
L
Linus Torvalds 已提交
1362 1363 1364
	struct super_block	*sb,
	struct fs_quota_stat	*fqs)
{
1365
	return -XFS_QM_QUOTACTL(XFS_M(sb), Q_XGETQSTAT, 0, (caddr_t)fqs);
L
Linus Torvalds 已提交
1366 1367 1368
}

STATIC int
1369
xfs_fs_setxstate(
L
Linus Torvalds 已提交
1370 1371 1372 1373
	struct super_block	*sb,
	unsigned int		flags,
	int			op)
{
1374
	return -XFS_QM_QUOTACTL(XFS_M(sb), op, 0, (caddr_t)&flags);
L
Linus Torvalds 已提交
1375 1376 1377
}

STATIC int
1378
xfs_fs_getxquota(
L
Linus Torvalds 已提交
1379 1380 1381 1382 1383
	struct super_block	*sb,
	int			type,
	qid_t			id,
	struct fs_disk_quota	*fdq)
{
1384
	return -XFS_QM_QUOTACTL(XFS_M(sb),
1385 1386 1387
				 (type == USRQUOTA) ? Q_XGETQUOTA :
				  ((type == GRPQUOTA) ? Q_XGETGQUOTA :
				   Q_XGETPQUOTA), id, (caddr_t)fdq);
L
Linus Torvalds 已提交
1388 1389 1390
}

STATIC int
1391
xfs_fs_setxquota(
L
Linus Torvalds 已提交
1392 1393 1394 1395 1396
	struct super_block	*sb,
	int			type,
	qid_t			id,
	struct fs_disk_quota	*fdq)
{
1397
	return -XFS_QM_QUOTACTL(XFS_M(sb),
1398 1399 1400
				 (type == USRQUOTA) ? Q_XSETQLIM :
				  ((type == GRPQUOTA) ? Q_XSETGQLIM :
				   Q_XSETPQLIM), id, (caddr_t)fdq);
L
Linus Torvalds 已提交
1401 1402
}

1403 1404 1405 1406 1407 1408 1409 1410 1411
/*
 * 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)
{
1412 1413
	int			error;

1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445
	/* 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);
	}
1446 1447 1448

	error = ENOMEM;

1449 1450
	mp->m_logbsize = ap->logbufsize;
	mp->m_fsname_len = strlen(ap->fsname) + 1;
1451 1452 1453 1454 1455

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

1456
	if (ap->rtname[0]) {
1457 1458 1459 1460
		mp->m_rtname = kstrdup(ap->rtname, GFP_KERNEL);
		if (!mp->m_rtname)
			goto out_free_fsname;

1461
	}
1462

1463
	if (ap->logname[0]) {
1464 1465 1466
		mp->m_logname = kstrdup(ap->logname, GFP_KERNEL);
		if (!mp->m_logname)
			goto out_free_rtname;
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 1535 1536 1537 1538
	}

	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;
1539 1540 1541 1542 1543 1544 1545 1546


 out_free_rtname:
	kfree(mp->m_rtname);
 out_free_fsname:
	kfree(mp->m_fsname);
 out:
	return error;
1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 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 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641
}

/*
 * 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
1642
STATIC int
1643
xfs_fs_fill_super(
L
Linus Torvalds 已提交
1644 1645 1646 1647
	struct super_block	*sb,
	void			*data,
	int			silent)
{
1648
	struct inode		*root;
1649
	struct xfs_mount	*mp = NULL;
1650
	struct xfs_mount_args	*args;
C
Christoph Hellwig 已提交
1651
	int			flags = 0, error = ENOMEM;
L
Linus Torvalds 已提交
1652

1653 1654 1655 1656
	args = xfs_args_allocate(sb, silent);
	if (!args)
		return -ENOMEM;

C
Christoph Hellwig 已提交
1657 1658 1659
	mp = kzalloc(sizeof(struct xfs_mount), GFP_KERNEL);
	if (!mp)
		goto out_free_args;
L
Linus Torvalds 已提交
1660

C
Christoph Hellwig 已提交
1661 1662 1663 1664
	spin_lock_init(&mp->m_sb_lock);
	mutex_init(&mp->m_ilock);
	mutex_init(&mp->m_growlock);
	atomic_set(&mp->m_active_trans, 0);
1665 1666 1667 1668
	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
1669 1670
	mp->m_super = sb;
	sb->s_fs_info = mp;
L
Linus Torvalds 已提交
1671

1672 1673 1674
	if (sb->s_flags & MS_RDONLY)
		mp->m_flags |= XFS_MOUNT_RDONLY;

1675 1676
	error = xfs_parseargs(mp, (char *)data, args, 0);
	if (error)
C
Christoph Hellwig 已提交
1677
		goto out_free_mp;
L
Linus Torvalds 已提交
1678 1679

	sb_min_blocksize(sb, BBSIZE);
1680
	sb->s_xattr = xfs_xattr_handlers;
1681 1682 1683
	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
1684

1685 1686
	error = xfs_dmops_get(mp, args);
	if (error)
C
Christoph Hellwig 已提交
1687
		goto out_free_mp;
1688
	error = xfs_qmops_get(mp, args);
1689
	if (error)
1690
		goto out_put_dmops;
L
Linus Torvalds 已提交
1691

1692 1693 1694
	if (args->flags & XFSMNT_QUIET)
		flags |= XFS_MFSI_QUIET;

1695 1696 1697
	error = xfs_open_devices(mp, args);
	if (error)
		goto out_put_qmops;
1698

C
Christoph Hellwig 已提交
1699 1700 1701
	if (xfs_icsb_init_counters(mp))
		mp->m_flags |= XFS_MOUNT_NO_PERCPU_SB;

1702 1703 1704 1705 1706
	/*
	 * Setup flags based on mount(2) options and then the superblock
	 */
	error = xfs_start_flags(args, mp);
	if (error)
1707
		goto out_free_fsname;
1708 1709
	error = xfs_readsb(mp, flags);
	if (error)
1710
		goto out_free_fsname;
1711 1712
	error = xfs_finish_flags(args, mp);
	if (error)
1713
		goto out_free_sb;
1714

1715
	error = xfs_setup_devices(mp);
1716
	if (error)
1717
		goto out_free_sb;
1718 1719 1720 1721 1722 1723

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

	error = xfs_filestream_mount(mp);
	if (error)
1724
		goto out_free_sb;
1725

C
Christoph Hellwig 已提交
1726
	error = xfs_mountfs(mp);
1727
	if (error)
1728
		goto out_filestream_unmount;
1729 1730 1731

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

L
Linus Torvalds 已提交
1732
	sb->s_dirt = 1;
C
Christoph Hellwig 已提交
1733 1734 1735
	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
1736 1737 1738 1739
	sb->s_maxbytes = xfs_max_file_offset(sb->s_blocksize_bits);
	sb->s_time_gran = 1;
	set_posix_acl_flag(sb);

1740
	root = igrab(VFS_I(mp->m_rootip));
1741
	if (!root) {
C
Christoph Hellwig 已提交
1742
		error = ENOENT;
L
Linus Torvalds 已提交
1743
		goto fail_unmount;
C
Christoph Hellwig 已提交
1744
	}
1745 1746
	if (is_bad_inode(root)) {
		error = EINVAL;
L
Linus Torvalds 已提交
1747 1748
		goto fail_vnrele;
	}
1749 1750 1751
	sb->s_root = d_alloc_root(root);
	if (!sb->s_root) {
		error = ENOMEM;
L
Linus Torvalds 已提交
1752 1753
		goto fail_vnrele;
	}
1754 1755 1756 1757 1758 1759

	mp->m_sync_work.w_syncer = xfs_sync_worker;
	mp->m_sync_work.w_mount = mp;
	mp->m_sync_task = kthread_run(xfssyncd, mp, "xfssyncd");
	if (IS_ERR(mp->m_sync_task)) {
		error = -PTR_ERR(mp->m_sync_task);
L
Linus Torvalds 已提交
1760
		goto fail_vnrele;
1761 1762
	}

1763
	xfs_itrace_exit(XFS_I(sb->s_root->d_inode));
L
Linus Torvalds 已提交
1764

1765
	kfree(args);
L
Linus Torvalds 已提交
1766 1767
	return 0;

1768 1769
 out_filestream_unmount:
	xfs_filestream_unmount(mp);
1770 1771
 out_free_sb:
	xfs_freesb(mp);
1772 1773
 out_free_fsname:
	xfs_free_fsname(mp);
C
Christoph Hellwig 已提交
1774
	xfs_icsb_destroy_counters(mp);
1775 1776
	xfs_close_devices(mp);
 out_put_qmops:
1777
	xfs_qmops_put(mp);
1778
 out_put_dmops:
1779
	xfs_dmops_put(mp);
C
Christoph Hellwig 已提交
1780 1781 1782 1783 1784
 out_free_mp:
	kfree(mp);
 out_free_args:
	kfree(args);
	return -error;
1785 1786

 fail_vnrele:
L
Linus Torvalds 已提交
1787 1788 1789 1790
	if (sb->s_root) {
		dput(sb->s_root);
		sb->s_root = NULL;
	} else {
1791
		iput(root);
L
Linus Torvalds 已提交
1792 1793
	}

1794
 fail_unmount:
1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805
	/*
	 * 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);

1806
	xfs_unmountfs(mp);
1807
	goto out_free_sb;
L
Linus Torvalds 已提交
1808 1809
}

1810
STATIC int
1811
xfs_fs_get_sb(
L
Linus Torvalds 已提交
1812 1813 1814
	struct file_system_type	*fs_type,
	int			flags,
	const char		*dev_name,
1815 1816
	void			*data,
	struct vfsmount		*mnt)
L
Linus Torvalds 已提交
1817
{
1818 1819
	return get_sb_bdev(fs_type, flags, dev_name, data, xfs_fs_fill_super,
			   mnt);
1820 1821
}

1822
static struct super_operations xfs_super_operations = {
1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833
	.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
1834 1835
};

1836
static struct quotactl_ops xfs_quotactl_operations = {
1837 1838 1839 1840 1841
	.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
1842 1843
};

A
Andrew Morton 已提交
1844
static struct file_system_type xfs_fs_type = {
L
Linus Torvalds 已提交
1845 1846
	.owner			= THIS_MODULE,
	.name			= "xfs",
1847
	.get_sb			= xfs_fs_get_sb,
L
Linus Torvalds 已提交
1848 1849 1850 1851
	.kill_sb		= kill_block_super,
	.fs_flags		= FS_REQUIRES_DEV,
};

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 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927
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
#ifdef XFS_BMBT_TRACE
	xfs_bmbt_trace_buf = ktrace_alloc(XFS_BMBT_TRACE_SIZE, KM_MAYFAIL);
	if (!xfs_bmbt_trace_buf)
		goto out_free_bmap_trace;
#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
#ifdef XFS_BMBT_TRACE
	ktrace_free(xfs_bmbt_trace_buf);
 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
#ifdef XFS_BMBT_TRACE
	ktrace_free(xfs_bmbt_trace_buf);
#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
1928
	xfs_vnode_zone = kmem_zone_init_flags(sizeof(struct inode), "xfs_vnode",
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 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077
					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 |
					KM_ZONE_SPREAD, NULL);
	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
2078 2079

STATIC int __init
2080
init_xfs_fs(void)
L
Linus Torvalds 已提交
2081 2082 2083 2084 2085 2086 2087 2088
{
	int			error;
	static char		message[] __initdata = KERN_INFO \
		XFS_VERSION_STRING " with " XFS_BUILD_OPTIONS " enabled\n";

	printk(message);

	ktrace_init(64);
2089 2090
	vn_init();
	xfs_dir_startup();
L
Linus Torvalds 已提交
2091

2092
	error = xfs_init_zones();
2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106
	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
2107

2108
	error = xfs_buf_init();
2109 2110 2111 2112 2113 2114 2115 2116 2117 2118
	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
2119 2120 2121 2122 2123

	vfs_initquota();

	error = register_filesystem(&xfs_fs_type);
	if (error)
2124
		goto out_sysctl_unregister;
L
Linus Torvalds 已提交
2125 2126
	return 0;

2127 2128 2129 2130 2131
 out_sysctl_unregister:
	xfs_sysctl_unregister();
 out_cleanup_procfs:
	xfs_cleanup_procfs();
 out_buf_terminate:
2132
	xfs_buf_terminate();
2133 2134 2135 2136 2137 2138 2139
 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:
2140
	xfs_destroy_zones();
2141
 out:
L
Linus Torvalds 已提交
2142 2143 2144 2145
	return error;
}

STATIC void __exit
2146
exit_xfs_fs(void)
L
Linus Torvalds 已提交
2147 2148 2149
{
	vfs_exitquota();
	unregister_filesystem(&xfs_fs_type);
2150 2151
	xfs_sysctl_unregister();
	xfs_cleanup_procfs();
2152
	xfs_buf_terminate();
2153 2154 2155
	xfs_filestream_uninit();
	xfs_mru_cache_uninit();
	xfs_free_trace_bufs();
2156
	xfs_destroy_zones();
L
Linus Torvalds 已提交
2157 2158 2159 2160 2161 2162 2163 2164 2165
	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");