xfs_iget.c 18.5 KB
Newer Older
L
Linus Torvalds 已提交
1
/*
2 3
 * Copyright (c) 2000-2005 Silicon Graphics, Inc.
 * 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_fs.h"
L
Linus Torvalds 已提交
20
#include "xfs_types.h"
21
#include "xfs_acl.h"
22
#include "xfs_bit.h"
L
Linus Torvalds 已提交
23
#include "xfs_log.h"
24
#include "xfs_inum.h"
L
Linus Torvalds 已提交
25 26 27 28 29
#include "xfs_trans.h"
#include "xfs_sb.h"
#include "xfs_ag.h"
#include "xfs_mount.h"
#include "xfs_bmap_btree.h"
30
#include "xfs_alloc_btree.h"
L
Linus Torvalds 已提交
31 32 33
#include "xfs_ialloc_btree.h"
#include "xfs_dinode.h"
#include "xfs_inode.h"
34 35
#include "xfs_btree.h"
#include "xfs_ialloc.h"
L
Linus Torvalds 已提交
36 37
#include "xfs_quota.h"
#include "xfs_utils.h"
38 39
#include "xfs_trans_priv.h"
#include "xfs_inode_item.h"
40 41
#include "xfs_bmap.h"
#include "xfs_btree_trace.h"
C
Christoph Hellwig 已提交
42
#include "xfs_trace.h"
43 44


45 46 47 48 49 50 51 52 53 54 55
/*
 * Define xfs inode iolock lockdep classes. We need to ensure that all active
 * inodes are considered the same for lockdep purposes, including inodes that
 * are recycled through the XFS_IRECLAIMABLE state. This is the the only way to
 * guarantee the locks are considered the same when there are multiple lock
 * initialisation siteѕ. Also, define a reclaimable inode class so it is
 * obvious in lockdep reports which class the report is against.
 */
static struct lock_class_key xfs_iolock_active;
struct lock_class_key xfs_iolock_reclaimable;

56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73
/*
 * Allocate and initialise an xfs_inode.
 */
STATIC struct xfs_inode *
xfs_inode_alloc(
	struct xfs_mount	*mp,
	xfs_ino_t		ino)
{
	struct xfs_inode	*ip;

	/*
	 * if this didn't occur in transactions, we could use
	 * KM_MAYFAIL and return NULL here on ENOMEM. Set the
	 * code up to do this anyway.
	 */
	ip = kmem_zone_alloc(xfs_inode_zone, KM_SLEEP);
	if (!ip)
		return NULL;
74 75 76 77
	if (inode_init_always(mp->m_super, VFS_I(ip))) {
		kmem_zone_free(xfs_inode_zone, ip);
		return NULL;
	}
78 79 80 81 82

	ASSERT(atomic_read(&ip->i_iocount) == 0);
	ASSERT(atomic_read(&ip->i_pincount) == 0);
	ASSERT(!spin_is_locked(&ip->i_flags_lock));
	ASSERT(completion_done(&ip->i_flush));
83 84

	mrlock_init(&ip->i_iolock, MRLOCK_BARRIER, "xfsio", ip->i_ino);
85 86
	lockdep_set_class_and_name(&ip->i_iolock.mr_lock,
			&xfs_iolock_active, "xfs_iolock_active");
87 88 89 90 91 92 93 94 95 96 97 98 99 100

	/* initialise the xfs inode */
	ip->i_ino = ino;
	ip->i_mount = mp;
	memset(&ip->i_imap, 0, sizeof(struct xfs_imap));
	ip->i_afp = NULL;
	memset(&ip->i_df, 0, sizeof(xfs_ifork_t));
	ip->i_flags = 0;
	ip->i_update_core = 0;
	ip->i_delayed_blks = 0;
	memset(&ip->i_d, 0, sizeof(xfs_icdinode_t));
	ip->i_size = 0;
	ip->i_new_size = 0;

D
Dave Chinner 已提交
101
	/* prevent anyone from using this yet */
C
Christoph Hellwig 已提交
102
	VFS_I(ip)->i_state = I_NEW;
103 104 105

	return ip;
}
L
Linus Torvalds 已提交
106

D
Dave Chinner 已提交
107 108 109 110 111 112 113 114 115 116 117 118
void
__xfs_inode_free(
	struct rcu_head		*head)
{
	struct inode		*inode = container_of((void *)head,
							struct inode, i_dentry);
	struct xfs_inode	*ip = XFS_I(inode);

	INIT_LIST_HEAD(&inode->i_dentry);
	kmem_zone_free(xfs_inode_zone, ip);
}

119
void
120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161
xfs_inode_free(
	struct xfs_inode	*ip)
{
	switch (ip->i_d.di_mode & S_IFMT) {
	case S_IFREG:
	case S_IFDIR:
	case S_IFLNK:
		xfs_idestroy_fork(ip, XFS_DATA_FORK);
		break;
	}

	if (ip->i_afp)
		xfs_idestroy_fork(ip, XFS_ATTR_FORK);

	if (ip->i_itemp) {
		/*
		 * Only if we are shutting down the fs will we see an
		 * inode still in the AIL. If it is there, we should remove
		 * it to prevent a use-after-free from occurring.
		 */
		xfs_log_item_t	*lip = &ip->i_itemp->ili_item;
		struct xfs_ail	*ailp = lip->li_ailp;

		ASSERT(((lip->li_flags & XFS_LI_IN_AIL) == 0) ||
				       XFS_FORCED_SHUTDOWN(ip->i_mount));
		if (lip->li_flags & XFS_LI_IN_AIL) {
			spin_lock(&ailp->xa_lock);
			if (lip->li_flags & XFS_LI_IN_AIL)
				xfs_trans_ail_delete(ailp, lip);
			else
				spin_unlock(&ailp->xa_lock);
		}
		xfs_inode_item_destroy(ip);
		ip->i_itemp = NULL;
	}

	/* asserts to verify all state is correct here */
	ASSERT(atomic_read(&ip->i_iocount) == 0);
	ASSERT(atomic_read(&ip->i_pincount) == 0);
	ASSERT(!spin_is_locked(&ip->i_flags_lock));
	ASSERT(completion_done(&ip->i_flush));

D
Dave Chinner 已提交
162
	call_rcu((struct rcu_head *)&VFS_I(ip)->i_dentry, __xfs_inode_free);
163 164
}

L
Linus Torvalds 已提交
165
/*
166
 * Check the validity of the inode we just found it the cache
L
Linus Torvalds 已提交
167
 */
168 169 170 171 172 173
static int
xfs_iget_cache_hit(
	struct xfs_perag	*pag,
	struct xfs_inode	*ip,
	int			flags,
	int			lock_flags) __releases(pag->pag_ici_lock)
L
Linus Torvalds 已提交
174
{
175
	struct inode		*inode = VFS_I(ip);
176
	struct xfs_mount	*mp = ip->i_mount;
177 178 179
	int			error;

	spin_lock(&ip->i_flags_lock);
180

181
	/*
182 183 184 185 186 187 188 189
	 * If we are racing with another cache hit that is currently
	 * instantiating this inode or currently recycling it out of
	 * reclaimabe state, wait for the initialisation to complete
	 * before continuing.
	 *
	 * XXX(hch): eventually we should do something equivalent to
	 *	     wait_on_inode to wait for these flags to be cleared
	 *	     instead of polling for it.
190
	 */
191
	if (ip->i_flags & (XFS_INEW|XFS_IRECLAIM)) {
C
Christoph Hellwig 已提交
192
		trace_xfs_iget_skip(ip);
193
		XFS_STATS_INC(xs_ig_frecycle);
194
		error = EAGAIN;
195 196
		goto out_error;
	}
197

198 199 200 201 202 203 204
	/*
	 * If lookup is racing with unlink return an error immediately.
	 */
	if (ip->i_d.di_mode == 0 && !(flags & XFS_IGET_CREATE)) {
		error = ENOENT;
		goto out_error;
	}
205

206 207 208 209 210
	/*
	 * If IRECLAIMABLE is set, we've torn down the VFS inode already.
	 * Need to carefully get it back into useable state.
	 */
	if (ip->i_flags & XFS_IRECLAIMABLE) {
C
Christoph Hellwig 已提交
211
		trace_xfs_iget_reclaim(ip);
212

213
		/*
214 215 216 217
		 * We need to set XFS_IRECLAIM to prevent xfs_reclaim_inode
		 * from stomping over us while we recycle the inode.  We can't
		 * clear the radix tree reclaimable tag yet as it requires
		 * pag_ici_lock to be held exclusive.
218
		 */
219
		ip->i_flags |= XFS_IRECLAIM;
220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235

		spin_unlock(&ip->i_flags_lock);
		read_unlock(&pag->pag_ici_lock);

		error = -inode_init_always(mp->m_super, inode);
		if (error) {
			/*
			 * Re-initializing the inode failed, and we are in deep
			 * trouble.  Try to re-add it to the reclaim list.
			 */
			read_lock(&pag->pag_ici_lock);
			spin_lock(&ip->i_flags_lock);

			ip->i_flags &= ~XFS_INEW;
			ip->i_flags |= XFS_IRECLAIMABLE;
			__xfs_inode_set_reclaim_tag(pag, ip);
236
			trace_xfs_iget_reclaim_fail(ip);
237
			goto out_error;
238
		}
239 240 241 242 243 244

		write_lock(&pag->pag_ici_lock);
		spin_lock(&ip->i_flags_lock);
		ip->i_flags &= ~(XFS_IRECLAIMABLE | XFS_IRECLAIM);
		ip->i_flags |= XFS_INEW;
		__xfs_inode_clear_reclaim_tag(mp, pag, ip);
C
Christoph Hellwig 已提交
245
		inode->i_state = I_NEW;
246 247 248 249 250 251

		ASSERT(!rwsem_is_locked(&ip->i_iolock.mr_lock));
		mrlock_init(&ip->i_iolock, MRLOCK_BARRIER, "xfsio", ip->i_ino);
		lockdep_set_class_and_name(&ip->i_iolock.mr_lock,
				&xfs_iolock_active, "xfs_iolock_active");

252 253
		spin_unlock(&ip->i_flags_lock);
		write_unlock(&pag->pag_ici_lock);
254
	} else {
255
		/* If the VFS inode is being torn down, pause and try again. */
256
		if (!igrab(inode)) {
257
			trace_xfs_iget_skip(ip);
258 259 260
			error = EAGAIN;
			goto out_error;
		}
L
Linus Torvalds 已提交
261

262 263 264
		/* We've got a live one. */
		spin_unlock(&ip->i_flags_lock);
		read_unlock(&pag->pag_ici_lock);
265
		trace_xfs_iget_hit(ip);
266
	}
267

268 269
	if (lock_flags != 0)
		xfs_ilock(ip, lock_flags);
270

271 272
	xfs_iflags_clear(ip, XFS_ISTALE);
	XFS_STATS_INC(xs_ig_found);
C
Christoph Hellwig 已提交
273

274
	return 0;
L
Linus Torvalds 已提交
275

276
out_error:
277
	spin_unlock(&ip->i_flags_lock);
278
	read_unlock(&pag->pag_ici_lock);
279 280 281 282 283 284 285 286 287 288 289 290
	return error;
}


static int
xfs_iget_cache_miss(
	struct xfs_mount	*mp,
	struct xfs_perag	*pag,
	xfs_trans_t		*tp,
	xfs_ino_t		ino,
	struct xfs_inode	**ipp,
	int			flags,
291
	int			lock_flags)
292 293 294 295
{
	struct xfs_inode	*ip;
	int			error;
	xfs_agino_t		agino = XFS_INO_TO_AGINO(mp, ino);
L
Linus Torvalds 已提交
296

297 298 299 300
	ip = xfs_inode_alloc(mp, ino);
	if (!ip)
		return ENOMEM;

301
	error = xfs_iread(mp, tp, ip, flags);
302
	if (error)
303
		goto out_destroy;
L
Linus Torvalds 已提交
304

305
	trace_xfs_iget_miss(ip);
L
Linus Torvalds 已提交
306

307
	if ((ip->i_d.di_mode == 0) && !(flags & XFS_IGET_CREATE)) {
308 309
		error = ENOENT;
		goto out_destroy;
L
Linus Torvalds 已提交
310 311 312
	}

	/*
313
	 * Preload the radix tree so we can insert safely under the
314 315
	 * write spinlock. Note that we cannot sleep inside the preload
	 * region.
L
Linus Torvalds 已提交
316
	 */
317
	if (radix_tree_preload(GFP_KERNEL)) {
318
		error = EAGAIN;
319 320 321 322 323 324 325 326 327 328
		goto out_destroy;
	}

	/*
	 * Because the inode hasn't been added to the radix-tree yet it can't
	 * be found by another thread, so we can do the non-sleeping lock here.
	 */
	if (lock_flags) {
		if (!xfs_ilock_nowait(ip, lock_flags))
			BUG();
329
	}
330

331
	write_lock(&pag->pag_ici_lock);
332 333

	/* insert the new inode */
334 335
	error = radix_tree_insert(&pag->pag_ici_root, agino, ip);
	if (unlikely(error)) {
336
		WARN_ON(error != -EEXIST);
337
		XFS_STATS_INC(xs_ig_dup);
338
		error = EAGAIN;
339
		goto out_preload_end;
L
Linus Torvalds 已提交
340 341
	}

342
	/* These values _must_ be set before releasing the radix tree lock! */
L
Linus Torvalds 已提交
343
	ip->i_udquot = ip->i_gdquot = NULL;
344
	xfs_iflags_set(ip, XFS_INEW);
L
Linus Torvalds 已提交
345

346 347
	write_unlock(&pag->pag_ici_lock);
	radix_tree_preload_end();
C
Christoph Hellwig 已提交
348

349 350 351
	*ipp = ip;
	return 0;

352
out_preload_end:
353 354
	write_unlock(&pag->pag_ici_lock);
	radix_tree_preload_end();
355 356
	if (lock_flags)
		xfs_iunlock(ip, lock_flags);
357
out_destroy:
358 359
	__destroy_inode(VFS_I(ip));
	xfs_inode_free(ip);
360 361 362 363 364 365
	return error;
}

/*
 * Look up an inode by number in the given file system.
 * The inode is looked up in the cache held in each AG.
366 367
 * If the inode is found in the cache, initialise the vfs inode
 * if necessary.
368 369
 *
 * If it is not in core, read it in from the file system's device,
370
 * add it to the cache and initialise the vfs inode.
371 372 373 374 375 376 377 378 379 380 381 382 383 384
 *
 * The inode is locked according to the value of the lock_flags parameter.
 * This flag parameter indicates how and if the inode's IO lock and inode lock
 * should be taken.
 *
 * mp -- the mount point structure for the current file system.  It points
 *       to the inode hash table.
 * tp -- a pointer to the current transaction if there is one.  This is
 *       simply passed through to the xfs_iread() call.
 * ino -- the number of the inode desired.  This is the unique identifier
 *        within the file system for the inode being requested.
 * lock_flags -- flags indicating how to lock the inode.  See the comment
 *		 for xfs_ilock() for a list of valid values.
 */
385 386
int
xfs_iget(
387 388 389 390 391
	xfs_mount_t	*mp,
	xfs_trans_t	*tp,
	xfs_ino_t	ino,
	uint		flags,
	uint		lock_flags,
392
	xfs_inode_t	**ipp)
393 394 395 396 397 398
{
	xfs_inode_t	*ip;
	int		error;
	xfs_perag_t	*pag;
	xfs_agino_t	agino;

399 400
	/* reject inode numbers outside existing AGs */
	if (XFS_INO_TO_AGNO(mp, ino) >= mp->m_sb.sb_agcount)
401 402 403
		return EINVAL;

	/* get the perag structure and ensure that it's inode capable */
D
Dave Chinner 已提交
404
	pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, ino));
405 406 407 408 409 410 411 412
	agino = XFS_INO_TO_AGINO(mp, ino);

again:
	error = 0;
	read_lock(&pag->pag_ici_lock);
	ip = radix_tree_lookup(&pag->pag_ici_root, agino);

	if (ip) {
413
		error = xfs_iget_cache_hit(pag, ip, flags, lock_flags);
414 415 416 417 418 419
		if (error)
			goto out_error_or_again;
	} else {
		read_unlock(&pag->pag_ici_lock);
		XFS_STATS_INC(xs_ig_missed);

420
		error = xfs_iget_cache_miss(mp, pag, tp, ino, &ip,
421 422 423 424
							flags, lock_flags);
		if (error)
			goto out_error_or_again;
	}
D
Dave Chinner 已提交
425
	xfs_perag_put(pag);
L
Linus Torvalds 已提交
426 427 428

	*ipp = ip;

429 430
	ASSERT(ip->i_df.if_ext_max ==
	       XFS_IFORK_DSIZE(ip) / sizeof(xfs_bmbt_rec_t));
L
Linus Torvalds 已提交
431 432 433 434
	/*
	 * If we have a real type for an on-disk inode, we can set ops(&unlock)
	 * now.	 If it's a new inode being created, xfs_ialloc will handle it.
	 */
435
	if (xfs_iflags_test(ip, XFS_INEW) && ip->i_d.di_mode != 0)
436
		xfs_setup_inode(ip);
L
Linus Torvalds 已提交
437
	return 0;
438 439 440 441 442 443

out_error_or_again:
	if (error == EAGAIN) {
		delay(1);
		goto again;
	}
D
Dave Chinner 已提交
444
	xfs_perag_put(pag);
445
	return error;
L
Linus Torvalds 已提交
446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516
}

/*
 * This is a wrapper routine around the xfs_ilock() routine
 * used to centralize some grungy code.  It is used in places
 * that wish to lock the inode solely for reading the extents.
 * The reason these places can't just call xfs_ilock(SHARED)
 * is that the inode lock also guards to bringing in of the
 * extents from disk for a file in b-tree format.  If the inode
 * is in b-tree format, then we need to lock the inode exclusively
 * until the extents are read in.  Locking it exclusively all
 * the time would limit our parallelism unnecessarily, though.
 * What we do instead is check to see if the extents have been
 * read in yet, and only lock the inode exclusively if they
 * have not.
 *
 * The function returns a value which should be given to the
 * corresponding xfs_iunlock_map_shared().  This value is
 * the mode in which the lock was actually taken.
 */
uint
xfs_ilock_map_shared(
	xfs_inode_t	*ip)
{
	uint	lock_mode;

	if ((ip->i_d.di_format == XFS_DINODE_FMT_BTREE) &&
	    ((ip->i_df.if_flags & XFS_IFEXTENTS) == 0)) {
		lock_mode = XFS_ILOCK_EXCL;
	} else {
		lock_mode = XFS_ILOCK_SHARED;
	}

	xfs_ilock(ip, lock_mode);

	return lock_mode;
}

/*
 * This is simply the unlock routine to go with xfs_ilock_map_shared().
 * All it does is call xfs_iunlock() with the given lock_mode.
 */
void
xfs_iunlock_map_shared(
	xfs_inode_t	*ip,
	unsigned int	lock_mode)
{
	xfs_iunlock(ip, lock_mode);
}

/*
 * The xfs inode contains 2 locks: a multi-reader lock called the
 * i_iolock and a multi-reader lock called the i_lock.  This routine
 * allows either or both of the locks to be obtained.
 *
 * The 2 locks should always be ordered so that the IO lock is
 * obtained first in order to prevent deadlock.
 *
 * ip -- the inode being locked
 * lock_flags -- this parameter indicates the inode's locks
 *       to be locked.  It can be:
 *		XFS_IOLOCK_SHARED,
 *		XFS_IOLOCK_EXCL,
 *		XFS_ILOCK_SHARED,
 *		XFS_ILOCK_EXCL,
 *		XFS_IOLOCK_SHARED | XFS_ILOCK_SHARED,
 *		XFS_IOLOCK_SHARED | XFS_ILOCK_EXCL,
 *		XFS_IOLOCK_EXCL | XFS_ILOCK_SHARED,
 *		XFS_IOLOCK_EXCL | XFS_ILOCK_EXCL
 */
void
C
Christoph Hellwig 已提交
517 518 519
xfs_ilock(
	xfs_inode_t		*ip,
	uint			lock_flags)
L
Linus Torvalds 已提交
520 521 522 523 524 525 526 527 528 529
{
	/*
	 * You can't set both SHARED and EXCL for the same lock,
	 * and only XFS_IOLOCK_SHARED, XFS_IOLOCK_EXCL, XFS_ILOCK_SHARED,
	 * and XFS_ILOCK_EXCL are valid values to set in lock_flags.
	 */
	ASSERT((lock_flags & (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL)) !=
	       (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL));
	ASSERT((lock_flags & (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)) !=
	       (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL));
530
	ASSERT((lock_flags & ~(XFS_LOCK_MASK | XFS_LOCK_DEP_MASK)) == 0);
L
Linus Torvalds 已提交
531

C
Christoph Hellwig 已提交
532
	if (lock_flags & XFS_IOLOCK_EXCL)
533
		mrupdate_nested(&ip->i_iolock, XFS_IOLOCK_DEP(lock_flags));
C
Christoph Hellwig 已提交
534
	else if (lock_flags & XFS_IOLOCK_SHARED)
535
		mraccess_nested(&ip->i_iolock, XFS_IOLOCK_DEP(lock_flags));
C
Christoph Hellwig 已提交
536 537

	if (lock_flags & XFS_ILOCK_EXCL)
538
		mrupdate_nested(&ip->i_lock, XFS_ILOCK_DEP(lock_flags));
C
Christoph Hellwig 已提交
539
	else if (lock_flags & XFS_ILOCK_SHARED)
540
		mraccess_nested(&ip->i_lock, XFS_ILOCK_DEP(lock_flags));
C
Christoph Hellwig 已提交
541

C
Christoph Hellwig 已提交
542
	trace_xfs_ilock(ip, lock_flags, _RET_IP_);
L
Linus Torvalds 已提交
543 544 545 546 547 548 549 550 551 552 553 554 555 556 557
}

/*
 * This is just like xfs_ilock(), except that the caller
 * is guaranteed not to sleep.  It returns 1 if it gets
 * the requested locks and 0 otherwise.  If the IO lock is
 * obtained but the inode lock cannot be, then the IO lock
 * is dropped before returning.
 *
 * ip -- the inode being locked
 * lock_flags -- this parameter indicates the inode's locks to be
 *       to be locked.  See the comment for xfs_ilock() for a list
 *	 of valid values.
 */
int
C
Christoph Hellwig 已提交
558 559 560
xfs_ilock_nowait(
	xfs_inode_t		*ip,
	uint			lock_flags)
L
Linus Torvalds 已提交
561 562 563 564 565 566 567 568 569 570
{
	/*
	 * You can't set both SHARED and EXCL for the same lock,
	 * and only XFS_IOLOCK_SHARED, XFS_IOLOCK_EXCL, XFS_ILOCK_SHARED,
	 * and XFS_ILOCK_EXCL are valid values to set in lock_flags.
	 */
	ASSERT((lock_flags & (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL)) !=
	       (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL));
	ASSERT((lock_flags & (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)) !=
	       (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL));
571
	ASSERT((lock_flags & ~(XFS_LOCK_MASK | XFS_LOCK_DEP_MASK)) == 0);
L
Linus Torvalds 已提交
572 573

	if (lock_flags & XFS_IOLOCK_EXCL) {
C
Christoph Hellwig 已提交
574 575
		if (!mrtryupdate(&ip->i_iolock))
			goto out;
L
Linus Torvalds 已提交
576
	} else if (lock_flags & XFS_IOLOCK_SHARED) {
C
Christoph Hellwig 已提交
577 578
		if (!mrtryaccess(&ip->i_iolock))
			goto out;
L
Linus Torvalds 已提交
579 580
	}
	if (lock_flags & XFS_ILOCK_EXCL) {
C
Christoph Hellwig 已提交
581 582
		if (!mrtryupdate(&ip->i_lock))
			goto out_undo_iolock;
L
Linus Torvalds 已提交
583
	} else if (lock_flags & XFS_ILOCK_SHARED) {
C
Christoph Hellwig 已提交
584 585
		if (!mrtryaccess(&ip->i_lock))
			goto out_undo_iolock;
L
Linus Torvalds 已提交
586
	}
C
Christoph Hellwig 已提交
587
	trace_xfs_ilock_nowait(ip, lock_flags, _RET_IP_);
L
Linus Torvalds 已提交
588
	return 1;
C
Christoph Hellwig 已提交
589 590 591 592 593 594 595 596

 out_undo_iolock:
	if (lock_flags & XFS_IOLOCK_EXCL)
		mrunlock_excl(&ip->i_iolock);
	else if (lock_flags & XFS_IOLOCK_SHARED)
		mrunlock_shared(&ip->i_iolock);
 out:
	return 0;
L
Linus Torvalds 已提交
597 598 599 600 601 602 603 604 605 606 607 608 609 610 611
}

/*
 * xfs_iunlock() is used to drop the inode locks acquired with
 * xfs_ilock() and xfs_ilock_nowait().  The caller must pass
 * in the flags given to xfs_ilock() or xfs_ilock_nowait() so
 * that we know which locks to drop.
 *
 * ip -- the inode being unlocked
 * lock_flags -- this parameter indicates the inode's locks to be
 *       to be unlocked.  See the comment for xfs_ilock() for a list
 *	 of valid values for this parameter.
 *
 */
void
C
Christoph Hellwig 已提交
612 613 614
xfs_iunlock(
	xfs_inode_t		*ip,
	uint			lock_flags)
L
Linus Torvalds 已提交
615 616 617 618 619 620 621 622 623 624
{
	/*
	 * You can't set both SHARED and EXCL for the same lock,
	 * and only XFS_IOLOCK_SHARED, XFS_IOLOCK_EXCL, XFS_ILOCK_SHARED,
	 * and XFS_ILOCK_EXCL are valid values to set in lock_flags.
	 */
	ASSERT((lock_flags & (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL)) !=
	       (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL));
	ASSERT((lock_flags & (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)) !=
	       (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL));
625 626
	ASSERT((lock_flags & ~(XFS_LOCK_MASK | XFS_IUNLOCK_NONOTIFY |
			XFS_LOCK_DEP_MASK)) == 0);
L
Linus Torvalds 已提交
627 628
	ASSERT(lock_flags != 0);

C
Christoph Hellwig 已提交
629 630 631 632
	if (lock_flags & XFS_IOLOCK_EXCL)
		mrunlock_excl(&ip->i_iolock);
	else if (lock_flags & XFS_IOLOCK_SHARED)
		mrunlock_shared(&ip->i_iolock);
L
Linus Torvalds 已提交
633

C
Christoph Hellwig 已提交
634 635 636 637
	if (lock_flags & XFS_ILOCK_EXCL)
		mrunlock_excl(&ip->i_lock);
	else if (lock_flags & XFS_ILOCK_SHARED)
		mrunlock_shared(&ip->i_lock);
L
Linus Torvalds 已提交
638

C
Christoph Hellwig 已提交
639 640
	if ((lock_flags & (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)) &&
	    !(lock_flags & XFS_IUNLOCK_NONOTIFY) && ip->i_itemp) {
L
Linus Torvalds 已提交
641 642 643 644 645
		/*
		 * Let the AIL know that this item has been unlocked in case
		 * it is in the AIL and anyone is waiting on it.  Don't do
		 * this if the caller has asked us not to.
		 */
646
		xfs_trans_unlocked_item(ip->i_itemp->ili_item.li_ailp,
C
Christoph Hellwig 已提交
647
					(xfs_log_item_t*)(ip->i_itemp));
L
Linus Torvalds 已提交
648
	}
C
Christoph Hellwig 已提交
649
	trace_xfs_iunlock(ip, lock_flags, _RET_IP_);
L
Linus Torvalds 已提交
650 651 652 653 654 655 656
}

/*
 * give up write locks.  the i/o lock cannot be held nested
 * if it is being demoted.
 */
void
C
Christoph Hellwig 已提交
657 658 659
xfs_ilock_demote(
	xfs_inode_t		*ip,
	uint			lock_flags)
L
Linus Torvalds 已提交
660 661 662 663
{
	ASSERT(lock_flags & (XFS_IOLOCK_EXCL|XFS_ILOCK_EXCL));
	ASSERT((lock_flags & ~(XFS_IOLOCK_EXCL|XFS_ILOCK_EXCL)) == 0);

C
Christoph Hellwig 已提交
664
	if (lock_flags & XFS_ILOCK_EXCL)
L
Linus Torvalds 已提交
665
		mrdemote(&ip->i_lock);
C
Christoph Hellwig 已提交
666
	if (lock_flags & XFS_IOLOCK_EXCL)
L
Linus Torvalds 已提交
667
		mrdemote(&ip->i_iolock);
C
Christoph Hellwig 已提交
668 669

	trace_xfs_ilock_demote(ip, lock_flags, _RET_IP_);
C
Christoph Hellwig 已提交
670 671 672 673 674 675 676 677
}

#ifdef DEBUG
int
xfs_isilocked(
	xfs_inode_t		*ip,
	uint			lock_flags)
{
C
Christoph Hellwig 已提交
678 679 680 681
	if (lock_flags & (XFS_ILOCK_EXCL|XFS_ILOCK_SHARED)) {
		if (!(lock_flags & XFS_ILOCK_SHARED))
			return !!ip->i_lock.mr_writer;
		return rwsem_is_locked(&ip->i_lock.mr_lock);
L
Linus Torvalds 已提交
682
	}
C
Christoph Hellwig 已提交
683

C
Christoph Hellwig 已提交
684 685 686 687
	if (lock_flags & (XFS_IOLOCK_EXCL|XFS_IOLOCK_SHARED)) {
		if (!(lock_flags & XFS_IOLOCK_SHARED))
			return !!ip->i_iolock.mr_writer;
		return rwsem_is_locked(&ip->i_iolock.mr_lock);
C
Christoph Hellwig 已提交
688 689
	}

C
Christoph Hellwig 已提交
690 691
	ASSERT(0);
	return 0;
L
Linus Torvalds 已提交
692
}
C
Christoph Hellwig 已提交
693
#endif