locks.c 73.3 KB
Newer Older
Linus Torvalds's avatar
Linus Torvalds committed
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62
/*
 *  linux/fs/locks.c
 *
 *  Provide support for fcntl()'s F_GETLK, F_SETLK, and F_SETLKW calls.
 *  Doug Evans (dje@spiff.uucp), August 07, 1992
 *
 *  Deadlock detection added.
 *  FIXME: one thing isn't handled yet:
 *	- mandatory locks (requires lots of changes elsewhere)
 *  Kelly Carmichael (kelly@[142.24.8.65]), September 17, 1994.
 *
 *  Miscellaneous edits, and a total rewrite of posix_lock_file() code.
 *  Kai Petzke (wpp@marie.physik.tu-berlin.de), 1994
 *  
 *  Converted file_lock_table to a linked list from an array, which eliminates
 *  the limits on how many active file locks are open.
 *  Chad Page (pageone@netcom.com), November 27, 1994
 * 
 *  Removed dependency on file descriptors. dup()'ed file descriptors now
 *  get the same locks as the original file descriptors, and a close() on
 *  any file descriptor removes ALL the locks on the file for the current
 *  process. Since locks still depend on the process id, locks are inherited
 *  after an exec() but not after a fork(). This agrees with POSIX, and both
 *  BSD and SVR4 practice.
 *  Andy Walker (andy@lysaker.kvaerner.no), February 14, 1995
 *
 *  Scrapped free list which is redundant now that we allocate locks
 *  dynamically with kmalloc()/kfree().
 *  Andy Walker (andy@lysaker.kvaerner.no), February 21, 1995
 *
 *  Implemented two lock personalities - FL_FLOCK and FL_POSIX.
 *
 *  FL_POSIX locks are created with calls to fcntl() and lockf() through the
 *  fcntl() system call. They have the semantics described above.
 *
 *  FL_FLOCK locks are created with calls to flock(), through the flock()
 *  system call, which is new. Old C libraries implement flock() via fcntl()
 *  and will continue to use the old, broken implementation.
 *
 *  FL_FLOCK locks follow the 4.4 BSD flock() semantics. They are associated
 *  with a file pointer (filp). As a result they can be shared by a parent
 *  process and its children after a fork(). They are removed when the last
 *  file descriptor referring to the file pointer is closed (unless explicitly
 *  unlocked). 
 *
 *  FL_FLOCK locks never deadlock, an existing lock is always removed before
 *  upgrading from shared to exclusive (or vice versa). When this happens
 *  any processes blocked by the current lock are woken up and allowed to
 *  run before the new lock is applied.
 *  Andy Walker (andy@lysaker.kvaerner.no), June 09, 1995
 *
 *  Removed some race conditions in flock_lock_file(), marked other possible
 *  races. Just grep for FIXME to see them. 
 *  Dmitry Gorodchanin (pgmdsg@ibi.com), February 09, 1996.
 *
 *  Addressed Dmitry's concerns. Deadlock checking no longer recursive.
 *  Lock allocation changed to GFP_ATOMIC as we can't afford to sleep
 *  once we've checked for blocking and deadlocking.
 *  Andy Walker (andy@lysaker.kvaerner.no), April 03, 1996.
 *
 *  Initial implementation of mandatory locks. SunOS turned out to be
 *  a rotten model, so I implemented the "obvious" semantics.
Paul Bolle's avatar
Paul Bolle committed
63
 *  See 'Documentation/filesystems/mandatory-locking.txt' for details.
Linus Torvalds's avatar
Linus Torvalds committed
64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118
 *  Andy Walker (andy@lysaker.kvaerner.no), April 06, 1996.
 *
 *  Don't allow mandatory locks on mmap()'ed files. Added simple functions to
 *  check if a file has mandatory locks, used by mmap(), open() and creat() to
 *  see if system call should be rejected. Ref. HP-UX/SunOS/Solaris Reference
 *  Manual, Section 2.
 *  Andy Walker (andy@lysaker.kvaerner.no), April 09, 1996.
 *
 *  Tidied up block list handling. Added '/proc/locks' interface.
 *  Andy Walker (andy@lysaker.kvaerner.no), April 24, 1996.
 *
 *  Fixed deadlock condition for pathological code that mixes calls to
 *  flock() and fcntl().
 *  Andy Walker (andy@lysaker.kvaerner.no), April 29, 1996.
 *
 *  Allow only one type of locking scheme (FL_POSIX or FL_FLOCK) to be in use
 *  for a given file at a time. Changed the CONFIG_LOCK_MANDATORY scheme to
 *  guarantee sensible behaviour in the case where file system modules might
 *  be compiled with different options than the kernel itself.
 *  Andy Walker (andy@lysaker.kvaerner.no), May 15, 1996.
 *
 *  Added a couple of missing wake_up() calls. Thanks to Thomas Meckel
 *  (Thomas.Meckel@mni.fh-giessen.de) for spotting this.
 *  Andy Walker (andy@lysaker.kvaerner.no), May 15, 1996.
 *
 *  Changed FL_POSIX locks to use the block list in the same way as FL_FLOCK
 *  locks. Changed process synchronisation to avoid dereferencing locks that
 *  have already been freed.
 *  Andy Walker (andy@lysaker.kvaerner.no), Sep 21, 1996.
 *
 *  Made the block list a circular list to minimise searching in the list.
 *  Andy Walker (andy@lysaker.kvaerner.no), Sep 25, 1996.
 *
 *  Made mandatory locking a mount option. Default is not to allow mandatory
 *  locking.
 *  Andy Walker (andy@lysaker.kvaerner.no), Oct 04, 1996.
 *
 *  Some adaptations for NFS support.
 *  Olaf Kirch (okir@monad.swb.de), Dec 1996,
 *
 *  Fixed /proc/locks interface so that we can't overrun the buffer we are handed.
 *  Andy Walker (andy@lysaker.kvaerner.no), May 12, 1997.
 *
 *  Use slab allocator instead of kmalloc/kfree.
 *  Use generic list implementation from <linux/list.h>.
 *  Sped up posix_locks_deadlock by only considering blocked locks.
 *  Matthew Wilcox <willy@debian.org>, March, 2000.
 *
 *  Leases and LOCK_MAND
 *  Matthew Wilcox <willy@debian.org>, June, 2000.
 *  Stephen Rothwell <sfr@canb.auug.org.au>, June, 2000.
 */

#include <linux/capability.h>
#include <linux/file.h>
Al Viro's avatar
Al Viro committed
119
#include <linux/fdtable.h>
Linus Torvalds's avatar
Linus Torvalds committed
120 121 122 123 124 125
#include <linux/fs.h>
#include <linux/init.h>
#include <linux/security.h>
#include <linux/slab.h>
#include <linux/syscalls.h>
#include <linux/time.h>
126
#include <linux/rcupdate.h>
127
#include <linux/pid_namespace.h>
128
#include <linux/hashtable.h>
129
#include <linux/percpu.h>
Linus Torvalds's avatar
Linus Torvalds committed
130

131 132 133
#define CREATE_TRACE_POINTS
#include <trace/events/filelock.h>

Linus Torvalds's avatar
Linus Torvalds committed
134 135 136 137
#include <asm/uaccess.h>

#define IS_POSIX(fl)	(fl->fl_flags & FL_POSIX)
#define IS_FLOCK(fl)	(fl->fl_flags & FL_FLOCK)
138
#define IS_LEASE(fl)	(fl->fl_flags & (FL_LEASE|FL_DELEG|FL_LAYOUT))
139
#define IS_OFDLCK(fl)	(fl->fl_flags & FL_OFDLCK)
Linus Torvalds's avatar
Linus Torvalds committed
140

141 142 143 144 145
static inline bool is_remote_lock(struct file *filp)
{
	return likely(!(filp->f_path.dentry->d_sb->s_flags & MS_NOREMOTELOCK));
}

J. Bruce Fields's avatar
J. Bruce Fields committed
146 147
static bool lease_breaking(struct file_lock *fl)
{
148 149 150 151 152 153 154 155 156 157
	return fl->fl_flags & (FL_UNLOCK_PENDING | FL_DOWNGRADE_PENDING);
}

static int target_leasetype(struct file_lock *fl)
{
	if (fl->fl_flags & FL_UNLOCK_PENDING)
		return F_UNLCK;
	if (fl->fl_flags & FL_DOWNGRADE_PENDING)
		return F_RDLCK;
	return fl->fl_type;
J. Bruce Fields's avatar
J. Bruce Fields committed
158 159
}

Linus Torvalds's avatar
Linus Torvalds committed
160 161 162
int leases_enable = 1;
int lease_break_time = 45;

163
/*
164
 * The global file_lock_list is only used for displaying /proc/locks, so we
165 166 167 168 169
 * keep a list on each CPU, with each list protected by its own spinlock.
 * Global serialization is done using file_rwsem.
 *
 * Note that alterations to the list also require that the relevant flc_lock is
 * held.
170
 */
171 172 173 174 175
struct file_lock_list_struct {
	spinlock_t		lock;
	struct hlist_head	hlist;
};
static DEFINE_PER_CPU(struct file_lock_list_struct, file_lock_list);
176
DEFINE_STATIC_PERCPU_RWSEM(file_rwsem);
177

178
/*
179
 * The blocked_hash is used to find POSIX lock loops for deadlock detection.
180
 * It is protected by blocked_lock_lock.
181 182 183 184 185 186 187
 *
 * We hash locks by lockowner in order to optimize searching for the lock a
 * particular lockowner is waiting on.
 *
 * FIXME: make this value scale via some heuristic? We generally will want more
 * buckets when we have more lockowners holding locks, but that's a little
 * difficult to determine without knowing what the workload will look like.
188
 */
189 190
#define BLOCKED_HASH_BITS	7
static DEFINE_HASHTABLE(blocked_hash, BLOCKED_HASH_BITS);
191

192
/*
193 194
 * This lock protects the blocked_hash. Generally, if you're accessing it, you
 * want to be holding this lock.
195 196 197 198 199 200
 *
 * In addition, it also protects the fl->fl_block list, and the fl->fl_next
 * pointer for file_lock structures that are acting as lock requests (in
 * contrast to those that are acting as records of acquired locks).
 *
 * Note that when we acquire this lock in order to change the above fields,
201
 * we often hold the flc_lock as well. In certain cases, when reading the fields
202
 * protected by this lock, we can skip acquiring it iff we already hold the
203
 * flc_lock.
204 205
 *
 * In particular, adding an entry to the fl_block list requires that you hold
206 207
 * both the flc_lock and the blocked_lock_lock (acquired in that order).
 * Deleting an entry from the list however only requires the file_lock_lock.
208
 */
209
static DEFINE_SPINLOCK(blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
210

211
static struct kmem_cache *flctx_cache __read_mostly;
212
static struct kmem_cache *filelock_cache __read_mostly;
Linus Torvalds's avatar
Linus Torvalds committed
213

214
static struct file_lock_context *
215
locks_get_lock_context(struct inode *inode, int type)
216
{
217
	struct file_lock_context *ctx;
218

219 220 221
	/* paired with cmpxchg() below */
	ctx = smp_load_acquire(&inode->i_flctx);
	if (likely(ctx) || type == F_UNLCK)
222 223
		goto out;

224 225
	ctx = kmem_cache_alloc(flctx_cache, GFP_KERNEL);
	if (!ctx)
226 227
		goto out;

228 229 230 231
	spin_lock_init(&ctx->flc_lock);
	INIT_LIST_HEAD(&ctx->flc_flock);
	INIT_LIST_HEAD(&ctx->flc_posix);
	INIT_LIST_HEAD(&ctx->flc_lease);
232 233 234 235 236

	/*
	 * Assign the pointer if it's not already assigned. If it is, then
	 * free the context we just allocated.
	 */
237 238 239 240
	if (cmpxchg(&inode->i_flctx, NULL, ctx)) {
		kmem_cache_free(flctx_cache, ctx);
		ctx = smp_load_acquire(&inode->i_flctx);
	}
241
out:
242
	trace_locks_get_lock_context(inode, type, ctx);
243
	return ctx;
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
static void
locks_dump_ctx_list(struct list_head *list, char *list_type)
{
	struct file_lock *fl;

	list_for_each_entry(fl, list, fl_list) {
		pr_warn("%s: fl_owner=%p fl_flags=0x%x fl_type=0x%x fl_pid=%u\n", list_type, fl->fl_owner, fl->fl_flags, fl->fl_type, fl->fl_pid);
	}
}

static void
locks_check_ctx_lists(struct inode *inode)
{
	struct file_lock_context *ctx = inode->i_flctx;

	if (unlikely(!list_empty(&ctx->flc_flock) ||
		     !list_empty(&ctx->flc_posix) ||
		     !list_empty(&ctx->flc_lease))) {
		pr_warn("Leaked locks on dev=0x%x:0x%x ino=0x%lx:\n",
			MAJOR(inode->i_sb->s_dev), MINOR(inode->i_sb->s_dev),
			inode->i_ino);
		locks_dump_ctx_list(&ctx->flc_flock, "FLOCK");
		locks_dump_ctx_list(&ctx->flc_posix, "POSIX");
		locks_dump_ctx_list(&ctx->flc_lease, "LEASE");
	}
}

273
void
274
locks_free_lock_context(struct inode *inode)
275
{
276 277
	struct file_lock_context *ctx = inode->i_flctx;

278 279
	if (unlikely(ctx)) {
		locks_check_ctx_lists(inode);
280 281 282 283
		kmem_cache_free(flctx_cache, ctx);
	}
}

Miklos Szeredi's avatar
Miklos Szeredi committed
284
static void locks_init_lock_heads(struct file_lock *fl)
Miklos Szeredi's avatar
Miklos Szeredi committed
285
{
286
	INIT_HLIST_NODE(&fl->fl_link);
287
	INIT_LIST_HEAD(&fl->fl_list);
Miklos Szeredi's avatar
Miklos Szeredi committed
288 289
	INIT_LIST_HEAD(&fl->fl_block);
	init_waitqueue_head(&fl->fl_wait);
Miklos Szeredi's avatar
Miklos Szeredi committed
290 291
}

Linus Torvalds's avatar
Linus Torvalds committed
292
/* Allocate an empty lock structure. */
293
struct file_lock *locks_alloc_lock(void)
Linus Torvalds's avatar
Linus Torvalds committed
294
{
Miklos Szeredi's avatar
Miklos Szeredi committed
295
	struct file_lock *fl = kmem_cache_zalloc(filelock_cache, GFP_KERNEL);
Miklos Szeredi's avatar
Miklos Szeredi committed
296 297

	if (fl)
Miklos Szeredi's avatar
Miklos Szeredi committed
298
		locks_init_lock_heads(fl);
Miklos Szeredi's avatar
Miklos Szeredi committed
299 300

	return fl;
Linus Torvalds's avatar
Linus Torvalds committed
301
}
302
EXPORT_SYMBOL_GPL(locks_alloc_lock);
Linus Torvalds's avatar
Linus Torvalds committed
303

304
void locks_release_private(struct file_lock *fl)
305 306 307 308 309 310 311
{
	if (fl->fl_ops) {
		if (fl->fl_ops->fl_release_private)
			fl->fl_ops->fl_release_private(fl);
		fl->fl_ops = NULL;
	}

312
	if (fl->fl_lmops) {
313 314 315 316
		if (fl->fl_lmops->lm_put_owner) {
			fl->fl_lmops->lm_put_owner(fl->fl_owner);
			fl->fl_owner = NULL;
		}
317 318
		fl->fl_lmops = NULL;
	}
319
}
320
EXPORT_SYMBOL_GPL(locks_release_private);
321

Linus Torvalds's avatar
Linus Torvalds committed
322
/* Free a lock which is not in use. */
323
void locks_free_lock(struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
324
{
Miklos Szeredi's avatar
Miklos Szeredi committed
325
	BUG_ON(waitqueue_active(&fl->fl_wait));
326
	BUG_ON(!list_empty(&fl->fl_list));
Miklos Szeredi's avatar
Miklos Szeredi committed
327
	BUG_ON(!list_empty(&fl->fl_block));
328
	BUG_ON(!hlist_unhashed(&fl->fl_link));
Linus Torvalds's avatar
Linus Torvalds committed
329

330
	locks_release_private(fl);
Linus Torvalds's avatar
Linus Torvalds committed
331 332
	kmem_cache_free(filelock_cache, fl);
}
333
EXPORT_SYMBOL(locks_free_lock);
Linus Torvalds's avatar
Linus Torvalds committed
334

335 336 337 338 339 340
static void
locks_dispose_list(struct list_head *dispose)
{
	struct file_lock *fl;

	while (!list_empty(dispose)) {
341 342
		fl = list_first_entry(dispose, struct file_lock, fl_list);
		list_del_init(&fl->fl_list);
343 344 345 346
		locks_free_lock(fl);
	}
}

Linus Torvalds's avatar
Linus Torvalds committed
347 348
void locks_init_lock(struct file_lock *fl)
{
Miklos Szeredi's avatar
Miklos Szeredi committed
349 350
	memset(fl, 0, sizeof(struct file_lock));
	locks_init_lock_heads(fl);
Linus Torvalds's avatar
Linus Torvalds committed
351 352 353 354 355 356 357
}

EXPORT_SYMBOL(locks_init_lock);

/*
 * Initialize a new lock from an existing file_lock structure.
 */
358
void locks_copy_conflock(struct file_lock *new, struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
359 360 361
{
	new->fl_owner = fl->fl_owner;
	new->fl_pid = fl->fl_pid;
362
	new->fl_file = NULL;
Linus Torvalds's avatar
Linus Torvalds committed
363 364 365 366
	new->fl_flags = fl->fl_flags;
	new->fl_type = fl->fl_type;
	new->fl_start = fl->fl_start;
	new->fl_end = fl->fl_end;
367
	new->fl_lmops = fl->fl_lmops;
368
	new->fl_ops = NULL;
369 370 371

	if (fl->fl_lmops) {
		if (fl->fl_lmops->lm_get_owner)
372
			fl->fl_lmops->lm_get_owner(fl->fl_owner);
373
	}
374
}
375
EXPORT_SYMBOL(locks_copy_conflock);
376 377 378

void locks_copy_lock(struct file_lock *new, struct file_lock *fl)
{
379 380
	/* "new" must be a freshly-initialized lock */
	WARN_ON_ONCE(new->fl_ops);
381

382
	locks_copy_conflock(new, fl);
383

384
	new->fl_file = fl->fl_file;
Linus Torvalds's avatar
Linus Torvalds committed
385
	new->fl_ops = fl->fl_ops;
386

387 388 389 390
	if (fl->fl_ops) {
		if (fl->fl_ops->fl_copy_lock)
			fl->fl_ops->fl_copy_lock(new, fl);
	}
Linus Torvalds's avatar
Linus Torvalds committed
391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409
}

EXPORT_SYMBOL(locks_copy_lock);

static inline int flock_translate_cmd(int cmd) {
	if (cmd & LOCK_MAND)
		return cmd & (LOCK_MAND | LOCK_RW);
	switch (cmd) {
	case LOCK_SH:
		return F_RDLCK;
	case LOCK_EX:
		return F_WRLCK;
	case LOCK_UN:
		return F_UNLCK;
	}
	return -EINVAL;
}

/* Fill in a file_lock structure with an appropriate FLOCK lock. */
410 411
static struct file_lock *
flock_make_lock(struct file *filp, unsigned int cmd)
Linus Torvalds's avatar
Linus Torvalds committed
412 413 414
{
	struct file_lock *fl;
	int type = flock_translate_cmd(cmd);
415

Linus Torvalds's avatar
Linus Torvalds committed
416
	if (type < 0)
417
		return ERR_PTR(type);
Linus Torvalds's avatar
Linus Torvalds committed
418 419 420
	
	fl = locks_alloc_lock();
	if (fl == NULL)
421
		return ERR_PTR(-ENOMEM);
Linus Torvalds's avatar
Linus Torvalds committed
422 423

	fl->fl_file = filp;
424
	fl->fl_owner = filp;
Linus Torvalds's avatar
Linus Torvalds committed
425 426 427 428 429
	fl->fl_pid = current->tgid;
	fl->fl_flags = FL_FLOCK;
	fl->fl_type = type;
	fl->fl_end = OFFSET_MAX;
	
430
	return fl;
Linus Torvalds's avatar
Linus Torvalds committed
431 432
}

433
static int assign_type(struct file_lock *fl, long type)
Linus Torvalds's avatar
Linus Torvalds committed
434 435 436 437 438 439 440 441 442 443 444 445 446
{
	switch (type) {
	case F_RDLCK:
	case F_WRLCK:
	case F_UNLCK:
		fl->fl_type = type;
		break;
	default:
		return -EINVAL;
	}
	return 0;
}

447 448
static int flock64_to_posix_lock(struct file *filp, struct file_lock *fl,
				 struct flock64 *l)
Linus Torvalds's avatar
Linus Torvalds committed
449 450
{
	switch (l->l_whence) {
451
	case SEEK_SET:
452
		fl->fl_start = 0;
Linus Torvalds's avatar
Linus Torvalds committed
453
		break;
454
	case SEEK_CUR:
455
		fl->fl_start = filp->f_pos;
Linus Torvalds's avatar
Linus Torvalds committed
456
		break;
457
	case SEEK_END:
458
		fl->fl_start = i_size_read(file_inode(filp));
Linus Torvalds's avatar
Linus Torvalds committed
459 460 461 462
		break;
	default:
		return -EINVAL;
	}
463 464 465 466 467
	if (l->l_start > OFFSET_MAX - fl->fl_start)
		return -EOVERFLOW;
	fl->fl_start += l->l_start;
	if (fl->fl_start < 0)
		return -EINVAL;
Linus Torvalds's avatar
Linus Torvalds committed
468 469 470

	/* POSIX-1996 leaves the case l->l_len < 0 undefined;
	   POSIX-2001 defines it. */
471
	if (l->l_len > 0) {
472 473 474 475
		if (l->l_len - 1 > OFFSET_MAX - fl->fl_start)
			return -EOVERFLOW;
		fl->fl_end = fl->fl_start + l->l_len - 1;

476
	} else if (l->l_len < 0) {
477
		if (fl->fl_start + l->l_len < 0)
478
			return -EINVAL;
479 480 481 482 483
		fl->fl_end = fl->fl_start - 1;
		fl->fl_start += l->l_len;
	} else
		fl->fl_end = OFFSET_MAX;

Linus Torvalds's avatar
Linus Torvalds committed
484 485 486 487 488 489 490 491 492 493
	fl->fl_owner = current->files;
	fl->fl_pid = current->tgid;
	fl->fl_file = filp;
	fl->fl_flags = FL_POSIX;
	fl->fl_ops = NULL;
	fl->fl_lmops = NULL;

	return assign_type(fl, l->l_type);
}

494 495 496 497 498
/* Verify a "struct flock" and copy it to a "struct file_lock" as a POSIX
 * style lock.
 */
static int flock_to_posix_lock(struct file *filp, struct file_lock *fl,
			       struct flock *l)
Linus Torvalds's avatar
Linus Torvalds committed
499
{
500 501 502 503 504 505 506 507
	struct flock64 ll = {
		.l_type = l->l_type,
		.l_whence = l->l_whence,
		.l_start = l->l_start,
		.l_len = l->l_len,
	};

	return flock64_to_posix_lock(filp, fl, &ll);
Linus Torvalds's avatar
Linus Torvalds committed
508 509 510
}

/* default lease lock manager operations */
511 512
static bool
lease_break_callback(struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
513 514
{
	kill_fasync(&fl->fl_fasync, SIGIO, POLL_MSG);
515
	return false;
Linus Torvalds's avatar
Linus Torvalds committed
516 517
}

518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534
static void
lease_setup(struct file_lock *fl, void **priv)
{
	struct file *filp = fl->fl_file;
	struct fasync_struct *fa = *priv;

	/*
	 * fasync_insert_entry() returns the old entry if any. If there was no
	 * old entry, then it used "priv" and inserted it into the fasync list.
	 * Clear the pointer to indicate that it shouldn't be freed.
	 */
	if (!fasync_insert_entry(fa->fa_fd, filp, &fl->fl_fasync, fa))
		*priv = NULL;

	__f_setown(filp, task_pid(current), PIDTYPE_PID, 0);
}

535
static const struct lock_manager_operations lease_manager_ops = {
536 537
	.lm_break = lease_break_callback,
	.lm_change = lease_modify,
538
	.lm_setup = lease_setup,
Linus Torvalds's avatar
Linus Torvalds committed
539 540 541 542 543
};

/*
 * Initialize a lease, use the default lock manager operations
 */
544
static int lease_init(struct file *filp, long type, struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
545
 {
546 547 548
	if (assign_type(fl, type) != 0)
		return -EINVAL;

549
	fl->fl_owner = filp;
Linus Torvalds's avatar
Linus Torvalds committed
550 551 552 553 554 555 556 557 558 559 560 561
	fl->fl_pid = current->tgid;

	fl->fl_file = filp;
	fl->fl_flags = FL_LEASE;
	fl->fl_start = 0;
	fl->fl_end = OFFSET_MAX;
	fl->fl_ops = NULL;
	fl->fl_lmops = &lease_manager_ops;
	return 0;
}

/* Allocate a file_lock initialised to this type of lease */
562
static struct file_lock *lease_alloc(struct file *filp, long type)
Linus Torvalds's avatar
Linus Torvalds committed
563 564
{
	struct file_lock *fl = locks_alloc_lock();
565
	int error = -ENOMEM;
Linus Torvalds's avatar
Linus Torvalds committed
566 567

	if (fl == NULL)
568
		return ERR_PTR(error);
Linus Torvalds's avatar
Linus Torvalds committed
569 570

	error = lease_init(filp, type, fl);
571 572
	if (error) {
		locks_free_lock(fl);
573
		return ERR_PTR(error);
574
	}
575
	return fl;
Linus Torvalds's avatar
Linus Torvalds committed
576 577 578 579 580 581 582 583 584 585 586 587 588
}

/* Check if two locks overlap each other.
 */
static inline int locks_overlap(struct file_lock *fl1, struct file_lock *fl2)
{
	return ((fl1->fl_end >= fl2->fl_start) &&
		(fl2->fl_end >= fl1->fl_start));
}

/*
 * Check whether two locks have the same owner.
 */
589
static int posix_same_owner(struct file_lock *fl1, struct file_lock *fl2)
Linus Torvalds's avatar
Linus Torvalds committed
590
{
591
	if (fl1->fl_lmops && fl1->fl_lmops->lm_compare_owner)
Linus Torvalds's avatar
Linus Torvalds committed
592
		return fl2->fl_lmops == fl1->fl_lmops &&
593
			fl1->fl_lmops->lm_compare_owner(fl1, fl2);
Linus Torvalds's avatar
Linus Torvalds committed
594 595 596
	return fl1->fl_owner == fl2->fl_owner;
}

597
/* Must be called with the flc_lock held! */
598
static void locks_insert_global_locks(struct file_lock *fl)
599
{
600 601
	struct file_lock_list_struct *fll = this_cpu_ptr(&file_lock_list);

602 603
	percpu_rwsem_assert_held(&file_rwsem);

604
	spin_lock(&fll->lock);
605
	fl->fl_link_cpu = smp_processor_id();
606 607
	hlist_add_head(&fl->fl_link, &fll->hlist);
	spin_unlock(&fll->lock);
608 609
}

610
/* Must be called with the flc_lock held! */
611
static void locks_delete_global_locks(struct file_lock *fl)
612
{
613 614
	struct file_lock_list_struct *fll;

615 616
	percpu_rwsem_assert_held(&file_rwsem);

617 618
	/*
	 * Avoid taking lock if already unhashed. This is safe since this check
619
	 * is done while holding the flc_lock, and new insertions into the list
620 621 622 623
	 * also require that it be held.
	 */
	if (hlist_unhashed(&fl->fl_link))
		return;
624 625 626

	fll = per_cpu_ptr(&file_lock_list, fl->fl_link_cpu);
	spin_lock(&fll->lock);
627
	hlist_del_init(&fl->fl_link);
628
	spin_unlock(&fll->lock);
629 630
}

631 632 633 634 635 636 637 638
static unsigned long
posix_owner_key(struct file_lock *fl)
{
	if (fl->fl_lmops && fl->fl_lmops->lm_owner_key)
		return fl->fl_lmops->lm_owner_key(fl);
	return (unsigned long)fl->fl_owner;
}

639
static void locks_insert_global_blocked(struct file_lock *waiter)
640
{
641 642
	lockdep_assert_held(&blocked_lock_lock);

643
	hash_add(blocked_hash, &waiter->fl_link, posix_owner_key(waiter));
644 645
}

646
static void locks_delete_global_blocked(struct file_lock *waiter)
647
{
648 649
	lockdep_assert_held(&blocked_lock_lock);

650
	hash_del(&waiter->fl_link);
651 652
}

Linus Torvalds's avatar
Linus Torvalds committed
653 654
/* Remove waiter from blocker's block list.
 * When blocker ends up pointing to itself then the list is empty.
655
 *
656
 * Must be called with blocked_lock_lock held.
Linus Torvalds's avatar
Linus Torvalds committed
657
 */
658
static void __locks_delete_block(struct file_lock *waiter)
Linus Torvalds's avatar
Linus Torvalds committed
659
{
660
	locks_delete_global_blocked(waiter);
Linus Torvalds's avatar
Linus Torvalds committed
661 662 663 664
	list_del_init(&waiter->fl_block);
	waiter->fl_next = NULL;
}

665
static void locks_delete_block(struct file_lock *waiter)
Linus Torvalds's avatar
Linus Torvalds committed
666
{
667
	spin_lock(&blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
668
	__locks_delete_block(waiter);
669
	spin_unlock(&blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
670 671 672 673 674 675
}

/* Insert waiter into blocker's block list.
 * We use a circular list so that processes can be easily woken up in
 * the order they blocked. The documentation doesn't require this but
 * it seems like the reasonable thing to do.
676
 *
677 678 679 680
 * Must be called with both the flc_lock and blocked_lock_lock held. The
 * fl_block list itself is protected by the blocked_lock_lock, but by ensuring
 * that the flc_lock is also held on insertions we can avoid taking the
 * blocked_lock_lock in some cases when we see that the fl_block list is empty.
Linus Torvalds's avatar
Linus Torvalds committed
681
 */
682 683
static void __locks_insert_block(struct file_lock *blocker,
					struct file_lock *waiter)
Linus Torvalds's avatar
Linus Torvalds committed
684
{
685
	BUG_ON(!list_empty(&waiter->fl_block));
Linus Torvalds's avatar
Linus Torvalds committed
686
	waiter->fl_next = blocker;
687
	list_add_tail(&waiter->fl_block, &blocker->fl_block);
688
	if (IS_POSIX(blocker) && !IS_OFDLCK(blocker))
689 690 691
		locks_insert_global_blocked(waiter);
}

692
/* Must be called with flc_lock held. */
693 694 695
static void locks_insert_block(struct file_lock *blocker,
					struct file_lock *waiter)
{
696
	spin_lock(&blocked_lock_lock);
697
	__locks_insert_block(blocker, waiter);
698
	spin_unlock(&blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
699 700
}

701 702 703
/*
 * Wake up processes blocked waiting for blocker.
 *
704
 * Must be called with the inode->flc_lock held!
Linus Torvalds's avatar
Linus Torvalds committed
705 706 707
 */
static void locks_wake_up_blocks(struct file_lock *blocker)
{
708 709
	/*
	 * Avoid taking global lock if list is empty. This is safe since new
710 711 712
	 * blocked requests are only added to the list under the flc_lock, and
	 * the flc_lock is always held here. Note that removal from the fl_block
	 * list does not require the flc_lock, so we must recheck list_empty()
713
	 * after acquiring the blocked_lock_lock.
714 715 716 717
	 */
	if (list_empty(&blocker->fl_block))
		return;

718
	spin_lock(&blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
719
	while (!list_empty(&blocker->fl_block)) {
720 721 722
		struct file_lock *waiter;

		waiter = list_first_entry(&blocker->fl_block,
Linus Torvalds's avatar
Linus Torvalds committed
723 724
				struct file_lock, fl_block);
		__locks_delete_block(waiter);
725 726
		if (waiter->fl_lmops && waiter->fl_lmops->lm_notify)
			waiter->fl_lmops->lm_notify(waiter);
Linus Torvalds's avatar
Linus Torvalds committed
727 728 729
		else
			wake_up(&waiter->fl_wait);
	}
730
	spin_unlock(&blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
731 732
}

733
static void
734
locks_insert_lock_ctx(struct file_lock *fl, struct list_head *before)
735 736 737 738 739 740
{
	fl->fl_nspid = get_pid(task_tgid(current));
	list_add_tail(&fl->fl_list, before);
	locks_insert_global_locks(fl);
}

741
static void
742
locks_unlink_lock_ctx(struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
743
{
744
	locks_delete_global_locks(fl);
745
	list_del_init(&fl->fl_list);
746 747 748 749
	if (fl->fl_nspid) {
		put_pid(fl->fl_nspid);
		fl->fl_nspid = NULL;
	}
Linus Torvalds's avatar
Linus Torvalds committed
750
	locks_wake_up_blocks(fl);
751 752
}

753
static void
754
locks_delete_lock_ctx(struct file_lock *fl, struct list_head *dispose)
755
{
756
	locks_unlink_lock_ctx(fl);
757
	if (dispose)
758
		list_add(&fl->fl_list, dispose);
759 760
	else
		locks_free_lock(fl);
Linus Torvalds's avatar
Linus Torvalds committed
761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782
}

/* Determine if lock sys_fl blocks lock caller_fl. Common functionality
 * checks for shared/exclusive status of overlapping locks.
 */
static int locks_conflict(struct file_lock *caller_fl, struct file_lock *sys_fl)
{
	if (sys_fl->fl_type == F_WRLCK)
		return 1;
	if (caller_fl->fl_type == F_WRLCK)
		return 1;
	return 0;
}

/* Determine if lock sys_fl blocks lock caller_fl. POSIX specific
 * checking before calling the locks_conflict().
 */
static int posix_locks_conflict(struct file_lock *caller_fl, struct file_lock *sys_fl)
{
	/* POSIX locks owned by the same process do not conflict with
	 * each other.
	 */
783
	if (posix_same_owner(caller_fl, sys_fl))
Linus Torvalds's avatar
Linus Torvalds committed
784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800
		return (0);

	/* Check whether they overlap */
	if (!locks_overlap(caller_fl, sys_fl))
		return 0;

	return (locks_conflict(caller_fl, sys_fl));
}

/* Determine if lock sys_fl blocks lock caller_fl. FLOCK specific
 * checking before calling the locks_conflict().
 */
static int flock_locks_conflict(struct file_lock *caller_fl, struct file_lock *sys_fl)
{
	/* FLOCK locks referring to the same filp do not conflict with
	 * each other.
	 */
801
	if (caller_fl->fl_file == sys_fl->fl_file)
Linus Torvalds's avatar
Linus Torvalds committed
802 803 804 805 806 807 808
		return (0);
	if ((caller_fl->fl_type & LOCK_MAND) || (sys_fl->fl_type & LOCK_MAND))
		return 0;

	return (locks_conflict(caller_fl, sys_fl));
}

809
void
810
posix_test_lock(struct file *filp, struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
811 812
{
	struct file_lock *cfl;
813
	struct file_lock_context *ctx;
814
	struct inode *inode = locks_inode(filp);
Linus Torvalds's avatar
Linus Torvalds committed
815

816
	ctx = smp_load_acquire(&inode->i_flctx);
817 818 819 820 821
	if (!ctx || list_empty_careful(&ctx->flc_posix)) {
		fl->fl_type = F_UNLCK;
		return;
	}

822
	spin_lock(&ctx->flc_lock);
823 824 825 826 827 828 829
	list_for_each_entry(cfl, &ctx->flc_posix, fl_list) {
		if (posix_locks_conflict(fl, cfl)) {
			locks_copy_conflock(fl, cfl);
			if (cfl->fl_nspid)
				fl->fl_pid = pid_vnr(cfl->fl_nspid);
			goto out;
		}
Linus Torvalds's avatar
Linus Torvalds committed
830
	}
831 832
	fl->fl_type = F_UNLCK;
out:
833
	spin_unlock(&ctx->flc_lock);
834
	return;
Linus Torvalds's avatar
Linus Torvalds committed
835 836 837
}
EXPORT_SYMBOL(posix_test_lock);

838 839 840 841 842
/*
 * Deadlock detection:
 *
 * We attempt to detect deadlocks that are due purely to posix file
 * locks.
Linus Torvalds's avatar
Linus Torvalds committed
843
 *
844 845 846 847 848 849 850
 * We assume that a task can be waiting for at most one lock at a time.
 * So for any acquired lock, the process holding that lock may be
 * waiting on at most one other lock.  That lock in turns may be held by
 * someone waiting for at most one other lock.  Given a requested lock
 * caller_fl which is about to wait for a conflicting lock block_fl, we
 * follow this chain of waiters to ensure we are not about to create a
 * cycle.
Linus Torvalds's avatar
Linus Torvalds committed
851
 *
852 853 854
 * Since we do this before we ever put a process to sleep on a lock, we
 * are ensured that there is never a cycle; that is what guarantees that
 * the while() loop in posix_locks_deadlock() eventually completes.
855
 *
856 857 858 859
 * Note: the above assumption may not be true when handling lock
 * requests from a broken NFS client. It may also fail in the presence
 * of tasks (such as posix threads) sharing the same open file table.
 * To handle those cases, we just bail out after a few iterations.
860
 *
861
 * For FL_OFDLCK locks, the owner is the filp, not the files_struct.
862 863 864 865
 * Because the owner is not even nominally tied to a thread of
 * execution, the deadlock detection below can't reasonably work well. Just
 * skip it for those.
 *
866
 * In principle, we could do a more limited deadlock detection on FL_OFDLCK
867 868
 * locks that just checks for the case where two tasks are attempting to
 * upgrade from read to write locks on the same inode.
Linus Torvalds's avatar
Linus Torvalds committed
869
 */
870 871 872

#define MAX_DEADLK_ITERATIONS 10

873 874 875 876 877
/* Find a lock that the owner of the given block_fl is blocking on. */
static struct file_lock *what_owner_is_waiting_for(struct file_lock *block_fl)
{
	struct file_lock *fl;

878
	hash_for_each_possible(blocked_hash, fl, fl_link, posix_owner_key(block_fl)) {
879 880 881 882 883 884
		if (posix_same_owner(fl, block_fl))
			return fl->fl_next;
	}
	return NULL;
}

885
/* Must be called with the blocked_lock_lock held! */
886
static int posix_locks_deadlock(struct file_lock *caller_fl,
Linus Torvalds's avatar
Linus Torvalds committed
887 888
				struct file_lock *block_fl)
{
889
	int i = 0;
Linus Torvalds's avatar
Linus Torvalds committed
890

891 892
	lockdep_assert_held(&blocked_lock_lock);

893 894
	/*
	 * This deadlock detector can't reasonably detect deadlocks with
895
	 * FL_OFDLCK locks, since they aren't owned by a process, per-se.
896
	 */
897
	if (IS_OFDLCK(caller_fl))
898 899
		return 0;

900 901 902 903 904
	while ((block_fl = what_owner_is_waiting_for(block_fl))) {
		if (i++ > MAX_DEADLK_ITERATIONS)
			return 0;
		if (posix_same_owner(caller_fl, block_fl))
			return 1;
Linus Torvalds's avatar
Linus Torvalds committed
905 906 907 908 909
	}
	return 0;
}

/* Try to create a FLOCK lock on filp. We always insert new FLOCK locks
910
 * after any leases, but before any posix locks.
911 912 913 914
 *
 * Note that if called with an FL_EXISTS argument, the caller may determine
 * whether or not a lock was successfully freed by testing the return
 * value for -ENOENT.
Linus Torvalds's avatar
Linus Torvalds committed
915
 */
916
static int flock_lock_inode(struct inode *inode, struct file_lock *request)
Linus Torvalds's avatar
Linus Torvalds committed
917
{
918
	struct file_lock *new_fl = NULL;
919 920
	struct file_lock *fl;
	struct file_lock_context *ctx;
Linus Torvalds's avatar
Linus Torvalds committed
921
	int error = 0;
922
	bool found = false;
923
	LIST_HEAD(dispose);
Linus Torvalds's avatar
Linus Torvalds committed
924

925 926 927 928 929 930
	ctx = locks_get_lock_context(inode, request->fl_type);
	if (!ctx) {
		if (request->fl_type != F_UNLCK)
			return -ENOMEM;
		return (request->fl_flags & FL_EXISTS) ? -ENOENT : 0;
	}
931

932
	if (!(request->fl_flags & FL_ACCESS) && (request->fl_type != F_UNLCK)) {
933
		new_fl = locks_alloc_lock();
934 935
		if (!new_fl)
			return -ENOMEM;
936 937
	}

938
	percpu_down_read_preempt_disable(&file_rwsem);
939
	spin_lock(&ctx->flc_lock);
940 941 942
	if (request->fl_flags & FL_ACCESS)
		goto find_conflict;

943
	list_for_each_entry(fl, &ctx->flc_flock, fl_list) {
944
		if (request->fl_file != fl->fl_file)
Linus Torvalds's avatar
Linus Torvalds committed
945
			continue;
946
		if (request->fl_type == fl->fl_type)
Linus Torvalds's avatar
Linus Torvalds committed
947
			goto out;
948
		found = true;
949
		locks_delete_lock_ctx(fl, &dispose);
Linus Torvalds's avatar
Linus Torvalds committed
950 951 952
		break;
	}

953 954 955
	if (request->fl_type == F_UNLCK) {
		if ((request->fl_flags & FL_EXISTS) && !found)
			error = -ENOENT;
956
		goto out;
957
	}
Linus Torvalds's avatar
Linus Torvalds committed
958

959
find_conflict:
960
	list_for_each_entry(fl, &ctx->flc_flock, fl_list) {
961
		if (!flock_locks_conflict(request, fl))
Linus Torvalds's avatar
Linus Torvalds committed
962 963
			continue;
		error = -EAGAIN;
964 965 966 967
		if (!(request->fl_flags & FL_SLEEP))
			goto out;
		error = FILE_LOCK_DEFERRED;
		locks_insert_block(fl, request);
Linus Torvalds's avatar
Linus Torvalds committed
968 969
		goto out;
	}
970 971
	if (request->fl_flags & FL_ACCESS)
		goto out;
972
	locks_copy_lock(new_fl, request);
973
	locks_insert_lock_ctx(new_fl, &ctx->flc_flock);
974
	new_fl = NULL;
975
	error = 0;
Linus Torvalds's avatar
Linus Torvalds committed
976 977

out:
978
	spin_unlock(&ctx->flc_lock);
979
	percpu_up_read_preempt_enable(&file_rwsem);
980 981
	if (new_fl)
		locks_free_lock(new_fl);
982
	locks_dispose_list(&dispose);
Linus Torvalds's avatar
Linus Torvalds committed
983 984 985
	return error;
}

986 987
static int posix_lock_inode(struct inode *inode, struct file_lock *request,
			    struct file_lock *conflock)
Linus Torvalds's avatar
Linus Torvalds committed
988
{
989
	struct file_lock *fl, *tmp;
990 991
	struct file_lock *new_fl = NULL;
	struct file_lock *new_fl2 = NULL;
Linus Torvalds's avatar
Linus Torvalds committed
992 993
	struct file_lock *left = NULL;
	struct file_lock *right = NULL;
994
	struct file_lock_context *ctx;
995 996
	int error;
	bool added = false;
997
	LIST_HEAD(dispose);
Linus Torvalds's avatar
Linus Torvalds committed
998

999
	ctx = locks_get_lock_context(inode, request->fl_type);
1000
	if (!ctx)
1001
		return (request->fl_type == F_UNLCK) ? 0 : -ENOMEM;
1002

Linus Torvalds's avatar
Linus Torvalds committed
1003 1004 1005
	/*
	 * We may need two file_lock structures for this operation,
	 * so we get them in advance to avoid races.
1006 1007
	 *
	 * In some cases we can be sure, that no new locks will be needed
Linus Torvalds's avatar
Linus Torvalds committed
1008
	 */
1009 1010 1011 1012 1013 1014
	if (!(request->fl_flags & FL_ACCESS) &&
	    (request->fl_type != F_UNLCK ||
	     request->fl_start != 0 || request->fl_end != OFFSET_MAX)) {
		new_fl = locks_alloc_lock();
		new_fl2 = locks_alloc_lock();
	}
Linus Torvalds's avatar
Linus Torvalds committed
1015

1016
	percpu_down_read_preempt_disable(&file_rwsem);
1017
	spin_lock(&ctx->flc_lock);
1018 1019 1020
	/*
	 * New lock request. Walk all POSIX locks and look for conflicts. If
	 * there are any, either return error or put the request on the
1021
	 * blocker's list of waiters and the global blocked_hash.
1022
	 */
Linus Torvalds's avatar
Linus Torvalds committed
1023
	if (request->fl_type != F_UNLCK) {
1024
		list_for_each_entry(fl, &ctx->flc_posix, fl_list) {
Linus Torvalds's avatar
Linus Torvalds committed
1025 1026
			if (!posix_locks_conflict(request, fl))
				continue;
1027
			if (conflock)
1028
				locks_copy_conflock(conflock, fl);
Linus Torvalds's avatar
Linus Torvalds committed
1029 1030 1031
			error = -EAGAIN;
			if (!(request->fl_flags & FL_SLEEP))
				goto out;
1032 1033 1034 1035
			/*
			 * Deadlock detection and insertion into the blocked
			 * locks list must be done while holding the same lock!
			 */
Linus Torvalds's avatar
Linus Torvalds committed
1036
			error = -EDEADLK;
1037
			spin_lock(&blocked_lock_lock);
1038 1039 1040 1041
			if (likely(!posix_locks_deadlock(request, fl))) {
				error = FILE_LOCK_DEFERRED;
				__locks_insert_block(fl, request);
			}
1042
			spin_unlock(&blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
1043 1044 1045 1046 1047 1048 1049 1050 1051
			goto out;
  		}
  	}

	/* If we're just looking for a conflict, we're done. */
	error = 0;
	if (request->fl_flags & FL_ACCESS)
		goto out;

1052 1053 1054 1055
	/* Find the first old lock with the same owner as the new lock */
	list_for_each_entry(fl, &ctx->flc_posix, fl_list) {
		if (posix_same_owner(request, fl))
			break;
Linus Torvalds's avatar
Linus Torvalds committed
1056 1057
	}

1058
	/* Process locks with this owner. */
1059 1060 1061 1062 1063
	list_for_each_entry_safe_from(fl, tmp, &ctx->flc_posix, fl_list) {
		if (!posix_same_owner(request, fl))
			break;

		/* Detect adjacent or overlapping regions (if same lock type) */
Linus Torvalds's avatar
Linus Torvalds committed
1064
		if (request->fl_type == fl->fl_type) {
Olaf Kirch's avatar
Olaf Kirch committed
1065 1066 1067 1068
			/* In all comparisons of start vs end, use
			 * "start - 1" rather than "end + 1". If end
			 * is OFFSET_MAX, end + 1 will become negative.
			 */
Linus Torvalds's avatar
Linus Torvalds committed
1069
			if (fl->fl_end < request->fl_start - 1)
1070
				continue;
Linus Torvalds's avatar
Linus Torvalds committed
1071 1072 1073
			/* If the next lock in the list has entirely bigger
			 * addresses than the new one, insert the lock here.
			 */
Olaf Kirch's avatar
Olaf Kirch committed
1074
			if (fl->fl_start - 1 > request->fl_end)
Linus Torvalds's avatar
Linus Torvalds committed
1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090
				break;

			/* If we come here, the new and old lock are of the
			 * same type and adjacent or overlapping. Make one
			 * lock yielding from the lower start address of both
			 * locks to the higher end address.
			 */
			if (fl->fl_start > request->fl_start)
				fl->fl_start = request->fl_start;
			else
				request->fl_start = fl->fl_start;
			if (fl->fl_end < request->fl_end)
				fl->fl_end = request->fl_end;
			else
				request->fl_end = fl->fl_end;
			if (added) {
1091
				locks_delete_lock_ctx(fl, &dispose);
Linus Torvalds's avatar
Linus Torvalds committed
1092 1093 1094
				continue;
			}
			request = fl;
1095
			added = true;
1096
		} else {
Linus Torvalds's avatar
Linus Torvalds committed
1097 1098 1099 1100
			/* Processing for different lock types is a bit
			 * more complex.
			 */
			if (fl->fl_end < request->fl_start)
1101
				continue;
Linus Torvalds's avatar
Linus Torvalds committed
1102 1103 1104
			if (fl->fl_start > request->fl_end)
				break;
			if (request->fl_type == F_UNLCK)
1105
				added = true;
Linus Torvalds's avatar
Linus Torvalds committed
1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119
			if (fl->fl_start < request->fl_start)
				left = fl;
			/* If the next lock in the list has a higher end
			 * address than the new one, insert the new one here.
			 */
			if (fl->fl_end > request->fl_end) {
				right = fl;
				break;
			}
			if (fl->fl_start >= request->fl_start) {
				/* The new lock completely replaces an old
				 * one (This may happen several times).
				 */
				if (added) {
1120
					locks_delete_lock_ctx(fl, &dispose);
Linus Torvalds's avatar
Linus Torvalds committed
1121 1122
					continue;
				}
1123 1124 1125 1126 1127 1128
				/*
				 * Replace the old lock with new_fl, and
				 * remove the old one. It's safe to do the
				 * insert here since we know that we won't be
				 * using new_fl later, and that the lock is
				 * just replacing an existing lock.
Linus Torvalds's avatar
Linus Torvalds committed
1129
				 */
1130 1131 1132 1133 1134 1135
				error = -ENOLCK;
				if (!new_fl)
					goto out;
				locks_copy_lock(new_fl, request);
				request = new_fl;
				new_fl = NULL;
1136 1137
				locks_insert_lock_ctx(request, &fl->fl_list);
				locks_delete_lock_ctx(fl, &dispose);
1138
				added = true;
Linus Torvalds's avatar
Linus Torvalds committed
1139 1140 1141 1142
			}
		}
	}

1143
	/*
1144 1145 1146
	 * The above code only modifies existing locks in case of merging or
	 * replacing. If new lock(s) need to be inserted all modifications are
	 * done below this, so it's safe yet to bail out.
1147 1148 1149 1150 1151
	 */
	error = -ENOLCK; /* "no luck" */
	if (right && left == right && !new_fl2)
		goto out;

Linus Torvalds's avatar
Linus Torvalds committed
1152 1153
	error = 0;
	if (!added) {
1154 1155 1156
		if (request->fl_type == F_UNLCK) {
			if (request->fl_flags & FL_EXISTS)
				error = -ENOENT;
Linus Torvalds's avatar
Linus Torvalds committed
1157
			goto out;
1158
		}
1159 1160 1161 1162 1163

		if (!new_fl) {
			error = -ENOLCK;
			goto out;
		}
Linus Torvalds's avatar
Linus Torvalds committed
1164
		locks_copy_lock(new_fl, request);
1165
		locks_insert_lock_ctx(new_fl, &fl->fl_list);
1166
		fl = new_fl;
Linus Torvalds's avatar
Linus Torvalds committed
1167 1168 1169 1170 1171 1172 1173 1174 1175 1176
		new_fl = NULL;
	}
	if (right) {
		if (left == right) {
			/* The new lock breaks the old one in two pieces,
			 * so we have to use the second new lock.
			 */
			left = new_fl2;
			new_fl2 = NULL;
			locks_copy_lock(left, right);
1177
			locks_insert_lock_ctx(left, &fl->fl_list);
Linus Torvalds's avatar
Linus Torvalds committed
1178 1179 1180 1181 1182 1183 1184 1185 1186
		}
		right->fl_start = request->fl_end + 1;
		locks_wake_up_blocks(right);
	}
	if (left) {
		left->fl_end = request->fl_start - 1;
		locks_wake_up_blocks(left);
	}
 out:
1187
	spin_unlock(&ctx->flc_lock);
1188
	percpu_up_read_preempt_enable(&file_rwsem);
Linus Torvalds's avatar
Linus Torvalds committed
1189 1190 1191 1192 1193 1194 1195
	/*
	 * Free any unused locks.
	 */
	if (new_fl)
		locks_free_lock(new_fl);
	if (new_fl2)
		locks_free_lock(new_fl2);
1196
	locks_dispose_list(&dispose);
1197 1198
	trace_posix_lock_inode(inode, request, error);

Linus Torvalds's avatar
Linus Torvalds committed
1199 1200 1201 1202 1203 1204 1205
	return error;
}

/**
 * posix_lock_file - Apply a POSIX-style lock to a file
 * @filp: The file to apply the lock to
 * @fl: The lock to be applied
1206
 * @conflock: Place to return a copy of the conflicting lock, if found.
Linus Torvalds's avatar
Linus Torvalds committed
1207 1208 1209 1210
 *
 * Add a POSIX style lock to a file.
 * We merge adjacent & overlapping locks whenever possible.
 * POSIX locks are sorted by owner task, then by starting address
1211 1212 1213 1214
 *
 * Note that if called with an FL_EXISTS argument, the caller may determine
 * whether or not a lock was successfully freed by testing the return
 * value for -ENOENT.
Linus Torvalds's avatar
Linus Torvalds committed
1215
 */