locks.c 73.7 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>

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

#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)
140
#define IS_REMOTELCK(fl)	(fl->fl_pid <= 0)
Linus Torvalds's avatar
Linus Torvalds committed
141

142 143
static inline bool is_remote_lock(struct file *filp)
{
144
	return likely(!(filp->f_path.dentry->d_sb->s_flags & SB_NOREMOTELOCK));
145 146
}

147 148
static bool lease_breaking(struct file_lock *fl)
{
149 150 151 152 153 154 155 156 157 158
	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;
159 160
}

Linus Torvalds's avatar
Linus Torvalds committed
161 162 163
int leases_enable = 1;
int lease_break_time = 45;

164
/*
165
 * The global file_lock_list is only used for displaying /proc/locks, so we
166 167 168 169 170
 * 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.
171
 */
172 173 174 175 176
struct file_lock_list_struct {
	spinlock_t		lock;
	struct hlist_head	hlist;
};
static DEFINE_PER_CPU(struct file_lock_list_struct, file_lock_list);
177
DEFINE_STATIC_PERCPU_RWSEM(file_rwsem);
178

179
/*
180
 * The blocked_hash is used to find POSIX lock loops for deadlock detection.
181
 * It is protected by blocked_lock_lock.
182 183 184 185 186 187 188
 *
 * 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.
189
 */
190 191
#define BLOCKED_HASH_BITS	7
static DEFINE_HASHTABLE(blocked_hash, BLOCKED_HASH_BITS);
192

193
/*
194 195
 * This lock protects the blocked_hash. Generally, if you're accessing it, you
 * want to be holding this lock.
196 197 198 199 200 201
 *
 * 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,
202
 * we often hold the flc_lock as well. In certain cases, when reading the fields
203
 * protected by this lock, we can skip acquiring it iff we already hold the
204
 * flc_lock.
205 206
 *
 * In particular, adding an entry to the fl_block list requires that you hold
207 208
 * 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.
209
 */
210
static DEFINE_SPINLOCK(blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
211

212
static struct kmem_cache *flctx_cache __read_mostly;
213
static struct kmem_cache *filelock_cache __read_mostly;
Linus Torvalds's avatar
Linus Torvalds committed
214

215
static struct file_lock_context *
216
locks_get_lock_context(struct inode *inode, int type)
217
{
218
	struct file_lock_context *ctx;
219

220 221 222
	/* paired with cmpxchg() below */
	ctx = smp_load_acquire(&inode->i_flctx);
	if (likely(ctx) || type == F_UNLCK)
223 224
		goto out;

225 226
	ctx = kmem_cache_alloc(flctx_cache, GFP_KERNEL);
	if (!ctx)
227 228
		goto out;

229 230 231 232
	spin_lock_init(&ctx->flc_lock);
	INIT_LIST_HEAD(&ctx->flc_flock);
	INIT_LIST_HEAD(&ctx->flc_posix);
	INIT_LIST_HEAD(&ctx->flc_lease);
233 234 235 236 237

	/*
	 * Assign the pointer if it's not already assigned. If it is, then
	 * free the context we just allocated.
	 */
238 239 240 241
	if (cmpxchg(&inode->i_flctx, NULL, ctx)) {
		kmem_cache_free(flctx_cache, ctx);
		ctx = smp_load_acquire(&inode->i_flctx);
	}
242
out:
243
	trace_locks_get_lock_context(inode, type, ctx);
244
	return ctx;
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
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");
	}
}

274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289
static void
locks_check_ctx_file_list(struct file *filp, struct list_head *list,
				char *list_type)
{
	struct file_lock *fl;
	struct inode *inode = locks_inode(filp);

	list_for_each_entry(fl, list, fl_list)
		if (fl->fl_file == filp)
			pr_warn("Leaked %s lock on dev=0x%x:0x%x ino=0x%lx "
				" fl_owner=%p fl_flags=0x%x fl_type=0x%x fl_pid=%u\n",
				list_type, MAJOR(inode->i_sb->s_dev),
				MINOR(inode->i_sb->s_dev), inode->i_ino,
				fl->fl_owner, fl->fl_flags, fl->fl_type, fl->fl_pid);
}

290
void
291
locks_free_lock_context(struct inode *inode)
292
{
293 294
	struct file_lock_context *ctx = inode->i_flctx;

295 296
	if (unlikely(ctx)) {
		locks_check_ctx_lists(inode);
297 298 299 300
		kmem_cache_free(flctx_cache, ctx);
	}
}

301
static void locks_init_lock_heads(struct file_lock *fl)
302
{
303
	INIT_HLIST_NODE(&fl->fl_link);
304
	INIT_LIST_HEAD(&fl->fl_list);
305 306
	INIT_LIST_HEAD(&fl->fl_block);
	init_waitqueue_head(&fl->fl_wait);
307 308
}

Linus Torvalds's avatar
Linus Torvalds committed
309
/* Allocate an empty lock structure. */
310
struct file_lock *locks_alloc_lock(void)
Linus Torvalds's avatar
Linus Torvalds committed
311
{
312
	struct file_lock *fl = kmem_cache_zalloc(filelock_cache, GFP_KERNEL);
313 314

	if (fl)
315
		locks_init_lock_heads(fl);
316 317

	return fl;
Linus Torvalds's avatar
Linus Torvalds committed
318
}
319
EXPORT_SYMBOL_GPL(locks_alloc_lock);
Linus Torvalds's avatar
Linus Torvalds committed
320

321
void locks_release_private(struct file_lock *fl)
322 323 324 325 326 327 328
{
	if (fl->fl_ops) {
		if (fl->fl_ops->fl_release_private)
			fl->fl_ops->fl_release_private(fl);
		fl->fl_ops = NULL;
	}

329
	if (fl->fl_lmops) {
330 331 332 333
		if (fl->fl_lmops->lm_put_owner) {
			fl->fl_lmops->lm_put_owner(fl->fl_owner);
			fl->fl_owner = NULL;
		}
334 335
		fl->fl_lmops = NULL;
	}
336
}
337
EXPORT_SYMBOL_GPL(locks_release_private);
338

Linus Torvalds's avatar
Linus Torvalds committed
339
/* Free a lock which is not in use. */
340
void locks_free_lock(struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
341
{
342
	BUG_ON(waitqueue_active(&fl->fl_wait));
343
	BUG_ON(!list_empty(&fl->fl_list));
344
	BUG_ON(!list_empty(&fl->fl_block));
345
	BUG_ON(!hlist_unhashed(&fl->fl_link));
Linus Torvalds's avatar
Linus Torvalds committed
346

347
	locks_release_private(fl);
Linus Torvalds's avatar
Linus Torvalds committed
348 349
	kmem_cache_free(filelock_cache, fl);
}
350
EXPORT_SYMBOL(locks_free_lock);
Linus Torvalds's avatar
Linus Torvalds committed
351

352 353 354 355 356 357
static void
locks_dispose_list(struct list_head *dispose)
{
	struct file_lock *fl;

	while (!list_empty(dispose)) {
358 359
		fl = list_first_entry(dispose, struct file_lock, fl_list);
		list_del_init(&fl->fl_list);
360 361 362 363
		locks_free_lock(fl);
	}
}

Linus Torvalds's avatar
Linus Torvalds committed
364 365
void locks_init_lock(struct file_lock *fl)
{
366 367
	memset(fl, 0, sizeof(struct file_lock));
	locks_init_lock_heads(fl);
Linus Torvalds's avatar
Linus Torvalds committed
368 369 370 371 372 373 374
}

EXPORT_SYMBOL(locks_init_lock);

/*
 * Initialize a new lock from an existing file_lock structure.
 */
375
void locks_copy_conflock(struct file_lock *new, struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
376 377 378
{
	new->fl_owner = fl->fl_owner;
	new->fl_pid = fl->fl_pid;
379
	new->fl_file = NULL;
Linus Torvalds's avatar
Linus Torvalds committed
380 381 382 383
	new->fl_flags = fl->fl_flags;
	new->fl_type = fl->fl_type;
	new->fl_start = fl->fl_start;
	new->fl_end = fl->fl_end;
384
	new->fl_lmops = fl->fl_lmops;
385
	new->fl_ops = NULL;
386 387 388

	if (fl->fl_lmops) {
		if (fl->fl_lmops->lm_get_owner)
389
			fl->fl_lmops->lm_get_owner(fl->fl_owner);
390
	}
391
}
392
EXPORT_SYMBOL(locks_copy_conflock);
393 394 395

void locks_copy_lock(struct file_lock *new, struct file_lock *fl)
{
396 397
	/* "new" must be a freshly-initialized lock */
	WARN_ON_ONCE(new->fl_ops);
398

399
	locks_copy_conflock(new, fl);
400

401
	new->fl_file = fl->fl_file;
Linus Torvalds's avatar
Linus Torvalds committed
402
	new->fl_ops = fl->fl_ops;
403

404 405 406 407
	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
408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426
}

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. */
427 428
static struct file_lock *
flock_make_lock(struct file *filp, unsigned int cmd)
Linus Torvalds's avatar
Linus Torvalds committed
429 430 431
{
	struct file_lock *fl;
	int type = flock_translate_cmd(cmd);
432

Linus Torvalds's avatar
Linus Torvalds committed
433
	if (type < 0)
434
		return ERR_PTR(type);
Linus Torvalds's avatar
Linus Torvalds committed
435 436 437
	
	fl = locks_alloc_lock();
	if (fl == NULL)
438
		return ERR_PTR(-ENOMEM);
Linus Torvalds's avatar
Linus Torvalds committed
439 440

	fl->fl_file = filp;
441
	fl->fl_owner = filp;
Linus Torvalds's avatar
Linus Torvalds committed
442 443 444 445 446
	fl->fl_pid = current->tgid;
	fl->fl_flags = FL_FLOCK;
	fl->fl_type = type;
	fl->fl_end = OFFSET_MAX;
	
447
	return fl;
Linus Torvalds's avatar
Linus Torvalds committed
448 449
}

450
static int assign_type(struct file_lock *fl, long type)
Linus Torvalds's avatar
Linus Torvalds committed
451 452 453 454 455 456 457 458 459 460 461 462 463
{
	switch (type) {
	case F_RDLCK:
	case F_WRLCK:
	case F_UNLCK:
		fl->fl_type = type;
		break;
	default:
		return -EINVAL;
	}
	return 0;
}

464 465
static int flock64_to_posix_lock(struct file *filp, struct file_lock *fl,
				 struct flock64 *l)
Linus Torvalds's avatar
Linus Torvalds committed
466 467
{
	switch (l->l_whence) {
468
	case SEEK_SET:
469
		fl->fl_start = 0;
Linus Torvalds's avatar
Linus Torvalds committed
470
		break;
471
	case SEEK_CUR:
472
		fl->fl_start = filp->f_pos;
Linus Torvalds's avatar
Linus Torvalds committed
473
		break;
474
	case SEEK_END:
475
		fl->fl_start = i_size_read(file_inode(filp));
Linus Torvalds's avatar
Linus Torvalds committed
476 477 478 479
		break;
	default:
		return -EINVAL;
	}
480 481 482 483 484
	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
485 486 487

	/* POSIX-1996 leaves the case l->l_len < 0 undefined;
	   POSIX-2001 defines it. */
488
	if (l->l_len > 0) {
489 490 491 492
		if (l->l_len - 1 > OFFSET_MAX - fl->fl_start)
			return -EOVERFLOW;
		fl->fl_end = fl->fl_start + l->l_len - 1;

493
	} else if (l->l_len < 0) {
494
		if (fl->fl_start + l->l_len < 0)
495
			return -EINVAL;
496 497 498 499 500
		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
501 502 503 504 505 506 507 508 509 510
	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);
}

511 512 513 514 515
/* 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
516
{
517 518 519 520 521 522 523 524
	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
525 526 527
}

/* default lease lock manager operations */
528 529
static bool
lease_break_callback(struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
530 531
{
	kill_fasync(&fl->fl_fasync, SIGIO, POLL_MSG);
532
	return false;
Linus Torvalds's avatar
Linus Torvalds committed
533 534
}

535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551
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);
}

552
static const struct lock_manager_operations lease_manager_ops = {
553 554
	.lm_break = lease_break_callback,
	.lm_change = lease_modify,
555
	.lm_setup = lease_setup,
Linus Torvalds's avatar
Linus Torvalds committed
556 557 558 559 560
};

/*
 * Initialize a lease, use the default lock manager operations
 */
561
static int lease_init(struct file *filp, long type, struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
562
 {
563 564 565
	if (assign_type(fl, type) != 0)
		return -EINVAL;

566
	fl->fl_owner = filp;
Linus Torvalds's avatar
Linus Torvalds committed
567 568 569 570 571 572 573 574 575 576 577 578
	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 */
579
static struct file_lock *lease_alloc(struct file *filp, long type)
Linus Torvalds's avatar
Linus Torvalds committed
580 581
{
	struct file_lock *fl = locks_alloc_lock();
582
	int error = -ENOMEM;
Linus Torvalds's avatar
Linus Torvalds committed
583 584

	if (fl == NULL)
585
		return ERR_PTR(error);
Linus Torvalds's avatar
Linus Torvalds committed
586 587

	error = lease_init(filp, type, fl);
588 589
	if (error) {
		locks_free_lock(fl);
590
		return ERR_PTR(error);
591
	}
592
	return fl;
Linus Torvalds's avatar
Linus Torvalds committed
593 594 595 596 597 598 599 600 601 602 603 604 605
}

/* 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.
 */
606
static int posix_same_owner(struct file_lock *fl1, struct file_lock *fl2)
Linus Torvalds's avatar
Linus Torvalds committed
607
{
608
	if (fl1->fl_lmops && fl1->fl_lmops->lm_compare_owner)
Linus Torvalds's avatar
Linus Torvalds committed
609
		return fl2->fl_lmops == fl1->fl_lmops &&
610
			fl1->fl_lmops->lm_compare_owner(fl1, fl2);
Linus Torvalds's avatar
Linus Torvalds committed
611 612 613
	return fl1->fl_owner == fl2->fl_owner;
}

614
/* Must be called with the flc_lock held! */
615
static void locks_insert_global_locks(struct file_lock *fl)
616
{
617 618
	struct file_lock_list_struct *fll = this_cpu_ptr(&file_lock_list);

619 620
	percpu_rwsem_assert_held(&file_rwsem);

621
	spin_lock(&fll->lock);
622
	fl->fl_link_cpu = smp_processor_id();
623 624
	hlist_add_head(&fl->fl_link, &fll->hlist);
	spin_unlock(&fll->lock);
625 626
}

627
/* Must be called with the flc_lock held! */
628
static void locks_delete_global_locks(struct file_lock *fl)
629
{
630 631
	struct file_lock_list_struct *fll;

632 633
	percpu_rwsem_assert_held(&file_rwsem);

634 635
	/*
	 * Avoid taking lock if already unhashed. This is safe since this check
636
	 * is done while holding the flc_lock, and new insertions into the list
637 638 639 640
	 * also require that it be held.
	 */
	if (hlist_unhashed(&fl->fl_link))
		return;
641 642 643

	fll = per_cpu_ptr(&file_lock_list, fl->fl_link_cpu);
	spin_lock(&fll->lock);
644
	hlist_del_init(&fl->fl_link);
645
	spin_unlock(&fll->lock);
646 647
}

648 649 650 651 652 653 654 655
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;
}

656
static void locks_insert_global_blocked(struct file_lock *waiter)
657
{
658 659
	lockdep_assert_held(&blocked_lock_lock);

660
	hash_add(blocked_hash, &waiter->fl_link, posix_owner_key(waiter));
661 662
}

663
static void locks_delete_global_blocked(struct file_lock *waiter)
664
{
665 666
	lockdep_assert_held(&blocked_lock_lock);

667
	hash_del(&waiter->fl_link);
668 669
}

Linus Torvalds's avatar
Linus Torvalds committed
670 671
/* Remove waiter from blocker's block list.
 * When blocker ends up pointing to itself then the list is empty.
672
 *
673
 * Must be called with blocked_lock_lock held.
Linus Torvalds's avatar
Linus Torvalds committed
674
 */
675
static void __locks_delete_block(struct file_lock *waiter)
Linus Torvalds's avatar
Linus Torvalds committed
676
{
677
	locks_delete_global_blocked(waiter);
Linus Torvalds's avatar
Linus Torvalds committed
678 679 680 681
	list_del_init(&waiter->fl_block);
	waiter->fl_next = NULL;
}

682
static void locks_delete_block(struct file_lock *waiter)
Linus Torvalds's avatar
Linus Torvalds committed
683
{
684
	spin_lock(&blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
685
	__locks_delete_block(waiter);
686
	spin_unlock(&blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
687 688 689 690 691 692
}

/* 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.
693
 *
694 695 696 697
 * 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
698
 */
699 700
static void __locks_insert_block(struct file_lock *blocker,
					struct file_lock *waiter)
Linus Torvalds's avatar
Linus Torvalds committed
701
{
702
	BUG_ON(!list_empty(&waiter->fl_block));
Linus Torvalds's avatar
Linus Torvalds committed
703
	waiter->fl_next = blocker;
704
	list_add_tail(&waiter->fl_block, &blocker->fl_block);
705
	if (IS_POSIX(blocker) && !IS_OFDLCK(blocker))
706 707 708
		locks_insert_global_blocked(waiter);
}

709
/* Must be called with flc_lock held. */
710 711 712
static void locks_insert_block(struct file_lock *blocker,
					struct file_lock *waiter)
{
713
	spin_lock(&blocked_lock_lock);
714
	__locks_insert_block(blocker, waiter);
715
	spin_unlock(&blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
716 717
}

718 719 720
/*
 * Wake up processes blocked waiting for blocker.
 *
721
 * Must be called with the inode->flc_lock held!
Linus Torvalds's avatar
Linus Torvalds committed
722 723 724
 */
static void locks_wake_up_blocks(struct file_lock *blocker)
{
725 726
	/*
	 * Avoid taking global lock if list is empty. This is safe since new
727 728 729
	 * 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()
730
	 * after acquiring the blocked_lock_lock.
731 732 733 734
	 */
	if (list_empty(&blocker->fl_block))
		return;

735
	spin_lock(&blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
736
	while (!list_empty(&blocker->fl_block)) {
737 738 739
		struct file_lock *waiter;

		waiter = list_first_entry(&blocker->fl_block,
Linus Torvalds's avatar
Linus Torvalds committed
740 741
				struct file_lock, fl_block);
		__locks_delete_block(waiter);
742 743
		if (waiter->fl_lmops && waiter->fl_lmops->lm_notify)
			waiter->fl_lmops->lm_notify(waiter);
Linus Torvalds's avatar
Linus Torvalds committed
744 745 746
		else
			wake_up(&waiter->fl_wait);
	}
747
	spin_unlock(&blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
748 749
}

750
static void
751
locks_insert_lock_ctx(struct file_lock *fl, struct list_head *before)
752 753 754 755 756
{
	list_add_tail(&fl->fl_list, before);
	locks_insert_global_locks(fl);
}

757
static void
758
locks_unlink_lock_ctx(struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
759
{
760
	locks_delete_global_locks(fl);
761
	list_del_init(&fl->fl_list);
Linus Torvalds's avatar
Linus Torvalds committed
762
	locks_wake_up_blocks(fl);
763 764
}

765
static void
766
locks_delete_lock_ctx(struct file_lock *fl, struct list_head *dispose)
767
{
768
	locks_unlink_lock_ctx(fl);
769
	if (dispose)
770
		list_add(&fl->fl_list, dispose);
771 772
	else
		locks_free_lock(fl);
Linus Torvalds's avatar
Linus Torvalds committed
773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794
}

/* 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.
	 */
795
	if (posix_same_owner(caller_fl, sys_fl))
Linus Torvalds's avatar
Linus Torvalds committed
796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812
		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.
	 */
813
	if (caller_fl->fl_file == sys_fl->fl_file)
Linus Torvalds's avatar
Linus Torvalds committed
814 815 816 817 818 819 820
		return (0);
	if ((caller_fl->fl_type & LOCK_MAND) || (sys_fl->fl_type & LOCK_MAND))
		return 0;

	return (locks_conflict(caller_fl, sys_fl));
}

821
void
822
posix_test_lock(struct file *filp, struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
823 824
{
	struct file_lock *cfl;
825
	struct file_lock_context *ctx;
826
	struct inode *inode = locks_inode(filp);
Linus Torvalds's avatar
Linus Torvalds committed
827

828
	ctx = smp_load_acquire(&inode->i_flctx);
829 830 831 832 833
	if (!ctx || list_empty_careful(&ctx->flc_posix)) {
		fl->fl_type = F_UNLCK;
		return;
	}

834
	spin_lock(&ctx->flc_lock);
835 836 837 838 839
	list_for_each_entry(cfl, &ctx->flc_posix, fl_list) {
		if (posix_locks_conflict(fl, cfl)) {
			locks_copy_conflock(fl, cfl);
			goto out;
		}
Linus Torvalds's avatar
Linus Torvalds committed
840
	}
841 842
	fl->fl_type = F_UNLCK;
out:
843
	spin_unlock(&ctx->flc_lock);
844
	return;
Linus Torvalds's avatar
Linus Torvalds committed
845 846 847
}
EXPORT_SYMBOL(posix_test_lock);

848 849 850 851 852
/*
 * Deadlock detection:
 *
 * We attempt to detect deadlocks that are due purely to posix file
 * locks.
Linus Torvalds's avatar
Linus Torvalds committed
853
 *
854 855 856 857 858 859 860
 * 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
861
 *
862 863 864
 * 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.
865
 *
866 867 868 869
 * 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.
870
 *
871
 * For FL_OFDLCK locks, the owner is the filp, not the files_struct.
872 873 874 875
 * 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.
 *
876
 * In principle, we could do a more limited deadlock detection on FL_OFDLCK
877 878
 * 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
879
 */
880 881 882

#define MAX_DEADLK_ITERATIONS 10

883 884 885 886 887
/* 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;

888
	hash_for_each_possible(blocked_hash, fl, fl_link, posix_owner_key(block_fl)) {
889 890 891 892 893 894
		if (posix_same_owner(fl, block_fl))
			return fl->fl_next;
	}
	return NULL;
}

895
/* Must be called with the blocked_lock_lock held! */
896
static int posix_locks_deadlock(struct file_lock *caller_fl,
Linus Torvalds's avatar
Linus Torvalds committed
897 898
				struct file_lock *block_fl)
{
899
	int i = 0;
Linus Torvalds's avatar
Linus Torvalds committed
900

901 902
	lockdep_assert_held(&blocked_lock_lock);

903 904
	/*
	 * This deadlock detector can't reasonably detect deadlocks with
905
	 * FL_OFDLCK locks, since they aren't owned by a process, per-se.
906
	 */
907
	if (IS_OFDLCK(caller_fl))
908 909
		return 0;

910 911 912 913 914
	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
915 916 917 918 919
	}
	return 0;
}

/* Try to create a FLOCK lock on filp. We always insert new FLOCK locks
920
 * after any leases, but before any posix locks.
921 922 923 924
 *
 * 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
925
 */
926
static int flock_lock_inode(struct inode *inode, struct file_lock *request)
Linus Torvalds's avatar
Linus Torvalds committed
927
{
928
	struct file_lock *new_fl = NULL;
929 930
	struct file_lock *fl;
	struct file_lock_context *ctx;
Linus Torvalds's avatar
Linus Torvalds committed
931
	int error = 0;
932
	bool found = false;
933
	LIST_HEAD(dispose);
Linus Torvalds's avatar
Linus Torvalds committed
934

935 936 937 938 939 940
	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;
	}
941

942
	if (!(request->fl_flags & FL_ACCESS) && (request->fl_type != F_UNLCK)) {
943
		new_fl = locks_alloc_lock();
944 945
		if (!new_fl)
			return -ENOMEM;
946 947
	}

948
	percpu_down_read_preempt_disable(&file_rwsem);
949
	spin_lock(&ctx->flc_lock);
950 951 952
	if (request->fl_flags & FL_ACCESS)
		goto find_conflict;

953
	list_for_each_entry(fl, &ctx->flc_flock, fl_list) {
954
		if (request->fl_file != fl->fl_file)
Linus Torvalds's avatar
Linus Torvalds committed
955
			continue;
956
		if (request->fl_type == fl->fl_type)
Linus Torvalds's avatar
Linus Torvalds committed
957
			goto out;
958
		found = true;
959
		locks_delete_lock_ctx(fl, &dispose);
Linus Torvalds's avatar
Linus Torvalds committed
960 961 962
		break;
	}

963 964 965
	if (request->fl_type == F_UNLCK) {
		if ((request->fl_flags & FL_EXISTS) && !found)
			error = -ENOENT;
966
		goto out;
967
	}
Linus Torvalds's avatar
Linus Torvalds committed
968

969
find_conflict:
970
	list_for_each_entry(fl, &ctx->flc_flock, fl_list) {
971
		if (!flock_locks_conflict(request, fl))
Linus Torvalds's avatar
Linus Torvalds committed
972 973
			continue;
		error = -EAGAIN;
974 975 976 977
		if (!(request->fl_flags & FL_SLEEP))
			goto out;
		error = FILE_LOCK_DEFERRED;
		locks_insert_block(fl, request);
Linus Torvalds's avatar
Linus Torvalds committed
978 979
		goto out;
	}
980 981
	if (request->fl_flags & FL_ACCESS)
		goto out;
982
	locks_copy_lock(new_fl, request);
983
	locks_insert_lock_ctx(new_fl, &ctx->flc_flock);
984
	new_fl = NULL;
985
	error = 0;
Linus Torvalds's avatar
Linus Torvalds committed
986 987

out:
988
	spin_unlock(&ctx->flc_lock);
989
	percpu_up_read_preempt_enable(&file_rwsem);
990 991
	if (new_fl)
		locks_free_lock(new_fl);
992
	locks_dispose_list(&dispose);
Linus Torvalds's avatar
Linus Torvalds committed
993 994 995
	return error;
}

996 997
static int posix_lock_inode(struct inode *inode, struct file_lock *request,
			    struct file_lock *conflock)
Linus Torvalds's avatar
Linus Torvalds committed
998
{
999
	struct file_lock *fl, *tmp;
1000 1001
	struct file_lock *new_fl = NULL;
	struct file_lock *new_fl2 = NULL;
Linus Torvalds's avatar
Linus Torvalds committed
1002 1003
	struct file_lock *left = NULL;
	struct file_lock *right = NULL;
1004
	struct file_lock_context *ctx;
1005 1006
	int error;
	bool added = false;
1007
	LIST_HEAD(dispose);
Linus Torvalds's avatar
Linus Torvalds committed
1008

1009
	ctx = locks_get_lock_context(inode, request->fl_type);
1010
	if (!ctx)
1011
		return (request->fl_type == F_UNLCK) ? 0 : -ENOMEM;
1012

Linus Torvalds's avatar
Linus Torvalds committed
1013 1014 1015
	/*
	 * We may need two file_lock structures for this operation,
	 * so we get them in advance to avoid races.
1016 1017
	 *
	 * In some cases we can be sure, that no new locks will be needed
Linus Torvalds's avatar
Linus Torvalds committed
1018
	 */
1019 1020 1021 1022 1023 1024
	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
1025

1026
	percpu_down_read_preempt_disable(&file_rwsem);
1027
	spin_lock(&ctx->flc_lock);
1028 1029 1030
	/*
	 * New lock request. Walk all POSIX locks and look for conflicts. If
	 * there are any, either return error or put the request on the
1031
	 * blocker's list of waiters and the global blocked_hash.
1032
	 */
Linus Torvalds's avatar
Linus Torvalds committed
1033
	if (request->fl_type != F_UNLCK) {
1034
		list_for_each_entry(fl, &ctx->flc_posix, fl_list) {
Linus Torvalds's avatar
Linus Torvalds committed
1035 1036
			if (!posix_locks_conflict(request, fl))
				continue;
1037
			if (conflock)
1038
				locks_copy_conflock(conflock, fl);
Linus Torvalds's avatar
Linus Torvalds committed
1039 1040 1041
			error = -EAGAIN;
			if (!(request->fl_flags & FL_SLEEP))
				goto out;
1042 1043 1044 1045
			/*
			 * Deadlock detection and insertion into the blocked
			 * locks list must be done while holding the same lock!
			 */
Linus Torvalds's avatar
Linus Torvalds committed
1046
			error = -EDEADLK;
1047
			spin_lock(&blocked_lock_lock);
1048 1049 1050 1051
			if (likely(!posix_locks_deadlock(request, fl))) {
				error = FILE_LOCK_DEFERRED;
				__locks_insert_block(fl, request);
			}
1052
			spin_unlock(&blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
1053 1054 1055 1056 1057 1058 1059 1060 1061
			goto out;
  		}
  	}

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

1062 1063 1064 1065
	/* 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
1066 1067
	}

1068
	/* Process locks with this owner. */
1069 1070 1071 1072 1073
	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
1074
		if (request->fl_type == fl->fl_type) {
1075 1076 1077 1078
			/* 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
1079
			if (fl->fl_end < request->fl_start - 1)
1080
				continue;
Linus Torvalds's avatar
Linus Torvalds committed
1081 1082 1083
			/* If the next lock in the list has entirely bigger
			 * addresses than the new one, insert the lock here.
			 */
1084
			if (fl->fl_start - 1 > request->fl_end)
Linus Torvalds's avatar
Linus Torvalds committed
1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100
				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) {
1101
				locks_delete_lock_ctx(fl, &dispose);
Linus Torvalds's avatar
Linus Torvalds committed
1102 1103 1104
				continue;
			}
			request = fl;
1105
			added = true;
1106
		} else {
Linus Torvalds's avatar
Linus Torvalds committed
1107 1108 1109 1110
			/* Processing for different lock types is a bit
			 * more complex.
			 */
			if (fl->fl_end < request->fl_start)
1111
				continue;
Linus Torvalds's avatar
Linus Torvalds committed
1112 1113 1114
			if (fl->fl_start > request->fl_end)
				break;
			if (request->fl_type == F_UNLCK)
1115
				added = true;
Linus Torvalds's avatar
Linus Torvalds committed
1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129
			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) {
1130
					locks_delete_lock_ctx(fl, &dispose);
Linus Torvalds's avatar
Linus Torvalds committed
1131 1132
					continue;
				}
1133 1134 1135 1136 1137 1138
				/*
				 * 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
1139
				 */
1140 1141 1142 1143 1144 1145
				error = -ENOLCK;
				if (!new_fl)
					goto out;
				locks_copy_lock(new_fl, request);
				request = new_fl;
				new_fl = NULL;
1146 1147
				locks_insert_lock_ctx(request, &fl->fl_list);
				locks_delete_lock_ctx(fl, &dispose);
1148
				added = true;
Linus Torvalds's avatar
Linus Torvalds committed
1149 1150 1151 1152
			}
		}
	}

1153
	/*
1154 1155 1156
	 * 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.
1157 1158 1159 1160 1161
	 */
	error = -ENOLCK; /* "no luck" */
	if (right && left == right && !new_fl2)
		goto out;

Linus Torvalds's avatar
Linus Torvalds committed
1162 1163
	error = 0;
	if (!added) {
1164 1165 1166
		if (request->fl_type == F_UNLCK) {
			if (request->fl_flags & FL_EXISTS)
				error = -ENOENT;
Linus Torvalds's avatar
Linus Torvalds committed
1167
			goto out;
1168
		}
1169 1170 1171 1172 1173

		if (!new_fl) {
			error = -ENOLCK;
			goto out;
		}
Linus Torvalds's avatar
Linus Torvalds committed
1174
		locks_copy_lock(new_fl, request);
1175
		locks_insert_lock_ctx(new_fl, &fl->fl_list);
1176
		fl = new_fl;
Linus Torvalds's avatar
Linus Torvalds committed
1177 1178 1179 1180 1181 1182 1183 1184 1185 1186
		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);
1187
			locks_insert_lock_ctx(left, &fl->fl_list);
Linus Torvalds's avatar
Linus Torvalds committed
1188 1189 1190 1191 1192 1193 1194 1195 1196
		}
		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:
1197
	spin_unlock(&ctx->flc_lock);
1198
	percpu_up_read_preempt_enable(&file_rwsem);
Linus Torvalds's avatar
Linus Torvalds committed
1199 1200 1201 1202 1203 1204 1205
	/*
	 * Free any unused locks.
	 */
	if (new_fl)
		locks_free_lock(new_fl);
	if (new_fl2)
		locks_free_lock(new_fl2);
1206
	locks_dispose_list(&dispose);
1207 1208
	trace_posix_lock_inode(inode, request, error);

Linus Torvalds's avatar
Linus Torvalds committed
1209 1210 1211 1212 1213 1214 1215
	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
1216
 * @conflock: Place to return a copy of the conflicting lock, if found.
Linus Torvalds's avatar
Linus Torvalds committed
1217 1218 1219 1220
 *
 * 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
1221 1222 1223 1224
 *
 * 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
1225
 */
1226
int posix_lock_file(struct file *filp, struct file_lock *fl,
1227 1228
			struct file_lock *conflock)
{
1229
	return posix_lock_inode(locks_inode(filp), fl, conflock);
Linus Torvalds's avatar
Linus Torvalds committed
1230
}
1231
EXPORT_SYMBOL(posix_lock_file);
Linus Torvalds's avatar
Linus Torvalds committed
1232 1233

/**
1234 1235
 * posix_lock_inode_wait - Apply a POSIX-style lock to a file
 * @inode: inode of file to which lock request should be applied
Linus Torvalds's avatar
Linus Torvalds committed
1236 1237
 * @fl: The lock to be applied
 *
1238
 * Apply a POSIX style lock request to an inode.
Linus Torvalds's avatar
Linus Torvalds committed
1239
 */
1240
static int posix_lock_inode_wait(struct inode *inode, struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
1241 1242 1243 1244
{
	int error;
	might_sleep ();
	for (;;) {
1245
		error = posix_lock_inode(inode, fl, NULL);
1246
		if (error != FILE_LOCK_DEFERRED)
Linus Torvalds's avatar
Linus Torvalds committed
1247 1248 1249 1250 1251 1252 1253 1254 1255 1256
			break;
		error = wait_event_interruptible(fl->fl_wait, !fl->fl_next);
		if (!error)
			continue;

		locks_delete_block(fl);
		break;
	}
	return error;
}
1257

1258
#ifdef CONFIG_MANDATORY_FILE_LOCKING
Linus Torvalds's avatar
Linus Torvalds committed
1259 1260
/**
 * locks_mandatory_locked - Check for an active lock
1261
 * @file: the file to check
Linus Torvalds's avatar
Linus Torvalds committed
1262 1263 1264 1265
 *
 * Searches the inode's list of locks to find any POSIX locks which conflict.
 * This function is called from locks_verify_locked() only.
 */
1266
int locks_mandatory_locked(struct file *file)
Linus Torvalds's avatar
Linus Torvalds committed
1267
{
1268
	int ret;
1269
	struct inode *inode = locks_inode(file);
1270
	struct file_lock_context *ctx;
Linus Torvalds's avatar
Linus Torvalds committed
1271 1272
	struct file_lock *fl;

1273
	ctx = smp_load_acquire(&inode->i_flctx);
1274 1275 1276
	if (!ctx || list_empty_careful(&ctx->flc_posix))
		return 0;

Linus Torvalds's avatar
Linus Torvalds committed
1277 1278 1279
	/*
	 * Search the lock list for this inode for any POSIX locks.
	 */
1280
	spin_lock(&ctx->flc_lock);
1281 1282
	ret = 0;
	list_for_each_entry(fl, &ctx->flc_posix, fl_list) {
1283
		if (fl->fl_owner != current->files &&
1284 1285
		    fl->fl_owner != file) {
			ret = -EAGAIN;
Linus Torvalds's avatar
Linus Torvalds committed
1286
			break;
1287
		}
Linus Torvalds's avatar
Linus Torvalds committed
1288
	}
1289
	spin_unlock(&ctx->flc_lock);
1290
	return ret;
Linus Torvalds's avatar
Linus Torvalds committed
1291 1292 1293 1294
}

/**
 * locks_mandatory_area - Check for a conflicting lock
1295
 * @inode:	the file to check
Linus Torvalds's avatar
Linus Torvalds committed
1296
 * @filp:       how the file was opened (if it was)
1297 1298 1299
 * @start:	first byte in the file to check
 * @end:	lastbyte in the file to check
 * @type:	%F_WRLCK for a write lock, else %F_RDLCK
Linus Torvalds's avatar
Linus Torvalds committed
1300 1301 1302
 *
 * Searches the inode's list of locks to find any POSIX locks which conflict.
 */
1303 1304
int locks_mandatory_area(struct inode *inode, struct file *filp, loff_t start,
			 loff_t end, unsigned char type)
Linus Torvalds's avatar
Linus Torvalds committed
1305 1306 1307
{
	struct file_lock fl;
	int error;
1308
	bool sleep = false;
Linus Torvalds's avatar
Linus Torvalds committed
1309 1310 1311 1312 1313 1314

	locks_init_lock(&fl);
	fl.fl_pid = current->tgid;
	fl.fl_file = filp;
	fl.fl_flags = FL_POSIX | FL_ACCESS;
	if (filp && !(filp->f_flags & O_NONBLOCK))
1315
		sleep = true;
1316 1317 1318
	fl.fl_type = type;
	fl.fl_start = start;
	fl.fl_end = end;
Linus Torvalds's avatar
Linus Torvalds committed
1319 1320

	for (;;) {
1321
		if (filp) {
1322
			fl.fl_owner = filp;
1323
			fl.fl_flags &= ~FL_SLEEP;
1324
			error = posix_lock_inode(inode, &fl, NULL);