locks.c 69.2 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
120 121 122 123 124 125 126
#include <linux/fs.h>
#include <linux/init.h>
#include <linux/module.h>
#include <linux/security.h>
#include <linux/slab.h>
#include <linux/syscalls.h>
#include <linux/time.h>
127
#include <linux/rcupdate.h>
128
#include <linux/pid_namespace.h>
129
#include <linux/hashtable.h>
130 131
#include <linux/percpu.h>
#include <linux/lglock.h>
Linus Torvalds's avatar
Linus Torvalds committed
132

133 134 135
#define CREATE_TRACE_POINTS
#include <trace/events/filelock.h>

Linus Torvalds's avatar
Linus Torvalds committed
136 137 138 139
#include <asm/uaccess.h>

#define IS_POSIX(fl)	(fl->fl_flags & FL_POSIX)
#define IS_FLOCK(fl)	(fl->fl_flags & FL_FLOCK)
140
#define IS_LEASE(fl)	(fl->fl_flags & (FL_LEASE|FL_DELEG|FL_LAYOUT))
141
#define IS_OFDLCK(fl)	(fl->fl_flags & FL_OFDLCK)
Linus Torvalds's avatar
Linus Torvalds committed
142

J. Bruce Fields's avatar
J. Bruce Fields committed
143 144
static bool lease_breaking(struct file_lock *fl)
{
145 146 147 148 149 150 151 152 153 154
	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
155 156
}

Linus Torvalds's avatar
Linus Torvalds committed
157 158 159
int leases_enable = 1;
int lease_break_time = 45;

160
/*
161 162 163
 * The global file_lock_list is only used for displaying /proc/locks, so we
 * keep a list on each CPU, with each list protected by its own spinlock via
 * the file_lock_lglock. Note that alterations to the list also require that
164
 * the relevant flc_lock is held.
165
 */
166 167
DEFINE_STATIC_LGLOCK(file_lock_lglock);
static DEFINE_PER_CPU(struct hlist_head, file_lock_list);
168

169
/*
170
 * The blocked_hash is used to find POSIX lock loops for deadlock detection.
171
 * It is protected by blocked_lock_lock.
172 173 174 175 176 177 178
 *
 * 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.
179
 */
180 181
#define BLOCKED_HASH_BITS	7
static DEFINE_HASHTABLE(blocked_hash, BLOCKED_HASH_BITS);
182

183
/*
184 185
 * This lock protects the blocked_hash. Generally, if you're accessing it, you
 * want to be holding this lock.
186 187 188 189 190 191
 *
 * 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,
192
 * we often hold the flc_lock as well. In certain cases, when reading the fields
193
 * protected by this lock, we can skip acquiring it iff we already hold the
194
 * flc_lock.
195 196
 *
 * In particular, adding an entry to the fl_block list requires that you hold
197 198
 * 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.
199
 */
200
static DEFINE_SPINLOCK(blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
201

202
static struct kmem_cache *flctx_cache __read_mostly;
203
static struct kmem_cache *filelock_cache __read_mostly;
Linus Torvalds's avatar
Linus Torvalds committed
204

205 206 207 208 209 210 211 212 213 214 215 216
static struct file_lock_context *
locks_get_lock_context(struct inode *inode)
{
	struct file_lock_context *new;

	if (likely(inode->i_flctx))
		goto out;

	new = kmem_cache_alloc(flctx_cache, GFP_KERNEL);
	if (!new)
		goto out;

217
	spin_lock_init(&new->flc_lock);
218
	INIT_LIST_HEAD(&new->flc_flock);
219
	INIT_LIST_HEAD(&new->flc_posix);
220
	INIT_LIST_HEAD(&new->flc_lease);
221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243

	/*
	 * Assign the pointer if it's not already assigned. If it is, then
	 * free the context we just allocated.
	 */
	spin_lock(&inode->i_lock);
	if (likely(!inode->i_flctx)) {
		inode->i_flctx = new;
		new = NULL;
	}
	spin_unlock(&inode->i_lock);

	if (new)
		kmem_cache_free(flctx_cache, new);
out:
	return inode->i_flctx;
}

void
locks_free_lock_context(struct file_lock_context *ctx)
{
	if (ctx) {
		WARN_ON_ONCE(!list_empty(&ctx->flc_flock));
244
		WARN_ON_ONCE(!list_empty(&ctx->flc_posix));
245
		WARN_ON_ONCE(!list_empty(&ctx->flc_lease));
246 247 248 249
		kmem_cache_free(flctx_cache, ctx);
	}
}

Miklos Szeredi's avatar
Miklos Szeredi committed
250
static void locks_init_lock_heads(struct file_lock *fl)
Miklos Szeredi's avatar
Miklos Szeredi committed
251
{
252
	INIT_HLIST_NODE(&fl->fl_link);
253
	INIT_LIST_HEAD(&fl->fl_list);
Miklos Szeredi's avatar
Miklos Szeredi committed
254 255
	INIT_LIST_HEAD(&fl->fl_block);
	init_waitqueue_head(&fl->fl_wait);
Miklos Szeredi's avatar
Miklos Szeredi committed
256 257
}

Linus Torvalds's avatar
Linus Torvalds committed
258
/* Allocate an empty lock structure. */
259
struct file_lock *locks_alloc_lock(void)
Linus Torvalds's avatar
Linus Torvalds committed
260
{
Miklos Szeredi's avatar
Miklos Szeredi committed
261
	struct file_lock *fl = kmem_cache_zalloc(filelock_cache, GFP_KERNEL);
Miklos Szeredi's avatar
Miklos Szeredi committed
262 263

	if (fl)
Miklos Szeredi's avatar
Miklos Szeredi committed
264
		locks_init_lock_heads(fl);
Miklos Szeredi's avatar
Miklos Szeredi committed
265 266

	return fl;
Linus Torvalds's avatar
Linus Torvalds committed
267
}
268
EXPORT_SYMBOL_GPL(locks_alloc_lock);
Linus Torvalds's avatar
Linus Torvalds committed
269

270
void locks_release_private(struct file_lock *fl)
271 272 273 274 275 276 277
{
	if (fl->fl_ops) {
		if (fl->fl_ops->fl_release_private)
			fl->fl_ops->fl_release_private(fl);
		fl->fl_ops = NULL;
	}

278 279 280 281 282
	if (fl->fl_lmops) {
		if (fl->fl_lmops->lm_put_owner)
			fl->fl_lmops->lm_put_owner(fl);
		fl->fl_lmops = NULL;
	}
283
}
284
EXPORT_SYMBOL_GPL(locks_release_private);
285

Linus Torvalds's avatar
Linus Torvalds committed
286
/* Free a lock which is not in use. */
287
void locks_free_lock(struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
288
{
Miklos Szeredi's avatar
Miklos Szeredi committed
289
	BUG_ON(waitqueue_active(&fl->fl_wait));
290
	BUG_ON(!list_empty(&fl->fl_list));
Miklos Szeredi's avatar
Miklos Szeredi committed
291
	BUG_ON(!list_empty(&fl->fl_block));
292
	BUG_ON(!hlist_unhashed(&fl->fl_link));
Linus Torvalds's avatar
Linus Torvalds committed
293

294
	locks_release_private(fl);
Linus Torvalds's avatar
Linus Torvalds committed
295 296
	kmem_cache_free(filelock_cache, fl);
}
297
EXPORT_SYMBOL(locks_free_lock);
Linus Torvalds's avatar
Linus Torvalds committed
298

299 300 301 302 303 304
static void
locks_dispose_list(struct list_head *dispose)
{
	struct file_lock *fl;

	while (!list_empty(dispose)) {
305 306
		fl = list_first_entry(dispose, struct file_lock, fl_list);
		list_del_init(&fl->fl_list);
307 308 309 310
		locks_free_lock(fl);
	}
}

Linus Torvalds's avatar
Linus Torvalds committed
311 312
void locks_init_lock(struct file_lock *fl)
{
Miklos Szeredi's avatar
Miklos Szeredi committed
313 314
	memset(fl, 0, sizeof(struct file_lock));
	locks_init_lock_heads(fl);
Linus Torvalds's avatar
Linus Torvalds committed
315 316 317 318 319 320 321
}

EXPORT_SYMBOL(locks_init_lock);

/*
 * Initialize a new lock from an existing file_lock structure.
 */
322
void locks_copy_conflock(struct file_lock *new, struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
323 324 325
{
	new->fl_owner = fl->fl_owner;
	new->fl_pid = fl->fl_pid;
326
	new->fl_file = NULL;
Linus Torvalds's avatar
Linus Torvalds committed
327 328 329 330
	new->fl_flags = fl->fl_flags;
	new->fl_type = fl->fl_type;
	new->fl_start = fl->fl_start;
	new->fl_end = fl->fl_end;
331
	new->fl_lmops = fl->fl_lmops;
332
	new->fl_ops = NULL;
333 334 335 336 337

	if (fl->fl_lmops) {
		if (fl->fl_lmops->lm_get_owner)
			fl->fl_lmops->lm_get_owner(new, fl);
	}
338
}
339
EXPORT_SYMBOL(locks_copy_conflock);
340 341 342

void locks_copy_lock(struct file_lock *new, struct file_lock *fl)
{
343 344
	/* "new" must be a freshly-initialized lock */
	WARN_ON_ONCE(new->fl_ops);
345

346
	locks_copy_conflock(new, fl);
347

348
	new->fl_file = fl->fl_file;
Linus Torvalds's avatar
Linus Torvalds committed
349
	new->fl_ops = fl->fl_ops;
350

351 352 353 354
	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
355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373
}

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. */
374 375
static struct file_lock *
flock_make_lock(struct file *filp, unsigned int cmd)
Linus Torvalds's avatar
Linus Torvalds committed
376 377 378
{
	struct file_lock *fl;
	int type = flock_translate_cmd(cmd);
379

Linus Torvalds's avatar
Linus Torvalds committed
380
	if (type < 0)
381
		return ERR_PTR(type);
Linus Torvalds's avatar
Linus Torvalds committed
382 383 384
	
	fl = locks_alloc_lock();
	if (fl == NULL)
385
		return ERR_PTR(-ENOMEM);
Linus Torvalds's avatar
Linus Torvalds committed
386 387

	fl->fl_file = filp;
388
	fl->fl_owner = filp;
Linus Torvalds's avatar
Linus Torvalds committed
389 390 391 392 393
	fl->fl_pid = current->tgid;
	fl->fl_flags = FL_FLOCK;
	fl->fl_type = type;
	fl->fl_end = OFFSET_MAX;
	
394
	return fl;
Linus Torvalds's avatar
Linus Torvalds committed
395 396
}

397
static int assign_type(struct file_lock *fl, long type)
Linus Torvalds's avatar
Linus Torvalds committed
398 399 400 401 402 403 404 405 406 407 408 409 410
{
	switch (type) {
	case F_RDLCK:
	case F_WRLCK:
	case F_UNLCK:
		fl->fl_type = type;
		break;
	default:
		return -EINVAL;
	}
	return 0;
}

411 412
static int flock64_to_posix_lock(struct file *filp, struct file_lock *fl,
				 struct flock64 *l)
Linus Torvalds's avatar
Linus Torvalds committed
413 414
{
	switch (l->l_whence) {
415
	case SEEK_SET:
416
		fl->fl_start = 0;
Linus Torvalds's avatar
Linus Torvalds committed
417
		break;
418
	case SEEK_CUR:
419
		fl->fl_start = filp->f_pos;
Linus Torvalds's avatar
Linus Torvalds committed
420
		break;
421
	case SEEK_END:
422
		fl->fl_start = i_size_read(file_inode(filp));
Linus Torvalds's avatar
Linus Torvalds committed
423 424 425 426
		break;
	default:
		return -EINVAL;
	}
427 428 429 430 431
	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
432 433 434

	/* POSIX-1996 leaves the case l->l_len < 0 undefined;
	   POSIX-2001 defines it. */
435
	if (l->l_len > 0) {
436 437 438 439
		if (l->l_len - 1 > OFFSET_MAX - fl->fl_start)
			return -EOVERFLOW;
		fl->fl_end = fl->fl_start + l->l_len - 1;

440
	} else if (l->l_len < 0) {
441
		if (fl->fl_start + l->l_len < 0)
442
			return -EINVAL;
443 444 445 446 447
		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
448 449 450 451 452 453 454 455 456 457
	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);
}

458 459 460 461 462
/* 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
463
{
464 465 466 467 468 469 470 471
	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
472 473 474
}

/* default lease lock manager operations */
475 476
static bool
lease_break_callback(struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
477 478
{
	kill_fasync(&fl->fl_fasync, SIGIO, POLL_MSG);
479
	return false;
Linus Torvalds's avatar
Linus Torvalds committed
480 481
}

482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498
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);
}

499
static const struct lock_manager_operations lease_manager_ops = {
500 501
	.lm_break = lease_break_callback,
	.lm_change = lease_modify,
502
	.lm_setup = lease_setup,
Linus Torvalds's avatar
Linus Torvalds committed
503 504 505 506 507
};

/*
 * Initialize a lease, use the default lock manager operations
 */
508
static int lease_init(struct file *filp, long type, struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
509
 {
510 511 512
	if (assign_type(fl, type) != 0)
		return -EINVAL;

513
	fl->fl_owner = filp;
Linus Torvalds's avatar
Linus Torvalds committed
514 515 516 517 518 519 520 521 522 523 524 525
	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 */
526
static struct file_lock *lease_alloc(struct file *filp, long type)
Linus Torvalds's avatar
Linus Torvalds committed
527 528
{
	struct file_lock *fl = locks_alloc_lock();
529
	int error = -ENOMEM;
Linus Torvalds's avatar
Linus Torvalds committed
530 531

	if (fl == NULL)
532
		return ERR_PTR(error);
Linus Torvalds's avatar
Linus Torvalds committed
533 534

	error = lease_init(filp, type, fl);
535 536
	if (error) {
		locks_free_lock(fl);
537
		return ERR_PTR(error);
538
	}
539
	return fl;
Linus Torvalds's avatar
Linus Torvalds committed
540 541 542 543 544 545 546 547 548 549 550 551 552
}

/* 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.
 */
553
static int posix_same_owner(struct file_lock *fl1, struct file_lock *fl2)
Linus Torvalds's avatar
Linus Torvalds committed
554
{
555
	if (fl1->fl_lmops && fl1->fl_lmops->lm_compare_owner)
Linus Torvalds's avatar
Linus Torvalds committed
556
		return fl2->fl_lmops == fl1->fl_lmops &&
557
			fl1->fl_lmops->lm_compare_owner(fl1, fl2);
Linus Torvalds's avatar
Linus Torvalds committed
558 559 560
	return fl1->fl_owner == fl2->fl_owner;
}

561
/* Must be called with the flc_lock held! */
562
static void locks_insert_global_locks(struct file_lock *fl)
563
{
564 565 566 567
	lg_local_lock(&file_lock_lglock);
	fl->fl_link_cpu = smp_processor_id();
	hlist_add_head(&fl->fl_link, this_cpu_ptr(&file_lock_list));
	lg_local_unlock(&file_lock_lglock);
568 569
}

570
/* Must be called with the flc_lock held! */
571
static void locks_delete_global_locks(struct file_lock *fl)
572
{
573 574
	/*
	 * Avoid taking lock if already unhashed. This is safe since this check
575
	 * is done while holding the flc_lock, and new insertions into the list
576 577 578 579 580
	 * also require that it be held.
	 */
	if (hlist_unhashed(&fl->fl_link))
		return;
	lg_local_lock_cpu(&file_lock_lglock, fl->fl_link_cpu);
581
	hlist_del_init(&fl->fl_link);
582
	lg_local_unlock_cpu(&file_lock_lglock, fl->fl_link_cpu);
583 584
}

585 586 587 588 589 590 591 592
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;
}

593
static void locks_insert_global_blocked(struct file_lock *waiter)
594
{
595
	hash_add(blocked_hash, &waiter->fl_link, posix_owner_key(waiter));
596 597
}

598
static void locks_delete_global_blocked(struct file_lock *waiter)
599
{
600
	hash_del(&waiter->fl_link);
601 602
}

Linus Torvalds's avatar
Linus Torvalds committed
603 604
/* Remove waiter from blocker's block list.
 * When blocker ends up pointing to itself then the list is empty.
605
 *
606
 * Must be called with blocked_lock_lock held.
Linus Torvalds's avatar
Linus Torvalds committed
607
 */
608
static void __locks_delete_block(struct file_lock *waiter)
Linus Torvalds's avatar
Linus Torvalds committed
609
{
610
	locks_delete_global_blocked(waiter);
Linus Torvalds's avatar
Linus Torvalds committed
611 612 613 614
	list_del_init(&waiter->fl_block);
	waiter->fl_next = NULL;
}

615
static void locks_delete_block(struct file_lock *waiter)
Linus Torvalds's avatar
Linus Torvalds committed
616
{
617
	spin_lock(&blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
618
	__locks_delete_block(waiter);
619
	spin_unlock(&blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
620 621 622 623 624 625
}

/* 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.
626
 *
627 628 629 630
 * 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
631
 */
632 633
static void __locks_insert_block(struct file_lock *blocker,
					struct file_lock *waiter)
Linus Torvalds's avatar
Linus Torvalds committed
634
{
635
	BUG_ON(!list_empty(&waiter->fl_block));
Linus Torvalds's avatar
Linus Torvalds committed
636
	waiter->fl_next = blocker;
637
	list_add_tail(&waiter->fl_block, &blocker->fl_block);
638
	if (IS_POSIX(blocker) && !IS_OFDLCK(blocker))
639 640 641
		locks_insert_global_blocked(waiter);
}

642
/* Must be called with flc_lock held. */
643 644 645
static void locks_insert_block(struct file_lock *blocker,
					struct file_lock *waiter)
{
646
	spin_lock(&blocked_lock_lock);
647
	__locks_insert_block(blocker, waiter);
648
	spin_unlock(&blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
649 650
}

651 652 653
/*
 * Wake up processes blocked waiting for blocker.
 *
654
 * Must be called with the inode->flc_lock held!
Linus Torvalds's avatar
Linus Torvalds committed
655 656 657
 */
static void locks_wake_up_blocks(struct file_lock *blocker)
{
658 659
	/*
	 * Avoid taking global lock if list is empty. This is safe since new
660 661 662
	 * 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()
663
	 * after acquiring the blocked_lock_lock.
664 665 666 667
	 */
	if (list_empty(&blocker->fl_block))
		return;

668
	spin_lock(&blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
669
	while (!list_empty(&blocker->fl_block)) {
670 671 672
		struct file_lock *waiter;

		waiter = list_first_entry(&blocker->fl_block,
Linus Torvalds's avatar
Linus Torvalds committed
673 674
				struct file_lock, fl_block);
		__locks_delete_block(waiter);
675 676
		if (waiter->fl_lmops && waiter->fl_lmops->lm_notify)
			waiter->fl_lmops->lm_notify(waiter);
Linus Torvalds's avatar
Linus Torvalds committed
677 678 679
		else
			wake_up(&waiter->fl_wait);
	}
680
	spin_unlock(&blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
681 682
}

683
static void
684 685
locks_insert_lock_ctx(struct file_lock *fl, int *counter,
		      struct list_head *before)
686 687 688
{
	fl->fl_nspid = get_pid(task_tgid(current));
	list_add_tail(&fl->fl_list, before);
689
	++*counter;
690 691 692
	locks_insert_global_locks(fl);
}

693
static void
694
locks_unlink_lock_ctx(struct file_lock *fl, int *counter)
Linus Torvalds's avatar
Linus Torvalds committed
695
{
696
	locks_delete_global_locks(fl);
697
	list_del_init(&fl->fl_list);
698
	--*counter;
699 700 701 702
	if (fl->fl_nspid) {
		put_pid(fl->fl_nspid);
		fl->fl_nspid = NULL;
	}
Linus Torvalds's avatar
Linus Torvalds committed
703
	locks_wake_up_blocks(fl);
704 705
}

706
static void
707 708
locks_delete_lock_ctx(struct file_lock *fl, int *counter,
		      struct list_head *dispose)
709
{
710
	locks_unlink_lock_ctx(fl, counter);
711
	if (dispose)
712
		list_add(&fl->fl_list, dispose);
713 714
	else
		locks_free_lock(fl);
Linus Torvalds's avatar
Linus Torvalds committed
715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762
}

/* 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.
	 */
	if (!IS_POSIX(sys_fl) || posix_same_owner(caller_fl, sys_fl))
		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.
	 */
	if (!IS_FLOCK(sys_fl) || (caller_fl->fl_file == sys_fl->fl_file))
		return (0);
	if ((caller_fl->fl_type & LOCK_MAND) || (sys_fl->fl_type & LOCK_MAND))
		return 0;

	return (locks_conflict(caller_fl, sys_fl));
}

763
void
764
posix_test_lock(struct file *filp, struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
765 766
{
	struct file_lock *cfl;
767
	struct file_lock_context *ctx;
768
	struct inode *inode = file_inode(filp);
Linus Torvalds's avatar
Linus Torvalds committed
769

770 771 772 773 774 775
	ctx = inode->i_flctx;
	if (!ctx || list_empty_careful(&ctx->flc_posix)) {
		fl->fl_type = F_UNLCK;
		return;
	}

776
	spin_lock(&ctx->flc_lock);
777 778 779 780 781 782 783
	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
784
	}
785 786
	fl->fl_type = F_UNLCK;
out:
787
	spin_unlock(&ctx->flc_lock);
788
	return;
Linus Torvalds's avatar
Linus Torvalds committed
789 790 791
}
EXPORT_SYMBOL(posix_test_lock);

792 793 794 795 796
/*
 * Deadlock detection:
 *
 * We attempt to detect deadlocks that are due purely to posix file
 * locks.
Linus Torvalds's avatar
Linus Torvalds committed
797
 *
798 799 800 801 802 803 804
 * 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
805
 *
806 807 808
 * 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.
809
 *
810 811 812 813
 * 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.
814
 *
815
 * For FL_OFDLCK locks, the owner is the filp, not the files_struct.
816 817 818 819
 * 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.
 *
820
 * In principle, we could do a more limited deadlock detection on FL_OFDLCK
821 822
 * 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
823
 */
824 825 826

#define MAX_DEADLK_ITERATIONS 10

827 828 829 830 831
/* 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;

832
	hash_for_each_possible(blocked_hash, fl, fl_link, posix_owner_key(block_fl)) {
833 834 835 836 837 838
		if (posix_same_owner(fl, block_fl))
			return fl->fl_next;
	}
	return NULL;
}

839
/* Must be called with the blocked_lock_lock held! */
840
static int posix_locks_deadlock(struct file_lock *caller_fl,
Linus Torvalds's avatar
Linus Torvalds committed
841 842
				struct file_lock *block_fl)
{
843
	int i = 0;
Linus Torvalds's avatar
Linus Torvalds committed
844

845 846
	/*
	 * This deadlock detector can't reasonably detect deadlocks with
847
	 * FL_OFDLCK locks, since they aren't owned by a process, per-se.
848
	 */
849
	if (IS_OFDLCK(caller_fl))
850 851
		return 0;

852 853 854 855 856
	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
857 858 859 860 861
	}
	return 0;
}

/* Try to create a FLOCK lock on filp. We always insert new FLOCK locks
862
 * after any leases, but before any posix locks.
863 864 865 866
 *
 * 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
867
 */
868
static int flock_lock_file(struct file *filp, struct file_lock *request)
Linus Torvalds's avatar
Linus Torvalds committed
869
{
870
	struct file_lock *new_fl = NULL;
871 872 873
	struct file_lock *fl;
	struct file_lock_context *ctx;
	struct inode *inode = file_inode(filp);
Linus Torvalds's avatar
Linus Torvalds committed
874
	int error = 0;
875
	bool found = false;
876
	LIST_HEAD(dispose);
Linus Torvalds's avatar
Linus Torvalds committed
877

878 879 880 881
	ctx = locks_get_lock_context(inode);
	if (!ctx)
		return -ENOMEM;

882
	if (!(request->fl_flags & FL_ACCESS) && (request->fl_type != F_UNLCK)) {
883
		new_fl = locks_alloc_lock();
884 885
		if (!new_fl)
			return -ENOMEM;
886 887
	}

888
	spin_lock(&ctx->flc_lock);
889 890 891
	if (request->fl_flags & FL_ACCESS)
		goto find_conflict;

892
	list_for_each_entry(fl, &ctx->flc_flock, fl_list) {
Linus Torvalds's avatar
Linus Torvalds committed
893 894
		if (filp != fl->fl_file)
			continue;
895
		if (request->fl_type == fl->fl_type)
Linus Torvalds's avatar
Linus Torvalds committed
896
			goto out;
897
		found = true;
898
		locks_delete_lock_ctx(fl, &ctx->flc_flock_cnt, &dispose);
Linus Torvalds's avatar
Linus Torvalds committed
899 900 901
		break;
	}

902 903 904
	if (request->fl_type == F_UNLCK) {
		if ((request->fl_flags & FL_EXISTS) && !found)
			error = -ENOENT;
905
		goto out;
906
	}
Linus Torvalds's avatar
Linus Torvalds committed
907 908 909 910 911

	/*
	 * If a higher-priority process was blocked on the old file lock,
	 * give it the opportunity to lock the file.
	 */
912
	if (found) {
913
		spin_unlock(&ctx->flc_lock);
914
		cond_resched();
915
		spin_lock(&ctx->flc_lock);
916
	}
Linus Torvalds's avatar
Linus Torvalds committed
917

918
find_conflict:
919
	list_for_each_entry(fl, &ctx->flc_flock, fl_list) {
920
		if (!flock_locks_conflict(request, fl))
Linus Torvalds's avatar
Linus Torvalds committed
921 922
			continue;
		error = -EAGAIN;
923 924 925 926
		if (!(request->fl_flags & FL_SLEEP))
			goto out;
		error = FILE_LOCK_DEFERRED;
		locks_insert_block(fl, request);
Linus Torvalds's avatar
Linus Torvalds committed
927 928
		goto out;
	}
929 930
	if (request->fl_flags & FL_ACCESS)
		goto out;
931
	locks_copy_lock(new_fl, request);
932
	locks_insert_lock_ctx(new_fl, &ctx->flc_flock_cnt, &ctx->flc_flock);
933
	new_fl = NULL;
934
	error = 0;
Linus Torvalds's avatar
Linus Torvalds committed
935 936

out:
937
	spin_unlock(&ctx->flc_lock);
938 939
	if (new_fl)
		locks_free_lock(new_fl);
940
	locks_dispose_list(&dispose);
Linus Torvalds's avatar
Linus Torvalds committed
941 942 943
	return error;
}

944
static int __posix_lock_file(struct inode *inode, struct file_lock *request, struct file_lock *conflock)
Linus Torvalds's avatar
Linus Torvalds committed
945
{
946
	struct file_lock *fl, *tmp;
947 948
	struct file_lock *new_fl = NULL;
	struct file_lock *new_fl2 = NULL;
Linus Torvalds's avatar
Linus Torvalds committed
949 950
	struct file_lock *left = NULL;
	struct file_lock *right = NULL;
951
	struct file_lock_context *ctx;
952 953
	int error;
	bool added = false;
954
	LIST_HEAD(dispose);
Linus Torvalds's avatar
Linus Torvalds committed
955

956 957 958 959
	ctx = locks_get_lock_context(inode);
	if (!ctx)
		return -ENOMEM;

Linus Torvalds's avatar
Linus Torvalds committed
960 961 962
	/*
	 * We may need two file_lock structures for this operation,
	 * so we get them in advance to avoid races.
963 964
	 *
	 * In some cases we can be sure, that no new locks will be needed
Linus Torvalds's avatar
Linus Torvalds committed
965
	 */
966 967 968 969 970 971
	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
972

973
	spin_lock(&ctx->flc_lock);
974 975 976
	/*
	 * New lock request. Walk all POSIX locks and look for conflicts. If
	 * there are any, either return error or put the request on the
977
	 * blocker's list of waiters and the global blocked_hash.
978
	 */
Linus Torvalds's avatar
Linus Torvalds committed
979
	if (request->fl_type != F_UNLCK) {
980
		list_for_each_entry(fl, &ctx->flc_posix, fl_list) {
Linus Torvalds's avatar
Linus Torvalds committed
981 982 983 984
			if (!IS_POSIX(fl))
				continue;
			if (!posix_locks_conflict(request, fl))
				continue;
985
			if (conflock)
986
				locks_copy_conflock(conflock, fl);
Linus Torvalds's avatar
Linus Torvalds committed
987 988 989
			error = -EAGAIN;
			if (!(request->fl_flags & FL_SLEEP))
				goto out;
990 991 992 993
			/*
			 * Deadlock detection and insertion into the blocked
			 * locks list must be done while holding the same lock!
			 */
Linus Torvalds's avatar
Linus Torvalds committed
994
			error = -EDEADLK;
995
			spin_lock(&blocked_lock_lock);
996 997 998 999
			if (likely(!posix_locks_deadlock(request, fl))) {
				error = FILE_LOCK_DEFERRED;
				__locks_insert_block(fl, request);
			}
1000
			spin_unlock(&blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
1001 1002 1003 1004 1005 1006 1007 1008 1009
			goto out;
  		}
  	}

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

1010 1011 1012 1013
	/* 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
1014 1015
	}

1016
	/* Process locks with this owner. */
1017 1018 1019 1020 1021
	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
1022
		if (request->fl_type == fl->fl_type) {
Olaf Kirch's avatar
Olaf Kirch committed
1023 1024 1025 1026
			/* 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
1027
			if (fl->fl_end < request->fl_start - 1)
1028
				continue;
Linus Torvalds's avatar
Linus Torvalds committed
1029 1030 1031
			/* 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
1032
			if (fl->fl_start - 1 > request->fl_end)
Linus Torvalds's avatar
Linus Torvalds committed
1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048
				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) {
1049 1050
				locks_delete_lock_ctx(fl, &ctx->flc_posix_cnt,
							&dispose);
Linus Torvalds's avatar
Linus Torvalds committed
1051 1052 1053
				continue;
			}
			request = fl;
1054
			added = true;
1055
		} else {
Linus Torvalds's avatar
Linus Torvalds committed
1056 1057 1058 1059
			/* Processing for different lock types is a bit
			 * more complex.
			 */
			if (fl->fl_end < request->fl_start)
1060
				continue;
Linus Torvalds's avatar
Linus Torvalds committed
1061 1062 1063
			if (fl->fl_start > request->fl_end)
				break;
			if (request->fl_type == F_UNLCK)
1064
				added = true;
Linus Torvalds's avatar
Linus Torvalds committed
1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078
			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) {
1079 1080
					locks_delete_lock_ctx(fl,
						&ctx->flc_posix_cnt, &dispose);
Linus Torvalds's avatar
Linus Torvalds committed
1081 1082
					continue;
				}
1083 1084 1085 1086 1087 1088
				/*
				 * 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
1089
				 */
1090 1091 1092 1093 1094 1095
				error = -ENOLCK;
				if (!new_fl)
					goto out;
				locks_copy_lock(new_fl, request);
				request = new_fl;
				new_fl = NULL;
1096 1097 1098 1099
				locks_insert_lock_ctx(request,
					&ctx->flc_posix_cnt, &fl->fl_list);
				locks_delete_lock_ctx(fl,
					&ctx->flc_posix_cnt, &dispose);
1100
				added = true;
Linus Torvalds's avatar
Linus Torvalds committed
1101 1102 1103 1104
			}
		}
	}

1105
	/*
1106 1107 1108
	 * 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.
1109 1110 1111 1112 1113
	 */
	error = -ENOLCK; /* "no luck" */
	if (right && left == right && !new_fl2)
		goto out;

Linus Torvalds's avatar
Linus Torvalds committed
1114 1115
	error = 0;
	if (!added) {
1116 1117 1118
		if (request->fl_type == F_UNLCK) {
			if (request->fl_flags & FL_EXISTS)
				error = -ENOENT;
Linus Torvalds's avatar
Linus Torvalds committed
1119
			goto out;
1120
		}
1121 1122 1123 1124 1125

		if (!new_fl) {
			error = -ENOLCK;
			goto out;
		}
Linus Torvalds's avatar
Linus Torvalds committed
1126
		locks_copy_lock(new_fl, request);
1127 1128
		locks_insert_lock_ctx(new_fl, &ctx->flc_posix_cnt,
					&fl->fl_list);
Linus Torvalds's avatar
Linus Torvalds committed
1129 1130 1131 1132 1133 1134 1135 1136 1137 1138
		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);
1139 1140
			locks_insert_lock_ctx(left, &ctx->flc_posix_cnt,
						&fl->fl_list);
Linus Torvalds's avatar
Linus Torvalds committed
1141 1142 1143 1144 1145 1146 1147 1148 1149
		}
		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:
1150
	spin_unlock(&ctx->flc_lock);
Linus Torvalds's avatar
Linus Torvalds committed
1151 1152 1153 1154 1155 1156 1157
	/*
	 * Free any unused locks.
	 */
	if (new_fl)
		locks_free_lock(new_fl);
	if (new_fl2)
		locks_free_lock(new_fl2);
1158
	locks_dispose_list(&dispose);
Linus Torvalds's avatar
Linus Torvalds committed
1159 1160 1161 1162 1163 1164 1165
	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
1166
 * @conflock: Place to return a copy of the conflicting lock, if found.
Linus Torvalds's avatar
Linus Torvalds committed
1167 1168 1169 1170
 *
 * 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
1171 1172 1173 1174
 *
 * 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
1175
 */
1176
int posix_lock_file(struct file *filp, struct file_lock *fl,
1177 1178
			struct file_lock *conflock)
{
Al Viro's avatar
Al Viro committed
1179
	return __posix_lock_file(file_inode(filp), fl, conflock);
Linus Torvalds's avatar
Linus Torvalds committed
1180
}
1181
EXPORT_SYMBOL(posix_lock_file);
Linus Torvalds's avatar
Linus Torvalds committed
1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196

/**
 * posix_lock_file_wait - Apply a POSIX-style lock to a file
 * @filp: The file to apply the lock to
 * @fl: The lock to be applied
 *
 * 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
 */
int posix_lock_file_wait(struct file *filp, struct file_lock *fl)
{
	int error;
	might_sleep ();
	for (;;) {
1197
		error = posix_lock_file(filp, fl, NULL);
1198
		if (error != FILE_LOCK_DEFERRED)
Linus Torvalds's avatar
Linus Torvalds committed
1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212
			break;
		error = wait_event_interruptible(fl->fl_wait, !fl->fl_next);
		if (!error)
			continue;

		locks_delete_block(fl);
		break;
	}
	return error;
}
EXPORT_SYMBOL(posix_lock_file_wait);

/**
 * locks_mandatory_locked - Check for an active lock
1213
 * @file: the file to check
Linus Torvalds's avatar
Linus Torvalds committed
1214 1215 1216 1217
 *
 * Searches the inode's list of locks to find any POSIX locks which conflict.
 * This function is called from locks_verify_locked() only.
 */
1218
int locks_mandatory_locked(struct file *file)
Linus Torvalds's avatar
Linus Torvalds committed
1219
{
1220
	int ret;
1221
	struct inode *inode = file_inode(file);
1222
	struct file_lock_context *ctx;
Linus Torvalds's avatar
Linus Torvalds committed
1223 1224
	struct file_lock *fl;

1225 1226 1227 1228
	ctx = inode->i_flctx;
	if (!ctx || list_empty_careful(&ctx->flc_posix))
		return 0;

Linus Torvalds's avatar
Linus Torvalds committed
1229 1230 1231
	/*
	 * Search the lock list for this inode for any POSIX locks.
	 */
1232
	spin_lock(&ctx->flc_lock);
1233 1234
	ret = 0;
	list_for_each_entry(fl, &ctx->flc_posix, fl_list) {
1235
		if (fl->fl_owner != current->files &&
1236 1237
		    fl->fl_owner != file) {
			ret = -EAGAIN;
Linus Torvalds's avatar
Linus Torvalds committed
1238
			break;
1239
		}
Linus Torvalds's avatar
Linus Torvalds committed
1240
	}
1241
	spin_unlock(&ctx->flc_lock);
1242
	return ret;
Linus Torvalds's avatar
Linus Torvalds committed
1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263
}

/**
 * locks_mandatory_area - Check for a conflicting lock
 * @read_write: %FLOCK_VERIFY_WRITE for exclusive access, %FLOCK_VERIFY_READ
 *		for shared
 * @inode:      the file to check
 * @filp:       how the file was opened (if it was)
 * @offset:     start of area to check
 * @count:      length of area to check
 *
 * Searches the inode's list of locks to find any POSIX locks which conflict.
 * This function is called from rw_verify_area() and
 * locks_verify_truncate().
 */
int locks_mandatory_area(int read_write, struct inode *inode,
			 struct file *filp, loff_t offset,
			 size_t count)
{
	struct file_lock fl;
	int error;
1264
	bool sleep = false;
Linus Torvalds's avatar
Linus Torvalds committed
1265 1266 1267 1268 1269 1270

	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))