locks.c 78.9 KB
Newer Older
1
// SPDX-License-Identifier: GPL-2.0-only
Linus Torvalds's avatar
Linus Torvalds committed
2 3 4 5 6 7 8 9 10 11 12 13 14
/*
 *  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
15
 *
Linus Torvalds's avatar
Linus Torvalds committed
16 17 18
 *  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
19
 *
Linus Torvalds's avatar
Linus Torvalds committed
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
 *  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
45
 *  unlocked).
Linus Torvalds's avatar
Linus Torvalds committed
46 47 48 49 50 51 52 53
 *
 *  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
54
 *  races. Just grep for FIXME to see them.
Linus Torvalds's avatar
Linus Torvalds committed
55 56 57 58 59 60 61 62 63
 *  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
64
 *  See 'Documentation/filesystems/mandatory-locking.txt' for details.
Linus Torvalds's avatar
Linus Torvalds committed
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
 *  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.
116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155
 *
 * Locking conflicts and dependencies:
 * If multiple threads attempt to lock the same byte (or flock the same file)
 * only one can be granted the lock, and other must wait their turn.
 * The first lock has been "applied" or "granted", the others are "waiting"
 * and are "blocked" by the "applied" lock..
 *
 * Waiting and applied locks are all kept in trees whose properties are:
 *
 *	- the root of a tree may be an applied or waiting lock.
 *	- every other node in the tree is a waiting lock that
 *	  conflicts with every ancestor of that node.
 *
 * Every such tree begins life as a waiting singleton which obviously
 * satisfies the above properties.
 *
 * The only ways we modify trees preserve these properties:
 *
 *	1. We may add a new leaf node, but only after first verifying that it
 *	   conflicts with all of its ancestors.
 *	2. We may remove the root of a tree, creating a new singleton
 *	   tree from the root and N new trees rooted in the immediate
 *	   children.
 *	3. If the root of a tree is not currently an applied lock, we may
 *	   apply it (if possible).
 *	4. We may upgrade the root of the tree (either extend its range,
 *	   or upgrade its entire range from read to write).
 *
 * When an applied lock is modified in a way that reduces or downgrades any
 * part of its range, we remove all its children (2 above).  This particularly
 * happens when a lock is unlocked.
 *
 * For each of those child trees we "wake up" the thread which is
 * waiting for the lock so it can continue handling as follows: if the
 * root of the tree applies, we do so (3).  If it doesn't, it must
 * conflict with some applied lock.  We remove (wake up) all of its children
 * (2), and add it is a new leaf to the tree rooted in the applied
 * lock (1).  We then repeat the process recursively with those
 * children.
 *
Linus Torvalds's avatar
Linus Torvalds committed
156 157 158 159
 */

#include <linux/capability.h>
#include <linux/file.h>
Al Viro's avatar
Al Viro committed
160
#include <linux/fdtable.h>
Linus Torvalds's avatar
Linus Torvalds committed
161 162 163 164 165 166
#include <linux/fs.h>
#include <linux/init.h>
#include <linux/security.h>
#include <linux/slab.h>
#include <linux/syscalls.h>
#include <linux/time.h>
167
#include <linux/rcupdate.h>
168
#include <linux/pid_namespace.h>
169
#include <linux/hashtable.h>
170
#include <linux/percpu.h>
Linus Torvalds's avatar
Linus Torvalds committed
171

172 173 174
#define CREATE_TRACE_POINTS
#include <trace/events/filelock.h>

175
#include <linux/uaccess.h>
Linus Torvalds's avatar
Linus Torvalds committed
176 177 178

#define IS_POSIX(fl)	(fl->fl_flags & FL_POSIX)
#define IS_FLOCK(fl)	(fl->fl_flags & FL_FLOCK)
179
#define IS_LEASE(fl)	(fl->fl_flags & (FL_LEASE|FL_DELEG|FL_LAYOUT))
180
#define IS_OFDLCK(fl)	(fl->fl_flags & FL_OFDLCK)
181
#define IS_REMOTELCK(fl)	(fl->fl_pid <= 0)
Linus Torvalds's avatar
Linus Torvalds committed
182

183 184
static bool lease_breaking(struct file_lock *fl)
{
185 186 187 188 189 190 191 192 193 194
	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;
195 196
}

Linus Torvalds's avatar
Linus Torvalds committed
197 198 199
int leases_enable = 1;
int lease_break_time = 45;

200
/*
201
 * The global file_lock_list is only used for displaying /proc/locks, so we
202 203 204 205 206
 * 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.
207
 */
208 209 210 211 212
struct file_lock_list_struct {
	spinlock_t		lock;
	struct hlist_head	hlist;
};
static DEFINE_PER_CPU(struct file_lock_list_struct, file_lock_list);
213
DEFINE_STATIC_PERCPU_RWSEM(file_rwsem);
214

215

216
/*
217
 * The blocked_hash is used to find POSIX lock loops for deadlock detection.
218
 * It is protected by blocked_lock_lock.
219 220 221 222 223 224 225
 *
 * 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.
226
 */
227 228
#define BLOCKED_HASH_BITS	7
static DEFINE_HASHTABLE(blocked_hash, BLOCKED_HASH_BITS);
229

230
/*
231 232
 * This lock protects the blocked_hash. Generally, if you're accessing it, you
 * want to be holding this lock.
233
 *
234 235 236
 * In addition, it also protects the fl->fl_blocked_requests list, and the
 * fl->fl_blocker pointer for file_lock structures that are acting as lock
 * requests (in contrast to those that are acting as records of acquired locks).
237 238
 *
 * Note that when we acquire this lock in order to change the above fields,
239
 * we often hold the flc_lock as well. In certain cases, when reading the fields
240
 * protected by this lock, we can skip acquiring it iff we already hold the
241
 * flc_lock.
242
 */
243
static DEFINE_SPINLOCK(blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
244

245
static struct kmem_cache *flctx_cache __read_mostly;
246
static struct kmem_cache *filelock_cache __read_mostly;
Linus Torvalds's avatar
Linus Torvalds committed
247

248
static struct file_lock_context *
249
locks_get_lock_context(struct inode *inode, int type)
250
{
251
	struct file_lock_context *ctx;
252

253 254 255
	/* paired with cmpxchg() below */
	ctx = smp_load_acquire(&inode->i_flctx);
	if (likely(ctx) || type == F_UNLCK)
256 257
		goto out;

258 259
	ctx = kmem_cache_alloc(flctx_cache, GFP_KERNEL);
	if (!ctx)
260 261
		goto out;

262 263 264 265
	spin_lock_init(&ctx->flc_lock);
	INIT_LIST_HEAD(&ctx->flc_flock);
	INIT_LIST_HEAD(&ctx->flc_posix);
	INIT_LIST_HEAD(&ctx->flc_lease);
266 267 268 269 270

	/*
	 * Assign the pointer if it's not already assigned. If it is, then
	 * free the context we just allocated.
	 */
271 272 273 274
	if (cmpxchg(&inode->i_flctx, NULL, ctx)) {
		kmem_cache_free(flctx_cache, ctx);
		ctx = smp_load_acquire(&inode->i_flctx);
	}
275
out:
276
	trace_locks_get_lock_context(inode, type, ctx);
277
	return ctx;
278 279
}

280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306
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");
	}
}

307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322
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);
}

323
void
324
locks_free_lock_context(struct inode *inode)
325
{
326 327
	struct file_lock_context *ctx = inode->i_flctx;

328 329
	if (unlikely(ctx)) {
		locks_check_ctx_lists(inode);
330 331 332 333
		kmem_cache_free(flctx_cache, ctx);
	}
}

334
static void locks_init_lock_heads(struct file_lock *fl)
335
{
336
	INIT_HLIST_NODE(&fl->fl_link);
337
	INIT_LIST_HEAD(&fl->fl_list);
338 339
	INIT_LIST_HEAD(&fl->fl_blocked_requests);
	INIT_LIST_HEAD(&fl->fl_blocked_member);
340
	init_waitqueue_head(&fl->fl_wait);
341 342
}

Linus Torvalds's avatar
Linus Torvalds committed
343
/* Allocate an empty lock structure. */
344
struct file_lock *locks_alloc_lock(void)
Linus Torvalds's avatar
Linus Torvalds committed
345
{
346
	struct file_lock *fl = kmem_cache_zalloc(filelock_cache, GFP_KERNEL);
347 348

	if (fl)
349
		locks_init_lock_heads(fl);
350 351

	return fl;
Linus Torvalds's avatar
Linus Torvalds committed
352
}
353
EXPORT_SYMBOL_GPL(locks_alloc_lock);
Linus Torvalds's avatar
Linus Torvalds committed
354

355
void locks_release_private(struct file_lock *fl)
356
{
357 358 359 360 361 362
	BUG_ON(waitqueue_active(&fl->fl_wait));
	BUG_ON(!list_empty(&fl->fl_list));
	BUG_ON(!list_empty(&fl->fl_blocked_requests));
	BUG_ON(!list_empty(&fl->fl_blocked_member));
	BUG_ON(!hlist_unhashed(&fl->fl_link));

363 364 365 366 367 368
	if (fl->fl_ops) {
		if (fl->fl_ops->fl_release_private)
			fl->fl_ops->fl_release_private(fl);
		fl->fl_ops = NULL;
	}

369
	if (fl->fl_lmops) {
370 371 372 373
		if (fl->fl_lmops->lm_put_owner) {
			fl->fl_lmops->lm_put_owner(fl->fl_owner);
			fl->fl_owner = NULL;
		}
374 375
		fl->fl_lmops = NULL;
	}
376
}
377
EXPORT_SYMBOL_GPL(locks_release_private);
378

Linus Torvalds's avatar
Linus Torvalds committed
379
/* Free a lock which is not in use. */
380
void locks_free_lock(struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
381
{
382
	locks_release_private(fl);
Linus Torvalds's avatar
Linus Torvalds committed
383 384
	kmem_cache_free(filelock_cache, fl);
}
385
EXPORT_SYMBOL(locks_free_lock);
Linus Torvalds's avatar
Linus Torvalds committed
386

387 388 389 390 391 392
static void
locks_dispose_list(struct list_head *dispose)
{
	struct file_lock *fl;

	while (!list_empty(dispose)) {
393 394
		fl = list_first_entry(dispose, struct file_lock, fl_list);
		list_del_init(&fl->fl_list);
395 396 397 398
		locks_free_lock(fl);
	}
}

Linus Torvalds's avatar
Linus Torvalds committed
399 400
void locks_init_lock(struct file_lock *fl)
{
401 402
	memset(fl, 0, sizeof(struct file_lock));
	locks_init_lock_heads(fl);
Linus Torvalds's avatar
Linus Torvalds committed
403 404 405 406 407 408
}
EXPORT_SYMBOL(locks_init_lock);

/*
 * Initialize a new lock from an existing file_lock structure.
 */
409
void locks_copy_conflock(struct file_lock *new, struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
410 411 412
{
	new->fl_owner = fl->fl_owner;
	new->fl_pid = fl->fl_pid;
413
	new->fl_file = NULL;
Linus Torvalds's avatar
Linus Torvalds committed
414 415 416 417
	new->fl_flags = fl->fl_flags;
	new->fl_type = fl->fl_type;
	new->fl_start = fl->fl_start;
	new->fl_end = fl->fl_end;
418
	new->fl_lmops = fl->fl_lmops;
419
	new->fl_ops = NULL;
420 421 422

	if (fl->fl_lmops) {
		if (fl->fl_lmops->lm_get_owner)
423
			fl->fl_lmops->lm_get_owner(fl->fl_owner);
424
	}
425
}
426
EXPORT_SYMBOL(locks_copy_conflock);
427 428 429

void locks_copy_lock(struct file_lock *new, struct file_lock *fl)
{
430 431
	/* "new" must be a freshly-initialized lock */
	WARN_ON_ONCE(new->fl_ops);
432

433
	locks_copy_conflock(new, fl);
434

435
	new->fl_file = fl->fl_file;
Linus Torvalds's avatar
Linus Torvalds committed
436
	new->fl_ops = fl->fl_ops;
437

438 439 440 441
	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
442 443 444
}
EXPORT_SYMBOL(locks_copy_lock);

445 446 447 448 449 450 451 452 453 454 455 456 457
static void locks_move_blocks(struct file_lock *new, struct file_lock *fl)
{
	struct file_lock *f;

	/*
	 * As ctx->flc_lock is held, new requests cannot be added to
	 * ->fl_blocked_requests, so we don't need a lock to check if it
	 * is empty.
	 */
	if (list_empty(&fl->fl_blocked_requests))
		return;
	spin_lock(&blocked_lock_lock);
	list_splice_init(&fl->fl_blocked_requests, &new->fl_blocked_requests);
458
	list_for_each_entry(f, &new->fl_blocked_requests, fl_blocked_member)
459 460 461 462
		f->fl_blocker = new;
	spin_unlock(&blocked_lock_lock);
}

Linus Torvalds's avatar
Linus Torvalds committed
463 464 465 466 467 468 469 470 471 472 473 474 475 476 477
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. */
478
static struct file_lock *
479
flock_make_lock(struct file *filp, unsigned int cmd, struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
480 481
{
	int type = flock_translate_cmd(cmd);
482

Linus Torvalds's avatar
Linus Torvalds committed
483
	if (type < 0)
484
		return ERR_PTR(type);
485

486 487 488 489 490 491 492
	if (fl == NULL) {
		fl = locks_alloc_lock();
		if (fl == NULL)
			return ERR_PTR(-ENOMEM);
	} else {
		locks_init_lock(fl);
	}
Linus Torvalds's avatar
Linus Torvalds committed
493 494

	fl->fl_file = filp;
495
	fl->fl_owner = filp;
Linus Torvalds's avatar
Linus Torvalds committed
496 497 498 499
	fl->fl_pid = current->tgid;
	fl->fl_flags = FL_FLOCK;
	fl->fl_type = type;
	fl->fl_end = OFFSET_MAX;
500

501
	return fl;
Linus Torvalds's avatar
Linus Torvalds committed
502 503
}

504
static int assign_type(struct file_lock *fl, long type)
Linus Torvalds's avatar
Linus Torvalds committed
505 506 507 508 509 510 511 512 513 514 515 516 517
{
	switch (type) {
	case F_RDLCK:
	case F_WRLCK:
	case F_UNLCK:
		fl->fl_type = type;
		break;
	default:
		return -EINVAL;
	}
	return 0;
}

518 519
static int flock64_to_posix_lock(struct file *filp, struct file_lock *fl,
				 struct flock64 *l)
Linus Torvalds's avatar
Linus Torvalds committed
520 521
{
	switch (l->l_whence) {
522
	case SEEK_SET:
523
		fl->fl_start = 0;
Linus Torvalds's avatar
Linus Torvalds committed
524
		break;
525
	case SEEK_CUR:
526
		fl->fl_start = filp->f_pos;
Linus Torvalds's avatar
Linus Torvalds committed
527
		break;
528
	case SEEK_END:
529
		fl->fl_start = i_size_read(file_inode(filp));
Linus Torvalds's avatar
Linus Torvalds committed
530 531 532 533
		break;
	default:
		return -EINVAL;
	}
534 535 536 537 538
	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
539 540 541

	/* POSIX-1996 leaves the case l->l_len < 0 undefined;
	   POSIX-2001 defines it. */
542
	if (l->l_len > 0) {
543 544 545 546
		if (l->l_len - 1 > OFFSET_MAX - fl->fl_start)
			return -EOVERFLOW;
		fl->fl_end = fl->fl_start + l->l_len - 1;

547
	} else if (l->l_len < 0) {
548
		if (fl->fl_start + l->l_len < 0)
549
			return -EINVAL;
550 551 552 553 554
		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
555 556 557 558 559 560 561 562 563 564
	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);
}

565 566 567 568 569
/* 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
570
{
571 572 573 574 575 576 577 578
	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
579 580 581
}

/* default lease lock manager operations */
582 583
static bool
lease_break_callback(struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
584 585
{
	kill_fasync(&fl->fl_fasync, SIGIO, POLL_MSG);
586
	return false;
Linus Torvalds's avatar
Linus Torvalds committed
587 588
}

589 590 591 592 593 594 595 596 597 598 599 600 601 602
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;

603
	__f_setown(filp, task_pid(current), PIDTYPE_TGID, 0);
604 605
}

606
static const struct lock_manager_operations lease_manager_ops = {
607 608
	.lm_break = lease_break_callback,
	.lm_change = lease_modify,
609
	.lm_setup = lease_setup,
Linus Torvalds's avatar
Linus Torvalds committed
610 611 612 613 614
};

/*
 * Initialize a lease, use the default lock manager operations
 */
615
static int lease_init(struct file *filp, long type, struct file_lock *fl)
616
{
617 618 619
	if (assign_type(fl, type) != 0)
		return -EINVAL;

620
	fl->fl_owner = filp;
Linus Torvalds's avatar
Linus Torvalds committed
621 622 623 624 625 626 627 628 629 630 631 632
	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 */
633
static struct file_lock *lease_alloc(struct file *filp, long type)
Linus Torvalds's avatar
Linus Torvalds committed
634 635
{
	struct file_lock *fl = locks_alloc_lock();
636
	int error = -ENOMEM;
Linus Torvalds's avatar
Linus Torvalds committed
637 638

	if (fl == NULL)
639
		return ERR_PTR(error);
Linus Torvalds's avatar
Linus Torvalds committed
640 641

	error = lease_init(filp, type, fl);
642 643
	if (error) {
		locks_free_lock(fl);
644
		return ERR_PTR(error);
645
	}
646
	return fl;
Linus Torvalds's avatar
Linus Torvalds committed
647 648 649 650 651 652 653 654 655 656 657 658 659
}

/* 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.
 */
660
static int posix_same_owner(struct file_lock *fl1, struct file_lock *fl2)
Linus Torvalds's avatar
Linus Torvalds committed
661 662 663 664
{
	return fl1->fl_owner == fl2->fl_owner;
}

665
/* Must be called with the flc_lock held! */
666
static void locks_insert_global_locks(struct file_lock *fl)
667
{
668 669
	struct file_lock_list_struct *fll = this_cpu_ptr(&file_lock_list);

670 671
	percpu_rwsem_assert_held(&file_rwsem);

672
	spin_lock(&fll->lock);
673
	fl->fl_link_cpu = smp_processor_id();
674 675
	hlist_add_head(&fl->fl_link, &fll->hlist);
	spin_unlock(&fll->lock);
676 677
}

678
/* Must be called with the flc_lock held! */
679
static void locks_delete_global_locks(struct file_lock *fl)
680
{
681 682
	struct file_lock_list_struct *fll;

683 684
	percpu_rwsem_assert_held(&file_rwsem);

685 686
	/*
	 * Avoid taking lock if already unhashed. This is safe since this check
687
	 * is done while holding the flc_lock, and new insertions into the list
688 689 690 691
	 * also require that it be held.
	 */
	if (hlist_unhashed(&fl->fl_link))
		return;
692 693 694

	fll = per_cpu_ptr(&file_lock_list, fl->fl_link_cpu);
	spin_lock(&fll->lock);
695
	hlist_del_init(&fl->fl_link);
696
	spin_unlock(&fll->lock);
697 698
}

699 700 701 702 703 704
static unsigned long
posix_owner_key(struct file_lock *fl)
{
	return (unsigned long)fl->fl_owner;
}

705
static void locks_insert_global_blocked(struct file_lock *waiter)
706
{
707 708
	lockdep_assert_held(&blocked_lock_lock);

709
	hash_add(blocked_hash, &waiter->fl_link, posix_owner_key(waiter));
710 711
}

712
static void locks_delete_global_blocked(struct file_lock *waiter)
713
{
714 715
	lockdep_assert_held(&blocked_lock_lock);

716
	hash_del(&waiter->fl_link);
717 718
}

Linus Torvalds's avatar
Linus Torvalds committed
719 720
/* Remove waiter from blocker's block list.
 * When blocker ends up pointing to itself then the list is empty.
721
 *
722
 * Must be called with blocked_lock_lock held.
Linus Torvalds's avatar
Linus Torvalds committed
723
 */
724
static void __locks_delete_block(struct file_lock *waiter)
Linus Torvalds's avatar
Linus Torvalds committed
725
{
726
	locks_delete_global_blocked(waiter);
727 728
	list_del_init(&waiter->fl_blocked_member);
	waiter->fl_blocker = NULL;
Linus Torvalds's avatar
Linus Torvalds committed
729 730
}

731 732 733 734 735 736 737 738 739 740 741 742 743 744 745
static void __locks_wake_up_blocks(struct file_lock *blocker)
{
	while (!list_empty(&blocker->fl_blocked_requests)) {
		struct file_lock *waiter;

		waiter = list_first_entry(&blocker->fl_blocked_requests,
					  struct file_lock, fl_blocked_member);
		__locks_delete_block(waiter);
		if (waiter->fl_lmops && waiter->fl_lmops->lm_notify)
			waiter->fl_lmops->lm_notify(waiter);
		else
			wake_up(&waiter->fl_wait);
	}
}

746 747 748 749 750 751 752
/**
 *	locks_delete_lock - stop waiting for a file lock
 *	@waiter: the lock which was waiting
 *
 *	lockd/nfsd need to disconnect the lock while working on it.
 */
int locks_delete_block(struct file_lock *waiter)
Linus Torvalds's avatar
Linus Torvalds committed
753
{
754 755
	int status = -ENOENT;

756 757 758 759 760 761 762 763 764 765 766 767 768
	/*
	 * If fl_blocker is NULL, it won't be set again as this thread
	 * "owns" the lock and is the only one that might try to claim
	 * the lock.  So it is safe to test fl_blocker locklessly.
	 * Also if fl_blocker is NULL, this waiter is not listed on
	 * fl_blocked_requests for some lock, so no other request can
	 * be added to the list of fl_blocked_requests for this
	 * request.  So if fl_blocker is NULL, it is safe to
	 * locklessly check if fl_blocked_requests is empty.  If both
	 * of these checks succeed, there is no need to take the lock.
	 */
	if (waiter->fl_blocker == NULL &&
	    list_empty(&waiter->fl_blocked_requests))
769
		return status;
770
	spin_lock(&blocked_lock_lock);
771 772
	if (waiter->fl_blocker)
		status = 0;
773
	__locks_wake_up_blocks(waiter);
Linus Torvalds's avatar
Linus Torvalds committed
774
	__locks_delete_block(waiter);
775
	spin_unlock(&blocked_lock_lock);
776
	return status;
Linus Torvalds's avatar
Linus Torvalds committed
777
}
778
EXPORT_SYMBOL(locks_delete_block);
Linus Torvalds's avatar
Linus Torvalds committed
779 780 781 782 783

/* 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.
784
 *
785
 * Must be called with both the flc_lock and blocked_lock_lock held. The
786 787 788 789
 * fl_blocked_requests 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_blocked_requests list is empty.
790 791 792 793
 *
 * Rather than just adding to the list, we check for conflicts with any existing
 * waiters, and add beneath any waiter that blocks the new waiter.
 * Thus wakeups don't happen until needed.
Linus Torvalds's avatar
Linus Torvalds committed
794
 */
795
static void __locks_insert_block(struct file_lock *blocker,
796 797 798
				 struct file_lock *waiter,
				 bool conflict(struct file_lock *,
					       struct file_lock *))
Linus Torvalds's avatar
Linus Torvalds committed
799
{
800
	struct file_lock *fl;
801
	BUG_ON(!list_empty(&waiter->fl_blocked_member));
802 803 804 805 806 807 808

new_blocker:
	list_for_each_entry(fl, &blocker->fl_blocked_requests, fl_blocked_member)
		if (conflict(fl, waiter)) {
			blocker =  fl;
			goto new_blocker;
		}
809 810
	waiter->fl_blocker = blocker;
	list_add_tail(&waiter->fl_blocked_member, &blocker->fl_blocked_requests);
811
	if (IS_POSIX(blocker) && !IS_OFDLCK(blocker))
812
		locks_insert_global_blocked(waiter);
813 814 815 816 817 818

	/* The requests in waiter->fl_blocked are known to conflict with
	 * waiter, but might not conflict with blocker, or the requests
	 * and lock which block it.  So they all need to be woken.
	 */
	__locks_wake_up_blocks(waiter);
819 820
}

821
/* Must be called with flc_lock held. */
822
static void locks_insert_block(struct file_lock *blocker,
823 824 825
			       struct file_lock *waiter,
			       bool conflict(struct file_lock *,
					     struct file_lock *))
826
{
827
	spin_lock(&blocked_lock_lock);
828
	__locks_insert_block(blocker, waiter, conflict);
829
	spin_unlock(&blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
830 831
}

832 833 834
/*
 * Wake up processes blocked waiting for blocker.
 *
835
 * Must be called with the inode->flc_lock held!
Linus Torvalds's avatar
Linus Torvalds committed
836 837 838
 */
static void locks_wake_up_blocks(struct file_lock *blocker)
{
839 840
	/*
	 * Avoid taking global lock if list is empty. This is safe since new
841
	 * blocked requests are only added to the list under the flc_lock, and
842 843 844
	 * the flc_lock is always held here. Note that removal from the
	 * fl_blocked_requests list does not require the flc_lock, so we must
	 * recheck list_empty() after acquiring the blocked_lock_lock.
845
	 */
846
	if (list_empty(&blocker->fl_blocked_requests))
847 848
		return;

849
	spin_lock(&blocked_lock_lock);
850
	__locks_wake_up_blocks(blocker);
851
	spin_unlock(&blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
852 853
}

854
static void
855
locks_insert_lock_ctx(struct file_lock *fl, struct list_head *before)
856 857 858 859 860
{
	list_add_tail(&fl->fl_list, before);
	locks_insert_global_locks(fl);
}

861
static void
862
locks_unlink_lock_ctx(struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
863
{
864
	locks_delete_global_locks(fl);
865
	list_del_init(&fl->fl_list);
Linus Torvalds's avatar
Linus Torvalds committed
866
	locks_wake_up_blocks(fl);
867 868
}

869
static void
870
locks_delete_lock_ctx(struct file_lock *fl, struct list_head *dispose)
871
{
872
	locks_unlink_lock_ctx(fl);
873
	if (dispose)
874
		list_add(&fl->fl_list, dispose);
875 876
	else
		locks_free_lock(fl);
Linus Torvalds's avatar
Linus Torvalds committed
877 878 879 880 881
}

/* Determine if lock sys_fl blocks lock caller_fl. Common functionality
 * checks for shared/exclusive status of overlapping locks.
 */
882 883
static bool locks_conflict(struct file_lock *caller_fl,
			   struct file_lock *sys_fl)
Linus Torvalds's avatar
Linus Torvalds committed
884 885
{
	if (sys_fl->fl_type == F_WRLCK)
886
		return true;
Linus Torvalds's avatar
Linus Torvalds committed
887
	if (caller_fl->fl_type == F_WRLCK)
888 889
		return true;
	return false;
Linus Torvalds's avatar
Linus Torvalds committed
890 891 892 893 894
}

/* Determine if lock sys_fl blocks lock caller_fl. POSIX specific
 * checking before calling the locks_conflict().
 */
895 896
static bool posix_locks_conflict(struct file_lock *caller_fl,
				 struct file_lock *sys_fl)
Linus Torvalds's avatar
Linus Torvalds committed
897 898 899 900
{
	/* POSIX locks owned by the same process do not conflict with
	 * each other.
	 */
901
	if (posix_same_owner(caller_fl, sys_fl))
902
		return false;
Linus Torvalds's avatar
Linus Torvalds committed
903 904 905

	/* Check whether they overlap */
	if (!locks_overlap(caller_fl, sys_fl))
906
		return false;
Linus Torvalds's avatar
Linus Torvalds committed
907

908
	return locks_conflict(caller_fl, sys_fl);
Linus Torvalds's avatar
Linus Torvalds committed
909 910 911 912 913
}

/* Determine if lock sys_fl blocks lock caller_fl. FLOCK specific
 * checking before calling the locks_conflict().
 */
914 915
static bool flock_locks_conflict(struct file_lock *caller_fl,
				 struct file_lock *sys_fl)
Linus Torvalds's avatar
Linus Torvalds committed
916 917 918 919
{
	/* FLOCK locks referring to the same filp do not conflict with
	 * each other.
	 */
920
	if (caller_fl->fl_file == sys_fl->fl_file)
921
		return false;
Linus Torvalds's avatar
Linus Torvalds committed
922
	if ((caller_fl->fl_type & LOCK_MAND) || (sys_fl->fl_type & LOCK_MAND))
923
		return false;
Linus Torvalds's avatar
Linus Torvalds committed
924

925
	return locks_conflict(caller_fl, sys_fl);
Linus Torvalds's avatar
Linus Torvalds committed
926 927
}

928
void
929
posix_test_lock(struct file *filp, struct file_lock *fl)
Linus Torvalds's avatar
Linus Torvalds committed
930 931
{
	struct file_lock *cfl;
932
	struct file_lock_context *ctx;
933
	struct inode *inode = locks_inode(filp);
Linus Torvalds's avatar
Linus Torvalds committed
934

935
	ctx = smp_load_acquire(&inode->i_flctx);
936 937 938 939 940
	if (!ctx || list_empty_careful(&ctx->flc_posix)) {
		fl->fl_type = F_UNLCK;
		return;
	}

941
	spin_lock(&ctx->flc_lock);
942 943 944 945 946
	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
947
	}
948 949
	fl->fl_type = F_UNLCK;
out:
950
	spin_unlock(&ctx->flc_lock);
951
	return;
Linus Torvalds's avatar
Linus Torvalds committed
952 953 954
}
EXPORT_SYMBOL(posix_test_lock);

955 956 957 958 959
/*
 * Deadlock detection:
 *
 * We attempt to detect deadlocks that are due purely to posix file
 * locks.
Linus Torvalds's avatar
Linus Torvalds committed
960
 *
961 962 963 964 965 966 967
 * 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
968
 *
969 970 971
 * 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.
972
 *
973 974 975 976
 * 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.
977
 *
978
 * For FL_OFDLCK locks, the owner is the filp, not the files_struct.
979 980 981 982
 * 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.
 *
983
 * In principle, we could do a more limited deadlock detection on FL_OFDLCK
984 985
 * 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
986
 */
987 988 989

#define MAX_DEADLK_ITERATIONS 10

990 991 992 993 994
/* 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;

995
	hash_for_each_possible(blocked_hash, fl, fl_link, posix_owner_key(block_fl)) {
996 997 998 999 1000
		if (posix_same_owner(fl, block_fl)) {
			while (fl->fl_blocker)
				fl = fl->fl_blocker;
			return fl;
		}
1001 1002 1003 1004
	}
	return NULL;
}

1005
/* Must be called with the blocked_lock_lock held! */
1006
static int posix_locks_deadlock(struct file_lock *caller_fl,
Linus Torvalds's avatar
Linus Torvalds committed
1007 1008
				struct file_lock *block_fl)
{
1009
	int i = 0;
Linus Torvalds's avatar
Linus Torvalds committed
1010

1011 1012
	lockdep_assert_held(&blocked_lock_lock);

1013 1014
	/*
	 * This deadlock detector can't reasonably detect deadlocks with
1015
	 * FL_OFDLCK locks, since they aren't owned by a process, per-se.
1016
	 */
1017
	if (IS_OFDLCK(caller_fl))
1018 1019
		return 0;

1020 1021 1022 1023 1024
	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
1025 1026 1027 1028 1029
	}
	return 0;
}

/* Try to create a FLOCK lock on filp. We always insert new FLOCK locks
1030
 * after any leases, but before any posix locks.
1031 1032 1033 1034
 *
 * 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
1035
 */
1036
static int flock_lock_inode(struct inode *inode, struct file_lock *request)
Linus Torvalds's avatar
Linus Torvalds committed
1037
{
1038
	struct file_lock *new_fl = NULL;
1039 1040
	struct file_lock *fl;
	struct file_lock_context *ctx;
Linus Torvalds's avatar
Linus Torvalds committed
1041
	int error = 0;
1042
	bool found = false;
1043
	LIST_HEAD(dispose);
Linus Torvalds's avatar
Linus Torvalds committed
1044

1045 1046 1047 1048 1049 1050
	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;
	}
1051

1052
	if (!(request->fl_flags & FL_ACCESS) && (request->fl_type != F_UNLCK)) {
1053
		new_fl = locks_alloc_lock();
1054 1055
		if (!new_fl)
			return -ENOMEM;
1056 1057
	}

1058
	percpu_down_read(&file_rwsem);
1059
	spin_lock(&ctx->flc_lock);
1060 1061 1062
	if (request->fl_flags & FL_ACCESS)
		goto find_conflict;

1063
	list_for_each_entry(fl, &ctx->flc_flock, fl_list) {
1064
		if (request->fl_file != fl->fl_file)
Linus Torvalds's avatar
Linus Torvalds committed
1065
			continue;
1066
		if (request->fl_type == fl->fl_type)
Linus Torvalds's avatar
Linus Torvalds committed
1067
			goto out;
1068
		found = true;
1069
		locks_delete_lock_ctx(fl, &dispose);
Linus Torvalds's avatar
Linus Torvalds committed
1070 1071 1072
		break;
	}

1073 1074 1075
	if (request->fl_type == F_UNLCK) {
		if ((request->fl_flags & FL_EXISTS) && !found)
			error = -ENOENT;
1076
		goto out;
1077
	}
Linus Torvalds's avatar
Linus Torvalds committed
1078

1079
find_conflict:
1080
	list_for_each_entry(fl, &ctx->flc_flock, fl_list) {
1081
		if (!flock_locks_conflict(request, fl))
Linus Torvalds's avatar
Linus Torvalds committed
1082 1083
			continue;
		error = -EAGAIN;
1084 1085 1086
		if (!(request->fl_flags & FL_SLEEP))
			goto out;
		error = FILE_LOCK_DEFERRED;
1087
		locks_insert_block(fl, request, flock_locks_conflict);
Linus Torvalds's avatar
Linus Torvalds committed
1088 1089
		goto out;
	}
1090 1091
	if (request->fl_flags & FL_ACCESS)
		goto out;
1092
	locks_copy_lock(new_fl, request);
1093
	locks_move_blocks(new_fl, request);
1094
	locks_insert_lock_ctx(new_fl, &ctx->flc_flock);
1095
	new_fl = NULL;
1096
	error = 0;
Linus Torvalds's avatar
Linus Torvalds committed
1097 1098

out:
1099
	spin_unlock(&ctx->flc_lock);
1100
	percpu_up_read(&file_rwsem);
1101 1102
	if (new_fl)
		locks_free_lock(new_fl);
1103
	locks_dispose_list(&dispose);
1104
	trace_flock_lock_inode(inode, request, error);
Linus Torvalds's avatar
Linus Torvalds committed
1105 1106 1107
	return error;
}

1108 1109
static int posix_lock_inode(struct inode *inode, struct file_lock *request,
			    struct file_lock *conflock)
Linus Torvalds's avatar
Linus Torvalds committed
1110
{
1111
	struct file_lock *fl, *tmp;
1112 1113
	struct file_lock *new_fl = NULL;
	struct file_lock *new_fl2 = NULL;
Linus Torvalds's avatar
Linus Torvalds committed
1114 1115
	struct file_lock *left = NULL;
	struct file_lock *right = NULL;
1116
	struct file_lock_context *ctx;
1117 1118
	int error;
	bool added = false;
1119
	LIST_HEAD(dispose);
Linus Torvalds's avatar
Linus Torvalds committed
1120

1121
	ctx = locks_get_lock_context(inode, request->fl_type);
1122
	if (!ctx)
1123
		return (request->fl_type == F_UNLCK) ? 0 : -ENOMEM;
1124

Linus Torvalds's avatar
Linus Torvalds committed
1125 1126 1127
	/*
	 * We may need two file_lock structures for this operation,
	 * so we get them in advance to avoid races.
1128 1129
	 *
	 * In some cases we can be sure, that no new locks will be needed
Linus Torvalds's avatar
Linus Torvalds committed
1130
	 */
1131 1132 1133 1134 1135 1136
	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
1137

1138
	percpu_down_read(&file_rwsem);
1139
	spin_lock(&ctx->flc_lock);
1140 1141 1142
	/*
	 * New lock request. Walk all POSIX locks and look for conflicts. If
	 * there are any, either return error or put the request on the
1143
	 * blocker's list of waiters and the global blocked_hash.
1144
	 */
Linus Torvalds's avatar
Linus Torvalds committed
1145
	if (request->fl_type != F_UNLCK) {
1146
		list_for_each_entry(fl, &ctx->flc_posix, fl_list) {
Linus Torvalds's avatar
Linus Torvalds committed
1147 1148
			if (!posix_locks_conflict(request, fl))
				continue;
1149
			if (conflock)
1150
				locks_copy_conflock(conflock, fl);
Linus Torvalds's avatar
Linus Torvalds committed
1151 1152 1153
			error = -EAGAIN;
			if (!(request->fl_flags & FL_SLEEP))
				goto out;
1154 1155 1156 1157
			/*
			 * Deadlock detection and insertion into the blocked
			 * locks list must be done while holding the same lock!
			 */
Linus Torvalds's avatar
Linus Torvalds committed
1158
			error = -EDEADLK;
1159
			spin_lock(&blocked_lock_lock);
1160 1161 1162 1163 1164
			/*
			 * Ensure that we don't find any locks blocked on this
			 * request during deadlock detection.
			 */
			__locks_wake_up_blocks(request);
1165 1166
			if (likely(!posix_locks_deadlock(request, fl))) {
				error = FILE_LOCK_DEFERRED;
1167 1168
				__locks_insert_block(fl, request,
						     posix_locks_conflict);
1169
			}
1170
			spin_unlock(&blocked_lock_lock);
Linus Torvalds's avatar
Linus Torvalds committed
1171
			goto out;
1172 1173
		}
	}
Linus Torvalds's avatar
Linus Torvalds committed
1174 1175 1176 1177 1178 1179

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

1180 1181 1182 1183
	/* 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
1184 1185
	}

1186
	/* Process locks with this owner. */
1187 1188 1189 1190 1191
	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
1192
		if (request->fl_type == fl->fl_type) {
1193 1194 1195 1196
			/* 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
1197
			if (fl->fl_end < request->fl_start - 1)
1198
				continue;
Linus Torvalds's avatar
Linus Torvalds committed
1199 1200 1201
			/* If the next lock in the list has entirely bigger
			 * addresses than the new one, insert the lock here.
			 */
1202
			if (fl->fl_start - 1 > request->fl_end)
Linus Torvalds's avatar
Linus Torvalds committed
1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218
				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) {
1219
				locks_delete_lock_ctx(fl, &dispose);
Linus Torvalds's avatar
Linus Torvalds committed
1220 1221 1222
				continue;
			}
			request = fl;
1223
			added = true;
1224
		} else {
Linus Torvalds's avatar
Linus Torvalds committed
1225 1226 1227 1228
			/* Processing for different lock types is a bit
			 * more complex.
			 */
			if (fl->fl_end < request->fl_start)
1229
				continue;
Linus Torvalds's avatar
Linus Torvalds committed
1230 1231 1232
			if (fl->fl_start > request->fl_end)
				break;
			if (request->fl_type == F_UNLCK)
1233
				added = true;
Linus Torvalds's avatar
Linus Torvalds committed
1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247
			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) {
1248
					locks_delete_lock_ctx(fl, &dispose);
Linus Torvalds's avatar
Linus Torvalds committed
1249 1250
					continue;
				}
1251 1252 1253 1254 1255 1256
				/*
				 * 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
1257
				 */
1258 1259 1260 1261 1262 1263
				error = -ENOLCK;
				if (!new_fl)
					goto out;
				locks_copy_lock(new_fl, request);
				request = new_fl;
				new_fl = NULL;
1264 1265
				locks_insert_lock_ctx(request, &fl->fl_list);
				locks_delete_lock_ctx(fl, &dispose);
1266
				added = true;
Linus Torvalds's avatar
Linus Torvalds committed
1267 1268 1269 1270
			}
		}
	}

1271
	/*
1272 1273 1274
	 * 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.
1275 1276 1277 1278 1279
	 */
	error = -ENOLCK; /* "no luck" */
	if (right && left == right && !new_fl2)
		goto out;

Linus Torvalds's avatar
Linus Torvalds committed
1280 1281
	error = 0;
	if (!added) {
1282 1283 1284
		if (request->fl_type == F_UNLCK) {
			if (request->fl_flags & FL_EXISTS)
				error = -ENOENT;
Linus Torvalds's avatar
Linus Torvalds committed
1285
			goto out;
1286
		}
1287 1288 1289 1290 1291

		if (!new_fl) {
			error = -ENOLCK;
			goto out;
		}
Linus Torvalds's avatar
Linus Torvalds committed
1292
		locks_copy_lock(new_fl, request);
1293
		locks_move_blocks(new_fl, request);
1294
		locks_insert_lock_ctx(new_fl, &fl->fl_list);
1295
		fl = new_fl;
Linus Torvalds's avatar
Linus Torvalds committed
1296 1297 1298 1299 1300 1301 1302 1303 1304 1305
		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);
1306
			locks_insert_lock_ctx(left, &fl->fl_list);
Linus Torvalds's avatar
Linus Torvalds committed
1307 1308 1309 1310 1311 1312 1313 1314 1315
		}
		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:
1316
	spin_unlock(&ctx->flc_lock);
1317
	percpu_up_read(&file_rwsem);
Linus Torvalds's avatar
Linus Torvalds committed
1318 1319 1320 1321 1322 1323 1324
	/*
	 * Free any unused locks.
	 */
	if (new_fl)
		locks_free_lock(new_fl);
	if (new_fl2)
		locks_free_lock(new_fl2);
1325
	locks_dispose_list(&dispose);