sem.c 46.6 KB
Newer Older
Linus Torvalds's avatar
Linus Torvalds committed
1
2
3
4
5
6
7
8
/*
 * linux/ipc/sem.c
 * Copyright (C) 1992 Krishna Balasubramanian
 * Copyright (C) 1995 Eric Schenk, Bruno Haible
 *
 * /proc/sysvipc/sem support (c) 1999 Dragos Acostachioaie <dragos@iname.com>
 *
 * SMP-threaded, sysctl's added
9
 * (c) 1999 Manfred Spraul <manfred@colorfullife.com>
Linus Torvalds's avatar
Linus Torvalds committed
10
 * Enforced range limit on SEM_UNDO
Alan Cox's avatar
Alan Cox committed
11
 * (c) 2001 Red Hat Inc
Linus Torvalds's avatar
Linus Torvalds committed
12
13
 * Lockless wakeup
 * (c) 2003 Manfred Spraul <manfred@colorfullife.com>
14
15
 * Further wakeup optimizations, documentation
 * (c) 2010 Manfred Spraul <manfred@colorfullife.com>
Steve Grubb's avatar
Steve Grubb committed
16
17
18
 *
 * support for audit of ipc object properties and permission changes
 * Dustin Kirkland <dustin.kirkland@us.ibm.com>
Kirill Korotaev's avatar
Kirill Korotaev committed
19
20
21
22
 *
 * namespaces support
 * OpenVZ, SWsoft Inc.
 * Pavel Emelianov <xemul@openvz.org>
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
63
64
65
66
67
68
69
70
71
72
73
 *
 * Implementation notes: (May 2010)
 * This file implements System V semaphores.
 *
 * User space visible behavior:
 * - FIFO ordering for semop() operations (just FIFO, not starvation
 *   protection)
 * - multiple semaphore operations that alter the same semaphore in
 *   one semop() are handled.
 * - sem_ctime (time of last semctl()) is updated in the IPC_SET, SETVAL and
 *   SETALL calls.
 * - two Linux specific semctl() commands: SEM_STAT, SEM_INFO.
 * - undo adjustments at process exit are limited to 0..SEMVMX.
 * - namespace are supported.
 * - SEMMSL, SEMMNS, SEMOPM and SEMMNI can be configured at runtine by writing
 *   to /proc/sys/kernel/sem.
 * - statistics about the usage are reported in /proc/sysvipc/sem.
 *
 * Internals:
 * - scalability:
 *   - all global variables are read-mostly.
 *   - semop() calls and semctl(RMID) are synchronized by RCU.
 *   - most operations do write operations (actually: spin_lock calls) to
 *     the per-semaphore array structure.
 *   Thus: Perfect SMP scaling between independent semaphore arrays.
 *         If multiple semaphores in one array are used, then cache line
 *         trashing on the semaphore array spinlock will limit the scaling.
 * - semncnt and semzcnt are calculated on demand in count_semncnt() and
 *   count_semzcnt()
 * - the task that performs a successful semop() scans the list of all
 *   sleeping tasks and completes any pending operations that can be fulfilled.
 *   Semaphores are actively given to waiting tasks (necessary for FIFO).
 *   (see update_queue())
 * - To improve the scalability, the actual wake-up calls are performed after
 *   dropping all locks. (see wake_up_sem_queue_prepare(),
 *   wake_up_sem_queue_do())
 * - All work is done by the waker, the woken up task does not have to do
 *   anything - not even acquiring a lock or dropping a refcount.
 * - A woken up task may not even touch the semaphore array anymore, it may
 *   have been destroyed already by a semctl(RMID).
 * - The synchronizations between wake-ups due to a timeout/signal and a
 *   wake-up due to a completed semaphore operation is achieved by using an
 *   intermediate state (IN_WAKEUP).
 * - UNDO values are stored in an array (one per process and per
 *   semaphore array, lazily allocated). For backwards compatibility, multiple
 *   modes for the UNDO variables are supported (per process, per thread)
 *   (see copy_semundo, CLONE_SYSVSEM)
 * - There are two lists of the pending operations: a per-array list
 *   and per-semaphore list (stored in the array). This allows to achieve FIFO
 *   ordering without always scanning all pending operations.
 *   The worst-case behavior is nevertheless O(N^2) for N wakeups.
Linus Torvalds's avatar
Linus Torvalds committed
74
75
76
77
78
79
80
81
82
83
 */

#include <linux/slab.h>
#include <linux/spinlock.h>
#include <linux/init.h>
#include <linux/proc_fs.h>
#include <linux/time.h>
#include <linux/security.h>
#include <linux/syscalls.h>
#include <linux/audit.h>
84
#include <linux/capability.h>
85
#include <linux/seq_file.h>
Nadia Derbey's avatar
Nadia Derbey committed
86
#include <linux/rwsem.h>
Kirill Korotaev's avatar
Kirill Korotaev committed
87
#include <linux/nsproxy.h>
88
#include <linux/ipc_namespace.h>
Ingo Molnar's avatar
Ingo Molnar committed
89

Linus Torvalds's avatar
Linus Torvalds committed
90
91
92
#include <asm/uaccess.h>
#include "util.h"

93
94
95
96
/* One semaphore structure for each semaphore in the system. */
struct sem {
	int	semval;		/* current value */
	int	sempid;		/* pid of last operation */
97
	spinlock_t	lock;	/* spinlock for fine-grained semtimedop */
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
	struct list_head sem_pending; /* pending single-sop operations */
};

/* One queue for each sleeping process in the system. */
struct sem_queue {
	struct list_head	list;	 /* queue of pending operations */
	struct task_struct	*sleeper; /* this process */
	struct sem_undo		*undo;	 /* undo structure */
	int			pid;	 /* process id of requesting process */
	int			status;	 /* completion status of operation */
	struct sembuf		*sops;	 /* array of pending operations */
	int			nsops;	 /* number of operations */
	int			alter;	 /* does *sops alter the array? */
};

/* Each task has a list of undo requests. They are executed automatically
 * when the process exits.
 */
struct sem_undo {
	struct list_head	list_proc;	/* per-process list: *
						 * all undos from one process
						 * rcu protected */
	struct rcu_head		rcu;		/* rcu struct for sem_undo */
	struct sem_undo_list	*ulp;		/* back ptr to sem_undo_list */
	struct list_head	list_id;	/* per semaphore array list:
						 * all undos for one array */
	int			semid;		/* semaphore set identifier */
	short			*semadj;	/* array of adjustments */
						/* one per semaphore */
};

/* sem_undo_list controls shared access to the list of sem_undo structures
 * that may be shared among all a CLONE_SYSVSEM task group.
 */
struct sem_undo_list {
	atomic_t		refcnt;
	spinlock_t		lock;
	struct list_head	list_proc;
};


139
#define sem_ids(ns)	((ns)->ids[IPC_SEM_IDS])
Kirill Korotaev's avatar
Kirill Korotaev committed
140

Nadia Derbey's avatar
Nadia Derbey committed
141
#define sem_checkid(sma, semid)	ipc_checkid(&sma->sem_perm, semid)
Linus Torvalds's avatar
Linus Torvalds committed
142

Nadia Derbey's avatar
Nadia Derbey committed
143
static int newary(struct ipc_namespace *, struct ipc_params *);
144
static void freeary(struct ipc_namespace *, struct kern_ipc_perm *);
Linus Torvalds's avatar
Linus Torvalds committed
145
#ifdef CONFIG_PROC_FS
146
static int sysvipc_sem_proc_show(struct seq_file *s, void *it);
Linus Torvalds's avatar
Linus Torvalds committed
147
148
149
150
151
152
153
154
155
156
157
158
159
160
#endif

#define SEMMSL_FAST	256 /* 512 bytes on stack */
#define SEMOPM_FAST	64  /* ~ 372 bytes on stack */

/*
 * linked list protection:
 *	sem_undo.id_next,
 *	sem_array.sem_pending{,last},
 *	sem_array.sem_undo: sem_lock() for read/write
 *	sem_undo.proc_next: only "current" is allowed to read/write that field.
 *	
 */

Kirill Korotaev's avatar
Kirill Korotaev committed
161
162
163
164
165
#define sc_semmsl	sem_ctls[0]
#define sc_semmns	sem_ctls[1]
#define sc_semopm	sem_ctls[2]
#define sc_semmni	sem_ctls[3]

166
void sem_init_ns(struct ipc_namespace *ns)
Kirill Korotaev's avatar
Kirill Korotaev committed
167
168
169
170
171
172
{
	ns->sc_semmsl = SEMMSL;
	ns->sc_semmns = SEMMNS;
	ns->sc_semopm = SEMOPM;
	ns->sc_semmni = SEMMNI;
	ns->used_sems = 0;
173
	ipc_init_ids(&ns->ids[IPC_SEM_IDS]);
Kirill Korotaev's avatar
Kirill Korotaev committed
174
175
}

176
#ifdef CONFIG_IPC_NS
Kirill Korotaev's avatar
Kirill Korotaev committed
177
178
void sem_exit_ns(struct ipc_namespace *ns)
{
179
	free_ipcs(ns, &sem_ids(ns), freeary);
180
	idr_destroy(&ns->ids[IPC_SEM_IDS].ipcs_idr);
Kirill Korotaev's avatar
Kirill Korotaev committed
181
}
182
#endif
Linus Torvalds's avatar
Linus Torvalds committed
183
184
185

void __init sem_init (void)
{
186
	sem_init_ns(&init_ipc_ns);
187
188
	ipc_init_proc_interface("sysvipc/sem",
				"       key      semid perms      nsems   uid   gid  cuid  cgid      otime      ctime\n",
Kirill Korotaev's avatar
Kirill Korotaev committed
189
				IPC_SEM_IDS, sysvipc_sem_proc_show);
Linus Torvalds's avatar
Linus Torvalds committed
190
191
}

192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
/*
 * If the request contains only one semaphore operation, and there are
 * no complex transactions pending, lock only the semaphore involved.
 * Otherwise, lock the entire semaphore array, since we either have
 * multiple semaphores in our own semops, or we need to look at
 * semaphores from other pending complex operations.
 *
 * Carefully guard against sma->complex_count changing between zero
 * and non-zero while we are spinning for the lock. The value of
 * sma->complex_count cannot change while we are holding the lock,
 * so sem_unlock should be fine.
 *
 * The global lock path checks that all the local locks have been released,
 * checking each local lock once. This means that the local lock paths
 * cannot start their critical sections while the global lock is held.
 */
static inline int sem_lock(struct sem_array *sma, struct sembuf *sops,
			      int nsops)
{
	int locknum;
 again:
	if (nsops == 1 && !sma->complex_count) {
		struct sem *sem = sma->sem_base + sops->sem_num;

		/* Lock just the semaphore we are interested in. */
		spin_lock(&sem->lock);

		/*
		 * If sma->complex_count was set while we were spinning,
		 * we may need to look at things we did not lock here.
		 */
		if (unlikely(sma->complex_count)) {
			spin_unlock(&sem->lock);
			goto lock_array;
		}

		/*
		 * Another process is holding the global lock on the
		 * sem_array; we cannot enter our critical section,
		 * but have to wait for the global lock to be released.
		 */
		if (unlikely(spin_is_locked(&sma->sem_perm.lock))) {
			spin_unlock(&sem->lock);
			spin_unlock_wait(&sma->sem_perm.lock);
			goto again;
		}

		locknum = sops->sem_num;
	} else {
		int i;
		/*
		 * Lock the semaphore array, and wait for all of the
		 * individual semaphore locks to go away.  The code
		 * above ensures no new single-lock holders will enter
		 * their critical section while the array lock is held.
		 */
 lock_array:
		spin_lock(&sma->sem_perm.lock);
		for (i = 0; i < sma->sem_nsems; i++) {
			struct sem *sem = sma->sem_base + i;
			spin_unlock_wait(&sem->lock);
		}
		locknum = -1;
	}
	return locknum;
}

static inline void sem_unlock(struct sem_array *sma, int locknum)
{
	if (locknum == -1) {
		spin_unlock(&sma->sem_perm.lock);
	} else {
		struct sem *sem = sma->sem_base + locknum;
		spin_unlock(&sem->lock);
	}
}

Nadia Derbey's avatar
Nadia Derbey committed
269
270
271
/*
 * sem_lock_(check_) routines are called in the paths where the rw_mutex
 * is not held.
272
273
 *
 * The caller holds the RCU read lock.
Nadia Derbey's avatar
Nadia Derbey committed
274
 */
275
276
static inline struct sem_array *sem_obtain_lock(struct ipc_namespace *ns,
			int id, struct sembuf *sops, int nsops, int *locknum)
277
{
278
279
	struct kern_ipc_perm *ipcp;
	struct sem_array *sma;
Nadia Derbey's avatar
Nadia Derbey committed
280

281
	ipcp = ipc_obtain_object(&sem_ids(ns), id);
282
283
	if (IS_ERR(ipcp))
		return ERR_CAST(ipcp);
284

285
286
	sma = container_of(ipcp, struct sem_array, sem_perm);
	*locknum = sem_lock(sma, sops, nsops);
287
288
289
290
291
292
293

	/* ipc_rmid() may have already freed the ID while sem_lock
	 * was spinning: verify that the structure is still valid
	 */
	if (!ipcp->deleted)
		return container_of(ipcp, struct sem_array, sem_perm);

294
	sem_unlock(sma, *locknum);
295
	return ERR_PTR(-EINVAL);
296
297
}

298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
static inline struct sem_array *sem_obtain_object(struct ipc_namespace *ns, int id)
{
	struct kern_ipc_perm *ipcp = ipc_obtain_object(&sem_ids(ns), id);

	if (IS_ERR(ipcp))
		return ERR_CAST(ipcp);

	return container_of(ipcp, struct sem_array, sem_perm);
}

static inline struct sem_array *sem_obtain_object_check(struct ipc_namespace *ns,
							int id)
{
	struct kern_ipc_perm *ipcp = ipc_obtain_object_check(&sem_ids(ns), id);

	if (IS_ERR(ipcp))
		return ERR_CAST(ipcp);
315

Nadia Derbey's avatar
Nadia Derbey committed
316
	return container_of(ipcp, struct sem_array, sem_perm);
317
318
}

319
320
static inline void sem_lock_and_putref(struct sem_array *sma)
{
321
	sem_lock(sma, NULL, -1);
322
323
324
325
326
	ipc_rcu_putref(sma);
}

static inline void sem_putref(struct sem_array *sma)
{
327
	ipc_rcu_putref(sma);
328
329
}

Nadia Derbey's avatar
Nadia Derbey committed
330
331
332
333
334
static inline void sem_rmid(struct ipc_namespace *ns, struct sem_array *s)
{
	ipc_rmid(&sem_ids(ns), &s->sem_perm);
}

Linus Torvalds's avatar
Linus Torvalds committed
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
/*
 * Lockless wakeup algorithm:
 * Without the check/retry algorithm a lockless wakeup is possible:
 * - queue.status is initialized to -EINTR before blocking.
 * - wakeup is performed by
 *	* unlinking the queue entry from sma->sem_pending
 *	* setting queue.status to IN_WAKEUP
 *	  This is the notification for the blocked thread that a
 *	  result value is imminent.
 *	* call wake_up_process
 *	* set queue.status to the final value.
 * - the previously blocked thread checks queue.status:
 *   	* if it's IN_WAKEUP, then it must wait until the value changes
 *   	* if it's not -EINTR, then the operation was completed by
 *   	  update_queue. semtimedop can return queue.status without
Ingo Molnar's avatar
Ingo Molnar committed
350
 *   	  performing any operation on the sem array.
Linus Torvalds's avatar
Linus Torvalds committed
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
 *   	* otherwise it must acquire the spinlock and check what's up.
 *
 * The two-stage algorithm is necessary to protect against the following
 * races:
 * - if queue.status is set after wake_up_process, then the woken up idle
 *   thread could race forward and try (and fail) to acquire sma->lock
 *   before update_queue had a chance to set queue.status
 * - if queue.status is written before wake_up_process and if the
 *   blocked process is woken up by a signal between writing
 *   queue.status and the wake_up_process, then the woken up
 *   process could return from semtimedop and die by calling
 *   sys_exit before wake_up_process is called. Then wake_up_process
 *   will oops, because the task structure is already invalid.
 *   (yes, this happened on s390 with sysv msg).
 *
 */
#define IN_WAKEUP	1

Nadia Derbey's avatar
Nadia Derbey committed
369
370
371
372
373
/**
 * newary - Create a new semaphore set
 * @ns: namespace
 * @params: ptr to the structure that contains key, semflg and nsems
 *
Nadia Derbey's avatar
Nadia Derbey committed
374
 * Called with sem_ids.rw_mutex held (as a writer)
Nadia Derbey's avatar
Nadia Derbey committed
375
376
 */

Nadia Derbey's avatar
Nadia Derbey committed
377
static int newary(struct ipc_namespace *ns, struct ipc_params *params)
Linus Torvalds's avatar
Linus Torvalds committed
378
379
380
381
382
{
	int id;
	int retval;
	struct sem_array *sma;
	int size;
Nadia Derbey's avatar
Nadia Derbey committed
383
384
385
	key_t key = params->key;
	int nsems = params->u.nsems;
	int semflg = params->flg;
386
	int i;
Linus Torvalds's avatar
Linus Torvalds committed
387
388
389

	if (!nsems)
		return -EINVAL;
Kirill Korotaev's avatar
Kirill Korotaev committed
390
	if (ns->used_sems + nsems > ns->sc_semmns)
Linus Torvalds's avatar
Linus Torvalds committed
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
		return -ENOSPC;

	size = sizeof (*sma) + nsems * sizeof (struct sem);
	sma = ipc_rcu_alloc(size);
	if (!sma) {
		return -ENOMEM;
	}
	memset (sma, 0, size);

	sma->sem_perm.mode = (semflg & S_IRWXUGO);
	sma->sem_perm.key = key;

	sma->sem_perm.security = NULL;
	retval = security_sem_alloc(sma);
	if (retval) {
		ipc_rcu_putref(sma);
		return retval;
	}

Kirill Korotaev's avatar
Kirill Korotaev committed
410
	id = ipc_addid(&sem_ids(ns), &sma->sem_perm, ns->sc_semmni);
411
	if (id < 0) {
Linus Torvalds's avatar
Linus Torvalds committed
412
413
		security_sem_free(sma);
		ipc_rcu_putref(sma);
414
		return id;
Linus Torvalds's avatar
Linus Torvalds committed
415
	}
Kirill Korotaev's avatar
Kirill Korotaev committed
416
	ns->used_sems += nsems;
Linus Torvalds's avatar
Linus Torvalds committed
417
418

	sma->sem_base = (struct sem *) &sma[1];
419

420
	for (i = 0; i < nsems; i++) {
421
		INIT_LIST_HEAD(&sma->sem_base[i].sem_pending);
422
423
		spin_lock_init(&sma->sem_base[i].lock);
	}
424
425

	sma->complex_count = 0;
426
	INIT_LIST_HEAD(&sma->sem_pending);
427
	INIT_LIST_HEAD(&sma->list_id);
Linus Torvalds's avatar
Linus Torvalds committed
428
429
	sma->sem_nsems = nsems;
	sma->sem_ctime = get_seconds();
430
	sem_unlock(sma, -1);
431
	rcu_read_unlock();
Linus Torvalds's avatar
Linus Torvalds committed
432

Nadia Derbey's avatar
Nadia Derbey committed
433
	return sma->sem_perm.id;
Linus Torvalds's avatar
Linus Torvalds committed
434
435
}

Nadia Derbey's avatar
Nadia Derbey committed
436

Nadia Derbey's avatar
Nadia Derbey committed
437
/*
Nadia Derbey's avatar
Nadia Derbey committed
438
 * Called with sem_ids.rw_mutex and ipcp locked.
Nadia Derbey's avatar
Nadia Derbey committed
439
 */
Nadia Derbey's avatar
Nadia Derbey committed
440
static inline int sem_security(struct kern_ipc_perm *ipcp, int semflg)
Nadia Derbey's avatar
Nadia Derbey committed
441
{
Nadia Derbey's avatar
Nadia Derbey committed
442
443
444
445
	struct sem_array *sma;

	sma = container_of(ipcp, struct sem_array, sem_perm);
	return security_sem_associate(sma, semflg);
Nadia Derbey's avatar
Nadia Derbey committed
446
447
}

Nadia Derbey's avatar
Nadia Derbey committed
448
/*
Nadia Derbey's avatar
Nadia Derbey committed
449
 * Called with sem_ids.rw_mutex and ipcp locked.
Nadia Derbey's avatar
Nadia Derbey committed
450
 */
Nadia Derbey's avatar
Nadia Derbey committed
451
452
static inline int sem_more_checks(struct kern_ipc_perm *ipcp,
				struct ipc_params *params)
Nadia Derbey's avatar
Nadia Derbey committed
453
{
Nadia Derbey's avatar
Nadia Derbey committed
454
455
456
457
	struct sem_array *sma;

	sma = container_of(ipcp, struct sem_array, sem_perm);
	if (params->u.nsems > sma->sem_nsems)
Nadia Derbey's avatar
Nadia Derbey committed
458
459
460
461
462
		return -EINVAL;

	return 0;
}

463
SYSCALL_DEFINE3(semget, key_t, key, int, nsems, int, semflg)
Linus Torvalds's avatar
Linus Torvalds committed
464
{
Kirill Korotaev's avatar
Kirill Korotaev committed
465
	struct ipc_namespace *ns;
Nadia Derbey's avatar
Nadia Derbey committed
466
467
	struct ipc_ops sem_ops;
	struct ipc_params sem_params;
Kirill Korotaev's avatar
Kirill Korotaev committed
468
469

	ns = current->nsproxy->ipc_ns;
Linus Torvalds's avatar
Linus Torvalds committed
470

Kirill Korotaev's avatar
Kirill Korotaev committed
471
	if (nsems < 0 || nsems > ns->sc_semmsl)
Linus Torvalds's avatar
Linus Torvalds committed
472
		return -EINVAL;
Nadia Derbey's avatar
Nadia Derbey committed
473

Nadia Derbey's avatar
Nadia Derbey committed
474
475
476
477
478
479
480
	sem_ops.getnew = newary;
	sem_ops.associate = sem_security;
	sem_ops.more_checks = sem_more_checks;

	sem_params.key = key;
	sem_params.flg = semflg;
	sem_params.u.nsems = nsems;
Linus Torvalds's avatar
Linus Torvalds committed
481

Nadia Derbey's avatar
Nadia Derbey committed
482
	return ipcget(ns, &sem_ids(ns), &sem_ops, &sem_params);
Linus Torvalds's avatar
Linus Torvalds committed
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
}

/*
 * Determine whether a sequence of semaphore operations would succeed
 * all at once. Return 0 if yes, 1 if need to sleep, else return error code.
 */

static int try_atomic_semop (struct sem_array * sma, struct sembuf * sops,
			     int nsops, struct sem_undo *un, int pid)
{
	int result, sem_op;
	struct sembuf *sop;
	struct sem * curr;

	for (sop = sops; sop < sops + nsops; sop++) {
		curr = sma->sem_base + sop->sem_num;
		sem_op = sop->sem_op;
		result = curr->semval;
  
		if (!sem_op && result)
			goto would_block;

		result += sem_op;
		if (result < 0)
			goto would_block;
		if (result > SEMVMX)
			goto out_of_range;
		if (sop->sem_flg & SEM_UNDO) {
			int undo = un->semadj[sop->sem_num] - sem_op;
			/*
	 		 *	Exceeding the undo range is an error.
			 */
			if (undo < (-SEMAEM - 1) || undo > SEMAEM)
				goto out_of_range;
		}
		curr->semval = result;
	}

	sop--;
	while (sop >= sops) {
		sma->sem_base[sop->sem_num].sempid = pid;
		if (sop->sem_flg & SEM_UNDO)
			un->semadj[sop->sem_num] -= sop->sem_op;
		sop--;
	}
	
	return 0;

out_of_range:
	result = -ERANGE;
	goto undo;

would_block:
	if (sop->sem_flg & IPC_NOWAIT)
		result = -EAGAIN;
	else
		result = 1;

undo:
	sop--;
	while (sop >= sops) {
		sma->sem_base[sop->sem_num].semval -= sop->sem_op;
		sop--;
	}

	return result;
}

551
552
553
554
555
/** wake_up_sem_queue_prepare(q, error): Prepare wake-up
 * @q: queue entry that must be signaled
 * @error: Error value for the signal
 *
 * Prepare the wake-up of the queue entry q.
Nick Piggin's avatar
Nick Piggin committed
556
 */
557
558
static void wake_up_sem_queue_prepare(struct list_head *pt,
				struct sem_queue *q, int error)
Nick Piggin's avatar
Nick Piggin committed
559
{
560
561
562
563
564
565
566
	if (list_empty(pt)) {
		/*
		 * Hold preempt off so that we don't get preempted and have the
		 * wakee busy-wait until we're scheduled back on.
		 */
		preempt_disable();
	}
Nick Piggin's avatar
Nick Piggin committed
567
	q->status = IN_WAKEUP;
568
569
	q->pid = error;

570
	list_add_tail(&q->list, pt);
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
}

/**
 * wake_up_sem_queue_do(pt) - do the actual wake-up
 * @pt: list of tasks to be woken up
 *
 * Do the actual wake-up.
 * The function is called without any locks held, thus the semaphore array
 * could be destroyed already and the tasks can disappear as soon as the
 * status is set to the actual return code.
 */
static void wake_up_sem_queue_do(struct list_head *pt)
{
	struct sem_queue *q, *t;
	int did_something;

	did_something = !list_empty(pt);
588
	list_for_each_entry_safe(q, t, pt, list) {
589
590
591
592
593
594
595
		wake_up_process(q->sleeper);
		/* q can disappear immediately after writing q->status. */
		smp_wmb();
		q->status = q->pid;
	}
	if (did_something)
		preempt_enable();
Nick Piggin's avatar
Nick Piggin committed
596
597
}

598
599
600
static void unlink_queue(struct sem_array *sma, struct sem_queue *q)
{
	list_del(&q->list);
601
	if (q->nsops > 1)
602
603
604
		sma->complex_count--;
}

605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
/** check_restart(sma, q)
 * @sma: semaphore array
 * @q: the operation that just completed
 *
 * update_queue is O(N^2) when it restarts scanning the whole queue of
 * waiting operations. Therefore this function checks if the restart is
 * really necessary. It is called after a previously waiting operation
 * was completed.
 */
static int check_restart(struct sem_array *sma, struct sem_queue *q)
{
	struct sem *curr;
	struct sem_queue *h;

	/* if the operation didn't modify the array, then no restart */
	if (q->alter == 0)
		return 0;

	/* pending complex operations are too difficult to analyse */
	if (sma->complex_count)
		return 1;

	/* we were a sleeping complex operation. Too difficult */
	if (q->nsops > 1)
		return 1;

	curr = sma->sem_base + q->sops[0].sem_num;

	/* No-one waits on this queue */
	if (list_empty(&curr->sem_pending))
		return 0;

	/* the new semaphore value */
	if (curr->semval) {
		/* It is impossible that someone waits for the new value:
		 * - q is a previously sleeping simple operation that
		 *   altered the array. It must be a decrement, because
		 *   simple increments never sleep.
		 * - The value is not 0, thus wait-for-zero won't proceed.
		 * - If there are older (higher priority) decrements
		 *   in the queue, then they have observed the original
		 *   semval value and couldn't proceed. The operation
		 *   decremented to value - thus they won't proceed either.
		 */
		BUG_ON(q->sops[0].sem_op >= 0);
		return 0;
	}
	/*
	 * semval is 0. Check if there are wait-for-zero semops.
654
	 * They must be the first entries in the per-semaphore queue
655
	 */
656
	h = list_first_entry(&curr->sem_pending, struct sem_queue, list);
657
658
659
660
661
662
663
664
665
666
667
	BUG_ON(h->nsops != 1);
	BUG_ON(h->sops[0].sem_num != q->sops[0].sem_num);

	/* Yes, there is a wait-for-zero semop. Restart */
	if (h->sops[0].sem_op == 0)
		return 1;

	/* Again - no-one is waiting for the new value. */
	return 0;
}

668
669
670
671
672

/**
 * update_queue(sma, semnum): Look for tasks that can be completed.
 * @sma: semaphore array.
 * @semnum: semaphore that was modified.
673
 * @pt: list head for the tasks that must be woken up.
674
675
 *
 * update_queue must be called after a semaphore in a semaphore array
676
677
678
 * was modified. If multiple semaphores were modified, update_queue must
 * be called with semnum = -1, as well as with the number of each modified
 * semaphore.
679
680
681
 * The tasks that must be woken up are added to @pt. The return code
 * is stored in q->pid.
 * The function return 1 if at least one semop was completed successfully.
Linus Torvalds's avatar
Linus Torvalds committed
682
 */
683
static int update_queue(struct sem_array *sma, int semnum, struct list_head *pt)
Linus Torvalds's avatar
Linus Torvalds committed
684
{
685
686
687
	struct sem_queue *q;
	struct list_head *walk;
	struct list_head *pending_list;
688
	int semop_completed = 0;
689

690
	if (semnum == -1)
691
		pending_list = &sma->sem_pending;
692
	else
693
		pending_list = &sma->sem_base[semnum].sem_pending;
694
695

again:
696
697
	walk = pending_list->next;
	while (walk != pending_list) {
698
		int error, restart;
699

700
		q = container_of(walk, struct sem_queue, list);
701
		walk = walk->next;
Linus Torvalds's avatar
Linus Torvalds committed
702

703
704
705
706
707
708
709
710
711
712
713
		/* If we are scanning the single sop, per-semaphore list of
		 * one semaphore and that semaphore is 0, then it is not
		 * necessary to scan the "alter" entries: simple increments
		 * that affect only one entry succeed immediately and cannot
		 * be in the  per semaphore pending queue, and decrements
		 * cannot be successful if the value is already 0.
		 */
		if (semnum != -1 && sma->sem_base[semnum].semval == 0 &&
				q->alter)
			break;

Linus Torvalds's avatar
Linus Torvalds committed
714
715
716
717
		error = try_atomic_semop(sma, q->sops, q->nsops,
					 q->undo, q->pid);

		/* Does q->sleeper still need to sleep? */
718
719
720
		if (error > 0)
			continue;

721
		unlink_queue(sma, q);
722

723
		if (error) {
724
			restart = 0;
725
726
		} else {
			semop_completed = 1;
727
			restart = check_restart(sma, q);
728
		}
729

730
		wake_up_sem_queue_prepare(pt, q, error);
731
		if (restart)
732
			goto again;
Linus Torvalds's avatar
Linus Torvalds committed
733
	}
734
	return semop_completed;
Linus Torvalds's avatar
Linus Torvalds committed
735
736
}

737
738
/**
 * do_smart_update(sma, sops, nsops, otime, pt) - optimized update_queue
739
740
741
 * @sma: semaphore array
 * @sops: operations that were performed
 * @nsops: number of operations
742
743
 * @otime: force setting otime
 * @pt: list head of the tasks that must be woken up.
744
745
746
 *
 * do_smart_update() does the required called to update_queue, based on the
 * actual changes that were performed on the semaphore array.
747
748
749
 * Note that the function does not do the actual wake-up: the caller is
 * responsible for calling wake_up_sem_queue_do(@pt).
 * It is safe to perform this call after dropping all locks.
750
 */
751
752
static void do_smart_update(struct sem_array *sma, struct sembuf *sops, int nsops,
			int otime, struct list_head *pt)
753
754
755
756
{
	int i;

	if (sma->complex_count || sops == NULL) {
757
758
		if (update_queue(sma, -1, pt))
			otime = 1;
759
760
761
762
763
764
765
766
	}

	if (!sops) {
		/* No semops; something special is going on. */
		for (i = 0; i < sma->sem_nsems; i++) {
			if (update_queue(sma, i, pt))
				otime = 1;
		}
767
		goto done;
768
769
	}

770
	/* Check the semaphores that were modified. */
771
772
773
774
	for (i = 0; i < nsops; i++) {
		if (sops[i].sem_op > 0 ||
			(sops[i].sem_op < 0 &&
				sma->sem_base[sops[i].sem_num].semval == 0))
775
776
			if (update_queue(sma, sops[i].sem_num, pt))
				otime = 1;
777
	}
778
779
780
done:
	if (otime)
		sma->sem_otime = get_seconds();
781
782
783
}


Linus Torvalds's avatar
Linus Torvalds committed
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
/* The following counts are associated to each semaphore:
 *   semncnt        number of tasks waiting on semval being nonzero
 *   semzcnt        number of tasks waiting on semval being zero
 * This model assumes that a task waits on exactly one semaphore.
 * Since semaphore operations are to be performed atomically, tasks actually
 * wait on a whole sequence of semaphores simultaneously.
 * The counts we return here are a rough approximation, but still
 * warrant that semncnt+semzcnt>0 if the task is on the pending queue.
 */
static int count_semncnt (struct sem_array * sma, ushort semnum)
{
	int semncnt;
	struct sem_queue * q;

	semncnt = 0;
799
	list_for_each_entry(q, &sma->sem_pending, list) {
Linus Torvalds's avatar
Linus Torvalds committed
800
801
802
803
804
805
806
807
808
809
810
		struct sembuf * sops = q->sops;
		int nsops = q->nsops;
		int i;
		for (i = 0; i < nsops; i++)
			if (sops[i].sem_num == semnum
			    && (sops[i].sem_op < 0)
			    && !(sops[i].sem_flg & IPC_NOWAIT))
				semncnt++;
	}
	return semncnt;
}
811

Linus Torvalds's avatar
Linus Torvalds committed
812
813
814
815
816
817
static int count_semzcnt (struct sem_array * sma, ushort semnum)
{
	int semzcnt;
	struct sem_queue * q;

	semzcnt = 0;
818
	list_for_each_entry(q, &sma->sem_pending, list) {
Linus Torvalds's avatar
Linus Torvalds committed
819
820
821
822
823
824
825
826
827
828
829
830
		struct sembuf * sops = q->sops;
		int nsops = q->nsops;
		int i;
		for (i = 0; i < nsops; i++)
			if (sops[i].sem_num == semnum
			    && (sops[i].sem_op == 0)
			    && !(sops[i].sem_flg & IPC_NOWAIT))
				semzcnt++;
	}
	return semzcnt;
}

Nadia Derbey's avatar
Nadia Derbey committed
831
832
833
/* Free a semaphore set. freeary() is called with sem_ids.rw_mutex locked
 * as a writer and the spinlock for this semaphore set hold. sem_ids.rw_mutex
 * remains locked on exit.
Linus Torvalds's avatar
Linus Torvalds committed
834
 */
835
static void freeary(struct ipc_namespace *ns, struct kern_ipc_perm *ipcp)
Linus Torvalds's avatar
Linus Torvalds committed
836
{
837
838
	struct sem_undo *un, *tu;
	struct sem_queue *q, *tq;
839
	struct sem_array *sma = container_of(ipcp, struct sem_array, sem_perm);
840
	struct list_head tasks;
841
	int i;
Linus Torvalds's avatar
Linus Torvalds committed
842

843
	/* Free the existing undo structures for this semaphore set.  */
844
	assert_spin_locked(&sma->sem_perm.lock);
845
846
847
	list_for_each_entry_safe(un, tu, &sma->list_id, list_id) {
		list_del(&un->list_id);
		spin_lock(&un->ulp->lock);
Linus Torvalds's avatar
Linus Torvalds committed
848
		un->semid = -1;
849
850
		list_del_rcu(&un->list_proc);
		spin_unlock(&un->ulp->lock);
851
		kfree_rcu(un, rcu);
852
	}
Linus Torvalds's avatar
Linus Torvalds committed
853
854

	/* Wake up all pending processes and let them fail with EIDRM. */
855
	INIT_LIST_HEAD(&tasks);
856
	list_for_each_entry_safe(q, tq, &sma->sem_pending, list) {
857
		unlink_queue(sma, q);
858
		wake_up_sem_queue_prepare(&tasks, q, -EIDRM);
Linus Torvalds's avatar
Linus Torvalds committed
859
	}
860
861
862
863
864
865
866
	for (i = 0; i < sma->sem_nsems; i++) {
		struct sem *sem = sma->sem_base + i;
		list_for_each_entry_safe(q, tq, &sem->sem_pending, list) {
			unlink_queue(sma, q);
			wake_up_sem_queue_prepare(&tasks, q, -EIDRM);
		}
	}
Linus Torvalds's avatar
Linus Torvalds committed
867

Nadia Derbey's avatar
Nadia Derbey committed
868
869
	/* Remove the semaphore set from the IDR */
	sem_rmid(ns, sma);
870
	sem_unlock(sma, -1);
871
	rcu_read_unlock();
Linus Torvalds's avatar
Linus Torvalds committed
872

873
	wake_up_sem_queue_do(&tasks);
Kirill Korotaev's avatar
Kirill Korotaev committed
874
	ns->used_sems -= sma->sem_nsems;
Linus Torvalds's avatar
Linus Torvalds committed
875
876
877
878
879
880
881
882
883
884
885
886
887
	security_sem_free(sma);
	ipc_rcu_putref(sma);
}

static unsigned long copy_semid_to_user(void __user *buf, struct semid64_ds *in, int version)
{
	switch(version) {
	case IPC_64:
		return copy_to_user(buf, in, sizeof(*in));
	case IPC_OLD:
	    {
		struct semid_ds out;

888
889
		memset(&out, 0, sizeof(out));

Linus Torvalds's avatar
Linus Torvalds committed
890
891
892
893
894
895
896
897
898
899
900
901
902
		ipc64_perm_to_ipc_perm(&in->sem_perm, &out.sem_perm);

		out.sem_otime	= in->sem_otime;
		out.sem_ctime	= in->sem_ctime;
		out.sem_nsems	= in->sem_nsems;

		return copy_to_user(buf, &out, sizeof(out));
	    }
	default:
		return -EINVAL;
	}
}

903
static int semctl_nolock(struct ipc_namespace *ns, int semid,
904
			 int cmd, int version, void __user *p)
Linus Torvalds's avatar
Linus Torvalds committed
905
{
906
	int err;
Linus Torvalds's avatar
Linus Torvalds committed
907
908
909
910
911
912
913
914
915
916
917
918
919
920
	struct sem_array *sma;

	switch(cmd) {
	case IPC_INFO:
	case SEM_INFO:
	{
		struct seminfo seminfo;
		int max_id;

		err = security_sem_semctl(NULL, cmd);
		if (err)
			return err;
		
		memset(&seminfo,0,sizeof(seminfo));
Kirill Korotaev's avatar
Kirill Korotaev committed
921
922
923
924
		seminfo.semmni = ns->sc_semmni;
		seminfo.semmns = ns->sc_semmns;
		seminfo.semmsl = ns->sc_semmsl;
		seminfo.semopm = ns->sc_semopm;
Linus Torvalds's avatar
Linus Torvalds committed
925
926
927
928
		seminfo.semvmx = SEMVMX;
		seminfo.semmnu = SEMMNU;
		seminfo.semmap = SEMMAP;
		seminfo.semume = SEMUME;
Nadia Derbey's avatar
Nadia Derbey committed
929
		down_read(&sem_ids(ns).rw_mutex);
Linus Torvalds's avatar
Linus Torvalds committed
930
		if (cmd == SEM_INFO) {
Kirill Korotaev's avatar
Kirill Korotaev committed
931
932
			seminfo.semusz = sem_ids(ns).in_use;
			seminfo.semaem = ns->used_sems;
Linus Torvalds's avatar
Linus Torvalds committed
933
934
935
936
		} else {
			seminfo.semusz = SEMUSZ;
			seminfo.semaem = SEMAEM;
		}
Nadia Derbey's avatar
Nadia Derbey committed
937
		max_id = ipc_get_maxid(&sem_ids(ns));
Nadia Derbey's avatar
Nadia Derbey committed
938
		up_read(&sem_ids(ns).rw_mutex);
939
		if (copy_to_user(p, &seminfo, sizeof(struct seminfo))) 
Linus Torvalds's avatar
Linus Torvalds committed
940
941
942
			return -EFAULT;
		return (max_id < 0) ? 0: max_id;
	}
943
	case IPC_STAT:
Linus Torvalds's avatar
Linus Torvalds committed
944
945
946
	case SEM_STAT:
	{
		struct semid64_ds tbuf;
947
948
949
		int id = 0;

		memset(&tbuf, 0, sizeof(tbuf));
Linus Torvalds's avatar
Linus Torvalds committed
950

951
		if (cmd == SEM_STAT) {
952
953
954
955
956
957
			rcu_read_lock();
			sma = sem_obtain_object(ns, semid);
			if (IS_ERR(sma)) {
				err = PTR_ERR(sma);
				goto out_unlock;
			}
958
959
			id = sma->sem_perm.id;
		} else {
960
961
962
963
964
965
			rcu_read_lock();
			sma = sem_obtain_object_check(ns, semid);
			if (IS_ERR(sma)) {
				err = PTR_ERR(sma);
				goto out_unlock;
			}
966
		}
Linus Torvalds's avatar
Linus Torvalds committed
967
968

		err = -EACCES;
969
		if (ipcperms(ns, &sma->sem_perm, S_IRUGO))
Linus Torvalds's avatar
Linus Torvalds committed
970
971
972
973
974
975
976
977
978
979
			goto out_unlock;

		err = security_sem_semctl(sma, cmd);
		if (err)
			goto out_unlock;

		kernel_to_ipc64_perm(&sma->sem_perm, &tbuf.sem_perm);
		tbuf.sem_otime  = sma->sem_otime;
		tbuf.sem_ctime  = sma->sem_ctime;
		tbuf.sem_nsems  = sma->sem_nsems;
980
		rcu_read_unlock();
981
		if (copy_semid_to_user(p, &tbuf, version))
Linus Torvalds's avatar
Linus Torvalds committed
982
983
984
985
986
987
988
			return -EFAULT;
		return id;
	}
	default:
		return -EINVAL;
	}
out_unlock:
989
	rcu_read_unlock();
Linus Torvalds's avatar
Linus Torvalds committed
990
991
992
	return err;
}

993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
static int semctl_setval(struct ipc_namespace *ns, int semid, int semnum,
		unsigned long arg)
{
	struct sem_undo *un;
	struct sem_array *sma;
	struct sem* curr;
	int err;
	struct list_head tasks;
	int val;
#if defined(CONFIG_64BIT) && defined(__BIG_ENDIAN)
	/* big-endian 64bit */
	val = arg >> 32;
#else
	/* 32bit or little-endian 64bit */
	val = arg;
#endif

1010
1011
	if (val > SEMVMX || val < 0)
		return -ERANGE;
1012
1013
1014

	INIT_LIST_HEAD(&tasks);

1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
	rcu_read_lock();
	sma = sem_obtain_object_check(ns, semid);
	if (IS_ERR(sma)) {
		rcu_read_unlock();
		return PTR_ERR(sma);
	}

	if (semnum < 0 || semnum >= sma->sem_nsems) {
		rcu_read_unlock();
		return -EINVAL;
	}


	if (ipcperms(ns, &sma->sem_perm, S_IWUGO)) {
		rcu_read_unlock();
		return -EACCES;
	}
1032
1033

	err = security_sem_semctl(sma, SETVAL);
1034
1035
1036
1037
	if (err) {
		rcu_read_unlock();
		return -EACCES;
	}
1038

1039
	sem_lock(sma, NULL, -1);
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051

	curr = &sma->sem_base[semnum];

	assert_spin_locked(&sma->sem_perm.lock);
	list_for_each_entry(un, &sma->list_id, list_id)
		un->semadj[semnum] = 0;

	curr->semval = val;
	curr->sempid = task_tgid_vnr(current);
	sma->sem_ctime = get_seconds();
	/* maybe some queued-up processes were waiting for this */
	do_smart_update(sma, NULL, 0, 0, &tasks);
1052
	sem_unlock(sma, -1);
1053
	rcu_read_unlock();
1054
	wake_up_sem_queue_do(&tasks);
1055
	return 0;
1056
1057
}

Kirill Korotaev's avatar
Kirill Korotaev committed
1058
static int semctl_main(struct ipc_namespace *ns, int semid, int semnum,
1059
		int cmd, void __user *p)
Linus Torvalds's avatar
Linus Torvalds committed
1060
1061
1062
{
	struct sem_array *sma;
	struct sem* curr;
1063
	int err, nsems;
Linus Torvalds's avatar
Linus Torvalds committed
1064
1065
	ushort fast_sem_io[SEMMSL_FAST];
	ushort* sem_io = fast_sem_io;
1066
	struct list_head tasks;
Linus Torvalds's avatar
Linus Torvalds committed
1067

1068
1069
1070
1071
1072
1073
	INIT_LIST_HEAD(&tasks);

	rcu_read_lock();
	sma = sem_obtain_object_check(ns, semid);
	if (IS_ERR(sma)) {
		rcu_read_unlock();
1074
		return PTR_ERR(sma);
1075
	}
Linus Torvalds's avatar
Linus Torvalds committed
1076
1077
1078
1079

	nsems = sma->sem_nsems;

	err = -EACCES;
1080
1081
	if (ipcperms(ns, &sma->sem_perm, cmd == SETALL ? S_IWUGO : S_IRUGO))
		goto out_rcu_wakeup;
Linus Torvalds's avatar
Linus Torvalds committed
1082
1083

	err = security_sem_semctl(sma, cmd);
1084
1085
	if (err)
		goto out_rcu_wakeup;
Linus Torvalds's avatar
Linus Torvalds committed
1086
1087
1088
1089
1090

	err = -EACCES;
	switch (cmd) {
	case GETALL:
	{
1091
		ushort __user *array = p;
Linus Torvalds's avatar
Linus Torvalds committed
1092
1093
		int i;

1094
		sem_lock(sma, NULL, -1);
Linus Torvalds's avatar
Linus Torvalds committed
1095
		if(nsems > SEMMSL_FAST) {
1096
1097
			if (!ipc_rcu_getref(sma)) {
				sem_unlock(sma, -1);
1098
				rcu_read_unlock();
1099
1100
1101
1102
				err = -EIDRM;
				goto out_free;
			}
			sem_unlock(sma, -1);
1103
			rcu_read_unlock();
Linus Torvalds's avatar
Linus Torvalds committed
1104
1105
			sem_io = ipc_alloc(sizeof(ushort)*nsems);
			if(sem_io == NULL) {
1106
				sem_putref(sma);
Linus Torvalds's avatar
Linus Torvalds committed
1107
1108
1109
				return -ENOMEM;
			}

1110
			rcu_read_lock();
1111
			sem_lock_and_putref(sma);
Linus Torvalds's avatar
Linus Torvalds committed
1112
			if (sma->sem_perm.deleted) {
1113
				sem_unlock(sma, -1);
1114
				rcu_read_unlock();
Linus Torvalds's avatar
Linus Torvalds committed
1115
1116
1117
				err = -EIDRM;
				goto out_free;
			}
1118
		}
Linus Torvalds's avatar
Linus Torvalds committed
1119
1120
		for (i = 0; i < sma->sem_nsems; i++)
			sem_io[i] = sma->sem_base[i].semval;
1121
		sem_unlock(sma, -1);
1122
		rcu_read_unlock();
Linus Torvalds's avatar
Linus Torvalds committed
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
		err = 0;
		if(copy_to_user(array, sem_io, nsems*sizeof(ushort)))
			err = -EFAULT;
		goto out_free;
	}
	case SETALL:
	{
		int i;
		struct sem_undo *un;

1133
1134
1135
1136
		if (!ipc_rcu_getref(sma)) {
			rcu_read_unlock();
			return -EIDRM;
		}
1137
		rcu_read_unlock();
Linus Torvalds's avatar
Linus Torvalds committed
1138
1139
1140
1141

		if(nsems > SEMMSL_FAST) {
			sem_io = ipc_alloc(sizeof(ushort)*nsems);
			if(sem_io == NULL) {
1142
				sem_putref(sma);
Linus Torvalds's avatar
Linus Torvalds committed
1143
1144
1145
1146
				return -ENOMEM;
			}
		}

1147
		if (copy_from_user (sem_io, p, nsems*sizeof(ushort))) {
1148
			sem_putref(sma);
Linus Torvalds's avatar
Linus Torvalds committed
1149
1150
1151
1152
1153
1154
			err = -EFAULT;
			goto out_free;
		}

		for (i = 0; i < nsems; i++) {
			if (sem_io[i] > SEMVMX) {
1155
				sem_putref(sma);
Linus Torvalds's avatar
Linus Torvalds committed
1156
1157
1158
1159
				err = -ERANGE;
				goto out_free;
			}
		}
1160
		rcu_read_lock();
1161
		sem_lock_and_putref(sma);
Linus Torvalds's avatar
Linus Torvalds committed
1162
		if (sma->sem_perm.deleted) {
1163
			sem_unlock(sma, -1);
1164
			rcu_read_unlock();
Linus Torvalds's avatar
Linus Torvalds committed
1165
1166
1167
1168
1169
1170
			err = -EIDRM;
			goto out_free;
		}

		for (i = 0; i < nsems; i++)
			sma->sem_base[i].semval = sem_io[i];
1171
1172
1173

		assert_spin_locked(&sma->sem_perm.lock);
		list_for_each_entry(un, &sma->list_id, list_id) {
Linus Torvalds's avatar
Linus Torvalds committed
1174
1175
			for (i = 0; i < nsems; i++)
				un->semadj[i] = 0;
1176
		}
Linus Torvalds's avatar
Linus Torvalds committed
1177
1178
		sma->sem_ctime = get_seconds();
		/* maybe some queued-up processes were waiting for this */
1179
		do_smart_update(sma, NULL, 0, 0, &tasks);
Linus Torvalds's avatar
Linus Torvalds committed
1180
1181
1182
		err = 0;
		goto out_unlock;
	}
1183
	/* GETVAL, GETPID, GETNCTN, GETZCNT: fall-through */
Linus Torvalds's avatar
Linus Torvalds committed
1184
1185
	}
	err = -EINVAL;
1186
1187
	if (semnum < 0 || semnum >= nsems)
		goto out_rcu_wakeup;
Linus Torvalds's avatar
Linus Torvalds committed
1188

1189
	sem_lock(sma, NULL, -1);
Linus Torvalds's avatar
Linus Torvalds committed
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
	curr = &sma->sem_base[semnum];

	switch (cmd) {
	case GETVAL:
		err = curr->semval;
		goto out_unlock;
	case GETPID:
		err = curr->sempid;
		goto out_unlock;
	case GETNCNT:
		err = count_semncnt(sma,semnum);
		goto out_unlock;
	case GETZCNT:
		err = count_semzcnt(sma,semnum);
		goto out_unlock;
	}
1206

Linus Torvalds's avatar
Linus Torvalds committed
1207
out_unlock:
1208
	sem_unlock(sma, -1);
1209
out_rcu_wakeup:
1210
	rcu_read_unlock();
1211
	wake_up_sem_queue_do(&tasks);
Linus Torvalds's avatar
Linus Torvalds committed
1212
1213
1214
1215
1216
1217
out_free:
	if(sem_io != fast_sem_io)
		ipc_free(sem_io, sizeof(ushort)*nsems);
	return err;
}

1218
1219
static inline unsigned long
copy_semid_from_user(struct semid64_ds *out, void __user *buf, int version)
Linus Torvalds's avatar
Linus Torvalds committed
1220
1221
1222
{
	switch(version) {
	case IPC_64:
1223
		if (copy_from_user(out, buf, sizeof(*out)))
Linus Torvalds's avatar
Linus Torvalds committed
1224
1225
1226
1227
1228
1229
1230
1231
1232
			return -EFAULT;
		return 0;
	case IPC_OLD:
	    {
		struct semid_ds tbuf_old;

		if(copy_from_user(&tbuf_old, buf, sizeof(tbuf_old)))
			return -EFAULT;

1233
1234
1235
		out->sem_perm.uid	= tbuf_old.sem_perm.uid;
		out->sem_perm.gid	= tbuf_old.sem_perm.gid;
		out->sem_perm.mode	= tbuf_old.sem_perm.mode;
Linus Torvalds's avatar
Linus Torvalds committed
1236
1237
1238
1239
1240
1241
1242
1243

		return 0;
	    }
	default:
		return -EINVAL;
	}
}

1244
1245
1246
1247
1248
/*
 * This function handles some semctl commands which require the rw_mutex
 * to be held in write mode.
 * NOTE: no locks must be held, the rw_mutex is taken inside this function.
 */
1249
static int semctl_down(struct ipc_namespace *ns, int semid,
1250
		       int cmd, int version, void __user *p)
Linus Torvalds's avatar
Linus Torvalds committed
1251
1252
1253
{
	struct sem_array *sma;
	int err;
1254
	struct semid64_ds semid64;
Linus Torvalds's avatar
Linus Torvalds committed
1255
1256
1257
	struct kern_ipc_perm *ipcp;

	if(cmd == IPC_SET) {
1258
		if (copy_semid_from_user(&semid64, p, version))
Linus Torvalds's avatar
Linus Torvalds committed
1259
1260
			return -EFAULT;
	}
Steve Grubb's avatar
Steve Grubb committed
1261

1262
1263
	ipcp = ipcctl_pre_down_nolock(ns, &sem_ids(ns), semid, cmd,
				      &semid64.sem_perm, 0);
1264
1265
	if (IS_ERR(ipcp))
		return PTR_ERR(ipcp);
Steve Grubb's avatar
Steve Grubb committed
1266

1267
	sma = container_of(ipcp, struct sem_array, sem_perm);
Linus Torvalds's avatar
Linus Torvalds committed
1268
1269

	err = security_sem_semctl(sma, cmd);
1270
1271
	if (err) {
		rcu_read_unlock();
1272
		goto out_up;
1273
	}
Linus Torvalds's avatar
Linus Torvalds committed
1274
1275
1276

	switch(cmd){
	case IPC_RMID:
1277
		sem_lock(sma, NULL, -1);
1278
		freeary(ns, ipcp);
1279
		goto out_up;
Linus Torvalds's avatar
Linus Torvalds committed
1280
	case IPC_SET:
1281
		sem_lock(sma, NULL, -1);
1282
1283
1284
		err = ipc_update_perm(&semid64.sem_perm, ipcp);
		if (err)
			goto out_unlock;
Linus Torvalds's avatar
Linus Torvalds committed
1285
1286
1287
		sma->sem_ctime = get_seconds();
		break;
	default:
1288
		rcu_read_unlock();
Linus Torvalds's avatar
Linus Torvalds committed
1289
		err = -EINVAL;
1290
		goto out_up;
Linus Torvalds's avatar
Linus Torvalds committed
1291
1292
1293
	}

out_unlock:
1294
	sem_unlock(sma, -1);
1295
	rcu_read_unlock();
1296
1297
out_up:
	up_write(&sem_ids(ns).rw_mutex);
Linus Torvalds's avatar
Linus Torvalds committed
1298
1299
1300
	return err;
}

1301
SYSCALL_DEFINE4(semctl, int, semid, int, semnum, int, cmd, unsigned long, arg)
Linus Torvalds's avatar
Linus Torvalds committed
1302
1303
{
	int version;
Kirill Korotaev's avatar
Kirill Korotaev committed
1304
	struct ipc_namespace *ns;
1305
	void __user *p = (void __user *)arg;
Linus Torvalds's avatar
Linus Torvalds committed
1306
1307
1308
1309
1310

	if (semid < 0)
		return -EINVAL;

	version = ipc_parse_version(&cmd);
Kirill Korotaev's avatar
Kirill Korotaev committed
1311
	ns = current->nsproxy->ipc_ns;