sem.c 38.0 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58
/*
 * linux/ipc/sem.c
 * Copyright (C) 1992 Krishna Balasubramanian
 * Copyright (C) 1995 Eric Schenk, Bruno Haible
 *
 * IMPLEMENTATION NOTES ON CODE REWRITE (Eric Schenk, January 1995):
 * This code underwent a massive rewrite in order to solve some problems
 * with the original code. In particular the original code failed to
 * wake up processes that were waiting for semval to go to 0 if the
 * value went to 0 and was then incremented rapidly enough. In solving
 * this problem I have also modified the implementation so that it
 * processes pending operations in a FIFO manner, thus give a guarantee
 * that processes waiting for a lock on the semaphore won't starve
 * unless another locking process fails to unlock.
 * In addition the following two changes in behavior have been introduced:
 * - The original implementation of semop returned the value
 *   last semaphore element examined on success. This does not
 *   match the manual page specifications, and effectively
 *   allows the user to read the semaphore even if they do not
 *   have read permissions. The implementation now returns 0
 *   on success as stated in the manual page.
 * - There is some confusion over whether the set of undo adjustments
 *   to be performed at exit should be done in an atomic manner.
 *   That is, if we are attempting to decrement the semval should we queue
 *   up and wait until we can do so legally?
 *   The original implementation attempted to do this.
 *   The current implementation does not do so. This is because I don't
 *   think it is the right thing (TM) to do, and because I couldn't
 *   see a clean way to get the old behavior with the new design.
 *   The POSIX standard and SVID should be consulted to determine
 *   what behavior is mandated.
 *
 * Further notes on refinement (Christoph Rohland, December 1998):
 * - The POSIX standard says, that the undo adjustments simply should
 *   redo. So the current implementation is o.K.
 * - The previous code had two flaws:
 *   1) It actively gave the semaphore to the next waiting process
 *      sleeping on the semaphore. Since this process did not have the
 *      cpu this led to many unnecessary context switches and bad
 *      performance. Now we only check which process should be able to
 *      get the semaphore and if this process wants to reduce some
 *      semaphore value we simply wake it up without doing the
 *      operation. So it has to try to get it later. Thus e.g. the
 *      running process may reacquire the semaphore during the current
 *      time slice. If it only waits for zero or increases the semaphore,
 *      we do the operation in advance and wake it up.
 *   2) It did not wake up all zero waiting processes. We try to do
 *      better but only get the semops right which only wait for zero or
 *      increase. If there are decrement operations in the operations
 *      array we do the same as before.
 *
 * With the incarnation of O(1) scheduler, it becomes unnecessary to perform
 * check/retry algorithm for waking up blocked processes as the new scheduler
 * is better at handling thread switch than the old one.
 *
 * /proc/sysvipc/sem support (c) 1999 Dragos Acostachioaie <dragos@iname.com>
 *
 * SMP-threaded, sysctl's added
59
 * (c) 1999 Manfred Spraul <manfred@colorfullife.com>
L
Linus Torvalds 已提交
60
 * Enforced range limit on SEM_UNDO
A
Alan Cox 已提交
61
 * (c) 2001 Red Hat Inc
L
Linus Torvalds 已提交
62 63
 * Lockless wakeup
 * (c) 2003 Manfred Spraul <manfred@colorfullife.com>
S
Steve Grubb 已提交
64 65 66
 *
 * support for audit of ipc object properties and permission changes
 * Dustin Kirkland <dustin.kirkland@us.ibm.com>
K
Kirill Korotaev 已提交
67 68 69 70
 *
 * namespaces support
 * OpenVZ, SWsoft Inc.
 * Pavel Emelianov <xemul@openvz.org>
L
Linus Torvalds 已提交
71 72 73 74 75 76 77 78 79 80
 */

#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>
81
#include <linux/capability.h>
82
#include <linux/seq_file.h>
N
Nadia Derbey 已提交
83
#include <linux/rwsem.h>
K
Kirill Korotaev 已提交
84
#include <linux/nsproxy.h>
85
#include <linux/ipc_namespace.h>
I
Ingo Molnar 已提交
86

L
Linus Torvalds 已提交
87 88 89
#include <asm/uaccess.h>
#include "util.h"

90
#define sem_ids(ns)	((ns)->ids[IPC_SEM_IDS])
K
Kirill Korotaev 已提交
91 92

#define sem_unlock(sma)		ipc_unlock(&(sma)->sem_perm)
N
Nadia Derbey 已提交
93
#define sem_checkid(sma, semid)	ipc_checkid(&sma->sem_perm, semid)
L
Linus Torvalds 已提交
94

N
Nadia Derbey 已提交
95
static int newary(struct ipc_namespace *, struct ipc_params *);
96
static void freeary(struct ipc_namespace *, struct kern_ipc_perm *);
L
Linus Torvalds 已提交
97
#ifdef CONFIG_PROC_FS
98
static int sysvipc_sem_proc_show(struct seq_file *s, void *it);
L
Linus Torvalds 已提交
99 100 101 102 103 104 105 106 107 108 109 110 111 112
#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.
 *	
 */

K
Kirill Korotaev 已提交
113 114 115 116 117
#define sc_semmsl	sem_ctls[0]
#define sc_semmns	sem_ctls[1]
#define sc_semopm	sem_ctls[2]
#define sc_semmni	sem_ctls[3]

118
void sem_init_ns(struct ipc_namespace *ns)
K
Kirill Korotaev 已提交
119 120 121 122 123 124
{
	ns->sc_semmsl = SEMMSL;
	ns->sc_semmns = SEMMNS;
	ns->sc_semopm = SEMOPM;
	ns->sc_semmni = SEMMNI;
	ns->used_sems = 0;
125
	ipc_init_ids(&ns->ids[IPC_SEM_IDS]);
K
Kirill Korotaev 已提交
126 127
}

128
#ifdef CONFIG_IPC_NS
K
Kirill Korotaev 已提交
129 130
void sem_exit_ns(struct ipc_namespace *ns)
{
131
	free_ipcs(ns, &sem_ids(ns), freeary);
S
Serge E. Hallyn 已提交
132
	idr_destroy(&ns->ids[IPC_SEM_IDS].ipcs_idr);
K
Kirill Korotaev 已提交
133
}
134
#endif
L
Linus Torvalds 已提交
135 136 137

void __init sem_init (void)
{
138
	sem_init_ns(&init_ipc_ns);
139 140
	ipc_init_proc_interface("sysvipc/sem",
				"       key      semid perms      nsems   uid   gid  cuid  cgid      otime      ctime\n",
K
Kirill Korotaev 已提交
141
				IPC_SEM_IDS, sysvipc_sem_proc_show);
L
Linus Torvalds 已提交
142 143
}

N
Nadia Derbey 已提交
144 145 146 147
/*
 * sem_lock_(check_) routines are called in the paths where the rw_mutex
 * is not held.
 */
148 149
static inline struct sem_array *sem_lock(struct ipc_namespace *ns, int id)
{
N
Nadia Derbey 已提交
150 151
	struct kern_ipc_perm *ipcp = ipc_lock(&sem_ids(ns), id);

152 153 154
	if (IS_ERR(ipcp))
		return (struct sem_array *)ipcp;

N
Nadia Derbey 已提交
155
	return container_of(ipcp, struct sem_array, sem_perm);
156 157 158 159 160
}

static inline struct sem_array *sem_lock_check(struct ipc_namespace *ns,
						int id)
{
N
Nadia Derbey 已提交
161 162
	struct kern_ipc_perm *ipcp = ipc_lock_check(&sem_ids(ns), id);

163 164 165
	if (IS_ERR(ipcp))
		return (struct sem_array *)ipcp;

N
Nadia Derbey 已提交
166
	return container_of(ipcp, struct sem_array, sem_perm);
167 168
}

169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187
static inline void sem_lock_and_putref(struct sem_array *sma)
{
	ipc_lock_by_ptr(&sma->sem_perm);
	ipc_rcu_putref(sma);
}

static inline void sem_getref_and_unlock(struct sem_array *sma)
{
	ipc_rcu_getref(sma);
	ipc_unlock(&(sma)->sem_perm);
}

static inline void sem_putref(struct sem_array *sma)
{
	ipc_lock_by_ptr(&sma->sem_perm);
	ipc_rcu_putref(sma);
	ipc_unlock(&(sma)->sem_perm);
}

N
Nadia Derbey 已提交
188 189 190 191 192
static inline void sem_rmid(struct ipc_namespace *ns, struct sem_array *s)
{
	ipc_rmid(&sem_ids(ns), &s->sem_perm);
}

L
Linus Torvalds 已提交
193 194 195 196 197 198 199 200 201 202 203 204 205 206 207
/*
 * 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
I
Ingo Molnar 已提交
208
 *   	  performing any operation on the sem array.
L
Linus Torvalds 已提交
209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226
 *   	* 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

N
Nadia Derbey 已提交
227 228 229 230 231
/**
 * newary - Create a new semaphore set
 * @ns: namespace
 * @params: ptr to the structure that contains key, semflg and nsems
 *
N
Nadia Derbey 已提交
232
 * Called with sem_ids.rw_mutex held (as a writer)
N
Nadia Derbey 已提交
233 234
 */

N
Nadia Derbey 已提交
235
static int newary(struct ipc_namespace *ns, struct ipc_params *params)
L
Linus Torvalds 已提交
236 237 238 239 240
{
	int id;
	int retval;
	struct sem_array *sma;
	int size;
N
Nadia Derbey 已提交
241 242 243
	key_t key = params->key;
	int nsems = params->u.nsems;
	int semflg = params->flg;
244
	int i;
L
Linus Torvalds 已提交
245 246 247

	if (!nsems)
		return -EINVAL;
K
Kirill Korotaev 已提交
248
	if (ns->used_sems + nsems > ns->sc_semmns)
L
Linus Torvalds 已提交
249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267
		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;
	}

K
Kirill Korotaev 已提交
268
	id = ipc_addid(&sem_ids(ns), &sma->sem_perm, ns->sc_semmni);
269
	if (id < 0) {
L
Linus Torvalds 已提交
270 271
		security_sem_free(sma);
		ipc_rcu_putref(sma);
272
		return id;
L
Linus Torvalds 已提交
273
	}
K
Kirill Korotaev 已提交
274
	ns->used_sems += nsems;
L
Linus Torvalds 已提交
275 276

	sma->sem_base = (struct sem *) &sma[1];
277 278 279 280 281

	for (i = 0; i < nsems; i++)
		INIT_LIST_HEAD(&sma->sem_base[i].sem_pending);

	sma->complex_count = 0;
282
	INIT_LIST_HEAD(&sma->sem_pending);
283
	INIT_LIST_HEAD(&sma->list_id);
L
Linus Torvalds 已提交
284 285 286 287
	sma->sem_nsems = nsems;
	sma->sem_ctime = get_seconds();
	sem_unlock(sma);

N
Nadia Derbey 已提交
288
	return sma->sem_perm.id;
L
Linus Torvalds 已提交
289 290
}

N
Nadia Derbey 已提交
291

N
Nadia Derbey 已提交
292
/*
N
Nadia Derbey 已提交
293
 * Called with sem_ids.rw_mutex and ipcp locked.
N
Nadia Derbey 已提交
294
 */
N
Nadia Derbey 已提交
295
static inline int sem_security(struct kern_ipc_perm *ipcp, int semflg)
N
Nadia Derbey 已提交
296
{
N
Nadia Derbey 已提交
297 298 299 300
	struct sem_array *sma;

	sma = container_of(ipcp, struct sem_array, sem_perm);
	return security_sem_associate(sma, semflg);
N
Nadia Derbey 已提交
301 302
}

N
Nadia Derbey 已提交
303
/*
N
Nadia Derbey 已提交
304
 * Called with sem_ids.rw_mutex and ipcp locked.
N
Nadia Derbey 已提交
305
 */
N
Nadia Derbey 已提交
306 307
static inline int sem_more_checks(struct kern_ipc_perm *ipcp,
				struct ipc_params *params)
N
Nadia Derbey 已提交
308
{
N
Nadia Derbey 已提交
309 310 311 312
	struct sem_array *sma;

	sma = container_of(ipcp, struct sem_array, sem_perm);
	if (params->u.nsems > sma->sem_nsems)
N
Nadia Derbey 已提交
313 314 315 316 317
		return -EINVAL;

	return 0;
}

318
SYSCALL_DEFINE3(semget, key_t, key, int, nsems, int, semflg)
L
Linus Torvalds 已提交
319
{
K
Kirill Korotaev 已提交
320
	struct ipc_namespace *ns;
N
Nadia Derbey 已提交
321 322
	struct ipc_ops sem_ops;
	struct ipc_params sem_params;
K
Kirill Korotaev 已提交
323 324

	ns = current->nsproxy->ipc_ns;
L
Linus Torvalds 已提交
325

K
Kirill Korotaev 已提交
326
	if (nsems < 0 || nsems > ns->sc_semmsl)
L
Linus Torvalds 已提交
327
		return -EINVAL;
N
Nadia Derbey 已提交
328

N
Nadia Derbey 已提交
329 330 331 332 333 334 335
	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;
L
Linus Torvalds 已提交
336

N
Nadia Derbey 已提交
337
	return ipcget(ns, &sem_ids(ns), &sem_ops, &sem_params);
L
Linus Torvalds 已提交
338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406
}

/*
 * 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--;
	}
	
	sma->sem_otime = get_seconds();
	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;
}

N
Nick Piggin 已提交
407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427
/*
 * Wake up a process waiting on the sem queue with a given error.
 * The queue is invalid (may not be accessed) after the function returns.
 */
static void wake_up_sem_queue(struct sem_queue *q, int error)
{
	/*
	 * Hold preempt off so that we don't get preempted and have the
	 * wakee busy-wait until we're scheduled back on. We're holding
	 * locks here so it may not strictly be needed, however if the
	 * locks become preemptible then this prevents such a problem.
	 */
	preempt_disable();
	q->status = IN_WAKEUP;
	wake_up_process(q->sleeper);
	/* hands-off: q can disappear immediately after writing q->status. */
	smp_wmb();
	q->status = error;
	preempt_enable();
}

428 429 430 431 432 433 434 435 436
static void unlink_queue(struct sem_array *sma, struct sem_queue *q)
{
	list_del(&q->list);
	if (q->nsops == 1)
		list_del(&q->simple_list);
	else
		sma->complex_count--;
}

437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499
/** 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.
	 * They must be the first entries in the per-semaphore simple queue
	 */
	h = list_first_entry(&curr->sem_pending, struct sem_queue, simple_list);
	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;
}

500 501 502 503 504 505 506 507 508

/**
 * update_queue(sma, semnum): Look for tasks that can be completed.
 * @sma: semaphore array.
 * @semnum: semaphore that was modified.
 *
 * update_queue must be called after a semaphore in a semaphore array
 * was modified. If multiple semaphore were modified, then @semnum
 * must be set to -1.
L
Linus Torvalds 已提交
509
 */
510
static void update_queue(struct sem_array *sma, int semnum)
L
Linus Torvalds 已提交
511
{
512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530
	struct sem_queue *q;
	struct list_head *walk;
	struct list_head *pending_list;
	int offset;

	/* if there are complex operations around, then knowing the semaphore
	 * that was modified doesn't help us. Assume that multiple semaphores
	 * were modified.
	 */
	if (sma->complex_count)
		semnum = -1;

	if (semnum == -1) {
		pending_list = &sma->sem_pending;
		offset = offsetof(struct sem_queue, list);
	} else {
		pending_list = &sma->sem_base[semnum].sem_pending;
		offset = offsetof(struct sem_queue, simple_list);
	}
N
Nick Piggin 已提交
531 532

again:
533 534
	walk = pending_list->next;
	while (walk != pending_list) {
535
		int error, restart;
536 537 538

		q = (struct sem_queue *)((char *)walk - offset);
		walk = walk->next;
L
Linus Torvalds 已提交
539

540 541 542 543 544 545 546 547 548 549 550
		/* 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;

L
Linus Torvalds 已提交
551 552 553 554
		error = try_atomic_semop(sma, q->sops, q->nsops,
					 q->undo, q->pid);

		/* Does q->sleeper still need to sleep? */
N
Nick Piggin 已提交
555 556 557
		if (error > 0)
			continue;

558
		unlink_queue(sma, q);
N
Nick Piggin 已提交
559

560 561 562 563 564
		if (error)
			restart = 0;
		else
			restart = check_restart(sma, q);

N
Nick Piggin 已提交
565
		wake_up_sem_queue(q, error);
566
		if (restart)
N
Nick Piggin 已提交
567
			goto again;
L
Linus Torvalds 已提交
568 569 570
	}
}

571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596
/** do_smart_update(sma, sops, nsops): Optimized update_queue
 * @sma: semaphore array
 * @sops: operations that were performed
 * @nsops: number of operations
 *
 * do_smart_update() does the required called to update_queue, based on the
 * actual changes that were performed on the semaphore array.
 */
static void do_smart_update(struct sem_array *sma, struct sembuf *sops, int nsops)
{
	int i;

	if (sma->complex_count || sops == NULL) {
		update_queue(sma, -1);
		return;
	}

	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))
			update_queue(sma, sops[i].sem_num);
	}
}


L
Linus Torvalds 已提交
597 598 599 600 601 602 603 604 605 606 607 608 609 610 611
/* 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;
612
	list_for_each_entry(q, &sma->sem_pending, list) {
L
Linus Torvalds 已提交
613 614 615 616 617 618 619 620 621 622 623
		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;
}
624

L
Linus Torvalds 已提交
625 626 627 628 629 630
static int count_semzcnt (struct sem_array * sma, ushort semnum)
{
	int semzcnt;
	struct sem_queue * q;

	semzcnt = 0;
631
	list_for_each_entry(q, &sma->sem_pending, list) {
L
Linus Torvalds 已提交
632 633 634 635 636 637 638 639 640 641 642 643
		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;
}

A
Adrian Bunk 已提交
644
static void free_un(struct rcu_head *head)
645 646 647 648 649
{
	struct sem_undo *un = container_of(head, struct sem_undo, rcu);
	kfree(un);
}

N
Nadia Derbey 已提交
650 651 652
/* 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.
L
Linus Torvalds 已提交
653
 */
654
static void freeary(struct ipc_namespace *ns, struct kern_ipc_perm *ipcp)
L
Linus Torvalds 已提交
655
{
656 657
	struct sem_undo *un, *tu;
	struct sem_queue *q, *tq;
658
	struct sem_array *sma = container_of(ipcp, struct sem_array, sem_perm);
L
Linus Torvalds 已提交
659

660
	/* Free the existing undo structures for this semaphore set.  */
661
	assert_spin_locked(&sma->sem_perm.lock);
662 663 664
	list_for_each_entry_safe(un, tu, &sma->list_id, list_id) {
		list_del(&un->list_id);
		spin_lock(&un->ulp->lock);
L
Linus Torvalds 已提交
665
		un->semid = -1;
666 667 668 669
		list_del_rcu(&un->list_proc);
		spin_unlock(&un->ulp->lock);
		call_rcu(&un->rcu, free_un);
	}
L
Linus Torvalds 已提交
670 671

	/* Wake up all pending processes and let them fail with EIDRM. */
672
	list_for_each_entry_safe(q, tq, &sma->sem_pending, list) {
673
		unlink_queue(sma, q);
N
Nick Piggin 已提交
674
		wake_up_sem_queue(q, -EIDRM);
L
Linus Torvalds 已提交
675 676
	}

N
Nadia Derbey 已提交
677 678
	/* Remove the semaphore set from the IDR */
	sem_rmid(ns, sma);
L
Linus Torvalds 已提交
679 680
	sem_unlock(sma);

K
Kirill Korotaev 已提交
681
	ns->used_sems -= sma->sem_nsems;
L
Linus Torvalds 已提交
682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707
	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;

		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;
	}
}

708 709
static int semctl_nolock(struct ipc_namespace *ns, int semid,
			 int cmd, int version, union semun arg)
L
Linus Torvalds 已提交
710
{
711
	int err;
L
Linus Torvalds 已提交
712 713 714 715 716 717 718 719 720 721 722 723 724 725
	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));
K
Kirill Korotaev 已提交
726 727 728 729
		seminfo.semmni = ns->sc_semmni;
		seminfo.semmns = ns->sc_semmns;
		seminfo.semmsl = ns->sc_semmsl;
		seminfo.semopm = ns->sc_semopm;
L
Linus Torvalds 已提交
730 731 732 733
		seminfo.semvmx = SEMVMX;
		seminfo.semmnu = SEMMNU;
		seminfo.semmap = SEMMAP;
		seminfo.semume = SEMUME;
N
Nadia Derbey 已提交
734
		down_read(&sem_ids(ns).rw_mutex);
L
Linus Torvalds 已提交
735
		if (cmd == SEM_INFO) {
K
Kirill Korotaev 已提交
736 737
			seminfo.semusz = sem_ids(ns).in_use;
			seminfo.semaem = ns->used_sems;
L
Linus Torvalds 已提交
738 739 740 741
		} else {
			seminfo.semusz = SEMUSZ;
			seminfo.semaem = SEMAEM;
		}
N
Nadia Derbey 已提交
742
		max_id = ipc_get_maxid(&sem_ids(ns));
N
Nadia Derbey 已提交
743
		up_read(&sem_ids(ns).rw_mutex);
L
Linus Torvalds 已提交
744 745 746 747
		if (copy_to_user (arg.__buf, &seminfo, sizeof(struct seminfo))) 
			return -EFAULT;
		return (max_id < 0) ? 0: max_id;
	}
748
	case IPC_STAT:
L
Linus Torvalds 已提交
749 750 751 752 753
	case SEM_STAT:
	{
		struct semid64_ds tbuf;
		int id;

754 755 756 757 758 759 760 761 762 763 764
		if (cmd == SEM_STAT) {
			sma = sem_lock(ns, semid);
			if (IS_ERR(sma))
				return PTR_ERR(sma);
			id = sma->sem_perm.id;
		} else {
			sma = sem_lock_check(ns, semid);
			if (IS_ERR(sma))
				return PTR_ERR(sma);
			id = 0;
		}
L
Linus Torvalds 已提交
765 766 767 768 769 770 771 772 773

		err = -EACCES;
		if (ipcperms (&sma->sem_perm, S_IRUGO))
			goto out_unlock;

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

774 775
		memset(&tbuf, 0, sizeof(tbuf));

L
Linus Torvalds 已提交
776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792
		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;
		sem_unlock(sma);
		if (copy_semid_to_user (arg.buf, &tbuf, version))
			return -EFAULT;
		return id;
	}
	default:
		return -EINVAL;
	}
out_unlock:
	sem_unlock(sma);
	return err;
}

K
Kirill Korotaev 已提交
793 794
static int semctl_main(struct ipc_namespace *ns, int semid, int semnum,
		int cmd, int version, union semun arg)
L
Linus Torvalds 已提交
795 796 797 798 799 800 801 802
{
	struct sem_array *sma;
	struct sem* curr;
	int err;
	ushort fast_sem_io[SEMMSL_FAST];
	ushort* sem_io = fast_sem_io;
	int nsems;

803 804 805
	sma = sem_lock_check(ns, semid);
	if (IS_ERR(sma))
		return PTR_ERR(sma);
L
Linus Torvalds 已提交
806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824

	nsems = sma->sem_nsems;

	err = -EACCES;
	if (ipcperms (&sma->sem_perm, (cmd==SETVAL||cmd==SETALL)?S_IWUGO:S_IRUGO))
		goto out_unlock;

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

	err = -EACCES;
	switch (cmd) {
	case GETALL:
	{
		ushort __user *array = arg.array;
		int i;

		if(nsems > SEMMSL_FAST) {
825
			sem_getref_and_unlock(sma);
L
Linus Torvalds 已提交
826 827 828

			sem_io = ipc_alloc(sizeof(ushort)*nsems);
			if(sem_io == NULL) {
829
				sem_putref(sma);
L
Linus Torvalds 已提交
830 831 832
				return -ENOMEM;
			}

833
			sem_lock_and_putref(sma);
L
Linus Torvalds 已提交
834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853
			if (sma->sem_perm.deleted) {
				sem_unlock(sma);
				err = -EIDRM;
				goto out_free;
			}
		}

		for (i = 0; i < sma->sem_nsems; i++)
			sem_io[i] = sma->sem_base[i].semval;
		sem_unlock(sma);
		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;

854
		sem_getref_and_unlock(sma);
L
Linus Torvalds 已提交
855 856 857 858

		if(nsems > SEMMSL_FAST) {
			sem_io = ipc_alloc(sizeof(ushort)*nsems);
			if(sem_io == NULL) {
859
				sem_putref(sma);
L
Linus Torvalds 已提交
860 861 862 863 864
				return -ENOMEM;
			}
		}

		if (copy_from_user (sem_io, arg.array, nsems*sizeof(ushort))) {
865
			sem_putref(sma);
L
Linus Torvalds 已提交
866 867 868 869 870 871
			err = -EFAULT;
			goto out_free;
		}

		for (i = 0; i < nsems; i++) {
			if (sem_io[i] > SEMVMX) {
872
				sem_putref(sma);
L
Linus Torvalds 已提交
873 874 875 876
				err = -ERANGE;
				goto out_free;
			}
		}
877
		sem_lock_and_putref(sma);
L
Linus Torvalds 已提交
878 879 880 881 882 883 884 885
		if (sma->sem_perm.deleted) {
			sem_unlock(sma);
			err = -EIDRM;
			goto out_free;
		}

		for (i = 0; i < nsems; i++)
			sma->sem_base[i].semval = sem_io[i];
886 887 888

		assert_spin_locked(&sma->sem_perm.lock);
		list_for_each_entry(un, &sma->list_id, list_id) {
L
Linus Torvalds 已提交
889 890
			for (i = 0; i < nsems; i++)
				un->semadj[i] = 0;
891
		}
L
Linus Torvalds 已提交
892 893
		sma->sem_ctime = get_seconds();
		/* maybe some queued-up processes were waiting for this */
894
		update_queue(sma, -1);
L
Linus Torvalds 已提交
895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922
		err = 0;
		goto out_unlock;
	}
	/* GETVAL, GETPID, GETNCTN, GETZCNT, SETVAL: fall-through */
	}
	err = -EINVAL;
	if(semnum < 0 || semnum >= nsems)
		goto out_unlock;

	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;
	case SETVAL:
	{
		int val = arg.val;
		struct sem_undo *un;
923

L
Linus Torvalds 已提交
924 925 926 927
		err = -ERANGE;
		if (val > SEMVMX || val < 0)
			goto out_unlock;

928 929
		assert_spin_locked(&sma->sem_perm.lock);
		list_for_each_entry(un, &sma->list_id, list_id)
L
Linus Torvalds 已提交
930
			un->semadj[semnum] = 0;
931

L
Linus Torvalds 已提交
932
		curr->semval = val;
933
		curr->sempid = task_tgid_vnr(current);
L
Linus Torvalds 已提交
934 935
		sma->sem_ctime = get_seconds();
		/* maybe some queued-up processes were waiting for this */
936
		update_queue(sma, semnum);
L
Linus Torvalds 已提交
937 938 939 940 941 942 943 944 945 946 947 948
		err = 0;
		goto out_unlock;
	}
	}
out_unlock:
	sem_unlock(sma);
out_free:
	if(sem_io != fast_sem_io)
		ipc_free(sem_io, sizeof(ushort)*nsems);
	return err;
}

949 950
static inline unsigned long
copy_semid_from_user(struct semid64_ds *out, void __user *buf, int version)
L
Linus Torvalds 已提交
951 952 953
{
	switch(version) {
	case IPC_64:
954
		if (copy_from_user(out, buf, sizeof(*out)))
L
Linus Torvalds 已提交
955 956 957 958 959 960 961 962 963
			return -EFAULT;
		return 0;
	case IPC_OLD:
	    {
		struct semid_ds tbuf_old;

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

964 965 966
		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;
L
Linus Torvalds 已提交
967 968 969 970 971 972 973 974

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

975 976 977 978 979
/*
 * 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.
 */
980 981
static int semctl_down(struct ipc_namespace *ns, int semid,
		       int cmd, int version, union semun arg)
L
Linus Torvalds 已提交
982 983 984
{
	struct sem_array *sma;
	int err;
985
	struct semid64_ds semid64;
L
Linus Torvalds 已提交
986 987 988
	struct kern_ipc_perm *ipcp;

	if(cmd == IPC_SET) {
989
		if (copy_semid_from_user(&semid64, arg.buf, version))
L
Linus Torvalds 已提交
990 991
			return -EFAULT;
	}
S
Steve Grubb 已提交
992

993 994 995
	ipcp = ipcctl_pre_down(&sem_ids(ns), semid, cmd, &semid64.sem_perm, 0);
	if (IS_ERR(ipcp))
		return PTR_ERR(ipcp);
S
Steve Grubb 已提交
996

997
	sma = container_of(ipcp, struct sem_array, sem_perm);
L
Linus Torvalds 已提交
998 999 1000 1001 1002 1003 1004

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

	switch(cmd){
	case IPC_RMID:
1005
		freeary(ns, ipcp);
1006
		goto out_up;
L
Linus Torvalds 已提交
1007
	case IPC_SET:
1008
		ipc_update_perm(&semid64.sem_perm, ipcp);
L
Linus Torvalds 已提交
1009 1010 1011 1012 1013 1014 1015 1016
		sma->sem_ctime = get_seconds();
		break;
	default:
		err = -EINVAL;
	}

out_unlock:
	sem_unlock(sma);
1017 1018
out_up:
	up_write(&sem_ids(ns).rw_mutex);
L
Linus Torvalds 已提交
1019 1020 1021
	return err;
}

1022
SYSCALL_DEFINE(semctl)(int semid, int semnum, int cmd, union semun arg)
L
Linus Torvalds 已提交
1023 1024 1025
{
	int err = -EINVAL;
	int version;
K
Kirill Korotaev 已提交
1026
	struct ipc_namespace *ns;
L
Linus Torvalds 已提交
1027 1028 1029 1030 1031

	if (semid < 0)
		return -EINVAL;

	version = ipc_parse_version(&cmd);
K
Kirill Korotaev 已提交
1032
	ns = current->nsproxy->ipc_ns;
L
Linus Torvalds 已提交
1033 1034 1035 1036

	switch(cmd) {
	case IPC_INFO:
	case SEM_INFO:
1037
	case IPC_STAT:
L
Linus Torvalds 已提交
1038
	case SEM_STAT:
1039
		err = semctl_nolock(ns, semid, cmd, version, arg);
L
Linus Torvalds 已提交
1040 1041 1042 1043 1044 1045 1046 1047
		return err;
	case GETALL:
	case GETVAL:
	case GETPID:
	case GETNCNT:
	case GETZCNT:
	case SETVAL:
	case SETALL:
K
Kirill Korotaev 已提交
1048
		err = semctl_main(ns,semid,semnum,cmd,version,arg);
L
Linus Torvalds 已提交
1049 1050 1051
		return err;
	case IPC_RMID:
	case IPC_SET:
1052
		err = semctl_down(ns, semid, cmd, version, arg);
L
Linus Torvalds 已提交
1053 1054 1055 1056 1057
		return err;
	default:
		return -EINVAL;
	}
}
1058 1059 1060 1061 1062 1063 1064
#ifdef CONFIG_HAVE_SYSCALL_WRAPPERS
asmlinkage long SyS_semctl(int semid, int semnum, int cmd, union semun arg)
{
	return SYSC_semctl((int) semid, (int) semnum, (int) cmd, arg);
}
SYSCALL_ALIAS(sys_semctl, SyS_semctl);
#endif
L
Linus Torvalds 已提交
1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082

/* If the task doesn't already have a undo_list, then allocate one
 * here.  We guarantee there is only one thread using this undo list,
 * and current is THE ONE
 *
 * If this allocation and assignment succeeds, but later
 * portions of this code fail, there is no need to free the sem_undo_list.
 * Just let it stay associated with the task, and it'll be freed later
 * at exit time.
 *
 * This can block, so callers must hold no locks.
 */
static inline int get_undo_list(struct sem_undo_list **undo_listp)
{
	struct sem_undo_list *undo_list;

	undo_list = current->sysvsem.undo_list;
	if (!undo_list) {
1083
		undo_list = kzalloc(sizeof(*undo_list), GFP_KERNEL);
L
Linus Torvalds 已提交
1084 1085
		if (undo_list == NULL)
			return -ENOMEM;
I
Ingo Molnar 已提交
1086
		spin_lock_init(&undo_list->lock);
L
Linus Torvalds 已提交
1087
		atomic_set(&undo_list->refcnt, 1);
1088 1089
		INIT_LIST_HEAD(&undo_list->list_proc);

L
Linus Torvalds 已提交
1090 1091 1092 1093 1094 1095
		current->sysvsem.undo_list = undo_list;
	}
	*undo_listp = undo_list;
	return 0;
}

1096
static struct sem_undo *__lookup_undo(struct sem_undo_list *ulp, int semid)
L
Linus Torvalds 已提交
1097
{
1098
	struct sem_undo *un;
1099

1100 1101 1102
	list_for_each_entry_rcu(un, &ulp->list_proc, list_proc) {
		if (un->semid == semid)
			return un;
L
Linus Torvalds 已提交
1103
	}
1104
	return NULL;
L
Linus Torvalds 已提交
1105 1106
}

1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120
static struct sem_undo *lookup_undo(struct sem_undo_list *ulp, int semid)
{
	struct sem_undo *un;

  	assert_spin_locked(&ulp->lock);

	un = __lookup_undo(ulp, semid);
	if (un) {
		list_del_rcu(&un->list_proc);
		list_add_rcu(&un->list_proc, &ulp->list_proc);
	}
	return un;
}

1121 1122 1123 1124 1125 1126 1127 1128
/**
 * find_alloc_undo - Lookup (and if not present create) undo array
 * @ns: namespace
 * @semid: semaphore array id
 *
 * The function looks up (and if not present creates) the undo structure.
 * The size of the undo structure depends on the size of the semaphore
 * array, thus the alloc path is not that straightforward.
1129 1130
 * Lifetime-rules: sem_undo is rcu-protected, on success, the function
 * performs a rcu_read_lock().
1131 1132
 */
static struct sem_undo *find_alloc_undo(struct ipc_namespace *ns, int semid)
L
Linus Torvalds 已提交
1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143
{
	struct sem_array *sma;
	struct sem_undo_list *ulp;
	struct sem_undo *un, *new;
	int nsems;
	int error;

	error = get_undo_list(&ulp);
	if (error)
		return ERR_PTR(error);

1144
	rcu_read_lock();
1145
	spin_lock(&ulp->lock);
L
Linus Torvalds 已提交
1146
	un = lookup_undo(ulp, semid);
1147
	spin_unlock(&ulp->lock);
L
Linus Torvalds 已提交
1148 1149
	if (likely(un!=NULL))
		goto out;
1150
	rcu_read_unlock();
L
Linus Torvalds 已提交
1151 1152

	/* no undo structure around - allocate one. */
1153
	/* step 1: figure out the size of the semaphore array */
1154 1155 1156 1157
	sma = sem_lock_check(ns, semid);
	if (IS_ERR(sma))
		return ERR_PTR(PTR_ERR(sma));

L
Linus Torvalds 已提交
1158
	nsems = sma->sem_nsems;
1159
	sem_getref_and_unlock(sma);
L
Linus Torvalds 已提交
1160

1161
	/* step 2: allocate new undo structure */
1162
	new = kzalloc(sizeof(struct sem_undo) + sizeof(short)*nsems, GFP_KERNEL);
L
Linus Torvalds 已提交
1163
	if (!new) {
1164
		sem_putref(sma);
L
Linus Torvalds 已提交
1165 1166 1167
		return ERR_PTR(-ENOMEM);
	}

1168
	/* step 3: Acquire the lock on semaphore array */
1169
	sem_lock_and_putref(sma);
L
Linus Torvalds 已提交
1170 1171 1172 1173 1174 1175
	if (sma->sem_perm.deleted) {
		sem_unlock(sma);
		kfree(new);
		un = ERR_PTR(-EIDRM);
		goto out;
	}
1176 1177 1178 1179 1180 1181 1182 1183 1184 1185
	spin_lock(&ulp->lock);

	/*
	 * step 4: check for races: did someone else allocate the undo struct?
	 */
	un = lookup_undo(ulp, semid);
	if (un) {
		kfree(new);
		goto success;
	}
1186 1187
	/* step 5: initialize & link new undo structure */
	new->semadj = (short *) &new[1];
1188
	new->ulp = ulp;
1189 1190
	new->semid = semid;
	assert_spin_locked(&ulp->lock);
1191
	list_add_rcu(&new->list_proc, &ulp->list_proc);
1192 1193
	assert_spin_locked(&sma->sem_perm.lock);
	list_add(&new->list_id, &sma->list_id);
1194
	un = new;
1195

1196
success:
1197
	spin_unlock(&ulp->lock);
1198 1199
	rcu_read_lock();
	sem_unlock(sma);
L
Linus Torvalds 已提交
1200 1201 1202 1203
out:
	return un;
}

1204 1205
SYSCALL_DEFINE4(semtimedop, int, semid, struct sembuf __user *, tsops,
		unsigned, nsops, const struct timespec __user *, timeout)
L
Linus Torvalds 已提交
1206 1207 1208 1209 1210 1211
{
	int error = -EINVAL;
	struct sem_array *sma;
	struct sembuf fast_sops[SEMOPM_FAST];
	struct sembuf* sops = fast_sops, *sop;
	struct sem_undo *un;
1212
	int undos = 0, alter = 0, max;
L
Linus Torvalds 已提交
1213 1214
	struct sem_queue queue;
	unsigned long jiffies_left = 0;
K
Kirill Korotaev 已提交
1215 1216 1217
	struct ipc_namespace *ns;

	ns = current->nsproxy->ipc_ns;
L
Linus Torvalds 已提交
1218 1219 1220

	if (nsops < 1 || semid < 0)
		return -EINVAL;
K
Kirill Korotaev 已提交
1221
	if (nsops > ns->sc_semopm)
L
Linus Torvalds 已提交
1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249
		return -E2BIG;
	if(nsops > SEMOPM_FAST) {
		sops = kmalloc(sizeof(*sops)*nsops,GFP_KERNEL);
		if(sops==NULL)
			return -ENOMEM;
	}
	if (copy_from_user (sops, tsops, nsops * sizeof(*tsops))) {
		error=-EFAULT;
		goto out_free;
	}
	if (timeout) {
		struct timespec _timeout;
		if (copy_from_user(&_timeout, timeout, sizeof(*timeout))) {
			error = -EFAULT;
			goto out_free;
		}
		if (_timeout.tv_sec < 0 || _timeout.tv_nsec < 0 ||
			_timeout.tv_nsec >= 1000000000L) {
			error = -EINVAL;
			goto out_free;
		}
		jiffies_left = timespec_to_jiffies(&_timeout);
	}
	max = 0;
	for (sop = sops; sop < sops + nsops; sop++) {
		if (sop->sem_num >= max)
			max = sop->sem_num;
		if (sop->sem_flg & SEM_UNDO)
1250 1251
			undos = 1;
		if (sop->sem_op != 0)
L
Linus Torvalds 已提交
1252 1253 1254 1255
			alter = 1;
	}

	if (undos) {
1256
		un = find_alloc_undo(ns, semid);
L
Linus Torvalds 已提交
1257 1258 1259 1260 1261 1262 1263
		if (IS_ERR(un)) {
			error = PTR_ERR(un);
			goto out_free;
		}
	} else
		un = NULL;

1264 1265
	sma = sem_lock_check(ns, semid);
	if (IS_ERR(sma)) {
1266 1267
		if (un)
			rcu_read_unlock();
1268
		error = PTR_ERR(sma);
L
Linus Torvalds 已提交
1269
		goto out_free;
1270 1271
	}

L
Linus Torvalds 已提交
1272
	/*
1273
	 * semid identifiers are not unique - find_alloc_undo may have
L
Linus Torvalds 已提交
1274
	 * allocated an undo structure, it was invalidated by an RMID
1275
	 * and now a new array with received the same id. Check and fail.
1276 1277
	 * This case can be detected checking un->semid. The existance of
	 * "un" itself is guaranteed by rcu.
L
Linus Torvalds 已提交
1278
	 */
1279
	error = -EIDRM;
1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295
	if (un) {
		if (un->semid == -1) {
			rcu_read_unlock();
			goto out_unlock_free;
		} else {
			/*
			 * rcu lock can be released, "un" cannot disappear:
			 * - sem_lock is acquired, thus IPC_RMID is
			 *   impossible.
			 * - exit_sem is impossible, it always operates on
			 *   current (or a dead task).
			 */

			rcu_read_unlock();
		}
	}
1296

L
Linus Torvalds 已提交
1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308
	error = -EFBIG;
	if (max >= sma->sem_nsems)
		goto out_unlock_free;

	error = -EACCES;
	if (ipcperms(&sma->sem_perm, alter ? S_IWUGO : S_IRUGO))
		goto out_unlock_free;

	error = security_sem_semop(sma, sops, nsops, alter);
	if (error)
		goto out_unlock_free;

1309
	error = try_atomic_semop (sma, sops, nsops, un, task_tgid_vnr(current));
L
Linus Torvalds 已提交
1310 1311
	if (error <= 0) {
		if (alter && error == 0)
1312
			do_smart_update(sma, sops, nsops);
1313

L
Linus Torvalds 已提交
1314 1315 1316 1317 1318 1319 1320 1321 1322 1323
		goto out_unlock_free;
	}

	/* We need to sleep on this operation, so we put the current
	 * task into the pending queue and go to sleep.
	 */
		
	queue.sops = sops;
	queue.nsops = nsops;
	queue.undo = un;
1324
	queue.pid = task_tgid_vnr(current);
L
Linus Torvalds 已提交
1325 1326
	queue.alter = alter;
	if (alter)
1327
		list_add_tail(&queue.list, &sma->sem_pending);
L
Linus Torvalds 已提交
1328
	else
1329
		list_add(&queue.list, &sma->sem_pending);
L
Linus Torvalds 已提交
1330

1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343
	if (nsops == 1) {
		struct sem *curr;
		curr = &sma->sem_base[sops->sem_num];

		if (alter)
			list_add_tail(&queue.simple_list, &curr->sem_pending);
		else
			list_add(&queue.simple_list, &curr->sem_pending);
	} else {
		INIT_LIST_HEAD(&queue.simple_list);
		sma->complex_count++;
	}

L
Linus Torvalds 已提交
1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365
	queue.status = -EINTR;
	queue.sleeper = current;
	current->state = TASK_INTERRUPTIBLE;
	sem_unlock(sma);

	if (timeout)
		jiffies_left = schedule_timeout(jiffies_left);
	else
		schedule();

	error = queue.status;
	while(unlikely(error == IN_WAKEUP)) {
		cpu_relax();
		error = queue.status;
	}

	if (error != -EINTR) {
		/* fast path: update_queue already obtained all requested
		 * resources */
		goto out_free;
	}

K
Kirill Korotaev 已提交
1366
	sma = sem_lock(ns, semid);
1367
	if (IS_ERR(sma)) {
L
Linus Torvalds 已提交
1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384
		error = -EIDRM;
		goto out_free;
	}

	/*
	 * If queue.status != -EINTR we are woken up by another process
	 */
	error = queue.status;
	if (error != -EINTR) {
		goto out_unlock_free;
	}

	/*
	 * If an interrupt occurred we have to clean up the queue
	 */
	if (timeout && jiffies_left == 0)
		error = -EAGAIN;
1385
	unlink_queue(sma, &queue);
L
Linus Torvalds 已提交
1386 1387 1388 1389 1390 1391 1392 1393 1394

out_unlock_free:
	sem_unlock(sma);
out_free:
	if(sops != fast_sops)
		kfree(sops);
	return error;
}

1395 1396
SYSCALL_DEFINE3(semop, int, semid, struct sembuf __user *, tsops,
		unsigned, nsops)
L
Linus Torvalds 已提交
1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435
{
	return sys_semtimedop(semid, tsops, nsops, NULL);
}

/* If CLONE_SYSVSEM is set, establish sharing of SEM_UNDO state between
 * parent and child tasks.
 */

int copy_semundo(unsigned long clone_flags, struct task_struct *tsk)
{
	struct sem_undo_list *undo_list;
	int error;

	if (clone_flags & CLONE_SYSVSEM) {
		error = get_undo_list(&undo_list);
		if (error)
			return error;
		atomic_inc(&undo_list->refcnt);
		tsk->sysvsem.undo_list = undo_list;
	} else 
		tsk->sysvsem.undo_list = NULL;

	return 0;
}

/*
 * add semadj values to semaphores, free undo structures.
 * undo structures are not freed when semaphore arrays are destroyed
 * so some of them may be out of date.
 * IMPLEMENTATION NOTE: There is some confusion over whether the
 * set of adjustments that needs to be done should be done in an atomic
 * manner or not. That is, if we are attempting to decrement the semval
 * should we queue up and wait until we can do so legally?
 * The original implementation attempted to do this (queue and wait).
 * The current implementation does not do so. The POSIX standard
 * and SVID should be consulted to determine what behavior is mandated.
 */
void exit_sem(struct task_struct *tsk)
{
1436
	struct sem_undo_list *ulp;
L
Linus Torvalds 已提交
1437

1438 1439
	ulp = tsk->sysvsem.undo_list;
	if (!ulp)
L
Linus Torvalds 已提交
1440
		return;
1441
	tsk->sysvsem.undo_list = NULL;
L
Linus Torvalds 已提交
1442

1443
	if (!atomic_dec_and_test(&ulp->refcnt))
L
Linus Torvalds 已提交
1444 1445
		return;

1446
	for (;;) {
L
Linus Torvalds 已提交
1447
		struct sem_array *sma;
1448 1449
		struct sem_undo *un;
		int semid;
1450 1451
		int i;

1452
		rcu_read_lock();
1453 1454
		un = list_entry_rcu(ulp->list_proc.next,
				    struct sem_undo, list_proc);
1455 1456 1457 1458 1459
		if (&un->list_proc == &ulp->list_proc)
			semid = -1;
		 else
			semid = un->semid;
		rcu_read_unlock();
1460

1461 1462
		if (semid == -1)
			break;
L
Linus Torvalds 已提交
1463

1464
		sma = sem_lock_check(tsk->nsproxy->ipc_ns, un->semid);
L
Linus Torvalds 已提交
1465

1466 1467 1468
		/* exit_sem raced with IPC_RMID, nothing to do */
		if (IS_ERR(sma))
			continue;
L
Linus Torvalds 已提交
1469

1470
		un = __lookup_undo(ulp, semid);
1471 1472 1473 1474 1475 1476 1477 1478 1479
		if (un == NULL) {
			/* exit_sem raced with IPC_RMID+semget() that created
			 * exactly the same semid. Nothing to do.
			 */
			sem_unlock(sma);
			continue;
		}

		/* remove un from the linked lists */
1480 1481 1482
		assert_spin_locked(&sma->sem_perm.lock);
		list_del(&un->list_id);

1483 1484 1485 1486
		spin_lock(&ulp->lock);
		list_del_rcu(&un->list_proc);
		spin_unlock(&ulp->lock);

1487 1488
		/* perform adjustments registered in un */
		for (i = 0; i < sma->sem_nsems; i++) {
I
Ingo Molnar 已提交
1489
			struct sem * semaphore = &sma->sem_base[i];
1490 1491
			if (un->semadj[i]) {
				semaphore->semval += un->semadj[i];
L
Linus Torvalds 已提交
1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504
				/*
				 * Range checks of the new semaphore value,
				 * not defined by sus:
				 * - Some unices ignore the undo entirely
				 *   (e.g. HP UX 11i 11.22, Tru64 V5.1)
				 * - some cap the value (e.g. FreeBSD caps
				 *   at 0, but doesn't enforce SEMVMX)
				 *
				 * Linux caps the semaphore value, both at 0
				 * and at SEMVMX.
				 *
				 * 	Manfred <manfred@colorfullife.com>
				 */
I
Ingo Molnar 已提交
1505 1506 1507 1508
				if (semaphore->semval < 0)
					semaphore->semval = 0;
				if (semaphore->semval > SEMVMX)
					semaphore->semval = SEMVMX;
1509
				semaphore->sempid = task_tgid_vnr(current);
L
Linus Torvalds 已提交
1510 1511 1512 1513
			}
		}
		sma->sem_otime = get_seconds();
		/* maybe some queued-up processes were waiting for this */
1514
		update_queue(sma, -1);
L
Linus Torvalds 已提交
1515
		sem_unlock(sma);
1516 1517

		call_rcu(&un->rcu, free_un);
L
Linus Torvalds 已提交
1518
	}
1519
	kfree(ulp);
L
Linus Torvalds 已提交
1520 1521 1522
}

#ifdef CONFIG_PROC_FS
1523
static int sysvipc_sem_proc_show(struct seq_file *s, void *it)
L
Linus Torvalds 已提交
1524
{
1525 1526 1527
	struct sem_array *sma = it;

	return seq_printf(s,
1528
			  "%10d %10d  %4o %10u %5u %5u %5u %5u %10lu %10lu\n",
1529
			  sma->sem_perm.key,
N
Nadia Derbey 已提交
1530
			  sma->sem_perm.id,
1531 1532 1533 1534 1535 1536 1537 1538
			  sma->sem_perm.mode,
			  sma->sem_nsems,
			  sma->sem_perm.uid,
			  sma->sem_perm.gid,
			  sma->sem_perm.cuid,
			  sma->sem_perm.cgid,
			  sma->sem_otime,
			  sma->sem_ctime);
L
Linus Torvalds 已提交
1539 1540
}
#endif