signal.c 96.8 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13
/*
 *  linux/kernel/signal.c
 *
 *  Copyright (C) 1991, 1992  Linus Torvalds
 *
 *  1997-11-02  Modified for POSIX.1b signals by Richard Henderson
 *
 *  2003-06-02  Jim Houston - Concurrent Computer Corp.
 *		Changes to use preallocated sigqueue structures
 *		to allow signals to be sent reliably.
 */

#include <linux/slab.h>
14
#include <linux/export.h>
L
Linus Torvalds 已提交
15
#include <linux/init.h>
16
#include <linux/sched/mm.h>
17
#include <linux/sched/user.h>
18
#include <linux/sched/debug.h>
19
#include <linux/sched/task.h>
20
#include <linux/sched/task_stack.h>
21
#include <linux/sched/cputime.h>
L
Linus Torvalds 已提交
22 23 24
#include <linux/fs.h>
#include <linux/tty.h>
#include <linux/binfmts.h>
25
#include <linux/coredump.h>
L
Linus Torvalds 已提交
26 27 28
#include <linux/security.h>
#include <linux/syscalls.h>
#include <linux/ptrace.h>
29
#include <linux/signal.h>
30
#include <linux/signalfd.h>
31
#include <linux/ratelimit.h>
32
#include <linux/tracehook.h>
33
#include <linux/capability.h>
34
#include <linux/freezer.h>
35 36
#include <linux/pid_namespace.h>
#include <linux/nsproxy.h>
37
#include <linux/user_namespace.h>
38
#include <linux/uprobes.h>
A
Al Viro 已提交
39
#include <linux/compat.h>
40
#include <linux/cn_proc.h>
41
#include <linux/compiler.h>
42
#include <linux/posix-timers.h>
43

44 45
#define CREATE_TRACE_POINTS
#include <trace/events/signal.h>
46

L
Linus Torvalds 已提交
47
#include <asm/param.h>
48
#include <linux/uaccess.h>
L
Linus Torvalds 已提交
49 50
#include <asm/unistd.h>
#include <asm/siginfo.h>
51
#include <asm/cacheflush.h>
52
#include "audit.h"	/* audit_signal_info() */
L
Linus Torvalds 已提交
53 54 55 56 57

/*
 * SLAB caches for signal bits.
 */

58
static struct kmem_cache *sigqueue_cachep;
L
Linus Torvalds 已提交
59

60 61
int print_fatal_signals __read_mostly;

62
static void __user *sig_handler(struct task_struct *t, int sig)
63
{
64 65
	return t->sighand->action[sig - 1].sa.sa_handler;
}
66

67 68
static int sig_handler_ignored(void __user *handler, int sig)
{
69 70 71 72
	/* Is it explicitly or implicitly ignored? */
	return handler == SIG_IGN ||
		(handler == SIG_DFL && sig_kernel_ignore(sig));
}
L
Linus Torvalds 已提交
73

74
static int sig_task_ignored(struct task_struct *t, int sig, bool force)
L
Linus Torvalds 已提交
75
{
76
	void __user *handler;
L
Linus Torvalds 已提交
77

78 79 80
	handler = sig_handler(t, sig);

	if (unlikely(t->signal->flags & SIGNAL_UNKILLABLE) &&
81
			handler == SIG_DFL && !force)
82 83 84 85 86
		return 1;

	return sig_handler_ignored(handler, sig);
}

87
static int sig_ignored(struct task_struct *t, int sig, bool force)
88
{
L
Linus Torvalds 已提交
89 90 91 92 93
	/*
	 * Blocked signals are never ignored, since the
	 * signal handler may change by the time it is
	 * unblocked.
	 */
94
	if (sigismember(&t->blocked, sig) || sigismember(&t->real_blocked, sig))
L
Linus Torvalds 已提交
95 96
		return 0;

97
	if (!sig_task_ignored(t, sig, force))
98 99 100 101 102
		return 0;

	/*
	 * Tracers may want to know about even ignored signals.
	 */
T
Tejun Heo 已提交
103
	return !t->ptrace;
L
Linus Torvalds 已提交
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
}

/*
 * Re-calculate pending state from the set of locally pending
 * signals, globally pending signals, and blocked signals.
 */
static inline int has_pending_signals(sigset_t *signal, sigset_t *blocked)
{
	unsigned long ready;
	long i;

	switch (_NSIG_WORDS) {
	default:
		for (i = _NSIG_WORDS, ready = 0; --i >= 0 ;)
			ready |= signal->sig[i] &~ blocked->sig[i];
		break;

	case 4: ready  = signal->sig[3] &~ blocked->sig[3];
		ready |= signal->sig[2] &~ blocked->sig[2];
		ready |= signal->sig[1] &~ blocked->sig[1];
		ready |= signal->sig[0] &~ blocked->sig[0];
		break;

	case 2: ready  = signal->sig[1] &~ blocked->sig[1];
		ready |= signal->sig[0] &~ blocked->sig[0];
		break;

	case 1: ready  = signal->sig[0] &~ blocked->sig[0];
	}
	return ready !=	0;
}

#define PENDING(p,b) has_pending_signals(&(p)->signal, (b))

R
Roland McGrath 已提交
138
static int recalc_sigpending_tsk(struct task_struct *t)
L
Linus Torvalds 已提交
139
{
140
	if ((t->jobctl & JOBCTL_PENDING_MASK) ||
L
Linus Torvalds 已提交
141
	    PENDING(&t->pending, &t->blocked) ||
R
Roland McGrath 已提交
142
	    PENDING(&t->signal->shared_pending, &t->blocked)) {
L
Linus Torvalds 已提交
143
		set_tsk_thread_flag(t, TIF_SIGPENDING);
R
Roland McGrath 已提交
144 145
		return 1;
	}
146 147 148 149 150
	/*
	 * We must never clear the flag in another thread, or in current
	 * when it's possible the current syscall is returning -ERESTART*.
	 * So we don't clear it here, and only callers who know they should do.
	 */
R
Roland McGrath 已提交
151 152 153 154 155 156 157 158 159 160 161
	return 0;
}

/*
 * After recalculating TIF_SIGPENDING, we need to make sure the task wakes up.
 * This is superfluous when called on current, the wakeup is a harmless no-op.
 */
void recalc_sigpending_and_wake(struct task_struct *t)
{
	if (recalc_sigpending_tsk(t))
		signal_wake_up(t, 0);
L
Linus Torvalds 已提交
162 163 164 165
}

void recalc_sigpending(void)
{
T
Tejun Heo 已提交
166
	if (!recalc_sigpending_tsk(current) && !freezing(current))
167 168
		clear_thread_flag(TIF_SIGPENDING);

L
Linus Torvalds 已提交
169 170 171 172
}

/* Given the mask, find the first available signal that should be serviced. */

173 174
#define SYNCHRONOUS_MASK \
	(sigmask(SIGSEGV) | sigmask(SIGBUS) | sigmask(SIGILL) | \
175
	 sigmask(SIGTRAP) | sigmask(SIGFPE) | sigmask(SIGSYS))
176

177
int next_signal(struct sigpending *pending, sigset_t *mask)
L
Linus Torvalds 已提交
178 179 180
{
	unsigned long i, *s, *m, x;
	int sig = 0;
181

L
Linus Torvalds 已提交
182 183
	s = pending->signal.sig;
	m = mask->sig;
184 185 186 187 188 189 190 191 192 193 194 195 196

	/*
	 * Handle the first word specially: it contains the
	 * synchronous signals that need to be dequeued first.
	 */
	x = *s &~ *m;
	if (x) {
		if (x & SYNCHRONOUS_MASK)
			x &= SYNCHRONOUS_MASK;
		sig = ffz(~x) + 1;
		return sig;
	}

L
Linus Torvalds 已提交
197 198
	switch (_NSIG_WORDS) {
	default:
199 200 201 202 203 204 205
		for (i = 1; i < _NSIG_WORDS; ++i) {
			x = *++s &~ *++m;
			if (!x)
				continue;
			sig = ffz(~x) + i*_NSIG_BPW + 1;
			break;
		}
L
Linus Torvalds 已提交
206 207
		break;

208 209 210
	case 2:
		x = s[1] &~ m[1];
		if (!x)
L
Linus Torvalds 已提交
211
			break;
212
		sig = ffz(~x) + _NSIG_BPW + 1;
L
Linus Torvalds 已提交
213 214
		break;

215 216
	case 1:
		/* Nothing to do */
L
Linus Torvalds 已提交
217 218
		break;
	}
219

L
Linus Torvalds 已提交
220 221 222
	return sig;
}

223 224 225 226 227 228 229 230 231 232
static inline void print_dropped_signal(int sig)
{
	static DEFINE_RATELIMIT_STATE(ratelimit_state, 5 * HZ, 10);

	if (!print_fatal_signals)
		return;

	if (!__ratelimit(&ratelimit_state))
		return;

233
	pr_info("%s/%d: reached RLIMIT_SIGPENDING, dropped signal %d\n",
234 235 236
				current->comm, current->pid, sig);
}

237
/**
238
 * task_set_jobctl_pending - set jobctl pending bits
239
 * @task: target task
240
 * @mask: pending bits to set
241
 *
242 243 244 245 246 247 248 249 250 251 252 253
 * Clear @mask from @task->jobctl.  @mask must be subset of
 * %JOBCTL_PENDING_MASK | %JOBCTL_STOP_CONSUME | %JOBCTL_STOP_SIGMASK |
 * %JOBCTL_TRAPPING.  If stop signo is being set, the existing signo is
 * cleared.  If @task is already being killed or exiting, this function
 * becomes noop.
 *
 * CONTEXT:
 * Must be called with @task->sighand->siglock held.
 *
 * RETURNS:
 * %true if @mask is set, %false if made noop because @task was dying.
 */
254
bool task_set_jobctl_pending(struct task_struct *task, unsigned long mask)
255 256 257 258 259 260 261 262 263 264 265 266 267 268 269
{
	BUG_ON(mask & ~(JOBCTL_PENDING_MASK | JOBCTL_STOP_CONSUME |
			JOBCTL_STOP_SIGMASK | JOBCTL_TRAPPING));
	BUG_ON((mask & JOBCTL_TRAPPING) && !(mask & JOBCTL_PENDING_MASK));

	if (unlikely(fatal_signal_pending(task) || (task->flags & PF_EXITING)))
		return false;

	if (mask & JOBCTL_STOP_SIGMASK)
		task->jobctl &= ~JOBCTL_STOP_SIGMASK;

	task->jobctl |= mask;
	return true;
}

270
/**
271
 * task_clear_jobctl_trapping - clear jobctl trapping bit
272 273
 * @task: target task
 *
274 275 276 277
 * If JOBCTL_TRAPPING is set, a ptracer is waiting for us to enter TRACED.
 * Clear it and wake up the ptracer.  Note that we don't need any further
 * locking.  @task->siglock guarantees that @task->parent points to the
 * ptracer.
278 279 280 281
 *
 * CONTEXT:
 * Must be called with @task->sighand->siglock held.
 */
282
void task_clear_jobctl_trapping(struct task_struct *task)
283
{
284 285
	if (unlikely(task->jobctl & JOBCTL_TRAPPING)) {
		task->jobctl &= ~JOBCTL_TRAPPING;
286
		smp_mb();	/* advised by wake_up_bit() */
287
		wake_up_bit(&task->jobctl, JOBCTL_TRAPPING_BIT);
288 289 290
	}
}

291
/**
292
 * task_clear_jobctl_pending - clear jobctl pending bits
293
 * @task: target task
294
 * @mask: pending bits to clear
295
 *
296 297 298
 * Clear @mask from @task->jobctl.  @mask must be subset of
 * %JOBCTL_PENDING_MASK.  If %JOBCTL_STOP_PENDING is being cleared, other
 * STOP bits are cleared together.
299
 *
300 301
 * If clearing of @mask leaves no stop or trap pending, this function calls
 * task_clear_jobctl_trapping().
302 303 304 305
 *
 * CONTEXT:
 * Must be called with @task->sighand->siglock held.
 */
306
void task_clear_jobctl_pending(struct task_struct *task, unsigned long mask)
307
{
308 309 310 311 312 313
	BUG_ON(mask & ~JOBCTL_PENDING_MASK);

	if (mask & JOBCTL_STOP_PENDING)
		mask |= JOBCTL_STOP_CONSUME | JOBCTL_STOP_DEQUEUED;

	task->jobctl &= ~mask;
314 315 316

	if (!(task->jobctl & JOBCTL_PENDING_MASK))
		task_clear_jobctl_trapping(task);
317 318 319 320 321 322
}

/**
 * task_participate_group_stop - participate in a group stop
 * @task: task participating in a group stop
 *
323
 * @task has %JOBCTL_STOP_PENDING set and is participating in a group stop.
324
 * Group stop states are cleared and the group stop count is consumed if
325
 * %JOBCTL_STOP_CONSUME was set.  If the consumption completes the group
326
 * stop, the appropriate %SIGNAL_* flags are set.
327 328 329
 *
 * CONTEXT:
 * Must be called with @task->sighand->siglock held.
330 331 332 333
 *
 * RETURNS:
 * %true if group stop completion should be notified to the parent, %false
 * otherwise.
334 335 336 337
 */
static bool task_participate_group_stop(struct task_struct *task)
{
	struct signal_struct *sig = task->signal;
338
	bool consume = task->jobctl & JOBCTL_STOP_CONSUME;
339

340
	WARN_ON_ONCE(!(task->jobctl & JOBCTL_STOP_PENDING));
341

342
	task_clear_jobctl_pending(task, JOBCTL_STOP_PENDING);
343 344 345 346 347 348 349

	if (!consume)
		return false;

	if (!WARN_ON_ONCE(sig->group_stop_count == 0))
		sig->group_stop_count--;

350 351 352 353 354
	/*
	 * Tell the caller to notify completion iff we are entering into a
	 * fresh group stop.  Read comment in do_signal_stop() for details.
	 */
	if (!sig->group_stop_count && !(sig->flags & SIGNAL_STOP_STOPPED)) {
355
		signal_set_stop_flags(sig, SIGNAL_STOP_STOPPED);
356 357 358 359 360
		return true;
	}
	return false;
}

361 362 363
/*
 * allocate a new signal queue record
 * - this may be called without locks if and only if t == current, otherwise an
364
 *   appropriate lock must be held to stop the target task from exiting
365
 */
366 367
static struct sigqueue *
__sigqueue_alloc(int sig, struct task_struct *t, gfp_t flags, int override_rlimit)
L
Linus Torvalds 已提交
368 369
{
	struct sigqueue *q = NULL;
370
	struct user_struct *user;
L
Linus Torvalds 已提交
371

372
	/*
373 374
	 * Protect access to @t credentials. This can go away when all
	 * callers hold rcu read lock.
375
	 */
376
	rcu_read_lock();
D
David Howells 已提交
377
	user = get_uid(__task_cred(t)->user);
378
	atomic_inc(&user->sigpending);
379
	rcu_read_unlock();
380

L
Linus Torvalds 已提交
381
	if (override_rlimit ||
382
	    atomic_read(&user->sigpending) <=
383
			task_rlimit(t, RLIMIT_SIGPENDING)) {
L
Linus Torvalds 已提交
384
		q = kmem_cache_alloc(sigqueue_cachep, flags);
385 386 387 388
	} else {
		print_dropped_signal(sig);
	}

L
Linus Torvalds 已提交
389
	if (unlikely(q == NULL)) {
390
		atomic_dec(&user->sigpending);
D
David Howells 已提交
391
		free_uid(user);
L
Linus Torvalds 已提交
392 393 394
	} else {
		INIT_LIST_HEAD(&q->list);
		q->flags = 0;
D
David Howells 已提交
395
		q->user = user;
L
Linus Torvalds 已提交
396
	}
D
David Howells 已提交
397 398

	return q;
L
Linus Torvalds 已提交
399 400
}

401
static void __sigqueue_free(struct sigqueue *q)
L
Linus Torvalds 已提交
402 403 404 405 406 407 408 409
{
	if (q->flags & SIGQUEUE_PREALLOC)
		return;
	atomic_dec(&q->user->sigpending);
	free_uid(q->user);
	kmem_cache_free(sigqueue_cachep, q);
}

410
void flush_sigqueue(struct sigpending *queue)
L
Linus Torvalds 已提交
411 412 413 414 415 416 417 418 419 420 421 422
{
	struct sigqueue *q;

	sigemptyset(&queue->signal);
	while (!list_empty(&queue->list)) {
		q = list_entry(queue->list.next, struct sigqueue , list);
		list_del_init(&q->list);
		__sigqueue_free(q);
	}
}

/*
423
 * Flush all pending signals for this kthread.
L
Linus Torvalds 已提交
424
 */
425
void flush_signals(struct task_struct *t)
L
Linus Torvalds 已提交
426 427 428 429
{
	unsigned long flags;

	spin_lock_irqsave(&t->sighand->siglock, flags);
430 431 432
	clear_tsk_thread_flag(t, TIF_SIGPENDING);
	flush_sigqueue(&t->pending);
	flush_sigqueue(&t->signal->shared_pending);
L
Linus Torvalds 已提交
433 434 435
	spin_unlock_irqrestore(&t->sighand->siglock, flags);
}

436
#ifdef CONFIG_POSIX_TIMERS
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
static void __flush_itimer_signals(struct sigpending *pending)
{
	sigset_t signal, retain;
	struct sigqueue *q, *n;

	signal = pending->signal;
	sigemptyset(&retain);

	list_for_each_entry_safe(q, n, &pending->list, list) {
		int sig = q->info.si_signo;

		if (likely(q->info.si_code != SI_TIMER)) {
			sigaddset(&retain, sig);
		} else {
			sigdelset(&signal, sig);
			list_del_init(&q->list);
			__sigqueue_free(q);
		}
	}

	sigorsets(&pending->signal, &signal, &retain);
}

void flush_itimer_signals(void)
{
	struct task_struct *tsk = current;
	unsigned long flags;

	spin_lock_irqsave(&tsk->sighand->siglock, flags);
	__flush_itimer_signals(&tsk->pending);
	__flush_itimer_signals(&tsk->signal->shared_pending);
	spin_unlock_irqrestore(&tsk->sighand->siglock, flags);
}
470
#endif
471

472 473 474 475 476 477 478 479 480 481
void ignore_signals(struct task_struct *t)
{
	int i;

	for (i = 0; i < _NSIG; ++i)
		t->sighand->action[i].sa.sa_handler = SIG_IGN;

	flush_signals(t);
}

L
Linus Torvalds 已提交
482 483 484 485 486 487 488 489 490 491 492 493 494
/*
 * Flush all handlers for a task.
 */

void
flush_signal_handlers(struct task_struct *t, int force_default)
{
	int i;
	struct k_sigaction *ka = &t->sighand->action[0];
	for (i = _NSIG ; i != 0 ; i--) {
		if (force_default || ka->sa.sa_handler != SIG_IGN)
			ka->sa.sa_handler = SIG_DFL;
		ka->sa.sa_flags = 0;
495
#ifdef __ARCH_HAS_SA_RESTORER
496 497
		ka->sa.sa_restorer = NULL;
#endif
L
Linus Torvalds 已提交
498 499 500 501 502
		sigemptyset(&ka->sa.sa_mask);
		ka++;
	}
}

503 504
int unhandled_signal(struct task_struct *tsk, int sig)
{
505
	void __user *handler = tsk->sighand->action[sig-1].sa.sa_handler;
506
	if (is_global_init(tsk))
507
		return 1;
508
	if (handler != SIG_IGN && handler != SIG_DFL)
509
		return 0;
T
Tejun Heo 已提交
510 511
	/* if ptraced, let the tracer determine */
	return !tsk->ptrace;
512 513
}

514 515
static void collect_signal(int sig, struct sigpending *list, siginfo_t *info,
			   bool *resched_timer)
L
Linus Torvalds 已提交
516 517 518 519 520 521 522 523 524
{
	struct sigqueue *q, *first = NULL;

	/*
	 * Collect the siginfo appropriate to this signal.  Check if
	 * there is another siginfo for the same signal.
	*/
	list_for_each_entry(q, &list->list, list) {
		if (q->info.si_signo == sig) {
525 526
			if (first)
				goto still_pending;
L
Linus Torvalds 已提交
527 528 529
			first = q;
		}
	}
530 531 532

	sigdelset(&list->signal, sig);

L
Linus Torvalds 已提交
533
	if (first) {
534
still_pending:
L
Linus Torvalds 已提交
535 536
		list_del_init(&first->list);
		copy_siginfo(info, &first->info);
537 538 539 540 541 542

		*resched_timer =
			(first->flags & SIGQUEUE_PREALLOC) &&
			(info->si_code == SI_TIMER) &&
			(info->si_sys_private);

L
Linus Torvalds 已提交
543 544
		__sigqueue_free(first);
	} else {
545 546 547 548
		/*
		 * Ok, it wasn't in the queue.  This must be
		 * a fast-pathed signal or we must have been
		 * out of queue space.  So zero out the info.
L
Linus Torvalds 已提交
549 550 551
		 */
		info->si_signo = sig;
		info->si_errno = 0;
552
		info->si_code = SI_USER;
L
Linus Torvalds 已提交
553 554 555 556 557 558
		info->si_pid = 0;
		info->si_uid = 0;
	}
}

static int __dequeue_signal(struct sigpending *pending, sigset_t *mask,
559
			siginfo_t *info, bool *resched_timer)
L
Linus Torvalds 已提交
560
{
561
	int sig = next_signal(pending, mask);
L
Linus Torvalds 已提交
562

563
	if (sig)
564
		collect_signal(sig, pending, info, resched_timer);
L
Linus Torvalds 已提交
565 566 567 568
	return sig;
}

/*
569
 * Dequeue a signal and return the element to the caller, which is
L
Linus Torvalds 已提交
570 571 572 573 574 575
 * expected to free it.
 *
 * All callers have to hold the siglock.
 */
int dequeue_signal(struct task_struct *tsk, sigset_t *mask, siginfo_t *info)
{
576
	bool resched_timer = false;
577
	int signr;
578 579 580 581

	/* We only dequeue private signals from ourselves, we don't let
	 * signalfd steal them
	 */
582
	signr = __dequeue_signal(&tsk->pending, mask, info, &resched_timer);
583
	if (!signr) {
L
Linus Torvalds 已提交
584
		signr = __dequeue_signal(&tsk->signal->shared_pending,
585
					 mask, info, &resched_timer);
586
#ifdef CONFIG_POSIX_TIMERS
587 588 589 590 591 592
		/*
		 * itimer signal ?
		 *
		 * itimers are process shared and we restart periodic
		 * itimers in the signal delivery path to prevent DoS
		 * attacks in the high resolution timer case. This is
593
		 * compliant with the old way of self-restarting
594 595 596 597 598 599 600 601 602 603
		 * itimers, as the SIGALRM is a legacy signal and only
		 * queued once. Changing the restart behaviour to
		 * restart the timer in the signal dequeue path is
		 * reducing the timer noise on heavy loaded !highres
		 * systems too.
		 */
		if (unlikely(signr == SIGALRM)) {
			struct hrtimer *tmr = &tsk->signal->real_timer;

			if (!hrtimer_is_queued(tmr) &&
T
Thomas Gleixner 已提交
604
			    tsk->signal->it_real_incr != 0) {
605 606 607 608 609
				hrtimer_forward(tmr, tmr->base->get_time(),
						tsk->signal->it_real_incr);
				hrtimer_restart(tmr);
			}
		}
610
#endif
611
	}
612

D
Davide Libenzi 已提交
613
	recalc_sigpending();
614 615 616 617
	if (!signr)
		return 0;

	if (unlikely(sig_kernel_stop(signr))) {
618 619 620 621 622 623 624 625 626 627 628 629
		/*
		 * Set a marker that we have dequeued a stop signal.  Our
		 * caller might release the siglock and then the pending
		 * stop signal it is about to process is no longer in the
		 * pending bitmasks, but must still be cleared by a SIGCONT
		 * (and overruled by a SIGKILL).  So those cases clear this
		 * shared flag after we've set it.  Note that this flag may
		 * remain set after the signal we return is ignored or
		 * handled.  That doesn't matter because its only purpose
		 * is to alert stop-signal processing code when another
		 * processor has come along and cleared the flag.
		 */
630
		current->jobctl |= JOBCTL_STOP_DEQUEUED;
631
	}
632
#ifdef CONFIG_POSIX_TIMERS
633
	if (resched_timer) {
L
Linus Torvalds 已提交
634 635 636 637 638 639 640
		/*
		 * Release the siglock to ensure proper locking order
		 * of timer locks outside of siglocks.  Note, we leave
		 * irqs disabled here, since the posix-timers code is
		 * about to disable them again anyway.
		 */
		spin_unlock(&tsk->sighand->siglock);
641
		posixtimer_rearm(info);
L
Linus Torvalds 已提交
642 643
		spin_lock(&tsk->sighand->siglock);
	}
644
#endif
L
Linus Torvalds 已提交
645 646 647 648 649 650 651 652 653 654 655 656 657 658
	return signr;
}

/*
 * Tell a process that it has a new active signal..
 *
 * NOTE! we rely on the previous spin_lock to
 * lock interrupts for us! We can only be called with
 * "siglock" held, and the local interrupt must
 * have been disabled when that got acquired!
 *
 * No need to set need_resched since signal event passing
 * goes through ->blocked
 */
659
void signal_wake_up_state(struct task_struct *t, unsigned int state)
L
Linus Torvalds 已提交
660 661 662
{
	set_tsk_thread_flag(t, TIF_SIGPENDING);
	/*
663
	 * TASK_WAKEKILL also means wake it up in the stopped/traced/killable
M
Matthew Wilcox 已提交
664
	 * case. We don't check t->state here because there is a race with it
L
Linus Torvalds 已提交
665 666 667 668
	 * executing another processor and just now entering stopped state.
	 * By using wake_up_state, we ensure the process will wake up and
	 * handle its death signal.
	 */
669
	if (!wake_up_state(t, state | TASK_INTERRUPTIBLE))
L
Linus Torvalds 已提交
670 671 672
		kick_process(t);
}

673 674 675 676 677 678
/*
 * Remove signals in mask from the pending set and queue.
 * Returns 1 if any signals were found.
 *
 * All callers must be holding the siglock.
 */
679
static int flush_sigqueue_mask(sigset_t *mask, struct sigpending *s)
680 681 682 683 684 685 686 687
{
	struct sigqueue *q, *n;
	sigset_t m;

	sigandsets(&m, mask, &s->signal);
	if (sigisemptyset(&m))
		return 0;

688
	sigandnsets(&s->signal, &s->signal, mask);
689 690 691 692 693 694 695 696
	list_for_each_entry_safe(q, n, &s->list, list) {
		if (sigismember(mask, q->info.si_signo)) {
			list_del_init(&q->list);
			__sigqueue_free(q);
		}
	}
	return 1;
}
L
Linus Torvalds 已提交
697

698 699 700 701 702 703 704 705 706 707 708
static inline int is_si_special(const struct siginfo *info)
{
	return info <= SEND_SIG_FORCED;
}

static inline bool si_fromuser(const struct siginfo *info)
{
	return info == SEND_SIG_NOINFO ||
		(!is_si_special(info) && SI_FROMUSER(info));
}

709 710 711 712 713 714 715 716
/*
 * called with RCU read lock from check_kill_permission()
 */
static int kill_ok_by_cred(struct task_struct *t)
{
	const struct cred *cred = current_cred();
	const struct cred *tcred = __task_cred(t);

717 718 719 720
	if (uid_eq(cred->euid, tcred->suid) ||
	    uid_eq(cred->euid, tcred->uid)  ||
	    uid_eq(cred->uid,  tcred->suid) ||
	    uid_eq(cred->uid,  tcred->uid))
721 722
		return 1;

723
	if (ns_capable(tcred->user_ns, CAP_KILL))
724 725 726 727 728
		return 1;

	return 0;
}

L
Linus Torvalds 已提交
729 730
/*
 * Bad permissions for sending the signal
731
 * - the caller must hold the RCU read lock
L
Linus Torvalds 已提交
732 733 734 735
 */
static int check_kill_permission(int sig, struct siginfo *info,
				 struct task_struct *t)
{
736
	struct pid *sid;
737 738
	int error;

739
	if (!valid_signal(sig))
740 741
		return -EINVAL;

742
	if (!si_fromuser(info))
743
		return 0;
A
Amy Griffis 已提交
744

745 746
	error = audit_signal_info(sig, t); /* Let audit system see the signal */
	if (error)
L
Linus Torvalds 已提交
747
		return error;
748

749
	if (!same_thread_group(current, t) &&
750
	    !kill_ok_by_cred(t)) {
751 752 753 754 755 756 757 758 759 760 761 762 763
		switch (sig) {
		case SIGCONT:
			sid = task_session(t);
			/*
			 * We don't return the error if sid == NULL. The
			 * task was unhashed, the caller must notice this.
			 */
			if (!sid || sid == task_session(current))
				break;
		default:
			return -EPERM;
		}
	}
764

A
Amy Griffis 已提交
765
	return security_task_kill(t, info, sig, 0);
L
Linus Torvalds 已提交
766 767
}

768 769 770 771 772 773 774 775
/**
 * ptrace_trap_notify - schedule trap to notify ptracer
 * @t: tracee wanting to notify tracer
 *
 * This function schedules sticky ptrace trap which is cleared on the next
 * TRAP_STOP to notify ptracer of an event.  @t must have been seized by
 * ptracer.
 *
T
Tejun Heo 已提交
776 777 778 779 780
 * If @t is running, STOP trap will be taken.  If trapped for STOP and
 * ptracer is listening for events, tracee is woken up so that it can
 * re-trap for the new event.  If trapped otherwise, STOP trap will be
 * eventually taken without returning to userland after the existing traps
 * are finished by PTRACE_CONT.
781 782 783 784 785 786 787 788 789 790
 *
 * CONTEXT:
 * Must be called with @task->sighand->siglock held.
 */
static void ptrace_trap_notify(struct task_struct *t)
{
	WARN_ON_ONCE(!(t->ptrace & PT_SEIZED));
	assert_spin_locked(&t->sighand->siglock);

	task_set_jobctl_pending(t, JOBCTL_TRAP_NOTIFY);
791
	ptrace_signal_wake_up(t, t->jobctl & JOBCTL_LISTENING);
792 793
}

L
Linus Torvalds 已提交
794
/*
795 796
 * Handle magic process-wide effects of stop/continue signals. Unlike
 * the signal actions, these happen immediately at signal-generation
L
Linus Torvalds 已提交
797 798
 * time regardless of blocking, ignoring, or handling.  This does the
 * actual continuing for SIGCONT, but not the actual stopping for stop
799 800 801 802
 * signals. The process stop is done as a signal action for SIG_DFL.
 *
 * Returns true if the signal should be actually delivered, otherwise
 * it should be dropped.
L
Linus Torvalds 已提交
803
 */
804
static bool prepare_signal(int sig, struct task_struct *p, bool force)
L
Linus Torvalds 已提交
805
{
806
	struct signal_struct *signal = p->signal;
L
Linus Torvalds 已提交
807
	struct task_struct *t;
808
	sigset_t flush;
L
Linus Torvalds 已提交
809

810
	if (signal->flags & (SIGNAL_GROUP_EXIT | SIGNAL_GROUP_COREDUMP)) {
811
		if (!(signal->flags & SIGNAL_GROUP_EXIT))
812
			return sig == SIGKILL;
L
Linus Torvalds 已提交
813
		/*
814
		 * The process is in the middle of dying, nothing to do.
L
Linus Torvalds 已提交
815
		 */
816
	} else if (sig_kernel_stop(sig)) {
L
Linus Torvalds 已提交
817 818 819
		/*
		 * This is a stop signal.  Remove SIGCONT from all queues.
		 */
820
		siginitset(&flush, sigmask(SIGCONT));
821
		flush_sigqueue_mask(&flush, &signal->shared_pending);
822
		for_each_thread(p, t)
823
			flush_sigqueue_mask(&flush, &t->pending);
L
Linus Torvalds 已提交
824
	} else if (sig == SIGCONT) {
825
		unsigned int why;
L
Linus Torvalds 已提交
826
		/*
827
		 * Remove all stop signals from all queues, wake all threads.
L
Linus Torvalds 已提交
828
		 */
829
		siginitset(&flush, SIG_KERNEL_STOP_MASK);
830
		flush_sigqueue_mask(&flush, &signal->shared_pending);
831
		for_each_thread(p, t) {
832
			flush_sigqueue_mask(&flush, &t->pending);
833
			task_clear_jobctl_pending(t, JOBCTL_STOP_PENDING);
834 835 836 837
			if (likely(!(t->ptrace & PT_SEIZED)))
				wake_up_state(t, __TASK_STOPPED);
			else
				ptrace_trap_notify(t);
838
		}
L
Linus Torvalds 已提交
839

840 841 842 843 844 845 846 847 848
		/*
		 * Notify the parent with CLD_CONTINUED if we were stopped.
		 *
		 * If we were in the middle of a group stop, we pretend it
		 * was already finished, and then continued. Since SIGCHLD
		 * doesn't queue we report only CLD_STOPPED, as if the next
		 * CLD_CONTINUED was dropped.
		 */
		why = 0;
849
		if (signal->flags & SIGNAL_STOP_STOPPED)
850
			why |= SIGNAL_CLD_CONTINUED;
851
		else if (signal->group_stop_count)
852 853 854
			why |= SIGNAL_CLD_STOPPED;

		if (why) {
855
			/*
856
			 * The first thread which returns from do_signal_stop()
857 858 859
			 * will take ->siglock, notice SIGNAL_CLD_MASK, and
			 * notify its parent. See get_signal_to_deliver().
			 */
860
			signal_set_stop_flags(signal, why | SIGNAL_STOP_CONTINUED);
861 862
			signal->group_stop_count = 0;
			signal->group_exit_code = 0;
L
Linus Torvalds 已提交
863 864
		}
	}
865

866
	return !sig_ignored(p, sig, force);
L
Linus Torvalds 已提交
867 868
}

869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889
/*
 * Test if P wants to take SIG.  After we've checked all threads with this,
 * it's equivalent to finding no threads not blocking SIG.  Any threads not
 * blocking SIG were ruled out because they are not running and already
 * have pending signals.  Such threads will dequeue from the shared queue
 * as soon as they're available, so putting the signal on the shared queue
 * will be equivalent to sending it to one such thread.
 */
static inline int wants_signal(int sig, struct task_struct *p)
{
	if (sigismember(&p->blocked, sig))
		return 0;
	if (p->flags & PF_EXITING)
		return 0;
	if (sig == SIGKILL)
		return 1;
	if (task_is_stopped_or_traced(p))
		return 0;
	return task_curr(p) || !signal_pending(p);
}

890
static void complete_signal(int sig, struct task_struct *p, int group)
891 892 893 894 895 896 897 898 899 900 901 902
{
	struct signal_struct *signal = p->signal;
	struct task_struct *t;

	/*
	 * Now find a thread we can wake up to take the signal off the queue.
	 *
	 * If the main thread wants the signal, it gets first crack.
	 * Probably the least surprising to the average bear.
	 */
	if (wants_signal(sig, p))
		t = p;
903
	else if (!group || thread_group_empty(p))
904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930
		/*
		 * There is just one thread and it does not need to be woken.
		 * It will dequeue unblocked signals before it runs again.
		 */
		return;
	else {
		/*
		 * Otherwise try to find a suitable thread.
		 */
		t = signal->curr_target;
		while (!wants_signal(sig, t)) {
			t = next_thread(t);
			if (t == signal->curr_target)
				/*
				 * No thread needs to be woken.
				 * Any eligible threads will see
				 * the signal in the queue soon.
				 */
				return;
		}
		signal->curr_target = t;
	}

	/*
	 * Found a killable thread.  If the signal will be fatal,
	 * then start taking the whole group down immediately.
	 */
931 932
	if (sig_fatal(p, sig) &&
	    !(signal->flags & (SIGNAL_UNKILLABLE | SIGNAL_GROUP_EXIT)) &&
933
	    !sigismember(&t->real_blocked, sig) &&
T
Tejun Heo 已提交
934
	    (sig == SIGKILL || !t->ptrace)) {
935 936 937 938 939 940 941 942 943 944 945 946 947 948 949
		/*
		 * This signal will be fatal to the whole group.
		 */
		if (!sig_kernel_coredump(sig)) {
			/*
			 * Start a group exit and wake everybody up.
			 * This way we don't have other threads
			 * running and doing things after a slower
			 * thread has the fatal signal pending.
			 */
			signal->flags = SIGNAL_GROUP_EXIT;
			signal->group_exit_code = sig;
			signal->group_stop_count = 0;
			t = p;
			do {
950
				task_clear_jobctl_pending(t, JOBCTL_PENDING_MASK);
951 952 953 954 955 956 957 958 959 960 961 962 963 964 965
				sigaddset(&t->pending.signal, SIGKILL);
				signal_wake_up(t, 1);
			} while_each_thread(p, t);
			return;
		}
	}

	/*
	 * The signal is already in the shared-pending queue.
	 * Tell the chosen thread to wake up and dequeue it.
	 */
	signal_wake_up(t, sig == SIGKILL);
	return;
}

966 967 968 969 970
static inline int legacy_queue(struct sigpending *signals, int sig)
{
	return (sig < SIGRTMIN) && sigismember(&signals->signal, sig);
}

971 972 973 974 975 976 977 978 979
#ifdef CONFIG_USER_NS
static inline void userns_fixup_signal_uid(struct siginfo *info, struct task_struct *t)
{
	if (current_user_ns() == task_cred_xxx(t, user_ns))
		return;

	if (SI_FROMKERNEL(info))
		return;

980 981 982 983
	rcu_read_lock();
	info->si_uid = from_kuid_munged(task_cred_xxx(t, user_ns),
					make_kuid(current_user_ns(), info->si_uid));
	rcu_read_unlock();
984 985 986 987 988 989 990 991
}
#else
static inline void userns_fixup_signal_uid(struct siginfo *info, struct task_struct *t)
{
	return;
}
#endif

992 993
static int __send_signal(int sig, struct siginfo *info, struct task_struct *t,
			int group, int from_ancestor_ns)
L
Linus Torvalds 已提交
994
{
995
	struct sigpending *pending;
996
	struct sigqueue *q;
997
	int override_rlimit;
998
	int ret = 0, result;
999

1000
	assert_spin_locked(&t->sighand->siglock);
1001

1002
	result = TRACE_SIGNAL_IGNORED;
1003 1004
	if (!prepare_signal(sig, t,
			from_ancestor_ns || (info == SEND_SIG_FORCED)))
1005
		goto ret;
1006 1007

	pending = group ? &t->signal->shared_pending : &t->pending;
1008 1009 1010 1011 1012
	/*
	 * Short-circuit ignored signals and support queuing
	 * exactly one non-rt signal, so that we can get more
	 * detailed information about the cause of the signal.
	 */
1013
	result = TRACE_SIGNAL_ALREADY_PENDING;
1014
	if (legacy_queue(pending, sig))
1015 1016 1017
		goto ret;

	result = TRACE_SIGNAL_DELIVERED;
L
Linus Torvalds 已提交
1018 1019 1020 1021
	/*
	 * fast-pathed signals for kernel-internal things like SIGSTOP
	 * or SIGKILL.
	 */
1022
	if (info == SEND_SIG_FORCED)
L
Linus Torvalds 已提交
1023 1024
		goto out_set;

1025 1026 1027 1028 1029 1030 1031 1032 1033
	/*
	 * Real-time signals must be queued if sent by sigqueue, or
	 * some other real-time mechanism.  It is implementation
	 * defined whether kill() does so.  We attempt to do so, on
	 * the principle of least surprise, but since kill is not
	 * allowed to fail with EAGAIN when low on memory we just
	 * make sure at least one signal gets delivered and don't
	 * pass on the info struct.
	 */
1034 1035 1036 1037 1038
	if (sig < SIGRTMIN)
		override_rlimit = (is_si_special(info) || info->si_code >= 0);
	else
		override_rlimit = 0;

1039
	q = __sigqueue_alloc(sig, t, GFP_ATOMIC | __GFP_NOTRACK_FALSE_POSITIVE,
1040
		override_rlimit);
L
Linus Torvalds 已提交
1041
	if (q) {
1042
		list_add_tail(&q->list, &pending->list);
L
Linus Torvalds 已提交
1043
		switch ((unsigned long) info) {
1044
		case (unsigned long) SEND_SIG_NOINFO:
L
Linus Torvalds 已提交
1045 1046 1047
			q->info.si_signo = sig;
			q->info.si_errno = 0;
			q->info.si_code = SI_USER;
1048
			q->info.si_pid = task_tgid_nr_ns(current,
1049
							task_active_pid_ns(t));
1050
			q->info.si_uid = from_kuid_munged(current_user_ns(), current_uid());
L
Linus Torvalds 已提交
1051
			break;
1052
		case (unsigned long) SEND_SIG_PRIV:
L
Linus Torvalds 已提交
1053 1054 1055 1056 1057 1058 1059 1060
			q->info.si_signo = sig;
			q->info.si_errno = 0;
			q->info.si_code = SI_KERNEL;
			q->info.si_pid = 0;
			q->info.si_uid = 0;
			break;
		default:
			copy_siginfo(&q->info, info);
1061 1062
			if (from_ancestor_ns)
				q->info.si_pid = 0;
L
Linus Torvalds 已提交
1063 1064
			break;
		}
1065 1066 1067

		userns_fixup_signal_uid(&q->info, t);

1068
	} else if (!is_si_special(info)) {
1069 1070 1071 1072 1073 1074
		if (sig >= SIGRTMIN && info->si_code != SI_USER) {
			/*
			 * Queue overflow, abort.  We may abort if the
			 * signal was rt and sent by user using something
			 * other than kill().
			 */
1075 1076 1077
			result = TRACE_SIGNAL_OVERFLOW_FAIL;
			ret = -EAGAIN;
			goto ret;
1078 1079 1080 1081 1082
		} else {
			/*
			 * This is a silent loss of information.  We still
			 * send the signal, but the *info bits are lost.
			 */
1083
			result = TRACE_SIGNAL_LOSE_INFO;
1084
		}
L
Linus Torvalds 已提交
1085 1086 1087
	}

out_set:
1088
	signalfd_notify(t, sig);
1089
	sigaddset(&pending->signal, sig);
1090
	complete_signal(sig, t, group);
1091 1092 1093
ret:
	trace_signal_generate(sig, info, t, group, result);
	return ret;
L
Linus Torvalds 已提交
1094 1095
}

1096 1097 1098
static int send_signal(int sig, struct siginfo *info, struct task_struct *t,
			int group)
{
1099 1100 1101
	int from_ancestor_ns = 0;

#ifdef CONFIG_PID_NS
1102 1103
	from_ancestor_ns = si_fromuser(info) &&
			   !task_pid_nr_ns(current, task_active_pid_ns(t));
1104 1105 1106
#endif

	return __send_signal(sig, info, t, group, from_ancestor_ns);
1107 1108
}

1109
static void print_fatal_signal(int signr)
I
Ingo Molnar 已提交
1110
{
1111
	struct pt_regs *regs = signal_pt_regs();
1112
	pr_info("potentially unexpected fatal signal %d.\n", signr);
I
Ingo Molnar 已提交
1113

A
Al Viro 已提交
1114
#if defined(__i386__) && !defined(__arch_um__)
1115
	pr_info("code at %08lx: ", regs->ip);
I
Ingo Molnar 已提交
1116 1117 1118 1119 1120
	{
		int i;
		for (i = 0; i < 16; i++) {
			unsigned char insn;

1121 1122
			if (get_user(insn, (unsigned char *)(regs->ip + i)))
				break;
1123
			pr_cont("%02x ", insn);
I
Ingo Molnar 已提交
1124 1125
		}
	}
1126
	pr_cont("\n");
I
Ingo Molnar 已提交
1127
#endif
1128
	preempt_disable();
I
Ingo Molnar 已提交
1129
	show_regs(regs);
1130
	preempt_enable();
I
Ingo Molnar 已提交
1131 1132 1133 1134 1135 1136 1137 1138 1139 1140
}

static int __init setup_print_fatal_signals(char *str)
{
	get_option (&str, &print_fatal_signals);

	return 1;
}

__setup("print-fatal-signals=", setup_print_fatal_signals);
L
Linus Torvalds 已提交
1141

1142 1143 1144 1145 1146 1147
int
__group_send_sig_info(int sig, struct siginfo *info, struct task_struct *p)
{
	return send_signal(sig, info, p, 1);
}

L
Linus Torvalds 已提交
1148 1149 1150
static int
specific_send_sig_info(int sig, struct siginfo *info, struct task_struct *t)
{
1151
	return send_signal(sig, info, t, 0);
L
Linus Torvalds 已提交
1152 1153
}

1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167
int do_send_sig_info(int sig, struct siginfo *info, struct task_struct *p,
			bool group)
{
	unsigned long flags;
	int ret = -ESRCH;

	if (lock_task_sighand(p, &flags)) {
		ret = send_signal(sig, info, p, group);
		unlock_task_sighand(p, &flags);
	}

	return ret;
}

L
Linus Torvalds 已提交
1168 1169 1170
/*
 * Force a signal that the process can't ignore: if necessary
 * we unblock the signal and change any SIG_IGN to SIG_DFL.
1171 1172 1173 1174 1175
 *
 * Note: If we unblock the signal, we always reset it to SIG_DFL,
 * since we do not want to have a signal handler that was blocked
 * be invoked when user space had explicitly blocked it.
 *
1176 1177
 * We don't want to have recursive SIGSEGV's etc, for example,
 * that is why we also clear SIGNAL_UNKILLABLE.
L
Linus Torvalds 已提交
1178 1179 1180 1181 1182
 */
int
force_sig_info(int sig, struct siginfo *info, struct task_struct *t)
{
	unsigned long int flags;
1183 1184
	int ret, blocked, ignored;
	struct k_sigaction *action;
L
Linus Torvalds 已提交
1185 1186

	spin_lock_irqsave(&t->sighand->siglock, flags);
1187 1188 1189 1190 1191 1192 1193
	action = &t->sighand->action[sig-1];
	ignored = action->sa.sa_handler == SIG_IGN;
	blocked = sigismember(&t->blocked, sig);
	if (blocked || ignored) {
		action->sa.sa_handler = SIG_DFL;
		if (blocked) {
			sigdelset(&t->blocked, sig);
R
Roland McGrath 已提交
1194
			recalc_sigpending_and_wake(t);
1195
		}
L
Linus Torvalds 已提交
1196
	}
1197 1198 1199 1200 1201
	/*
	 * Don't clear SIGNAL_UNKILLABLE for traced tasks, users won't expect
	 * debugging to leave init killable.
	 */
	if (action->sa.sa_handler == SIG_DFL && !t->ptrace)
1202
		t->signal->flags &= ~SIGNAL_UNKILLABLE;
L
Linus Torvalds 已提交
1203 1204 1205 1206 1207 1208 1209 1210 1211
	ret = specific_send_sig_info(sig, info, t);
	spin_unlock_irqrestore(&t->sighand->siglock, flags);

	return ret;
}

/*
 * Nuke all other threads in the group.
 */
1212
int zap_other_threads(struct task_struct *p)
L
Linus Torvalds 已提交
1213
{
1214 1215
	struct task_struct *t = p;
	int count = 0;
L
Linus Torvalds 已提交
1216 1217 1218

	p->signal->group_stop_count = 0;

1219
	while_each_thread(p, t) {
1220
		task_clear_jobctl_pending(t, JOBCTL_PENDING_MASK);
1221 1222 1223
		count++;

		/* Don't bother with already dead threads */
L
Linus Torvalds 已提交
1224 1225 1226 1227 1228
		if (t->exit_state)
			continue;
		sigaddset(&t->pending.signal, SIGKILL);
		signal_wake_up(t, 1);
	}
1229 1230

	return count;
L
Linus Torvalds 已提交
1231 1232
}

1233 1234
struct sighand_struct *__lock_task_sighand(struct task_struct *tsk,
					   unsigned long *flags)
1235 1236 1237 1238
{
	struct sighand_struct *sighand;

	for (;;) {
1239 1240 1241 1242
		/*
		 * Disable interrupts early to avoid deadlocks.
		 * See rcu_read_unlock() comment header for details.
		 */
1243 1244
		local_irq_save(*flags);
		rcu_read_lock();
1245
		sighand = rcu_dereference(tsk->sighand);
1246 1247 1248
		if (unlikely(sighand == NULL)) {
			rcu_read_unlock();
			local_irq_restore(*flags);
1249
			break;
1250
		}
1251 1252
		/*
		 * This sighand can be already freed and even reused, but
1253
		 * we rely on SLAB_TYPESAFE_BY_RCU and sighand_ctor() which
1254 1255 1256 1257 1258 1259 1260 1261
		 * initializes ->siglock: this slab can't go away, it has
		 * the same object type, ->siglock can't be reinitialized.
		 *
		 * We need to ensure that tsk->sighand is still the same
		 * after we take the lock, we can race with de_thread() or
		 * __exit_signal(). In the latter case the next iteration
		 * must see ->sighand == NULL.
		 */
1262 1263 1264
		spin_lock(&sighand->siglock);
		if (likely(sighand == tsk->sighand)) {
			rcu_read_unlock();
1265
			break;
1266 1267 1268 1269
		}
		spin_unlock(&sighand->siglock);
		rcu_read_unlock();
		local_irq_restore(*flags);
1270 1271 1272 1273 1274
	}

	return sighand;
}

1275 1276 1277
/*
 * send signal info to all the members of a group
 */
L
Linus Torvalds 已提交
1278 1279
int group_send_sig_info(int sig, struct siginfo *info, struct task_struct *p)
{
1280 1281 1282 1283 1284
	int ret;

	rcu_read_lock();
	ret = check_kill_permission(sig, info, p);
	rcu_read_unlock();
1285

1286 1287
	if (!ret && sig)
		ret = do_send_sig_info(sig, info, p, true);
L
Linus Torvalds 已提交
1288 1289 1290 1291 1292

	return ret;
}

/*
1293
 * __kill_pgrp_info() sends a signal to a process group: this is what the tty
L
Linus Torvalds 已提交
1294
 * control characters do (^C, ^Z etc)
1295
 * - the caller must hold at least a readlock on tasklist_lock
L
Linus Torvalds 已提交
1296
 */
1297
int __kill_pgrp_info(int sig, struct siginfo *info, struct pid *pgrp)
L
Linus Torvalds 已提交
1298 1299 1300 1301 1302 1303
{
	struct task_struct *p = NULL;
	int retval, success;

	success = 0;
	retval = -ESRCH;
1304
	do_each_pid_task(pgrp, PIDTYPE_PGID, p) {
L
Linus Torvalds 已提交
1305 1306 1307
		int err = group_send_sig_info(sig, info, p);
		success |= !err;
		retval = err;
1308
	} while_each_pid_task(pgrp, PIDTYPE_PGID, p);
L
Linus Torvalds 已提交
1309 1310 1311
	return success ? 0 : retval;
}

1312
int kill_pid_info(int sig, struct siginfo *info, struct pid *pid)
L
Linus Torvalds 已提交
1313
{
1314
	int error = -ESRCH;
L
Linus Torvalds 已提交
1315 1316
	struct task_struct *p;

1317 1318 1319 1320 1321 1322 1323 1324
	for (;;) {
		rcu_read_lock();
		p = pid_task(pid, PIDTYPE_PID);
		if (p)
			error = group_send_sig_info(sig, info, p);
		rcu_read_unlock();
		if (likely(!p || error != -ESRCH))
			return error;
1325

1326 1327 1328 1329 1330 1331
		/*
		 * The task was unhashed in between, try again.  If it
		 * is dead, pid_task() will return NULL, if we race with
		 * de_thread() it will find the new leader.
		 */
	}
L
Linus Torvalds 已提交
1332 1333
}

1334
static int kill_proc_info(int sig, struct siginfo *info, pid_t pid)
1335 1336 1337
{
	int error;
	rcu_read_lock();
1338
	error = kill_pid_info(sig, info, find_vpid(pid));
1339 1340 1341 1342
	rcu_read_unlock();
	return error;
}

1343 1344 1345 1346
static int kill_as_cred_perm(const struct cred *cred,
			     struct task_struct *target)
{
	const struct cred *pcred = __task_cred(target);
1347 1348
	if (!uid_eq(cred->euid, pcred->suid) && !uid_eq(cred->euid, pcred->uid) &&
	    !uid_eq(cred->uid,  pcred->suid) && !uid_eq(cred->uid,  pcred->uid))
1349 1350 1351 1352
		return 0;
	return 1;
}

1353
/* like kill_pid_info(), but doesn't use uid/euid of "current" */
1354 1355
int kill_pid_info_as_cred(int sig, struct siginfo *info, struct pid *pid,
			 const struct cred *cred, u32 secid)
1356 1357 1358
{
	int ret = -EINVAL;
	struct task_struct *p;
1359
	unsigned long flags;
1360 1361 1362 1363

	if (!valid_signal(sig))
		return ret;

1364
	rcu_read_lock();
1365
	p = pid_task(pid, PIDTYPE_PID);
1366 1367 1368 1369
	if (!p) {
		ret = -ESRCH;
		goto out_unlock;
	}
1370
	if (si_fromuser(info) && !kill_as_cred_perm(cred, p)) {
1371 1372 1373
		ret = -EPERM;
		goto out_unlock;
	}
1374 1375 1376
	ret = security_task_kill(p, info, sig, secid);
	if (ret)
		goto out_unlock;
1377 1378 1379 1380 1381 1382 1383

	if (sig) {
		if (lock_task_sighand(p, &flags)) {
			ret = __send_signal(sig, info, p, 1, 0);
			unlock_task_sighand(p, &flags);
		} else
			ret = -ESRCH;
1384 1385
	}
out_unlock:
1386
	rcu_read_unlock();
1387 1388
	return ret;
}
1389
EXPORT_SYMBOL_GPL(kill_pid_info_as_cred);
L
Linus Torvalds 已提交
1390 1391 1392 1393 1394 1395 1396 1397

/*
 * kill_something_info() interprets pid in interesting ways just like kill(2).
 *
 * POSIX specifies that kill(-1,sig) is unspecified, but what we have
 * is probably wrong.  Should make it like BSD or SYSV.
 */

1398
static int kill_something_info(int sig, struct siginfo *info, pid_t pid)
L
Linus Torvalds 已提交
1399
{
1400
	int ret;
1401 1402 1403 1404 1405 1406 1407 1408

	if (pid > 0) {
		rcu_read_lock();
		ret = kill_pid_info(sig, info, find_vpid(pid));
		rcu_read_unlock();
		return ret;
	}

1409 1410 1411 1412
	/* -INT_MIN is undefined.  Exclude this case to avoid a UBSAN warning */
	if (pid == INT_MIN)
		return -ESRCH;

1413 1414 1415 1416 1417
	read_lock(&tasklist_lock);
	if (pid != -1) {
		ret = __kill_pgrp_info(sig, info,
				pid ? find_vpid(-pid) : task_pgrp(current));
	} else {
L
Linus Torvalds 已提交
1418 1419 1420 1421
		int retval = 0, count = 0;
		struct task_struct * p;

		for_each_process(p) {
1422 1423
			if (task_pid_vnr(p) > 1 &&
					!same_thread_group(p, current)) {
L
Linus Torvalds 已提交
1424 1425 1426 1427 1428 1429
				int err = group_send_sig_info(sig, info, p);
				++count;
				if (err != -EPERM)
					retval = err;
			}
		}
1430
		ret = count ? retval : -ESRCH;
L
Linus Torvalds 已提交
1431
	}
1432 1433
	read_unlock(&tasklist_lock);

1434
	return ret;
L
Linus Torvalds 已提交
1435 1436 1437 1438 1439 1440
}

/*
 * These are for backward compatibility with the rest of the kernel source.
 */

1441
int send_sig_info(int sig, struct siginfo *info, struct task_struct *p)
L
Linus Torvalds 已提交
1442 1443 1444 1445 1446
{
	/*
	 * Make sure legacy kernel users don't send in bad values
	 * (normal paths check this in check_kill_permission).
	 */
1447
	if (!valid_signal(sig))
L
Linus Torvalds 已提交
1448 1449
		return -EINVAL;

1450
	return do_send_sig_info(sig, info, p, false);
L
Linus Torvalds 已提交
1451 1452
}

1453 1454 1455
#define __si_special(priv) \
	((priv) ? SEND_SIG_PRIV : SEND_SIG_NOINFO)

L
Linus Torvalds 已提交
1456 1457 1458
int
send_sig(int sig, struct task_struct *p, int priv)
{
1459
	return send_sig_info(sig, __si_special(priv), p);
L
Linus Torvalds 已提交
1460 1461 1462 1463 1464
}

void
force_sig(int sig, struct task_struct *p)
{
1465
	force_sig_info(sig, SEND_SIG_PRIV, p);
L
Linus Torvalds 已提交
1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486
}

/*
 * When things go south during signal handling, we
 * will force a SIGSEGV. And if the signal that caused
 * the problem was already a SIGSEGV, we'll want to
 * make sure we don't even try to deliver the signal..
 */
int
force_sigsegv(int sig, struct task_struct *p)
{
	if (sig == SIGSEGV) {
		unsigned long flags;
		spin_lock_irqsave(&p->sighand->siglock, flags);
		p->sighand->action[sig - 1].sa.sa_handler = SIG_DFL;
		spin_unlock_irqrestore(&p->sighand->siglock, flags);
	}
	force_sig(SIGSEGV, p);
	return 0;
}

1487 1488
int kill_pgrp(struct pid *pid, int sig, int priv)
{
1489 1490 1491 1492 1493 1494 1495
	int ret;

	read_lock(&tasklist_lock);
	ret = __kill_pgrp_info(sig, __si_special(priv), pid);
	read_unlock(&tasklist_lock);

	return ret;
1496 1497 1498 1499 1500 1501 1502 1503 1504
}
EXPORT_SYMBOL(kill_pgrp);

int kill_pid(struct pid *pid, int sig, int priv)
{
	return kill_pid_info(sig, __si_special(priv), pid);
}
EXPORT_SYMBOL(kill_pid);

L
Linus Torvalds 已提交
1505 1506 1507 1508
/*
 * These functions support sending signals using preallocated sigqueue
 * structures.  This is needed "because realtime applications cannot
 * afford to lose notifications of asynchronous events, like timer
1509
 * expirations or I/O completions".  In the case of POSIX Timers
L
Linus Torvalds 已提交
1510 1511 1512 1513 1514 1515
 * we allocate the sigqueue structure from the timer_create.  If this
 * allocation fails we are able to report the failure to the application
 * with an EAGAIN error.
 */
struct sigqueue *sigqueue_alloc(void)
{
1516
	struct sigqueue *q = __sigqueue_alloc(-1, current, GFP_KERNEL, 0);
L
Linus Torvalds 已提交
1517

1518
	if (q)
L
Linus Torvalds 已提交
1519
		q->flags |= SIGQUEUE_PREALLOC;
1520 1521

	return q;
L
Linus Torvalds 已提交
1522 1523 1524 1525 1526
}

void sigqueue_free(struct sigqueue *q)
{
	unsigned long flags;
1527 1528
	spinlock_t *lock = &current->sighand->siglock;

L
Linus Torvalds 已提交
1529 1530
	BUG_ON(!(q->flags & SIGQUEUE_PREALLOC));
	/*
1531 1532
	 * We must hold ->siglock while testing q->list
	 * to serialize with collect_signal() or with
1533
	 * __exit_signal()->flush_sigqueue().
L
Linus Torvalds 已提交
1534
	 */
1535
	spin_lock_irqsave(lock, flags);
1536 1537 1538 1539 1540
	q->flags &= ~SIGQUEUE_PREALLOC;
	/*
	 * If it is queued it will be freed when dequeued,
	 * like the "regular" sigqueue.
	 */
1541
	if (!list_empty(&q->list))
1542
		q = NULL;
1543 1544
	spin_unlock_irqrestore(lock, flags);

1545 1546
	if (q)
		__sigqueue_free(q);
L
Linus Torvalds 已提交
1547 1548
}

1549
int send_sigqueue(struct sigqueue *q, struct task_struct *t, int group)
1550
{
1551
	int sig = q->info.si_signo;
1552
	struct sigpending *pending;
1553
	unsigned long flags;
1554
	int ret, result;
1555

1556
	BUG_ON(!(q->flags & SIGQUEUE_PREALLOC));
1557 1558 1559 1560 1561

	ret = -1;
	if (!likely(lock_task_sighand(t, &flags)))
		goto ret;

1562
	ret = 1; /* the signal is ignored */
1563
	result = TRACE_SIGNAL_IGNORED;
1564
	if (!prepare_signal(sig, t, false))
1565 1566 1567
		goto out;

	ret = 0;
1568 1569 1570 1571 1572 1573 1574
	if (unlikely(!list_empty(&q->list))) {
		/*
		 * If an SI_TIMER entry is already queue just increment
		 * the overrun count.
		 */
		BUG_ON(q->info.si_code != SI_TIMER);
		q->info.si_overrun++;
1575
		result = TRACE_SIGNAL_ALREADY_PENDING;
1576
		goto out;
1577
	}
1578
	q->info.si_overrun = 0;
1579 1580

	signalfd_notify(t, sig);
1581
	pending = group ? &t->signal->shared_pending : &t->pending;
1582 1583
	list_add_tail(&q->list, &pending->list);
	sigaddset(&pending->signal, sig);
1584
	complete_signal(sig, t, group);
1585
	result = TRACE_SIGNAL_DELIVERED;
1586
out:
1587
	trace_signal_generate(sig, &q->info, t, group, result);
1588 1589 1590
	unlock_task_sighand(t, &flags);
ret:
	return ret;
1591 1592
}

L
Linus Torvalds 已提交
1593 1594 1595
/*
 * Let a parent know about the death of a child.
 * For a stopped/continued status change, use do_notify_parent_cldstop instead.
R
Roland McGrath 已提交
1596
 *
1597 1598
 * Returns true if our parent ignored us and so we've switched to
 * self-reaping.
L
Linus Torvalds 已提交
1599
 */
1600
bool do_notify_parent(struct task_struct *tsk, int sig)
L
Linus Torvalds 已提交
1601 1602 1603 1604
{
	struct siginfo info;
	unsigned long flags;
	struct sighand_struct *psig;
1605
	bool autoreap = false;
1606
	u64 utime, stime;
L
Linus Torvalds 已提交
1607 1608 1609 1610

	BUG_ON(sig == -1);

 	/* do_notify_parent_cldstop should have been called instead.  */
M
Matthew Wilcox 已提交
1611
 	BUG_ON(task_is_stopped_or_traced(tsk));
L
Linus Torvalds 已提交
1612

T
Tejun Heo 已提交
1613
	BUG_ON(!tsk->ptrace &&
L
Linus Torvalds 已提交
1614 1615
	       (tsk->group_leader != tsk || !thread_group_empty(tsk)));

1616 1617 1618 1619 1620 1621 1622 1623 1624
	if (sig != SIGCHLD) {
		/*
		 * This is only possible if parent == real_parent.
		 * Check if it has changed security domain.
		 */
		if (tsk->parent_exec_id != tsk->parent->self_exec_id)
			sig = SIGCHLD;
	}

L
Linus Torvalds 已提交
1625 1626
	info.si_signo = sig;
	info.si_errno = 0;
1627
	/*
1628 1629
	 * We are under tasklist_lock here so our parent is tied to
	 * us and cannot change.
1630
	 *
1631 1632
	 * task_active_pid_ns will always return the same pid namespace
	 * until a task passes through release_task.
1633 1634 1635 1636 1637 1638
	 *
	 * write_lock() currently calls preempt_disable() which is the
	 * same as rcu_read_lock(), but according to Oleg, this is not
	 * correct to rely on this
	 */
	rcu_read_lock();
1639
	info.si_pid = task_pid_nr_ns(tsk, task_active_pid_ns(tsk->parent));
1640 1641
	info.si_uid = from_kuid_munged(task_cred_xxx(tsk->parent, user_ns),
				       task_uid(tsk));
1642 1643
	rcu_read_unlock();

1644 1645 1646
	task_cputime(tsk, &utime, &stime);
	info.si_utime = nsec_to_clock_t(utime + tsk->signal->utime);
	info.si_stime = nsec_to_clock_t(stime + tsk->signal->stime);
L
Linus Torvalds 已提交
1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659

	info.si_status = tsk->exit_code & 0x7f;
	if (tsk->exit_code & 0x80)
		info.si_code = CLD_DUMPED;
	else if (tsk->exit_code & 0x7f)
		info.si_code = CLD_KILLED;
	else {
		info.si_code = CLD_EXITED;
		info.si_status = tsk->exit_code >> 8;
	}

	psig = tsk->parent->sighand;
	spin_lock_irqsave(&psig->siglock, flags);
T
Tejun Heo 已提交
1660
	if (!tsk->ptrace && sig == SIGCHLD &&
L
Linus Torvalds 已提交
1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677
	    (psig->action[SIGCHLD-1].sa.sa_handler == SIG_IGN ||
	     (psig->action[SIGCHLD-1].sa.sa_flags & SA_NOCLDWAIT))) {
		/*
		 * We are exiting and our parent doesn't care.  POSIX.1
		 * defines special semantics for setting SIGCHLD to SIG_IGN
		 * or setting the SA_NOCLDWAIT flag: we should be reaped
		 * automatically and not left for our parent's wait4 call.
		 * Rather than having the parent do it as a magic kind of
		 * signal handler, we just set this to tell do_exit that we
		 * can be cleaned up without becoming a zombie.  Note that
		 * we still call __wake_up_parent in this case, because a
		 * blocked sys_wait4 might now return -ECHILD.
		 *
		 * Whether we send SIGCHLD or not for SA_NOCLDWAIT
		 * is implementation-defined: we do (if you don't want
		 * it, just use SIG_IGN instead).
		 */
1678
		autoreap = true;
L
Linus Torvalds 已提交
1679
		if (psig->action[SIGCHLD-1].sa.sa_handler == SIG_IGN)
1680
			sig = 0;
L
Linus Torvalds 已提交
1681
	}
1682
	if (valid_signal(sig) && sig)
L
Linus Torvalds 已提交
1683 1684 1685
		__group_send_sig_info(sig, &info, tsk->parent);
	__wake_up_parent(tsk, tsk->parent);
	spin_unlock_irqrestore(&psig->siglock, flags);
R
Roland McGrath 已提交
1686

1687
	return autoreap;
L
Linus Torvalds 已提交
1688 1689
}

1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704
/**
 * do_notify_parent_cldstop - notify parent of stopped/continued state change
 * @tsk: task reporting the state change
 * @for_ptracer: the notification is for ptracer
 * @why: CLD_{CONTINUED|STOPPED|TRAPPED} to report
 *
 * Notify @tsk's parent that the stopped/continued state has changed.  If
 * @for_ptracer is %false, @tsk's group leader notifies to its real parent.
 * If %true, @tsk reports to @tsk->parent which should be the ptracer.
 *
 * CONTEXT:
 * Must be called with tasklist_lock at least read locked.
 */
static void do_notify_parent_cldstop(struct task_struct *tsk,
				     bool for_ptracer, int why)
L
Linus Torvalds 已提交
1705 1706 1707
{
	struct siginfo info;
	unsigned long flags;
1708
	struct task_struct *parent;
L
Linus Torvalds 已提交
1709
	struct sighand_struct *sighand;
1710
	u64 utime, stime;
L
Linus Torvalds 已提交
1711

1712
	if (for_ptracer) {
1713
		parent = tsk->parent;
1714
	} else {
1715 1716 1717 1718
		tsk = tsk->group_leader;
		parent = tsk->real_parent;
	}

L
Linus Torvalds 已提交
1719 1720
	info.si_signo = SIGCHLD;
	info.si_errno = 0;
1721
	/*
1722
	 * see comment in do_notify_parent() about the following 4 lines
1723 1724
	 */
	rcu_read_lock();
1725
	info.si_pid = task_pid_nr_ns(tsk, task_active_pid_ns(parent));
1726
	info.si_uid = from_kuid_munged(task_cred_xxx(parent, user_ns), task_uid(tsk));
1727 1728
	rcu_read_unlock();

1729 1730 1731
	task_cputime(tsk, &utime, &stime);
	info.si_utime = nsec_to_clock_t(utime);
	info.si_stime = nsec_to_clock_t(stime);
L
Linus Torvalds 已提交
1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759

 	info.si_code = why;
 	switch (why) {
 	case CLD_CONTINUED:
 		info.si_status = SIGCONT;
 		break;
 	case CLD_STOPPED:
 		info.si_status = tsk->signal->group_exit_code & 0x7f;
 		break;
 	case CLD_TRAPPED:
 		info.si_status = tsk->exit_code & 0x7f;
 		break;
 	default:
 		BUG();
 	}

	sighand = parent->sighand;
	spin_lock_irqsave(&sighand->siglock, flags);
	if (sighand->action[SIGCHLD-1].sa.sa_handler != SIG_IGN &&
	    !(sighand->action[SIGCHLD-1].sa.sa_flags & SA_NOCLDSTOP))
		__group_send_sig_info(SIGCHLD, &info, parent);
	/*
	 * Even if SIGCHLD is not generated, we must wake up wait4 calls.
	 */
	__wake_up_parent(tsk, parent);
	spin_unlock_irqrestore(&sighand->siglock, flags);
}

1760 1761
static inline int may_ptrace_stop(void)
{
T
Tejun Heo 已提交
1762
	if (!likely(current->ptrace))
1763 1764 1765 1766 1767 1768 1769
		return 0;
	/*
	 * Are we in the middle of do_coredump?
	 * If so and our tracer is also part of the coredump stopping
	 * is a deadlock situation, and pointless because our tracer
	 * is dead so don't allow us to stop.
	 * If SIGKILL was already sent before the caller unlocked
1770
	 * ->siglock we must see ->core_state != NULL. Otherwise it
1771
	 * is safe to enter schedule().
1772 1773 1774 1775
	 *
	 * This is almost outdated, a task with the pending SIGKILL can't
	 * block in TASK_TRACED. But PTRACE_EVENT_EXIT can be reported
	 * after SIGKILL was already dequeued.
1776
	 */
1777
	if (unlikely(current->mm->core_state) &&
1778 1779 1780 1781 1782 1783
	    unlikely(current->mm == current->parent->mm))
		return 0;

	return 1;
}

R
Roland McGrath 已提交
1784
/*
1785
 * Return non-zero if there is a SIGKILL that should be waking us up.
R
Roland McGrath 已提交
1786 1787 1788 1789
 * Called with the siglock held.
 */
static int sigkill_pending(struct task_struct *tsk)
{
1790 1791
	return	sigismember(&tsk->pending.signal, SIGKILL) ||
		sigismember(&tsk->signal->shared_pending.signal, SIGKILL);
R
Roland McGrath 已提交
1792 1793
}

L
Linus Torvalds 已提交
1794 1795 1796 1797 1798 1799 1800 1801
/*
 * This must be called with current->sighand->siglock held.
 *
 * This should be the path for all ptrace stops.
 * We always set current->last_siginfo while stopped here.
 * That makes it a way to test a stopped process for
 * being ptrace-stopped vs being job-control-stopped.
 *
1802 1803
 * If we actually decide not to stop at all because the tracer
 * is gone, we keep current->exit_code unless clear_code.
L
Linus Torvalds 已提交
1804
 */
T
Tejun Heo 已提交
1805
static void ptrace_stop(int exit_code, int why, int clear_code, siginfo_t *info)
1806 1807
	__releases(&current->sighand->siglock)
	__acquires(&current->sighand->siglock)
L
Linus Torvalds 已提交
1808
{
1809 1810
	bool gstop_done = false;

R
Roland McGrath 已提交
1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825
	if (arch_ptrace_stop_needed(exit_code, info)) {
		/*
		 * The arch code has something special to do before a
		 * ptrace stop.  This is allowed to block, e.g. for faults
		 * on user stack pages.  We can't keep the siglock while
		 * calling arch_ptrace_stop, so we must release it now.
		 * To preserve proper semantics, we must do this before
		 * any signal bookkeeping like checking group_stop_count.
		 * Meanwhile, a SIGKILL could come in before we retake the
		 * siglock.  That must prevent us from sleeping in TASK_TRACED.
		 * So after regaining the lock, we must check for SIGKILL.
		 */
		spin_unlock_irq(&current->sighand->siglock);
		arch_ptrace_stop(exit_code, info);
		spin_lock_irq(&current->sighand->siglock);
1826 1827
		if (sigkill_pending(current))
			return;
R
Roland McGrath 已提交
1828 1829
	}

L
Linus Torvalds 已提交
1830
	/*
1831 1832 1833 1834 1835
	 * We're committing to trapping.  TRACED should be visible before
	 * TRAPPING is cleared; otherwise, the tracer might fail do_wait().
	 * Also, transition to TRACED and updates to ->jobctl should be
	 * atomic with respect to siglock and should be done after the arch
	 * hook as siglock is released and regrabbed across it.
L
Linus Torvalds 已提交
1836
	 */
1837
	set_current_state(TASK_TRACED);
L
Linus Torvalds 已提交
1838 1839 1840 1841

	current->last_siginfo = info;
	current->exit_code = exit_code;

1842
	/*
1843 1844
	 * If @why is CLD_STOPPED, we're trapping to participate in a group
	 * stop.  Do the bookkeeping.  Note that if SIGCONT was delievered
1845 1846 1847
	 * across siglock relocks since INTERRUPT was scheduled, PENDING
	 * could be clear now.  We act as if SIGCONT is received after
	 * TASK_TRACED is entered - ignore it.
1848
	 */
1849
	if (why == CLD_STOPPED && (current->jobctl & JOBCTL_STOP_PENDING))
1850
		gstop_done = task_participate_group_stop(current);
1851

1852
	/* any trap clears pending STOP trap, STOP trap clears NOTIFY */
1853
	task_clear_jobctl_pending(current, JOBCTL_TRAP_STOP);
1854 1855
	if (info && info->si_code >> 8 == PTRACE_EVENT_STOP)
		task_clear_jobctl_pending(current, JOBCTL_TRAP_NOTIFY);
1856

1857
	/* entering a trap, clear TRAPPING */
1858
	task_clear_jobctl_trapping(current);
1859

L
Linus Torvalds 已提交
1860 1861
	spin_unlock_irq(&current->sighand->siglock);
	read_lock(&tasklist_lock);
1862
	if (may_ptrace_stop()) {
1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873
		/*
		 * Notify parents of the stop.
		 *
		 * While ptraced, there are two parents - the ptracer and
		 * the real_parent of the group_leader.  The ptracer should
		 * know about every stop while the real parent is only
		 * interested in the completion of group stop.  The states
		 * for the two don't interact with each other.  Notify
		 * separately unless they're gonna be duplicates.
		 */
		do_notify_parent_cldstop(current, true, why);
1874
		if (gstop_done && ptrace_reparented(current))
1875 1876
			do_notify_parent_cldstop(current, false, why);

M
Miklos Szeredi 已提交
1877 1878 1879 1880 1881 1882 1883
		/*
		 * Don't want to allow preemption here, because
		 * sys_ptrace() needs this task to be inactive.
		 *
		 * XXX: implement read_unlock_no_resched().
		 */
		preempt_disable();
L
Linus Torvalds 已提交
1884
		read_unlock(&tasklist_lock);
M
Miklos Szeredi 已提交
1885
		preempt_enable_no_resched();
1886
		freezable_schedule();
L
Linus Torvalds 已提交
1887 1888 1889
	} else {
		/*
		 * By the time we got the lock, our tracer went away.
1890
		 * Don't drop the lock yet, another tracer may come.
1891 1892 1893
		 *
		 * If @gstop_done, the ptracer went away between group stop
		 * completion and here.  During detach, it would have set
1894 1895 1896
		 * JOBCTL_STOP_PENDING on us and we'll re-enter
		 * TASK_STOPPED in do_signal_stop() on return, so notifying
		 * the real parent of the group stop completion is enough.
L
Linus Torvalds 已提交
1897
		 */
1898 1899 1900
		if (gstop_done)
			do_notify_parent_cldstop(current, false, why);

1901
		/* tasklist protects us from ptrace_freeze_traced() */
1902
		__set_current_state(TASK_RUNNING);
1903 1904
		if (clear_code)
			current->exit_code = 0;
1905
		read_unlock(&tasklist_lock);
L
Linus Torvalds 已提交
1906 1907 1908 1909 1910 1911 1912 1913 1914 1915
	}

	/*
	 * We are back.  Now reacquire the siglock before touching
	 * last_siginfo, so that we are sure to have synchronized with
	 * any signal-sending on another CPU that wants to examine it.
	 */
	spin_lock_irq(&current->sighand->siglock);
	current->last_siginfo = NULL;

T
Tejun Heo 已提交
1916 1917 1918
	/* LISTENING can be set only during STOP traps, clear it */
	current->jobctl &= ~JOBCTL_LISTENING;

L
Linus Torvalds 已提交
1919 1920 1921
	/*
	 * Queued signals ignored us while we were stopped for tracing.
	 * So check for any that we should take before resuming user mode.
1922
	 * This sets TIF_SIGPENDING, but never clears it.
L
Linus Torvalds 已提交
1923
	 */
1924
	recalc_sigpending_tsk(current);
L
Linus Torvalds 已提交
1925 1926
}

T
Tejun Heo 已提交
1927
static void ptrace_do_notify(int signr, int exit_code, int why)
L
Linus Torvalds 已提交
1928 1929 1930 1931
{
	siginfo_t info;

	memset(&info, 0, sizeof info);
T
Tejun Heo 已提交
1932
	info.si_signo = signr;
L
Linus Torvalds 已提交
1933
	info.si_code = exit_code;
1934
	info.si_pid = task_pid_vnr(current);
1935
	info.si_uid = from_kuid_munged(current_user_ns(), current_uid());
L
Linus Torvalds 已提交
1936 1937

	/* Let the debugger run.  */
T
Tejun Heo 已提交
1938 1939 1940 1941 1942 1943
	ptrace_stop(exit_code, why, 1, &info);
}

void ptrace_notify(int exit_code)
{
	BUG_ON((exit_code & (0x7f | ~0xffff)) != SIGTRAP);
1944 1945
	if (unlikely(current->task_works))
		task_work_run();
T
Tejun Heo 已提交
1946

L
Linus Torvalds 已提交
1947
	spin_lock_irq(&current->sighand->siglock);
T
Tejun Heo 已提交
1948
	ptrace_do_notify(SIGTRAP, exit_code, CLD_TRAPPED);
L
Linus Torvalds 已提交
1949 1950 1951
	spin_unlock_irq(&current->sighand->siglock);
}

1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972
/**
 * do_signal_stop - handle group stop for SIGSTOP and other stop signals
 * @signr: signr causing group stop if initiating
 *
 * If %JOBCTL_STOP_PENDING is not set yet, initiate group stop with @signr
 * and participate in it.  If already set, participate in the existing
 * group stop.  If participated in a group stop (and thus slept), %true is
 * returned with siglock released.
 *
 * If ptraced, this function doesn't handle stop itself.  Instead,
 * %JOBCTL_TRAP_STOP is scheduled and %false is returned with siglock
 * untouched.  The caller must ensure that INTERRUPT trap handling takes
 * places afterwards.
 *
 * CONTEXT:
 * Must be called with @current->sighand->siglock held, which is released
 * on %true return.
 *
 * RETURNS:
 * %false if group stop is already cancelled or ptrace trap is scheduled.
 * %true if participated in group stop.
L
Linus Torvalds 已提交
1973
 */
1974 1975
static bool do_signal_stop(int signr)
	__releases(&current->sighand->siglock)
L
Linus Torvalds 已提交
1976 1977 1978
{
	struct signal_struct *sig = current->signal;

1979
	if (!(current->jobctl & JOBCTL_STOP_PENDING)) {
1980
		unsigned long gstop = JOBCTL_STOP_PENDING | JOBCTL_STOP_CONSUME;
1981 1982
		struct task_struct *t;

1983 1984
		/* signr will be recorded in task->jobctl for retries */
		WARN_ON_ONCE(signr & ~JOBCTL_STOP_SIGMASK);
1985

1986
		if (!likely(current->jobctl & JOBCTL_STOP_DEQUEUED) ||
1987
		    unlikely(signal_group_exit(sig)))
1988
			return false;
L
Linus Torvalds 已提交
1989
		/*
1990 1991 1992 1993 1994 1995 1996
		 * There is no group stop already in progress.  We must
		 * initiate one now.
		 *
		 * While ptraced, a task may be resumed while group stop is
		 * still in effect and then receive a stop signal and
		 * initiate another group stop.  This deviates from the
		 * usual behavior as two consecutive stop signals can't
1997 1998
		 * cause two group stops when !ptraced.  That is why we
		 * also check !task_is_stopped(t) below.
1999 2000 2001 2002 2003 2004 2005 2006
		 *
		 * The condition can be distinguished by testing whether
		 * SIGNAL_STOP_STOPPED is already set.  Don't generate
		 * group_exit_code in such case.
		 *
		 * This is not necessary for SIGNAL_STOP_CONTINUED because
		 * an intervening stop signal is required to cause two
		 * continued events regardless of ptrace.
L
Linus Torvalds 已提交
2007
		 */
2008 2009
		if (!(sig->flags & SIGNAL_STOP_STOPPED))
			sig->group_exit_code = signr;
L
Linus Torvalds 已提交
2010

2011 2012 2013 2014
		sig->group_stop_count = 0;

		if (task_set_jobctl_pending(current, signr | gstop))
			sig->group_stop_count++;
L
Linus Torvalds 已提交
2015

2016 2017
		t = current;
		while_each_thread(current, t) {
L
Linus Torvalds 已提交
2018
			/*
2019 2020 2021
			 * Setting state to TASK_STOPPED for a group
			 * stop is always done with the siglock held,
			 * so this check has no races.
L
Linus Torvalds 已提交
2022
			 */
2023 2024
			if (!task_is_stopped(t) &&
			    task_set_jobctl_pending(t, signr | gstop)) {
2025
				sig->group_stop_count++;
2026 2027 2028 2029
				if (likely(!(t->ptrace & PT_SEIZED)))
					signal_wake_up(t, 0);
				else
					ptrace_trap_notify(t);
2030
			}
2031
		}
L
Linus Torvalds 已提交
2032
	}
2033

T
Tejun Heo 已提交
2034
	if (likely(!current->ptrace)) {
2035
		int notify = 0;
L
Linus Torvalds 已提交
2036

2037 2038 2039 2040 2041 2042 2043 2044
		/*
		 * If there are no other threads in the group, or if there
		 * is a group stop in progress and we are the last to stop,
		 * report to the parent.
		 */
		if (task_participate_group_stop(current))
			notify = CLD_STOPPED;

2045
		__set_current_state(TASK_STOPPED);
2046 2047
		spin_unlock_irq(&current->sighand->siglock);

2048 2049 2050 2051 2052 2053 2054 2055 2056
		/*
		 * Notify the parent of the group stop completion.  Because
		 * we're not holding either the siglock or tasklist_lock
		 * here, ptracer may attach inbetween; however, this is for
		 * group stop and should always be delivered to the real
		 * parent of the group leader.  The new ptracer will get
		 * its notification when this task transitions into
		 * TASK_TRACED.
		 */
2057 2058
		if (notify) {
			read_lock(&tasklist_lock);
2059
			do_notify_parent_cldstop(current, false, notify);
2060 2061 2062 2063
			read_unlock(&tasklist_lock);
		}

		/* Now we don't run again until woken by SIGCONT or SIGKILL */
2064
		freezable_schedule();
2065
		return true;
2066
	} else {
2067 2068 2069 2070 2071 2072
		/*
		 * While ptraced, group stop is handled by STOP trap.
		 * Schedule it and let the caller deal with it.
		 */
		task_set_jobctl_pending(current, JOBCTL_TRAP_STOP);
		return false;
2073
	}
2074
}
L
Linus Torvalds 已提交
2075

2076 2077 2078
/**
 * do_jobctl_trap - take care of ptrace jobctl traps
 *
T
Tejun Heo 已提交
2079 2080 2081 2082 2083 2084 2085
 * When PT_SEIZED, it's used for both group stop and explicit
 * SEIZE/INTERRUPT traps.  Both generate PTRACE_EVENT_STOP trap with
 * accompanying siginfo.  If stopped, lower eight bits of exit_code contain
 * the stop signal; otherwise, %SIGTRAP.
 *
 * When !PT_SEIZED, it's used only for group stop trap with stop signal
 * number as exit_code and no siginfo.
2086 2087 2088 2089 2090 2091 2092
 *
 * CONTEXT:
 * Must be called with @current->sighand->siglock held, which may be
 * released and re-acquired before returning with intervening sleep.
 */
static void do_jobctl_trap(void)
{
T
Tejun Heo 已提交
2093
	struct signal_struct *signal = current->signal;
2094
	int signr = current->jobctl & JOBCTL_STOP_SIGMASK;
2095

T
Tejun Heo 已提交
2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106
	if (current->ptrace & PT_SEIZED) {
		if (!signal->group_stop_count &&
		    !(signal->flags & SIGNAL_STOP_STOPPED))
			signr = SIGTRAP;
		WARN_ON_ONCE(!signr);
		ptrace_do_notify(signr, signr | (PTRACE_EVENT_STOP << 8),
				 CLD_STOPPED);
	} else {
		WARN_ON_ONCE(!signr);
		ptrace_stop(signr, CLD_STOPPED, 0, NULL);
		current->exit_code = 0;
2107
	}
L
Linus Torvalds 已提交
2108 2109
}

2110
static int ptrace_signal(int signr, siginfo_t *info)
R
Roland McGrath 已提交
2111
{
2112 2113 2114 2115 2116 2117 2118 2119 2120 2121
	/*
	 * We do not check sig_kernel_stop(signr) but set this marker
	 * unconditionally because we do not know whether debugger will
	 * change signr. This flag has no meaning unless we are going
	 * to stop after return from ptrace_stop(). In this case it will
	 * be checked in do_signal_stop(), we should only stop if it was
	 * not cleared by SIGCONT while we were sleeping. See also the
	 * comment in dequeue_signal().
	 */
	current->jobctl |= JOBCTL_STOP_DEQUEUED;
T
Tejun Heo 已提交
2122
	ptrace_stop(signr, CLD_TRAPPED, 0, info);
R
Roland McGrath 已提交
2123 2124 2125 2126 2127 2128 2129 2130

	/* We're back.  Did the debugger cancel the sig?  */
	signr = current->exit_code;
	if (signr == 0)
		return signr;

	current->exit_code = 0;

2131 2132 2133 2134 2135 2136
	/*
	 * Update the siginfo structure if the signal has
	 * changed.  If the debugger wanted something
	 * specific in the siginfo structure then it should
	 * have updated *info via PTRACE_SETSIGINFO.
	 */
R
Roland McGrath 已提交
2137 2138 2139 2140
	if (signr != info->si_signo) {
		info->si_signo = signr;
		info->si_errno = 0;
		info->si_code = SI_USER;
2141
		rcu_read_lock();
R
Roland McGrath 已提交
2142
		info->si_pid = task_pid_vnr(current->parent);
2143 2144
		info->si_uid = from_kuid_munged(current_user_ns(),
						task_uid(current->parent));
2145
		rcu_read_unlock();
R
Roland McGrath 已提交
2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156
	}

	/* If the (new) signal is now blocked, requeue it.  */
	if (sigismember(&current->blocked, signr)) {
		specific_send_sig_info(signr, info, current);
		signr = 0;
	}

	return signr;
}

2157
int get_signal(struct ksignal *ksig)
L
Linus Torvalds 已提交
2158
{
2159 2160 2161
	struct sighand_struct *sighand = current->sighand;
	struct signal_struct *signal = current->signal;
	int signr;
L
Linus Torvalds 已提交
2162

2163 2164
	if (unlikely(current->task_works))
		task_work_run();
2165

2166 2167 2168
	if (unlikely(uprobe_deny_signal()))
		return 0;

R
Roland McGrath 已提交
2169
	/*
2170 2171 2172
	 * Do this once, we can't return to user-mode if freezing() == T.
	 * do_signal_stop() and ptrace_stop() do freezable_schedule() and
	 * thus do not need another check after return.
R
Roland McGrath 已提交
2173
	 */
2174 2175
	try_to_freeze();

2176
relock:
2177
	spin_lock_irq(&sighand->siglock);
2178 2179 2180 2181 2182
	/*
	 * Every stopped thread goes here after wakeup. Check to see if
	 * we should notify the parent, prepare_signal(SIGCONT) encodes
	 * the CLD_ si_code into SIGNAL_CLD_MASK bits.
	 */
2183
	if (unlikely(signal->flags & SIGNAL_CLD_MASK)) {
2184 2185 2186 2187 2188 2189 2190
		int why;

		if (signal->flags & SIGNAL_CLD_CONTINUED)
			why = CLD_CONTINUED;
		else
			why = CLD_STOPPED;

2191
		signal->flags &= ~SIGNAL_CLD_MASK;
2192

2193
		spin_unlock_irq(&sighand->siglock);
R
Roland McGrath 已提交
2194

2195 2196 2197 2198 2199 2200 2201 2202
		/*
		 * Notify the parent that we're continuing.  This event is
		 * always per-process and doesn't make whole lot of sense
		 * for ptracers, who shouldn't consume the state via
		 * wait(2) either, but, for backward compatibility, notify
		 * the ptracer of the group leader too unless it's gonna be
		 * a duplicate.
		 */
T
Tejun Heo 已提交
2203
		read_lock(&tasklist_lock);
2204 2205
		do_notify_parent_cldstop(current, false, why);

2206 2207 2208
		if (ptrace_reparented(current->group_leader))
			do_notify_parent_cldstop(current->group_leader,
						true, why);
T
Tejun Heo 已提交
2209
		read_unlock(&tasklist_lock);
2210

2211 2212 2213
		goto relock;
	}

L
Linus Torvalds 已提交
2214 2215
	for (;;) {
		struct k_sigaction *ka;
2216

T
Tejun Heo 已提交
2217 2218
		if (unlikely(current->jobctl & JOBCTL_STOP_PENDING) &&
		    do_signal_stop(0))
2219
			goto relock;
2220

2221 2222 2223 2224 2225
		if (unlikely(current->jobctl & JOBCTL_TRAP_MASK)) {
			do_jobctl_trap();
			spin_unlock_irq(&sighand->siglock);
			goto relock;
		}
L
Linus Torvalds 已提交
2226

2227
		signr = dequeue_signal(current, &current->blocked, &ksig->info);
2228

T
Tejun Heo 已提交
2229 2230
		if (!signr)
			break; /* will return 0 */
2231

2232
		if (unlikely(current->ptrace) && signr != SIGKILL) {
2233
			signr = ptrace_signal(signr, &ksig->info);
T
Tejun Heo 已提交
2234 2235
			if (!signr)
				continue;
L
Linus Torvalds 已提交
2236 2237
		}

T
Tejun Heo 已提交
2238 2239
		ka = &sighand->action[signr-1];

2240
		/* Trace actually delivered signals. */
2241
		trace_signal_deliver(signr, &ksig->info, ka);
2242

L
Linus Torvalds 已提交
2243 2244 2245 2246
		if (ka->sa.sa_handler == SIG_IGN) /* Do nothing.  */
			continue;
		if (ka->sa.sa_handler != SIG_DFL) {
			/* Run the handler.  */
2247
			ksig->ka = *ka;
L
Linus Torvalds 已提交
2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260

			if (ka->sa.sa_flags & SA_ONESHOT)
				ka->sa.sa_handler = SIG_DFL;

			break; /* will return non-zero "signr" value */
		}

		/*
		 * Now we are doing the default action for this signal.
		 */
		if (sig_kernel_ignore(signr)) /* Default is nothing. */
			continue;

2261
		/*
2262
		 * Global init gets no signals it doesn't want.
2263 2264 2265 2266 2267 2268 2269
		 * Container-init gets no signals it doesn't want from same
		 * container.
		 *
		 * Note that if global/container-init sees a sig_kernel_only()
		 * signal here, the signal must have been generated internally
		 * or must have come from an ancestor namespace. In either
		 * case, the signal cannot be dropped.
2270
		 */
2271
		if (unlikely(signal->flags & SIGNAL_UNKILLABLE) &&
2272
				!sig_kernel_only(signr))
L
Linus Torvalds 已提交
2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286
			continue;

		if (sig_kernel_stop(signr)) {
			/*
			 * The default action is to stop all threads in
			 * the thread group.  The job control signals
			 * do nothing in an orphaned pgrp, but SIGSTOP
			 * always works.  Note that siglock needs to be
			 * dropped during the call to is_orphaned_pgrp()
			 * because of lock ordering with tasklist_lock.
			 * This allows an intervening SIGCONT to be posted.
			 * We need to check for that and bail out if necessary.
			 */
			if (signr != SIGSTOP) {
2287
				spin_unlock_irq(&sighand->siglock);
L
Linus Torvalds 已提交
2288 2289 2290

				/* signals can be posted during this window */

2291
				if (is_current_pgrp_orphaned())
L
Linus Torvalds 已提交
2292 2293
					goto relock;

2294
				spin_lock_irq(&sighand->siglock);
L
Linus Torvalds 已提交
2295 2296
			}

2297
			if (likely(do_signal_stop(ksig->info.si_signo))) {
L
Linus Torvalds 已提交
2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308
				/* It released the siglock.  */
				goto relock;
			}

			/*
			 * We didn't actually stop, due to a race
			 * with SIGCONT or something like that.
			 */
			continue;
		}

2309
		spin_unlock_irq(&sighand->siglock);
L
Linus Torvalds 已提交
2310 2311 2312 2313 2314

		/*
		 * Anything else is fatal, maybe with a core dump.
		 */
		current->flags |= PF_SIGNALED;
2315

L
Linus Torvalds 已提交
2316
		if (sig_kernel_coredump(signr)) {
2317
			if (print_fatal_signals)
2318
				print_fatal_signal(ksig->info.si_signo);
2319
			proc_coredump_connector(current);
L
Linus Torvalds 已提交
2320 2321 2322 2323 2324 2325 2326 2327
			/*
			 * If it was able to dump core, this kills all
			 * other threads in the group and synchronizes with
			 * their demise.  If we lost the race with another
			 * thread getting here, it set group_exit_code
			 * first and our do_group_exit call below will use
			 * that value and ignore the one we pass it.
			 */
2328
			do_coredump(&ksig->info);
L
Linus Torvalds 已提交
2329 2330 2331 2332 2333
		}

		/*
		 * Death signals, no core dump.
		 */
2334
		do_group_exit(ksig->info.si_signo);
L
Linus Torvalds 已提交
2335 2336
		/* NOTREACHED */
	}
2337
	spin_unlock_irq(&sighand->siglock);
2338 2339 2340

	ksig->sig = signr;
	return ksig->sig > 0;
L
Linus Torvalds 已提交
2341 2342
}

2343
/**
A
Al Viro 已提交
2344
 * signal_delivered - 
R
Richard Weinberger 已提交
2345
 * @ksig:		kernel signal struct
A
Al Viro 已提交
2346
 * @stepping:		nonzero if debugger single-step or block-step in use
2347
 *
2348
 * This function should be called when a signal has successfully been
R
Richard Weinberger 已提交
2349
 * delivered. It updates the blocked signals accordingly (@ksig->ka.sa.sa_mask
A
Al Viro 已提交
2350
 * is always blocked, and the signal itself is blocked unless %SA_NODEFER
R
Richard Weinberger 已提交
2351
 * is set in @ksig->ka.sa.sa_flags.  Tracing is notified.
2352
 */
R
Richard Weinberger 已提交
2353
static void signal_delivered(struct ksignal *ksig, int stepping)
2354 2355 2356
{
	sigset_t blocked;

2357 2358 2359 2360 2361 2362
	/* A signal was successfully delivered, and the
	   saved sigmask was stored on the signal frame,
	   and will be restored by sigreturn.  So we can
	   simply clear the restore sigmask flag.  */
	clear_restore_sigmask();

R
Richard Weinberger 已提交
2363 2364 2365
	sigorsets(&blocked, &current->blocked, &ksig->ka.sa.sa_mask);
	if (!(ksig->ka.sa.sa_flags & SA_NODEFER))
		sigaddset(&blocked, ksig->sig);
2366
	set_current_blocked(&blocked);
2367
	tracehook_signal_handler(stepping);
2368 2369
}

A
Al Viro 已提交
2370 2371 2372 2373 2374
void signal_setup_done(int failed, struct ksignal *ksig, int stepping)
{
	if (failed)
		force_sigsegv(ksig->sig, current);
	else
R
Richard Weinberger 已提交
2375
		signal_delivered(ksig, stepping);
A
Al Viro 已提交
2376 2377
}

2378 2379
/*
 * It could be that complete_signal() picked us to notify about the
2380 2381
 * group-wide signal. Other threads should be notified now to take
 * the shared signals in @which since we will not.
2382
 */
2383
static void retarget_shared_pending(struct task_struct *tsk, sigset_t *which)
2384
{
2385
	sigset_t retarget;
2386 2387
	struct task_struct *t;

2388 2389 2390 2391
	sigandsets(&retarget, &tsk->signal->shared_pending.signal, which);
	if (sigisemptyset(&retarget))
		return;

2392 2393
	t = tsk;
	while_each_thread(tsk, t) {
2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406
		if (t->flags & PF_EXITING)
			continue;

		if (!has_pending_signals(&retarget, &t->blocked))
			continue;
		/* Remove the signals this thread can handle. */
		sigandsets(&retarget, &retarget, &t->blocked);

		if (!signal_pending(t))
			signal_wake_up(t, 0);

		if (sigisemptyset(&retarget))
			break;
2407 2408 2409
	}
}

O
Oleg Nesterov 已提交
2410 2411 2412
void exit_signals(struct task_struct *tsk)
{
	int group_stop = 0;
2413
	sigset_t unblocked;
O
Oleg Nesterov 已提交
2414

2415 2416 2417 2418
	/*
	 * @tsk is about to have PF_EXITING set - lock out users which
	 * expect stable threadgroup.
	 */
2419
	cgroup_threadgroup_change_begin(tsk);
2420

2421 2422
	if (thread_group_empty(tsk) || signal_group_exit(tsk->signal)) {
		tsk->flags |= PF_EXITING;
2423
		cgroup_threadgroup_change_end(tsk);
2424
		return;
O
Oleg Nesterov 已提交
2425 2426
	}

2427
	spin_lock_irq(&tsk->sighand->siglock);
O
Oleg Nesterov 已提交
2428 2429 2430 2431 2432
	/*
	 * From now this task is not visible for group-wide signals,
	 * see wants_signal(), do_signal_stop().
	 */
	tsk->flags |= PF_EXITING;
2433

2434
	cgroup_threadgroup_change_end(tsk);
2435

2436 2437 2438
	if (!signal_pending(tsk))
		goto out;

2439 2440 2441
	unblocked = tsk->blocked;
	signotset(&unblocked);
	retarget_shared_pending(tsk, &unblocked);
2442

2443
	if (unlikely(tsk->jobctl & JOBCTL_STOP_PENDING) &&
2444
	    task_participate_group_stop(tsk))
T
Tejun Heo 已提交
2445
		group_stop = CLD_STOPPED;
2446
out:
O
Oleg Nesterov 已提交
2447 2448
	spin_unlock_irq(&tsk->sighand->siglock);

2449 2450 2451 2452
	/*
	 * If group stop has completed, deliver the notification.  This
	 * should always go to the real parent of the group leader.
	 */
2453
	if (unlikely(group_stop)) {
O
Oleg Nesterov 已提交
2454
		read_lock(&tasklist_lock);
2455
		do_notify_parent_cldstop(tsk, false, group_stop);
O
Oleg Nesterov 已提交
2456 2457 2458 2459
		read_unlock(&tasklist_lock);
	}
}

L
Linus Torvalds 已提交
2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471
EXPORT_SYMBOL(recalc_sigpending);
EXPORT_SYMBOL_GPL(dequeue_signal);
EXPORT_SYMBOL(flush_signals);
EXPORT_SYMBOL(force_sig);
EXPORT_SYMBOL(send_sig);
EXPORT_SYMBOL(send_sig_info);
EXPORT_SYMBOL(sigprocmask);

/*
 * System call entry points.
 */

2472 2473 2474
/**
 *  sys_restart_syscall - restart a system call
 */
2475
SYSCALL_DEFINE0(restart_syscall)
L
Linus Torvalds 已提交
2476
{
2477
	struct restart_block *restart = &current->restart_block;
L
Linus Torvalds 已提交
2478 2479 2480 2481 2482 2483 2484 2485
	return restart->fn(restart);
}

long do_no_restart_syscall(struct restart_block *param)
{
	return -EINTR;
}

2486 2487 2488 2489 2490
static void __set_task_blocked(struct task_struct *tsk, const sigset_t *newset)
{
	if (signal_pending(tsk) && !thread_group_empty(tsk)) {
		sigset_t newblocked;
		/* A set of now blocked but previously unblocked signals. */
2491
		sigandnsets(&newblocked, newset, &current->blocked);
2492 2493 2494 2495 2496 2497
		retarget_shared_pending(tsk, &newblocked);
	}
	tsk->blocked = *newset;
	recalc_sigpending();
}

2498 2499 2500 2501 2502 2503
/**
 * set_current_blocked - change current->blocked mask
 * @newset: new mask
 *
 * It is wrong to change ->blocked directly, this helper should be used
 * to ensure the process can't miss a shared signal we are going to block.
L
Linus Torvalds 已提交
2504
 */
2505 2506 2507
void set_current_blocked(sigset_t *newset)
{
	sigdelsetmask(newset, sigmask(SIGKILL) | sigmask(SIGSTOP));
2508
	__set_current_blocked(newset);
2509 2510 2511
}

void __set_current_blocked(const sigset_t *newset)
2512 2513 2514
{
	struct task_struct *tsk = current;

2515 2516 2517 2518 2519 2520 2521
	/*
	 * In case the signal mask hasn't changed, there is nothing we need
	 * to do. The current->blocked shouldn't be modified by other task.
	 */
	if (sigequalsets(&tsk->blocked, newset))
		return;

2522
	spin_lock_irq(&tsk->sighand->siglock);
2523
	__set_task_blocked(tsk, newset);
2524 2525
	spin_unlock_irq(&tsk->sighand->siglock);
}
L
Linus Torvalds 已提交
2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536

/*
 * This is also useful for kernel threads that want to temporarily
 * (or permanently) block certain signals.
 *
 * NOTE! Unlike the user-mode sys_sigprocmask(), the kernel
 * interface happily blocks "unblockable" signals like SIGKILL
 * and friends.
 */
int sigprocmask(int how, sigset_t *set, sigset_t *oldset)
{
2537 2538
	struct task_struct *tsk = current;
	sigset_t newset;
L
Linus Torvalds 已提交
2539

2540
	/* Lockless, only current can change ->blocked, never from irq */
2541
	if (oldset)
2542
		*oldset = tsk->blocked;
2543

L
Linus Torvalds 已提交
2544 2545
	switch (how) {
	case SIG_BLOCK:
2546
		sigorsets(&newset, &tsk->blocked, set);
L
Linus Torvalds 已提交
2547 2548
		break;
	case SIG_UNBLOCK:
2549
		sigandnsets(&newset, &tsk->blocked, set);
L
Linus Torvalds 已提交
2550 2551
		break;
	case SIG_SETMASK:
2552
		newset = *set;
L
Linus Torvalds 已提交
2553 2554
		break;
	default:
2555
		return -EINVAL;
L
Linus Torvalds 已提交
2556
	}
2557

2558
	__set_current_blocked(&newset);
2559
	return 0;
L
Linus Torvalds 已提交
2560 2561
}

2562 2563 2564
/**
 *  sys_rt_sigprocmask - change the list of currently blocked signals
 *  @how: whether to add, remove, or set signals
R
Randy Dunlap 已提交
2565
 *  @nset: stores pending signals
2566 2567 2568
 *  @oset: previous value of signal mask if non-null
 *  @sigsetsize: size of sigset_t type
 */
2569
SYSCALL_DEFINE4(rt_sigprocmask, int, how, sigset_t __user *, nset,
2570
		sigset_t __user *, oset, size_t, sigsetsize)
L
Linus Torvalds 已提交
2571 2572
{
	sigset_t old_set, new_set;
2573
	int error;
L
Linus Torvalds 已提交
2574 2575 2576

	/* XXX: Don't preclude handling different sized sigset_t's.  */
	if (sigsetsize != sizeof(sigset_t))
2577
		return -EINVAL;
L
Linus Torvalds 已提交
2578

2579 2580 2581 2582 2583
	old_set = current->blocked;

	if (nset) {
		if (copy_from_user(&new_set, nset, sizeof(sigset_t)))
			return -EFAULT;
L
Linus Torvalds 已提交
2584 2585
		sigdelsetmask(&new_set, sigmask(SIGKILL)|sigmask(SIGSTOP));

2586
		error = sigprocmask(how, &new_set, NULL);
L
Linus Torvalds 已提交
2587
		if (error)
2588 2589
			return error;
	}
L
Linus Torvalds 已提交
2590

2591 2592 2593
	if (oset) {
		if (copy_to_user(oset, &old_set, sizeof(sigset_t)))
			return -EFAULT;
L
Linus Torvalds 已提交
2594
	}
2595 2596

	return 0;
L
Linus Torvalds 已提交
2597 2598
}

A
Al Viro 已提交
2599 2600 2601
#ifdef CONFIG_COMPAT
COMPAT_SYSCALL_DEFINE4(rt_sigprocmask, int, how, compat_sigset_t __user *, nset,
		compat_sigset_t __user *, oset, compat_size_t, sigsetsize)
L
Linus Torvalds 已提交
2602
{
A
Al Viro 已提交
2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626
#ifdef __BIG_ENDIAN
	sigset_t old_set = current->blocked;

	/* XXX: Don't preclude handling different sized sigset_t's.  */
	if (sigsetsize != sizeof(sigset_t))
		return -EINVAL;

	if (nset) {
		compat_sigset_t new32;
		sigset_t new_set;
		int error;
		if (copy_from_user(&new32, nset, sizeof(compat_sigset_t)))
			return -EFAULT;

		sigset_from_compat(&new_set, &new32);
		sigdelsetmask(&new_set, sigmask(SIGKILL)|sigmask(SIGSTOP));

		error = sigprocmask(how, &new_set, NULL);
		if (error)
			return error;
	}
	if (oset) {
		compat_sigset_t old32;
		sigset_to_compat(&old32, &old_set);
A
Al Viro 已提交
2627
		if (copy_to_user(oset, &old32, sizeof(compat_sigset_t)))
A
Al Viro 已提交
2628 2629 2630 2631 2632 2633 2634 2635 2636
			return -EFAULT;
	}
	return 0;
#else
	return sys_rt_sigprocmask(how, (sigset_t __user *)nset,
				  (sigset_t __user *)oset, sigsetsize);
#endif
}
#endif
L
Linus Torvalds 已提交
2637

A
Al Viro 已提交
2638
static int do_sigpending(void *set, unsigned long sigsetsize)
L
Linus Torvalds 已提交
2639 2640
{
	if (sigsetsize > sizeof(sigset_t))
A
Al Viro 已提交
2641
		return -EINVAL;
L
Linus Torvalds 已提交
2642 2643

	spin_lock_irq(&current->sighand->siglock);
A
Al Viro 已提交
2644
	sigorsets(set, &current->pending.signal,
L
Linus Torvalds 已提交
2645 2646 2647 2648
		  &current->signal->shared_pending.signal);
	spin_unlock_irq(&current->sighand->siglock);

	/* Outside the lock because only this thread touches it.  */
A
Al Viro 已提交
2649 2650
	sigandsets(set, &current->blocked, set);
	return 0;
2651
}
L
Linus Torvalds 已提交
2652

2653 2654 2655
/**
 *  sys_rt_sigpending - examine a pending signal that has been raised
 *			while blocked
2656
 *  @uset: stores pending signals
2657 2658
 *  @sigsetsize: size of sigset_t type or larger
 */
A
Al Viro 已提交
2659
SYSCALL_DEFINE2(rt_sigpending, sigset_t __user *, uset, size_t, sigsetsize)
L
Linus Torvalds 已提交
2660
{
A
Al Viro 已提交
2661 2662 2663 2664 2665 2666 2667 2668 2669 2670
	sigset_t set;
	int err = do_sigpending(&set, sigsetsize);
	if (!err && copy_to_user(uset, &set, sigsetsize))
		err = -EFAULT;
	return err;
}

#ifdef CONFIG_COMPAT
COMPAT_SYSCALL_DEFINE2(rt_sigpending, compat_sigset_t __user *, uset,
		compat_size_t, sigsetsize)
L
Linus Torvalds 已提交
2671
{
A
Al Viro 已提交
2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685
#ifdef __BIG_ENDIAN
	sigset_t set;
	int err = do_sigpending(&set, sigsetsize);
	if (!err) {
		compat_sigset_t set32;
		sigset_to_compat(&set32, &set);
		/* we can get here only if sigsetsize <= sizeof(set) */
		if (copy_to_user(uset, &set32, sigsetsize))
			err = -EFAULT;
	}
	return err;
#else
	return sys_rt_sigpending((sigset_t __user *)uset, sigsetsize);
#endif
L
Linus Torvalds 已提交
2686
}
A
Al Viro 已提交
2687
#endif
L
Linus Torvalds 已提交
2688

2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700
enum siginfo_layout siginfo_layout(int sig, int si_code)
{
	enum siginfo_layout layout = SIL_KILL;
	if ((si_code > SI_USER) && (si_code < SI_KERNEL)) {
		static const struct {
			unsigned char limit, layout;
		} filter[] = {
			[SIGILL]  = { NSIGILL,  SIL_FAULT },
			[SIGFPE]  = { NSIGFPE,  SIL_FAULT },
			[SIGSEGV] = { NSIGSEGV, SIL_FAULT },
			[SIGBUS]  = { NSIGBUS,  SIL_FAULT },
			[SIGTRAP] = { NSIGTRAP, SIL_FAULT },
2701
#if defined(SIGEMT) && defined(NSIGEMT)
2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733
			[SIGEMT]  = { NSIGEMT,  SIL_FAULT },
#endif
			[SIGCHLD] = { NSIGCHLD, SIL_CHLD },
			[SIGPOLL] = { NSIGPOLL, SIL_POLL },
#ifdef __ARCH_SIGSYS
			[SIGSYS]  = { NSIGSYS,  SIL_SYS },
#endif
		};
		if ((sig < ARRAY_SIZE(filter)) && (si_code <= filter[sig].limit))
			layout = filter[sig].layout;
		else if (si_code <= NSIGPOLL)
			layout = SIL_POLL;
	} else {
		if (si_code == SI_TIMER)
			layout = SIL_TIMER;
		else if (si_code == SI_SIGIO)
			layout = SIL_POLL;
		else if (si_code < 0)
			layout = SIL_RT;
		/* Tests to support buggy kernel ABIs */
#ifdef TRAP_FIXME
		if ((sig == SIGTRAP) && (si_code == TRAP_FIXME))
			layout = SIL_FAULT;
#endif
#ifdef FPE_FIXME
		if ((sig == SIGFPE) && (si_code == FPE_FIXME))
			layout = SIL_FAULT;
#endif
	}
	return layout;
}

L
Linus Torvalds 已提交
2734 2735
#ifndef HAVE_ARCH_COPY_SIGINFO_TO_USER

A
Al Viro 已提交
2736
int copy_siginfo_to_user(siginfo_t __user *to, const siginfo_t *from)
L
Linus Torvalds 已提交
2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747
{
	int err;

	if (!access_ok (VERIFY_WRITE, to, sizeof(siginfo_t)))
		return -EFAULT;
	if (from->si_code < 0)
		return __copy_to_user(to, from, sizeof(siginfo_t))
			? -EFAULT : 0;
	/*
	 * If you change siginfo_t structure, please be sure
	 * this code is fixed accordingly.
2748 2749
	 * Please remember to update the signalfd_copyinfo() function
	 * inside fs/signalfd.c too, in case siginfo_t changes.
L
Linus Torvalds 已提交
2750 2751 2752 2753 2754 2755
	 * It should never copy any pad contained in the structure
	 * to avoid security leaks, but must copy the generic
	 * 3 ints plus the relevant union member.
	 */
	err = __put_user(from->si_signo, &to->si_signo);
	err |= __put_user(from->si_errno, &to->si_errno);
2756 2757 2758
	err |= __put_user(from->si_code, &to->si_code);
	switch (siginfo_layout(from->si_signo, from->si_code)) {
	case SIL_KILL:
L
Linus Torvalds 已提交
2759 2760 2761
		err |= __put_user(from->si_pid, &to->si_pid);
		err |= __put_user(from->si_uid, &to->si_uid);
		break;
2762 2763
	case SIL_TIMER:
		/* Unreached SI_TIMER is negative */
L
Linus Torvalds 已提交
2764
		break;
2765
	case SIL_POLL:
L
Linus Torvalds 已提交
2766 2767 2768
		err |= __put_user(from->si_band, &to->si_band);
		err |= __put_user(from->si_fd, &to->si_fd);
		break;
2769
	case SIL_FAULT:
L
Linus Torvalds 已提交
2770 2771 2772
		err |= __put_user(from->si_addr, &to->si_addr);
#ifdef __ARCH_SI_TRAPNO
		err |= __put_user(from->si_trapno, &to->si_trapno);
A
Andi Kleen 已提交
2773 2774
#endif
#ifdef BUS_MCEERR_AO
2775
		/*
A
Andi Kleen 已提交
2776
		 * Other callers might not initialize the si_lsb field,
2777
		 * so check explicitly for the right codes here.
A
Andi Kleen 已提交
2778
		 */
2779 2780
		if (from->si_signo == SIGBUS &&
		    (from->si_code == BUS_MCEERR_AR || from->si_code == BUS_MCEERR_AO))
A
Andi Kleen 已提交
2781
			err |= __put_user(from->si_addr_lsb, &to->si_addr_lsb);
2782 2783
#endif
#ifdef SEGV_BNDERR
2784 2785 2786 2787
		if (from->si_signo == SIGSEGV && from->si_code == SEGV_BNDERR) {
			err |= __put_user(from->si_lower, &to->si_lower);
			err |= __put_user(from->si_upper, &to->si_upper);
		}
2788 2789 2790 2791
#endif
#ifdef SEGV_PKUERR
		if (from->si_signo == SIGSEGV && from->si_code == SEGV_PKUERR)
			err |= __put_user(from->si_pkey, &to->si_pkey);
L
Linus Torvalds 已提交
2792 2793
#endif
		break;
2794
	case SIL_CHLD:
L
Linus Torvalds 已提交
2795 2796 2797 2798 2799 2800
		err |= __put_user(from->si_pid, &to->si_pid);
		err |= __put_user(from->si_uid, &to->si_uid);
		err |= __put_user(from->si_status, &to->si_status);
		err |= __put_user(from->si_utime, &to->si_utime);
		err |= __put_user(from->si_stime, &to->si_stime);
		break;
2801
	case SIL_RT:
L
Linus Torvalds 已提交
2802 2803 2804 2805
		err |= __put_user(from->si_pid, &to->si_pid);
		err |= __put_user(from->si_uid, &to->si_uid);
		err |= __put_user(from->si_ptr, &to->si_ptr);
		break;
2806
#ifdef __ARCH_SIGSYS
2807
	case SIL_SYS:
2808 2809 2810 2811 2812
		err |= __put_user(from->si_call_addr, &to->si_call_addr);
		err |= __put_user(from->si_syscall, &to->si_syscall);
		err |= __put_user(from->si_arch, &to->si_arch);
		break;
#endif
L
Linus Torvalds 已提交
2813 2814 2815 2816 2817 2818
	}
	return err;
}

#endif

2819 2820 2821 2822 2823 2824
/**
 *  do_sigtimedwait - wait for queued signals specified in @which
 *  @which: queued signals to wait for
 *  @info: if non-null, the signal's siginfo is returned here
 *  @ts: upper bound on process time suspension
 */
2825
static int do_sigtimedwait(const sigset_t *which, siginfo_t *info,
2826
		    const struct timespec *ts)
2827
{
T
Thomas Gleixner 已提交
2828
	ktime_t *to = NULL, timeout = KTIME_MAX;
2829 2830
	struct task_struct *tsk = current;
	sigset_t mask = *which;
2831
	int sig, ret = 0;
2832 2833 2834 2835

	if (ts) {
		if (!timespec_valid(ts))
			return -EINVAL;
2836 2837
		timeout = timespec_to_ktime(*ts);
		to = &timeout;
2838 2839 2840 2841 2842 2843 2844 2845 2846 2847
	}

	/*
	 * Invert the set of allowed signals to get those we want to block.
	 */
	sigdelsetmask(&mask, sigmask(SIGKILL) | sigmask(SIGSTOP));
	signotset(&mask);

	spin_lock_irq(&tsk->sighand->siglock);
	sig = dequeue_signal(tsk, &mask, info);
T
Thomas Gleixner 已提交
2848
	if (!sig && timeout) {
2849 2850 2851
		/*
		 * None ready, temporarily unblock those we're interested
		 * while we are sleeping in so that we'll be awakened when
2852 2853
		 * they arrive. Unblocking is always fine, we can avoid
		 * set_current_blocked().
2854 2855 2856 2857 2858 2859
		 */
		tsk->real_blocked = tsk->blocked;
		sigandsets(&tsk->blocked, &tsk->blocked, &mask);
		recalc_sigpending();
		spin_unlock_irq(&tsk->sighand->siglock);

2860 2861 2862
		__set_current_state(TASK_INTERRUPTIBLE);
		ret = freezable_schedule_hrtimeout_range(to, tsk->timer_slack_ns,
							 HRTIMER_MODE_REL);
2863
		spin_lock_irq(&tsk->sighand->siglock);
2864
		__set_task_blocked(tsk, &tsk->real_blocked);
2865
		sigemptyset(&tsk->real_blocked);
2866
		sig = dequeue_signal(tsk, &mask, info);
2867 2868 2869 2870 2871
	}
	spin_unlock_irq(&tsk->sighand->siglock);

	if (sig)
		return sig;
2872
	return ret ? -EINTR : -EAGAIN;
2873 2874
}

2875 2876 2877 2878 2879 2880 2881 2882
/**
 *  sys_rt_sigtimedwait - synchronously wait for queued signals specified
 *			in @uthese
 *  @uthese: queued signals to wait for
 *  @uinfo: if non-null, the signal's siginfo is returned here
 *  @uts: upper bound on process time suspension
 *  @sigsetsize: size of sigset_t type
 */
2883 2884 2885
SYSCALL_DEFINE4(rt_sigtimedwait, const sigset_t __user *, uthese,
		siginfo_t __user *, uinfo, const struct timespec __user *, uts,
		size_t, sigsetsize)
L
Linus Torvalds 已提交
2886 2887 2888 2889
{
	sigset_t these;
	struct timespec ts;
	siginfo_t info;
2890
	int ret;
L
Linus Torvalds 已提交
2891 2892 2893 2894 2895 2896 2897

	/* XXX: Don't preclude handling different sized sigset_t's.  */
	if (sigsetsize != sizeof(sigset_t))
		return -EINVAL;

	if (copy_from_user(&these, uthese, sizeof(these)))
		return -EFAULT;
2898

L
Linus Torvalds 已提交
2899 2900 2901 2902 2903
	if (uts) {
		if (copy_from_user(&ts, uts, sizeof(ts)))
			return -EFAULT;
	}

2904
	ret = do_sigtimedwait(&these, &info, uts ? &ts : NULL);
L
Linus Torvalds 已提交
2905

2906 2907 2908
	if (ret > 0 && uinfo) {
		if (copy_siginfo_to_user(uinfo, &info))
			ret = -EFAULT;
L
Linus Torvalds 已提交
2909 2910 2911 2912 2913
	}

	return ret;
}

2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947
#ifdef CONFIG_COMPAT
COMPAT_SYSCALL_DEFINE4(rt_sigtimedwait, compat_sigset_t __user *, uthese,
		struct compat_siginfo __user *, uinfo,
		struct compat_timespec __user *, uts, compat_size_t, sigsetsize)
{
	compat_sigset_t s32;
	sigset_t s;
	struct timespec t;
	siginfo_t info;
	long ret;

	if (sigsetsize != sizeof(sigset_t))
		return -EINVAL;

	if (copy_from_user(&s32, uthese, sizeof(compat_sigset_t)))
		return -EFAULT;
	sigset_from_compat(&s, &s32);

	if (uts) {
		if (compat_get_timespec(&t, uts))
			return -EFAULT;
	}

	ret = do_sigtimedwait(&s, &info, uts ? &t : NULL);

	if (ret > 0 && uinfo) {
		if (copy_siginfo_to_user32(uinfo, &info))
			ret = -EFAULT;
	}

	return ret;
}
#endif

2948 2949 2950 2951 2952
/**
 *  sys_kill - send a signal to a process
 *  @pid: the PID of the process
 *  @sig: signal to be sent
 */
2953
SYSCALL_DEFINE2(kill, pid_t, pid, int, sig)
L
Linus Torvalds 已提交
2954 2955 2956 2957 2958 2959
{
	struct siginfo info;

	info.si_signo = sig;
	info.si_errno = 0;
	info.si_code = SI_USER;
2960
	info.si_pid = task_tgid_vnr(current);
2961
	info.si_uid = from_kuid_munged(current_user_ns(), current_uid());
L
Linus Torvalds 已提交
2962 2963 2964 2965

	return kill_something_info(sig, &info, pid);
}

T
Thomas Gleixner 已提交
2966 2967
static int
do_send_specific(pid_t tgid, pid_t pid, int sig, struct siginfo *info)
L
Linus Torvalds 已提交
2968 2969
{
	struct task_struct *p;
T
Thomas Gleixner 已提交
2970
	int error = -ESRCH;
L
Linus Torvalds 已提交
2971

2972
	rcu_read_lock();
2973
	p = find_task_by_vpid(pid);
2974
	if (p && (tgid <= 0 || task_tgid_vnr(p) == tgid)) {
T
Thomas Gleixner 已提交
2975
		error = check_kill_permission(sig, info, p);
L
Linus Torvalds 已提交
2976 2977 2978 2979
		/*
		 * The null signal is a permissions and process existence
		 * probe.  No signal is actually delivered.
		 */
2980 2981 2982 2983 2984 2985 2986 2987 2988
		if (!error && sig) {
			error = do_send_sig_info(sig, info, p, false);
			/*
			 * If lock_task_sighand() failed we pretend the task
			 * dies after receiving the signal. The window is tiny,
			 * and the signal is private anyway.
			 */
			if (unlikely(error == -ESRCH))
				error = 0;
L
Linus Torvalds 已提交
2989 2990
		}
	}
2991
	rcu_read_unlock();
2992

L
Linus Torvalds 已提交
2993 2994 2995
	return error;
}

T
Thomas Gleixner 已提交
2996 2997
static int do_tkill(pid_t tgid, pid_t pid, int sig)
{
2998
	struct siginfo info = {};
T
Thomas Gleixner 已提交
2999 3000 3001 3002 3003

	info.si_signo = sig;
	info.si_errno = 0;
	info.si_code = SI_TKILL;
	info.si_pid = task_tgid_vnr(current);
3004
	info.si_uid = from_kuid_munged(current_user_ns(), current_uid());
T
Thomas Gleixner 已提交
3005 3006 3007 3008

	return do_send_specific(tgid, pid, sig, &info);
}

3009 3010 3011 3012 3013 3014
/**
 *  sys_tgkill - send signal to one specific thread
 *  @tgid: the thread group ID of the thread
 *  @pid: the PID of the thread
 *  @sig: signal to be sent
 *
3015
 *  This syscall also checks the @tgid and returns -ESRCH even if the PID
3016 3017 3018
 *  exists but it's not belonging to the target process anymore. This
 *  method solves the problem of threads exiting and PIDs getting reused.
 */
3019
SYSCALL_DEFINE3(tgkill, pid_t, tgid, pid_t, pid, int, sig)
3020 3021 3022 3023 3024 3025 3026 3027
{
	/* This is only valid for single tasks */
	if (pid <= 0 || tgid <= 0)
		return -EINVAL;

	return do_tkill(tgid, pid, sig);
}

3028 3029 3030 3031 3032
/**
 *  sys_tkill - send signal to one specific task
 *  @pid: the PID of the task
 *  @sig: signal to be sent
 *
L
Linus Torvalds 已提交
3033 3034
 *  Send a signal to only one task, even if it's a CLONE_THREAD task.
 */
3035
SYSCALL_DEFINE2(tkill, pid_t, pid, int, sig)
L
Linus Torvalds 已提交
3036 3037 3038 3039 3040
{
	/* This is only valid for single tasks */
	if (pid <= 0)
		return -EINVAL;

3041
	return do_tkill(0, pid, sig);
L
Linus Torvalds 已提交
3042 3043
}

A
Al Viro 已提交
3044 3045 3046 3047 3048
static int do_rt_sigqueueinfo(pid_t pid, int sig, siginfo_t *info)
{
	/* Not even root can pretend to send signals from the kernel.
	 * Nor can they impersonate a kill()/tgkill(), which adds source info.
	 */
3049
	if ((info->si_code >= 0 || info->si_code == SI_TKILL) &&
3050
	    (task_pid_vnr(current) != pid))
A
Al Viro 已提交
3051
		return -EPERM;
3052

A
Al Viro 已提交
3053 3054 3055 3056 3057 3058
	info->si_signo = sig;

	/* POSIX.1b doesn't mention process groups.  */
	return kill_proc_info(sig, info, pid);
}

3059 3060 3061 3062 3063 3064
/**
 *  sys_rt_sigqueueinfo - send signal information to a signal
 *  @pid: the PID of the thread
 *  @sig: signal to be sent
 *  @uinfo: signal info to be sent
 */
3065 3066
SYSCALL_DEFINE3(rt_sigqueueinfo, pid_t, pid, int, sig,
		siginfo_t __user *, uinfo)
L
Linus Torvalds 已提交
3067 3068 3069 3070
{
	siginfo_t info;
	if (copy_from_user(&info, uinfo, sizeof(siginfo_t)))
		return -EFAULT;
A
Al Viro 已提交
3071 3072
	return do_rt_sigqueueinfo(pid, sig, &info);
}
L
Linus Torvalds 已提交
3073

A
Al Viro 已提交
3074 3075 3076 3077 3078 3079
#ifdef CONFIG_COMPAT
COMPAT_SYSCALL_DEFINE3(rt_sigqueueinfo,
			compat_pid_t, pid,
			int, sig,
			struct compat_siginfo __user *, uinfo)
{
3080
	siginfo_t info = {};
A
Al Viro 已提交
3081 3082 3083 3084
	int ret = copy_siginfo_from_user32(&info, uinfo);
	if (unlikely(ret))
		return ret;
	return do_rt_sigqueueinfo(pid, sig, &info);
L
Linus Torvalds 已提交
3085
}
A
Al Viro 已提交
3086
#endif
L
Linus Torvalds 已提交
3087

3088
static int do_rt_tgsigqueueinfo(pid_t tgid, pid_t pid, int sig, siginfo_t *info)
3089 3090 3091 3092 3093 3094
{
	/* This is only valid for single tasks */
	if (pid <= 0 || tgid <= 0)
		return -EINVAL;

	/* Not even root can pretend to send signals from the kernel.
3095 3096
	 * Nor can they impersonate a kill()/tgkill(), which adds source info.
	 */
3097 3098
	if ((info->si_code >= 0 || info->si_code == SI_TKILL) &&
	    (task_pid_vnr(current) != pid))
3099
		return -EPERM;
3100

3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112 3113 3114 3115 3116
	info->si_signo = sig;

	return do_send_specific(tgid, pid, sig, info);
}

SYSCALL_DEFINE4(rt_tgsigqueueinfo, pid_t, tgid, pid_t, pid, int, sig,
		siginfo_t __user *, uinfo)
{
	siginfo_t info;

	if (copy_from_user(&info, uinfo, sizeof(siginfo_t)))
		return -EFAULT;

	return do_rt_tgsigqueueinfo(tgid, pid, sig, &info);
}

3117 3118 3119 3120 3121 3122 3123
#ifdef CONFIG_COMPAT
COMPAT_SYSCALL_DEFINE4(rt_tgsigqueueinfo,
			compat_pid_t, tgid,
			compat_pid_t, pid,
			int, sig,
			struct compat_siginfo __user *, uinfo)
{
3124
	siginfo_t info = {};
3125 3126 3127 3128 3129 3130 3131

	if (copy_siginfo_from_user32(&info, uinfo))
		return -EFAULT;
	return do_rt_tgsigqueueinfo(tgid, pid, sig, &info);
}
#endif

3132
/*
3133
 * For kthreads only, must not be used if cloned with CLONE_SIGHAND
3134
 */
3135
void kernel_sigaction(int sig, __sighandler_t action)
3136
{
3137
	spin_lock_irq(&current->sighand->siglock);
3138 3139 3140
	current->sighand->action[sig - 1].sa.sa_handler = action;
	if (action == SIG_IGN) {
		sigset_t mask;
3141

3142 3143
		sigemptyset(&mask);
		sigaddset(&mask, sig);
3144

3145 3146 3147 3148
		flush_sigqueue_mask(&mask, &current->signal->shared_pending);
		flush_sigqueue_mask(&mask, &current->pending);
		recalc_sigpending();
	}
3149 3150
	spin_unlock_irq(&current->sighand->siglock);
}
3151
EXPORT_SYMBOL(kernel_sigaction);
3152

3153 3154 3155 3156 3157
void __weak sigaction_compat_abi(struct k_sigaction *act,
		struct k_sigaction *oact)
{
}

3158
int do_sigaction(int sig, struct k_sigaction *act, struct k_sigaction *oact)
L
Linus Torvalds 已提交
3159
{
3160
	struct task_struct *p = current, *t;
L
Linus Torvalds 已提交
3161
	struct k_sigaction *k;
3162
	sigset_t mask;
L
Linus Torvalds 已提交
3163

3164
	if (!valid_signal(sig) || sig < 1 || (act && sig_kernel_only(sig)))
L
Linus Torvalds 已提交
3165 3166
		return -EINVAL;

3167
	k = &p->sighand->action[sig-1];
L
Linus Torvalds 已提交
3168

3169
	spin_lock_irq(&p->sighand->siglock);
L
Linus Torvalds 已提交
3170 3171 3172
	if (oact)
		*oact = *k;

3173 3174
	sigaction_compat_abi(act, oact);

L
Linus Torvalds 已提交
3175
	if (act) {
3176 3177
		sigdelsetmask(&act->sa.sa_mask,
			      sigmask(SIGKILL) | sigmask(SIGSTOP));
3178
		*k = *act;
L
Linus Torvalds 已提交
3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189
		/*
		 * POSIX 3.3.1.3:
		 *  "Setting a signal action to SIG_IGN for a signal that is
		 *   pending shall cause the pending signal to be discarded,
		 *   whether or not it is blocked."
		 *
		 *  "Setting a signal action to SIG_DFL for a signal that is
		 *   pending and whose default action is to ignore the signal
		 *   (for example, SIGCHLD), shall cause the pending signal to
		 *   be discarded, whether or not it is blocked"
		 */
3190
		if (sig_handler_ignored(sig_handler(p, sig), sig)) {
3191 3192
			sigemptyset(&mask);
			sigaddset(&mask, sig);
3193 3194
			flush_sigqueue_mask(&mask, &p->signal->shared_pending);
			for_each_thread(p, t)
3195
				flush_sigqueue_mask(&mask, &t->pending);
L
Linus Torvalds 已提交
3196 3197 3198
		}
	}

3199
	spin_unlock_irq(&p->sighand->siglock);
L
Linus Torvalds 已提交
3200 3201 3202
	return 0;
}

3203
static int
3204
do_sigaltstack (const stack_t *ss, stack_t *oss, unsigned long sp)
L
Linus Torvalds 已提交
3205
{
3206
	struct task_struct *t = current;
L
Linus Torvalds 已提交
3207

3208 3209 3210 3211 3212 3213 3214
	if (oss) {
		memset(oss, 0, sizeof(stack_t));
		oss->ss_sp = (void __user *) t->sas_ss_sp;
		oss->ss_size = t->sas_ss_size;
		oss->ss_flags = sas_ss_flags(sp) |
			(current->sas_ss_flags & SS_FLAG_BITS);
	}
L
Linus Torvalds 已提交
3215

3216 3217 3218 3219
	if (ss) {
		void __user *ss_sp = ss->ss_sp;
		size_t ss_size = ss->ss_size;
		unsigned ss_flags = ss->ss_flags;
3220
		int ss_mode;
L
Linus Torvalds 已提交
3221

3222 3223
		if (unlikely(on_sig_stack(sp)))
			return -EPERM;
L
Linus Torvalds 已提交
3224

3225
		ss_mode = ss_flags & ~SS_FLAG_BITS;
3226 3227 3228
		if (unlikely(ss_mode != SS_DISABLE && ss_mode != SS_ONSTACK &&
				ss_mode != 0))
			return -EINVAL;
L
Linus Torvalds 已提交
3229

3230
		if (ss_mode == SS_DISABLE) {
L
Linus Torvalds 已提交
3231 3232 3233
			ss_size = 0;
			ss_sp = NULL;
		} else {
3234 3235
			if (unlikely(ss_size < MINSIGSTKSZ))
				return -ENOMEM;
L
Linus Torvalds 已提交
3236 3237
		}

3238 3239 3240
		t->sas_ss_sp = (unsigned long) ss_sp;
		t->sas_ss_size = ss_size;
		t->sas_ss_flags = ss_flags;
L
Linus Torvalds 已提交
3241
	}
3242
	return 0;
L
Linus Torvalds 已提交
3243
}
3244

3245 3246
SYSCALL_DEFINE2(sigaltstack,const stack_t __user *,uss, stack_t __user *,uoss)
{
3247 3248 3249 3250 3251 3252 3253 3254 3255
	stack_t new, old;
	int err;
	if (uss && copy_from_user(&new, uss, sizeof(stack_t)))
		return -EFAULT;
	err = do_sigaltstack(uss ? &new : NULL, uoss ? &old : NULL,
			      current_user_stack_pointer());
	if (!err && uoss && copy_to_user(uoss, &old, sizeof(stack_t)))
		err = -EFAULT;
	return err;
3256
}
L
Linus Torvalds 已提交
3257

A
Al Viro 已提交
3258 3259
int restore_altstack(const stack_t __user *uss)
{
3260 3261 3262 3263
	stack_t new;
	if (copy_from_user(&new, uss, sizeof(stack_t)))
		return -EFAULT;
	(void)do_sigaltstack(&new, NULL, current_user_stack_pointer());
A
Al Viro 已提交
3264
	/* squash all but EFAULT for now */
3265
	return 0;
A
Al Viro 已提交
3266 3267
}

3268 3269 3270
int __save_altstack(stack_t __user *uss, unsigned long sp)
{
	struct task_struct *t = current;
3271 3272
	int err = __put_user((void __user *)t->sas_ss_sp, &uss->ss_sp) |
		__put_user(t->sas_ss_flags, &uss->ss_flags) |
3273
		__put_user(t->sas_ss_size, &uss->ss_size);
3274 3275 3276 3277 3278
	if (err)
		return err;
	if (t->sas_ss_flags & SS_AUTODISARM)
		sas_ss_reset(t);
	return 0;
3279 3280
}

A
Al Viro 已提交
3281
#ifdef CONFIG_COMPAT
3282 3283 3284
COMPAT_SYSCALL_DEFINE2(sigaltstack,
			const compat_stack_t __user *, uss_ptr,
			compat_stack_t __user *, uoss_ptr)
A
Al Viro 已提交
3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296
{
	stack_t uss, uoss;
	int ret;

	if (uss_ptr) {
		compat_stack_t uss32;
		if (copy_from_user(&uss32, uss_ptr, sizeof(compat_stack_t)))
			return -EFAULT;
		uss.ss_sp = compat_ptr(uss32.ss_sp);
		uss.ss_flags = uss32.ss_flags;
		uss.ss_size = uss32.ss_size;
	}
3297
	ret = do_sigaltstack(uss_ptr ? &uss : NULL, &uoss,
A
Al Viro 已提交
3298 3299
			     compat_user_stack_pointer());
	if (ret >= 0 && uoss_ptr)  {
3300 3301 3302 3303 3304 3305
		compat_stack_t old;
		memset(&old, 0, sizeof(old));
		old.ss_sp = ptr_to_compat(uoss.ss_sp);
		old.ss_flags = uoss.ss_flags;
		old.ss_size = uoss.ss_size;
		if (copy_to_user(uoss_ptr, &old, sizeof(compat_stack_t)))
A
Al Viro 已提交
3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316
			ret = -EFAULT;
	}
	return ret;
}

int compat_restore_altstack(const compat_stack_t __user *uss)
{
	int err = compat_sys_sigaltstack(uss, NULL);
	/* squash all but -EFAULT for now */
	return err == -EFAULT ? err : 0;
}
3317 3318 3319

int __compat_save_altstack(compat_stack_t __user *uss, unsigned long sp)
{
3320
	int err;
3321
	struct task_struct *t = current;
3322 3323 3324
	err = __put_user(ptr_to_compat((void __user *)t->sas_ss_sp),
			 &uss->ss_sp) |
		__put_user(t->sas_ss_flags, &uss->ss_flags) |
3325
		__put_user(t->sas_ss_size, &uss->ss_size);
3326 3327 3328 3329 3330
	if (err)
		return err;
	if (t->sas_ss_flags & SS_AUTODISARM)
		sas_ss_reset(t);
	return 0;
3331
}
A
Al Viro 已提交
3332
#endif
L
Linus Torvalds 已提交
3333 3334 3335

#ifdef __ARCH_WANT_SYS_SIGPENDING

3336 3337 3338 3339
/**
 *  sys_sigpending - examine pending signals
 *  @set: where mask of pending signal is returned
 */
3340
SYSCALL_DEFINE1(sigpending, old_sigset_t __user *, set)
L
Linus Torvalds 已提交
3341
{
A
Al Viro 已提交
3342
	return sys_rt_sigpending((sigset_t __user *)set, sizeof(old_sigset_t)); 
L
Linus Torvalds 已提交
3343 3344
}

A
Al Viro 已提交
3345 3346 3347
#ifdef CONFIG_COMPAT
COMPAT_SYSCALL_DEFINE1(sigpending, compat_old_sigset_t __user *, set32)
{
3348
#ifdef __BIG_ENDIAN
A
Al Viro 已提交
3349
	sigset_t set;
3350 3351 3352
	int err = do_sigpending(&set, sizeof(set.sig[0]));
	if (!err)
		err = put_user(set.sig[0], set32);
A
Al Viro 已提交
3353
	return err;
3354 3355 3356
#else
	return sys_rt_sigpending((sigset_t __user *)set32, sizeof(*set32));
#endif
A
Al Viro 已提交
3357 3358 3359
}
#endif

L
Linus Torvalds 已提交
3360 3361 3362
#endif

#ifdef __ARCH_WANT_SYS_SIGPROCMASK
3363 3364 3365
/**
 *  sys_sigprocmask - examine and change blocked signals
 *  @how: whether to add, remove, or set signals
3366
 *  @nset: signals to add or remove (if non-null)
3367 3368
 *  @oset: previous value of signal mask if non-null
 *
3369 3370 3371
 * Some platforms have their own version with special arguments;
 * others support only sys_rt_sigprocmask.
 */
L
Linus Torvalds 已提交
3372

3373
SYSCALL_DEFINE3(sigprocmask, int, how, old_sigset_t __user *, nset,
3374
		old_sigset_t __user *, oset)
L
Linus Torvalds 已提交
3375 3376
{
	old_sigset_t old_set, new_set;
3377
	sigset_t new_blocked;
L
Linus Torvalds 已提交
3378

3379
	old_set = current->blocked.sig[0];
L
Linus Torvalds 已提交
3380

3381 3382 3383
	if (nset) {
		if (copy_from_user(&new_set, nset, sizeof(*nset)))
			return -EFAULT;
L
Linus Torvalds 已提交
3384

3385
		new_blocked = current->blocked;
L
Linus Torvalds 已提交
3386 3387 3388

		switch (how) {
		case SIG_BLOCK:
3389
			sigaddsetmask(&new_blocked, new_set);
L
Linus Torvalds 已提交
3390 3391
			break;
		case SIG_UNBLOCK:
3392
			sigdelsetmask(&new_blocked, new_set);
L
Linus Torvalds 已提交
3393 3394
			break;
		case SIG_SETMASK:
3395
			new_blocked.sig[0] = new_set;
L
Linus Torvalds 已提交
3396
			break;
3397 3398
		default:
			return -EINVAL;
L
Linus Torvalds 已提交
3399 3400
		}

3401
		set_current_blocked(&new_blocked);
3402 3403 3404
	}

	if (oset) {
L
Linus Torvalds 已提交
3405
		if (copy_to_user(oset, &old_set, sizeof(*oset)))
3406
			return -EFAULT;
L
Linus Torvalds 已提交
3407
	}
3408 3409

	return 0;
L
Linus Torvalds 已提交
3410 3411 3412
}
#endif /* __ARCH_WANT_SYS_SIGPROCMASK */

3413
#ifndef CONFIG_ODD_RT_SIGACTION
3414 3415 3416
/**
 *  sys_rt_sigaction - alter an action taken by a process
 *  @sig: signal to be sent
3417 3418
 *  @act: new sigaction
 *  @oact: used to save the previous sigaction
3419 3420
 *  @sigsetsize: size of sigset_t type
 */
3421 3422 3423 3424
SYSCALL_DEFINE4(rt_sigaction, int, sig,
		const struct sigaction __user *, act,
		struct sigaction __user *, oact,
		size_t, sigsetsize)
L
Linus Torvalds 已提交
3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446
{
	struct k_sigaction new_sa, old_sa;
	int ret = -EINVAL;

	/* XXX: Don't preclude handling different sized sigset_t's.  */
	if (sigsetsize != sizeof(sigset_t))
		goto out;

	if (act) {
		if (copy_from_user(&new_sa.sa, act, sizeof(new_sa.sa)))
			return -EFAULT;
	}

	ret = do_sigaction(sig, act ? &new_sa : NULL, oact ? &old_sa : NULL);

	if (!ret && oact) {
		if (copy_to_user(oact, &old_sa.sa, sizeof(old_sa.sa)))
			return -EFAULT;
	}
out:
	return ret;
}
A
Al Viro 已提交
3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472
#ifdef CONFIG_COMPAT
COMPAT_SYSCALL_DEFINE4(rt_sigaction, int, sig,
		const struct compat_sigaction __user *, act,
		struct compat_sigaction __user *, oact,
		compat_size_t, sigsetsize)
{
	struct k_sigaction new_ka, old_ka;
	compat_sigset_t mask;
#ifdef __ARCH_HAS_SA_RESTORER
	compat_uptr_t restorer;
#endif
	int ret;

	/* XXX: Don't preclude handling different sized sigset_t's.  */
	if (sigsetsize != sizeof(compat_sigset_t))
		return -EINVAL;

	if (act) {
		compat_uptr_t handler;
		ret = get_user(handler, &act->sa_handler);
		new_ka.sa.sa_handler = compat_ptr(handler);
#ifdef __ARCH_HAS_SA_RESTORER
		ret |= get_user(restorer, &act->sa_restorer);
		new_ka.sa.sa_restorer = compat_ptr(restorer);
#endif
		ret |= copy_from_user(&mask, &act->sa_mask, sizeof(mask));
3473
		ret |= get_user(new_ka.sa.sa_flags, &act->sa_flags);
A
Al Viro 已提交
3474 3475 3476 3477 3478 3479 3480 3481 3482 3483 3484
		if (ret)
			return -EFAULT;
		sigset_from_compat(&new_ka.sa.sa_mask, &mask);
	}

	ret = do_sigaction(sig, act ? &new_ka : NULL, oact ? &old_ka : NULL);
	if (!ret && oact) {
		sigset_to_compat(&mask, &old_ka.sa.sa_mask);
		ret = put_user(ptr_to_compat(old_ka.sa.sa_handler), 
			       &oact->sa_handler);
		ret |= copy_to_user(&oact->sa_mask, &mask, sizeof(mask));
3485
		ret |= put_user(old_ka.sa.sa_flags, &oact->sa_flags);
A
Al Viro 已提交
3486 3487 3488 3489 3490 3491 3492 3493
#ifdef __ARCH_HAS_SA_RESTORER
		ret |= put_user(ptr_to_compat(old_ka.sa.sa_restorer),
				&oact->sa_restorer);
#endif
	}
	return ret;
}
#endif
3494
#endif /* !CONFIG_ODD_RT_SIGACTION */
L
Linus Torvalds 已提交
3495

3496 3497 3498 3499 3500 3501 3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524 3525 3526 3527 3528 3529 3530 3531 3532 3533 3534 3535 3536 3537 3538 3539 3540 3541 3542 3543 3544 3545 3546 3547 3548 3549 3550 3551 3552 3553 3554 3555 3556 3557 3558 3559 3560 3561 3562 3563 3564 3565 3566 3567 3568 3569 3570 3571 3572
#ifdef CONFIG_OLD_SIGACTION
SYSCALL_DEFINE3(sigaction, int, sig,
		const struct old_sigaction __user *, act,
	        struct old_sigaction __user *, oact)
{
	struct k_sigaction new_ka, old_ka;
	int ret;

	if (act) {
		old_sigset_t mask;
		if (!access_ok(VERIFY_READ, act, sizeof(*act)) ||
		    __get_user(new_ka.sa.sa_handler, &act->sa_handler) ||
		    __get_user(new_ka.sa.sa_restorer, &act->sa_restorer) ||
		    __get_user(new_ka.sa.sa_flags, &act->sa_flags) ||
		    __get_user(mask, &act->sa_mask))
			return -EFAULT;
#ifdef __ARCH_HAS_KA_RESTORER
		new_ka.ka_restorer = NULL;
#endif
		siginitset(&new_ka.sa.sa_mask, mask);
	}

	ret = do_sigaction(sig, act ? &new_ka : NULL, oact ? &old_ka : NULL);

	if (!ret && oact) {
		if (!access_ok(VERIFY_WRITE, oact, sizeof(*oact)) ||
		    __put_user(old_ka.sa.sa_handler, &oact->sa_handler) ||
		    __put_user(old_ka.sa.sa_restorer, &oact->sa_restorer) ||
		    __put_user(old_ka.sa.sa_flags, &oact->sa_flags) ||
		    __put_user(old_ka.sa.sa_mask.sig[0], &oact->sa_mask))
			return -EFAULT;
	}

	return ret;
}
#endif
#ifdef CONFIG_COMPAT_OLD_SIGACTION
COMPAT_SYSCALL_DEFINE3(sigaction, int, sig,
		const struct compat_old_sigaction __user *, act,
	        struct compat_old_sigaction __user *, oact)
{
	struct k_sigaction new_ka, old_ka;
	int ret;
	compat_old_sigset_t mask;
	compat_uptr_t handler, restorer;

	if (act) {
		if (!access_ok(VERIFY_READ, act, sizeof(*act)) ||
		    __get_user(handler, &act->sa_handler) ||
		    __get_user(restorer, &act->sa_restorer) ||
		    __get_user(new_ka.sa.sa_flags, &act->sa_flags) ||
		    __get_user(mask, &act->sa_mask))
			return -EFAULT;

#ifdef __ARCH_HAS_KA_RESTORER
		new_ka.ka_restorer = NULL;
#endif
		new_ka.sa.sa_handler = compat_ptr(handler);
		new_ka.sa.sa_restorer = compat_ptr(restorer);
		siginitset(&new_ka.sa.sa_mask, mask);
	}

	ret = do_sigaction(sig, act ? &new_ka : NULL, oact ? &old_ka : NULL);

	if (!ret && oact) {
		if (!access_ok(VERIFY_WRITE, oact, sizeof(*oact)) ||
		    __put_user(ptr_to_compat(old_ka.sa.sa_handler),
			       &oact->sa_handler) ||
		    __put_user(ptr_to_compat(old_ka.sa.sa_restorer),
			       &oact->sa_restorer) ||
		    __put_user(old_ka.sa.sa_flags, &oact->sa_flags) ||
		    __put_user(old_ka.sa.sa_mask.sig[0], &oact->sa_mask))
			return -EFAULT;
	}
	return ret;
}
#endif
L
Linus Torvalds 已提交
3573

3574
#ifdef CONFIG_SGETMASK_SYSCALL
L
Linus Torvalds 已提交
3575 3576 3577 3578

/*
 * For backwards compatibility.  Functionality superseded by sigprocmask.
 */
3579
SYSCALL_DEFINE0(sgetmask)
L
Linus Torvalds 已提交
3580 3581 3582 3583 3584
{
	/* SMP safe */
	return current->blocked.sig[0];
}

3585
SYSCALL_DEFINE1(ssetmask, int, newmask)
L
Linus Torvalds 已提交
3586
{
3587 3588
	int old = current->blocked.sig[0];
	sigset_t newset;
L
Linus Torvalds 已提交
3589

3590
	siginitset(&newset, newmask);
3591
	set_current_blocked(&newset);
L
Linus Torvalds 已提交
3592 3593 3594

	return old;
}
3595
#endif /* CONFIG_SGETMASK_SYSCALL */
L
Linus Torvalds 已提交
3596 3597 3598 3599 3600

#ifdef __ARCH_WANT_SYS_SIGNAL
/*
 * For backwards compatibility.  Functionality superseded by sigaction.
 */
3601
SYSCALL_DEFINE2(signal, int, sig, __sighandler_t, handler)
L
Linus Torvalds 已提交
3602 3603 3604 3605 3606 3607
{
	struct k_sigaction new_sa, old_sa;
	int ret;

	new_sa.sa.sa_handler = handler;
	new_sa.sa.sa_flags = SA_ONESHOT | SA_NOMASK;
3608
	sigemptyset(&new_sa.sa.sa_mask);
L
Linus Torvalds 已提交
3609 3610 3611 3612 3613 3614 3615 3616 3617

	ret = do_sigaction(sig, &new_sa, &old_sa);

	return ret ? ret : (unsigned long)old_sa.sa.sa_handler;
}
#endif /* __ARCH_WANT_SYS_SIGNAL */

#ifdef __ARCH_WANT_SYS_PAUSE

3618
SYSCALL_DEFINE0(pause)
L
Linus Torvalds 已提交
3619
{
3620
	while (!signal_pending(current)) {
3621
		__set_current_state(TASK_INTERRUPTIBLE);
3622 3623
		schedule();
	}
L
Linus Torvalds 已提交
3624 3625 3626 3627 3628
	return -ERESTARTNOHAND;
}

#endif

3629
static int sigsuspend(sigset_t *set)
A
Al Viro 已提交
3630 3631 3632 3633
{
	current->saved_sigmask = current->blocked;
	set_current_blocked(set);

3634 3635 3636 3637
	while (!signal_pending(current)) {
		__set_current_state(TASK_INTERRUPTIBLE);
		schedule();
	}
A
Al Viro 已提交
3638 3639 3640 3641
	set_restore_sigmask();
	return -ERESTARTNOHAND;
}

3642 3643 3644 3645 3646 3647
/**
 *  sys_rt_sigsuspend - replace the signal mask for a value with the
 *	@unewset value until a signal is received
 *  @unewset: new signal mask value
 *  @sigsetsize: size of sigset_t type
 */
3648
SYSCALL_DEFINE2(rt_sigsuspend, sigset_t __user *, unewset, size_t, sigsetsize)
3649 3650 3651 3652 3653 3654 3655 3656 3657
{
	sigset_t newset;

	/* XXX: Don't preclude handling different sized sigset_t's.  */
	if (sigsetsize != sizeof(sigset_t))
		return -EINVAL;

	if (copy_from_user(&newset, unewset, sizeof(newset)))
		return -EFAULT;
A
Al Viro 已提交
3658
	return sigsuspend(&newset);
3659
}
A
Al Viro 已提交
3660 3661 3662 3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 3673 3674 3675 3676 3677 3678 3679 3680 3681
 
#ifdef CONFIG_COMPAT
COMPAT_SYSCALL_DEFINE2(rt_sigsuspend, compat_sigset_t __user *, unewset, compat_size_t, sigsetsize)
{
#ifdef __BIG_ENDIAN
	sigset_t newset;
	compat_sigset_t newset32;

	/* XXX: Don't preclude handling different sized sigset_t's.  */
	if (sigsetsize != sizeof(sigset_t))
		return -EINVAL;

	if (copy_from_user(&newset32, unewset, sizeof(compat_sigset_t)))
		return -EFAULT;
	sigset_from_compat(&newset, &newset32);
	return sigsuspend(&newset);
#else
	/* on little-endian bitmaps don't care about granularity */
	return sys_rt_sigsuspend((sigset_t __user *)unewset, sigsetsize);
#endif
}
#endif
3682

3683 3684 3685 3686 3687 3688 3689 3690 3691 3692 3693 3694 3695 3696 3697 3698
#ifdef CONFIG_OLD_SIGSUSPEND
SYSCALL_DEFINE1(sigsuspend, old_sigset_t, mask)
{
	sigset_t blocked;
	siginitset(&blocked, mask);
	return sigsuspend(&blocked);
}
#endif
#ifdef CONFIG_OLD_SIGSUSPEND3
SYSCALL_DEFINE3(sigsuspend, int, unused1, int, unused2, old_sigset_t, mask)
{
	sigset_t blocked;
	siginitset(&blocked, mask);
	return sigsuspend(&blocked);
}
#endif
3699

3700
__weak const char *arch_vma_name(struct vm_area_struct *vma)
3701 3702 3703 3704
{
	return NULL;
}

L
Linus Torvalds 已提交
3705 3706
void __init signals_init(void)
{
3707 3708 3709 3710
	/* If this check fails, the __ARCH_SI_PREAMBLE_SIZE value is wrong! */
	BUILD_BUG_ON(__ARCH_SI_PREAMBLE_SIZE
		!= offsetof(struct siginfo, _sifields._pad));

3711
	sigqueue_cachep = KMEM_CACHE(sigqueue, SLAB_PANIC);
L
Linus Torvalds 已提交
3712
}
3713 3714 3715 3716 3717 3718 3719 3720 3721 3722 3723 3724 3725 3726 3727 3728 3729 3730 3731 3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751 3752

#ifdef CONFIG_KGDB_KDB
#include <linux/kdb.h>
/*
 * kdb_send_sig_info - Allows kdb to send signals without exposing
 * signal internals.  This function checks if the required locks are
 * available before calling the main signal code, to avoid kdb
 * deadlocks.
 */
void
kdb_send_sig_info(struct task_struct *t, struct siginfo *info)
{
	static struct task_struct *kdb_prev_t;
	int sig, new_t;
	if (!spin_trylock(&t->sighand->siglock)) {
		kdb_printf("Can't do kill command now.\n"
			   "The sigmask lock is held somewhere else in "
			   "kernel, try again later\n");
		return;
	}
	spin_unlock(&t->sighand->siglock);
	new_t = kdb_prev_t != t;
	kdb_prev_t = t;
	if (t->state != TASK_RUNNING && new_t) {
		kdb_printf("Process is not RUNNING, sending a signal from "
			   "kdb risks deadlock\n"
			   "on the run queue locks. "
			   "The signal has _not_ been sent.\n"
			   "Reissue the kill command if you want to risk "
			   "the deadlock.\n");
		return;
	}
	sig = info->si_signo;
	if (send_sig_info(sig, info, t))
		kdb_printf("Fail to deliver Signal %d to process %d.\n",
			   sig, t->pid);
	else
		kdb_printf("Signal %d is sent to process %d.\n", sig, t->pid);
}
#endif	/* CONFIG_KGDB_KDB */