signal.c 95.8 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 && sig_kernel_only(sig)))
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
	/*
98 99 100
	 * Tracers may want to know about even ignored signal unless it
	 * is SIGKILL which can't be reported anyway but can be ignored
	 * by SIGNAL_UNKILLABLE task.
101
	 */
102 103 104 105
	if (t->ptrace && sig != SIGKILL)
		return 0;

	return sig_task_ignored(t, sig, force);
L
Linus Torvalds 已提交
106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139
}

/*
 * 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
140
static int recalc_sigpending_tsk(struct task_struct *t)
L
Linus Torvalds 已提交
141
{
142
	if ((t->jobctl & JOBCTL_PENDING_MASK) ||
L
Linus Torvalds 已提交
143
	    PENDING(&t->pending, &t->blocked) ||
R
Roland McGrath 已提交
144
	    PENDING(&t->signal->shared_pending, &t->blocked)) {
L
Linus Torvalds 已提交
145
		set_tsk_thread_flag(t, TIF_SIGPENDING);
R
Roland McGrath 已提交
146 147
		return 1;
	}
148 149 150 151 152
	/*
	 * 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
153 154 155 156 157 158 159 160 161 162 163
	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
164 165 166 167
}

void recalc_sigpending(void)
{
T
Tejun Heo 已提交
168
	if (!recalc_sigpending_tsk(current) && !freezing(current))
169 170
		clear_thread_flag(TIF_SIGPENDING);

L
Linus Torvalds 已提交
171 172 173 174
}

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

175 176
#define SYNCHRONOUS_MASK \
	(sigmask(SIGSEGV) | sigmask(SIGBUS) | sigmask(SIGILL) | \
177
	 sigmask(SIGTRAP) | sigmask(SIGFPE) | sigmask(SIGSYS))
178

179
int next_signal(struct sigpending *pending, sigset_t *mask)
L
Linus Torvalds 已提交
180 181 182
{
	unsigned long i, *s, *m, x;
	int sig = 0;
183

L
Linus Torvalds 已提交
184 185
	s = pending->signal.sig;
	m = mask->sig;
186 187 188 189 190 191 192 193 194 195 196 197 198

	/*
	 * 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
199 200
	switch (_NSIG_WORDS) {
	default:
201 202 203 204 205 206 207
		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
208 209
		break;

210 211 212
	case 2:
		x = s[1] &~ m[1];
		if (!x)
L
Linus Torvalds 已提交
213
			break;
214
		sig = ffz(~x) + _NSIG_BPW + 1;
L
Linus Torvalds 已提交
215 216
		break;

217 218
	case 1:
		/* Nothing to do */
L
Linus Torvalds 已提交
219 220
		break;
	}
221

L
Linus Torvalds 已提交
222 223 224
	return sig;
}

225 226 227 228 229 230 231 232 233 234
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;

235
	pr_info("%s/%d: reached RLIMIT_SIGPENDING, dropped signal %d\n",
236 237 238
				current->comm, current->pid, sig);
}

239
/**
240
 * task_set_jobctl_pending - set jobctl pending bits
241
 * @task: target task
242
 * @mask: pending bits to set
243
 *
244 245 246 247 248 249 250 251 252 253 254 255
 * 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.
 */
256
bool task_set_jobctl_pending(struct task_struct *task, unsigned long mask)
257 258 259 260 261 262 263 264 265 266 267 268 269 270 271
{
	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;
}

272
/**
273
 * task_clear_jobctl_trapping - clear jobctl trapping bit
274 275
 * @task: target task
 *
276 277 278 279
 * 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.
280 281 282 283
 *
 * CONTEXT:
 * Must be called with @task->sighand->siglock held.
 */
284
void task_clear_jobctl_trapping(struct task_struct *task)
285
{
286 287
	if (unlikely(task->jobctl & JOBCTL_TRAPPING)) {
		task->jobctl &= ~JOBCTL_TRAPPING;
288
		smp_mb();	/* advised by wake_up_bit() */
289
		wake_up_bit(&task->jobctl, JOBCTL_TRAPPING_BIT);
290 291 292
	}
}

293
/**
294
 * task_clear_jobctl_pending - clear jobctl pending bits
295
 * @task: target task
296
 * @mask: pending bits to clear
297
 *
298 299 300
 * 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.
301
 *
302 303
 * If clearing of @mask leaves no stop or trap pending, this function calls
 * task_clear_jobctl_trapping().
304 305 306 307
 *
 * CONTEXT:
 * Must be called with @task->sighand->siglock held.
 */
308
void task_clear_jobctl_pending(struct task_struct *task, unsigned long mask)
309
{
310 311 312 313 314 315
	BUG_ON(mask & ~JOBCTL_PENDING_MASK);

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

	task->jobctl &= ~mask;
316 317 318

	if (!(task->jobctl & JOBCTL_PENDING_MASK))
		task_clear_jobctl_trapping(task);
319 320 321 322 323 324
}

/**
 * task_participate_group_stop - participate in a group stop
 * @task: task participating in a group stop
 *
325
 * @task has %JOBCTL_STOP_PENDING set and is participating in a group stop.
326
 * Group stop states are cleared and the group stop count is consumed if
327
 * %JOBCTL_STOP_CONSUME was set.  If the consumption completes the group
328
 * stop, the appropriate %SIGNAL_* flags are set.
329 330 331
 *
 * CONTEXT:
 * Must be called with @task->sighand->siglock held.
332 333 334 335
 *
 * RETURNS:
 * %true if group stop completion should be notified to the parent, %false
 * otherwise.
336 337 338 339
 */
static bool task_participate_group_stop(struct task_struct *task)
{
	struct signal_struct *sig = task->signal;
340
	bool consume = task->jobctl & JOBCTL_STOP_CONSUME;
341

342
	WARN_ON_ONCE(!(task->jobctl & JOBCTL_STOP_PENDING));
343

344
	task_clear_jobctl_pending(task, JOBCTL_STOP_PENDING);
345 346 347 348 349 350 351

	if (!consume)
		return false;

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

352 353 354 355 356
	/*
	 * 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)) {
357
		signal_set_stop_flags(sig, SIGNAL_STOP_STOPPED);
358 359 360 361 362
		return true;
	}
	return false;
}

363 364 365
/*
 * allocate a new signal queue record
 * - this may be called without locks if and only if t == current, otherwise an
366
 *   appropriate lock must be held to stop the target task from exiting
367
 */
368 369
static struct sigqueue *
__sigqueue_alloc(int sig, struct task_struct *t, gfp_t flags, int override_rlimit)
L
Linus Torvalds 已提交
370 371
{
	struct sigqueue *q = NULL;
372
	struct user_struct *user;
L
Linus Torvalds 已提交
373

374
	/*
375 376
	 * Protect access to @t credentials. This can go away when all
	 * callers hold rcu read lock.
377
	 */
378
	rcu_read_lock();
D
David Howells 已提交
379
	user = get_uid(__task_cred(t)->user);
380
	atomic_inc(&user->sigpending);
381
	rcu_read_unlock();
382

L
Linus Torvalds 已提交
383
	if (override_rlimit ||
384
	    atomic_read(&user->sigpending) <=
385
			task_rlimit(t, RLIMIT_SIGPENDING)) {
L
Linus Torvalds 已提交
386
		q = kmem_cache_alloc(sigqueue_cachep, flags);
387 388 389 390
	} else {
		print_dropped_signal(sig);
	}

L
Linus Torvalds 已提交
391
	if (unlikely(q == NULL)) {
392
		atomic_dec(&user->sigpending);
D
David Howells 已提交
393
		free_uid(user);
L
Linus Torvalds 已提交
394 395 396
	} else {
		INIT_LIST_HEAD(&q->list);
		q->flags = 0;
D
David Howells 已提交
397
		q->user = user;
L
Linus Torvalds 已提交
398
	}
D
David Howells 已提交
399 400

	return q;
L
Linus Torvalds 已提交
401 402
}

403
static void __sigqueue_free(struct sigqueue *q)
L
Linus Torvalds 已提交
404 405 406 407 408 409 410 411
{
	if (q->flags & SIGQUEUE_PREALLOC)
		return;
	atomic_dec(&q->user->sigpending);
	free_uid(q->user);
	kmem_cache_free(sigqueue_cachep, q);
}

412
void flush_sigqueue(struct sigpending *queue)
L
Linus Torvalds 已提交
413 414 415 416 417 418 419 420 421 422 423 424
{
	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);
	}
}

/*
425
 * Flush all pending signals for this kthread.
L
Linus Torvalds 已提交
426
 */
427
void flush_signals(struct task_struct *t)
L
Linus Torvalds 已提交
428 429 430 431
{
	unsigned long flags;

	spin_lock_irqsave(&t->sighand->siglock, flags);
432 433 434
	clear_tsk_thread_flag(t, TIF_SIGPENDING);
	flush_sigqueue(&t->pending);
	flush_sigqueue(&t->signal->shared_pending);
L
Linus Torvalds 已提交
435 436 437
	spin_unlock_irqrestore(&t->sighand->siglock, flags);
}

438
#ifdef CONFIG_POSIX_TIMERS
439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471
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);
}
472
#endif
473

474 475 476 477 478 479 480 481 482 483
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
484 485 486 487 488 489 490 491 492 493 494 495 496
/*
 * 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;
497
#ifdef __ARCH_HAS_SA_RESTORER
498 499
		ka->sa.sa_restorer = NULL;
#endif
L
Linus Torvalds 已提交
500 501 502 503 504
		sigemptyset(&ka->sa.sa_mask);
		ka++;
	}
}

505 506
int unhandled_signal(struct task_struct *tsk, int sig)
{
507
	void __user *handler = tsk->sighand->action[sig-1].sa.sa_handler;
508
	if (is_global_init(tsk))
509
		return 1;
510
	if (handler != SIG_IGN && handler != SIG_DFL)
511
		return 0;
T
Tejun Heo 已提交
512 513
	/* if ptraced, let the tracer determine */
	return !tsk->ptrace;
514 515
}

516 517
static void collect_signal(int sig, struct sigpending *list, siginfo_t *info,
			   bool *resched_timer)
L
Linus Torvalds 已提交
518 519 520 521 522 523 524 525 526
{
	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) {
527 528
			if (first)
				goto still_pending;
L
Linus Torvalds 已提交
529 530 531
			first = q;
		}
	}
532 533 534

	sigdelset(&list->signal, sig);

L
Linus Torvalds 已提交
535
	if (first) {
536
still_pending:
L
Linus Torvalds 已提交
537 538
		list_del_init(&first->list);
		copy_siginfo(info, &first->info);
539 540 541 542 543 544

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

L
Linus Torvalds 已提交
545 546
		__sigqueue_free(first);
	} else {
547 548 549 550
		/*
		 * 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
551 552 553
		 */
		info->si_signo = sig;
		info->si_errno = 0;
554
		info->si_code = SI_USER;
L
Linus Torvalds 已提交
555 556 557 558 559 560
		info->si_pid = 0;
		info->si_uid = 0;
	}
}

static int __dequeue_signal(struct sigpending *pending, sigset_t *mask,
561
			siginfo_t *info, bool *resched_timer)
L
Linus Torvalds 已提交
562
{
563
	int sig = next_signal(pending, mask);
L
Linus Torvalds 已提交
564

565
	if (sig)
566
		collect_signal(sig, pending, info, resched_timer);
L
Linus Torvalds 已提交
567 568 569 570
	return sig;
}

/*
571
 * Dequeue a signal and return the element to the caller, which is
L
Linus Torvalds 已提交
572 573 574 575 576 577
 * expected to free it.
 *
 * All callers have to hold the siglock.
 */
int dequeue_signal(struct task_struct *tsk, sigset_t *mask, siginfo_t *info)
{
578
	bool resched_timer = false;
579
	int signr;
580 581 582 583

	/* We only dequeue private signals from ourselves, we don't let
	 * signalfd steal them
	 */
584
	signr = __dequeue_signal(&tsk->pending, mask, info, &resched_timer);
585
	if (!signr) {
L
Linus Torvalds 已提交
586
		signr = __dequeue_signal(&tsk->signal->shared_pending,
587
					 mask, info, &resched_timer);
588
#ifdef CONFIG_POSIX_TIMERS
589 590 591 592 593 594
		/*
		 * 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
595
		 * compliant with the old way of self-restarting
596 597 598 599 600 601 602 603 604 605
		 * 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
606
			    tsk->signal->it_real_incr != 0) {
607 608 609 610 611
				hrtimer_forward(tmr, tmr->base->get_time(),
						tsk->signal->it_real_incr);
				hrtimer_restart(tmr);
			}
		}
612
#endif
613
	}
614

D
Davide Libenzi 已提交
615
	recalc_sigpending();
616 617 618 619
	if (!signr)
		return 0;

	if (unlikely(sig_kernel_stop(signr))) {
620 621 622 623 624 625 626 627 628 629 630 631
		/*
		 * 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.
		 */
632
		current->jobctl |= JOBCTL_STOP_DEQUEUED;
633
	}
634
#ifdef CONFIG_POSIX_TIMERS
635
	if (resched_timer) {
L
Linus Torvalds 已提交
636 637 638 639 640 641 642
		/*
		 * 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);
643
		posixtimer_rearm(info);
L
Linus Torvalds 已提交
644 645
		spin_lock(&tsk->sighand->siglock);
	}
646
#endif
L
Linus Torvalds 已提交
647 648 649 650 651 652 653 654 655 656 657 658 659 660
	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
 */
661
void signal_wake_up_state(struct task_struct *t, unsigned int state)
L
Linus Torvalds 已提交
662 663 664
{
	set_tsk_thread_flag(t, TIF_SIGPENDING);
	/*
665
	 * TASK_WAKEKILL also means wake it up in the stopped/traced/killable
M
Matthew Wilcox 已提交
666
	 * case. We don't check t->state here because there is a race with it
L
Linus Torvalds 已提交
667 668 669 670
	 * 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.
	 */
671
	if (!wake_up_state(t, state | TASK_INTERRUPTIBLE))
L
Linus Torvalds 已提交
672 673 674
		kick_process(t);
}

675 676 677 678 679 680
/*
 * Remove signals in mask from the pending set and queue.
 * Returns 1 if any signals were found.
 *
 * All callers must be holding the siglock.
 */
681
static int flush_sigqueue_mask(sigset_t *mask, struct sigpending *s)
682 683 684 685 686 687 688 689
{
	struct sigqueue *q, *n;
	sigset_t m;

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

690
	sigandnsets(&s->signal, &s->signal, mask);
691 692 693 694 695 696 697 698
	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
699

700 701 702 703 704 705 706 707 708 709 710
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));
}

711 712 713 714 715 716 717 718
/*
 * 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);

719 720 721 722
	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))
723 724
		return 1;

725
	if (ns_capable(tcred->user_ns, CAP_KILL))
726 727 728 729 730
		return 1;

	return 0;
}

L
Linus Torvalds 已提交
731 732
/*
 * Bad permissions for sending the signal
733
 * - the caller must hold the RCU read lock
L
Linus Torvalds 已提交
734 735 736 737
 */
static int check_kill_permission(int sig, struct siginfo *info,
				 struct task_struct *t)
{
738
	struct pid *sid;
739 740
	int error;

741
	if (!valid_signal(sig))
742 743
		return -EINVAL;

744
	if (!si_fromuser(info))
745
		return 0;
A
Amy Griffis 已提交
746

747 748
	error = audit_signal_info(sig, t); /* Let audit system see the signal */
	if (error)
L
Linus Torvalds 已提交
749
		return error;
750

751
	if (!same_thread_group(current, t) &&
752
	    !kill_ok_by_cred(t)) {
753 754 755 756 757 758 759 760 761 762 763 764 765
		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;
		}
	}
766

A
Amy Griffis 已提交
767
	return security_task_kill(t, info, sig, 0);
L
Linus Torvalds 已提交
768 769
}

770 771 772 773 774 775 776 777
/**
 * 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
778 779 780 781 782
 * 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.
783 784 785 786 787 788 789 790 791 792
 *
 * 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);
793
	ptrace_signal_wake_up(t, t->jobctl & JOBCTL_LISTENING);
794 795
}

L
Linus Torvalds 已提交
796
/*
797 798
 * 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
799 800
 * time regardless of blocking, ignoring, or handling.  This does the
 * actual continuing for SIGCONT, but not the actual stopping for stop
801 802 803 804
 * 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
805
 */
806
static bool prepare_signal(int sig, struct task_struct *p, bool force)
L
Linus Torvalds 已提交
807
{
808
	struct signal_struct *signal = p->signal;
L
Linus Torvalds 已提交
809
	struct task_struct *t;
810
	sigset_t flush;
L
Linus Torvalds 已提交
811

812
	if (signal->flags & (SIGNAL_GROUP_EXIT | SIGNAL_GROUP_COREDUMP)) {
813
		if (!(signal->flags & SIGNAL_GROUP_EXIT))
814
			return sig == SIGKILL;
L
Linus Torvalds 已提交
815
		/*
816
		 * The process is in the middle of dying, nothing to do.
L
Linus Torvalds 已提交
817
		 */
818
	} else if (sig_kernel_stop(sig)) {
L
Linus Torvalds 已提交
819 820 821
		/*
		 * This is a stop signal.  Remove SIGCONT from all queues.
		 */
822
		siginitset(&flush, sigmask(SIGCONT));
823
		flush_sigqueue_mask(&flush, &signal->shared_pending);
824
		for_each_thread(p, t)
825
			flush_sigqueue_mask(&flush, &t->pending);
L
Linus Torvalds 已提交
826
	} else if (sig == SIGCONT) {
827
		unsigned int why;
L
Linus Torvalds 已提交
828
		/*
829
		 * Remove all stop signals from all queues, wake all threads.
L
Linus Torvalds 已提交
830
		 */
831
		siginitset(&flush, SIG_KERNEL_STOP_MASK);
832
		flush_sigqueue_mask(&flush, &signal->shared_pending);
833
		for_each_thread(p, t) {
834
			flush_sigqueue_mask(&flush, &t->pending);
835
			task_clear_jobctl_pending(t, JOBCTL_STOP_PENDING);
836 837 838 839
			if (likely(!(t->ptrace & PT_SEIZED)))
				wake_up_state(t, __TASK_STOPPED);
			else
				ptrace_trap_notify(t);
840
		}
L
Linus Torvalds 已提交
841

842 843 844 845 846 847 848 849 850
		/*
		 * 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;
851
		if (signal->flags & SIGNAL_STOP_STOPPED)
852
			why |= SIGNAL_CLD_CONTINUED;
853
		else if (signal->group_stop_count)
854 855 856
			why |= SIGNAL_CLD_STOPPED;

		if (why) {
857
			/*
858
			 * The first thread which returns from do_signal_stop()
859 860 861
			 * will take ->siglock, notice SIGNAL_CLD_MASK, and
			 * notify its parent. See get_signal_to_deliver().
			 */
862
			signal_set_stop_flags(signal, why | SIGNAL_STOP_CONTINUED);
863 864
			signal->group_stop_count = 0;
			signal->group_exit_code = 0;
L
Linus Torvalds 已提交
865 866
		}
	}
867

868
	return !sig_ignored(p, sig, force);
L
Linus Torvalds 已提交
869 870
}

871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891
/*
 * 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);
}

892
static void complete_signal(int sig, struct task_struct *p, int group)
893 894 895 896 897 898 899 900 901 902 903 904
{
	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;
905
	else if (!group || thread_group_empty(p))
906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932
		/*
		 * 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.
	 */
933
	if (sig_fatal(p, sig) &&
934
	    !(signal->flags & SIGNAL_GROUP_EXIT) &&
935
	    !sigismember(&t->real_blocked, sig) &&
936
	    (sig == SIGKILL || !p->ptrace)) {
937 938 939 940 941 942 943 944 945 946 947 948 949 950 951
		/*
		 * 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 {
952
				task_clear_jobctl_pending(t, JOBCTL_PENDING_MASK);
953 954 955 956 957 958 959 960 961 962 963 964 965 966 967
				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;
}

968 969 970 971 972
static inline int legacy_queue(struct sigpending *signals, int sig)
{
	return (sig < SIGRTMIN) && sigismember(&signals->signal, sig);
}

973 974 975 976 977 978 979 980 981
#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;

982 983 984 985
	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();
986 987 988 989 990 991 992 993
}
#else
static inline void userns_fixup_signal_uid(struct siginfo *info, struct task_struct *t)
{
	return;
}
#endif

994 995
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
996
{
997
	struct sigpending *pending;
998
	struct sigqueue *q;
999
	int override_rlimit;
1000
	int ret = 0, result;
1001

1002
	assert_spin_locked(&t->sighand->siglock);
1003

1004
	result = TRACE_SIGNAL_IGNORED;
1005 1006
	if (!prepare_signal(sig, t,
			from_ancestor_ns || (info == SEND_SIG_FORCED)))
1007
		goto ret;
1008 1009

	pending = group ? &t->signal->shared_pending : &t->pending;
1010 1011 1012 1013 1014
	/*
	 * 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.
	 */
1015
	result = TRACE_SIGNAL_ALREADY_PENDING;
1016
	if (legacy_queue(pending, sig))
1017 1018 1019
		goto ret;

	result = TRACE_SIGNAL_DELIVERED;
L
Linus Torvalds 已提交
1020 1021 1022 1023
	/*
	 * fast-pathed signals for kernel-internal things like SIGSTOP
	 * or SIGKILL.
	 */
1024
	if (info == SEND_SIG_FORCED)
L
Linus Torvalds 已提交
1025 1026
		goto out_set;

1027 1028 1029 1030 1031 1032 1033 1034 1035
	/*
	 * 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.
	 */
1036 1037 1038 1039 1040
	if (sig < SIGRTMIN)
		override_rlimit = (is_si_special(info) || info->si_code >= 0);
	else
		override_rlimit = 0;

1041
	q = __sigqueue_alloc(sig, t, GFP_ATOMIC, override_rlimit);
L
Linus Torvalds 已提交
1042
	if (q) {
1043
		list_add_tail(&q->list, &pending->list);
L
Linus Torvalds 已提交
1044
		switch ((unsigned long) info) {
1045
		case (unsigned long) SEND_SIG_NOINFO:
L
Linus Torvalds 已提交
1046 1047 1048
			q->info.si_signo = sig;
			q->info.si_errno = 0;
			q->info.si_code = SI_USER;
1049
			q->info.si_pid = task_tgid_nr_ns(current,
1050
							task_active_pid_ns(t));
1051
			q->info.si_uid = from_kuid_munged(current_user_ns(), current_uid());
L
Linus Torvalds 已提交
1052
			break;
1053
		case (unsigned long) SEND_SIG_PRIV:
L
Linus Torvalds 已提交
1054 1055 1056 1057 1058 1059 1060 1061
			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);
1062 1063
			if (from_ancestor_ns)
				q->info.si_pid = 0;
L
Linus Torvalds 已提交
1064 1065
			break;
		}
1066 1067 1068

		userns_fixup_signal_uid(&q->info, t);

1069
	} else if (!is_si_special(info)) {
1070 1071 1072 1073 1074 1075
		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().
			 */
1076 1077 1078
			result = TRACE_SIGNAL_OVERFLOW_FAIL;
			ret = -EAGAIN;
			goto ret;
1079 1080 1081 1082 1083
		} else {
			/*
			 * This is a silent loss of information.  We still
			 * send the signal, but the *info bits are lost.
			 */
1084
			result = TRACE_SIGNAL_LOSE_INFO;
1085
		}
L
Linus Torvalds 已提交
1086 1087 1088
	}

out_set:
1089
	signalfd_notify(t, sig);
1090
	sigaddset(&pending->signal, sig);
1091
	complete_signal(sig, t, group);
1092 1093 1094
ret:
	trace_signal_generate(sig, info, t, group, result);
	return ret;
L
Linus Torvalds 已提交
1095 1096
}

1097 1098 1099
static int send_signal(int sig, struct siginfo *info, struct task_struct *t,
			int group)
{
1100 1101 1102
	int from_ancestor_ns = 0;

#ifdef CONFIG_PID_NS
1103 1104
	from_ancestor_ns = si_fromuser(info) &&
			   !task_pid_nr_ns(current, task_active_pid_ns(t));
1105 1106 1107
#endif

	return __send_signal(sig, info, t, group, from_ancestor_ns);
1108 1109
}

1110
static void print_fatal_signal(int signr)
I
Ingo Molnar 已提交
1111
{
1112
	struct pt_regs *regs = signal_pt_regs();
1113
	pr_info("potentially unexpected fatal signal %d.\n", signr);
I
Ingo Molnar 已提交
1114

A
Al Viro 已提交
1115
#if defined(__i386__) && !defined(__arch_um__)
1116
	pr_info("code at %08lx: ", regs->ip);
I
Ingo Molnar 已提交
1117 1118 1119 1120 1121
	{
		int i;
		for (i = 0; i < 16; i++) {
			unsigned char insn;

1122 1123
			if (get_user(insn, (unsigned char *)(regs->ip + i)))
				break;
1124
			pr_cont("%02x ", insn);
I
Ingo Molnar 已提交
1125 1126
		}
	}
1127
	pr_cont("\n");
I
Ingo Molnar 已提交
1128
#endif
1129
	preempt_disable();
I
Ingo Molnar 已提交
1130
	show_regs(regs);
1131
	preempt_enable();
I
Ingo Molnar 已提交
1132 1133 1134 1135 1136 1137 1138 1139 1140 1141
}

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
1142

1143 1144 1145 1146 1147 1148
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
1149 1150 1151
static int
specific_send_sig_info(int sig, struct siginfo *info, struct task_struct *t)
{
1152
	return send_signal(sig, info, t, 0);
L
Linus Torvalds 已提交
1153 1154
}

1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168
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
1169 1170 1171
/*
 * Force a signal that the process can't ignore: if necessary
 * we unblock the signal and change any SIG_IGN to SIG_DFL.
1172 1173 1174 1175 1176
 *
 * 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.
 *
1177 1178
 * 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
1179 1180 1181 1182 1183
 */
int
force_sig_info(int sig, struct siginfo *info, struct task_struct *t)
{
	unsigned long int flags;
1184 1185
	int ret, blocked, ignored;
	struct k_sigaction *action;
L
Linus Torvalds 已提交
1186 1187

	spin_lock_irqsave(&t->sighand->siglock, flags);
1188 1189 1190 1191 1192 1193 1194
	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
1195
			recalc_sigpending_and_wake(t);
1196
		}
L
Linus Torvalds 已提交
1197
	}
1198 1199 1200 1201 1202
	/*
	 * 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)
1203
		t->signal->flags &= ~SIGNAL_UNKILLABLE;
L
Linus Torvalds 已提交
1204 1205 1206 1207 1208 1209 1210 1211 1212
	ret = specific_send_sig_info(sig, info, t);
	spin_unlock_irqrestore(&t->sighand->siglock, flags);

	return ret;
}

/*
 * Nuke all other threads in the group.
 */
1213
int zap_other_threads(struct task_struct *p)
L
Linus Torvalds 已提交
1214
{
1215 1216
	struct task_struct *t = p;
	int count = 0;
L
Linus Torvalds 已提交
1217 1218 1219

	p->signal->group_stop_count = 0;

1220
	while_each_thread(p, t) {
1221
		task_clear_jobctl_pending(t, JOBCTL_PENDING_MASK);
1222 1223 1224
		count++;

		/* Don't bother with already dead threads */
L
Linus Torvalds 已提交
1225 1226 1227 1228 1229
		if (t->exit_state)
			continue;
		sigaddset(&t->pending.signal, SIGKILL);
		signal_wake_up(t, 1);
	}
1230 1231

	return count;
L
Linus Torvalds 已提交
1232 1233
}

1234 1235
struct sighand_struct *__lock_task_sighand(struct task_struct *tsk,
					   unsigned long *flags)
1236 1237 1238 1239
{
	struct sighand_struct *sighand;

	for (;;) {
1240 1241 1242 1243
		/*
		 * Disable interrupts early to avoid deadlocks.
		 * See rcu_read_unlock() comment header for details.
		 */
1244 1245
		local_irq_save(*flags);
		rcu_read_lock();
1246
		sighand = rcu_dereference(tsk->sighand);
1247 1248 1249
		if (unlikely(sighand == NULL)) {
			rcu_read_unlock();
			local_irq_restore(*flags);
1250
			break;
1251
		}
1252 1253
		/*
		 * This sighand can be already freed and even reused, but
1254
		 * we rely on SLAB_TYPESAFE_BY_RCU and sighand_ctor() which
1255 1256 1257 1258 1259 1260 1261 1262
		 * 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.
		 */
1263 1264 1265
		spin_lock(&sighand->siglock);
		if (likely(sighand == tsk->sighand)) {
			rcu_read_unlock();
1266
			break;
1267 1268 1269 1270
		}
		spin_unlock(&sighand->siglock);
		rcu_read_unlock();
		local_irq_restore(*flags);
1271 1272 1273 1274 1275
	}

	return sighand;
}

1276 1277 1278
/*
 * send signal info to all the members of a group
 */
L
Linus Torvalds 已提交
1279 1280
int group_send_sig_info(int sig, struct siginfo *info, struct task_struct *p)
{
1281 1282 1283 1284 1285
	int ret;

	rcu_read_lock();
	ret = check_kill_permission(sig, info, p);
	rcu_read_unlock();
1286

1287 1288
	if (!ret && sig)
		ret = do_send_sig_info(sig, info, p, true);
L
Linus Torvalds 已提交
1289 1290 1291 1292 1293

	return ret;
}

/*
1294
 * __kill_pgrp_info() sends a signal to a process group: this is what the tty
L
Linus Torvalds 已提交
1295
 * control characters do (^C, ^Z etc)
1296
 * - the caller must hold at least a readlock on tasklist_lock
L
Linus Torvalds 已提交
1297
 */
1298
int __kill_pgrp_info(int sig, struct siginfo *info, struct pid *pgrp)
L
Linus Torvalds 已提交
1299 1300 1301 1302 1303 1304
{
	struct task_struct *p = NULL;
	int retval, success;

	success = 0;
	retval = -ESRCH;
1305
	do_each_pid_task(pgrp, PIDTYPE_PGID, p) {
L
Linus Torvalds 已提交
1306 1307 1308
		int err = group_send_sig_info(sig, info, p);
		success |= !err;
		retval = err;
1309
	} while_each_pid_task(pgrp, PIDTYPE_PGID, p);
L
Linus Torvalds 已提交
1310 1311 1312
	return success ? 0 : retval;
}

1313
int kill_pid_info(int sig, struct siginfo *info, struct pid *pid)
L
Linus Torvalds 已提交
1314
{
1315
	int error = -ESRCH;
L
Linus Torvalds 已提交
1316 1317
	struct task_struct *p;

1318 1319 1320 1321 1322 1323 1324 1325
	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;
1326

1327 1328 1329 1330 1331 1332
		/*
		 * 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
1333 1334
}

1335
static int kill_proc_info(int sig, struct siginfo *info, pid_t pid)
1336 1337 1338
{
	int error;
	rcu_read_lock();
1339
	error = kill_pid_info(sig, info, find_vpid(pid));
1340 1341 1342 1343
	rcu_read_unlock();
	return error;
}

1344 1345 1346 1347
static int kill_as_cred_perm(const struct cred *cred,
			     struct task_struct *target)
{
	const struct cred *pcred = __task_cred(target);
1348 1349
	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))
1350 1351 1352 1353
		return 0;
	return 1;
}

1354
/* like kill_pid_info(), but doesn't use uid/euid of "current" */
1355 1356
int kill_pid_info_as_cred(int sig, struct siginfo *info, struct pid *pid,
			 const struct cred *cred, u32 secid)
1357 1358 1359
{
	int ret = -EINVAL;
	struct task_struct *p;
1360
	unsigned long flags;
1361 1362 1363 1364

	if (!valid_signal(sig))
		return ret;

1365
	rcu_read_lock();
1366
	p = pid_task(pid, PIDTYPE_PID);
1367 1368 1369 1370
	if (!p) {
		ret = -ESRCH;
		goto out_unlock;
	}
1371
	if (si_fromuser(info) && !kill_as_cred_perm(cred, p)) {
1372 1373 1374
		ret = -EPERM;
		goto out_unlock;
	}
1375 1376 1377
	ret = security_task_kill(p, info, sig, secid);
	if (ret)
		goto out_unlock;
1378 1379 1380 1381 1382 1383 1384

	if (sig) {
		if (lock_task_sighand(p, &flags)) {
			ret = __send_signal(sig, info, p, 1, 0);
			unlock_task_sighand(p, &flags);
		} else
			ret = -ESRCH;
1385 1386
	}
out_unlock:
1387
	rcu_read_unlock();
1388 1389
	return ret;
}
1390
EXPORT_SYMBOL_GPL(kill_pid_info_as_cred);
L
Linus Torvalds 已提交
1391 1392 1393 1394 1395 1396 1397 1398

/*
 * 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.
 */

1399
static int kill_something_info(int sig, struct siginfo *info, pid_t pid)
L
Linus Torvalds 已提交
1400
{
1401
	int ret;
1402 1403 1404 1405 1406 1407 1408 1409

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

1410 1411 1412 1413
	/* -INT_MIN is undefined.  Exclude this case to avoid a UBSAN warning */
	if (pid == INT_MIN)
		return -ESRCH;

1414 1415 1416 1417 1418
	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
1419 1420 1421 1422
		int retval = 0, count = 0;
		struct task_struct * p;

		for_each_process(p) {
1423 1424
			if (task_pid_vnr(p) > 1 &&
					!same_thread_group(p, current)) {
L
Linus Torvalds 已提交
1425 1426 1427 1428 1429 1430
				int err = group_send_sig_info(sig, info, p);
				++count;
				if (err != -EPERM)
					retval = err;
			}
		}
1431
		ret = count ? retval : -ESRCH;
L
Linus Torvalds 已提交
1432
	}
1433 1434
	read_unlock(&tasklist_lock);

1435
	return ret;
L
Linus Torvalds 已提交
1436 1437 1438 1439 1440 1441
}

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

1442
int send_sig_info(int sig, struct siginfo *info, struct task_struct *p)
L
Linus Torvalds 已提交
1443 1444 1445 1446 1447
{
	/*
	 * Make sure legacy kernel users don't send in bad values
	 * (normal paths check this in check_kill_permission).
	 */
1448
	if (!valid_signal(sig))
L
Linus Torvalds 已提交
1449 1450
		return -EINVAL;

1451
	return do_send_sig_info(sig, info, p, false);
L
Linus Torvalds 已提交
1452 1453
}

1454 1455 1456
#define __si_special(priv) \
	((priv) ? SEND_SIG_PRIV : SEND_SIG_NOINFO)

L
Linus Torvalds 已提交
1457 1458 1459
int
send_sig(int sig, struct task_struct *p, int priv)
{
1460
	return send_sig_info(sig, __si_special(priv), p);
L
Linus Torvalds 已提交
1461 1462 1463 1464 1465
}

void
force_sig(int sig, struct task_struct *p)
{
1466
	force_sig_info(sig, SEND_SIG_PRIV, p);
L
Linus Torvalds 已提交
1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487
}

/*
 * 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;
}

1488 1489
int kill_pgrp(struct pid *pid, int sig, int priv)
{
1490 1491 1492 1493 1494 1495 1496
	int ret;

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

	return ret;
1497 1498 1499 1500 1501 1502 1503 1504 1505
}
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
1506 1507 1508 1509
/*
 * 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
1510
 * expirations or I/O completions".  In the case of POSIX Timers
L
Linus Torvalds 已提交
1511 1512 1513 1514 1515 1516
 * 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)
{
1517
	struct sigqueue *q = __sigqueue_alloc(-1, current, GFP_KERNEL, 0);
L
Linus Torvalds 已提交
1518

1519
	if (q)
L
Linus Torvalds 已提交
1520
		q->flags |= SIGQUEUE_PREALLOC;
1521 1522

	return q;
L
Linus Torvalds 已提交
1523 1524 1525 1526 1527
}

void sigqueue_free(struct sigqueue *q)
{
	unsigned long flags;
1528 1529
	spinlock_t *lock = &current->sighand->siglock;

L
Linus Torvalds 已提交
1530 1531
	BUG_ON(!(q->flags & SIGQUEUE_PREALLOC));
	/*
1532 1533
	 * We must hold ->siglock while testing q->list
	 * to serialize with collect_signal() or with
1534
	 * __exit_signal()->flush_sigqueue().
L
Linus Torvalds 已提交
1535
	 */
1536
	spin_lock_irqsave(lock, flags);
1537 1538 1539 1540 1541
	q->flags &= ~SIGQUEUE_PREALLOC;
	/*
	 * If it is queued it will be freed when dequeued,
	 * like the "regular" sigqueue.
	 */
1542
	if (!list_empty(&q->list))
1543
		q = NULL;
1544 1545
	spin_unlock_irqrestore(lock, flags);

1546 1547
	if (q)
		__sigqueue_free(q);
L
Linus Torvalds 已提交
1548 1549
}

1550
int send_sigqueue(struct sigqueue *q, struct task_struct *t, int group)
1551
{
1552
	int sig = q->info.si_signo;
1553
	struct sigpending *pending;
1554
	unsigned long flags;
1555
	int ret, result;
1556

1557
	BUG_ON(!(q->flags & SIGQUEUE_PREALLOC));
1558 1559 1560 1561 1562

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

1563
	ret = 1; /* the signal is ignored */
1564
	result = TRACE_SIGNAL_IGNORED;
1565
	if (!prepare_signal(sig, t, false))
1566 1567 1568
		goto out;

	ret = 0;
1569 1570 1571 1572 1573 1574 1575
	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++;
1576
		result = TRACE_SIGNAL_ALREADY_PENDING;
1577
		goto out;
1578
	}
1579
	q->info.si_overrun = 0;
1580 1581

	signalfd_notify(t, sig);
1582
	pending = group ? &t->signal->shared_pending : &t->pending;
1583 1584
	list_add_tail(&q->list, &pending->list);
	sigaddset(&pending->signal, sig);
1585
	complete_signal(sig, t, group);
1586
	result = TRACE_SIGNAL_DELIVERED;
1587
out:
1588
	trace_signal_generate(sig, &q->info, t, group, result);
1589 1590 1591
	unlock_task_sighand(t, &flags);
ret:
	return ret;
1592 1593
}

L
Linus Torvalds 已提交
1594 1595 1596
/*
 * 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
1597
 *
1598 1599
 * Returns true if our parent ignored us and so we've switched to
 * self-reaping.
L
Linus Torvalds 已提交
1600
 */
1601
bool do_notify_parent(struct task_struct *tsk, int sig)
L
Linus Torvalds 已提交
1602 1603 1604 1605
{
	struct siginfo info;
	unsigned long flags;
	struct sighand_struct *psig;
1606
	bool autoreap = false;
1607
	u64 utime, stime;
L
Linus Torvalds 已提交
1608 1609 1610 1611

	BUG_ON(sig == -1);

 	/* do_notify_parent_cldstop should have been called instead.  */
M
Matthew Wilcox 已提交
1612
 	BUG_ON(task_is_stopped_or_traced(tsk));
L
Linus Torvalds 已提交
1613

T
Tejun Heo 已提交
1614
	BUG_ON(!tsk->ptrace &&
L
Linus Torvalds 已提交
1615 1616
	       (tsk->group_leader != tsk || !thread_group_empty(tsk)));

1617 1618 1619 1620 1621 1622 1623 1624 1625
	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
1626 1627
	info.si_signo = sig;
	info.si_errno = 0;
1628
	/*
1629 1630
	 * We are under tasklist_lock here so our parent is tied to
	 * us and cannot change.
1631
	 *
1632 1633
	 * task_active_pid_ns will always return the same pid namespace
	 * until a task passes through release_task.
1634 1635 1636 1637 1638 1639
	 *
	 * 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();
1640
	info.si_pid = task_pid_nr_ns(tsk, task_active_pid_ns(tsk->parent));
1641 1642
	info.si_uid = from_kuid_munged(task_cred_xxx(tsk->parent, user_ns),
				       task_uid(tsk));
1643 1644
	rcu_read_unlock();

1645 1646 1647
	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
1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660

	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
1661
	if (!tsk->ptrace && sig == SIGCHLD &&
L
Linus Torvalds 已提交
1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678
	    (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).
		 */
1679
		autoreap = true;
L
Linus Torvalds 已提交
1680
		if (psig->action[SIGCHLD-1].sa.sa_handler == SIG_IGN)
1681
			sig = 0;
L
Linus Torvalds 已提交
1682
	}
1683
	if (valid_signal(sig) && sig)
L
Linus Torvalds 已提交
1684 1685 1686
		__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
1687

1688
	return autoreap;
L
Linus Torvalds 已提交
1689 1690
}

1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705
/**
 * 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
1706 1707 1708
{
	struct siginfo info;
	unsigned long flags;
1709
	struct task_struct *parent;
L
Linus Torvalds 已提交
1710
	struct sighand_struct *sighand;
1711
	u64 utime, stime;
L
Linus Torvalds 已提交
1712

1713
	if (for_ptracer) {
1714
		parent = tsk->parent;
1715
	} else {
1716 1717 1718 1719
		tsk = tsk->group_leader;
		parent = tsk->real_parent;
	}

L
Linus Torvalds 已提交
1720 1721
	info.si_signo = SIGCHLD;
	info.si_errno = 0;
1722
	/*
1723
	 * see comment in do_notify_parent() about the following 4 lines
1724 1725
	 */
	rcu_read_lock();
1726
	info.si_pid = task_pid_nr_ns(tsk, task_active_pid_ns(parent));
1727
	info.si_uid = from_kuid_munged(task_cred_xxx(parent, user_ns), task_uid(tsk));
1728 1729
	rcu_read_unlock();

1730 1731 1732
	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
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 1760

 	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);
}

1761 1762
static inline int may_ptrace_stop(void)
{
T
Tejun Heo 已提交
1763
	if (!likely(current->ptrace))
1764 1765 1766 1767 1768 1769 1770
		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
1771
	 * ->siglock we must see ->core_state != NULL. Otherwise it
1772
	 * is safe to enter schedule().
1773 1774 1775 1776
	 *
	 * 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.
1777
	 */
1778
	if (unlikely(current->mm->core_state) &&
1779 1780 1781 1782 1783 1784
	    unlikely(current->mm == current->parent->mm))
		return 0;

	return 1;
}

R
Roland McGrath 已提交
1785
/*
1786
 * Return non-zero if there is a SIGKILL that should be waking us up.
R
Roland McGrath 已提交
1787 1788 1789 1790
 * Called with the siglock held.
 */
static int sigkill_pending(struct task_struct *tsk)
{
1791 1792
	return	sigismember(&tsk->pending.signal, SIGKILL) ||
		sigismember(&tsk->signal->shared_pending.signal, SIGKILL);
R
Roland McGrath 已提交
1793 1794
}

L
Linus Torvalds 已提交
1795 1796 1797 1798 1799 1800 1801 1802
/*
 * 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.
 *
1803 1804
 * 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
1805
 */
T
Tejun Heo 已提交
1806
static void ptrace_stop(int exit_code, int why, int clear_code, siginfo_t *info)
1807 1808
	__releases(&current->sighand->siglock)
	__acquires(&current->sighand->siglock)
L
Linus Torvalds 已提交
1809
{
1810 1811
	bool gstop_done = false;

R
Roland McGrath 已提交
1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826
	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);
1827 1828
		if (sigkill_pending(current))
			return;
R
Roland McGrath 已提交
1829 1830
	}

L
Linus Torvalds 已提交
1831
	/*
1832 1833 1834 1835 1836
	 * 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
1837
	 */
1838
	set_current_state(TASK_TRACED);
L
Linus Torvalds 已提交
1839 1840 1841 1842

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

1843
	/*
1844 1845
	 * If @why is CLD_STOPPED, we're trapping to participate in a group
	 * stop.  Do the bookkeeping.  Note that if SIGCONT was delievered
1846 1847 1848
	 * 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.
1849
	 */
1850
	if (why == CLD_STOPPED && (current->jobctl & JOBCTL_STOP_PENDING))
1851
		gstop_done = task_participate_group_stop(current);
1852

1853
	/* any trap clears pending STOP trap, STOP trap clears NOTIFY */
1854
	task_clear_jobctl_pending(current, JOBCTL_TRAP_STOP);
1855 1856
	if (info && info->si_code >> 8 == PTRACE_EVENT_STOP)
		task_clear_jobctl_pending(current, JOBCTL_TRAP_NOTIFY);
1857

1858
	/* entering a trap, clear TRAPPING */
1859
	task_clear_jobctl_trapping(current);
1860

L
Linus Torvalds 已提交
1861 1862
	spin_unlock_irq(&current->sighand->siglock);
	read_lock(&tasklist_lock);
1863
	if (may_ptrace_stop()) {
1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874
		/*
		 * 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);
1875
		if (gstop_done && ptrace_reparented(current))
1876 1877
			do_notify_parent_cldstop(current, false, why);

M
Miklos Szeredi 已提交
1878 1879 1880 1881 1882 1883 1884
		/*
		 * 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
1885
		read_unlock(&tasklist_lock);
M
Miklos Szeredi 已提交
1886
		preempt_enable_no_resched();
1887
		freezable_schedule();
L
Linus Torvalds 已提交
1888 1889 1890
	} else {
		/*
		 * By the time we got the lock, our tracer went away.
1891
		 * Don't drop the lock yet, another tracer may come.
1892 1893 1894
		 *
		 * If @gstop_done, the ptracer went away between group stop
		 * completion and here.  During detach, it would have set
1895 1896 1897
		 * 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
1898
		 */
1899 1900 1901
		if (gstop_done)
			do_notify_parent_cldstop(current, false, why);

1902
		/* tasklist protects us from ptrace_freeze_traced() */
1903
		__set_current_state(TASK_RUNNING);
1904 1905
		if (clear_code)
			current->exit_code = 0;
1906
		read_unlock(&tasklist_lock);
L
Linus Torvalds 已提交
1907 1908 1909 1910 1911 1912 1913 1914 1915 1916
	}

	/*
	 * 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
1917 1918 1919
	/* LISTENING can be set only during STOP traps, clear it */
	current->jobctl &= ~JOBCTL_LISTENING;

L
Linus Torvalds 已提交
1920 1921 1922
	/*
	 * Queued signals ignored us while we were stopped for tracing.
	 * So check for any that we should take before resuming user mode.
1923
	 * This sets TIF_SIGPENDING, but never clears it.
L
Linus Torvalds 已提交
1924
	 */
1925
	recalc_sigpending_tsk(current);
L
Linus Torvalds 已提交
1926 1927
}

T
Tejun Heo 已提交
1928
static void ptrace_do_notify(int signr, int exit_code, int why)
L
Linus Torvalds 已提交
1929 1930 1931 1932
{
	siginfo_t info;

	memset(&info, 0, sizeof info);
T
Tejun Heo 已提交
1933
	info.si_signo = signr;
L
Linus Torvalds 已提交
1934
	info.si_code = exit_code;
1935
	info.si_pid = task_pid_vnr(current);
1936
	info.si_uid = from_kuid_munged(current_user_ns(), current_uid());
L
Linus Torvalds 已提交
1937 1938

	/* Let the debugger run.  */
T
Tejun Heo 已提交
1939 1940 1941 1942 1943 1944
	ptrace_stop(exit_code, why, 1, &info);
}

void ptrace_notify(int exit_code)
{
	BUG_ON((exit_code & (0x7f | ~0xffff)) != SIGTRAP);
1945 1946
	if (unlikely(current->task_works))
		task_work_run();
T
Tejun Heo 已提交
1947

L
Linus Torvalds 已提交
1948
	spin_lock_irq(&current->sighand->siglock);
T
Tejun Heo 已提交
1949
	ptrace_do_notify(SIGTRAP, exit_code, CLD_TRAPPED);
L
Linus Torvalds 已提交
1950 1951 1952
	spin_unlock_irq(&current->sighand->siglock);
}

1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973
/**
 * 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
1974
 */
1975 1976
static bool do_signal_stop(int signr)
	__releases(&current->sighand->siglock)
L
Linus Torvalds 已提交
1977 1978 1979
{
	struct signal_struct *sig = current->signal;

1980
	if (!(current->jobctl & JOBCTL_STOP_PENDING)) {
1981
		unsigned long gstop = JOBCTL_STOP_PENDING | JOBCTL_STOP_CONSUME;
1982 1983
		struct task_struct *t;

1984 1985
		/* signr will be recorded in task->jobctl for retries */
		WARN_ON_ONCE(signr & ~JOBCTL_STOP_SIGMASK);
1986

1987
		if (!likely(current->jobctl & JOBCTL_STOP_DEQUEUED) ||
1988
		    unlikely(signal_group_exit(sig)))
1989
			return false;
L
Linus Torvalds 已提交
1990
		/*
1991 1992 1993 1994 1995 1996 1997
		 * 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
1998 1999
		 * cause two group stops when !ptraced.  That is why we
		 * also check !task_is_stopped(t) below.
2000 2001 2002 2003 2004 2005 2006 2007
		 *
		 * 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
2008
		 */
2009 2010
		if (!(sig->flags & SIGNAL_STOP_STOPPED))
			sig->group_exit_code = signr;
L
Linus Torvalds 已提交
2011

2012 2013 2014 2015
		sig->group_stop_count = 0;

		if (task_set_jobctl_pending(current, signr | gstop))
			sig->group_stop_count++;
L
Linus Torvalds 已提交
2016

2017 2018
		t = current;
		while_each_thread(current, t) {
L
Linus Torvalds 已提交
2019
			/*
2020 2021 2022
			 * 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
2023
			 */
2024 2025
			if (!task_is_stopped(t) &&
			    task_set_jobctl_pending(t, signr | gstop)) {
2026
				sig->group_stop_count++;
2027 2028 2029 2030
				if (likely(!(t->ptrace & PT_SEIZED)))
					signal_wake_up(t, 0);
				else
					ptrace_trap_notify(t);
2031
			}
2032
		}
L
Linus Torvalds 已提交
2033
	}
2034

T
Tejun Heo 已提交
2035
	if (likely(!current->ptrace)) {
2036
		int notify = 0;
L
Linus Torvalds 已提交
2037

2038 2039 2040 2041 2042 2043 2044 2045
		/*
		 * 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;

2046
		__set_current_state(TASK_STOPPED);
2047 2048
		spin_unlock_irq(&current->sighand->siglock);

2049 2050 2051 2052 2053 2054 2055 2056 2057
		/*
		 * 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.
		 */
2058 2059
		if (notify) {
			read_lock(&tasklist_lock);
2060
			do_notify_parent_cldstop(current, false, notify);
2061 2062 2063 2064
			read_unlock(&tasklist_lock);
		}

		/* Now we don't run again until woken by SIGCONT or SIGKILL */
2065
		freezable_schedule();
2066
		return true;
2067
	} else {
2068 2069 2070 2071 2072 2073
		/*
		 * 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;
2074
	}
2075
}
L
Linus Torvalds 已提交
2076

2077 2078 2079
/**
 * do_jobctl_trap - take care of ptrace jobctl traps
 *
T
Tejun Heo 已提交
2080 2081 2082 2083 2084 2085 2086
 * 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.
2087 2088 2089 2090 2091 2092 2093
 *
 * 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
2094
	struct signal_struct *signal = current->signal;
2095
	int signr = current->jobctl & JOBCTL_STOP_SIGMASK;
2096

T
Tejun Heo 已提交
2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107
	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;
2108
	}
L
Linus Torvalds 已提交
2109 2110
}

2111
static int ptrace_signal(int signr, siginfo_t *info)
R
Roland McGrath 已提交
2112
{
2113 2114 2115 2116 2117 2118 2119 2120 2121 2122
	/*
	 * 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
2123
	ptrace_stop(signr, CLD_TRAPPED, 0, info);
R
Roland McGrath 已提交
2124 2125 2126 2127 2128 2129 2130 2131

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

	current->exit_code = 0;

2132 2133 2134 2135 2136 2137
	/*
	 * 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
2138 2139 2140 2141
	if (signr != info->si_signo) {
		info->si_signo = signr;
		info->si_errno = 0;
		info->si_code = SI_USER;
2142
		rcu_read_lock();
R
Roland McGrath 已提交
2143
		info->si_pid = task_pid_vnr(current->parent);
2144 2145
		info->si_uid = from_kuid_munged(current_user_ns(),
						task_uid(current->parent));
2146
		rcu_read_unlock();
R
Roland McGrath 已提交
2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157
	}

	/* 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;
}

2158
int get_signal(struct ksignal *ksig)
L
Linus Torvalds 已提交
2159
{
2160 2161 2162
	struct sighand_struct *sighand = current->sighand;
	struct signal_struct *signal = current->signal;
	int signr;
L
Linus Torvalds 已提交
2163

2164 2165
	if (unlikely(current->task_works))
		task_work_run();
2166

2167 2168 2169
	if (unlikely(uprobe_deny_signal()))
		return 0;

R
Roland McGrath 已提交
2170
	/*
2171 2172 2173
	 * 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
2174
	 */
2175 2176
	try_to_freeze();

2177
relock:
2178
	spin_lock_irq(&sighand->siglock);
2179 2180 2181 2182 2183
	/*
	 * 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.
	 */
2184
	if (unlikely(signal->flags & SIGNAL_CLD_MASK)) {
2185 2186 2187 2188 2189 2190 2191
		int why;

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

2192
		signal->flags &= ~SIGNAL_CLD_MASK;
2193

2194
		spin_unlock_irq(&sighand->siglock);
R
Roland McGrath 已提交
2195

2196 2197 2198 2199 2200 2201 2202 2203
		/*
		 * 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
2204
		read_lock(&tasklist_lock);
2205 2206
		do_notify_parent_cldstop(current, false, why);

2207 2208 2209
		if (ptrace_reparented(current->group_leader))
			do_notify_parent_cldstop(current->group_leader,
						true, why);
T
Tejun Heo 已提交
2210
		read_unlock(&tasklist_lock);
2211

2212 2213 2214
		goto relock;
	}

L
Linus Torvalds 已提交
2215 2216
	for (;;) {
		struct k_sigaction *ka;
2217

T
Tejun Heo 已提交
2218 2219
		if (unlikely(current->jobctl & JOBCTL_STOP_PENDING) &&
		    do_signal_stop(0))
2220
			goto relock;
2221

2222 2223 2224 2225 2226
		if (unlikely(current->jobctl & JOBCTL_TRAP_MASK)) {
			do_jobctl_trap();
			spin_unlock_irq(&sighand->siglock);
			goto relock;
		}
L
Linus Torvalds 已提交
2227

2228
		signr = dequeue_signal(current, &current->blocked, &ksig->info);
2229

T
Tejun Heo 已提交
2230 2231
		if (!signr)
			break; /* will return 0 */
2232

2233
		if (unlikely(current->ptrace) && signr != SIGKILL) {
2234
			signr = ptrace_signal(signr, &ksig->info);
T
Tejun Heo 已提交
2235 2236
			if (!signr)
				continue;
L
Linus Torvalds 已提交
2237 2238
		}

T
Tejun Heo 已提交
2239 2240
		ka = &sighand->action[signr-1];

2241
		/* Trace actually delivered signals. */
2242
		trace_signal_deliver(signr, &ksig->info, ka);
2243

L
Linus Torvalds 已提交
2244 2245 2246 2247
		if (ka->sa.sa_handler == SIG_IGN) /* Do nothing.  */
			continue;
		if (ka->sa.sa_handler != SIG_DFL) {
			/* Run the handler.  */
2248
			ksig->ka = *ka;
L
Linus Torvalds 已提交
2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261

			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;

2262
		/*
2263
		 * Global init gets no signals it doesn't want.
2264 2265 2266 2267 2268 2269 2270
		 * 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.
2271
		 */
2272
		if (unlikely(signal->flags & SIGNAL_UNKILLABLE) &&
2273
				!sig_kernel_only(signr))
L
Linus Torvalds 已提交
2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287
			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) {
2288
				spin_unlock_irq(&sighand->siglock);
L
Linus Torvalds 已提交
2289 2290 2291

				/* signals can be posted during this window */

2292
				if (is_current_pgrp_orphaned())
L
Linus Torvalds 已提交
2293 2294
					goto relock;

2295
				spin_lock_irq(&sighand->siglock);
L
Linus Torvalds 已提交
2296 2297
			}

2298
			if (likely(do_signal_stop(ksig->info.si_signo))) {
L
Linus Torvalds 已提交
2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309
				/* It released the siglock.  */
				goto relock;
			}

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

2310
		spin_unlock_irq(&sighand->siglock);
L
Linus Torvalds 已提交
2311 2312 2313 2314 2315

		/*
		 * Anything else is fatal, maybe with a core dump.
		 */
		current->flags |= PF_SIGNALED;
2316

L
Linus Torvalds 已提交
2317
		if (sig_kernel_coredump(signr)) {
2318
			if (print_fatal_signals)
2319
				print_fatal_signal(ksig->info.si_signo);
2320
			proc_coredump_connector(current);
L
Linus Torvalds 已提交
2321 2322 2323 2324 2325 2326 2327 2328
			/*
			 * 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.
			 */
2329
			do_coredump(&ksig->info);
L
Linus Torvalds 已提交
2330 2331 2332 2333 2334
		}

		/*
		 * Death signals, no core dump.
		 */
2335
		do_group_exit(ksig->info.si_signo);
L
Linus Torvalds 已提交
2336 2337
		/* NOTREACHED */
	}
2338
	spin_unlock_irq(&sighand->siglock);
2339 2340 2341

	ksig->sig = signr;
	return ksig->sig > 0;
L
Linus Torvalds 已提交
2342 2343
}

2344
/**
A
Al Viro 已提交
2345
 * signal_delivered - 
R
Richard Weinberger 已提交
2346
 * @ksig:		kernel signal struct
A
Al Viro 已提交
2347
 * @stepping:		nonzero if debugger single-step or block-step in use
2348
 *
2349
 * This function should be called when a signal has successfully been
R
Richard Weinberger 已提交
2350
 * delivered. It updates the blocked signals accordingly (@ksig->ka.sa.sa_mask
A
Al Viro 已提交
2351
 * is always blocked, and the signal itself is blocked unless %SA_NODEFER
R
Richard Weinberger 已提交
2352
 * is set in @ksig->ka.sa.sa_flags.  Tracing is notified.
2353
 */
R
Richard Weinberger 已提交
2354
static void signal_delivered(struct ksignal *ksig, int stepping)
2355 2356 2357
{
	sigset_t blocked;

2358 2359 2360 2361 2362 2363
	/* 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
2364 2365 2366
	sigorsets(&blocked, &current->blocked, &ksig->ka.sa.sa_mask);
	if (!(ksig->ka.sa.sa_flags & SA_NODEFER))
		sigaddset(&blocked, ksig->sig);
2367
	set_current_blocked(&blocked);
2368
	tracehook_signal_handler(stepping);
2369 2370
}

A
Al Viro 已提交
2371 2372 2373 2374 2375
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
2376
		signal_delivered(ksig, stepping);
A
Al Viro 已提交
2377 2378
}

2379 2380
/*
 * It could be that complete_signal() picked us to notify about the
2381 2382
 * group-wide signal. Other threads should be notified now to take
 * the shared signals in @which since we will not.
2383
 */
2384
static void retarget_shared_pending(struct task_struct *tsk, sigset_t *which)
2385
{
2386
	sigset_t retarget;
2387 2388
	struct task_struct *t;

2389 2390 2391 2392
	sigandsets(&retarget, &tsk->signal->shared_pending.signal, which);
	if (sigisemptyset(&retarget))
		return;

2393 2394
	t = tsk;
	while_each_thread(tsk, t) {
2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407
		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;
2408 2409 2410
	}
}

O
Oleg Nesterov 已提交
2411 2412 2413
void exit_signals(struct task_struct *tsk)
{
	int group_stop = 0;
2414
	sigset_t unblocked;
O
Oleg Nesterov 已提交
2415

2416 2417 2418 2419
	/*
	 * @tsk is about to have PF_EXITING set - lock out users which
	 * expect stable threadgroup.
	 */
2420
	cgroup_threadgroup_change_begin(tsk);
2421

2422 2423
	if (thread_group_empty(tsk) || signal_group_exit(tsk->signal)) {
		tsk->flags |= PF_EXITING;
2424
		cgroup_threadgroup_change_end(tsk);
2425
		return;
O
Oleg Nesterov 已提交
2426 2427
	}

2428
	spin_lock_irq(&tsk->sighand->siglock);
O
Oleg Nesterov 已提交
2429 2430 2431 2432 2433
	/*
	 * From now this task is not visible for group-wide signals,
	 * see wants_signal(), do_signal_stop().
	 */
	tsk->flags |= PF_EXITING;
2434

2435
	cgroup_threadgroup_change_end(tsk);
2436

2437 2438 2439
	if (!signal_pending(tsk))
		goto out;

2440 2441 2442
	unblocked = tsk->blocked;
	signotset(&unblocked);
	retarget_shared_pending(tsk, &unblocked);
2443

2444
	if (unlikely(tsk->jobctl & JOBCTL_STOP_PENDING) &&
2445
	    task_participate_group_stop(tsk))
T
Tejun Heo 已提交
2446
		group_stop = CLD_STOPPED;
2447
out:
O
Oleg Nesterov 已提交
2448 2449
	spin_unlock_irq(&tsk->sighand->siglock);

2450 2451 2452 2453
	/*
	 * If group stop has completed, deliver the notification.  This
	 * should always go to the real parent of the group leader.
	 */
2454
	if (unlikely(group_stop)) {
O
Oleg Nesterov 已提交
2455
		read_lock(&tasklist_lock);
2456
		do_notify_parent_cldstop(tsk, false, group_stop);
O
Oleg Nesterov 已提交
2457 2458 2459 2460
		read_unlock(&tasklist_lock);
	}
}

L
Linus Torvalds 已提交
2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472
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.
 */

2473 2474 2475
/**
 *  sys_restart_syscall - restart a system call
 */
2476
SYSCALL_DEFINE0(restart_syscall)
L
Linus Torvalds 已提交
2477
{
2478
	struct restart_block *restart = &current->restart_block;
L
Linus Torvalds 已提交
2479 2480 2481 2482 2483 2484 2485 2486
	return restart->fn(restart);
}

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

2487 2488 2489 2490 2491
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. */
2492
		sigandnsets(&newblocked, newset, &current->blocked);
2493 2494 2495 2496 2497 2498
		retarget_shared_pending(tsk, &newblocked);
	}
	tsk->blocked = *newset;
	recalc_sigpending();
}

2499 2500 2501 2502 2503 2504
/**
 * 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
2505
 */
2506 2507 2508
void set_current_blocked(sigset_t *newset)
{
	sigdelsetmask(newset, sigmask(SIGKILL) | sigmask(SIGSTOP));
2509
	__set_current_blocked(newset);
2510 2511 2512
}

void __set_current_blocked(const sigset_t *newset)
2513 2514 2515
{
	struct task_struct *tsk = current;

2516 2517 2518 2519 2520 2521 2522
	/*
	 * 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;

2523
	spin_lock_irq(&tsk->sighand->siglock);
2524
	__set_task_blocked(tsk, newset);
2525 2526
	spin_unlock_irq(&tsk->sighand->siglock);
}
L
Linus Torvalds 已提交
2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537

/*
 * 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)
{
2538 2539
	struct task_struct *tsk = current;
	sigset_t newset;
L
Linus Torvalds 已提交
2540

2541
	/* Lockless, only current can change ->blocked, never from irq */
2542
	if (oldset)
2543
		*oldset = tsk->blocked;
2544

L
Linus Torvalds 已提交
2545 2546
	switch (how) {
	case SIG_BLOCK:
2547
		sigorsets(&newset, &tsk->blocked, set);
L
Linus Torvalds 已提交
2548 2549
		break;
	case SIG_UNBLOCK:
2550
		sigandnsets(&newset, &tsk->blocked, set);
L
Linus Torvalds 已提交
2551 2552
		break;
	case SIG_SETMASK:
2553
		newset = *set;
L
Linus Torvalds 已提交
2554 2555
		break;
	default:
2556
		return -EINVAL;
L
Linus Torvalds 已提交
2557
	}
2558

2559
	__set_current_blocked(&newset);
2560
	return 0;
L
Linus Torvalds 已提交
2561 2562
}

2563 2564 2565
/**
 *  sys_rt_sigprocmask - change the list of currently blocked signals
 *  @how: whether to add, remove, or set signals
R
Randy Dunlap 已提交
2566
 *  @nset: stores pending signals
2567 2568 2569
 *  @oset: previous value of signal mask if non-null
 *  @sigsetsize: size of sigset_t type
 */
2570
SYSCALL_DEFINE4(rt_sigprocmask, int, how, sigset_t __user *, nset,
2571
		sigset_t __user *, oset, size_t, sigsetsize)
L
Linus Torvalds 已提交
2572 2573
{
	sigset_t old_set, new_set;
2574
	int error;
L
Linus Torvalds 已提交
2575 2576 2577

	/* XXX: Don't preclude handling different sized sigset_t's.  */
	if (sigsetsize != sizeof(sigset_t))
2578
		return -EINVAL;
L
Linus Torvalds 已提交
2579

2580 2581 2582 2583 2584
	old_set = current->blocked;

	if (nset) {
		if (copy_from_user(&new_set, nset, sizeof(sigset_t)))
			return -EFAULT;
L
Linus Torvalds 已提交
2585 2586
		sigdelsetmask(&new_set, sigmask(SIGKILL)|sigmask(SIGSTOP));

2587
		error = sigprocmask(how, &new_set, NULL);
L
Linus Torvalds 已提交
2588
		if (error)
2589 2590
			return error;
	}
L
Linus Torvalds 已提交
2591

2592 2593 2594
	if (oset) {
		if (copy_to_user(oset, &old_set, sizeof(sigset_t)))
			return -EFAULT;
L
Linus Torvalds 已提交
2595
	}
2596 2597

	return 0;
L
Linus Torvalds 已提交
2598 2599
}

A
Al Viro 已提交
2600 2601 2602
#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
2603
{
A
Al Viro 已提交
2604 2605 2606 2607 2608 2609 2610 2611 2612
	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) {
		sigset_t new_set;
		int error;
A
Al Viro 已提交
2613
		if (get_compat_sigset(&new_set, nset))
A
Al Viro 已提交
2614 2615 2616 2617 2618 2619 2620
			return -EFAULT;
		sigdelsetmask(&new_set, sigmask(SIGKILL)|sigmask(SIGSTOP));

		error = sigprocmask(how, &new_set, NULL);
		if (error)
			return error;
	}
2621
	return oset ? put_compat_sigset(oset, &old_set, sizeof(*oset)) : 0;
A
Al Viro 已提交
2622 2623
}
#endif
L
Linus Torvalds 已提交
2624

2625
static int do_sigpending(sigset_t *set)
L
Linus Torvalds 已提交
2626 2627
{
	spin_lock_irq(&current->sighand->siglock);
A
Al Viro 已提交
2628
	sigorsets(set, &current->pending.signal,
L
Linus Torvalds 已提交
2629 2630 2631 2632
		  &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
2633 2634
	sigandsets(set, &current->blocked, set);
	return 0;
2635
}
L
Linus Torvalds 已提交
2636

2637 2638 2639
/**
 *  sys_rt_sigpending - examine a pending signal that has been raised
 *			while blocked
2640
 *  @uset: stores pending signals
2641 2642
 *  @sigsetsize: size of sigset_t type or larger
 */
A
Al Viro 已提交
2643
SYSCALL_DEFINE2(rt_sigpending, sigset_t __user *, uset, size_t, sigsetsize)
L
Linus Torvalds 已提交
2644
{
A
Al Viro 已提交
2645
	sigset_t set;
2646 2647 2648 2649 2650 2651
	int err;

	if (sigsetsize > sizeof(*uset))
		return -EINVAL;

	err = do_sigpending(&set);
A
Al Viro 已提交
2652 2653 2654 2655 2656 2657 2658 2659
	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
2660
{
A
Al Viro 已提交
2661
	sigset_t set;
2662 2663 2664 2665 2666 2667
	int err;

	if (sigsetsize > sizeof(*uset))
		return -EINVAL;

	err = do_sigpending(&set);
2668 2669
	if (!err)
		err = put_compat_sigset(uset, &set, sigsetsize);
A
Al Viro 已提交
2670
	return err;
L
Linus Torvalds 已提交
2671
}
A
Al Viro 已提交
2672
#endif
L
Linus Torvalds 已提交
2673

2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685
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 },
2686
#if defined(SIGEMT) && defined(NSIGEMT)
2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718
			[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
2719 2720
#ifndef HAVE_ARCH_COPY_SIGINFO_TO_USER

A
Al Viro 已提交
2721
int copy_siginfo_to_user(siginfo_t __user *to, const siginfo_t *from)
L
Linus Torvalds 已提交
2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732
{
	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.
2733 2734
	 * Please remember to update the signalfd_copyinfo() function
	 * inside fs/signalfd.c too, in case siginfo_t changes.
L
Linus Torvalds 已提交
2735 2736 2737 2738 2739 2740
	 * 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);
2741 2742 2743
	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
2744 2745 2746
		err |= __put_user(from->si_pid, &to->si_pid);
		err |= __put_user(from->si_uid, &to->si_uid);
		break;
2747 2748
	case SIL_TIMER:
		/* Unreached SI_TIMER is negative */
L
Linus Torvalds 已提交
2749
		break;
2750
	case SIL_POLL:
L
Linus Torvalds 已提交
2751 2752 2753
		err |= __put_user(from->si_band, &to->si_band);
		err |= __put_user(from->si_fd, &to->si_fd);
		break;
2754
	case SIL_FAULT:
L
Linus Torvalds 已提交
2755 2756 2757
		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
2758 2759
#endif
#ifdef BUS_MCEERR_AO
2760
		/*
A
Andi Kleen 已提交
2761
		 * Other callers might not initialize the si_lsb field,
2762
		 * so check explicitly for the right codes here.
A
Andi Kleen 已提交
2763
		 */
2764 2765
		if (from->si_signo == SIGBUS &&
		    (from->si_code == BUS_MCEERR_AR || from->si_code == BUS_MCEERR_AO))
A
Andi Kleen 已提交
2766
			err |= __put_user(from->si_addr_lsb, &to->si_addr_lsb);
2767 2768
#endif
#ifdef SEGV_BNDERR
2769 2770 2771 2772
		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);
		}
2773 2774 2775 2776
#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
2777 2778
#endif
		break;
2779
	case SIL_CHLD:
L
Linus Torvalds 已提交
2780 2781 2782 2783 2784 2785
		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;
2786
	case SIL_RT:
L
Linus Torvalds 已提交
2787 2788 2789 2790
		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;
2791
#ifdef __ARCH_SIGSYS
2792
	case SIL_SYS:
2793 2794 2795 2796 2797
		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
2798 2799 2800 2801 2802 2803
	}
	return err;
}

#endif

2804 2805 2806 2807 2808 2809
/**
 *  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
 */
2810
static int do_sigtimedwait(const sigset_t *which, siginfo_t *info,
2811
		    const struct timespec *ts)
2812
{
T
Thomas Gleixner 已提交
2813
	ktime_t *to = NULL, timeout = KTIME_MAX;
2814 2815
	struct task_struct *tsk = current;
	sigset_t mask = *which;
2816
	int sig, ret = 0;
2817 2818 2819 2820

	if (ts) {
		if (!timespec_valid(ts))
			return -EINVAL;
2821 2822
		timeout = timespec_to_ktime(*ts);
		to = &timeout;
2823 2824 2825 2826 2827 2828 2829 2830 2831 2832
	}

	/*
	 * 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
2833
	if (!sig && timeout) {
2834 2835 2836
		/*
		 * None ready, temporarily unblock those we're interested
		 * while we are sleeping in so that we'll be awakened when
2837 2838
		 * they arrive. Unblocking is always fine, we can avoid
		 * set_current_blocked().
2839 2840 2841 2842 2843 2844
		 */
		tsk->real_blocked = tsk->blocked;
		sigandsets(&tsk->blocked, &tsk->blocked, &mask);
		recalc_sigpending();
		spin_unlock_irq(&tsk->sighand->siglock);

2845 2846 2847
		__set_current_state(TASK_INTERRUPTIBLE);
		ret = freezable_schedule_hrtimeout_range(to, tsk->timer_slack_ns,
							 HRTIMER_MODE_REL);
2848
		spin_lock_irq(&tsk->sighand->siglock);
2849
		__set_task_blocked(tsk, &tsk->real_blocked);
2850
		sigemptyset(&tsk->real_blocked);
2851
		sig = dequeue_signal(tsk, &mask, info);
2852 2853 2854 2855 2856
	}
	spin_unlock_irq(&tsk->sighand->siglock);

	if (sig)
		return sig;
2857
	return ret ? -EINTR : -EAGAIN;
2858 2859
}

2860 2861 2862 2863 2864 2865 2866 2867
/**
 *  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
 */
2868 2869 2870
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
2871 2872 2873 2874
{
	sigset_t these;
	struct timespec ts;
	siginfo_t info;
2875
	int ret;
L
Linus Torvalds 已提交
2876 2877 2878 2879 2880 2881 2882

	/* 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;
2883

L
Linus Torvalds 已提交
2884 2885 2886 2887 2888
	if (uts) {
		if (copy_from_user(&ts, uts, sizeof(ts)))
			return -EFAULT;
	}

2889
	ret = do_sigtimedwait(&these, &info, uts ? &ts : NULL);
L
Linus Torvalds 已提交
2890

2891 2892 2893
	if (ret > 0 && uinfo) {
		if (copy_siginfo_to_user(uinfo, &info))
			ret = -EFAULT;
L
Linus Torvalds 已提交
2894 2895 2896 2897 2898
	}

	return ret;
}

2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911
#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)
{
	sigset_t s;
	struct timespec t;
	siginfo_t info;
	long ret;

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

A
Al Viro 已提交
2912
	if (get_compat_sigset(&s, uthese))
2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930
		return -EFAULT;

	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

2931 2932 2933 2934 2935
/**
 *  sys_kill - send a signal to a process
 *  @pid: the PID of the process
 *  @sig: signal to be sent
 */
2936
SYSCALL_DEFINE2(kill, pid_t, pid, int, sig)
L
Linus Torvalds 已提交
2937 2938 2939 2940 2941 2942
{
	struct siginfo info;

	info.si_signo = sig;
	info.si_errno = 0;
	info.si_code = SI_USER;
2943
	info.si_pid = task_tgid_vnr(current);
2944
	info.si_uid = from_kuid_munged(current_user_ns(), current_uid());
L
Linus Torvalds 已提交
2945 2946 2947 2948

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

T
Thomas Gleixner 已提交
2949 2950
static int
do_send_specific(pid_t tgid, pid_t pid, int sig, struct siginfo *info)
L
Linus Torvalds 已提交
2951 2952
{
	struct task_struct *p;
T
Thomas Gleixner 已提交
2953
	int error = -ESRCH;
L
Linus Torvalds 已提交
2954

2955
	rcu_read_lock();
2956
	p = find_task_by_vpid(pid);
2957
	if (p && (tgid <= 0 || task_tgid_vnr(p) == tgid)) {
T
Thomas Gleixner 已提交
2958
		error = check_kill_permission(sig, info, p);
L
Linus Torvalds 已提交
2959 2960 2961 2962
		/*
		 * The null signal is a permissions and process existence
		 * probe.  No signal is actually delivered.
		 */
2963 2964 2965 2966 2967 2968 2969 2970 2971
		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
2972 2973
		}
	}
2974
	rcu_read_unlock();
2975

L
Linus Torvalds 已提交
2976 2977 2978
	return error;
}

T
Thomas Gleixner 已提交
2979 2980
static int do_tkill(pid_t tgid, pid_t pid, int sig)
{
2981
	struct siginfo info = {};
T
Thomas Gleixner 已提交
2982 2983 2984 2985 2986

	info.si_signo = sig;
	info.si_errno = 0;
	info.si_code = SI_TKILL;
	info.si_pid = task_tgid_vnr(current);
2987
	info.si_uid = from_kuid_munged(current_user_ns(), current_uid());
T
Thomas Gleixner 已提交
2988 2989 2990 2991

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

2992 2993 2994 2995 2996 2997
/**
 *  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
 *
2998
 *  This syscall also checks the @tgid and returns -ESRCH even if the PID
2999 3000 3001
 *  exists but it's not belonging to the target process anymore. This
 *  method solves the problem of threads exiting and PIDs getting reused.
 */
3002
SYSCALL_DEFINE3(tgkill, pid_t, tgid, pid_t, pid, int, sig)
3003 3004 3005 3006 3007 3008 3009 3010
{
	/* This is only valid for single tasks */
	if (pid <= 0 || tgid <= 0)
		return -EINVAL;

	return do_tkill(tgid, pid, sig);
}

3011 3012 3013 3014 3015
/**
 *  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
3016 3017
 *  Send a signal to only one task, even if it's a CLONE_THREAD task.
 */
3018
SYSCALL_DEFINE2(tkill, pid_t, pid, int, sig)
L
Linus Torvalds 已提交
3019 3020 3021 3022 3023
{
	/* This is only valid for single tasks */
	if (pid <= 0)
		return -EINVAL;

3024
	return do_tkill(0, pid, sig);
L
Linus Torvalds 已提交
3025 3026
}

A
Al Viro 已提交
3027 3028 3029 3030 3031
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.
	 */
3032
	if ((info->si_code >= 0 || info->si_code == SI_TKILL) &&
3033
	    (task_pid_vnr(current) != pid))
A
Al Viro 已提交
3034
		return -EPERM;
3035

A
Al Viro 已提交
3036 3037 3038 3039 3040 3041
	info->si_signo = sig;

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

3042 3043 3044 3045 3046 3047
/**
 *  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
 */
3048 3049
SYSCALL_DEFINE3(rt_sigqueueinfo, pid_t, pid, int, sig,
		siginfo_t __user *, uinfo)
L
Linus Torvalds 已提交
3050 3051 3052 3053
{
	siginfo_t info;
	if (copy_from_user(&info, uinfo, sizeof(siginfo_t)))
		return -EFAULT;
A
Al Viro 已提交
3054 3055
	return do_rt_sigqueueinfo(pid, sig, &info);
}
L
Linus Torvalds 已提交
3056

A
Al Viro 已提交
3057 3058 3059 3060 3061 3062
#ifdef CONFIG_COMPAT
COMPAT_SYSCALL_DEFINE3(rt_sigqueueinfo,
			compat_pid_t, pid,
			int, sig,
			struct compat_siginfo __user *, uinfo)
{
3063
	siginfo_t info = {};
A
Al Viro 已提交
3064 3065 3066 3067
	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
3068
}
A
Al Viro 已提交
3069
#endif
L
Linus Torvalds 已提交
3070

3071
static int do_rt_tgsigqueueinfo(pid_t tgid, pid_t pid, int sig, siginfo_t *info)
3072 3073 3074 3075 3076 3077
{
	/* 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.
3078 3079
	 * Nor can they impersonate a kill()/tgkill(), which adds source info.
	 */
3080 3081
	if ((info->si_code >= 0 || info->si_code == SI_TKILL) &&
	    (task_pid_vnr(current) != pid))
3082
		return -EPERM;
3083

3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099
	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);
}

3100 3101 3102 3103 3104 3105 3106
#ifdef CONFIG_COMPAT
COMPAT_SYSCALL_DEFINE4(rt_tgsigqueueinfo,
			compat_pid_t, tgid,
			compat_pid_t, pid,
			int, sig,
			struct compat_siginfo __user *, uinfo)
{
3107
	siginfo_t info = {};
3108 3109 3110 3111 3112 3113 3114

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

3115
/*
3116
 * For kthreads only, must not be used if cloned with CLONE_SIGHAND
3117
 */
3118
void kernel_sigaction(int sig, __sighandler_t action)
3119
{
3120
	spin_lock_irq(&current->sighand->siglock);
3121 3122 3123
	current->sighand->action[sig - 1].sa.sa_handler = action;
	if (action == SIG_IGN) {
		sigset_t mask;
3124

3125 3126
		sigemptyset(&mask);
		sigaddset(&mask, sig);
3127

3128 3129 3130 3131
		flush_sigqueue_mask(&mask, &current->signal->shared_pending);
		flush_sigqueue_mask(&mask, &current->pending);
		recalc_sigpending();
	}
3132 3133
	spin_unlock_irq(&current->sighand->siglock);
}
3134
EXPORT_SYMBOL(kernel_sigaction);
3135

3136 3137 3138 3139 3140
void __weak sigaction_compat_abi(struct k_sigaction *act,
		struct k_sigaction *oact)
{
}

3141
int do_sigaction(int sig, struct k_sigaction *act, struct k_sigaction *oact)
L
Linus Torvalds 已提交
3142
{
3143
	struct task_struct *p = current, *t;
L
Linus Torvalds 已提交
3144
	struct k_sigaction *k;
3145
	sigset_t mask;
L
Linus Torvalds 已提交
3146

3147
	if (!valid_signal(sig) || sig < 1 || (act && sig_kernel_only(sig)))
L
Linus Torvalds 已提交
3148 3149
		return -EINVAL;

3150
	k = &p->sighand->action[sig-1];
L
Linus Torvalds 已提交
3151

3152
	spin_lock_irq(&p->sighand->siglock);
L
Linus Torvalds 已提交
3153 3154 3155
	if (oact)
		*oact = *k;

3156 3157
	sigaction_compat_abi(act, oact);

L
Linus Torvalds 已提交
3158
	if (act) {
3159 3160
		sigdelsetmask(&act->sa.sa_mask,
			      sigmask(SIGKILL) | sigmask(SIGSTOP));
3161
		*k = *act;
L
Linus Torvalds 已提交
3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172
		/*
		 * 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"
		 */
3173
		if (sig_handler_ignored(sig_handler(p, sig), sig)) {
3174 3175
			sigemptyset(&mask);
			sigaddset(&mask, sig);
3176 3177
			flush_sigqueue_mask(&mask, &p->signal->shared_pending);
			for_each_thread(p, t)
3178
				flush_sigqueue_mask(&mask, &t->pending);
L
Linus Torvalds 已提交
3179 3180 3181
		}
	}

3182
	spin_unlock_irq(&p->sighand->siglock);
L
Linus Torvalds 已提交
3183 3184 3185
	return 0;
}

3186
static int
3187
do_sigaltstack (const stack_t *ss, stack_t *oss, unsigned long sp)
L
Linus Torvalds 已提交
3188
{
3189
	struct task_struct *t = current;
L
Linus Torvalds 已提交
3190

3191 3192 3193 3194 3195 3196 3197
	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
3198

3199 3200 3201 3202
	if (ss) {
		void __user *ss_sp = ss->ss_sp;
		size_t ss_size = ss->ss_size;
		unsigned ss_flags = ss->ss_flags;
3203
		int ss_mode;
L
Linus Torvalds 已提交
3204

3205 3206
		if (unlikely(on_sig_stack(sp)))
			return -EPERM;
L
Linus Torvalds 已提交
3207

3208
		ss_mode = ss_flags & ~SS_FLAG_BITS;
3209 3210 3211
		if (unlikely(ss_mode != SS_DISABLE && ss_mode != SS_ONSTACK &&
				ss_mode != 0))
			return -EINVAL;
L
Linus Torvalds 已提交
3212

3213
		if (ss_mode == SS_DISABLE) {
L
Linus Torvalds 已提交
3214 3215 3216
			ss_size = 0;
			ss_sp = NULL;
		} else {
3217 3218
			if (unlikely(ss_size < MINSIGSTKSZ))
				return -ENOMEM;
L
Linus Torvalds 已提交
3219 3220
		}

3221 3222 3223
		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
3224
	}
3225
	return 0;
L
Linus Torvalds 已提交
3226
}
3227

3228 3229
SYSCALL_DEFINE2(sigaltstack,const stack_t __user *,uss, stack_t __user *,uoss)
{
3230 3231 3232 3233 3234 3235 3236 3237 3238
	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;
3239
}
L
Linus Torvalds 已提交
3240

A
Al Viro 已提交
3241 3242
int restore_altstack(const stack_t __user *uss)
{
3243 3244 3245 3246
	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
3247
	/* squash all but EFAULT for now */
3248
	return 0;
A
Al Viro 已提交
3249 3250
}

3251 3252 3253
int __save_altstack(stack_t __user *uss, unsigned long sp)
{
	struct task_struct *t = current;
3254 3255
	int err = __put_user((void __user *)t->sas_ss_sp, &uss->ss_sp) |
		__put_user(t->sas_ss_flags, &uss->ss_flags) |
3256
		__put_user(t->sas_ss_size, &uss->ss_size);
3257 3258 3259 3260 3261
	if (err)
		return err;
	if (t->sas_ss_flags & SS_AUTODISARM)
		sas_ss_reset(t);
	return 0;
3262 3263
}

A
Al Viro 已提交
3264
#ifdef CONFIG_COMPAT
3265 3266 3267
COMPAT_SYSCALL_DEFINE2(sigaltstack,
			const compat_stack_t __user *, uss_ptr,
			compat_stack_t __user *, uoss_ptr)
A
Al Viro 已提交
3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279
{
	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;
	}
3280
	ret = do_sigaltstack(uss_ptr ? &uss : NULL, &uoss,
A
Al Viro 已提交
3281 3282
			     compat_user_stack_pointer());
	if (ret >= 0 && uoss_ptr)  {
3283 3284 3285 3286 3287 3288
		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
3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299
			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;
}
3300 3301 3302

int __compat_save_altstack(compat_stack_t __user *uss, unsigned long sp)
{
3303
	int err;
3304
	struct task_struct *t = current;
3305 3306 3307
	err = __put_user(ptr_to_compat((void __user *)t->sas_ss_sp),
			 &uss->ss_sp) |
		__put_user(t->sas_ss_flags, &uss->ss_flags) |
3308
		__put_user(t->sas_ss_size, &uss->ss_size);
3309 3310 3311 3312 3313
	if (err)
		return err;
	if (t->sas_ss_flags & SS_AUTODISARM)
		sas_ss_reset(t);
	return 0;
3314
}
A
Al Viro 已提交
3315
#endif
L
Linus Torvalds 已提交
3316 3317 3318

#ifdef __ARCH_WANT_SYS_SIGPENDING

3319 3320 3321 3322
/**
 *  sys_sigpending - examine pending signals
 *  @set: where mask of pending signal is returned
 */
3323
SYSCALL_DEFINE1(sigpending, old_sigset_t __user *, set)
L
Linus Torvalds 已提交
3324
{
A
Al Viro 已提交
3325
	return sys_rt_sigpending((sigset_t __user *)set, sizeof(old_sigset_t)); 
L
Linus Torvalds 已提交
3326 3327
}

A
Al Viro 已提交
3328 3329 3330 3331
#ifdef CONFIG_COMPAT
COMPAT_SYSCALL_DEFINE1(sigpending, compat_old_sigset_t __user *, set32)
{
	sigset_t set;
3332
	int err = do_sigpending(&set);
3333 3334
	if (!err)
		err = put_user(set.sig[0], set32);
A
Al Viro 已提交
3335 3336 3337 3338
	return err;
}
#endif

L
Linus Torvalds 已提交
3339 3340 3341
#endif

#ifdef __ARCH_WANT_SYS_SIGPROCMASK
3342 3343 3344
/**
 *  sys_sigprocmask - examine and change blocked signals
 *  @how: whether to add, remove, or set signals
3345
 *  @nset: signals to add or remove (if non-null)
3346 3347
 *  @oset: previous value of signal mask if non-null
 *
3348 3349 3350
 * Some platforms have their own version with special arguments;
 * others support only sys_rt_sigprocmask.
 */
L
Linus Torvalds 已提交
3351

3352
SYSCALL_DEFINE3(sigprocmask, int, how, old_sigset_t __user *, nset,
3353
		old_sigset_t __user *, oset)
L
Linus Torvalds 已提交
3354 3355
{
	old_sigset_t old_set, new_set;
3356
	sigset_t new_blocked;
L
Linus Torvalds 已提交
3357

3358
	old_set = current->blocked.sig[0];
L
Linus Torvalds 已提交
3359

3360 3361 3362
	if (nset) {
		if (copy_from_user(&new_set, nset, sizeof(*nset)))
			return -EFAULT;
L
Linus Torvalds 已提交
3363

3364
		new_blocked = current->blocked;
L
Linus Torvalds 已提交
3365 3366 3367

		switch (how) {
		case SIG_BLOCK:
3368
			sigaddsetmask(&new_blocked, new_set);
L
Linus Torvalds 已提交
3369 3370
			break;
		case SIG_UNBLOCK:
3371
			sigdelsetmask(&new_blocked, new_set);
L
Linus Torvalds 已提交
3372 3373
			break;
		case SIG_SETMASK:
3374
			new_blocked.sig[0] = new_set;
L
Linus Torvalds 已提交
3375
			break;
3376 3377
		default:
			return -EINVAL;
L
Linus Torvalds 已提交
3378 3379
		}

3380
		set_current_blocked(&new_blocked);
3381 3382 3383
	}

	if (oset) {
L
Linus Torvalds 已提交
3384
		if (copy_to_user(oset, &old_set, sizeof(*oset)))
3385
			return -EFAULT;
L
Linus Torvalds 已提交
3386
	}
3387 3388

	return 0;
L
Linus Torvalds 已提交
3389 3390 3391
}
#endif /* __ARCH_WANT_SYS_SIGPROCMASK */

3392
#ifndef CONFIG_ODD_RT_SIGACTION
3393 3394 3395
/**
 *  sys_rt_sigaction - alter an action taken by a process
 *  @sig: signal to be sent
3396 3397
 *  @act: new sigaction
 *  @oact: used to save the previous sigaction
3398 3399
 *  @sigsetsize: size of sigset_t type
 */
3400 3401 3402 3403
SYSCALL_DEFINE4(rt_sigaction, int, sig,
		const struct sigaction __user *, act,
		struct sigaction __user *, oact,
		size_t, sigsetsize)
L
Linus Torvalds 已提交
3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425
{
	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
3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449
#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;
#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
A
Al Viro 已提交
3450
		ret |= get_compat_sigset(&new_ka.sa.sa_mask, &act->sa_mask);
3451
		ret |= get_user(new_ka.sa.sa_flags, &act->sa_flags);
A
Al Viro 已提交
3452 3453 3454 3455 3456 3457 3458 3459
		if (ret)
			return -EFAULT;
	}

	ret = do_sigaction(sig, act ? &new_ka : NULL, oact ? &old_ka : NULL);
	if (!ret && oact) {
		ret = put_user(ptr_to_compat(old_ka.sa.sa_handler), 
			       &oact->sa_handler);
3460 3461
		ret |= put_compat_sigset(&oact->sa_mask, &old_ka.sa.sa_mask,
					 sizeof(oact->sa_mask));
3462
		ret |= put_user(old_ka.sa.sa_flags, &oact->sa_flags);
A
Al Viro 已提交
3463 3464 3465 3466 3467 3468 3469 3470
#ifdef __ARCH_HAS_SA_RESTORER
		ret |= put_user(ptr_to_compat(old_ka.sa.sa_restorer),
				&oact->sa_restorer);
#endif
	}
	return ret;
}
#endif
3471
#endif /* !CONFIG_ODD_RT_SIGACTION */
L
Linus Torvalds 已提交
3472

3473 3474 3475 3476 3477 3478 3479 3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 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
#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
3550

3551
#ifdef CONFIG_SGETMASK_SYSCALL
L
Linus Torvalds 已提交
3552 3553 3554 3555

/*
 * For backwards compatibility.  Functionality superseded by sigprocmask.
 */
3556
SYSCALL_DEFINE0(sgetmask)
L
Linus Torvalds 已提交
3557 3558 3559 3560 3561
{
	/* SMP safe */
	return current->blocked.sig[0];
}

3562
SYSCALL_DEFINE1(ssetmask, int, newmask)
L
Linus Torvalds 已提交
3563
{
3564 3565
	int old = current->blocked.sig[0];
	sigset_t newset;
L
Linus Torvalds 已提交
3566

3567
	siginitset(&newset, newmask);
3568
	set_current_blocked(&newset);
L
Linus Torvalds 已提交
3569 3570 3571

	return old;
}
3572
#endif /* CONFIG_SGETMASK_SYSCALL */
L
Linus Torvalds 已提交
3573 3574 3575 3576 3577

#ifdef __ARCH_WANT_SYS_SIGNAL
/*
 * For backwards compatibility.  Functionality superseded by sigaction.
 */
3578
SYSCALL_DEFINE2(signal, int, sig, __sighandler_t, handler)
L
Linus Torvalds 已提交
3579 3580 3581 3582 3583 3584
{
	struct k_sigaction new_sa, old_sa;
	int ret;

	new_sa.sa.sa_handler = handler;
	new_sa.sa.sa_flags = SA_ONESHOT | SA_NOMASK;
3585
	sigemptyset(&new_sa.sa.sa_mask);
L
Linus Torvalds 已提交
3586 3587 3588 3589 3590 3591 3592 3593 3594

	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

3595
SYSCALL_DEFINE0(pause)
L
Linus Torvalds 已提交
3596
{
3597
	while (!signal_pending(current)) {
3598
		__set_current_state(TASK_INTERRUPTIBLE);
3599 3600
		schedule();
	}
L
Linus Torvalds 已提交
3601 3602 3603 3604 3605
	return -ERESTARTNOHAND;
}

#endif

3606
static int sigsuspend(sigset_t *set)
A
Al Viro 已提交
3607 3608 3609 3610
{
	current->saved_sigmask = current->blocked;
	set_current_blocked(set);

3611 3612 3613 3614
	while (!signal_pending(current)) {
		__set_current_state(TASK_INTERRUPTIBLE);
		schedule();
	}
A
Al Viro 已提交
3615 3616 3617 3618
	set_restore_sigmask();
	return -ERESTARTNOHAND;
}

3619 3620 3621 3622 3623 3624
/**
 *  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
 */
3625
SYSCALL_DEFINE2(rt_sigsuspend, sigset_t __user *, unewset, size_t, sigsetsize)
3626 3627 3628 3629 3630 3631 3632 3633 3634
{
	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
3635
	return sigsuspend(&newset);
3636
}
A
Al Viro 已提交
3637 3638 3639 3640 3641 3642 3643 3644 3645 3646
 
#ifdef CONFIG_COMPAT
COMPAT_SYSCALL_DEFINE2(rt_sigsuspend, compat_sigset_t __user *, unewset, compat_size_t, sigsetsize)
{
	sigset_t newset;

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

A
Al Viro 已提交
3647
	if (get_compat_sigset(&newset, unewset))
A
Al Viro 已提交
3648 3649 3650 3651
		return -EFAULT;
	return sigsuspend(&newset);
}
#endif
3652

3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 3667 3668
#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
3669

3670
__weak const char *arch_vma_name(struct vm_area_struct *vma)
3671 3672 3673 3674
{
	return NULL;
}

L
Linus Torvalds 已提交
3675 3676
void __init signals_init(void)
{
3677 3678 3679 3680
	/* If this check fails, the __ARCH_SI_PREAMBLE_SIZE value is wrong! */
	BUILD_BUG_ON(__ARCH_SI_PREAMBLE_SIZE
		!= offsetof(struct siginfo, _sifields._pad));

3681
	sigqueue_cachep = KMEM_CACHE(sigqueue, SLAB_PANIC);
L
Linus Torvalds 已提交
3682
}
3683 3684 3685 3686

#ifdef CONFIG_KGDB_KDB
#include <linux/kdb.h>
/*
3687
 * kdb_send_sig - Allows kdb to send signals without exposing
3688 3689 3690 3691
 * signal internals.  This function checks if the required locks are
 * available before calling the main signal code, to avoid kdb
 * deadlocks.
 */
3692
void kdb_send_sig(struct task_struct *t, int sig)
3693 3694
{
	static struct task_struct *kdb_prev_t;
3695
	int new_t, ret;
3696 3697 3698 3699 3700 3701 3702 3703 3704
	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;
	}
	new_t = kdb_prev_t != t;
	kdb_prev_t = t;
	if (t->state != TASK_RUNNING && new_t) {
3705
		spin_unlock(&t->sighand->siglock);
3706 3707 3708 3709 3710 3711 3712 3713
		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;
	}
3714 3715 3716
	ret = send_signal(sig, SEND_SIG_PRIV, t, false);
	spin_unlock(&t->sighand->siglock);
	if (ret)
3717 3718 3719 3720 3721 3722
		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 */