signal.c 86.1 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 16 17 18 19
#include <linux/init.h>
#include <linux/sched.h>
#include <linux/fs.h>
#include <linux/tty.h>
#include <linux/binfmts.h>
20
#include <linux/coredump.h>
L
Linus Torvalds 已提交
21 22 23
#include <linux/security.h>
#include <linux/syscalls.h>
#include <linux/ptrace.h>
24
#include <linux/signal.h>
25
#include <linux/signalfd.h>
26
#include <linux/ratelimit.h>
27
#include <linux/tracehook.h>
28
#include <linux/capability.h>
29
#include <linux/freezer.h>
30 31
#include <linux/pid_namespace.h>
#include <linux/nsproxy.h>
32
#include <linux/user_namespace.h>
33
#include <linux/uprobes.h>
34 35
#define CREATE_TRACE_POINTS
#include <trace/events/signal.h>
36

L
Linus Torvalds 已提交
37 38 39 40
#include <asm/param.h>
#include <asm/uaccess.h>
#include <asm/unistd.h>
#include <asm/siginfo.h>
41
#include <asm/cacheflush.h>
42
#include "audit.h"	/* audit_signal_info() */
L
Linus Torvalds 已提交
43 44 45 46 47

/*
 * SLAB caches for signal bits.
 */

48
static struct kmem_cache *sigqueue_cachep;
L
Linus Torvalds 已提交
49

50 51
int print_fatal_signals __read_mostly;

52
static void __user *sig_handler(struct task_struct *t, int sig)
53
{
54 55
	return t->sighand->action[sig - 1].sa.sa_handler;
}
56

57 58
static int sig_handler_ignored(void __user *handler, int sig)
{
59 60 61 62
	/* Is it explicitly or implicitly ignored? */
	return handler == SIG_IGN ||
		(handler == SIG_DFL && sig_kernel_ignore(sig));
}
L
Linus Torvalds 已提交
63

64
static int sig_task_ignored(struct task_struct *t, int sig, bool force)
L
Linus Torvalds 已提交
65
{
66
	void __user *handler;
L
Linus Torvalds 已提交
67

68 69 70
	handler = sig_handler(t, sig);

	if (unlikely(t->signal->flags & SIGNAL_UNKILLABLE) &&
71
			handler == SIG_DFL && !force)
72 73 74 75 76
		return 1;

	return sig_handler_ignored(handler, sig);
}

77
static int sig_ignored(struct task_struct *t, int sig, bool force)
78
{
L
Linus Torvalds 已提交
79 80 81 82 83
	/*
	 * Blocked signals are never ignored, since the
	 * signal handler may change by the time it is
	 * unblocked.
	 */
84
	if (sigismember(&t->blocked, sig) || sigismember(&t->real_blocked, sig))
L
Linus Torvalds 已提交
85 86
		return 0;

87
	if (!sig_task_ignored(t, sig, force))
88 89 90 91 92
		return 0;

	/*
	 * Tracers may want to know about even ignored signals.
	 */
T
Tejun Heo 已提交
93
	return !t->ptrace;
L
Linus Torvalds 已提交
94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127
}

/*
 * 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
128
static int recalc_sigpending_tsk(struct task_struct *t)
L
Linus Torvalds 已提交
129
{
130
	if ((t->jobctl & JOBCTL_PENDING_MASK) ||
L
Linus Torvalds 已提交
131
	    PENDING(&t->pending, &t->blocked) ||
R
Roland McGrath 已提交
132
	    PENDING(&t->signal->shared_pending, &t->blocked)) {
L
Linus Torvalds 已提交
133
		set_tsk_thread_flag(t, TIF_SIGPENDING);
R
Roland McGrath 已提交
134 135
		return 1;
	}
136 137 138 139 140
	/*
	 * 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
141 142 143 144 145 146 147 148 149 150 151
	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
152 153 154 155
}

void recalc_sigpending(void)
{
T
Tejun Heo 已提交
156
	if (!recalc_sigpending_tsk(current) && !freezing(current))
157 158
		clear_thread_flag(TIF_SIGPENDING);

L
Linus Torvalds 已提交
159 160 161 162
}

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

163 164
#define SYNCHRONOUS_MASK \
	(sigmask(SIGSEGV) | sigmask(SIGBUS) | sigmask(SIGILL) | \
165
	 sigmask(SIGTRAP) | sigmask(SIGFPE) | sigmask(SIGSYS))
166

167
int next_signal(struct sigpending *pending, sigset_t *mask)
L
Linus Torvalds 已提交
168 169 170
{
	unsigned long i, *s, *m, x;
	int sig = 0;
171

L
Linus Torvalds 已提交
172 173
	s = pending->signal.sig;
	m = mask->sig;
174 175 176 177 178 179 180 181 182 183 184 185 186

	/*
	 * 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
187 188
	switch (_NSIG_WORDS) {
	default:
189 190 191 192 193 194 195
		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
196 197
		break;

198 199 200
	case 2:
		x = s[1] &~ m[1];
		if (!x)
L
Linus Torvalds 已提交
201
			break;
202
		sig = ffz(~x) + _NSIG_BPW + 1;
L
Linus Torvalds 已提交
203 204
		break;

205 206
	case 1:
		/* Nothing to do */
L
Linus Torvalds 已提交
207 208
		break;
	}
209

L
Linus Torvalds 已提交
210 211 212
	return sig;
}

213 214 215 216 217 218 219 220 221 222 223 224 225 226
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;

	printk(KERN_INFO "%s/%d: reached RLIMIT_SIGPENDING, dropped signal %d\n",
				current->comm, current->pid, sig);
}

227
/**
228
 * task_set_jobctl_pending - set jobctl pending bits
229
 * @task: target task
230
 * @mask: pending bits to set
231
 *
232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259
 * 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.
 */
bool task_set_jobctl_pending(struct task_struct *task, unsigned int mask)
{
	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;
}

260
/**
261
 * task_clear_jobctl_trapping - clear jobctl trapping bit
262 263
 * @task: target task
 *
264 265 266 267
 * 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.
268 269 270 271
 *
 * CONTEXT:
 * Must be called with @task->sighand->siglock held.
 */
272
void task_clear_jobctl_trapping(struct task_struct *task)
273
{
274 275
	if (unlikely(task->jobctl & JOBCTL_TRAPPING)) {
		task->jobctl &= ~JOBCTL_TRAPPING;
276
		wake_up_bit(&task->jobctl, JOBCTL_TRAPPING_BIT);
277 278 279
	}
}

280
/**
281
 * task_clear_jobctl_pending - clear jobctl pending bits
282
 * @task: target task
283
 * @mask: pending bits to clear
284
 *
285 286 287
 * 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.
288
 *
289 290
 * If clearing of @mask leaves no stop or trap pending, this function calls
 * task_clear_jobctl_trapping().
291 292 293 294
 *
 * CONTEXT:
 * Must be called with @task->sighand->siglock held.
 */
295
void task_clear_jobctl_pending(struct task_struct *task, unsigned int mask)
296
{
297 298 299 300 301 302
	BUG_ON(mask & ~JOBCTL_PENDING_MASK);

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

	task->jobctl &= ~mask;
303 304 305

	if (!(task->jobctl & JOBCTL_PENDING_MASK))
		task_clear_jobctl_trapping(task);
306 307 308 309 310 311
}

/**
 * task_participate_group_stop - participate in a group stop
 * @task: task participating in a group stop
 *
312
 * @task has %JOBCTL_STOP_PENDING set and is participating in a group stop.
313
 * Group stop states are cleared and the group stop count is consumed if
314
 * %JOBCTL_STOP_CONSUME was set.  If the consumption completes the group
315
 * stop, the appropriate %SIGNAL_* flags are set.
316 317 318
 *
 * CONTEXT:
 * Must be called with @task->sighand->siglock held.
319 320 321 322
 *
 * RETURNS:
 * %true if group stop completion should be notified to the parent, %false
 * otherwise.
323 324 325 326
 */
static bool task_participate_group_stop(struct task_struct *task)
{
	struct signal_struct *sig = task->signal;
327
	bool consume = task->jobctl & JOBCTL_STOP_CONSUME;
328

329
	WARN_ON_ONCE(!(task->jobctl & JOBCTL_STOP_PENDING));
330

331
	task_clear_jobctl_pending(task, JOBCTL_STOP_PENDING);
332 333 334 335 336 337 338

	if (!consume)
		return false;

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

339 340 341 342 343
	/*
	 * 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)) {
344 345 346 347 348 349
		sig->flags = SIGNAL_STOP_STOPPED;
		return true;
	}
	return false;
}

350 351 352
/*
 * allocate a new signal queue record
 * - this may be called without locks if and only if t == current, otherwise an
353
 *   appropriate lock must be held to stop the target task from exiting
354
 */
355 356
static struct sigqueue *
__sigqueue_alloc(int sig, struct task_struct *t, gfp_t flags, int override_rlimit)
L
Linus Torvalds 已提交
357 358
{
	struct sigqueue *q = NULL;
359
	struct user_struct *user;
L
Linus Torvalds 已提交
360

361
	/*
362 363
	 * Protect access to @t credentials. This can go away when all
	 * callers hold rcu read lock.
364
	 */
365
	rcu_read_lock();
D
David Howells 已提交
366
	user = get_uid(__task_cred(t)->user);
367
	atomic_inc(&user->sigpending);
368
	rcu_read_unlock();
369

L
Linus Torvalds 已提交
370
	if (override_rlimit ||
371
	    atomic_read(&user->sigpending) <=
372
			task_rlimit(t, RLIMIT_SIGPENDING)) {
L
Linus Torvalds 已提交
373
		q = kmem_cache_alloc(sigqueue_cachep, flags);
374 375 376 377
	} else {
		print_dropped_signal(sig);
	}

L
Linus Torvalds 已提交
378
	if (unlikely(q == NULL)) {
379
		atomic_dec(&user->sigpending);
D
David Howells 已提交
380
		free_uid(user);
L
Linus Torvalds 已提交
381 382 383
	} else {
		INIT_LIST_HEAD(&q->list);
		q->flags = 0;
D
David Howells 已提交
384
		q->user = user;
L
Linus Torvalds 已提交
385
	}
D
David Howells 已提交
386 387

	return q;
L
Linus Torvalds 已提交
388 389
}

390
static void __sigqueue_free(struct sigqueue *q)
L
Linus Torvalds 已提交
391 392 393 394 395 396 397 398
{
	if (q->flags & SIGQUEUE_PREALLOC)
		return;
	atomic_dec(&q->user->sigpending);
	free_uid(q->user);
	kmem_cache_free(sigqueue_cachep, q);
}

399
void flush_sigqueue(struct sigpending *queue)
L
Linus Torvalds 已提交
400 401 402 403 404 405 406 407 408 409 410 411 412 413
{
	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);
	}
}

/*
 * Flush all pending signals for a task.
 */
414 415 416 417 418 419 420
void __flush_signals(struct task_struct *t)
{
	clear_tsk_thread_flag(t, TIF_SIGPENDING);
	flush_sigqueue(&t->pending);
	flush_sigqueue(&t->signal->shared_pending);
}

421
void flush_signals(struct task_struct *t)
L
Linus Torvalds 已提交
422 423 424 425
{
	unsigned long flags;

	spin_lock_irqsave(&t->sighand->siglock, flags);
426
	__flush_signals(t);
L
Linus Torvalds 已提交
427 428 429
	spin_unlock_irqrestore(&t->sighand->siglock, flags);
}

430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463
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);
}

464 465 466 467 468 469 470 471 472 473
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
474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491
/*
 * 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;
		sigemptyset(&ka->sa.sa_mask);
		ka++;
	}
}

492 493
int unhandled_signal(struct task_struct *tsk, int sig)
{
494
	void __user *handler = tsk->sighand->action[sig-1].sa.sa_handler;
495
	if (is_global_init(tsk))
496
		return 1;
497
	if (handler != SIG_IGN && handler != SIG_DFL)
498
		return 0;
T
Tejun Heo 已提交
499 500
	/* if ptraced, let the tracer determine */
	return !tsk->ptrace;
501 502
}

503 504
/*
 * Notify the system that a driver wants to block all signals for this
L
Linus Torvalds 已提交
505 506 507 508 509
 * process, and wants to be notified if any signals at all were to be
 * sent/acted upon.  If the notifier routine returns non-zero, then the
 * signal will be acted upon after all.  If the notifier routine returns 0,
 * then then signal will be blocked.  Only one block per process is
 * allowed.  priv is a pointer to private data that the notifier routine
510 511
 * can use to determine if the signal should be blocked or not.
 */
L
Linus Torvalds 已提交
512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537
void
block_all_signals(int (*notifier)(void *priv), void *priv, sigset_t *mask)
{
	unsigned long flags;

	spin_lock_irqsave(&current->sighand->siglock, flags);
	current->notifier_mask = mask;
	current->notifier_data = priv;
	current->notifier = notifier;
	spin_unlock_irqrestore(&current->sighand->siglock, flags);
}

/* Notify the system that blocking has ended. */

void
unblock_all_signals(void)
{
	unsigned long flags;

	spin_lock_irqsave(&current->sighand->siglock, flags);
	current->notifier = NULL;
	current->notifier_data = NULL;
	recalc_sigpending();
	spin_unlock_irqrestore(&current->sighand->siglock, flags);
}

538
static void collect_signal(int sig, struct sigpending *list, siginfo_t *info)
L
Linus Torvalds 已提交
539 540 541 542 543 544 545 546 547
{
	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) {
548 549
			if (first)
				goto still_pending;
L
Linus Torvalds 已提交
550 551 552
			first = q;
		}
	}
553 554 555

	sigdelset(&list->signal, sig);

L
Linus Torvalds 已提交
556
	if (first) {
557
still_pending:
L
Linus Torvalds 已提交
558 559 560 561
		list_del_init(&first->list);
		copy_siginfo(info, &first->info);
		__sigqueue_free(first);
	} else {
562 563 564 565
		/*
		 * 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
566 567 568
		 */
		info->si_signo = sig;
		info->si_errno = 0;
569
		info->si_code = SI_USER;
L
Linus Torvalds 已提交
570 571 572 573 574 575 576 577
		info->si_pid = 0;
		info->si_uid = 0;
	}
}

static int __dequeue_signal(struct sigpending *pending, sigset_t *mask,
			siginfo_t *info)
{
578
	int sig = next_signal(pending, mask);
L
Linus Torvalds 已提交
579 580 581 582 583 584 585 586 587 588 589

	if (sig) {
		if (current->notifier) {
			if (sigismember(current->notifier_mask, sig)) {
				if (!(current->notifier)(current->notifier_data)) {
					clear_thread_flag(TIF_SIGPENDING);
					return 0;
				}
			}
		}

590
		collect_signal(sig, pending, info);
L
Linus Torvalds 已提交
591 592 593 594 595 596
	}

	return sig;
}

/*
597
 * Dequeue a signal and return the element to the caller, which is
L
Linus Torvalds 已提交
598 599 600 601 602 603
 * expected to free it.
 *
 * All callers have to hold the siglock.
 */
int dequeue_signal(struct task_struct *tsk, sigset_t *mask, siginfo_t *info)
{
604
	int signr;
605 606 607 608

	/* We only dequeue private signals from ourselves, we don't let
	 * signalfd steal them
	 */
D
Davide Libenzi 已提交
609
	signr = __dequeue_signal(&tsk->pending, mask, info);
610
	if (!signr) {
L
Linus Torvalds 已提交
611 612
		signr = __dequeue_signal(&tsk->signal->shared_pending,
					 mask, info);
613 614 615 616 617 618
		/*
		 * 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
619
		 * compliant with the old way of self-restarting
620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636
		 * 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) &&
			    tsk->signal->it_real_incr.tv64 != 0) {
				hrtimer_forward(tmr, tmr->base->get_time(),
						tsk->signal->it_real_incr);
				hrtimer_restart(tmr);
			}
		}
	}
637

D
Davide Libenzi 已提交
638
	recalc_sigpending();
639 640 641 642
	if (!signr)
		return 0;

	if (unlikely(sig_kernel_stop(signr))) {
643 644 645 646 647 648 649 650 651 652 653 654
		/*
		 * 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.
		 */
655
		current->jobctl |= JOBCTL_STOP_DEQUEUED;
656
	}
657
	if ((info->si_code & __SI_MASK) == __SI_TIMER && info->si_sys_private) {
L
Linus Torvalds 已提交
658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688
		/*
		 * 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);
		do_schedule_next_timer(info);
		spin_lock(&tsk->sighand->siglock);
	}
	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
 */
void signal_wake_up(struct task_struct *t, int resume)
{
	unsigned int mask;

	set_tsk_thread_flag(t, TIF_SIGPENDING);

	/*
M
Matthew Wilcox 已提交
689 690
	 * For SIGKILL, we want to wake it up in the stopped/traced/killable
	 * case. We don't check t->state here because there is a race with it
L
Linus Torvalds 已提交
691 692 693 694 695 696
	 * 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.
	 */
	mask = TASK_INTERRUPTIBLE;
	if (resume)
M
Matthew Wilcox 已提交
697
		mask |= TASK_WAKEKILL;
L
Linus Torvalds 已提交
698 699 700 701
	if (!wake_up_state(t, mask))
		kick_process(t);
}

702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719
/*
 * Remove signals in mask from the pending set and queue.
 * Returns 1 if any signals were found.
 *
 * All callers must be holding the siglock.
 *
 * This version takes a sigset mask and looks at all signals,
 * not just those in the first mask word.
 */
static int rm_from_queue_full(sigset_t *mask, struct sigpending *s)
{
	struct sigqueue *q, *n;
	sigset_t m;

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

720
	sigandnsets(&s->signal, &s->signal, mask);
721 722 723 724 725 726 727 728
	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
729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752
/*
 * Remove signals in mask from the pending set and queue.
 * Returns 1 if any signals were found.
 *
 * All callers must be holding the siglock.
 */
static int rm_from_queue(unsigned long mask, struct sigpending *s)
{
	struct sigqueue *q, *n;

	if (!sigtestsetmask(&s->signal, mask))
		return 0;

	sigdelsetmask(&s->signal, mask);
	list_for_each_entry_safe(q, n, &s->list, list) {
		if (q->info.si_signo < SIGRTMIN &&
		    (mask & sigmask(q->info.si_signo))) {
			list_del_init(&q->list);
			__sigqueue_free(q);
		}
	}
	return 1;
}

753 754 755 756 757 758 759 760 761 762 763
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));
}

764 765 766 767 768 769 770 771
/*
 * 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);

772 773 774 775
	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))
776 777
		return 1;

778
	if (ns_capable(tcred->user_ns, CAP_KILL))
779 780 781 782 783
		return 1;

	return 0;
}

L
Linus Torvalds 已提交
784 785
/*
 * Bad permissions for sending the signal
786
 * - the caller must hold the RCU read lock
L
Linus Torvalds 已提交
787 788 789 790
 */
static int check_kill_permission(int sig, struct siginfo *info,
				 struct task_struct *t)
{
791
	struct pid *sid;
792 793
	int error;

794
	if (!valid_signal(sig))
795 796
		return -EINVAL;

797
	if (!si_fromuser(info))
798
		return 0;
A
Amy Griffis 已提交
799

800 801
	error = audit_signal_info(sig, t); /* Let audit system see the signal */
	if (error)
L
Linus Torvalds 已提交
802
		return error;
803

804
	if (!same_thread_group(current, t) &&
805
	    !kill_ok_by_cred(t)) {
806 807 808 809 810 811 812 813 814 815 816 817 818
		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;
		}
	}
819

A
Amy Griffis 已提交
820
	return security_task_kill(t, info, sig, 0);
L
Linus Torvalds 已提交
821 822
}

823 824 825 826 827 828 829 830
/**
 * 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
831 832 833 834 835
 * 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.
836 837 838 839 840 841 842 843 844 845
 *
 * 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);
T
Tejun Heo 已提交
846
	signal_wake_up(t, t->jobctl & JOBCTL_LISTENING);
847 848
}

L
Linus Torvalds 已提交
849
/*
850 851
 * 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
852 853
 * time regardless of blocking, ignoring, or handling.  This does the
 * actual continuing for SIGCONT, but not the actual stopping for stop
854 855 856 857
 * 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
858
 */
859
static int prepare_signal(int sig, struct task_struct *p, bool force)
L
Linus Torvalds 已提交
860
{
861
	struct signal_struct *signal = p->signal;
L
Linus Torvalds 已提交
862 863
	struct task_struct *t;

864
	if (unlikely(signal->flags & SIGNAL_GROUP_EXIT)) {
L
Linus Torvalds 已提交
865
		/*
866
		 * The process is in the middle of dying, nothing to do.
L
Linus Torvalds 已提交
867
		 */
868
	} else if (sig_kernel_stop(sig)) {
L
Linus Torvalds 已提交
869 870 871
		/*
		 * This is a stop signal.  Remove SIGCONT from all queues.
		 */
872
		rm_from_queue(sigmask(SIGCONT), &signal->shared_pending);
L
Linus Torvalds 已提交
873 874 875
		t = p;
		do {
			rm_from_queue(sigmask(SIGCONT), &t->pending);
876
		} while_each_thread(p, t);
L
Linus Torvalds 已提交
877
	} else if (sig == SIGCONT) {
878
		unsigned int why;
L
Linus Torvalds 已提交
879
		/*
880
		 * Remove all stop signals from all queues, wake all threads.
L
Linus Torvalds 已提交
881
		 */
882
		rm_from_queue(SIG_KERNEL_STOP_MASK, &signal->shared_pending);
L
Linus Torvalds 已提交
883 884
		t = p;
		do {
885
			task_clear_jobctl_pending(t, JOBCTL_STOP_PENDING);
L
Linus Torvalds 已提交
886
			rm_from_queue(SIG_KERNEL_STOP_MASK, &t->pending);
887 888 889 890
			if (likely(!(t->ptrace & PT_SEIZED)))
				wake_up_state(t, __TASK_STOPPED);
			else
				ptrace_trap_notify(t);
891
		} while_each_thread(p, t);
L
Linus Torvalds 已提交
892

893 894 895 896 897 898 899 900 901
		/*
		 * 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;
902
		if (signal->flags & SIGNAL_STOP_STOPPED)
903
			why |= SIGNAL_CLD_CONTINUED;
904
		else if (signal->group_stop_count)
905 906 907
			why |= SIGNAL_CLD_STOPPED;

		if (why) {
908
			/*
909
			 * The first thread which returns from do_signal_stop()
910 911 912
			 * will take ->siglock, notice SIGNAL_CLD_MASK, and
			 * notify its parent. See get_signal_to_deliver().
			 */
913 914 915
			signal->flags = why | SIGNAL_STOP_CONTINUED;
			signal->group_stop_count = 0;
			signal->group_exit_code = 0;
L
Linus Torvalds 已提交
916 917
		}
	}
918

919
	return !sig_ignored(p, sig, force);
L
Linus Torvalds 已提交
920 921
}

922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942
/*
 * 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);
}

943
static void complete_signal(int sig, struct task_struct *p, int group)
944 945 946 947 948 949 950 951 952 953 954 955
{
	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;
956
	else if (!group || thread_group_empty(p))
957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983
		/*
		 * 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.
	 */
984 985
	if (sig_fatal(p, sig) &&
	    !(signal->flags & (SIGNAL_UNKILLABLE | SIGNAL_GROUP_EXIT)) &&
986
	    !sigismember(&t->real_blocked, sig) &&
T
Tejun Heo 已提交
987
	    (sig == SIGKILL || !t->ptrace)) {
988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002
		/*
		 * 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 {
1003
				task_clear_jobctl_pending(t, JOBCTL_PENDING_MASK);
1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018
				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;
}

1019 1020 1021 1022 1023
static inline int legacy_queue(struct sigpending *signals, int sig)
{
	return (sig < SIGRTMIN) && sigismember(&signals->signal, sig);
}

1024 1025 1026 1027 1028 1029 1030 1031 1032
#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;

1033 1034 1035 1036
	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();
1037 1038 1039 1040 1041 1042 1043 1044
}
#else
static inline void userns_fixup_signal_uid(struct siginfo *info, struct task_struct *t)
{
	return;
}
#endif

1045 1046
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
1047
{
1048
	struct sigpending *pending;
1049
	struct sigqueue *q;
1050
	int override_rlimit;
1051
	int ret = 0, result;
1052

1053
	assert_spin_locked(&t->sighand->siglock);
1054

1055
	result = TRACE_SIGNAL_IGNORED;
1056 1057
	if (!prepare_signal(sig, t,
			from_ancestor_ns || (info == SEND_SIG_FORCED)))
1058
		goto ret;
1059 1060

	pending = group ? &t->signal->shared_pending : &t->pending;
1061 1062 1063 1064 1065
	/*
	 * 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.
	 */
1066
	result = TRACE_SIGNAL_ALREADY_PENDING;
1067
	if (legacy_queue(pending, sig))
1068 1069 1070
		goto ret;

	result = TRACE_SIGNAL_DELIVERED;
L
Linus Torvalds 已提交
1071 1072 1073 1074
	/*
	 * fast-pathed signals for kernel-internal things like SIGSTOP
	 * or SIGKILL.
	 */
1075
	if (info == SEND_SIG_FORCED)
L
Linus Torvalds 已提交
1076 1077
		goto out_set;

1078 1079 1080 1081 1082 1083 1084 1085 1086
	/*
	 * 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.
	 */
1087 1088 1089 1090 1091
	if (sig < SIGRTMIN)
		override_rlimit = (is_si_special(info) || info->si_code >= 0);
	else
		override_rlimit = 0;

1092
	q = __sigqueue_alloc(sig, t, GFP_ATOMIC | __GFP_NOTRACK_FALSE_POSITIVE,
1093
		override_rlimit);
L
Linus Torvalds 已提交
1094
	if (q) {
1095
		list_add_tail(&q->list, &pending->list);
L
Linus Torvalds 已提交
1096
		switch ((unsigned long) info) {
1097
		case (unsigned long) SEND_SIG_NOINFO:
L
Linus Torvalds 已提交
1098 1099 1100
			q->info.si_signo = sig;
			q->info.si_errno = 0;
			q->info.si_code = SI_USER;
1101
			q->info.si_pid = task_tgid_nr_ns(current,
1102
							task_active_pid_ns(t));
1103
			q->info.si_uid = from_kuid_munged(current_user_ns(), current_uid());
L
Linus Torvalds 已提交
1104
			break;
1105
		case (unsigned long) SEND_SIG_PRIV:
L
Linus Torvalds 已提交
1106 1107 1108 1109 1110 1111 1112 1113
			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);
1114 1115
			if (from_ancestor_ns)
				q->info.si_pid = 0;
L
Linus Torvalds 已提交
1116 1117
			break;
		}
1118 1119 1120

		userns_fixup_signal_uid(&q->info, t);

1121
	} else if (!is_si_special(info)) {
1122 1123 1124 1125 1126 1127
		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().
			 */
1128 1129 1130
			result = TRACE_SIGNAL_OVERFLOW_FAIL;
			ret = -EAGAIN;
			goto ret;
1131 1132 1133 1134 1135
		} else {
			/*
			 * This is a silent loss of information.  We still
			 * send the signal, but the *info bits are lost.
			 */
1136
			result = TRACE_SIGNAL_LOSE_INFO;
1137
		}
L
Linus Torvalds 已提交
1138 1139 1140
	}

out_set:
1141
	signalfd_notify(t, sig);
1142
	sigaddset(&pending->signal, sig);
1143
	complete_signal(sig, t, group);
1144 1145 1146
ret:
	trace_signal_generate(sig, info, t, group, result);
	return ret;
L
Linus Torvalds 已提交
1147 1148
}

1149 1150 1151
static int send_signal(int sig, struct siginfo *info, struct task_struct *t,
			int group)
{
1152 1153 1154
	int from_ancestor_ns = 0;

#ifdef CONFIG_PID_NS
1155 1156
	from_ancestor_ns = si_fromuser(info) &&
			   !task_pid_nr_ns(current, task_active_pid_ns(t));
1157 1158 1159
#endif

	return __send_signal(sig, info, t, group, from_ancestor_ns);
1160 1161
}

1162
static void print_fatal_signal(int signr)
I
Ingo Molnar 已提交
1163
{
1164
	struct pt_regs *regs = signal_pt_regs();
I
Ingo Molnar 已提交
1165
	printk("%s/%d: potentially unexpected fatal signal %d.\n",
1166
		current->comm, task_pid_nr(current), signr);
I
Ingo Molnar 已提交
1167

A
Al Viro 已提交
1168
#if defined(__i386__) && !defined(__arch_um__)
1169
	printk("code at %08lx: ", regs->ip);
I
Ingo Molnar 已提交
1170 1171 1172 1173 1174
	{
		int i;
		for (i = 0; i < 16; i++) {
			unsigned char insn;

1175 1176
			if (get_user(insn, (unsigned char *)(regs->ip + i)))
				break;
I
Ingo Molnar 已提交
1177 1178 1179 1180 1181
			printk("%02x ", insn);
		}
	}
#endif
	printk("\n");
1182
	preempt_disable();
I
Ingo Molnar 已提交
1183
	show_regs(regs);
1184
	preempt_enable();
I
Ingo Molnar 已提交
1185 1186 1187 1188 1189 1190 1191 1192 1193 1194
}

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
1195

1196 1197 1198 1199 1200 1201
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
1202 1203 1204
static int
specific_send_sig_info(int sig, struct siginfo *info, struct task_struct *t)
{
1205
	return send_signal(sig, info, t, 0);
L
Linus Torvalds 已提交
1206 1207
}

1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221
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
1222 1223 1224
/*
 * Force a signal that the process can't ignore: if necessary
 * we unblock the signal and change any SIG_IGN to SIG_DFL.
1225 1226 1227 1228 1229
 *
 * 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.
 *
1230 1231
 * 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
1232 1233 1234 1235 1236
 */
int
force_sig_info(int sig, struct siginfo *info, struct task_struct *t)
{
	unsigned long int flags;
1237 1238
	int ret, blocked, ignored;
	struct k_sigaction *action;
L
Linus Torvalds 已提交
1239 1240

	spin_lock_irqsave(&t->sighand->siglock, flags);
1241 1242 1243 1244 1245 1246 1247
	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
1248
			recalc_sigpending_and_wake(t);
1249
		}
L
Linus Torvalds 已提交
1250
	}
1251 1252
	if (action->sa.sa_handler == SIG_DFL)
		t->signal->flags &= ~SIGNAL_UNKILLABLE;
L
Linus Torvalds 已提交
1253 1254 1255 1256 1257 1258 1259 1260 1261
	ret = specific_send_sig_info(sig, info, t);
	spin_unlock_irqrestore(&t->sighand->siglock, flags);

	return ret;
}

/*
 * Nuke all other threads in the group.
 */
1262
int zap_other_threads(struct task_struct *p)
L
Linus Torvalds 已提交
1263
{
1264 1265
	struct task_struct *t = p;
	int count = 0;
L
Linus Torvalds 已提交
1266 1267 1268

	p->signal->group_stop_count = 0;

1269
	while_each_thread(p, t) {
1270
		task_clear_jobctl_pending(t, JOBCTL_PENDING_MASK);
1271 1272 1273
		count++;

		/* Don't bother with already dead threads */
L
Linus Torvalds 已提交
1274 1275 1276 1277 1278
		if (t->exit_state)
			continue;
		sigaddset(&t->pending.signal, SIGKILL);
		signal_wake_up(t, 1);
	}
1279 1280

	return count;
L
Linus Torvalds 已提交
1281 1282
}

1283 1284
struct sighand_struct *__lock_task_sighand(struct task_struct *tsk,
					   unsigned long *flags)
1285 1286 1287 1288
{
	struct sighand_struct *sighand;

	for (;;) {
1289 1290
		local_irq_save(*flags);
		rcu_read_lock();
1291
		sighand = rcu_dereference(tsk->sighand);
1292 1293 1294
		if (unlikely(sighand == NULL)) {
			rcu_read_unlock();
			local_irq_restore(*flags);
1295
			break;
1296
		}
1297

1298 1299 1300
		spin_lock(&sighand->siglock);
		if (likely(sighand == tsk->sighand)) {
			rcu_read_unlock();
1301
			break;
1302 1303 1304 1305
		}
		spin_unlock(&sighand->siglock);
		rcu_read_unlock();
		local_irq_restore(*flags);
1306 1307 1308 1309 1310
	}

	return sighand;
}

1311 1312 1313
/*
 * send signal info to all the members of a group
 */
L
Linus Torvalds 已提交
1314 1315
int group_send_sig_info(int sig, struct siginfo *info, struct task_struct *p)
{
1316 1317 1318 1319 1320
	int ret;

	rcu_read_lock();
	ret = check_kill_permission(sig, info, p);
	rcu_read_unlock();
1321

1322 1323
	if (!ret && sig)
		ret = do_send_sig_info(sig, info, p, true);
L
Linus Torvalds 已提交
1324 1325 1326 1327 1328

	return ret;
}

/*
1329
 * __kill_pgrp_info() sends a signal to a process group: this is what the tty
L
Linus Torvalds 已提交
1330
 * control characters do (^C, ^Z etc)
1331
 * - the caller must hold at least a readlock on tasklist_lock
L
Linus Torvalds 已提交
1332
 */
1333
int __kill_pgrp_info(int sig, struct siginfo *info, struct pid *pgrp)
L
Linus Torvalds 已提交
1334 1335 1336 1337 1338 1339
{
	struct task_struct *p = NULL;
	int retval, success;

	success = 0;
	retval = -ESRCH;
1340
	do_each_pid_task(pgrp, PIDTYPE_PGID, p) {
L
Linus Torvalds 已提交
1341 1342 1343
		int err = group_send_sig_info(sig, info, p);
		success |= !err;
		retval = err;
1344
	} while_each_pid_task(pgrp, PIDTYPE_PGID, p);
L
Linus Torvalds 已提交
1345 1346 1347
	return success ? 0 : retval;
}

1348
int kill_pid_info(int sig, struct siginfo *info, struct pid *pid)
L
Linus Torvalds 已提交
1349
{
1350
	int error = -ESRCH;
L
Linus Torvalds 已提交
1351 1352
	struct task_struct *p;

I
Ingo Molnar 已提交
1353
	rcu_read_lock();
1354
retry:
1355
	p = pid_task(pid, PIDTYPE_PID);
1356
	if (p) {
L
Linus Torvalds 已提交
1357
		error = group_send_sig_info(sig, info, p);
1358 1359 1360 1361 1362 1363 1364 1365 1366
		if (unlikely(error == -ESRCH))
			/*
			 * 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.
			 */
			goto retry;
	}
I
Ingo Molnar 已提交
1367
	rcu_read_unlock();
1368

L
Linus Torvalds 已提交
1369 1370 1371
	return error;
}

1372
int kill_proc_info(int sig, struct siginfo *info, pid_t pid)
1373 1374 1375
{
	int error;
	rcu_read_lock();
1376
	error = kill_pid_info(sig, info, find_vpid(pid));
1377 1378 1379 1380
	rcu_read_unlock();
	return error;
}

1381 1382 1383 1384
static int kill_as_cred_perm(const struct cred *cred,
			     struct task_struct *target)
{
	const struct cred *pcred = __task_cred(target);
1385 1386
	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))
1387 1388 1389 1390
		return 0;
	return 1;
}

1391
/* like kill_pid_info(), but doesn't use uid/euid of "current" */
1392 1393
int kill_pid_info_as_cred(int sig, struct siginfo *info, struct pid *pid,
			 const struct cred *cred, u32 secid)
1394 1395 1396
{
	int ret = -EINVAL;
	struct task_struct *p;
1397
	unsigned long flags;
1398 1399 1400 1401

	if (!valid_signal(sig))
		return ret;

1402
	rcu_read_lock();
1403
	p = pid_task(pid, PIDTYPE_PID);
1404 1405 1406 1407
	if (!p) {
		ret = -ESRCH;
		goto out_unlock;
	}
1408
	if (si_fromuser(info) && !kill_as_cred_perm(cred, p)) {
1409 1410 1411
		ret = -EPERM;
		goto out_unlock;
	}
1412 1413 1414
	ret = security_task_kill(p, info, sig, secid);
	if (ret)
		goto out_unlock;
1415 1416 1417 1418 1419 1420 1421

	if (sig) {
		if (lock_task_sighand(p, &flags)) {
			ret = __send_signal(sig, info, p, 1, 0);
			unlock_task_sighand(p, &flags);
		} else
			ret = -ESRCH;
1422 1423
	}
out_unlock:
1424
	rcu_read_unlock();
1425 1426
	return ret;
}
1427
EXPORT_SYMBOL_GPL(kill_pid_info_as_cred);
L
Linus Torvalds 已提交
1428 1429 1430 1431 1432 1433 1434 1435

/*
 * 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.
 */

1436
static int kill_something_info(int sig, struct siginfo *info, pid_t pid)
L
Linus Torvalds 已提交
1437
{
1438
	int ret;
1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451

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

	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
1452 1453 1454 1455
		int retval = 0, count = 0;
		struct task_struct * p;

		for_each_process(p) {
1456 1457
			if (task_pid_vnr(p) > 1 &&
					!same_thread_group(p, current)) {
L
Linus Torvalds 已提交
1458 1459 1460 1461 1462 1463
				int err = group_send_sig_info(sig, info, p);
				++count;
				if (err != -EPERM)
					retval = err;
			}
		}
1464
		ret = count ? retval : -ESRCH;
L
Linus Torvalds 已提交
1465
	}
1466 1467
	read_unlock(&tasklist_lock);

1468
	return ret;
L
Linus Torvalds 已提交
1469 1470 1471 1472 1473 1474
}

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

1475
int send_sig_info(int sig, struct siginfo *info, struct task_struct *p)
L
Linus Torvalds 已提交
1476 1477 1478 1479 1480
{
	/*
	 * Make sure legacy kernel users don't send in bad values
	 * (normal paths check this in check_kill_permission).
	 */
1481
	if (!valid_signal(sig))
L
Linus Torvalds 已提交
1482 1483
		return -EINVAL;

1484
	return do_send_sig_info(sig, info, p, false);
L
Linus Torvalds 已提交
1485 1486
}

1487 1488 1489
#define __si_special(priv) \
	((priv) ? SEND_SIG_PRIV : SEND_SIG_NOINFO)

L
Linus Torvalds 已提交
1490 1491 1492
int
send_sig(int sig, struct task_struct *p, int priv)
{
1493
	return send_sig_info(sig, __si_special(priv), p);
L
Linus Torvalds 已提交
1494 1495 1496 1497 1498
}

void
force_sig(int sig, struct task_struct *p)
{
1499
	force_sig_info(sig, SEND_SIG_PRIV, p);
L
Linus Torvalds 已提交
1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520
}

/*
 * 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;
}

1521 1522
int kill_pgrp(struct pid *pid, int sig, int priv)
{
1523 1524 1525 1526 1527 1528 1529
	int ret;

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

	return ret;
1530 1531 1532 1533 1534 1535 1536 1537 1538
}
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
1539 1540 1541 1542
/*
 * 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
1543
 * expirations or I/O completions".  In the case of POSIX Timers
L
Linus Torvalds 已提交
1544 1545 1546 1547 1548 1549
 * 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)
{
1550
	struct sigqueue *q = __sigqueue_alloc(-1, current, GFP_KERNEL, 0);
L
Linus Torvalds 已提交
1551

1552
	if (q)
L
Linus Torvalds 已提交
1553
		q->flags |= SIGQUEUE_PREALLOC;
1554 1555

	return q;
L
Linus Torvalds 已提交
1556 1557 1558 1559 1560
}

void sigqueue_free(struct sigqueue *q)
{
	unsigned long flags;
1561 1562
	spinlock_t *lock = &current->sighand->siglock;

L
Linus Torvalds 已提交
1563 1564
	BUG_ON(!(q->flags & SIGQUEUE_PREALLOC));
	/*
1565 1566
	 * We must hold ->siglock while testing q->list
	 * to serialize with collect_signal() or with
1567
	 * __exit_signal()->flush_sigqueue().
L
Linus Torvalds 已提交
1568
	 */
1569
	spin_lock_irqsave(lock, flags);
1570 1571 1572 1573 1574
	q->flags &= ~SIGQUEUE_PREALLOC;
	/*
	 * If it is queued it will be freed when dequeued,
	 * like the "regular" sigqueue.
	 */
1575
	if (!list_empty(&q->list))
1576
		q = NULL;
1577 1578
	spin_unlock_irqrestore(lock, flags);

1579 1580
	if (q)
		__sigqueue_free(q);
L
Linus Torvalds 已提交
1581 1582
}

1583
int send_sigqueue(struct sigqueue *q, struct task_struct *t, int group)
1584
{
1585
	int sig = q->info.si_signo;
1586
	struct sigpending *pending;
1587
	unsigned long flags;
1588
	int ret, result;
1589

1590
	BUG_ON(!(q->flags & SIGQUEUE_PREALLOC));
1591 1592 1593 1594 1595

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

1596
	ret = 1; /* the signal is ignored */
1597
	result = TRACE_SIGNAL_IGNORED;
1598
	if (!prepare_signal(sig, t, false))
1599 1600 1601
		goto out;

	ret = 0;
1602 1603 1604 1605 1606 1607 1608
	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++;
1609
		result = TRACE_SIGNAL_ALREADY_PENDING;
1610
		goto out;
1611
	}
1612
	q->info.si_overrun = 0;
1613 1614

	signalfd_notify(t, sig);
1615
	pending = group ? &t->signal->shared_pending : &t->pending;
1616 1617
	list_add_tail(&q->list, &pending->list);
	sigaddset(&pending->signal, sig);
1618
	complete_signal(sig, t, group);
1619
	result = TRACE_SIGNAL_DELIVERED;
1620
out:
1621
	trace_signal_generate(sig, &q->info, t, group, result);
1622 1623 1624
	unlock_task_sighand(t, &flags);
ret:
	return ret;
1625 1626
}

L
Linus Torvalds 已提交
1627 1628 1629
/*
 * 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
1630
 *
1631 1632
 * Returns true if our parent ignored us and so we've switched to
 * self-reaping.
L
Linus Torvalds 已提交
1633
 */
1634
bool do_notify_parent(struct task_struct *tsk, int sig)
L
Linus Torvalds 已提交
1635 1636 1637 1638
{
	struct siginfo info;
	unsigned long flags;
	struct sighand_struct *psig;
1639
	bool autoreap = false;
L
Linus Torvalds 已提交
1640 1641 1642 1643

	BUG_ON(sig == -1);

 	/* do_notify_parent_cldstop should have been called instead.  */
M
Matthew Wilcox 已提交
1644
 	BUG_ON(task_is_stopped_or_traced(tsk));
L
Linus Torvalds 已提交
1645

T
Tejun Heo 已提交
1646
	BUG_ON(!tsk->ptrace &&
L
Linus Torvalds 已提交
1647 1648
	       (tsk->group_leader != tsk || !thread_group_empty(tsk)));

1649 1650 1651 1652 1653 1654 1655 1656 1657
	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
1658 1659
	info.si_signo = sig;
	info.si_errno = 0;
1660
	/*
1661 1662
	 * We are under tasklist_lock here so our parent is tied to
	 * us and cannot change.
1663
	 *
1664 1665
	 * task_active_pid_ns will always return the same pid namespace
	 * until a task passes through release_task.
1666 1667 1668 1669 1670 1671
	 *
	 * 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();
1672
	info.si_pid = task_pid_nr_ns(tsk, task_active_pid_ns(tsk->parent));
1673 1674
	info.si_uid = from_kuid_munged(task_cred_xxx(tsk->parent, user_ns),
				       task_uid(tsk));
1675 1676
	rcu_read_unlock();

1677 1678
	info.si_utime = cputime_to_clock_t(tsk->utime + tsk->signal->utime);
	info.si_stime = cputime_to_clock_t(tsk->stime + tsk->signal->stime);
L
Linus Torvalds 已提交
1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691

	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
1692
	if (!tsk->ptrace && sig == SIGCHLD &&
L
Linus Torvalds 已提交
1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709
	    (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).
		 */
1710
		autoreap = true;
L
Linus Torvalds 已提交
1711
		if (psig->action[SIGCHLD-1].sa.sa_handler == SIG_IGN)
1712
			sig = 0;
L
Linus Torvalds 已提交
1713
	}
1714
	if (valid_signal(sig) && sig)
L
Linus Torvalds 已提交
1715 1716 1717
		__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
1718

1719
	return autoreap;
L
Linus Torvalds 已提交
1720 1721
}

1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736
/**
 * 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
1737 1738 1739
{
	struct siginfo info;
	unsigned long flags;
1740
	struct task_struct *parent;
L
Linus Torvalds 已提交
1741 1742
	struct sighand_struct *sighand;

1743
	if (for_ptracer) {
1744
		parent = tsk->parent;
1745
	} else {
1746 1747 1748 1749
		tsk = tsk->group_leader;
		parent = tsk->real_parent;
	}

L
Linus Torvalds 已提交
1750 1751
	info.si_signo = SIGCHLD;
	info.si_errno = 0;
1752
	/*
1753
	 * see comment in do_notify_parent() about the following 4 lines
1754 1755
	 */
	rcu_read_lock();
1756
	info.si_pid = task_pid_nr_ns(tsk, parent->nsproxy->pid_ns);
1757
	info.si_uid = from_kuid_munged(task_cred_xxx(parent, user_ns), task_uid(tsk));
1758 1759
	rcu_read_unlock();

1760 1761
	info.si_utime = cputime_to_clock_t(tsk->utime);
	info.si_stime = cputime_to_clock_t(tsk->stime);
L
Linus Torvalds 已提交
1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789

 	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);
}

1790 1791
static inline int may_ptrace_stop(void)
{
T
Tejun Heo 已提交
1792
	if (!likely(current->ptrace))
1793 1794 1795 1796 1797 1798 1799
		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
1800
	 * ->siglock we must see ->core_state != NULL. Otherwise it
1801 1802
	 * is safe to enter schedule().
	 */
1803
	if (unlikely(current->mm->core_state) &&
1804 1805 1806 1807 1808 1809
	    unlikely(current->mm == current->parent->mm))
		return 0;

	return 1;
}

R
Roland McGrath 已提交
1810
/*
1811
 * Return non-zero if there is a SIGKILL that should be waking us up.
R
Roland McGrath 已提交
1812 1813 1814 1815
 * Called with the siglock held.
 */
static int sigkill_pending(struct task_struct *tsk)
{
1816 1817
	return	sigismember(&tsk->pending.signal, SIGKILL) ||
		sigismember(&tsk->signal->shared_pending.signal, SIGKILL);
R
Roland McGrath 已提交
1818 1819
}

L
Linus Torvalds 已提交
1820 1821 1822 1823 1824 1825 1826 1827
/*
 * 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.
 *
1828 1829
 * 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
1830
 */
T
Tejun Heo 已提交
1831
static void ptrace_stop(int exit_code, int why, int clear_code, siginfo_t *info)
1832 1833
	__releases(&current->sighand->siglock)
	__acquires(&current->sighand->siglock)
L
Linus Torvalds 已提交
1834
{
1835 1836
	bool gstop_done = false;

R
Roland McGrath 已提交
1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851
	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);
1852 1853
		if (sigkill_pending(current))
			return;
R
Roland McGrath 已提交
1854 1855
	}

L
Linus Torvalds 已提交
1856
	/*
1857 1858 1859 1860 1861
	 * 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
1862
	 */
1863
	set_current_state(TASK_TRACED);
L
Linus Torvalds 已提交
1864 1865 1866 1867

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

1868
	/*
1869 1870
	 * If @why is CLD_STOPPED, we're trapping to participate in a group
	 * stop.  Do the bookkeeping.  Note that if SIGCONT was delievered
1871 1872 1873
	 * 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.
1874
	 */
1875
	if (why == CLD_STOPPED && (current->jobctl & JOBCTL_STOP_PENDING))
1876
		gstop_done = task_participate_group_stop(current);
1877

1878
	/* any trap clears pending STOP trap, STOP trap clears NOTIFY */
1879
	task_clear_jobctl_pending(current, JOBCTL_TRAP_STOP);
1880 1881
	if (info && info->si_code >> 8 == PTRACE_EVENT_STOP)
		task_clear_jobctl_pending(current, JOBCTL_TRAP_NOTIFY);
1882

1883
	/* entering a trap, clear TRAPPING */
1884
	task_clear_jobctl_trapping(current);
1885

L
Linus Torvalds 已提交
1886 1887
	spin_unlock_irq(&current->sighand->siglock);
	read_lock(&tasklist_lock);
1888
	if (may_ptrace_stop()) {
1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899
		/*
		 * 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);
1900
		if (gstop_done && ptrace_reparented(current))
1901 1902
			do_notify_parent_cldstop(current, false, why);

M
Miklos Szeredi 已提交
1903 1904 1905 1906 1907 1908 1909
		/*
		 * 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
1910
		read_unlock(&tasklist_lock);
M
Miklos Szeredi 已提交
1911
		preempt_enable_no_resched();
L
Linus Torvalds 已提交
1912 1913 1914 1915
		schedule();
	} else {
		/*
		 * By the time we got the lock, our tracer went away.
1916
		 * Don't drop the lock yet, another tracer may come.
1917 1918 1919
		 *
		 * If @gstop_done, the ptracer went away between group stop
		 * completion and here.  During detach, it would have set
1920 1921 1922
		 * 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
1923
		 */
1924 1925 1926
		if (gstop_done)
			do_notify_parent_cldstop(current, false, why);

1927
		__set_current_state(TASK_RUNNING);
1928 1929
		if (clear_code)
			current->exit_code = 0;
1930
		read_unlock(&tasklist_lock);
L
Linus Torvalds 已提交
1931 1932
	}

R
Roland McGrath 已提交
1933 1934 1935 1936 1937 1938 1939
	/*
	 * While in TASK_TRACED, we were considered "frozen enough".
	 * Now that we woke up, it's crucial if we're supposed to be
	 * frozen that we freeze now before running anything substantial.
	 */
	try_to_freeze();

L
Linus Torvalds 已提交
1940 1941 1942 1943 1944 1945 1946 1947
	/*
	 * 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
1948 1949 1950
	/* LISTENING can be set only during STOP traps, clear it */
	current->jobctl &= ~JOBCTL_LISTENING;

L
Linus Torvalds 已提交
1951 1952 1953
	/*
	 * Queued signals ignored us while we were stopped for tracing.
	 * So check for any that we should take before resuming user mode.
1954
	 * This sets TIF_SIGPENDING, but never clears it.
L
Linus Torvalds 已提交
1955
	 */
1956
	recalc_sigpending_tsk(current);
L
Linus Torvalds 已提交
1957 1958
}

T
Tejun Heo 已提交
1959
static void ptrace_do_notify(int signr, int exit_code, int why)
L
Linus Torvalds 已提交
1960 1961 1962 1963
{
	siginfo_t info;

	memset(&info, 0, sizeof info);
T
Tejun Heo 已提交
1964
	info.si_signo = signr;
L
Linus Torvalds 已提交
1965
	info.si_code = exit_code;
1966
	info.si_pid = task_pid_vnr(current);
1967
	info.si_uid = from_kuid_munged(current_user_ns(), current_uid());
L
Linus Torvalds 已提交
1968 1969

	/* Let the debugger run.  */
T
Tejun Heo 已提交
1970 1971 1972 1973 1974 1975
	ptrace_stop(exit_code, why, 1, &info);
}

void ptrace_notify(int exit_code)
{
	BUG_ON((exit_code & (0x7f | ~0xffff)) != SIGTRAP);
1976 1977
	if (unlikely(current->task_works))
		task_work_run();
T
Tejun Heo 已提交
1978

L
Linus Torvalds 已提交
1979
	spin_lock_irq(&current->sighand->siglock);
T
Tejun Heo 已提交
1980
	ptrace_do_notify(SIGTRAP, exit_code, CLD_TRAPPED);
L
Linus Torvalds 已提交
1981 1982 1983
	spin_unlock_irq(&current->sighand->siglock);
}

1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
/**
 * 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
2005
 */
2006 2007
static bool do_signal_stop(int signr)
	__releases(&current->sighand->siglock)
L
Linus Torvalds 已提交
2008 2009 2010
{
	struct signal_struct *sig = current->signal;

2011 2012
	if (!(current->jobctl & JOBCTL_STOP_PENDING)) {
		unsigned int gstop = JOBCTL_STOP_PENDING | JOBCTL_STOP_CONSUME;
2013 2014
		struct task_struct *t;

2015 2016
		/* signr will be recorded in task->jobctl for retries */
		WARN_ON_ONCE(signr & ~JOBCTL_STOP_SIGMASK);
2017

2018
		if (!likely(current->jobctl & JOBCTL_STOP_DEQUEUED) ||
2019
		    unlikely(signal_group_exit(sig)))
2020
			return false;
L
Linus Torvalds 已提交
2021
		/*
2022 2023 2024 2025 2026 2027 2028
		 * 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
2029 2030
		 * cause two group stops when !ptraced.  That is why we
		 * also check !task_is_stopped(t) below.
2031 2032 2033 2034 2035 2036 2037 2038
		 *
		 * 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
2039
		 */
2040 2041
		if (!(sig->flags & SIGNAL_STOP_STOPPED))
			sig->group_exit_code = signr;
L
Linus Torvalds 已提交
2042

2043 2044 2045 2046
		sig->group_stop_count = 0;

		if (task_set_jobctl_pending(current, signr | gstop))
			sig->group_stop_count++;
L
Linus Torvalds 已提交
2047

2048 2049
		for (t = next_thread(current); t != current;
		     t = next_thread(t)) {
L
Linus Torvalds 已提交
2050
			/*
2051 2052 2053
			 * 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
2054
			 */
2055 2056
			if (!task_is_stopped(t) &&
			    task_set_jobctl_pending(t, signr | gstop)) {
2057
				sig->group_stop_count++;
2058 2059 2060 2061
				if (likely(!(t->ptrace & PT_SEIZED)))
					signal_wake_up(t, 0);
				else
					ptrace_trap_notify(t);
2062
			}
2063
		}
L
Linus Torvalds 已提交
2064
	}
2065

T
Tejun Heo 已提交
2066
	if (likely(!current->ptrace)) {
2067
		int notify = 0;
L
Linus Torvalds 已提交
2068

2069 2070 2071 2072 2073 2074 2075 2076
		/*
		 * 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;

2077
		__set_current_state(TASK_STOPPED);
2078 2079
		spin_unlock_irq(&current->sighand->siglock);

2080 2081 2082 2083 2084 2085 2086 2087 2088
		/*
		 * 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.
		 */
2089 2090
		if (notify) {
			read_lock(&tasklist_lock);
2091
			do_notify_parent_cldstop(current, false, notify);
2092 2093 2094 2095 2096
			read_unlock(&tasklist_lock);
		}

		/* Now we don't run again until woken by SIGCONT or SIGKILL */
		schedule();
2097
		return true;
2098
	} else {
2099 2100 2101 2102 2103 2104
		/*
		 * 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;
2105
	}
2106
}
L
Linus Torvalds 已提交
2107

2108 2109 2110
/**
 * do_jobctl_trap - take care of ptrace jobctl traps
 *
T
Tejun Heo 已提交
2111 2112 2113 2114 2115 2116 2117
 * 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.
2118 2119 2120 2121 2122 2123 2124
 *
 * 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
2125
	struct signal_struct *signal = current->signal;
2126
	int signr = current->jobctl & JOBCTL_STOP_SIGMASK;
2127

T
Tejun Heo 已提交
2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138
	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;
2139
	}
L
Linus Torvalds 已提交
2140 2141
}

2142
static int ptrace_signal(int signr, siginfo_t *info)
R
Roland McGrath 已提交
2143
{
2144
	ptrace_signal_deliver();
2145 2146 2147 2148 2149 2150 2151 2152 2153 2154
	/*
	 * 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
2155
	ptrace_stop(signr, CLD_TRAPPED, 0, info);
R
Roland McGrath 已提交
2156 2157 2158 2159 2160 2161 2162 2163

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

	current->exit_code = 0;

2164 2165 2166 2167 2168 2169
	/*
	 * 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
2170 2171 2172 2173
	if (signr != info->si_signo) {
		info->si_signo = signr;
		info->si_errno = 0;
		info->si_code = SI_USER;
2174
		rcu_read_lock();
R
Roland McGrath 已提交
2175
		info->si_pid = task_pid_vnr(current->parent);
2176 2177
		info->si_uid = from_kuid_munged(current_user_ns(),
						task_uid(current->parent));
2178
		rcu_read_unlock();
R
Roland McGrath 已提交
2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189
	}

	/* 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;
}

L
Linus Torvalds 已提交
2190 2191 2192
int get_signal_to_deliver(siginfo_t *info, struct k_sigaction *return_ka,
			  struct pt_regs *regs, void *cookie)
{
2193 2194 2195
	struct sighand_struct *sighand = current->sighand;
	struct signal_struct *signal = current->signal;
	int signr;
L
Linus Torvalds 已提交
2196

2197 2198
	if (unlikely(current->task_works))
		task_work_run();
2199

2200 2201 2202
	if (unlikely(uprobe_deny_signal()))
		return 0;

R
Roland McGrath 已提交
2203 2204 2205 2206 2207 2208 2209
relock:
	/*
	 * We'll jump back here after any time we were stopped in TASK_STOPPED.
	 * While in TASK_STOPPED, we were considered "frozen enough".
	 * Now that we woke up, it's crucial if we're supposed to be
	 * frozen that we freeze now before running anything substantial.
	 */
2210 2211
	try_to_freeze();

2212
	spin_lock_irq(&sighand->siglock);
2213 2214 2215 2216 2217
	/*
	 * 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.
	 */
2218
	if (unlikely(signal->flags & SIGNAL_CLD_MASK)) {
2219 2220 2221 2222 2223 2224 2225
		int why;

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

2226
		signal->flags &= ~SIGNAL_CLD_MASK;
2227

2228
		spin_unlock_irq(&sighand->siglock);
R
Roland McGrath 已提交
2229

2230 2231 2232 2233 2234 2235 2236 2237
		/*
		 * 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
2238
		read_lock(&tasklist_lock);
2239 2240
		do_notify_parent_cldstop(current, false, why);

2241 2242 2243
		if (ptrace_reparented(current->group_leader))
			do_notify_parent_cldstop(current->group_leader,
						true, why);
T
Tejun Heo 已提交
2244
		read_unlock(&tasklist_lock);
2245

2246 2247 2248
		goto relock;
	}

L
Linus Torvalds 已提交
2249 2250
	for (;;) {
		struct k_sigaction *ka;
2251

T
Tejun Heo 已提交
2252 2253
		if (unlikely(current->jobctl & JOBCTL_STOP_PENDING) &&
		    do_signal_stop(0))
2254
			goto relock;
2255

2256 2257 2258 2259 2260
		if (unlikely(current->jobctl & JOBCTL_TRAP_MASK)) {
			do_jobctl_trap();
			spin_unlock_irq(&sighand->siglock);
			goto relock;
		}
L
Linus Torvalds 已提交
2261

T
Tejun Heo 已提交
2262
		signr = dequeue_signal(current, &current->blocked, info);
2263

T
Tejun Heo 已提交
2264 2265
		if (!signr)
			break; /* will return 0 */
2266

2267
		if (unlikely(current->ptrace) && signr != SIGKILL) {
2268
			signr = ptrace_signal(signr, info);
T
Tejun Heo 已提交
2269 2270
			if (!signr)
				continue;
L
Linus Torvalds 已提交
2271 2272
		}

T
Tejun Heo 已提交
2273 2274
		ka = &sighand->action[signr-1];

2275 2276 2277
		/* Trace actually delivered signals. */
		trace_signal_deliver(signr, info, ka);

L
Linus Torvalds 已提交
2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295
		if (ka->sa.sa_handler == SIG_IGN) /* Do nothing.  */
			continue;
		if (ka->sa.sa_handler != SIG_DFL) {
			/* Run the handler.  */
			*return_ka = *ka;

			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;

2296
		/*
2297
		 * Global init gets no signals it doesn't want.
2298 2299 2300 2301 2302 2303 2304
		 * 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.
2305
		 */
2306
		if (unlikely(signal->flags & SIGNAL_UNKILLABLE) &&
2307
				!sig_kernel_only(signr))
L
Linus Torvalds 已提交
2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321
			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) {
2322
				spin_unlock_irq(&sighand->siglock);
L
Linus Torvalds 已提交
2323 2324 2325

				/* signals can be posted during this window */

2326
				if (is_current_pgrp_orphaned())
L
Linus Torvalds 已提交
2327 2328
					goto relock;

2329
				spin_lock_irq(&sighand->siglock);
L
Linus Torvalds 已提交
2330 2331
			}

2332
			if (likely(do_signal_stop(info->si_signo))) {
L
Linus Torvalds 已提交
2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343
				/* It released the siglock.  */
				goto relock;
			}

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

2344
		spin_unlock_irq(&sighand->siglock);
L
Linus Torvalds 已提交
2345 2346 2347 2348 2349

		/*
		 * Anything else is fatal, maybe with a core dump.
		 */
		current->flags |= PF_SIGNALED;
2350

L
Linus Torvalds 已提交
2351
		if (sig_kernel_coredump(signr)) {
2352
			if (print_fatal_signals)
2353
				print_fatal_signal(info->si_signo);
L
Linus Torvalds 已提交
2354 2355 2356 2357 2358 2359 2360 2361
			/*
			 * 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.
			 */
2362
			do_coredump(info);
L
Linus Torvalds 已提交
2363 2364 2365 2366 2367
		}

		/*
		 * Death signals, no core dump.
		 */
2368
		do_group_exit(info->si_signo);
L
Linus Torvalds 已提交
2369 2370
		/* NOTREACHED */
	}
2371
	spin_unlock_irq(&sighand->siglock);
L
Linus Torvalds 已提交
2372 2373 2374
	return signr;
}

2375
/**
A
Al Viro 已提交
2376 2377 2378 2379 2380 2381
 * signal_delivered - 
 * @sig:		number of signal being delivered
 * @info:		siginfo_t of signal being delivered
 * @ka:			sigaction setting that chose the handler
 * @regs:		user register state
 * @stepping:		nonzero if debugger single-step or block-step in use
2382 2383
 *
 * This function should be called when a signal has succesfully been
A
Al Viro 已提交
2384 2385 2386
 * delivered. It updates the blocked signals accordingly (@ka->sa.sa_mask
 * is always blocked, and the signal itself is blocked unless %SA_NODEFER
 * is set in @ka->sa.sa_flags.  Tracing is notified.
2387
 */
A
Al Viro 已提交
2388 2389
void signal_delivered(int sig, siginfo_t *info, struct k_sigaction *ka,
			struct pt_regs *regs, int stepping)
2390 2391 2392
{
	sigset_t blocked;

2393 2394 2395 2396 2397 2398
	/* 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();

2399 2400
	sigorsets(&blocked, &current->blocked, &ka->sa.sa_mask);
	if (!(ka->sa.sa_flags & SA_NODEFER))
A
Al Viro 已提交
2401
		sigaddset(&blocked, sig);
2402
	set_current_blocked(&blocked);
A
Al Viro 已提交
2403
	tracehook_signal_handler(sig, info, ka, regs, stepping);
2404 2405
}

2406 2407
/*
 * It could be that complete_signal() picked us to notify about the
2408 2409
 * group-wide signal. Other threads should be notified now to take
 * the shared signals in @which since we will not.
2410
 */
2411
static void retarget_shared_pending(struct task_struct *tsk, sigset_t *which)
2412
{
2413
	sigset_t retarget;
2414 2415
	struct task_struct *t;

2416 2417 2418 2419
	sigandsets(&retarget, &tsk->signal->shared_pending.signal, which);
	if (sigisemptyset(&retarget))
		return;

2420 2421
	t = tsk;
	while_each_thread(tsk, t) {
2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434
		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;
2435 2436 2437
	}
}

O
Oleg Nesterov 已提交
2438 2439 2440
void exit_signals(struct task_struct *tsk)
{
	int group_stop = 0;
2441
	sigset_t unblocked;
O
Oleg Nesterov 已提交
2442

2443 2444 2445 2446 2447 2448
	/*
	 * @tsk is about to have PF_EXITING set - lock out users which
	 * expect stable threadgroup.
	 */
	threadgroup_change_begin(tsk);

2449 2450
	if (thread_group_empty(tsk) || signal_group_exit(tsk->signal)) {
		tsk->flags |= PF_EXITING;
2451
		threadgroup_change_end(tsk);
2452
		return;
O
Oleg Nesterov 已提交
2453 2454
	}

2455
	spin_lock_irq(&tsk->sighand->siglock);
O
Oleg Nesterov 已提交
2456 2457 2458 2459 2460
	/*
	 * From now this task is not visible for group-wide signals,
	 * see wants_signal(), do_signal_stop().
	 */
	tsk->flags |= PF_EXITING;
2461 2462 2463

	threadgroup_change_end(tsk);

2464 2465 2466
	if (!signal_pending(tsk))
		goto out;

2467 2468 2469
	unblocked = tsk->blocked;
	signotset(&unblocked);
	retarget_shared_pending(tsk, &unblocked);
2470

2471
	if (unlikely(tsk->jobctl & JOBCTL_STOP_PENDING) &&
2472
	    task_participate_group_stop(tsk))
T
Tejun Heo 已提交
2473
		group_stop = CLD_STOPPED;
2474
out:
O
Oleg Nesterov 已提交
2475 2476
	spin_unlock_irq(&tsk->sighand->siglock);

2477 2478 2479 2480
	/*
	 * If group stop has completed, deliver the notification.  This
	 * should always go to the real parent of the group leader.
	 */
2481
	if (unlikely(group_stop)) {
O
Oleg Nesterov 已提交
2482
		read_lock(&tasklist_lock);
2483
		do_notify_parent_cldstop(tsk, false, group_stop);
O
Oleg Nesterov 已提交
2484 2485 2486 2487
		read_unlock(&tasklist_lock);
	}
}

L
Linus Torvalds 已提交
2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502
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);
EXPORT_SYMBOL(block_all_signals);
EXPORT_SYMBOL(unblock_all_signals);


/*
 * System call entry points.
 */

2503 2504 2505
/**
 *  sys_restart_syscall - restart a system call
 */
2506
SYSCALL_DEFINE0(restart_syscall)
L
Linus Torvalds 已提交
2507 2508 2509 2510 2511 2512 2513 2514 2515 2516
{
	struct restart_block *restart = &current_thread_info()->restart_block;
	return restart->fn(restart);
}

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

2517 2518 2519 2520 2521
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. */
2522
		sigandnsets(&newblocked, newset, &current->blocked);
2523 2524 2525 2526 2527 2528
		retarget_shared_pending(tsk, &newblocked);
	}
	tsk->blocked = *newset;
	recalc_sigpending();
}

2529 2530 2531 2532 2533 2534
/**
 * 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
2535
 */
2536 2537 2538 2539 2540 2541 2542 2543 2544 2545
void set_current_blocked(sigset_t *newset)
{
	struct task_struct *tsk = current;
	sigdelsetmask(newset, sigmask(SIGKILL) | sigmask(SIGSTOP));
	spin_lock_irq(&tsk->sighand->siglock);
	__set_task_blocked(tsk, newset);
	spin_unlock_irq(&tsk->sighand->siglock);
}

void __set_current_blocked(const sigset_t *newset)
2546 2547 2548 2549
{
	struct task_struct *tsk = current;

	spin_lock_irq(&tsk->sighand->siglock);
2550
	__set_task_blocked(tsk, newset);
2551 2552
	spin_unlock_irq(&tsk->sighand->siglock);
}
L
Linus Torvalds 已提交
2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563

/*
 * 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)
{
2564 2565
	struct task_struct *tsk = current;
	sigset_t newset;
L
Linus Torvalds 已提交
2566

2567
	/* Lockless, only current can change ->blocked, never from irq */
2568
	if (oldset)
2569
		*oldset = tsk->blocked;
2570

L
Linus Torvalds 已提交
2571 2572
	switch (how) {
	case SIG_BLOCK:
2573
		sigorsets(&newset, &tsk->blocked, set);
L
Linus Torvalds 已提交
2574 2575
		break;
	case SIG_UNBLOCK:
2576
		sigandnsets(&newset, &tsk->blocked, set);
L
Linus Torvalds 已提交
2577 2578
		break;
	case SIG_SETMASK:
2579
		newset = *set;
L
Linus Torvalds 已提交
2580 2581
		break;
	default:
2582
		return -EINVAL;
L
Linus Torvalds 已提交
2583
	}
2584

2585
	__set_current_blocked(&newset);
2586
	return 0;
L
Linus Torvalds 已提交
2587 2588
}

2589 2590 2591
/**
 *  sys_rt_sigprocmask - change the list of currently blocked signals
 *  @how: whether to add, remove, or set signals
R
Randy Dunlap 已提交
2592
 *  @nset: stores pending signals
2593 2594 2595
 *  @oset: previous value of signal mask if non-null
 *  @sigsetsize: size of sigset_t type
 */
2596
SYSCALL_DEFINE4(rt_sigprocmask, int, how, sigset_t __user *, nset,
2597
		sigset_t __user *, oset, size_t, sigsetsize)
L
Linus Torvalds 已提交
2598 2599
{
	sigset_t old_set, new_set;
2600
	int error;
L
Linus Torvalds 已提交
2601 2602 2603

	/* XXX: Don't preclude handling different sized sigset_t's.  */
	if (sigsetsize != sizeof(sigset_t))
2604
		return -EINVAL;
L
Linus Torvalds 已提交
2605

2606 2607 2608 2609 2610
	old_set = current->blocked;

	if (nset) {
		if (copy_from_user(&new_set, nset, sizeof(sigset_t)))
			return -EFAULT;
L
Linus Torvalds 已提交
2611 2612
		sigdelsetmask(&new_set, sigmask(SIGKILL)|sigmask(SIGSTOP));

2613
		error = sigprocmask(how, &new_set, NULL);
L
Linus Torvalds 已提交
2614
		if (error)
2615 2616
			return error;
	}
L
Linus Torvalds 已提交
2617

2618 2619 2620
	if (oset) {
		if (copy_to_user(oset, &old_set, sizeof(sigset_t)))
			return -EFAULT;
L
Linus Torvalds 已提交
2621
	}
2622 2623

	return 0;
L
Linus Torvalds 已提交
2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647
}

long do_sigpending(void __user *set, unsigned long sigsetsize)
{
	long error = -EINVAL;
	sigset_t pending;

	if (sigsetsize > sizeof(sigset_t))
		goto out;

	spin_lock_irq(&current->sighand->siglock);
	sigorsets(&pending, &current->pending.signal,
		  &current->signal->shared_pending.signal);
	spin_unlock_irq(&current->sighand->siglock);

	/* Outside the lock because only this thread touches it.  */
	sigandsets(&pending, &current->blocked, &pending);

	error = -EFAULT;
	if (!copy_to_user(set, &pending, sigsetsize))
		error = 0;

out:
	return error;
2648
}
L
Linus Torvalds 已提交
2649

2650 2651 2652 2653 2654 2655
/**
 *  sys_rt_sigpending - examine a pending signal that has been raised
 *			while blocked
 *  @set: stores pending signals
 *  @sigsetsize: size of sigset_t type or larger
 */
2656
SYSCALL_DEFINE2(rt_sigpending, sigset_t __user *, set, size_t, sigsetsize)
L
Linus Torvalds 已提交
2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674
{
	return do_sigpending(set, sigsetsize);
}

#ifndef HAVE_ARCH_COPY_SIGINFO_TO_USER

int copy_siginfo_to_user(siginfo_t __user *to, siginfo_t *from)
{
	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.
2675 2676
	 * Please remember to update the signalfd_copyinfo() function
	 * inside fs/signalfd.c too, in case siginfo_t changes.
L
Linus Torvalds 已提交
2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701
	 * 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);
	err |= __put_user((short)from->si_code, &to->si_code);
	switch (from->si_code & __SI_MASK) {
	case __SI_KILL:
		err |= __put_user(from->si_pid, &to->si_pid);
		err |= __put_user(from->si_uid, &to->si_uid);
		break;
	case __SI_TIMER:
		 err |= __put_user(from->si_tid, &to->si_tid);
		 err |= __put_user(from->si_overrun, &to->si_overrun);
		 err |= __put_user(from->si_ptr, &to->si_ptr);
		break;
	case __SI_POLL:
		err |= __put_user(from->si_band, &to->si_band);
		err |= __put_user(from->si_fd, &to->si_fd);
		break;
	case __SI_FAULT:
		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
2702 2703
#endif
#ifdef BUS_MCEERR_AO
2704
		/*
A
Andi Kleen 已提交
2705
		 * Other callers might not initialize the si_lsb field,
2706
		 * so check explicitly for the right codes here.
A
Andi Kleen 已提交
2707 2708 2709
		 */
		if (from->si_code == BUS_MCEERR_AR || from->si_code == BUS_MCEERR_AO)
			err |= __put_user(from->si_addr_lsb, &to->si_addr_lsb);
L
Linus Torvalds 已提交
2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724
#endif
		break;
	case __SI_CHLD:
		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;
	case __SI_RT: /* This is not generated by the kernel as of now. */
	case __SI_MESGQ: /* But this is */
		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;
2725 2726 2727 2728 2729 2730 2731
#ifdef __ARCH_SIGSYS
	case __SI_SYS:
		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
2732 2733 2734 2735 2736 2737 2738 2739 2740 2741
	default: /* this is just in case for now ... */
		err |= __put_user(from->si_pid, &to->si_pid);
		err |= __put_user(from->si_uid, &to->si_uid);
		break;
	}
	return err;
}

#endif

2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779
/**
 *  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
 */
int do_sigtimedwait(const sigset_t *which, siginfo_t *info,
			const struct timespec *ts)
{
	struct task_struct *tsk = current;
	long timeout = MAX_SCHEDULE_TIMEOUT;
	sigset_t mask = *which;
	int sig;

	if (ts) {
		if (!timespec_valid(ts))
			return -EINVAL;
		timeout = timespec_to_jiffies(ts);
		/*
		 * We can be close to the next tick, add another one
		 * to ensure we will wait at least the time asked for.
		 */
		if (ts->tv_sec || ts->tv_nsec)
			timeout++;
	}

	/*
	 * 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);
	if (!sig && timeout) {
		/*
		 * None ready, temporarily unblock those we're interested
		 * while we are sleeping in so that we'll be awakened when
2780 2781
		 * they arrive. Unblocking is always fine, we can avoid
		 * set_current_blocked().
2782 2783 2784 2785 2786 2787 2788 2789 2790
		 */
		tsk->real_blocked = tsk->blocked;
		sigandsets(&tsk->blocked, &tsk->blocked, &mask);
		recalc_sigpending();
		spin_unlock_irq(&tsk->sighand->siglock);

		timeout = schedule_timeout_interruptible(timeout);

		spin_lock_irq(&tsk->sighand->siglock);
2791
		__set_task_blocked(tsk, &tsk->real_blocked);
2792
		siginitset(&tsk->real_blocked, 0);
2793
		sig = dequeue_signal(tsk, &mask, info);
2794 2795 2796 2797 2798 2799 2800 2801
	}
	spin_unlock_irq(&tsk->sighand->siglock);

	if (sig)
		return sig;
	return timeout ? -EINTR : -EAGAIN;
}

2802 2803 2804 2805 2806 2807 2808 2809
/**
 *  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
 */
2810 2811 2812
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
2813 2814 2815 2816
{
	sigset_t these;
	struct timespec ts;
	siginfo_t info;
2817
	int ret;
L
Linus Torvalds 已提交
2818 2819 2820 2821 2822 2823 2824

	/* 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;
2825

L
Linus Torvalds 已提交
2826 2827 2828 2829 2830
	if (uts) {
		if (copy_from_user(&ts, uts, sizeof(ts)))
			return -EFAULT;
	}

2831
	ret = do_sigtimedwait(&these, &info, uts ? &ts : NULL);
L
Linus Torvalds 已提交
2832

2833 2834 2835
	if (ret > 0 && uinfo) {
		if (copy_siginfo_to_user(uinfo, &info))
			ret = -EFAULT;
L
Linus Torvalds 已提交
2836 2837 2838 2839 2840
	}

	return ret;
}

2841 2842 2843 2844 2845
/**
 *  sys_kill - send a signal to a process
 *  @pid: the PID of the process
 *  @sig: signal to be sent
 */
2846
SYSCALL_DEFINE2(kill, pid_t, pid, int, sig)
L
Linus Torvalds 已提交
2847 2848 2849 2850 2851 2852
{
	struct siginfo info;

	info.si_signo = sig;
	info.si_errno = 0;
	info.si_code = SI_USER;
2853
	info.si_pid = task_tgid_vnr(current);
2854
	info.si_uid = from_kuid_munged(current_user_ns(), current_uid());
L
Linus Torvalds 已提交
2855 2856 2857 2858

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

T
Thomas Gleixner 已提交
2859 2860
static int
do_send_specific(pid_t tgid, pid_t pid, int sig, struct siginfo *info)
L
Linus Torvalds 已提交
2861 2862
{
	struct task_struct *p;
T
Thomas Gleixner 已提交
2863
	int error = -ESRCH;
L
Linus Torvalds 已提交
2864

2865
	rcu_read_lock();
2866
	p = find_task_by_vpid(pid);
2867
	if (p && (tgid <= 0 || task_tgid_vnr(p) == tgid)) {
T
Thomas Gleixner 已提交
2868
		error = check_kill_permission(sig, info, p);
L
Linus Torvalds 已提交
2869 2870 2871 2872
		/*
		 * The null signal is a permissions and process existence
		 * probe.  No signal is actually delivered.
		 */
2873 2874 2875 2876 2877 2878 2879 2880 2881
		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
2882 2883
		}
	}
2884
	rcu_read_unlock();
2885

L
Linus Torvalds 已提交
2886 2887 2888
	return error;
}

T
Thomas Gleixner 已提交
2889 2890 2891 2892 2893 2894 2895 2896
static int do_tkill(pid_t tgid, pid_t pid, int sig)
{
	struct siginfo info;

	info.si_signo = sig;
	info.si_errno = 0;
	info.si_code = SI_TKILL;
	info.si_pid = task_tgid_vnr(current);
2897
	info.si_uid = from_kuid_munged(current_user_ns(), current_uid());
T
Thomas Gleixner 已提交
2898 2899 2900 2901

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

2902 2903 2904 2905 2906 2907
/**
 *  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
 *
2908
 *  This syscall also checks the @tgid and returns -ESRCH even if the PID
2909 2910 2911
 *  exists but it's not belonging to the target process anymore. This
 *  method solves the problem of threads exiting and PIDs getting reused.
 */
2912
SYSCALL_DEFINE3(tgkill, pid_t, tgid, pid_t, pid, int, sig)
2913 2914 2915 2916 2917 2918 2919 2920
{
	/* This is only valid for single tasks */
	if (pid <= 0 || tgid <= 0)
		return -EINVAL;

	return do_tkill(tgid, pid, sig);
}

2921 2922 2923 2924 2925
/**
 *  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
2926 2927
 *  Send a signal to only one task, even if it's a CLONE_THREAD task.
 */
2928
SYSCALL_DEFINE2(tkill, pid_t, pid, int, sig)
L
Linus Torvalds 已提交
2929 2930 2931 2932 2933
{
	/* This is only valid for single tasks */
	if (pid <= 0)
		return -EINVAL;

2934
	return do_tkill(0, pid, sig);
L
Linus Torvalds 已提交
2935 2936
}

2937 2938 2939 2940 2941 2942
/**
 *  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
 */
2943 2944
SYSCALL_DEFINE3(rt_sigqueueinfo, pid_t, pid, int, sig,
		siginfo_t __user *, uinfo)
L
Linus Torvalds 已提交
2945 2946 2947 2948 2949 2950 2951
{
	siginfo_t info;

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

	/* Not even root can pretend to send signals from the kernel.
2952 2953
	 * Nor can they impersonate a kill()/tgkill(), which adds source info.
	 */
2954
	if (info.si_code >= 0 || info.si_code == SI_TKILL) {
2955 2956
		/* We used to allow any < 0 si_code */
		WARN_ON_ONCE(info.si_code < 0);
L
Linus Torvalds 已提交
2957
		return -EPERM;
2958
	}
L
Linus Torvalds 已提交
2959 2960 2961 2962 2963 2964
	info.si_signo = sig;

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

2965 2966 2967 2968 2969 2970 2971
long do_rt_tgsigqueueinfo(pid_t tgid, pid_t pid, int sig, siginfo_t *info)
{
	/* 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.
2972 2973
	 * Nor can they impersonate a kill()/tgkill(), which adds source info.
	 */
2974
	if (info->si_code >= 0 || info->si_code == SI_TKILL) {
2975 2976
		/* We used to allow any < 0 si_code */
		WARN_ON_ONCE(info->si_code < 0);
2977
		return -EPERM;
2978
	}
2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994
	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);
}

2995
int do_sigaction(int sig, struct k_sigaction *act, struct k_sigaction *oact)
L
Linus Torvalds 已提交
2996
{
2997
	struct task_struct *t = current;
L
Linus Torvalds 已提交
2998
	struct k_sigaction *k;
2999
	sigset_t mask;
L
Linus Torvalds 已提交
3000

3001
	if (!valid_signal(sig) || sig < 1 || (act && sig_kernel_only(sig)))
L
Linus Torvalds 已提交
3002 3003
		return -EINVAL;

3004
	k = &t->sighand->action[sig-1];
L
Linus Torvalds 已提交
3005 3006 3007 3008 3009 3010

	spin_lock_irq(&current->sighand->siglock);
	if (oact)
		*oact = *k;

	if (act) {
3011 3012
		sigdelsetmask(&act->sa.sa_mask,
			      sigmask(SIGKILL) | sigmask(SIGSTOP));
3013
		*k = *act;
L
Linus Torvalds 已提交
3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024
		/*
		 * 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"
		 */
3025
		if (sig_handler_ignored(sig_handler(t, sig), sig)) {
3026 3027 3028
			sigemptyset(&mask);
			sigaddset(&mask, sig);
			rm_from_queue_full(&mask, &t->signal->shared_pending);
L
Linus Torvalds 已提交
3029
			do {
3030
				rm_from_queue_full(&mask, &t->pending);
L
Linus Torvalds 已提交
3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045
				t = next_thread(t);
			} while (t != current);
		}
	}

	spin_unlock_irq(&current->sighand->siglock);
	return 0;
}

int 
do_sigaltstack (const stack_t __user *uss, stack_t __user *uoss, unsigned long sp)
{
	stack_t oss;
	int error;

3046 3047 3048
	oss.ss_sp = (void __user *) current->sas_ss_sp;
	oss.ss_size = current->sas_ss_size;
	oss.ss_flags = sas_ss_flags(sp);
L
Linus Torvalds 已提交
3049 3050 3051 3052 3053 3054 3055

	if (uss) {
		void __user *ss_sp;
		size_t ss_size;
		int ss_flags;

		error = -EFAULT;
L
Linus Torvalds 已提交
3056 3057 3058 3059 3060 3061
		if (!access_ok(VERIFY_READ, uss, sizeof(*uss)))
			goto out;
		error = __get_user(ss_sp, &uss->ss_sp) |
			__get_user(ss_flags, &uss->ss_flags) |
			__get_user(ss_size, &uss->ss_size);
		if (error)
L
Linus Torvalds 已提交
3062 3063 3064 3065 3066 3067 3068 3069
			goto out;

		error = -EPERM;
		if (on_sig_stack(sp))
			goto out;

		error = -EINVAL;
		/*
3070
		 * Note - this code used to test ss_flags incorrectly:
L
Linus Torvalds 已提交
3071 3072 3073
		 *  	  old code may have been written using ss_flags==0
		 *	  to mean ss_flags==SS_ONSTACK (as this was the only
		 *	  way that worked) - this fix preserves that older
3074
		 *	  mechanism.
L
Linus Torvalds 已提交
3075 3076 3077 3078 3079 3080 3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091
		 */
		if (ss_flags != SS_DISABLE && ss_flags != SS_ONSTACK && ss_flags != 0)
			goto out;

		if (ss_flags == SS_DISABLE) {
			ss_size = 0;
			ss_sp = NULL;
		} else {
			error = -ENOMEM;
			if (ss_size < MINSIGSTKSZ)
				goto out;
		}

		current->sas_ss_sp = (unsigned long) ss_sp;
		current->sas_ss_size = ss_size;
	}

3092
	error = 0;
L
Linus Torvalds 已提交
3093 3094
	if (uoss) {
		error = -EFAULT;
3095
		if (!access_ok(VERIFY_WRITE, uoss, sizeof(*uoss)))
L
Linus Torvalds 已提交
3096
			goto out;
3097 3098 3099
		error = __put_user(oss.ss_sp, &uoss->ss_sp) |
			__put_user(oss.ss_size, &uoss->ss_size) |
			__put_user(oss.ss_flags, &uoss->ss_flags);
L
Linus Torvalds 已提交
3100 3101 3102 3103 3104 3105
	}

out:
	return error;
}

A
Al Viro 已提交
3106 3107 3108 3109 3110 3111 3112
int restore_altstack(const stack_t __user *uss)
{
	int err = do_sigaltstack(uss, NULL, current_user_stack_pointer());
	/* squash all but EFAULT for now */
	return err == -EFAULT ? err : 0;
}

L
Linus Torvalds 已提交
3113 3114
#ifdef __ARCH_WANT_SYS_SIGPENDING

3115 3116 3117 3118
/**
 *  sys_sigpending - examine pending signals
 *  @set: where mask of pending signal is returned
 */
3119
SYSCALL_DEFINE1(sigpending, old_sigset_t __user *, set)
L
Linus Torvalds 已提交
3120 3121 3122 3123 3124 3125 3126
{
	return do_sigpending(set, sizeof(*set));
}

#endif

#ifdef __ARCH_WANT_SYS_SIGPROCMASK
3127 3128 3129
/**
 *  sys_sigprocmask - examine and change blocked signals
 *  @how: whether to add, remove, or set signals
3130
 *  @nset: signals to add or remove (if non-null)
3131 3132
 *  @oset: previous value of signal mask if non-null
 *
3133 3134 3135
 * Some platforms have their own version with special arguments;
 * others support only sys_rt_sigprocmask.
 */
L
Linus Torvalds 已提交
3136

3137
SYSCALL_DEFINE3(sigprocmask, int, how, old_sigset_t __user *, nset,
3138
		old_sigset_t __user *, oset)
L
Linus Torvalds 已提交
3139 3140
{
	old_sigset_t old_set, new_set;
3141
	sigset_t new_blocked;
L
Linus Torvalds 已提交
3142

3143
	old_set = current->blocked.sig[0];
L
Linus Torvalds 已提交
3144

3145 3146 3147
	if (nset) {
		if (copy_from_user(&new_set, nset, sizeof(*nset)))
			return -EFAULT;
L
Linus Torvalds 已提交
3148 3149
		new_set &= ~(sigmask(SIGKILL) | sigmask(SIGSTOP));

3150
		new_blocked = current->blocked;
L
Linus Torvalds 已提交
3151 3152 3153

		switch (how) {
		case SIG_BLOCK:
3154
			sigaddsetmask(&new_blocked, new_set);
L
Linus Torvalds 已提交
3155 3156
			break;
		case SIG_UNBLOCK:
3157
			sigdelsetmask(&new_blocked, new_set);
L
Linus Torvalds 已提交
3158 3159
			break;
		case SIG_SETMASK:
3160
			new_blocked.sig[0] = new_set;
L
Linus Torvalds 已提交
3161
			break;
3162 3163
		default:
			return -EINVAL;
L
Linus Torvalds 已提交
3164 3165
		}

3166
		__set_current_blocked(&new_blocked);
3167 3168 3169
	}

	if (oset) {
L
Linus Torvalds 已提交
3170
		if (copy_to_user(oset, &old_set, sizeof(*oset)))
3171
			return -EFAULT;
L
Linus Torvalds 已提交
3172
	}
3173 3174

	return 0;
L
Linus Torvalds 已提交
3175 3176 3177 3178
}
#endif /* __ARCH_WANT_SYS_SIGPROCMASK */

#ifdef __ARCH_WANT_SYS_RT_SIGACTION
3179 3180 3181
/**
 *  sys_rt_sigaction - alter an action taken by a process
 *  @sig: signal to be sent
3182 3183
 *  @act: new sigaction
 *  @oact: used to save the previous sigaction
3184 3185
 *  @sigsetsize: size of sigset_t type
 */
3186 3187 3188 3189
SYSCALL_DEFINE4(rt_sigaction, int, sig,
		const struct sigaction __user *, act,
		struct sigaction __user *, oact,
		size_t, sigsetsize)
L
Linus Torvalds 已提交
3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218
{
	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;
}
#endif /* __ARCH_WANT_SYS_RT_SIGACTION */

#ifdef __ARCH_WANT_SYS_SGETMASK

/*
 * For backwards compatibility.  Functionality superseded by sigprocmask.
 */
3219
SYSCALL_DEFINE0(sgetmask)
L
Linus Torvalds 已提交
3220 3221 3222 3223 3224
{
	/* SMP safe */
	return current->blocked.sig[0];
}

3225
SYSCALL_DEFINE1(ssetmask, int, newmask)
L
Linus Torvalds 已提交
3226
{
3227 3228
	int old = current->blocked.sig[0];
	sigset_t newset;
L
Linus Torvalds 已提交
3229

3230
	set_current_blocked(&newset);
L
Linus Torvalds 已提交
3231 3232 3233 3234 3235 3236 3237 3238 3239

	return old;
}
#endif /* __ARCH_WANT_SGETMASK */

#ifdef __ARCH_WANT_SYS_SIGNAL
/*
 * For backwards compatibility.  Functionality superseded by sigaction.
 */
3240
SYSCALL_DEFINE2(signal, int, sig, __sighandler_t, handler)
L
Linus Torvalds 已提交
3241 3242 3243 3244 3245 3246
{
	struct k_sigaction new_sa, old_sa;
	int ret;

	new_sa.sa.sa_handler = handler;
	new_sa.sa.sa_flags = SA_ONESHOT | SA_NOMASK;
3247
	sigemptyset(&new_sa.sa.sa_mask);
L
Linus Torvalds 已提交
3248 3249 3250 3251 3252 3253 3254 3255 3256

	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

3257
SYSCALL_DEFINE0(pause)
L
Linus Torvalds 已提交
3258
{
3259 3260 3261 3262
	while (!signal_pending(current)) {
		current->state = TASK_INTERRUPTIBLE;
		schedule();
	}
L
Linus Torvalds 已提交
3263 3264 3265 3266 3267
	return -ERESTARTNOHAND;
}

#endif

A
Al Viro 已提交
3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278
int sigsuspend(sigset_t *set)
{
	current->saved_sigmask = current->blocked;
	set_current_blocked(set);

	current->state = TASK_INTERRUPTIBLE;
	schedule();
	set_restore_sigmask();
	return -ERESTARTNOHAND;
}

3279
#ifdef __ARCH_WANT_SYS_RT_SIGSUSPEND
3280 3281 3282 3283 3284 3285
/**
 *  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
 */
3286
SYSCALL_DEFINE2(rt_sigsuspend, sigset_t __user *, unewset, size_t, sigsetsize)
3287 3288 3289 3290 3291 3292 3293 3294 3295
{
	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
3296
	return sigsuspend(&newset);
3297 3298 3299
}
#endif /* __ARCH_WANT_SYS_RT_SIGSUSPEND */

3300 3301 3302 3303 3304
__attribute__((weak)) const char *arch_vma_name(struct vm_area_struct *vma)
{
	return NULL;
}

L
Linus Torvalds 已提交
3305 3306
void __init signals_init(void)
{
3307
	sigqueue_cachep = KMEM_CACHE(sigqueue, SLAB_PANIC);
L
Linus Torvalds 已提交
3308
}
3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348

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