signal.c 87.5 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>
A
Al Viro 已提交
34
#include <linux/compat.h>
35 36
#define CREATE_TRACE_POINTS
#include <trace/events/signal.h>
37

L
Linus Torvalds 已提交
38 39 40 41
#include <asm/param.h>
#include <asm/uaccess.h>
#include <asm/unistd.h>
#include <asm/siginfo.h>
42
#include <asm/cacheflush.h>
43
#include "audit.h"	/* audit_signal_info() */
L
Linus Torvalds 已提交
44 45 46 47 48

/*
 * SLAB caches for signal bits.
 */

49
static struct kmem_cache *sigqueue_cachep;
L
Linus Torvalds 已提交
50

51 52
int print_fatal_signals __read_mostly;

53
static void __user *sig_handler(struct task_struct *t, int sig)
54
{
55 56
	return t->sighand->action[sig - 1].sa.sa_handler;
}
57

58 59
static int sig_handler_ignored(void __user *handler, int sig)
{
60 61 62 63
	/* Is it explicitly or implicitly ignored? */
	return handler == SIG_IGN ||
		(handler == SIG_DFL && sig_kernel_ignore(sig));
}
L
Linus Torvalds 已提交
64

65
static int sig_task_ignored(struct task_struct *t, int sig, bool force)
L
Linus Torvalds 已提交
66
{
67
	void __user *handler;
L
Linus Torvalds 已提交
68

69 70 71
	handler = sig_handler(t, sig);

	if (unlikely(t->signal->flags & SIGNAL_UNKILLABLE) &&
72
			handler == SIG_DFL && !force)
73 74 75 76 77
		return 1;

	return sig_handler_ignored(handler, sig);
}

78
static int sig_ignored(struct task_struct *t, int sig, bool force)
79
{
L
Linus Torvalds 已提交
80 81 82 83 84
	/*
	 * Blocked signals are never ignored, since the
	 * signal handler may change by the time it is
	 * unblocked.
	 */
85
	if (sigismember(&t->blocked, sig) || sigismember(&t->real_blocked, sig))
L
Linus Torvalds 已提交
86 87
		return 0;

88
	if (!sig_task_ignored(t, sig, force))
89 90 91 92 93
		return 0;

	/*
	 * Tracers may want to know about even ignored signals.
	 */
T
Tejun Heo 已提交
94
	return !t->ptrace;
L
Linus Torvalds 已提交
95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128
}

/*
 * 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
129
static int recalc_sigpending_tsk(struct task_struct *t)
L
Linus Torvalds 已提交
130
{
131
	if ((t->jobctl & JOBCTL_PENDING_MASK) ||
L
Linus Torvalds 已提交
132
	    PENDING(&t->pending, &t->blocked) ||
R
Roland McGrath 已提交
133
	    PENDING(&t->signal->shared_pending, &t->blocked)) {
L
Linus Torvalds 已提交
134
		set_tsk_thread_flag(t, TIF_SIGPENDING);
R
Roland McGrath 已提交
135 136
		return 1;
	}
137 138 139 140 141
	/*
	 * 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
142 143 144 145 146 147 148 149 150 151 152
	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
153 154 155 156
}

void recalc_sigpending(void)
{
T
Tejun Heo 已提交
157
	if (!recalc_sigpending_tsk(current) && !freezing(current))
158 159
		clear_thread_flag(TIF_SIGPENDING);

L
Linus Torvalds 已提交
160 161 162 163
}

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

164 165
#define SYNCHRONOUS_MASK \
	(sigmask(SIGSEGV) | sigmask(SIGBUS) | sigmask(SIGILL) | \
166
	 sigmask(SIGTRAP) | sigmask(SIGFPE) | sigmask(SIGSYS))
167

168
int next_signal(struct sigpending *pending, sigset_t *mask)
L
Linus Torvalds 已提交
169 170 171
{
	unsigned long i, *s, *m, x;
	int sig = 0;
172

L
Linus Torvalds 已提交
173 174
	s = pending->signal.sig;
	m = mask->sig;
175 176 177 178 179 180 181 182 183 184 185 186 187

	/*
	 * 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
188 189
	switch (_NSIG_WORDS) {
	default:
190 191 192 193 194 195 196
		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
197 198
		break;

199 200 201
	case 2:
		x = s[1] &~ m[1];
		if (!x)
L
Linus Torvalds 已提交
202
			break;
203
		sig = ffz(~x) + _NSIG_BPW + 1;
L
Linus Torvalds 已提交
204 205
		break;

206 207
	case 1:
		/* Nothing to do */
L
Linus Torvalds 已提交
208 209
		break;
	}
210

L
Linus Torvalds 已提交
211 212 213
	return sig;
}

214 215 216 217 218 219 220 221 222 223 224 225 226 227
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);
}

228
/**
229
 * task_set_jobctl_pending - set jobctl pending bits
230
 * @task: target task
231
 * @mask: pending bits to set
232
 *
233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260
 * 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;
}

261
/**
262
 * task_clear_jobctl_trapping - clear jobctl trapping bit
263 264
 * @task: target task
 *
265 266 267 268
 * 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.
269 270 271 272
 *
 * CONTEXT:
 * Must be called with @task->sighand->siglock held.
 */
273
void task_clear_jobctl_trapping(struct task_struct *task)
274
{
275 276
	if (unlikely(task->jobctl & JOBCTL_TRAPPING)) {
		task->jobctl &= ~JOBCTL_TRAPPING;
277
		wake_up_bit(&task->jobctl, JOBCTL_TRAPPING_BIT);
278 279 280
	}
}

281
/**
282
 * task_clear_jobctl_pending - clear jobctl pending bits
283
 * @task: target task
284
 * @mask: pending bits to clear
285
 *
286 287 288
 * 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.
289
 *
290 291
 * If clearing of @mask leaves no stop or trap pending, this function calls
 * task_clear_jobctl_trapping().
292 293 294 295
 *
 * CONTEXT:
 * Must be called with @task->sighand->siglock held.
 */
296
void task_clear_jobctl_pending(struct task_struct *task, unsigned int mask)
297
{
298 299 300 301 302 303
	BUG_ON(mask & ~JOBCTL_PENDING_MASK);

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

	task->jobctl &= ~mask;
304 305 306

	if (!(task->jobctl & JOBCTL_PENDING_MASK))
		task_clear_jobctl_trapping(task);
307 308 309 310 311 312
}

/**
 * task_participate_group_stop - participate in a group stop
 * @task: task participating in a group stop
 *
313
 * @task has %JOBCTL_STOP_PENDING set and is participating in a group stop.
314
 * Group stop states are cleared and the group stop count is consumed if
315
 * %JOBCTL_STOP_CONSUME was set.  If the consumption completes the group
316
 * stop, the appropriate %SIGNAL_* flags are set.
317 318 319
 *
 * CONTEXT:
 * Must be called with @task->sighand->siglock held.
320 321 322 323
 *
 * RETURNS:
 * %true if group stop completion should be notified to the parent, %false
 * otherwise.
324 325 326 327
 */
static bool task_participate_group_stop(struct task_struct *task)
{
	struct signal_struct *sig = task->signal;
328
	bool consume = task->jobctl & JOBCTL_STOP_CONSUME;
329

330
	WARN_ON_ONCE(!(task->jobctl & JOBCTL_STOP_PENDING));
331

332
	task_clear_jobctl_pending(task, JOBCTL_STOP_PENDING);
333 334 335 336 337 338 339

	if (!consume)
		return false;

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

340 341 342 343 344
	/*
	 * 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)) {
345 346 347 348 349 350
		sig->flags = SIGNAL_STOP_STOPPED;
		return true;
	}
	return false;
}

351 352 353
/*
 * allocate a new signal queue record
 * - this may be called without locks if and only if t == current, otherwise an
354
 *   appropriate lock must be held to stop the target task from exiting
355
 */
356 357
static struct sigqueue *
__sigqueue_alloc(int sig, struct task_struct *t, gfp_t flags, int override_rlimit)
L
Linus Torvalds 已提交
358 359
{
	struct sigqueue *q = NULL;
360
	struct user_struct *user;
L
Linus Torvalds 已提交
361

362
	/*
363 364
	 * Protect access to @t credentials. This can go away when all
	 * callers hold rcu read lock.
365
	 */
366
	rcu_read_lock();
D
David Howells 已提交
367
	user = get_uid(__task_cred(t)->user);
368
	atomic_inc(&user->sigpending);
369
	rcu_read_unlock();
370

L
Linus Torvalds 已提交
371
	if (override_rlimit ||
372
	    atomic_read(&user->sigpending) <=
373
			task_rlimit(t, RLIMIT_SIGPENDING)) {
L
Linus Torvalds 已提交
374
		q = kmem_cache_alloc(sigqueue_cachep, flags);
375 376 377 378
	} else {
		print_dropped_signal(sig);
	}

L
Linus Torvalds 已提交
379
	if (unlikely(q == NULL)) {
380
		atomic_dec(&user->sigpending);
D
David Howells 已提交
381
		free_uid(user);
L
Linus Torvalds 已提交
382 383 384
	} else {
		INIT_LIST_HEAD(&q->list);
		q->flags = 0;
D
David Howells 已提交
385
		q->user = user;
L
Linus Torvalds 已提交
386
	}
D
David Howells 已提交
387 388

	return q;
L
Linus Torvalds 已提交
389 390
}

391
static void __sigqueue_free(struct sigqueue *q)
L
Linus Torvalds 已提交
392 393 394 395 396 397 398 399
{
	if (q->flags & SIGQUEUE_PREALLOC)
		return;
	atomic_dec(&q->user->sigpending);
	free_uid(q->user);
	kmem_cache_free(sigqueue_cachep, q);
}

400
void flush_sigqueue(struct sigpending *queue)
L
Linus Torvalds 已提交
401 402 403 404 405 406 407 408 409 410 411 412 413 414
{
	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.
 */
415 416 417 418 419 420 421
void __flush_signals(struct task_struct *t)
{
	clear_tsk_thread_flag(t, TIF_SIGPENDING);
	flush_sigqueue(&t->pending);
	flush_sigqueue(&t->signal->shared_pending);
}

422
void flush_signals(struct task_struct *t)
L
Linus Torvalds 已提交
423 424 425 426
{
	unsigned long flags;

	spin_lock_irqsave(&t->sighand->siglock, flags);
427
	__flush_signals(t);
L
Linus Torvalds 已提交
428 429 430
	spin_unlock_irqrestore(&t->sighand->siglock, flags);
}

431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464
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);
}

465 466 467 468 469 470 471 472 473 474
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
475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492
/*
 * 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++;
	}
}

493 494
int unhandled_signal(struct task_struct *tsk, int sig)
{
495
	void __user *handler = tsk->sighand->action[sig-1].sa.sa_handler;
496
	if (is_global_init(tsk))
497
		return 1;
498
	if (handler != SIG_IGN && handler != SIG_DFL)
499
		return 0;
T
Tejun Heo 已提交
500 501
	/* if ptraced, let the tracer determine */
	return !tsk->ptrace;
502 503
}

504 505
/*
 * Notify the system that a driver wants to block all signals for this
L
Linus Torvalds 已提交
506 507 508 509 510
 * 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
511 512
 * can use to determine if the signal should be blocked or not.
 */
L
Linus Torvalds 已提交
513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538
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);
}

539
static void collect_signal(int sig, struct sigpending *list, siginfo_t *info)
L
Linus Torvalds 已提交
540 541 542 543 544 545 546 547 548
{
	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) {
549 550
			if (first)
				goto still_pending;
L
Linus Torvalds 已提交
551 552 553
			first = q;
		}
	}
554 555 556

	sigdelset(&list->signal, sig);

L
Linus Torvalds 已提交
557
	if (first) {
558
still_pending:
L
Linus Torvalds 已提交
559 560 561 562
		list_del_init(&first->list);
		copy_siginfo(info, &first->info);
		__sigqueue_free(first);
	} else {
563 564 565 566
		/*
		 * 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
567 568 569
		 */
		info->si_signo = sig;
		info->si_errno = 0;
570
		info->si_code = SI_USER;
L
Linus Torvalds 已提交
571 572 573 574 575 576 577 578
		info->si_pid = 0;
		info->si_uid = 0;
	}
}

static int __dequeue_signal(struct sigpending *pending, sigset_t *mask,
			siginfo_t *info)
{
579
	int sig = next_signal(pending, mask);
L
Linus Torvalds 已提交
580 581 582 583 584 585 586 587 588 589 590

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

591
		collect_signal(sig, pending, info);
L
Linus Torvalds 已提交
592 593 594 595 596 597
	}

	return sig;
}

/*
598
 * Dequeue a signal and return the element to the caller, which is
L
Linus Torvalds 已提交
599 600 601 602 603 604
 * expected to free it.
 *
 * All callers have to hold the siglock.
 */
int dequeue_signal(struct task_struct *tsk, sigset_t *mask, siginfo_t *info)
{
605
	int signr;
606 607 608 609

	/* We only dequeue private signals from ourselves, we don't let
	 * signalfd steal them
	 */
D
Davide Libenzi 已提交
610
	signr = __dequeue_signal(&tsk->pending, mask, info);
611
	if (!signr) {
L
Linus Torvalds 已提交
612 613
		signr = __dequeue_signal(&tsk->signal->shared_pending,
					 mask, info);
614 615 616 617 618 619
		/*
		 * 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
620
		 * compliant with the old way of self-restarting
621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637
		 * 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);
			}
		}
	}
638

D
Davide Libenzi 已提交
639
	recalc_sigpending();
640 641 642 643
	if (!signr)
		return 0;

	if (unlikely(sig_kernel_stop(signr))) {
644 645 646 647 648 649 650 651 652 653 654 655
		/*
		 * 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.
		 */
656
		current->jobctl |= JOBCTL_STOP_DEQUEUED;
657
	}
658
	if ((info->si_code & __SI_MASK) == __SI_TIMER && info->si_sys_private) {
L
Linus Torvalds 已提交
659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689
		/*
		 * 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
690 691
	 * 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
692 693 694 695 696 697
	 * 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
698
		mask |= TASK_WAKEKILL;
L
Linus Torvalds 已提交
699 700 701 702
	if (!wake_up_state(t, mask))
		kick_process(t);
}

703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720
/*
 * 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;

721
	sigandnsets(&s->signal, &s->signal, mask);
722 723 724 725 726 727 728 729
	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
730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753
/*
 * 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;
}

754 755 756 757 758 759 760 761 762 763 764
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));
}

765 766 767 768 769 770 771 772
/*
 * 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);

773 774 775 776
	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))
777 778
		return 1;

779
	if (ns_capable(tcred->user_ns, CAP_KILL))
780 781 782 783 784
		return 1;

	return 0;
}

L
Linus Torvalds 已提交
785 786
/*
 * Bad permissions for sending the signal
787
 * - the caller must hold the RCU read lock
L
Linus Torvalds 已提交
788 789 790 791
 */
static int check_kill_permission(int sig, struct siginfo *info,
				 struct task_struct *t)
{
792
	struct pid *sid;
793 794
	int error;

795
	if (!valid_signal(sig))
796 797
		return -EINVAL;

798
	if (!si_fromuser(info))
799
		return 0;
A
Amy Griffis 已提交
800

801 802
	error = audit_signal_info(sig, t); /* Let audit system see the signal */
	if (error)
L
Linus Torvalds 已提交
803
		return error;
804

805
	if (!same_thread_group(current, t) &&
806
	    !kill_ok_by_cred(t)) {
807 808 809 810 811 812 813 814 815 816 817 818 819
		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;
		}
	}
820

A
Amy Griffis 已提交
821
	return security_task_kill(t, info, sig, 0);
L
Linus Torvalds 已提交
822 823
}

824 825 826 827 828 829 830 831
/**
 * 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
832 833 834 835 836
 * 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.
837 838 839 840 841 842 843 844 845 846
 *
 * 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
847
	signal_wake_up(t, t->jobctl & JOBCTL_LISTENING);
848 849
}

L
Linus Torvalds 已提交
850
/*
851 852
 * 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
853 854
 * time regardless of blocking, ignoring, or handling.  This does the
 * actual continuing for SIGCONT, but not the actual stopping for stop
855 856 857 858
 * 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
859
 */
860
static int prepare_signal(int sig, struct task_struct *p, bool force)
L
Linus Torvalds 已提交
861
{
862
	struct signal_struct *signal = p->signal;
L
Linus Torvalds 已提交
863 864
	struct task_struct *t;

865
	if (unlikely(signal->flags & SIGNAL_GROUP_EXIT)) {
L
Linus Torvalds 已提交
866
		/*
867
		 * The process is in the middle of dying, nothing to do.
L
Linus Torvalds 已提交
868
		 */
869
	} else if (sig_kernel_stop(sig)) {
L
Linus Torvalds 已提交
870 871 872
		/*
		 * This is a stop signal.  Remove SIGCONT from all queues.
		 */
873
		rm_from_queue(sigmask(SIGCONT), &signal->shared_pending);
L
Linus Torvalds 已提交
874 875 876
		t = p;
		do {
			rm_from_queue(sigmask(SIGCONT), &t->pending);
877
		} while_each_thread(p, t);
L
Linus Torvalds 已提交
878
	} else if (sig == SIGCONT) {
879
		unsigned int why;
L
Linus Torvalds 已提交
880
		/*
881
		 * Remove all stop signals from all queues, wake all threads.
L
Linus Torvalds 已提交
882
		 */
883
		rm_from_queue(SIG_KERNEL_STOP_MASK, &signal->shared_pending);
L
Linus Torvalds 已提交
884 885
		t = p;
		do {
886
			task_clear_jobctl_pending(t, JOBCTL_STOP_PENDING);
L
Linus Torvalds 已提交
887
			rm_from_queue(SIG_KERNEL_STOP_MASK, &t->pending);
888 889 890 891
			if (likely(!(t->ptrace & PT_SEIZED)))
				wake_up_state(t, __TASK_STOPPED);
			else
				ptrace_trap_notify(t);
892
		} while_each_thread(p, t);
L
Linus Torvalds 已提交
893

894 895 896 897 898 899 900 901 902
		/*
		 * 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;
903
		if (signal->flags & SIGNAL_STOP_STOPPED)
904
			why |= SIGNAL_CLD_CONTINUED;
905
		else if (signal->group_stop_count)
906 907 908
			why |= SIGNAL_CLD_STOPPED;

		if (why) {
909
			/*
910
			 * The first thread which returns from do_signal_stop()
911 912 913
			 * will take ->siglock, notice SIGNAL_CLD_MASK, and
			 * notify its parent. See get_signal_to_deliver().
			 */
914 915 916
			signal->flags = why | SIGNAL_STOP_CONTINUED;
			signal->group_stop_count = 0;
			signal->group_exit_code = 0;
L
Linus Torvalds 已提交
917 918
		}
	}
919

920
	return !sig_ignored(p, sig, force);
L
Linus Torvalds 已提交
921 922
}

923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943
/*
 * 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);
}

944
static void complete_signal(int sig, struct task_struct *p, int group)
945 946 947 948 949 950 951 952 953 954 955 956
{
	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;
957
	else if (!group || thread_group_empty(p))
958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984
		/*
		 * 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.
	 */
985 986
	if (sig_fatal(p, sig) &&
	    !(signal->flags & (SIGNAL_UNKILLABLE | SIGNAL_GROUP_EXIT)) &&
987
	    !sigismember(&t->real_blocked, sig) &&
T
Tejun Heo 已提交
988
	    (sig == SIGKILL || !t->ptrace)) {
989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003
		/*
		 * 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 {
1004
				task_clear_jobctl_pending(t, JOBCTL_PENDING_MASK);
1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019
				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;
}

1020 1021 1022 1023 1024
static inline int legacy_queue(struct sigpending *signals, int sig)
{
	return (sig < SIGRTMIN) && sigismember(&signals->signal, sig);
}

1025 1026 1027 1028 1029 1030 1031 1032 1033
#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;

1034 1035 1036 1037
	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();
1038 1039 1040 1041 1042 1043 1044 1045
}
#else
static inline void userns_fixup_signal_uid(struct siginfo *info, struct task_struct *t)
{
	return;
}
#endif

1046 1047
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
1048
{
1049
	struct sigpending *pending;
1050
	struct sigqueue *q;
1051
	int override_rlimit;
1052
	int ret = 0, result;
1053

1054
	assert_spin_locked(&t->sighand->siglock);
1055

1056
	result = TRACE_SIGNAL_IGNORED;
1057 1058
	if (!prepare_signal(sig, t,
			from_ancestor_ns || (info == SEND_SIG_FORCED)))
1059
		goto ret;
1060 1061

	pending = group ? &t->signal->shared_pending : &t->pending;
1062 1063 1064 1065 1066
	/*
	 * 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.
	 */
1067
	result = TRACE_SIGNAL_ALREADY_PENDING;
1068
	if (legacy_queue(pending, sig))
1069 1070 1071
		goto ret;

	result = TRACE_SIGNAL_DELIVERED;
L
Linus Torvalds 已提交
1072 1073 1074 1075
	/*
	 * fast-pathed signals for kernel-internal things like SIGSTOP
	 * or SIGKILL.
	 */
1076
	if (info == SEND_SIG_FORCED)
L
Linus Torvalds 已提交
1077 1078
		goto out_set;

1079 1080 1081 1082 1083 1084 1085 1086 1087
	/*
	 * 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.
	 */
1088 1089 1090 1091 1092
	if (sig < SIGRTMIN)
		override_rlimit = (is_si_special(info) || info->si_code >= 0);
	else
		override_rlimit = 0;

1093
	q = __sigqueue_alloc(sig, t, GFP_ATOMIC | __GFP_NOTRACK_FALSE_POSITIVE,
1094
		override_rlimit);
L
Linus Torvalds 已提交
1095
	if (q) {
1096
		list_add_tail(&q->list, &pending->list);
L
Linus Torvalds 已提交
1097
		switch ((unsigned long) info) {
1098
		case (unsigned long) SEND_SIG_NOINFO:
L
Linus Torvalds 已提交
1099 1100 1101
			q->info.si_signo = sig;
			q->info.si_errno = 0;
			q->info.si_code = SI_USER;
1102
			q->info.si_pid = task_tgid_nr_ns(current,
1103
							task_active_pid_ns(t));
1104
			q->info.si_uid = from_kuid_munged(current_user_ns(), current_uid());
L
Linus Torvalds 已提交
1105
			break;
1106
		case (unsigned long) SEND_SIG_PRIV:
L
Linus Torvalds 已提交
1107 1108 1109 1110 1111 1112 1113 1114
			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);
1115 1116
			if (from_ancestor_ns)
				q->info.si_pid = 0;
L
Linus Torvalds 已提交
1117 1118
			break;
		}
1119 1120 1121

		userns_fixup_signal_uid(&q->info, t);

1122
	} else if (!is_si_special(info)) {
1123 1124 1125 1126 1127 1128
		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().
			 */
1129 1130 1131
			result = TRACE_SIGNAL_OVERFLOW_FAIL;
			ret = -EAGAIN;
			goto ret;
1132 1133 1134 1135 1136
		} else {
			/*
			 * This is a silent loss of information.  We still
			 * send the signal, but the *info bits are lost.
			 */
1137
			result = TRACE_SIGNAL_LOSE_INFO;
1138
		}
L
Linus Torvalds 已提交
1139 1140 1141
	}

out_set:
1142
	signalfd_notify(t, sig);
1143
	sigaddset(&pending->signal, sig);
1144
	complete_signal(sig, t, group);
1145 1146 1147
ret:
	trace_signal_generate(sig, info, t, group, result);
	return ret;
L
Linus Torvalds 已提交
1148 1149
}

1150 1151 1152
static int send_signal(int sig, struct siginfo *info, struct task_struct *t,
			int group)
{
1153 1154 1155
	int from_ancestor_ns = 0;

#ifdef CONFIG_PID_NS
1156 1157
	from_ancestor_ns = si_fromuser(info) &&
			   !task_pid_nr_ns(current, task_active_pid_ns(t));
1158 1159 1160
#endif

	return __send_signal(sig, info, t, group, from_ancestor_ns);
1161 1162
}

1163
static void print_fatal_signal(int signr)
I
Ingo Molnar 已提交
1164
{
1165
	struct pt_regs *regs = signal_pt_regs();
I
Ingo Molnar 已提交
1166
	printk("%s/%d: potentially unexpected fatal signal %d.\n",
1167
		current->comm, task_pid_nr(current), signr);
I
Ingo Molnar 已提交
1168

A
Al Viro 已提交
1169
#if defined(__i386__) && !defined(__arch_um__)
1170
	printk("code at %08lx: ", regs->ip);
I
Ingo Molnar 已提交
1171 1172 1173 1174 1175
	{
		int i;
		for (i = 0; i < 16; i++) {
			unsigned char insn;

1176 1177
			if (get_user(insn, (unsigned char *)(regs->ip + i)))
				break;
I
Ingo Molnar 已提交
1178 1179 1180 1181 1182
			printk("%02x ", insn);
		}
	}
#endif
	printk("\n");
1183
	preempt_disable();
I
Ingo Molnar 已提交
1184
	show_regs(regs);
1185
	preempt_enable();
I
Ingo Molnar 已提交
1186 1187 1188 1189 1190 1191 1192 1193 1194 1195
}

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
1196

1197 1198 1199 1200 1201 1202
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
1203 1204 1205
static int
specific_send_sig_info(int sig, struct siginfo *info, struct task_struct *t)
{
1206
	return send_signal(sig, info, t, 0);
L
Linus Torvalds 已提交
1207 1208
}

1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222
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
1223 1224 1225
/*
 * Force a signal that the process can't ignore: if necessary
 * we unblock the signal and change any SIG_IGN to SIG_DFL.
1226 1227 1228 1229 1230
 *
 * 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.
 *
1231 1232
 * 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
1233 1234 1235 1236 1237
 */
int
force_sig_info(int sig, struct siginfo *info, struct task_struct *t)
{
	unsigned long int flags;
1238 1239
	int ret, blocked, ignored;
	struct k_sigaction *action;
L
Linus Torvalds 已提交
1240 1241

	spin_lock_irqsave(&t->sighand->siglock, flags);
1242 1243 1244 1245 1246 1247 1248
	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
1249
			recalc_sigpending_and_wake(t);
1250
		}
L
Linus Torvalds 已提交
1251
	}
1252 1253
	if (action->sa.sa_handler == SIG_DFL)
		t->signal->flags &= ~SIGNAL_UNKILLABLE;
L
Linus Torvalds 已提交
1254 1255 1256 1257 1258 1259 1260 1261 1262
	ret = specific_send_sig_info(sig, info, t);
	spin_unlock_irqrestore(&t->sighand->siglock, flags);

	return ret;
}

/*
 * Nuke all other threads in the group.
 */
1263
int zap_other_threads(struct task_struct *p)
L
Linus Torvalds 已提交
1264
{
1265 1266
	struct task_struct *t = p;
	int count = 0;
L
Linus Torvalds 已提交
1267 1268 1269

	p->signal->group_stop_count = 0;

1270
	while_each_thread(p, t) {
1271
		task_clear_jobctl_pending(t, JOBCTL_PENDING_MASK);
1272 1273 1274
		count++;

		/* Don't bother with already dead threads */
L
Linus Torvalds 已提交
1275 1276 1277 1278 1279
		if (t->exit_state)
			continue;
		sigaddset(&t->pending.signal, SIGKILL);
		signal_wake_up(t, 1);
	}
1280 1281

	return count;
L
Linus Torvalds 已提交
1282 1283
}

1284 1285
struct sighand_struct *__lock_task_sighand(struct task_struct *tsk,
					   unsigned long *flags)
1286 1287 1288 1289
{
	struct sighand_struct *sighand;

	for (;;) {
1290 1291
		local_irq_save(*flags);
		rcu_read_lock();
1292
		sighand = rcu_dereference(tsk->sighand);
1293 1294 1295
		if (unlikely(sighand == NULL)) {
			rcu_read_unlock();
			local_irq_restore(*flags);
1296
			break;
1297
		}
1298

1299 1300 1301
		spin_lock(&sighand->siglock);
		if (likely(sighand == tsk->sighand)) {
			rcu_read_unlock();
1302
			break;
1303 1304 1305 1306
		}
		spin_unlock(&sighand->siglock);
		rcu_read_unlock();
		local_irq_restore(*flags);
1307 1308 1309 1310 1311
	}

	return sighand;
}

1312 1313 1314
/*
 * send signal info to all the members of a group
 */
L
Linus Torvalds 已提交
1315 1316
int group_send_sig_info(int sig, struct siginfo *info, struct task_struct *p)
{
1317 1318 1319 1320 1321
	int ret;

	rcu_read_lock();
	ret = check_kill_permission(sig, info, p);
	rcu_read_unlock();
1322

1323 1324
	if (!ret && sig)
		ret = do_send_sig_info(sig, info, p, true);
L
Linus Torvalds 已提交
1325 1326 1327 1328 1329

	return ret;
}

/*
1330
 * __kill_pgrp_info() sends a signal to a process group: this is what the tty
L
Linus Torvalds 已提交
1331
 * control characters do (^C, ^Z etc)
1332
 * - the caller must hold at least a readlock on tasklist_lock
L
Linus Torvalds 已提交
1333
 */
1334
int __kill_pgrp_info(int sig, struct siginfo *info, struct pid *pgrp)
L
Linus Torvalds 已提交
1335 1336 1337 1338 1339 1340
{
	struct task_struct *p = NULL;
	int retval, success;

	success = 0;
	retval = -ESRCH;
1341
	do_each_pid_task(pgrp, PIDTYPE_PGID, p) {
L
Linus Torvalds 已提交
1342 1343 1344
		int err = group_send_sig_info(sig, info, p);
		success |= !err;
		retval = err;
1345
	} while_each_pid_task(pgrp, PIDTYPE_PGID, p);
L
Linus Torvalds 已提交
1346 1347 1348
	return success ? 0 : retval;
}

1349
int kill_pid_info(int sig, struct siginfo *info, struct pid *pid)
L
Linus Torvalds 已提交
1350
{
1351
	int error = -ESRCH;
L
Linus Torvalds 已提交
1352 1353
	struct task_struct *p;

I
Ingo Molnar 已提交
1354
	rcu_read_lock();
1355
retry:
1356
	p = pid_task(pid, PIDTYPE_PID);
1357
	if (p) {
L
Linus Torvalds 已提交
1358
		error = group_send_sig_info(sig, info, p);
1359 1360 1361 1362 1363 1364 1365 1366 1367
		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
1368
	rcu_read_unlock();
1369

L
Linus Torvalds 已提交
1370 1371 1372
	return error;
}

1373
int kill_proc_info(int sig, struct siginfo *info, pid_t pid)
1374 1375 1376
{
	int error;
	rcu_read_lock();
1377
	error = kill_pid_info(sig, info, find_vpid(pid));
1378 1379 1380 1381
	rcu_read_unlock();
	return error;
}

1382 1383 1384 1385
static int kill_as_cred_perm(const struct cred *cred,
			     struct task_struct *target)
{
	const struct cred *pcred = __task_cred(target);
1386 1387
	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))
1388 1389 1390 1391
		return 0;
	return 1;
}

1392
/* like kill_pid_info(), but doesn't use uid/euid of "current" */
1393 1394
int kill_pid_info_as_cred(int sig, struct siginfo *info, struct pid *pid,
			 const struct cred *cred, u32 secid)
1395 1396 1397
{
	int ret = -EINVAL;
	struct task_struct *p;
1398
	unsigned long flags;
1399 1400 1401 1402

	if (!valid_signal(sig))
		return ret;

1403
	rcu_read_lock();
1404
	p = pid_task(pid, PIDTYPE_PID);
1405 1406 1407 1408
	if (!p) {
		ret = -ESRCH;
		goto out_unlock;
	}
1409
	if (si_fromuser(info) && !kill_as_cred_perm(cred, p)) {
1410 1411 1412
		ret = -EPERM;
		goto out_unlock;
	}
1413 1414 1415
	ret = security_task_kill(p, info, sig, secid);
	if (ret)
		goto out_unlock;
1416 1417 1418 1419 1420 1421 1422

	if (sig) {
		if (lock_task_sighand(p, &flags)) {
			ret = __send_signal(sig, info, p, 1, 0);
			unlock_task_sighand(p, &flags);
		} else
			ret = -ESRCH;
1423 1424
	}
out_unlock:
1425
	rcu_read_unlock();
1426 1427
	return ret;
}
1428
EXPORT_SYMBOL_GPL(kill_pid_info_as_cred);
L
Linus Torvalds 已提交
1429 1430 1431 1432 1433 1434 1435 1436

/*
 * 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.
 */

1437
static int kill_something_info(int sig, struct siginfo *info, pid_t pid)
L
Linus Torvalds 已提交
1438
{
1439
	int ret;
1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452

	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
1453 1454 1455 1456
		int retval = 0, count = 0;
		struct task_struct * p;

		for_each_process(p) {
1457 1458
			if (task_pid_vnr(p) > 1 &&
					!same_thread_group(p, current)) {
L
Linus Torvalds 已提交
1459 1460 1461 1462 1463 1464
				int err = group_send_sig_info(sig, info, p);
				++count;
				if (err != -EPERM)
					retval = err;
			}
		}
1465
		ret = count ? retval : -ESRCH;
L
Linus Torvalds 已提交
1466
	}
1467 1468
	read_unlock(&tasklist_lock);

1469
	return ret;
L
Linus Torvalds 已提交
1470 1471 1472 1473 1474 1475
}

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

1476
int send_sig_info(int sig, struct siginfo *info, struct task_struct *p)
L
Linus Torvalds 已提交
1477 1478 1479 1480 1481
{
	/*
	 * Make sure legacy kernel users don't send in bad values
	 * (normal paths check this in check_kill_permission).
	 */
1482
	if (!valid_signal(sig))
L
Linus Torvalds 已提交
1483 1484
		return -EINVAL;

1485
	return do_send_sig_info(sig, info, p, false);
L
Linus Torvalds 已提交
1486 1487
}

1488 1489 1490
#define __si_special(priv) \
	((priv) ? SEND_SIG_PRIV : SEND_SIG_NOINFO)

L
Linus Torvalds 已提交
1491 1492 1493
int
send_sig(int sig, struct task_struct *p, int priv)
{
1494
	return send_sig_info(sig, __si_special(priv), p);
L
Linus Torvalds 已提交
1495 1496 1497 1498 1499
}

void
force_sig(int sig, struct task_struct *p)
{
1500
	force_sig_info(sig, SEND_SIG_PRIV, p);
L
Linus Torvalds 已提交
1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521
}

/*
 * 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;
}

1522 1523
int kill_pgrp(struct pid *pid, int sig, int priv)
{
1524 1525 1526 1527 1528 1529 1530
	int ret;

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

	return ret;
1531 1532 1533 1534 1535 1536 1537 1538 1539
}
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
1540 1541 1542 1543
/*
 * 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
1544
 * expirations or I/O completions".  In the case of POSIX Timers
L
Linus Torvalds 已提交
1545 1546 1547 1548 1549 1550
 * 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)
{
1551
	struct sigqueue *q = __sigqueue_alloc(-1, current, GFP_KERNEL, 0);
L
Linus Torvalds 已提交
1552

1553
	if (q)
L
Linus Torvalds 已提交
1554
		q->flags |= SIGQUEUE_PREALLOC;
1555 1556

	return q;
L
Linus Torvalds 已提交
1557 1558 1559 1560 1561
}

void sigqueue_free(struct sigqueue *q)
{
	unsigned long flags;
1562 1563
	spinlock_t *lock = &current->sighand->siglock;

L
Linus Torvalds 已提交
1564 1565
	BUG_ON(!(q->flags & SIGQUEUE_PREALLOC));
	/*
1566 1567
	 * We must hold ->siglock while testing q->list
	 * to serialize with collect_signal() or with
1568
	 * __exit_signal()->flush_sigqueue().
L
Linus Torvalds 已提交
1569
	 */
1570
	spin_lock_irqsave(lock, flags);
1571 1572 1573 1574 1575
	q->flags &= ~SIGQUEUE_PREALLOC;
	/*
	 * If it is queued it will be freed when dequeued,
	 * like the "regular" sigqueue.
	 */
1576
	if (!list_empty(&q->list))
1577
		q = NULL;
1578 1579
	spin_unlock_irqrestore(lock, flags);

1580 1581
	if (q)
		__sigqueue_free(q);
L
Linus Torvalds 已提交
1582 1583
}

1584
int send_sigqueue(struct sigqueue *q, struct task_struct *t, int group)
1585
{
1586
	int sig = q->info.si_signo;
1587
	struct sigpending *pending;
1588
	unsigned long flags;
1589
	int ret, result;
1590

1591
	BUG_ON(!(q->flags & SIGQUEUE_PREALLOC));
1592 1593 1594 1595 1596

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

1597
	ret = 1; /* the signal is ignored */
1598
	result = TRACE_SIGNAL_IGNORED;
1599
	if (!prepare_signal(sig, t, false))
1600 1601 1602
		goto out;

	ret = 0;
1603 1604 1605 1606 1607 1608 1609
	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++;
1610
		result = TRACE_SIGNAL_ALREADY_PENDING;
1611
		goto out;
1612
	}
1613
	q->info.si_overrun = 0;
1614 1615

	signalfd_notify(t, sig);
1616
	pending = group ? &t->signal->shared_pending : &t->pending;
1617 1618
	list_add_tail(&q->list, &pending->list);
	sigaddset(&pending->signal, sig);
1619
	complete_signal(sig, t, group);
1620
	result = TRACE_SIGNAL_DELIVERED;
1621
out:
1622
	trace_signal_generate(sig, &q->info, t, group, result);
1623 1624 1625
	unlock_task_sighand(t, &flags);
ret:
	return ret;
1626 1627
}

L
Linus Torvalds 已提交
1628 1629 1630
/*
 * 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
1631
 *
1632 1633
 * Returns true if our parent ignored us and so we've switched to
 * self-reaping.
L
Linus Torvalds 已提交
1634
 */
1635
bool do_notify_parent(struct task_struct *tsk, int sig)
L
Linus Torvalds 已提交
1636 1637 1638 1639
{
	struct siginfo info;
	unsigned long flags;
	struct sighand_struct *psig;
1640
	bool autoreap = false;
L
Linus Torvalds 已提交
1641 1642 1643 1644

	BUG_ON(sig == -1);

 	/* do_notify_parent_cldstop should have been called instead.  */
M
Matthew Wilcox 已提交
1645
 	BUG_ON(task_is_stopped_or_traced(tsk));
L
Linus Torvalds 已提交
1646

T
Tejun Heo 已提交
1647
	BUG_ON(!tsk->ptrace &&
L
Linus Torvalds 已提交
1648 1649
	       (tsk->group_leader != tsk || !thread_group_empty(tsk)));

1650 1651 1652 1653 1654 1655 1656 1657 1658
	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
1659 1660
	info.si_signo = sig;
	info.si_errno = 0;
1661
	/*
1662 1663
	 * We are under tasklist_lock here so our parent is tied to
	 * us and cannot change.
1664
	 *
1665 1666
	 * task_active_pid_ns will always return the same pid namespace
	 * until a task passes through release_task.
1667 1668 1669 1670 1671 1672
	 *
	 * 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();
1673
	info.si_pid = task_pid_nr_ns(tsk, task_active_pid_ns(tsk->parent));
1674 1675
	info.si_uid = from_kuid_munged(task_cred_xxx(tsk->parent, user_ns),
				       task_uid(tsk));
1676 1677
	rcu_read_unlock();

1678 1679
	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
1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692

	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
1693
	if (!tsk->ptrace && sig == SIGCHLD &&
L
Linus Torvalds 已提交
1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710
	    (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).
		 */
1711
		autoreap = true;
L
Linus Torvalds 已提交
1712
		if (psig->action[SIGCHLD-1].sa.sa_handler == SIG_IGN)
1713
			sig = 0;
L
Linus Torvalds 已提交
1714
	}
1715
	if (valid_signal(sig) && sig)
L
Linus Torvalds 已提交
1716 1717 1718
		__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
1719

1720
	return autoreap;
L
Linus Torvalds 已提交
1721 1722
}

1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737
/**
 * 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
1738 1739 1740
{
	struct siginfo info;
	unsigned long flags;
1741
	struct task_struct *parent;
L
Linus Torvalds 已提交
1742 1743
	struct sighand_struct *sighand;

1744
	if (for_ptracer) {
1745
		parent = tsk->parent;
1746
	} else {
1747 1748 1749 1750
		tsk = tsk->group_leader;
		parent = tsk->real_parent;
	}

L
Linus Torvalds 已提交
1751 1752
	info.si_signo = SIGCHLD;
	info.si_errno = 0;
1753
	/*
1754
	 * see comment in do_notify_parent() about the following 4 lines
1755 1756
	 */
	rcu_read_lock();
1757
	info.si_pid = task_pid_nr_ns(tsk, parent->nsproxy->pid_ns);
1758
	info.si_uid = from_kuid_munged(task_cred_xxx(parent, user_ns), task_uid(tsk));
1759 1760
	rcu_read_unlock();

1761 1762
	info.si_utime = cputime_to_clock_t(tsk->utime);
	info.si_stime = cputime_to_clock_t(tsk->stime);
L
Linus Torvalds 已提交
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 1790

 	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);
}

1791 1792
static inline int may_ptrace_stop(void)
{
T
Tejun Heo 已提交
1793
	if (!likely(current->ptrace))
1794 1795 1796 1797 1798 1799 1800
		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
1801
	 * ->siglock we must see ->core_state != NULL. Otherwise it
1802 1803
	 * is safe to enter schedule().
	 */
1804
	if (unlikely(current->mm->core_state) &&
1805 1806 1807 1808 1809 1810
	    unlikely(current->mm == current->parent->mm))
		return 0;

	return 1;
}

R
Roland McGrath 已提交
1811
/*
1812
 * Return non-zero if there is a SIGKILL that should be waking us up.
R
Roland McGrath 已提交
1813 1814 1815 1816
 * Called with the siglock held.
 */
static int sigkill_pending(struct task_struct *tsk)
{
1817 1818
	return	sigismember(&tsk->pending.signal, SIGKILL) ||
		sigismember(&tsk->signal->shared_pending.signal, SIGKILL);
R
Roland McGrath 已提交
1819 1820
}

L
Linus Torvalds 已提交
1821 1822 1823 1824 1825 1826 1827 1828
/*
 * 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.
 *
1829 1830
 * 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
1831
 */
T
Tejun Heo 已提交
1832
static void ptrace_stop(int exit_code, int why, int clear_code, siginfo_t *info)
1833 1834
	__releases(&current->sighand->siglock)
	__acquires(&current->sighand->siglock)
L
Linus Torvalds 已提交
1835
{
1836 1837
	bool gstop_done = false;

R
Roland McGrath 已提交
1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852
	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);
1853 1854
		if (sigkill_pending(current))
			return;
R
Roland McGrath 已提交
1855 1856
	}

L
Linus Torvalds 已提交
1857
	/*
1858 1859 1860 1861 1862
	 * 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
1863
	 */
1864
	set_current_state(TASK_TRACED);
L
Linus Torvalds 已提交
1865 1866 1867 1868

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

1869
	/*
1870 1871
	 * If @why is CLD_STOPPED, we're trapping to participate in a group
	 * stop.  Do the bookkeeping.  Note that if SIGCONT was delievered
1872 1873 1874
	 * 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.
1875
	 */
1876
	if (why == CLD_STOPPED && (current->jobctl & JOBCTL_STOP_PENDING))
1877
		gstop_done = task_participate_group_stop(current);
1878

1879
	/* any trap clears pending STOP trap, STOP trap clears NOTIFY */
1880
	task_clear_jobctl_pending(current, JOBCTL_TRAP_STOP);
1881 1882
	if (info && info->si_code >> 8 == PTRACE_EVENT_STOP)
		task_clear_jobctl_pending(current, JOBCTL_TRAP_NOTIFY);
1883

1884
	/* entering a trap, clear TRAPPING */
1885
	task_clear_jobctl_trapping(current);
1886

L
Linus Torvalds 已提交
1887 1888
	spin_unlock_irq(&current->sighand->siglock);
	read_lock(&tasklist_lock);
1889
	if (may_ptrace_stop()) {
1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900
		/*
		 * 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);
1901
		if (gstop_done && ptrace_reparented(current))
1902 1903
			do_notify_parent_cldstop(current, false, why);

M
Miklos Szeredi 已提交
1904 1905 1906 1907 1908 1909 1910
		/*
		 * 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
1911
		read_unlock(&tasklist_lock);
M
Miklos Szeredi 已提交
1912
		preempt_enable_no_resched();
L
Linus Torvalds 已提交
1913 1914 1915 1916
		schedule();
	} else {
		/*
		 * By the time we got the lock, our tracer went away.
1917
		 * Don't drop the lock yet, another tracer may come.
1918 1919 1920
		 *
		 * If @gstop_done, the ptracer went away between group stop
		 * completion and here.  During detach, it would have set
1921 1922 1923
		 * 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
1924
		 */
1925 1926 1927
		if (gstop_done)
			do_notify_parent_cldstop(current, false, why);

1928
		__set_current_state(TASK_RUNNING);
1929 1930
		if (clear_code)
			current->exit_code = 0;
1931
		read_unlock(&tasklist_lock);
L
Linus Torvalds 已提交
1932 1933
	}

R
Roland McGrath 已提交
1934 1935 1936 1937 1938 1939 1940
	/*
	 * 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
1941 1942 1943 1944 1945 1946 1947 1948
	/*
	 * 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
1949 1950 1951
	/* LISTENING can be set only during STOP traps, clear it */
	current->jobctl &= ~JOBCTL_LISTENING;

L
Linus Torvalds 已提交
1952 1953 1954
	/*
	 * Queued signals ignored us while we were stopped for tracing.
	 * So check for any that we should take before resuming user mode.
1955
	 * This sets TIF_SIGPENDING, but never clears it.
L
Linus Torvalds 已提交
1956
	 */
1957
	recalc_sigpending_tsk(current);
L
Linus Torvalds 已提交
1958 1959
}

T
Tejun Heo 已提交
1960
static void ptrace_do_notify(int signr, int exit_code, int why)
L
Linus Torvalds 已提交
1961 1962 1963 1964
{
	siginfo_t info;

	memset(&info, 0, sizeof info);
T
Tejun Heo 已提交
1965
	info.si_signo = signr;
L
Linus Torvalds 已提交
1966
	info.si_code = exit_code;
1967
	info.si_pid = task_pid_vnr(current);
1968
	info.si_uid = from_kuid_munged(current_user_ns(), current_uid());
L
Linus Torvalds 已提交
1969 1970

	/* Let the debugger run.  */
T
Tejun Heo 已提交
1971 1972 1973 1974 1975 1976
	ptrace_stop(exit_code, why, 1, &info);
}

void ptrace_notify(int exit_code)
{
	BUG_ON((exit_code & (0x7f | ~0xffff)) != SIGTRAP);
1977 1978
	if (unlikely(current->task_works))
		task_work_run();
T
Tejun Heo 已提交
1979

L
Linus Torvalds 已提交
1980
	spin_lock_irq(&current->sighand->siglock);
T
Tejun Heo 已提交
1981
	ptrace_do_notify(SIGTRAP, exit_code, CLD_TRAPPED);
L
Linus Torvalds 已提交
1982 1983 1984
	spin_unlock_irq(&current->sighand->siglock);
}

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
/**
 * 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
2006
 */
2007 2008
static bool do_signal_stop(int signr)
	__releases(&current->sighand->siglock)
L
Linus Torvalds 已提交
2009 2010 2011
{
	struct signal_struct *sig = current->signal;

2012 2013
	if (!(current->jobctl & JOBCTL_STOP_PENDING)) {
		unsigned int gstop = JOBCTL_STOP_PENDING | JOBCTL_STOP_CONSUME;
2014 2015
		struct task_struct *t;

2016 2017
		/* signr will be recorded in task->jobctl for retries */
		WARN_ON_ONCE(signr & ~JOBCTL_STOP_SIGMASK);
2018

2019
		if (!likely(current->jobctl & JOBCTL_STOP_DEQUEUED) ||
2020
		    unlikely(signal_group_exit(sig)))
2021
			return false;
L
Linus Torvalds 已提交
2022
		/*
2023 2024 2025 2026 2027 2028 2029
		 * 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
2030 2031
		 * cause two group stops when !ptraced.  That is why we
		 * also check !task_is_stopped(t) below.
2032 2033 2034 2035 2036 2037 2038 2039
		 *
		 * 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
2040
		 */
2041 2042
		if (!(sig->flags & SIGNAL_STOP_STOPPED))
			sig->group_exit_code = signr;
L
Linus Torvalds 已提交
2043

2044 2045 2046 2047
		sig->group_stop_count = 0;

		if (task_set_jobctl_pending(current, signr | gstop))
			sig->group_stop_count++;
L
Linus Torvalds 已提交
2048

2049 2050
		for (t = next_thread(current); t != current;
		     t = next_thread(t)) {
L
Linus Torvalds 已提交
2051
			/*
2052 2053 2054
			 * 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
2055
			 */
2056 2057
			if (!task_is_stopped(t) &&
			    task_set_jobctl_pending(t, signr | gstop)) {
2058
				sig->group_stop_count++;
2059 2060 2061 2062
				if (likely(!(t->ptrace & PT_SEIZED)))
					signal_wake_up(t, 0);
				else
					ptrace_trap_notify(t);
2063
			}
2064
		}
L
Linus Torvalds 已提交
2065
	}
2066

T
Tejun Heo 已提交
2067
	if (likely(!current->ptrace)) {
2068
		int notify = 0;
L
Linus Torvalds 已提交
2069

2070 2071 2072 2073 2074 2075 2076 2077
		/*
		 * 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;

2078
		__set_current_state(TASK_STOPPED);
2079 2080
		spin_unlock_irq(&current->sighand->siglock);

2081 2082 2083 2084 2085 2086 2087 2088 2089
		/*
		 * 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.
		 */
2090 2091
		if (notify) {
			read_lock(&tasklist_lock);
2092
			do_notify_parent_cldstop(current, false, notify);
2093 2094 2095 2096 2097
			read_unlock(&tasklist_lock);
		}

		/* Now we don't run again until woken by SIGCONT or SIGKILL */
		schedule();
2098
		return true;
2099
	} else {
2100 2101 2102 2103 2104 2105
		/*
		 * 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;
2106
	}
2107
}
L
Linus Torvalds 已提交
2108

2109 2110 2111
/**
 * do_jobctl_trap - take care of ptrace jobctl traps
 *
T
Tejun Heo 已提交
2112 2113 2114 2115 2116 2117 2118
 * 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.
2119 2120 2121 2122 2123 2124 2125
 *
 * 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
2126
	struct signal_struct *signal = current->signal;
2127
	int signr = current->jobctl & JOBCTL_STOP_SIGMASK;
2128

T
Tejun Heo 已提交
2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139
	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;
2140
	}
L
Linus Torvalds 已提交
2141 2142
}

2143
static int ptrace_signal(int signr, siginfo_t *info)
R
Roland McGrath 已提交
2144
{
2145
	ptrace_signal_deliver();
2146 2147 2148 2149 2150 2151 2152 2153 2154 2155
	/*
	 * 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
2156
	ptrace_stop(signr, CLD_TRAPPED, 0, info);
R
Roland McGrath 已提交
2157 2158 2159 2160 2161 2162 2163 2164

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

	current->exit_code = 0;

2165 2166 2167 2168 2169 2170
	/*
	 * 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
2171 2172 2173 2174
	if (signr != info->si_signo) {
		info->si_signo = signr;
		info->si_errno = 0;
		info->si_code = SI_USER;
2175
		rcu_read_lock();
R
Roland McGrath 已提交
2176
		info->si_pid = task_pid_vnr(current->parent);
2177 2178
		info->si_uid = from_kuid_munged(current_user_ns(),
						task_uid(current->parent));
2179
		rcu_read_unlock();
R
Roland McGrath 已提交
2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190
	}

	/* 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
2191 2192 2193
int get_signal_to_deliver(siginfo_t *info, struct k_sigaction *return_ka,
			  struct pt_regs *regs, void *cookie)
{
2194 2195 2196
	struct sighand_struct *sighand = current->sighand;
	struct signal_struct *signal = current->signal;
	int signr;
L
Linus Torvalds 已提交
2197

2198 2199
	if (unlikely(current->task_works))
		task_work_run();
2200

2201 2202 2203
	if (unlikely(uprobe_deny_signal()))
		return 0;

R
Roland McGrath 已提交
2204 2205 2206 2207 2208 2209 2210
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.
	 */
2211 2212
	try_to_freeze();

2213
	spin_lock_irq(&sighand->siglock);
2214 2215 2216 2217 2218
	/*
	 * 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.
	 */
2219
	if (unlikely(signal->flags & SIGNAL_CLD_MASK)) {
2220 2221 2222 2223 2224 2225 2226
		int why;

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

2227
		signal->flags &= ~SIGNAL_CLD_MASK;
2228

2229
		spin_unlock_irq(&sighand->siglock);
R
Roland McGrath 已提交
2230

2231 2232 2233 2234 2235 2236 2237 2238
		/*
		 * 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
2239
		read_lock(&tasklist_lock);
2240 2241
		do_notify_parent_cldstop(current, false, why);

2242 2243 2244
		if (ptrace_reparented(current->group_leader))
			do_notify_parent_cldstop(current->group_leader,
						true, why);
T
Tejun Heo 已提交
2245
		read_unlock(&tasklist_lock);
2246

2247 2248 2249
		goto relock;
	}

L
Linus Torvalds 已提交
2250 2251
	for (;;) {
		struct k_sigaction *ka;
2252

T
Tejun Heo 已提交
2253 2254
		if (unlikely(current->jobctl & JOBCTL_STOP_PENDING) &&
		    do_signal_stop(0))
2255
			goto relock;
2256

2257 2258 2259 2260 2261
		if (unlikely(current->jobctl & JOBCTL_TRAP_MASK)) {
			do_jobctl_trap();
			spin_unlock_irq(&sighand->siglock);
			goto relock;
		}
L
Linus Torvalds 已提交
2262

T
Tejun Heo 已提交
2263
		signr = dequeue_signal(current, &current->blocked, info);
2264

T
Tejun Heo 已提交
2265 2266
		if (!signr)
			break; /* will return 0 */
2267

2268
		if (unlikely(current->ptrace) && signr != SIGKILL) {
2269
			signr = ptrace_signal(signr, info);
T
Tejun Heo 已提交
2270 2271
			if (!signr)
				continue;
L
Linus Torvalds 已提交
2272 2273
		}

T
Tejun Heo 已提交
2274 2275
		ka = &sighand->action[signr-1];

2276 2277 2278
		/* Trace actually delivered signals. */
		trace_signal_deliver(signr, info, ka);

L
Linus Torvalds 已提交
2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296
		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;

2297
		/*
2298
		 * Global init gets no signals it doesn't want.
2299 2300 2301 2302 2303 2304 2305
		 * 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.
2306
		 */
2307
		if (unlikely(signal->flags & SIGNAL_UNKILLABLE) &&
2308
				!sig_kernel_only(signr))
L
Linus Torvalds 已提交
2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322
			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) {
2323
				spin_unlock_irq(&sighand->siglock);
L
Linus Torvalds 已提交
2324 2325 2326

				/* signals can be posted during this window */

2327
				if (is_current_pgrp_orphaned())
L
Linus Torvalds 已提交
2328 2329
					goto relock;

2330
				spin_lock_irq(&sighand->siglock);
L
Linus Torvalds 已提交
2331 2332
			}

2333
			if (likely(do_signal_stop(info->si_signo))) {
L
Linus Torvalds 已提交
2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344
				/* It released the siglock.  */
				goto relock;
			}

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

2345
		spin_unlock_irq(&sighand->siglock);
L
Linus Torvalds 已提交
2346 2347 2348 2349 2350

		/*
		 * Anything else is fatal, maybe with a core dump.
		 */
		current->flags |= PF_SIGNALED;
2351

L
Linus Torvalds 已提交
2352
		if (sig_kernel_coredump(signr)) {
2353
			if (print_fatal_signals)
2354
				print_fatal_signal(info->si_signo);
L
Linus Torvalds 已提交
2355 2356 2357 2358 2359 2360 2361 2362
			/*
			 * 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.
			 */
2363
			do_coredump(info);
L
Linus Torvalds 已提交
2364 2365 2366 2367 2368
		}

		/*
		 * Death signals, no core dump.
		 */
2369
		do_group_exit(info->si_signo);
L
Linus Torvalds 已提交
2370 2371
		/* NOTREACHED */
	}
2372
	spin_unlock_irq(&sighand->siglock);
L
Linus Torvalds 已提交
2373 2374 2375
	return signr;
}

2376
/**
A
Al Viro 已提交
2377 2378 2379 2380 2381 2382
 * 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
2383 2384
 *
 * This function should be called when a signal has succesfully been
A
Al Viro 已提交
2385 2386 2387
 * 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.
2388
 */
A
Al Viro 已提交
2389 2390
void signal_delivered(int sig, siginfo_t *info, struct k_sigaction *ka,
			struct pt_regs *regs, int stepping)
2391 2392 2393
{
	sigset_t blocked;

2394 2395 2396 2397 2398 2399
	/* 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();

2400 2401
	sigorsets(&blocked, &current->blocked, &ka->sa.sa_mask);
	if (!(ka->sa.sa_flags & SA_NODEFER))
A
Al Viro 已提交
2402
		sigaddset(&blocked, sig);
2403
	set_current_blocked(&blocked);
A
Al Viro 已提交
2404
	tracehook_signal_handler(sig, info, ka, regs, stepping);
2405 2406
}

2407 2408
/*
 * It could be that complete_signal() picked us to notify about the
2409 2410
 * group-wide signal. Other threads should be notified now to take
 * the shared signals in @which since we will not.
2411
 */
2412
static void retarget_shared_pending(struct task_struct *tsk, sigset_t *which)
2413
{
2414
	sigset_t retarget;
2415 2416
	struct task_struct *t;

2417 2418 2419 2420
	sigandsets(&retarget, &tsk->signal->shared_pending.signal, which);
	if (sigisemptyset(&retarget))
		return;

2421 2422
	t = tsk;
	while_each_thread(tsk, t) {
2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435
		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;
2436 2437 2438
	}
}

O
Oleg Nesterov 已提交
2439 2440 2441
void exit_signals(struct task_struct *tsk)
{
	int group_stop = 0;
2442
	sigset_t unblocked;
O
Oleg Nesterov 已提交
2443

2444 2445 2446 2447 2448 2449
	/*
	 * @tsk is about to have PF_EXITING set - lock out users which
	 * expect stable threadgroup.
	 */
	threadgroup_change_begin(tsk);

2450 2451
	if (thread_group_empty(tsk) || signal_group_exit(tsk->signal)) {
		tsk->flags |= PF_EXITING;
2452
		threadgroup_change_end(tsk);
2453
		return;
O
Oleg Nesterov 已提交
2454 2455
	}

2456
	spin_lock_irq(&tsk->sighand->siglock);
O
Oleg Nesterov 已提交
2457 2458 2459 2460 2461
	/*
	 * From now this task is not visible for group-wide signals,
	 * see wants_signal(), do_signal_stop().
	 */
	tsk->flags |= PF_EXITING;
2462 2463 2464

	threadgroup_change_end(tsk);

2465 2466 2467
	if (!signal_pending(tsk))
		goto out;

2468 2469 2470
	unblocked = tsk->blocked;
	signotset(&unblocked);
	retarget_shared_pending(tsk, &unblocked);
2471

2472
	if (unlikely(tsk->jobctl & JOBCTL_STOP_PENDING) &&
2473
	    task_participate_group_stop(tsk))
T
Tejun Heo 已提交
2474
		group_stop = CLD_STOPPED;
2475
out:
O
Oleg Nesterov 已提交
2476 2477
	spin_unlock_irq(&tsk->sighand->siglock);

2478 2479 2480 2481
	/*
	 * If group stop has completed, deliver the notification.  This
	 * should always go to the real parent of the group leader.
	 */
2482
	if (unlikely(group_stop)) {
O
Oleg Nesterov 已提交
2483
		read_lock(&tasklist_lock);
2484
		do_notify_parent_cldstop(tsk, false, group_stop);
O
Oleg Nesterov 已提交
2485 2486 2487 2488
		read_unlock(&tasklist_lock);
	}
}

L
Linus Torvalds 已提交
2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503
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.
 */

2504 2505 2506
/**
 *  sys_restart_syscall - restart a system call
 */
2507
SYSCALL_DEFINE0(restart_syscall)
L
Linus Torvalds 已提交
2508 2509 2510 2511 2512 2513 2514 2515 2516 2517
{
	struct restart_block *restart = &current_thread_info()->restart_block;
	return restart->fn(restart);
}

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

2518 2519 2520 2521 2522
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. */
2523
		sigandnsets(&newblocked, newset, &current->blocked);
2524 2525 2526 2527 2528 2529
		retarget_shared_pending(tsk, &newblocked);
	}
	tsk->blocked = *newset;
	recalc_sigpending();
}

2530 2531 2532 2533 2534 2535
/**
 * 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
2536
 */
2537 2538 2539 2540 2541 2542 2543 2544 2545 2546
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)
2547 2548 2549 2550
{
	struct task_struct *tsk = current;

	spin_lock_irq(&tsk->sighand->siglock);
2551
	__set_task_blocked(tsk, newset);
2552 2553
	spin_unlock_irq(&tsk->sighand->siglock);
}
L
Linus Torvalds 已提交
2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564

/*
 * 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)
{
2565 2566
	struct task_struct *tsk = current;
	sigset_t newset;
L
Linus Torvalds 已提交
2567

2568
	/* Lockless, only current can change ->blocked, never from irq */
2569
	if (oldset)
2570
		*oldset = tsk->blocked;
2571

L
Linus Torvalds 已提交
2572 2573
	switch (how) {
	case SIG_BLOCK:
2574
		sigorsets(&newset, &tsk->blocked, set);
L
Linus Torvalds 已提交
2575 2576
		break;
	case SIG_UNBLOCK:
2577
		sigandnsets(&newset, &tsk->blocked, set);
L
Linus Torvalds 已提交
2578 2579
		break;
	case SIG_SETMASK:
2580
		newset = *set;
L
Linus Torvalds 已提交
2581 2582
		break;
	default:
2583
		return -EINVAL;
L
Linus Torvalds 已提交
2584
	}
2585

2586
	__set_current_blocked(&newset);
2587
	return 0;
L
Linus Torvalds 已提交
2588 2589
}

2590 2591 2592
/**
 *  sys_rt_sigprocmask - change the list of currently blocked signals
 *  @how: whether to add, remove, or set signals
R
Randy Dunlap 已提交
2593
 *  @nset: stores pending signals
2594 2595 2596
 *  @oset: previous value of signal mask if non-null
 *  @sigsetsize: size of sigset_t type
 */
2597
SYSCALL_DEFINE4(rt_sigprocmask, int, how, sigset_t __user *, nset,
2598
		sigset_t __user *, oset, size_t, sigsetsize)
L
Linus Torvalds 已提交
2599 2600
{
	sigset_t old_set, new_set;
2601
	int error;
L
Linus Torvalds 已提交
2602 2603 2604

	/* XXX: Don't preclude handling different sized sigset_t's.  */
	if (sigsetsize != sizeof(sigset_t))
2605
		return -EINVAL;
L
Linus Torvalds 已提交
2606

2607 2608 2609 2610 2611
	old_set = current->blocked;

	if (nset) {
		if (copy_from_user(&new_set, nset, sizeof(sigset_t)))
			return -EFAULT;
L
Linus Torvalds 已提交
2612 2613
		sigdelsetmask(&new_set, sigmask(SIGKILL)|sigmask(SIGSTOP));

2614
		error = sigprocmask(how, &new_set, NULL);
L
Linus Torvalds 已提交
2615
		if (error)
2616 2617
			return error;
	}
L
Linus Torvalds 已提交
2618

2619 2620 2621
	if (oset) {
		if (copy_to_user(oset, &old_set, sizeof(sigset_t)))
			return -EFAULT;
L
Linus Torvalds 已提交
2622
	}
2623 2624

	return 0;
L
Linus Torvalds 已提交
2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648
}

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;
2649
}
L
Linus Torvalds 已提交
2650

2651 2652 2653 2654 2655 2656
/**
 *  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
 */
2657
SYSCALL_DEFINE2(rt_sigpending, sigset_t __user *, set, size_t, sigsetsize)
L
Linus Torvalds 已提交
2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675
{
	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.
2676 2677
	 * Please remember to update the signalfd_copyinfo() function
	 * inside fs/signalfd.c too, in case siginfo_t changes.
L
Linus Torvalds 已提交
2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702
	 * 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
2703 2704
#endif
#ifdef BUS_MCEERR_AO
2705
		/*
A
Andi Kleen 已提交
2706
		 * Other callers might not initialize the si_lsb field,
2707
		 * so check explicitly for the right codes here.
A
Andi Kleen 已提交
2708 2709 2710
		 */
		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
2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725
#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;
2726 2727 2728 2729 2730 2731 2732
#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
2733 2734 2735 2736 2737 2738 2739 2740 2741 2742
	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

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 2780
/**
 *  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
2781 2782
		 * they arrive. Unblocking is always fine, we can avoid
		 * set_current_blocked().
2783 2784 2785 2786 2787 2788 2789 2790 2791
		 */
		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);
2792
		__set_task_blocked(tsk, &tsk->real_blocked);
2793
		siginitset(&tsk->real_blocked, 0);
2794
		sig = dequeue_signal(tsk, &mask, info);
2795 2796 2797 2798 2799 2800 2801 2802
	}
	spin_unlock_irq(&tsk->sighand->siglock);

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

2803 2804 2805 2806 2807 2808 2809 2810
/**
 *  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
 */
2811 2812 2813
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
2814 2815 2816 2817
{
	sigset_t these;
	struct timespec ts;
	siginfo_t info;
2818
	int ret;
L
Linus Torvalds 已提交
2819 2820 2821 2822 2823 2824 2825

	/* 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;
2826

L
Linus Torvalds 已提交
2827 2828 2829 2830 2831
	if (uts) {
		if (copy_from_user(&ts, uts, sizeof(ts)))
			return -EFAULT;
	}

2832
	ret = do_sigtimedwait(&these, &info, uts ? &ts : NULL);
L
Linus Torvalds 已提交
2833

2834 2835 2836
	if (ret > 0 && uinfo) {
		if (copy_siginfo_to_user(uinfo, &info))
			ret = -EFAULT;
L
Linus Torvalds 已提交
2837 2838 2839 2840 2841
	}

	return ret;
}

2842 2843 2844 2845 2846
/**
 *  sys_kill - send a signal to a process
 *  @pid: the PID of the process
 *  @sig: signal to be sent
 */
2847
SYSCALL_DEFINE2(kill, pid_t, pid, int, sig)
L
Linus Torvalds 已提交
2848 2849 2850 2851 2852 2853
{
	struct siginfo info;

	info.si_signo = sig;
	info.si_errno = 0;
	info.si_code = SI_USER;
2854
	info.si_pid = task_tgid_vnr(current);
2855
	info.si_uid = from_kuid_munged(current_user_ns(), current_uid());
L
Linus Torvalds 已提交
2856 2857 2858 2859

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

T
Thomas Gleixner 已提交
2860 2861
static int
do_send_specific(pid_t tgid, pid_t pid, int sig, struct siginfo *info)
L
Linus Torvalds 已提交
2862 2863
{
	struct task_struct *p;
T
Thomas Gleixner 已提交
2864
	int error = -ESRCH;
L
Linus Torvalds 已提交
2865

2866
	rcu_read_lock();
2867
	p = find_task_by_vpid(pid);
2868
	if (p && (tgid <= 0 || task_tgid_vnr(p) == tgid)) {
T
Thomas Gleixner 已提交
2869
		error = check_kill_permission(sig, info, p);
L
Linus Torvalds 已提交
2870 2871 2872 2873
		/*
		 * The null signal is a permissions and process existence
		 * probe.  No signal is actually delivered.
		 */
2874 2875 2876 2877 2878 2879 2880 2881 2882
		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
2883 2884
		}
	}
2885
	rcu_read_unlock();
2886

L
Linus Torvalds 已提交
2887 2888 2889
	return error;
}

T
Thomas Gleixner 已提交
2890 2891 2892 2893 2894 2895 2896 2897
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);
2898
	info.si_uid = from_kuid_munged(current_user_ns(), current_uid());
T
Thomas Gleixner 已提交
2899 2900 2901 2902

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

2903 2904 2905 2906 2907 2908
/**
 *  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
 *
2909
 *  This syscall also checks the @tgid and returns -ESRCH even if the PID
2910 2911 2912
 *  exists but it's not belonging to the target process anymore. This
 *  method solves the problem of threads exiting and PIDs getting reused.
 */
2913
SYSCALL_DEFINE3(tgkill, pid_t, tgid, pid_t, pid, int, sig)
2914 2915 2916 2917 2918 2919 2920 2921
{
	/* This is only valid for single tasks */
	if (pid <= 0 || tgid <= 0)
		return -EINVAL;

	return do_tkill(tgid, pid, sig);
}

2922 2923 2924 2925 2926
/**
 *  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
2927 2928
 *  Send a signal to only one task, even if it's a CLONE_THREAD task.
 */
2929
SYSCALL_DEFINE2(tkill, pid_t, pid, int, sig)
L
Linus Torvalds 已提交
2930 2931 2932 2933 2934
{
	/* This is only valid for single tasks */
	if (pid <= 0)
		return -EINVAL;

2935
	return do_tkill(0, pid, sig);
L
Linus Torvalds 已提交
2936 2937
}

2938 2939 2940 2941 2942 2943
/**
 *  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
 */
2944 2945
SYSCALL_DEFINE3(rt_sigqueueinfo, pid_t, pid, int, sig,
		siginfo_t __user *, uinfo)
L
Linus Torvalds 已提交
2946 2947 2948 2949 2950 2951 2952
{
	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.
2953 2954
	 * Nor can they impersonate a kill()/tgkill(), which adds source info.
	 */
2955
	if (info.si_code >= 0 || info.si_code == SI_TKILL) {
2956 2957
		/* We used to allow any < 0 si_code */
		WARN_ON_ONCE(info.si_code < 0);
L
Linus Torvalds 已提交
2958
		return -EPERM;
2959
	}
L
Linus Torvalds 已提交
2960 2961 2962 2963 2964 2965
	info.si_signo = sig;

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

2966 2967 2968 2969 2970 2971 2972
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.
2973 2974
	 * Nor can they impersonate a kill()/tgkill(), which adds source info.
	 */
2975
	if (info->si_code >= 0 || info->si_code == SI_TKILL) {
2976 2977
		/* We used to allow any < 0 si_code */
		WARN_ON_ONCE(info->si_code < 0);
2978
		return -EPERM;
2979
	}
2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995
	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);
}

2996
int do_sigaction(int sig, struct k_sigaction *act, struct k_sigaction *oact)
L
Linus Torvalds 已提交
2997
{
2998
	struct task_struct *t = current;
L
Linus Torvalds 已提交
2999
	struct k_sigaction *k;
3000
	sigset_t mask;
L
Linus Torvalds 已提交
3001

3002
	if (!valid_signal(sig) || sig < 1 || (act && sig_kernel_only(sig)))
L
Linus Torvalds 已提交
3003 3004
		return -EINVAL;

3005
	k = &t->sighand->action[sig-1];
L
Linus Torvalds 已提交
3006 3007 3008 3009 3010 3011

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

	if (act) {
3012 3013
		sigdelsetmask(&act->sa.sa_mask,
			      sigmask(SIGKILL) | sigmask(SIGSTOP));
3014
		*k = *act;
L
Linus Torvalds 已提交
3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025
		/*
		 * 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"
		 */
3026
		if (sig_handler_ignored(sig_handler(t, sig), sig)) {
3027 3028 3029
			sigemptyset(&mask);
			sigaddset(&mask, sig);
			rm_from_queue_full(&mask, &t->signal->shared_pending);
L
Linus Torvalds 已提交
3030
			do {
3031
				rm_from_queue_full(&mask, &t->pending);
L
Linus Torvalds 已提交
3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046
				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;

3047 3048 3049
	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
3050 3051 3052 3053 3054 3055 3056

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

		error = -EFAULT;
L
Linus Torvalds 已提交
3057 3058 3059 3060 3061 3062
		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
3063 3064 3065 3066 3067 3068 3069 3070
			goto out;

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

		error = -EINVAL;
		/*
3071
		 * Note - this code used to test ss_flags incorrectly:
L
Linus Torvalds 已提交
3072 3073 3074
		 *  	  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
3075
		 *	  mechanism.
L
Linus Torvalds 已提交
3076 3077 3078 3079 3080 3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091 3092
		 */
		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;
	}

3093
	error = 0;
L
Linus Torvalds 已提交
3094 3095
	if (uoss) {
		error = -EFAULT;
3096
		if (!access_ok(VERIFY_WRITE, uoss, sizeof(*uoss)))
L
Linus Torvalds 已提交
3097
			goto out;
3098 3099 3100
		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
3101 3102 3103 3104 3105
	}

out:
	return error;
}
3106 3107 3108 3109 3110 3111
#ifdef CONFIG_GENERIC_SIGALTSTACK
SYSCALL_DEFINE2(sigaltstack,const stack_t __user *,uss, stack_t __user *,uoss)
{
	return do_sigaltstack(uss, uoss, current_user_stack_pointer());
}
#endif
L
Linus Torvalds 已提交
3112

A
Al Viro 已提交
3113 3114 3115 3116 3117 3118 3119
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;
}

A
Al Viro 已提交
3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163
#ifdef CONFIG_COMPAT
#ifdef CONFIG_GENERIC_SIGALTSTACK
asmlinkage long compat_sys_sigaltstack(const compat_stack_t __user *uss_ptr,
				       compat_stack_t __user *uoss_ptr)
{
	stack_t uss, uoss;
	int ret;
	mm_segment_t seg;

	if (uss_ptr) {
		compat_stack_t uss32;

		memset(&uss, 0, sizeof(stack_t));
		if (copy_from_user(&uss32, uss_ptr, sizeof(compat_stack_t)))
			return -EFAULT;
		uss.ss_sp = compat_ptr(uss32.ss_sp);
		uss.ss_flags = uss32.ss_flags;
		uss.ss_size = uss32.ss_size;
	}
	seg = get_fs();
	set_fs(KERNEL_DS);
	ret = do_sigaltstack((stack_t __force __user *) (uss_ptr ? &uss : NULL),
			     (stack_t __force __user *) &uoss,
			     compat_user_stack_pointer());
	set_fs(seg);
	if (ret >= 0 && uoss_ptr)  {
		if (!access_ok(VERIFY_WRITE, uoss_ptr, sizeof(compat_stack_t)) ||
		    __put_user(ptr_to_compat(uoss.ss_sp), &uoss_ptr->ss_sp) ||
		    __put_user(uoss.ss_flags, &uoss_ptr->ss_flags) ||
		    __put_user(uoss.ss_size, &uoss_ptr->ss_size))
			ret = -EFAULT;
	}
	return ret;
}

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

L
Linus Torvalds 已提交
3164 3165
#ifdef __ARCH_WANT_SYS_SIGPENDING

3166 3167 3168 3169
/**
 *  sys_sigpending - examine pending signals
 *  @set: where mask of pending signal is returned
 */
3170
SYSCALL_DEFINE1(sigpending, old_sigset_t __user *, set)
L
Linus Torvalds 已提交
3171 3172 3173 3174 3175 3176 3177
{
	return do_sigpending(set, sizeof(*set));
}

#endif

#ifdef __ARCH_WANT_SYS_SIGPROCMASK
3178 3179 3180
/**
 *  sys_sigprocmask - examine and change blocked signals
 *  @how: whether to add, remove, or set signals
3181
 *  @nset: signals to add or remove (if non-null)
3182 3183
 *  @oset: previous value of signal mask if non-null
 *
3184 3185 3186
 * Some platforms have their own version with special arguments;
 * others support only sys_rt_sigprocmask.
 */
L
Linus Torvalds 已提交
3187

3188
SYSCALL_DEFINE3(sigprocmask, int, how, old_sigset_t __user *, nset,
3189
		old_sigset_t __user *, oset)
L
Linus Torvalds 已提交
3190 3191
{
	old_sigset_t old_set, new_set;
3192
	sigset_t new_blocked;
L
Linus Torvalds 已提交
3193

3194
	old_set = current->blocked.sig[0];
L
Linus Torvalds 已提交
3195

3196 3197 3198
	if (nset) {
		if (copy_from_user(&new_set, nset, sizeof(*nset)))
			return -EFAULT;
L
Linus Torvalds 已提交
3199 3200
		new_set &= ~(sigmask(SIGKILL) | sigmask(SIGSTOP));

3201
		new_blocked = current->blocked;
L
Linus Torvalds 已提交
3202 3203 3204

		switch (how) {
		case SIG_BLOCK:
3205
			sigaddsetmask(&new_blocked, new_set);
L
Linus Torvalds 已提交
3206 3207
			break;
		case SIG_UNBLOCK:
3208
			sigdelsetmask(&new_blocked, new_set);
L
Linus Torvalds 已提交
3209 3210
			break;
		case SIG_SETMASK:
3211
			new_blocked.sig[0] = new_set;
L
Linus Torvalds 已提交
3212
			break;
3213 3214
		default:
			return -EINVAL;
L
Linus Torvalds 已提交
3215 3216
		}

3217
		__set_current_blocked(&new_blocked);
3218 3219 3220
	}

	if (oset) {
L
Linus Torvalds 已提交
3221
		if (copy_to_user(oset, &old_set, sizeof(*oset)))
3222
			return -EFAULT;
L
Linus Torvalds 已提交
3223
	}
3224 3225

	return 0;
L
Linus Torvalds 已提交
3226 3227 3228 3229
}
#endif /* __ARCH_WANT_SYS_SIGPROCMASK */

#ifdef __ARCH_WANT_SYS_RT_SIGACTION
3230 3231 3232
/**
 *  sys_rt_sigaction - alter an action taken by a process
 *  @sig: signal to be sent
3233 3234
 *  @act: new sigaction
 *  @oact: used to save the previous sigaction
3235 3236
 *  @sigsetsize: size of sigset_t type
 */
3237 3238 3239 3240
SYSCALL_DEFINE4(rt_sigaction, int, sig,
		const struct sigaction __user *, act,
		struct sigaction __user *, oact,
		size_t, sigsetsize)
L
Linus Torvalds 已提交
3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269
{
	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.
 */
3270
SYSCALL_DEFINE0(sgetmask)
L
Linus Torvalds 已提交
3271 3272 3273 3274 3275
{
	/* SMP safe */
	return current->blocked.sig[0];
}

3276
SYSCALL_DEFINE1(ssetmask, int, newmask)
L
Linus Torvalds 已提交
3277
{
3278 3279
	int old = current->blocked.sig[0];
	sigset_t newset;
L
Linus Torvalds 已提交
3280

3281
	set_current_blocked(&newset);
L
Linus Torvalds 已提交
3282 3283 3284 3285 3286 3287 3288 3289 3290

	return old;
}
#endif /* __ARCH_WANT_SGETMASK */

#ifdef __ARCH_WANT_SYS_SIGNAL
/*
 * For backwards compatibility.  Functionality superseded by sigaction.
 */
3291
SYSCALL_DEFINE2(signal, int, sig, __sighandler_t, handler)
L
Linus Torvalds 已提交
3292 3293 3294 3295 3296 3297
{
	struct k_sigaction new_sa, old_sa;
	int ret;

	new_sa.sa.sa_handler = handler;
	new_sa.sa.sa_flags = SA_ONESHOT | SA_NOMASK;
3298
	sigemptyset(&new_sa.sa.sa_mask);
L
Linus Torvalds 已提交
3299 3300 3301 3302 3303 3304 3305 3306 3307

	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

3308
SYSCALL_DEFINE0(pause)
L
Linus Torvalds 已提交
3309
{
3310 3311 3312 3313
	while (!signal_pending(current)) {
		current->state = TASK_INTERRUPTIBLE;
		schedule();
	}
L
Linus Torvalds 已提交
3314 3315 3316 3317 3318
	return -ERESTARTNOHAND;
}

#endif

A
Al Viro 已提交
3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329
int sigsuspend(sigset_t *set)
{
	current->saved_sigmask = current->blocked;
	set_current_blocked(set);

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

3330
#ifdef __ARCH_WANT_SYS_RT_SIGSUSPEND
3331 3332 3333 3334 3335 3336
/**
 *  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
 */
3337
SYSCALL_DEFINE2(rt_sigsuspend, sigset_t __user *, unewset, size_t, sigsetsize)
3338 3339 3340 3341 3342 3343 3344 3345 3346
{
	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
3347
	return sigsuspend(&newset);
3348 3349 3350
}
#endif /* __ARCH_WANT_SYS_RT_SIGSUSPEND */

3351 3352 3353 3354 3355
__attribute__((weak)) const char *arch_vma_name(struct vm_area_struct *vma)
{
	return NULL;
}

L
Linus Torvalds 已提交
3356 3357
void __init signals_init(void)
{
3358
	sigqueue_cachep = KMEM_CACHE(sigqueue, SLAB_PANIC);
L
Linus Torvalds 已提交
3359
}
3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399

#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 */