exit.c 44.3 KB
Newer Older
Linus Torvalds's avatar
Linus Torvalds committed
1 2 3 4 5 6 7 8
/*
 *  linux/kernel/exit.c
 *
 *  Copyright (C) 1991, 1992  Linus Torvalds
 */

#include <linux/mm.h>
#include <linux/slab.h>
9
#include <linux/sched/autogroup.h>
10
#include <linux/sched/mm.h>
11
#include <linux/sched/stat.h>
12
#include <linux/sched/task.h>
13
#include <linux/sched/task_stack.h>
14
#include <linux/sched/cputime.h>
Linus Torvalds's avatar
Linus Torvalds committed
15 16
#include <linux/interrupt.h>
#include <linux/module.h>
17
#include <linux/capability.h>
Linus Torvalds's avatar
Linus Torvalds committed
18 19 20
#include <linux/completion.h>
#include <linux/personality.h>
#include <linux/tty.h>
21
#include <linux/iocontext.h>
Linus Torvalds's avatar
Linus Torvalds committed
22 23 24
#include <linux/key.h>
#include <linux/cpu.h>
#include <linux/acct.h>
25
#include <linux/tsacct_kern.h>
Linus Torvalds's avatar
Linus Torvalds committed
26
#include <linux/file.h>
Al Viro's avatar
Al Viro committed
27
#include <linux/fdtable.h>
28
#include <linux/freezer.h>
Linus Torvalds's avatar
Linus Torvalds committed
29
#include <linux/binfmts.h>
30
#include <linux/nsproxy.h>
31
#include <linux/pid_namespace.h>
Linus Torvalds's avatar
Linus Torvalds committed
32 33 34 35
#include <linux/ptrace.h>
#include <linux/profile.h>
#include <linux/mount.h>
#include <linux/proc_fs.h>
36
#include <linux/kthread.h>
Linus Torvalds's avatar
Linus Torvalds committed
37
#include <linux/mempolicy.h>
38
#include <linux/taskstats_kern.h>
39
#include <linux/delayacct.h>
40
#include <linux/cgroup.h>
Linus Torvalds's avatar
Linus Torvalds committed
41
#include <linux/syscalls.h>
42
#include <linux/signal.h>
43
#include <linux/posix-timers.h>
44
#include <linux/cn_proc.h>
45
#include <linux/mutex.h>
46
#include <linux/futex.h>
47
#include <linux/pipe_fs_i.h>
48
#include <linux/audit.h> /* for audit_free() */
49
#include <linux/resource.h>
50
#include <linux/blkdev.h>
51
#include <linux/task_io_accounting_ops.h>
Roland McGrath's avatar
Roland McGrath committed
52
#include <linux/tracehook.h>
53
#include <linux/fs_struct.h>
54
#include <linux/init_task.h>
55
#include <linux/perf_event.h>
56
#include <trace/events/sched.h>
57
#include <linux/hw_breakpoint.h>
58
#include <linux/oom.h>
59
#include <linux/writeback.h>
Al Viro's avatar
Al Viro committed
60
#include <linux/shm.h>
61
#include <linux/kcov.h>
62
#include <linux/random.h>
63
#include <linux/rcuwait.h>
64
#include <linux/compat.h>
Linus Torvalds's avatar
Linus Torvalds committed
65

66
#include <linux/uaccess.h>
Linus Torvalds's avatar
Linus Torvalds committed
67 68 69 70
#include <asm/unistd.h>
#include <asm/pgtable.h>
#include <asm/mmu_context.h>

71
static void __unhash_process(struct task_struct *p, bool group_dead)
Linus Torvalds's avatar
Linus Torvalds committed
72 73
{
	nr_threads--;
74
	detach_pid(p, PIDTYPE_PID);
75
	if (group_dead) {
Linus Torvalds's avatar
Linus Torvalds committed
76 77
		detach_pid(p, PIDTYPE_PGID);
		detach_pid(p, PIDTYPE_SID);
78

79
		list_del_rcu(&p->tasks);
80
		list_del_init(&p->sibling);
81
		__this_cpu_dec(process_counts);
Linus Torvalds's avatar
Linus Torvalds committed
82
	}
83
	list_del_rcu(&p->thread_group);
84
	list_del_rcu(&p->thread_node);
Linus Torvalds's avatar
Linus Torvalds committed
85 86
}

87 88 89 90 91 92
/*
 * This function expects the tasklist_lock write-locked.
 */
static void __exit_signal(struct task_struct *tsk)
{
	struct signal_struct *sig = tsk->signal;
93
	bool group_dead = thread_group_leader(tsk);
94
	struct sighand_struct *sighand;
95
	struct tty_struct *uninitialized_var(tty);
96
	u64 utime, stime;
97

98
	sighand = rcu_dereference_check(tsk->sighand,
99
					lockdep_tasklist_lock_is_held());
100 101
	spin_lock(&sighand->siglock);

102
#ifdef CONFIG_POSIX_TIMERS
103
	posix_cpu_timers_exit(tsk);
104
	if (group_dead) {
105
		posix_cpu_timers_exit_group(tsk);
106
	} else {
107 108 109 110 111 112 113
		/*
		 * This can only happen if the caller is de_thread().
		 * FIXME: this is the temporary hack, we should teach
		 * posix-cpu-timers to handle this case correctly.
		 */
		if (unlikely(has_group_leader_pid(tsk)))
			posix_cpu_timers_exit_group(tsk);
114 115
	}
#endif
116

117 118 119 120
	if (group_dead) {
		tty = sig->tty;
		sig->tty = NULL;
	} else {
121 122 123 124
		/*
		 * If there is any task waiting for the group exit
		 * then notify it:
		 */
125
		if (sig->notify_count > 0 && !--sig->notify_count)
126
			wake_up_process(sig->group_exit_task);
127

128 129 130 131
		if (tsk == sig->curr_target)
			sig->curr_target = next_thread(tsk);
	}

132 133 134
	add_device_randomness((const void*) &tsk->se.sum_exec_runtime,
			      sizeof(unsigned long long));

135
	/*
136 137 138 139
	 * Accumulate here the counters for all threads as they die. We could
	 * skip the group leader because it is the last user of signal_struct,
	 * but we want to avoid the race with thread_group_cputime() which can
	 * see the empty ->thread_head list.
140 141
	 */
	task_cputime(tsk, &utime, &stime);
142
	write_seqlock(&sig->stats_lock);
143 144 145 146 147 148 149 150 151 152 153
	sig->utime += utime;
	sig->stime += stime;
	sig->gtime += task_gtime(tsk);
	sig->min_flt += tsk->min_flt;
	sig->maj_flt += tsk->maj_flt;
	sig->nvcsw += tsk->nvcsw;
	sig->nivcsw += tsk->nivcsw;
	sig->inblock += task_io_get_inblock(tsk);
	sig->oublock += task_io_get_oublock(tsk);
	task_io_accounting_add(&sig->ioac, &tsk->ioac);
	sig->sum_sched_runtime += tsk->se.sum_exec_runtime;
154
	sig->nr_threads--;
155
	__unhash_process(tsk, group_dead);
156
	write_sequnlock(&sig->stats_lock);
157

158 159 160 161 162
	/*
	 * Do this under ->siglock, we can race with another thread
	 * doing sigqueue_free() if we have SIGQUEUE_PREALLOC signals.
	 */
	flush_sigqueue(&tsk->pending);
163
	tsk->sighand = NULL;
164 165
	spin_unlock(&sighand->siglock);

166
	__cleanup_sighand(sighand);
167
	clear_tsk_thread_flag(tsk, TIF_SIGPENDING);
168
	if (group_dead) {
169
		flush_sigqueue(&sig->shared_pending);
170
		tty_kref_put(tty);
171 172 173
	}
}

174 175
static void delayed_put_task_struct(struct rcu_head *rhp)
{
176 177
	struct task_struct *tsk = container_of(rhp, struct task_struct, rcu);

178
	perf_event_delayed_put(tsk);
179 180
	trace_sched_process_free(tsk);
	put_task_struct(tsk);
181 182
}

Roland McGrath's avatar
Roland McGrath committed
183

184
void release_task(struct task_struct *p)
Linus Torvalds's avatar
Linus Torvalds committed
185
{
186
	struct task_struct *leader;
Linus Torvalds's avatar
Linus Torvalds committed
187
	int zap_leader;
188
repeat:
189
	/* don't need to get the RCU readlock here - the process is dead and
190 191
	 * can't be modifying its own credentials. But shut RCU-lockdep up */
	rcu_read_lock();
192
	atomic_dec(&__task_cred(p)->user->processes);
193
	rcu_read_unlock();
194

195
	proc_flush_task(p);
196

Linus Torvalds's avatar
Linus Torvalds committed
197
	write_lock_irq(&tasklist_lock);
198
	ptrace_release_task(p);
Linus Torvalds's avatar
Linus Torvalds committed
199
	__exit_signal(p);
200

Linus Torvalds's avatar
Linus Torvalds committed
201 202 203 204 205 206 207
	/*
	 * If we are the last non-leader member of the thread
	 * group, and the leader is zombie, then notify the
	 * group leader's parent process. (if it wants notification.)
	 */
	zap_leader = 0;
	leader = p->group_leader;
208 209
	if (leader != p && thread_group_empty(leader)
			&& leader->exit_state == EXIT_ZOMBIE) {
Linus Torvalds's avatar
Linus Torvalds committed
210 211 212 213
		/*
		 * If we were the last child thread and the leader has
		 * exited already, and the leader's parent ignores SIGCHLD,
		 * then we are the one who should release the leader.
Roland McGrath's avatar
Roland McGrath committed
214
		 */
215
		zap_leader = do_notify_parent(leader, leader->exit_signal);
Roland McGrath's avatar
Roland McGrath committed
216 217
		if (zap_leader)
			leader->exit_state = EXIT_DEAD;
Linus Torvalds's avatar
Linus Torvalds committed
218 219 220 221
	}

	write_unlock_irq(&tasklist_lock);
	release_thread(p);
222
	call_rcu(&p->rcu, delayed_put_task_struct);
Linus Torvalds's avatar
Linus Torvalds committed
223 224 225 226 227 228

	p = leader;
	if (unlikely(zap_leader))
		goto repeat;
}

229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291
/*
 * Note that if this function returns a valid task_struct pointer (!NULL)
 * task->usage must remain >0 for the duration of the RCU critical section.
 */
struct task_struct *task_rcu_dereference(struct task_struct **ptask)
{
	struct sighand_struct *sighand;
	struct task_struct *task;

	/*
	 * We need to verify that release_task() was not called and thus
	 * delayed_put_task_struct() can't run and drop the last reference
	 * before rcu_read_unlock(). We check task->sighand != NULL,
	 * but we can read the already freed and reused memory.
	 */
retry:
	task = rcu_dereference(*ptask);
	if (!task)
		return NULL;

	probe_kernel_address(&task->sighand, sighand);

	/*
	 * Pairs with atomic_dec_and_test() in put_task_struct(). If this task
	 * was already freed we can not miss the preceding update of this
	 * pointer.
	 */
	smp_rmb();
	if (unlikely(task != READ_ONCE(*ptask)))
		goto retry;

	/*
	 * We've re-checked that "task == *ptask", now we have two different
	 * cases:
	 *
	 * 1. This is actually the same task/task_struct. In this case
	 *    sighand != NULL tells us it is still alive.
	 *
	 * 2. This is another task which got the same memory for task_struct.
	 *    We can't know this of course, and we can not trust
	 *    sighand != NULL.
	 *
	 *    In this case we actually return a random value, but this is
	 *    correct.
	 *
	 *    If we return NULL - we can pretend that we actually noticed that
	 *    *ptask was updated when the previous task has exited. Or pretend
	 *    that probe_slab_address(&sighand) reads NULL.
	 *
	 *    If we return the new task (because sighand is not NULL for any
	 *    reason) - this is fine too. This (new) task can't go away before
	 *    another gp pass.
	 *
	 *    And note: We could even eliminate the false positive if re-read
	 *    task->sighand once again to avoid the falsely NULL. But this case
	 *    is very unlikely so we don't care.
	 */
	if (!sighand)
		return NULL;

	return task;
}

292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320
void rcuwait_wake_up(struct rcuwait *w)
{
	struct task_struct *task;

	rcu_read_lock();

	/*
	 * Order condition vs @task, such that everything prior to the load
	 * of @task is visible. This is the condition as to why the user called
	 * rcuwait_trywake() in the first place. Pairs with set_current_state()
	 * barrier (A) in rcuwait_wait_event().
	 *
	 *    WAIT                WAKE
	 *    [S] tsk = current	  [S] cond = true
	 *        MB (A)	      MB (B)
	 *    [L] cond		  [L] tsk
	 */
	smp_rmb(); /* (B) */

	/*
	 * Avoid using task_rcu_dereference() magic as long as we are careful,
	 * see comment in rcuwait_wait_event() regarding ->exit_state.
	 */
	task = rcu_dereference(w->task);
	if (task)
		wake_up_process(task);
	rcu_read_unlock();
}

Linus Torvalds's avatar
Linus Torvalds committed
321 322 323 324 325 326 327 328
/*
 * Determine if a process group is "orphaned", according to the POSIX
 * definition in 2.2.2.52.  Orphaned process groups are not to be affected
 * by terminal-generated stop signals.  Newly orphaned process groups are
 * to receive a SIGHUP and a SIGCONT.
 *
 * "I ask you, have you ever known what it is to be an orphan?"
 */
329 330
static int will_become_orphaned_pgrp(struct pid *pgrp,
					struct task_struct *ignored_task)
Linus Torvalds's avatar
Linus Torvalds committed
331 332 333
{
	struct task_struct *p;

334
	do_each_pid_task(pgrp, PIDTYPE_PGID, p) {
335 336 337
		if ((p == ignored_task) ||
		    (p->exit_state && thread_group_empty(p)) ||
		    is_global_init(p->real_parent))
Linus Torvalds's avatar
Linus Torvalds committed
338
			continue;
339

340
		if (task_pgrp(p->real_parent) != pgrp &&
341 342
		    task_session(p->real_parent) == task_session(p))
			return 0;
343
	} while_each_pid_task(pgrp, PIDTYPE_PGID, p);
344 345

	return 1;
Linus Torvalds's avatar
Linus Torvalds committed
346 347
}

348
int is_current_pgrp_orphaned(void)
Linus Torvalds's avatar
Linus Torvalds committed
349 350 351 352
{
	int retval;

	read_lock(&tasklist_lock);
353
	retval = will_become_orphaned_pgrp(task_pgrp(current), NULL);
Linus Torvalds's avatar
Linus Torvalds committed
354 355 356 357 358
	read_unlock(&tasklist_lock);

	return retval;
}

359
static bool has_stopped_jobs(struct pid *pgrp)
Linus Torvalds's avatar
Linus Torvalds committed
360 361 362
{
	struct task_struct *p;

363
	do_each_pid_task(pgrp, PIDTYPE_PGID, p) {
364 365
		if (p->signal->flags & SIGNAL_STOP_STOPPED)
			return true;
366
	} while_each_pid_task(pgrp, PIDTYPE_PGID, p);
367 368

	return false;
Linus Torvalds's avatar
Linus Torvalds committed
369 370
}

371 372 373 374 375 376 377 378 379 380 381 382
/*
 * Check to see if any process groups have become orphaned as
 * a result of our exiting, and if they have any stopped jobs,
 * send them a SIGHUP and then a SIGCONT. (POSIX 3.2.2.2)
 */
static void
kill_orphaned_pgrp(struct task_struct *tsk, struct task_struct *parent)
{
	struct pid *pgrp = task_pgrp(tsk);
	struct task_struct *ignored_task = tsk;

	if (!parent)
383 384 385
		/* exit: our father is in a different pgrp than
		 * we are and we were the only connection outside.
		 */
386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401
		parent = tsk->real_parent;
	else
		/* reparent: our child is in a different pgrp than
		 * we are, and it was the only connection outside.
		 */
		ignored_task = NULL;

	if (task_pgrp(parent) != pgrp &&
	    task_session(parent) == task_session(tsk) &&
	    will_become_orphaned_pgrp(pgrp, ignored_task) &&
	    has_stopped_jobs(pgrp)) {
		__kill_pgrp_info(SIGHUP, SEND_SIG_PRIV, pgrp);
		__kill_pgrp_info(SIGCONT, SEND_SIG_PRIV, pgrp);
	}
}

402
#ifdef CONFIG_MEMCG
403
/*
404
 * A task is exiting.   If it owned this mm, find a new owner for the mm.
405 406 407 408 409 410
 */
void mm_update_next_owner(struct mm_struct *mm)
{
	struct task_struct *c, *g, *p = current;

retry:
411 412 413 414 415
	/*
	 * If the exiting or execing task is not the owner, it's
	 * someone else's problem.
	 */
	if (mm->owner != p)
416
		return;
417 418 419 420 421 422 423 424 425
	/*
	 * The current owner is exiting/execing and there are no other
	 * candidates.  Do not leave the mm pointing to a possibly
	 * freed task structure.
	 */
	if (atomic_read(&mm->mm_users) <= 1) {
		mm->owner = NULL;
		return;
	}
426 427 428 429 430 431 432 433 434 435 436 437 438

	read_lock(&tasklist_lock);
	/*
	 * Search in the children
	 */
	list_for_each_entry(c, &p->children, sibling) {
		if (c->mm == mm)
			goto assign_new_owner;
	}

	/*
	 * Search in the siblings
	 */
439
	list_for_each_entry(c, &p->real_parent->children, sibling) {
440 441 442 443 444
		if (c->mm == mm)
			goto assign_new_owner;
	}

	/*
445
	 * Search through everything else, we should not get here often.
446
	 */
447 448 449 450 451 452 453 454 455
	for_each_process(g) {
		if (g->flags & PF_KTHREAD)
			continue;
		for_each_thread(g, c) {
			if (c->mm == mm)
				goto assign_new_owner;
			if (c->mm)
				break;
		}
456
	}
457
	read_unlock(&tasklist_lock);
458 459 460
	/*
	 * We found no owner yet mm_users > 1: this implies that we are
	 * most likely racing with swapoff (try_to_unuse()) or /proc or
461
	 * ptrace or page migration (get_task_mm()).  Mark owner as NULL.
462 463
	 */
	mm->owner = NULL;
464 465 466 467 468 469 470 471 472 473
	return;

assign_new_owner:
	BUG_ON(c == p);
	get_task_struct(c);
	/*
	 * The task_lock protects c->mm from changing.
	 * We always want mm->owner->mm == mm
	 */
	task_lock(c);
474 475 476 477 478
	/*
	 * Delay read_unlock() till we have the task_lock()
	 * to ensure that c does not slip away underneath us
	 */
	read_unlock(&tasklist_lock);
479 480 481 482 483 484 485 486 487
	if (c->mm != mm) {
		task_unlock(c);
		put_task_struct(c);
		goto retry;
	}
	mm->owner = c;
	task_unlock(c);
	put_task_struct(c);
}
488
#endif /* CONFIG_MEMCG */
489

Linus Torvalds's avatar
Linus Torvalds committed
490 491 492 493
/*
 * Turn us into a lazy TLB process if we
 * aren't already..
 */
494
static void exit_mm(void)
Linus Torvalds's avatar
Linus Torvalds committed
495
{
496
	struct mm_struct *mm = current->mm;
497
	struct core_state *core_state;
Linus Torvalds's avatar
Linus Torvalds committed
498

499
	mm_release(current, mm);
Linus Torvalds's avatar
Linus Torvalds committed
500 501
	if (!mm)
		return;
502
	sync_mm_rss(mm);
Linus Torvalds's avatar
Linus Torvalds committed
503 504
	/*
	 * Serialize with any possible pending coredump.
505
	 * We must hold mmap_sem around checking core_state
Linus Torvalds's avatar
Linus Torvalds committed
506
	 * and clearing tsk->mm.  The core-inducing thread
507
	 * will increment ->nr_threads for each thread in the
Linus Torvalds's avatar
Linus Torvalds committed
508 509 510
	 * group with ->mm != NULL.
	 */
	down_read(&mm->mmap_sem);
511 512 513
	core_state = mm->core_state;
	if (core_state) {
		struct core_thread self;
514

Linus Torvalds's avatar
Linus Torvalds committed
515 516
		up_read(&mm->mmap_sem);

517
		self.task = current;
518 519 520 521 522 523 524
		self.next = xchg(&core_state->dumper.next, &self);
		/*
		 * Implies mb(), the result of xchg() must be visible
		 * to core_state->dumper.
		 */
		if (atomic_dec_and_test(&core_state->nr_threads))
			complete(&core_state->startup);
Linus Torvalds's avatar
Linus Torvalds committed
525

526
		for (;;) {
527
			set_current_state(TASK_UNINTERRUPTIBLE);
528 529
			if (!self.task) /* see coredump_finish() */
				break;
530
			freezable_schedule();
531
		}
532
		__set_current_state(TASK_RUNNING);
Linus Torvalds's avatar
Linus Torvalds committed
533 534
		down_read(&mm->mmap_sem);
	}
535
	mmgrab(mm);
536
	BUG_ON(mm != current->active_mm);
Linus Torvalds's avatar
Linus Torvalds committed
537
	/* more a memory barrier than a real lock */
538 539
	task_lock(current);
	current->mm = NULL;
Linus Torvalds's avatar
Linus Torvalds committed
540 541
	up_read(&mm->mmap_sem);
	enter_lazy_tlb(mm, current);
542
	task_unlock(current);
543
	mm_update_next_owner(mm);
Linus Torvalds's avatar
Linus Torvalds committed
544
	mmput(mm);
545
	if (test_thread_flag(TIF_MEMDIE))
546
		exit_oom_victim();
Linus Torvalds's avatar
Linus Torvalds committed
547 548
}

549 550 551 552 553 554 555 556 557 558 559
static struct task_struct *find_alive_thread(struct task_struct *p)
{
	struct task_struct *t;

	for_each_thread(p, t) {
		if (!(t->flags & PF_EXITING))
			return t;
	}
	return NULL;
}

560 561 562 563 564 565 566 567 568 569
static struct task_struct *find_child_reaper(struct task_struct *father)
	__releases(&tasklist_lock)
	__acquires(&tasklist_lock)
{
	struct pid_namespace *pid_ns = task_active_pid_ns(father);
	struct task_struct *reaper = pid_ns->child_reaper;

	if (likely(reaper != father))
		return reaper;

570 571
	reaper = find_alive_thread(father);
	if (reaper) {
572 573 574 575 576 577 578 579 580 581 582 583 584 585 586
		pid_ns->child_reaper = reaper;
		return reaper;
	}

	write_unlock_irq(&tasklist_lock);
	if (unlikely(pid_ns == &init_pid_ns)) {
		panic("Attempted to kill init! exitcode=0x%08x\n",
			father->signal->group_exit_code ?: father->exit_code);
	}
	zap_pid_ns_processes(pid_ns);
	write_lock_irq(&tasklist_lock);

	return father;
}

Linus Torvalds's avatar
Linus Torvalds committed
587
/*
588 589 590 591 592
 * When we die, we re-parent all our children, and try to:
 * 1. give them to another thread in our thread group, if such a member exists
 * 2. give it to the first ancestor process which prctl'd itself as a
 *    child_subreaper for its children (like a service manager)
 * 3. give it to the init process (PID 1) in our pid namespace
Linus Torvalds's avatar
Linus Torvalds committed
593
 */
594 595
static struct task_struct *find_new_reaper(struct task_struct *father,
					   struct task_struct *child_reaper)
Linus Torvalds's avatar
Linus Torvalds committed
596
{
597
	struct task_struct *thread, *reaper;
Linus Torvalds's avatar
Linus Torvalds committed
598

599 600
	thread = find_alive_thread(father);
	if (thread)
601
		return thread;
Linus Torvalds's avatar
Linus Torvalds committed
602

603
	if (father->signal->has_child_subreaper) {
604
		unsigned int ns_level = task_pid(father)->level;
605
		/*
606
		 * Find the first ->is_child_subreaper ancestor in our pid_ns.
607 608 609 610 611
		 * We can't check reaper != child_reaper to ensure we do not
		 * cross the namespaces, the exiting parent could be injected
		 * by setns() + fork().
		 * We check pid->level, this is slightly more efficient than
		 * task_active_pid_ns(reaper) != task_active_pid_ns(father).
612
		 */
613 614
		for (reaper = father->real_parent;
		     task_pid(reaper)->level == ns_level;
615
		     reaper = reaper->real_parent) {
616
			if (reaper == &init_task)
617 618 619
				break;
			if (!reaper->signal->is_child_subreaper)
				continue;
620 621 622
			thread = find_alive_thread(reaper);
			if (thread)
				return thread;
623
		}
Linus Torvalds's avatar
Linus Torvalds committed
624
	}
625

626
	return child_reaper;
627 628
}

629 630 631
/*
* Any that need to be release_task'd are put on the @dead list.
 */
632
static void reparent_leader(struct task_struct *father, struct task_struct *p,
633 634
				struct list_head *dead)
{
635
	if (unlikely(p->exit_state == EXIT_DEAD))
636 637
		return;

638
	/* We don't want people slaying init. */
639 640 641
	p->exit_signal = SIGCHLD;

	/* If it has exited notify the new parent about this child's death. */
Tejun Heo's avatar
Tejun Heo committed
642
	if (!p->ptrace &&
643
	    p->exit_state == EXIT_ZOMBIE && thread_group_empty(p)) {
644
		if (do_notify_parent(p, p->exit_signal)) {
645
			p->exit_state = EXIT_DEAD;
646
			list_add(&p->ptrace_entry, dead);
647 648 649 650 651 652
		}
	}

	kill_orphaned_pgrp(p, father);
}

653 654 655 656 657 658 659 660 661 662
/*
 * This does two things:
 *
 * A.  Make init inherit all the child processes
 * B.  Check to see if any process groups have become orphaned
 *	as a result of our exiting, and if they have any stopped
 *	jobs, send them a SIGHUP and then a SIGCONT.  (POSIX 3.2.2.2)
 */
static void forget_original_parent(struct task_struct *father,
					struct list_head *dead)
Linus Torvalds's avatar
Linus Torvalds committed
663
{
664
	struct task_struct *p, *t, *reaper;
665

666
	if (unlikely(!list_empty(&father->ptraced)))
667
		exit_ptrace(father, dead);
Roland McGrath's avatar
Roland McGrath committed
668

669
	/* Can drop and reacquire tasklist_lock */
670
	reaper = find_child_reaper(father);
671
	if (list_empty(&father->children))
672
		return;
673 674

	reaper = find_new_reaper(father, reaper);
675
	list_for_each_entry(p, &father->children, sibling) {
676
		for_each_thread(p, t) {
677
			t->real_parent = reaper;
678 679
			BUG_ON((!t->ptrace) != (t->parent == father));
			if (likely(!t->ptrace))
680 681 682 683
				t->parent = t->real_parent;
			if (t->pdeath_signal)
				group_send_sig_info(t->pdeath_signal,
						    SEND_SIG_NOINFO, t);
684
		}
685 686 687 688 689
		/*
		 * If this is a threaded reparent there is no need to
		 * notify anyone anything has happened.
		 */
		if (!same_thread_group(reaper, father))
690
			reparent_leader(father, p, dead);
Linus Torvalds's avatar
Linus Torvalds committed
691
	}
692
	list_splice_tail_init(&father->children, &reaper->children);
Linus Torvalds's avatar
Linus Torvalds committed
693 694 695 696 697 698
}

/*
 * Send signals to all our closest relatives so that they know
 * to properly mourn us..
 */
699
static void exit_notify(struct task_struct *tsk, int group_dead)
Linus Torvalds's avatar
Linus Torvalds committed
700
{
701
	bool autoreap;
702 703
	struct task_struct *p, *n;
	LIST_HEAD(dead);
Linus Torvalds's avatar
Linus Torvalds committed
704

705
	write_lock_irq(&tasklist_lock);
706 707
	forget_original_parent(tsk, &dead);

708 709
	if (group_dead)
		kill_orphaned_pgrp(tsk->group_leader, NULL);
Linus Torvalds's avatar
Linus Torvalds committed
710

711 712 713 714 715 716 717 718 719 720 721 722
	if (unlikely(tsk->ptrace)) {
		int sig = thread_group_leader(tsk) &&
				thread_group_empty(tsk) &&
				!ptrace_reparented(tsk) ?
			tsk->exit_signal : SIGCHLD;
		autoreap = do_notify_parent(tsk, sig);
	} else if (thread_group_leader(tsk)) {
		autoreap = thread_group_empty(tsk) &&
			do_notify_parent(tsk, tsk->exit_signal);
	} else {
		autoreap = true;
	}
Linus Torvalds's avatar
Linus Torvalds committed
723

724
	tsk->exit_state = autoreap ? EXIT_DEAD : EXIT_ZOMBIE;
725 726
	if (tsk->exit_state == EXIT_DEAD)
		list_add(&tsk->ptrace_entry, &dead);
Linus Torvalds's avatar
Linus Torvalds committed
727

728 729
	/* mt-exec, de_thread() is waiting for group leader */
	if (unlikely(tsk->signal->notify_count < 0))
730
		wake_up_process(tsk->signal->group_exit_task);
Linus Torvalds's avatar
Linus Torvalds committed
731 732
	write_unlock_irq(&tasklist_lock);

733 734 735 736
	list_for_each_entry_safe(p, n, &dead, ptrace_entry) {
		list_del_init(&p->ptrace_entry);
		release_task(p);
	}
Linus Torvalds's avatar
Linus Torvalds committed
737 738
}

739 740 741 742 743 744 745
#ifdef CONFIG_DEBUG_STACK_USAGE
static void check_stack_usage(void)
{
	static DEFINE_SPINLOCK(low_water_lock);
	static int lowest_to_date = THREAD_SIZE;
	unsigned long free;

746
	free = stack_not_used(current);
747 748 749 750 751 752

	if (free >= lowest_to_date)
		return;

	spin_lock(&low_water_lock);
	if (free < lowest_to_date) {
753
		pr_info("%s (%d) used greatest stack depth: %lu bytes left\n",
754
			current->comm, task_pid_nr(current), free);
755 756 757 758 759 760 761 762
		lowest_to_date = free;
	}
	spin_unlock(&low_water_lock);
}
#else
static inline void check_stack_usage(void) {}
#endif

763
void __noreturn do_exit(long code)
Linus Torvalds's avatar
Linus Torvalds committed
764 765 766 767 768
{
	struct task_struct *tsk = current;
	int group_dead;

	profile_task_exit(tsk);
769
	kcov_task_exit(tsk);
Linus Torvalds's avatar
Linus Torvalds committed
770

771
	WARN_ON(blk_needs_flush_plug(tsk));
772

Linus Torvalds's avatar
Linus Torvalds committed
773 774 775 776 777
	if (unlikely(in_interrupt()))
		panic("Aiee, killing interrupt handler!");
	if (unlikely(!tsk->pid))
		panic("Attempted to kill the idle task!");

778 779 780 781 782 783 784 785 786
	/*
	 * If do_exit is called because this processes oopsed, it's possible
	 * that get_fs() was left as KERNEL_DS, so reset it to USER_DS before
	 * continuing. Amongst other possible reasons, this is to prevent
	 * mm_release()->clear_child_tid() from writing to a user-controlled
	 * kernel address.
	 */
	set_fs(USER_DS);

787
	ptrace_event(PTRACE_EVENT_EXIT, code);
Linus Torvalds's avatar
Linus Torvalds committed
788

789 790
	validate_creds_for_do_exit(tsk);

791 792 793 794 795
	/*
	 * We're taking recursive faults here in do_exit. Safest is to just
	 * leave this task alone and wait for reboot.
	 */
	if (unlikely(tsk->flags & PF_EXITING)) {
796
		pr_alert("Fixing recursive fault but reboot is needed!\n");
797 798 799 800 801 802 803 804 805 806
		/*
		 * We can do this unlocked here. The futex code uses
		 * this flag just to verify whether the pi state
		 * cleanup has been done or not. In the worst case it
		 * loops once more. We pretend that the cleanup was
		 * done as there is no way to return. Either the
		 * OWNER_DIED bit is set by now or we push the blocked
		 * task into the wait for ever nirwana as well.
		 */
		tsk->flags |= PF_EXITPIDONE;
807 808 809 810
		set_current_state(TASK_UNINTERRUPTIBLE);
		schedule();
	}

811
	exit_signals(tsk);  /* sets PF_EXITING */
812
	/*
813 814
	 * Ensure that all new tsk->pi_lock acquisitions must observe
	 * PF_EXITING. Serializes against futex.c:attach_to_pi_owner().
815
	 */
816
	smp_mb();
817 818 819 820
	/*
	 * Ensure that we must observe the pi_state in exit_mm() ->
	 * mm_release() -> exit_pi_state_list().
	 */
821 822
	raw_spin_lock_irq(&tsk->pi_lock);
	raw_spin_unlock_irq(&tsk->pi_lock);
Linus Torvalds's avatar
Linus Torvalds committed
823

824
	if (unlikely(in_atomic())) {
825 826 827
		pr_info("note: %s[%d] exited with preempt_count %d\n",
			current->comm, task_pid_nr(current),
			preempt_count());
828 829
		preempt_count_set(PREEMPT_ENABLED);
	}
Linus Torvalds's avatar
Linus Torvalds committed
830

831 832 833
	/* sync mm's RSS info before statistics gathering */
	if (tsk->mm)
		sync_mm_rss(tsk->mm);
834
	acct_update_integrals(tsk);
Linus Torvalds's avatar
Linus Torvalds committed
835
	group_dead = atomic_dec_and_test(&tsk->signal->live);
836
	if (group_dead) {
837
#ifdef CONFIG_POSIX_TIMERS
838
		hrtimer_cancel(&tsk->signal->real_timer);
839
		exit_itimers(tsk->signal);
840
#endif
841 842
		if (tsk->mm)
			setmax_mm_hiwater_rss(&tsk->signal->maxrss, tsk->mm);
843
	}
844
	acct_collect(code, group_dead);
845 846
	if (group_dead)
		tty_audit_exit();
847
	audit_free(tsk);
848

849
	tsk->exit_code = code;
850
	taskstats_exit(tsk, group_dead);
851

852
	exit_mm();
Linus Torvalds's avatar
Linus Torvalds committed
853

854
	if (group_dead)
855
		acct_process();
856 857
	trace_sched_process_exit(tsk);

Linus Torvalds's avatar
Linus Torvalds committed
858
	exit_sem(tsk);
859
	exit_shm(tsk);
860 861
	exit_files(tsk);
	exit_fs(tsk);
862 863
	if (group_dead)
		disassociate_ctty(1);
864
	exit_task_namespaces(tsk);
865
	exit_task_work(tsk);
866
	exit_thread(tsk);
867 868 869 870 871 872 873 874 875

	/*
	 * Flush inherited counters to the parent - before the parent
	 * gets woken up by child-exit notifications.
	 *
	 * because of cgroup mode, must be called before cgroup_exit()
	 */
	perf_event_exit_task(tsk);

876
	sched_autogroup_exit_task(tsk);
877
	cgroup_exit(tsk);
Linus Torvalds's avatar
Linus Torvalds committed
878

879 880 881
	/*
	 * FIXME: do that only when needed, using sched_exit tracepoint
	 */
882
	flush_ptrace_hw_breakpoint(tsk);
883

884
	exit_tasks_rcu_start();
885
	exit_notify(tsk, group_dead);
886
	proc_exit_connector(tsk);
887
	mpol_put_task_policy(tsk);
888
#ifdef CONFIG_FUTEX
889 890
	if (unlikely(current->pi_state_cache))
		kfree(current->pi_state_cache);
891
#endif
892
	/*
893
	 * Make sure we are holding no locks:
894
	 */
895
	debug_check_no_locks_held();
896 897 898 899 900 901
	/*
	 * We can do this unlocked here. The futex code uses this flag
	 * just to verify whether the pi state cleanup has been done
	 * or not. In the worst case it loops once more.
	 */
	tsk->flags |= PF_EXITPIDONE;
Linus Torvalds's avatar
Linus Torvalds committed
902

903
	if (tsk->io_context)
904
		exit_io_context(tsk);
905

906
	if (tsk->splice_pipe)
907
		free_pipe_info(tsk->splice_pipe);
908

909 910 911
	if (tsk->task_frag.page)
		put_page(tsk->task_frag.page);

912 913
	validate_creds_for_do_exit(tsk);

914
	check_stack_usage();
915
	preempt_disable();
916 917
	if (tsk->nr_dirtied)
		__this_cpu_add(dirty_throttle_leaks, tsk->nr_dirtied);
918
	exit_rcu();
919
	exit_tasks_rcu_finish();
920

921
	lockdep_free_task(tsk);
922
	do_task_dead();
Linus Torvalds's avatar
Linus Torvalds committed
923
}
924 925
EXPORT_SYMBOL_GPL(do_exit);

926
void complete_and_exit(struct completion *comp, long code)
Linus Torvalds's avatar
Linus Torvalds committed
927 928 929
{
	if (comp)
		complete(comp);
930

Linus Torvalds's avatar
Linus Torvalds committed
931 932 933 934
	do_exit(code);
}
EXPORT_SYMBOL(complete_and_exit);

935
SYSCALL_DEFINE1(exit, int, error_code)
Linus Torvalds's avatar
Linus Torvalds committed
936 937 938 939 940 941 942 943
{
	do_exit((error_code&0xff)<<8);
}

/*
 * Take down every thread in the group.  This is called by fatal signals
 * as well as by sys_exit_group (below).
 */
944
void
Linus Torvalds's avatar
Linus Torvalds committed
945 946
do_group_exit(int exit_code)
{
947 948
	struct signal_struct *sig = current->signal;

Linus Torvalds's avatar
Linus Torvalds committed
949 950
	BUG_ON(exit_code & 0x80); /* core dumps don't get here */

951 952
	if (signal_group_exit(sig))
		exit_code = sig->group_exit_code;
Linus Torvalds's avatar
Linus Torvalds committed
953 954
	else if (!thread_group_empty(current)) {
		struct sighand_struct *const sighand = current->sighand;
955

Linus Torvalds's avatar
Linus Torvalds committed
956
		spin_lock_irq(&sighand->siglock);
957
		if (signal_group_exit(sig))
Linus Torvalds's avatar
Linus Torvalds committed
958 959 960 961
			/* Another thread got here before we took the lock.  */
			exit_code = sig->group_exit_code;
		else {
			sig->group_exit_code = exit_code;
962
			sig->flags = SIGNAL_GROUP_EXIT;
Linus Torvalds's avatar
Linus Torvalds committed
963 964 965 966 967 968 969 970 971 972 973 974 975 976
			zap_other_threads(current);
		}
		spin_unlock_irq(&sighand->siglock);
	}

	do_exit(exit_code);
	/* NOTREACHED */
}

/*
 * this kills every thread in the thread group. Note that any externally
 * wait4()-ing process will get the correct exit code - even if this
 * thread is not the thread group leader.
 */
977
SYSCALL_DEFINE1(exit_group, int, error_code)
Linus Torvalds's avatar
Linus Torvalds committed
978 979
{
	do_group_exit((error_code & 0xff) << 8);
980 981
	/* NOTREACHED */
	return 0;
Linus Torvalds's avatar
Linus Torvalds committed
982 983
}

984 985 986 987 988 989 990
struct waitid_info {
	pid_t pid;
	uid_t uid;
	int status;
	int cause;
};

991 992 993
struct wait_opts {
	enum pid_type		wo_type;
	int			wo_flags;
994
	struct pid		*wo_pid;
995

996
	struct waitid_info	*wo_info;
997
	int			wo_stat;
998
	struct rusage		*wo_rusage;
999

1000
	wait_queue_entry_t		child_wait;
1001 1002 1003
	int			notask_error;
};

1004 1005
static inline
struct pid *task_pid_type(struct task_struct *task, enum pid_type type)
1006
{
1007 1008 1009
	if (type != PIDTYPE_PID)
		task = task->group_leader;
	return task->pids[type].pid;
1010 1011
}

1012
static int eligible_pid(struct wait_opts *wo, struct task_struct *p)
Linus Torvalds's avatar
Linus Torvalds committed
1013
{
1014 1015 1016
	return	wo->wo_type == PIDTYPE_MAX ||
		task_pid_type(p, wo->wo_type) == wo->wo_pid;
}
Linus Torvalds's avatar
Linus Torvalds committed
1017

1018 1019
static int
eligible_child(struct wait_opts *wo, bool ptrace, struct task_struct *p)
1020 1021 1022
{
	if (!eligible_pid(wo, p))
		return 0;
1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039

	/*
	 * Wait for all children (clone and not) if __WALL is set or
	 * if it is traced by us.
	 */
	if (ptrace || (wo->wo_flags & __WALL))
		return 1;

	/*
	 * Otherwise, wait for clone children *only* if __WCLONE is set;
	 * otherwise, wait for non-clone children *only*.
	 *
	 * Note: a "clone" child here is one that reports to its parent
	 * using a signal other than SIGCHLD, or a non-leader thread which
	 * we can only see if it is traced by us.
	 */
	if ((p->exit_signal != SIGCHLD) ^ !!(wo->wo_flags & __WCLONE))
Linus Torvalds's avatar
Linus Torvalds committed
1040 1041
		return 0;

1042
	return 1;
Linus Torvalds's avatar
Linus Torvalds committed
1043 1044 1045 1046 1047 1048 1049 1050
}

/*
 * Handle sys_wait4 work for one task in state EXIT_ZOMBIE.  We hold
 * read_lock(&tasklist_lock) on entry.  If we return zero, we still hold
 * the lock and this task is uninteresting.  If we return nonzero, we have
 * released the lock and the system call should return.
 */
1051
static int wait_task_zombie(struct wait_opts *wo, struct task_struct *p)
Linus Torvalds's avatar
Linus Torvalds committed
1052
{
1053
	int state, status;
1054
	pid_t pid = task_pid_vnr(p);
1055
	uid_t uid = from_kuid_munged(current_user_ns(), task_uid(p));
1056
	struct waitid_info *infop;
Linus Torvalds's avatar
Linus Torvalds committed
1057

1058
	if (!likely(wo->wo_flags & WEXITED))
Roland McGrath's avatar
Roland McGrath committed
1059 1060
		return 0;

1061
	if (unlikely(wo->wo_flags & WNOWAIT)) {
1062
		status = p->exit_code;
Linus Torvalds's avatar
Linus Torvalds committed
1063 1064
		get_task_struct(p);
		read_unlock(&tasklist_lock);
1065
		sched_annotate_sleep();
1066 1067
		if (wo->wo_rusage)
			getrusage(p, RUSAGE_BOTH, wo->wo_rusage);
Al Viro's avatar
Al Viro committed
1068
		put_task_struct(p);
1069
		goto out_info;
Linus Torvalds's avatar
Linus Torvalds committed
1070 1071
	}
	/*
1072
	 * Move the task's state to DEAD/TRACE, only one thread can do this.
Linus Torvalds's avatar
Linus Torvalds committed
1073
	 */
1074 1075
	state = (ptrace_reparented(p) && thread_group_leader(p)) ?
		EXIT_TRACE : EXIT_DEAD;
1076
	if (cmpxchg(&p->exit_state, EXIT_ZOMBIE, state) != EXIT_ZOMBIE)
Linus Torvalds's avatar
Linus Torvalds committed
1077
		return 0;
1078 1079 1080 1081 1082
	/*
	 * We own this thread, nobody else can reap it.
	 */
	read_unlock(&tasklist_lock);
	sched_annotate_sleep();
1083

1084
	/*
1085
	 * Check thread_group_leader() to exclude the traced sub-threads.
1086
	 */
1087
	if (state == EXIT_DEAD && thread_group_leader(p)) {
1088 1089
		struct signal_struct *sig = p->signal;
		struct signal_struct *psig = current->signal;
1090
		unsigned long maxrss;
1091
		u64 tgutime, tgstime;
1092

Linus Torvalds's avatar
Linus Torvalds committed
1093 1094 1095 1096 1097 1098 1099 1100
		/*
		 * The resource counters for the group leader are in its
		 * own task_struct.  Those for dead threads in the group
		 * are in its signal_struct, as are those for the child
		 * processes it has previously reaped.  All these
		 * accumulate in the parent's signal_struct c* fields.
		 *
		 * We don't bother to take a lock here to protect these
1101 1102 1103 1104 1105 1106 1107
		 * p->signal fields because the whole thread group is dead
		 * and nobody can change them.
		 *
		 * psig->stats_lock also protects us from our sub-theads
		 * which can reap other children at the same time. Until
		 * we change k_getrusage()-like users to rely on this lock
		 * we have to take ->siglock as well.
1108
		 *
1109 1110 1111
		 * We use thread_group_cputime_adjusted() to get times for
		 * the thread group, which consolidates times for all threads
		 * in the group including the group leader.
Linus Torvalds's avatar
Linus Torvalds committed
1112
		 */
1113
		thread_group_cputime_adjusted(p, &tgutime, &tgstime);
1114
		spin_lock_irq(&current->sighand->siglock);
1115
		write_seqlock(&psig->stats_lock);
1116 1117
		psig->cutime += tgutime + sig->cutime;
		psig->cstime += tgstime + sig->cstime;
1118
		psig->cgtime += task_gtime(p) + sig->gtime + sig->cgtime;
1119 1120 1121 1122 1123 1124 1125 1126
		psig->cmin_flt +=
			p->min_flt + sig->min_flt + sig->cmin_flt;
		psig->cmaj_flt +=
			p->maj_flt + sig->maj_flt + sig->cmaj_flt;
		psig->cnvcsw +=
			p->nvcsw + sig->nvcsw + sig->cnvcsw;
		psig->cnivcsw +=
			p->nivcsw + sig->nivcsw + sig->cnivcsw;
1127 1128 1129 1130 1131 1132
		psig->cinblock +=
			task_io_get_inblock(p) +
			sig->inblock + sig->cinblock;
		psig->coublock +=
			task_io_get_oublock(p) +
			sig->oublock + sig->coublock;
1133 1134 1135
		maxrss = max(sig->maxrss, sig->cmaxrss);
		if (psig->cmaxrss < maxrss)
			psig->cmaxrss = maxrss;
1136 1137
		task_io_accounting_add(&psig->ioac, &p->ioac);
		task_io_accounting_add(&psig->ioac, &sig->ioac);
1138
		write_sequnlock(&psig->stats_lock);
1139
		spin_unlock_irq(&current->sighand->siglock);
Linus Torvalds's avatar
Linus Torvalds committed
1140 1141
	}

1142 1143
	if (wo->wo_rusage)
		getrusage(p, RUSAGE_BOTH, wo->wo_rusage);
Linus Torvalds's avatar
Linus Torvalds committed
1144 1145
	status = (p->signal->flags & SIGNAL_GROUP_EXIT)
		? p->signal->group_exit_code : p->exit_code;
1146
	wo->wo_stat = status;
1147

1148
	if (state == EXIT_TRACE) {
Linus Torvalds's avatar
Linus Torvalds committed
1149
		write_lock_irq(&tasklist_lock);
1150 1151
		/* We dropped tasklist, ptracer could die and untrace */
		ptrace_unlink(p);
1152 1153 1154 1155 1156

		/* If parent wants a zombie, don't release it now */
		state = EXIT_ZOMBIE;
		if (do_notify_parent(p, p->exit_signal))
			state = EXIT_DEAD;
1157
		p->exit_state = state;
Linus Torvalds's avatar
Linus Torvalds committed
1158 1159
		write_unlock_irq(&tasklist_lock);
	}
1160
	if (state == EXIT_DEAD)
Linus Torvalds's avatar
Linus Torvalds committed
1161
		release_task(p);
1162

1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176
out_info:
	infop = wo->wo_info;
	if (infop) {
		if ((status & 0x7f) == 0) {
			infop->cause = CLD_EXITED;
			infop->status = status >> 8;
		} else {
			infop->cause = (status & 0x80) ? CLD_DUMPED : CLD_KILLED;
			infop->status = status & 0x7f;
		}
		infop->pid = pid;
		infop->uid = uid;
	}

1177
	return pid;
Linus Torvalds's avatar
Linus Torvalds committed
1178 1179
}

1180 1181 1182
static int *task_stopped_code(struct task_struct *p, bool ptrace)
{
	if (ptrace) {
1183
		if (task_is_traced(p) && !(p->jobctl & JOBCTL_LISTENING))
1184 1185 1186 1187 1188 1189 1190 1191
			return &p->exit_code;
	} else {
		if (p->signal->flags & SIGNAL_STOP_STOPPED)
			return &p->signal->group_exit_code;
	}
	return NULL;
}

1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208
/**
 * wait_task_stopped - Wait for %TASK_STOPPED or %TASK_TRACED
 * @wo: wait options
 * @ptrace: is the wait for ptrace
 * @p: task to wait for
 *
 * Handle sys_wait4() work for %p in state %TASK_STOPPED or %TASK_TRACED.
 *
 * CONTEXT:
 * read_lock(&tasklist_lock), which is released if return value is
 * non-zero.  Also, grabs and releases @p->sighand->siglock.
 *
 * RETURNS:
 * 0 if wait condition didn't exist and search for other wait conditions
 * should continue.  Non-zero return, -errno on failure and @p's pid on
 * success, implies that tasklist_lock is released and wait condition
 * search should terminate.
Linus Torvalds's avatar
Linus Torvalds committed
1209
 */
1210 1211
static int wait_task_stopped(struct wait_opts *wo,
				int ptrace, struct task_struct *p)
Linus Torvalds's avatar
Linus Torvalds committed
1212
{
1213 1214
	struct waitid_info *infop;
	int exit_code, *p_code, why;
1215
	uid_t uid = 0; /* unneeded, required by compiler */
1216
	pid_t pid;
Linus Torvalds's avatar
Linus Torvalds committed
1217

1218 1219 1220
	/*
	 * Traditionally we see ptrace'd stopped tasks regardless of options.
	 */
1221
	if (!ptrace && !(wo->wo_flags & WUNTRACED))
Roland McGrath's avatar
Roland McGrath committed
1222 1223
		return 0;

1224 1225 1226
	if (!task_stopped_code(p, ptrace))
		return 0;

1227 1228 1229
	exit_code = 0;
	spin_lock_irq(&p->sighand->siglock);

1230 1231
	p_code = task_stopped_code(p, ptrace);
	if (unlikely(!p_code))
1232 1233
		goto unlock_sig;

1234
	exit_code = *p_code;
1235 1236 1237
	if (!exit_code)
		goto unlock_sig;

1238
	if (!unlikely(wo->wo_flags & WNOWAIT))
1239
		*p_code = 0;
1240

1241
	uid = from_kuid_munged(current_user_ns(), task_uid(p));
1242 1243 1244
unlock_sig:
	spin_unlock_irq(&p->sighand->siglock);
	if (!exit_code)
Linus Torvalds's avatar
Linus Torvalds committed
1245 1246 1247 1248 1249 1250 1251 1252 1253 1254
		return 0;

	/*
	 * Now we are pretty sure this task is interesting.
	 * Make sure it doesn't get reaped out from under us while we
	 * give up the lock and then examine it below.  We don't want to
	 * keep holding onto the tasklist_lock while we call getrusage and
	 * possibly take page faults for user memory.
	 */
	get_task_struct(p);
1255
	pid = task_pid_vnr(p);
Roland McGrath's avatar
Roland McGrath committed
1256
	why = ptrace ? CLD_TRAPPED : CLD_STOPPED;
Linus Torvalds's avatar
Linus Torvalds committed
1257
	read_unlock(&tasklist_lock);