exit.c 45.2 KB
Newer Older
Linus Torvalds's avatar
Linus Torvalds committed
1 2 3 4 5 6 7 8 9 10
/*
 *  linux/kernel/exit.c
 *
 *  Copyright (C) 1991, 1992  Linus Torvalds
 */

#include <linux/mm.h>
#include <linux/slab.h>
#include <linux/interrupt.h>
#include <linux/module.h>
11
#include <linux/capability.h>
Linus Torvalds's avatar
Linus Torvalds committed
12 13 14
#include <linux/completion.h>
#include <linux/personality.h>
#include <linux/tty.h>
15
#include <linux/mnt_namespace.h>
16
#include <linux/iocontext.h>
Linus Torvalds's avatar
Linus Torvalds committed
17 18 19 20
#include <linux/key.h>
#include <linux/security.h>
#include <linux/cpu.h>
#include <linux/acct.h>
21
#include <linux/tsacct_kern.h>
Linus Torvalds's avatar
Linus Torvalds committed
22
#include <linux/file.h>
Al Viro's avatar
Al Viro committed
23
#include <linux/fdtable.h>
Linus Torvalds's avatar
Linus Torvalds committed
24
#include <linux/binfmts.h>
25
#include <linux/nsproxy.h>
26
#include <linux/pid_namespace.h>
Linus Torvalds's avatar
Linus Torvalds committed
27 28 29 30
#include <linux/ptrace.h>
#include <linux/profile.h>
#include <linux/mount.h>
#include <linux/proc_fs.h>
31
#include <linux/kthread.h>
Linus Torvalds's avatar
Linus Torvalds committed
32
#include <linux/mempolicy.h>
33
#include <linux/taskstats_kern.h>
34
#include <linux/delayacct.h>
35
#include <linux/freezer.h>
36
#include <linux/cgroup.h>
Linus Torvalds's avatar
Linus Torvalds committed
37
#include <linux/syscalls.h>
38
#include <linux/signal.h>
39
#include <linux/posix-timers.h>
40
#include <linux/cn_proc.h>
41
#include <linux/mutex.h>
42
#include <linux/futex.h>
43
#include <linux/pipe_fs_i.h>
44
#include <linux/audit.h> /* for audit_free() */
Adrian Bunk's avatar
Adrian Bunk committed
45
#include <linux/resource.h>
46
#include <linux/blkdev.h>
47
#include <linux/task_io_accounting_ops.h>
Roland McGrath's avatar
Roland McGrath committed
48
#include <linux/tracehook.h>
49
#include <linux/init_task.h>
50
#include <trace/sched.h>
Linus Torvalds's avatar
Linus Torvalds committed
51 52 53 54 55

#include <asm/uaccess.h>
#include <asm/unistd.h>
#include <asm/pgtable.h>
#include <asm/mmu_context.h>
56
#include "cred-internals.h"
Linus Torvalds's avatar
Linus Torvalds committed
57

58 59 60
DEFINE_TRACE(sched_process_free);
DEFINE_TRACE(sched_process_exit);
DEFINE_TRACE(sched_process_wait);
Linus Torvalds's avatar
Linus Torvalds committed
61

62 63
static void exit_mm(struct task_struct * tsk);

64 65 66 67 68
static inline int task_detached(struct task_struct *p)
{
	return p->exit_signal == -1;
}

Linus Torvalds's avatar
Linus Torvalds committed
69 70 71 72 73 74 75
static void __unhash_process(struct task_struct *p)
{
	nr_threads--;
	detach_pid(p, PIDTYPE_PID);
	if (thread_group_leader(p)) {
		detach_pid(p, PIDTYPE_PGID);
		detach_pid(p, PIDTYPE_SID);
76

77
		list_del_rcu(&p->tasks);
78
		__get_cpu_var(process_counts)--;
Linus Torvalds's avatar
Linus Torvalds committed
79
	}
80
	list_del_rcu(&p->thread_group);
Roland McGrath's avatar
Roland McGrath committed
81
	list_del_init(&p->sibling);
Linus Torvalds's avatar
Linus Torvalds committed
82 83
}

84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105
/*
 * This function expects the tasklist_lock write-locked.
 */
static void __exit_signal(struct task_struct *tsk)
{
	struct signal_struct *sig = tsk->signal;
	struct sighand_struct *sighand;

	BUG_ON(!sig);
	BUG_ON(!atomic_read(&sig->count));

	sighand = rcu_dereference(tsk->sighand);
	spin_lock(&sighand->siglock);

	posix_cpu_timers_exit(tsk);
	if (atomic_dec_and_test(&sig->count))
		posix_cpu_timers_exit_group(tsk);
	else {
		/*
		 * If there is any task waiting for the group exit
		 * then notify it:
		 */
106
		if (sig->group_exit_task && atomic_read(&sig->count) == sig->notify_count)
107
			wake_up_process(sig->group_exit_task);
108

109 110 111 112 113 114 115 116 117 118 119 120
		if (tsk == sig->curr_target)
			sig->curr_target = next_thread(tsk);
		/*
		 * Accumulate here the counters for all threads but the
		 * group leader as they die, so they can be added into
		 * the process-wide totals when those are taken.
		 * The group leader stays around as a zombie as long
		 * as there are other threads.  When it gets reaped,
		 * the exit.c code will add its counts into these totals.
		 * We won't ever get here for the group leader, since it
		 * will have been the last reference on the signal_struct.
		 */
121 122
		sig->utime = cputime_add(sig->utime, task_utime(tsk));
		sig->stime = cputime_add(sig->stime, task_stime(tsk));
123
		sig->gtime = cputime_add(sig->gtime, task_gtime(tsk));
124 125 126 127
		sig->min_flt += tsk->min_flt;
		sig->maj_flt += tsk->maj_flt;
		sig->nvcsw += tsk->nvcsw;
		sig->nivcsw += tsk->nivcsw;
128 129
		sig->inblock += task_io_get_inblock(tsk);
		sig->oublock += task_io_get_oublock(tsk);
130
		task_io_accounting_add(&sig->ioac, &tsk->ioac);
131
		sig->sum_sched_runtime += tsk->se.sum_exec_runtime;
132 133 134
		sig = NULL; /* Marker for below. */
	}

135 136
	__unhash_process(tsk);

137 138 139 140 141 142
	/*
	 * Do this under ->siglock, we can race with another thread
	 * doing sigqueue_free() if we have SIGQUEUE_PREALLOC signals.
	 */
	flush_sigqueue(&tsk->pending);

143
	tsk->signal = NULL;
144
	tsk->sighand = NULL;
145 146
	spin_unlock(&sighand->siglock);

147
	__cleanup_sighand(sighand);
148 149 150
	clear_tsk_thread_flag(tsk,TIF_SIGPENDING);
	if (sig) {
		flush_sigqueue(&sig->shared_pending);
151
		taskstats_tgid_free(sig);
152 153 154 155 156
		/*
		 * Make sure ->signal can't go away under rq->lock,
		 * see account_group_exec_runtime().
		 */
		task_rq_unlock_wait(tsk);
157 158 159 160
		__cleanup_signal(sig);
	}
}

161 162
static void delayed_put_task_struct(struct rcu_head *rhp)
{
163 164 165 166
	struct task_struct *tsk = container_of(rhp, struct task_struct, rcu);

	trace_sched_process_free(tsk);
	put_task_struct(tsk);
167 168
}

Roland McGrath's avatar
Roland McGrath committed
169

Linus Torvalds's avatar
Linus Torvalds committed
170 171
void release_task(struct task_struct * p)
{
172
	struct task_struct *leader;
Linus Torvalds's avatar
Linus Torvalds committed
173
	int zap_leader;
174
repeat:
Roland McGrath's avatar
Roland McGrath committed
175
	tracehook_prepare_release_task(p);
176 177 178 179
	/* don't need to get the RCU readlock here - the process is dead and
	 * can't be modifying its own credentials */
	atomic_dec(&__task_cred(p)->user->processes);

180
	proc_flush_task(p);
Linus Torvalds's avatar
Linus Torvalds committed
181
	write_lock_irq(&tasklist_lock);
Roland McGrath's avatar
Roland McGrath committed
182
	tracehook_finish_release_task(p);
Linus Torvalds's avatar
Linus Torvalds committed
183
	__exit_signal(p);
184

Linus Torvalds's avatar
Linus Torvalds committed
185 186 187 188 189 190 191 192
	/*
	 * 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;
	if (leader != p && thread_group_empty(leader) && leader->exit_state == EXIT_ZOMBIE) {
193
		BUG_ON(task_detached(leader));
Linus Torvalds's avatar
Linus Torvalds committed
194 195 196 197 198 199 200 201 202
		do_notify_parent(leader, leader->exit_signal);
		/*
		 * 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.
		 *
		 * do_notify_parent() will have marked it self-reaping in
		 * that case.
		 */
203
		zap_leader = task_detached(leader);
Roland McGrath's avatar
Roland McGrath committed
204 205 206 207 208 209 210

		/*
		 * This maintains the invariant that release_task()
		 * only runs on a task in EXIT_DEAD, just for sanity.
		 */
		if (zap_leader)
			leader->exit_state = EXIT_DEAD;
Linus Torvalds's avatar
Linus Torvalds committed
211 212 213 214
	}

	write_unlock_irq(&tasklist_lock);
	release_thread(p);
215
	call_rcu(&p->rcu, delayed_put_task_struct);
Linus Torvalds's avatar
Linus Torvalds committed
216 217 218 219 220 221 222 223 224 225

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

/*
 * This checks not only the pgrp, but falls back on the pid if no
 * satisfactory pgrp is found. I dunno - gdb doesn't work correctly
 * without this...
226 227
 *
 * The caller must hold rcu lock or the tasklist lock.
Linus Torvalds's avatar
Linus Torvalds committed
228
 */
229
struct pid *session_of_pgrp(struct pid *pgrp)
Linus Torvalds's avatar
Linus Torvalds committed
230 231
{
	struct task_struct *p;
232
	struct pid *sid = NULL;
233

234
	p = pid_task(pgrp, PIDTYPE_PGID);
235
	if (p == NULL)
236
		p = pid_task(pgrp, PIDTYPE_PID);
237
	if (p != NULL)
238
		sid = task_session(p);
239

Linus Torvalds's avatar
Linus Torvalds committed
240 241 242 243 244 245 246 247 248 249 250
	return sid;
}

/*
 * 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?"
 */
251
static int will_become_orphaned_pgrp(struct pid *pgrp, struct task_struct *ignored_task)
Linus Torvalds's avatar
Linus Torvalds committed
252 253 254
{
	struct task_struct *p;

255
	do_each_pid_task(pgrp, PIDTYPE_PGID, p) {
256 257 258
		if ((p == ignored_task) ||
		    (p->exit_state && thread_group_empty(p)) ||
		    is_global_init(p->real_parent))
Linus Torvalds's avatar
Linus Torvalds committed
259
			continue;
260

261
		if (task_pgrp(p->real_parent) != pgrp &&
262 263
		    task_session(p->real_parent) == task_session(p))
			return 0;
264
	} while_each_pid_task(pgrp, PIDTYPE_PGID, p);
265 266

	return 1;
Linus Torvalds's avatar
Linus Torvalds committed
267 268
}

269
int is_current_pgrp_orphaned(void)
Linus Torvalds's avatar
Linus Torvalds committed
270 271 272 273
{
	int retval;

	read_lock(&tasklist_lock);
274
	retval = will_become_orphaned_pgrp(task_pgrp(current), NULL);
Linus Torvalds's avatar
Linus Torvalds committed
275 276 277 278 279
	read_unlock(&tasklist_lock);

	return retval;
}

280
static int has_stopped_jobs(struct pid *pgrp)
Linus Torvalds's avatar
Linus Torvalds committed
281 282 283 284
{
	int retval = 0;
	struct task_struct *p;

285
	do_each_pid_task(pgrp, PIDTYPE_PGID, p) {
Matthew Wilcox's avatar
Matthew Wilcox committed
286
		if (!task_is_stopped(p))
Linus Torvalds's avatar
Linus Torvalds committed
287 288 289
			continue;
		retval = 1;
		break;
290
	} while_each_pid_task(pgrp, PIDTYPE_PGID, p);
Linus Torvalds's avatar
Linus Torvalds committed
291 292 293
	return retval;
}

294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324
/*
 * 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)
		 /* exit: our father is in a different pgrp than
		  * we are and we were the only connection outside.
		  */
		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);
	}
}

Linus Torvalds's avatar
Linus Torvalds committed
325
/**
326
 * reparent_to_kthreadd - Reparent the calling kernel thread to kthreadd
Linus Torvalds's avatar
Linus Torvalds committed
327 328
 *
 * If a kernel thread is launched as a result of a system call, or if
329 330
 * it ever exits, it should generally reparent itself to kthreadd so it
 * isn't in the way of other processes and is correctly cleaned up on exit.
Linus Torvalds's avatar
Linus Torvalds committed
331 332 333 334
 *
 * The various task state such as scheduling policy and priority may have
 * been inherited from a user process, so we reset them to sane values here.
 *
335
 * NOTE that reparent_to_kthreadd() gives the caller full capabilities.
Linus Torvalds's avatar
Linus Torvalds committed
336
 */
337
static void reparent_to_kthreadd(void)
Linus Torvalds's avatar
Linus Torvalds committed
338 339 340 341 342
{
	write_lock_irq(&tasklist_lock);

	ptrace_unlink(current);
	/* Reparent to init */
343
	current->real_parent = current->parent = kthreadd_task;
Roland McGrath's avatar
Roland McGrath committed
344
	list_move_tail(&current->sibling, &current->real_parent->children);
Linus Torvalds's avatar
Linus Torvalds committed
345 346 347 348

	/* Set the exit signal to SIGCHLD so we signal init on exit */
	current->exit_signal = SIGCHLD;

349
	if (task_nice(current) < 0)
Linus Torvalds's avatar
Linus Torvalds committed
350 351 352 353 354 355
		set_user_nice(current, 0);
	/* cpus_allowed? */
	/* rt_priority? */
	/* signals? */
	memcpy(current->signal->rlim, init_task.signal->rlim,
	       sizeof(current->signal->rlim));
356 357 358

	atomic_inc(&init_cred.usage);
	commit_creds(&init_cred);
Linus Torvalds's avatar
Linus Torvalds committed
359 360 361
	write_unlock_irq(&tasklist_lock);
}

362
void __set_special_pids(struct pid *pid)
Linus Torvalds's avatar
Linus Torvalds committed
363
{
364
	struct task_struct *curr = current->group_leader;
365
	pid_t nr = pid_nr(pid);
Linus Torvalds's avatar
Linus Torvalds committed
366

367
	if (task_session(curr) != pid) {
368
		change_pid(curr, PIDTYPE_SID, pid);
369
		set_task_session(curr, nr);
Linus Torvalds's avatar
Linus Torvalds committed
370
	}
371
	if (task_pgrp(curr) != pid) {
372
		change_pid(curr, PIDTYPE_PGID, pid);
373
		set_task_pgrp(curr, nr);
Linus Torvalds's avatar
Linus Torvalds committed
374 375 376
	}
}

377
static void set_special_pids(struct pid *pid)
Linus Torvalds's avatar
Linus Torvalds committed
378 379
{
	write_lock_irq(&tasklist_lock);
380
	__set_special_pids(pid);
Linus Torvalds's avatar
Linus Torvalds committed
381 382 383 384 385 386 387 388 389 390
	write_unlock_irq(&tasklist_lock);
}

/*
 * Let kernel threads use this to say that they
 * allow a certain signal (since daemonize() will
 * have disabled all of them by default).
 */
int allow_signal(int sig)
{
391
	if (!valid_signal(sig) || sig < 1)
Linus Torvalds's avatar
Linus Torvalds committed
392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411
		return -EINVAL;

	spin_lock_irq(&current->sighand->siglock);
	sigdelset(&current->blocked, sig);
	if (!current->mm) {
		/* Kernel threads handle their own signals.
		   Let the signal code know it'll be handled, so
		   that they don't get converted to SIGKILL or
		   just silently dropped */
		current->sighand->action[(sig)-1].sa.sa_handler = (void __user *)2;
	}
	recalc_sigpending();
	spin_unlock_irq(&current->sighand->siglock);
	return 0;
}

EXPORT_SYMBOL(allow_signal);

int disallow_signal(int sig)
{
412
	if (!valid_signal(sig) || sig < 1)
Linus Torvalds's avatar
Linus Torvalds committed
413 414 415
		return -EINVAL;

	spin_lock_irq(&current->sighand->siglock);
416
	current->sighand->action[(sig)-1].sa.sa_handler = SIG_IGN;
Linus Torvalds's avatar
Linus Torvalds committed
417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443
	recalc_sigpending();
	spin_unlock_irq(&current->sighand->siglock);
	return 0;
}

EXPORT_SYMBOL(disallow_signal);

/*
 *	Put all the gunge required to become a kernel thread without
 *	attached user resources in one place where it belongs.
 */

void daemonize(const char *name, ...)
{
	va_list args;
	sigset_t blocked;

	va_start(args, name);
	vsnprintf(current->comm, sizeof(current->comm), name, args);
	va_end(args);

	/*
	 * If we were started as result of loading a module, close all of the
	 * user space pages.  We don't need them, and if we didn't close them
	 * they would be locked into memory.
	 */
	exit_mm(current);
444 445 446 447
	/*
	 * We don't want to have TIF_FREEZE set if the system-wide hibernation
	 * or suspend transition begins right now.
	 */
Oleg Nesterov's avatar
Oleg Nesterov committed
448
	current->flags |= (PF_NOFREEZE | PF_KTHREAD);
Linus Torvalds's avatar
Linus Torvalds committed
449

450 451 452 453
	if (current->nsproxy != &init_nsproxy) {
		get_nsproxy(&init_nsproxy);
		switch_task_namespaces(current, &init_nsproxy);
	}
454
	set_special_pids(&init_struct_pid);
455
	proc_clear_tty(current);
Linus Torvalds's avatar
Linus Torvalds committed
456 457 458 459 460 461 462 463

	/* Block and flush all signals */
	sigfillset(&blocked);
	sigprocmask(SIG_BLOCK, &blocked, NULL);
	flush_signals(current);

	/* Become as one with the init task */

464
	daemonize_fs_struct();
465
	exit_files(current);
Linus Torvalds's avatar
Linus Torvalds committed
466 467 468
	current->files = init_task.files;
	atomic_inc(&current->files->count);

469
	reparent_to_kthreadd();
Linus Torvalds's avatar
Linus Torvalds committed
470 471 472 473
}

EXPORT_SYMBOL(daemonize);

474
static void close_files(struct files_struct * files)
Linus Torvalds's avatar
Linus Torvalds committed
475 476
{
	int i, j;
477
	struct fdtable *fdt;
Linus Torvalds's avatar
Linus Torvalds committed
478 479

	j = 0;
480 481 482 483 484 485

	/*
	 * It is safe to dereference the fd table without RCU or
	 * ->file_lock because this is the last reference to the
	 * files structure.
	 */
486
	fdt = files_fdtable(files);
Linus Torvalds's avatar
Linus Torvalds committed
487 488 489
	for (;;) {
		unsigned long set;
		i = j * __NFDBITS;
490
		if (i >= fdt->max_fds)
Linus Torvalds's avatar
Linus Torvalds committed
491
			break;
492
		set = fdt->open_fds->fds_bits[j++];
Linus Torvalds's avatar
Linus Torvalds committed
493 494
		while (set) {
			if (set & 1) {
495
				struct file * file = xchg(&fdt->fd[i], NULL);
496
				if (file) {
Linus Torvalds's avatar
Linus Torvalds committed
497
					filp_close(file, files);
498 499
					cond_resched();
				}
Linus Torvalds's avatar
Linus Torvalds committed
500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519
			}
			i++;
			set >>= 1;
		}
	}
}

struct files_struct *get_files_struct(struct task_struct *task)
{
	struct files_struct *files;

	task_lock(task);
	files = task->files;
	if (files)
		atomic_inc(&files->count);
	task_unlock(task);

	return files;
}

520
void put_files_struct(struct files_struct *files)
Linus Torvalds's avatar
Linus Torvalds committed
521
{
522 523
	struct fdtable *fdt;

Linus Torvalds's avatar
Linus Torvalds committed
524 525 526 527
	if (atomic_dec_and_test(&files->count)) {
		close_files(files);
		/*
		 * Free the fd and fdset arrays if we expanded them.
528 529 530
		 * If the fdtable was embedded, pass files for freeing
		 * at the end of the RCU grace period. Otherwise,
		 * you can free files immediately.
Linus Torvalds's avatar
Linus Torvalds committed
531
		 */
532
		fdt = files_fdtable(files);
533
		if (fdt != &files->fdtab)
534
			kmem_cache_free(files_cachep, files);
535
		free_fdtable(fdt);
Linus Torvalds's avatar
Linus Torvalds committed
536 537 538
	}
}

539
void reset_files_struct(struct files_struct *files)
540
{
541
	struct task_struct *tsk = current;
542 543 544 545 546 547 548 549 550
	struct files_struct *old;

	old = tsk->files;
	task_lock(tsk);
	tsk->files = files;
	task_unlock(tsk);
	put_files_struct(old);
}

551
void exit_files(struct task_struct *tsk)
Linus Torvalds's avatar
Linus Torvalds committed
552 553 554 555 556 557 558 559 560 561 562
{
	struct files_struct * files = tsk->files;

	if (files) {
		task_lock(tsk);
		tsk->files = NULL;
		task_unlock(tsk);
		put_files_struct(files);
	}
}

563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615
#ifdef CONFIG_MM_OWNER
/*
 * Task p is exiting and it owned mm, lets find a new owner for it
 */
static inline int
mm_need_new_owner(struct mm_struct *mm, struct task_struct *p)
{
	/*
	 * If there are other users of the mm and the owner (us) is exiting
	 * we need to find a new owner to take on the responsibility.
	 */
	if (atomic_read(&mm->mm_users) <= 1)
		return 0;
	if (mm->owner != p)
		return 0;
	return 1;
}

void mm_update_next_owner(struct mm_struct *mm)
{
	struct task_struct *c, *g, *p = current;

retry:
	if (!mm_need_new_owner(mm, p))
		return;

	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
	 */
	list_for_each_entry(c, &p->parent->children, sibling) {
		if (c->mm == mm)
			goto assign_new_owner;
	}

	/*
	 * Search through everything else. We should not get
	 * here often
	 */
	do_each_thread(g, c) {
		if (c->mm == mm)
			goto assign_new_owner;
	} while_each_thread(g, c);

	read_unlock(&tasklist_lock);
616 617 618
	/*
	 * We found no owner yet mm_users > 1: this implies that we are
	 * most likely racing with swapoff (try_to_unuse()) or /proc or
619
	 * ptrace or page migration (get_task_mm()).  Mark owner as NULL.
620 621
	 */
	mm->owner = NULL;
622 623 624 625 626 627 628 629 630 631
	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);
632 633 634 635 636
	/*
	 * Delay read_unlock() till we have the task_lock()
	 * to ensure that c does not slip away underneath us
	 */
	read_unlock(&tasklist_lock);
637 638 639 640 641 642 643 644 645 646 647
	if (c->mm != mm) {
		task_unlock(c);
		put_task_struct(c);
		goto retry;
	}
	mm->owner = c;
	task_unlock(c);
	put_task_struct(c);
}
#endif /* CONFIG_MM_OWNER */

Linus Torvalds's avatar
Linus Torvalds committed
648 649 650 651
/*
 * Turn us into a lazy TLB process if we
 * aren't already..
 */
652
static void exit_mm(struct task_struct * tsk)
Linus Torvalds's avatar
Linus Torvalds committed
653 654
{
	struct mm_struct *mm = tsk->mm;
655
	struct core_state *core_state;
Linus Torvalds's avatar
Linus Torvalds committed
656 657 658 659 660 661

	mm_release(tsk, mm);
	if (!mm)
		return;
	/*
	 * Serialize with any possible pending coredump.
662
	 * We must hold mmap_sem around checking core_state
Linus Torvalds's avatar
Linus Torvalds committed
663
	 * and clearing tsk->mm.  The core-inducing thread
664
	 * will increment ->nr_threads for each thread in the
Linus Torvalds's avatar
Linus Torvalds committed
665 666 667
	 * group with ->mm != NULL.
	 */
	down_read(&mm->mmap_sem);
668 669 670
	core_state = mm->core_state;
	if (core_state) {
		struct core_thread self;
Linus Torvalds's avatar
Linus Torvalds committed
671 672
		up_read(&mm->mmap_sem);

673 674 675 676 677 678 679 680
		self.task = tsk;
		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
681

Oleg Nesterov's avatar
Oleg Nesterov committed
682 683 684 685 686 687 688
		for (;;) {
			set_task_state(tsk, TASK_UNINTERRUPTIBLE);
			if (!self.task) /* see coredump_finish() */
				break;
			schedule();
		}
		__set_task_state(tsk, TASK_RUNNING);
Linus Torvalds's avatar
Linus Torvalds committed
689 690 691
		down_read(&mm->mmap_sem);
	}
	atomic_inc(&mm->mm_count);
692
	BUG_ON(mm != tsk->active_mm);
Linus Torvalds's avatar
Linus Torvalds committed
693 694 695 696 697
	/* more a memory barrier than a real lock */
	task_lock(tsk);
	tsk->mm = NULL;
	up_read(&mm->mmap_sem);
	enter_lazy_tlb(mm, current);
698 699
	/* We don't want this task to be frozen prematurely */
	clear_freeze_flag(tsk);
Linus Torvalds's avatar
Linus Torvalds committed
700
	task_unlock(tsk);
701
	mm_update_next_owner(mm);
Linus Torvalds's avatar
Linus Torvalds committed
702 703 704
	mmput(mm);
}

705 706 707 708 709 710
/*
 * Return nonzero if @parent's children should reap themselves.
 *
 * Called with write_lock_irq(&tasklist_lock) held.
 */
static int ignoring_children(struct task_struct *parent)
Linus Torvalds's avatar
Linus Torvalds committed
711
{
712 713 714 715 716 717 718 719 720
	int ret;
	struct sighand_struct *psig = parent->sighand;
	unsigned long flags;
	spin_lock_irqsave(&psig->siglock, flags);
	ret = (psig->action[SIGCHLD-1].sa.sa_handler == SIG_IGN ||
	       (psig->action[SIGCHLD-1].sa.sa_flags & SA_NOCLDWAIT));
	spin_unlock_irqrestore(&psig->siglock, flags);
	return ret;
}
721

Roland McGrath's avatar
Roland McGrath committed
722 723 724 725 726 727 728
/*
 * Detach all tasks we were using ptrace on.
 * Any that need to be release_task'd are put on the @dead list.
 *
 * Called with write_lock(&tasklist_lock) held.
 */
static void ptrace_exit(struct task_struct *parent, struct list_head *dead)
Linus Torvalds's avatar
Linus Torvalds committed
729
{
Roland McGrath's avatar
Roland McGrath committed
730
	struct task_struct *p, *n;
731
	int ign = -1;
732

Roland McGrath's avatar
Roland McGrath committed
733 734 735 736 737 738 739 740 741 742 743 744 745 746
	list_for_each_entry_safe(p, n, &parent->ptraced, ptrace_entry) {
		__ptrace_unlink(p);

		if (p->exit_state != EXIT_ZOMBIE)
			continue;

		/*
		 * If it's a zombie, our attachedness prevented normal
		 * parent notification or self-reaping.  Do notification
		 * now if it would have happened earlier.  If it should
		 * reap itself, add it to the @dead list.  We can't call
		 * release_task() here because we already hold tasklist_lock.
		 *
		 * If it's our own child, there is no notification to do.
747 748
		 * But if our normal children self-reap, then this child
		 * was prevented by ptrace and we must reap it now.
Linus Torvalds's avatar
Linus Torvalds committed
749
		 */
Roland McGrath's avatar
Roland McGrath committed
750 751 752
		if (!task_detached(p) && thread_group_empty(p)) {
			if (!same_thread_group(p->real_parent, parent))
				do_notify_parent(p, p->exit_signal);
753 754 755 756 757 758
			else {
				if (ign < 0)
					ign = ignoring_children(parent);
				if (ign)
					p->exit_signal = -1;
			}
Roland McGrath's avatar
Roland McGrath committed
759
		}
Linus Torvalds's avatar
Linus Torvalds committed
760

Roland McGrath's avatar
Roland McGrath committed
761
		if (task_detached(p)) {
Linus Torvalds's avatar
Linus Torvalds committed
762
			/*
Roland McGrath's avatar
Roland McGrath committed
763
			 * Mark it as in the process of being reaped.
Linus Torvalds's avatar
Linus Torvalds committed
764
			 */
Roland McGrath's avatar
Roland McGrath committed
765 766
			p->exit_state = EXIT_DEAD;
			list_add(&p->ptrace_entry, dead);
Linus Torvalds's avatar
Linus Torvalds committed
767 768
		}
	}
Roland McGrath's avatar
Roland McGrath committed
769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795
}

/*
 * Finish up exit-time ptrace cleanup.
 *
 * Called without locks.
 */
static void ptrace_exit_finish(struct task_struct *parent,
			       struct list_head *dead)
{
	struct task_struct *p, *n;

	BUG_ON(!list_empty(&parent->ptraced));

	list_for_each_entry_safe(p, n, dead, ptrace_entry) {
		list_del_init(&p->ptrace_entry);
		release_task(p);
	}
}

static void reparent_thread(struct task_struct *p, struct task_struct *father)
{
	if (p->pdeath_signal)
		/* We already hold the tasklist_lock here.  */
		group_send_sig_info(p->pdeath_signal, SEND_SIG_NOINFO, p);

	list_move_tail(&p->sibling, &p->real_parent->children);
Linus Torvalds's avatar
Linus Torvalds committed
796

797 798 799
	/* If this is a threaded reparent there is no need to
	 * notify anyone anything has happened.
	 */
800
	if (same_thread_group(p->real_parent, father))
801 802 803
		return;

	/* We don't want people slaying init.  */
804
	if (!task_detached(p))
805 806 807 808 809
		p->exit_signal = SIGCHLD;

	/* If we'd notified the old parent about this child's death,
	 * also notify the new parent.
	 */
Roland McGrath's avatar
Roland McGrath committed
810 811
	if (!ptrace_reparented(p) &&
	    p->exit_state == EXIT_ZOMBIE &&
812
	    !task_detached(p) && thread_group_empty(p))
813 814
		do_notify_parent(p, p->exit_signal);

815
	kill_orphaned_pgrp(p, father);
Linus Torvalds's avatar
Linus Torvalds committed
816 817 818 819 820 821
}

/*
 * When we die, we re-parent all our children.
 * Try to give them to another thread in our thread
 * group, and if no such member exists, give it to
822 823
 * the child reaper process (ie "init") in our pid
 * space.
Linus Torvalds's avatar
Linus Torvalds committed
824
 */
825
static struct task_struct *find_new_reaper(struct task_struct *father)
Linus Torvalds's avatar
Linus Torvalds committed
826
{
827 828
	struct pid_namespace *pid_ns = task_active_pid_ns(father);
	struct task_struct *thread;
Linus Torvalds's avatar
Linus Torvalds committed
829

830 831 832 833 834 835 836 837
	thread = father;
	while_each_thread(father, thread) {
		if (thread->flags & PF_EXITING)
			continue;
		if (unlikely(pid_ns->child_reaper == father))
			pid_ns->child_reaper = thread;
		return thread;
	}
Linus Torvalds's avatar
Linus Torvalds committed
838

839 840 841 842
	if (unlikely(pid_ns->child_reaper == father)) {
		write_unlock_irq(&tasklist_lock);
		if (unlikely(pid_ns == &init_pid_ns))
			panic("Attempted to kill init!");
Linus Torvalds's avatar
Linus Torvalds committed
843

844 845
		zap_pid_ns_processes(pid_ns);
		write_lock_irq(&tasklist_lock);
Linus Torvalds's avatar
Linus Torvalds committed
846
		/*
847 848 849
		 * We can not clear ->child_reaper or leave it alone.
		 * There may by stealth EXIT_DEAD tasks on ->children,
		 * forget_original_parent() must move them somewhere.
Linus Torvalds's avatar
Linus Torvalds committed
850
		 */
851
		pid_ns->child_reaper = init_pid_ns.child_reaper;
Linus Torvalds's avatar
Linus Torvalds committed
852
	}
853

854 855 856
	return pid_ns->child_reaper;
}

857
static void forget_original_parent(struct task_struct *father)
Linus Torvalds's avatar
Linus Torvalds committed
858
{
859
	struct task_struct *p, *n, *reaper;
Roland McGrath's avatar
Roland McGrath committed
860
	LIST_HEAD(ptrace_dead);
861 862

	write_lock_irq(&tasklist_lock);
863
	reaper = find_new_reaper(father);
Roland McGrath's avatar
Roland McGrath committed
864 865 866 867 868
	/*
	 * First clean up ptrace if we were using it.
	 */
	ptrace_exit(father, &ptrace_dead);

869
	list_for_each_entry_safe(p, n, &father->children, sibling) {
870
		p->real_parent = reaper;
Roland McGrath's avatar
Roland McGrath committed
871 872 873 874 875
		if (p->parent == father) {
			BUG_ON(p->ptrace);
			p->parent = p->real_parent;
		}
		reparent_thread(p, father);
Linus Torvalds's avatar
Linus Torvalds committed
876
	}
877 878 879 880

	write_unlock_irq(&tasklist_lock);
	BUG_ON(!list_empty(&father->children));

Roland McGrath's avatar
Roland McGrath committed
881
	ptrace_exit_finish(father, &ptrace_dead);
Linus Torvalds's avatar
Linus Torvalds committed
882 883 884 885 886 887
}

/*
 * Send signals to all our closest relatives so that they know
 * to properly mourn us..
 */
888
static void exit_notify(struct task_struct *tsk, int group_dead)
Linus Torvalds's avatar
Linus Torvalds committed
889
{
Roland McGrath's avatar
Roland McGrath committed
890 891
	int signal;
	void *cookie;
Linus Torvalds's avatar
Linus Torvalds committed
892 893 894 895 896 897 898 899 900

	/*
	 * 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)
	 */
901
	forget_original_parent(tsk);
902
	exit_task_namespaces(tsk);
Linus Torvalds's avatar
Linus Torvalds committed
903

904
	write_lock_irq(&tasklist_lock);
905 906
	if (group_dead)
		kill_orphaned_pgrp(tsk->group_leader, NULL);
Linus Torvalds's avatar
Linus Torvalds committed
907

908
	/* Let father know we died
Linus Torvalds's avatar
Linus Torvalds committed
909 910
	 *
	 * Thread signals are configurable, but you aren't going to use
911
	 * that to send signals to arbitary processes.
Linus Torvalds's avatar
Linus Torvalds committed
912 913 914 915 916 917 918 919 920 921
	 * That stops right now.
	 *
	 * If the parent exec id doesn't match the exec id we saved
	 * when we started then we know the parent has changed security
	 * domain.
	 *
	 * If our self_exec id doesn't match our parent_exec_id then
	 * we have changed execution domain as these two values started
	 * the same after a fork.
	 */
922
	if (tsk->exit_signal != SIGCHLD && !task_detached(tsk) &&
923
	    (tsk->parent_exec_id != tsk->real_parent->self_exec_id ||
924 925
	     tsk->self_exec_id != tsk->parent_exec_id) &&
	    !capable(CAP_KILL))
Linus Torvalds's avatar
Linus Torvalds committed
926 927
		tsk->exit_signal = SIGCHLD;

Roland McGrath's avatar
Roland McGrath committed
928
	signal = tracehook_notify_death(tsk, &cookie, group_dead);
929
	if (signal >= 0)
Roland McGrath's avatar
Roland McGrath committed
930
		signal = do_notify_parent(tsk, signal);
Linus Torvalds's avatar
Linus Torvalds committed
931

932
	tsk->exit_state = signal == DEATH_REAP ? EXIT_DEAD : EXIT_ZOMBIE;
Linus Torvalds's avatar
Linus Torvalds committed
933

934
	/* mt-exec, de_thread() is waiting for us */
935
	if (thread_group_leader(tsk) &&
936 937
	    tsk->signal->group_exit_task &&
	    tsk->signal->notify_count < 0)
938 939
		wake_up_process(tsk->signal->group_exit_task);

Linus Torvalds's avatar
Linus Torvalds committed
940 941
	write_unlock_irq(&tasklist_lock);

Roland McGrath's avatar
Roland McGrath committed
942 943
	tracehook_report_death(tsk, signal, cookie, group_dead);

Linus Torvalds's avatar
Linus Torvalds committed
944
	/* If the process is dead, release it - nobody will wait for it */
945
	if (signal == DEATH_REAP)
Linus Torvalds's avatar
Linus Torvalds committed
946 947 948
		release_task(tsk);
}

949 950 951 952 953 954 955
#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;

956
	free = stack_not_used(current);
957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973

	if (free >= lowest_to_date)
		return;

	spin_lock(&low_water_lock);
	if (free < lowest_to_date) {
		printk(KERN_WARNING "%s used greatest stack depth: %lu bytes "
				"left\n",
				current->comm, free);
		lowest_to_date = free;
	}
	spin_unlock(&low_water_lock);
}
#else
static inline void check_stack_usage(void) {}
#endif

974
NORET_TYPE void do_exit(long code)
Linus Torvalds's avatar
Linus Torvalds committed
975 976 977 978 979 980
{
	struct task_struct *tsk = current;
	int group_dead;

	profile_task_exit(tsk);

981 982
	WARN_ON(atomic_read(&tsk->fs_excl));

Linus Torvalds's avatar
Linus Torvalds committed
983 984 985 986 987
	if (unlikely(in_interrupt()))
		panic("Aiee, killing interrupt handler!");
	if (unlikely(!tsk->pid))
		panic("Attempted to kill the idle task!");

Roland McGrath's avatar
Roland McGrath committed
988
	tracehook_report_exit(&code);
Linus Torvalds's avatar
Linus Torvalds committed
989

990 991 992 993 994 995 996
	/*
	 * 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)) {
		printk(KERN_ALERT
			"Fixing recursive fault but reboot is needed!\n");
997 998 999 1000 1001 1002 1003 1004 1005 1006
		/*
		 * 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;
1007 1008 1009 1010
		set_current_state(TASK_UNINTERRUPTIBLE);
		schedule();
	}

Oleg Nesterov's avatar
Oleg Nesterov committed
1011
	exit_signals(tsk);  /* sets PF_EXITING */
1012 1013 1014 1015
	/*
	 * tsk->flags are checked in the futex code to protect against
	 * an exiting task cleaning up the robust pi futexes.
	 */
1016 1017
	smp_mb();
	spin_unlock_wait(&tsk->pi_lock);
Linus Torvalds's avatar
Linus Torvalds committed
1018 1019 1020

	if (unlikely(in_atomic()))
		printk(KERN_INFO "note: %s[%d] exited with preempt_count %d\n",
1021
				current->comm, task_pid_nr(current),
Linus Torvalds's avatar
Linus Torvalds committed
1022 1023 1024
				preempt_count());

	acct_update_integrals(tsk);
1025

Linus Torvalds's avatar
Linus Torvalds committed
1026
	group_dead = atomic_dec_and_test(&tsk->signal->live);
1027
	if (group_dead) {
1028
		hrtimer_cancel(&tsk->signal->real_timer);
1029
		exit_itimers(tsk->signal);
1030
	}
1031
	acct_collect(code, group_dead);
Miloslav Trmac's avatar
Miloslav Trmac committed
1032 1033
	if (group_dead)
		tty_audit_exit();
1034 1035
	if (unlikely(tsk->audit_context))
		audit_free(tsk);
1036

1037
	tsk->exit_code = code;
1038
	taskstats_exit(tsk, group_dead);
1039

Linus Torvalds's avatar
Linus Torvalds committed
1040 1041
	exit_mm(tsk);

1042
	if (group_dead)
1043
		acct_process();
1044 1045
	trace_sched_process_exit(tsk);

Linus Torvalds's avatar
Linus Torvalds committed
1046
	exit_sem(tsk);
1047 1048
	exit_files(tsk);
	exit_fs(tsk);
1049
	check_stack_usage();
Linus Torvalds's avatar
Linus Torvalds committed
1050
	exit_thread();
1051
	cgroup_exit(tsk, 1);
Linus Torvalds's avatar
Linus Torvalds committed
1052 1053 1054 1055

	if (group_dead && tsk->signal->leader)
		disassociate_ctty(1);

1056
	module_put(task_thread_info(tsk)->exec_domain->module);
Linus Torvalds's avatar
Linus Torvalds committed
1057 1058 1059
	if (tsk->binfmt)
		module_put(tsk->binfmt->module);

1060
	proc_exit_connector(tsk);
1061
	exit_notify(tsk, group_dead);
Linus Torvalds's avatar
Linus Torvalds committed
1062
#ifdef CONFIG_NUMA
1063
	mpol_put(tsk->mempolicy);
Linus Torvalds's avatar
Linus Torvalds committed
1064 1065
	tsk->mempolicy = NULL;
#endif
1066
#ifdef CONFIG_FUTEX
1067 1068 1069 1070 1071 1072 1073 1074
	/*
	 * This must happen late, after the PID is not
	 * hashed anymore:
	 */
	if (unlikely(!list_empty(&tsk->pi_state_list)))
		exit_pi_state_list(tsk);
	if (unlikely(current->pi_state_cache))
		kfree(current->pi_state_cache);
1075
#endif
1076
	/*
1077
	 * Make sure we are holding no locks:
1078
	 */
1079
	debug_check_no_locks_held(tsk);
1080 1081 1082 1083 1084 1085
	/*
	 * 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
1086

1087 1088 1089
	if (tsk->io_context)
		exit_io_context();

1090 1091 1092
	if (tsk->splice_pipe)
		__free_pipe_info(tsk->splice_pipe);

Coywolf Qi Hunt's avatar
Coywolf Qi Hunt committed
1093
	preempt_disable();
Oleg Nesterov's avatar
Oleg Nesterov committed
1094
	/* causes final put_task_struct in finish_task_switch(). */
1095
	tsk->state = TASK_DEAD;
Linus Torvalds's avatar
Linus Torvalds committed
1096 1097 1098
	schedule();
	BUG();
	/* Avoid "noreturn function does return".  */
Alan Cox's avatar
Alan Cox committed
1099 1100
	for (;;)
		cpu_relax();	/* For when BUG is null */
Linus Torvalds's avatar
Linus Torvalds committed
1101 1102
}

1103 1104
EXPORT_SYMBOL_GPL(do_exit);

Linus Torvalds's avatar
Linus Torvalds committed
1105 1106 1107 1108
NORET_TYPE void complete_and_exit(struct completion *comp, long code)
{
	if (comp)
		complete(comp);
Oleg Nesterov's avatar
Oleg Nesterov committed
1109

Linus Torvalds's avatar
Linus Torvalds committed
1110 1111 1112 1113 1114
	do_exit(code);
}

EXPORT_SYMBOL(complete_and_exit);

1115
SYSCALL_DEFINE1(exit, int, error_code)
Linus Torvalds's avatar
Linus Torvalds committed
1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126
{
	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).
 */
NORET_TYPE void
do_group_exit(int exit_code)
{
1127 1128
	struct signal_struct *sig = current->signal;

Linus Torvalds's avatar
Linus Torvalds committed
1129 1130
	BUG_ON(exit_code & 0x80); /* core dumps don't get here */

1131 1132
	if (signal_group_exit(sig))
		exit_code = sig->group_exit_code;
Linus Torvalds's avatar
Linus Torvalds committed
1133 1134 1135
	else if (!thread_group_empty(current)) {
		struct sighand_struct *const sighand = current->sighand;
		spin_lock_irq(&sighand->siglock);
1136
		if (signal_group_exit(sig))
Linus Torvalds's avatar
Linus Torvalds committed
1137 1138 1139 1140
			/* Another thread got here before we took the lock.  */
			exit_code = sig->group_exit_code;
		else {
			sig->group_exit_code = exit_code;
1141
			sig->flags = SIGNAL_GROUP_EXIT;
Linus Torvalds's avatar
Linus Torvalds committed
1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155
			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.
 */
1156
SYSCALL_DEFINE1(exit_group, int, error_code)
Linus Torvalds's avatar
Linus Torvalds committed
1157 1158
{
	do_group_exit((error_code & 0xff) << 8);
1159 1160
	/* NOTREACHED */
	return 0;
Linus Torvalds's avatar
Linus Torvalds committed
1161 1162
}

Eric W. Biederman's avatar
Eric W. Biederman committed
1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174
static struct pid *task_pid_type(struct task_struct *task, enum pid_type type)
{
	struct pid *pid = NULL;
	if (type == PIDTYPE_PID)
		pid = task->pids[type].pid;
	else if (type < PIDTYPE_MAX)
		pid = task->group_leader->pids[type].pid;
	return pid;
}

static int eligible_child(enum pid_type type, struct pid *pid, int options,
			  struct task_struct *p)
Linus Torvalds's avatar
Linus Torvalds committed
1175
{
1176 1177
	int err;

Eric W. Biederman's avatar
Eric W. Biederman committed
1178 1179
	if (type < PIDTYPE_MAX) {
		if (task_pid_type(p, type) != pid)
Linus Torvalds's avatar
Linus Torvalds committed
1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191
			return 0;
	}

	/* Wait for all children (clone and not) if __WALL is set;
	 * 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.) */
	if (((p->exit_signal != SIGCHLD) ^ ((options & __WCLONE) != 0))
	    && !(options & __WALL))
		return 0;

1192
	err = security_task_wait(p);
1193 1194
	if (err)
		return err;
Linus Torvalds's avatar
Linus Torvalds committed
1195

1196
	return 1;
Linus Torvalds's avatar
Linus Torvalds committed
1197 1198
}

1199
static int wait_noreap_copyout(struct task_struct *p, pid_t pid, uid_t uid,
Linus Torvalds's avatar
Linus Torvalds committed
1200 1201 1202 1203 1204
			       int why, int status,
			       struct siginfo __user *infop,
			       struct rusage __user *rusagep)
{
	int retval = rusagep ? getrusage(p, RUSAGE_BOTH, rusagep) : 0;
1205

Linus Torvalds's avatar
Linus Torvalds committed
1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229
	put_task_struct(p);
	if (!retval)
		retval = put_user(SIGCHLD, &infop->si_signo);
	if (!retval)
		retval = put_user(0, &infop->si_errno);
	if (!retval)
		retval = put_user((short)why, &infop->si_code);
	if (!retval)
		retval = put_user(pid, &infop->si_pid);
	if (!retval)
		retval = put_user(uid, &infop->si_uid);
	if (!retval)
		retval = put_user(status, &infop->si_status);
	if (!retval)
		retval = pid;
	return retval;
}

/*
 * 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.
 */
Roland McGrath's avatar
Roland McGrath committed
1230
static int wait_task_zombie(struct task_struct *p, int options,
Linus Torvalds's avatar
Linus Torvalds committed
1231 1232 1233 1234
			    struct siginfo __user *infop,
			    int __user *stat_addr, struct rusage __user *ru)
{
	unsigned long state;
1235
	int retval, status, traced;
1236
	pid_t pid = task_pid_vnr(p);
1237
	uid_t uid = __task_cred(p)->uid;
Linus Torvalds's avatar
Linus Torvalds committed
1238

Roland McGrath's avatar
Roland McGrath committed
1239 1240 1241 1242
	if (!likely(options & WEXITED))
		return 0;

	if (unlikely(options & WNOWAIT)) {
Linus Torvalds's avatar
Linus Torvalds committed
1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268
		int exit_code = p->exit_code;
		int why, status;

		get_task_struct(p);
		read_unlock(&tasklist_lock);
		if ((exit_code & 0x7f) == 0) {
			why = CLD_EXITED;
			status = exit_code >> 8;
		} else {
			why = (exit_code & 0x80) ? CLD_DUMPED : CLD_KILLED;
			status = exit_code & 0x7f;
		}
		return wait_noreap_copyout(p, pid, uid, why,
					   status, infop, ru);
	}

	/*
	 * Try to move the task's state to DEAD
	 * only one thread is allowed to do this:
	 */
	state = xchg(&p->exit_state, EXIT_DEAD);
	if (state != EXIT_ZOMBIE) {
		BUG_ON(state != EXIT_DEAD);
		return 0;
	}

1269
	traced = ptrace_reparented(p);
1270 1271

	if (likely(!traced)) {
1272 1273
		struct signal_struct *psig;
		struct signal_struct *sig;
1274
		struct task_cputime cputime;
1275

Linus Torvalds's avatar
Linus Torvalds committed
1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289
		/*
		 * 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
		 * p->signal fields, because they are only touched by
		 * __exit_signal, which runs with tasklist_lock
		 * write-locked anyway, and so is excluded here.  We do
		 * need to protect the access to p->parent->signal fields,
		 * as other threads in the parent group can be right
		 * here reaping other children at the same time.
1290 1291 1292 1293
		 *
		 * We use thread_group_cputime() 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
1294
		 */
1295
		thread_group_cputime(p, &cputime);
Linus Torvalds's avatar
Linus Torvalds committed
1296
		spin_lock_irq(&p->parent->sighand->siglock);
1297 1298 1299 1300
		psig = p->parent->signal;
		sig = p->signal;
		psig->cutime =
			cputime_add(psig->cutime,
1301 1302
			cputime_add(cputime.utime,
				    sig->cutime));
1303 1304
		psig->cstime =
			cputime_add(psig->cstime,
1305 1306
			cputime_add(cputime.stime,
				    sig->cstime));
1307 1308 1309 1310 1311
		psig->cgtime =
			cputime_add(psig->cgtime,
			cputime_add(p->gtime,
			cputime_add(sig->gtime,
				    sig->cgtime)));
1312 1313 1314 1315 1316 1317 1318 1319
		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;
1320 1321 1322 1323 1324 1325
		psig->cinblock +=
			task_io_get_inblock(p) +
			sig->inblock + sig->cinblock;
		psig->coublock +=
			task_io_get_oublock(p) +
			sig->oublock + sig->coublock;
1326 1327
		task_io_accounting_add(&psig->ioac, &p->ioac);
		task_io_accounting_add(&psig->ioac, &sig->ioac);
Linus Torvalds's avatar
Linus Torvalds committed
1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360
		spin_unlock_irq(&p->parent->sighand->siglock);
	}

	/*
	 * Now we are sure this task is interesting, and no other
	 * thread can reap it because we set its state to EXIT_DEAD.
	 */
	read_unlock(&tasklist_lock);

	retval = ru ? getrusage(p, RUSAGE_BOTH, ru) : 0;
	status = (p->signal->flags & SIGNAL_GROUP_EXIT)
		? p->signal->group_exit_code : p->exit_code;
	if (!retval && stat_addr)
		retval = put_user(status, stat_addr);
	if (!retval && infop)
		retval = put_user(SIGCHLD, &infop->si_signo);
	if (!retval && infop)
		retval = put_user(0, &infop->si_errno);
	if (!retval && infop) {
		int why;

		if ((status & 0x7f) == 0) {
			why = CLD_EXITED;
			status >>= 8;
		} else {
			why = (status & 0x80) ? CLD_DUMPED : CLD_KILLED;
			status &= 0x7f;
		}
		retval = put_user((short)why, &infop->si_code);
		if (!retval)
			retval = put_user(status, &infop->si_status);
	}
	if (!retval && infop)
1361