exit.c 45.8 KB
Newer Older
L
Linus Torvalds 已提交
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>
L
Linus Torvalds 已提交
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>
L
Linus Torvalds 已提交
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>
L
Linus Torvalds 已提交
22
#include <linux/file.h>
A
Al Viro 已提交
23
#include <linux/fdtable.h>
L
Linus Torvalds 已提交
24
#include <linux/binfmts.h>
S
Serge E. Hallyn 已提交
25
#include <linux/nsproxy.h>
26
#include <linux/pid_namespace.h>
L
Linus Torvalds 已提交
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>
L
Linus Torvalds 已提交
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>
L
Linus Torvalds 已提交
37
#include <linux/syscalls.h>
38
#include <linux/signal.h>
39
#include <linux/posix-timers.h>
M
Matt Helsley 已提交
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() */
A
Adrian Bunk 已提交
45
#include <linux/resource.h>
46
#include <linux/blkdev.h>
47
#include <linux/task_io_accounting_ops.h>
R
Roland McGrath 已提交
48
#include <linux/tracehook.h>
49
#include <trace/sched.h>
L
Linus Torvalds 已提交
50 51 52 53 54 55

#include <asm/uaccess.h>
#include <asm/unistd.h>
#include <asm/pgtable.h>
#include <asm/mmu_context.h>

A
Adrian Bunk 已提交
56 57
static void exit_mm(struct task_struct * tsk);

58 59 60 61 62
static inline int task_detached(struct task_struct *p)
{
	return p->exit_signal == -1;
}

L
Linus Torvalds 已提交
63 64 65 66 67 68 69
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);
70

71
		list_del_rcu(&p->tasks);
72
		__get_cpu_var(process_counts)--;
L
Linus Torvalds 已提交
73
	}
O
Oleg Nesterov 已提交
74
	list_del_rcu(&p->thread_group);
R
Roland McGrath 已提交
75
	list_del_init(&p->sibling);
L
Linus Torvalds 已提交
76 77
}

78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99
/*
 * 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:
		 */
100
		if (sig->group_exit_task && atomic_read(&sig->count) == sig->notify_count)
101
			wake_up_process(sig->group_exit_task);
102

103 104 105 106 107 108 109 110 111 112 113 114
		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.
		 */
115
		sig->gtime = cputime_add(sig->gtime, task_gtime(tsk));
116 117 118 119
		sig->min_flt += tsk->min_flt;
		sig->maj_flt += tsk->maj_flt;
		sig->nvcsw += tsk->nvcsw;
		sig->nivcsw += tsk->nivcsw;
120 121
		sig->inblock += task_io_get_inblock(tsk);
		sig->oublock += task_io_get_oublock(tsk);
122
		task_io_accounting_add(&sig->ioac, &tsk->ioac);
123 124 125
		sig = NULL; /* Marker for below. */
	}

126 127
	__unhash_process(tsk);

128 129 130 131 132 133
	/*
	 * Do this under ->siglock, we can race with another thread
	 * doing sigqueue_free() if we have SIGQUEUE_PREALLOC signals.
	 */
	flush_sigqueue(&tsk->pending);

134
	tsk->signal = NULL;
135
	tsk->sighand = NULL;
136 137
	spin_unlock(&sighand->siglock);

138
	__cleanup_sighand(sighand);
139 140 141
	clear_tsk_thread_flag(tsk,TIF_SIGPENDING);
	if (sig) {
		flush_sigqueue(&sig->shared_pending);
142
		taskstats_tgid_free(sig);
143 144 145 146 147
		/*
		 * Make sure ->signal can't go away under rq->lock,
		 * see account_group_exec_runtime().
		 */
		task_rq_unlock_wait(tsk);
148 149 150 151
		__cleanup_signal(sig);
	}
}

152 153
static void delayed_put_task_struct(struct rcu_head *rhp)
{
154 155 156 157
	struct task_struct *tsk = container_of(rhp, struct task_struct, rcu);

	trace_sched_process_free(tsk);
	put_task_struct(tsk);
158 159
}

R
Roland McGrath 已提交
160

L
Linus Torvalds 已提交
161 162
void release_task(struct task_struct * p)
{
163
	struct task_struct *leader;
L
Linus Torvalds 已提交
164
	int zap_leader;
165
repeat:
R
Roland McGrath 已提交
166
	tracehook_prepare_release_task(p);
L
Linus Torvalds 已提交
167
	atomic_dec(&p->user->processes);
168
	proc_flush_task(p);
L
Linus Torvalds 已提交
169
	write_lock_irq(&tasklist_lock);
R
Roland McGrath 已提交
170
	tracehook_finish_release_task(p);
L
Linus Torvalds 已提交
171
	__exit_signal(p);
172

L
Linus Torvalds 已提交
173 174 175 176 177 178 179 180
	/*
	 * 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) {
181
		BUG_ON(task_detached(leader));
L
Linus Torvalds 已提交
182 183 184 185 186 187 188 189 190
		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.
		 */
191
		zap_leader = task_detached(leader);
R
Roland McGrath 已提交
192 193 194 195 196 197 198

		/*
		 * 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;
L
Linus Torvalds 已提交
199 200 201 202
	}

	write_unlock_irq(&tasklist_lock);
	release_thread(p);
203
	call_rcu(&p->rcu, delayed_put_task_struct);
L
Linus Torvalds 已提交
204 205 206 207 208 209 210 211 212 213

	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...
214 215
 *
 * The caller must hold rcu lock or the tasklist lock.
L
Linus Torvalds 已提交
216
 */
217
struct pid *session_of_pgrp(struct pid *pgrp)
L
Linus Torvalds 已提交
218 219
{
	struct task_struct *p;
220
	struct pid *sid = NULL;
221

222
	p = pid_task(pgrp, PIDTYPE_PGID);
223
	if (p == NULL)
224
		p = pid_task(pgrp, PIDTYPE_PID);
225
	if (p != NULL)
226
		sid = task_session(p);
227

L
Linus Torvalds 已提交
228 229 230 231 232 233 234 235 236 237 238
	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?"
 */
239
static int will_become_orphaned_pgrp(struct pid *pgrp, struct task_struct *ignored_task)
L
Linus Torvalds 已提交
240 241 242
{
	struct task_struct *p;

243
	do_each_pid_task(pgrp, PIDTYPE_PGID, p) {
244 245 246
		if ((p == ignored_task) ||
		    (p->exit_state && thread_group_empty(p)) ||
		    is_global_init(p->real_parent))
L
Linus Torvalds 已提交
247
			continue;
248

249
		if (task_pgrp(p->real_parent) != pgrp &&
250 251
		    task_session(p->real_parent) == task_session(p))
			return 0;
252
	} while_each_pid_task(pgrp, PIDTYPE_PGID, p);
253 254

	return 1;
L
Linus Torvalds 已提交
255 256
}

257
int is_current_pgrp_orphaned(void)
L
Linus Torvalds 已提交
258 259 260 261
{
	int retval;

	read_lock(&tasklist_lock);
262
	retval = will_become_orphaned_pgrp(task_pgrp(current), NULL);
L
Linus Torvalds 已提交
263 264 265 266 267
	read_unlock(&tasklist_lock);

	return retval;
}

268
static int has_stopped_jobs(struct pid *pgrp)
L
Linus Torvalds 已提交
269 270 271 272
{
	int retval = 0;
	struct task_struct *p;

273
	do_each_pid_task(pgrp, PIDTYPE_PGID, p) {
M
Matthew Wilcox 已提交
274
		if (!task_is_stopped(p))
L
Linus Torvalds 已提交
275 276 277
			continue;
		retval = 1;
		break;
278
	} while_each_pid_task(pgrp, PIDTYPE_PGID, p);
L
Linus Torvalds 已提交
279 280 281
	return retval;
}

282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312
/*
 * 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);
	}
}

L
Linus Torvalds 已提交
313
/**
314
 * reparent_to_kthreadd - Reparent the calling kernel thread to kthreadd
L
Linus Torvalds 已提交
315 316
 *
 * If a kernel thread is launched as a result of a system call, or if
317 318
 * 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.
L
Linus Torvalds 已提交
319 320 321 322
 *
 * 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.
 *
323
 * NOTE that reparent_to_kthreadd() gives the caller full capabilities.
L
Linus Torvalds 已提交
324
 */
325
static void reparent_to_kthreadd(void)
L
Linus Torvalds 已提交
326 327 328 329 330
{
	write_lock_irq(&tasklist_lock);

	ptrace_unlink(current);
	/* Reparent to init */
331
	current->real_parent = current->parent = kthreadd_task;
R
Roland McGrath 已提交
332
	list_move_tail(&current->sibling, &current->real_parent->children);
L
Linus Torvalds 已提交
333 334 335 336

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

337
	if (task_nice(current) < 0)
L
Linus Torvalds 已提交
338 339 340 341 342 343 344 345 346 347 348 349
		set_user_nice(current, 0);
	/* cpus_allowed? */
	/* rt_priority? */
	/* signals? */
	security_task_reparent_to_init(current);
	memcpy(current->signal->rlim, init_task.signal->rlim,
	       sizeof(current->signal->rlim));
	atomic_inc(&(INIT_USER->__count));
	write_unlock_irq(&tasklist_lock);
	switch_uid(INIT_USER);
}

350
void __set_special_pids(struct pid *pid)
L
Linus Torvalds 已提交
351
{
352
	struct task_struct *curr = current->group_leader;
353
	pid_t nr = pid_nr(pid);
L
Linus Torvalds 已提交
354

355
	if (task_session(curr) != pid) {
356
		change_pid(curr, PIDTYPE_SID, pid);
357
		set_task_session(curr, nr);
L
Linus Torvalds 已提交
358
	}
359
	if (task_pgrp(curr) != pid) {
360
		change_pid(curr, PIDTYPE_PGID, pid);
361
		set_task_pgrp(curr, nr);
L
Linus Torvalds 已提交
362 363 364
	}
}

365
static void set_special_pids(struct pid *pid)
L
Linus Torvalds 已提交
366 367
{
	write_lock_irq(&tasklist_lock);
368
	__set_special_pids(pid);
L
Linus Torvalds 已提交
369 370 371 372 373 374 375 376 377 378
	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)
{
379
	if (!valid_signal(sig) || sig < 1)
L
Linus Torvalds 已提交
380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399
		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)
{
400
	if (!valid_signal(sig) || sig < 1)
L
Linus Torvalds 已提交
401 402 403
		return -EINVAL;

	spin_lock_irq(&current->sighand->siglock);
404
	current->sighand->action[(sig)-1].sa.sa_handler = SIG_IGN;
L
Linus Torvalds 已提交
405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432
	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;
	struct fs_struct *fs;
	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);
433 434 435 436
	/*
	 * We don't want to have TIF_FREEZE set if the system-wide hibernation
	 * or suspend transition begins right now.
	 */
O
Oleg Nesterov 已提交
437
	current->flags |= (PF_NOFREEZE | PF_KTHREAD);
L
Linus Torvalds 已提交
438

439 440 441 442
	if (current->nsproxy != &init_nsproxy) {
		get_nsproxy(&init_nsproxy);
		switch_task_namespaces(current, &init_nsproxy);
	}
443
	set_special_pids(&init_struct_pid);
444
	proc_clear_tty(current);
L
Linus Torvalds 已提交
445 446 447 448 449 450 451 452 453 454 455 456

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

	/* Become as one with the init task */

	exit_fs(current);	/* current->fs->count--; */
	fs = init_task.fs;
	current->fs = fs;
	atomic_inc(&fs->count);
S
Serge E. Hallyn 已提交
457

458
	exit_files(current);
L
Linus Torvalds 已提交
459 460 461
	current->files = init_task.files;
	atomic_inc(&current->files->count);

462
	reparent_to_kthreadd();
L
Linus Torvalds 已提交
463 464 465 466
}

EXPORT_SYMBOL(daemonize);

467
static void close_files(struct files_struct * files)
L
Linus Torvalds 已提交
468 469
{
	int i, j;
470
	struct fdtable *fdt;
L
Linus Torvalds 已提交
471 472

	j = 0;
473 474 475 476 477 478

	/*
	 * It is safe to dereference the fd table without RCU or
	 * ->file_lock because this is the last reference to the
	 * files structure.
	 */
479
	fdt = files_fdtable(files);
L
Linus Torvalds 已提交
480 481 482
	for (;;) {
		unsigned long set;
		i = j * __NFDBITS;
483
		if (i >= fdt->max_fds)
L
Linus Torvalds 已提交
484
			break;
485
		set = fdt->open_fds->fds_bits[j++];
L
Linus Torvalds 已提交
486 487
		while (set) {
			if (set & 1) {
488
				struct file * file = xchg(&fdt->fd[i], NULL);
489
				if (file) {
L
Linus Torvalds 已提交
490
					filp_close(file, files);
491 492
					cond_resched();
				}
L
Linus Torvalds 已提交
493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512
			}
			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;
}

513
void put_files_struct(struct files_struct *files)
L
Linus Torvalds 已提交
514
{
515 516
	struct fdtable *fdt;

L
Linus Torvalds 已提交
517 518 519 520
	if (atomic_dec_and_test(&files->count)) {
		close_files(files);
		/*
		 * Free the fd and fdset arrays if we expanded them.
521 522 523
		 * If the fdtable was embedded, pass files for freeing
		 * at the end of the RCU grace period. Otherwise,
		 * you can free files immediately.
L
Linus Torvalds 已提交
524
		 */
525
		fdt = files_fdtable(files);
526
		if (fdt != &files->fdtab)
527
			kmem_cache_free(files_cachep, files);
528
		free_fdtable(fdt);
L
Linus Torvalds 已提交
529 530 531
	}
}

532
void reset_files_struct(struct files_struct *files)
533
{
534
	struct task_struct *tsk = current;
535 536 537 538 539 540 541 542 543
	struct files_struct *old;

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

544
void exit_files(struct task_struct *tsk)
L
Linus Torvalds 已提交
545 546 547 548 549 550 551 552 553 554 555
{
	struct files_struct * files = tsk->files;

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

556
void put_fs_struct(struct fs_struct *fs)
L
Linus Torvalds 已提交
557 558 559
{
	/* No need to hold fs->lock if we are killing it */
	if (atomic_dec_and_test(&fs->count)) {
J
Jan Blunck 已提交
560 561
		path_put(&fs->root);
		path_put(&fs->pwd);
L
Linus Torvalds 已提交
562 563 564 565
		kmem_cache_free(fs_cachep, fs);
	}
}

566
void exit_fs(struct task_struct *tsk)
L
Linus Torvalds 已提交
567 568 569 570 571 572 573
{
	struct fs_struct * fs = tsk->fs;

	if (fs) {
		task_lock(tsk);
		tsk->fs = NULL;
		task_unlock(tsk);
574
		put_fs_struct(fs);
L
Linus Torvalds 已提交
575 576 577 578 579
	}
}

EXPORT_SYMBOL_GPL(exit_fs);

580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632
#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);
633 634 635 636 637 638 639 640 641 642
	/*
	 * We found no owner yet mm_users > 1: this implies that we are
	 * most likely racing with swapoff (try_to_unuse()) or /proc or
	 * ptrace or page migration (get_task_mm()).  Mark owner as NULL,
	 * so that subsystems can understand the callback and take action.
	 */
	down_write(&mm->mmap_sem);
	cgroup_mm_owner_callbacks(mm->owner, NULL);
	mm->owner = NULL;
	up_write(&mm->mmap_sem);
643 644 645 646 647
	return;

assign_new_owner:
	BUG_ON(c == p);
	get_task_struct(c);
648 649
	read_unlock(&tasklist_lock);
	down_write(&mm->mmap_sem);
650 651 652 653 654 655 656
	/*
	 * The task_lock protects c->mm from changing.
	 * We always want mm->owner->mm == mm
	 */
	task_lock(c);
	if (c->mm != mm) {
		task_unlock(c);
657
		up_write(&mm->mmap_sem);
658 659 660 661 662 663
		put_task_struct(c);
		goto retry;
	}
	cgroup_mm_owner_callbacks(mm->owner, c);
	mm->owner = c;
	task_unlock(c);
664
	up_write(&mm->mmap_sem);
665 666 667 668
	put_task_struct(c);
}
#endif /* CONFIG_MM_OWNER */

L
Linus Torvalds 已提交
669 670 671 672
/*
 * Turn us into a lazy TLB process if we
 * aren't already..
 */
A
Adrian Bunk 已提交
673
static void exit_mm(struct task_struct * tsk)
L
Linus Torvalds 已提交
674 675
{
	struct mm_struct *mm = tsk->mm;
676
	struct core_state *core_state;
L
Linus Torvalds 已提交
677 678 679 680 681 682

	mm_release(tsk, mm);
	if (!mm)
		return;
	/*
	 * Serialize with any possible pending coredump.
683
	 * We must hold mmap_sem around checking core_state
L
Linus Torvalds 已提交
684
	 * and clearing tsk->mm.  The core-inducing thread
685
	 * will increment ->nr_threads for each thread in the
L
Linus Torvalds 已提交
686 687 688
	 * group with ->mm != NULL.
	 */
	down_read(&mm->mmap_sem);
689 690 691
	core_state = mm->core_state;
	if (core_state) {
		struct core_thread self;
L
Linus Torvalds 已提交
692
		up_read(&mm->mmap_sem);
693

694 695 696 697 698 699 700 701
		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);
L
Linus Torvalds 已提交
702

O
Oleg Nesterov 已提交
703 704 705 706 707 708 709
		for (;;) {
			set_task_state(tsk, TASK_UNINTERRUPTIBLE);
			if (!self.task) /* see coredump_finish() */
				break;
			schedule();
		}
		__set_task_state(tsk, TASK_RUNNING);
L
Linus Torvalds 已提交
710 711 712
		down_read(&mm->mmap_sem);
	}
	atomic_inc(&mm->mm_count);
E
Eric Sesterhenn 已提交
713
	BUG_ON(mm != tsk->active_mm);
L
Linus Torvalds 已提交
714 715 716 717 718
	/* more a memory barrier than a real lock */
	task_lock(tsk);
	tsk->mm = NULL;
	up_read(&mm->mmap_sem);
	enter_lazy_tlb(mm, current);
719 720
	/* We don't want this task to be frozen prematurely */
	clear_freeze_flag(tsk);
L
Linus Torvalds 已提交
721
	task_unlock(tsk);
722
	mm_update_next_owner(mm);
L
Linus Torvalds 已提交
723 724 725
	mmput(mm);
}

726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742
/*
 * 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)
{
	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;
}

R
Roland McGrath 已提交
743 744 745 746 747 748 749
/*
 * 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)
L
Linus Torvalds 已提交
750
{
R
Roland McGrath 已提交
751
	struct task_struct *p, *n;
752
	int ign = -1;
753

R
Roland McGrath 已提交
754 755 756 757 758 759 760 761 762 763 764 765 766 767
	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.
768 769
		 * But if our normal children self-reap, then this child
		 * was prevented by ptrace and we must reap it now.
L
Linus Torvalds 已提交
770
		 */
R
Roland McGrath 已提交
771 772 773
		if (!task_detached(p) && thread_group_empty(p)) {
			if (!same_thread_group(p->real_parent, parent))
				do_notify_parent(p, p->exit_signal);
774 775 776 777 778 779
			else {
				if (ign < 0)
					ign = ignoring_children(parent);
				if (ign)
					p->exit_signal = -1;
			}
R
Roland McGrath 已提交
780
		}
L
Linus Torvalds 已提交
781

R
Roland McGrath 已提交
782
		if (task_detached(p)) {
L
Linus Torvalds 已提交
783
			/*
R
Roland McGrath 已提交
784
			 * Mark it as in the process of being reaped.
L
Linus Torvalds 已提交
785
			 */
R
Roland McGrath 已提交
786 787
			p->exit_state = EXIT_DEAD;
			list_add(&p->ptrace_entry, dead);
L
Linus Torvalds 已提交
788 789
		}
	}
R
Roland McGrath 已提交
790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816
}

/*
 * 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);
L
Linus Torvalds 已提交
817

818 819 820
	/* If this is a threaded reparent there is no need to
	 * notify anyone anything has happened.
	 */
821
	if (same_thread_group(p->real_parent, father))
822 823 824
		return;

	/* We don't want people slaying init.  */
825
	if (!task_detached(p))
826 827 828 829 830
		p->exit_signal = SIGCHLD;

	/* If we'd notified the old parent about this child's death,
	 * also notify the new parent.
	 */
R
Roland McGrath 已提交
831 832
	if (!ptrace_reparented(p) &&
	    p->exit_state == EXIT_ZOMBIE &&
833
	    !task_detached(p) && thread_group_empty(p))
834 835
		do_notify_parent(p, p->exit_signal);

836
	kill_orphaned_pgrp(p, father);
L
Linus Torvalds 已提交
837 838 839 840 841 842
}

/*
 * 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
843 844
 * the child reaper process (ie "init") in our pid
 * space.
L
Linus Torvalds 已提交
845
 */
846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877
static struct task_struct *find_new_reaper(struct task_struct *father)
{
	struct pid_namespace *pid_ns = task_active_pid_ns(father);
	struct task_struct *thread;

	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;
	}

	if (unlikely(pid_ns->child_reaper == father)) {
		write_unlock_irq(&tasklist_lock);
		if (unlikely(pid_ns == &init_pid_ns))
			panic("Attempted to kill init!");

		zap_pid_ns_processes(pid_ns);
		write_lock_irq(&tasklist_lock);
		/*
		 * 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.
		 */
		pid_ns->child_reaper = init_pid_ns.child_reaper;
	}

	return pid_ns->child_reaper;
}

878
static void forget_original_parent(struct task_struct *father)
L
Linus Torvalds 已提交
879
{
880
	struct task_struct *p, *n, *reaper;
R
Roland McGrath 已提交
881
	LIST_HEAD(ptrace_dead);
882 883

	write_lock_irq(&tasklist_lock);
884
	reaper = find_new_reaper(father);
R
Roland McGrath 已提交
885 886 887 888 889
	/*
	 * First clean up ptrace if we were using it.
	 */
	ptrace_exit(father, &ptrace_dead);

890
	list_for_each_entry_safe(p, n, &father->children, sibling) {
891
		p->real_parent = reaper;
R
Roland McGrath 已提交
892 893 894 895 896
		if (p->parent == father) {
			BUG_ON(p->ptrace);
			p->parent = p->real_parent;
		}
		reparent_thread(p, father);
L
Linus Torvalds 已提交
897
	}
898 899 900 901

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

R
Roland McGrath 已提交
902
	ptrace_exit_finish(father, &ptrace_dead);
L
Linus Torvalds 已提交
903 904 905 906 907 908
}

/*
 * Send signals to all our closest relatives so that they know
 * to properly mourn us..
 */
909
static void exit_notify(struct task_struct *tsk, int group_dead)
L
Linus Torvalds 已提交
910
{
R
Roland McGrath 已提交
911 912
	int signal;
	void *cookie;
L
Linus Torvalds 已提交
913 914 915 916 917 918 919 920 921

	/*
	 * 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)
	 */
922
	forget_original_parent(tsk);
923
	exit_task_namespaces(tsk);
L
Linus Torvalds 已提交
924

I
Ingo Molnar 已提交
925 926 927 928 929 930
	/*
	 * Flush inherited counters to the parent - before the parent
	 * gets woken up by child-exit notifications.
	 */
	perf_counter_exit_task(tsk);

931
	write_lock_irq(&tasklist_lock);
932 933
	if (group_dead)
		kill_orphaned_pgrp(tsk->group_leader, NULL);
L
Linus Torvalds 已提交
934

935
	/* Let father know we died
L
Linus Torvalds 已提交
936 937
	 *
	 * Thread signals are configurable, but you aren't going to use
938
	 * that to send signals to arbitary processes.
L
Linus Torvalds 已提交
939 940 941 942 943 944 945 946 947 948
	 * 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.
	 */
949
	if (tsk->exit_signal != SIGCHLD && !task_detached(tsk) &&
950
	    (tsk->parent_exec_id != tsk->real_parent->self_exec_id ||
951 952
	     tsk->self_exec_id != tsk->parent_exec_id) &&
	    !capable(CAP_KILL))
L
Linus Torvalds 已提交
953 954
		tsk->exit_signal = SIGCHLD;

R
Roland McGrath 已提交
955
	signal = tracehook_notify_death(tsk, &cookie, group_dead);
956
	if (signal >= 0)
R
Roland McGrath 已提交
957
		signal = do_notify_parent(tsk, signal);
L
Linus Torvalds 已提交
958

959
	tsk->exit_state = signal == DEATH_REAP ? EXIT_DEAD : EXIT_ZOMBIE;
L
Linus Torvalds 已提交
960

961
	/* mt-exec, de_thread() is waiting for us */
962
	if (thread_group_leader(tsk) &&
963 964
	    tsk->signal->group_exit_task &&
	    tsk->signal->notify_count < 0)
965 966
		wake_up_process(tsk->signal->group_exit_task);

L
Linus Torvalds 已提交
967 968
	write_unlock_irq(&tasklist_lock);

R
Roland McGrath 已提交
969 970
	tracehook_report_death(tsk, signal, cookie, group_dead);

L
Linus Torvalds 已提交
971
	/* If the process is dead, release it - nobody will wait for it */
972
	if (signal == DEATH_REAP)
L
Linus Torvalds 已提交
973 974 975
		release_task(tsk);
}

976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003
#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 *n = end_of_stack(current);
	unsigned long free;

	while (*n == 0)
		n++;
	free = (unsigned long)n - (unsigned long)end_of_stack(current);

	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

1004
NORET_TYPE void do_exit(long code)
L
Linus Torvalds 已提交
1005 1006 1007 1008 1009 1010
{
	struct task_struct *tsk = current;
	int group_dead;

	profile_task_exit(tsk);

1011 1012
	WARN_ON(atomic_read(&tsk->fs_excl));

L
Linus Torvalds 已提交
1013 1014 1015 1016 1017
	if (unlikely(in_interrupt()))
		panic("Aiee, killing interrupt handler!");
	if (unlikely(!tsk->pid))
		panic("Attempted to kill the idle task!");

R
Roland McGrath 已提交
1018
	tracehook_report_exit(&code);
L
Linus Torvalds 已提交
1019

1020 1021 1022 1023 1024 1025 1026
	/*
	 * 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");
1027 1028 1029 1030 1031 1032 1033 1034 1035 1036
		/*
		 * 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;
1037 1038
		if (tsk->io_context)
			exit_io_context();
1039 1040 1041 1042
		set_current_state(TASK_UNINTERRUPTIBLE);
		schedule();
	}

O
Oleg Nesterov 已提交
1043
	exit_signals(tsk);  /* sets PF_EXITING */
1044 1045 1046 1047
	/*
	 * tsk->flags are checked in the futex code to protect against
	 * an exiting task cleaning up the robust pi futexes.
	 */
1048 1049
	smp_mb();
	spin_unlock_wait(&tsk->pi_lock);
L
Linus Torvalds 已提交
1050 1051 1052

	if (unlikely(in_atomic()))
		printk(KERN_INFO "note: %s[%d] exited with preempt_count %d\n",
1053
				current->comm, task_pid_nr(current),
L
Linus Torvalds 已提交
1054 1055 1056
				preempt_count());

	acct_update_integrals(tsk);
1057 1058 1059 1060
	if (tsk->mm) {
		update_hiwater_rss(tsk->mm);
		update_hiwater_vm(tsk->mm);
	}
L
Linus Torvalds 已提交
1061
	group_dead = atomic_dec_and_test(&tsk->signal->live);
1062
	if (group_dead) {
1063
		hrtimer_cancel(&tsk->signal->real_timer);
1064
		exit_itimers(tsk->signal);
1065
	}
1066
	acct_collect(code, group_dead);
M
Miloslav Trmac 已提交
1067 1068
	if (group_dead)
		tty_audit_exit();
1069 1070
	if (unlikely(tsk->audit_context))
		audit_free(tsk);
1071

1072
	tsk->exit_code = code;
1073
	taskstats_exit(tsk, group_dead);
1074

L
Linus Torvalds 已提交
1075 1076
	exit_mm(tsk);

1077
	if (group_dead)
1078
		acct_process();
1079 1080
	trace_sched_process_exit(tsk);

L
Linus Torvalds 已提交
1081
	exit_sem(tsk);
1082 1083
	exit_files(tsk);
	exit_fs(tsk);
1084
	check_stack_usage();
L
Linus Torvalds 已提交
1085
	exit_thread();
1086
	cgroup_exit(tsk, 1);
L
Linus Torvalds 已提交
1087 1088 1089 1090 1091
	exit_keys(tsk);

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

A
Al Viro 已提交
1092
	module_put(task_thread_info(tsk)->exec_domain->module);
L
Linus Torvalds 已提交
1093 1094 1095
	if (tsk->binfmt)
		module_put(tsk->binfmt->module);

M
Matt Helsley 已提交
1096
	proc_exit_connector(tsk);
1097
	exit_notify(tsk, group_dead);
L
Linus Torvalds 已提交
1098
#ifdef CONFIG_NUMA
1099
	mpol_put(tsk->mempolicy);
L
Linus Torvalds 已提交
1100 1101
	tsk->mempolicy = NULL;
#endif
1102
#ifdef CONFIG_FUTEX
1103 1104 1105 1106
	if (unlikely(!list_empty(&tsk->pi_state_list)))
		exit_pi_state_list(tsk);
	if (unlikely(current->pi_state_cache))
		kfree(current->pi_state_cache);
1107
#endif
1108
	/*
1109
	 * Make sure we are holding no locks:
1110
	 */
1111
	debug_check_no_locks_held(tsk);
1112 1113 1114 1115 1116 1117
	/*
	 * 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;
L
Linus Torvalds 已提交
1118

1119 1120 1121
	if (tsk->io_context)
		exit_io_context();

1122 1123 1124
	if (tsk->splice_pipe)
		__free_pipe_info(tsk->splice_pipe);

I
Ingo Molnar 已提交
1125 1126 1127 1128 1129 1130
	/*
	 * These must happen late, after the PID is not
	 * hashed anymore, but still at a point that may sleep:
	 */
	perf_counter_exit_task(tsk);

C
Coywolf Qi Hunt 已提交
1131
	preempt_disable();
O
Oleg Nesterov 已提交
1132
	/* causes final put_task_struct in finish_task_switch(). */
1133
	tsk->state = TASK_DEAD;
C
Coywolf Qi Hunt 已提交
1134

L
Linus Torvalds 已提交
1135 1136 1137
	schedule();
	BUG();
	/* Avoid "noreturn function does return".  */
A
Alan Cox 已提交
1138 1139
	for (;;)
		cpu_relax();	/* For when BUG is null */
L
Linus Torvalds 已提交
1140 1141
}

1142 1143
EXPORT_SYMBOL_GPL(do_exit);

L
Linus Torvalds 已提交
1144 1145 1146 1147
NORET_TYPE void complete_and_exit(struct completion *comp, long code)
{
	if (comp)
		complete(comp);
O
Oleg Nesterov 已提交
1148

L
Linus Torvalds 已提交
1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165
	do_exit(code);
}

EXPORT_SYMBOL(complete_and_exit);

asmlinkage long sys_exit(int error_code)
{
	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)
{
1166 1167
	struct signal_struct *sig = current->signal;

L
Linus Torvalds 已提交
1168 1169
	BUG_ON(exit_code & 0x80); /* core dumps don't get here */

1170 1171
	if (signal_group_exit(sig))
		exit_code = sig->group_exit_code;
L
Linus Torvalds 已提交
1172 1173 1174
	else if (!thread_group_empty(current)) {
		struct sighand_struct *const sighand = current->sighand;
		spin_lock_irq(&sighand->siglock);
1175
		if (signal_group_exit(sig))
L
Linus Torvalds 已提交
1176 1177 1178 1179
			/* Another thread got here before we took the lock.  */
			exit_code = sig->group_exit_code;
		else {
			sig->group_exit_code = exit_code;
1180
			sig->flags = SIGNAL_GROUP_EXIT;
L
Linus Torvalds 已提交
1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199
			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.
 */
asmlinkage void sys_exit_group(int error_code)
{
	do_group_exit((error_code & 0xff) << 8);
}

E
Eric W. Biederman 已提交
1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211
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)
L
Linus Torvalds 已提交
1212
{
1213 1214
	int err;

E
Eric W. Biederman 已提交
1215 1216
	if (type < PIDTYPE_MAX) {
		if (task_pid_type(p, type) != pid)
L
Linus Torvalds 已提交
1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228
			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;

1229
	err = security_task_wait(p);
1230 1231
	if (err)
		return err;
L
Linus Torvalds 已提交
1232

1233
	return 1;
L
Linus Torvalds 已提交
1234 1235
}

1236
static int wait_noreap_copyout(struct task_struct *p, pid_t pid, uid_t uid,
L
Linus Torvalds 已提交
1237 1238 1239 1240 1241
			       int why, int status,
			       struct siginfo __user *infop,
			       struct rusage __user *rusagep)
{
	int retval = rusagep ? getrusage(p, RUSAGE_BOTH, rusagep) : 0;
1242

L
Linus Torvalds 已提交
1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266
	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.
 */
R
Roland McGrath 已提交
1267
static int wait_task_zombie(struct task_struct *p, int options,
L
Linus Torvalds 已提交
1268 1269 1270 1271
			    struct siginfo __user *infop,
			    int __user *stat_addr, struct rusage __user *ru)
{
	unsigned long state;
1272
	int retval, status, traced;
1273
	pid_t pid = task_pid_vnr(p);
L
Linus Torvalds 已提交
1274

R
Roland McGrath 已提交
1275 1276 1277 1278
	if (!likely(options & WEXITED))
		return 0;

	if (unlikely(options & WNOWAIT)) {
L
Linus Torvalds 已提交
1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305
		uid_t uid = p->uid;
		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;
	}

1306
	traced = ptrace_reparented(p);
1307 1308

	if (likely(!traced)) {
1309 1310
		struct signal_struct *psig;
		struct signal_struct *sig;
1311
		struct task_cputime cputime;
1312

L
Linus Torvalds 已提交
1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326
		/*
		 * 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.
1327 1328 1329 1330
		 *
		 * 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.
L
Linus Torvalds 已提交
1331 1332
		 */
		spin_lock_irq(&p->parent->sighand->siglock);
1333 1334
		psig = p->parent->signal;
		sig = p->signal;
1335
		thread_group_cputime(p, &cputime);
1336 1337
		psig->cutime =
			cputime_add(psig->cutime,
1338 1339
			cputime_add(cputime.utime,
				    sig->cutime));
1340 1341
		psig->cstime =
			cputime_add(psig->cstime,
1342 1343
			cputime_add(cputime.stime,
				    sig->cstime));
1344 1345 1346 1347 1348
		psig->cgtime =
			cputime_add(psig->cgtime,
			cputime_add(p->gtime,
			cputime_add(sig->gtime,
				    sig->cgtime)));
1349 1350 1351 1352 1353 1354 1355 1356
		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;
1357 1358 1359 1360 1361 1362
		psig->cinblock +=
			task_io_get_inblock(p) +
			sig->inblock + sig->cinblock;
		psig->coublock +=
			task_io_get_oublock(p) +
			sig->oublock + sig->coublock;
1363 1364
		task_io_accounting_add(&psig->ioac, &p->ioac);
		task_io_accounting_add(&psig->ioac, &sig->ioac);
L
Linus Torvalds 已提交
1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397
		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)
1398
		retval = put_user(pid, &infop->si_pid);
L
Linus Torvalds 已提交
1399 1400
	if (!retval && infop)
		retval = put_user(p->uid, &infop->si_uid);
1401
	if (!retval)
1402
		retval = pid;
1403 1404

	if (traced) {
L
Linus Torvalds 已提交
1405
		write_lock_irq(&tasklist_lock);
1406 1407 1408 1409 1410 1411 1412
		/* We dropped tasklist, ptracer could die and untrace */
		ptrace_unlink(p);
		/*
		 * If this is not a detached task, notify the parent.
		 * If it's still not detached after that, don't release
		 * it now.
		 */
1413
		if (!task_detached(p)) {
1414
			do_notify_parent(p, p->exit_signal);
1415
			if (!task_detached(p)) {
1416 1417
				p->exit_state = EXIT_ZOMBIE;
				p = NULL;
L
Linus Torvalds 已提交
1418 1419 1420 1421 1422 1423
			}
		}
		write_unlock_irq(&tasklist_lock);
	}
	if (p != NULL)
		release_task(p);
1424

L
Linus Torvalds 已提交
1425 1426 1427 1428 1429 1430 1431 1432 1433
	return retval;
}

/*
 * Handle sys_wait4 work for one task in state TASK_STOPPED.  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.
 */
R
Roland McGrath 已提交
1434
static int wait_task_stopped(int ptrace, struct task_struct *p,
R
Roland McGrath 已提交
1435
			     int options, struct siginfo __user *infop,
L
Linus Torvalds 已提交
1436 1437
			     int __user *stat_addr, struct rusage __user *ru)
{
1438 1439
	int retval, exit_code, why;
	uid_t uid = 0; /* unneeded, required by compiler */
1440
	pid_t pid;
L
Linus Torvalds 已提交
1441

R
Roland McGrath 已提交
1442
	if (!(options & WUNTRACED))
R
Roland McGrath 已提交
1443 1444
		return 0;

1445 1446 1447 1448 1449 1450
	exit_code = 0;
	spin_lock_irq(&p->sighand->siglock);

	if (unlikely(!task_is_stopped_or_traced(p)))
		goto unlock_sig;

R
Roland McGrath 已提交
1451
	if (!ptrace && p->signal->group_stop_count > 0)
L
Linus Torvalds 已提交
1452 1453 1454 1455
		/*
		 * A group stop is in progress and this is the group leader.
		 * We won't report until all threads have stopped.
		 */
1456 1457 1458 1459 1460 1461
		goto unlock_sig;

	exit_code = p->exit_code;
	if (!exit_code)
		goto unlock_sig;

R
Roland McGrath 已提交
1462
	if (!unlikely(options & WNOWAIT))
1463 1464 1465 1466 1467 1468
		p->exit_code = 0;

	uid = p->uid;
unlock_sig:
	spin_unlock_irq(&p->sighand->siglock);
	if (!exit_code)
L
Linus Torvalds 已提交
1469 1470 1471 1472 1473 1474 1475 1476 1477 1478
		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);
1479
	pid = task_pid_vnr(p);
R
Roland McGrath 已提交
1480
	why = ptrace ? CLD_TRAPPED : CLD_STOPPED;
L
Linus Torvalds 已提交
1481 1482
	read_unlock(&tasklist_lock);

R
Roland McGrath 已提交
1483
	if (unlikely(options & WNOWAIT))
L
Linus Torvalds 已提交
1484
		return wait_noreap_copyout(p, pid, uid,
1485
					   why, exit_code,
L
Linus Torvalds 已提交
1486 1487 1488 1489 1490 1491 1492 1493 1494 1495
					   infop, ru);

	retval = ru ? getrusage(p, RUSAGE_BOTH, ru) : 0;
	if (!retval && stat_addr)
		retval = put_user((exit_code << 8) | 0x7f, 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)
R
Roland McGrath 已提交
1496
		retval = put_user((short)why, &infop->si_code);
L
Linus Torvalds 已提交
1497 1498 1499
	if (!retval && infop)
		retval = put_user(exit_code, &infop->si_status);
	if (!retval && infop)
1500
		retval = put_user(pid, &infop->si_pid);
L
Linus Torvalds 已提交
1501
	if (!retval && infop)
1502
		retval = put_user(uid, &infop->si_uid);
L
Linus Torvalds 已提交
1503
	if (!retval)
1504
		retval = pid;
L
Linus Torvalds 已提交
1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516
	put_task_struct(p);

	BUG_ON(!retval);
	return retval;
}

/*
 * Handle do_wait work for one task in a live, non-stopped state.
 * 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.
 */
R
Roland McGrath 已提交
1517
static int wait_task_continued(struct task_struct *p, int options,
L
Linus Torvalds 已提交
1518 1519 1520 1521 1522 1523 1524
			       struct siginfo __user *infop,
			       int __user *stat_addr, struct rusage __user *ru)
{
	int retval;
	pid_t pid;
	uid_t uid;

R
Roland McGrath 已提交
1525 1526 1527
	if (!unlikely(options & WCONTINUED))
		return 0;

L
Linus Torvalds 已提交
1528 1529 1530 1531 1532 1533 1534 1535 1536
	if (!(p->signal->flags & SIGNAL_STOP_CONTINUED))
		return 0;

	spin_lock_irq(&p->sighand->siglock);
	/* Re-check with the lock held.  */
	if (!(p->signal->flags & SIGNAL_STOP_CONTINUED)) {
		spin_unlock_irq(&p->sighand->siglock);
		return 0;
	}
R
Roland McGrath 已提交
1537
	if (!unlikely(options & WNOWAIT))
L
Linus Torvalds 已提交
1538 1539 1540
		p->signal->flags &= ~SIGNAL_STOP_CONTINUED;
	spin_unlock_irq(&p->sighand->siglock);

1541
	pid = task_pid_vnr(p);
L
Linus Torvalds 已提交
1542 1543 1544 1545 1546 1547 1548 1549 1550 1551
	uid = p->uid;
	get_task_struct(p);
	read_unlock(&tasklist_lock);

	if (!infop) {
		retval = ru ? getrusage(p, RUSAGE_BOTH, ru) : 0;
		put_task_struct(p);
		if (!retval && stat_addr)
			retval = put_user(0xffff, stat_addr);
		if (!retval)
1552
			retval = pid;
L
Linus Torvalds 已提交
1553 1554 1555 1556 1557 1558 1559 1560 1561 1562
	} else {
		retval = wait_noreap_copyout(p, pid, uid,
					     CLD_CONTINUED, SIGCONT,
					     infop, ru);
		BUG_ON(retval == 0);
	}

	return retval;
}

R
Roland McGrath 已提交
1563 1564 1565 1566 1567 1568
/*
 * Consider @p for a wait by @parent.
 *
 * -ECHILD should be in *@notask_error before the first call.
 * Returns nonzero for a final return, when we have unlocked tasklist_lock.
 * Returns zero if the search for a child should continue;
1569 1570
 * then *@notask_error is 0 if @p is an eligible child,
 * or another error from security_task_wait(), or still -ECHILD.
R
Roland McGrath 已提交
1571
 */
R
Roland McGrath 已提交
1572
static int wait_consider_task(struct task_struct *parent, int ptrace,
R
Roland McGrath 已提交
1573 1574 1575 1576 1577 1578
			      struct task_struct *p, int *notask_error,
			      enum pid_type type, struct pid *pid, int options,
			      struct siginfo __user *infop,
			      int __user *stat_addr, struct rusage __user *ru)
{
	int ret = eligible_child(type, pid, options, p);
1579
	if (!ret)
R
Roland McGrath 已提交
1580 1581
		return ret;

1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593
	if (unlikely(ret < 0)) {
		/*
		 * If we have not yet seen any eligible child,
		 * then let this error code replace -ECHILD.
		 * A permission error will give the user a clue
		 * to look for security policy problems, rather
		 * than for mysterious wait bugs.
		 */
		if (*notask_error)
			*notask_error = ret;
	}

R
Roland McGrath 已提交
1594 1595 1596 1597 1598 1599 1600 1601 1602
	if (likely(!ptrace) && unlikely(p->ptrace)) {
		/*
		 * This child is hidden by ptrace.
		 * We aren't allowed to see it now, but eventually we will.
		 */
		*notask_error = 0;
		return 0;
	}

R
Roland McGrath 已提交
1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618
	if (p->exit_state == EXIT_DEAD)
		return 0;

	/*
	 * We don't reap group leaders with subthreads.
	 */
	if (p->exit_state == EXIT_ZOMBIE && !delay_group_leader(p))
		return wait_task_zombie(p, options, infop, stat_addr, ru);

	/*
	 * It's stopped or running now, so it might
	 * later continue, exit, or stop again.
	 */
	*notask_error = 0;

	if (task_is_stopped_or_traced(p))
R
Roland McGrath 已提交
1619 1620
		return wait_task_stopped(ptrace, p, options,
					 infop, stat_addr, ru);
R
Roland McGrath 已提交
1621 1622 1623 1624 1625 1626 1627 1628 1629 1630

	return wait_task_continued(p, options, infop, stat_addr, ru);
}

/*
 * Do the work of do_wait() for one thread in the group, @tsk.
 *
 * -ECHILD should be in *@notask_error before the first call.
 * Returns nonzero for a final return, when we have unlocked tasklist_lock.
 * Returns zero if the search for a child should continue; then
1631 1632
 * *@notask_error is 0 if there were any eligible children,
 * or another error from security_task_wait(), or still -ECHILD.
R
Roland McGrath 已提交
1633 1634 1635 1636 1637 1638 1639 1640 1641
 */
static int do_wait_thread(struct task_struct *tsk, int *notask_error,
			  enum pid_type type, struct pid *pid, int options,
			  struct siginfo __user *infop, int __user *stat_addr,
			  struct rusage __user *ru)
{
	struct task_struct *p;

	list_for_each_entry(p, &tsk->children, sibling) {
R
Roland McGrath 已提交
1642 1643 1644 1645 1646 1647 1648 1649 1650 1651
		/*
		 * Do not consider detached threads.
		 */
		if (!task_detached(p)) {
			int ret = wait_consider_task(tsk, 0, p, notask_error,
						     type, pid, options,
						     infop, stat_addr, ru);
			if (ret)
				return ret;
		}
R
Roland McGrath 已提交
1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664
	}

	return 0;
}

static int ptrace_do_wait(struct task_struct *tsk, int *notask_error,
			  enum pid_type type, struct pid *pid, int options,
			  struct siginfo __user *infop, int __user *stat_addr,
			  struct rusage __user *ru)
{
	struct task_struct *p;

	/*
R
Roland McGrath 已提交
1665
	 * Traditionally we see ptrace'd stopped tasks regardless of options.
R
Roland McGrath 已提交
1666
	 */
R
Roland McGrath 已提交
1667
	options |= WUNTRACED;
R
Roland McGrath 已提交
1668

R
Roland McGrath 已提交
1669 1670 1671 1672 1673
	list_for_each_entry(p, &tsk->ptraced, ptrace_entry) {
		int ret = wait_consider_task(tsk, 1, p, notask_error,
					     type, pid, options,
					     infop, stat_addr, ru);
		if (ret)
R
Roland McGrath 已提交
1674 1675 1676 1677 1678 1679
			return ret;
	}

	return 0;
}

E
Eric W. Biederman 已提交
1680 1681 1682
static long do_wait(enum pid_type type, struct pid *pid, int options,
		    struct siginfo __user *infop, int __user *stat_addr,
		    struct rusage __user *ru)
L
Linus Torvalds 已提交
1683 1684 1685
{
	DECLARE_WAITQUEUE(wait, current);
	struct task_struct *tsk;
R
Roland McGrath 已提交
1686
	int retval;
L
Linus Torvalds 已提交
1687

1688 1689
	trace_sched_process_wait(pid);

L
Linus Torvalds 已提交
1690 1691
	add_wait_queue(&current->signal->wait_chldexit,&wait);
repeat:
R
Roland McGrath 已提交
1692 1693 1694 1695 1696
	/*
	 * If there is nothing that can match our critiera just get out.
	 * We will clear @retval to zero if we see any child that might later
	 * match our criteria, even if we are not able to reap it yet.
	 */
E
Eric W. Biederman 已提交
1697 1698 1699 1700
	retval = -ECHILD;
	if ((type < PIDTYPE_MAX) && (!pid || hlist_empty(&pid->tasks[type])))
		goto end;

L
Linus Torvalds 已提交
1701 1702 1703 1704
	current->state = TASK_INTERRUPTIBLE;
	read_lock(&tasklist_lock);
	tsk = current;
	do {
R
Roland McGrath 已提交
1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717
		int tsk_result = do_wait_thread(tsk, &retval,
						type, pid, options,
						infop, stat_addr, ru);
		if (!tsk_result)
			tsk_result = ptrace_do_wait(tsk, &retval,
						    type, pid, options,
						    infop, stat_addr, ru);
		if (tsk_result) {
			/*
			 * tasklist_lock is unlocked and we have a final result.
			 */
			retval = tsk_result;
			goto end;
L
Linus Torvalds 已提交
1718
		}
R
Roland McGrath 已提交
1719

L
Linus Torvalds 已提交
1720 1721 1722
		if (options & __WNOTHREAD)
			break;
		tsk = next_thread(tsk);
E
Eric Sesterhenn 已提交
1723
		BUG_ON(tsk->signal != current->signal);
L
Linus Torvalds 已提交
1724 1725
	} while (tsk != current);
	read_unlock(&tasklist_lock);
O
Oleg Nesterov 已提交
1726

R
Roland McGrath 已提交
1727
	if (!retval && !(options & WNOHANG)) {
L
Linus Torvalds 已提交
1728
		retval = -ERESTARTSYS;
R
Roland McGrath 已提交
1729 1730 1731 1732
		if (!signal_pending(current)) {
			schedule();
			goto repeat;
		}
L
Linus Torvalds 已提交
1733
	}
R
Roland McGrath 已提交
1734

L
Linus Torvalds 已提交
1735 1736 1737 1738 1739
end:
	current->state = TASK_RUNNING;
	remove_wait_queue(&current->signal->wait_chldexit,&wait);
	if (infop) {
		if (retval > 0)
1740
			retval = 0;
L
Linus Torvalds 已提交
1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763
		else {
			/*
			 * For a WNOHANG return, clear out all the fields
			 * we would set so the user can easily tell the
			 * difference.
			 */
			if (!retval)
				retval = put_user(0, &infop->si_signo);
			if (!retval)
				retval = put_user(0, &infop->si_errno);
			if (!retval)
				retval = put_user(0, &infop->si_code);
			if (!retval)
				retval = put_user(0, &infop->si_pid);
			if (!retval)
				retval = put_user(0, &infop->si_uid);
			if (!retval)
				retval = put_user(0, &infop->si_status);
		}
	}
	return retval;
}

E
Eric W. Biederman 已提交
1764
asmlinkage long sys_waitid(int which, pid_t upid,
L
Linus Torvalds 已提交
1765 1766 1767
			   struct siginfo __user *infop, int options,
			   struct rusage __user *ru)
{
E
Eric W. Biederman 已提交
1768 1769
	struct pid *pid = NULL;
	enum pid_type type;
L
Linus Torvalds 已提交
1770 1771 1772 1773 1774 1775 1776 1777 1778
	long ret;

	if (options & ~(WNOHANG|WNOWAIT|WEXITED|WSTOPPED|WCONTINUED))
		return -EINVAL;
	if (!(options & (WEXITED|WSTOPPED|WCONTINUED)))
		return -EINVAL;

	switch (which) {
	case P_ALL:
E
Eric W. Biederman 已提交
1779
		type = PIDTYPE_MAX;
L
Linus Torvalds 已提交
1780 1781
		break;
	case P_PID:
E
Eric W. Biederman 已提交
1782 1783
		type = PIDTYPE_PID;
		if (upid <= 0)
L
Linus Torvalds 已提交
1784 1785 1786
			return -EINVAL;
		break;
	case P_PGID:
E
Eric W. Biederman 已提交
1787 1788
		type = PIDTYPE_PGID;
		if (upid <= 0)
L
Linus Torvalds 已提交
1789 1790 1791 1792 1793 1794
			return -EINVAL;
		break;
	default:
		return -EINVAL;
	}

E
Eric W. Biederman 已提交
1795 1796 1797 1798
	if (type < PIDTYPE_MAX)
		pid = find_get_pid(upid);
	ret = do_wait(type, pid, options, infop, NULL, ru);
	put_pid(pid);
L
Linus Torvalds 已提交
1799 1800

	/* avoid REGPARM breakage on x86: */
1801
	asmlinkage_protect(5, ret, which, upid, infop, options, ru);
L
Linus Torvalds 已提交
1802 1803 1804
	return ret;
}

E
Eric W. Biederman 已提交
1805
asmlinkage long sys_wait4(pid_t upid, int __user *stat_addr,
L
Linus Torvalds 已提交
1806 1807
			  int options, struct rusage __user *ru)
{
E
Eric W. Biederman 已提交
1808 1809
	struct pid *pid = NULL;
	enum pid_type type;
L
Linus Torvalds 已提交
1810 1811 1812 1813 1814
	long ret;

	if (options & ~(WNOHANG|WUNTRACED|WCONTINUED|
			__WNOTHREAD|__WCLONE|__WALL))
		return -EINVAL;
E
Eric W. Biederman 已提交
1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830

	if (upid == -1)
		type = PIDTYPE_MAX;
	else if (upid < 0) {
		type = PIDTYPE_PGID;
		pid = find_get_pid(-upid);
	} else if (upid == 0) {
		type = PIDTYPE_PGID;
		pid = get_pid(task_pgrp(current));
	} else /* upid > 0 */ {
		type = PIDTYPE_PID;
		pid = find_get_pid(upid);
	}

	ret = do_wait(type, pid, options | WEXITED, NULL, stat_addr, ru);
	put_pid(pid);
L
Linus Torvalds 已提交
1831 1832

	/* avoid REGPARM breakage on x86: */
1833
	asmlinkage_protect(4, ret, upid, stat_addr, options, ru);
L
Linus Torvalds 已提交
1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848
	return ret;
}

#ifdef __ARCH_WANT_SYS_WAITPID

/*
 * sys_waitpid() remains for compatibility. waitpid() should be
 * implemented by calling sys_wait4() from libc.a.
 */
asmlinkage long sys_waitpid(pid_t pid, int __user *stat_addr, int options)
{
	return sys_wait4(pid, stat_addr, options, NULL);
}

#endif