sys.c 42.1 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
/*
 *  linux/kernel/sys.c
 *
 *  Copyright (C) 1991, 1992  Linus Torvalds
 */

#include <linux/module.h>
#include <linux/mm.h>
#include <linux/utsname.h>
#include <linux/mman.h>
#include <linux/smp_lock.h>
#include <linux/notifier.h>
#include <linux/reboot.h>
#include <linux/prctl.h>
#include <linux/highuid.h>
#include <linux/fs.h>
17
#include <linux/resource.h>
18 19
#include <linux/kernel.h>
#include <linux/kexec.h>
L
Linus Torvalds 已提交
20
#include <linux/workqueue.h>
21
#include <linux/capability.h>
L
Linus Torvalds 已提交
22 23 24 25 26 27 28 29
#include <linux/device.h>
#include <linux/key.h>
#include <linux/times.h>
#include <linux/posix-timers.h>
#include <linux/security.h>
#include <linux/dcookies.h>
#include <linux/suspend.h>
#include <linux/tty.h>
30
#include <linux/signal.h>
M
Matt Helsley 已提交
31
#include <linux/cn_proc.h>
32
#include <linux/getcpu.h>
33
#include <linux/task_io_accounting_ops.h>
34
#include <linux/seccomp.h>
M
Mark Lord 已提交
35
#include <linux/cpu.h>
36
#include <linux/ptrace.h>
L
Linus Torvalds 已提交
37 38 39

#include <linux/compat.h>
#include <linux/syscalls.h>
40
#include <linux/kprobes.h>
41
#include <linux/user_namespace.h>
L
Linus Torvalds 已提交
42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64

#include <asm/uaccess.h>
#include <asm/io.h>
#include <asm/unistd.h>

#ifndef SET_UNALIGN_CTL
# define SET_UNALIGN_CTL(a,b)	(-EINVAL)
#endif
#ifndef GET_UNALIGN_CTL
# define GET_UNALIGN_CTL(a,b)	(-EINVAL)
#endif
#ifndef SET_FPEMU_CTL
# define SET_FPEMU_CTL(a,b)	(-EINVAL)
#endif
#ifndef GET_FPEMU_CTL
# define GET_FPEMU_CTL(a,b)	(-EINVAL)
#endif
#ifndef SET_FPEXC_CTL
# define SET_FPEXC_CTL(a,b)	(-EINVAL)
#endif
#ifndef GET_FPEXC_CTL
# define GET_FPEXC_CTL(a,b)	(-EINVAL)
#endif
65 66 67 68 69 70
#ifndef GET_ENDIAN
# define GET_ENDIAN(a,b)	(-EINVAL)
#endif
#ifndef SET_ENDIAN
# define SET_ENDIAN(a,b)	(-EINVAL)
#endif
71 72 73 74 75 76
#ifndef GET_TSC_CTL
# define GET_TSC_CTL(a)		(-EINVAL)
#endif
#ifndef SET_TSC_CTL
# define SET_TSC_CTL(a)		(-EINVAL)
#endif
L
Linus Torvalds 已提交
77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106

/*
 * this is where the system-wide overflow UID and GID are defined, for
 * architectures that now have 32-bit UID/GID but didn't in the past
 */

int overflowuid = DEFAULT_OVERFLOWUID;
int overflowgid = DEFAULT_OVERFLOWGID;

#ifdef CONFIG_UID16
EXPORT_SYMBOL(overflowuid);
EXPORT_SYMBOL(overflowgid);
#endif

/*
 * the same as above, but for filesystems which can only store a 16-bit
 * UID and GID. as such, this is needed on all architectures
 */

int fs_overflowuid = DEFAULT_FS_OVERFLOWUID;
int fs_overflowgid = DEFAULT_FS_OVERFLOWUID;

EXPORT_SYMBOL(fs_overflowuid);
EXPORT_SYMBOL(fs_overflowgid);

/*
 * this indicates whether you can reboot with ctrl-alt-del: the default is yes
 */

int C_A_D = 1;
107 108
struct pid *cad_pid;
EXPORT_SYMBOL(cad_pid);
L
Linus Torvalds 已提交
109

110 111 112 113 114 115
/*
 * If set, this is used for preparing the system to power off.
 */

void (*pm_power_off_prepare)(void);

116 117 118 119
/*
 * set the priority of a task
 * - the caller must hold the RCU read lock
 */
L
Linus Torvalds 已提交
120 121
static int set_one_prio(struct task_struct *p, int niceval, int error)
{
122
	const struct cred *cred = current_cred(), *pcred = __task_cred(p);
L
Linus Torvalds 已提交
123 124
	int no_nice;

125 126
	if (pcred->uid  != cred->euid &&
	    pcred->euid != cred->euid && !capable(CAP_SYS_NICE)) {
L
Linus Torvalds 已提交
127 128 129
		error = -EPERM;
		goto out;
	}
M
Matt Mackall 已提交
130
	if (niceval < task_nice(p) && !can_nice(p, niceval)) {
L
Linus Torvalds 已提交
131 132 133 134 135 136 137 138 139 140 141 142 143 144 145
		error = -EACCES;
		goto out;
	}
	no_nice = security_task_setnice(p, niceval);
	if (no_nice) {
		error = no_nice;
		goto out;
	}
	if (error == -ESRCH)
		error = 0;
	set_user_nice(p, niceval);
out:
	return error;
}

146
SYSCALL_DEFINE3(setpriority, int, which, int, who, int, niceval)
L
Linus Torvalds 已提交
147 148 149
{
	struct task_struct *g, *p;
	struct user_struct *user;
150
	const struct cred *cred = current_cred();
L
Linus Torvalds 已提交
151
	int error = -EINVAL;
152
	struct pid *pgrp;
L
Linus Torvalds 已提交
153

154
	if (which > PRIO_USER || which < PRIO_PROCESS)
L
Linus Torvalds 已提交
155 156 157 158 159 160 161 162 163 164 165 166
		goto out;

	/* normalize: avoid signed division (rounding problems) */
	error = -ESRCH;
	if (niceval < -20)
		niceval = -20;
	if (niceval > 19)
		niceval = 19;

	read_lock(&tasklist_lock);
	switch (which) {
		case PRIO_PROCESS:
167
			if (who)
168
				p = find_task_by_vpid(who);
169 170
			else
				p = current;
L
Linus Torvalds 已提交
171 172 173 174
			if (p)
				error = set_one_prio(p, niceval, error);
			break;
		case PRIO_PGRP:
175
			if (who)
176
				pgrp = find_vpid(who);
177 178
			else
				pgrp = task_pgrp(current);
179
			do_each_pid_thread(pgrp, PIDTYPE_PGID, p) {
L
Linus Torvalds 已提交
180
				error = set_one_prio(p, niceval, error);
181
			} while_each_pid_thread(pgrp, PIDTYPE_PGID, p);
L
Linus Torvalds 已提交
182 183
			break;
		case PRIO_USER:
D
David Howells 已提交
184
			user = (struct user_struct *) cred->user;
L
Linus Torvalds 已提交
185
			if (!who)
186 187 188 189
				who = cred->uid;
			else if ((who != cred->uid) &&
				 !(user = find_user(who)))
				goto out_unlock;	/* No processes for this user */
L
Linus Torvalds 已提交
190 191

			do_each_thread(g, p)
192
				if (__task_cred(p)->uid == who)
L
Linus Torvalds 已提交
193 194
					error = set_one_prio(p, niceval, error);
			while_each_thread(g, p);
195
			if (who != cred->uid)
L
Linus Torvalds 已提交
196 197 198 199 200 201 202 203 204 205 206 207 208 209 210
				free_uid(user);		/* For find_user() */
			break;
	}
out_unlock:
	read_unlock(&tasklist_lock);
out:
	return error;
}

/*
 * Ugh. To avoid negative return values, "getpriority()" will
 * not return the normal nice-value, but a negated value that
 * has been offset by 20 (ie it returns 40..1 instead of -20..19)
 * to stay compatible.
 */
211
SYSCALL_DEFINE2(getpriority, int, which, int, who)
L
Linus Torvalds 已提交
212 213 214
{
	struct task_struct *g, *p;
	struct user_struct *user;
215
	const struct cred *cred = current_cred();
L
Linus Torvalds 已提交
216
	long niceval, retval = -ESRCH;
217
	struct pid *pgrp;
L
Linus Torvalds 已提交
218

219
	if (which > PRIO_USER || which < PRIO_PROCESS)
L
Linus Torvalds 已提交
220 221 222 223 224
		return -EINVAL;

	read_lock(&tasklist_lock);
	switch (which) {
		case PRIO_PROCESS:
225
			if (who)
226
				p = find_task_by_vpid(who);
227 228
			else
				p = current;
L
Linus Torvalds 已提交
229 230 231 232 233 234 235
			if (p) {
				niceval = 20 - task_nice(p);
				if (niceval > retval)
					retval = niceval;
			}
			break;
		case PRIO_PGRP:
236
			if (who)
237
				pgrp = find_vpid(who);
238 239
			else
				pgrp = task_pgrp(current);
240
			do_each_pid_thread(pgrp, PIDTYPE_PGID, p) {
L
Linus Torvalds 已提交
241 242 243
				niceval = 20 - task_nice(p);
				if (niceval > retval)
					retval = niceval;
244
			} while_each_pid_thread(pgrp, PIDTYPE_PGID, p);
L
Linus Torvalds 已提交
245 246
			break;
		case PRIO_USER:
247
			user = (struct user_struct *) cred->user;
L
Linus Torvalds 已提交
248
			if (!who)
249 250 251 252
				who = cred->uid;
			else if ((who != cred->uid) &&
				 !(user = find_user(who)))
				goto out_unlock;	/* No processes for this user */
L
Linus Torvalds 已提交
253 254

			do_each_thread(g, p)
255
				if (__task_cred(p)->uid == who) {
L
Linus Torvalds 已提交
256 257 258 259 260
					niceval = 20 - task_nice(p);
					if (niceval > retval)
						retval = niceval;
				}
			while_each_thread(g, p);
261
			if (who != cred->uid)
L
Linus Torvalds 已提交
262 263 264 265 266 267 268 269 270
				free_uid(user);		/* for find_user() */
			break;
	}
out_unlock:
	read_unlock(&tasklist_lock);

	return retval;
}

271 272 273 274 275 276 277 278
/**
 *	emergency_restart - reboot the system
 *
 *	Without shutting down any hardware or taking any locks
 *	reboot the system.  This is called when we know we are in
 *	trouble so this is our best effort to reboot.  This is
 *	safe to call in interrupt context.
 */
279 280 281 282 283 284
void emergency_restart(void)
{
	machine_emergency_restart();
}
EXPORT_SYMBOL_GPL(emergency_restart);

285
void kernel_restart_prepare(char *cmd)
286
{
287
	blocking_notifier_call_chain(&reboot_notifier_list, SYS_RESTART, cmd);
288 289
	system_state = SYSTEM_RESTART;
	device_shutdown();
290
	sysdev_shutdown();
291
}
292 293 294 295

/**
 *	kernel_restart - reboot the system
 *	@cmd: pointer to buffer containing command to execute for restart
296
 *		or %NULL
297 298 299 300
 *
 *	Shutdown everything and perform a clean reboot.
 *	This is not safe to call in interrupt context.
 */
301 302 303
void kernel_restart(char *cmd)
{
	kernel_restart_prepare(cmd);
304
	if (!cmd)
305
		printk(KERN_EMERG "Restarting system.\n");
306
	else
307 308 309 310 311
		printk(KERN_EMERG "Restarting system with command '%s'.\n", cmd);
	machine_restart(cmd);
}
EXPORT_SYMBOL_GPL(kernel_restart);

312
static void kernel_shutdown_prepare(enum system_states state)
313
{
314
	blocking_notifier_call_chain(&reboot_notifier_list,
315 316 317 318
		(state == SYSTEM_HALT)?SYS_HALT:SYS_POWER_OFF, NULL);
	system_state = state;
	device_shutdown();
}
319 320 321 322 323 324 325
/**
 *	kernel_halt - halt the system
 *
 *	Shutdown everything and perform a clean system halt.
 */
void kernel_halt(void)
{
326
	kernel_shutdown_prepare(SYSTEM_HALT);
327
	sysdev_shutdown();
328 329 330
	printk(KERN_EMERG "System halted.\n");
	machine_halt();
}
331

332 333
EXPORT_SYMBOL_GPL(kernel_halt);

334 335 336 337 338 339 340
/**
 *	kernel_power_off - power_off the system
 *
 *	Shutdown everything and perform a clean system power_off.
 */
void kernel_power_off(void)
{
341
	kernel_shutdown_prepare(SYSTEM_POWER_OFF);
342 343
	if (pm_power_off_prepare)
		pm_power_off_prepare();
M
Mark Lord 已提交
344
	disable_nonboot_cpus();
345
	sysdev_shutdown();
346 347 348 349
	printk(KERN_EMERG "Power down.\n");
	machine_power_off();
}
EXPORT_SYMBOL_GPL(kernel_power_off);
L
Linus Torvalds 已提交
350 351 352 353 354 355 356 357
/*
 * Reboot system call: for obvious reasons only root may call it,
 * and even root needs to set up some magic numbers in the registers
 * so that some mistake won't make this reboot the whole machine.
 * You can also set the meaning of the ctrl-alt-del-key here.
 *
 * reboot doesn't sync: do that yourself before calling this.
 */
358 359
SYSCALL_DEFINE4(reboot, int, magic1, int, magic2, unsigned int, cmd,
		void __user *, arg)
L
Linus Torvalds 已提交
360 361 362 363 364 365 366 367 368 369 370 371 372 373 374
{
	char buffer[256];

	/* We only trust the superuser with rebooting the system. */
	if (!capable(CAP_SYS_BOOT))
		return -EPERM;

	/* For safety, we require "magic" arguments. */
	if (magic1 != LINUX_REBOOT_MAGIC1 ||
	    (magic2 != LINUX_REBOOT_MAGIC2 &&
	                magic2 != LINUX_REBOOT_MAGIC2A &&
			magic2 != LINUX_REBOOT_MAGIC2B &&
	                magic2 != LINUX_REBOOT_MAGIC2C))
		return -EINVAL;

375 376 377 378 379 380
	/* Instead of trying to make the power_off code look like
	 * halt when pm_power_off is not set do it the easy way.
	 */
	if ((cmd == LINUX_REBOOT_CMD_POWER_OFF) && !pm_power_off)
		cmd = LINUX_REBOOT_CMD_HALT;

L
Linus Torvalds 已提交
381 382 383
	lock_kernel();
	switch (cmd) {
	case LINUX_REBOOT_CMD_RESTART:
384
		kernel_restart(NULL);
L
Linus Torvalds 已提交
385 386 387 388 389 390 391 392 393 394 395
		break;

	case LINUX_REBOOT_CMD_CAD_ON:
		C_A_D = 1;
		break;

	case LINUX_REBOOT_CMD_CAD_OFF:
		C_A_D = 0;
		break;

	case LINUX_REBOOT_CMD_HALT:
396
		kernel_halt();
L
Linus Torvalds 已提交
397 398 399 400 401
		unlock_kernel();
		do_exit(0);
		break;

	case LINUX_REBOOT_CMD_POWER_OFF:
402
		kernel_power_off();
L
Linus Torvalds 已提交
403 404 405 406 407 408 409 410 411 412 413
		unlock_kernel();
		do_exit(0);
		break;

	case LINUX_REBOOT_CMD_RESTART2:
		if (strncpy_from_user(&buffer[0], arg, sizeof(buffer) - 1) < 0) {
			unlock_kernel();
			return -EFAULT;
		}
		buffer[sizeof(buffer) - 1] = '\0';

414
		kernel_restart(buffer);
L
Linus Torvalds 已提交
415 416
		break;

H
Huang Ying 已提交
417
#ifdef CONFIG_KEXEC
418
	case LINUX_REBOOT_CMD_KEXEC:
H
Huang Ying 已提交
419 420 421 422 423 424 425
		{
			int ret;
			ret = kernel_kexec();
			unlock_kernel();
			return ret;
		}
#endif
426

427
#ifdef CONFIG_HIBERNATION
L
Linus Torvalds 已提交
428 429
	case LINUX_REBOOT_CMD_SW_SUSPEND:
		{
430
			int ret = hibernate();
L
Linus Torvalds 已提交
431 432 433 434 435 436 437 438 439 440 441 442 443
			unlock_kernel();
			return ret;
		}
#endif

	default:
		unlock_kernel();
		return -EINVAL;
	}
	unlock_kernel();
	return 0;
}

444
static void deferred_cad(struct work_struct *dummy)
L
Linus Torvalds 已提交
445
{
446
	kernel_restart(NULL);
L
Linus Torvalds 已提交
447 448 449 450 451 452 453 454 455
}

/*
 * This function gets called by ctrl-alt-del - ie the keyboard interrupt.
 * As it's called within an interrupt, it may NOT sync: the only choice
 * is whether to reboot at once, or just ignore the ctrl-alt-del.
 */
void ctrl_alt_del(void)
{
456
	static DECLARE_WORK(cad_work, deferred_cad);
L
Linus Torvalds 已提交
457 458 459 460

	if (C_A_D)
		schedule_work(&cad_work);
	else
461
		kill_cad_pid(SIGINT, 1);
L
Linus Torvalds 已提交
462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481
}
	
/*
 * Unprivileged users may change the real gid to the effective gid
 * or vice versa.  (BSD-style)
 *
 * If you set the real gid at all, or set the effective gid to a value not
 * equal to the real gid, then the saved gid is set to the new effective gid.
 *
 * This makes it possible for a setgid program to completely drop its
 * privileges, which is often a useful assertion to make when you are doing
 * a security audit over a program.
 *
 * The general idea is that a program which uses just setregid() will be
 * 100% compatible with BSD.  A program which uses just setgid() will be
 * 100% compatible with POSIX with saved IDs. 
 *
 * SMP: There are not races, the GIDs are checked only by filesystem
 *      operations (as far as semantic preservation is concerned).
 */
482
SYSCALL_DEFINE2(setregid, gid_t, rgid, gid_t, egid)
L
Linus Torvalds 已提交
483
{
D
David Howells 已提交
484 485
	const struct cred *old;
	struct cred *new;
L
Linus Torvalds 已提交
486 487
	int retval;

D
David Howells 已提交
488 489 490 491 492
	new = prepare_creds();
	if (!new)
		return -ENOMEM;
	old = current_cred();

L
Linus Torvalds 已提交
493 494
	retval = security_task_setgid(rgid, egid, (gid_t)-1, LSM_SETID_RE);
	if (retval)
D
David Howells 已提交
495
		goto error;
L
Linus Torvalds 已提交
496

D
David Howells 已提交
497
	retval = -EPERM;
L
Linus Torvalds 已提交
498
	if (rgid != (gid_t) -1) {
D
David Howells 已提交
499 500
		if (old->gid == rgid ||
		    old->egid == rgid ||
L
Linus Torvalds 已提交
501
		    capable(CAP_SETGID))
D
David Howells 已提交
502
			new->gid = rgid;
L
Linus Torvalds 已提交
503
		else
D
David Howells 已提交
504
			goto error;
L
Linus Torvalds 已提交
505 506
	}
	if (egid != (gid_t) -1) {
D
David Howells 已提交
507 508 509
		if (old->gid == egid ||
		    old->egid == egid ||
		    old->sgid == egid ||
L
Linus Torvalds 已提交
510
		    capable(CAP_SETGID))
D
David Howells 已提交
511
			new->egid = egid;
512
		else
D
David Howells 已提交
513
			goto error;
L
Linus Torvalds 已提交
514
	}
D
David Howells 已提交
515

L
Linus Torvalds 已提交
516
	if (rgid != (gid_t) -1 ||
D
David Howells 已提交
517 518 519 520 521 522 523 524 525
	    (egid != (gid_t) -1 && egid != old->gid))
		new->sgid = new->egid;
	new->fsgid = new->egid;

	return commit_creds(new);

error:
	abort_creds(new);
	return retval;
L
Linus Torvalds 已提交
526 527 528 529 530 531 532
}

/*
 * setgid() is implemented like SysV w/ SAVED_IDS 
 *
 * SMP: Same implicit races as above.
 */
533
SYSCALL_DEFINE1(setgid, gid_t, gid)
L
Linus Torvalds 已提交
534
{
D
David Howells 已提交
535 536
	const struct cred *old;
	struct cred *new;
L
Linus Torvalds 已提交
537 538
	int retval;

D
David Howells 已提交
539 540 541 542 543
	new = prepare_creds();
	if (!new)
		return -ENOMEM;
	old = current_cred();

L
Linus Torvalds 已提交
544 545
	retval = security_task_setgid(gid, (gid_t)-1, (gid_t)-1, LSM_SETID_ID);
	if (retval)
D
David Howells 已提交
546
		goto error;
L
Linus Torvalds 已提交
547

D
David Howells 已提交
548 549 550 551 552
	retval = -EPERM;
	if (capable(CAP_SETGID))
		new->gid = new->egid = new->sgid = new->fsgid = gid;
	else if (gid == old->gid || gid == old->sgid)
		new->egid = new->fsgid = gid;
L
Linus Torvalds 已提交
553
	else
D
David Howells 已提交
554
		goto error;
L
Linus Torvalds 已提交
555

D
David Howells 已提交
556 557 558 559 560
	return commit_creds(new);

error:
	abort_creds(new);
	return retval;
L
Linus Torvalds 已提交
561 562
}
  
D
David Howells 已提交
563 564 565 566
/*
 * change the user struct in a credentials set to match the new UID
 */
static int set_user(struct cred *new)
L
Linus Torvalds 已提交
567 568 569
{
	struct user_struct *new_user;

570
	new_user = alloc_uid(current_user_ns(), new->uid);
L
Linus Torvalds 已提交
571 572 573 574 575
	if (!new_user)
		return -EAGAIN;

	if (atomic_read(&new_user->processes) >=
				current->signal->rlim[RLIMIT_NPROC].rlim_cur &&
576
			new_user != INIT_USER) {
L
Linus Torvalds 已提交
577 578 579 580
		free_uid(new_user);
		return -EAGAIN;
	}

D
David Howells 已提交
581 582
	free_uid(new->user);
	new->user = new_user;
L
Linus Torvalds 已提交
583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600
	return 0;
}

/*
 * Unprivileged users may change the real uid to the effective uid
 * or vice versa.  (BSD-style)
 *
 * If you set the real uid at all, or set the effective uid to a value not
 * equal to the real uid, then the saved uid is set to the new effective uid.
 *
 * This makes it possible for a setuid program to completely drop its
 * privileges, which is often a useful assertion to make when you are doing
 * a security audit over a program.
 *
 * The general idea is that a program which uses just setreuid() will be
 * 100% compatible with BSD.  A program which uses just setuid() will be
 * 100% compatible with POSIX with saved IDs. 
 */
601
SYSCALL_DEFINE2(setreuid, uid_t, ruid, uid_t, euid)
L
Linus Torvalds 已提交
602
{
D
David Howells 已提交
603 604
	const struct cred *old;
	struct cred *new;
L
Linus Torvalds 已提交
605 606
	int retval;

D
David Howells 已提交
607 608 609 610 611
	new = prepare_creds();
	if (!new)
		return -ENOMEM;
	old = current_cred();

L
Linus Torvalds 已提交
612 613
	retval = security_task_setuid(ruid, euid, (uid_t)-1, LSM_SETID_RE);
	if (retval)
D
David Howells 已提交
614
		goto error;
L
Linus Torvalds 已提交
615

D
David Howells 已提交
616
	retval = -EPERM;
L
Linus Torvalds 已提交
617
	if (ruid != (uid_t) -1) {
D
David Howells 已提交
618 619 620
		new->uid = ruid;
		if (old->uid != ruid &&
		    old->euid != ruid &&
L
Linus Torvalds 已提交
621
		    !capable(CAP_SETUID))
D
David Howells 已提交
622
			goto error;
L
Linus Torvalds 已提交
623 624 625
	}

	if (euid != (uid_t) -1) {
D
David Howells 已提交
626 627 628 629
		new->euid = euid;
		if (old->uid != euid &&
		    old->euid != euid &&
		    old->suid != euid &&
L
Linus Torvalds 已提交
630
		    !capable(CAP_SETUID))
D
David Howells 已提交
631
			goto error;
L
Linus Torvalds 已提交
632 633
	}

D
David Howells 已提交
634 635 636
	retval = -EAGAIN;
	if (new->uid != old->uid && set_user(new) < 0)
		goto error;
L
Linus Torvalds 已提交
637 638

	if (ruid != (uid_t) -1 ||
D
David Howells 已提交
639 640 641
	    (euid != (uid_t) -1 && euid != old->uid))
		new->suid = new->euid;
	new->fsuid = new->euid;
L
Linus Torvalds 已提交
642

D
David Howells 已提交
643 644 645
	retval = security_task_fix_setuid(new, old, LSM_SETID_RE);
	if (retval < 0)
		goto error;
L
Linus Torvalds 已提交
646

D
David Howells 已提交
647
	return commit_creds(new);
L
Linus Torvalds 已提交
648

D
David Howells 已提交
649 650 651 652
error:
	abort_creds(new);
	return retval;
}
L
Linus Torvalds 已提交
653 654 655 656 657 658 659 660 661 662 663 664
		
/*
 * setuid() is implemented like SysV with SAVED_IDS 
 * 
 * Note that SAVED_ID's is deficient in that a setuid root program
 * like sendmail, for example, cannot set its uid to be a normal 
 * user and then switch back, because if you're root, setuid() sets
 * the saved uid too.  If you don't like this, blame the bright people
 * in the POSIX committee and/or USG.  Note that the BSD-style setreuid()
 * will allow a root program to temporarily drop privileges and be able to
 * regain them by swapping the real and effective uid.  
 */
665
SYSCALL_DEFINE1(setuid, uid_t, uid)
L
Linus Torvalds 已提交
666
{
D
David Howells 已提交
667 668
	const struct cred *old;
	struct cred *new;
L
Linus Torvalds 已提交
669 670
	int retval;

D
David Howells 已提交
671 672 673 674 675
	new = prepare_creds();
	if (!new)
		return -ENOMEM;
	old = current_cred();

L
Linus Torvalds 已提交
676 677
	retval = security_task_setuid(uid, (uid_t)-1, (uid_t)-1, LSM_SETID_ID);
	if (retval)
D
David Howells 已提交
678
		goto error;
L
Linus Torvalds 已提交
679

D
David Howells 已提交
680
	retval = -EPERM;
L
Linus Torvalds 已提交
681
	if (capable(CAP_SETUID)) {
D
David Howells 已提交
682 683 684 685 686 687 688
		new->suid = new->uid = uid;
		if (uid != old->uid && set_user(new) < 0) {
			retval = -EAGAIN;
			goto error;
		}
	} else if (uid != old->uid && uid != new->suid) {
		goto error;
L
Linus Torvalds 已提交
689 690
	}

D
David Howells 已提交
691 692 693 694 695 696 697
	new->fsuid = new->euid = uid;

	retval = security_task_fix_setuid(new, old, LSM_SETID_ID);
	if (retval < 0)
		goto error;

	return commit_creds(new);
L
Linus Torvalds 已提交
698

D
David Howells 已提交
699 700 701
error:
	abort_creds(new);
	return retval;
L
Linus Torvalds 已提交
702 703 704 705 706 707 708
}


/*
 * This function implements a generic ability to update ruid, euid,
 * and suid.  This allows you to implement the 4.4 compatible seteuid().
 */
709
SYSCALL_DEFINE3(setresuid, uid_t, ruid, uid_t, euid, uid_t, suid)
L
Linus Torvalds 已提交
710
{
D
David Howells 已提交
711 712
	const struct cred *old;
	struct cred *new;
L
Linus Torvalds 已提交
713 714
	int retval;

D
David Howells 已提交
715 716 717 718
	new = prepare_creds();
	if (!new)
		return -ENOMEM;

L
Linus Torvalds 已提交
719 720
	retval = security_task_setuid(ruid, euid, suid, LSM_SETID_RES);
	if (retval)
D
David Howells 已提交
721 722
		goto error;
	old = current_cred();
L
Linus Torvalds 已提交
723

D
David Howells 已提交
724
	retval = -EPERM;
L
Linus Torvalds 已提交
725
	if (!capable(CAP_SETUID)) {
D
David Howells 已提交
726 727 728 729 730 731 732 733 734
		if (ruid != (uid_t) -1 && ruid != old->uid &&
		    ruid != old->euid  && ruid != old->suid)
			goto error;
		if (euid != (uid_t) -1 && euid != old->uid &&
		    euid != old->euid  && euid != old->suid)
			goto error;
		if (suid != (uid_t) -1 && suid != old->uid &&
		    suid != old->euid  && suid != old->suid)
			goto error;
L
Linus Torvalds 已提交
735
	}
D
David Howells 已提交
736 737

	retval = -EAGAIN;
L
Linus Torvalds 已提交
738
	if (ruid != (uid_t) -1) {
D
David Howells 已提交
739 740 741
		new->uid = ruid;
		if (ruid != old->uid && set_user(new) < 0)
			goto error;
L
Linus Torvalds 已提交
742
	}
D
David Howells 已提交
743 744
	if (euid != (uid_t) -1)
		new->euid = euid;
L
Linus Torvalds 已提交
745
	if (suid != (uid_t) -1)
D
David Howells 已提交
746 747
		new->suid = suid;
	new->fsuid = new->euid;
L
Linus Torvalds 已提交
748

D
David Howells 已提交
749 750 751
	retval = security_task_fix_setuid(new, old, LSM_SETID_RES);
	if (retval < 0)
		goto error;
L
Linus Torvalds 已提交
752

D
David Howells 已提交
753 754 755 756 757
	return commit_creds(new);

error:
	abort_creds(new);
	return retval;
L
Linus Torvalds 已提交
758 759
}

760
SYSCALL_DEFINE3(getresuid, uid_t __user *, ruid, uid_t __user *, euid, uid_t __user *, suid)
L
Linus Torvalds 已提交
761
{
762
	const struct cred *cred = current_cred();
L
Linus Torvalds 已提交
763 764
	int retval;

765 766
	if (!(retval   = put_user(cred->uid,  ruid)) &&
	    !(retval   = put_user(cred->euid, euid)))
767
		retval = put_user(cred->suid, suid);
L
Linus Torvalds 已提交
768 769 770 771 772 773 774

	return retval;
}

/*
 * Same as above, but for rgid, egid, sgid.
 */
775
SYSCALL_DEFINE3(setresgid, gid_t, rgid, gid_t, egid, gid_t, sgid)
L
Linus Torvalds 已提交
776
{
D
David Howells 已提交
777 778
	const struct cred *old;
	struct cred *new;
L
Linus Torvalds 已提交
779 780
	int retval;

D
David Howells 已提交
781 782 783 784 785
	new = prepare_creds();
	if (!new)
		return -ENOMEM;
	old = current_cred();

L
Linus Torvalds 已提交
786 787
	retval = security_task_setgid(rgid, egid, sgid, LSM_SETID_RES);
	if (retval)
D
David Howells 已提交
788
		goto error;
L
Linus Torvalds 已提交
789

D
David Howells 已提交
790
	retval = -EPERM;
L
Linus Torvalds 已提交
791
	if (!capable(CAP_SETGID)) {
D
David Howells 已提交
792 793 794 795 796 797 798 799 800
		if (rgid != (gid_t) -1 && rgid != old->gid &&
		    rgid != old->egid  && rgid != old->sgid)
			goto error;
		if (egid != (gid_t) -1 && egid != old->gid &&
		    egid != old->egid  && egid != old->sgid)
			goto error;
		if (sgid != (gid_t) -1 && sgid != old->gid &&
		    sgid != old->egid  && sgid != old->sgid)
			goto error;
L
Linus Torvalds 已提交
801
	}
D
David Howells 已提交
802

L
Linus Torvalds 已提交
803
	if (rgid != (gid_t) -1)
D
David Howells 已提交
804 805 806
		new->gid = rgid;
	if (egid != (gid_t) -1)
		new->egid = egid;
L
Linus Torvalds 已提交
807
	if (sgid != (gid_t) -1)
D
David Howells 已提交
808 809
		new->sgid = sgid;
	new->fsgid = new->egid;
L
Linus Torvalds 已提交
810

D
David Howells 已提交
811 812 813 814 815
	return commit_creds(new);

error:
	abort_creds(new);
	return retval;
L
Linus Torvalds 已提交
816 817
}

818
SYSCALL_DEFINE3(getresgid, gid_t __user *, rgid, gid_t __user *, egid, gid_t __user *, sgid)
L
Linus Torvalds 已提交
819
{
820
	const struct cred *cred = current_cred();
L
Linus Torvalds 已提交
821 822
	int retval;

823 824
	if (!(retval   = put_user(cred->gid,  rgid)) &&
	    !(retval   = put_user(cred->egid, egid)))
825
		retval = put_user(cred->sgid, sgid);
L
Linus Torvalds 已提交
826 827 828 829 830 831 832 833 834 835 836

	return retval;
}


/*
 * "setfsuid()" sets the fsuid - the uid used for filesystem checks. This
 * is used for "access()" and for the NFS daemon (letting nfsd stay at
 * whatever uid it wants to). It normally shadows "euid", except when
 * explicitly set by setfsuid() or for access..
 */
837
SYSCALL_DEFINE1(setfsuid, uid_t, uid)
L
Linus Torvalds 已提交
838
{
D
David Howells 已提交
839 840 841 842 843 844 845 846 847
	const struct cred *old;
	struct cred *new;
	uid_t old_fsuid;

	new = prepare_creds();
	if (!new)
		return current_fsuid();
	old = current_cred();
	old_fsuid = old->fsuid;
L
Linus Torvalds 已提交
848

D
David Howells 已提交
849 850
	if (security_task_setuid(uid, (uid_t)-1, (uid_t)-1, LSM_SETID_FS) < 0)
		goto error;
L
Linus Torvalds 已提交
851

D
David Howells 已提交
852 853
	if (uid == old->uid  || uid == old->euid  ||
	    uid == old->suid || uid == old->fsuid ||
854 855
	    capable(CAP_SETUID)) {
		if (uid != old_fsuid) {
D
David Howells 已提交
856 857 858
			new->fsuid = uid;
			if (security_task_fix_setuid(new, old, LSM_SETID_FS) == 0)
				goto change_okay;
L
Linus Torvalds 已提交
859 860 861
		}
	}

D
David Howells 已提交
862 863 864
error:
	abort_creds(new);
	return old_fsuid;
L
Linus Torvalds 已提交
865

D
David Howells 已提交
866 867
change_okay:
	commit_creds(new);
L
Linus Torvalds 已提交
868 869 870 871
	return old_fsuid;
}

/*
872
 * Samma på svenska..
L
Linus Torvalds 已提交
873
 */
874
SYSCALL_DEFINE1(setfsgid, gid_t, gid)
L
Linus Torvalds 已提交
875
{
D
David Howells 已提交
876 877 878 879 880 881 882 883 884
	const struct cred *old;
	struct cred *new;
	gid_t old_fsgid;

	new = prepare_creds();
	if (!new)
		return current_fsgid();
	old = current_cred();
	old_fsgid = old->fsgid;
L
Linus Torvalds 已提交
885 886

	if (security_task_setgid(gid, (gid_t)-1, (gid_t)-1, LSM_SETID_FS))
D
David Howells 已提交
887
		goto error;
L
Linus Torvalds 已提交
888

D
David Howells 已提交
889 890
	if (gid == old->gid  || gid == old->egid  ||
	    gid == old->sgid || gid == old->fsgid ||
891 892
	    capable(CAP_SETGID)) {
		if (gid != old_fsgid) {
D
David Howells 已提交
893 894
			new->fsgid = gid;
			goto change_okay;
L
Linus Torvalds 已提交
895 896
		}
	}
D
David Howells 已提交
897 898 899 900 901 902 903

error:
	abort_creds(new);
	return old_fsgid;

change_okay:
	commit_creds(new);
L
Linus Torvalds 已提交
904 905 906
	return old_fsgid;
}

907 908 909 910 911 912
void do_sys_times(struct tms *tms)
{
	struct task_cputime cputime;
	cputime_t cutime, cstime;

	thread_group_cputime(current, &cputime);
913
	spin_lock_irq(&current->sighand->siglock);
914 915 916 917 918 919 920 921 922
	cutime = current->signal->cutime;
	cstime = current->signal->cstime;
	spin_unlock_irq(&current->sighand->siglock);
	tms->tms_utime = cputime_to_clock_t(cputime.utime);
	tms->tms_stime = cputime_to_clock_t(cputime.stime);
	tms->tms_cutime = cputime_to_clock_t(cutime);
	tms->tms_cstime = cputime_to_clock_t(cstime);
}

923
SYSCALL_DEFINE1(times, struct tms __user *, tbuf)
L
Linus Torvalds 已提交
924 925 926
{
	if (tbuf) {
		struct tms tmp;
927 928

		do_sys_times(&tmp);
L
Linus Torvalds 已提交
929 930 931
		if (copy_to_user(tbuf, &tmp, sizeof(struct tms)))
			return -EFAULT;
	}
932
	force_successful_syscall_return();
L
Linus Torvalds 已提交
933 934 935 936 937 938 939 940 941 942 943 944 945 946 947
	return (long) jiffies_64_to_clock_t(get_jiffies_64());
}

/*
 * This needs some heavy checking ...
 * I just haven't the stomach for it. I also don't fully
 * understand sessions/pgrp etc. Let somebody who does explain it.
 *
 * OK, I think I have the protection semantics right.... this is really
 * only important on a multi-user system anyway, to make sure one user
 * can't send a signal to a process owned by another.  -TYT, 12/12/91
 *
 * Auch. Had to add the 'did_exec' flag to conform completely to POSIX.
 * LBT 04.03.94
 */
948
SYSCALL_DEFINE2(setpgid, pid_t, pid, pid_t, pgid)
L
Linus Torvalds 已提交
949 950
{
	struct task_struct *p;
951
	struct task_struct *group_leader = current->group_leader;
952 953
	struct pid *pgrp;
	int err;
L
Linus Torvalds 已提交
954 955

	if (!pid)
956
		pid = task_pid_vnr(group_leader);
L
Linus Torvalds 已提交
957 958 959 960 961 962 963 964 965 966 967
	if (!pgid)
		pgid = pid;
	if (pgid < 0)
		return -EINVAL;

	/* From this point forward we keep holding onto the tasklist lock
	 * so that our parent does not change from under us. -DaveM
	 */
	write_lock_irq(&tasklist_lock);

	err = -ESRCH;
968
	p = find_task_by_vpid(pid);
L
Linus Torvalds 已提交
969 970 971 972 973 974 975
	if (!p)
		goto out;

	err = -EINVAL;
	if (!thread_group_leader(p))
		goto out;

976
	if (same_thread_group(p->real_parent, group_leader)) {
L
Linus Torvalds 已提交
977
		err = -EPERM;
978
		if (task_session(p) != task_session(group_leader))
L
Linus Torvalds 已提交
979 980 981 982 983 984
			goto out;
		err = -EACCES;
		if (p->did_exec)
			goto out;
	} else {
		err = -ESRCH;
985
		if (p != group_leader)
L
Linus Torvalds 已提交
986 987 988 989 990 991 992
			goto out;
	}

	err = -EPERM;
	if (p->signal->leader)
		goto out;

993
	pgrp = task_pid(p);
L
Linus Torvalds 已提交
994
	if (pgid != pid) {
995
		struct task_struct *g;
L
Linus Torvalds 已提交
996

997 998
		pgrp = find_vpid(pgid);
		g = pid_task(pgrp, PIDTYPE_PGID);
999
		if (!g || task_session(g) != task_session(group_leader))
1000
			goto out;
L
Linus Torvalds 已提交
1001 1002 1003 1004 1005 1006
	}

	err = security_task_setpgid(p, pgid);
	if (err)
		goto out;

1007
	if (task_pgrp(p) != pgrp) {
1008
		change_pid(p, PIDTYPE_PGID, pgrp);
1009
		set_task_pgrp(p, pid_nr(pgrp));
L
Linus Torvalds 已提交
1010 1011 1012 1013 1014 1015 1016 1017 1018
	}

	err = 0;
out:
	/* All paths lead to here, thus we are safe. -DaveM */
	write_unlock_irq(&tasklist_lock);
	return err;
}

1019
SYSCALL_DEFINE1(getpgid, pid_t, pid)
L
Linus Torvalds 已提交
1020
{
1021 1022 1023 1024 1025
	struct task_struct *p;
	struct pid *grp;
	int retval;

	rcu_read_lock();
1026
	if (!pid)
1027
		grp = task_pgrp(current);
1028
	else {
L
Linus Torvalds 已提交
1029
		retval = -ESRCH;
1030 1031 1032 1033 1034 1035 1036 1037 1038 1039
		p = find_task_by_vpid(pid);
		if (!p)
			goto out;
		grp = task_pgrp(p);
		if (!grp)
			goto out;

		retval = security_task_getpgid(p);
		if (retval)
			goto out;
L
Linus Torvalds 已提交
1040
	}
1041 1042 1043 1044
	retval = pid_vnr(grp);
out:
	rcu_read_unlock();
	return retval;
L
Linus Torvalds 已提交
1045 1046 1047 1048
}

#ifdef __ARCH_WANT_SYS_GETPGRP

1049
SYSCALL_DEFINE0(getpgrp)
L
Linus Torvalds 已提交
1050
{
1051
	return sys_getpgid(0);
L
Linus Torvalds 已提交
1052 1053 1054 1055
}

#endif

1056
SYSCALL_DEFINE1(getsid, pid_t, pid)
L
Linus Torvalds 已提交
1057
{
1058 1059 1060 1061 1062
	struct task_struct *p;
	struct pid *sid;
	int retval;

	rcu_read_lock();
1063
	if (!pid)
1064
		sid = task_session(current);
1065
	else {
L
Linus Torvalds 已提交
1066
		retval = -ESRCH;
1067 1068 1069 1070 1071 1072 1073 1074 1075 1076
		p = find_task_by_vpid(pid);
		if (!p)
			goto out;
		sid = task_session(p);
		if (!sid)
			goto out;

		retval = security_task_getsid(p);
		if (retval)
			goto out;
L
Linus Torvalds 已提交
1077
	}
1078 1079 1080 1081
	retval = pid_vnr(sid);
out:
	rcu_read_unlock();
	return retval;
L
Linus Torvalds 已提交
1082 1083
}

1084
SYSCALL_DEFINE0(setsid)
L
Linus Torvalds 已提交
1085
{
1086
	struct task_struct *group_leader = current->group_leader;
1087 1088
	struct pid *sid = task_pid(group_leader);
	pid_t session = pid_vnr(sid);
L
Linus Torvalds 已提交
1089 1090 1091
	int err = -EPERM;

	write_lock_irq(&tasklist_lock);
1092 1093 1094 1095
	/* Fail if I am already a session leader */
	if (group_leader->signal->leader)
		goto out;

1096 1097
	/* Fail if a process group id already exists that equals the
	 * proposed session id.
1098
	 */
1099
	if (pid_task(sid, PIDTYPE_PGID))
L
Linus Torvalds 已提交
1100 1101
		goto out;

1102
	group_leader->signal->leader = 1;
1103
	__set_special_pids(sid);
1104

A
Alan Cox 已提交
1105
	proc_clear_tty(group_leader);
1106

1107
	err = session;
L
Linus Torvalds 已提交
1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135
out:
	write_unlock_irq(&tasklist_lock);
	return err;
}

/*
 * Supplementary group IDs
 */

/* init to 2 - one for init_task, one to ensure it is never freed */
struct group_info init_groups = { .usage = ATOMIC_INIT(2) };

struct group_info *groups_alloc(int gidsetsize)
{
	struct group_info *group_info;
	int nblocks;
	int i;

	nblocks = (gidsetsize + NGROUPS_PER_BLOCK - 1) / NGROUPS_PER_BLOCK;
	/* Make sure we always allocate at least one indirect block pointer */
	nblocks = nblocks ? : 1;
	group_info = kmalloc(sizeof(*group_info) + nblocks*sizeof(gid_t *), GFP_USER);
	if (!group_info)
		return NULL;
	group_info->ngroups = gidsetsize;
	group_info->nblocks = nblocks;
	atomic_set(&group_info->usage, 1);

1136
	if (gidsetsize <= NGROUPS_SMALL)
L
Linus Torvalds 已提交
1137
		group_info->blocks[0] = group_info->small_block;
1138
	else {
L
Linus Torvalds 已提交
1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172
		for (i = 0; i < nblocks; i++) {
			gid_t *b;
			b = (void *)__get_free_page(GFP_USER);
			if (!b)
				goto out_undo_partial_alloc;
			group_info->blocks[i] = b;
		}
	}
	return group_info;

out_undo_partial_alloc:
	while (--i >= 0) {
		free_page((unsigned long)group_info->blocks[i]);
	}
	kfree(group_info);
	return NULL;
}

EXPORT_SYMBOL(groups_alloc);

void groups_free(struct group_info *group_info)
{
	if (group_info->blocks[0] != group_info->small_block) {
		int i;
		for (i = 0; i < group_info->nblocks; i++)
			free_page((unsigned long)group_info->blocks[i]);
	}
	kfree(group_info);
}

EXPORT_SYMBOL(groups_free);

/* export the group_info to a user-space array */
static int groups_to_user(gid_t __user *grouplist,
D
David Howells 已提交
1173
			  const struct group_info *group_info)
L
Linus Torvalds 已提交
1174 1175
{
	int i;
1176
	unsigned int count = group_info->ngroups;
L
Linus Torvalds 已提交
1177 1178

	for (i = 0; i < group_info->nblocks; i++) {
1179 1180
		unsigned int cp_count = min(NGROUPS_PER_BLOCK, count);
		unsigned int len = cp_count * sizeof(*grouplist);
L
Linus Torvalds 已提交
1181

1182
		if (copy_to_user(grouplist, group_info->blocks[i], len))
L
Linus Torvalds 已提交
1183 1184
			return -EFAULT;

1185
		grouplist += NGROUPS_PER_BLOCK;
L
Linus Torvalds 已提交
1186 1187 1188 1189 1190 1191 1192 1193
		count -= cp_count;
	}
	return 0;
}

/* fill a group_info from a user-space array - it must be allocated already */
static int groups_from_user(struct group_info *group_info,
    gid_t __user *grouplist)
1194
{
L
Linus Torvalds 已提交
1195
	int i;
1196
	unsigned int count = group_info->ngroups;
L
Linus Torvalds 已提交
1197 1198

	for (i = 0; i < group_info->nblocks; i++) {
1199 1200
		unsigned int cp_count = min(NGROUPS_PER_BLOCK, count);
		unsigned int len = cp_count * sizeof(*grouplist);
L
Linus Torvalds 已提交
1201

1202
		if (copy_from_user(group_info->blocks[i], grouplist, len))
L
Linus Torvalds 已提交
1203 1204
			return -EFAULT;

1205
		grouplist += NGROUPS_PER_BLOCK;
L
Linus Torvalds 已提交
1206 1207 1208 1209 1210
		count -= cp_count;
	}
	return 0;
}

1211
/* a simple Shell sort */
L
Linus Torvalds 已提交
1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240
static void groups_sort(struct group_info *group_info)
{
	int base, max, stride;
	int gidsetsize = group_info->ngroups;

	for (stride = 1; stride < gidsetsize; stride = 3 * stride + 1)
		; /* nothing */
	stride /= 3;

	while (stride) {
		max = gidsetsize - stride;
		for (base = 0; base < max; base++) {
			int left = base;
			int right = left + stride;
			gid_t tmp = GROUP_AT(group_info, right);

			while (left >= 0 && GROUP_AT(group_info, left) > tmp) {
				GROUP_AT(group_info, right) =
				    GROUP_AT(group_info, left);
				right = left;
				left -= stride;
			}
			GROUP_AT(group_info, right) = tmp;
		}
		stride /= 3;
	}
}

/* a simple bsearch */
1241
int groups_search(const struct group_info *group_info, gid_t grp)
L
Linus Torvalds 已提交
1242
{
1243
	unsigned int left, right;
L
Linus Torvalds 已提交
1244 1245 1246 1247 1248 1249 1250

	if (!group_info)
		return 0;

	left = 0;
	right = group_info->ngroups;
	while (left < right) {
1251
		unsigned int mid = (left+right)/2;
L
Linus Torvalds 已提交
1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262
		int cmp = grp - GROUP_AT(group_info, mid);
		if (cmp > 0)
			left = mid + 1;
		else if (cmp < 0)
			right = mid;
		else
			return 1;
	}
	return 0;
}

1263
/**
D
David Howells 已提交
1264 1265 1266
 * set_groups - Change a group subscription in a set of credentials
 * @new: The newly prepared set of credentials to alter
 * @group_info: The group list to install
1267
 *
D
David Howells 已提交
1268 1269
 * Validate a group subscription and, if valid, insert it into a set
 * of credentials.
1270
 */
D
David Howells 已提交
1271
int set_groups(struct cred *new, struct group_info *group_info)
L
Linus Torvalds 已提交
1272 1273 1274 1275 1276 1277 1278
{
	int retval;

	retval = security_task_setgroups(group_info);
	if (retval)
		return retval;

D
David Howells 已提交
1279
	put_group_info(new->group_info);
L
Linus Torvalds 已提交
1280 1281
	groups_sort(group_info);
	get_group_info(group_info);
D
David Howells 已提交
1282
	new->group_info = group_info;
L
Linus Torvalds 已提交
1283 1284 1285
	return 0;
}

1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296
EXPORT_SYMBOL(set_groups);

/**
 * set_current_groups - Change current's group subscription
 * @group_info: The group list to impose
 *
 * Validate a group subscription and, if valid, impose it upon current's task
 * security record.
 */
int set_current_groups(struct group_info *group_info)
{
D
David Howells 已提交
1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310
	struct cred *new;
	int ret;

	new = prepare_creds();
	if (!new)
		return -ENOMEM;

	ret = set_groups(new, group_info);
	if (ret < 0) {
		abort_creds(new);
		return ret;
	}

	return commit_creds(new);
1311 1312
}

L
Linus Torvalds 已提交
1313 1314
EXPORT_SYMBOL(set_current_groups);

1315
SYSCALL_DEFINE2(getgroups, int, gidsetsize, gid_t __user *, grouplist)
L
Linus Torvalds 已提交
1316
{
1317 1318
	const struct cred *cred = current_cred();
	int i;
L
Linus Torvalds 已提交
1319 1320 1321 1322 1323

	if (gidsetsize < 0)
		return -EINVAL;

	/* no need to grab task_lock here; it cannot change */
1324
	i = cred->group_info->ngroups;
L
Linus Torvalds 已提交
1325 1326 1327 1328 1329
	if (gidsetsize) {
		if (i > gidsetsize) {
			i = -EINVAL;
			goto out;
		}
1330
		if (groups_to_user(grouplist, cred->group_info)) {
L
Linus Torvalds 已提交
1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343
			i = -EFAULT;
			goto out;
		}
	}
out:
	return i;
}

/*
 *	SMP: Our groups are copy-on-write. We can set them safely
 *	without another task interfering.
 */
 
1344
SYSCALL_DEFINE2(setgroups, int, gidsetsize, gid_t __user *, grouplist)
L
Linus Torvalds 已提交
1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373
{
	struct group_info *group_info;
	int retval;

	if (!capable(CAP_SETGID))
		return -EPERM;
	if ((unsigned)gidsetsize > NGROUPS_MAX)
		return -EINVAL;

	group_info = groups_alloc(gidsetsize);
	if (!group_info)
		return -ENOMEM;
	retval = groups_from_user(group_info, grouplist);
	if (retval) {
		put_group_info(group_info);
		return retval;
	}

	retval = set_current_groups(group_info);
	put_group_info(group_info);

	return retval;
}

/*
 * Check whether we're fsgid/egid or in the supplemental group..
 */
int in_group_p(gid_t grp)
{
1374
	const struct cred *cred = current_cred();
L
Linus Torvalds 已提交
1375
	int retval = 1;
1376

1377 1378
	if (grp != cred->fsgid)
		retval = groups_search(cred->group_info, grp);
L
Linus Torvalds 已提交
1379 1380 1381 1382 1383 1384 1385
	return retval;
}

EXPORT_SYMBOL(in_group_p);

int in_egroup_p(gid_t grp)
{
1386
	const struct cred *cred = current_cred();
L
Linus Torvalds 已提交
1387
	int retval = 1;
1388

1389 1390
	if (grp != cred->egid)
		retval = groups_search(cred->group_info, grp);
L
Linus Torvalds 已提交
1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402
	return retval;
}

EXPORT_SYMBOL(in_egroup_p);

DECLARE_RWSEM(uts_sem);

asmlinkage long sys_newuname(struct new_utsname __user * name)
{
	int errno = 0;

	down_read(&uts_sem);
1403
	if (copy_to_user(name, utsname(), sizeof *name))
L
Linus Torvalds 已提交
1404 1405 1406 1407 1408
		errno = -EFAULT;
	up_read(&uts_sem);
	return errno;
}

1409
SYSCALL_DEFINE2(sethostname, char __user *, name, int, len)
L
Linus Torvalds 已提交
1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420
{
	int errno;
	char tmp[__NEW_UTS_LEN];

	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;
	if (len < 0 || len > __NEW_UTS_LEN)
		return -EINVAL;
	down_write(&uts_sem);
	errno = -EFAULT;
	if (!copy_from_user(tmp, name, len)) {
1421 1422 1423 1424
		struct new_utsname *u = utsname();

		memcpy(u->nodename, tmp, len);
		memset(u->nodename + len, 0, sizeof(u->nodename) - len);
L
Linus Torvalds 已提交
1425 1426 1427 1428 1429 1430 1431 1432
		errno = 0;
	}
	up_write(&uts_sem);
	return errno;
}

#ifdef __ARCH_WANT_SYS_GETHOSTNAME

1433
SYSCALL_DEFINE2(gethostname, char __user *, name, int, len)
L
Linus Torvalds 已提交
1434 1435
{
	int i, errno;
1436
	struct new_utsname *u;
L
Linus Torvalds 已提交
1437 1438 1439 1440

	if (len < 0)
		return -EINVAL;
	down_read(&uts_sem);
1441 1442
	u = utsname();
	i = 1 + strlen(u->nodename);
L
Linus Torvalds 已提交
1443 1444 1445
	if (i > len)
		i = len;
	errno = 0;
1446
	if (copy_to_user(name, u->nodename, i))
L
Linus Torvalds 已提交
1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457
		errno = -EFAULT;
	up_read(&uts_sem);
	return errno;
}

#endif

/*
 * Only setdomainname; getdomainname can be implemented by calling
 * uname()
 */
1458
SYSCALL_DEFINE2(setdomainname, char __user *, name, int, len)
L
Linus Torvalds 已提交
1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470
{
	int errno;
	char tmp[__NEW_UTS_LEN];

	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;
	if (len < 0 || len > __NEW_UTS_LEN)
		return -EINVAL;

	down_write(&uts_sem);
	errno = -EFAULT;
	if (!copy_from_user(tmp, name, len)) {
1471 1472 1473 1474
		struct new_utsname *u = utsname();

		memcpy(u->domainname, tmp, len);
		memset(u->domainname + len, 0, sizeof(u->domainname) - len);
L
Linus Torvalds 已提交
1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508
		errno = 0;
	}
	up_write(&uts_sem);
	return errno;
}

asmlinkage long sys_getrlimit(unsigned int resource, struct rlimit __user *rlim)
{
	if (resource >= RLIM_NLIMITS)
		return -EINVAL;
	else {
		struct rlimit value;
		task_lock(current->group_leader);
		value = current->signal->rlim[resource];
		task_unlock(current->group_leader);
		return copy_to_user(rlim, &value, sizeof(*rlim)) ? -EFAULT : 0;
	}
}

#ifdef __ARCH_WANT_SYS_OLD_GETRLIMIT

/*
 *	Back compatibility for getrlimit. Needed for some apps.
 */
 
asmlinkage long sys_old_getrlimit(unsigned int resource, struct rlimit __user *rlim)
{
	struct rlimit x;
	if (resource >= RLIM_NLIMITS)
		return -EINVAL;

	task_lock(current->group_leader);
	x = current->signal->rlim[resource];
	task_unlock(current->group_leader);
1509
	if (x.rlim_cur > 0x7FFFFFFF)
L
Linus Torvalds 已提交
1510
		x.rlim_cur = 0x7FFFFFFF;
1511
	if (x.rlim_max > 0x7FFFFFFF)
L
Linus Torvalds 已提交
1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524
		x.rlim_max = 0x7FFFFFFF;
	return copy_to_user(rlim, &x, sizeof(x))?-EFAULT:0;
}

#endif

asmlinkage long sys_setrlimit(unsigned int resource, struct rlimit __user *rlim)
{
	struct rlimit new_rlim, *old_rlim;
	int retval;

	if (resource >= RLIM_NLIMITS)
		return -EINVAL;
A
Andrew Morton 已提交
1525
	if (copy_from_user(&new_rlim, rlim, sizeof(*rlim)))
L
Linus Torvalds 已提交
1526 1527 1528 1529 1530
		return -EFAULT;
	old_rlim = current->signal->rlim + resource;
	if ((new_rlim.rlim_max > old_rlim->rlim_max) &&
	    !capable(CAP_SYS_RESOURCE))
		return -EPERM;
1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542

	if (resource == RLIMIT_NOFILE) {
		if (new_rlim.rlim_max == RLIM_INFINITY)
			new_rlim.rlim_max = sysctl_nr_open;
		if (new_rlim.rlim_cur == RLIM_INFINITY)
			new_rlim.rlim_cur = sysctl_nr_open;
		if (new_rlim.rlim_max > sysctl_nr_open)
			return -EPERM;
	}

	if (new_rlim.rlim_cur > new_rlim.rlim_max)
		return -EINVAL;
L
Linus Torvalds 已提交
1543 1544 1545 1546 1547

	retval = security_task_setrlimit(resource, &new_rlim);
	if (retval)
		return retval;

1548 1549 1550 1551 1552 1553 1554 1555 1556 1557
	if (resource == RLIMIT_CPU && new_rlim.rlim_cur == 0) {
		/*
		 * The caller is asking for an immediate RLIMIT_CPU
		 * expiry.  But we use the zero value to mean "it was
		 * never set".  So let's cheat and make it one second
		 * instead
		 */
		new_rlim.rlim_cur = 1;
	}

L
Linus Torvalds 已提交
1558 1559 1560 1561
	task_lock(current->group_leader);
	*old_rlim = new_rlim;
	task_unlock(current->group_leader);

A
Andrew Morton 已提交
1562 1563
	if (resource != RLIMIT_CPU)
		goto out;
1564 1565 1566 1567 1568 1569 1570

	/*
	 * RLIMIT_CPU handling.   Note that the kernel fails to return an error
	 * code if it rejected the user's attempt to set RLIMIT_CPU.  This is a
	 * very long-standing error, and fixing it now risks breakage of
	 * applications, so we live with it
	 */
A
Andrew Morton 已提交
1571 1572 1573
	if (new_rlim.rlim_cur == RLIM_INFINITY)
		goto out;

1574
	update_rlimit_cpu(new_rlim.rlim_cur);
A
Andrew Morton 已提交
1575
out:
L
Linus Torvalds 已提交
1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593
	return 0;
}

/*
 * It would make sense to put struct rusage in the task_struct,
 * except that would make the task_struct be *really big*.  After
 * task_struct gets moved into malloc'ed memory, it would
 * make sense to do this.  It will make moving the rest of the information
 * a lot simpler!  (Which we're not doing right now because we're not
 * measuring them yet).
 *
 * When sampling multiple threads for RUSAGE_SELF, under SMP we might have
 * races with threads incrementing their own counters.  But since word
 * reads are atomic, we either get new values or old values and we don't
 * care which for the sums.  We always take the siglock to protect reading
 * the c* fields from p->signal from races with exit.c updating those
 * fields when reaping, so a sample either gets all the additions of a
 * given child after it's reaped, or none so this sample is before reaping.
1594
 *
1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608
 * Locking:
 * We need to take the siglock for CHILDEREN, SELF and BOTH
 * for  the cases current multithreaded, non-current single threaded
 * non-current multithreaded.  Thread traversal is now safe with
 * the siglock held.
 * Strictly speaking, we donot need to take the siglock if we are current and
 * single threaded,  as no one else can take our signal_struct away, no one
 * else can  reap the  children to update signal->c* counters, and no one else
 * can race with the signal-> fields. If we do not take any lock, the
 * signal-> fields could be read out of order while another thread was just
 * exiting. So we should  place a read memory barrier when we avoid the lock.
 * On the writer side,  write memory barrier is implied in  __exit_signal
 * as __exit_signal releases  the siglock spinlock after updating the signal->
 * fields. But we don't do this yet to keep things simple.
1609
 *
L
Linus Torvalds 已提交
1610 1611
 */

1612
static void accumulate_thread_rusage(struct task_struct *t, struct rusage *r)
S
Sripathi Kodi 已提交
1613 1614 1615 1616 1617 1618 1619 1620 1621
{
	r->ru_nvcsw += t->nvcsw;
	r->ru_nivcsw += t->nivcsw;
	r->ru_minflt += t->min_flt;
	r->ru_majflt += t->maj_flt;
	r->ru_inblock += task_io_get_inblock(t);
	r->ru_oublock += task_io_get_oublock(t);
}

L
Linus Torvalds 已提交
1622 1623 1624 1625 1626
static void k_getrusage(struct task_struct *p, int who, struct rusage *r)
{
	struct task_struct *t;
	unsigned long flags;
	cputime_t utime, stime;
1627
	struct task_cputime cputime;
L
Linus Torvalds 已提交
1628 1629

	memset((char *) r, 0, sizeof *r);
1630
	utime = stime = cputime_zero;
L
Linus Torvalds 已提交
1631

S
Sripathi Kodi 已提交
1632
	if (who == RUSAGE_THREAD) {
1633 1634
		utime = task_utime(current);
		stime = task_stime(current);
1635
		accumulate_thread_rusage(p, r);
S
Sripathi Kodi 已提交
1636 1637 1638
		goto out;
	}

1639
	if (!lock_task_sighand(p, &flags))
1640
		return;
O
Oleg Nesterov 已提交
1641

L
Linus Torvalds 已提交
1642
	switch (who) {
O
Oleg Nesterov 已提交
1643
		case RUSAGE_BOTH:
L
Linus Torvalds 已提交
1644 1645 1646 1647 1648 1649 1650
		case RUSAGE_CHILDREN:
			utime = p->signal->cutime;
			stime = p->signal->cstime;
			r->ru_nvcsw = p->signal->cnvcsw;
			r->ru_nivcsw = p->signal->cnivcsw;
			r->ru_minflt = p->signal->cmin_flt;
			r->ru_majflt = p->signal->cmaj_flt;
1651 1652
			r->ru_inblock = p->signal->cinblock;
			r->ru_oublock = p->signal->coublock;
O
Oleg Nesterov 已提交
1653 1654 1655 1656

			if (who == RUSAGE_CHILDREN)
				break;

L
Linus Torvalds 已提交
1657
		case RUSAGE_SELF:
1658 1659 1660
			thread_group_cputime(p, &cputime);
			utime = cputime_add(utime, cputime.utime);
			stime = cputime_add(stime, cputime.stime);
L
Linus Torvalds 已提交
1661 1662 1663 1664
			r->ru_nvcsw += p->signal->nvcsw;
			r->ru_nivcsw += p->signal->nivcsw;
			r->ru_minflt += p->signal->min_flt;
			r->ru_majflt += p->signal->maj_flt;
1665 1666
			r->ru_inblock += p->signal->inblock;
			r->ru_oublock += p->signal->oublock;
L
Linus Torvalds 已提交
1667 1668
			t = p;
			do {
1669
				accumulate_thread_rusage(t, r);
L
Linus Torvalds 已提交
1670 1671 1672
				t = next_thread(t);
			} while (t != p);
			break;
O
Oleg Nesterov 已提交
1673

L
Linus Torvalds 已提交
1674 1675 1676
		default:
			BUG();
	}
1677 1678
	unlock_task_sighand(p, &flags);

S
Sripathi Kodi 已提交
1679
out:
O
Oleg Nesterov 已提交
1680 1681
	cputime_to_timeval(utime, &r->ru_utime);
	cputime_to_timeval(stime, &r->ru_stime);
L
Linus Torvalds 已提交
1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692
}

int getrusage(struct task_struct *p, int who, struct rusage __user *ru)
{
	struct rusage r;
	k_getrusage(p, who, &r);
	return copy_to_user(ru, &r, sizeof(r)) ? -EFAULT : 0;
}

asmlinkage long sys_getrusage(int who, struct rusage __user *ru)
{
S
Sripathi Kodi 已提交
1693 1694
	if (who != RUSAGE_SELF && who != RUSAGE_CHILDREN &&
	    who != RUSAGE_THREAD)
L
Linus Torvalds 已提交
1695 1696 1697 1698 1699 1700 1701 1702 1703
		return -EINVAL;
	return getrusage(current, who, ru);
}

asmlinkage long sys_umask(int mask)
{
	mask = xchg(&current->fs->umask, mask & S_IRWXUGO);
	return mask;
}
1704

L
Linus Torvalds 已提交
1705 1706 1707
asmlinkage long sys_prctl(int option, unsigned long arg2, unsigned long arg3,
			  unsigned long arg4, unsigned long arg5)
{
1708 1709 1710
	struct task_struct *me = current;
	unsigned char comm[sizeof(me->comm)];
	long error;
L
Linus Torvalds 已提交
1711

D
David Howells 已提交
1712 1713
	error = security_task_prctl(option, arg2, arg3, arg4, arg5);
	if (error != -ENOSYS)
L
Linus Torvalds 已提交
1714 1715
		return error;

D
David Howells 已提交
1716
	error = 0;
L
Linus Torvalds 已提交
1717 1718
	switch (option) {
		case PR_SET_PDEATHSIG:
1719
			if (!valid_signal(arg2)) {
L
Linus Torvalds 已提交
1720 1721 1722
				error = -EINVAL;
				break;
			}
1723 1724
			me->pdeath_signal = arg2;
			error = 0;
L
Linus Torvalds 已提交
1725 1726
			break;
		case PR_GET_PDEATHSIG:
1727
			error = put_user(me->pdeath_signal, (int __user *)arg2);
L
Linus Torvalds 已提交
1728 1729
			break;
		case PR_GET_DUMPABLE:
1730
			error = get_dumpable(me->mm);
L
Linus Torvalds 已提交
1731 1732
			break;
		case PR_SET_DUMPABLE:
1733
			if (arg2 < 0 || arg2 > 1) {
L
Linus Torvalds 已提交
1734 1735 1736
				error = -EINVAL;
				break;
			}
1737 1738
			set_dumpable(me->mm, arg2);
			error = 0;
L
Linus Torvalds 已提交
1739 1740 1741
			break;

		case PR_SET_UNALIGN:
1742
			error = SET_UNALIGN_CTL(me, arg2);
L
Linus Torvalds 已提交
1743 1744
			break;
		case PR_GET_UNALIGN:
1745
			error = GET_UNALIGN_CTL(me, arg2);
L
Linus Torvalds 已提交
1746 1747
			break;
		case PR_SET_FPEMU:
1748
			error = SET_FPEMU_CTL(me, arg2);
L
Linus Torvalds 已提交
1749 1750
			break;
		case PR_GET_FPEMU:
1751
			error = GET_FPEMU_CTL(me, arg2);
L
Linus Torvalds 已提交
1752 1753
			break;
		case PR_SET_FPEXC:
1754
			error = SET_FPEXC_CTL(me, arg2);
L
Linus Torvalds 已提交
1755 1756
			break;
		case PR_GET_FPEXC:
1757
			error = GET_FPEXC_CTL(me, arg2);
L
Linus Torvalds 已提交
1758 1759 1760 1761 1762
			break;
		case PR_GET_TIMING:
			error = PR_TIMING_STATISTICAL;
			break;
		case PR_SET_TIMING:
1763
			if (arg2 != PR_TIMING_STATISTICAL)
L
Linus Torvalds 已提交
1764
				error = -EINVAL;
1765 1766
			else
				error = 0;
L
Linus Torvalds 已提交
1767 1768
			break;

1769 1770 1771 1772
		case PR_SET_NAME:
			comm[sizeof(me->comm)-1] = 0;
			if (strncpy_from_user(comm, (char __user *)arg2,
					      sizeof(me->comm) - 1) < 0)
L
Linus Torvalds 已提交
1773
				return -EFAULT;
1774
			set_task_comm(me, comm);
L
Linus Torvalds 已提交
1775
			return 0;
1776 1777 1778 1779
		case PR_GET_NAME:
			get_task_comm(comm, me);
			if (copy_to_user((char __user *)arg2, comm,
					 sizeof(comm)))
L
Linus Torvalds 已提交
1780 1781
				return -EFAULT;
			return 0;
1782
		case PR_GET_ENDIAN:
1783
			error = GET_ENDIAN(me, arg2);
1784 1785
			break;
		case PR_SET_ENDIAN:
1786
			error = SET_ENDIAN(me, arg2);
1787 1788
			break;

1789 1790 1791 1792 1793 1794
		case PR_GET_SECCOMP:
			error = prctl_get_seccomp();
			break;
		case PR_SET_SECCOMP:
			error = prctl_set_seccomp(arg2);
			break;
1795 1796 1797 1798 1799 1800
		case PR_GET_TSC:
			error = GET_TSC_CTL(arg2);
			break;
		case PR_SET_TSC:
			error = SET_TSC_CTL(arg2);
			break;
1801 1802 1803 1804 1805 1806 1807 1808 1809
		case PR_GET_TIMERSLACK:
			error = current->timer_slack_ns;
			break;
		case PR_SET_TIMERSLACK:
			if (arg2 <= 0)
				current->timer_slack_ns =
					current->default_timer_slack_ns;
			else
				current->timer_slack_ns = arg2;
1810
			error = 0;
1811
			break;
L
Linus Torvalds 已提交
1812 1813 1814 1815 1816 1817
		default:
			error = -EINVAL;
			break;
	}
	return error;
}
1818 1819

asmlinkage long sys_getcpu(unsigned __user *cpup, unsigned __user *nodep,
1820
			   struct getcpu_cache __user *unused)
1821 1822 1823 1824 1825 1826 1827 1828 1829
{
	int err = 0;
	int cpu = raw_smp_processor_id();
	if (cpup)
		err |= put_user(cpu, cpup);
	if (nodep)
		err |= put_user(cpu_to_node(cpu), nodep);
	return err ? -EFAULT : 0;
}
1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862

char poweroff_cmd[POWEROFF_CMD_PATH_LEN] = "/sbin/poweroff";

static void argv_cleanup(char **argv, char **envp)
{
	argv_free(argv);
}

/**
 * orderly_poweroff - Trigger an orderly system poweroff
 * @force: force poweroff if command execution fails
 *
 * This may be called from any context to trigger a system shutdown.
 * If the orderly shutdown fails, it will force an immediate shutdown.
 */
int orderly_poweroff(bool force)
{
	int argc;
	char **argv = argv_split(GFP_ATOMIC, poweroff_cmd, &argc);
	static char *envp[] = {
		"HOME=/",
		"PATH=/sbin:/bin:/usr/sbin:/usr/bin",
		NULL
	};
	int ret = -ENOMEM;
	struct subprocess_info *info;

	if (argv == NULL) {
		printk(KERN_WARNING "%s failed to allocate memory for \"%s\"\n",
		       __func__, poweroff_cmd);
		goto out;
	}

1863
	info = call_usermodehelper_setup(argv[0], argv, envp, GFP_ATOMIC);
1864 1865 1866 1867 1868 1869 1870
	if (info == NULL) {
		argv_free(argv);
		goto out;
	}

	call_usermodehelper_setcleanup(info, argv_cleanup);

1871
	ret = call_usermodehelper_exec(info, UMH_NO_WAIT);
1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887

  out:
	if (ret && force) {
		printk(KERN_WARNING "Failed to start orderly shutdown: "
		       "forcing the issue\n");

		/* I guess this should try to kick off some daemon to
		   sync and poweroff asap.  Or not even bother syncing
		   if we're doing an emergency shutdown? */
		emergency_sync();
		kernel_power_off();
	}

	return ret;
}
EXPORT_SYMBOL_GPL(orderly_poweroff);