sys.c 50.6 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6
/*
 *  linux/kernel/sys.c
 *
 *  Copyright (C) 1991, 1992  Linus Torvalds
 */

7
#include <linux/export.h>
L
Linus Torvalds 已提交
8 9 10 11 12 13 14
#include <linux/mm.h>
#include <linux/utsname.h>
#include <linux/mman.h>
#include <linux/reboot.h>
#include <linux/prctl.h>
#include <linux/highuid.h>
#include <linux/fs.h>
15
#include <linux/kmod.h>
16
#include <linux/perf_event.h>
17
#include <linux/resource.h>
18
#include <linux/kernel.h>
L
Linus Torvalds 已提交
19
#include <linux/workqueue.h>
20
#include <linux/capability.h>
L
Linus Torvalds 已提交
21 22 23 24 25 26 27 28
#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>
29
#include <linux/signal.h>
M
Matt Helsley 已提交
30
#include <linux/cn_proc.h>
31
#include <linux/getcpu.h>
32
#include <linux/task_io_accounting_ops.h>
33
#include <linux/seccomp.h>
M
Mark Lord 已提交
34
#include <linux/cpu.h>
35
#include <linux/personality.h>
36
#include <linux/ptrace.h>
37
#include <linux/fs_struct.h>
38 39
#include <linux/file.h>
#include <linux/mount.h>
40
#include <linux/gfp.h>
41
#include <linux/syscore_ops.h>
42 43
#include <linux/version.h>
#include <linux/ctype.h>
L
Linus Torvalds 已提交
44 45 46

#include <linux/compat.h>
#include <linux/syscalls.h>
47
#include <linux/kprobes.h>
48
#include <linux/user_namespace.h>
49
#include <linux/binfmts.h>
L
Linus Torvalds 已提交
50

51 52 53 54 55
#include <linux/sched.h>
#include <linux/rcupdate.h>
#include <linux/uidgid.h>
#include <linux/cred.h>

56
#include <linux/kmsg_dump.h>
57 58
/* Move somewhere else to avoid recompiling? */
#include <generated/utsrelease.h>
59

L
Linus Torvalds 已提交
60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81
#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
82 83 84 85 86 87
#ifndef GET_ENDIAN
# define GET_ENDIAN(a,b)	(-EINVAL)
#endif
#ifndef SET_ENDIAN
# define SET_ENDIAN(a,b)	(-EINVAL)
#endif
88 89 90 91 92 93
#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
94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116

/*
 * 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;

EXPORT_SYMBOL(overflowuid);
EXPORT_SYMBOL(overflowgid);

/*
 * 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);

117 118 119 120 121 122 123 124 125 126
/*
 * Returns true if current's euid is same as p's uid or euid,
 * or has CAP_SYS_NICE to p's user_ns.
 *
 * Called with rcu_read_lock, creds are safe
 */
static bool set_one_prio_perm(struct task_struct *p)
{
	const struct cred *cred = current_cred(), *pcred = __task_cred(p);

127 128
	if (uid_eq(pcred->uid,  cred->euid) ||
	    uid_eq(pcred->euid, cred->euid))
129
		return true;
130
	if (ns_capable(pcred->user_ns, CAP_SYS_NICE))
131 132 133 134
		return true;
	return false;
}

135 136 137 138
/*
 * set the priority of a task
 * - the caller must hold the RCU read lock
 */
L
Linus Torvalds 已提交
139 140 141 142
static int set_one_prio(struct task_struct *p, int niceval, int error)
{
	int no_nice;

143
	if (!set_one_prio_perm(p)) {
L
Linus Torvalds 已提交
144 145 146
		error = -EPERM;
		goto out;
	}
M
Matt Mackall 已提交
147
	if (niceval < task_nice(p) && !can_nice(p, niceval)) {
L
Linus Torvalds 已提交
148 149 150 151 152 153 154 155 156 157 158 159 160 161 162
		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;
}

163
SYSCALL_DEFINE3(setpriority, int, which, int, who, int, niceval)
L
Linus Torvalds 已提交
164 165 166
{
	struct task_struct *g, *p;
	struct user_struct *user;
167
	const struct cred *cred = current_cred();
L
Linus Torvalds 已提交
168
	int error = -EINVAL;
169
	struct pid *pgrp;
170
	kuid_t uid;
L
Linus Torvalds 已提交
171

172
	if (which > PRIO_USER || which < PRIO_PROCESS)
L
Linus Torvalds 已提交
173 174 175 176
		goto out;

	/* normalize: avoid signed division (rounding problems) */
	error = -ESRCH;
177 178 179 180
	if (niceval < MIN_NICE)
		niceval = MIN_NICE;
	if (niceval > MAX_NICE)
		niceval = MAX_NICE;
L
Linus Torvalds 已提交
181

182
	rcu_read_lock();
L
Linus Torvalds 已提交
183 184 185
	read_lock(&tasklist_lock);
	switch (which) {
		case PRIO_PROCESS:
186
			if (who)
187
				p = find_task_by_vpid(who);
188 189
			else
				p = current;
L
Linus Torvalds 已提交
190 191 192 193
			if (p)
				error = set_one_prio(p, niceval, error);
			break;
		case PRIO_PGRP:
194
			if (who)
195
				pgrp = find_vpid(who);
196 197
			else
				pgrp = task_pgrp(current);
198
			do_each_pid_thread(pgrp, PIDTYPE_PGID, p) {
L
Linus Torvalds 已提交
199
				error = set_one_prio(p, niceval, error);
200
			} while_each_pid_thread(pgrp, PIDTYPE_PGID, p);
L
Linus Torvalds 已提交
201 202
			break;
		case PRIO_USER:
203
			uid = make_kuid(cred->user_ns, who);
204
			user = cred->user;
L
Linus Torvalds 已提交
205
			if (!who)
206 207
				uid = cred->uid;
			else if (!uid_eq(uid, cred->uid) &&
208
				 !(user = find_user(uid)))
209
				goto out_unlock;	/* No processes for this user */
L
Linus Torvalds 已提交
210

211
			do_each_thread(g, p) {
212
				if (uid_eq(task_uid(p), uid))
L
Linus Torvalds 已提交
213
					error = set_one_prio(p, niceval, error);
214
			} while_each_thread(g, p);
215
			if (!uid_eq(uid, cred->uid))
L
Linus Torvalds 已提交
216 217 218 219 220
				free_uid(user);		/* For find_user() */
			break;
	}
out_unlock:
	read_unlock(&tasklist_lock);
221
	rcu_read_unlock();
L
Linus Torvalds 已提交
222 223 224 225 226 227 228 229 230 231
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.
 */
232
SYSCALL_DEFINE2(getpriority, int, which, int, who)
L
Linus Torvalds 已提交
233 234 235
{
	struct task_struct *g, *p;
	struct user_struct *user;
236
	const struct cred *cred = current_cred();
L
Linus Torvalds 已提交
237
	long niceval, retval = -ESRCH;
238
	struct pid *pgrp;
239
	kuid_t uid;
L
Linus Torvalds 已提交
240

241
	if (which > PRIO_USER || which < PRIO_PROCESS)
L
Linus Torvalds 已提交
242 243
		return -EINVAL;

244
	rcu_read_lock();
L
Linus Torvalds 已提交
245 246 247
	read_lock(&tasklist_lock);
	switch (which) {
		case PRIO_PROCESS:
248
			if (who)
249
				p = find_task_by_vpid(who);
250 251
			else
				p = current;
L
Linus Torvalds 已提交
252
			if (p) {
253
				niceval = nice_to_rlimit(task_nice(p));
L
Linus Torvalds 已提交
254 255 256 257 258
				if (niceval > retval)
					retval = niceval;
			}
			break;
		case PRIO_PGRP:
259
			if (who)
260
				pgrp = find_vpid(who);
261 262
			else
				pgrp = task_pgrp(current);
263
			do_each_pid_thread(pgrp, PIDTYPE_PGID, p) {
264
				niceval = nice_to_rlimit(task_nice(p));
L
Linus Torvalds 已提交
265 266
				if (niceval > retval)
					retval = niceval;
267
			} while_each_pid_thread(pgrp, PIDTYPE_PGID, p);
L
Linus Torvalds 已提交
268 269
			break;
		case PRIO_USER:
270
			uid = make_kuid(cred->user_ns, who);
271
			user = cred->user;
L
Linus Torvalds 已提交
272
			if (!who)
273 274
				uid = cred->uid;
			else if (!uid_eq(uid, cred->uid) &&
275
				 !(user = find_user(uid)))
276
				goto out_unlock;	/* No processes for this user */
L
Linus Torvalds 已提交
277

278
			do_each_thread(g, p) {
279
				if (uid_eq(task_uid(p), uid)) {
280
					niceval = nice_to_rlimit(task_nice(p));
L
Linus Torvalds 已提交
281 282 283
					if (niceval > retval)
						retval = niceval;
				}
284
			} while_each_thread(g, p);
285
			if (!uid_eq(uid, cred->uid))
L
Linus Torvalds 已提交
286 287 288 289 290
				free_uid(user);		/* for find_user() */
			break;
	}
out_unlock:
	read_unlock(&tasklist_lock);
291
	rcu_read_unlock();
L
Linus Torvalds 已提交
292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313

	return retval;
}

/*
 * 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).
 */
314
SYSCALL_DEFINE2(setregid, gid_t, rgid, gid_t, egid)
L
Linus Torvalds 已提交
315
{
316
	struct user_namespace *ns = current_user_ns();
D
David Howells 已提交
317 318
	const struct cred *old;
	struct cred *new;
L
Linus Torvalds 已提交
319
	int retval;
320 321 322 323 324 325 326 327 328
	kgid_t krgid, kegid;

	krgid = make_kgid(ns, rgid);
	kegid = make_kgid(ns, egid);

	if ((rgid != (gid_t) -1) && !gid_valid(krgid))
		return -EINVAL;
	if ((egid != (gid_t) -1) && !gid_valid(kegid))
		return -EINVAL;
L
Linus Torvalds 已提交
329

D
David Howells 已提交
330 331 332 333 334 335
	new = prepare_creds();
	if (!new)
		return -ENOMEM;
	old = current_cred();

	retval = -EPERM;
L
Linus Torvalds 已提交
336
	if (rgid != (gid_t) -1) {
337 338
		if (gid_eq(old->gid, krgid) ||
		    gid_eq(old->egid, krgid) ||
339
		    ns_capable(old->user_ns, CAP_SETGID))
340
			new->gid = krgid;
L
Linus Torvalds 已提交
341
		else
D
David Howells 已提交
342
			goto error;
L
Linus Torvalds 已提交
343 344
	}
	if (egid != (gid_t) -1) {
345 346 347
		if (gid_eq(old->gid, kegid) ||
		    gid_eq(old->egid, kegid) ||
		    gid_eq(old->sgid, kegid) ||
348
		    ns_capable(old->user_ns, CAP_SETGID))
349
			new->egid = kegid;
350
		else
D
David Howells 已提交
351
			goto error;
L
Linus Torvalds 已提交
352
	}
D
David Howells 已提交
353

L
Linus Torvalds 已提交
354
	if (rgid != (gid_t) -1 ||
355
	    (egid != (gid_t) -1 && !gid_eq(kegid, old->gid)))
D
David Howells 已提交
356 357 358 359 360 361 362 363
		new->sgid = new->egid;
	new->fsgid = new->egid;

	return commit_creds(new);

error:
	abort_creds(new);
	return retval;
L
Linus Torvalds 已提交
364 365 366 367 368 369 370
}

/*
 * setgid() is implemented like SysV w/ SAVED_IDS 
 *
 * SMP: Same implicit races as above.
 */
371
SYSCALL_DEFINE1(setgid, gid_t, gid)
L
Linus Torvalds 已提交
372
{
373
	struct user_namespace *ns = current_user_ns();
D
David Howells 已提交
374 375
	const struct cred *old;
	struct cred *new;
L
Linus Torvalds 已提交
376
	int retval;
377 378 379 380 381
	kgid_t kgid;

	kgid = make_kgid(ns, gid);
	if (!gid_valid(kgid))
		return -EINVAL;
L
Linus Torvalds 已提交
382

D
David Howells 已提交
383 384 385 386 387 388
	new = prepare_creds();
	if (!new)
		return -ENOMEM;
	old = current_cred();

	retval = -EPERM;
389
	if (ns_capable(old->user_ns, CAP_SETGID))
390 391 392
		new->gid = new->egid = new->sgid = new->fsgid = kgid;
	else if (gid_eq(kgid, old->gid) || gid_eq(kgid, old->sgid))
		new->egid = new->fsgid = kgid;
L
Linus Torvalds 已提交
393
	else
D
David Howells 已提交
394
		goto error;
L
Linus Torvalds 已提交
395

D
David Howells 已提交
396 397 398 399 400
	return commit_creds(new);

error:
	abort_creds(new);
	return retval;
L
Linus Torvalds 已提交
401
}
402

D
David Howells 已提交
403 404 405 406
/*
 * 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
407 408 409
{
	struct user_struct *new_user;

410
	new_user = alloc_uid(new->uid);
L
Linus Torvalds 已提交
411 412 413
	if (!new_user)
		return -EAGAIN;

414 415 416 417 418 419 420
	/*
	 * We don't fail in case of NPROC limit excess here because too many
	 * poorly written programs don't check set*uid() return code, assuming
	 * it never fails if called by root.  We may still enforce NPROC limit
	 * for programs doing set*uid()+execve() by harmlessly deferring the
	 * failure to the execve() stage.
	 */
421
	if (atomic_read(&new_user->processes) >= rlimit(RLIMIT_NPROC) &&
422 423 424 425
			new_user != INIT_USER)
		current->flags |= PF_NPROC_EXCEEDED;
	else
		current->flags &= ~PF_NPROC_EXCEEDED;
L
Linus Torvalds 已提交
426

D
David Howells 已提交
427 428
	free_uid(new->user);
	new->user = new_user;
L
Linus Torvalds 已提交
429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446
	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. 
 */
447
SYSCALL_DEFINE2(setreuid, uid_t, ruid, uid_t, euid)
L
Linus Torvalds 已提交
448
{
449
	struct user_namespace *ns = current_user_ns();
D
David Howells 已提交
450 451
	const struct cred *old;
	struct cred *new;
L
Linus Torvalds 已提交
452
	int retval;
453 454 455 456 457 458 459 460 461
	kuid_t kruid, keuid;

	kruid = make_kuid(ns, ruid);
	keuid = make_kuid(ns, euid);

	if ((ruid != (uid_t) -1) && !uid_valid(kruid))
		return -EINVAL;
	if ((euid != (uid_t) -1) && !uid_valid(keuid))
		return -EINVAL;
L
Linus Torvalds 已提交
462

D
David Howells 已提交
463 464 465 466 467 468
	new = prepare_creds();
	if (!new)
		return -ENOMEM;
	old = current_cred();

	retval = -EPERM;
L
Linus Torvalds 已提交
469
	if (ruid != (uid_t) -1) {
470 471 472
		new->uid = kruid;
		if (!uid_eq(old->uid, kruid) &&
		    !uid_eq(old->euid, kruid) &&
473
		    !ns_capable(old->user_ns, CAP_SETUID))
D
David Howells 已提交
474
			goto error;
L
Linus Torvalds 已提交
475 476 477
	}

	if (euid != (uid_t) -1) {
478 479 480 481
		new->euid = keuid;
		if (!uid_eq(old->uid, keuid) &&
		    !uid_eq(old->euid, keuid) &&
		    !uid_eq(old->suid, keuid) &&
482
		    !ns_capable(old->user_ns, CAP_SETUID))
D
David Howells 已提交
483
			goto error;
L
Linus Torvalds 已提交
484 485
	}

486
	if (!uid_eq(new->uid, old->uid)) {
487 488 489 490
		retval = set_user(new);
		if (retval < 0)
			goto error;
	}
L
Linus Torvalds 已提交
491
	if (ruid != (uid_t) -1 ||
492
	    (euid != (uid_t) -1 && !uid_eq(keuid, old->uid)))
D
David Howells 已提交
493 494
		new->suid = new->euid;
	new->fsuid = new->euid;
L
Linus Torvalds 已提交
495

D
David Howells 已提交
496 497 498
	retval = security_task_fix_setuid(new, old, LSM_SETID_RE);
	if (retval < 0)
		goto error;
L
Linus Torvalds 已提交
499

D
David Howells 已提交
500
	return commit_creds(new);
L
Linus Torvalds 已提交
501

D
David Howells 已提交
502 503 504 505
error:
	abort_creds(new);
	return retval;
}
L
Linus Torvalds 已提交
506 507 508 509 510 511 512 513 514 515 516 517
		
/*
 * 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.  
 */
518
SYSCALL_DEFINE1(setuid, uid_t, uid)
L
Linus Torvalds 已提交
519
{
520
	struct user_namespace *ns = current_user_ns();
D
David Howells 已提交
521 522
	const struct cred *old;
	struct cred *new;
L
Linus Torvalds 已提交
523
	int retval;
524 525 526 527 528
	kuid_t kuid;

	kuid = make_kuid(ns, uid);
	if (!uid_valid(kuid))
		return -EINVAL;
L
Linus Torvalds 已提交
529

D
David Howells 已提交
530 531 532 533 534 535
	new = prepare_creds();
	if (!new)
		return -ENOMEM;
	old = current_cred();

	retval = -EPERM;
536
	if (ns_capable(old->user_ns, CAP_SETUID)) {
537 538
		new->suid = new->uid = kuid;
		if (!uid_eq(kuid, old->uid)) {
539 540 541
			retval = set_user(new);
			if (retval < 0)
				goto error;
D
David Howells 已提交
542
		}
543
	} else if (!uid_eq(kuid, old->uid) && !uid_eq(kuid, new->suid)) {
D
David Howells 已提交
544
		goto error;
L
Linus Torvalds 已提交
545 546
	}

547
	new->fsuid = new->euid = kuid;
D
David Howells 已提交
548 549 550 551

	retval = security_task_fix_setuid(new, old, LSM_SETID_ID);
	if (retval < 0)
		goto error;
L
Linus Torvalds 已提交
552

D
David Howells 已提交
553
	return commit_creds(new);
L
Linus Torvalds 已提交
554

D
David Howells 已提交
555 556 557
error:
	abort_creds(new);
	return retval;
L
Linus Torvalds 已提交
558 559 560 561 562 563 564
}


/*
 * This function implements a generic ability to update ruid, euid,
 * and suid.  This allows you to implement the 4.4 compatible seteuid().
 */
565
SYSCALL_DEFINE3(setresuid, uid_t, ruid, uid_t, euid, uid_t, suid)
L
Linus Torvalds 已提交
566
{
567
	struct user_namespace *ns = current_user_ns();
D
David Howells 已提交
568 569
	const struct cred *old;
	struct cred *new;
L
Linus Torvalds 已提交
570
	int retval;
571 572 573 574 575 576 577 578 579 580 581 582 583 584
	kuid_t kruid, keuid, ksuid;

	kruid = make_kuid(ns, ruid);
	keuid = make_kuid(ns, euid);
	ksuid = make_kuid(ns, suid);

	if ((ruid != (uid_t) -1) && !uid_valid(kruid))
		return -EINVAL;

	if ((euid != (uid_t) -1) && !uid_valid(keuid))
		return -EINVAL;

	if ((suid != (uid_t) -1) && !uid_valid(ksuid))
		return -EINVAL;
L
Linus Torvalds 已提交
585

D
David Howells 已提交
586 587 588 589 590
	new = prepare_creds();
	if (!new)
		return -ENOMEM;

	old = current_cred();
L
Linus Torvalds 已提交
591

D
David Howells 已提交
592
	retval = -EPERM;
593
	if (!ns_capable(old->user_ns, CAP_SETUID)) {
594 595
		if (ruid != (uid_t) -1        && !uid_eq(kruid, old->uid) &&
		    !uid_eq(kruid, old->euid) && !uid_eq(kruid, old->suid))
D
David Howells 已提交
596
			goto error;
597 598
		if (euid != (uid_t) -1        && !uid_eq(keuid, old->uid) &&
		    !uid_eq(keuid, old->euid) && !uid_eq(keuid, old->suid))
D
David Howells 已提交
599
			goto error;
600 601
		if (suid != (uid_t) -1        && !uid_eq(ksuid, old->uid) &&
		    !uid_eq(ksuid, old->euid) && !uid_eq(ksuid, old->suid))
D
David Howells 已提交
602
			goto error;
L
Linus Torvalds 已提交
603
	}
D
David Howells 已提交
604

L
Linus Torvalds 已提交
605
	if (ruid != (uid_t) -1) {
606 607
		new->uid = kruid;
		if (!uid_eq(kruid, old->uid)) {
608 609 610 611
			retval = set_user(new);
			if (retval < 0)
				goto error;
		}
L
Linus Torvalds 已提交
612
	}
D
David Howells 已提交
613
	if (euid != (uid_t) -1)
614
		new->euid = keuid;
L
Linus Torvalds 已提交
615
	if (suid != (uid_t) -1)
616
		new->suid = ksuid;
D
David Howells 已提交
617
	new->fsuid = new->euid;
L
Linus Torvalds 已提交
618

D
David Howells 已提交
619 620 621
	retval = security_task_fix_setuid(new, old, LSM_SETID_RES);
	if (retval < 0)
		goto error;
L
Linus Torvalds 已提交
622

D
David Howells 已提交
623
	return commit_creds(new);
L
Linus Torvalds 已提交
624

D
David Howells 已提交
625 626 627
error:
	abort_creds(new);
	return retval;
L
Linus Torvalds 已提交
628 629
}

630
SYSCALL_DEFINE3(getresuid, uid_t __user *, ruidp, uid_t __user *, euidp, uid_t __user *, suidp)
L
Linus Torvalds 已提交
631
{
632
	const struct cred *cred = current_cred();
L
Linus Torvalds 已提交
633
	int retval;
634 635 636 637 638
	uid_t ruid, euid, suid;

	ruid = from_kuid_munged(cred->user_ns, cred->uid);
	euid = from_kuid_munged(cred->user_ns, cred->euid);
	suid = from_kuid_munged(cred->user_ns, cred->suid);
L
Linus Torvalds 已提交
639

640 641 642
	if (!(retval   = put_user(ruid, ruidp)) &&
	    !(retval   = put_user(euid, euidp)))
		retval = put_user(suid, suidp);
L
Linus Torvalds 已提交
643 644 645 646 647 648 649

	return retval;
}

/*
 * Same as above, but for rgid, egid, sgid.
 */
650
SYSCALL_DEFINE3(setresgid, gid_t, rgid, gid_t, egid, gid_t, sgid)
L
Linus Torvalds 已提交
651
{
652
	struct user_namespace *ns = current_user_ns();
D
David Howells 已提交
653 654
	const struct cred *old;
	struct cred *new;
L
Linus Torvalds 已提交
655
	int retval;
656 657 658 659 660 661 662 663 664 665 666 667
	kgid_t krgid, kegid, ksgid;

	krgid = make_kgid(ns, rgid);
	kegid = make_kgid(ns, egid);
	ksgid = make_kgid(ns, sgid);

	if ((rgid != (gid_t) -1) && !gid_valid(krgid))
		return -EINVAL;
	if ((egid != (gid_t) -1) && !gid_valid(kegid))
		return -EINVAL;
	if ((sgid != (gid_t) -1) && !gid_valid(ksgid))
		return -EINVAL;
L
Linus Torvalds 已提交
668

D
David Howells 已提交
669 670 671 672 673 674
	new = prepare_creds();
	if (!new)
		return -ENOMEM;
	old = current_cred();

	retval = -EPERM;
675
	if (!ns_capable(old->user_ns, CAP_SETGID)) {
676 677
		if (rgid != (gid_t) -1        && !gid_eq(krgid, old->gid) &&
		    !gid_eq(krgid, old->egid) && !gid_eq(krgid, old->sgid))
D
David Howells 已提交
678
			goto error;
679 680
		if (egid != (gid_t) -1        && !gid_eq(kegid, old->gid) &&
		    !gid_eq(kegid, old->egid) && !gid_eq(kegid, old->sgid))
D
David Howells 已提交
681
			goto error;
682 683
		if (sgid != (gid_t) -1        && !gid_eq(ksgid, old->gid) &&
		    !gid_eq(ksgid, old->egid) && !gid_eq(ksgid, old->sgid))
D
David Howells 已提交
684
			goto error;
L
Linus Torvalds 已提交
685
	}
D
David Howells 已提交
686

L
Linus Torvalds 已提交
687
	if (rgid != (gid_t) -1)
688
		new->gid = krgid;
D
David Howells 已提交
689
	if (egid != (gid_t) -1)
690
		new->egid = kegid;
L
Linus Torvalds 已提交
691
	if (sgid != (gid_t) -1)
692
		new->sgid = ksgid;
D
David Howells 已提交
693
	new->fsgid = new->egid;
L
Linus Torvalds 已提交
694

D
David Howells 已提交
695 696 697 698 699
	return commit_creds(new);

error:
	abort_creds(new);
	return retval;
L
Linus Torvalds 已提交
700 701
}

702
SYSCALL_DEFINE3(getresgid, gid_t __user *, rgidp, gid_t __user *, egidp, gid_t __user *, sgidp)
L
Linus Torvalds 已提交
703
{
704
	const struct cred *cred = current_cred();
L
Linus Torvalds 已提交
705
	int retval;
706 707 708 709 710
	gid_t rgid, egid, sgid;

	rgid = from_kgid_munged(cred->user_ns, cred->gid);
	egid = from_kgid_munged(cred->user_ns, cred->egid);
	sgid = from_kgid_munged(cred->user_ns, cred->sgid);
L
Linus Torvalds 已提交
711

712 713 714
	if (!(retval   = put_user(rgid, rgidp)) &&
	    !(retval   = put_user(egid, egidp)))
		retval = put_user(sgid, sgidp);
L
Linus Torvalds 已提交
715 716 717 718 719 720 721 722 723 724 725

	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..
 */
726
SYSCALL_DEFINE1(setfsuid, uid_t, uid)
L
Linus Torvalds 已提交
727
{
D
David Howells 已提交
728 729 730
	const struct cred *old;
	struct cred *new;
	uid_t old_fsuid;
731 732 733 734 735 736 737 738
	kuid_t kuid;

	old = current_cred();
	old_fsuid = from_kuid_munged(old->user_ns, old->fsuid);

	kuid = make_kuid(old->user_ns, uid);
	if (!uid_valid(kuid))
		return old_fsuid;
L
Linus Torvalds 已提交
739

D
David Howells 已提交
740 741
	new = prepare_creds();
	if (!new)
742
		return old_fsuid;
L
Linus Torvalds 已提交
743

744 745
	if (uid_eq(kuid, old->uid)  || uid_eq(kuid, old->euid)  ||
	    uid_eq(kuid, old->suid) || uid_eq(kuid, old->fsuid) ||
746
	    ns_capable(old->user_ns, CAP_SETUID)) {
747 748
		if (!uid_eq(kuid, old->fsuid)) {
			new->fsuid = kuid;
D
David Howells 已提交
749 750
			if (security_task_fix_setuid(new, old, LSM_SETID_FS) == 0)
				goto change_okay;
L
Linus Torvalds 已提交
751 752 753
		}
	}

D
David Howells 已提交
754 755
	abort_creds(new);
	return old_fsuid;
L
Linus Torvalds 已提交
756

D
David Howells 已提交
757 758
change_okay:
	commit_creds(new);
L
Linus Torvalds 已提交
759 760 761 762
	return old_fsuid;
}

/*
763
 * Samma på svenska..
L
Linus Torvalds 已提交
764
 */
765
SYSCALL_DEFINE1(setfsgid, gid_t, gid)
L
Linus Torvalds 已提交
766
{
D
David Howells 已提交
767 768 769
	const struct cred *old;
	struct cred *new;
	gid_t old_fsgid;
770 771 772 773 774 775 776 777
	kgid_t kgid;

	old = current_cred();
	old_fsgid = from_kgid_munged(old->user_ns, old->fsgid);

	kgid = make_kgid(old->user_ns, gid);
	if (!gid_valid(kgid))
		return old_fsgid;
D
David Howells 已提交
778 779 780

	new = prepare_creds();
	if (!new)
781
		return old_fsgid;
L
Linus Torvalds 已提交
782

783 784
	if (gid_eq(kgid, old->gid)  || gid_eq(kgid, old->egid)  ||
	    gid_eq(kgid, old->sgid) || gid_eq(kgid, old->fsgid) ||
785
	    ns_capable(old->user_ns, CAP_SETGID)) {
786 787
		if (!gid_eq(kgid, old->fsgid)) {
			new->fsgid = kgid;
D
David Howells 已提交
788
			goto change_okay;
L
Linus Torvalds 已提交
789 790
		}
	}
D
David Howells 已提交
791 792 793 794 795 796

	abort_creds(new);
	return old_fsgid;

change_okay:
	commit_creds(new);
L
Linus Torvalds 已提交
797 798 799
	return old_fsgid;
}

800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860
/**
 * sys_getpid - return the thread group id of the current process
 *
 * Note, despite the name, this returns the tgid not the pid.  The tgid and
 * the pid are identical unless CLONE_THREAD was specified on clone() in
 * which case the tgid is the same in all threads of the same group.
 *
 * This is SMP safe as current->tgid does not change.
 */
SYSCALL_DEFINE0(getpid)
{
	return task_tgid_vnr(current);
}

/* Thread ID - the internal kernel "pid" */
SYSCALL_DEFINE0(gettid)
{
	return task_pid_vnr(current);
}

/*
 * Accessing ->real_parent is not SMP-safe, it could
 * change from under us. However, we can use a stale
 * value of ->real_parent under rcu_read_lock(), see
 * release_task()->call_rcu(delayed_put_task_struct).
 */
SYSCALL_DEFINE0(getppid)
{
	int pid;

	rcu_read_lock();
	pid = task_tgid_vnr(rcu_dereference(current->real_parent));
	rcu_read_unlock();

	return pid;
}

SYSCALL_DEFINE0(getuid)
{
	/* Only we change this so SMP safe */
	return from_kuid_munged(current_user_ns(), current_uid());
}

SYSCALL_DEFINE0(geteuid)
{
	/* Only we change this so SMP safe */
	return from_kuid_munged(current_user_ns(), current_euid());
}

SYSCALL_DEFINE0(getgid)
{
	/* Only we change this so SMP safe */
	return from_kgid_munged(current_user_ns(), current_gid());
}

SYSCALL_DEFINE0(getegid)
{
	/* Only we change this so SMP safe */
	return from_kgid_munged(current_user_ns(), current_egid());
}

861 862
void do_sys_times(struct tms *tms)
{
863
	cputime_t tgutime, tgstime, cutime, cstime;
864

865
	spin_lock_irq(&current->sighand->siglock);
866
	thread_group_cputime_adjusted(current, &tgutime, &tgstime);
867 868 869
	cutime = current->signal->cutime;
	cstime = current->signal->cstime;
	spin_unlock_irq(&current->sighand->siglock);
870 871
	tms->tms_utime = cputime_to_clock_t(tgutime);
	tms->tms_stime = cputime_to_clock_t(tgstime);
872 873 874 875
	tms->tms_cutime = cputime_to_clock_t(cutime);
	tms->tms_cstime = cputime_to_clock_t(cstime);
}

876
SYSCALL_DEFINE1(times, struct tms __user *, tbuf)
L
Linus Torvalds 已提交
877 878 879
{
	if (tbuf) {
		struct tms tmp;
880 881

		do_sys_times(&tmp);
L
Linus Torvalds 已提交
882 883 884
		if (copy_to_user(tbuf, &tmp, sizeof(struct tms)))
			return -EFAULT;
	}
885
	force_successful_syscall_return();
L
Linus Torvalds 已提交
886 887 888 889 890 891 892 893 894 895 896 897
	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
 *
O
Oleg Nesterov 已提交
898
 * !PF_FORKNOEXEC check to conform completely to POSIX.
L
Linus Torvalds 已提交
899
 */
900
SYSCALL_DEFINE2(setpgid, pid_t, pid, pid_t, pgid)
L
Linus Torvalds 已提交
901 902
{
	struct task_struct *p;
903
	struct task_struct *group_leader = current->group_leader;
904 905
	struct pid *pgrp;
	int err;
L
Linus Torvalds 已提交
906 907

	if (!pid)
908
		pid = task_pid_vnr(group_leader);
L
Linus Torvalds 已提交
909 910 911 912
	if (!pgid)
		pgid = pid;
	if (pgid < 0)
		return -EINVAL;
913
	rcu_read_lock();
L
Linus Torvalds 已提交
914 915 916 917 918 919 920

	/* 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;
921
	p = find_task_by_vpid(pid);
L
Linus Torvalds 已提交
922 923 924 925 926 927 928
	if (!p)
		goto out;

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

929
	if (same_thread_group(p->real_parent, group_leader)) {
L
Linus Torvalds 已提交
930
		err = -EPERM;
931
		if (task_session(p) != task_session(group_leader))
L
Linus Torvalds 已提交
932 933
			goto out;
		err = -EACCES;
O
Oleg Nesterov 已提交
934
		if (!(p->flags & PF_FORKNOEXEC))
L
Linus Torvalds 已提交
935 936 937
			goto out;
	} else {
		err = -ESRCH;
938
		if (p != group_leader)
L
Linus Torvalds 已提交
939 940 941 942 943 944 945
			goto out;
	}

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

946
	pgrp = task_pid(p);
L
Linus Torvalds 已提交
947
	if (pgid != pid) {
948
		struct task_struct *g;
L
Linus Torvalds 已提交
949

950 951
		pgrp = find_vpid(pgid);
		g = pid_task(pgrp, PIDTYPE_PGID);
952
		if (!g || task_session(g) != task_session(group_leader))
953
			goto out;
L
Linus Torvalds 已提交
954 955 956 957 958 959
	}

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

960
	if (task_pgrp(p) != pgrp)
961
		change_pid(p, PIDTYPE_PGID, pgrp);
L
Linus Torvalds 已提交
962 963 964 965 966

	err = 0;
out:
	/* All paths lead to here, thus we are safe. -DaveM */
	write_unlock_irq(&tasklist_lock);
967
	rcu_read_unlock();
L
Linus Torvalds 已提交
968 969 970
	return err;
}

971
SYSCALL_DEFINE1(getpgid, pid_t, pid)
L
Linus Torvalds 已提交
972
{
973 974 975 976 977
	struct task_struct *p;
	struct pid *grp;
	int retval;

	rcu_read_lock();
978
	if (!pid)
979
		grp = task_pgrp(current);
980
	else {
L
Linus Torvalds 已提交
981
		retval = -ESRCH;
982 983 984 985 986 987 988 989 990 991
		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
992
	}
993 994 995 996
	retval = pid_vnr(grp);
out:
	rcu_read_unlock();
	return retval;
L
Linus Torvalds 已提交
997 998 999 1000
}

#ifdef __ARCH_WANT_SYS_GETPGRP

1001
SYSCALL_DEFINE0(getpgrp)
L
Linus Torvalds 已提交
1002
{
1003
	return sys_getpgid(0);
L
Linus Torvalds 已提交
1004 1005 1006 1007
}

#endif

1008
SYSCALL_DEFINE1(getsid, pid_t, pid)
L
Linus Torvalds 已提交
1009
{
1010 1011 1012 1013 1014
	struct task_struct *p;
	struct pid *sid;
	int retval;

	rcu_read_lock();
1015
	if (!pid)
1016
		sid = task_session(current);
1017
	else {
L
Linus Torvalds 已提交
1018
		retval = -ESRCH;
1019 1020 1021 1022 1023 1024 1025 1026 1027 1028
		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
1029
	}
1030 1031 1032 1033
	retval = pid_vnr(sid);
out:
	rcu_read_unlock();
	return retval;
L
Linus Torvalds 已提交
1034 1035
}

1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046
static void set_special_pids(struct pid *pid)
{
	struct task_struct *curr = current->group_leader;

	if (task_session(curr) != pid)
		change_pid(curr, PIDTYPE_SID, pid);

	if (task_pgrp(curr) != pid)
		change_pid(curr, PIDTYPE_PGID, pid);
}

1047
SYSCALL_DEFINE0(setsid)
L
Linus Torvalds 已提交
1048
{
1049
	struct task_struct *group_leader = current->group_leader;
1050 1051
	struct pid *sid = task_pid(group_leader);
	pid_t session = pid_vnr(sid);
L
Linus Torvalds 已提交
1052 1053 1054
	int err = -EPERM;

	write_lock_irq(&tasklist_lock);
1055 1056 1057 1058
	/* Fail if I am already a session leader */
	if (group_leader->signal->leader)
		goto out;

1059 1060
	/* Fail if a process group id already exists that equals the
	 * proposed session id.
1061
	 */
1062
	if (pid_task(sid, PIDTYPE_PGID))
L
Linus Torvalds 已提交
1063 1064
		goto out;

1065
	group_leader->signal->leader = 1;
1066
	set_special_pids(sid);
1067

A
Alan Cox 已提交
1068
	proc_clear_tty(group_leader);
1069

1070
	err = session;
L
Linus Torvalds 已提交
1071 1072
out:
	write_unlock_irq(&tasklist_lock);
1073
	if (err > 0) {
1074
		proc_sid_connector(group_leader);
1075 1076
		sched_autogroup_create_attach(group_leader);
	}
L
Linus Torvalds 已提交
1077 1078 1079 1080 1081
	return err;
}

DECLARE_RWSEM(uts_sem);

1082 1083
#ifdef COMPAT_UTS_MACHINE
#define override_architecture(name) \
1084
	(personality(current->personality) == PER_LINUX32 && \
1085 1086 1087 1088 1089 1090
	 copy_to_user(name->machine, COMPAT_UTS_MACHINE, \
		      sizeof(COMPAT_UTS_MACHINE)))
#else
#define override_architecture(name)	0
#endif

1091 1092 1093 1094
/*
 * Work around broken programs that cannot handle "Linux 3.0".
 * Instead we map 3.x to 2.6.40+x, so e.g. 3.0 would be 2.6.40
 */
1095
static int override_release(char __user *release, size_t len)
1096 1097 1098 1099
{
	int ret = 0;

	if (current->personality & UNAME26) {
1100 1101
		const char *rest = UTS_RELEASE;
		char buf[65] = { 0 };
1102 1103
		int ndots = 0;
		unsigned v;
1104
		size_t copy;
1105 1106 1107 1108 1109 1110 1111 1112 1113

		while (*rest) {
			if (*rest == '.' && ++ndots >= 3)
				break;
			if (!isdigit(*rest) && *rest != '.')
				break;
			rest++;
		}
		v = ((LINUX_VERSION_CODE >> 8) & 0xff) + 40;
K
Kees Cook 已提交
1114
		copy = clamp_t(size_t, len, 1, sizeof(buf));
1115 1116
		copy = scnprintf(buf, copy, "2.6.%u%s", v, rest);
		ret = copy_to_user(release, buf, copy + 1);
1117 1118 1119 1120
	}
	return ret;
}

1121
SYSCALL_DEFINE1(newuname, struct new_utsname __user *, name)
L
Linus Torvalds 已提交
1122 1123 1124 1125
{
	int errno = 0;

	down_read(&uts_sem);
1126
	if (copy_to_user(name, utsname(), sizeof *name))
L
Linus Torvalds 已提交
1127 1128
		errno = -EFAULT;
	up_read(&uts_sem);
1129

1130 1131
	if (!errno && override_release(name->release, sizeof(name->release)))
		errno = -EFAULT;
1132 1133
	if (!errno && override_architecture(name))
		errno = -EFAULT;
L
Linus Torvalds 已提交
1134 1135 1136
	return errno;
}

C
Christoph Hellwig 已提交
1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152
#ifdef __ARCH_WANT_SYS_OLD_UNAME
/*
 * Old cruft
 */
SYSCALL_DEFINE1(uname, struct old_utsname __user *, name)
{
	int error = 0;

	if (!name)
		return -EFAULT;

	down_read(&uts_sem);
	if (copy_to_user(name, utsname(), sizeof(*name)))
		error = -EFAULT;
	up_read(&uts_sem);

1153 1154
	if (!error && override_release(name->release, sizeof(name->release)))
		error = -EFAULT;
C
Christoph Hellwig 已提交
1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188
	if (!error && override_architecture(name))
		error = -EFAULT;
	return error;
}

SYSCALL_DEFINE1(olduname, struct oldold_utsname __user *, name)
{
	int error;

	if (!name)
		return -EFAULT;
	if (!access_ok(VERIFY_WRITE, name, sizeof(struct oldold_utsname)))
		return -EFAULT;

	down_read(&uts_sem);
	error = __copy_to_user(&name->sysname, &utsname()->sysname,
			       __OLD_UTS_LEN);
	error |= __put_user(0, name->sysname + __OLD_UTS_LEN);
	error |= __copy_to_user(&name->nodename, &utsname()->nodename,
				__OLD_UTS_LEN);
	error |= __put_user(0, name->nodename + __OLD_UTS_LEN);
	error |= __copy_to_user(&name->release, &utsname()->release,
				__OLD_UTS_LEN);
	error |= __put_user(0, name->release + __OLD_UTS_LEN);
	error |= __copy_to_user(&name->version, &utsname()->version,
				__OLD_UTS_LEN);
	error |= __put_user(0, name->version + __OLD_UTS_LEN);
	error |= __copy_to_user(&name->machine, &utsname()->machine,
				__OLD_UTS_LEN);
	error |= __put_user(0, name->machine + __OLD_UTS_LEN);
	up_read(&uts_sem);

	if (!error && override_architecture(name))
		error = -EFAULT;
1189 1190
	if (!error && override_release(name->release, sizeof(name->release)))
		error = -EFAULT;
C
Christoph Hellwig 已提交
1191 1192 1193 1194
	return error ? -EFAULT : 0;
}
#endif

1195
SYSCALL_DEFINE2(sethostname, char __user *, name, int, len)
L
Linus Torvalds 已提交
1196 1197 1198 1199
{
	int errno;
	char tmp[__NEW_UTS_LEN];

1200
	if (!ns_capable(current->nsproxy->uts_ns->user_ns, CAP_SYS_ADMIN))
L
Linus Torvalds 已提交
1201
		return -EPERM;
1202

L
Linus Torvalds 已提交
1203 1204 1205 1206 1207
	if (len < 0 || len > __NEW_UTS_LEN)
		return -EINVAL;
	down_write(&uts_sem);
	errno = -EFAULT;
	if (!copy_from_user(tmp, name, len)) {
1208 1209 1210 1211
		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
1212
		errno = 0;
1213
		uts_proc_notify(UTS_PROC_HOSTNAME);
L
Linus Torvalds 已提交
1214 1215 1216 1217 1218 1219 1220
	}
	up_write(&uts_sem);
	return errno;
}

#ifdef __ARCH_WANT_SYS_GETHOSTNAME

1221
SYSCALL_DEFINE2(gethostname, char __user *, name, int, len)
L
Linus Torvalds 已提交
1222 1223
{
	int i, errno;
1224
	struct new_utsname *u;
L
Linus Torvalds 已提交
1225 1226 1227 1228

	if (len < 0)
		return -EINVAL;
	down_read(&uts_sem);
1229 1230
	u = utsname();
	i = 1 + strlen(u->nodename);
L
Linus Torvalds 已提交
1231 1232 1233
	if (i > len)
		i = len;
	errno = 0;
1234
	if (copy_to_user(name, u->nodename, i))
L
Linus Torvalds 已提交
1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245
		errno = -EFAULT;
	up_read(&uts_sem);
	return errno;
}

#endif

/*
 * Only setdomainname; getdomainname can be implemented by calling
 * uname()
 */
1246
SYSCALL_DEFINE2(setdomainname, char __user *, name, int, len)
L
Linus Torvalds 已提交
1247 1248 1249 1250
{
	int errno;
	char tmp[__NEW_UTS_LEN];

1251
	if (!ns_capable(current->nsproxy->uts_ns->user_ns, CAP_SYS_ADMIN))
L
Linus Torvalds 已提交
1252 1253 1254 1255 1256 1257 1258
		return -EPERM;
	if (len < 0 || len > __NEW_UTS_LEN)
		return -EINVAL;

	down_write(&uts_sem);
	errno = -EFAULT;
	if (!copy_from_user(tmp, name, len)) {
1259 1260 1261 1262
		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
1263
		errno = 0;
1264
		uts_proc_notify(UTS_PROC_DOMAINNAME);
L
Linus Torvalds 已提交
1265 1266 1267 1268 1269
	}
	up_write(&uts_sem);
	return errno;
}

1270
SYSCALL_DEFINE2(getrlimit, unsigned int, resource, struct rlimit __user *, rlim)
L
Linus Torvalds 已提交
1271
{
1272 1273 1274 1275 1276 1277 1278 1279
	struct rlimit value;
	int ret;

	ret = do_prlimit(current, resource, NULL, &value);
	if (!ret)
		ret = copy_to_user(rlim, &value, sizeof(*rlim)) ? -EFAULT : 0;

	return ret;
L
Linus Torvalds 已提交
1280 1281 1282 1283 1284 1285 1286 1287
}

#ifdef __ARCH_WANT_SYS_OLD_GETRLIMIT

/*
 *	Back compatibility for getrlimit. Needed for some apps.
 */
 
1288 1289
SYSCALL_DEFINE2(old_getrlimit, unsigned int, resource,
		struct rlimit __user *, rlim)
L
Linus Torvalds 已提交
1290 1291 1292 1293 1294 1295 1296 1297
{
	struct rlimit x;
	if (resource >= RLIM_NLIMITS)
		return -EINVAL;

	task_lock(current->group_leader);
	x = current->signal->rlim[resource];
	task_unlock(current->group_leader);
1298
	if (x.rlim_cur > 0x7FFFFFFF)
L
Linus Torvalds 已提交
1299
		x.rlim_cur = 0x7FFFFFFF;
1300
	if (x.rlim_max > 0x7FFFFFFF)
L
Linus Torvalds 已提交
1301 1302 1303 1304 1305 1306
		x.rlim_max = 0x7FFFFFFF;
	return copy_to_user(rlim, &x, sizeof(x))?-EFAULT:0;
}

#endif

1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339
static inline bool rlim64_is_infinity(__u64 rlim64)
{
#if BITS_PER_LONG < 64
	return rlim64 >= ULONG_MAX;
#else
	return rlim64 == RLIM64_INFINITY;
#endif
}

static void rlim_to_rlim64(const struct rlimit *rlim, struct rlimit64 *rlim64)
{
	if (rlim->rlim_cur == RLIM_INFINITY)
		rlim64->rlim_cur = RLIM64_INFINITY;
	else
		rlim64->rlim_cur = rlim->rlim_cur;
	if (rlim->rlim_max == RLIM_INFINITY)
		rlim64->rlim_max = RLIM64_INFINITY;
	else
		rlim64->rlim_max = rlim->rlim_max;
}

static void rlim64_to_rlim(const struct rlimit64 *rlim64, struct rlimit *rlim)
{
	if (rlim64_is_infinity(rlim64->rlim_cur))
		rlim->rlim_cur = RLIM_INFINITY;
	else
		rlim->rlim_cur = (unsigned long)rlim64->rlim_cur;
	if (rlim64_is_infinity(rlim64->rlim_max))
		rlim->rlim_max = RLIM_INFINITY;
	else
		rlim->rlim_max = (unsigned long)rlim64->rlim_max;
}

1340
/* make sure you are allowed to change @tsk limits before calling this */
1341 1342
int do_prlimit(struct task_struct *tsk, unsigned int resource,
		struct rlimit *new_rlim, struct rlimit *old_rlim)
L
Linus Torvalds 已提交
1343
{
1344
	struct rlimit *rlim;
1345
	int retval = 0;
L
Linus Torvalds 已提交
1346 1347 1348

	if (resource >= RLIM_NLIMITS)
		return -EINVAL;
1349 1350 1351 1352 1353 1354 1355
	if (new_rlim) {
		if (new_rlim->rlim_cur > new_rlim->rlim_max)
			return -EINVAL;
		if (resource == RLIMIT_NOFILE &&
				new_rlim->rlim_max > sysctl_nr_open)
			return -EPERM;
	}
L
Linus Torvalds 已提交
1356

1357 1358 1359 1360 1361 1362 1363
	/* protect tsk->signal and tsk->sighand from disappearing */
	read_lock(&tasklist_lock);
	if (!tsk->sighand) {
		retval = -ESRCH;
		goto out;
	}

1364
	rlim = tsk->signal->rlim + resource;
1365
	task_lock(tsk->group_leader);
1366
	if (new_rlim) {
1367 1368
		/* Keep the capable check against init_user_ns until
		   cgroups can contain all limits */
1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389
		if (new_rlim->rlim_max > rlim->rlim_max &&
				!capable(CAP_SYS_RESOURCE))
			retval = -EPERM;
		if (!retval)
			retval = security_task_setrlimit(tsk->group_leader,
					resource, new_rlim);
		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;
		}
	}
	if (!retval) {
		if (old_rlim)
			*old_rlim = *rlim;
		if (new_rlim)
			*rlim = *new_rlim;
1390
	}
J
Jiri Slaby 已提交
1391
	task_unlock(tsk->group_leader);
L
Linus Torvalds 已提交
1392

1393 1394 1395 1396 1397 1398
	/*
	 * 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
	 */
1399 1400 1401
	 if (!retval && new_rlim && resource == RLIMIT_CPU &&
			 new_rlim->rlim_cur != RLIM_INFINITY)
		update_rlimit_cpu(tsk, new_rlim->rlim_cur);
A
Andrew Morton 已提交
1402
out:
1403
	read_unlock(&tasklist_lock);
1404
	return retval;
L
Linus Torvalds 已提交
1405 1406
}

1407 1408 1409 1410 1411
/* rcu lock must be held */
static int check_prlimit_permission(struct task_struct *task)
{
	const struct cred *cred = current_cred(), *tcred;

1412 1413
	if (current == task)
		return 0;
1414

1415
	tcred = __task_cred(task);
1416 1417 1418 1419 1420 1421
	if (uid_eq(cred->uid, tcred->euid) &&
	    uid_eq(cred->uid, tcred->suid) &&
	    uid_eq(cred->uid, tcred->uid)  &&
	    gid_eq(cred->gid, tcred->egid) &&
	    gid_eq(cred->gid, tcred->sgid) &&
	    gid_eq(cred->gid, tcred->gid))
1422
		return 0;
1423
	if (ns_capable(tcred->user_ns, CAP_SYS_RESOURCE))
1424 1425 1426
		return 0;

	return -EPERM;
1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470
}

SYSCALL_DEFINE4(prlimit64, pid_t, pid, unsigned int, resource,
		const struct rlimit64 __user *, new_rlim,
		struct rlimit64 __user *, old_rlim)
{
	struct rlimit64 old64, new64;
	struct rlimit old, new;
	struct task_struct *tsk;
	int ret;

	if (new_rlim) {
		if (copy_from_user(&new64, new_rlim, sizeof(new64)))
			return -EFAULT;
		rlim64_to_rlim(&new64, &new);
	}

	rcu_read_lock();
	tsk = pid ? find_task_by_vpid(pid) : current;
	if (!tsk) {
		rcu_read_unlock();
		return -ESRCH;
	}
	ret = check_prlimit_permission(tsk);
	if (ret) {
		rcu_read_unlock();
		return ret;
	}
	get_task_struct(tsk);
	rcu_read_unlock();

	ret = do_prlimit(tsk, resource, new_rlim ? &new : NULL,
			old_rlim ? &old : NULL);

	if (!ret && old_rlim) {
		rlim_to_rlim64(&old, &old64);
		if (copy_to_user(old_rlim, &old64, sizeof(old64)))
			ret = -EFAULT;
	}

	put_task_struct(tsk);
	return ret;
}

J
Jiri Slaby 已提交
1471 1472 1473 1474 1475 1476
SYSCALL_DEFINE2(setrlimit, unsigned int, resource, struct rlimit __user *, rlim)
{
	struct rlimit new_rlim;

	if (copy_from_user(&new_rlim, rlim, sizeof(*rlim)))
		return -EFAULT;
1477
	return do_prlimit(current, resource, &new_rlim, NULL);
J
Jiri Slaby 已提交
1478 1479
}

L
Linus Torvalds 已提交
1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494
/*
 * 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.
1495
 *
1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509
 * 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.
1510
 *
L
Linus Torvalds 已提交
1511 1512
 */

1513
static void accumulate_thread_rusage(struct task_struct *t, struct rusage *r)
S
Sripathi Kodi 已提交
1514 1515 1516 1517 1518 1519 1520 1521 1522
{
	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
1523 1524 1525 1526
static void k_getrusage(struct task_struct *p, int who, struct rusage *r)
{
	struct task_struct *t;
	unsigned long flags;
1527
	cputime_t tgutime, tgstime, utime, stime;
J
Jiri Pirko 已提交
1528
	unsigned long maxrss = 0;
L
Linus Torvalds 已提交
1529 1530

	memset((char *) r, 0, sizeof *r);
1531
	utime = stime = 0;
L
Linus Torvalds 已提交
1532

S
Sripathi Kodi 已提交
1533
	if (who == RUSAGE_THREAD) {
1534
		task_cputime_adjusted(current, &utime, &stime);
1535
		accumulate_thread_rusage(p, r);
J
Jiri Pirko 已提交
1536
		maxrss = p->signal->maxrss;
S
Sripathi Kodi 已提交
1537 1538 1539
		goto out;
	}

1540
	if (!lock_task_sighand(p, &flags))
1541
		return;
O
Oleg Nesterov 已提交
1542

L
Linus Torvalds 已提交
1543
	switch (who) {
O
Oleg Nesterov 已提交
1544
		case RUSAGE_BOTH:
L
Linus Torvalds 已提交
1545 1546 1547 1548 1549 1550 1551
		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;
1552 1553
			r->ru_inblock = p->signal->cinblock;
			r->ru_oublock = p->signal->coublock;
J
Jiri Pirko 已提交
1554
			maxrss = p->signal->cmaxrss;
O
Oleg Nesterov 已提交
1555 1556 1557 1558

			if (who == RUSAGE_CHILDREN)
				break;

L
Linus Torvalds 已提交
1559
		case RUSAGE_SELF:
1560
			thread_group_cputime_adjusted(p, &tgutime, &tgstime);
1561 1562
			utime += tgutime;
			stime += tgstime;
L
Linus Torvalds 已提交
1563 1564 1565 1566
			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;
1567 1568
			r->ru_inblock += p->signal->inblock;
			r->ru_oublock += p->signal->oublock;
J
Jiri Pirko 已提交
1569 1570
			if (maxrss < p->signal->maxrss)
				maxrss = p->signal->maxrss;
L
Linus Torvalds 已提交
1571 1572
			t = p;
			do {
1573
				accumulate_thread_rusage(t, r);
1574
			} while_each_thread(p, t);
L
Linus Torvalds 已提交
1575
			break;
O
Oleg Nesterov 已提交
1576

L
Linus Torvalds 已提交
1577 1578 1579
		default:
			BUG();
	}
1580 1581
	unlock_task_sighand(p, &flags);

S
Sripathi Kodi 已提交
1582
out:
O
Oleg Nesterov 已提交
1583 1584
	cputime_to_timeval(utime, &r->ru_utime);
	cputime_to_timeval(stime, &r->ru_stime);
J
Jiri Pirko 已提交
1585 1586 1587 1588 1589 1590 1591 1592 1593

	if (who != RUSAGE_CHILDREN) {
		struct mm_struct *mm = get_task_mm(p);
		if (mm) {
			setmax_mm_hiwater_rss(&maxrss, mm);
			mmput(mm);
		}
	}
	r->ru_maxrss = maxrss * (PAGE_SIZE / 1024); /* convert pages to KBs */
L
Linus Torvalds 已提交
1594 1595 1596 1597 1598 1599 1600 1601 1602
}

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;
}

1603
SYSCALL_DEFINE2(getrusage, int, who, struct rusage __user *, ru)
L
Linus Torvalds 已提交
1604
{
S
Sripathi Kodi 已提交
1605 1606
	if (who != RUSAGE_SELF && who != RUSAGE_CHILDREN &&
	    who != RUSAGE_THREAD)
L
Linus Torvalds 已提交
1607 1608 1609 1610
		return -EINVAL;
	return getrusage(current, who, ru);
}

1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624
#ifdef CONFIG_COMPAT
COMPAT_SYSCALL_DEFINE2(getrusage, int, who, struct compat_rusage __user *, ru)
{
	struct rusage r;

	if (who != RUSAGE_SELF && who != RUSAGE_CHILDREN &&
	    who != RUSAGE_THREAD)
		return -EINVAL;

	k_getrusage(current, who, &r);
	return put_compat_rusage(&r, ru);
}
#endif

1625
SYSCALL_DEFINE1(umask, int, mask)
L
Linus Torvalds 已提交
1626 1627 1628 1629
{
	mask = xchg(&current->fs->umask, mask & S_IRWXUGO);
	return mask;
}
1630

1631 1632
static int prctl_set_mm_exe_file(struct mm_struct *mm, unsigned int fd)
{
1633
	struct fd exe;
A
Al Viro 已提交
1634
	struct inode *inode;
1635
	int err;
1636

1637 1638
	exe = fdget(fd);
	if (!exe.file)
1639 1640
		return -EBADF;

A
Al Viro 已提交
1641
	inode = file_inode(exe.file);
1642 1643 1644 1645 1646 1647 1648

	/*
	 * Because the original mm->exe_file points to executable file, make
	 * sure that this one is executable as well, to avoid breaking an
	 * overall picture.
	 */
	err = -EACCES;
A
Al Viro 已提交
1649
	if (!S_ISREG(inode->i_mode)	||
1650
	    exe.file->f_path.mnt->mnt_flags & MNT_NOEXEC)
1651 1652
		goto exit;

A
Al Viro 已提交
1653
	err = inode_permission(inode, MAY_EXEC);
1654 1655 1656
	if (err)
		goto exit;

1657 1658 1659
	down_write(&mm->mmap_sem);

	/*
1660
	 * Forbid mm->exe_file change if old file still mapped.
1661 1662
	 */
	err = -EBUSY;
1663 1664 1665 1666 1667 1668 1669 1670
	if (mm->exe_file) {
		struct vm_area_struct *vma;

		for (vma = mm->mmap; vma; vma = vma->vm_next)
			if (vma->vm_file &&
			    path_equal(&vma->vm_file->f_path,
				       &mm->exe_file->f_path))
				goto exit_unlock;
1671 1672
	}

1673 1674 1675 1676 1677 1678
	/*
	 * The symlink can be changed only once, just to disallow arbitrary
	 * transitions malicious software might bring in. This means one
	 * could make a snapshot over all processes running and monitor
	 * /proc/pid/exe changes to notice unusual activity if needed.
	 */
1679 1680 1681 1682
	err = -EPERM;
	if (test_and_set_bit(MMF_EXE_FILE_CHANGED, &mm->flags))
		goto exit_unlock;

1683
	err = 0;
1684
	set_mm_exe_file(mm, exe.file);	/* this grabs a reference to exe.file */
1685
exit_unlock:
1686 1687 1688
	up_write(&mm->mmap_sem);

exit:
1689
	fdput(exe);
1690 1691 1692
	return err;
}

1693 1694 1695 1696 1697
static int prctl_set_mm(int opt, unsigned long addr,
			unsigned long arg4, unsigned long arg5)
{
	unsigned long rlim = rlimit(RLIMIT_DATA);
	struct mm_struct *mm = current->mm;
1698 1699
	struct vm_area_struct *vma;
	int error;
1700

1701
	if (arg5 || (arg4 && opt != PR_SET_MM_AUXV))
1702 1703
		return -EINVAL;

1704
	if (!capable(CAP_SYS_RESOURCE))
1705 1706
		return -EPERM;

1707 1708 1709
	if (opt == PR_SET_MM_EXE_FILE)
		return prctl_set_mm_exe_file(mm, (unsigned int)addr);

1710
	if (addr >= TASK_SIZE || addr < mmap_min_addr)
1711 1712
		return -EINVAL;

1713 1714
	error = -EINVAL;

1715 1716 1717 1718 1719
	down_read(&mm->mmap_sem);
	vma = find_vma(mm, addr);

	switch (opt) {
	case PR_SET_MM_START_CODE:
1720 1721
		mm->start_code = addr;
		break;
1722
	case PR_SET_MM_END_CODE:
1723
		mm->end_code = addr;
1724 1725
		break;
	case PR_SET_MM_START_DATA:
1726
		mm->start_data = addr;
1727
		break;
1728 1729
	case PR_SET_MM_END_DATA:
		mm->end_data = addr;
1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755
		break;

	case PR_SET_MM_START_BRK:
		if (addr <= mm->end_data)
			goto out;

		if (rlim < RLIM_INFINITY &&
		    (mm->brk - addr) +
		    (mm->end_data - mm->start_data) > rlim)
			goto out;

		mm->start_brk = addr;
		break;

	case PR_SET_MM_BRK:
		if (addr <= mm->end_data)
			goto out;

		if (rlim < RLIM_INFINITY &&
		    (addr - mm->start_brk) +
		    (mm->end_data - mm->start_data) > rlim)
			goto out;

		mm->brk = addr;
		break;

1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813
	/*
	 * If command line arguments and environment
	 * are placed somewhere else on stack, we can
	 * set them up here, ARG_START/END to setup
	 * command line argumets and ENV_START/END
	 * for environment.
	 */
	case PR_SET_MM_START_STACK:
	case PR_SET_MM_ARG_START:
	case PR_SET_MM_ARG_END:
	case PR_SET_MM_ENV_START:
	case PR_SET_MM_ENV_END:
		if (!vma) {
			error = -EFAULT;
			goto out;
		}
		if (opt == PR_SET_MM_START_STACK)
			mm->start_stack = addr;
		else if (opt == PR_SET_MM_ARG_START)
			mm->arg_start = addr;
		else if (opt == PR_SET_MM_ARG_END)
			mm->arg_end = addr;
		else if (opt == PR_SET_MM_ENV_START)
			mm->env_start = addr;
		else if (opt == PR_SET_MM_ENV_END)
			mm->env_end = addr;
		break;

	/*
	 * This doesn't move auxiliary vector itself
	 * since it's pinned to mm_struct, but allow
	 * to fill vector with new values. It's up
	 * to a caller to provide sane values here
	 * otherwise user space tools which use this
	 * vector might be unhappy.
	 */
	case PR_SET_MM_AUXV: {
		unsigned long user_auxv[AT_VECTOR_SIZE];

		if (arg4 > sizeof(user_auxv))
			goto out;
		up_read(&mm->mmap_sem);

		if (copy_from_user(user_auxv, (const void __user *)addr, arg4))
			return -EFAULT;

		/* Make sure the last entry is always AT_NULL */
		user_auxv[AT_VECTOR_SIZE - 2] = 0;
		user_auxv[AT_VECTOR_SIZE - 1] = 0;

		BUILD_BUG_ON(sizeof(user_auxv) != sizeof(mm->saved_auxv));

		task_lock(current);
		memcpy(mm->saved_auxv, user_auxv, arg4);
		task_unlock(current);

		return 0;
	}
1814 1815 1816 1817 1818 1819 1820 1821 1822
	default:
		goto out;
	}

	error = 0;
out:
	up_read(&mm->mmap_sem);
	return error;
}
1823

1824
#ifdef CONFIG_CHECKPOINT_RESTORE
1825 1826 1827 1828
static int prctl_get_tid_address(struct task_struct *me, int __user **tid_addr)
{
	return put_user(me->clear_child_tid, tid_addr);
}
1829
#else
1830 1831 1832 1833
static int prctl_get_tid_address(struct task_struct *me, int __user **tid_addr)
{
	return -EINVAL;
}
1834 1835
#endif

1836 1837
SYSCALL_DEFINE5(prctl, int, option, unsigned long, arg2, unsigned long, arg3,
		unsigned long, arg4, unsigned long, arg5)
L
Linus Torvalds 已提交
1838
{
1839 1840 1841
	struct task_struct *me = current;
	unsigned char comm[sizeof(me->comm)];
	long error;
L
Linus Torvalds 已提交
1842

D
David Howells 已提交
1843 1844
	error = security_task_prctl(option, arg2, arg3, arg4, arg5);
	if (error != -ENOSYS)
L
Linus Torvalds 已提交
1845 1846
		return error;

D
David Howells 已提交
1847
	error = 0;
L
Linus Torvalds 已提交
1848
	switch (option) {
1849 1850 1851
	case PR_SET_PDEATHSIG:
		if (!valid_signal(arg2)) {
			error = -EINVAL;
L
Linus Torvalds 已提交
1852
			break;
1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864
		}
		me->pdeath_signal = arg2;
		break;
	case PR_GET_PDEATHSIG:
		error = put_user(me->pdeath_signal, (int __user *)arg2);
		break;
	case PR_GET_DUMPABLE:
		error = get_dumpable(me->mm);
		break;
	case PR_SET_DUMPABLE:
		if (arg2 != SUID_DUMP_DISABLE && arg2 != SUID_DUMP_USER) {
			error = -EINVAL;
L
Linus Torvalds 已提交
1865
			break;
1866 1867 1868
		}
		set_dumpable(me->mm, arg2);
		break;
L
Linus Torvalds 已提交
1869

1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937
	case PR_SET_UNALIGN:
		error = SET_UNALIGN_CTL(me, arg2);
		break;
	case PR_GET_UNALIGN:
		error = GET_UNALIGN_CTL(me, arg2);
		break;
	case PR_SET_FPEMU:
		error = SET_FPEMU_CTL(me, arg2);
		break;
	case PR_GET_FPEMU:
		error = GET_FPEMU_CTL(me, arg2);
		break;
	case PR_SET_FPEXC:
		error = SET_FPEXC_CTL(me, arg2);
		break;
	case PR_GET_FPEXC:
		error = GET_FPEXC_CTL(me, arg2);
		break;
	case PR_GET_TIMING:
		error = PR_TIMING_STATISTICAL;
		break;
	case PR_SET_TIMING:
		if (arg2 != PR_TIMING_STATISTICAL)
			error = -EINVAL;
		break;
	case PR_SET_NAME:
		comm[sizeof(me->comm) - 1] = 0;
		if (strncpy_from_user(comm, (char __user *)arg2,
				      sizeof(me->comm) - 1) < 0)
			return -EFAULT;
		set_task_comm(me, comm);
		proc_comm_connector(me);
		break;
	case PR_GET_NAME:
		get_task_comm(comm, me);
		if (copy_to_user((char __user *)arg2, comm, sizeof(comm)))
			return -EFAULT;
		break;
	case PR_GET_ENDIAN:
		error = GET_ENDIAN(me, arg2);
		break;
	case PR_SET_ENDIAN:
		error = SET_ENDIAN(me, arg2);
		break;
	case PR_GET_SECCOMP:
		error = prctl_get_seccomp();
		break;
	case PR_SET_SECCOMP:
		error = prctl_set_seccomp(arg2, (char __user *)arg3);
		break;
	case PR_GET_TSC:
		error = GET_TSC_CTL(arg2);
		break;
	case PR_SET_TSC:
		error = SET_TSC_CTL(arg2);
		break;
	case PR_TASK_PERF_EVENTS_DISABLE:
		error = perf_event_task_disable();
		break;
	case PR_TASK_PERF_EVENTS_ENABLE:
		error = perf_event_task_enable();
		break;
	case PR_GET_TIMERSLACK:
		error = current->timer_slack_ns;
		break;
	case PR_SET_TIMERSLACK:
		if (arg2 <= 0)
			current->timer_slack_ns =
1938
					current->default_timer_slack_ns;
1939 1940 1941 1942 1943 1944 1945 1946 1947
		else
			current->timer_slack_ns = arg2;
		break;
	case PR_MCE_KILL:
		if (arg4 | arg5)
			return -EINVAL;
		switch (arg2) {
		case PR_MCE_KILL_CLEAR:
			if (arg3 != 0)
1948
				return -EINVAL;
1949
			current->flags &= ~PF_MCE_PROCESS;
1950
			break;
1951 1952 1953 1954 1955 1956 1957 1958 1959
		case PR_MCE_KILL_SET:
			current->flags |= PF_MCE_PROCESS;
			if (arg3 == PR_MCE_KILL_EARLY)
				current->flags |= PF_MCE_EARLY;
			else if (arg3 == PR_MCE_KILL_LATE)
				current->flags &= ~PF_MCE_EARLY;
			else if (arg3 == PR_MCE_KILL_DEFAULT)
				current->flags &=
						~(PF_MCE_EARLY|PF_MCE_PROCESS);
1960
			else
1961 1962
				return -EINVAL;
			break;
L
Linus Torvalds 已提交
1963
		default:
1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992
			return -EINVAL;
		}
		break;
	case PR_MCE_KILL_GET:
		if (arg2 | arg3 | arg4 | arg5)
			return -EINVAL;
		if (current->flags & PF_MCE_PROCESS)
			error = (current->flags & PF_MCE_EARLY) ?
				PR_MCE_KILL_EARLY : PR_MCE_KILL_LATE;
		else
			error = PR_MCE_KILL_DEFAULT;
		break;
	case PR_SET_MM:
		error = prctl_set_mm(arg2, arg3, arg4, arg5);
		break;
	case PR_GET_TID_ADDRESS:
		error = prctl_get_tid_address(me, (int __user **)arg2);
		break;
	case PR_SET_CHILD_SUBREAPER:
		me->signal->is_child_subreaper = !!arg2;
		break;
	case PR_GET_CHILD_SUBREAPER:
		error = put_user(me->signal->is_child_subreaper,
				 (int __user *)arg2);
		break;
	case PR_SET_NO_NEW_PRIVS:
		if (arg2 != 1 || arg3 || arg4 || arg5)
			return -EINVAL;

1993
		task_set_no_new_privs(current);
1994 1995 1996 1997
		break;
	case PR_GET_NO_NEW_PRIVS:
		if (arg2 || arg3 || arg4 || arg5)
			return -EINVAL;
1998
		return task_no_new_privs(current) ? 1 : 0;
1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
	case PR_GET_THP_DISABLE:
		if (arg2 || arg3 || arg4 || arg5)
			return -EINVAL;
		error = !!(me->mm->def_flags & VM_NOHUGEPAGE);
		break;
	case PR_SET_THP_DISABLE:
		if (arg3 || arg4 || arg5)
			return -EINVAL;
		down_write(&me->mm->mmap_sem);
		if (arg2)
			me->mm->def_flags |= VM_NOHUGEPAGE;
		else
			me->mm->def_flags &= ~VM_NOHUGEPAGE;
		up_write(&me->mm->mmap_sem);
		break;
2014 2015 2016
	default:
		error = -EINVAL;
		break;
L
Linus Torvalds 已提交
2017 2018 2019
	}
	return error;
}
2020

2021 2022
SYSCALL_DEFINE3(getcpu, unsigned __user *, cpup, unsigned __user *, nodep,
		struct getcpu_cache __user *, unused)
2023 2024 2025 2026 2027 2028 2029 2030 2031
{
	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;
}
2032

2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044
/**
 * do_sysinfo - fill in sysinfo struct
 * @info: pointer to buffer to fill
 */
static int do_sysinfo(struct sysinfo *info)
{
	unsigned long mem_total, sav_total;
	unsigned int mem_unit, bitcount;
	struct timespec tp;

	memset(info, 0, sizeof(struct sysinfo));

2045
	get_monotonic_boottime(&tp);
2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175
	info->uptime = tp.tv_sec + (tp.tv_nsec ? 1 : 0);

	get_avenrun(info->loads, 0, SI_LOAD_SHIFT - FSHIFT);

	info->procs = nr_threads;

	si_meminfo(info);
	si_swapinfo(info);

	/*
	 * If the sum of all the available memory (i.e. ram + swap)
	 * is less than can be stored in a 32 bit unsigned long then
	 * we can be binary compatible with 2.2.x kernels.  If not,
	 * well, in that case 2.2.x was broken anyways...
	 *
	 *  -Erik Andersen <andersee@debian.org>
	 */

	mem_total = info->totalram + info->totalswap;
	if (mem_total < info->totalram || mem_total < info->totalswap)
		goto out;
	bitcount = 0;
	mem_unit = info->mem_unit;
	while (mem_unit > 1) {
		bitcount++;
		mem_unit >>= 1;
		sav_total = mem_total;
		mem_total <<= 1;
		if (mem_total < sav_total)
			goto out;
	}

	/*
	 * If mem_total did not overflow, multiply all memory values by
	 * info->mem_unit and set it to 1.  This leaves things compatible
	 * with 2.2.x, and also retains compatibility with earlier 2.4.x
	 * kernels...
	 */

	info->mem_unit = 1;
	info->totalram <<= bitcount;
	info->freeram <<= bitcount;
	info->sharedram <<= bitcount;
	info->bufferram <<= bitcount;
	info->totalswap <<= bitcount;
	info->freeswap <<= bitcount;
	info->totalhigh <<= bitcount;
	info->freehigh <<= bitcount;

out:
	return 0;
}

SYSCALL_DEFINE1(sysinfo, struct sysinfo __user *, info)
{
	struct sysinfo val;

	do_sysinfo(&val);

	if (copy_to_user(info, &val, sizeof(struct sysinfo)))
		return -EFAULT;

	return 0;
}

#ifdef CONFIG_COMPAT
struct compat_sysinfo {
	s32 uptime;
	u32 loads[3];
	u32 totalram;
	u32 freeram;
	u32 sharedram;
	u32 bufferram;
	u32 totalswap;
	u32 freeswap;
	u16 procs;
	u16 pad;
	u32 totalhigh;
	u32 freehigh;
	u32 mem_unit;
	char _f[20-2*sizeof(u32)-sizeof(int)];
};

COMPAT_SYSCALL_DEFINE1(sysinfo, struct compat_sysinfo __user *, info)
{
	struct sysinfo s;

	do_sysinfo(&s);

	/* Check to see if any memory value is too large for 32-bit and scale
	 *  down if needed
	 */
	if ((s.totalram >> 32) || (s.totalswap >> 32)) {
		int bitcount = 0;

		while (s.mem_unit < PAGE_SIZE) {
			s.mem_unit <<= 1;
			bitcount++;
		}

		s.totalram >>= bitcount;
		s.freeram >>= bitcount;
		s.sharedram >>= bitcount;
		s.bufferram >>= bitcount;
		s.totalswap >>= bitcount;
		s.freeswap >>= bitcount;
		s.totalhigh >>= bitcount;
		s.freehigh >>= bitcount;
	}

	if (!access_ok(VERIFY_WRITE, info, sizeof(struct compat_sysinfo)) ||
	    __put_user(s.uptime, &info->uptime) ||
	    __put_user(s.loads[0], &info->loads[0]) ||
	    __put_user(s.loads[1], &info->loads[1]) ||
	    __put_user(s.loads[2], &info->loads[2]) ||
	    __put_user(s.totalram, &info->totalram) ||
	    __put_user(s.freeram, &info->freeram) ||
	    __put_user(s.sharedram, &info->sharedram) ||
	    __put_user(s.bufferram, &info->bufferram) ||
	    __put_user(s.totalswap, &info->totalswap) ||
	    __put_user(s.freeswap, &info->freeswap) ||
	    __put_user(s.procs, &info->procs) ||
	    __put_user(s.totalhigh, &info->totalhigh) ||
	    __put_user(s.freehigh, &info->freehigh) ||
	    __put_user(s.mem_unit, &info->mem_unit))
		return -EFAULT;

	return 0;
}
#endif /* CONFIG_COMPAT */