namei.c 73.3 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 17 18 19 20 21 22 23
/*
 *  linux/fs/namei.c
 *
 *  Copyright (C) 1991, 1992  Linus Torvalds
 */

/*
 * Some corrections by tytso.
 */

/* [Feb 1997 T. Schoebel-Theuer] Complete rewrite of the pathname
 * lookup logic.
 */
/* [Feb-Apr 2000, AV] Rewrite to the new namespace architecture.
 */

#include <linux/init.h>
#include <linux/module.h>
#include <linux/slab.h>
#include <linux/fs.h>
#include <linux/namei.h>
#include <linux/quotaops.h>
#include <linux/pagemap.h>
R
Robert Love 已提交
24
#include <linux/fsnotify.h>
L
Linus Torvalds 已提交
25 26 27 28 29
#include <linux/personality.h>
#include <linux/security.h>
#include <linux/syscalls.h>
#include <linux/mount.h>
#include <linux/audit.h>
30
#include <linux/capability.h>
31
#include <linux/file.h>
32
#include <linux/fcntl.h>
33
#include <linux/device_cgroup.h>
L
Linus Torvalds 已提交
34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 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
#include <asm/namei.h>
#include <asm/uaccess.h>

#define ACC_MODE(x) ("\000\004\002\006"[(x)&O_ACCMODE])

/* [Feb-1997 T. Schoebel-Theuer]
 * Fundamental changes in the pathname lookup mechanisms (namei)
 * were necessary because of omirr.  The reason is that omirr needs
 * to know the _real_ pathname, not the user-supplied one, in case
 * of symlinks (and also when transname replacements occur).
 *
 * The new code replaces the old recursive symlink resolution with
 * an iterative one (in case of non-nested symlink chains).  It does
 * this with calls to <fs>_follow_link().
 * As a side effect, dir_namei(), _namei() and follow_link() are now 
 * replaced with a single function lookup_dentry() that can handle all 
 * the special cases of the former code.
 *
 * With the new dcache, the pathname is stored at each inode, at least as
 * long as the refcount of the inode is positive.  As a side effect, the
 * size of the dcache depends on the inode cache and thus is dynamic.
 *
 * [29-Apr-1998 C. Scott Ananian] Updated above description of symlink
 * resolution to correspond with current state of the code.
 *
 * Note that the symlink resolution is not *completely* iterative.
 * There is still a significant amount of tail- and mid- recursion in
 * the algorithm.  Also, note that <fs>_readlink() is not used in
 * lookup_dentry(): lookup_dentry() on the result of <fs>_readlink()
 * may return different results than <fs>_follow_link().  Many virtual
 * filesystems (including /proc) exhibit this behavior.
 */

/* [24-Feb-97 T. Schoebel-Theuer] Side effects caused by new implementation:
 * New symlink semantics: when open() is called with flags O_CREAT | O_EXCL
 * and the name already exists in form of a symlink, try to create the new
 * name indicated by the symlink. The old code always complained that the
 * name already exists, due to not following the symlink even if its target
 * is nonexistent.  The new semantics affects also mknod() and link() when
 * the name is a symlink pointing to a non-existant name.
 *
 * I don't know which semantics is the right one, since I have no access
 * to standards. But I found by trial that HP-UX 9.0 has the full "new"
 * semantics implemented, while SunOS 4.1.1 and Solaris (SunOS 5.4) have the
 * "old" one. Personally, I think the new semantics is much more logical.
 * Note that "ln old new" where "new" is a symlink pointing to a non-existing
 * file does succeed in both HP-UX and SunOs, but not in Solaris
 * and in the old Linux semantics.
 */

/* [16-Dec-97 Kevin Buhr] For security reasons, we change some symlink
 * semantics.  See the comments in "open_namei" and "do_link" below.
 *
 * [10-Sep-98 Alan Modra] Another symlink change.
 */

/* [Feb-Apr 2000 AV] Complete rewrite. Rules for symlinks:
 *	inside the path - always follow.
 *	in the last component in creation/removal/renaming - never follow.
 *	if LOOKUP_FOLLOW passed - follow.
 *	if the pathname has trailing slashes - follow.
 *	otherwise - don't follow.
 * (applied in that order).
 *
 * [Jun 2000 AV] Inconsistent behaviour of open() in case if flags==O_CREAT
 * restored for 2.4. This is the last surviving part of old 4.2BSD bug.
 * During the 2.4 we need to fix the userland stuff depending on it -
 * hopefully we will be able to get rid of that wart in 2.5. So far only
 * XEmacs seems to be relying on it...
 */
/*
 * [Sep 2001 AV] Single-semaphore locking scheme (kudos to David Holland)
106
 * implemented.  Let's see if raised priority of ->s_vfs_rename_mutex gives
L
Linus Torvalds 已提交
107 108 109
 * any extra contention...
 */

110
static int __link_path_walk(const char *name, struct nameidata *nd);
111

L
Linus Torvalds 已提交
112 113 114 115 116 117 118
/* In order to reduce some races, while at the same time doing additional
 * checking and hopefully speeding things up, we copy filenames to the
 * kernel data space before using them..
 *
 * POSIX.1 2.4: an empty pathname is invalid (ENOENT).
 * PATH_MAX includes the nul terminator --RR.
 */
119
static int do_getname(const char __user *filename, char *page)
L
Linus Torvalds 已提交
120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162
{
	int retval;
	unsigned long len = PATH_MAX;

	if (!segment_eq(get_fs(), KERNEL_DS)) {
		if ((unsigned long) filename >= TASK_SIZE)
			return -EFAULT;
		if (TASK_SIZE - (unsigned long) filename < PATH_MAX)
			len = TASK_SIZE - (unsigned long) filename;
	}

	retval = strncpy_from_user(page, filename, len);
	if (retval > 0) {
		if (retval < len)
			return 0;
		return -ENAMETOOLONG;
	} else if (!retval)
		retval = -ENOENT;
	return retval;
}

char * getname(const char __user * filename)
{
	char *tmp, *result;

	result = ERR_PTR(-ENOMEM);
	tmp = __getname();
	if (tmp)  {
		int retval = do_getname(filename, tmp);

		result = tmp;
		if (retval < 0) {
			__putname(tmp);
			result = ERR_PTR(retval);
		}
	}
	audit_getname(result);
	return result;
}

#ifdef CONFIG_AUDITSYSCALL
void putname(const char *name)
{
163
	if (unlikely(!audit_dummy_context()))
L
Linus Torvalds 已提交
164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187
		audit_putname(name);
	else
		__putname(name);
}
EXPORT_SYMBOL(putname);
#endif


/**
 * generic_permission  -  check for access rights on a Posix-like filesystem
 * @inode:	inode to check access rights for
 * @mask:	right to check for (%MAY_READ, %MAY_WRITE, %MAY_EXEC)
 * @check_acl:	optional callback to check for Posix ACLs
 *
 * Used to check for read/write/execute permissions on a file.
 * We use "fsuid" for this, letting us set arbitrary permissions
 * for filesystem access without changing the "normal" uids which
 * are used for other things..
 */
int generic_permission(struct inode *inode, int mask,
		int (*check_acl)(struct inode *inode, int mask))
{
	umode_t			mode = inode->i_mode;

188 189
	mask &= MAY_READ | MAY_WRITE | MAY_EXEC;

L
Linus Torvalds 已提交
190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207
	if (current->fsuid == inode->i_uid)
		mode >>= 6;
	else {
		if (IS_POSIXACL(inode) && (mode & S_IRWXG) && check_acl) {
			int error = check_acl(inode, mask);
			if (error == -EACCES)
				goto check_capabilities;
			else if (error != -EAGAIN)
				return error;
		}

		if (in_group_p(inode->i_gid))
			mode >>= 3;
	}

	/*
	 * If the DACs are ok we don't need any capability check.
	 */
208
	if ((mask & ~mode) == 0)
L
Linus Torvalds 已提交
209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232
		return 0;

 check_capabilities:
	/*
	 * Read/write DACs are always overridable.
	 * Executable DACs are overridable if at least one exec bit is set.
	 */
	if (!(mask & MAY_EXEC) ||
	    (inode->i_mode & S_IXUGO) || S_ISDIR(inode->i_mode))
		if (capable(CAP_DAC_OVERRIDE))
			return 0;

	/*
	 * Searching includes executable on directories, else just read.
	 */
	if (mask == MAY_READ || (S_ISDIR(inode->i_mode) && !(mask & MAY_WRITE)))
		if (capable(CAP_DAC_READ_SEARCH))
			return 0;

	return -EACCES;
}

int permission(struct inode *inode, int mask, struct nameidata *nd)
{
233
	int retval;
234 235 236
	struct vfsmount *mnt = NULL;

	if (nd)
237
		mnt = nd->path.mnt;
L
Linus Torvalds 已提交
238 239

	if (mask & MAY_WRITE) {
240
		umode_t mode = inode->i_mode;
L
Linus Torvalds 已提交
241 242 243 244 245 246 247 248 249 250 251 252 253 254 255

		/*
		 * Nobody gets write access to a read-only fs.
		 */
		if (IS_RDONLY(inode) &&
		    (S_ISREG(mode) || S_ISDIR(mode) || S_ISLNK(mode)))
			return -EROFS;

		/*
		 * Nobody gets write access to an immutable file.
		 */
		if (IS_IMMUTABLE(inode))
			return -EACCES;
	}

256 257 258 259 260
	if ((mask & MAY_EXEC) && S_ISREG(inode->i_mode)) {
		/*
		 * MAY_EXEC on regular files is denied if the fs is mounted
		 * with the "noexec" flag.
		 */
261
		if (mnt && (mnt->mnt_flags & MNT_NOEXEC))
262 263
			return -EACCES;
	}
264

L
Linus Torvalds 已提交
265
	/* Ordinary permission routines do not understand MAY_APPEND. */
266
	if (inode->i_op && inode->i_op->permission) {
267 268 269 270 271 272 273 274 275 276
		int extra = 0;
		if (nd) {
			if (nd->flags & LOOKUP_ACCESS)
				extra |= MAY_ACCESS;
			if (nd->flags & LOOKUP_CHDIR)
				extra |= MAY_CHDIR;
			if (nd->flags & LOOKUP_OPEN)
				extra |= MAY_OPEN;
		}
		retval = inode->i_op->permission(inode, mask | extra);
277 278 279 280 281 282 283 284 285 286 287 288 289
		if (!retval) {
			/*
			 * Exec permission on a regular file is denied if none
			 * of the execute bits are set.
			 *
			 * This check should be done by the ->permission()
			 * method.
			 */
			if ((mask & MAY_EXEC) && S_ISREG(inode->i_mode) &&
			    !(inode->i_mode & S_IXUGO))
				return -EACCES;
		}
	} else {
290
		retval = generic_permission(inode, mask, NULL);
291
	}
L
Linus Torvalds 已提交
292 293 294
	if (retval)
		return retval;

295 296 297 298
	retval = devcgroup_inode_permission(inode, mask);
	if (retval)
		return retval;

299 300
	return security_inode_permission(inode,
			mask & (MAY_READ|MAY_WRITE|MAY_EXEC), nd);
L
Linus Torvalds 已提交
301 302
}

303 304 305 306 307 308 309 310 311 312 313 314
/**
 * vfs_permission  -  check for access rights to a given path
 * @nd:		lookup result that describes the path
 * @mask:	right to check for (%MAY_READ, %MAY_WRITE, %MAY_EXEC)
 *
 * Used to check for read/write/execute permissions on a path.
 * We use "fsuid" for this, letting us set arbitrary permissions
 * for filesystem access without changing the "normal" uids which
 * are used for other things.
 */
int vfs_permission(struct nameidata *nd, int mask)
{
315
	return permission(nd->path.dentry->d_inode, mask, nd);
316 317
}

318 319 320 321 322 323 324 325 326 327 328 329 330 331
/**
 * file_permission  -  check for additional access rights to a given file
 * @file:	file to check access rights for
 * @mask:	right to check for (%MAY_READ, %MAY_WRITE, %MAY_EXEC)
 *
 * Used to check for read/write/execute permissions on an already opened
 * file.
 *
 * Note:
 *	Do not use this function in new code.  All access checks should
 *	be done using vfs_permission().
 */
int file_permission(struct file *file, int mask)
{
332
	return permission(file->f_path.dentry->d_inode, mask, NULL);
333 334
}

L
Linus Torvalds 已提交
335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367
/*
 * get_write_access() gets write permission for a file.
 * put_write_access() releases this write permission.
 * This is used for regular files.
 * We cannot support write (and maybe mmap read-write shared) accesses and
 * MAP_DENYWRITE mmappings simultaneously. The i_writecount field of an inode
 * can have the following values:
 * 0: no writers, no VM_DENYWRITE mappings
 * < 0: (-i_writecount) vm_area_structs with VM_DENYWRITE set exist
 * > 0: (i_writecount) users are writing to the file.
 *
 * Normally we operate on that counter with atomic_{inc,dec} and it's safe
 * except for the cases where we don't hold i_writecount yet. Then we need to
 * use {get,deny}_write_access() - these functions check the sign and refuse
 * to do the change if sign is wrong. Exclusion between them is provided by
 * the inode->i_lock spinlock.
 */

int get_write_access(struct inode * inode)
{
	spin_lock(&inode->i_lock);
	if (atomic_read(&inode->i_writecount) < 0) {
		spin_unlock(&inode->i_lock);
		return -ETXTBSY;
	}
	atomic_inc(&inode->i_writecount);
	spin_unlock(&inode->i_lock);

	return 0;
}

int deny_write_access(struct file * file)
{
368
	struct inode *inode = file->f_path.dentry->d_inode;
L
Linus Torvalds 已提交
369 370 371 372 373 374 375 376 377 378 379 380

	spin_lock(&inode->i_lock);
	if (atomic_read(&inode->i_writecount) > 0) {
		spin_unlock(&inode->i_lock);
		return -ETXTBSY;
	}
	atomic_dec(&inode->i_writecount);
	spin_unlock(&inode->i_lock);

	return 0;
}

J
Jan Blunck 已提交
381 382 383 384 385 386 387 388 389 390 391 392 393
/**
 * path_get - get a reference to a path
 * @path: path to get the reference to
 *
 * Given a path increment the reference count to the dentry and the vfsmount.
 */
void path_get(struct path *path)
{
	mntget(path->mnt);
	dget(path->dentry);
}
EXPORT_SYMBOL(path_get);

J
Jan Blunck 已提交
394 395 396 397 398 399 400
/**
 * path_put - put a reference to a path
 * @path: path to put the reference to
 *
 * Given a path decrement the reference count to the dentry and the vfsmount.
 */
void path_put(struct path *path)
L
Linus Torvalds 已提交
401
{
J
Jan Blunck 已提交
402 403
	dput(path->dentry);
	mntput(path->mnt);
L
Linus Torvalds 已提交
404
}
J
Jan Blunck 已提交
405
EXPORT_SYMBOL(path_put);
L
Linus Torvalds 已提交
406

407 408 409 410 411 412
/**
 * release_open_intent - free up open intent resources
 * @nd: pointer to nameidata
 */
void release_open_intent(struct nameidata *nd)
{
413
	if (nd->intent.open.file->f_path.dentry == NULL)
414 415 416 417 418
		put_filp(nd->intent.open.file);
	else
		fput(nd->intent.open.file);
}

419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442
static inline struct dentry *
do_revalidate(struct dentry *dentry, struct nameidata *nd)
{
	int status = dentry->d_op->d_revalidate(dentry, nd);
	if (unlikely(status <= 0)) {
		/*
		 * The dentry failed validation.
		 * If d_revalidate returned 0 attempt to invalidate
		 * the dentry otherwise d_revalidate is asking us
		 * to return a fail status.
		 */
		if (!status) {
			if (!d_invalidate(dentry)) {
				dput(dentry);
				dentry = NULL;
			}
		} else {
			dput(dentry);
			dentry = ERR_PTR(status);
		}
	}
	return dentry;
}

L
Linus Torvalds 已提交
443 444 445 446 447 448 449 450 451 452 453 454 455 456
/*
 * Internal lookup() using the new generic dcache.
 * SMP-safe
 */
static struct dentry * cached_lookup(struct dentry * parent, struct qstr * name, struct nameidata *nd)
{
	struct dentry * dentry = __d_lookup(parent, name);

	/* lockess __d_lookup may fail due to concurrent d_move() 
	 * in some unrelated directory, so try with d_lookup
	 */
	if (!dentry)
		dentry = d_lookup(parent, name);

457 458 459
	if (dentry && dentry->d_op && dentry->d_op->d_revalidate)
		dentry = do_revalidate(dentry, nd);

L
Linus Torvalds 已提交
460 461 462 463 464 465 466 467 468 469 470 471 472
	return dentry;
}

/*
 * Short-cut version of permission(), for calling by
 * path_walk(), when dcache lock is held.  Combines parts
 * of permission() and generic_permission(), and tests ONLY for
 * MAY_EXEC permission.
 *
 * If appropriate, check DAC only.  If not appropriate, or
 * short-cut DAC fails, then call permission() to do more
 * complete permission check.
 */
473
static int exec_permission_lite(struct inode *inode,
L
Linus Torvalds 已提交
474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515
				       struct nameidata *nd)
{
	umode_t	mode = inode->i_mode;

	if (inode->i_op && inode->i_op->permission)
		return -EAGAIN;

	if (current->fsuid == inode->i_uid)
		mode >>= 6;
	else if (in_group_p(inode->i_gid))
		mode >>= 3;

	if (mode & MAY_EXEC)
		goto ok;

	if ((inode->i_mode & S_IXUGO) && capable(CAP_DAC_OVERRIDE))
		goto ok;

	if (S_ISDIR(inode->i_mode) && capable(CAP_DAC_OVERRIDE))
		goto ok;

	if (S_ISDIR(inode->i_mode) && capable(CAP_DAC_READ_SEARCH))
		goto ok;

	return -EACCES;
ok:
	return security_inode_permission(inode, MAY_EXEC, nd);
}

/*
 * This is called when everything else fails, and we actually have
 * to go to the low-level filesystem to find out what we should do..
 *
 * We get the directory semaphore, and after getting that we also
 * make sure that nobody added the entry to the dcache in the meantime..
 * SMP-safe
 */
static struct dentry * real_lookup(struct dentry * parent, struct qstr * name, struct nameidata *nd)
{
	struct dentry * result;
	struct inode *dir = parent->d_inode;

516
	mutex_lock(&dir->i_mutex);
L
Linus Torvalds 已提交
517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532
	/*
	 * First re-do the cached lookup just in case it was created
	 * while we waited for the directory semaphore..
	 *
	 * FIXME! This could use version numbering or similar to
	 * avoid unnecessary cache lookups.
	 *
	 * The "dcache_lock" is purely to protect the RCU list walker
	 * from concurrent renames at this point (we mustn't get false
	 * negatives from the RCU list walk here, unlike the optimistic
	 * fast walk).
	 *
	 * so doing d_lookup() (with seqlock), instead of lockfree __d_lookup
	 */
	result = d_lookup(parent, name);
	if (!result) {
533 534 535 536 537 538 539 540
		struct dentry *dentry;

		/* Don't create child dentry for a dead directory. */
		result = ERR_PTR(-ENOENT);
		if (IS_DEADDIR(dir))
			goto out_unlock;

		dentry = d_alloc(parent, name);
L
Linus Torvalds 已提交
541 542 543 544 545 546 547 548
		result = ERR_PTR(-ENOMEM);
		if (dentry) {
			result = dir->i_op->lookup(dir, dentry, nd);
			if (result)
				dput(dentry);
			else
				result = dentry;
		}
549
out_unlock:
550
		mutex_unlock(&dir->i_mutex);
L
Linus Torvalds 已提交
551 552 553 554 555 556 557
		return result;
	}

	/*
	 * Uhhuh! Nasty case: the cache was re-populated while
	 * we waited on the semaphore. Need to revalidate.
	 */
558
	mutex_unlock(&dir->i_mutex);
L
Linus Torvalds 已提交
559
	if (result->d_op && result->d_op->d_revalidate) {
560 561
		result = do_revalidate(result, nd);
		if (!result)
L
Linus Torvalds 已提交
562 563 564 565 566 567 568 569
			result = ERR_PTR(-ENOENT);
	}
	return result;
}

static int __emul_lookup_dentry(const char *, struct nameidata *);

/* SMP-safe */
570
static __always_inline int
L
Linus Torvalds 已提交
571 572
walk_init_root(const char *name, struct nameidata *nd)
{
573 574 575
	struct fs_struct *fs = current->fs;

	read_lock(&fs->lock);
J
Jan Blunck 已提交
576 577 578
	if (fs->altroot.dentry && !(nd->flags & LOOKUP_NOALT)) {
		nd->path = fs->altroot;
		path_get(&fs->altroot);
579
		read_unlock(&fs->lock);
L
Linus Torvalds 已提交
580 581
		if (__emul_lookup_dentry(name,nd))
			return 0;
582
		read_lock(&fs->lock);
L
Linus Torvalds 已提交
583
	}
J
Jan Blunck 已提交
584 585
	nd->path = fs->root;
	path_get(&fs->root);
586
	read_unlock(&fs->lock);
L
Linus Torvalds 已提交
587 588 589
	return 1;
}

590 591 592 593 594 595 596 597 598 599 600 601 602
/*
 * Wrapper to retry pathname resolution whenever the underlying
 * file system returns an ESTALE.
 *
 * Retry the whole path once, forcing real lookup requests
 * instead of relying on the dcache.
 */
static __always_inline int link_path_walk(const char *name, struct nameidata *nd)
{
	struct path save = nd->path;
	int result;

	/* make sure the stuff we saved doesn't go away */
603
	path_get(&save);
604 605 606 607 608

	result = __link_path_walk(name, nd);
	if (result == -ESTALE) {
		/* nd->path had been dropped */
		nd->path = save;
609
		path_get(&nd->path);
610 611 612 613 614 615 616 617 618
		nd->flags |= LOOKUP_REVAL;
		result = __link_path_walk(name, nd);
	}

	path_put(&save);

	return result;
}

619
static __always_inline int __vfs_follow_link(struct nameidata *nd, const char *link)
L
Linus Torvalds 已提交
620 621 622 623 624 625 626
{
	int res = 0;
	char *name;
	if (IS_ERR(link))
		goto fail;

	if (*link == '/') {
J
Jan Blunck 已提交
627
		path_put(&nd->path);
L
Linus Torvalds 已提交
628 629 630 631 632 633 634 635 636 637 638 639 640 641 642
		if (!walk_init_root(link, nd))
			/* weird __emul_prefix() stuff did it */
			goto out;
	}
	res = link_path_walk(link, nd);
out:
	if (nd->depth || res || nd->last_type!=LAST_NORM)
		return res;
	/*
	 * If it is an iterative symlinks resolution in open_namei() we
	 * have to copy the last component. And all that crap because of
	 * bloody create() on broken symlinks. Furrfu...
	 */
	name = __getname();
	if (unlikely(!name)) {
J
Jan Blunck 已提交
643
		path_put(&nd->path);
L
Linus Torvalds 已提交
644 645 646 647 648 649
		return -ENOMEM;
	}
	strcpy(name, nd->last.name);
	nd->last.name = name;
	return 0;
fail:
J
Jan Blunck 已提交
650
	path_put(&nd->path);
L
Linus Torvalds 已提交
651 652 653
	return PTR_ERR(link);
}

J
Jan Blunck 已提交
654
static void path_put_conditional(struct path *path, struct nameidata *nd)
655 656
{
	dput(path->dentry);
657
	if (path->mnt != nd->path.mnt)
658 659 660 661 662
		mntput(path->mnt);
}

static inline void path_to_nameidata(struct path *path, struct nameidata *nd)
{
663 664 665 666 667
	dput(nd->path.dentry);
	if (nd->path.mnt != path->mnt)
		mntput(nd->path.mnt);
	nd->path.mnt = path->mnt;
	nd->path.dentry = path->dentry;
668 669
}

670
static __always_inline int __do_follow_link(struct path *path, struct nameidata *nd)
L
Linus Torvalds 已提交
671 672
{
	int error;
673
	void *cookie;
A
Al Viro 已提交
674
	struct dentry *dentry = path->dentry;
L
Linus Torvalds 已提交
675

A
Al Viro 已提交
676
	touch_atime(path->mnt, dentry);
L
Linus Torvalds 已提交
677
	nd_set_link(nd, NULL);
A
Al Viro 已提交
678

679
	if (path->mnt != nd->path.mnt) {
680 681 682 683
		path_to_nameidata(path, nd);
		dget(dentry);
	}
	mntget(path->mnt);
684 685 686
	cookie = dentry->d_inode->i_op->follow_link(dentry, nd);
	error = PTR_ERR(cookie);
	if (!IS_ERR(cookie)) {
L
Linus Torvalds 已提交
687
		char *s = nd_get_link(nd);
688
		error = 0;
L
Linus Torvalds 已提交
689 690 691
		if (s)
			error = __vfs_follow_link(nd, s);
		if (dentry->d_inode->i_op->put_link)
692
			dentry->d_inode->i_op->put_link(dentry, nd, cookie);
L
Linus Torvalds 已提交
693
	}
694
	path_put(path);
L
Linus Torvalds 已提交
695 696 697 698 699 700 701 702 703 704 705

	return error;
}

/*
 * This limits recursive symlink follows to 8, while
 * limiting consecutive symlinks to 40.
 *
 * Without that kind of total limit, nasty chains of consecutive
 * symlinks can cause almost arbitrarily long lookups. 
 */
A
Al Viro 已提交
706
static inline int do_follow_link(struct path *path, struct nameidata *nd)
L
Linus Torvalds 已提交
707 708 709 710 711 712 713 714
{
	int err = -ELOOP;
	if (current->link_count >= MAX_NESTED_LINKS)
		goto loop;
	if (current->total_link_count >= 40)
		goto loop;
	BUG_ON(nd->depth >= MAX_NESTED_LINKS);
	cond_resched();
A
Al Viro 已提交
715
	err = security_inode_follow_link(path->dentry, nd);
L
Linus Torvalds 已提交
716 717 718 719 720
	if (err)
		goto loop;
	current->link_count++;
	current->total_link_count++;
	nd->depth++;
A
Al Viro 已提交
721
	err = __do_follow_link(path, nd);
A
Al Viro 已提交
722 723
	current->link_count--;
	nd->depth--;
L
Linus Torvalds 已提交
724 725
	return err;
loop:
J
Jan Blunck 已提交
726 727
	path_put_conditional(path, nd);
	path_put(&nd->path);
L
Linus Torvalds 已提交
728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753
	return err;
}

int follow_up(struct vfsmount **mnt, struct dentry **dentry)
{
	struct vfsmount *parent;
	struct dentry *mountpoint;
	spin_lock(&vfsmount_lock);
	parent=(*mnt)->mnt_parent;
	if (parent == *mnt) {
		spin_unlock(&vfsmount_lock);
		return 0;
	}
	mntget(parent);
	mountpoint=dget((*mnt)->mnt_mountpoint);
	spin_unlock(&vfsmount_lock);
	dput(*dentry);
	*dentry = mountpoint;
	mntput(*mnt);
	*mnt = parent;
	return 1;
}

/* no need for dcache_lock, as serialization is taken care in
 * namespace.c
 */
A
Al Viro 已提交
754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770
static int __follow_mount(struct path *path)
{
	int res = 0;
	while (d_mountpoint(path->dentry)) {
		struct vfsmount *mounted = lookup_mnt(path->mnt, path->dentry);
		if (!mounted)
			break;
		dput(path->dentry);
		if (res)
			mntput(path->mnt);
		path->mnt = mounted;
		path->dentry = dget(mounted->mnt_root);
		res = 1;
	}
	return res;
}

A
Al Viro 已提交
771
static void follow_mount(struct vfsmount **mnt, struct dentry **dentry)
L
Linus Torvalds 已提交
772 773 774 775 776
{
	while (d_mountpoint(*dentry)) {
		struct vfsmount *mounted = lookup_mnt(*mnt, *dentry);
		if (!mounted)
			break;
A
Al Viro 已提交
777
		dput(*dentry);
L
Linus Torvalds 已提交
778 779 780 781 782 783 784 785 786
		mntput(*mnt);
		*mnt = mounted;
		*dentry = dget(mounted->mnt_root);
	}
}

/* no need for dcache_lock, as serialization is taken care in
 * namespace.c
 */
A
Al Viro 已提交
787
int follow_down(struct vfsmount **mnt, struct dentry **dentry)
L
Linus Torvalds 已提交
788 789 790 791 792
{
	struct vfsmount *mounted;

	mounted = lookup_mnt(*mnt, *dentry);
	if (mounted) {
A
Al Viro 已提交
793
		dput(*dentry);
L
Linus Torvalds 已提交
794 795 796 797 798 799 800 801
		mntput(*mnt);
		*mnt = mounted;
		*dentry = dget(mounted->mnt_root);
		return 1;
	}
	return 0;
}

802
static __always_inline void follow_dotdot(struct nameidata *nd)
L
Linus Torvalds 已提交
803
{
804 805
	struct fs_struct *fs = current->fs;

L
Linus Torvalds 已提交
806 807
	while(1) {
		struct vfsmount *parent;
808
		struct dentry *old = nd->path.dentry;
L
Linus Torvalds 已提交
809

810
                read_lock(&fs->lock);
J
Jan Blunck 已提交
811 812
		if (nd->path.dentry == fs->root.dentry &&
		    nd->path.mnt == fs->root.mnt) {
813
                        read_unlock(&fs->lock);
L
Linus Torvalds 已提交
814 815
			break;
		}
816
                read_unlock(&fs->lock);
L
Linus Torvalds 已提交
817
		spin_lock(&dcache_lock);
818 819
		if (nd->path.dentry != nd->path.mnt->mnt_root) {
			nd->path.dentry = dget(nd->path.dentry->d_parent);
L
Linus Torvalds 已提交
820 821 822 823 824 825
			spin_unlock(&dcache_lock);
			dput(old);
			break;
		}
		spin_unlock(&dcache_lock);
		spin_lock(&vfsmount_lock);
826 827
		parent = nd->path.mnt->mnt_parent;
		if (parent == nd->path.mnt) {
L
Linus Torvalds 已提交
828 829 830 831
			spin_unlock(&vfsmount_lock);
			break;
		}
		mntget(parent);
832
		nd->path.dentry = dget(nd->path.mnt->mnt_mountpoint);
L
Linus Torvalds 已提交
833 834
		spin_unlock(&vfsmount_lock);
		dput(old);
835 836
		mntput(nd->path.mnt);
		nd->path.mnt = parent;
L
Linus Torvalds 已提交
837
	}
838
	follow_mount(&nd->path.mnt, &nd->path.dentry);
L
Linus Torvalds 已提交
839 840 841 842 843 844 845 846 847 848
}

/*
 *  It's more convoluted than I'd like it to be, but... it's still fairly
 *  small and for now I'd prefer to have fast path as straight as possible.
 *  It _is_ time-critical.
 */
static int do_lookup(struct nameidata *nd, struct qstr *name,
		     struct path *path)
{
849 850
	struct vfsmount *mnt = nd->path.mnt;
	struct dentry *dentry = __d_lookup(nd->path.dentry, name);
L
Linus Torvalds 已提交
851 852 853 854 855 856 857 858

	if (!dentry)
		goto need_lookup;
	if (dentry->d_op && dentry->d_op->d_revalidate)
		goto need_revalidate;
done:
	path->mnt = mnt;
	path->dentry = dentry;
A
Al Viro 已提交
859
	__follow_mount(path);
L
Linus Torvalds 已提交
860 861 862
	return 0;

need_lookup:
863
	dentry = real_lookup(nd->path.dentry, name, nd);
L
Linus Torvalds 已提交
864 865 866 867 868
	if (IS_ERR(dentry))
		goto fail;
	goto done;

need_revalidate:
869 870 871 872 873 874
	dentry = do_revalidate(dentry, nd);
	if (!dentry)
		goto need_lookup;
	if (IS_ERR(dentry))
		goto fail;
	goto done;
L
Linus Torvalds 已提交
875 876 877 878 879 880 881

fail:
	return PTR_ERR(dentry);
}

/*
 * Name resolution.
882 883
 * This is the basic name resolution function, turning a pathname into
 * the final dentry. We expect 'base' to be positive and a directory.
L
Linus Torvalds 已提交
884
 *
885 886
 * Returns 0 and nd will have valid dentry and mnt on success.
 * Returns error and drops reference to input namei data on failure.
L
Linus Torvalds 已提交
887
 */
888
static int __link_path_walk(const char *name, struct nameidata *nd)
L
Linus Torvalds 已提交
889 890 891 892 893 894 895 896 897 898 899
{
	struct path next;
	struct inode *inode;
	int err;
	unsigned int lookup_flags = nd->flags;
	
	while (*name=='/')
		name++;
	if (!*name)
		goto return_reval;

900
	inode = nd->path.dentry->d_inode;
L
Linus Torvalds 已提交
901
	if (nd->depth)
902
		lookup_flags = LOOKUP_FOLLOW | (nd->flags & LOOKUP_CONTINUE);
L
Linus Torvalds 已提交
903 904 905 906 907 908 909

	/* At this point we know we have a real path component. */
	for(;;) {
		unsigned long hash;
		struct qstr this;
		unsigned int c;

910
		nd->flags |= LOOKUP_CONTINUE;
L
Linus Torvalds 已提交
911
		err = exec_permission_lite(inode, nd);
912 913
		if (err == -EAGAIN)
			err = vfs_permission(nd, MAY_EXEC);
L
Linus Torvalds 已提交
914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946
 		if (err)
			break;

		this.name = name;
		c = *(const unsigned char *)name;

		hash = init_name_hash();
		do {
			name++;
			hash = partial_name_hash(c, hash);
			c = *(const unsigned char *)name;
		} while (c && (c != '/'));
		this.len = name - (const char *) this.name;
		this.hash = end_name_hash(hash);

		/* remove trailing slashes? */
		if (!c)
			goto last_component;
		while (*++name == '/');
		if (!*name)
			goto last_with_slashes;

		/*
		 * "." and ".." are special - ".." especially so because it has
		 * to be able to know about the current root directory and
		 * parent relationships.
		 */
		if (this.name[0] == '.') switch (this.len) {
			default:
				break;
			case 2:	
				if (this.name[1] != '.')
					break;
A
Al Viro 已提交
947
				follow_dotdot(nd);
948
				inode = nd->path.dentry->d_inode;
L
Linus Torvalds 已提交
949 950 951 952 953 954 955 956
				/* fallthrough */
			case 1:
				continue;
		}
		/*
		 * See if the low-level filesystem might want
		 * to use its own hash..
		 */
957 958 959
		if (nd->path.dentry->d_op && nd->path.dentry->d_op->d_hash) {
			err = nd->path.dentry->d_op->d_hash(nd->path.dentry,
							    &this);
L
Linus Torvalds 已提交
960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976
			if (err < 0)
				break;
		}
		/* This does the actual lookups.. */
		err = do_lookup(nd, &this, &next);
		if (err)
			break;

		err = -ENOENT;
		inode = next.dentry->d_inode;
		if (!inode)
			goto out_dput;
		err = -ENOTDIR; 
		if (!inode->i_op)
			goto out_dput;

		if (inode->i_op->follow_link) {
A
Al Viro 已提交
977
			err = do_follow_link(&next, nd);
L
Linus Torvalds 已提交
978 979 980
			if (err)
				goto return_err;
			err = -ENOENT;
981
			inode = nd->path.dentry->d_inode;
L
Linus Torvalds 已提交
982 983 984 985 986
			if (!inode)
				break;
			err = -ENOTDIR; 
			if (!inode->i_op)
				break;
M
Miklos Szeredi 已提交
987 988
		} else
			path_to_nameidata(&next, nd);
L
Linus Torvalds 已提交
989 990 991 992 993 994 995 996 997
		err = -ENOTDIR; 
		if (!inode->i_op->lookup)
			break;
		continue;
		/* here ends the main loop */

last_with_slashes:
		lookup_flags |= LOOKUP_FOLLOW | LOOKUP_DIRECTORY;
last_component:
998 999
		/* Clear LOOKUP_CONTINUE iff it was previously unset */
		nd->flags &= lookup_flags | ~LOOKUP_CONTINUE;
L
Linus Torvalds 已提交
1000 1001 1002 1003 1004 1005 1006 1007
		if (lookup_flags & LOOKUP_PARENT)
			goto lookup_parent;
		if (this.name[0] == '.') switch (this.len) {
			default:
				break;
			case 2:	
				if (this.name[1] != '.')
					break;
A
Al Viro 已提交
1008
				follow_dotdot(nd);
1009
				inode = nd->path.dentry->d_inode;
L
Linus Torvalds 已提交
1010 1011 1012 1013
				/* fallthrough */
			case 1:
				goto return_reval;
		}
1014 1015 1016
		if (nd->path.dentry->d_op && nd->path.dentry->d_op->d_hash) {
			err = nd->path.dentry->d_op->d_hash(nd->path.dentry,
							    &this);
L
Linus Torvalds 已提交
1017 1018 1019 1020 1021 1022 1023 1024 1025
			if (err < 0)
				break;
		}
		err = do_lookup(nd, &this, &next);
		if (err)
			break;
		inode = next.dentry->d_inode;
		if ((lookup_flags & LOOKUP_FOLLOW)
		    && inode && inode->i_op && inode->i_op->follow_link) {
A
Al Viro 已提交
1026
			err = do_follow_link(&next, nd);
L
Linus Torvalds 已提交
1027 1028
			if (err)
				goto return_err;
1029
			inode = nd->path.dentry->d_inode;
M
Miklos Szeredi 已提交
1030 1031
		} else
			path_to_nameidata(&next, nd);
L
Linus Torvalds 已提交
1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056
		err = -ENOENT;
		if (!inode)
			break;
		if (lookup_flags & LOOKUP_DIRECTORY) {
			err = -ENOTDIR; 
			if (!inode->i_op || !inode->i_op->lookup)
				break;
		}
		goto return_base;
lookup_parent:
		nd->last = this;
		nd->last_type = LAST_NORM;
		if (this.name[0] != '.')
			goto return_base;
		if (this.len == 1)
			nd->last_type = LAST_DOT;
		else if (this.len == 2 && this.name[1] == '.')
			nd->last_type = LAST_DOTDOT;
		else
			goto return_base;
return_reval:
		/*
		 * We bypassed the ordinary revalidation routines.
		 * We may need to check the cached dentry for staleness.
		 */
1057 1058
		if (nd->path.dentry && nd->path.dentry->d_sb &&
		    (nd->path.dentry->d_sb->s_type->fs_flags & FS_REVAL_DOT)) {
L
Linus Torvalds 已提交
1059 1060
			err = -ESTALE;
			/* Note: we do not d_invalidate() */
1061 1062
			if (!nd->path.dentry->d_op->d_revalidate(
					nd->path.dentry, nd))
L
Linus Torvalds 已提交
1063 1064 1065 1066 1067
				break;
		}
return_base:
		return 0;
out_dput:
J
Jan Blunck 已提交
1068
		path_put_conditional(&next, nd);
L
Linus Torvalds 已提交
1069 1070
		break;
	}
J
Jan Blunck 已提交
1071
	path_put(&nd->path);
L
Linus Torvalds 已提交
1072 1073 1074 1075
return_err:
	return err;
}

1076
static int path_walk(const char *name, struct nameidata *nd)
L
Linus Torvalds 已提交
1077 1078 1079 1080 1081
{
	current->total_link_count = 0;
	return link_path_walk(name, nd);
}

1082 1083 1084 1085
/* 
 * SMP-safe: Returns 1 and nd will have valid dentry and mnt, if
 * everything is done. Returns 0 and drops input nd, if lookup failed;
 */
L
Linus Torvalds 已提交
1086 1087 1088 1089 1090
static int __emul_lookup_dentry(const char *name, struct nameidata *nd)
{
	if (path_walk(name, nd))
		return 0;		/* something went wrong... */

1091 1092
	if (!nd->path.dentry->d_inode ||
	    S_ISDIR(nd->path.dentry->d_inode->i_mode)) {
1093
		struct path old_path = nd->path;
L
Linus Torvalds 已提交
1094 1095
		struct qstr last = nd->last;
		int last_type = nd->last_type;
1096 1097
		struct fs_struct *fs = current->fs;

L
Linus Torvalds 已提交
1098
		/*
1099 1100
		 * NAME was not found in alternate root or it's a directory.
		 * Try to find it in the normal root:
L
Linus Torvalds 已提交
1101 1102
		 */
		nd->last_type = LAST_ROOT;
1103
		read_lock(&fs->lock);
J
Jan Blunck 已提交
1104 1105
		nd->path = fs->root;
		path_get(&fs->root);
1106
		read_unlock(&fs->lock);
L
Linus Torvalds 已提交
1107
		if (path_walk(name, nd) == 0) {
1108
			if (nd->path.dentry->d_inode) {
1109
				path_put(&old_path);
L
Linus Torvalds 已提交
1110 1111
				return 1;
			}
J
Jan Blunck 已提交
1112
			path_put(&nd->path);
L
Linus Torvalds 已提交
1113
		}
1114
		nd->path = old_path;
L
Linus Torvalds 已提交
1115 1116 1117 1118 1119 1120 1121 1122 1123 1124
		nd->last = last;
		nd->last_type = last_type;
	}
	return 1;
}

void set_fs_altroot(void)
{
	char *emul = __emul_prefix();
	struct nameidata nd;
J
Jan Blunck 已提交
1125
	struct path path = {}, old_path;
L
Linus Torvalds 已提交
1126
	int err;
1127
	struct fs_struct *fs = current->fs;
L
Linus Torvalds 已提交
1128 1129 1130 1131

	if (!emul)
		goto set_it;
	err = path_lookup(emul, LOOKUP_FOLLOW|LOOKUP_DIRECTORY|LOOKUP_NOALT, &nd);
J
Jan Blunck 已提交
1132 1133
	if (!err)
		path = nd.path;
L
Linus Torvalds 已提交
1134
set_it:
1135
	write_lock(&fs->lock);
J
Jan Blunck 已提交
1136 1137
	old_path = fs->altroot;
	fs->altroot = path;
1138
	write_unlock(&fs->lock);
J
Jan Blunck 已提交
1139 1140
	if (old_path.dentry)
		path_put(&old_path);
L
Linus Torvalds 已提交
1141 1142
}

1143
/* Returns 0 and nd will be valid on success; Retuns error, otherwise. */
1144
static int do_path_lookup(int dfd, const char *name,
1145
				unsigned int flags, struct nameidata *nd)
L
Linus Torvalds 已提交
1146
{
1147
	int retval = 0;
1148 1149
	int fput_needed;
	struct file *file;
1150
	struct fs_struct *fs = current->fs;
L
Linus Torvalds 已提交
1151 1152 1153 1154 1155 1156

	nd->last_type = LAST_ROOT; /* if there are only slashes... */
	nd->flags = flags;
	nd->depth = 0;

	if (*name=='/') {
1157
		read_lock(&fs->lock);
J
Jan Blunck 已提交
1158 1159 1160
		if (fs->altroot.dentry && !(nd->flags & LOOKUP_NOALT)) {
			nd->path = fs->altroot;
			path_get(&fs->altroot);
1161
			read_unlock(&fs->lock);
L
Linus Torvalds 已提交
1162
			if (__emul_lookup_dentry(name,nd))
1163
				goto out; /* found in altroot */
1164
			read_lock(&fs->lock);
L
Linus Torvalds 已提交
1165
		}
J
Jan Blunck 已提交
1166 1167
		nd->path = fs->root;
		path_get(&fs->root);
1168
		read_unlock(&fs->lock);
1169
	} else if (dfd == AT_FDCWD) {
1170
		read_lock(&fs->lock);
J
Jan Blunck 已提交
1171 1172
		nd->path = fs->pwd;
		path_get(&fs->pwd);
1173
		read_unlock(&fs->lock);
1174 1175 1176 1177
	} else {
		struct dentry *dentry;

		file = fget_light(dfd, &fput_needed);
1178 1179
		retval = -EBADF;
		if (!file)
1180
			goto out_fail;
1181

1182
		dentry = file->f_path.dentry;
1183

1184 1185
		retval = -ENOTDIR;
		if (!S_ISDIR(dentry->d_inode->i_mode))
1186
			goto fput_fail;
1187 1188

		retval = file_permission(file, MAY_EXEC);
1189
		if (retval)
1190
			goto fput_fail;
1191

J
Jan Blunck 已提交
1192 1193
		nd->path = file->f_path;
		path_get(&file->f_path);
1194 1195

		fput_light(file, fput_needed);
L
Linus Torvalds 已提交
1196
	}
1197 1198

	retval = path_walk(name, nd);
1199
out:
1200 1201 1202
	if (unlikely(!retval && !audit_dummy_context() && nd->path.dentry &&
				nd->path.dentry->d_inode))
		audit_inode(name, nd->path.dentry);
1203
out_fail:
1204 1205
	return retval;

1206
fput_fail:
1207
	fput_light(file, fput_needed);
1208
	goto out_fail;
L
Linus Torvalds 已提交
1209 1210
}

1211
int path_lookup(const char *name, unsigned int flags,
1212 1213 1214 1215 1216
			struct nameidata *nd)
{
	return do_path_lookup(AT_FDCWD, name, flags, nd);
}

1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235
/**
 * vfs_path_lookup - lookup a file path relative to a dentry-vfsmount pair
 * @dentry:  pointer to dentry of the base directory
 * @mnt: pointer to vfs mount of the base directory
 * @name: pointer to file name
 * @flags: lookup flags
 * @nd: pointer to nameidata
 */
int vfs_path_lookup(struct dentry *dentry, struct vfsmount *mnt,
		    const char *name, unsigned int flags,
		    struct nameidata *nd)
{
	int retval;

	/* same as do_path_lookup */
	nd->last_type = LAST_ROOT;
	nd->flags = flags;
	nd->depth = 0;

1236 1237 1238
	nd->path.dentry = dentry;
	nd->path.mnt = mnt;
	path_get(&nd->path);
1239 1240

	retval = path_walk(name, nd);
1241 1242 1243
	if (unlikely(!retval && !audit_dummy_context() && nd->path.dentry &&
				nd->path.dentry->d_inode))
		audit_inode(name, nd->path.dentry);
1244 1245 1246 1247 1248

	return retval;

}

1249 1250 1251
static int __path_lookup_intent_open(int dfd, const char *name,
		unsigned int lookup_flags, struct nameidata *nd,
		int open_flags, int create_mode)
1252 1253 1254 1255 1256 1257 1258 1259 1260
{
	struct file *filp = get_empty_filp();
	int err;

	if (filp == NULL)
		return -ENFILE;
	nd->intent.open.file = filp;
	nd->intent.open.flags = open_flags;
	nd->intent.open.create_mode = create_mode;
1261
	err = do_path_lookup(dfd, name, lookup_flags|LOOKUP_OPEN, nd);
1262 1263 1264
	if (IS_ERR(nd->intent.open.file)) {
		if (err == 0) {
			err = PTR_ERR(nd->intent.open.file);
J
Jan Blunck 已提交
1265
			path_put(&nd->path);
1266 1267 1268 1269 1270 1271 1272 1273
		}
	} else if (err != 0)
		release_open_intent(nd);
	return err;
}

/**
 * path_lookup_open - lookup a file path with open intent
1274
 * @dfd: the directory to use as base, or AT_FDCWD
1275 1276 1277 1278 1279
 * @name: pointer to file name
 * @lookup_flags: lookup intent flags
 * @nd: pointer to nameidata
 * @open_flags: open intent flags
 */
1280
int path_lookup_open(int dfd, const char *name, unsigned int lookup_flags,
1281 1282
		struct nameidata *nd, int open_flags)
{
1283
	return __path_lookup_intent_open(dfd, name, lookup_flags, nd,
1284 1285 1286 1287 1288
			open_flags, 0);
}

/**
 * path_lookup_create - lookup a file path with open + create intent
1289
 * @dfd: the directory to use as base, or AT_FDCWD
1290 1291 1292 1293 1294 1295
 * @name: pointer to file name
 * @lookup_flags: lookup intent flags
 * @nd: pointer to nameidata
 * @open_flags: open intent flags
 * @create_mode: create intent flags
 */
1296 1297 1298
static int path_lookup_create(int dfd, const char *name,
			      unsigned int lookup_flags, struct nameidata *nd,
			      int open_flags, int create_mode)
1299
{
1300 1301
	return __path_lookup_intent_open(dfd, name, lookup_flags|LOOKUP_CREATE,
			nd, open_flags, create_mode);
1302 1303 1304 1305 1306 1307 1308 1309 1310
}

int __user_path_lookup_open(const char __user *name, unsigned int lookup_flags,
		struct nameidata *nd, int open_flags)
{
	char *tmp = getname(name);
	int err = PTR_ERR(tmp);

	if (!IS_ERR(tmp)) {
1311
		err = __path_lookup_intent_open(AT_FDCWD, tmp, lookup_flags, nd, open_flags, 0);
1312 1313 1314 1315 1316
		putname(tmp);
	}
	return err;
}

1317 1318
static struct dentry *__lookup_hash(struct qstr *name,
		struct dentry *base, struct nameidata *nd)
L
Linus Torvalds 已提交
1319
{
1320
	struct dentry *dentry;
L
Linus Torvalds 已提交
1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338
	struct inode *inode;
	int err;

	inode = base->d_inode;

	/*
	 * See if the low-level filesystem might want
	 * to use its own hash..
	 */
	if (base->d_op && base->d_op->d_hash) {
		err = base->d_op->d_hash(base, name);
		dentry = ERR_PTR(err);
		if (err < 0)
			goto out;
	}

	dentry = cached_lookup(base, name, nd);
	if (!dentry) {
1339 1340 1341 1342 1343 1344 1345 1346
		struct dentry *new;

		/* Don't create child dentry for a dead directory. */
		dentry = ERR_PTR(-ENOENT);
		if (IS_DEADDIR(inode))
			goto out;

		new = d_alloc(base, name);
L
Linus Torvalds 已提交
1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359
		dentry = ERR_PTR(-ENOMEM);
		if (!new)
			goto out;
		dentry = inode->i_op->lookup(inode, new, nd);
		if (!dentry)
			dentry = new;
		else
			dput(new);
	}
out:
	return dentry;
}

1360 1361 1362 1363 1364
/*
 * Restricted form of lookup. Doesn't follow links, single-component only,
 * needs parent already locked. Doesn't follow mounts.
 * SMP-safe.
 */
1365
static struct dentry *lookup_hash(struct nameidata *nd)
1366 1367 1368
{
	int err;

1369
	err = permission(nd->path.dentry->d_inode, MAY_EXEC, nd);
1370
	if (err)
1371
		return ERR_PTR(err);
1372
	return __lookup_hash(&nd->last, nd->path.dentry, nd);
L
Linus Torvalds 已提交
1373 1374
}

1375 1376
static int __lookup_one_len(const char *name, struct qstr *this,
		struct dentry *base, int len)
L
Linus Torvalds 已提交
1377 1378 1379 1380
{
	unsigned long hash;
	unsigned int c;

1381 1382
	this->name = name;
	this->len = len;
L
Linus Torvalds 已提交
1383
	if (!len)
1384
		return -EACCES;
L
Linus Torvalds 已提交
1385 1386 1387 1388 1389

	hash = init_name_hash();
	while (len--) {
		c = *(const unsigned char *)name++;
		if (c == '/' || c == '\0')
1390
			return -EACCES;
L
Linus Torvalds 已提交
1391 1392
		hash = partial_name_hash(c, hash);
	}
1393 1394 1395
	this->hash = end_name_hash(hash);
	return 0;
}
L
Linus Torvalds 已提交
1396

1397
/**
1398
 * lookup_one_len - filesystem helper to lookup single pathname component
1399 1400 1401 1402
 * @name:	pathname component to lookup
 * @base:	base directory to lookup from
 * @len:	maximum length @len should be interpreted to
 *
1403 1404
 * Note that this routine is purely a helper for filesystem usage and should
 * not be called by generic code.  Also note that by using this function the
1405 1406 1407
 * nameidata argument is passed to the filesystem methods and a filesystem
 * using this helper needs to be prepared for that.
 */
1408 1409 1410 1411 1412 1413
struct dentry *lookup_one_len(const char *name, struct dentry *base, int len)
{
	int err;
	struct qstr this;

	err = __lookup_one_len(name, &this, base, len);
1414 1415 1416 1417
	if (err)
		return ERR_PTR(err);

	err = permission(base->d_inode, MAY_EXEC, NULL);
1418 1419
	if (err)
		return ERR_PTR(err);
1420
	return __lookup_hash(&this, base, NULL);
1421 1422
}

1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434
/**
 * lookup_one_noperm - bad hack for sysfs
 * @name:	pathname component to lookup
 * @base:	base directory to lookup from
 *
 * This is a variant of lookup_one_len that doesn't perform any permission
 * checks.   It's a horrible hack to work around the braindead sysfs
 * architecture and should not be used anywhere else.
 *
 * DON'T USE THIS FUNCTION EVER, thanks.
 */
struct dentry *lookup_one_noperm(const char *name, struct dentry *base)
1435 1436 1437 1438
{
	int err;
	struct qstr this;

1439
	err = __lookup_one_len(name, &this, base, strlen(name));
1440 1441
	if (err)
		return ERR_PTR(err);
1442
	return __lookup_hash(&this, base, NULL);
L
Linus Torvalds 已提交
1443 1444
}

1445
int __user_walk_fd(int dfd, const char __user *name, unsigned flags,
1446
			    struct nameidata *nd)
L
Linus Torvalds 已提交
1447 1448 1449 1450 1451
{
	char *tmp = getname(name);
	int err = PTR_ERR(tmp);

	if (!IS_ERR(tmp)) {
1452
		err = do_path_lookup(dfd, tmp, flags, nd);
L
Linus Torvalds 已提交
1453 1454 1455 1456 1457
		putname(tmp);
	}
	return err;
}

1458
int __user_walk(const char __user *name, unsigned flags, struct nameidata *nd)
1459 1460 1461 1462
{
	return __user_walk_fd(AT_FDCWD, name, flags, nd);
}

L
Linus Torvalds 已提交
1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496
/*
 * It's inline, so penalty for filesystems that don't use sticky bit is
 * minimal.
 */
static inline int check_sticky(struct inode *dir, struct inode *inode)
{
	if (!(dir->i_mode & S_ISVTX))
		return 0;
	if (inode->i_uid == current->fsuid)
		return 0;
	if (dir->i_uid == current->fsuid)
		return 0;
	return !capable(CAP_FOWNER);
}

/*
 *	Check whether we can remove a link victim from directory dir, check
 *  whether the type of victim is right.
 *  1. We can't do it if dir is read-only (done in permission())
 *  2. We should have write and exec permissions on dir
 *  3. We can't remove anything from append-only dir
 *  4. We can't do anything with immutable dir (done in permission())
 *  5. If the sticky bit on dir is set we should either
 *	a. be owner of dir, or
 *	b. be owner of victim, or
 *	c. have CAP_FOWNER capability
 *  6. If the victim is append-only or immutable we can't do antyhing with
 *     links pointing to it.
 *  7. If we were asked to remove a directory and victim isn't one - ENOTDIR.
 *  8. If we were asked to remove a non-directory and victim isn't one - EISDIR.
 *  9. We can't remove a root or mountpoint.
 * 10. We don't allow removal of NFS sillyrenamed files; it's handled by
 *     nfs_async_unlink().
 */
1497
static int may_delete(struct inode *dir,struct dentry *victim,int isdir)
L
Linus Torvalds 已提交
1498 1499 1500 1501 1502 1503 1504
{
	int error;

	if (!victim->d_inode)
		return -ENOENT;

	BUG_ON(victim->d_parent->d_inode != dir);
1505
	audit_inode_child(victim->d_name.name, victim, dir);
L
Linus Torvalds 已提交
1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570

	error = permission(dir,MAY_WRITE | MAY_EXEC, NULL);
	if (error)
		return error;
	if (IS_APPEND(dir))
		return -EPERM;
	if (check_sticky(dir, victim->d_inode)||IS_APPEND(victim->d_inode)||
	    IS_IMMUTABLE(victim->d_inode))
		return -EPERM;
	if (isdir) {
		if (!S_ISDIR(victim->d_inode->i_mode))
			return -ENOTDIR;
		if (IS_ROOT(victim))
			return -EBUSY;
	} else if (S_ISDIR(victim->d_inode->i_mode))
		return -EISDIR;
	if (IS_DEADDIR(dir))
		return -ENOENT;
	if (victim->d_flags & DCACHE_NFSFS_RENAMED)
		return -EBUSY;
	return 0;
}

/*	Check whether we can create an object with dentry child in directory
 *  dir.
 *  1. We can't do it if child already exists (open has special treatment for
 *     this case, but since we are inlined it's OK)
 *  2. We can't do it if dir is read-only (done in permission())
 *  3. We should have write and exec permissions on dir
 *  4. We can't do it if dir is immutable (done in permission())
 */
static inline int may_create(struct inode *dir, struct dentry *child,
			     struct nameidata *nd)
{
	if (child->d_inode)
		return -EEXIST;
	if (IS_DEADDIR(dir))
		return -ENOENT;
	return permission(dir,MAY_WRITE | MAY_EXEC, nd);
}

/* 
 * O_DIRECTORY translates into forcing a directory lookup.
 */
static inline int lookup_flags(unsigned int f)
{
	unsigned long retval = LOOKUP_FOLLOW;

	if (f & O_NOFOLLOW)
		retval &= ~LOOKUP_FOLLOW;
	
	if (f & O_DIRECTORY)
		retval |= LOOKUP_DIRECTORY;

	return retval;
}

/*
 * p1 and p2 should be directories on the same fs.
 */
struct dentry *lock_rename(struct dentry *p1, struct dentry *p2)
{
	struct dentry *p;

	if (p1 == p2) {
I
Ingo Molnar 已提交
1571
		mutex_lock_nested(&p1->d_inode->i_mutex, I_MUTEX_PARENT);
L
Linus Torvalds 已提交
1572 1573 1574
		return NULL;
	}

1575
	mutex_lock(&p1->d_inode->i_sb->s_vfs_rename_mutex);
L
Linus Torvalds 已提交
1576 1577 1578

	for (p = p1; p->d_parent != p; p = p->d_parent) {
		if (p->d_parent == p2) {
I
Ingo Molnar 已提交
1579 1580
			mutex_lock_nested(&p2->d_inode->i_mutex, I_MUTEX_PARENT);
			mutex_lock_nested(&p1->d_inode->i_mutex, I_MUTEX_CHILD);
L
Linus Torvalds 已提交
1581 1582 1583 1584 1585 1586
			return p;
		}
	}

	for (p = p2; p->d_parent != p; p = p->d_parent) {
		if (p->d_parent == p1) {
I
Ingo Molnar 已提交
1587 1588
			mutex_lock_nested(&p1->d_inode->i_mutex, I_MUTEX_PARENT);
			mutex_lock_nested(&p2->d_inode->i_mutex, I_MUTEX_CHILD);
L
Linus Torvalds 已提交
1589 1590 1591 1592
			return p;
		}
	}

I
Ingo Molnar 已提交
1593 1594
	mutex_lock_nested(&p1->d_inode->i_mutex, I_MUTEX_PARENT);
	mutex_lock_nested(&p2->d_inode->i_mutex, I_MUTEX_CHILD);
L
Linus Torvalds 已提交
1595 1596 1597 1598 1599
	return NULL;
}

void unlock_rename(struct dentry *p1, struct dentry *p2)
{
1600
	mutex_unlock(&p1->d_inode->i_mutex);
L
Linus Torvalds 已提交
1601
	if (p1 != p2) {
1602
		mutex_unlock(&p2->d_inode->i_mutex);
1603
		mutex_unlock(&p1->d_inode->i_sb->s_vfs_rename_mutex);
L
Linus Torvalds 已提交
1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623
	}
}

int vfs_create(struct inode *dir, struct dentry *dentry, int mode,
		struct nameidata *nd)
{
	int error = may_create(dir, dentry, nd);

	if (error)
		return error;

	if (!dir->i_op || !dir->i_op->create)
		return -EACCES;	/* shouldn't it be ENOSYS? */
	mode &= S_IALLUGO;
	mode |= S_IFREG;
	error = security_inode_create(dir, dentry, mode);
	if (error)
		return error;
	DQUOT_INIT(dir);
	error = dir->i_op->create(dir, dentry, mode, nd);
1624
	if (!error)
1625
		fsnotify_create(dir, dentry);
L
Linus Torvalds 已提交
1626 1627 1628 1629 1630
	return error;
}

int may_open(struct nameidata *nd, int acc_mode, int flag)
{
1631
	struct dentry *dentry = nd->path.dentry;
L
Linus Torvalds 已提交
1632 1633 1634 1635 1636 1637 1638 1639 1640
	struct inode *inode = dentry->d_inode;
	int error;

	if (!inode)
		return -ENOENT;

	if (S_ISLNK(inode->i_mode))
		return -ELOOP;
	
1641
	if (S_ISDIR(inode->i_mode) && (acc_mode & MAY_WRITE))
L
Linus Torvalds 已提交
1642 1643 1644 1645 1646 1647 1648 1649 1650 1651
		return -EISDIR;

	/*
	 * FIFO's, sockets and device files are special: they don't
	 * actually live on the filesystem itself, and as such you
	 * can write to them even if the filesystem is read-only.
	 */
	if (S_ISFIFO(inode->i_mode) || S_ISSOCK(inode->i_mode)) {
	    	flag &= ~O_TRUNC;
	} else if (S_ISBLK(inode->i_mode) || S_ISCHR(inode->i_mode)) {
1652
		if (nd->path.mnt->mnt_flags & MNT_NODEV)
L
Linus Torvalds 已提交
1653 1654 1655
			return -EACCES;

		flag &= ~O_TRUNC;
1656
	}
1657 1658 1659 1660

	error = vfs_permission(nd, acc_mode);
	if (error)
		return error;
L
Linus Torvalds 已提交
1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672
	/*
	 * An append-only file must be opened in append mode for writing.
	 */
	if (IS_APPEND(inode)) {
		if  ((flag & FMODE_WRITE) && !(flag & O_APPEND))
			return -EPERM;
		if (flag & O_TRUNC)
			return -EPERM;
	}

	/* O_NOATIME can only be set by the owner or superuser */
	if (flag & O_NOATIME)
1673
		if (!is_owner_or_cap(inode))
L
Linus Torvalds 已提交
1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693
			return -EPERM;

	/*
	 * Ensure there are no outstanding leases on the file.
	 */
	error = break_lease(inode, flag);
	if (error)
		return error;

	if (flag & O_TRUNC) {
		error = get_write_access(inode);
		if (error)
			return error;

		/*
		 * Refuse to truncate files with mandatory locks held on them.
		 */
		error = locks_verify_locked(inode);
		if (!error) {
			DQUOT_INIT(inode);
1694 1695 1696 1697

			error = do_truncate(dentry, 0,
					    ATTR_MTIME|ATTR_CTIME|ATTR_OPEN,
					    NULL);
L
Linus Torvalds 已提交
1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708
		}
		put_write_access(inode);
		if (error)
			return error;
	} else
		if (flag & FMODE_WRITE)
			DQUOT_INIT(inode);

	return 0;
}

1709 1710 1711 1712 1713 1714
/*
 * Be careful about ever adding any more callers of this
 * function.  Its flags must be in the namei format, not
 * what get passed to sys_open().
 */
static int __open_namei_create(struct nameidata *nd, struct path *path,
1715 1716 1717
				int flag, int mode)
{
	int error;
1718
	struct dentry *dir = nd->path.dentry;
1719 1720 1721 1722 1723

	if (!IS_POSIXACL(dir->d_inode))
		mode &= ~current->fs->umask;
	error = vfs_create(dir->d_inode, path->dentry, mode, nd);
	mutex_unlock(&dir->d_inode->i_mutex);
1724 1725
	dput(nd->path.dentry);
	nd->path.dentry = path->dentry;
1726 1727 1728 1729 1730 1731
	if (error)
		return error;
	/* Don't check for write permission, don't truncate */
	return may_open(nd, 0, flag & ~O_TRUNC);
}

1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755
/*
 * Note that while the flag value (low two bits) for sys_open means:
 *	00 - read-only
 *	01 - write-only
 *	10 - read-write
 *	11 - special
 * it is changed into
 *	00 - no permissions needed
 *	01 - read-permission
 *	10 - write-permission
 *	11 - read-write
 * for the internal routines (ie open_namei()/follow_link() etc)
 * This is more logical, and also allows the 00 "no perm needed"
 * to be used for symlinks (where the permissions are checked
 * later).
 *
*/
static inline int open_to_namei_flags(int flag)
{
	if ((flag+1) & O_ACCMODE)
		flag++;
	return flag;
}

1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766
static int open_will_write_to_fs(int flag, struct inode *inode)
{
	/*
	 * We'll never write to the fs underlying
	 * a device file.
	 */
	if (special_file(inode->i_mode))
		return 0;
	return (flag & O_TRUNC);
}

L
Linus Torvalds 已提交
1767
/*
1768 1769 1770
 * Note that the low bits of the passed in "open_flag"
 * are not the same as in the local variable "flag". See
 * open_to_namei_flags() for more details.
L
Linus Torvalds 已提交
1771
 */
1772 1773
struct file *do_filp_open(int dfd, const char *pathname,
		int open_flag, int mode)
L
Linus Torvalds 已提交
1774
{
1775
	struct file *filp;
1776
	struct nameidata nd;
1777
	int acc_mode, error;
A
Al Viro 已提交
1778
	struct path path;
L
Linus Torvalds 已提交
1779 1780
	struct dentry *dir;
	int count = 0;
1781
	int will_write;
1782
	int flag = open_to_namei_flags(open_flag);
L
Linus Torvalds 已提交
1783 1784 1785

	acc_mode = ACC_MODE(flag);

1786 1787 1788 1789
	/* O_TRUNC implies we need access checks for write permissions */
	if (flag & O_TRUNC)
		acc_mode |= MAY_WRITE;

L
Linus Torvalds 已提交
1790 1791 1792 1793 1794 1795 1796 1797 1798
	/* Allow the LSM permission hook to distinguish append 
	   access from general write access. */
	if (flag & O_APPEND)
		acc_mode |= MAY_APPEND;

	/*
	 * The simplest case - just a plain lookup.
	 */
	if (!(flag & O_CREAT)) {
1799
		error = path_lookup_open(dfd, pathname, lookup_flags(flag),
1800
					 &nd, flag);
L
Linus Torvalds 已提交
1801
		if (error)
1802
			return ERR_PTR(error);
L
Linus Torvalds 已提交
1803 1804 1805 1806 1807 1808
		goto ok;
	}

	/*
	 * Create - we need to know the parent.
	 */
1809 1810
	error = path_lookup_create(dfd, pathname, LOOKUP_PARENT,
				   &nd, flag, mode);
L
Linus Torvalds 已提交
1811
	if (error)
1812
		return ERR_PTR(error);
L
Linus Torvalds 已提交
1813 1814 1815 1816 1817 1818 1819

	/*
	 * We have the parent and last component. First of all, check
	 * that we are not asked to creat(2) an obvious directory - that
	 * will not do.
	 */
	error = -EISDIR;
1820
	if (nd.last_type != LAST_NORM || nd.last.name[nd.last.len])
L
Linus Torvalds 已提交
1821 1822
		goto exit;

1823 1824
	dir = nd.path.dentry;
	nd.flags &= ~LOOKUP_PARENT;
1825
	mutex_lock(&dir->d_inode->i_mutex);
1826 1827
	path.dentry = lookup_hash(&nd);
	path.mnt = nd.path.mnt;
L
Linus Torvalds 已提交
1828 1829

do_last:
A
Al Viro 已提交
1830 1831
	error = PTR_ERR(path.dentry);
	if (IS_ERR(path.dentry)) {
1832
		mutex_unlock(&dir->d_inode->i_mutex);
L
Linus Torvalds 已提交
1833 1834 1835
		goto exit;
	}

1836 1837
	if (IS_ERR(nd.intent.open.file)) {
		error = PTR_ERR(nd.intent.open.file);
1838
		goto exit_mutex_unlock;
1839 1840
	}

L
Linus Torvalds 已提交
1841
	/* Negative dentry, just create the file */
A
Al Viro 已提交
1842
	if (!path.dentry->d_inode) {
1843 1844 1845 1846 1847 1848 1849 1850
		/*
		 * This write is needed to ensure that a
		 * ro->rw transition does not occur between
		 * the time when the file is created and when
		 * a permanent write count is taken through
		 * the 'struct file' in nameidata_to_filp().
		 */
		error = mnt_want_write(nd.path.mnt);
L
Linus Torvalds 已提交
1851
		if (error)
1852 1853 1854 1855
			goto exit_mutex_unlock;
		error = __open_namei_create(&nd, &path, flag, mode);
		if (error) {
			mnt_drop_write(nd.path.mnt);
L
Linus Torvalds 已提交
1856
			goto exit;
1857 1858 1859 1860
		}
		filp = nameidata_to_filp(&nd, open_flag);
		mnt_drop_write(nd.path.mnt);
		return filp;
L
Linus Torvalds 已提交
1861 1862 1863 1864 1865
	}

	/*
	 * It already exists.
	 */
1866
	mutex_unlock(&dir->d_inode->i_mutex);
1867
	audit_inode(pathname, path.dentry);
L
Linus Torvalds 已提交
1868 1869 1870 1871 1872

	error = -EEXIST;
	if (flag & O_EXCL)
		goto exit_dput;

A
Al Viro 已提交
1873
	if (__follow_mount(&path)) {
L
Linus Torvalds 已提交
1874
		error = -ELOOP;
A
Al Viro 已提交
1875 1876
		if (flag & O_NOFOLLOW)
			goto exit_dput;
L
Linus Torvalds 已提交
1877
	}
1878

L
Linus Torvalds 已提交
1879
	error = -ENOENT;
A
Al Viro 已提交
1880
	if (!path.dentry->d_inode)
L
Linus Torvalds 已提交
1881
		goto exit_dput;
A
Al Viro 已提交
1882
	if (path.dentry->d_inode->i_op && path.dentry->d_inode->i_op->follow_link)
L
Linus Torvalds 已提交
1883 1884
		goto do_link;

1885
	path_to_nameidata(&path, &nd);
L
Linus Torvalds 已提交
1886
	error = -EISDIR;
A
Al Viro 已提交
1887
	if (path.dentry->d_inode && S_ISDIR(path.dentry->d_inode->i_mode))
L
Linus Torvalds 已提交
1888 1889
		goto exit;
ok:
1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905
	/*
	 * Consider:
	 * 1. may_open() truncates a file
	 * 2. a rw->ro mount transition occurs
	 * 3. nameidata_to_filp() fails due to
	 *    the ro mount.
	 * That would be inconsistent, and should
	 * be avoided. Taking this mnt write here
	 * ensures that (2) can not occur.
	 */
	will_write = open_will_write_to_fs(flag, nd.path.dentry->d_inode);
	if (will_write) {
		error = mnt_want_write(nd.path.mnt);
		if (error)
			goto exit;
	}
1906
	error = may_open(&nd, acc_mode, flag);
1907 1908 1909
	if (error) {
		if (will_write)
			mnt_drop_write(nd.path.mnt);
L
Linus Torvalds 已提交
1910
		goto exit;
1911 1912 1913 1914 1915 1916 1917 1918 1919 1920
	}
	filp = nameidata_to_filp(&nd, open_flag);
	/*
	 * It is now safe to drop the mnt write
	 * because the filp has had a write taken
	 * on its behalf.
	 */
	if (will_write)
		mnt_drop_write(nd.path.mnt);
	return filp;
L
Linus Torvalds 已提交
1921

1922 1923
exit_mutex_unlock:
	mutex_unlock(&dir->d_inode->i_mutex);
L
Linus Torvalds 已提交
1924
exit_dput:
1925
	path_put_conditional(&path, &nd);
L
Linus Torvalds 已提交
1926
exit:
1927 1928 1929 1930
	if (!IS_ERR(nd.intent.open.file))
		release_open_intent(&nd);
	path_put(&nd.path);
	return ERR_PTR(error);
L
Linus Torvalds 已提交
1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945

do_link:
	error = -ELOOP;
	if (flag & O_NOFOLLOW)
		goto exit_dput;
	/*
	 * This is subtle. Instead of calling do_follow_link() we do the
	 * thing by hands. The reason is that this way we have zero link_count
	 * and path_walk() (called from ->follow_link) honoring LOOKUP_PARENT.
	 * After that we have the parent and last component, i.e.
	 * we are in the same situation as after the first path_walk().
	 * Well, almost - if the last component is normal we get its copy
	 * stored in nd->last.name and we will have to putname() it when we
	 * are done. Procfs-like symlinks just set LAST_BIND.
	 */
1946 1947
	nd.flags |= LOOKUP_PARENT;
	error = security_inode_follow_link(path.dentry, &nd);
L
Linus Torvalds 已提交
1948 1949
	if (error)
		goto exit_dput;
1950
	error = __do_follow_link(&path, &nd);
K
Kirill Korotaev 已提交
1951 1952 1953 1954 1955
	if (error) {
		/* Does someone understand code flow here? Or it is only
		 * me so stupid? Anathema to whoever designed this non-sense
		 * with "intent.open".
		 */
1956 1957
		release_open_intent(&nd);
		return ERR_PTR(error);
K
Kirill Korotaev 已提交
1958
	}
1959 1960
	nd.flags &= ~LOOKUP_PARENT;
	if (nd.last_type == LAST_BIND)
L
Linus Torvalds 已提交
1961 1962
		goto ok;
	error = -EISDIR;
1963
	if (nd.last_type != LAST_NORM)
L
Linus Torvalds 已提交
1964
		goto exit;
1965 1966
	if (nd.last.name[nd.last.len]) {
		__putname(nd.last.name);
L
Linus Torvalds 已提交
1967 1968 1969 1970
		goto exit;
	}
	error = -ELOOP;
	if (count++==32) {
1971
		__putname(nd.last.name);
L
Linus Torvalds 已提交
1972 1973
		goto exit;
	}
1974
	dir = nd.path.dentry;
1975
	mutex_lock(&dir->d_inode->i_mutex);
1976 1977 1978
	path.dentry = lookup_hash(&nd);
	path.mnt = nd.path.mnt;
	__putname(nd.last.name);
L
Linus Torvalds 已提交
1979 1980 1981
	goto do_last;
}

1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998
/**
 * filp_open - open file and return file pointer
 *
 * @filename:	path to open
 * @flags:	open flags as per the open(2) second argument
 * @mode:	mode for the new file if O_CREAT is set, else ignored
 *
 * This is the helper to open a file from kernelspace if you really
 * have to.  But in generally you should not do this, so please move
 * along, nothing to see here..
 */
struct file *filp_open(const char *filename, int flags, int mode)
{
	return do_filp_open(AT_FDCWD, filename, flags, mode);
}
EXPORT_SYMBOL(filp_open);

L
Linus Torvalds 已提交
1999 2000 2001 2002 2003 2004 2005
/**
 * lookup_create - lookup a dentry, creating it if it doesn't exist
 * @nd: nameidata info
 * @is_dir: directory flag
 *
 * Simple function to lookup and return a dentry and create it
 * if it doesn't exist.  Is SMP-safe.
2006
 *
2007
 * Returns with nd->path.dentry->d_inode->i_mutex locked.
L
Linus Torvalds 已提交
2008 2009 2010
 */
struct dentry *lookup_create(struct nameidata *nd, int is_dir)
{
2011
	struct dentry *dentry = ERR_PTR(-EEXIST);
L
Linus Torvalds 已提交
2012

2013
	mutex_lock_nested(&nd->path.dentry->d_inode->i_mutex, I_MUTEX_PARENT);
2014 2015 2016 2017
	/*
	 * Yucky last component or no last component at all?
	 * (foo/., foo/.., /////)
	 */
L
Linus Torvalds 已提交
2018 2019 2020
	if (nd->last_type != LAST_NORM)
		goto fail;
	nd->flags &= ~LOOKUP_PARENT;
2021 2022
	nd->flags |= LOOKUP_CREATE;
	nd->intent.open.flags = O_EXCL;
2023 2024 2025 2026

	/*
	 * Do the final lookup.
	 */
2027
	dentry = lookup_hash(nd);
L
Linus Torvalds 已提交
2028 2029
	if (IS_ERR(dentry))
		goto fail;
2030

2031 2032
	if (dentry->d_inode)
		goto eexist;
2033 2034 2035 2036 2037 2038
	/*
	 * Special case - lookup gave negative, but... we had foo/bar/
	 * From the vfs_mknod() POV we just have a negative dentry -
	 * all is fine. Let's be bastards - you had / on the end, you've
	 * been asking for (non-existent) directory. -ENOENT for you.
	 */
2039 2040 2041 2042
	if (unlikely(!is_dir && nd->last.name[nd->last.len])) {
		dput(dentry);
		dentry = ERR_PTR(-ENOENT);
	}
L
Linus Torvalds 已提交
2043
	return dentry;
2044
eexist:
L
Linus Torvalds 已提交
2045
	dput(dentry);
2046
	dentry = ERR_PTR(-EEXIST);
L
Linus Torvalds 已提交
2047 2048 2049
fail:
	return dentry;
}
2050
EXPORT_SYMBOL_GPL(lookup_create);
L
Linus Torvalds 已提交
2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064

int vfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t dev)
{
	int error = may_create(dir, dentry, NULL);

	if (error)
		return error;

	if ((S_ISCHR(mode) || S_ISBLK(mode)) && !capable(CAP_MKNOD))
		return -EPERM;

	if (!dir->i_op || !dir->i_op->mknod)
		return -EPERM;

2065 2066 2067 2068
	error = devcgroup_inode_mknod(mode, dev);
	if (error)
		return error;

L
Linus Torvalds 已提交
2069 2070 2071 2072 2073 2074
	error = security_inode_mknod(dir, dentry, mode, dev);
	if (error)
		return error;

	DQUOT_INIT(dir);
	error = dir->i_op->mknod(dir, dentry, mode, dev);
2075
	if (!error)
2076
		fsnotify_create(dir, dentry);
L
Linus Torvalds 已提交
2077 2078 2079
	return error;
}

2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096
static int may_mknod(mode_t mode)
{
	switch (mode & S_IFMT) {
	case S_IFREG:
	case S_IFCHR:
	case S_IFBLK:
	case S_IFIFO:
	case S_IFSOCK:
	case 0: /* zero mode translates to S_IFREG */
		return 0;
	case S_IFDIR:
		return -EPERM;
	default:
		return -EINVAL;
	}
}

2097 2098
asmlinkage long sys_mknodat(int dfd, const char __user *filename, int mode,
				unsigned dev)
L
Linus Torvalds 已提交
2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110
{
	int error = 0;
	char * tmp;
	struct dentry * dentry;
	struct nameidata nd;

	if (S_ISDIR(mode))
		return -EPERM;
	tmp = getname(filename);
	if (IS_ERR(tmp))
		return PTR_ERR(tmp);

2111
	error = do_path_lookup(dfd, tmp, LOOKUP_PARENT, &nd);
L
Linus Torvalds 已提交
2112 2113 2114
	if (error)
		goto out;
	dentry = lookup_create(&nd, 0);
2115 2116 2117 2118
	if (IS_ERR(dentry)) {
		error = PTR_ERR(dentry);
		goto out_unlock;
	}
2119
	if (!IS_POSIXACL(nd.path.dentry->d_inode))
L
Linus Torvalds 已提交
2120
		mode &= ~current->fs->umask;
2121 2122 2123 2124 2125 2126 2127
	error = may_mknod(mode);
	if (error)
		goto out_dput;
	error = mnt_want_write(nd.path.mnt);
	if (error)
		goto out_dput;
	switch (mode & S_IFMT) {
L
Linus Torvalds 已提交
2128
		case 0: case S_IFREG:
2129
			error = vfs_create(nd.path.dentry->d_inode,dentry,mode,&nd);
L
Linus Torvalds 已提交
2130 2131
			break;
		case S_IFCHR: case S_IFBLK:
2132
			error = vfs_mknod(nd.path.dentry->d_inode,dentry,mode,
L
Linus Torvalds 已提交
2133 2134 2135
					new_decode_dev(dev));
			break;
		case S_IFIFO: case S_IFSOCK:
2136
			error = vfs_mknod(nd.path.dentry->d_inode,dentry,mode,0);
L
Linus Torvalds 已提交
2137 2138
			break;
	}
2139 2140 2141 2142
	mnt_drop_write(nd.path.mnt);
out_dput:
	dput(dentry);
out_unlock:
2143
	mutex_unlock(&nd.path.dentry->d_inode->i_mutex);
J
Jan Blunck 已提交
2144
	path_put(&nd.path);
L
Linus Torvalds 已提交
2145 2146 2147 2148 2149 2150
out:
	putname(tmp);

	return error;
}

2151 2152 2153 2154 2155
asmlinkage long sys_mknod(const char __user *filename, int mode, unsigned dev)
{
	return sys_mknodat(AT_FDCWD, filename, mode, dev);
}

L
Linus Torvalds 已提交
2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172
int vfs_mkdir(struct inode *dir, struct dentry *dentry, int mode)
{
	int error = may_create(dir, dentry, NULL);

	if (error)
		return error;

	if (!dir->i_op || !dir->i_op->mkdir)
		return -EPERM;

	mode &= (S_IRWXUGO|S_ISVTX);
	error = security_inode_mkdir(dir, dentry, mode);
	if (error)
		return error;

	DQUOT_INIT(dir);
	error = dir->i_op->mkdir(dir, dentry, mode);
2173
	if (!error)
2174
		fsnotify_mkdir(dir, dentry);
L
Linus Torvalds 已提交
2175 2176 2177
	return error;
}

2178
asmlinkage long sys_mkdirat(int dfd, const char __user *pathname, int mode)
L
Linus Torvalds 已提交
2179 2180 2181
{
	int error = 0;
	char * tmp;
2182 2183
	struct dentry *dentry;
	struct nameidata nd;
L
Linus Torvalds 已提交
2184 2185 2186

	tmp = getname(pathname);
	error = PTR_ERR(tmp);
2187 2188
	if (IS_ERR(tmp))
		goto out_err;
L
Linus Torvalds 已提交
2189

2190 2191 2192 2193 2194 2195 2196
	error = do_path_lookup(dfd, tmp, LOOKUP_PARENT, &nd);
	if (error)
		goto out;
	dentry = lookup_create(&nd, 1);
	error = PTR_ERR(dentry);
	if (IS_ERR(dentry))
		goto out_unlock;
L
Linus Torvalds 已提交
2197

2198
	if (!IS_POSIXACL(nd.path.dentry->d_inode))
2199
		mode &= ~current->fs->umask;
2200 2201 2202
	error = mnt_want_write(nd.path.mnt);
	if (error)
		goto out_dput;
2203
	error = vfs_mkdir(nd.path.dentry->d_inode, dentry, mode);
2204 2205
	mnt_drop_write(nd.path.mnt);
out_dput:
2206 2207
	dput(dentry);
out_unlock:
2208
	mutex_unlock(&nd.path.dentry->d_inode->i_mutex);
J
Jan Blunck 已提交
2209
	path_put(&nd.path);
2210 2211 2212
out:
	putname(tmp);
out_err:
L
Linus Torvalds 已提交
2213 2214 2215
	return error;
}

2216 2217 2218 2219 2220
asmlinkage long sys_mkdir(const char __user *pathname, int mode)
{
	return sys_mkdirat(AT_FDCWD, pathname, mode);
}

L
Linus Torvalds 已提交
2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238
/*
 * We try to drop the dentry early: we should have
 * a usage count of 2 if we're the only user of this
 * dentry, and if that is true (possibly after pruning
 * the dcache), then we drop the dentry now.
 *
 * A low-level filesystem can, if it choses, legally
 * do a
 *
 *	if (!d_unhashed(dentry))
 *		return -EBUSY;
 *
 * if it cannot handle the case of removing a directory
 * that is still in use by something else..
 */
void dentry_unhash(struct dentry *dentry)
{
	dget(dentry);
2239
	shrink_dcache_parent(dentry);
L
Linus Torvalds 已提交
2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259
	spin_lock(&dcache_lock);
	spin_lock(&dentry->d_lock);
	if (atomic_read(&dentry->d_count) == 2)
		__d_drop(dentry);
	spin_unlock(&dentry->d_lock);
	spin_unlock(&dcache_lock);
}

int vfs_rmdir(struct inode *dir, struct dentry *dentry)
{
	int error = may_delete(dir, dentry, 1);

	if (error)
		return error;

	if (!dir->i_op || !dir->i_op->rmdir)
		return -EPERM;

	DQUOT_INIT(dir);

2260
	mutex_lock(&dentry->d_inode->i_mutex);
L
Linus Torvalds 已提交
2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271
	dentry_unhash(dentry);
	if (d_mountpoint(dentry))
		error = -EBUSY;
	else {
		error = security_inode_rmdir(dir, dentry);
		if (!error) {
			error = dir->i_op->rmdir(dir, dentry);
			if (!error)
				dentry->d_inode->i_flags |= S_DEAD;
		}
	}
2272
	mutex_unlock(&dentry->d_inode->i_mutex);
L
Linus Torvalds 已提交
2273 2274 2275 2276 2277 2278 2279 2280
	if (!error) {
		d_delete(dentry);
	}
	dput(dentry);

	return error;
}

2281
static long do_rmdir(int dfd, const char __user *pathname)
L
Linus Torvalds 已提交
2282 2283 2284 2285 2286 2287 2288 2289 2290 2291
{
	int error = 0;
	char * name;
	struct dentry *dentry;
	struct nameidata nd;

	name = getname(pathname);
	if(IS_ERR(name))
		return PTR_ERR(name);

2292
	error = do_path_lookup(dfd, name, LOOKUP_PARENT, &nd);
L
Linus Torvalds 已提交
2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306
	if (error)
		goto exit;

	switch(nd.last_type) {
		case LAST_DOTDOT:
			error = -ENOTEMPTY;
			goto exit1;
		case LAST_DOT:
			error = -EINVAL;
			goto exit1;
		case LAST_ROOT:
			error = -EBUSY;
			goto exit1;
	}
2307
	mutex_lock_nested(&nd.path.dentry->d_inode->i_mutex, I_MUTEX_PARENT);
2308
	dentry = lookup_hash(&nd);
L
Linus Torvalds 已提交
2309
	error = PTR_ERR(dentry);
2310 2311
	if (IS_ERR(dentry))
		goto exit2;
2312 2313 2314
	error = mnt_want_write(nd.path.mnt);
	if (error)
		goto exit3;
2315
	error = vfs_rmdir(nd.path.dentry->d_inode, dentry);
2316 2317
	mnt_drop_write(nd.path.mnt);
exit3:
2318 2319
	dput(dentry);
exit2:
2320
	mutex_unlock(&nd.path.dentry->d_inode->i_mutex);
L
Linus Torvalds 已提交
2321
exit1:
J
Jan Blunck 已提交
2322
	path_put(&nd.path);
L
Linus Torvalds 已提交
2323 2324 2325 2326 2327
exit:
	putname(name);
	return error;
}

2328 2329 2330 2331 2332
asmlinkage long sys_rmdir(const char __user *pathname)
{
	return do_rmdir(AT_FDCWD, pathname);
}

L
Linus Torvalds 已提交
2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344
int vfs_unlink(struct inode *dir, struct dentry *dentry)
{
	int error = may_delete(dir, dentry, 0);

	if (error)
		return error;

	if (!dir->i_op || !dir->i_op->unlink)
		return -EPERM;

	DQUOT_INIT(dir);

2345
	mutex_lock(&dentry->d_inode->i_mutex);
L
Linus Torvalds 已提交
2346 2347 2348 2349 2350 2351 2352
	if (d_mountpoint(dentry))
		error = -EBUSY;
	else {
		error = security_inode_unlink(dir, dentry);
		if (!error)
			error = dir->i_op->unlink(dir, dentry);
	}
2353
	mutex_unlock(&dentry->d_inode->i_mutex);
L
Linus Torvalds 已提交
2354 2355 2356

	/* We don't d_delete() NFS sillyrenamed files--they still exist. */
	if (!error && !(dentry->d_flags & DCACHE_NFSFS_RENAMED)) {
2357
		fsnotify_link_count(dentry->d_inode);
J
John McCutchan 已提交
2358
		d_delete(dentry);
L
Linus Torvalds 已提交
2359
	}
R
Robert Love 已提交
2360

L
Linus Torvalds 已提交
2361 2362 2363 2364 2365
	return error;
}

/*
 * Make sure that the actual truncation of the file will occur outside its
2366
 * directory's i_mutex.  Truncate can take a long time if there is a lot of
L
Linus Torvalds 已提交
2367 2368 2369
 * writeout happening, and we don't want to prevent access to the directory
 * while waiting on the I/O.
 */
2370
static long do_unlinkat(int dfd, const char __user *pathname)
L
Linus Torvalds 已提交
2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381
{
	int error = 0;
	char * name;
	struct dentry *dentry;
	struct nameidata nd;
	struct inode *inode = NULL;

	name = getname(pathname);
	if(IS_ERR(name))
		return PTR_ERR(name);

2382
	error = do_path_lookup(dfd, name, LOOKUP_PARENT, &nd);
L
Linus Torvalds 已提交
2383 2384 2385 2386 2387
	if (error)
		goto exit;
	error = -EISDIR;
	if (nd.last_type != LAST_NORM)
		goto exit1;
2388
	mutex_lock_nested(&nd.path.dentry->d_inode->i_mutex, I_MUTEX_PARENT);
2389
	dentry = lookup_hash(&nd);
L
Linus Torvalds 已提交
2390 2391 2392 2393 2394 2395 2396 2397
	error = PTR_ERR(dentry);
	if (!IS_ERR(dentry)) {
		/* Why not before? Because we want correct error value */
		if (nd.last.name[nd.last.len])
			goto slashes;
		inode = dentry->d_inode;
		if (inode)
			atomic_inc(&inode->i_count);
2398 2399 2400
		error = mnt_want_write(nd.path.mnt);
		if (error)
			goto exit2;
2401
		error = vfs_unlink(nd.path.dentry->d_inode, dentry);
2402
		mnt_drop_write(nd.path.mnt);
L
Linus Torvalds 已提交
2403 2404 2405
	exit2:
		dput(dentry);
	}
2406
	mutex_unlock(&nd.path.dentry->d_inode->i_mutex);
L
Linus Torvalds 已提交
2407 2408 2409
	if (inode)
		iput(inode);	/* truncate the inode here */
exit1:
J
Jan Blunck 已提交
2410
	path_put(&nd.path);
L
Linus Torvalds 已提交
2411 2412 2413 2414 2415 2416 2417 2418 2419 2420
exit:
	putname(name);
	return error;

slashes:
	error = !dentry->d_inode ? -ENOENT :
		S_ISDIR(dentry->d_inode->i_mode) ? -EISDIR : -ENOTDIR;
	goto exit2;
}

2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436
asmlinkage long sys_unlinkat(int dfd, const char __user *pathname, int flag)
{
	if ((flag & ~AT_REMOVEDIR) != 0)
		return -EINVAL;

	if (flag & AT_REMOVEDIR)
		return do_rmdir(dfd, pathname);

	return do_unlinkat(dfd, pathname);
}

asmlinkage long sys_unlink(const char __user *pathname)
{
	return do_unlinkat(AT_FDCWD, pathname);
}

L
Linus Torvalds 已提交
2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452
int vfs_symlink(struct inode *dir, struct dentry *dentry, const char *oldname, int mode)
{
	int error = may_create(dir, dentry, NULL);

	if (error)
		return error;

	if (!dir->i_op || !dir->i_op->symlink)
		return -EPERM;

	error = security_inode_symlink(dir, dentry, oldname);
	if (error)
		return error;

	DQUOT_INIT(dir);
	error = dir->i_op->symlink(dir, dentry, oldname);
2453
	if (!error)
2454
		fsnotify_create(dir, dentry);
L
Linus Torvalds 已提交
2455 2456 2457
	return error;
}

2458 2459
asmlinkage long sys_symlinkat(const char __user *oldname,
			      int newdfd, const char __user *newname)
L
Linus Torvalds 已提交
2460 2461 2462 2463
{
	int error = 0;
	char * from;
	char * to;
2464 2465
	struct dentry *dentry;
	struct nameidata nd;
L
Linus Torvalds 已提交
2466 2467 2468 2469 2470 2471

	from = getname(oldname);
	if(IS_ERR(from))
		return PTR_ERR(from);
	to = getname(newname);
	error = PTR_ERR(to);
2472 2473
	if (IS_ERR(to))
		goto out_putname;
L
Linus Torvalds 已提交
2474

2475 2476 2477 2478 2479 2480 2481 2482
	error = do_path_lookup(newdfd, to, LOOKUP_PARENT, &nd);
	if (error)
		goto out;
	dentry = lookup_create(&nd, 0);
	error = PTR_ERR(dentry);
	if (IS_ERR(dentry))
		goto out_unlock;

2483 2484 2485
	error = mnt_want_write(nd.path.mnt);
	if (error)
		goto out_dput;
2486
	error = vfs_symlink(nd.path.dentry->d_inode, dentry, from, S_IALLUGO);
2487 2488
	mnt_drop_write(nd.path.mnt);
out_dput:
2489 2490
	dput(dentry);
out_unlock:
2491
	mutex_unlock(&nd.path.dentry->d_inode->i_mutex);
J
Jan Blunck 已提交
2492
	path_put(&nd.path);
L
Linus Torvalds 已提交
2493
out:
2494 2495
	putname(to);
out_putname:
L
Linus Torvalds 已提交
2496 2497 2498 2499
	putname(from);
	return error;
}

2500 2501 2502 2503 2504
asmlinkage long sys_symlink(const char __user *oldname, const char __user *newname)
{
	return sys_symlinkat(oldname, AT_FDCWD, newname);
}

L
Linus Torvalds 已提交
2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526
int vfs_link(struct dentry *old_dentry, struct inode *dir, struct dentry *new_dentry)
{
	struct inode *inode = old_dentry->d_inode;
	int error;

	if (!inode)
		return -ENOENT;

	error = may_create(dir, new_dentry, NULL);
	if (error)
		return error;

	if (dir->i_sb != inode->i_sb)
		return -EXDEV;

	/*
	 * A link to an append-only or immutable file cannot be created.
	 */
	if (IS_APPEND(inode) || IS_IMMUTABLE(inode))
		return -EPERM;
	if (!dir->i_op || !dir->i_op->link)
		return -EPERM;
2527
	if (S_ISDIR(inode->i_mode))
L
Linus Torvalds 已提交
2528 2529 2530 2531 2532 2533
		return -EPERM;

	error = security_inode_link(old_dentry, dir, new_dentry);
	if (error)
		return error;

2534
	mutex_lock(&inode->i_mutex);
L
Linus Torvalds 已提交
2535 2536
	DQUOT_INIT(dir);
	error = dir->i_op->link(old_dentry, dir, new_dentry);
2537
	mutex_unlock(&inode->i_mutex);
2538
	if (!error)
2539
		fsnotify_link(dir, inode, new_dentry);
L
Linus Torvalds 已提交
2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551
	return error;
}

/*
 * Hardlinks are often used in delicate situations.  We avoid
 * security-related surprises by not following symlinks on the
 * newname.  --KAB
 *
 * We don't follow them on the oldname either to be compatible
 * with linux 2.0, and to avoid hard-linking to directories
 * and other special files.  --ADM
 */
2552
asmlinkage long sys_linkat(int olddfd, const char __user *oldname,
2553 2554
			   int newdfd, const char __user *newname,
			   int flags)
L
Linus Torvalds 已提交
2555 2556 2557 2558 2559 2560
{
	struct dentry *new_dentry;
	struct nameidata nd, old_nd;
	int error;
	char * to;

2561
	if ((flags & ~AT_SYMLINK_FOLLOW) != 0)
2562 2563
		return -EINVAL;

L
Linus Torvalds 已提交
2564 2565 2566 2567
	to = getname(newname);
	if (IS_ERR(to))
		return PTR_ERR(to);

2568 2569 2570
	error = __user_walk_fd(olddfd, oldname,
			       flags & AT_SYMLINK_FOLLOW ? LOOKUP_FOLLOW : 0,
			       &old_nd);
L
Linus Torvalds 已提交
2571 2572
	if (error)
		goto exit;
2573
	error = do_path_lookup(newdfd, to, LOOKUP_PARENT, &nd);
L
Linus Torvalds 已提交
2574 2575 2576
	if (error)
		goto out;
	error = -EXDEV;
2577
	if (old_nd.path.mnt != nd.path.mnt)
L
Linus Torvalds 已提交
2578 2579 2580
		goto out_release;
	new_dentry = lookup_create(&nd, 0);
	error = PTR_ERR(new_dentry);
2581 2582
	if (IS_ERR(new_dentry))
		goto out_unlock;
2583 2584 2585
	error = mnt_want_write(nd.path.mnt);
	if (error)
		goto out_dput;
2586
	error = vfs_link(old_nd.path.dentry, nd.path.dentry->d_inode, new_dentry);
2587 2588
	mnt_drop_write(nd.path.mnt);
out_dput:
2589 2590
	dput(new_dentry);
out_unlock:
2591
	mutex_unlock(&nd.path.dentry->d_inode->i_mutex);
L
Linus Torvalds 已提交
2592
out_release:
J
Jan Blunck 已提交
2593
	path_put(&nd.path);
L
Linus Torvalds 已提交
2594
out:
J
Jan Blunck 已提交
2595
	path_put(&old_nd.path);
L
Linus Torvalds 已提交
2596 2597 2598 2599 2600 2601
exit:
	putname(to);

	return error;
}

2602 2603
asmlinkage long sys_link(const char __user *oldname, const char __user *newname)
{
2604
	return sys_linkat(AT_FDCWD, oldname, AT_FDCWD, newname, 0);
2605 2606
}

L
Linus Torvalds 已提交
2607 2608 2609 2610 2611 2612 2613
/*
 * The worst of all namespace operations - renaming directory. "Perverted"
 * doesn't even start to describe it. Somebody in UCB had a heck of a trip...
 * Problems:
 *	a) we can get into loop creation. Check is done in is_subdir().
 *	b) race potential - two innocent renames can create a loop together.
 *	   That's where 4.4 screws up. Current fix: serialization on
2614
 *	   sb->s_vfs_rename_mutex. We might be more accurate, but that's another
L
Linus Torvalds 已提交
2615 2616
 *	   story.
 *	c) we have to lock _three_ objects - parents and victim (if it exists).
2617
 *	   And that - after we got ->i_mutex on parents (until then we don't know
L
Linus Torvalds 已提交
2618 2619
 *	   whether the target exists).  Solution: try to be smart with locking
 *	   order for inodes.  We rely on the fact that tree topology may change
2620
 *	   only under ->s_vfs_rename_mutex _and_ that parent of the object we
L
Linus Torvalds 已提交
2621 2622 2623
 *	   move will be locked.  Thus we can rank directories by the tree
 *	   (ancestors first) and rank all non-directories after them.
 *	   That works since everybody except rename does "lock parent, lookup,
2624
 *	   lock child" and rename is under ->s_vfs_rename_mutex.
L
Linus Torvalds 已提交
2625 2626 2627 2628 2629 2630 2631 2632 2633
 *	   HOWEVER, it relies on the assumption that any object with ->lookup()
 *	   has no more than 1 dentry.  If "hybrid" objects will ever appear,
 *	   we'd better make sure that there's no link(2) for them.
 *	d) some filesystems don't support opened-but-unlinked directories,
 *	   either because of layout or because they are not ready to deal with
 *	   all cases correctly. The latter will be fixed (taking this sort of
 *	   stuff into VFS), but the former is not going away. Solution: the same
 *	   trick as in rmdir().
 *	e) conversion from fhandle to dentry may come in the wrong moment - when
2634
 *	   we are removing the target. Solution: we will have to grab ->i_mutex
L
Linus Torvalds 已提交
2635
 *	   in the fhandle_to_dentry code. [FIXME - current nfsfh.c relies on
2636
 *	   ->i_mutex on parents, which works but leads to some truely excessive
L
Linus Torvalds 已提交
2637 2638
 *	   locking].
 */
A
Adrian Bunk 已提交
2639 2640
static int vfs_rename_dir(struct inode *old_dir, struct dentry *old_dentry,
			  struct inode *new_dir, struct dentry *new_dentry)
L
Linus Torvalds 已提交
2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660
{
	int error = 0;
	struct inode *target;

	/*
	 * If we are going to change the parent - check write permissions,
	 * we'll need to flip '..'.
	 */
	if (new_dir != old_dir) {
		error = permission(old_dentry->d_inode, MAY_WRITE, NULL);
		if (error)
			return error;
	}

	error = security_inode_rename(old_dir, old_dentry, new_dir, new_dentry);
	if (error)
		return error;

	target = new_dentry->d_inode;
	if (target) {
2661
		mutex_lock(&target->i_mutex);
L
Linus Torvalds 已提交
2662 2663 2664 2665 2666 2667 2668 2669 2670
		dentry_unhash(new_dentry);
	}
	if (d_mountpoint(old_dentry)||d_mountpoint(new_dentry))
		error = -EBUSY;
	else 
		error = old_dir->i_op->rename(old_dir, old_dentry, new_dir, new_dentry);
	if (target) {
		if (!error)
			target->i_flags |= S_DEAD;
2671
		mutex_unlock(&target->i_mutex);
L
Linus Torvalds 已提交
2672 2673 2674 2675
		if (d_unhashed(new_dentry))
			d_rehash(new_dentry);
		dput(new_dentry);
	}
2676
	if (!error)
2677 2678
		if (!(old_dir->i_sb->s_type->fs_flags & FS_RENAME_DOES_D_MOVE))
			d_move(old_dentry,new_dentry);
L
Linus Torvalds 已提交
2679 2680 2681
	return error;
}

A
Adrian Bunk 已提交
2682 2683
static int vfs_rename_other(struct inode *old_dir, struct dentry *old_dentry,
			    struct inode *new_dir, struct dentry *new_dentry)
L
Linus Torvalds 已提交
2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694
{
	struct inode *target;
	int error;

	error = security_inode_rename(old_dir, old_dentry, new_dir, new_dentry);
	if (error)
		return error;

	dget(new_dentry);
	target = new_dentry->d_inode;
	if (target)
2695
		mutex_lock(&target->i_mutex);
L
Linus Torvalds 已提交
2696 2697 2698 2699 2700
	if (d_mountpoint(old_dentry)||d_mountpoint(new_dentry))
		error = -EBUSY;
	else
		error = old_dir->i_op->rename(old_dir, old_dentry, new_dir, new_dentry);
	if (!error) {
2701
		if (!(old_dir->i_sb->s_type->fs_flags & FS_RENAME_DOES_D_MOVE))
L
Linus Torvalds 已提交
2702 2703 2704
			d_move(old_dentry, new_dentry);
	}
	if (target)
2705
		mutex_unlock(&target->i_mutex);
L
Linus Torvalds 已提交
2706 2707 2708 2709 2710 2711 2712 2713 2714
	dput(new_dentry);
	return error;
}

int vfs_rename(struct inode *old_dir, struct dentry *old_dentry,
	       struct inode *new_dir, struct dentry *new_dentry)
{
	int error;
	int is_dir = S_ISDIR(old_dentry->d_inode->i_mode);
R
Robert Love 已提交
2715
	const char *old_name;
L
Linus Torvalds 已提交
2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736

	if (old_dentry->d_inode == new_dentry->d_inode)
 		return 0;
 
	error = may_delete(old_dir, old_dentry, is_dir);
	if (error)
		return error;

	if (!new_dentry->d_inode)
		error = may_create(new_dir, new_dentry, NULL);
	else
		error = may_delete(new_dir, new_dentry, is_dir);
	if (error)
		return error;

	if (!old_dir->i_op || !old_dir->i_op->rename)
		return -EPERM;

	DQUOT_INIT(old_dir);
	DQUOT_INIT(new_dir);

R
Robert Love 已提交
2737 2738
	old_name = fsnotify_oldname_init(old_dentry->d_name.name);

L
Linus Torvalds 已提交
2739 2740 2741 2742 2743
	if (is_dir)
		error = vfs_rename_dir(old_dir,old_dentry,new_dir,new_dentry);
	else
		error = vfs_rename_other(old_dir,old_dentry,new_dir,new_dentry);
	if (!error) {
R
Robert Love 已提交
2744
		const char *new_name = old_dentry->d_name.name;
2745
		fsnotify_move(old_dir, new_dir, old_name, new_name, is_dir,
2746
			      new_dentry->d_inode, old_dentry);
L
Linus Torvalds 已提交
2747
	}
R
Robert Love 已提交
2748 2749
	fsnotify_oldname_free(old_name);

L
Linus Torvalds 已提交
2750 2751 2752
	return error;
}

2753 2754
static int do_rename(int olddfd, const char *oldname,
			int newdfd, const char *newname)
L
Linus Torvalds 已提交
2755 2756 2757 2758 2759 2760 2761
{
	int error = 0;
	struct dentry * old_dir, * new_dir;
	struct dentry * old_dentry, *new_dentry;
	struct dentry * trap;
	struct nameidata oldnd, newnd;

2762
	error = do_path_lookup(olddfd, oldname, LOOKUP_PARENT, &oldnd);
L
Linus Torvalds 已提交
2763 2764 2765
	if (error)
		goto exit;

2766
	error = do_path_lookup(newdfd, newname, LOOKUP_PARENT, &newnd);
L
Linus Torvalds 已提交
2767 2768 2769 2770
	if (error)
		goto exit1;

	error = -EXDEV;
2771
	if (oldnd.path.mnt != newnd.path.mnt)
L
Linus Torvalds 已提交
2772 2773
		goto exit2;

2774
	old_dir = oldnd.path.dentry;
L
Linus Torvalds 已提交
2775 2776 2777 2778
	error = -EBUSY;
	if (oldnd.last_type != LAST_NORM)
		goto exit2;

2779
	new_dir = newnd.path.dentry;
L
Linus Torvalds 已提交
2780 2781 2782 2783 2784
	if (newnd.last_type != LAST_NORM)
		goto exit2;

	trap = lock_rename(new_dir, old_dir);

2785
	old_dentry = lookup_hash(&oldnd);
L
Linus Torvalds 已提交
2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804
	error = PTR_ERR(old_dentry);
	if (IS_ERR(old_dentry))
		goto exit3;
	/* source must exist */
	error = -ENOENT;
	if (!old_dentry->d_inode)
		goto exit4;
	/* unless the source is a directory trailing slashes give -ENOTDIR */
	if (!S_ISDIR(old_dentry->d_inode->i_mode)) {
		error = -ENOTDIR;
		if (oldnd.last.name[oldnd.last.len])
			goto exit4;
		if (newnd.last.name[newnd.last.len])
			goto exit4;
	}
	/* source should not be ancestor of target */
	error = -EINVAL;
	if (old_dentry == trap)
		goto exit4;
2805
	new_dentry = lookup_hash(&newnd);
L
Linus Torvalds 已提交
2806 2807 2808 2809 2810 2811 2812 2813
	error = PTR_ERR(new_dentry);
	if (IS_ERR(new_dentry))
		goto exit4;
	/* target should not be an ancestor of source */
	error = -ENOTEMPTY;
	if (new_dentry == trap)
		goto exit5;

2814 2815 2816
	error = mnt_want_write(oldnd.path.mnt);
	if (error)
		goto exit5;
L
Linus Torvalds 已提交
2817 2818
	error = vfs_rename(old_dir->d_inode, old_dentry,
				   new_dir->d_inode, new_dentry);
2819
	mnt_drop_write(oldnd.path.mnt);
L
Linus Torvalds 已提交
2820 2821 2822 2823 2824 2825 2826
exit5:
	dput(new_dentry);
exit4:
	dput(old_dentry);
exit3:
	unlock_rename(new_dir, old_dir);
exit2:
J
Jan Blunck 已提交
2827
	path_put(&newnd.path);
L
Linus Torvalds 已提交
2828
exit1:
J
Jan Blunck 已提交
2829
	path_put(&oldnd.path);
L
Linus Torvalds 已提交
2830 2831 2832 2833
exit:
	return error;
}

2834 2835
asmlinkage long sys_renameat(int olddfd, const char __user *oldname,
			     int newdfd, const char __user *newname)
L
Linus Torvalds 已提交
2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846
{
	int error;
	char * from;
	char * to;

	from = getname(oldname);
	if(IS_ERR(from))
		return PTR_ERR(from);
	to = getname(newname);
	error = PTR_ERR(to);
	if (!IS_ERR(to)) {
2847
		error = do_rename(olddfd, from, newdfd, to);
L
Linus Torvalds 已提交
2848 2849 2850 2851 2852 2853
		putname(to);
	}
	putname(from);
	return error;
}

2854 2855 2856 2857 2858
asmlinkage long sys_rename(const char __user *oldname, const char __user *newname)
{
	return sys_renameat(AT_FDCWD, oldname, AT_FDCWD, newname);
}

L
Linus Torvalds 已提交
2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883
int vfs_readlink(struct dentry *dentry, char __user *buffer, int buflen, const char *link)
{
	int len;

	len = PTR_ERR(link);
	if (IS_ERR(link))
		goto out;

	len = strlen(link);
	if (len > (unsigned) buflen)
		len = buflen;
	if (copy_to_user(buffer, link, len))
		len = -EFAULT;
out:
	return len;
}

/*
 * A helper for ->readlink().  This should be used *ONLY* for symlinks that
 * have ->follow_link() touching nd only in nd_set_link().  Using (or not
 * using) it for any given inode is up to filesystem.
 */
int generic_readlink(struct dentry *dentry, char __user *buffer, int buflen)
{
	struct nameidata nd;
2884
	void *cookie;
2885
	int res;
2886

L
Linus Torvalds 已提交
2887
	nd.depth = 0;
2888
	cookie = dentry->d_inode->i_op->follow_link(dentry, &nd);
2889 2890 2891 2892 2893 2894 2895
	if (IS_ERR(cookie))
		return PTR_ERR(cookie);

	res = vfs_readlink(dentry, buffer, buflen, nd_get_link(&nd));
	if (dentry->d_inode->i_op->put_link)
		dentry->d_inode->i_op->put_link(dentry, &nd, cookie);
	return res;
L
Linus Torvalds 已提交
2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907
}

int vfs_follow_link(struct nameidata *nd, const char *link)
{
	return __vfs_follow_link(nd, link);
}

/* get the link contents into pagecache */
static char *page_getlink(struct dentry * dentry, struct page **ppage)
{
	struct page * page;
	struct address_space *mapping = dentry->d_inode->i_mapping;
2908
	page = read_mapping_page(mapping, 0, NULL);
L
Linus Torvalds 已提交
2909
	if (IS_ERR(page))
2910
		return (char*)page;
L
Linus Torvalds 已提交
2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926
	*ppage = page;
	return kmap(page);
}

int page_readlink(struct dentry *dentry, char __user *buffer, int buflen)
{
	struct page *page = NULL;
	char *s = page_getlink(dentry, &page);
	int res = vfs_readlink(dentry,buffer,buflen,s);
	if (page) {
		kunmap(page);
		page_cache_release(page);
	}
	return res;
}

2927
void *page_follow_link_light(struct dentry *dentry, struct nameidata *nd)
L
Linus Torvalds 已提交
2928
{
2929
	struct page *page = NULL;
L
Linus Torvalds 已提交
2930
	nd_set_link(nd, page_getlink(dentry, &page));
2931
	return page;
L
Linus Torvalds 已提交
2932 2933
}

2934
void page_put_link(struct dentry *dentry, struct nameidata *nd, void *cookie)
L
Linus Torvalds 已提交
2935
{
2936 2937 2938
	struct page *page = cookie;

	if (page) {
L
Linus Torvalds 已提交
2939 2940 2941 2942 2943
		kunmap(page);
		page_cache_release(page);
	}
}

2944 2945
int __page_symlink(struct inode *inode, const char *symname, int len,
		gfp_t gfp_mask)
L
Linus Torvalds 已提交
2946 2947
{
	struct address_space *mapping = inode->i_mapping;
2948
	struct page *page;
2949
	void *fsdata;
2950
	int err;
L
Linus Torvalds 已提交
2951 2952
	char *kaddr;

2953
retry:
2954 2955
	err = pagecache_write_begin(NULL, mapping, 0, len-1,
				AOP_FLAG_UNINTERRUPTIBLE, &page, &fsdata);
L
Linus Torvalds 已提交
2956
	if (err)
2957 2958
		goto fail;

L
Linus Torvalds 已提交
2959 2960 2961
	kaddr = kmap_atomic(page, KM_USER0);
	memcpy(kaddr, symname, len-1);
	kunmap_atomic(kaddr, KM_USER0);
2962 2963 2964

	err = pagecache_write_end(NULL, mapping, 0, len-1, len-1,
							page, fsdata);
L
Linus Torvalds 已提交
2965 2966
	if (err < 0)
		goto fail;
2967 2968 2969
	if (err < len-1)
		goto retry;

L
Linus Torvalds 已提交
2970 2971 2972 2973 2974 2975
	mark_inode_dirty(inode);
	return 0;
fail:
	return err;
}

2976 2977 2978 2979 2980 2981
int page_symlink(struct inode *inode, const char *symname, int len)
{
	return __page_symlink(inode, symname, len,
			mapping_gfp_mask(inode->i_mapping));
}

2982
const struct inode_operations page_symlink_inode_operations = {
L
Linus Torvalds 已提交
2983 2984 2985 2986 2987 2988
	.readlink	= generic_readlink,
	.follow_link	= page_follow_link_light,
	.put_link	= page_put_link,
};

EXPORT_SYMBOL(__user_walk);
2989
EXPORT_SYMBOL(__user_walk_fd);
L
Linus Torvalds 已提交
2990 2991 2992 2993 2994 2995 2996 2997 2998
EXPORT_SYMBOL(follow_down);
EXPORT_SYMBOL(follow_up);
EXPORT_SYMBOL(get_write_access); /* binfmt_aout */
EXPORT_SYMBOL(getname);
EXPORT_SYMBOL(lock_rename);
EXPORT_SYMBOL(lookup_one_len);
EXPORT_SYMBOL(page_follow_link_light);
EXPORT_SYMBOL(page_put_link);
EXPORT_SYMBOL(page_readlink);
2999
EXPORT_SYMBOL(__page_symlink);
L
Linus Torvalds 已提交
3000 3001 3002
EXPORT_SYMBOL(page_symlink);
EXPORT_SYMBOL(page_symlink_inode_operations);
EXPORT_SYMBOL(path_lookup);
3003
EXPORT_SYMBOL(vfs_path_lookup);
L
Linus Torvalds 已提交
3004
EXPORT_SYMBOL(permission);
3005
EXPORT_SYMBOL(vfs_permission);
3006
EXPORT_SYMBOL(file_permission);
L
Linus Torvalds 已提交
3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020
EXPORT_SYMBOL(unlock_rename);
EXPORT_SYMBOL(vfs_create);
EXPORT_SYMBOL(vfs_follow_link);
EXPORT_SYMBOL(vfs_link);
EXPORT_SYMBOL(vfs_mkdir);
EXPORT_SYMBOL(vfs_mknod);
EXPORT_SYMBOL(generic_permission);
EXPORT_SYMBOL(vfs_readlink);
EXPORT_SYMBOL(vfs_rename);
EXPORT_SYMBOL(vfs_rmdir);
EXPORT_SYMBOL(vfs_symlink);
EXPORT_SYMBOL(vfs_unlink);
EXPORT_SYMBOL(dentry_unhash);
EXPORT_SYMBOL(generic_readlink);