open.c 27.9 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9
/*
 *  linux/fs/open.c
 *
 *  Copyright (C) 1991, 1992  Linus Torvalds
 */

#include <linux/string.h>
#include <linux/mm.h>
#include <linux/file.h>
A
Al Viro 已提交
10
#include <linux/fdtable.h>
L
Linus Torvalds 已提交
11
#include <linux/quotaops.h>
R
Robert Love 已提交
12
#include <linux/fsnotify.h>
L
Linus Torvalds 已提交
13 14 15 16 17
#include <linux/module.h>
#include <linux/slab.h>
#include <linux/tty.h>
#include <linux/namei.h>
#include <linux/backing-dev.h>
18
#include <linux/capability.h>
19
#include <linux/securebits.h>
L
Linus Torvalds 已提交
20 21 22
#include <linux/security.h>
#include <linux/mount.h>
#include <linux/vfs.h>
23
#include <linux/fcntl.h>
L
Linus Torvalds 已提交
24 25
#include <asm/uaccess.h>
#include <linux/fs.h>
26
#include <linux/personality.h>
L
Linus Torvalds 已提交
27 28
#include <linux/pagemap.h>
#include <linux/syscalls.h>
29
#include <linux/rcupdate.h>
30
#include <linux/audit.h>
31
#include <linux/falloc.h>
L
Linus Torvalds 已提交
32

33
int vfs_statfs(struct dentry *dentry, struct kstatfs *buf)
L
Linus Torvalds 已提交
34 35 36
{
	int retval = -ENODEV;

37
	if (dentry) {
L
Linus Torvalds 已提交
38
		retval = -ENOSYS;
39
		if (dentry->d_sb->s_op->statfs) {
L
Linus Torvalds 已提交
40
			memset(buf, 0, sizeof(*buf));
41
			retval = security_sb_statfs(dentry);
L
Linus Torvalds 已提交
42 43
			if (retval)
				return retval;
44
			retval = dentry->d_sb->s_op->statfs(dentry, buf);
L
Linus Torvalds 已提交
45 46 47 48 49 50 51 52 53
			if (retval == 0 && buf->f_frsize == 0)
				buf->f_frsize = buf->f_bsize;
		}
	}
	return retval;
}

EXPORT_SYMBOL(vfs_statfs);

54
static int vfs_statfs_native(struct dentry *dentry, struct statfs *buf)
L
Linus Torvalds 已提交
55 56 57 58
{
	struct kstatfs st;
	int retval;

59
	retval = vfs_statfs(dentry, &st);
L
Linus Torvalds 已提交
60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96
	if (retval)
		return retval;

	if (sizeof(*buf) == sizeof(st))
		memcpy(buf, &st, sizeof(st));
	else {
		if (sizeof buf->f_blocks == 4) {
			if ((st.f_blocks | st.f_bfree | st.f_bavail) &
			    0xffffffff00000000ULL)
				return -EOVERFLOW;
			/*
			 * f_files and f_ffree may be -1; it's okay to stuff
			 * that into 32 bits
			 */
			if (st.f_files != -1 &&
			    (st.f_files & 0xffffffff00000000ULL))
				return -EOVERFLOW;
			if (st.f_ffree != -1 &&
			    (st.f_ffree & 0xffffffff00000000ULL))
				return -EOVERFLOW;
		}

		buf->f_type = st.f_type;
		buf->f_bsize = st.f_bsize;
		buf->f_blocks = st.f_blocks;
		buf->f_bfree = st.f_bfree;
		buf->f_bavail = st.f_bavail;
		buf->f_files = st.f_files;
		buf->f_ffree = st.f_ffree;
		buf->f_fsid = st.f_fsid;
		buf->f_namelen = st.f_namelen;
		buf->f_frsize = st.f_frsize;
		memset(buf->f_spare, 0, sizeof(buf->f_spare));
	}
	return 0;
}

97
static int vfs_statfs64(struct dentry *dentry, struct statfs64 *buf)
L
Linus Torvalds 已提交
98 99 100 101
{
	struct kstatfs st;
	int retval;

102
	retval = vfs_statfs(dentry, &st);
L
Linus Torvalds 已提交
103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131
	if (retval)
		return retval;

	if (sizeof(*buf) == sizeof(st))
		memcpy(buf, &st, sizeof(st));
	else {
		buf->f_type = st.f_type;
		buf->f_bsize = st.f_bsize;
		buf->f_blocks = st.f_blocks;
		buf->f_bfree = st.f_bfree;
		buf->f_bavail = st.f_bavail;
		buf->f_files = st.f_files;
		buf->f_ffree = st.f_ffree;
		buf->f_fsid = st.f_fsid;
		buf->f_namelen = st.f_namelen;
		buf->f_frsize = st.f_frsize;
		memset(buf->f_spare, 0, sizeof(buf->f_spare));
	}
	return 0;
}

asmlinkage long sys_statfs(const char __user * path, struct statfs __user * buf)
{
	struct nameidata nd;
	int error;

	error = user_path_walk(path, &nd);
	if (!error) {
		struct statfs tmp;
132
		error = vfs_statfs_native(nd.path.dentry, &tmp);
L
Linus Torvalds 已提交
133 134
		if (!error && copy_to_user(buf, &tmp, sizeof(tmp)))
			error = -EFAULT;
J
Jan Blunck 已提交
135
		path_put(&nd.path);
L
Linus Torvalds 已提交
136 137 138 139 140 141 142 143 144 145 146 147 148 149 150
	}
	return error;
}


asmlinkage long sys_statfs64(const char __user *path, size_t sz, struct statfs64 __user *buf)
{
	struct nameidata nd;
	long error;

	if (sz != sizeof(*buf))
		return -EINVAL;
	error = user_path_walk(path, &nd);
	if (!error) {
		struct statfs64 tmp;
151
		error = vfs_statfs64(nd.path.dentry, &tmp);
L
Linus Torvalds 已提交
152 153
		if (!error && copy_to_user(buf, &tmp, sizeof(tmp)))
			error = -EFAULT;
J
Jan Blunck 已提交
154
		path_put(&nd.path);
L
Linus Torvalds 已提交
155 156 157 158 159 160 161 162 163 164 165 166 167 168 169
	}
	return error;
}


asmlinkage long sys_fstatfs(unsigned int fd, struct statfs __user * buf)
{
	struct file * file;
	struct statfs tmp;
	int error;

	error = -EBADF;
	file = fget(fd);
	if (!file)
		goto out;
170
	error = vfs_statfs_native(file->f_path.dentry, &tmp);
L
Linus Torvalds 已提交
171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190
	if (!error && copy_to_user(buf, &tmp, sizeof(tmp)))
		error = -EFAULT;
	fput(file);
out:
	return error;
}

asmlinkage long sys_fstatfs64(unsigned int fd, size_t sz, struct statfs64 __user *buf)
{
	struct file * file;
	struct statfs64 tmp;
	int error;

	if (sz != sizeof(*buf))
		return -EINVAL;

	error = -EBADF;
	file = fget(fd);
	if (!file)
		goto out;
191
	error = vfs_statfs64(file->f_path.dentry, &tmp);
L
Linus Torvalds 已提交
192 193 194 195 196 197 198
	if (!error && copy_to_user(buf, &tmp, sizeof(tmp)))
		error = -EFAULT;
	fput(file);
out:
	return error;
}

199 200
int do_truncate(struct dentry *dentry, loff_t length, unsigned int time_attrs,
	struct file *filp)
L
Linus Torvalds 已提交
201 202 203 204 205 206 207 208 209
{
	int err;
	struct iattr newattrs;

	/* Not pretty: "inode->i_size" shouldn't really be signed. But it is. */
	if (length < 0)
		return -EINVAL;

	newattrs.ia_size = length;
210
	newattrs.ia_valid = ATTR_SIZE | time_attrs;
211 212 213 214
	if (filp) {
		newattrs.ia_file = filp;
		newattrs.ia_valid |= ATTR_FILE;
	}
L
Linus Torvalds 已提交
215

216 217 218
	/* Remove suid/sgid on truncate too */
	newattrs.ia_valid |= should_remove_suid(dentry);

219
	mutex_lock(&dentry->d_inode->i_mutex);
L
Linus Torvalds 已提交
220
	err = notify_change(dentry, &newattrs);
221
	mutex_unlock(&dentry->d_inode->i_mutex);
L
Linus Torvalds 已提交
222 223 224
	return err;
}

225
static long do_sys_truncate(const char __user * path, loff_t length)
L
Linus Torvalds 已提交
226 227 228 229 230 231 232 233 234 235 236 237
{
	struct nameidata nd;
	struct inode * inode;
	int error;

	error = -EINVAL;
	if (length < 0)	/* sorry, but loff_t says... */
		goto out;

	error = user_path_walk(path, &nd);
	if (error)
		goto out;
238
	inode = nd.path.dentry->d_inode;
L
Linus Torvalds 已提交
239 240 241 242 243 244 245 246 247 248

	/* For directories it's -EISDIR, for other non-regulars - -EINVAL */
	error = -EISDIR;
	if (S_ISDIR(inode->i_mode))
		goto dput_and_out;

	error = -EINVAL;
	if (!S_ISREG(inode->i_mode))
		goto dput_and_out;

249
	error = mnt_want_write(nd.path.mnt);
L
Linus Torvalds 已提交
250 251 252
	if (error)
		goto dput_and_out;

253 254 255
	error = vfs_permission(&nd, MAY_WRITE);
	if (error)
		goto mnt_drop_write_and_out;
L
Linus Torvalds 已提交
256 257 258

	error = -EPERM;
	if (IS_IMMUTABLE(inode) || IS_APPEND(inode))
259
		goto mnt_drop_write_and_out;
L
Linus Torvalds 已提交
260

261
	error = get_write_access(inode);
L
Linus Torvalds 已提交
262
	if (error)
263
		goto mnt_drop_write_and_out;
L
Linus Torvalds 已提交
264

265 266 267 268 269
	/*
	 * Make sure that there are no leases.  get_write_access() protects
	 * against the truncate racing with a lease-granting setlease().
	 */
	error = break_lease(inode, FMODE_WRITE);
L
Linus Torvalds 已提交
270
	if (error)
271
		goto put_write_and_out;
L
Linus Torvalds 已提交
272 273 274 275

	error = locks_verify_truncate(inode, NULL, length);
	if (!error) {
		DQUOT_INIT(inode);
276
		error = do_truncate(nd.path.dentry, length, 0, NULL);
L
Linus Torvalds 已提交
277 278
	}

279 280
put_write_and_out:
	put_write_access(inode);
281 282
mnt_drop_write_and_out:
	mnt_drop_write(nd.path.mnt);
L
Linus Torvalds 已提交
283
dput_and_out:
J
Jan Blunck 已提交
284
	path_put(&nd.path);
L
Linus Torvalds 已提交
285 286 287 288 289 290 291 292 293 294
out:
	return error;
}

asmlinkage long sys_truncate(const char __user * path, unsigned long length)
{
	/* on 32-bit boxen it will cut the range 2^31--2^32-1 off */
	return do_sys_truncate(path, (long)length);
}

295
static long do_sys_ftruncate(unsigned int fd, loff_t length, int small)
L
Linus Torvalds 已提交
296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313
{
	struct inode * inode;
	struct dentry *dentry;
	struct file * file;
	int error;

	error = -EINVAL;
	if (length < 0)
		goto out;
	error = -EBADF;
	file = fget(fd);
	if (!file)
		goto out;

	/* explicitly opened as large or we are on 64-bit box */
	if (file->f_flags & O_LARGEFILE)
		small = 0;

314
	dentry = file->f_path.dentry;
L
Linus Torvalds 已提交
315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330
	inode = dentry->d_inode;
	error = -EINVAL;
	if (!S_ISREG(inode->i_mode) || !(file->f_mode & FMODE_WRITE))
		goto out_putf;

	error = -EINVAL;
	/* Cannot ftruncate over 2^31 bytes without large file support */
	if (small && length > MAX_NON_LFS)
		goto out_putf;

	error = -EPERM;
	if (IS_APPEND(inode))
		goto out_putf;

	error = locks_verify_truncate(inode, file, length);
	if (!error)
331
		error = do_truncate(dentry, length, ATTR_MTIME|ATTR_CTIME, file);
L
Linus Torvalds 已提交
332 333 334 335 336 337 338 339
out_putf:
	fput(file);
out:
	return error;
}

asmlinkage long sys_ftruncate(unsigned int fd, unsigned long length)
{
340
	long ret = do_sys_ftruncate(fd, length, 1);
341
	/* avoid REGPARM breakage on x86: */
342
	asmlinkage_protect(2, ret, fd, length);
343
	return ret;
L
Linus Torvalds 已提交
344 345 346 347 348 349 350 351 352 353 354
}

/* LFS versions of truncate are only needed on 32 bit machines */
#if BITS_PER_LONG == 32
asmlinkage long sys_truncate64(const char __user * path, loff_t length)
{
	return do_sys_truncate(path, length);
}

asmlinkage long sys_ftruncate64(unsigned int fd, loff_t length)
{
355
	long ret = do_sys_ftruncate(fd, length, 0);
356
	/* avoid REGPARM breakage on x86: */
357
	asmlinkage_protect(2, ret, fd, length);
358
	return ret;
L
Linus Torvalds 已提交
359 360 361
}
#endif

362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411
asmlinkage long sys_fallocate(int fd, int mode, loff_t offset, loff_t len)
{
	struct file *file;
	struct inode *inode;
	long ret = -EINVAL;

	if (offset < 0 || len <= 0)
		goto out;

	/* Return error if mode is not supported */
	ret = -EOPNOTSUPP;
	if (mode && !(mode & FALLOC_FL_KEEP_SIZE))
		goto out;

	ret = -EBADF;
	file = fget(fd);
	if (!file)
		goto out;
	if (!(file->f_mode & FMODE_WRITE))
		goto out_fput;
	/*
	 * Revalidate the write permissions, in case security policy has
	 * changed since the files were opened.
	 */
	ret = security_file_permission(file, MAY_WRITE);
	if (ret)
		goto out_fput;

	inode = file->f_path.dentry->d_inode;

	ret = -ESPIPE;
	if (S_ISFIFO(inode->i_mode))
		goto out_fput;

	ret = -ENODEV;
	/*
	 * Let individual file system decide if it supports preallocation
	 * for directories or not.
	 */
	if (!S_ISREG(inode->i_mode) && !S_ISDIR(inode->i_mode))
		goto out_fput;

	ret = -EFBIG;
	/* Check for wrap through zero too */
	if (((offset + len) > inode->i_sb->s_maxbytes) || ((offset + len) < 0))
		goto out_fput;

	if (inode->i_op && inode->i_op->fallocate)
		ret = inode->i_op->fallocate(inode, mode, offset, len);
	else
412
		ret = -EOPNOTSUPP;
413 414 415 416 417 418 419

out_fput:
	fput(file);
out:
	return ret;
}

L
Linus Torvalds 已提交
420 421 422 423 424
/*
 * access() needs to use the real uid/gid, not the effective uid/gid.
 * We do this by temporarily clearing all FS-related capabilities and
 * switching the fsuid/fsgid around to the real ones.
 */
425
asmlinkage long sys_faccessat(int dfd, const char __user *filename, int mode)
L
Linus Torvalds 已提交
426 427 428
{
	struct nameidata nd;
	int old_fsuid, old_fsgid;
429
	kernel_cap_t uninitialized_var(old_cap);  /* !SECURE_NO_SETUID_FIXUP */
L
Linus Torvalds 已提交
430 431 432 433 434 435 436 437 438 439 440
	int res;

	if (mode & ~S_IRWXO)	/* where's F_OK, X_OK, W_OK, R_OK? */
		return -EINVAL;

	old_fsuid = current->fsuid;
	old_fsgid = current->fsgid;

	current->fsuid = current->uid;
	current->fsgid = current->gid;

441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457
	if (!issecure(SECURE_NO_SETUID_FIXUP)) {
		/*
		 * Clear the capabilities if we switch to a non-root user
		 */
#ifndef CONFIG_SECURITY_FILE_CAPABILITIES
		/*
		 * FIXME: There is a race here against sys_capset.  The
		 * capabilities can change yet we will restore the old
		 * value below.  We should hold task_capabilities_lock,
		 * but we cannot because user_path_walk can sleep.
		 */
#endif /* ndef CONFIG_SECURITY_FILE_CAPABILITIES */
		if (current->uid)
			old_cap = cap_set_effective(__cap_empty_set);
		else
			old_cap = cap_set_effective(current->cap_permitted);
	}
L
Linus Torvalds 已提交
458

459
	res = __user_walk_fd(dfd, filename, LOOKUP_FOLLOW|LOOKUP_ACCESS, &nd);
460 461 462 463 464 465
	if (res)
		goto out;

	res = vfs_permission(&nd, mode);
	/* SuS v2 requires we report a read only fs too */
	if(res || !(mode & S_IWOTH) ||
466
	   special_file(nd.path.dentry->d_inode->i_mode))
467
		goto out_path_release;
468 469 470 471 472 473 474 475 476 477 478
	/*
	 * This is a rare case where using __mnt_is_readonly()
	 * is OK without a mnt_want/drop_write() pair.  Since
	 * no actual write to the fs is performed here, we do
	 * not need to telegraph to that to anyone.
	 *
	 * By doing this, we accept that this access is
	 * inherently racy and know that the fs may change
	 * state before we even see this result.
	 */
	if (__mnt_is_readonly(nd.path.mnt))
479
		res = -EROFS;
L
Linus Torvalds 已提交
480

481
out_path_release:
J
Jan Blunck 已提交
482
	path_put(&nd.path);
483
out:
L
Linus Torvalds 已提交
484 485
	current->fsuid = old_fsuid;
	current->fsgid = old_fsgid;
486 487 488

	if (!issecure(SECURE_NO_SETUID_FIXUP))
		cap_set_effective(old_cap);
L
Linus Torvalds 已提交
489 490 491 492

	return res;
}

493 494 495 496 497
asmlinkage long sys_access(const char __user *filename, int mode)
{
	return sys_faccessat(AT_FDCWD, filename, mode);
}

L
Linus Torvalds 已提交
498 499 500 501 502
asmlinkage long sys_chdir(const char __user * filename)
{
	struct nameidata nd;
	int error;

503 504
	error = __user_walk(filename,
			    LOOKUP_FOLLOW|LOOKUP_DIRECTORY|LOOKUP_CHDIR, &nd);
L
Linus Torvalds 已提交
505 506 507
	if (error)
		goto out;

508
	error = vfs_permission(&nd, MAY_EXEC);
L
Linus Torvalds 已提交
509 510 511
	if (error)
		goto dput_and_out;

512
	set_fs_pwd(current->fs, &nd.path);
L
Linus Torvalds 已提交
513 514

dput_and_out:
J
Jan Blunck 已提交
515
	path_put(&nd.path);
L
Linus Torvalds 已提交
516 517 518 519 520 521 522 523 524 525 526 527 528 529 530
out:
	return error;
}

asmlinkage long sys_fchdir(unsigned int fd)
{
	struct file *file;
	struct inode *inode;
	int error;

	error = -EBADF;
	file = fget(fd);
	if (!file)
		goto out;

531
	inode = file->f_path.dentry->d_inode;
L
Linus Torvalds 已提交
532 533 534 535 536

	error = -ENOTDIR;
	if (!S_ISDIR(inode->i_mode))
		goto out_putf;

537
	error = file_permission(file, MAY_EXEC);
L
Linus Torvalds 已提交
538
	if (!error)
539
		set_fs_pwd(current->fs, &file->f_path);
L
Linus Torvalds 已提交
540 541 542 543 544 545 546 547 548 549 550 551 552 553 554
out_putf:
	fput(file);
out:
	return error;
}

asmlinkage long sys_chroot(const char __user * filename)
{
	struct nameidata nd;
	int error;

	error = __user_walk(filename, LOOKUP_FOLLOW | LOOKUP_DIRECTORY | LOOKUP_NOALT, &nd);
	if (error)
		goto out;

555
	error = vfs_permission(&nd, MAY_EXEC);
L
Linus Torvalds 已提交
556 557 558 559 560 561 562
	if (error)
		goto dput_and_out;

	error = -EPERM;
	if (!capable(CAP_SYS_CHROOT))
		goto dput_and_out;

563
	set_fs_root(current->fs, &nd.path);
L
Linus Torvalds 已提交
564 565 566
	set_fs_altroot();
	error = 0;
dput_and_out:
J
Jan Blunck 已提交
567
	path_put(&nd.path);
L
Linus Torvalds 已提交
568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583
out:
	return error;
}

asmlinkage long sys_fchmod(unsigned int fd, mode_t mode)
{
	struct inode * inode;
	struct dentry * dentry;
	struct file * file;
	int err = -EBADF;
	struct iattr newattrs;

	file = fget(fd);
	if (!file)
		goto out;

584
	dentry = file->f_path.dentry;
L
Linus Torvalds 已提交
585 586
	inode = dentry->d_inode;

587
	audit_inode(NULL, dentry);
588

589 590
	err = mnt_want_write(file->f_path.mnt);
	if (err)
L
Linus Torvalds 已提交
591 592 593
		goto out_putf;
	err = -EPERM;
	if (IS_IMMUTABLE(inode) || IS_APPEND(inode))
594
		goto out_drop_write;
595
	mutex_lock(&inode->i_mutex);
L
Linus Torvalds 已提交
596 597 598 599 600
	if (mode == (mode_t) -1)
		mode = inode->i_mode;
	newattrs.ia_mode = (mode & S_IALLUGO) | (inode->i_mode & ~S_IALLUGO);
	newattrs.ia_valid = ATTR_MODE | ATTR_CTIME;
	err = notify_change(dentry, &newattrs);
601
	mutex_unlock(&inode->i_mutex);
L
Linus Torvalds 已提交
602

603 604
out_drop_write:
	mnt_drop_write(file->f_path.mnt);
L
Linus Torvalds 已提交
605 606 607 608 609 610
out_putf:
	fput(file);
out:
	return err;
}

611 612
asmlinkage long sys_fchmodat(int dfd, const char __user *filename,
			     mode_t mode)
L
Linus Torvalds 已提交
613 614 615 616 617 618
{
	struct nameidata nd;
	struct inode * inode;
	int error;
	struct iattr newattrs;

619
	error = __user_walk_fd(dfd, filename, LOOKUP_FOLLOW, &nd);
L
Linus Torvalds 已提交
620 621
	if (error)
		goto out;
622
	inode = nd.path.dentry->d_inode;
L
Linus Torvalds 已提交
623

624 625
	error = mnt_want_write(nd.path.mnt);
	if (error)
L
Linus Torvalds 已提交
626 627 628 629
		goto dput_and_out;

	error = -EPERM;
	if (IS_IMMUTABLE(inode) || IS_APPEND(inode))
630
		goto out_drop_write;
L
Linus Torvalds 已提交
631

632
	mutex_lock(&inode->i_mutex);
L
Linus Torvalds 已提交
633 634 635 636
	if (mode == (mode_t) -1)
		mode = inode->i_mode;
	newattrs.ia_mode = (mode & S_IALLUGO) | (inode->i_mode & ~S_IALLUGO);
	newattrs.ia_valid = ATTR_MODE | ATTR_CTIME;
637
	error = notify_change(nd.path.dentry, &newattrs);
638
	mutex_unlock(&inode->i_mutex);
L
Linus Torvalds 已提交
639

640 641
out_drop_write:
	mnt_drop_write(nd.path.mnt);
L
Linus Torvalds 已提交
642
dput_and_out:
J
Jan Blunck 已提交
643
	path_put(&nd.path);
L
Linus Torvalds 已提交
644 645 646 647
out:
	return error;
}

648 649 650 651 652
asmlinkage long sys_chmod(const char __user *filename, mode_t mode)
{
	return sys_fchmodat(AT_FDCWD, filename, mode);
}

L
Linus Torvalds 已提交
653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676
static int chown_common(struct dentry * dentry, uid_t user, gid_t group)
{
	struct inode * inode;
	int error;
	struct iattr newattrs;

	error = -ENOENT;
	if (!(inode = dentry->d_inode)) {
		printk(KERN_ERR "chown_common: NULL inode\n");
		goto out;
	}
	error = -EPERM;
	if (IS_IMMUTABLE(inode) || IS_APPEND(inode))
		goto out;
	newattrs.ia_valid =  ATTR_CTIME;
	if (user != (uid_t) -1) {
		newattrs.ia_valid |= ATTR_UID;
		newattrs.ia_uid = user;
	}
	if (group != (gid_t) -1) {
		newattrs.ia_valid |= ATTR_GID;
		newattrs.ia_gid = group;
	}
	if (!S_ISDIR(inode->i_mode))
677 678
		newattrs.ia_valid |=
			ATTR_KILL_SUID | ATTR_KILL_SGID | ATTR_KILL_PRIV;
679
	mutex_lock(&inode->i_mutex);
L
Linus Torvalds 已提交
680
	error = notify_change(dentry, &newattrs);
681
	mutex_unlock(&inode->i_mutex);
L
Linus Torvalds 已提交
682 683 684 685 686 687 688 689 690 691
out:
	return error;
}

asmlinkage long sys_chown(const char __user * filename, uid_t user, gid_t group)
{
	struct nameidata nd;
	int error;

	error = user_path_walk(filename, &nd);
692 693
	if (error)
		goto out;
694 695 696
	error = mnt_want_write(nd.path.mnt);
	if (error)
		goto out_release;
697
	error = chown_common(nd.path.dentry, user, group);
698 699
	mnt_drop_write(nd.path.mnt);
out_release:
J
Jan Blunck 已提交
700
	path_put(&nd.path);
701
out:
L
Linus Torvalds 已提交
702 703 704
	return error;
}

705 706 707 708 709 710 711 712 713 714 715 716
asmlinkage long sys_fchownat(int dfd, const char __user *filename, uid_t user,
			     gid_t group, int flag)
{
	struct nameidata nd;
	int error = -EINVAL;
	int follow;

	if ((flag & ~AT_SYMLINK_NOFOLLOW) != 0)
		goto out;

	follow = (flag & AT_SYMLINK_NOFOLLOW) ? 0 : LOOKUP_FOLLOW;
	error = __user_walk_fd(dfd, filename, follow, &nd);
717 718
	if (error)
		goto out;
719 720 721
	error = mnt_want_write(nd.path.mnt);
	if (error)
		goto out_release;
722
	error = chown_common(nd.path.dentry, user, group);
723 724
	mnt_drop_write(nd.path.mnt);
out_release:
J
Jan Blunck 已提交
725
	path_put(&nd.path);
726 727 728 729
out:
	return error;
}

L
Linus Torvalds 已提交
730 731 732 733 734 735
asmlinkage long sys_lchown(const char __user * filename, uid_t user, gid_t group)
{
	struct nameidata nd;
	int error;

	error = user_path_walk_link(filename, &nd);
736 737
	if (error)
		goto out;
738 739 740
	error = mnt_want_write(nd.path.mnt);
	if (error)
		goto out_release;
741
	error = chown_common(nd.path.dentry, user, group);
742 743
	mnt_drop_write(nd.path.mnt);
out_release:
J
Jan Blunck 已提交
744
	path_put(&nd.path);
745
out:
L
Linus Torvalds 已提交
746 747 748 749 750 751 752 753
	return error;
}


asmlinkage long sys_fchown(unsigned int fd, uid_t user, gid_t group)
{
	struct file * file;
	int error = -EBADF;
754
	struct dentry * dentry;
L
Linus Torvalds 已提交
755 756

	file = fget(fd);
757 758 759
	if (!file)
		goto out;

760 761 762
	error = mnt_want_write(file->f_path.mnt);
	if (error)
		goto out_fput;
763
	dentry = file->f_path.dentry;
764
	audit_inode(NULL, dentry);
765
	error = chown_common(dentry, user, group);
766 767
	mnt_drop_write(file->f_path.mnt);
out_fput:
768 769
	fput(file);
out:
L
Linus Torvalds 已提交
770 771 772
	return error;
}

773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801
/*
 * You have to be very careful that these write
 * counts get cleaned up in error cases and
 * upon __fput().  This should probably never
 * be called outside of __dentry_open().
 */
static inline int __get_file_write_access(struct inode *inode,
					  struct vfsmount *mnt)
{
	int error;
	error = get_write_access(inode);
	if (error)
		return error;
	/*
	 * Do not take mount writer counts on
	 * special files since no writes to
	 * the mount itself will occur.
	 */
	if (!special_file(inode->i_mode)) {
		/*
		 * Balanced in __fput()
		 */
		error = mnt_want_write(mnt);
		if (error)
			put_write_access(inode);
	}
	return error;
}

802
static struct file *__dentry_open(struct dentry *dentry, struct vfsmount *mnt,
803 804
					int flags, struct file *f,
					int (*open)(struct inode *, struct file *))
L
Linus Torvalds 已提交
805 806 807 808 809
{
	struct inode *inode;
	int error;

	f->f_flags = flags;
810 811
	f->f_mode = ((flags+1) & O_ACCMODE) | FMODE_LSEEK |
				FMODE_PREAD | FMODE_PWRITE;
L
Linus Torvalds 已提交
812 813
	inode = dentry->d_inode;
	if (f->f_mode & FMODE_WRITE) {
814
		error = __get_file_write_access(inode, mnt);
L
Linus Torvalds 已提交
815 816
		if (error)
			goto cleanup_file;
817 818
		if (!special_file(inode->i_mode))
			file_take_write(f);
L
Linus Torvalds 已提交
819 820 821
	}

	f->f_mapping = inode->i_mapping;
822 823
	f->f_path.dentry = dentry;
	f->f_path.mnt = mnt;
L
Linus Torvalds 已提交
824 825 826 827
	f->f_pos = 0;
	f->f_op = fops_get(inode->i_fop);
	file_move(f, &inode->i_sb->s_files);

828 829 830 831
	error = security_dentry_open(f);
	if (error)
		goto cleanup_all;

832 833 834 835
	if (!open && f->f_op)
		open = f->f_op->open;
	if (open) {
		error = open(inode, f);
L
Linus Torvalds 已提交
836 837 838
		if (error)
			goto cleanup_all;
	}
839

L
Linus Torvalds 已提交
840 841 842 843 844 845
	f->f_flags &= ~(O_CREAT | O_EXCL | O_NOCTTY | O_TRUNC);

	file_ra_state_init(&f->f_ra, f->f_mapping->host->i_mapping);

	/* NB: we're sure to have correct a_ops only after f_op->open */
	if (f->f_flags & O_DIRECT) {
846 847
		if (!f->f_mapping->a_ops ||
		    ((!f->f_mapping->a_ops->direct_IO) &&
848
		    (!f->f_mapping->a_ops->get_xip_mem))) {
L
Linus Torvalds 已提交
849 850 851 852 853 854 855 856 857
			fput(f);
			f = ERR_PTR(-EINVAL);
		}
	}

	return f;

cleanup_all:
	fops_put(f->f_op);
858
	if (f->f_mode & FMODE_WRITE) {
L
Linus Torvalds 已提交
859
		put_write_access(inode);
860 861 862 863 864 865 866 867
		if (!special_file(inode->i_mode)) {
			/*
			 * We don't consider this a real
			 * mnt_want/drop_write() pair
			 * because it all happenend right
			 * here, so just reset the state.
			 */
			file_reset_write(f);
868
			mnt_drop_write(mnt);
869
		}
870
	}
L
Linus Torvalds 已提交
871
	file_kill(f);
872 873
	f->f_path.dentry = NULL;
	f->f_path.mnt = NULL;
L
Linus Torvalds 已提交
874 875 876 877 878 879 880
cleanup_file:
	put_filp(f);
	dput(dentry);
	mntput(mnt);
	return ERR_PTR(error);
}

881 882 883 884 885 886 887 888 889 890
/**
 * lookup_instantiate_filp - instantiates the open intent filp
 * @nd: pointer to nameidata
 * @dentry: pointer to dentry
 * @open: open callback
 *
 * Helper for filesystems that want to use lookup open intents and pass back
 * a fully instantiated struct file to the caller.
 * This function is meant to be called from within a filesystem's
 * lookup method.
891 892 893 894
 * Beware of calling it for non-regular files! Those ->open methods might block
 * (e.g. in fifo_open), leaving you with parent locked (and in case of fifo,
 * leading to a deadlock, as nobody can open that fifo anymore, because
 * another process to open fifo will block on locked parent when doing lookup).
895 896 897 898 899 900 901 902 903 904 905 906
 * Note that in case of error, nd->intent.open.file is destroyed, but the
 * path information remains valid.
 * If the open callback is set to NULL, then the standard f_op->open()
 * filesystem callback is substituted.
 */
struct file *lookup_instantiate_filp(struct nameidata *nd, struct dentry *dentry,
		int (*open)(struct inode *, struct file *))
{
	if (IS_ERR(nd->intent.open.file))
		goto out;
	if (IS_ERR(dentry))
		goto out_err;
907
	nd->intent.open.file = __dentry_open(dget(dentry), mntget(nd->path.mnt),
908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933
					     nd->intent.open.flags - 1,
					     nd->intent.open.file,
					     open);
out:
	return nd->intent.open.file;
out_err:
	release_open_intent(nd);
	nd->intent.open.file = (struct file *)dentry;
	goto out;
}
EXPORT_SYMBOL_GPL(lookup_instantiate_filp);

/**
 * nameidata_to_filp - convert a nameidata to an open filp.
 * @nd: pointer to nameidata
 * @flags: open flags
 *
 * Note that this function destroys the original nameidata
 */
struct file *nameidata_to_filp(struct nameidata *nd, int flags)
{
	struct file *filp;

	/* Pick up the filp from the open intent */
	filp = nd->intent.open.file;
	/* Has the filesystem initialised the file for us? */
934
	if (filp->f_path.dentry == NULL)
935 936
		filp = __dentry_open(nd->path.dentry, nd->path.mnt, flags, filp,
				     NULL);
937
	else
J
Jan Blunck 已提交
938
		path_put(&nd->path);
939 940 941
	return filp;
}

942 943 944 945
/*
 * dentry_open() will have done dput(dentry) and mntput(mnt) if it returns an
 * error.
 */
946 947 948 949 950
struct file *dentry_open(struct dentry *dentry, struct vfsmount *mnt, int flags)
{
	int error;
	struct file *f;

951 952 953 954 955 956 957 958 959 960 961 962
	/*
	 * We must always pass in a valid mount pointer.   Historically
	 * callers got away with not passing it, but we must enforce this at
	 * the earliest possible point now to avoid strange problems deep in the
	 * filesystem stack.
	 */
	if (!mnt) {
		printk(KERN_WARNING "%s called with NULL vfsmount\n", __func__);
		dump_stack();
		return ERR_PTR(-EINVAL);
	}

963 964
	error = -ENFILE;
	f = get_empty_filp();
965 966 967
	if (f == NULL) {
		dput(dentry);
		mntput(mnt);
968
		return ERR_PTR(error);
969
	}
970

971
	return __dentry_open(dentry, mnt, flags, f, NULL);
972
}
L
Linus Torvalds 已提交
973 974 975 976 977
EXPORT_SYMBOL(dentry_open);

/*
 * Find an empty file descriptor entry, and mark it busy.
 */
U
Ulrich Drepper 已提交
978
int get_unused_fd_flags(int flags)
L
Linus Torvalds 已提交
979 980 981
{
	struct files_struct * files = current->files;
	int fd, error;
982
	struct fdtable *fdt;
L
Linus Torvalds 已提交
983 984 985 986 987

  	error = -EMFILE;
	spin_lock(&files->file_lock);

repeat:
988
	fdt = files_fdtable(files);
989
	fd = find_next_zero_bit(fdt->open_fds->fds_bits, fdt->max_fds,
990
				files->next_fd);
L
Linus Torvalds 已提交
991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012

	/*
	 * N.B. For clone tasks sharing a files structure, this test
	 * will limit the total number of files that can be opened.
	 */
	if (fd >= current->signal->rlim[RLIMIT_NOFILE].rlim_cur)
		goto out;

	/* Do we need to expand the fd array or fd set?  */
	error = expand_files(files, fd);
	if (error < 0)
		goto out;

	if (error) {
		/*
	 	 * If we needed to expand the fs array we
		 * might have blocked - try again.
		 */
		error = -EMFILE;
		goto repeat;
	}

1013
	FD_SET(fd, fdt->open_fds);
U
Ulrich Drepper 已提交
1014 1015 1016 1017
	if (flags & O_CLOEXEC)
		FD_SET(fd, fdt->close_on_exec);
	else
		FD_CLR(fd, fdt->close_on_exec);
1018
	files->next_fd = fd + 1;
L
Linus Torvalds 已提交
1019 1020
#if 1
	/* Sanity check */
1021
	if (fdt->fd[fd] != NULL) {
L
Linus Torvalds 已提交
1022
		printk(KERN_WARNING "get_unused_fd: slot %d not NULL!\n", fd);
1023
		fdt->fd[fd] = NULL;
L
Linus Torvalds 已提交
1024 1025 1026 1027 1028 1029 1030 1031 1032
	}
#endif
	error = fd;

out:
	spin_unlock(&files->file_lock);
	return error;
}

U
Ulrich Drepper 已提交
1033 1034 1035 1036 1037
int get_unused_fd(void)
{
	return get_unused_fd_flags(0);
}

L
Linus Torvalds 已提交
1038 1039
EXPORT_SYMBOL(get_unused_fd);

1040
static void __put_unused_fd(struct files_struct *files, unsigned int fd)
L
Linus Torvalds 已提交
1041
{
1042 1043
	struct fdtable *fdt = files_fdtable(files);
	__FD_CLR(fd, fdt->open_fds);
1044 1045
	if (fd < files->next_fd)
		files->next_fd = fd;
L
Linus Torvalds 已提交
1046 1047
}

1048
void put_unused_fd(unsigned int fd)
L
Linus Torvalds 已提交
1049 1050 1051 1052 1053 1054 1055 1056 1057 1058
{
	struct files_struct *files = current->files;
	spin_lock(&files->file_lock);
	__put_unused_fd(files, fd);
	spin_unlock(&files->file_lock);
}

EXPORT_SYMBOL(put_unused_fd);

/*
1059
 * Install a file pointer in the fd array.
L
Linus Torvalds 已提交
1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070
 *
 * The VFS is full of places where we drop the files lock between
 * setting the open_fds bitmap and installing the file in the file
 * array.  At any such point, we are vulnerable to a dup2() race
 * installing a file in the array before us.  We need to detect this and
 * fput() the struct file we are about to overwrite in this case.
 *
 * It should never happen - if we allow dup2() do it, _really_ bad things
 * will follow.
 */

1071
void fd_install(unsigned int fd, struct file *file)
L
Linus Torvalds 已提交
1072 1073
{
	struct files_struct *files = current->files;
1074
	struct fdtable *fdt;
L
Linus Torvalds 已提交
1075
	spin_lock(&files->file_lock);
1076
	fdt = files_fdtable(files);
1077 1078
	BUG_ON(fdt->fd[fd] != NULL);
	rcu_assign_pointer(fdt->fd[fd], file);
L
Linus Torvalds 已提交
1079 1080 1081 1082 1083
	spin_unlock(&files->file_lock);
}

EXPORT_SYMBOL(fd_install);

1084
long do_sys_open(int dfd, const char __user *filename, int flags, int mode)
L
Linus Torvalds 已提交
1085
{
1086 1087
	char *tmp = getname(filename);
	int fd = PTR_ERR(tmp);
L
Linus Torvalds 已提交
1088 1089

	if (!IS_ERR(tmp)) {
U
Ulrich Drepper 已提交
1090
		fd = get_unused_fd_flags(flags);
L
Linus Torvalds 已提交
1091
		if (fd >= 0) {
1092
			struct file *f = do_filp_open(dfd, tmp, flags, mode);
T
Telemaque Ndizihiwe 已提交
1093 1094 1095 1096
			if (IS_ERR(f)) {
				put_unused_fd(fd);
				fd = PTR_ERR(f);
			} else {
1097
				fsnotify_open(f->f_path.dentry);
T
Telemaque Ndizihiwe 已提交
1098 1099
				fd_install(fd, f);
			}
L
Linus Torvalds 已提交
1100 1101 1102 1103 1104
		}
		putname(tmp);
	}
	return fd;
}
1105 1106 1107

asmlinkage long sys_open(const char __user *filename, int flags, int mode)
{
1108 1109
	long ret;

1110 1111 1112
	if (force_o_largefile())
		flags |= O_LARGEFILE;

1113 1114
	ret = do_sys_open(AT_FDCWD, filename, flags, mode);
	/* avoid REGPARM breakage on x86: */
1115
	asmlinkage_protect(3, ret, filename, flags, mode);
1116
	return ret;
1117
}
L
Linus Torvalds 已提交
1118

1119 1120 1121
asmlinkage long sys_openat(int dfd, const char __user *filename, int flags,
			   int mode)
{
1122 1123
	long ret;

1124 1125 1126
	if (force_o_largefile())
		flags |= O_LARGEFILE;

1127 1128
	ret = do_sys_open(dfd, filename, flags, mode);
	/* avoid REGPARM breakage on x86: */
1129
	asmlinkage_protect(4, ret, dfd, filename, flags, mode);
1130
	return ret;
1131 1132
}

L
Linus Torvalds 已提交
1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151
#ifndef __alpha__

/*
 * For backward compatibility?  Maybe this should be moved
 * into arch/i386 instead?
 */
asmlinkage long sys_creat(const char __user * pathname, int mode)
{
	return sys_open(pathname, O_CREAT | O_WRONLY | O_TRUNC, mode);
}

#endif

/*
 * "id" is the POSIX thread ID. We use the
 * files pointer for this..
 */
int filp_close(struct file *filp, fl_owner_t id)
{
1152
	int retval = 0;
L
Linus Torvalds 已提交
1153 1154 1155

	if (!file_count(filp)) {
		printk(KERN_ERR "VFS: Close: file count is 0\n");
1156
		return 0;
L
Linus Torvalds 已提交
1157 1158
	}

1159
	if (filp->f_op && filp->f_op->flush)
1160
		retval = filp->f_op->flush(filp, id);
L
Linus Torvalds 已提交
1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178

	dnotify_flush(filp, id);
	locks_remove_posix(filp, id);
	fput(filp);
	return retval;
}

EXPORT_SYMBOL(filp_close);

/*
 * Careful here! We test whether the file pointer is NULL before
 * releasing the fd. This ensures that one clone task can't release
 * an fd while another clone is opening it.
 */
asmlinkage long sys_close(unsigned int fd)
{
	struct file * filp;
	struct files_struct *files = current->files;
1179
	struct fdtable *fdt;
1180
	int retval;
L
Linus Torvalds 已提交
1181 1182

	spin_lock(&files->file_lock);
1183 1184
	fdt = files_fdtable(files);
	if (fd >= fdt->max_fds)
L
Linus Torvalds 已提交
1185
		goto out_unlock;
1186
	filp = fdt->fd[fd];
L
Linus Torvalds 已提交
1187 1188
	if (!filp)
		goto out_unlock;
1189
	rcu_assign_pointer(fdt->fd[fd], NULL);
1190
	FD_CLR(fd, fdt->close_on_exec);
L
Linus Torvalds 已提交
1191 1192
	__put_unused_fd(files, fd);
	spin_unlock(&files->file_lock);
1193 1194 1195 1196 1197 1198 1199 1200 1201 1202
	retval = filp_close(filp, files);

	/* can't restart close syscall because file table entry was cleared */
	if (unlikely(retval == -ERESTARTSYS ||
		     retval == -ERESTARTNOINTR ||
		     retval == -ERESTARTNOHAND ||
		     retval == -ERESTART_RESTARTBLOCK))
		retval = -EINTR;

	return retval;
L
Linus Torvalds 已提交
1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217

out_unlock:
	spin_unlock(&files->file_lock);
	return -EBADF;
}

EXPORT_SYMBOL(sys_close);

/*
 * This routine simulates a hangup on the tty, to arrange that users
 * are given clean terminals at login time.
 */
asmlinkage long sys_vhangup(void)
{
	if (capable(CAP_SYS_TTY_CONFIG)) {
1218
		/* XXX: this needs locking */
L
Linus Torvalds 已提交
1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233
		tty_vhangup(current->signal->tty);
		return 0;
	}
	return -EPERM;
}

/*
 * Called when an inode is about to be open.
 * We use this to disallow opening large files on 32bit systems if
 * the caller didn't specify O_LARGEFILE.  On 64bit systems we force
 * on this flag in sys_open.
 */
int generic_file_open(struct inode * inode, struct file * filp)
{
	if (!(filp->f_flags & O_LARGEFILE) && i_size_read(inode) > MAX_NON_LFS)
1234
		return -EOVERFLOW;
L
Linus Torvalds 已提交
1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250
	return 0;
}

EXPORT_SYMBOL(generic_file_open);

/*
 * This is used by subsystems that don't want seekable
 * file descriptors
 */
int nonseekable_open(struct inode *inode, struct file *filp)
{
	filp->f_mode &= ~(FMODE_LSEEK | FMODE_PREAD | FMODE_PWRITE);
	return 0;
}

EXPORT_SYMBOL(nonseekable_open);