1// SPDX-License-Identifier: GPL-2.0-only 2/* 3 * 4 * Copyright (C) 2011 Novell Inc. 5 */ 6 7#include <linux/fs.h> 8#include <linux/namei.h> 9#include <linux/xattr.h> 10#include <linux/security.h> 11#include <linux/cred.h> 12#include <linux/module.h> 13#include <linux/posix_acl.h> 14#include <linux/posix_acl_xattr.h> 15#include <linux/atomic.h> 16#include <linux/ratelimit.h> 17#include "overlayfs.h" 18 19static unsigned short ovl_redirect_max = 256; 20module_param_named(redirect_max, ovl_redirect_max, ushort, 0644); 21MODULE_PARM_DESC(redirect_max, 22 "Maximum length of absolute redirect xattr value"); 23 24static int ovl_set_redirect(struct dentry *dentry, bool samedir); 25 26int ovl_cleanup(struct ovl_fs *ofs, struct inode *wdir, struct dentry *wdentry) 27{ 28 int err; 29 30 dget(wdentry); 31 if (d_is_dir(wdentry)) 32 err = ovl_do_rmdir(ofs, wdir, wdentry); 33 else 34 err = ovl_do_unlink(ofs, wdir, wdentry); 35 dput(wdentry); 36 37 if (err) { 38 pr_err("cleanup of '%pd2' failed (%i)\n", 39 wdentry, err); 40 } 41 42 return err; 43} 44 45struct dentry *ovl_lookup_temp(struct ovl_fs *ofs, struct dentry *workdir) 46{ 47 struct dentry *temp; 48 char name[20]; 49 static atomic_t temp_id = ATOMIC_INIT(0); 50 51 /* counter is allowed to wrap, since temp dentries are ephemeral */ 52 snprintf(name, sizeof(name), "#%x", atomic_inc_return(&temp_id)); 53 54 temp = ovl_lookup_upper(ofs, name, workdir, strlen(name)); 55 if (!IS_ERR(temp) && temp->d_inode) { 56 pr_err("workdir/%s already exists\n", name); 57 dput(temp); 58 temp = ERR_PTR(-EIO); 59 } 60 61 return temp; 62} 63 64/* caller holds i_mutex on workdir */ 65static struct dentry *ovl_whiteout(struct ovl_fs *ofs) 66{ 67 int err; 68 struct dentry *whiteout; 69 struct dentry *workdir = ofs->workdir; 70 struct inode *wdir = workdir->d_inode; 71 72 if (!ofs->whiteout) { 73 whiteout = ovl_lookup_temp(ofs, workdir); 74 if (IS_ERR(whiteout)) 75 goto out; 76 77 err = ovl_do_whiteout(ofs, wdir, whiteout); 78 if (err) { 79 dput(whiteout); 80 whiteout = ERR_PTR(err); 81 goto out; 82 } 83 ofs->whiteout = whiteout; 84 } 85 86 if (!ofs->no_shared_whiteout) { 87 whiteout = ovl_lookup_temp(ofs, workdir); 88 if (IS_ERR(whiteout)) 89 goto out; 90 91 err = ovl_do_link(ofs, ofs->whiteout, wdir, whiteout); 92 if (!err) 93 goto out; 94 95 if (err != -EMLINK) { 96 pr_warn("Failed to link whiteout - disabling whiteout inode sharing(nlink=%u, err=%i)\n", 97 ofs->whiteout->d_inode->i_nlink, err); 98 ofs->no_shared_whiteout = true; 99 } 100 dput(whiteout); 101 } 102 whiteout = ofs->whiteout; 103 ofs->whiteout = NULL; 104out: 105 return whiteout; 106} 107 108/* Caller must hold i_mutex on both workdir and dir */ 109int ovl_cleanup_and_whiteout(struct ovl_fs *ofs, struct inode *dir, 110 struct dentry *dentry) 111{ 112 struct inode *wdir = ofs->workdir->d_inode; 113 struct dentry *whiteout; 114 int err; 115 int flags = 0; 116 117 whiteout = ovl_whiteout(ofs); 118 err = PTR_ERR(whiteout); 119 if (IS_ERR(whiteout)) 120 return err; 121 122 if (d_is_dir(dentry)) 123 flags = RENAME_EXCHANGE; 124 125 err = ovl_do_rename(ofs, wdir, whiteout, dir, dentry, flags); 126 if (err) 127 goto kill_whiteout; 128 if (flags) 129 ovl_cleanup(ofs, wdir, dentry); 130 131out: 132 dput(whiteout); 133 return err; 134 135kill_whiteout: 136 ovl_cleanup(ofs, wdir, whiteout); 137 goto out; 138} 139 140int ovl_mkdir_real(struct ovl_fs *ofs, struct inode *dir, 141 struct dentry **newdentry, umode_t mode) 142{ 143 int err; 144 struct dentry *d, *dentry = *newdentry; 145 146 err = ovl_do_mkdir(ofs, dir, dentry, mode); 147 if (err) 148 return err; 149 150 if (likely(!d_unhashed(dentry))) 151 return 0; 152 153 /* 154 * vfs_mkdir() may succeed and leave the dentry passed 155 * to it unhashed and negative. If that happens, try to 156 * lookup a new hashed and positive dentry. 157 */ 158 d = ovl_lookup_upper(ofs, dentry->d_name.name, dentry->d_parent, 159 dentry->d_name.len); 160 if (IS_ERR(d)) { 161 pr_warn("failed lookup after mkdir (%pd2, err=%i).\n", 162 dentry, err); 163 return PTR_ERR(d); 164 } 165 dput(dentry); 166 *newdentry = d; 167 168 return 0; 169} 170 171struct dentry *ovl_create_real(struct ovl_fs *ofs, struct inode *dir, 172 struct dentry *newdentry, struct ovl_cattr *attr) 173{ 174 int err; 175 176 if (IS_ERR(newdentry)) 177 return newdentry; 178 179 err = -ESTALE; 180 if (newdentry->d_inode) 181 goto out; 182 183 if (attr->hardlink) { 184 err = ovl_do_link(ofs, attr->hardlink, dir, newdentry); 185 } else { 186 switch (attr->mode & S_IFMT) { 187 case S_IFREG: 188 err = ovl_do_create(ofs, dir, newdentry, attr->mode); 189 break; 190 191 case S_IFDIR: 192 /* mkdir is special... */ 193 err = ovl_mkdir_real(ofs, dir, &newdentry, attr->mode); 194 break; 195 196 case S_IFCHR: 197 case S_IFBLK: 198 case S_IFIFO: 199 case S_IFSOCK: 200 err = ovl_do_mknod(ofs, dir, newdentry, attr->mode, 201 attr->rdev); 202 break; 203 204 case S_IFLNK: 205 err = ovl_do_symlink(ofs, dir, newdentry, attr->link); 206 break; 207 208 default: 209 err = -EPERM; 210 } 211 } 212 if (!err && WARN_ON(!newdentry->d_inode)) { 213 /* 214 * Not quite sure if non-instantiated dentry is legal or not. 215 * VFS doesn't seem to care so check and warn here. 216 */ 217 err = -EIO; 218 } 219out: 220 if (err) { 221 dput(newdentry); 222 return ERR_PTR(err); 223 } 224 return newdentry; 225} 226 227struct dentry *ovl_create_temp(struct ovl_fs *ofs, struct dentry *workdir, 228 struct ovl_cattr *attr) 229{ 230 return ovl_create_real(ofs, d_inode(workdir), 231 ovl_lookup_temp(ofs, workdir), attr); 232} 233 234static int ovl_set_opaque_xerr(struct dentry *dentry, struct dentry *upper, 235 int xerr) 236{ 237 struct ovl_fs *ofs = OVL_FS(dentry->d_sb); 238 int err; 239 240 err = ovl_check_setxattr(ofs, upper, OVL_XATTR_OPAQUE, "y", 1, xerr); 241 if (!err) 242 ovl_dentry_set_opaque(dentry); 243 244 return err; 245} 246 247static int ovl_set_opaque(struct dentry *dentry, struct dentry *upperdentry) 248{ 249 /* 250 * Fail with -EIO when trying to create opaque dir and upper doesn't 251 * support xattrs. ovl_rename() calls ovl_set_opaque_xerr(-EXDEV) to 252 * return a specific error for noxattr case. 253 */ 254 return ovl_set_opaque_xerr(dentry, upperdentry, -EIO); 255} 256 257/* 258 * Common operations required to be done after creation of file on upper. 259 * If @hardlink is false, then @inode is a pre-allocated inode, we may or 260 * may not use to instantiate the new dentry. 261 */ 262static int ovl_instantiate(struct dentry *dentry, struct inode *inode, 263 struct dentry *newdentry, bool hardlink) 264{ 265 struct ovl_inode_params oip = { 266 .upperdentry = newdentry, 267 .newinode = inode, 268 }; 269 270 ovl_dir_modified(dentry->d_parent, false); 271 ovl_dentry_set_upper_alias(dentry); 272 ovl_dentry_init_reval(dentry, newdentry, NULL); 273 274 if (!hardlink) { 275 /* 276 * ovl_obtain_alias() can be called after ovl_create_real() 277 * and before we get here, so we may get an inode from cache 278 * with the same real upperdentry that is not the inode we 279 * pre-allocated. In this case we will use the cached inode 280 * to instantiate the new dentry. 281 * 282 * XXX: if we ever use ovl_obtain_alias() to decode directory 283 * file handles, need to use ovl_get_inode_locked() and 284 * d_instantiate_new() here to prevent from creating two 285 * hashed directory inode aliases. 286 */ 287 inode = ovl_get_inode(dentry->d_sb, &oip); 288 if (IS_ERR(inode)) 289 return PTR_ERR(inode); 290 if (inode == oip.newinode) 291 ovl_set_flag(OVL_UPPERDATA, inode); 292 } else { 293 WARN_ON(ovl_inode_real(inode) != d_inode(newdentry)); 294 dput(newdentry); 295 inc_nlink(inode); 296 } 297 298 d_instantiate(dentry, inode); 299 if (inode != oip.newinode) { 300 pr_warn_ratelimited("newly created inode found in cache (%pd2)\n", 301 dentry); 302 } 303 304 /* Force lookup of new upper hardlink to find its lower */ 305 if (hardlink) 306 d_drop(dentry); 307 308 return 0; 309} 310 311static bool ovl_type_merge(struct dentry *dentry) 312{ 313 return OVL_TYPE_MERGE(ovl_path_type(dentry)); 314} 315 316static bool ovl_type_origin(struct dentry *dentry) 317{ 318 return OVL_TYPE_ORIGIN(ovl_path_type(dentry)); 319} 320 321static int ovl_create_upper(struct dentry *dentry, struct inode *inode, 322 struct ovl_cattr *attr) 323{ 324 struct ovl_fs *ofs = OVL_FS(dentry->d_sb); 325 struct dentry *upperdir = ovl_dentry_upper(dentry->d_parent); 326 struct inode *udir = upperdir->d_inode; 327 struct dentry *newdentry; 328 int err; 329 330 if (!attr->hardlink && !IS_POSIXACL(udir)) 331 attr->mode &= ~current_umask(); 332 333 inode_lock_nested(udir, I_MUTEX_PARENT); 334 newdentry = ovl_create_real(ofs, udir, 335 ovl_lookup_upper(ofs, dentry->d_name.name, 336 upperdir, dentry->d_name.len), 337 attr); 338 err = PTR_ERR(newdentry); 339 if (IS_ERR(newdentry)) 340 goto out_unlock; 341 342 if (ovl_type_merge(dentry->d_parent) && d_is_dir(newdentry) && 343 !ovl_allow_offline_changes(ofs)) { 344 /* Setting opaque here is just an optimization, allow to fail */ 345 ovl_set_opaque(dentry, newdentry); 346 } 347 348 err = ovl_instantiate(dentry, inode, newdentry, !!attr->hardlink); 349 if (err) 350 goto out_cleanup; 351out_unlock: 352 inode_unlock(udir); 353 return err; 354 355out_cleanup: 356 ovl_cleanup(ofs, udir, newdentry); 357 dput(newdentry); 358 goto out_unlock; 359} 360 361static struct dentry *ovl_clear_empty(struct dentry *dentry, 362 struct list_head *list) 363{ 364 struct ovl_fs *ofs = OVL_FS(dentry->d_sb); 365 struct dentry *workdir = ovl_workdir(dentry); 366 struct inode *wdir = workdir->d_inode; 367 struct dentry *upperdir = ovl_dentry_upper(dentry->d_parent); 368 struct inode *udir = upperdir->d_inode; 369 struct path upperpath; 370 struct dentry *upper; 371 struct dentry *opaquedir; 372 struct kstat stat; 373 int err; 374 375 if (WARN_ON(!workdir)) 376 return ERR_PTR(-EROFS); 377 378 err = ovl_lock_rename_workdir(workdir, upperdir); 379 if (err) 380 goto out; 381 382 ovl_path_upper(dentry, &upperpath); 383 err = vfs_getattr(&upperpath, &stat, 384 STATX_BASIC_STATS, AT_STATX_SYNC_AS_STAT); 385 if (err) 386 goto out_unlock; 387 388 err = -ESTALE; 389 if (!S_ISDIR(stat.mode)) 390 goto out_unlock; 391 upper = upperpath.dentry; 392 if (upper->d_parent->d_inode != udir) 393 goto out_unlock; 394 395 opaquedir = ovl_create_temp(ofs, workdir, OVL_CATTR(stat.mode)); 396 err = PTR_ERR(opaquedir); 397 if (IS_ERR(opaquedir)) 398 goto out_unlock; 399 400 err = ovl_copy_xattr(dentry->d_sb, &upperpath, opaquedir); 401 if (err) 402 goto out_cleanup; 403 404 err = ovl_set_opaque(dentry, opaquedir); 405 if (err) 406 goto out_cleanup; 407 408 inode_lock(opaquedir->d_inode); 409 err = ovl_set_attr(ofs, opaquedir, &stat); 410 inode_unlock(opaquedir->d_inode); 411 if (err) 412 goto out_cleanup; 413 414 err = ovl_do_rename(ofs, wdir, opaquedir, udir, upper, RENAME_EXCHANGE); 415 if (err) 416 goto out_cleanup; 417 418 ovl_cleanup_whiteouts(ofs, upper, list); 419 ovl_cleanup(ofs, wdir, upper); 420 unlock_rename(workdir, upperdir); 421 422 /* dentry's upper doesn't match now, get rid of it */ 423 d_drop(dentry); 424 425 return opaquedir; 426 427out_cleanup: 428 ovl_cleanup(ofs, wdir, opaquedir); 429 dput(opaquedir); 430out_unlock: 431 unlock_rename(workdir, upperdir); 432out: 433 return ERR_PTR(err); 434} 435 436static int ovl_set_upper_acl(struct ovl_fs *ofs, struct dentry *upperdentry, 437 const char *acl_name, struct posix_acl *acl) 438{ 439 if (!IS_ENABLED(CONFIG_FS_POSIX_ACL) || !acl) 440 return 0; 441 442 return ovl_do_set_acl(ofs, upperdentry, acl_name, acl); 443} 444 445static int ovl_create_over_whiteout(struct dentry *dentry, struct inode *inode, 446 struct ovl_cattr *cattr) 447{ 448 struct ovl_fs *ofs = OVL_FS(dentry->d_sb); 449 struct dentry *workdir = ovl_workdir(dentry); 450 struct inode *wdir = workdir->d_inode; 451 struct dentry *upperdir = ovl_dentry_upper(dentry->d_parent); 452 struct inode *udir = upperdir->d_inode; 453 struct dentry *upper; 454 struct dentry *newdentry; 455 int err; 456 struct posix_acl *acl, *default_acl; 457 bool hardlink = !!cattr->hardlink; 458 459 if (WARN_ON(!workdir)) 460 return -EROFS; 461 462 if (!hardlink) { 463 err = posix_acl_create(dentry->d_parent->d_inode, 464 &cattr->mode, &default_acl, &acl); 465 if (err) 466 return err; 467 } 468 469 err = ovl_lock_rename_workdir(workdir, upperdir); 470 if (err) 471 goto out; 472 473 upper = ovl_lookup_upper(ofs, dentry->d_name.name, upperdir, 474 dentry->d_name.len); 475 err = PTR_ERR(upper); 476 if (IS_ERR(upper)) 477 goto out_unlock; 478 479 err = -ESTALE; 480 if (d_is_negative(upper) || !IS_WHITEOUT(d_inode(upper))) 481 goto out_dput; 482 483 newdentry = ovl_create_temp(ofs, workdir, cattr); 484 err = PTR_ERR(newdentry); 485 if (IS_ERR(newdentry)) 486 goto out_dput; 487 488 /* 489 * mode could have been mutilated due to umask (e.g. sgid directory) 490 */ 491 if (!hardlink && 492 !S_ISLNK(cattr->mode) && 493 newdentry->d_inode->i_mode != cattr->mode) { 494 struct iattr attr = { 495 .ia_valid = ATTR_MODE, 496 .ia_mode = cattr->mode, 497 }; 498 inode_lock(newdentry->d_inode); 499 err = ovl_do_notify_change(ofs, newdentry, &attr); 500 inode_unlock(newdentry->d_inode); 501 if (err) 502 goto out_cleanup; 503 } 504 if (!hardlink) { 505 err = ovl_set_upper_acl(ofs, newdentry, 506 XATTR_NAME_POSIX_ACL_ACCESS, acl); 507 if (err) 508 goto out_cleanup; 509 510 err = ovl_set_upper_acl(ofs, newdentry, 511 XATTR_NAME_POSIX_ACL_DEFAULT, default_acl); 512 if (err) 513 goto out_cleanup; 514 } 515 516 if (!hardlink && S_ISDIR(cattr->mode)) { 517 err = ovl_set_opaque(dentry, newdentry); 518 if (err) 519 goto out_cleanup; 520 521 err = ovl_do_rename(ofs, wdir, newdentry, udir, upper, 522 RENAME_EXCHANGE); 523 if (err) 524 goto out_cleanup; 525 526 ovl_cleanup(ofs, wdir, upper); 527 } else { 528 err = ovl_do_rename(ofs, wdir, newdentry, udir, upper, 0); 529 if (err) 530 goto out_cleanup; 531 } 532 err = ovl_instantiate(dentry, inode, newdentry, hardlink); 533 if (err) { 534 ovl_cleanup(ofs, udir, newdentry); 535 dput(newdentry); 536 } 537out_dput: 538 dput(upper); 539out_unlock: 540 unlock_rename(workdir, upperdir); 541out: 542 if (!hardlink) { 543 posix_acl_release(acl); 544 posix_acl_release(default_acl); 545 } 546 return err; 547 548out_cleanup: 549 ovl_cleanup(ofs, wdir, newdentry); 550 dput(newdentry); 551 goto out_dput; 552} 553 554static int ovl_create_or_link(struct dentry *dentry, struct inode *inode, 555 struct ovl_cattr *attr, bool origin) 556{ 557 int err; 558 const struct cred *old_cred; 559 struct cred *override_cred; 560 struct dentry *parent = dentry->d_parent; 561 562 err = ovl_copy_up(parent); 563 if (err) 564 return err; 565 566 old_cred = ovl_override_creds(dentry->d_sb); 567 568 /* 569 * When linking a file with copy up origin into a new parent, mark the 570 * new parent dir "impure". 571 */ 572 if (origin) { 573 err = ovl_set_impure(parent, ovl_dentry_upper(parent)); 574 if (err) 575 goto out_revert_creds; 576 } 577 578 if (!attr->hardlink) { 579 err = -ENOMEM; 580 override_cred = prepare_creds(); 581 if (!override_cred) 582 goto out_revert_creds; 583 /* 584 * In the creation cases(create, mkdir, mknod, symlink), 585 * ovl should transfer current's fs{u,g}id to underlying 586 * fs. Because underlying fs want to initialize its new 587 * inode owner using current's fs{u,g}id. And in this 588 * case, the @inode is a new inode that is initialized 589 * in inode_init_owner() to current's fs{u,g}id. So use 590 * the inode's i_{u,g}id to override the cred's fs{u,g}id. 591 * 592 * But in the other hardlink case, ovl_link() does not 593 * create a new inode, so just use the ovl mounter's 594 * fs{u,g}id. 595 */ 596 override_cred->fsuid = inode->i_uid; 597 override_cred->fsgid = inode->i_gid; 598 err = security_dentry_create_files_as(dentry, 599 attr->mode, &dentry->d_name, old_cred, 600 override_cred); 601 if (err) { 602 put_cred(override_cred); 603 goto out_revert_creds; 604 } 605 put_cred(override_creds(override_cred)); 606 put_cred(override_cred); 607 } 608 609 if (!ovl_dentry_is_whiteout(dentry)) 610 err = ovl_create_upper(dentry, inode, attr); 611 else 612 err = ovl_create_over_whiteout(dentry, inode, attr); 613 614out_revert_creds: 615 revert_creds(old_cred); 616 return err; 617} 618 619static int ovl_create_object(struct dentry *dentry, int mode, dev_t rdev, 620 const char *link) 621{ 622 int err; 623 struct inode *inode; 624 struct ovl_cattr attr = { 625 .rdev = rdev, 626 .link = link, 627 }; 628 629 err = ovl_want_write(dentry); 630 if (err) 631 goto out; 632 633 /* Preallocate inode to be used by ovl_get_inode() */ 634 err = -ENOMEM; 635 inode = ovl_new_inode(dentry->d_sb, mode, rdev); 636 if (!inode) 637 goto out_drop_write; 638 639 spin_lock(&inode->i_lock); 640 inode->i_state |= I_CREATING; 641 spin_unlock(&inode->i_lock); 642 643 inode_init_owner(&nop_mnt_idmap, inode, dentry->d_parent->d_inode, mode); 644 attr.mode = inode->i_mode; 645 646 err = ovl_create_or_link(dentry, inode, &attr, false); 647 /* Did we end up using the preallocated inode? */ 648 if (inode != d_inode(dentry)) 649 iput(inode); 650 651out_drop_write: 652 ovl_drop_write(dentry); 653out: 654 return err; 655} 656 657static int ovl_create(struct mnt_idmap *idmap, struct inode *dir, 658 struct dentry *dentry, umode_t mode, bool excl) 659{ 660 return ovl_create_object(dentry, (mode & 07777) | S_IFREG, 0, NULL); 661} 662 663static int ovl_mkdir(struct mnt_idmap *idmap, struct inode *dir, 664 struct dentry *dentry, umode_t mode) 665{ 666 return ovl_create_object(dentry, (mode & 07777) | S_IFDIR, 0, NULL); 667} 668 669static int ovl_mknod(struct mnt_idmap *idmap, struct inode *dir, 670 struct dentry *dentry, umode_t mode, dev_t rdev) 671{ 672 /* Don't allow creation of "whiteout" on overlay */ 673 if (S_ISCHR(mode) && rdev == WHITEOUT_DEV) 674 return -EPERM; 675 676 return ovl_create_object(dentry, mode, rdev, NULL); 677} 678 679static int ovl_symlink(struct mnt_idmap *idmap, struct inode *dir, 680 struct dentry *dentry, const char *link) 681{ 682 return ovl_create_object(dentry, S_IFLNK, 0, link); 683} 684 685static int ovl_set_link_redirect(struct dentry *dentry) 686{ 687 const struct cred *old_cred; 688 int err; 689 690 old_cred = ovl_override_creds(dentry->d_sb); 691 err = ovl_set_redirect(dentry, false); 692 revert_creds(old_cred); 693 694 return err; 695} 696 697static int ovl_link(struct dentry *old, struct inode *newdir, 698 struct dentry *new) 699{ 700 int err; 701 struct inode *inode; 702 703 err = ovl_want_write(old); 704 if (err) 705 goto out; 706 707 err = ovl_copy_up(old); 708 if (err) 709 goto out_drop_write; 710 711 err = ovl_copy_up(new->d_parent); 712 if (err) 713 goto out_drop_write; 714 715 if (ovl_is_metacopy_dentry(old)) { 716 err = ovl_set_link_redirect(old); 717 if (err) 718 goto out_drop_write; 719 } 720 721 err = ovl_nlink_start(old); 722 if (err) 723 goto out_drop_write; 724 725 inode = d_inode(old); 726 ihold(inode); 727 728 err = ovl_create_or_link(new, inode, 729 &(struct ovl_cattr) {.hardlink = ovl_dentry_upper(old)}, 730 ovl_type_origin(old)); 731 if (err) 732 iput(inode); 733 734 ovl_nlink_end(old); 735out_drop_write: 736 ovl_drop_write(old); 737out: 738 return err; 739} 740 741static bool ovl_matches_upper(struct dentry *dentry, struct dentry *upper) 742{ 743 return d_inode(ovl_dentry_upper(dentry)) == d_inode(upper); 744} 745 746static int ovl_remove_and_whiteout(struct dentry *dentry, 747 struct list_head *list) 748{ 749 struct ovl_fs *ofs = OVL_FS(dentry->d_sb); 750 struct dentry *workdir = ovl_workdir(dentry); 751 struct dentry *upperdir = ovl_dentry_upper(dentry->d_parent); 752 struct dentry *upper; 753 struct dentry *opaquedir = NULL; 754 int err; 755 756 if (WARN_ON(!workdir)) 757 return -EROFS; 758 759 if (!list_empty(list)) { 760 opaquedir = ovl_clear_empty(dentry, list); 761 err = PTR_ERR(opaquedir); 762 if (IS_ERR(opaquedir)) 763 goto out; 764 } 765 766 err = ovl_lock_rename_workdir(workdir, upperdir); 767 if (err) 768 goto out_dput; 769 770 upper = ovl_lookup_upper(ofs, dentry->d_name.name, upperdir, 771 dentry->d_name.len); 772 err = PTR_ERR(upper); 773 if (IS_ERR(upper)) 774 goto out_unlock; 775 776 err = -ESTALE; 777 if ((opaquedir && upper != opaquedir) || 778 (!opaquedir && ovl_dentry_upper(dentry) && 779 !ovl_matches_upper(dentry, upper))) { 780 goto out_dput_upper; 781 } 782 783 err = ovl_cleanup_and_whiteout(ofs, d_inode(upperdir), upper); 784 if (err) 785 goto out_d_drop; 786 787 ovl_dir_modified(dentry->d_parent, true); 788out_d_drop: 789 d_drop(dentry); 790out_dput_upper: 791 dput(upper); 792out_unlock: 793 unlock_rename(workdir, upperdir); 794out_dput: 795 dput(opaquedir); 796out: 797 return err; 798} 799 800static int ovl_remove_upper(struct dentry *dentry, bool is_dir, 801 struct list_head *list) 802{ 803 struct ovl_fs *ofs = OVL_FS(dentry->d_sb); 804 struct dentry *upperdir = ovl_dentry_upper(dentry->d_parent); 805 struct inode *dir = upperdir->d_inode; 806 struct dentry *upper; 807 struct dentry *opaquedir = NULL; 808 int err; 809 810 if (!list_empty(list)) { 811 opaquedir = ovl_clear_empty(dentry, list); 812 err = PTR_ERR(opaquedir); 813 if (IS_ERR(opaquedir)) 814 goto out; 815 } 816 817 inode_lock_nested(dir, I_MUTEX_PARENT); 818 upper = ovl_lookup_upper(ofs, dentry->d_name.name, upperdir, 819 dentry->d_name.len); 820 err = PTR_ERR(upper); 821 if (IS_ERR(upper)) 822 goto out_unlock; 823 824 err = -ESTALE; 825 if ((opaquedir && upper != opaquedir) || 826 (!opaquedir && !ovl_matches_upper(dentry, upper))) 827 goto out_dput_upper; 828 829 if (is_dir) 830 err = ovl_do_rmdir(ofs, dir, upper); 831 else 832 err = ovl_do_unlink(ofs, dir, upper); 833 ovl_dir_modified(dentry->d_parent, ovl_type_origin(dentry)); 834 835 /* 836 * Keeping this dentry hashed would mean having to release 837 * upperpath/lowerpath, which could only be done if we are the 838 * sole user of this dentry. Too tricky... Just unhash for 839 * now. 840 */ 841 if (!err) 842 d_drop(dentry); 843out_dput_upper: 844 dput(upper); 845out_unlock: 846 inode_unlock(dir); 847 dput(opaquedir); 848out: 849 return err; 850} 851 852static bool ovl_pure_upper(struct dentry *dentry) 853{ 854 return !ovl_dentry_lower(dentry) && 855 !ovl_test_flag(OVL_WHITEOUTS, d_inode(dentry)); 856} 857 858static void ovl_drop_nlink(struct dentry *dentry) 859{ 860 struct inode *inode = d_inode(dentry); 861 struct dentry *alias; 862 863 /* Try to find another, hashed alias */ 864 spin_lock(&inode->i_lock); 865 hlist_for_each_entry(alias, &inode->i_dentry, d_u.d_alias) { 866 if (alias != dentry && !d_unhashed(alias)) 867 break; 868 } 869 spin_unlock(&inode->i_lock); 870 871 /* 872 * Changes to underlying layers may cause i_nlink to lose sync with 873 * reality. In this case prevent the link count from going to zero 874 * prematurely. 875 */ 876 if (inode->i_nlink > !!alias) 877 drop_nlink(inode); 878} 879 880static int ovl_do_remove(struct dentry *dentry, bool is_dir) 881{ 882 int err; 883 const struct cred *old_cred; 884 bool lower_positive = ovl_lower_positive(dentry); 885 LIST_HEAD(list); 886 887 /* No need to clean pure upper removed by vfs_rmdir() */ 888 if (is_dir && (lower_positive || !ovl_pure_upper(dentry))) { 889 err = ovl_check_empty_dir(dentry, &list); 890 if (err) 891 goto out; 892 } 893 894 err = ovl_want_write(dentry); 895 if (err) 896 goto out; 897 898 err = ovl_copy_up(dentry->d_parent); 899 if (err) 900 goto out_drop_write; 901 902 err = ovl_nlink_start(dentry); 903 if (err) 904 goto out_drop_write; 905 906 old_cred = ovl_override_creds(dentry->d_sb); 907 if (!lower_positive) 908 err = ovl_remove_upper(dentry, is_dir, &list); 909 else 910 err = ovl_remove_and_whiteout(dentry, &list); 911 revert_creds(old_cred); 912 if (!err) { 913 if (is_dir) 914 clear_nlink(dentry->d_inode); 915 else 916 ovl_drop_nlink(dentry); 917 } 918 ovl_nlink_end(dentry); 919 920 /* 921 * Copy ctime 922 * 923 * Note: we fail to update ctime if there was no copy-up, only a 924 * whiteout 925 */ 926 if (ovl_dentry_upper(dentry)) 927 ovl_copyattr(d_inode(dentry)); 928 929out_drop_write: 930 ovl_drop_write(dentry); 931out: 932 ovl_cache_free(&list); 933 return err; 934} 935 936static int ovl_unlink(struct inode *dir, struct dentry *dentry) 937{ 938 return ovl_do_remove(dentry, false); 939} 940 941static int ovl_rmdir(struct inode *dir, struct dentry *dentry) 942{ 943 return ovl_do_remove(dentry, true); 944} 945 946static bool ovl_type_merge_or_lower(struct dentry *dentry) 947{ 948 enum ovl_path_type type = ovl_path_type(dentry); 949 950 return OVL_TYPE_MERGE(type) || !OVL_TYPE_UPPER(type); 951} 952 953static bool ovl_can_move(struct dentry *dentry) 954{ 955 return ovl_redirect_dir(OVL_FS(dentry->d_sb)) || 956 !d_is_dir(dentry) || !ovl_type_merge_or_lower(dentry); 957} 958 959static char *ovl_get_redirect(struct dentry *dentry, bool abs_redirect) 960{ 961 char *buf, *ret; 962 struct dentry *d, *tmp; 963 int buflen = ovl_redirect_max + 1; 964 965 if (!abs_redirect) { 966 ret = kstrndup(dentry->d_name.name, dentry->d_name.len, 967 GFP_KERNEL); 968 goto out; 969 } 970 971 buf = ret = kmalloc(buflen, GFP_KERNEL); 972 if (!buf) 973 goto out; 974 975 buflen--; 976 buf[buflen] = '\0'; 977 for (d = dget(dentry); !IS_ROOT(d);) { 978 const char *name; 979 int thislen; 980 981 spin_lock(&d->d_lock); 982 name = ovl_dentry_get_redirect(d); 983 if (name) { 984 thislen = strlen(name); 985 } else { 986 name = d->d_name.name; 987 thislen = d->d_name.len; 988 } 989 990 /* If path is too long, fall back to userspace move */ 991 if (thislen + (name[0] != '/') > buflen) { 992 ret = ERR_PTR(-EXDEV); 993 spin_unlock(&d->d_lock); 994 goto out_put; 995 } 996 997 buflen -= thislen; 998 memcpy(&buf[buflen], name, thislen); 999 spin_unlock(&d->d_lock); 1000 tmp = dget_parent(d); 1001 1002 dput(d); 1003 d = tmp; 1004 1005 /* Absolute redirect: finished */ 1006 if (buf[buflen] == '/') 1007 break; 1008 buflen--; 1009 buf[buflen] = '/'; 1010 } 1011 ret = kstrdup(&buf[buflen], GFP_KERNEL); 1012out_put: 1013 dput(d); 1014 kfree(buf); 1015out: 1016 return ret ? ret : ERR_PTR(-ENOMEM); 1017} 1018 1019static bool ovl_need_absolute_redirect(struct dentry *dentry, bool samedir) 1020{ 1021 struct dentry *lowerdentry; 1022 1023 if (!samedir) 1024 return true; 1025 1026 if (d_is_dir(dentry)) 1027 return false; 1028 1029 /* 1030 * For non-dir hardlinked files, we need absolute redirects 1031 * in general as two upper hardlinks could be in different 1032 * dirs. We could put a relative redirect now and convert 1033 * it to absolute redirect later. But when nlink > 1 and 1034 * indexing is on, that means relative redirect needs to be 1035 * converted to absolute during copy up of another lower 1036 * hardllink as well. 1037 * 1038 * So without optimizing too much, just check if lower is 1039 * a hard link or not. If lower is hard link, put absolute 1040 * redirect. 1041 */ 1042 lowerdentry = ovl_dentry_lower(dentry); 1043 return (d_inode(lowerdentry)->i_nlink > 1); 1044} 1045 1046static int ovl_set_redirect(struct dentry *dentry, bool samedir) 1047{ 1048 int err; 1049 struct ovl_fs *ofs = OVL_FS(dentry->d_sb); 1050 const char *redirect = ovl_dentry_get_redirect(dentry); 1051 bool absolute_redirect = ovl_need_absolute_redirect(dentry, samedir); 1052 1053 if (redirect && (!absolute_redirect || redirect[0] == '/')) 1054 return 0; 1055 1056 redirect = ovl_get_redirect(dentry, absolute_redirect); 1057 if (IS_ERR(redirect)) 1058 return PTR_ERR(redirect); 1059 1060 err = ovl_check_setxattr(ofs, ovl_dentry_upper(dentry), 1061 OVL_XATTR_REDIRECT, 1062 redirect, strlen(redirect), -EXDEV); 1063 if (!err) { 1064 spin_lock(&dentry->d_lock); 1065 ovl_dentry_set_redirect(dentry, redirect); 1066 spin_unlock(&dentry->d_lock); 1067 } else { 1068 kfree(redirect); 1069 pr_warn_ratelimited("failed to set redirect (%i)\n", 1070 err); 1071 /* Fall back to userspace copy-up */ 1072 err = -EXDEV; 1073 } 1074 return err; 1075} 1076 1077static int ovl_rename(struct mnt_idmap *idmap, struct inode *olddir, 1078 struct dentry *old, struct inode *newdir, 1079 struct dentry *new, unsigned int flags) 1080{ 1081 int err; 1082 struct dentry *old_upperdir; 1083 struct dentry *new_upperdir; 1084 struct dentry *olddentry; 1085 struct dentry *newdentry; 1086 struct dentry *trap; 1087 bool old_opaque; 1088 bool new_opaque; 1089 bool cleanup_whiteout = false; 1090 bool update_nlink = false; 1091 bool overwrite = !(flags & RENAME_EXCHANGE); 1092 bool is_dir = d_is_dir(old); 1093 bool new_is_dir = d_is_dir(new); 1094 bool samedir = olddir == newdir; 1095 struct dentry *opaquedir = NULL; 1096 const struct cred *old_cred = NULL; 1097 struct ovl_fs *ofs = OVL_FS(old->d_sb); 1098 LIST_HEAD(list); 1099 1100 err = -EINVAL; 1101 if (flags & ~(RENAME_EXCHANGE | RENAME_NOREPLACE)) 1102 goto out; 1103 1104 flags &= ~RENAME_NOREPLACE; 1105 1106 /* Don't copy up directory trees */ 1107 err = -EXDEV; 1108 if (!ovl_can_move(old)) 1109 goto out; 1110 if (!overwrite && !ovl_can_move(new)) 1111 goto out; 1112 1113 if (overwrite && new_is_dir && !ovl_pure_upper(new)) { 1114 err = ovl_check_empty_dir(new, &list); 1115 if (err) 1116 goto out; 1117 } 1118 1119 if (overwrite) { 1120 if (ovl_lower_positive(old)) { 1121 if (!ovl_dentry_is_whiteout(new)) { 1122 /* Whiteout source */ 1123 flags |= RENAME_WHITEOUT; 1124 } else { 1125 /* Switch whiteouts */ 1126 flags |= RENAME_EXCHANGE; 1127 } 1128 } else if (is_dir && ovl_dentry_is_whiteout(new)) { 1129 flags |= RENAME_EXCHANGE; 1130 cleanup_whiteout = true; 1131 } 1132 } 1133 1134 err = ovl_want_write(old); 1135 if (err) 1136 goto out; 1137 1138 err = ovl_copy_up(old); 1139 if (err) 1140 goto out_drop_write; 1141 1142 err = ovl_copy_up(new->d_parent); 1143 if (err) 1144 goto out_drop_write; 1145 if (!overwrite) { 1146 err = ovl_copy_up(new); 1147 if (err) 1148 goto out_drop_write; 1149 } else if (d_inode(new)) { 1150 err = ovl_nlink_start(new); 1151 if (err) 1152 goto out_drop_write; 1153 1154 update_nlink = true; 1155 } 1156 1157 old_cred = ovl_override_creds(old->d_sb); 1158 1159 if (!list_empty(&list)) { 1160 opaquedir = ovl_clear_empty(new, &list); 1161 err = PTR_ERR(opaquedir); 1162 if (IS_ERR(opaquedir)) { 1163 opaquedir = NULL; 1164 goto out_revert_creds; 1165 } 1166 } 1167 1168 old_upperdir = ovl_dentry_upper(old->d_parent); 1169 new_upperdir = ovl_dentry_upper(new->d_parent); 1170 1171 if (!samedir) { 1172 /* 1173 * When moving a merge dir or non-dir with copy up origin into 1174 * a new parent, we are marking the new parent dir "impure". 1175 * When ovl_iterate() iterates an "impure" upper dir, it will 1176 * lookup the origin inodes of the entries to fill d_ino. 1177 */ 1178 if (ovl_type_origin(old)) { 1179 err = ovl_set_impure(new->d_parent, new_upperdir); 1180 if (err) 1181 goto out_revert_creds; 1182 } 1183 if (!overwrite && ovl_type_origin(new)) { 1184 err = ovl_set_impure(old->d_parent, old_upperdir); 1185 if (err) 1186 goto out_revert_creds; 1187 } 1188 } 1189 1190 trap = lock_rename(new_upperdir, old_upperdir); 1191 1192 olddentry = ovl_lookup_upper(ofs, old->d_name.name, old_upperdir, 1193 old->d_name.len); 1194 err = PTR_ERR(olddentry); 1195 if (IS_ERR(olddentry)) 1196 goto out_unlock; 1197 1198 err = -ESTALE; 1199 if (!ovl_matches_upper(old, olddentry)) 1200 goto out_dput_old; 1201 1202 newdentry = ovl_lookup_upper(ofs, new->d_name.name, new_upperdir, 1203 new->d_name.len); 1204 err = PTR_ERR(newdentry); 1205 if (IS_ERR(newdentry)) 1206 goto out_dput_old; 1207 1208 old_opaque = ovl_dentry_is_opaque(old); 1209 new_opaque = ovl_dentry_is_opaque(new); 1210 1211 err = -ESTALE; 1212 if (d_inode(new) && ovl_dentry_upper(new)) { 1213 if (opaquedir) { 1214 if (newdentry != opaquedir) 1215 goto out_dput; 1216 } else { 1217 if (!ovl_matches_upper(new, newdentry)) 1218 goto out_dput; 1219 } 1220 } else { 1221 if (!d_is_negative(newdentry)) { 1222 if (!new_opaque || !ovl_is_whiteout(newdentry)) 1223 goto out_dput; 1224 } else { 1225 if (flags & RENAME_EXCHANGE) 1226 goto out_dput; 1227 } 1228 } 1229 1230 if (olddentry == trap) 1231 goto out_dput; 1232 if (newdentry == trap) 1233 goto out_dput; 1234 1235 if (olddentry->d_inode == newdentry->d_inode) 1236 goto out_dput; 1237 1238 err = 0; 1239 if (ovl_type_merge_or_lower(old)) 1240 err = ovl_set_redirect(old, samedir); 1241 else if (is_dir && !old_opaque && ovl_type_merge(new->d_parent)) 1242 err = ovl_set_opaque_xerr(old, olddentry, -EXDEV); 1243 if (err) 1244 goto out_dput; 1245 1246 if (!overwrite && ovl_type_merge_or_lower(new)) 1247 err = ovl_set_redirect(new, samedir); 1248 else if (!overwrite && new_is_dir && !new_opaque && 1249 ovl_type_merge(old->d_parent)) 1250 err = ovl_set_opaque_xerr(new, newdentry, -EXDEV); 1251 if (err) 1252 goto out_dput; 1253 1254 err = ovl_do_rename(ofs, old_upperdir->d_inode, olddentry, 1255 new_upperdir->d_inode, newdentry, flags); 1256 if (err) 1257 goto out_dput; 1258 1259 if (cleanup_whiteout) 1260 ovl_cleanup(ofs, old_upperdir->d_inode, newdentry); 1261 1262 if (overwrite && d_inode(new)) { 1263 if (new_is_dir) 1264 clear_nlink(d_inode(new)); 1265 else 1266 ovl_drop_nlink(new); 1267 } 1268 1269 ovl_dir_modified(old->d_parent, ovl_type_origin(old) || 1270 (!overwrite && ovl_type_origin(new))); 1271 ovl_dir_modified(new->d_parent, ovl_type_origin(old) || 1272 (d_inode(new) && ovl_type_origin(new))); 1273 1274 /* copy ctime: */ 1275 ovl_copyattr(d_inode(old)); 1276 if (d_inode(new) && ovl_dentry_upper(new)) 1277 ovl_copyattr(d_inode(new)); 1278 1279out_dput: 1280 dput(newdentry); 1281out_dput_old: 1282 dput(olddentry); 1283out_unlock: 1284 unlock_rename(new_upperdir, old_upperdir); 1285out_revert_creds: 1286 revert_creds(old_cred); 1287 if (update_nlink) 1288 ovl_nlink_end(new); 1289out_drop_write: 1290 ovl_drop_write(old); 1291out: 1292 dput(opaquedir); 1293 ovl_cache_free(&list); 1294 return err; 1295} 1296 1297const struct inode_operations ovl_dir_inode_operations = { 1298 .lookup = ovl_lookup, 1299 .mkdir = ovl_mkdir, 1300 .symlink = ovl_symlink, 1301 .unlink = ovl_unlink, 1302 .rmdir = ovl_rmdir, 1303 .rename = ovl_rename, 1304 .link = ovl_link, 1305 .setattr = ovl_setattr, 1306 .create = ovl_create, 1307 .mknod = ovl_mknod, 1308 .permission = ovl_permission, 1309 .getattr = ovl_getattr, 1310 .listxattr = ovl_listxattr, 1311 .get_inode_acl = ovl_get_inode_acl, 1312 .get_acl = ovl_get_acl, 1313 .set_acl = ovl_set_acl, 1314 .update_time = ovl_update_time, 1315 .fileattr_get = ovl_fileattr_get, 1316 .fileattr_set = ovl_fileattr_set, 1317}; 1318