diff options
author | Linus Torvalds <torvalds@linux-foundation.org> | 2009-12-16 12:04:02 -0800 |
---|---|---|
committer | Linus Torvalds <torvalds@linux-foundation.org> | 2009-12-16 12:04:02 -0800 |
commit | bac5e54c29f352d962a2447d22735316b347b9f1 (patch) | |
tree | 7642993fa93164835ffaa2dacd341388193f1979 /fs/namei.c | |
parent | 529e89430d6c0d64db8ac474cb95e68e2527c79a (diff) | |
parent | c05c4edd876b7ae92787d1295868afcb89b6a348 (diff) | |
download | op-kernel-dev-bac5e54c29f352d962a2447d22735316b347b9f1.zip op-kernel-dev-bac5e54c29f352d962a2447d22735316b347b9f1.tar.gz |
Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6
* 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6: (38 commits)
direct I/O fallback sync simplification
ocfs: stop using do_sync_mapping_range
cleanup blockdev_direct_IO locking
make generic_acl slightly more generic
sanitize xattr handler prototypes
libfs: move EXPORT_SYMBOL for d_alloc_name
vfs: force reval of target when following LAST_BIND symlinks (try #7)
ima: limit imbalance msg
Untangling ima mess, part 3: kill dead code in ima
Untangling ima mess, part 2: deal with counters
Untangling ima mess, part 1: alloc_file()
O_TRUNC open shouldn't fail after file truncation
ima: call ima_inode_free ima_inode_free
IMA: clean up the IMA counts updating code
ima: only insert at inode creation time
ima: valid return code from ima_inode_alloc
fs: move get_empty_filp() deffinition to internal.h
Sanitize exec_permission_lite()
Kill cached_lookup() and real_lookup()
Kill path_lookup_open()
...
Trivial conflicts in fs/direct-io.c
Diffstat (limited to 'fs/namei.c')
-rw-r--r-- | fs/namei.c | 466 |
1 files changed, 233 insertions, 233 deletions
@@ -35,6 +35,8 @@ #include <linux/fs_struct.h> #include <asm/uaccess.h> +#include "internal.h" + #define ACC_MODE(x) ("\000\004\002\006"[(x)&O_ACCMODE]) /* [Feb-1997 T. Schoebel-Theuer] @@ -108,8 +110,6 @@ * any extra contention... */ -static int __link_path_walk(const char *name, struct nameidata *nd); - /* 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.. @@ -414,36 +414,55 @@ do_revalidate(struct dentry *dentry, struct nameidata *nd) } /* - * Internal lookup() using the new generic dcache. - * SMP-safe + * force_reval_path - force revalidation of a dentry + * + * In some situations the path walking code will trust dentries without + * revalidating them. This causes problems for filesystems that depend on + * d_revalidate to handle file opens (e.g. NFSv4). When FS_REVAL_DOT is set + * (which indicates that it's possible for the dentry to go stale), force + * a d_revalidate call before proceeding. + * + * Returns 0 if the revalidation was successful. If the revalidation fails, + * either return the error returned by d_revalidate or -ESTALE if the + * revalidation it just returned 0. If d_revalidate returns 0, we attempt to + * invalidate the dentry. It's up to the caller to handle putting references + * to the path if necessary. */ -static struct dentry * cached_lookup(struct dentry * parent, struct qstr * name, struct nameidata *nd) +static int +force_reval_path(struct path *path, struct nameidata *nd) { - struct dentry * dentry = __d_lookup(parent, name); + int status; + struct dentry *dentry = path->dentry; - /* lockess __d_lookup may fail due to concurrent d_move() - * in some unrelated directory, so try with d_lookup + /* + * only check on filesystems where it's possible for the dentry to + * become stale. It's assumed that if this flag is set then the + * d_revalidate op will also be defined. */ - if (!dentry) - dentry = d_lookup(parent, name); + if (!(dentry->d_sb->s_type->fs_flags & FS_REVAL_DOT)) + return 0; - if (dentry && dentry->d_op && dentry->d_op->d_revalidate) - dentry = do_revalidate(dentry, nd); + status = dentry->d_op->d_revalidate(dentry, nd); + if (status > 0) + return 0; - return dentry; + if (!status) { + d_invalidate(dentry); + status = -ESTALE; + } + return status; } /* - * 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. + * Short-cut version of permission(), for calling on directories + * during pathname resolution. 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 + * short-cut DAC fails, then call ->permission() to do more * complete permission check. */ -static int exec_permission_lite(struct inode *inode) +static int exec_permission(struct inode *inode) { int ret; @@ -465,99 +484,6 @@ ok: return security_inode_permission(inode, MAY_EXEC); } -/* - * 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; - - mutex_lock(&dir->i_mutex); - /* - * 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) { - 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); - result = ERR_PTR(-ENOMEM); - if (dentry) { - result = dir->i_op->lookup(dir, dentry, nd); - if (result) - dput(dentry); - else - result = dentry; - } -out_unlock: - mutex_unlock(&dir->i_mutex); - return result; - } - - /* - * Uhhuh! Nasty case: the cache was re-populated while - * we waited on the semaphore. Need to revalidate. - */ - mutex_unlock(&dir->i_mutex); - if (result->d_op && result->d_op->d_revalidate) { - result = do_revalidate(result, nd); - if (!result) - result = ERR_PTR(-ENOENT); - } - return result; -} - -/* - * 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 */ - path_get(&save); - - result = __link_path_walk(name, nd); - if (result == -ESTALE) { - /* nd->path had been dropped */ - nd->path = save; - path_get(&nd->path); - nd->flags |= LOOKUP_REVAL; - result = __link_path_walk(name, nd); - } - - path_put(&save); - - return result; -} - static __always_inline void set_root(struct nameidata *nd) { if (!nd->root.mnt) { @@ -569,6 +495,8 @@ static __always_inline void set_root(struct nameidata *nd) } } +static int link_path_walk(const char *, struct nameidata *); + static __always_inline int __vfs_follow_link(struct nameidata *nd, const char *link) { int res = 0; @@ -641,11 +569,14 @@ static __always_inline int __do_follow_link(struct path *path, struct nameidata error = 0; if (s) error = __vfs_follow_link(nd, s); + else if (nd->last_type == LAST_BIND) { + error = force_reval_path(&nd->path, nd); + if (error) + path_put(&nd->path); + } if (dentry->d_inode->i_op->put_link) dentry->d_inode->i_op->put_link(dentry, nd, cookie); } - path_put(path); - return error; } @@ -672,6 +603,7 @@ static inline int do_follow_link(struct path *path, struct nameidata *nd) current->total_link_count++; nd->depth++; err = __do_follow_link(path, nd); + path_put(path); current->link_count--; nd->depth--; return err; @@ -797,8 +729,19 @@ static int do_lookup(struct nameidata *nd, struct qstr *name, struct path *path) { struct vfsmount *mnt = nd->path.mnt; - struct dentry *dentry = __d_lookup(nd->path.dentry, name); + struct dentry *dentry, *parent; + struct inode *dir; + /* + * See if the low-level filesystem might want + * to use its own hash.. + */ + if (nd->path.dentry->d_op && nd->path.dentry->d_op->d_hash) { + int err = nd->path.dentry->d_op->d_hash(nd->path.dentry, name); + if (err < 0) + return err; + } + dentry = __d_lookup(nd->path.dentry, name); if (!dentry) goto need_lookup; if (dentry->d_op && dentry->d_op->d_revalidate) @@ -810,7 +753,59 @@ done: return 0; need_lookup: - dentry = real_lookup(nd->path.dentry, name, nd); + parent = nd->path.dentry; + dir = parent->d_inode; + + mutex_lock(&dir->i_mutex); + /* + * 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 + */ + dentry = d_lookup(parent, name); + if (!dentry) { + struct dentry *new; + + /* Don't create child dentry for a dead directory. */ + dentry = ERR_PTR(-ENOENT); + if (IS_DEADDIR(dir)) + goto out_unlock; + + new = d_alloc(parent, name); + dentry = ERR_PTR(-ENOMEM); + if (new) { + dentry = dir->i_op->lookup(dir, new, nd); + if (dentry) + dput(new); + else + dentry = new; + } +out_unlock: + mutex_unlock(&dir->i_mutex); + if (IS_ERR(dentry)) + goto fail; + goto done; + } + + /* + * Uhhuh! Nasty case: the cache was re-populated while + * we waited on the semaphore. Need to revalidate. + */ + mutex_unlock(&dir->i_mutex); + if (dentry->d_op && dentry->d_op->d_revalidate) { + dentry = do_revalidate(dentry, nd); + if (!dentry) + dentry = ERR_PTR(-ENOENT); + } if (IS_ERR(dentry)) goto fail; goto done; @@ -835,7 +830,7 @@ fail: * Returns 0 and nd will have valid dentry and mnt on success. * Returns error and drops reference to input namei data on failure. */ -static int __link_path_walk(const char *name, struct nameidata *nd) +static int link_path_walk(const char *name, struct nameidata *nd) { struct path next; struct inode *inode; @@ -858,7 +853,7 @@ static int __link_path_walk(const char *name, struct nameidata *nd) unsigned int c; nd->flags |= LOOKUP_CONTINUE; - err = exec_permission_lite(inode); + err = exec_permission(inode); if (err) break; @@ -898,16 +893,6 @@ static int __link_path_walk(const char *name, struct nameidata *nd) case 1: continue; } - /* - * See if the low-level filesystem might want - * to use its own hash.. - */ - 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); - if (err < 0) - break; - } /* This does the actual lookups.. */ err = do_lookup(nd, &this, &next); if (err) @@ -953,12 +938,6 @@ last_component: case 1: goto return_reval; } - 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); - if (err < 0) - break; - } err = do_lookup(nd, &this, &next); if (err) break; @@ -1017,8 +996,27 @@ return_err: static int path_walk(const char *name, struct nameidata *nd) { + struct path save = nd->path; + int result; + current->total_link_count = 0; - return link_path_walk(name, nd); + + /* make sure the stuff we saved doesn't go away */ + path_get(&save); + + result = link_path_walk(name, nd); + if (result == -ESTALE) { + /* nd->path had been dropped */ + current->total_link_count = 0; + nd->path = save; + path_get(&nd->path); + nd->flags |= LOOKUP_REVAL; + result = link_path_walk(name, nd); + } + + path_put(&save); + + return result; } static int path_init(int dfd, const char *name, unsigned int flags, struct nameidata *nd) @@ -1141,36 +1139,6 @@ int vfs_path_lookup(struct dentry *dentry, struct vfsmount *mnt, return retval; } -/** - * path_lookup_open - lookup a file path with open intent - * @dfd: the directory to use as base, or AT_FDCWD - * @name: pointer to file name - * @lookup_flags: lookup intent flags - * @nd: pointer to nameidata - * @open_flags: open intent flags - */ -static int path_lookup_open(int dfd, const char *name, - unsigned int lookup_flags, struct nameidata *nd, int open_flags) -{ - 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 = 0; - err = do_path_lookup(dfd, name, lookup_flags|LOOKUP_OPEN, nd); - if (IS_ERR(nd->intent.open.file)) { - if (err == 0) { - err = PTR_ERR(nd->intent.open.file); - path_put(&nd->path); - } - } else if (err != 0) - release_open_intent(nd); - return err; -} - static struct dentry *__lookup_hash(struct qstr *name, struct dentry *base, struct nameidata *nd) { @@ -1191,7 +1159,17 @@ static struct dentry *__lookup_hash(struct qstr *name, goto out; } - dentry = cached_lookup(base, name, nd); + dentry = __d_lookup(base, 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(base, name); + + if (dentry && dentry->d_op && dentry->d_op->d_revalidate) + dentry = do_revalidate(dentry, nd); + if (!dentry) { struct dentry *new; @@ -1223,7 +1201,7 @@ static struct dentry *lookup_hash(struct nameidata *nd) { int err; - err = inode_permission(nd->path.dentry->d_inode, MAY_EXEC); + err = exec_permission(nd->path.dentry->d_inode); if (err) return ERR_PTR(err); return __lookup_hash(&nd->last, nd->path.dentry, nd); @@ -1273,7 +1251,7 @@ struct dentry *lookup_one_len(const char *name, struct dentry *base, int len) if (err) return ERR_PTR(err); - err = inode_permission(base->d_inode, MAY_EXEC); + err = exec_permission(base->d_inode); if (err) return ERR_PTR(err); return __lookup_hash(&this, base, NULL); @@ -1511,69 +1489,45 @@ int may_open(struct path *path, int acc_mode, int flag) if (error) return error; - error = ima_path_check(path, acc_mode ? - acc_mode & (MAY_READ | MAY_WRITE | MAY_EXEC) : - ACC_MODE(flag) & (MAY_READ | MAY_WRITE), - IMA_COUNT_UPDATE); - - if (error) - return error; /* * An append-only file must be opened in append mode for writing. */ if (IS_APPEND(inode)) { - error = -EPERM; if ((flag & FMODE_WRITE) && !(flag & O_APPEND)) - goto err_out; + return -EPERM; if (flag & O_TRUNC) - goto err_out; + return -EPERM; } /* O_NOATIME can only be set by the owner or superuser */ - if (flag & O_NOATIME) - if (!is_owner_or_cap(inode)) { - error = -EPERM; - goto err_out; - } + if (flag & O_NOATIME && !is_owner_or_cap(inode)) + return -EPERM; /* * Ensure there are no outstanding leases on the file. */ - error = break_lease(inode, flag); - if (error) - goto err_out; - - if (flag & O_TRUNC) { - error = get_write_access(inode); - if (error) - goto err_out; - - /* - * Refuse to truncate files with mandatory locks held on them. - */ - error = locks_verify_locked(inode); - if (!error) - error = security_path_truncate(path, 0, - ATTR_MTIME|ATTR_CTIME|ATTR_OPEN); - if (!error) { - vfs_dq_init(inode); - - error = do_truncate(dentry, 0, - ATTR_MTIME|ATTR_CTIME|ATTR_OPEN, - NULL); - } - put_write_access(inode); - if (error) - goto err_out; - } else - if (flag & FMODE_WRITE) - vfs_dq_init(inode); + return break_lease(inode, flag); +} - return 0; -err_out: - ima_counts_put(path, acc_mode ? - acc_mode & (MAY_READ | MAY_WRITE | MAY_EXEC) : - ACC_MODE(flag) & (MAY_READ | MAY_WRITE)); +static int handle_truncate(struct path *path) +{ + struct inode *inode = path->dentry->d_inode; + int 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) + error = security_path_truncate(path, 0, + ATTR_MTIME|ATTR_CTIME|ATTR_OPEN); + if (!error) { + error = do_truncate(path->dentry, 0, + ATTR_MTIME|ATTR_CTIME|ATTR_OPEN, + NULL); + } + put_write_access(inode); return error; } @@ -1628,7 +1582,7 @@ static inline int open_to_namei_flags(int flag) return flag; } -static int open_will_write_to_fs(int flag, struct inode *inode) +static int open_will_truncate(int flag, struct inode *inode) { /* * We'll never write to the fs underlying @@ -1650,10 +1604,10 @@ struct file *do_filp_open(int dfd, const char *pathname, struct file *filp; struct nameidata nd; int error; - struct path path; + struct path path, save; struct dentry *dir; int count = 0; - int will_write; + int will_truncate; int flag = open_to_namei_flags(open_flag); /* @@ -1681,8 +1635,22 @@ struct file *do_filp_open(int dfd, const char *pathname, * The simplest case - just a plain lookup. */ if (!(flag & O_CREAT)) { - error = path_lookup_open(dfd, pathname, lookup_flags(flag), - &nd, flag); + filp = get_empty_filp(); + + if (filp == NULL) + return ERR_PTR(-ENFILE); + nd.intent.open.file = filp; + nd.intent.open.flags = flag; + nd.intent.open.create_mode = 0; + error = do_path_lookup(dfd, pathname, + lookup_flags(flag)|LOOKUP_OPEN, &nd); + if (IS_ERR(nd.intent.open.file)) { + if (error == 0) { + error = PTR_ERR(nd.intent.open.file); + path_put(&nd.path); + } + } else if (error) + release_open_intent(&nd); if (error) return ERR_PTR(error); goto ok; @@ -1758,13 +1726,17 @@ do_last: goto exit; } filp = nameidata_to_filp(&nd, open_flag); - if (IS_ERR(filp)) - ima_counts_put(&nd.path, - acc_mode & (MAY_READ | MAY_WRITE | - MAY_EXEC)); mnt_drop_write(nd.path.mnt); if (nd.root.mnt) path_put(&nd.root); + if (!IS_ERR(filp)) { + error = ima_path_check(&filp->f_path, filp->f_mode & + (MAY_READ | MAY_WRITE | MAY_EXEC)); + if (error) { + fput(filp); + filp = ERR_PTR(error); + } + } return filp; } @@ -1805,28 +1777,45 @@ ok: * 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) { + will_truncate = open_will_truncate(flag, nd.path.dentry->d_inode); + if (will_truncate) { error = mnt_want_write(nd.path.mnt); if (error) goto exit; } error = may_open(&nd.path, acc_mode, flag); if (error) { - if (will_write) + if (will_truncate) mnt_drop_write(nd.path.mnt); goto exit; } filp = nameidata_to_filp(&nd, open_flag); - if (IS_ERR(filp)) - ima_counts_put(&nd.path, - acc_mode & (MAY_READ | MAY_WRITE | MAY_EXEC)); + if (!IS_ERR(filp)) { + error = ima_path_check(&filp->f_path, filp->f_mode & + (MAY_READ | MAY_WRITE | MAY_EXEC)); + if (error) { + fput(filp); + filp = ERR_PTR(error); + } + } + if (!IS_ERR(filp)) { + if (acc_mode & MAY_WRITE) + vfs_dq_init(nd.path.dentry->d_inode); + + if (will_truncate) { + error = handle_truncate(&nd.path); + if (error) { + fput(filp); + filp = ERR_PTR(error); + } + } + } /* * It is now safe to drop the mnt write * because the filp has had a write taken * on its behalf. */ - if (will_write) + if (will_truncate) mnt_drop_write(nd.path.mnt); if (nd.root.mnt) path_put(&nd.root); @@ -1863,7 +1852,18 @@ do_link: error = security_inode_follow_link(path.dentry, &nd); if (error) goto exit_dput; + save = nd.path; + path_get(&save); error = __do_follow_link(&path, &nd); + if (error == -ESTALE) { + /* nd.path had been dropped */ + nd.path = save; + path_get(&nd.path); + nd.flags |= LOOKUP_REVAL; + error = __do_follow_link(&path, &nd); + } + path_put(&save); + path_put(&path); if (error) { /* Does someone understand code flow here? Or it is only * me so stupid? Anathema to whoever designed this non-sense |