diff options
Diffstat (limited to 'fs/configfs')
| -rw-r--r-- | fs/configfs/Kconfig | 11 | ||||
| -rw-r--r-- | fs/configfs/configfs_internal.h | 58 | ||||
| -rw-r--r-- | fs/configfs/dir.c | 1192 | ||||
| -rw-r--r-- | fs/configfs/file.c | 120 | ||||
| -rw-r--r-- | fs/configfs/inode.c | 231 | ||||
| -rw-r--r-- | fs/configfs/item.c | 87 | ||||
| -rw-r--r-- | fs/configfs/mount.c | 79 | ||||
| -rw-r--r-- | fs/configfs/symlink.c | 79 |
8 files changed, 1342 insertions, 515 deletions
diff --git a/fs/configfs/Kconfig b/fs/configfs/Kconfig new file mode 100644 index 00000000000..9febcdefdfd --- /dev/null +++ b/fs/configfs/Kconfig @@ -0,0 +1,11 @@ +config CONFIGFS_FS + tristate "Userspace-driven configuration filesystem" + select SYSFS + help + configfs is a RAM-based filesystem that provides the converse + of sysfs's functionality. Where sysfs is a filesystem-based + view of kernel objects, configfs is a filesystem-based manager + of kernel objects, or config_items. + + Both sysfs and configfs can and should exist together on the + same system. One is not a replacement for the other. diff --git a/fs/configfs/configfs_internal.h b/fs/configfs/configfs_internal.h index 8899d9c5f6b..bd4a3c16709 100644 --- a/fs/configfs/configfs_internal.h +++ b/fs/configfs/configfs_internal.h @@ -24,56 +24,78 @@ * configfs Copyright (C) 2005 Oracle. All rights reserved. */ +#ifdef pr_fmt +#undef pr_fmt +#endif + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/slab.h> #include <linux/list.h> +#include <linux/spinlock.h> struct configfs_dirent { atomic_t s_count; + int s_dependent_count; struct list_head s_sibling; struct list_head s_children; struct list_head s_links; - void * s_element; + void * s_element; int s_type; umode_t s_mode; struct dentry * s_dentry; + struct iattr * s_iattr; +#ifdef CONFIG_LOCKDEP + int s_depth; +#endif }; #define CONFIGFS_ROOT 0x0001 #define CONFIGFS_DIR 0x0002 -#define CONFIGFS_ITEM_ATTR 0x0004 -#define CONFIGFS_ITEM_LINK 0x0020 +#define CONFIGFS_ITEM_ATTR 0x0004 +#define CONFIGFS_ITEM_LINK 0x0020 #define CONFIGFS_USET_DIR 0x0040 #define CONFIGFS_USET_DEFAULT 0x0080 #define CONFIGFS_USET_DROPPING 0x0100 +#define CONFIGFS_USET_IN_MKDIR 0x0200 +#define CONFIGFS_USET_CREATING 0x0400 #define CONFIGFS_NOT_PINNED (CONFIGFS_ITEM_ATTR) -extern struct vfsmount * configfs_mount; +extern struct mutex configfs_symlink_mutex; +extern spinlock_t configfs_dirent_lock; + +extern struct kmem_cache *configfs_dir_cachep; extern int configfs_is_root(struct config_item *item); -extern struct inode * configfs_new_inode(mode_t mode); -extern int configfs_create(struct dentry *, int mode, int (*init)(struct inode *)); +extern struct inode * configfs_new_inode(umode_t mode, struct configfs_dirent *, struct super_block *); +extern int configfs_create(struct dentry *, umode_t mode, int (*init)(struct inode *)); +extern int configfs_inode_init(void); +extern void configfs_inode_exit(void); extern int configfs_create_file(struct config_item *, const struct configfs_attribute *); extern int configfs_make_dirent(struct configfs_dirent *, struct dentry *, void *, umode_t, int); +extern int configfs_dirent_is_ready(struct configfs_dirent *); extern int configfs_add_file(struct dentry *, const struct configfs_attribute *, int); extern void configfs_hash_and_remove(struct dentry * dir, const char * name); extern const unsigned char * configfs_get_name(struct configfs_dirent *sd); extern void configfs_drop_dentry(struct configfs_dirent *sd, struct dentry *parent); +extern int configfs_setattr(struct dentry *dentry, struct iattr *iattr); -extern int configfs_pin_fs(void); +extern struct dentry *configfs_pin_fs(void); extern void configfs_release_fs(void); extern struct rw_semaphore configfs_rename_sem; -extern struct super_block * configfs_sb; -extern struct file_operations configfs_dir_operations; -extern struct file_operations configfs_file_operations; -extern struct file_operations bin_fops; -extern struct inode_operations configfs_dir_inode_operations; -extern struct inode_operations configfs_symlink_inode_operations; +extern const struct file_operations configfs_dir_operations; +extern const struct file_operations configfs_file_operations; +extern const struct file_operations bin_fops; +extern const struct inode_operations configfs_dir_inode_operations; +extern const struct inode_operations configfs_root_inode_operations; +extern const struct inode_operations configfs_symlink_inode_operations; +extern const struct dentry_operations configfs_dentry_ops; extern int configfs_symlink(struct inode *dir, struct dentry *dentry, const char *symname); @@ -104,7 +126,7 @@ static inline struct config_item *configfs_get_config_item(struct dentry *dentry { struct config_item * item = NULL; - spin_lock(&dcache_lock); + spin_lock(&dentry->d_lock); if (!d_unhashed(dentry)) { struct configfs_dirent * sd = dentry->d_fsdata; if (sd->s_type & CONFIGFS_ITEM_LINK) { @@ -113,15 +135,17 @@ static inline struct config_item *configfs_get_config_item(struct dentry *dentry } else item = config_item_get(sd->s_element); } - spin_unlock(&dcache_lock); + spin_unlock(&dentry->d_lock); return item; } static inline void release_configfs_dirent(struct configfs_dirent * sd) { - if (!(sd->s_type & CONFIGFS_ROOT)) - kfree(sd); + if (!(sd->s_type & CONFIGFS_ROOT)) { + kfree(sd->s_iattr); + kmem_cache_free(configfs_dir_cachep, sd); + } } static inline struct configfs_dirent * configfs_get(struct configfs_dirent * sd) diff --git a/fs/configfs/dir.c b/fs/configfs/dir.c index b668ec61527..668dcabc569 100644 --- a/fs/configfs/dir.c +++ b/fs/configfs/dir.c @@ -30,79 +30,212 @@ #include <linux/mount.h> #include <linux/module.h> #include <linux/slab.h> +#include <linux/err.h> #include <linux/configfs.h> #include "configfs_internal.h" DECLARE_RWSEM(configfs_rename_sem); +/* + * Protects mutations of configfs_dirent linkage together with proper i_mutex + * Also protects mutations of symlinks linkage to target configfs_dirent + * Mutators of configfs_dirent linkage must *both* have the proper inode locked + * and configfs_dirent_lock locked, in that order. + * This allows one to safely traverse configfs_dirent trees and symlinks without + * having to lock inodes. + * + * Protects setting of CONFIGFS_USET_DROPPING: checking the flag + * unlocked is not reliable unless in detach_groups() called from + * rmdir()/unregister() and from configfs_attach_group() + */ +DEFINE_SPINLOCK(configfs_dirent_lock); static void configfs_d_iput(struct dentry * dentry, struct inode * inode) { - struct configfs_dirent * sd = dentry->d_fsdata; + struct configfs_dirent *sd = dentry->d_fsdata; if (sd) { - BUG_ON(sd->s_dentry != dentry); - sd->s_dentry = NULL; + /* Coordinate with configfs_readdir */ + spin_lock(&configfs_dirent_lock); + /* Coordinate with configfs_attach_attr where will increase + * sd->s_count and update sd->s_dentry to new allocated one. + * Only set sd->dentry to null when this dentry is the only + * sd owner. + * If not do so, configfs_d_iput may run just after + * configfs_attach_attr and set sd->s_dentry to null + * even it's still in use. + */ + if (atomic_read(&sd->s_count) <= 2) + sd->s_dentry = NULL; + + spin_unlock(&configfs_dirent_lock); configfs_put(sd); } iput(inode); } +const struct dentry_operations configfs_dentry_ops = { + .d_iput = configfs_d_iput, + .d_delete = always_delete_dentry, +}; + +#ifdef CONFIG_LOCKDEP + /* - * We _must_ delete our dentries on last dput, as the chain-to-parent - * behavior is required to clear the parents of default_groups. + * Helpers to make lockdep happy with our recursive locking of default groups' + * inodes (see configfs_attach_group() and configfs_detach_group()). + * We put default groups i_mutexes in separate classes according to their depth + * from the youngest non-default group ancestor. + * + * For a non-default group A having default groups A/B, A/C, and A/C/D, default + * groups A/B and A/C will have their inode's mutex in class + * default_group_class[0], and default group A/C/D will be in + * default_group_class[1]. + * + * The lock classes are declared and assigned in inode.c, according to the + * s_depth value. + * The s_depth value is initialized to -1, adjusted to >= 0 when attaching + * default groups, and reset to -1 when all default groups are attached. During + * attachment, if configfs_create() sees s_depth > 0, the lock class of the new + * inode's mutex is set to default_group_class[s_depth - 1]. */ -static int configfs_d_delete(struct dentry *dentry) + +static void configfs_init_dirent_depth(struct configfs_dirent *sd) { - return 1; + sd->s_depth = -1; } -static struct dentry_operations configfs_dentry_ops = { - .d_iput = configfs_d_iput, - /* simple_delete_dentry() isn't exported */ - .d_delete = configfs_d_delete, -}; +static void configfs_set_dir_dirent_depth(struct configfs_dirent *parent_sd, + struct configfs_dirent *sd) +{ + int parent_depth = parent_sd->s_depth; + + if (parent_depth >= 0) + sd->s_depth = parent_depth + 1; +} + +static void +configfs_adjust_dir_dirent_depth_before_populate(struct configfs_dirent *sd) +{ + /* + * item's i_mutex class is already setup, so s_depth is now only + * used to set new sub-directories s_depth, which is always done + * with item's i_mutex locked. + */ + /* + * sd->s_depth == -1 iff we are a non default group. + * else (we are a default group) sd->s_depth > 0 (see + * create_dir()). + */ + if (sd->s_depth == -1) + /* + * We are a non default group and we are going to create + * default groups. + */ + sd->s_depth = 0; +} + +static void +configfs_adjust_dir_dirent_depth_after_populate(struct configfs_dirent *sd) +{ + /* We will not create default groups anymore. */ + sd->s_depth = -1; +} + +#else /* CONFIG_LOCKDEP */ + +static void configfs_init_dirent_depth(struct configfs_dirent *sd) +{ +} + +static void configfs_set_dir_dirent_depth(struct configfs_dirent *parent_sd, + struct configfs_dirent *sd) +{ +} + +static void +configfs_adjust_dir_dirent_depth_before_populate(struct configfs_dirent *sd) +{ +} + +static void +configfs_adjust_dir_dirent_depth_after_populate(struct configfs_dirent *sd) +{ +} + +#endif /* CONFIG_LOCKDEP */ /* * Allocates a new configfs_dirent and links it to the parent configfs_dirent */ -static struct configfs_dirent *configfs_new_dirent(struct configfs_dirent * parent_sd, - void * element) +static struct configfs_dirent *configfs_new_dirent(struct configfs_dirent *parent_sd, + void *element, int type) { struct configfs_dirent * sd; - sd = kmalloc(sizeof(*sd), GFP_KERNEL); + sd = kmem_cache_zalloc(configfs_dir_cachep, GFP_KERNEL); if (!sd) - return NULL; + return ERR_PTR(-ENOMEM); - memset(sd, 0, sizeof(*sd)); atomic_set(&sd->s_count, 1); INIT_LIST_HEAD(&sd->s_links); INIT_LIST_HEAD(&sd->s_children); - list_add(&sd->s_sibling, &parent_sd->s_children); sd->s_element = element; + sd->s_type = type; + configfs_init_dirent_depth(sd); + spin_lock(&configfs_dirent_lock); + if (parent_sd->s_type & CONFIGFS_USET_DROPPING) { + spin_unlock(&configfs_dirent_lock); + kmem_cache_free(configfs_dir_cachep, sd); + return ERR_PTR(-ENOENT); + } + list_add(&sd->s_sibling, &parent_sd->s_children); + spin_unlock(&configfs_dirent_lock); return sd; } +/* + * + * Return -EEXIST if there is already a configfs element with the same + * name for the same parent. + * + * called with parent inode's i_mutex held + */ +static int configfs_dirent_exists(struct configfs_dirent *parent_sd, + const unsigned char *new) +{ + struct configfs_dirent * sd; + + list_for_each_entry(sd, &parent_sd->s_children, s_sibling) { + if (sd->s_element) { + const unsigned char *existing = configfs_get_name(sd); + if (strcmp(existing, new)) + continue; + else + return -EEXIST; + } + } + + return 0; +} + + int configfs_make_dirent(struct configfs_dirent * parent_sd, struct dentry * dentry, void * element, umode_t mode, int type) { struct configfs_dirent * sd; - sd = configfs_new_dirent(parent_sd, element); - if (!sd) - return -ENOMEM; + sd = configfs_new_dirent(parent_sd, element, type); + if (IS_ERR(sd)) + return PTR_ERR(sd); sd->s_mode = mode; - sd->s_type = type; sd->s_dentry = dentry; - if (dentry) { + if (dentry) dentry->d_fsdata = configfs_get(sd); - dentry->d_op = &configfs_dentry_ops; - } return 0; } @@ -113,11 +246,11 @@ static int init_dir(struct inode * inode) inode->i_fop = &configfs_dir_operations; /* directory inodes start off with i_nlink == 2 (for "." entry) */ - inode->i_nlink++; + inc_nlink(inode); return 0; } -static int init_file(struct inode * inode) +static int configfs_init_file(struct inode * inode) { inode->i_size = PAGE_SIZE; inode->i_fop = &configfs_file_operations; @@ -130,19 +263,31 @@ static int init_symlink(struct inode * inode) return 0; } -static int create_dir(struct config_item * k, struct dentry * p, - struct dentry * d) +static int create_dir(struct config_item *k, struct dentry *d) { int error; umode_t mode = S_IFDIR| S_IRWXU | S_IRUGO | S_IXUGO; + struct dentry *p = d->d_parent; - error = configfs_create(d, mode, init_dir); - if (!error) { + BUG_ON(!k); + + error = configfs_dirent_exists(p->d_fsdata, d->d_name.name); + if (!error) error = configfs_make_dirent(p->d_fsdata, d, k, mode, - CONFIGFS_DIR); + CONFIGFS_DIR | CONFIGFS_USET_CREATING); + if (!error) { + configfs_set_dir_dirent_depth(p->d_fsdata, d->d_fsdata); + error = configfs_create(d, mode, init_dir); if (!error) { - p->d_inode->i_nlink++; - (d)->d_op = &configfs_dentry_ops; + inc_nlink(p->d_inode); + } else { + struct configfs_dirent *sd = d->d_fsdata; + if (sd) { + spin_lock(&configfs_dirent_lock); + list_del_init(&sd->s_sibling); + spin_unlock(&configfs_dirent_lock); + configfs_put(sd); + } } } return error; @@ -153,28 +298,57 @@ static int create_dir(struct config_item * k, struct dentry * p, * configfs_create_dir - create a directory for an config_item. * @item: config_itemwe're creating directory for. * @dentry: config_item's dentry. + * + * Note: user-created entries won't be allowed under this new directory + * until it is validated by configfs_dir_set_ready() */ static int configfs_create_dir(struct config_item * item, struct dentry *dentry) { - struct dentry * parent; - int error = 0; - - BUG_ON(!item); - - if (item->ci_parent) - parent = item->ci_parent->ci_dentry; - else if (configfs_mount && configfs_mount->mnt_sb) - parent = configfs_mount->mnt_sb->s_root; - else - return -EFAULT; - - error = create_dir(item,parent,dentry); + int error = create_dir(item, dentry); if (!error) item->ci_dentry = dentry; return error; } +/* + * Allow userspace to create new entries under a new directory created with + * configfs_create_dir(), and under all of its chidlren directories recursively. + * @sd configfs_dirent of the new directory to validate + * + * Caller must hold configfs_dirent_lock. + */ +static void configfs_dir_set_ready(struct configfs_dirent *sd) +{ + struct configfs_dirent *child_sd; + + sd->s_type &= ~CONFIGFS_USET_CREATING; + list_for_each_entry(child_sd, &sd->s_children, s_sibling) + if (child_sd->s_type & CONFIGFS_USET_CREATING) + configfs_dir_set_ready(child_sd); +} + +/* + * Check that a directory does not belong to a directory hierarchy being + * attached and not validated yet. + * @sd configfs_dirent of the directory to check + * + * @return non-zero iff the directory was validated + * + * Note: takes configfs_dirent_lock, so the result may change from false to true + * in two consecutive calls, but never from true to false. + */ +int configfs_dirent_is_ready(struct configfs_dirent *sd) +{ + int ret; + + spin_lock(&configfs_dirent_lock); + ret = !(sd->s_type & CONFIGFS_USET_CREATING); + spin_unlock(&configfs_dirent_lock); + + return ret; +} + int configfs_create_link(struct configfs_symlink *sl, struct dentry *parent, struct dentry *dentry) @@ -182,12 +356,19 @@ int configfs_create_link(struct configfs_symlink *sl, int err = 0; umode_t mode = S_IFLNK | S_IRWXUGO; - err = configfs_create(dentry, mode, init_symlink); + err = configfs_make_dirent(parent->d_fsdata, dentry, sl, mode, + CONFIGFS_ITEM_LINK); if (!err) { - err = configfs_make_dirent(parent->d_fsdata, dentry, sl, - mode, CONFIGFS_ITEM_LINK); - if (!err) - dentry->d_op = &configfs_dentry_ops; + err = configfs_create(dentry, mode, init_symlink); + if (err) { + struct configfs_dirent *sd = dentry->d_fsdata; + if (sd) { + spin_lock(&configfs_dirent_lock); + list_del_init(&sd->s_sibling); + spin_unlock(&configfs_dirent_lock); + configfs_put(sd); + } + } } return err; } @@ -198,13 +379,14 @@ static void remove_dir(struct dentry * d) struct configfs_dirent * sd; sd = d->d_fsdata; - list_del_init(&sd->s_sibling); + spin_lock(&configfs_dirent_lock); + list_del_init(&sd->s_sibling); + spin_unlock(&configfs_dirent_lock); configfs_put(sd); if (d->d_inode) simple_rmdir(parent->d_inode,d); - pr_debug(" o %s removing done (%d)\n",d->d_name.name, - atomic_read(&d->d_count)); + pr_debug(" o %s removing done (%d)\n",d->d_name.name, d_count(d)); dput(parent); } @@ -216,6 +398,8 @@ static void remove_dir(struct dentry * d) * The only thing special about this is that we remove any files in * the directory before we remove the directory, and we've inlined * what used to be configfs_rmdir() below, instead of calling separately. + * + * Caller holds the mutex of the item's inode */ static void configfs_remove_dir(struct config_item * item) @@ -241,13 +425,18 @@ static int configfs_attach_attr(struct configfs_dirent * sd, struct dentry * den struct configfs_attribute * attr = sd->s_element; int error; - error = configfs_create(dentry, (attr->ca_mode & S_IALLUGO) | S_IFREG, init_file); - if (error) - return error; - - dentry->d_op = &configfs_dentry_ops; + spin_lock(&configfs_dirent_lock); dentry->d_fsdata = configfs_get(sd); sd->s_dentry = dentry; + spin_unlock(&configfs_dirent_lock); + + error = configfs_create(dentry, (attr->ca_mode & S_IALLUGO) | S_IFREG, + configfs_init_file); + if (error) { + configfs_put(sd); + return error; + } + d_rehash(dentry); return 0; @@ -255,12 +444,24 @@ static int configfs_attach_attr(struct configfs_dirent * sd, struct dentry * den static struct dentry * configfs_lookup(struct inode *dir, struct dentry *dentry, - struct nameidata *nd) + unsigned int flags) { struct configfs_dirent * parent_sd = dentry->d_parent->d_fsdata; struct configfs_dirent * sd; int found = 0; - int err = 0; + int err; + + /* + * Fake invisibility if dir belongs to a group/default groups hierarchy + * being attached + * + * This forbids userspace to read/write attributes of items which may + * not complete their initialization, since the dentries of the + * attributes won't be instantiated. + */ + err = -ENOENT; + if (!configfs_dirent_is_ready(parent_sd)) + goto out; list_for_each_entry(sd, &parent_sd->s_children, s_sibling) { if (sd->s_type & CONFIGFS_NOT_PINNED) { @@ -280,42 +481,57 @@ static struct dentry * configfs_lookup(struct inode *dir, * If it doesn't exist and it isn't a NOT_PINNED item, * it must be negative. */ - return simple_lookup(dir, dentry, nd); + if (dentry->d_name.len > NAME_MAX) + return ERR_PTR(-ENAMETOOLONG); + d_add(dentry, NULL); + return NULL; } +out: return ERR_PTR(err); } /* * Only subdirectories count here. Files (CONFIGFS_NOT_PINNED) are - * attributes and are removed by rmdir(). We recurse, taking i_mutex - * on all children that are candidates for default detach. If the - * result is clean, then configfs_detach_group() will handle dropping - * i_mutex. If there is an error, the caller will clean up the i_mutex - * holders via configfs_detach_rollback(). + * attributes and are removed by rmdir(). We recurse, setting + * CONFIGFS_USET_DROPPING on all children that are candidates for + * default detach. + * If there is an error, the caller will reset the flags via + * configfs_detach_rollback(). */ -static int configfs_detach_prep(struct dentry *dentry) +static int configfs_detach_prep(struct dentry *dentry, struct mutex **wait_mutex) { struct configfs_dirent *parent_sd = dentry->d_fsdata; struct configfs_dirent *sd; int ret; + /* Mark that we're trying to drop the group */ + parent_sd->s_type |= CONFIGFS_USET_DROPPING; + ret = -EBUSY; if (!list_empty(&parent_sd->s_links)) goto out; ret = 0; list_for_each_entry(sd, &parent_sd->s_children, s_sibling) { - if (sd->s_type & CONFIGFS_NOT_PINNED) + if (!sd->s_element || + (sd->s_type & CONFIGFS_NOT_PINNED)) continue; if (sd->s_type & CONFIGFS_USET_DEFAULT) { - mutex_lock(&sd->s_dentry->d_inode->i_mutex); - /* Mark that we've taken i_mutex */ - sd->s_type |= CONFIGFS_USET_DROPPING; + /* Abort if racing with mkdir() */ + if (sd->s_type & CONFIGFS_USET_IN_MKDIR) { + if (wait_mutex) + *wait_mutex = &sd->s_dentry->d_inode->i_mutex; + return -EAGAIN; + } - ret = configfs_detach_prep(sd->s_dentry); + /* + * Yup, recursive. If there's a problem, blame + * deep nesting of default_groups + */ + ret = configfs_detach_prep(sd->s_dentry, wait_mutex); if (!ret) - continue; + continue; } else ret = -ENOTEMPTY; @@ -327,7 +543,7 @@ out: } /* - * Walk the tree, dropping i_mutex wherever CONFIGFS_USET_DROPPING is + * Walk the tree, resetting CONFIGFS_USET_DROPPING wherever it was * set. */ static void configfs_detach_rollback(struct dentry *dentry) @@ -335,16 +551,11 @@ static void configfs_detach_rollback(struct dentry *dentry) struct configfs_dirent *parent_sd = dentry->d_fsdata; struct configfs_dirent *sd; - list_for_each_entry(sd, &parent_sd->s_children, s_sibling) { - if (sd->s_type & CONFIGFS_USET_DEFAULT) { - configfs_detach_rollback(sd->s_dentry); + parent_sd->s_type &= ~CONFIGFS_USET_DROPPING; - if (sd->s_type & CONFIGFS_USET_DROPPING) { - sd->s_type &= ~CONFIGFS_USET_DROPPING; - mutex_unlock(&sd->s_dentry->d_inode->i_mutex); - } - } - } + list_for_each_entry(sd, &parent_sd->s_children, s_sibling) + if (sd->s_type & CONFIGFS_USET_DEFAULT) + configfs_detach_rollback(sd->s_dentry); } static void detach_attrs(struct config_item * item) @@ -363,7 +574,9 @@ static void detach_attrs(struct config_item * item) list_for_each_entry_safe(sd, tmp, &parent_sd->s_children, s_sibling) { if (!sd->s_element || !(sd->s_type & CONFIGFS_NOT_PINNED)) continue; + spin_lock(&configfs_dirent_lock); list_del_init(&sd->s_sibling); + spin_unlock(&configfs_dirent_lock); configfs_drop_dentry(sd, dentry); configfs_put(sd); } @@ -419,16 +632,13 @@ static void detach_groups(struct config_group *group) child = sd->s_dentry; + mutex_lock(&child->d_inode->i_mutex); + configfs_detach_group(sd->s_element); child->d_inode->i_flags |= S_DEAD; + dont_mount(child); - /* - * From rmdir/unregister, a configfs_detach_prep() pass - * has taken our i_mutex for us. Drop it. - * From mkdir/register cleanup, there is no sem held. - */ - if (sd->s_type & CONFIGFS_USET_DROPPING) - mutex_unlock(&child->d_inode->i_mutex); + mutex_unlock(&child->d_inode->i_mutex); d_delete(child); dput(child); @@ -452,19 +662,15 @@ static int create_default_group(struct config_group *parent_group, struct config_group *group) { int ret; - struct qstr name; struct configfs_dirent *sd; /* We trust the caller holds a reference to parent */ struct dentry *child, *parent = parent_group->cg_item.ci_dentry; if (!group->cg_item.ci_name) group->cg_item.ci_name = group->cg_item.ci_namebuf; - name.name = group->cg_item.ci_name; - name.len = strlen(name.name); - name.hash = full_name_hash(name.name, name.len); ret = -ENOMEM; - child = d_alloc(parent, &name); + child = d_alloc_name(parent, group->cg_item.ci_name); if (child) { d_add(child, NULL); @@ -474,7 +680,8 @@ static int create_default_group(struct config_group *parent_group, sd = child->d_fsdata; sd->s_type |= CONFIGFS_USET_DEFAULT; } else { - d_delete(child); + BUG_ON(child->d_inode); + d_drop(child); dput(child); } } @@ -485,40 +692,27 @@ static int create_default_group(struct config_group *parent_group, static int populate_groups(struct config_group *group) { struct config_group *new_group; - struct dentry *dentry = group->cg_item.ci_dentry; int ret = 0; int i; - if (group && group->default_groups) { - /* FYI, we're faking mkdir here - * I'm not sure we need this semaphore, as we're called - * from our parent's mkdir. That holds our parent's - * i_mutex, so afaik lookup cannot continue through our - * parent to find us, let alone mess with our tree. - * That said, taking our i_mutex is closer to mkdir - * emulation, and shouldn't hurt. */ - mutex_lock(&dentry->d_inode->i_mutex); - + if (group->default_groups) { for (i = 0; group->default_groups[i]; i++) { new_group = group->default_groups[i]; ret = create_default_group(group, new_group); - if (ret) + if (ret) { + detach_groups(group); break; + } } - - mutex_unlock(&dentry->d_inode->i_mutex); } - if (ret) - detach_groups(group); - return ret; } /* * All of link_obj/unlink_obj/link_group/unlink_group require that - * subsys->su_sem is held. + * subsys->su_mutex is held. */ static void unlink_obj(struct config_item *item) @@ -531,20 +725,34 @@ static void unlink_obj(struct config_item *item) item->ci_group = NULL; item->ci_parent = NULL; + + /* Drop the reference for ci_entry */ config_item_put(item); + /* Drop the reference for ci_parent */ config_group_put(group); } } static void link_obj(struct config_item *parent_item, struct config_item *item) { - /* Parent seems redundant with group, but it makes certain - * traversals much nicer. */ + /* + * Parent seems redundant with group, but it makes certain + * traversals much nicer. + */ item->ci_parent = parent_item; + + /* + * We hold a reference on the parent for the child's ci_parent + * link. + */ item->ci_group = config_group_get(to_config_group(parent_item)); list_add_tail(&item->ci_entry, &item->ci_group->cg_children); + /* + * We hold a reference on the child for ci_entry on the parent's + * cg_children + */ config_item_get(item); } @@ -613,7 +821,16 @@ static int configfs_attach_item(struct config_item *parent_item, if (!ret) { ret = populate_attrs(item); if (ret) { + /* + * We are going to remove an inode and its dentry but + * the VFS may already have hit and used them. Thus, + * we must lock them as rmdir() would. + */ + mutex_lock(&dentry->d_inode->i_mutex); configfs_remove_dir(item); + dentry->d_inode->i_flags |= S_DEAD; + dont_mount(dentry); + mutex_unlock(&dentry->d_inode->i_mutex); d_delete(dentry); } } @@ -621,6 +838,7 @@ static int configfs_attach_item(struct config_item *parent_item, return ret; } +/* Caller holds the mutex of the item's inode */ static void configfs_detach_item(struct config_item *item) { detach_attrs(item); @@ -639,16 +857,33 @@ static int configfs_attach_group(struct config_item *parent_item, sd = dentry->d_fsdata; sd->s_type |= CONFIGFS_USET_DIR; + /* + * FYI, we're faking mkdir in populate_groups() + * We must lock the group's inode to avoid races with the VFS + * which can already hit the inode and try to add/remove entries + * under it. + * + * We must also lock the inode to remove it safely in case of + * error, as rmdir() would. + */ + mutex_lock_nested(&dentry->d_inode->i_mutex, I_MUTEX_CHILD); + configfs_adjust_dir_dirent_depth_before_populate(sd); ret = populate_groups(to_config_group(item)); if (ret) { configfs_detach_item(item); - d_delete(dentry); + dentry->d_inode->i_flags |= S_DEAD; + dont_mount(dentry); } + configfs_adjust_dir_dirent_depth_after_populate(sd); + mutex_unlock(&dentry->d_inode->i_mutex); + if (ret) + d_delete(dentry); } return ret; } +/* Caller holds the mutex of the group's inode */ static void configfs_detach_group(struct config_item *item) { detach_groups(to_config_group(item)); @@ -656,6 +891,28 @@ static void configfs_detach_group(struct config_item *item) } /* + * After the item has been detached from the filesystem view, we are + * ready to tear it out of the hierarchy. Notify the client before + * we do that so they can perform any cleanup that requires + * navigating the hierarchy. A client does not need to provide this + * callback. The subsystem semaphore MUST be held by the caller, and + * references must be valid for both items. It also assumes the + * caller has validated ci_type. + */ +static void client_disconnect_notify(struct config_item *parent_item, + struct config_item *item) +{ + struct config_item_type *type; + + type = parent_item->ci_type; + BUG_ON(!type); + + if (type->ct_group_ops && type->ct_group_ops->disconnect_notify) + type->ct_group_ops->disconnect_notify(to_config_group(parent_item), + item); +} + +/* * Drop the initial reference from make_item()/make_group() * This function assumes that reference is held on item * and that item holds a valid reference to the parent. Also, it @@ -669,33 +926,270 @@ static void client_drop_item(struct config_item *parent_item, type = parent_item->ci_type; BUG_ON(!type); + /* + * If ->drop_item() exists, it is responsible for the + * config_item_put(). + */ if (type->ct_group_ops && type->ct_group_ops->drop_item) type->ct_group_ops->drop_item(to_config_group(parent_item), - item); + item); else config_item_put(item); } +#ifdef DEBUG +static void configfs_dump_one(struct configfs_dirent *sd, int level) +{ + pr_info("%*s\"%s\":\n", level, " ", configfs_get_name(sd)); + +#define type_print(_type) if (sd->s_type & _type) pr_info("%*s %s\n", level, " ", #_type); + type_print(CONFIGFS_ROOT); + type_print(CONFIGFS_DIR); + type_print(CONFIGFS_ITEM_ATTR); + type_print(CONFIGFS_ITEM_LINK); + type_print(CONFIGFS_USET_DIR); + type_print(CONFIGFS_USET_DEFAULT); + type_print(CONFIGFS_USET_DROPPING); +#undef type_print +} + +static int configfs_dump(struct configfs_dirent *sd, int level) +{ + struct configfs_dirent *child_sd; + int ret = 0; + + configfs_dump_one(sd, level); + + if (!(sd->s_type & (CONFIGFS_DIR|CONFIGFS_ROOT))) + return 0; + + list_for_each_entry(child_sd, &sd->s_children, s_sibling) { + ret = configfs_dump(child_sd, level + 2); + if (ret) + break; + } + + return ret; +} +#endif + + +/* + * configfs_depend_item() and configfs_undepend_item() + * + * WARNING: Do not call these from a configfs callback! + * + * This describes these functions and their helpers. + * + * Allow another kernel system to depend on a config_item. If this + * happens, the item cannot go away until the dependent can live without + * it. The idea is to give client modules as simple an interface as + * possible. When a system asks them to depend on an item, they just + * call configfs_depend_item(). If the item is live and the client + * driver is in good shape, we'll happily do the work for them. + * + * Why is the locking complex? Because configfs uses the VFS to handle + * all locking, but this function is called outside the normal + * VFS->configfs path. So it must take VFS locks to prevent the + * VFS->configfs stuff (configfs_mkdir(), configfs_rmdir(), etc). This is + * why you can't call these functions underneath configfs callbacks. + * + * Note, btw, that this can be called at *any* time, even when a configfs + * subsystem isn't registered, or when configfs is loading or unloading. + * Just like configfs_register_subsystem(). So we take the same + * precautions. We pin the filesystem. We lock configfs_dirent_lock. + * If we can find the target item in the + * configfs tree, it must be part of the subsystem tree as well, so we + * do not need the subsystem semaphore. Holding configfs_dirent_lock helps + * locking out mkdir() and rmdir(), who might be racing us. + */ + +/* + * configfs_depend_prep() + * + * Only subdirectories count here. Files (CONFIGFS_NOT_PINNED) are + * attributes. This is similar but not the same to configfs_detach_prep(). + * Note that configfs_detach_prep() expects the parent to be locked when it + * is called, but we lock the parent *inside* configfs_depend_prep(). We + * do that so we can unlock it if we find nothing. + * + * Here we do a depth-first search of the dentry hierarchy looking for + * our object. + * We deliberately ignore items tagged as dropping since they are virtually + * dead, as well as items in the middle of attachment since they virtually + * do not exist yet. This completes the locking out of racing mkdir() and + * rmdir(). + * Note: subdirectories in the middle of attachment start with s_type = + * CONFIGFS_DIR|CONFIGFS_USET_CREATING set by create_dir(). When + * CONFIGFS_USET_CREATING is set, we ignore the item. The actual set of + * s_type is in configfs_new_dirent(), which has configfs_dirent_lock. + * + * If the target is not found, -ENOENT is bubbled up. + * + * This adds a requirement that all config_items be unique! + * + * This is recursive. There isn't + * much on the stack, though, so folks that need this function - be careful + * about your stack! Patches will be accepted to make it iterative. + */ +static int configfs_depend_prep(struct dentry *origin, + struct config_item *target) +{ + struct configfs_dirent *child_sd, *sd; + int ret = 0; + + BUG_ON(!origin || !origin->d_fsdata); + sd = origin->d_fsdata; -static int configfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) + if (sd->s_element == target) /* Boo-yah */ + goto out; + + list_for_each_entry(child_sd, &sd->s_children, s_sibling) { + if ((child_sd->s_type & CONFIGFS_DIR) && + !(child_sd->s_type & CONFIGFS_USET_DROPPING) && + !(child_sd->s_type & CONFIGFS_USET_CREATING)) { + ret = configfs_depend_prep(child_sd->s_dentry, + target); + if (!ret) + goto out; /* Child path boo-yah */ + } + } + + /* We looped all our children and didn't find target */ + ret = -ENOENT; + +out: + return ret; +} + +int configfs_depend_item(struct configfs_subsystem *subsys, + struct config_item *target) { int ret; - struct config_group *group; - struct config_item *item; + struct configfs_dirent *p, *root_sd, *subsys_sd = NULL; + struct config_item *s_item = &subsys->su_group.cg_item; + struct dentry *root; + + /* + * Pin the configfs filesystem. This means we can safely access + * the root of the configfs filesystem. + */ + root = configfs_pin_fs(); + if (IS_ERR(root)) + return PTR_ERR(root); + + /* + * Next, lock the root directory. We're going to check that the + * subsystem is really registered, and so we need to lock out + * configfs_[un]register_subsystem(). + */ + mutex_lock(&root->d_inode->i_mutex); + + root_sd = root->d_fsdata; + + list_for_each_entry(p, &root_sd->s_children, s_sibling) { + if (p->s_type & CONFIGFS_DIR) { + if (p->s_element == s_item) { + subsys_sd = p; + break; + } + } + } + + if (!subsys_sd) { + ret = -ENOENT; + goto out_unlock_fs; + } + + /* Ok, now we can trust subsys/s_item */ + + spin_lock(&configfs_dirent_lock); + /* Scan the tree, return 0 if found */ + ret = configfs_depend_prep(subsys_sd->s_dentry, target); + if (ret) + goto out_unlock_dirent_lock; + + /* + * We are sure that the item is not about to be removed by rmdir(), and + * not in the middle of attachment by mkdir(). + */ + p = target->ci_dentry->d_fsdata; + p->s_dependent_count += 1; + +out_unlock_dirent_lock: + spin_unlock(&configfs_dirent_lock); +out_unlock_fs: + mutex_unlock(&root->d_inode->i_mutex); + + /* + * If we succeeded, the fs is pinned via other methods. If not, + * we're done with it anyway. So release_fs() is always right. + */ + configfs_release_fs(); + + return ret; +} +EXPORT_SYMBOL(configfs_depend_item); + +/* + * Release the dependent linkage. This is much simpler than + * configfs_depend_item() because we know that that the client driver is + * pinned, thus the subsystem is pinned, and therefore configfs is pinned. + */ +void configfs_undepend_item(struct configfs_subsystem *subsys, + struct config_item *target) +{ + struct configfs_dirent *sd; + + /* + * Since we can trust everything is pinned, we just need + * configfs_dirent_lock. + */ + spin_lock(&configfs_dirent_lock); + + sd = target->ci_dentry->d_fsdata; + BUG_ON(sd->s_dependent_count < 1); + + sd->s_dependent_count -= 1; + + /* + * After this unlock, we cannot trust the item to stay alive! + * DO NOT REFERENCE item after this unlock. + */ + spin_unlock(&configfs_dirent_lock); +} +EXPORT_SYMBOL(configfs_undepend_item); + +static int configfs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) +{ + int ret = 0; + int module_got = 0; + struct config_group *group = NULL; + struct config_item *item = NULL; struct config_item *parent_item; struct configfs_subsystem *subsys; struct configfs_dirent *sd; struct config_item_type *type; - struct module *owner; + struct module *subsys_owner = NULL, *new_item_owner = NULL; char *name; - if (dentry->d_parent == configfs_sb->s_root) - return -EPERM; - sd = dentry->d_parent->d_fsdata; - if (!(sd->s_type & CONFIGFS_USET_DIR)) - return -EPERM; + /* + * Fake invisibility if dir belongs to a group/default groups hierarchy + * being attached + */ + if (!configfs_dirent_is_ready(sd)) { + ret = -ENOENT; + goto out; + } + + if (!(sd->s_type & CONFIGFS_USET_DIR)) { + ret = -EPERM; + goto out; + } + + /* Get a working ref for the duration of this function */ parent_item = configfs_get_config_item(dentry->d_parent); type = parent_item->ci_type; subsys = to_config_group(parent_item)->cg_subsys; @@ -704,69 +1198,140 @@ static int configfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) if (!type || !type->ct_group_ops || (!type->ct_group_ops->make_group && !type->ct_group_ops->make_item)) { - config_item_put(parent_item); - return -EPERM; /* What lack-of-mkdir returns */ + ret = -EPERM; /* Lack-of-mkdir returns -EPERM */ + goto out_put; + } + + /* + * The subsystem may belong to a different module than the item + * being created. We don't want to safely pin the new item but + * fail to pin the subsystem it sits under. + */ + if (!subsys->su_group.cg_item.ci_type) { + ret = -EINVAL; + goto out_put; + } + subsys_owner = subsys->su_group.cg_item.ci_type->ct_owner; + if (!try_module_get(subsys_owner)) { + ret = -EINVAL; + goto out_put; } name = kmalloc(dentry->d_name.len + 1, GFP_KERNEL); if (!name) { - config_item_put(parent_item); - return -ENOMEM; + ret = -ENOMEM; + goto out_subsys_put; } + snprintf(name, dentry->d_name.len + 1, "%s", dentry->d_name.name); - down(&subsys->su_sem); - group = NULL; - item = NULL; + mutex_lock(&subsys->su_mutex); if (type->ct_group_ops->make_group) { group = type->ct_group_ops->make_group(to_config_group(parent_item), name); - if (group) { + if (!group) + group = ERR_PTR(-ENOMEM); + if (!IS_ERR(group)) { link_group(to_config_group(parent_item), group); item = &group->cg_item; - } + } else + ret = PTR_ERR(group); } else { item = type->ct_group_ops->make_item(to_config_group(parent_item), name); - if (item) + if (!item) + item = ERR_PTR(-ENOMEM); + if (!IS_ERR(item)) link_obj(parent_item, item); + else + ret = PTR_ERR(item); } - up(&subsys->su_sem); + mutex_unlock(&subsys->su_mutex); kfree(name); - if (!item) { - config_item_put(parent_item); - return -ENOMEM; + if (ret) { + /* + * If ret != 0, then link_obj() was never called. + * There are no extra references to clean up. + */ + goto out_subsys_put; } - ret = -EINVAL; + /* + * link_obj() has been called (via link_group() for groups). + * From here on out, errors must clean that up. + */ + type = item->ci_type; - if (type) { - owner = type->ct_owner; - if (try_module_get(owner)) { - if (group) { - ret = configfs_attach_group(parent_item, - item, - dentry); - } else { - ret = configfs_attach_item(parent_item, - item, - dentry); - } + if (!type) { + ret = -EINVAL; + goto out_unlink; + } - if (ret) { - down(&subsys->su_sem); - if (group) - unlink_group(group); - else - unlink_obj(item); - client_drop_item(parent_item, item); - up(&subsys->su_sem); + new_item_owner = type->ct_owner; + if (!try_module_get(new_item_owner)) { + ret = -EINVAL; + goto out_unlink; + } - config_item_put(parent_item); - module_put(owner); - } - } + /* + * I hate doing it this way, but if there is + * an error, module_put() probably should + * happen after any cleanup. + */ + module_got = 1; + + /* + * Make racing rmdir() fail if it did not tag parent with + * CONFIGFS_USET_DROPPING + * Note: if CONFIGFS_USET_DROPPING is already set, attach_group() will + * fail and let rmdir() terminate correctly + */ + spin_lock(&configfs_dirent_lock); + /* This will make configfs_detach_prep() fail */ + sd->s_type |= CONFIGFS_USET_IN_MKDIR; + spin_unlock(&configfs_dirent_lock); + + if (group) + ret = configfs_attach_group(parent_item, item, dentry); + else + ret = configfs_attach_item(parent_item, item, dentry); + + spin_lock(&configfs_dirent_lock); + sd->s_type &= ~CONFIGFS_USET_IN_MKDIR; + if (!ret) + configfs_dir_set_ready(dentry->d_fsdata); + spin_unlock(&configfs_dirent_lock); + +out_unlink: + if (ret) { + /* Tear down everything we built up */ + mutex_lock(&subsys->su_mutex); + + client_disconnect_notify(parent_item, item); + if (group) + unlink_group(group); + else + unlink_obj(item); + client_drop_item(parent_item, item); + + mutex_unlock(&subsys->su_mutex); + + if (module_got) + module_put(new_item_owner); } +out_subsys_put: + if (ret) + module_put(subsys_owner); + +out_put: + /* + * link_obj()/link_group() took a reference from child->parent, + * so the parent is safely pinned. We can drop our working + * reference. + */ + config_item_put(parent_item); + +out: return ret; } @@ -776,16 +1341,14 @@ static int configfs_rmdir(struct inode *dir, struct dentry *dentry) struct config_item *item; struct configfs_subsystem *subsys; struct configfs_dirent *sd; - struct module *owner = NULL; + struct module *subsys_owner = NULL, *dead_item_owner = NULL; int ret; - if (dentry->d_parent == configfs_sb->s_root) - return -EPERM; - sd = dentry->d_fsdata; if (sd->s_type & CONFIGFS_USET_DEFAULT) return -EPERM; + /* Get a working ref until we have the child */ parent_item = configfs_get_config_item(dentry->d_parent); subsys = to_config_group(parent_item)->cg_subsys; BUG_ON(!subsys); @@ -795,50 +1358,92 @@ static int configfs_rmdir(struct inode *dir, struct dentry *dentry) return -EINVAL; } - ret = configfs_detach_prep(dentry); - if (ret) { - configfs_detach_rollback(dentry); - config_item_put(parent_item); - return ret; - } + /* configfs_mkdir() shouldn't have allowed this */ + BUG_ON(!subsys->su_group.cg_item.ci_type); + subsys_owner = subsys->su_group.cg_item.ci_type->ct_owner; + + /* + * Ensure that no racing symlink() will make detach_prep() fail while + * the new link is temporarily attached + */ + do { + struct mutex *wait_mutex; + mutex_lock(&configfs_symlink_mutex); + spin_lock(&configfs_dirent_lock); + /* + * Here's where we check for dependents. We're protected by + * configfs_dirent_lock. + * If no dependent, atomically tag the item as dropping. + */ + ret = sd->s_dependent_count ? -EBUSY : 0; + if (!ret) { + ret = configfs_detach_prep(dentry, &wait_mutex); + if (ret) + configfs_detach_rollback(dentry); + } + spin_unlock(&configfs_dirent_lock); + mutex_unlock(&configfs_symlink_mutex); + + if (ret) { + if (ret != -EAGAIN) { + config_item_put(parent_item); + return ret; + } + + /* Wait until the racing operation terminates */ + mutex_lock(wait_mutex); + mutex_unlock(wait_mutex); + } + } while (ret == -EAGAIN); + + /* Get a working ref for the duration of this function */ item = configfs_get_config_item(dentry); /* Drop reference from above, item already holds one. */ config_item_put(parent_item); if (item->ci_type) - owner = item->ci_type->ct_owner; + dead_item_owner = item->ci_type->ct_owner; if (sd->s_type & CONFIGFS_USET_DIR) { configfs_detach_group(item); - down(&subsys->su_sem); + mutex_lock(&subsys->su_mutex); + client_disconnect_notify(parent_item, item); unlink_group(to_config_group(item)); } else { configfs_detach_item(item); - down(&subsys->su_sem); + mutex_lock(&subsys->su_mutex); + client_disconnect_notify(parent_item, item); unlink_obj(item); } client_drop_item(parent_item, item); - up(&subsys->su_sem); + mutex_unlock(&subsys->su_mutex); /* Drop our reference from above */ config_item_put(item); - module_put(owner); + module_put(dead_item_owner); + module_put(subsys_owner); return 0; } -struct inode_operations configfs_dir_inode_operations = { +const struct inode_operations configfs_dir_inode_operations = { .mkdir = configfs_mkdir, .rmdir = configfs_rmdir, .symlink = configfs_symlink, .unlink = configfs_unlink, .lookup = configfs_lookup, + .setattr = configfs_setattr, +}; + +const struct inode_operations configfs_root_inode_operations = { + .lookup = configfs_lookup, + .setattr = configfs_setattr, }; #if 0 @@ -860,7 +1465,7 @@ int configfs_rename_dir(struct config_item * item, const char *new_name) new_dentry = lookup_one_len(new_name, parent, strlen(new_name)); if (!IS_ERR(new_dentry)) { - if (!new_dentry->d_inode) { + if (!new_dentry->d_inode) { error = config_item_set_name(item, "%s", new_name); if (!error) { d_add(new_dentry, NULL); @@ -881,24 +1486,37 @@ int configfs_rename_dir(struct config_item * item, const char *new_name) static int configfs_dir_open(struct inode *inode, struct file *file) { - struct dentry * dentry = file->f_dentry; + struct dentry * dentry = file->f_path.dentry; struct configfs_dirent * parent_sd = dentry->d_fsdata; + int err; mutex_lock(&dentry->d_inode->i_mutex); - file->private_data = configfs_new_dirent(parent_sd, NULL); + /* + * Fake invisibility if dir belongs to a group/default groups hierarchy + * being attached + */ + err = -ENOENT; + if (configfs_dirent_is_ready(parent_sd)) { + file->private_data = configfs_new_dirent(parent_sd, NULL, 0); + if (IS_ERR(file->private_data)) + err = PTR_ERR(file->private_data); + else + err = 0; + } mutex_unlock(&dentry->d_inode->i_mutex); - return file->private_data ? 0 : -ENOMEM; - + return err; } static int configfs_dir_close(struct inode *inode, struct file *file) { - struct dentry * dentry = file->f_dentry; + struct dentry * dentry = file->f_path.dentry; struct configfs_dirent * cursor = file->private_data; mutex_lock(&dentry->d_inode->i_mutex); + spin_lock(&configfs_dirent_lock); list_del_init(&cursor->s_sibling); + spin_unlock(&configfs_dirent_lock); mutex_unlock(&dentry->d_inode->i_mutex); release_configfs_dirent(cursor); @@ -912,78 +1530,83 @@ static inline unsigned char dt_type(struct configfs_dirent *sd) return (sd->s_mode >> 12) & 15; } -static int configfs_readdir(struct file * filp, void * dirent, filldir_t filldir) +static int configfs_readdir(struct file *file, struct dir_context *ctx) { - struct dentry *dentry = filp->f_dentry; + struct dentry *dentry = file->f_path.dentry; + struct super_block *sb = dentry->d_sb; struct configfs_dirent * parent_sd = dentry->d_fsdata; - struct configfs_dirent *cursor = filp->private_data; + struct configfs_dirent *cursor = file->private_data; struct list_head *p, *q = &cursor->s_sibling; - ino_t ino; - int i = filp->f_pos; + ino_t ino = 0; + + if (!dir_emit_dots(file, ctx)) + return 0; + if (ctx->pos == 2) { + spin_lock(&configfs_dirent_lock); + list_move(q, &parent_sd->s_children); + spin_unlock(&configfs_dirent_lock); + } + for (p = q->next; p != &parent_sd->s_children; p = p->next) { + struct configfs_dirent *next; + const char *name; + int len; + struct inode *inode = NULL; + + next = list_entry(p, struct configfs_dirent, s_sibling); + if (!next->s_element) + continue; - switch (i) { - case 0: - ino = dentry->d_inode->i_ino; - if (filldir(dirent, ".", 1, i, ino, DT_DIR) < 0) - break; - filp->f_pos++; - i++; - /* fallthrough */ - case 1: - ino = parent_ino(dentry); - if (filldir(dirent, "..", 2, i, ino, DT_DIR) < 0) - break; - filp->f_pos++; - i++; - /* fallthrough */ - default: - if (filp->f_pos == 2) { - list_del(q); - list_add(q, &parent_sd->s_children); - } - for (p=q->next; p!= &parent_sd->s_children; p=p->next) { - struct configfs_dirent *next; - const char * name; - int len; + name = configfs_get_name(next); + len = strlen(name); - next = list_entry(p, struct configfs_dirent, - s_sibling); - if (!next->s_element) - continue; - - name = configfs_get_name(next); - len = strlen(name); - if (next->s_dentry) - ino = next->s_dentry->d_inode->i_ino; - else - ino = iunique(configfs_sb, 2); - - if (filldir(dirent, name, len, filp->f_pos, ino, - dt_type(next)) < 0) - return 0; - - list_del(q); - list_add(q, p); - p = q; - filp->f_pos++; - } + /* + * We'll have a dentry and an inode for + * PINNED items and for open attribute + * files. We lock here to prevent a race + * with configfs_d_iput() clearing + * s_dentry before calling iput(). + * + * Why do we go to the trouble? If + * someone has an attribute file open, + * the inode number should match until + * they close it. Beyond that, we don't + * care. + */ + spin_lock(&configfs_dirent_lock); + dentry = next->s_dentry; + if (dentry) + inode = dentry->d_inode; + if (inode) + ino = inode->i_ino; + spin_unlock(&configfs_dirent_lock); + if (!inode) + ino = iunique(sb, 2); + + if (!dir_emit(ctx, name, len, ino, dt_type(next))) + return 0; + + spin_lock(&configfs_dirent_lock); + list_move(q, p); + spin_unlock(&configfs_dirent_lock); + p = q; + ctx->pos++; } return 0; } -static loff_t configfs_dir_lseek(struct file * file, loff_t offset, int origin) +static loff_t configfs_dir_lseek(struct file *file, loff_t offset, int whence) { - struct dentry * dentry = file->f_dentry; + struct dentry * dentry = file->f_path.dentry; mutex_lock(&dentry->d_inode->i_mutex); - switch (origin) { + switch (whence) { case 1: offset += file->f_pos; case 0: if (offset >= 0) break; default: - mutex_unlock(&file->f_dentry->d_inode->i_mutex); + mutex_unlock(&file_inode(file)->i_mutex); return -EINVAL; } if (offset != file->f_pos) { @@ -994,6 +1617,7 @@ static loff_t configfs_dir_lseek(struct file * file, loff_t offset, int origin) struct list_head *p; loff_t n = file->f_pos - 2; + spin_lock(&configfs_dirent_lock); list_del(&cursor->s_sibling); p = sd->s_children.next; while (n && p != &sd->s_children) { @@ -1005,65 +1629,64 @@ static loff_t configfs_dir_lseek(struct file * file, loff_t offset, int origin) p = p->next; } list_add_tail(&cursor->s_sibling, p); + spin_unlock(&configfs_dirent_lock); } } mutex_unlock(&dentry->d_inode->i_mutex); return offset; } -struct file_operations configfs_dir_operations = { +const struct file_operations configfs_dir_operations = { .open = configfs_dir_open, .release = configfs_dir_close, .llseek = configfs_dir_lseek, .read = generic_read_dir, - .readdir = configfs_readdir, + .iterate = configfs_readdir, }; int configfs_register_subsystem(struct configfs_subsystem *subsys) { int err; struct config_group *group = &subsys->su_group; - struct qstr name; struct dentry *dentry; + struct dentry *root; struct configfs_dirent *sd; - err = configfs_pin_fs(); - if (err) - return err; + root = configfs_pin_fs(); + if (IS_ERR(root)) + return PTR_ERR(root); if (!group->cg_item.ci_name) group->cg_item.ci_name = group->cg_item.ci_namebuf; - sd = configfs_sb->s_root->d_fsdata; + sd = root->d_fsdata; link_group(to_config_group(sd->s_element), group); - mutex_lock(&configfs_sb->s_root->d_inode->i_mutex); - - name.name = group->cg_item.ci_name; - name.len = strlen(name.name); - name.hash = full_name_hash(name.name, name.len); + mutex_lock_nested(&root->d_inode->i_mutex, I_MUTEX_PARENT); err = -ENOMEM; - dentry = d_alloc(configfs_sb->s_root, &name); - if (!dentry) - goto out_release; - - d_add(dentry, NULL); - - err = configfs_attach_group(sd->s_element, &group->cg_item, - dentry); - if (!err) - dentry = NULL; - else - d_delete(dentry); + dentry = d_alloc_name(root, group->cg_item.ci_name); + if (dentry) { + d_add(dentry, NULL); + + err = configfs_attach_group(sd->s_element, &group->cg_item, + dentry); + if (err) { + BUG_ON(dentry->d_inode); + d_drop(dentry); + dput(dentry); + } else { + spin_lock(&configfs_dirent_lock); + configfs_dir_set_ready(dentry->d_fsdata); + spin_unlock(&configfs_dirent_lock); + } + } - mutex_unlock(&configfs_sb->s_root->d_inode->i_mutex); + mutex_unlock(&root->d_inode->i_mutex); - if (dentry) { - dput(dentry); -out_release: - unlink_group(group); - configfs_release_fs(); + if (err) { + unlink_group(group); + configfs_release_fs(); } return err; @@ -1073,24 +1696,31 @@ void configfs_unregister_subsystem(struct configfs_subsystem *subsys) { struct config_group *group = &subsys->su_group; struct dentry *dentry = group->cg_item.ci_dentry; + struct dentry *root = dentry->d_sb->s_root; - if (dentry->d_parent != configfs_sb->s_root) { - printk(KERN_ERR "configfs: Tried to unregister non-subsystem!\n"); + if (dentry->d_parent != root) { + pr_err("Tried to unregister non-subsystem!\n"); return; } - mutex_lock(&configfs_sb->s_root->d_inode->i_mutex); - mutex_lock(&dentry->d_inode->i_mutex); - if (configfs_detach_prep(dentry)) { - printk(KERN_ERR "configfs: Tried to unregister non-empty subsystem!\n"); + mutex_lock_nested(&root->d_inode->i_mutex, + I_MUTEX_PARENT); + mutex_lock_nested(&dentry->d_inode->i_mutex, I_MUTEX_CHILD); + mutex_lock(&configfs_symlink_mutex); + spin_lock(&configfs_dirent_lock); + if (configfs_detach_prep(dentry, NULL)) { + pr_err("Tried to unregister non-empty subsystem!\n"); } + spin_unlock(&configfs_dirent_lock); + mutex_unlock(&configfs_symlink_mutex); configfs_detach_group(&group->cg_item); dentry->d_inode->i_flags |= S_DEAD; + dont_mount(dentry); mutex_unlock(&dentry->d_inode->i_mutex); d_delete(dentry); - mutex_unlock(&configfs_sb->s_root->d_inode->i_mutex); + mutex_unlock(&root->d_inode->i_mutex); dput(dentry); diff --git a/fs/configfs/file.c b/fs/configfs/file.c index c26cd61f13a..1d1c41f1014 100644 --- a/fs/configfs/file.c +++ b/fs/configfs/file.c @@ -26,21 +26,27 @@ #include <linux/fs.h> #include <linux/module.h> -#include <linux/dnotify.h> #include <linux/slab.h> +#include <linux/mutex.h> #include <asm/uaccess.h> -#include <asm/semaphore.h> #include <linux/configfs.h> #include "configfs_internal.h" +/* + * A simple attribute can only be 4096 characters. Why 4k? Because the + * original code limited it to PAGE_SIZE. That's a bad idea, though, + * because an attribute of 16k on ia64 won't work on x86. So we limit to + * 4k, our minimum common page size. + */ +#define SIMPLE_ATTR_SIZE 4096 struct configfs_buffer { size_t count; loff_t pos; char * page; struct configfs_item_operations * ops; - struct semaphore sem; + struct mutex mutex; int needs_read_fill; }; @@ -70,7 +76,7 @@ static int fill_read_buffer(struct dentry * dentry, struct configfs_buffer * buf count = ops->show_attribute(item,attr,buffer->page); buffer->needs_read_fill = 0; - BUG_ON(count > (ssize_t)PAGE_SIZE); + BUG_ON(count > (ssize_t)SIMPLE_ATTR_SIZE); if (count >= 0) buffer->count = count; else @@ -78,36 +84,6 @@ static int fill_read_buffer(struct dentry * dentry, struct configfs_buffer * buf return ret; } - -/** - * flush_read_buffer - push buffer to userspace. - * @buffer: data buffer for file. - * @userbuf: user-passed buffer. - * @count: number of bytes requested. - * @ppos: file position. - * - * Copy the buffer we filled in fill_read_buffer() to userspace. - * This is done at the reader's leisure, copying and advancing - * the amount they specify each time. - * This may be called continuously until the buffer is empty. - */ -static int flush_read_buffer(struct configfs_buffer * buffer, char __user * buf, - size_t count, loff_t * ppos) -{ - int error; - - if (*ppos > buffer->count) - return 0; - - if (count > (buffer->count - *ppos)) - count = buffer->count - *ppos; - - error = copy_to_user(buf,buffer->page + *ppos,count); - if (!error) - *ppos += count; - return error ? -EFAULT : count; -} - /** * configfs_read_file - read an attribute. * @file: file pointer. @@ -133,16 +109,17 @@ configfs_read_file(struct file *file, char __user *buf, size_t count, loff_t *pp struct configfs_buffer * buffer = file->private_data; ssize_t retval = 0; - down(&buffer->sem); + mutex_lock(&buffer->mutex); if (buffer->needs_read_fill) { - if ((retval = fill_read_buffer(file->f_dentry,buffer))) + if ((retval = fill_read_buffer(file->f_path.dentry,buffer))) goto out; } - pr_debug("%s: count = %d, ppos = %lld, buf = %s\n", - __FUNCTION__,count,*ppos,buffer->page); - retval = flush_read_buffer(buffer,buf,count,ppos); + pr_debug("%s: count = %zd, ppos = %lld, buf = %s\n", + __func__, count, *ppos, buffer->page); + retval = simple_read_from_buffer(buf, count, ppos, buffer->page, + buffer->count); out: - up(&buffer->sem); + mutex_unlock(&buffer->mutex); return retval; } @@ -150,7 +127,7 @@ out: /** * fill_write_buffer - copy buffer from userspace. * @buffer: data buffer for file. - * @userbuf: data from user. + * @buf: data from user. * @count: number of bytes in @userbuf. * * Allocate @buffer->page if it hasn't been already, then @@ -163,22 +140,26 @@ fill_write_buffer(struct configfs_buffer * buffer, const char __user * buf, size int error; if (!buffer->page) - buffer->page = (char *)get_zeroed_page(GFP_KERNEL); + buffer->page = (char *)__get_free_pages(GFP_KERNEL, 0); if (!buffer->page) return -ENOMEM; - if (count > PAGE_SIZE) - count = PAGE_SIZE; + if (count >= SIMPLE_ATTR_SIZE) + count = SIMPLE_ATTR_SIZE - 1; error = copy_from_user(buffer->page,buf,count); buffer->needs_read_fill = 1; + /* if buf is assumed to contain a string, terminate it by \0, + * so e.g. sscanf() can scan the string easily */ + buffer->page[count] = 0; return error ? -EFAULT : count; } /** * flush_write_buffer - push buffer to config_item. - * @file: file pointer. + * @dentry: dentry to the attribute * @buffer: data buffer for file. + * @count: number of bytes * * Get the correct pointers for the config_item and the attribute we're * dealing with, then call the store() method for the attribute, @@ -217,21 +198,22 @@ static ssize_t configfs_write_file(struct file *file, const char __user *buf, size_t count, loff_t *ppos) { struct configfs_buffer * buffer = file->private_data; - - down(&buffer->sem); - count = fill_write_buffer(buffer,buf,count); - if (count > 0) - count = flush_write_buffer(file->f_dentry,buffer,count); - if (count > 0) - *ppos += count; - up(&buffer->sem); - return count; + ssize_t len; + + mutex_lock(&buffer->mutex); + len = fill_write_buffer(buffer, buf, count); + if (len > 0) + len = flush_write_buffer(file->f_path.dentry, buffer, len); + if (len > 0) + *ppos += len; + mutex_unlock(&buffer->mutex); + return len; } static int check_perm(struct inode * inode, struct file * file) { - struct config_item *item = configfs_get_config_item(file->f_dentry->d_parent); - struct configfs_attribute * attr = to_attr(file->f_dentry); + struct config_item *item = configfs_get_config_item(file->f_path.dentry->d_parent); + struct configfs_attribute * attr = to_attr(file->f_path.dentry); struct configfs_buffer * buffer; struct configfs_item_operations * ops = NULL; int error = 0; @@ -273,15 +255,15 @@ static int check_perm(struct inode * inode, struct file * file) /* No error? Great, allocate a buffer for the file, and store it * it in file->private_data for easy access. */ - buffer = kmalloc(sizeof(struct configfs_buffer),GFP_KERNEL); - if (buffer) { - memset(buffer,0,sizeof(struct configfs_buffer)); - init_MUTEX(&buffer->sem); - buffer->needs_read_fill = 1; - buffer->ops = ops; - file->private_data = buffer; - } else + buffer = kzalloc(sizeof(struct configfs_buffer),GFP_KERNEL); + if (!buffer) { error = -ENOMEM; + goto Enomem; + } + mutex_init(&buffer->mutex); + buffer->needs_read_fill = 1; + buffer->ops = ops; + file->private_data = buffer; goto Done; Einval: @@ -289,6 +271,7 @@ static int check_perm(struct inode * inode, struct file * file) goto Done; Eaccess: error = -EACCES; + Enomem: module_put(attr->ca_owner); Done: if (error && item) @@ -303,8 +286,8 @@ static int configfs_open_file(struct inode * inode, struct file * filp) static int configfs_release(struct inode * inode, struct file * filp) { - struct config_item * item = to_item(filp->f_dentry->d_parent); - struct configfs_attribute * attr = to_attr(filp->f_dentry); + struct config_item * item = to_item(filp->f_path.dentry->d_parent); + struct configfs_attribute * attr = to_attr(filp->f_path.dentry); struct module * owner = attr->ca_owner; struct configfs_buffer * buffer = filp->private_data; @@ -316,12 +299,13 @@ static int configfs_release(struct inode * inode, struct file * filp) if (buffer) { if (buffer->page) free_page((unsigned long)buffer->page); + mutex_destroy(&buffer->mutex); kfree(buffer); } return 0; } -struct file_operations configfs_file_operations = { +const struct file_operations configfs_file_operations = { .read = configfs_read_file, .write = configfs_write_file, .llseek = generic_file_llseek, @@ -336,7 +320,7 @@ int configfs_add_file(struct dentry * dir, const struct configfs_attribute * att umode_t mode = (attr->ca_mode & S_IALLUGO) | S_IFREG; int error = 0; - mutex_lock(&dir->d_inode->i_mutex); + mutex_lock_nested(&dir->d_inode->i_mutex, I_MUTEX_NORMAL); error = configfs_make_dirent(parent_sd, NULL, (void *) attr, mode, type); mutex_unlock(&dir->d_inode->i_mutex); diff --git a/fs/configfs/inode.c b/fs/configfs/inode.c index 6577c588de9..5946ad98053 100644 --- a/fs/configfs/inode.c +++ b/fs/configfs/inode.c @@ -23,7 +23,8 @@ * * configfs Copyright (C) 2005 Oracle. All rights reserved. * - * Please see Documentation/filesystems/configfs.txt for more information. + * Please see Documentation/filesystems/configfs/configfs.txt for more + * information. */ #undef DEBUG @@ -31,70 +32,189 @@ #include <linux/pagemap.h> #include <linux/namei.h> #include <linux/backing-dev.h> +#include <linux/capability.h> +#include <linux/sched.h> +#include <linux/lockdep.h> +#include <linux/slab.h> #include <linux/configfs.h> #include "configfs_internal.h" -extern struct super_block * configfs_sb; +#ifdef CONFIG_LOCKDEP +static struct lock_class_key default_group_class[MAX_LOCK_DEPTH]; +#endif -static struct address_space_operations configfs_aops = { +static const struct address_space_operations configfs_aops = { .readpage = simple_readpage, - .prepare_write = simple_prepare_write, - .commit_write = simple_commit_write + .write_begin = simple_write_begin, + .write_end = simple_write_end, }; static struct backing_dev_info configfs_backing_dev_info = { + .name = "configfs", .ra_pages = 0, /* No readahead */ - .capabilities = BDI_CAP_NO_ACCT_DIRTY | BDI_CAP_NO_WRITEBACK, + .capabilities = BDI_CAP_NO_ACCT_AND_WRITEBACK, }; -struct inode * configfs_new_inode(mode_t mode) +static const struct inode_operations configfs_inode_operations ={ + .setattr = configfs_setattr, +}; + +int configfs_setattr(struct dentry * dentry, struct iattr * iattr) +{ + struct inode * inode = dentry->d_inode; + struct configfs_dirent * sd = dentry->d_fsdata; + struct iattr * sd_iattr; + unsigned int ia_valid = iattr->ia_valid; + int error; + + if (!sd) + return -EINVAL; + + sd_iattr = sd->s_iattr; + if (!sd_iattr) { + /* setting attributes for the first time, allocate now */ + sd_iattr = kzalloc(sizeof(struct iattr), GFP_KERNEL); + if (!sd_iattr) + return -ENOMEM; + /* assign default attributes */ + sd_iattr->ia_mode = sd->s_mode; + sd_iattr->ia_uid = GLOBAL_ROOT_UID; + sd_iattr->ia_gid = GLOBAL_ROOT_GID; + sd_iattr->ia_atime = sd_iattr->ia_mtime = sd_iattr->ia_ctime = CURRENT_TIME; + sd->s_iattr = sd_iattr; + } + /* attributes were changed atleast once in past */ + + error = simple_setattr(dentry, iattr); + if (error) + return error; + + if (ia_valid & ATTR_UID) + sd_iattr->ia_uid = iattr->ia_uid; + if (ia_valid & ATTR_GID) + sd_iattr->ia_gid = iattr->ia_gid; + if (ia_valid & ATTR_ATIME) + sd_iattr->ia_atime = timespec_trunc(iattr->ia_atime, + inode->i_sb->s_time_gran); + if (ia_valid & ATTR_MTIME) + sd_iattr->ia_mtime = timespec_trunc(iattr->ia_mtime, + inode->i_sb->s_time_gran); + if (ia_valid & ATTR_CTIME) + sd_iattr->ia_ctime = timespec_trunc(iattr->ia_ctime, + inode->i_sb->s_time_gran); + if (ia_valid & ATTR_MODE) { + umode_t mode = iattr->ia_mode; + + if (!in_group_p(inode->i_gid) && !capable(CAP_FSETID)) + mode &= ~S_ISGID; + sd_iattr->ia_mode = sd->s_mode = mode; + } + + return error; +} + +static inline void set_default_inode_attr(struct inode * inode, umode_t mode) +{ + inode->i_mode = mode; + inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; +} + +static inline void set_inode_attr(struct inode * inode, struct iattr * iattr) +{ + inode->i_mode = iattr->ia_mode; + inode->i_uid = iattr->ia_uid; + inode->i_gid = iattr->ia_gid; + inode->i_atime = iattr->ia_atime; + inode->i_mtime = iattr->ia_mtime; + inode->i_ctime = iattr->ia_ctime; +} + +struct inode *configfs_new_inode(umode_t mode, struct configfs_dirent *sd, + struct super_block *s) { - struct inode * inode = new_inode(configfs_sb); + struct inode * inode = new_inode(s); if (inode) { - inode->i_mode = mode; - inode->i_uid = 0; - inode->i_gid = 0; - inode->i_blksize = PAGE_CACHE_SIZE; - inode->i_blocks = 0; - inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; + inode->i_ino = get_next_ino(); inode->i_mapping->a_ops = &configfs_aops; inode->i_mapping->backing_dev_info = &configfs_backing_dev_info; + inode->i_op = &configfs_inode_operations; + + if (sd->s_iattr) { + /* sysfs_dirent has non-default attributes + * get them for the new inode from persistent copy + * in sysfs_dirent + */ + set_inode_attr(inode, sd->s_iattr); + } else + set_default_inode_attr(inode, mode); } return inode; } -int configfs_create(struct dentry * dentry, int mode, int (*init)(struct inode *)) +#ifdef CONFIG_LOCKDEP + +static void configfs_set_inode_lock_class(struct configfs_dirent *sd, + struct inode *inode) +{ + int depth = sd->s_depth; + + if (depth > 0) { + if (depth <= ARRAY_SIZE(default_group_class)) { + lockdep_set_class(&inode->i_mutex, + &default_group_class[depth - 1]); + } else { + /* + * In practice the maximum level of locking depth is + * already reached. Just inform about possible reasons. + */ + pr_info("Too many levels of inodes for the locking correctness validator.\n"); + pr_info("Spurious warnings may appear.\n"); + } + } +} + +#else /* CONFIG_LOCKDEP */ + +static void configfs_set_inode_lock_class(struct configfs_dirent *sd, + struct inode *inode) +{ +} + +#endif /* CONFIG_LOCKDEP */ + +int configfs_create(struct dentry * dentry, umode_t mode, int (*init)(struct inode *)) { int error = 0; - struct inode * inode = NULL; - if (dentry) { - if (!dentry->d_inode) { - if ((inode = configfs_new_inode(mode))) { - if (dentry->d_parent && dentry->d_parent->d_inode) { - struct inode *p_inode = dentry->d_parent->d_inode; - p_inode->i_mtime = p_inode->i_ctime = CURRENT_TIME; - } - goto Proceed; - } - else - error = -ENOMEM; - } else - error = -EEXIST; - } else - error = -ENOENT; - goto Done; + struct inode *inode = NULL; + struct configfs_dirent *sd; + struct inode *p_inode; - Proceed: - if (init) + if (!dentry) + return -ENOENT; + + if (dentry->d_inode) + return -EEXIST; + + sd = dentry->d_fsdata; + inode = configfs_new_inode(mode, sd, dentry->d_sb); + if (!inode) + return -ENOMEM; + + p_inode = dentry->d_parent->d_inode; + p_inode->i_mtime = p_inode->i_ctime = CURRENT_TIME; + configfs_set_inode_lock_class(sd, inode); + + if (init) { error = init(inode); - if (!error) { - d_instantiate(dentry, inode); - if (S_ISDIR(mode) || S_ISLNK(mode)) - dget(dentry); /* pin link and directory dentries in core */ - } else - iput(inode); - Done: + if (error) { + iput(inode); + return error; + } + } + d_instantiate(dentry, inode); + if (S_ISDIR(mode) || S_ISLNK(mode)) + dget(dentry); /* pin link and directory dentries in core */ return error; } @@ -103,10 +223,9 @@ int configfs_create(struct dentry * dentry, int mode, int (*init)(struct inode * */ const unsigned char * configfs_get_name(struct configfs_dirent *sd) { - struct attribute * attr; + struct configfs_attribute *attr; - if (!sd || !sd->s_element) - BUG(); + BUG_ON(!sd || !sd->s_element); /* These always have a dentry, so use that */ if (sd->s_type & (CONFIGFS_DIR | CONFIGFS_ITEM_LINK)) @@ -114,7 +233,7 @@ const unsigned char * configfs_get_name(struct configfs_dirent *sd) if (sd->s_type & CONFIGFS_ITEM_ATTR) { attr = sd->s_element; - return attr->name; + return attr->ca_name; } return NULL; } @@ -129,14 +248,14 @@ void configfs_drop_dentry(struct configfs_dirent * sd, struct dentry * parent) struct dentry * dentry = sd->s_dentry; if (dentry) { - spin_lock(&dcache_lock); + spin_lock(&dentry->d_lock); if (!(d_unhashed(dentry) && dentry->d_inode)) { - dget_locked(dentry); + dget_dlock(dentry); __d_drop(dentry); - spin_unlock(&dcache_lock); + spin_unlock(&dentry->d_lock); simple_unlink(parent->d_inode, dentry); } else - spin_unlock(&dcache_lock); + spin_unlock(&dentry->d_lock); } } @@ -145,12 +264,18 @@ void configfs_hash_and_remove(struct dentry * dir, const char * name) struct configfs_dirent * sd; struct configfs_dirent * parent_sd = dir->d_fsdata; + if (dir->d_inode == NULL) + /* no inode means this hasn't been made visible yet */ + return; + mutex_lock(&dir->d_inode->i_mutex); list_for_each_entry(sd, &parent_sd->s_children, s_sibling) { if (!sd->s_element) continue; if (!strcmp(configfs_get_name(sd), name)) { + spin_lock(&configfs_dirent_lock); list_del_init(&sd->s_sibling); + spin_unlock(&configfs_dirent_lock); configfs_drop_dentry(sd, dir); configfs_put(sd); break; @@ -159,4 +284,12 @@ void configfs_hash_and_remove(struct dentry * dir, const char * name) mutex_unlock(&dir->d_inode->i_mutex); } +int __init configfs_inode_init(void) +{ + return bdi_init(&configfs_backing_dev_info); +} +void configfs_inode_exit(void) +{ + bdi_destroy(&configfs_backing_dev_info); +} diff --git a/fs/configfs/item.c b/fs/configfs/item.c index e07485ac50a..e65f9ffbb99 100644 --- a/fs/configfs/item.c +++ b/fs/configfs/item.c @@ -19,11 +19,11 @@ * Boston, MA 021110-1307, USA. * * Based on kobject: - * kobject is Copyright (c) 2002-2003 Patrick Mochel + * kobject is Copyright (c) 2002-2003 Patrick Mochel * * configfs Copyright (C) 2005 Oracle. All rights reserved. * - * Please see the file Documentation/filesystems/configfs.txt for + * Please see the file Documentation/filesystems/configfs/configfs.txt for * critical information about using the config_item interface. */ @@ -35,9 +35,9 @@ #include <linux/configfs.h> -static inline struct config_item * to_item(struct list_head * entry) +static inline struct config_item *to_item(struct list_head *entry) { - return container_of(entry,struct config_item,ci_entry); + return container_of(entry, struct config_item, ci_entry); } /* Evil kernel */ @@ -47,35 +47,35 @@ static void config_item_release(struct kref *kref); * config_item_init - initialize item. * @item: item in question. */ -void config_item_init(struct config_item * item) +void config_item_init(struct config_item *item) { kref_init(&item->ci_kref); INIT_LIST_HEAD(&item->ci_entry); } +EXPORT_SYMBOL(config_item_init); /** * config_item_set_name - Set the name of an item * @item: item. - * @name: name. + * @fmt: The vsnprintf()'s format string. * * If strlen(name) >= CONFIGFS_ITEM_NAME_LEN, then use a * dynamically allocated string that @item->ci_name points to. * Otherwise, use the static @item->ci_namebuf array. */ - -int config_item_set_name(struct config_item * item, const char * fmt, ...) +int config_item_set_name(struct config_item *item, const char *fmt, ...) { int error = 0; int limit = CONFIGFS_ITEM_NAME_LEN; int need; va_list args; - char * name; + char *name; /* * First, try the static array */ - va_start(args,fmt); - need = vsnprintf(item->ci_namebuf,limit,fmt,args); + va_start(args, fmt); + need = vsnprintf(item->ci_namebuf, limit, fmt, args); va_end(args); if (need < limit) name = item->ci_namebuf; @@ -84,13 +84,13 @@ int config_item_set_name(struct config_item * item, const char * fmt, ...) * Need more space? Allocate it and try again */ limit = need + 1; - name = kmalloc(limit,GFP_KERNEL); + name = kmalloc(limit, GFP_KERNEL); if (!name) { error = -ENOMEM; goto Done; } - va_start(args,fmt); - need = vsnprintf(name,limit,fmt,args); + va_start(args, fmt); + need = vsnprintf(name, limit, fmt, args); va_end(args); /* Still? Give up. */ @@ -110,7 +110,6 @@ int config_item_set_name(struct config_item * item, const char * fmt, ...) Done: return error; } - EXPORT_SYMBOL(config_item_set_name); void config_item_init_type_name(struct config_item *item, @@ -132,25 +131,21 @@ void config_group_init_type_name(struct config_group *group, const char *name, } EXPORT_SYMBOL(config_group_init_type_name); -struct config_item * config_item_get(struct config_item * item) +struct config_item *config_item_get(struct config_item *item) { if (item) kref_get(&item->ci_kref); return item; } +EXPORT_SYMBOL(config_item_get); -/** - * config_item_cleanup - free config_item resources. - * @item: item. - */ - -void config_item_cleanup(struct config_item * item) +static void config_item_cleanup(struct config_item *item) { - struct config_item_type * t = item->ci_type; - struct config_group * s = item->ci_group; - struct config_item * parent = item->ci_parent; + struct config_item_type *t = item->ci_type; + struct config_group *s = item->ci_group; + struct config_item *parent = item->ci_parent; - pr_debug("config_item %s: cleaning up\n",config_item_name(item)); + pr_debug("config_item %s: cleaning up\n", config_item_name(item)); if (item->ci_name != item->ci_namebuf) kfree(item->ci_name); item->ci_name = NULL; @@ -173,55 +168,47 @@ static void config_item_release(struct kref *kref) * * Decrement the refcount, and if 0, call config_item_cleanup(). */ -void config_item_put(struct config_item * item) +void config_item_put(struct config_item *item) { if (item) kref_put(&item->ci_kref, config_item_release); } - +EXPORT_SYMBOL(config_item_put); /** * config_group_init - initialize a group for use - * @k: group + * @group: config_group */ - void config_group_init(struct config_group *group) { config_item_init(&group->cg_item); INIT_LIST_HEAD(&group->cg_children); } - +EXPORT_SYMBOL(config_group_init); /** - * config_group_find_obj - search for item in group. + * config_group_find_item - search for item in group. * @group: group we're looking in. * @name: item's name. * - * Lock group via @group->cg_subsys, and iterate over @group->cg_list, - * looking for a matching config_item. If matching item is found - * take a reference and return the item. + * Iterate over @group->cg_list, looking for a matching config_item. + * If matching item is found take a reference and return the item. + * Caller must have locked group via @group->cg_subsys->su_mtx. */ - -struct config_item * config_group_find_obj(struct config_group * group, const char * name) +struct config_item *config_group_find_item(struct config_group *group, + const char *name) { - struct list_head * entry; - struct config_item * ret = NULL; + struct list_head *entry; + struct config_item *ret = NULL; - /* XXX LOCKING! */ - list_for_each(entry,&group->cg_children) { - struct config_item * item = to_item(entry); + list_for_each(entry, &group->cg_children) { + struct config_item *item = to_item(entry); if (config_item_name(item) && - !strcmp(config_item_name(item), name)) { + !strcmp(config_item_name(item), name)) { ret = config_item_get(item); break; } } return ret; } - - -EXPORT_SYMBOL(config_item_init); -EXPORT_SYMBOL(config_group_init); -EXPORT_SYMBOL(config_item_get); -EXPORT_SYMBOL(config_item_put); - +EXPORT_SYMBOL(config_group_find_item); diff --git a/fs/configfs/mount.c b/fs/configfs/mount.c index 1a2f6f6a4d9..f6c28583339 100644 --- a/fs/configfs/mount.c +++ b/fs/configfs/mount.c @@ -29,6 +29,7 @@ #include <linux/mount.h> #include <linux/pagemap.h> #include <linux/init.h> +#include <linux/slab.h> #include <linux/configfs.h> #include "configfs_internal.h" @@ -36,11 +37,11 @@ /* Random magic number */ #define CONFIGFS_MAGIC 0x62656570 -struct vfsmount * configfs_mount = NULL; -struct super_block * configfs_sb = NULL; +static struct vfsmount *configfs_mount = NULL; +struct kmem_cache *configfs_dir_cachep; static int configfs_mnt_count = 0; -static struct super_operations configfs_ops = { +static const struct super_operations configfs_ops = { .statfs = simple_statfs, .drop_inode = generic_delete_inode, }; @@ -62,6 +63,7 @@ static struct configfs_dirent configfs_root = { .s_children = LIST_HEAD_INIT(configfs_root.s_children), .s_element = &configfs_root_group.cg_item, .s_type = CONFIGFS_ROOT, + .s_iattr = NULL, }; static int configfs_fill_super(struct super_block *sb, void *data, int silent) @@ -73,49 +75,52 @@ static int configfs_fill_super(struct super_block *sb, void *data, int silent) sb->s_blocksize_bits = PAGE_CACHE_SHIFT; sb->s_magic = CONFIGFS_MAGIC; sb->s_op = &configfs_ops; - configfs_sb = sb; + sb->s_time_gran = 1; - inode = configfs_new_inode(S_IFDIR | S_IRWXU | S_IRUGO | S_IXUGO); + inode = configfs_new_inode(S_IFDIR | S_IRWXU | S_IRUGO | S_IXUGO, + &configfs_root, sb); if (inode) { - inode->i_op = &configfs_dir_inode_operations; + inode->i_op = &configfs_root_inode_operations; inode->i_fop = &configfs_dir_operations; /* directory inodes start off with i_nlink == 2 (for "." entry) */ - inode->i_nlink++; + inc_nlink(inode); } else { - pr_debug("configfs: could not get root inode\n"); + pr_debug("could not get root inode\n"); return -ENOMEM; } - root = d_alloc_root(inode); + root = d_make_root(inode); if (!root) { - pr_debug("%s: could not get root dentry!\n",__FUNCTION__); - iput(inode); + pr_debug("%s: could not get root dentry!\n",__func__); return -ENOMEM; } config_group_init(&configfs_root_group); configfs_root_group.cg_item.ci_dentry = root; root->d_fsdata = &configfs_root; sb->s_root = root; + sb->s_d_op = &configfs_dentry_ops; /* the rest get that */ return 0; } -static struct super_block *configfs_get_sb(struct file_system_type *fs_type, +static struct dentry *configfs_do_mount(struct file_system_type *fs_type, int flags, const char *dev_name, void *data) { - return get_sb_single(fs_type, flags, data, configfs_fill_super); + return mount_single(fs_type, flags, data, configfs_fill_super); } static struct file_system_type configfs_fs_type = { .owner = THIS_MODULE, .name = "configfs", - .get_sb = configfs_get_sb, + .mount = configfs_do_mount, .kill_sb = kill_litter_super, }; +MODULE_ALIAS_FS("configfs"); -int configfs_pin_fs(void) +struct dentry *configfs_pin_fs(void) { - return simple_pin_fs("configfs", &configfs_mount, + int err = simple_pin_fs(&configfs_fs_type, &configfs_mount, &configfs_mnt_count); + return err ? ERR_PTR(err) : configfs_mount->mnt_root; } void configfs_release_fs(void) @@ -124,35 +129,55 @@ void configfs_release_fs(void) } -static decl_subsys(config, NULL, NULL); +static struct kobject *config_kobj; static int __init configfs_init(void) { - int err; + int err = -ENOMEM; - kset_set_kset_s(&config_subsys, kernel_subsys); - err = subsystem_register(&config_subsys); + configfs_dir_cachep = kmem_cache_create("configfs_dir_cache", + sizeof(struct configfs_dirent), + 0, 0, NULL); + if (!configfs_dir_cachep) + goto out; + + config_kobj = kobject_create_and_add("config", kernel_kobj); + if (!config_kobj) + goto out2; + + err = configfs_inode_init(); if (err) - return err; + goto out3; err = register_filesystem(&configfs_fs_type); - if (err) { - printk(KERN_ERR "configfs: Unable to register filesystem!\n"); - subsystem_unregister(&config_subsys); - } + if (err) + goto out4; + return 0; +out4: + pr_err("Unable to register filesystem!\n"); + configfs_inode_exit(); +out3: + kobject_put(config_kobj); +out2: + kmem_cache_destroy(configfs_dir_cachep); + configfs_dir_cachep = NULL; +out: return err; } static void __exit configfs_exit(void) { unregister_filesystem(&configfs_fs_type); - subsystem_unregister(&config_subsys); + kobject_put(config_kobj); + kmem_cache_destroy(configfs_dir_cachep); + configfs_dir_cachep = NULL; + configfs_inode_exit(); } MODULE_AUTHOR("Oracle"); MODULE_LICENSE("GPL"); -MODULE_VERSION("0.0.1"); +MODULE_VERSION("0.0.2"); MODULE_DESCRIPTION("Simple RAM filesystem for user driven kernel subsystem configuration."); module_init(configfs_init); diff --git a/fs/configfs/symlink.c b/fs/configfs/symlink.c index 50f5840521a..cc9f2546ea4 100644 --- a/fs/configfs/symlink.c +++ b/fs/configfs/symlink.c @@ -27,10 +27,14 @@ #include <linux/fs.h> #include <linux/module.h> #include <linux/namei.h> +#include <linux/slab.h> #include <linux/configfs.h> #include "configfs_internal.h" +/* Protects attachments of new symlinks */ +DEFINE_MUTEX(configfs_symlink_mutex); + static int item_depth(struct config_item * item) { struct config_item * p = item; @@ -66,47 +70,62 @@ static void fill_item_path(struct config_item * item, char * buffer, int length) } static int create_link(struct config_item *parent_item, - struct config_item *item, + struct config_item *item, struct dentry *dentry) { struct configfs_dirent *target_sd = item->ci_dentry->d_fsdata; struct configfs_symlink *sl; int ret; + ret = -ENOENT; + if (!configfs_dirent_is_ready(target_sd)) + goto out; ret = -ENOMEM; sl = kmalloc(sizeof(struct configfs_symlink), GFP_KERNEL); if (sl) { sl->sl_target = config_item_get(item); - /* FIXME: needs a lock, I'd bet */ + spin_lock(&configfs_dirent_lock); + if (target_sd->s_type & CONFIGFS_USET_DROPPING) { + spin_unlock(&configfs_dirent_lock); + config_item_put(item); + kfree(sl); + return -ENOENT; + } list_add(&sl->sl_list, &target_sd->s_links); + spin_unlock(&configfs_dirent_lock); ret = configfs_create_link(sl, parent_item->ci_dentry, dentry); if (ret) { + spin_lock(&configfs_dirent_lock); list_del_init(&sl->sl_list); + spin_unlock(&configfs_dirent_lock); config_item_put(item); kfree(sl); } } +out: return ret; } -static int get_target(const char *symname, struct nameidata *nd, - struct config_item **target) +static int get_target(const char *symname, struct path *path, + struct config_item **target, struct super_block *sb) { int ret; - ret = path_lookup(symname, LOOKUP_FOLLOW|LOOKUP_DIRECTORY, nd); + ret = kern_path(symname, LOOKUP_FOLLOW|LOOKUP_DIRECTORY, path); if (!ret) { - if (nd->dentry->d_sb == configfs_sb) { - *target = configfs_get_config_item(nd->dentry); + if (path->dentry->d_sb == sb) { + *target = configfs_get_config_item(path->dentry); if (!*target) { ret = -ENOENT; - path_release(nd); + path_put(path); } - } else + } else { ret = -EPERM; + path_put(path); + } } return ret; @@ -116,32 +135,45 @@ static int get_target(const char *symname, struct nameidata *nd, int configfs_symlink(struct inode *dir, struct dentry *dentry, const char *symname) { int ret; - struct nameidata nd; + struct path path; + struct configfs_dirent *sd; struct config_item *parent_item; - struct config_item *target_item; + struct config_item *target_item = NULL; struct config_item_type *type; - ret = -EPERM; /* What lack-of-symlink returns */ - if (dentry->d_parent == configfs_sb->s_root) + sd = dentry->d_parent->d_fsdata; + /* + * Fake invisibility if dir belongs to a group/default groups hierarchy + * being attached + */ + ret = -ENOENT; + if (!configfs_dirent_is_ready(sd)) goto out; parent_item = configfs_get_config_item(dentry->d_parent); type = parent_item->ci_type; + ret = -EPERM; if (!type || !type->ct_item_ops || !type->ct_item_ops->allow_link) goto out_put; - ret = get_target(symname, &nd, &target_item); + ret = get_target(symname, &path, &target_item, dentry->d_sb); if (ret) goto out_put; ret = type->ct_item_ops->allow_link(parent_item, target_item); - if (!ret) + if (!ret) { + mutex_lock(&configfs_symlink_mutex); ret = create_link(parent_item, target_item, dentry); + mutex_unlock(&configfs_symlink_mutex); + if (ret && type->ct_item_ops->drop_link) + type->ct_item_ops->drop_link(parent_item, + target_item); + } config_item_put(target_item); - path_release(&nd); + path_put(&path); out_put: config_item_put(parent_item); @@ -162,15 +194,14 @@ int configfs_unlink(struct inode *dir, struct dentry *dentry) if (!(sd->s_type & CONFIGFS_ITEM_LINK)) goto out; - if (dentry->d_parent == configfs_sb->s_root) - BUG(); - sl = sd->s_element; parent_item = configfs_get_config_item(dentry->d_parent); type = parent_item->ci_type; + spin_lock(&configfs_dirent_lock); list_del_init(&sd->s_sibling); + spin_unlock(&configfs_dirent_lock); configfs_drop_dentry(sd, dentry->d_parent); dput(dentry); configfs_put(sd); @@ -185,8 +216,9 @@ int configfs_unlink(struct inode *dir, struct dentry *dentry) type->ct_item_ops->drop_link(parent_item, sl->sl_target); - /* FIXME: Needs lock */ + spin_lock(&configfs_dirent_lock); list_del_init(&sl->sl_list); + spin_unlock(&configfs_dirent_lock); /* Put reference from create_link() */ config_item_put(sl->sl_target); @@ -211,13 +243,13 @@ static int configfs_get_target_path(struct config_item * item, struct config_ite if (size > PATH_MAX) return -ENAMETOOLONG; - pr_debug("%s: depth = %d, size = %d\n", __FUNCTION__, depth, size); + pr_debug("%s: depth = %d, size = %d\n", __func__, depth, size); for (s = path; depth--; s += 3) strcpy(s,"../"); fill_item_path(target, path, size); - pr_debug("%s: path = '%s'\n", __FUNCTION__, path); + pr_debug("%s: path = '%s'\n", __func__, path); return 0; } @@ -273,9 +305,10 @@ static void configfs_put_link(struct dentry *dentry, struct nameidata *nd, } } -struct inode_operations configfs_symlink_inode_operations = { +const struct inode_operations configfs_symlink_inode_operations = { .follow_link = configfs_follow_link, .readlink = generic_readlink, .put_link = configfs_put_link, + .setattr = configfs_setattr, }; |