1
linux/fs/sysfs/dir.c

794 lines
18 KiB
C
Raw Normal View History

/*
* dir.c - Operations for sysfs directories.
*/
#undef DEBUG
#include <linux/fs.h>
#include <linux/mount.h>
#include <linux/module.h>
#include <linux/kobject.h>
#include <linux/namei.h>
#include <linux/idr.h>
#include <asm/semaphore.h>
#include "sysfs.h"
DECLARE_RWSEM(sysfs_rename_sem);
spinlock_t sysfs_lock = SPIN_LOCK_UNLOCKED;
spinlock_t kobj_sysfs_assoc_lock = SPIN_LOCK_UNLOCKED;
static spinlock_t sysfs_ino_lock = SPIN_LOCK_UNLOCKED;
static DEFINE_IDA(sysfs_ino_ida);
int sysfs_alloc_ino(ino_t *pino)
{
int ino, rc;
retry:
spin_lock(&sysfs_ino_lock);
rc = ida_get_new_above(&sysfs_ino_ida, 2, &ino);
spin_unlock(&sysfs_ino_lock);
if (rc == -EAGAIN) {
if (ida_pre_get(&sysfs_ino_ida, GFP_KERNEL))
goto retry;
rc = -ENOMEM;
}
*pino = ino;
return rc;
}
static void sysfs_free_ino(ino_t ino)
{
spin_lock(&sysfs_ino_lock);
ida_remove(&sysfs_ino_ida, ino);
spin_unlock(&sysfs_ino_lock);
}
void release_sysfs_dirent(struct sysfs_dirent * sd)
{
struct sysfs_dirent *parent_sd;
repeat:
parent_sd = sd->s_parent;
sysfs: implement sysfs_dirent active reference and immediate disconnect sysfs: implement sysfs_dirent active reference and immediate disconnect Opening a sysfs node references its associated kobject, so userland can arbitrarily prolong lifetime of a kobject which complicates lifetime rules in drivers. This patch implements active reference and makes the association between kobject and sysfs immediately breakable. Now each sysfs_dirent has two reference counts - s_count and s_active. s_count is a regular reference count which guarantees that the containing sysfs_dirent is accessible. As long as s_count reference is held, all sysfs internal fields in sysfs_dirent are accessible including s_parent and s_name. The newly added s_active is active reference count. This is acquired by invoking sysfs_get_active() and it's the caller's responsibility to ensure sysfs_dirent itself is accessible (should be holding s_count one way or the other). Dereferencing sysfs_dirent to access objects out of sysfs proper requires active reference. This includes access to the associated kobjects, attributes and ops. The active references can be drained and denied by calling sysfs_deactivate(). All active sysfs_dirents must be deactivated after deletion but before the default reference is dropped. This enables immediate disconnect of sysfs nodes. Once a sysfs_dirent is deleted, it won't access any entity external to sysfs proper. Because attr/bin_attr ops access both the node itself and its parent for kobject, they need to hold active references to both. sysfs_get/put_active_two() helpers are provided to help grabbing both references. Parent's is acquired first and released last. Unlike other operations, mmapped area lingers on after mmap() is finished and the module implement implementing it and kobj need to stay referenced till all the mapped pages are gone. This is accomplished by holding one set of active references to the bin_attr and its parent if there have been any mmap during lifetime of an openfile. The references are dropped when the openfile is released. This change makes sysfs lifetime rules independent from both kobject's and module's. It not only fixes several race conditions caused by sysfs not holding onto the proper module when referencing kobject, but also helps fixing and simplifying lifetime management in driver model and drivers by taking sysfs out of the equation. Please read the following message for more info. http://article.gmane.org/gmane.linux.kernel/510293 Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2007-06-13 11:45:16 -07:00
/* If @sd is being released after deletion, s_active is write
* locked. If @sd is cursor for directory walk or being
* released prematurely, s_active has no reader or writer.
*
* sysfs_deactivate() lies to lockdep that s_active is
* unlocked immediately. Lie one more time to cover the
* previous lie.
*/
if (!down_write_trylock(&sd->s_active))
rwsem_acquire(&sd->s_active.dep_map,
SYSFS_S_ACTIVE_DEACTIVATE, 0, _RET_IP_);
up_write(&sd->s_active);
if (sd->s_type & SYSFS_KOBJ_LINK)
sysfs_put(sd->s_elem.symlink.target_sd);
if (sd->s_type & SYSFS_COPY_NAME)
kfree(sd->s_name);
kfree(sd->s_iattr);
sysfs_free_ino(sd->s_ino);
kmem_cache_free(sysfs_dir_cachep, sd);
sd = parent_sd;
if (sd && atomic_dec_and_test(&sd->s_count))
goto repeat;
}
static void sysfs_d_iput(struct dentry * dentry, struct inode * inode)
{
struct sysfs_dirent * sd = dentry->d_fsdata;
if (sd) {
/* sd->s_dentry is protected with sysfs_lock. This
* allows sysfs_drop_dentry() to dereference it.
*/
spin_lock(&sysfs_lock);
/* The dentry might have been deleted or another
* lookup could have happened updating sd->s_dentry to
* point the new dentry. Ignore if it isn't pointing
* to this dentry.
*/
if (sd->s_dentry == dentry)
sd->s_dentry = NULL;
spin_unlock(&sysfs_lock);
sysfs_put(sd);
}
iput(inode);
}
static struct dentry_operations sysfs_dentry_ops = {
.d_iput = sysfs_d_iput,
};
struct sysfs_dirent *sysfs_new_dirent(const char *name, umode_t mode, int type)
{
char *dup_name = NULL;
struct sysfs_dirent *sd = NULL;
if (type & SYSFS_COPY_NAME) {
name = dup_name = kstrdup(name, GFP_KERNEL);
if (!name)
goto err_out;
}
sd = kmem_cache_zalloc(sysfs_dir_cachep, GFP_KERNEL);
if (!sd)
goto err_out;
if (sysfs_alloc_ino(&sd->s_ino))
goto err_out;
atomic_set(&sd->s_count, 1);
atomic_set(&sd->s_event, 1);
sysfs: implement sysfs_dirent active reference and immediate disconnect sysfs: implement sysfs_dirent active reference and immediate disconnect Opening a sysfs node references its associated kobject, so userland can arbitrarily prolong lifetime of a kobject which complicates lifetime rules in drivers. This patch implements active reference and makes the association between kobject and sysfs immediately breakable. Now each sysfs_dirent has two reference counts - s_count and s_active. s_count is a regular reference count which guarantees that the containing sysfs_dirent is accessible. As long as s_count reference is held, all sysfs internal fields in sysfs_dirent are accessible including s_parent and s_name. The newly added s_active is active reference count. This is acquired by invoking sysfs_get_active() and it's the caller's responsibility to ensure sysfs_dirent itself is accessible (should be holding s_count one way or the other). Dereferencing sysfs_dirent to access objects out of sysfs proper requires active reference. This includes access to the associated kobjects, attributes and ops. The active references can be drained and denied by calling sysfs_deactivate(). All active sysfs_dirents must be deactivated after deletion but before the default reference is dropped. This enables immediate disconnect of sysfs nodes. Once a sysfs_dirent is deleted, it won't access any entity external to sysfs proper. Because attr/bin_attr ops access both the node itself and its parent for kobject, they need to hold active references to both. sysfs_get/put_active_two() helpers are provided to help grabbing both references. Parent's is acquired first and released last. Unlike other operations, mmapped area lingers on after mmap() is finished and the module implement implementing it and kobj need to stay referenced till all the mapped pages are gone. This is accomplished by holding one set of active references to the bin_attr and its parent if there have been any mmap during lifetime of an openfile. The references are dropped when the openfile is released. This change makes sysfs lifetime rules independent from both kobject's and module's. It not only fixes several race conditions caused by sysfs not holding onto the proper module when referencing kobject, but also helps fixing and simplifying lifetime management in driver model and drivers by taking sysfs out of the equation. Please read the following message for more info. http://article.gmane.org/gmane.linux.kernel/510293 Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2007-06-13 11:45:16 -07:00
init_rwsem(&sd->s_active);
INIT_LIST_HEAD(&sd->s_children);
INIT_LIST_HEAD(&sd->s_sibling);
sd->s_name = name;
sd->s_mode = mode;
sd->s_type = type;
return sd;
err_out:
kfree(dup_name);
kmem_cache_free(sysfs_dir_cachep, sd);
return NULL;
}
static void sysfs_attach_dentry(struct sysfs_dirent *sd, struct dentry *dentry)
{
dentry->d_op = &sysfs_dentry_ops;
dentry->d_fsdata = sysfs_get(sd);
/* protect sd->s_dentry against sysfs_d_iput */
spin_lock(&sysfs_lock);
sd->s_dentry = dentry;
spin_unlock(&sysfs_lock);
d_rehash(dentry);
}
void sysfs_attach_dirent(struct sysfs_dirent *sd,
struct sysfs_dirent *parent_sd, struct dentry *dentry)
{
if (dentry)
sysfs_attach_dentry(sd, dentry);
if (parent_sd) {
sd->s_parent = sysfs_get(parent_sd);
list_add(&sd->s_sibling, &parent_sd->s_children);
}
}
/*
*
* Return -EEXIST if there is already a sysfs element with the same name for
* the same parent.
*
* called with parent inode's i_mutex held
*/
int sysfs_dirent_exist(struct sysfs_dirent *parent_sd,
const unsigned char *new)
{
struct sysfs_dirent * sd;
list_for_each_entry(sd, &parent_sd->s_children, s_sibling) {
if (sd->s_type) {
if (strcmp(sd->s_name, new))
continue;
else
return -EEXIST;
}
}
return 0;
}
static int init_dir(struct inode * inode)
{
inode->i_op = &sysfs_dir_inode_operations;
inode->i_fop = &sysfs_dir_operations;
/* directory inodes start off with i_nlink == 2 (for "." entry) */
inc_nlink(inode);
return 0;
}
static int init_file(struct inode * inode)
{
inode->i_size = PAGE_SIZE;
inode->i_fop = &sysfs_file_operations;
return 0;
}
static int init_symlink(struct inode * inode)
{
inode->i_op = &sysfs_symlink_inode_operations;
return 0;
}
static int create_dir(struct kobject *kobj, struct dentry *parent,
const char *name, struct dentry **p_dentry)
{
int error;
umode_t mode = S_IFDIR| S_IRWXU | S_IRUGO | S_IXUGO;
struct dentry *dentry;
struct sysfs_dirent *sd;
mutex_lock(&parent->d_inode->i_mutex);
dentry = lookup_one_len(name, parent, strlen(name));
if (IS_ERR(dentry)) {
error = PTR_ERR(dentry);
goto out_unlock;
}
error = -EEXIST;
if (sysfs_dirent_exist(parent->d_fsdata, name))
goto out_dput;
error = -ENOMEM;
sd = sysfs_new_dirent(name, mode, SYSFS_DIR);
if (!sd)
goto out_drop;
sd->s_elem.dir.kobj = kobj;
error = sysfs_create(sd, dentry, mode, init_dir);
if (error)
goto out_sput;
inc_nlink(parent->d_inode);
sysfs_attach_dirent(sd, parent->d_fsdata, dentry);
*p_dentry = dentry;
error = 0;
goto out_dput;
out_sput:
list_del_init(&sd->s_sibling);
sysfs_put(sd);
out_drop:
d_drop(dentry);
out_dput:
dput(dentry);
out_unlock:
mutex_unlock(&parent->d_inode->i_mutex);
return error;
}
int sysfs_create_subdir(struct kobject * k, const char * n, struct dentry ** d)
{
return create_dir(k,k->dentry,n,d);
}
/**
* sysfs_create_dir - create a directory for an object.
* @kobj: object we're creating directory for.
* @shadow_parent: parent parent object.
*/
int sysfs_create_dir(struct kobject * kobj, struct dentry *shadow_parent)
{
struct dentry * dentry = NULL;
struct dentry * parent;
int error = 0;
BUG_ON(!kobj);
if (shadow_parent)
parent = shadow_parent;
else if (kobj->parent)
parent = kobj->parent->dentry;
else if (sysfs_mount && sysfs_mount->mnt_sb)
parent = sysfs_mount->mnt_sb->s_root;
else
return -EFAULT;
error = create_dir(kobj,parent,kobject_name(kobj),&dentry);
if (!error)
kobj->dentry = dentry;
return error;
}
/* attaches attribute's sysfs_dirent to the dentry corresponding to the
* attribute file
*/
static int sysfs_attach_attr(struct sysfs_dirent * sd, struct dentry * dentry)
{
struct attribute * attr = NULL;
struct bin_attribute * bin_attr = NULL;
int (* init) (struct inode *) = NULL;
int error = 0;
if (sd->s_type & SYSFS_KOBJ_BIN_ATTR) {
bin_attr = sd->s_elem.bin_attr.bin_attr;
attr = &bin_attr->attr;
} else {
attr = sd->s_elem.attr.attr;
init = init_file;
}
error = sysfs_create(sd, dentry,
(attr->mode & S_IALLUGO) | S_IFREG, init);
if (error)
return error;
if (bin_attr) {
dentry->d_inode->i_size = bin_attr->size;
dentry->d_inode->i_fop = &bin_fops;
}
sysfs_attach_dentry(sd, dentry);
return 0;
}
static int sysfs_attach_link(struct sysfs_dirent * sd, struct dentry * dentry)
{
int err;
err = sysfs_create(sd, dentry, S_IFLNK|S_IRWXUGO, init_symlink);
if (!err)
sysfs_attach_dentry(sd, dentry);
return err;
}
static struct dentry * sysfs_lookup(struct inode *dir, struct dentry *dentry,
struct nameidata *nd)
{
struct sysfs_dirent * parent_sd = dentry->d_parent->d_fsdata;
struct sysfs_dirent * sd;
int err = 0;
list_for_each_entry(sd, &parent_sd->s_children, s_sibling) {
if (sd->s_type & SYSFS_NOT_PINNED) {
if (strcmp(sd->s_name, dentry->d_name.name))
continue;
if (sd->s_type & SYSFS_KOBJ_LINK)
err = sysfs_attach_link(sd, dentry);
else
err = sysfs_attach_attr(sd, dentry);
break;
}
}
return ERR_PTR(err);
}
const struct inode_operations sysfs_dir_inode_operations = {
.lookup = sysfs_lookup,
.setattr = sysfs_setattr,
};
static void remove_dir(struct dentry * d)
{
struct dentry *parent = d->d_parent;
struct sysfs_dirent *sd = d->d_fsdata;
mutex_lock(&parent->d_inode->i_mutex);
list_del_init(&sd->s_sibling);
pr_debug(" o %s removing done (%d)\n",d->d_name.name,
atomic_read(&d->d_count));
mutex_unlock(&parent->d_inode->i_mutex);
sysfs: implement sysfs_dirent active reference and immediate disconnect sysfs: implement sysfs_dirent active reference and immediate disconnect Opening a sysfs node references its associated kobject, so userland can arbitrarily prolong lifetime of a kobject which complicates lifetime rules in drivers. This patch implements active reference and makes the association between kobject and sysfs immediately breakable. Now each sysfs_dirent has two reference counts - s_count and s_active. s_count is a regular reference count which guarantees that the containing sysfs_dirent is accessible. As long as s_count reference is held, all sysfs internal fields in sysfs_dirent are accessible including s_parent and s_name. The newly added s_active is active reference count. This is acquired by invoking sysfs_get_active() and it's the caller's responsibility to ensure sysfs_dirent itself is accessible (should be holding s_count one way or the other). Dereferencing sysfs_dirent to access objects out of sysfs proper requires active reference. This includes access to the associated kobjects, attributes and ops. The active references can be drained and denied by calling sysfs_deactivate(). All active sysfs_dirents must be deactivated after deletion but before the default reference is dropped. This enables immediate disconnect of sysfs nodes. Once a sysfs_dirent is deleted, it won't access any entity external to sysfs proper. Because attr/bin_attr ops access both the node itself and its parent for kobject, they need to hold active references to both. sysfs_get/put_active_two() helpers are provided to help grabbing both references. Parent's is acquired first and released last. Unlike other operations, mmapped area lingers on after mmap() is finished and the module implement implementing it and kobj need to stay referenced till all the mapped pages are gone. This is accomplished by holding one set of active references to the bin_attr and its parent if there have been any mmap during lifetime of an openfile. The references are dropped when the openfile is released. This change makes sysfs lifetime rules independent from both kobject's and module's. It not only fixes several race conditions caused by sysfs not holding onto the proper module when referencing kobject, but also helps fixing and simplifying lifetime management in driver model and drivers by taking sysfs out of the equation. Please read the following message for more info. http://article.gmane.org/gmane.linux.kernel/510293 Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2007-06-13 11:45:16 -07:00
sysfs_drop_dentry(sd);
sysfs: implement sysfs_dirent active reference and immediate disconnect sysfs: implement sysfs_dirent active reference and immediate disconnect Opening a sysfs node references its associated kobject, so userland can arbitrarily prolong lifetime of a kobject which complicates lifetime rules in drivers. This patch implements active reference and makes the association between kobject and sysfs immediately breakable. Now each sysfs_dirent has two reference counts - s_count and s_active. s_count is a regular reference count which guarantees that the containing sysfs_dirent is accessible. As long as s_count reference is held, all sysfs internal fields in sysfs_dirent are accessible including s_parent and s_name. The newly added s_active is active reference count. This is acquired by invoking sysfs_get_active() and it's the caller's responsibility to ensure sysfs_dirent itself is accessible (should be holding s_count one way or the other). Dereferencing sysfs_dirent to access objects out of sysfs proper requires active reference. This includes access to the associated kobjects, attributes and ops. The active references can be drained and denied by calling sysfs_deactivate(). All active sysfs_dirents must be deactivated after deletion but before the default reference is dropped. This enables immediate disconnect of sysfs nodes. Once a sysfs_dirent is deleted, it won't access any entity external to sysfs proper. Because attr/bin_attr ops access both the node itself and its parent for kobject, they need to hold active references to both. sysfs_get/put_active_two() helpers are provided to help grabbing both references. Parent's is acquired first and released last. Unlike other operations, mmapped area lingers on after mmap() is finished and the module implement implementing it and kobj need to stay referenced till all the mapped pages are gone. This is accomplished by holding one set of active references to the bin_attr and its parent if there have been any mmap during lifetime of an openfile. The references are dropped when the openfile is released. This change makes sysfs lifetime rules independent from both kobject's and module's. It not only fixes several race conditions caused by sysfs not holding onto the proper module when referencing kobject, but also helps fixing and simplifying lifetime management in driver model and drivers by taking sysfs out of the equation. Please read the following message for more info. http://article.gmane.org/gmane.linux.kernel/510293 Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2007-06-13 11:45:16 -07:00
sysfs_deactivate(sd);
sysfs_put(sd);
}
void sysfs_remove_subdir(struct dentry * d)
{
remove_dir(d);
}
static void __sysfs_remove_dir(struct dentry *dentry)
{
sysfs: implement sysfs_dirent active reference and immediate disconnect sysfs: implement sysfs_dirent active reference and immediate disconnect Opening a sysfs node references its associated kobject, so userland can arbitrarily prolong lifetime of a kobject which complicates lifetime rules in drivers. This patch implements active reference and makes the association between kobject and sysfs immediately breakable. Now each sysfs_dirent has two reference counts - s_count and s_active. s_count is a regular reference count which guarantees that the containing sysfs_dirent is accessible. As long as s_count reference is held, all sysfs internal fields in sysfs_dirent are accessible including s_parent and s_name. The newly added s_active is active reference count. This is acquired by invoking sysfs_get_active() and it's the caller's responsibility to ensure sysfs_dirent itself is accessible (should be holding s_count one way or the other). Dereferencing sysfs_dirent to access objects out of sysfs proper requires active reference. This includes access to the associated kobjects, attributes and ops. The active references can be drained and denied by calling sysfs_deactivate(). All active sysfs_dirents must be deactivated after deletion but before the default reference is dropped. This enables immediate disconnect of sysfs nodes. Once a sysfs_dirent is deleted, it won't access any entity external to sysfs proper. Because attr/bin_attr ops access both the node itself and its parent for kobject, they need to hold active references to both. sysfs_get/put_active_two() helpers are provided to help grabbing both references. Parent's is acquired first and released last. Unlike other operations, mmapped area lingers on after mmap() is finished and the module implement implementing it and kobj need to stay referenced till all the mapped pages are gone. This is accomplished by holding one set of active references to the bin_attr and its parent if there have been any mmap during lifetime of an openfile. The references are dropped when the openfile is released. This change makes sysfs lifetime rules independent from both kobject's and module's. It not only fixes several race conditions caused by sysfs not holding onto the proper module when referencing kobject, but also helps fixing and simplifying lifetime management in driver model and drivers by taking sysfs out of the equation. Please read the following message for more info. http://article.gmane.org/gmane.linux.kernel/510293 Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2007-06-13 11:45:16 -07:00
LIST_HEAD(removed);
struct sysfs_dirent * parent_sd;
struct sysfs_dirent * sd, * tmp;
if (!dentry)
return;
pr_debug("sysfs %s: removing dir\n",dentry->d_name.name);
mutex_lock(&dentry->d_inode->i_mutex);
parent_sd = dentry->d_fsdata;
list_for_each_entry_safe(sd, tmp, &parent_sd->s_children, s_sibling) {
if (!sd->s_type || !(sd->s_type & SYSFS_NOT_PINNED))
continue;
sysfs: implement sysfs_dirent active reference and immediate disconnect sysfs: implement sysfs_dirent active reference and immediate disconnect Opening a sysfs node references its associated kobject, so userland can arbitrarily prolong lifetime of a kobject which complicates lifetime rules in drivers. This patch implements active reference and makes the association between kobject and sysfs immediately breakable. Now each sysfs_dirent has two reference counts - s_count and s_active. s_count is a regular reference count which guarantees that the containing sysfs_dirent is accessible. As long as s_count reference is held, all sysfs internal fields in sysfs_dirent are accessible including s_parent and s_name. The newly added s_active is active reference count. This is acquired by invoking sysfs_get_active() and it's the caller's responsibility to ensure sysfs_dirent itself is accessible (should be holding s_count one way or the other). Dereferencing sysfs_dirent to access objects out of sysfs proper requires active reference. This includes access to the associated kobjects, attributes and ops. The active references can be drained and denied by calling sysfs_deactivate(). All active sysfs_dirents must be deactivated after deletion but before the default reference is dropped. This enables immediate disconnect of sysfs nodes. Once a sysfs_dirent is deleted, it won't access any entity external to sysfs proper. Because attr/bin_attr ops access both the node itself and its parent for kobject, they need to hold active references to both. sysfs_get/put_active_two() helpers are provided to help grabbing both references. Parent's is acquired first and released last. Unlike other operations, mmapped area lingers on after mmap() is finished and the module implement implementing it and kobj need to stay referenced till all the mapped pages are gone. This is accomplished by holding one set of active references to the bin_attr and its parent if there have been any mmap during lifetime of an openfile. The references are dropped when the openfile is released. This change makes sysfs lifetime rules independent from both kobject's and module's. It not only fixes several race conditions caused by sysfs not holding onto the proper module when referencing kobject, but also helps fixing and simplifying lifetime management in driver model and drivers by taking sysfs out of the equation. Please read the following message for more info. http://article.gmane.org/gmane.linux.kernel/510293 Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2007-06-13 11:45:16 -07:00
list_move(&sd->s_sibling, &removed);
}
mutex_unlock(&dentry->d_inode->i_mutex);
sysfs: implement sysfs_dirent active reference and immediate disconnect sysfs: implement sysfs_dirent active reference and immediate disconnect Opening a sysfs node references its associated kobject, so userland can arbitrarily prolong lifetime of a kobject which complicates lifetime rules in drivers. This patch implements active reference and makes the association between kobject and sysfs immediately breakable. Now each sysfs_dirent has two reference counts - s_count and s_active. s_count is a regular reference count which guarantees that the containing sysfs_dirent is accessible. As long as s_count reference is held, all sysfs internal fields in sysfs_dirent are accessible including s_parent and s_name. The newly added s_active is active reference count. This is acquired by invoking sysfs_get_active() and it's the caller's responsibility to ensure sysfs_dirent itself is accessible (should be holding s_count one way or the other). Dereferencing sysfs_dirent to access objects out of sysfs proper requires active reference. This includes access to the associated kobjects, attributes and ops. The active references can be drained and denied by calling sysfs_deactivate(). All active sysfs_dirents must be deactivated after deletion but before the default reference is dropped. This enables immediate disconnect of sysfs nodes. Once a sysfs_dirent is deleted, it won't access any entity external to sysfs proper. Because attr/bin_attr ops access both the node itself and its parent for kobject, they need to hold active references to both. sysfs_get/put_active_two() helpers are provided to help grabbing both references. Parent's is acquired first and released last. Unlike other operations, mmapped area lingers on after mmap() is finished and the module implement implementing it and kobj need to stay referenced till all the mapped pages are gone. This is accomplished by holding one set of active references to the bin_attr and its parent if there have been any mmap during lifetime of an openfile. The references are dropped when the openfile is released. This change makes sysfs lifetime rules independent from both kobject's and module's. It not only fixes several race conditions caused by sysfs not holding onto the proper module when referencing kobject, but also helps fixing and simplifying lifetime management in driver model and drivers by taking sysfs out of the equation. Please read the following message for more info. http://article.gmane.org/gmane.linux.kernel/510293 Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2007-06-13 11:45:16 -07:00
list_for_each_entry_safe(sd, tmp, &removed, s_sibling) {
list_del_init(&sd->s_sibling);
sysfs_drop_dentry(sd);
sysfs: implement sysfs_dirent active reference and immediate disconnect sysfs: implement sysfs_dirent active reference and immediate disconnect Opening a sysfs node references its associated kobject, so userland can arbitrarily prolong lifetime of a kobject which complicates lifetime rules in drivers. This patch implements active reference and makes the association between kobject and sysfs immediately breakable. Now each sysfs_dirent has two reference counts - s_count and s_active. s_count is a regular reference count which guarantees that the containing sysfs_dirent is accessible. As long as s_count reference is held, all sysfs internal fields in sysfs_dirent are accessible including s_parent and s_name. The newly added s_active is active reference count. This is acquired by invoking sysfs_get_active() and it's the caller's responsibility to ensure sysfs_dirent itself is accessible (should be holding s_count one way or the other). Dereferencing sysfs_dirent to access objects out of sysfs proper requires active reference. This includes access to the associated kobjects, attributes and ops. The active references can be drained and denied by calling sysfs_deactivate(). All active sysfs_dirents must be deactivated after deletion but before the default reference is dropped. This enables immediate disconnect of sysfs nodes. Once a sysfs_dirent is deleted, it won't access any entity external to sysfs proper. Because attr/bin_attr ops access both the node itself and its parent for kobject, they need to hold active references to both. sysfs_get/put_active_two() helpers are provided to help grabbing both references. Parent's is acquired first and released last. Unlike other operations, mmapped area lingers on after mmap() is finished and the module implement implementing it and kobj need to stay referenced till all the mapped pages are gone. This is accomplished by holding one set of active references to the bin_attr and its parent if there have been any mmap during lifetime of an openfile. The references are dropped when the openfile is released. This change makes sysfs lifetime rules independent from both kobject's and module's. It not only fixes several race conditions caused by sysfs not holding onto the proper module when referencing kobject, but also helps fixing and simplifying lifetime management in driver model and drivers by taking sysfs out of the equation. Please read the following message for more info. http://article.gmane.org/gmane.linux.kernel/510293 Signed-off-by: Tejun Heo <htejun@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2007-06-13 11:45:16 -07:00
sysfs_deactivate(sd);
sysfs_put(sd);
}
remove_dir(dentry);
}
/**
* sysfs_remove_dir - remove an object's directory.
* @kobj: object.
*
* 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 sysfs_rmdir() below, instead of calling separately.
*/
void sysfs_remove_dir(struct kobject * kobj)
{
struct dentry *d = kobj->dentry;
spin_lock(&kobj_sysfs_assoc_lock);
kobj->dentry = NULL;
spin_unlock(&kobj_sysfs_assoc_lock);
__sysfs_remove_dir(d);
}
int sysfs_rename_dir(struct kobject * kobj, struct dentry *new_parent,
const char *new_name)
{
struct sysfs_dirent *sd = kobj->dentry->d_fsdata;
struct sysfs_dirent *parent_sd = new_parent->d_fsdata;
struct dentry *new_dentry;
char *dup_name;
int error;
if (!new_parent)
return -EFAULT;
down_write(&sysfs_rename_sem);
mutex_lock(&new_parent->d_inode->i_mutex);
new_dentry = lookup_one_len(new_name, new_parent, strlen(new_name));
if (IS_ERR(new_dentry)) {
error = PTR_ERR(new_dentry);
goto out_unlock;
}
/* By allowing two different directories with the same
* d_parent we allow this routine to move between different
* shadows of the same directory
*/
error = -EINVAL;
if (kobj->dentry->d_parent->d_inode != new_parent->d_inode ||
new_dentry->d_parent->d_inode != new_parent->d_inode ||
new_dentry == kobj->dentry)
goto out_dput;
error = -EEXIST;
if (new_dentry->d_inode)
goto out_dput;
/* rename kobject and sysfs_dirent */
error = -ENOMEM;
new_name = dup_name = kstrdup(new_name, GFP_KERNEL);
if (!new_name)
goto out_drop;
error = kobject_set_name(kobj, "%s", new_name);
if (error)
goto out_free;
kfree(sd->s_name);
sd->s_name = new_name;
/* move under the new parent */
d_add(new_dentry, NULL);
d_move(kobj->dentry, new_dentry);
list_del_init(&sd->s_sibling);
list_add(&sd->s_sibling, &parent_sd->s_children);
error = 0;
goto out_unlock;
out_free:
kfree(dup_name);
out_drop:
d_drop(new_dentry);
out_dput:
dput(new_dentry);
out_unlock:
mutex_unlock(&new_parent->d_inode->i_mutex);
up_write(&sysfs_rename_sem);
return error;
}
int sysfs_move_dir(struct kobject *kobj, struct kobject *new_parent)
{
struct dentry *old_parent_dentry, *new_parent_dentry, *new_dentry;
struct sysfs_dirent *new_parent_sd, *sd;
int error;
old_parent_dentry = kobj->parent ?
kobj->parent->dentry : sysfs_mount->mnt_sb->s_root;
new_parent_dentry = new_parent ?
new_parent->dentry : sysfs_mount->mnt_sb->s_root;
if (old_parent_dentry->d_inode == new_parent_dentry->d_inode)
return 0; /* nothing to move */
again:
mutex_lock(&old_parent_dentry->d_inode->i_mutex);
if (!mutex_trylock(&new_parent_dentry->d_inode->i_mutex)) {
mutex_unlock(&old_parent_dentry->d_inode->i_mutex);
goto again;
}
new_parent_sd = new_parent_dentry->d_fsdata;
sd = kobj->dentry->d_fsdata;
new_dentry = lookup_one_len(kobj->name, new_parent_dentry,
strlen(kobj->name));
if (IS_ERR(new_dentry)) {
error = PTR_ERR(new_dentry);
goto out;
} else
error = 0;
d_add(new_dentry, NULL);
d_move(kobj->dentry, new_dentry);
dput(new_dentry);
/* Remove from old parent's list and insert into new parent's list. */
list_del_init(&sd->s_sibling);
list_add(&sd->s_sibling, &new_parent_sd->s_children);
out:
mutex_unlock(&new_parent_dentry->d_inode->i_mutex);
mutex_unlock(&old_parent_dentry->d_inode->i_mutex);
return error;
}
static int sysfs_dir_open(struct inode *inode, struct file *file)
{
struct dentry * dentry = file->f_path.dentry;
struct sysfs_dirent * parent_sd = dentry->d_fsdata;
struct sysfs_dirent * sd;
mutex_lock(&dentry->d_inode->i_mutex);
sd = sysfs_new_dirent("_DIR_", 0, 0);
if (sd)
sysfs_attach_dirent(sd, parent_sd, NULL);
mutex_unlock(&dentry->d_inode->i_mutex);
file->private_data = sd;
return sd ? 0 : -ENOMEM;
}
static int sysfs_dir_close(struct inode *inode, struct file *file)
{
struct dentry * dentry = file->f_path.dentry;
struct sysfs_dirent * cursor = file->private_data;
mutex_lock(&dentry->d_inode->i_mutex);
list_del_init(&cursor->s_sibling);
mutex_unlock(&dentry->d_inode->i_mutex);
release_sysfs_dirent(cursor);
return 0;
}
/* Relationship between s_mode and the DT_xxx types */
static inline unsigned char dt_type(struct sysfs_dirent *sd)
{
return (sd->s_mode >> 12) & 15;
}
static int sysfs_readdir(struct file * filp, void * dirent, filldir_t filldir)
{
struct dentry *dentry = filp->f_path.dentry;
struct sysfs_dirent * parent_sd = dentry->d_fsdata;
struct sysfs_dirent *cursor = filp->private_data;
struct list_head *p, *q = &cursor->s_sibling;
ino_t ino;
int i = filp->f_pos;
switch (i) {
case 0:
ino = parent_sd->s_ino;
if (filldir(dirent, ".", 1, i, ino, DT_DIR) < 0)
break;
filp->f_pos++;
i++;
/* fallthrough */
case 1:
if (parent_sd->s_parent)
ino = parent_sd->s_parent->s_ino;
else
ino = parent_sd->s_ino;
if (filldir(dirent, "..", 2, i, ino, DT_DIR) < 0)
break;
filp->f_pos++;
i++;
/* fallthrough */
default:
if (filp->f_pos == 2)
list_move(q, &parent_sd->s_children);
for (p=q->next; p!= &parent_sd->s_children; p=p->next) {
struct sysfs_dirent *next;
const char * name;
int len;
next = list_entry(p, struct sysfs_dirent,
s_sibling);
if (!next->s_type)
continue;
name = next->s_name;
len = strlen(name);
ino = next->s_ino;
if (filldir(dirent, name, len, filp->f_pos, ino,
dt_type(next)) < 0)
return 0;
list_move(q, p);
p = q;
filp->f_pos++;
}
}
return 0;
}
static loff_t sysfs_dir_lseek(struct file * file, loff_t offset, int origin)
{
struct dentry * dentry = file->f_path.dentry;
mutex_lock(&dentry->d_inode->i_mutex);
switch (origin) {
case 1:
offset += file->f_pos;
case 0:
if (offset >= 0)
break;
default:
mutex_unlock(&file->f_path.dentry->d_inode->i_mutex);
return -EINVAL;
}
if (offset != file->f_pos) {
file->f_pos = offset;
if (file->f_pos >= 2) {
struct sysfs_dirent *sd = dentry->d_fsdata;
struct sysfs_dirent *cursor = file->private_data;
struct list_head *p;
loff_t n = file->f_pos - 2;
list_del(&cursor->s_sibling);
p = sd->s_children.next;
while (n && p != &sd->s_children) {
struct sysfs_dirent *next;
next = list_entry(p, struct sysfs_dirent,
s_sibling);
if (next->s_type)
n--;
p = p->next;
}
list_add_tail(&cursor->s_sibling, p);
}
}
mutex_unlock(&dentry->d_inode->i_mutex);
return offset;
}
/**
* sysfs_make_shadowed_dir - Setup so a directory can be shadowed
* @kobj: object we're creating shadow of.
*/
int sysfs_make_shadowed_dir(struct kobject *kobj,
void * (*follow_link)(struct dentry *, struct nameidata *))
{
struct inode *inode;
struct inode_operations *i_op;
inode = kobj->dentry->d_inode;
if (inode->i_op != &sysfs_dir_inode_operations)
return -EINVAL;
i_op = kmalloc(sizeof(*i_op), GFP_KERNEL);
if (!i_op)
return -ENOMEM;
memcpy(i_op, &sysfs_dir_inode_operations, sizeof(*i_op));
i_op->follow_link = follow_link;
/* Locking of inode->i_op?
* Since setting i_op is a single word write and they
* are atomic we should be ok here.
*/
inode->i_op = i_op;
return 0;
}
/**
* sysfs_create_shadow_dir - create a shadow directory for an object.
* @kobj: object we're creating directory for.
*
* sysfs_make_shadowed_dir must already have been called on this
* directory.
*/
struct dentry *sysfs_create_shadow_dir(struct kobject *kobj)
{
struct dentry *dir = kobj->dentry;
struct inode *inode = dir->d_inode;
struct dentry *parent = dir->d_parent;
struct sysfs_dirent *parent_sd = parent->d_fsdata;
struct dentry *shadow;
struct sysfs_dirent *sd;
shadow = ERR_PTR(-EINVAL);
if (!sysfs_is_shadowed_inode(inode))
goto out;
shadow = d_alloc(parent, &dir->d_name);
if (!shadow)
goto nomem;
sd = sysfs_new_dirent("_SHADOW_", inode->i_mode, SYSFS_DIR);
if (!sd)
goto nomem;
sd->s_elem.dir.kobj = kobj;
/* point to parent_sd but don't attach to it */
sd->s_parent = sysfs_get(parent_sd);
sysfs_attach_dirent(sd, NULL, shadow);
d_instantiate(shadow, igrab(inode));
inc_nlink(inode);
inc_nlink(parent->d_inode);
dget(shadow); /* Extra count - pin the dentry in core */
out:
return shadow;
nomem:
dput(shadow);
shadow = ERR_PTR(-ENOMEM);
goto out;
}
/**
* sysfs_remove_shadow_dir - remove an object's directory.
* @shadow: dentry of shadow directory
*
* 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 sysfs_rmdir() below, instead of calling separately.
*/
void sysfs_remove_shadow_dir(struct dentry *shadow)
{
__sysfs_remove_dir(shadow);
}
const struct file_operations sysfs_dir_operations = {
.open = sysfs_dir_open,
.release = sysfs_dir_close,
.llseek = sysfs_dir_lseek,
.read = generic_read_dir,
.readdir = sysfs_readdir,
};