* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6: (25 commits)
cifs: remove unnecessary dentry_unhash on rmdir/rename_dir
ocfs2: remove unnecessary dentry_unhash on rmdir/rename_dir
exofs: remove unnecessary dentry_unhash on rmdir/rename_dir
nfs: remove unnecessary dentry_unhash on rmdir/rename_dir
ext2: remove unnecessary dentry_unhash on rmdir/rename_dir
ext3: remove unnecessary dentry_unhash on rmdir/rename_dir
ext4: remove unnecessary dentry_unhash on rmdir/rename_dir
btrfs: remove unnecessary dentry_unhash in rmdir/rename_dir
ceph: remove unnecessary dentry_unhash calls
vfs: clean up vfs_rename_other
vfs: clean up vfs_rename_dir
vfs: clean up vfs_rmdir
vfs: fix vfs_rename_dir for FS_RENAME_DOES_D_MOVE filesystems
libfs: drop unneeded dentry_unhash
vfs: update dentry_unhash() comment
vfs: push dentry_unhash on rename_dir into file systems
vfs: push dentry_unhash on rmdir into file systems
vfs: remove dget() from dentry_unhash()
vfs: dentry_unhash immediately prior to rmdir
vfs: Block mmapped writes while the fs is frozen
...
def_bool n
config EXPORTFS
- bool
+ tristate
config FILE_LOCKING
bool "Enable POSIX file locking API" if EXPERT
config TMPFS_POSIX_ACL
bool "Tmpfs POSIX Access Control Lists"
depends on TMPFS
+ select TMPFS_XATTR
select GENERIC_ACL
help
POSIX Access Control Lists (ACLs) support permissions for users and
If you don't know what Access Control Lists are, say N.
+config TMPFS_XATTR
+ bool "Tmpfs extended attributes"
+ depends on TMPFS
+ default n
+ help
+ Extended attributes are name:value pairs associated with inodes by
+ the kernel or by users (see the attr(5) manual page, or visit
+ <http://acl.bestbits.at/> for details).
+
+ Currently this enables support for the trusted.* and
+ security.* namespaces.
+
+ You need this for POSIX ACL support on tmpfs.
+
+ If unsure, say N.
+
config HUGETLBFS
bool "HugeTLB file system support"
depends on X86 || IA64 || SPARC64 || (S390 && 64BIT) || \
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);
+ /* Coordinate with configfs_readdir */
+ spin_lock(&configfs_dirent_lock);
sd->s_dentry = NULL;
+ spin_unlock(&configfs_dirent_lock);
configfs_put(sd);
}
iput(inode);
sd = child->d_fsdata;
sd->s_type |= CONFIGFS_USET_DEFAULT;
} else {
- d_delete(child);
+ BUG_ON(child->d_inode);
+ d_drop(child);
dput(child);
}
}
struct module *subsys_owner = NULL, *dead_item_owner = NULL;
int ret;
+ dentry_unhash(dentry);
+
if (dentry->d_parent == configfs_sb->s_root)
return -EPERM;
struct configfs_dirent * parent_sd = dentry->d_fsdata;
struct configfs_dirent *cursor = filp->private_data;
struct list_head *p, *q = &cursor->s_sibling;
- ino_t ino;
+ ino_t ino = 0;
int i = filp->f_pos;
switch (i) {
struct configfs_dirent *next;
const char * name;
int len;
+ struct inode *inode = NULL;
next = list_entry(p, struct configfs_dirent,
s_sibling);
name = configfs_get_name(next);
len = strlen(name);
- if (next->s_dentry)
- ino = next->s_dentry->d_inode->i_ino;
- else
+
+ /*
+ * 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(configfs_sb, 2);
if (filldir(dirent, name, len, filp->f_pos, ino,
err = configfs_attach_group(sd->s_element, &group->cg_item,
dentry);
if (err) {
- d_delete(dentry);
+ BUG_ON(dentry->d_inode);
+ d_drop(dentry);
dput(dentry);
} else {
spin_lock(&configfs_dirent_lock);
struct fat_slot_info sinfo;
int err;
+ dentry_unhash(dentry);
+
lock_super(sb);
/*
* Check whether the directory is not in use, then check
old_inode = old_dentry->d_inode;
new_inode = new_dentry->d_inode;
+ if (new_inode && S_ISDIR(new_inode->i_mode))
+ dentry_unhash(new_dentry);
+
err = fat_scan(old_dir, old_name, &old_sinfo);
if (err) {
err = -EIO;
static void setup(struct super_block *sb)
{
+ MSDOS_SB(sb)->dir_ops = &msdos_dir_inode_operations;
sb->s_d_op = &msdos_dentry_operations;
sb->s_flags |= MS_NOATIME;
}
static int msdos_fill_super(struct super_block *sb, void *data, int silent)
{
- return fat_fill_super(sb, data, silent, &msdos_dir_inode_operations,
- 0, setup);
+ return fat_fill_super(sb, data, silent, 0, setup);
}
static struct dentry *msdos_mount(struct file_system_type *fs_type,
struct fat_slot_info sinfo;
int err;
+ dentry_unhash(dentry);
+
lock_super(sb);
err = fat_dir_empty(inode);
int err, is_dir, update_dotdot, corrupt = 0;
struct super_block *sb = old_dir->i_sb;
+ if (new_dentry->d_inode && S_ISDIR(new_dentry->d_inode->i_mode))
+ dentry_unhash(new_dentry);
+
old_sinfo.bh = sinfo.bh = dotdot_bh = NULL;
old_inode = old_dentry->d_inode;
new_inode = new_dentry->d_inode;
static void setup(struct super_block *sb)
{
+ MSDOS_SB(sb)->dir_ops = &vfat_dir_inode_operations;
if (MSDOS_SB(sb)->options.name_check != 's')
sb->s_d_op = &vfat_ci_dentry_ops;
else
static int vfat_fill_super(struct super_block *sb, void *data, int silent)
{
- return fat_fill_super(sb, data, silent, &vfat_dir_inode_operations,
- 1, setup);
+ return fat_fill_super(sb, data, silent, 1, setup);
}
static struct dentry *vfat_mount(struct file_system_type *fs_type,
if (!inode)
return 0;
- if (nd->flags & LOOKUP_RCU)
+ if (nd && (nd->flags & LOOKUP_RCU))
return -ECHILD;
fc = get_fuse_conn(inode);
if (IS_ERR(req))
return PTR_ERR(req);
+ dentry_unhash(entry);
+
req->in.h.opcode = FUSE_RMDIR;
req->in.h.nodeid = get_node_id(dir);
req->in.numargs = 1;
struct fuse_rename_in inarg;
struct fuse_conn *fc = get_fuse_conn(olddir);
struct fuse_req *req = fuse_get_req(fc);
+
+ if (newent->d_inode && S_ISDIR(newent->d_inode->i_mode))
+ dentry_unhash(newent);
+
if (IS_ERR(req))
return PTR_ERR(req);
fnode = hpfs_alloc_fnode(dir->i_sb, hpfs_i(dir)->i_dno, &fno, &bh);
if (!fnode)
goto bail;
- dnode = hpfs_alloc_dnode(dir->i_sb, fno, &dno, &qbh0, 1);
+ dnode = hpfs_alloc_dnode(dir->i_sb, fno, &dno, &qbh0);
if (!dnode)
goto bail1;
memset(&dee, 0, sizeof dee);
if (!(mode & 0222)) dee.read_only = 1;
/*dee.archive = 0;*/
dee.hidden = name[0] == '.';
- dee.fnode = fno;
- dee.creation_date = dee.write_date = dee.read_date = gmt_to_local(dir->i_sb, get_seconds());
+ dee.fnode = cpu_to_le32(fno);
+ dee.creation_date = dee.write_date = dee.read_date = cpu_to_le32(gmt_to_local(dir->i_sb, get_seconds()));
result = new_inode(dir->i_sb);
if (!result)
goto bail2;
result->i_ino = fno;
hpfs_i(result)->i_parent_dir = dir->i_ino;
hpfs_i(result)->i_dno = dno;
- result->i_ctime.tv_sec = result->i_mtime.tv_sec = result->i_atime.tv_sec = local_to_gmt(dir->i_sb, dee.creation_date);
+ result->i_ctime.tv_sec = result->i_mtime.tv_sec = result->i_atime.tv_sec = local_to_gmt(dir->i_sb, le32_to_cpu(dee.creation_date));
result->i_ctime.tv_nsec = 0;
result->i_mtime.tv_nsec = 0;
result->i_atime.tv_nsec = 0;
if (dee.read_only)
result->i_mode &= ~0222;
- mutex_lock(&hpfs_i(dir)->i_mutex);
- r = hpfs_add_dirent(dir, name, len, &dee, 0);
+ r = hpfs_add_dirent(dir, name, len, &dee);
if (r == 1)
goto bail3;
if (r == -1) {
}
fnode->len = len;
memcpy(fnode->name, name, len > 15 ? 15 : len);
- fnode->up = dir->i_ino;
+ fnode->up = cpu_to_le32(dir->i_ino);
fnode->dirflag = 1;
fnode->btree.n_free_nodes = 7;
fnode->btree.n_used_nodes = 1;
- fnode->btree.first_free = 0x14;
- fnode->u.external[0].disk_secno = dno;
- fnode->u.external[0].file_secno = -1;
+ fnode->btree.first_free = cpu_to_le16(0x14);
+ fnode->u.external[0].disk_secno = cpu_to_le32(dno);
+ fnode->u.external[0].file_secno = cpu_to_le32(-1);
dnode->root_dnode = 1;
- dnode->up = fno;
+ dnode->up = cpu_to_le32(fno);
de = hpfs_add_de(dir->i_sb, dnode, "\001\001", 2, 0);
- de->creation_date = de->write_date = de->read_date = gmt_to_local(dir->i_sb, get_seconds());
+ de->creation_date = de->write_date = de->read_date = cpu_to_le32(gmt_to_local(dir->i_sb, get_seconds()));
if (!(mode & 0222)) de->read_only = 1;
de->first = de->directory = 1;
/*de->hidden = de->system = 0;*/
- de->fnode = fno;
+ de->fnode = cpu_to_le32(fno);
mark_buffer_dirty(bh);
brelse(bh);
hpfs_mark_4buffers_dirty(&qbh0);
hpfs_write_inode_nolock(result);
}
d_instantiate(dentry, result);
- mutex_unlock(&hpfs_i(dir)->i_mutex);
hpfs_unlock(dir->i_sb);
return 0;
bail3:
- mutex_unlock(&hpfs_i(dir)->i_mutex);
iput(result);
bail2:
hpfs_brelse4(&qbh0);
if (!(mode & 0222)) dee.read_only = 1;
dee.archive = 1;
dee.hidden = name[0] == '.';
- dee.fnode = fno;
- dee.creation_date = dee.write_date = dee.read_date = gmt_to_local(dir->i_sb, get_seconds());
+ dee.fnode = cpu_to_le32(fno);
+ dee.creation_date = dee.write_date = dee.read_date = cpu_to_le32(gmt_to_local(dir->i_sb, get_seconds()));
result = new_inode(dir->i_sb);
if (!result)
result->i_op = &hpfs_file_iops;
result->i_fop = &hpfs_file_ops;
result->i_nlink = 1;
- hpfs_decide_conv(result, name, len);
hpfs_i(result)->i_parent_dir = dir->i_ino;
- result->i_ctime.tv_sec = result->i_mtime.tv_sec = result->i_atime.tv_sec = local_to_gmt(dir->i_sb, dee.creation_date);
+ result->i_ctime.tv_sec = result->i_mtime.tv_sec = result->i_atime.tv_sec = local_to_gmt(dir->i_sb, le32_to_cpu(dee.creation_date));
result->i_ctime.tv_nsec = 0;
result->i_mtime.tv_nsec = 0;
result->i_atime.tv_nsec = 0;
result->i_data.a_ops = &hpfs_aops;
hpfs_i(result)->mmu_private = 0;
- mutex_lock(&hpfs_i(dir)->i_mutex);
- r = hpfs_add_dirent(dir, name, len, &dee, 0);
+ r = hpfs_add_dirent(dir, name, len, &dee);
if (r == 1)
goto bail2;
if (r == -1) {
}
fnode->len = len;
memcpy(fnode->name, name, len > 15 ? 15 : len);
- fnode->up = dir->i_ino;
+ fnode->up = cpu_to_le32(dir->i_ino);
mark_buffer_dirty(bh);
brelse(bh);
hpfs_write_inode_nolock(result);
}
d_instantiate(dentry, result);
- mutex_unlock(&hpfs_i(dir)->i_mutex);
hpfs_unlock(dir->i_sb);
return 0;
bail2:
- mutex_unlock(&hpfs_i(dir)->i_mutex);
iput(result);
bail1:
brelse(bh);
if (!(mode & 0222)) dee.read_only = 1;
dee.archive = 1;
dee.hidden = name[0] == '.';
- dee.fnode = fno;
- dee.creation_date = dee.write_date = dee.read_date = gmt_to_local(dir->i_sb, get_seconds());
+ dee.fnode = cpu_to_le32(fno);
+ dee.creation_date = dee.write_date = dee.read_date = cpu_to_le32(gmt_to_local(dir->i_sb, get_seconds()));
result = new_inode(dir->i_sb);
if (!result)
hpfs_init_inode(result);
result->i_ino = fno;
hpfs_i(result)->i_parent_dir = dir->i_ino;
- result->i_ctime.tv_sec = result->i_mtime.tv_sec = result->i_atime.tv_sec = local_to_gmt(dir->i_sb, dee.creation_date);
+ result->i_ctime.tv_sec = result->i_mtime.tv_sec = result->i_atime.tv_sec = local_to_gmt(dir->i_sb, le32_to_cpu(dee.creation_date));
result->i_ctime.tv_nsec = 0;
result->i_mtime.tv_nsec = 0;
result->i_atime.tv_nsec = 0;
result->i_blocks = 1;
init_special_inode(result, mode, rdev);
- mutex_lock(&hpfs_i(dir)->i_mutex);
- r = hpfs_add_dirent(dir, name, len, &dee, 0);
+ r = hpfs_add_dirent(dir, name, len, &dee);
if (r == 1)
goto bail2;
if (r == -1) {
}
fnode->len = len;
memcpy(fnode->name, name, len > 15 ? 15 : len);
- fnode->up = dir->i_ino;
+ fnode->up = cpu_to_le32(dir->i_ino);
mark_buffer_dirty(bh);
insert_inode_hash(result);
hpfs_write_inode_nolock(result);
d_instantiate(dentry, result);
- mutex_unlock(&hpfs_i(dir)->i_mutex);
brelse(bh);
hpfs_unlock(dir->i_sb);
return 0;
bail2:
- mutex_unlock(&hpfs_i(dir)->i_mutex);
iput(result);
bail1:
brelse(bh);
memset(&dee, 0, sizeof dee);
dee.archive = 1;
dee.hidden = name[0] == '.';
- dee.fnode = fno;
- dee.creation_date = dee.write_date = dee.read_date = gmt_to_local(dir->i_sb, get_seconds());
+ dee.fnode = cpu_to_le32(fno);
+ dee.creation_date = dee.write_date = dee.read_date = cpu_to_le32(gmt_to_local(dir->i_sb, get_seconds()));
result = new_inode(dir->i_sb);
if (!result)
result->i_ino = fno;
hpfs_init_inode(result);
hpfs_i(result)->i_parent_dir = dir->i_ino;
- result->i_ctime.tv_sec = result->i_mtime.tv_sec = result->i_atime.tv_sec = local_to_gmt(dir->i_sb, dee.creation_date);
+ result->i_ctime.tv_sec = result->i_mtime.tv_sec = result->i_atime.tv_sec = local_to_gmt(dir->i_sb, le32_to_cpu(dee.creation_date));
result->i_ctime.tv_nsec = 0;
result->i_mtime.tv_nsec = 0;
result->i_atime.tv_nsec = 0;
result->i_op = &page_symlink_inode_operations;
result->i_data.a_ops = &hpfs_symlink_aops;
- mutex_lock(&hpfs_i(dir)->i_mutex);
- r = hpfs_add_dirent(dir, name, len, &dee, 0);
+ r = hpfs_add_dirent(dir, name, len, &dee);
if (r == 1)
goto bail2;
if (r == -1) {
}
fnode->len = len;
memcpy(fnode->name, name, len > 15 ? 15 : len);
- fnode->up = dir->i_ino;
+ fnode->up = cpu_to_le32(dir->i_ino);
hpfs_set_ea(result, fnode, "SYMLINK", symlink, strlen(symlink));
mark_buffer_dirty(bh);
brelse(bh);
hpfs_write_inode_nolock(result);
d_instantiate(dentry, result);
- mutex_unlock(&hpfs_i(dir)->i_mutex);
hpfs_unlock(dir->i_sb);
return 0;
bail2:
- mutex_unlock(&hpfs_i(dir)->i_mutex);
iput(result);
bail1:
brelse(bh);
struct hpfs_dirent *de;
struct inode *inode = dentry->d_inode;
dnode_secno dno;
- fnode_secno fno;
int r;
int rep = 0;
int err;
hpfs_lock(dir->i_sb);
hpfs_adjust_length(name, &len);
again:
- mutex_lock(&hpfs_i(inode)->i_parent_mutex);
- mutex_lock(&hpfs_i(dir)->i_mutex);
err = -ENOENT;
de = map_dirent(dir, hpfs_i(dir)->i_dno, name, len, &dno, &qbh);
if (!de)
if (de->directory)
goto out1;
- fno = de->fnode;
r = hpfs_remove_dirent(dir, dno, de, &qbh, 1);
switch (r) {
case 1:
if (rep++)
break;
- mutex_unlock(&hpfs_i(dir)->i_mutex);
- mutex_unlock(&hpfs_i(inode)->i_parent_mutex);
dentry_unhash(dentry);
if (!d_unhashed(dentry)) {
- dput(dentry);
hpfs_unlock(dir->i_sb);
return -ENOSPC;
}
!S_ISREG(inode->i_mode) ||
get_write_access(inode)) {
d_rehash(dentry);
- dput(dentry);
} else {
struct iattr newattrs;
/*printk("HPFS: truncating file before delete.\n");*/
newattrs.ia_valid = ATTR_SIZE | ATTR_CTIME;
err = notify_change(dentry, &newattrs);
put_write_access(inode);
- dput(dentry);
if (!err)
goto again;
}
out1:
hpfs_brelse4(&qbh);
out:
- mutex_unlock(&hpfs_i(dir)->i_mutex);
- mutex_unlock(&hpfs_i(inode)->i_parent_mutex);
hpfs_unlock(dir->i_sb);
return err;
}
struct hpfs_dirent *de;
struct inode *inode = dentry->d_inode;
dnode_secno dno;
- fnode_secno fno;
int n_items = 0;
int err;
int r;
+ dentry_unhash(dentry);
+
hpfs_adjust_length(name, &len);
hpfs_lock(dir->i_sb);
- mutex_lock(&hpfs_i(inode)->i_parent_mutex);
- mutex_lock(&hpfs_i(dir)->i_mutex);
err = -ENOENT;
de = map_dirent(dir, hpfs_i(dir)->i_dno, name, len, &dno, &qbh);
if (!de)
if (n_items)
goto out1;
- fno = de->fnode;
r = hpfs_remove_dirent(dir, dno, de, &qbh, 1);
switch (r) {
case 1:
out1:
hpfs_brelse4(&qbh);
out:
- mutex_unlock(&hpfs_i(dir)->i_mutex);
- mutex_unlock(&hpfs_i(inode)->i_parent_mutex);
hpfs_unlock(dir->i_sb);
return err;
}
struct buffer_head *bh;
struct fnode *fnode;
int err;
+
+ if (new_inode && S_ISDIR(new_inode->i_mode))
+ dentry_unhash(new_dentry);
+
if ((err = hpfs_chk_name(new_name, &new_len))) return err;
err = 0;
hpfs_adjust_length(old_name, &old_len);
hpfs_lock(i->i_sb);
/* order doesn't matter, due to VFS exclusion */
- mutex_lock(&hpfs_i(i)->i_parent_mutex);
- if (new_inode)
- mutex_lock(&hpfs_i(new_inode)->i_parent_mutex);
- mutex_lock(&hpfs_i(old_dir)->i_mutex);
- if (new_dir != old_dir)
- mutex_lock(&hpfs_i(new_dir)->i_mutex);
/* Erm? Moving over the empty non-busy directory is perfectly legal */
if (new_inode && S_ISDIR(new_inode->i_mode)) {
if (new_dir == old_dir) hpfs_brelse4(&qbh);
- hpfs_lock_creation(i->i_sb);
- if ((r = hpfs_add_dirent(new_dir, new_name, new_len, &de, 1))) {
- hpfs_unlock_creation(i->i_sb);
+ if ((r = hpfs_add_dirent(new_dir, new_name, new_len, &de))) {
if (r == -1) hpfs_error(new_dir->i_sb, "hpfs_rename: dirent already exists!");
err = r == 1 ? -ENOSPC : -EFSERROR;
if (new_dir != old_dir) hpfs_brelse4(&qbh);
if (new_dir == old_dir)
if (!(dep = map_dirent(old_dir, hpfs_i(old_dir)->i_dno, old_name, old_len, &dno, &qbh))) {
- hpfs_unlock_creation(i->i_sb);
hpfs_error(i->i_sb, "lookup succeeded but map dirent failed at #2");
err = -ENOENT;
goto end1;
}
if ((r = hpfs_remove_dirent(old_dir, dno, dep, &qbh, 0))) {
- hpfs_unlock_creation(i->i_sb);
hpfs_error(i->i_sb, "hpfs_rename: could not remove dirent");
err = r == 2 ? -ENOSPC : -EFSERROR;
goto end1;
}
- hpfs_unlock_creation(i->i_sb);
-
+
end:
hpfs_i(i)->i_parent_dir = new_dir->i_ino;
if (S_ISDIR(i->i_mode)) {
drop_nlink(old_dir);
}
if ((fnode = hpfs_map_fnode(i->i_sb, i->i_ino, &bh))) {
- fnode->up = new_dir->i_ino;
+ fnode->up = cpu_to_le32(new_dir->i_ino);
fnode->len = new_len;
memcpy(fnode->name, new_name, new_len>15?15:new_len);
if (new_len < 15) memset(&fnode->name[new_len], 0, 15 - new_len);
mark_buffer_dirty(bh);
brelse(bh);
}
- hpfs_i(i)->i_conv = hpfs_sb(i->i_sb)->sb_conv;
- hpfs_decide_conv(i, new_name, new_len);
end1:
- if (old_dir != new_dir)
- mutex_unlock(&hpfs_i(new_dir)->i_mutex);
- mutex_unlock(&hpfs_i(old_dir)->i_mutex);
- mutex_unlock(&hpfs_i(i)->i_parent_mutex);
- if (new_inode)
- mutex_unlock(&hpfs_i(new_inode)->i_parent_mutex);
hpfs_unlock(i->i_sb);
return err;
}
static int acl_permission_check(struct inode *inode, int mask, unsigned int flags,
int (*check_acl)(struct inode *inode, int mask, unsigned int flags))
{
- umode_t mode = inode->i_mode;
+ unsigned int mode = inode->i_mode;
mask &= MAY_READ | MAY_WRITE | MAY_EXEC;
}
EXPORT_SYMBOL(path_put);
- /**
- * nameidata_drop_rcu - drop this nameidata out of rcu-walk
- * @nd: nameidata pathwalk data to drop
- * Returns: 0 on success, -ECHILD on failure
- *
+ /*
* Path walking has 2 modes, rcu-walk and ref-walk (see
- * Documentation/filesystems/path-lookup.txt). __drop_rcu* functions attempt
- * to drop out of rcu-walk mode and take normal reference counts on dentries
- * and vfsmounts to transition to rcu-walk mode. __drop_rcu* functions take
- * refcounts at the last known good point before rcu-walk got stuck, so
- * ref-walk may continue from there. If this is not successful (eg. a seqcount
- * has changed), then failure is returned and path walk restarts from the
- * beginning in ref-walk mode.
- *
- * nameidata_drop_rcu attempts to drop the current nd->path and nd->root into
- * ref-walk. Must be called from rcu-walk context.
+ * Documentation/filesystems/path-lookup.txt). In situations when we can't
+ * continue in RCU mode, we attempt to drop out of rcu-walk mode and grab
+ * normal reference counts on dentries and vfsmounts to transition to rcu-walk
+ * mode. Refcounts are grabbed at the last known good point before rcu-walk
+ * got stuck, so ref-walk may continue from there. If this is not successful
+ * (eg. a seqcount has changed), then failure is returned and it's up to caller
+ * to restart the path walk from the beginning in ref-walk mode.
*/
- static int nameidata_drop_rcu(struct nameidata *nd)
- {
- struct fs_struct *fs = current->fs;
- struct dentry *dentry = nd->path.dentry;
- int want_root = 0;
-
- BUG_ON(!(nd->flags & LOOKUP_RCU));
- if (nd->root.mnt && !(nd->flags & LOOKUP_ROOT)) {
- want_root = 1;
- spin_lock(&fs->lock);
- if (nd->root.mnt != fs->root.mnt ||
- nd->root.dentry != fs->root.dentry)
- goto err_root;
- }
- spin_lock(&dentry->d_lock);
- if (!__d_rcu_to_refcount(dentry, nd->seq))
- goto err;
- BUG_ON(nd->inode != dentry->d_inode);
- spin_unlock(&dentry->d_lock);
- if (want_root) {
- path_get(&nd->root);
- spin_unlock(&fs->lock);
- }
- mntget(nd->path.mnt);
-
- rcu_read_unlock();
- br_read_unlock(vfsmount_lock);
- nd->flags &= ~LOOKUP_RCU;
- return 0;
- err:
- spin_unlock(&dentry->d_lock);
- err_root:
- if (want_root)
- spin_unlock(&fs->lock);
- return -ECHILD;
- }
-
- /* Try to drop out of rcu-walk mode if we were in it, otherwise do nothing. */
- static inline int nameidata_drop_rcu_maybe(struct nameidata *nd)
- {
- if (nd->flags & LOOKUP_RCU)
- return nameidata_drop_rcu(nd);
- return 0;
- }
/**
- * nameidata_dentry_drop_rcu - drop nameidata and dentry out of rcu-walk
- * @nd: nameidata pathwalk data to drop
- * @dentry: dentry to drop
+ * unlazy_walk - try to switch to ref-walk mode.
+ * @nd: nameidata pathwalk data
+ * @dentry: child of nd->path.dentry or NULL
* Returns: 0 on success, -ECHILD on failure
*
- * nameidata_dentry_drop_rcu attempts to drop the current nd->path and nd->root,
- * and dentry into ref-walk. @dentry must be a path found by a do_lookup call on
- * @nd. Must be called from rcu-walk context.
+ * unlazy_walk attempts to legitimize the current nd->path, nd->root and dentry
+ * for ref-walk mode. @dentry must be a path found by a do_lookup call on
+ * @nd or NULL. Must be called from rcu-walk context.
*/
- static int nameidata_dentry_drop_rcu(struct nameidata *nd, struct dentry *dentry)
+ static int unlazy_walk(struct nameidata *nd, struct dentry *dentry)
{
struct fs_struct *fs = current->fs;
struct dentry *parent = nd->path.dentry;
goto err_root;
}
spin_lock(&parent->d_lock);
- spin_lock_nested(&dentry->d_lock, DENTRY_D_LOCK_NESTED);
- if (!__d_rcu_to_refcount(dentry, nd->seq))
- goto err;
- /*
- * If the sequence check on the child dentry passed, then the child has
- * not been removed from its parent. This means the parent dentry must
- * be valid and able to take a reference at this point.
- */
- BUG_ON(!IS_ROOT(dentry) && dentry->d_parent != parent);
- BUG_ON(!parent->d_count);
- parent->d_count++;
- spin_unlock(&dentry->d_lock);
+ if (!dentry) {
+ if (!__d_rcu_to_refcount(parent, nd->seq))
+ goto err_parent;
+ BUG_ON(nd->inode != parent->d_inode);
+ } else {
+ spin_lock_nested(&dentry->d_lock, DENTRY_D_LOCK_NESTED);
+ if (!__d_rcu_to_refcount(dentry, nd->seq))
+ goto err_child;
+ /*
+ * If the sequence check on the child dentry passed, then
+ * the child has not been removed from its parent. This
+ * means the parent dentry must be valid and able to take
+ * a reference at this point.
+ */
+ BUG_ON(!IS_ROOT(dentry) && dentry->d_parent != parent);
+ BUG_ON(!parent->d_count);
+ parent->d_count++;
+ spin_unlock(&dentry->d_lock);
+ }
spin_unlock(&parent->d_lock);
if (want_root) {
path_get(&nd->root);
br_read_unlock(vfsmount_lock);
nd->flags &= ~LOOKUP_RCU;
return 0;
- err:
+
+ err_child:
spin_unlock(&dentry->d_lock);
+ err_parent:
spin_unlock(&parent->d_lock);
err_root:
if (want_root)
return -ECHILD;
}
- /* Try to drop out of rcu-walk mode if we were in it, otherwise do nothing. */
- static inline int nameidata_dentry_drop_rcu_maybe(struct nameidata *nd, struct dentry *dentry)
- {
- if (nd->flags & LOOKUP_RCU) {
- if (unlikely(nameidata_dentry_drop_rcu(nd, dentry))) {
- nd->flags &= ~LOOKUP_RCU;
- if (!(nd->flags & LOOKUP_ROOT))
- nd->root.mnt = NULL;
- rcu_read_unlock();
- br_read_unlock(vfsmount_lock);
- return -ECHILD;
- }
- }
- return 0;
- }
-
- /**
- * nameidata_drop_rcu_last - drop nameidata ending path walk out of rcu-walk
- * @nd: nameidata pathwalk data to drop
- * Returns: 0 on success, -ECHILD on failure
- *
- * nameidata_drop_rcu_last attempts to drop the current nd->path into ref-walk.
- * nd->path should be the final element of the lookup, so nd->root is discarded.
- * Must be called from rcu-walk context.
- */
- static int nameidata_drop_rcu_last(struct nameidata *nd)
- {
- struct dentry *dentry = nd->path.dentry;
-
- BUG_ON(!(nd->flags & LOOKUP_RCU));
- nd->flags &= ~LOOKUP_RCU;
- if (!(nd->flags & LOOKUP_ROOT))
- nd->root.mnt = NULL;
- spin_lock(&dentry->d_lock);
- if (!__d_rcu_to_refcount(dentry, nd->seq))
- goto err_unlock;
- BUG_ON(nd->inode != dentry->d_inode);
- spin_unlock(&dentry->d_lock);
-
- mntget(nd->path.mnt);
-
- rcu_read_unlock();
- br_read_unlock(vfsmount_lock);
-
- return 0;
-
- err_unlock:
- spin_unlock(&dentry->d_lock);
- rcu_read_unlock();
- br_read_unlock(vfsmount_lock);
- return -ECHILD;
- }
-
/**
* release_open_intent - free up open intent resources
* @nd: pointer to nameidata
return dentry;
}
- /*
- * handle_reval_path - force revalidation of a dentry
- *
- * In some situations the path walking code will trust dentries without
- * revalidating them. This causes problems for filesystems that depend on
- * d_revalidate to handle file opens (e.g. NFSv4). When FS_REVAL_DOT is set
- * (which indicates that it's possible for the dentry to go stale), force
- * a d_revalidate call before proceeding.
+ /**
+ * complete_walk - successful completion of path walk
+ * @nd: pointer nameidata
*
- * Returns 0 if the revalidation was successful. If the revalidation fails,
- * either return the error returned by d_revalidate or -ESTALE if the
- * revalidation it just returned 0. If d_revalidate returns 0, we attempt to
- * invalidate the dentry. It's up to the caller to handle putting references
- * to the path if necessary.
+ * If we had been in RCU mode, drop out of it and legitimize nd->path.
+ * Revalidate the final result, unless we'd already done that during
+ * the path walk or the filesystem doesn't ask for it. Return 0 on
+ * success, -error on failure. In case of failure caller does not
+ * need to drop nd->path.
*/
- static inline int handle_reval_path(struct nameidata *nd)
+ static int complete_walk(struct nameidata *nd)
{
struct dentry *dentry = nd->path.dentry;
int status;
+ if (nd->flags & LOOKUP_RCU) {
+ nd->flags &= ~LOOKUP_RCU;
+ if (!(nd->flags & LOOKUP_ROOT))
+ nd->root.mnt = NULL;
+ spin_lock(&dentry->d_lock);
+ if (unlikely(!__d_rcu_to_refcount(dentry, nd->seq))) {
+ spin_unlock(&dentry->d_lock);
+ rcu_read_unlock();
+ br_read_unlock(vfsmount_lock);
+ return -ECHILD;
+ }
+ BUG_ON(nd->inode != dentry->d_inode);
+ spin_unlock(&dentry->d_lock);
+ mntget(nd->path.mnt);
+ rcu_read_unlock();
+ br_read_unlock(vfsmount_lock);
+ }
+
if (likely(!(nd->flags & LOOKUP_JUMPED)))
return 0;
if (!status)
status = -ESTALE;
+ path_put(&nd->path);
return status;
}
if (likely(__follow_mount_rcu(nd, path, inode, false)))
return 0;
unlazy:
- if (dentry) {
- if (nameidata_dentry_drop_rcu(nd, dentry))
- return -ECHILD;
- } else {
- if (nameidata_drop_rcu(nd))
- return -ECHILD;
- }
+ if (unlazy_walk(nd, dentry))
+ return -ECHILD;
} else {
dentry = __d_lookup(parent, name);
}
int err = exec_permission(nd->inode, IPERM_FLAG_RCU);
if (err != -ECHILD)
return err;
- if (nameidata_drop_rcu(nd))
+ if (unlazy_walk(nd, NULL))
return -ECHILD;
}
return exec_permission(nd->inode, 0);
return -ENOENT;
}
if (unlikely(inode->i_op->follow_link) && follow) {
- if (nameidata_dentry_drop_rcu_maybe(nd, path->dentry))
- return -ECHILD;
+ if (nd->flags & LOOKUP_RCU) {
+ if (unlikely(unlazy_walk(nd, path->dentry))) {
+ terminate_walk(nd);
+ return -ECHILD;
+ }
+ }
BUG_ON(inode != path->dentry->d_inode);
return 1;
}
{
int res;
- BUG_ON(nd->depth >= MAX_NESTED_LINKS);
if (unlikely(current->link_count >= MAX_NESTED_LINKS)) {
path_put_conditional(path, nd);
path_put(&nd->path);
return -ELOOP;
}
+ BUG_ON(nd->depth >= MAX_NESTED_LINKS);
nd->depth++;
current->link_count++;
}
}
- if (nd->flags & LOOKUP_RCU) {
- /* went all way through without dropping RCU */
- BUG_ON(err);
- if (nameidata_drop_rcu_last(nd))
- err = -ECHILD;
- }
-
- if (!err) {
- err = handle_reval_path(nd);
- if (err)
- path_put(&nd->path);
- }
+ if (!err)
+ err = complete_walk(nd);
if (!err && nd->flags & LOOKUP_DIRECTORY) {
if (!nd->inode->i_op->lookup) {
return ERR_PTR(error);
/* fallthrough */
case LAST_ROOT:
- if (nd->flags & LOOKUP_RCU) {
- if (nameidata_drop_rcu_last(nd))
- return ERR_PTR(-ECHILD);
- }
- error = handle_reval_path(nd);
+ error = complete_walk(nd);
if (error)
- goto exit;
+ return ERR_PTR(error);
audit_inode(pathname, nd->path.dentry);
if (open_flag & O_CREAT) {
error = -EISDIR;
}
goto ok;
case LAST_BIND:
- /* can't be RCU mode here */
- error = handle_reval_path(nd);
+ error = complete_walk(nd);
if (error)
- goto exit;
+ return ERR_PTR(error);
audit_inode(pathname, dir);
goto ok;
}
if (error) /* symlink */
return NULL;
/* sayonara */
- if (nd->flags & LOOKUP_RCU) {
- if (nameidata_drop_rcu_last(nd))
- return ERR_PTR(-ECHILD);
- }
+ error = complete_walk(nd);
+ if (error)
+ return ERR_PTR(-ECHILD);
error = -ENOTDIR;
if (nd->flags & LOOKUP_DIRECTORY) {
}
/* create side of things */
-
- if (nd->flags & LOOKUP_RCU) {
- if (nameidata_drop_rcu_last(nd))
- return ERR_PTR(-ECHILD);
- }
+ error = complete_walk(nd);
+ if (error)
+ return ERR_PTR(error);
audit_inode(pathname, dir);
error = -EISDIR;
}
/*
- * We try to drop the dentry early: we should have
- * a usage count of 2 if we're the only user of this
- * dentry, and if that is true (possibly after pruning
- * the dcache), then we drop the dentry now.
+ * The dentry_unhash() helper will try to drop the dentry early: we
+ * should have a usage count of 2 if we're the only user of this
+ * dentry, and if that is true (possibly after pruning the dcache),
+ * then we drop the dentry now.
*
* A low-level filesystem can, if it choses, legally
* do a
*/
void dentry_unhash(struct dentry *dentry)
{
- dget(dentry);
shrink_dcache_parent(dentry);
spin_lock(&dentry->d_lock);
- if (dentry->d_count == 2)
+ if (dentry->d_count == 1)
__d_drop(dentry);
spin_unlock(&dentry->d_lock);
}
return -EPERM;
mutex_lock(&dentry->d_inode->i_mutex);
- dentry_unhash(dentry);
+
+ error = -EBUSY;
if (d_mountpoint(dentry))
- error = -EBUSY;
- else {
- error = security_inode_rmdir(dir, dentry);
- if (!error) {
- error = dir->i_op->rmdir(dir, dentry);
- if (!error) {
- dentry->d_inode->i_flags |= S_DEAD;
- dont_mount(dentry);
- }
- }
- }
+ goto out;
+
+ error = security_inode_rmdir(dir, dentry);
+ if (error)
+ goto out;
+
+ error = dir->i_op->rmdir(dir, dentry);
+ if (error)
+ goto out;
+
+ dentry->d_inode->i_flags |= S_DEAD;
+ dont_mount(dentry);
+
+ out:
mutex_unlock(&dentry->d_inode->i_mutex);
- if (!error) {
+ if (!error)
d_delete(dentry);
- }
- dput(dentry);
-
return error;
}
* HOWEVER, it relies on the assumption that any object with ->lookup()
* has no more than 1 dentry. If "hybrid" objects will ever appear,
* we'd better make sure that there's no link(2) for them.
- * d) some filesystems don't support opened-but-unlinked directories,
- * either because of layout or because they are not ready to deal with
- * all cases correctly. The latter will be fixed (taking this sort of
- * stuff into VFS), but the former is not going away. Solution: the same
- * trick as in rmdir().
- * e) conversion from fhandle to dentry may come in the wrong moment - when
+ * d) conversion from fhandle to dentry may come in the wrong moment - when
* we are removing the target. Solution: we will have to grab ->i_mutex
* in the fhandle_to_dentry code. [FIXME - current nfsfh.c relies on
* ->i_mutex on parents, which works but leads to some truly excessive
struct inode *new_dir, struct dentry *new_dentry)
{
int error = 0;
- struct inode *target;
+ struct inode *target = new_dentry->d_inode;
/*
* If we are going to change the parent - check write permissions,
if (error)
return error;
- target = new_dentry->d_inode;
if (target)
mutex_lock(&target->i_mutex);
- if (d_mountpoint(old_dentry)||d_mountpoint(new_dentry))
- error = -EBUSY;
- else {
- if (target)
- dentry_unhash(new_dentry);
- error = old_dir->i_op->rename(old_dir, old_dentry, new_dir, new_dentry);
- }
+
+ error = -EBUSY;
+ if (d_mountpoint(old_dentry) || d_mountpoint(new_dentry))
+ goto out;
+
+ error = old_dir->i_op->rename(old_dir, old_dentry, new_dir, new_dentry);
+ if (error)
+ goto out;
+
if (target) {
- if (!error) {
- target->i_flags |= S_DEAD;
- dont_mount(new_dentry);
- }
- mutex_unlock(&target->i_mutex);
- if (d_unhashed(new_dentry))
- d_rehash(new_dentry);
- dput(new_dentry);
+ target->i_flags |= S_DEAD;
+ dont_mount(new_dentry);
}
+ out:
+ if (target)
+ mutex_unlock(&target->i_mutex);
if (!error)
if (!(old_dir->i_sb->s_type->fs_flags & FS_RENAME_DOES_D_MOVE))
d_move(old_dentry,new_dentry);
static int vfs_rename_other(struct inode *old_dir, struct dentry *old_dentry,
struct inode *new_dir, struct dentry *new_dentry)
{
- struct inode *target;
+ struct inode *target = new_dentry->d_inode;
int error;
error = security_inode_rename(old_dir, old_dentry, new_dir, new_dentry);
return error;
dget(new_dentry);
- target = new_dentry->d_inode;
if (target)
mutex_lock(&target->i_mutex);
+
+ error = -EBUSY;
if (d_mountpoint(old_dentry)||d_mountpoint(new_dentry))
- error = -EBUSY;
- else
- error = old_dir->i_op->rename(old_dir, old_dentry, new_dir, new_dentry);
- if (!error) {
- if (target)
- dont_mount(new_dentry);
- if (!(old_dir->i_sb->s_type->fs_flags & FS_RENAME_DOES_D_MOVE))
- d_move(old_dentry, new_dentry);
- }
+ goto out;
+
+ error = old_dir->i_op->rename(old_dir, old_dentry, new_dir, new_dentry);
+ if (error)
+ goto out;
+
+ if (target)
+ dont_mount(new_dentry);
+ if (!(old_dir->i_sb->s_type->fs_flags & FS_RENAME_DOES_D_MOVE))
+ d_move(old_dentry, new_dentry);
+ out:
if (target)
mutex_unlock(&target->i_mutex);
dput(new_dentry);
ubifs_release_budget(c, &req);
else {
/* We've deleted something - clean the "no space" flags */
- c->nospace = c->nospace_rp = 0;
+ c->bi.nospace = c->bi.nospace_rp = 0;
smp_wmb();
}
return 0;
struct ubifs_inode *dir_ui = ubifs_inode(dir);
struct ubifs_budget_req req = { .mod_dent = 1, .dirtied_ino = 2 };
+ dentry_unhash(dentry);
+
/*
* Budget request settings: deletion direntry, deletion inode and
* changing the parent inode. If budgeting fails, go ahead anyway
ubifs_release_budget(c, &req);
else {
/* We've deleted something - clean the "no space" flags */
- c->nospace = c->nospace_rp = 0;
+ c->bi.nospace = c->bi.nospace_rp = 0;
smp_wmb();
}
return 0;
.dirtied_ino_d = ALIGN(old_inode_ui->data_len, 8) };
struct timespec time;
+ if (new_inode && S_ISDIR(new_inode->i_mode))
+ dentry_unhash(new_dentry);
+
/*
* Budget request settings: deletion direntry, new direntry, removing
* the old inode, and changing old and new parent directory inodes.