static int btrfs_defrag_file(struct file *file,
struct btrfs_ioctl_defrag_range_args *range)
{
struct inode *inode = fdentry(file)->d_inode;
struct btrfs_root *root = BTRFS_I(inode)->root;
struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
struct btrfs_ordered_extent *ordered;
struct page *page;
unsigned long last_index;
unsigned long ra_pages = root->fs_info->bdi.ra_pages;
unsigned long total_read = 0;
u64 page_start;
u64 page_end;
u64 last_len = 0;
u64 skip = 0;
u64 defrag_end = 0;
unsigned long i;
int ret;
if (inode->i_size == 0)
return 0;
if (range->start + range->len > range->start) {
last_index = min_t(u64, inode->i_size - 1,
range->start + range->len - 1) >> PAGE_CACHE_SHIFT;
} else {
/*
* @value: "" makes the attribute to empty, NULL removes it
*/
int __btrfs_setxattr(struct btrfs_trans_handle *trans,
struct inode *inode, const char *name,
const void *value, size_t size, int flags)
{
struct btrfs_root *root = BTRFS_I(inode)->root;
int ret;
if (trans)
return do_setxattr(trans, inode, name, value, size, flags);
trans = btrfs_start_transaction(root, 2);
if (IS_ERR(trans))
return PTR_ERR(trans);
ret = do_setxattr(trans, inode, name, value, size, flags);
if (ret)
goto out;
inode_inc_iversion(inode);
inode->i_ctime = current_fs_time(inode->i_sb);
set_bit(BTRFS_INODE_COPY_EVERYTHING, &BTRFS_I(inode)->runtime_flags);
ret = btrfs_update_inode(trans, root, inode);
BUG_ON(ret);
out:
btrfs_end_transaction(trans, root);
return ret;
}
int btrfs_getxattr(struct inode *inode, const char *name,
void *buffer, size_t size)
{
struct btrfs_dir_item *di;
struct btrfs_root *root = BTRFS_I(inode)->root;
struct btrfs_path *path;
struct extent_buffer *leaf;
int ret = 0;
unsigned long data_ptr;
path = btrfs_alloc_path();
if (!path)
return -ENOMEM;
/* lookup the xattr by name */
di = btrfs_lookup_xattr(NULL, root, path, btrfs_ino(BTRFS_I(inode)),
name, strlen(name), 0);
if (!di) {
ret = -ENODATA;
goto out;
} else if (IS_ERR(di)) {
ret = PTR_ERR(di);
goto out;
}
leaf = path->nodes[0];
/* if size is 0, that means we want the size of the attr */
if (!size) {
ret = btrfs_dir_data_len(leaf, di);
goto out;
}
/* now get the data out of our dir_item */
if (btrfs_dir_data_len(leaf, di) > size) {
ret = -ERANGE;
goto out;
}
/*
* The way things are packed into the leaf is like this
* |struct btrfs_dir_item|name|data|
* where name is the xattr name, so security.foo, and data is the
* content of the xattr. data_ptr points to the location in memory
* where the data starts in the in memory leaf
*/
data_ptr = (unsigned long)((char *)(di + 1) +
btrfs_dir_name_len(leaf, di));
read_extent_buffer(leaf, buffer, data_ptr,
btrfs_dir_data_len(leaf, di));
ret = btrfs_dir_data_len(leaf, di);
out:
btrfs_free_path(path);
return ret;
}
/*
* Create a new subvolume below @parent. This is largely modeled after
* sys_mkdirat and vfs_mkdir, but we only do a single component lookup
* inside this filesystem so it's quite a bit simpler.
*/
static noinline int btrfs_mksubvol(struct path *parent,
char *name, int namelen,
struct btrfs_root *snap_src)
{
struct inode *dir = parent->dentry->d_inode;
struct dentry *dentry;
int error;
mutex_lock_nested(&dir->i_mutex, I_MUTEX_PARENT);
dentry = lookup_one_len(name, parent->dentry, namelen);
error = PTR_ERR(dentry);
if (IS_ERR(dentry))
goto out_unlock;
error = -EEXIST;
if (dentry->d_inode)
goto out_dput;
error = mnt_want_write(parent->mnt);
if (error)
goto out_dput;
error = btrfs_may_create(dir, dentry);
if (error)
goto out_drop_write;
down_read(&BTRFS_I(dir)->root->fs_info->subvol_sem);
if (btrfs_root_refs(&BTRFS_I(dir)->root->root_item) == 0)
goto out_up_read;
if (snap_src) {
error = create_snapshot(snap_src, dentry,
name, namelen);
} else {
error = create_subvol(BTRFS_I(dir)->root, dentry,
name, namelen);
}
if (!error)
fsnotify_mkdir(dir, dentry);
out_up_read:
up_read(&BTRFS_I(dir)->root->fs_info->subvol_sem);
out_drop_write:
mnt_drop_write(parent->mnt);
out_dput:
dput(dentry);
out_unlock:
mutex_unlock(&dir->i_mutex);
return error;
}
static struct dentry *btrfs_get_parent(struct dentry *child)
{
struct inode *dir = d_inode(child);
struct btrfs_fs_info *fs_info = btrfs_sb(dir->i_sb);
struct btrfs_root *root = BTRFS_I(dir)->root;
struct btrfs_path *path;
struct extent_buffer *leaf;
struct btrfs_root_ref *ref;
struct btrfs_key key;
struct btrfs_key found_key;
int ret;
path = btrfs_alloc_path();
if (!path)
return ERR_PTR(-ENOMEM);
if (btrfs_ino(BTRFS_I(dir)) == BTRFS_FIRST_FREE_OBJECTID) {
key.objectid = root->root_key.objectid;
key.type = BTRFS_ROOT_BACKREF_KEY;
key.offset = (u64)-1;
root = fs_info->tree_root;
} else {
key.objectid = btrfs_ino(BTRFS_I(dir));
key.type = BTRFS_INODE_REF_KEY;
key.offset = (u64)-1;
}
ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
if (ret < 0)
goto fail;
BUG_ON(ret == 0); /* Key with offset of -1 found */
if (path->slots[0] == 0) {
ret = -ENOENT;
goto fail;
}
path->slots[0]--;
leaf = path->nodes[0];
btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
if (found_key.objectid != key.objectid || found_key.type != key.type) {
ret = -ENOENT;
goto fail;
}
if (found_key.type == BTRFS_ROOT_BACKREF_KEY) {
ref = btrfs_item_ptr(leaf, path->slots[0],
struct btrfs_root_ref);
key.objectid = btrfs_root_ref_dirid(leaf, ref);
} else {
static int btrfs_encode_fh(struct dentry *dentry, u32 *fh, int *max_len,
int connectable)
{
struct btrfs_fid *fid = (struct btrfs_fid *)fh;
struct inode *inode = dentry->d_inode;
int len = *max_len;
int type;
if (connectable && (len < BTRFS_FID_SIZE_CONNECTABLE)) {
*max_len = BTRFS_FID_SIZE_CONNECTABLE;
return 255;
} else if (len < BTRFS_FID_SIZE_NON_CONNECTABLE) {
*max_len = BTRFS_FID_SIZE_NON_CONNECTABLE;
return 255;
}
len = BTRFS_FID_SIZE_NON_CONNECTABLE;
type = FILEID_BTRFS_WITHOUT_PARENT;
fid->objectid = btrfs_ino(inode);
fid->root_objectid = BTRFS_I(inode)->root->objectid;
fid->gen = inode->i_generation;
if (connectable && !S_ISDIR(inode->i_mode)) {
struct inode *parent;
u64 parent_root_id;
spin_lock(&dentry->d_lock);
parent = dentry->d_parent->d_inode;
fid->parent_objectid = BTRFS_I(parent)->location.objectid;
fid->parent_gen = parent->i_generation;
parent_root_id = BTRFS_I(parent)->root->objectid;
spin_unlock(&dentry->d_lock);
if (parent_root_id != fid->root_objectid) {
fid->parent_root_objectid = parent_root_id;
len = BTRFS_FID_SIZE_CONNECTABLE_ROOT;
type = FILEID_BTRFS_WITH_PARENT_ROOT;
} else {
len = BTRFS_FID_SIZE_CONNECTABLE;
type = FILEID_BTRFS_WITH_PARENT;
}
}
*max_len = len;
return type;
}
int __btrfs_setxattr(struct btrfs_trans_handle *trans,
struct inode *inode, const char *name,
const void *value, size_t size, int flags)
{
struct btrfs_root *root = BTRFS_I(inode)->root;
int ret;
if (trans)
return do_setxattr(trans, inode, name, value, size, flags);
ret = btrfs_reserve_metadata_space(root, 2);
if (ret)
return ret;
trans = btrfs_start_transaction(root, 1);
if (!trans) {
ret = -ENOMEM;
goto out;
}
btrfs_set_trans_block_group(trans, inode);
ret = do_setxattr(trans, inode, name, value, size, flags);
if (ret)
goto out;
inode->i_ctime = CURRENT_TIME;
ret = btrfs_update_inode(trans, root, inode);
BUG_ON(ret);
out:
btrfs_end_transaction_throttle(trans, root);
btrfs_unreserve_metadata_space(root, 2);
return ret;
}
int btrfs_removexattr(struct dentry *dentry, const char *name)
{
struct btrfs_root *root = BTRFS_I(dentry->d_inode)->root;
/*
* The permission on security.* and system.* is not checked
* in permission().
*/
if (btrfs_root_readonly(root))
return -EROFS;
/*
* If this is a request for a synthetic attribute in the system.*
* namespace use the generic infrastructure to resolve a handler
* for it via sb->s_xattr.
*/
if (!strncmp(name, XATTR_SYSTEM_PREFIX, XATTR_SYSTEM_PREFIX_LEN))
return generic_removexattr(dentry, name);
if (!btrfs_is_valid_xattr(name))
return -EOPNOTSUPP;
return __btrfs_setxattr(NULL, dentry->d_inode, name, NULL, 0,
XATTR_REPLACE);
}
int btrfs_setxattr(struct dentry *dentry, const char *name, const void *value,
size_t size, int flags)
{
struct btrfs_root *root = BTRFS_I(dentry->d_inode)->root;
/*
* The permission on security.* and system.* is not checked
* in permission().
*/
if (btrfs_root_readonly(root))
return -EROFS;
/*
* If this is a request for a synthetic attribute in the system.*
* namespace use the generic infrastructure to resolve a handler
* for it via sb->s_xattr.
*/
if (!strncmp(name, XATTR_SYSTEM_PREFIX, XATTR_SYSTEM_PREFIX_LEN))
return generic_setxattr(dentry, name, value, size, flags);
if (!btrfs_is_valid_xattr(name))
return -EOPNOTSUPP;
if (!strncmp(name, XATTR_BTRFS_PREFIX, XATTR_BTRFS_PREFIX_LEN))
return btrfs_set_prop(dentry->d_inode, name,
value, size, flags);
if (size == 0)
value = ""; /* empty EA, do not remove */
return __btrfs_setxattr(NULL, dentry->d_inode, name, value, size,
flags);
}
static int btrfs_statfs(struct dentry *dentry, struct kstatfs *buf)
{
struct btrfs_root *root = btrfs_sb(dentry->d_sb);
struct btrfs_super_block *disk_super = &root->fs_info->super_copy;
int bits = dentry->d_sb->s_blocksize_bits;
__be32 *fsid = (__be32 *)root->fs_info->fsid;
buf->f_namelen = BTRFS_NAME_LEN;
buf->f_blocks = btrfs_super_total_bytes(disk_super) >> bits;
buf->f_bfree = buf->f_blocks -
(btrfs_super_bytes_used(disk_super) >> bits);
buf->f_bavail = buf->f_bfree;
buf->f_bsize = dentry->d_sb->s_blocksize;
buf->f_type = BTRFS_SUPER_MAGIC;
/* We treat it as constant endianness (it doesn't matter _which_)
because we want the fsid to come out the same whether mounted
on a big-endian or little-endian host */
buf->f_fsid.val[0] = be32_to_cpu(fsid[0]) ^ be32_to_cpu(fsid[2]);
buf->f_fsid.val[1] = be32_to_cpu(fsid[1]) ^ be32_to_cpu(fsid[3]);
/* Mask in the root object ID too, to disambiguate subvols */
buf->f_fsid.val[0] ^= BTRFS_I(dentry->d_inode)->root->objectid >> 32;
buf->f_fsid.val[1] ^= BTRFS_I(dentry->d_inode)->root->objectid;
return 0;
}
int btrfs_removexattr(struct dentry *dentry, const char *name)
{
struct btrfs_root *root = BTRFS_I(dentry->d_inode)->root;
/*
*/
if (btrfs_root_readonly(root))
return -EROFS;
/*
*/
if (!strncmp(name, XATTR_SYSTEM_PREFIX, XATTR_SYSTEM_PREFIX_LEN))
return generic_removexattr(dentry, name);
if (!btrfs_is_valid_xattr(name))
return -EOPNOTSUPP;
return __btrfs_setxattr(NULL, dentry->d_inode, name, NULL, 0,
XATTR_REPLACE);
}
int btrfs_setxattr(struct dentry *dentry, const char *name, const void *value,
size_t size, int flags)
{
struct btrfs_root *root = BTRFS_I(dentry->d_inode)->root;
/*
*/
if (btrfs_root_readonly(root))
return -EROFS;
/*
*/
if (!strncmp(name, XATTR_SYSTEM_PREFIX, XATTR_SYSTEM_PREFIX_LEN))
return generic_setxattr(dentry, name, value, size, flags);
if (!btrfs_is_valid_xattr(name))
return -EOPNOTSUPP;
if (size == 0)
value = ""; /* */
return __btrfs_setxattr(NULL, dentry->d_inode, name, value, size,
flags);
}
/*
* @value: "" makes the attribute to empty, NULL removes it
*/
int __btrfs_setxattr(struct btrfs_trans_handle *trans,
struct inode *inode, const char *name,
const void *value, size_t size, int flags)
{
struct btrfs_root *root = BTRFS_I(inode)->root;
int ret;
if (trans)
return do_setxattr(trans, inode, name, value, size, flags);
trans = btrfs_start_transaction(root, 2);
if (IS_ERR(trans))
return PTR_ERR(trans);
ret = do_setxattr(trans, inode, name, value, size, flags);
if (ret)
goto out;
inode->i_ctime = CURRENT_TIME;
ret = btrfs_update_inode(trans, root, inode);
BUG_ON(ret);
out:
btrfs_end_transaction_throttle(trans, root);
return ret;
}
/*
* after copy_from_user, pages need to be dirtied and we need to make
* sure holes are created between the current EOF and the start of
* any next extents (if required).
*
* this also makes the decision about creating an inline extent vs
* doing real data extents, marking pages dirty and delalloc as required.
*/
static noinline int dirty_and_release_pages(struct btrfs_trans_handle *trans,
struct btrfs_root *root,
struct file *file,
struct page **pages,
size_t num_pages,
loff_t pos,
size_t write_bytes)
{
int err = 0;
int i;
struct inode *inode = fdentry(file)->d_inode;
struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
u64 hint_byte;
u64 num_bytes;
u64 start_pos;
u64 end_of_last_block;
u64 end_pos = pos + write_bytes;
loff_t isize = i_size_read(inode);
start_pos = pos & ~((u64)root->sectorsize - 1);
num_bytes = (write_bytes + pos - start_pos +
root->sectorsize - 1) & ~((u64)root->sectorsize - 1);
end_of_last_block = start_pos + num_bytes - 1;
lock_extent(io_tree, start_pos, end_of_last_block, GFP_NOFS);
trans = btrfs_join_transaction(root, 1);
if (!trans) {
err = -ENOMEM;
goto out_unlock;
}
btrfs_set_trans_block_group(trans, inode);
hint_byte = 0;
set_extent_uptodate(io_tree, start_pos, end_of_last_block, GFP_NOFS);
/* check for reserved extents on each page, we don't want
* to reset the delalloc bit on things that already have
* extents reserved.
*/
btrfs_set_extent_delalloc(inode, start_pos, end_of_last_block);
for (i = 0; i < num_pages; i++) {
struct page *p = pages[i];
SetPageUptodate(p);
ClearPageChecked(p);
set_page_dirty(p);
}
if (end_pos > isize) {
i_size_write(inode, end_pos);
/* we've only changed i_size in ram, and we haven't updated
* the disk i_size. There is no need to log the inode
* at this time.
*/
}
err = btrfs_end_transaction(trans, root);
out_unlock:
unlock_extent(io_tree, start_pos, end_of_last_block, GFP_NOFS);
return err;
}
int btrfs_removexattr(struct dentry *dentry, const char *name)
{
struct btrfs_root *root = BTRFS_I(d_inode(dentry))->root;
if (btrfs_root_readonly(root))
return -EROFS;
return generic_removexattr(dentry, name);
}
/*
* Inherit flags from the parent inode.
*
* Unlike extN we don't have any flags we don't want to inherit currently.
*/
void btrfs_inherit_iflags(struct inode *inode, struct inode *dir)
{
unsigned int flags;
if (!dir)
return;
flags = BTRFS_I(dir)->flags;
if (S_ISREG(inode->i_mode))
flags &= ~BTRFS_INODE_DIRSYNC;
else if (!S_ISDIR(inode->i_mode))
flags &= (BTRFS_INODE_NODUMP | BTRFS_INODE_NOATIME);
BTRFS_I(inode)->flags = flags;
btrfs_update_iflags(inode);
}
static int do_setxattr(struct btrfs_trans_handle *trans,
struct inode *inode, const char *name,
const void *value, size_t size, int flags)
{
struct btrfs_dir_item *di;
struct btrfs_root *root = BTRFS_I(inode)->root;
struct btrfs_path *path;
size_t name_len = strlen(name);
int ret = 0;
if (name_len + size > BTRFS_MAX_XATTR_SIZE(root))
return -ENOSPC;
path = btrfs_alloc_path();
if (!path)
return -ENOMEM;
/* first lets see if we already have this xattr */
di = btrfs_lookup_xattr(trans, root, path, btrfs_ino(inode), name,
strlen(name), -1);
if (IS_ERR(di)) {
ret = PTR_ERR(di);
goto out;
}
/* ok we already have this xattr, lets remove it */
if (di) {
/* if we want create only exit */
if (flags & XATTR_CREATE) {
ret = -EEXIST;
goto out;
}
ret = btrfs_delete_one_dir_name(trans, root, path, di);
BUG_ON(ret);
btrfs_release_path(path);
/* if we don't have a value then we are removing the xattr */
if (!value)
goto out;
} else {
btrfs_release_path(path);
if (flags & XATTR_REPLACE) {
/* we couldn't find the attr to replace */
ret = -ENODATA;
goto out;
}
}
/* ok we have to create a completely new xattr */
ret = btrfs_insert_xattr_item(trans, root, path, btrfs_ino(inode),
name, name_len, value, size);
BUG_ON(ret);
out:
btrfs_free_path(path);
return ret;
}
static struct dentry *btrfs_get_parent(struct dentry *child)
{
struct inode *dir = child->d_inode;
struct btrfs_root *root = BTRFS_I(dir)->root;
struct btrfs_key key;
struct btrfs_path *path;
struct extent_buffer *leaf;
int slot;
u64 objectid;
int ret;
path = btrfs_alloc_path();
key.objectid = dir->i_ino;
btrfs_set_key_type(&key, BTRFS_INODE_REF_KEY);
key.offset = (u64)-1;
ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
if (ret < 0) {
/* Error */
btrfs_free_path(path);
return ERR_PTR(ret);
}
leaf = path->nodes[0];
slot = path->slots[0];
if (ret) {
/* btrfs_search_slot() returns the slot where we'd want to
insert a backref for parent inode #0xFFFFFFFFFFFFFFFF.
The _real_ backref, telling us what the parent inode
_actually_ is, will be in the slot _before_ the one
that btrfs_search_slot() returns. */
if (!slot) {
/* Unless there is _no_ key in the tree before... */
btrfs_free_path(path);
return ERR_PTR(-EIO);
}
slot--;
}
btrfs_item_key_to_cpu(leaf, &key, slot);
btrfs_free_path(path);
if (key.objectid != dir->i_ino || key.type != BTRFS_INODE_REF_KEY)
return ERR_PTR(-EINVAL);
objectid = key.offset;
/* If we are already at the root of a subvol, return the real root */
if (objectid == dir->i_ino)
return dget(dir->i_sb->s_root);
/* Build a new key for the inode item */
key.objectid = objectid;
btrfs_set_key_type(&key, BTRFS_INODE_ITEM_KEY);
key.offset = 0;
return d_obtain_alias(btrfs_iget(root->fs_info->sb, &key, root, NULL));
}
static int btrfs_ioctl_getflags(struct file *file, void __user *arg)
{
struct btrfs_inode *ip = BTRFS_I(file->f_path.dentry->d_inode);
unsigned int flags = btrfs_flags_to_ioctl(ip->flags);
if (copy_to_user(arg, &flags, sizeof(flags)))
return -EFAULT;
return 0;
}
int btrfs_setxattr(struct dentry *dentry, const char *name, const void *value,
size_t size, int flags)
{
struct btrfs_root *root = BTRFS_I(d_inode(dentry))->root;
if (btrfs_root_readonly(root))
return -EROFS;
return generic_setxattr(dentry, name, value, size, flags);
}
ssize_t __btrfs_getxattr(struct inode *inode, const char *name,
void *buffer, size_t size)
{
struct btrfs_dir_item *di;
struct btrfs_root *root = BTRFS_I(inode)->root;
struct btrfs_path *path;
struct extent_buffer *leaf;
int ret = 0;
unsigned long data_ptr;
path = btrfs_alloc_path();
if (!path)
return -ENOMEM;
/* */
di = btrfs_lookup_xattr(NULL, root, path, btrfs_ino(inode), name,
strlen(name), 0);
if (!di) {
ret = -ENODATA;
goto out;
} else if (IS_ERR(di)) {
ret = PTR_ERR(di);
goto out;
}
leaf = path->nodes[0];
/* */
if (!size) {
ret = btrfs_dir_data_len(leaf, di);
goto out;
}
/* */
if (btrfs_dir_data_len(leaf, di) > size) {
ret = -ERANGE;
goto out;
}
/*
*/
data_ptr = (unsigned long)((char *)(di + 1) +
btrfs_dir_name_len(leaf, di));
read_extent_buffer(leaf, buffer, data_ptr,
btrfs_dir_data_len(leaf, di));
ret = btrfs_dir_data_len(leaf, di);
out:
btrfs_free_path(path);
return ret;
}
static int btrfs_encode_fh(struct inode *inode, u32 *fh, int *max_len,
struct inode *parent)
{
struct btrfs_fid *fid = (struct btrfs_fid *)fh;
int len = *max_len;
int type;
if (parent && (len < BTRFS_FID_SIZE_CONNECTABLE)) {
*max_len = BTRFS_FID_SIZE_CONNECTABLE;
return FILEID_INVALID;
} else if (len < BTRFS_FID_SIZE_NON_CONNECTABLE) {
*max_len = BTRFS_FID_SIZE_NON_CONNECTABLE;
return FILEID_INVALID;
}
len = BTRFS_FID_SIZE_NON_CONNECTABLE;
type = FILEID_BTRFS_WITHOUT_PARENT;
fid->objectid = btrfs_ino(BTRFS_I(inode));
fid->root_objectid = BTRFS_I(inode)->root->objectid;
fid->gen = inode->i_generation;
if (parent) {
u64 parent_root_id;
fid->parent_objectid = BTRFS_I(parent)->location.objectid;
fid->parent_gen = parent->i_generation;
parent_root_id = BTRFS_I(parent)->root->objectid;
if (parent_root_id != fid->root_objectid) {
fid->parent_root_objectid = parent_root_id;
len = BTRFS_FID_SIZE_CONNECTABLE_ROOT;
type = FILEID_BTRFS_WITH_PARENT_ROOT;
} else {
len = BTRFS_FID_SIZE_CONNECTABLE;
type = FILEID_BTRFS_WITH_PARENT;
}
}
*max_len = len;
return type;
}
/* pop a record for an inode into the defrag tree. The lock
* must be held already
*
* If you're inserting a record for an older transid than an
* existing record, the transid already in the tree is lowered
*
* If an existing record is found the defrag item you
* pass in is freed
*/
static void __btrfs_add_inode_defrag(struct inode *inode,
struct inode_defrag *defrag)
{
struct btrfs_root *root = BTRFS_I(inode)->root;
struct inode_defrag *entry;
struct rb_node **p;
struct rb_node *parent = NULL;
int ret;
p = &root->fs_info->defrag_inodes.rb_node;
while (*p) {
parent = *p;
entry = rb_entry(parent, struct inode_defrag, rb_node);
ret = __compare_inode_defrag(defrag, entry);
if (ret < 0)
p = &parent->rb_left;
else if (ret > 0)
p = &parent->rb_right;
else {
/* if we're reinserting an entry for
* an old defrag run, make sure to
* lower the transid of our existing record
*/
if (defrag->transid < entry->transid)
entry->transid = defrag->transid;
if (defrag->last_offset > entry->last_offset)
entry->last_offset = defrag->last_offset;
goto exists;
}
}
set_bit(BTRFS_INODE_IN_DEFRAG, &BTRFS_I(inode)->runtime_flags);
rb_link_node(&defrag->rb_node, parent, p);
rb_insert_color(&defrag->rb_node, &root->fs_info->defrag_inodes);
return;
exists:
kfree(defrag);
return;
}
/*
* insert a defrag record for this inode if auto defrag is
* enabled
*/
int btrfs_add_inode_defrag(struct btrfs_trans_handle *trans,
struct inode *inode)
{
struct btrfs_root *root = BTRFS_I(inode)->root;
struct inode_defrag *defrag;
u64 transid;
if (!btrfs_test_opt(root, AUTO_DEFRAG))
return 0;
if (btrfs_fs_closing(root->fs_info))
return 0;
if (test_bit(BTRFS_INODE_IN_DEFRAG, &BTRFS_I(inode)->runtime_flags))
return 0;
if (trans)
transid = trans->transid;
else
transid = BTRFS_I(inode)->root->last_trans;
defrag = kzalloc(sizeof(*defrag), GFP_NOFS);
if (!defrag)
return -ENOMEM;
defrag->ino = btrfs_ino(inode);
defrag->transid = transid;
defrag->root = root->root_key.objectid;
spin_lock(&root->fs_info->defrag_inodes_lock);
if (!test_bit(BTRFS_INODE_IN_DEFRAG, &BTRFS_I(inode)->runtime_flags))
__btrfs_add_inode_defrag(inode, defrag);
else
kfree(defrag);
spin_unlock(&root->fs_info->defrag_inodes_lock);
return 0;
}
static noinline int finish_pending_snapshot(struct btrfs_fs_info *fs_info,
struct btrfs_pending_snapshot *pending)
{
int ret;
int namelen;
u64 index = 0;
struct btrfs_trans_handle *trans;
struct inode *parent_inode;
struct inode *inode;
struct btrfs_root *parent_root;
parent_inode = pending->dentry->d_parent->d_inode;
parent_root = BTRFS_I(parent_inode)->root;
trans = btrfs_join_transaction(parent_root, 1);
/*
* insert the directory item
*/
namelen = strlen(pending->name);
ret = btrfs_set_inode_index(parent_inode, &index);
ret = btrfs_insert_dir_item(trans, parent_root,
pending->name, namelen,
parent_inode->i_ino,
&pending->root_key, BTRFS_FT_DIR, index);
if (ret)
goto fail;
btrfs_i_size_write(parent_inode, parent_inode->i_size + namelen * 2);
ret = btrfs_update_inode(trans, parent_root, parent_inode);
BUG_ON(ret);
ret = btrfs_add_root_ref(trans, parent_root->fs_info->tree_root,
pending->root_key.objectid,
parent_root->root_key.objectid,
parent_inode->i_ino, index, pending->name,
namelen);
BUG_ON(ret);
inode = btrfs_lookup_dentry(parent_inode, pending->dentry);
d_instantiate(pending->dentry, inode);
fail:
btrfs_end_transaction(trans, fs_info->fs_root);
return ret;
}
/*
* Update inode->i_flags based on the btrfs internal flags.
*/
void btrfs_update_iflags(struct inode *inode)
{
struct btrfs_inode *ip = BTRFS_I(inode);
inode->i_flags &= ~(S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC);
if (ip->flags & BTRFS_INODE_SYNC)
inode->i_flags |= S_SYNC;
if (ip->flags & BTRFS_INODE_IMMUTABLE)
inode->i_flags |= S_IMMUTABLE;
if (ip->flags & BTRFS_INODE_APPEND)
inode->i_flags |= S_APPEND;
if (ip->flags & BTRFS_INODE_NOATIME)
inode->i_flags |= S_NOATIME;
if (ip->flags & BTRFS_INODE_DIRSYNC)
inode->i_flags |= S_DIRSYNC;
}
int btrfs_dedup_search(struct inode *inode, u64 file_pos,
struct btrfs_dedup_hash *hash)
{
struct btrfs_fs_info *fs_info = BTRFS_I(inode)->root->fs_info;
struct btrfs_dedup_info *dedup_info = fs_info->dedup_info;
int ret = 0;
if (WARN_ON(!dedup_info || !hash))
return 0;
if (dedup_info->backend < BTRFS_DEDUP_BACKEND_LAST) {
ret = generic_search(inode, file_pos, hash);
if (ret == 0) {
hash->num_bytes = 0;
hash->bytenr = 0;
}
return ret;
}
return -EINVAL;
}
static int generic_search(struct inode *inode, u64 file_pos,
struct btrfs_dedup_hash *hash)
{
int ret;
struct btrfs_root *root = BTRFS_I(inode)->root;
struct btrfs_fs_info *fs_info = root->fs_info;
struct btrfs_trans_handle *trans;
struct btrfs_delayed_ref_root *delayed_refs;
struct btrfs_delayed_ref_head *head;
struct btrfs_dedup_info *dedup_info = fs_info->dedup_info;
u64 bytenr;
u64 tmp_bytenr;
u32 num_bytes;
trans = btrfs_join_transaction(root);
if (IS_ERR(trans))
return PTR_ERR(trans);
again:
mutex_lock(&dedup_info->lock);
ret = generic_search_hash(dedup_info, hash->hash, &bytenr, &num_bytes);
if (ret <= 0)
goto out;
delayed_refs = &trans->transaction->delayed_refs;
spin_lock(&delayed_refs->lock);
head = btrfs_find_delayed_ref_head(trans, bytenr);
if (!head) {
/*
* We can safely insert a new delayed_ref as long as we
* hold delayed_refs->lock.
* Only need to use atomic inc_extent_ref()
*/
ret = btrfs_inc_extent_ref_atomic(trans, root, bytenr,
num_bytes, 0, root->root_key.objectid,
btrfs_ino(inode), file_pos);
spin_unlock(&delayed_refs->lock);
if (ret == 0) {
hash->bytenr = bytenr;
hash->num_bytes = num_bytes;
ret = 1;
}
goto out;
}
/*
* We can't lock ref head with dedup_info->lock hold or we will cause
* ABBA dead lock.
*/
mutex_unlock(&dedup_info->lock);
ret = btrfs_delayed_ref_lock(trans, head);
spin_unlock(&delayed_refs->lock);
if (ret == -EAGAIN)
goto again;
mutex_lock(&dedup_info->lock);
/*
* Search again to ensure the hash is still here and bytenr didn't
* change
*/
ret = generic_search_hash(dedup_info, hash->hash, &tmp_bytenr,
&num_bytes);
if (ret <= 0) {
mutex_unlock(&head->mutex);
goto out;
}
if (tmp_bytenr != bytenr) {
mutex_unlock(&head->mutex);
mutex_unlock(&dedup_info->lock);
goto again;
}
hash->bytenr = bytenr;
hash->num_bytes = num_bytes;
/*
* Increase the extent ref right now, to avoid delayed ref run
* Or we may increase ref on non-exist extent.
*/
btrfs_inc_extent_ref(trans, root, bytenr, num_bytes, 0,
root->root_key.objectid,
btrfs_ino(inode), file_pos);
mutex_unlock(&head->mutex);
out:
mutex_unlock(&dedup_info->lock);
btrfs_end_transaction(trans, root);
return ret;
}
static int do_setxattr(struct btrfs_trans_handle *trans,
struct inode *inode, const char *name,
const void *value, size_t size, int flags)
{
struct btrfs_dir_item *di;
struct btrfs_root *root = BTRFS_I(inode)->root;
struct btrfs_path *path;
size_t name_len = strlen(name);
int ret = 0;
if (name_len + size > BTRFS_MAX_XATTR_SIZE(root))
return -ENOSPC;
path = btrfs_alloc_path();
if (!path)
return -ENOMEM;
if (flags & XATTR_REPLACE) {
di = btrfs_lookup_xattr(trans, root, path, btrfs_ino(inode), name,
name_len, -1);
if (IS_ERR(di)) {
ret = PTR_ERR(di);
goto out;
} else if (!di) {
ret = -ENODATA;
goto out;
}
ret = btrfs_delete_one_dir_name(trans, root, path, di);
if (ret)
goto out;
btrfs_release_path(path);
/*
* remove the attribute
*/
if (!value)
goto out;
}
again:
ret = btrfs_insert_xattr_item(trans, root, path, btrfs_ino(inode),
name, name_len, value, size);
if (ret == -EEXIST) {
if (flags & XATTR_CREATE)
goto out;
/*
* We can't use the path we already have since we won't have the
* proper locking for a delete, so release the path and
* re-lookup to delete the thing.
*/
btrfs_release_path(path);
di = btrfs_lookup_xattr(trans, root, path, btrfs_ino(inode),
name, name_len, -1);
if (IS_ERR(di)) {
ret = PTR_ERR(di);
goto out;
} else if (!di) {
/* Shouldn't happen but just in case... */
btrfs_release_path(path);
goto again;
}
ret = btrfs_delete_one_dir_name(trans, root, path, di);
if (ret)
goto out;
/*
* We have a value to set, so go back and try to insert it now.
*/
if (value) {
btrfs_release_path(path);
goto again;
}
}
out:
btrfs_free_path(path);
return ret;
}
ssize_t btrfs_listxattr(struct dentry *dentry, char *buffer, size_t size)
{
struct btrfs_key key, found_key;
struct inode *inode = dentry->d_inode;
struct btrfs_root *root = BTRFS_I(inode)->root;
struct btrfs_path *path;
struct extent_buffer *leaf;
struct btrfs_dir_item *di;
int ret = 0, slot;
size_t total_size = 0, size_left = size;
unsigned long name_ptr;
size_t name_len;
/*
* ok we want all objects associated with this id.
* NOTE: we set key.offset = 0; because we want to start with the
* first xattr that we find and walk forward
*/
key.objectid = btrfs_ino(inode);
btrfs_set_key_type(&key, BTRFS_XATTR_ITEM_KEY);
key.offset = 0;
path = btrfs_alloc_path();
if (!path)
return -ENOMEM;
path->reada = 2;
/* search for our xattrs */
ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
if (ret < 0)
goto err;
while (1) {
leaf = path->nodes[0];
slot = path->slots[0];
/* this is where we start walking through the path */
if (slot >= btrfs_header_nritems(leaf)) {
/*
* if we've reached the last slot in this leaf we need
* to go to the next leaf and reset everything
*/
ret = btrfs_next_leaf(root, path);
if (ret < 0)
goto err;
else if (ret > 0)
break;
continue;
}
btrfs_item_key_to_cpu(leaf, &found_key, slot);
/* check to make sure this item is what we want */
if (found_key.objectid != key.objectid)
break;
if (btrfs_key_type(&found_key) != BTRFS_XATTR_ITEM_KEY)
break;
di = btrfs_item_ptr(leaf, slot, struct btrfs_dir_item);
if (verify_dir_item(root, leaf, di))
continue;
name_len = btrfs_dir_name_len(leaf, di);
total_size += name_len + 1;
/* we are just looking for how big our buffer needs to be */
if (!size)
goto next;
if (!buffer || (name_len + 1) > size_left) {
ret = -ERANGE;
goto err;
}
name_ptr = (unsigned long)(di + 1);
read_extent_buffer(leaf, buffer, name_ptr, name_len);
buffer[name_len] = '\0';
size_left -= name_len + 1;
buffer += name_len + 1;
next:
path->slots[0]++;
}
ret = total_size;
err:
btrfs_free_path(path);
return ret;
}
