#include <linux/time.h>
#include <linux/init.h>
#include <linux/string.h>
-#include <linux/smp_lock.h>
#include <linux/backing-dev.h>
#include <linux/mpage.h>
#include <linux/swap.h>
#include "ordered-data.h"
#include "xattr.h"
#include "tree-log.h"
-#include "ref-cache.h"
#include "compression.h"
#include "locking.h"
struct btrfs_root *root;
};
-static struct inode_operations btrfs_dir_inode_operations;
-static struct inode_operations btrfs_symlink_inode_operations;
-static struct inode_operations btrfs_dir_ro_inode_operations;
-static struct inode_operations btrfs_special_inode_operations;
-static struct inode_operations btrfs_file_inode_operations;
-static struct address_space_operations btrfs_aops;
-static struct address_space_operations btrfs_symlink_aops;
-static struct file_operations btrfs_dir_file_operations;
+static const struct inode_operations btrfs_dir_inode_operations;
+static const struct inode_operations btrfs_symlink_inode_operations;
+static const struct inode_operations btrfs_dir_ro_inode_operations;
+static const struct inode_operations btrfs_special_inode_operations;
+static const struct inode_operations btrfs_file_inode_operations;
+static const struct address_space_operations btrfs_aops;
+static const struct address_space_operations btrfs_symlink_aops;
+static const struct file_operations btrfs_dir_file_operations;
static struct extent_io_ops btrfs_extent_io_ops;
static struct kmem_cache *btrfs_inode_cachep;
struct kmem_cache *btrfs_trans_handle_cachep;
struct kmem_cache *btrfs_transaction_cachep;
-struct kmem_cache *btrfs_bit_radix_cachep;
struct kmem_cache *btrfs_path_cachep;
#define S_SHIFT 12
}
ret = btrfs_drop_extents(trans, root, inode, start,
- aligned_end, start, &hint_byte);
+ aligned_end, aligned_end, start,
+ &hint_byte, 1);
BUG_ON(ret);
if (isize > actual_end)
inline_len, compressed_size,
compressed_pages);
BUG_ON(ret);
- btrfs_drop_extent_cache(inode, start, aligned_end, 0);
+ btrfs_drop_extent_cache(inode, start, aligned_end - 1, 0);
return 0;
}
* inode has not been flagged as nocompress. This flag can
* change at any time if we discover bad compression ratios.
*/
- if (!btrfs_test_flag(inode, NOCOMPRESS) &&
+ if (!(BTRFS_I(inode)->flags & BTRFS_INODE_NOCOMPRESS) &&
btrfs_test_opt(root, COMPRESS)) {
WARN_ON(pages);
pages = kzalloc(sizeof(struct page *) * nr_pages, GFP_NOFS);
extent_clear_unlock_delalloc(inode,
&BTRFS_I(inode)->io_tree,
start, end, NULL, 1, 0,
- 0, 1, 1, 1);
+ 0, 1, 1, 1, 0);
ret = 0;
goto free_pages_out;
}
nr_pages_ret = 0;
/* flag the file so we don't compress in the future */
- btrfs_set_flag(inode, NOCOMPRESS);
+ BTRFS_I(inode)->flags |= BTRFS_INODE_NOCOMPRESS;
}
if (will_compress) {
*num_added += 1;
set_bit(EXTENT_FLAG_COMPRESSED, &em->flags);
while (1) {
- spin_lock(&em_tree->lock);
+ write_lock(&em_tree->lock);
ret = add_extent_mapping(em_tree, em);
- spin_unlock(&em_tree->lock);
+ write_unlock(&em_tree->lock);
if (ret != -EEXIST) {
free_extent_map(em);
break;
async_extent->start,
async_extent->start +
async_extent->ram_size - 1,
- NULL, 1, 1, 0, 1, 1, 0);
+ NULL, 1, 1, 0, 1, 1, 0, 0);
ret = btrfs_submit_compressed_write(inode,
async_extent->start,
extent_clear_unlock_delalloc(inode,
&BTRFS_I(inode)->io_tree,
start, end, NULL, 1, 1,
- 1, 1, 1, 1);
+ 1, 1, 1, 1, 0);
*nr_written = *nr_written +
(end - start + PAGE_CACHE_SIZE) / PAGE_CACHE_SIZE;
*page_started = 1;
BUG_ON(disk_num_bytes >
btrfs_super_total_bytes(&root->fs_info->super_copy));
+
+ read_lock(&BTRFS_I(inode)->extent_tree.lock);
+ em = search_extent_mapping(&BTRFS_I(inode)->extent_tree,
+ start, num_bytes);
+ if (em) {
+ alloc_hint = em->block_start;
+ free_extent_map(em);
+ }
+ read_unlock(&BTRFS_I(inode)->extent_tree.lock);
btrfs_drop_extent_cache(inode, start, start + num_bytes - 1, 0);
while (disk_num_bytes > 0) {
em = alloc_extent_map(GFP_NOFS);
em->start = start;
em->orig_start = em->start;
-
ram_size = ins.offset;
em->len = ins.offset;
set_bit(EXTENT_FLAG_PINNED, &em->flags);
while (1) {
- spin_lock(&em_tree->lock);
+ write_lock(&em_tree->lock);
ret = add_extent_mapping(em_tree, em);
- spin_unlock(&em_tree->lock);
+ write_unlock(&em_tree->lock);
if (ret != -EEXIST) {
free_extent_map(em);
break;
/* we're not doing compressed IO, don't unlock the first
* page (which the caller expects to stay locked), don't
* clear any dirty bits and don't set any writeback bits
+ *
+ * Do set the Private2 bit so we know this page was properly
+ * setup for writepage
*/
extent_clear_unlock_delalloc(inode, &BTRFS_I(inode)->io_tree,
start, start + ram_size - 1,
locked_page, unlock, 1,
- 1, 0, 0, 0);
+ 1, 0, 0, 0, 1);
disk_num_bytes -= cur_alloc_size;
num_bytes -= cur_alloc_size;
alloc_hint = ins.objectid + ins.offset;
int limit = 10 * 1024 * 1042;
clear_extent_bit(&BTRFS_I(inode)->io_tree, start, end, EXTENT_LOCKED |
- EXTENT_DELALLOC, 1, 0, GFP_NOFS);
+ EXTENT_DELALLOC, 1, 0, NULL, GFP_NOFS);
while (start < end) {
async_cow = kmalloc(sizeof(*async_cow), GFP_NOFS);
async_cow->inode = inode;
async_cow->locked_page = locked_page;
async_cow->start = start;
- if (btrfs_test_flag(inode, NOCOMPRESS))
+ if (BTRFS_I(inode)->flags & BTRFS_INODE_NOCOMPRESS)
cur_end = end;
else
cur_end = min(end, start + 512 * 1024 - 1);
u64 cow_start;
u64 cur_offset;
u64 extent_end;
+ u64 extent_offset;
u64 disk_bytenr;
u64 num_bytes;
int extent_type;
if (extent_type == BTRFS_FILE_EXTENT_REG ||
extent_type == BTRFS_FILE_EXTENT_PREALLOC) {
disk_bytenr = btrfs_file_extent_disk_bytenr(leaf, fi);
+ extent_offset = btrfs_file_extent_offset(leaf, fi);
extent_end = found_key.offset +
btrfs_file_extent_num_bytes(leaf, fi);
if (extent_end <= start) {
if (btrfs_extent_readonly(root, disk_bytenr))
goto out_check;
if (btrfs_cross_ref_exist(trans, root, inode->i_ino,
- disk_bytenr))
+ found_key.offset -
+ extent_offset, disk_bytenr))
goto out_check;
- disk_bytenr += btrfs_file_extent_offset(leaf, fi);
+ disk_bytenr += extent_offset;
disk_bytenr += cur_offset - found_key.offset;
num_bytes = min(end + 1, extent_end) - cur_offset;
/*
em->bdev = root->fs_info->fs_devices->latest_bdev;
set_bit(EXTENT_FLAG_PINNED, &em->flags);
while (1) {
- spin_lock(&em_tree->lock);
+ write_lock(&em_tree->lock);
ret = add_extent_mapping(em_tree, em);
- spin_unlock(&em_tree->lock);
+ write_unlock(&em_tree->lock);
if (ret != -EEXIST) {
free_extent_map(em);
break;
extent_clear_unlock_delalloc(inode, &BTRFS_I(inode)->io_tree,
cur_offset, cur_offset + num_bytes - 1,
- locked_page, 1, 1, 1, 0, 0, 0);
+ locked_page, 1, 1, 1, 0, 0, 0, 1);
cur_offset = extent_end;
if (cur_offset > end)
break;
int ret;
struct btrfs_root *root = BTRFS_I(inode)->root;
- if (btrfs_test_flag(inode, NODATACOW))
+ if (BTRFS_I(inode)->flags & BTRFS_INODE_NODATACOW)
ret = run_delalloc_nocow(inode, locked_page, start, end,
page_started, 1, nr_written);
- else if (btrfs_test_flag(inode, PREALLOC))
+ else if (BTRFS_I(inode)->flags & BTRFS_INODE_PREALLOC)
ret = run_delalloc_nocow(inode, locked_page, start, end,
page_started, 0, nr_written);
else if (!btrfs_test_opt(root, COMPRESS))
int ret = 0;
int skip_sum;
- skip_sum = btrfs_test_flag(inode, NODATASUM);
+ skip_sum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;
ret = btrfs_bio_wq_end_io(root->fs_info, bio, 0);
BUG_ON(ret);
lock_extent(&BTRFS_I(inode)->io_tree, page_start, page_end, GFP_NOFS);
/* already ordered? We're done */
- if (test_range_bit(&BTRFS_I(inode)->io_tree, page_start, page_end,
- EXTENT_ORDERED, 0)) {
+ if (PagePrivate2(page))
goto out;
- }
ordered = btrfs_lookup_ordered_extent(inode, page_start);
if (ordered) {
struct inode *inode = page->mapping->host;
struct btrfs_writepage_fixup *fixup;
struct btrfs_root *root = BTRFS_I(inode)->root;
- int ret;
- ret = test_range_bit(&BTRFS_I(inode)->io_tree, start, end,
- EXTENT_ORDERED, 0);
- if (ret)
+ /* this page is properly in the ordered list */
+ if (TestClearPagePrivate2(page))
return 0;
if (PageChecked(page))
struct inode *inode, u64 file_pos,
u64 disk_bytenr, u64 disk_num_bytes,
u64 num_bytes, u64 ram_bytes,
+ u64 locked_end,
u8 compression, u8 encryption,
u16 other_encoding, int extent_type)
{
BUG_ON(!path);
path->leave_spinning = 1;
+
+ /*
+ * we may be replacing one extent in the tree with another.
+ * The new extent is pinned in the extent map, and we don't want
+ * to drop it from the cache until it is completely in the btree.
+ *
+ * So, tell btrfs_drop_extents to leave this extent in the cache.
+ * the caller is expected to unpin it and allow it to be merged
+ * with the others.
+ */
ret = btrfs_drop_extents(trans, root, inode, file_pos,
- file_pos + num_bytes, file_pos, &hint);
+ file_pos + num_bytes, locked_end,
+ file_pos, &hint, 0);
BUG_ON(ret);
ins.objectid = inode->i_ino;
btrfs_mark_buffer_dirty(leaf);
inode_add_bytes(inode, num_bytes);
- btrfs_drop_extent_cache(inode, file_pos, file_pos + num_bytes - 1, 0);
ins.objectid = disk_bytenr;
ins.offset = disk_num_bytes;
ins.type = BTRFS_EXTENT_ITEM_KEY;
- ret = btrfs_alloc_reserved_extent(trans, root, leaf->start,
- root->root_key.objectid,
- trans->transid, inode->i_ino, &ins);
+ ret = btrfs_alloc_reserved_file_extent(trans, root,
+ root->root_key.objectid,
+ inode->i_ino, file_pos, &ins);
BUG_ON(ret);
btrfs_free_path(path);
ordered_extent->disk_len,
ordered_extent->len,
ordered_extent->len,
+ ordered_extent->file_offset +
+ ordered_extent->len,
compressed, 0, 0,
BTRFS_FILE_EXTENT_REG);
+ unpin_extent_cache(&BTRFS_I(inode)->extent_tree,
+ ordered_extent->file_offset,
+ ordered_extent->len);
BUG_ON(ret);
}
unlock_extent(io_tree, ordered_extent->file_offset,
static int btrfs_writepage_end_io_hook(struct page *page, u64 start, u64 end,
struct extent_state *state, int uptodate)
{
+ ClearPagePrivate2(page);
return btrfs_finish_ordered_io(page->mapping->host, start, end);
}
failrec->last_mirror = 0;
failrec->bio_flags = 0;
- spin_lock(&em_tree->lock);
+ read_lock(&em_tree->lock);
em = lookup_extent_mapping(em_tree, start, failrec->len);
if (em->start > start || em->start + em->len < start) {
free_extent_map(em);
em = NULL;
}
- spin_unlock(&em_tree->lock);
+ read_unlock(&em_tree->lock);
if (!em || IS_ERR(em)) {
kfree(failrec);
ClearPageChecked(page);
goto good;
}
- if (btrfs_test_flag(inode, NODATASUM))
+
+ if (BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM)
return 0;
if (root->root_key.objectid == BTRFS_DATA_RELOC_TREE_OBJECTID &&
- test_range_bit(io_tree, start, end, EXTENT_NODATASUM, 1)) {
+ test_range_bit(io_tree, start, end, EXTENT_NODATASUM, 1, NULL)) {
clear_extent_bits(io_tree, start, end, EXTENT_NODATASUM,
GFP_NOFS);
return 0;
return 0;
zeroit:
- printk(KERN_INFO "btrfs csum failed ino %lu off %llu csum %u "
- "private %llu\n", page->mapping->host->i_ino,
- (unsigned long long)start, csum,
- (unsigned long long)private);
+ if (printk_ratelimit()) {
+ printk(KERN_INFO "btrfs csum failed ino %lu off %llu csum %u "
+ "private %llu\n", page->mapping->host->i_ino,
+ (unsigned long long)start, csum,
+ (unsigned long long)private);
+ }
memset(kaddr + offset, 1, end - start + 1);
flush_dcache_page(page);
kunmap_atomic(kaddr, KM_USER0);
* crossing root thing. we store the inode number in the
* offset of the orphan item.
*/
- inode = btrfs_iget_locked(root->fs_info->sb,
- found_key.offset, root);
- if (!inode)
+ found_key.objectid = found_key.offset;
+ found_key.type = BTRFS_INODE_ITEM_KEY;
+ found_key.offset = 0;
+ inode = btrfs_iget(root->fs_info->sb, &found_key, root);
+ if (IS_ERR(inode))
break;
- if (inode->i_state & I_NEW) {
- BTRFS_I(inode)->root = root;
-
- /* have to set the location manually */
- BTRFS_I(inode)->location.objectid = inode->i_ino;
- BTRFS_I(inode)->location.type = BTRFS_INODE_ITEM_KEY;
- BTRFS_I(inode)->location.offset = 0;
-
- btrfs_read_locked_inode(inode);
- unlock_new_inode(inode);
- }
-
/*
* add this inode to the orphan list so btrfs_orphan_del does
* the proper thing when we hit it
}
/*
+ * very simple check to peek ahead in the leaf looking for xattrs. If we
+ * don't find any xattrs, we know there can't be any acls.
+ *
+ * slot is the slot the inode is in, objectid is the objectid of the inode
+ */
+static noinline int acls_after_inode_item(struct extent_buffer *leaf,
+ int slot, u64 objectid)
+{
+ u32 nritems = btrfs_header_nritems(leaf);
+ struct btrfs_key found_key;
+ int scanned = 0;
+
+ slot++;
+ while (slot < nritems) {
+ btrfs_item_key_to_cpu(leaf, &found_key, slot);
+
+ /* we found a different objectid, there must not be acls */
+ if (found_key.objectid != objectid)
+ return 0;
+
+ /* we found an xattr, assume we've got an acl */
+ if (found_key.type == BTRFS_XATTR_ITEM_KEY)
+ return 1;
+
+ /*
+ * we found a key greater than an xattr key, there can't
+ * be any acls later on
+ */
+ if (found_key.type > BTRFS_XATTR_ITEM_KEY)
+ return 0;
+
+ slot++;
+ scanned++;
+
+ /*
+ * it goes inode, inode backrefs, xattrs, extents,
+ * so if there are a ton of hard links to an inode there can
+ * be a lot of backrefs. Don't waste time searching too hard,
+ * this is just an optimization
+ */
+ if (scanned >= 8)
+ break;
+ }
+ /* we hit the end of the leaf before we found an xattr or
+ * something larger than an xattr. We have to assume the inode
+ * has acls
+ */
+ return 1;
+}
+
+/*
* read an inode from the btree into the in-memory inode
*/
-void btrfs_read_locked_inode(struct inode *inode)
+static void btrfs_read_locked_inode(struct inode *inode)
{
struct btrfs_path *path;
struct extent_buffer *leaf;
struct btrfs_timespec *tspec;
struct btrfs_root *root = BTRFS_I(inode)->root;
struct btrfs_key location;
+ int maybe_acls;
u64 alloc_group_block;
u32 rdev;
int ret;
alloc_group_block = btrfs_inode_block_group(leaf, inode_item);
+ /*
+ * try to precache a NULL acl entry for files that don't have
+ * any xattrs or acls
+ */
+ maybe_acls = acls_after_inode_item(leaf, path->slots[0], inode->i_ino);
+ if (!maybe_acls)
+ cache_no_acl(inode);
+
BTRFS_I(inode)->block_group = btrfs_find_block_group(root, 0,
alloc_group_block, 0);
btrfs_free_path(path);
init_special_inode(inode, inode->i_mode, rdev);
break;
}
+
+ btrfs_update_iflags(inode);
return;
make_bad:
btrfs_update_inode(trans, root, dir);
btrfs_drop_nlink(inode);
ret = btrfs_update_inode(trans, root, inode);
- dir->i_sb->s_dirt = 1;
out:
return ret;
}
return ret;
}
+int btrfs_unlink_subvol(struct btrfs_trans_handle *trans,
+ struct btrfs_root *root,
+ struct inode *dir, u64 objectid,
+ const char *name, int name_len)
+{
+ struct btrfs_path *path;
+ struct extent_buffer *leaf;
+ struct btrfs_dir_item *di;
+ struct btrfs_key key;
+ u64 index;
+ int ret;
+
+ path = btrfs_alloc_path();
+ if (!path)
+ return -ENOMEM;
+
+ di = btrfs_lookup_dir_item(trans, root, path, dir->i_ino,
+ name, name_len, -1);
+ BUG_ON(!di || IS_ERR(di));
+
+ leaf = path->nodes[0];
+ btrfs_dir_item_key_to_cpu(leaf, di, &key);
+ WARN_ON(key.type != BTRFS_ROOT_ITEM_KEY || key.objectid != objectid);
+ ret = btrfs_delete_one_dir_name(trans, root, path, di);
+ BUG_ON(ret);
+ btrfs_release_path(root, path);
+
+ ret = btrfs_del_root_ref(trans, root->fs_info->tree_root,
+ objectid, root->root_key.objectid,
+ dir->i_ino, &index, name, name_len);
+ if (ret < 0) {
+ BUG_ON(ret != -ENOENT);
+ di = btrfs_search_dir_index_item(root, path, dir->i_ino,
+ name, name_len);
+ BUG_ON(!di || IS_ERR(di));
+
+ leaf = path->nodes[0];
+ btrfs_item_key_to_cpu(leaf, &key, path->slots[0]);
+ btrfs_release_path(root, path);
+ index = key.offset;
+ }
+
+ di = btrfs_lookup_dir_index_item(trans, root, path, dir->i_ino,
+ index, name, name_len, -1);
+ BUG_ON(!di || IS_ERR(di));
+
+ leaf = path->nodes[0];
+ btrfs_dir_item_key_to_cpu(leaf, di, &key);
+ WARN_ON(key.type != BTRFS_ROOT_ITEM_KEY || key.objectid != objectid);
+ ret = btrfs_delete_one_dir_name(trans, root, path, di);
+ BUG_ON(ret);
+ btrfs_release_path(root, path);
+
+ btrfs_i_size_write(dir, dir->i_size - name_len * 2);
+ dir->i_mtime = dir->i_ctime = CURRENT_TIME;
+ ret = btrfs_update_inode(trans, root, dir);
+ BUG_ON(ret);
+ dir->i_sb->s_dirt = 1;
+
+ btrfs_free_path(path);
+ return 0;
+}
+
static int btrfs_rmdir(struct inode *dir, struct dentry *dentry)
{
struct inode *inode = dentry->d_inode;
struct btrfs_trans_handle *trans;
unsigned long nr = 0;
- /*
- * the FIRST_FREE_OBJECTID check makes sure we don't try to rmdir
- * the root of a subvolume or snapshot
- */
if (inode->i_size > BTRFS_EMPTY_DIR_SIZE ||
- inode->i_ino == BTRFS_FIRST_FREE_OBJECTID) {
+ inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)
return -ENOTEMPTY;
- }
trans = btrfs_start_transaction(root, 1);
btrfs_set_trans_block_group(trans, dir);
+ if (unlikely(inode->i_ino == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID)) {
+ err = btrfs_unlink_subvol(trans, root, dir,
+ BTRFS_I(inode)->location.objectid,
+ dentry->d_name.name,
+ dentry->d_name.len);
+ goto out;
+ }
+
err = btrfs_orphan_add(trans, inode);
if (err)
- goto fail_trans;
+ goto out;
/* now the directory is empty */
err = btrfs_unlink_inode(trans, root, dir, dentry->d_inode,
dentry->d_name.name, dentry->d_name.len);
if (!err)
btrfs_i_size_write(inode, 0);
-
-fail_trans:
+out:
nr = trans->blocks_used;
ret = btrfs_end_transaction_throttle(trans, root);
btrfs_btree_balance_dirty(root, nr);
struct btrfs_file_extent_item *fi;
u64 extent_start = 0;
u64 extent_num_bytes = 0;
+ u64 extent_offset = 0;
u64 item_end = 0;
- u64 root_gen = 0;
- u64 root_owner = 0;
int found_extent;
int del_item;
int pending_del_nr = 0;
if (root->ref_cows)
btrfs_drop_extent_cache(inode, new_size & (~mask), (u64)-1, 0);
path = btrfs_alloc_path();
- path->reada = -1;
BUG_ON(!path);
+ path->reada = -1;
/* FIXME, add redo link to tree so we don't leak on crash */
key.objectid = inode->i_ino;
extent_num_bytes =
btrfs_file_extent_disk_num_bytes(leaf,
fi);
+ extent_offset = found_key.offset -
+ btrfs_file_extent_offset(leaf, fi);
+
/* FIXME blocksize != 4096 */
num_dec = btrfs_file_extent_num_bytes(leaf, fi);
if (extent_start != 0) {
if (root->ref_cows)
inode_sub_bytes(inode, num_dec);
}
- root_gen = btrfs_header_generation(leaf);
- root_owner = btrfs_header_owner(leaf);
}
} else if (extent_type == BTRFS_FILE_EXTENT_INLINE) {
/*
} else {
break;
}
- if (found_extent) {
+ if (found_extent && root->ref_cows) {
btrfs_set_path_blocking(path);
ret = btrfs_free_extent(trans, root, extent_start,
- extent_num_bytes,
- leaf->start, root_owner,
- root_gen, inode->i_ino, 0);
+ extent_num_bytes, 0,
+ btrfs_header_owner(leaf),
+ inode->i_ino, extent_offset);
BUG_ON(ret);
}
next:
pending_del_nr);
}
btrfs_free_path(path);
- inode->i_sb->s_dirt = 1;
return ret;
}
err = btrfs_drop_extents(trans, root, inode,
cur_offset,
cur_offset + hole_size,
- cur_offset, &hint_byte);
+ block_end,
+ cur_offset, &hint_byte, 1);
if (err)
break;
err = btrfs_insert_file_extent(trans, root,
}
btrfs_wait_ordered_range(inode, 0, (u64)-1);
+ if (inode->i_nlink > 0) {
+ BUG_ON(btrfs_root_refs(&root->root_item) != 0);
+ goto no_delete;
+ }
+
btrfs_i_size_write(inode, 0);
trans = btrfs_join_transaction(root, 1);
* is kind of like crossing a mount point.
*/
static int fixup_tree_root_location(struct btrfs_root *root,
- struct btrfs_key *location,
- struct btrfs_root **sub_root,
- struct dentry *dentry)
+ struct inode *dir,
+ struct dentry *dentry,
+ struct btrfs_key *location,
+ struct btrfs_root **sub_root)
{
- struct btrfs_root_item *ri;
+ struct btrfs_path *path;
+ struct btrfs_root *new_root;
+ struct btrfs_root_ref *ref;
+ struct extent_buffer *leaf;
+ int ret;
+ int err = 0;
- if (btrfs_key_type(location) != BTRFS_ROOT_ITEM_KEY)
- return 0;
- if (location->objectid == BTRFS_ROOT_TREE_OBJECTID)
- return 0;
+ path = btrfs_alloc_path();
+ if (!path) {
+ err = -ENOMEM;
+ goto out;
+ }
- *sub_root = btrfs_read_fs_root(root->fs_info, location,
- dentry->d_name.name,
- dentry->d_name.len);
- if (IS_ERR(*sub_root))
- return PTR_ERR(*sub_root);
+ err = -ENOENT;
+ ret = btrfs_find_root_ref(root->fs_info->tree_root, path,
+ BTRFS_I(dir)->root->root_key.objectid,
+ location->objectid);
+ if (ret) {
+ if (ret < 0)
+ err = ret;
+ goto out;
+ }
- ri = &(*sub_root)->root_item;
- location->objectid = btrfs_root_dirid(ri);
- btrfs_set_key_type(location, BTRFS_INODE_ITEM_KEY);
+ leaf = path->nodes[0];
+ ref = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_root_ref);
+ if (btrfs_root_ref_dirid(leaf, ref) != dir->i_ino ||
+ btrfs_root_ref_name_len(leaf, ref) != dentry->d_name.len)
+ goto out;
+
+ ret = memcmp_extent_buffer(leaf, dentry->d_name.name,
+ (unsigned long)(ref + 1),
+ dentry->d_name.len);
+ if (ret)
+ goto out;
+
+ btrfs_release_path(root->fs_info->tree_root, path);
+
+ new_root = btrfs_read_fs_root_no_name(root->fs_info, location);
+ if (IS_ERR(new_root)) {
+ err = PTR_ERR(new_root);
+ goto out;
+ }
+
+ if (btrfs_root_refs(&new_root->root_item) == 0) {
+ err = -ENOENT;
+ goto out;
+ }
+
+ *sub_root = new_root;
+ location->objectid = btrfs_root_dirid(&new_root->root_item);
+ location->type = BTRFS_INODE_ITEM_KEY;
location->offset = 0;
+ err = 0;
+out:
+ btrfs_free_path(path);
+ return err;
+}
+
+static void inode_tree_add(struct inode *inode)
+{
+ struct btrfs_root *root = BTRFS_I(inode)->root;
+ struct btrfs_inode *entry;
+ struct rb_node **p;
+ struct rb_node *parent;
+again:
+ p = &root->inode_tree.rb_node;
+ parent = NULL;
+
+ if (hlist_unhashed(&inode->i_hash))
+ return;
+
+ spin_lock(&root->inode_lock);
+ while (*p) {
+ parent = *p;
+ entry = rb_entry(parent, struct btrfs_inode, rb_node);
+
+ if (inode->i_ino < entry->vfs_inode.i_ino)
+ p = &parent->rb_left;
+ else if (inode->i_ino > entry->vfs_inode.i_ino)
+ p = &parent->rb_right;
+ else {
+ WARN_ON(!(entry->vfs_inode.i_state &
+ (I_WILL_FREE | I_FREEING | I_CLEAR)));
+ rb_erase(parent, &root->inode_tree);
+ RB_CLEAR_NODE(parent);
+ spin_unlock(&root->inode_lock);
+ goto again;
+ }
+ }
+ rb_link_node(&BTRFS_I(inode)->rb_node, parent, p);
+ rb_insert_color(&BTRFS_I(inode)->rb_node, &root->inode_tree);
+ spin_unlock(&root->inode_lock);
+}
+
+static void inode_tree_del(struct inode *inode)
+{
+ struct btrfs_root *root = BTRFS_I(inode)->root;
+ int empty = 0;
+
+ spin_lock(&root->inode_lock);
+ if (!RB_EMPTY_NODE(&BTRFS_I(inode)->rb_node)) {
+ rb_erase(&BTRFS_I(inode)->rb_node, &root->inode_tree);
+ RB_CLEAR_NODE(&BTRFS_I(inode)->rb_node);
+ empty = RB_EMPTY_ROOT(&root->inode_tree);
+ }
+ spin_unlock(&root->inode_lock);
+
+ if (empty && btrfs_root_refs(&root->root_item) == 0) {
+ synchronize_srcu(&root->fs_info->subvol_srcu);
+ spin_lock(&root->inode_lock);
+ empty = RB_EMPTY_ROOT(&root->inode_tree);
+ spin_unlock(&root->inode_lock);
+ if (empty)
+ btrfs_add_dead_root(root);
+ }
+}
+
+int btrfs_invalidate_inodes(struct btrfs_root *root)
+{
+ struct rb_node *node;
+ struct rb_node *prev;
+ struct btrfs_inode *entry;
+ struct inode *inode;
+ u64 objectid = 0;
+
+ WARN_ON(btrfs_root_refs(&root->root_item) != 0);
+
+ spin_lock(&root->inode_lock);
+again:
+ node = root->inode_tree.rb_node;
+ prev = NULL;
+ while (node) {
+ prev = node;
+ entry = rb_entry(node, struct btrfs_inode, rb_node);
+
+ if (objectid < entry->vfs_inode.i_ino)
+ node = node->rb_left;
+ else if (objectid > entry->vfs_inode.i_ino)
+ node = node->rb_right;
+ else
+ break;
+ }
+ if (!node) {
+ while (prev) {
+ entry = rb_entry(prev, struct btrfs_inode, rb_node);
+ if (objectid <= entry->vfs_inode.i_ino) {
+ node = prev;
+ break;
+ }
+ prev = rb_next(prev);
+ }
+ }
+ while (node) {
+ entry = rb_entry(node, struct btrfs_inode, rb_node);
+ objectid = entry->vfs_inode.i_ino + 1;
+ inode = igrab(&entry->vfs_inode);
+ if (inode) {
+ spin_unlock(&root->inode_lock);
+ if (atomic_read(&inode->i_count) > 1)
+ d_prune_aliases(inode);
+ /*
+ * btrfs_drop_inode will remove it from
+ * the inode cache when its usage count
+ * hits zero.
+ */
+ iput(inode);
+ cond_resched();
+ spin_lock(&root->inode_lock);
+ goto again;
+ }
+
+ if (cond_resched_lock(&root->inode_lock))
+ goto again;
+ node = rb_next(node);
+ }
+ spin_unlock(&root->inode_lock);
return 0;
}
{
struct btrfs_inode *bi = BTRFS_I(inode);
- bi->i_acl = NULL;
- bi->i_default_acl = NULL;
-
bi->generation = 0;
bi->sequence = 0;
bi->last_trans = 0;
bi->flags = 0;
bi->index_cnt = (u64)-1;
bi->last_unlink_trans = 0;
+ bi->ordered_data_close = 0;
extent_map_tree_init(&BTRFS_I(inode)->extent_tree, GFP_NOFS);
extent_io_tree_init(&BTRFS_I(inode)->io_tree,
inode->i_mapping, GFP_NOFS);
inode->i_mapping, GFP_NOFS);
INIT_LIST_HEAD(&BTRFS_I(inode)->delalloc_inodes);
INIT_LIST_HEAD(&BTRFS_I(inode)->ordered_operations);
+ RB_CLEAR_NODE(&BTRFS_I(inode)->rb_node);
btrfs_ordered_inode_tree_init(&BTRFS_I(inode)->ordered_tree);
mutex_init(&BTRFS_I(inode)->extent_mutex);
mutex_init(&BTRFS_I(inode)->log_mutex);
args->root == BTRFS_I(inode)->root;
}
-struct inode *btrfs_ilookup(struct super_block *s, u64 objectid,
- struct btrfs_root *root, int wait)
-{
- struct inode *inode;
- struct btrfs_iget_args args;
- args.ino = objectid;
- args.root = root;
-
- if (wait) {
- inode = ilookup5(s, objectid, btrfs_find_actor,
- (void *)&args);
- } else {
- inode = ilookup5_nowait(s, objectid, btrfs_find_actor,
- (void *)&args);
- }
- return inode;
-}
-
-struct inode *btrfs_iget_locked(struct super_block *s, u64 objectid,
- struct btrfs_root *root)
+static struct inode *btrfs_iget_locked(struct super_block *s,
+ u64 objectid,
+ struct btrfs_root *root)
{
struct inode *inode;
struct btrfs_iget_args args;
* Returns in *is_new if the inode was read from disk
*/
struct inode *btrfs_iget(struct super_block *s, struct btrfs_key *location,
- struct btrfs_root *root, int *is_new)
+ struct btrfs_root *root)
{
struct inode *inode;
inode = btrfs_iget_locked(s, location->objectid, root);
if (!inode)
- return ERR_PTR(-EACCES);
+ return ERR_PTR(-ENOMEM);
if (inode->i_state & I_NEW) {
BTRFS_I(inode)->root = root;
memcpy(&BTRFS_I(inode)->location, location, sizeof(*location));
btrfs_read_locked_inode(inode);
+
+ inode_tree_add(inode);
unlock_new_inode(inode);
- if (is_new)
- *is_new = 1;
- } else {
- if (is_new)
- *is_new = 0;
}
return inode;
}
+static struct inode *new_simple_dir(struct super_block *s,
+ struct btrfs_key *key,
+ struct btrfs_root *root)
+{
+ struct inode *inode = new_inode(s);
+
+ if (!inode)
+ return ERR_PTR(-ENOMEM);
+
+ init_btrfs_i(inode);
+
+ BTRFS_I(inode)->root = root;
+ memcpy(&BTRFS_I(inode)->location, key, sizeof(*key));
+ BTRFS_I(inode)->dummy_inode = 1;
+
+ inode->i_ino = BTRFS_EMPTY_SUBVOL_DIR_OBJECTID;
+ inode->i_op = &simple_dir_inode_operations;
+ inode->i_fop = &simple_dir_operations;
+ inode->i_mode = S_IFDIR | S_IRUGO | S_IWUSR | S_IXUGO;
+ inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME;
+
+ return inode;
+}
+
struct inode *btrfs_lookup_dentry(struct inode *dir, struct dentry *dentry)
{
struct inode *inode;
- struct btrfs_inode *bi = BTRFS_I(dir);
- struct btrfs_root *root = bi->root;
+ struct btrfs_root *root = BTRFS_I(dir)->root;
struct btrfs_root *sub_root = root;
struct btrfs_key location;
- int ret, new;
+ int index;
+ int ret;
+
+ dentry->d_op = &btrfs_dentry_operations;
if (dentry->d_name.len > BTRFS_NAME_LEN)
return ERR_PTR(-ENAMETOOLONG);
if (ret < 0)
return ERR_PTR(ret);
- inode = NULL;
- if (location.objectid) {
- ret = fixup_tree_root_location(root, &location, &sub_root,
- dentry);
- if (ret < 0)
- return ERR_PTR(ret);
- if (ret > 0)
- return ERR_PTR(-ENOENT);
- inode = btrfs_iget(dir->i_sb, &location, sub_root, &new);
- if (IS_ERR(inode))
- return ERR_CAST(inode);
+ if (location.objectid == 0)
+ return NULL;
+
+ if (location.type == BTRFS_INODE_ITEM_KEY) {
+ inode = btrfs_iget(dir->i_sb, &location, root);
+ return inode;
}
+
+ BUG_ON(location.type != BTRFS_ROOT_ITEM_KEY);
+
+ index = srcu_read_lock(&root->fs_info->subvol_srcu);
+ ret = fixup_tree_root_location(root, dir, dentry,
+ &location, &sub_root);
+ if (ret < 0) {
+ if (ret != -ENOENT)
+ inode = ERR_PTR(ret);
+ else
+ inode = new_simple_dir(dir->i_sb, &location, sub_root);
+ } else {
+ inode = btrfs_iget(dir->i_sb, &location, sub_root);
+ }
+ srcu_read_unlock(&root->fs_info->subvol_srcu, index);
+
return inode;
}
+static int btrfs_dentry_delete(struct dentry *dentry)
+{
+ struct btrfs_root *root;
+
+ if (!dentry->d_inode)
+ return 0;
+
+ root = BTRFS_I(dentry->d_inode)->root;
+ if (btrfs_root_refs(&root->root_item) == 0)
+ return 1;
+ return 0;
+}
+
static struct dentry *btrfs_lookup(struct inode *dir, struct dentry *dentry,
struct nameidata *nd)
{
struct inode *inode;
- if (dentry->d_name.len > BTRFS_NAME_LEN)
- return ERR_PTR(-ENAMETOOLONG);
-
inode = btrfs_lookup_dentry(dir, dentry);
if (IS_ERR(inode))
return ERR_CAST(inode);
owner = 1;
BTRFS_I(inode)->block_group =
btrfs_find_block_group(root, 0, alloc_hint, owner);
- if ((mode & S_IFREG)) {
- if (btrfs_test_opt(root, NODATASUM))
- btrfs_set_flag(inode, NODATASUM);
- if (btrfs_test_opt(root, NODATACOW))
- btrfs_set_flag(inode, NODATACOW);
- }
key[0].objectid = objectid;
btrfs_set_key_type(&key[0], BTRFS_INODE_ITEM_KEY);
if (ret != 0)
goto fail;
- if (objectid > root->highest_inode)
- root->highest_inode = objectid;
-
inode->i_uid = current_fsuid();
if (dir && (dir->i_mode & S_ISGID)) {
location->offset = 0;
btrfs_set_key_type(location, BTRFS_INODE_ITEM_KEY);
+ btrfs_inherit_iflags(inode, dir);
+
+ if ((mode & S_IFREG)) {
+ if (btrfs_test_opt(root, NODATASUM))
+ BTRFS_I(inode)->flags |= BTRFS_INODE_NODATASUM;
+ if (btrfs_test_opt(root, NODATACOW))
+ BTRFS_I(inode)->flags |= BTRFS_INODE_NODATACOW;
+ }
+
insert_inode_hash(inode);
+ inode_tree_add(inode);
return inode;
fail:
if (dir)
struct inode *parent_inode, struct inode *inode,
const char *name, int name_len, int add_backref, u64 index)
{
- int ret;
+ int ret = 0;
struct btrfs_key key;
struct btrfs_root *root = BTRFS_I(parent_inode)->root;
- key.objectid = inode->i_ino;
- btrfs_set_key_type(&key, BTRFS_INODE_ITEM_KEY);
- key.offset = 0;
+ if (unlikely(inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)) {
+ memcpy(&key, &BTRFS_I(inode)->root->root_key, sizeof(key));
+ } else {
+ key.objectid = inode->i_ino;
+ btrfs_set_key_type(&key, BTRFS_INODE_ITEM_KEY);
+ key.offset = 0;
+ }
+
+ if (unlikely(inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)) {
+ ret = btrfs_add_root_ref(trans, root->fs_info->tree_root,
+ key.objectid, root->root_key.objectid,
+ parent_inode->i_ino,
+ index, name, name_len);
+ } else if (add_backref) {
+ ret = btrfs_insert_inode_ref(trans, root,
+ name, name_len, inode->i_ino,
+ parent_inode->i_ino, index);
+ }
- ret = btrfs_insert_dir_item(trans, root, name, name_len,
- parent_inode->i_ino,
- &key, btrfs_inode_type(inode),
- index);
if (ret == 0) {
- if (add_backref) {
- ret = btrfs_insert_inode_ref(trans, root,
- name, name_len,
- inode->i_ino,
- parent_inode->i_ino,
- index);
- }
+ ret = btrfs_insert_dir_item(trans, root, name, name_len,
+ parent_inode->i_ino, &key,
+ btrfs_inode_type(inode), index);
+ BUG_ON(ret);
+
btrfs_i_size_write(parent_inode, parent_inode->i_size +
name_len * 2);
parent_inode->i_mtime = parent_inode->i_ctime = CURRENT_TIME;
init_special_inode(inode, inode->i_mode, rdev);
btrfs_update_inode(trans, root, inode);
}
- dir->i_sb->s_dirt = 1;
btrfs_update_inode_block_group(trans, inode);
btrfs_update_inode_block_group(trans, dir);
out_unlock:
inode->i_op = &btrfs_file_inode_operations;
BTRFS_I(inode)->io_tree.ops = &btrfs_extent_io_ops;
}
- dir->i_sb->s_dirt = 1;
btrfs_update_inode_block_group(trans, inode);
btrfs_update_inode_block_group(trans, dir);
out_unlock:
err = btrfs_add_nondir(trans, dentry, inode, 1, index);
- if (err)
- drop_inode = 1;
-
- dir->i_sb->s_dirt = 1;
- btrfs_update_inode_block_group(trans, dir);
- err = btrfs_update_inode(trans, root, inode);
-
- if (err)
+ if (err) {
drop_inode = 1;
+ } else {
+ btrfs_update_inode_block_group(trans, dir);
+ err = btrfs_update_inode(trans, root, inode);
+ BUG_ON(err);
+ btrfs_log_new_name(trans, inode, NULL, dentry->d_parent);
+ }
nr = trans->blocks_used;
-
- btrfs_log_new_name(trans, inode, NULL, dentry->d_parent);
btrfs_end_transaction_throttle(trans, root);
fail:
if (drop_inode) {
d_instantiate(dentry, inode);
drop_on_err = 0;
- dir->i_sb->s_dirt = 1;
btrfs_update_inode_block_group(trans, inode);
btrfs_update_inode_block_group(trans, dir);
int compressed;
again:
- spin_lock(&em_tree->lock);
+ read_lock(&em_tree->lock);
em = lookup_extent_mapping(em_tree, start, len);
if (em)
em->bdev = root->fs_info->fs_devices->latest_bdev;
- spin_unlock(&em_tree->lock);
+ read_unlock(&em_tree->lock);
if (em) {
if (em->start > start || em->start + em->len <= start)
map = kmap(page);
read_extent_buffer(leaf, map + pg_offset, ptr,
copy_size);
+ if (pg_offset + copy_size < PAGE_CACHE_SIZE) {
+ memset(map + pg_offset + copy_size, 0,
+ PAGE_CACHE_SIZE - pg_offset -
+ copy_size);
+ }
kunmap(page);
}
flush_dcache_page(page);
}
err = 0;
- spin_lock(&em_tree->lock);
+ write_lock(&em_tree->lock);
ret = add_extent_mapping(em_tree, em);
/* it is possible that someone inserted the extent into the tree
* while we had the lock dropped. It is also possible that
err = 0;
}
}
- spin_unlock(&em_tree->lock);
+ write_unlock(&em_tree->lock);
out:
if (path)
btrfs_free_path(path);
}
if (err) {
free_extent_map(em);
- WARN_ON(1);
return ERR_PTR(err);
}
return em;
u64 page_start = page_offset(page);
u64 page_end = page_start + PAGE_CACHE_SIZE - 1;
+
+ /*
+ * we have the page locked, so new writeback can't start,
+ * and the dirty bit won't be cleared while we are here.
+ *
+ * Wait for IO on this page so that we can safely clear
+ * the PagePrivate2 bit and do ordered accounting
+ */
wait_on_page_writeback(page);
+
tree = &BTRFS_I(page->mapping->host)->io_tree;
if (offset) {
btrfs_releasepage(page, GFP_NOFS);
return;
}
-
lock_extent(tree, page_start, page_end, GFP_NOFS);
ordered = btrfs_lookup_ordered_extent(page->mapping->host,
page_offset(page));
*/
clear_extent_bit(tree, page_start, page_end,
EXTENT_DIRTY | EXTENT_DELALLOC |
- EXTENT_LOCKED, 1, 0, GFP_NOFS);
- btrfs_finish_ordered_io(page->mapping->host,
- page_start, page_end);
+ EXTENT_LOCKED, 1, 0, NULL, GFP_NOFS);
+ /*
+ * whoever cleared the private bit is responsible
+ * for the finish_ordered_io
+ */
+ if (TestClearPagePrivate2(page)) {
+ btrfs_finish_ordered_io(page->mapping->host,
+ page_start, page_end);
+ }
btrfs_put_ordered_extent(ordered);
lock_extent(tree, page_start, page_end, GFP_NOFS);
}
clear_extent_bit(tree, page_start, page_end,
- EXTENT_LOCKED | EXTENT_DIRTY | EXTENT_DELALLOC |
- EXTENT_ORDERED,
- 1, 1, GFP_NOFS);
+ EXTENT_LOCKED | EXTENT_DIRTY | EXTENT_DELALLOC,
+ 1, 1, NULL, GFP_NOFS);
__btrfs_releasepage(page, GFP_NOFS);
ClearPageChecked(page);
}
ClearPageChecked(page);
set_page_dirty(page);
+ SetPageUptodate(page);
BTRFS_I(inode)->last_trans = root->fs_info->generation + 1;
unlock_extent(io_tree, page_start, page_end, GFP_NOFS);
out_unlock:
+ if (!ret)
+ return VM_FAULT_LOCKED;
unlock_page(page);
out:
return ret;
* create a new subvolume directory/inode (helper for the ioctl).
*/
int btrfs_create_subvol_root(struct btrfs_trans_handle *trans,
- struct btrfs_root *new_root, struct dentry *dentry,
+ struct btrfs_root *new_root,
u64 new_dirid, u64 alloc_hint)
{
struct inode *inode;
- int error;
+ int err;
u64 index = 0;
inode = btrfs_new_inode(trans, new_root, NULL, "..", 2, new_dirid,
inode->i_nlink = 1;
btrfs_i_size_write(inode, 0);
- error = btrfs_update_inode(trans, new_root, inode);
- if (error)
- return error;
+ err = btrfs_update_inode(trans, new_root, inode);
+ BUG_ON(err);
- d_instantiate(dentry, inode);
+ iput(inode);
return 0;
}
ei->last_trans = 0;
ei->logged_trans = 0;
btrfs_ordered_inode_tree_init(&ei->ordered_tree);
- ei->i_acl = BTRFS_ACL_NOT_CACHED;
- ei->i_default_acl = BTRFS_ACL_NOT_CACHED;
INIT_LIST_HEAD(&ei->i_orphan);
INIT_LIST_HEAD(&ei->ordered_operations);
return &ei->vfs_inode;
WARN_ON(!list_empty(&inode->i_dentry));
WARN_ON(inode->i_data.nrpages);
- if (BTRFS_I(inode)->i_acl &&
- BTRFS_I(inode)->i_acl != BTRFS_ACL_NOT_CACHED)
- posix_acl_release(BTRFS_I(inode)->i_acl);
- if (BTRFS_I(inode)->i_default_acl &&
- BTRFS_I(inode)->i_default_acl != BTRFS_ACL_NOT_CACHED)
- posix_acl_release(BTRFS_I(inode)->i_default_acl);
-
/*
* Make sure we're properly removed from the ordered operation
* lists.
btrfs_put_ordered_extent(ordered);
}
}
+ inode_tree_del(inode);
btrfs_drop_extent_cache(inode, 0, (u64)-1, 0);
kmem_cache_free(btrfs_inode_cachep, BTRFS_I(inode));
}
+void btrfs_drop_inode(struct inode *inode)
+{
+ struct btrfs_root *root = BTRFS_I(inode)->root;
+
+ if (inode->i_nlink > 0 && btrfs_root_refs(&root->root_item) == 0)
+ generic_delete_inode(inode);
+ else
+ generic_drop_inode(inode);
+}
+
static void init_once(void *foo)
{
struct btrfs_inode *ei = (struct btrfs_inode *) foo;
kmem_cache_destroy(btrfs_trans_handle_cachep);
if (btrfs_transaction_cachep)
kmem_cache_destroy(btrfs_transaction_cachep);
- if (btrfs_bit_radix_cachep)
- kmem_cache_destroy(btrfs_bit_radix_cachep);
if (btrfs_path_cachep)
kmem_cache_destroy(btrfs_path_cachep);
}
if (!btrfs_path_cachep)
goto fail;
- btrfs_bit_radix_cachep = kmem_cache_create("btrfs_radix", 256, 0,
- SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD |
- SLAB_DESTROY_BY_RCU, NULL);
- if (!btrfs_bit_radix_cachep)
- goto fail;
return 0;
fail:
btrfs_destroy_cachep();
{
struct btrfs_trans_handle *trans;
struct btrfs_root *root = BTRFS_I(old_dir)->root;
+ struct btrfs_root *dest = BTRFS_I(new_dir)->root;
struct inode *new_inode = new_dentry->d_inode;
struct inode *old_inode = old_dentry->d_inode;
struct timespec ctime = CURRENT_TIME;
u64 index = 0;
+ u64 root_objectid;
int ret;
- /* we're not allowed to rename between subvolumes */
- if (BTRFS_I(old_inode)->root->root_key.objectid !=
- BTRFS_I(new_dir)->root->root_key.objectid)
+ if (new_dir->i_ino == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID)
+ return -EPERM;
+
+ /* we only allow rename subvolume link between subvolumes */
+ if (old_inode->i_ino != BTRFS_FIRST_FREE_OBJECTID && root != dest)
return -EXDEV;
- if (S_ISDIR(old_inode->i_mode) && new_inode &&
- new_inode->i_size > BTRFS_EMPTY_DIR_SIZE) {
+ if (old_inode->i_ino == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID ||
+ (new_inode && new_inode->i_ino == BTRFS_FIRST_FREE_OBJECTID))
return -ENOTEMPTY;
- }
- /* to rename a snapshot or subvolume, we need to juggle the
- * backrefs. This isn't coded yet
- */
- if (old_inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)
- return -EXDEV;
+ if (S_ISDIR(old_inode->i_mode) && new_inode &&
+ new_inode->i_size > BTRFS_EMPTY_DIR_SIZE)
+ return -ENOTEMPTY;
ret = btrfs_check_metadata_free_space(root);
if (ret)
- goto out_unlock;
+ return ret;
/*
* we're using rename to replace one file with another.
* and the replacement file is large. Start IO on it now so
* we don't add too much work to the end of the transaction
*/
- if (new_inode && old_inode && S_ISREG(old_inode->i_mode) &&
- new_inode->i_size &&
+ if (new_inode && S_ISREG(old_inode->i_mode) && new_inode->i_size &&
old_inode->i_size > BTRFS_ORDERED_OPERATIONS_FLUSH_LIMIT)
filemap_flush(old_inode->i_mapping);
+ /* close the racy window with snapshot create/destroy ioctl */
+ if (old_inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)
+ down_read(&root->fs_info->subvol_sem);
+
trans = btrfs_start_transaction(root, 1);
+ btrfs_set_trans_block_group(trans, new_dir);
+
+ if (dest != root)
+ btrfs_record_root_in_trans(trans, dest);
+
+ ret = btrfs_set_inode_index(new_dir, &index);
+ if (ret)
+ goto out_fail;
+ if (unlikely(old_inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)) {
+ /* force full log commit if subvolume involved. */
+ root->fs_info->last_trans_log_full_commit = trans->transid;
+ } else {
+ ret = btrfs_insert_inode_ref(trans, dest,
+ new_dentry->d_name.name,
+ new_dentry->d_name.len,
+ old_inode->i_ino,
+ new_dir->i_ino, index);
+ if (ret)
+ goto out_fail;
+ /*
+ * this is an ugly little race, but the rename is required
+ * to make sure that if we crash, the inode is either at the
+ * old name or the new one. pinning the log transaction lets
+ * us make sure we don't allow a log commit to come in after
+ * we unlink the name but before we add the new name back in.
+ */
+ btrfs_pin_log_trans(root);
+ }
/*
* make sure the inode gets flushed if it is replacing
* something.
btrfs_add_ordered_operation(trans, root, old_inode);
}
- /*
- * this is an ugly little race, but the rename is required to make
- * sure that if we crash, the inode is either at the old name
- * or the new one. pinning the log transaction lets us make sure
- * we don't allow a log commit to come in after we unlink the
- * name but before we add the new name back in.
- */
- btrfs_pin_log_trans(root);
-
- btrfs_set_trans_block_group(trans, new_dir);
-
- btrfs_inc_nlink(old_dentry->d_inode);
old_dir->i_ctime = old_dir->i_mtime = ctime;
new_dir->i_ctime = new_dir->i_mtime = ctime;
old_inode->i_ctime = ctime;
if (old_dentry->d_parent != new_dentry->d_parent)
btrfs_record_unlink_dir(trans, old_dir, old_inode, 1);
- ret = btrfs_unlink_inode(trans, root, old_dir, old_dentry->d_inode,
- old_dentry->d_name.name,
- old_dentry->d_name.len);
- if (ret)
- goto out_fail;
+ if (unlikely(old_inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)) {
+ root_objectid = BTRFS_I(old_inode)->root->root_key.objectid;
+ ret = btrfs_unlink_subvol(trans, root, old_dir, root_objectid,
+ old_dentry->d_name.name,
+ old_dentry->d_name.len);
+ } else {
+ btrfs_inc_nlink(old_dentry->d_inode);
+ ret = btrfs_unlink_inode(trans, root, old_dir,
+ old_dentry->d_inode,
+ old_dentry->d_name.name,
+ old_dentry->d_name.len);
+ }
+ BUG_ON(ret);
if (new_inode) {
new_inode->i_ctime = CURRENT_TIME;
- ret = btrfs_unlink_inode(trans, root, new_dir,
- new_dentry->d_inode,
- new_dentry->d_name.name,
- new_dentry->d_name.len);
- if (ret)
- goto out_fail;
+ if (unlikely(new_inode->i_ino ==
+ BTRFS_EMPTY_SUBVOL_DIR_OBJECTID)) {
+ root_objectid = BTRFS_I(new_inode)->location.objectid;
+ ret = btrfs_unlink_subvol(trans, dest, new_dir,
+ root_objectid,
+ new_dentry->d_name.name,
+ new_dentry->d_name.len);
+ BUG_ON(new_inode->i_nlink == 0);
+ } else {
+ ret = btrfs_unlink_inode(trans, dest, new_dir,
+ new_dentry->d_inode,
+ new_dentry->d_name.name,
+ new_dentry->d_name.len);
+ }
+ BUG_ON(ret);
if (new_inode->i_nlink == 0) {
ret = btrfs_orphan_add(trans, new_dentry->d_inode);
- if (ret)
- goto out_fail;
+ BUG_ON(ret);
}
-
}
- ret = btrfs_set_inode_index(new_dir, &index);
- if (ret)
- goto out_fail;
- ret = btrfs_add_link(trans, new_dentry->d_parent->d_inode,
- old_inode, new_dentry->d_name.name,
- new_dentry->d_name.len, 1, index);
- if (ret)
- goto out_fail;
+ ret = btrfs_add_link(trans, new_dir, old_inode,
+ new_dentry->d_name.name,
+ new_dentry->d_name.len, 0, index);
+ BUG_ON(ret);
- btrfs_log_new_name(trans, old_inode, old_dir,
- new_dentry->d_parent);
+ if (old_inode->i_ino != BTRFS_FIRST_FREE_OBJECTID) {
+ btrfs_log_new_name(trans, old_inode, old_dir,
+ new_dentry->d_parent);
+ btrfs_end_log_trans(root);
+ }
out_fail:
-
- /* this btrfs_end_log_trans just allows the current
- * log-sub transaction to complete
- */
- btrfs_end_log_trans(root);
btrfs_end_transaction_throttle(trans, root);
-out_unlock:
+
+ if (old_inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)
+ up_read(&root->fs_info->subvol_sem);
return ret;
}
inode->i_op = &btrfs_file_inode_operations;
BTRFS_I(inode)->io_tree.ops = &btrfs_extent_io_ops;
}
- dir->i_sb->s_dirt = 1;
btrfs_update_inode_block_group(trans, inode);
btrfs_update_inode_block_group(trans, dir);
if (drop_inode)
static int prealloc_file_range(struct btrfs_trans_handle *trans,
struct inode *inode, u64 start, u64 end,
- u64 alloc_hint, int mode)
+ u64 locked_end, u64 alloc_hint, int mode)
{
struct btrfs_root *root = BTRFS_I(inode)->root;
struct btrfs_key ins;
ret = insert_reserved_file_extent(trans, inode,
cur_offset, ins.objectid,
ins.offset, ins.offset,
- ins.offset, 0, 0, 0,
+ ins.offset, locked_end,
+ 0, 0, 0,
BTRFS_FILE_EXTENT_PREALLOC);
BUG_ON(ret);
+ btrfs_drop_extent_cache(inode, cur_offset,
+ cur_offset + ins.offset -1, 0);
num_bytes -= ins.offset;
cur_offset += ins.offset;
alloc_hint = ins.objectid + ins.offset;
out:
if (cur_offset > start) {
inode->i_ctime = CURRENT_TIME;
- btrfs_set_flag(inode, PREALLOC);
+ BTRFS_I(inode)->flags |= BTRFS_INODE_PREALLOC;
if (!(mode & FALLOC_FL_KEEP_SIZE) &&
cur_offset > i_size_read(inode))
btrfs_i_size_write(inode, cur_offset);
u64 alloc_start;
u64 alloc_end;
u64 alloc_hint = 0;
+ u64 locked_end;
u64 mask = BTRFS_I(inode)->root->sectorsize - 1;
struct extent_map *em;
struct btrfs_trans_handle *trans;
+ struct btrfs_root *root;
int ret;
alloc_start = offset & ~mask;
goto out;
}
+ root = BTRFS_I(inode)->root;
+
+ ret = btrfs_check_data_free_space(root, inode,
+ alloc_end - alloc_start);
+ if (ret)
+ goto out;
+
+ locked_end = alloc_end - 1;
while (1) {
struct btrfs_ordered_extent *ordered;
trans = btrfs_start_transaction(BTRFS_I(inode)->root, 1);
if (!trans) {
ret = -EIO;
- goto out;
+ goto out_free;
}
/* the extent lock is ordered inside the running
* transaction
*/
- lock_extent(&BTRFS_I(inode)->io_tree, alloc_start,
- alloc_end - 1, GFP_NOFS);
+ lock_extent(&BTRFS_I(inode)->io_tree, alloc_start, locked_end,
+ GFP_NOFS);
ordered = btrfs_lookup_first_ordered_extent(inode,
alloc_end - 1);
if (ordered &&
ordered->file_offset < alloc_end) {
btrfs_put_ordered_extent(ordered);
unlock_extent(&BTRFS_I(inode)->io_tree,
- alloc_start, alloc_end - 1, GFP_NOFS);
+ alloc_start, locked_end, GFP_NOFS);
btrfs_end_transaction(trans, BTRFS_I(inode)->root);
/*
last_byte = (last_byte + mask) & ~mask;
if (em->block_start == EXTENT_MAP_HOLE) {
ret = prealloc_file_range(trans, inode, cur_offset,
- last_byte, alloc_hint, mode);
+ last_byte, locked_end + 1,
+ alloc_hint, mode);
if (ret < 0) {
free_extent_map(em);
break;
break;
}
}
- unlock_extent(&BTRFS_I(inode)->io_tree, alloc_start, alloc_end - 1,
+ unlock_extent(&BTRFS_I(inode)->io_tree, alloc_start, locked_end,
GFP_NOFS);
btrfs_end_transaction(trans, BTRFS_I(inode)->root);
+out_free:
+ btrfs_free_reserved_data_space(root, inode, alloc_end - alloc_start);
out:
mutex_unlock(&inode->i_mutex);
return ret;
static int btrfs_permission(struct inode *inode, int mask)
{
- if (btrfs_test_flag(inode, READONLY) && (mask & MAY_WRITE))
+ if ((BTRFS_I(inode)->flags & BTRFS_INODE_READONLY) && (mask & MAY_WRITE))
return -EACCES;
return generic_permission(inode, mask, btrfs_check_acl);
}
-static struct inode_operations btrfs_dir_inode_operations = {
+static const struct inode_operations btrfs_dir_inode_operations = {
.getattr = btrfs_getattr,
.lookup = btrfs_lookup,
.create = btrfs_create,
.removexattr = btrfs_removexattr,
.permission = btrfs_permission,
};
-static struct inode_operations btrfs_dir_ro_inode_operations = {
+static const struct inode_operations btrfs_dir_ro_inode_operations = {
.lookup = btrfs_lookup,
.permission = btrfs_permission,
};
-static struct file_operations btrfs_dir_file_operations = {
+
+static const struct file_operations btrfs_dir_file_operations = {
.llseek = generic_file_llseek,
.read = generic_read_dir,
.readdir = btrfs_real_readdir,
*
* For now we're avoiding this by dropping bmap.
*/
-static struct address_space_operations btrfs_aops = {
+static const struct address_space_operations btrfs_aops = {
.readpage = btrfs_readpage,
.writepage = btrfs_writepage,
.writepages = btrfs_writepages,
.invalidatepage = btrfs_invalidatepage,
.releasepage = btrfs_releasepage,
.set_page_dirty = btrfs_set_page_dirty,
+ .error_remove_page = generic_error_remove_page,
};
-static struct address_space_operations btrfs_symlink_aops = {
+static const struct address_space_operations btrfs_symlink_aops = {
.readpage = btrfs_readpage,
.writepage = btrfs_writepage,
.invalidatepage = btrfs_invalidatepage,
.releasepage = btrfs_releasepage,
};
-static struct inode_operations btrfs_file_inode_operations = {
+static const struct inode_operations btrfs_file_inode_operations = {
.truncate = btrfs_truncate,
.getattr = btrfs_getattr,
.setattr = btrfs_setattr,
.fallocate = btrfs_fallocate,
.fiemap = btrfs_fiemap,
};
-static struct inode_operations btrfs_special_inode_operations = {
+static const struct inode_operations btrfs_special_inode_operations = {
.getattr = btrfs_getattr,
.setattr = btrfs_setattr,
.permission = btrfs_permission,
.listxattr = btrfs_listxattr,
.removexattr = btrfs_removexattr,
};
-static struct inode_operations btrfs_symlink_inode_operations = {
+static const struct inode_operations btrfs_symlink_inode_operations = {
.readlink = generic_readlink,
.follow_link = page_follow_link_light,
.put_link = page_put_link,
.listxattr = btrfs_listxattr,
.removexattr = btrfs_removexattr,
};
+
+struct dentry_operations btrfs_dentry_operations = {
+ .d_delete = btrfs_dentry_delete,
+};
git://ftp.safe.ca / safe / jmp / linux-2.6 / blobdiff