#include "space-info.h"
#include "zoned.h"
#include "inode-item.h"
+#include "fs.h"
+#include "accessors.h"
+#include "extent-tree.h"
+#include "root-tree.h"
+#include "dir-item.h"
+#include "file-item.h"
+#include "file.h"
+#include "orphan.h"
#define MAX_CONFLICT_INODES 10
* pin down any logged extents, so we have to read the block.
*/
if (btrfs_fs_incompat(fs_info, MIXED_GROUPS)) {
- ret = btrfs_read_extent_buffer(eb, gen, level, NULL);
+ struct btrfs_tree_parent_check check = {
+ .level = level,
+ .transid = gen
+ };
+
+ ret = btrfs_read_extent_buffer(eb, &check);
if (ret)
return ret;
}
return do_overwrite_item(trans, root, path, eb, slot, key);
}
+static int read_alloc_one_name(struct extent_buffer *eb, void *start, int len,
+ struct fscrypt_str *name)
+{
+ char *buf;
+
+ buf = kmalloc(len, GFP_NOFS);
+ if (!buf)
+ return -ENOMEM;
+
+ read_extent_buffer(eb, buf, (unsigned long)start, len);
+ name->name = buf;
+ name->len = len;
+ return 0;
+}
+
/*
* simple helper to read an inode off the disk from a given root
* This can only be called for subvolume roots and not for the log
*/
ret = btrfs_qgroup_trace_extent(trans,
btrfs_file_extent_disk_bytenr(eb, item),
- btrfs_file_extent_disk_num_bytes(eb, item),
- GFP_NOFS);
+ btrfs_file_extent_disk_num_bytes(eb, item));
if (ret < 0)
goto out;
static int unlink_inode_for_log_replay(struct btrfs_trans_handle *trans,
struct btrfs_inode *dir,
struct btrfs_inode *inode,
- const char *name,
- int name_len)
+ const struct fscrypt_str *name)
{
int ret;
- ret = btrfs_unlink_inode(trans, dir, inode, name, name_len);
+ ret = btrfs_unlink_inode(trans, dir, inode, name);
if (ret)
return ret;
/*
{
struct btrfs_root *root = dir->root;
struct inode *inode;
- char *name;
- int name_len;
+ struct fscrypt_str name;
struct extent_buffer *leaf;
struct btrfs_key location;
int ret;
leaf = path->nodes[0];
btrfs_dir_item_key_to_cpu(leaf, di, &location);
- name_len = btrfs_dir_name_len(leaf, di);
- name = kmalloc(name_len, GFP_NOFS);
- if (!name)
+ ret = read_alloc_one_name(leaf, di + 1, btrfs_dir_name_len(leaf, di), &name);
+ if (ret)
return -ENOMEM;
- read_extent_buffer(leaf, name, (unsigned long)(di + 1), name_len);
btrfs_release_path(path);
inode = read_one_inode(root, location.objectid);
if (ret)
goto out;
- ret = unlink_inode_for_log_replay(trans, dir, BTRFS_I(inode), name,
- name_len);
+ ret = unlink_inode_for_log_replay(trans, dir, BTRFS_I(inode), &name);
out:
- kfree(name);
+ kfree(name.name);
iput(inode);
return ret;
}
static noinline int inode_in_dir(struct btrfs_root *root,
struct btrfs_path *path,
u64 dirid, u64 objectid, u64 index,
- const char *name, int name_len)
+ struct fscrypt_str *name)
{
struct btrfs_dir_item *di;
struct btrfs_key location;
int ret = 0;
di = btrfs_lookup_dir_index_item(NULL, root, path, dirid,
- index, name, name_len, 0);
+ index, name, 0);
if (IS_ERR(di)) {
ret = PTR_ERR(di);
goto out;
}
btrfs_release_path(path);
- di = btrfs_lookup_dir_item(NULL, root, path, dirid, name, name_len, 0);
+ di = btrfs_lookup_dir_item(NULL, root, path, dirid, name, 0);
if (IS_ERR(di)) {
ret = PTR_ERR(di);
goto out;
static noinline int backref_in_log(struct btrfs_root *log,
struct btrfs_key *key,
u64 ref_objectid,
- const char *name, int namelen)
+ const struct fscrypt_str *name)
{
struct btrfs_path *path;
int ret;
if (key->type == BTRFS_INODE_EXTREF_KEY)
ret = !!btrfs_find_name_in_ext_backref(path->nodes[0],
path->slots[0],
- ref_objectid,
- name, namelen);
+ ref_objectid, name);
else
ret = !!btrfs_find_name_in_backref(path->nodes[0],
- path->slots[0],
- name, namelen);
+ path->slots[0], name);
out:
btrfs_free_path(path);
return ret;
struct btrfs_inode *dir,
struct btrfs_inode *inode,
u64 inode_objectid, u64 parent_objectid,
- u64 ref_index, char *name, int namelen)
+ u64 ref_index, struct fscrypt_str *name)
{
int ret;
- char *victim_name;
- int victim_name_len;
struct extent_buffer *leaf;
struct btrfs_dir_item *di;
struct btrfs_key search_key;
ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
ptr_end = ptr + btrfs_item_size(leaf, path->slots[0]);
while (ptr < ptr_end) {
- victim_ref = (struct btrfs_inode_ref *)ptr;
- victim_name_len = btrfs_inode_ref_name_len(leaf,
- victim_ref);
- victim_name = kmalloc(victim_name_len, GFP_NOFS);
- if (!victim_name)
- return -ENOMEM;
+ struct fscrypt_str victim_name;
- read_extent_buffer(leaf, victim_name,
- (unsigned long)(victim_ref + 1),
- victim_name_len);
+ victim_ref = (struct btrfs_inode_ref *)ptr;
+ ret = read_alloc_one_name(leaf, (victim_ref + 1),
+ btrfs_inode_ref_name_len(leaf, victim_ref),
+ &victim_name);
+ if (ret)
+ return ret;
ret = backref_in_log(log_root, &search_key,
- parent_objectid, victim_name,
- victim_name_len);
+ parent_objectid, &victim_name);
if (ret < 0) {
- kfree(victim_name);
+ kfree(victim_name.name);
return ret;
} else if (!ret) {
inc_nlink(&inode->vfs_inode);
btrfs_release_path(path);
ret = unlink_inode_for_log_replay(trans, dir, inode,
- victim_name, victim_name_len);
- kfree(victim_name);
+ &victim_name);
+ kfree(victim_name.name);
if (ret)
return ret;
goto again;
}
- kfree(victim_name);
+ kfree(victim_name.name);
- ptr = (unsigned long)(victim_ref + 1) + victim_name_len;
+ ptr = (unsigned long)(victim_ref + 1) + victim_name.len;
}
}
btrfs_release_path(path);
/* Same search but for extended refs */
- extref = btrfs_lookup_inode_extref(NULL, root, path, name, namelen,
+ extref = btrfs_lookup_inode_extref(NULL, root, path, name,
inode_objectid, parent_objectid, 0,
0);
if (IS_ERR(extref)) {
base = btrfs_item_ptr_offset(leaf, path->slots[0]);
while (cur_offset < item_size) {
- extref = (struct btrfs_inode_extref *)(base + cur_offset);
+ struct fscrypt_str victim_name;
- victim_name_len = btrfs_inode_extref_name_len(leaf, extref);
+ extref = (struct btrfs_inode_extref *)(base + cur_offset);
if (btrfs_inode_extref_parent(leaf, extref) != parent_objectid)
goto next;
- victim_name = kmalloc(victim_name_len, GFP_NOFS);
- if (!victim_name)
- return -ENOMEM;
- read_extent_buffer(leaf, victim_name, (unsigned long)&extref->name,
- victim_name_len);
+ ret = read_alloc_one_name(leaf, &extref->name,
+ btrfs_inode_extref_name_len(leaf, extref),
+ &victim_name);
+ if (ret)
+ return ret;
search_key.objectid = inode_objectid;
search_key.type = BTRFS_INODE_EXTREF_KEY;
search_key.offset = btrfs_extref_hash(parent_objectid,
- victim_name,
- victim_name_len);
+ victim_name.name,
+ victim_name.len);
ret = backref_in_log(log_root, &search_key,
- parent_objectid, victim_name,
- victim_name_len);
+ parent_objectid, &victim_name);
if (ret < 0) {
- kfree(victim_name);
+ kfree(victim_name.name);
return ret;
} else if (!ret) {
ret = -ENOENT;
ret = unlink_inode_for_log_replay(trans,
BTRFS_I(victim_parent),
- inode,
- victim_name,
- victim_name_len);
+ inode, &victim_name);
}
iput(victim_parent);
- kfree(victim_name);
+ kfree(victim_name.name);
if (ret)
return ret;
goto again;
}
- kfree(victim_name);
+ kfree(victim_name.name);
next:
- cur_offset += victim_name_len + sizeof(*extref);
+ cur_offset += victim_name.len + sizeof(*extref);
}
}
btrfs_release_path(path);
/* look for a conflicting sequence number */
di = btrfs_lookup_dir_index_item(trans, root, path, btrfs_ino(dir),
- ref_index, name, namelen, 0);
+ ref_index, name, 0);
if (IS_ERR(di)) {
return PTR_ERR(di);
} else if (di) {
btrfs_release_path(path);
/* look for a conflicting name */
- di = btrfs_lookup_dir_item(trans, root, path, btrfs_ino(dir),
- name, namelen, 0);
+ di = btrfs_lookup_dir_item(trans, root, path, btrfs_ino(dir), name, 0);
if (IS_ERR(di)) {
return PTR_ERR(di);
} else if (di) {
}
static int extref_get_fields(struct extent_buffer *eb, unsigned long ref_ptr,
- u32 *namelen, char **name, u64 *index,
+ struct fscrypt_str *name, u64 *index,
u64 *parent_objectid)
{
struct btrfs_inode_extref *extref;
+ int ret;
extref = (struct btrfs_inode_extref *)ref_ptr;
- *namelen = btrfs_inode_extref_name_len(eb, extref);
- *name = kmalloc(*namelen, GFP_NOFS);
- if (*name == NULL)
- return -ENOMEM;
-
- read_extent_buffer(eb, *name, (unsigned long)&extref->name,
- *namelen);
+ ret = read_alloc_one_name(eb, &extref->name,
+ btrfs_inode_extref_name_len(eb, extref), name);
+ if (ret)
+ return ret;
if (index)
*index = btrfs_inode_extref_index(eb, extref);
}
static int ref_get_fields(struct extent_buffer *eb, unsigned long ref_ptr,
- u32 *namelen, char **name, u64 *index)
+ struct fscrypt_str *name, u64 *index)
{
struct btrfs_inode_ref *ref;
+ int ret;
ref = (struct btrfs_inode_ref *)ref_ptr;
- *namelen = btrfs_inode_ref_name_len(eb, ref);
- *name = kmalloc(*namelen, GFP_NOFS);
- if (*name == NULL)
- return -ENOMEM;
-
- read_extent_buffer(eb, *name, (unsigned long)(ref + 1), *namelen);
+ ret = read_alloc_one_name(eb, ref + 1, btrfs_inode_ref_name_len(eb, ref),
+ name);
+ if (ret)
+ return ret;
if (index)
*index = btrfs_inode_ref_index(eb, ref);
ref_ptr = btrfs_item_ptr_offset(eb, path->slots[0]);
ref_end = ref_ptr + btrfs_item_size(eb, path->slots[0]);
while (ref_ptr < ref_end) {
- char *name = NULL;
- int namelen;
+ struct fscrypt_str name;
u64 parent_id;
if (key->type == BTRFS_INODE_EXTREF_KEY) {
- ret = extref_get_fields(eb, ref_ptr, &namelen, &name,
+ ret = extref_get_fields(eb, ref_ptr, &name,
NULL, &parent_id);
} else {
parent_id = key->offset;
- ret = ref_get_fields(eb, ref_ptr, &namelen, &name,
- NULL);
+ ret = ref_get_fields(eb, ref_ptr, &name, NULL);
}
if (ret)
goto out;
if (key->type == BTRFS_INODE_EXTREF_KEY)
ret = !!btrfs_find_name_in_ext_backref(log_eb, log_slot,
- parent_id, name,
- namelen);
+ parent_id, &name);
else
- ret = !!btrfs_find_name_in_backref(log_eb, log_slot,
- name, namelen);
+ ret = !!btrfs_find_name_in_backref(log_eb, log_slot, &name);
if (!ret) {
struct inode *dir;
dir = read_one_inode(root, parent_id);
if (!dir) {
ret = -ENOENT;
- kfree(name);
+ kfree(name.name);
goto out;
}
ret = unlink_inode_for_log_replay(trans, BTRFS_I(dir),
- inode, name, namelen);
- kfree(name);
+ inode, &name);
+ kfree(name.name);
iput(dir);
if (ret)
goto out;
goto again;
}
- kfree(name);
- ref_ptr += namelen;
+ kfree(name.name);
+ ref_ptr += name.len;
if (key->type == BTRFS_INODE_EXTREF_KEY)
ref_ptr += sizeof(struct btrfs_inode_extref);
else
struct inode *inode = NULL;
unsigned long ref_ptr;
unsigned long ref_end;
- char *name = NULL;
- int namelen;
+ struct fscrypt_str name;
int ret;
int log_ref_ver = 0;
u64 parent_objectid;
while (ref_ptr < ref_end) {
if (log_ref_ver) {
- ret = extref_get_fields(eb, ref_ptr, &namelen, &name,
+ ret = extref_get_fields(eb, ref_ptr, &name,
&ref_index, &parent_objectid);
/*
* parent object can change from one array
goto out;
}
} else {
- ret = ref_get_fields(eb, ref_ptr, &namelen, &name,
- &ref_index);
+ ret = ref_get_fields(eb, ref_ptr, &name, &ref_index);
}
if (ret)
goto out;
ret = inode_in_dir(root, path, btrfs_ino(BTRFS_I(dir)),
- btrfs_ino(BTRFS_I(inode)), ref_index,
- name, namelen);
+ btrfs_ino(BTRFS_I(inode)), ref_index, &name);
if (ret < 0) {
goto out;
} else if (ret == 0) {
ret = __add_inode_ref(trans, root, path, log,
BTRFS_I(dir), BTRFS_I(inode),
inode_objectid, parent_objectid,
- ref_index, name, namelen);
+ ref_index, &name);
if (ret) {
if (ret == 1)
ret = 0;
/* insert our name */
ret = btrfs_add_link(trans, BTRFS_I(dir), BTRFS_I(inode),
- name, namelen, 0, ref_index);
+ &name, 0, ref_index);
if (ret)
goto out;
}
/* Else, ret == 1, we already have a perfect match, we're done. */
- ref_ptr = (unsigned long)(ref_ptr + ref_struct_size) + namelen;
- kfree(name);
- name = NULL;
+ ref_ptr = (unsigned long)(ref_ptr + ref_struct_size) + name.len;
+ kfree(name.name);
+ name.name = NULL;
if (log_ref_ver) {
iput(dir);
dir = NULL;
ret = overwrite_item(trans, root, path, eb, slot, key);
out:
btrfs_release_path(path);
- kfree(name);
+ kfree(name.name);
iput(dir);
iput(inode);
return ret;
static noinline int insert_one_name(struct btrfs_trans_handle *trans,
struct btrfs_root *root,
u64 dirid, u64 index,
- char *name, int name_len,
+ const struct fscrypt_str *name,
struct btrfs_key *location)
{
struct inode *inode;
}
ret = btrfs_add_link(trans, BTRFS_I(dir), BTRFS_I(inode), name,
- name_len, 1, index);
+ 1, index);
/* FIXME, put inode into FIXUP list */
struct btrfs_path *path,
struct btrfs_dir_item *dst_di,
const struct btrfs_key *log_key,
- u8 log_type,
+ u8 log_flags,
bool exists)
{
struct btrfs_key found_key;
if (found_key.objectid == log_key->objectid &&
found_key.type == log_key->type &&
found_key.offset == log_key->offset &&
- btrfs_dir_type(path->nodes[0], dst_di) == log_type)
+ btrfs_dir_flags(path->nodes[0], dst_di) == log_flags)
return 1;
/*
struct btrfs_dir_item *di,
struct btrfs_key *key)
{
- char *name;
- int name_len;
+ struct fscrypt_str name;
struct btrfs_dir_item *dir_dst_di;
struct btrfs_dir_item *index_dst_di;
bool dir_dst_matches = false;
struct btrfs_key log_key;
struct btrfs_key search_key;
struct inode *dir;
- u8 log_type;
+ u8 log_flags;
bool exists;
int ret;
bool update_size = true;
if (!dir)
return -EIO;
- name_len = btrfs_dir_name_len(eb, di);
- name = kmalloc(name_len, GFP_NOFS);
- if (!name) {
- ret = -ENOMEM;
+ ret = read_alloc_one_name(eb, di + 1, btrfs_dir_name_len(eb, di), &name);
+ if (ret)
goto out;
- }
-
- log_type = btrfs_dir_type(eb, di);
- read_extent_buffer(eb, name, (unsigned long)(di + 1),
- name_len);
+ log_flags = btrfs_dir_flags(eb, di);
btrfs_dir_item_key_to_cpu(eb, di, &log_key);
ret = btrfs_lookup_inode(trans, root, path, &log_key, 0);
btrfs_release_path(path);
ret = 0;
dir_dst_di = btrfs_lookup_dir_item(trans, root, path, key->objectid,
- name, name_len, 1);
+ &name, 1);
if (IS_ERR(dir_dst_di)) {
ret = PTR_ERR(dir_dst_di);
goto out;
} else if (dir_dst_di) {
ret = delete_conflicting_dir_entry(trans, BTRFS_I(dir), path,
- dir_dst_di, &log_key, log_type,
- exists);
+ dir_dst_di, &log_key,
+ log_flags, exists);
if (ret < 0)
goto out;
dir_dst_matches = (ret == 1);
index_dst_di = btrfs_lookup_dir_index_item(trans, root, path,
key->objectid, key->offset,
- name, name_len, 1);
+ &name, 1);
if (IS_ERR(index_dst_di)) {
ret = PTR_ERR(index_dst_di);
goto out;
} else if (index_dst_di) {
ret = delete_conflicting_dir_entry(trans, BTRFS_I(dir), path,
index_dst_di, &log_key,
- log_type, exists);
+ log_flags, exists);
if (ret < 0)
goto out;
index_dst_matches = (ret == 1);
search_key.objectid = log_key.objectid;
search_key.type = BTRFS_INODE_REF_KEY;
search_key.offset = key->objectid;
- ret = backref_in_log(root->log_root, &search_key, 0, name, name_len);
+ ret = backref_in_log(root->log_root, &search_key, 0, &name);
if (ret < 0) {
goto out;
} else if (ret) {
search_key.objectid = log_key.objectid;
search_key.type = BTRFS_INODE_EXTREF_KEY;
search_key.offset = key->objectid;
- ret = backref_in_log(root->log_root, &search_key, key->objectid, name,
- name_len);
+ ret = backref_in_log(root->log_root, &search_key, key->objectid, &name);
if (ret < 0) {
goto out;
} else if (ret) {
}
btrfs_release_path(path);
ret = insert_one_name(trans, root, key->objectid, key->offset,
- name, name_len, &log_key);
+ &name, &log_key);
if (ret && ret != -ENOENT && ret != -EEXIST)
goto out;
if (!ret)
out:
if (!ret && update_size) {
- btrfs_i_size_write(BTRFS_I(dir), dir->i_size + name_len * 2);
+ btrfs_i_size_write(BTRFS_I(dir), dir->i_size + name.len * 2);
ret = btrfs_update_inode(trans, root, BTRFS_I(dir));
}
- kfree(name);
+ kfree(name.name);
iput(dir);
if (!ret && name_added)
ret = 1;
* to ever delete the parent directory has it would result in stale
* dentries that can never be deleted.
*/
- if (ret == 1 && btrfs_dir_type(eb, di) != BTRFS_FT_DIR) {
+ if (ret == 1 && btrfs_dir_ftype(eb, di) != BTRFS_FT_DIR) {
struct btrfs_path *fixup_path;
struct btrfs_key di_key;
struct extent_buffer *eb;
int slot;
struct btrfs_dir_item *di;
- int name_len;
- char *name;
+ struct fscrypt_str name;
struct inode *inode = NULL;
struct btrfs_key location;
eb = path->nodes[0];
slot = path->slots[0];
di = btrfs_item_ptr(eb, slot, struct btrfs_dir_item);
- name_len = btrfs_dir_name_len(eb, di);
- name = kmalloc(name_len, GFP_NOFS);
- if (!name) {
- ret = -ENOMEM;
+ ret = read_alloc_one_name(eb, di + 1, btrfs_dir_name_len(eb, di), &name);
+ if (ret)
goto out;
- }
-
- read_extent_buffer(eb, name, (unsigned long)(di + 1), name_len);
if (log) {
struct btrfs_dir_item *log_di;
log_di = btrfs_lookup_dir_index_item(trans, log, log_path,
dir_key->objectid,
- dir_key->offset,
- name, name_len, 0);
+ dir_key->offset, &name, 0);
if (IS_ERR(log_di)) {
ret = PTR_ERR(log_di);
goto out;
inc_nlink(inode);
ret = unlink_inode_for_log_replay(trans, BTRFS_I(dir), BTRFS_I(inode),
- name, name_len);
+ &name);
/*
* Unlike dir item keys, dir index keys can only have one name (entry) in
* them, as there are no key collisions since each key has a unique offset
out:
btrfs_release_path(path);
btrfs_release_path(log_path);
- kfree(name);
+ kfree(name.name);
iput(inode);
return ret;
}
struct walk_control *wc, u64 gen, int level)
{
int nritems;
+ struct btrfs_tree_parent_check check = {
+ .transid = gen,
+ .level = level
+ };
struct btrfs_path *path;
struct btrfs_root *root = wc->replay_dest;
struct btrfs_key key;
int i;
int ret;
- ret = btrfs_read_extent_buffer(eb, gen, level, NULL);
+ ret = btrfs_read_extent_buffer(eb, &check);
if (ret)
return ret;
int ret = 0;
while (*level > 0) {
- struct btrfs_key first_key;
+ struct btrfs_tree_parent_check check = { 0 };
cur = path->nodes[*level];
bytenr = btrfs_node_blockptr(cur, path->slots[*level]);
ptr_gen = btrfs_node_ptr_generation(cur, path->slots[*level]);
- btrfs_node_key_to_cpu(cur, &first_key, path->slots[*level]);
+ check.transid = ptr_gen;
+ check.level = *level - 1;
+ check.has_first_key = true;
+ btrfs_node_key_to_cpu(cur, &check.first_key, path->slots[*level]);
blocksize = fs_info->nodesize;
next = btrfs_find_create_tree_block(fs_info, bytenr,
path->slots[*level]++;
if (wc->free) {
- ret = btrfs_read_extent_buffer(next, ptr_gen,
- *level - 1, &first_key);
+ ret = btrfs_read_extent_buffer(next, &check);
if (ret) {
free_extent_buffer(next);
return ret;
free_extent_buffer(next);
continue;
}
- ret = btrfs_read_extent_buffer(next, ptr_gen, *level - 1, &first_key);
+ ret = btrfs_read_extent_buffer(next, &check);
if (ret) {
free_extent_buffer(next);
return ret;
struct btrfs_root *log,
struct btrfs_path *path,
u64 dir_ino,
- const char *name, int name_len,
+ const struct fscrypt_str *name,
u64 index)
{
struct btrfs_dir_item *di;
* for dir item keys.
*/
di = btrfs_lookup_dir_index_item(trans, log, path, dir_ino,
- index, name, name_len, -1);
+ index, name, -1);
if (IS_ERR(di))
return PTR_ERR(di);
else if (!di)
*/
void btrfs_del_dir_entries_in_log(struct btrfs_trans_handle *trans,
struct btrfs_root *root,
- const char *name, int name_len,
+ const struct fscrypt_str *name,
struct btrfs_inode *dir, u64 index)
{
struct btrfs_path *path;
}
ret = del_logged_dentry(trans, root->log_root, path, btrfs_ino(dir),
- name, name_len, index);
+ name, index);
btrfs_free_path(path);
out_unlock:
mutex_unlock(&dir->log_mutex);
/* see comments for btrfs_del_dir_entries_in_log */
void btrfs_del_inode_ref_in_log(struct btrfs_trans_handle *trans,
struct btrfs_root *root,
- const char *name, int name_len,
+ const struct fscrypt_str *name,
struct btrfs_inode *inode, u64 dirid)
{
struct btrfs_root *log;
log = root->log_root;
mutex_lock(&inode->log_mutex);
- ret = btrfs_del_inode_ref(trans, log, name, name_len, btrfs_ino(inode),
+ ret = btrfs_del_inode_ref(trans, log, name, btrfs_ino(inode),
dirid, &index);
mutex_unlock(&inode->log_mutex);
if (ret < 0 && ret != -ENOENT)
u64 *last_old_dentry_offset)
{
struct btrfs_root *log = inode->root->log_root;
- struct extent_buffer *src = path->nodes[0];
- const int nritems = btrfs_header_nritems(src);
+ struct extent_buffer *src;
+ const int nritems = btrfs_header_nritems(path->nodes[0]);
const u64 ino = btrfs_ino(inode);
bool last_found = false;
int batch_start = 0;
int batch_size = 0;
int i;
- for (i = path->slots[0]; i < nritems; i++) {
+ /*
+ * We need to clone the leaf, release the read lock on it, and use the
+ * clone before modifying the log tree. See the comment at copy_items()
+ * about why we need to do this.
+ */
+ src = btrfs_clone_extent_buffer(path->nodes[0]);
+ if (!src)
+ return -ENOMEM;
+
+ i = path->slots[0];
+ btrfs_release_path(path);
+ path->nodes[0] = src;
+ path->slots[0] = i;
+
+ for (; i < nritems; i++) {
struct btrfs_dir_item *di;
struct btrfs_key key;
int ret;
{
struct btrfs_root *log = inode->root->log_root;
struct btrfs_file_extent_item *extent;
- struct extent_buffer *src = src_path->nodes[0];
+ struct extent_buffer *src;
int ret = 0;
struct btrfs_key *ins_keys;
u32 *ins_sizes;
const bool skip_csum = (inode->flags & BTRFS_INODE_NODATASUM);
const u64 i_size = i_size_read(&inode->vfs_inode);
+ /*
+ * To keep lockdep happy and avoid deadlocks, clone the source leaf and
+ * use the clone. This is because otherwise we would be changing the log
+ * tree, to insert items from the subvolume tree or insert csum items,
+ * while holding a read lock on a leaf from the subvolume tree, which
+ * creates a nasty lock dependency when COWing log tree nodes/leaves:
+ *
+ * 1) Modifying the log tree triggers an extent buffer allocation while
+ * holding a write lock on a parent extent buffer from the log tree.
+ * Allocating the pages for an extent buffer, or the extent buffer
+ * struct, can trigger inode eviction and finally the inode eviction
+ * will trigger a release/remove of a delayed node, which requires
+ * taking the delayed node's mutex;
+ *
+ * 2) Allocating a metadata extent for a log tree can trigger the async
+ * reclaim thread and make us wait for it to release enough space and
+ * unblock our reservation ticket. The reclaim thread can start
+ * flushing delayed items, and that in turn results in the need to
+ * lock delayed node mutexes and in the need to write lock extent
+ * buffers of a subvolume tree - all this while holding a write lock
+ * on the parent extent buffer in the log tree.
+ *
+ * So one task in scenario 1) running in parallel with another task in
+ * scenario 2) could lead to a deadlock, one wanting to lock a delayed
+ * node mutex while having a read lock on a leaf from the subvolume,
+ * while the other is holding the delayed node's mutex and wants to
+ * write lock the same subvolume leaf for flushing delayed items.
+ */
+ src = btrfs_clone_extent_buffer(src_path->nodes[0]);
+ if (!src)
+ return -ENOMEM;
+
+ i = src_path->slots[0];
+ btrfs_release_path(src_path);
+ src_path->nodes[0] = src;
+ src_path->slots[0] = i;
+
ins_data = kmalloc(nr * sizeof(struct btrfs_key) +
nr * sizeof(u32), GFP_NOFS);
if (!ins_data)
u32 this_len;
unsigned long name_ptr;
struct btrfs_dir_item *di;
+ struct fscrypt_str name_str;
if (key->type == BTRFS_INODE_REF_KEY) {
struct btrfs_inode_ref *iref;
}
read_extent_buffer(eb, name, name_ptr, this_name_len);
+
+ name_str.name = name;
+ name_str.len = this_name_len;
di = btrfs_lookup_dir_item(NULL, inode->root, search_path,
- parent, name, this_name_len, 0);
+ parent, &name_str, 0);
if (di && !IS_ERR(di)) {
struct btrfs_key di_key;
}
di = btrfs_item_ptr(leaf, i, struct btrfs_dir_item);
- type = btrfs_dir_type(leaf, di);
+ type = btrfs_dir_ftype(leaf, di);
if (btrfs_dir_transid(leaf, di) < trans->transid)
continue;
btrfs_dir_item_key_to_cpu(leaf, di, &di_key);
}
if (!need_log_inode(trans, BTRFS_I(di_inode))) {
- btrfs_add_delayed_iput(di_inode);
+ btrfs_add_delayed_iput(BTRFS_I(di_inode));
break;
}
log_mode = LOG_INODE_ALL;
ret = btrfs_log_inode(trans, BTRFS_I(di_inode),
log_mode, ctx);
- btrfs_add_delayed_iput(di_inode);
+ btrfs_add_delayed_iput(BTRFS_I(di_inode));
if (ret)
goto out;
if (ctx->log_new_dentries) {
* so that the log ends up with the new name and without the old name.
*/
if (!need_log_inode(trans, BTRFS_I(inode))) {
- btrfs_add_delayed_iput(inode);
+ btrfs_add_delayed_iput(BTRFS_I(inode));
return 0;
}
- btrfs_add_delayed_iput(inode);
+ btrfs_add_delayed_iput(BTRFS_I(inode));
ino_elem = kmalloc(sizeof(*ino_elem), GFP_NOFS);
if (!ino_elem)
*/
ret = btrfs_log_inode(trans, BTRFS_I(inode),
LOG_INODE_ALL, ctx);
- btrfs_add_delayed_iput(inode);
+ btrfs_add_delayed_iput(BTRFS_I(inode));
if (ret)
break;
continue;
* that, we can avoid doing it again.
*/
if (!need_log_inode(trans, BTRFS_I(inode))) {
- btrfs_add_delayed_iput(inode);
+ btrfs_add_delayed_iput(BTRFS_I(inode));
continue;
}
* log with the new name before we unpin it.
*/
ret = btrfs_log_inode(trans, BTRFS_I(inode), LOG_INODE_EXISTS, ctx);
- btrfs_add_delayed_iput(inode);
+ btrfs_add_delayed_iput(BTRFS_I(inode));
if (ret)
break;
}
}
if (!need_log_inode(trans, BTRFS_I(di_inode))) {
- btrfs_add_delayed_iput(di_inode);
+ btrfs_add_delayed_iput(BTRFS_I(di_inode));
continue;
}
- if (btrfs_stack_dir_type(dir_item) == BTRFS_FT_DIR)
+ if (btrfs_stack_dir_ftype(dir_item) == BTRFS_FT_DIR)
log_mode = LOG_INODE_ALL;
ctx->log_new_dentries = false;
if (!ret && ctx->log_new_dentries)
ret = log_new_dir_dentries(trans, BTRFS_I(di_inode), ctx);
- btrfs_add_delayed_iput(di_inode);
+ btrfs_add_delayed_iput(BTRFS_I(di_inode));
if (ret)
break;
}
if (!need_log_inode(trans, BTRFS_I(dir_inode))) {
- btrfs_add_delayed_iput(dir_inode);
+ btrfs_add_delayed_iput(BTRFS_I(dir_inode));
continue;
}
if (!ret && ctx->log_new_dentries)
ret = log_new_dir_dentries(trans,
BTRFS_I(dir_inode), ctx);
- btrfs_add_delayed_iput(dir_inode);
+ btrfs_add_delayed_iput(BTRFS_I(dir_inode));
if (ret)
goto out;
}
need_log_inode(trans, BTRFS_I(inode)))
ret = btrfs_log_inode(trans, BTRFS_I(inode),
LOG_INODE_EXISTS, ctx);
- btrfs_add_delayed_iput(inode);
+ btrfs_add_delayed_iput(BTRFS_I(inode));
if (ret)
return ret;
mutex_unlock(&dir->log_mutex);
}
-/**
+/*
* Update the log after adding a new name for an inode.
*
* @trans: Transaction handle.
if (old_dir && old_dir->logged_trans == trans->transid) {
struct btrfs_root *log = old_dir->root->log_root;
struct btrfs_path *path;
+ struct fscrypt_name fname;
ASSERT(old_dir_index >= BTRFS_DIR_START_INDEX);
+ ret = fscrypt_setup_filename(&old_dir->vfs_inode,
+ &old_dentry->d_name, 0, &fname);
+ if (ret)
+ goto out;
/*
* We have two inodes to update in the log, the old directory and
* the inode that got renamed, so we must pin the log to prevent
path = btrfs_alloc_path();
if (!path) {
ret = -ENOMEM;
+ fscrypt_free_filename(&fname);
goto out;
}
*/
mutex_lock(&old_dir->log_mutex);
ret = del_logged_dentry(trans, log, path, btrfs_ino(old_dir),
- old_dentry->d_name.name,
- old_dentry->d_name.len, old_dir_index);
+ &fname.disk_name, old_dir_index);
if (ret > 0) {
/*
* The dentry does not exist in the log, so record its
mutex_unlock(&old_dir->log_mutex);
btrfs_free_path(path);
+ fscrypt_free_filename(&fname);
if (ret < 0)
goto out;
}