summaryrefslogtreecommitdiff
path: root/fs/btrfs
diff options
context:
space:
mode:
Diffstat (limited to 'fs/btrfs')
-rw-r--r--fs/btrfs/btrfs_inode.h2
-rw-r--r--fs/btrfs/extent_io.c40
-rw-r--r--fs/btrfs/inode.c56
-rw-r--r--fs/btrfs/qgroup.c6
-rw-r--r--fs/btrfs/super.c9
-rw-r--r--fs/btrfs/tree-log.c78
-rw-r--r--fs/btrfs/volumes.c5
7 files changed, 132 insertions, 64 deletions
diff --git a/fs/btrfs/btrfs_inode.h b/fs/btrfs/btrfs_inode.h
index b99fb0273292..0387b9f43a52 100644
--- a/fs/btrfs/btrfs_inode.h
+++ b/fs/btrfs/btrfs_inode.h
@@ -248,7 +248,7 @@ struct btrfs_inode {
u64 new_delalloc_bytes;
/*
* The offset of the last dir index key that was logged.
- * This is used only for directories.
+ * This is used only for directories. Protected by 'log_mutex'.
*/
u64 last_dir_index_offset;
};
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index c953297aa89a..b21cb72835cc 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -111,6 +111,24 @@ struct btrfs_bio_ctrl {
*/
unsigned long submit_bitmap;
struct readahead_control *ractl;
+
+ /*
+ * The start offset of the last used extent map by a read operation.
+ *
+ * This is for proper compressed read merge.
+ * U64_MAX means we are starting the read and have made no progress yet.
+ *
+ * The current btrfs_bio_is_contig() only uses disk_bytenr as
+ * the condition to check if the read can be merged with previous
+ * bio, which is not correct. E.g. two file extents pointing to the
+ * same extent but with different offset.
+ *
+ * So here we need to do extra checks to only merge reads that are
+ * covered by the same extent map.
+ * Just extent_map::start will be enough, as they are unique
+ * inside the same inode.
+ */
+ u64 last_em_start;
};
static void submit_one_bio(struct btrfs_bio_ctrl *bio_ctrl)
@@ -909,7 +927,7 @@ static void btrfs_readahead_expand(struct readahead_control *ractl,
* return 0 on success, otherwise return error
*/
static int btrfs_do_readpage(struct folio *folio, struct extent_map **em_cached,
- struct btrfs_bio_ctrl *bio_ctrl, u64 *prev_em_start)
+ struct btrfs_bio_ctrl *bio_ctrl)
{
struct inode *inode = folio->mapping->host;
struct btrfs_fs_info *fs_info = inode_to_fs_info(inode);
@@ -1019,12 +1037,11 @@ static int btrfs_do_readpage(struct folio *folio, struct extent_map **em_cached,
* non-optimal behavior (submitting 2 bios for the same extent).
*/
if (compress_type != BTRFS_COMPRESS_NONE &&
- prev_em_start && *prev_em_start != (u64)-1 &&
- *prev_em_start != em->start)
+ bio_ctrl->last_em_start != U64_MAX &&
+ bio_ctrl->last_em_start != em->start)
force_bio_submit = true;
- if (prev_em_start)
- *prev_em_start = em->start;
+ bio_ctrl->last_em_start = em->start;
btrfs_free_extent_map(em);
em = NULL;
@@ -1238,12 +1255,15 @@ int btrfs_read_folio(struct file *file, struct folio *folio)
const u64 start = folio_pos(folio);
const u64 end = start + folio_size(folio) - 1;
struct extent_state *cached_state = NULL;
- struct btrfs_bio_ctrl bio_ctrl = { .opf = REQ_OP_READ };
+ struct btrfs_bio_ctrl bio_ctrl = {
+ .opf = REQ_OP_READ,
+ .last_em_start = U64_MAX,
+ };
struct extent_map *em_cached = NULL;
int ret;
lock_extents_for_read(inode, start, end, &cached_state);
- ret = btrfs_do_readpage(folio, &em_cached, &bio_ctrl, NULL);
+ ret = btrfs_do_readpage(folio, &em_cached, &bio_ctrl);
btrfs_unlock_extent(&inode->io_tree, start, end, &cached_state);
btrfs_free_extent_map(em_cached);
@@ -2583,7 +2603,8 @@ void btrfs_readahead(struct readahead_control *rac)
{
struct btrfs_bio_ctrl bio_ctrl = {
.opf = REQ_OP_READ | REQ_RAHEAD,
- .ractl = rac
+ .ractl = rac,
+ .last_em_start = U64_MAX,
};
struct folio *folio;
struct btrfs_inode *inode = BTRFS_I(rac->mapping->host);
@@ -2591,12 +2612,11 @@ void btrfs_readahead(struct readahead_control *rac)
const u64 end = start + readahead_length(rac) - 1;
struct extent_state *cached_state = NULL;
struct extent_map *em_cached = NULL;
- u64 prev_em_start = (u64)-1;
lock_extents_for_read(inode, start, end, &cached_state);
while ((folio = readahead_folio(rac)) != NULL)
- btrfs_do_readpage(folio, &em_cached, &bio_ctrl, &prev_em_start);
+ btrfs_do_readpage(folio, &em_cached, &bio_ctrl);
btrfs_unlock_extent(&inode->io_tree, start, end, &cached_state);
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 9e4aec7330cb..e7218e78bff4 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -5696,7 +5696,17 @@ static void btrfs_del_inode_from_root(struct btrfs_inode *inode)
bool empty = false;
xa_lock(&root->inodes);
- entry = __xa_erase(&root->inodes, btrfs_ino(inode));
+ /*
+ * This btrfs_inode is being freed and has already been unhashed at this
+ * point. It's possible that another btrfs_inode has already been
+ * allocated for the same inode and inserted itself into the root, so
+ * don't delete it in that case.
+ *
+ * Note that this shouldn't need to allocate memory, so the gfp flags
+ * don't really matter.
+ */
+ entry = __xa_cmpxchg(&root->inodes, btrfs_ino(inode), inode, NULL,
+ GFP_ATOMIC);
if (entry == inode)
empty = xa_empty(&root->inodes);
xa_unlock(&root->inodes);
@@ -6805,7 +6815,6 @@ static int btrfs_link(struct dentry *old_dentry, struct inode *dir,
struct fscrypt_name fname;
u64 index;
int ret;
- int drop_inode = 0;
/* do not allow sys_link's with other subvols of the same device */
if (btrfs_root_id(root) != btrfs_root_id(BTRFS_I(inode)->root))
@@ -6837,44 +6846,44 @@ static int btrfs_link(struct dentry *old_dentry, struct inode *dir,
/* There are several dir indexes for this inode, clear the cache. */
BTRFS_I(inode)->dir_index = 0ULL;
- inc_nlink(inode);
inode_inc_iversion(inode);
inode_set_ctime_current(inode);
- ihold(inode);
set_bit(BTRFS_INODE_COPY_EVERYTHING, &BTRFS_I(inode)->runtime_flags);
ret = btrfs_add_link(trans, BTRFS_I(dir), BTRFS_I(inode),
&fname.disk_name, 1, index);
+ if (ret)
+ goto fail;
+ /* Link added now we update the inode item with the new link count. */
+ inc_nlink(inode);
+ ret = btrfs_update_inode(trans, BTRFS_I(inode));
if (ret) {
- drop_inode = 1;
- } else {
- struct dentry *parent = dentry->d_parent;
+ btrfs_abort_transaction(trans, ret);
+ goto fail;
+ }
- ret = btrfs_update_inode(trans, BTRFS_I(inode));
- if (ret)
+ if (inode->i_nlink == 1) {
+ /*
+ * If the new hard link count is 1, it's a file created with the
+ * open(2) O_TMPFILE flag.
+ */
+ ret = btrfs_orphan_del(trans, BTRFS_I(inode));
+ if (ret) {
+ btrfs_abort_transaction(trans, ret);
goto fail;
- if (inode->i_nlink == 1) {
- /*
- * If new hard link count is 1, it's a file created
- * with open(2) O_TMPFILE flag.
- */
- ret = btrfs_orphan_del(trans, BTRFS_I(inode));
- if (ret)
- goto fail;
}
- d_instantiate(dentry, inode);
- btrfs_log_new_name(trans, old_dentry, NULL, 0, parent);
}
+ /* Grab reference for the new dentry passed to d_instantiate(). */
+ ihold(inode);
+ d_instantiate(dentry, inode);
+ btrfs_log_new_name(trans, old_dentry, NULL, 0, dentry->d_parent);
+
fail:
fscrypt_free_filename(&fname);
if (trans)
btrfs_end_transaction(trans);
- if (drop_inode) {
- inode_dec_link_count(inode);
- iput(inode);
- }
btrfs_btree_balance_dirty(fs_info);
return ret;
}
@@ -7830,6 +7839,7 @@ struct inode *btrfs_alloc_inode(struct super_block *sb)
ei->last_sub_trans = 0;
ei->logged_trans = 0;
ei->delalloc_bytes = 0;
+ /* new_delalloc_bytes and last_dir_index_offset are in a union. */
ei->new_delalloc_bytes = 0;
ei->defrag_bytes = 0;
ei->disk_i_size = 0;
diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c
index ccaa9a3cf1ce..da102da169fd 100644
--- a/fs/btrfs/qgroup.c
+++ b/fs/btrfs/qgroup.c
@@ -1455,6 +1455,7 @@ static int __qgroup_excl_accounting(struct btrfs_fs_info *fs_info, u64 ref_root,
struct btrfs_qgroup *qgroup;
LIST_HEAD(qgroup_list);
u64 num_bytes = src->excl;
+ u64 num_bytes_cmpr = src->excl_cmpr;
int ret = 0;
qgroup = find_qgroup_rb(fs_info, ref_root);
@@ -1466,11 +1467,12 @@ static int __qgroup_excl_accounting(struct btrfs_fs_info *fs_info, u64 ref_root,
struct btrfs_qgroup_list *glist;
qgroup->rfer += sign * num_bytes;
- qgroup->rfer_cmpr += sign * num_bytes;
+ qgroup->rfer_cmpr += sign * num_bytes_cmpr;
WARN_ON(sign < 0 && qgroup->excl < num_bytes);
+ WARN_ON(sign < 0 && qgroup->excl_cmpr < num_bytes_cmpr);
qgroup->excl += sign * num_bytes;
- qgroup->excl_cmpr += sign * num_bytes;
+ qgroup->excl_cmpr += sign * num_bytes_cmpr;
if (sign > 0)
qgroup_rsv_add_by_qgroup(fs_info, qgroup, src);
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index a262b494a89f..df1f6cc3fe21 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -299,9 +299,12 @@ static int btrfs_parse_compress(struct btrfs_fs_context *ctx,
btrfs_set_opt(ctx->mount_opt, COMPRESS);
btrfs_clear_opt(ctx->mount_opt, NODATACOW);
btrfs_clear_opt(ctx->mount_opt, NODATASUM);
- } else if (btrfs_match_compress_type(string, "lzo", false)) {
+ } else if (btrfs_match_compress_type(string, "lzo", true)) {
ctx->compress_type = BTRFS_COMPRESS_LZO;
- ctx->compress_level = 0;
+ ctx->compress_level = btrfs_compress_str2level(BTRFS_COMPRESS_LZO,
+ string + 3);
+ if (string[3] == ':' && string[4])
+ btrfs_warn(NULL, "Compression level ignored for LZO");
btrfs_set_opt(ctx->mount_opt, COMPRESS);
btrfs_clear_opt(ctx->mount_opt, NODATACOW);
btrfs_clear_opt(ctx->mount_opt, NODATASUM);
@@ -1079,7 +1082,7 @@ static int btrfs_show_options(struct seq_file *seq, struct dentry *dentry)
seq_printf(seq, ",compress-force=%s", compress_type);
else
seq_printf(seq, ",compress=%s", compress_type);
- if (info->compress_level)
+ if (info->compress_level && info->compress_type != BTRFS_COMPRESS_LZO)
seq_printf(seq, ":%d", info->compress_level);
}
if (btrfs_test_opt(info, NOSSD))
diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index 69e11557fd13..7d5d90845ca9 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -3340,6 +3340,31 @@ int btrfs_free_log_root_tree(struct btrfs_trans_handle *trans,
return 0;
}
+static bool mark_inode_as_not_logged(const struct btrfs_trans_handle *trans,
+ struct btrfs_inode *inode)
+{
+ bool ret = false;
+
+ /*
+ * Do this only if ->logged_trans is still 0 to prevent races with
+ * concurrent logging as we may see the inode not logged when
+ * inode_logged() is called but it gets logged after inode_logged() did
+ * not find it in the log tree and we end up setting ->logged_trans to a
+ * value less than trans->transid after the concurrent logging task has
+ * set it to trans->transid. As a consequence, subsequent rename, unlink
+ * and link operations may end up not logging new names and removing old
+ * names from the log.
+ */
+ spin_lock(&inode->lock);
+ if (inode->logged_trans == 0)
+ inode->logged_trans = trans->transid - 1;
+ else if (inode->logged_trans == trans->transid)
+ ret = true;
+ spin_unlock(&inode->lock);
+
+ return ret;
+}
+
/*
* Check if an inode was logged in the current transaction. This correctly deals
* with the case where the inode was logged but has a logged_trans of 0, which
@@ -3357,15 +3382,32 @@ static int inode_logged(const struct btrfs_trans_handle *trans,
struct btrfs_key key;
int ret;
- if (inode->logged_trans == trans->transid)
+ /*
+ * Quick lockless call, since once ->logged_trans is set to the current
+ * transaction, we never set it to a lower value anywhere else.
+ */
+ if (data_race(inode->logged_trans) == trans->transid)
return 1;
/*
- * If logged_trans is not 0, then we know the inode logged was not logged
- * in this transaction, so we can return false right away.
+ * If logged_trans is not 0 and not trans->transid, then we know the
+ * inode was not logged in this transaction, so we can return false
+ * right away. We take the lock to avoid a race caused by load/store
+ * tearing with a concurrent btrfs_log_inode() call or a concurrent task
+ * in this function further below - an update to trans->transid can be
+ * teared into two 32 bits updates for example, in which case we could
+ * see a positive value that is not trans->transid and assume the inode
+ * was not logged when it was.
*/
- if (inode->logged_trans > 0)
+ spin_lock(&inode->lock);
+ if (inode->logged_trans == trans->transid) {
+ spin_unlock(&inode->lock);
+ return 1;
+ } else if (inode->logged_trans > 0) {
+ spin_unlock(&inode->lock);
return 0;
+ }
+ spin_unlock(&inode->lock);
/*
* If no log tree was created for this root in this transaction, then
@@ -3374,10 +3416,8 @@ static int inode_logged(const struct btrfs_trans_handle *trans,
* transaction's ID, to avoid the search below in a future call in case
* a log tree gets created after this.
*/
- if (!test_bit(BTRFS_ROOT_HAS_LOG_TREE, &inode->root->state)) {
- inode->logged_trans = trans->transid - 1;
- return 0;
- }
+ if (!test_bit(BTRFS_ROOT_HAS_LOG_TREE, &inode->root->state))
+ return mark_inode_as_not_logged(trans, inode);
/*
* We have a log tree and the inode's logged_trans is 0. We can't tell
@@ -3431,8 +3471,7 @@ static int inode_logged(const struct btrfs_trans_handle *trans,
* Set logged_trans to a value greater than 0 and less then the
* current transaction to avoid doing the search in future calls.
*/
- inode->logged_trans = trans->transid - 1;
- return 0;
+ return mark_inode_as_not_logged(trans, inode);
}
/*
@@ -3440,20 +3479,9 @@ static int inode_logged(const struct btrfs_trans_handle *trans,
* the current transacion's ID, to avoid future tree searches as long as
* the inode is not evicted again.
*/
+ spin_lock(&inode->lock);
inode->logged_trans = trans->transid;
-
- /*
- * If it's a directory, then we must set last_dir_index_offset to the
- * maximum possible value, so that the next attempt to log the inode does
- * not skip checking if dir index keys found in modified subvolume tree
- * leaves have been logged before, otherwise it would result in attempts
- * to insert duplicate dir index keys in the log tree. This must be done
- * because last_dir_index_offset is an in-memory only field, not persisted
- * in the inode item or any other on-disk structure, so its value is lost
- * once the inode is evicted.
- */
- if (S_ISDIR(inode->vfs_inode.i_mode))
- inode->last_dir_index_offset = (u64)-1;
+ spin_unlock(&inode->lock);
return 1;
}
@@ -4045,7 +4073,7 @@ done:
/*
* If the inode was logged before and it was evicted, then its
- * last_dir_index_offset is (u64)-1, so we don't the value of the last index
+ * last_dir_index_offset is 0, so we don't know the value of the last index
* key offset. If that's the case, search for it and update the inode. This
* is to avoid lookups in the log tree every time we try to insert a dir index
* key from a leaf changed in the current transaction, and to allow us to always
@@ -4061,7 +4089,7 @@ static int update_last_dir_index_offset(struct btrfs_inode *inode,
lockdep_assert_held(&inode->log_mutex);
- if (inode->last_dir_index_offset != (u64)-1)
+ if (inode->last_dir_index_offset != 0)
return 0;
if (!ctx->logged_before) {
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index fa7a929a0461..c6e3efd6f602 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -2722,6 +2722,11 @@ int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path
goto error;
}
+ if (bdev_nr_bytes(file_bdev(bdev_file)) <= BTRFS_DEVICE_RANGE_RESERVED) {
+ ret = -EINVAL;
+ goto error;
+ }
+
if (fs_devices->seeding) {
seeding_dev = true;
down_write(&sb->s_umount);
generated by cgit v1.2.3 (git 2.46.0) at 2025-09-12 22:22:00 +0000