diff options
| author | Chris Mason <chris.mason@oracle.com> | 2009-04-24 22:39:24 +0400 |
|---|---|---|
| committer | Chris Mason <chris.mason@oracle.com> | 2009-04-24 23:46:05 +0400 |
| commit | e980b50cda1610f1c17978d9b7fd311a9dd93877 (patch) | |
| tree | a5ab72fa4b791758e25fd6ece6f1aedd29954e7b /fs/btrfs/file.c | |
| parent | 9601e3f6336f6ca66929f451b1f66085e68e36e3 (diff) | |
| download | linux-e980b50cda1610f1c17978d9b7fd311a9dd93877.tar.xz | |
Btrfs: fix fallocate deadlock on inode extent lock
The btrfs fallocate call takes an extent lock on the entire range
being fallocated, and then runs through insert_reserved_extent on each
extent as they are allocated.
The problem with this is that btrfs_drop_extents may decide to try
and take the same extent lock fallocate was already holding. The solution
used here is to push down knowledge of the range that is already locked
going into btrfs_drop_extents.
It turns out that at least one other caller had the same bug.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Diffstat (limited to 'fs/btrfs/file.c')
| -rw-r--r-- | fs/btrfs/file.c | 11 |
1 files changed, 6 insertions, 5 deletions
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 482f8db2cfd0..da3ed965c956 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -363,15 +363,16 @@ out: */ noinline int btrfs_drop_extents(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct inode *inode, - u64 start, u64 end, u64 inline_limit, u64 *hint_byte) + u64 start, u64 end, u64 locked_end, + u64 inline_limit, u64 *hint_byte) { u64 extent_end = 0; - u64 locked_end = end; u64 search_start = start; u64 leaf_start; u64 ram_bytes = 0; u64 orig_parent = 0; u64 disk_bytenr = 0; + u64 orig_locked_end = locked_end; u8 compression; u8 encryption; u16 other_encoding = 0; @@ -684,9 +685,9 @@ next_slot: } out: btrfs_free_path(path); - if (locked_end > end) { - unlock_extent(&BTRFS_I(inode)->io_tree, end, locked_end - 1, - GFP_NOFS); + if (locked_end > orig_locked_end) { + unlock_extent(&BTRFS_I(inode)->io_tree, orig_locked_end, + locked_end - 1, GFP_NOFS); } btrfs_check_file(root, inode); return ret; |