ext4: fix a memory leak of ext4_free_data
authorChunguang Xu <brookxu@tencent.com>
Sat, 7 Nov 2020 15:58:18 +0000 (23:58 +0800)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 30 Dec 2020 10:54:13 +0000 (11:54 +0100)
commit cca415537244f6102cbb09b5b90db6ae2c953bdd upstream.

When freeing metadata, we will create an ext4_free_data and
insert it into the pending free list.  After the current
transaction is committed, the object will be freed.

ext4_mb_free_metadata() will check whether the area to be freed
overlaps with the pending free list. If true, return directly. At this
time, ext4_free_data is leaked.  Fortunately, the probability of this
problem is small, since it only occurs if the file system is corrupted
such that a block is claimed by more one inode and those inodes are
deleted within a single jbd2 transaction.

Signed-off-by: Chunguang Xu <brookxu@tencent.com>
Link: https://lore.kernel.org/r/1604764698-4269-8-git-send-email-brookxu@tencent.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
fs/ext4/mballoc.c

index 24af9ed..37a619b 100644 (file)
@@ -5126,6 +5126,7 @@ ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b,
                                ext4_group_first_block_no(sb, group) +
                                EXT4_C2B(sbi, cluster),
                                "Block already on to-be-freed list");
+                       kmem_cache_free(ext4_free_data_cachep, new_entry);
                        return 0;
                }
        }