ocfs2: fix a deadlock when commit trans
authorJoseph Qi <joseph.qi@linux.alibaba.com>
Sat, 29 Jan 2022 21:41:27 +0000 (13:41 -0800)
committerLinus Torvalds <torvalds@linux-foundation.org>
Sun, 30 Jan 2022 07:56:58 +0000 (09:56 +0200)
commit 6f1b228529ae introduces a regression which can deadlock as
follows:

  Task1:                              Task2:
  jbd2_journal_commit_transaction     ocfs2_test_bg_bit_allocatable
  spin_lock(&jh->b_state_lock)        jbd_lock_bh_journal_head
  __jbd2_journal_remove_checkpoint    spin_lock(&jh->b_state_lock)
  jbd2_journal_put_journal_head
  jbd_lock_bh_journal_head

Task1 and Task2 lock bh->b_state and jh->b_state_lock in different
order, which finally result in a deadlock.

So use jbd2_journal_[grab|put]_journal_head instead in
ocfs2_test_bg_bit_allocatable() to fix it.

Link: https://lkml.kernel.org/r/20220121071205.100648-3-joseph.qi@linux.alibaba.com
Fixes: 6f1b228529ae ("ocfs2: fix race between searching chunks and release journal_head from buffer_head")
Signed-off-by: Joseph Qi <joseph.qi@linux.alibaba.com>
Reported-by: Gautham Ananthakrishna <gautham.ananthakrishna@oracle.com>
Tested-by: Gautham Ananthakrishna <gautham.ananthakrishna@oracle.com>
Reported-by: Saeed Mirzamohammadi <saeed.mirzamohammadi@oracle.com>
Cc: "Theodore Ts'o" <tytso@mit.edu>
Cc: Andreas Dilger <adilger.kernel@dilger.ca>
Cc: Changwei Ge <gechangwei@live.cn>
Cc: Gang He <ghe@suse.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Jun Piao <piaojun@huawei.com>
Cc: Junxiao Bi <junxiao.bi@oracle.com>
Cc: Mark Fasheh <mark@fasheh.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
fs/ocfs2/suballoc.c

index 481017e1dac5ac64100850f4b8b59f84f2a7c9f6..166c8918c825a60b755816c384ec22bbed3d1b70 100644 (file)
@@ -1251,26 +1251,23 @@ static int ocfs2_test_bg_bit_allocatable(struct buffer_head *bg_bh,
 {
        struct ocfs2_group_desc *bg = (struct ocfs2_group_desc *) bg_bh->b_data;
        struct journal_head *jh;
-       int ret = 1;
+       int ret;
 
        if (ocfs2_test_bit(nr, (unsigned long *)bg->bg_bitmap))
                return 0;
 
-       if (!buffer_jbd(bg_bh))
+       jh = jbd2_journal_grab_journal_head(bg_bh);
+       if (!jh)
                return 1;
 
-       jbd_lock_bh_journal_head(bg_bh);
-       if (buffer_jbd(bg_bh)) {
-               jh = bh2jh(bg_bh);
-               spin_lock(&jh->b_state_lock);
-               bg = (struct ocfs2_group_desc *) jh->b_committed_data;
-               if (bg)
-                       ret = !ocfs2_test_bit(nr, (unsigned long *)bg->bg_bitmap);
-               else
-                       ret = 1;
-               spin_unlock(&jh->b_state_lock);
-       }
-       jbd_unlock_bh_journal_head(bg_bh);
+       spin_lock(&jh->b_state_lock);
+       bg = (struct ocfs2_group_desc *) jh->b_committed_data;
+       if (bg)
+               ret = !ocfs2_test_bit(nr, (unsigned long *)bg->bg_bitmap);
+       else
+               ret = 1;
+       spin_unlock(&jh->b_state_lock);
+       jbd2_journal_put_journal_head(jh);
 
        return ret;
 }