Skip to content

Commit

Permalink
ocfs2: call ocfs2_journal_access_di() before ocfs2_journal_dirty() in…
Browse files Browse the repository at this point in the history
… ocfs2_write_end_nolock()

1: After we call ocfs2_journal_access_di() in ocfs2_write_begin(),
   jbd2_journal_restart() may also be called, in this function transaction
   A's t_updates-- and obtains a new transaction B.  If
   jbd2_journal_commit_transaction() is happened to commit transaction A,
   when t_updates==0, it will continue to complete commit and unfile
   buffer.

   So when jbd2_journal_dirty_metadata(), the handle is pointed a new
   transaction B, and the buffer head's journal head is already freed,
   jh->b_transaction == NULL, jh->b_next_transaction == NULL, it returns
   EINVAL, So it triggers the BUG_ON(status).

thread 1                                          jbd2
ocfs2_write_begin                     jbd2_journal_commit_transaction
ocfs2_write_begin_nolock
  ocfs2_start_trans
    jbd2__journal_start(t_updates+1,
                       transaction A)
    ocfs2_journal_access_di
    ocfs2_write_cluster_by_desc
      ocfs2_mark_extent_written
        ocfs2_change_extent_flag
          ocfs2_split_extent
            ocfs2_extend_rotate_transaction
              jbd2_journal_restart
              (t_updates-1,transaction B) t_updates==0
                                        __jbd2_journal_refile_buffer
                                        (jh->b_transaction = NULL)
ocfs2_write_end
ocfs2_write_end_nolock
    ocfs2_journal_dirty
        jbd2_journal_dirty_metadata(bug)
   ocfs2_commit_trans

2.  In ext4, I found that: jbd2_journal_get_write_access() called by
   ext4_write_end.

ext4_write_begin
    ext4_journal_start
        __ext4_journal_start_sb
            ext4_journal_check_start
            jbd2__journal_start

ext4_write_end
    ext4_mark_inode_dirty
        ext4_reserve_inode_write
            ext4_journal_get_write_access
                jbd2_journal_get_write_access
        ext4_mark_iloc_dirty
            ext4_do_update_inode
                ext4_handle_dirty_metadata
                    jbd2_journal_dirty_metadata

3. So I think we should put ocfs2_journal_access_di before
   ocfs2_journal_dirty in the ocfs2_write_end.  and it works well after my
   modification.

Signed-off-by: vicky <vicky.yangwenfang@huawei.com>
Reviewed-by: Mark Fasheh <mfasheh@suse.de>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Zhangguanghui <zhang.guanghui@h3c.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
  • Loading branch information
yangwenfang authored and torvalds committed Sep 4, 2015
1 parent 40476b8 commit 7f27ec9
Showing 1 changed file with 11 additions and 5 deletions.
16 changes: 11 additions & 5 deletions fs/ocfs2/aops.c
Expand Up @@ -2207,10 +2207,7 @@ int ocfs2_write_begin_nolock(struct file *filp,
if (ret)
goto out_commit;
}
/*
* We don't want this to fail in ocfs2_write_end(), so do it
* here.
*/

ret = ocfs2_journal_access_di(handle, INODE_CACHE(inode), wc->w_di_bh,
OCFS2_JOURNAL_ACCESS_WRITE);
if (ret) {
Expand Down Expand Up @@ -2367,7 +2364,7 @@ int ocfs2_write_end_nolock(struct address_space *mapping,
loff_t pos, unsigned len, unsigned copied,
struct page *page, void *fsdata)
{
int i;
int i, ret;
unsigned from, to, start = pos & (PAGE_CACHE_SIZE - 1);
struct inode *inode = mapping->host;
struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
Expand All @@ -2376,6 +2373,14 @@ int ocfs2_write_end_nolock(struct address_space *mapping,
handle_t *handle = wc->w_handle;
struct page *tmppage;

ret = ocfs2_journal_access_di(handle, INODE_CACHE(inode), wc->w_di_bh,
OCFS2_JOURNAL_ACCESS_WRITE);
if (ret) {
copied = ret;
mlog_errno(ret);
goto out;
}

if (OCFS2_I(inode)->ip_dyn_features & OCFS2_INLINE_DATA_FL) {
ocfs2_write_end_inline(inode, pos, len, &copied, di, wc);
goto out_write_size;
Expand Down Expand Up @@ -2431,6 +2436,7 @@ int ocfs2_write_end_nolock(struct address_space *mapping,
ocfs2_update_inode_fsync_trans(handle, inode, 1);
ocfs2_journal_dirty(handle, wc->w_di_bh);

out:
/* unlock pages before dealloc since it needs acquiring j_trans_barrier
* lock, or it will cause a deadlock since journal commit threads holds
* this lock and will ask for the page lock when flushing the data.
Expand Down

0 comments on commit 7f27ec9

Please sign in to comment.