| Message ID | 558CE8B4.8070904@huawei.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On 2015/6/26 13:52, Joseph Qi wrote: > During direct io the inode will be added to orphan first and then > deleted from orphan. There is a race window that the orphan entry will > be deleted twice and thus trigger the BUG when validating > OCFS2_DIO_ORPHANED_FL in ocfs2_del_inode_from_orphan. > > ocfs2_direct_IO_write > ... > ocfs2_add_inode_to_orphan > >>>>>>>> race window, another node may rm the file and then down, > this node take care of orphan recovery and clear flag > OCFS2_DIO_ORPHANED_FL > ocfs2_del_inode_from_orphan > > Take inode mutex lock to fix this race case. > > Reported-by: Yiwen Jiang <jiangyiwen@huawei.com> > Signed-off-by: Joseph Qi <joseph.qi@huawei.com> Reviewed-by: Weiwei Wang <wangww631@huawei.com> > --- > fs/ocfs2/journal.c | 2 ++ > 1 file changed, 2 insertions(+) > > diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c > index 8017032..f8206d1 100644 > --- a/fs/ocfs2/journal.c > +++ b/fs/ocfs2/journal.c > @@ -2172,6 +2172,7 @@ static int ocfs2_recover_orphans(struct ocfs2_super *osb, > iter = oi->ip_next_orphan; > oi->ip_next_orphan = NULL; > > + mutex_lock(&inode->i_mutex); > ret = ocfs2_rw_lock(inode, 1); > if (ret < 0) { > mlog_errno(ret); > @@ -2218,6 +2219,7 @@ unlock_inode: > unlock_rw: > ocfs2_rw_unlock(inode, 1); > next: > + mutex_unlock(&inode->i_mutex); > iput(inode); > brelse(di_bh); > di_bh = NULL; >
diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c index 8017032..f8206d1 100644 --- a/fs/ocfs2/journal.c +++ b/fs/ocfs2/journal.c @@ -2172,6 +2172,7 @@ static int ocfs2_recover_orphans(struct ocfs2_super *osb, iter = oi->ip_next_orphan; oi->ip_next_orphan = NULL; + mutex_lock(&inode->i_mutex); ret = ocfs2_rw_lock(inode, 1); if (ret < 0) { mlog_errno(ret); @@ -2218,6 +2219,7 @@ unlock_inode: unlock_rw: ocfs2_rw_unlock(inode, 1); next: + mutex_unlock(&inode->i_mutex); iput(inode); brelse(di_bh); di_bh = NULL;
During direct io the inode will be added to orphan first and then deleted from orphan. There is a race window that the orphan entry will be deleted twice and thus trigger the BUG when validating OCFS2_DIO_ORPHANED_FL in ocfs2_del_inode_from_orphan. ocfs2_direct_IO_write ... ocfs2_add_inode_to_orphan >>>>>>>> race window, another node may rm the file and then down, this node take care of orphan recovery and clear flag OCFS2_DIO_ORPHANED_FL ocfs2_del_inode_from_orphan Take inode mutex lock to fix this race case. Reported-by: Yiwen Jiang <jiangyiwen@huawei.com> Signed-off-by: Joseph Qi <joseph.qi@huawei.com> --- fs/ocfs2/journal.c | 2 ++ 1 file changed, 2 insertions(+)