The code was using dlm->spinlock instead of dlm->ast_lock to protect the
ast_list. This patch fixes the issue.
Signed-off-by: Sunil Mushran <sunil.mushran@oracle.com>
Acked-by: Joel Becker <joel.becker@oracle.com>
Signed-off-by: Mark Fasheh <mfasheh@suse.com>
        else
                BUG_ON(res->owner == dlm->node_num);
 
-       spin_lock(&dlm->spinlock);
+       spin_lock(&dlm->ast_lock);
        /* We want to be sure that we're not freeing a lock
         * that still has AST's pending... */
        in_use = !list_empty(&lock->ast_list);
-       spin_unlock(&dlm->spinlock);
+       spin_unlock(&dlm->ast_lock);
        if (in_use) {
               mlog(ML_ERROR, "lockres %.*s: Someone is calling dlmunlock "
                    "while waiting for an ast!", res->lockname.len,