On 5/30/2016 11:09 PM, Or Gerlitz wrote:

On Mon, May 23, 2016 at 3:20 PM, Yishai Hadas [off-list ref] wrote:

quoted

+int ib_destroy_wq(struct ib_wq *wq)
+{
+       int err;
+       struct ib_cq *cq = wq->cq;
+       struct ib_pd *pd = wq->pd;
+
+       if (atomic_read(&wq->usecnt))
+               return -EBUSY;
+
+       err = wq->device->destroy_wq(wq);
+       if (!err) {
+               atomic_dec(&pd->usecnt);
+               atomic_dec(&cq->usecnt);
+       }
+       return err;

by the reverse unwinding principle

quoted

+               atomic_inc(&pd->usecnt);
+               atomic_inc(&wq_attr->cq->usecnt);

-->

quoted

+               atomic_dec(&cq->usecnt);
+               atomic_dec(&pd->usecnt);

The PD and the CQ are two unrelated objects, we are not dealing with 
locking order so the reverse principle is not required. Look also at 
ib_create/destroy QP for PD/CQ usage, behavior is the same.
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html