diff --git a/drivers/infiniband/hw/mthca/mthca_provider.h b/drivers/infiniband/hw/mthca/mthca_provider.h
index 8de2887ba15c..9a5bece3fa5c 100644
--- a/drivers/infiniband/hw/mthca/mthca_provider.h
+++ b/drivers/infiniband/hw/mthca/mthca_provider.h
@@ -136,8 +136,8 @@ struct mthca_ah {
  * We have one global lock that protects dev->cq/qp_table.  Each
  * struct mthca_cq/qp also has its own lock.  An individual qp lock
  * may be taken inside of an individual cq lock.  Both cqs attached to
- * a qp may be locked, with the send cq locked first.  No other
- * nesting should be done.
+ * a qp may be locked, with the cq with the lower cqn locked first.
+ * No other nesting should be done.
  *
  * Each struct mthca_cq/qp also has an ref count, protected by the
  * corresponding table lock.  The pointer from the cq/qp_table to the
diff --git a/drivers/infiniband/hw/mthca/mthca_qp.c b/drivers/infiniband/hw/mthca/mthca_qp.c
index 157b4f8ac407..2e8f6f36e0a5 100644
--- a/drivers/infiniband/hw/mthca/mthca_qp.c
+++ b/drivers/infiniband/hw/mthca/mthca_qp.c
@@ -1263,6 +1263,32 @@ int mthca_alloc_qp(struct mthca_dev *dev,
 	return 0;
 }
 
+static void mthca_lock_cqs(struct mthca_cq *send_cq, struct mthca_cq *recv_cq)
+{
+	if (send_cq == recv_cq)
+		spin_lock_irq(&send_cq->lock);
+	else if (send_cq->cqn < recv_cq->cqn) {
+		spin_lock_irq(&send_cq->lock);
+		spin_lock_nested(&recv_cq->lock, SINGLE_DEPTH_NESTING);
+	} else {
+		spin_lock_irq(&recv_cq->lock);
+		spin_lock_nested(&send_cq->lock, SINGLE_DEPTH_NESTING);
+	}
+}
+
+static void mthca_unlock_cqs(struct mthca_cq *send_cq, struct mthca_cq *recv_cq)
+{
+	if (send_cq == recv_cq)
+		spin_unlock_irq(&send_cq->lock);
+	else if (send_cq->cqn < recv_cq->cqn) {
+		spin_unlock(&recv_cq->lock);
+		spin_unlock_irq(&send_cq->lock);
+	} else {
+		spin_unlock(&send_cq->lock);
+		spin_unlock_irq(&recv_cq->lock);
+	}
+}
+
 int mthca_alloc_sqp(struct mthca_dev *dev,
 		    struct mthca_pd *pd,
 		    struct mthca_cq *send_cq,
@@ -1315,17 +1341,13 @@ int mthca_alloc_sqp(struct mthca_dev *dev,
 	 * Lock CQs here, so that CQ polling code can do QP lookup
 	 * without taking a lock.
 	 */
-	spin_lock_irq(&send_cq->lock);
-	if (send_cq != recv_cq)
-		spin_lock(&recv_cq->lock);
+	mthca_lock_cqs(send_cq, recv_cq);
 
 	spin_lock(&dev->qp_table.lock);
 	mthca_array_clear(&dev->qp_table.qp, mqpn);
 	spin_unlock(&dev->qp_table.lock);
 
-	if (send_cq != recv_cq)
-		spin_unlock(&recv_cq->lock);
-	spin_unlock_irq(&send_cq->lock);
+	mthca_unlock_cqs(send_cq, recv_cq);
 
  err_out:
 	dma_free_coherent(&dev->pdev->dev, sqp->header_buf_size,
@@ -1359,9 +1381,7 @@ void mthca_free_qp(struct mthca_dev *dev,
 	 * Lock CQs here, so that CQ polling code can do QP lookup
 	 * without taking a lock.
 	 */
-	spin_lock_irq(&send_cq->lock);
-	if (send_cq != recv_cq)
-		spin_lock(&recv_cq->lock);
+	mthca_lock_cqs(send_cq, recv_cq);
 
 	spin_lock(&dev->qp_table.lock);
 	mthca_array_clear(&dev->qp_table.qp,
@@ -1369,9 +1389,7 @@ void mthca_free_qp(struct mthca_dev *dev,
 	--qp->refcount;
 	spin_unlock(&dev->qp_table.lock);
 
-	if (send_cq != recv_cq)
-		spin_unlock(&recv_cq->lock);
-	spin_unlock_irq(&send_cq->lock);
+	mthca_unlock_cqs(send_cq, recv_cq);
 
 	wait_event(qp->wait, !get_qp_refcount(dev, qp));