1 files changed, 48 insertions, 20 deletions

diff --git a/drivers/infiniband/hw/mthca/mthca_qp.c b/drivers/infiniband/hw/mthca/mthca_qp.c
index 16c387d8170c..2e8f6f36e0a5 100644
--- a/drivers/infiniband/hw/mthca/mthca_qp.c
+++ b/drivers/infiniband/hw/mthca/mthca_qp.c
@@ -99,6 +99,10 @@ enum {
         MTHCA_QP_BIT_RSC = 1 <<  3
 };
 
+enum {
+        MTHCA_SEND_DOORBELL_FENCE = 1 << 5
+};
+
 struct mthca_qp_path {
         __be32 port_pkey;
         u8     rnr_retry;
@@ -222,9 +226,8 @@ static void *get_send_wqe(struct mthca_qp *qp, int n)
                          (PAGE_SIZE - 1));
 }
 
-static void mthca_wq_init(struct mthca_wq *wq)
+static void mthca_wq_reset(struct mthca_wq *wq)
 {
-        spin_lock_init(&wq->lock);
         wq->next_ind  = 0;
         wq->last_comp = wq->max - 1;
         wq->head      = 0;
@@ -845,10 +848,10 @@ int mthca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask)
                         mthca_cq_clean(dev, to_mcq(qp->ibqp.recv_cq), qp->qpn,
                                        qp->ibqp.srq ? to_msrq(qp->ibqp.srq) : NULL);
 
-                mthca_wq_init(&qp->sq);
+                mthca_wq_reset(&qp->sq);
                 qp->sq.last = get_send_wqe(qp, qp->sq.max - 1);
 
-                mthca_wq_init(&qp->rq);
+                mthca_wq_reset(&qp->rq);
                 qp->rq.last = get_recv_wqe(qp, qp->rq.max - 1);
 
                 if (mthca_is_memfree(dev)) {
@@ -1112,8 +1115,11 @@ static int mthca_alloc_qp_common(struct mthca_dev *dev,
         qp->atomic_rd_en = 0;
         qp->resp_depth   = 0;
         qp->sq_policy    = send_policy;
-        mthca_wq_init(&qp->sq);
-        mthca_wq_init(&qp->rq);
+        mthca_wq_reset(&qp->sq);
+        mthca_wq_reset(&qp->rq);
+
+        spin_lock_init(&qp->sq.lock);
+        spin_lock_init(&qp->rq.lock);
 
         ret = mthca_map_memfree(dev, qp);
         if (ret)
@@ -1257,6 +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,
@@ -1309,17 +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,
@@ -1353,9 +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,
@@ -1363,9 +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));
 
@@ -1500,7 +1524,7 @@ int mthca_tavor_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
         int i;
         int size;
         int size0 = 0;
-        u32 f0 = 0;
+        u32 f0;
         int ind;
         u8 op0 = 0;
 
@@ -1684,6 +1708,8 @@ int mthca_tavor_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
                 if (!size0) {
                         size0 = size;
                         op0   = mthca_opcode[wr->opcode];
+                        f0    = wr->send_flags & IB_SEND_FENCE ?
+                                MTHCA_SEND_DOORBELL_FENCE : 0;
                 }
 
                 ++ind;
@@ -1841,7 +1867,7 @@ int mthca_arbel_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
         int i;
         int size;
         int size0 = 0;
-        u32 f0 = 0;
+        u32 f0;
         int ind;
         u8 op0 = 0;
 
@@ -2049,6 +2075,8 @@ int mthca_arbel_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
                 if (!size0) {
                         size0 = size;
                         op0   = mthca_opcode[wr->opcode];
+                        f0    = wr->send_flags & IB_SEND_FENCE ?
+                                MTHCA_SEND_DOORBELL_FENCE : 0;
                 }
 
                 ++ind;