Merge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma

Pull rdma updates from Jason Gunthorpe:
 "A very quiet cycle with few notable changes. Mostly the usual list of
  one or two patches to drivers changing something that isn't quite rc
  worthy. The subsystem seems to be seeing a larger number of rework and
  cleanup style patches right now, I feel that several vendors are
  prepping their drivers for new silicon.

  Summary:

   - Driver updates and cleanup for qedr, bnxt_re, hns, siw, mlx5, mlx4,
     rxe, i40iw

   - Larger series doing cleanup and rework for hns and hfi1.

   - Some general reworking of the CM code to make it a little more
     understandable

   - Unify the different code paths connected to the uverbs FD scheme

   - New UAPI ioctls conversions for get context and get async fd

   - Trace points for CQ and CM portions of the RDMA stack

   - mlx5 driver support for virtio-net formatted rings as RDMA raw
     ethernet QPs

   - verbs support for setting the PCI-E relaxed ordering bit on DMA
     traffic connected to a MR

   - A couple of bug fixes that came too late to make rc7"

* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma: (108 commits)
  RDMA/core: Make the entire API tree static
  RDMA/efa: Mask access flags with the correct optional range
  RDMA/cma: Fix unbalanced cm_id reference count during address resolve
  RDMA/umem: Fix ib_umem_find_best_pgsz()
  IB/mlx4: Fix leak in id_map_find_del
  IB/opa_vnic: Spelling correction of 'erorr' to 'error'
  IB/hfi1: Fix logical condition in msix_request_irq
  RDMA/cm: Remove CM message structs
  RDMA/cm: Use IBA functions for complex structure members
  RDMA/cm: Use IBA functions for simple structure members
  RDMA/cm: Use IBA functions for swapping get/set acessors
  RDMA/cm: Use IBA functions for simple get/set acessors
  RDMA/cm: Add SET/GET implementations to hide IBA wire format
  RDMA/cm: Add accessors for CM_REQ transport_type
  IB/mlx5: Return the administrative GUID if exists
  RDMA/core: Ensure that rdma_user_mmap_entry_remove() is a fence
  IB/mlx4: Fix memory leak in add_gid error flow
  IB/mlx5: Expose RoCE accelerator counters
  RDMA/mlx5: Set relaxed ordering when requested
  RDMA/core: Add the core support field to METHOD_GET_CONTEXT
  ...

1 files changed, 25 insertions, 26 deletions

diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
index 2a2b2112f886..c6e66586e533 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
@@ -74,8 +74,8 @@ static int hns_roce_v1_post_send(struct ib_qp *ibqp,
 	unsigned long flags = 0;
 	void *wqe = NULL;
 	__le32 doorbell[2];
+	u32 wqe_idx = 0;
 	int nreq = 0;
-	u32 ind = 0;
 	int ret = 0;
 	u8 *smac;
 	int loopback;
@@ -88,7 +88,7 @@ static int hns_roce_v1_post_send(struct ib_qp *ibqp,
 	}
 
 	spin_lock_irqsave(&qp->sq.lock, flags);
-	ind = qp->sq_next_wqe;
+
 	for (nreq = 0; wr; ++nreq, wr = wr->next) {
 		if (hns_roce_wq_overflow(&qp->sq, nreq, qp->ibqp.send_cq)) {
 			ret = -ENOMEM;
@@ -96,6 +96,8 @@ static int hns_roce_v1_post_send(struct ib_qp *ibqp,
 			goto out;
 		}
 
+		wqe_idx = (qp->sq.head + nreq) & (qp->sq.wqe_cnt - 1);
+
 		if (unlikely(wr->num_sge > qp->sq.max_gs)) {
 			dev_err(dev, "num_sge=%d > qp->sq.max_gs=%d\n",
 				wr->num_sge, qp->sq.max_gs);
@@ -104,9 +106,8 @@ static int hns_roce_v1_post_send(struct ib_qp *ibqp,
 			goto out;
 		}
 
-		wqe = get_send_wqe(qp, ind & (qp->sq.wqe_cnt - 1));
-		qp->sq.wrid[(qp->sq.head + nreq) & (qp->sq.wqe_cnt - 1)] =
-								      wr->wr_id;
+		wqe = get_send_wqe(qp, wqe_idx);
+		qp->sq.wrid[wqe_idx] = wr->wr_id;
 
 		/* Corresponding to the RC and RD type wqe process separately */
 		if (ibqp->qp_type == IB_QPT_GSI) {
@@ -210,7 +211,6 @@ static int hns_roce_v1_post_send(struct ib_qp *ibqp,
 				       cpu_to_le32((wr->sg_list[1].addr) >> 32);
 			ud_sq_wqe->l_key1 =
 				       cpu_to_le32(wr->sg_list[1].lkey);
-			ind++;
 		} else if (ibqp->qp_type == IB_QPT_RC) {
 			u32 tmp_len = 0;
 
@@ -308,7 +308,6 @@ static int hns_roce_v1_post_send(struct ib_qp *ibqp,
 				ctrl->flag |= cpu_to_le32(wr->num_sge <<
 					      HNS_ROCE_WQE_SGE_NUM_BIT);
 			}
-			ind++;
 		}
 	}
 
@@ -336,7 +335,6 @@ out:
 		doorbell[1] = sq_db.u32_8;
 
 		hns_roce_write64_k(doorbell, qp->sq.db_reg_l);
-		qp->sq_next_wqe = ind;
 	}
 
 	spin_unlock_irqrestore(&qp->sq.lock, flags);
@@ -348,12 +346,6 @@ static int hns_roce_v1_post_recv(struct ib_qp *ibqp,
 				 const struct ib_recv_wr *wr,
 				 const struct ib_recv_wr **bad_wr)
 {
-	int ret = 0;
-	int nreq = 0;
-	int ind = 0;
-	int i = 0;
-	u32 reg_val;
-	unsigned long flags = 0;
 	struct hns_roce_rq_wqe_ctrl *ctrl = NULL;
 	struct hns_roce_wqe_data_seg *scat = NULL;
 	struct hns_roce_qp *hr_qp = to_hr_qp(ibqp);
@@ -361,9 +353,14 @@ static int hns_roce_v1_post_recv(struct ib_qp *ibqp,
 	struct device *dev = &hr_dev->pdev->dev;
 	struct hns_roce_rq_db rq_db;
 	__le32 doorbell[2] = {0};
+	unsigned long flags = 0;
+	unsigned int wqe_idx;
+	int ret = 0;
+	int nreq = 0;
+	int i = 0;
+	u32 reg_val;
 
 	spin_lock_irqsave(&hr_qp->rq.lock, flags);
-	ind = hr_qp->rq.head & (hr_qp->rq.wqe_cnt - 1);
 
 	for (nreq = 0; wr; ++nreq, wr = wr->next) {
 		if (hns_roce_wq_overflow(&hr_qp->rq, nreq,
@@ -373,6 +370,8 @@ static int hns_roce_v1_post_recv(struct ib_qp *ibqp,
 			goto out;
 		}
 
+		wqe_idx = (hr_qp->rq.head + nreq) & (hr_qp->rq.wqe_cnt - 1);
+
 		if (unlikely(wr->num_sge > hr_qp->rq.max_gs)) {
 			dev_err(dev, "rq:num_sge=%d > qp->sq.max_gs=%d\n",
 				wr->num_sge, hr_qp->rq.max_gs);
@@ -381,7 +380,7 @@ static int hns_roce_v1_post_recv(struct ib_qp *ibqp,
 			goto out;
 		}
 
-		ctrl = get_recv_wqe(hr_qp, ind);
+		ctrl = get_recv_wqe(hr_qp, wqe_idx);
 
 		roce_set_field(ctrl->rwqe_byte_12,
 			       RQ_WQE_CTRL_RWQE_BYTE_12_RWQE_SGE_NUM_M,
@@ -393,9 +392,7 @@ static int hns_roce_v1_post_recv(struct ib_qp *ibqp,
 		for (i = 0; i < wr->num_sge; i++)
 			set_data_seg(scat + i, wr->sg_list + i);
 
-		hr_qp->rq.wrid[ind] = wr->wr_id;
-
-		ind = (ind + 1) & (hr_qp->rq.wqe_cnt - 1);
+		hr_qp->rq.wrid[wqe_idx] = wr->wr_id;
 	}
 
 out:
@@ -2701,7 +2698,6 @@ static int hns_roce_v1_m_sqp(struct ib_qp *ibqp, const struct ib_qp_attr *attr,
 		hr_qp->rq.tail = 0;
 		hr_qp->sq.head = 0;
 		hr_qp->sq.tail = 0;
-		hr_qp->sq_next_wqe = 0;
 	}
 
 	kfree(context);
@@ -3315,7 +3311,6 @@ static int hns_roce_v1_m_qp(struct ib_qp *ibqp, const struct ib_qp_attr *attr,
 		hr_qp->rq.tail = 0;
 		hr_qp->sq.head = 0;
 		hr_qp->sq.tail = 0;
-		hr_qp->sq_next_wqe = 0;
 	}
 out:
 	kfree(context);
@@ -3614,14 +3609,18 @@ int hns_roce_v1_destroy_qp(struct ib_qp *ibqp, struct ib_udata *udata)
 	if (ret)
 		return ret;
 
-	send_cq = to_hr_cq(hr_qp->ibqp.send_cq);
-	recv_cq = to_hr_cq(hr_qp->ibqp.recv_cq);
+	send_cq = hr_qp->ibqp.send_cq ? to_hr_cq(hr_qp->ibqp.send_cq) : NULL;
+	recv_cq = hr_qp->ibqp.recv_cq ? to_hr_cq(hr_qp->ibqp.recv_cq) : NULL;
 
 	hns_roce_lock_cqs(send_cq, recv_cq);
 	if (!udata) {
-		__hns_roce_v1_cq_clean(recv_cq, hr_qp->qpn, hr_qp->ibqp.srq ?
-				       to_hr_srq(hr_qp->ibqp.srq) : NULL);
-		if (send_cq != recv_cq)
+		if (recv_cq)
+			__hns_roce_v1_cq_clean(recv_cq, hr_qp->qpn,
+					       (hr_qp->ibqp.srq ?
+						to_hr_srq(hr_qp->ibqp.srq) :
+						NULL));
+
+		if (send_cq && send_cq != recv_cq)
 			__hns_roce_v1_cq_clean(send_cq, hr_qp->qpn, NULL);
 	}
 	hns_roce_unlock_cqs(send_cq, recv_cq);