20 files changed, 227 insertions, 131 deletions

diff --git a/drivers/infiniband/hw/efa/efa_verbs.c b/drivers/infiniband/hw/efa/efa_verbs.c
index 08313f7c73bc..7dd082441333 100644
--- a/drivers/infiniband/hw/efa/efa_verbs.c
+++ b/drivers/infiniband/hw/efa/efa_verbs.c
@@ -212,6 +212,7 @@ int efa_query_device(struct ib_device *ibdev,
 	props->max_send_sge = dev_attr->max_sq_sge;
 	props->max_recv_sge = dev_attr->max_rq_sge;
 	props->max_sge_rd = dev_attr->max_wr_rdma_sge;
+	props->max_pkeys = 1;
 
 	if (udata && udata->outlen) {
 		resp.max_sq_sge = dev_attr->max_sq_sge;
diff --git a/drivers/infiniband/hw/hfi1/debugfs.c b/drivers/infiniband/hw/hfi1/debugfs.c
index 4633a0ce1a8c..2ced236e1553 100644
--- a/drivers/infiniband/hw/hfi1/debugfs.c
+++ b/drivers/infiniband/hw/hfi1/debugfs.c
@@ -985,15 +985,10 @@ static ssize_t qsfp2_debugfs_read(struct file *file, char __user *buf,
 static int __i2c_debugfs_open(struct inode *in, struct file *fp, u32 target)
 {
 	struct hfi1_pportdata *ppd;
-	int ret;
 
 	ppd = private2ppd(fp);
 
-	ret = acquire_chip_resource(ppd->dd, i2c_target(target), 0);
-	if (ret) /* failed - release the module */
-		module_put(THIS_MODULE);
-
-	return ret;
+	return acquire_chip_resource(ppd->dd, i2c_target(target), 0);
 }
 
 static int i2c1_debugfs_open(struct inode *in, struct file *fp)
@@ -1013,7 +1008,6 @@ static int __i2c_debugfs_release(struct inode *in, struct file *fp, u32 target)
 	ppd = private2ppd(fp);
 
 	release_chip_resource(ppd->dd, i2c_target(target));
-	module_put(THIS_MODULE);
 
 	return 0;
 }
@@ -1031,18 +1025,10 @@ static int i2c2_debugfs_release(struct inode *in, struct file *fp)
 static int __qsfp_debugfs_open(struct inode *in, struct file *fp, u32 target)
 {
 	struct hfi1_pportdata *ppd;
-	int ret;
-
-	if (!try_module_get(THIS_MODULE))
-		return -ENODEV;
 
 	ppd = private2ppd(fp);
 
-	ret = acquire_chip_resource(ppd->dd, i2c_target(target), 0);
-	if (ret) /* failed - release the module */
-		module_put(THIS_MODULE);
-
-	return ret;
+	return acquire_chip_resource(ppd->dd, i2c_target(target), 0);
 }
 
 static int qsfp1_debugfs_open(struct inode *in, struct file *fp)
@@ -1062,7 +1048,6 @@ static int __qsfp_debugfs_release(struct inode *in, struct file *fp, u32 target)
 	ppd = private2ppd(fp);
 
 	release_chip_resource(ppd->dd, i2c_target(target));
-	module_put(THIS_MODULE);
 
 	return 0;
 }
diff --git a/drivers/infiniband/hw/hfi1/init.c b/drivers/infiniband/hw/hfi1/init.c
index 5eed4360695f..cb7ad1288821 100644
--- a/drivers/infiniband/hw/hfi1/init.c
+++ b/drivers/infiniband/hw/hfi1/init.c
@@ -831,6 +831,29 @@ wq_error:
 }
 
 /**
+ * destroy_workqueues - destroy per port workqueues
+ * @dd: the hfi1_ib device
+ */
+static void destroy_workqueues(struct hfi1_devdata *dd)
+{
+	int pidx;
+	struct hfi1_pportdata *ppd;
+
+	for (pidx = 0; pidx < dd->num_pports; ++pidx) {
+		ppd = dd->pport + pidx;
+
+		if (ppd->hfi1_wq) {
+			destroy_workqueue(ppd->hfi1_wq);
+			ppd->hfi1_wq = NULL;
+		}
+		if (ppd->link_wq) {
+			destroy_workqueue(ppd->link_wq);
+			ppd->link_wq = NULL;
+		}
+	}
+}
+
+/**
  * enable_general_intr() - Enable the IRQs that will be handled by the
  * general interrupt handler.
  * @dd: valid devdata
@@ -1103,15 +1126,10 @@ static void shutdown_device(struct hfi1_devdata *dd)
 		 * We can't count on interrupts since we are stopping.
 		 */
 		hfi1_quiet_serdes(ppd);
-
-		if (ppd->hfi1_wq) {
-			destroy_workqueue(ppd->hfi1_wq);
-			ppd->hfi1_wq = NULL;
-		}
-		if (ppd->link_wq) {
-			destroy_workqueue(ppd->link_wq);
-			ppd->link_wq = NULL;
-		}
+		if (ppd->hfi1_wq)
+			flush_workqueue(ppd->hfi1_wq);
+		if (ppd->link_wq)
+			flush_workqueue(ppd->link_wq);
 	}
 	sdma_exit(dd);
 }
@@ -1756,6 +1774,7 @@ static void remove_one(struct pci_dev *pdev)
 	 * clear dma engines, etc.
 	 */
 	shutdown_device(dd);
+	destroy_workqueues(dd);
 
 	stop_timers(dd);
 
diff --git a/drivers/infiniband/hw/hfi1/iowait.h b/drivers/infiniband/hw/hfi1/iowait.h
index 07847cb72169..d580aa17ae37 100644
--- a/drivers/infiniband/hw/hfi1/iowait.h
+++ b/drivers/infiniband/hw/hfi1/iowait.h
@@ -399,7 +399,7 @@ static inline void iowait_get_priority(struct iowait *w)
  * @wait_head: the wait queue
  *
  * This function is called to insert an iowait struct into a
- * wait queue after a resource (eg, sdma decriptor or pio
+ * wait queue after a resource (eg, sdma descriptor or pio
  * buffer) is run out.
  */
 static inline void iowait_queue(bool pkts_sent, struct iowait *w,
diff --git a/drivers/infiniband/hw/hfi1/ipoib.h b/drivers/infiniband/hw/hfi1/ipoib.h
index 185c9b02c974..b8c9d0a003fb 100644
--- a/drivers/infiniband/hw/hfi1/ipoib.h
+++ b/drivers/infiniband/hw/hfi1/ipoib.h
@@ -67,6 +67,9 @@ struct hfi1_ipoib_circ_buf {
  * @sde: sdma engine
  * @tx_list: tx request list
  * @sent_txreqs: count of txreqs posted to sdma
+ * @stops: count of stops of queue
+ * @ring_full: ring has been filled
+ * @no_desc: descriptor shortage seen
  * @flow: tracks when list needs to be flushed for a flow change
  * @q_idx: ipoib Tx queue index
  * @pkts_sent: indicator packets have been sent from this queue
@@ -80,6 +83,9 @@ struct hfi1_ipoib_txq {
 	struct sdma_engine *sde;
 	struct list_head tx_list;
 	u64 sent_txreqs;
+	atomic_t stops;
+	atomic_t ring_full;
+	atomic_t no_desc;
 	union hfi1_ipoib_flow flow;
 	u8 q_idx;
 	bool pkts_sent;
diff --git a/drivers/infiniband/hw/hfi1/ipoib_tx.c b/drivers/infiniband/hw/hfi1/ipoib_tx.c
index 883cb9d48022..9df292b51a05 100644
--- a/drivers/infiniband/hw/hfi1/ipoib_tx.c
+++ b/drivers/infiniband/hw/hfi1/ipoib_tx.c
@@ -55,23 +55,48 @@ static u64 hfi1_ipoib_txreqs(const u64 sent, const u64 completed)
 	return sent - completed;
 }
 
-static void hfi1_ipoib_check_queue_depth(struct hfi1_ipoib_txq *txq)
+static u64 hfi1_ipoib_used(struct hfi1_ipoib_txq *txq)
 {
-	if (unlikely(hfi1_ipoib_txreqs(++txq->sent_txreqs,
-				       atomic64_read(&txq->complete_txreqs)) >=
-	    min_t(unsigned int, txq->priv->netdev->tx_queue_len,
-		  txq->tx_ring.max_items - 1)))
+	return hfi1_ipoib_txreqs(txq->sent_txreqs,
+				 atomic64_read(&txq->complete_txreqs));
+}
+
+static void hfi1_ipoib_stop_txq(struct hfi1_ipoib_txq *txq)
+{
+	if (atomic_inc_return(&txq->stops) == 1)
 		netif_stop_subqueue(txq->priv->netdev, txq->q_idx);
 }
 
+static void hfi1_ipoib_wake_txq(struct hfi1_ipoib_txq *txq)
+{
+	if (atomic_dec_and_test(&txq->stops))
+		netif_wake_subqueue(txq->priv->netdev, txq->q_idx);
+}
+
+static uint hfi1_ipoib_ring_hwat(struct hfi1_ipoib_txq *txq)
+{
+	return min_t(uint, txq->priv->netdev->tx_queue_len,
+		     txq->tx_ring.max_items - 1);
+}
+
+static uint hfi1_ipoib_ring_lwat(struct hfi1_ipoib_txq *txq)
+{
+	return min_t(uint, txq->priv->netdev->tx_queue_len,
+		     txq->tx_ring.max_items) >> 1;
+}
+
+static void hfi1_ipoib_check_queue_depth(struct hfi1_ipoib_txq *txq)
+{
+	++txq->sent_txreqs;
+	if (hfi1_ipoib_used(txq) >= hfi1_ipoib_ring_hwat(txq) &&
+	    !atomic_xchg(&txq->ring_full, 1))
+		hfi1_ipoib_stop_txq(txq);
+}
+
 static void hfi1_ipoib_check_queue_stopped(struct hfi1_ipoib_txq *txq)
 {
 	struct net_device *dev = txq->priv->netdev;
 
-	/* If the queue is already running just return */
-	if (likely(!__netif_subqueue_stopped(dev, txq->q_idx)))
-		return;
-
 	/* If shutting down just return as queue state is irrelevant */
 	if (unlikely(dev->reg_state != NETREG_REGISTERED))
 		return;
@@ -86,11 +111,9 @@ static void hfi1_ipoib_check_queue_stopped(struct hfi1_ipoib_txq *txq)
 	 * Use the minimum of the current tx_queue_len or the rings max txreqs
 	 * to protect against ring overflow.
 	 */
-	if (hfi1_ipoib_txreqs(txq->sent_txreqs,
-			      atomic64_read(&txq->complete_txreqs))
-	    < min_t(unsigned int, dev->tx_queue_len,
-		    txq->tx_ring.max_items) >> 1)
-		netif_wake_subqueue(dev, txq->q_idx);
+	if (hfi1_ipoib_used(txq) < hfi1_ipoib_ring_lwat(txq) &&
+	    atomic_xchg(&txq->ring_full, 0))
+		hfi1_ipoib_wake_txq(txq);
 }
 
 static void hfi1_ipoib_free_tx(struct ipoib_txreq *tx, int budget)
@@ -364,11 +387,12 @@ static struct ipoib_txreq *hfi1_ipoib_send_dma_common(struct net_device *dev,
 	if (unlikely(!tx))
 		return ERR_PTR(-ENOMEM);
 
-	/* so that we can test if the sdma decriptors are there */
+	/* so that we can test if the sdma descriptors are there */
 	tx->txreq.num_desc = 0;
 	tx->priv = priv;
 	tx->txq = txp->txq;
 	tx->skb = skb;
+	INIT_LIST_HEAD(&tx->txreq.list);
 
 	hfi1_ipoib_build_ib_tx_headers(tx, txp);
 
@@ -469,6 +493,7 @@ static int hfi1_ipoib_send_dma_single(struct net_device *dev,
 
 	ret = hfi1_ipoib_submit_tx(txq, tx);
 	if (likely(!ret)) {
+tx_ok:
 		trace_sdma_output_ibhdr(tx->priv->dd,
 					&tx->sdma_hdr.hdr,
 					ib_is_sc5(txp->flow.sc5));
@@ -478,20 +503,8 @@ static int hfi1_ipoib_send_dma_single(struct net_device *dev,
 
 	txq->pkts_sent = false;
 
-	if (ret == -EBUSY) {
-		list_add_tail(&tx->txreq.list, &txq->tx_list);
-
-		trace_sdma_output_ibhdr(tx->priv->dd,
-					&tx->sdma_hdr.hdr,
-					ib_is_sc5(txp->flow.sc5));
-		hfi1_ipoib_check_queue_depth(txq);
-		return NETDEV_TX_OK;
-	}
-
-	if (ret == -ECOMM) {
-		hfi1_ipoib_check_queue_depth(txq);
-		return NETDEV_TX_OK;
-	}
+	if (ret == -EBUSY || ret == -ECOMM)
+		goto tx_ok;
 
 	sdma_txclean(priv->dd, &tx->txreq);
 	dev_kfree_skb_any(skb);
@@ -509,9 +522,17 @@ static int hfi1_ipoib_send_dma_list(struct net_device *dev,
 	struct ipoib_txreq *tx;
 
 	/* Has the flow change ? */
-	if (txq->flow.as_int != txp->flow.as_int)
-		(void)hfi1_ipoib_flush_tx_list(dev, txq);
-
+	if (txq->flow.as_int != txp->flow.as_int) {
+		int ret;
+
+		ret = hfi1_ipoib_flush_tx_list(dev, txq);
+		if (unlikely(ret)) {
+			if (ret == -EBUSY)
+				++dev->stats.tx_dropped;
+			dev_kfree_skb_any(skb);
+			return NETDEV_TX_OK;
+		}
+	}
 	tx = hfi1_ipoib_send_dma_common(dev, skb, txp);
 	if (IS_ERR(tx)) {
 		int ret = PTR_ERR(tx);
@@ -610,10 +631,14 @@ static int hfi1_ipoib_sdma_sleep(struct sdma_engine *sde,
 			return -EAGAIN;
 		}
 
-		netif_stop_subqueue(txq->priv->netdev, txq->q_idx);
-
-		if (list_empty(&txq->wait.list))
+		if (list_empty(&txreq->list))
+			/* came from non-list submit */
+			list_add_tail(&txreq->list, &txq->tx_list);
+		if (list_empty(&txq->wait.list)) {
+			if (!atomic_xchg(&txq->no_desc, 1))
+				hfi1_ipoib_stop_txq(txq);
 			iowait_queue(pkts_sent, wait->iow, &sde->dmawait);
+		}
 
 		write_sequnlock(&sde->waitlock);
 		return -EBUSY;
@@ -648,9 +673,9 @@ static void hfi1_ipoib_flush_txq(struct work_struct *work)
 	struct net_device *dev = txq->priv->netdev;
 
 	if (likely(dev->reg_state == NETREG_REGISTERED) &&
-	    likely(__netif_subqueue_stopped(dev, txq->q_idx)) &&
 	    likely(!hfi1_ipoib_flush_tx_list(dev, txq)))
-		netif_wake_subqueue(dev, txq->q_idx);
+		if (atomic_xchg(&txq->no_desc, 0))
+			hfi1_ipoib_wake_txq(txq);
 }
 
 int hfi1_ipoib_txreq_init(struct hfi1_ipoib_dev_priv *priv)
@@ -704,6 +729,9 @@ int hfi1_ipoib_txreq_init(struct hfi1_ipoib_dev_priv *priv)
 		txq->sde = NULL;
 		INIT_LIST_HEAD(&txq->tx_list);
 		atomic64_set(&txq->complete_txreqs, 0);
+		atomic_set(&txq->stops, 0);
+		atomic_set(&txq->ring_full, 0);
+		atomic_set(&txq->no_desc, 0);
 		txq->q_idx = i;
 		txq->flow.tx_queue = 0xff;
 		txq->flow.sc5 = 0xff;
@@ -769,7 +797,7 @@ static void hfi1_ipoib_drain_tx_list(struct hfi1_ipoib_txq *txq)
 		atomic64_inc(complete_txreqs);
 	}
 
-	if (hfi1_ipoib_txreqs(txq->sent_txreqs, atomic64_read(complete_txreqs)))
+	if (hfi1_ipoib_used(txq))
 		dd_dev_warn(txq->priv->dd,
 			    "txq %d not empty found %llu requests\n",
 			    txq->q_idx,
diff --git a/drivers/infiniband/hw/hfi1/netdev_rx.c b/drivers/infiniband/hw/hfi1/netdev_rx.c
index 63688e85e8da..6d263c9749b3 100644
--- a/drivers/infiniband/hw/hfi1/netdev_rx.c
+++ b/drivers/infiniband/hw/hfi1/netdev_rx.c
@@ -373,7 +373,7 @@ void hfi1_netdev_free(struct hfi1_devdata *dd)
 {
 	if (dd->dummy_netdev) {
 		dd_dev_info(dd, "hfi1 netdev freed\n");
-		free_netdev(dd->dummy_netdev);
+		kfree(dd->dummy_netdev);
 		dd->dummy_netdev = NULL;
 	}
 }
diff --git a/drivers/infiniband/hw/hfi1/qp.c b/drivers/infiniband/hw/hfi1/qp.c
index 0c2ae9f7b3e8..be62284e42d9 100644
--- a/drivers/infiniband/hw/hfi1/qp.c
+++ b/drivers/infiniband/hw/hfi1/qp.c
@@ -195,7 +195,7 @@ static inline int verbs_mtu_enum_to_int(struct ib_device *dev, enum ib_mtu mtu)
 {
 	/* Constraining 10KB packets to 8KB packets */
 	if (mtu == (enum ib_mtu)OPA_MTU_10240)
-		mtu = OPA_MTU_8192;
+		mtu = (enum ib_mtu)OPA_MTU_8192;
 	return opa_mtu_enum_to_int((enum opa_mtu)mtu);
 }
 
@@ -367,7 +367,10 @@ bool _hfi1_schedule_send(struct rvt_qp *qp)
 	struct hfi1_ibport *ibp =
 		to_iport(qp->ibqp.device, qp->port_num);
 	struct hfi1_pportdata *ppd = ppd_from_ibp(ibp);
-	struct hfi1_devdata *dd = dd_from_ibdev(qp->ibqp.device);
+	struct hfi1_devdata *dd = ppd->dd;
+
+	if (dd->flags & HFI1_SHUTDOWN)
+		return true;
 
 	return iowait_schedule(&priv->s_iowait, ppd->hfi1_wq,
 			       priv->s_sde ?
diff --git a/drivers/infiniband/hw/hfi1/tid_rdma.c b/drivers/infiniband/hw/hfi1/tid_rdma.c
index 243b4ba0b6f6..facff133139a 100644
--- a/drivers/infiniband/hw/hfi1/tid_rdma.c
+++ b/drivers/infiniband/hw/hfi1/tid_rdma.c
@@ -5406,7 +5406,10 @@ static bool _hfi1_schedule_tid_send(struct rvt_qp *qp)
 	struct hfi1_ibport *ibp =
 		to_iport(qp->ibqp.device, qp->port_num);
 	struct hfi1_pportdata *ppd = ppd_from_ibp(ibp);
-	struct hfi1_devdata *dd = dd_from_ibdev(qp->ibqp.device);
+	struct hfi1_devdata *dd = ppd->dd;
+
+	if ((dd->flags & HFI1_SHUTDOWN))
+		return true;
 
 	return iowait_tid_schedule(&priv->s_iowait, ppd->hfi1_wq,
 				   priv->s_sde ?
diff --git a/drivers/infiniband/hw/hfi1/verbs_txreq.h b/drivers/infiniband/hw/hfi1/verbs_txreq.h
index bfa6e081cb56..d2d526c5a756 100644
--- a/drivers/infiniband/hw/hfi1/verbs_txreq.h
+++ b/drivers/infiniband/hw/hfi1/verbs_txreq.h
@@ -91,7 +91,7 @@ static inline struct verbs_txreq *get_txreq(struct hfi1_ibdev *dev,
 	tx->mr = NULL;
 	tx->sde = priv->s_sde;
 	tx->psc = priv->s_sendcontext;
-	/* so that we can test if the sdma decriptors are there */
+	/* so that we can test if the sdma descriptors are there */
 	tx->txreq.num_desc = 0;
 	/* Set the header type */
 	tx->phdr.hdr.hdr_type = priv->hdr_type;
diff --git a/drivers/infiniband/hw/hns/hns_roce_device.h b/drivers/infiniband/hw/hns/hns_roce_device.h
index a77fa6730b2d..479fa557993e 100644
--- a/drivers/infiniband/hw/hns/hns_roce_device.h
+++ b/drivers/infiniband/hw/hns/hns_roce_device.h
@@ -898,13 +898,14 @@ struct hns_roce_hw {
 	int (*set_mac)(struct hns_roce_dev *hr_dev, u8 phy_port, u8 *addr);
 	void (*set_mtu)(struct hns_roce_dev *hr_dev, u8 phy_port,
 			enum ib_mtu mtu);
-	int (*write_mtpt)(void *mb_buf, struct hns_roce_mr *mr,
-			  unsigned long mtpt_idx);
+	int (*write_mtpt)(struct hns_roce_dev *hr_dev, void *mb_buf,
+			  struct hns_roce_mr *mr, unsigned long mtpt_idx);
 	int (*rereg_write_mtpt)(struct hns_roce_dev *hr_dev,
 				struct hns_roce_mr *mr, int flags, u32 pdn,
 				int mr_access_flags, u64 iova, u64 size,
 				void *mb_buf);
-	int (*frmr_write_mtpt)(void *mb_buf, struct hns_roce_mr *mr);
+	int (*frmr_write_mtpt)(struct hns_roce_dev *hr_dev, void *mb_buf,
+			       struct hns_roce_mr *mr);
 	int (*mw_write_mtpt)(void *mb_buf, struct hns_roce_mw *mw);
 	void (*write_cqc)(struct hns_roce_dev *hr_dev,
 			  struct hns_roce_cq *hr_cq, void *mb_buf, u64 *mtts,
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
index d02207cd30df..cf39f560b800 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
@@ -1756,10 +1756,10 @@ static void hns_roce_v1_set_mtu(struct hns_roce_dev *hr_dev, u8 phy_port,
 		   val);
 }
 
-static int hns_roce_v1_write_mtpt(void *mb_buf, struct hns_roce_mr *mr,
+static int hns_roce_v1_write_mtpt(struct hns_roce_dev *hr_dev, void *mb_buf,
+				  struct hns_roce_mr *mr,
 				  unsigned long mtpt_idx)
 {
-	struct hns_roce_dev *hr_dev = to_hr_dev(mr->ibmr.device);
 	u64 pages[HNS_ROCE_MAX_INNER_MTPT_NUM] = { 0 };
 	struct ib_device *ibdev = &hr_dev->ib_dev;
 	struct hns_roce_v1_mpt_entry *mpt_entry;
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
index c597d7281629..0618ced45bf8 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
@@ -910,7 +910,7 @@ static int hns_roce_v2_rst_process_cmd(struct hns_roce_dev *hr_dev)
 	instance_stage = handle->rinfo.instance_state;
 	reset_stage = handle->rinfo.reset_state;
 	reset_cnt = ops->ae_dev_reset_cnt(handle);
-	hw_resetting = ops->get_hw_reset_stat(handle);
+	hw_resetting = ops->get_cmdq_stat(handle);
 	sw_resetting = ops->ae_dev_resetting(handle);
 
 	if (reset_cnt != hr_dev->reset_cnt)
@@ -2529,10 +2529,10 @@ static int hns_roce_v2_set_mac(struct hns_roce_dev *hr_dev, u8 phy_port,
 	return hns_roce_cmq_send(hr_dev, &desc, 1);
 }
 
-static int set_mtpt_pbl(struct hns_roce_v2_mpt_entry *mpt_entry,
+static int set_mtpt_pbl(struct hns_roce_dev *hr_dev,
+			struct hns_roce_v2_mpt_entry *mpt_entry,
 			struct hns_roce_mr *mr)
 {
-	struct hns_roce_dev *hr_dev = to_hr_dev(mr->ibmr.device);
 	u64 pages[HNS_ROCE_V2_MAX_INNER_MTPT_NUM] = { 0 };
 	struct ib_device *ibdev = &hr_dev->ib_dev;
 	dma_addr_t pbl_ba;
@@ -2571,7 +2571,8 @@ static int set_mtpt_pbl(struct hns_roce_v2_mpt_entry *mpt_entry,
 	return 0;
 }
 
-static int hns_roce_v2_write_mtpt(void *mb_buf, struct hns_roce_mr *mr,
+static int hns_roce_v2_write_mtpt(struct hns_roce_dev *hr_dev,
+				  void *mb_buf, struct hns_roce_mr *mr,
 				  unsigned long mtpt_idx)
 {
 	struct hns_roce_v2_mpt_entry *mpt_entry;
@@ -2620,7 +2621,7 @@ static int hns_roce_v2_write_mtpt(void *mb_buf, struct hns_roce_mr *mr,
 	if (mr->type == MR_TYPE_DMA)
 		return 0;
 
-	ret = set_mtpt_pbl(mpt_entry, mr);
+	ret = set_mtpt_pbl(hr_dev, mpt_entry, mr);
 
 	return ret;
 }
@@ -2666,15 +2667,15 @@ static int hns_roce_v2_rereg_write_mtpt(struct hns_roce_dev *hr_dev,
 		mr->iova = iova;
 		mr->size = size;
 
-		ret = set_mtpt_pbl(mpt_entry, mr);
+		ret = set_mtpt_pbl(hr_dev, mpt_entry, mr);
 	}
 
 	return ret;
 }
 
-static int hns_roce_v2_frmr_write_mtpt(void *mb_buf, struct hns_roce_mr *mr)
+static int hns_roce_v2_frmr_write_mtpt(struct hns_roce_dev *hr_dev,
+				       void *mb_buf, struct hns_roce_mr *mr)
 {
-	struct hns_roce_dev *hr_dev = to_hr_dev(mr->ibmr.device);
 	struct ib_device *ibdev = &hr_dev->ib_dev;
 	struct hns_roce_v2_mpt_entry *mpt_entry;
 	dma_addr_t pbl_ba = 0;
@@ -3953,6 +3954,15 @@ static int config_qp_sq_buf(struct hns_roce_dev *hr_dev,
 	return 0;
 }
 
+static inline enum ib_mtu get_mtu(struct ib_qp *ibqp,
+				  const struct ib_qp_attr *attr)
+{
+	if (ibqp->qp_type == IB_QPT_GSI || ibqp->qp_type == IB_QPT_UD)
+		return IB_MTU_4096;
+
+	return attr->path_mtu;
+}
+
 static int modify_qp_init_to_rtr(struct ib_qp *ibqp,
 				 const struct ib_qp_attr *attr, int attr_mask,
 				 struct hns_roce_v2_qp_context *context,
@@ -3964,6 +3974,7 @@ static int modify_qp_init_to_rtr(struct ib_qp *ibqp,
 	struct ib_device *ibdev = &hr_dev->ib_dev;
 	dma_addr_t trrl_ba;
 	dma_addr_t irrl_ba;
+	enum ib_mtu mtu;
 	u8 port_num;
 	u64 *mtts;
 	u8 *dmac;
@@ -4061,23 +4072,23 @@ static int modify_qp_init_to_rtr(struct ib_qp *ibqp,
 	roce_set_field(qpc_mask->byte_52_udpspn_dmac, V2_QPC_BYTE_52_DMAC_M,
 		       V2_QPC_BYTE_52_DMAC_S, 0);
 
-	/* mtu*(2^LP_PKTN_INI) should not bigger than 1 message length 64kb */
+	mtu = get_mtu(ibqp, attr);
+
+	if (attr_mask & IB_QP_PATH_MTU) {
+		roce_set_field(context->byte_24_mtu_tc, V2_QPC_BYTE_24_MTU_M,
+			       V2_QPC_BYTE_24_MTU_S, mtu);
+		roce_set_field(qpc_mask->byte_24_mtu_tc, V2_QPC_BYTE_24_MTU_M,
+			       V2_QPC_BYTE_24_MTU_S, 0);
+	}
+
+#define MAX_LP_MSG_LEN 65536
+	/* MTU*(2^LP_PKTN_INI) shouldn't be bigger than 64kb */
 	roce_set_field(context->byte_56_dqpn_err, V2_QPC_BYTE_56_LP_PKTN_INI_M,
 		       V2_QPC_BYTE_56_LP_PKTN_INI_S,
-		       ilog2(hr_dev->caps.max_sq_inline / IB_MTU_4096));
+		       ilog2(MAX_LP_MSG_LEN / ib_mtu_enum_to_int(mtu)));
 	roce_set_field(qpc_mask->byte_56_dqpn_err, V2_QPC_BYTE_56_LP_PKTN_INI_M,
 		       V2_QPC_BYTE_56_LP_PKTN_INI_S, 0);
 
-	if (ibqp->qp_type == IB_QPT_GSI || ibqp->qp_type == IB_QPT_UD)
-		roce_set_field(context->byte_24_mtu_tc, V2_QPC_BYTE_24_MTU_M,
-			       V2_QPC_BYTE_24_MTU_S, IB_MTU_4096);
-	else if (attr_mask & IB_QP_PATH_MTU)
-		roce_set_field(context->byte_24_mtu_tc, V2_QPC_BYTE_24_MTU_M,
-			       V2_QPC_BYTE_24_MTU_S, attr->path_mtu);
-
-	roce_set_field(qpc_mask->byte_24_mtu_tc, V2_QPC_BYTE_24_MTU_M,
-		       V2_QPC_BYTE_24_MTU_S, 0);
-
 	roce_set_bit(qpc_mask->byte_108_rx_reqepsn,
 		     V2_QPC_BYTE_108_RX_REQ_PSN_ERR_S, 0);
 	roce_set_field(qpc_mask->byte_96_rx_reqmsn, V2_QPC_BYTE_96_RX_REQ_MSN_M,
diff --git a/drivers/infiniband/hw/hns/hns_roce_mr.c b/drivers/infiniband/hw/hns/hns_roce_mr.c
index 4c0bbb12770d..6b226a5eb7db 100644
--- a/drivers/infiniband/hw/hns/hns_roce_mr.c
+++ b/drivers/infiniband/hw/hns/hns_roce_mr.c
@@ -120,7 +120,7 @@ static int alloc_mr_pbl(struct hns_roce_dev *hr_dev, struct hns_roce_mr *mr,
 
 	mr->pbl_hop_num = is_fast ? 1 : hr_dev->caps.pbl_hop_num;
 	buf_attr.page_shift = is_fast ? PAGE_SHIFT :
-			      hr_dev->caps.pbl_buf_pg_sz + HNS_HW_PAGE_SHIFT;
+			      hr_dev->caps.pbl_buf_pg_sz + PAGE_SHIFT;
 	buf_attr.region[0].size = length;
 	buf_attr.region[0].hopnum = mr->pbl_hop_num;
 	buf_attr.region_count = 1;
@@ -180,9 +180,10 @@ static int hns_roce_mr_enable(struct hns_roce_dev *hr_dev,
 	}
 
 	if (mr->type != MR_TYPE_FRMR)
-		ret = hr_dev->hw->write_mtpt(mailbox->buf, mr, mtpt_idx);
+		ret = hr_dev->hw->write_mtpt(hr_dev, mailbox->buf, mr,
+					     mtpt_idx);
 	else
-		ret = hr_dev->hw->frmr_write_mtpt(mailbox->buf, mr);
+		ret = hr_dev->hw->frmr_write_mtpt(hr_dev, mailbox->buf, mr);
 	if (ret) {
 		dev_err(dev, "Write mtpt fail!\n");
 		goto err_page;
diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index 343a8b8361e7..6f99ed03d88e 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -511,7 +511,7 @@ static int mlx5_query_port_roce(struct ib_device *device, u8 port_num,
 					   mdev_port_num);
 	if (err)
 		goto out;
-	ext = MLX5_CAP_PCAM_FEATURE(dev->mdev, ptys_extended_ethernet);
+	ext = !!MLX5_GET_ETH_PROTO(ptys_reg, out, true, eth_proto_capability);
 	eth_prot_oper = MLX5_GET_ETH_PROTO(ptys_reg, out, ext, eth_proto_oper);
 
 	props->active_width     = IB_WIDTH_4X;
diff --git a/drivers/infiniband/hw/mlx5/odp.c b/drivers/infiniband/hw/mlx5/odp.c
index 7d2ec9ee5097..1ab676b66894 100644
--- a/drivers/infiniband/hw/mlx5/odp.c
+++ b/drivers/infiniband/hw/mlx5/odp.c
@@ -601,6 +601,23 @@ void mlx5_ib_free_implicit_mr(struct mlx5_ib_mr *imr)
 	 */
 	synchronize_srcu(&dev->odp_srcu);
 
+	/*
+	 * All work on the prefetch list must be completed, xa_erase() prevented
+	 * new work from being created.
+	 */
+	wait_event(imr->q_deferred_work, !atomic_read(&imr->num_deferred_work));
+
+	/*
+	 * At this point it is forbidden for any other thread to enter
+	 * pagefault_mr() on this imr. It is already forbidden to call
+	 * pagefault_mr() on an implicit child. Due to this additions to
+	 * implicit_children are prevented.
+	 */
+
+	/*
+	 * Block destroy_unused_implicit_child_mr() from incrementing
+	 * num_deferred_work.
+	 */
 	xa_lock(&imr->implicit_children);
 	xa_for_each (&imr->implicit_children, idx, mtt) {
 		__xa_erase(&imr->implicit_children, idx);
@@ -609,9 +626,8 @@ void mlx5_ib_free_implicit_mr(struct mlx5_ib_mr *imr)
 	xa_unlock(&imr->implicit_children);
 
 	/*
-	 * num_deferred_work can only be incremented inside the odp_srcu, or
-	 * under xa_lock while the child is in the xarray. Thus at this point
-	 * it is only decreasing, and all work holding it is now on the wq.
+	 * Wait for any concurrent destroy_unused_implicit_child_mr() to
+	 * complete.
 	 */
 	wait_event(imr->q_deferred_work, !atomic_read(&imr->num_deferred_work));
 
diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c
index 81bf6b975e0e..e050eade97a1 100644
--- a/drivers/infiniband/hw/mlx5/qp.c
+++ b/drivers/infiniband/hw/mlx5/qp.c
@@ -1862,7 +1862,7 @@ static int create_xrc_tgt_qp(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp,
 	if (!in)
 		return -ENOMEM;
 
-	if (MLX5_CAP_GEN(mdev, ece_support))
+	if (MLX5_CAP_GEN(mdev, ece_support) && ucmd)
 		MLX5_SET(create_qp_in, in, ece, ucmd->ece_options);
 	qpc = MLX5_ADDR_OF(create_qp_in, in, qpc);
 
@@ -2341,18 +2341,18 @@ static void destroy_qp_common(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp,
 	unsigned long flags;
 	int err;
 
-	if (qp->ibqp.rwq_ind_tbl) {
+	if (qp->is_rss) {
 		destroy_rss_raw_qp_tir(dev, qp);
 		return;
 	}
 
-	base = (qp->ibqp.qp_type == IB_QPT_RAW_PACKET ||
+	base = (qp->type == IB_QPT_RAW_PACKET ||
 		qp->flags & IB_QP_CREATE_SOURCE_QPN) ?
-	       &qp->raw_packet_qp.rq.base :
-	       &qp->trans_qp.base;
+		       &qp->raw_packet_qp.rq.base :
+		       &qp->trans_qp.base;
 
 	if (qp->state != IB_QPS_RESET) {
-		if (qp->ibqp.qp_type != IB_QPT_RAW_PACKET &&
+		if (qp->type != IB_QPT_RAW_PACKET &&
 		    !(qp->flags & IB_QP_CREATE_SOURCE_QPN)) {
 			err = mlx5_core_qp_modify(dev, MLX5_CMD_OP_2RST_QP, 0,
 						  NULL, &base->mqp, NULL);
@@ -2368,8 +2368,8 @@ static void destroy_qp_common(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp,
 				     base->mqp.qpn);
 	}
 
-	get_cqs(qp->ibqp.qp_type, qp->ibqp.send_cq, qp->ibqp.recv_cq,
-		&send_cq, &recv_cq);
+	get_cqs(qp->type, qp->ibqp.send_cq, qp->ibqp.recv_cq, &send_cq,
+		&recv_cq);
 
 	spin_lock_irqsave(&dev->reset_flow_resource_lock, flags);
 	mlx5_ib_lock_cqs(send_cq, recv_cq);
@@ -2391,7 +2391,7 @@ static void destroy_qp_common(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp,
 	mlx5_ib_unlock_cqs(send_cq, recv_cq);
 	spin_unlock_irqrestore(&dev->reset_flow_resource_lock, flags);
 
-	if (qp->ibqp.qp_type == IB_QPT_RAW_PACKET ||
+	if (qp->type == IB_QPT_RAW_PACKET ||
 	    qp->flags & IB_QP_CREATE_SOURCE_QPN) {
 		destroy_raw_packet_qp(dev, qp);
 	} else {
@@ -2668,6 +2668,13 @@ static int process_create_flags(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp,
 	if (qp_type == IB_QPT_RAW_PACKET && attr->rwq_ind_tbl)
 		return (create_flags) ? -EINVAL : 0;
 
+	process_create_flag(dev, &create_flags, IB_QP_CREATE_NETIF_QP,
+			    mlx5_get_flow_namespace(dev->mdev,
+						    MLX5_FLOW_NAMESPACE_BYPASS),
+			    qp);
+	process_create_flag(dev, &create_flags,
+			    IB_QP_CREATE_INTEGRITY_EN,
+			    MLX5_CAP_GEN(mdev, sho), qp);
 	process_create_flag(dev, &create_flags,
 			    IB_QP_CREATE_BLOCK_MULTICAST_LOOPBACK,
 			    MLX5_CAP_GEN(mdev, block_lb_mc), qp);
@@ -2873,7 +2880,6 @@ static int mlx5_ib_destroy_dct(struct mlx5_ib_qp *mqp)
 static int check_ucmd_data(struct mlx5_ib_dev *dev,
 			   struct mlx5_create_qp_params *params)
 {
-	struct ib_qp_init_attr *attr = params->attr;
 	struct ib_udata *udata = params->udata;
 	size_t size, last;
 	int ret;
@@ -2885,14 +2891,7 @@ static int check_ucmd_data(struct mlx5_ib_dev *dev,
 		 */
 		last = sizeof(struct mlx5_ib_create_qp_rss);
 	else
-		/* IB_QPT_RAW_PACKET doesn't have ECE data */
-		switch (attr->qp_type) {
-		case IB_QPT_RAW_PACKET:
-			last = offsetof(struct mlx5_ib_create_qp, ece_options);
-			break;
-		default:
-			last = offsetof(struct mlx5_ib_create_qp, reserved);
-		}
+		last = offsetof(struct mlx5_ib_create_qp, reserved);
 
 	if (udata->inlen <= last)
 		return 0;
@@ -2907,7 +2906,7 @@ static int check_ucmd_data(struct mlx5_ib_dev *dev,
 	if (!ret)
 		mlx5_ib_dbg(
 			dev,
-			"udata is not cleared, inlen = %lu, ucmd = %lu, last = %lu, size = %lu\n",
+			"udata is not cleared, inlen = %zu, ucmd = %zu, last = %zu, size = %zu\n",
 			udata->inlen, params->ucmd_size, last, size);
 	return ret ? 0 : -EINVAL;
 }
@@ -3002,10 +3001,19 @@ struct ib_qp *mlx5_ib_create_qp(struct ib_pd *pd, struct ib_qp_init_attr *attr,
 	return &qp->ibqp;
 
 destroy_qp:
-	if (qp->type == MLX5_IB_QPT_DCT)
+	if (qp->type == MLX5_IB_QPT_DCT) {
 		mlx5_ib_destroy_dct(qp);
-	else
+	} else {
+		/*
+		 * These lines below are temp solution till QP allocation
+		 * will be moved to be under IB/core responsiblity.
+		 */
+		qp->ibqp.send_cq = attr->send_cq;
+		qp->ibqp.recv_cq = attr->recv_cq;
+		qp->ibqp.pd = pd;
 		destroy_qp_common(dev, qp, udata);
+	}
+
 	qp = NULL;
 free_qp:
 	kfree(qp);
@@ -4162,8 +4170,6 @@ static int mlx5_ib_modify_dct(struct ib_qp *ibqp, struct ib_qp_attr *attr,
 
 		if (udata->outlen < min_resp_len)
 			return -EINVAL;
-		resp.response_length = min_resp_len;
-
 		/*
 		 * If we don't have enough space for the ECE options,
 		 * simply indicate it with resp.response_length.
@@ -4384,8 +4390,7 @@ static void to_rdma_ah_attr(struct mlx5_ib_dev *ibdev,
 				MLX5_GET(ads, path, src_addr_index),
 				MLX5_GET(ads, path, hop_limit),
 				MLX5_GET(ads, path, tclass));
-		memcpy(ah_attr, MLX5_ADDR_OF(ads, path, rgid_rip),
-		       MLX5_FLD_SZ_BYTES(ads, rgid_rip));
+		rdma_ah_set_dgid_raw(ah_attr, MLX5_ADDR_OF(ads, path, rgid_rip));
 	}
 }
 
diff --git a/drivers/infiniband/hw/mlx5/qpc.c b/drivers/infiniband/hw/mlx5/qpc.c
index c19d91d6dce8..7c3968ef9cd1 100644
--- a/drivers/infiniband/hw/mlx5/qpc.c
+++ b/drivers/infiniband/hw/mlx5/qpc.c
@@ -346,6 +346,9 @@ static int get_ece_from_mbox(void *out, u16 opcode)
 	int ece = 0;
 
 	switch (opcode) {
+	case MLX5_CMD_OP_INIT2INIT_QP:
+		ece = MLX5_GET(init2init_qp_out, out, ece);
+		break;
 	case MLX5_CMD_OP_INIT2RTR_QP:
 		ece = MLX5_GET(init2rtr_qp_out, out, ece);
 		break;
@@ -355,6 +358,9 @@ static int get_ece_from_mbox(void *out, u16 opcode)
 	case MLX5_CMD_OP_RTS2RTS_QP:
 		ece = MLX5_GET(rts2rts_qp_out, out, ece);
 		break;
+	case MLX5_CMD_OP_RST2INIT_QP:
+		ece = MLX5_GET(rst2init_qp_out, out, ece);
+		break;
 	default:
 		break;
 	}
@@ -406,6 +412,7 @@ static int modify_qp_mbox_alloc(struct mlx5_core_dev *dev, u16 opcode, int qpn,
 			return -ENOMEM;
 		MOD_QP_IN_SET_QPC(rst2init_qp, mbox->in, opcode, qpn,
 				  opt_param_mask, qpc, uid);
+		MLX5_SET(rst2init_qp_in, mbox->in, ece, ece);
 		break;
 	case MLX5_CMD_OP_INIT2RTR_QP:
 		if (MBOX_ALLOC(mbox, init2rtr_qp))
@@ -439,6 +446,7 @@ static int modify_qp_mbox_alloc(struct mlx5_core_dev *dev, u16 opcode, int qpn,
 			return -ENOMEM;
 		MOD_QP_IN_SET_QPC(init2init_qp, mbox->in, opcode, qpn,
 				  opt_param_mask, qpc, uid);
+		MLX5_SET(init2init_qp_in, mbox->in, ece, ece);
 		break;
 	default:
 		return -EINVAL;
diff --git a/drivers/infiniband/hw/mlx5/srq_cmd.c b/drivers/infiniband/hw/mlx5/srq_cmd.c
index 6f5eadc4d183..37aaacebd3f2 100644
--- a/drivers/infiniband/hw/mlx5/srq_cmd.c
+++ b/drivers/infiniband/hw/mlx5/srq_cmd.c
@@ -83,11 +83,11 @@ struct mlx5_core_srq *mlx5_cmd_get_srq(struct mlx5_ib_dev *dev, u32 srqn)
 	struct mlx5_srq_table *table = &dev->srq_table;
 	struct mlx5_core_srq *srq;
 
-	xa_lock(&table->array);
+	xa_lock_irq(&table->array);
 	srq = xa_load(&table->array, srqn);
 	if (srq)
 		refcount_inc(&srq->common.refcount);
-	xa_unlock(&table->array);
+	xa_unlock_irq(&table->array);
 
 	return srq;
 }
diff --git a/drivers/infiniband/hw/qedr/qedr_iw_cm.c b/drivers/infiniband/hw/qedr/qedr_iw_cm.c
index 792eecd206b6..97fc7dd353b0 100644
--- a/drivers/infiniband/hw/qedr/qedr_iw_cm.c
+++ b/drivers/infiniband/hw/qedr/qedr_iw_cm.c
@@ -150,8 +150,17 @@ qedr_iw_issue_event(void *context,
 	if (params->cm_info) {
 		event.ird = params->cm_info->ird;
 		event.ord = params->cm_info->ord;
-		event.private_data_len = params->cm_info->private_data_len;
-		event.private_data = (void *)params->cm_info->private_data;
+		/* Only connect_request and reply have valid private data
+		 * the rest of the events this may be left overs from
+		 * connection establishment. CONNECT_REQUEST is issued via
+		 * qedr_iw_mpa_request
+		 */
+		if (event_type == IW_CM_EVENT_CONNECT_REPLY) {
+			event.private_data_len =
+				params->cm_info->private_data_len;
+			event.private_data =
+				(void *)params->cm_info->private_data;
+		}
 	}
 
 	if (ep->cm_id)