From 4e045572e2c2be674ed7e43cca7ca105e8a22f56 Mon Sep 17 00:00:00 2001
From: Mike Marciniszyn <mike.marciniszyn@intel.com>
Date: Mon, 10 Oct 2016 06:14:28 -0700
Subject: IB/hfi1: Add unique txwait_lock for txreq events

Profiling suggests that the read_seqbegin() in
the txreq put logic is colliding with other uses
of the iowait lock.

The packet at a time use of this lock dictates a unique
lock to avoid reader/writer collisions when the number
of vTxWait events is low.

In order to support a unique lock the iowait struct embedded
in the QP is extended to remember the lock that protects the queue
head.

The QP destroy removes that QP from any wait list.  It doesn't
need to know the head because of the linked list API, but it does
need to know the lock required to protect the head.

This also opens up the wait logic to have unique per resources locks
which needs to be in future refinement.

Reviewed-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/iowait.h      |  8 ++++++++
 drivers/infiniband/hw/hfi1/pio.c         |  1 +
 drivers/infiniband/hw/hfi1/qp.c          | 11 ++++++++---
 drivers/infiniband/hw/hfi1/verbs.c       |  4 ++++
 drivers/infiniband/hw/hfi1/verbs.h       | 13 +++++++------
 drivers/infiniband/hw/hfi1/verbs_txreq.c | 13 +++++++------
 6 files changed, 35 insertions(+), 15 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/iowait.h b/drivers/infiniband/hw/hfi1/iowait.h
index 2ec6ef38d389..d9740ddea6f1 100644
--- a/drivers/infiniband/hw/hfi1/iowait.h
+++ b/drivers/infiniband/hw/hfi1/iowait.h
@@ -64,6 +64,7 @@ struct sdma_engine;
 /**
  * struct iowait - linkage for delayed progress/waiting
  * @list: used to add/insert into QP/PQ wait lists
+ * @lock: uses to record the list head lock
  * @tx_head: overflow list of sdma_txreq's
  * @sleep: no space callback
  * @wakeup: space callback wakeup
@@ -91,6 +92,11 @@ struct sdma_engine;
  * so sleeping is not allowed.
  *
  * The wait_dma member along with the iow
+ *
+ * The lock field is used by waiters to record
+ * the seqlock_t that guards the list head.
+ * Waiters explicity know that, but the destroy
+ * code that unwaits QPs does not.
  */
 
 struct iowait {
@@ -103,6 +109,7 @@ struct iowait {
 		unsigned seq);
 	void (*wakeup)(struct iowait *wait, int reason);
 	void (*sdma_drained)(struct iowait *wait);
+	seqlock_t *lock;
 	struct work_struct iowork;
 	wait_queue_head_t wait_dma;
 	wait_queue_head_t wait_pio;
@@ -141,6 +148,7 @@ static inline void iowait_init(
 	void (*sdma_drained)(struct iowait *wait))
 {
 	wait->count = 0;
+	wait->lock = NULL;
 	INIT_LIST_HEAD(&wait->list);
 	INIT_LIST_HEAD(&wait->tx_head);
 	INIT_WORK(&wait->iowork, func);
diff --git a/drivers/infiniband/hw/hfi1/pio.c b/drivers/infiniband/hw/hfi1/pio.c
index 50a3a36d9363..385e4dcf2cd3 100644
--- a/drivers/infiniband/hw/hfi1/pio.c
+++ b/drivers/infiniband/hw/hfi1/pio.c
@@ -1580,6 +1580,7 @@ static void sc_piobufavail(struct send_context *sc)
 		qp = iowait_to_qp(wait);
 		priv = qp->priv;
 		list_del_init(&priv->s_iowait.list);
+		priv->s_iowait.lock = NULL;
 		/* refcount held until actual wake up */
 		qps[n++] = qp;
 	}
diff --git a/drivers/infiniband/hw/hfi1/qp.c b/drivers/infiniband/hw/hfi1/qp.c
index 9fc75e7e8781..d752d6768a49 100644
--- a/drivers/infiniband/hw/hfi1/qp.c
+++ b/drivers/infiniband/hw/hfi1/qp.c
@@ -196,15 +196,18 @@ static void flush_tx_list(struct rvt_qp *qp)
 static void flush_iowait(struct rvt_qp *qp)
 {
 	struct hfi1_qp_priv *priv = qp->priv;
-	struct hfi1_ibdev *dev = to_idev(qp->ibqp.device);
 	unsigned long flags;
+	seqlock_t *lock = priv->s_iowait.lock;
 
-	write_seqlock_irqsave(&dev->iowait_lock, flags);
+	if (!lock)
+		return;
+	write_seqlock_irqsave(lock, flags);
 	if (!list_empty(&priv->s_iowait.list)) {
 		list_del_init(&priv->s_iowait.list);
+		priv->s_iowait.lock = NULL;
 		rvt_put_qp(qp);
 	}
-	write_sequnlock_irqrestore(&dev->iowait_lock, flags);
+	write_sequnlock_irqrestore(lock, flags);
 }
 
 static inline int opa_mtu_enum_to_int(int mtu)
@@ -543,6 +546,7 @@ static int iowait_sleep(
 			ibp->rvp.n_dmawait++;
 			qp->s_flags |= RVT_S_WAIT_DMA_DESC;
 			list_add_tail(&priv->s_iowait.list, &sde->dmawait);
+			priv->s_iowait.lock = &dev->iowait_lock;
 			trace_hfi1_qpsleep(qp, RVT_S_WAIT_DMA_DESC);
 			rvt_get_qp(qp);
 		}
@@ -964,6 +968,7 @@ void notify_error_qp(struct rvt_qp *qp)
 	if (!list_empty(&priv->s_iowait.list) && !(qp->s_flags & RVT_S_BUSY)) {
 		qp->s_flags &= ~RVT_S_ANY_WAIT_IO;
 		list_del_init(&priv->s_iowait.list);
+		priv->s_iowait.lock = NULL;
 		rvt_put_qp(qp);
 	}
 	write_sequnlock(&dev->iowait_lock);
diff --git a/drivers/infiniband/hw/hfi1/verbs.c b/drivers/infiniband/hw/hfi1/verbs.c
index 4b7a16ceb362..feecacb62162 100644
--- a/drivers/infiniband/hw/hfi1/verbs.c
+++ b/drivers/infiniband/hw/hfi1/verbs.c
@@ -694,6 +694,7 @@ static void mem_timer(unsigned long data)
 		qp = iowait_to_qp(wait);
 		priv = qp->priv;
 		list_del_init(&priv->s_iowait.list);
+		priv->s_iowait.lock = NULL;
 		/* refcount held until actual wake up */
 		if (!list_empty(list))
 			mod_timer(&dev->mem_timer, jiffies + 1);
@@ -769,6 +770,7 @@ static int wait_kmem(struct hfi1_ibdev *dev,
 				mod_timer(&dev->mem_timer, jiffies + 1);
 			qp->s_flags |= RVT_S_WAIT_KMEM;
 			list_add_tail(&priv->s_iowait.list, &dev->memwait);
+			priv->s_iowait.lock = &dev->iowait_lock;
 			trace_hfi1_qpsleep(qp, RVT_S_WAIT_KMEM);
 			rvt_get_qp(qp);
 		}
@@ -980,6 +982,7 @@ static int pio_wait(struct rvt_qp *qp,
 			qp->s_flags |= flag;
 			was_empty = list_empty(&sc->piowait);
 			list_add_tail(&priv->s_iowait.list, &sc->piowait);
+			priv->s_iowait.lock = &dev->iowait_lock;
 			trace_hfi1_qpsleep(qp, RVT_S_WAIT_PIO);
 			rvt_get_qp(qp);
 			/* counting: only call wantpiobuf_intr if first user */
@@ -1632,6 +1635,7 @@ int hfi1_register_ib_device(struct hfi1_devdata *dd)
 	setup_timer(&dev->mem_timer, mem_timer, (unsigned long)dev);
 
 	seqlock_init(&dev->iowait_lock);
+	seqlock_init(&dev->txwait_lock);
 	INIT_LIST_HEAD(&dev->txwait);
 	INIT_LIST_HEAD(&dev->memwait);
 
diff --git a/drivers/infiniband/hw/hfi1/verbs.h b/drivers/infiniband/hw/hfi1/verbs.h
index 1c3815d89eb7..7a8af39e99d4 100644
--- a/drivers/infiniband/hw/hfi1/verbs.h
+++ b/drivers/infiniband/hw/hfi1/verbs.h
@@ -180,18 +180,19 @@ struct hfi1_ibdev {
 	struct rvt_dev_info rdi; /* Must be first */
 
 	/* QP numbers are shared by all IB ports */
-	/* protect wait lists */
-	seqlock_t iowait_lock;
+	/* protect txwait list */
+	seqlock_t txwait_lock ____cacheline_aligned_in_smp;
 	struct list_head txwait;        /* list for wait verbs_txreq */
 	struct list_head memwait;       /* list for wait kernel memory */
-	struct list_head txreq_free;
 	struct kmem_cache *verbs_txreq_cache;
-	struct timer_list mem_timer;
+	u64 n_txwait;
+	u64 n_kmem_wait;
 
+	/* protect iowait lists */
+	seqlock_t iowait_lock ____cacheline_aligned_in_smp;
 	u64 n_piowait;
 	u64 n_piodrain;
-	u64 n_txwait;
-	u64 n_kmem_wait;
+	struct timer_list mem_timer;
 
 #ifdef CONFIG_DEBUG_FS
 	/* per HFI debugfs */
diff --git a/drivers/infiniband/hw/hfi1/verbs_txreq.c b/drivers/infiniband/hw/hfi1/verbs_txreq.c
index 094ab829ec42..5d23172c470f 100644
--- a/drivers/infiniband/hw/hfi1/verbs_txreq.c
+++ b/drivers/infiniband/hw/hfi1/verbs_txreq.c
@@ -72,22 +72,22 @@ void hfi1_put_txreq(struct verbs_txreq *tx)
 	kmem_cache_free(dev->verbs_txreq_cache, tx);
 
 	do {
-		seq = read_seqbegin(&dev->iowait_lock);
+		seq = read_seqbegin(&dev->txwait_lock);
 		if (!list_empty(&dev->txwait)) {
 			struct iowait *wait;
 
-			write_seqlock_irqsave(&dev->iowait_lock, flags);
+			write_seqlock_irqsave(&dev->txwait_lock, flags);
 			wait = list_first_entry(&dev->txwait, struct iowait,
 						list);
 			qp = iowait_to_qp(wait);
 			priv = qp->priv;
 			list_del_init(&priv->s_iowait.list);
 			/* refcount held until actual wake up */
-			write_sequnlock_irqrestore(&dev->iowait_lock, flags);
+			write_sequnlock_irqrestore(&dev->txwait_lock, flags);
 			hfi1_qp_wakeup(qp, RVT_S_WAIT_TX);
 			break;
 		}
-	} while (read_seqretry(&dev->iowait_lock, seq));
+	} while (read_seqretry(&dev->txwait_lock, seq));
 }
 
 struct verbs_txreq *__get_txreq(struct hfi1_ibdev *dev,
@@ -96,7 +96,7 @@ struct verbs_txreq *__get_txreq(struct hfi1_ibdev *dev,
 {
 	struct verbs_txreq *tx = ERR_PTR(-EBUSY);
 
-	write_seqlock(&dev->iowait_lock);
+	write_seqlock(&dev->txwait_lock);
 	if (ib_rvt_state_ops[qp->state] & RVT_PROCESS_RECV_OK) {
 		struct hfi1_qp_priv *priv;
 
@@ -108,13 +108,14 @@ struct verbs_txreq *__get_txreq(struct hfi1_ibdev *dev,
 			dev->n_txwait++;
 			qp->s_flags |= RVT_S_WAIT_TX;
 			list_add_tail(&priv->s_iowait.list, &dev->txwait);
+			priv->s_iowait.lock = &dev->txwait_lock;
 			trace_hfi1_qpsleep(qp, RVT_S_WAIT_TX);
 			rvt_get_qp(qp);
 		}
 		qp->s_flags &= ~RVT_S_BUSY;
 	}
 out:
-	write_sequnlock(&dev->iowait_lock);
+	write_sequnlock(&dev->txwait_lock);
 	return tx;
 }
 
-- 
cgit v1.2.1


From 63df8e09e185796ba532cff7c31f88811dd6cae6 Mon Sep 17 00:00:00 2001
From: Mike Marciniszyn <mike.marciniszyn@intel.com>
Date: Mon, 10 Oct 2016 06:14:34 -0700
Subject: IB/hfi1: Inline sdma_txclean() for verbs pio

Short circuit sdma_txclean() by adding an __sdma_txclean()
that is only called when the tx has sdma mappings.

Convert internal calls to __sdma_txclean().

This removes a call from the critical path.

Reviewed-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/sdma.c | 18 +++++++++---------
 drivers/infiniband/hw/hfi1/sdma.h | 12 +++++++++---
 2 files changed, 18 insertions(+), 12 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/sdma.c b/drivers/infiniband/hw/hfi1/sdma.c
index fd39bcaa062d..7102a076146d 100644
--- a/drivers/infiniband/hw/hfi1/sdma.c
+++ b/drivers/infiniband/hw/hfi1/sdma.c
@@ -375,7 +375,7 @@ static inline void complete_tx(struct sdma_engine *sde,
 			   sde->head_sn, tx->sn);
 	sde->head_sn++;
 #endif
-	sdma_txclean(sde->dd, tx);
+	__sdma_txclean(sde->dd, tx);
 	if (complete)
 		(*complete)(tx, res);
 	if (wait && iowait_sdma_dec(wait))
@@ -1643,7 +1643,7 @@ static inline u8 ahg_mode(struct sdma_txreq *tx)
 }
 
 /**
- * sdma_txclean() - clean tx of mappings, descp *kmalloc's
+ * __sdma_txclean() - clean tx of mappings, descp *kmalloc's
  * @dd: hfi1_devdata for unmapping
  * @tx: tx request to clean
  *
@@ -1653,7 +1653,7 @@ static inline u8 ahg_mode(struct sdma_txreq *tx)
  * The code can be called multiple times without issue.
  *
  */
-void sdma_txclean(
+void __sdma_txclean(
 	struct hfi1_devdata *dd,
 	struct sdma_txreq *tx)
 {
@@ -3080,7 +3080,7 @@ static int _extend_sdma_tx_descs(struct hfi1_devdata *dd, struct sdma_txreq *tx)
 		tx->descp[i] = tx->descs[i];
 	return 0;
 enomem:
-	sdma_txclean(dd, tx);
+	__sdma_txclean(dd, tx);
 	return -ENOMEM;
 }
 
@@ -3109,14 +3109,14 @@ int ext_coal_sdma_tx_descs(struct hfi1_devdata *dd, struct sdma_txreq *tx,
 
 	rval = _extend_sdma_tx_descs(dd, tx);
 	if (rval) {
-		sdma_txclean(dd, tx);
+		__sdma_txclean(dd, tx);
 		return rval;
 	}
 
 	/* If coalesce buffer is allocated, copy data into it */
 	if (tx->coalesce_buf) {
 		if (type == SDMA_MAP_NONE) {
-			sdma_txclean(dd, tx);
+			__sdma_txclean(dd, tx);
 			return -EINVAL;
 		}
 
@@ -3124,7 +3124,7 @@ int ext_coal_sdma_tx_descs(struct hfi1_devdata *dd, struct sdma_txreq *tx,
 			kvaddr = kmap(page);
 			kvaddr += offset;
 		} else if (WARN_ON(!kvaddr)) {
-			sdma_txclean(dd, tx);
+			__sdma_txclean(dd, tx);
 			return -EINVAL;
 		}
 
@@ -3154,7 +3154,7 @@ int ext_coal_sdma_tx_descs(struct hfi1_devdata *dd, struct sdma_txreq *tx,
 				      DMA_TO_DEVICE);
 
 		if (unlikely(dma_mapping_error(&dd->pcidev->dev, addr))) {
-			sdma_txclean(dd, tx);
+			__sdma_txclean(dd, tx);
 			return -ENOSPC;
 		}
 
@@ -3196,7 +3196,7 @@ int _pad_sdma_tx_descs(struct hfi1_devdata *dd, struct sdma_txreq *tx)
 	if ((unlikely(tx->num_desc == tx->desc_limit))) {
 		rval = _extend_sdma_tx_descs(dd, tx);
 		if (rval) {
-			sdma_txclean(dd, tx);
+			__sdma_txclean(dd, tx);
 			return rval;
 		}
 	}
diff --git a/drivers/infiniband/hw/hfi1/sdma.h b/drivers/infiniband/hw/hfi1/sdma.h
index 56257ea3598f..21f1e2834f37 100644
--- a/drivers/infiniband/hw/hfi1/sdma.h
+++ b/drivers/infiniband/hw/hfi1/sdma.h
@@ -667,7 +667,13 @@ int ext_coal_sdma_tx_descs(struct hfi1_devdata *dd, struct sdma_txreq *tx,
 			   int type, void *kvaddr, struct page *page,
 			   unsigned long offset, u16 len);
 int _pad_sdma_tx_descs(struct hfi1_devdata *, struct sdma_txreq *);
-void sdma_txclean(struct hfi1_devdata *, struct sdma_txreq *);
+void __sdma_txclean(struct hfi1_devdata *, struct sdma_txreq *);
+
+static inline void sdma_txclean(struct hfi1_devdata *dd, struct sdma_txreq *tx)
+{
+	if (tx->num_desc)
+		__sdma_txclean(dd, tx);
+}
 
 /* helpers used by public routines */
 static inline void _sdma_close_tx(struct hfi1_devdata *dd,
@@ -753,7 +759,7 @@ static inline int sdma_txadd_page(
 		       DMA_TO_DEVICE);
 
 	if (unlikely(dma_mapping_error(&dd->pcidev->dev, addr))) {
-		sdma_txclean(dd, tx);
+		__sdma_txclean(dd, tx);
 		return -ENOSPC;
 	}
 
@@ -834,7 +840,7 @@ static inline int sdma_txadd_kvaddr(
 		       DMA_TO_DEVICE);
 
 	if (unlikely(dma_mapping_error(&dd->pcidev->dev, addr))) {
-		sdma_txclean(dd, tx);
+		__sdma_txclean(dd, tx);
 		return -ENOSPC;
 	}
 
-- 
cgit v1.2.1


From 99f80d2f5fb6d4165186390ecba83952803b667b Mon Sep 17 00:00:00 2001
From: Mike Marciniszyn <mike.marciniszyn@intel.com>
Date: Mon, 10 Oct 2016 06:14:39 -0700
Subject: IB/hfi1: Optimize lkey validation structures

Profiling shows that the key validation is susceptible
to cache line trading when accessing the lkey table.

Fix by separating out the read mostly fields from the write
fields.   In addition the shift amount, which is function
of the lkey table size, is precomputed and stored with the
table pointer.   Since both the shift and table pointer
are in the same read mostly cacheline, this saves a cache
line in this hot path.

Reviewed-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rdmavt/mr.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rdmavt/mr.c b/drivers/infiniband/sw/rdmavt/mr.c
index 46b64970058e..4acf179adf3b 100644
--- a/drivers/infiniband/sw/rdmavt/mr.c
+++ b/drivers/infiniband/sw/rdmavt/mr.c
@@ -84,6 +84,7 @@ int rvt_driver_mr_init(struct rvt_dev_info *rdi)
 		lkey_table_size = rdi->dparms.lkey_table_size;
 	}
 	rdi->lkey_table.max = 1 << lkey_table_size;
+	rdi->lkey_table.shift = 32 - lkey_table_size;
 	lk_tab_size = rdi->lkey_table.max * sizeof(*rdi->lkey_table.table);
 	rdi->lkey_table.table = (struct rvt_mregion __rcu **)
 			       vmalloc_node(lk_tab_size, rdi->dparms.node);
@@ -774,7 +775,6 @@ int rvt_lkey_ok(struct rvt_lkey_table *rkt, struct rvt_pd *pd,
 	struct rvt_mregion *mr;
 	unsigned n, m;
 	size_t off;
-	struct rvt_dev_info *dev = ib_to_rvt(pd->ibpd.device);
 
 	/*
 	 * We use LKEY == zero for kernel virtual addresses
@@ -782,6 +782,8 @@ int rvt_lkey_ok(struct rvt_lkey_table *rkt, struct rvt_pd *pd,
 	 */
 	rcu_read_lock();
 	if (sge->lkey == 0) {
+		struct rvt_dev_info *dev = ib_to_rvt(pd->ibpd.device);
+
 		if (pd->user)
 			goto bail;
 		mr = rcu_dereference(dev->dma_mr);
@@ -798,8 +800,7 @@ int rvt_lkey_ok(struct rvt_lkey_table *rkt, struct rvt_pd *pd,
 		isge->n = 0;
 		goto ok;
 	}
-	mr = rcu_dereference(
-		rkt->table[(sge->lkey >> (32 - dev->dparms.lkey_table_size))]);
+	mr = rcu_dereference(rkt->table[sge->lkey >> rkt->shift]);
 	if (unlikely(!mr || atomic_read(&mr->lkey_invalid) ||
 		     mr->lkey != sge->lkey || mr->pd != &pd->ibpd))
 		goto bail;
@@ -899,8 +900,7 @@ int rvt_rkey_ok(struct rvt_qp *qp, struct rvt_sge *sge,
 		goto ok;
 	}
 
-	mr = rcu_dereference(
-		rkt->table[(rkey >> (32 - dev->dparms.lkey_table_size))]);
+	mr = rcu_dereference(rkt->table[rkey >> rkt->shift]);
 	if (unlikely(!mr || atomic_read(&mr->lkey_invalid) ||
 		     mr->lkey != rkey || qp->ibqp.pd != mr->pd))
 		goto bail;
-- 
cgit v1.2.1


From 99c7abfb626d0ce4e6e190dc7aeeb55eb21e335f Mon Sep 17 00:00:00 2001
From: Mike Marciniszyn <mike.marciniszyn@intel.com>
Date: Mon, 17 Oct 2016 04:19:13 -0700
Subject: IB/hfi1: Optimize pio cachelines

Move buffers_allocated pcpu pointer to allocator line.

Move hw_free pointer to releaser line.

Fill other holes revealed by pahole.

Reviewed-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/pio.h | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/pio.h b/drivers/infiniband/hw/hfi1/pio.h
index e709eaf743b5..bd19507b6bb0 100644
--- a/drivers/infiniband/hw/hfi1/pio.h
+++ b/drivers/infiniband/hw/hfi1/pio.h
@@ -104,7 +104,6 @@ struct send_context {
 	void __iomem *base_addr;	/* start of PIO memory */
 	union pio_shadow_ring *sr;	/* shadow ring */
 
-	volatile __le64 *hw_free;	/* HW free counter */
 	struct work_struct halt_work;	/* halted context work queue entry */
 	unsigned long flags;		/* flags */
 	int node;			/* context home node */
@@ -116,19 +115,20 @@ struct send_context {
 	u32 group;			/* credit return group */
 	/* allocator fields */
 	spinlock_t alloc_lock ____cacheline_aligned_in_smp;
+	u32 sr_head;			/* shadow ring head */
 	unsigned long fill;		/* official alloc count */
 	unsigned long alloc_free;	/* copy of free (less cache thrash) */
-	u32 sr_head;			/* shadow ring head */
+	u32 __percpu *buffers_allocated;/* count of buffers allocated */
 	/* releaser fields */
 	spinlock_t release_lock ____cacheline_aligned_in_smp;
-	unsigned long free;		/* official free count */
 	u32 sr_tail;			/* shadow ring tail */
+	unsigned long free;		/* official free count */
+	volatile __le64 *hw_free;	/* HW free counter */
 	/* list for PIO waiters */
 	struct list_head piowait  ____cacheline_aligned_in_smp;
 	spinlock_t credit_ctrl_lock ____cacheline_aligned_in_smp;
-	u64 credit_ctrl;		/* cache for credit control */
 	u32 credit_intr_count;		/* count of credit intr users */
-	u32 __percpu *buffers_allocated;/* count of buffers allocated */
+	u64 credit_ctrl;		/* cache for credit control */
 	wait_queue_head_t halt_wait;    /* wait until kernel sees interrupt */
 };
 
-- 
cgit v1.2.1


From a6cd5f08e085176fe8dbd4c57217058e53dfc4d9 Mon Sep 17 00:00:00 2001
From: Jakub Pawlak <jakub.pawlak@intel.com>
Date: Mon, 17 Oct 2016 04:19:30 -0700
Subject: IB/hfi1: Unify access to GUID entries

This patch consolidates the node GUIDs and the port GUID handling
and unifies access to these items. The knowledge of hfi1 GUIDs'
design and their location are kept in accessors to centralize access.

Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Reviewed-by: Brian Welty <brian.welty@intel.com>
Signed-off-by: Jakub Pawlak <jakub.pawlak@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/chip.c  |  4 ++--
 drivers/infiniband/hw/hfi1/hfi.h   | 19 +++++++++++++++++--
 drivers/infiniband/hw/hfi1/mad.c   | 19 ++++++++++---------
 drivers/infiniband/hw/hfi1/ruc.c   | 16 +++-------------
 drivers/infiniband/hw/hfi1/verbs.c | 15 +++++++--------
 drivers/infiniband/hw/hfi1/verbs.h |  3 ---
 6 files changed, 39 insertions(+), 37 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/chip.c b/drivers/infiniband/hw/hfi1/chip.c
index 9bf5f23544d4..f87d80567572 100644
--- a/drivers/infiniband/hw/hfi1/chip.c
+++ b/drivers/infiniband/hw/hfi1/chip.c
@@ -9567,11 +9567,11 @@ int bringup_serdes(struct hfi1_pportdata *ppd)
 	if (HFI1_CAP_IS_KSET(EXTENDED_PSN))
 		add_rcvctrl(dd, RCV_CTRL_RCV_EXTENDED_PSN_ENABLE_SMASK);
 
-	guid = ppd->guid;
+	guid = ppd->guids[HFI1_PORT_GUID_INDEX];
 	if (!guid) {
 		if (dd->base_guid)
 			guid = dd->base_guid + ppd->port - 1;
-		ppd->guid = guid;
+		ppd->guids[HFI1_PORT_GUID_INDEX] = guid;
 	}
 
 	/* Set linkinit_reason on power up per OPA spec */
diff --git a/drivers/infiniband/hw/hfi1/hfi.h b/drivers/infiniband/hw/hfi1/hfi.h
index 7eef11b316ff..a2ea643a79fc 100644
--- a/drivers/infiniband/hw/hfi1/hfi.h
+++ b/drivers/infiniband/hw/hfi1/hfi.h
@@ -512,6 +512,9 @@ struct rvt_sge_state;
 #define HFI1_MIN_VLS_SUPPORTED 1
 #define HFI1_MAX_VLS_SUPPORTED 8
 
+#define HFI1_GUIDS_PER_PORT  5
+#define HFI1_PORT_GUID_INDEX 0
+
 static inline void incr_cntr64(u64 *cntr)
 {
 	if (*cntr < (u64)-1LL)
@@ -582,8 +585,9 @@ struct hfi1_pportdata {
 	u32 port_type;
 	struct qsfp_data qsfp_info;
 
-	/* GUID for this interface, in host order */
-	u64 guid;
+	/* GUIDs for this interface, in host order, guids[0] is a port guid */
+	u64 guids[HFI1_GUIDS_PER_PORT];
+
 	/* GUID for peer interface, in host order */
 	u64 neighbor_guid;
 
@@ -1632,6 +1636,17 @@ static inline u16 hfi1_get_pkey(struct hfi1_ibport *ibp, unsigned index)
 	return ret;
 }
 
+/*
+ * Return the indexed GUID from the port GUIDs table.
+ */
+static inline __be64 get_sguid(struct hfi1_ibport *ibp, unsigned int index)
+{
+	struct hfi1_pportdata *ppd = ppd_from_ibp(ibp);
+
+	WARN_ON(index >= HFI1_GUIDS_PER_PORT);
+	return cpu_to_be64(ppd->guids[index]);
+}
+
 /*
  * Called by readers of cc_state only, must call under rcu_read_lock().
  */
diff --git a/drivers/infiniband/hw/hfi1/mad.c b/drivers/infiniband/hw/hfi1/mad.c
index 9487c9bb8920..0ef62e67f283 100644
--- a/drivers/infiniband/hw/hfi1/mad.c
+++ b/drivers/infiniband/hw/hfi1/mad.c
@@ -336,20 +336,20 @@ static int __subn_get_opa_nodeinfo(struct opa_smp *smp, u32 am, u8 *data,
 	ni = (struct opa_node_info *)data;
 
 	/* GUID 0 is illegal */
-	if (am || pidx >= dd->num_pports || dd->pport[pidx].guid == 0) {
+	if (am || pidx >= dd->num_pports || ibdev->node_guid == 0 ||
+	    get_sguid(to_iport(ibdev, port), HFI1_PORT_GUID_INDEX) == 0) {
 		smp->status |= IB_SMP_INVALID_FIELD;
 		return reply((struct ib_mad_hdr *)smp);
 	}
 
-	ni->port_guid = cpu_to_be64(dd->pport[pidx].guid);
+	ni->port_guid = get_sguid(to_iport(ibdev, port), HFI1_PORT_GUID_INDEX);
 	ni->base_version = OPA_MGMT_BASE_VERSION;
 	ni->class_version = OPA_SMI_CLASS_VERSION;
 	ni->node_type = 1;     /* channel adapter */
 	ni->num_ports = ibdev->phys_port_cnt;
 	/* This is already in network order */
 	ni->system_image_guid = ib_hfi1_sys_image_guid;
-	/* Use first-port GUID as node */
-	ni->node_guid = cpu_to_be64(dd->pport->guid);
+	ni->node_guid = ibdev->node_guid;
 	ni->partition_cap = cpu_to_be16(hfi1_get_npkeys(dd));
 	ni->device_id = cpu_to_be16(dd->pcidev->device);
 	ni->revision = cpu_to_be32(dd->minrev);
@@ -373,19 +373,20 @@ static int subn_get_nodeinfo(struct ib_smp *smp, struct ib_device *ibdev,
 
 	/* GUID 0 is illegal */
 	if (smp->attr_mod || pidx >= dd->num_pports ||
-	    dd->pport[pidx].guid == 0)
+	    ibdev->node_guid == 0 ||
+	    get_sguid(to_iport(ibdev, port), HFI1_PORT_GUID_INDEX) == 0) {
 		smp->status |= IB_SMP_INVALID_FIELD;
-	else
-		nip->port_guid = cpu_to_be64(dd->pport[pidx].guid);
+		return reply((struct ib_mad_hdr *)smp);
+	}
 
+	nip->port_guid = get_sguid(to_iport(ibdev, port), HFI1_PORT_GUID_INDEX);
 	nip->base_version = OPA_MGMT_BASE_VERSION;
 	nip->class_version = OPA_SMI_CLASS_VERSION;
 	nip->node_type = 1;     /* channel adapter */
 	nip->num_ports = ibdev->phys_port_cnt;
 	/* This is already in network order */
 	nip->sys_guid = ib_hfi1_sys_image_guid;
-	 /* Use first-port GUID as node */
-	nip->node_guid = cpu_to_be64(dd->pport->guid);
+	nip->node_guid = ibdev->node_guid;
 	nip->partition_cap = cpu_to_be16(hfi1_get_npkeys(dd));
 	nip->device_id = cpu_to_be16(dd->pcidev->device);
 	nip->revision = cpu_to_be32(dd->minrev);
diff --git a/drivers/infiniband/hw/hfi1/ruc.c b/drivers/infiniband/hw/hfi1/ruc.c
index a1576aea4756..5a70e91b5191 100644
--- a/drivers/infiniband/hw/hfi1/ruc.c
+++ b/drivers/infiniband/hw/hfi1/ruc.c
@@ -239,16 +239,6 @@ bail:
 	return ret;
 }
 
-static __be64 get_sguid(struct hfi1_ibport *ibp, unsigned index)
-{
-	if (!index) {
-		struct hfi1_pportdata *ppd = ppd_from_ibp(ibp);
-
-		return cpu_to_be64(ppd->guid);
-	}
-	return ibp->guids[index - 1];
-}
-
 static int gid_ok(union ib_gid *gid, __be64 gid_prefix, __be64 id)
 {
 	return (gid->global.interface_id == id &&
@@ -699,9 +689,9 @@ u32 hfi1_make_grh(struct hfi1_ibport *ibp, struct ib_grh *hdr,
 	/* The SGID is 32-bit aligned. */
 	hdr->sgid.global.subnet_prefix = ibp->rvp.gid_prefix;
 	hdr->sgid.global.interface_id =
-		grh->sgid_index && grh->sgid_index < ARRAY_SIZE(ibp->guids) ?
-		ibp->guids[grh->sgid_index - 1] :
-			cpu_to_be64(ppd_from_ibp(ibp)->guid);
+		grh->sgid_index < HFI1_GUIDS_PER_PORT ?
+		get_sguid(ibp, grh->sgid_index) :
+		get_sguid(ibp, HFI1_PORT_GUID_INDEX);
 	hdr->dgid = grh->dgid;
 
 	/* GRH header size in 32-bit words. */
diff --git a/drivers/infiniband/hw/hfi1/verbs.c b/drivers/infiniband/hw/hfi1/verbs.c
index feecacb62162..f1dead369258 100644
--- a/drivers/infiniband/hw/hfi1/verbs.c
+++ b/drivers/infiniband/hw/hfi1/verbs.c
@@ -1486,15 +1486,11 @@ static int hfi1_get_guid_be(struct rvt_dev_info *rdi, struct rvt_ibport *rvp,
 			    int guid_index, __be64 *guid)
 {
 	struct hfi1_ibport *ibp = container_of(rvp, struct hfi1_ibport, rvp);
-	struct hfi1_pportdata *ppd = ppd_from_ibp(ibp);
 
-	if (guid_index == 0)
-		*guid = cpu_to_be64(ppd->guid);
-	else if (guid_index < HFI1_GUIDS_PER_PORT)
-		*guid = ibp->guids[guid_index - 1];
-	else
+	if (guid_index >= HFI1_GUIDS_PER_PORT)
 		return -EINVAL;
 
+	*guid = get_sguid(ibp, guid_index);
 	return 0;
 }
 
@@ -1623,6 +1619,7 @@ int hfi1_register_ib_device(struct hfi1_devdata *dd)
 	struct hfi1_ibdev *dev = &dd->verbs_dev;
 	struct ib_device *ibdev = &dev->rdi.ibdev;
 	struct hfi1_pportdata *ppd = dd->pport;
+	struct hfi1_ibport *ibp = &ppd->ibport_data;
 	unsigned i;
 	int ret;
 	size_t lcpysz = IB_DEVICE_NAME_MAX;
@@ -1643,17 +1640,19 @@ int hfi1_register_ib_device(struct hfi1_devdata *dd)
 	if (ret)
 		goto err_verbs_txreq;
 
+	/* Use first-port GUID as node guid */
+	ibdev->node_guid = get_sguid(ibp, HFI1_PORT_GUID_INDEX);
+
 	/*
 	 * The system image GUID is supposed to be the same for all
 	 * HFIs in a single system but since there can be other
 	 * device types in the system, we can't be sure this is unique.
 	 */
 	if (!ib_hfi1_sys_image_guid)
-		ib_hfi1_sys_image_guid = cpu_to_be64(ppd->guid);
+		ib_hfi1_sys_image_guid = ibdev->node_guid;
 	lcpysz = strlcpy(ibdev->name, class_name(), lcpysz);
 	strlcpy(ibdev->name + lcpysz, "_%d", IB_DEVICE_NAME_MAX - lcpysz);
 	ibdev->owner = THIS_MODULE;
-	ibdev->node_guid = cpu_to_be64(ppd->guid);
 	ibdev->phys_port_cnt = dd->num_pports;
 	ibdev->dma_device = &dd->pcidev->dev;
 	ibdev->modify_device = modify_device;
diff --git a/drivers/infiniband/hw/hfi1/verbs.h b/drivers/infiniband/hw/hfi1/verbs.h
index 7a8af39e99d4..e6b893010e6d 100644
--- a/drivers/infiniband/hw/hfi1/verbs.h
+++ b/drivers/infiniband/hw/hfi1/verbs.h
@@ -73,7 +73,6 @@ struct hfi1_packet;
 #include "iowait.h"
 
 #define HFI1_MAX_RDMA_ATOMIC     16
-#define HFI1_GUIDS_PER_PORT	5
 
 /*
  * Increment this value if any changes that break userspace ABI
@@ -169,8 +168,6 @@ struct hfi1_ibport {
 	struct rvt_qp __rcu *qp[2];
 	struct rvt_ibport rvp;
 
-	__be64 guids[HFI1_GUIDS_PER_PORT	- 1];	/* writable GUIDs */
-
 	/* the first 16 entries are sl_to_vl for !OPA */
 	u8 sl_to_sc[32];
 	u8 sc_to_sl[32];
-- 
cgit v1.2.1


From 6e768f0682e26e7683c4af5b0de04a2e08bc67db Mon Sep 17 00:00:00 2001
From: Sebastian Sanchez <sebastian.sanchez@intel.com>
Date: Mon, 17 Oct 2016 04:19:35 -0700
Subject: IB/hfi1: Optimize devdata cachelines

Profiling shows hot path struct members that need
to be in a minimum set of cachelines.

Group these struct member in the same cacheline:
	sc2vl_lock
	sc2vl
	rhf_rcv_function_map
	rcv_limit
	rhf_offset

Group these struct member in the same cacheline:
	process_pio_send
	process_dma_send
	pport
	rcd
	int_counter
	flags
	num_pports
	first_user_ctxt

Fill holes in struct hfi1_devdata revealed by pahole.

Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/hfi.h | 108 ++++++++++++++++++++-------------------
 1 file changed, 55 insertions(+), 53 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/hfi.h b/drivers/infiniband/hw/hfi1/hfi.h
index a2ea643a79fc..25e44c4af995 100644
--- a/drivers/infiniband/hw/hfi1/hfi.h
+++ b/drivers/infiniband/hw/hfi1/hfi.h
@@ -852,32 +852,29 @@ struct hfi1_devdata {
 	u8 __iomem *kregend;
 	/* physical address of chip for io_remap, etc. */
 	resource_size_t physaddr;
-	/* receive context data */
-	struct hfi1_ctxtdata **rcd;
+	/* Per VL data. Enough for all VLs but not all elements are set/used. */
+	struct per_vl_data vld[PER_VL_SEND_CONTEXTS];
 	/* send context data */
 	struct send_context_info *send_contexts;
 	/* map hardware send contexts to software index */
 	u8 *hw_to_sw;
 	/* spinlock for allocating and releasing send context resources */
 	spinlock_t sc_lock;
-	/* Per VL data. Enough for all VLs but not all elements are set/used. */
-	struct per_vl_data vld[PER_VL_SEND_CONTEXTS];
 	/* lock for pio_map */
 	spinlock_t pio_map_lock;
+	/* Send Context initialization lock. */
+	spinlock_t sc_init_lock;
+	/* lock for sdma_map */
+	spinlock_t                          sde_map_lock;
 	/* array of kernel send contexts */
 	struct send_context **kernel_send_context;
 	/* array of vl maps */
 	struct pio_vl_map __rcu *pio_map;
-	/* seqlock for sc2vl */
-	seqlock_t sc2vl_lock;
-	u64 sc2vl[4];
-	/* Send Context initialization lock. */
-	spinlock_t sc_init_lock;
+	/* default flags to last descriptor */
+	u64 default_desc1;
 
 	/* fields common to all SDMA engines */
 
-	/* default flags to last descriptor */
-	u64 default_desc1;
 	volatile __le64                    *sdma_heads_dma; /* DMA'ed by chip */
 	dma_addr_t                          sdma_heads_phys;
 	void                               *sdma_pad_dma; /* DMA'ed by chip */
@@ -888,8 +885,6 @@ struct hfi1_devdata {
 	u32                                 chip_sdma_engines;
 	/* num used */
 	u32                                 num_sdma;
-	/* lock for sdma_map */
-	spinlock_t                          sde_map_lock;
 	/* array of engines sized by num_sdma */
 	struct sdma_engine                 *per_sdma;
 	/* array of vl maps */
@@ -898,14 +893,11 @@ struct hfi1_devdata {
 	wait_queue_head_t		  sdma_unfreeze_wq;
 	atomic_t			  sdma_unfreeze_count;
 
+	u32 lcb_access_count;		/* count of LCB users */
+
 	/* common data between shared ASIC HFIs in this OS */
 	struct hfi1_asic_data *asic_data;
 
-	/* hfi1_pportdata, points to array of (physical) port-specific
-	 * data structs, indexed by pidx (0..n-1)
-	 */
-	struct hfi1_pportdata *pport;
-
 	/* mem-mapped pointer to base of PIO buffers */
 	void __iomem *piobase;
 	/*
@@ -922,20 +914,13 @@ struct hfi1_devdata {
 	/* send context numbers and sizes for each type */
 	struct sc_config_sizes sc_sizes[SC_MAX];
 
-	u32 lcb_access_count;		/* count of LCB users */
-
 	char *boardname; /* human readable board info */
 
-	/* device (not port) flags, basically device capabilities */
-	u32 flags;
-
 	/* reset value */
 	u64 z_int_counter;
 	u64 z_rcv_limit;
 	u64 z_send_schedule;
-	/* percpu int_counter */
-	u64 __percpu *int_counter;
-	u64 __percpu *rcv_limit;
+
 	u64 __percpu *send_schedule;
 	/* number of receive contexts in use by the driver */
 	u32 num_rcv_contexts;
@@ -950,6 +935,7 @@ struct hfi1_devdata {
 	/* base receive interrupt timeout, in CSR units */
 	u32 rcv_intr_timeout_csr;
 
+	u32 freezelen; /* max length of freezemsg */
 	u64 __iomem *egrtidbase;
 	spinlock_t sendctrl_lock; /* protect changes to SendCtrl */
 	spinlock_t rcvctrl_lock; /* protect changes to RcvCtrl */
@@ -971,7 +957,6 @@ struct hfi1_devdata {
 	 * IB link status cheaply
 	 */
 	struct hfi1_status *status;
-	u32 freezelen; /* max length of freezemsg */
 
 	/* revision register shadow */
 	u64 revision;
@@ -999,6 +984,8 @@ struct hfi1_devdata {
 	u16 rcvegrbufsize_shift;
 	/* both sides of the PCIe link are gen3 capable */
 	u8 link_gen3_capable;
+	/* default link down value (poll/sleep) */
+	u8 link_default;
 	/* localbus width (1, 2,4,8,16,32) from config space  */
 	u32 lbus_width;
 	/* localbus speed in MHz */
@@ -1034,8 +1021,6 @@ struct hfi1_devdata {
 	u8 hfi1_id;
 	/* implementation code */
 	u8 icode;
-	/* default link down value (poll/sleep) */
-	u8 link_default;
 	/* vAU of this device */
 	u8 vau;
 	/* vCU of this device */
@@ -1046,27 +1031,17 @@ struct hfi1_devdata {
 	u16 vl15_init;
 
 	/* Misc small ints */
-	/* Number of physical ports available */
-	u8 num_pports;
-	/* Lowest context number which can be used by user processes */
-	u8 first_user_ctxt;
 	u8 n_krcv_queues;
 	u8 qos_shift;
-	u8 qpn_mask;
 
-	u16 rhf_offset; /* offset of RHF within receive header entry */
 	u16 irev;	/* implementation revision */
 	u16 dc8051_ver; /* 8051 firmware version */
 
+	spinlock_t hfi1_diag_trans_lock; /* protect diag observer ops */
 	struct platform_config platform_config;
 	struct platform_config_cache pcfg_cache;
 
 	struct diag_client *diag_client;
-	spinlock_t hfi1_diag_trans_lock; /* protect diag observer ops */
-
-	u8 psxmitwait_supported;
-	/* cycle length of PS* counters in HW (in picoseconds) */
-	u16 psxmitwait_check_rate;
 
 	/* MSI-X information */
 	struct hfi1_msix_entry *msix_entries;
@@ -1081,6 +1056,9 @@ struct hfi1_devdata {
 
 	struct rcv_array_data rcv_entries;
 
+	/* cycle length of PS* counters in HW (in picoseconds) */
+	u16 psxmitwait_check_rate;
+
 	/*
 	 * 64 bit synthetic counters
 	 */
@@ -1113,11 +1091,11 @@ struct hfi1_devdata {
 	struct err_info_rcvport err_info_rcvport;
 	struct err_info_constraint err_info_rcv_constraint;
 	struct err_info_constraint err_info_xmit_constraint;
-	u8 err_info_uncorrectable;
-	u8 err_info_fmconfig;
 
 	atomic_t drop_packet;
 	u8 do_drop;
+	u8 err_info_uncorrectable;
+	u8 err_info_fmconfig;
 
 	/*
 	 * Software counters for the status bits defined by the
@@ -1140,47 +1118,71 @@ struct hfi1_devdata {
 	u64 sw_cce_err_status_aggregate;
 	/* Software counter that aggregates all bypass packet rcv errors */
 	u64 sw_rcv_bypass_packet_errors;
-	/* receive interrupt functions */
-	rhf_rcv_function_ptr *rhf_rcv_function_map;
+	/* receive interrupt function */
 	rhf_rcv_function_ptr normal_rhf_rcv_functions[8];
 
+	/* Save the enabled LCB error bits */
+	u64 lcb_err_en;
+
 	/*
 	 * Handlers for outgoing data so that snoop/capture does not
 	 * have to have its hooks in the send path
 	 */
-	send_routine process_pio_send;
+	send_routine process_pio_send ____cacheline_aligned_in_smp;
 	send_routine process_dma_send;
 	void (*pio_inline_send)(struct hfi1_devdata *dd, struct pio_buf *pbuf,
 				u64 pbc, const void *from, size_t count);
+	/* hfi1_pportdata, points to array of (physical) port-specific
+	 * data structs, indexed by pidx (0..n-1)
+	 */
+	struct hfi1_pportdata *pport;
+	/* receive context data */
+	struct hfi1_ctxtdata **rcd;
+	u64 __percpu *int_counter;
+	/* device (not port) flags, basically device capabilities */
+	u16 flags;
+	/* Number of physical ports available */
+	u8 num_pports;
+	/* Lowest context number which can be used by user processes */
+	u8 first_user_ctxt;
+	/* adding a new field here would make it part of this cacheline */
+
+	/* seqlock for sc2vl */
+	seqlock_t sc2vl_lock ____cacheline_aligned_in_smp;
+	u64 sc2vl[4];
+	/* receive interrupt functions */
+	rhf_rcv_function_ptr *rhf_rcv_function_map;
+	u64 __percpu *rcv_limit;
+	u16 rhf_offset; /* offset of RHF within receive header entry */
+	/* adding a new field here would make it part of this cacheline */
 
 	/* OUI comes from the HW. Used everywhere as 3 separate bytes. */
 	u8 oui1;
 	u8 oui2;
 	u8 oui3;
+	u8 dc_shutdown;
+
 	/* Timer and counter used to detect RcvBufOvflCnt changes */
 	struct timer_list rcverr_timer;
-	u32 rcv_ovfl_cnt;
 
 	wait_queue_head_t event_queue;
 
-	/* Save the enabled LCB error bits */
-	u64 lcb_err_en;
-	u8 dc_shutdown;
-
 	/* receive context tail dummy address */
 	__le64 *rcvhdrtail_dummy_kvaddr;
 	dma_addr_t rcvhdrtail_dummy_dma;
 
-	bool eprom_available;	/* true if EPROM is available for this device */
-	bool aspm_supported;	/* Does HW support ASPM */
-	bool aspm_enabled;	/* ASPM state: enabled/disabled */
+	u32 rcv_ovfl_cnt;
 	/* Serialize ASPM enable/disable between multiple verbs contexts */
 	spinlock_t aspm_lock;
 	/* Number of verbs contexts which have disabled ASPM */
 	atomic_t aspm_disabled_cnt;
 
 	struct hfi1_affinity *affinity;
+	bool eprom_available;	/* true if EPROM is available for this device */
+	bool aspm_supported;	/* Does HW support ASPM */
+	bool aspm_enabled;	/* ASPM state: enabled/disabled */
 	struct rhashtable sdma_rht;
+
 	struct kobject kobj;
 };
 
-- 
cgit v1.2.1


From fe4d924396a861937256293ff4a84b76b84854d8 Mon Sep 17 00:00:00 2001
From: Easwar Hariharan <easwar.hariharan@intel.com>
Date: Mon, 17 Oct 2016 04:19:47 -0700
Subject: IB/hfi1: Add active channel and backplane support for integrated
 devices

Use scratch registers within the HFI1 device to recover signal
integrity information that is then used to tune the channel. While
there, update error messages to better convey the result of falling
back to a backup file.

Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Reviewed-by: Dean Luick <dean.luick@intel.com>
Signed-off-by: Easwar Hariharan <easwar.hariharan@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/chip_registers.h |   3 +
 drivers/infiniband/hw/hfi1/firmware.c       | 136 +++++++++++++++++++++----
 drivers/infiniband/hw/hfi1/hfi.h            |  16 ++-
 drivers/infiniband/hw/hfi1/platform.c       | 149 +++++++++++++++++++++++-----
 drivers/infiniband/hw/hfi1/platform.h       | 117 ++++++++++++++++++++++
 5 files changed, 376 insertions(+), 45 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/chip_registers.h b/drivers/infiniband/hw/hfi1/chip_registers.h
index 5b9993899789..5bfa839d1c48 100644
--- a/drivers/infiniband/hw/hfi1/chip_registers.h
+++ b/drivers/infiniband/hw/hfi1/chip_registers.h
@@ -415,6 +415,9 @@
 #define ASIC_CFG_SBUS_REQUEST_DATA_IN_SHIFT 32
 #define ASIC_CFG_SBUS_REQUEST_RECEIVER_ADDR_SHIFT 0
 #define ASIC_CFG_SCRATCH (ASIC + 0x000000000020)
+#define ASIC_CFG_SCRATCH_1 (ASIC_CFG_SCRATCH + 0x08)
+#define ASIC_CFG_SCRATCH_2 (ASIC_CFG_SCRATCH + 0x10)
+#define ASIC_CFG_SCRATCH_3 (ASIC_CFG_SCRATCH + 0x18)
 #define ASIC_CFG_THERM_POLL_EN (ASIC + 0x000000000050)
 #define ASIC_EEP_ADDR_CMD (ASIC + 0x000000000308)
 #define ASIC_EEP_ADDR_CMD_EP_ADDR_MASK 0xFFFFFFull
diff --git a/drivers/infiniband/hw/hfi1/firmware.c b/drivers/infiniband/hw/hfi1/firmware.c
index 13db8eb4f4ec..9a36c8c2f8a1 100644
--- a/drivers/infiniband/hw/hfi1/firmware.c
+++ b/drivers/infiniband/hw/hfi1/firmware.c
@@ -707,6 +707,9 @@ static int obtain_firmware(struct hfi1_devdata *dd)
 				       &dd->pcidev->dev);
 		if (err) {
 			platform_config = NULL;
+			dd_dev_err(dd,
+				   "%s: No default platform config file found\n",
+				   __func__);
 			goto done;
 		}
 		dd->platform_config.data = platform_config->data;
@@ -1761,8 +1764,17 @@ int parse_platform_config(struct hfi1_devdata *dd)
 	u32 record_idx = 0, table_type = 0, table_length_dwords = 0;
 	int ret = -EINVAL; /* assume failure */
 
+	/*
+	 * For integrated devices that did not fall back to the default file,
+	 * the SI tuning information for active channels is acquired from the
+	 * scratch register bitmap, thus there is no platform config to parse.
+	 * Skip parsing in these situations.
+	 */
+	if (is_integrated(dd) && !platform_config_load)
+		return 0;
+
 	if (!dd->platform_config.data) {
-		dd_dev_info(dd, "%s: Missing config file\n", __func__);
+		dd_dev_err(dd, "%s: Missing config file\n", __func__);
 		goto bail;
 	}
 	ptr = (u32 *)dd->platform_config.data;
@@ -1770,7 +1782,7 @@ int parse_platform_config(struct hfi1_devdata *dd)
 	magic_num = *ptr;
 	ptr++;
 	if (magic_num != PLATFORM_CONFIG_MAGIC_NUM) {
-		dd_dev_info(dd, "%s: Bad config file\n", __func__);
+		dd_dev_err(dd, "%s: Bad config file\n", __func__);
 		goto bail;
 	}
 
@@ -1797,9 +1809,9 @@ int parse_platform_config(struct hfi1_devdata *dd)
 		header1 = *ptr;
 		header2 = *(ptr + 1);
 		if (header1 != ~header2) {
-			dd_dev_info(dd, "%s: Failed validation at offset %ld\n",
-				    __func__, (ptr - (u32 *)
-					       dd->platform_config.data));
+			dd_dev_err(dd, "%s: Failed validation at offset %ld\n",
+				   __func__, (ptr - (u32 *)
+					      dd->platform_config.data));
 			goto bail;
 		}
 
@@ -1841,11 +1853,11 @@ int parse_platform_config(struct hfi1_devdata *dd)
 							table_length_dwords;
 				break;
 			default:
-				dd_dev_info(dd,
-					    "%s: Unknown data table %d, offset %ld\n",
-					    __func__, table_type,
-					    (ptr - (u32 *)
-					     dd->platform_config.data));
+				dd_dev_err(dd,
+					   "%s: Unknown data table %d, offset %ld\n",
+					   __func__, table_type,
+					   (ptr - (u32 *)
+					    dd->platform_config.data));
 				goto bail; /* We don't trust this file now */
 			}
 			pcfgcache->config_tables[table_type].table = ptr;
@@ -1865,11 +1877,11 @@ int parse_platform_config(struct hfi1_devdata *dd)
 			case PLATFORM_CONFIG_VARIABLE_SETTINGS_TABLE:
 				break;
 			default:
-				dd_dev_info(dd,
-					    "%s: Unknown meta table %d, offset %ld\n",
-					    __func__, table_type,
-					    (ptr -
-					     (u32 *)dd->platform_config.data));
+				dd_dev_err(dd,
+					   "%s: Unknown meta table %d, offset %ld\n",
+					   __func__, table_type,
+					   (ptr -
+					    (u32 *)dd->platform_config.data));
 				goto bail; /* We don't trust this file now */
 			}
 			pcfgcache->config_tables[table_type].table_metadata =
@@ -1884,10 +1896,9 @@ int parse_platform_config(struct hfi1_devdata *dd)
 		/* Jump the table */
 		ptr += table_length_dwords;
 		if (crc != *ptr) {
-			dd_dev_info(dd, "%s: Failed CRC check at offset %ld\n",
-				    __func__, (ptr -
-					       (u32 *)
-					       dd->platform_config.data));
+			dd_dev_err(dd, "%s: Failed CRC check at offset %ld\n",
+				   __func__, (ptr -
+				   (u32 *)dd->platform_config.data));
 			goto bail;
 		}
 		/* Jump the CRC DWORD */
@@ -1901,6 +1912,84 @@ bail:
 	return ret;
 }
 
+static void get_integrated_platform_config_field(
+		struct hfi1_devdata *dd,
+		enum platform_config_table_type_encoding table_type,
+		int field_index, u32 *data)
+{
+	struct hfi1_pportdata *ppd = dd->pport;
+	u8 *cache = ppd->qsfp_info.cache;
+	u32 tx_preset = 0;
+
+	switch (table_type) {
+	case PLATFORM_CONFIG_SYSTEM_TABLE:
+		if (field_index == SYSTEM_TABLE_QSFP_POWER_CLASS_MAX)
+			*data = ppd->max_power_class;
+		else if (field_index == SYSTEM_TABLE_QSFP_ATTENUATION_DEFAULT_25G)
+			*data = ppd->default_atten;
+		break;
+	case PLATFORM_CONFIG_PORT_TABLE:
+		if (field_index == PORT_TABLE_PORT_TYPE)
+			*data = ppd->port_type;
+		else if (field_index == PORT_TABLE_LOCAL_ATTEN_25G)
+			*data = ppd->local_atten;
+		else if (field_index == PORT_TABLE_REMOTE_ATTEN_25G)
+			*data = ppd->remote_atten;
+		break;
+	case PLATFORM_CONFIG_RX_PRESET_TABLE:
+		if (field_index == RX_PRESET_TABLE_QSFP_RX_CDR_APPLY)
+			*data = (ppd->rx_preset & QSFP_RX_CDR_APPLY_SMASK) >>
+				QSFP_RX_CDR_APPLY_SHIFT;
+		else if (field_index == RX_PRESET_TABLE_QSFP_RX_EMP_APPLY)
+			*data = (ppd->rx_preset & QSFP_RX_EMP_APPLY_SMASK) >>
+				QSFP_RX_EMP_APPLY_SHIFT;
+		else if (field_index == RX_PRESET_TABLE_QSFP_RX_AMP_APPLY)
+			*data = (ppd->rx_preset & QSFP_RX_AMP_APPLY_SMASK) >>
+				QSFP_RX_AMP_APPLY_SHIFT;
+		else if (field_index == RX_PRESET_TABLE_QSFP_RX_CDR)
+			*data = (ppd->rx_preset & QSFP_RX_CDR_SMASK) >>
+				QSFP_RX_CDR_SHIFT;
+		else if (field_index == RX_PRESET_TABLE_QSFP_RX_EMP)
+			*data = (ppd->rx_preset & QSFP_RX_EMP_SMASK) >>
+				QSFP_RX_EMP_SHIFT;
+		else if (field_index == RX_PRESET_TABLE_QSFP_RX_AMP)
+			*data = (ppd->rx_preset & QSFP_RX_AMP_SMASK) >>
+				QSFP_RX_AMP_SHIFT;
+		break;
+	case PLATFORM_CONFIG_TX_PRESET_TABLE:
+		if (cache[QSFP_EQ_INFO_OFFS] & 0x4)
+			tx_preset = ppd->tx_preset_eq;
+		else
+			tx_preset = ppd->tx_preset_noeq;
+		if (field_index == TX_PRESET_TABLE_PRECUR)
+			*data = (tx_preset & TX_PRECUR_SMASK) >>
+				TX_PRECUR_SHIFT;
+		else if (field_index == TX_PRESET_TABLE_ATTN)
+			*data = (tx_preset & TX_ATTN_SMASK) >>
+				TX_ATTN_SHIFT;
+		else if (field_index == TX_PRESET_TABLE_POSTCUR)
+			*data = (tx_preset & TX_POSTCUR_SMASK) >>
+				TX_POSTCUR_SHIFT;
+		else if (field_index == TX_PRESET_TABLE_QSFP_TX_CDR_APPLY)
+			*data = (tx_preset & QSFP_TX_CDR_APPLY_SMASK) >>
+				QSFP_TX_CDR_APPLY_SHIFT;
+		else if (field_index == TX_PRESET_TABLE_QSFP_TX_EQ_APPLY)
+			*data = (tx_preset & QSFP_TX_EQ_APPLY_SMASK) >>
+				QSFP_TX_EQ_APPLY_SHIFT;
+		else if (field_index == TX_PRESET_TABLE_QSFP_TX_CDR)
+			*data = (tx_preset & QSFP_TX_CDR_SMASK) >>
+				QSFP_TX_CDR_SHIFT;
+		else if (field_index == TX_PRESET_TABLE_QSFP_TX_EQ)
+			*data = (tx_preset & QSFP_TX_EQ_SMASK) >>
+				QSFP_TX_EQ_SHIFT;
+		break;
+	case PLATFORM_CONFIG_QSFP_ATTEN_TABLE:
+	case PLATFORM_CONFIG_VARIABLE_SETTINGS_TABLE:
+	default:
+		break;
+	}
+}
+
 static int get_platform_fw_field_metadata(struct hfi1_devdata *dd, int table,
 					  int field, u32 *field_len_bits,
 					  u32 *field_start_bits)
@@ -1976,6 +2065,15 @@ int get_platform_config_field(struct hfi1_devdata *dd,
 	else
 		return -EINVAL;
 
+	if (is_integrated(dd) && !platform_config_load) {
+		/*
+		 * Use saved configuration from ppd for integrated platforms
+		 */
+		get_integrated_platform_config_field(dd, table_type,
+						     field_index, data);
+		return 0;
+	}
+
 	ret = get_platform_fw_field_metadata(dd, table_type, field_index,
 					     &field_len_bits,
 					     &field_start_bits);
diff --git a/drivers/infiniband/hw/hfi1/hfi.h b/drivers/infiniband/hw/hfi1/hfi.h
index 25e44c4af995..a66d198199ff 100644
--- a/drivers/infiniband/hw/hfi1/hfi.h
+++ b/drivers/infiniband/hw/hfi1/hfi.h
@@ -582,8 +582,16 @@ struct hfi1_pportdata {
 	struct kobject vl2mtu_kobj;
 
 	/* PHY support */
-	u32 port_type;
 	struct qsfp_data qsfp_info;
+	/* Values for SI tuning of SerDes */
+	u32 port_type;
+	u32 tx_preset_eq;
+	u32 tx_preset_noeq;
+	u32 rx_preset;
+	u8  local_atten;
+	u8  remote_atten;
+	u8  default_atten;
+	u8  max_power_class;
 
 	/* GUIDs for this interface, in host order, guids[0] is a port guid */
 	u64 guids[HFI1_GUIDS_PER_PORT];
@@ -2020,6 +2028,12 @@ static inline u32 qsfp_resource(struct hfi1_devdata *dd)
 	return i2c_target(dd->hfi1_id);
 }
 
+/* Is this device integrated or discrete? */
+static inline bool is_integrated(struct hfi1_devdata *dd)
+{
+	return dd->pcidev->device == PCI_DEVICE_ID_INTEL1;
+}
+
 int hfi1_tempsense_rd(struct hfi1_devdata *dd, struct hfi1_temp *temp);
 
 #define DD_DEV_ENTRY(dd)       __string(dev, dev_name(&(dd)->pcidev->dev))
diff --git a/drivers/infiniband/hw/hfi1/platform.c b/drivers/infiniband/hw/hfi1/platform.c
index 202433178864..0e7043801bb1 100644
--- a/drivers/infiniband/hw/hfi1/platform.c
+++ b/drivers/infiniband/hw/hfi1/platform.c
@@ -49,6 +49,90 @@
 #include "efivar.h"
 #include "eprom.h"
 
+static int validate_scratch_checksum(struct hfi1_devdata *dd)
+{
+	u64 checksum = 0, temp_scratch = 0;
+	int i, j, version;
+
+	temp_scratch = read_csr(dd, ASIC_CFG_SCRATCH);
+	version = (temp_scratch & BITMAP_VERSION_SMASK) >> BITMAP_VERSION_SHIFT;
+
+	/* Prevent power on default of all zeroes from passing checksum */
+	if (!version)
+		return 0;
+
+	/*
+	 * ASIC scratch 0 only contains the checksum and bitmap version as
+	 * fields of interest, both of which are handled separately from the
+	 * loop below, so skip it
+	 */
+	checksum += version;
+	for (i = 1; i < ASIC_NUM_SCRATCH; i++) {
+		temp_scratch = read_csr(dd, ASIC_CFG_SCRATCH + (8 * i));
+		for (j = sizeof(u64); j != 0; j -= 2) {
+			checksum += (temp_scratch & 0xFFFF);
+			temp_scratch >>= 16;
+		}
+	}
+
+	while (checksum >> 16)
+		checksum = (checksum & CHECKSUM_MASK) + (checksum >> 16);
+
+	temp_scratch = read_csr(dd, ASIC_CFG_SCRATCH);
+	temp_scratch &= CHECKSUM_SMASK;
+	temp_scratch >>= CHECKSUM_SHIFT;
+
+	if (checksum + temp_scratch == 0xFFFF)
+		return 1;
+	return 0;
+}
+
+static void save_platform_config_fields(struct hfi1_devdata *dd)
+{
+	struct hfi1_pportdata *ppd = dd->pport;
+	u64 temp_scratch = 0, temp_dest = 0;
+
+	temp_scratch = read_csr(dd, ASIC_CFG_SCRATCH_1);
+
+	temp_dest = temp_scratch &
+		    (dd->hfi1_id ? PORT1_PORT_TYPE_SMASK :
+		     PORT0_PORT_TYPE_SMASK);
+	ppd->port_type = temp_dest >>
+			 (dd->hfi1_id ? PORT1_PORT_TYPE_SHIFT :
+			  PORT0_PORT_TYPE_SHIFT);
+
+	temp_dest = temp_scratch &
+		    (dd->hfi1_id ? PORT1_LOCAL_ATTEN_SMASK :
+		     PORT0_LOCAL_ATTEN_SMASK);
+	ppd->local_atten = temp_dest >>
+			   (dd->hfi1_id ? PORT1_LOCAL_ATTEN_SHIFT :
+			    PORT0_LOCAL_ATTEN_SHIFT);
+
+	temp_dest = temp_scratch &
+		    (dd->hfi1_id ? PORT1_REMOTE_ATTEN_SMASK :
+		     PORT0_REMOTE_ATTEN_SMASK);
+	ppd->remote_atten = temp_dest >>
+			    (dd->hfi1_id ? PORT1_REMOTE_ATTEN_SHIFT :
+			     PORT0_REMOTE_ATTEN_SHIFT);
+
+	temp_dest = temp_scratch &
+		    (dd->hfi1_id ? PORT1_DEFAULT_ATTEN_SMASK :
+		     PORT0_DEFAULT_ATTEN_SMASK);
+	ppd->default_atten = temp_dest >>
+			     (dd->hfi1_id ? PORT1_DEFAULT_ATTEN_SHIFT :
+			      PORT0_DEFAULT_ATTEN_SHIFT);
+
+	temp_scratch = read_csr(dd, dd->hfi1_id ? ASIC_CFG_SCRATCH_3 :
+				ASIC_CFG_SCRATCH_2);
+
+	ppd->tx_preset_eq = (temp_scratch & TX_EQ_SMASK) >> TX_EQ_SHIFT;
+	ppd->tx_preset_noeq = (temp_scratch & TX_NO_EQ_SMASK) >> TX_NO_EQ_SHIFT;
+	ppd->rx_preset = (temp_scratch & RX_SMASK) >> RX_SHIFT;
+
+	ppd->max_power_class = (temp_scratch & QSFP_MAX_POWER_SMASK) >>
+				QSFP_MAX_POWER_SHIFT;
+}
+
 void get_platform_config(struct hfi1_devdata *dd)
 {
 	int ret = 0;
@@ -56,38 +140,49 @@ void get_platform_config(struct hfi1_devdata *dd)
 	u8 *temp_platform_config = NULL;
 	u32 esize;
 
-	ret = eprom_read_platform_config(dd, (void **)&temp_platform_config,
-					 &esize);
-	if (!ret) {
-		/* success */
-		size = esize;
-		goto success;
+	if (is_integrated(dd)) {
+		if (validate_scratch_checksum(dd)) {
+			save_platform_config_fields(dd);
+			return;
+		}
+		dd_dev_err(dd, "%s: Config bitmap corrupted/uninitialized\n",
+			   __func__);
+		dd_dev_err(dd,
+			   "%s: Please update your BIOS to support active channels\n",
+			   __func__);
+	} else {
+		ret = eprom_read_platform_config(dd,
+						 (void **)&temp_platform_config,
+						 &esize);
+		if (!ret) {
+			/* success */
+			dd->platform_config.data = temp_platform_config;
+			dd->platform_config.size = esize;
+			return;
+		}
+		/* fail, try EFI variable */
+
+		ret = read_hfi1_efi_var(dd, "configuration", &size,
+					(void **)&temp_platform_config);
+		if (!ret) {
+			dd->platform_config.data = temp_platform_config;
+			dd->platform_config.size = size;
+			return;
+		}
 	}
-	/* fail, try EFI variable */
-
-	ret = read_hfi1_efi_var(dd, "configuration", &size,
-				(void **)&temp_platform_config);
-	if (!ret)
-		goto success;
-
-	dd_dev_info(dd,
-		    "%s: Failed to get platform config from UEFI, falling back to request firmware\n",
-		    __func__);
+	dd_dev_err(dd,
+		   "%s: Failed to get platform config, falling back to sub-optimal default file\n",
+		   __func__);
 	/* fall back to request firmware */
 	platform_config_load = 1;
-	return;
-
-success:
-	dd->platform_config.data = temp_platform_config;
-	dd->platform_config.size = size;
 }
 
 void free_platform_config(struct hfi1_devdata *dd)
 {
 	if (!platform_config_load) {
 		/*
-		 * was loaded from EFI, release memory
-		 * allocated by read_efi_var
+		 * was loaded from EFI or the EPROM, release memory
+		 * allocated by read_efi_var/eprom_read_platform_config
 		 */
 		kfree(dd->platform_config.data);
 	}
@@ -100,12 +195,16 @@ void free_platform_config(struct hfi1_devdata *dd)
 void get_port_type(struct hfi1_pportdata *ppd)
 {
 	int ret;
+	u32 temp;
 
 	ret = get_platform_config_field(ppd->dd, PLATFORM_CONFIG_PORT_TABLE, 0,
-					PORT_TABLE_PORT_TYPE, &ppd->port_type,
+					PORT_TABLE_PORT_TYPE, &temp,
 					4);
-	if (ret)
+	if (ret) {
 		ppd->port_type = PORT_TYPE_UNKNOWN;
+		return;
+	}
+	ppd->port_type = temp;
 }
 
 int set_qsfp_tx(struct hfi1_pportdata *ppd, int on)
diff --git a/drivers/infiniband/hw/hfi1/platform.h b/drivers/infiniband/hw/hfi1/platform.h
index e2c21613c326..31155401a95b 100644
--- a/drivers/infiniband/hw/hfi1/platform.h
+++ b/drivers/infiniband/hw/hfi1/platform.h
@@ -295,6 +295,123 @@ enum link_tuning_encoding {
 	OPA_UNKNOWN_TUNING
 };
 
+/*
+ * Shifts and masks for the link SI tuning values stuffed into the ASIC scratch
+ * registers for integrated platforms
+ */
+#define PORT0_PORT_TYPE_SHIFT		0
+#define PORT0_LOCAL_ATTEN_SHIFT		4
+#define PORT0_REMOTE_ATTEN_SHIFT	10
+#define PORT0_DEFAULT_ATTEN_SHIFT	32
+
+#define PORT1_PORT_TYPE_SHIFT		16
+#define PORT1_LOCAL_ATTEN_SHIFT		20
+#define PORT1_REMOTE_ATTEN_SHIFT	26
+#define PORT1_DEFAULT_ATTEN_SHIFT	40
+
+#define PORT0_PORT_TYPE_MASK		0xFUL
+#define PORT0_LOCAL_ATTEN_MASK		0x3FUL
+#define PORT0_REMOTE_ATTEN_MASK		0x3FUL
+#define PORT0_DEFAULT_ATTEN_MASK	0xFFUL
+
+#define PORT1_PORT_TYPE_MASK		0xFUL
+#define PORT1_LOCAL_ATTEN_MASK		0x3FUL
+#define PORT1_REMOTE_ATTEN_MASK		0x3FUL
+#define PORT1_DEFAULT_ATTEN_MASK	0xFFUL
+
+#define PORT0_PORT_TYPE_SMASK		(PORT0_PORT_TYPE_MASK << \
+					 PORT0_PORT_TYPE_SHIFT)
+#define PORT0_LOCAL_ATTEN_SMASK		(PORT0_LOCAL_ATTEN_MASK << \
+					 PORT0_LOCAL_ATTEN_SHIFT)
+#define PORT0_REMOTE_ATTEN_SMASK	(PORT0_REMOTE_ATTEN_MASK << \
+					 PORT0_REMOTE_ATTEN_SHIFT)
+#define PORT0_DEFAULT_ATTEN_SMASK	(PORT0_DEFAULT_ATTEN_MASK << \
+					 PORT0_DEFAULT_ATTEN_SHIFT)
+
+#define PORT1_PORT_TYPE_SMASK		(PORT1_PORT_TYPE_MASK << \
+					 PORT1_PORT_TYPE_SHIFT)
+#define PORT1_LOCAL_ATTEN_SMASK		(PORT1_LOCAL_ATTEN_MASK << \
+					 PORT1_LOCAL_ATTEN_SHIFT)
+#define PORT1_REMOTE_ATTEN_SMASK	(PORT1_REMOTE_ATTEN_MASK << \
+					 PORT1_REMOTE_ATTEN_SHIFT)
+#define PORT1_DEFAULT_ATTEN_SMASK	(PORT1_DEFAULT_ATTEN_MASK << \
+					 PORT1_DEFAULT_ATTEN_SHIFT)
+
+#define QSFP_MAX_POWER_SHIFT		0
+#define TX_NO_EQ_SHIFT			4
+#define TX_EQ_SHIFT			25
+#define RX_SHIFT			46
+
+#define QSFP_MAX_POWER_MASK		0xFUL
+#define TX_NO_EQ_MASK			0x1FFFFFUL
+#define TX_EQ_MASK			0x1FFFFFUL
+#define RX_MASK				0xFFFFUL
+
+#define QSFP_MAX_POWER_SMASK		(QSFP_MAX_POWER_MASK << \
+					 QSFP_MAX_POWER_SHIFT)
+#define TX_NO_EQ_SMASK			(TX_NO_EQ_MASK << TX_NO_EQ_SHIFT)
+#define TX_EQ_SMASK			(TX_EQ_MASK << TX_EQ_SHIFT)
+#define RX_SMASK			(RX_MASK << RX_SHIFT)
+
+#define TX_PRECUR_SHIFT			0
+#define TX_ATTN_SHIFT			4
+#define QSFP_TX_CDR_APPLY_SHIFT		9
+#define QSFP_TX_EQ_APPLY_SHIFT		10
+#define QSFP_TX_CDR_SHIFT		11
+#define QSFP_TX_EQ_SHIFT		12
+#define TX_POSTCUR_SHIFT		16
+
+#define TX_PRECUR_MASK			0xFUL
+#define TX_ATTN_MASK			0x1FUL
+#define QSFP_TX_CDR_APPLY_MASK		0x1UL
+#define QSFP_TX_EQ_APPLY_MASK		0x1UL
+#define QSFP_TX_CDR_MASK		0x1UL
+#define QSFP_TX_EQ_MASK			0xFUL
+#define TX_POSTCUR_MASK			0x1FUL
+
+#define TX_PRECUR_SMASK			(TX_PRECUR_MASK << TX_PRECUR_SHIFT)
+#define TX_ATTN_SMASK			(TX_ATTN_MASK << TX_ATTN_SHIFT)
+#define QSFP_TX_CDR_APPLY_SMASK		(QSFP_TX_CDR_APPLY_MASK << \
+					 QSFP_TX_CDR_APPLY_SHIFT)
+#define QSFP_TX_EQ_APPLY_SMASK		(QSFP_TX_EQ_APPLY_MASK << \
+					 QSFP_TX_EQ_APPLY_SHIFT)
+#define QSFP_TX_CDR_SMASK		(QSFP_TX_CDR_MASK << QSFP_TX_CDR_SHIFT)
+#define QSFP_TX_EQ_SMASK		(QSFP_TX_EQ_MASK << QSFP_TX_EQ_SHIFT)
+#define TX_POSTCUR_SMASK		(TX_POSTCUR_MASK << TX_POSTCUR_SHIFT)
+
+#define QSFP_RX_CDR_APPLY_SHIFT		0
+#define QSFP_RX_EMP_APPLY_SHIFT		1
+#define QSFP_RX_AMP_APPLY_SHIFT		2
+#define QSFP_RX_CDR_SHIFT		3
+#define QSFP_RX_EMP_SHIFT		4
+#define QSFP_RX_AMP_SHIFT		8
+
+#define QSFP_RX_CDR_APPLY_MASK		0x1UL
+#define QSFP_RX_EMP_APPLY_MASK		0x1UL
+#define QSFP_RX_AMP_APPLY_MASK		0x1UL
+#define QSFP_RX_CDR_MASK		0x1UL
+#define QSFP_RX_EMP_MASK		0xFUL
+#define QSFP_RX_AMP_MASK		0x3UL
+
+#define QSFP_RX_CDR_APPLY_SMASK		(QSFP_RX_CDR_APPLY_MASK << \
+					 QSFP_RX_CDR_APPLY_SHIFT)
+#define QSFP_RX_EMP_APPLY_SMASK		(QSFP_RX_EMP_APPLY_MASK << \
+					 QSFP_RX_EMP_APPLY_SHIFT)
+#define QSFP_RX_AMP_APPLY_SMASK		(QSFP_RX_AMP_APPLY_MASK << \
+					 QSFP_RX_AMP_APPLY_SHIFT)
+#define QSFP_RX_CDR_SMASK		(QSFP_RX_CDR_MASK << QSFP_RX_CDR_SHIFT)
+#define QSFP_RX_EMP_SMASK		(QSFP_RX_EMP_MASK << QSFP_RX_EMP_SHIFT)
+#define QSFP_RX_AMP_SMASK		(QSFP_RX_AMP_MASK << QSFP_RX_AMP_SHIFT)
+
+#define BITMAP_VERSION			1
+#define BITMAP_VERSION_SHIFT		44
+#define BITMAP_VERSION_MASK		0xFUL
+#define BITMAP_VERSION_SMASK		(BITMAP_VERSION_MASK << \
+					 BITMAP_VERSION_SHIFT)
+#define CHECKSUM_SHIFT			48
+#define CHECKSUM_MASK			0xFFFFUL
+#define CHECKSUM_SMASK			(CHECKSUM_MASK << CHECKSUM_SHIFT)
+
 /* platform.c */
 void get_platform_config(struct hfi1_devdata *dd);
 void free_platform_config(struct hfi1_devdata *dd);
-- 
cgit v1.2.1


From 2474d775d9e2f935ff6840c8b21b4262afacc821 Mon Sep 17 00:00:00 2001
From: Sebastian Sanchez <sebastian.sanchez@intel.com>
Date: Tue, 25 Oct 2016 13:12:28 -0700
Subject: IB/hfi1: Get rid of divide in pio buffer allocator

The div instruction shows costly in profiles.

Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/pio.c | 11 +++++++----
 drivers/infiniband/hw/hfi1/pio.h |  1 +
 2 files changed, 8 insertions(+), 4 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/pio.c b/drivers/infiniband/hw/hfi1/pio.c
index 385e4dcf2cd3..516fac38d31e 100644
--- a/drivers/infiniband/hw/hfi1/pio.c
+++ b/drivers/infiniband/hw/hfi1/pio.c
@@ -1249,6 +1249,7 @@ int sc_enable(struct send_context *sc)
 	sc->free = 0;
 	sc->alloc_free = 0;
 	sc->fill = 0;
+	sc->fill_wrap = 0;
 	sc->sr_head = 0;
 	sc->sr_tail = 0;
 	sc->flags = 0;
@@ -1392,7 +1393,7 @@ struct pio_buf *sc_buffer_alloc(struct send_context *sc, u32 dw_len,
 	unsigned long flags;
 	unsigned long avail;
 	unsigned long blocks = dwords_to_blocks(dw_len);
-	unsigned long start_fill;
+	u32 fill_wrap;
 	int trycount = 0;
 	u32 head, next;
 
@@ -1435,8 +1436,11 @@ retry:
 	head = sc->sr_head;
 
 	/* "allocate" the buffer */
-	start_fill = sc->fill;
 	sc->fill += blocks;
+	fill_wrap = sc->fill_wrap;
+	sc->fill_wrap += blocks;
+	if (sc->fill_wrap >= sc->credits)
+		sc->fill_wrap = sc->fill_wrap - sc->credits;
 
 	/*
 	 * Fill the parts that the releaser looks at before moving the head.
@@ -1465,8 +1469,7 @@ retry:
 	spin_unlock_irqrestore(&sc->alloc_lock, flags);
 
 	/* finish filling in the buffer outside the lock */
-	pbuf->start = sc->base_addr + ((start_fill % sc->credits)
-							* PIO_BLOCK_SIZE);
+	pbuf->start = sc->base_addr + fill_wrap * PIO_BLOCK_SIZE;
 	pbuf->size = sc->credits * PIO_BLOCK_SIZE;
 	pbuf->end = sc->base_addr + pbuf->size;
 	pbuf->block_count = blocks;
diff --git a/drivers/infiniband/hw/hfi1/pio.h b/drivers/infiniband/hw/hfi1/pio.h
index bd19507b6bb0..498b548055e0 100644
--- a/drivers/infiniband/hw/hfi1/pio.h
+++ b/drivers/infiniband/hw/hfi1/pio.h
@@ -119,6 +119,7 @@ struct send_context {
 	unsigned long fill;		/* official alloc count */
 	unsigned long alloc_free;	/* copy of free (less cache thrash) */
 	u32 __percpu *buffers_allocated;/* count of buffers allocated */
+	u32 fill_wrap;			/* tracks fill within ring */
 	/* releaser fields */
 	spinlock_t release_lock ____cacheline_aligned_in_smp;
 	u32 sr_tail;			/* shadow ring tail */
-- 
cgit v1.2.1


From 8af8d2970ed98493a2db88dfcad88b0065e55e79 Mon Sep 17 00:00:00 2001
From: Sebastian Sanchez <sebastian.sanchez@intel.com>
Date: Tue, 25 Oct 2016 13:12:34 -0700
Subject: IB/hfi1: Optimize pio_buf and send_context structs

Both pio_buf and send_context structs have oversized
fields and have cachelines that can be optimized.

Reduce oversized fields for both structs.
Make sure pio_buf struct fits within a cacheline.
Move read-only fields to their own cacheline in
send_context struct.

All of this will avoid cacheline trading as the ring
progresses and pio buffers/send contexts are used.

Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/pio.c      |  5 ++---
 drivers/infiniband/hw/hfi1/pio.h      | 29 +++++++++++++++--------------
 drivers/infiniband/hw/hfi1/pio_copy.c | 22 +++++++++++-----------
 3 files changed, 28 insertions(+), 28 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/pio.c b/drivers/infiniband/hw/hfi1/pio.c
index 516fac38d31e..86a7f365b624 100644
--- a/drivers/infiniband/hw/hfi1/pio.c
+++ b/drivers/infiniband/hw/hfi1/pio.c
@@ -765,6 +765,7 @@ struct send_context *sc_alloc(struct hfi1_devdata *dd, int type,
 	sc->hw_context = hw_context;
 	cr_group_addresses(sc, &dma);
 	sc->credits = sci->credits;
+	sc->size = sc->credits * PIO_BLOCK_SIZE;
 
 /* PIO Send Memory Address details */
 #define PIO_ADDR_CONTEXT_MASK 0xfful
@@ -1470,9 +1471,7 @@ retry:
 
 	/* finish filling in the buffer outside the lock */
 	pbuf->start = sc->base_addr + fill_wrap * PIO_BLOCK_SIZE;
-	pbuf->size = sc->credits * PIO_BLOCK_SIZE;
-	pbuf->end = sc->base_addr + pbuf->size;
-	pbuf->block_count = blocks;
+	pbuf->end = sc->base_addr + sc->size;
 	pbuf->qw_written = 0;
 	pbuf->carry_bytes = 0;
 	pbuf->carry.val64 = 0;
diff --git a/drivers/infiniband/hw/hfi1/pio.h b/drivers/infiniband/hw/hfi1/pio.h
index 498b548055e0..867e5ffc3595 100644
--- a/drivers/infiniband/hw/hfi1/pio.h
+++ b/drivers/infiniband/hw/hfi1/pio.h
@@ -83,43 +83,43 @@ struct pio_buf {
 	void *arg;		/* argument for cb */
 	void __iomem *start;	/* buffer start address */
 	void __iomem *end;	/* context end address */
-	unsigned long size;	/* context size, in bytes */
 	unsigned long sent_at;	/* buffer is sent when <= free */
-	u32 block_count;	/* size of buffer, in blocks */
-	u32 qw_written;		/* QW written so far */
-	u32 carry_bytes;	/* number of valid bytes in carry */
 	union mix carry;	/* pending unwritten bytes */
+	u16 qw_written;		/* QW written so far */
+	u8 carry_bytes;	/* number of valid bytes in carry */
 };
 
 /* cache line aligned pio buffer array */
 union pio_shadow_ring {
 	struct pio_buf pbuf;
-	u64 unused[16];		/* cache line spacer */
 } ____cacheline_aligned;
 
 /* per-NUMA send context */
 struct send_context {
 	/* read-only after init */
 	struct hfi1_devdata *dd;		/* device */
-	void __iomem *base_addr;	/* start of PIO memory */
 	union pio_shadow_ring *sr;	/* shadow ring */
+	void __iomem *base_addr;	/* start of PIO memory */
+	u32 __percpu *buffers_allocated;/* count of buffers allocated */
+	u32 size;			/* context size, in bytes */
 
-	struct work_struct halt_work;	/* halted context work queue entry */
-	unsigned long flags;		/* flags */
 	int node;			/* context home node */
-	int type;			/* context type */
-	u32 sw_index;			/* software index number */
-	u32 hw_context;			/* hardware context number */
-	u32 credits;			/* number of blocks in context */
 	u32 sr_size;			/* size of the shadow ring */
-	u32 group;			/* credit return group */
+	u16 flags;			/* flags */
+	u8  type;			/* context type */
+	u8  sw_index;			/* software index number */
+	u8  hw_context;			/* hardware context number */
+	u8  group;			/* credit return group */
+
 	/* allocator fields */
 	spinlock_t alloc_lock ____cacheline_aligned_in_smp;
 	u32 sr_head;			/* shadow ring head */
 	unsigned long fill;		/* official alloc count */
 	unsigned long alloc_free;	/* copy of free (less cache thrash) */
-	u32 __percpu *buffers_allocated;/* count of buffers allocated */
 	u32 fill_wrap;			/* tracks fill within ring */
+	u32 credits;			/* number of blocks in context */
+	/* adding a new field here would make it part of this cacheline */
+
 	/* releaser fields */
 	spinlock_t release_lock ____cacheline_aligned_in_smp;
 	u32 sr_tail;			/* shadow ring tail */
@@ -131,6 +131,7 @@ struct send_context {
 	u32 credit_intr_count;		/* count of credit intr users */
 	u64 credit_ctrl;		/* cache for credit control */
 	wait_queue_head_t halt_wait;    /* wait until kernel sees interrupt */
+	struct work_struct halt_work;	/* halted context work queue entry */
 };
 
 /* send context flags */
diff --git a/drivers/infiniband/hw/hfi1/pio_copy.c b/drivers/infiniband/hw/hfi1/pio_copy.c
index aa7773643107..03024cec78dd 100644
--- a/drivers/infiniband/hw/hfi1/pio_copy.c
+++ b/drivers/infiniband/hw/hfi1/pio_copy.c
@@ -129,8 +129,8 @@ void pio_copy(struct hfi1_devdata *dd, struct pio_buf *pbuf, u64 pbc,
 				dest += sizeof(u64);
 			}
 
-			dest -= pbuf->size;
-			dend -= pbuf->size;
+			dest -= pbuf->sc->size;
+			dend -= pbuf->sc->size;
 		}
 
 		/* write 8-byte non-SOP, non-wrap chunk data */
@@ -361,8 +361,8 @@ void seg_pio_copy_start(struct pio_buf *pbuf, u64 pbc,
 				dest += sizeof(u64);
 			}
 
-			dest -= pbuf->size;
-			dend -= pbuf->size;
+			dest -= pbuf->sc->size;
+			dend -= pbuf->sc->size;
 		}
 
 		/* write 8-byte non-SOP, non-wrap chunk data */
@@ -458,8 +458,8 @@ static void mid_copy_mix(struct pio_buf *pbuf, const void *from, size_t nbytes)
 			dest += sizeof(u64);
 		}
 
-		dest -= pbuf->size;
-		dend -= pbuf->size;
+		dest -= pbuf->sc->size;
+		dend -= pbuf->sc->size;
 	}
 
 	/* write 8-byte non-SOP, non-wrap chunk data */
@@ -492,7 +492,7 @@ static void mid_copy_mix(struct pio_buf *pbuf, const void *from, size_t nbytes)
 		 */
 		/* adjust if we have wrapped */
 		if (dest >= pbuf->end)
-			dest -= pbuf->size;
+			dest -= pbuf->sc->size;
 		/* jump to the SOP range if within the first block */
 		else if (pbuf->qw_written < PIO_BLOCK_QWS)
 			dest += SOP_DISTANCE;
@@ -584,8 +584,8 @@ static void mid_copy_straight(struct pio_buf *pbuf,
 			dest += sizeof(u64);
 		}
 
-		dest -= pbuf->size;
-		dend -= pbuf->size;
+		dest -= pbuf->sc->size;
+		dend -= pbuf->sc->size;
 	}
 
 	/* write 8-byte non-SOP, non-wrap chunk data */
@@ -666,7 +666,7 @@ void seg_pio_copy_mid(struct pio_buf *pbuf, const void *from, size_t nbytes)
 			 */
 			/* adjust if we've wrapped */
 			if (dest >= pbuf->end)
-				dest -= pbuf->size;
+				dest -= pbuf->sc->size;
 			/* jump to SOP range if within the first block */
 			else if (pbuf->qw_written < PIO_BLOCK_QWS)
 				dest += SOP_DISTANCE;
@@ -719,7 +719,7 @@ void seg_pio_copy_end(struct pio_buf *pbuf)
 	 */
 	/* adjust if we have wrapped */
 	if (dest >= pbuf->end)
-		dest -= pbuf->size;
+		dest -= pbuf->sc->size;
 	/* jump to the SOP range if within the first block */
 	else if (pbuf->qw_written < PIO_BLOCK_QWS)
 		dest += SOP_DISTANCE;
-- 
cgit v1.2.1


From 8ecc7985b4b15f1f14bce31d8ab45dc426df7da3 Mon Sep 17 00:00:00 2001
From: Mark Bloch <markb@mellanox.com>
Date: Thu, 27 Oct 2016 16:36:30 +0300
Subject: IB/core: Save QP in ib_flow structure

When we create flow steering rule, we need to save the related QP in the
ib_flow struct. this QP is used in destroy flow.

Move the QP assignment from ib_uverbs_ex_create_flow into ib_create_flow,
this would allow both kernel and userspace consumers to use it.

This bug wasn't seen in the wild because there are no kernel consumers
currently in the kernel.

Fixes: 319a441d1361 ("IB/core: Add receive flow steering support")
Signed-off-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/uverbs_cmd.c | 1 -
 drivers/infiniband/core/verbs.c      | 4 +++-
 2 files changed, 3 insertions(+), 2 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c
index cb3f515a2285..2690c9263ee4 100644
--- a/drivers/infiniband/core/uverbs_cmd.c
+++ b/drivers/infiniband/core/uverbs_cmd.c
@@ -3745,7 +3745,6 @@ int ib_uverbs_ex_create_flow(struct ib_uverbs_file *file,
 		err = PTR_ERR(flow_id);
 		goto err_free;
 	}
-	flow_id->qp = qp;
 	flow_id->uobject = uobj;
 	uobj->object = flow_id;
 
diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c
index 83687646da68..b7f9f3d95975 100644
--- a/drivers/infiniband/core/verbs.c
+++ b/drivers/infiniband/core/verbs.c
@@ -1734,8 +1734,10 @@ struct ib_flow *ib_create_flow(struct ib_qp *qp,
 		return ERR_PTR(-ENOSYS);
 
 	flow_id = qp->device->create_flow(qp, flow_attr, domain);
-	if (!IS_ERR(flow_id))
+	if (!IS_ERR(flow_id)) {
 		atomic_inc(&qp->usecnt);
+		flow_id->qp = qp;
+	}
 	return flow_id;
 }
 EXPORT_SYMBOL(ib_create_flow);
-- 
cgit v1.2.1


From 2d2215888d758c194efb38332aaf5e2069ac578e Mon Sep 17 00:00:00 2001
From: Max Gurtovoy <maxg@mellanox.com>
Date: Thu, 27 Oct 2016 16:36:36 +0300
Subject: IB/mlx5: Replace numerical constant with predefined MACRO

Replace the pre-defined macro signifying inline umr instead
of the numerical constant.

Signed-off-by: Max Gurtovoy <maxg@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/qp.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c
index 41f4c2afbcdd..00cffbfe6c35 100644
--- a/drivers/infiniband/hw/mlx5/qp.c
+++ b/drivers/infiniband/hw/mlx5/qp.c
@@ -3068,7 +3068,7 @@ static void set_linv_umr_seg(struct mlx5_wqe_umr_ctrl_seg *umr)
 {
 	memset(umr, 0, sizeof(*umr));
 	umr->mkey_mask = cpu_to_be64(MLX5_MKEY_MASK_FREE);
-	umr->flags = 1 << 7;
+	umr->flags = MLX5_UMR_INLINE;
 }
 
 static __be64 get_umr_reg_mr_mask(void)
-- 
cgit v1.2.1


From 578e72647ba5a77a09004606cb572f0881c34e0d Mon Sep 17 00:00:00 2001
From: Maor Gottlieb <maorg@mellanox.com>
Date: Thu, 27 Oct 2016 16:36:37 +0300
Subject: IB/mlx5: Fix atomic cap in indirect UMR

Remove from the driver the limitation imposed by firmware check
to not allow change of atomic permissions for indirect UMRs.
In order to avoid failures on old firmware, we only ask for change
of atomic permissions if atomic operations are supported.

Fixes: 968e78dd9644 ('IB/mlx5: Enhance UMR support to allow partial page table update')
Signed-off-by: Eli Cohen <eli@mellanox.com>
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/qp.c | 12 +++++++-----
 1 file changed, 7 insertions(+), 5 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c
index 00cffbfe6c35..5bdf20c676fe 100644
--- a/drivers/infiniband/hw/mlx5/qp.c
+++ b/drivers/infiniband/hw/mlx5/qp.c
@@ -3071,7 +3071,7 @@ static void set_linv_umr_seg(struct mlx5_wqe_umr_ctrl_seg *umr)
 	umr->flags = MLX5_UMR_INLINE;
 }
 
-static __be64 get_umr_reg_mr_mask(void)
+static __be64 get_umr_reg_mr_mask(int atomic)
 {
 	u64 result;
 
@@ -3084,9 +3084,11 @@ static __be64 get_umr_reg_mr_mask(void)
 		 MLX5_MKEY_MASK_KEY		|
 		 MLX5_MKEY_MASK_RR		|
 		 MLX5_MKEY_MASK_RW		|
-		 MLX5_MKEY_MASK_A		|
 		 MLX5_MKEY_MASK_FREE;
 
+	if (atomic)
+		result |= MLX5_MKEY_MASK_A;
+
 	return cpu_to_be64(result);
 }
 
@@ -3147,7 +3149,7 @@ static __be64 get_umr_update_pd_mask(void)
 }
 
 static void set_reg_umr_segment(struct mlx5_wqe_umr_ctrl_seg *umr,
-				struct ib_send_wr *wr)
+				struct ib_send_wr *wr, int atomic)
 {
 	struct mlx5_umr_wr *umrwr = umr_wr(wr);
 
@@ -3172,7 +3174,7 @@ static void set_reg_umr_segment(struct mlx5_wqe_umr_ctrl_seg *umr,
 		if (wr->send_flags & MLX5_IB_SEND_UMR_UPDATE_PD)
 			umr->mkey_mask |= get_umr_update_pd_mask();
 		if (!umr->mkey_mask)
-			umr->mkey_mask = get_umr_reg_mr_mask();
+			umr->mkey_mask = get_umr_reg_mr_mask(atomic);
 	} else {
 		umr->mkey_mask = get_umr_unreg_mr_mask();
 	}
@@ -4025,7 +4027,7 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 			}
 			qp->sq.wr_data[idx] = MLX5_IB_WR_UMR;
 			ctrl->imm = cpu_to_be32(umr_wr(wr)->mkey);
-			set_reg_umr_segment(seg, wr);
+			set_reg_umr_segment(seg, wr, !!(MLX5_CAP_GEN(mdev, atomic)));
 			seg += sizeof(struct mlx5_wqe_umr_ctrl_seg);
 			size += sizeof(struct mlx5_wqe_umr_ctrl_seg) / 16;
 			if (unlikely((seg == qend)))
-- 
cgit v1.2.1


From 86695a6582e3b1c4895de2bde4e1022b3a8fbda0 Mon Sep 17 00:00:00 2001
From: Maor Gottlieb <maorg@mellanox.com>
Date: Thu, 27 Oct 2016 16:36:38 +0300
Subject: IB/mlx5: Put non zero value in max_ah

We put INT_MAX since this is the max value that can be held.
Though there is no hardware limitation, this is practically
a large enough number so we can use it.

Fixes: e126ba97dba9 ('mlx5: Add driver for Mellanox Connect-IB adapters')
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/main.c | 1 +
 1 file changed, 1 insertion(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index 8e0dbd51944e..16ac41d89dfc 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -643,6 +643,7 @@ static int mlx5_ib_query_device(struct ib_device *ibdev,
 	props->max_total_mcast_qp_attach = props->max_mcast_qp_attach *
 					   props->max_mcast_grp;
 	props->max_map_per_fmr = INT_MAX; /* no limit in ConnectIB */
+	props->max_ah = INT_MAX;
 	props->hca_core_clock = MLX5_CAP_GEN(mdev, device_frequency_khz);
 	props->timestamp_mask = 0x7FFFFFFFFFFFFFFFULL;
 
-- 
cgit v1.2.1


From acbda523884dcf45613bf6818d8ead5180df35c2 Mon Sep 17 00:00:00 2001
From: Eli Cohen <eli@mellanox.com>
Date: Thu, 27 Oct 2016 16:36:43 +0300
Subject: IB/mlx5: Wait for all async command completions to complete

Wait before continuing unload till all pending mkey async creation requests
are done.

Fixes: e126ba97dba9 ('mlx5: Add driver for Mellanox Connect-IB adapters')
Signed-off-by: Eli Cohen <eli@mellanox.com>
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/mr.c | 28 ++++++++++++++++++++++++++++
 1 file changed, 28 insertions(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/mr.c b/drivers/infiniband/hw/mlx5/mr.c
index d4ad672b905b..881166a9d868 100644
--- a/drivers/infiniband/hw/mlx5/mr.c
+++ b/drivers/infiniband/hw/mlx5/mr.c
@@ -645,6 +645,33 @@ int mlx5_mr_cache_init(struct mlx5_ib_dev *dev)
 	return 0;
 }
 
+static void wait_for_async_commands(struct mlx5_ib_dev *dev)
+{
+	struct mlx5_mr_cache *cache = &dev->cache;
+	struct mlx5_cache_ent *ent;
+	int total = 0;
+	int i;
+	int j;
+
+	for (i = 0; i < MAX_MR_CACHE_ENTRIES; i++) {
+		ent = &cache->ent[i];
+		for (j = 0 ; j < 1000; j++) {
+			if (!ent->pending)
+				break;
+			msleep(50);
+		}
+	}
+	for (i = 0; i < MAX_MR_CACHE_ENTRIES; i++) {
+		ent = &cache->ent[i];
+		total += ent->pending;
+	}
+
+	if (total)
+		mlx5_ib_warn(dev, "aborted while there are %d pending mr requests\n", total);
+	else
+		mlx5_ib_warn(dev, "done with all pending requests\n");
+}
+
 int mlx5_mr_cache_cleanup(struct mlx5_ib_dev *dev)
 {
 	int i;
@@ -658,6 +685,7 @@ int mlx5_mr_cache_cleanup(struct mlx5_ib_dev *dev)
 		clean_keys(dev, i);
 
 	destroy_workqueue(dev->cache.wq);
+	wait_for_async_commands(dev);
 	del_timer_sync(&dev->delay_timer);
 
 	return 0;
-- 
cgit v1.2.1


From 288c01b746aab484651391ca6d64b585d3eb5ec6 Mon Sep 17 00:00:00 2001
From: Eli Cohen <eli@mellanox.com>
Date: Thu, 27 Oct 2016 16:36:45 +0300
Subject: IB/mlx5: Fix reported max SGE calculation

Add the 512 bytes limit of RDMA READ and the size of remote
address to the max SGE calculation.

Fixes: e126ba97dba9 ('mlx5: Add driver for Mellanox Connect-IB adapters')
Signed-off-by: Eli Cohen <eli@mellanox.com>
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/main.c |  8 +++++---
 drivers/infiniband/hw/mlx5/qp.c   | 29 ++++++++++++++++++++++++++++-
 2 files changed, 33 insertions(+), 4 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index 16ac41d89dfc..46facd30e672 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -496,6 +496,7 @@ static int mlx5_ib_query_device(struct ib_device *ibdev,
 	struct mlx5_ib_dev *dev = to_mdev(ibdev);
 	struct mlx5_core_dev *mdev = dev->mdev;
 	int err = -ENOMEM;
+	int max_sq_desc;
 	int max_rq_sg;
 	int max_sq_sg;
 	u64 min_page_size = 1ull << MLX5_CAP_GEN(mdev, log_pg_sz);
@@ -618,9 +619,10 @@ static int mlx5_ib_query_device(struct ib_device *ibdev,
 	props->max_qp_wr	   = 1 << MLX5_CAP_GEN(mdev, log_max_qp_sz);
 	max_rq_sg =  MLX5_CAP_GEN(mdev, max_wqe_sz_rq) /
 		     sizeof(struct mlx5_wqe_data_seg);
-	max_sq_sg = (MLX5_CAP_GEN(mdev, max_wqe_sz_sq) -
-		     sizeof(struct mlx5_wqe_ctrl_seg)) /
-		     sizeof(struct mlx5_wqe_data_seg);
+	max_sq_desc = min_t(int, MLX5_CAP_GEN(mdev, max_wqe_sz_sq), 512);
+	max_sq_sg = (max_sq_desc - sizeof(struct mlx5_wqe_ctrl_seg) -
+		     sizeof(struct mlx5_wqe_raddr_seg)) /
+		sizeof(struct mlx5_wqe_data_seg);
 	props->max_sge = min(max_rq_sg, max_sq_sg);
 	props->max_sge_rd	   = MLX5_MAX_SGE_RD;
 	props->max_cq		   = 1 << MLX5_CAP_GEN(mdev, log_max_cq);
diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c
index 5bdf20c676fe..c84b07f9f111 100644
--- a/drivers/infiniband/hw/mlx5/qp.c
+++ b/drivers/infiniband/hw/mlx5/qp.c
@@ -352,6 +352,29 @@ static int calc_send_wqe(struct ib_qp_init_attr *attr)
 		return ALIGN(max_t(int, inl_size, size), MLX5_SEND_WQE_BB);
 }
 
+static int get_send_sge(struct ib_qp_init_attr *attr, int wqe_size)
+{
+	int max_sge;
+
+	if (attr->qp_type == IB_QPT_RC)
+		max_sge = (min_t(int, wqe_size, 512) -
+			   sizeof(struct mlx5_wqe_ctrl_seg) -
+			   sizeof(struct mlx5_wqe_raddr_seg)) /
+			sizeof(struct mlx5_wqe_data_seg);
+	else if (attr->qp_type == IB_QPT_XRC_INI)
+		max_sge = (min_t(int, wqe_size, 512) -
+			   sizeof(struct mlx5_wqe_ctrl_seg) -
+			   sizeof(struct mlx5_wqe_xrc_seg) -
+			   sizeof(struct mlx5_wqe_raddr_seg)) /
+			sizeof(struct mlx5_wqe_data_seg);
+	else
+		max_sge = (wqe_size - sq_overhead(attr)) /
+			sizeof(struct mlx5_wqe_data_seg);
+
+	return min_t(int, max_sge, wqe_size - sq_overhead(attr) /
+		     sizeof(struct mlx5_wqe_data_seg));
+}
+
 static int calc_sq_size(struct mlx5_ib_dev *dev, struct ib_qp_init_attr *attr,
 			struct mlx5_ib_qp *qp)
 {
@@ -388,7 +411,11 @@ static int calc_sq_size(struct mlx5_ib_dev *dev, struct ib_qp_init_attr *attr,
 		return -ENOMEM;
 	}
 	qp->sq.wqe_shift = ilog2(MLX5_SEND_WQE_BB);
-	qp->sq.max_gs = attr->cap.max_send_sge;
+	qp->sq.max_gs = get_send_sge(attr, wqe_size);
+	if (qp->sq.max_gs < attr->cap.max_send_sge)
+		return -ENOMEM;
+
+	attr->cap.max_send_sge = qp->sq.max_gs;
 	qp->sq.max_post = wq_size / wqe_size;
 	attr->cap.max_send_wr = qp->sq.max_post;
 
-- 
cgit v1.2.1


From 762f899ae7875554284af92b821be8c083227092 Mon Sep 17 00:00:00 2001
From: Majd Dibbiny <majd@mellanox.com>
Date: Thu, 27 Oct 2016 16:36:47 +0300
Subject: IB/mlx5: Limit mkey page size to 2GB

The maximum page size in the mkey context is 2GB.

Until today, we didn't enforce this requirement in the code,
and therefore, if we got a page size larger than 2GB, we
have passed zeros in the log_page_shift instead of the actual value
and the registration failed.

This patch limits the driver to use compound pages of 2GB for mkeys.

Fixes: e126ba97dba9 ('mlx5: Add driver for Mellanox Connect-IB adapters')
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Majd Dibbiny <majd@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/cq.c      | 4 ++--
 drivers/infiniband/hw/mlx5/mem.c     | 7 ++++++-
 drivers/infiniband/hw/mlx5/mlx5_ib.h | 6 +++++-
 drivers/infiniband/hw/mlx5/mr.c      | 3 ++-
 drivers/infiniband/hw/mlx5/qp.c      | 4 ++--
 drivers/infiniband/hw/mlx5/srq.c     | 2 +-
 6 files changed, 18 insertions(+), 8 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/cq.c b/drivers/infiniband/hw/mlx5/cq.c
index 79d017baf6f4..9e0598b5615f 100644
--- a/drivers/infiniband/hw/mlx5/cq.c
+++ b/drivers/infiniband/hw/mlx5/cq.c
@@ -770,7 +770,7 @@ static int create_cq_user(struct mlx5_ib_dev *dev, struct ib_udata *udata,
 	if (err)
 		goto err_umem;
 
-	mlx5_ib_cont_pages(cq->buf.umem, ucmd.buf_addr, &npages, &page_shift,
+	mlx5_ib_cont_pages(cq->buf.umem, ucmd.buf_addr, 0, &npages, &page_shift,
 			   &ncont, NULL);
 	mlx5_ib_dbg(dev, "addr 0x%llx, size %u, npages %d, page_shift %d, ncont %d\n",
 		    ucmd.buf_addr, entries * ucmd.cqe_size, npages, page_shift, ncont);
@@ -1125,7 +1125,7 @@ static int resize_user(struct mlx5_ib_dev *dev, struct mlx5_ib_cq *cq,
 		return err;
 	}
 
-	mlx5_ib_cont_pages(umem, ucmd.buf_addr, &npages, page_shift,
+	mlx5_ib_cont_pages(umem, ucmd.buf_addr, 0, &npages, page_shift,
 			   npas, NULL);
 
 	cq->resize_umem = umem;
diff --git a/drivers/infiniband/hw/mlx5/mem.c b/drivers/infiniband/hw/mlx5/mem.c
index 996b54e366b0..6851357c16f4 100644
--- a/drivers/infiniband/hw/mlx5/mem.c
+++ b/drivers/infiniband/hw/mlx5/mem.c
@@ -37,12 +37,15 @@
 
 /* @umem: umem object to scan
  * @addr: ib virtual address requested by the user
+ * @max_page_shift: high limit for page_shift - 0 means no limit
  * @count: number of PAGE_SIZE pages covered by umem
  * @shift: page shift for the compound pages found in the region
  * @ncont: number of compund pages
  * @order: log2 of the number of compound pages
  */
-void mlx5_ib_cont_pages(struct ib_umem *umem, u64 addr, int *count, int *shift,
+void mlx5_ib_cont_pages(struct ib_umem *umem, u64 addr,
+			unsigned long max_page_shift,
+			int *count, int *shift,
 			int *ncont, int *order)
 {
 	unsigned long tmp;
@@ -72,6 +75,8 @@ void mlx5_ib_cont_pages(struct ib_umem *umem, u64 addr, int *count, int *shift,
 	addr = addr >> page_shift;
 	tmp = (unsigned long)addr;
 	m = find_first_bit(&tmp, BITS_PER_LONG);
+	if (max_page_shift)
+		m = min_t(unsigned long, max_page_shift - page_shift, m);
 	skip = 1 << m;
 	mask = skip - 1;
 	i = 0;
diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h
index d5d007740159..95937e7d2584 100644
--- a/drivers/infiniband/hw/mlx5/mlx5_ib.h
+++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h
@@ -63,6 +63,8 @@ pr_warn("%s:%s:%d:(pid %d): " format, (dev)->ib_dev.name, __func__,	\
 #define MLX5_IB_DEFAULT_UIDX 0xffffff
 #define MLX5_USER_ASSIGNED_UIDX_MASK __mlx5_mask(qpc, user_index)
 
+#define MLX5_MKEY_PAGE_SHIFT_MASK __mlx5_mask(mkc, log_page_size)
+
 enum {
 	MLX5_IB_MMAP_CMD_SHIFT	= 8,
 	MLX5_IB_MMAP_CMD_MASK	= 0xff,
@@ -823,7 +825,9 @@ int mlx5_ib_query_port(struct ib_device *ibdev, u8 port,
 		       struct ib_port_attr *props);
 int mlx5_ib_init_fmr(struct mlx5_ib_dev *dev);
 void mlx5_ib_cleanup_fmr(struct mlx5_ib_dev *dev);
-void mlx5_ib_cont_pages(struct ib_umem *umem, u64 addr, int *count, int *shift,
+void mlx5_ib_cont_pages(struct ib_umem *umem, u64 addr,
+			unsigned long max_page_shift,
+			int *count, int *shift,
 			int *ncont, int *order);
 void __mlx5_ib_populate_pas(struct mlx5_ib_dev *dev, struct ib_umem *umem,
 			    int page_shift, size_t offset, size_t num_pages,
diff --git a/drivers/infiniband/hw/mlx5/mr.c b/drivers/infiniband/hw/mlx5/mr.c
index 881166a9d868..6cbda901f259 100644
--- a/drivers/infiniband/hw/mlx5/mr.c
+++ b/drivers/infiniband/hw/mlx5/mr.c
@@ -855,7 +855,8 @@ static struct ib_umem *mr_umem_get(struct ib_pd *pd, u64 start, u64 length,
 		return (void *)umem;
 	}
 
-	mlx5_ib_cont_pages(umem, start, npages, page_shift, ncont, order);
+	mlx5_ib_cont_pages(umem, start, MLX5_MKEY_PAGE_SHIFT_MASK, npages,
+			   page_shift, ncont, order);
 	if (!*npages) {
 		mlx5_ib_warn(dev, "avoid zero region\n");
 		ib_umem_release(umem);
diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c
index c84b07f9f111..aa27688f5ae9 100644
--- a/drivers/infiniband/hw/mlx5/qp.c
+++ b/drivers/infiniband/hw/mlx5/qp.c
@@ -675,7 +675,7 @@ static int mlx5_ib_umem_get(struct mlx5_ib_dev *dev,
 		return PTR_ERR(*umem);
 	}
 
-	mlx5_ib_cont_pages(*umem, addr, npages, page_shift, ncont, NULL);
+	mlx5_ib_cont_pages(*umem, addr, 0, npages, page_shift, ncont, NULL);
 
 	err = mlx5_ib_get_buf_offset(addr, *page_shift, offset);
 	if (err) {
@@ -728,7 +728,7 @@ static int create_user_rq(struct mlx5_ib_dev *dev, struct ib_pd *pd,
 		return err;
 	}
 
-	mlx5_ib_cont_pages(rwq->umem, ucmd->buf_addr, &npages, &page_shift,
+	mlx5_ib_cont_pages(rwq->umem, ucmd->buf_addr, 0, &npages, &page_shift,
 			   &ncont, NULL);
 	err = mlx5_ib_get_buf_offset(ucmd->buf_addr, page_shift,
 				     &rwq->rq_page_offset);
diff --git a/drivers/infiniband/hw/mlx5/srq.c b/drivers/infiniband/hw/mlx5/srq.c
index 3857dbd9c956..f384db5367fb 100644
--- a/drivers/infiniband/hw/mlx5/srq.c
+++ b/drivers/infiniband/hw/mlx5/srq.c
@@ -118,7 +118,7 @@ static int create_srq_user(struct ib_pd *pd, struct mlx5_ib_srq *srq,
 		return err;
 	}
 
-	mlx5_ib_cont_pages(srq->umem, ucmd.buf_addr, &npages,
+	mlx5_ib_cont_pages(srq->umem, ucmd.buf_addr, 0, &npages,
 			   &page_shift, &ncont, NULL);
 	err = mlx5_ib_get_buf_offset(ucmd.buf_addr, page_shift,
 				     &offset);
-- 
cgit v1.2.1


From 0b59970e7d96edcb3c7f651d9d48e1a59af3c3b0 Mon Sep 17 00:00:00 2001
From: Kamal Heib <kamalh@mellanox.com>
Date: Thu, 10 Nov 2016 10:16:48 +0200
Subject: IB/IPoIB: Remove can't use GFP_NOIO warning

Remove the warning print of "can't use of GFP_NOIO" to avoid prints in
each QP creation when devices aren't supporting IB_QP_CREATE_USE_GFP_NOIO.

This print become more annoying when the IPoIB interface is configured
to work in connected mode.

Fixes: 09b93088d750 ('IB: Add a QP creation flag to use GFP_NOIO allocations')
Signed-off-by: Kamal Heib <kamalh@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Reviewed-by: Yuval Shaia <yuval.shaia@oracle.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/ulp/ipoib/ipoib_cm.c | 2 --
 1 file changed, 2 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/ulp/ipoib/ipoib_cm.c b/drivers/infiniband/ulp/ipoib/ipoib_cm.c
index 4ad297d3de89..50c9772bde32 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_cm.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_cm.c
@@ -1053,8 +1053,6 @@ static struct ib_qp *ipoib_cm_create_tx_qp(struct net_device *dev, struct ipoib_
 
 	tx_qp = ib_create_qp(priv->pd, &attr);
 	if (PTR_ERR(tx_qp) == -EINVAL) {
-		ipoib_warn(priv, "can't use GFP_NOIO for QPs on device %s, using GFP_KERNEL\n",
-			   priv->ca->name);
 		attr.create_flags &= ~IB_QP_CREATE_USE_GFP_NOIO;
 		tx_qp = ib_create_qp(priv->pd, &attr);
 	}
-- 
cgit v1.2.1


From af4295c117b82a521b05d0daf39ce879d26e6cb1 Mon Sep 17 00:00:00 2001
From: Maor Gottlieb <maorg@mellanox.com>
Date: Thu, 10 Nov 2016 11:30:53 +0200
Subject: IB/mlx4: Set traffic class in AH

Set traffic class within sl_tclass_flowlabel when create iboe AH.
Without this the TOS value will be empty when running VLAN tagged
traffic, because the TOS value is taken from the traffic class in the
address handle attributes.

Fixes: 9106c4106974 ('IB/mlx4: Fix SL to 802.1Q priority-bits mapping for IBoE')
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/ah.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx4/ah.c b/drivers/infiniband/hw/mlx4/ah.c
index 5fc623362731..6be7dc320ff7 100644
--- a/drivers/infiniband/hw/mlx4/ah.c
+++ b/drivers/infiniband/hw/mlx4/ah.c
@@ -111,7 +111,9 @@ static struct ib_ah *create_iboe_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr
 		       !(1 << ah->av.eth.stat_rate & dev->caps.stat_rate_support))
 			--ah->av.eth.stat_rate;
 	}
-
+	ah->av.eth.sl_tclass_flowlabel |=
+			cpu_to_be32((ah_attr->grh.traffic_class << 20) |
+				    ah_attr->grh.flow_label);
 	/*
 	 * HW requires multicast LID so we just choose one.
 	 */
@@ -119,7 +121,7 @@ static struct ib_ah *create_iboe_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr
 		ah->av.ib.dlid = cpu_to_be16(0xc000);
 
 	memcpy(ah->av.eth.dgid, ah_attr->grh.dgid.raw, 16);
-	ah->av.eth.sl_tclass_flowlabel = cpu_to_be32(ah_attr->sl << 29);
+	ah->av.eth.sl_tclass_flowlabel |= cpu_to_be32(ah_attr->sl << 29);
 
 	return &ah->ibah;
 }
-- 
cgit v1.2.1


From 850d8fd765079d223d47de89f1f03ab95d1036cb Mon Sep 17 00:00:00 2001
From: Moni Shoua <monis@mellanox.com>
Date: Thu, 10 Nov 2016 11:30:56 +0200
Subject: IB/mlx4: Handle IPv4 header when demultiplexing MAD

When MAD arrives to the hypervisor, we need to identify which slave it
should be sent by destination GID. When L3 protocol is IPv4 the
GRH is replaced by an IPv4 header. This patch detects when IPv4 header
needs to be parsed instead of GRH.

Fixes: b6ffaeffaea4 ('mlx4: In RoCE allow guests to have multiple GIDS')
Signed-off-by: Moni Shoua <monis@mellanox.com>
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/verbs.c  | 16 +++++++++-------
 drivers/infiniband/hw/mlx4/mad.c | 33 ++++++++++++++++++++++++++++++---
 2 files changed, 39 insertions(+), 10 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c
index b7f9f3d95975..98e10c5f2a96 100644
--- a/drivers/infiniband/core/verbs.c
+++ b/drivers/infiniband/core/verbs.c
@@ -328,7 +328,7 @@ struct ib_ah *ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr)
 }
 EXPORT_SYMBOL(ib_create_ah);
 
-static int ib_get_header_version(const union rdma_network_hdr *hdr)
+int ib_get_rdma_header_version(const union rdma_network_hdr *hdr)
 {
 	const struct iphdr *ip4h = (struct iphdr *)&hdr->roce4grh;
 	struct iphdr ip4h_checked;
@@ -359,6 +359,7 @@ static int ib_get_header_version(const union rdma_network_hdr *hdr)
 		return 4;
 	return 6;
 }
+EXPORT_SYMBOL(ib_get_rdma_header_version);
 
 static enum rdma_network_type ib_get_net_type_by_grh(struct ib_device *device,
 						     u8 port_num,
@@ -369,7 +370,7 @@ static enum rdma_network_type ib_get_net_type_by_grh(struct ib_device *device,
 	if (rdma_protocol_ib(device, port_num))
 		return RDMA_NETWORK_IB;
 
-	grh_version = ib_get_header_version((union rdma_network_hdr *)grh);
+	grh_version = ib_get_rdma_header_version((union rdma_network_hdr *)grh);
 
 	if (grh_version == 4)
 		return RDMA_NETWORK_IPV4;
@@ -415,9 +416,9 @@ static int get_sgid_index_from_eth(struct ib_device *device, u8 port_num,
 				     &context, gid_index);
 }
 
-static int get_gids_from_rdma_hdr(union rdma_network_hdr *hdr,
-				  enum rdma_network_type net_type,
-				  union ib_gid *sgid, union ib_gid *dgid)
+int ib_get_gids_from_rdma_hdr(const union rdma_network_hdr *hdr,
+			      enum rdma_network_type net_type,
+			      union ib_gid *sgid, union ib_gid *dgid)
 {
 	struct sockaddr_in  src_in;
 	struct sockaddr_in  dst_in;
@@ -447,6 +448,7 @@ static int get_gids_from_rdma_hdr(union rdma_network_hdr *hdr,
 		return -EINVAL;
 	}
 }
+EXPORT_SYMBOL(ib_get_gids_from_rdma_hdr);
 
 int ib_init_ah_from_wc(struct ib_device *device, u8 port_num,
 		       const struct ib_wc *wc, const struct ib_grh *grh,
@@ -469,8 +471,8 @@ int ib_init_ah_from_wc(struct ib_device *device, u8 port_num,
 			net_type = ib_get_net_type_by_grh(device, port_num, grh);
 		gid_type = ib_network_to_gid_type(net_type);
 	}
-	ret = get_gids_from_rdma_hdr((union rdma_network_hdr *)grh, net_type,
-				     &sgid, &dgid);
+	ret = ib_get_gids_from_rdma_hdr((union rdma_network_hdr *)grh, net_type,
+					&sgid, &dgid);
 	if (ret)
 		return ret;
 
diff --git a/drivers/infiniband/hw/mlx4/mad.c b/drivers/infiniband/hw/mlx4/mad.c
index 1672907ff219..b8e90130db37 100644
--- a/drivers/infiniband/hw/mlx4/mad.c
+++ b/drivers/infiniband/hw/mlx4/mad.c
@@ -39,6 +39,8 @@
 #include <linux/mlx4/cmd.h>
 #include <linux/gfp.h>
 #include <rdma/ib_pma.h>
+#include <linux/ip.h>
+#include <net/ipv6.h>
 
 #include <linux/mlx4/driver.h>
 #include "mlx4_ib.h"
@@ -480,6 +482,23 @@ static int find_slave_port_pkey_ix(struct mlx4_ib_dev *dev, int slave,
 	return -EINVAL;
 }
 
+static int get_gids_from_l3_hdr(struct ib_grh *grh, union ib_gid *sgid,
+				union ib_gid *dgid)
+{
+	int version = ib_get_rdma_header_version((const union rdma_network_hdr *)grh);
+	enum rdma_network_type net_type;
+
+	if (version == 4)
+		net_type = RDMA_NETWORK_IPV4;
+	else if (version == 6)
+		net_type = RDMA_NETWORK_IPV6;
+	else
+		return -EINVAL;
+
+	return ib_get_gids_from_rdma_hdr((union rdma_network_hdr *)grh, net_type,
+					 sgid, dgid);
+}
+
 int mlx4_ib_send_to_slave(struct mlx4_ib_dev *dev, int slave, u8 port,
 			  enum ib_qp_type dest_qpt, struct ib_wc *wc,
 			  struct ib_grh *grh, struct ib_mad *mad)
@@ -538,7 +557,10 @@ int mlx4_ib_send_to_slave(struct mlx4_ib_dev *dev, int slave, u8 port,
 	memset(&attr, 0, sizeof attr);
 	attr.port_num = port;
 	if (is_eth) {
-		memcpy(&attr.grh.dgid.raw[0], &grh->dgid.raw[0], 16);
+		union ib_gid sgid;
+
+		if (get_gids_from_l3_hdr(grh, &sgid, &attr.grh.dgid))
+			return -EINVAL;
 		attr.ah_flags = IB_AH_GRH;
 	}
 	ah = ib_create_ah(tun_ctx->pd, &attr);
@@ -651,6 +673,11 @@ static int mlx4_ib_demux_mad(struct ib_device *ibdev, u8 port,
 		is_eth = 1;
 
 	if (is_eth) {
+		union ib_gid dgid;
+		union ib_gid sgid;
+
+		if (get_gids_from_l3_hdr(grh, &sgid, &dgid))
+			return -EINVAL;
 		if (!(wc->wc_flags & IB_WC_GRH)) {
 			mlx4_ib_warn(ibdev, "RoCE grh not present.\n");
 			return -EINVAL;
@@ -659,10 +686,10 @@ static int mlx4_ib_demux_mad(struct ib_device *ibdev, u8 port,
 			mlx4_ib_warn(ibdev, "RoCE mgmt class is not CM\n");
 			return -EINVAL;
 		}
-		err = mlx4_get_slave_from_roce_gid(dev->dev, port, grh->dgid.raw, &slave);
+		err = mlx4_get_slave_from_roce_gid(dev->dev, port, dgid.raw, &slave);
 		if (err && mlx4_is_mf_bonded(dev->dev)) {
 			other_port = (port == 1) ? 2 : 1;
-			err = mlx4_get_slave_from_roce_gid(dev->dev, other_port, grh->dgid.raw, &slave);
+			err = mlx4_get_slave_from_roce_gid(dev->dev, other_port, dgid.raw, &slave);
 			if (!err) {
 				port = other_port;
 				pr_debug("resolved slave %d from gid %pI6 wire port %d other %d\n",
-- 
cgit v1.2.1


From befcabcd530e4ffb6f016638f693b7d94986d2ba Mon Sep 17 00:00:00 2001
From: Jack Morgenstein <jackm@dev.mellanox.co.il>
Date: Thu, 10 Nov 2016 11:30:57 +0200
Subject: IB/mlx4: Handle well-known-gid in mad_demux processing

If OpenSM runs over a ConnectX-3, and there are ConnectX-4 or Connect-IB
VFs active on the network, the OpenSM will receive QP1 packets containing
a GRH where the destination GID is the "Well-Known GID" -- which is not a
GID in the HCA Port's GID Table.

This GID must be tested-for separately -- and packets which contain
this destination GID should be routed to slave 0 (the PF).

Fixes: 37bfc7c1e83f ('IB/mlx4: SR-IOV multiplex and demultiplex MADs')
Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/mad.c | 16 ++++++++++++----
 1 file changed, 12 insertions(+), 4 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx4/mad.c b/drivers/infiniband/hw/mlx4/mad.c
index b8e90130db37..7bcae4115408 100644
--- a/drivers/infiniband/hw/mlx4/mad.c
+++ b/drivers/infiniband/hw/mlx4/mad.c
@@ -729,10 +729,18 @@ static int mlx4_ib_demux_mad(struct ib_device *ibdev, u8 port,
 
 	/* If a grh is present, we demux according to it */
 	if (wc->wc_flags & IB_WC_GRH) {
-		slave = mlx4_ib_find_real_gid(ibdev, port, grh->dgid.global.interface_id);
-		if (slave < 0) {
-			mlx4_ib_warn(ibdev, "failed matching grh\n");
-			return -ENOENT;
+		if (grh->dgid.global.interface_id ==
+			cpu_to_be64(IB_SA_WELL_KNOWN_GUID) &&
+		    grh->dgid.global.subnet_prefix == cpu_to_be64(
+			atomic64_read(&dev->sriov.demux[port - 1].subnet_prefix))) {
+			slave = 0;
+		} else {
+			slave = mlx4_ib_find_real_gid(ibdev, port,
+						      grh->dgid.global.interface_id);
+			if (slave < 0) {
+				mlx4_ib_warn(ibdev, "failed matching grh\n");
+				return -ENOENT;
+			}
 		}
 	}
 	/* Class-specific handling */
-- 
cgit v1.2.1


From 731e0415b4af3a133d0316e4dc8ef0ea57dc3fdf Mon Sep 17 00:00:00 2001
From: Maor Gottlieb <maorg@mellanox.com>
Date: Thu, 10 Nov 2016 11:30:58 +0200
Subject: IB/mlx4: Put non zero value in max_ah device attribute

Use INT_MAX since this is the max value the attribute can hold, though
hardware capability is unlimited.

Fixes: 225c7b1feef1 ('IB/mlx4: Add a driver Mellanox ConnectX InfiniBand adapters')
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Reviewed-by: Yuval Shaia <yuval.shaia@oracle.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/main.c | 1 +
 1 file changed, 1 insertion(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index b597e8227591..05ab3cb34cb3 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -547,6 +547,7 @@ static int mlx4_ib_query_device(struct ib_device *ibdev,
 	props->max_map_per_fmr = dev->dev->caps.max_fmr_maps;
 	props->hca_core_clock = dev->dev->caps.hca_core_clock * 1000UL;
 	props->timestamp_mask = 0xFFFFFFFFFFFFULL;
+	props->max_ah = INT_MAX;
 
 	if (!mlx4_is_slave(dev->dev))
 		err = mlx4_get_internal_clock_params(dev->dev, &clock_params);
-- 
cgit v1.2.1


From 6fa26208206c406fa529cd73f7ae6bf4181e270b Mon Sep 17 00:00:00 2001
From: Saeed Mahameed <saeedm@mellanox.com>
Date: Thu, 10 Nov 2016 11:30:59 +0200
Subject: IB/mlx4: Fix port query for 56Gb Ethernet links

Report the correct speed in the port attributes when using a 56Gbps
ethernet link.  Without this change the field is incorrectly set to 10.

Fixes: a9c766bb75ee ('IB/mlx4: Fix info returned when querying IBoE ports')
Fixes: 2e96691c31ec ('IB: Use central enum for speed instead of hard-coded values')
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/main.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index 05ab3cb34cb3..4054a1bfabb5 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -698,9 +698,11 @@ static int eth_link_query_port(struct ib_device *ibdev, u8 port,
 	if (err)
 		goto out;
 
-	props->active_width	=  (((u8 *)mailbox->buf)[5] == 0x40) ?
-						IB_WIDTH_4X : IB_WIDTH_1X;
-	props->active_speed	= IB_SPEED_QDR;
+	props->active_width	=  (((u8 *)mailbox->buf)[5] == 0x40) ||
+				   (((u8 *)mailbox->buf)[5] == 0x20 /*56Gb*/) ?
+					   IB_WIDTH_4X : IB_WIDTH_1X;
+	props->active_speed	=  (((u8 *)mailbox->buf)[5] == 0x20 /*56Gb*/) ?
+					   IB_SPEED_FDR : IB_SPEED_QDR;
 	props->port_cap_flags	= IB_PORT_CM_SUP | IB_PORT_IP_BASED_GIDS;
 	props->gid_tbl_len	= mdev->dev->caps.gid_table_len[port];
 	props->max_msg_sz	= mdev->dev->caps.max_msg_sz;
-- 
cgit v1.2.1


From 2716243212241855cd9070883779f6e58967dec5 Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leon@kernel.org>
Date: Thu, 3 Nov 2016 16:44:09 +0200
Subject: IB/mad: Remove debug prints after allocation failure

The prints after [k|v][m|z|c]alloc() functions are not needed,
because in case of failure, allocator will print their internal
error prints anyway.

Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/mad.c | 40 ++++++----------------------------------
 1 file changed, 6 insertions(+), 34 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/core/mad.c b/drivers/infiniband/core/mad.c
index 40cbd6bdb73b..dadce12e9056 100644
--- a/drivers/infiniband/core/mad.c
+++ b/drivers/infiniband/core/mad.c
@@ -816,7 +816,6 @@ static int handle_outgoing_dr_smp(struct ib_mad_agent_private *mad_agent_priv,
 	local = kmalloc(sizeof *local, GFP_ATOMIC);
 	if (!local) {
 		ret = -ENOMEM;
-		dev_err(&device->dev, "No memory for ib_mad_local_private\n");
 		goto out;
 	}
 	local->mad_priv = NULL;
@@ -824,7 +823,6 @@ static int handle_outgoing_dr_smp(struct ib_mad_agent_private *mad_agent_priv,
 	mad_priv = alloc_mad_private(mad_size, GFP_ATOMIC);
 	if (!mad_priv) {
 		ret = -ENOMEM;
-		dev_err(&device->dev, "No memory for local response MAD\n");
 		kfree(local);
 		goto out;
 	}
@@ -947,9 +945,6 @@ static int alloc_send_rmpp_list(struct ib_mad_send_wr_private *send_wr,
 	for (left = send_buf->data_len + pad; left > 0; left -= seg_size) {
 		seg = kmalloc(sizeof (*seg) + seg_size, gfp_mask);
 		if (!seg) {
-			dev_err(&send_buf->mad_agent->device->dev,
-				"alloc_send_rmpp_segs: RMPP mem alloc failed for len %zd, gfp %#x\n",
-				sizeof (*seg) + seg_size, gfp_mask);
 			free_send_rmpp_list(send_wr);
 			return -ENOMEM;
 		}
@@ -1362,12 +1357,7 @@ static int allocate_method_table(struct ib_mad_mgmt_method_table **method)
 {
 	/* Allocate management method table */
 	*method = kzalloc(sizeof **method, GFP_ATOMIC);
-	if (!*method) {
-		pr_err("No memory for ib_mad_mgmt_method_table\n");
-		return -ENOMEM;
-	}
-
-	return 0;
+	return (*method) ? 0 : (-ENOMEM);
 }
 
 /*
@@ -1458,8 +1448,6 @@ static int add_nonoui_reg_req(struct ib_mad_reg_req *mad_reg_req,
 		/* Allocate management class table for "new" class version */
 		*class = kzalloc(sizeof **class, GFP_ATOMIC);
 		if (!*class) {
-			dev_err(&agent_priv->agent.device->dev,
-				"No memory for ib_mad_mgmt_class_table\n");
 			ret = -ENOMEM;
 			goto error1;
 		}
@@ -1524,22 +1512,16 @@ static int add_oui_reg_req(struct ib_mad_reg_req *mad_reg_req,
 	if (!*vendor_table) {
 		/* Allocate mgmt vendor class table for "new" class version */
 		vendor = kzalloc(sizeof *vendor, GFP_ATOMIC);
-		if (!vendor) {
-			dev_err(&agent_priv->agent.device->dev,
-				"No memory for ib_mad_mgmt_vendor_class_table\n");
+		if (!vendor)
 			goto error1;
-		}
 
 		*vendor_table = vendor;
 	}
 	if (!(*vendor_table)->vendor_class[vclass]) {
 		/* Allocate table for this management vendor class */
 		vendor_class = kzalloc(sizeof *vendor_class, GFP_ATOMIC);
-		if (!vendor_class) {
-			dev_err(&agent_priv->agent.device->dev,
-				"No memory for ib_mad_mgmt_vendor_class\n");
+		if (!vendor_class)
 			goto error2;
-		}
 
 		(*vendor_table)->vendor_class[vclass] = vendor_class;
 	}
@@ -2238,11 +2220,8 @@ static void ib_mad_recv_done(struct ib_cq *cq, struct ib_wc *wc)
 
 	mad_size = recv->mad_size;
 	response = alloc_mad_private(mad_size, GFP_KERNEL);
-	if (!response) {
-		dev_err(&port_priv->device->dev,
-			"%s: no memory for response buffer\n", __func__);
+	if (!response)
 		goto out;
-	}
 
 	if (rdma_cap_ib_switch(port_priv->device))
 		port_num = wc->port_num;
@@ -2869,8 +2848,6 @@ static int ib_mad_post_receive_mads(struct ib_mad_qp_info *qp_info,
 			mad_priv = alloc_mad_private(port_mad_size(qp_info->port_priv),
 						     GFP_ATOMIC);
 			if (!mad_priv) {
-				dev_err(&qp_info->port_priv->device->dev,
-					"No memory for receive buffer\n");
 				ret = -ENOMEM;
 				break;
 			}
@@ -2961,11 +2938,8 @@ static int ib_mad_port_start(struct ib_mad_port_private *port_priv)
 	u16 pkey_index;
 
 	attr = kmalloc(sizeof *attr, GFP_KERNEL);
-	if (!attr) {
-		dev_err(&port_priv->device->dev,
-			"Couldn't kmalloc ib_qp_attr\n");
+	if (!attr)
 		return -ENOMEM;
-	}
 
 	ret = ib_find_pkey(port_priv->device, port_priv->port_num,
 			   IB_DEFAULT_PKEY_FULL, &pkey_index);
@@ -3135,10 +3109,8 @@ static int ib_mad_port_open(struct ib_device *device,
 
 	/* Create new device info */
 	port_priv = kzalloc(sizeof *port_priv, GFP_KERNEL);
-	if (!port_priv) {
-		dev_err(&device->dev, "No memory for ib_mad_port_private\n");
+	if (!port_priv)
 		return -ENOMEM;
-	}
 
 	port_priv->device = device;
 	port_priv->port_num = port_num;
-- 
cgit v1.2.1


From a0b3455fcb2d1b3e486349a4a51803c3cb3847b5 Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leon@kernel.org>
Date: Thu, 3 Nov 2016 16:44:10 +0200
Subject: IB/core: Remove debug prints after allocation failure

The prints after [k|v][m|z|c]alloc() functions are not needed,
because in case of failure, allocator will print their internal
error prints anyway.

Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/agent.c         |  1 -
 drivers/infiniband/core/device.c        |  5 +----
 drivers/infiniband/core/fmr_pool.c      |  1 -
 drivers/infiniband/core/iwpm_msg.c      |  1 -
 drivers/infiniband/core/iwpm_util.c     | 12 ++++--------
 drivers/infiniband/core/roce_gid_mgmt.c | 21 ++++++---------------
 6 files changed, 11 insertions(+), 30 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/core/agent.c b/drivers/infiniband/core/agent.c
index 4fa524dfb6cf..11dacd97a667 100644
--- a/drivers/infiniband/core/agent.c
+++ b/drivers/infiniband/core/agent.c
@@ -156,7 +156,6 @@ int ib_agent_port_open(struct ib_device *device, int port_num)
 	/* Create new device info */
 	port_priv = kzalloc(sizeof *port_priv, GFP_KERNEL);
 	if (!port_priv) {
-		dev_err(&device->dev, "No memory for ib_agent_port_private\n");
 		ret = -ENOMEM;
 		goto error1;
 	}
diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c
index 760ef603a468..571974cd3919 100644
--- a/drivers/infiniband/core/device.c
+++ b/drivers/infiniband/core/device.c
@@ -254,11 +254,8 @@ static int add_client_context(struct ib_device *device, struct ib_client *client
 	unsigned long flags;
 
 	context = kmalloc(sizeof *context, GFP_KERNEL);
-	if (!context) {
-		pr_warn("Couldn't allocate client context for %s/%s\n",
-			device->name, client->name);
+	if (!context)
 		return -ENOMEM;
-	}
 
 	context->client = client;
 	context->data   = NULL;
diff --git a/drivers/infiniband/core/fmr_pool.c b/drivers/infiniband/core/fmr_pool.c
index cdbb1f1a6d97..cdfad5f26212 100644
--- a/drivers/infiniband/core/fmr_pool.c
+++ b/drivers/infiniband/core/fmr_pool.c
@@ -247,7 +247,6 @@ struct ib_fmr_pool *ib_create_fmr_pool(struct ib_pd             *pd,
 			kmalloc(IB_FMR_HASH_SIZE * sizeof *pool->cache_bucket,
 				GFP_KERNEL);
 		if (!pool->cache_bucket) {
-			pr_warn(PFX "Failed to allocate cache in pool\n");
 			ret = -ENOMEM;
 			goto out_free_pool;
 		}
diff --git a/drivers/infiniband/core/iwpm_msg.c b/drivers/infiniband/core/iwpm_msg.c
index 1c41b95cefec..a0e7c16d8bd8 100644
--- a/drivers/infiniband/core/iwpm_msg.c
+++ b/drivers/infiniband/core/iwpm_msg.c
@@ -604,7 +604,6 @@ int iwpm_remote_info_cb(struct sk_buff *skb, struct netlink_callback *cb)
 	}
 	rem_info = kzalloc(sizeof(struct iwpm_remote_info), GFP_ATOMIC);
 	if (!rem_info) {
-		pr_err("%s: Unable to allocate a remote info\n", __func__);
 		ret = -ENOMEM;
 		return ret;
 	}
diff --git a/drivers/infiniband/core/iwpm_util.c b/drivers/infiniband/core/iwpm_util.c
index ade71e7f0131..3ef51a96bbf1 100644
--- a/drivers/infiniband/core/iwpm_util.c
+++ b/drivers/infiniband/core/iwpm_util.c
@@ -62,7 +62,6 @@ int iwpm_init(u8 nl_client)
 					sizeof(struct hlist_head), GFP_KERNEL);
 		if (!iwpm_hash_bucket) {
 			ret = -ENOMEM;
-			pr_err("%s Unable to create mapinfo hash table\n", __func__);
 			goto init_exit;
 		}
 		iwpm_reminfo_bucket = kzalloc(IWPM_REMINFO_HASH_SIZE *
@@ -70,7 +69,6 @@ int iwpm_init(u8 nl_client)
 		if (!iwpm_reminfo_bucket) {
 			kfree(iwpm_hash_bucket);
 			ret = -ENOMEM;
-			pr_err("%s Unable to create reminfo hash table\n", __func__);
 			goto init_exit;
 		}
 	}
@@ -128,10 +126,9 @@ int iwpm_create_mapinfo(struct sockaddr_storage *local_sockaddr,
 	if (!iwpm_valid_client(nl_client))
 		return ret;
 	map_info = kzalloc(sizeof(struct iwpm_mapping_info), GFP_KERNEL);
-	if (!map_info) {
-		pr_err("%s: Unable to allocate a mapping info\n", __func__);
+	if (!map_info)
 		return -ENOMEM;
-	}
+
 	memcpy(&map_info->local_sockaddr, local_sockaddr,
 	       sizeof(struct sockaddr_storage));
 	memcpy(&map_info->mapped_sockaddr, mapped_sockaddr,
@@ -309,10 +306,9 @@ struct iwpm_nlmsg_request *iwpm_get_nlmsg_request(__u32 nlmsg_seq,
 	unsigned long flags;
 
 	nlmsg_request = kzalloc(sizeof(struct iwpm_nlmsg_request), gfp);
-	if (!nlmsg_request) {
-		pr_err("%s Unable to allocate a nlmsg_request\n", __func__);
+	if (!nlmsg_request)
 		return NULL;
-	}
+
 	spin_lock_irqsave(&iwpm_nlmsg_req_lock, flags);
 	list_add_tail(&nlmsg_request->inprocess_list, &iwpm_nlmsg_req_list);
 	spin_unlock_irqrestore(&iwpm_nlmsg_req_lock, flags);
diff --git a/drivers/infiniband/core/roce_gid_mgmt.c b/drivers/infiniband/core/roce_gid_mgmt.c
index 06556c34606d..c86ddcea7675 100644
--- a/drivers/infiniband/core/roce_gid_mgmt.c
+++ b/drivers/infiniband/core/roce_gid_mgmt.c
@@ -304,10 +304,9 @@ static void enum_netdev_ipv4_ips(struct ib_device *ib_dev,
 	for_ifa(in_dev) {
 		struct sin_list *entry = kzalloc(sizeof(*entry), GFP_ATOMIC);
 
-		if (!entry) {
-			pr_warn("roce_gid_mgmt: couldn't allocate entry for IPv4 update\n");
+		if (!entry)
 			continue;
-		}
+
 		entry->ip.sin_family = AF_INET;
 		entry->ip.sin_addr.s_addr = ifa->ifa_address;
 		list_add_tail(&entry->list, &sin_list);
@@ -348,10 +347,8 @@ static void enum_netdev_ipv6_ips(struct ib_device *ib_dev,
 	list_for_each_entry(ifp, &in6_dev->addr_list, if_list) {
 		struct sin6_list *entry = kzalloc(sizeof(*entry), GFP_ATOMIC);
 
-		if (!entry) {
-			pr_warn("roce_gid_mgmt: couldn't allocate entry for IPv6 update\n");
+		if (!entry)
 			continue;
-		}
 
 		entry->sin6.sin6_family = AF_INET6;
 		entry->sin6.sin6_addr = ifp->addr;
@@ -459,10 +456,8 @@ static void handle_netdev_upper(struct ib_device *ib_dev, u8 port,
 		struct upper_list *entry = kmalloc(sizeof(*entry),
 						   GFP_ATOMIC);
 
-		if (!entry) {
-			pr_info("roce_gid_mgmt: couldn't allocate entry to delete ndev\n");
+		if (!entry)
 			continue;
-		}
 
 		list_add_tail(&entry->list, &upper_list);
 		dev_hold(upper);
@@ -555,10 +550,8 @@ static int netdevice_queue_work(struct netdev_event_work_cmd *cmds,
 	struct netdev_event_work *ndev_work =
 		kmalloc(sizeof(*ndev_work), GFP_KERNEL);
 
-	if (!ndev_work) {
-		pr_warn("roce_gid_mgmt: can't allocate work for netdevice_event\n");
+	if (!ndev_work)
 		return NOTIFY_DONE;
-	}
 
 	memcpy(ndev_work->cmds, cmds, sizeof(ndev_work->cmds));
 	for (i = 0; i < ARRAY_SIZE(ndev_work->cmds) && ndev_work->cmds[i].cb; i++) {
@@ -692,10 +685,8 @@ static int addr_event(struct notifier_block *this, unsigned long event,
 	}
 
 	work = kmalloc(sizeof(*work), GFP_ATOMIC);
-	if (!work) {
-		pr_warn("roce_gid_mgmt: Couldn't allocate work for addr_event\n");
+	if (!work)
 		return NOTIFY_DONE;
-	}
 
 	INIT_WORK(&work->work, update_gid_event_work_handler);
 
-- 
cgit v1.2.1


From aa6aae38f7fb2c030f326a6dd10b58fff1851dfa Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leon@kernel.org>
Date: Thu, 3 Nov 2016 16:44:11 +0200
Subject: IB/core: Release allocated memory in cache setup failure

The failure in ib_cache_setup_one function during
ib_register_device will leave leaked allocated memory.

Fixes: 03db3a2d81e6 ("IB/core: Add RoCE GID table management")
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/cache.c | 16 +++++++---------
 1 file changed, 7 insertions(+), 9 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/core/cache.c b/drivers/infiniband/core/cache.c
index 1a2984c28b95..ae04826e82fc 100644
--- a/drivers/infiniband/core/cache.c
+++ b/drivers/infiniband/core/cache.c
@@ -770,12 +770,8 @@ static int _gid_table_setup_one(struct ib_device *ib_dev)
 	int err = 0;
 
 	table = kcalloc(ib_dev->phys_port_cnt, sizeof(*table), GFP_KERNEL);
-
-	if (!table) {
-		pr_warn("failed to allocate ib gid cache for %s\n",
-			ib_dev->name);
+	if (!table)
 		return -ENOMEM;
-	}
 
 	for (port = 0; port < ib_dev->phys_port_cnt; port++) {
 		u8 rdma_port = port + rdma_start_port(ib_dev);
@@ -1170,14 +1166,13 @@ int ib_cache_setup_one(struct ib_device *device)
 					  GFP_KERNEL);
 	if (!device->cache.pkey_cache ||
 	    !device->cache.lmc_cache) {
-		pr_warn("Couldn't allocate cache for %s\n", device->name);
-		return -ENOMEM;
+		err = -ENOMEM;
+		goto free;
 	}
 
 	err = gid_table_setup_one(device);
 	if (err)
-		/* Allocated memory will be cleaned in the release function */
-		return err;
+		goto free;
 
 	for (p = 0; p <= rdma_end_port(device) - rdma_start_port(device); ++p)
 		ib_cache_update(device, p + rdma_start_port(device));
@@ -1192,6 +1187,9 @@ int ib_cache_setup_one(struct ib_device *device)
 
 err:
 	gid_table_cleanup_one(device);
+free:
+	kfree(device->cache.pkey_cache);
+	kfree(device->cache.lmc_cache);
 	return err;
 }
 
-- 
cgit v1.2.1


From 15d4626e498c09b66c0f74a107a83bd95abb175c Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leon@kernel.org>
Date: Thu, 3 Nov 2016 16:44:12 +0200
Subject: IB/mlx4: Remove debug prints after allocation failure

The prints after [k|v][m|z|c]alloc() functions are not needed,
because in case of failure, allocator will print their internal
error prints anyway.

Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/alias_GUID.c |  4 +---
 drivers/infiniband/hw/mlx4/cm.c         |  4 +---
 drivers/infiniband/hw/mlx4/mad.c        |  9 ++-------
 drivers/infiniband/hw/mlx4/main.c       | 16 +++-------------
 drivers/infiniband/hw/mlx4/mcg.c        |  5 +----
 5 files changed, 8 insertions(+), 30 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx4/alias_GUID.c b/drivers/infiniband/hw/mlx4/alias_GUID.c
index 5e9939045852..06020c54db20 100644
--- a/drivers/infiniband/hw/mlx4/alias_GUID.c
+++ b/drivers/infiniband/hw/mlx4/alias_GUID.c
@@ -755,10 +755,8 @@ static void alias_guid_work(struct work_struct *work)
 	struct mlx4_ib_dev *dev = container_of(ib_sriov, struct mlx4_ib_dev, sriov);
 
 	rec = kzalloc(sizeof *rec, GFP_KERNEL);
-	if (!rec) {
-		pr_err("alias_guid_work: No Memory\n");
+	if (!rec)
 		return;
-	}
 
 	pr_debug("starting [port: %d]...\n", sriov_alias_port->port + 1);
 	ret = get_next_record_to_update(dev, sriov_alias_port->port, rec);
diff --git a/drivers/infiniband/hw/mlx4/cm.c b/drivers/infiniband/hw/mlx4/cm.c
index 39a488889fc7..d64845335e87 100644
--- a/drivers/infiniband/hw/mlx4/cm.c
+++ b/drivers/infiniband/hw/mlx4/cm.c
@@ -247,10 +247,8 @@ id_map_alloc(struct ib_device *ibdev, int slave_id, u32 sl_cm_id)
 	struct mlx4_ib_sriov *sriov = &to_mdev(ibdev)->sriov;
 
 	ent = kmalloc(sizeof (struct id_map_entry), GFP_KERNEL);
-	if (!ent) {
-		mlx4_ib_warn(ibdev, "Couldn't allocate id cache entry - out of memory\n");
+	if (!ent)
 		return ERR_PTR(-ENOMEM);
-	}
 
 	ent->sl_cm_id = sl_cm_id;
 	ent->slave_id = slave_id;
diff --git a/drivers/infiniband/hw/mlx4/mad.c b/drivers/infiniband/hw/mlx4/mad.c
index 1672907ff219..b0cd66336fcb 100644
--- a/drivers/infiniband/hw/mlx4/mad.c
+++ b/drivers/infiniband/hw/mlx4/mad.c
@@ -1102,10 +1102,8 @@ static void handle_slaves_guid_change(struct mlx4_ib_dev *dev, u8 port_num,
 
 	in_mad  = kmalloc(sizeof *in_mad, GFP_KERNEL);
 	out_mad = kmalloc(sizeof *out_mad, GFP_KERNEL);
-	if (!in_mad || !out_mad) {
-		mlx4_ib_warn(&dev->ib_dev, "failed to allocate memory for guid info mads\n");
+	if (!in_mad || !out_mad)
 		goto out;
-	}
 
 	guid_tbl_blk_num  *= 4;
 
@@ -1916,11 +1914,8 @@ static int alloc_pv_object(struct mlx4_ib_dev *dev, int slave, int port,
 
 	*ret_ctx = NULL;
 	ctx = kzalloc(sizeof (struct mlx4_ib_demux_pv_ctx), GFP_KERNEL);
-	if (!ctx) {
-		pr_err("failed allocating pv resource context "
-		       "for port %d, slave %d\n", port, slave);
+	if (!ctx)
 		return -ENOMEM;
-	}
 
 	ctx->ib_dev = &dev->ib_dev;
 	ctx->port = port;
diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index b597e8227591..1b54786d13d0 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -2814,11 +2814,8 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
 			kmalloc(BITS_TO_LONGS(ibdev->steer_qpn_count) *
 				sizeof(long),
 				GFP_KERNEL);
-		if (!ibdev->ib_uc_qpns_bitmap) {
-			dev_err(&dev->persist->pdev->dev,
-				"bit map alloc failed\n");
+		if (!ibdev->ib_uc_qpns_bitmap)
 			goto err_steer_qp_release;
-		}
 
 		bitmap_zero(ibdev->ib_uc_qpns_bitmap, ibdev->steer_qpn_count);
 
@@ -3055,15 +3052,12 @@ static void do_slave_init(struct mlx4_ib_dev *ibdev, int slave, int do_init)
 	first_port = find_first_bit(actv_ports.ports, dev->caps.num_ports);
 
 	dm = kcalloc(ports, sizeof(*dm), GFP_ATOMIC);
-	if (!dm) {
-		pr_err("failed to allocate memory for tunneling qp update\n");
+	if (!dm)
 		return;
-	}
 
 	for (i = 0; i < ports; i++) {
 		dm[i] = kmalloc(sizeof (struct mlx4_ib_demux_work), GFP_ATOMIC);
 		if (!dm[i]) {
-			pr_err("failed to allocate memory for tunneling qp update work struct\n");
 			while (--i >= 0)
 				kfree(dm[i]);
 			goto out;
@@ -3223,8 +3217,6 @@ void mlx4_sched_ib_sl2vl_update_work(struct mlx4_ib_dev *ibdev,
 		ew->port = port;
 		ew->ib_dev = ibdev;
 		queue_work(wq, &ew->work);
-	} else {
-		pr_err("failed to allocate memory for sl2vl update work\n");
 	}
 }
 
@@ -3284,10 +3276,8 @@ static void mlx4_ib_event(struct mlx4_dev *dev, void *ibdev_ptr,
 
 	case MLX4_DEV_EVENT_PORT_MGMT_CHANGE:
 		ew = kmalloc(sizeof *ew, GFP_ATOMIC);
-		if (!ew) {
-			pr_err("failed to allocate memory for events work\n");
+		if (!ew)
 			break;
-		}
 
 		INIT_WORK(&ew->work, handle_port_mgmt_change_event);
 		memcpy(&ew->ib_eqe, eqe, sizeof *eqe);
diff --git a/drivers/infiniband/hw/mlx4/mcg.c b/drivers/infiniband/hw/mlx4/mcg.c
index a21d37f02f35..e010fe459e67 100644
--- a/drivers/infiniband/hw/mlx4/mcg.c
+++ b/drivers/infiniband/hw/mlx4/mcg.c
@@ -1142,7 +1142,6 @@ void mlx4_ib_mcg_port_cleanup(struct mlx4_ib_demux_ctx *ctx, int destroy_wq)
 	work = kmalloc(sizeof *work, GFP_KERNEL);
 	if (!work) {
 		ctx->flushing = 0;
-		mcg_warn("failed allocating work for cleanup\n");
 		return;
 	}
 
@@ -1202,10 +1201,8 @@ static int push_deleteing_req(struct mcast_group *group, int slave)
 		return 0;
 
 	req = kzalloc(sizeof *req, GFP_KERNEL);
-	if (!req) {
-		mcg_warn_group(group, "failed allocation - may leave stall groups\n");
+	if (!req)
 		return -ENOMEM;
-	}
 
 	if (!list_empty(&group->func[slave].pending)) {
 		pend_req = list_entry(group->func[slave].pending.prev, struct mcast_req, group_list);
-- 
cgit v1.2.1


From 0886d8f0b7c9bb2f85fa8b71c4842f92308de641 Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leon@kernel.org>
Date: Thu, 3 Nov 2016 16:44:13 +0200
Subject: IB/mlx5: Remove debug prints after allocation failure

The prints after [k|v][m|z|c]alloc() functions are not needed,
because in case of failure, allocator will print their internal
error prints anyway.

Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/srq.c | 2 --
 1 file changed, 2 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/srq.c b/drivers/infiniband/hw/mlx5/srq.c
index 3857dbd9c956..c676133750b7 100644
--- a/drivers/infiniband/hw/mlx5/srq.c
+++ b/drivers/infiniband/hw/mlx5/srq.c
@@ -203,8 +203,6 @@ static int create_srq_kernel(struct mlx5_ib_dev *dev, struct mlx5_ib_srq *srq,
 
 	srq->wrid = kmalloc(srq->msrq.max * sizeof(u64), GFP_KERNEL);
 	if (!srq->wrid) {
-		mlx5_ib_dbg(dev, "kmalloc failed %lu\n",
-			    (unsigned long)(srq->msrq.max * sizeof(u64)));
 		err = -ENOMEM;
 		goto err_in;
 	}
-- 
cgit v1.2.1


From 5ce9f115bd796be4f0dd8230e6b8baeb2c558311 Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leon@kernel.org>
Date: Thu, 3 Nov 2016 16:44:14 +0200
Subject: IB/hfi1: Remove debug prints after allocation failure

The prints after [k|v][m|z|c]alloc() functions are not needed,
because in case of failure, allocator will print their internal
error prints anyway.

Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/pio.c | 1 -
 1 file changed, 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/pio.c b/drivers/infiniband/hw/hfi1/pio.c
index 50a3a36d9363..d15ffed48a39 100644
--- a/drivers/infiniband/hw/hfi1/pio.c
+++ b/drivers/infiniband/hw/hfi1/pio.c
@@ -2053,7 +2053,6 @@ int init_credit_return(struct hfi1_devdata *dd)
 		sizeof(struct credit_return_base),
 		GFP_KERNEL);
 	if (!dd->cr_base) {
-		dd_dev_err(dd, "Unable to allocate credit return base\n");
 		ret = -ENOMEM;
 		goto done;
 	}
-- 
cgit v1.2.1


From 51ad2bae213fdd4c63d7c3f2906bfe4e244bad46 Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leon@kernel.org>
Date: Thu, 3 Nov 2016 16:44:15 +0200
Subject: IB/cxgb3: Remove debug prints after allocation failure

The prints after [k|v][m|z|c]alloc() functions are not needed,
because in case of failure, allocator will print their internal
error prints anyway.

Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/cxgb3/cxio_dbg.c | 20 ++++++++------------
 1 file changed, 8 insertions(+), 12 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/cxgb3/cxio_dbg.c b/drivers/infiniband/hw/cxgb3/cxio_dbg.c
index 8bca6b4ec9af..445e89e5e7cf 100644
--- a/drivers/infiniband/hw/cxgb3/cxio_dbg.c
+++ b/drivers/infiniband/hw/cxgb3/cxio_dbg.c
@@ -45,10 +45,9 @@ void cxio_dump_tpt(struct cxio_rdev *rdev, u32 stag)
 	int size = 32;
 
 	m = kmalloc(sizeof(*m) + size, GFP_ATOMIC);
-	if (!m) {
-		PDBG("%s couldn't allocate memory.\n", __func__);
+	if (!m)
 		return;
-	}
+
 	m->mem_id = MEM_PMRX;
 	m->addr = (stag>>8) * 32 + rdev->rnic_info.tpt_base;
 	m->len = size;
@@ -82,10 +81,9 @@ void cxio_dump_pbl(struct cxio_rdev *rdev, u32 pbl_addr, uint len, u8 shift)
 	size = npages * sizeof(u64);
 
 	m = kmalloc(sizeof(*m) + size, GFP_ATOMIC);
-	if (!m) {
-		PDBG("%s couldn't allocate memory.\n", __func__);
+	if (!m)
 		return;
-	}
+
 	m->mem_id = MEM_PMRX;
 	m->addr = pbl_addr;
 	m->len = size;
@@ -144,10 +142,9 @@ void cxio_dump_rqt(struct cxio_rdev *rdev, u32 hwtid, int nents)
 	int rc;
 
 	m = kmalloc(sizeof(*m) + size, GFP_ATOMIC);
-	if (!m) {
-		PDBG("%s couldn't allocate memory.\n", __func__);
+	if (!m)
 		return;
-	}
+
 	m->mem_id = MEM_PMRX;
 	m->addr = ((hwtid)<<10) + rdev->rnic_info.rqt_base;
 	m->len = size;
@@ -177,10 +174,9 @@ void cxio_dump_tcb(struct cxio_rdev *rdev, u32 hwtid)
 	int rc;
 
 	m = kmalloc(sizeof(*m) + size, GFP_ATOMIC);
-	if (!m) {
-		PDBG("%s couldn't allocate memory.\n", __func__);
+	if (!m)
 		return;
-	}
+
 	m->mem_id = MEM_CM;
 	m->addr = hwtid * size;
 	m->len = size;
-- 
cgit v1.2.1


From 9a88f96f21b27b0fccb69601fb725f03fbab841b Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leon@kernel.org>
Date: Thu, 3 Nov 2016 16:44:16 +0200
Subject: IB/cxgb4: Remove debug prints after allocation failure

The prints after [k|v][m|z|c]alloc() functions are not needed,
because in case of failure, allocator will print their internal
error prints anyway.

Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/cxgb4/device.c | 4 ----
 1 file changed, 4 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/cxgb4/device.c b/drivers/infiniband/hw/cxgb4/device.c
index 93e3d270a98a..11a25ec1ce03 100644
--- a/drivers/infiniband/hw/cxgb4/device.c
+++ b/drivers/infiniband/hw/cxgb4/device.c
@@ -841,8 +841,6 @@ static int c4iw_rdev_open(struct c4iw_rdev *rdev)
 		if (rdev->wr_log) {
 			rdev->wr_log_size = 1 << c4iw_wr_log_size_order;
 			atomic_set(&rdev->wr_log_idx, 0);
-		} else {
-			pr_err(MOD "error allocating wr_log. Logging disabled\n");
 		}
 	}
 
@@ -1424,8 +1422,6 @@ static void recover_queues(struct uld_ctx *ctx)
 
 	qp_list.qps = kzalloc(count * sizeof *qp_list.qps, GFP_ATOMIC);
 	if (!qp_list.qps) {
-		printk(KERN_ERR MOD "%s: Fatal error - DB overflow recovery failed\n",
-		       pci_name(ctx->lldi.pdev));
 		spin_unlock_irq(&ctx->dev->lock);
 		return;
 	}
-- 
cgit v1.2.1


From 315b41480bb956b223a04c6556f04c9b7c74c8c2 Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leon@kernel.org>
Date: Thu, 3 Nov 2016 16:44:17 +0200
Subject: IB/i40iw: Remove debug prints after allocation failure

The prints after [k|v][m|z|c]alloc() functions are not needed,
because in case of failure, allocator will print their internal
error prints anyway.

Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_cm.c    | 2 --
 drivers/infiniband/hw/i40iw/i40iw_main.c  | 5 ++---
 drivers/infiniband/hw/i40iw/i40iw_verbs.c | 1 -
 3 files changed, 2 insertions(+), 6 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.c b/drivers/infiniband/hw/i40iw/i40iw_cm.c
index 85637696f6e9..47d1bbce9561 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_cm.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_cm.c
@@ -1675,7 +1675,6 @@ static enum i40iw_status_code i40iw_add_mqh_6(struct i40iw_device *iwdev,
 					    "Allocating child listener %p\n",
 					    child_listen_node);
 				if (!child_listen_node) {
-					i40iw_pr_err("listener memory allocation\n");
 					ret = I40IW_ERR_NO_MEMORY;
 					goto exit;
 				}
@@ -1751,7 +1750,6 @@ static enum i40iw_status_code i40iw_add_mqh_4(
 					    "Allocating child listener %p\n",
 					    child_listen_node);
 				if (!child_listen_node) {
-					i40iw_pr_err("listener memory allocation\n");
 					in_dev_put(idev);
 					ret = I40IW_ERR_NO_MEMORY;
 					goto exit;
diff --git a/drivers/infiniband/hw/i40iw/i40iw_main.c b/drivers/infiniband/hw/i40iw/i40iw_main.c
index ac2f3cd9478c..a6ad913a30dd 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_main.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_main.c
@@ -1301,10 +1301,9 @@ static enum i40iw_status_code i40iw_initialize_dev(struct i40iw_device *iwdev,
 	size = sizeof(struct i40iw_hmc_pble_rsrc) + sizeof(struct i40iw_hmc_info) +
 				(sizeof(struct i40iw_hmc_obj_info) * I40IW_HMC_IW_MAX);
 	iwdev->hmc_info_mem = kzalloc(size, GFP_KERNEL);
-	if (!iwdev->hmc_info_mem) {
-		i40iw_pr_err("memory alloc fail\n");
+	if (!iwdev->hmc_info_mem)
 		return I40IW_ERR_NO_MEMORY;
-	}
+
 	iwdev->pble_rsrc = (struct i40iw_hmc_pble_rsrc *)iwdev->hmc_info_mem;
 	dev->hmc_info = &iwdev->hw.hmc;
 	dev->hmc_info->hmc_obj = (struct i40iw_hmc_obj_info *)(iwdev->pble_rsrc + 1);
diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
index 6329c971c22f..62e068b9bf4c 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
@@ -2654,7 +2654,6 @@ static struct i40iw_ib_device *i40iw_init_rdma_device(struct i40iw_device *iwdev
 	iwibdev->ibdev.iwcm = kzalloc(sizeof(*iwibdev->ibdev.iwcm), GFP_KERNEL);
 	if (!iwibdev->ibdev.iwcm) {
 		ib_dealloc_device(&iwibdev->ibdev);
-		i40iw_pr_err("iwcm == NULL\n");
 		return NULL;
 	}
 
-- 
cgit v1.2.1


From c40a83b9786e99a78c849f2f33448426bf8fa0f2 Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leon@kernel.org>
Date: Thu, 3 Nov 2016 16:44:18 +0200
Subject: IB/qib: Remove debug prints after allocation failure

The prints after [k|v][m|z|c]alloc() functions are not needed,
because in case of failure, allocator will print their internal
error prints anyway.

Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/qib/qib_diag.c    |  6 +----
 drivers/infiniband/hw/qib/qib_eeprom.c  |  6 +----
 drivers/infiniband/hw/qib/qib_iba6120.c |  8 +-----
 drivers/infiniband/hw/qib/qib_iba7220.c |  8 +-----
 drivers/infiniband/hw/qib/qib_iba7322.c | 22 ++++-----------
 drivers/infiniband/hw/qib/qib_init.c    | 47 ++++++---------------------------
 6 files changed, 17 insertions(+), 80 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/qib/qib_diag.c b/drivers/infiniband/hw/qib/qib_diag.c
index 8c34b23e5bf6..775018b32b0d 100644
--- a/drivers/infiniband/hw/qib/qib_diag.c
+++ b/drivers/infiniband/hw/qib/qib_diag.c
@@ -609,8 +609,6 @@ static ssize_t qib_diagpkt_write(struct file *fp,
 
 	tmpbuf = vmalloc(plen);
 	if (!tmpbuf) {
-		qib_devinfo(dd->pcidev,
-			"Unable to allocate tmp buffer, failing\n");
 		ret = -ENOMEM;
 		goto bail;
 	}
@@ -702,10 +700,8 @@ int qib_register_observer(struct qib_devdata *dd,
 	if (!dd || !op)
 		return -EINVAL;
 	olp = vmalloc(sizeof(*olp));
-	if (!olp) {
-		pr_err("vmalloc for observer failed\n");
+	if (!olp)
 		return -ENOMEM;
-	}
 
 	spin_lock_irqsave(&dd->qib_diag_trans_lock, flags);
 	olp->op = op;
diff --git a/drivers/infiniband/hw/qib/qib_eeprom.c b/drivers/infiniband/hw/qib/qib_eeprom.c
index 311ee6c3dd5e..33a2e74c8495 100644
--- a/drivers/infiniband/hw/qib/qib_eeprom.c
+++ b/drivers/infiniband/hw/qib/qib_eeprom.c
@@ -182,12 +182,8 @@ void qib_get_eeprom_info(struct qib_devdata *dd)
 	 * */
 	len = sizeof(struct qib_flash);
 	buf = vmalloc(len);
-	if (!buf) {
-		qib_dev_err(dd,
-			"Couldn't allocate memory to read %u bytes from eeprom for GUID\n",
-			len);
+	if (!buf)
 		goto bail;
-	}
 
 	/*
 	 * Use "public" eeprom read function, which does locking and
diff --git a/drivers/infiniband/hw/qib/qib_iba6120.c b/drivers/infiniband/hw/qib/qib_iba6120.c
index a3733f25280f..92399d3ffd15 100644
--- a/drivers/infiniband/hw/qib/qib_iba6120.c
+++ b/drivers/infiniband/hw/qib/qib_iba6120.c
@@ -1759,9 +1759,7 @@ static void pe_boardname(struct qib_devdata *dd)
 	}
 	namelen = strlen(n) + 1;
 	dd->boardname = kmalloc(namelen, GFP_KERNEL);
-	if (!dd->boardname)
-		qib_dev_err(dd, "Failed allocation for board name: %s\n", n);
-	else
+	if (dd->boardname)
 		snprintf(dd->boardname, namelen, "%s", n);
 
 	if (dd->majrev != 4 || !dd->minrev || dd->minrev > 2)
@@ -2533,8 +2531,6 @@ static void init_6120_cntrnames(struct qib_devdata *dd)
 		dd->cspec->cntrnamelen = 1 + s - cntr6120names;
 	dd->cspec->cntrs = kmalloc(dd->cspec->ncntrs
 		* sizeof(u64), GFP_KERNEL);
-	if (!dd->cspec->cntrs)
-		qib_dev_err(dd, "Failed allocation for counters\n");
 
 	for (i = 0, s = (char *)portcntr6120names; s; i++)
 		s = strchr(s + 1, '\n');
@@ -2542,8 +2538,6 @@ static void init_6120_cntrnames(struct qib_devdata *dd)
 	dd->cspec->portcntrnamelen = sizeof(portcntr6120names) - 1;
 	dd->cspec->portcntrs = kmalloc(dd->cspec->nportcntrs
 		* sizeof(u64), GFP_KERNEL);
-	if (!dd->cspec->portcntrs)
-		qib_dev_err(dd, "Failed allocation for portcounters\n");
 }
 
 static u32 qib_read_6120cntrs(struct qib_devdata *dd, loff_t pos, char **namep,
diff --git a/drivers/infiniband/hw/qib/qib_iba7220.c b/drivers/infiniband/hw/qib/qib_iba7220.c
index 00b2af211157..e55e31a69195 100644
--- a/drivers/infiniband/hw/qib/qib_iba7220.c
+++ b/drivers/infiniband/hw/qib/qib_iba7220.c
@@ -2070,9 +2070,7 @@ static void qib_7220_boardname(struct qib_devdata *dd)
 
 	namelen = strlen(n) + 1;
 	dd->boardname = kmalloc(namelen, GFP_KERNEL);
-	if (!dd->boardname)
-		qib_dev_err(dd, "Failed allocation for board name: %s\n", n);
-	else
+	if (dd->boardname)
 		snprintf(dd->boardname, namelen, "%s", n);
 
 	if (dd->majrev != 5 || !dd->minrev || dd->minrev > 2)
@@ -3179,8 +3177,6 @@ static void init_7220_cntrnames(struct qib_devdata *dd)
 		dd->cspec->cntrnamelen = 1 + s - cntr7220names;
 	dd->cspec->cntrs = kmalloc(dd->cspec->ncntrs
 		* sizeof(u64), GFP_KERNEL);
-	if (!dd->cspec->cntrs)
-		qib_dev_err(dd, "Failed allocation for counters\n");
 
 	for (i = 0, s = (char *)portcntr7220names; s; i++)
 		s = strchr(s + 1, '\n');
@@ -3188,8 +3184,6 @@ static void init_7220_cntrnames(struct qib_devdata *dd)
 	dd->cspec->portcntrnamelen = sizeof(portcntr7220names) - 1;
 	dd->cspec->portcntrs = kmalloc(dd->cspec->nportcntrs
 		* sizeof(u64), GFP_KERNEL);
-	if (!dd->cspec->portcntrs)
-		qib_dev_err(dd, "Failed allocation for portcounters\n");
 }
 
 static u32 qib_read_7220cntrs(struct qib_devdata *dd, loff_t pos, char **namep,
diff --git a/drivers/infiniband/hw/qib/qib_iba7322.c b/drivers/infiniband/hw/qib/qib_iba7322.c
index ded27172320e..c4a3616062f1 100644
--- a/drivers/infiniband/hw/qib/qib_iba7322.c
+++ b/drivers/infiniband/hw/qib/qib_iba7322.c
@@ -3627,9 +3627,7 @@ static unsigned qib_7322_boardname(struct qib_devdata *dd)
 
 	namelen = strlen(n) + 1;
 	dd->boardname = kmalloc(namelen, GFP_KERNEL);
-	if (!dd->boardname)
-		qib_dev_err(dd, "Failed allocation for board name: %s\n", n);
-	else
+	if (dd->boardname)
 		snprintf(dd->boardname, namelen, "%s", n);
 
 	snprintf(dd->boardversion, sizeof(dd->boardversion),
@@ -3656,7 +3654,7 @@ static unsigned qib_7322_boardname(struct qib_devdata *dd)
 static int qib_do_7322_reset(struct qib_devdata *dd)
 {
 	u64 val;
-	u64 *msix_vecsave;
+	u64 *msix_vecsave = NULL;
 	int i, msix_entries, ret = 1;
 	u16 cmdval;
 	u8 int_line, clinesz;
@@ -3677,10 +3675,7 @@ static int qib_do_7322_reset(struct qib_devdata *dd)
 		/* can be up to 512 bytes, too big for stack */
 		msix_vecsave = kmalloc(2 * dd->cspec->num_msix_entries *
 			sizeof(u64), GFP_KERNEL);
-		if (!msix_vecsave)
-			qib_dev_err(dd, "No mem to save MSIx data\n");
-	} else
-		msix_vecsave = NULL;
+	}
 
 	/*
 	 * Core PCI (as of 2.6.18) doesn't save or rewrite the full vector
@@ -5043,8 +5038,6 @@ static void init_7322_cntrnames(struct qib_devdata *dd)
 		dd->cspec->cntrnamelen = 1 + s - cntr7322names;
 	dd->cspec->cntrs = kmalloc(dd->cspec->ncntrs
 		* sizeof(u64), GFP_KERNEL);
-	if (!dd->cspec->cntrs)
-		qib_dev_err(dd, "Failed allocation for counters\n");
 
 	for (i = 0, s = (char *)portcntr7322names; s; i++)
 		s = strchr(s + 1, '\n');
@@ -5053,9 +5046,6 @@ static void init_7322_cntrnames(struct qib_devdata *dd)
 	for (i = 0; i < dd->num_pports; ++i) {
 		dd->pport[i].cpspec->portcntrs = kmalloc(dd->cspec->nportcntrs
 			* sizeof(u64), GFP_KERNEL);
-		if (!dd->pport[i].cpspec->portcntrs)
-			qib_dev_err(dd,
-				"Failed allocation for portcounters\n");
 	}
 }
 
@@ -6461,7 +6451,6 @@ static int qib_init_7322_variables(struct qib_devdata *dd)
 		sizeof(*dd->cspec->sendibchk), GFP_KERNEL);
 	if (!dd->cspec->sendchkenable || !dd->cspec->sendgrhchk ||
 		!dd->cspec->sendibchk) {
-		qib_dev_err(dd, "Failed allocation for hdrchk bitmaps\n");
 		ret = -ENOMEM;
 		goto bail;
 	}
@@ -7338,10 +7327,9 @@ struct qib_devdata *qib_init_iba7322_funcs(struct pci_dev *pdev,
 	tabsize = actual_cnt;
 	dd->cspec->msix_entries = kzalloc(tabsize *
 			sizeof(struct qib_msix_entry), GFP_KERNEL);
-	if (!dd->cspec->msix_entries) {
-		qib_dev_err(dd, "No memory for MSIx table\n");
+	if (!dd->cspec->msix_entries)
 		tabsize = 0;
-	}
+
 	for (i = 0; i < tabsize; i++)
 		dd->cspec->msix_entries[i].msix.entry = i;
 
diff --git a/drivers/infiniband/hw/qib/qib_init.c b/drivers/infiniband/hw/qib/qib_init.c
index 1730aa839a47..b50240b1d5a4 100644
--- a/drivers/infiniband/hw/qib/qib_init.c
+++ b/drivers/infiniband/hw/qib/qib_init.c
@@ -133,11 +133,8 @@ int qib_create_ctxts(struct qib_devdata *dd)
 	 * cleanup iterates across all possible ctxts.
 	 */
 	dd->rcd = kcalloc(dd->ctxtcnt, sizeof(*dd->rcd), GFP_KERNEL);
-	if (!dd->rcd) {
-		qib_dev_err(dd,
-			"Unable to allocate ctxtdata array, failing\n");
+	if (!dd->rcd)
 		return -ENOMEM;
-	}
 
 	/* create (one or more) kctxt */
 	for (i = 0; i < dd->first_user_ctxt; ++i) {
@@ -265,39 +262,23 @@ int qib_init_pportdata(struct qib_pportdata *ppd, struct qib_devdata *dd,
 	size = IB_CC_TABLE_CAP_DEFAULT * sizeof(struct ib_cc_table_entry)
 		* IB_CCT_ENTRIES;
 	ppd->ccti_entries = kzalloc(size, GFP_KERNEL);
-	if (!ppd->ccti_entries) {
-		qib_dev_err(dd,
-		  "failed to allocate congestion control table for port %d!\n",
-		  port);
+	if (!ppd->ccti_entries)
 		goto bail;
-	}
 
 	size = IB_CC_CCS_ENTRIES * sizeof(struct ib_cc_congestion_entry);
 	ppd->congestion_entries = kzalloc(size, GFP_KERNEL);
-	if (!ppd->congestion_entries) {
-		qib_dev_err(dd,
-		 "failed to allocate congestion setting list for port %d!\n",
-		 port);
+	if (!ppd->congestion_entries)
 		goto bail_1;
-	}
 
 	size = sizeof(struct cc_table_shadow);
 	ppd->ccti_entries_shadow = kzalloc(size, GFP_KERNEL);
-	if (!ppd->ccti_entries_shadow) {
-		qib_dev_err(dd,
-		 "failed to allocate shadow ccti list for port %d!\n",
-		 port);
+	if (!ppd->ccti_entries_shadow)
 		goto bail_2;
-	}
 
 	size = sizeof(struct ib_cc_congestion_setting_attr);
 	ppd->congestion_entries_shadow = kzalloc(size, GFP_KERNEL);
-	if (!ppd->congestion_entries_shadow) {
-		qib_dev_err(dd,
-		 "failed to allocate shadow congestion setting list for port %d!\n",
-		 port);
+	if (!ppd->congestion_entries_shadow)
 		goto bail_3;
-	}
 
 	return 0;
 
@@ -391,18 +372,12 @@ static void init_shadow_tids(struct qib_devdata *dd)
 	dma_addr_t *addrs;
 
 	pages = vzalloc(dd->cfgctxts * dd->rcvtidcnt * sizeof(struct page *));
-	if (!pages) {
-		qib_dev_err(dd,
-			"failed to allocate shadow page * array, no expected sends!\n");
+	if (!pages)
 		goto bail;
-	}
 
 	addrs = vzalloc(dd->cfgctxts * dd->rcvtidcnt * sizeof(dma_addr_t));
-	if (!addrs) {
-		qib_dev_err(dd,
-			"failed to allocate shadow dma handle array, no expected sends!\n");
+	if (!addrs)
 		goto bail_free;
-	}
 
 	dd->pageshadow = pages;
 	dd->physshadow = addrs;
@@ -1026,11 +1001,8 @@ static void qib_verify_pioperf(struct qib_devdata *dd)
 	cnt = 1024;
 
 	addr = vmalloc(cnt);
-	if (!addr) {
-		qib_devinfo(dd->pcidev,
-			 "Couldn't get memory for checking PIO perf, skipping\n");
+	if (!addr)
 		goto done;
-	}
 
 	preempt_disable();  /* we want reasonably accurate elapsed time */
 	msecs = 1 + jiffies_to_msecs(jiffies);
@@ -1172,9 +1144,6 @@ struct qib_devdata *qib_alloc_devdata(struct pci_dev *pdev, size_t extra)
 				      sizeof(long), GFP_KERNEL);
 		if (qib_cpulist)
 			qib_cpulist_count = count;
-		else
-			qib_early_err(&pdev->dev,
-				"Could not alloc cpulist info, cpu affinity might be wrong\n");
 	}
 #ifdef CONFIG_DEBUG_FS
 	qib_dbg_ibdev_init(&dd->verbs_dev);
-- 
cgit v1.2.1


From 2e65835a1ba0607cb212ec687e509ee6c24a68ed Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leon@kernel.org>
Date: Thu, 3 Nov 2016 16:44:19 +0200
Subject: IB/nes: Remove debug prints after allocation failure

The prints after [k|v][m|z|c]alloc() functions are not needed,
because in case of failure, allocator will print their internal
error prints anyway.

Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/nes/nes.c       |  1 -
 drivers/infiniband/hw/nes/nes_cm.c    |  4 +---
 drivers/infiniband/hw/nes/nes_hw.c    |  6 ++----
 drivers/infiniband/hw/nes/nes_mgt.c   | 10 +++-------
 drivers/infiniband/hw/nes/nes_verbs.c |  4 ----
 5 files changed, 6 insertions(+), 19 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/nes/nes.c b/drivers/infiniband/hw/nes/nes.c
index 35cbb17bec12..9badd0224ec5 100644
--- a/drivers/infiniband/hw/nes/nes.c
+++ b/drivers/infiniband/hw/nes/nes.c
@@ -516,7 +516,6 @@ static int nes_probe(struct pci_dev *pcidev, const struct pci_device_id *ent)
 	/* Allocate hardware structure */
 	nesdev = kzalloc(sizeof(struct nes_device), GFP_KERNEL);
 	if (!nesdev) {
-		printk(KERN_ERR PFX "%s: Unable to alloc hardware struct\n", pci_name(pcidev));
 		ret = -ENOMEM;
 		goto bail2;
 	}
diff --git a/drivers/infiniband/hw/nes/nes_cm.c b/drivers/infiniband/hw/nes/nes_cm.c
index 57db9b332f44..8e703479e7ce 100644
--- a/drivers/infiniband/hw/nes/nes_cm.c
+++ b/drivers/infiniband/hw/nes/nes_cm.c
@@ -2282,10 +2282,8 @@ static struct nes_cm_listener *mini_cm_listen(struct nes_cm_core *cm_core,
 	if (!listener) {
 		/* create a CM listen node (1/2 node to compare incoming traffic to) */
 		listener = kzalloc(sizeof(*listener), GFP_ATOMIC);
-		if (!listener) {
-			nes_debug(NES_DBG_CM, "Not creating listener memory allocation failed\n");
+		if (!listener)
 			return NULL;
-		}
 
 		listener->loc_addr = cm_info->loc_addr;
 		listener->loc_port = cm_info->loc_port;
diff --git a/drivers/infiniband/hw/nes/nes_hw.c b/drivers/infiniband/hw/nes/nes_hw.c
index a1c6481d8038..19acd13c6cb1 100644
--- a/drivers/infiniband/hw/nes/nes_hw.c
+++ b/drivers/infiniband/hw/nes/nes_hw.c
@@ -351,9 +351,8 @@ struct nes_adapter *nes_init_adapter(struct nes_device *nesdev, u8 hw_rev) {
 
 	/* allocate a new adapter struct */
 	nesadapter = kzalloc(adapter_size, GFP_KERNEL);
-	if (nesadapter == NULL) {
+	if (!nesadapter)
 		return NULL;
-	}
 
 	nes_debug(NES_DBG_INIT, "Allocating new nesadapter @ %p, size = %u (actual size = %u).\n",
 			nesadapter, (u32)sizeof(struct nes_adapter), adapter_size);
@@ -1007,8 +1006,7 @@ int nes_init_cqp(struct nes_device *nesdev)
 	/* Allocate a twice the number of CQP requests as the SQ size */
 	nesdev->nes_cqp_requests = kzalloc(sizeof(struct nes_cqp_request) *
 			2 * NES_CQP_SQ_SIZE, GFP_KERNEL);
-	if (nesdev->nes_cqp_requests == NULL) {
-		nes_debug(NES_DBG_INIT, "Unable to allocate memory CQP request entries.\n");
+	if (!nesdev->nes_cqp_requests) {
 		pci_free_consistent(nesdev->pcidev, nesdev->cqp_mem_size, nesdev->cqp.sq_vbase,
 				nesdev->cqp.sq_pbase);
 		return -ENOMEM;
diff --git a/drivers/infiniband/hw/nes/nes_mgt.c b/drivers/infiniband/hw/nes/nes_mgt.c
index 416645259b0f..33624f17c347 100644
--- a/drivers/infiniband/hw/nes/nes_mgt.c
+++ b/drivers/infiniband/hw/nes/nes_mgt.c
@@ -320,8 +320,7 @@ static int get_fpdu_info(struct nes_device *nesdev, struct nes_qp *nesqp,
 
 	/* Found one */
 	fpdu_info = kzalloc(sizeof(*fpdu_info), GFP_ATOMIC);
-	if (fpdu_info == NULL) {
-		nes_debug(NES_DBG_PAU, "Failed to alloc a fpdu_info.\n");
+	if (!fpdu_info) {
 		rc = -ENOMEM;
 		goto out;
 	}
@@ -729,8 +728,7 @@ static int nes_change_quad_hash(struct nes_device *nesdev,
 	}
 
 	qh_chg = kmalloc(sizeof *qh_chg, GFP_ATOMIC);
-	if (qh_chg == NULL) {
-		nes_debug(NES_DBG_PAU, "Failed to get a cqp_request.\n");
+	if (!qh_chg) {
 		ret = -ENOMEM;
 		goto chg_qh_err;
 	}
@@ -880,10 +878,8 @@ int nes_init_mgt_qp(struct nes_device *nesdev, struct net_device *netdev, struct
 
 	/* Allocate space the all mgt QPs once */
 	mgtvnic = kzalloc(NES_MGT_QP_COUNT * sizeof(struct nes_vnic_mgt), GFP_KERNEL);
-	if (mgtvnic == NULL) {
-		nes_debug(NES_DBG_INIT, "Unable to allocate memory for mgt structure\n");
+	if (!mgtvnic)
 		return -ENOMEM;
-	}
 
 	/* Allocate fragment, RQ, and CQ; Reuse CEQ based on the PCI function */
 	/* We are not sending from this NIC so sq is not allocated */
diff --git a/drivers/infiniband/hw/nes/nes_verbs.c b/drivers/infiniband/hw/nes/nes_verbs.c
index bd69125731c1..42ab31d06ef9 100644
--- a/drivers/infiniband/hw/nes/nes_verbs.c
+++ b/drivers/infiniband/hw/nes/nes_verbs.c
@@ -1075,7 +1075,6 @@ static struct ib_qp *nes_create_qp(struct ib_pd *ibpd,
 			mem = kzalloc(sizeof(*nesqp)+NES_SW_CONTEXT_ALIGN-1, GFP_KERNEL);
 			if (!mem) {
 				nes_free_resource(nesadapter, nesadapter->allocated_qps, qp_num);
-				nes_debug(NES_DBG_QP, "Unable to allocate QP\n");
 				return ERR_PTR(-ENOMEM);
 			}
 			u64nesqp = (unsigned long)mem;
@@ -1475,7 +1474,6 @@ static struct ib_cq *nes_create_cq(struct ib_device *ibdev,
 	nescq = kzalloc(sizeof(struct nes_cq), GFP_KERNEL);
 	if (!nescq) {
 		nes_free_resource(nesadapter, nesadapter->allocated_cqs, cq_num);
-		nes_debug(NES_DBG_CQ, "Unable to allocate nes_cq struct\n");
 		return ERR_PTR(-ENOMEM);
 	}
 
@@ -2408,7 +2406,6 @@ static struct ib_mr *nes_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 			}
 			nespbl = kzalloc(sizeof(*nespbl), GFP_KERNEL);
 			if (!nespbl) {
-				nes_debug(NES_DBG_MR, "Unable to allocate PBL\n");
 				ib_umem_release(region);
 				return ERR_PTR(-ENOMEM);
 			}
@@ -2416,7 +2413,6 @@ static struct ib_mr *nes_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 			if (!nesmr) {
 				ib_umem_release(region);
 				kfree(nespbl);
-				nes_debug(NES_DBG_MR, "Unable to allocate nesmr\n");
 				return ERR_PTR(-ENOMEM);
 			}
 			nesmr->region = region;
-- 
cgit v1.2.1


From 870b2852455266996e55b3152a5ef40b45e853ec Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leon@kernel.org>
Date: Thu, 3 Nov 2016 16:44:20 +0200
Subject: IB/mthca: Remove debug prints after allocation failure

The prints after [k|v][m|z|c]alloc() functions are not needed,
because in case of failure, allocator will print their internal
error prints anyway.

Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mthca/mthca_reset.c | 4 ----
 1 file changed, 4 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mthca/mthca_reset.c b/drivers/infiniband/hw/mthca/mthca_reset.c
index 6727af27c017..2a6979e4ae1c 100644
--- a/drivers/infiniband/hw/mthca/mthca_reset.c
+++ b/drivers/infiniband/hw/mthca/mthca_reset.c
@@ -96,8 +96,6 @@ int mthca_reset(struct mthca_dev *mdev)
 	hca_header = kmalloc(256, GFP_KERNEL);
 	if (!hca_header) {
 		err = -ENOMEM;
-		mthca_err(mdev, "Couldn't allocate memory to save HCA "
-			  "PCI header, aborting.\n");
 		goto put_dev;
 	}
 
@@ -119,8 +117,6 @@ int mthca_reset(struct mthca_dev *mdev)
 		bridge_header = kmalloc(256, GFP_KERNEL);
 		if (!bridge_header) {
 			err = -ENOMEM;
-			mthca_err(mdev, "Couldn't allocate memory to save HCA "
-				  "bridge PCI header, aborting.\n");
 			goto free_hca;
 		}
 
-- 
cgit v1.2.1


From 02d93f8e6b8bfef8453c4dfa61220335cabbb138 Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leon@kernel.org>
Date: Thu, 3 Nov 2016 16:44:21 +0200
Subject: IB/usninc: Remove and fix debug prints after allocation failure

This patch removes unneeded prints after allocation failure
and moves one debug print into the appropriate place.

Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/usnic/usnic_ib_qp_grp.c | 10 +---------
 drivers/infiniband/hw/usnic/usnic_vnic.c      | 22 ++++++----------------
 2 files changed, 7 insertions(+), 25 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/usnic/usnic_ib_qp_grp.c b/drivers/infiniband/hw/usnic/usnic_ib_qp_grp.c
index 5b0248adf4ce..0e813ecbcd7d 100644
--- a/drivers/infiniband/hw/usnic/usnic_ib_qp_grp.c
+++ b/drivers/infiniband/hw/usnic/usnic_ib_qp_grp.c
@@ -228,8 +228,6 @@ create_roce_custom_flow(struct usnic_ib_qp_grp *qp_grp,
 
 	flow = usnic_fwd_alloc_flow(qp_grp->ufdev, &filter, &uaction);
 	if (IS_ERR_OR_NULL(flow)) {
-		usnic_err("Unable to alloc flow failed with err %ld\n",
-				PTR_ERR(flow));
 		err = flow ? PTR_ERR(flow) : -EFAULT;
 		goto out_unreserve_port;
 	}
@@ -303,8 +301,6 @@ create_udp_flow(struct usnic_ib_qp_grp *qp_grp,
 
 	flow = usnic_fwd_alloc_flow(qp_grp->ufdev, &filter, &uaction);
 	if (IS_ERR_OR_NULL(flow)) {
-		usnic_err("Unable to alloc flow failed with err %ld\n",
-				PTR_ERR(flow));
 		err = flow ? PTR_ERR(flow) : -EFAULT;
 		goto out_put_sock;
 	}
@@ -694,18 +690,14 @@ usnic_ib_qp_grp_create(struct usnic_fwd_dev *ufdev, struct usnic_ib_vf *vf,
 	}
 
 	qp_grp = kzalloc(sizeof(*qp_grp), GFP_ATOMIC);
-	if (!qp_grp) {
-		usnic_err("Unable to alloc qp_grp - Out of memory\n");
+	if (!qp_grp)
 		return NULL;
-	}
 
 	qp_grp->res_chunk_list = alloc_res_chunk_list(vf->vnic, res_spec,
 							qp_grp);
 	if (IS_ERR_OR_NULL(qp_grp->res_chunk_list)) {
 		err = qp_grp->res_chunk_list ?
 				PTR_ERR(qp_grp->res_chunk_list) : -ENOMEM;
-		usnic_err("Unable to alloc res for %d with err %d\n",
-				qp_grp->grp_id, err);
 		goto out_free_qp_grp;
 	}
 
diff --git a/drivers/infiniband/hw/usnic/usnic_vnic.c b/drivers/infiniband/hw/usnic/usnic_vnic.c
index 887510718690..e7b0030254da 100644
--- a/drivers/infiniband/hw/usnic/usnic_vnic.c
+++ b/drivers/infiniband/hw/usnic/usnic_vnic.c
@@ -241,17 +241,12 @@ usnic_vnic_get_resources(struct usnic_vnic *vnic, enum usnic_vnic_res_type type,
 		return ERR_PTR(-EINVAL);
 
 	ret = kzalloc(sizeof(*ret), GFP_ATOMIC);
-	if (!ret) {
-		usnic_err("Failed to allocate chunk for %s - Out of memory\n",
-				usnic_vnic_pci_name(vnic));
+	if (!ret)
 		return ERR_PTR(-ENOMEM);
-	}
 
 	if (cnt > 0) {
 		ret->res = kcalloc(cnt, sizeof(*(ret->res)), GFP_ATOMIC);
 		if (!ret->res) {
-			usnic_err("Failed to allocate resources for %s. Out of memory\n",
-					usnic_vnic_pci_name(vnic));
 			kfree(ret);
 			return ERR_PTR(-ENOMEM);
 		}
@@ -311,8 +306,10 @@ static int usnic_vnic_alloc_res_chunk(struct usnic_vnic *vnic,
 	struct usnic_vnic_res *res;
 
 	cnt = vnic_dev_get_res_count(vnic->vdev, _to_vnic_res_type(type));
-	if (cnt < 1)
+	if (cnt < 1) {
+		usnic_err("Wrong res count with cnt %d\n", cnt);
 		return -EINVAL;
+	}
 
 	chunk->cnt = chunk->free_cnt = cnt;
 	chunk->res = kzalloc(sizeof(*(chunk->res))*cnt, GFP_KERNEL);
@@ -384,12 +381,8 @@ static int usnic_vnic_discover_resources(struct pci_dev *pdev,
 			res_type < USNIC_VNIC_RES_TYPE_MAX; res_type++) {
 		err = usnic_vnic_alloc_res_chunk(vnic, res_type,
 						&vnic->chunks[res_type]);
-		if (err) {
-			usnic_err("Failed to alloc res %s with err %d\n",
-					usnic_vnic_res_type_to_str(res_type),
-					err);
+		if (err)
 			goto out_clean_chunks;
-		}
 	}
 
 	return 0;
@@ -454,11 +447,8 @@ struct usnic_vnic *usnic_vnic_alloc(struct pci_dev *pdev)
 	}
 
 	vnic = kzalloc(sizeof(*vnic), GFP_KERNEL);
-	if (!vnic) {
-		usnic_err("Failed to alloc vnic for %s - out of memory\n",
-				pci_name(pdev));
+	if (!vnic)
 		return ERR_PTR(-ENOMEM);
-	}
 
 	spin_lock_init(&vnic->res_lock);
 
-- 
cgit v1.2.1


From 740c330ee677f4e79c3865f85c6af5a5c961fe34 Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leon@kernel.org>
Date: Thu, 3 Nov 2016 16:44:22 +0200
Subject: IB/ocrdma: Remove and fix debug prints after allocation failure

The prints after [k|v][m|z|c]alloc() functions are not needed,
because in case of failure, allocator will print their internal
error prints anyway.

Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/ocrdma/ocrdma_hw.c    | 5 ++---
 drivers/infiniband/hw/ocrdma/ocrdma_stats.c | 4 +---
 2 files changed, 3 insertions(+), 6 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_hw.c b/drivers/infiniband/hw/ocrdma/ocrdma_hw.c
index 67fc0b6857e1..12b27e7bc6a7 100644
--- a/drivers/infiniband/hw/ocrdma/ocrdma_hw.c
+++ b/drivers/infiniband/hw/ocrdma/ocrdma_hw.c
@@ -1596,10 +1596,9 @@ void ocrdma_alloc_pd_pool(struct ocrdma_dev *dev)
 
 	dev->pd_mgr = kzalloc(sizeof(struct ocrdma_pd_resource_mgr),
 			      GFP_KERNEL);
-	if (!dev->pd_mgr) {
-		pr_err("%s(%d)Memory allocation failure.\n", __func__, dev->id);
+	if (!dev->pd_mgr)
 		return;
-	}
+
 	status = ocrdma_mbx_alloc_pd_range(dev);
 	if (status) {
 		pr_err("%s(%d) Unable to initialize PD pool, using default.\n",
diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_stats.c b/drivers/infiniband/hw/ocrdma/ocrdma_stats.c
index 8bef09a8c49f..f8e4b0a6486f 100644
--- a/drivers/infiniband/hw/ocrdma/ocrdma_stats.c
+++ b/drivers/infiniband/hw/ocrdma/ocrdma_stats.c
@@ -84,10 +84,8 @@ bool ocrdma_alloc_stats_resources(struct ocrdma_dev *dev)
 
 	/* Alloc debugfs mem */
 	mem->debugfs_mem = kzalloc(OCRDMA_MAX_DBGFS_MEM, GFP_KERNEL);
-	if (!mem->debugfs_mem) {
-		pr_err("%s: stats debugfs mem allocation failed\n", __func__);
+	if (!mem->debugfs_mem)
 		return false;
-	}
 
 	return true;
 }
-- 
cgit v1.2.1


From 907610bfdf1a7f3d173a9593fde70f67d63476d1 Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leon@kernel.org>
Date: Thu, 3 Nov 2016 16:44:23 +0200
Subject: IB/rxe: Remove and fix debug prints after allocation failure

The prints after [k|v][m|z|c]alloc() functions are not needed,
because in case of failure, allocator will print their internal
error prints anyway.

Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rxe/rxe_pool.c | 1 -
 1 file changed, 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rxe/rxe_pool.c b/drivers/infiniband/sw/rxe/rxe_pool.c
index 6bac0717c540..d723947a8542 100644
--- a/drivers/infiniband/sw/rxe/rxe_pool.c
+++ b/drivers/infiniband/sw/rxe/rxe_pool.c
@@ -180,7 +180,6 @@ static int rxe_pool_init_index(struct rxe_pool *pool, u32 max, u32 min)
 	size = BITS_TO_LONGS(max - min + 1) * sizeof(long);
 	pool->table = kmalloc(size, GFP_KERNEL);
 	if (!pool->table) {
-		pr_warn("no memory for bit table\n");
 		err = -ENOMEM;
 		goto out;
 	}
-- 
cgit v1.2.1


From 93b80f29044639ce98baabd3431af9d78e9d1223 Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leon@kernel.org>
Date: Thu, 3 Nov 2016 16:44:24 +0200
Subject: IB/isert: Remove and fix debug prints after allocation failure

The prints after [k|v][m|z|c]alloc() functions are not needed,
because in case of failure, allocator will print their internal
error prints anyway.

Signed-off-by: Leon Romanovsky <leon@kernel.org>
Acked-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/ulp/isert/ib_isert.c | 23 ++++++-----------------
 1 file changed, 6 insertions(+), 17 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/ulp/isert/ib_isert.c b/drivers/infiniband/ulp/isert/ib_isert.c
index 6dd43f63238e..225cb8255563 100644
--- a/drivers/infiniband/ulp/isert/ib_isert.c
+++ b/drivers/infiniband/ulp/isert/ib_isert.c
@@ -184,7 +184,7 @@ isert_alloc_rx_descriptors(struct isert_conn *isert_conn)
 	isert_conn->rx_descs = kzalloc(ISERT_QP_MAX_RECV_DTOS *
 				sizeof(struct iser_rx_desc), GFP_KERNEL);
 	if (!isert_conn->rx_descs)
-		goto fail;
+		return -ENOMEM;
 
 	rx_desc = isert_conn->rx_descs;
 
@@ -213,9 +213,7 @@ dma_map_fail:
 	}
 	kfree(isert_conn->rx_descs);
 	isert_conn->rx_descs = NULL;
-fail:
 	isert_err("conn %p failed to allocate rx descriptors\n", isert_conn);
-
 	return -ENOMEM;
 }
 
@@ -269,10 +267,8 @@ isert_alloc_comps(struct isert_device *device)
 
 	device->comps = kcalloc(device->comps_used, sizeof(struct isert_comp),
 				GFP_KERNEL);
-	if (!device->comps) {
-		isert_err("Unable to allocate completion contexts\n");
+	if (!device->comps)
 		return -ENOMEM;
-	}
 
 	max_cqe = min(ISER_MAX_CQ_LEN, device->ib_device->attrs.max_cqe);
 
@@ -432,10 +428,8 @@ isert_alloc_login_buf(struct isert_conn *isert_conn,
 
 	isert_conn->login_req_buf = kzalloc(sizeof(*isert_conn->login_req_buf),
 			GFP_KERNEL);
-	if (!isert_conn->login_req_buf) {
-		isert_err("Unable to allocate isert_conn->login_buf\n");
+	if (!isert_conn->login_req_buf)
 		return -ENOMEM;
-	}
 
 	isert_conn->login_req_dma = ib_dma_map_single(ib_dev,
 				isert_conn->login_req_buf,
@@ -1276,11 +1270,8 @@ isert_handle_text_cmd(struct isert_conn *isert_conn, struct isert_cmd *isert_cmd
 
 	if (payload_length) {
 		text_in = kzalloc(payload_length, GFP_KERNEL);
-		if (!text_in) {
-			isert_err("Unable to allocate text_in of payload_length: %u\n",
-				  payload_length);
+		if (!text_in)
 			return -ENOMEM;
-		}
 	}
 	cmd->text_in_ptr = text_in;
 
@@ -2307,10 +2298,9 @@ isert_setup_np(struct iscsi_np *np,
 	int ret;
 
 	isert_np = kzalloc(sizeof(struct isert_np), GFP_KERNEL);
-	if (!isert_np) {
-		isert_err("Unable to allocate struct isert_np\n");
+	if (!isert_np)
 		return -ENOMEM;
-	}
+
 	sema_init(&isert_np->sem, 0);
 	mutex_init(&isert_np->mutex);
 	INIT_LIST_HEAD(&isert_np->accepted);
@@ -2651,7 +2641,6 @@ static int __init isert_init(void)
 					WQ_UNBOUND | WQ_HIGHPRI, 0);
 	if (!isert_comp_wq) {
 		isert_err("Unable to allocate isert_comp_wq\n");
-		ret = -ENOMEM;
 		return -ENOMEM;
 	}
 
-- 
cgit v1.2.1


From 74226649f42d34a8ade2799bfb4f3941f4adfa95 Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leon@kernel.org>
Date: Thu, 3 Nov 2016 16:44:25 +0200
Subject: IB/ipoib: Remove and fix debug prints after allocation failure

The prints after [k|v][m|z|c]alloc() functions are not needed,
because in case of failure, allocator will print their internal
error prints anyway.

Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/ulp/ipoib/ipoib_cm.c   | 8 +-------
 drivers/infiniband/ulp/ipoib/ipoib_ib.c   | 5 +----
 drivers/infiniband/ulp/ipoib/ipoib_main.c | 5 +----
 3 files changed, 3 insertions(+), 15 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/ulp/ipoib/ipoib_cm.c b/drivers/infiniband/ulp/ipoib/ipoib_cm.c
index 4ad297d3de89..44f152e431cb 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_cm.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_cm.c
@@ -355,11 +355,8 @@ static int ipoib_cm_nonsrq_init_rx(struct net_device *dev, struct ib_cm_id *cm_i
 	int i;
 
 	rx->rx_ring = vzalloc(ipoib_recvq_size * sizeof *rx->rx_ring);
-	if (!rx->rx_ring) {
-		printk(KERN_WARNING "%s: failed to allocate CM non-SRQ ring (%d entries)\n",
-		       priv->ca->name, ipoib_recvq_size);
+	if (!rx->rx_ring)
 		return -ENOMEM;
-	}
 
 	t = kmalloc(sizeof *t, GFP_KERNEL);
 	if (!t) {
@@ -1133,7 +1130,6 @@ static int ipoib_cm_tx_init(struct ipoib_cm_tx *p, u32 qpn,
 	p->tx_ring = __vmalloc(ipoib_sendq_size * sizeof *p->tx_ring,
 			       GFP_NOIO, PAGE_KERNEL);
 	if (!p->tx_ring) {
-		ipoib_warn(priv, "failed to allocate tx ring\n");
 		ret = -ENOMEM;
 		goto err_tx;
 	}
@@ -1549,8 +1545,6 @@ static void ipoib_cm_create_srq(struct net_device *dev, int max_sge)
 
 	priv->cm.srq_ring = vzalloc(ipoib_recvq_size * sizeof *priv->cm.srq_ring);
 	if (!priv->cm.srq_ring) {
-		printk(KERN_WARNING "%s: failed to allocate CM SRQ ring (%d entries)\n",
-		       priv->ca->name, ipoib_recvq_size);
 		ib_destroy_srq(priv->cm.srq);
 		priv->cm.srq = NULL;
 		return;
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_ib.c b/drivers/infiniband/ulp/ipoib/ipoib_ib.c
index be11d5d5b8c1..43cf8b8a8d2e 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_ib.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_ib.c
@@ -418,11 +418,8 @@ static void ipoib_ib_handle_tx_wc(struct net_device *dev, struct ib_wc *wc)
 			   "(status=%d, wrid=%d vend_err %x)\n",
 			   wc->status, wr_id, wc->vendor_err);
 		qp_work = kzalloc(sizeof(*qp_work), GFP_ATOMIC);
-		if (!qp_work) {
-			ipoib_warn(priv, "%s Failed alloc ipoib_qp_state_validate for qp: 0x%x\n",
-				   __func__, priv->qp->qp_num);
+		if (!qp_work)
 			return;
-		}
 
 		INIT_WORK(&qp_work->work, ipoib_qp_state_validate_work);
 		qp_work->priv = priv;
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c
index 5636fc3da6b8..423b30dfe2d8 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_main.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c
@@ -1594,11 +1594,8 @@ int ipoib_dev_init(struct net_device *dev, struct ib_device *ca, int port)
 	/* Allocate RX/TX "rings" to hold queued skbs */
 	priv->rx_ring =	kzalloc(ipoib_recvq_size * sizeof *priv->rx_ring,
 				GFP_KERNEL);
-	if (!priv->rx_ring) {
-		printk(KERN_WARNING "%s: failed to allocate RX ring (%d entries)\n",
-		       ca->name, ipoib_recvq_size);
+	if (!priv->rx_ring)
 		goto out;
-	}
 
 	priv->tx_ring = vzalloc(ipoib_sendq_size * sizeof *priv->tx_ring);
 	if (!priv->tx_ring) {
-- 
cgit v1.2.1


From f73a1dbc45a58ea3ae5c3d60c71e58a8a9563914 Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leon@kernel.org>
Date: Mon, 21 Nov 2016 19:38:20 +0200
Subject: infiniband: remove WARN that is not kernel bug

On Mon, Nov 21, 2016 at 09:52:53AM -0700, Jason Gunthorpe wrote:
> On Mon, Nov 21, 2016 at 02:14:08PM +0200, Leon Romanovsky wrote:
> > >
> > > In ib_ucm_write function there is a wrong prefix:
> > >
> > > + pr_err_once("ucm_write: process %d (%s) tried to do something hinky\n",
> >
> > I did it intentionally to have the same errors for all flows.
>
> Lets actually use a good message too please?
>
>  pr_err_once("ucm_write: process %d (%s) changed security contexts after opening FD, this is not allowed.\n",
>
> Jason

>From 70f95b2d35aea42e5b97e7d27ab2f4e8effcbe67 Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leonro@mellanox.com>
Date: Mon, 21 Nov 2016 13:30:59 +0200
Subject: [PATCH rdma-next V2] IB/{core, qib}: Remove WARN that is not kernel bug

WARNINGs mean kernel bugs, in this case, they are placed
to mark programming errors and/or malicious attempts.

BUG/WARNs that are not kernel bugs hinder automated testing efforts.

Signed-off-by: Dmitry Vyukov <dvyukov@google.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/ucm.c            | 5 ++++-
 drivers/infiniband/core/ucma.c           | 5 ++++-
 drivers/infiniband/core/uverbs_main.c    | 5 ++++-
 drivers/infiniband/hw/qib/qib_file_ops.c | 5 ++++-
 4 files changed, 16 insertions(+), 4 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/core/ucm.c b/drivers/infiniband/core/ucm.c
index 7713ef089c3c..579f9a7f6283 100644
--- a/drivers/infiniband/core/ucm.c
+++ b/drivers/infiniband/core/ucm.c
@@ -1104,8 +1104,11 @@ static ssize_t ib_ucm_write(struct file *filp, const char __user *buf,
 	struct ib_ucm_cmd_hdr hdr;
 	ssize_t result;
 
-	if (WARN_ON_ONCE(!ib_safe_file_access(filp)))
+	if (!ib_safe_file_access(filp)) {
+		pr_err_once("ucm_write: process %d (%s) changed security contexts after opening file descriptor, this is not allowed.\n",
+			    task_tgid_vnr(current), current->comm);
 		return -EACCES;
+	}
 
 	if (len < sizeof(hdr))
 		return -EINVAL;
diff --git a/drivers/infiniband/core/ucma.c b/drivers/infiniband/core/ucma.c
index 9520154f1d7c..e12f8faf8c23 100644
--- a/drivers/infiniband/core/ucma.c
+++ b/drivers/infiniband/core/ucma.c
@@ -1584,8 +1584,11 @@ static ssize_t ucma_write(struct file *filp, const char __user *buf,
 	struct rdma_ucm_cmd_hdr hdr;
 	ssize_t ret;
 
-	if (WARN_ON_ONCE(!ib_safe_file_access(filp)))
+	if (!ib_safe_file_access(filp)) {
+		pr_err_once("ucma_write: process %d (%s) changed security contexts after opening file descriptor, this is not allowed.\n",
+			    task_tgid_vnr(current), current->comm);
 		return -EACCES;
+	}
 
 	if (len < sizeof(hdr))
 		return -EINVAL;
diff --git a/drivers/infiniband/core/uverbs_main.c b/drivers/infiniband/core/uverbs_main.c
index 0012fa58c105..3efec031c253 100644
--- a/drivers/infiniband/core/uverbs_main.c
+++ b/drivers/infiniband/core/uverbs_main.c
@@ -749,8 +749,11 @@ static ssize_t ib_uverbs_write(struct file *filp, const char __user *buf,
 	int srcu_key;
 	ssize_t ret;
 
-	if (WARN_ON_ONCE(!ib_safe_file_access(filp)))
+	if (!ib_safe_file_access(filp)) {
+		pr_err_once("uverbs_write: process %d (%s) changed security contexts after opening file descriptor, this is not allowed.\n",
+			    task_tgid_vnr(current), current->comm);
 		return -EACCES;
+	}
 
 	if (count < sizeof hdr)
 		return -EINVAL;
diff --git a/drivers/infiniband/hw/qib/qib_file_ops.c b/drivers/infiniband/hw/qib/qib_file_ops.c
index 382466a90da7..2d1eacf1dfed 100644
--- a/drivers/infiniband/hw/qib/qib_file_ops.c
+++ b/drivers/infiniband/hw/qib/qib_file_ops.c
@@ -2066,8 +2066,11 @@ static ssize_t qib_write(struct file *fp, const char __user *data,
 	ssize_t ret = 0;
 	void *dest;
 
-	if (WARN_ON_ONCE(!ib_safe_file_access(fp)))
+	if (!ib_safe_file_access(fp)) {
+		pr_err_once("qib_write: process %d (%s) changed security contexts after opening file descriptor, this is not allowed.\n",
+			    task_tgid_vnr(current), current->comm);
 		return -EACCES;
+	}
 
 	if (count < sizeof(cmd.type)) {
 		ret = -EINVAL;
-- 
cgit v1.2.1


From 9eefa953f475897636f9290d833e45009d58a55d Mon Sep 17 00:00:00 2001
From: Lijun Ou <oulijun@huawei.com>
Date: Wed, 23 Nov 2016 19:40:59 +0000
Subject: IB/hns: Add the interface for querying QP1

In old code, It only added the interface for querying non-specific
QP. This patch mainly adds an interface for querying QP1.

Signed-off-by: Lijun Ou <oulijun@huawei.com>
Reviewed-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Signed-off-by: Salil Mehta  <salil.mehta@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_hw_v1.c | 83 +++++++++++++++++++++++++++++-
 drivers/infiniband/hw/hns/hns_roce_hw_v1.h |  6 ++-
 2 files changed, 86 insertions(+), 3 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
index 71232e5fabf6..7f2c26bd6232 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
@@ -2630,8 +2630,78 @@ static int hns_roce_v1_query_qpc(struct hns_roce_dev *hr_dev,
 	return ret;
 }
 
-int hns_roce_v1_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr,
-			 int qp_attr_mask, struct ib_qp_init_attr *qp_init_attr)
+static int hns_roce_v1_q_sqp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr,
+			     int qp_attr_mask,
+			     struct ib_qp_init_attr *qp_init_attr)
+{
+	struct hns_roce_dev *hr_dev = to_hr_dev(ibqp->device);
+	struct hns_roce_qp *hr_qp = to_hr_qp(ibqp);
+	struct hns_roce_sqp_context context;
+	u32 addr;
+
+	mutex_lock(&hr_qp->mutex);
+
+	if (hr_qp->state == IB_QPS_RESET) {
+		qp_attr->qp_state = IB_QPS_RESET;
+		goto done;
+	}
+
+	addr = ROCEE_QP1C_CFG0_0_REG +
+		hr_qp->port * sizeof(struct hns_roce_sqp_context);
+	context.qp1c_bytes_4 = roce_read(hr_dev, addr);
+	context.sq_rq_bt_l = roce_read(hr_dev, addr + 1);
+	context.qp1c_bytes_12 = roce_read(hr_dev, addr + 2);
+	context.qp1c_bytes_16 = roce_read(hr_dev, addr + 3);
+	context.qp1c_bytes_20 = roce_read(hr_dev, addr + 4);
+	context.cur_rq_wqe_ba_l = roce_read(hr_dev, addr + 5);
+	context.qp1c_bytes_28 = roce_read(hr_dev, addr + 6);
+	context.qp1c_bytes_32 = roce_read(hr_dev, addr + 7);
+	context.cur_sq_wqe_ba_l = roce_read(hr_dev, addr + 8);
+	context.qp1c_bytes_40 = roce_read(hr_dev, addr + 9);
+
+	hr_qp->state = roce_get_field(context.qp1c_bytes_4,
+				      QP1C_BYTES_4_QP_STATE_M,
+				      QP1C_BYTES_4_QP_STATE_S);
+	qp_attr->qp_state	= hr_qp->state;
+	qp_attr->path_mtu	= IB_MTU_256;
+	qp_attr->path_mig_state	= IB_MIG_ARMED;
+	qp_attr->qkey		= QKEY_VAL;
+	qp_attr->rq_psn		= 0;
+	qp_attr->sq_psn		= 0;
+	qp_attr->dest_qp_num	= 1;
+	qp_attr->qp_access_flags = 6;
+
+	qp_attr->pkey_index = roce_get_field(context.qp1c_bytes_20,
+					     QP1C_BYTES_20_PKEY_IDX_M,
+					     QP1C_BYTES_20_PKEY_IDX_S);
+	qp_attr->port_num = hr_qp->port + 1;
+	qp_attr->sq_draining = 0;
+	qp_attr->max_rd_atomic = 0;
+	qp_attr->max_dest_rd_atomic = 0;
+	qp_attr->min_rnr_timer = 0;
+	qp_attr->timeout = 0;
+	qp_attr->retry_cnt = 0;
+	qp_attr->rnr_retry = 0;
+	qp_attr->alt_timeout = 0;
+
+done:
+	qp_attr->cur_qp_state = qp_attr->qp_state;
+	qp_attr->cap.max_recv_wr = hr_qp->rq.wqe_cnt;
+	qp_attr->cap.max_recv_sge = hr_qp->rq.max_gs;
+	qp_attr->cap.max_send_wr = hr_qp->sq.wqe_cnt;
+	qp_attr->cap.max_send_sge = hr_qp->sq.max_gs;
+	qp_attr->cap.max_inline_data = 0;
+	qp_init_attr->cap = qp_attr->cap;
+	qp_init_attr->create_flags = 0;
+
+	mutex_unlock(&hr_qp->mutex);
+
+	return 0;
+}
+
+static int hns_roce_v1_q_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr,
+			    int qp_attr_mask,
+			    struct ib_qp_init_attr *qp_init_attr)
 {
 	struct hns_roce_dev *hr_dev = to_hr_dev(ibqp->device);
 	struct hns_roce_qp *hr_qp = to_hr_qp(ibqp);
@@ -2767,6 +2837,15 @@ out:
 	return ret;
 }
 
+int hns_roce_v1_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr,
+			 int qp_attr_mask, struct ib_qp_init_attr *qp_init_attr)
+{
+	struct hns_roce_qp *hr_qp = to_hr_qp(ibqp);
+
+	return hr_qp->doorbell_qpn <= 1 ?
+		hns_roce_v1_q_sqp(ibqp, qp_attr, qp_attr_mask, qp_init_attr) :
+		hns_roce_v1_q_qp(ibqp, qp_attr, qp_attr_mask, qp_init_attr);
+}
 static void hns_roce_v1_destroy_qp_common(struct hns_roce_dev *hr_dev,
 					  struct hns_roce_qp *hr_qp,
 					  int is_user)
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.h b/drivers/infiniband/hw/hns/hns_roce_hw_v1.h
index 539b0a3b92b0..2e1878bf4836 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.h
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.h
@@ -480,13 +480,17 @@ struct hns_roce_sqp_context {
 	u32 qp1c_bytes_12;
 	u32 qp1c_bytes_16;
 	u32 qp1c_bytes_20;
-	u32 qp1c_bytes_28;
 	u32 cur_rq_wqe_ba_l;
+	u32 qp1c_bytes_28;
 	u32 qp1c_bytes_32;
 	u32 cur_sq_wqe_ba_l;
 	u32 qp1c_bytes_40;
 };
 
+#define QP1C_BYTES_4_QP_STATE_S 0
+#define QP1C_BYTES_4_QP_STATE_M   \
+	(((1UL << 3) - 1) << QP1C_BYTES_4_QP_STATE_S)
+
 #define QP1C_BYTES_4_SQ_WQE_SHIFT_S 8
 #define QP1C_BYTES_4_SQ_WQE_SHIFT_M   \
 	(((1UL << 4) - 1) << QP1C_BYTES_4_SQ_WQE_SHIFT_S)
-- 
cgit v1.2.1


From 8f3e9f3ea0871a988ec9540a4cde48a20c1cb98b Mon Sep 17 00:00:00 2001
From: "Wei Hu (Xavier)" <xavier.huwei@huawei.com>
Date: Wed, 23 Nov 2016 19:41:00 +0000
Subject: IB/hns: Add code for refreshing CQ CI using TPTR

This patch added the code for refreshing CQ CI using TPTR in hip06
SoC.

We will send a doorbell to hardware for refreshing CQ CI when user
succeed to poll a cqe. But it will be failed if the doorbell has
been blocked. So hardware will read a special buffer called TPTR
to get the lastest CI value when the cq is almost full.

This patch support the special CI buffer as follows:
a) Alloc the memory for TPTR in the hns_roce_tptr_init function and
   free it in hns_roce_tptr_free function, these two functions will
   be called in probe function and in the remove function.
b) Add the code for computing offset(every cq need 2 bytes) and
   write the dma addr to every cq context to notice hardware in the
   function named hns_roce_v1_write_cqc.
c) Add code for mapping TPTR buffer to user space in function named
   hns_roce_mmap. The mapping distinguish TPTR and UAR of user mode
   by vm_pgoff(0: UAR, 1: TPTR, others:invaild) in hip06.
d) Alloc the code for refreshing CQ CI using TPTR in the function
   named hns_roce_v1_poll_cq.
e) Add some variable definitions to the related structure.

Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Signed-off-by: Dongdong Huang(Donald) <hdd.huang@huawei.com>
Signed-off-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Salil Mehta  <salil.mehta@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_common.h |  2 -
 drivers/infiniband/hw/hns/hns_roce_cq.c     |  9 ++++
 drivers/infiniband/hw/hns/hns_roce_device.h |  6 ++-
 drivers/infiniband/hw/hns/hns_roce_hw_v1.c  | 79 +++++++++++++++++++++++++----
 drivers/infiniband/hw/hns/hns_roce_hw_v1.h  |  9 ++++
 drivers/infiniband/hw/hns/hns_roce_main.c   | 13 +++--
 6 files changed, 103 insertions(+), 15 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hns/hns_roce_common.h b/drivers/infiniband/hw/hns/hns_roce_common.h
index 297016103aa7..0dcb620e91dd 100644
--- a/drivers/infiniband/hw/hns/hns_roce_common.h
+++ b/drivers/infiniband/hw/hns/hns_roce_common.h
@@ -253,8 +253,6 @@
 #define ROCEE_VENDOR_ID_REG			0x0
 #define ROCEE_VENDOR_PART_ID_REG		0x4
 
-#define ROCEE_HW_VERSION_REG			0x8
-
 #define ROCEE_SYS_IMAGE_GUID_L_REG		0xC
 #define ROCEE_SYS_IMAGE_GUID_H_REG		0x10
 
diff --git a/drivers/infiniband/hw/hns/hns_roce_cq.c b/drivers/infiniband/hw/hns/hns_roce_cq.c
index 097365932b09..5dc8d92e79fd 100644
--- a/drivers/infiniband/hw/hns/hns_roce_cq.c
+++ b/drivers/infiniband/hw/hns/hns_roce_cq.c
@@ -349,6 +349,15 @@ struct ib_cq *hns_roce_ib_create_cq(struct ib_device *ib_dev,
 		goto err_mtt;
 	}
 
+	/*
+	 * For the QP created by kernel space, tptr value should be initialized
+	 * to zero; For the QP created by user space, it will cause synchronous
+	 * problems if tptr is set to zero here, so we initialze it in user
+	 * space.
+	 */
+	if (!context)
+		*hr_cq->tptr_addr = 0;
+
 	/* Get created cq handler and carry out event */
 	hr_cq->comp = hns_roce_ib_cq_comp;
 	hr_cq->event = hns_roce_ib_cq_event;
diff --git a/drivers/infiniband/hw/hns/hns_roce_device.h b/drivers/infiniband/hw/hns/hns_roce_device.h
index 341731553a60..7242b1438873 100644
--- a/drivers/infiniband/hw/hns/hns_roce_device.h
+++ b/drivers/infiniband/hw/hns/hns_roce_device.h
@@ -37,6 +37,8 @@
 
 #define DRV_NAME "hns_roce"
 
+#define HNS_ROCE_HW_VER1	('h' << 24 | 'i' << 16 | '0' << 8 | '6')
+
 #define MAC_ADDR_OCTET_NUM			6
 #define HNS_ROCE_MAX_MSG_LEN			0x80000000
 
@@ -296,7 +298,7 @@ struct hns_roce_cq {
 	u32				cq_depth;
 	u32				cons_index;
 	void __iomem			*cq_db_l;
-	void __iomem			*tptr_addr;
+	u16				*tptr_addr;
 	unsigned long			cqn;
 	u32				vector;
 	atomic_t			refcount;
@@ -553,6 +555,8 @@ struct hns_roce_dev {
 
 	int			cmd_mod;
 	int			loop_idc;
+	dma_addr_t		tptr_dma_addr; /*only for hw v1*/
+	u32			tptr_size; /*only for hw v1*/
 	struct hns_roce_hw	*hw;
 };
 
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
index 7f2c26bd6232..957f0de866cd 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
@@ -849,6 +849,45 @@ static void hns_roce_bt_free(struct hns_roce_dev *hr_dev)
 		priv->bt_table.qpc_buf.buf, priv->bt_table.qpc_buf.map);
 }
 
+static int hns_roce_tptr_init(struct hns_roce_dev *hr_dev)
+{
+	struct device *dev = &hr_dev->pdev->dev;
+	struct hns_roce_buf_list *tptr_buf;
+	struct hns_roce_v1_priv *priv;
+
+	priv = (struct hns_roce_v1_priv *)hr_dev->hw->priv;
+	tptr_buf = &priv->tptr_table.tptr_buf;
+
+	/*
+	 * This buffer will be used for CQ's tptr(tail pointer), also
+	 * named ci(customer index). Every CQ will use 2 bytes to save
+	 * cqe ci in hip06. Hardware will read this area to get new ci
+	 * when the queue is almost full.
+	 */
+	tptr_buf->buf = dma_alloc_coherent(dev, HNS_ROCE_V1_TPTR_BUF_SIZE,
+					   &tptr_buf->map, GFP_KERNEL);
+	if (!tptr_buf->buf)
+		return -ENOMEM;
+
+	hr_dev->tptr_dma_addr = tptr_buf->map;
+	hr_dev->tptr_size = HNS_ROCE_V1_TPTR_BUF_SIZE;
+
+	return 0;
+}
+
+static void hns_roce_tptr_free(struct hns_roce_dev *hr_dev)
+{
+	struct device *dev = &hr_dev->pdev->dev;
+	struct hns_roce_buf_list *tptr_buf;
+	struct hns_roce_v1_priv *priv;
+
+	priv = (struct hns_roce_v1_priv *)hr_dev->hw->priv;
+	tptr_buf = &priv->tptr_table.tptr_buf;
+
+	dma_free_coherent(dev, HNS_ROCE_V1_TPTR_BUF_SIZE,
+			  tptr_buf->buf, tptr_buf->map);
+}
+
 /**
  * hns_roce_v1_reset - reset RoCE
  * @hr_dev: RoCE device struct pointer
@@ -906,12 +945,11 @@ void hns_roce_v1_profile(struct hns_roce_dev *hr_dev)
 	hr_dev->vendor_id = le32_to_cpu(roce_read(hr_dev, ROCEE_VENDOR_ID_REG));
 	hr_dev->vendor_part_id = le32_to_cpu(roce_read(hr_dev,
 					     ROCEE_VENDOR_PART_ID_REG));
-	hr_dev->hw_rev = le32_to_cpu(roce_read(hr_dev, ROCEE_HW_VERSION_REG));
-
 	hr_dev->sys_image_guid = le32_to_cpu(roce_read(hr_dev,
 					     ROCEE_SYS_IMAGE_GUID_L_REG)) |
 				((u64)le32_to_cpu(roce_read(hr_dev,
 					    ROCEE_SYS_IMAGE_GUID_H_REG)) << 32);
+	hr_dev->hw_rev		= HNS_ROCE_HW_VER1;
 
 	caps->num_qps		= HNS_ROCE_V1_MAX_QP_NUM;
 	caps->max_wqes		= HNS_ROCE_V1_MAX_WQE_NUM;
@@ -1009,8 +1047,17 @@ int hns_roce_v1_init(struct hns_roce_dev *hr_dev)
 		goto error_failed_bt_init;
 	}
 
+	ret = hns_roce_tptr_init(hr_dev);
+	if (ret) {
+		dev_err(dev, "tptr init failed!\n");
+		goto error_failed_tptr_init;
+	}
+
 	return 0;
 
+error_failed_tptr_init:
+	hns_roce_bt_free(hr_dev);
+
 error_failed_bt_init:
 	hns_roce_port_enable(hr_dev, HNS_ROCE_PORT_DOWN);
 	hns_roce_raq_free(hr_dev);
@@ -1022,6 +1069,7 @@ error_failed_raq_init:
 
 void hns_roce_v1_exit(struct hns_roce_dev *hr_dev)
 {
+	hns_roce_tptr_free(hr_dev);
 	hns_roce_bt_free(hr_dev);
 	hns_roce_port_enable(hr_dev, HNS_ROCE_PORT_DOWN);
 	hns_roce_raq_free(hr_dev);
@@ -1339,14 +1387,21 @@ void hns_roce_v1_write_cqc(struct hns_roce_dev *hr_dev,
 			   dma_addr_t dma_handle, int nent, u32 vector)
 {
 	struct hns_roce_cq_context *cq_context = NULL;
-	void __iomem *tptr_addr;
+	struct hns_roce_buf_list *tptr_buf;
+	struct hns_roce_v1_priv *priv;
+	dma_addr_t tptr_dma_addr;
+	int offset;
+
+	priv = (struct hns_roce_v1_priv *)hr_dev->hw->priv;
+	tptr_buf = &priv->tptr_table.tptr_buf;
 
 	cq_context = mb_buf;
 	memset(cq_context, 0, sizeof(*cq_context));
 
-	tptr_addr = 0;
-	hr_dev->priv_addr = tptr_addr;
-	hr_cq->tptr_addr = tptr_addr;
+	/* Get the tptr for this CQ. */
+	offset = hr_cq->cqn * HNS_ROCE_V1_TPTR_ENTRY_SIZE;
+	tptr_dma_addr = tptr_buf->map + offset;
+	hr_cq->tptr_addr = (u16 *)(tptr_buf->buf + offset);
 
 	/* Register cq_context members */
 	roce_set_field(cq_context->cqc_byte_4,
@@ -1390,10 +1445,10 @@ void hns_roce_v1_write_cqc(struct hns_roce_dev *hr_dev,
 	roce_set_field(cq_context->cqc_byte_20,
 		       CQ_CONTEXT_CQC_BYTE_20_CQE_TPTR_ADDR_H_M,
 		       CQ_CONTEXT_CQC_BYTE_20_CQE_TPTR_ADDR_H_S,
-		       (u64)tptr_addr >> 44);
+		       tptr_dma_addr >> 44);
 	cq_context->cqc_byte_20 = cpu_to_le32(cq_context->cqc_byte_20);
 
-	cq_context->cqe_tptr_addr_l = (u32)((u64)tptr_addr >> 12);
+	cq_context->cqe_tptr_addr_l = (u32)(tptr_dma_addr >> 12);
 
 	roce_set_field(cq_context->cqc_byte_32,
 		       CQ_CONTEXT_CQC_BYTE_32_CUR_CQE_BA1_H_M,
@@ -1659,8 +1714,14 @@ int hns_roce_v1_poll_cq(struct ib_cq *ibcq, int num_entries, struct ib_wc *wc)
 			break;
 	}
 
-	if (npolled)
+	if (npolled) {
+		*hr_cq->tptr_addr = hr_cq->cons_index &
+			((hr_cq->cq_depth << 1) - 1);
+
+		/* Memroy barrier */
+		wmb();
 		hns_roce_v1_cq_set_ci(hr_cq, hr_cq->cons_index);
+	}
 
 	spin_unlock_irqrestore(&hr_cq->lock, flags);
 
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.h b/drivers/infiniband/hw/hns/hns_roce_hw_v1.h
index 2e1878bf4836..6004c7f39542 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.h
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.h
@@ -104,6 +104,10 @@
 
 #define HNS_ROCE_BT_RSV_BUF_SIZE			(1 << 17)
 
+#define HNS_ROCE_V1_TPTR_ENTRY_SIZE			2
+#define HNS_ROCE_V1_TPTR_BUF_SIZE	\
+	(HNS_ROCE_V1_TPTR_ENTRY_SIZE * HNS_ROCE_V1_MAX_CQ_NUM)
+
 #define HNS_ROCE_ODB_POLL_MODE				0
 
 #define HNS_ROCE_SDB_NORMAL_MODE			0
@@ -983,10 +987,15 @@ struct hns_roce_bt_table {
 	struct hns_roce_buf_list cqc_buf;
 };
 
+struct hns_roce_tptr_table {
+	struct hns_roce_buf_list tptr_buf;
+};
+
 struct hns_roce_v1_priv {
 	struct hns_roce_db_table  db_table;
 	struct hns_roce_raq_table raq_table;
 	struct hns_roce_bt_table  bt_table;
+	struct hns_roce_tptr_table tptr_table;
 };
 
 int hns_dsaf_roce_reset(struct fwnode_handle *dsaf_fwnode, bool dereset);
diff --git a/drivers/infiniband/hw/hns/hns_roce_main.c b/drivers/infiniband/hw/hns/hns_roce_main.c
index 764e35a54457..67701719bad1 100644
--- a/drivers/infiniband/hw/hns/hns_roce_main.c
+++ b/drivers/infiniband/hw/hns/hns_roce_main.c
@@ -549,6 +549,8 @@ static int hns_roce_dealloc_ucontext(struct ib_ucontext *ibcontext)
 static int hns_roce_mmap(struct ib_ucontext *context,
 			 struct vm_area_struct *vma)
 {
+	struct hns_roce_dev *hr_dev = to_hr_dev(context->device);
+
 	if (((vma->vm_end - vma->vm_start) % PAGE_SIZE) != 0)
 		return -EINVAL;
 
@@ -558,10 +560,15 @@ static int hns_roce_mmap(struct ib_ucontext *context,
 				       to_hr_ucontext(context)->uar.pfn,
 				       PAGE_SIZE, vma->vm_page_prot))
 			return -EAGAIN;
-
-	} else {
+	} else if (vma->vm_pgoff == 1 && hr_dev->hw_rev == HNS_ROCE_HW_VER1) {
+		/* vm_pgoff: 1 -- TPTR */
+		if (io_remap_pfn_range(vma, vma->vm_start,
+				       hr_dev->tptr_dma_addr >> PAGE_SHIFT,
+				       hr_dev->tptr_size,
+				       vma->vm_page_prot))
+			return -EAGAIN;
+	} else
 		return -EINVAL;
-	}
 
 	return 0;
 }
-- 
cgit v1.2.1


From 8d497eb0f325e98fa87b59ac86069deea64e8d5d Mon Sep 17 00:00:00 2001
From: "Wei Hu (Xavier)" <xavier.huwei@huawei.com>
Date: Wed, 23 Nov 2016 19:41:01 +0000
Subject: IB/hns: Optimize the logic of allocating memory using APIs

This patch modified the logic of allocating memory using APIs in
hns RoCE driver. We used kcalloc instead of kmalloc_array and
bitmap_zero. And When kcalloc failed, call vzalloc to alloc
memory.

Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Signed-off-by: Ping Zhang <zhangping5@huawei.com>
Signed-off-by: Salil Mehta  <salil.mehta@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_mr.c | 16 +++++++++-------
 1 file changed, 9 insertions(+), 7 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hns/hns_roce_mr.c b/drivers/infiniband/hw/hns/hns_roce_mr.c
index fb87883ead34..d87d189cb3d8 100644
--- a/drivers/infiniband/hw/hns/hns_roce_mr.c
+++ b/drivers/infiniband/hw/hns/hns_roce_mr.c
@@ -137,11 +137,13 @@ static int hns_roce_buddy_init(struct hns_roce_buddy *buddy, int max_order)
 
 	for (i = 0; i <= buddy->max_order; ++i) {
 		s = BITS_TO_LONGS(1 << (buddy->max_order - i));
-		buddy->bits[i] = kmalloc_array(s, sizeof(long), GFP_KERNEL);
-		if (!buddy->bits[i])
-			goto err_out_free;
-
-		bitmap_zero(buddy->bits[i], 1 << (buddy->max_order - i));
+		buddy->bits[i] = kcalloc(s, sizeof(long), GFP_KERNEL |
+					 __GFP_NOWARN);
+		if (!buddy->bits[i]) {
+			buddy->bits[i] = vzalloc(s * sizeof(long));
+			if (!buddy->bits[i])
+				goto err_out_free;
+		}
 	}
 
 	set_bit(0, buddy->bits[buddy->max_order]);
@@ -151,7 +153,7 @@ static int hns_roce_buddy_init(struct hns_roce_buddy *buddy, int max_order)
 
 err_out_free:
 	for (i = 0; i <= buddy->max_order; ++i)
-		kfree(buddy->bits[i]);
+		kvfree(buddy->bits[i]);
 
 err_out:
 	kfree(buddy->bits);
@@ -164,7 +166,7 @@ static void hns_roce_buddy_cleanup(struct hns_roce_buddy *buddy)
 	int i;
 
 	for (i = 0; i <= buddy->max_order; ++i)
-		kfree(buddy->bits[i]);
+		kvfree(buddy->bits[i]);
 
 	kfree(buddy->bits);
 	kfree(buddy->num_free);
-- 
cgit v1.2.1


From 543bfe6c3c1602d1781504047795a4266076233b Mon Sep 17 00:00:00 2001
From: Lijun Ou <oulijun@huawei.com>
Date: Wed, 23 Nov 2016 19:41:02 +0000
Subject: IB/hns: add self loopback for CM

This patch mainly adds self loopback support for CM.

Signed-off-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Peter Chen <luck.chen@huawei.com>
Reviewed-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Signed-off-by: Salil Mehta  <salil.mehta@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_hw_v1.c | 11 +++++++++++
 drivers/infiniband/hw/hns/hns_roce_hw_v1.h |  2 ++
 2 files changed, 13 insertions(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
index 957f0de866cd..48c71e839c6a 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
@@ -32,6 +32,7 @@
 
 #include <linux/platform_device.h>
 #include <linux/acpi.h>
+#include <linux/etherdevice.h>
 #include <rdma/ib_umem.h>
 #include "hns_roce_common.h"
 #include "hns_roce_device.h"
@@ -72,6 +73,8 @@ int hns_roce_v1_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 	int nreq = 0;
 	u32 ind = 0;
 	int ret = 0;
+	u8 *smac;
+	int loopback;
 
 	if (unlikely(ibqp->qp_type != IB_QPT_GSI &&
 		ibqp->qp_type != IB_QPT_RC)) {
@@ -129,6 +132,14 @@ int hns_roce_v1_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 				       UD_SEND_WQE_U32_8_DMAC_5_M,
 				       UD_SEND_WQE_U32_8_DMAC_5_S,
 				       ah->av.mac[5]);
+
+			smac = (u8 *)hr_dev->dev_addr[qp->port];
+			loopback = ether_addr_equal_unaligned(ah->av.mac,
+							      smac) ? 1 : 0;
+			roce_set_bit(ud_sq_wqe->u32_8,
+				     UD_SEND_WQE_U32_8_LOOPBACK_INDICATOR_S,
+				     loopback);
+
 			roce_set_field(ud_sq_wqe->u32_8,
 				       UD_SEND_WQE_U32_8_OPERATION_TYPE_M,
 				       UD_SEND_WQE_U32_8_OPERATION_TYPE_S,
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.h b/drivers/infiniband/hw/hns/hns_roce_hw_v1.h
index 6004c7f39542..cf28f1b6492c 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.h
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.h
@@ -440,6 +440,8 @@ struct hns_roce_ud_send_wqe {
 #define UD_SEND_WQE_U32_8_DMAC_5_M   \
 	(((1UL << 8) - 1) << UD_SEND_WQE_U32_8_DMAC_5_S)
 
+#define UD_SEND_WQE_U32_8_LOOPBACK_INDICATOR_S 22
+
 #define UD_SEND_WQE_U32_8_OPERATION_TYPE_S 16
 #define UD_SEND_WQE_U32_8_OPERATION_TYPE_M   \
 	(((1UL << 4) - 1) << UD_SEND_WQE_U32_8_OPERATION_TYPE_S)
-- 
cgit v1.2.1


From 80596c6717a7b0e9969ffbd03c7e1dd33112f7e9 Mon Sep 17 00:00:00 2001
From: Lijun Ou <oulijun@huawei.com>
Date: Wed, 23 Nov 2016 19:41:03 +0000
Subject: IB/hns: Modify the condition of notifying hardware loopback

This patch modified the condition of notifying hardware loopback.

In hip06, RoCE Engine has several ports, one QP is related
to one port. hardware only support loopback in the same port,
not in the different ports.

So, If QP related to port N, the dmac in the QP context equals
the smac of the local port N or the loop_idc is 1, we should
set loopback bit in QP context to notify hardware.

Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Signed-off-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Salil Mehta  <salil.mehta@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_hw_v1.c | 24 +++++++-----------------
 1 file changed, 7 insertions(+), 17 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
index 48c71e839c6a..c1a9dac25d7b 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
@@ -2244,24 +2244,14 @@ static int hns_roce_v1_m_qp(struct ib_qp *ibqp, const struct ib_qp_attr *attr,
 			     QP_CONTEXT_QPC_BYTE_32_SIGNALING_TYPE_S,
 			     hr_qp->sq_signal_bits);
 
-		for (port = 0; port < hr_dev->caps.num_ports; port++) {
-			smac = (u8 *)hr_dev->dev_addr[port];
-			dev_dbg(dev, "smac: %2x: %2x: %2x: %2x: %2x: %2x\n",
-				smac[0], smac[1], smac[2], smac[3], smac[4],
-				smac[5]);
-			if ((dmac[0] == smac[0]) && (dmac[1] == smac[1]) &&
-			    (dmac[2] == smac[2]) && (dmac[3] == smac[3]) &&
-			    (dmac[4] == smac[4]) && (dmac[5] == smac[5])) {
-				roce_set_bit(context->qpc_bytes_32,
-				    QP_CONTEXT_QPC_BYTE_32_LOOPBACK_INDICATOR_S,
-				    1);
-				break;
-			}
-		}
-
-		if (hr_dev->loop_idc == 0x1)
+		port = (attr_mask & IB_QP_PORT) ? (attr->port_num - 1) :
+			hr_qp->port;
+		smac = (u8 *)hr_dev->dev_addr[port];
+		/* when dmac equals smac or loop_idc is 1, it should loopback */
+		if (ether_addr_equal_unaligned(dmac, smac) ||
+		    hr_dev->loop_idc == 0x1)
 			roce_set_bit(context->qpc_bytes_32,
-				QP_CONTEXT_QPC_BYTE_32_LOOPBACK_INDICATOR_S, 1);
+			      QP_CONTEXT_QPC_BYTE_32_LOOPBACK_INDICATOR_S, 1);
 
 		roce_set_bit(context->qpc_bytes_32,
 			     QP_CONTEXT_QPC_BYTE_32_GLOBAL_HEADER_S,
-- 
cgit v1.2.1


From 1dec243ac00c6e10217c994c5a500bc85fe0f649 Mon Sep 17 00:00:00 2001
From: Lijun Ou <oulijun@huawei.com>
Date: Wed, 23 Nov 2016 19:41:04 +0000
Subject: IB/hns: Fix the bug for qp state in hns_roce_v1_m_qp()

In old code, the value of qp state from qpc was assigned for
attr->qp_state. The value may be an error while attr_mask &
IB_QP_STATE is zero.

Signed-off-by: Lijun Ou <oulijun@huawei.com>
Reviewed-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Signed-off-by: Salil Mehta  <salil.mehta@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_hw_v1.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
index c1a9dac25d7b..528cb84975b4 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
@@ -2571,7 +2571,7 @@ static int hns_roce_v1_m_qp(struct ib_qp *ibqp, const struct ib_qp_attr *attr,
 	/* Every status migrate must change state */
 	roce_set_field(context->qpc_bytes_144,
 		       QP_CONTEXT_QPC_BYTES_144_QP_STATE_M,
-		       QP_CONTEXT_QPC_BYTES_144_QP_STATE_S, attr->qp_state);
+		       QP_CONTEXT_QPC_BYTES_144_QP_STATE_S, new_state);
 
 	/* SW pass context to HW */
 	ret = hns_roce_v1_qp_modify(hr_dev, &hr_qp->mtt,
-- 
cgit v1.2.1


From 6b877c32bc40bfa9fa13659ac4e1413b67ff1c3d Mon Sep 17 00:00:00 2001
From: "Wei Hu (Xavier)" <xavier.huwei@huawei.com>
Date: Wed, 23 Nov 2016 19:41:05 +0000
Subject: IB/hns: Modify the macro for the timeout when cmd process

This patch modified the macro for the timeout when cmd is
processing as follows:
Before modification:
 enum {
	HNS_ROCE_CMD_TIME_CLASS_A       = 10000,
	HNS_ROCE_CMD_TIME_CLASS_B       = 10000,
	HNS_ROCE_CMD_TIME_CLASS_C       = 10000,
 };
After modification:
 #define HNS_ROCE_CMD_TIMEOUT_MSECS	10000

Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Signed-off-by: Salil Mehta  <salil.mehta@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_cmd.h   | 7 +------
 drivers/infiniband/hw/hns/hns_roce_cq.c    | 4 ++--
 drivers/infiniband/hw/hns/hns_roce_hw_v1.c | 8 ++++----
 drivers/infiniband/hw/hns/hns_roce_mr.c    | 4 ++--
 4 files changed, 9 insertions(+), 14 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hns/hns_roce_cmd.h b/drivers/infiniband/hw/hns/hns_roce_cmd.h
index e3997d312c55..ed14ad3d06ef 100644
--- a/drivers/infiniband/hw/hns/hns_roce_cmd.h
+++ b/drivers/infiniband/hw/hns/hns_roce_cmd.h
@@ -34,6 +34,7 @@
 #define _HNS_ROCE_CMD_H
 
 #define HNS_ROCE_MAILBOX_SIZE		4096
+#define HNS_ROCE_CMD_TIMEOUT_MSECS	10000
 
 enum {
 	/* TPT commands */
@@ -57,12 +58,6 @@ enum {
 	HNS_ROCE_CMD_QUERY_QP		= 0x22,
 };
 
-enum {
-	HNS_ROCE_CMD_TIME_CLASS_A	= 10000,
-	HNS_ROCE_CMD_TIME_CLASS_B	= 10000,
-	HNS_ROCE_CMD_TIME_CLASS_C	= 10000,
-};
-
 struct hns_roce_cmd_mailbox {
 	void		       *buf;
 	dma_addr_t		dma;
diff --git a/drivers/infiniband/hw/hns/hns_roce_cq.c b/drivers/infiniband/hw/hns/hns_roce_cq.c
index 5dc8d92e79fd..461a27330e83 100644
--- a/drivers/infiniband/hw/hns/hns_roce_cq.c
+++ b/drivers/infiniband/hw/hns/hns_roce_cq.c
@@ -77,7 +77,7 @@ static int hns_roce_sw2hw_cq(struct hns_roce_dev *dev,
 			     unsigned long cq_num)
 {
 	return hns_roce_cmd_mbox(dev, mailbox->dma, 0, cq_num, 0,
-			    HNS_ROCE_CMD_SW2HW_CQ, HNS_ROCE_CMD_TIME_CLASS_A);
+			    HNS_ROCE_CMD_SW2HW_CQ, HNS_ROCE_CMD_TIMEOUT_MSECS);
 }
 
 static int hns_roce_cq_alloc(struct hns_roce_dev *hr_dev, int nent,
@@ -176,7 +176,7 @@ static int hns_roce_hw2sw_cq(struct hns_roce_dev *dev,
 {
 	return hns_roce_cmd_mbox(dev, 0, mailbox ? mailbox->dma : 0, cq_num,
 				 mailbox ? 0 : 1, HNS_ROCE_CMD_HW2SW_CQ,
-				 HNS_ROCE_CMD_TIME_CLASS_A);
+				 HNS_ROCE_CMD_TIMEOUT_MSECS);
 }
 
 static void hns_roce_free_cq(struct hns_roce_dev *hr_dev,
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
index 528cb84975b4..f7e0fdf33bb1 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
@@ -1871,12 +1871,12 @@ static int hns_roce_v1_qp_modify(struct hns_roce_dev *hr_dev,
 	if (op[cur_state][new_state] == HNS_ROCE_CMD_2RST_QP)
 		return hns_roce_cmd_mbox(hr_dev, 0, 0, hr_qp->qpn, 2,
 					 HNS_ROCE_CMD_2RST_QP,
-					 HNS_ROCE_CMD_TIME_CLASS_A);
+					 HNS_ROCE_CMD_TIMEOUT_MSECS);
 
 	if (op[cur_state][new_state] == HNS_ROCE_CMD_2ERR_QP)
 		return hns_roce_cmd_mbox(hr_dev, 0, 0, hr_qp->qpn, 2,
 					 HNS_ROCE_CMD_2ERR_QP,
-					 HNS_ROCE_CMD_TIME_CLASS_A);
+					 HNS_ROCE_CMD_TIMEOUT_MSECS);
 
 	mailbox = hns_roce_alloc_cmd_mailbox(hr_dev);
 	if (IS_ERR(mailbox))
@@ -1886,7 +1886,7 @@ static int hns_roce_v1_qp_modify(struct hns_roce_dev *hr_dev,
 
 	ret = hns_roce_cmd_mbox(hr_dev, mailbox->dma, 0, hr_qp->qpn, 0,
 				op[cur_state][new_state],
-				HNS_ROCE_CMD_TIME_CLASS_C);
+				HNS_ROCE_CMD_TIMEOUT_MSECS);
 
 	hns_roce_free_cmd_mailbox(hr_dev, mailbox);
 	return ret;
@@ -2681,7 +2681,7 @@ static int hns_roce_v1_query_qpc(struct hns_roce_dev *hr_dev,
 
 	ret = hns_roce_cmd_mbox(hr_dev, 0, mailbox->dma, hr_qp->qpn, 0,
 				HNS_ROCE_CMD_QUERY_QP,
-				HNS_ROCE_CMD_TIME_CLASS_A);
+				HNS_ROCE_CMD_TIMEOUT_MSECS);
 	if (!ret)
 		memcpy(hr_context, mailbox->buf, sizeof(*hr_context));
 	else
diff --git a/drivers/infiniband/hw/hns/hns_roce_mr.c b/drivers/infiniband/hw/hns/hns_roce_mr.c
index d87d189cb3d8..a5bd64586848 100644
--- a/drivers/infiniband/hw/hns/hns_roce_mr.c
+++ b/drivers/infiniband/hw/hns/hns_roce_mr.c
@@ -53,7 +53,7 @@ static int hns_roce_sw2hw_mpt(struct hns_roce_dev *hr_dev,
 {
 	return hns_roce_cmd_mbox(hr_dev, mailbox->dma, 0, mpt_index, 0,
 				 HNS_ROCE_CMD_SW2HW_MPT,
-				 HNS_ROCE_CMD_TIME_CLASS_B);
+				 HNS_ROCE_CMD_TIMEOUT_MSECS);
 }
 
 static int hns_roce_hw2sw_mpt(struct hns_roce_dev *hr_dev,
@@ -62,7 +62,7 @@ static int hns_roce_hw2sw_mpt(struct hns_roce_dev *hr_dev,
 {
 	return hns_roce_cmd_mbox(hr_dev, 0, mailbox ? mailbox->dma : 0,
 				 mpt_index, !mailbox, HNS_ROCE_CMD_HW2SW_MPT,
-				 HNS_ROCE_CMD_TIME_CLASS_B);
+				 HNS_ROCE_CMD_TIMEOUT_MSECS);
 }
 
 static int hns_roce_buddy_alloc(struct hns_roce_buddy *buddy, int order,
-- 
cgit v1.2.1


From dd783a212c023b309e91ab9752c07e22f759461e Mon Sep 17 00:00:00 2001
From: "Wei Hu (Xavier)" <xavier.huwei@huawei.com>
Date: Wed, 23 Nov 2016 19:41:06 +0000
Subject: IB/hns: Modify query info named port_num when querying RC QP

This patch modified the output query info qp_attr->port_num
to fix bug in hip06.

Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Signed-off-by: Salil Mehta  <salil.mehta@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_hw_v1.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
index f7e0fdf33bb1..cc64230d44c0 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
@@ -2857,9 +2857,7 @@ static int hns_roce_v1_q_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr,
 	qp_attr->pkey_index = roce_get_field(context->qpc_bytes_12,
 			      QP_CONTEXT_QPC_BYTES_12_P_KEY_INDEX_M,
 			      QP_CONTEXT_QPC_BYTES_12_P_KEY_INDEX_S);
-	qp_attr->port_num = (u8)roce_get_field(context->qpc_bytes_156,
-			     QP_CONTEXT_QPC_BYTES_156_PORT_NUM_M,
-			     QP_CONTEXT_QPC_BYTES_156_PORT_NUM_S) + 1;
+	qp_attr->port_num = hr_qp->port + 1;
 	qp_attr->sq_draining = 0;
 	qp_attr->max_rd_atomic = roce_get_field(context->qpc_bytes_156,
 				 QP_CONTEXT_QPC_BYTES_156_INITIATOR_DEPTH_M,
-- 
cgit v1.2.1


From 5e6ff78a229c2f231f2f743b017987621e469858 Mon Sep 17 00:00:00 2001
From: "Wei Hu (Xavier)" <xavier.huwei@huawei.com>
Date: Wed, 23 Nov 2016 19:41:07 +0000
Subject: IB/hns: Change qpn allocation to round-robin mode.

When using CM to establish connections, qp number that was freed
just now will be rejected by ib core. To fix these problem, We
change qpn allocation to round-robin mode. We added the round-robin
mode for allocating resources using bitmap. We use round-robin mode
for qp number and non round-robing mode for other resources like
cq number, pd number etc.

Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Signed-off-by: Salil Mehta  <salil.mehta@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_alloc.c  | 11 +++++++----
 drivers/infiniband/hw/hns/hns_roce_cq.c     |  4 ++--
 drivers/infiniband/hw/hns/hns_roce_device.h |  9 +++++++--
 drivers/infiniband/hw/hns/hns_roce_mr.c     |  2 +-
 drivers/infiniband/hw/hns/hns_roce_pd.c     |  5 +++--
 drivers/infiniband/hw/hns/hns_roce_qp.c     |  2 +-
 6 files changed, 21 insertions(+), 12 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hns/hns_roce_alloc.c b/drivers/infiniband/hw/hns/hns_roce_alloc.c
index 863a17a2de40..605962f2828c 100644
--- a/drivers/infiniband/hw/hns/hns_roce_alloc.c
+++ b/drivers/infiniband/hw/hns/hns_roce_alloc.c
@@ -61,9 +61,10 @@ int hns_roce_bitmap_alloc(struct hns_roce_bitmap *bitmap, unsigned long *obj)
 	return ret;
 }
 
-void hns_roce_bitmap_free(struct hns_roce_bitmap *bitmap, unsigned long obj)
+void hns_roce_bitmap_free(struct hns_roce_bitmap *bitmap, unsigned long obj,
+			  int rr)
 {
-	hns_roce_bitmap_free_range(bitmap, obj, 1);
+	hns_roce_bitmap_free_range(bitmap, obj, 1, rr);
 }
 
 int hns_roce_bitmap_alloc_range(struct hns_roce_bitmap *bitmap, int cnt,
@@ -106,7 +107,8 @@ int hns_roce_bitmap_alloc_range(struct hns_roce_bitmap *bitmap, int cnt,
 }
 
 void hns_roce_bitmap_free_range(struct hns_roce_bitmap *bitmap,
-				unsigned long obj, int cnt)
+				unsigned long obj, int cnt,
+				int rr)
 {
 	int i;
 
@@ -116,7 +118,8 @@ void hns_roce_bitmap_free_range(struct hns_roce_bitmap *bitmap,
 	for (i = 0; i < cnt; i++)
 		clear_bit(obj + i, bitmap->table);
 
-	bitmap->last = min(bitmap->last, obj);
+	if (!rr)
+		bitmap->last = min(bitmap->last, obj);
 	bitmap->top = (bitmap->top + bitmap->max + bitmap->reserved_top)
 		       & bitmap->mask;
 	spin_unlock(&bitmap->lock);
diff --git a/drivers/infiniband/hw/hns/hns_roce_cq.c b/drivers/infiniband/hw/hns/hns_roce_cq.c
index 461a27330e83..c9f6c3dce83c 100644
--- a/drivers/infiniband/hw/hns/hns_roce_cq.c
+++ b/drivers/infiniband/hw/hns/hns_roce_cq.c
@@ -166,7 +166,7 @@ err_put:
 	hns_roce_table_put(hr_dev, &cq_table->table, hr_cq->cqn);
 
 err_out:
-	hns_roce_bitmap_free(&cq_table->bitmap, hr_cq->cqn);
+	hns_roce_bitmap_free(&cq_table->bitmap, hr_cq->cqn, BITMAP_NO_RR);
 	return ret;
 }
 
@@ -204,7 +204,7 @@ static void hns_roce_free_cq(struct hns_roce_dev *hr_dev,
 	spin_unlock_irq(&cq_table->lock);
 
 	hns_roce_table_put(hr_dev, &cq_table->table, hr_cq->cqn);
-	hns_roce_bitmap_free(&cq_table->bitmap, hr_cq->cqn);
+	hns_roce_bitmap_free(&cq_table->bitmap, hr_cq->cqn, BITMAP_NO_RR);
 }
 
 static int hns_roce_ib_get_cq_umem(struct hns_roce_dev *hr_dev,
diff --git a/drivers/infiniband/hw/hns/hns_roce_device.h b/drivers/infiniband/hw/hns/hns_roce_device.h
index 7242b1438873..593a42a198f6 100644
--- a/drivers/infiniband/hw/hns/hns_roce_device.h
+++ b/drivers/infiniband/hw/hns/hns_roce_device.h
@@ -72,6 +72,9 @@
 #define HNS_ROCE_MAX_GID_NUM			16
 #define HNS_ROCE_GID_SIZE			16
 
+#define BITMAP_NO_RR				0
+#define BITMAP_RR				1
+
 #define MR_TYPE_MR				0x00
 #define MR_TYPE_DMA				0x03
 
@@ -661,7 +664,8 @@ void hns_roce_cleanup_cq_table(struct hns_roce_dev *hr_dev);
 void hns_roce_cleanup_qp_table(struct hns_roce_dev *hr_dev);
 
 int hns_roce_bitmap_alloc(struct hns_roce_bitmap *bitmap, unsigned long *obj);
-void hns_roce_bitmap_free(struct hns_roce_bitmap *bitmap, unsigned long obj);
+void hns_roce_bitmap_free(struct hns_roce_bitmap *bitmap, unsigned long obj,
+			 int rr);
 int hns_roce_bitmap_init(struct hns_roce_bitmap *bitmap, u32 num, u32 mask,
 			 u32 reserved_bot, u32 resetrved_top);
 void hns_roce_bitmap_cleanup(struct hns_roce_bitmap *bitmap);
@@ -669,7 +673,8 @@ void hns_roce_cleanup_bitmap(struct hns_roce_dev *hr_dev);
 int hns_roce_bitmap_alloc_range(struct hns_roce_bitmap *bitmap, int cnt,
 				int align, unsigned long *obj);
 void hns_roce_bitmap_free_range(struct hns_roce_bitmap *bitmap,
-				unsigned long obj, int cnt);
+				unsigned long obj, int cnt,
+				int rr);
 
 struct ib_ah *hns_roce_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr);
 int hns_roce_query_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr);
diff --git a/drivers/infiniband/hw/hns/hns_roce_mr.c b/drivers/infiniband/hw/hns/hns_roce_mr.c
index a5bd64586848..9b8a1ad4ee6c 100644
--- a/drivers/infiniband/hw/hns/hns_roce_mr.c
+++ b/drivers/infiniband/hw/hns/hns_roce_mr.c
@@ -289,7 +289,7 @@ static void hns_roce_mr_free(struct hns_roce_dev *hr_dev,
 	}
 
 	hns_roce_bitmap_free(&hr_dev->mr_table.mtpt_bitmap,
-			     key_to_hw_index(mr->key));
+			     key_to_hw_index(mr->key), BITMAP_NO_RR);
 }
 
 static int hns_roce_mr_enable(struct hns_roce_dev *hr_dev,
diff --git a/drivers/infiniband/hw/hns/hns_roce_pd.c b/drivers/infiniband/hw/hns/hns_roce_pd.c
index 05db7d59812a..a64500fa1145 100644
--- a/drivers/infiniband/hw/hns/hns_roce_pd.c
+++ b/drivers/infiniband/hw/hns/hns_roce_pd.c
@@ -40,7 +40,7 @@ static int hns_roce_pd_alloc(struct hns_roce_dev *hr_dev, unsigned long *pdn)
 
 static void hns_roce_pd_free(struct hns_roce_dev *hr_dev, unsigned long pdn)
 {
-	hns_roce_bitmap_free(&hr_dev->pd_bitmap, pdn);
+	hns_roce_bitmap_free(&hr_dev->pd_bitmap, pdn, BITMAP_NO_RR);
 }
 
 int hns_roce_init_pd_table(struct hns_roce_dev *hr_dev)
@@ -121,7 +121,8 @@ int hns_roce_uar_alloc(struct hns_roce_dev *hr_dev, struct hns_roce_uar *uar)
 
 void hns_roce_uar_free(struct hns_roce_dev *hr_dev, struct hns_roce_uar *uar)
 {
-	hns_roce_bitmap_free(&hr_dev->uar_table.bitmap, uar->index);
+	hns_roce_bitmap_free(&hr_dev->uar_table.bitmap, uar->index,
+			     BITMAP_NO_RR);
 }
 
 int hns_roce_init_uar_table(struct hns_roce_dev *hr_dev)
diff --git a/drivers/infiniband/hw/hns/hns_roce_qp.c b/drivers/infiniband/hw/hns/hns_roce_qp.c
index e86dd8d06777..4775b5c725a9 100644
--- a/drivers/infiniband/hw/hns/hns_roce_qp.c
+++ b/drivers/infiniband/hw/hns/hns_roce_qp.c
@@ -250,7 +250,7 @@ void hns_roce_release_range_qp(struct hns_roce_dev *hr_dev, int base_qpn,
 	if (base_qpn < SQP_NUM)
 		return;
 
-	hns_roce_bitmap_free_range(&qp_table->bitmap, base_qpn, cnt);
+	hns_roce_bitmap_free_range(&qp_table->bitmap, base_qpn, cnt, BITMAP_RR);
 }
 
 static int hns_roce_set_rq_size(struct hns_roce_dev *hr_dev,
-- 
cgit v1.2.1


From 82547469782a952452c84c055c7911e635c77cd0 Mon Sep 17 00:00:00 2001
From: Shaobo Xu <xushaobo2@huawei.com>
Date: Wed, 23 Nov 2016 19:41:08 +0000
Subject: IB/hns: Implement the add_gid/del_gid and optimize the GIDs
 management

IB core has implemented the calculation of GIDs and the management
of GID tables, and it is now responsible to supply query function
for GIDs. So the calculation of GIDs and the management of GID
tables in the RoCE driver is redundant.

The patch is to implement the add_gid/del_gid to set the GIDs in
the RoCE driver, remove the redundant calculation and management of
GIDs in the notifier call of the net device and the inet, and
update the query_gid.

Signed-off-by: Shaobo Xu <xushaobo2@huawei.com>
Reviewed-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Signed-off-by: Salil Mehta  <salil.mehta@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_device.h |   2 -
 drivers/infiniband/hw/hns/hns_roce_main.c   | 270 +++++-----------------------
 2 files changed, 48 insertions(+), 224 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hns/hns_roce_device.h b/drivers/infiniband/hw/hns/hns_roce_device.h
index 593a42a198f6..9ef1cc3ec930 100644
--- a/drivers/infiniband/hw/hns/hns_roce_device.h
+++ b/drivers/infiniband/hw/hns/hns_roce_device.h
@@ -429,8 +429,6 @@ struct hns_roce_ib_iboe {
 	struct net_device      *netdevs[HNS_ROCE_MAX_PORTS];
 	struct notifier_block	nb;
 	struct notifier_block	nb_inet;
-	/* 16 GID is shared by 6 port in v1 engine. */
-	union ib_gid		gid_table[HNS_ROCE_MAX_GID_NUM];
 	u8			phy_port[HNS_ROCE_MAX_PORTS];
 };
 
diff --git a/drivers/infiniband/hw/hns/hns_roce_main.c b/drivers/infiniband/hw/hns/hns_roce_main.c
index 67701719bad1..795ef97bfcbf 100644
--- a/drivers/infiniband/hw/hns/hns_roce_main.c
+++ b/drivers/infiniband/hw/hns/hns_roce_main.c
@@ -35,51 +35,12 @@
 #include <rdma/ib_addr.h>
 #include <rdma/ib_smi.h>
 #include <rdma/ib_user_verbs.h>
+#include <rdma/ib_cache.h>
 #include "hns_roce_common.h"
 #include "hns_roce_device.h"
 #include "hns_roce_user.h"
 #include "hns_roce_hem.h"
 
-/**
- * hns_roce_addrconf_ifid_eui48 - Get default gid.
- * @eui: eui.
- * @vlan_id:  gid
- * @dev:  net device
- * Description:
- *    MAC convert to GID
- *        gid[0..7] = fe80 0000 0000 0000
- *        gid[8] = mac[0] ^ 2
- *        gid[9] = mac[1]
- *        gid[10] = mac[2]
- *        gid[11] = ff        (VLAN ID high byte (4 MS bits))
- *        gid[12] = fe        (VLAN ID low byte)
- *        gid[13] = mac[3]
- *        gid[14] = mac[4]
- *        gid[15] = mac[5]
- */
-static void hns_roce_addrconf_ifid_eui48(u8 *eui, u16 vlan_id,
-					 struct net_device *dev)
-{
-	memcpy(eui, dev->dev_addr, 3);
-	memcpy(eui + 5, dev->dev_addr + 3, 3);
-	if (vlan_id < 0x1000) {
-		eui[3] = vlan_id >> 8;
-		eui[4] = vlan_id & 0xff;
-	} else {
-		eui[3] = 0xff;
-		eui[4] = 0xfe;
-	}
-	eui[0] ^= 2;
-}
-
-static void hns_roce_make_default_gid(struct net_device *dev, union ib_gid *gid)
-{
-	memset(gid, 0, sizeof(*gid));
-	gid->raw[0] = 0xFE;
-	gid->raw[1] = 0x80;
-	hns_roce_addrconf_ifid_eui48(&gid->raw[8], 0xffff, dev);
-}
-
 /**
  * hns_get_gid_index - Get gid index.
  * @hr_dev: pointer to structure hns_roce_dev.
@@ -96,30 +57,6 @@ int hns_get_gid_index(struct hns_roce_dev *hr_dev, u8 port, int gid_index)
 	return gid_index * hr_dev->caps.num_ports + port;
 }
 
-static int hns_roce_set_gid(struct hns_roce_dev *hr_dev, u8 port, int gid_index,
-		     union ib_gid *gid)
-{
-	struct device *dev = &hr_dev->pdev->dev;
-	u8 gid_idx = 0;
-
-	if (gid_index >= hr_dev->caps.gid_table_len[port]) {
-		dev_err(dev, "gid_index %d illegal, port %d gid range: 0~%d\n",
-			gid_index, port, hr_dev->caps.gid_table_len[port] - 1);
-		return -EINVAL;
-	}
-
-	gid_idx = hns_get_gid_index(hr_dev, port, gid_index);
-
-	if (!memcmp(gid, &hr_dev->iboe.gid_table[gid_idx], sizeof(*gid)))
-		return -EINVAL;
-
-	memcpy(&hr_dev->iboe.gid_table[gid_idx], gid, sizeof(*gid));
-
-	hr_dev->hw->set_gid(hr_dev, port, gid_index, gid);
-
-	return 0;
-}
-
 static void hns_roce_set_mac(struct hns_roce_dev *hr_dev, u8 port, u8 *addr)
 {
 	u8 phy_port;
@@ -147,15 +84,44 @@ static void hns_roce_set_mtu(struct hns_roce_dev *hr_dev, u8 port, int mtu)
 	hr_dev->hw->set_mtu(hr_dev, phy_port, tmp);
 }
 
-static void hns_roce_update_gids(struct hns_roce_dev *hr_dev, int port)
+static int hns_roce_add_gid(struct ib_device *device, u8 port_num,
+			    unsigned int index, const union ib_gid *gid,
+			    const struct ib_gid_attr *attr, void **context)
+{
+	struct hns_roce_dev *hr_dev = to_hr_dev(device);
+	u8 port = port_num - 1;
+	unsigned long flags;
+
+	if (port >= hr_dev->caps.num_ports)
+		return -EINVAL;
+
+	spin_lock_irqsave(&hr_dev->iboe.lock, flags);
+
+	hr_dev->hw->set_gid(hr_dev, port, index, (union ib_gid *)gid);
+
+	spin_unlock_irqrestore(&hr_dev->iboe.lock, flags);
+
+	return 0;
+}
+
+static int hns_roce_del_gid(struct ib_device *device, u8 port_num,
+			    unsigned int index, void **context)
 {
-	struct ib_event event;
+	struct hns_roce_dev *hr_dev = to_hr_dev(device);
+	union ib_gid zgid = { {0} };
+	u8 port = port_num - 1;
+	unsigned long flags;
+
+	if (port >= hr_dev->caps.num_ports)
+		return -EINVAL;
 
-	/* Refresh gid in ib_cache */
-	event.device = &hr_dev->ib_dev;
-	event.element.port_num = port + 1;
-	event.event = IB_EVENT_GID_CHANGE;
-	ib_dispatch_event(&event);
+	spin_lock_irqsave(&hr_dev->iboe.lock, flags);
+
+	hr_dev->hw->set_gid(hr_dev, port, index, &zgid);
+
+	spin_unlock_irqrestore(&hr_dev->iboe.lock, flags);
+
+	return 0;
 }
 
 static int handle_en_event(struct hns_roce_dev *hr_dev, u8 port,
@@ -164,8 +130,6 @@ static int handle_en_event(struct hns_roce_dev *hr_dev, u8 port,
 	struct device *dev = &hr_dev->pdev->dev;
 	struct net_device *netdev;
 	unsigned long flags;
-	union ib_gid gid;
-	int ret = 0;
 
 	netdev = hr_dev->iboe.netdevs[port];
 	if (!netdev) {
@@ -181,10 +145,6 @@ static int handle_en_event(struct hns_roce_dev *hr_dev, u8 port,
 	case NETDEV_REGISTER:
 	case NETDEV_CHANGEADDR:
 		hns_roce_set_mac(hr_dev, port, netdev->dev_addr);
-		hns_roce_make_default_gid(netdev, &gid);
-		ret = hns_roce_set_gid(hr_dev, port, 0, &gid);
-		if (!ret)
-			hns_roce_update_gids(hr_dev, port);
 		break;
 	case NETDEV_DOWN:
 		/*
@@ -197,7 +157,7 @@ static int handle_en_event(struct hns_roce_dev *hr_dev, u8 port,
 	}
 
 	spin_unlock_irqrestore(&hr_dev->iboe.lock, flags);
-	return ret;
+	return 0;
 }
 
 static int hns_roce_netdev_event(struct notifier_block *self,
@@ -224,118 +184,17 @@ static int hns_roce_netdev_event(struct notifier_block *self,
 	return NOTIFY_DONE;
 }
 
-static void hns_roce_addr_event(int event, struct net_device *event_netdev,
-				struct hns_roce_dev *hr_dev, union ib_gid *gid)
-{
-	struct hns_roce_ib_iboe *iboe = NULL;
-	int gid_table_len = 0;
-	unsigned long flags;
-	union ib_gid zgid;
-	u8 gid_idx = 0;
-	u8 port = 0;
-	int i = 0;
-	int free;
-	struct net_device *real_dev = rdma_vlan_dev_real_dev(event_netdev) ?
-				      rdma_vlan_dev_real_dev(event_netdev) :
-				      event_netdev;
-
-	if (event != NETDEV_UP && event != NETDEV_DOWN)
-		return;
-
-	iboe = &hr_dev->iboe;
-	while (port < hr_dev->caps.num_ports) {
-		if (real_dev == iboe->netdevs[port])
-			break;
-		port++;
-	}
-
-	if (port >= hr_dev->caps.num_ports) {
-		dev_dbg(&hr_dev->pdev->dev, "can't find netdev\n");
-		return;
-	}
-
-	memset(zgid.raw, 0, sizeof(zgid.raw));
-	free = -1;
-	gid_table_len = hr_dev->caps.gid_table_len[port];
-
-	spin_lock_irqsave(&hr_dev->iboe.lock, flags);
-
-	for (i = 0; i < gid_table_len; i++) {
-		gid_idx = hns_get_gid_index(hr_dev, port, i);
-		if (!memcmp(gid->raw, iboe->gid_table[gid_idx].raw,
-			    sizeof(gid->raw)))
-			break;
-		if (free < 0 && !memcmp(zgid.raw,
-			iboe->gid_table[gid_idx].raw, sizeof(zgid.raw)))
-			free = i;
-	}
-
-	if (i >= gid_table_len) {
-		if (free < 0) {
-			spin_unlock_irqrestore(&hr_dev->iboe.lock, flags);
-			dev_dbg(&hr_dev->pdev->dev,
-				"gid_index overflow, port(%d)\n", port);
-			return;
-		}
-		if (!hns_roce_set_gid(hr_dev, port, free, gid))
-			hns_roce_update_gids(hr_dev, port);
-	} else if (event == NETDEV_DOWN) {
-		if (!hns_roce_set_gid(hr_dev, port, i, &zgid))
-			hns_roce_update_gids(hr_dev, port);
-	}
-
-	spin_unlock_irqrestore(&hr_dev->iboe.lock, flags);
-}
-
-static int hns_roce_inet_event(struct notifier_block *self, unsigned long event,
-			       void *ptr)
-{
-	struct in_ifaddr *ifa = ptr;
-	struct hns_roce_dev *hr_dev;
-	struct net_device *dev = ifa->ifa_dev->dev;
-	union ib_gid gid;
-
-	ipv6_addr_set_v4mapped(ifa->ifa_address, (struct in6_addr *)&gid);
-
-	hr_dev = container_of(self, struct hns_roce_dev, iboe.nb_inet);
-
-	hns_roce_addr_event(event, dev, hr_dev, &gid);
-
-	return NOTIFY_DONE;
-}
-
-static int hns_roce_setup_mtu_gids(struct hns_roce_dev *hr_dev)
+static int hns_roce_setup_mtu_mac(struct hns_roce_dev *hr_dev)
 {
-	struct in_ifaddr *ifa_list = NULL;
-	union ib_gid gid = {{0} };
-	u32 ipaddr = 0;
-	int index = 0;
-	int ret = 0;
-	u8 i = 0;
+	u8 i;
 
 	for (i = 0; i < hr_dev->caps.num_ports; i++) {
 		hns_roce_set_mtu(hr_dev, i,
 				 ib_mtu_enum_to_int(hr_dev->caps.max_mtu));
 		hns_roce_set_mac(hr_dev, i, hr_dev->iboe.netdevs[i]->dev_addr);
-
-		if (hr_dev->iboe.netdevs[i]->ip_ptr) {
-			ifa_list = hr_dev->iboe.netdevs[i]->ip_ptr->ifa_list;
-			index = 1;
-			while (ifa_list) {
-				ipaddr = ifa_list->ifa_address;
-				ipv6_addr_set_v4mapped(ipaddr,
-						       (struct in6_addr *)&gid);
-				ret = hns_roce_set_gid(hr_dev, i, index, &gid);
-				if (ret)
-					break;
-				index++;
-				ifa_list = ifa_list->ifa_next;
-			}
-			hns_roce_update_gids(hr_dev, i);
-		}
 	}
 
-	return ret;
+	return 0;
 }
 
 static int hns_roce_query_device(struct ib_device *ib_dev,
@@ -444,31 +303,6 @@ static enum rdma_link_layer hns_roce_get_link_layer(struct ib_device *device,
 static int hns_roce_query_gid(struct ib_device *ib_dev, u8 port_num, int index,
 			      union ib_gid *gid)
 {
-	struct hns_roce_dev *hr_dev = to_hr_dev(ib_dev);
-	struct device *dev = &hr_dev->pdev->dev;
-	u8 gid_idx = 0;
-	u8 port;
-
-	if (port_num < 1 || port_num > hr_dev->caps.num_ports ||
-	    index >= hr_dev->caps.gid_table_len[port_num - 1]) {
-		dev_err(dev,
-			"port_num %d index %d illegal! correct range: port_num 1~%d index 0~%d!\n",
-			port_num, index, hr_dev->caps.num_ports,
-			hr_dev->caps.gid_table_len[port_num - 1] - 1);
-		return -EINVAL;
-	}
-
-	port = port_num - 1;
-	gid_idx = hns_get_gid_index(hr_dev, port, index);
-	if (gid_idx >= HNS_ROCE_MAX_GID_NUM) {
-		dev_err(dev, "port_num %d index %d illegal! total gid num %d!\n",
-			port_num, index, HNS_ROCE_MAX_GID_NUM);
-		return -EINVAL;
-	}
-
-	memcpy(gid->raw, hr_dev->iboe.gid_table[gid_idx].raw,
-	       HNS_ROCE_GID_SIZE);
-
 	return 0;
 }
 
@@ -646,6 +480,8 @@ static int hns_roce_register_device(struct hns_roce_dev *hr_dev)
 	ib_dev->get_link_layer		= hns_roce_get_link_layer;
 	ib_dev->get_netdev		= hns_roce_get_netdev;
 	ib_dev->query_gid		= hns_roce_query_gid;
+	ib_dev->add_gid			= hns_roce_add_gid;
+	ib_dev->del_gid			= hns_roce_del_gid;
 	ib_dev->query_pkey		= hns_roce_query_pkey;
 	ib_dev->alloc_ucontext		= hns_roce_alloc_ucontext;
 	ib_dev->dealloc_ucontext	= hns_roce_dealloc_ucontext;
@@ -688,32 +524,22 @@ static int hns_roce_register_device(struct hns_roce_dev *hr_dev)
 		return ret;
 	}
 
-	ret = hns_roce_setup_mtu_gids(hr_dev);
+	ret = hns_roce_setup_mtu_mac(hr_dev);
 	if (ret) {
-		dev_err(dev, "roce_setup_mtu_gids failed!\n");
-		goto error_failed_setup_mtu_gids;
+		dev_err(dev, "setup_mtu_mac failed!\n");
+		goto error_failed_setup_mtu_mac;
 	}
 
 	iboe->nb.notifier_call = hns_roce_netdev_event;
 	ret = register_netdevice_notifier(&iboe->nb);
 	if (ret) {
 		dev_err(dev, "register_netdevice_notifier failed!\n");
-		goto error_failed_setup_mtu_gids;
-	}
-
-	iboe->nb_inet.notifier_call = hns_roce_inet_event;
-	ret = register_inetaddr_notifier(&iboe->nb_inet);
-	if (ret) {
-		dev_err(dev, "register inet addr notifier failed!\n");
-		goto error_failed_register_inetaddr_notifier;
+		goto error_failed_setup_mtu_mac;
 	}
 
 	return 0;
 
-error_failed_register_inetaddr_notifier:
-	unregister_netdevice_notifier(&iboe->nb);
-
-error_failed_setup_mtu_gids:
+error_failed_setup_mtu_mac:
 	ib_unregister_device(ib_dev);
 
 	return ret;
-- 
cgit v1.2.1


From e84e40be8e7bc29599da7056b340490d25ff87b5 Mon Sep 17 00:00:00 2001
From: Salil <salil.mehta@huawei.com>
Date: Wed, 23 Nov 2016 19:41:09 +0000
Subject: IB/hns: Fix for Checkpatch.pl comment style errors

This patch correct the comment style errors caught by
checkpatch.pl script

Signed-off-by: Salil Mehta  <salil.mehta@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_cmd.c    |  8 ++---
 drivers/infiniband/hw/hns/hns_roce_device.h | 28 +++++++--------
 drivers/infiniband/hw/hns/hns_roce_eq.c     |  6 ++--
 drivers/infiniband/hw/hns/hns_roce_hem.c    |  6 ++--
 drivers/infiniband/hw/hns/hns_roce_hw_v1.c  | 56 ++++++++++++++---------------
 drivers/infiniband/hw/hns/hns_roce_main.c   | 28 +++++++--------
 6 files changed, 66 insertions(+), 66 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hns/hns_roce_cmd.c b/drivers/infiniband/hw/hns/hns_roce_cmd.c
index 2a0b6c05da5f..8c1f7a6f84d2 100644
--- a/drivers/infiniband/hw/hns/hns_roce_cmd.c
+++ b/drivers/infiniband/hw/hns/hns_roce_cmd.c
@@ -216,10 +216,10 @@ static int __hns_roce_cmd_mbox_wait(struct hns_roce_dev *hr_dev, u64 in_param,
 		goto out;
 
 	/*
-	* It is timeout when wait_for_completion_timeout return 0
-	* The return value is the time limit set in advance
-	* how many seconds showing
-	*/
+	 * It is timeout when wait_for_completion_timeout return 0
+	 * The return value is the time limit set in advance
+	 * how many seconds showing
+	 */
 	if (!wait_for_completion_timeout(&context->done,
 					 msecs_to_jiffies(timeout))) {
 		dev_err(dev, "[cmd]wait_for_completion_timeout timeout\n");
diff --git a/drivers/infiniband/hw/hns/hns_roce_device.h b/drivers/infiniband/hw/hns/hns_roce_device.h
index 9ef1cc3ec930..e48464ddd2b0 100644
--- a/drivers/infiniband/hw/hns/hns_roce_device.h
+++ b/drivers/infiniband/hw/hns/hns_roce_device.h
@@ -201,9 +201,9 @@ struct hns_roce_bitmap {
 /* Order = 0: bitmap is biggest, order = max bitmap is least (only a bit) */
 /* Every bit repesent to a partner free/used status in bitmap */
 /*
-* Initial, bits of other bitmap are all 0 except that a bit of max_order is 1
-* Bit = 1 represent to idle and available; bit = 0: not available
-*/
+ * Initial, bits of other bitmap are all 0 except that a bit of max_order is 1
+ * Bit = 1 represent to idle and available; bit = 0: not available
+ */
 struct hns_roce_buddy {
 	/* Members point to every order level bitmap */
 	unsigned long **bits;
@@ -365,25 +365,25 @@ struct hns_roce_cmdq {
 	struct mutex		hcr_mutex;
 	struct semaphore	poll_sem;
 	/*
-	* Event mode: cmd register mutex protection,
-	* ensure to not exceed max_cmds and user use limit region
-	*/
+	 * Event mode: cmd register mutex protection,
+	 * ensure to not exceed max_cmds and user use limit region
+	 */
 	struct semaphore	event_sem;
 	int			max_cmds;
 	spinlock_t		context_lock;
 	int			free_head;
 	struct hns_roce_cmd_context *context;
 	/*
-	* Result of get integer part
-	* which max_comds compute according a power of 2
-	*/
+	 * Result of get integer part
+	 * which max_comds compute according a power of 2
+	 */
 	u16			token_mask;
 	/*
-	* Process whether use event mode, init default non-zero
-	* After the event queue of cmd event ready,
-	* can switch into event mode
-	* close device, switch into poll mode(non event mode)
-	*/
+	 * Process whether use event mode, init default non-zero
+	 * After the event queue of cmd event ready,
+	 * can switch into event mode
+	 * close device, switch into poll mode(non event mode)
+	 */
 	u8			use_events;
 	u8			toggle;
 };
diff --git a/drivers/infiniband/hw/hns/hns_roce_eq.c b/drivers/infiniband/hw/hns/hns_roce_eq.c
index 21e21b03cfb5..50f864935a0e 100644
--- a/drivers/infiniband/hw/hns/hns_roce_eq.c
+++ b/drivers/infiniband/hw/hns/hns_roce_eq.c
@@ -371,9 +371,9 @@ static int hns_roce_aeq_ovf_int(struct hns_roce_dev *hr_dev,
 	int i = 0;
 
 	/**
-	* AEQ overflow ECC mult bit err CEQ overflow alarm
-	* must clear interrupt, mask irq, clear irq, cancel mask operation
-	*/
+	 * AEQ overflow ECC mult bit err CEQ overflow alarm
+	 * must clear interrupt, mask irq, clear irq, cancel mask operation
+	 */
 	aeshift_val = roce_read(hr_dev, ROCEE_CAEP_AEQC_AEQE_SHIFT_REG);
 
 	if (roce_get_bit(aeshift_val,
diff --git a/drivers/infiniband/hw/hns/hns_roce_hem.c b/drivers/infiniband/hw/hns/hns_roce_hem.c
index 250d8f280390..c5104e0b2916 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hem.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hem.c
@@ -80,9 +80,9 @@ struct hns_roce_hem *hns_roce_alloc_hem(struct hns_roce_dev *hr_dev, int npages,
 			--order;
 
 		/*
-		* Alloc memory one time. If failed, don't alloc small block
-		* memory, directly return fail.
-		*/
+		 * Alloc memory one time. If failed, don't alloc small block
+		 * memory, directly return fail.
+		 */
 		mem = &chunk->mem[chunk->npages];
 		buf = dma_alloc_coherent(&hr_dev->pdev->dev, PAGE_SIZE << order,
 				&sg_dma_address(mem), gfp_mask);
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
index cc64230d44c0..125ab90157f6 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
@@ -1352,9 +1352,9 @@ static void __hns_roce_v1_cq_clean(struct hns_roce_cq *hr_cq, u32 qpn,
 	}
 
 	/*
-	* Now backwards through the CQ, removing CQ entries
-	* that match our QP by overwriting them with next entries.
-	*/
+	 * Now backwards through the CQ, removing CQ entries
+	 * that match our QP by overwriting them with next entries.
+	 */
 	while ((int) --prod_index - (int) hr_cq->cons_index >= 0) {
 		cqe = get_cqe(hr_cq, prod_index & hr_cq->ib_cq.cqe);
 		if ((roce_get_field(cqe->cqe_byte_16, CQE_BYTE_16_LOCAL_QPN_M,
@@ -1376,9 +1376,9 @@ static void __hns_roce_v1_cq_clean(struct hns_roce_cq *hr_cq, u32 qpn,
 	if (nfreed) {
 		hr_cq->cons_index += nfreed;
 		/*
-		* Make sure update of buffer contents is done before
-		* updating consumer index.
-		*/
+		 * Make sure update of buffer contents is done before
+		 * updating consumer index.
+		 */
 		wmb();
 
 		hns_roce_v1_cq_set_ci(hr_cq, hr_cq->cons_index);
@@ -1473,7 +1473,7 @@ void hns_roce_v1_write_cqc(struct hns_roce_dev *hr_dev,
 	roce_set_bit(cq_context->cqc_byte_32,
 		     CQ_CQNTEXT_CQC_BYTE_32_TYPE_OF_COMPLETION_NOTIFICATION_S,
 		     0);
-	/*The initial value of cq's ci is 0 */
+	/* The initial value of cq's ci is 0 */
 	roce_set_field(cq_context->cqc_byte_32,
 		       CQ_CONTEXT_CQC_BYTE_32_CQ_CONS_IDX_M,
 		       CQ_CONTEXT_CQC_BYTE_32_CQ_CONS_IDX_S, 0);
@@ -1490,9 +1490,9 @@ int hns_roce_v1_req_notify_cq(struct ib_cq *ibcq, enum ib_cq_notify_flags flags)
 	notification_flag = (flags & IB_CQ_SOLICITED_MASK) ==
 			    IB_CQ_SOLICITED ? CQ_DB_REQ_NOT : CQ_DB_REQ_NOT_SOL;
 	/*
-	* flags = 0; Notification Flag = 1, next
-	* flags = 1; Notification Flag = 0, solocited
-	*/
+	 * flags = 0; Notification Flag = 1, next
+	 * flags = 1; Notification Flag = 0, solocited
+	 */
 	doorbell[0] = hr_cq->cons_index & ((hr_cq->cq_depth << 1) - 1);
 	roce_set_bit(doorbell[1], ROCEE_DB_OTHERS_H_ROCEE_DB_OTH_HW_SYNS_S, 1);
 	roce_set_field(doorbell[1], ROCEE_DB_OTHERS_H_ROCEE_DB_OTH_CMD_M,
@@ -1647,10 +1647,10 @@ static int hns_roce_v1_poll_one(struct hns_roce_cq *hr_cq,
 		wq = &(*cur_qp)->sq;
 		if ((*cur_qp)->sq_signal_bits) {
 			/*
-			* If sg_signal_bit is 1,
-			* firstly tail pointer updated to wqe
-			* which current cqe correspond to
-			*/
+			 * If sg_signal_bit is 1,
+			 * firstly tail pointer updated to wqe
+			 * which current cqe correspond to
+			 */
 			wqe_ctr = (u16)roce_get_field(cqe->cqe_byte_4,
 						      CQE_BYTE_4_WQE_INDEX_M,
 						      CQE_BYTE_4_WQE_INDEX_S);
@@ -2072,11 +2072,11 @@ static int hns_roce_v1_m_qp(struct ib_qp *ibqp, const struct ib_qp_attr *attr,
 	}
 
 	/*
-	*Reset to init
-	*	Mandatory param:
-	*	IB_QP_STATE | IB_QP_PKEY_INDEX | IB_QP_PORT | IB_QP_ACCESS_FLAGS
-	*	Optional param: NA
-	*/
+	 * Reset to init
+	 *	Mandatory param:
+	 *	IB_QP_STATE | IB_QP_PKEY_INDEX | IB_QP_PORT | IB_QP_ACCESS_FLAGS
+	 *	Optional param: NA
+	 */
 	if (cur_state == IB_QPS_RESET && new_state == IB_QPS_INIT) {
 		roce_set_field(context->qpc_bytes_4,
 			       QP_CONTEXT_QPC_BYTES_4_TRANSPORT_SERVICE_TYPE_M,
@@ -2584,9 +2584,9 @@ static int hns_roce_v1_m_qp(struct ib_qp *ibqp, const struct ib_qp_attr *attr,
 	}
 
 	/*
-	* Use rst2init to instead of init2init with drv,
-	* need to hw to flash RQ HEAD by DB again
-	*/
+	 * Use rst2init to instead of init2init with drv,
+	 * need to hw to flash RQ HEAD by DB again
+	 */
 	if (cur_state == IB_QPS_INIT && new_state == IB_QPS_INIT) {
 		/* Memory barrier */
 		wmb();
@@ -2921,9 +2921,9 @@ static void hns_roce_v1_destroy_qp_common(struct hns_roce_dev *hr_dev,
 	if (hr_qp->ibqp.qp_type == IB_QPT_RC) {
 		if (hr_qp->state != IB_QPS_RESET) {
 			/*
-			* Set qp to ERR,
-			* waiting for hw complete processing all dbs
-			*/
+			 * Set qp to ERR,
+			 * waiting for hw complete processing all dbs
+			 */
 			if (hns_roce_v1_qp_modify(hr_dev, NULL,
 					to_hns_roce_state(
 						(enum ib_qp_state)hr_qp->state),
@@ -2936,9 +2936,9 @@ static void hns_roce_v1_destroy_qp_common(struct hns_roce_dev *hr_dev,
 			sdbisusepr_val = roce_read(hr_dev,
 					 ROCEE_SDB_ISSUE_PTR_REG);
 			/*
-			* Query db process status,
-			* until hw process completely
-			*/
+			 * Query db process status,
+			 * until hw process completely
+			 */
 			end = msecs_to_jiffies(
 			      HNS_ROCE_QP_DESTROY_TIMEOUT_MSECS) + jiffies;
 			do {
diff --git a/drivers/infiniband/hw/hns/hns_roce_main.c b/drivers/infiniband/hw/hns/hns_roce_main.c
index 795ef97bfcbf..914d0ac7881c 100644
--- a/drivers/infiniband/hw/hns/hns_roce_main.c
+++ b/drivers/infiniband/hw/hns/hns_roce_main.c
@@ -148,8 +148,8 @@ static int handle_en_event(struct hns_roce_dev *hr_dev, u8 port,
 		break;
 	case NETDEV_DOWN:
 		/*
-		* In v1 engine, only support all ports closed together.
-		*/
+		 * In v1 engine, only support all ports closed together.
+		 */
 		break;
 	default:
 		dev_dbg(dev, "NETDEV event = 0x%x!\n", (u32)(event));
@@ -773,10 +773,10 @@ err_unmap_mtt:
 }
 
 /**
-* hns_roce_setup_hca - setup host channel adapter
-* @hr_dev: pointer to hns roce device
-* Return : int
-*/
+ * hns_roce_setup_hca - setup host channel adapter
+ * @hr_dev: pointer to hns roce device
+ * Return : int
+ */
 static int hns_roce_setup_hca(struct hns_roce_dev *hr_dev)
 {
 	int ret;
@@ -841,11 +841,11 @@ err_uar_table_free:
 }
 
 /**
-* hns_roce_probe - RoCE driver entrance
-* @pdev: pointer to platform device
-* Return : int
-*
-*/
+ * hns_roce_probe - RoCE driver entrance
+ * @pdev: pointer to platform device
+ * Return : int
+ *
+ */
 static int hns_roce_probe(struct platform_device *pdev)
 {
 	int ret;
@@ -958,9 +958,9 @@ error_failed_get_cfg:
 }
 
 /**
-* hns_roce_remove - remove RoCE device
-* @pdev: pointer to platform device
-*/
+ * hns_roce_remove - remove RoCE device
+ * @pdev: pointer to platform device
+ */
 static int hns_roce_remove(struct platform_device *pdev)
 {
 	struct hns_roce_dev *hr_dev = platform_get_drvdata(pdev);
-- 
cgit v1.2.1


From d838c481e025db374171e16d5cc463b85cffec9f Mon Sep 17 00:00:00 2001
From: "Wei Hu (Xavier)" <xavier.huwei@huawei.com>
Date: Tue, 29 Nov 2016 23:10:25 +0000
Subject: IB/hns: Fix the bug when destroy qp

If send queue is still working when qp is in reset state by modify qp
in destroy qp function, hardware will hold on and don't work in hip06
SoC. In current codes, RoCE driver check hardware pointer of sending and
hardware pointer of processing to ensure that hardware has processed all
the dbs of this qp. But while the environment of wire becomes not good,
The checking time maybe too long.

In order to solve this problem, RoCE driver created a workqueue at probe
function. If there is a timeout when checking the status of qp, driver
initialize work entry and push it into the workqueue, Work function will
finish checking and release the related resources later.

Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Signed-off-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Dongdong Huang(Donald) <hdd.huang@huawei.com>
Signed-off-by: Salil Mehta <salil.mehta@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_common.h |  40 +++
 drivers/infiniband/hw/hns/hns_roce_hw_v1.c  | 435 ++++++++++++++++++++++------
 drivers/infiniband/hw/hns/hns_roce_hw_v1.h  |  23 ++
 3 files changed, 402 insertions(+), 96 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hns/hns_roce_common.h b/drivers/infiniband/hw/hns/hns_roce_common.h
index 0dcb620e91dd..a0556323a678 100644
--- a/drivers/infiniband/hw/hns/hns_roce_common.h
+++ b/drivers/infiniband/hw/hns/hns_roce_common.h
@@ -57,6 +57,32 @@
 #define roce_set_bit(origin, shift, val) \
 	roce_set_field((origin), (1ul << (shift)), (shift), (val))
 
+/*
+ * roce_hw_index_cmp_lt - Compare two hardware index values in hisilicon
+ *                        SOC, check if a is less than b.
+ * @a: hardware index value
+ * @b: hardware index value
+ * @bits: the number of bits of a and b, range: 0~31.
+ *
+ * Hardware index increases continuously till max value, and then restart
+ * from zero, again and again. Because the bits of reg field is often
+ * limited, the reg field can only hold the low bits of the hardware index
+ * in hisilicon SOC.
+ * In some scenes we need to compare two values(a,b) getted from two reg
+ * fields in this driver, for example:
+ * If a equals 0xfffe, b equals 0x1 and bits equals 16, we think b has
+ * incresed from 0xffff to 0x1 and a is less than b.
+ * If a equals 0xfffe, b equals 0x0xf001 and bits equals 16, we think a
+ * is bigger than b.
+ *
+ * Return true on a less than b, otherwise false.
+ */
+#define roce_hw_index_mask(bits)	((1ul << (bits)) - 1)
+#define roce_hw_index_shift(bits)	(32 - (bits))
+#define roce_hw_index_cmp_lt(a, b, bits) \
+	((int)((((a) - (b)) & roce_hw_index_mask(bits)) << \
+		roce_hw_index_shift(bits)) < 0)
+
 #define ROCEE_GLB_CFG_ROCEE_DB_SQ_MODE_S 3
 #define ROCEE_GLB_CFG_ROCEE_DB_OTH_MODE_S 4
 
@@ -245,10 +271,22 @@
 #define ROCEE_SDB_SEND_PTR_SDB_SEND_PTR_M   \
 	(((1UL << 28) - 1) << ROCEE_SDB_SEND_PTR_SDB_SEND_PTR_S)
 
+#define ROCEE_SDB_PTR_CMP_BITS 28
+
 #define ROCEE_SDB_INV_CNT_SDB_INV_CNT_S 0
 #define ROCEE_SDB_INV_CNT_SDB_INV_CNT_M   \
 	(((1UL << 16) - 1) << ROCEE_SDB_INV_CNT_SDB_INV_CNT_S)
 
+#define ROCEE_SDB_RETRY_CNT_SDB_RETRY_CT_S	0
+#define ROCEE_SDB_RETRY_CNT_SDB_RETRY_CT_M	\
+	(((1UL << 16) - 1) << ROCEE_SDB_RETRY_CNT_SDB_RETRY_CT_S)
+
+#define ROCEE_SDB_CNT_CMP_BITS 16
+
+#define ROCEE_TSP_BP_ST_QH_FIFO_ENTRY_S	20
+
+#define ROCEE_CNT_CLR_CE_CNT_CLR_CE_S 0
+
 /*************ROCEE_REG DEFINITION****************/
 #define ROCEE_VENDOR_ID_REG			0x0
 #define ROCEE_VENDOR_PART_ID_REG		0x4
@@ -317,6 +355,8 @@
 #define ROCEE_SDB_ISSUE_PTR_REG			0x758
 #define ROCEE_SDB_SEND_PTR_REG			0x75C
 #define ROCEE_SDB_INV_CNT_REG			0x9A4
+#define ROCEE_SDB_RETRY_CNT_REG			0x9AC
+#define ROCEE_TSP_BP_ST_REG			0x9EC
 #define ROCEE_ECC_UCERR_ALM0_REG		0xB34
 #define ROCEE_ECC_CERR_ALM0_REG			0xB40
 
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
index 125ab90157f6..aee1d01ca70c 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
@@ -948,6 +948,38 @@ int hns_roce_v1_reset(struct hns_roce_dev *hr_dev, bool dereset)
 	return ret;
 }
 
+static int hns_roce_des_qp_init(struct hns_roce_dev *hr_dev)
+{
+	struct device *dev = &hr_dev->pdev->dev;
+	struct hns_roce_v1_priv *priv;
+	struct hns_roce_des_qp *des_qp;
+
+	priv = (struct hns_roce_v1_priv *)hr_dev->hw->priv;
+	des_qp = &priv->des_qp;
+
+	des_qp->requeue_flag = 1;
+	des_qp->qp_wq = create_singlethread_workqueue("hns_roce_destroy_qp");
+	if (!des_qp->qp_wq) {
+		dev_err(dev, "Create destroy qp workqueue failed!\n");
+		return -ENOMEM;
+	}
+
+	return 0;
+}
+
+static void hns_roce_des_qp_free(struct hns_roce_dev *hr_dev)
+{
+	struct hns_roce_v1_priv *priv;
+	struct hns_roce_des_qp *des_qp;
+
+	priv = (struct hns_roce_v1_priv *)hr_dev->hw->priv;
+	des_qp = &priv->des_qp;
+
+	des_qp->requeue_flag = 0;
+	flush_workqueue(des_qp->qp_wq);
+	destroy_workqueue(des_qp->qp_wq);
+}
+
 void hns_roce_v1_profile(struct hns_roce_dev *hr_dev)
 {
 	int i = 0;
@@ -1050,8 +1082,6 @@ int hns_roce_v1_init(struct hns_roce_dev *hr_dev)
 		goto error_failed_raq_init;
 	}
 
-	hns_roce_port_enable(hr_dev, HNS_ROCE_PORT_UP);
-
 	ret = hns_roce_bt_init(hr_dev);
 	if (ret) {
 		dev_err(dev, "bt init failed!\n");
@@ -1064,13 +1094,23 @@ int hns_roce_v1_init(struct hns_roce_dev *hr_dev)
 		goto error_failed_tptr_init;
 	}
 
+	ret = hns_roce_des_qp_init(hr_dev);
+	if (ret) {
+		dev_err(dev, "des qp init failed!\n");
+		goto error_failed_des_qp_init;
+	}
+
+	hns_roce_port_enable(hr_dev, HNS_ROCE_PORT_UP);
+
 	return 0;
 
+error_failed_des_qp_init:
+	hns_roce_tptr_free(hr_dev);
+
 error_failed_tptr_init:
 	hns_roce_bt_free(hr_dev);
 
 error_failed_bt_init:
-	hns_roce_port_enable(hr_dev, HNS_ROCE_PORT_DOWN);
 	hns_roce_raq_free(hr_dev);
 
 error_failed_raq_init:
@@ -1080,9 +1120,10 @@ error_failed_raq_init:
 
 void hns_roce_v1_exit(struct hns_roce_dev *hr_dev)
 {
+	hns_roce_port_enable(hr_dev, HNS_ROCE_PORT_DOWN);
+	hns_roce_des_qp_free(hr_dev);
 	hns_roce_tptr_free(hr_dev);
 	hns_roce_bt_free(hr_dev);
-	hns_roce_port_enable(hr_dev, HNS_ROCE_PORT_DOWN);
 	hns_roce_raq_free(hr_dev);
 	hns_roce_db_free(hr_dev);
 }
@@ -2906,132 +2947,334 @@ int hns_roce_v1_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr,
 		hns_roce_v1_q_sqp(ibqp, qp_attr, qp_attr_mask, qp_init_attr) :
 		hns_roce_v1_q_qp(ibqp, qp_attr, qp_attr_mask, qp_init_attr);
 }
-static void hns_roce_v1_destroy_qp_common(struct hns_roce_dev *hr_dev,
-					  struct hns_roce_qp *hr_qp,
-					  int is_user)
+
+static int check_qp_db_process_status(struct hns_roce_dev *hr_dev,
+				      struct hns_roce_qp *hr_qp,
+				      u32 sdb_issue_ptr,
+				      u32 *sdb_inv_cnt,
+				      u32 *wait_stage)
 {
-	u32 sdbinvcnt;
-	unsigned long end = 0;
-	u32 sdbinvcnt_val;
-	u32 sdbsendptr_val;
-	u32 sdbisusepr_val;
-	struct hns_roce_cq *send_cq, *recv_cq;
 	struct device *dev = &hr_dev->pdev->dev;
+	u32 sdb_retry_cnt, old_retry;
+	u32 sdb_send_ptr, old_send;
+	u32 success_flags = 0;
+	u32 cur_cnt, old_cnt;
+	unsigned long end;
+	u32 send_ptr;
+	u32 inv_cnt;
+	u32 tsp_st;
+
+	if (*wait_stage > HNS_ROCE_V1_DB_STAGE2 ||
+	    *wait_stage < HNS_ROCE_V1_DB_STAGE1) {
+		dev_err(dev, "QP(0x%lx) db status wait stage(%d) error!\n",
+			hr_qp->qpn, *wait_stage);
+		return -EINVAL;
+	}
 
-	if (hr_qp->ibqp.qp_type == IB_QPT_RC) {
-		if (hr_qp->state != IB_QPS_RESET) {
-			/*
-			 * Set qp to ERR,
-			 * waiting for hw complete processing all dbs
-			 */
-			if (hns_roce_v1_qp_modify(hr_dev, NULL,
-					to_hns_roce_state(
-						(enum ib_qp_state)hr_qp->state),
-						HNS_ROCE_QP_STATE_ERR, NULL,
-						hr_qp))
-				dev_err(dev, "modify QP %06lx to ERR failed.\n",
-					hr_qp->qpn);
-
-			/* Record issued doorbell */
-			sdbisusepr_val = roce_read(hr_dev,
-					 ROCEE_SDB_ISSUE_PTR_REG);
-			/*
-			 * Query db process status,
-			 * until hw process completely
-			 */
-			end = msecs_to_jiffies(
-			      HNS_ROCE_QP_DESTROY_TIMEOUT_MSECS) + jiffies;
-			do {
-				sdbsendptr_val = roce_read(hr_dev,
+	/* Calculate the total timeout for the entire verification process */
+	end = msecs_to_jiffies(HNS_ROCE_V1_CHECK_DB_TIMEOUT_MSECS) + jiffies;
+
+	if (*wait_stage == HNS_ROCE_V1_DB_STAGE1) {
+		/* Query db process status, until hw process completely */
+		sdb_send_ptr = roce_read(hr_dev, ROCEE_SDB_SEND_PTR_REG);
+		while (roce_hw_index_cmp_lt(sdb_send_ptr, sdb_issue_ptr,
+					    ROCEE_SDB_PTR_CMP_BITS)) {
+			if (!time_before(jiffies, end)) {
+				dev_dbg(dev, "QP(0x%lx) db process stage1 timeout. issue 0x%x send 0x%x.\n",
+					hr_qp->qpn, sdb_issue_ptr,
+					sdb_send_ptr);
+				return 0;
+			}
+
+			msleep(HNS_ROCE_V1_CHECK_DB_SLEEP_MSECS);
+			sdb_send_ptr = roce_read(hr_dev,
 						 ROCEE_SDB_SEND_PTR_REG);
-				if (!time_before(jiffies, end)) {
-					dev_err(dev, "destroy qp(0x%lx) timeout!!!",
-						hr_qp->qpn);
-					break;
-				}
-			} while ((short)(roce_get_field(sdbsendptr_val,
-					ROCEE_SDB_SEND_PTR_SDB_SEND_PTR_M,
-					ROCEE_SDB_SEND_PTR_SDB_SEND_PTR_S) -
-				roce_get_field(sdbisusepr_val,
-					ROCEE_SDB_ISSUE_PTR_SDB_ISSUE_PTR_M,
-					ROCEE_SDB_ISSUE_PTR_SDB_ISSUE_PTR_S)
-				) < 0);
+		}
 
-			/* Get list pointer */
-			sdbinvcnt = roce_read(hr_dev, ROCEE_SDB_INV_CNT_REG);
+		if (roce_get_field(sdb_issue_ptr,
+				   ROCEE_SDB_ISSUE_PTR_SDB_ISSUE_PTR_M,
+				   ROCEE_SDB_ISSUE_PTR_SDB_ISSUE_PTR_S) ==
+		    roce_get_field(sdb_send_ptr,
+				   ROCEE_SDB_SEND_PTR_SDB_SEND_PTR_M,
+				   ROCEE_SDB_SEND_PTR_SDB_SEND_PTR_S)) {
+			old_send = roce_read(hr_dev, ROCEE_SDB_SEND_PTR_REG);
+			old_retry = roce_read(hr_dev, ROCEE_SDB_RETRY_CNT_REG);
 
-			/* Query db's list status, until hw reversal */
 			do {
-				sdbinvcnt_val = roce_read(hr_dev,
-						ROCEE_SDB_INV_CNT_REG);
+				tsp_st = roce_read(hr_dev, ROCEE_TSP_BP_ST_REG);
+				if (roce_get_bit(tsp_st,
+					ROCEE_TSP_BP_ST_QH_FIFO_ENTRY_S) == 1) {
+					*wait_stage = HNS_ROCE_V1_DB_WAIT_OK;
+					return 0;
+				}
+
 				if (!time_before(jiffies, end)) {
-					dev_err(dev, "destroy qp(0x%lx) timeout!!!",
-						hr_qp->qpn);
-					dev_err(dev, "SdbInvCnt = 0x%x\n",
-						sdbinvcnt_val);
-					break;
+					dev_dbg(dev, "QP(0x%lx) db process stage1 timeout when send ptr equals issue ptr.\n"
+						     "issue 0x%x send 0x%x.\n",
+						hr_qp->qpn, sdb_issue_ptr,
+						sdb_send_ptr);
+					return 0;
 				}
-			} while ((short)(roce_get_field(sdbinvcnt_val,
-				  ROCEE_SDB_INV_CNT_SDB_INV_CNT_M,
-				  ROCEE_SDB_INV_CNT_SDB_INV_CNT_S) -
-				  (sdbinvcnt + SDB_INV_CNT_OFFSET)) < 0);
-
-			/* Modify qp to reset before destroying qp */
-			if (hns_roce_v1_qp_modify(hr_dev, NULL,
-					to_hns_roce_state(
-					(enum ib_qp_state)hr_qp->state),
-					HNS_ROCE_QP_STATE_RST, NULL, hr_qp))
-				dev_err(dev, "modify QP %06lx to RESET failed.\n",
-					hr_qp->qpn);
+
+				msleep(HNS_ROCE_V1_CHECK_DB_SLEEP_MSECS);
+
+				sdb_send_ptr = roce_read(hr_dev,
+							ROCEE_SDB_SEND_PTR_REG);
+				sdb_retry_cnt =	roce_read(hr_dev,
+						       ROCEE_SDB_RETRY_CNT_REG);
+				cur_cnt = roce_get_field(sdb_send_ptr,
+					ROCEE_SDB_SEND_PTR_SDB_SEND_PTR_M,
+					ROCEE_SDB_SEND_PTR_SDB_SEND_PTR_S) +
+					roce_get_field(sdb_retry_cnt,
+					ROCEE_SDB_RETRY_CNT_SDB_RETRY_CT_M,
+					ROCEE_SDB_RETRY_CNT_SDB_RETRY_CT_S);
+				if (!roce_get_bit(tsp_st,
+					ROCEE_CNT_CLR_CE_CNT_CLR_CE_S)) {
+					old_cnt = roce_get_field(old_send,
+					ROCEE_SDB_SEND_PTR_SDB_SEND_PTR_M,
+					ROCEE_SDB_SEND_PTR_SDB_SEND_PTR_S) +
+					roce_get_field(old_retry,
+					ROCEE_SDB_RETRY_CNT_SDB_RETRY_CT_M,
+					ROCEE_SDB_RETRY_CNT_SDB_RETRY_CT_S);
+					if (cur_cnt - old_cnt > SDB_ST_CMP_VAL)
+						success_flags = 1;
+				} else {
+					old_cnt = roce_get_field(old_send,
+					ROCEE_SDB_SEND_PTR_SDB_SEND_PTR_M,
+					ROCEE_SDB_SEND_PTR_SDB_SEND_PTR_S);
+					if (cur_cnt - old_cnt > SDB_ST_CMP_VAL)
+						success_flags = 1;
+					else {
+					    send_ptr = roce_get_field(old_send,
+					    ROCEE_SDB_SEND_PTR_SDB_SEND_PTR_M,
+					    ROCEE_SDB_SEND_PTR_SDB_SEND_PTR_S) +
+					    roce_get_field(sdb_retry_cnt,
+					    ROCEE_SDB_RETRY_CNT_SDB_RETRY_CT_M,
+					    ROCEE_SDB_RETRY_CNT_SDB_RETRY_CT_S);
+					    roce_set_field(old_send,
+					    ROCEE_SDB_SEND_PTR_SDB_SEND_PTR_M,
+					    ROCEE_SDB_SEND_PTR_SDB_SEND_PTR_S,
+						send_ptr);
+					}
+				}
+			} while (!success_flags);
+		}
+
+		*wait_stage = HNS_ROCE_V1_DB_STAGE2;
+
+		/* Get list pointer */
+		*sdb_inv_cnt = roce_read(hr_dev, ROCEE_SDB_INV_CNT_REG);
+		dev_dbg(dev, "QP(0x%lx) db process stage2. inv cnt = 0x%x.\n",
+			hr_qp->qpn, *sdb_inv_cnt);
+	}
+
+	if (*wait_stage == HNS_ROCE_V1_DB_STAGE2) {
+		/* Query db's list status, until hw reversal */
+		inv_cnt = roce_read(hr_dev, ROCEE_SDB_INV_CNT_REG);
+		while (roce_hw_index_cmp_lt(inv_cnt,
+					    *sdb_inv_cnt + SDB_INV_CNT_OFFSET,
+					    ROCEE_SDB_CNT_CMP_BITS)) {
+			if (!time_before(jiffies, end)) {
+				dev_dbg(dev, "QP(0x%lx) db process stage2 timeout. inv cnt 0x%x.\n",
+					hr_qp->qpn, inv_cnt);
+				return 0;
+			}
+
+			msleep(HNS_ROCE_V1_CHECK_DB_SLEEP_MSECS);
+			inv_cnt = roce_read(hr_dev, ROCEE_SDB_INV_CNT_REG);
 		}
+
+		*wait_stage = HNS_ROCE_V1_DB_WAIT_OK;
+	}
+
+	return 0;
+}
+
+static int check_qp_reset_state(struct hns_roce_dev *hr_dev,
+				struct hns_roce_qp *hr_qp,
+				struct hns_roce_qp_work *qp_work_entry,
+				int *is_timeout)
+{
+	struct device *dev = &hr_dev->pdev->dev;
+	u32 sdb_issue_ptr;
+	int ret;
+
+	if (hr_qp->state != IB_QPS_RESET) {
+		/* Set qp to ERR, waiting for hw complete processing all dbs */
+		ret = hns_roce_v1_modify_qp(&hr_qp->ibqp, NULL, 0, hr_qp->state,
+					    IB_QPS_ERR);
+		if (ret) {
+			dev_err(dev, "Modify QP(0x%lx) to ERR failed!\n",
+				hr_qp->qpn);
+			return ret;
+		}
+
+		/* Record issued doorbell */
+		sdb_issue_ptr = roce_read(hr_dev, ROCEE_SDB_ISSUE_PTR_REG);
+		qp_work_entry->sdb_issue_ptr = sdb_issue_ptr;
+		qp_work_entry->db_wait_stage = HNS_ROCE_V1_DB_STAGE1;
+
+		/* Query db process status, until hw process completely */
+		ret = check_qp_db_process_status(hr_dev, hr_qp, sdb_issue_ptr,
+						 &qp_work_entry->sdb_inv_cnt,
+						 &qp_work_entry->db_wait_stage);
+		if (ret) {
+			dev_err(dev, "Check QP(0x%lx) db process status failed!\n",
+				hr_qp->qpn);
+			return ret;
+		}
+
+		if (qp_work_entry->db_wait_stage != HNS_ROCE_V1_DB_WAIT_OK) {
+			qp_work_entry->sche_cnt = 0;
+			*is_timeout = 1;
+			return 0;
+		}
+
+		/* Modify qp to reset before destroying qp */
+		ret = hns_roce_v1_modify_qp(&hr_qp->ibqp, NULL, 0, hr_qp->state,
+					    IB_QPS_RESET);
+		if (ret) {
+			dev_err(dev, "Modify QP(0x%lx) to RST failed!\n",
+				hr_qp->qpn);
+			return ret;
+		}
+	}
+
+	return 0;
+}
+
+static void hns_roce_v1_destroy_qp_work_fn(struct work_struct *work)
+{
+	struct hns_roce_qp_work *qp_work_entry;
+	struct hns_roce_v1_priv *priv;
+	struct hns_roce_dev *hr_dev;
+	struct hns_roce_qp *hr_qp;
+	struct device *dev;
+	int ret;
+
+	qp_work_entry = container_of(work, struct hns_roce_qp_work, work);
+	hr_dev = to_hr_dev(qp_work_entry->ib_dev);
+	dev = &hr_dev->pdev->dev;
+	priv = (struct hns_roce_v1_priv *)hr_dev->hw->priv;
+	hr_qp = qp_work_entry->qp;
+
+	dev_dbg(dev, "Schedule destroy QP(0x%lx) work.\n", hr_qp->qpn);
+
+	qp_work_entry->sche_cnt++;
+
+	/* Query db process status, until hw process completely */
+	ret = check_qp_db_process_status(hr_dev, hr_qp,
+					 qp_work_entry->sdb_issue_ptr,
+					 &qp_work_entry->sdb_inv_cnt,
+					 &qp_work_entry->db_wait_stage);
+	if (ret) {
+		dev_err(dev, "Check QP(0x%lx) db process status failed!\n",
+			hr_qp->qpn);
+		return;
+	}
+
+	if (qp_work_entry->db_wait_stage != HNS_ROCE_V1_DB_WAIT_OK &&
+	    priv->des_qp.requeue_flag) {
+		queue_work(priv->des_qp.qp_wq, work);
+		return;
+	}
+
+	/* Modify qp to reset before destroying qp */
+	ret = hns_roce_v1_modify_qp(&hr_qp->ibqp, NULL, 0, hr_qp->state,
+				    IB_QPS_RESET);
+	if (ret) {
+		dev_err(dev, "Modify QP(0x%lx) to RST failed!\n", hr_qp->qpn);
+		return;
+	}
+
+	hns_roce_qp_remove(hr_dev, hr_qp);
+	hns_roce_qp_free(hr_dev, hr_qp);
+
+	if (hr_qp->ibqp.qp_type == IB_QPT_RC) {
+		/* RC QP, release QPN */
+		hns_roce_release_range_qp(hr_dev, hr_qp->qpn, 1);
+		kfree(hr_qp);
+	} else
+		kfree(hr_to_hr_sqp(hr_qp));
+
+	kfree(qp_work_entry);
+
+	dev_dbg(dev, "Accomplished destroy QP(0x%lx) work.\n", hr_qp->qpn);
+}
+
+int hns_roce_v1_destroy_qp(struct ib_qp *ibqp)
+{
+	struct hns_roce_dev *hr_dev = to_hr_dev(ibqp->device);
+	struct hns_roce_qp *hr_qp = to_hr_qp(ibqp);
+	struct device *dev = &hr_dev->pdev->dev;
+	struct hns_roce_qp_work qp_work_entry;
+	struct hns_roce_qp_work *qp_work;
+	struct hns_roce_v1_priv *priv;
+	struct hns_roce_cq *send_cq, *recv_cq;
+	int is_user = !!ibqp->pd->uobject;
+	int is_timeout = 0;
+	int ret;
+
+	ret = check_qp_reset_state(hr_dev, hr_qp, &qp_work_entry, &is_timeout);
+	if (ret) {
+		dev_err(dev, "QP reset state check failed(%d)!\n", ret);
+		return ret;
 	}
 
 	send_cq = to_hr_cq(hr_qp->ibqp.send_cq);
 	recv_cq = to_hr_cq(hr_qp->ibqp.recv_cq);
 
 	hns_roce_lock_cqs(send_cq, recv_cq);
-
 	if (!is_user) {
 		__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)
 			__hns_roce_v1_cq_clean(send_cq, hr_qp->qpn, NULL);
 	}
-
-	hns_roce_qp_remove(hr_dev, hr_qp);
-
 	hns_roce_unlock_cqs(send_cq, recv_cq);
 
-	hns_roce_qp_free(hr_dev, hr_qp);
+	if (!is_timeout) {
+		hns_roce_qp_remove(hr_dev, hr_qp);
+		hns_roce_qp_free(hr_dev, hr_qp);
 
-	/* Not special_QP, free their QPN */
-	if ((hr_qp->ibqp.qp_type == IB_QPT_RC) ||
-	    (hr_qp->ibqp.qp_type == IB_QPT_UC) ||
-	    (hr_qp->ibqp.qp_type == IB_QPT_UD))
-		hns_roce_release_range_qp(hr_dev, hr_qp->qpn, 1);
+		/* RC QP, release QPN */
+		if (hr_qp->ibqp.qp_type == IB_QPT_RC)
+			hns_roce_release_range_qp(hr_dev, hr_qp->qpn, 1);
+	}
 
 	hns_roce_mtt_cleanup(hr_dev, &hr_qp->mtt);
 
-	if (is_user) {
+	if (is_user)
 		ib_umem_release(hr_qp->umem);
-	} else {
+	else {
 		kfree(hr_qp->sq.wrid);
 		kfree(hr_qp->rq.wrid);
+
 		hns_roce_buf_free(hr_dev, hr_qp->buff_size, &hr_qp->hr_buf);
 	}
-}
 
-int hns_roce_v1_destroy_qp(struct ib_qp *ibqp)
-{
-	struct hns_roce_dev *hr_dev = to_hr_dev(ibqp->device);
-	struct hns_roce_qp *hr_qp = to_hr_qp(ibqp);
-
-	hns_roce_v1_destroy_qp_common(hr_dev, hr_qp, !!ibqp->pd->uobject);
-
-	if (hr_qp->ibqp.qp_type == IB_QPT_GSI)
-		kfree(hr_to_hr_sqp(hr_qp));
-	else
-		kfree(hr_qp);
+	if (!is_timeout) {
+		if (hr_qp->ibqp.qp_type == IB_QPT_RC)
+			kfree(hr_qp);
+		else
+			kfree(hr_to_hr_sqp(hr_qp));
+	} else {
+		qp_work = kzalloc(sizeof(*qp_work), GFP_KERNEL);
+		if (!qp_work)
+			return -ENOMEM;
+
+		INIT_WORK(&qp_work->work, hns_roce_v1_destroy_qp_work_fn);
+		qp_work->ib_dev	= &hr_dev->ib_dev;
+		qp_work->qp		= hr_qp;
+		qp_work->db_wait_stage	= qp_work_entry.db_wait_stage;
+		qp_work->sdb_issue_ptr	= qp_work_entry.sdb_issue_ptr;
+		qp_work->sdb_inv_cnt	= qp_work_entry.sdb_inv_cnt;
+		qp_work->sche_cnt	= qp_work_entry.sche_cnt;
+
+		priv = (struct hns_roce_v1_priv *)hr_dev->hw->priv;
+		queue_work(priv->des_qp.qp_wq, &qp_work->work);
+		dev_dbg(dev, "Begin destroy QP(0x%lx) work.\n", hr_qp->qpn);
+	}
 
 	return 0;
 }
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.h b/drivers/infiniband/hw/hns/hns_roce_hw_v1.h
index cf28f1b6492c..1d250c026c10 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.h
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.h
@@ -102,6 +102,12 @@
 #define HNS_ROCE_V1_EXT_ODB_ALFUL	\
 	(HNS_ROCE_V1_EXT_ODB_DEPTH - HNS_ROCE_V1_DB_RSVD)
 
+#define HNS_ROCE_V1_DB_WAIT_OK				0
+#define HNS_ROCE_V1_DB_STAGE1				1
+#define HNS_ROCE_V1_DB_STAGE2				2
+#define HNS_ROCE_V1_CHECK_DB_TIMEOUT_MSECS		10000
+#define HNS_ROCE_V1_CHECK_DB_SLEEP_MSECS		20
+
 #define HNS_ROCE_BT_RSV_BUF_SIZE			(1 << 17)
 
 #define HNS_ROCE_V1_TPTR_ENTRY_SIZE			2
@@ -144,6 +150,7 @@
 #define SQ_PSN_SHIFT					8
 #define QKEY_VAL					0x80010000
 #define SDB_INV_CNT_OFFSET				8
+#define SDB_ST_CMP_VAL					8
 
 struct hns_roce_cq_context {
 	u32 cqc_byte_4;
@@ -993,11 +1000,27 @@ struct hns_roce_tptr_table {
 	struct hns_roce_buf_list tptr_buf;
 };
 
+struct hns_roce_qp_work {
+	struct	work_struct work;
+	struct	ib_device *ib_dev;
+	struct	hns_roce_qp *qp;
+	u32	db_wait_stage;
+	u32	sdb_issue_ptr;
+	u32	sdb_inv_cnt;
+	u32	sche_cnt;
+};
+
+struct hns_roce_des_qp {
+	struct workqueue_struct	*qp_wq;
+	int	requeue_flag;
+};
+
 struct hns_roce_v1_priv {
 	struct hns_roce_db_table  db_table;
 	struct hns_roce_raq_table raq_table;
 	struct hns_roce_bt_table  bt_table;
 	struct hns_roce_tptr_table tptr_table;
+	struct hns_roce_des_qp des_qp;
 };
 
 int hns_dsaf_roce_reset(struct fwnode_handle *dsaf_fwnode, bool dereset);
-- 
cgit v1.2.1


From bfcc681bd09d6cd96aa0ec667533a867253731aa Mon Sep 17 00:00:00 2001
From: Shaobo Xu <xushaobo2@huawei.com>
Date: Tue, 29 Nov 2016 23:10:26 +0000
Subject: IB/hns: Fix the bug when free mr

If the resources of mr are freed while executing the user case, hardware
can not been notified in hip06 SoC. Then hardware will hold on when it
reads the payload by the PA which has been released.

In order to slove this problem, RoCE driver creates 8 reserved loopback
QPs to ensure zero wqe when free mr. When the mac address is reset, in
order to avoid loopback failure, we need to release the reserved loopback
QPs and recreate them.

Signed-off-by: Shaobo Xu <xushaobo2@huawei.com>
Reviewed-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Signed-off-by: Salil Mehta <salil.mehta@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_cmd.h    |   5 -
 drivers/infiniband/hw/hns/hns_roce_device.h |  10 +
 drivers/infiniband/hw/hns/hns_roce_hw_v1.c  | 485 ++++++++++++++++++++++++++++
 drivers/infiniband/hw/hns/hns_roce_hw_v1.h  |  34 ++
 drivers/infiniband/hw/hns/hns_roce_main.c   |   5 +-
 drivers/infiniband/hw/hns/hns_roce_mr.c     |  21 +-
 6 files changed, 545 insertions(+), 15 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hns/hns_roce_cmd.h b/drivers/infiniband/hw/hns/hns_roce_cmd.h
index ed14ad3d06ef..f5a9ee2fc53d 100644
--- a/drivers/infiniband/hw/hns/hns_roce_cmd.h
+++ b/drivers/infiniband/hw/hns/hns_roce_cmd.h
@@ -58,11 +58,6 @@ enum {
 	HNS_ROCE_CMD_QUERY_QP		= 0x22,
 };
 
-struct hns_roce_cmd_mailbox {
-	void		       *buf;
-	dma_addr_t		dma;
-};
-
 int hns_roce_cmd_mbox(struct hns_roce_dev *hr_dev, u64 in_param, u64 out_param,
 		      unsigned long in_modifier, u8 op_modifier, u16 op,
 		      unsigned long timeout);
diff --git a/drivers/infiniband/hw/hns/hns_roce_device.h b/drivers/infiniband/hw/hns/hns_roce_device.h
index e48464ddd2b0..1050829da642 100644
--- a/drivers/infiniband/hw/hns/hns_roce_device.h
+++ b/drivers/infiniband/hw/hns/hns_roce_device.h
@@ -388,6 +388,11 @@ struct hns_roce_cmdq {
 	u8			toggle;
 };
 
+struct hns_roce_cmd_mailbox {
+	void		       *buf;
+	dma_addr_t		dma;
+};
+
 struct hns_roce_dev;
 
 struct hns_roce_qp {
@@ -522,6 +527,7 @@ struct hns_roce_hw {
 			 struct ib_recv_wr **bad_recv_wr);
 	int (*req_notify_cq)(struct ib_cq *ibcq, enum ib_cq_notify_flags flags);
 	int (*poll_cq)(struct ib_cq *ibcq, int num_entries, struct ib_wc *wc);
+	int (*dereg_mr)(struct hns_roce_dev *hr_dev, struct hns_roce_mr *mr);
 	void	*priv;
 };
 
@@ -688,6 +694,10 @@ struct ib_mr *hns_roce_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 				   u64 virt_addr, int access_flags,
 				   struct ib_udata *udata);
 int hns_roce_dereg_mr(struct ib_mr *ibmr);
+int hns_roce_hw2sw_mpt(struct hns_roce_dev *hr_dev,
+		       struct hns_roce_cmd_mailbox *mailbox,
+		       unsigned long mpt_index);
+unsigned long key_to_hw_index(u32 key);
 
 void hns_roce_buf_free(struct hns_roce_dev *hr_dev, u32 size,
 		       struct hns_roce_buf *buf);
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
index aee1d01ca70c..f67a3bfd4c55 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
@@ -295,6 +295,8 @@ out:
 		roce_set_field(sq_db.u32_4, SQ_DOORBELL_U32_4_SQ_HEAD_M,
 			       SQ_DOORBELL_U32_4_SQ_HEAD_S,
 			      (qp->sq.head & ((qp->sq.wqe_cnt << 1) - 1)));
+		roce_set_field(sq_db.u32_4, SQ_DOORBELL_U32_4_SL_M,
+			       SQ_DOORBELL_U32_4_SL_S, qp->sl);
 		roce_set_field(sq_db.u32_4, SQ_DOORBELL_U32_4_PORT_M,
 			       SQ_DOORBELL_U32_4_PORT_S, qp->phy_port);
 		roce_set_field(sq_db.u32_8, SQ_DOORBELL_U32_8_QPN_M,
@@ -622,6 +624,213 @@ ext_sdb_buf_fail_out:
 	return ret;
 }
 
+static struct hns_roce_qp *hns_roce_v1_create_lp_qp(struct hns_roce_dev *hr_dev,
+						    struct ib_pd *pd)
+{
+	struct device *dev = &hr_dev->pdev->dev;
+	struct ib_qp_init_attr init_attr;
+	struct ib_qp *qp;
+
+	memset(&init_attr, 0, sizeof(struct ib_qp_init_attr));
+	init_attr.qp_type		= IB_QPT_RC;
+	init_attr.sq_sig_type		= IB_SIGNAL_ALL_WR;
+	init_attr.cap.max_recv_wr	= HNS_ROCE_MIN_WQE_NUM;
+	init_attr.cap.max_send_wr	= HNS_ROCE_MIN_WQE_NUM;
+
+	qp = hns_roce_create_qp(pd, &init_attr, NULL);
+	if (IS_ERR(qp)) {
+		dev_err(dev, "Create loop qp for mr free failed!");
+		return NULL;
+	}
+
+	return to_hr_qp(qp);
+}
+
+static int hns_roce_v1_rsv_lp_qp(struct hns_roce_dev *hr_dev)
+{
+	struct hns_roce_caps *caps = &hr_dev->caps;
+	struct device *dev = &hr_dev->pdev->dev;
+	struct ib_cq_init_attr cq_init_attr;
+	struct hns_roce_free_mr *free_mr;
+	struct ib_qp_attr attr = { 0 };
+	struct hns_roce_v1_priv *priv;
+	struct hns_roce_qp *hr_qp;
+	struct ib_cq *cq;
+	struct ib_pd *pd;
+	u64 subnet_prefix;
+	int attr_mask = 0;
+	int i;
+	int ret;
+	u8 phy_port;
+	u8 sl;
+
+	priv = (struct hns_roce_v1_priv *)hr_dev->hw->priv;
+	free_mr = &priv->free_mr;
+
+	/* Reserved cq for loop qp */
+	cq_init_attr.cqe		= HNS_ROCE_MIN_WQE_NUM * 2;
+	cq_init_attr.comp_vector	= 0;
+	cq = hns_roce_ib_create_cq(&hr_dev->ib_dev, &cq_init_attr, NULL, NULL);
+	if (IS_ERR(cq)) {
+		dev_err(dev, "Create cq for reseved loop qp failed!");
+		return -ENOMEM;
+	}
+	free_mr->mr_free_cq = to_hr_cq(cq);
+	free_mr->mr_free_cq->ib_cq.device		= &hr_dev->ib_dev;
+	free_mr->mr_free_cq->ib_cq.uobject		= NULL;
+	free_mr->mr_free_cq->ib_cq.comp_handler		= NULL;
+	free_mr->mr_free_cq->ib_cq.event_handler	= NULL;
+	free_mr->mr_free_cq->ib_cq.cq_context		= NULL;
+	atomic_set(&free_mr->mr_free_cq->ib_cq.usecnt, 0);
+
+	pd = hns_roce_alloc_pd(&hr_dev->ib_dev, NULL, NULL);
+	if (IS_ERR(pd)) {
+		dev_err(dev, "Create pd for reseved loop qp failed!");
+		ret = -ENOMEM;
+		goto alloc_pd_failed;
+	}
+	free_mr->mr_free_pd = to_hr_pd(pd);
+	free_mr->mr_free_pd->ibpd.device  = &hr_dev->ib_dev;
+	free_mr->mr_free_pd->ibpd.uobject = NULL;
+	atomic_set(&free_mr->mr_free_pd->ibpd.usecnt, 0);
+
+	attr.qp_access_flags	= IB_ACCESS_REMOTE_WRITE;
+	attr.pkey_index		= 0;
+	attr.min_rnr_timer	= 0;
+	/* Disable read ability */
+	attr.max_dest_rd_atomic = 0;
+	attr.max_rd_atomic	= 0;
+	/* Use arbitrary values as rq_psn and sq_psn */
+	attr.rq_psn		= 0x0808;
+	attr.sq_psn		= 0x0808;
+	attr.retry_cnt		= 7;
+	attr.rnr_retry		= 7;
+	attr.timeout		= 0x12;
+	attr.path_mtu		= IB_MTU_256;
+	attr.ah_attr.ah_flags		= 1;
+	attr.ah_attr.static_rate	= 3;
+	attr.ah_attr.grh.sgid_index	= 0;
+	attr.ah_attr.grh.hop_limit	= 1;
+	attr.ah_attr.grh.flow_label	= 0;
+	attr.ah_attr.grh.traffic_class	= 0;
+
+	subnet_prefix = cpu_to_be64(0xfe80000000000000LL);
+	for (i = 0; i < HNS_ROCE_V1_RESV_QP; i++) {
+		free_mr->mr_free_qp[i] = hns_roce_v1_create_lp_qp(hr_dev, pd);
+		if (IS_ERR(free_mr->mr_free_qp[i])) {
+			dev_err(dev, "Create loop qp failed!\n");
+			goto create_lp_qp_failed;
+		}
+		hr_qp = free_mr->mr_free_qp[i];
+
+		sl = i / caps->num_ports;
+
+		if (caps->num_ports == HNS_ROCE_MAX_PORTS)
+			phy_port = (i >= HNS_ROCE_MAX_PORTS) ? (i - 2) :
+				(i % caps->num_ports);
+		else
+			phy_port = i % caps->num_ports;
+
+		hr_qp->port		= phy_port + 1;
+		hr_qp->phy_port		= phy_port;
+		hr_qp->ibqp.qp_type	= IB_QPT_RC;
+		hr_qp->ibqp.device	= &hr_dev->ib_dev;
+		hr_qp->ibqp.uobject	= NULL;
+		atomic_set(&hr_qp->ibqp.usecnt, 0);
+		hr_qp->ibqp.pd		= pd;
+		hr_qp->ibqp.recv_cq	= cq;
+		hr_qp->ibqp.send_cq	= cq;
+
+		attr.ah_attr.port_num	= phy_port + 1;
+		attr.ah_attr.sl		= sl;
+		attr.port_num		= phy_port + 1;
+
+		attr.dest_qp_num	= hr_qp->qpn;
+		memcpy(attr.ah_attr.dmac, hr_dev->dev_addr[phy_port],
+		       MAC_ADDR_OCTET_NUM);
+
+		memcpy(attr.ah_attr.grh.dgid.raw,
+			&subnet_prefix, sizeof(u64));
+		memcpy(&attr.ah_attr.grh.dgid.raw[8],
+		       hr_dev->dev_addr[phy_port], 3);
+		memcpy(&attr.ah_attr.grh.dgid.raw[13],
+		       hr_dev->dev_addr[phy_port] + 3, 3);
+		attr.ah_attr.grh.dgid.raw[11] = 0xff;
+		attr.ah_attr.grh.dgid.raw[12] = 0xfe;
+		attr.ah_attr.grh.dgid.raw[8] ^= 2;
+
+		attr_mask |= IB_QP_PORT;
+
+		ret = hr_dev->hw->modify_qp(&hr_qp->ibqp, &attr, attr_mask,
+					    IB_QPS_RESET, IB_QPS_INIT);
+		if (ret) {
+			dev_err(dev, "modify qp failed(%d)!\n", ret);
+			goto create_lp_qp_failed;
+		}
+
+		ret = hr_dev->hw->modify_qp(&hr_qp->ibqp, &attr, attr_mask,
+					    IB_QPS_INIT, IB_QPS_RTR);
+		if (ret) {
+			dev_err(dev, "modify qp failed(%d)!\n", ret);
+			goto create_lp_qp_failed;
+		}
+
+		ret = hr_dev->hw->modify_qp(&hr_qp->ibqp, &attr, attr_mask,
+					    IB_QPS_RTR, IB_QPS_RTS);
+		if (ret) {
+			dev_err(dev, "modify qp failed(%d)!\n", ret);
+			goto create_lp_qp_failed;
+		}
+	}
+
+	return 0;
+
+create_lp_qp_failed:
+	for (i -= 1; i >= 0; i--) {
+		hr_qp = free_mr->mr_free_qp[i];
+		if (hns_roce_v1_destroy_qp(&hr_qp->ibqp))
+			dev_err(dev, "Destroy qp %d for mr free failed!\n", i);
+	}
+
+	if (hns_roce_dealloc_pd(pd))
+		dev_err(dev, "Destroy pd for create_lp_qp failed!\n");
+
+alloc_pd_failed:
+	if (hns_roce_ib_destroy_cq(cq))
+		dev_err(dev, "Destroy cq for create_lp_qp failed!\n");
+
+	return -EINVAL;
+}
+
+static void hns_roce_v1_release_lp_qp(struct hns_roce_dev *hr_dev)
+{
+	struct device *dev = &hr_dev->pdev->dev;
+	struct hns_roce_free_mr *free_mr;
+	struct hns_roce_v1_priv *priv;
+	struct hns_roce_qp *hr_qp;
+	int ret;
+	int i;
+
+	priv = (struct hns_roce_v1_priv *)hr_dev->hw->priv;
+	free_mr = &priv->free_mr;
+
+	for (i = 0; i < HNS_ROCE_V1_RESV_QP; i++) {
+		hr_qp = free_mr->mr_free_qp[i];
+		ret = hns_roce_v1_destroy_qp(&hr_qp->ibqp);
+		if (ret)
+			dev_err(dev, "Destroy qp %d for mr free failed(%d)!\n",
+				i, ret);
+	}
+
+	ret = hns_roce_ib_destroy_cq(&free_mr->mr_free_cq->ib_cq);
+	if (ret)
+		dev_err(dev, "Destroy cq for mr_free failed(%d)!\n", ret);
+
+	ret = hns_roce_dealloc_pd(&free_mr->mr_free_pd->ibpd);
+	if (ret)
+		dev_err(dev, "Destroy pd for mr_free failed(%d)!\n", ret);
+}
+
 static int hns_roce_db_init(struct hns_roce_dev *hr_dev)
 {
 	struct device *dev = &hr_dev->pdev->dev;
@@ -659,6 +868,223 @@ static int hns_roce_db_init(struct hns_roce_dev *hr_dev)
 	return 0;
 }
 
+void hns_roce_v1_recreate_lp_qp_work_fn(struct work_struct *work)
+{
+	struct hns_roce_recreate_lp_qp_work *lp_qp_work;
+	struct hns_roce_dev *hr_dev;
+
+	lp_qp_work = container_of(work, struct hns_roce_recreate_lp_qp_work,
+				  work);
+	hr_dev = to_hr_dev(lp_qp_work->ib_dev);
+
+	hns_roce_v1_release_lp_qp(hr_dev);
+
+	if (hns_roce_v1_rsv_lp_qp(hr_dev))
+		dev_err(&hr_dev->pdev->dev, "create reserver qp failed\n");
+
+	if (lp_qp_work->comp_flag)
+		complete(lp_qp_work->comp);
+
+	kfree(lp_qp_work);
+}
+
+static int hns_roce_v1_recreate_lp_qp(struct hns_roce_dev *hr_dev)
+{
+	struct device *dev = &hr_dev->pdev->dev;
+	struct hns_roce_recreate_lp_qp_work *lp_qp_work;
+	struct hns_roce_free_mr *free_mr;
+	struct hns_roce_v1_priv *priv;
+	struct completion comp;
+	unsigned long end =
+	  msecs_to_jiffies(HNS_ROCE_V1_RECREATE_LP_QP_TIMEOUT_MSECS) + jiffies;
+
+	priv = (struct hns_roce_v1_priv *)hr_dev->hw->priv;
+	free_mr = &priv->free_mr;
+
+	lp_qp_work = kzalloc(sizeof(struct hns_roce_recreate_lp_qp_work),
+			     GFP_KERNEL);
+
+	INIT_WORK(&(lp_qp_work->work), hns_roce_v1_recreate_lp_qp_work_fn);
+
+	lp_qp_work->ib_dev = &(hr_dev->ib_dev);
+	lp_qp_work->comp = &comp;
+	lp_qp_work->comp_flag = 1;
+
+	init_completion(lp_qp_work->comp);
+
+	queue_work(free_mr->free_mr_wq, &(lp_qp_work->work));
+
+	while (time_before_eq(jiffies, end)) {
+		if (try_wait_for_completion(&comp))
+			return 0;
+		msleep(HNS_ROCE_V1_RECREATE_LP_QP_WAIT_VALUE);
+	}
+
+	lp_qp_work->comp_flag = 0;
+	if (try_wait_for_completion(&comp))
+		return 0;
+
+	dev_warn(dev, "recreate lp qp failed 20s timeout and return failed!\n");
+	return -ETIMEDOUT;
+}
+
+static int hns_roce_v1_send_lp_wqe(struct hns_roce_qp *hr_qp)
+{
+	struct hns_roce_dev *hr_dev = to_hr_dev(hr_qp->ibqp.device);
+	struct device *dev = &hr_dev->pdev->dev;
+	struct ib_send_wr send_wr, *bad_wr;
+	int ret;
+
+	memset(&send_wr, 0, sizeof(send_wr));
+	send_wr.next	= NULL;
+	send_wr.num_sge	= 0;
+	send_wr.send_flags = 0;
+	send_wr.sg_list	= NULL;
+	send_wr.wr_id	= (unsigned long long)&send_wr;
+	send_wr.opcode	= IB_WR_RDMA_WRITE;
+
+	ret = hns_roce_v1_post_send(&hr_qp->ibqp, &send_wr, &bad_wr);
+	if (ret) {
+		dev_err(dev, "Post write wqe for mr free failed(%d)!", ret);
+		return ret;
+	}
+
+	return 0;
+}
+
+static void hns_roce_v1_mr_free_work_fn(struct work_struct *work)
+{
+	struct hns_roce_mr_free_work *mr_work;
+	struct ib_wc wc[HNS_ROCE_V1_RESV_QP];
+	struct hns_roce_free_mr *free_mr;
+	struct hns_roce_cq *mr_free_cq;
+	struct hns_roce_v1_priv *priv;
+	struct hns_roce_dev *hr_dev;
+	struct hns_roce_mr *hr_mr;
+	struct hns_roce_qp *hr_qp;
+	struct device *dev;
+	unsigned long end =
+		msecs_to_jiffies(HNS_ROCE_V1_FREE_MR_TIMEOUT_MSECS) + jiffies;
+	int i;
+	int ret;
+	int ne;
+
+	mr_work = container_of(work, struct hns_roce_mr_free_work, work);
+	hr_mr = (struct hns_roce_mr *)mr_work->mr;
+	hr_dev = to_hr_dev(mr_work->ib_dev);
+	dev = &hr_dev->pdev->dev;
+
+	priv = (struct hns_roce_v1_priv *)hr_dev->hw->priv;
+	free_mr = &priv->free_mr;
+	mr_free_cq = free_mr->mr_free_cq;
+
+	for (i = 0; i < HNS_ROCE_V1_RESV_QP; i++) {
+		hr_qp = free_mr->mr_free_qp[i];
+		ret = hns_roce_v1_send_lp_wqe(hr_qp);
+		if (ret) {
+			dev_err(dev,
+			     "Send wqe (qp:0x%lx) for mr free failed(%d)!\n",
+			     hr_qp->qpn, ret);
+			goto free_work;
+		}
+	}
+
+	ne = HNS_ROCE_V1_RESV_QP;
+	do {
+		ret = hns_roce_v1_poll_cq(&mr_free_cq->ib_cq, ne, wc);
+		if (ret < 0) {
+			dev_err(dev,
+			   "(qp:0x%lx) starts, Poll cqe failed(%d) for mr 0x%x free! Remain %d cqe\n",
+			   hr_qp->qpn, ret, hr_mr->key, ne);
+			goto free_work;
+		}
+		ne -= ret;
+		msleep(HNS_ROCE_V1_FREE_MR_WAIT_VALUE);
+	} while (ne && time_before_eq(jiffies, end));
+
+	if (ne != 0)
+		dev_err(dev,
+			"Poll cqe for mr 0x%x free timeout! Remain %d cqe\n",
+			hr_mr->key, ne);
+
+free_work:
+	if (mr_work->comp_flag)
+		complete(mr_work->comp);
+	kfree(mr_work);
+}
+
+int hns_roce_v1_dereg_mr(struct hns_roce_dev *hr_dev, struct hns_roce_mr *mr)
+{
+	struct device *dev = &hr_dev->pdev->dev;
+	struct hns_roce_mr_free_work *mr_work;
+	struct hns_roce_free_mr *free_mr;
+	struct hns_roce_v1_priv *priv;
+	struct completion comp;
+	unsigned long end =
+		msecs_to_jiffies(HNS_ROCE_V1_FREE_MR_TIMEOUT_MSECS) + jiffies;
+	unsigned long start = jiffies;
+	int npages;
+	int ret = 0;
+
+	priv = (struct hns_roce_v1_priv *)hr_dev->hw->priv;
+	free_mr = &priv->free_mr;
+
+	if (mr->enabled) {
+		if (hns_roce_hw2sw_mpt(hr_dev, NULL, key_to_hw_index(mr->key)
+				       & (hr_dev->caps.num_mtpts - 1)))
+			dev_warn(dev, "HW2SW_MPT failed!\n");
+	}
+
+	mr_work = kzalloc(sizeof(*mr_work), GFP_KERNEL);
+	if (!mr_work) {
+		ret = -ENOMEM;
+		goto free_mr;
+	}
+
+	INIT_WORK(&(mr_work->work), hns_roce_v1_mr_free_work_fn);
+
+	mr_work->ib_dev = &(hr_dev->ib_dev);
+	mr_work->comp = &comp;
+	mr_work->comp_flag = 1;
+	mr_work->mr = (void *)mr;
+	init_completion(mr_work->comp);
+
+	queue_work(free_mr->free_mr_wq, &(mr_work->work));
+
+	while (time_before_eq(jiffies, end)) {
+		if (try_wait_for_completion(&comp))
+			goto free_mr;
+		msleep(HNS_ROCE_V1_FREE_MR_WAIT_VALUE);
+	}
+
+	mr_work->comp_flag = 0;
+	if (try_wait_for_completion(&comp))
+		goto free_mr;
+
+	dev_warn(dev, "Free mr work 0x%x over 50s and failed!\n", mr->key);
+	ret = -ETIMEDOUT;
+
+free_mr:
+	dev_dbg(dev, "Free mr 0x%x use 0x%x us.\n",
+		mr->key, jiffies_to_usecs(jiffies) - jiffies_to_usecs(start));
+
+	if (mr->size != ~0ULL) {
+		npages = ib_umem_page_count(mr->umem);
+		dma_free_coherent(dev, npages * 8, mr->pbl_buf,
+				  mr->pbl_dma_addr);
+	}
+
+	hns_roce_bitmap_free(&hr_dev->mr_table.mtpt_bitmap,
+			     key_to_hw_index(mr->key), 0);
+
+	if (mr->umem)
+		ib_umem_release(mr->umem);
+
+	kfree(mr);
+
+	return ret;
+}
+
 static void hns_roce_db_free(struct hns_roce_dev *hr_dev)
 {
 	struct device *dev = &hr_dev->pdev->dev;
@@ -899,6 +1325,46 @@ static void hns_roce_tptr_free(struct hns_roce_dev *hr_dev)
 			  tptr_buf->buf, tptr_buf->map);
 }
 
+static int hns_roce_free_mr_init(struct hns_roce_dev *hr_dev)
+{
+	struct device *dev = &hr_dev->pdev->dev;
+	struct hns_roce_free_mr *free_mr;
+	struct hns_roce_v1_priv *priv;
+	int ret = 0;
+
+	priv = (struct hns_roce_v1_priv *)hr_dev->hw->priv;
+	free_mr = &priv->free_mr;
+
+	free_mr->free_mr_wq = create_singlethread_workqueue("hns_roce_free_mr");
+	if (!free_mr->free_mr_wq) {
+		dev_err(dev, "Create free mr workqueue failed!\n");
+		return -ENOMEM;
+	}
+
+	ret = hns_roce_v1_rsv_lp_qp(hr_dev);
+	if (ret) {
+		dev_err(dev, "Reserved loop qp failed(%d)!\n", ret);
+		flush_workqueue(free_mr->free_mr_wq);
+		destroy_workqueue(free_mr->free_mr_wq);
+	}
+
+	return ret;
+}
+
+static void hns_roce_free_mr_free(struct hns_roce_dev *hr_dev)
+{
+	struct hns_roce_free_mr *free_mr;
+	struct hns_roce_v1_priv *priv;
+
+	priv = (struct hns_roce_v1_priv *)hr_dev->hw->priv;
+	free_mr = &priv->free_mr;
+
+	flush_workqueue(free_mr->free_mr_wq);
+	destroy_workqueue(free_mr->free_mr_wq);
+
+	hns_roce_v1_release_lp_qp(hr_dev);
+}
+
 /**
  * hns_roce_v1_reset - reset RoCE
  * @hr_dev: RoCE device struct pointer
@@ -1100,10 +1566,19 @@ int hns_roce_v1_init(struct hns_roce_dev *hr_dev)
 		goto error_failed_des_qp_init;
 	}
 
+	ret = hns_roce_free_mr_init(hr_dev);
+	if (ret) {
+		dev_err(dev, "free mr init failed!\n");
+		goto error_failed_free_mr_init;
+	}
+
 	hns_roce_port_enable(hr_dev, HNS_ROCE_PORT_UP);
 
 	return 0;
 
+error_failed_free_mr_init:
+	hns_roce_des_qp_free(hr_dev);
+
 error_failed_des_qp_init:
 	hns_roce_tptr_free(hr_dev);
 
@@ -1121,6 +1596,7 @@ error_failed_raq_init:
 void hns_roce_v1_exit(struct hns_roce_dev *hr_dev)
 {
 	hns_roce_port_enable(hr_dev, HNS_ROCE_PORT_DOWN);
+	hns_roce_free_mr_free(hr_dev);
 	hns_roce_des_qp_free(hr_dev);
 	hns_roce_tptr_free(hr_dev);
 	hns_roce_bt_free(hr_dev);
@@ -1161,6 +1637,14 @@ void hns_roce_v1_set_mac(struct hns_roce_dev *hr_dev, u8 phy_port, u8 *addr)
 	u32 *p;
 	u32 val;
 
+	/*
+	 * When mac changed, loopback may fail
+	 * because of smac not equal to dmac.
+	 * We Need to release and create reserved qp again.
+	 */
+	if (hr_dev->hw->dereg_mr && hns_roce_v1_recreate_lp_qp(hr_dev))
+		dev_warn(&hr_dev->pdev->dev, "recreate lp qp timeout!\n");
+
 	p = (u32 *)(&addr[0]);
 	reg_smac_l = *p;
 	roce_raw_write(reg_smac_l, hr_dev->reg_base + ROCEE_SMAC_L_0_REG +
@@ -3299,5 +3783,6 @@ struct hns_roce_hw hns_roce_hw_v1 = {
 	.post_recv = hns_roce_v1_post_recv,
 	.req_notify_cq = hns_roce_v1_req_notify_cq,
 	.poll_cq = hns_roce_v1_poll_cq,
+	.dereg_mr = hns_roce_v1_dereg_mr,
 	.priv = &hr_v1_priv,
 };
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.h b/drivers/infiniband/hw/hns/hns_roce_hw_v1.h
index 1d250c026c10..b213b5e6fef1 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.h
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.h
@@ -58,6 +58,7 @@
 #define HNS_ROCE_V1_PHY_UAR_NUM				8
 
 #define HNS_ROCE_V1_GID_NUM				16
+#define HNS_ROCE_V1_RESV_QP				8
 
 #define HNS_ROCE_V1_NUM_COMP_EQE			0x8000
 #define HNS_ROCE_V1_NUM_ASYNC_EQE			0x400
@@ -107,6 +108,10 @@
 #define HNS_ROCE_V1_DB_STAGE2				2
 #define HNS_ROCE_V1_CHECK_DB_TIMEOUT_MSECS		10000
 #define HNS_ROCE_V1_CHECK_DB_SLEEP_MSECS		20
+#define HNS_ROCE_V1_FREE_MR_TIMEOUT_MSECS		50000
+#define HNS_ROCE_V1_RECREATE_LP_QP_TIMEOUT_MSECS	10000
+#define HNS_ROCE_V1_FREE_MR_WAIT_VALUE			5
+#define HNS_ROCE_V1_RECREATE_LP_QP_WAIT_VALUE		20
 
 #define HNS_ROCE_BT_RSV_BUF_SIZE			(1 << 17)
 
@@ -969,6 +974,10 @@ struct hns_roce_sq_db {
 #define SQ_DOORBELL_U32_4_SQ_HEAD_M   \
 	(((1UL << 15) - 1) << SQ_DOORBELL_U32_4_SQ_HEAD_S)
 
+#define SQ_DOORBELL_U32_4_SL_S 16
+#define SQ_DOORBELL_U32_4_SL_M   \
+	(((1UL << 2) - 1) << SQ_DOORBELL_U32_4_SL_S)
+
 #define SQ_DOORBELL_U32_4_PORT_S 18
 #define SQ_DOORBELL_U32_4_PORT_M  (((1UL << 3) - 1) << SQ_DOORBELL_U32_4_PORT_S)
 
@@ -1015,14 +1024,39 @@ struct hns_roce_des_qp {
 	int	requeue_flag;
 };
 
+struct hns_roce_mr_free_work {
+	struct	work_struct work;
+	struct	ib_device *ib_dev;
+	struct	completion *comp;
+	int	comp_flag;
+	void	*mr;
+};
+
+struct hns_roce_recreate_lp_qp_work {
+	struct	work_struct work;
+	struct	ib_device *ib_dev;
+	struct	completion *comp;
+	int	comp_flag;
+};
+
+struct hns_roce_free_mr {
+	struct workqueue_struct *free_mr_wq;
+	struct hns_roce_qp *mr_free_qp[HNS_ROCE_V1_RESV_QP];
+	struct hns_roce_cq *mr_free_cq;
+	struct hns_roce_pd *mr_free_pd;
+};
+
 struct hns_roce_v1_priv {
 	struct hns_roce_db_table  db_table;
 	struct hns_roce_raq_table raq_table;
 	struct hns_roce_bt_table  bt_table;
 	struct hns_roce_tptr_table tptr_table;
 	struct hns_roce_des_qp des_qp;
+	struct hns_roce_free_mr free_mr;
 };
 
 int hns_dsaf_roce_reset(struct fwnode_handle *dsaf_fwnode, bool dereset);
+int hns_roce_v1_poll_cq(struct ib_cq *ibcq, int num_entries, struct ib_wc *wc);
+int hns_roce_v1_destroy_qp(struct ib_qp *ibqp);
 
 #endif
diff --git a/drivers/infiniband/hw/hns/hns_roce_main.c b/drivers/infiniband/hw/hns/hns_roce_main.c
index 914d0ac7881c..0cedec0c5576 100644
--- a/drivers/infiniband/hw/hns/hns_roce_main.c
+++ b/drivers/infiniband/hw/hns/hns_roce_main.c
@@ -129,7 +129,6 @@ static int handle_en_event(struct hns_roce_dev *hr_dev, u8 port,
 {
 	struct device *dev = &hr_dev->pdev->dev;
 	struct net_device *netdev;
-	unsigned long flags;
 
 	netdev = hr_dev->iboe.netdevs[port];
 	if (!netdev) {
@@ -137,7 +136,7 @@ static int handle_en_event(struct hns_roce_dev *hr_dev, u8 port,
 		return -ENODEV;
 	}
 
-	spin_lock_irqsave(&hr_dev->iboe.lock, flags);
+	spin_lock_bh(&hr_dev->iboe.lock);
 
 	switch (event) {
 	case NETDEV_UP:
@@ -156,7 +155,7 @@ static int handle_en_event(struct hns_roce_dev *hr_dev, u8 port,
 		break;
 	}
 
-	spin_unlock_irqrestore(&hr_dev->iboe.lock, flags);
+	spin_unlock_bh(&hr_dev->iboe.lock);
 	return 0;
 }
 
diff --git a/drivers/infiniband/hw/hns/hns_roce_mr.c b/drivers/infiniband/hw/hns/hns_roce_mr.c
index 9b8a1ad4ee6c..4139abee3b54 100644
--- a/drivers/infiniband/hw/hns/hns_roce_mr.c
+++ b/drivers/infiniband/hw/hns/hns_roce_mr.c
@@ -42,7 +42,7 @@ static u32 hw_index_to_key(unsigned long ind)
 	return (u32)(ind >> 24) | (ind << 8);
 }
 
-static unsigned long key_to_hw_index(u32 key)
+unsigned long key_to_hw_index(u32 key)
 {
 	return (key << 24) | (key >> 8);
 }
@@ -56,7 +56,7 @@ static int hns_roce_sw2hw_mpt(struct hns_roce_dev *hr_dev,
 				 HNS_ROCE_CMD_TIMEOUT_MSECS);
 }
 
-static int hns_roce_hw2sw_mpt(struct hns_roce_dev *hr_dev,
+int hns_roce_hw2sw_mpt(struct hns_roce_dev *hr_dev,
 			      struct hns_roce_cmd_mailbox *mailbox,
 			      unsigned long mpt_index)
 {
@@ -607,13 +607,20 @@ err_free:
 
 int hns_roce_dereg_mr(struct ib_mr *ibmr)
 {
+	struct hns_roce_dev *hr_dev = to_hr_dev(ibmr->device);
 	struct hns_roce_mr *mr = to_hr_mr(ibmr);
+	int ret = 0;
 
-	hns_roce_mr_free(to_hr_dev(ibmr->device), mr);
-	if (mr->umem)
-		ib_umem_release(mr->umem);
+	if (hr_dev->hw->dereg_mr) {
+		ret = hr_dev->hw->dereg_mr(hr_dev, mr);
+	} else {
+		hns_roce_mr_free(hr_dev, mr);
 
-	kfree(mr);
+		if (mr->umem)
+			ib_umem_release(mr->umem);
 
-	return 0;
+		kfree(mr);
+	}
+
+	return ret;
 }
-- 
cgit v1.2.1


From 9daed0affa1378cbfbc549de6a2b25c778c15bda Mon Sep 17 00:00:00 2001
From: "Wei Hu (Xavier)" <xavier.huwei@huawei.com>
Date: Tue, 29 Nov 2016 23:10:27 +0000
Subject: IB/hns: Fix the bug of setting port mtu

In hns_roce driver, we need not call iboe_get_mtu to reduce
IB headers from effective IBoE MTU because hr_dev->caps.max_mtu
has already been reduced.

Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Signed-off-by: Salil Mehta <salil.mehta@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_main.c | 16 ++--------------
 1 file changed, 2 insertions(+), 14 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hns/hns_roce_main.c b/drivers/infiniband/hw/hns/hns_roce_main.c
index 0cedec0c5576..5e620f9f13d2 100644
--- a/drivers/infiniband/hw/hns/hns_roce_main.c
+++ b/drivers/infiniband/hw/hns/hns_roce_main.c
@@ -72,18 +72,6 @@ static void hns_roce_set_mac(struct hns_roce_dev *hr_dev, u8 port, u8 *addr)
 	hr_dev->hw->set_mac(hr_dev, phy_port, addr);
 }
 
-static void hns_roce_set_mtu(struct hns_roce_dev *hr_dev, u8 port, int mtu)
-{
-	u8 phy_port = hr_dev->iboe.phy_port[port];
-	enum ib_mtu tmp;
-
-	tmp = iboe_get_mtu(mtu);
-	if (!tmp)
-		tmp = IB_MTU_256;
-
-	hr_dev->hw->set_mtu(hr_dev, phy_port, tmp);
-}
-
 static int hns_roce_add_gid(struct ib_device *device, u8 port_num,
 			    unsigned int index, const union ib_gid *gid,
 			    const struct ib_gid_attr *attr, void **context)
@@ -188,8 +176,8 @@ static int hns_roce_setup_mtu_mac(struct hns_roce_dev *hr_dev)
 	u8 i;
 
 	for (i = 0; i < hr_dev->caps.num_ports; i++) {
-		hns_roce_set_mtu(hr_dev, i,
-				 ib_mtu_enum_to_int(hr_dev->caps.max_mtu));
+		hr_dev->hw->set_mtu(hr_dev, hr_dev->iboe.phy_port[i],
+				    hr_dev->caps.max_mtu);
 		hns_roce_set_mac(hr_dev, i, hr_dev->iboe.netdevs[i]->dev_addr);
 	}
 
-- 
cgit v1.2.1


From 19a408efa099e6ebe5458d4fc1ba4630d9daf9d4 Mon Sep 17 00:00:00 2001
From: "Wei Hu (Xavier)" <xavier.huwei@huawei.com>
Date: Tue, 29 Nov 2016 23:10:28 +0000
Subject: IB/hns: Delete the redundant memset operation

It deleted the redundant memset operation because the memory allocated
by ib_alloc_device has been set zero.

Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Signed-off-by: Salil Mehta <salil.mehta@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_main.c | 3 ---
 1 file changed, 3 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hns/hns_roce_main.c b/drivers/infiniband/hw/hns/hns_roce_main.c
index 5e620f9f13d2..28a8f24366df 100644
--- a/drivers/infiniband/hw/hns/hns_roce_main.c
+++ b/drivers/infiniband/hw/hns/hns_roce_main.c
@@ -843,9 +843,6 @@ static int hns_roce_probe(struct platform_device *pdev)
 	if (!hr_dev)
 		return -ENOMEM;
 
-	memset((u8 *)hr_dev + sizeof(struct ib_device), 0,
-		sizeof(struct hns_roce_dev) - sizeof(struct ib_device));
-
 	hr_dev->pdev = pdev;
 	platform_set_drvdata(pdev, hr_dev);
 
-- 
cgit v1.2.1


From afb6b092d6507b17c75df394b1d1b2a6e1674627 Mon Sep 17 00:00:00 2001
From: Shaobo Xu <xushaobo2@huawei.com>
Date: Tue, 29 Nov 2016 23:10:29 +0000
Subject: IB/hns: Fix the bug when free cq

If the resources of cq are freed while executing the user case, hardware
can not been notified in hip06 SoC. Then hardware will hold on when it
writes the cq buffer which has been released.

In order to slove this problem, RoCE driver checks the CQE counter, and
ensure that the outstanding CQE have been written. Then the cq buffer
can be released.

Signed-off-by: Shaobo Xu <xushaobo2@huawei.com>
Reviewed-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Signed-off-by: Salil Mehta <salil.mehta@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_common.h |  2 ++
 drivers/infiniband/hw/hns/hns_roce_cq.c     | 27 +++++++++------
 drivers/infiniband/hw/hns/hns_roce_device.h |  8 +++++
 drivers/infiniband/hw/hns/hns_roce_hw_v1.c  | 53 +++++++++++++++++++++++++++++
 4 files changed, 79 insertions(+), 11 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hns/hns_roce_common.h b/drivers/infiniband/hw/hns/hns_roce_common.h
index a0556323a678..4af403e1348c 100644
--- a/drivers/infiniband/hw/hns/hns_roce_common.h
+++ b/drivers/infiniband/hw/hns/hns_roce_common.h
@@ -354,6 +354,8 @@
 
 #define ROCEE_SDB_ISSUE_PTR_REG			0x758
 #define ROCEE_SDB_SEND_PTR_REG			0x75C
+#define ROCEE_CAEP_CQE_WCMD_EMPTY		0x850
+#define ROCEE_SCAEP_WR_CQE_CNT			0x8D0
 #define ROCEE_SDB_INV_CNT_REG			0x9A4
 #define ROCEE_SDB_RETRY_CNT_REG			0x9AC
 #define ROCEE_TSP_BP_ST_REG			0x9EC
diff --git a/drivers/infiniband/hw/hns/hns_roce_cq.c b/drivers/infiniband/hw/hns/hns_roce_cq.c
index c9f6c3dce83c..ff9a6a3c826a 100644
--- a/drivers/infiniband/hw/hns/hns_roce_cq.c
+++ b/drivers/infiniband/hw/hns/hns_roce_cq.c
@@ -179,8 +179,7 @@ static int hns_roce_hw2sw_cq(struct hns_roce_dev *dev,
 				 HNS_ROCE_CMD_TIMEOUT_MSECS);
 }
 
-static void hns_roce_free_cq(struct hns_roce_dev *hr_dev,
-			     struct hns_roce_cq *hr_cq)
+void hns_roce_free_cq(struct hns_roce_dev *hr_dev, struct hns_roce_cq *hr_cq)
 {
 	struct hns_roce_cq_table *cq_table = &hr_dev->cq_table;
 	struct device *dev = &hr_dev->pdev->dev;
@@ -392,19 +391,25 @@ int hns_roce_ib_destroy_cq(struct ib_cq *ib_cq)
 {
 	struct hns_roce_dev *hr_dev = to_hr_dev(ib_cq->device);
 	struct hns_roce_cq *hr_cq = to_hr_cq(ib_cq);
+	int ret = 0;
 
-	hns_roce_free_cq(hr_dev, hr_cq);
-	hns_roce_mtt_cleanup(hr_dev, &hr_cq->hr_buf.hr_mtt);
+	if (hr_dev->hw->destroy_cq) {
+		ret = hr_dev->hw->destroy_cq(ib_cq);
+	} else {
+		hns_roce_free_cq(hr_dev, hr_cq);
+		hns_roce_mtt_cleanup(hr_dev, &hr_cq->hr_buf.hr_mtt);
 
-	if (ib_cq->uobject)
-		ib_umem_release(hr_cq->umem);
-	else
-		/* Free the buff of stored cq */
-		hns_roce_ib_free_cq_buf(hr_dev, &hr_cq->hr_buf, ib_cq->cqe);
+		if (ib_cq->uobject)
+			ib_umem_release(hr_cq->umem);
+		else
+			/* Free the buff of stored cq */
+			hns_roce_ib_free_cq_buf(hr_dev, &hr_cq->hr_buf,
+						ib_cq->cqe);
 
-	kfree(hr_cq);
+		kfree(hr_cq);
+	}
 
-	return 0;
+	return ret;
 }
 
 void hns_roce_cq_completion(struct hns_roce_dev *hr_dev, u32 cqn)
diff --git a/drivers/infiniband/hw/hns/hns_roce_device.h b/drivers/infiniband/hw/hns/hns_roce_device.h
index 1050829da642..d4f0fce98587 100644
--- a/drivers/infiniband/hw/hns/hns_roce_device.h
+++ b/drivers/infiniband/hw/hns/hns_roce_device.h
@@ -56,6 +56,12 @@
 #define HNS_ROCE_MAX_INNER_MTPT_NUM		0x7
 #define HNS_ROCE_MAX_MTPT_PBL_NUM		0x100000
 
+#define HNS_ROCE_EACH_FREE_CQ_WAIT_MSECS	20
+#define HNS_ROCE_MAX_FREE_CQ_WAIT_CNT	\
+	(5000 / HNS_ROCE_EACH_FREE_CQ_WAIT_MSECS)
+#define HNS_ROCE_CQE_WCMD_EMPTY_BIT		0x2
+#define HNS_ROCE_MIN_CQE_CNT			16
+
 #define HNS_ROCE_MAX_IRQ_NUM			34
 
 #define HNS_ROCE_COMP_VEC_NUM			32
@@ -528,6 +534,7 @@ struct hns_roce_hw {
 	int (*req_notify_cq)(struct ib_cq *ibcq, enum ib_cq_notify_flags flags);
 	int (*poll_cq)(struct ib_cq *ibcq, int num_entries, struct ib_wc *wc);
 	int (*dereg_mr)(struct hns_roce_dev *hr_dev, struct hns_roce_mr *mr);
+	int (*destroy_cq)(struct ib_cq *ibcq);
 	void	*priv;
 };
 
@@ -734,6 +741,7 @@ struct ib_cq *hns_roce_ib_create_cq(struct ib_device *ib_dev,
 				    struct ib_udata *udata);
 
 int hns_roce_ib_destroy_cq(struct ib_cq *ib_cq);
+void hns_roce_free_cq(struct hns_roce_dev *hr_dev, struct hns_roce_cq *hr_cq);
 
 void hns_roce_cq_completion(struct hns_roce_dev *hr_dev, u32 cqn);
 void hns_roce_cq_event(struct hns_roce_dev *hr_dev, u32 cqn, int event_type);
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
index f67a3bfd4c55..b8111b0c8877 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
@@ -3763,6 +3763,58 @@ int hns_roce_v1_destroy_qp(struct ib_qp *ibqp)
 	return 0;
 }
 
+int hns_roce_v1_destroy_cq(struct ib_cq *ibcq)
+{
+	struct hns_roce_dev *hr_dev = to_hr_dev(ibcq->device);
+	struct hns_roce_cq *hr_cq = to_hr_cq(ibcq);
+	struct device *dev = &hr_dev->pdev->dev;
+	u32 cqe_cnt_ori;
+	u32 cqe_cnt_cur;
+	u32 cq_buf_size;
+	int wait_time = 0;
+	int ret = 0;
+
+	hns_roce_free_cq(hr_dev, hr_cq);
+
+	/*
+	 * Before freeing cq buffer, we need to ensure that the outstanding CQE
+	 * have been written by checking the CQE counter.
+	 */
+	cqe_cnt_ori = roce_read(hr_dev, ROCEE_SCAEP_WR_CQE_CNT);
+	while (1) {
+		if (roce_read(hr_dev, ROCEE_CAEP_CQE_WCMD_EMPTY) &
+		    HNS_ROCE_CQE_WCMD_EMPTY_BIT)
+			break;
+
+		cqe_cnt_cur = roce_read(hr_dev, ROCEE_SCAEP_WR_CQE_CNT);
+		if ((cqe_cnt_cur - cqe_cnt_ori) >= HNS_ROCE_MIN_CQE_CNT)
+			break;
+
+		msleep(HNS_ROCE_EACH_FREE_CQ_WAIT_MSECS);
+		if (wait_time > HNS_ROCE_MAX_FREE_CQ_WAIT_CNT) {
+			dev_warn(dev, "Destroy cq 0x%lx timeout!\n",
+				hr_cq->cqn);
+			ret = -ETIMEDOUT;
+			break;
+		}
+		wait_time++;
+	}
+
+	hns_roce_mtt_cleanup(hr_dev, &hr_cq->hr_buf.hr_mtt);
+
+	if (ibcq->uobject)
+		ib_umem_release(hr_cq->umem);
+	else {
+		/* Free the buff of stored cq */
+		cq_buf_size = (ibcq->cqe + 1) * hr_dev->caps.cq_entry_sz;
+		hns_roce_buf_free(hr_dev, cq_buf_size, &hr_cq->hr_buf.hr_buf);
+	}
+
+	kfree(hr_cq);
+
+	return ret;
+}
+
 struct hns_roce_v1_priv hr_v1_priv;
 
 struct hns_roce_hw hns_roce_hw_v1 = {
@@ -3784,5 +3836,6 @@ struct hns_roce_hw hns_roce_hw_v1 = {
 	.req_notify_cq = hns_roce_v1_req_notify_cq,
 	.poll_cq = hns_roce_v1_poll_cq,
 	.dereg_mr = hns_roce_v1_dereg_mr,
+	.destroy_cq = hns_roce_v1_destroy_cq,
 	.priv = &hr_v1_priv,
 };
-- 
cgit v1.2.1


From 3b5184be89449275e5396512ab0127fc19ce359e Mon Sep 17 00:00:00 2001
From: Lijun Ou <oulijun@huawei.com>
Date: Tue, 29 Nov 2016 23:10:30 +0000
Subject: IB/hns: Fix the IB device name

This patch mainly fix the name for IB device in order
to match with libhns.

Signed-off-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Salil Mehta <salil.mehta@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_main.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hns/hns_roce_main.c b/drivers/infiniband/hw/hns/hns_roce_main.c
index 28a8f24366df..eddb053ff0ba 100644
--- a/drivers/infiniband/hw/hns/hns_roce_main.c
+++ b/drivers/infiniband/hw/hns/hns_roce_main.c
@@ -433,7 +433,7 @@ static int hns_roce_register_device(struct hns_roce_dev *hr_dev)
 	spin_lock_init(&iboe->lock);
 
 	ib_dev = &hr_dev->ib_dev;
-	strlcpy(ib_dev->name, "hisi_%d", IB_DEVICE_NAME_MAX);
+	strlcpy(ib_dev->name, "hns_%d", IB_DEVICE_NAME_MAX);
 
 	ib_dev->owner			= THIS_MODULE;
 	ib_dev->node_type		= RDMA_NODE_IB_CA;
-- 
cgit v1.2.1


From 4d4099584c2c4dca6c04d78ded4cc81f50cc3634 Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leon@kernel.org>
Date: Wed, 19 Oct 2016 20:13:07 +0300
Subject: IB/hns: Move HNS RoCE user vendor structures

This patch moves HNS vendor's specific structures to
common UAPI folder which will be visible to all consumers.

Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hns/hns_roce_cq.c   |  2 +-
 drivers/infiniband/hw/hns/hns_roce_main.c |  2 +-
 drivers/infiniband/hw/hns/hns_roce_qp.c   |  2 +-
 drivers/infiniband/hw/hns/hns_roce_user.h | 53 -------------------------------
 4 files changed, 3 insertions(+), 56 deletions(-)
 delete mode 100644 drivers/infiniband/hw/hns/hns_roce_user.h

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hns/hns_roce_cq.c b/drivers/infiniband/hw/hns/hns_roce_cq.c
index ff9a6a3c826a..589496c8fb9e 100644
--- a/drivers/infiniband/hw/hns/hns_roce_cq.c
+++ b/drivers/infiniband/hw/hns/hns_roce_cq.c
@@ -35,7 +35,7 @@
 #include "hns_roce_device.h"
 #include "hns_roce_cmd.h"
 #include "hns_roce_hem.h"
-#include "hns_roce_user.h"
+#include <rdma/hns-abi.h>
 #include "hns_roce_common.h"
 
 static void hns_roce_ib_cq_comp(struct hns_roce_cq *hr_cq)
diff --git a/drivers/infiniband/hw/hns/hns_roce_main.c b/drivers/infiniband/hw/hns/hns_roce_main.c
index eddb053ff0ba..4953d9cb83a7 100644
--- a/drivers/infiniband/hw/hns/hns_roce_main.c
+++ b/drivers/infiniband/hw/hns/hns_roce_main.c
@@ -38,7 +38,7 @@
 #include <rdma/ib_cache.h>
 #include "hns_roce_common.h"
 #include "hns_roce_device.h"
-#include "hns_roce_user.h"
+#include <rdma/hns-abi.h>
 #include "hns_roce_hem.h"
 
 /**
diff --git a/drivers/infiniband/hw/hns/hns_roce_qp.c b/drivers/infiniband/hw/hns/hns_roce_qp.c
index 4775b5c725a9..f036f32f15d3 100644
--- a/drivers/infiniband/hw/hns/hns_roce_qp.c
+++ b/drivers/infiniband/hw/hns/hns_roce_qp.c
@@ -37,7 +37,7 @@
 #include "hns_roce_common.h"
 #include "hns_roce_device.h"
 #include "hns_roce_hem.h"
-#include "hns_roce_user.h"
+#include <rdma/hns-abi.h>
 
 #define SQP_NUM				(2 * HNS_ROCE_MAX_PORTS)
 
diff --git a/drivers/infiniband/hw/hns/hns_roce_user.h b/drivers/infiniband/hw/hns/hns_roce_user.h
deleted file mode 100644
index a28f761a9f65..000000000000
--- a/drivers/infiniband/hw/hns/hns_roce_user.h
+++ /dev/null
@@ -1,53 +0,0 @@
-/*
- * Copyright (c) 2016 Hisilicon Limited.
- *
- * This software is available to you under a choice of one of two
- * licenses.  You may choose to be licensed under the terms of the GNU
- * General Public License (GPL) Version 2, available from the file
- * COPYING in the main directory of this source tree, or the
- * OpenIB.org BSD license below:
- *
- *     Redistribution and use in source and binary forms, with or
- *     without modification, are permitted provided that the following
- *     conditions are met:
- *
- *      - Redistributions of source code must retain the above
- *        copyright notice, this list of conditions and the following
- *        disclaimer.
- *
- *      - Redistributions in binary form must reproduce the above
- *        copyright notice, this list of conditions and the following
- *        disclaimer in the documentation and/or other materials
- *        provided with the distribution.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
- * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
- * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
- * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
- * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
- * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
- * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
- * SOFTWARE.
- */
-
-#ifndef _HNS_ROCE_USER_H
-#define _HNS_ROCE_USER_H
-
-struct hns_roce_ib_create_cq {
-	__u64   buf_addr;
-};
-
-struct hns_roce_ib_create_qp {
-	__u64	buf_addr;
-	__u64   db_addr;
-	__u8    log_sq_bb_count;
-	__u8    log_sq_stride;
-	__u8    sq_no_prefetch;
-	__u8    reserved[5];
-};
-
-struct hns_roce_ib_alloc_ucontext_resp {
-	__u32	qp_tab_size;
-};
-
-#endif /*_HNS_ROCE_USER_H */
-- 
cgit v1.2.1


From 0fc2dc58896f182daeeb4a7b5fc8d763afec3117 Mon Sep 17 00:00:00 2001
From: Henry Orosco <henry.orosco@intel.com>
Date: Mon, 10 Oct 2016 21:12:10 -0500
Subject: i40iw: Add Quality of Service support

Add support for QoS on QPs. Upon device initialization,
a map is created from user priority to queue set
handles. On QP creation, use ToS to look up the queue
set handle for use with the QP.

Signed-off-by: Faisal Latif <faisal.latif@intel.com>
Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw.h       |   9 ++
 drivers/infiniband/hw/i40iw/i40iw_cm.c    |  30 +++++-
 drivers/infiniband/hw/i40iw/i40iw_cm.h    |   2 +-
 drivers/infiniband/hw/i40iw/i40iw_ctrl.c  | 151 +++++++++++++++++++++++++++++-
 drivers/infiniband/hw/i40iw/i40iw_d.h     |   2 +
 drivers/infiniband/hw/i40iw/i40iw_hw.c    |  25 ++---
 drivers/infiniband/hw/i40iw/i40iw_main.c  |  66 +++++++++++--
 drivers/infiniband/hw/i40iw/i40iw_osdep.h |   2 +
 drivers/infiniband/hw/i40iw/i40iw_p.h     |   2 +
 drivers/infiniband/hw/i40iw/i40iw_puda.c  |   3 +-
 drivers/infiniband/hw/i40iw/i40iw_type.h  |  18 +++-
 drivers/infiniband/hw/i40iw/i40iw_utils.c |  45 +++++++++
 drivers/infiniband/hw/i40iw/i40iw_verbs.c |   6 +-
 13 files changed, 325 insertions(+), 36 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw.h b/drivers/infiniband/hw/i40iw/i40iw.h
index 8ec09e470f84..4a0c12bd9caa 100644
--- a/drivers/infiniband/hw/i40iw/i40iw.h
+++ b/drivers/infiniband/hw/i40iw/i40iw.h
@@ -210,6 +210,12 @@ struct i40iw_msix_vector {
 	u32 ceq_id;
 };
 
+struct l2params_work {
+	struct work_struct work;
+	struct i40iw_device *iwdev;
+	struct i40iw_l2params l2params;
+};
+
 #define I40IW_MSIX_TABLE_SIZE   65
 
 struct virtchnl_work {
@@ -514,6 +520,9 @@ void i40iw_add_pdusecount(struct i40iw_pd *iwpd);
 void i40iw_hw_modify_qp(struct i40iw_device *iwdev, struct i40iw_qp *iwqp,
 			struct i40iw_modify_qp_info *info, bool wait);
 
+void i40iw_qp_suspend_resume(struct i40iw_sc_dev *dev,
+			     struct i40iw_sc_qp *qp,
+			     bool suspend);
 enum i40iw_status_code i40iw_manage_qhash(struct i40iw_device *iwdev,
 					  struct i40iw_cm_info *cminfo,
 					  enum i40iw_quad_entry_type etype,
diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.c b/drivers/infiniband/hw/i40iw/i40iw_cm.c
index 85637696f6e9..24b22e99003c 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_cm.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_cm.c
@@ -221,6 +221,7 @@ static void i40iw_get_addr_info(struct i40iw_cm_node *cm_node,
 	memcpy(cm_info->rem_addr, cm_node->rem_addr, sizeof(cm_info->rem_addr));
 	cm_info->loc_port = cm_node->loc_port;
 	cm_info->rem_port = cm_node->rem_port;
+	cm_info->user_pri = cm_node->user_pri;
 }
 
 /**
@@ -396,6 +397,7 @@ static struct i40iw_puda_buf *i40iw_form_cm_frame(struct i40iw_cm_node *cm_node,
 	u32 opts_len = 0;
 	u32 pd_len = 0;
 	u32 hdr_len = 0;
+	u16 vtag;
 
 	sqbuf = i40iw_puda_get_bufpool(dev->ilq);
 	if (!sqbuf)
@@ -445,7 +447,8 @@ static struct i40iw_puda_buf *i40iw_form_cm_frame(struct i40iw_cm_node *cm_node,
 		ether_addr_copy(ethh->h_source, cm_node->loc_mac);
 		if (cm_node->vlan_id < VLAN_TAG_PRESENT) {
 			((struct vlan_ethhdr *)ethh)->h_vlan_proto = htons(ETH_P_8021Q);
-			((struct vlan_ethhdr *)ethh)->h_vlan_TCI = htons(cm_node->vlan_id);
+			vtag = (cm_node->user_pri << VLAN_PRIO_SHIFT) | cm_node->vlan_id;
+			((struct vlan_ethhdr *)ethh)->h_vlan_TCI = htons(vtag);
 
 			((struct vlan_ethhdr *)ethh)->h_vlan_encapsulated_proto = htons(ETH_P_IP);
 		} else {
@@ -474,7 +477,8 @@ static struct i40iw_puda_buf *i40iw_form_cm_frame(struct i40iw_cm_node *cm_node,
 		ether_addr_copy(ethh->h_source, cm_node->loc_mac);
 		if (cm_node->vlan_id < VLAN_TAG_PRESENT) {
 			((struct vlan_ethhdr *)ethh)->h_vlan_proto = htons(ETH_P_8021Q);
-			((struct vlan_ethhdr *)ethh)->h_vlan_TCI = htons(cm_node->vlan_id);
+			vtag = (cm_node->user_pri << VLAN_PRIO_SHIFT) | cm_node->vlan_id;
+			((struct vlan_ethhdr *)ethh)->h_vlan_TCI = htons(vtag);
 			((struct vlan_ethhdr *)ethh)->h_vlan_encapsulated_proto = htons(ETH_P_IPV6);
 		} else {
 			ethh->h_proto = htons(ETH_P_IPV6);
@@ -1880,6 +1884,7 @@ static int i40iw_dec_refcnt_listen(struct i40iw_cm_core *cm_core,
 			nfo.loc_port = listener->loc_port;
 			nfo.ipv4 = listener->ipv4;
 			nfo.vlan_id = listener->vlan_id;
+			nfo.user_pri = listener->user_pri;
 
 			if (!list_empty(&listener->child_listen_list)) {
 				i40iw_del_multiple_qhash(listener->iwdev, &nfo, listener);
@@ -2138,6 +2143,11 @@ static struct i40iw_cm_node *i40iw_make_cm_node(
 	/* set our node specific transport info */
 	cm_node->ipv4 = cm_info->ipv4;
 	cm_node->vlan_id = cm_info->vlan_id;
+	if ((cm_node->vlan_id == I40IW_NO_VLAN) && iwdev->dcb)
+		cm_node->vlan_id = 0;
+	cm_node->user_pri = cm_info->user_pri;
+	if (listener)
+		cm_node->user_pri = listener->user_pri;
 	memcpy(cm_node->loc_addr, cm_info->loc_addr, sizeof(cm_node->loc_addr));
 	memcpy(cm_node->rem_addr, cm_info->rem_addr, sizeof(cm_node->rem_addr));
 	cm_node->loc_port = cm_info->loc_port;
@@ -3055,6 +3065,7 @@ void i40iw_receive_ilq(struct i40iw_sc_dev *dev, struct i40iw_puda_buf *rbuf)
 	struct i40iw_device *iwdev = (struct i40iw_device *)dev->back_dev;
 	struct i40iw_cm_core *cm_core = &iwdev->cm_core;
 	struct vlan_ethhdr *ethh;
+	u16 vtag;
 
 	/* if vlan, then maclen = 18 else 14 */
 	iph = (struct iphdr *)rbuf->iph;
@@ -3068,7 +3079,9 @@ void i40iw_receive_ilq(struct i40iw_sc_dev *dev, struct i40iw_puda_buf *rbuf)
 	ethh = (struct vlan_ethhdr *)rbuf->mem.va;
 
 	if (ethh->h_vlan_proto == htons(ETH_P_8021Q)) {
-		cm_info.vlan_id = ntohs(ethh->h_vlan_TCI) & VLAN_VID_MASK;
+		vtag = ntohs(ethh->h_vlan_TCI);
+		cm_info.user_pri = (vtag & VLAN_PRIO_MASK) >> VLAN_PRIO_SHIFT;
+		cm_info.vlan_id = vtag & VLAN_VID_MASK;
 		i40iw_debug(cm_core->dev,
 			    I40IW_DEBUG_CM,
 			    "%s vlan_id=%d\n",
@@ -3309,6 +3322,8 @@ static void i40iw_cm_init_tsa_conn(struct i40iw_qp *iwqp,
 
 	ctx_info->tcp_info_valid = true;
 	ctx_info->iwarp_info_valid = true;
+	ctx_info->add_to_qoslist = true;
+	ctx_info->user_pri = cm_node->user_pri;
 
 	i40iw_init_tcp_ctx(cm_node, &tcp_info, iwqp);
 	if (cm_node->snd_mark_en) {
@@ -3326,6 +3341,7 @@ static void i40iw_cm_init_tsa_conn(struct i40iw_qp *iwqp,
 	/* once tcp_info is set, no need to do it again */
 	ctx_info->tcp_info_valid = false;
 	ctx_info->iwarp_info_valid = false;
+	ctx_info->add_to_qoslist = false;
 }
 
 /**
@@ -3759,6 +3775,9 @@ int i40iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param)
 		i40iw_netdev_vlan_ipv6(cm_info.loc_addr, &cm_info.vlan_id, NULL);
 	}
 	cm_info.cm_id = cm_id;
+	cm_info.user_pri = rt_tos2priority(cm_id->tos);
+	i40iw_debug(&iwdev->sc_dev, I40IW_DEBUG_DCB, "%s TOS:[%d] UP:[%d]\n",
+		    __func__, cm_id->tos, cm_info.user_pri);
 	if ((cm_info.ipv4 && (laddr->sin_addr.s_addr != raddr->sin_addr.s_addr)) ||
 	    (!cm_info.ipv4 && memcmp(laddr6->sin6_addr.in6_u.u6_addr32,
 				     raddr6->sin6_addr.in6_u.u6_addr32,
@@ -3904,6 +3923,11 @@ int i40iw_create_listen(struct iw_cm_id *cm_id, int backlog)
 
 	cm_id->provider_data = cm_listen_node;
 
+	cm_listen_node->user_pri = rt_tos2priority(cm_id->tos);
+	cm_info.user_pri = cm_listen_node->user_pri;
+	i40iw_debug(&iwdev->sc_dev, I40IW_DEBUG_DCB, "%s TOS:[%d] UP:[%d]\n",
+		    __func__, cm_id->tos, cm_listen_node->user_pri);
+
 	if (!cm_listen_node->reused_node) {
 		if (wildcard) {
 			if (cm_info.ipv4)
diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.h b/drivers/infiniband/hw/i40iw/i40iw_cm.h
index e9046d9f9645..945ed26f8dba 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_cm.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_cm.h
@@ -368,7 +368,7 @@ struct i40iw_cm_info {
 	u32 rem_addr[4];
 	u16 vlan_id;
 	int backlog;
-	u16 user_pri;
+	u8 user_pri;
 	bool ipv4;
 };
 
diff --git a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
index 2c4b4d072d6a..31c4a0c26786 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
@@ -222,6 +222,133 @@ static enum i40iw_status_code i40iw_sc_parse_fpm_query_buf(
 	return 0;
 }
 
+/**
+ * i40iw_fill_qos_list - Change all unknown qs handles to available ones
+ * @qs_list: list of qs_handles to be fixed with valid qs_handles
+ */
+static void i40iw_fill_qos_list(u16 *qs_list)
+{
+	u16 qshandle = qs_list[0];
+	int i;
+
+	for (i = 0; i < I40IW_MAX_USER_PRIORITY; i++) {
+		if (qs_list[i] == QS_HANDLE_UNKNOWN)
+			qs_list[i] = qshandle;
+		else
+			qshandle = qs_list[i];
+	}
+}
+
+/**
+ * i40iw_qp_from_entry - Given entry, get to the qp structure
+ * @entry: Points to list of qp structure
+ */
+static struct i40iw_sc_qp *i40iw_qp_from_entry(struct list_head *entry)
+{
+	if (!entry)
+		return NULL;
+
+	return (struct i40iw_sc_qp *)((char *)entry - offsetof(struct i40iw_sc_qp, list));
+}
+
+/**
+ * i40iw_get_qp - get the next qp from the list given current qp
+ * @head: Listhead of qp's
+ * @qp: current qp
+ */
+static struct i40iw_sc_qp *i40iw_get_qp(struct list_head *head, struct i40iw_sc_qp *qp)
+{
+	struct list_head *entry = NULL;
+	struct list_head *lastentry;
+
+	if (list_empty(head))
+		return NULL;
+
+	if (!qp) {
+		entry = head->next;
+	} else {
+		lastentry = &qp->list;
+		entry = (lastentry != head) ? lastentry->next : NULL;
+	}
+
+	return i40iw_qp_from_entry(entry);
+}
+
+/**
+ * i40iw_change_l2params - given the new l2 parameters, change all qp
+ * @dev: IWARP device pointer
+ * @l2params: New paramaters from l2
+ */
+void i40iw_change_l2params(struct i40iw_sc_dev *dev, struct i40iw_l2params *l2params)
+{
+	struct i40iw_sc_qp *qp = NULL;
+	bool qs_handle_change = false;
+	bool mss_change = false;
+	unsigned long flags;
+	u16 qs_handle;
+	int i;
+
+	if (dev->mss != l2params->mss) {
+		mss_change = true;
+		dev->mss = l2params->mss;
+	}
+
+	i40iw_fill_qos_list(l2params->qs_handle_list);
+	for (i = 0; i < I40IW_MAX_USER_PRIORITY; i++) {
+		qs_handle = l2params->qs_handle_list[i];
+		if (dev->qos[i].qs_handle != qs_handle)
+			qs_handle_change = true;
+		else if (!mss_change)
+			continue;       /* no MSS nor qs handle change */
+		spin_lock_irqsave(&dev->qos[i].lock, flags);
+		qp = i40iw_get_qp(&dev->qos[i].qplist, qp);
+		while (qp) {
+			if (mss_change)
+				i40iw_qp_mss_modify(dev, qp);
+			if (qs_handle_change) {
+				qp->qs_handle = qs_handle;
+				/* issue cqp suspend command */
+				i40iw_qp_suspend_resume(dev, qp, true);
+			}
+			qp = i40iw_get_qp(&dev->qos[i].qplist, qp);
+		}
+		spin_unlock_irqrestore(&dev->qos[i].lock, flags);
+		dev->qos[i].qs_handle = qs_handle;
+	}
+}
+
+/**
+ * i40iw_qp_rem_qos - remove qp from qos lists during destroy qp
+ * @dev: IWARP device pointer
+ * @qp: qp to be removed from qos
+ */
+static void i40iw_qp_rem_qos(struct i40iw_sc_dev *dev, struct i40iw_sc_qp *qp)
+{
+	unsigned long flags;
+
+	if (!qp->on_qoslist)
+		return;
+	spin_lock_irqsave(&dev->qos[qp->user_pri].lock, flags);
+	list_del(&qp->list);
+	spin_unlock_irqrestore(&dev->qos[qp->user_pri].lock, flags);
+}
+
+/**
+ * i40iw_qp_add_qos - called during setctx fot qp to be added to qos
+ * @dev: IWARP device pointer
+ * @qp: qp to be added to qos
+ */
+void i40iw_qp_add_qos(struct i40iw_sc_dev *dev, struct i40iw_sc_qp *qp)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&dev->qos[qp->user_pri].lock, flags);
+	qp->qs_handle = dev->qos[qp->user_pri].qs_handle;
+	list_add(&qp->list, &dev->qos[qp->user_pri].qplist);
+	qp->on_qoslist = true;
+	spin_unlock_irqrestore(&dev->qos[qp->user_pri].lock, flags);
+}
+
 /**
  * i40iw_sc_pd_init - initialize sc pd struct
  * @dev: sc device struct
@@ -1082,7 +1209,7 @@ static enum i40iw_status_code i40iw_sc_manage_qhash_table_entry(
 			      LS_64(info->dest_ip[2], I40IW_CQPSQ_QHASH_ADDR2) |
 			      LS_64(info->dest_ip[3], I40IW_CQPSQ_QHASH_ADDR3));
 	}
-	qw2 = LS_64(cqp->dev->qs_handle, I40IW_CQPSQ_QHASH_QS_HANDLE);
+	qw2 = LS_64(cqp->dev->qos[info->user_pri].qs_handle, I40IW_CQPSQ_QHASH_QS_HANDLE);
 	if (info->vlan_valid)
 		qw2 |= LS_64(info->vlan_id, I40IW_CQPSQ_QHASH_VLANID);
 	set_64bit_val(wqe, 16, qw2);
@@ -2151,7 +2278,7 @@ static enum i40iw_status_code i40iw_sc_qp_init(struct i40iw_sc_qp *qp,
 	qp->rq_tph_en = info->rq_tph_en;
 	qp->rcv_tph_en = info->rcv_tph_en;
 	qp->xmit_tph_en = info->xmit_tph_en;
-	qp->qs_handle = qp->pd->dev->qs_handle;
+	qp->qs_handle = qp->pd->dev->qos[qp->user_pri].qs_handle;
 	qp->exception_lan_queue = qp->pd->dev->exception_lan_queue;
 
 	return 0;
@@ -2296,6 +2423,7 @@ static enum i40iw_status_code i40iw_sc_qp_destroy(
 	struct i40iw_sc_cqp *cqp;
 	u64 header;
 
+	i40iw_qp_rem_qos(qp->pd->dev, qp);
 	cqp = qp->pd->dev->cqp;
 	wqe = i40iw_sc_cqp_get_next_send_wqe(cqp, scratch);
 	if (!wqe)
@@ -2447,6 +2575,12 @@ static enum i40iw_status_code i40iw_sc_qp_setctx(
 
 	iw = info->iwarp_info;
 	tcp = info->tcp_info;
+	if (info->add_to_qoslist) {
+		qp->user_pri = info->user_pri;
+		i40iw_qp_add_qos(qp->pd->dev, qp);
+		i40iw_debug(qp->dev, I40IW_DEBUG_DCB, "%s qp[%d] UP[%d] qset[%d]\n",
+			    __func__, qp->qp_uk.qp_id, qp->user_pri, qp->qs_handle);
+	}
 	qw0 = LS_64(qp->qp_uk.rq_wqe_size, I40IWQPC_RQWQESIZE) |
 	      LS_64(info->err_rq_idx_valid, I40IWQPC_ERR_RQ_IDX_VALID) |
 	      LS_64(qp->rcv_tph_en, I40IWQPC_RCVTPHEN) |
@@ -3959,7 +4093,7 @@ enum i40iw_status_code i40iw_process_cqp_cmd(struct i40iw_sc_dev *dev,
 					     struct cqp_commands_info *pcmdinfo)
 {
 	enum i40iw_status_code status = 0;
-	unsigned long	flags;
+	unsigned long flags;
 
 	spin_lock_irqsave(&dev->cqp_lock, flags);
 	if (list_empty(&dev->cqp_cmd_head) && !i40iw_ring_full(dev->cqp))
@@ -3978,7 +4112,7 @@ enum i40iw_status_code i40iw_process_bh(struct i40iw_sc_dev *dev)
 {
 	enum i40iw_status_code status = 0;
 	struct cqp_commands_info *pcmdinfo;
-	unsigned long	flags;
+	unsigned long flags;
 
 	spin_lock_irqsave(&dev->cqp_lock, flags);
 	while (!list_empty(&dev->cqp_cmd_head) && !i40iw_ring_full(dev->cqp)) {
@@ -4742,6 +4876,7 @@ enum i40iw_status_code i40iw_device_init(struct i40iw_sc_dev *dev,
 	u16 hmc_fcn = 0;
 	enum i40iw_status_code ret_code = 0;
 	u8 db_size;
+	int i;
 
 	spin_lock_init(&dev->cqp_lock);
 	INIT_LIST_HEAD(&dev->cqp_cmd_head);             /* for the cqp commands backlog. */
@@ -4757,7 +4892,13 @@ enum i40iw_status_code i40iw_device_init(struct i40iw_sc_dev *dev,
 		return ret_code;
 	}
 	dev->hmc_fn_id = info->hmc_fn_id;
-	dev->qs_handle = info->qs_handle;
+	i40iw_fill_qos_list(info->l2params.qs_handle_list);
+	for (i = 0; i < I40IW_MAX_USER_PRIORITY; i++) {
+		dev->qos[i].qs_handle = info->l2params.qs_handle_list[i];
+		i40iw_debug(dev, I40IW_DEBUG_DCB, "qset[%d]: %d\n", i, dev->qos[i].qs_handle);
+		spin_lock_init(&dev->qos[i].lock);
+		INIT_LIST_HEAD(&dev->qos[i].qplist);
+	}
 	dev->exception_lan_queue = info->exception_lan_queue;
 	dev->is_pf = info->is_pf;
 
diff --git a/drivers/infiniband/hw/i40iw/i40iw_d.h b/drivers/infiniband/hw/i40iw/i40iw_d.h
index 2fac1db0e0a0..e184c0e99cb1 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_d.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_d.h
@@ -74,6 +74,8 @@
 #define RS_32_1(val, bits)      (u32)(val >> bits)
 #define I40E_HI_DWORD(x)        ((u32)((((x) >> 16) >> 16) & 0xFFFFFFFF))
 
+#define QS_HANDLE_UNKNOWN       0xffff
+
 #define LS_64(val, field) (((u64)val << field ## _SHIFT) & (field ## _MASK))
 
 #define RS_64(val, field) ((u64)(val & field ## _MASK) >> field ## _SHIFT)
diff --git a/drivers/infiniband/hw/i40iw/i40iw_hw.c b/drivers/infiniband/hw/i40iw/i40iw_hw.c
index 0c92a40b3e86..b94727ffc862 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_hw.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_hw.c
@@ -359,6 +359,9 @@ void i40iw_process_aeq(struct i40iw_device *iwdev)
 				continue;
 			i40iw_cm_disconn(iwqp);
 			break;
+		case I40IW_AE_QP_SUSPEND_COMPLETE:
+			i40iw_qp_suspend_resume(dev, &iwqp->sc_qp, false);
+			break;
 		case I40IW_AE_TERMINATE_SENT:
 			i40iw_terminate_send_fin(qp);
 			break;
@@ -404,19 +407,18 @@ void i40iw_process_aeq(struct i40iw_device *iwdev)
 		case I40IW_AE_LCE_CQ_CATASTROPHIC:
 		case I40IW_AE_UDA_XMIT_DGRAM_TOO_LONG:
 		case I40IW_AE_UDA_XMIT_IPADDR_MISMATCH:
-		case I40IW_AE_QP_SUSPEND_COMPLETE:
 			ctx_info->err_rq_idx_valid = false;
 		default:
-				if (!info->sq && ctx_info->err_rq_idx_valid) {
-					ctx_info->err_rq_idx = info->wqe_idx;
-					ctx_info->tcp_info_valid = false;
-					ctx_info->iwarp_info_valid = false;
-					ret = dev->iw_priv_qp_ops->qp_setctx(&iwqp->sc_qp,
-									     iwqp->host_ctx.va,
-									     ctx_info);
-				}
-				i40iw_terminate_connection(qp, info);
-				break;
+			if (!info->sq && ctx_info->err_rq_idx_valid) {
+				ctx_info->err_rq_idx = info->wqe_idx;
+				ctx_info->tcp_info_valid = false;
+				ctx_info->iwarp_info_valid = false;
+				ret = dev->iw_priv_qp_ops->qp_setctx(&iwqp->sc_qp,
+								     iwqp->host_ctx.va,
+								     ctx_info);
+			}
+			i40iw_terminate_connection(qp, info);
+			break;
 		}
 		if (info->qp)
 			i40iw_rem_ref(&iwqp->ibqp);
@@ -560,6 +562,7 @@ enum i40iw_status_code i40iw_manage_qhash(struct i40iw_device *iwdev,
 	}
 
 	info->ipv4_valid = cminfo->ipv4;
+	info->user_pri = cminfo->user_pri;
 	ether_addr_copy(info->mac_addr, iwdev->netdev->dev_addr);
 	info->qp_num = cpu_to_le32(dev->ilq->qp_id);
 	info->dest_port = cpu_to_le16(cminfo->loc_port);
diff --git a/drivers/infiniband/hw/i40iw/i40iw_main.c b/drivers/infiniband/hw/i40iw/i40iw_main.c
index ac2f3cd9478c..40aac87c128c 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_main.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_main.c
@@ -939,7 +939,7 @@ static enum i40iw_status_code i40iw_initialize_ilq(struct i40iw_device *iwdev)
 	info.rq_size = 8192;
 	info.buf_size = 1024;
 	info.tx_buf_cnt = 16384;
-	info.mss = iwdev->mss;
+	info.mss = iwdev->sc_dev.mss;
 	info.receive = i40iw_receive_ilq;
 	info.xmit_complete = i40iw_free_sqbuf;
 	status = i40iw_puda_create_rsrc(&iwdev->sc_dev, &info);
@@ -967,7 +967,7 @@ static enum i40iw_status_code i40iw_initialize_ieq(struct i40iw_device *iwdev)
 	info.sq_size = 8192;
 	info.rq_size = 8192;
 	info.buf_size = 2048;
-	info.mss = iwdev->mss;
+	info.mss = iwdev->sc_dev.mss;
 	info.tx_buf_cnt = 16384;
 	status = i40iw_puda_create_rsrc(&iwdev->sc_dev, &info);
 	if (status)
@@ -1296,6 +1296,9 @@ static enum i40iw_status_code i40iw_initialize_dev(struct i40iw_device *iwdev,
 	struct i40iw_device_init_info info;
 	struct i40iw_dma_mem mem;
 	u32 size;
+	u16 last_qset = I40IW_NO_QSET;
+	u16 qset;
+	u32 i;
 
 	memset(&info, 0, sizeof(info));
 	size = sizeof(struct i40iw_hmc_pble_rsrc) + sizeof(struct i40iw_hmc_info) +
@@ -1325,7 +1328,16 @@ static enum i40iw_status_code i40iw_initialize_dev(struct i40iw_device *iwdev,
 	info.bar0 = ldev->hw_addr;
 	info.hw = &iwdev->hw;
 	info.debug_mask = debug;
-	info.qs_handle = ldev->params.qos.prio_qos[0].qs_handle;
+	info.l2params.mss =
+		(ldev->params.mtu) ? ldev->params.mtu - I40IW_MTU_TO_MSS : I40IW_DEFAULT_MSS;
+	for (i = 0; i < I40E_CLIENT_MAX_USER_PRIORITY; i++) {
+		qset = ldev->params.qos.prio_qos[i].qs_handle;
+		info.l2params.qs_handle_list[i] = qset;
+		if (last_qset == I40IW_NO_QSET)
+			last_qset = qset;
+		else if ((qset != last_qset) && (qset != I40IW_NO_QSET))
+			iwdev->dcb = true;
+	}
 	info.exception_lan_queue = 1;
 	info.vchnl_send = i40iw_virtchnl_send;
 	status = i40iw_device_init(&iwdev->sc_dev, &info);
@@ -1416,6 +1428,8 @@ static void i40iw_deinit_device(struct i40iw_device *iwdev, bool reset, bool del
 	struct i40iw_sc_dev *dev = &iwdev->sc_dev;
 
 	i40iw_pr_info("state = %d\n", iwdev->init_state);
+	if (iwdev->param_wq)
+		destroy_workqueue(iwdev->param_wq);
 
 	switch (iwdev->init_state) {
 	case RDMA_DEV_REGISTERED:
@@ -1630,6 +1644,9 @@ static int i40iw_open(struct i40e_info *ldev, struct i40e_client *client)
 		iwdev->init_state = RDMA_DEV_REGISTERED;
 		iwdev->iw_status = 1;
 		i40iw_port_ibevent(iwdev);
+		iwdev->param_wq = alloc_ordered_workqueue("l2params", WQ_MEM_RECLAIM);
+		if(iwdev->param_wq == NULL)
+			break;
 		i40iw_pr_info("i40iw_open completed\n");
 		return 0;
 	} while (0);
@@ -1640,25 +1657,58 @@ static int i40iw_open(struct i40e_info *ldev, struct i40e_client *client)
 }
 
 /**
- * i40iw_l2param_change : handle qs handles for qos and mss change
+ * i40iw_l2params_worker - worker for l2 params change
+ * @work: work pointer for l2 params
+ */
+static void i40iw_l2params_worker(struct work_struct *work)
+{
+	struct l2params_work *dwork =
+	    container_of(work, struct l2params_work, work);
+	struct i40iw_device *iwdev = dwork->iwdev;
+
+	i40iw_change_l2params(&iwdev->sc_dev, &dwork->l2params);
+	atomic_dec(&iwdev->params_busy);
+	kfree(work);
+}
+
+/**
+ * i40iw_l2param_change - handle qs handles for qos and mss change
  * @ldev: lan device information
  * @client: client for paramater change
  * @params: new parameters from L2
  */
-static void i40iw_l2param_change(struct i40e_info *ldev,
-				 struct i40e_client *client,
+static void i40iw_l2param_change(struct i40e_info *ldev, struct i40e_client *client,
 				 struct i40e_params *params)
 {
 	struct i40iw_handler *hdl;
+	struct i40iw_l2params *l2params;
+	struct l2params_work *work;
 	struct i40iw_device *iwdev;
+	int i;
 
 	hdl = i40iw_find_i40e_handler(ldev);
 	if (!hdl)
 		return;
 
 	iwdev = &hdl->device;
-	if (params->mtu)
-		iwdev->mss = params->mtu - I40IW_MTU_TO_MSS;
+
+	if (atomic_read(&iwdev->params_busy))
+		return;
+
+
+	work = kzalloc(sizeof(*work), GFP_ATOMIC);
+	if (!work)
+		return;
+
+	atomic_inc(&iwdev->params_busy);
+
+	work->iwdev = iwdev;
+	l2params = &work->l2params;
+	for (i = 0; i < I40E_CLIENT_MAX_USER_PRIORITY; i++)
+		l2params->qs_handle_list[i] = params->qos.prio_qos[i].qs_handle;
+
+	INIT_WORK(&work->work, i40iw_l2params_worker);
+	queue_work(iwdev->param_wq, &work->work);
 }
 
 /**
diff --git a/drivers/infiniband/hw/i40iw/i40iw_osdep.h b/drivers/infiniband/hw/i40iw/i40iw_osdep.h
index 80f422bf3967..a6b18cd42e67 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_osdep.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_osdep.h
@@ -198,6 +198,8 @@ enum i40iw_status_code i40iw_cqp_manage_vf_pble_bp(struct i40iw_sc_dev *dev,
 void i40iw_cqp_spawn_worker(struct i40iw_sc_dev *dev,
 			    struct i40iw_virtchnl_work_info *work_info, u32 iw_vf_idx);
 void *i40iw_remove_head(struct list_head *list);
+void i40iw_qp_suspend_resume(struct i40iw_sc_dev *dev, struct i40iw_sc_qp *qp, bool suspend);
+void i40iw_qp_mss_modify(struct i40iw_sc_dev *dev, struct i40iw_sc_qp *qp);
 
 void i40iw_term_modify_qp(struct i40iw_sc_qp *qp, u8 next_state, u8 term, u8 term_len);
 void i40iw_terminate_done(struct i40iw_sc_qp *qp, int timeout_occurred);
diff --git a/drivers/infiniband/hw/i40iw/i40iw_p.h b/drivers/infiniband/hw/i40iw/i40iw_p.h
index a0b8ca10d67e..c9e8cb8c4410 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_p.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_p.h
@@ -65,6 +65,8 @@ enum i40iw_status_code i40iw_pf_init_vfhmc(struct i40iw_sc_dev *dev, u8 vf_hmc_f
 					   u32 *vf_cnt_array);
 
 /* cqp misc functions */
+void i40iw_change_l2params(struct i40iw_sc_dev *dev, struct i40iw_l2params *l2params);
+void i40iw_qp_add_qos(struct i40iw_sc_dev *dev, struct i40iw_sc_qp *qp);
 
 void i40iw_terminate_send_fin(struct i40iw_sc_qp *qp);
 
diff --git a/drivers/infiniband/hw/i40iw/i40iw_puda.c b/drivers/infiniband/hw/i40iw/i40iw_puda.c
index c62d354f7810..7541b0dada59 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_puda.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_puda.c
@@ -608,7 +608,8 @@ static enum i40iw_status_code i40iw_puda_qp_create(struct i40iw_puda_rsrc *rsrc)
 		ukqp->wqe_alloc_reg = (u32 __iomem *)(i40iw_get_hw_addr(qp->pd->dev) +
 						    I40E_VFPE_WQEALLOC1);
 
-	qp->qs_handle = qp->dev->qs_handle;
+	qp->user_pri = 0;
+	i40iw_qp_add_qos(rsrc->dev, qp);
 	i40iw_puda_qp_setctx(rsrc);
 	ret = i40iw_puda_qp_wqe(rsrc);
 	if (ret)
diff --git a/drivers/infiniband/hw/i40iw/i40iw_type.h b/drivers/infiniband/hw/i40iw/i40iw_type.h
index 2b1a04e9ca3c..b6f448a8d2e0 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_type.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_type.h
@@ -397,6 +397,9 @@ struct i40iw_sc_qp {
 	bool virtual_map;
 	bool flush_sq;
 	bool flush_rq;
+	u8 user_pri;
+	struct list_head list;
+	bool on_qoslist;
 	bool sq_flush;
 	enum i40iw_flush_opcode flush_code;
 	enum i40iw_term_eventtypes eventtype;
@@ -424,6 +427,12 @@ struct i40iw_vchnl_vf_msg_buffer {
 	char parm_buffer[I40IW_VCHNL_MAX_VF_MSG_SIZE - 1];
 };
 
+struct i40iw_qos {
+	struct list_head qplist;
+	spinlock_t lock;	/* qos list */
+	u16 qs_handle;
+};
+
 struct i40iw_vfdev {
 	struct i40iw_sc_dev *pf_dev;
 	u8 *hmc_info_mem;
@@ -482,7 +491,8 @@ struct i40iw_sc_dev {
 	const struct i40iw_vf_cqp_ops *iw_vf_cqp_ops;
 
 	struct i40iw_hmc_fpm_misc hmc_fpm_misc;
-	u16 qs_handle;
+	struct i40iw_qos qos[I40IW_MAX_USER_PRIORITY];
+	u16 mss;
 	u32 debug_mask;
 	u16 exception_lan_queue;
 	u8 hmc_fn_id;
@@ -564,7 +574,7 @@ struct i40iw_device_init_info {
 	struct i40iw_hw *hw;
 	void __iomem *bar0;
 	enum i40iw_status_code (*vchnl_send)(struct i40iw_sc_dev *, u32, u8 *, u16);
-	u16 qs_handle;
+	struct i40iw_l2params l2params;
 	u16 exception_lan_queue;
 	u8 hmc_fn_id;
 	bool is_pf;
@@ -722,6 +732,8 @@ struct i40iw_qp_host_ctx_info {
 	bool iwarp_info_valid;
 	bool err_rq_idx_valid;
 	u16 err_rq_idx;
+	bool add_to_qoslist;
+	u8 user_pri;
 };
 
 struct i40iw_aeqe_info {
@@ -886,7 +898,7 @@ struct i40iw_qhash_table_info {
 	bool ipv4_valid;
 	u8 mac_addr[6];
 	u16 vlan_id;
-	u16 qs_handle;
+	u8 user_pri;
 	u32 qp_num;
 	u32 dest_ip[4];
 	u32 src_ip[4];
diff --git a/drivers/infiniband/hw/i40iw/i40iw_utils.c b/drivers/infiniband/hw/i40iw/i40iw_utils.c
index 6fd043b1d714..cd9890263cd2 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_utils.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_utils.c
@@ -711,6 +711,51 @@ enum i40iw_status_code i40iw_cqp_sds_cmd(struct i40iw_sc_dev *dev,
 	return status;
 }
 
+/**
+ * i40iw_qp_suspend_resume - cqp command for suspend/resume
+ * @dev: hardware control device structure
+ * @qp: hardware control qp
+ * @suspend: flag if suspend or resume
+ */
+void i40iw_qp_suspend_resume(struct i40iw_sc_dev *dev, struct i40iw_sc_qp *qp, bool suspend)
+{
+	struct i40iw_device *iwdev = (struct i40iw_device *)dev->back_dev;
+	struct i40iw_cqp_request *cqp_request;
+	struct i40iw_sc_cqp *cqp = dev->cqp;
+	struct cqp_commands_info *cqp_info;
+	enum i40iw_status_code status;
+
+	cqp_request = i40iw_get_cqp_request(&iwdev->cqp, false);
+	if (!cqp_request)
+		return;
+
+	cqp_info = &cqp_request->info;
+	cqp_info->cqp_cmd = (suspend) ? OP_SUSPEND : OP_RESUME;
+	cqp_info->in.u.suspend_resume.cqp = cqp;
+	cqp_info->in.u.suspend_resume.qp = qp;
+	cqp_info->in.u.suspend_resume.scratch = (uintptr_t)cqp_request;
+	status = i40iw_handle_cqp_op(iwdev, cqp_request);
+	if (status)
+		i40iw_pr_err("CQP-OP QP Suspend/Resume fail");
+}
+
+/**
+ * i40iw_qp_mss_modify - modify mss for qp
+ * @dev: hardware control device structure
+ * @qp: hardware control qp
+ */
+void i40iw_qp_mss_modify(struct i40iw_sc_dev *dev, struct i40iw_sc_qp *qp)
+{
+	struct i40iw_device *iwdev = (struct i40iw_device *)dev->back_dev;
+	struct i40iw_qp *iwqp = (struct i40iw_qp *)qp->back_qp;
+	struct i40iw_modify_qp_info info;
+
+	memset(&info, 0, sizeof(info));
+	info.mss_change = true;
+	info.new_mss = dev->mss;
+	i40iw_hw_modify_qp(iwdev, iwqp, &info, false);
+}
+
 /**
  * i40iw_term_modify_qp - modify qp for term message
  * @qp: hardware control qp
diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
index 6329c971c22f..56e1c2c6ec34 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
@@ -254,7 +254,6 @@ static void i40iw_alloc_push_page(struct i40iw_device *iwdev, struct i40iw_sc_qp
 {
 	struct i40iw_cqp_request *cqp_request;
 	struct cqp_commands_info *cqp_info;
-	struct i40iw_sc_dev *dev = &iwdev->sc_dev;
 	enum i40iw_status_code status;
 
 	if (qp->push_idx != I40IW_INVALID_PUSH_PAGE_INDEX)
@@ -270,7 +269,7 @@ static void i40iw_alloc_push_page(struct i40iw_device *iwdev, struct i40iw_sc_qp
 	cqp_info->cqp_cmd = OP_MANAGE_PUSH_PAGE;
 	cqp_info->post_sq = 1;
 
-	cqp_info->in.u.manage_push_page.info.qs_handle = dev->qs_handle;
+	cqp_info->in.u.manage_push_page.info.qs_handle = qp->qs_handle;
 	cqp_info->in.u.manage_push_page.info.free_page = 0;
 	cqp_info->in.u.manage_push_page.cqp = &iwdev->cqp.sc_cqp;
 	cqp_info->in.u.manage_push_page.scratch = (uintptr_t)cqp_request;
@@ -292,7 +291,6 @@ static void i40iw_dealloc_push_page(struct i40iw_device *iwdev, struct i40iw_sc_
 {
 	struct i40iw_cqp_request *cqp_request;
 	struct cqp_commands_info *cqp_info;
-	struct i40iw_sc_dev *dev = &iwdev->sc_dev;
 	enum i40iw_status_code status;
 
 	if (qp->push_idx == I40IW_INVALID_PUSH_PAGE_INDEX)
@@ -307,7 +305,7 @@ static void i40iw_dealloc_push_page(struct i40iw_device *iwdev, struct i40iw_sc_
 	cqp_info->post_sq = 1;
 
 	cqp_info->in.u.manage_push_page.info.push_idx = qp->push_idx;
-	cqp_info->in.u.manage_push_page.info.qs_handle = dev->qs_handle;
+	cqp_info->in.u.manage_push_page.info.qs_handle = qp->qs_handle;
 	cqp_info->in.u.manage_push_page.info.free_page = 1;
 	cqp_info->in.u.manage_push_page.cqp = &iwdev->cqp.sc_cqp;
 	cqp_info->in.u.manage_push_page.scratch = (uintptr_t)cqp_request;
-- 
cgit v1.2.1


From d62d563424e3da0c0a1176f38c0d49c7ad91fbc1 Mon Sep 17 00:00:00 2001
From: Henry Orosco <henry.orosco@intel.com>
Date: Wed, 19 Oct 2016 15:32:53 -0500
Subject: i40iw: Enable message packing

Remove the parameter to disable message packing and
always enable it.

Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_ctrl.c | 5 -----
 drivers/infiniband/hw/i40iw/i40iw_main.c | 2 +-
 drivers/infiniband/hw/i40iw/i40iw_type.h | 2 +-
 3 files changed, 2 insertions(+), 7 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
index 31c4a0c26786..6c6a1ef02d46 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
@@ -429,12 +429,10 @@ static enum i40iw_status_code i40iw_sc_cqp_init(struct i40iw_sc_cqp *cqp,
 /**
  * i40iw_sc_cqp_create - create cqp during bringup
  * @cqp: struct for cqp hw
- * @disable_pfpdus: if pfpdu to be disabled
  * @maj_err: If error, major err number
  * @min_err: If error, minor err number
  */
 static enum i40iw_status_code i40iw_sc_cqp_create(struct i40iw_sc_cqp *cqp,
-						  bool disable_pfpdus,
 						  u16 *maj_err,
 						  u16 *min_err)
 {
@@ -453,9 +451,6 @@ static enum i40iw_status_code i40iw_sc_cqp_create(struct i40iw_sc_cqp *cqp,
 	temp = LS_64(cqp->hw_sq_size, I40IW_CQPHC_SQSIZE) |
 	       LS_64(cqp->struct_ver, I40IW_CQPHC_SVER);
 
-	if (disable_pfpdus)
-		temp |= LS_64(1, I40IW_CQPHC_DISABLE_PFPDUS);
-
 	set_64bit_val(cqp->host_ctx, 0, temp);
 	set_64bit_val(cqp->host_ctx, 8, cqp->sq_pa);
 	temp = LS_64(cqp->enabled_vf_count, I40IW_CQPHC_ENABLED_VFS) |
diff --git a/drivers/infiniband/hw/i40iw/i40iw_main.c b/drivers/infiniband/hw/i40iw/i40iw_main.c
index 40aac87c128c..e6abdafd1b6f 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_main.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_main.c
@@ -603,7 +603,7 @@ static enum i40iw_status_code i40iw_create_cqp(struct i40iw_device *iwdev)
 		i40iw_pr_err("cqp init status %d\n", status);
 		goto exit;
 	}
-	status = dev->cqp_ops->cqp_create(dev->cqp, true, &maj_err, &min_err);
+	status = dev->cqp_ops->cqp_create(dev->cqp, &maj_err, &min_err);
 	if (status) {
 		i40iw_pr_err("cqp create status %d maj_err %d min_err %d\n",
 			     status, maj_err, min_err);
diff --git a/drivers/infiniband/hw/i40iw/i40iw_type.h b/drivers/infiniband/hw/i40iw/i40iw_type.h
index b6f448a8d2e0..d1847e693025 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_type.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_type.h
@@ -988,7 +988,7 @@ struct i40iw_cqp_query_fpm_values {
 struct i40iw_cqp_ops {
 	enum i40iw_status_code (*cqp_init)(struct i40iw_sc_cqp *,
 					   struct i40iw_cqp_init_info *);
-	enum i40iw_status_code (*cqp_create)(struct i40iw_sc_cqp *, bool, u16 *, u16 *);
+	enum i40iw_status_code (*cqp_create)(struct i40iw_sc_cqp *, u16 *, u16 *);
 	void (*cqp_post_sq)(struct i40iw_sc_cqp *);
 	u64 *(*cqp_get_next_send_wqe)(struct i40iw_sc_cqp *, u64 scratch);
 	enum i40iw_status_code (*cqp_destroy)(struct i40iw_sc_cqp *);
-- 
cgit v1.2.1


From 7581e96ca4de26da7237d507ac9cf519753e1787 Mon Sep 17 00:00:00 2001
From: Henry Orosco <henry.orosco@intel.com>
Date: Wed, 19 Oct 2016 15:33:32 -0500
Subject: i40iw: Remove workaround for pre-production errata

Pre-production silicon incorrectly truncates 4 bytes of the MPA
packet in UDP loopback case. Remove the workaround as it is no
longer necessary.

Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_cm.c    | 26 +++-----------------------
 drivers/infiniband/hw/i40iw/i40iw_cm.h    |  2 --
 drivers/infiniband/hw/i40iw/i40iw_utils.c |  2 +-
 3 files changed, 4 insertions(+), 26 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.c b/drivers/infiniband/hw/i40iw/i40iw_cm.c
index 24b22e99003c..9e447b5951fc 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_cm.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_cm.c
@@ -361,15 +361,6 @@ static void i40iw_cleanup_retrans_entry(struct i40iw_cm_node *cm_node)
 	spin_unlock_irqrestore(&cm_node->retrans_list_lock, flags);
 }
 
-static bool is_remote_ne020_or_chelsio(struct i40iw_cm_node *cm_node)
-{
-	if ((cm_node->rem_mac[0] == 0x0) &&
-	    (((cm_node->rem_mac[1] == 0x12) && (cm_node->rem_mac[2] == 0x55)) ||
-	     ((cm_node->rem_mac[1] == 0x07 && (cm_node->rem_mac[2] == 0x43)))))
-		return true;
-	return false;
-}
-
 /**
  * i40iw_form_cm_frame - get a free packet and build frame
  * @cm_node: connection's node ionfo to use in frame
@@ -410,11 +401,8 @@ static struct i40iw_puda_buf *i40iw_form_cm_frame(struct i40iw_cm_node *cm_node,
 	if (hdr)
 		hdr_len = hdr->size;
 
-	if (pdata) {
+	if (pdata)
 		pd_len = pdata->size;
-		if (!is_remote_ne020_or_chelsio(cm_node))
-			pd_len += MPA_ZERO_PAD_LEN;
-	}
 
 	if (cm_node->vlan_id < VLAN_TAG_PRESENT)
 		eth_hlen += 4;
@@ -3587,7 +3575,7 @@ int i40iw_accept(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param)
 	iwqp->cm_node = (void *)cm_node;
 	cm_node->iwqp = iwqp;
 
-	buf_len = conn_param->private_data_len + I40IW_MAX_IETF_SIZE + MPA_ZERO_PAD_LEN;
+	buf_len = conn_param->private_data_len + I40IW_MAX_IETF_SIZE;
 
 	status = i40iw_allocate_dma_mem(dev->hw, &iwqp->ietf_mem, buf_len, 1);
 
@@ -3621,18 +3609,10 @@ int i40iw_accept(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param)
 		iwqp->lsmm_mr = ibmr;
 		if (iwqp->page)
 			iwqp->sc_qp.qp_uk.sq_base = kmap(iwqp->page);
-		if (is_remote_ne020_or_chelsio(cm_node))
-			dev->iw_priv_qp_ops->qp_send_lsmm(
-							&iwqp->sc_qp,
+		dev->iw_priv_qp_ops->qp_send_lsmm(&iwqp->sc_qp,
 							iwqp->ietf_mem.va,
 							(accept.size + conn_param->private_data_len),
 							ibmr->lkey);
-		else
-			dev->iw_priv_qp_ops->qp_send_lsmm(
-							&iwqp->sc_qp,
-							iwqp->ietf_mem.va,
-							(accept.size + conn_param->private_data_len + MPA_ZERO_PAD_LEN),
-							ibmr->lkey);
 
 	} else {
 		if (iwqp->page)
diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.h b/drivers/infiniband/hw/i40iw/i40iw_cm.h
index 945ed26f8dba..24615c24cb04 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_cm.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_cm.h
@@ -56,8 +56,6 @@
 
 #define I40IW_MAX_IETF_SIZE      32
 
-#define MPA_ZERO_PAD_LEN	4
-
 /* IETF RTR MSG Fields               */
 #define IETF_PEER_TO_PEER       0x8000
 #define IETF_FLPDU_ZERO_LEN     0x4000
diff --git a/drivers/infiniband/hw/i40iw/i40iw_utils.c b/drivers/infiniband/hw/i40iw/i40iw_utils.c
index cd9890263cd2..4e880e8689eb 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_utils.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_utils.c
@@ -1253,7 +1253,7 @@ enum i40iw_status_code i40iw_puda_get_tcpip_info(struct i40iw_puda_completion_in
 
 	buf->totallen = pkt_len + buf->maclen;
 
-	if (info->payload_len < buf->totallen - 4) {
+	if (info->payload_len < buf->totallen) {
 		i40iw_pr_err("payload_len = 0x%x totallen expected0x%x\n",
 			     info->payload_len, buf->totallen);
 		return I40IW_ERR_INVALID_SIZE;
-- 
cgit v1.2.1


From 7cba2cc13e12c824ad7e414b3834dc3df05fbf46 Mon Sep 17 00:00:00 2001
From: Henry Orosco <henry.orosco@intel.com>
Date: Wed, 19 Oct 2016 15:33:58 -0500
Subject: i40iw: Set MAX IRD, MAX ORD size to max supported value

Set the MAX_IRD and MAX_ORD size negotiated to the maximum
supported values.

Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_ctrl.c | 5 -----
 drivers/infiniband/hw/i40iw/i40iw_user.h | 5 ++---
 2 files changed, 2 insertions(+), 8 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
index 6c6a1ef02d46..6bf2a19f3cba 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
@@ -2621,11 +2621,6 @@ static enum i40iw_status_code i40iw_sc_qp_setctx(
 			      152,
 			      LS_64(iw->last_byte_sent, I40IWQPC_LASTBYTESENT));
 
-		/*
-		* Hard-code IRD_SIZE to hw-limit, 128, in qpctx, i.e matching an
-		*advertisable IRD of 64
-		*/
-		iw->ird_size = I40IW_QPCTX_ENCD_MAXIRD;
 		set_64bit_val(qp_ctx,
 			      160,
 			      LS_64(iw->ord_size, I40IWQPC_ORDSIZE) |
diff --git a/drivers/infiniband/hw/i40iw/i40iw_user.h b/drivers/infiniband/hw/i40iw/i40iw_user.h
index 276bcefffd7e..e65c2baa7db2 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_user.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_user.h
@@ -72,10 +72,9 @@ enum i40iw_device_capabilities_const {
 	I40IW_MAX_SQ_PAYLOAD_SIZE =		2145386496,
 	I40IW_MAX_INLINE_DATA_SIZE =		48,
 	I40IW_MAX_PUSHMODE_INLINE_DATA_SIZE =	48,
-	I40IW_MAX_IRD_SIZE =			32,
-	I40IW_QPCTX_ENCD_MAXIRD =		3,
+	I40IW_MAX_IRD_SIZE =			63,
+	I40IW_MAX_ORD_SIZE =			127,
 	I40IW_MAX_WQ_ENTRIES =			2048,
-	I40IW_MAX_ORD_SIZE =			32,
 	I40IW_Q2_BUFFER_SIZE =			(248 + 100),
 	I40IW_QP_CTX_SIZE =			248
 };
-- 
cgit v1.2.1


From 68583ca2a183c6368f4c333fa989685fba7cf325 Mon Sep 17 00:00:00 2001
From: Henry Orosco <henry.orosco@intel.com>
Date: Sat, 19 Nov 2016 20:26:25 -0600
Subject: i40iw: Convert page_size to encoded value

Passed in page_size was used as encoded value for writing
the WQE and passed in value was usually 4096. This was
working out since bit 0 was 0 and implies 4KB pages,
but would not work for other page sizes.

Signed-off-by: Mustafa Ismail <mustafa.ismail@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_ctrl.c | 12 +++++++++---
 drivers/infiniband/hw/i40iw/i40iw_type.h |  5 +++++
 2 files changed, 14 insertions(+), 3 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
index 6bf2a19f3cba..8417452a7a9c 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
@@ -2747,7 +2747,9 @@ static enum i40iw_status_code i40iw_sc_alloc_stag(
 	u64 *wqe;
 	struct i40iw_sc_cqp *cqp;
 	u64 header;
+	enum i40iw_page_size page_size;
 
+	page_size = (info->page_size == 0x200000) ? I40IW_PAGE_SIZE_2M : I40IW_PAGE_SIZE_4K;
 	cqp = dev->cqp;
 	wqe = i40iw_sc_cqp_get_next_send_wqe(cqp, scratch);
 	if (!wqe)
@@ -2767,7 +2769,7 @@ static enum i40iw_status_code i40iw_sc_alloc_stag(
 		 LS_64(1, I40IW_CQPSQ_STAG_MR) |
 		 LS_64(info->access_rights, I40IW_CQPSQ_STAG_ARIGHTS) |
 		 LS_64(info->chunk_size, I40IW_CQPSQ_STAG_LPBLSIZE) |
-		 LS_64(info->page_size, I40IW_CQPSQ_STAG_HPAGESIZE) |
+		 LS_64(page_size, I40IW_CQPSQ_STAG_HPAGESIZE) |
 		 LS_64(info->remote_access, I40IW_CQPSQ_STAG_REMACCENABLED) |
 		 LS_64(info->use_hmc_fcn_index, I40IW_CQPSQ_STAG_USEHMCFNIDX) |
 		 LS_64(info->use_pf_rid, I40IW_CQPSQ_STAG_USEPFRID) |
@@ -2803,7 +2805,9 @@ static enum i40iw_status_code i40iw_sc_mr_reg_non_shared(
 	u32 pble_obj_cnt;
 	bool remote_access;
 	u8 addr_type;
+	enum i40iw_page_size page_size;
 
+	page_size = (info->page_size == 0x200000) ? I40IW_PAGE_SIZE_2M : I40IW_PAGE_SIZE_4K;
 	if (info->access_rights & (I40IW_ACCESS_FLAGS_REMOTEREAD_ONLY |
 				   I40IW_ACCESS_FLAGS_REMOTEWRITE_ONLY))
 		remote_access = true;
@@ -2846,7 +2850,7 @@ static enum i40iw_status_code i40iw_sc_mr_reg_non_shared(
 	header = LS_64(I40IW_CQP_OP_REG_MR, I40IW_CQPSQ_OPCODE) |
 		 LS_64(1, I40IW_CQPSQ_STAG_MR) |
 		 LS_64(info->chunk_size, I40IW_CQPSQ_STAG_LPBLSIZE) |
-		 LS_64(info->page_size, I40IW_CQPSQ_STAG_HPAGESIZE) |
+		 LS_64(page_size, I40IW_CQPSQ_STAG_HPAGESIZE) |
 		 LS_64(info->access_rights, I40IW_CQPSQ_STAG_ARIGHTS) |
 		 LS_64(remote_access, I40IW_CQPSQ_STAG_REMACCENABLED) |
 		 LS_64(addr_type, I40IW_CQPSQ_STAG_VABASEDTO) |
@@ -3061,7 +3065,9 @@ enum i40iw_status_code i40iw_sc_mr_fast_register(
 	u64 temp, header;
 	u64 *wqe;
 	u32 wqe_idx;
+	enum i40iw_page_size page_size;
 
+	page_size = (info->page_size == 0x200000) ? I40IW_PAGE_SIZE_2M : I40IW_PAGE_SIZE_4K;
 	wqe = i40iw_qp_get_next_send_wqe(&qp->qp_uk, &wqe_idx, I40IW_QP_WQE_MIN_SIZE,
 					 0, info->wr_id);
 	if (!wqe)
@@ -3088,7 +3094,7 @@ enum i40iw_status_code i40iw_sc_mr_fast_register(
 		 LS_64(info->stag_idx, I40IWQPSQ_STAGINDEX) |
 		 LS_64(I40IWQP_OP_FAST_REGISTER, I40IWQPSQ_OPCODE) |
 		 LS_64(info->chunk_size, I40IWQPSQ_LPBLSIZE) |
-		 LS_64(info->page_size, I40IWQPSQ_HPAGESIZE) |
+		 LS_64(page_size, I40IWQPSQ_HPAGESIZE) |
 		 LS_64(info->access_rights, I40IWQPSQ_STAGRIGHTS) |
 		 LS_64(info->addr_type, I40IWQPSQ_VABASEDTO) |
 		 LS_64(info->read_fence, I40IWQPSQ_READFENCE) |
diff --git a/drivers/infiniband/hw/i40iw/i40iw_type.h b/drivers/infiniband/hw/i40iw/i40iw_type.h
index d1847e693025..928d91b956a8 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_type.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_type.h
@@ -74,6 +74,11 @@ struct i40iw_priv_qp_ops;
 struct i40iw_priv_cq_ops;
 struct i40iw_hmc_ops;
 
+enum i40iw_page_size {
+	I40IW_PAGE_SIZE_4K,
+	I40IW_PAGE_SIZE_2M
+};
+
 enum i40iw_resource_indicator_type {
 	I40IW_RSRC_INDICATOR_TYPE_ADAPTER = 0,
 	I40IW_RSRC_INDICATOR_TYPE_CQ,
-- 
cgit v1.2.1


From e69c5093617afdbd2ab02c289d0adaac044dff66 Mon Sep 17 00:00:00 2001
From: Henry Orosco <henry.orosco@intel.com>
Date: Wed, 9 Nov 2016 21:24:48 -0600
Subject: i40iw: Use vector when creating CQs

Assign each CEQ vector to a different CPU when possible, then
when creating a CQ, use the vector for the CEQ id. This
allows completion work to be distributed over multiple cores.

Signed-off-by: Mustafa Ismail <mustafa.ismail@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_main.c  | 8 +++++++-
 drivers/infiniband/hw/i40iw/i40iw_verbs.c | 5 +++--
 2 files changed, 10 insertions(+), 3 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_main.c b/drivers/infiniband/hw/i40iw/i40iw_main.c
index e6abdafd1b6f..ed24831344b1 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_main.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_main.c
@@ -270,6 +270,7 @@ static void i40iw_disable_irq(struct i40iw_sc_dev *dev,
 		i40iw_wr32(dev->hw, I40E_PFINT_DYN_CTLN(msix_vec->idx - 1), 0);
 	else
 		i40iw_wr32(dev->hw, I40E_VFINT_DYN_CTLN1(msix_vec->idx - 1), 0);
+	irq_set_affinity_hint(msix_vec->irq, NULL);
 	free_irq(msix_vec->irq, dev_id);
 }
 
@@ -688,6 +689,7 @@ static enum i40iw_status_code i40iw_configure_ceq_vector(struct i40iw_device *iw
 							 struct i40iw_msix_vector *msix_vec)
 {
 	enum i40iw_status_code status;
+	cpumask_t mask;
 
 	if (iwdev->msix_shared && !ceq_id) {
 		tasklet_init(&iwdev->dpc_tasklet, i40iw_dpc, (unsigned long)iwdev);
@@ -697,12 +699,15 @@ static enum i40iw_status_code i40iw_configure_ceq_vector(struct i40iw_device *iw
 		status = request_irq(msix_vec->irq, i40iw_ceq_handler, 0, "CEQ", iwceq);
 	}
 
+	cpumask_clear(&mask);
+	cpumask_set_cpu(msix_vec->cpu_affinity, &mask);
+	irq_set_affinity_hint(msix_vec->irq, &mask);
+
 	if (status) {
 		i40iw_pr_err("ceq irq config fail\n");
 		return I40IW_ERR_CONFIG;
 	}
 	msix_vec->ceq_id = ceq_id;
-	msix_vec->cpu_affinity = 0;
 
 	return 0;
 }
@@ -1396,6 +1401,7 @@ static enum i40iw_status_code i40iw_save_msix_info(struct i40iw_device *iwdev,
 	for (i = 0, ceq_idx = 0; i < iwdev->msix_count; i++, iw_qvinfo++) {
 		iwdev->iw_msixtbl[i].idx = ldev->msix_entries[i].entry;
 		iwdev->iw_msixtbl[i].irq = ldev->msix_entries[i].vector;
+		iwdev->iw_msixtbl[i].cpu_affinity = ceq_idx;
 		if (i == 0) {
 			iw_qvinfo->aeq_idx = 0;
 			if (iwdev->msix_shared)
diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
index 56e1c2c6ec34..aacaa0fcfb77 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
@@ -1135,7 +1135,8 @@ static struct ib_cq *i40iw_create_cq(struct ib_device *ibdev,
 	ukinfo->cq_id = cq_num;
 	iwcq->ibcq.cqe = info.cq_uk_init_info.cq_size;
 	info.ceqe_mask = 0;
-	info.ceq_id = 0;
+	if (attr->comp_vector < iwdev->ceqs_count)
+		info.ceq_id = attr->comp_vector;
 	info.ceq_id_valid = true;
 	info.ceqe_mask = 1;
 	info.type = I40IW_CQ_TYPE_IWARP;
@@ -2619,7 +2620,7 @@ static struct i40iw_ib_device *i40iw_init_rdma_device(struct i40iw_device *iwdev
 	    (1ull << IB_USER_VERBS_CMD_POST_RECV) |
 	    (1ull << IB_USER_VERBS_CMD_POST_SEND);
 	iwibdev->ibdev.phys_port_cnt = 1;
-	iwibdev->ibdev.num_comp_vectors = 1;
+	iwibdev->ibdev.num_comp_vectors = iwdev->ceqs_count;
 	iwibdev->ibdev.dma_device = &pcidev->dev;
 	iwibdev->ibdev.dev.parent = &pcidev->dev;
 	iwibdev->ibdev.query_port = i40iw_query_port;
-- 
cgit v1.2.1


From 01d0b36798732d826fbf84de8961a09b3a2fbf3f Mon Sep 17 00:00:00 2001
From: Henry Orosco <henry.orosco@intel.com>
Date: Wed, 9 Nov 2016 21:26:39 -0600
Subject: i40iw: Correct values for max_recv_sge, max_send_sge

When creating QPs, ensure init_attr->cap.max_recv_sge
is clipped to MAX_FRAG_COUNT.

Expose MAX_FRAG_COUNT for max_recv_sge and max_send_sge in
i40iw_query_qp().

Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Reviewed-By: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_verbs.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
index aacaa0fcfb77..6b516d6156dc 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
@@ -608,6 +608,9 @@ static struct ib_qp *i40iw_create_qp(struct ib_pd *ibpd,
 	if (init_attr->cap.max_send_sge > I40IW_MAX_WQ_FRAGMENT_COUNT)
 		init_attr->cap.max_send_sge = I40IW_MAX_WQ_FRAGMENT_COUNT;
 
+	if (init_attr->cap.max_recv_sge > I40IW_MAX_WQ_FRAGMENT_COUNT)
+		init_attr->cap.max_recv_sge = I40IW_MAX_WQ_FRAGMENT_COUNT;
+
 	memset(&init_info, 0, sizeof(init_info));
 
 	sq_size = init_attr->cap.max_send_wr;
@@ -813,8 +816,9 @@ static int i40iw_query_qp(struct ib_qp *ibqp,
 	attr->qp_access_flags = 0;
 	attr->cap.max_send_wr = qp->qp_uk.sq_size;
 	attr->cap.max_recv_wr = qp->qp_uk.rq_size;
-	attr->cap.max_recv_sge = 1;
 	attr->cap.max_inline_data = I40IW_MAX_INLINE_DATA_SIZE;
+	attr->cap.max_send_sge = I40IW_MAX_WQ_FRAGMENT_COUNT;
+	attr->cap.max_recv_sge = I40IW_MAX_WQ_FRAGMENT_COUNT;
 	init_attr->event_handler = iwqp->ibqp.event_handler;
 	init_attr->qp_context = iwqp->ibqp.qp_context;
 	init_attr->send_cq = iwqp->ibqp.send_cq;
-- 
cgit v1.2.1


From c38d7e0d08421a53cea4e09b76b3453d499fbd67 Mon Sep 17 00:00:00 2001
From: Henry Orosco <henry.orosco@intel.com>
Date: Wed, 9 Nov 2016 21:27:02 -0600
Subject: i40iw: Fix for LAN handler removal

If i40iw_open() fails for any reason, the LAN handler
is not being removed. Modify i40iw_deinit_device()
to always remove the handler.

Signed-off-by: Mustafa Ismail <mustafa.ismail@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_main.c | 10 ++++------
 1 file changed, 4 insertions(+), 6 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_main.c b/drivers/infiniband/hw/i40iw/i40iw_main.c
index ed24831344b1..db9fd31b5c72 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_main.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_main.c
@@ -1422,12 +1422,11 @@ static enum i40iw_status_code i40iw_save_msix_info(struct i40iw_device *iwdev,
  * i40iw_deinit_device - clean up the device resources
  * @iwdev: iwarp device
  * @reset: true if called before reset
- * @del_hdl: true if delete hdl entry
  *
  * Destroy the ib device interface, remove the mac ip entry and ipv4/ipv6 addresses,
  * destroy the device queues and free the pble and the hmc objects
  */
-static void i40iw_deinit_device(struct i40iw_device *iwdev, bool reset, bool del_hdl)
+static void i40iw_deinit_device(struct i40iw_device *iwdev, bool reset)
 {
 	struct i40e_info *ldev = iwdev->ldev;
 
@@ -1492,8 +1491,7 @@ static void i40iw_deinit_device(struct i40iw_device *iwdev, bool reset, bool del
 		break;
 	}
 
-	if (del_hdl)
-		i40iw_del_handler(i40iw_find_i40e_handler(ldev));
+	i40iw_del_handler(i40iw_find_i40e_handler(ldev));
 	kfree(iwdev->hdl);
 }
 
@@ -1658,7 +1656,7 @@ static int i40iw_open(struct i40e_info *ldev, struct i40e_client *client)
 	} while (0);
 
 	i40iw_pr_err("status = %d last completion = %d\n", status, iwdev->init_state);
-	i40iw_deinit_device(iwdev, false, false);
+	i40iw_deinit_device(iwdev, false);
 	return -ERESTART;
 }
 
@@ -1736,7 +1734,7 @@ static void i40iw_close(struct i40e_info *ldev, struct i40e_client *client, bool
 
 	iwdev = &hdl->device;
 	destroy_workqueue(iwdev->virtchnl_wq);
-	i40iw_deinit_device(iwdev, reset, true);
+	i40iw_deinit_device(iwdev, reset);
 }
 
 /**
-- 
cgit v1.2.1


From e7f9774af591d346990f1d6dfca0ee9caeb52756 Mon Sep 17 00:00:00 2001
From: Henry Orosco <henry.orosco@intel.com>
Date: Wed, 9 Nov 2016 21:28:02 -0600
Subject: i40iw: Optimize inline data copy

Use memcpy for inline data copy in sends
and writes instead of byte by byte copy.

Signed-off-by: Mustafa Ismail <mustafa.ismail@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_uk.c | 24 ++++++++++--------------
 1 file changed, 10 insertions(+), 14 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_uk.c b/drivers/infiniband/hw/i40iw/i40iw_uk.c
index 4d28c3cb03cc..47cb2e063ee3 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_uk.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_uk.c
@@ -430,7 +430,7 @@ static enum i40iw_status_code i40iw_inline_rdma_write(struct i40iw_qp_uk *qp,
 	struct i40iw_inline_rdma_write *op_info;
 	u64 *push;
 	u64 header = 0;
-	u32 i, wqe_idx;
+	u32 wqe_idx;
 	enum i40iw_status_code ret_code;
 	bool read_fence = false;
 	u8 wqe_size;
@@ -465,14 +465,12 @@ static enum i40iw_status_code i40iw_inline_rdma_write(struct i40iw_qp_uk *qp,
 	src = (u8 *)(op_info->data);
 
 	if (op_info->len <= 16) {
-		for (i = 0; i < op_info->len; i++, src++, dest++)
-			*dest = *src;
+		memcpy(dest, src, op_info->len);
 	} else {
-		for (i = 0; i < 16; i++, src++, dest++)
-			*dest = *src;
+		memcpy(dest, src, 16);
+		src += 16;
 		dest = (u8 *)wqe + 32;
-		for (; i < op_info->len; i++, src++, dest++)
-			*dest = *src;
+		memcpy(dest, src, op_info->len - 16);
 	}
 
 	wmb(); /* make sure WQE is populated before valid bit is set */
@@ -507,7 +505,7 @@ static enum i40iw_status_code i40iw_inline_send(struct i40iw_qp_uk *qp,
 	u8 *dest, *src;
 	struct i40iw_post_inline_send *op_info;
 	u64 header;
-	u32 wqe_idx, i;
+	u32 wqe_idx;
 	enum i40iw_status_code ret_code;
 	bool read_fence = false;
 	u8 wqe_size;
@@ -540,14 +538,12 @@ static enum i40iw_status_code i40iw_inline_send(struct i40iw_qp_uk *qp,
 	src = (u8 *)(op_info->data);
 
 	if (op_info->len <= 16) {
-		for (i = 0; i < op_info->len; i++, src++, dest++)
-			*dest = *src;
+		memcpy(dest, src, op_info->len);
 	} else {
-		for (i = 0; i < 16; i++, src++, dest++)
-			*dest = *src;
+		memcpy(dest, src, 16);
+		src += 16;
 		dest = (u8 *)wqe + 32;
-		for (; i < op_info->len; i++, src++, dest++)
-			*dest = *src;
+		memcpy(dest, src, op_info->len - 16);
 	}
 
 	wmb(); /* make sure WQE is populated before valid bit is set */
-- 
cgit v1.2.1


From 85a87c90ee90217da1b05a77bbb47ebe31a2f124 Mon Sep 17 00:00:00 2001
From: Henry Orosco <henry.orosco@intel.com>
Date: Wed, 9 Nov 2016 21:30:28 -0600
Subject: i40iw: Query device accounts for internal rsrc

Some resources are consumed internally and not available to the user.
After hw is initialized, figure out how many resources are consumed
and subtract those numbers from the initial max device capability in
i40iw_query_device().

Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw.h       |  4 ++++
 drivers/infiniband/hw/i40iw/i40iw_hw.c    |  2 +-
 drivers/infiniband/hw/i40iw/i40iw_main.c  | 15 +++++++++++++++
 drivers/infiniband/hw/i40iw/i40iw_user.h  |  3 ++-
 drivers/infiniband/hw/i40iw/i40iw_verbs.c |  8 ++++----
 5 files changed, 26 insertions(+), 6 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw.h b/drivers/infiniband/hw/i40iw/i40iw.h
index 4a0c12bd9caa..dac9a6bcc631 100644
--- a/drivers/infiniband/hw/i40iw/i40iw.h
+++ b/drivers/infiniband/hw/i40iw/i40iw.h
@@ -303,6 +303,10 @@ struct i40iw_device {
 	u32 mr_stagmask;
 	u32 mpa_version;
 	bool dcb;
+	u32 used_pds;
+	u32 used_cqs;
+	u32 used_mrs;
+	u32 used_qps;
 };
 
 struct i40iw_ib_device {
diff --git a/drivers/infiniband/hw/i40iw/i40iw_hw.c b/drivers/infiniband/hw/i40iw/i40iw_hw.c
index b94727ffc862..5e2c16c725e3 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_hw.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_hw.c
@@ -62,7 +62,7 @@ u32 i40iw_initialize_hw_resources(struct i40iw_device *iwdev)
 	max_mr = iwdev->sc_dev.hmc_info->hmc_obj[I40IW_HMC_IW_MR].cnt;
 	arp_table_size = iwdev->sc_dev.hmc_info->hmc_obj[I40IW_HMC_IW_ARP].cnt;
 	iwdev->max_cqe = 0xFFFFF;
-	num_pds = max_qp * 4;
+	num_pds = I40IW_MAX_PDS;
 	resources_size = sizeof(struct i40iw_arp_entry) * arp_table_size;
 	resources_size += sizeof(unsigned long) * BITS_TO_LONGS(max_qp);
 	resources_size += sizeof(unsigned long) * BITS_TO_LONGS(max_mr);
diff --git a/drivers/infiniband/hw/i40iw/i40iw_main.c b/drivers/infiniband/hw/i40iw/i40iw_main.c
index db9fd31b5c72..9d3b9ee20ba7 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_main.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_main.c
@@ -1557,6 +1557,20 @@ exit:
 	return status;
 }
 
+/**
+ * i40iw_get_used_rsrc - determine resources used internally
+ * @iwdev: iwarp device
+ *
+ * Called after internal allocations
+ */
+static void i40iw_get_used_rsrc(struct i40iw_device *iwdev)
+{
+	iwdev->used_pds = find_next_zero_bit(iwdev->allocated_pds, iwdev->max_pd, 0);
+	iwdev->used_qps = find_next_zero_bit(iwdev->allocated_qps, iwdev->max_qp, 0);
+	iwdev->used_cqs = find_next_zero_bit(iwdev->allocated_cqs, iwdev->max_cq, 0);
+	iwdev->used_mrs = find_next_zero_bit(iwdev->allocated_mrs, iwdev->max_mr, 0);
+}
+
 /**
  * i40iw_open - client interface operation open for iwarp/uda device
  * @ldev: lan device information
@@ -1629,6 +1643,7 @@ static int i40iw_open(struct i40e_info *ldev, struct i40e_client *client)
 		status = i40iw_initialize_hw_resources(iwdev);
 		if (status)
 			break;
+		i40iw_get_used_rsrc(iwdev);
 		dev->ccq_ops->ccq_arm(dev->ccq);
 		status = i40iw_hmc_init_pble(&iwdev->sc_dev, iwdev->pble_rsrc);
 		if (status)
diff --git a/drivers/infiniband/hw/i40iw/i40iw_user.h b/drivers/infiniband/hw/i40iw/i40iw_user.h
index e65c2baa7db2..66263fced68f 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_user.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_user.h
@@ -76,7 +76,8 @@ enum i40iw_device_capabilities_const {
 	I40IW_MAX_ORD_SIZE =			127,
 	I40IW_MAX_WQ_ENTRIES =			2048,
 	I40IW_Q2_BUFFER_SIZE =			(248 + 100),
-	I40IW_QP_CTX_SIZE =			248
+	I40IW_QP_CTX_SIZE =			248,
+	I40IW_MAX_PDS = 			32768
 };
 
 #define i40iw_handle void *
diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
index 6b516d6156dc..d20ee118f366 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
@@ -67,13 +67,13 @@ static int i40iw_query_device(struct ib_device *ibdev,
 	props->vendor_part_id = iwdev->ldev->pcidev->device;
 	props->hw_ver = (u32)iwdev->sc_dev.hw_rev;
 	props->max_mr_size = I40IW_MAX_OUTBOUND_MESSAGE_SIZE;
-	props->max_qp = iwdev->max_qp;
+	props->max_qp = iwdev->max_qp - iwdev->used_qps;
 	props->max_qp_wr = (I40IW_MAX_WQ_ENTRIES >> 2) - 1;
 	props->max_sge = I40IW_MAX_WQ_FRAGMENT_COUNT;
-	props->max_cq = iwdev->max_cq;
+	props->max_cq = iwdev->max_cq - iwdev->used_cqs;
 	props->max_cqe = iwdev->max_cqe;
-	props->max_mr = iwdev->max_mr;
-	props->max_pd = iwdev->max_pd;
+	props->max_mr = iwdev->max_mr - iwdev->used_mrs;
+	props->max_pd = iwdev->max_pd - iwdev->used_pds;
 	props->max_sge_rd = I40IW_MAX_SGE_RD;
 	props->max_qp_rd_atom = I40IW_MAX_IRD_SIZE;
 	props->max_qp_init_rd_atom = props->max_qp_rd_atom;
-- 
cgit v1.2.1


From 799749979dbf41a878a00abdae00cf484c21a5b2 Mon Sep 17 00:00:00 2001
From: Henry Orosco <henry.orosco@intel.com>
Date: Wed, 9 Nov 2016 21:32:03 -0600
Subject: i40iw: Remove checks for more than 48 bytes inline data

Remove dead code, which isn't executed because we
return error if the data size is greater than 48 bytes.

Inline data size greater than 48 bytes isn't supported
and the maximum WQE size is 64 bytes.

Signed-off-by: Tatyana Nikolova <tatyana.e.nikolova@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_uk.c | 6 +-----
 1 file changed, 1 insertion(+), 5 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_uk.c b/drivers/infiniband/hw/i40iw/i40iw_uk.c
index 47cb2e063ee3..5d9c3bfb7310 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_uk.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_uk.c
@@ -1186,12 +1186,8 @@ enum i40iw_status_code i40iw_inline_data_size_to_wqesize(u32 data_size,
 
 	if (data_size <= 16)
 		*wqe_size = I40IW_QP_WQE_MIN_SIZE;
-	else if (data_size <= 48)
-		*wqe_size = 64;
-	else if (data_size <= 80)
-		*wqe_size = 96;
 	else
-		*wqe_size = 128;
+		*wqe_size = 64;
 
 	return 0;
 }
-- 
cgit v1.2.1


From 1ad19f739f494eda2f8e9611ab6c3056244b70fc Mon Sep 17 00:00:00 2001
From: Henry Orosco <henry.orosco@intel.com>
Date: Wed, 9 Nov 2016 21:32:25 -0600
Subject: i40iw: Remove NULL check for cm_node->iwdev

It is not necessary to check cm_node->iwdev in
i40iw_rem_ref_cm_node() as it can never be NULL after
a successful call out of i40iw_make_cm_node().

Signed-off-by: Chien Tin Tung <chien.tin.tung@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_cm.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.c b/drivers/infiniband/hw/i40iw/i40iw_cm.c
index 9e447b5951fc..cbd77eb30385 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_cm.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_cm.c
@@ -2234,7 +2234,7 @@ static void i40iw_rem_ref_cm_node(struct i40iw_cm_node *cm_node)
 		i40iw_dec_refcnt_listen(cm_core, cm_node->listener, 0, true);
 	} else {
 		if (!i40iw_listen_port_in_use(cm_core, cm_node->loc_port) &&
-		    cm_node->apbvt_set && cm_node->iwdev) {
+		    cm_node->apbvt_set) {
 			i40iw_manage_apbvt(cm_node->iwdev,
 					   cm_node->loc_port,
 					   I40IW_MANAGE_APBVT_DEL);
-- 
cgit v1.2.1


From e67791858e7b8d1389833386cb2dd0ca30d06862 Mon Sep 17 00:00:00 2001
From: Henry Orosco <henry.orosco@intel.com>
Date: Wed, 9 Nov 2016 21:33:32 -0600
Subject: i40iw: Use actual page size

In i40iw_post_send, use the actual page size instead of
encoded page size. This is to be consistent with the
rest of the file.

Signed-off-by: Mustafa Ismail <mustafa.ismail@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_verbs.c | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
index d20ee118f366..dcf08b8994ec 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
@@ -2145,7 +2145,6 @@ static int i40iw_post_send(struct ib_qp *ibqp,
 		case IB_WR_REG_MR:
 		{
 			struct i40iw_mr *iwmr = to_iwmr(reg_wr(ib_wr)->mr);
-			int page_shift = ilog2(reg_wr(ib_wr)->mr->page_size);
 			int flags = reg_wr(ib_wr)->access;
 			struct i40iw_pble_alloc *palloc = &iwmr->iwpbl.pble_alloc;
 			struct i40iw_sc_dev *dev = &iwqp->iwdev->sc_dev;
@@ -2156,6 +2155,7 @@ static int i40iw_post_send(struct ib_qp *ibqp,
 			info.access_rights |= i40iw_get_user_access(flags);
 			info.stag_key = reg_wr(ib_wr)->key & 0xff;
 			info.stag_idx = reg_wr(ib_wr)->key >> 8;
+			info.page_size = reg_wr(ib_wr)->mr->page_size;
 			info.wr_id = ib_wr->wr_id;
 
 			info.addr_type = I40IW_ADDR_TYPE_VA_BASED;
@@ -2169,9 +2169,6 @@ static int i40iw_post_send(struct ib_qp *ibqp,
 			if (iwmr->npages > I40IW_MIN_PAGES_PER_FMR)
 				info.chunk_size = 1;
 
-			if (page_shift == 21)
-				info.page_size = 1; /* 2M page */
-
 			ret = dev->iw_priv_qp_ops->iw_mr_fast_register(&iwqp->sc_qp, &info, true);
 			if (ret)
 				err = -ENOMEM;
-- 
cgit v1.2.1


From 5ebcb0ff54e594668e506583fa7344d101e3d05e Mon Sep 17 00:00:00 2001
From: Henry Orosco <henry.orosco@intel.com>
Date: Wed, 9 Nov 2016 21:34:02 -0600
Subject: i40iw: Use runtime check for IS_ENABLED(CONFIG_IPV6)

To be consistent, use the runtime check instead of
conditional compile.

Signed-off-by: Mustafa Ismail <mustafa.ismail@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_cm.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.c b/drivers/infiniband/hw/i40iw/i40iw_cm.c
index cbd77eb30385..b60e34653ec1 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_cm.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_cm.c
@@ -1583,9 +1583,10 @@ static enum i40iw_status_code i40iw_del_multiple_qhash(
 static struct net_device *i40iw_netdev_vlan_ipv6(u32 *addr, u16 *vlan_id, u8 *mac)
 {
 	struct net_device *ip_dev = NULL;
-#if IS_ENABLED(CONFIG_IPV6)
 	struct in6_addr laddr6;
 
+	if (!IS_ENABLED(CONFIG_IPV6))
+		return NULL;
 	i40iw_copy_ip_htonl(laddr6.in6_u.u6_addr32, addr);
 	if (vlan_id)
 		*vlan_id = I40IW_NO_VLAN;
@@ -1602,7 +1603,6 @@ static struct net_device *i40iw_netdev_vlan_ipv6(u32 *addr, u16 *vlan_id, u8 *ma
 		}
 	}
 	rcu_read_unlock();
-#endif
 	return ip_dev;
 }
 
-- 
cgit v1.2.1


From dfd9c43b3ce97e1b66a6dc1d9dcc95db9a27cc4b Mon Sep 17 00:00:00 2001
From: Henry Orosco <henry.orosco@intel.com>
Date: Wed, 9 Nov 2016 21:42:26 -0600
Subject: i40iw: Remove check on return from device_init_pestat()

Remove unnecessary check for return code from
device_init_pestat() and change func to void.

Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_ctrl.c     | 10 ++--------
 drivers/infiniband/hw/i40iw/i40iw_p.h        |  2 +-
 drivers/infiniband/hw/i40iw/i40iw_virtchnl.c |  6 +-----
 3 files changed, 4 insertions(+), 14 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
index 8417452a7a9c..5dde358349ee 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
@@ -4853,10 +4853,9 @@ static const struct i40iw_device_pestat_ops iw_device_pestat_ops = {
  * i40iw_device_init_pestat - Initialize the pestat structure
  * @dev: pestat struct
  */
-enum i40iw_status_code i40iw_device_init_pestat(struct i40iw_dev_pestat *devstat)
+void i40iw_device_init_pestat(struct i40iw_dev_pestat *devstat)
 {
 	devstat->ops = iw_device_pestat_ops;
-	return 0;
 }
 
 /**
@@ -4881,12 +4880,7 @@ enum i40iw_status_code i40iw_device_init(struct i40iw_sc_dev *dev,
 
 	dev->debug_mask = info->debug_mask;
 
-	ret_code = i40iw_device_init_pestat(&dev->dev_pestat);
-	if (ret_code) {
-		i40iw_debug(dev, I40IW_DEBUG_DEV,
-			    "%s: i40iw_device_init_pestat failed\n", __func__);
-		return ret_code;
-	}
+	i40iw_device_init_pestat(&dev->dev_pestat);
 	dev->hmc_fn_id = info->hmc_fn_id;
 	i40iw_fill_qos_list(info->l2params.qs_handle_list);
 	for (i = 0; i < I40IW_MAX_USER_PRIORITY; i++) {
diff --git a/drivers/infiniband/hw/i40iw/i40iw_p.h b/drivers/infiniband/hw/i40iw/i40iw_p.h
index c9e8cb8c4410..2a4bd3233b41 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_p.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_p.h
@@ -47,7 +47,7 @@ void i40iw_debug_buf(struct i40iw_sc_dev *dev, enum i40iw_debug_flag mask,
 enum i40iw_status_code i40iw_device_init(struct i40iw_sc_dev *dev,
 					 struct i40iw_device_init_info *info);
 
-enum i40iw_status_code i40iw_device_init_pestat(struct i40iw_dev_pestat *);
+void i40iw_device_init_pestat(struct i40iw_dev_pestat *devstat);
 
 void i40iw_sc_cqp_post_sq(struct i40iw_sc_cqp *cqp);
 
diff --git a/drivers/infiniband/hw/i40iw/i40iw_virtchnl.c b/drivers/infiniband/hw/i40iw/i40iw_virtchnl.c
index 3041003c94d2..dbd39c4af946 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_virtchnl.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_virtchnl.c
@@ -496,11 +496,7 @@ enum i40iw_status_code i40iw_vchnl_recv_pf(struct i40iw_sc_dev *dev,
 				i40iw_debug(dev, I40IW_DEBUG_VIRT,
 					    "VF%u error CQP HMC Function operation.\n",
 					    vf_id);
-			ret_code = i40iw_device_init_pestat(&vf_dev->dev_pestat);
-			if (ret_code)
-				i40iw_debug(dev, I40IW_DEBUG_VIRT,
-					    "VF%u - i40iw_device_init_pestat failed\n",
-					    vf_id);
+			i40iw_device_init_pestat(&vf_dev->dev_pestat);
 			vf_dev->dev_pestat.ops.iw_hw_stat_init(&vf_dev->dev_pestat,
 							      (u8)vf_dev->pmf_index,
 							      dev->hw, false);
-- 
cgit v1.2.1


From 78e945aace5b9aaf19404799cd29b4d155806053 Mon Sep 17 00:00:00 2001
From: Henry Orosco <henry.orosco@intel.com>
Date: Wed, 9 Nov 2016 22:20:31 -0600
Subject: i40iw: Remove variable flush_code and check to set qp->sq_flush

The flush_code variable in i40iw_bld_terminate_hdr() is obsolete and
the check to set qp->sq_flush is unreachable. Currently flush code is
populated in setup_term_hdr() and both SQ and RQ are flushed always
as part of the tear down flow.

Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_ctrl.c | 4 ----
 1 file changed, 4 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
index 5dde358349ee..a13503715f20 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
@@ -4185,7 +4185,6 @@ static int i40iw_bld_terminate_hdr(struct i40iw_sc_qp *qp,
 	u16 ddp_seg_len;
 	int copy_len = 0;
 	u8 is_tagged = 0;
-	enum i40iw_flush_opcode flush_code = FLUSH_INVALID;
 	u32 opcode;
 	struct i40iw_terminate_hdr *termhdr;
 
@@ -4358,9 +4357,6 @@ static int i40iw_bld_terminate_hdr(struct i40iw_sc_qp *qp,
 	if (copy_len)
 		memcpy(termhdr + 1, pkt, copy_len);
 
-	if (flush_code && !info->in_rdrsp_wr)
-		qp->sq_flush = (info->sq) ? true : false;
-
 	return sizeof(struct i40iw_terminate_hdr) + copy_len;
 }
 
-- 
cgit v1.2.1


From d4165e3abdf16707602c10f0d678d4d564a87e35 Mon Sep 17 00:00:00 2001
From: Henry Orosco <henry.orosco@intel.com>
Date: Tue, 22 Nov 2016 09:44:20 -0600
Subject: i40iw: Fix incorrect assignment of SQ head

The SQ head is incorrectly incremented when the number
of WQEs required is greater than the number available.
The fix is to use the I40IW_RING_MOV_HEAD_BY_COUNT
macro. This checks for the SQ full condition first and
only if SQ has room for the request, then we move the
head appropriately.

Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_uk.c | 10 ++++------
 1 file changed, 4 insertions(+), 6 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_uk.c b/drivers/infiniband/hw/i40iw/i40iw_uk.c
index 5d9c3bfb7310..4376cd628774 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_uk.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_uk.c
@@ -175,12 +175,10 @@ u64 *i40iw_qp_get_next_send_wqe(struct i40iw_qp_uk *qp,
 		if (!*wqe_idx)
 			qp->swqe_polarity = !qp->swqe_polarity;
 	}
-
-	for (i = 0; i < wqe_size / I40IW_QP_WQE_MIN_SIZE; i++) {
-		I40IW_RING_MOVE_HEAD(qp->sq_ring, ret_code);
-		if (ret_code)
-			return NULL;
-	}
+	I40IW_RING_MOVE_HEAD_BY_COUNT(qp->sq_ring,
+				      wqe_size / I40IW_QP_WQE_MIN_SIZE, ret_code);
+	if (ret_code)
+		return NULL;
 
 	wqe = qp->sq_base[*wqe_idx].elem;
 
-- 
cgit v1.2.1


From b6a529da69ce880ee4d0f3730ef46ead7f7cd0d3 Mon Sep 17 00:00:00 2001
From: Henry Orosco <henry.orosco@intel.com>
Date: Wed, 30 Nov 2016 14:56:14 -0600
Subject: i40iw: Utilize physically mapped memory regions

Add support to use physically mapped WQ's and MR's if determined
that the OS registered user-memory for the region is physically
contiguous. This feature will eliminate the need for unnecessarily
setting up and using PBL's when not required.

Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_verbs.c | 102 +++++++++++++++++++++++++++---
 1 file changed, 93 insertions(+), 9 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
index dcf08b8994ec..43bae5b14a71 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
@@ -1355,11 +1355,63 @@ static void i40iw_copy_user_pgaddrs(struct i40iw_mr *iwmr,
 	}
 }
 
+/**
+ * i40iw_check_mem_contiguous - check if pbls stored in arr are contiguous
+ * @arr: lvl1 pbl array
+ * @npages: page count
+ * pg_size: page size
+ *
+ */
+static bool i40iw_check_mem_contiguous(u64 *arr, u32 npages, u32 pg_size)
+{
+	u32 pg_idx;
+
+	for (pg_idx = 0; pg_idx < npages; pg_idx++) {
+		if ((*arr + (pg_size * pg_idx)) != arr[pg_idx])
+			return false;
+	}
+	return true;
+}
+
+/**
+ * i40iw_check_mr_contiguous - check if MR is physically contiguous
+ * @palloc: pbl allocation struct
+ * pg_size: page size
+ */
+static bool i40iw_check_mr_contiguous(struct i40iw_pble_alloc *palloc, u32 pg_size)
+{
+	struct i40iw_pble_level2 *lvl2 = &palloc->level2;
+	struct i40iw_pble_info *leaf = lvl2->leaf;
+	u64 *arr = NULL;
+	u64 *start_addr = NULL;
+	int i;
+	bool ret;
+
+	if (palloc->level == I40IW_LEVEL_1) {
+		arr = (u64 *)palloc->level1.addr;
+		ret = i40iw_check_mem_contiguous(arr, palloc->total_cnt, pg_size);
+		return ret;
+	}
+
+	start_addr = (u64 *)leaf->addr;
+
+	for (i = 0; i < lvl2->leaf_cnt; i++, leaf++) {
+		arr = (u64 *)leaf->addr;
+		if ((*start_addr + (i * pg_size * PBLE_PER_PAGE)) != *arr)
+			return false;
+		ret = i40iw_check_mem_contiguous(arr, leaf->cnt, pg_size);
+		if (!ret)
+			return false;
+	}
+
+	return true;
+}
+
 /**
  * i40iw_setup_pbles - copy user pg address to pble's
  * @iwdev: iwarp device
  * @iwmr: mr pointer for this memory registration
- * @use_pbles: flag if to use pble's or memory (level 0)
+ * @use_pbles: flag if to use pble's
  */
 static int i40iw_setup_pbles(struct i40iw_device *iwdev,
 			     struct i40iw_mr *iwmr,
@@ -1372,9 +1424,6 @@ static int i40iw_setup_pbles(struct i40iw_device *iwdev,
 	enum i40iw_status_code status;
 	enum i40iw_pble_level level = I40IW_LEVEL_1;
 
-	if (!use_pbles && (iwmr->page_cnt > MAX_SAVE_PAGE_ADDRS))
-		return -ENOMEM;
-
 	if (use_pbles) {
 		mutex_lock(&iwdev->pbl_mutex);
 		status = i40iw_get_pble(&iwdev->sc_dev, iwdev->pble_rsrc, palloc, iwmr->page_cnt);
@@ -1391,6 +1440,10 @@ static int i40iw_setup_pbles(struct i40iw_device *iwdev,
 	}
 
 	i40iw_copy_user_pgaddrs(iwmr, pbl, level);
+
+	if (use_pbles)
+		iwmr->pgaddrmem[0] = *pbl;
+
 	return 0;
 }
 
@@ -1412,14 +1465,18 @@ static int i40iw_handle_q_mem(struct i40iw_device *iwdev,
 	struct i40iw_cq_mr *cqmr = &iwpbl->cq_mr;
 	struct i40iw_hmc_pble *hmc_p;
 	u64 *arr = iwmr->pgaddrmem;
+	u32 pg_size;
 	int err;
 	int total;
+	bool ret = true;
 
 	total = req->sq_pages + req->rq_pages + req->cq_pages;
+	pg_size = iwmr->region->page_size;
 
 	err = i40iw_setup_pbles(iwdev, iwmr, use_pbles);
 	if (err)
 		return err;
+
 	if (use_pbles && (palloc->level != I40IW_LEVEL_1)) {
 		i40iw_free_pble(iwdev->pble_rsrc, palloc);
 		iwpbl->pbl_allocated = false;
@@ -1428,26 +1485,44 @@ static int i40iw_handle_q_mem(struct i40iw_device *iwdev,
 
 	if (use_pbles)
 		arr = (u64 *)palloc->level1.addr;
-	if (req->reg_type == IW_MEMREG_TYPE_QP) {
+
+	if (iwmr->type == IW_MEMREG_TYPE_QP) {
 		hmc_p = &qpmr->sq_pbl;
 		qpmr->shadow = (dma_addr_t)arr[total];
+
 		if (use_pbles) {
+			ret = i40iw_check_mem_contiguous(arr, req->sq_pages, pg_size);
+			if (ret)
+				ret = i40iw_check_mem_contiguous(&arr[req->sq_pages], req->rq_pages, pg_size);
+		}
+
+		if (!ret) {
 			hmc_p->idx = palloc->level1.idx;
 			hmc_p = &qpmr->rq_pbl;
 			hmc_p->idx = palloc->level1.idx + req->sq_pages;
 		} else {
 			hmc_p->addr = arr[0];
 			hmc_p = &qpmr->rq_pbl;
-			hmc_p->addr = arr[1];
+			hmc_p->addr = arr[req->sq_pages];
 		}
 	} else {		/* CQ */
 		hmc_p = &cqmr->cq_pbl;
 		cqmr->shadow = (dma_addr_t)arr[total];
+
 		if (use_pbles)
+			ret = i40iw_check_mem_contiguous(arr, req->cq_pages, pg_size);
+
+		if (!ret)
 			hmc_p->idx = palloc->level1.idx;
 		else
 			hmc_p->addr = arr[0];
 	}
+
+	if (use_pbles && ret) {
+		i40iw_free_pble(iwdev->pble_rsrc, palloc);
+		iwpbl->pbl_allocated = false;
+	}
+
 	return err;
 }
 
@@ -1646,7 +1721,7 @@ static int i40iw_hwreg_mr(struct i40iw_device *iwdev,
 	stag_info->pd_id = iwpd->sc_pd.pd_id;
 	stag_info->addr_type = I40IW_ADDR_TYPE_VA_BASED;
 
-	if (iwmr->page_cnt > 1) {
+	if (iwpbl->pbl_allocated) {
 		if (palloc->level == I40IW_LEVEL_1) {
 			stag_info->first_pm_pbl_index = palloc->level1.idx;
 			stag_info->chunk_size = 1;
@@ -1702,6 +1777,7 @@ static struct ib_mr *i40iw_reg_user_mr(struct ib_pd *pd,
 	bool use_pbles = false;
 	unsigned long flags;
 	int err = -ENOSYS;
+	int ret;
 
 	if (length > I40IW_MAX_MR_SIZE)
 		return ERR_PTR(-EINVAL);
@@ -1758,13 +1834,21 @@ static struct ib_mr *i40iw_reg_user_mr(struct ib_pd *pd,
 		spin_unlock_irqrestore(&ucontext->cq_reg_mem_list_lock, flags);
 		break;
 	case IW_MEMREG_TYPE_MEM:
+		use_pbles = (iwmr->page_cnt != 1);
 		access = I40IW_ACCESS_FLAGS_LOCALREAD;
 
-		use_pbles = (iwmr->page_cnt != 1);
 		err = i40iw_setup_pbles(iwdev, iwmr, use_pbles);
 		if (err)
 			goto error;
 
+		if (use_pbles) {
+			ret = i40iw_check_mr_contiguous(palloc, region->page_size);
+			if (ret) {
+				i40iw_free_pble(iwdev->pble_rsrc, palloc);
+				iwpbl->pbl_allocated = false;
+			}
+		}
+
 		access |= i40iw_get_user_access(acc);
 		stag = i40iw_create_stag(iwdev);
 		if (!stag) {
@@ -1792,7 +1876,7 @@ static struct ib_mr *i40iw_reg_user_mr(struct ib_pd *pd,
 	return &iwmr->ibmr;
 
 error:
-	if (palloc->level != I40IW_LEVEL_0)
+	if (palloc->level != I40IW_LEVEL_0 && iwpbl->pbl_allocated)
 		i40iw_free_pble(iwdev->pble_rsrc, palloc);
 	ib_umem_release(region);
 	kfree(iwmr);
-- 
cgit v1.2.1


From f26c7c83395b72f30d111f4e3adb3437c0a30b77 Mon Sep 17 00:00:00 2001
From: Henry Orosco <henry.orosco@intel.com>
Date: Wed, 30 Nov 2016 14:57:40 -0600
Subject: i40iw: Add 2MB page support

Add support to allow each independent memory region to
be configured for 2MB page size in addition to 4KB
page size.

Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_verbs.c | 59 +++++++++++++++++++++++++------
 drivers/infiniband/hw/i40iw/i40iw_verbs.h |  2 ++
 2 files changed, 51 insertions(+), 10 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
index 43bae5b14a71..1c2f0a19bd63 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
@@ -37,6 +37,7 @@
 #include <linux/random.h>
 #include <linux/highmem.h>
 #include <linux/time.h>
+#include <linux/hugetlb.h>
 #include <asm/byteorder.h>
 #include <net/ip.h>
 #include <rdma/ib_verbs.h>
@@ -1305,13 +1306,11 @@ static u32 i40iw_create_stag(struct i40iw_device *iwdev)
 
 /**
  * i40iw_next_pbl_addr - Get next pbl address
- * @palloc: Poiner to allocated pbles
  * @pbl: pointer to a pble
  * @pinfo: info pointer
  * @idx: index
  */
-static inline u64 *i40iw_next_pbl_addr(struct i40iw_pble_alloc *palloc,
-				       u64 *pbl,
+static inline u64 *i40iw_next_pbl_addr(u64 *pbl,
 				       struct i40iw_pble_info **pinfo,
 				       u32 *idx)
 {
@@ -1339,9 +1338,11 @@ static void i40iw_copy_user_pgaddrs(struct i40iw_mr *iwmr,
 	struct i40iw_pble_alloc *palloc = &iwpbl->pble_alloc;
 	struct i40iw_pble_info *pinfo;
 	struct scatterlist *sg;
+	u64 pg_addr = 0;
 	u32 idx = 0;
 
 	pinfo = (level == I40IW_LEVEL_1) ? NULL : palloc->level2.leaf;
+
 	pg_shift = ffs(region->page_size) - 1;
 	for_each_sg(region->sg_head.sgl, sg, region->nmap, entry) {
 		chunk_pages = sg_dma_len(sg) >> pg_shift;
@@ -1349,8 +1350,35 @@ static void i40iw_copy_user_pgaddrs(struct i40iw_mr *iwmr,
 		    !iwpbl->qp_mr.sq_page)
 			iwpbl->qp_mr.sq_page = sg_page(sg);
 		for (i = 0; i < chunk_pages; i++) {
-			*pbl = cpu_to_le64(sg_dma_address(sg) + region->page_size * i);
-			pbl = i40iw_next_pbl_addr(palloc, pbl, &pinfo, &idx);
+			pg_addr = sg_dma_address(sg) + region->page_size * i;
+
+			if ((entry + i) == 0)
+				*pbl = cpu_to_le64(pg_addr & iwmr->page_msk);
+			else if (!(pg_addr & ~iwmr->page_msk))
+				*pbl = cpu_to_le64(pg_addr);
+			else
+				continue;
+			pbl = i40iw_next_pbl_addr(pbl, &pinfo, &idx);
+		}
+	}
+}
+
+/**
+ * i40iw_set_hugetlb_params - set MR pg size and mask to huge pg values.
+ * @addr: virtual address
+ * @iwmr: mr pointer for this memory registration
+ */
+static void i40iw_set_hugetlb_values(u64 addr, struct i40iw_mr *iwmr)
+{
+	struct vm_area_struct *vma;
+	struct hstate *h;
+
+	vma = find_vma(current->mm, addr);
+	if (vma && is_vm_hugetlb_page(vma)) {
+		h = hstate_vma(vma);
+		if (huge_page_size(h) == 0x200000) {
+			iwmr->page_size = huge_page_size(h);
+			iwmr->page_msk = huge_page_mask(h);
 		}
 	}
 }
@@ -1471,7 +1499,7 @@ static int i40iw_handle_q_mem(struct i40iw_device *iwdev,
 	bool ret = true;
 
 	total = req->sq_pages + req->rq_pages + req->cq_pages;
-	pg_size = iwmr->region->page_size;
+	pg_size = iwmr->page_size;
 
 	err = i40iw_setup_pbles(iwdev, iwmr, use_pbles);
 	if (err)
@@ -1720,6 +1748,7 @@ static int i40iw_hwreg_mr(struct i40iw_device *iwdev,
 	stag_info->access_rights = access;
 	stag_info->pd_id = iwpd->sc_pd.pd_id;
 	stag_info->addr_type = I40IW_ADDR_TYPE_VA_BASED;
+	stag_info->page_size = iwmr->page_size;
 
 	if (iwpbl->pbl_allocated) {
 		if (palloc->level == I40IW_LEVEL_1) {
@@ -1778,6 +1807,7 @@ static struct ib_mr *i40iw_reg_user_mr(struct ib_pd *pd,
 	unsigned long flags;
 	int err = -ENOSYS;
 	int ret;
+	int pg_shift;
 
 	if (length > I40IW_MAX_MR_SIZE)
 		return ERR_PTR(-EINVAL);
@@ -1802,9 +1832,17 @@ static struct ib_mr *i40iw_reg_user_mr(struct ib_pd *pd,
 	iwmr->ibmr.pd = pd;
 	iwmr->ibmr.device = pd->device;
 	ucontext = to_ucontext(pd->uobject->context);
-	region_length = region->length + (start & 0xfff);
-	pbl_depth = region_length >> 12;
-	pbl_depth += (region_length & (4096 - 1)) ? 1 : 0;
+
+	iwmr->page_size = region->page_size;
+	iwmr->page_msk = PAGE_MASK;
+
+	if (region->hugetlb && (req.reg_type == IW_MEMREG_TYPE_MEM))
+		i40iw_set_hugetlb_values(start, iwmr);
+
+	region_length = region->length + (start & (iwmr->page_size - 1));
+	pg_shift = ffs(iwmr->page_size) - 1;
+	pbl_depth = region_length >> pg_shift;
+	pbl_depth += (region_length & (iwmr->page_size - 1)) ? 1 : 0;
 	iwmr->length = region->length;
 
 	iwpbl->user_base = virt;
@@ -1842,7 +1880,7 @@ static struct ib_mr *i40iw_reg_user_mr(struct ib_pd *pd,
 			goto error;
 
 		if (use_pbles) {
-			ret = i40iw_check_mr_contiguous(palloc, region->page_size);
+			ret = i40iw_check_mr_contiguous(palloc, iwmr->page_size);
 			if (ret) {
 				i40iw_free_pble(iwdev->pble_rsrc, palloc);
 				iwpbl->pbl_allocated = false;
@@ -1865,6 +1903,7 @@ static struct ib_mr *i40iw_reg_user_mr(struct ib_pd *pd,
 			i40iw_free_stag(iwdev, stag);
 			goto error;
 		}
+
 		break;
 	default:
 		goto error;
diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.h b/drivers/infiniband/hw/i40iw/i40iw_verbs.h
index 0069be8a5a38..6549c939500f 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_verbs.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.h
@@ -92,6 +92,8 @@ struct i40iw_mr {
 	struct ib_umem *region;
 	u16 type;
 	u32 page_cnt;
+	u32 page_size;
+	u64 page_msk;
 	u32 npages;
 	u32 stag;
 	u64 length;
-- 
cgit v1.2.1


From d59659340c61e777208524f77c268fe6edc6fe37 Mon Sep 17 00:00:00 2001
From: Mustafa Ismail <mustafa.ismail@intel.com>
Date: Wed, 30 Nov 2016 14:59:26 -0600
Subject: i40iw: Add missing cleanup on device close

On i40iw device close, disconnect all connected QPs by moving
them to error state; and block further QPs, PDs and CQs from
being created. Additionally, make sure all resources have been
freed before deallocating the ibdev as part of the device close.

Signed-off-by: Mustafa Ismail <mustafa.ismail@intel.com>
Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw.h       |  5 +++++
 drivers/infiniband/hw/i40iw/i40iw_cm.c    | 31 +++++++++++++++++++++++++++++++
 drivers/infiniband/hw/i40iw/i40iw_cm.h    |  2 ++
 drivers/infiniband/hw/i40iw/i40iw_d.h     |  2 ++
 drivers/infiniband/hw/i40iw/i40iw_main.c  |  4 ++++
 drivers/infiniband/hw/i40iw/i40iw_utils.c | 21 +++++++++++++++++++++
 drivers/infiniband/hw/i40iw/i40iw_verbs.c | 25 +++++++++++++++++++++++++
 7 files changed, 90 insertions(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw.h b/drivers/infiniband/hw/i40iw/i40iw.h
index dac9a6bcc631..c795c6160261 100644
--- a/drivers/infiniband/hw/i40iw/i40iw.h
+++ b/drivers/infiniband/hw/i40iw/i40iw.h
@@ -303,10 +303,13 @@ struct i40iw_device {
 	u32 mr_stagmask;
 	u32 mpa_version;
 	bool dcb;
+	bool closing;
 	u32 used_pds;
 	u32 used_cqs;
 	u32 used_mrs;
 	u32 used_qps;
+	wait_queue_head_t close_wq;
+	atomic64_t use_count;
 };
 
 struct i40iw_ib_device {
@@ -521,6 +524,8 @@ int i40iw_modify_qp(struct ib_qp *, struct ib_qp_attr *, int, struct ib_udata *)
 
 void i40iw_rem_pdusecount(struct i40iw_pd *iwpd, struct i40iw_device *iwdev);
 void i40iw_add_pdusecount(struct i40iw_pd *iwpd);
+void i40iw_rem_devusecount(struct i40iw_device *iwdev);
+void i40iw_add_devusecount(struct i40iw_device *iwdev);
 void i40iw_hw_modify_qp(struct i40iw_device *iwdev, struct i40iw_qp *iwqp,
 			struct i40iw_modify_qp_info *info, bool wait);
 
diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.c b/drivers/infiniband/hw/i40iw/i40iw_cm.c
index b60e34653ec1..11ef0b09c843 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_cm.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_cm.c
@@ -4128,3 +4128,34 @@ static void i40iw_cm_post_event(struct i40iw_cm_event *event)
 
 	queue_work(event->cm_node->cm_core->event_wq, &event->event_work);
 }
+
+/**
+ * i40iw_cm_disconnect_all - disconnect all connected qp's
+ * @iwdev: device pointer
+ */
+void i40iw_cm_disconnect_all(struct i40iw_device *iwdev)
+{
+	struct i40iw_cm_core *cm_core = &iwdev->cm_core;
+	struct list_head *list_core_temp;
+	struct list_head *list_node;
+	struct i40iw_cm_node *cm_node;
+	unsigned long flags;
+	struct list_head connected_list;
+	struct ib_qp_attr attr;
+
+	INIT_LIST_HEAD(&connected_list);
+	spin_lock_irqsave(&cm_core->ht_lock, flags);
+	list_for_each_safe(list_node, list_core_temp, &cm_core->connected_nodes) {
+		cm_node = container_of(list_node, struct i40iw_cm_node, list);
+		atomic_inc(&cm_node->ref_count);
+		list_add(&cm_node->connected_entry, &connected_list);
+	}
+	spin_unlock_irqrestore(&cm_core->ht_lock, flags);
+
+	list_for_each_safe(list_node, list_core_temp, &connected_list) {
+		cm_node = container_of(list_node, struct i40iw_cm_node, connected_entry);
+		attr.qp_state = IB_QPS_ERR;
+		i40iw_modify_qp(&cm_node->iwqp->ibqp, &attr, IB_QP_STATE, NULL);
+		i40iw_rem_ref_cm_node(cm_node);
+	}
+}
diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.h b/drivers/infiniband/hw/i40iw/i40iw_cm.h
index 24615c24cb04..0381b7f5e20d 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_cm.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_cm.h
@@ -339,6 +339,7 @@ struct i40iw_cm_node {
 	int accept_pend;
 	struct list_head timer_entry;
 	struct list_head reset_entry;
+	struct list_head connected_entry;
 	atomic_t passive_state;
 	bool qhash_set;
 	u8 user_pri;
@@ -443,4 +444,5 @@ int i40iw_arp_table(struct i40iw_device *iwdev,
 		    u8 *mac_addr,
 		    u32 action);
 
+void i40iw_cm_disconnect_all(struct i40iw_device *iwdev);
 #endif /* I40IW_CM_H */
diff --git a/drivers/infiniband/hw/i40iw/i40iw_d.h b/drivers/infiniband/hw/i40iw/i40iw_d.h
index e184c0e99cb1..1bd4badb26d3 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_d.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_d.h
@@ -35,6 +35,8 @@
 #ifndef I40IW_D_H
 #define I40IW_D_H
 
+#define I40IW_FIRST_USER_QP_ID  2
+
 #define I40IW_DB_ADDR_OFFSET    (4 * 1024 * 1024 - 64 * 1024)
 #define I40IW_VF_DB_ADDR_OFFSET (64 * 1024)
 
diff --git a/drivers/infiniband/hw/i40iw/i40iw_main.c b/drivers/infiniband/hw/i40iw/i40iw_main.c
index 9d3b9ee20ba7..d86bb6e98f07 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_main.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_main.c
@@ -1546,6 +1546,7 @@ static enum i40iw_status_code i40iw_setup_init_state(struct i40iw_handler *hdl,
 
 	init_waitqueue_head(&iwdev->vchnl_waitq);
 	init_waitqueue_head(&dev->vf_reqs);
+	init_waitqueue_head(&iwdev->close_wq);
 
 	status = i40iw_initialize_dev(iwdev, ldev);
 exit:
@@ -1748,6 +1749,9 @@ static void i40iw_close(struct i40e_info *ldev, struct i40e_client *client, bool
 		return;
 
 	iwdev = &hdl->device;
+	iwdev->closing = true;
+
+	i40iw_cm_disconnect_all(iwdev);
 	destroy_workqueue(iwdev->virtchnl_wq);
 	i40iw_deinit_device(iwdev, reset);
 }
diff --git a/drivers/infiniband/hw/i40iw/i40iw_utils.c b/drivers/infiniband/hw/i40iw/i40iw_utils.c
index 4e880e8689eb..58151280828d 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_utils.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_utils.c
@@ -392,6 +392,7 @@ static void i40iw_free_qp(struct i40iw_cqp_request *cqp_request, u32 num)
 
 	i40iw_rem_pdusecount(iwqp->iwpd, iwdev);
 	i40iw_free_qp_resources(iwdev, iwqp, qp_num);
+	i40iw_rem_devusecount(iwdev);
 }
 
 /**
@@ -458,6 +459,26 @@ enum i40iw_status_code i40iw_handle_cqp_op(struct i40iw_device *iwdev,
 	return status;
 }
 
+/**
+ * i40iw_add_devusecount - add dev refcount
+ * @iwdev: dev for refcount
+ */
+void i40iw_add_devusecount(struct i40iw_device *iwdev)
+{
+	atomic64_inc(&iwdev->use_count);
+}
+
+/**
+ * i40iw_rem_devusecount - decrement refcount for dev
+ * @iwdev: device
+ */
+void i40iw_rem_devusecount(struct i40iw_device *iwdev)
+{
+	if (!atomic64_dec_and_test(&iwdev->use_count))
+		return;
+	wake_up(&iwdev->close_wq);
+}
+
 /**
  * i40iw_add_pdusecount - add pd refcount
  * @iwpd: pd for refcount
diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
index 1c2f0a19bd63..bc24086989e3 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
@@ -336,6 +336,9 @@ static struct ib_pd *i40iw_alloc_pd(struct ib_device *ibdev,
 	u32 pd_id = 0;
 	int err;
 
+	if (iwdev->closing)
+		return ERR_PTR(-ENODEV);
+
 	err = i40iw_alloc_resource(iwdev, iwdev->allocated_pds,
 				   iwdev->max_pd, &pd_id, &iwdev->next_pd);
 	if (err) {
@@ -601,6 +604,9 @@ static struct ib_qp *i40iw_create_qp(struct ib_pd *ibpd,
 	struct i40iwarp_offload_info *iwarp_info;
 	unsigned long flags;
 
+	if (iwdev->closing)
+		return ERR_PTR(-ENODEV);
+
 	if (init_attr->create_flags)
 		return ERR_PTR(-EINVAL);
 	if (init_attr->cap.max_inline_data > I40IW_MAX_INLINE_DATA_SIZE)
@@ -776,6 +782,7 @@ static struct ib_qp *i40iw_create_qp(struct ib_pd *ibpd,
 	iwqp->sig_all = (init_attr->sq_sig_type == IB_SIGNAL_ALL_WR) ? 1 : 0;
 	iwdev->qp_table[qp_num] = iwqp;
 	i40iw_add_pdusecount(iwqp->iwpd);
+	i40iw_add_devusecount(iwdev);
 	if (ibpd->uobject && udata) {
 		memset(&uresp, 0, sizeof(uresp));
 		uresp.actual_sq_size = sq_size;
@@ -887,6 +894,11 @@ int i40iw_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
 	spin_lock_irqsave(&iwqp->lock, flags);
 
 	if (attr_mask & IB_QP_STATE) {
+		if (iwdev->closing && attr->qp_state != IB_QPS_ERR) {
+			err = -EINVAL;
+			goto exit;
+		}
+
 		switch (attr->qp_state) {
 		case IB_QPS_INIT:
 		case IB_QPS_RTR:
@@ -1086,6 +1098,7 @@ static int i40iw_destroy_cq(struct ib_cq *ib_cq)
 	cq_wq_destroy(iwdev, cq);
 	cq_free_resources(iwdev, iwcq);
 	kfree(iwcq);
+	i40iw_rem_devusecount(iwdev);
 	return 0;
 }
 
@@ -1116,6 +1129,9 @@ static struct ib_cq *i40iw_create_cq(struct ib_device *ibdev,
 	int err_code;
 	int entries = attr->cqe;
 
+	if (iwdev->closing)
+		return ERR_PTR(-ENODEV);
+
 	if (entries > iwdev->max_cqe)
 		return ERR_PTR(-EINVAL);
 
@@ -1233,6 +1249,7 @@ static struct ib_cq *i40iw_create_cq(struct ib_device *ibdev,
 		}
 	}
 
+	i40iw_add_devusecount(iwdev);
 	return (struct ib_cq *)iwcq;
 
 cq_destroy:
@@ -1270,6 +1287,7 @@ static void i40iw_free_stag(struct i40iw_device *iwdev, u32 stag)
 
 	stag_idx = (stag & iwdev->mr_stagmask) >> I40IW_CQPSQ_STAG_IDX_SHIFT;
 	i40iw_free_resource(iwdev, iwdev->allocated_mrs, stag_idx);
+	i40iw_rem_devusecount(iwdev);
 }
 
 /**
@@ -1300,6 +1318,7 @@ static u32 i40iw_create_stag(struct i40iw_device *iwdev)
 		stag = stag_index << I40IW_CQPSQ_STAG_IDX_SHIFT;
 		stag |= driver_key;
 		stag += (u32)consumer_key;
+		i40iw_add_devusecount(iwdev);
 	}
 	return stag;
 }
@@ -1809,6 +1828,9 @@ static struct ib_mr *i40iw_reg_user_mr(struct ib_pd *pd,
 	int ret;
 	int pg_shift;
 
+	if (iwdev->closing)
+		return ERR_PTR(-ENODEV);
+
 	if (length > I40IW_MAX_MR_SIZE)
 		return ERR_PTR(-EINVAL);
 	region = ib_umem_get(pd->uobject->context, start, length, acc, 0);
@@ -2842,6 +2864,9 @@ void i40iw_destroy_rdma_device(struct i40iw_ib_device *iwibdev)
 	i40iw_unregister_rdma_device(iwibdev);
 	kfree(iwibdev->ibdev.iwcm);
 	iwibdev->ibdev.iwcm = NULL;
+	wait_event_timeout(iwibdev->iwdev->close_wq,
+			   !atomic64_read(&iwibdev->iwdev->use_count),
+			   I40IW_EVENT_TIMEOUT);
 	ib_dealloc_device(&iwibdev->ibdev);
 }
 
-- 
cgit v1.2.1


From e5e74b61b16503acbd914f673b783fa2a1532a64 Mon Sep 17 00:00:00 2001
From: Mustafa Ismail <mustafa.ismail@intel.com>
Date: Wed, 30 Nov 2016 15:07:30 -0600
Subject: i40iw: Add IP addr handling on netdev events

Disable listeners and disconnect all connected QPs on
a netdev interface down event. On an interface up event,
the listeners are re-enabled.

Signed-off-by: Mustafa Ismail <mustafa.ismail@intel.com>
Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_cm.c    | 138 +++++++++++++++++++++++++++++-
 drivers/infiniband/hw/i40iw/i40iw_cm.h    |   2 +
 drivers/infiniband/hw/i40iw/i40iw_utils.c |  58 ++++---------
 3 files changed, 156 insertions(+), 42 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.c b/drivers/infiniband/hw/i40iw/i40iw_cm.c
index 11ef0b09c843..93ae764fc44e 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_cm.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_cm.c
@@ -1556,9 +1556,15 @@ static enum i40iw_status_code i40iw_del_multiple_qhash(
 		memcpy(cm_info->loc_addr, child_listen_node->loc_addr,
 		       sizeof(cm_info->loc_addr));
 		cm_info->vlan_id = child_listen_node->vlan_id;
-		ret = i40iw_manage_qhash(iwdev, cm_info,
-					 I40IW_QHASH_TYPE_TCP_SYN,
-					 I40IW_QHASH_MANAGE_TYPE_DELETE, NULL, false);
+		if (child_listen_node->qhash_set) {
+			ret = i40iw_manage_qhash(iwdev, cm_info,
+						 I40IW_QHASH_TYPE_TCP_SYN,
+						 I40IW_QHASH_MANAGE_TYPE_DELETE,
+						 NULL, false);
+			child_listen_node->qhash_set = false;
+		} else {
+			ret = I40IW_SUCCESS;
+		}
 		i40iw_debug(&iwdev->sc_dev,
 			    I40IW_DEBUG_CM,
 			    "freed pointer = %p\n",
@@ -1687,6 +1693,7 @@ static enum i40iw_status_code i40iw_add_mqh_6(struct i40iw_device *iwdev,
 							 I40IW_QHASH_MANAGE_TYPE_ADD,
 							 NULL, true);
 				if (!ret) {
+					child_listen_node->qhash_set = true;
 					spin_lock_irqsave(&iwdev->cm_core.listen_list_lock, flags);
 					list_add(&child_listen_node->child_listen_list,
 						 &cm_parent_listen_node->child_listen_list);
@@ -1765,6 +1772,7 @@ static enum i40iw_status_code i40iw_add_mqh_4(
 							 NULL,
 							 true);
 				if (!ret) {
+					child_listen_node->qhash_set = true;
 					spin_lock_irqsave(&iwdev->cm_core.listen_list_lock, flags);
 					list_add(&child_listen_node->child_listen_list,
 						 &cm_parent_listen_node->child_listen_list);
@@ -4129,6 +4137,73 @@ static void i40iw_cm_post_event(struct i40iw_cm_event *event)
 	queue_work(event->cm_node->cm_core->event_wq, &event->event_work);
 }
 
+/**
+ * i40iw_qhash_ctrl - enable/disable qhash for list
+ * @iwdev: device pointer
+ * @parent_listen_node: parent listen node
+ * @nfo: cm info node
+ * @ipaddr: Pointer to IPv4 or IPv6 address
+ * @ipv4: flag indicating IPv4 when true
+ * @ifup: flag indicating interface up when true
+ *
+ * Enables or disables the qhash for the node in the child
+ * listen list that matches ipaddr. If no matching IP was found
+ * it will allocate and add a new child listen node to the
+ * parent listen node. The listen_list_lock is assumed to be
+ * held when called.
+ */
+static void i40iw_qhash_ctrl(struct i40iw_device *iwdev,
+			     struct i40iw_cm_listener *parent_listen_node,
+			     struct i40iw_cm_info *nfo,
+			     u32 *ipaddr, bool ipv4, bool ifup)
+{
+	struct list_head *child_listen_list = &parent_listen_node->child_listen_list;
+	struct i40iw_cm_listener *child_listen_node;
+	struct list_head *pos, *tpos;
+	enum i40iw_status_code ret;
+	bool node_allocated = false;
+	enum i40iw_quad_hash_manage_type op =
+		ifup ? I40IW_QHASH_MANAGE_TYPE_ADD : I40IW_QHASH_MANAGE_TYPE_DELETE;
+
+	list_for_each_safe(pos, tpos, child_listen_list) {
+		child_listen_node =
+			list_entry(pos,
+				   struct i40iw_cm_listener,
+				   child_listen_list);
+		if (!memcmp(child_listen_node->loc_addr, ipaddr, ipv4 ? 4 : 16))
+			goto set_qhash;
+	}
+
+	/* if not found then add a child listener if interface is going up */
+	if (!ifup)
+		return;
+	child_listen_node = kzalloc(sizeof(*child_listen_node), GFP_ATOMIC);
+	if (!child_listen_node)
+		return;
+	node_allocated = true;
+	memcpy(child_listen_node, parent_listen_node, sizeof(*child_listen_node));
+
+	memcpy(child_listen_node->loc_addr, ipaddr,  ipv4 ? 4 : 16);
+
+set_qhash:
+	memcpy(nfo->loc_addr,
+	       child_listen_node->loc_addr,
+	       sizeof(nfo->loc_addr));
+	nfo->vlan_id = child_listen_node->vlan_id;
+	ret = i40iw_manage_qhash(iwdev, nfo,
+				 I40IW_QHASH_TYPE_TCP_SYN,
+				 op,
+				 NULL, false);
+	if (!ret) {
+		child_listen_node->qhash_set = ifup;
+		if (node_allocated)
+			list_add(&child_listen_node->child_listen_list,
+				 &parent_listen_node->child_listen_list);
+	} else if (node_allocated) {
+		kfree(child_listen_node);
+	}
+}
+
 /**
  * i40iw_cm_disconnect_all - disconnect all connected qp's
  * @iwdev: device pointer
@@ -4159,3 +4234,60 @@ void i40iw_cm_disconnect_all(struct i40iw_device *iwdev)
 		i40iw_rem_ref_cm_node(cm_node);
 	}
 }
+
+/**
+ * i40iw_ifdown_notify - process an ifdown on an interface
+ * @iwdev: device pointer
+ * @ipaddr: Pointer to IPv4 or IPv6 address
+ * @ipv4: flag indicating IPv4 when true
+ * @ifup: flag indicating interface up when true
+ */
+void i40iw_if_notify(struct i40iw_device *iwdev, struct net_device *netdev,
+		     u32 *ipaddr, bool ipv4, bool ifup)
+{
+	struct i40iw_cm_core *cm_core = &iwdev->cm_core;
+	unsigned long flags;
+	struct i40iw_cm_listener *listen_node;
+	static const u32 ip_zero[4] = { 0, 0, 0, 0 };
+	struct i40iw_cm_info nfo;
+	u16 vlan_id = rdma_vlan_dev_vlan_id(netdev);
+	enum i40iw_status_code ret;
+	enum i40iw_quad_hash_manage_type op =
+		ifup ? I40IW_QHASH_MANAGE_TYPE_ADD : I40IW_QHASH_MANAGE_TYPE_DELETE;
+
+	/* Disable or enable qhash for listeners */
+	spin_lock_irqsave(&cm_core->listen_list_lock, flags);
+	list_for_each_entry(listen_node, &cm_core->listen_nodes, list) {
+		if (vlan_id == listen_node->vlan_id &&
+		    (!memcmp(listen_node->loc_addr, ipaddr, ipv4 ? 4 : 16) ||
+		    !memcmp(listen_node->loc_addr, ip_zero, ipv4 ? 4 : 16))) {
+			memcpy(nfo.loc_addr, listen_node->loc_addr,
+			       sizeof(nfo.loc_addr));
+			nfo.loc_port = listen_node->loc_port;
+			nfo.ipv4 = listen_node->ipv4;
+			nfo.vlan_id = listen_node->vlan_id;
+			nfo.user_pri = listen_node->user_pri;
+			if (!list_empty(&listen_node->child_listen_list)) {
+				i40iw_qhash_ctrl(iwdev,
+						 listen_node,
+						 &nfo,
+						 ipaddr, ipv4, ifup);
+			} else if (memcmp(listen_node->loc_addr, ip_zero,
+					  ipv4 ? 4 : 16)) {
+				ret = i40iw_manage_qhash(iwdev,
+							 &nfo,
+							 I40IW_QHASH_TYPE_TCP_SYN,
+							 op,
+							 NULL,
+							 false);
+				if (!ret)
+					listen_node->qhash_set = ifup;
+			}
+		}
+	}
+	spin_unlock_irqrestore(&cm_core->listen_list_lock, flags);
+
+	/* disconnect any connected qp's on ifdown */
+	if (!ifup)
+		i40iw_cm_disconnect_all(iwdev);
+}
diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.h b/drivers/infiniband/hw/i40iw/i40iw_cm.h
index 0381b7f5e20d..49ed7a52a84d 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_cm.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_cm.h
@@ -444,5 +444,7 @@ int i40iw_arp_table(struct i40iw_device *iwdev,
 		    u8 *mac_addr,
 		    u32 action);
 
+void i40iw_if_notify(struct i40iw_device *iwdev, struct net_device *netdev,
+		     u32 *ipaddr, bool ipv4, bool ifup);
 void i40iw_cm_disconnect_all(struct i40iw_device *iwdev);
 #endif /* I40IW_CM_H */
diff --git a/drivers/infiniband/hw/i40iw/i40iw_utils.c b/drivers/infiniband/hw/i40iw/i40iw_utils.c
index 58151280828d..641f00f3dda1 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_utils.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_utils.c
@@ -153,6 +153,7 @@ int i40iw_inetaddr_event(struct notifier_block *notifier,
 	struct i40iw_device *iwdev;
 	struct i40iw_handler *hdl;
 	u32 local_ipaddr;
+	u32 action = I40IW_ARP_ADD;
 
 	hdl = i40iw_find_netdev(event_netdev);
 	if (!hdl)
@@ -164,44 +165,25 @@ int i40iw_inetaddr_event(struct notifier_block *notifier,
 	if (netdev != event_netdev)
 		return NOTIFY_DONE;
 
+	if (upper_dev)
+		local_ipaddr = ntohl(
+			((struct in_device *)upper_dev->ip_ptr)->ifa_list->ifa_address);
+	else
+		local_ipaddr = ntohl(ifa->ifa_address);
 	switch (event) {
 	case NETDEV_DOWN:
-		if (upper_dev)
-			local_ipaddr = ntohl(
-				((struct in_device *)upper_dev->ip_ptr)->ifa_list->ifa_address);
-		else
-			local_ipaddr = ntohl(ifa->ifa_address);
-		i40iw_manage_arp_cache(iwdev,
-				       netdev->dev_addr,
-				       &local_ipaddr,
-				       true,
-				       I40IW_ARP_DELETE);
-		return NOTIFY_OK;
+		action = I40IW_ARP_DELETE;
+		/* Fall through */
 	case NETDEV_UP:
-		if (upper_dev)
-			local_ipaddr = ntohl(
-				((struct in_device *)upper_dev->ip_ptr)->ifa_list->ifa_address);
-		else
-			local_ipaddr = ntohl(ifa->ifa_address);
-		i40iw_manage_arp_cache(iwdev,
-				       netdev->dev_addr,
-				       &local_ipaddr,
-				       true,
-				       I40IW_ARP_ADD);
-		break;
+		/* Fall through */
 	case NETDEV_CHANGEADDR:
-		/* Add the address to the IP table */
-		if (upper_dev)
-			local_ipaddr = ntohl(
-				((struct in_device *)upper_dev->ip_ptr)->ifa_list->ifa_address);
-		else
-			local_ipaddr = ntohl(ifa->ifa_address);
-
 		i40iw_manage_arp_cache(iwdev,
 				       netdev->dev_addr,
 				       &local_ipaddr,
 				       true,
-				       I40IW_ARP_ADD);
+				       action);
+		i40iw_if_notify(iwdev, netdev, &local_ipaddr, true,
+				(action == I40IW_ARP_ADD) ? true : false);
 		break;
 	default:
 		break;
@@ -225,6 +207,7 @@ int i40iw_inet6addr_event(struct notifier_block *notifier,
 	struct i40iw_device *iwdev;
 	struct i40iw_handler *hdl;
 	u32 local_ipaddr6[4];
+	u32 action = I40IW_ARP_ADD;
 
 	hdl = i40iw_find_netdev(event_netdev);
 	if (!hdl)
@@ -235,24 +218,21 @@ int i40iw_inet6addr_event(struct notifier_block *notifier,
 	if (netdev != event_netdev)
 		return NOTIFY_DONE;
 
+	i40iw_copy_ip_ntohl(local_ipaddr6, ifa->addr.in6_u.u6_addr32);
 	switch (event) {
 	case NETDEV_DOWN:
-		i40iw_copy_ip_ntohl(local_ipaddr6, ifa->addr.in6_u.u6_addr32);
-		i40iw_manage_arp_cache(iwdev,
-				       netdev->dev_addr,
-				       local_ipaddr6,
-				       false,
-				       I40IW_ARP_DELETE);
-		return NOTIFY_OK;
+		action = I40IW_ARP_DELETE;
+		/* Fall through */
 	case NETDEV_UP:
 		/* Fall through */
 	case NETDEV_CHANGEADDR:
-		i40iw_copy_ip_ntohl(local_ipaddr6, ifa->addr.in6_u.u6_addr32);
 		i40iw_manage_arp_cache(iwdev,
 				       netdev->dev_addr,
 				       local_ipaddr6,
 				       false,
-				       I40IW_ARP_ADD);
+				       action);
+		i40iw_if_notify(iwdev, netdev, local_ipaddr6, false,
+				(action == I40IW_ARP_ADD) ? true : false);
 		break;
 	default:
 		break;
-- 
cgit v1.2.1


From a05e15135b67d71f30f70ab45dede4706f988439 Mon Sep 17 00:00:00 2001
From: Mustafa Ismail <mustafa.ismail@intel.com>
Date: Wed, 30 Nov 2016 15:08:34 -0600
Subject: i40iw: Replace list_for_each_entry macro with safe version

Use list_for_each_entry_safe macro for the IPv6 addr list
as IPv6 addresses can be deleted while going through the
list.

Signed-off-by: Mustafa Ismail <mustafa.ismail@intel.com>
Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_cm.c   | 4 ++--
 drivers/infiniband/hw/i40iw/i40iw_main.c | 4 ++--
 2 files changed, 4 insertions(+), 4 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.c b/drivers/infiniband/hw/i40iw/i40iw_cm.c
index 93ae764fc44e..9a14880fe85a 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_cm.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_cm.c
@@ -1644,7 +1644,7 @@ static enum i40iw_status_code i40iw_add_mqh_6(struct i40iw_device *iwdev,
 {
 	struct net_device *ip_dev;
 	struct inet6_dev *idev;
-	struct inet6_ifaddr *ifp;
+	struct inet6_ifaddr *ifp, *tmp;
 	enum i40iw_status_code ret = 0;
 	struct i40iw_cm_listener *child_listen_node;
 	unsigned long flags;
@@ -1659,7 +1659,7 @@ static enum i40iw_status_code i40iw_add_mqh_6(struct i40iw_device *iwdev,
 				i40iw_pr_err("idev == NULL\n");
 				break;
 			}
-			list_for_each_entry(ifp, &idev->addr_list, if_list) {
+			list_for_each_entry_safe(ifp, tmp, &idev->addr_list, if_list) {
 				i40iw_debug(&iwdev->sc_dev,
 					    I40IW_DEBUG_CM,
 					    "IP=%pI6, vlan_id=%d, MAC=%pM\n",
diff --git a/drivers/infiniband/hw/i40iw/i40iw_main.c b/drivers/infiniband/hw/i40iw/i40iw_main.c
index d86bb6e98f07..4ce05b8e8f0a 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_main.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_main.c
@@ -1164,7 +1164,7 @@ static void i40iw_add_ipv6_addr(struct i40iw_device *iwdev)
 {
 	struct net_device *ip_dev;
 	struct inet6_dev *idev;
-	struct inet6_ifaddr *ifp;
+	struct inet6_ifaddr *ifp, *tmp;
 	u32 local_ipaddr6[4];
 
 	rcu_read_lock();
@@ -1177,7 +1177,7 @@ static void i40iw_add_ipv6_addr(struct i40iw_device *iwdev)
 				i40iw_pr_err("ipv6 inet device not found\n");
 				break;
 			}
-			list_for_each_entry(ifp, &idev->addr_list, if_list) {
+			list_for_each_entry_safe(ifp, tmp, &idev->addr_list, if_list) {
 				i40iw_pr_info("IP=%pI6, vlan_id=%d, MAC=%pM\n", &ifp->addr,
 					      rdma_vlan_dev_vlan_id(ip_dev), ip_dev->dev_addr);
 				i40iw_copy_ip_ntohl(local_ipaddr6,
-- 
cgit v1.2.1


From e0b010da87e3aaf7ca9d28ba5d141924a7f8c66d Mon Sep 17 00:00:00 2001
From: Shiraz Saleem <shiraz.saleem@intel.com>
Date: Wed, 30 Nov 2016 15:09:07 -0600
Subject: i40iw: Add NULL check for ibqp event handler

Add NULL check for ibqp event handler before calling it to report
QP events, as it might not initialized.

Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Faisal Latif <faisal.latif@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_cm.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.c b/drivers/infiniband/hw/i40iw/i40iw_cm.c
index 9a14880fe85a..13b6dee4d57c 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_cm.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_cm.c
@@ -3474,7 +3474,7 @@ static void i40iw_cm_disconn_true(struct i40iw_qp *iwqp)
 		/* Flush the queues */
 		i40iw_flush_wqes(iwdev, iwqp);
 
-		if (qp->term_flags) {
+		if (qp->term_flags && iwqp->ibqp.event_handler) {
 			ibevent.device = iwqp->ibqp.device;
 			ibevent.event = (qp->eventtype == TERM_EVENT_QP_FATAL) ?
 					IB_EVENT_QP_FATAL : IB_EVENT_QP_ACCESS_ERR;
-- 
cgit v1.2.1


From 7eb2bde7f3900f044ab351e450adc41623ff2f5c Mon Sep 17 00:00:00 2001
From: Shiraz Saleem <shiraz.saleem@intel.com>
Date: Wed, 30 Nov 2016 15:09:34 -0600
Subject: i40iw: Set TOS field in IP header

Set the TOS field in IP header with the value passed in
from application. If there is mismatch between the remote
client's TOS and listener, set the listener Tos to the higher
of the two values.

Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Faisal Latif <faisal.latif@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_cm.c | 25 +++++++++++++++++++------
 drivers/infiniband/hw/i40iw/i40iw_cm.h |  3 +++
 2 files changed, 22 insertions(+), 6 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.c b/drivers/infiniband/hw/i40iw/i40iw_cm.c
index 13b6dee4d57c..0c92037d4f7c 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_cm.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_cm.c
@@ -445,7 +445,7 @@ static struct i40iw_puda_buf *i40iw_form_cm_frame(struct i40iw_cm_node *cm_node,
 
 		iph->version = IPVERSION;
 		iph->ihl = 5;	/* 5 * 4Byte words, IP headr len */
-		iph->tos = 0;
+		iph->tos = cm_node->tos;
 		iph->tot_len = htons(packetsize);
 		iph->id = htons(++cm_node->tcp_cntxt.loc_id);
 
@@ -472,7 +472,8 @@ static struct i40iw_puda_buf *i40iw_form_cm_frame(struct i40iw_cm_node *cm_node,
 			ethh->h_proto = htons(ETH_P_IPV6);
 		}
 		ip6h->version = 6;
-		ip6h->flow_lbl[0] = 0;
+		ip6h->priority = cm_node->tos >> 4;
+		ip6h->flow_lbl[0] = cm_node->tos << 4;
 		ip6h->flow_lbl[1] = 0;
 		ip6h->flow_lbl[2] = 0;
 		ip6h->payload_len = htons(packetsize - sizeof(*ip6h));
@@ -2141,9 +2142,18 @@ static struct i40iw_cm_node *i40iw_make_cm_node(
 	cm_node->vlan_id = cm_info->vlan_id;
 	if ((cm_node->vlan_id == I40IW_NO_VLAN) && iwdev->dcb)
 		cm_node->vlan_id = 0;
+	cm_node->tos = cm_info->tos;
 	cm_node->user_pri = cm_info->user_pri;
-	if (listener)
-		cm_node->user_pri = listener->user_pri;
+	if (listener) {
+		if (listener->tos != cm_info->tos)
+			i40iw_debug(&iwdev->sc_dev, I40IW_DEBUG_DCB,
+				    "application TOS[%d] and remote client TOS[%d] mismatch\n",
+				     listener->tos, cm_info->tos);
+		cm_node->tos = max(listener->tos, cm_info->tos);
+		cm_node->user_pri = rt_tos2priority(cm_node->tos);
+		i40iw_debug(&iwdev->sc_dev, I40IW_DEBUG_DCB, "listener: TOS:[%d] UP:[%d]\n",
+			    cm_node->tos, cm_node->user_pri);
+	}
 	memcpy(cm_node->loc_addr, cm_info->loc_addr, sizeof(cm_node->loc_addr));
 	memcpy(cm_node->rem_addr, cm_info->rem_addr, sizeof(cm_node->rem_addr));
 	cm_node->loc_port = cm_info->loc_port;
@@ -3092,6 +3102,7 @@ void i40iw_receive_ilq(struct i40iw_sc_dev *dev, struct i40iw_puda_buf *rbuf)
 		cm_info.loc_addr[0] = ntohl(iph->daddr);
 		cm_info.rem_addr[0] = ntohl(iph->saddr);
 		cm_info.ipv4 = true;
+		cm_info.tos = iph->tos;
 	} else {
 		ip6h = (struct ipv6hdr *)rbuf->iph;
 		i40iw_copy_ip_ntohl(cm_info.loc_addr,
@@ -3099,6 +3110,7 @@ void i40iw_receive_ilq(struct i40iw_sc_dev *dev, struct i40iw_puda_buf *rbuf)
 		i40iw_copy_ip_ntohl(cm_info.rem_addr,
 				    ip6h->saddr.in6_u.u6_addr32);
 		cm_info.ipv4 = false;
+		cm_info.tos = (ip6h->priority << 4) | (ip6h->flow_lbl[0] >> 4);
 	}
 	cm_info.loc_port = ntohs(tcph->dest);
 	cm_info.rem_port = ntohs(tcph->source);
@@ -3331,6 +3343,7 @@ static void i40iw_cm_init_tsa_conn(struct i40iw_qp *iwqp,
 	cm_node->state = I40IW_CM_STATE_OFFLOADED;
 	tcp_info.tcp_state = I40IW_TCP_STATE_ESTABLISHED;
 	tcp_info.src_mac_addr_idx = iwdev->mac_ip_table_idx;
+	tcp_info.tos = cm_node->tos;
 
 	dev->iw_priv_qp_ops->qp_setctx(&iwqp->sc_qp, (u64 *)(iwqp->host_ctx.va), ctx_info);
 
@@ -3763,6 +3776,7 @@ int i40iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param)
 		i40iw_netdev_vlan_ipv6(cm_info.loc_addr, &cm_info.vlan_id, NULL);
 	}
 	cm_info.cm_id = cm_id;
+	cm_info.tos = cm_id->tos;
 	cm_info.user_pri = rt_tos2priority(cm_id->tos);
 	i40iw_debug(&iwdev->sc_dev, I40IW_DEBUG_DCB, "%s TOS:[%d] UP:[%d]\n",
 		    __func__, cm_id->tos, cm_info.user_pri);
@@ -3911,10 +3925,9 @@ int i40iw_create_listen(struct iw_cm_id *cm_id, int backlog)
 
 	cm_id->provider_data = cm_listen_node;
 
+	cm_listen_node->tos = cm_id->tos;
 	cm_listen_node->user_pri = rt_tos2priority(cm_id->tos);
 	cm_info.user_pri = cm_listen_node->user_pri;
-	i40iw_debug(&iwdev->sc_dev, I40IW_DEBUG_DCB, "%s TOS:[%d] UP:[%d]\n",
-		    __func__, cm_id->tos, cm_listen_node->user_pri);
 
 	if (!cm_listen_node->reused_node) {
 		if (wildcard) {
diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.h b/drivers/infiniband/hw/i40iw/i40iw_cm.h
index 49ed7a52a84d..2e52e38ffcf3 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_cm.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_cm.h
@@ -297,6 +297,7 @@ struct i40iw_cm_listener {
 	enum i40iw_cm_listener_state listener_state;
 	u32 reused_node;
 	u8 user_pri;
+	u8 tos;
 	u16 vlan_id;
 	bool qhash_set;
 	bool ipv4;
@@ -343,6 +344,7 @@ struct i40iw_cm_node {
 	atomic_t passive_state;
 	bool qhash_set;
 	u8 user_pri;
+	u8 tos;
 	bool ipv4;
 	bool snd_mark_en;
 	u16 lsmm_size;
@@ -368,6 +370,7 @@ struct i40iw_cm_info {
 	u16 vlan_id;
 	int backlog;
 	u8 user_pri;
+	u8 tos;
 	bool ipv4;
 };
 
-- 
cgit v1.2.1


From fd4e906b2e2c04056e8c1773b5b6e06d307239e6 Mon Sep 17 00:00:00 2001
From: Shiraz Saleem <shiraz.saleem@intel.com>
Date: Wed, 30 Nov 2016 15:12:11 -0600
Subject: i40iw: Fill in IRD value when on connect request

IRD is not populated on connect request and application is
getting 0 for the value. Fill in the correct value on
connect request.

Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Faisal Latif <faisal.latif@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_cm.c | 1 +
 1 file changed, 1 insertion(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.c b/drivers/infiniband/hw/i40iw/i40iw_cm.c
index 0c92037d4f7c..2f14de77cb1a 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_cm.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_cm.c
@@ -272,6 +272,7 @@ static int i40iw_send_cm_event(struct i40iw_cm_node *cm_node,
 		event.provider_data = (void *)cm_node;
 		event.private_data = (void *)cm_node->pdata_buf;
 		event.private_data_len = (u8)cm_node->pdata.size;
+		event.ird = cm_node->ird_size;
 		break;
 	case IW_CM_EVENT_CONNECT_REPLY:
 		i40iw_get_cmevent_info(cm_node, cm_id, &event);
-- 
cgit v1.2.1


From bf69f494c337cf3c43d3358ad66642dbde50fe03 Mon Sep 17 00:00:00 2001
From: Shiraz Saleem <shiraz.saleem@intel.com>
Date: Wed, 30 Nov 2016 15:12:35 -0600
Subject: i40iw: Correctly fail loopback connection if no listener

Fail the connect and return the proper error code if a client
is started with local IP address and there is no corresponding
loopback listener.

Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Faisal Latif <faisal.latif@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_cm.c | 62 +++++++++++++++++++---------------
 1 file changed, 35 insertions(+), 27 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.c b/drivers/infiniband/hw/i40iw/i40iw_cm.c
index 2f14de77cb1a..25af89a3cdce 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_cm.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_cm.c
@@ -2878,7 +2878,7 @@ static struct i40iw_cm_node *i40iw_create_cm_node(
 	/* create a CM connection node */
 	cm_node = i40iw_make_cm_node(cm_core, iwdev, cm_info, NULL);
 	if (!cm_node)
-		return NULL;
+		return ERR_PTR(-ENOMEM);
 	/* set our node side to client (active) side */
 	cm_node->tcp_cntxt.client = 1;
 	cm_node->tcp_cntxt.rcv_wscale = I40IW_CM_DEFAULT_RCV_WND_SCALE;
@@ -2891,7 +2891,8 @@ static struct i40iw_cm_node *i40iw_create_cm_node(
 						cm_node->vlan_id,
 						I40IW_CM_LISTENER_ACTIVE_STATE);
 		if (!loopback_remotelistener) {
-			i40iw_create_event(cm_node, I40IW_CM_EVENT_ABORTED);
+			i40iw_rem_ref_cm_node(cm_node);
+			return ERR_PTR(-ECONNREFUSED);
 		} else {
 			loopback_cm_info = *cm_info;
 			loopback_cm_info.loc_port = cm_info->rem_port;
@@ -2904,7 +2905,7 @@ static struct i40iw_cm_node *i40iw_create_cm_node(
 								 loopback_remotelistener);
 			if (!loopback_remotenode) {
 				i40iw_rem_ref_cm_node(cm_node);
-				return NULL;
+				return ERR_PTR(-ENOMEM);
 			}
 			cm_core->stats_loopbacks++;
 			loopback_remotenode->loopbackpartner = cm_node;
@@ -3732,6 +3733,7 @@ int i40iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param)
 	struct sockaddr_in6 *raddr6;
 	bool qhash_set = false;
 	int apbvt_set = 0;
+	int err = 0;
 	enum i40iw_status_code status;
 
 	ibqp = i40iw_get_qp(cm_id->device, conn_param->qpn);
@@ -3812,8 +3814,11 @@ int i40iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param)
 				       conn_param->private_data_len,
 				       (void *)conn_param->private_data,
 				       &cm_info);
-	if (!cm_node)
-		goto err;
+
+	if (IS_ERR(cm_node)) {
+		err = PTR_ERR(cm_node);
+		goto err_out;
+	}
 
 	i40iw_record_ird_ord(cm_node, (u16)conn_param->ird, (u16)conn_param->ord);
 	if (cm_node->send_rdma0_op == SEND_RDMA_READ_ZERO &&
@@ -3827,10 +3832,12 @@ int i40iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param)
 	iwqp->cm_id = cm_id;
 	i40iw_add_ref(&iwqp->ibqp);
 
-	if (cm_node->state == I40IW_CM_STATE_SYN_SENT) {
-		if (i40iw_send_syn(cm_node, 0)) {
+	if (cm_node->state != I40IW_CM_STATE_OFFLOADED) {
+		cm_node->state = I40IW_CM_STATE_SYN_SENT;
+		err = i40iw_send_syn(cm_node, 0);
+		if (err) {
 			i40iw_rem_ref_cm_node(cm_node);
-			goto err;
+			goto err_out;
 		}
 	}
 
@@ -3842,24 +3849,25 @@ int i40iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param)
 		    cm_node->cm_id);
 	return 0;
 
-err:
-	if (cm_node) {
-		if (cm_node->ipv4)
-			i40iw_debug(cm_node->dev,
-				    I40IW_DEBUG_CM,
-				    "Api - connect() FAILED: dest addr=%pI4",
-				    cm_node->rem_addr);
-		else
-			i40iw_debug(cm_node->dev, I40IW_DEBUG_CM,
-				    "Api - connect() FAILED: dest addr=%pI6",
-				    cm_node->rem_addr);
-	}
-	i40iw_manage_qhash(iwdev,
-			   &cm_info,
-			   I40IW_QHASH_TYPE_TCP_ESTABLISHED,
-			   I40IW_QHASH_MANAGE_TYPE_DELETE,
-			   NULL,
-			   false);
+err_out:
+	if (cm_info.ipv4)
+		i40iw_debug(&iwdev->sc_dev,
+			    I40IW_DEBUG_CM,
+			    "Api - connect() FAILED: dest addr=%pI4",
+			    cm_info.rem_addr);
+	else
+		i40iw_debug(&iwdev->sc_dev,
+			    I40IW_DEBUG_CM,
+			    "Api - connect() FAILED: dest addr=%pI6",
+			    cm_info.rem_addr);
+
+	if (qhash_set)
+		i40iw_manage_qhash(iwdev,
+				   &cm_info,
+				   I40IW_QHASH_TYPE_TCP_ESTABLISHED,
+				   I40IW_QHASH_MANAGE_TYPE_DELETE,
+				   NULL,
+				   false);
 
 	if (apbvt_set && !i40iw_listen_port_in_use(&iwdev->cm_core,
 						   cm_info.loc_port))
@@ -3868,7 +3876,7 @@ err:
 				   I40IW_MANAGE_APBVT_DEL);
 	cm_id->rem_ref(cm_id);
 	iwdev->cm_core.stats_connect_errs++;
-	return -ENOMEM;
+	return err;
 }
 
 /**
-- 
cgit v1.2.1


From 1ef936b229c54e73a3cce9f4eb8dd5a146add073 Mon Sep 17 00:00:00 2001
From: Henry Orosco <henry.orosco@intel.com>
Date: Wed, 30 Nov 2016 15:13:47 -0600
Subject: i40iw: Code cleanup, remove check of PBLE pages

Remove check for zero 'pages' of unallocated pbles calculated in
add_pble_pool(); as it can never be true.

Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_pble.c | 4 ----
 1 file changed, 4 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_pble.c b/drivers/infiniband/hw/i40iw/i40iw_pble.c
index 85993dc44f6e..c87ba1617087 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_pble.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_pble.c
@@ -353,10 +353,6 @@ static enum i40iw_status_code add_pble_pool(struct i40iw_sc_dev *dev,
 	pages = (idx->rel_pd_idx) ? (I40IW_HMC_PD_CNT_IN_SD -
 			idx->rel_pd_idx) : I40IW_HMC_PD_CNT_IN_SD;
 	pages = min(pages, pble_rsrc->unallocated_pble >> PBLE_512_SHIFT);
-	if (!pages) {
-		ret_code = I40IW_ERR_NO_PBLCHUNKS_AVAILABLE;
-		goto error;
-	}
 	info.chunk = chunk;
 	info.hmc_info = hmc_info;
 	info.pages = pages;
-- 
cgit v1.2.1


From 78300cf8152f87adb20fbe71a600e0d8d72aabe8 Mon Sep 17 00:00:00 2001
From: Henry Orosco <henry.orosco@intel.com>
Date: Wed, 30 Nov 2016 15:14:15 -0600
Subject: i40iw: Add request for reset on CQP timeout

When CQP times out, send a request to LAN driver for reset.

Signed-off-by: Mustafa Ismail <mustafa.ismail@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw.h       |  1 +
 drivers/infiniband/hw/i40iw/i40iw_utils.c | 10 +++++++++-
 2 files changed, 10 insertions(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw.h b/drivers/infiniband/hw/i40iw/i40iw.h
index c795c6160261..ef188e6b1f89 100644
--- a/drivers/infiniband/hw/i40iw/i40iw.h
+++ b/drivers/infiniband/hw/i40iw/i40iw.h
@@ -304,6 +304,7 @@ struct i40iw_device {
 	u32 mpa_version;
 	bool dcb;
 	bool closing;
+	bool reset;
 	u32 used_pds;
 	u32 used_cqs;
 	u32 used_mrs;
diff --git a/drivers/infiniband/hw/i40iw/i40iw_utils.c b/drivers/infiniband/hw/i40iw/i40iw_utils.c
index 641f00f3dda1..4a08ffb75d2e 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_utils.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_utils.c
@@ -396,7 +396,10 @@ static int i40iw_wait_event(struct i40iw_device *iwdev,
 		i40iw_pr_err("error cqp command 0x%x timed out ret = %d\n",
 			     info->cqp_cmd, timeout_ret);
 		err_code = -ETIME;
-		i40iw_request_reset(iwdev);
+		if (!iwdev->reset) {
+			iwdev->reset = true;
+			i40iw_request_reset(iwdev);
+		}
 		goto done;
 	}
 	cqp_error = cqp_request->compl_info.error;
@@ -426,6 +429,11 @@ enum i40iw_status_code i40iw_handle_cqp_op(struct i40iw_device *iwdev,
 	struct cqp_commands_info *info = &cqp_request->info;
 	int err_code = 0;
 
+	if (iwdev->reset) {
+		i40iw_free_cqp_request(&iwdev->cqp, cqp_request);
+		return I40IW_ERR_CQP_COMPL_ERROR;
+	}
+
 	status = i40iw_process_cqp_cmd(dev, info);
 	if (status) {
 		i40iw_pr_err("error cqp command 0x%x failed\n", info->cqp_cmd);
-- 
cgit v1.2.1


From 15f7e3c21b76598bc6e5816d2577ce843b2b963f Mon Sep 17 00:00:00 2001
From: Wei Yongjun <weiyongjun1@huawei.com>
Date: Sat, 17 Sep 2016 00:41:37 +0000
Subject: iw_cxgb4: Fix error return code in c4iw_rdev_open()

Fix to return error code -ENOMEM from the __get_free_page() error
handling case instead of 0, as done elsewhere in this function.

Fixes: 05eb23893c2c ("cxgb4/iw_cxgb4: Doorbell Drop Avoidance Bug Fixes")
Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Acked-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/cxgb4/device.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/cxgb4/device.c b/drivers/infiniband/hw/cxgb4/device.c
index 93e3d270a98a..b99dc9e0ffb2 100644
--- a/drivers/infiniband/hw/cxgb4/device.c
+++ b/drivers/infiniband/hw/cxgb4/device.c
@@ -828,8 +828,10 @@ static int c4iw_rdev_open(struct c4iw_rdev *rdev)
 	}
 	rdev->status_page = (struct t4_dev_status_page *)
 			    __get_free_page(GFP_KERNEL);
-	if (!rdev->status_page)
+	if (!rdev->status_page) {
+		err = -ENOMEM;
 		goto destroy_ocqp_pool;
+	}
 	rdev->status_page->qp_start = rdev->lldi.vr->qp.start;
 	rdev->status_page->qp_size = rdev->lldi.vr->qp.size;
 	rdev->status_page->cq_start = rdev->lldi.vr->cq.start;
-- 
cgit v1.2.1


From 62aeddbf28fa63872e8f13f47177338b0f1fd8b5 Mon Sep 17 00:00:00 2001
From: Dean Luick <dean.luick@intel.com>
Date: Wed, 7 Dec 2016 19:32:15 -0800
Subject: IB/hfi1: Read new EPROM format

Add the ability to read the new EPROM format.

Reviewed-by: Easwar Hariharan <easwar.hariharan@intel.com>
Signed-off-by: Dean Luick <dean.luick@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/eprom.c | 211 +++++++++++++++++++++++++++++++++++--
 1 file changed, 203 insertions(+), 8 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/eprom.c b/drivers/infiniband/hw/hfi1/eprom.c
index e70c223801b4..26da124c88e2 100644
--- a/drivers/infiniband/hw/hfi1/eprom.c
+++ b/drivers/infiniband/hw/hfi1/eprom.c
@@ -207,6 +207,40 @@ done_asic:
 /* magic character sequence that trails an image */
 #define IMAGE_TRAIL_MAGIC "egamiAPO"
 
+/* EPROM file types */
+#define HFI1_EFT_PLATFORM_CONFIG 2
+
+/* segment size - 128 KiB */
+#define SEG_SIZE (128 * 1024)
+
+struct hfi1_eprom_footer {
+	u32 oprom_size;		/* size of the oprom, in bytes */
+	u16 num_table_entries;
+	u16 version;		/* version of this footer */
+	u32 magic;		/* must be last */
+};
+
+struct hfi1_eprom_table_entry {
+	u32 type;		/* file type */
+	u32 offset;		/* file offset from start of EPROM */
+	u32 size;		/* file size, in bytes */
+};
+
+/*
+ * Calculate the max number of table entries that will fit within a directory
+ * buffer of size 'dir_size'.
+ */
+#define MAX_TABLE_ENTRIES(dir_size) \
+	(((dir_size) - sizeof(struct hfi1_eprom_footer)) / \
+		sizeof(struct hfi1_eprom_table_entry))
+
+#define DIRECTORY_SIZE(n) (sizeof(struct hfi1_eprom_footer) + \
+	(sizeof(struct hfi1_eprom_table_entry) * (n)))
+
+#define MAGIC4(a, b, c, d) ((d) << 24 | (c) << 16 | (b) << 8 | (a))
+#define FOOTER_MAGIC MAGIC4('e', 'p', 'r', 'm')
+#define FOOTER_VERSION 1
+
 /*
  * Read all of partition 1.  The actual file is at the front.  Adjust
  * the returned size if a trailing image magic is found.
@@ -241,6 +275,167 @@ static int read_partition_platform_config(struct hfi1_devdata *dd, void **data,
 	return 0;
 }
 
+/*
+ * The segment magic has been checked.  There is a footer and table of
+ * contents present.
+ *
+ * directory is a u32 aligned buffer of size EP_PAGE_SIZE.
+ */
+static int read_segment_platform_config(struct hfi1_devdata *dd,
+					void *directory, void **data, u32 *size)
+{
+	struct hfi1_eprom_footer *footer;
+	struct hfi1_eprom_table_entry *table;
+	struct hfi1_eprom_table_entry *entry;
+	void *buffer = NULL;
+	void *table_buffer = NULL;
+	int ret, i;
+	u32 directory_size;
+	u32 seg_base, seg_offset;
+	u32 bytes_available, ncopied, to_copy;
+
+	/* the footer is at the end of the directory */
+	footer = (struct hfi1_eprom_footer *)
+			(directory + EP_PAGE_SIZE - sizeof(*footer));
+
+	/* make sure the structure version is supported */
+	if (footer->version != FOOTER_VERSION)
+		return -EINVAL;
+
+	/* oprom size cannot be larger than a segment */
+	if (footer->oprom_size >= SEG_SIZE)
+		return -EINVAL;
+
+	/* the file table must fit in a segment with the oprom */
+	if (footer->num_table_entries >
+			MAX_TABLE_ENTRIES(SEG_SIZE - footer->oprom_size))
+		return -EINVAL;
+
+	/* find the file table start, which precedes the footer */
+	directory_size = DIRECTORY_SIZE(footer->num_table_entries);
+	if (directory_size <= EP_PAGE_SIZE) {
+		/* the file table fits into the directory buffer handed in */
+		table = (struct hfi1_eprom_table_entry *)
+				(directory + EP_PAGE_SIZE - directory_size);
+	} else {
+		/* need to allocate and read more */
+		table_buffer = kmalloc(directory_size, GFP_KERNEL);
+		if (!table_buffer)
+			return -ENOMEM;
+		ret = read_length(dd, SEG_SIZE - directory_size,
+				  directory_size, table_buffer);
+		if (ret)
+			goto done;
+		table = table_buffer;
+	}
+
+	/* look for the platform configuration file in the table */
+	for (entry = NULL, i = 0; i < footer->num_table_entries; i++) {
+		if (table[i].type == HFI1_EFT_PLATFORM_CONFIG) {
+			entry = &table[i];
+			break;
+		}
+	}
+	if (!entry) {
+		ret = -ENOENT;
+		goto done;
+	}
+
+	/*
+	 * Sanity check on the configuration file size - it should never
+	 * be larger than 4 KiB.
+	 */
+	if (entry->size > (4 * 1024)) {
+		dd_dev_err(dd, "Bad configuration file size 0x%x\n",
+			   entry->size);
+		ret = -EINVAL;
+		goto done;
+	}
+
+	/* check for bogus offset and size that wrap when added together */
+	if (entry->offset + entry->size < entry->offset) {
+		dd_dev_err(dd,
+			   "Bad configuration file start + size 0x%x+0x%x\n",
+			   entry->offset, entry->size);
+		ret = -EINVAL;
+		goto done;
+	}
+
+	/* allocate the buffer to return */
+	buffer = kmalloc(entry->size, GFP_KERNEL);
+	if (!buffer) {
+		ret = -ENOMEM;
+		goto done;
+	}
+
+	/*
+	 * Extract the file by looping over segments until it is fully read.
+	 */
+	seg_offset = entry->offset % SEG_SIZE;
+	seg_base = entry->offset - seg_offset;
+	ncopied = 0;
+	while (ncopied < entry->size) {
+		/* calculate data bytes available in this segment */
+
+		/* start with the bytes from the current offset to the end */
+		bytes_available = SEG_SIZE - seg_offset;
+		/* subtract off footer and table from segment 0 */
+		if (seg_base == 0) {
+			/*
+			 * Sanity check: should not have a starting point
+			 * at or within the directory.
+			 */
+			if (bytes_available <= directory_size) {
+				dd_dev_err(dd,
+					   "Bad configuration file - offset 0x%x within footer+table\n",
+					   entry->offset);
+				ret = -EINVAL;
+				goto done;
+			}
+			bytes_available -= directory_size;
+		}
+
+		/* calculate bytes wanted */
+		to_copy = entry->size - ncopied;
+
+		/* max out at the available bytes in this segment */
+		if (to_copy > bytes_available)
+			to_copy = bytes_available;
+
+		/*
+		 * Read from the EPROM.
+		 *
+		 * The sanity check for entry->offset is done in read_length().
+		 * The EPROM offset is validated against what the hardware
+		 * addressing supports.  In addition, if the offset is larger
+		 * than the actual EPROM, it silently wraps.  It will work
+		 * fine, though the reader may not get what they expected
+		 * from the EPROM.
+		 */
+		ret = read_length(dd, seg_base + seg_offset, to_copy,
+				  buffer + ncopied);
+		if (ret)
+			goto done;
+
+		ncopied += to_copy;
+
+		/* set up for next segment */
+		seg_offset = footer->oprom_size;
+		seg_base += SEG_SIZE;
+	}
+
+	/* success */
+	ret = 0;
+	*data = buffer;
+	*size = entry->size;
+
+done:
+	kfree(table_buffer);
+	if (ret)
+		kfree(buffer);
+	return ret;
+}
+
 /*
  * Read the platform configuration file from the EPROM.
  *
@@ -253,6 +448,7 @@ static int read_partition_platform_config(struct hfi1_devdata *dd, void **data,
  *   -EBUSY   - not able to acquire access to the EPROM
  *   -ENOENT  - no recognizable file written
  *   -ENOMEM  - buffer could not be allocated
+ *   -EINVAL  - invalid EPROM contentents found
  */
 int eprom_read_platform_config(struct hfi1_devdata *dd, void **data, u32 *size)
 {
@@ -266,21 +462,20 @@ int eprom_read_platform_config(struct hfi1_devdata *dd, void **data, u32 *size)
 	if (ret)
 		return -EBUSY;
 
-	/* read the last page of P0 for the EPROM format magic */
-	ret = read_length(dd, P1_START - EP_PAGE_SIZE, EP_PAGE_SIZE, directory);
+	/* read the last page of the segment for the EPROM format magic */
+	ret = read_length(dd, SEG_SIZE - EP_PAGE_SIZE, EP_PAGE_SIZE, directory);
 	if (ret)
 		goto done;
 
-	/* last dword of P0 contains a magic indicator */
-	if (directory[EP_PAGE_DWORDS - 1] == 0) {
+	/* last dword of the segment contains a magic value */
+	if (directory[EP_PAGE_DWORDS - 1] == FOOTER_MAGIC) {
+		/* segment format */
+		ret = read_segment_platform_config(dd, directory, data, size);
+	} else {
 		/* partition format */
 		ret = read_partition_platform_config(dd, data, size);
-		goto done;
 	}
 
-	/* nothing recognized */
-	ret = -ENOENT;
-
 done:
 	release_chip_resource(dd, CR_EPROM);
 	return ret;
-- 
cgit v1.2.1


From d7cf4ccf6fe52694142f65cce82decb3fe833f48 Mon Sep 17 00:00:00 2001
From: Dean Luick <dean.luick@intel.com>
Date: Wed, 7 Dec 2016 19:32:22 -0800
Subject: IB/hfi1: Fix dc8051 multiple qword memory reads

When reading multiple dc8051 data memory locations
at once, the read enabled field must be toggled
at every address change.  Do that by writing only
the address first, then writing the enable.

Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Dean Luick <dean.luick@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/firmware.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/firmware.c b/drivers/infiniband/hw/hfi1/firmware.c
index 9a36c8c2f8a1..bcf7da4599c4 100644
--- a/drivers/infiniband/hw/hfi1/firmware.c
+++ b/drivers/infiniband/hw/hfi1/firmware.c
@@ -263,11 +263,13 @@ static int __read_8051_data(struct hfi1_devdata *dd, u32 addr, u64 *result)
 	u64 reg;
 	int count;
 
-	/* start the read at the given address */
-	reg = ((addr & DC_DC8051_CFG_RAM_ACCESS_CTRL_ADDRESS_MASK)
-			<< DC_DC8051_CFG_RAM_ACCESS_CTRL_ADDRESS_SHIFT)
-		| DC_DC8051_CFG_RAM_ACCESS_CTRL_READ_ENA_SMASK;
+	/* step 1: set the address, clear enable */
+	reg = (addr & DC_DC8051_CFG_RAM_ACCESS_CTRL_ADDRESS_MASK)
+			<< DC_DC8051_CFG_RAM_ACCESS_CTRL_ADDRESS_SHIFT;
 	write_csr(dd, DC_DC8051_CFG_RAM_ACCESS_CTRL, reg);
+	/* step 2: enable */
+	write_csr(dd, DC_DC8051_CFG_RAM_ACCESS_CTRL,
+		  reg | DC_DC8051_CFG_RAM_ACCESS_CTRL_READ_ENA_SMASK);
 
 	/* wait until ACCESS_COMPLETED is set */
 	count = 0;
-- 
cgit v1.2.1


From 53e91d264be9e9adbf34b83b8ce5c12452f6765e Mon Sep 17 00:00:00 2001
From: Sebastian Sanchez <sebastian.sanchez@intel.com>
Date: Wed, 7 Dec 2016 19:32:28 -0800
Subject: IB/hfi1: Use non-atomic __test_and_clear_bit in hot path

qp->r_aflags is already protected by qp->r_lock, therefore,
test_and_clear_bit() doesn't need to be atomic. Profile
shows this function call is costly.

Change the test_and_clear_bit() call to use the non-atomic
variant.

Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/rc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/rc.c b/drivers/infiniband/hw/hfi1/rc.c
index 8bc5013f39a1..aaca8588e199 100644
--- a/drivers/infiniband/hw/hfi1/rc.c
+++ b/drivers/infiniband/hw/hfi1/rc.c
@@ -2295,7 +2295,7 @@ send_last:
 		hfi1_copy_sge(&qp->r_sge, data, tlen, 1, copy_last);
 		rvt_put_ss(&qp->r_sge);
 		qp->r_msn++;
-		if (!test_and_clear_bit(RVT_R_WRID_VALID, &qp->r_aflags))
+		if (!__test_and_clear_bit(RVT_R_WRID_VALID, &qp->r_aflags))
 			break;
 		wc.wr_id = qp->r_wr_id;
 		wc.status = IB_WC_SUCCESS;
-- 
cgit v1.2.1


From 1b9e7749335281c086243c2dc75d2ad7c2b32abe Mon Sep 17 00:00:00 2001
From: Dean Luick <dean.luick@intel.com>
Date: Wed, 7 Dec 2016 19:32:34 -0800
Subject: IB/hfi1: Export 8051 memory and LCB registers via debugfs

Both the 8051 memory and LCB register access require multiple
steps and coordination with the driver.  This cannot be safely
done with resource0 alone.

The 8051 memory is exported read-only.  LCB is exported read/write.

Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Dean Luick <dean.luick@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/debugfs.c | 110 +++++++++++++++++++++++++++++++++++
 1 file changed, 110 insertions(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/debugfs.c b/drivers/infiniband/hw/hfi1/debugfs.c
index 632ba21759ab..8725f4c086cf 100644
--- a/drivers/infiniband/hw/hfi1/debugfs.c
+++ b/drivers/infiniband/hw/hfi1/debugfs.c
@@ -541,6 +541,114 @@ static ssize_t asic_flags_write(struct file *file, const char __user *buf,
 	return ret;
 }
 
+/* read the dc8051 memory */
+static ssize_t dc8051_memory_read(struct file *file, char __user *buf,
+				  size_t count, loff_t *ppos)
+{
+	struct hfi1_pportdata *ppd = private2ppd(file);
+	ssize_t rval;
+	void *tmp;
+	loff_t start, end;
+
+	/* the checks below expect the position to be positive */
+	if (*ppos < 0)
+		return -EINVAL;
+
+	tmp = kzalloc(DC8051_DATA_MEM_SIZE, GFP_KERNEL);
+	if (!tmp)
+		return -ENOMEM;
+
+	/*
+	 * Fill in the requested portion of the temporary buffer from the
+	 * 8051 memory.  The 8051 memory read is done in terms of 8 bytes.
+	 * Adjust start and end to fit.  Skip reading anything if out of
+	 * range.
+	 */
+	start = *ppos & ~0x7;	/* round down */
+	if (start < DC8051_DATA_MEM_SIZE) {
+		end = (*ppos + count + 7) & ~0x7; /* round up */
+		if (end > DC8051_DATA_MEM_SIZE)
+			end = DC8051_DATA_MEM_SIZE;
+		rval = read_8051_data(ppd->dd, start, end - start,
+				      (u64 *)(tmp + start));
+		if (rval)
+			goto done;
+	}
+
+	rval = simple_read_from_buffer(buf, count, ppos, tmp,
+				       DC8051_DATA_MEM_SIZE);
+done:
+	kfree(tmp);
+	return rval;
+}
+
+static ssize_t debugfs_lcb_read(struct file *file, char __user *buf,
+				size_t count, loff_t *ppos)
+{
+	struct hfi1_pportdata *ppd = private2ppd(file);
+	struct hfi1_devdata *dd = ppd->dd;
+	unsigned long total, csr_off;
+	u64 data;
+
+	if (*ppos < 0)
+		return -EINVAL;
+	/* only read 8 byte quantities */
+	if ((count % 8) != 0)
+		return -EINVAL;
+	/* offset must be 8-byte aligned */
+	if ((*ppos % 8) != 0)
+		return -EINVAL;
+	/* do nothing if out of range or zero count */
+	if (*ppos >= (LCB_END - LCB_START) || !count)
+		return 0;
+	/* reduce count if needed */
+	if (*ppos + count > LCB_END - LCB_START)
+		count = (LCB_END - LCB_START) - *ppos;
+
+	csr_off = LCB_START + *ppos;
+	for (total = 0; total < count; total += 8, csr_off += 8) {
+		if (read_lcb_csr(dd, csr_off, (u64 *)&data))
+			break; /* failed */
+		if (put_user(data, (unsigned long __user *)(buf + total)))
+			break;
+	}
+	*ppos += total;
+	return total;
+}
+
+static ssize_t debugfs_lcb_write(struct file *file, const char __user *buf,
+				 size_t count, loff_t *ppos)
+{
+	struct hfi1_pportdata *ppd = private2ppd(file);
+	struct hfi1_devdata *dd = ppd->dd;
+	unsigned long total, csr_off, data;
+
+	if (*ppos < 0)
+		return -EINVAL;
+	/* only write 8 byte quantities */
+	if ((count % 8) != 0)
+		return -EINVAL;
+	/* offset must be 8-byte aligned */
+	if ((*ppos % 8) != 0)
+		return -EINVAL;
+	/* do nothing if out of range or zero count */
+	if (*ppos >= (LCB_END - LCB_START) || !count)
+		return 0;
+	/* reduce count if needed */
+	if (*ppos + count > LCB_END - LCB_START)
+		count = (LCB_END - LCB_START) - *ppos;
+
+	csr_off = LCB_START + *ppos;
+	for (total = 0; total < count; total += 8, csr_off += 8) {
+		if (get_user(data, (unsigned long __user *)(buf + total)))
+			break;
+		if (write_lcb_csr(dd, csr_off, data))
+			break; /* failed */
+	}
+	*ppos += total;
+	return total;
+}
+
 /*
  * read the per-port QSFP data for ppd
  */
@@ -931,6 +1039,8 @@ static const struct counter_info port_cntr_ops[] = {
 	DEBUGFS_XOPS("qsfp2", qsfp2_debugfs_read, qsfp2_debugfs_write,
 		     qsfp2_debugfs_open, qsfp2_debugfs_release),
 	DEBUGFS_OPS("asic_flags", asic_flags_read, asic_flags_write),
+	DEBUGFS_OPS("dc8051_memory", dc8051_memory_read, NULL),
+	DEBUGFS_OPS("lcb", debugfs_lcb_read, debugfs_lcb_write),
 };
 
 static void *_sdma_cpu_list_seq_start(struct seq_file *s, loff_t *pos)
-- 
cgit v1.2.1


From e730139b3464cc740c33131c872f7d173744ef11 Mon Sep 17 00:00:00 2001
From: Jakub Pawlak <jakub.pawlak@intel.com>
Date: Wed, 7 Dec 2016 19:32:41 -0800
Subject: IB/hfi1: Disable header suppression for short packets

For the received packets with payload less or equal 8DWS
RxDmaDataFifoRdUncErr is not reported. There is set RHF.EccErr
if the header is not suppressed. When such packet is detected
on the send side the header suppression mechanism is disabled
by clearing SH bit in the packet header.

Reviewed-by: Mitko Haralanov <mitko.haralanov@intel.com>
Signed-off-by: Jakub Pawlak <jakub.pawlak@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/user_sdma.c | 60 ++++++++++++++++++++++++----------
 1 file changed, 43 insertions(+), 17 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/user_sdma.c b/drivers/infiniband/hw/hfi1/user_sdma.c
index a761f804111e..663980ef01a8 100644
--- a/drivers/infiniband/hw/hfi1/user_sdma.c
+++ b/drivers/infiniband/hw/hfi1/user_sdma.c
@@ -115,6 +115,7 @@ MODULE_PARM_DESC(sdma_comp_size, "Size of User SDMA completion ring. Default: 12
 #define KDETH_HCRC_LOWER_MASK     0xff
 
 #define AHG_KDETH_INTR_SHIFT 12
+#define AHG_KDETH_SH_SHIFT   13
 
 #define PBC2LRH(x) ((((x) & 0xfff) << 2) - 4)
 #define LRH2PBC(x) ((((x) >> 2) + 1) & 0xfff)
@@ -144,8 +145,9 @@ MODULE_PARM_DESC(sdma_comp_size, "Size of User SDMA completion ring. Default: 12
 #define KDETH_OM_LARGE     64
 #define KDETH_OM_MAX_SIZE  (1 << ((KDETH_OM_LARGE / KDETH_OM_SMALL) + 1))
 
-/* Last packet in the request */
-#define TXREQ_FLAGS_REQ_LAST_PKT BIT(0)
+/* Tx request flag bits */
+#define TXREQ_FLAGS_REQ_ACK   BIT(0)      /* Set the ACK bit in the header */
+#define TXREQ_FLAGS_REQ_DISABLE_SH BIT(1) /* Disable header suppression */
 
 /* SDMA request flag bits */
 #define SDMA_REQ_FOR_THREAD 1
@@ -943,8 +945,13 @@ static int user_sdma_send_pkts(struct user_sdma_request *req, unsigned maxpkts)
 		tx->busycount = 0;
 		INIT_LIST_HEAD(&tx->list);
 
+		/*
+		 * For the last packet set the ACK request
+		 * and disable header suppression.
+		 */
 		if (req->seqnum == req->info.npkts - 1)
-			tx->flags |= TXREQ_FLAGS_REQ_LAST_PKT;
+			tx->flags |= (TXREQ_FLAGS_REQ_ACK |
+				      TXREQ_FLAGS_REQ_DISABLE_SH);
 
 		/*
 		 * Calculate the payload size - this is min of the fragment
@@ -963,11 +970,22 @@ static int user_sdma_send_pkts(struct user_sdma_request *req, unsigned maxpkts)
 			}
 
 			datalen = compute_data_length(req, tx);
+
+			/*
+			 * Disable header suppression for the payload <= 8DWS.
+			 * If there is an uncorrectable error in the receive
+			 * data FIFO when the received payload size is less than
+			 * or equal to 8DWS then the RxDmaDataFifoRdUncErr is
+			 * not reported.There is set RHF.EccErr if the header
+			 * is not suppressed.
+			 */
 			if (!datalen) {
 				SDMA_DBG(req,
 					 "Request has data but pkt len is 0");
 				ret = -EFAULT;
 				goto free_tx;
+			} else if (datalen <= 32) {
+				tx->flags |= TXREQ_FLAGS_REQ_DISABLE_SH;
 			}
 		}
 
@@ -990,6 +1008,10 @@ static int user_sdma_send_pkts(struct user_sdma_request *req, unsigned maxpkts)
 						LRH2PBC(lrhlen);
 					tx->hdr.pbc[0] = cpu_to_le16(pbclen);
 				}
+				ret = check_header_template(req, &tx->hdr,
+							    lrhlen, datalen);
+				if (ret)
+					goto free_tx;
 				ret = sdma_txinit_ahg(&tx->txreq,
 						      SDMA_TXREQ_F_AHG_COPY,
 						      sizeof(tx->hdr) + datalen,
@@ -1351,7 +1373,7 @@ static int set_txreq_header(struct user_sdma_request *req,
 				req->seqnum));
 
 	/* Set ACK request on last packet */
-	if (unlikely(tx->flags & TXREQ_FLAGS_REQ_LAST_PKT))
+	if (unlikely(tx->flags & TXREQ_FLAGS_REQ_ACK))
 		hdr->bth[2] |= cpu_to_be32(1UL << 31);
 
 	/* Set the new offset */
@@ -1384,8 +1406,8 @@ static int set_txreq_header(struct user_sdma_request *req,
 		/* Set KDETH.TID based on value for this TID */
 		KDETH_SET(hdr->kdeth.ver_tid_offset, TID,
 			  EXP_TID_GET(tidval, IDX));
-		/* Clear KDETH.SH only on the last packet */
-		if (unlikely(tx->flags & TXREQ_FLAGS_REQ_LAST_PKT))
+		/* Clear KDETH.SH when DISABLE_SH flag is set */
+		if (unlikely(tx->flags & TXREQ_FLAGS_REQ_DISABLE_SH))
 			KDETH_SET(hdr->kdeth.ver_tid_offset, SH, 0);
 		/*
 		 * Set the KDETH.OFFSET and KDETH.OM based on size of
@@ -1429,7 +1451,7 @@ static int set_txreq_header_ahg(struct user_sdma_request *req,
 	/* BTH.PSN and BTH.A */
 	val32 = (be32_to_cpu(hdr->bth[2]) + req->seqnum) &
 		(HFI1_CAP_IS_KSET(EXTENDED_PSN) ? 0x7fffffff : 0xffffff);
-	if (unlikely(tx->flags & TXREQ_FLAGS_REQ_LAST_PKT))
+	if (unlikely(tx->flags & TXREQ_FLAGS_REQ_ACK))
 		val32 |= 1UL << 31;
 	AHG_HEADER_SET(req->ahg, diff, 6, 0, 16, cpu_to_be16(val32 >> 16));
 	AHG_HEADER_SET(req->ahg, diff, 6, 16, 16, cpu_to_be16(val32 & 0xffff));
@@ -1468,19 +1490,23 @@ static int set_txreq_header_ahg(struct user_sdma_request *req,
 		AHG_HEADER_SET(req->ahg, diff, 7, 0, 16,
 			       ((!!(req->omfactor - KDETH_OM_SMALL)) << 15 |
 				((req->tidoffset / req->omfactor) & 0x7fff)));
-		/* KDETH.TIDCtrl, KDETH.TID */
+		/* KDETH.TIDCtrl, KDETH.TID, KDETH.Intr, KDETH.SH */
 		val = cpu_to_le16(((EXP_TID_GET(tidval, CTRL) & 0x3) << 10) |
-					(EXP_TID_GET(tidval, IDX) & 0x3ff));
-		/* Clear KDETH.SH on last packet */
-		if (unlikely(tx->flags & TXREQ_FLAGS_REQ_LAST_PKT)) {
-			val |= cpu_to_le16(KDETH_GET(hdr->kdeth.ver_tid_offset,
-						     INTR) <<
-					   AHG_KDETH_INTR_SHIFT);
-			val &= cpu_to_le16(~(1U << 13));
-			AHG_HEADER_SET(req->ahg, diff, 7, 16, 14, val);
+				   (EXP_TID_GET(tidval, IDX) & 0x3ff));
+
+		if (unlikely(tx->flags & TXREQ_FLAGS_REQ_DISABLE_SH)) {
+			val |= cpu_to_le16((KDETH_GET(hdr->kdeth.ver_tid_offset,
+						      INTR) <<
+					    AHG_KDETH_INTR_SHIFT));
 		} else {
-			AHG_HEADER_SET(req->ahg, diff, 7, 16, 12, val);
+			val |= KDETH_GET(hdr->kdeth.ver_tid_offset, SH) ?
+			       cpu_to_le16(0x1 << AHG_KDETH_SH_SHIFT) :
+			       cpu_to_le16((KDETH_GET(hdr->kdeth.ver_tid_offset,
+						      INTR) <<
+					     AHG_KDETH_INTR_SHIFT));
 		}
+
+		AHG_HEADER_SET(req->ahg, diff, 7, 16, 14, val);
 	}
 
 	trace_hfi1_sdma_user_header_ahg(pq->dd, pq->ctxt, pq->subctxt,
-- 
cgit v1.2.1


From cf4c2f8c9d0483f7585883d3f6733ebf09ea837c Mon Sep 17 00:00:00 2001
From: Dennis Dalessandro <dennis.dalessandro@intel.com>
Date: Wed, 7 Dec 2016 19:32:47 -0800
Subject: IB/rdmavt: Fix trace hierarchy

Split rdmavt traces into separate files to preserve the original
hierarchy since only one trace sub system may now be defined per header
file.

Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rdmavt/trace.h     | 140 +------------------------------
 drivers/infiniband/sw/rdmavt/trace_qp.h  |  96 +++++++++++++++++++++
 drivers/infiniband/sw/rdmavt/trace_rvt.h |  81 ++++++++++++++++++
 drivers/infiniband/sw/rdmavt/trace_tx.h  | 132 +++++++++++++++++++++++++++++
 4 files changed, 312 insertions(+), 137 deletions(-)
 create mode 100644 drivers/infiniband/sw/rdmavt/trace_qp.h
 create mode 100644 drivers/infiniband/sw/rdmavt/trace_rvt.h
 create mode 100644 drivers/infiniband/sw/rdmavt/trace_tx.h

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rdmavt/trace.h b/drivers/infiniband/sw/rdmavt/trace.h
index 6c0457db5499..11a2afbc8e76 100644
--- a/drivers/infiniband/sw/rdmavt/trace.h
+++ b/drivers/infiniband/sw/rdmavt/trace.h
@@ -45,143 +45,9 @@
  *
  */
 
-#undef TRACE_SYSTEM_VAR
-#define TRACE_SYSTEM_VAR rdmavt
-
-#if !defined(__RDMAVT_TRACE_H) || defined(TRACE_HEADER_MULTI_READ)
-#define __RDMAVT_TRACE_H
-
-#include <linux/tracepoint.h>
-#include <linux/trace_seq.h>
-
-#include <rdma/ib_verbs.h>
-#include <rdma/rdma_vt.h>
-
 #define RDI_DEV_ENTRY(rdi)   __string(dev, rdi->driver_f.get_card_name(rdi))
 #define RDI_DEV_ASSIGN(rdi)  __assign_str(dev, rdi->driver_f.get_card_name(rdi))
 
-#undef TRACE_SYSTEM
-#define TRACE_SYSTEM rdmavt
-
-TRACE_EVENT(rvt_dbg,
-	TP_PROTO(struct rvt_dev_info *rdi,
-		 const char *msg),
-	TP_ARGS(rdi, msg),
-	TP_STRUCT__entry(
-		RDI_DEV_ENTRY(rdi)
-		__string(msg, msg)
-	),
-	TP_fast_assign(
-		RDI_DEV_ASSIGN(rdi);
-		__assign_str(msg, msg);
-	),
-	TP_printk("[%s]: %s", __get_str(dev), __get_str(msg))
-);
-
-#undef TRACE_SYSTEM
-#define TRACE_SYSTEM rvt_qphash
-DECLARE_EVENT_CLASS(rvt_qphash_template,
-	TP_PROTO(struct rvt_qp *qp, u32 bucket),
-	TP_ARGS(qp, bucket),
-	TP_STRUCT__entry(
-		RDI_DEV_ENTRY(ib_to_rvt(qp->ibqp.device))
-		__field(u32, qpn)
-		__field(u32, bucket)
-	),
-	TP_fast_assign(
-		RDI_DEV_ASSIGN(ib_to_rvt(qp->ibqp.device))
-		__entry->qpn = qp->ibqp.qp_num;
-		__entry->bucket = bucket;
-	),
-	TP_printk(
-		"[%s] qpn 0x%x bucket %u",
-		__get_str(dev),
-		__entry->qpn,
-		__entry->bucket
-	)
-);
-
-DEFINE_EVENT(rvt_qphash_template, rvt_qpinsert,
-	TP_PROTO(struct rvt_qp *qp, u32 bucket),
-	TP_ARGS(qp, bucket));
-
-DEFINE_EVENT(rvt_qphash_template, rvt_qpremove,
-	TP_PROTO(struct rvt_qp *qp, u32 bucket),
-	TP_ARGS(qp, bucket));
-
-#undef TRACE_SYSTEM
-#define TRACE_SYSTEM rvt_tx
-
-#define wr_opcode_name(opcode) { IB_WR_##opcode, #opcode  }
-#define show_wr_opcode(opcode)                             \
-__print_symbolic(opcode,                                   \
-	wr_opcode_name(RDMA_WRITE),                        \
-	wr_opcode_name(RDMA_WRITE_WITH_IMM),               \
-	wr_opcode_name(SEND),                              \
-	wr_opcode_name(SEND_WITH_IMM),                     \
-	wr_opcode_name(RDMA_READ),                         \
-	wr_opcode_name(ATOMIC_CMP_AND_SWP),                \
-	wr_opcode_name(ATOMIC_FETCH_AND_ADD),              \
-	wr_opcode_name(LSO),                               \
-	wr_opcode_name(SEND_WITH_INV),                     \
-	wr_opcode_name(RDMA_READ_WITH_INV),                \
-	wr_opcode_name(LOCAL_INV),                         \
-	wr_opcode_name(MASKED_ATOMIC_CMP_AND_SWP),         \
-	wr_opcode_name(MASKED_ATOMIC_FETCH_AND_ADD))
-
-#define POS_PRN \
-"[%s] wr_id %llx qpn %x psn 0x%x lpsn 0x%x length %u opcode 0x%.2x,%s size %u avail %u head %u last %u"
-
-TRACE_EVENT(
-	rvt_post_one_wr,
-	TP_PROTO(struct rvt_qp *qp, struct rvt_swqe *wqe),
-	TP_ARGS(qp, wqe),
-	TP_STRUCT__entry(
-		RDI_DEV_ENTRY(ib_to_rvt(qp->ibqp.device))
-		__field(u64, wr_id)
-		__field(u32, qpn)
-		__field(u32, psn)
-		__field(u32, lpsn)
-		__field(u32, length)
-		__field(u32, opcode)
-		__field(u32, size)
-		__field(u32, avail)
-		__field(u32, head)
-		__field(u32, last)
-	),
-	TP_fast_assign(
-		RDI_DEV_ASSIGN(ib_to_rvt(qp->ibqp.device))
-		__entry->wr_id = wqe->wr.wr_id;
-		__entry->qpn = qp->ibqp.qp_num;
-		__entry->psn = wqe->psn;
-		__entry->lpsn = wqe->lpsn;
-		__entry->length = wqe->length;
-		__entry->opcode = wqe->wr.opcode;
-		__entry->size = qp->s_size;
-		__entry->avail = qp->s_avail;
-		__entry->head = qp->s_head;
-		__entry->last = qp->s_last;
-	),
-	TP_printk(
-		POS_PRN,
-		__get_str(dev),
-		__entry->wr_id,
-		__entry->qpn,
-		__entry->psn,
-		__entry->lpsn,
-		__entry->length,
-		__entry->opcode, show_wr_opcode(__entry->opcode),
-		__entry->size,
-		__entry->avail,
-		__entry->head,
-		__entry->last
-	)
-);
-
-#endif /* __RDMAVT_TRACE_H */
-
-#undef TRACE_INCLUDE_PATH
-#undef TRACE_INCLUDE_FILE
-#define TRACE_INCLUDE_PATH .
-#define TRACE_INCLUDE_FILE trace
-#include <trace/define_trace.h>
+#include "trace_rvt.h"
+#include "trace_qp.h"
+#include "trace_tx.h"
diff --git a/drivers/infiniband/sw/rdmavt/trace_qp.h b/drivers/infiniband/sw/rdmavt/trace_qp.h
new file mode 100644
index 000000000000..4c77a3119bda
--- /dev/null
+++ b/drivers/infiniband/sw/rdmavt/trace_qp.h
@@ -0,0 +1,96 @@
+/*
+ * Copyright(c) 2016 Intel Corporation.
+ *
+ * This file is provided under a dual BSD/GPLv2 license.  When using or
+ * redistributing this file, you may do so under either license.
+ *
+ * GPL LICENSE SUMMARY
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of version 2 of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * BSD LICENSE
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ *  - Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ *  - Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in
+ *    the documentation and/or other materials provided with the
+ *    distribution.
+ *  - Neither the name of Intel Corporation nor the names of its
+ *    contributors may be used to endorse or promote products derived
+ *    from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ */
+#if !defined(__RVT_TRACE_QP_H) || defined(TRACE_HEADER_MULTI_READ)
+#define __RVT_TRACE_QP_H
+
+#include <linux/tracepoint.h>
+#include <linux/trace_seq.h>
+
+#include <rdma/ib_verbs.h>
+#include <rdma/rdma_vt.h>
+
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM rvt_qp
+
+DECLARE_EVENT_CLASS(rvt_qphash_template,
+	TP_PROTO(struct rvt_qp *qp, u32 bucket),
+	TP_ARGS(qp, bucket),
+	TP_STRUCT__entry(
+		RDI_DEV_ENTRY(ib_to_rvt(qp->ibqp.device))
+		__field(u32, qpn)
+		__field(u32, bucket)
+	),
+	TP_fast_assign(
+		RDI_DEV_ASSIGN(ib_to_rvt(qp->ibqp.device))
+		__entry->qpn = qp->ibqp.qp_num;
+		__entry->bucket = bucket;
+	),
+	TP_printk(
+		"[%s] qpn 0x%x bucket %u",
+		__get_str(dev),
+		__entry->qpn,
+		__entry->bucket
+	)
+);
+
+DEFINE_EVENT(rvt_qphash_template, rvt_qpinsert,
+	TP_PROTO(struct rvt_qp *qp, u32 bucket),
+	TP_ARGS(qp, bucket));
+
+DEFINE_EVENT(rvt_qphash_template, rvt_qpremove,
+	TP_PROTO(struct rvt_qp *qp, u32 bucket),
+	TP_ARGS(qp, bucket));
+
+
+#endif /* __RVT_TRACE_QP_H */
+
+#undef TRACE_INCLUDE_PATH
+#undef TRACE_INCLUDE_FILE
+#define TRACE_INCLUDE_PATH .
+#define TRACE_INCLUDE_FILE trace_qp
+#include <trace/define_trace.h>
+
diff --git a/drivers/infiniband/sw/rdmavt/trace_rvt.h b/drivers/infiniband/sw/rdmavt/trace_rvt.h
new file mode 100644
index 000000000000..746f33461d9a
--- /dev/null
+++ b/drivers/infiniband/sw/rdmavt/trace_rvt.h
@@ -0,0 +1,81 @@
+/*
+ * Copyright(c) 2016 Intel Corporation.
+ *
+ * This file is provided under a dual BSD/GPLv2 license.  When using or
+ * redistributing this file, you may do so under either license.
+ *
+ * GPL LICENSE SUMMARY
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of version 2 of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * BSD LICENSE
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ *  - Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ *  - Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in
+ *    the documentation and/or other materials provided with the
+ *    distribution.
+ *  - Neither the name of Intel Corporation nor the names of its
+ *    contributors may be used to endorse or promote products derived
+ *    from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ */
+#if !defined(__RVT_TRACE_RVT_H) || defined(TRACE_HEADER_MULTI_READ)
+#define __RVT_TRACE_RVT_H
+
+#include <linux/tracepoint.h>
+#include <linux/trace_seq.h>
+
+#include <rdma/ib_verbs.h>
+#include <rdma/rdma_vt.h>
+
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM rvt
+
+TRACE_EVENT(rvt_dbg,
+	TP_PROTO(struct rvt_dev_info *rdi,
+		 const char *msg),
+	TP_ARGS(rdi, msg),
+	TP_STRUCT__entry(
+		RDI_DEV_ENTRY(rdi)
+		__string(msg, msg)
+	),
+	TP_fast_assign(
+		RDI_DEV_ASSIGN(rdi);
+		__assign_str(msg, msg);
+	),
+	TP_printk("[%s]: %s", __get_str(dev), __get_str(msg))
+);
+
+#endif /* __RVT_TRACE_MISC_H */
+
+#undef TRACE_INCLUDE_PATH
+#undef TRACE_INCLUDE_FILE
+#define TRACE_INCLUDE_PATH .
+#define TRACE_INCLUDE_FILE trace_rvt
+#include <trace/define_trace.h>
+
diff --git a/drivers/infiniband/sw/rdmavt/trace_tx.h b/drivers/infiniband/sw/rdmavt/trace_tx.h
new file mode 100644
index 000000000000..0e03173662d8
--- /dev/null
+++ b/drivers/infiniband/sw/rdmavt/trace_tx.h
@@ -0,0 +1,132 @@
+/*
+ * Copyright(c) 2016 Intel Corporation.
+ *
+ * This file is provided under a dual BSD/GPLv2 license.  When using or
+ * redistributing this file, you may do so under either license.
+ *
+ * GPL LICENSE SUMMARY
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of version 2 of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * BSD LICENSE
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ *  - Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ *  - Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in
+ *    the documentation and/or other materials provided with the
+ *    distribution.
+ *  - Neither the name of Intel Corporation nor the names of its
+ *    contributors may be used to endorse or promote products derived
+ *    from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ */
+#if !defined(__RVT_TRACE_TX_H) || defined(TRACE_HEADER_MULTI_READ)
+#define __RVT_TRACE_TX_H
+
+#include <linux/tracepoint.h>
+#include <linux/trace_seq.h>
+
+#include <rdma/ib_verbs.h>
+#include <rdma/rdma_vt.h>
+
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM rvt_tx
+
+#define wr_opcode_name(opcode) { IB_WR_##opcode, #opcode  }
+#define show_wr_opcode(opcode)                             \
+__print_symbolic(opcode,                                   \
+	wr_opcode_name(RDMA_WRITE),                        \
+	wr_opcode_name(RDMA_WRITE_WITH_IMM),               \
+	wr_opcode_name(SEND),                              \
+	wr_opcode_name(SEND_WITH_IMM),                     \
+	wr_opcode_name(RDMA_READ),                         \
+	wr_opcode_name(ATOMIC_CMP_AND_SWP),                \
+	wr_opcode_name(ATOMIC_FETCH_AND_ADD),              \
+	wr_opcode_name(LSO),                               \
+	wr_opcode_name(SEND_WITH_INV),                     \
+	wr_opcode_name(RDMA_READ_WITH_INV),                \
+	wr_opcode_name(LOCAL_INV),                         \
+	wr_opcode_name(MASKED_ATOMIC_CMP_AND_SWP),         \
+	wr_opcode_name(MASKED_ATOMIC_FETCH_AND_ADD))
+
+#define POS_PRN \
+"[%s] wr_id %llx qpn %x psn 0x%x lpsn 0x%x length %u opcode 0x%.2x,%s size %u avail %u head %u last %u"
+
+TRACE_EVENT(
+	rvt_post_one_wr,
+	TP_PROTO(struct rvt_qp *qp, struct rvt_swqe *wqe),
+	TP_ARGS(qp, wqe),
+	TP_STRUCT__entry(
+		RDI_DEV_ENTRY(ib_to_rvt(qp->ibqp.device))
+		__field(u64, wr_id)
+		__field(u32, qpn)
+		__field(u32, psn)
+		__field(u32, lpsn)
+		__field(u32, length)
+		__field(u32, opcode)
+		__field(u32, size)
+		__field(u32, avail)
+		__field(u32, head)
+		__field(u32, last)
+	),
+	TP_fast_assign(
+		RDI_DEV_ASSIGN(ib_to_rvt(qp->ibqp.device))
+		__entry->wr_id = wqe->wr.wr_id;
+		__entry->qpn = qp->ibqp.qp_num;
+		__entry->psn = wqe->psn;
+		__entry->lpsn = wqe->lpsn;
+		__entry->length = wqe->length;
+		__entry->opcode = wqe->wr.opcode;
+		__entry->size = qp->s_size;
+		__entry->avail = qp->s_avail;
+		__entry->head = qp->s_head;
+		__entry->last = qp->s_last;
+	),
+	TP_printk(
+		POS_PRN,
+		__get_str(dev),
+		__entry->wr_id,
+		__entry->qpn,
+		__entry->psn,
+		__entry->lpsn,
+		__entry->length,
+		__entry->opcode, show_wr_opcode(__entry->opcode),
+		__entry->size,
+		__entry->avail,
+		__entry->head,
+		__entry->last
+	)
+);
+
+#endif /* __RVT_TRACE_TX_H */
+
+#undef TRACE_INCLUDE_PATH
+#undef TRACE_INCLUDE_FILE
+#define TRACE_INCLUDE_PATH .
+#define TRACE_INCLUDE_FILE trace_tx
+#include <trace/define_trace.h>
+
-- 
cgit v1.2.1


From b7481944b06e99dc84f4c7da2681ac89528b2020 Mon Sep 17 00:00:00 2001
From: Jianxin Xiong <jianxin.xiong@intel.com>
Date: Wed, 7 Dec 2016 19:32:53 -0800
Subject: IB/hfi1: Show statistics counters under IB stats interface

Previously tools like hfi1stats had to access these counters through
debugfs, which often caused permission issue for non-root users. It is
not always acceptable to change the debugfs mounting permission due
to security concerns. When exposed under the IB stats interface, the
counters are universally readable by default.

Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Signed-off-by: Jianxin Xiong <jianxin.xiong@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/verbs.c | 154 +++++++++++++++++++++++++++++++++++++
 1 file changed, 154 insertions(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/verbs.c b/drivers/infiniband/hw/hfi1/verbs.c
index f1dead369258..09132582e09c 100644
--- a/drivers/infiniband/hw/hfi1/verbs.c
+++ b/drivers/infiniband/hw/hfi1/verbs.c
@@ -1609,6 +1609,154 @@ static void hfi1_get_dev_fw_str(struct ib_device *ibdev, char *str,
 		 dc8051_ver_min(ver));
 }
 
+static const char * const driver_cntr_names[] = {
+	/* must be element 0*/
+	"DRIVER_KernIntr",
+	"DRIVER_ErrorIntr",
+	"DRIVER_Tx_Errs",
+	"DRIVER_Rcv_Errs",
+	"DRIVER_HW_Errs",
+	"DRIVER_NoPIOBufs",
+	"DRIVER_CtxtsOpen",
+	"DRIVER_RcvLen_Errs",
+	"DRIVER_EgrBufFull",
+	"DRIVER_EgrHdrFull"
+};
+
+static const char **dev_cntr_names;
+static const char **port_cntr_names;
+static int num_driver_cntrs = ARRAY_SIZE(driver_cntr_names);
+static int num_dev_cntrs;
+static int num_port_cntrs;
+static int cntr_names_initialized;
+
+/*
+ * Convert a list of names separated by '\n' into an array of NULL terminated
+ * strings. Optionally some entries can be reserved in the array to hold extra
+ * external strings.
+ */
+static int init_cntr_names(const char *names_in,
+			   const int names_len,
+			   int num_extra_names,
+			   int *num_cntrs,
+			   const char ***cntr_names)
+{
+	char *names_out, *p, **q;
+	int i, n;
+
+	n = 0;
+	for (i = 0; i < names_len; i++)
+		if (names_in[i] == '\n')
+			n++;
+
+	names_out = kmalloc((n + num_extra_names) * sizeof(char *) + names_len,
+			    GFP_KERNEL);
+	if (!names_out) {
+		*num_cntrs = 0;
+		*cntr_names = NULL;
+		return -ENOMEM;
+	}
+
+	p = names_out + (n + num_extra_names) * sizeof(char *);
+	memcpy(p, names_in, names_len);
+
+	q = (char **)names_out;
+	for (i = 0; i < n; i++) {
+		q[i] = p;
+		p = strchr(p, '\n');
+		*p++ = '\0';
+	}
+
+	*num_cntrs = n;
+	*cntr_names = (const char **)names_out;
+	return 0;
+}
+
+static struct rdma_hw_stats *alloc_hw_stats(struct ib_device *ibdev,
+					    u8 port_num)
+{
+	int i, err;
+
+	if (!cntr_names_initialized) {
+		struct hfi1_devdata *dd = dd_from_ibdev(ibdev);
+
+		err = init_cntr_names(dd->cntrnames,
+				      dd->cntrnameslen,
+				      num_driver_cntrs,
+				      &num_dev_cntrs,
+				      &dev_cntr_names);
+		if (err)
+			return NULL;
+
+		for (i = 0; i < num_driver_cntrs; i++)
+			dev_cntr_names[num_dev_cntrs + i] =
+				driver_cntr_names[i];
+
+		err = init_cntr_names(dd->portcntrnames,
+				      dd->portcntrnameslen,
+				      0,
+				      &num_port_cntrs,
+				      &port_cntr_names);
+		if (err) {
+			kfree(dev_cntr_names);
+			dev_cntr_names = NULL;
+			return NULL;
+		}
+		cntr_names_initialized = 1;
+	}
+
+	if (!port_num)
+		return rdma_alloc_hw_stats_struct(
+				dev_cntr_names,
+				num_dev_cntrs + num_driver_cntrs,
+				RDMA_HW_STATS_DEFAULT_LIFESPAN);
+	else
+		return rdma_alloc_hw_stats_struct(
+				port_cntr_names,
+				num_port_cntrs,
+				RDMA_HW_STATS_DEFAULT_LIFESPAN);
+}
+
+static u64 hfi1_sps_ints(void)
+{
+	unsigned long flags;
+	struct hfi1_devdata *dd;
+	u64 sps_ints = 0;
+
+	spin_lock_irqsave(&hfi1_devs_lock, flags);
+	list_for_each_entry(dd, &hfi1_dev_list, list) {
+		sps_ints += get_all_cpu_total(dd->int_counter);
+	}
+	spin_unlock_irqrestore(&hfi1_devs_lock, flags);
+	return sps_ints;
+}
+
+static int get_hw_stats(struct ib_device *ibdev, struct rdma_hw_stats *stats,
+			u8 port, int index)
+{
+	u64 *values;
+	int count;
+
+	if (!port) {
+		u64 *stats = (u64 *)&hfi1_stats;
+		int i;
+
+		hfi1_read_cntrs(dd_from_ibdev(ibdev), NULL, &values);
+		values[num_dev_cntrs] = hfi1_sps_ints();
+		for (i = 1; i < num_driver_cntrs; i++)
+			values[num_dev_cntrs + i] = stats[i];
+		count = num_dev_cntrs + num_driver_cntrs;
+	} else {
+		struct hfi1_ibport *ibp = to_iport(ibdev, port);
+
+		hfi1_read_portcntrs(ppd_from_ibp(ibp), NULL, &values);
+		count = num_port_cntrs;
+	}
+
+	memcpy(stats->value, values, count * sizeof(u64));
+	return count;
+}
+
 /**
  * hfi1_register_ib_device - register our device with the infiniband core
  * @dd: the device data structure
@@ -1656,6 +1804,8 @@ int hfi1_register_ib_device(struct hfi1_devdata *dd)
 	ibdev->phys_port_cnt = dd->num_pports;
 	ibdev->dma_device = &dd->pcidev->dev;
 	ibdev->modify_device = modify_device;
+	ibdev->alloc_hw_stats = alloc_hw_stats;
+	ibdev->get_hw_stats = get_hw_stats;
 
 	/* keep process mad in the driver */
 	ibdev->process_mad = hfi1_process_mad;
@@ -1770,6 +1920,10 @@ void hfi1_unregister_ib_device(struct hfi1_devdata *dd)
 
 	del_timer_sync(&dev->mem_timer);
 	verbs_txreq_exit(dev);
+
+	kfree(dev_cntr_names);
+	kfree(port_cntr_names);
+	cntr_names_initialized = 0;
 }
 
 void hfi1_cnp_rcv(struct hfi1_packet *packet)
-- 
cgit v1.2.1


From e922ae06e90a37ab0b212f844e8aed9b6021cf21 Mon Sep 17 00:00:00 2001
From: Don Hiatt <don.hiatt@intel.com>
Date: Wed, 7 Dec 2016 19:33:00 -0800
Subject: IB/hfi1: Remove dependence on qp->s_cur_size

The qp->s_cur_size field assumes that the S_BUSY bit protects
the field from modification after the slock is dropped. Scaling the
send engine to multiple cores would break that assumption.

Correct the issue by carrying the payload size in the txreq structure.

Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Don Hiatt <don.hiatt@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/rc.c          | 4 ++--
 drivers/infiniband/hw/hfi1/ruc.c         | 4 ++--
 drivers/infiniband/hw/hfi1/uc.c          | 2 +-
 drivers/infiniband/hw/hfi1/ud.c          | 2 +-
 drivers/infiniband/hw/hfi1/verbs.c       | 6 +++---
 drivers/infiniband/hw/hfi1/verbs_txreq.h | 1 +
 6 files changed, 10 insertions(+), 9 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/rc.c b/drivers/infiniband/hw/hfi1/rc.c
index aaca8588e199..e69161e007db 100644
--- a/drivers/infiniband/hw/hfi1/rc.c
+++ b/drivers/infiniband/hw/hfi1/rc.c
@@ -351,7 +351,7 @@ normal:
 	qp->s_rdma_ack_cnt++;
 	qp->s_hdrwords = hwords;
 	ps->s_txreq->sde = priv->s_sde;
-	qp->s_cur_size = len;
+	ps->s_txreq->s_cur_size = len;
 	hfi1_make_ruc_header(qp, ohdr, bth0, bth2, middle, ps);
 	/* pbc */
 	ps->s_txreq->hdr_dwords = qp->s_hdrwords + 2;
@@ -802,7 +802,7 @@ int hfi1_make_rc_req(struct rvt_qp *qp, struct hfi1_pkt_state *ps)
 	qp->s_hdrwords = hwords;
 	ps->s_txreq->sde = priv->s_sde;
 	qp->s_cur_sge = ss;
-	qp->s_cur_size = len;
+	ps->s_txreq->s_cur_size = len;
 	hfi1_make_ruc_header(
 		qp,
 		ohdr,
diff --git a/drivers/infiniband/hw/hfi1/ruc.c b/drivers/infiniband/hw/hfi1/ruc.c
index 5a70e91b5191..2ec3e908691b 100644
--- a/drivers/infiniband/hw/hfi1/ruc.c
+++ b/drivers/infiniband/hw/hfi1/ruc.c
@@ -767,8 +767,8 @@ void hfi1_make_ruc_header(struct rvt_qp *qp, struct ib_other_headers *ohdr,
 	u32 bth1;
 
 	/* Construct the header. */
-	extra_bytes = -qp->s_cur_size & 3;
-	nwords = (qp->s_cur_size + extra_bytes) >> 2;
+	extra_bytes = -ps->s_txreq->s_cur_size & 3;
+	nwords = (ps->s_txreq->s_cur_size + extra_bytes) >> 2;
 	lrh0 = HFI1_LRH_BTH;
 	if (unlikely(qp->remote_ah_attr.ah_flags & IB_AH_GRH)) {
 		qp->s_hdrwords += hfi1_make_grh(ibp,
diff --git a/drivers/infiniband/hw/hfi1/uc.c b/drivers/infiniband/hw/hfi1/uc.c
index 5e6d1bac4914..d062c3537f7c 100644
--- a/drivers/infiniband/hw/hfi1/uc.c
+++ b/drivers/infiniband/hw/hfi1/uc.c
@@ -259,7 +259,7 @@ int hfi1_make_uc_req(struct rvt_qp *qp, struct hfi1_pkt_state *ps)
 	qp->s_hdrwords = hwords;
 	ps->s_txreq->sde = priv->s_sde;
 	qp->s_cur_sge = &qp->s_sge;
-	qp->s_cur_size = len;
+	ps->s_txreq->s_cur_size = len;
 	hfi1_make_ruc_header(qp, ohdr, bth0 | (qp->s_state << 24),
 			     mask_psn(qp->s_psn++), middle, ps);
 	/* pbc */
diff --git a/drivers/infiniband/hw/hfi1/ud.c b/drivers/infiniband/hw/hfi1/ud.c
index 97ae24b6314c..d742ac1af947 100644
--- a/drivers/infiniband/hw/hfi1/ud.c
+++ b/drivers/infiniband/hw/hfi1/ud.c
@@ -354,7 +354,7 @@ int hfi1_make_ud_req(struct rvt_qp *qp, struct hfi1_pkt_state *ps)
 
 	/* header size in 32-bit words LRH+BTH+DETH = (8+12+8)/4. */
 	qp->s_hdrwords = 7;
-	qp->s_cur_size = wqe->length;
+	ps->s_txreq->s_cur_size = wqe->length;
 	qp->s_cur_sge = &qp->s_sge;
 	qp->s_srate = ah_attr->static_rate;
 	qp->srate_mbps = ib_rate_to_mbps(qp->s_srate);
diff --git a/drivers/infiniband/hw/hfi1/verbs.c b/drivers/infiniband/hw/hfi1/verbs.c
index 09132582e09c..65f2f1d76fc8 100644
--- a/drivers/infiniband/hw/hfi1/verbs.c
+++ b/drivers/infiniband/hw/hfi1/verbs.c
@@ -895,7 +895,7 @@ int hfi1_verbs_send_dma(struct rvt_qp *qp, struct hfi1_pkt_state *ps,
 	struct hfi1_ahg_info *ahg_info = priv->s_ahg;
 	u32 hdrwords = qp->s_hdrwords;
 	struct rvt_sge_state *ss = qp->s_cur_sge;
-	u32 len = qp->s_cur_size;
+	u32 len = ps->s_txreq->s_cur_size;
 	u32 plen = hdrwords + ((len + 3) >> 2) + 2; /* includes pbc */
 	struct hfi1_ibdev *dev = ps->dev;
 	struct hfi1_pportdata *ppd = ps->ppd;
@@ -1012,7 +1012,7 @@ int hfi1_verbs_send_pio(struct rvt_qp *qp, struct hfi1_pkt_state *ps,
 	struct hfi1_qp_priv *priv = qp->priv;
 	u32 hdrwords = qp->s_hdrwords;
 	struct rvt_sge_state *ss = qp->s_cur_sge;
-	u32 len = qp->s_cur_size;
+	u32 len = ps->s_txreq->s_cur_size;
 	u32 dwords = (len + 3) >> 2;
 	u32 plen = hdrwords + dwords + 2; /* includes pbc */
 	struct hfi1_pportdata *ppd = ps->ppd;
@@ -1240,7 +1240,7 @@ static inline send_routine get_send_routine(struct rvt_qp *qp,
 		u8 op = get_opcode(h);
 
 		if (piothreshold &&
-		    qp->s_cur_size <= min(piothreshold, qp->pmtu) &&
+		    tx->s_cur_size <= min(piothreshold, qp->pmtu) &&
 		    (BIT(op & OPMASK) & pio_opmask[op >> 5]) &&
 		    iowait_sdma_pending(&priv->s_iowait) == 0 &&
 		    !sdma_txreq_built(&tx->txreq))
diff --git a/drivers/infiniband/hw/hfi1/verbs_txreq.h b/drivers/infiniband/hw/hfi1/verbs_txreq.h
index 5660897593ba..76216f2ef35a 100644
--- a/drivers/infiniband/hw/hfi1/verbs_txreq.h
+++ b/drivers/infiniband/hw/hfi1/verbs_txreq.h
@@ -65,6 +65,7 @@ struct verbs_txreq {
 	struct sdma_engine     *sde;
 	struct send_context     *psc;
 	u16                     hdr_dwords;
+	u16			s_cur_size;
 };
 
 struct hfi1_ibdev;
-- 
cgit v1.2.1


From 6e40b59cfa99710fcbf849439783e8fc98d183bf Mon Sep 17 00:00:00 2001
From: Tadeusz Struk <tadeusz.struk@intel.com>
Date: Wed, 7 Dec 2016 19:33:06 -0800
Subject: IB/hfi1: Remove definition of unused hfi1_affinity struct

The struct hfi1_affinity is not used anymore.
We use the struct hfi1_affinity_node and hfi1_affinity_node_list
instead.

Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Tadeusz Struk <tadeusz.struk@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/affinity.h | 8 --------
 drivers/infiniband/hw/hfi1/hfi.h      | 1 -
 2 files changed, 9 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/affinity.h b/drivers/infiniband/hw/hfi1/affinity.h
index b89ea3c0ee1a..ef4ebc358619 100644
--- a/drivers/infiniband/hw/hfi1/affinity.h
+++ b/drivers/infiniband/hw/hfi1/affinity.h
@@ -70,14 +70,6 @@ struct cpu_mask_set {
 	uint gen;
 };
 
-struct hfi1_affinity {
-	struct cpu_mask_set def_intr;
-	struct cpu_mask_set rcv_intr;
-	struct cpumask real_cpu_mask;
-	/* spin lock to protect affinity struct */
-	spinlock_t lock;
-};
-
 struct hfi1_msix_entry;
 
 /* Initialize non-HT cpu cores mask */
diff --git a/drivers/infiniband/hw/hfi1/hfi.h b/drivers/infiniband/hw/hfi1/hfi.h
index a66d198199ff..4163596ce4c9 100644
--- a/drivers/infiniband/hw/hfi1/hfi.h
+++ b/drivers/infiniband/hw/hfi1/hfi.h
@@ -1185,7 +1185,6 @@ struct hfi1_devdata {
 	/* Number of verbs contexts which have disabled ASPM */
 	atomic_t aspm_disabled_cnt;
 
-	struct hfi1_affinity *affinity;
 	bool eprom_available;	/* true if EPROM is available for this device */
 	bool aspm_supported;	/* Does HW support ASPM */
 	bool aspm_enabled;	/* ASPM state: enabled/disabled */
-- 
cgit v1.2.1


From 5213006ade47153f5f2c2e5a42e84312afccabfd Mon Sep 17 00:00:00 2001
From: Dean Luick <dean.luick@intel.com>
Date: Wed, 7 Dec 2016 19:33:13 -0800
Subject: IB/hfi1: Add special setting for low power AOC

Low power QSFP AOC cables require a different SerDes
Tx PLL bandwidth setting than the default.  The
8051 firmware does not know the details, so the driver
needs to tell the firmware through a special setting.

Reviewed-by: Easwar Hariharan <easwar.hariharan@intel.com>
Signed-off-by: Dean Luick <dean.luick@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/platform.c | 44 ++++++++++++++++++++++++++++++++++-
 1 file changed, 43 insertions(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/platform.c b/drivers/infiniband/hw/hfi1/platform.c
index 0e7043801bb1..838fe84e285a 100644
--- a/drivers/infiniband/hw/hfi1/platform.c
+++ b/drivers/infiniband/hw/hfi1/platform.c
@@ -637,6 +637,38 @@ static void apply_tx_lanes(struct hfi1_pportdata *ppd, u8 field_id,
 	}
 }
 
+/*
+ * Return a special SerDes setting for low power AOC cables.  The power class
+ * threshold and setting being used were all found by empirical testing.
+ *
+ * Summary of the logic:
+ *
+ * if (QSFP and QSFP_TYPE == AOC and QSFP_POWER_CLASS < 4)
+ *     return 0xe
+ * return 0; // leave at default
+ */
+static u8 aoc_low_power_setting(struct hfi1_pportdata *ppd)
+{
+	u8 *cache = ppd->qsfp_info.cache;
+	int power_class;
+
+	/* QSFP only */
+	if (ppd->port_type != PORT_TYPE_QSFP)
+		return 0; /* leave at default */
+
+	/* active optical cables only */
+	switch ((cache[QSFP_MOD_TECH_OFFS] & 0xF0) >> 4) {
+	case 0x0 ... 0x9: /* fallthrough */
+	case 0xC: /* fallthrough */
+	case 0xE:
+		/* active AOC */
+		power_class = get_qsfp_power_class(cache[QSFP_MOD_PWR_OFFS]);
+		if (power_class < QSFP_POWER_CLASS_4)
+			return 0xe;
+	}
+	return 0; /* leave at default */
+}
+
 static void apply_tunings(
 		struct hfi1_pportdata *ppd, u32 tx_preset_index,
 		u8 tuning_method, u32 total_atten, u8 limiting_active)
@@ -705,7 +737,17 @@ static void apply_tunings(
 		tx_preset_index, TX_PRESET_TABLE_POSTCUR, &tx_preset, 4);
 	postcur = tx_preset;
 
-	config_data = precur | (attn << 8) | (postcur << 16);
+	/*
+	 * NOTES:
+	 * o The aoc_low_power_setting is applied to all lanes even
+	 *   though only lane 0's value is examined by the firmware.
+	 * o A lingering low power setting after a cable swap does
+	 *   not occur.  On cable unplug the 8051 is reset and
+	 *   restarted on cable insert.  This resets all settings to
+	 *   their default, erasing any previous low power setting.
+	 */
+	config_data = precur | (attn << 8) | (postcur << 16) |
+			(aoc_low_power_setting(ppd) << 24);
 
 	apply_tx_lanes(ppd, TX_EQ_SETTINGS, config_data,
 		       "Applying TX settings");
-- 
cgit v1.2.1


From fcb29a6668a254104d38c0132d1853d0644af066 Mon Sep 17 00:00:00 2001
From: Mike Marciniszyn <mike.marciniszyn@intel.com>
Date: Wed, 7 Dec 2016 19:33:20 -0800
Subject: IB/rdmavt: Add trace of MR segs

Add tracing of MR segment information.

Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rdmavt/mr.c       |   4 ++
 drivers/infiniband/sw/rdmavt/trace.h    |   1 +
 drivers/infiniband/sw/rdmavt/trace_mr.h | 112 ++++++++++++++++++++++++++++++++
 3 files changed, 117 insertions(+)
 create mode 100644 drivers/infiniband/sw/rdmavt/trace_mr.h

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rdmavt/mr.c b/drivers/infiniband/sw/rdmavt/mr.c
index 4acf179adf3b..562cf78b6ae9 100644
--- a/drivers/infiniband/sw/rdmavt/mr.c
+++ b/drivers/infiniband/sw/rdmavt/mr.c
@@ -51,6 +51,7 @@
 #include <rdma/rdma_vt.h>
 #include "vt.h"
 #include "mr.h"
+#include "trace.h"
 
 /**
  * rvt_driver_mr_init - Init MR resources per driver
@@ -403,6 +404,7 @@ struct ib_mr *rvt_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 		}
 		mr->mr.map[m]->segs[n].vaddr = vaddr;
 		mr->mr.map[m]->segs[n].length = umem->page_size;
+		trace_rvt_mr_user_seg(&mr->mr, m, n, vaddr, umem->page_size);
 		n++;
 		if (n == RVT_SEGSZ) {
 			m++;
@@ -507,6 +509,7 @@ static int rvt_set_page(struct ib_mr *ibmr, u64 addr)
 	n = mapped_segs % RVT_SEGSZ;
 	mr->mr.map[m]->segs[n].vaddr = (void *)addr;
 	mr->mr.map[m]->segs[n].length = ps;
+	trace_rvt_mr_page_seg(&mr->mr, m, n, (void *)addr, ps);
 	mr->mr.length += ps;
 
 	return 0;
@@ -693,6 +696,7 @@ int rvt_map_phys_fmr(struct ib_fmr *ibfmr, u64 *page_list,
 	for (i = 0; i < list_len; i++) {
 		fmr->mr.map[m]->segs[n].vaddr = (void *)page_list[i];
 		fmr->mr.map[m]->segs[n].length = ps;
+		trace_rvt_mr_fmr_seg(&fmr->mr, m, n, (void *)page_list[i], ps);
 		if (++n == RVT_SEGSZ) {
 			m++;
 			n = 0;
diff --git a/drivers/infiniband/sw/rdmavt/trace.h b/drivers/infiniband/sw/rdmavt/trace.h
index 11a2afbc8e76..e2d23acb6a7d 100644
--- a/drivers/infiniband/sw/rdmavt/trace.h
+++ b/drivers/infiniband/sw/rdmavt/trace.h
@@ -51,3 +51,4 @@
 #include "trace_rvt.h"
 #include "trace_qp.h"
 #include "trace_tx.h"
+#include "trace_mr.h"
diff --git a/drivers/infiniband/sw/rdmavt/trace_mr.h b/drivers/infiniband/sw/rdmavt/trace_mr.h
new file mode 100644
index 000000000000..3318a6c36373
--- /dev/null
+++ b/drivers/infiniband/sw/rdmavt/trace_mr.h
@@ -0,0 +1,112 @@
+/*
+ * Copyright(c) 2016 Intel Corporation.
+ *
+ * This file is provided under a dual BSD/GPLv2 license.  When using or
+ * redistributing this file, you may do so under either license.
+ *
+ * GPL LICENSE SUMMARY
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of version 2 of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * BSD LICENSE
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ *  - Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ *  - Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in
+ *    the documentation and/or other materials provided with the
+ *    distribution.
+ *  - Neither the name of Intel Corporation nor the names of its
+ *    contributors may be used to endorse or promote products derived
+ *    from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ */
+#if !defined(__RVT_TRACE_MR_H) || defined(TRACE_HEADER_MULTI_READ)
+#define __RVT_TRACE_MR_H
+
+#include <linux/tracepoint.h>
+#include <linux/trace_seq.h>
+
+#include <rdma/ib_verbs.h>
+#include <rdma/rdma_vt.h>
+#include <rdma/rdmavt_mr.h>
+
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM rvt_mr
+DECLARE_EVENT_CLASS(
+	rvt_mr_template,
+	TP_PROTO(struct rvt_mregion *mr, u16 m, u16 n, void *v, size_t len),
+	TP_ARGS(mr, m, n, v, len),
+	TP_STRUCT__entry(
+		RDI_DEV_ENTRY(ib_to_rvt(mr->pd->device))
+		__field(void *, vaddr)
+		__field(struct page *, page)
+		__field(size_t, len)
+		__field(u32, lkey)
+		__field(u16, m)
+		__field(u16, n)
+	),
+	TP_fast_assign(
+		RDI_DEV_ASSIGN(ib_to_rvt(mr->pd->device));
+		__entry->vaddr = v;
+		__entry->page = virt_to_page(v);
+		__entry->m = m;
+		__entry->n = n;
+		__entry->len = len;
+	),
+	TP_printk(
+		"[%s] vaddr %p page %p m %u n %u len %ld",
+		__get_str(dev),
+		__entry->vaddr,
+		__entry->page,
+		__entry->m,
+		__entry->n,
+		__entry->len
+	)
+);
+
+DEFINE_EVENT(
+	rvt_mr_template, rvt_mr_page_seg,
+	TP_PROTO(struct rvt_mregion *mr, u16 m, u16 n, void *v, size_t len),
+	TP_ARGS(mr, m, n, v, len));
+
+DEFINE_EVENT(
+	rvt_mr_template, rvt_mr_fmr_seg,
+	TP_PROTO(struct rvt_mregion *mr, u16 m, u16 n, void *v, size_t len),
+	TP_ARGS(mr, m, n, v, len));
+
+DEFINE_EVENT(
+	rvt_mr_template, rvt_mr_user_seg,
+	TP_PROTO(struct rvt_mregion *mr, u16 m, u16 n, void *v, size_t len),
+	TP_ARGS(mr, m, n, v, len));
+
+#endif /* __RVT_TRACE_MR_H */
+
+#undef TRACE_INCLUDE_PATH
+#undef TRACE_INCLUDE_FILE
+#define TRACE_INCLUDE_PATH .
+#define TRACE_INCLUDE_FILE trace_mr
+#include <trace/define_trace.h>
-- 
cgit v1.2.1


From b777f154a0c21e1187c1806ababf9c5ba3e49eea Mon Sep 17 00:00:00 2001
From: Mitko Haralanov <mitko.haralanov@intel.com>
Date: Wed, 7 Dec 2016 19:33:27 -0800
Subject: IB/hfi1: Remove usage of qp->s_cur_sge

The s_cur_sge field in the qp structure holds a pointer to the
SGE of the currently processed WQE. It assumes the protection
of the RVT_S_BUSY flag to prevent the changing of this field
while the send engine is using it. This scheme works as long
as there is only one instance of the send engine running at a
time.

Scaling of the send engine to multiple cores would break this
assumption as there could be multiple instances of the send engine
running on different CPUs. This opens a window where the QP's
RVT_S_BUSY flag is not set but the send engine is still running.

To prevent accidental changing of the s_cur_sge pointer, the QP's
dependence on it is removed. The SGE pointer is now stored in the
verbs_txreq, which is a per-packet data structure. This ensures
that each individual packet has it's own pointer, which is setup
while the RVT_S_BUSY flag is set.

Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Mitko Haralanov <mitko.haralanov@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/rc.c    | 10 +++++-----
 drivers/infiniband/hw/hfi1/uc.c    |  2 +-
 drivers/infiniband/hw/hfi1/ud.c    |  2 +-
 drivers/infiniband/hw/hfi1/verbs.c | 14 ++++++--------
 4 files changed, 13 insertions(+), 15 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/rc.c b/drivers/infiniband/hw/hfi1/rc.c
index e69161e007db..c996a373837e 100644
--- a/drivers/infiniband/hw/hfi1/rc.c
+++ b/drivers/infiniband/hw/hfi1/rc.c
@@ -276,7 +276,7 @@ static int make_rc_ack(struct hfi1_ibdev *dev, struct rvt_qp *qp,
 				rvt_get_mr(ps->s_txreq->mr);
 			qp->s_ack_rdma_sge.sge = e->rdma_sge;
 			qp->s_ack_rdma_sge.num_sge = 1;
-			qp->s_cur_sge = &qp->s_ack_rdma_sge;
+			ps->s_txreq->ss = &qp->s_ack_rdma_sge;
 			if (len > pmtu) {
 				len = pmtu;
 				qp->s_ack_state = OP(RDMA_READ_RESPONSE_FIRST);
@@ -290,7 +290,7 @@ static int make_rc_ack(struct hfi1_ibdev *dev, struct rvt_qp *qp,
 			bth2 = mask_psn(qp->s_ack_rdma_psn++);
 		} else {
 			/* COMPARE_SWAP or FETCH_ADD */
-			qp->s_cur_sge = NULL;
+			ps->s_txreq->ss = NULL;
 			len = 0;
 			qp->s_ack_state = OP(ATOMIC_ACKNOWLEDGE);
 			ohdr->u.at.aeth = hfi1_compute_aeth(qp);
@@ -306,7 +306,7 @@ static int make_rc_ack(struct hfi1_ibdev *dev, struct rvt_qp *qp,
 		qp->s_ack_state = OP(RDMA_READ_RESPONSE_MIDDLE);
 		/* FALLTHROUGH */
 	case OP(RDMA_READ_RESPONSE_MIDDLE):
-		qp->s_cur_sge = &qp->s_ack_rdma_sge;
+		ps->s_txreq->ss = &qp->s_ack_rdma_sge;
 		ps->s_txreq->mr = qp->s_ack_rdma_sge.sge.mr;
 		if (ps->s_txreq->mr)
 			rvt_get_mr(ps->s_txreq->mr);
@@ -335,7 +335,7 @@ normal:
 		 */
 		qp->s_ack_state = OP(SEND_ONLY);
 		qp->s_flags &= ~RVT_S_ACK_PENDING;
-		qp->s_cur_sge = NULL;
+		ps->s_txreq->ss = NULL;
 		if (qp->s_nak_state)
 			ohdr->u.aeth =
 				cpu_to_be32((qp->r_msn & HFI1_MSN_MASK) |
@@ -801,7 +801,7 @@ int hfi1_make_rc_req(struct rvt_qp *qp, struct hfi1_pkt_state *ps)
 	qp->s_len -= len;
 	qp->s_hdrwords = hwords;
 	ps->s_txreq->sde = priv->s_sde;
-	qp->s_cur_sge = ss;
+	ps->s_txreq->ss = ss;
 	ps->s_txreq->s_cur_size = len;
 	hfi1_make_ruc_header(
 		qp,
diff --git a/drivers/infiniband/hw/hfi1/uc.c b/drivers/infiniband/hw/hfi1/uc.c
index d062c3537f7c..b141a78ae38b 100644
--- a/drivers/infiniband/hw/hfi1/uc.c
+++ b/drivers/infiniband/hw/hfi1/uc.c
@@ -258,7 +258,7 @@ int hfi1_make_uc_req(struct rvt_qp *qp, struct hfi1_pkt_state *ps)
 	qp->s_len -= len;
 	qp->s_hdrwords = hwords;
 	ps->s_txreq->sde = priv->s_sde;
-	qp->s_cur_sge = &qp->s_sge;
+	ps->s_txreq->ss = &qp->s_sge;
 	ps->s_txreq->s_cur_size = len;
 	hfi1_make_ruc_header(qp, ohdr, bth0 | (qp->s_state << 24),
 			     mask_psn(qp->s_psn++), middle, ps);
diff --git a/drivers/infiniband/hw/hfi1/ud.c b/drivers/infiniband/hw/hfi1/ud.c
index d742ac1af947..c071955c0272 100644
--- a/drivers/infiniband/hw/hfi1/ud.c
+++ b/drivers/infiniband/hw/hfi1/ud.c
@@ -355,7 +355,7 @@ int hfi1_make_ud_req(struct rvt_qp *qp, struct hfi1_pkt_state *ps)
 	/* header size in 32-bit words LRH+BTH+DETH = (8+12+8)/4. */
 	qp->s_hdrwords = 7;
 	ps->s_txreq->s_cur_size = wqe->length;
-	qp->s_cur_sge = &qp->s_sge;
+	ps->s_txreq->ss = &qp->s_sge;
 	qp->s_srate = ah_attr->static_rate;
 	qp->srate_mbps = ib_rate_to_mbps(qp->s_srate);
 	qp->s_wqe = wqe;
diff --git a/drivers/infiniband/hw/hfi1/verbs.c b/drivers/infiniband/hw/hfi1/verbs.c
index 65f2f1d76fc8..3b7bfd817647 100644
--- a/drivers/infiniband/hw/hfi1/verbs.c
+++ b/drivers/infiniband/hw/hfi1/verbs.c
@@ -790,10 +790,10 @@ static int wait_kmem(struct hfi1_ibdev *dev,
  */
 static noinline int build_verbs_ulp_payload(
 	struct sdma_engine *sde,
-	struct rvt_sge_state *ss,
 	u32 length,
 	struct verbs_txreq *tx)
 {
+	struct rvt_sge_state *ss = tx->ss;
 	struct rvt_sge *sg_list = ss->sg_list;
 	struct rvt_sge sge = ss->sge;
 	u8 num_sge = ss->num_sge;
@@ -837,7 +837,6 @@ bail_txadd:
 /* New API */
 static int build_verbs_tx_desc(
 	struct sdma_engine *sde,
-	struct rvt_sge_state *ss,
 	u32 length,
 	struct verbs_txreq *tx,
 	struct hfi1_ahg_info *ahg_info,
@@ -881,9 +880,9 @@ static int build_verbs_tx_desc(
 			goto bail_txadd;
 	}
 
-	/* add the ulp payload - if any.  ss can be NULL for acks */
-	if (ss)
-		ret = build_verbs_ulp_payload(sde, ss, length, tx);
+	/* add the ulp payload - if any. tx->ss can be NULL for acks */
+	if (tx->ss)
+		ret = build_verbs_ulp_payload(sde, length, tx);
 bail_txadd:
 	return ret;
 }
@@ -894,7 +893,6 @@ int hfi1_verbs_send_dma(struct rvt_qp *qp, struct hfi1_pkt_state *ps,
 	struct hfi1_qp_priv *priv = qp->priv;
 	struct hfi1_ahg_info *ahg_info = priv->s_ahg;
 	u32 hdrwords = qp->s_hdrwords;
-	struct rvt_sge_state *ss = qp->s_cur_sge;
 	u32 len = ps->s_txreq->s_cur_size;
 	u32 plen = hdrwords + ((len + 3) >> 2) + 2; /* includes pbc */
 	struct hfi1_ibdev *dev = ps->dev;
@@ -920,7 +918,7 @@ int hfi1_verbs_send_dma(struct rvt_qp *qp, struct hfi1_pkt_state *ps,
 					 plen);
 		}
 		tx->wqe = qp->s_wqe;
-		ret = build_verbs_tx_desc(tx->sde, ss, len, tx, ahg_info, pbc);
+		ret = build_verbs_tx_desc(tx->sde, len, tx, ahg_info, pbc);
 		if (unlikely(ret))
 			goto bail_build;
 	}
@@ -1011,7 +1009,7 @@ int hfi1_verbs_send_pio(struct rvt_qp *qp, struct hfi1_pkt_state *ps,
 {
 	struct hfi1_qp_priv *priv = qp->priv;
 	u32 hdrwords = qp->s_hdrwords;
-	struct rvt_sge_state *ss = qp->s_cur_sge;
+	struct rvt_sge_state *ss = ps->s_txreq->ss;
 	u32 len = ps->s_txreq->s_cur_size;
 	u32 dwords = (len + 3) >> 2;
 	u32 plen = hdrwords + dwords + 2; /* includes pbc */
-- 
cgit v1.2.1


From 9b86071c5ef7a8c0a0d6d541cb79df5a8b115d91 Mon Sep 17 00:00:00 2001
From: Sebastian Sanchez <sebastian.sanchez@intel.com>
Date: Wed, 7 Dec 2016 19:33:33 -0800
Subject: IB/hfi1: Remove critical section gap in sc_buffer_alloc()

In sc_buffer_alloc(), the sc->alloc_lock is released
before calling sc_release_update(), and it is reacquired
after the function call. This causes CPU lock trading.
Fix it by not dropping the lock before calling
sc_release_update().

Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/pio.c | 2 --
 1 file changed, 2 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/pio.c b/drivers/infiniband/hw/hfi1/pio.c
index 86a7f365b624..75c4dea84c4b 100644
--- a/drivers/infiniband/hw/hfi1/pio.c
+++ b/drivers/infiniband/hw/hfi1/pio.c
@@ -1419,9 +1419,7 @@ retry:
 			(sc->fill - sc->alloc_free);
 		if (blocks > avail) {
 			/* still no room, actively update */
-			spin_unlock_irqrestore(&sc->alloc_lock, flags);
 			sc_release_update(sc);
-			spin_lock_irqsave(&sc->alloc_lock, flags);
 			sc->alloc_free = ACCESS_ONCE(sc->free);
 			trycount++;
 			goto retry;
-- 
cgit v1.2.1


From 008016746720d4295e38e940e22dbd116c90c95b Mon Sep 17 00:00:00 2001
From: Dean Luick <dean.luick@intel.com>
Date: Wed, 7 Dec 2016 19:33:40 -0800
Subject: IB/hfi1: Preserve external device completed bit

The driver should not change the external device request
completed bit when not actually doing an external device
request.

Reviewed-by: Easwar Hariharan <easwar.hariharan@intel.com>
Signed-off-by: Dean Luick <dean.luick@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/chip.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/chip.c b/drivers/infiniband/hw/hfi1/chip.c
index f87d80567572..37d8af50cc13 100644
--- a/drivers/infiniband/hw/hfi1/chip.c
+++ b/drivers/infiniband/hw/hfi1/chip.c
@@ -8488,7 +8488,10 @@ static int do_8051_command(
 	 */
 	if (type == HCMD_WRITE_LCB_CSR) {
 		in_data |= ((*out_data) & 0xffffffffffull) << 8;
-		reg = ((((*out_data) >> 40) & 0xff) <<
+		/* must preserve COMPLETED - it is tied to hardware */
+		reg = read_csr(dd, DC_DC8051_CFG_EXT_DEV_0);
+		reg &= DC_DC8051_CFG_EXT_DEV_0_COMPLETED_SMASK;
+		reg |= ((((*out_data) >> 40) & 0xff) <<
 				DC_DC8051_CFG_EXT_DEV_0_RETURN_CODE_SHIFT)
 		      | ((((*out_data) >> 48) & 0xffff) <<
 				DC_DC8051_CFG_EXT_DEV_0_RSP_DATA_SHIFT);
-- 
cgit v1.2.1


From b44980f8794faf3561d970fe8c05e34765fbdc9b Mon Sep 17 00:00:00 2001
From: Sebastian Sanchez <sebastian.sanchez@intel.com>
Date: Wed, 7 Dec 2016 19:33:46 -0800
Subject: IB/hfi1: Replace qp->refcount release code with standard driver
 wrapper

Some parts of the code don't use the standard release
wrapper rvt_put_qp() for decrementing and testing
the refcount to then try to use a resource.
Replace this code with the standard driver wrapper.

Fixes: Commit 4d6f85c3fa55 ("IB/rdmavt, IB/qib, IB/hfi1: Use new QP put get routines")
Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/driver.c  | 3 +--
 drivers/infiniband/sw/rdmavt/mcast.c | 5 ++---
 2 files changed, 3 insertions(+), 5 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/driver.c b/drivers/infiniband/hw/hfi1/driver.c
index 6563e4d38b80..d4261163bd25 100644
--- a/drivers/infiniband/hw/hfi1/driver.c
+++ b/drivers/infiniband/hw/hfi1/driver.c
@@ -793,8 +793,7 @@ static inline void process_rcv_qp_work(struct hfi1_packet *packet)
 				hfi1_schedule_send(qp);
 			spin_unlock_irqrestore(&qp->s_lock, flags);
 		}
-		if (atomic_dec_and_test(&qp->refcount))
-			wake_up(&qp->wait);
+		rvt_put_qp(qp);
 	}
 }
 
diff --git a/drivers/infiniband/sw/rdmavt/mcast.c b/drivers/infiniband/sw/rdmavt/mcast.c
index 983d319ac976..05c8c2afb0e3 100644
--- a/drivers/infiniband/sw/rdmavt/mcast.c
+++ b/drivers/infiniband/sw/rdmavt/mcast.c
@@ -81,7 +81,7 @@ static struct rvt_mcast_qp *rvt_mcast_qp_alloc(struct rvt_qp *qp)
 		goto bail;
 
 	mqp->qp = qp;
-	atomic_inc(&qp->refcount);
+	rvt_get_qp(qp);
 
 bail:
 	return mqp;
@@ -92,8 +92,7 @@ static void rvt_mcast_qp_free(struct rvt_mcast_qp *mqp)
 	struct rvt_qp *qp = mqp->qp;
 
 	/* Notify hfi1_destroy_qp() if it is waiting. */
-	if (atomic_dec_and_test(&qp->refcount))
-		wake_up(&qp->wait);
+	rvt_put_qp(qp);
 
 	kfree(mqp);
 }
-- 
cgit v1.2.1


From f84dfa26e61c6f31a3a0366532296911df763170 Mon Sep 17 00:00:00 2001
From: Sebastian Sanchez <sebastian.sanchez@intel.com>
Date: Wed, 7 Dec 2016 19:33:53 -0800
Subject: IB/hfi1: Use reference count wrapper for MRs

Some parts of the code don't use the standard driver
wrapper for memory region reference counters. Use the
standard driver wrapper throughout the code.

Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rdmavt/mr.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rdmavt/mr.c b/drivers/infiniband/sw/rdmavt/mr.c
index 562cf78b6ae9..52fd15276ee6 100644
--- a/drivers/infiniband/sw/rdmavt/mr.c
+++ b/drivers/infiniband/sw/rdmavt/mr.c
@@ -793,7 +793,7 @@ int rvt_lkey_ok(struct rvt_lkey_table *rkt, struct rvt_pd *pd,
 		mr = rcu_dereference(dev->dma_mr);
 		if (!mr)
 			goto bail;
-		atomic_inc(&mr->refcount);
+		rvt_get_mr(mr);
 		rcu_read_unlock();
 
 		isge->mr = mr;
@@ -814,7 +814,7 @@ int rvt_lkey_ok(struct rvt_lkey_table *rkt, struct rvt_pd *pd,
 		     off + sge->length > mr->length ||
 		     (mr->access_flags & acc) != acc))
 		goto bail;
-	atomic_inc(&mr->refcount);
+	rvt_get_mr(mr);
 	rcu_read_unlock();
 
 	off += mr->offset;
@@ -892,7 +892,7 @@ int rvt_rkey_ok(struct rvt_qp *qp, struct rvt_sge *sge,
 		mr = rcu_dereference(rdi->dma_mr);
 		if (!mr)
 			goto bail;
-		atomic_inc(&mr->refcount);
+		rvt_get_mr(mr);
 		rcu_read_unlock();
 
 		sge->mr = mr;
@@ -913,7 +913,7 @@ int rvt_rkey_ok(struct rvt_qp *qp, struct rvt_sge *sge,
 	if (unlikely(vaddr < mr->iova || off + len > mr->length ||
 		     (mr->access_flags & acc) == 0))
 		goto bail;
-	atomic_inc(&mr->refcount);
+	rvt_get_mr(mr);
 	rcu_read_unlock();
 
 	off += mr->offset;
-- 
cgit v1.2.1


From 238b1862b4608ee8c28733be67e08f0447963090 Mon Sep 17 00:00:00 2001
From: Sebastian Sanchez <sebastian.sanchez@intel.com>
Date: Wed, 7 Dec 2016 19:34:00 -0800
Subject: IB/qib: Use standard refcount wrapper for QPs

Use the standard driver wrapper for QP reference counters.
This makes the code more maintainable.

Fixes: Commit 4d6f85c3fa55 ("IB/rdmavt, IB/qib, IB/hfi1: Use new QP put get routines")
Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Sebastian Sanchez <sebastian.sanchez@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/qib/qib_driver.c |  3 +--
 drivers/infiniband/hw/qib/qib_verbs.c  | 20 ++++++++------------
 2 files changed, 9 insertions(+), 14 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/qib/qib_driver.c b/drivers/infiniband/hw/qib/qib_driver.c
index 728e0a030d2e..2b5982f743ef 100644
--- a/drivers/infiniband/hw/qib/qib_driver.c
+++ b/drivers/infiniband/hw/qib/qib_driver.c
@@ -420,8 +420,7 @@ static u32 qib_rcv_hdrerr(struct qib_ctxtdata *rcd, struct qib_pportdata *ppd,
 						if (list_empty(&qp->rspwait)) {
 							qp->r_flags |=
 								RVT_R_RSP_NAK;
-							atomic_inc(
-								&qp->refcount);
+							rvt_get_qp(qp);
 							list_add_tail(
 							 &qp->rspwait,
 							 &rcd->qp_wait_list);
diff --git a/drivers/infiniband/hw/qib/qib_verbs.c b/drivers/infiniband/hw/qib/qib_verbs.c
index 954f15064514..204482468953 100644
--- a/drivers/infiniband/hw/qib/qib_verbs.c
+++ b/drivers/infiniband/hw/qib/qib_verbs.c
@@ -464,7 +464,7 @@ static void mem_timer(unsigned long data)
 		priv = list_entry(list->next, struct qib_qp_priv, iowait);
 		qp = priv->owner;
 		list_del_init(&priv->iowait);
-		atomic_inc(&qp->refcount);
+		rvt_get_qp(qp);
 		if (!list_empty(list))
 			mod_timer(&dev->mem_timer, jiffies + 1);
 	}
@@ -477,8 +477,7 @@ static void mem_timer(unsigned long data)
 			qib_schedule_send(qp);
 		}
 		spin_unlock_irqrestore(&qp->s_lock, flags);
-		if (atomic_dec_and_test(&qp->refcount))
-			wake_up(&qp->wait);
+		rvt_put_qp(qp);
 	}
 }
 
@@ -762,7 +761,7 @@ void qib_put_txreq(struct qib_verbs_txreq *tx)
 				  iowait);
 		qp = priv->owner;
 		list_del_init(&priv->iowait);
-		atomic_inc(&qp->refcount);
+		rvt_get_qp(qp);
 		spin_unlock_irqrestore(&dev->rdi.pending_lock, flags);
 
 		spin_lock_irqsave(&qp->s_lock, flags);
@@ -772,8 +771,7 @@ void qib_put_txreq(struct qib_verbs_txreq *tx)
 		}
 		spin_unlock_irqrestore(&qp->s_lock, flags);
 
-		if (atomic_dec_and_test(&qp->refcount))
-			wake_up(&qp->wait);
+		rvt_put_qp(qp);
 	} else
 		spin_unlock_irqrestore(&dev->rdi.pending_lock, flags);
 }
@@ -808,7 +806,7 @@ void qib_verbs_sdma_desc_avail(struct qib_pportdata *ppd, unsigned avail)
 			break;
 		avail -= qpp->s_tx->txreq.sg_count;
 		list_del_init(&qpp->iowait);
-		atomic_inc(&qp->refcount);
+		rvt_get_qp(qp);
 		qps[n++] = qp;
 	}
 
@@ -822,8 +820,7 @@ void qib_verbs_sdma_desc_avail(struct qib_pportdata *ppd, unsigned avail)
 			qib_schedule_send(qp);
 		}
 		spin_unlock(&qp->s_lock);
-		if (atomic_dec_and_test(&qp->refcount))
-			wake_up(&qp->wait);
+		rvt_put_qp(qp);
 	}
 }
 
@@ -1288,7 +1285,7 @@ void qib_ib_piobufavail(struct qib_devdata *dd)
 		priv = list_entry(list->next, struct qib_qp_priv, iowait);
 		qp = priv->owner;
 		list_del_init(&priv->iowait);
-		atomic_inc(&qp->refcount);
+		rvt_get_qp(qp);
 		qps[n++] = qp;
 	}
 	dd->f_wantpiobuf_intr(dd, 0);
@@ -1306,8 +1303,7 @@ full:
 		spin_unlock_irqrestore(&qp->s_lock, flags);
 
 		/* Notify qib_destroy_qp() if it is waiting. */
-		if (atomic_dec_and_test(&qp->refcount))
-			wake_up(&qp->wait);
+		rvt_put_qp(qp);
 	}
 }
 
-- 
cgit v1.2.1


From f2dc9cdce83c4aac2f8b6c803b0327df1c7d44a6 Mon Sep 17 00:00:00 2001
From: Mike Marciniszyn <mike.marciniszyn@intel.com>
Date: Wed, 7 Dec 2016 19:34:06 -0800
Subject: IB/rdmavt: Add a send completion helper

This is for use by client drivers to drive
send completions into a CQ.

A new exported table allows for the mapping
of ib_wr_opcode into a ib_wc_opcode.

Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rdmavt/qp.c | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rdmavt/qp.c b/drivers/infiniband/sw/rdmavt/qp.c
index 6500c3b5a89c..9e14addd690c 100644
--- a/drivers/infiniband/sw/rdmavt/qp.c
+++ b/drivers/infiniband/sw/rdmavt/qp.c
@@ -76,6 +76,23 @@ const int ib_rvt_state_ops[IB_QPS_ERR + 1] = {
 };
 EXPORT_SYMBOL(ib_rvt_state_ops);
 
+/*
+ * Translate ib_wr_opcode into ib_wc_opcode.
+ */
+const enum ib_wc_opcode ib_rvt_wc_opcode[] = {
+	[IB_WR_RDMA_WRITE] = IB_WC_RDMA_WRITE,
+	[IB_WR_RDMA_WRITE_WITH_IMM] = IB_WC_RDMA_WRITE,
+	[IB_WR_SEND] = IB_WC_SEND,
+	[IB_WR_SEND_WITH_IMM] = IB_WC_SEND,
+	[IB_WR_RDMA_READ] = IB_WC_RDMA_READ,
+	[IB_WR_ATOMIC_CMP_AND_SWP] = IB_WC_COMP_SWAP,
+	[IB_WR_ATOMIC_FETCH_AND_ADD] = IB_WC_FETCH_ADD,
+	[IB_WR_SEND_WITH_INV] = IB_WC_SEND,
+	[IB_WR_LOCAL_INV] = IB_WC_LOCAL_INV,
+	[IB_WR_REG_MR] = IB_WC_REG_MR
+};
+EXPORT_SYMBOL(ib_rvt_wc_opcode);
+
 static void get_map_page(struct rvt_qpn_table *qpt,
 			 struct rvt_qpn_map *map,
 			 gfp_t gfp)
-- 
cgit v1.2.1


From 0771da5a6e9d67e48207a81ddf6c8e739e9493bd Mon Sep 17 00:00:00 2001
From: Mike Marciniszyn <mike.marciniszyn@intel.com>
Date: Wed, 7 Dec 2016 19:34:12 -0800
Subject: IB/hfi1,IB/qib: Use new send completion helper

Convert cq completion returns in both rdmavt drivers
to use the new helper.

Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/rc.c       | 26 ++------------------------
 drivers/infiniband/hw/hfi1/ruc.c      | 17 +----------------
 drivers/infiniband/hw/hfi1/verbs.c    | 16 ----------------
 drivers/infiniband/hw/qib/qib_rc.c    | 26 ++------------------------
 drivers/infiniband/hw/qib/qib_ruc.c   | 17 +----------------
 drivers/infiniband/hw/qib/qib_verbs.c | 13 -------------
 6 files changed, 6 insertions(+), 109 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/rc.c b/drivers/infiniband/hw/hfi1/rc.c
index c996a373837e..3df1c0650571 100644
--- a/drivers/infiniband/hw/hfi1/rc.c
+++ b/drivers/infiniband/hw/hfi1/rc.c
@@ -1146,7 +1146,6 @@ void hfi1_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr)
 {
 	struct ib_other_headers *ohdr;
 	struct rvt_swqe *wqe;
-	struct ib_wc wc;
 	unsigned i;
 	u32 opcode;
 	u32 psn;
@@ -1200,17 +1199,7 @@ void hfi1_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr)
 
 			rvt_put_mr(sge->mr);
 		}
-		/* Post a send completion queue entry if requested. */
-		if (!(qp->s_flags & RVT_S_SIGNAL_REQ_WR) ||
-		    (wqe->wr.send_flags & IB_SEND_SIGNALED)) {
-			memset(&wc, 0, sizeof(wc));
-			wc.wr_id = wqe->wr.wr_id;
-			wc.status = IB_WC_SUCCESS;
-			wc.opcode = ib_hfi1_wc_opcode[wqe->wr.opcode];
-			wc.byte_len = wqe->length;
-			wc.qp = &qp->ibqp;
-			rvt_cq_enter(ibcq_to_rvtcq(qp->ibqp.send_cq), &wc, 0);
-		}
+		rvt_qp_swqe_complete(qp, wqe, IB_WC_SUCCESS);
 	}
 	/*
 	 * If we were waiting for sends to complete before re-sending,
@@ -1240,7 +1229,6 @@ static struct rvt_swqe *do_rc_completion(struct rvt_qp *qp,
 					 struct rvt_swqe *wqe,
 					 struct hfi1_ibport *ibp)
 {
-	struct ib_wc wc;
 	unsigned i;
 
 	lockdep_assert_held(&qp->s_lock);
@@ -1264,17 +1252,7 @@ static struct rvt_swqe *do_rc_completion(struct rvt_qp *qp,
 		qp->s_last = s_last;
 		/* see post_send() */
 		barrier();
-		/* Post a send completion queue entry if requested. */
-		if (!(qp->s_flags & RVT_S_SIGNAL_REQ_WR) ||
-		    (wqe->wr.send_flags & IB_SEND_SIGNALED)) {
-			memset(&wc, 0, sizeof(wc));
-			wc.wr_id = wqe->wr.wr_id;
-			wc.status = IB_WC_SUCCESS;
-			wc.opcode = ib_hfi1_wc_opcode[wqe->wr.opcode];
-			wc.byte_len = wqe->length;
-			wc.qp = &qp->ibqp;
-			rvt_cq_enter(ibcq_to_rvtcq(qp->ibqp.send_cq), &wc, 0);
-		}
+		rvt_qp_swqe_complete(qp, wqe, IB_WC_SUCCESS);
 	} else {
 		struct hfi1_pportdata *ppd = ppd_from_ibp(ibp);
 
diff --git a/drivers/infiniband/hw/hfi1/ruc.c b/drivers/infiniband/hw/hfi1/ruc.c
index 2ec3e908691b..03c89315f1c7 100644
--- a/drivers/infiniband/hw/hfi1/ruc.c
+++ b/drivers/infiniband/hw/hfi1/ruc.c
@@ -964,22 +964,7 @@ void hfi1_send_complete(struct rvt_qp *qp, struct rvt_swqe *wqe,
 	    qp->ibqp.qp_type == IB_QPT_GSI)
 		atomic_dec(&ibah_to_rvtah(wqe->ud_wr.ah)->refcount);
 
-	/* See ch. 11.2.4.1 and 10.7.3.1 */
-	if (!(qp->s_flags & RVT_S_SIGNAL_REQ_WR) ||
-	    (wqe->wr.send_flags & IB_SEND_SIGNALED) ||
-	    status != IB_WC_SUCCESS) {
-		struct ib_wc wc;
-
-		memset(&wc, 0, sizeof(wc));
-		wc.wr_id = wqe->wr.wr_id;
-		wc.status = status;
-		wc.opcode = ib_hfi1_wc_opcode[wqe->wr.opcode];
-		wc.qp = &qp->ibqp;
-		if (status == IB_WC_SUCCESS)
-			wc.byte_len = wqe->length;
-		rvt_cq_enter(ibcq_to_rvtcq(qp->ibqp.send_cq), &wc,
-			     status != IB_WC_SUCCESS);
-	}
+	rvt_qp_swqe_complete(qp, wqe, status);
 
 	if (qp->s_acked == old_last)
 		qp->s_acked = last;
diff --git a/drivers/infiniband/hw/hfi1/verbs.c b/drivers/infiniband/hw/hfi1/verbs.c
index 3b7bfd817647..95ed4d6da510 100644
--- a/drivers/infiniband/hw/hfi1/verbs.c
+++ b/drivers/infiniband/hw/hfi1/verbs.c
@@ -296,22 +296,6 @@ static inline int wss_exceeds_threshold(void)
 	return atomic_read(&wss.total_count) >= wss.threshold;
 }
 
-/*
- * Translate ib_wr_opcode into ib_wc_opcode.
- */
-const enum ib_wc_opcode ib_hfi1_wc_opcode[] = {
-	[IB_WR_RDMA_WRITE] = IB_WC_RDMA_WRITE,
-	[IB_WR_RDMA_WRITE_WITH_IMM] = IB_WC_RDMA_WRITE,
-	[IB_WR_SEND] = IB_WC_SEND,
-	[IB_WR_SEND_WITH_IMM] = IB_WC_SEND,
-	[IB_WR_RDMA_READ] = IB_WC_RDMA_READ,
-	[IB_WR_ATOMIC_CMP_AND_SWP] = IB_WC_COMP_SWAP,
-	[IB_WR_ATOMIC_FETCH_AND_ADD] = IB_WC_FETCH_ADD,
-	[IB_WR_SEND_WITH_INV] = IB_WC_SEND,
-	[IB_WR_LOCAL_INV] = IB_WC_LOCAL_INV,
-	[IB_WR_REG_MR] = IB_WC_REG_MR
-};
-
 /*
  * Length of header by opcode, 0 --> not supported
  */
diff --git a/drivers/infiniband/hw/qib/qib_rc.c b/drivers/infiniband/hw/qib/qib_rc.c
index 2097512e75aa..d54990514730 100644
--- a/drivers/infiniband/hw/qib/qib_rc.c
+++ b/drivers/infiniband/hw/qib/qib_rc.c
@@ -941,7 +941,6 @@ void qib_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr)
 {
 	struct ib_other_headers *ohdr;
 	struct rvt_swqe *wqe;
-	struct ib_wc wc;
 	unsigned i;
 	u32 opcode;
 	u32 psn;
@@ -993,17 +992,7 @@ void qib_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr)
 
 			rvt_put_mr(sge->mr);
 		}
-		/* Post a send completion queue entry if requested. */
-		if (!(qp->s_flags & RVT_S_SIGNAL_REQ_WR) ||
-		    (wqe->wr.send_flags & IB_SEND_SIGNALED)) {
-			memset(&wc, 0, sizeof(wc));
-			wc.wr_id = wqe->wr.wr_id;
-			wc.status = IB_WC_SUCCESS;
-			wc.opcode = ib_qib_wc_opcode[wqe->wr.opcode];
-			wc.byte_len = wqe->length;
-			wc.qp = &qp->ibqp;
-			rvt_cq_enter(ibcq_to_rvtcq(qp->ibqp.send_cq), &wc, 0);
-		}
+		rvt_qp_swqe_complete(qp, wqe, IB_WC_SUCCESS);
 	}
 	/*
 	 * If we were waiting for sends to complete before resending,
@@ -1032,7 +1021,6 @@ static struct rvt_swqe *do_rc_completion(struct rvt_qp *qp,
 					 struct rvt_swqe *wqe,
 					 struct qib_ibport *ibp)
 {
-	struct ib_wc wc;
 	unsigned i;
 
 	/*
@@ -1055,17 +1043,7 @@ static struct rvt_swqe *do_rc_completion(struct rvt_qp *qp,
 		qp->s_last = s_last;
 		/* see post_send() */
 		barrier();
-		/* Post a send completion queue entry if requested. */
-		if (!(qp->s_flags & RVT_S_SIGNAL_REQ_WR) ||
-		    (wqe->wr.send_flags & IB_SEND_SIGNALED)) {
-			memset(&wc, 0, sizeof(wc));
-			wc.wr_id = wqe->wr.wr_id;
-			wc.status = IB_WC_SUCCESS;
-			wc.opcode = ib_qib_wc_opcode[wqe->wr.opcode];
-			wc.byte_len = wqe->length;
-			wc.qp = &qp->ibqp;
-			rvt_cq_enter(ibcq_to_rvtcq(qp->ibqp.send_cq), &wc, 0);
-		}
+		rvt_qp_swqe_complete(qp, wqe, IB_WC_SUCCESS);
 	} else
 		this_cpu_inc(*ibp->rvp.rc_delayed_comp);
 
diff --git a/drivers/infiniband/hw/qib/qib_ruc.c b/drivers/infiniband/hw/qib/qib_ruc.c
index de1bde5950f5..87e3eb5a836b 100644
--- a/drivers/infiniband/hw/qib/qib_ruc.c
+++ b/drivers/infiniband/hw/qib/qib_ruc.c
@@ -815,22 +815,7 @@ void qib_send_complete(struct rvt_qp *qp, struct rvt_swqe *wqe,
 	    qp->ibqp.qp_type == IB_QPT_GSI)
 		atomic_dec(&ibah_to_rvtah(wqe->ud_wr.ah)->refcount);
 
-	/* See ch. 11.2.4.1 and 10.7.3.1 */
-	if (!(qp->s_flags & RVT_S_SIGNAL_REQ_WR) ||
-	    (wqe->wr.send_flags & IB_SEND_SIGNALED) ||
-	    status != IB_WC_SUCCESS) {
-		struct ib_wc wc;
-
-		memset(&wc, 0, sizeof(wc));
-		wc.wr_id = wqe->wr.wr_id;
-		wc.status = status;
-		wc.opcode = ib_qib_wc_opcode[wqe->wr.opcode];
-		wc.qp = &qp->ibqp;
-		if (status == IB_WC_SUCCESS)
-			wc.byte_len = wqe->length;
-		rvt_cq_enter(ibcq_to_rvtcq(qp->ibqp.send_cq), &wc,
-			     status != IB_WC_SUCCESS);
-	}
+	rvt_qp_swqe_complete(qp, wqe, status);
 
 	if (qp->s_acked == old_last)
 		qp->s_acked = last;
diff --git a/drivers/infiniband/hw/qib/qib_verbs.c b/drivers/infiniband/hw/qib/qib_verbs.c
index 204482468953..4b54c0ddd08a 100644
--- a/drivers/infiniband/hw/qib/qib_verbs.c
+++ b/drivers/infiniband/hw/qib/qib_verbs.c
@@ -113,19 +113,6 @@ static unsigned int ib_qib_disable_sma;
 module_param_named(disable_sma, ib_qib_disable_sma, uint, S_IWUSR | S_IRUGO);
 MODULE_PARM_DESC(disable_sma, "Disable the SMA");
 
-/*
- * Translate ib_wr_opcode into ib_wc_opcode.
- */
-const enum ib_wc_opcode ib_qib_wc_opcode[] = {
-	[IB_WR_RDMA_WRITE] = IB_WC_RDMA_WRITE,
-	[IB_WR_RDMA_WRITE_WITH_IMM] = IB_WC_RDMA_WRITE,
-	[IB_WR_SEND] = IB_WC_SEND,
-	[IB_WR_SEND_WITH_IMM] = IB_WC_SEND,
-	[IB_WR_RDMA_READ] = IB_WC_RDMA_READ,
-	[IB_WR_ATOMIC_CMP_AND_SWP] = IB_WC_COMP_SWAP,
-	[IB_WR_ATOMIC_FETCH_AND_ADD] = IB_WC_FETCH_ADD
-};
-
 /*
  * System image GUID.
  */
-- 
cgit v1.2.1


From 9d8145a604937780898c0e4bdb124a57988fc2ed Mon Sep 17 00:00:00 2001
From: Harish Chegondi <harish.chegondi@intel.com>
Date: Wed, 7 Dec 2016 19:34:19 -0800
Subject: IB/hfi1: Avoid credit return allocation for cpu-less NUMA nodes

Do not allocate credit return base and DMA memory for
NUMA nodes without CPUs.

Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Harish Chegondi <harish.chegondi@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/affinity.c |  3 ++-
 drivers/infiniband/hw/hfi1/affinity.h |  1 +
 drivers/infiniband/hw/hfi1/pio.c      | 20 +++-----------------
 3 files changed, 6 insertions(+), 18 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/affinity.c b/drivers/infiniband/hw/hfi1/affinity.c
index a26a9a0bfc41..4962b6ef1f34 100644
--- a/drivers/infiniband/hw/hfi1/affinity.c
+++ b/drivers/infiniband/hw/hfi1/affinity.c
@@ -125,6 +125,7 @@ int node_affinity_init(void)
 				cpumask_weight(topology_sibling_cpumask(
 					cpumask_first(&node_affinity.proc.mask)
 					));
+	node_affinity.num_possible_nodes = num_possible_nodes();
 	node_affinity.num_online_nodes = num_online_nodes();
 	node_affinity.num_online_cpus = num_online_cpus();
 
@@ -135,7 +136,7 @@ int node_affinity_init(void)
 	 */
 	init_real_cpu_mask();
 
-	hfi1_per_node_cntr = kcalloc(num_possible_nodes(),
+	hfi1_per_node_cntr = kcalloc(node_affinity.num_possible_nodes,
 				     sizeof(*hfi1_per_node_cntr), GFP_KERNEL);
 	if (!hfi1_per_node_cntr)
 		return -ENOMEM;
diff --git a/drivers/infiniband/hw/hfi1/affinity.h b/drivers/infiniband/hw/hfi1/affinity.h
index ef4ebc358619..c9453b3d47b4 100644
--- a/drivers/infiniband/hw/hfi1/affinity.h
+++ b/drivers/infiniband/hw/hfi1/affinity.h
@@ -111,6 +111,7 @@ struct hfi1_affinity_node_list {
 	struct cpumask real_cpu_mask;
 	struct cpu_mask_set proc;
 	int num_core_siblings;
+	int num_possible_nodes;
 	int num_online_nodes;
 	int num_online_cpus;
 	struct mutex lock; /* protects affinity nodes */
diff --git a/drivers/infiniband/hw/hfi1/pio.c b/drivers/infiniband/hw/hfi1/pio.c
index 75c4dea84c4b..3d5dac2f5266 100644
--- a/drivers/infiniband/hw/hfi1/pio.c
+++ b/drivers/infiniband/hw/hfi1/pio.c
@@ -2036,21 +2036,10 @@ freesc15:
 int init_credit_return(struct hfi1_devdata *dd)
 {
 	int ret;
-	int num_numa;
 	int i;
 
-	num_numa = num_online_nodes();
-	/* enforce the expectation that the numas are compact */
-	for (i = 0; i < num_numa; i++) {
-		if (!node_online(i)) {
-			dd_dev_err(dd, "NUMA nodes are not compact\n");
-			ret = -EINVAL;
-			goto done;
-		}
-	}
-
 	dd->cr_base = kcalloc(
-		num_numa,
+		node_affinity.num_possible_nodes,
 		sizeof(struct credit_return_base),
 		GFP_KERNEL);
 	if (!dd->cr_base) {
@@ -2058,7 +2047,7 @@ int init_credit_return(struct hfi1_devdata *dd)
 		ret = -ENOMEM;
 		goto done;
 	}
-	for (i = 0; i < num_numa; i++) {
+	for_each_node_with_cpus(i) {
 		int bytes = TXE_NUM_CONTEXTS * sizeof(struct credit_return);
 
 		set_dev_node(&dd->pcidev->dev, i);
@@ -2085,14 +2074,11 @@ done:
 
 void free_credit_return(struct hfi1_devdata *dd)
 {
-	int num_numa;
 	int i;
 
 	if (!dd->cr_base)
 		return;
-
-	num_numa = num_online_nodes();
-	for (i = 0; i < num_numa; i++) {
+	for (i = 0; i < node_affinity.num_possible_nodes; i++) {
 		if (dd->cr_base[i].va) {
 			dma_free_coherent(&dd->pcidev->dev,
 					  TXE_NUM_CONTEXTS *
-- 
cgit v1.2.1


From c64607aa8abd16ba584523bdf83a708fae91986a Mon Sep 17 00:00:00 2001
From: Mike Marciniszyn <mike.marciniszyn@intel.com>
Date: Wed, 7 Dec 2016 19:34:31 -0800
Subject: IB/hfi1,IB/qib: use rvt swqe mr deref helper

Convert to use new swqe put routine.

Reviewed-by: Brian Welty <brian.welty@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/rc.c     | 15 ++-------------
 drivers/infiniband/hw/hfi1/ruc.c    |  7 +------
 drivers/infiniband/hw/qib/qib_rc.c  | 15 ++-------------
 drivers/infiniband/hw/qib/qib_ruc.c |  7 +------
 4 files changed, 6 insertions(+), 38 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/rc.c b/drivers/infiniband/hw/hfi1/rc.c
index 3df1c0650571..ddc02237e5f6 100644
--- a/drivers/infiniband/hw/hfi1/rc.c
+++ b/drivers/infiniband/hw/hfi1/rc.c
@@ -1146,7 +1146,6 @@ void hfi1_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr)
 {
 	struct ib_other_headers *ohdr;
 	struct rvt_swqe *wqe;
-	unsigned i;
 	u32 opcode;
 	u32 psn;
 
@@ -1194,11 +1193,7 @@ void hfi1_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr)
 		qp->s_last = s_last;
 		/* see post_send() */
 		barrier();
-		for (i = 0; i < wqe->wr.num_sge; i++) {
-			struct rvt_sge *sge = &wqe->sg_list[i];
-
-			rvt_put_mr(sge->mr);
-		}
+		rvt_put_swqe(wqe);
 		rvt_qp_swqe_complete(qp, wqe, IB_WC_SUCCESS);
 	}
 	/*
@@ -1229,8 +1224,6 @@ static struct rvt_swqe *do_rc_completion(struct rvt_qp *qp,
 					 struct rvt_swqe *wqe,
 					 struct hfi1_ibport *ibp)
 {
-	unsigned i;
-
 	lockdep_assert_held(&qp->s_lock);
 	/*
 	 * Don't decrement refcount and don't generate a
@@ -1241,11 +1234,7 @@ static struct rvt_swqe *do_rc_completion(struct rvt_qp *qp,
 	    cmp_psn(qp->s_sending_psn, qp->s_sending_hpsn) > 0) {
 		u32 s_last;
 
-		for (i = 0; i < wqe->wr.num_sge; i++) {
-			struct rvt_sge *sge = &wqe->sg_list[i];
-
-			rvt_put_mr(sge->mr);
-		}
+		rvt_put_swqe(wqe);
 		s_last = qp->s_last;
 		if (++s_last >= qp->s_size)
 			s_last = 0;
diff --git a/drivers/infiniband/hw/hfi1/ruc.c b/drivers/infiniband/hw/hfi1/ruc.c
index 03c89315f1c7..717ed4b159d3 100644
--- a/drivers/infiniband/hw/hfi1/ruc.c
+++ b/drivers/infiniband/hw/hfi1/ruc.c
@@ -942,7 +942,6 @@ void hfi1_send_complete(struct rvt_qp *qp, struct rvt_swqe *wqe,
 			enum ib_wc_status status)
 {
 	u32 old_last, last;
-	unsigned i;
 
 	if (!(ib_rvt_state_ops[qp->state] & RVT_PROCESS_OR_FLUSH_SEND))
 		return;
@@ -954,11 +953,7 @@ void hfi1_send_complete(struct rvt_qp *qp, struct rvt_swqe *wqe,
 	qp->s_last = last;
 	/* See post_send() */
 	barrier();
-	for (i = 0; i < wqe->wr.num_sge; i++) {
-		struct rvt_sge *sge = &wqe->sg_list[i];
-
-		rvt_put_mr(sge->mr);
-	}
+	rvt_put_swqe(wqe);
 	if (qp->ibqp.qp_type == IB_QPT_UD ||
 	    qp->ibqp.qp_type == IB_QPT_SMI ||
 	    qp->ibqp.qp_type == IB_QPT_GSI)
diff --git a/drivers/infiniband/hw/qib/qib_rc.c b/drivers/infiniband/hw/qib/qib_rc.c
index d54990514730..e23ee6ce435a 100644
--- a/drivers/infiniband/hw/qib/qib_rc.c
+++ b/drivers/infiniband/hw/qib/qib_rc.c
@@ -941,7 +941,6 @@ void qib_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr)
 {
 	struct ib_other_headers *ohdr;
 	struct rvt_swqe *wqe;
-	unsigned i;
 	u32 opcode;
 	u32 psn;
 
@@ -987,11 +986,7 @@ void qib_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr)
 		qp->s_last = s_last;
 		/* see post_send() */
 		barrier();
-		for (i = 0; i < wqe->wr.num_sge; i++) {
-			struct rvt_sge *sge = &wqe->sg_list[i];
-
-			rvt_put_mr(sge->mr);
-		}
+		rvt_put_swqe(wqe);
 		rvt_qp_swqe_complete(qp, wqe, IB_WC_SUCCESS);
 	}
 	/*
@@ -1021,8 +1016,6 @@ static struct rvt_swqe *do_rc_completion(struct rvt_qp *qp,
 					 struct rvt_swqe *wqe,
 					 struct qib_ibport *ibp)
 {
-	unsigned i;
-
 	/*
 	 * Don't decrement refcount and don't generate a
 	 * completion if the SWQE is being resent until the send
@@ -1032,11 +1025,7 @@ static struct rvt_swqe *do_rc_completion(struct rvt_qp *qp,
 	    qib_cmp24(qp->s_sending_psn, qp->s_sending_hpsn) > 0) {
 		u32 s_last;
 
-		for (i = 0; i < wqe->wr.num_sge; i++) {
-			struct rvt_sge *sge = &wqe->sg_list[i];
-
-			rvt_put_mr(sge->mr);
-		}
+		rvt_put_swqe(wqe);
 		s_last = qp->s_last;
 		if (++s_last >= qp->s_size)
 			s_last = 0;
diff --git a/drivers/infiniband/hw/qib/qib_ruc.c b/drivers/infiniband/hw/qib/qib_ruc.c
index 87e3eb5a836b..e54a2feeeb10 100644
--- a/drivers/infiniband/hw/qib/qib_ruc.c
+++ b/drivers/infiniband/hw/qib/qib_ruc.c
@@ -793,7 +793,6 @@ void qib_send_complete(struct rvt_qp *qp, struct rvt_swqe *wqe,
 		       enum ib_wc_status status)
 {
 	u32 old_last, last;
-	unsigned i;
 
 	if (!(ib_rvt_state_ops[qp->state] & RVT_PROCESS_OR_FLUSH_SEND))
 		return;
@@ -805,11 +804,7 @@ void qib_send_complete(struct rvt_qp *qp, struct rvt_swqe *wqe,
 	qp->s_last = last;
 	/* See post_send() */
 	barrier();
-	for (i = 0; i < wqe->wr.num_sge; i++) {
-		struct rvt_sge *sge = &wqe->sg_list[i];
-
-		rvt_put_mr(sge->mr);
-	}
+	rvt_put_swqe(wqe);
 	if (qp->ibqp.qp_type == IB_QPT_UD ||
 	    qp->ibqp.qp_type == IB_QPT_SMI ||
 	    qp->ibqp.qp_type == IB_QPT_GSI)
-- 
cgit v1.2.1


From 5dc806052a9fc8c44e111646f9a9f436877749d0 Mon Sep 17 00:00:00 2001
From: Mike Marciniszyn <mike.marciniszyn@intel.com>
Date: Wed, 7 Dec 2016 19:34:37 -0800
Subject: IB/rdmavt, IB/hfi1, IB/qib: Add inlines for mtu division

Add rvt_div_round_up_mtu() and rvt_div_mtu() routines to
do the computation based on the pmtu and the log_pmtu.

Change divides in qib, hfi1 to use the new inlines.

Reviewed-by: Kaike Wan <kaike.wan@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/rc.c    | 3 +--
 drivers/infiniband/hw/qib/qib_rc.c | 3 +--
 2 files changed, 2 insertions(+), 4 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/rc.c b/drivers/infiniband/hw/hfi1/rc.c
index ddc02237e5f6..9db260fe782a 100644
--- a/drivers/infiniband/hw/hfi1/rc.c
+++ b/drivers/infiniband/hw/hfi1/rc.c
@@ -2377,8 +2377,7 @@ send_last:
 			 * Update the next expected PSN.  We add 1 later
 			 * below, so only add the remainder here.
 			 */
-			if (len > pmtu)
-				qp->r_psn += (len - 1) / pmtu;
+			qp->r_psn += rvt_div_mtu(qp, len - 1);
 		} else {
 			e->rdma_sge.mr = NULL;
 			e->rdma_sge.vaddr = NULL;
diff --git a/drivers/infiniband/hw/qib/qib_rc.c b/drivers/infiniband/hw/qib/qib_rc.c
index e23ee6ce435a..031433cb7206 100644
--- a/drivers/infiniband/hw/qib/qib_rc.c
+++ b/drivers/infiniband/hw/qib/qib_rc.c
@@ -2079,8 +2079,7 @@ send_last:
 			 * Update the next expected PSN.  We add 1 later
 			 * below, so only add the remainder here.
 			 */
-			if (len > pmtu)
-				qp->r_psn += (len - 1) / pmtu;
+			qp->r_psn += rvt_div_mtu(qp, len - 1);
 		} else {
 			e->rdma_sge.mr = NULL;
 			e->rdma_sge.vaddr = NULL;
-- 
cgit v1.2.1


From 0fc859a6576aaa747c5359c2bf622aac3158e2c7 Mon Sep 17 00:00:00 2001
From: Bhumika Goyal <bhumirks@gmail.com>
Date: Sat, 19 Nov 2016 15:17:48 +0530
Subject: IB/hfi1: constify mmu_notifier_ops structure

Declare the structure mmu_notifier_ops as const as it is only stored in
the ops field of a mmu_notifier structure. The ops field is of type
const struct mmu_notifier_ops *, so mmu_notifier_ops structures having
this property can be declared as const.
Done using coccinelle:
@r1 disable optional_qualifier @
identifier i;
position p;
@@
static struct mmu_notifier_ops i@p = {...};

@ok1@
identifier r1.i;
position p;
struct mmu_rb_handler handler;
@@
handler.mn.ops=&i@p

@bad@
position p!={r1.p,ok1.p};
identifier r1.i;
@@
i@p

@depends on !bad disable optional_qualifier@
identifier r1.i;
@@
static
+const
struct mmu_notifier_ops i={...};

@depends on !bad disable optional_qualifier@
identifier r1.i;
@@
+const
struct mmu_notifier_ops i;

File size before:
   text	   data	    bss	    dec	    hex	filename
   3566	     72	     16	   3654	    e46
drivers/infiniband/hw/hfi1/mmu_rb.o

File size after:
   text	   data	    bss	    dec	    hex	filename
   3658	      0	     16	   3674	    e5a
drivers/infiniband/hw/hfi1/mmu_rb.o

Signed-off-by: Bhumika Goyal <bhumirks@gmail.com>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/mmu_rb.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/mmu_rb.c b/drivers/infiniband/hw/hfi1/mmu_rb.c
index 7ad30898fc19..ccbf52c8ff6f 100644
--- a/drivers/infiniband/hw/hfi1/mmu_rb.c
+++ b/drivers/infiniband/hw/hfi1/mmu_rb.c
@@ -81,7 +81,7 @@ static void do_remove(struct mmu_rb_handler *handler,
 		      struct list_head *del_list);
 static void handle_remove(struct work_struct *work);
 
-static struct mmu_notifier_ops mn_opts = {
+static const struct mmu_notifier_ops mn_opts = {
 	.invalidate_page = mmu_notifier_page,
 	.invalidate_range_start = mmu_notifier_range_start,
 };
-- 
cgit v1.2.1


From 66431b0e8657e2406742105f89175f571340090b Mon Sep 17 00:00:00 2001
From: Bart Van Assche <bart.vanassche@sandisk.com>
Date: Mon, 5 Dec 2016 16:48:11 -0800
Subject: IB/hfi1: Define platform_config_table_limits once

Defining static data structures in a header file is wrong because
this causes the data structure to be instantiated once in every .c
file it is included in. Hence move the definition of a static
array from a header file into the only .c file in which it is used.

Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Cc: Dennis Dalessandro <dennis.dalessandro@intel.com>
Cc: Dean Luick <dean.luick@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/hfi1/firmware.c | 10 ++++++++++
 drivers/infiniband/hw/hfi1/platform.h | 10 ----------
 2 files changed, 10 insertions(+), 10 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/hfi1/firmware.c b/drivers/infiniband/hw/hfi1/firmware.c
index bcf7da4599c4..0dd50cdb039a 100644
--- a/drivers/infiniband/hw/hfi1/firmware.c
+++ b/drivers/infiniband/hw/hfi1/firmware.c
@@ -239,6 +239,16 @@ static const u8 all_fabric_serdes_broadcast = 0xe1;
 const u8 pcie_serdes_broadcast[2] = { 0xe2, 0xe3 };
 static const u8 all_pcie_serdes_broadcast = 0xe0;
 
+static const u32 platform_config_table_limits[PLATFORM_CONFIG_TABLE_MAX] = {
+	0,
+	SYSTEM_TABLE_MAX,
+	PORT_TABLE_MAX,
+	RX_PRESET_TABLE_MAX,
+	TX_PRESET_TABLE_MAX,
+	QSFP_ATTEN_TABLE_MAX,
+	VARIABLE_SETTINGS_TABLE_MAX
+};
+
 /* forwards */
 static void dispose_one_firmware(struct firmware_details *fdet);
 static int load_fabric_serdes_firmware(struct hfi1_devdata *dd,
diff --git a/drivers/infiniband/hw/hfi1/platform.h b/drivers/infiniband/hw/hfi1/platform.h
index 31155401a95b..eed0aa9124fa 100644
--- a/drivers/infiniband/hw/hfi1/platform.h
+++ b/drivers/infiniband/hw/hfi1/platform.h
@@ -168,16 +168,6 @@ struct platform_config_cache {
 	struct platform_config_data config_tables[PLATFORM_CONFIG_TABLE_MAX];
 };
 
-static const u32 platform_config_table_limits[PLATFORM_CONFIG_TABLE_MAX] = {
-	0,
-	SYSTEM_TABLE_MAX,
-	PORT_TABLE_MAX,
-	RX_PRESET_TABLE_MAX,
-	TX_PRESET_TABLE_MAX,
-	QSFP_ATTEN_TABLE_MAX,
-	VARIABLE_SETTINGS_TABLE_MAX
-};
-
 /* This section defines default values and encodings for the
  * fields defined for each table above
  */
-- 
cgit v1.2.1


From a0fa72683e78979ef1123d679b1c40ae28bd9096 Mon Sep 17 00:00:00 2001
From: Arnd Bergmann <arnd@arndb.de>
Date: Mon, 19 Sep 2016 13:57:26 +0200
Subject: IB/rxe: avoid putting a large struct rxe_qp on stack

A race condition fix added an rxe_qp structure to the stack in order
to be able to perform rollback in rxe_requester(), but the structure
is large enough to trigger the warning for possible stack overflow:

drivers/infiniband/sw/rxe/rxe_req.c: In function 'rxe_requester':
drivers/infiniband/sw/rxe/rxe_req.c:757:1: error: the frame size of 2064 bytes is larger than 1024 bytes [-Werror=frame-larger-than=]

This changes the rollback function to only save the psn inside
the qp, which is the only field we access in the rollback_qp
anyway.

Fixes: 3050b9985024 ("IB/rxe: Fix race condition between requester and completer")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rxe/rxe_req.c | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rxe/rxe_req.c b/drivers/infiniband/sw/rxe/rxe_req.c
index 832846b73ea0..205222909e53 100644
--- a/drivers/infiniband/sw/rxe/rxe_req.c
+++ b/drivers/infiniband/sw/rxe/rxe_req.c
@@ -548,23 +548,23 @@ static void update_wqe_psn(struct rxe_qp *qp,
 static void save_state(struct rxe_send_wqe *wqe,
 		       struct rxe_qp *qp,
 		       struct rxe_send_wqe *rollback_wqe,
-		       struct rxe_qp *rollback_qp)
+		       u32 *rollback_psn)
 {
 	rollback_wqe->state     = wqe->state;
 	rollback_wqe->first_psn = wqe->first_psn;
 	rollback_wqe->last_psn  = wqe->last_psn;
-	rollback_qp->req.psn    = qp->req.psn;
+	*rollback_psn		= qp->req.psn;
 }
 
 static void rollback_state(struct rxe_send_wqe *wqe,
 			   struct rxe_qp *qp,
 			   struct rxe_send_wqe *rollback_wqe,
-			   struct rxe_qp *rollback_qp)
+			   u32 rollback_psn)
 {
 	wqe->state     = rollback_wqe->state;
 	wqe->first_psn = rollback_wqe->first_psn;
 	wqe->last_psn  = rollback_wqe->last_psn;
-	qp->req.psn    = rollback_qp->req.psn;
+	qp->req.psn    = rollback_psn;
 }
 
 static void update_state(struct rxe_qp *qp, struct rxe_send_wqe *wqe,
@@ -593,8 +593,8 @@ int rxe_requester(void *arg)
 	int mtu;
 	int opcode;
 	int ret;
-	struct rxe_qp rollback_qp;
 	struct rxe_send_wqe rollback_wqe;
+	u32 rollback_psn;
 
 next_wqe:
 	if (unlikely(!qp->valid || qp->req.state == QP_STATE_ERROR))
@@ -718,7 +718,7 @@ next_wqe:
 	 * rxe_xmit_packet().
 	 * Otherwise, completer might initiate an unjustified retry flow.
 	 */
-	save_state(wqe, qp, &rollback_wqe, &rollback_qp);
+	save_state(wqe, qp, &rollback_wqe, &rollback_psn);
 	update_wqe_state(qp, wqe, &pkt);
 	update_wqe_psn(qp, wqe, &pkt, payload);
 	ret = rxe_xmit_packet(to_rdev(qp->ibqp.device), qp, &pkt, skb);
@@ -726,7 +726,7 @@ next_wqe:
 		qp->need_req_skb = 1;
 		kfree_skb(skb);
 
-		rollback_state(wqe, qp, &rollback_wqe, &rollback_qp);
+		rollback_state(wqe, qp, &rollback_wqe, rollback_psn);
 
 		if (ret == -EAGAIN) {
 			rxe_run_task(&qp->req.task, 1);
-- 
cgit v1.2.1


From 4ac4707102d9ea1ffc9b5735891f6c5ee3d236e5 Mon Sep 17 00:00:00 2001
From: Wei Yongjun <weiyongjun1@huawei.com>
Date: Sat, 29 Oct 2016 16:19:33 +0000
Subject: IB/rxe: Use DEFINE_SPINLOCK() for spinlock

spinlock can be initialized automatically with DEFINE_SPINLOCK()
rather than explicitly calling spin_lock_init().

Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Reviewed-by: Leon Romanosky <leonro@mellanox.com>
Reviewed-by: Moni Shoua <monis@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rxe/rxe_net.c | 6 +-----
 1 file changed, 1 insertion(+), 5 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rxe/rxe_net.c b/drivers/infiniband/sw/rxe/rxe_net.c
index b8258e4f0aea..4cb63780fa74 100644
--- a/drivers/infiniband/sw/rxe/rxe_net.c
+++ b/drivers/infiniband/sw/rxe/rxe_net.c
@@ -46,7 +46,7 @@
 #include "rxe_loc.h"
 
 static LIST_HEAD(rxe_dev_list);
-static spinlock_t dev_list_lock; /* spinlock for device list */
+static DEFINE_SPINLOCK(dev_list_lock); /* spinlock for device list */
 
 struct rxe_dev *net_to_rxe(struct net_device *ndev)
 {
@@ -663,8 +663,6 @@ struct notifier_block rxe_net_notifier = {
 
 int rxe_net_ipv4_init(void)
 {
-	spin_lock_init(&dev_list_lock);
-
 	recv_sockets.sk4 = rxe_setup_udp_tunnel(&init_net,
 				htons(ROCE_V2_UDP_DPORT), false);
 	if (IS_ERR(recv_sockets.sk4)) {
@@ -680,8 +678,6 @@ int rxe_net_ipv6_init(void)
 {
 #if IS_ENABLED(CONFIG_IPV6)
 
-	spin_lock_init(&dev_list_lock);
-
 	recv_sockets.sk6 = rxe_setup_udp_tunnel(&init_net,
 						htons(ROCE_V2_UDP_DPORT), true);
 	if (IS_ERR(recv_sockets.sk6)) {
-- 
cgit v1.2.1


From 95db9d05b717b1ec242799103c09acda8d728974 Mon Sep 17 00:00:00 2001
From: Dan Carpenter <dan.carpenter@oracle.com>
Date: Thu, 17 Nov 2016 14:00:05 +0300
Subject: IB/rxe: Remove unneeded cast in rxe_srq_from_attr()

It makes me nervous when we cast pointer parameters.  I would estimate
that around 50% of the time, it indicates a bug.  Here the cast is not
needed becaue u32 and and unsigned int are the same thing.  Removing the
cast makes the code more robust and future proof in case any of the
types change.

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Leon Romanovsky <leonro@mellanox.com>
Reviewed-by: Yuval Shaia <yuval.shaia@oracle.com>
Acked-by: Moni Shoua <monis@mellanox.com>
Reviewed-by: Yuval Shaia <yuval.shaia@oracle.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rxe/rxe_srq.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rxe/rxe_srq.c b/drivers/infiniband/sw/rxe/rxe_srq.c
index 2a6e3cd2d4e8..efc832a2d7c6 100644
--- a/drivers/infiniband/sw/rxe/rxe_srq.c
+++ b/drivers/infiniband/sw/rxe/rxe_srq.c
@@ -169,7 +169,7 @@ int rxe_srq_from_attr(struct rxe_dev *rxe, struct rxe_srq *srq,
 			}
 		}
 
-		err = rxe_queue_resize(q, (unsigned int *)&attr->max_wr,
+		err = rxe_queue_resize(q, &attr->max_wr,
 				       rcv_wqe_size(srq->rq.max_sge),
 				       srq->rq.queue->ip ?
 						srq->rq.queue->ip->context :
-- 
cgit v1.2.1


From 6e9bb530ff8b07b3bbce79c937fc9dcb32da4eb9 Mon Sep 17 00:00:00 2001
From: Andrew Boyer <andrew.boyer@dell.com>
Date: Wed, 23 Nov 2016 12:39:17 -0500
Subject: IB/rxe: Remove buffer used for printing IP address

Avoid smashing the stack when an ICRC error occurs on an IPv6 network.

Signed-off-by: Andrew Boyer <andrew.boyer@dell.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rxe/rxe_recv.c | 11 +++++------
 1 file changed, 5 insertions(+), 6 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rxe/rxe_recv.c b/drivers/infiniband/sw/rxe/rxe_recv.c
index 46f062842a9a..252b4d637d45 100644
--- a/drivers/infiniband/sw/rxe/rxe_recv.c
+++ b/drivers/infiniband/sw/rxe/rxe_recv.c
@@ -391,16 +391,15 @@ int rxe_rcv(struct sk_buff *skb)
 			     payload_size(pkt));
 	calc_icrc = cpu_to_be32(~calc_icrc);
 	if (unlikely(calc_icrc != pack_icrc)) {
-		char saddr[sizeof(struct in6_addr)];
-
 		if (skb->protocol == htons(ETH_P_IPV6))
-			sprintf(saddr, "%pI6", &ipv6_hdr(skb)->saddr);
+			pr_warn_ratelimited("bad ICRC from %pI6c\n",
+					    &ipv6_hdr(skb)->saddr);
 		else if (skb->protocol == htons(ETH_P_IP))
-			sprintf(saddr, "%pI4", &ip_hdr(skb)->saddr);
+			pr_warn_ratelimited("bad ICRC from %pI4\n",
+					    &ip_hdr(skb)->saddr);
 		else
-			sprintf(saddr, "unknown");
+			pr_warn_ratelimited("bad ICRC from unknown\n");
 
-		pr_warn_ratelimited("bad ICRC from %s\n", saddr);
 		goto drop;
 	}
 
-- 
cgit v1.2.1


From dd753d87436ce7ef2a958d684b38d5acc99c9f5c Mon Sep 17 00:00:00 2001
From: Andrew Boyer <andrew.boyer@dell.com>
Date: Wed, 23 Nov 2016 12:39:18 -0500
Subject: IB/rxe: Advance the consumer pointer before posting the CQE

A simple userspace application might poll the CQ, find a completion,
and then attempt to post a new WQE to the SQ. A spurious error can
occur if the userspace application detects a full SQ in the instant
before the kernel is able to advance the SQ consumer pointer.

This is noticeable when using single-entry SQs with ibv_rc_pingpong
if lots of kernel and userspace library debugging is enabled.

Signed-off-by: Andrew Boyer <andrew.boyer@dell.com>
Reviewed-by: Yonatan Cohen <yonatanc@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rxe/rxe_comp.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rxe/rxe_comp.c b/drivers/infiniband/sw/rxe/rxe_comp.c
index 6c5e29db88e3..d46c49b33b13 100644
--- a/drivers/infiniband/sw/rxe/rxe_comp.c
+++ b/drivers/infiniband/sw/rxe/rxe_comp.c
@@ -420,11 +420,12 @@ static void do_complete(struct rxe_qp *qp, struct rxe_send_wqe *wqe)
 	    (wqe->wr.send_flags & IB_SEND_SIGNALED) ||
 	    (qp->req.state == QP_STATE_ERROR)) {
 		make_send_cqe(qp, wqe, &cqe);
+		advance_consumer(qp->sq.queue);
 		rxe_cq_post(qp->scq, &cqe, 0);
+	} else {
+		advance_consumer(qp->sq.queue);
 	}
 
-	advance_consumer(qp->sq.queue);
-
 	/*
 	 * we completed something so let req run again
 	 * if it is trying to fence
-- 
cgit v1.2.1


From 2a7a85487e5432424eef7a394ed26ef1d8f0d192 Mon Sep 17 00:00:00 2001
From: Andrew Boyer <andrew.boyer@dell.com>
Date: Wed, 23 Nov 2016 12:39:19 -0500
Subject: IB/rxe: Don't update the response PSN unless it's going forwards

A client might post a read followed by a send. The partner receives
and acknowledges both transactions, posting an RCQ entry for the
send, but something goes wrong with the read ACK. When the client
retries the read, the partner's responder processes the duplicate
read but incorrectly resets the PSN to the value preceding the
original send. When the duplicate send arrives, the responder cannot
tell that it is a duplicate, so the responder generates a duplicate
RCQ entry, confusing the client.

Signed-off-by: Andrew Boyer <andrew.boyer@dell.com>
Reviewed-by: Yonatan Cohen <yonatanc@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rxe/rxe_resp.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rxe/rxe_resp.c b/drivers/infiniband/sw/rxe/rxe_resp.c
index dd3d88adc003..cb3fd4cb0daa 100644
--- a/drivers/infiniband/sw/rxe/rxe_resp.c
+++ b/drivers/infiniband/sw/rxe/rxe_resp.c
@@ -742,7 +742,8 @@ static enum resp_states read_reply(struct rxe_qp *qp,
 	} else {
 		qp->resp.res = NULL;
 		qp->resp.opcode = -1;
-		qp->resp.psn = res->cur_psn;
+		if (psn_compare(res->cur_psn, qp->resp.psn) >= 0)
+			qp->resp.psn = res->cur_psn;
 		state = RESPST_CLEANUP;
 	}
 
-- 
cgit v1.2.1


From d38eb801aa145aedf4b97e8e0bb2e65763aa6149 Mon Sep 17 00:00:00 2001
From: Andrew Boyer <andrew.boyer@dell.com>
Date: Wed, 23 Nov 2016 12:39:20 -0500
Subject: IB/rxe: Unblock loopback by moving skb_out increment

skb_out is decremented in rxe_skb_tx_dtor(), which is not called in the
loopback() path. Move the increment to the send() path rather than
rxe_xmit_packet().

Signed-off-by: Andrew Boyer <andrew.boyer@dell.com>
Acked-by: Moni Shoua <monis@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rxe/rxe_loc.h | 2 --
 drivers/infiniband/sw/rxe/rxe_net.c | 2 ++
 2 files changed, 2 insertions(+), 2 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rxe/rxe_loc.h b/drivers/infiniband/sw/rxe/rxe_loc.h
index 73849a5a91b3..efe4c6a35442 100644
--- a/drivers/infiniband/sw/rxe/rxe_loc.h
+++ b/drivers/infiniband/sw/rxe/rxe_loc.h
@@ -266,8 +266,6 @@ static inline int rxe_xmit_packet(struct rxe_dev *rxe, struct rxe_qp *qp,
 		return err;
 	}
 
-	atomic_inc(&qp->skb_out);
-
 	if ((qp_type(qp) != IB_QPT_RC) &&
 	    (pkt->mask & RXE_END_MASK)) {
 		pkt->wqe->state = wqe_state_done;
diff --git a/drivers/infiniband/sw/rxe/rxe_net.c b/drivers/infiniband/sw/rxe/rxe_net.c
index 4cb63780fa74..a576603304f7 100644
--- a/drivers/infiniband/sw/rxe/rxe_net.c
+++ b/drivers/infiniband/sw/rxe/rxe_net.c
@@ -459,6 +459,8 @@ static int send(struct rxe_dev *rxe, struct rxe_pkt_info *pkt,
 		return -EAGAIN;
 	}
 
+	if (pkt->qp)
+		atomic_inc(&pkt->qp->skb_out);
 	kfree_skb(skb);
 
 	return 0;
-- 
cgit v1.2.1


From d4fb59256ac03d84f68e36c430b58d6fc76dd651 Mon Sep 17 00:00:00 2001
From: Andrew Boyer <andrew.boyer@dell.com>
Date: Wed, 23 Nov 2016 12:39:21 -0500
Subject: IB/rxe: Add support for zero-byte operations

The last_psn algorithm fails in the zero-byte case: it calculates
first_psn = N, last_psn = N-1. This makes the operation unretryable since
the res structure will fail the (first_psn <= psn <= last_psn) test in
find_resource().

While here, use BTH_PSN_MASK to mask the calculated last_psn.

Signed-off-by: Andrew Boyer <andrew.boyer@dell.com>
Reviewed-by: Moni Shoua <monis@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rxe/rxe_mr.c   |  3 +++
 drivers/infiniband/sw/rxe/rxe_resp.c | 18 +++++++++++++++---
 2 files changed, 18 insertions(+), 3 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rxe/rxe_mr.c b/drivers/infiniband/sw/rxe/rxe_mr.c
index 1869152f1d23..d0faca294006 100644
--- a/drivers/infiniband/sw/rxe/rxe_mr.c
+++ b/drivers/infiniband/sw/rxe/rxe_mr.c
@@ -355,6 +355,9 @@ int rxe_mem_copy(struct rxe_mem *mem, u64 iova, void *addr, int length,
 	size_t			offset;
 	u32			crc = crcp ? (*crcp) : 0;
 
+	if (length == 0)
+		return 0;
+
 	if (mem->type == RXE_MEM_TYPE_DMA) {
 		u8 *src, *dest;
 
diff --git a/drivers/infiniband/sw/rxe/rxe_resp.c b/drivers/infiniband/sw/rxe/rxe_resp.c
index cb3fd4cb0daa..a5e9ce34171b 100644
--- a/drivers/infiniband/sw/rxe/rxe_resp.c
+++ b/drivers/infiniband/sw/rxe/rxe_resp.c
@@ -444,6 +444,13 @@ static enum resp_states check_rkey(struct rxe_qp *qp,
 		return RESPST_EXECUTE;
 	}
 
+	/* A zero-byte op is not required to set an addr or rkey. */
+	if ((pkt->mask & (RXE_READ_MASK | RXE_WRITE_OR_SEND)) &&
+	    (pkt->mask & RXE_RETH_MASK) &&
+	    reth_len(pkt) == 0) {
+		return RESPST_EXECUTE;
+	}
+
 	va	= qp->resp.va;
 	rkey	= qp->resp.rkey;
 	resid	= qp->resp.resid;
@@ -680,9 +687,14 @@ static enum resp_states read_reply(struct rxe_qp *qp,
 		res->read.va_org	= qp->resp.va;
 
 		res->first_psn		= req_pkt->psn;
-		res->last_psn		= req_pkt->psn +
-					  (reth_len(req_pkt) + mtu - 1) /
-					  mtu - 1;
+
+		if (reth_len(req_pkt)) {
+			res->last_psn	= (req_pkt->psn +
+					   (reth_len(req_pkt) + mtu - 1) /
+					   mtu - 1) & BTH_PSN_MASK;
+		} else {
+			res->last_psn	= res->first_psn;
+		}
 		res->cur_psn		= req_pkt->psn;
 
 		res->read.resid		= qp->resp.resid;
-- 
cgit v1.2.1


From accacb8f51c299965939ac56926d1c718e2691a1 Mon Sep 17 00:00:00 2001
From: Andrew Boyer <andrew.boyer@dell.com>
Date: Wed, 23 Nov 2016 12:39:22 -0500
Subject: IB/rxe: Add support for IB_CQ_REPORT_MISSED_EVENTS

Peek at the CQ after arming it so that we can return a hint.
This avoids missed completions due to a race between posting
CQEs and arming the CQ.

For example, CM teardown waits on MAD requests to complete with
ib_cq_poll_work(). Without this fix, the last completion might be
left on the CQ, hanging the kthread doing the teardown.

The console backtraces look like this:

[ 4199.911284] Call Trace:
[ 4199.911401]  [<ffffffff9657fe95>] schedule+0x35/0x80
[ 4199.911556]  [<ffffffff965830df>] schedule_timeout+0x22f/0x2c0
[ 4199.911727]  [<ffffffff9657f7a8>] ? __schedule+0x368/0xa20
[ 4199.911891]  [<ffffffff96580903>] wait_for_completion+0xb3/0x130
[ 4199.912067]  [<ffffffff960a17e0>] ? wake_up_q+0x70/0x70
[ 4199.912243]  [<ffffffffc074a06d>] cm_destroy_id+0x13d/0x450 [ib_cm]
[ 4199.912422]  [<ffffffff961615d5>] ? printk+0x57/0x73
[ 4199.912578]  [<ffffffffc074a390>] ib_destroy_cm_id+0x10/0x20 [ib_cm]
[ 4199.912759]  [<ffffffffc076098c>] rdma_destroy_id+0xac/0x340 [rdma_cm]
[ 4199.912941]  [<ffffffffc076f2cc>] 0xffffffffc076f2cc

Signed-off-by: Andrew Boyer <andrew.boyer@dell.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rxe/rxe_verbs.c | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rxe/rxe_verbs.c b/drivers/infiniband/sw/rxe/rxe_verbs.c
index 19841c863daf..de39b0a081de 100644
--- a/drivers/infiniband/sw/rxe/rxe_verbs.c
+++ b/drivers/infiniband/sw/rxe/rxe_verbs.c
@@ -1007,11 +1007,19 @@ static int rxe_peek_cq(struct ib_cq *ibcq, int wc_cnt)
 static int rxe_req_notify_cq(struct ib_cq *ibcq, enum ib_cq_notify_flags flags)
 {
 	struct rxe_cq *cq = to_rcq(ibcq);
+	unsigned long irq_flags;
+	int ret = 0;
 
+	spin_lock_irqsave(&cq->cq_lock, irq_flags);
 	if (cq->notify != IB_CQ_NEXT_COMP)
 		cq->notify = flags & IB_CQ_SOLICITED_MASK;
 
-	return 0;
+	if ((flags & IB_CQ_REPORT_MISSED_EVENTS) && !queue_empty(cq->queue))
+		ret = 1;
+
+	spin_unlock_irqrestore(&cq->cq_lock, irq_flags);
+
+	return ret;
 }
 
 static struct ib_mr *rxe_get_dma_mr(struct ib_pd *ibpd, int access)
-- 
cgit v1.2.1


From 5b9ea16c5488e577b26cc198ac63550b746ce3b9 Mon Sep 17 00:00:00 2001
From: Andrew Boyer <andrew.boyer@dell.com>
Date: Wed, 23 Nov 2016 12:39:23 -0500
Subject: IB/rxe: Fix ref leak in rxe_create_qp()

The udata->inlen error path needs to clean up the ref
added by rxe_alloc().

Signed-off-by: Andrew Boyer <andrew.boyer@dell.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rxe/rxe_verbs.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rxe/rxe_verbs.c b/drivers/infiniband/sw/rxe/rxe_verbs.c
index de39b0a081de..071430c594b7 100644
--- a/drivers/infiniband/sw/rxe/rxe_verbs.c
+++ b/drivers/infiniband/sw/rxe/rxe_verbs.c
@@ -564,7 +564,7 @@ static struct ib_qp *rxe_create_qp(struct ib_pd *ibpd,
 	if (udata) {
 		if (udata->inlen) {
 			err = -EINVAL;
-			goto err1;
+			goto err2;
 		}
 		qp->is_user = 1;
 	}
@@ -573,12 +573,13 @@ static struct ib_qp *rxe_create_qp(struct ib_pd *ibpd,
 
 	err = rxe_qp_from_init(rxe, qp, pd, init, udata, ibpd);
 	if (err)
-		goto err2;
+		goto err3;
 
 	return &qp->ibqp;
 
-err2:
+err3:
 	rxe_drop_index(qp);
+err2:
 	rxe_drop_ref(qp);
 err1:
 	return ERR_PTR(err);
-- 
cgit v1.2.1


From 5407f530122aa63cf304eb0874c938b3bdb8d3fb Mon Sep 17 00:00:00 2001
From: Andrew Boyer <andrew.boyer@dell.com>
Date: Wed, 23 Nov 2016 12:39:24 -0500
Subject: IB/rxe: Fix ref leak in duplicate_request()

A ref was added after the call to skb_clone().

Signed-off-by: Andrew Boyer <andrew.boyer@dell.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rxe/rxe_resp.c | 1 +
 1 file changed, 1 insertion(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rxe/rxe_resp.c b/drivers/infiniband/sw/rxe/rxe_resp.c
index a5e9ce34171b..8643797fb530 100644
--- a/drivers/infiniband/sw/rxe/rxe_resp.c
+++ b/drivers/infiniband/sw/rxe/rxe_resp.c
@@ -1145,6 +1145,7 @@ static enum resp_states duplicate_request(struct rxe_qp *qp,
 					     pkt, skb_copy);
 			if (rc) {
 				pr_err("Failed resending result. This flow is not handled - skb ignored\n");
+				rxe_drop_ref(qp);
 				kfree_skb(skb_copy);
 				rc = RESPST_CLEANUP;
 				goto out;
-- 
cgit v1.2.1


From 07bf9627d5f1c0334fc543a5435a31a3b5907944 Mon Sep 17 00:00:00 2001
From: Andrew Boyer <andrew.boyer@dell.com>
Date: Mon, 5 Dec 2016 08:43:20 -0500
Subject: IB/rxe: Wait for tasklets to finish before tearing down QP

The system may crash when a malformed request is received and
the error is detected by the responder.

NodeA: $ ibv_rc_pingpong -g 0 -d rxe0 -i 1 -n 1 -s 50000
NodeB: $ ibv_rc_pingpong -g 0 -d rxe0 -i 1 -n 1 -s 1024 <NodeA_ip>

The responder generates a receive error on node B since the incoming
SEND is oversized. If the client tears down the QP before the responder
or the completer finish running, a page fault may occur.

The fix makes the destroy operation spin until the tasks complete, which
appears to be original intent of the design.

Signed-off-by: Andrew Boyer <andrew.boyer@dell.com>
Reviewed-by: Yuval Shaia <yuval.shaia@oracle.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rxe/rxe_task.c | 19 +++++++++++++++++++
 drivers/infiniband/sw/rxe/rxe_task.h |  1 +
 2 files changed, 20 insertions(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rxe/rxe_task.c b/drivers/infiniband/sw/rxe/rxe_task.c
index 1e19bf828a6e..d2a14a1bdc7f 100644
--- a/drivers/infiniband/sw/rxe/rxe_task.c
+++ b/drivers/infiniband/sw/rxe/rxe_task.c
@@ -121,6 +121,7 @@ int rxe_init_task(void *obj, struct rxe_task *task,
 	task->arg	= arg;
 	task->func	= func;
 	snprintf(task->name, sizeof(task->name), "%s", name);
+	task->destroyed	= false;
 
 	tasklet_init(&task->tasklet, rxe_do_task, (unsigned long)task);
 
@@ -132,11 +133,29 @@ int rxe_init_task(void *obj, struct rxe_task *task,
 
 void rxe_cleanup_task(struct rxe_task *task)
 {
+	unsigned long flags;
+	bool idle;
+
+	/*
+	 * Mark the task, then wait for it to finish. It might be
+	 * running in a non-tasklet (direct call) context.
+	 */
+	task->destroyed = true;
+
+	do {
+		spin_lock_irqsave(&task->state_lock, flags);
+		idle = (task->state == TASK_STATE_START);
+		spin_unlock_irqrestore(&task->state_lock, flags);
+	} while (!idle);
+
 	tasklet_kill(&task->tasklet);
 }
 
 void rxe_run_task(struct rxe_task *task, int sched)
 {
+	if (task->destroyed)
+		return;
+
 	if (sched)
 		tasklet_schedule(&task->tasklet);
 	else
diff --git a/drivers/infiniband/sw/rxe/rxe_task.h b/drivers/infiniband/sw/rxe/rxe_task.h
index d14aa6daed05..08ff42d451c6 100644
--- a/drivers/infiniband/sw/rxe/rxe_task.h
+++ b/drivers/infiniband/sw/rxe/rxe_task.h
@@ -54,6 +54,7 @@ struct rxe_task {
 	int			(*func)(void *arg);
 	int			ret;
 	char			name[16];
+	bool			destroyed;
 };
 
 /*
-- 
cgit v1.2.1


From 37f69f43fb5aba4288d38ea32bbe0dfdb412c763 Mon Sep 17 00:00:00 2001
From: Andrew Boyer <andrew.boyer@dell.com>
Date: Mon, 5 Dec 2016 08:43:21 -0500
Subject: IB/rxe: Hold refs when running tasklets

It might be possible for all of a QP's references to be dropped
while one of that QP's tasklets is running.

For example, the completer might run during QP destroy.
If qp->valid is false, it will drop all of the packets on
the resp_pkts list, potentially removing the last reference.
Then it tries to advance the SQ consumer pointer. If the
SQ's buffer has already been destroyed, the system will
panic.

To be safe, hold a reference on the QP for the duration
of each tasklet.

Signed-off-by: Andrew Boyer <andrew.boyer@dell.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rxe/rxe_comp.c | 4 ++++
 drivers/infiniband/sw/rxe/rxe_req.c  | 5 ++++-
 drivers/infiniband/sw/rxe/rxe_resp.c | 3 +++
 3 files changed, 11 insertions(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rxe/rxe_comp.c b/drivers/infiniband/sw/rxe/rxe_comp.c
index d46c49b33b13..cd27cbde7652 100644
--- a/drivers/infiniband/sw/rxe/rxe_comp.c
+++ b/drivers/infiniband/sw/rxe/rxe_comp.c
@@ -511,6 +511,8 @@ int rxe_completer(void *arg)
 	struct rxe_pkt_info *pkt = NULL;
 	enum comp_state state;
 
+	rxe_add_ref(qp);
+
 	if (!qp->valid) {
 		while ((skb = skb_dequeue(&qp->resp_pkts))) {
 			rxe_drop_ref(qp);
@@ -740,11 +742,13 @@ exit:
 	/* we come here if we are done with processing and want the task to
 	 * exit from the loop calling us
 	 */
+	rxe_drop_ref(qp);
 	return -EAGAIN;
 
 done:
 	/* we come here if we have processed a packet we want the task to call
 	 * us again to see if there is anything else to do
 	 */
+	rxe_drop_ref(qp);
 	return 0;
 }
diff --git a/drivers/infiniband/sw/rxe/rxe_req.c b/drivers/infiniband/sw/rxe/rxe_req.c
index 205222909e53..b246653cf713 100644
--- a/drivers/infiniband/sw/rxe/rxe_req.c
+++ b/drivers/infiniband/sw/rxe/rxe_req.c
@@ -596,6 +596,8 @@ int rxe_requester(void *arg)
 	struct rxe_send_wqe rollback_wqe;
 	u32 rollback_psn;
 
+	rxe_add_ref(qp);
+
 next_wqe:
 	if (unlikely(!qp->valid || qp->req.state == QP_STATE_ERROR))
 		goto exit;
@@ -750,9 +752,10 @@ complete:
 		while (rxe_completer(qp) == 0)
 			;
 	}
-
+	rxe_drop_ref(qp);
 	return 0;
 
 exit:
+	rxe_drop_ref(qp);
 	return -EAGAIN;
 }
diff --git a/drivers/infiniband/sw/rxe/rxe_resp.c b/drivers/infiniband/sw/rxe/rxe_resp.c
index 8643797fb530..7a36ec9dbc0c 100644
--- a/drivers/infiniband/sw/rxe/rxe_resp.c
+++ b/drivers/infiniband/sw/rxe/rxe_resp.c
@@ -1212,6 +1212,8 @@ int rxe_responder(void *arg)
 	struct rxe_pkt_info *pkt = NULL;
 	int ret = 0;
 
+	rxe_add_ref(qp);
+
 	qp->resp.aeth_syndrome = AETH_ACK_UNLIMITED;
 
 	if (!qp->valid) {
@@ -1400,5 +1402,6 @@ int rxe_responder(void *arg)
 exit:
 	ret = -EAGAIN;
 done:
+	rxe_drop_ref(qp);
 	return ret;
 }
-- 
cgit v1.2.1


From 5e589171225b6aeac5eaca0b0887bd83dc9376d8 Mon Sep 17 00:00:00 2001
From: Thomas Huth <thuth@redhat.com>
Date: Wed, 5 Oct 2016 13:55:38 +0200
Subject: i40iw: Remove macros I40IW_STAG_KEY_FROM_STAG and
 I40IW_STAG_INDEX_FROM_STAG

The macros I40IW_STAG_KEY_FROM_STAG and I40IW_STAG_INDEX_FROM_STAG are
apparently bad - they are using the logical "&&" operation which
does not make sense here. It should have been a bitwise "&" instead.
Since the macros seem to be completely unused, let's simply remove
them so that nobody accidentially uses them in the future. And while
we're at it, also remove the unused macro I40IW_CREATE_STAG.

Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Leon Romanovsky <leonro@mellanox.com>
Acked-by: Faisal Latif <faisal.latif@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_user.h | 6 ------
 1 file changed, 6 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_user.h b/drivers/infiniband/hw/i40iw/i40iw_user.h
index 66263fced68f..80d9f464f65e 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_user.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_user.h
@@ -96,12 +96,6 @@ enum i40iw_device_capabilities_const {
 #define i40iw_physical_fragment u64
 #define i40iw_address_list u64 *
 
-#define I40IW_CREATE_STAG(index, key)       (((index) << 8) + (key))
-
-#define I40IW_STAG_KEY_FROM_STAG(stag)      ((stag) && 0x000000FF)
-
-#define I40IW_STAG_INDEX_FROM_STAG(stag)    (((stag) && 0xFFFFFF00) >> 8)
-
 #define	I40IW_MAX_MR_SIZE	0x10000000000L
 
 struct i40iw_qp_uk;
-- 
cgit v1.2.1


From 91c42b72f8e8b45961ff05a05009b644e6316ca2 Mon Sep 17 00:00:00 2001
From: Shiraz Saleem <shiraz.saleem@intel.com>
Date: Fri, 11 Nov 2016 10:55:41 -0600
Subject: i40iw: Use correct src address in memcpy to rdma stats counters

hw_stats is a pointer to i40_iw_dev_stats struct in i40iw_get_hw_stats().
Use hw_stats and not &hw_stats in the memcpy to copy the i40iw device stats
data into rdma_hw_stats counters.

Fixes: b40f4757daa1 ("IB/core: Make device counter infrastructure dynamic")

Cc: stable@vger.kernel.org # 4.7+
Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Faisal Latif <faisal.latif@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_verbs.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
index bc24086989e3..206d72b8a8bd 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
@@ -2646,7 +2646,7 @@ static int i40iw_get_hw_stats(struct ib_device *ibdev,
 			return -ENOSYS;
 	}
 
-	memcpy(&stats->value[0], &hw_stats, sizeof(*hw_stats));
+	memcpy(&stats->value[0], hw_stats, sizeof(*hw_stats));
 
 	return stats->num_counters;
 }
-- 
cgit v1.2.1


From f4a87ca12a1c203913a5cc889ec49b817a1f45fc Mon Sep 17 00:00:00 2001
From: Mustafa Ismail <mustafa.ismail@intel.com>
Date: Tue, 6 Dec 2016 15:49:30 -0600
Subject: i40iw: Fix double free of QP

A QP can be double freed if i40iw_cm_disconn() is
called while it is currently being freed by
i40iw_rem_ref(). The fix in i40iw_cm_disconn() will
first check if the QP is already freed before
making another request for the QP to be freed.

Signed-off-by: Mustafa Ismail <mustafa.ismail@intel.com>
Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw.h    |  2 +-
 drivers/infiniband/hw/i40iw/i40iw_cm.c | 20 ++++++++++++++++----
 drivers/infiniband/hw/i40iw/i40iw_hw.c |  4 +++-
 3 files changed, 20 insertions(+), 6 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw.h b/drivers/infiniband/hw/i40iw/i40iw.h
index ef188e6b1f89..51b828026b53 100644
--- a/drivers/infiniband/hw/i40iw/i40iw.h
+++ b/drivers/infiniband/hw/i40iw/i40iw.h
@@ -512,7 +512,7 @@ u32 i40iw_initialize_hw_resources(struct i40iw_device *iwdev);
 
 int i40iw_register_rdma_device(struct i40iw_device *iwdev);
 void i40iw_port_ibevent(struct i40iw_device *iwdev);
-int i40iw_cm_disconn(struct i40iw_qp *);
+void i40iw_cm_disconn(struct i40iw_qp *iwqp);
 void i40iw_cm_disconn_worker(void *);
 int mini_cm_recv_pkt(struct i40iw_cm_core *, struct i40iw_device *,
 		     struct sk_buff *);
diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.c b/drivers/infiniband/hw/i40iw/i40iw_cm.c
index 25af89a3cdce..ff95feaee105 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_cm.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_cm.c
@@ -3359,21 +3359,33 @@ static void i40iw_cm_init_tsa_conn(struct i40iw_qp *iwqp,
  * i40iw_cm_disconn - when a connection is being closed
  * @iwqp: associate qp for the connection
  */
-int i40iw_cm_disconn(struct i40iw_qp *iwqp)
+void i40iw_cm_disconn(struct i40iw_qp *iwqp)
 {
 	struct disconn_work *work;
 	struct i40iw_device *iwdev = iwqp->iwdev;
 	struct i40iw_cm_core *cm_core = &iwdev->cm_core;
+	unsigned long flags;
 
 	work = kzalloc(sizeof(*work), GFP_ATOMIC);
 	if (!work)
-		return -ENOMEM;	/* Timer will clean up */
-
+		return;	/* Timer will clean up */
+
+	spin_lock_irqsave(&iwdev->qptable_lock, flags);
+	if (!iwdev->qp_table[iwqp->ibqp.qp_num]) {
+		spin_unlock_irqrestore(&iwdev->qptable_lock, flags);
+		i40iw_debug(&iwdev->sc_dev, I40IW_DEBUG_CM,
+			    "%s qp_id %d is already freed\n",
+			     __func__, iwqp->ibqp.qp_num);
+		kfree(work);
+		return;
+	}
 	i40iw_add_ref(&iwqp->ibqp);
+	spin_unlock_irqrestore(&iwdev->qptable_lock, flags);
+
 	work->iwqp = iwqp;
 	INIT_WORK(&work->work, i40iw_disconnect_worker);
 	queue_work(cm_core->disconn_wq, &work->work);
-	return 0;
+	return;
 }
 
 /**
diff --git a/drivers/infiniband/hw/i40iw/i40iw_hw.c b/drivers/infiniband/hw/i40iw/i40iw_hw.c
index 5e2c16c725e3..b2854b11a240 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_hw.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_hw.c
@@ -308,7 +308,9 @@ void i40iw_process_aeq(struct i40iw_device *iwdev)
 			iwqp = iwdev->qp_table[info->qp_cq_id];
 			if (!iwqp) {
 				spin_unlock_irqrestore(&iwdev->qptable_lock, flags);
-				i40iw_pr_err("qp_id %d is already freed\n", info->qp_cq_id);
+				i40iw_debug(dev, I40IW_DEBUG_AEQ,
+					    "%s qp_id %d is already freed\n",
+					    __func__, info->qp_cq_id);
 				continue;
 			}
 			i40iw_add_ref(&iwqp->ibqp);
-- 
cgit v1.2.1


From 1cda28bb5b503bab734072d97a41b2e7eda6b6b9 Mon Sep 17 00:00:00 2001
From: Shiraz Saleem <shiraz.saleem@intel.com>
Date: Tue, 6 Dec 2016 15:49:31 -0600
Subject: i40iw: Fix QP flush to not hang on empty queues or failure

When flush QP and there are no pending work requests, signal completion
to unblock i40iw_drain_sq and i40iw_drain_rq which are waiting on
completion for iwqp->sq_drained and iwqp->sq_drained respectively.
Also, signal completion if flush QP fails to prevent the drain SQ or RQ
from being blocked indefintely.

Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw.h    |  9 ++++++---
 drivers/infiniband/hw/i40iw/i40iw_hw.c | 26 ++++++++++++++++++++++++--
 2 files changed, 30 insertions(+), 5 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw.h b/drivers/infiniband/hw/i40iw/i40iw.h
index 51b828026b53..2aab85bbcbc7 100644
--- a/drivers/infiniband/hw/i40iw/i40iw.h
+++ b/drivers/infiniband/hw/i40iw/i40iw.h
@@ -112,9 +112,12 @@
 #define I40IW_DRV_OPT_MCAST_LOGPORT_MAP    0x00000800
 
 #define IW_HMC_OBJ_TYPE_NUM ARRAY_SIZE(iw_hmc_obj_types)
-#define IW_CFG_FPM_QP_COUNT		32768
-#define I40IW_MAX_PAGES_PER_FMR		512
-#define I40IW_MIN_PAGES_PER_FMR		1
+#define IW_CFG_FPM_QP_COUNT               32768
+#define I40IW_MAX_PAGES_PER_FMR           512
+#define I40IW_MIN_PAGES_PER_FMR           1
+#define I40IW_CQP_COMPL_RQ_WQE_FLUSHED    2
+#define I40IW_CQP_COMPL_SQ_WQE_FLUSHED    3
+#define I40IW_CQP_COMPL_RQ_SQ_WQE_FLUSHED 4
 
 #define I40IW_MTU_TO_MSS		40
 #define I40IW_DEFAULT_MSS		1460
diff --git a/drivers/infiniband/hw/i40iw/i40iw_hw.c b/drivers/infiniband/hw/i40iw/i40iw_hw.c
index b2854b11a240..4394a6713bdf 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_hw.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_hw.c
@@ -622,6 +622,7 @@ enum i40iw_status_code i40iw_hw_flush_wqes(struct i40iw_device *iwdev,
 	struct i40iw_qp_flush_info *hw_info;
 	struct i40iw_cqp_request *cqp_request;
 	struct cqp_commands_info *cqp_info;
+	struct i40iw_qp *iwqp = (struct i40iw_qp *)qp->back_qp;
 
 	cqp_request = i40iw_get_cqp_request(&iwdev->cqp, wait);
 	if (!cqp_request)
@@ -636,9 +637,30 @@ enum i40iw_status_code i40iw_hw_flush_wqes(struct i40iw_device *iwdev,
 	cqp_info->in.u.qp_flush_wqes.qp = qp;
 	cqp_info->in.u.qp_flush_wqes.scratch = (uintptr_t)cqp_request;
 	status = i40iw_handle_cqp_op(iwdev, cqp_request);
-	if (status)
+	if (status) {
 		i40iw_pr_err("CQP-OP Flush WQE's fail");
-	return status;
+		complete(&iwqp->sq_drained);
+		complete(&iwqp->rq_drained);
+		return status;
+	}
+	if (!cqp_request->compl_info.maj_err_code) {
+		switch (cqp_request->compl_info.min_err_code) {
+		case I40IW_CQP_COMPL_RQ_WQE_FLUSHED:
+			complete(&iwqp->sq_drained);
+			break;
+		case I40IW_CQP_COMPL_SQ_WQE_FLUSHED:
+			complete(&iwqp->rq_drained);
+			break;
+		case I40IW_CQP_COMPL_RQ_SQ_WQE_FLUSHED:
+			break;
+		default:
+			complete(&iwqp->sq_drained);
+			complete(&iwqp->rq_drained);
+			break;
+		}
+	}
+
+	return 0;
 }
 
 /**
-- 
cgit v1.2.1


From fd90d4d4c2dc815ef5a5f5d50a9c65c266c68ace Mon Sep 17 00:00:00 2001
From: Mustafa Ismail <mustafa.ismail@intel.com>
Date: Tue, 6 Dec 2016 15:49:32 -0600
Subject: i40iw: Fix memory leak in CQP destroy when in reset

On a device close, the control QP (CQP) is destroyed by calling
cqp_destroy which destroys the CQP and frees its SD buffer memory.
However, if the reset flag is true, cqp_destroy is never called and
leads to a memory leak on SD buffer memory. Fix this by always calling
cqp_destroy, on device close, regardless of reset. The exception to this
when CQP create fails. In this case, the SD buffer memory is already
freed on an error check and there is no need to call cqp_destroy.

Signed-off-by: Mustafa Ismail <mustafa.ismail@intel.com>
Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_main.c | 9 +++------
 1 file changed, 3 insertions(+), 6 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_main.c b/drivers/infiniband/hw/i40iw/i40iw_main.c
index 4ce05b8e8f0a..85d8fa63d5b5 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_main.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_main.c
@@ -237,14 +237,11 @@ static irqreturn_t i40iw_irq_handler(int irq, void *data)
  */
 static void i40iw_destroy_cqp(struct i40iw_device *iwdev, bool free_hwcqp)
 {
-	enum i40iw_status_code status = 0;
 	struct i40iw_sc_dev *dev = &iwdev->sc_dev;
 	struct i40iw_cqp *cqp = &iwdev->cqp;
 
-	if (free_hwcqp && dev->cqp_ops->cqp_destroy)
-		status = dev->cqp_ops->cqp_destroy(dev->cqp);
-	if (status)
-		i40iw_pr_err("destroy cqp failed");
+	if (free_hwcqp)
+		dev->cqp_ops->cqp_destroy(dev->cqp);
 
 	i40iw_free_dma_mem(dev->hw, &cqp->sq);
 	kfree(cqp->scratch_array);
@@ -1475,7 +1472,7 @@ static void i40iw_deinit_device(struct i40iw_device *iwdev, bool reset)
 		i40iw_del_hmc_objects(dev, dev->hmc_info, true, reset);
 		/* fallthrough */
 	case CQP_CREATED:
-		i40iw_destroy_cqp(iwdev, !reset);
+		i40iw_destroy_cqp(iwdev, true);
 		/* fallthrough */
 	case INITIAL_STATE:
 		i40iw_cleanup_cm_core(&iwdev->cm_core);
-- 
cgit v1.2.1


From d627b506313c887e7159580cde926f5d14797aaa Mon Sep 17 00:00:00 2001
From: Shiraz Saleem <shiraz.saleem@intel.com>
Date: Tue, 6 Dec 2016 15:49:33 -0600
Subject: i40iw: Fix race condition in terminate timer's handler

Add a QP reference when terminate timer is started to ensure
the destroy QP doesn't race ahead to free the QP while it is being
referenced in the terminate timer's handler.

Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_cm.c    | 2 +-
 drivers/infiniband/hw/i40iw/i40iw_utils.c | 5 ++++-
 drivers/infiniband/hw/i40iw/i40iw_verbs.c | 2 +-
 3 files changed, 6 insertions(+), 3 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.c b/drivers/infiniband/hw/i40iw/i40iw_cm.c
index ff95feaee105..a217d2f34914 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_cm.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_cm.c
@@ -3471,7 +3471,7 @@ static void i40iw_cm_disconn_true(struct i40iw_qp *iwqp)
 		 *terminate-handler to issue cm_disconn which can re-free
 		 *a QP even after its refcnt=0.
 		 */
-		del_timer(&iwqp->terminate_timer);
+		i40iw_terminate_del_timer(qp);
 		if (!iwqp->flush_issued) {
 			iwqp->flush_issued = 1;
 			issue_flush = 1;
diff --git a/drivers/infiniband/hw/i40iw/i40iw_utils.c b/drivers/infiniband/hw/i40iw/i40iw_utils.c
index 4a08ffb75d2e..7d4af7759970 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_utils.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_utils.c
@@ -823,6 +823,7 @@ static void i40iw_terminate_timeout(unsigned long context)
 	struct i40iw_sc_qp *qp = (struct i40iw_sc_qp *)&iwqp->sc_qp;
 
 	i40iw_terminate_done(qp, 1);
+	i40iw_rem_ref(&iwqp->ibqp);
 }
 
 /**
@@ -834,6 +835,7 @@ void i40iw_terminate_start_timer(struct i40iw_sc_qp *qp)
 	struct i40iw_qp *iwqp;
 
 	iwqp = (struct i40iw_qp *)qp->back_qp;
+	i40iw_add_ref(&iwqp->ibqp);
 	init_timer(&iwqp->terminate_timer);
 	iwqp->terminate_timer.function = i40iw_terminate_timeout;
 	iwqp->terminate_timer.expires = jiffies + HZ;
@@ -850,7 +852,8 @@ void i40iw_terminate_del_timer(struct i40iw_sc_qp *qp)
 	struct i40iw_qp *iwqp;
 
 	iwqp = (struct i40iw_qp *)qp->back_qp;
-	del_timer(&iwqp->terminate_timer);
+	if (del_timer(&iwqp->terminate_timer))
+		i40iw_rem_ref(&iwqp->ibqp);
 }
 
 /**
diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
index 206d72b8a8bd..18526e6f9d85 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
@@ -959,7 +959,7 @@ int i40iw_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
 				goto exit;
 			}
 			if (iwqp->sc_qp.term_flags)
-				del_timer(&iwqp->terminate_timer);
+				i40iw_terminate_del_timer(&iwqp->sc_qp);
 			info.next_iwarp_state = I40IW_QP_STATE_ERROR;
 			if ((iwqp->hw_tcp_state > I40IW_TCP_STATE_CLOSED) &&
 			    iwdev->iw_status &&
-- 
cgit v1.2.1


From 6b0805c25657f9b702607ed4617c2821343158c0 Mon Sep 17 00:00:00 2001
From: Mustafa Ismail <mustafa.ismail@intel.com>
Date: Tue, 6 Dec 2016 15:49:34 -0600
Subject: i40iw: Assign MSS only when it is a new MTU

Currently we are changing the MSS regardless of whether
there is a change or not in MTU. Fix to make the
assignment of MSS dependent on an MTU change.

Signed-off-by: Mustafa Ismail <mustafa.ismail@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_main.c | 2 ++
 1 file changed, 2 insertions(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_main.c b/drivers/infiniband/hw/i40iw/i40iw_main.c
index 85d8fa63d5b5..cf9d288a79b3 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_main.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_main.c
@@ -1724,6 +1724,8 @@ static void i40iw_l2param_change(struct i40e_info *ldev, struct i40e_client *cli
 	for (i = 0; i < I40E_CLIENT_MAX_USER_PRIORITY; i++)
 		l2params->qs_handle_list[i] = params->qos.prio_qos[i].qs_handle;
 
+	l2params->mss = (params->mtu) ? params->mtu - I40IW_MTU_TO_MSS : iwdev->mss;
+
 	INIT_WORK(&work->work, i40iw_l2params_worker);
 	queue_work(iwdev->param_wq, &work->work);
 }
-- 
cgit v1.2.1


From 0cc0d851ccf1746466822c1b7ce02c980406d57f Mon Sep 17 00:00:00 2001
From: Mustafa Ismail <mustafa.ismail@intel.com>
Date: Tue, 6 Dec 2016 15:49:35 -0600
Subject: i40iw: Fix incorrect check for error

In i40iw_ieq_handle_partial() the check for !status is incorrect.

Signed-off-by: Mustafa Ismail <mustafa.ismail@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw_puda.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw_puda.c b/drivers/infiniband/hw/i40iw/i40iw_puda.c
index 7541b0dada59..c3d28baf8e3a 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_puda.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_puda.c
@@ -1132,7 +1132,7 @@ static enum i40iw_status_code i40iw_ieq_handle_partial(struct i40iw_puda_rsrc *i
 	list_add(&buf->list, &pbufl);
 
 	status = i40iw_ieq_create_pbufl(pfpdu, rxlist, &pbufl, buf, fpdu_len);
-	if (!status)
+	if (status)
 		goto error;
 
 	txbuf = i40iw_puda_get_bufpool(ieq);
-- 
cgit v1.2.1


From d6f7bbcc2e419c8afd4a426af78b3dac44632268 Mon Sep 17 00:00:00 2001
From: Henry Orosco <henry.orosco@intel.com>
Date: Tue, 6 Dec 2016 16:16:20 -0600
Subject: i40iw: Reorganize structures to align with HW capabilities

Some resources are incorrectly organized and at odds with
HW capabilities. Specifically, ILQ, IEQ, QPs, MSS, QOS
and statistics belong in a VSI.

Signed-off-by: Faisal Latif <faisal.latif@intel.com>
Signed-off-by: Mustafa Ismail <mustafa.ismail@intel.com>
Signed-off-by: Henry Orosco <henry.orosco@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/i40iw/i40iw.h          |   7 +-
 drivers/infiniband/hw/i40iw/i40iw_cm.c       |  36 +-
 drivers/infiniband/hw/i40iw/i40iw_ctrl.c     | 527 +++++++++++++++------------
 drivers/infiniband/hw/i40iw/i40iw_d.h        |  21 +-
 drivers/infiniband/hw/i40iw/i40iw_hw.c       |   4 +-
 drivers/infiniband/hw/i40iw/i40iw_main.c     |  53 ++-
 drivers/infiniband/hw/i40iw/i40iw_osdep.h    |   6 +-
 drivers/infiniband/hw/i40iw/i40iw_p.h        |  23 +-
 drivers/infiniband/hw/i40iw/i40iw_puda.c     | 268 ++++++++------
 drivers/infiniband/hw/i40iw/i40iw_puda.h     |  20 +-
 drivers/infiniband/hw/i40iw/i40iw_type.h     |  79 ++--
 drivers/infiniband/hw/i40iw/i40iw_utils.c    | 150 +++++++-
 drivers/infiniband/hw/i40iw/i40iw_verbs.c    |  17 +-
 drivers/infiniband/hw/i40iw/i40iw_virtchnl.c |  29 +-
 14 files changed, 775 insertions(+), 465 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/i40iw/i40iw.h b/drivers/infiniband/hw/i40iw/i40iw.h
index 2aab85bbcbc7..da2eb5a281fa 100644
--- a/drivers/infiniband/hw/i40iw/i40iw.h
+++ b/drivers/infiniband/hw/i40iw/i40iw.h
@@ -236,6 +236,7 @@ struct i40iw_device {
 	struct net_device *netdev;
 	wait_queue_head_t vchnl_waitq;
 	struct i40iw_sc_dev sc_dev;
+	struct i40iw_sc_vsi vsi;
 	struct i40iw_handler *hdl;
 	struct i40e_info *ldev;
 	struct i40e_client *client;
@@ -289,7 +290,6 @@ struct i40iw_device {
 	u32 sd_type;
 	struct workqueue_struct *param_wq;
 	atomic_t params_busy;
-	u32 mss;
 	enum init_completion_state init_state;
 	u16 mac_ip_table_idx;
 	atomic_t vchnl_msgs;
@@ -525,6 +525,7 @@ enum i40iw_status_code i40iw_handle_cqp_op(struct i40iw_device *iwdev,
 enum i40iw_status_code i40iw_add_mac_addr(struct i40iw_device *iwdev,
 					  u8 *mac_addr, u8 *mac_index);
 int i40iw_modify_qp(struct ib_qp *, struct ib_qp_attr *, int, struct ib_udata *);
+void i40iw_cq_wq_destroy(struct i40iw_device *iwdev, struct i40iw_sc_cq *cq);
 
 void i40iw_rem_pdusecount(struct i40iw_pd *iwpd, struct i40iw_device *iwdev);
 void i40iw_add_pdusecount(struct i40iw_pd *iwpd);
@@ -542,8 +543,8 @@ enum i40iw_status_code i40iw_manage_qhash(struct i40iw_device *iwdev,
 					  enum i40iw_quad_hash_manage_type mtype,
 					  void *cmnode,
 					  bool wait);
-void i40iw_receive_ilq(struct i40iw_sc_dev *dev, struct i40iw_puda_buf *rbuf);
-void i40iw_free_sqbuf(struct i40iw_sc_dev *dev, void *bufp);
+void i40iw_receive_ilq(struct i40iw_sc_vsi *vsi, struct i40iw_puda_buf *rbuf);
+void i40iw_free_sqbuf(struct i40iw_sc_vsi *vsi, void *bufp);
 void i40iw_free_qp_resources(struct i40iw_device *iwdev,
 			     struct i40iw_qp *iwqp,
 			     u32 qp_num);
diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.c b/drivers/infiniband/hw/i40iw/i40iw_cm.c
index a217d2f34914..e4820bef90be 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_cm.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_cm.c
@@ -68,13 +68,13 @@ static void i40iw_disconnect_worker(struct work_struct *work);
 
 /**
  * i40iw_free_sqbuf - put back puda buffer if refcount = 0
- * @dev: FPK device
+ * @vsi: pointer to vsi structure
  * @buf: puda buffer to free
  */
-void i40iw_free_sqbuf(struct i40iw_sc_dev *dev, void *bufp)
+void i40iw_free_sqbuf(struct i40iw_sc_vsi *vsi, void *bufp)
 {
 	struct i40iw_puda_buf *buf = (struct i40iw_puda_buf *)bufp;
-	struct i40iw_puda_rsrc *ilq = dev->ilq;
+	struct i40iw_puda_rsrc *ilq = vsi->ilq;
 
 	if (!atomic_dec_return(&buf->refcount))
 		i40iw_puda_ret_bufpool(ilq, buf);
@@ -337,13 +337,13 @@ static struct i40iw_cm_event *i40iw_create_event(struct i40iw_cm_node *cm_node,
  */
 static void i40iw_free_retrans_entry(struct i40iw_cm_node *cm_node)
 {
-	struct i40iw_sc_dev *dev = cm_node->dev;
+	struct i40iw_device *iwdev = cm_node->iwdev;
 	struct i40iw_timer_entry *send_entry;
 
 	send_entry = cm_node->send_entry;
 	if (send_entry) {
 		cm_node->send_entry = NULL;
-		i40iw_free_sqbuf(dev, (void *)send_entry->sqbuf);
+		i40iw_free_sqbuf(&iwdev->vsi, (void *)send_entry->sqbuf);
 		kfree(send_entry);
 		atomic_dec(&cm_node->ref_count);
 	}
@@ -377,7 +377,7 @@ static struct i40iw_puda_buf *i40iw_form_cm_frame(struct i40iw_cm_node *cm_node,
 						  u8 flags)
 {
 	struct i40iw_puda_buf *sqbuf;
-	struct i40iw_sc_dev *dev = cm_node->dev;
+	struct i40iw_sc_vsi *vsi = &cm_node->iwdev->vsi;
 	u8 *buf;
 
 	struct tcphdr *tcph;
@@ -391,7 +391,7 @@ static struct i40iw_puda_buf *i40iw_form_cm_frame(struct i40iw_cm_node *cm_node,
 	u32 hdr_len = 0;
 	u16 vtag;
 
-	sqbuf = i40iw_puda_get_bufpool(dev->ilq);
+	sqbuf = i40iw_puda_get_bufpool(vsi->ilq);
 	if (!sqbuf)
 		return NULL;
 	buf = sqbuf->mem.va;
@@ -1059,7 +1059,7 @@ int i40iw_schedule_cm_timer(struct i40iw_cm_node *cm_node,
 			    int send_retrans,
 			    int close_when_complete)
 {
-	struct i40iw_sc_dev *dev = cm_node->dev;
+	struct i40iw_sc_vsi *vsi = &cm_node->iwdev->vsi;
 	struct i40iw_cm_core *cm_core = cm_node->cm_core;
 	struct i40iw_timer_entry *new_send;
 	int ret = 0;
@@ -1068,7 +1068,7 @@ int i40iw_schedule_cm_timer(struct i40iw_cm_node *cm_node,
 
 	new_send = kzalloc(sizeof(*new_send), GFP_ATOMIC);
 	if (!new_send) {
-		i40iw_free_sqbuf(cm_node->dev, (void *)sqbuf);
+		i40iw_free_sqbuf(vsi, (void *)sqbuf);
 		return -ENOMEM;
 	}
 	new_send->retrycount = I40IW_DEFAULT_RETRYS;
@@ -1083,7 +1083,7 @@ int i40iw_schedule_cm_timer(struct i40iw_cm_node *cm_node,
 		new_send->timetosend += (HZ / 10);
 		if (cm_node->close_entry) {
 			kfree(new_send);
-			i40iw_free_sqbuf(cm_node->dev, (void *)sqbuf);
+			i40iw_free_sqbuf(vsi, (void *)sqbuf);
 			i40iw_pr_err("already close entry\n");
 			return -EINVAL;
 		}
@@ -1098,7 +1098,7 @@ int i40iw_schedule_cm_timer(struct i40iw_cm_node *cm_node,
 		new_send->timetosend = jiffies + I40IW_RETRY_TIMEOUT;
 
 		atomic_inc(&sqbuf->refcount);
-		i40iw_puda_send_buf(dev->ilq, sqbuf);
+		i40iw_puda_send_buf(vsi->ilq, sqbuf);
 		if (!send_retrans) {
 			i40iw_cleanup_retrans_entry(cm_node);
 			if (close_when_complete)
@@ -1195,6 +1195,7 @@ static void i40iw_cm_timer_tick(unsigned long pass)
 	struct i40iw_cm_node *cm_node;
 	struct i40iw_timer_entry *send_entry, *close_entry;
 	struct list_head *list_core_temp;
+	struct i40iw_sc_vsi *vsi;
 	struct list_head *list_node;
 	struct i40iw_cm_core *cm_core = (struct i40iw_cm_core *)pass;
 	u32 settimer = 0;
@@ -1270,9 +1271,10 @@ static void i40iw_cm_timer_tick(unsigned long pass)
 		cm_node->cm_core->stats_pkt_retrans++;
 		spin_unlock_irqrestore(&cm_node->retrans_list_lock, flags);
 
+		vsi = &cm_node->iwdev->vsi;
 		dev = cm_node->dev;
 		atomic_inc(&send_entry->sqbuf->refcount);
-		i40iw_puda_send_buf(dev->ilq, send_entry->sqbuf);
+		i40iw_puda_send_buf(vsi->ilq, send_entry->sqbuf);
 		spin_lock_irqsave(&cm_node->retrans_list_lock, flags);
 		if (send_entry->send_retrans) {
 			send_entry->retranscount--;
@@ -1373,10 +1375,11 @@ int i40iw_send_syn(struct i40iw_cm_node *cm_node, u32 sendack)
 static void i40iw_send_ack(struct i40iw_cm_node *cm_node)
 {
 	struct i40iw_puda_buf *sqbuf;
+	struct i40iw_sc_vsi *vsi = &cm_node->iwdev->vsi;
 
 	sqbuf = i40iw_form_cm_frame(cm_node, NULL, NULL, NULL, SET_ACK);
 	if (sqbuf)
-		i40iw_puda_send_buf(cm_node->dev->ilq, sqbuf);
+		i40iw_puda_send_buf(vsi->ilq, sqbuf);
 	else
 		i40iw_pr_err("no sqbuf\n");
 }
@@ -2179,7 +2182,7 @@ static struct i40iw_cm_node *i40iw_make_cm_node(
 			I40IW_CM_DEFAULT_RCV_WND_SCALED >> I40IW_CM_DEFAULT_RCV_WND_SCALE;
 	ts = current_kernel_time();
 	cm_node->tcp_cntxt.loc_seq_num = ts.tv_nsec;
-	cm_node->tcp_cntxt.mss = iwdev->mss;
+	cm_node->tcp_cntxt.mss = iwdev->vsi.mss;
 
 	cm_node->iwdev = iwdev;
 	cm_node->dev = &iwdev->sc_dev;
@@ -3059,10 +3062,10 @@ static int i40iw_cm_close(struct i40iw_cm_node *cm_node)
 /**
  * i40iw_receive_ilq - recv an ETHERNET packet, and process it
  * through CM
- * @dev: FPK dev struct
+ * @vsi: pointer to the vsi structure
  * @rbuf: receive buffer
  */
-void i40iw_receive_ilq(struct i40iw_sc_dev *dev, struct i40iw_puda_buf *rbuf)
+void i40iw_receive_ilq(struct i40iw_sc_vsi *vsi, struct i40iw_puda_buf *rbuf)
 {
 	struct i40iw_cm_node *cm_node;
 	struct i40iw_cm_listener *listener;
@@ -3070,6 +3073,7 @@ void i40iw_receive_ilq(struct i40iw_sc_dev *dev, struct i40iw_puda_buf *rbuf)
 	struct ipv6hdr *ip6h;
 	struct tcphdr *tcph;
 	struct i40iw_cm_info cm_info;
+	struct i40iw_sc_dev *dev = vsi->dev;
 	struct i40iw_device *iwdev = (struct i40iw_device *)dev->back_dev;
 	struct i40iw_cm_core *cm_core = &iwdev->cm_core;
 	struct vlan_ethhdr *ethh;
diff --git a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
index a13503715f20..392f78384a60 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c
@@ -103,6 +103,7 @@ static enum i40iw_status_code i40iw_cqp_poll_registers(
 		if (newtail != tail) {
 			/* SUCCESS */
 			I40IW_RING_MOVE_TAIL(cqp->sq_ring);
+			cqp->dev->cqp_cmd_stats[OP_COMPLETED_COMMANDS]++;
 			return 0;
 		}
 		udelay(I40IW_SLEEP_COUNT);
@@ -276,11 +277,12 @@ static struct i40iw_sc_qp *i40iw_get_qp(struct list_head *head, struct i40iw_sc_
 
 /**
  * i40iw_change_l2params - given the new l2 parameters, change all qp
- * @dev: IWARP device pointer
+ * @vsi: pointer to the vsi structure
  * @l2params: New paramaters from l2
  */
-void i40iw_change_l2params(struct i40iw_sc_dev *dev, struct i40iw_l2params *l2params)
+void i40iw_change_l2params(struct i40iw_sc_vsi *vsi, struct i40iw_l2params *l2params)
 {
+	struct i40iw_sc_dev *dev = vsi->dev;
 	struct i40iw_sc_qp *qp = NULL;
 	bool qs_handle_change = false;
 	bool mss_change = false;
@@ -288,20 +290,20 @@ void i40iw_change_l2params(struct i40iw_sc_dev *dev, struct i40iw_l2params *l2pa
 	u16 qs_handle;
 	int i;
 
-	if (dev->mss != l2params->mss) {
+	if (vsi->mss != l2params->mss) {
 		mss_change = true;
-		dev->mss = l2params->mss;
+		vsi->mss = l2params->mss;
 	}
 
 	i40iw_fill_qos_list(l2params->qs_handle_list);
 	for (i = 0; i < I40IW_MAX_USER_PRIORITY; i++) {
 		qs_handle = l2params->qs_handle_list[i];
-		if (dev->qos[i].qs_handle != qs_handle)
+		if (vsi->qos[i].qs_handle != qs_handle)
 			qs_handle_change = true;
 		else if (!mss_change)
 			continue;       /* no MSS nor qs handle change */
-		spin_lock_irqsave(&dev->qos[i].lock, flags);
-		qp = i40iw_get_qp(&dev->qos[i].qplist, qp);
+		spin_lock_irqsave(&vsi->qos[i].lock, flags);
+		qp = i40iw_get_qp(&vsi->qos[i].qplist, qp);
 		while (qp) {
 			if (mss_change)
 				i40iw_qp_mss_modify(dev, qp);
@@ -310,43 +312,45 @@ void i40iw_change_l2params(struct i40iw_sc_dev *dev, struct i40iw_l2params *l2pa
 				/* issue cqp suspend command */
 				i40iw_qp_suspend_resume(dev, qp, true);
 			}
-			qp = i40iw_get_qp(&dev->qos[i].qplist, qp);
+			qp = i40iw_get_qp(&vsi->qos[i].qplist, qp);
 		}
-		spin_unlock_irqrestore(&dev->qos[i].lock, flags);
-		dev->qos[i].qs_handle = qs_handle;
+		spin_unlock_irqrestore(&vsi->qos[i].lock, flags);
+		vsi->qos[i].qs_handle = qs_handle;
 	}
 }
 
 /**
  * i40iw_qp_rem_qos - remove qp from qos lists during destroy qp
- * @dev: IWARP device pointer
  * @qp: qp to be removed from qos
  */
-static void i40iw_qp_rem_qos(struct i40iw_sc_dev *dev, struct i40iw_sc_qp *qp)
+static void i40iw_qp_rem_qos(struct i40iw_sc_qp *qp)
 {
+	struct i40iw_sc_vsi *vsi = qp->vsi;
 	unsigned long flags;
 
 	if (!qp->on_qoslist)
 		return;
-	spin_lock_irqsave(&dev->qos[qp->user_pri].lock, flags);
+	spin_lock_irqsave(&vsi->qos[qp->user_pri].lock, flags);
 	list_del(&qp->list);
-	spin_unlock_irqrestore(&dev->qos[qp->user_pri].lock, flags);
+	spin_unlock_irqrestore(&vsi->qos[qp->user_pri].lock, flags);
 }
 
 /**
  * i40iw_qp_add_qos - called during setctx fot qp to be added to qos
- * @dev: IWARP device pointer
  * @qp: qp to be added to qos
  */
-void i40iw_qp_add_qos(struct i40iw_sc_dev *dev, struct i40iw_sc_qp *qp)
+void i40iw_qp_add_qos(struct i40iw_sc_qp *qp)
 {
+	struct i40iw_sc_vsi *vsi = qp->vsi;
 	unsigned long flags;
 
-	spin_lock_irqsave(&dev->qos[qp->user_pri].lock, flags);
-	qp->qs_handle = dev->qos[qp->user_pri].qs_handle;
-	list_add(&qp->list, &dev->qos[qp->user_pri].qplist);
+	if (qp->on_qoslist)
+		return;
+	spin_lock_irqsave(&vsi->qos[qp->user_pri].lock, flags);
+	qp->qs_handle = vsi->qos[qp->user_pri].qs_handle;
+	list_add(&qp->list, &vsi->qos[qp->user_pri].qplist);
 	qp->on_qoslist = true;
-	spin_unlock_irqrestore(&dev->qos[qp->user_pri].lock, flags);
+	spin_unlock_irqrestore(&vsi->qos[qp->user_pri].lock, flags);
 }
 
 /**
@@ -419,6 +423,9 @@ static enum i40iw_status_code i40iw_sc_cqp_init(struct i40iw_sc_cqp *cqp,
 	info->dev->cqp = cqp;
 
 	I40IW_RING_INIT(cqp->sq_ring, cqp->sq_size);
+	cqp->dev->cqp_cmd_stats[OP_REQUESTED_COMMANDS] = 0;
+	cqp->dev->cqp_cmd_stats[OP_COMPLETED_COMMANDS] = 0;
+
 	i40iw_debug(cqp->dev, I40IW_DEBUG_WQE,
 		    "%s: sq_size[%04d] hw_sq_size[%04d] sq_base[%p] sq_pa[%llxh] cqp[%p] polarity[x%04X]\n",
 		    __func__, cqp->sq_size, cqp->hw_sq_size,
@@ -546,6 +553,7 @@ u64 *i40iw_sc_cqp_get_next_send_wqe(struct i40iw_sc_cqp *cqp, u64 scratch)
 		return NULL;
 	}
 	I40IW_ATOMIC_RING_MOVE_HEAD(cqp->sq_ring, wqe_idx, ret_code);
+	cqp->dev->cqp_cmd_stats[OP_REQUESTED_COMMANDS]++;
 	if (ret_code)
 		return NULL;
 	if (!wqe_idx)
@@ -681,6 +689,8 @@ static enum i40iw_status_code i40iw_sc_ccq_get_cqe_info(
 		      I40IW_RING_GETCURRENT_HEAD(ccq->cq_uk.cq_ring));
 	wmb(); /* write shadow area before tail */
 	I40IW_RING_MOVE_TAIL(cqp->sq_ring);
+	ccq->dev->cqp_cmd_stats[OP_COMPLETED_COMMANDS]++;
+
 	return ret_code;
 }
 
@@ -1173,6 +1183,7 @@ static enum i40iw_status_code i40iw_sc_manage_qhash_table_entry(
 	u64 qw1 = 0;
 	u64 qw2 = 0;
 	u64 temp;
+	struct i40iw_sc_vsi *vsi = info->vsi;
 
 	wqe = i40iw_sc_cqp_get_next_send_wqe(cqp, scratch);
 	if (!wqe)
@@ -1204,7 +1215,7 @@ static enum i40iw_status_code i40iw_sc_manage_qhash_table_entry(
 			      LS_64(info->dest_ip[2], I40IW_CQPSQ_QHASH_ADDR2) |
 			      LS_64(info->dest_ip[3], I40IW_CQPSQ_QHASH_ADDR3));
 	}
-	qw2 = LS_64(cqp->dev->qos[info->user_pri].qs_handle, I40IW_CQPSQ_QHASH_QS_HANDLE);
+	qw2 = LS_64(vsi->qos[info->user_pri].qs_handle, I40IW_CQPSQ_QHASH_QS_HANDLE);
 	if (info->vlan_valid)
 		qw2 |= LS_64(info->vlan_id, I40IW_CQPSQ_QHASH_VLANID);
 	set_64bit_val(wqe, 16, qw2);
@@ -2225,6 +2236,7 @@ static enum i40iw_status_code i40iw_sc_qp_init(struct i40iw_sc_qp *qp,
 	u32 offset;
 
 	qp->dev = info->pd->dev;
+	qp->vsi = info->vsi;
 	qp->sq_pa = info->sq_pa;
 	qp->rq_pa = info->rq_pa;
 	qp->hw_host_ctx_pa = info->host_ctx_pa;
@@ -2273,7 +2285,7 @@ static enum i40iw_status_code i40iw_sc_qp_init(struct i40iw_sc_qp *qp,
 	qp->rq_tph_en = info->rq_tph_en;
 	qp->rcv_tph_en = info->rcv_tph_en;
 	qp->xmit_tph_en = info->xmit_tph_en;
-	qp->qs_handle = qp->pd->dev->qos[qp->user_pri].qs_handle;
+	qp->qs_handle = qp->vsi->qos[qp->user_pri].qs_handle;
 	qp->exception_lan_queue = qp->pd->dev->exception_lan_queue;
 
 	return 0;
@@ -2418,7 +2430,7 @@ static enum i40iw_status_code i40iw_sc_qp_destroy(
 	struct i40iw_sc_cqp *cqp;
 	u64 header;
 
-	i40iw_qp_rem_qos(qp->pd->dev, qp);
+	i40iw_qp_rem_qos(qp);
 	cqp = qp->pd->dev->cqp;
 	wqe = i40iw_sc_cqp_get_next_send_wqe(cqp, scratch);
 	if (!wqe)
@@ -2566,13 +2578,17 @@ static enum i40iw_status_code i40iw_sc_qp_setctx(
 {
 	struct i40iwarp_offload_info *iw;
 	struct i40iw_tcp_offload_info *tcp;
+	struct i40iw_sc_vsi *vsi;
+	struct i40iw_sc_dev *dev;
 	u64 qw0, qw3, qw7 = 0;
 
 	iw = info->iwarp_info;
 	tcp = info->tcp_info;
+	vsi = qp->vsi;
+	dev = qp->dev;
 	if (info->add_to_qoslist) {
 		qp->user_pri = info->user_pri;
-		i40iw_qp_add_qos(qp->pd->dev, qp);
+		i40iw_qp_add_qos(qp);
 		i40iw_debug(qp->dev, I40IW_DEBUG_DCB, "%s qp[%d] UP[%d] qset[%d]\n",
 			    __func__, qp->qp_uk.qp_id, qp->user_pri, qp->qs_handle);
 	}
@@ -2616,7 +2632,10 @@ static enum i40iw_status_code i40iw_sc_qp_setctx(
 		       LS_64(iw->rdmap_ver, I40IWQPC_RDMAP_VER);
 
 		qw7 |= LS_64(iw->pd_id, I40IWQPC_PDIDX);
-		set_64bit_val(qp_ctx, 144, qp->q2_pa);
+		set_64bit_val(qp_ctx,
+			      144,
+			      LS_64(qp->q2_pa, I40IWQPC_Q2ADDR) |
+			      LS_64(vsi->fcn_id, I40IWQPC_STAT_INDEX));
 		set_64bit_val(qp_ctx,
 			      152,
 			      LS_64(iw->last_byte_sent, I40IWQPC_LASTBYTESENT));
@@ -2631,6 +2650,9 @@ static enum i40iw_status_code i40iw_sc_qp_setctx(
 			      LS_64(iw->bind_en, I40IWQPC_BINDEN) |
 			      LS_64(iw->fast_reg_en, I40IWQPC_FASTREGEN) |
 			      LS_64(iw->priv_mode_en, I40IWQPC_PRIVEN) |
+			      LS_64((((vsi->stats_fcn_id_alloc) &&
+				      (dev->is_pf) && (vsi->fcn_id >= I40IW_FIRST_NON_PF_STAT)) ? 1 : 0),
+				    I40IWQPC_USESTATSINSTANCE) |
 			      LS_64(1, I40IWQPC_IWARPMODE) |
 			      LS_64(iw->rcv_mark_en, I40IWQPC_RCVMARKERS) |
 			      LS_64(iw->align_hdrs, I40IWQPC_ALIGNHDRS) |
@@ -4447,286 +4469,370 @@ void i40iw_terminate_received(struct i40iw_sc_qp *qp, struct i40iw_aeqe_info *in
 }
 
 /**
- * i40iw_hw_stat_init - Initiliaze HW stats table
- * @devstat: pestat struct
+ * i40iw_sc_vsi_init - Initialize virtual device
+ * @vsi: pointer to the vsi structure
+ * @info: parameters to initialize vsi
+ **/
+void i40iw_sc_vsi_init(struct i40iw_sc_vsi *vsi, struct i40iw_vsi_init_info *info)
+{
+	int i;
+
+	vsi->dev = info->dev;
+	vsi->back_vsi = info->back_vsi;
+	vsi->mss = info->params->mss;
+	i40iw_fill_qos_list(info->params->qs_handle_list);
+
+	for (i = 0; i < I40IW_MAX_USER_PRIORITY; i++) {
+		vsi->qos[i].qs_handle =
+			info->params->qs_handle_list[i];
+			i40iw_debug(vsi->dev, I40IW_DEBUG_DCB, "qset[%d]: %d\n", i, vsi->qos[i].qs_handle);
+		spin_lock_init(&vsi->qos[i].lock);
+		INIT_LIST_HEAD(&vsi->qos[i].qplist);
+	}
+}
+
+/**
+ * i40iw_hw_stats_init - Initiliaze HW stats table
+ * @stats: pestat struct
  * @fcn_idx: PCI fn id
- * @hw: PF i40iw_hw structure.
  * @is_pf: Is it a PF?
  *
- * Populate the HW stat table with register offset addr for each
- * stat. And start the perioidic stats timer.
+ * Populate the HW stats table with register offset addr for each
+ * stats. And start the perioidic stats timer.
  */
-static void i40iw_hw_stat_init(struct i40iw_dev_pestat *devstat,
-			       u8 fcn_idx,
-			       struct i40iw_hw *hw, bool is_pf)
+void i40iw_hw_stats_init(struct i40iw_vsi_pestat *stats, u8 fcn_idx, bool is_pf)
 {
-	u32 stat_reg_offset;
-	u32 stat_index;
-	struct i40iw_dev_hw_stat_offsets *stat_table =
-		&devstat->hw_stat_offsets;
-	struct i40iw_dev_hw_stats *last_rd_stats = &devstat->last_read_hw_stats;
-
-	devstat->hw = hw;
+	u32 stats_reg_offset;
+	u32 stats_index;
+	struct i40iw_dev_hw_stats_offsets *stats_table =
+		&stats->hw_stats_offsets;
+	struct i40iw_dev_hw_stats *last_rd_stats = &stats->last_read_hw_stats;
 
 	if (is_pf) {
-		stat_table->stat_offset_32[I40IW_HW_STAT_INDEX_IP4RXDISCARD] =
+		stats_table->stats_offset_32[I40IW_HW_STAT_INDEX_IP4RXDISCARD] =
 				I40E_GLPES_PFIP4RXDISCARD(fcn_idx);
-		stat_table->stat_offset_32[I40IW_HW_STAT_INDEX_IP4RXTRUNC] =
+		stats_table->stats_offset_32[I40IW_HW_STAT_INDEX_IP4RXTRUNC] =
 				I40E_GLPES_PFIP4RXTRUNC(fcn_idx);
-		stat_table->stat_offset_32[I40IW_HW_STAT_INDEX_IP4TXNOROUTE] =
+		stats_table->stats_offset_32[I40IW_HW_STAT_INDEX_IP4TXNOROUTE] =
 				I40E_GLPES_PFIP4TXNOROUTE(fcn_idx);
-		stat_table->stat_offset_32[I40IW_HW_STAT_INDEX_IP6RXDISCARD] =
+		stats_table->stats_offset_32[I40IW_HW_STAT_INDEX_IP6RXDISCARD] =
 				I40E_GLPES_PFIP6RXDISCARD(fcn_idx);
-		stat_table->stat_offset_32[I40IW_HW_STAT_INDEX_IP6RXTRUNC] =
+		stats_table->stats_offset_32[I40IW_HW_STAT_INDEX_IP6RXTRUNC] =
 				I40E_GLPES_PFIP6RXTRUNC(fcn_idx);
-		stat_table->stat_offset_32[I40IW_HW_STAT_INDEX_IP6TXNOROUTE] =
+		stats_table->stats_offset_32[I40IW_HW_STAT_INDEX_IP6TXNOROUTE] =
 				I40E_GLPES_PFIP6TXNOROUTE(fcn_idx);
-		stat_table->stat_offset_32[I40IW_HW_STAT_INDEX_TCPRTXSEG] =
+		stats_table->stats_offset_32[I40IW_HW_STAT_INDEX_TCPRTXSEG] =
 				I40E_GLPES_PFTCPRTXSEG(fcn_idx);
-		stat_table->stat_offset_32[I40IW_HW_STAT_INDEX_TCPRXOPTERR] =
+		stats_table->stats_offset_32[I40IW_HW_STAT_INDEX_TCPRXOPTERR] =
 				I40E_GLPES_PFTCPRXOPTERR(fcn_idx);
-		stat_table->stat_offset_32[I40IW_HW_STAT_INDEX_TCPRXPROTOERR] =
+		stats_table->stats_offset_32[I40IW_HW_STAT_INDEX_TCPRXPROTOERR] =
 				I40E_GLPES_PFTCPRXPROTOERR(fcn_idx);
 
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP4RXOCTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP4RXOCTS] =
 				I40E_GLPES_PFIP4RXOCTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP4RXPKTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP4RXPKTS] =
 				I40E_GLPES_PFIP4RXPKTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP4RXFRAGS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP4RXFRAGS] =
 				I40E_GLPES_PFIP4RXFRAGSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP4RXMCPKTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP4RXMCPKTS] =
 				I40E_GLPES_PFIP4RXMCPKTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP4TXOCTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP4TXOCTS] =
 				I40E_GLPES_PFIP4TXOCTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP4TXPKTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP4TXPKTS] =
 				I40E_GLPES_PFIP4TXPKTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP4TXFRAGS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP4TXFRAGS] =
 				I40E_GLPES_PFIP4TXFRAGSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP4TXMCPKTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP4TXMCPKTS] =
 				I40E_GLPES_PFIP4TXMCPKTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP6RXOCTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP6RXOCTS] =
 				I40E_GLPES_PFIP6RXOCTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP6RXPKTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP6RXPKTS] =
 				I40E_GLPES_PFIP6RXPKTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP6RXFRAGS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP6RXFRAGS] =
 				I40E_GLPES_PFIP6RXFRAGSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP6RXMCPKTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP6RXMCPKTS] =
 				I40E_GLPES_PFIP6RXMCPKTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP6TXOCTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP6TXOCTS] =
 				I40E_GLPES_PFIP6TXOCTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP6TXPKTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP6TXPKTS] =
 				I40E_GLPES_PFIP6TXPKTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP6TXPKTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP6TXPKTS] =
 				I40E_GLPES_PFIP6TXPKTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP6TXFRAGS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP6TXFRAGS] =
 				I40E_GLPES_PFIP6TXFRAGSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_TCPRXSEGS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_TCPRXSEGS] =
 				I40E_GLPES_PFTCPRXSEGSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_TCPTXSEG] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_TCPTXSEG] =
 				I40E_GLPES_PFTCPTXSEGLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_RDMARXRDS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_RDMARXRDS] =
 				I40E_GLPES_PFRDMARXRDSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_RDMARXSNDS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_RDMARXSNDS] =
 				I40E_GLPES_PFRDMARXSNDSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_RDMARXWRS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_RDMARXWRS] =
 				I40E_GLPES_PFRDMARXWRSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_RDMATXRDS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_RDMATXRDS] =
 				I40E_GLPES_PFRDMATXRDSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_RDMATXSNDS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_RDMATXSNDS] =
 				I40E_GLPES_PFRDMATXSNDSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_RDMATXWRS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_RDMATXWRS] =
 				I40E_GLPES_PFRDMATXWRSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_RDMAVBND] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_RDMAVBND] =
 				I40E_GLPES_PFRDMAVBNDLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_RDMAVINV] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_RDMAVINV] =
 				I40E_GLPES_PFRDMAVINVLO(fcn_idx);
 	} else {
-		stat_table->stat_offset_32[I40IW_HW_STAT_INDEX_IP4RXDISCARD] =
+		stats_table->stats_offset_32[I40IW_HW_STAT_INDEX_IP4RXDISCARD] =
 				I40E_GLPES_VFIP4RXDISCARD(fcn_idx);
-		stat_table->stat_offset_32[I40IW_HW_STAT_INDEX_IP4RXTRUNC] =
+		stats_table->stats_offset_32[I40IW_HW_STAT_INDEX_IP4RXTRUNC] =
 				I40E_GLPES_VFIP4RXTRUNC(fcn_idx);
-		stat_table->stat_offset_32[I40IW_HW_STAT_INDEX_IP4TXNOROUTE] =
+		stats_table->stats_offset_32[I40IW_HW_STAT_INDEX_IP4TXNOROUTE] =
 				I40E_GLPES_VFIP4TXNOROUTE(fcn_idx);
-		stat_table->stat_offset_32[I40IW_HW_STAT_INDEX_IP6RXDISCARD] =
+		stats_table->stats_offset_32[I40IW_HW_STAT_INDEX_IP6RXDISCARD] =
 				I40E_GLPES_VFIP6RXDISCARD(fcn_idx);
-		stat_table->stat_offset_32[I40IW_HW_STAT_INDEX_IP6RXTRUNC] =
+		stats_table->stats_offset_32[I40IW_HW_STAT_INDEX_IP6RXTRUNC] =
 				I40E_GLPES_VFIP6RXTRUNC(fcn_idx);
-		stat_table->stat_offset_32[I40IW_HW_STAT_INDEX_IP6TXNOROUTE] =
+		stats_table->stats_offset_32[I40IW_HW_STAT_INDEX_IP6TXNOROUTE] =
 				I40E_GLPES_VFIP6TXNOROUTE(fcn_idx);
-		stat_table->stat_offset_32[I40IW_HW_STAT_INDEX_TCPRTXSEG] =
+		stats_table->stats_offset_32[I40IW_HW_STAT_INDEX_TCPRTXSEG] =
 				I40E_GLPES_VFTCPRTXSEG(fcn_idx);
-		stat_table->stat_offset_32[I40IW_HW_STAT_INDEX_TCPRXOPTERR] =
+		stats_table->stats_offset_32[I40IW_HW_STAT_INDEX_TCPRXOPTERR] =
 				I40E_GLPES_VFTCPRXOPTERR(fcn_idx);
-		stat_table->stat_offset_32[I40IW_HW_STAT_INDEX_TCPRXPROTOERR] =
+		stats_table->stats_offset_32[I40IW_HW_STAT_INDEX_TCPRXPROTOERR] =
 				I40E_GLPES_VFTCPRXPROTOERR(fcn_idx);
 
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP4RXOCTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP4RXOCTS] =
 				I40E_GLPES_VFIP4RXOCTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP4RXPKTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP4RXPKTS] =
 				I40E_GLPES_VFIP4RXPKTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP4RXFRAGS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP4RXFRAGS] =
 				I40E_GLPES_VFIP4RXFRAGSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP4RXMCPKTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP4RXMCPKTS] =
 				I40E_GLPES_VFIP4RXMCPKTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP4TXOCTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP4TXOCTS] =
 				I40E_GLPES_VFIP4TXOCTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP4TXPKTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP4TXPKTS] =
 				I40E_GLPES_VFIP4TXPKTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP4TXFRAGS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP4TXFRAGS] =
 				I40E_GLPES_VFIP4TXFRAGSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP4TXMCPKTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP4TXMCPKTS] =
 				I40E_GLPES_VFIP4TXMCPKTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP6RXOCTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP6RXOCTS] =
 				I40E_GLPES_VFIP6RXOCTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP6RXPKTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP6RXPKTS] =
 				I40E_GLPES_VFIP6RXPKTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP6RXFRAGS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP6RXFRAGS] =
 				I40E_GLPES_VFIP6RXFRAGSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP6RXMCPKTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP6RXMCPKTS] =
 				I40E_GLPES_VFIP6RXMCPKTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP6TXOCTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP6TXOCTS] =
 				I40E_GLPES_VFIP6TXOCTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP6TXPKTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP6TXPKTS] =
 				I40E_GLPES_VFIP6TXPKTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP6TXPKTS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP6TXPKTS] =
 				I40E_GLPES_VFIP6TXPKTSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_IP6TXFRAGS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_IP6TXFRAGS] =
 				I40E_GLPES_VFIP6TXFRAGSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_TCPRXSEGS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_TCPRXSEGS] =
 				I40E_GLPES_VFTCPRXSEGSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_TCPTXSEG] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_TCPTXSEG] =
 				I40E_GLPES_VFTCPTXSEGLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_RDMARXRDS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_RDMARXRDS] =
 				I40E_GLPES_VFRDMARXRDSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_RDMARXSNDS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_RDMARXSNDS] =
 				I40E_GLPES_VFRDMARXSNDSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_RDMARXWRS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_RDMARXWRS] =
 				I40E_GLPES_VFRDMARXWRSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_RDMATXRDS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_RDMATXRDS] =
 				I40E_GLPES_VFRDMATXRDSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_RDMATXSNDS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_RDMATXSNDS] =
 				I40E_GLPES_VFRDMATXSNDSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_RDMATXWRS] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_RDMATXWRS] =
 				I40E_GLPES_VFRDMATXWRSLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_RDMAVBND] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_RDMAVBND] =
 				I40E_GLPES_VFRDMAVBNDLO(fcn_idx);
-		stat_table->stat_offset_64[I40IW_HW_STAT_INDEX_RDMAVINV] =
+		stats_table->stats_offset_64[I40IW_HW_STAT_INDEX_RDMAVINV] =
 				I40E_GLPES_VFRDMAVINVLO(fcn_idx);
 	}
 
-	for (stat_index = 0; stat_index < I40IW_HW_STAT_INDEX_MAX_64;
-	     stat_index++) {
-		stat_reg_offset = stat_table->stat_offset_64[stat_index];
-		last_rd_stats->stat_value_64[stat_index] =
-			readq(devstat->hw->hw_addr + stat_reg_offset);
+	for (stats_index = 0; stats_index < I40IW_HW_STAT_INDEX_MAX_64;
+	     stats_index++) {
+		stats_reg_offset = stats_table->stats_offset_64[stats_index];
+		last_rd_stats->stats_value_64[stats_index] =
+			readq(stats->hw->hw_addr + stats_reg_offset);
 	}
 
-	for (stat_index = 0; stat_index < I40IW_HW_STAT_INDEX_MAX_32;
-	     stat_index++) {
-		stat_reg_offset = stat_table->stat_offset_32[stat_index];
-		last_rd_stats->stat_value_32[stat_index] =
-			i40iw_rd32(devstat->hw, stat_reg_offset);
+	for (stats_index = 0; stats_index < I40IW_HW_STAT_INDEX_MAX_32;
+	     stats_index++) {
+		stats_reg_offset = stats_table->stats_offset_32[stats_index];
+		last_rd_stats->stats_value_32[stats_index] =
+			i40iw_rd32(stats->hw, stats_reg_offset);
 	}
 }
 
 /**
- * i40iw_hw_stat_read_32 - Read 32-bit HW stat counters and accommodates for roll-overs.
- * @devstat: pestat struct
- * @index: index in HW stat table which contains offset reg-addr
- * @value: hw stat value
+ * i40iw_hw_stats_read_32 - Read 32-bit HW stats counters and accommodates for roll-overs.
+ * @stat: pestat struct
+ * @index: index in HW stats table which contains offset reg-addr
+ * @value: hw stats value
  */
-static void i40iw_hw_stat_read_32(struct i40iw_dev_pestat *devstat,
-				  enum i40iw_hw_stat_index_32b index,
-				  u64 *value)
+void i40iw_hw_stats_read_32(struct i40iw_vsi_pestat *stats,
+			    enum i40iw_hw_stats_index_32b index,
+			    u64 *value)
 {
-	struct i40iw_dev_hw_stat_offsets *stat_table =
-		&devstat->hw_stat_offsets;
-	struct i40iw_dev_hw_stats *last_rd_stats = &devstat->last_read_hw_stats;
-	struct i40iw_dev_hw_stats *hw_stats = &devstat->hw_stats;
-	u64 new_stat_value = 0;
-	u32 stat_reg_offset = stat_table->stat_offset_32[index];
-
-	new_stat_value = i40iw_rd32(devstat->hw, stat_reg_offset);
+	struct i40iw_dev_hw_stats_offsets *stats_table =
+		&stats->hw_stats_offsets;
+	struct i40iw_dev_hw_stats *last_rd_stats = &stats->last_read_hw_stats;
+	struct i40iw_dev_hw_stats *hw_stats = &stats->hw_stats;
+	u64 new_stats_value = 0;
+	u32 stats_reg_offset = stats_table->stats_offset_32[index];
+
+	new_stats_value = i40iw_rd32(stats->hw, stats_reg_offset);
 	/*roll-over case */
-	if (new_stat_value < last_rd_stats->stat_value_32[index])
-		hw_stats->stat_value_32[index] += new_stat_value;
+	if (new_stats_value < last_rd_stats->stats_value_32[index])
+		hw_stats->stats_value_32[index] += new_stats_value;
 	else
-		hw_stats->stat_value_32[index] +=
-			new_stat_value - last_rd_stats->stat_value_32[index];
-	last_rd_stats->stat_value_32[index] = new_stat_value;
-	*value = hw_stats->stat_value_32[index];
+		hw_stats->stats_value_32[index] +=
+			new_stats_value - last_rd_stats->stats_value_32[index];
+	last_rd_stats->stats_value_32[index] = new_stats_value;
+	*value = hw_stats->stats_value_32[index];
 }
 
 /**
- * i40iw_hw_stat_read_64 - Read HW stat counters (greater than 32-bit) and accommodates for roll-overs.
- * @devstat: pestat struct
- * @index: index in HW stat table which contains offset reg-addr
- * @value: hw stat value
+ * i40iw_hw_stats_read_64 - Read HW stats counters (greater than 32-bit) and accommodates for roll-overs.
+ * @stats: pestat struct
+ * @index: index in HW stats table which contains offset reg-addr
+ * @value: hw stats value
  */
-static void i40iw_hw_stat_read_64(struct i40iw_dev_pestat *devstat,
-				  enum i40iw_hw_stat_index_64b index,
-				  u64 *value)
+void i40iw_hw_stats_read_64(struct i40iw_vsi_pestat *stats,
+			    enum i40iw_hw_stats_index_64b index,
+			    u64 *value)
 {
-	struct i40iw_dev_hw_stat_offsets *stat_table =
-		&devstat->hw_stat_offsets;
-	struct i40iw_dev_hw_stats *last_rd_stats = &devstat->last_read_hw_stats;
-	struct i40iw_dev_hw_stats *hw_stats = &devstat->hw_stats;
-	u64 new_stat_value = 0;
-	u32 stat_reg_offset = stat_table->stat_offset_64[index];
-
-	new_stat_value = readq(devstat->hw->hw_addr + stat_reg_offset);
+	struct i40iw_dev_hw_stats_offsets *stats_table =
+		&stats->hw_stats_offsets;
+	struct i40iw_dev_hw_stats *last_rd_stats = &stats->last_read_hw_stats;
+	struct i40iw_dev_hw_stats *hw_stats = &stats->hw_stats;
+	u64 new_stats_value = 0;
+	u32 stats_reg_offset = stats_table->stats_offset_64[index];
+
+	new_stats_value = readq(stats->hw->hw_addr + stats_reg_offset);
 	/*roll-over case */
-	if (new_stat_value < last_rd_stats->stat_value_64[index])
-		hw_stats->stat_value_64[index] += new_stat_value;
+	if (new_stats_value < last_rd_stats->stats_value_64[index])
+		hw_stats->stats_value_64[index] += new_stats_value;
 	else
-		hw_stats->stat_value_64[index] +=
-			new_stat_value - last_rd_stats->stat_value_64[index];
-	last_rd_stats->stat_value_64[index] = new_stat_value;
-	*value = hw_stats->stat_value_64[index];
+		hw_stats->stats_value_64[index] +=
+			new_stats_value - last_rd_stats->stats_value_64[index];
+	last_rd_stats->stats_value_64[index] = new_stats_value;
+	*value = hw_stats->stats_value_64[index];
 }
 
 /**
- * i40iw_hw_stat_read_all - read all HW stat counters
- * @devstat: pestat struct
- * @stat_values: hw stats structure
+ * i40iw_hw_stats_read_all - read all HW stat counters
+ * @stats: pestat struct
+ * @stats_values: hw stats structure
  *
  * Read all the HW stat counters and populates hw_stats structure
- * of passed-in dev's pestat as well as copy created in stat_values.
+ * of passed-in vsi's pestat as well as copy created in stat_values.
  */
-static void i40iw_hw_stat_read_all(struct i40iw_dev_pestat *devstat,
-				   struct i40iw_dev_hw_stats *stat_values)
+void i40iw_hw_stats_read_all(struct i40iw_vsi_pestat *stats,
+			     struct i40iw_dev_hw_stats *stats_values)
 {
-	u32 stat_index;
-
-	for (stat_index = 0; stat_index < I40IW_HW_STAT_INDEX_MAX_32;
-	     stat_index++)
-		i40iw_hw_stat_read_32(devstat, stat_index,
-				      &stat_values->stat_value_32[stat_index]);
-	for (stat_index = 0; stat_index < I40IW_HW_STAT_INDEX_MAX_64;
-	     stat_index++)
-		i40iw_hw_stat_read_64(devstat, stat_index,
-				      &stat_values->stat_value_64[stat_index]);
+	u32 stats_index;
+	unsigned long flags;
+
+	spin_lock_irqsave(&stats->lock, flags);
+
+	for (stats_index = 0; stats_index < I40IW_HW_STAT_INDEX_MAX_32;
+	     stats_index++)
+		i40iw_hw_stats_read_32(stats, stats_index,
+				       &stats_values->stats_value_32[stats_index]);
+	for (stats_index = 0; stats_index < I40IW_HW_STAT_INDEX_MAX_64;
+	     stats_index++)
+		i40iw_hw_stats_read_64(stats, stats_index,
+				       &stats_values->stats_value_64[stats_index]);
+	spin_unlock_irqrestore(&stats->lock, flags);
 }
 
 /**
- * i40iw_hw_stat_refresh_all - Update all HW stat structs
- * @devstat: pestat struct
- * @stat_values: hw stats structure
+ * i40iw_hw_stats_refresh_all - Update all HW stats structs
+ * @stats: pestat struct
  *
- * Read all the HW stat counters to refresh values in hw_stats structure
+ * Read all the HW stats counters to refresh values in hw_stats structure
  * of passed-in dev's pestat
  */
-static void i40iw_hw_stat_refresh_all(struct i40iw_dev_pestat *devstat)
+void i40iw_hw_stats_refresh_all(struct i40iw_vsi_pestat *stats)
+{
+	u64 stats_value;
+	u32 stats_index;
+	unsigned long flags;
+
+	spin_lock_irqsave(&stats->lock, flags);
+
+	for (stats_index = 0; stats_index < I40IW_HW_STAT_INDEX_MAX_32;
+	     stats_index++)
+		i40iw_hw_stats_read_32(stats, stats_index, &stats_value);
+	for (stats_index = 0; stats_index < I40IW_HW_STAT_INDEX_MAX_64;
+	     stats_index++)
+		i40iw_hw_stats_read_64(stats, stats_index, &stats_value);
+	spin_unlock_irqrestore(&stats->lock, flags);
+}
+
+/**
+ * i40iw_get_fcn_id - Return the function id
+ * @dev: pointer to the device
+ */
+static u8 i40iw_get_fcn_id(struct i40iw_sc_dev *dev)
+{
+	u8 fcn_id = I40IW_INVALID_FCN_ID;
+	u8 i;
+
+	for (i = I40IW_FIRST_NON_PF_STAT; i < I40IW_MAX_STATS_COUNT; i++)
+		if (!dev->fcn_id_array[i]) {
+			fcn_id = i;
+			dev->fcn_id_array[i] = true;
+			break;
+		}
+	return fcn_id;
+}
+
+/**
+ * i40iw_vsi_stats_init - Initialize the vsi statistics
+ * @vsi: pointer to the vsi structure
+ * @info: The info structure used for initialization
+ */
+enum i40iw_status_code i40iw_vsi_stats_init(struct i40iw_sc_vsi *vsi, struct i40iw_vsi_stats_info *info)
 {
-	u64 stat_value;
-	u32 stat_index;
-
-	for (stat_index = 0; stat_index < I40IW_HW_STAT_INDEX_MAX_32;
-	     stat_index++)
-		i40iw_hw_stat_read_32(devstat, stat_index, &stat_value);
-	for (stat_index = 0; stat_index < I40IW_HW_STAT_INDEX_MAX_64;
-	     stat_index++)
-		i40iw_hw_stat_read_64(devstat, stat_index, &stat_value);
+	u8 fcn_id = info->fcn_id;
+
+	if (info->alloc_fcn_id)
+		fcn_id = i40iw_get_fcn_id(vsi->dev);
+
+	if (fcn_id == I40IW_INVALID_FCN_ID)
+		return I40IW_ERR_NOT_READY;
+
+	vsi->pestat = info->pestat;
+	vsi->pestat->hw = vsi->dev->hw;
+
+	if (info->stats_initialize) {
+		i40iw_hw_stats_init(vsi->pestat, fcn_id, true);
+		spin_lock_init(&vsi->pestat->lock);
+		i40iw_hw_stats_start_timer(vsi);
+	}
+	vsi->stats_fcn_id_alloc = info->alloc_fcn_id;
+	vsi->fcn_id = fcn_id;
+	return I40IW_SUCCESS;
+}
+
+/**
+ * i40iw_vsi_stats_free - Free the vsi stats
+ * @vsi: pointer to the vsi structure
+ */
+void i40iw_vsi_stats_free(struct i40iw_sc_vsi *vsi)
+{
+	u8 fcn_id = vsi->fcn_id;
+
+	if ((vsi->stats_fcn_id_alloc) && (fcn_id != I40IW_INVALID_FCN_ID))
+		vsi->dev->fcn_id_array[fcn_id] = false;
+	i40iw_hw_stats_stop_timer(vsi);
 }
 
 static struct i40iw_cqp_ops iw_cqp_ops = {
@@ -4837,23 +4943,6 @@ static struct i40iw_hmc_ops iw_hmc_ops = {
 	NULL
 };
 
-static const struct i40iw_device_pestat_ops iw_device_pestat_ops = {
-	i40iw_hw_stat_init,
-	i40iw_hw_stat_read_32,
-	i40iw_hw_stat_read_64,
-	i40iw_hw_stat_read_all,
-	i40iw_hw_stat_refresh_all
-};
-
-/**
- * i40iw_device_init_pestat - Initialize the pestat structure
- * @dev: pestat struct
- */
-void i40iw_device_init_pestat(struct i40iw_dev_pestat *devstat)
-{
-	devstat->ops = iw_device_pestat_ops;
-}
-
 /**
  * i40iw_device_init - Initialize IWARP device
  * @dev: IWARP device pointer
@@ -4867,7 +4956,6 @@ enum i40iw_status_code i40iw_device_init(struct i40iw_sc_dev *dev,
 	u16 hmc_fcn = 0;
 	enum i40iw_status_code ret_code = 0;
 	u8 db_size;
-	int i;
 
 	spin_lock_init(&dev->cqp_lock);
 	INIT_LIST_HEAD(&dev->cqp_cmd_head);             /* for the cqp commands backlog. */
@@ -4876,15 +4964,7 @@ enum i40iw_status_code i40iw_device_init(struct i40iw_sc_dev *dev,
 
 	dev->debug_mask = info->debug_mask;
 
-	i40iw_device_init_pestat(&dev->dev_pestat);
 	dev->hmc_fn_id = info->hmc_fn_id;
-	i40iw_fill_qos_list(info->l2params.qs_handle_list);
-	for (i = 0; i < I40IW_MAX_USER_PRIORITY; i++) {
-		dev->qos[i].qs_handle = info->l2params.qs_handle_list[i];
-		i40iw_debug(dev, I40IW_DEBUG_DCB, "qset[%d]: %d\n", i, dev->qos[i].qs_handle);
-		spin_lock_init(&dev->qos[i].lock);
-		INIT_LIST_HEAD(&dev->qos[i].qplist);
-	}
 	dev->exception_lan_queue = info->exception_lan_queue;
 	dev->is_pf = info->is_pf;
 
@@ -4897,15 +4977,10 @@ enum i40iw_status_code i40iw_device_init(struct i40iw_sc_dev *dev,
 	dev->hw = info->hw;
 	dev->hw->hw_addr = info->bar0;
 
-	val = i40iw_rd32(dev->hw, I40E_GLPCI_DREVID);
-	dev->hw_rev = (u8)RS_32(val, I40E_GLPCI_DREVID_DEFAULT_REVID);
-
 	if (dev->is_pf) {
-		dev->dev_pestat.ops.iw_hw_stat_init(&dev->dev_pestat,
-			dev->hmc_fn_id, dev->hw, true);
-		spin_lock_init(&dev->dev_pestat.stats_lock);
-		/*start the periodic stats_timer */
-		i40iw_hw_stats_start_timer(dev);
+		val = i40iw_rd32(dev->hw, I40E_GLPCI_DREVID);
+		dev->hw_rev = (u8)RS_32(val, I40E_GLPCI_DREVID_DEFAULT_REVID);
+
 		val = i40iw_rd32(dev->hw, I40E_GLPCI_LBARCTRL);
 		db_size = (u8)RS_32(val, I40E_GLPCI_LBARCTRL_PE_DB_SIZE);
 		if ((db_size != I40IW_PE_DB_SIZE_4M) &&
diff --git a/drivers/infiniband/hw/i40iw/i40iw_d.h b/drivers/infiniband/hw/i40iw/i40iw_d.h
index 1bd4badb26d3..a39ac12b6a7e 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_d.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_d.h
@@ -69,6 +69,9 @@
 #define I40IW_STAG_TYPE_NONSHARED 1
 
 #define I40IW_MAX_USER_PRIORITY 8
+#define I40IW_MAX_STATS_COUNT 16
+#define I40IW_FIRST_NON_PF_STAT	4
+
 
 #define LS_64_1(val, bits)      ((u64)(uintptr_t)val << bits)
 #define RS_64_1(val, bits)      ((u64)(uintptr_t)val >> bits)
@@ -1203,8 +1206,11 @@
 #define I40IWQPC_RXCQNUM_SHIFT 32
 #define I40IWQPC_RXCQNUM_MASK (0x1ffffULL << I40IWQPC_RXCQNUM_SHIFT)
 
-#define I40IWQPC_Q2ADDR_SHIFT I40IW_CQPHC_QPCTX_SHIFT
-#define I40IWQPC_Q2ADDR_MASK I40IW_CQPHC_QPCTX_MASK
+#define I40IWQPC_STAT_INDEX_SHIFT 0
+#define I40IWQPC_STAT_INDEX_MASK (0x1fULL << I40IWQPC_STAT_INDEX_SHIFT)
+
+#define I40IWQPC_Q2ADDR_SHIFT 0
+#define I40IWQPC_Q2ADDR_MASK (0xffffffffffffff00ULL << I40IWQPC_Q2ADDR_SHIFT)
 
 #define I40IWQPC_LASTBYTESENT_SHIFT 0
 #define I40IWQPC_LASTBYTESENT_MASK (0xffUL << I40IWQPC_LASTBYTESENT_SHIFT)
@@ -1236,11 +1242,8 @@
 #define I40IWQPC_PRIVEN_SHIFT 25
 #define I40IWQPC_PRIVEN_MASK (1UL << I40IWQPC_PRIVEN_SHIFT)
 
-#define I40IWQPC_LSMMPRESENT_SHIFT 26
-#define I40IWQPC_LSMMPRESENT_MASK (1UL << I40IWQPC_LSMMPRESENT_SHIFT)
-
-#define I40IWQPC_ADJUSTFORLSMM_SHIFT 27
-#define I40IWQPC_ADJUSTFORLSMM_MASK (1UL << I40IWQPC_ADJUSTFORLSMM_SHIFT)
+#define I40IWQPC_USESTATSINSTANCE_SHIFT 26
+#define I40IWQPC_USESTATSINSTANCE_MASK (1UL << I40IWQPC_USESTATSINSTANCE_SHIFT)
 
 #define I40IWQPC_IWARPMODE_SHIFT 28
 #define I40IWQPC_IWARPMODE_MASK (1UL << I40IWQPC_IWARPMODE_SHIFT)
@@ -1717,6 +1720,8 @@ enum i40iw_alignment {
 #define OP_MANAGE_VF_PBLE_BP                    28
 #define OP_QUERY_FPM_VALUES                     29
 #define OP_COMMIT_FPM_VALUES                    30
-#define OP_SIZE_CQP_STAT_ARRAY                  31
+#define OP_REQUESTED_COMMANDS                   31
+#define OP_COMPLETED_COMMANDS                   32
+#define OP_SIZE_CQP_STAT_ARRAY                  33
 
 #endif
diff --git a/drivers/infiniband/hw/i40iw/i40iw_hw.c b/drivers/infiniband/hw/i40iw/i40iw_hw.c
index 4394a6713bdf..476867a3f584 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_hw.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_hw.c
@@ -542,6 +542,7 @@ enum i40iw_status_code i40iw_manage_qhash(struct i40iw_device *iwdev,
 {
 	struct i40iw_qhash_table_info *info;
 	struct i40iw_sc_dev *dev = &iwdev->sc_dev;
+	struct i40iw_sc_vsi *vsi = &iwdev->vsi;
 	enum i40iw_status_code status;
 	struct i40iw_cqp *iwcqp = &iwdev->cqp;
 	struct i40iw_cqp_request *cqp_request;
@@ -554,6 +555,7 @@ enum i40iw_status_code i40iw_manage_qhash(struct i40iw_device *iwdev,
 	info = &cqp_info->in.u.manage_qhash_table_entry.info;
 	memset(info, 0, sizeof(*info));
 
+	info->vsi = &iwdev->vsi;
 	info->manage = mtype;
 	info->entry_type = etype;
 	if (cminfo->vlan_id != 0xFFFF) {
@@ -566,7 +568,7 @@ enum i40iw_status_code i40iw_manage_qhash(struct i40iw_device *iwdev,
 	info->ipv4_valid = cminfo->ipv4;
 	info->user_pri = cminfo->user_pri;
 	ether_addr_copy(info->mac_addr, iwdev->netdev->dev_addr);
-	info->qp_num = cpu_to_le32(dev->ilq->qp_id);
+	info->qp_num = cpu_to_le32(vsi->ilq->qp_id);
 	info->dest_port = cpu_to_le16(cminfo->loc_port);
 	info->dest_ip[0] = cpu_to_le32(cminfo->loc_addr[0]);
 	info->dest_ip[1] = cpu_to_le32(cminfo->loc_addr[1]);
diff --git a/drivers/infiniband/hw/i40iw/i40iw_main.c b/drivers/infiniband/hw/i40iw/i40iw_main.c
index cf9d288a79b3..2bdb8b0c8c88 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_main.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_main.c
@@ -932,6 +932,7 @@ static enum i40iw_status_code i40iw_initialize_ilq(struct i40iw_device *iwdev)
 	struct i40iw_puda_rsrc_info info;
 	enum i40iw_status_code status;
 
+	memset(&info, 0, sizeof(info));
 	info.type = I40IW_PUDA_RSRC_TYPE_ILQ;
 	info.cq_id = 1;
 	info.qp_id = 0;
@@ -941,10 +942,9 @@ static enum i40iw_status_code i40iw_initialize_ilq(struct i40iw_device *iwdev)
 	info.rq_size = 8192;
 	info.buf_size = 1024;
 	info.tx_buf_cnt = 16384;
-	info.mss = iwdev->sc_dev.mss;
 	info.receive = i40iw_receive_ilq;
 	info.xmit_complete = i40iw_free_sqbuf;
-	status = i40iw_puda_create_rsrc(&iwdev->sc_dev, &info);
+	status = i40iw_puda_create_rsrc(&iwdev->vsi, &info);
 	if (status)
 		i40iw_pr_err("ilq create fail\n");
 	return status;
@@ -961,6 +961,7 @@ static enum i40iw_status_code i40iw_initialize_ieq(struct i40iw_device *iwdev)
 	struct i40iw_puda_rsrc_info info;
 	enum i40iw_status_code status;
 
+	memset(&info, 0, sizeof(info));
 	info.type = I40IW_PUDA_RSRC_TYPE_IEQ;
 	info.cq_id = 2;
 	info.qp_id = iwdev->sc_dev.exception_lan_queue;
@@ -969,9 +970,8 @@ static enum i40iw_status_code i40iw_initialize_ieq(struct i40iw_device *iwdev)
 	info.sq_size = 8192;
 	info.rq_size = 8192;
 	info.buf_size = 2048;
-	info.mss = iwdev->sc_dev.mss;
 	info.tx_buf_cnt = 16384;
-	status = i40iw_puda_create_rsrc(&iwdev->sc_dev, &info);
+	status = i40iw_puda_create_rsrc(&iwdev->vsi, &info);
 	if (status)
 		i40iw_pr_err("ieq create fail\n");
 	return status;
@@ -1296,12 +1296,16 @@ static enum i40iw_status_code i40iw_initialize_dev(struct i40iw_device *iwdev,
 	enum i40iw_status_code status;
 	struct i40iw_sc_dev *dev = &iwdev->sc_dev;
 	struct i40iw_device_init_info info;
+	struct i40iw_vsi_init_info vsi_info;
 	struct i40iw_dma_mem mem;
+	struct i40iw_l2params l2params;
 	u32 size;
+	struct i40iw_vsi_stats_info stats_info;
 	u16 last_qset = I40IW_NO_QSET;
 	u16 qset;
 	u32 i;
 
+	memset(&l2params, 0, sizeof(l2params));
 	memset(&info, 0, sizeof(info));
 	size = sizeof(struct i40iw_hmc_pble_rsrc) + sizeof(struct i40iw_hmc_info) +
 				(sizeof(struct i40iw_hmc_obj_info) * I40IW_HMC_IW_MAX);
@@ -1330,16 +1334,17 @@ static enum i40iw_status_code i40iw_initialize_dev(struct i40iw_device *iwdev,
 	info.bar0 = ldev->hw_addr;
 	info.hw = &iwdev->hw;
 	info.debug_mask = debug;
-	info.l2params.mss =
+	l2params.mss =
 		(ldev->params.mtu) ? ldev->params.mtu - I40IW_MTU_TO_MSS : I40IW_DEFAULT_MSS;
 	for (i = 0; i < I40E_CLIENT_MAX_USER_PRIORITY; i++) {
 		qset = ldev->params.qos.prio_qos[i].qs_handle;
-		info.l2params.qs_handle_list[i] = qset;
+		l2params.qs_handle_list[i] = qset;
 		if (last_qset == I40IW_NO_QSET)
 			last_qset = qset;
 		else if ((qset != last_qset) && (qset != I40IW_NO_QSET))
 			iwdev->dcb = true;
 	}
+	i40iw_pr_info("DCB is set/clear = %d\n", iwdev->dcb);
 	info.exception_lan_queue = 1;
 	info.vchnl_send = i40iw_virtchnl_send;
 	status = i40iw_device_init(&iwdev->sc_dev, &info);
@@ -1348,6 +1353,20 @@ exit:
 		kfree(iwdev->hmc_info_mem);
 		iwdev->hmc_info_mem = NULL;
 	}
+	memset(&vsi_info, 0, sizeof(vsi_info));
+	vsi_info.dev = &iwdev->sc_dev;
+	vsi_info.back_vsi = (void *)iwdev;
+	vsi_info.params = &l2params;
+	i40iw_sc_vsi_init(&iwdev->vsi, &vsi_info);
+
+	if (dev->is_pf) {
+		memset(&stats_info, 0, sizeof(stats_info));
+		stats_info.fcn_id = ldev->fid;
+		stats_info.pestat = kzalloc(sizeof(*stats_info.pestat), GFP_KERNEL);
+		stats_info.stats_initialize = true;
+		if (stats_info.pestat)
+			i40iw_vsi_stats_init(&iwdev->vsi, &stats_info);
+	}
 	return status;
 }
 
@@ -1457,10 +1476,10 @@ static void i40iw_deinit_device(struct i40iw_device *iwdev, bool reset)
 		i40iw_destroy_aeq(iwdev, reset);
 		/* fallthrough */
 	case IEQ_CREATED:
-		i40iw_puda_dele_resources(dev, I40IW_PUDA_RSRC_TYPE_IEQ, reset);
+		i40iw_puda_dele_resources(&iwdev->vsi, I40IW_PUDA_RSRC_TYPE_IEQ, reset);
 		/* fallthrough */
 	case ILQ_CREATED:
-		i40iw_puda_dele_resources(dev, I40IW_PUDA_RSRC_TYPE_ILQ, reset);
+		i40iw_puda_dele_resources(&iwdev->vsi, I40IW_PUDA_RSRC_TYPE_ILQ, reset);
 		/* fallthrough */
 	case CCQ_CREATED:
 		i40iw_destroy_ccq(iwdev, reset);
@@ -1476,9 +1495,10 @@ static void i40iw_deinit_device(struct i40iw_device *iwdev, bool reset)
 		/* fallthrough */
 	case INITIAL_STATE:
 		i40iw_cleanup_cm_core(&iwdev->cm_core);
-		if (dev->is_pf)
-			i40iw_hw_stats_del_timer(dev);
-
+		if (iwdev->vsi.pestat) {
+			i40iw_vsi_stats_free(&iwdev->vsi);
+			kfree(iwdev->vsi.pestat);
+		}
 		i40iw_del_init_mem(iwdev);
 		break;
 	case INVALID_STATE:
@@ -1523,7 +1543,6 @@ static enum i40iw_status_code i40iw_setup_init_state(struct i40iw_handler *hdl,
 	iwdev->max_enabled_vfs = iwdev->max_rdma_vfs;
 	iwdev->netdev = ldev->netdev;
 	hdl->client = client;
-	iwdev->mss = (!ldev->params.mtu) ? I40IW_DEFAULT_MSS : ldev->params.mtu - I40IW_MTU_TO_MSS;
 	if (!ldev->ftype)
 		iwdev->db_start = pci_resource_start(ldev->pcidev, 0) + I40IW_DB_ADDR_OFFSET;
 	else
@@ -1683,7 +1702,7 @@ static void i40iw_l2params_worker(struct work_struct *work)
 	    container_of(work, struct l2params_work, work);
 	struct i40iw_device *iwdev = dwork->iwdev;
 
-	i40iw_change_l2params(&iwdev->sc_dev, &dwork->l2params);
+	i40iw_change_l2params(&iwdev->vsi, &dwork->l2params);
 	atomic_dec(&iwdev->params_busy);
 	kfree(work);
 }
@@ -1724,7 +1743,7 @@ static void i40iw_l2param_change(struct i40e_info *ldev, struct i40e_client *cli
 	for (i = 0; i < I40E_CLIENT_MAX_USER_PRIORITY; i++)
 		l2params->qs_handle_list[i] = params->qos.prio_qos[i].qs_handle;
 
-	l2params->mss = (params->mtu) ? params->mtu - I40IW_MTU_TO_MSS : iwdev->mss;
+	l2params->mss = (params->mtu) ? params->mtu - I40IW_MTU_TO_MSS : iwdev->vsi.mss;
 
 	INIT_WORK(&work->work, i40iw_l2params_worker);
 	queue_work(iwdev->param_wq, &work->work);
@@ -1773,21 +1792,23 @@ static void i40iw_vf_reset(struct i40e_info *ldev, struct i40e_client *client, u
 	struct i40iw_vfdev *tmp_vfdev;
 	unsigned int i;
 	unsigned long flags;
+	struct i40iw_device *iwdev;
 
 	hdl = i40iw_find_i40e_handler(ldev);
 	if (!hdl)
 		return;
 
 	dev = &hdl->device.sc_dev;
+	iwdev = (struct i40iw_device *)dev->back_dev;
 
 	for (i = 0; i < I40IW_MAX_PE_ENABLED_VF_COUNT; i++) {
 		if (!dev->vf_dev[i] || (dev->vf_dev[i]->vf_id != vf_id))
 			continue;
 		/* free all resources allocated on behalf of vf */
 		tmp_vfdev = dev->vf_dev[i];
-		spin_lock_irqsave(&dev->dev_pestat.stats_lock, flags);
+		spin_lock_irqsave(&iwdev->vsi.pestat->lock, flags);
 		dev->vf_dev[i] = NULL;
-		spin_unlock_irqrestore(&dev->dev_pestat.stats_lock, flags);
+		spin_unlock_irqrestore(&iwdev->vsi.pestat->lock, flags);
 		i40iw_del_hmc_objects(dev, &tmp_vfdev->hmc_info, false, false);
 		/* remove vf hmc function */
 		memset(&hmc_fcn_info, 0, sizeof(hmc_fcn_info));
diff --git a/drivers/infiniband/hw/i40iw/i40iw_osdep.h b/drivers/infiniband/hw/i40iw/i40iw_osdep.h
index a6b18cd42e67..aa66c1c63dfa 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_osdep.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_osdep.h
@@ -209,9 +209,9 @@ void i40iw_terminate_del_timer(struct i40iw_sc_qp *qp);
 enum i40iw_status_code i40iw_hw_manage_vf_pble_bp(struct i40iw_device *iwdev,
 						  struct i40iw_manage_vf_pble_info *info,
 						  bool wait);
-struct i40iw_dev_pestat;
-void i40iw_hw_stats_start_timer(struct i40iw_sc_dev *);
-void i40iw_hw_stats_del_timer(struct i40iw_sc_dev *);
+struct i40iw_sc_vsi;
+void i40iw_hw_stats_start_timer(struct i40iw_sc_vsi *vsi);
+void i40iw_hw_stats_stop_timer(struct i40iw_sc_vsi *vsi);
 #define i40iw_mmiowb() mmiowb()
 void i40iw_wr32(struct i40iw_hw *hw, u32 reg, u32 value);
 u32  i40iw_rd32(struct i40iw_hw *hw, u32 reg);
diff --git a/drivers/infiniband/hw/i40iw/i40iw_p.h b/drivers/infiniband/hw/i40iw/i40iw_p.h
index 2a4bd3233b41..28a92fee0822 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_p.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_p.h
@@ -47,8 +47,6 @@ void i40iw_debug_buf(struct i40iw_sc_dev *dev, enum i40iw_debug_flag mask,
 enum i40iw_status_code i40iw_device_init(struct i40iw_sc_dev *dev,
 					 struct i40iw_device_init_info *info);
 
-void i40iw_device_init_pestat(struct i40iw_dev_pestat *devstat);
-
 void i40iw_sc_cqp_post_sq(struct i40iw_sc_cqp *cqp);
 
 u64 *i40iw_sc_cqp_get_next_send_wqe(struct i40iw_sc_cqp *cqp, u64 scratch);
@@ -64,9 +62,24 @@ enum i40iw_status_code i40iw_sc_init_iw_hmc(struct i40iw_sc_dev *dev,
 enum i40iw_status_code i40iw_pf_init_vfhmc(struct i40iw_sc_dev *dev, u8 vf_hmc_fn_id,
 					   u32 *vf_cnt_array);
 
-/* cqp misc functions */
-void i40iw_change_l2params(struct i40iw_sc_dev *dev, struct i40iw_l2params *l2params);
-void i40iw_qp_add_qos(struct i40iw_sc_dev *dev, struct i40iw_sc_qp *qp);
+/* stats functions */
+void i40iw_hw_stats_refresh_all(struct i40iw_vsi_pestat *stats);
+void i40iw_hw_stats_read_all(struct i40iw_vsi_pestat *stats, struct i40iw_dev_hw_stats *stats_values);
+void i40iw_hw_stats_read_32(struct i40iw_vsi_pestat *stats,
+			    enum i40iw_hw_stats_index_32b index,
+			    u64 *value);
+void i40iw_hw_stats_read_64(struct i40iw_vsi_pestat *stats,
+			    enum i40iw_hw_stats_index_64b index,
+			    u64 *value);
+void i40iw_hw_stats_init(struct i40iw_vsi_pestat *stats, u8 index, bool is_pf);
+
+/* vsi misc functions */
+enum i40iw_status_code i40iw_vsi_stats_init(struct i40iw_sc_vsi *vsi, struct i40iw_vsi_stats_info *info);
+void i40iw_vsi_stats_free(struct i40iw_sc_vsi *vsi);
+void i40iw_sc_vsi_init(struct i40iw_sc_vsi *vsi, struct i40iw_vsi_init_info *info);
+
+void i40iw_change_l2params(struct i40iw_sc_vsi *vsi, struct i40iw_l2params *l2params);
+void i40iw_qp_add_qos(struct i40iw_sc_qp *qp);
 
 void i40iw_terminate_send_fin(struct i40iw_sc_qp *qp);
 
diff --git a/drivers/infiniband/hw/i40iw/i40iw_puda.c b/drivers/infiniband/hw/i40iw/i40iw_puda.c
index c3d28baf8e3a..449ba8c81ce7 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_puda.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_puda.c
@@ -42,12 +42,13 @@
 #include "i40iw_p.h"
 #include "i40iw_puda.h"
 
-static void i40iw_ieq_receive(struct i40iw_sc_dev *dev,
+static void i40iw_ieq_receive(struct i40iw_sc_vsi *vsi,
 			      struct i40iw_puda_buf *buf);
-static void i40iw_ieq_tx_compl(struct i40iw_sc_dev *dev, void *sqwrid);
+static void i40iw_ieq_tx_compl(struct i40iw_sc_vsi *vsi, void *sqwrid);
 static void i40iw_ilq_putback_rcvbuf(struct i40iw_sc_qp *qp, u32 wqe_idx);
 static enum i40iw_status_code i40iw_puda_replenish_rq(struct i40iw_puda_rsrc
 						      *rsrc, bool initial);
+static void i40iw_ieq_cleanup_qp(struct i40iw_puda_rsrc *ieq, struct i40iw_sc_qp *qp);
 /**
  * i40iw_puda_get_listbuf - get buffer from puda list
  * @list: list to use for buffers (ILQ or IEQ)
@@ -292,7 +293,7 @@ enum i40iw_status_code i40iw_puda_poll_completion(struct i40iw_sc_dev *dev,
 	unsigned long	flags;
 
 	if ((cq_type == I40IW_CQ_TYPE_ILQ) || (cq_type == I40IW_CQ_TYPE_IEQ)) {
-		rsrc = (cq_type == I40IW_CQ_TYPE_ILQ) ? dev->ilq : dev->ieq;
+		rsrc = (cq_type == I40IW_CQ_TYPE_ILQ) ? cq->vsi->ilq : cq->vsi->ieq;
 	} else {
 		i40iw_debug(dev, I40IW_DEBUG_PUDA, "%s qp_type error\n", __func__);
 		return I40IW_ERR_BAD_PTR;
@@ -335,7 +336,7 @@ enum i40iw_status_code i40iw_puda_poll_completion(struct i40iw_sc_dev *dev,
 		rsrc->stats_pkt_rcvd++;
 		rsrc->compl_rxwqe_idx = info.wqe_idx;
 		i40iw_debug(dev, I40IW_DEBUG_PUDA, "%s RQ completion\n", __func__);
-		rsrc->receive(rsrc->dev, buf);
+		rsrc->receive(rsrc->vsi, buf);
 		if (cq_type == I40IW_CQ_TYPE_ILQ)
 			i40iw_ilq_putback_rcvbuf(&rsrc->qp, info.wqe_idx);
 		else
@@ -345,12 +346,12 @@ enum i40iw_status_code i40iw_puda_poll_completion(struct i40iw_sc_dev *dev,
 		i40iw_debug(dev, I40IW_DEBUG_PUDA, "%s SQ completion\n", __func__);
 		sqwrid = (void *)(uintptr_t)qp->sq_wrtrk_array[info.wqe_idx].wrid;
 		I40IW_RING_SET_TAIL(qp->sq_ring, info.wqe_idx);
-		rsrc->xmit_complete(rsrc->dev, sqwrid);
+		rsrc->xmit_complete(rsrc->vsi, sqwrid);
 		spin_lock_irqsave(&rsrc->bufpool_lock, flags);
 		rsrc->tx_wqe_avail_cnt++;
 		spin_unlock_irqrestore(&rsrc->bufpool_lock, flags);
-		if (!list_empty(&dev->ilq->txpend))
-			i40iw_puda_send_buf(dev->ilq, NULL);
+		if (!list_empty(&rsrc->vsi->ilq->txpend))
+			i40iw_puda_send_buf(rsrc->vsi->ilq, NULL);
 	}
 
 done:
@@ -513,10 +514,8 @@ static void i40iw_puda_qp_setctx(struct i40iw_puda_rsrc *rsrc)
  * i40iw_puda_qp_wqe - setup wqe for qp create
  * @rsrc: resource for qp
  */
-static enum i40iw_status_code i40iw_puda_qp_wqe(struct i40iw_puda_rsrc *rsrc)
+static enum i40iw_status_code i40iw_puda_qp_wqe(struct i40iw_sc_dev *dev, struct i40iw_sc_qp *qp)
 {
-	struct i40iw_sc_qp *qp = &rsrc->qp;
-	struct i40iw_sc_dev *dev = rsrc->dev;
 	struct i40iw_sc_cqp *cqp;
 	u64 *wqe;
 	u64 header;
@@ -582,6 +581,7 @@ static enum i40iw_status_code i40iw_puda_qp_create(struct i40iw_puda_rsrc *rsrc)
 	qp->back_qp = (void *)rsrc;
 	qp->sq_pa = mem->pa;
 	qp->rq_pa = qp->sq_pa + sq_size;
+	qp->vsi = rsrc->vsi;
 	ukqp->sq_base = mem->va;
 	ukqp->rq_base = &ukqp->sq_base[rsrc->sq_size];
 	ukqp->shadow_area = ukqp->rq_base[rsrc->rq_size].elem;
@@ -609,14 +609,61 @@ static enum i40iw_status_code i40iw_puda_qp_create(struct i40iw_puda_rsrc *rsrc)
 						    I40E_VFPE_WQEALLOC1);
 
 	qp->user_pri = 0;
-	i40iw_qp_add_qos(rsrc->dev, qp);
+	i40iw_qp_add_qos(qp);
 	i40iw_puda_qp_setctx(rsrc);
-	ret = i40iw_puda_qp_wqe(rsrc);
+	if (rsrc->ceq_valid)
+		ret = i40iw_cqp_qp_create_cmd(rsrc->dev, qp);
+	else
+		ret = i40iw_puda_qp_wqe(rsrc->dev, qp);
 	if (ret)
 		i40iw_free_dma_mem(rsrc->dev->hw, &rsrc->qpmem);
 	return ret;
 }
 
+/**
+ * i40iw_puda_cq_wqe - setup wqe for cq create
+ * @rsrc: resource for cq
+ */
+static enum i40iw_status_code i40iw_puda_cq_wqe(struct i40iw_sc_dev *dev, struct i40iw_sc_cq *cq)
+{
+	u64 *wqe;
+	struct i40iw_sc_cqp *cqp;
+	u64 header;
+	struct i40iw_ccq_cqe_info compl_info;
+	enum i40iw_status_code status = 0;
+
+	cqp = dev->cqp;
+	wqe = i40iw_sc_cqp_get_next_send_wqe(cqp, 0);
+	if (!wqe)
+		return I40IW_ERR_RING_FULL;
+
+	set_64bit_val(wqe, 0, cq->cq_uk.cq_size);
+	set_64bit_val(wqe, 8, RS_64_1(cq, 1));
+	set_64bit_val(wqe, 16,
+		      LS_64(cq->shadow_read_threshold,
+			    I40IW_CQPSQ_CQ_SHADOW_READ_THRESHOLD));
+	set_64bit_val(wqe, 32, cq->cq_pa);
+
+	set_64bit_val(wqe, 40, cq->shadow_area_pa);
+
+	header = cq->cq_uk.cq_id |
+	    LS_64(I40IW_CQP_OP_CREATE_CQ, I40IW_CQPSQ_OPCODE) |
+	    LS_64(1, I40IW_CQPSQ_CQ_CHKOVERFLOW) |
+	    LS_64(1, I40IW_CQPSQ_CQ_ENCEQEMASK) |
+	    LS_64(1, I40IW_CQPSQ_CQ_CEQIDVALID) |
+	    LS_64(cqp->polarity, I40IW_CQPSQ_WQEVALID);
+	set_64bit_val(wqe, 24, header);
+
+	i40iw_debug_buf(dev, I40IW_DEBUG_PUDA, "PUDA CQE",
+			wqe, I40IW_CQP_WQE_SIZE * 8);
+
+	i40iw_sc_cqp_post_sq(dev->cqp);
+	status = dev->cqp_ops->poll_for_cqp_op_done(dev->cqp,
+						 I40IW_CQP_OP_CREATE_CQ,
+						 &compl_info);
+	return status;
+}
+
 /**
  * i40iw_puda_cq_create - create cq for resource
  * @rsrc: resource for which cq to create
@@ -625,18 +672,13 @@ static enum i40iw_status_code i40iw_puda_cq_create(struct i40iw_puda_rsrc *rsrc)
 {
 	struct i40iw_sc_dev *dev = rsrc->dev;
 	struct i40iw_sc_cq *cq = &rsrc->cq;
-	u64 *wqe;
-	struct i40iw_sc_cqp *cqp;
-	u64 header;
 	enum i40iw_status_code ret = 0;
 	u32 tsize, cqsize;
-	u32 shadow_read_threshold = 128;
 	struct i40iw_dma_mem *mem;
-	struct i40iw_ccq_cqe_info compl_info;
 	struct i40iw_cq_init_info info;
 	struct i40iw_cq_uk_init_info *init_info = &info.cq_uk_init_info;
 
-	cq->back_cq = (void *)rsrc;
+	cq->vsi = rsrc->vsi;
 	cqsize = rsrc->cq_size * (sizeof(struct i40iw_cqe));
 	tsize = cqsize + sizeof(struct i40iw_cq_shadow_area);
 	ret = i40iw_allocate_dma_mem(dev->hw, &rsrc->cqmem, tsize,
@@ -657,43 +699,84 @@ static enum i40iw_status_code i40iw_puda_cq_create(struct i40iw_puda_rsrc *rsrc)
 	init_info->shadow_area = (u64 *)((u8 *)mem->va + cqsize);
 	init_info->cq_size = rsrc->cq_size;
 	init_info->cq_id = rsrc->cq_id;
+	info.ceqe_mask = true;
+	info.ceq_id_valid = true;
 	ret = dev->iw_priv_cq_ops->cq_init(cq, &info);
 	if (ret)
 		goto error;
-	cqp = dev->cqp;
-	wqe = i40iw_sc_cqp_get_next_send_wqe(cqp, 0);
-	if (!wqe) {
-		ret = I40IW_ERR_RING_FULL;
-		goto error;
-	}
+	if (rsrc->ceq_valid)
+		ret = i40iw_cqp_cq_create_cmd(dev, cq);
+	else
+		ret = i40iw_puda_cq_wqe(dev, cq);
+error:
+	if (ret)
+		i40iw_free_dma_mem(dev->hw, &rsrc->cqmem);
+	return ret;
+}
 
-	set_64bit_val(wqe, 0, rsrc->cq_size);
-	set_64bit_val(wqe, 8, RS_64_1(cq, 1));
-	set_64bit_val(wqe, 16, LS_64(shadow_read_threshold, I40IW_CQPSQ_CQ_SHADOW_READ_THRESHOLD));
-	set_64bit_val(wqe, 32, cq->cq_pa);
+/**
+ * i40iw_puda_free_qp - free qp for resource
+ * @rsrc: resource for which qp to free
+ */
+static void i40iw_puda_free_qp(struct i40iw_puda_rsrc *rsrc)
+{
+	enum i40iw_status_code ret;
+	struct i40iw_ccq_cqe_info compl_info;
+	struct i40iw_sc_dev *dev = rsrc->dev;
 
-	set_64bit_val(wqe, 40, cq->shadow_area_pa);
+	if (rsrc->ceq_valid) {
+		i40iw_cqp_qp_destroy_cmd(dev, &rsrc->qp);
+		return;
+	}
 
-	header = rsrc->cq_id |
-	    LS_64(I40IW_CQP_OP_CREATE_CQ, I40IW_CQPSQ_OPCODE) |
-	    LS_64(1, I40IW_CQPSQ_CQ_CHKOVERFLOW) |
-	    LS_64(1, I40IW_CQPSQ_CQ_ENCEQEMASK) |
-	    LS_64(1, I40IW_CQPSQ_CQ_CEQIDVALID) |
-	    LS_64(cqp->polarity, I40IW_CQPSQ_WQEVALID);
-	set_64bit_val(wqe, 24, header);
+	ret = dev->iw_priv_qp_ops->qp_destroy(&rsrc->qp,
+			0, false, true, true);
+	if (ret)
+		i40iw_debug(dev, I40IW_DEBUG_PUDA,
+			    "%s error puda qp destroy wqe\n",
+			    __func__);
 
-	i40iw_debug_buf(dev, I40IW_DEBUG_PUDA, "PUDA CQE",
-			wqe, I40IW_CQP_WQE_SIZE * 8);
+	if (!ret) {
+		ret = dev->cqp_ops->poll_for_cqp_op_done(dev->cqp,
+				I40IW_CQP_OP_DESTROY_QP,
+				&compl_info);
+		if (ret)
+			i40iw_debug(dev, I40IW_DEBUG_PUDA,
+				    "%s error puda qp destroy failed\n",
+				    __func__);
+	}
+}
 
-	i40iw_sc_cqp_post_sq(dev->cqp);
-	ret = dev->cqp_ops->poll_for_cqp_op_done(dev->cqp,
-						 I40IW_CQP_OP_CREATE_CQ,
-						 &compl_info);
+/**
+ * i40iw_puda_free_cq - free cq for resource
+ * @rsrc: resource for which cq to free
+ */
+static void i40iw_puda_free_cq(struct i40iw_puda_rsrc *rsrc)
+{
+	enum i40iw_status_code ret;
+	struct i40iw_ccq_cqe_info compl_info;
+	struct i40iw_sc_dev *dev = rsrc->dev;
+
+	if (rsrc->ceq_valid) {
+		i40iw_cqp_cq_destroy_cmd(dev, &rsrc->cq);
+		return;
+	}
+	ret = dev->iw_priv_cq_ops->cq_destroy(&rsrc->cq, 0, true);
 
-error:
 	if (ret)
-		i40iw_free_dma_mem(dev->hw, &rsrc->cqmem);
-	return ret;
+		i40iw_debug(dev, I40IW_DEBUG_PUDA,
+			    "%s error ieq cq destroy\n",
+			    __func__);
+
+	if (!ret) {
+		ret = dev->cqp_ops->poll_for_cqp_op_done(dev->cqp,
+				I40IW_CQP_OP_DESTROY_CQ,
+				&compl_info);
+		if (ret)
+			i40iw_debug(dev, I40IW_DEBUG_PUDA,
+				    "%s error ieq qp destroy done\n",
+				    __func__);
+	}
 }
 
 /**
@@ -702,25 +785,24 @@ error:
  * @type: type of resource to dele
  * @reset: true if reset chip
  */
-void i40iw_puda_dele_resources(struct i40iw_sc_dev *dev,
+void i40iw_puda_dele_resources(struct i40iw_sc_vsi *vsi,
 			       enum puda_resource_type type,
 			       bool reset)
 {
-	struct i40iw_ccq_cqe_info compl_info;
+	struct i40iw_sc_dev *dev = vsi->dev;
 	struct i40iw_puda_rsrc *rsrc;
 	struct i40iw_puda_buf *buf = NULL;
 	struct i40iw_puda_buf *nextbuf = NULL;
 	struct i40iw_virt_mem *vmem;
-	enum i40iw_status_code ret;
 
 	switch (type) {
 	case I40IW_PUDA_RSRC_TYPE_ILQ:
-		rsrc = dev->ilq;
-		vmem = &dev->ilq_mem;
+		rsrc = vsi->ilq;
+		vmem = &vsi->ilq_mem;
 		break;
 	case I40IW_PUDA_RSRC_TYPE_IEQ:
-		rsrc = dev->ieq;
-		vmem = &dev->ieq_mem;
+		rsrc = vsi->ieq;
+		vmem = &vsi->ieq_mem;
 		break;
 	default:
 		i40iw_debug(dev, I40IW_DEBUG_PUDA, "%s: error resource type = 0x%x\n",
@@ -732,45 +814,14 @@ void i40iw_puda_dele_resources(struct i40iw_sc_dev *dev,
 	case PUDA_HASH_CRC_COMPLETE:
 		i40iw_free_hash_desc(rsrc->hash_desc);
 	case PUDA_QP_CREATED:
-		do {
-			if (reset)
-				break;
-			ret = dev->iw_priv_qp_ops->qp_destroy(&rsrc->qp,
-							      0, false, true, true);
-			if (ret)
-				i40iw_debug(rsrc->dev, I40IW_DEBUG_PUDA,
-					    "%s error ieq qp destroy\n",
-					    __func__);
-
-			ret = dev->cqp_ops->poll_for_cqp_op_done(dev->cqp,
-								 I40IW_CQP_OP_DESTROY_QP,
-								 &compl_info);
-			if (ret)
-				i40iw_debug(rsrc->dev, I40IW_DEBUG_PUDA,
-					    "%s error ieq qp destroy done\n",
-					    __func__);
-		} while (0);
+		if (!reset)
+			i40iw_puda_free_qp(rsrc);
 
 		i40iw_free_dma_mem(dev->hw, &rsrc->qpmem);
 		/* fallthrough */
 	case PUDA_CQ_CREATED:
-		do {
-			if (reset)
-				break;
-			ret = dev->iw_priv_cq_ops->cq_destroy(&rsrc->cq, 0, true);
-			if (ret)
-				i40iw_debug(rsrc->dev, I40IW_DEBUG_PUDA,
-					    "%s error ieq cq destroy\n",
-					    __func__);
-
-			ret = dev->cqp_ops->poll_for_cqp_op_done(dev->cqp,
-								 I40IW_CQP_OP_DESTROY_CQ,
-								 &compl_info);
-			if (ret)
-				i40iw_debug(rsrc->dev, I40IW_DEBUG_PUDA,
-					    "%s error ieq qp destroy done\n",
-					    __func__);
-		} while (0);
+		if (!reset)
+			i40iw_puda_free_cq(rsrc);
 
 		i40iw_free_dma_mem(dev->hw, &rsrc->cqmem);
 		break;
@@ -826,9 +877,10 @@ static enum i40iw_status_code i40iw_puda_allocbufs(struct i40iw_puda_rsrc *rsrc,
  * @dev: iwarp device
  * @info: resource information
  */
-enum i40iw_status_code i40iw_puda_create_rsrc(struct i40iw_sc_dev *dev,
+enum i40iw_status_code i40iw_puda_create_rsrc(struct i40iw_sc_vsi *vsi,
 					      struct i40iw_puda_rsrc_info *info)
 {
+	struct i40iw_sc_dev *dev = vsi->dev;
 	enum i40iw_status_code ret = 0;
 	struct i40iw_puda_rsrc *rsrc;
 	u32 pudasize;
@@ -841,10 +893,10 @@ enum i40iw_status_code i40iw_puda_create_rsrc(struct i40iw_sc_dev *dev,
 	rqwridsize = info->rq_size * 8;
 	switch (info->type) {
 	case I40IW_PUDA_RSRC_TYPE_ILQ:
-		vmem = &dev->ilq_mem;
+		vmem = &vsi->ilq_mem;
 		break;
 	case I40IW_PUDA_RSRC_TYPE_IEQ:
-		vmem = &dev->ieq_mem;
+		vmem = &vsi->ieq_mem;
 		break;
 	default:
 		return I40IW_NOT_SUPPORTED;
@@ -857,22 +909,22 @@ enum i40iw_status_code i40iw_puda_create_rsrc(struct i40iw_sc_dev *dev,
 	rsrc = (struct i40iw_puda_rsrc *)vmem->va;
 	spin_lock_init(&rsrc->bufpool_lock);
 	if (info->type == I40IW_PUDA_RSRC_TYPE_ILQ) {
-		dev->ilq = (struct i40iw_puda_rsrc *)vmem->va;
-		dev->ilq_count = info->count;
+		vsi->ilq = (struct i40iw_puda_rsrc *)vmem->va;
+		vsi->ilq_count = info->count;
 		rsrc->receive = info->receive;
 		rsrc->xmit_complete = info->xmit_complete;
 	} else {
-		vmem = &dev->ieq_mem;
-		dev->ieq_count = info->count;
-		dev->ieq = (struct i40iw_puda_rsrc *)vmem->va;
+		vmem = &vsi->ieq_mem;
+		vsi->ieq_count = info->count;
+		vsi->ieq = (struct i40iw_puda_rsrc *)vmem->va;
 		rsrc->receive = i40iw_ieq_receive;
 		rsrc->xmit_complete = i40iw_ieq_tx_compl;
 	}
 
+	rsrc->ceq_valid = info->ceq_valid;
 	rsrc->type = info->type;
 	rsrc->sq_wrtrk_array = (struct i40iw_sq_uk_wr_trk_info *)((u8 *)vmem->va + pudasize);
 	rsrc->rq_wrid_array = (u64 *)((u8 *)vmem->va + pudasize + sqwridsize);
-	rsrc->mss = info->mss;
 	/* Initialize all ieq lists */
 	INIT_LIST_HEAD(&rsrc->bufpool);
 	INIT_LIST_HEAD(&rsrc->txpend);
@@ -886,6 +938,7 @@ enum i40iw_status_code i40iw_puda_create_rsrc(struct i40iw_sc_dev *dev,
 	rsrc->cq_size = info->rq_size + info->sq_size;
 	rsrc->buf_size = info->buf_size;
 	rsrc->dev = dev;
+	rsrc->vsi = vsi;
 
 	ret = i40iw_puda_cq_create(rsrc);
 	if (!ret) {
@@ -920,7 +973,7 @@ enum i40iw_status_code i40iw_puda_create_rsrc(struct i40iw_sc_dev *dev,
 	dev->ccq_ops->ccq_arm(&rsrc->cq);
 	return ret;
  error:
-	i40iw_puda_dele_resources(dev, info->type, false);
+	i40iw_puda_dele_resources(vsi, info->type, false);
 
 	return ret;
 }
@@ -1333,7 +1386,7 @@ static void i40iw_ieq_handle_exception(struct i40iw_puda_rsrc *ieq,
 	}
 	if (pfpdu->mode && (fps != pfpdu->fps)) {
 		/* clean up qp as it is new partial sequence */
-		i40iw_ieq_cleanup_qp(ieq->dev, qp);
+		i40iw_ieq_cleanup_qp(ieq, qp);
 		i40iw_debug(ieq->dev, I40IW_DEBUG_IEQ,
 			    "%s: restarting new partial\n", __func__);
 		pfpdu->mode = false;
@@ -1345,7 +1398,7 @@ static void i40iw_ieq_handle_exception(struct i40iw_puda_rsrc *ieq,
 		pfpdu->rcv_nxt = fps;
 		pfpdu->fps = fps;
 		pfpdu->mode = true;
-		pfpdu->max_fpdu_data = ieq->mss;
+		pfpdu->max_fpdu_data = ieq->vsi->mss;
 		pfpdu->pmode_count++;
 		INIT_LIST_HEAD(rxlist);
 		i40iw_ieq_check_first_buf(buf, fps);
@@ -1380,14 +1433,14 @@ static void i40iw_ieq_handle_exception(struct i40iw_puda_rsrc *ieq,
  * @dev: iwarp device
  * @buf: exception buffer received
  */
-static void i40iw_ieq_receive(struct i40iw_sc_dev *dev,
+static void i40iw_ieq_receive(struct i40iw_sc_vsi *vsi,
 			      struct i40iw_puda_buf *buf)
 {
-	struct i40iw_puda_rsrc *ieq = dev->ieq;
+	struct i40iw_puda_rsrc *ieq = vsi->ieq;
 	struct i40iw_sc_qp *qp = NULL;
 	u32 wqe_idx = ieq->compl_rxwqe_idx;
 
-	qp = i40iw_ieq_get_qp(dev, buf);
+	qp = i40iw_ieq_get_qp(vsi->dev, buf);
 	if (!qp) {
 		ieq->stats_bad_qp_id++;
 		i40iw_puda_ret_bufpool(ieq, buf);
@@ -1405,12 +1458,12 @@ static void i40iw_ieq_receive(struct i40iw_sc_dev *dev,
 
 /**
  * i40iw_ieq_tx_compl - put back after sending completed exception buffer
- * @dev: iwarp device
+ * @vsi: pointer to the vsi structure
  * @sqwrid: pointer to puda buffer
  */
-static void i40iw_ieq_tx_compl(struct i40iw_sc_dev *dev, void *sqwrid)
+static void i40iw_ieq_tx_compl(struct i40iw_sc_vsi *vsi, void *sqwrid)
 {
-	struct i40iw_puda_rsrc *ieq = dev->ieq;
+	struct i40iw_puda_rsrc *ieq = vsi->ieq;
 	struct i40iw_puda_buf *buf = (struct i40iw_puda_buf *)sqwrid;
 
 	i40iw_puda_ret_bufpool(ieq, buf);
@@ -1422,15 +1475,14 @@ static void i40iw_ieq_tx_compl(struct i40iw_sc_dev *dev, void *sqwrid)
 
 /**
  * i40iw_ieq_cleanup_qp - qp is being destroyed
- * @dev: iwarp device
+ * @ieq: ieq resource
  * @qp: all pending fpdu buffers
  */
-void i40iw_ieq_cleanup_qp(struct i40iw_sc_dev *dev, struct i40iw_sc_qp *qp)
+static void i40iw_ieq_cleanup_qp(struct i40iw_puda_rsrc *ieq, struct i40iw_sc_qp *qp)
 {
 	struct i40iw_puda_buf *buf;
 	struct i40iw_pfpdu *pfpdu = &qp->pfpdu;
 	struct list_head *rxlist = &pfpdu->rxlist;
-	struct i40iw_puda_rsrc *ieq = dev->ieq;
 
 	if (!pfpdu->mode)
 		return;
diff --git a/drivers/infiniband/hw/i40iw/i40iw_puda.h b/drivers/infiniband/hw/i40iw/i40iw_puda.h
index 52bf7826ce4e..dba05ce7d392 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_puda.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_puda.h
@@ -100,6 +100,7 @@ struct i40iw_puda_rsrc_info {
 	enum puda_resource_type type;	/* ILQ or IEQ */
 	u32 count;
 	u16 pd_id;
+	bool ceq_valid;
 	u32 cq_id;
 	u32 qp_id;
 	u32 sq_size;
@@ -107,8 +108,8 @@ struct i40iw_puda_rsrc_info {
 	u16 buf_size;
 	u16 mss;
 	u32 tx_buf_cnt;		/* total bufs allocated will be rq_size + tx_buf_cnt */
-	void (*receive)(struct i40iw_sc_dev *, struct i40iw_puda_buf *);
-	void (*xmit_complete)(struct i40iw_sc_dev *, void *);
+	void (*receive)(struct i40iw_sc_vsi *, struct i40iw_puda_buf *);
+	void (*xmit_complete)(struct i40iw_sc_vsi *, void *);
 };
 
 struct i40iw_puda_rsrc {
@@ -116,6 +117,7 @@ struct i40iw_puda_rsrc {
 	struct i40iw_sc_qp qp;
 	struct i40iw_sc_pd sc_pd;
 	struct i40iw_sc_dev *dev;
+	struct i40iw_sc_vsi *vsi;
 	struct i40iw_dma_mem cqmem;
 	struct i40iw_dma_mem qpmem;
 	struct i40iw_virt_mem ilq_mem;
@@ -123,6 +125,7 @@ struct i40iw_puda_rsrc {
 	enum puda_resource_type type;
 	u16 buf_size;		/*buffer must be max datalen + tcpip hdr + mac */
 	u16 mss;
+	bool ceq_valid;
 	u32 cq_id;
 	u32 qp_id;
 	u32 sq_size;
@@ -142,8 +145,8 @@ struct i40iw_puda_rsrc {
 	u32 avail_buf_count;		/* snapshot of currently available buffers */
 	spinlock_t bufpool_lock;
 	struct i40iw_puda_buf *alloclist;
-	void (*receive)(struct i40iw_sc_dev *, struct i40iw_puda_buf *);
-	void (*xmit_complete)(struct i40iw_sc_dev *, void *);
+	void (*receive)(struct i40iw_sc_vsi *, struct i40iw_puda_buf *);
+	void (*xmit_complete)(struct i40iw_sc_vsi *, void *);
 	/* puda stats */
 	u64 stats_buf_alloc_fail;
 	u64 stats_pkt_rcvd;
@@ -160,14 +163,13 @@ void i40iw_puda_send_buf(struct i40iw_puda_rsrc *rsrc,
 			 struct i40iw_puda_buf *buf);
 enum i40iw_status_code i40iw_puda_send(struct i40iw_sc_qp *qp,
 				       struct i40iw_puda_send_info *info);
-enum i40iw_status_code i40iw_puda_create_rsrc(struct i40iw_sc_dev *dev,
+enum i40iw_status_code i40iw_puda_create_rsrc(struct i40iw_sc_vsi *vsi,
 					      struct i40iw_puda_rsrc_info *info);
-void i40iw_puda_dele_resources(struct i40iw_sc_dev *dev,
+void i40iw_puda_dele_resources(struct i40iw_sc_vsi *vsi,
 			       enum puda_resource_type type,
 			       bool reset);
 enum i40iw_status_code i40iw_puda_poll_completion(struct i40iw_sc_dev *dev,
 						  struct i40iw_sc_cq *cq, u32 *compl_err);
-void i40iw_ieq_cleanup_qp(struct i40iw_sc_dev *dev, struct i40iw_sc_qp *qp);
 
 struct i40iw_sc_qp *i40iw_ieq_get_qp(struct i40iw_sc_dev *dev,
 				     struct i40iw_puda_buf *buf);
@@ -180,4 +182,8 @@ void i40iw_ieq_mpa_crc_ae(struct i40iw_sc_dev *dev, struct i40iw_sc_qp *qp);
 void i40iw_free_hash_desc(struct shash_desc *desc);
 void i40iw_ieq_update_tcpip_info(struct i40iw_puda_buf *buf, u16 length,
 				 u32 seqnum);
+enum i40iw_status_code i40iw_cqp_qp_create_cmd(struct i40iw_sc_dev *dev, struct i40iw_sc_qp *qp);
+enum i40iw_status_code i40iw_cqp_cq_create_cmd(struct i40iw_sc_dev *dev, struct i40iw_sc_cq *cq);
+void i40iw_cqp_qp_destroy_cmd(struct i40iw_sc_dev *dev, struct i40iw_sc_qp *qp);
+void i40iw_cqp_cq_destroy_cmd(struct i40iw_sc_dev *dev, struct i40iw_sc_cq *cq);
 #endif
diff --git a/drivers/infiniband/hw/i40iw/i40iw_type.h b/drivers/infiniband/hw/i40iw/i40iw_type.h
index 928d91b956a8..f3f8e9cc3c05 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_type.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_type.h
@@ -61,7 +61,7 @@ struct i40iw_cq_shadow_area {
 
 struct i40iw_sc_dev;
 struct i40iw_hmc_info;
-struct i40iw_dev_pestat;
+struct i40iw_vsi_pestat;
 
 struct i40iw_cqp_ops;
 struct i40iw_ccq_ops;
@@ -191,7 +191,7 @@ enum i40iw_debug_flag {
 	I40IW_DEBUG_ALL		= 0xFFFFFFFF
 };
 
-enum i40iw_hw_stat_index_32b {
+enum i40iw_hw_stats_index_32b {
 	I40IW_HW_STAT_INDEX_IP4RXDISCARD = 0,
 	I40IW_HW_STAT_INDEX_IP4RXTRUNC,
 	I40IW_HW_STAT_INDEX_IP4TXNOROUTE,
@@ -204,7 +204,7 @@ enum i40iw_hw_stat_index_32b {
 	I40IW_HW_STAT_INDEX_MAX_32
 };
 
-enum i40iw_hw_stat_index_64b {
+enum i40iw_hw_stats_index_64b {
 	I40IW_HW_STAT_INDEX_IP4RXOCTS = 0,
 	I40IW_HW_STAT_INDEX_IP4RXPKTS,
 	I40IW_HW_STAT_INDEX_IP4RXFRAGS,
@@ -234,32 +234,23 @@ enum i40iw_hw_stat_index_64b {
 	I40IW_HW_STAT_INDEX_MAX_64
 };
 
-struct i40iw_dev_hw_stat_offsets {
-	u32 stat_offset_32[I40IW_HW_STAT_INDEX_MAX_32];
-	u32 stat_offset_64[I40IW_HW_STAT_INDEX_MAX_64];
+struct i40iw_dev_hw_stats_offsets {
+	u32 stats_offset_32[I40IW_HW_STAT_INDEX_MAX_32];
+	u32 stats_offset_64[I40IW_HW_STAT_INDEX_MAX_64];
 };
 
 struct i40iw_dev_hw_stats {
-	u64 stat_value_32[I40IW_HW_STAT_INDEX_MAX_32];
-	u64 stat_value_64[I40IW_HW_STAT_INDEX_MAX_64];
+	u64 stats_value_32[I40IW_HW_STAT_INDEX_MAX_32];
+	u64 stats_value_64[I40IW_HW_STAT_INDEX_MAX_64];
 };
 
-struct i40iw_device_pestat_ops {
-	void (*iw_hw_stat_init)(struct i40iw_dev_pestat *, u8, struct i40iw_hw *, bool);
-	void (*iw_hw_stat_read_32)(struct i40iw_dev_pestat *, enum i40iw_hw_stat_index_32b, u64 *);
-	void (*iw_hw_stat_read_64)(struct i40iw_dev_pestat *, enum i40iw_hw_stat_index_64b, u64 *);
-	void (*iw_hw_stat_read_all)(struct i40iw_dev_pestat *, struct i40iw_dev_hw_stats *);
-	void (*iw_hw_stat_refresh_all)(struct i40iw_dev_pestat *);
-};
-
-struct i40iw_dev_pestat {
+struct i40iw_vsi_pestat {
 	struct i40iw_hw *hw;
-	struct i40iw_device_pestat_ops ops;
 	struct i40iw_dev_hw_stats hw_stats;
 	struct i40iw_dev_hw_stats last_read_hw_stats;
-	struct i40iw_dev_hw_stat_offsets hw_stat_offsets;
+	struct i40iw_dev_hw_stats_offsets hw_stats_offsets;
 	struct timer_list stats_timer;
-	spinlock_t stats_lock; /* rdma stats lock */
+	spinlock_t lock; /* rdma stats lock */
 };
 
 struct i40iw_hw {
@@ -355,6 +346,7 @@ struct i40iw_sc_cq {
 	u64 cq_pa;
 	u64 shadow_area_pa;
 	struct i40iw_sc_dev *dev;
+	struct i40iw_sc_vsi *vsi;
 	void *pbl_list;
 	void *back_cq;
 	u32 ceq_id;
@@ -378,6 +370,7 @@ struct i40iw_sc_qp {
 	u64 shadow_area_pa;
 	u64 q2_pa;
 	struct i40iw_sc_dev *dev;
+	struct i40iw_sc_vsi *vsi;
 	struct i40iw_sc_pd *pd;
 	u64 *hw_host_ctx;
 	void *llp_stream_handle;
@@ -441,7 +434,7 @@ struct i40iw_qos {
 struct i40iw_vfdev {
 	struct i40iw_sc_dev *pf_dev;
 	u8 *hmc_info_mem;
-	struct i40iw_dev_pestat dev_pestat;
+	struct i40iw_vsi_pestat pestat;
 	struct i40iw_hmc_pble_info *pble_info;
 	struct i40iw_hmc_info hmc_info;
 	struct i40iw_vchnl_vf_msg_buffer vf_msg_buffer;
@@ -455,11 +448,28 @@ struct i40iw_vfdev {
 	bool stats_initialized;
 };
 
+#define I40IW_INVALID_FCN_ID 0xff
+struct i40iw_sc_vsi {
+	struct i40iw_sc_dev *dev;
+	void *back_vsi; /* Owned by OS */
+	u32 ilq_count;
+	struct i40iw_virt_mem ilq_mem;
+	struct i40iw_puda_rsrc *ilq;
+	u32 ieq_count;
+	struct i40iw_virt_mem ieq_mem;
+	struct i40iw_puda_rsrc *ieq;
+	u16 mss;
+	u8 fcn_id;
+	bool stats_fcn_id_alloc;
+	struct i40iw_qos qos[I40IW_MAX_USER_PRIORITY];
+	struct i40iw_vsi_pestat *pestat;
+};
+
 struct i40iw_sc_dev {
 	struct list_head cqp_cmd_head;	/* head of the CQP command list */
 	spinlock_t cqp_lock; /* cqp list sync */
 	struct i40iw_dev_uk dev_uk;
-	struct i40iw_dev_pestat dev_pestat;
+	bool fcn_id_array[I40IW_MAX_STATS_COUNT];
 	struct i40iw_dma_mem vf_fpm_query_buf[I40IW_MAX_PE_ENABLED_VF_COUNT];
 	u64 fpm_query_buf_pa;
 	u64 fpm_commit_buf_pa;
@@ -486,18 +496,9 @@ struct i40iw_sc_dev {
 	struct i40iw_cqp_misc_ops *cqp_misc_ops;
 	struct i40iw_hmc_ops *hmc_ops;
 	struct i40iw_vchnl_if vchnl_if;
-	u32 ilq_count;
-	struct i40iw_virt_mem ilq_mem;
-	struct i40iw_puda_rsrc *ilq;
-	u32 ieq_count;
-	struct i40iw_virt_mem ieq_mem;
-	struct i40iw_puda_rsrc *ieq;
-
 	const struct i40iw_vf_cqp_ops *iw_vf_cqp_ops;
 
 	struct i40iw_hmc_fpm_misc hmc_fpm_misc;
-	struct i40iw_qos qos[I40IW_MAX_USER_PRIORITY];
-	u16 mss;
 	u32 debug_mask;
 	u16 exception_lan_queue;
 	u8 hmc_fn_id;
@@ -571,6 +572,19 @@ struct i40iw_l2params {
 	u16 mss;
 };
 
+struct i40iw_vsi_init_info {
+	struct i40iw_sc_dev *dev;
+	void  *back_vsi;
+	struct i40iw_l2params *params;
+};
+
+struct i40iw_vsi_stats_info {
+	struct i40iw_vsi_pestat *pestat;
+	u8 fcn_id;
+	bool alloc_fcn_id;
+	bool stats_initialize;
+};
+
 struct i40iw_device_init_info {
 	u64 fpm_query_buf_pa;
 	u64 fpm_commit_buf_pa;
@@ -579,7 +593,6 @@ struct i40iw_device_init_info {
 	struct i40iw_hw *hw;
 	void __iomem *bar0;
 	enum i40iw_status_code (*vchnl_send)(struct i40iw_sc_dev *, u32, u8 *, u16);
-	struct i40iw_l2params l2params;
 	u16 exception_lan_queue;
 	u8 hmc_fn_id;
 	bool is_pf;
@@ -831,6 +844,7 @@ struct i40iw_register_shared_stag {
 struct i40iw_qp_init_info {
 	struct i40iw_qp_uk_init_info qp_uk_init_info;
 	struct i40iw_sc_pd *pd;
+	struct i40iw_sc_vsi *vsi;
 	u64 *host_ctx;
 	u8 *q2;
 	u64 sq_pa;
@@ -897,6 +911,7 @@ enum i40iw_quad_hash_manage_type {
 };
 
 struct i40iw_qhash_table_info {
+	struct i40iw_sc_vsi *vsi;
 	enum i40iw_quad_hash_manage_type manage;
 	enum i40iw_quad_entry_type entry_type;
 	bool vlan_valid;
diff --git a/drivers/infiniband/hw/i40iw/i40iw_utils.c b/drivers/infiniband/hw/i40iw/i40iw_utils.c
index 7d4af7759970..0f5d43d1f5fc 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_utils.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_utils.c
@@ -761,7 +761,7 @@ void i40iw_qp_mss_modify(struct i40iw_sc_dev *dev, struct i40iw_sc_qp *qp)
 
 	memset(&info, 0, sizeof(info));
 	info.mss_change = true;
-	info.new_mss = dev->mss;
+	info.new_mss = qp->vsi->mss;
 	i40iw_hw_modify_qp(iwdev, iwqp, &info, false);
 }
 
@@ -1067,6 +1067,116 @@ enum i40iw_status_code i40iw_vf_wait_vchnl_resp(struct i40iw_sc_dev *dev)
 	return 0;
 }
 
+/**
+ * i40iw_cqp_cq_create_cmd - create a cq for the cqp
+ * @dev: device pointer
+ * @cq: pointer to created cq
+ */
+enum i40iw_status_code i40iw_cqp_cq_create_cmd(struct i40iw_sc_dev *dev,
+					       struct i40iw_sc_cq *cq)
+{
+	struct i40iw_device *iwdev = (struct i40iw_device *)dev->back_dev;
+	struct i40iw_cqp *iwcqp = &iwdev->cqp;
+	struct i40iw_cqp_request *cqp_request;
+	struct cqp_commands_info *cqp_info;
+	enum i40iw_status_code status;
+
+	cqp_request = i40iw_get_cqp_request(iwcqp, true);
+	if (!cqp_request)
+		return I40IW_ERR_NO_MEMORY;
+
+	cqp_info = &cqp_request->info;
+	cqp_info->cqp_cmd = OP_CQ_CREATE;
+	cqp_info->post_sq = 1;
+	cqp_info->in.u.cq_create.cq = cq;
+	cqp_info->in.u.cq_create.scratch = (uintptr_t)cqp_request;
+	status = i40iw_handle_cqp_op(iwdev, cqp_request);
+	if (status)
+		i40iw_pr_err("CQP-OP Create QP fail");
+
+	return status;
+}
+
+/**
+ * i40iw_cqp_qp_create_cmd - create a qp for the cqp
+ * @dev: device pointer
+ * @qp: pointer to created qp
+ */
+enum i40iw_status_code i40iw_cqp_qp_create_cmd(struct i40iw_sc_dev *dev,
+					       struct i40iw_sc_qp *qp)
+{
+	struct i40iw_device *iwdev = (struct i40iw_device *)dev->back_dev;
+	struct i40iw_cqp *iwcqp = &iwdev->cqp;
+	struct i40iw_cqp_request *cqp_request;
+	struct cqp_commands_info *cqp_info;
+	struct i40iw_create_qp_info *qp_info;
+	enum i40iw_status_code status;
+
+	cqp_request = i40iw_get_cqp_request(iwcqp, true);
+	if (!cqp_request)
+		return I40IW_ERR_NO_MEMORY;
+
+	cqp_info = &cqp_request->info;
+	qp_info = &cqp_request->info.in.u.qp_create.info;
+
+	memset(qp_info, 0, sizeof(*qp_info));
+
+	qp_info->cq_num_valid = true;
+	qp_info->next_iwarp_state = I40IW_QP_STATE_RTS;
+
+	cqp_info->cqp_cmd = OP_QP_CREATE;
+	cqp_info->post_sq = 1;
+	cqp_info->in.u.qp_create.qp = qp;
+	cqp_info->in.u.qp_create.scratch = (uintptr_t)cqp_request;
+	status = i40iw_handle_cqp_op(iwdev, cqp_request);
+	if (status)
+		i40iw_pr_err("CQP-OP QP create fail");
+	return status;
+}
+
+/**
+ * i40iw_cqp_cq_destroy_cmd - destroy the cqp cq
+ * @dev: device pointer
+ * @cq: pointer to cq
+ */
+void i40iw_cqp_cq_destroy_cmd(struct i40iw_sc_dev *dev, struct i40iw_sc_cq *cq)
+{
+	struct i40iw_device *iwdev = (struct i40iw_device *)dev->back_dev;
+
+	i40iw_cq_wq_destroy(iwdev, cq);
+}
+
+/**
+ * i40iw_cqp_qp_destroy_cmd - destroy the cqp
+ * @dev: device pointer
+ * @qp: pointer to qp
+ */
+void i40iw_cqp_qp_destroy_cmd(struct i40iw_sc_dev *dev, struct i40iw_sc_qp *qp)
+{
+	struct i40iw_device *iwdev = (struct i40iw_device *)dev->back_dev;
+	struct i40iw_cqp *iwcqp = &iwdev->cqp;
+	struct i40iw_cqp_request *cqp_request;
+	struct cqp_commands_info *cqp_info;
+	enum i40iw_status_code status;
+
+	cqp_request = i40iw_get_cqp_request(iwcqp, true);
+	if (!cqp_request)
+		return;
+
+	cqp_info = &cqp_request->info;
+	memset(cqp_info, 0, sizeof(*cqp_info));
+
+	cqp_info->cqp_cmd = OP_QP_DESTROY;
+	cqp_info->post_sq = 1;
+	cqp_info->in.u.qp_destroy.qp = qp;
+	cqp_info->in.u.qp_destroy.scratch = (uintptr_t)cqp_request;
+	cqp_info->in.u.qp_destroy.remove_hash_idx = true;
+	status = i40iw_handle_cqp_op(iwdev, cqp_request);
+	if (status)
+		i40iw_pr_err("CQP QP_DESTROY fail");
+}
+
+
 /**
  * i40iw_ieq_mpa_crc_ae - generate AE for crc error
  * @dev: hardware control device structure
@@ -1281,27 +1391,29 @@ enum i40iw_status_code i40iw_puda_get_tcpip_info(struct i40iw_puda_completion_in
 
 /**
  * i40iw_hw_stats_timeout - Stats timer-handler which updates all HW stats
- * @dev: hardware control device structure
+ * @vsi: pointer to the vsi structure
  */
-static void i40iw_hw_stats_timeout(unsigned long dev)
+static void i40iw_hw_stats_timeout(unsigned long vsi)
 {
-	struct i40iw_sc_dev *pf_dev = (struct i40iw_sc_dev *)dev;
-	struct i40iw_dev_pestat *pf_devstat = &pf_dev->dev_pestat;
-	struct i40iw_dev_pestat *vf_devstat = NULL;
+	struct i40iw_sc_vsi *sc_vsi =  (struct i40iw_sc_vsi *)vsi;
+	struct i40iw_sc_dev *pf_dev = sc_vsi->dev;
+	struct i40iw_vsi_pestat *pf_devstat = sc_vsi->pestat;
+	struct i40iw_vsi_pestat *vf_devstat = NULL;
 	u16 iw_vf_idx;
 	unsigned long flags;
 
 	/*PF*/
-	pf_devstat->ops.iw_hw_stat_read_all(pf_devstat, &pf_devstat->hw_stats);
+	i40iw_hw_stats_read_all(pf_devstat, &pf_devstat->hw_stats);
+
 	for (iw_vf_idx = 0; iw_vf_idx < I40IW_MAX_PE_ENABLED_VF_COUNT; iw_vf_idx++) {
-		spin_lock_irqsave(&pf_devstat->stats_lock, flags);
+		spin_lock_irqsave(&pf_devstat->lock, flags);
 		if (pf_dev->vf_dev[iw_vf_idx]) {
 			if (pf_dev->vf_dev[iw_vf_idx]->stats_initialized) {
-				vf_devstat = &pf_dev->vf_dev[iw_vf_idx]->dev_pestat;
-				vf_devstat->ops.iw_hw_stat_read_all(vf_devstat, &vf_devstat->hw_stats);
+				vf_devstat = &pf_dev->vf_dev[iw_vf_idx]->pestat;
+				i40iw_hw_stats_read_all(vf_devstat, &vf_devstat->hw_stats);
 			}
 		}
-		spin_unlock_irqrestore(&pf_devstat->stats_lock, flags);
+		spin_unlock_irqrestore(&pf_devstat->lock, flags);
 	}
 
 	mod_timer(&pf_devstat->stats_timer,
@@ -1310,26 +1422,26 @@ static void i40iw_hw_stats_timeout(unsigned long dev)
 
 /**
  * i40iw_hw_stats_start_timer - Start periodic stats timer
- * @dev: hardware control device structure
+ * @vsi: pointer to the vsi structure
  */
-void i40iw_hw_stats_start_timer(struct i40iw_sc_dev *dev)
+void i40iw_hw_stats_start_timer(struct i40iw_sc_vsi *vsi)
 {
-	struct i40iw_dev_pestat *devstat = &dev->dev_pestat;
+	struct i40iw_vsi_pestat *devstat = vsi->pestat;
 
 	init_timer(&devstat->stats_timer);
 	devstat->stats_timer.function = i40iw_hw_stats_timeout;
-	devstat->stats_timer.data = (unsigned long)dev;
+	devstat->stats_timer.data = (unsigned long)vsi;
 	mod_timer(&devstat->stats_timer,
 		  jiffies + msecs_to_jiffies(STATS_TIMER_DELAY));
 }
 
 /**
- * i40iw_hw_stats_del_timer - Delete periodic stats timer
- * @dev: hardware control device structure
+ * i40iw_hw_stats_stop_timer - Delete periodic stats timer
+ * @vsi: pointer to the vsi structure
  */
-void i40iw_hw_stats_del_timer(struct i40iw_sc_dev *dev)
+void i40iw_hw_stats_stop_timer(struct i40iw_sc_vsi *vsi)
 {
-	struct i40iw_dev_pestat *devstat = &dev->dev_pestat;
+	struct i40iw_vsi_pestat *devstat = vsi->pestat;
 
 	del_timer_sync(&devstat->stats_timer);
 }
diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
index 18526e6f9d85..855e49957a1a 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
@@ -623,6 +623,7 @@ static struct ib_qp *i40iw_create_qp(struct ib_pd *ibpd,
 	sq_size = init_attr->cap.max_send_wr;
 	rq_size = init_attr->cap.max_recv_wr;
 
+	init_info.vsi = &iwdev->vsi;
 	init_info.qp_uk_init_info.sq_size = sq_size;
 	init_info.qp_uk_init_info.rq_size = rq_size;
 	init_info.qp_uk_init_info.max_sq_frag_cnt = init_attr->cap.max_send_sge;
@@ -1052,11 +1053,11 @@ static void cq_free_resources(struct i40iw_device *iwdev, struct i40iw_cq *iwcq)
 }
 
 /**
- * cq_wq_destroy - send cq destroy cqp
+ * i40iw_cq_wq_destroy - send cq destroy cqp
  * @iwdev: iwarp device
  * @cq: hardware control cq
  */
-static void cq_wq_destroy(struct i40iw_device *iwdev, struct i40iw_sc_cq *cq)
+void i40iw_cq_wq_destroy(struct i40iw_device *iwdev, struct i40iw_sc_cq *cq)
 {
 	enum i40iw_status_code status;
 	struct i40iw_cqp_request *cqp_request;
@@ -1095,7 +1096,7 @@ static int i40iw_destroy_cq(struct ib_cq *ib_cq)
 	iwcq = to_iwcq(ib_cq);
 	iwdev = to_iwdev(ib_cq->device);
 	cq = &iwcq->sc_cq;
-	cq_wq_destroy(iwdev, cq);
+	i40iw_cq_wq_destroy(iwdev, cq);
 	cq_free_resources(iwdev, iwcq);
 	kfree(iwcq);
 	i40iw_rem_devusecount(iwdev);
@@ -1253,7 +1254,7 @@ static struct ib_cq *i40iw_create_cq(struct ib_device *ibdev,
 	return (struct ib_cq *)iwcq;
 
 cq_destroy:
-	cq_wq_destroy(iwdev, cq);
+	i40iw_cq_wq_destroy(iwdev, cq);
 cq_free_resources:
 	cq_free_resources(iwdev, iwcq);
 error:
@@ -2632,15 +2633,11 @@ static int i40iw_get_hw_stats(struct ib_device *ibdev,
 {
 	struct i40iw_device *iwdev = to_iwdev(ibdev);
 	struct i40iw_sc_dev *dev = &iwdev->sc_dev;
-	struct i40iw_dev_pestat *devstat = &dev->dev_pestat;
+	struct i40iw_vsi_pestat *devstat = iwdev->vsi.pestat;
 	struct i40iw_dev_hw_stats *hw_stats = &devstat->hw_stats;
-	unsigned long flags;
 
 	if (dev->is_pf) {
-		spin_lock_irqsave(&devstat->stats_lock, flags);
-		devstat->ops.iw_hw_stat_read_all(devstat,
-			&devstat->hw_stats);
-		spin_unlock_irqrestore(&devstat->stats_lock, flags);
+		i40iw_hw_stats_read_all(devstat, &devstat->hw_stats);
 	} else {
 		if (i40iw_vchnl_vf_get_pe_stats(dev, &devstat->hw_stats))
 			return -ENOSYS;
diff --git a/drivers/infiniband/hw/i40iw/i40iw_virtchnl.c b/drivers/infiniband/hw/i40iw/i40iw_virtchnl.c
index dbd39c4af946..f4d13683a403 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_virtchnl.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_virtchnl.c
@@ -402,6 +402,19 @@ del_out:
 	vchnl_pf_send_error_resp(vf_dev->pf_dev, vf_dev->vf_id, vchnl_msg, (u16)ret_code);
 }
 
+/**
+ * i40iw_vf_init_pestat - Initialize stats for VF
+ * @devL pointer to the VF Device
+ * @stats: Statistics structure pointer
+ * @index: Stats index
+ */
+static void i40iw_vf_init_pestat(struct i40iw_sc_dev *dev, struct i40iw_vsi_pestat *stats, u16 index)
+{
+	stats->hw = dev->hw;
+	i40iw_hw_stats_init(stats, (u8)index, false);
+	spin_lock_init(&stats->lock);
+}
+
 /**
  * i40iw_vchnl_recv_pf - Receive PF virtual channel messages
  * @dev: IWARP device pointer
@@ -421,9 +434,8 @@ enum i40iw_status_code i40iw_vchnl_recv_pf(struct i40iw_sc_dev *dev,
 	u16 first_avail_iw_vf = I40IW_MAX_PE_ENABLED_VF_COUNT;
 	struct i40iw_virt_mem vf_dev_mem;
 	struct i40iw_virtchnl_work_info work_info;
-	struct i40iw_dev_pestat *devstat;
+	struct i40iw_vsi_pestat *stats;
 	enum i40iw_status_code ret_code;
-	unsigned long flags;
 
 	if (!dev || !msg || !len)
 		return I40IW_ERR_PARAM;
@@ -496,10 +508,7 @@ enum i40iw_status_code i40iw_vchnl_recv_pf(struct i40iw_sc_dev *dev,
 				i40iw_debug(dev, I40IW_DEBUG_VIRT,
 					    "VF%u error CQP HMC Function operation.\n",
 					    vf_id);
-			i40iw_device_init_pestat(&vf_dev->dev_pestat);
-			vf_dev->dev_pestat.ops.iw_hw_stat_init(&vf_dev->dev_pestat,
-							      (u8)vf_dev->pmf_index,
-							      dev->hw, false);
+			i40iw_vf_init_pestat(dev, &vf_dev->pestat, vf_dev->pmf_index);
 			vf_dev->stats_initialized = true;
 		} else {
 			if (vf_dev) {
@@ -530,12 +539,10 @@ enum i40iw_status_code i40iw_vchnl_recv_pf(struct i40iw_sc_dev *dev,
 	case I40IW_VCHNL_OP_GET_STATS:
 		if (!vf_dev)
 			return I40IW_ERR_BAD_PTR;
-		devstat = &vf_dev->dev_pestat;
-		spin_lock_irqsave(&dev->dev_pestat.stats_lock, flags);
-		devstat->ops.iw_hw_stat_read_all(devstat, &devstat->hw_stats);
-		spin_unlock_irqrestore(&dev->dev_pestat.stats_lock, flags);
+		stats = &vf_dev->pestat;
+		i40iw_hw_stats_read_all(stats, &stats->hw_stats);
 		vf_dev->msg_count--;
-		vchnl_pf_send_get_pe_stats_resp(dev, vf_id, vchnl_msg, &devstat->hw_stats);
+		vchnl_pf_send_get_pe_stats_resp(dev, vf_id, vchnl_msg, &stats->hw_stats);
 		break;
 	default:
 		i40iw_debug(dev, I40IW_DEBUG_VIRT,
-- 
cgit v1.2.1


From 1f22e454df2eb99ba6b7ace3f594f6805cdf5cbc Mon Sep 17 00:00:00 2001
From: Eran Ben Elisha <eranbe@mellanox.com>
Date: Thu, 10 Nov 2016 11:31:00 +0200
Subject: IB/mlx4: When no DMFS for IPoIB, don't allow NET_IF QPs

According to the firmware spec, FLOW_STEERING_IB_UC_QP_RANGE command is
supported only if dmfs_ipoib bit is set.

If it isn't set we want to ensure allocating NET_IF QPs fail. We do so
by filling out the allocation bitmap. By thus, the NET_IF QPs allocating
function won't find any free QP and will fail.

Fixes: c1c98501121e ('IB/mlx4: Add support for steerable IB UD QPs')
Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/main.c | 21 +++++++++++++--------
 1 file changed, 13 insertions(+), 8 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index 4054a1bfabb5..4a67ffc1ddfb 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -2823,14 +2823,19 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
 			goto err_steer_qp_release;
 		}
 
-		bitmap_zero(ibdev->ib_uc_qpns_bitmap, ibdev->steer_qpn_count);
-
-		err = mlx4_FLOW_STEERING_IB_UC_QP_RANGE(
-				dev, ibdev->steer_qpn_base,
-				ibdev->steer_qpn_base +
-				ibdev->steer_qpn_count - 1);
-		if (err)
-			goto err_steer_free_bitmap;
+		if (dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_DMFS_IPOIB) {
+			bitmap_zero(ibdev->ib_uc_qpns_bitmap,
+				    ibdev->steer_qpn_count);
+			err = mlx4_FLOW_STEERING_IB_UC_QP_RANGE(
+					dev, ibdev->steer_qpn_base,
+					ibdev->steer_qpn_base +
+					ibdev->steer_qpn_count - 1);
+			if (err)
+				goto err_steer_free_bitmap;
+		} else {
+			bitmap_fill(ibdev->ib_uc_qpns_bitmap,
+				    ibdev->steer_qpn_count);
+		}
 	}
 
 	for (j = 1; j <= ibdev->dev->caps.num_ports; j++)
-- 
cgit v1.2.1


From bf08e884bfd5be068fd2ccf2bc450f085d8dd853 Mon Sep 17 00:00:00 2001
From: Eran Ben Elisha <eranbe@mellanox.com>
Date: Thu, 10 Nov 2016 11:31:01 +0200
Subject: IB/mlx4: Check if GRH is available before using it

Before reading GRH attributes, need to make sure AH contains GRH,
and in addition, initialize GID type.

Fixes: dbf727de7440 ('IB/core: Use GID table in AH creation and dmac resolution')
Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/qp.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index 570bc866b1d6..0914731c9aac 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -1764,14 +1764,14 @@ static int __mlx4_ib_modify_qp(struct ib_qp *ibqp,
 		u8 port_num = mlx4_is_bonded(to_mdev(ibqp->device)->dev) ? 1 :
 			attr_mask & IB_QP_PORT ? attr->port_num : qp->port;
 		union ib_gid gid;
-		struct ib_gid_attr gid_attr;
+		struct ib_gid_attr gid_attr = {.gid_type = IB_GID_TYPE_IB};
 		u16 vlan = 0xffff;
 		u8 smac[ETH_ALEN];
 		int status = 0;
 		int is_eth = rdma_cap_eth_ah(&dev->ib_dev, port_num) &&
 			attr->ah_attr.ah_flags & IB_AH_GRH;
 
-		if (is_eth) {
+		if (is_eth && attr->ah_attr.ah_flags & IB_AH_GRH) {
 			int index = attr->ah_attr.grh.sgid_index;
 
 			status = ib_get_cached_gid(ibqp->device, port_num,
-- 
cgit v1.2.1


From d680ebed91e0b45c43ae03a880a0b43211096161 Mon Sep 17 00:00:00 2001
From: Yonatan Cohen <yonatanc@mellanox.com>
Date: Wed, 16 Nov 2016 10:39:16 +0200
Subject: IB/rxe: Increase max number of completions to 32k

Increase limit of max CQE from 8K to 32K to allow demanding
applications to work over SoftRoCE with same configuration
as most RoCEv2 HW vendors have.

Fixes: 8700e3e7c485 ("Soft RoCE driver")
Signed-off-by: Yonatan Cohen <yonatanc@mellanox.com>
Reviewed-by: Moni Shoua <monis@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rxe/rxe_param.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rxe/rxe_param.h b/drivers/infiniband/sw/rxe/rxe_param.h
index f459c43a77c8..13ed2cc6eaa2 100644
--- a/drivers/infiniband/sw/rxe/rxe_param.h
+++ b/drivers/infiniband/sw/rxe/rxe_param.h
@@ -82,7 +82,7 @@ enum rxe_device_param {
 	RXE_MAX_SGE			= 32,
 	RXE_MAX_SGE_RD			= 32,
 	RXE_MAX_CQ			= 16384,
-	RXE_MAX_LOG_CQE			= 13,
+	RXE_MAX_LOG_CQE			= 15,
 	RXE_MAX_MR			= 2 * 1024,
 	RXE_MAX_PD			= 0x7ffc,
 	RXE_MAX_QP_RD_ATOM		= 128,
-- 
cgit v1.2.1


From 191ded4a4d991acf17207e0b4370fef070bce3e9 Mon Sep 17 00:00:00 2001
From: Bodong Wang <bodong@mellanox.com>
Date: Mon, 31 Oct 2016 12:15:21 +0200
Subject: IB/mlx5: Report mlx5 multi packet WQE caps during query

The capabilities whether hardware support multi packet WQE or not is
exposed to user space through query_device by uhw.

Signed-off-by: Bodong Wang <bodong@mellanox.com>
Reviewed-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/main.c | 11 +++++++++++
 1 file changed, 11 insertions(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index 46facd30e672..a16207c33333 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -672,6 +672,17 @@ static int mlx5_ib_query_device(struct ib_device *ibdev,
 			1 << MLX5_CAP_GEN(dev->mdev, log_max_rq);
 	}
 
+	if (field_avail(typeof(resp), mlx5_ib_support_multi_pkt_send_wqes,
+			uhw->outlen)) {
+		resp.mlx5_ib_support_multi_pkt_send_wqes =
+			MLX5_CAP_ETH(mdev, multi_pkt_send_wqe);
+		resp.response_length +=
+			sizeof(resp.mlx5_ib_support_multi_pkt_send_wqes);
+	}
+
+	if (field_avail(typeof(resp), reserved, uhw->outlen))
+		resp.response_length += sizeof(resp.reserved);
+
 	if (uhw->outlen) {
 		err = ib_copy_to_udata(uhw, &resp, resp.response_length);
 
-- 
cgit v1.2.1


From 7e43a2a5bae39fedaa7cce21d637e0c8d96d8e54 Mon Sep 17 00:00:00 2001
From: Bodong Wang <bodong@mellanox.com>
Date: Mon, 31 Oct 2016 12:16:44 +0200
Subject: IB/mlx5: Report mlx5 CQE compression caps during query

The capabilities include:
- Max number of compressed and aggregated CQEs in a single session,
  while zero means unsupported.
- For Responder, there are two formats of mini CQE: mini CQE with Rx
  hash and mini CQE with checksum. They're mutual exclusive.

Signed-off-by: Bodong Wang <bodong@mellanox.com>
Reviewed-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/main.c | 10 ++++++++++
 1 file changed, 10 insertions(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index a16207c33333..2687a93c4af2 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -683,6 +683,16 @@ static int mlx5_ib_query_device(struct ib_device *ibdev,
 	if (field_avail(typeof(resp), reserved, uhw->outlen))
 		resp.response_length += sizeof(resp.reserved);
 
+	if (field_avail(typeof(resp), cqe_comp_caps, uhw->outlen)) {
+		resp.cqe_comp_caps.max_num =
+			MLX5_CAP_GEN(dev->mdev, cqe_compression) ?
+			MLX5_CAP_GEN(dev->mdev, cqe_compression_max_num) : 0;
+		resp.cqe_comp_caps.supported_format =
+			MLX5_IB_CQE_RES_FORMAT_HASH |
+			MLX5_IB_CQE_RES_FORMAT_CSUM;
+		resp.response_length += sizeof(resp.cqe_comp_caps);
+	}
+
 	if (uhw->outlen) {
 		err = ib_copy_to_udata(uhw, &resp, resp.response_length);
 
-- 
cgit v1.2.1


From 1cbe6fc86ccfe05a910be4883da7c7bd28c190fe Mon Sep 17 00:00:00 2001
From: Bodong Wang <bodong@mellanox.com>
Date: Mon, 31 Oct 2016 12:16:45 +0200
Subject: IB/mlx5: Add support for CQE compressing

CQE compressing reduces PCI overhead by coalescing and compressing
multiple CQEs into a single merged CQE. Successful compressing
improves message rate especially for small packet traffic.

CQE compressing is supported for all 64B CQE formats (with certain
limitations) generated by RQ/Responder or by SQ/Requestor.

Signed-off-by: Bodong Wang <bodong@mellanox.com>
Reviewed-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/cq.c | 30 +++++++++++++++++++++++++++++-
 1 file changed, 29 insertions(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/cq.c b/drivers/infiniband/hw/mlx5/cq.c
index 9e0598b5615f..d72a4367c891 100644
--- a/drivers/infiniband/hw/mlx5/cq.c
+++ b/drivers/infiniband/hw/mlx5/cq.c
@@ -731,7 +731,7 @@ static int create_cq_user(struct mlx5_ib_dev *dev, struct ib_udata *udata,
 			  int entries, u32 **cqb,
 			  int *cqe_size, int *index, int *inlen)
 {
-	struct mlx5_ib_create_cq ucmd;
+	struct mlx5_ib_create_cq ucmd = {};
 	size_t ucmdlen;
 	int page_shift;
 	__be64 *pas;
@@ -792,8 +792,36 @@ static int create_cq_user(struct mlx5_ib_dev *dev, struct ib_udata *udata,
 
 	*index = to_mucontext(context)->uuari.uars[0].index;
 
+	if (ucmd.cqe_comp_en == 1) {
+		if (unlikely((*cqe_size != 64) ||
+			     !MLX5_CAP_GEN(dev->mdev, cqe_compression))) {
+			err = -EOPNOTSUPP;
+			mlx5_ib_warn(dev, "CQE compression is not supported for size %d!\n",
+				     *cqe_size);
+			goto err_cqb;
+		}
+
+		if (unlikely(!ucmd.cqe_comp_res_format ||
+			     !(ucmd.cqe_comp_res_format <
+			       MLX5_IB_CQE_RES_RESERVED) ||
+			     (ucmd.cqe_comp_res_format &
+			      (ucmd.cqe_comp_res_format - 1)))) {
+			err = -EOPNOTSUPP;
+			mlx5_ib_warn(dev, "CQE compression res format %d is not supported!\n",
+				     ucmd.cqe_comp_res_format);
+			goto err_cqb;
+		}
+
+		MLX5_SET(cqc, cqc, cqe_comp_en, 1);
+		MLX5_SET(cqc, cqc, mini_cqe_res_format,
+			 ilog2(ucmd.cqe_comp_res_format));
+	}
+
 	return 0;
 
+err_cqb:
+	kfree(cqb);
+
 err_db:
 	mlx5_ib_db_unmap_user(to_mucontext(context), &cq->db);
 
-- 
cgit v1.2.1


From 0dbf3332b7b683db33a385a3ce9baab157e3ff9a Mon Sep 17 00:00:00 2001
From: Moses Reuben <mosesr@mellanox.com>
Date: Mon, 14 Nov 2016 19:04:47 +0200
Subject: IB/core: Add flow spec tunneling support

In order to support tunneling, that can be used by the QP,
both struct ib_flow_spec_tunnel and struct ib_flow_tunnel_filter can be
used to more IP or UDP based tunneling protocols (e.g NVGRE, GRE, etc).

IB_FLOW_SPEC_VXLAN_TUNNEL type flow specification is added to use this
functionality and match specific Vxlan packets.

In similar to IPv6, we check overflow of the vni value by
comparing with the maximum size.

Signed-off-by: Moses Reuben <mosesr@mellanox.com>
Reviewed-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/uverbs_cmd.c | 15 +++++++++++++++
 1 file changed, 15 insertions(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c
index 2690c9263ee4..deb06fa3548b 100644
--- a/drivers/infiniband/core/uverbs_cmd.c
+++ b/drivers/infiniband/core/uverbs_cmd.c
@@ -3175,6 +3175,21 @@ static int kern_spec_to_ib_spec(struct ib_uverbs_flow_spec *kern_spec,
 		memcpy(&ib_spec->tcp_udp.val, kern_spec_val, actual_filter_sz);
 		memcpy(&ib_spec->tcp_udp.mask, kern_spec_mask, actual_filter_sz);
 		break;
+	case IB_FLOW_SPEC_VXLAN_TUNNEL:
+		ib_filter_sz = offsetof(struct ib_flow_tunnel_filter, real_sz);
+		actual_filter_sz = spec_filter_size(kern_spec_mask,
+						    kern_filter_sz,
+						    ib_filter_sz);
+		if (actual_filter_sz <= 0)
+			return -EINVAL;
+		ib_spec->tunnel.size = sizeof(struct ib_flow_spec_tunnel);
+		memcpy(&ib_spec->tunnel.val, kern_spec_val, actual_filter_sz);
+		memcpy(&ib_spec->tunnel.mask, kern_spec_mask, actual_filter_sz);
+
+		if ((ntohl(ib_spec->tunnel.mask.tunnel_id)) >= BIT(24) ||
+		    (ntohl(ib_spec->tunnel.val.tunnel_id)) >= BIT(24))
+			return -EINVAL;
+		break;
 	default:
 		return -EINVAL;
 	}
-- 
cgit v1.2.1


From ffb30d8f107b27820a069ae6772ab48e58cc0b2f Mon Sep 17 00:00:00 2001
From: Moses Reuben <mosesr@mellanox.com>
Date: Mon, 14 Nov 2016 19:04:50 +0200
Subject: IB/mlx5: Support Vxlan tunneling specification

Add support to receive specific Vxlan packet in ConnectX-4.

Signed-off-by: Moses Reuben <mosesr@mellanox.com>
Reviewed-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/main.c | 11 +++++++++++
 1 file changed, 11 insertions(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index 2687a93c4af2..a9e5851365d3 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -1539,6 +1539,7 @@ static void set_tos(void *outer_c, void *outer_v, u8 mask, u8 val)
 #define LAST_IPV4_FIELD tos
 #define LAST_IPV6_FIELD traffic_class
 #define LAST_TCP_UDP_FIELD src_port
+#define LAST_TUNNEL_FIELD tunnel_id
 
 /* Field is the last supported field */
 #define FIELDS_NOT_SUPPORTED(filter, field)\
@@ -1722,6 +1723,16 @@ static int parse_flow_attr(u32 *match_c, u32 *match_v,
 		MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, udp_dport,
 			 ntohs(ib_spec->tcp_udp.val.dst_port));
 		break;
+	case IB_FLOW_SPEC_VXLAN_TUNNEL:
+		if (FIELDS_NOT_SUPPORTED(ib_spec->tunnel.mask,
+					 LAST_TUNNEL_FIELD))
+			return -ENOTSUPP;
+
+		MLX5_SET(fte_match_set_misc, misc_params_c, vxlan_vni,
+			 ntohl(ib_spec->tunnel.mask.tunnel_id));
+		MLX5_SET(fte_match_set_misc, misc_params_v, vxlan_vni,
+			 ntohl(ib_spec->tunnel.val.tunnel_id));
+		break;
 	default:
 		return -EINVAL;
 	}
-- 
cgit v1.2.1


From fbf46860b19ddb485f00bef1ad1a43aabc9f71ad Mon Sep 17 00:00:00 2001
From: Moses Reuben <mosesr@mellanox.com>
Date: Mon, 14 Nov 2016 19:04:51 +0200
Subject: IB/core: Introduce inner flow steering

For a tunneled packet which contains external and internal headers,
we refer to the external headers as "outer fields" and the internal
headers as "inner fields".

Example of a tunneled packet:

{ L2 | L3 | L4 | tunnel header | L2 | L3 | l4 | data }
  |     |    |         |         |    |    |
{       outer fields           }{ inner fields }

This patch introduces a new flag for flow steering rules
- IB_FLOW_SPEC_INNER - which specifies that the rule applies
to the inner fields, rather than to the outer fields of the packet.

Signed-off-by: Moses Reuben <mosesr@mellanox.com>
Reviewed-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/uverbs_cmd.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c
index deb06fa3548b..d84dcde7a8bb 100644
--- a/drivers/infiniband/core/uverbs_cmd.c
+++ b/drivers/infiniband/core/uverbs_cmd.c
@@ -3124,8 +3124,10 @@ static int kern_spec_to_ib_spec(struct ib_uverbs_flow_spec *kern_spec,
 	kern_spec_val = (void *)kern_spec +
 		sizeof(struct ib_uverbs_flow_spec_hdr);
 	kern_spec_mask = kern_spec_val + kern_filter_sz;
+	if (ib_spec->type == (IB_FLOW_SPEC_INNER | IB_FLOW_SPEC_VXLAN_TUNNEL))
+		return -EINVAL;
 
-	switch (ib_spec->type) {
+	switch (ib_spec->type & ~IB_FLOW_SPEC_INNER) {
 	case IB_FLOW_SPEC_ETH:
 		ib_filter_sz = offsetof(struct ib_flow_eth_filter, real_sz);
 		actual_filter_sz = spec_filter_size(kern_spec_mask,
-- 
cgit v1.2.1


From 2d1e697e9b716b8a692bc9c197e5f4ffd10d7307 Mon Sep 17 00:00:00 2001
From: Moses Reuben <mosesr@mellanox.com>
Date: Mon, 14 Nov 2016 19:04:52 +0200
Subject: IB/mlx5: Add support to match inner packet fields

Add support to match packet fields which are tunneled,
i.e. support matching the header of the inner packet which is the result of
or bit operation of the original header and the IB_FLOW_SPEC_INNER type.

The combination of IB_FLOW_SPEC_INNER | IB_FLOW_SPEC_VXLAN_TUNNEL is not
needed to be checked, because the IB core has this check already.

Signed-off-by: Moses Reuben <mosesr@mellanox.com>
Reviewed-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/main.c | 131 +++++++++++++++++++++++---------------
 1 file changed, 78 insertions(+), 53 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index a9e5851365d3..f7a299431e6c 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -1526,6 +1526,22 @@ static void set_proto(void *outer_c, void *outer_v, u8 mask, u8 val)
 	MLX5_SET(fte_match_set_lyr_2_4, outer_v, ip_protocol, val);
 }
 
+static void set_flow_label(void *misc_c, void *misc_v, u8 mask, u8 val,
+			   bool inner)
+{
+	if (inner) {
+		MLX5_SET(fte_match_set_misc,
+			 misc_c, inner_ipv6_flow_label, mask);
+		MLX5_SET(fte_match_set_misc,
+			 misc_v, inner_ipv6_flow_label, val);
+	} else {
+		MLX5_SET(fte_match_set_misc,
+			 misc_c, outer_ipv6_flow_label, mask);
+		MLX5_SET(fte_match_set_misc,
+			 misc_v, outer_ipv6_flow_label, val);
+	}
+}
+
 static void set_tos(void *outer_c, void *outer_v, u8 mask, u8 val)
 {
 	MLX5_SET(fte_match_set_lyr_2_4, outer_c, ip_ecn, mask);
@@ -1552,155 +1568,164 @@ static void set_tos(void *outer_c, void *outer_v, u8 mask, u8 val)
 static int parse_flow_attr(u32 *match_c, u32 *match_v,
 			   const union ib_flow_spec *ib_spec)
 {
-	void *outer_headers_c = MLX5_ADDR_OF(fte_match_param, match_c,
-					     outer_headers);
-	void *outer_headers_v = MLX5_ADDR_OF(fte_match_param, match_v,
-					     outer_headers);
 	void *misc_params_c = MLX5_ADDR_OF(fte_match_param, match_c,
 					   misc_parameters);
 	void *misc_params_v = MLX5_ADDR_OF(fte_match_param, match_v,
 					   misc_parameters);
+	void *headers_c;
+	void *headers_v;
+
+	if (ib_spec->type & IB_FLOW_SPEC_INNER) {
+		headers_c = MLX5_ADDR_OF(fte_match_param, match_c,
+					 inner_headers);
+		headers_v = MLX5_ADDR_OF(fte_match_param, match_v,
+					 inner_headers);
+	} else {
+		headers_c = MLX5_ADDR_OF(fte_match_param, match_c,
+					 outer_headers);
+		headers_v = MLX5_ADDR_OF(fte_match_param, match_v,
+					 outer_headers);
+	}
 
-	switch (ib_spec->type) {
+	switch (ib_spec->type & ~IB_FLOW_SPEC_INNER) {
 	case IB_FLOW_SPEC_ETH:
 		if (FIELDS_NOT_SUPPORTED(ib_spec->eth.mask, LAST_ETH_FIELD))
 			return -ENOTSUPP;
 
-		ether_addr_copy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_c,
+		ether_addr_copy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, headers_c,
 					     dmac_47_16),
 				ib_spec->eth.mask.dst_mac);
-		ether_addr_copy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_v,
+		ether_addr_copy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, headers_v,
 					     dmac_47_16),
 				ib_spec->eth.val.dst_mac);
 
-		ether_addr_copy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_c,
+		ether_addr_copy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, headers_c,
 					     smac_47_16),
 				ib_spec->eth.mask.src_mac);
-		ether_addr_copy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_v,
+		ether_addr_copy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, headers_v,
 					     smac_47_16),
 				ib_spec->eth.val.src_mac);
 
 		if (ib_spec->eth.mask.vlan_tag) {
-			MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c,
+			MLX5_SET(fte_match_set_lyr_2_4, headers_c,
 				 vlan_tag, 1);
-			MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v,
+			MLX5_SET(fte_match_set_lyr_2_4, headers_v,
 				 vlan_tag, 1);
 
-			MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c,
+			MLX5_SET(fte_match_set_lyr_2_4, headers_c,
 				 first_vid, ntohs(ib_spec->eth.mask.vlan_tag));
-			MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v,
+			MLX5_SET(fte_match_set_lyr_2_4, headers_v,
 				 first_vid, ntohs(ib_spec->eth.val.vlan_tag));
 
-			MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c,
+			MLX5_SET(fte_match_set_lyr_2_4, headers_c,
 				 first_cfi,
 				 ntohs(ib_spec->eth.mask.vlan_tag) >> 12);
-			MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v,
+			MLX5_SET(fte_match_set_lyr_2_4, headers_v,
 				 first_cfi,
 				 ntohs(ib_spec->eth.val.vlan_tag) >> 12);
 
-			MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c,
+			MLX5_SET(fte_match_set_lyr_2_4, headers_c,
 				 first_prio,
 				 ntohs(ib_spec->eth.mask.vlan_tag) >> 13);
-			MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v,
+			MLX5_SET(fte_match_set_lyr_2_4, headers_v,
 				 first_prio,
 				 ntohs(ib_spec->eth.val.vlan_tag) >> 13);
 		}
-		MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c,
+		MLX5_SET(fte_match_set_lyr_2_4, headers_c,
 			 ethertype, ntohs(ib_spec->eth.mask.ether_type));
-		MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v,
+		MLX5_SET(fte_match_set_lyr_2_4, headers_v,
 			 ethertype, ntohs(ib_spec->eth.val.ether_type));
 		break;
 	case IB_FLOW_SPEC_IPV4:
 		if (FIELDS_NOT_SUPPORTED(ib_spec->ipv4.mask, LAST_IPV4_FIELD))
 			return -ENOTSUPP;
 
-		MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c,
+		MLX5_SET(fte_match_set_lyr_2_4, headers_c,
 			 ethertype, 0xffff);
-		MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v,
+		MLX5_SET(fte_match_set_lyr_2_4, headers_v,
 			 ethertype, ETH_P_IP);
 
-		memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_c,
+		memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, headers_c,
 				    src_ipv4_src_ipv6.ipv4_layout.ipv4),
 		       &ib_spec->ipv4.mask.src_ip,
 		       sizeof(ib_spec->ipv4.mask.src_ip));
-		memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_v,
+		memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, headers_v,
 				    src_ipv4_src_ipv6.ipv4_layout.ipv4),
 		       &ib_spec->ipv4.val.src_ip,
 		       sizeof(ib_spec->ipv4.val.src_ip));
-		memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_c,
+		memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, headers_c,
 				    dst_ipv4_dst_ipv6.ipv4_layout.ipv4),
 		       &ib_spec->ipv4.mask.dst_ip,
 		       sizeof(ib_spec->ipv4.mask.dst_ip));
-		memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_v,
+		memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, headers_v,
 				    dst_ipv4_dst_ipv6.ipv4_layout.ipv4),
 		       &ib_spec->ipv4.val.dst_ip,
 		       sizeof(ib_spec->ipv4.val.dst_ip));
 
-		set_tos(outer_headers_c, outer_headers_v,
+		set_tos(headers_c, headers_v,
 			ib_spec->ipv4.mask.tos, ib_spec->ipv4.val.tos);
 
-		set_proto(outer_headers_c, outer_headers_v,
+		set_proto(headers_c, headers_v,
 			  ib_spec->ipv4.mask.proto, ib_spec->ipv4.val.proto);
 		break;
 	case IB_FLOW_SPEC_IPV6:
 		if (FIELDS_NOT_SUPPORTED(ib_spec->ipv6.mask, LAST_IPV6_FIELD))
 			return -ENOTSUPP;
 
-		MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c,
+		MLX5_SET(fte_match_set_lyr_2_4, headers_c,
 			 ethertype, 0xffff);
-		MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v,
+		MLX5_SET(fte_match_set_lyr_2_4, headers_v,
 			 ethertype, ETH_P_IPV6);
 
-		memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_c,
+		memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, headers_c,
 				    src_ipv4_src_ipv6.ipv6_layout.ipv6),
 		       &ib_spec->ipv6.mask.src_ip,
 		       sizeof(ib_spec->ipv6.mask.src_ip));
-		memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_v,
+		memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, headers_v,
 				    src_ipv4_src_ipv6.ipv6_layout.ipv6),
 		       &ib_spec->ipv6.val.src_ip,
 		       sizeof(ib_spec->ipv6.val.src_ip));
-		memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_c,
+		memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, headers_c,
 				    dst_ipv4_dst_ipv6.ipv6_layout.ipv6),
 		       &ib_spec->ipv6.mask.dst_ip,
 		       sizeof(ib_spec->ipv6.mask.dst_ip));
-		memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_v,
+		memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, headers_v,
 				    dst_ipv4_dst_ipv6.ipv6_layout.ipv6),
 		       &ib_spec->ipv6.val.dst_ip,
 		       sizeof(ib_spec->ipv6.val.dst_ip));
 
-		set_tos(outer_headers_c, outer_headers_v,
+		set_tos(headers_c, headers_v,
 			ib_spec->ipv6.mask.traffic_class,
 			ib_spec->ipv6.val.traffic_class);
 
-		set_proto(outer_headers_c, outer_headers_v,
+		set_proto(headers_c, headers_v,
 			  ib_spec->ipv6.mask.next_hdr,
 			  ib_spec->ipv6.val.next_hdr);
 
-		MLX5_SET(fte_match_set_misc, misc_params_c,
-			 outer_ipv6_flow_label,
-			 ntohl(ib_spec->ipv6.mask.flow_label));
-		MLX5_SET(fte_match_set_misc, misc_params_v,
-			 outer_ipv6_flow_label,
-			 ntohl(ib_spec->ipv6.val.flow_label));
+		set_flow_label(misc_params_c, misc_params_v,
+			       ntohl(ib_spec->ipv6.mask.flow_label),
+			       ntohl(ib_spec->ipv6.val.flow_label),
+			       ib_spec->type & IB_FLOW_SPEC_INNER);
+
 		break;
 	case IB_FLOW_SPEC_TCP:
 		if (FIELDS_NOT_SUPPORTED(ib_spec->tcp_udp.mask,
 					 LAST_TCP_UDP_FIELD))
 			return -ENOTSUPP;
 
-		MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, ip_protocol,
+		MLX5_SET(fte_match_set_lyr_2_4, headers_c, ip_protocol,
 			 0xff);
-		MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, ip_protocol,
+		MLX5_SET(fte_match_set_lyr_2_4, headers_v, ip_protocol,
 			 IPPROTO_TCP);
 
-		MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, tcp_sport,
+		MLX5_SET(fte_match_set_lyr_2_4, headers_c, tcp_sport,
 			 ntohs(ib_spec->tcp_udp.mask.src_port));
-		MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, tcp_sport,
+		MLX5_SET(fte_match_set_lyr_2_4, headers_v, tcp_sport,
 			 ntohs(ib_spec->tcp_udp.val.src_port));
 
-		MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, tcp_dport,
+		MLX5_SET(fte_match_set_lyr_2_4, headers_c, tcp_dport,
 			 ntohs(ib_spec->tcp_udp.mask.dst_port));
-		MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, tcp_dport,
+		MLX5_SET(fte_match_set_lyr_2_4, headers_v, tcp_dport,
 			 ntohs(ib_spec->tcp_udp.val.dst_port));
 		break;
 	case IB_FLOW_SPEC_UDP:
@@ -1708,19 +1733,19 @@ static int parse_flow_attr(u32 *match_c, u32 *match_v,
 					 LAST_TCP_UDP_FIELD))
 			return -ENOTSUPP;
 
-		MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, ip_protocol,
+		MLX5_SET(fte_match_set_lyr_2_4, headers_c, ip_protocol,
 			 0xff);
-		MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, ip_protocol,
+		MLX5_SET(fte_match_set_lyr_2_4, headers_v, ip_protocol,
 			 IPPROTO_UDP);
 
-		MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, udp_sport,
+		MLX5_SET(fte_match_set_lyr_2_4, headers_c, udp_sport,
 			 ntohs(ib_spec->tcp_udp.mask.src_port));
-		MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, udp_sport,
+		MLX5_SET(fte_match_set_lyr_2_4, headers_v, udp_sport,
 			 ntohs(ib_spec->tcp_udp.val.src_port));
 
-		MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, udp_dport,
+		MLX5_SET(fte_match_set_lyr_2_4, headers_c, udp_dport,
 			 ntohs(ib_spec->tcp_udp.mask.dst_port));
-		MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, udp_dport,
+		MLX5_SET(fte_match_set_lyr_2_4, headers_v, udp_dport,
 			 ntohs(ib_spec->tcp_udp.val.dst_port));
 		break;
 	case IB_FLOW_SPEC_VXLAN_TUNNEL:
-- 
cgit v1.2.1


From c90ea9d8e51196d9c528e57d9ab09ee7d41f0ba0 Mon Sep 17 00:00:00 2001
From: Moni Shoua <monis@mellanox.com>
Date: Wed, 23 Nov 2016 08:23:22 +0200
Subject: IB/core: Change ib_resolve_eth_dmac to use it in create AH

The function ib_resolve_eth_dmac() requires struct qp_attr * and
qp_attr_mask as parameters while the function might be useful to resolve
dmac for address handles. This patch changes the signature of the
function so it can be used in the flow of creating an address handle.

Signed-off-by: Moni Shoua <monis@mellanox.com>
Reviewed-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/core_priv.h  |  3 --
 drivers/infiniband/core/uverbs_cmd.c |  8 ++--
 drivers/infiniband/core/verbs.c      | 84 ++++++++++++++++++------------------
 3 files changed, 47 insertions(+), 48 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/core/core_priv.h b/drivers/infiniband/core/core_priv.h
index 19d499dcab76..1acc95b3aaa3 100644
--- a/drivers/infiniband/core/core_priv.h
+++ b/drivers/infiniband/core/core_priv.h
@@ -72,9 +72,6 @@ void ib_device_unregister_sysfs(struct ib_device *device);
 void ib_cache_setup(void);
 void ib_cache_cleanup(void);
 
-int ib_resolve_eth_dmac(struct ib_qp *qp,
-			struct ib_qp_attr *qp_attr, int *qp_attr_mask);
-
 typedef void (*roce_netdev_callback)(struct ib_device *device, u8 port,
 	      struct net_device *idev, void *cookie);
 
diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c
index d84dcde7a8bb..77bb15f66414 100644
--- a/drivers/infiniband/core/uverbs_cmd.c
+++ b/drivers/infiniband/core/uverbs_cmd.c
@@ -2402,9 +2402,11 @@ ssize_t ib_uverbs_modify_qp(struct ib_uverbs_file *file,
 	attr->alt_ah_attr.port_num 	    = cmd.alt_dest.port_num;
 
 	if (qp->real_qp == qp) {
-		ret = ib_resolve_eth_dmac(qp, attr, &cmd.attr_mask);
-		if (ret)
-			goto release_qp;
+		if (cmd.attr_mask & IB_QP_AV) {
+			ret = ib_resolve_eth_dmac(qp->device, &attr->ah_attr);
+			if (ret)
+				goto release_qp;
+		}
 		ret = qp->device->modify_qp(qp, attr,
 			modify_qp_mask(qp->qp_type, cmd.attr_mask), &udata);
 	} else {
diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c
index 98e10c5f2a96..cacee169cd55 100644
--- a/drivers/infiniband/core/verbs.c
+++ b/drivers/infiniband/core/verbs.c
@@ -1198,66 +1198,66 @@ int ib_modify_qp_is_ok(enum ib_qp_state cur_state, enum ib_qp_state next_state,
 }
 EXPORT_SYMBOL(ib_modify_qp_is_ok);
 
-int ib_resolve_eth_dmac(struct ib_qp *qp,
-			struct ib_qp_attr *qp_attr, int *qp_attr_mask)
+int ib_resolve_eth_dmac(struct ib_device *device,
+			struct ib_ah_attr *ah_attr)
 {
 	int           ret = 0;
 
-	if (*qp_attr_mask & IB_QP_AV) {
-		if (qp_attr->ah_attr.port_num < rdma_start_port(qp->device) ||
-		    qp_attr->ah_attr.port_num > rdma_end_port(qp->device))
-			return -EINVAL;
-
-		if (!rdma_cap_eth_ah(qp->device, qp_attr->ah_attr.port_num))
-			return 0;
-
-		if (rdma_link_local_addr((struct in6_addr *)qp_attr->ah_attr.grh.dgid.raw)) {
-			rdma_get_ll_mac((struct in6_addr *)qp_attr->ah_attr.grh.dgid.raw,
-					qp_attr->ah_attr.dmac);
-		} else {
-			union ib_gid		sgid;
-			struct ib_gid_attr	sgid_attr;
-			int			ifindex;
-			int			hop_limit;
-
-			ret = ib_query_gid(qp->device,
-					   qp_attr->ah_attr.port_num,
-					   qp_attr->ah_attr.grh.sgid_index,
-					   &sgid, &sgid_attr);
-
-			if (ret || !sgid_attr.ndev) {
-				if (!ret)
-					ret = -ENXIO;
-				goto out;
-			}
+	if (ah_attr->port_num < rdma_start_port(device) ||
+	    ah_attr->port_num > rdma_end_port(device))
+		return -EINVAL;
 
-			ifindex = sgid_attr.ndev->ifindex;
+	if (!rdma_cap_eth_ah(device, ah_attr->port_num))
+		return 0;
 
-			ret = rdma_addr_find_l2_eth_by_grh(&sgid,
-							   &qp_attr->ah_attr.grh.dgid,
-							   qp_attr->ah_attr.dmac,
-							   NULL, &ifindex, &hop_limit);
+	if (rdma_link_local_addr((struct in6_addr *)ah_attr->grh.dgid.raw)) {
+		rdma_get_ll_mac((struct in6_addr *)ah_attr->grh.dgid.raw,
+				ah_attr->dmac);
+	} else {
+		union ib_gid		sgid;
+		struct ib_gid_attr	sgid_attr;
+		int			ifindex;
+		int			hop_limit;
+
+		ret = ib_query_gid(device,
+				   ah_attr->port_num,
+				   ah_attr->grh.sgid_index,
+				   &sgid, &sgid_attr);
+
+		if (ret || !sgid_attr.ndev) {
+			if (!ret)
+				ret = -ENXIO;
+			goto out;
+		}
 
-			dev_put(sgid_attr.ndev);
+		ifindex = sgid_attr.ndev->ifindex;
 
-			qp_attr->ah_attr.grh.hop_limit = hop_limit;
-		}
+		ret = rdma_addr_find_l2_eth_by_grh(&sgid,
+						   &ah_attr->grh.dgid,
+						   ah_attr->dmac,
+						   NULL, &ifindex, &hop_limit);
+
+		dev_put(sgid_attr.ndev);
+
+		ah_attr->grh.hop_limit = hop_limit;
 	}
 out:
 	return ret;
 }
 EXPORT_SYMBOL(ib_resolve_eth_dmac);
 
-
 int ib_modify_qp(struct ib_qp *qp,
 		 struct ib_qp_attr *qp_attr,
 		 int qp_attr_mask)
 {
-	int ret;
 
-	ret = ib_resolve_eth_dmac(qp, qp_attr, &qp_attr_mask);
-	if (ret)
-		return ret;
+	if (qp_attr_mask & IB_QP_AV) {
+		int ret;
+
+		ret = ib_resolve_eth_dmac(qp->device, &qp_attr->ah_attr);
+		if (ret)
+			return ret;
+	}
 
 	return qp->device->modify_qp(qp->real_qp, qp_attr, qp_attr_mask, NULL);
 }
-- 
cgit v1.2.1


From 6ad279c5a2e55bf2bd100b4222090d4717de88d5 Mon Sep 17 00:00:00 2001
From: Moni Shoua <monis@mellanox.com>
Date: Wed, 23 Nov 2016 08:23:23 +0200
Subject: IB/mlx5: Report that device has udata response in create_ah

To make mlx5 user driver aware of whether kernel driver returns dmac
in user data response add a new flag that will be returned back to
user-space through alloc_ucontext.

Signed-off-by: Moni Shoua <monis@mellanox.com>
Reviewed-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/main.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index f7a299431e6c..f2ef2a170d9f 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -1117,7 +1117,8 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev,
 		resp.response_length += sizeof(resp.cqe_version);
 
 	if (field_avail(typeof(resp), cmds_supp_uhw, udata->outlen)) {
-		resp.cmds_supp_uhw |= MLX5_USER_CMDS_SUPP_UHW_QUERY_DEVICE;
+		resp.cmds_supp_uhw |= MLX5_USER_CMDS_SUPP_UHW_QUERY_DEVICE |
+				      MLX5_USER_CMDS_SUPP_UHW_CREATE_AH;
 		resp.response_length += sizeof(resp.cmds_supp_uhw);
 	}
 
-- 
cgit v1.2.1


From 477864c8fcd953e5a988073ca5be18bb7fd93410 Mon Sep 17 00:00:00 2001
From: Moni Shoua <monis@mellanox.com>
Date: Wed, 23 Nov 2016 08:23:24 +0200
Subject: IB/core: Let create_ah return extended response to user

Add struct ib_udata to the signature of create_ah callback that is
implemented by IB device drivers. This allows HW drivers to return extra
data to the userspace library.
This patch prepares the ground for mlx5 driver to resolve destination
mac address for a given GID and return it to userspace.
This patch was previously submitted by Knut Omang as a part of the
patch set to support Oracle's Infiniband HCA (SIF).

Signed-off-by: Knut Omang <knut.omang@oracle.com>
Signed-off-by: Moni Shoua <monis@mellanox.com>
Reviewed-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/uverbs_cmd.c         | 11 ++++++++++-
 drivers/infiniband/core/verbs.c              |  2 +-
 drivers/infiniband/hw/cxgb3/iwch_provider.c  |  3 ++-
 drivers/infiniband/hw/cxgb4/provider.c       |  4 +++-
 drivers/infiniband/hw/hns/hns_roce_ah.c      |  3 ++-
 drivers/infiniband/hw/hns/hns_roce_device.h  |  3 ++-
 drivers/infiniband/hw/i40iw/i40iw_verbs.c    |  4 +++-
 drivers/infiniband/hw/mlx4/ah.c              |  4 +++-
 drivers/infiniband/hw/mlx4/mlx4_ib.h         |  3 ++-
 drivers/infiniband/hw/mlx5/ah.c              |  4 +++-
 drivers/infiniband/hw/mlx5/mlx5_ib.h         |  3 ++-
 drivers/infiniband/hw/mthca/mthca_provider.c |  4 +++-
 drivers/infiniband/hw/nes/nes_verbs.c        |  3 ++-
 drivers/infiniband/hw/ocrdma/ocrdma_ah.c     |  3 ++-
 drivers/infiniband/hw/ocrdma/ocrdma_ah.h     |  4 +++-
 drivers/infiniband/hw/qedr/verbs.c           |  3 ++-
 drivers/infiniband/hw/qedr/verbs.h           |  3 ++-
 drivers/infiniband/hw/usnic/usnic_ib_verbs.c |  4 +++-
 drivers/infiniband/hw/usnic/usnic_ib_verbs.h |  4 +++-
 drivers/infiniband/sw/rxe/rxe_verbs.c        |  4 +++-
 20 files changed, 56 insertions(+), 20 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c
index 77bb15f66414..cdb935ddab69 100644
--- a/drivers/infiniband/core/uverbs_cmd.c
+++ b/drivers/infiniband/core/uverbs_cmd.c
@@ -2877,6 +2877,7 @@ ssize_t ib_uverbs_create_ah(struct ib_uverbs_file *file,
 	struct ib_ah			*ah;
 	struct ib_ah_attr		attr;
 	int ret;
+	struct ib_udata                   udata;
 
 	if (out_len < sizeof resp)
 		return -ENOSPC;
@@ -2884,6 +2885,10 @@ ssize_t ib_uverbs_create_ah(struct ib_uverbs_file *file,
 	if (copy_from_user(&cmd, buf, sizeof cmd))
 		return -EFAULT;
 
+	INIT_UDATA(&udata, buf + sizeof(cmd),
+		   (unsigned long)cmd.response + sizeof(resp),
+		   in_len - sizeof(cmd), out_len - sizeof(resp));
+
 	uobj = kmalloc(sizeof *uobj, GFP_KERNEL);
 	if (!uobj)
 		return -ENOMEM;
@@ -2910,12 +2915,16 @@ ssize_t ib_uverbs_create_ah(struct ib_uverbs_file *file,
 	memset(&attr.dmac, 0, sizeof(attr.dmac));
 	memcpy(attr.grh.dgid.raw, cmd.attr.grh.dgid, 16);
 
-	ah = ib_create_ah(pd, &attr);
+	ah = pd->device->create_ah(pd, &attr, &udata);
+
 	if (IS_ERR(ah)) {
 		ret = PTR_ERR(ah);
 		goto err_put;
 	}
 
+	ah->device  = pd->device;
+	ah->pd      = pd;
+	atomic_inc(&pd->usecnt);
 	ah->uobject  = uobj;
 	uobj->object = ah;
 
diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c
index cacee169cd55..c976f29f7ad2 100644
--- a/drivers/infiniband/core/verbs.c
+++ b/drivers/infiniband/core/verbs.c
@@ -315,7 +315,7 @@ struct ib_ah *ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr)
 {
 	struct ib_ah *ah;
 
-	ah = pd->device->create_ah(pd, ah_attr);
+	ah = pd->device->create_ah(pd, ah_attr, NULL);
 
 	if (!IS_ERR(ah)) {
 		ah->device  = pd->device;
diff --git a/drivers/infiniband/hw/cxgb3/iwch_provider.c b/drivers/infiniband/hw/cxgb3/iwch_provider.c
index cba57bb53dba..9d5fe1853da4 100644
--- a/drivers/infiniband/hw/cxgb3/iwch_provider.c
+++ b/drivers/infiniband/hw/cxgb3/iwch_provider.c
@@ -62,7 +62,8 @@
 #include "common.h"
 
 static struct ib_ah *iwch_ah_create(struct ib_pd *pd,
-				    struct ib_ah_attr *ah_attr)
+				    struct ib_ah_attr *ah_attr,
+				    struct ib_udata *udata)
 {
 	return ERR_PTR(-ENOSYS);
 }
diff --git a/drivers/infiniband/hw/cxgb4/provider.c b/drivers/infiniband/hw/cxgb4/provider.c
index 645e606a17c5..49b51b7e0fd7 100644
--- a/drivers/infiniband/hw/cxgb4/provider.c
+++ b/drivers/infiniband/hw/cxgb4/provider.c
@@ -59,7 +59,9 @@ module_param(fastreg_support, int, 0644);
 MODULE_PARM_DESC(fastreg_support, "Advertise fastreg support (default=1)");
 
 static struct ib_ah *c4iw_ah_create(struct ib_pd *pd,
-				    struct ib_ah_attr *ah_attr)
+				    struct ib_ah_attr *ah_attr,
+				    struct ib_udata *udata)
+
 {
 	return ERR_PTR(-ENOSYS);
 }
diff --git a/drivers/infiniband/hw/hns/hns_roce_ah.c b/drivers/infiniband/hw/hns/hns_roce_ah.c
index 24f79ee39fdf..0ac294db3b29 100644
--- a/drivers/infiniband/hw/hns/hns_roce_ah.c
+++ b/drivers/infiniband/hw/hns/hns_roce_ah.c
@@ -39,7 +39,8 @@
 #define HNS_ROCE_VLAN_SL_BIT_MASK	7
 #define HNS_ROCE_VLAN_SL_SHIFT		13
 
-struct ib_ah *hns_roce_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *ah_attr)
+struct ib_ah *hns_roce_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *ah_attr,
+				 struct ib_udata *udata)
 {
 	struct hns_roce_dev *hr_dev = to_hr_dev(ibpd->device);
 	struct device *dev = &hr_dev->pdev->dev;
diff --git a/drivers/infiniband/hw/hns/hns_roce_device.h b/drivers/infiniband/hw/hns/hns_roce_device.h
index 341731553a60..470615f7b517 100644
--- a/drivers/infiniband/hw/hns/hns_roce_device.h
+++ b/drivers/infiniband/hw/hns/hns_roce_device.h
@@ -667,7 +667,8 @@ int hns_roce_bitmap_alloc_range(struct hns_roce_bitmap *bitmap, int cnt,
 void hns_roce_bitmap_free_range(struct hns_roce_bitmap *bitmap,
 				unsigned long obj, int cnt);
 
-struct ib_ah *hns_roce_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr);
+struct ib_ah *hns_roce_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr,
+				 struct ib_udata *udata);
 int hns_roce_query_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr);
 int hns_roce_destroy_ah(struct ib_ah *ah);
 
diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
index 6329c971c22f..f03fc1527d70 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
@@ -2562,7 +2562,9 @@ static int i40iw_query_pkey(struct ib_device *ibdev,
  * @ah_attr: address handle attributes
  */
 static struct ib_ah *i40iw_create_ah(struct ib_pd *ibpd,
-				     struct ib_ah_attr *attr)
+				     struct ib_ah_attr *attr,
+				     struct ib_udata *udata)
+
 {
 	return ERR_PTR(-ENOSYS);
 }
diff --git a/drivers/infiniband/hw/mlx4/ah.c b/drivers/infiniband/hw/mlx4/ah.c
index 6be7dc320ff7..20c6d17ac8b8 100644
--- a/drivers/infiniband/hw/mlx4/ah.c
+++ b/drivers/infiniband/hw/mlx4/ah.c
@@ -126,7 +126,9 @@ static struct ib_ah *create_iboe_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr
 	return &ah->ibah;
 }
 
-struct ib_ah *mlx4_ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr)
+struct ib_ah *mlx4_ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr,
+				struct ib_udata *udata)
+
 {
 	struct mlx4_ib_ah *ah;
 	struct ib_ah *ret;
diff --git a/drivers/infiniband/hw/mlx4/mlx4_ib.h b/drivers/infiniband/hw/mlx4/mlx4_ib.h
index 35141f451e5c..7f3d976d81ed 100644
--- a/drivers/infiniband/hw/mlx4/mlx4_ib.h
+++ b/drivers/infiniband/hw/mlx4/mlx4_ib.h
@@ -742,7 +742,8 @@ int mlx4_ib_arm_cq(struct ib_cq *cq, enum ib_cq_notify_flags flags);
 void __mlx4_ib_cq_clean(struct mlx4_ib_cq *cq, u32 qpn, struct mlx4_ib_srq *srq);
 void mlx4_ib_cq_clean(struct mlx4_ib_cq *cq, u32 qpn, struct mlx4_ib_srq *srq);
 
-struct ib_ah *mlx4_ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr);
+struct ib_ah *mlx4_ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr,
+				struct ib_udata *udata);
 int mlx4_ib_query_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr);
 int mlx4_ib_destroy_ah(struct ib_ah *ah);
 
diff --git a/drivers/infiniband/hw/mlx5/ah.c b/drivers/infiniband/hw/mlx5/ah.c
index 745efa4cfc71..ecac9ea2c85f 100644
--- a/drivers/infiniband/hw/mlx5/ah.c
+++ b/drivers/infiniband/hw/mlx5/ah.c
@@ -64,7 +64,9 @@ static struct ib_ah *create_ib_ah(struct mlx5_ib_dev *dev,
 	return &ah->ibah;
 }
 
-struct ib_ah *mlx5_ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr)
+struct ib_ah *mlx5_ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr,
+				struct ib_udata *udata)
+
 {
 	struct mlx5_ib_ah *ah;
 	struct mlx5_ib_dev *dev = to_mdev(pd->device);
diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h
index 95937e7d2584..ff05afa864ee 100644
--- a/drivers/infiniband/hw/mlx5/mlx5_ib.h
+++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h
@@ -739,7 +739,8 @@ void mlx5_ib_free_srq_wqe(struct mlx5_ib_srq *srq, int wqe_index);
 int mlx5_MAD_IFC(struct mlx5_ib_dev *dev, int ignore_mkey, int ignore_bkey,
 		 u8 port, const struct ib_wc *in_wc, const struct ib_grh *in_grh,
 		 const void *in_mad, void *response_mad);
-struct ib_ah *mlx5_ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr);
+struct ib_ah *mlx5_ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr,
+				struct ib_udata *udata);
 int mlx5_ib_query_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr);
 int mlx5_ib_destroy_ah(struct ib_ah *ah);
 struct ib_srq *mlx5_ib_create_srq(struct ib_pd *pd,
diff --git a/drivers/infiniband/hw/mthca/mthca_provider.c b/drivers/infiniband/hw/mthca/mthca_provider.c
index 358930a41e36..d31708742ba5 100644
--- a/drivers/infiniband/hw/mthca/mthca_provider.c
+++ b/drivers/infiniband/hw/mthca/mthca_provider.c
@@ -410,7 +410,9 @@ static int mthca_dealloc_pd(struct ib_pd *pd)
 }
 
 static struct ib_ah *mthca_ah_create(struct ib_pd *pd,
-				     struct ib_ah_attr *ah_attr)
+				     struct ib_ah_attr *ah_attr,
+				     struct ib_udata *udata)
+
 {
 	int err;
 	struct mthca_ah *ah;
diff --git a/drivers/infiniband/hw/nes/nes_verbs.c b/drivers/infiniband/hw/nes/nes_verbs.c
index bd69125731c1..0bb857c7e439 100644
--- a/drivers/infiniband/hw/nes/nes_verbs.c
+++ b/drivers/infiniband/hw/nes/nes_verbs.c
@@ -771,7 +771,8 @@ static int nes_dealloc_pd(struct ib_pd *ibpd)
 /**
  * nes_create_ah
  */
-static struct ib_ah *nes_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr)
+static struct ib_ah *nes_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr,
+				   struct ib_udata *udata)
 {
 	return ERR_PTR(-ENOSYS);
 }
diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_ah.c b/drivers/infiniband/hw/ocrdma/ocrdma_ah.c
index 797362a297b2..14d33b0f3950 100644
--- a/drivers/infiniband/hw/ocrdma/ocrdma_ah.c
+++ b/drivers/infiniband/hw/ocrdma/ocrdma_ah.c
@@ -154,7 +154,8 @@ static inline int set_av_attr(struct ocrdma_dev *dev, struct ocrdma_ah *ah,
 	return status;
 }
 
-struct ib_ah *ocrdma_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr)
+struct ib_ah *ocrdma_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr,
+			       struct ib_udata *udata)
 {
 	u32 *ahid_addr;
 	int status;
diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_ah.h b/drivers/infiniband/hw/ocrdma/ocrdma_ah.h
index 3856dd4c7e3d..0704a24b17c8 100644
--- a/drivers/infiniband/hw/ocrdma/ocrdma_ah.h
+++ b/drivers/infiniband/hw/ocrdma/ocrdma_ah.h
@@ -50,7 +50,9 @@ enum {
 	OCRDMA_AH_L3_TYPE_MASK		= 0x03,
 	OCRDMA_AH_L3_TYPE_SHIFT		= 0x1D /* 29 bits */
 };
-struct ib_ah *ocrdma_create_ah(struct ib_pd *, struct ib_ah_attr *);
+
+struct ib_ah *ocrdma_create_ah(struct ib_pd *, struct ib_ah_attr *,
+			       struct ib_udata *);
 int ocrdma_destroy_ah(struct ib_ah *);
 int ocrdma_query_ah(struct ib_ah *, struct ib_ah_attr *);
 int ocrdma_modify_ah(struct ib_ah *, struct ib_ah_attr *);
diff --git a/drivers/infiniband/hw/qedr/verbs.c b/drivers/infiniband/hw/qedr/verbs.c
index a61514296767..ccff6c6e3f33 100644
--- a/drivers/infiniband/hw/qedr/verbs.c
+++ b/drivers/infiniband/hw/qedr/verbs.c
@@ -2094,7 +2094,8 @@ int qedr_destroy_qp(struct ib_qp *ibqp)
 	return rc;
 }
 
-struct ib_ah *qedr_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr)
+struct ib_ah *qedr_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr,
+			     struct ib_udata *udata)
 {
 	struct qedr_ah *ah;
 
diff --git a/drivers/infiniband/hw/qedr/verbs.h b/drivers/infiniband/hw/qedr/verbs.h
index a9b5e67bb81e..070677ca4d19 100644
--- a/drivers/infiniband/hw/qedr/verbs.h
+++ b/drivers/infiniband/hw/qedr/verbs.h
@@ -70,7 +70,8 @@ int qedr_query_qp(struct ib_qp *, struct ib_qp_attr *qp_attr,
 		  int qp_attr_mask, struct ib_qp_init_attr *);
 int qedr_destroy_qp(struct ib_qp *ibqp);
 
-struct ib_ah *qedr_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr);
+struct ib_ah *qedr_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr,
+			     struct ib_udata *udata);
 int qedr_destroy_ah(struct ib_ah *ibah);
 
 int qedr_dereg_mr(struct ib_mr *);
diff --git a/drivers/infiniband/hw/usnic/usnic_ib_verbs.c b/drivers/infiniband/hw/usnic/usnic_ib_verbs.c
index a5bfbba6bbac..fd2a50eb4c91 100644
--- a/drivers/infiniband/hw/usnic/usnic_ib_verbs.c
+++ b/drivers/infiniband/hw/usnic/usnic_ib_verbs.c
@@ -738,7 +738,9 @@ int usnic_ib_mmap(struct ib_ucontext *context,
 
 /* In ib callbacks section -  Start of stub funcs */
 struct ib_ah *usnic_ib_create_ah(struct ib_pd *pd,
-					struct ib_ah_attr *ah_attr)
+				 struct ib_ah_attr *ah_attr,
+				 struct ib_udata *udata)
+
 {
 	usnic_dbg("\n");
 	return ERR_PTR(-EPERM);
diff --git a/drivers/infiniband/hw/usnic/usnic_ib_verbs.h b/drivers/infiniband/hw/usnic/usnic_ib_verbs.h
index 0d9d2e6a14d5..0ed8e072329e 100644
--- a/drivers/infiniband/hw/usnic/usnic_ib_verbs.h
+++ b/drivers/infiniband/hw/usnic/usnic_ib_verbs.h
@@ -75,7 +75,9 @@ int usnic_ib_dealloc_ucontext(struct ib_ucontext *ibcontext);
 int usnic_ib_mmap(struct ib_ucontext *context,
 			struct vm_area_struct *vma);
 struct ib_ah *usnic_ib_create_ah(struct ib_pd *pd,
-					struct ib_ah_attr *ah_attr);
+				 struct ib_ah_attr *ah_attr,
+				 struct ib_udata *udata);
+
 int usnic_ib_destroy_ah(struct ib_ah *ah);
 int usnic_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 			struct ib_send_wr **bad_wr);
diff --git a/drivers/infiniband/sw/rxe/rxe_verbs.c b/drivers/infiniband/sw/rxe/rxe_verbs.c
index 19841c863daf..187d85ccfe58 100644
--- a/drivers/infiniband/sw/rxe/rxe_verbs.c
+++ b/drivers/infiniband/sw/rxe/rxe_verbs.c
@@ -316,7 +316,9 @@ static int rxe_init_av(struct rxe_dev *rxe, struct ib_ah_attr *attr,
 	return err;
 }
 
-static struct ib_ah *rxe_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr)
+static struct ib_ah *rxe_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr,
+				   struct ib_udata *udata)
+
 {
 	int err;
 	struct rxe_dev *rxe = to_rdev(ibpd->device);
-- 
cgit v1.2.1


From 5097e71f3edafad3e7d8d4f9c4a137d9aad0fae2 Mon Sep 17 00:00:00 2001
From: Moni Shoua <monis@mellanox.com>
Date: Wed, 23 Nov 2016 08:23:25 +0200
Subject: IB/mlx5: Use kernel driver to help userspace create ah

Resolving a MAC address for a given IP address in userspace is inefficient.
This patch lets mlx5 user driver using the kernel driver to resolve the mac
and get the answer in the private section of the response.

Signed-off-by: Moni Shoua <monis@mellanox.com>
Reviewed-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/ah.c | 21 +++++++++++++++++++++
 1 file changed, 21 insertions(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/ah.c b/drivers/infiniband/hw/mlx5/ah.c
index ecac9ea2c85f..d090e96f6f01 100644
--- a/drivers/infiniband/hw/mlx5/ah.c
+++ b/drivers/infiniband/hw/mlx5/ah.c
@@ -77,6 +77,27 @@ struct ib_ah *mlx5_ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr,
 	if (ll == IB_LINK_LAYER_ETHERNET && !(ah_attr->ah_flags & IB_AH_GRH))
 		return ERR_PTR(-EINVAL);
 
+	if (ll == IB_LINK_LAYER_ETHERNET && udata) {
+		int err;
+		struct mlx5_ib_create_ah_resp resp = {};
+		u32 min_resp_len = offsetof(typeof(resp), dmac) +
+				   sizeof(resp.dmac);
+
+		if (udata->outlen < min_resp_len)
+			return ERR_PTR(-EINVAL);
+
+		resp.response_length = min_resp_len;
+
+		err = ib_resolve_eth_dmac(pd->device, ah_attr);
+		if (err)
+			return ERR_PTR(err);
+
+		memcpy(resp.dmac, ah_attr->dmac, ETH_ALEN);
+		err = ib_copy_to_udata(udata, &resp, resp.response_length);
+		if (err)
+			return ERR_PTR(err);
+	}
+
 	ah = kzalloc(sizeof(*ah), GFP_ATOMIC);
 	if (!ah)
 		return ERR_PTR(-ENOMEM);
-- 
cgit v1.2.1


From 41c450fd8da549c5f7cced6650354095b0d4312a Mon Sep 17 00:00:00 2001
From: Moni Shoua <monis@mellanox.com>
Date: Wed, 23 Nov 2016 08:23:26 +0200
Subject: IB/mlx5: Make create/destroy_ah available to userspace

Advertise that create_ah and destroy_ah verbs are accessible from
uverbs interface.

Signed-off-by: Moni Shoua <monis@mellanox.com>
Reviewed-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/main.c | 2 ++
 1 file changed, 2 insertions(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index f2ef2a170d9f..7e5b80361724 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -3053,6 +3053,8 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev)
 		(1ull << IB_USER_VERBS_CMD_QUERY_PORT)		|
 		(1ull << IB_USER_VERBS_CMD_ALLOC_PD)		|
 		(1ull << IB_USER_VERBS_CMD_DEALLOC_PD)		|
+		(1ull << IB_USER_VERBS_CMD_CREATE_AH)		|
+		(1ull << IB_USER_VERBS_CMD_DESTROY_AH)		|
 		(1ull << IB_USER_VERBS_CMD_REG_MR)		|
 		(1ull << IB_USER_VERBS_CMD_REREG_MR)		|
 		(1ull << IB_USER_VERBS_CMD_DEREG_MR)		|
-- 
cgit v1.2.1


From c482af646d0809a8d5e1b7f4398cce3592589b98 Mon Sep 17 00:00:00 2001
From: Jack Morgenstein <jackm@dev.mellanox.co.il>
Date: Sun, 27 Nov 2016 15:18:19 +0200
Subject: IB/mlx4: Fix out-of-range array index in destroy qp flow

For non-special QPs, the port value becomes non-zero only at the
RESET-to-INIT transition. If the QP has not undergone that transition,
its port number value is still zero.

If such a QP is destroyed before being moved out of the RESET state,
subtracting one from the qp port number results in a negative value.
Using that negative value as an index into the qp1_proxy array
results in an out-of-bounds array reference.

Fix this by testing that the QP type is one that uses qp1_proxy before
using the port number. For special QPs of all types, the port number is
specified at QP creation time.

Fixes: 9433c188915c ("IB/mlx4: Invoke UPDATE_QP for proxy QP1 on MAC changes")
Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/qp.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index 0914731c9aac..c22454383976 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -1280,7 +1280,8 @@ static int _mlx4_ib_destroy_qp(struct ib_qp *qp)
 	if (is_qp0(dev, mqp))
 		mlx4_CLOSE_PORT(dev->dev, mqp->port);
 
-	if (dev->qp1_proxy[mqp->port - 1] == mqp) {
+	if (mqp->mlx4_ib_qp_type == MLX4_IB_QPT_PROXY_GSI &&
+	    dev->qp1_proxy[mqp->port - 1] == mqp) {
 		mutex_lock(&dev->qp1_proxy_lock[mqp->port - 1]);
 		dev->qp1_proxy[mqp->port - 1] = NULL;
 		mutex_unlock(&dev->qp1_proxy_lock[mqp->port - 1]);
-- 
cgit v1.2.1


From c73b7911de97fad3ab9032a110af48d6ab2da48f Mon Sep 17 00:00:00 2001
From: Maor Gottlieb <maorg@mellanox.com>
Date: Sun, 27 Nov 2016 15:18:20 +0200
Subject: IB/mlx5: Assign SRQ type earlier

Move the SRQ type assignment to be before actually using it
in create_srq_user() and in create_srq_kernel() functions.

Fixes: af1ba291c5e4 ('{net, IB}/mlx5: Refactor internal SRQ API')
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Reviewed-by: Majd Dibbiny <majd@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/srq.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/srq.c b/drivers/infiniband/hw/mlx5/srq.c
index f384db5367fb..83588db2785f 100644
--- a/drivers/infiniband/hw/mlx5/srq.c
+++ b/drivers/infiniband/hw/mlx5/srq.c
@@ -282,6 +282,7 @@ struct ib_srq *mlx5_ib_create_srq(struct ib_pd *pd,
 	mlx5_ib_dbg(dev, "desc_size 0x%x, req wr 0x%x, srq size 0x%x, max_gs 0x%x, max_avail_gather 0x%x\n",
 		    desc_size, init_attr->attr.max_wr, srq->msrq.max, srq->msrq.max_gs,
 		    srq->msrq.max_avail_gather);
+	in.type = init_attr->srq_type;
 
 	if (pd->uobject)
 		err = create_srq_user(pd, srq, &in, udata, buf_size);
@@ -294,7 +295,6 @@ struct ib_srq *mlx5_ib_create_srq(struct ib_pd *pd,
 		goto err_srq;
 	}
 
-	in.type = init_attr->srq_type;
 	in.log_size = ilog2(srq->msrq.max);
 	in.wqe_shift = srq->msrq.wqe_shift - 4;
 	if (srq->wq_sig)
-- 
cgit v1.2.1


From afd02cd3a9b6c04b41d946b5d7f6e17b3fc30c6b Mon Sep 17 00:00:00 2001
From: Eli Cohen <eli@mellanox.com>
Date: Sun, 27 Nov 2016 15:18:21 +0200
Subject: IB/mlx5: Avoid system crash when enabling many VFs

When enabling many VFs, the total amount of DMA mappings increase
significantly. This causes DMA allocations to take a lot of time
since they are serialized in the kernel.

As a result the driver enters into fatal condition due to
timeout and the system hangs. To recover from this we disable
MR cache for VFs.

PFs will still have a full cache and VFs cache can be manipulated
as usual after driver load.

Fixes: e126ba97dba9 ('mlx5: Add driver for Mellanox Connect-IB adapters')
Signed-off-by: Eli Cohen <eli@mellanox.com>
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/mr.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/mr.c b/drivers/infiniband/hw/mlx5/mr.c
index 6cbda901f259..5258ac870e6a 100644
--- a/drivers/infiniband/hw/mlx5/mr.c
+++ b/drivers/infiniband/hw/mlx5/mr.c
@@ -627,7 +627,8 @@ int mlx5_mr_cache_init(struct mlx5_ib_dev *dev)
 		ent->order = i + 2;
 		ent->dev = dev;
 
-		if (dev->mdev->profile->mask & MLX5_PROF_MASK_MR_CACHE)
+		if ((dev->mdev->profile->mask & MLX5_PROF_MASK_MR_CACHE) &&
+		    (mlx5_core_is_pf(dev->mdev)))
 			limit = dev->mdev->profile->mr_cache[i].limit;
 		else
 			limit = 0;
-- 
cgit v1.2.1


From b216af408c985092a79472ad10e6f216cb2973fc Mon Sep 17 00:00:00 2001
From: Maor Gottlieb <maorg@mellanox.com>
Date: Sun, 27 Nov 2016 15:18:22 +0200
Subject: IB/mlx5: Use u64 for UMR length

The fast_registration length is used to convey length for memory
registrations through UMR which can be of any size up to 2^64.

Change the length type to be u64.

Fixes: 968e78dd9644 ('IB/mlx5: Enhance UMR support to allow partial page table update')
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/mlx5_ib.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h
index ff05afa864ee..df3d6af3f683 100644
--- a/drivers/infiniband/hw/mlx5/mlx5_ib.h
+++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h
@@ -420,7 +420,7 @@ struct mlx5_umr_wr {
 	struct ib_pd		       *pd;
 	unsigned int			page_shift;
 	unsigned int			npages;
-	u32				length;
+	u64				length;
 	int				access_flags;
 	u32				mkey;
 };
-- 
cgit v1.2.1


From d012f5d6f8597f936f44c79e46345fda86dcff4d Mon Sep 17 00:00:00 2001
From: Or Gerlitz <ogerlitz@mellanox.com>
Date: Sun, 27 Nov 2016 16:51:34 +0200
Subject: IB/mlx5: Refactor registration to netdev notifier

Refactor the netdev notifier registration into a small helper function.

This is a pre-step towards having mlx5 IB device over an Ethernet port
which doesn't support RoCE. Also, renamed the de-registration helper
and the new helper as netdev notifier and not roce, to make it clear
this is not only used with roce.

This patch doesn't change any functionality.

Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/main.c | 29 ++++++++++++++++++++---------
 1 file changed, 20 insertions(+), 9 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index 7e5b80361724..0ee1e4ab02fa 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -2843,7 +2843,21 @@ static void mlx5_roce_lag_cleanup(struct mlx5_ib_dev *dev)
 	}
 }
 
-static void mlx5_remove_roce_notifier(struct mlx5_ib_dev *dev)
+static int mlx5_add_netdev_notifier(struct mlx5_ib_dev *dev)
+{
+	int err;
+
+	dev->roce.nb.notifier_call = mlx5_netdev_event;
+	err = register_netdevice_notifier(&dev->roce.nb);
+	if (err) {
+		dev->roce.nb.notifier_call = NULL;
+		return err;
+	}
+
+	return 0;
+}
+
+static void mlx5_remove_netdev_notifier(struct mlx5_ib_dev *dev)
 {
 	if (dev->roce.nb.notifier_call) {
 		unregister_netdevice_notifier(&dev->roce.nb);
@@ -2855,12 +2869,9 @@ static int mlx5_enable_roce(struct mlx5_ib_dev *dev)
 {
 	int err;
 
-	dev->roce.nb.notifier_call = mlx5_netdev_event;
-	err = register_netdevice_notifier(&dev->roce.nb);
-	if (err) {
-		dev->roce.nb.notifier_call = NULL;
+	err = mlx5_add_netdev_notifier(dev);
+	if (err)
 		return err;
-	}
 
 	err = mlx5_nic_vport_enable_roce(dev->mdev);
 	if (err)
@@ -2876,7 +2887,7 @@ err_disable_roce:
 	mlx5_nic_vport_disable_roce(dev->mdev);
 
 err_unregister_netdevice_notifier:
-	mlx5_remove_roce_notifier(dev);
+	mlx5_remove_netdev_notifier(dev);
 	return err;
 }
 
@@ -3242,7 +3253,7 @@ err_rsrc:
 err_disable_roce:
 	if (ll == IB_LINK_LAYER_ETHERNET) {
 		mlx5_disable_roce(dev);
-		mlx5_remove_roce_notifier(dev);
+		mlx5_remove_netdev_notifier(dev);
 	}
 
 err_free_port:
@@ -3259,7 +3270,7 @@ static void mlx5_ib_remove(struct mlx5_core_dev *mdev, void *context)
 	struct mlx5_ib_dev *dev = context;
 	enum rdma_link_layer ll = mlx5_ib_port_link_layer(&dev->ib_dev, 1);
 
-	mlx5_remove_roce_notifier(dev);
+	mlx5_remove_netdev_notifier(dev);
 	ib_unregister_device(&dev->ib_dev);
 	mlx5_ib_dealloc_q_counters(dev);
 	destroy_umrc_res(dev);
-- 
cgit v1.2.1


From 45f95acd63222dd1dc752fa904536327b10f1082 Mon Sep 17 00:00:00 2001
From: Or Gerlitz <ogerlitz@mellanox.com>
Date: Sun, 27 Nov 2016 16:51:35 +0200
Subject: IB/mlx5: Rename RoCE related helpers to reflect being Eth ones

This is a pre-step towards having mlx5 IB device also over Eth ports where
RoCE is not supported. We change the roce enable/disable and roce_lag
init/fini function names to have _eth instead of _roce.

This patch doesn't change any functionality.

Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/main.c | 22 +++++++++++-----------
 1 file changed, 11 insertions(+), 11 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index 0ee1e4ab02fa..b5b7459e9af0 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -2802,7 +2802,7 @@ static void get_dev_fw_str(struct ib_device *ibdev, char *str,
 		       fw_rev_min(dev->mdev), fw_rev_sub(dev->mdev));
 }
 
-static int mlx5_roce_lag_init(struct mlx5_ib_dev *dev)
+static int mlx5_eth_lag_init(struct mlx5_ib_dev *dev)
 {
 	struct mlx5_core_dev *mdev = dev->mdev;
 	struct mlx5_flow_namespace *ns = mlx5_get_flow_namespace(mdev,
@@ -2831,7 +2831,7 @@ err_destroy_vport_lag:
 	return err;
 }
 
-static void mlx5_roce_lag_cleanup(struct mlx5_ib_dev *dev)
+static void mlx5_eth_lag_cleanup(struct mlx5_ib_dev *dev)
 {
 	struct mlx5_core_dev *mdev = dev->mdev;
 
@@ -2865,7 +2865,7 @@ static void mlx5_remove_netdev_notifier(struct mlx5_ib_dev *dev)
 	}
 }
 
-static int mlx5_enable_roce(struct mlx5_ib_dev *dev)
+static int mlx5_enable_eth(struct mlx5_ib_dev *dev)
 {
 	int err;
 
@@ -2877,7 +2877,7 @@ static int mlx5_enable_roce(struct mlx5_ib_dev *dev)
 	if (err)
 		goto err_unregister_netdevice_notifier;
 
-	err = mlx5_roce_lag_init(dev);
+	err = mlx5_eth_lag_init(dev);
 	if (err)
 		goto err_disable_roce;
 
@@ -2891,9 +2891,9 @@ err_unregister_netdevice_notifier:
 	return err;
 }
 
-static void mlx5_disable_roce(struct mlx5_ib_dev *dev)
+static void mlx5_disable_eth(struct mlx5_ib_dev *dev)
 {
-	mlx5_roce_lag_cleanup(dev);
+	mlx5_eth_lag_cleanup(dev);
 	mlx5_nic_vport_disable_roce(dev->mdev);
 }
 
@@ -3199,14 +3199,14 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev)
 	spin_lock_init(&dev->reset_flow_resource_lock);
 
 	if (ll == IB_LINK_LAYER_ETHERNET) {
-		err = mlx5_enable_roce(dev);
+		err = mlx5_enable_eth(dev);
 		if (err)
 			goto err_dealloc;
 	}
 
 	err = create_dev_resources(&dev->devr);
 	if (err)
-		goto err_disable_roce;
+		goto err_disable_eth;
 
 	err = mlx5_ib_odp_init_one(dev);
 	if (err)
@@ -3250,9 +3250,9 @@ err_odp:
 err_rsrc:
 	destroy_dev_resources(&dev->devr);
 
-err_disable_roce:
+err_disable_eth:
 	if (ll == IB_LINK_LAYER_ETHERNET) {
-		mlx5_disable_roce(dev);
+		mlx5_disable_eth(dev);
 		mlx5_remove_netdev_notifier(dev);
 	}
 
@@ -3277,7 +3277,7 @@ static void mlx5_ib_remove(struct mlx5_core_dev *mdev, void *context)
 	mlx5_ib_odp_remove_one(dev);
 	destroy_dev_resources(&dev->devr);
 	if (ll == IB_LINK_LAYER_ETHERNET)
-		mlx5_disable_roce(dev);
+		mlx5_disable_eth(dev);
 	kfree(dev->port);
 	ib_dealloc_device(&dev->ib_dev);
 }
-- 
cgit v1.2.1


From ca5b91d63192ceaa41a6145f8c923debb64c71fa Mon Sep 17 00:00:00 2001
From: Or Gerlitz <ogerlitz@mellanox.com>
Date: Sun, 27 Nov 2016 16:51:36 +0200
Subject: IB/mlx5: Support RAW Ethernet when RoCE is disabled

On some environments, such as certain SRIOV VF configurations, RoCE is
not supported for mlx5 Ethernet ports. Currently, the driver will not
open IB device on that port.

This is problematic, since we do want user-space RAW Ethernet (RAW_PACKET
QPs) functionality to remain in place. For that end, enhance the relevant
driver flows such that we do create a device instance in that case.

Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/main.c | 22 +++++++++++++---------
 1 file changed, 13 insertions(+), 9 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index b5b7459e9af0..8b013f8b832a 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -2779,6 +2779,8 @@ static int mlx5_port_immutable(struct ib_device *ibdev, u8 port_num,
 			       struct ib_port_immutable *immutable)
 {
 	struct ib_port_attr attr;
+	struct mlx5_ib_dev *dev = to_mdev(ibdev);
+	enum rdma_link_layer ll = mlx5_ib_port_link_layer(ibdev, port_num);
 	int err;
 
 	err = mlx5_ib_query_port(ibdev, port_num, &attr);
@@ -2788,7 +2790,8 @@ static int mlx5_port_immutable(struct ib_device *ibdev, u8 port_num,
 	immutable->pkey_tbl_len = attr.pkey_tbl_len;
 	immutable->gid_tbl_len = attr.gid_tbl_len;
 	immutable->core_cap_flags = get_core_cap_flags(ibdev);
-	immutable->max_mad_size = IB_MGMT_MAD_SIZE;
+	if ((ll == IB_LINK_LAYER_INFINIBAND) || MLX5_CAP_GEN(dev->mdev, roce))
+		immutable->max_mad_size = IB_MGMT_MAD_SIZE;
 
 	return 0;
 }
@@ -2873,9 +2876,11 @@ static int mlx5_enable_eth(struct mlx5_ib_dev *dev)
 	if (err)
 		return err;
 
-	err = mlx5_nic_vport_enable_roce(dev->mdev);
-	if (err)
-		goto err_unregister_netdevice_notifier;
+	if (MLX5_CAP_GEN(dev->mdev, roce)) {
+		err = mlx5_nic_vport_enable_roce(dev->mdev);
+		if (err)
+			goto err_unregister_netdevice_notifier;
+	}
 
 	err = mlx5_eth_lag_init(dev);
 	if (err)
@@ -2884,7 +2889,8 @@ static int mlx5_enable_eth(struct mlx5_ib_dev *dev)
 	return 0;
 
 err_disable_roce:
-	mlx5_nic_vport_disable_roce(dev->mdev);
+	if (MLX5_CAP_GEN(dev->mdev, roce))
+		mlx5_nic_vport_disable_roce(dev->mdev);
 
 err_unregister_netdevice_notifier:
 	mlx5_remove_netdev_notifier(dev);
@@ -2894,7 +2900,8 @@ err_unregister_netdevice_notifier:
 static void mlx5_disable_eth(struct mlx5_ib_dev *dev)
 {
 	mlx5_eth_lag_cleanup(dev);
-	mlx5_nic_vport_disable_roce(dev->mdev);
+	if (MLX5_CAP_GEN(dev->mdev, roce))
+		mlx5_nic_vport_disable_roce(dev->mdev);
 }
 
 static void mlx5_ib_dealloc_q_counters(struct mlx5_ib_dev *dev)
@@ -3016,9 +3023,6 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev)
 	port_type_cap = MLX5_CAP_GEN(mdev, port_type);
 	ll = mlx5_port_type_cap_to_rdma_ll(port_type_cap);
 
-	if ((ll == IB_LINK_LAYER_ETHERNET) && !MLX5_CAP_GEN(mdev, roce))
-		return NULL;
-
 	printk_once(KERN_INFO "%s", mlx5_version);
 
 	dev = (struct mlx5_ib_dev *)ib_alloc_device(sizeof(*dev));
-- 
cgit v1.2.1


From d949167d68b304c0a00331cf33ef49a29b65d85f Mon Sep 17 00:00:00 2001
From: Bodong Wang <bodong@mellanox.com>
Date: Thu, 1 Dec 2016 13:43:13 +0200
Subject: IB/mlx5: Report mlx5 packet pacing capabilities when querying device

Enable mlx5 based hardware to report packet pacing capabilities
from kernel to user space. Packet pacing allows to limit the rate to any
number between the maximum and minimum, based on user settings.

The capabilities are exposed to user space through query_device by uhw.
The following capabilities are reported:

1. The maximum and minimum rate limit in kbps supported by packet pacing.
2. Bitmap showing which QP types are supported by packet pacing operation.

Signed-off-by: Bodong Wang <bodong@mellanox.com>
Reviewed-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/main.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index 8b013f8b832a..6c194000903d 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -693,6 +693,19 @@ static int mlx5_ib_query_device(struct ib_device *ibdev,
 		resp.response_length += sizeof(resp.cqe_comp_caps);
 	}
 
+	if (field_avail(typeof(resp), packet_pacing_caps, uhw->outlen)) {
+		if (MLX5_CAP_QOS(mdev, packet_pacing) &&
+		    MLX5_CAP_GEN(mdev, qos)) {
+			resp.packet_pacing_caps.qp_rate_limit_max =
+				MLX5_CAP_QOS(mdev, packet_pacing_max_rate);
+			resp.packet_pacing_caps.qp_rate_limit_min =
+				MLX5_CAP_QOS(mdev, packet_pacing_min_rate);
+			resp.packet_pacing_caps.supported_qpts |=
+				1 << IB_QPT_RAW_PACKET;
+		}
+		resp.response_length += sizeof(resp.packet_pacing_caps);
+	}
+
 	if (uhw->outlen) {
 		err = ib_copy_to_udata(uhw, &resp, resp.response_length);
 
-- 
cgit v1.2.1


From 528e5a1bd3f0e9b760cb3a1062fce7513712a15d Mon Sep 17 00:00:00 2001
From: Bodong Wang <bodong@mellanox.com>
Date: Thu, 1 Dec 2016 13:43:14 +0200
Subject: IB/core: Support rate limit for packet pacing

Add new member rate_limit to ib_qp_attr which holds the packet pacing rate
in kbps, 0 means unlimited.

IB_QP_RATE_LIMIT is added to ib_attr_mask and could be used by RAW
QPs when changing QP state from RTR to RTS, RTS to RTS.

Signed-off-by: Bodong Wang <bodong@mellanox.com>
Reviewed-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/verbs.c | 2 ++
 1 file changed, 2 insertions(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c
index c976f29f7ad2..71580cc28c9e 100644
--- a/drivers/infiniband/core/verbs.c
+++ b/drivers/infiniband/core/verbs.c
@@ -1016,6 +1016,7 @@ static const struct {
 						 IB_QP_QKEY),
 				 [IB_QPT_GSI] = (IB_QP_CUR_STATE		|
 						 IB_QP_QKEY),
+				 [IB_QPT_RAW_PACKET] = IB_QP_RATE_LIMIT,
 			 }
 		}
 	},
@@ -1049,6 +1050,7 @@ static const struct {
 						IB_QP_QKEY),
 				[IB_QPT_GSI] = (IB_QP_CUR_STATE			|
 						IB_QP_QKEY),
+				[IB_QPT_RAW_PACKET] = IB_QP_RATE_LIMIT,
 			}
 		},
 		[IB_QPS_SQD]   = {
-- 
cgit v1.2.1


From 189aba99e70030cfb56bd8f199bc5b077a1bc6ff Mon Sep 17 00:00:00 2001
From: Bodong Wang <bodong@mellanox.com>
Date: Thu, 1 Dec 2016 13:43:15 +0200
Subject: IB/uverbs: Extend modify_qp and support packet pacing

An new uverbs command ib_uverbs_ex_modify_qp is added to support more QP
attributes. User driver should choose to call the legacy/extended API
based on input mask.

IB_USER_LAST_QP_ATTR_MASK is added to indicated the maximum bit position
which supports legacy ib_uverbs_modify_qp.
IB_USER_LEGACY_LAST_QP_ATTR_MASK indicates the maximum bit position
which supports ib_uverbs_ex_modify_qp, the value of this mask should be
updated if new mask is added later.

Along with this change, rate_limit is supported by the extended command,
user driver could use it to control packet packing.

Signed-off-by: Bodong Wang <bodong@mellanox.com>
Reviewed-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/uverbs.h      |   1 +
 drivers/infiniband/core/uverbs_cmd.c  | 192 ++++++++++++++++++++++------------
 drivers/infiniband/core/uverbs_main.c |   1 +
 3 files changed, 125 insertions(+), 69 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/core/uverbs.h b/drivers/infiniband/core/uverbs.h
index df26a741cda6..455034ac994e 100644
--- a/drivers/infiniband/core/uverbs.h
+++ b/drivers/infiniband/core/uverbs.h
@@ -289,5 +289,6 @@ IB_UVERBS_DECLARE_EX_CMD(modify_wq);
 IB_UVERBS_DECLARE_EX_CMD(destroy_wq);
 IB_UVERBS_DECLARE_EX_CMD(create_rwq_ind_table);
 IB_UVERBS_DECLARE_EX_CMD(destroy_rwq_ind_table);
+IB_UVERBS_DECLARE_EX_CMD(modify_qp);
 
 #endif /* UVERBS_H */
diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c
index cdb935ddab69..09b649159e6c 100644
--- a/drivers/infiniband/core/uverbs_cmd.c
+++ b/drivers/infiniband/core/uverbs_cmd.c
@@ -2328,96 +2328,88 @@ static int modify_qp_mask(enum ib_qp_type qp_type, int mask)
 	}
 }
 
-ssize_t ib_uverbs_modify_qp(struct ib_uverbs_file *file,
-			    struct ib_device *ib_dev,
-			    const char __user *buf, int in_len,
-			    int out_len)
+static int modify_qp(struct ib_uverbs_file *file,
+		     struct ib_uverbs_ex_modify_qp *cmd, struct ib_udata *udata)
 {
-	struct ib_uverbs_modify_qp cmd;
-	struct ib_udata            udata;
-	struct ib_qp              *qp;
-	struct ib_qp_attr         *attr;
-	int                        ret;
-
-	if (copy_from_user(&cmd, buf, sizeof cmd))
-		return -EFAULT;
-
-	INIT_UDATA(&udata, buf + sizeof cmd, NULL, in_len - sizeof cmd,
-		   out_len);
+	struct ib_qp_attr *attr;
+	struct ib_qp *qp;
+	int ret;
 
 	attr = kmalloc(sizeof *attr, GFP_KERNEL);
 	if (!attr)
 		return -ENOMEM;
 
-	qp = idr_read_qp(cmd.qp_handle, file->ucontext);
+	qp = idr_read_qp(cmd->base.qp_handle, file->ucontext);
 	if (!qp) {
 		ret = -EINVAL;
 		goto out;
 	}
 
-	attr->qp_state 		  = cmd.qp_state;
-	attr->cur_qp_state 	  = cmd.cur_qp_state;
-	attr->path_mtu 		  = cmd.path_mtu;
-	attr->path_mig_state 	  = cmd.path_mig_state;
-	attr->qkey 		  = cmd.qkey;
-	attr->rq_psn 		  = cmd.rq_psn;
-	attr->sq_psn 		  = cmd.sq_psn;
-	attr->dest_qp_num 	  = cmd.dest_qp_num;
-	attr->qp_access_flags 	  = cmd.qp_access_flags;
-	attr->pkey_index 	  = cmd.pkey_index;
-	attr->alt_pkey_index 	  = cmd.alt_pkey_index;
-	attr->en_sqd_async_notify = cmd.en_sqd_async_notify;
-	attr->max_rd_atomic 	  = cmd.max_rd_atomic;
-	attr->max_dest_rd_atomic  = cmd.max_dest_rd_atomic;
-	attr->min_rnr_timer 	  = cmd.min_rnr_timer;
-	attr->port_num 		  = cmd.port_num;
-	attr->timeout 		  = cmd.timeout;
-	attr->retry_cnt 	  = cmd.retry_cnt;
-	attr->rnr_retry 	  = cmd.rnr_retry;
-	attr->alt_port_num 	  = cmd.alt_port_num;
-	attr->alt_timeout 	  = cmd.alt_timeout;
-
-	memcpy(attr->ah_attr.grh.dgid.raw, cmd.dest.dgid, 16);
-	attr->ah_attr.grh.flow_label        = cmd.dest.flow_label;
-	attr->ah_attr.grh.sgid_index        = cmd.dest.sgid_index;
-	attr->ah_attr.grh.hop_limit         = cmd.dest.hop_limit;
-	attr->ah_attr.grh.traffic_class     = cmd.dest.traffic_class;
-	attr->ah_attr.dlid 	    	    = cmd.dest.dlid;
-	attr->ah_attr.sl   	    	    = cmd.dest.sl;
-	attr->ah_attr.src_path_bits 	    = cmd.dest.src_path_bits;
-	attr->ah_attr.static_rate   	    = cmd.dest.static_rate;
-	attr->ah_attr.ah_flags 	    	    = cmd.dest.is_global ? IB_AH_GRH : 0;
-	attr->ah_attr.port_num 	    	    = cmd.dest.port_num;
-
-	memcpy(attr->alt_ah_attr.grh.dgid.raw, cmd.alt_dest.dgid, 16);
-	attr->alt_ah_attr.grh.flow_label    = cmd.alt_dest.flow_label;
-	attr->alt_ah_attr.grh.sgid_index    = cmd.alt_dest.sgid_index;
-	attr->alt_ah_attr.grh.hop_limit     = cmd.alt_dest.hop_limit;
-	attr->alt_ah_attr.grh.traffic_class = cmd.alt_dest.traffic_class;
-	attr->alt_ah_attr.dlid 	    	    = cmd.alt_dest.dlid;
-	attr->alt_ah_attr.sl   	    	    = cmd.alt_dest.sl;
-	attr->alt_ah_attr.src_path_bits     = cmd.alt_dest.src_path_bits;
-	attr->alt_ah_attr.static_rate       = cmd.alt_dest.static_rate;
-	attr->alt_ah_attr.ah_flags 	    = cmd.alt_dest.is_global ? IB_AH_GRH : 0;
-	attr->alt_ah_attr.port_num 	    = cmd.alt_dest.port_num;
+	attr->qp_state		  = cmd->base.qp_state;
+	attr->cur_qp_state	  = cmd->base.cur_qp_state;
+	attr->path_mtu		  = cmd->base.path_mtu;
+	attr->path_mig_state	  = cmd->base.path_mig_state;
+	attr->qkey		  = cmd->base.qkey;
+	attr->rq_psn		  = cmd->base.rq_psn;
+	attr->sq_psn		  = cmd->base.sq_psn;
+	attr->dest_qp_num	  = cmd->base.dest_qp_num;
+	attr->qp_access_flags	  = cmd->base.qp_access_flags;
+	attr->pkey_index	  = cmd->base.pkey_index;
+	attr->alt_pkey_index	  = cmd->base.alt_pkey_index;
+	attr->en_sqd_async_notify = cmd->base.en_sqd_async_notify;
+	attr->max_rd_atomic	  = cmd->base.max_rd_atomic;
+	attr->max_dest_rd_atomic  = cmd->base.max_dest_rd_atomic;
+	attr->min_rnr_timer	  = cmd->base.min_rnr_timer;
+	attr->port_num		  = cmd->base.port_num;
+	attr->timeout		  = cmd->base.timeout;
+	attr->retry_cnt		  = cmd->base.retry_cnt;
+	attr->rnr_retry		  = cmd->base.rnr_retry;
+	attr->alt_port_num	  = cmd->base.alt_port_num;
+	attr->alt_timeout	  = cmd->base.alt_timeout;
+	attr->rate_limit	  = cmd->rate_limit;
+
+	memcpy(attr->ah_attr.grh.dgid.raw, cmd->base.dest.dgid, 16);
+	attr->ah_attr.grh.flow_label	= cmd->base.dest.flow_label;
+	attr->ah_attr.grh.sgid_index	= cmd->base.dest.sgid_index;
+	attr->ah_attr.grh.hop_limit	= cmd->base.dest.hop_limit;
+	attr->ah_attr.grh.traffic_class	= cmd->base.dest.traffic_class;
+	attr->ah_attr.dlid		= cmd->base.dest.dlid;
+	attr->ah_attr.sl		= cmd->base.dest.sl;
+	attr->ah_attr.src_path_bits	= cmd->base.dest.src_path_bits;
+	attr->ah_attr.static_rate	= cmd->base.dest.static_rate;
+	attr->ah_attr.ah_flags		= cmd->base.dest.is_global ?
+					  IB_AH_GRH : 0;
+	attr->ah_attr.port_num		= cmd->base.dest.port_num;
+
+	memcpy(attr->alt_ah_attr.grh.dgid.raw, cmd->base.alt_dest.dgid, 16);
+	attr->alt_ah_attr.grh.flow_label    = cmd->base.alt_dest.flow_label;
+	attr->alt_ah_attr.grh.sgid_index    = cmd->base.alt_dest.sgid_index;
+	attr->alt_ah_attr.grh.hop_limit     = cmd->base.alt_dest.hop_limit;
+	attr->alt_ah_attr.grh.traffic_class = cmd->base.alt_dest.traffic_class;
+	attr->alt_ah_attr.dlid		    = cmd->base.alt_dest.dlid;
+	attr->alt_ah_attr.sl		    = cmd->base.alt_dest.sl;
+	attr->alt_ah_attr.src_path_bits	    = cmd->base.alt_dest.src_path_bits;
+	attr->alt_ah_attr.static_rate	    = cmd->base.alt_dest.static_rate;
+	attr->alt_ah_attr.ah_flags	    = cmd->base.alt_dest.is_global ?
+					      IB_AH_GRH : 0;
+	attr->alt_ah_attr.port_num	    = cmd->base.alt_dest.port_num;
 
 	if (qp->real_qp == qp) {
-		if (cmd.attr_mask & IB_QP_AV) {
+		if (cmd->base.attr_mask & IB_QP_AV) {
 			ret = ib_resolve_eth_dmac(qp->device, &attr->ah_attr);
 			if (ret)
 				goto release_qp;
 		}
 		ret = qp->device->modify_qp(qp, attr,
-			modify_qp_mask(qp->qp_type, cmd.attr_mask), &udata);
+					    modify_qp_mask(qp->qp_type,
+							   cmd->base.attr_mask),
+					    udata);
 	} else {
-		ret = ib_modify_qp(qp, attr, modify_qp_mask(qp->qp_type, cmd.attr_mask));
+		ret = ib_modify_qp(qp, attr,
+				   modify_qp_mask(qp->qp_type,
+						  cmd->base.attr_mask));
 	}
 
-	if (ret)
-		goto release_qp;
-
-	ret = in_len;
-
 release_qp:
 	put_qp_read(qp);
 
@@ -2427,6 +2419,68 @@ out:
 	return ret;
 }
 
+ssize_t ib_uverbs_modify_qp(struct ib_uverbs_file *file,
+			    struct ib_device *ib_dev,
+			    const char __user *buf, int in_len,
+			    int out_len)
+{
+	struct ib_uverbs_ex_modify_qp cmd = {};
+	struct ib_udata udata;
+	int ret;
+
+	if (copy_from_user(&cmd.base, buf, sizeof(cmd.base)))
+		return -EFAULT;
+
+	if (cmd.base.attr_mask &
+	    ~((IB_USER_LEGACY_LAST_QP_ATTR_MASK << 1) - 1))
+		return -EOPNOTSUPP;
+
+	INIT_UDATA(&udata, buf + sizeof(cmd.base), NULL,
+		   in_len - sizeof(cmd.base), out_len);
+
+	ret = modify_qp(file, &cmd, &udata);
+	if (ret)
+		return ret;
+
+	return in_len;
+}
+
+int ib_uverbs_ex_modify_qp(struct ib_uverbs_file *file,
+			   struct ib_device *ib_dev,
+			   struct ib_udata *ucore,
+			   struct ib_udata *uhw)
+{
+	struct ib_uverbs_ex_modify_qp cmd = {};
+	int ret;
+
+	/*
+	 * Last bit is reserved for extending the attr_mask by
+	 * using another field.
+	 */
+	BUILD_BUG_ON(IB_USER_LAST_QP_ATTR_MASK == (1 << 31));
+
+	if (ucore->inlen < sizeof(cmd.base))
+		return -EINVAL;
+
+	ret = ib_copy_from_udata(&cmd, ucore, min(sizeof(cmd), ucore->inlen));
+	if (ret)
+		return ret;
+
+	if (cmd.base.attr_mask &
+	    ~((IB_USER_LAST_QP_ATTR_MASK << 1) - 1))
+		return -EOPNOTSUPP;
+
+	if (ucore->inlen > sizeof(cmd)) {
+		if (ib_is_udata_cleared(ucore, sizeof(cmd),
+					ucore->inlen - sizeof(cmd)))
+			return -EOPNOTSUPP;
+	}
+
+	ret = modify_qp(file, &cmd, uhw);
+
+	return ret;
+}
+
 ssize_t ib_uverbs_destroy_qp(struct ib_uverbs_file *file,
 			     struct ib_device *ib_dev,
 			     const char __user *buf, int in_len,
diff --git a/drivers/infiniband/core/uverbs_main.c b/drivers/infiniband/core/uverbs_main.c
index 0012fa58c105..80839c31c887 100644
--- a/drivers/infiniband/core/uverbs_main.c
+++ b/drivers/infiniband/core/uverbs_main.c
@@ -137,6 +137,7 @@ static int (*uverbs_ex_cmd_table[])(struct ib_uverbs_file *file,
 	[IB_USER_VERBS_EX_CMD_DESTROY_WQ]       = ib_uverbs_ex_destroy_wq,
 	[IB_USER_VERBS_EX_CMD_CREATE_RWQ_IND_TBL] = ib_uverbs_ex_create_rwq_ind_table,
 	[IB_USER_VERBS_EX_CMD_DESTROY_RWQ_IND_TBL] = ib_uverbs_ex_destroy_rwq_ind_table,
+	[IB_USER_VERBS_EX_CMD_MODIFY_QP]        = ib_uverbs_ex_modify_qp,
 };
 
 static void ib_uverbs_add_one(struct ib_device *device);
-- 
cgit v1.2.1


From 7d29f349a4b9dcf5bc9dcc05630d6a7f6b6b3ccd Mon Sep 17 00:00:00 2001
From: Bodong Wang <bodong@mellanox.com>
Date: Thu, 1 Dec 2016 13:43:16 +0200
Subject: IB/mlx5: Properly adjust rate limit on QP state transitions

- Add MODIFY_QP_EX CMD to extend modify_qp.
- Rate limit will be updated in the following state transactions: RTR2RTS,
  RTS2RTS. The limit will be removed when SQ is in RST and ERR state.

Signed-off-by: Bodong Wang <bodong@mellanox.com>
Reviewed-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/main.c    |  3 +-
 drivers/infiniband/hw/mlx5/mlx5_ib.h |  1 +
 drivers/infiniband/hw/mlx5/qp.c      | 74 ++++++++++++++++++++++++++++++++----
 3 files changed, 69 insertions(+), 9 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index 6c194000903d..cda541ced141 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -3105,7 +3105,8 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev)
 	dev->ib_dev.uverbs_ex_cmd_mask =
 		(1ull << IB_USER_VERBS_EX_CMD_QUERY_DEVICE)	|
 		(1ull << IB_USER_VERBS_EX_CMD_CREATE_CQ)	|
-		(1ull << IB_USER_VERBS_EX_CMD_CREATE_QP);
+		(1ull << IB_USER_VERBS_EX_CMD_CREATE_QP)	|
+		(1ull << IB_USER_VERBS_EX_CMD_MODIFY_QP);
 
 	dev->ib_dev.query_device	= mlx5_ib_query_device;
 	dev->ib_dev.query_port		= mlx5_ib_query_port;
diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h
index df3d6af3f683..ab8961cc8bca 100644
--- a/drivers/infiniband/hw/mlx5/mlx5_ib.h
+++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h
@@ -389,6 +389,7 @@ struct mlx5_ib_qp {
 	struct list_head	qps_list;
 	struct list_head	cq_recv_list;
 	struct list_head	cq_send_list;
+	u32			rate_limit;
 };
 
 struct mlx5_ib_cq_buf {
diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c
index aa27688f5ae9..a69524fb6032 100644
--- a/drivers/infiniband/hw/mlx5/qp.c
+++ b/drivers/infiniband/hw/mlx5/qp.c
@@ -78,12 +78,14 @@ struct mlx5_wqe_eth_pad {
 
 enum raw_qp_set_mask_map {
 	MLX5_RAW_QP_MOD_SET_RQ_Q_CTR_ID		= 1UL << 0,
+	MLX5_RAW_QP_RATE_LIMIT			= 1UL << 1,
 };
 
 struct mlx5_modify_raw_qp_param {
 	u16 operation;
 
 	u32 set_mask; /* raw_qp_set_mask_map */
+	u32 rate_limit;
 	u8 rq_q_ctr_id;
 };
 
@@ -2470,8 +2472,14 @@ out:
 }
 
 static int modify_raw_packet_qp_sq(struct mlx5_core_dev *dev,
-				   struct mlx5_ib_sq *sq, int new_state)
+				   struct mlx5_ib_sq *sq,
+				   int new_state,
+				   const struct mlx5_modify_raw_qp_param *raw_qp_param)
 {
+	struct mlx5_ib_qp *ibqp = sq->base.container_mibqp;
+	u32 old_rate = ibqp->rate_limit;
+	u32 new_rate = old_rate;
+	u16 rl_index = 0;
 	void *in;
 	void *sqc;
 	int inlen;
@@ -2487,10 +2495,44 @@ static int modify_raw_packet_qp_sq(struct mlx5_core_dev *dev,
 	sqc = MLX5_ADDR_OF(modify_sq_in, in, ctx);
 	MLX5_SET(sqc, sqc, state, new_state);
 
+	if (raw_qp_param->set_mask & MLX5_RAW_QP_RATE_LIMIT) {
+		if (new_state != MLX5_SQC_STATE_RDY)
+			pr_warn("%s: Rate limit can only be changed when SQ is moving to RDY\n",
+				__func__);
+		else
+			new_rate = raw_qp_param->rate_limit;
+	}
+
+	if (old_rate != new_rate) {
+		if (new_rate) {
+			err = mlx5_rl_add_rate(dev, new_rate, &rl_index);
+			if (err) {
+				pr_err("Failed configuring rate %u: %d\n",
+				       new_rate, err);
+				goto out;
+			}
+		}
+
+		MLX5_SET64(modify_sq_in, in, modify_bitmask, 1);
+		MLX5_SET(sqc, sqc, packet_pacing_rate_limit_index, rl_index);
+	}
+
 	err = mlx5_core_modify_sq(dev, sq->base.mqp.qpn, in, inlen);
-	if (err)
+	if (err) {
+		/* Remove new rate from table if failed */
+		if (new_rate &&
+		    old_rate != new_rate)
+			mlx5_rl_remove_rate(dev, new_rate);
 		goto out;
+	}
 
+	/* Only remove the old rate after new rate was set */
+	if ((old_rate &&
+	    (old_rate != new_rate)) ||
+	    (new_state != MLX5_SQC_STATE_RDY))
+		mlx5_rl_remove_rate(dev, old_rate);
+
+	ibqp->rate_limit = new_rate;
 	sq->state = new_state;
 
 out:
@@ -2505,6 +2547,8 @@ static int modify_raw_packet_qp(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp,
 	struct mlx5_ib_raw_packet_qp *raw_packet_qp = &qp->raw_packet_qp;
 	struct mlx5_ib_rq *rq = &raw_packet_qp->rq;
 	struct mlx5_ib_sq *sq = &raw_packet_qp->sq;
+	int modify_rq = !!qp->rq.wqe_cnt;
+	int modify_sq = !!qp->sq.wqe_cnt;
 	int rq_state;
 	int sq_state;
 	int err;
@@ -2522,10 +2566,18 @@ static int modify_raw_packet_qp(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp,
 		rq_state = MLX5_RQC_STATE_RST;
 		sq_state = MLX5_SQC_STATE_RST;
 		break;
-	case MLX5_CMD_OP_INIT2INIT_QP:
-	case MLX5_CMD_OP_INIT2RTR_QP:
 	case MLX5_CMD_OP_RTR2RTS_QP:
 	case MLX5_CMD_OP_RTS2RTS_QP:
+		if (raw_qp_param->set_mask ==
+		    MLX5_RAW_QP_RATE_LIMIT) {
+			modify_rq = 0;
+			sq_state = sq->state;
+		} else {
+			return raw_qp_param->set_mask ? -EINVAL : 0;
+		}
+		break;
+	case MLX5_CMD_OP_INIT2INIT_QP:
+	case MLX5_CMD_OP_INIT2RTR_QP:
 		if (raw_qp_param->set_mask)
 			return -EINVAL;
 		else
@@ -2535,13 +2587,13 @@ static int modify_raw_packet_qp(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp,
 		return -EINVAL;
 	}
 
-	if (qp->rq.wqe_cnt) {
-		err = modify_raw_packet_qp_rq(dev, rq, rq_state, raw_qp_param);
+	if (modify_rq) {
+		err =  modify_raw_packet_qp_rq(dev, rq, rq_state, raw_qp_param);
 		if (err)
 			return err;
 	}
 
-	if (qp->sq.wqe_cnt) {
+	if (modify_sq) {
 		if (tx_affinity) {
 			err = modify_raw_packet_tx_affinity(dev->mdev, sq,
 							    tx_affinity);
@@ -2549,7 +2601,7 @@ static int modify_raw_packet_qp(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp,
 				return err;
 		}
 
-		return modify_raw_packet_qp_sq(dev->mdev, sq, sq_state);
+		return modify_raw_packet_qp_sq(dev->mdev, sq, sq_state, raw_qp_param);
 	}
 
 	return 0;
@@ -2804,6 +2856,12 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp,
 			raw_qp_param.rq_q_ctr_id = mibport->q_cnt_id;
 			raw_qp_param.set_mask |= MLX5_RAW_QP_MOD_SET_RQ_Q_CTR_ID;
 		}
+
+		if (attr_mask & IB_QP_RATE_LIMIT) {
+			raw_qp_param.rate_limit = attr->rate_limit;
+			raw_qp_param.set_mask |= MLX5_RAW_QP_RATE_LIMIT;
+		}
+
 		err = modify_raw_packet_qp(dev, qp, &raw_qp_param, tx_affinity);
 	} else {
 		err = mlx5_core_qp_modify(dev->mdev, op, optpar, context,
-- 
cgit v1.2.1


From 9fa240bbfc4200b080c8fad12579659c2c2f36b5 Mon Sep 17 00:00:00 2001
From: Hal Rosenstock <hal@dev.mellanox.co.il>
Date: Tue, 18 Oct 2016 13:20:29 -0400
Subject: IB/mad: Eliminate redundant SM class version defines for OPA

and rename class version define to indicate SM rather than SMP or SMI

Signed-off-by: Hal Rosenstock <hal@mellanox.com>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/mad.c    |  4 ++--
 drivers/infiniband/hw/hfi1/mad.c | 12 ++++++------
 2 files changed, 8 insertions(+), 8 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/core/mad.c b/drivers/infiniband/core/mad.c
index 40cbd6bdb73b..bc2a9c2caa3c 100644
--- a/drivers/infiniband/core/mad.c
+++ b/drivers/infiniband/core/mad.c
@@ -769,7 +769,7 @@ static int handle_outgoing_dr_smp(struct ib_mad_agent_private *mad_agent_priv,
 	 * If we are at the start of the LID routed part, don't update the
 	 * hop_ptr or hop_cnt.  See section 14.2.2, Vol 1 IB spec.
 	 */
-	if (opa && smp->class_version == OPA_SMP_CLASS_VERSION) {
+	if (opa && smp->class_version == OPA_SM_CLASS_VERSION) {
 		u32 opa_drslid;
 
 		if ((opa_get_smp_direction(opa_smp)
@@ -2167,7 +2167,7 @@ handle_smi(struct ib_mad_port_private *port_priv,
 	struct ib_mad_hdr *mad_hdr = (struct ib_mad_hdr *)recv->mad;
 
 	if (opa && mad_hdr->base_version == OPA_MGMT_BASE_VERSION &&
-	    mad_hdr->class_version == OPA_SMI_CLASS_VERSION)
+	    mad_hdr->class_version == OPA_SM_CLASS_VERSION)
 		return handle_opa_smi(port_priv, qp_info, wc, port_num, recv,
 				      response);
 
diff --git a/drivers/infiniband/hw/hfi1/mad.c b/drivers/infiniband/hw/hfi1/mad.c
index 9487c9bb8920..320e4daa5fc6 100644
--- a/drivers/infiniband/hw/hfi1/mad.c
+++ b/drivers/infiniband/hw/hfi1/mad.c
@@ -128,7 +128,7 @@ static void send_trap(struct hfi1_ibport *ibp, void *data, unsigned len)
 	smp = send_buf->mad;
 	smp->base_version = OPA_MGMT_BASE_VERSION;
 	smp->mgmt_class = IB_MGMT_CLASS_SUBN_LID_ROUTED;
-	smp->class_version = OPA_SMI_CLASS_VERSION;
+	smp->class_version = OPA_SM_CLASS_VERSION;
 	smp->method = IB_MGMT_METHOD_TRAP;
 	ibp->rvp.tid++;
 	smp->tid = cpu_to_be64(ibp->rvp.tid);
@@ -343,7 +343,7 @@ static int __subn_get_opa_nodeinfo(struct opa_smp *smp, u32 am, u8 *data,
 
 	ni->port_guid = cpu_to_be64(dd->pport[pidx].guid);
 	ni->base_version = OPA_MGMT_BASE_VERSION;
-	ni->class_version = OPA_SMI_CLASS_VERSION;
+	ni->class_version = OPA_SM_CLASS_VERSION;
 	ni->node_type = 1;     /* channel adapter */
 	ni->num_ports = ibdev->phys_port_cnt;
 	/* This is already in network order */
@@ -379,7 +379,7 @@ static int subn_get_nodeinfo(struct ib_smp *smp, struct ib_device *ibdev,
 		nip->port_guid = cpu_to_be64(dd->pport[pidx].guid);
 
 	nip->base_version = OPA_MGMT_BASE_VERSION;
-	nip->class_version = OPA_SMI_CLASS_VERSION;
+	nip->class_version = OPA_SM_CLASS_VERSION;
 	nip->node_type = 1;     /* channel adapter */
 	nip->num_ports = ibdev->phys_port_cnt;
 	/* This is already in network order */
@@ -2302,7 +2302,7 @@ static int pma_get_opa_classportinfo(struct opa_pma_mad *pmp,
 		pmp->mad_hdr.status |= IB_SMP_INVALID_FIELD;
 
 	p->base_version = OPA_MGMT_BASE_VERSION;
-	p->class_version = OPA_SMI_CLASS_VERSION;
+	p->class_version = OPA_SM_CLASS_VERSION;
 	/*
 	 * Expected response time is 4.096 usec. * 2^18 == 1.073741824 sec.
 	 */
@@ -4022,7 +4022,7 @@ static int process_subn_opa(struct ib_device *ibdev, int mad_flags,
 
 	am = be32_to_cpu(smp->attr_mod);
 	attr_id = smp->attr_id;
-	if (smp->class_version != OPA_SMI_CLASS_VERSION) {
+	if (smp->class_version != OPA_SM_CLASS_VERSION) {
 		smp->status |= IB_SMP_UNSUP_VERSION;
 		ret = reply((struct ib_mad_hdr *)smp);
 		return ret;
@@ -4232,7 +4232,7 @@ static int process_perf_opa(struct ib_device *ibdev, u8 port,
 
 	*out_mad = *in_mad;
 
-	if (pmp->mad_hdr.class_version != OPA_SMI_CLASS_VERSION) {
+	if (pmp->mad_hdr.class_version != OPA_SM_CLASS_VERSION) {
 		pmp->mad_hdr.status |= IB_SMP_UNSUP_VERSION;
 		return reply((struct ib_mad_hdr *)pmp);
 	}
-- 
cgit v1.2.1


From ea7ef2accdeaf825165cccd60b848765677bf1f2 Mon Sep 17 00:00:00 2001
From: Colin Ian King <colin.king@canonical.com>
Date: Tue, 18 Oct 2016 19:39:28 +0100
Subject: qedr: return -EINVAL if pd is null and avoid null ptr dereference

Currently, if pd is null then we hit a null pointer derference
on accessing pd->pd_id.  Instead of just printing an error message
we should also return -EINVAL immediately.

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/qedr/verbs.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/qedr/verbs.c b/drivers/infiniband/hw/qedr/verbs.c
index a61514296767..b2a0eb8f73d2 100644
--- a/drivers/infiniband/hw/qedr/verbs.c
+++ b/drivers/infiniband/hw/qedr/verbs.c
@@ -511,8 +511,10 @@ int qedr_dealloc_pd(struct ib_pd *ibpd)
 	struct qedr_dev *dev = get_qedr_dev(ibpd->device);
 	struct qedr_pd *pd = get_qedr_pd(ibpd);
 
-	if (!pd)
+	if (!pd) {
 		pr_err("Invalid PD received in dealloc_pd\n");
+		return -EINVAL;
+	}
 
 	DP_DEBUG(dev, QEDR_MSG_INIT, "Deallocating PD %d\n", pd->pd_id);
 	dev->ops->rdma_dealloc_pd(dev->rdma_ctx, pd->pd_id);
-- 
cgit v1.2.1


From 181d80151f9c6ff3c765c1bd2e4e200ada23c2f3 Mon Sep 17 00:00:00 2001
From: Wei Yongjun <weiyongjun1@huawei.com>
Date: Fri, 28 Oct 2016 16:33:47 +0000
Subject: qedr: Fix possible memory leak in qedr_create_qp()

'qp' is malloced in qedr_create_qp() and should be freed before leaving
from the error handling cases, otherwise it will cause memory leak.

Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Acked-by: Ram Amrani <Ram.Amrani@cavium.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/qedr/verbs.c | 12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/qedr/verbs.c b/drivers/infiniband/hw/qedr/verbs.c
index b2a0eb8f73d2..8c6c5ccaf57a 100644
--- a/drivers/infiniband/hw/qedr/verbs.c
+++ b/drivers/infiniband/hw/qedr/verbs.c
@@ -1479,6 +1479,7 @@ struct ib_qp *qedr_create_qp(struct ib_pd *ibpd,
 	struct qedr_ucontext *ctx = NULL;
 	struct qedr_create_qp_ureq ureq;
 	struct qedr_qp *qp;
+	struct ib_qp *ibqp;
 	int rc = 0;
 
 	DP_DEBUG(dev, QEDR_MSG_QP, "create qp: called from %s, pd=%p\n",
@@ -1488,13 +1489,13 @@ struct ib_qp *qedr_create_qp(struct ib_pd *ibpd,
 	if (rc)
 		return ERR_PTR(rc);
 
+	if (attrs->srq)
+		return ERR_PTR(-EINVAL);
+
 	qp = kzalloc(sizeof(*qp), GFP_KERNEL);
 	if (!qp)
 		return ERR_PTR(-ENOMEM);
 
-	if (attrs->srq)
-		return ERR_PTR(-EINVAL);
-
 	DP_DEBUG(dev, QEDR_MSG_QP,
 		 "create qp: sq_cq=%p, sq_icid=%d, rq_cq=%p, rq_icid=%d\n",
 		 get_qedr_cq(attrs->send_cq),
@@ -1510,7 +1511,10 @@ struct ib_qp *qedr_create_qp(struct ib_pd *ibpd,
 			       "create qp: unexpected udata when creating GSI QP\n");
 			goto err0;
 		}
-		return qedr_create_gsi_qp(dev, attrs, qp);
+		ibqp = qedr_create_gsi_qp(dev, attrs, qp);
+		if (IS_ERR(ibqp))
+			kfree(qp);
+		return ibqp;
 	}
 
 	memset(&in_params, 0, sizeof(in_params));
-- 
cgit v1.2.1


From aafec388a1b7fc533a81c03b4a81c6e5f3e5688f Mon Sep 17 00:00:00 2001
From: Wei Yongjun <weiyongjun1@huawei.com>
Date: Sat, 29 Oct 2016 16:19:53 +0000
Subject: qedr: Use list_move_tail instead of list_del/list_add_tail

Using list_move_tail() instead of list_del() + list_add_tail().

Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Reviewed-by: Leon Romanovsky <leonro@mellanox.com>
Acked-by: Ram Amrani <Ram.Amrani@cavium.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/qedr/verbs.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/qedr/verbs.c b/drivers/infiniband/hw/qedr/verbs.c
index 8c6c5ccaf57a..18e2f60a9d4f 100644
--- a/drivers/infiniband/hw/qedr/verbs.c
+++ b/drivers/infiniband/hw/qedr/verbs.c
@@ -2419,8 +2419,7 @@ static void handle_completed_mrs(struct qedr_dev *dev, struct mr_info *info)
 		 */
 		pbl = list_first_entry(&info->inuse_pbl_list,
 				       struct qedr_pbl, list_entry);
-		list_del(&pbl->list_entry);
-		list_add_tail(&pbl->list_entry, &info->free_pbl_list);
+		list_move_tail(&pbl->list_entry, &info->free_pbl_list);
 		info->completed_handled++;
 	}
 }
-- 
cgit v1.2.1


From aecb66b2b01a87b4b723267b9193c8f00d661c1f Mon Sep 17 00:00:00 2001
From: Wei Yongjun <weiyongjun1@huawei.com>
Date: Wed, 2 Nov 2016 13:11:32 +0000
Subject: qedr: remove pointless NULL check in qedr_post_send()

Remove pointless NULL check for 'wr' in qedr_post_send().

Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Acked-by: Ram Amrani <Ram.Amrani@cavium.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/qedr/verbs.c | 5 -----
 1 file changed, 5 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/qedr/verbs.c b/drivers/infiniband/hw/qedr/verbs.c
index 18e2f60a9d4f..12f35150f32f 100644
--- a/drivers/infiniband/hw/qedr/verbs.c
+++ b/drivers/infiniband/hw/qedr/verbs.c
@@ -2986,11 +2986,6 @@ int qedr_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 		return -EINVAL;
 	}
 
-	if (!wr) {
-		DP_ERR(dev, "Got an empty post send.\n");
-		return -EINVAL;
-	}
-
 	while (wr) {
 		rc = __qedr_post_send(ibqp, wr, bad_wr);
 		if (rc)
-- 
cgit v1.2.1


From 77a5db13153906a7e00740b10b2730e53385c5a8 Mon Sep 17 00:00:00 2001
From: Steve Wise <swise@opengridcomputing.com>
Date: Wed, 26 Oct 2016 12:36:40 -0700
Subject: rdma_cm: add rdma_reject_msg() helper function

rdma_reject_msg() returns a pointer to a string message associated with
the transport reject reason codes.

Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Reviewed-by: Bart Van Assche <bart.vanassche@sandisk.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/cm.c   | 48 ++++++++++++++++++++++++++++++++++++++++++
 drivers/infiniband/core/cma.c  | 14 ++++++++++++
 drivers/infiniband/core/iwcm.c | 21 ++++++++++++++++++
 3 files changed, 83 insertions(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/core/cm.c b/drivers/infiniband/core/cm.c
index c99525512b34..6c64d0ce64b3 100644
--- a/drivers/infiniband/core/cm.c
+++ b/drivers/infiniband/core/cm.c
@@ -57,6 +57,54 @@ MODULE_AUTHOR("Sean Hefty");
 MODULE_DESCRIPTION("InfiniBand CM");
 MODULE_LICENSE("Dual BSD/GPL");
 
+static const char * const ibcm_rej_reason_strs[] = {
+	[IB_CM_REJ_NO_QP]			= "no QP",
+	[IB_CM_REJ_NO_EEC]			= "no EEC",
+	[IB_CM_REJ_NO_RESOURCES]		= "no resources",
+	[IB_CM_REJ_TIMEOUT]			= "timeout",
+	[IB_CM_REJ_UNSUPPORTED]			= "unsupported",
+	[IB_CM_REJ_INVALID_COMM_ID]		= "invalid comm ID",
+	[IB_CM_REJ_INVALID_COMM_INSTANCE]	= "invalid comm instance",
+	[IB_CM_REJ_INVALID_SERVICE_ID]		= "invalid service ID",
+	[IB_CM_REJ_INVALID_TRANSPORT_TYPE]	= "invalid transport type",
+	[IB_CM_REJ_STALE_CONN]			= "stale conn",
+	[IB_CM_REJ_RDC_NOT_EXIST]		= "RDC not exist",
+	[IB_CM_REJ_INVALID_GID]			= "invalid GID",
+	[IB_CM_REJ_INVALID_LID]			= "invalid LID",
+	[IB_CM_REJ_INVALID_SL]			= "invalid SL",
+	[IB_CM_REJ_INVALID_TRAFFIC_CLASS]	= "invalid traffic class",
+	[IB_CM_REJ_INVALID_HOP_LIMIT]		= "invalid hop limit",
+	[IB_CM_REJ_INVALID_PACKET_RATE]		= "invalid packet rate",
+	[IB_CM_REJ_INVALID_ALT_GID]		= "invalid alt GID",
+	[IB_CM_REJ_INVALID_ALT_LID]		= "invalid alt LID",
+	[IB_CM_REJ_INVALID_ALT_SL]		= "invalid alt SL",
+	[IB_CM_REJ_INVALID_ALT_TRAFFIC_CLASS]	= "invalid alt traffic class",
+	[IB_CM_REJ_INVALID_ALT_HOP_LIMIT]	= "invalid alt hop limit",
+	[IB_CM_REJ_INVALID_ALT_PACKET_RATE]	= "invalid alt packet rate",
+	[IB_CM_REJ_PORT_CM_REDIRECT]		= "port CM redirect",
+	[IB_CM_REJ_PORT_REDIRECT]		= "port redirect",
+	[IB_CM_REJ_INVALID_MTU]			= "invalid MTU",
+	[IB_CM_REJ_INSUFFICIENT_RESP_RESOURCES]	= "insufficient resp resources",
+	[IB_CM_REJ_CONSUMER_DEFINED]		= "consumer defined",
+	[IB_CM_REJ_INVALID_RNR_RETRY]		= "invalid RNR retry",
+	[IB_CM_REJ_DUPLICATE_LOCAL_COMM_ID]	= "duplicate local comm ID",
+	[IB_CM_REJ_INVALID_CLASS_VERSION]	= "invalid class version",
+	[IB_CM_REJ_INVALID_FLOW_LABEL]		= "invalid flow label",
+	[IB_CM_REJ_INVALID_ALT_FLOW_LABEL]	= "invalid alt flow label",
+};
+
+const char *__attribute_const__ ibcm_reject_msg(int reason)
+{
+	size_t index = reason;
+
+	if (index < ARRAY_SIZE(ibcm_rej_reason_strs) &&
+	    ibcm_rej_reason_strs[index])
+		return ibcm_rej_reason_strs[index];
+	else
+		return "unrecognized reason";
+}
+EXPORT_SYMBOL(ibcm_reject_msg);
+
 static void cm_add_one(struct ib_device *device);
 static void cm_remove_one(struct ib_device *device, void *client_data);
 
diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c
index 36bf50ebb187..877e4dc9e382 100644
--- a/drivers/infiniband/core/cma.c
+++ b/drivers/infiniband/core/cma.c
@@ -101,6 +101,20 @@ const char *__attribute_const__ rdma_event_msg(enum rdma_cm_event_type event)
 }
 EXPORT_SYMBOL(rdma_event_msg);
 
+const char *__attribute_const__ rdma_reject_msg(struct rdma_cm_id *id,
+						int reason)
+{
+	if (rdma_ib_or_roce(id->device, id->port_num))
+		return ibcm_reject_msg(reason);
+
+	if (rdma_protocol_iwarp(id->device, id->port_num))
+		return iwcm_reject_msg(reason);
+
+	WARN_ON_ONCE(1);
+	return "unrecognized transport";
+}
+EXPORT_SYMBOL(rdma_reject_msg);
+
 static void cma_add_one(struct ib_device *device);
 static void cma_remove_one(struct ib_device *device, void *client_data);
 
diff --git a/drivers/infiniband/core/iwcm.c b/drivers/infiniband/core/iwcm.c
index 5495e22839a7..31661b5c1743 100644
--- a/drivers/infiniband/core/iwcm.c
+++ b/drivers/infiniband/core/iwcm.c
@@ -59,6 +59,27 @@ MODULE_AUTHOR("Tom Tucker");
 MODULE_DESCRIPTION("iWARP CM");
 MODULE_LICENSE("Dual BSD/GPL");
 
+static const char * const iwcm_rej_reason_strs[] = {
+	[ECONNRESET]			= "reset by remote host",
+	[ECONNREFUSED]			= "refused by remote application",
+	[ETIMEDOUT]			= "setup timeout",
+};
+
+const char *__attribute_const__ iwcm_reject_msg(int reason)
+{
+	size_t index;
+
+	/* iWARP uses negative errnos */
+	index = -reason;
+
+	if (index < ARRAY_SIZE(iwcm_rej_reason_strs) &&
+	    iwcm_rej_reason_strs[index])
+		return iwcm_rej_reason_strs[index];
+	else
+		return "unrecognized reason";
+}
+EXPORT_SYMBOL(iwcm_reject_msg);
+
 static struct ibnl_client_cbs iwcm_nl_cb_table[] = {
 	[RDMA_NL_IWPM_REG_PID] = {.dump = iwpm_register_pid_cb},
 	[RDMA_NL_IWPM_ADD_MAPPING] = {.dump = iwpm_add_mapping_cb},
-- 
cgit v1.2.1


From 5042a73d3e9de7bcc2a31adea08ee95bbce998dc Mon Sep 17 00:00:00 2001
From: Steve Wise <swise@opengridcomputing.com>
Date: Wed, 26 Oct 2016 12:36:47 -0700
Subject: rdma_cm: add rdma_is_consumer_reject() helper function

Return true if the peer consumer application rejected the
connection attempt.

Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Reviewed-by: Bart Van Assche <bart.vanassche@sandisk.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/cma.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c
index 877e4dc9e382..bd1d4fe47740 100644
--- a/drivers/infiniband/core/cma.c
+++ b/drivers/infiniband/core/cma.c
@@ -115,6 +115,19 @@ const char *__attribute_const__ rdma_reject_msg(struct rdma_cm_id *id,
 }
 EXPORT_SYMBOL(rdma_reject_msg);
 
+bool rdma_is_consumer_reject(struct rdma_cm_id *id, int reason)
+{
+	if (rdma_ib_or_roce(id->device, id->port_num))
+		return reason == IB_CM_REJ_CONSUMER_DEFINED;
+
+	if (rdma_protocol_iwarp(id->device, id->port_num))
+		return reason == -ECONNREFUSED;
+
+	WARN_ON_ONCE(1);
+	return false;
+}
+EXPORT_SYMBOL(rdma_is_consumer_reject);
+
 static void cma_add_one(struct ib_device *device);
 static void cma_remove_one(struct ib_device *device, void *client_data);
 
-- 
cgit v1.2.1


From 5f24410408fd093734ce758f2fe3a66fe543de22 Mon Sep 17 00:00:00 2001
From: Steve Wise <swise@opengridcomputing.com>
Date: Wed, 26 Oct 2016 12:36:47 -0700
Subject: rdma_cm: add rdma_consumer_reject_data helper function

rdma_consumer_reject_data() will return the private data pointer
and length if any is available.

Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/cma.c | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c
index bd1d4fe47740..8c30e3dedebe 100644
--- a/drivers/infiniband/core/cma.c
+++ b/drivers/infiniband/core/cma.c
@@ -128,6 +128,22 @@ bool rdma_is_consumer_reject(struct rdma_cm_id *id, int reason)
 }
 EXPORT_SYMBOL(rdma_is_consumer_reject);
 
+const void *rdma_consumer_reject_data(struct rdma_cm_id *id,
+				      struct rdma_cm_event *ev, u8 *data_len)
+{
+	const void *p;
+
+	if (rdma_is_consumer_reject(id, ev->status)) {
+		*data_len = ev->param.conn.private_data_len;
+		p = ev->param.conn.private_data;
+	} else {
+		*data_len = 0;
+		p = NULL;
+	}
+	return p;
+}
+EXPORT_SYMBOL(rdma_consumer_reject_data);
+
 static void cma_add_one(struct ib_device *device);
 static void cma_remove_one(struct ib_device *device, void *client_data);
 
-- 
cgit v1.2.1


From 97540bb90acfab268b256a58c3e51cd06b2d1654 Mon Sep 17 00:00:00 2001
From: Steve Wise <swise@opengridcomputing.com>
Date: Wed, 26 Oct 2016 12:36:47 -0700
Subject: ib_iser: log the connection reject message

Acked-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/ulp/iser/iser_verbs.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/ulp/iser/iser_verbs.c b/drivers/infiniband/ulp/iser/iser_verbs.c
index a4b791dfaa1d..8ae7a3beddb7 100644
--- a/drivers/infiniband/ulp/iser/iser_verbs.c
+++ b/drivers/infiniband/ulp/iser/iser_verbs.c
@@ -890,11 +890,14 @@ static int iser_cma_handler(struct rdma_cm_id *cma_id, struct rdma_cm_event *eve
 	case RDMA_CM_EVENT_ESTABLISHED:
 		iser_connected_handler(cma_id, event->param.conn.private_data);
 		break;
+	case RDMA_CM_EVENT_REJECTED:
+		iser_info("Connection rejected: %s\n",
+			 rdma_reject_msg(cma_id, event->status));
+		/* FALLTHROUGH */
 	case RDMA_CM_EVENT_ADDR_ERROR:
 	case RDMA_CM_EVENT_ROUTE_ERROR:
 	case RDMA_CM_EVENT_CONNECT_ERROR:
 	case RDMA_CM_EVENT_UNREACHABLE:
-	case RDMA_CM_EVENT_REJECTED:
 		iser_connect_error(cma_id);
 		break;
 	case RDMA_CM_EVENT_DISCONNECTED:
-- 
cgit v1.2.1


From 1e38a366ee86ac2b7a8110775cb3353649b18b70 Mon Sep 17 00:00:00 2001
From: Steve Wise <swise@opengridcomputing.com>
Date: Wed, 26 Oct 2016 12:36:48 -0700
Subject: ib_isert: log the connection reject message

Acked-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/ulp/isert/ib_isert.c | 2 ++
 1 file changed, 2 insertions(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/ulp/isert/ib_isert.c b/drivers/infiniband/ulp/isert/ib_isert.c
index 6dd43f63238e..37016f01b5a5 100644
--- a/drivers/infiniband/ulp/isert/ib_isert.c
+++ b/drivers/infiniband/ulp/isert/ib_isert.c
@@ -795,6 +795,8 @@ isert_cma_handler(struct rdma_cm_id *cma_id, struct rdma_cm_event *event)
 		 */
 		return 1;
 	case RDMA_CM_EVENT_REJECTED:       /* FALLTHRU */
+		isert_info("Connection rejected: %s\n",
+			   rdma_reject_msg(cma_id, event->status));
 	case RDMA_CM_EVENT_UNREACHABLE:    /* FALLTHRU */
 	case RDMA_CM_EVENT_CONNECT_ERROR:
 		ret = isert_connect_error(cma_id);
-- 
cgit v1.2.1


From 14ab8896f5d993c5f427504276e3c42ccb3cc354 Mon Sep 17 00:00:00 2001
From: Arnd Bergmann <arnd@arndb.de>
Date: Mon, 24 Oct 2016 22:48:21 +0200
Subject: IB/mlx5: avoid bogus -Wmaybe-uninitialized warning
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

We get a false-positive warning in linux-next for the mlx5 driver:

infiniband/hw/mlx5/mr.c: In function ‘mlx5_ib_reg_user_mr’:
infiniband/hw/mlx5/mr.c:1172:5: error: ‘order’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
infiniband/hw/mlx5/mr.c:1161:6: note: ‘order’ was declared here
infiniband/hw/mlx5/mr.c:1173:6: error: ‘ncont’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
infiniband/hw/mlx5/mr.c:1160:6: note: ‘ncont’ was declared here
infiniband/hw/mlx5/mr.c:1173:6: error: ‘page_shift’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
infiniband/hw/mlx5/mr.c:1158:6: note: ‘page_shift’ was declared here
infiniband/hw/mlx5/mr.c:1143:13: error: ‘npages’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
infiniband/hw/mlx5/mr.c:1159:6: note: ‘npages’ was declared here

I had a trivial workaround for gcc-5 or higher, but that didn't work
on gcc-4.9 unfortunately.

The only way I found to avoid the warnings for gcc-4.9, short of
initializing each of the arguments first was to change the calling
conventions to separate the error code from the umem pointer. This
avoids casting the error codes from one pointer to another incompatible
pointer, and lets gcc figure out when that the data is actually valid
whenever we return successfully.

Acked-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/mr.c | 39 +++++++++++++++++++++------------------
 1 file changed, 21 insertions(+), 18 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/mr.c b/drivers/infiniband/hw/mlx5/mr.c
index 5258ac870e6a..67985c69f9b9 100644
--- a/drivers/infiniband/hw/mlx5/mr.c
+++ b/drivers/infiniband/hw/mlx5/mr.c
@@ -844,30 +844,34 @@ static void prep_umr_unreg_wqe(struct mlx5_ib_dev *dev,
 	umrwr->mkey = key;
 }
 
-static struct ib_umem *mr_umem_get(struct ib_pd *pd, u64 start, u64 length,
-				   int access_flags, int *npages,
-				   int *page_shift, int *ncont, int *order)
+static int mr_umem_get(struct ib_pd *pd, u64 start, u64 length,
+		       int access_flags, struct ib_umem **umem,
+		       int *npages, int *page_shift, int *ncont,
+		       int *order)
 {
 	struct mlx5_ib_dev *dev = to_mdev(pd->device);
-	struct ib_umem *umem = ib_umem_get(pd->uobject->context, start, length,
-					   access_flags, 0);
-	if (IS_ERR(umem)) {
+	int err;
+
+	*umem = ib_umem_get(pd->uobject->context, start, length,
+			    access_flags, 0);
+	err = PTR_ERR_OR_ZERO(*umem);
+	if (err < 0) {
 		mlx5_ib_err(dev, "umem get failed (%ld)\n", PTR_ERR(umem));
-		return (void *)umem;
+		return err;
 	}
 
-	mlx5_ib_cont_pages(umem, start, MLX5_MKEY_PAGE_SHIFT_MASK, npages,
+	mlx5_ib_cont_pages(*umem, start, MLX5_MKEY_PAGE_SHIFT_MASK, npages,
 			   page_shift, ncont, order);
 	if (!*npages) {
 		mlx5_ib_warn(dev, "avoid zero region\n");
-		ib_umem_release(umem);
-		return ERR_PTR(-EINVAL);
+		ib_umem_release(*umem);
+		return -EINVAL;
 	}
 
 	mlx5_ib_dbg(dev, "npages %d, ncont %d, order %d, page_shift %d\n",
 		    *npages, *ncont, *order, *page_shift);
 
-	return umem;
+	return 0;
 }
 
 static void mlx5_ib_umr_done(struct ib_cq *cq, struct ib_wc *wc)
@@ -1193,11 +1197,11 @@ struct ib_mr *mlx5_ib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 
 	mlx5_ib_dbg(dev, "start 0x%llx, virt_addr 0x%llx, length 0x%llx, access_flags 0x%x\n",
 		    start, virt_addr, length, access_flags);
-	umem = mr_umem_get(pd, start, length, access_flags, &npages,
+	err = mr_umem_get(pd, start, length, access_flags, &umem, &npages,
 			   &page_shift, &ncont, &order);
 
-	if (IS_ERR(umem))
-		return (void *)umem;
+        if (err < 0)
+		return ERR_PTR(err);
 
 	if (use_umr(order)) {
 		mr = reg_umr(pd, umem, virt_addr, length, ncont, page_shift,
@@ -1371,10 +1375,9 @@ int mlx5_ib_rereg_user_mr(struct ib_mr *ib_mr, int flags, u64 start,
 		 */
 		flags |= IB_MR_REREG_TRANS;
 		ib_umem_release(mr->umem);
-		mr->umem = mr_umem_get(pd, addr, len, access_flags, &npages,
-				       &page_shift, &ncont, &order);
-		if (IS_ERR(mr->umem)) {
-			err = PTR_ERR(mr->umem);
+		err = mr_umem_get(pd, addr, len, access_flags, &mr->umem,
+				  &npages, &page_shift, &ncont, &order);
+		if (err < 0) {
 			mr->umem = NULL;
 			return err;
 		}
-- 
cgit v1.2.1


From 6efaf10f163d9a60d1d4b2a049b194a53537ba1b Mon Sep 17 00:00:00 2001
From: Petr Mladek <pmladek@suse.com>
Date: Wed, 19 Oct 2016 14:07:19 +0200
Subject: IB/rdmavt: Avoid queuing work into a destroyed cq kthread worker

The memory barrier is not enough to protect queuing works into
a destroyed cq kthread. Just imagine the following situation:

CPU1				CPU2

rvt_cq_enter()
  worker =  cq->rdi->worker;

				rvt_cq_exit()
				  rdi->worker = NULL;
				  smp_wmb();
				  kthread_flush_worker(worker);
				  kthread_stop(worker->task);
				  kfree(worker);

				  // nothing queued yet =>
				  // nothing flushed and
				  // happily stopped and freed

  if (likely(worker)) {
     // true => read before CPU2 acted
     cq->notify = RVT_CQ_NONE;
     cq->triggered++;
     kthread_queue_work(worker, &cq->comptask);

  BANG: worker has been flushed/stopped/freed in the meantime.

This patch solves this by protecting the critical sections by
rdi->n_cqs_lock. It seems that this lock is not much contended
and looks reasonable for this purpose.

One catch is that rvt_cq_enter() might be called from IRQ context.
Therefore we must always take the lock with IRQs disabled to avoid
a possible deadlock.

Signed-off-by: Petr Mladek <pmladek@suse.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rdmavt/cq.c | 30 ++++++++++++++++--------------
 1 file changed, 16 insertions(+), 14 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rdmavt/cq.c b/drivers/infiniband/sw/rdmavt/cq.c
index 6d9904a4a0ab..223ec4589fc7 100644
--- a/drivers/infiniband/sw/rdmavt/cq.c
+++ b/drivers/infiniband/sw/rdmavt/cq.c
@@ -119,18 +119,17 @@ void rvt_cq_enter(struct rvt_cq *cq, struct ib_wc *entry, bool solicited)
 	if (cq->notify == IB_CQ_NEXT_COMP ||
 	    (cq->notify == IB_CQ_SOLICITED &&
 	     (solicited || entry->status != IB_WC_SUCCESS))) {
-		struct kthread_worker *worker;
 		/*
 		 * This will cause send_complete() to be called in
 		 * another thread.
 		 */
-		smp_read_barrier_depends(); /* see rvt_cq_exit */
-		worker = cq->rdi->worker;
-		if (likely(worker)) {
+		spin_lock(&cq->rdi->n_cqs_lock);
+		if (likely(cq->rdi->worker)) {
 			cq->notify = RVT_CQ_NONE;
 			cq->triggered++;
-			kthread_queue_work(worker, &cq->comptask);
+			kthread_queue_work(cq->rdi->worker, &cq->comptask);
 		}
+		spin_unlock(&cq->rdi->n_cqs_lock);
 	}
 
 	spin_unlock_irqrestore(&cq->lock, flags);
@@ -240,15 +239,15 @@ struct ib_cq *rvt_create_cq(struct ib_device *ibdev,
 		}
 	}
 
-	spin_lock(&rdi->n_cqs_lock);
+	spin_lock_irq(&rdi->n_cqs_lock);
 	if (rdi->n_cqs_allocated == rdi->dparms.props.max_cq) {
-		spin_unlock(&rdi->n_cqs_lock);
+		spin_unlock_irq(&rdi->n_cqs_lock);
 		ret = ERR_PTR(-ENOMEM);
 		goto bail_ip;
 	}
 
 	rdi->n_cqs_allocated++;
-	spin_unlock(&rdi->n_cqs_lock);
+	spin_unlock_irq(&rdi->n_cqs_lock);
 
 	if (cq->ip) {
 		spin_lock_irq(&rdi->pending_lock);
@@ -296,9 +295,9 @@ int rvt_destroy_cq(struct ib_cq *ibcq)
 	struct rvt_dev_info *rdi = cq->rdi;
 
 	kthread_flush_work(&cq->comptask);
-	spin_lock(&rdi->n_cqs_lock);
+	spin_lock_irq(&rdi->n_cqs_lock);
 	rdi->n_cqs_allocated--;
-	spin_unlock(&rdi->n_cqs_lock);
+	spin_unlock_irq(&rdi->n_cqs_lock);
 	if (cq->ip)
 		kref_put(&cq->ip->ref, rvt_release_mmap_info);
 	else
@@ -541,12 +540,15 @@ void rvt_cq_exit(struct rvt_dev_info *rdi)
 {
 	struct kthread_worker *worker;
 
-	worker = rdi->worker;
-	if (!worker)
+	/* block future queuing from send_complete() */
+	spin_lock_irq(&rdi->n_cqs_lock);
+	if (!rdi->worker) {
+		spin_unlock_irq(&rdi->n_cqs_lock);
 		return;
-	/* blocks future queuing from send_complete() */
+	}
 	rdi->worker = NULL;
-	smp_wmb(); /* See rdi_cq_enter */
+	spin_unlock_irq(&rdi->n_cqs_lock);
+
 	kthread_flush_worker(worker);
 	kthread_stop(worker->task);
 	kfree(worker);
-- 
cgit v1.2.1


From f5eabf5e5129e8ab5b3e7f50b24444aca1680e64 Mon Sep 17 00:00:00 2001
From: Petr Mladek <pmladek@suse.com>
Date: Wed, 19 Oct 2016 14:07:20 +0200
Subject: IB/rdmavt: Handle the kthread worker using the new API

Use the new API to create and destroy the cq kthread worker.
The API hides some implementation details.

In particular, kthread_create_worker() allocates and initializes
struct kthread_worker. It runs the kthread the right way and stores
task_struct into the worker structure. In addition, the *on_cpu()
variant binds the kthread to the given cpu and the related memory
node.

kthread_destroy_worker() flushes all pending works, stops
the kthread and frees the structure.

This patch does not change the existing behavior. Note that we must
use the on_cpu() variant because the function starts the kthread
and it must bind it to the right CPU before waking. The numa node
is associated for given CPU as well.

Signed-off-by: Petr Mladek <pmladek@suse.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rdmavt/cq.c | 34 +++++++++++-----------------------
 1 file changed, 11 insertions(+), 23 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rdmavt/cq.c b/drivers/infiniband/sw/rdmavt/cq.c
index 223ec4589fc7..4d0b6992e847 100644
--- a/drivers/infiniband/sw/rdmavt/cq.c
+++ b/drivers/infiniband/sw/rdmavt/cq.c
@@ -503,33 +503,23 @@ int rvt_poll_cq(struct ib_cq *ibcq, int num_entries, struct ib_wc *entry)
  */
 int rvt_driver_cq_init(struct rvt_dev_info *rdi)
 {
-	int ret = 0;
 	int cpu;
-	struct task_struct *task;
+	struct kthread_worker *worker;
 
 	if (rdi->worker)
 		return 0;
+
 	spin_lock_init(&rdi->n_cqs_lock);
-	rdi->worker = kzalloc(sizeof(*rdi->worker), GFP_KERNEL);
-	if (!rdi->worker)
-		return -ENOMEM;
-	kthread_init_worker(rdi->worker);
-	task = kthread_create_on_node(
-		kthread_worker_fn,
-		rdi->worker,
-		rdi->dparms.node,
-		"%s", rdi->dparms.cq_name);
-	if (IS_ERR(task)) {
-		kfree(rdi->worker);
-		rdi->worker = NULL;
-		return PTR_ERR(task);
-	}
 
-	set_user_nice(task, MIN_NICE);
 	cpu = cpumask_first(cpumask_of_node(rdi->dparms.node));
-	kthread_bind(task, cpu);
-	wake_up_process(task);
-	return ret;
+	worker = kthread_create_worker_on_cpu(cpu, 0,
+					      "%s", rdi->dparms.cq_name);
+	if (IS_ERR(worker))
+		return PTR_ERR(worker);
+
+	set_user_nice(worker->task, MIN_NICE);
+	rdi->worker = worker;
+	return 0;
 }
 
 /**
@@ -549,7 +539,5 @@ void rvt_cq_exit(struct rvt_dev_info *rdi)
 	rdi->worker = NULL;
 	spin_unlock_irq(&rdi->n_cqs_lock);
 
-	kthread_flush_worker(worker);
-	kthread_stop(worker->task);
-	kfree(worker);
+	kthread_destroy_worker(worker);
 }
-- 
cgit v1.2.1


From 22dccc5454a39427de7b87a080d026b6bf66a7b9 Mon Sep 17 00:00:00 2001
From: Jim Foraker <foraker1@llnl.gov>
Date: Tue, 1 Nov 2016 13:44:12 -0700
Subject: IB/rdmavt: Only put mmap_info ref if it exists

rvt_create_qp() creates qp->ip only when a qp creation request comes from
userspace (udata is not NULL).  If we exceed the number of available
queue pairs however, the error path always attempts to put a kref to this
structure.  If the requestor is inside the kernel, this leads to a crash.

We fix this by checking that qp->ip is not NULL before caling kref_put().

Signed-off-by: Jim Foraker <foraker1@llnl.gov>
Acked-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Acked-by: Jonathan Toppins <jtoppins@redhat.com>
Acked-by: Alex Estrin <alex.estrin@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/sw/rdmavt/qp.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/sw/rdmavt/qp.c b/drivers/infiniband/sw/rdmavt/qp.c
index 9e14addd690c..2a13ac660f2b 100644
--- a/drivers/infiniband/sw/rdmavt/qp.c
+++ b/drivers/infiniband/sw/rdmavt/qp.c
@@ -901,7 +901,8 @@ struct ib_qp *rvt_create_qp(struct ib_pd *ibpd,
 	return ret;
 
 bail_ip:
-	kref_put(&qp->ip->ref, rvt_release_mmap_info);
+	if (qp->ip)
+		kref_put(&qp->ip->ref, rvt_release_mmap_info);
 
 bail_qpn:
 	free_qpn(&rdi->qp_dev->qpn_table, qp->ibqp.qp_num);
-- 
cgit v1.2.1


From def4a6ffc9d178d6bb14178f56873c4831fb05a7 Mon Sep 17 00:00:00 2001
From: Alexey Khoroshilov <khoroshilov@ispras.ru>
Date: Sat, 22 Oct 2016 01:01:21 +0300
Subject: IB/isert: do not ignore errors in dma_map_single()

There are several places, where errors in dma_map_single() are
ignored. The patch fixes them.

Found by Linux Driver Verification project (linuxtesting.org).

Signed-off-by: Alexey Khoroshilov <khoroshilov@ispras.ru>
Acked-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/ulp/isert/ib_isert.c | 6 ++++++
 1 file changed, 6 insertions(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/ulp/isert/ib_isert.c b/drivers/infiniband/ulp/isert/ib_isert.c
index 6dd43f63238e..f0ba5f83b02c 100644
--- a/drivers/infiniband/ulp/isert/ib_isert.c
+++ b/drivers/infiniband/ulp/isert/ib_isert.c
@@ -1851,6 +1851,8 @@ isert_put_response(struct iscsi_conn *conn, struct iscsi_cmd *cmd)
 		isert_cmd->pdu_buf_dma = ib_dma_map_single(ib_dev,
 				(void *)cmd->sense_buffer, pdu_len,
 				DMA_TO_DEVICE);
+		if (ib_dma_mapping_error(ib_dev, isert_cmd->pdu_buf_dma))
+			return -ENOMEM;
 
 		isert_cmd->pdu_buf_len = pdu_len;
 		tx_dsg->addr	= isert_cmd->pdu_buf_dma;
@@ -1978,6 +1980,8 @@ isert_put_reject(struct iscsi_cmd *cmd, struct iscsi_conn *conn)
 	isert_cmd->pdu_buf_dma = ib_dma_map_single(ib_dev,
 			(void *)cmd->buf_ptr, ISCSI_HDR_LEN,
 			DMA_TO_DEVICE);
+	if (ib_dma_mapping_error(ib_dev, isert_cmd->pdu_buf_dma))
+		return -ENOMEM;
 	isert_cmd->pdu_buf_len = ISCSI_HDR_LEN;
 	tx_dsg->addr	= isert_cmd->pdu_buf_dma;
 	tx_dsg->length	= ISCSI_HDR_LEN;
@@ -2018,6 +2022,8 @@ isert_put_text_rsp(struct iscsi_cmd *cmd, struct iscsi_conn *conn)
 
 		isert_cmd->pdu_buf_dma = ib_dma_map_single(ib_dev,
 				txt_rsp_buf, txt_rsp_len, DMA_TO_DEVICE);
+		if (ib_dma_mapping_error(ib_dev, isert_cmd->pdu_buf_dma))
+			return -ENOMEM;
 
 		isert_cmd->pdu_buf_len = txt_rsp_len;
 		tx_dsg->addr	= isert_cmd->pdu_buf_dma;
-- 
cgit v1.2.1


From 24dc08c3c9891a79f2754f99b7bffe65745af0f3 Mon Sep 17 00:00:00 2001
From: Philippe Reynes <tremyfr@gmail.com>
Date: Tue, 25 Oct 2016 17:29:47 +0200
Subject: IB/nes: use new api ethtool_{get|set}_link_ksettings

The ethtool api {get|set}_settings is deprecated.
We move this driver to new api {get|set}_link_ksettings.

Signed-off-by: Philippe Reynes <tremyfr@gmail.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/nes/nes_nic.c | 76 ++++++++++++++++++++-----------------
 1 file changed, 42 insertions(+), 34 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/nes/nes_nic.c b/drivers/infiniband/hw/nes/nes_nic.c
index 2b27d1351cf7..311b4ba1cade 100644
--- a/drivers/infiniband/hw/nes/nes_nic.c
+++ b/drivers/infiniband/hw/nes/nes_nic.c
@@ -1465,7 +1465,8 @@ static int nes_netdev_set_pauseparam(struct net_device *netdev,
 /**
  * nes_netdev_get_settings
  */
-static int nes_netdev_get_settings(struct net_device *netdev, struct ethtool_cmd *et_cmd)
+static int nes_netdev_get_link_ksettings(struct net_device *netdev,
+					 struct ethtool_link_ksettings *cmd)
 {
 	struct nes_vnic *nesvnic = netdev_priv(netdev);
 	struct nes_device *nesdev = nesvnic->nesdev;
@@ -1474,54 +1475,59 @@ static int nes_netdev_get_settings(struct net_device *netdev, struct ethtool_cmd
 	u8 phy_type = nesadapter->phy_type[mac_index];
 	u8 phy_index = nesadapter->phy_index[mac_index];
 	u16 phy_data;
+	u32 supported, advertising;
 
-	et_cmd->duplex = DUPLEX_FULL;
-	et_cmd->port   = PORT_MII;
-	et_cmd->maxtxpkt = 511;
-	et_cmd->maxrxpkt = 511;
+	cmd->base.duplex = DUPLEX_FULL;
+	cmd->base.port   = PORT_MII;
 
 	if (nesadapter->OneG_Mode) {
-		ethtool_cmd_speed_set(et_cmd, SPEED_1000);
+		cmd->base.speed = SPEED_1000;
 		if (phy_type == NES_PHY_TYPE_PUMA_1G) {
-			et_cmd->supported   = SUPPORTED_1000baseT_Full;
-			et_cmd->advertising = ADVERTISED_1000baseT_Full;
-			et_cmd->autoneg     = AUTONEG_DISABLE;
-			et_cmd->transceiver = XCVR_INTERNAL;
-			et_cmd->phy_address = mac_index;
+			supported   = SUPPORTED_1000baseT_Full;
+			advertising = ADVERTISED_1000baseT_Full;
+			cmd->base.autoneg     = AUTONEG_DISABLE;
+			cmd->base.phy_address = mac_index;
 		} else {
 			unsigned long flags;
-			et_cmd->supported   = SUPPORTED_1000baseT_Full
-					    | SUPPORTED_Autoneg;
-			et_cmd->advertising = ADVERTISED_1000baseT_Full
-					    | ADVERTISED_Autoneg;
+
+			supported = SUPPORTED_1000baseT_Full
+				| SUPPORTED_Autoneg;
+			advertising = ADVERTISED_1000baseT_Full
+				| ADVERTISED_Autoneg;
 			spin_lock_irqsave(&nesadapter->phy_lock, flags);
 			nes_read_1G_phy_reg(nesdev, 0, phy_index, &phy_data);
 			spin_unlock_irqrestore(&nesadapter->phy_lock, flags);
 			if (phy_data & 0x1000)
-				et_cmd->autoneg = AUTONEG_ENABLE;
+				cmd->base.autoneg = AUTONEG_ENABLE;
 			else
-				et_cmd->autoneg = AUTONEG_DISABLE;
-			et_cmd->transceiver = XCVR_EXTERNAL;
-			et_cmd->phy_address = phy_index;
+				cmd->base.autoneg = AUTONEG_DISABLE;
+			cmd->base.phy_address = phy_index;
 		}
+		ethtool_convert_legacy_u32_to_link_mode(
+			cmd->link_modes.supported, supported);
+		ethtool_convert_legacy_u32_to_link_mode(
+			cmd->link_modes.advertising, advertising);
 		return 0;
 	}
 	if ((phy_type == NES_PHY_TYPE_ARGUS) ||
 	    (phy_type == NES_PHY_TYPE_SFP_D) ||
 	    (phy_type == NES_PHY_TYPE_KR)) {
-		et_cmd->transceiver = XCVR_EXTERNAL;
-		et_cmd->port        = PORT_FIBRE;
-		et_cmd->supported   = SUPPORTED_FIBRE;
-		et_cmd->advertising = ADVERTISED_FIBRE;
-		et_cmd->phy_address = phy_index;
+		cmd->base.port        = PORT_FIBRE;
+		supported   = SUPPORTED_FIBRE;
+		advertising = ADVERTISED_FIBRE;
+		cmd->base.phy_address = phy_index;
 	} else {
-		et_cmd->transceiver = XCVR_INTERNAL;
-		et_cmd->supported   = SUPPORTED_10000baseT_Full;
-		et_cmd->advertising = ADVERTISED_10000baseT_Full;
-		et_cmd->phy_address = mac_index;
+		supported   = SUPPORTED_10000baseT_Full;
+		advertising = ADVERTISED_10000baseT_Full;
+		cmd->base.phy_address = mac_index;
 	}
-	ethtool_cmd_speed_set(et_cmd, SPEED_10000);
-	et_cmd->autoneg = AUTONEG_DISABLE;
+	cmd->base.speed = SPEED_10000;
+	cmd->base.autoneg = AUTONEG_DISABLE;
+	ethtool_convert_legacy_u32_to_link_mode(cmd->link_modes.supported,
+						supported);
+	ethtool_convert_legacy_u32_to_link_mode(cmd->link_modes.advertising,
+						advertising);
+
 	return 0;
 }
 
@@ -1529,7 +1535,9 @@ static int nes_netdev_get_settings(struct net_device *netdev, struct ethtool_cmd
 /**
  * nes_netdev_set_settings
  */
-static int nes_netdev_set_settings(struct net_device *netdev, struct ethtool_cmd *et_cmd)
+static int
+nes_netdev_set_link_ksettings(struct net_device *netdev,
+			      const struct ethtool_link_ksettings *cmd)
 {
 	struct nes_vnic *nesvnic = netdev_priv(netdev);
 	struct nes_device *nesdev = nesvnic->nesdev;
@@ -1543,7 +1551,7 @@ static int nes_netdev_set_settings(struct net_device *netdev, struct ethtool_cmd
 
 		spin_lock_irqsave(&nesadapter->phy_lock, flags);
 		nes_read_1G_phy_reg(nesdev, 0, phy_index, &phy_data);
-		if (et_cmd->autoneg) {
+		if (cmd->base.autoneg) {
 			/* Turn on Full duplex, Autoneg, and restart autonegotiation */
 			phy_data |= 0x1300;
 		} else {
@@ -1560,8 +1568,6 @@ static int nes_netdev_set_settings(struct net_device *netdev, struct ethtool_cmd
 
 static const struct ethtool_ops nes_ethtool_ops = {
 	.get_link = ethtool_op_get_link,
-	.get_settings = nes_netdev_get_settings,
-	.set_settings = nes_netdev_set_settings,
 	.get_strings = nes_netdev_get_strings,
 	.get_sset_count = nes_netdev_get_sset_count,
 	.get_ethtool_stats = nes_netdev_get_ethtool_stats,
@@ -1570,6 +1576,8 @@ static const struct ethtool_ops nes_ethtool_ops = {
 	.set_coalesce = nes_netdev_set_coalesce,
 	.get_pauseparam = nes_netdev_get_pauseparam,
 	.set_pauseparam = nes_netdev_set_pauseparam,
+	.get_link_ksettings = nes_netdev_get_link_ksettings,
+	.set_link_ksettings = nes_netdev_set_link_ksettings,
 };
 
 static void nes_vlan_mode(struct net_device *netdev, struct nes_device *nesdev, netdev_features_t features)
-- 
cgit v1.2.1


From 9315bc9a133011fdb084f2626b86db3ebb64661f Mon Sep 17 00:00:00 2001
From: Hans Westgaard Ry <hans.westgaard.ry@oracle.com>
Date: Fri, 28 Oct 2016 13:14:29 +0200
Subject: IB/core: Issue DREQ when receiving REQ/REP for stale QP
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

from "InfiBand Architecture Specifications Volume 1":

  A QP is said to have a stale connection when only one side has
  connection information. A stale connection may result if the remote CM
  had dropped the connection and sent a DREQ but the DREQ was never
  received by the local CM. Alternatively the remote CM may have lost
  all record of past connections because its node crashed and rebooted,
  while the local CM did not become aware of the remote node's reboot
  and therefore did not clean up stale connections.

and:

   A local CM may receive a REQ/REP for a stale connection. It shall
   abort the connection issuing REJ to the REQ/REP. It shall then issue
   DREQ with "DREQ:remote QPN” set to the remote QPN from the REQ/REP.

This patch solves a problem with reuse of QPN. Current codebase, that
is IPoIB, relies on a REAP-mechanism to do cleanup of the structures
in CM. A problem with this is the timeconstants governing this
mechanism; they are up to 768 seconds and the interface may look
inresponsive in that period.  Issuing a DREQ (and receiving a DREP)
does the necessary cleanup and the interface comes up.

Signed-off-by: Hans Westgaard Ry <hans.westgaard.ry@oracle.com>
Reviewed-by: Håkon Bugge <haakon.bugge@oracle.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/cm.c | 24 +++++++++++++++++++++++-
 1 file changed, 23 insertions(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/core/cm.c b/drivers/infiniband/core/cm.c
index c99525512b34..c97e4d586b44 100644
--- a/drivers/infiniband/core/cm.c
+++ b/drivers/infiniband/core/cm.c
@@ -1519,6 +1519,7 @@ static struct cm_id_private * cm_match_req(struct cm_work *work,
 	struct cm_id_private *listen_cm_id_priv, *cur_cm_id_priv;
 	struct cm_timewait_info *timewait_info;
 	struct cm_req_msg *req_msg;
+	struct ib_cm_id *cm_id;
 
 	req_msg = (struct cm_req_msg *)work->mad_recv_wc->recv_buf.mad;
 
@@ -1540,10 +1541,18 @@ static struct cm_id_private * cm_match_req(struct cm_work *work,
 	timewait_info = cm_insert_remote_qpn(cm_id_priv->timewait_info);
 	if (timewait_info) {
 		cm_cleanup_timewait(cm_id_priv->timewait_info);
+		cur_cm_id_priv = cm_get_id(timewait_info->work.local_id,
+					   timewait_info->work.remote_id);
+
 		spin_unlock_irq(&cm.lock);
 		cm_issue_rej(work->port, work->mad_recv_wc,
 			     IB_CM_REJ_STALE_CONN, CM_MSG_RESPONSE_REQ,
 			     NULL, 0);
+		if (cur_cm_id_priv) {
+			cm_id = &cur_cm_id_priv->id;
+			ib_send_cm_dreq(cm_id, NULL, 0);
+			cm_deref_id(cur_cm_id_priv);
+		}
 		return NULL;
 	}
 
@@ -1919,6 +1928,9 @@ static int cm_rep_handler(struct cm_work *work)
 	struct cm_id_private *cm_id_priv;
 	struct cm_rep_msg *rep_msg;
 	int ret;
+	struct cm_id_private *cur_cm_id_priv;
+	struct ib_cm_id *cm_id;
+	struct cm_timewait_info *timewait_info;
 
 	rep_msg = (struct cm_rep_msg *)work->mad_recv_wc->recv_buf.mad;
 	cm_id_priv = cm_acquire_id(rep_msg->remote_comm_id, 0);
@@ -1953,16 +1965,26 @@ static int cm_rep_handler(struct cm_work *work)
 		goto error;
 	}
 	/* Check for a stale connection. */
-	if (cm_insert_remote_qpn(cm_id_priv->timewait_info)) {
+	timewait_info = cm_insert_remote_qpn(cm_id_priv->timewait_info);
+	if (timewait_info) {
 		rb_erase(&cm_id_priv->timewait_info->remote_id_node,
 			 &cm.remote_id_table);
 		cm_id_priv->timewait_info->inserted_remote_id = 0;
+		cur_cm_id_priv = cm_get_id(timewait_info->work.local_id,
+					   timewait_info->work.remote_id);
+
 		spin_unlock(&cm.lock);
 		spin_unlock_irq(&cm_id_priv->lock);
 		cm_issue_rej(work->port, work->mad_recv_wc,
 			     IB_CM_REJ_STALE_CONN, CM_MSG_RESPONSE_REP,
 			     NULL, 0);
 		ret = -EINVAL;
+		if (cur_cm_id_priv) {
+			cm_id = &cur_cm_id_priv->id;
+			ib_send_cm_dreq(cm_id, NULL, 0);
+			cm_deref_id(cur_cm_id_priv);
+		}
+
 		goto error;
 	}
 	spin_unlock(&cm.lock);
-- 
cgit v1.2.1


From 5f4c7e4eb5f36974ed46a485290f5d01ace5fdba Mon Sep 17 00:00:00 2001
From: Julia Lawall <Julia.Lawall@lip6.fr>
Date: Fri, 11 Nov 2016 20:04:26 +0100
Subject: IB/usnic: simplify IS_ERR_OR_NULL to IS_ERR

The function usnic_ib_qp_grp_get_chunk only returns an ERR_PTR value or a
valid pointer, never NULL.  The same is true of get_qp_res_chunk, which
just returns the result of calling usnic_ib_qp_grp_get_chunk.  Simplify
IS_ERR_OR_NULL to IS_ERR in both cases.

The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)

// <smpl>
@@
expression t,e;
@@

t = \(usnic_ib_qp_grp_get_chunk(...)\|get_qp_res_chunk(...)\)
... when != t=e
- IS_ERR_OR_NULL(t)
+ IS_ERR(t)

@@
expression t,e,e1;
@@

t = \(usnic_ib_qp_grp_get_chunk(...)\|get_qp_res_chunk(...)\)
... when != t=e
?- t ? PTR_ERR(t) : e1
+ PTR_ERR(t)
... when any
// </smpl>

Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr>
Reviewed-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/usnic/usnic_ib_qp_grp.c | 12 ++++++------
 drivers/infiniband/hw/usnic/usnic_ib_verbs.c  | 12 ++++++------
 2 files changed, 12 insertions(+), 12 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/usnic/usnic_ib_qp_grp.c b/drivers/infiniband/hw/usnic/usnic_ib_qp_grp.c
index 5b0248adf4ce..2b1a3819627f 100644
--- a/drivers/infiniband/hw/usnic/usnic_ib_qp_grp.c
+++ b/drivers/infiniband/hw/usnic/usnic_ib_qp_grp.c
@@ -117,10 +117,10 @@ static int enable_qp_grp(struct usnic_ib_qp_grp *qp_grp)
 	vnic_idx = usnic_vnic_get_index(qp_grp->vf->vnic);
 
 	res_chunk = get_qp_res_chunk(qp_grp);
-	if (IS_ERR_OR_NULL(res_chunk)) {
+	if (IS_ERR(res_chunk)) {
 		usnic_err("Unable to get qp res with err %ld\n",
 				PTR_ERR(res_chunk));
-		return res_chunk ? PTR_ERR(res_chunk) : -ENOMEM;
+		return PTR_ERR(res_chunk);
 	}
 
 	for (i = 0; i < res_chunk->cnt; i++) {
@@ -158,10 +158,10 @@ static int disable_qp_grp(struct usnic_ib_qp_grp *qp_grp)
 	vnic_idx = usnic_vnic_get_index(qp_grp->vf->vnic);
 
 	res_chunk = get_qp_res_chunk(qp_grp);
-	if (IS_ERR_OR_NULL(res_chunk)) {
+	if (IS_ERR(res_chunk)) {
 		usnic_err("Unable to get qp res with err %ld\n",
 			PTR_ERR(res_chunk));
-		return res_chunk ? PTR_ERR(res_chunk) : -ENOMEM;
+		return PTR_ERR(res_chunk);
 	}
 
 	for (i = 0; i < res_chunk->cnt; i++) {
@@ -186,11 +186,11 @@ static int init_filter_action(struct usnic_ib_qp_grp *qp_grp,
 	struct usnic_vnic_res_chunk *res_chunk;
 
 	res_chunk = usnic_ib_qp_grp_get_chunk(qp_grp, USNIC_VNIC_RES_TYPE_RQ);
-	if (IS_ERR_OR_NULL(res_chunk)) {
+	if (IS_ERR(res_chunk)) {
 		usnic_err("Unable to get %s with err %ld\n",
 			usnic_vnic_res_type_to_str(USNIC_VNIC_RES_TYPE_RQ),
 			PTR_ERR(res_chunk));
-		return res_chunk ? PTR_ERR(res_chunk) : -ENOMEM;
+		return PTR_ERR(res_chunk);
 	}
 
 	uaction->vnic_idx = usnic_vnic_get_index(qp_grp->vf->vnic);
diff --git a/drivers/infiniband/hw/usnic/usnic_ib_verbs.c b/drivers/infiniband/hw/usnic/usnic_ib_verbs.c
index a5bfbba6bbac..79766db1dc6c 100644
--- a/drivers/infiniband/hw/usnic/usnic_ib_verbs.c
+++ b/drivers/infiniband/hw/usnic/usnic_ib_verbs.c
@@ -87,12 +87,12 @@ static int usnic_ib_fill_create_qp_resp(struct usnic_ib_qp_grp *qp_grp,
 	resp.bar_len = bar->len;
 
 	chunk = usnic_ib_qp_grp_get_chunk(qp_grp, USNIC_VNIC_RES_TYPE_RQ);
-	if (IS_ERR_OR_NULL(chunk)) {
+	if (IS_ERR(chunk)) {
 		usnic_err("Failed to get chunk %s for qp_grp %d with err %ld\n",
 			usnic_vnic_res_type_to_str(USNIC_VNIC_RES_TYPE_RQ),
 			qp_grp->grp_id,
 			PTR_ERR(chunk));
-		return chunk ? PTR_ERR(chunk) : -ENOMEM;
+		return PTR_ERR(chunk);
 	}
 
 	WARN_ON(chunk->type != USNIC_VNIC_RES_TYPE_RQ);
@@ -101,12 +101,12 @@ static int usnic_ib_fill_create_qp_resp(struct usnic_ib_qp_grp *qp_grp,
 		resp.rq_idx[i] = chunk->res[i]->vnic_idx;
 
 	chunk = usnic_ib_qp_grp_get_chunk(qp_grp, USNIC_VNIC_RES_TYPE_WQ);
-	if (IS_ERR_OR_NULL(chunk)) {
+	if (IS_ERR(chunk)) {
 		usnic_err("Failed to get chunk %s for qp_grp %d with err %ld\n",
 			usnic_vnic_res_type_to_str(USNIC_VNIC_RES_TYPE_WQ),
 			qp_grp->grp_id,
 			PTR_ERR(chunk));
-		return chunk ? PTR_ERR(chunk) : -ENOMEM;
+		return PTR_ERR(chunk);
 	}
 
 	WARN_ON(chunk->type != USNIC_VNIC_RES_TYPE_WQ);
@@ -115,12 +115,12 @@ static int usnic_ib_fill_create_qp_resp(struct usnic_ib_qp_grp *qp_grp,
 		resp.wq_idx[i] = chunk->res[i]->vnic_idx;
 
 	chunk = usnic_ib_qp_grp_get_chunk(qp_grp, USNIC_VNIC_RES_TYPE_CQ);
-	if (IS_ERR_OR_NULL(chunk)) {
+	if (IS_ERR(chunk)) {
 		usnic_err("Failed to get chunk %s for qp_grp %d with err %ld\n",
 			usnic_vnic_res_type_to_str(USNIC_VNIC_RES_TYPE_CQ),
 			qp_grp->grp_id,
 			PTR_ERR(chunk));
-		return chunk ? PTR_ERR(chunk) : -ENOMEM;
+		return PTR_ERR(chunk);
 	}
 
 	WARN_ON(chunk->type != USNIC_VNIC_RES_TYPE_CQ);
-- 
cgit v1.2.1


From 0d38c240f97602d9a4553252bb710521f49bb264 Mon Sep 17 00:00:00 2001
From: Bart Van Assche <bart.vanassche@sandisk.com>
Date: Fri, 11 Nov 2016 16:36:06 -0800
Subject: IB/srpt: Report login failures only once

Report the following message only once if no ACL has been configured
yet for an initiator port:

"Rejected login because no ACL has been configured yet for initiator %s.\n"

Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Cc: Nicholas Bellinger <nab@linux-iscsi.org>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Sagi Grimberg <sagig@grimberg.me>
Reviewed-by: Max Gurtovoy <maxg@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/ulp/srpt/ib_srpt.c | 22 +++++++++-------------
 1 file changed, 9 insertions(+), 13 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/ulp/srpt/ib_srpt.c b/drivers/infiniband/ulp/srpt/ib_srpt.c
index 0b1f69ed2e92..d21ba9d857c3 100644
--- a/drivers/infiniband/ulp/srpt/ib_srpt.c
+++ b/drivers/infiniband/ulp/srpt/ib_srpt.c
@@ -1840,7 +1840,6 @@ static int srpt_cm_req_recv(struct ib_cm_id *cm_id,
 	struct srpt_rdma_ch *ch, *tmp_ch;
 	u32 it_iu_len;
 	int i, ret = 0;
-	unsigned char *p;
 
 	WARN_ON_ONCE(irqs_disabled());
 
@@ -1994,21 +1993,18 @@ static int srpt_cm_req_recv(struct ib_cm_id *cm_id,
 			be64_to_cpu(*(__be64 *)(ch->i_port_id + 8)));
 
 	pr_debug("registering session %s\n", ch->sess_name);
-	p = &ch->sess_name[0];
 
-try_again:
 	ch->sess = target_alloc_session(&sport->port_tpg_1, 0, 0,
-					TARGET_PROT_NORMAL, p, ch, NULL);
+					TARGET_PROT_NORMAL, ch->sess_name, ch,
+					NULL);
+	/* Retry without leading "0x" */
+	if (IS_ERR(ch->sess))
+		ch->sess = target_alloc_session(&sport->port_tpg_1, 0, 0,
+						TARGET_PROT_NORMAL,
+						ch->sess_name + 2, ch, NULL);
 	if (IS_ERR(ch->sess)) {
-		pr_info("Rejected login because no ACL has been"
-			" configured yet for initiator %s.\n", p);
-		/*
-		 * XXX: Hack to retry of ch->i_port_id without leading '0x'
-		 */
-		if (p == &ch->sess_name[0]) {
-			p += 2;
-			goto try_again;
-		}
+		pr_info("Rejected login because no ACL has been configured yet for initiator %s.\n",
+			ch->sess_name);
 		rej->reason = cpu_to_be32((PTR_ERR(ch->sess) == -ENOMEM) ?
 				SRP_LOGIN_REJ_INSUFFICIENT_RESOURCES :
 				SRP_LOGIN_REJ_CHANNEL_LIMIT_REACHED);
-- 
cgit v1.2.1


From b42dde478bcaa8113b0d1cd82ad0048372c5599d Mon Sep 17 00:00:00 2001
From: Bart Van Assche <bart.vanassche@sandisk.com>
Date: Mon, 14 Nov 2016 08:44:11 -0800
Subject: IB/mlx4: Rework special QP creation error path

The special QP creation error path relies on offset_of(struct mlx4_ib_sqp,
qp) == 0. Remove this assumption because that makes the QP creation
code easier to understand.

Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Cc: Yishai Hadas <yishaih@mellanox.com>
Reviewed-by: Laurence Oberman <loberman@redhat.com>
Reviewed-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/qp.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index c22454383976..c068add8838b 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -644,7 +644,7 @@ static int create_qp_common(struct mlx4_ib_dev *dev, struct ib_pd *pd,
 	int qpn;
 	int err;
 	struct ib_qp_cap backup_cap;
-	struct mlx4_ib_sqp *sqp;
+	struct mlx4_ib_sqp *sqp = NULL;
 	struct mlx4_ib_qp *qp;
 	enum mlx4_ib_qp_type qp_type = (enum mlx4_ib_qp_type) init_attr->qp_type;
 	struct mlx4_ib_cq *mcq;
@@ -933,7 +933,9 @@ err_db:
 		mlx4_db_free(dev->dev, &qp->db);
 
 err:
-	if (!*caller_qp)
+	if (sqp)
+		kfree(sqp);
+	else if (!*caller_qp)
 		kfree(qp);
 	return err;
 }
-- 
cgit v1.2.1


From 2fe2f378dd45847d2643638c07a7658822087836 Mon Sep 17 00:00:00 2001
From: Bart Van Assche <bart.vanassche@sandisk.com>
Date: Mon, 21 Nov 2016 10:21:17 -0800
Subject: IB/mad: Fix an array index check

The array ib_mad_mgmt_class_table.method_table has MAX_MGMT_CLASS
(80) elements. Hence compare the array index with that value instead
of with IB_MGMT_MAX_METHODS (128). This patch avoids that Coverity
reports the following:

Overrunning array class->method_table of 80 8-byte elements at element index 127 (byte offset 1016) using index convert_mgmt_class(mad_hdr->mgmt_class) (which evaluates to 127).

Fixes: commit b7ab0b19a85f ("IB/mad: Verify mgmt class in received MADs")
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Cc: Sean Hefty <sean.hefty@intel.com>
Cc: <stable@vger.kernel.org>
Reviewed-by: Hal Rosenstock <hal@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/mad.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/core/mad.c b/drivers/infiniband/core/mad.c
index bc2a9c2caa3c..d120f6ff1ea6 100644
--- a/drivers/infiniband/core/mad.c
+++ b/drivers/infiniband/core/mad.c
@@ -1746,7 +1746,7 @@ find_mad_agent(struct ib_mad_port_private *port_priv,
 			if (!class)
 				goto out;
 			if (convert_mgmt_class(mad_hdr->mgmt_class) >=
-			    IB_MGMT_MAX_METHODS)
+			    ARRAY_SIZE(class->method_table))
 				goto out;
 			method = class->method_table[convert_mgmt_class(
 							mad_hdr->mgmt_class)];
-- 
cgit v1.2.1


From 11b642b84e8c43e8597de031678d15c08dd057bc Mon Sep 17 00:00:00 2001
From: Bart Van Assche <bart.vanassche@sandisk.com>
Date: Mon, 21 Nov 2016 10:21:41 -0800
Subject: IPoIB: Avoid reading an uninitialized member variable

This patch avoids that Coverity reports the following:

    Using uninitialized value port_attr.state when calling printk

Fixes: commit 94232d9ce817 ("IPoIB: Start multicast join process only on active ports")
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Cc: Erez Shitrit <erezsh@mellanox.com>
Cc: <stable@vger.kernel.org>
Reviewed-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/ulp/ipoib/ipoib_multicast.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
index d3394b6add24..2a3980c2c670 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
@@ -575,8 +575,11 @@ void ipoib_mcast_join_task(struct work_struct *work)
 	if (!test_bit(IPOIB_FLAG_OPER_UP, &priv->flags))
 		return;
 
-	if (ib_query_port(priv->ca, priv->port, &port_attr) ||
-	    port_attr.state != IB_PORT_ACTIVE) {
+	if (ib_query_port(priv->ca, priv->port, &port_attr)) {
+		ipoib_dbg(priv, "ib_query_port() failed\n");
+		return;
+	}
+	if (port_attr.state != IB_PORT_ACTIVE) {
 		ipoib_dbg(priv, "port state is not ACTIVE (state = %d) suspending join task\n",
 			  port_attr.state);
 		return;
-- 
cgit v1.2.1


From d3a2418ee36a59bc02e9d454723f3175dcf4bfd9 Mon Sep 17 00:00:00 2001
From: Bart Van Assche <bart.vanassche@sandisk.com>
Date: Mon, 21 Nov 2016 10:22:17 -0800
Subject: IB/multicast: Check ib_find_pkey() return value

This patch avoids that Coverity complains about not checking the
ib_find_pkey() return value.

Fixes: commit 547af76521b3 ("IB/multicast: Report errors on multicast groups if P_key changes")
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Cc: Sean Hefty <sean.hefty@intel.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/multicast.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/core/multicast.c b/drivers/infiniband/core/multicast.c
index e51b739f6ea3..322cb67b07a9 100644
--- a/drivers/infiniband/core/multicast.c
+++ b/drivers/infiniband/core/multicast.c
@@ -518,8 +518,11 @@ static void join_handler(int status, struct ib_sa_mcmember_rec *rec,
 		process_join_error(group, status);
 	else {
 		int mgids_changed, is_mgid0;
-		ib_find_pkey(group->port->dev->device, group->port->port_num,
-			     be16_to_cpu(rec->pkey), &pkey_index);
+
+		if (ib_find_pkey(group->port->dev->device,
+				 group->port->port_num, be16_to_cpu(rec->pkey),
+				 &pkey_index))
+			pkey_index = MCAST_INVALID_PKEY_INDEX;
 
 		spin_lock_irq(&group->port->lock);
 		if (group->state == MCAST_BUSY &&
-- 
cgit v1.2.1


From 1a1faf7a8a251d134d375b7783a614ee79e932f2 Mon Sep 17 00:00:00 2001
From: Bart Van Assche <bart.vanassche@sandisk.com>
Date: Mon, 21 Nov 2016 13:56:46 -0800
Subject: IB/srp: Fix CONFIG_DYNAMIC_DEBUG=n build

Avoid that the kernel build fails as follows if dynamic debug support
is disabled:

drivers/infiniband/ulp/srp/ib_srp.c:2272:3: error: implicit declaration of function 'DEFINE_DYNAMIC_DEBUG_METADATA'
drivers/infiniband/ulp/srp/ib_srp.c:2272:33: error: 'ddm' undeclared (first use in this function)
drivers/infiniband/ulp/srp/ib_srp.c:2275:39: error: '_DPRINTK_FLAGS_PRINT' undeclared (first use in this function)

Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/ulp/srp/ib_srp.c | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c
index d980fb458ad4..8bb720c68af0 100644
--- a/drivers/infiniband/ulp/srp/ib_srp.c
+++ b/drivers/infiniband/ulp/srp/ib_srp.c
@@ -64,6 +64,11 @@ MODULE_LICENSE("Dual BSD/GPL");
 MODULE_VERSION(DRV_VERSION);
 MODULE_INFO(release_date, DRV_RELDATE);
 
+#if !defined(CONFIG_DYNAMIC_DEBUG)
+#define DEFINE_DYNAMIC_DEBUG_METADATA(name, fmt)
+#define DYNAMIC_DEBUG_BRANCH(descriptor) false
+#endif
+
 static unsigned int srp_sg_tablesize;
 static unsigned int cmd_sg_entries;
 static unsigned int indirect_sg_entries;
@@ -1556,7 +1561,6 @@ static int srp_map_idb(struct srp_rdma_ch *ch, struct srp_request *req,
 	return 0;
 }
 
-#if defined(DYNAMIC_DATA_DEBUG)
 static void srp_check_mapping(struct srp_map_state *state,
 			      struct srp_rdma_ch *ch, struct srp_request *req,
 			      struct scatterlist *scat, int count)
@@ -1580,7 +1584,6 @@ static void srp_check_mapping(struct srp_map_state *state,
 		       scsi_bufflen(req->scmnd), desc_len, mr_len,
 		       state->ndesc, state->nmdesc);
 }
-#endif
 
 /**
  * srp_map_data() - map SCSI data buffer onto an SRP request
@@ -1669,14 +1672,12 @@ static int srp_map_data(struct scsi_cmnd *scmnd, struct srp_rdma_ch *ch,
 	if (ret < 0)
 		goto unmap;
 
-#if defined(DYNAMIC_DEBUG)
 	{
 		DEFINE_DYNAMIC_DEBUG_METADATA(ddm,
 			"Memory mapping consistency check");
-		if (unlikely(ddm.flags & _DPRINTK_FLAGS_PRINT))
+		if (DYNAMIC_DEBUG_BRANCH(ddm))
 			srp_check_mapping(&state, ch, req, scat, count);
 	}
-#endif
 
 	/* We've mapped the request, now pull as much of the indirect
 	 * descriptor table as we can into the command buffer. If this
-- 
cgit v1.2.1


From 042dd765bdf401c0ccdeb16717b0c2a0b1405f18 Mon Sep 17 00:00:00 2001
From: Bart Van Assche <bart.vanassche@sandisk.com>
Date: Mon, 21 Nov 2016 13:57:07 -0800
Subject: IB/srp: Introduce a local variable in srp_add_one()

This patch makes the srp_add_one() code more compact and does not
change any functionality.

Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/ulp/srp/ib_srp.c | 17 +++++++++--------
 1 file changed, 9 insertions(+), 8 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c
index 8bb720c68af0..c216c6eaef26 100644
--- a/drivers/infiniband/ulp/srp/ib_srp.c
+++ b/drivers/infiniband/ulp/srp/ib_srp.c
@@ -3527,6 +3527,7 @@ free_host:
 static void srp_add_one(struct ib_device *device)
 {
 	struct srp_device *srp_dev;
+	struct ib_device_attr *attr = &device->attrs;
 	struct srp_host *host;
 	int mr_page_shift, p;
 	u64 max_pages_per_mr;
@@ -3541,25 +3542,25 @@ static void srp_add_one(struct ib_device *device)
 	 * minimum of 4096 bytes. We're unlikely to build large sglists
 	 * out of smaller entries.
 	 */
-	mr_page_shift		= max(12, ffs(device->attrs.page_size_cap) - 1);
+	mr_page_shift		= max(12, ffs(attr->page_size_cap) - 1);
 	srp_dev->mr_page_size	= 1 << mr_page_shift;
 	srp_dev->mr_page_mask	= ~((u64) srp_dev->mr_page_size - 1);
-	max_pages_per_mr	= device->attrs.max_mr_size;
+	max_pages_per_mr	= attr->max_mr_size;
 	do_div(max_pages_per_mr, srp_dev->mr_page_size);
 	pr_debug("%s: %llu / %u = %llu <> %u\n", __func__,
-		 device->attrs.max_mr_size, srp_dev->mr_page_size,
+		 attr->max_mr_size, srp_dev->mr_page_size,
 		 max_pages_per_mr, SRP_MAX_PAGES_PER_MR);
 	srp_dev->max_pages_per_mr = min_t(u64, SRP_MAX_PAGES_PER_MR,
 					  max_pages_per_mr);
 
 	srp_dev->has_fmr = (device->alloc_fmr && device->dealloc_fmr &&
 			    device->map_phys_fmr && device->unmap_fmr);
-	srp_dev->has_fr = (device->attrs.device_cap_flags &
+	srp_dev->has_fr = (attr->device_cap_flags &
 			   IB_DEVICE_MEM_MGT_EXTENSIONS);
 	if (!never_register && !srp_dev->has_fmr && !srp_dev->has_fr) {
 		dev_warn(&device->dev, "neither FMR nor FR is supported\n");
 	} else if (!never_register &&
-		   device->attrs.max_mr_size >= 2 * srp_dev->mr_page_size) {
+		   attr->max_mr_size >= 2 * srp_dev->mr_page_size) {
 		srp_dev->use_fast_reg = (srp_dev->has_fr &&
 					 (!srp_dev->has_fmr || prefer_fr));
 		srp_dev->use_fmr = !srp_dev->use_fast_reg && srp_dev->has_fmr;
@@ -3572,13 +3573,13 @@ static void srp_add_one(struct ib_device *device)
 	if (srp_dev->use_fast_reg) {
 		srp_dev->max_pages_per_mr =
 			min_t(u32, srp_dev->max_pages_per_mr,
-			      device->attrs.max_fast_reg_page_list_len);
+			      attr->max_fast_reg_page_list_len);
 	}
 	srp_dev->mr_max_size	= srp_dev->mr_page_size *
 				   srp_dev->max_pages_per_mr;
 	pr_debug("%s: mr_page_shift = %d, device->max_mr_size = %#llx, device->max_fast_reg_page_list_len = %u, max_pages_per_mr = %d, mr_max_size = %#x\n",
-		 device->name, mr_page_shift, device->attrs.max_mr_size,
-		 device->attrs.max_fast_reg_page_list_len,
+		 device->name, mr_page_shift, attr->max_mr_size,
+		 attr->max_fast_reg_page_list_len,
 		 srp_dev->max_pages_per_mr, srp_dev->mr_max_size);
 
 	INIT_LIST_HEAD(&srp_dev->dev_list);
-- 
cgit v1.2.1


From 3787d9908c4e05af0322613fe7f8c617c1ddb1d5 Mon Sep 17 00:00:00 2001
From: Bart Van Assche <bart.vanassche@sandisk.com>
Date: Mon, 21 Nov 2016 13:57:24 -0800
Subject: IB/srp: Make login failures easier to debug

If login fails because memory region allocation failed it can be
hard to figure out what happened. Make it easier to figure out
why login failed by logging a message if ib_alloc_mr() fails.

Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/ulp/srp/ib_srp.c | 3 +++
 1 file changed, 3 insertions(+)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c
index c216c6eaef26..81cb27f2f0e6 100644
--- a/drivers/infiniband/ulp/srp/ib_srp.c
+++ b/drivers/infiniband/ulp/srp/ib_srp.c
@@ -389,6 +389,9 @@ static struct srp_fr_pool *srp_create_fr_pool(struct ib_device *device,
 				 max_page_list_len);
 		if (IS_ERR(mr)) {
 			ret = PTR_ERR(mr);
+			if (ret == -ENOMEM)
+				pr_info("%s: ib_alloc_mr() failed. Try to reduce max_cmd_per_lun, max_sect or ch_count\n",
+					dev_name(&device->dev));
 			goto destroy_pool;
 		}
 		d->mr = mr;
-- 
cgit v1.2.1


From 290081b45342ef902ed756f929fa9e4feb9f7dab Mon Sep 17 00:00:00 2001
From: Bart Van Assche <bart.vanassche@sandisk.com>
Date: Mon, 21 Nov 2016 13:57:41 -0800
Subject: IB/srp: Make mapping failures easier to debug

Make it easier to figure out what is going on if memory mapping
fails because more memory regions than mr_per_cmd are needed.

Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/ulp/srp/ib_srp.c | 12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c
index 81cb27f2f0e6..bb9d73daadfa 100644
--- a/drivers/infiniband/ulp/srp/ib_srp.c
+++ b/drivers/infiniband/ulp/srp/ib_srp.c
@@ -1274,8 +1274,12 @@ static int srp_map_finish_fmr(struct srp_map_state *state,
 	struct ib_pool_fmr *fmr;
 	u64 io_addr = 0;
 
-	if (state->fmr.next >= state->fmr.end)
+	if (state->fmr.next >= state->fmr.end) {
+		shost_printk(KERN_ERR, ch->target->scsi_host,
+			     PFX "Out of MRs (mr_per_cmd = %d)\n",
+			     ch->target->mr_per_cmd);
 		return -ENOMEM;
+	}
 
 	WARN_ON_ONCE(!dev->use_fmr);
 
@@ -1331,8 +1335,12 @@ static int srp_map_finish_fr(struct srp_map_state *state,
 	u32 rkey;
 	int n, err;
 
-	if (state->fr.next >= state->fr.end)
+	if (state->fr.next >= state->fr.end) {
+		shost_printk(KERN_ERR, ch->target->scsi_host,
+			     PFX "Out of MRs (mr_per_cmd = %d)\n",
+			     ch->target->mr_per_cmd);
 		return -ENOMEM;
+	}
 
 	WARN_ON_ONCE(!dev->use_fast_reg);
 
-- 
cgit v1.2.1


From 4fa354c9dbfef9226a690d8ee319b046f3067a6a Mon Sep 17 00:00:00 2001
From: Bart Van Assche <bart.vanassche@sandisk.com>
Date: Mon, 21 Nov 2016 13:58:18 -0800
Subject: IB/srp: Make writing the add_target sysfs attr interruptible

Avoid that shutdown of srp_daemon is delayed if add_target_mutex is
held by another process.

Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/ulp/srp/ib_srp.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c
index bb9d73daadfa..8ddc07123193 100644
--- a/drivers/infiniband/ulp/srp/ib_srp.c
+++ b/drivers/infiniband/ulp/srp/ib_srp.c
@@ -3299,7 +3299,9 @@ static ssize_t srp_create_target(struct device *dev,
 	 */
 	scsi_host_get(target->scsi_host);
 
-	mutex_lock(&host->add_target_mutex);
+	ret = mutex_lock_interruptible(&host->add_target_mutex);
+	if (ret < 0)
+		goto put;
 
 	ret = srp_parse_options(buf, target);
 	if (ret)
@@ -3455,6 +3457,7 @@ connected:
 out:
 	mutex_unlock(&host->add_target_mutex);
 
+put:
 	scsi_host_put(target->scsi_host);
 	if (ret < 0)
 		scsi_host_put(target->scsi_host);
-- 
cgit v1.2.1


From 626bc02d4d33510b6ecb6f37c577f844cc6cfc57 Mon Sep 17 00:00:00 2001
From: Bart Van Assche <bart.vanassche@sandisk.com>
Date: Mon, 5 Dec 2016 17:18:08 -0800
Subject: mlx5: Use { } instead of { 0 } to init struct

Detected by sparse.

Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Cc: Eli Cohen <eli@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/main.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index cda541ced141..b81736d625fc 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -127,7 +127,7 @@ static int mlx5_netdev_event(struct notifier_block *this,
 
 		if ((upper == ndev || (!upper && ndev == ibdev->roce.netdev))
 		    && ibdev->ib_active) {
-			struct ib_event ibev = {0};
+			struct ib_event ibev = { };
 
 			ibev.device = &ibdev->ib_dev;
 			ibev.event = (event == NETDEV_UP) ?
-- 
cgit v1.2.1


From 3d6bdf1625857ba50f433bd140dada387432f051 Mon Sep 17 00:00:00 2001
From: Bart Van Assche <bart.vanassche@sandisk.com>
Date: Mon, 5 Dec 2016 17:18:27 -0800
Subject: mlx5: Remove a set-but-not-used variable

This has been detected by building the mlx5 driver with W=1.

Fixes: 1a412fb1caa2 ('net/mlx5: Fixes: 1a412fb1caa2 (IB/mlx5: Modify QP
commands via mlx5 ifc')
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Cc: Eli Cohen <eli@mellanox.com>
Acked-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/qp.c | 7 -------
 1 file changed, 7 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c
index a69524fb6032..30f6066b0560 100644
--- a/drivers/infiniband/hw/mlx5/qp.c
+++ b/drivers/infiniband/hw/mlx5/qp.c
@@ -2657,7 +2657,6 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp,
 	struct mlx5_ib_port *mibport = NULL;
 	enum mlx5_qp_state mlx5_cur, mlx5_new;
 	enum mlx5_qp_optpar optpar;
-	int sqd_event;
 	int mlx5_st;
 	int err;
 	u16 op;
@@ -2804,12 +2803,6 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp,
 	if (qp->rq.wqe_cnt && cur_state == IB_QPS_RESET && new_state == IB_QPS_INIT)
 		context->db_rec_addr = cpu_to_be64(qp->db.dma);
 
-	if (cur_state == IB_QPS_RTS && new_state == IB_QPS_SQD	&&
-	    attr_mask & IB_QP_EN_SQD_ASYNC_NOTIFY && attr->en_sqd_async_notify)
-		sqd_event = 1;
-	else
-		sqd_event = 0;
-
 	if (cur_state == IB_QPS_RESET && new_state == IB_QPS_INIT) {
 		u8 port_num = (attr_mask & IB_QP_PORT ? attr->port_num :
 			       qp->port) - 1;
-- 
cgit v1.2.1


From 1974ab9d9d5b6eeafa629f793fdf59958646cb9d Mon Sep 17 00:00:00 2001
From: Bart Van Assche <bart.vanassche@sandisk.com>
Date: Mon, 5 Dec 2016 17:19:52 -0800
Subject: mlx5, calc_sq_size(): Make a debug message more informative

Make it clear that qp->sq.wqe_cnt is not the number of WQEs.

Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Cc: Eli Cohen <eli@mellanox.com>
Acked-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx5/qp.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c
index 30f6066b0560..cc24f2d429b9 100644
--- a/drivers/infiniband/hw/mlx5/qp.c
+++ b/drivers/infiniband/hw/mlx5/qp.c
@@ -407,7 +407,8 @@ static int calc_sq_size(struct mlx5_ib_dev *dev, struct ib_qp_init_attr *attr,
 	wq_size = roundup_pow_of_two(attr->cap.max_send_wr * wqe_size);
 	qp->sq.wqe_cnt = wq_size / MLX5_SEND_WQE_BB;
 	if (qp->sq.wqe_cnt > (1 << MLX5_CAP_GEN(dev->mdev, log_max_qp_sz))) {
-		mlx5_ib_dbg(dev, "wqe count(%d) exceeds limits(%d)\n",
+		mlx5_ib_dbg(dev, "send queue size (%d * %d / %d -> %d) exceeds limits(%d)\n",
+			    attr->cap.max_send_wr, wqe_size, MLX5_SEND_WQE_BB,
 			    qp->sq.wqe_cnt,
 			    1 << MLX5_CAP_GEN(dev->mdev, log_max_qp_sz));
 		return -ENOMEM;
-- 
cgit v1.2.1


From 7ceb740c540dde362b3055ad92c6a38009eb7a83 Mon Sep 17 00:00:00 2001
From: Souptick Joarder <jrdr.linux@gmail.com>
Date: Fri, 2 Dec 2016 00:11:59 +0530
Subject: IB/mthca: Replace pci_pool_alloc by pci_pool_zalloc

In mthca_create_ah(), pci_pool_alloc() followed by memset will be
replaced by pci_pool_zalloc()

Signed-off-by: Souptick joarder <jrdr.linux@gmail.com>
Reviewed-by: Yuval Shaia <yuval.shaia@oracle.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mthca/mthca_av.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mthca/mthca_av.c b/drivers/infiniband/hw/mthca/mthca_av.c
index bcac294042f5..c9f0f364f484 100644
--- a/drivers/infiniband/hw/mthca/mthca_av.c
+++ b/drivers/infiniband/hw/mthca/mthca_av.c
@@ -186,8 +186,8 @@ int mthca_create_ah(struct mthca_dev *dev,
 
 on_hca_fail:
 	if (ah->type == MTHCA_AH_PCI_POOL) {
-		ah->av = pci_pool_alloc(dev->av_table.pool,
-					GFP_ATOMIC, &ah->avdma);
+		ah->av = pci_pool_zalloc(dev->av_table.pool,
+					 GFP_ATOMIC, &ah->avdma);
 		if (!ah->av)
 			return -ENOMEM;
 
@@ -196,8 +196,6 @@ on_hca_fail:
 
 	ah->key = pd->ntmr.ibmr.lkey;
 
-	memset(av, 0, MTHCA_AV_SIZE);
-
 	av->port_pd = cpu_to_be32(pd->pd_num | (ah_attr->port_num << 24));
 	av->g_slid  = ah_attr->src_path_bits;
 	av->dlid    = cpu_to_be16(ah_attr->dlid);
-- 
cgit v1.2.1


From 17069d32a3408e69d257a3fe26f08de0336d958d Mon Sep 17 00:00:00 2001
From: Sebastian Ott <sebott@linux.vnet.ibm.com>
Date: Fri, 2 Dec 2016 14:45:26 +0100
Subject: IB/core: fix unmap_sg argument

__ib_umem_release calls dma_unmap_sg with a different number of
sg_entries than ib_umem_get uses for dma_map_sg. This might cause
trouble for implementations that merge sglist entries and results
in the following dma debug complaint:

DMA-API: device driver frees DMA sg list with different entry
         count [map count=2] [unmap count=1]

Fix it by using the correct value.

Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/umem.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/core/umem.c b/drivers/infiniband/core/umem.c
index 224ad274ea0b..0120e7ff449f 100644
--- a/drivers/infiniband/core/umem.c
+++ b/drivers/infiniband/core/umem.c
@@ -51,7 +51,7 @@ static void __ib_umem_release(struct ib_device *dev, struct ib_umem *umem, int d
 
 	if (umem->nmap > 0)
 		ib_dma_unmap_sg(dev, umem->sg_head.sgl,
-				umem->nmap,
+				umem->npages,
 				DMA_BIDIRECTIONAL);
 
 	for_each_sg(umem->sg_head.sgl, sg, umem->npages, i) {
-- 
cgit v1.2.1


From 6a3a1056d66e6a64446930b0d9de2430d835d38f Mon Sep 17 00:00:00 2001
From: Zhouyi Zhou <zhouzhouyi@gmail.com>
Date: Wed, 7 Dec 2016 15:30:05 +0800
Subject: infiniband: nes: return value of skb_linearize should be handled

Return value of skb_linearize should be handled in function
nes_netdev_start_xmit.

Compiled in x86_64
Signed-off-by: Zhouyi Zhou <yizhouzhou@ict.ac.cn>
Reviewed-by: Yuval Shaia <yuval.shaia@oracle.com>
Reviewed-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/nes/nes_nic.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/nes/nes_nic.c b/drivers/infiniband/hw/nes/nes_nic.c
index 311b4ba1cade..4dcfe669ebad 100644
--- a/drivers/infiniband/hw/nes/nes_nic.c
+++ b/drivers/infiniband/hw/nes/nes_nic.c
@@ -662,10 +662,14 @@ tso_sq_no_longer_full:
 				nesnic->sq_head &= nesnic->sq_size-1;
 			}
 		} else {
-			nesvnic->linearized_skbs++;
 			hoffset = skb_transport_header(skb) - skb->data;
 			nhoffset = skb_network_header(skb) - skb->data;
-			skb_linearize(skb);
+			if (skb_linearize(skb)) {
+				nesvnic->tx_sw_dropped++;
+				kfree_skb(skb);
+				return NETDEV_TX_OK;
+			}
+			nesvnic->linearized_skbs++;
 			skb_set_transport_header(skb, hoffset);
 			skb_set_network_header(skb, nhoffset);
 			if (!nes_nic_send(skb, netdev))
-- 
cgit v1.2.1


From 5b4c9cd7e4790f37b595aeb4bf6fcbf7e3ba9e2c Mon Sep 17 00:00:00 2001
From: Pan Bian <bianpan2016@163.com>
Date: Sat, 3 Dec 2016 21:10:21 +0800
Subject: IB/ocrdma: fix bad initialization

In function ocrdma_mbx_create_ah_tbl(), returns the value of status on
errors. However, because status is initialized with 0, 0 will be
returned even if on error paths. This patch initialize status with
"-ENOMEM".

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=188831

Signed-off-by: Pan Bian <bianpan2016@163.com>
Reviewed-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/ocrdma/ocrdma_hw.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_hw.c b/drivers/infiniband/hw/ocrdma/ocrdma_hw.c
index 67fc0b6857e1..6876a7129e81 100644
--- a/drivers/infiniband/hw/ocrdma/ocrdma_hw.c
+++ b/drivers/infiniband/hw/ocrdma/ocrdma_hw.c
@@ -1642,7 +1642,7 @@ static int ocrdma_build_q_conf(u32 *num_entries, int entry_size,
 static int ocrdma_mbx_create_ah_tbl(struct ocrdma_dev *dev)
 {
 	int i;
-	int status = 0;
+	int status = -ENOMEM;
 	int max_ah;
 	struct ocrdma_create_ah_tbl *cmd;
 	struct ocrdma_create_ah_tbl_rsp *rsp;
-- 
cgit v1.2.1


From 46d0703fac3ffa12ec36f22f386d96d0f474c9c2 Mon Sep 17 00:00:00 2001
From: Pan Bian <bianpan2016@163.com>
Date: Sun, 4 Dec 2016 14:45:38 +0800
Subject: IB/mlx4: fix improper return value

If uhw->inlen is non-zero, the value of variable err is 0 if the copy
succeeds. Then, if kzalloc() or kmalloc() returns a NULL pointer, it
will return 0 to the callers. As a result, the callers cannot detect the
errors. This patch fixes the bug, assign "-ENOMEM" to err before the
NULL pointer checks, and remove the initialization of err at the
beginning.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=189031
Signed-off-by: Pan Bian <bianpan2016@163.com>
Reviewed-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/main.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index b597e8227591..a87c395e185a 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -430,7 +430,7 @@ static int mlx4_ib_query_device(struct ib_device *ibdev,
 	struct mlx4_ib_dev *dev = to_mdev(ibdev);
 	struct ib_smp *in_mad  = NULL;
 	struct ib_smp *out_mad = NULL;
-	int err = -ENOMEM;
+	int err;
 	int have_ib_ports;
 	struct mlx4_uverbs_ex_query_device cmd;
 	struct mlx4_uverbs_ex_query_device_resp resp = {.comp_mask = 0};
@@ -455,6 +455,7 @@ static int mlx4_ib_query_device(struct ib_device *ibdev,
 		sizeof(resp.response_length);
 	in_mad  = kzalloc(sizeof *in_mad, GFP_KERNEL);
 	out_mad = kmalloc(sizeof *out_mad, GFP_KERNEL);
+	err = -ENOMEM;
 	if (!in_mad || !out_mad)
 		goto out;
 
-- 
cgit v1.2.1


From 29c8d9eba550c6d73d17cc1618a9f5f2a7345aa1 Mon Sep 17 00:00:00 2001
From: Adit Ranadive <aditr@vmware.com>
Date: Sun, 2 Oct 2016 19:10:22 -0700
Subject: IB: Add vmw_pvrdma driver

This patch series adds a driver for a paravirtual RDMA device. The
device is developed for VMware's Virtual Machines and allows existing RDMA
applications to continue to use existing Verbs API when deployed in VMs
on ESXi. We recently did a presentation in the OFA Workshop [1] regarding
this device.

Description and RDMA Support
============================
The virtual device is exposed as a dual function PCIe device. One part
is a virtual network device (VMXNet3) which provides networking properties
like MAC, IP addresses to the RDMA part of the device. The networking
properties are used to register GIDs required by RDMA applications to
communicate.

These patches add support and the all required infrastructure for
letting applications use such a device. We support the mandatory Verbs API as
well as the base memory management extensions (Local Inv, Send with Inv and
Fast Register Work Requests). We currently support both Reliable Connected
and Unreliable Datagram QPs but do not support Shared Receive Queues
(SRQs).

Also, we support the following types of Work Requests:
 o Send/Receive (with or without Immediate Data)
 o RDMA Write (with or without Immediate Data)
 o RDMA Read
 o Local Invalidate
 o Send with Invalidate
 o Fast Register Work Requests

This version only adds support for version 1 of RoCE. We will add RoCEv2
support in a future patch. We do support registration of both MAC-based
and IP-based GIDs. I have also created a git tree for our user-level driver
[2].

Testing
=======
We have tested this internally for various types of Guest OS - Red Hat,
Centos, Ubuntu 12.04/14.04/16.04, Oracle Enterprise Linux, SLES 12
using backported versions of this driver. The tests included several
runs of the performance tests (included with OFED), Intel MPI PingPong
benchmark on OpenMPI, krping for FRWRs. Mellanox has been kind enough
to test the backported version of the driver internally on their hardware
using a VMware provided ESX build. I have also applied and tested this
with Doug's k.o/for-4.9 branch (commit 5603910b). Note, that this patch
series should be applied all together. I split out the commits so that
it may be easier to review.

PVRDMA Resources
================
[1] OFA Workshop Presentation -
https://openfabrics.org/images/eventpresos/2016presentations/102parardma.pdf

[2] Libpvrdma User-level library -
http://git.openfabrics.org/?p=~aditr/libpvrdma.git;a=summary

Reviewed-by: Jorgen Hansen <jhansen@vmware.com>
Reviewed-by: George Zhang <georgezhang@vmware.com>
Reviewed-by: Aditya Sarwade <asarwade@vmware.com>
Reviewed-by: Bryan Tan <bryantan@vmware.com>
Reviewed-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Adit Ranadive <aditr@vmware.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/Kconfig                         |    1 +
 drivers/infiniband/hw/Makefile                     |    1 +
 drivers/infiniband/hw/vmw_pvrdma/Kconfig           |    7 +
 drivers/infiniband/hw/vmw_pvrdma/Makefile          |    3 +
 drivers/infiniband/hw/vmw_pvrdma/pvrdma.h          |  474 ++++++++
 drivers/infiniband/hw/vmw_pvrdma/pvrdma_cmd.c      |  119 ++
 drivers/infiniband/hw/vmw_pvrdma/pvrdma_cq.c       |  425 +++++++
 drivers/infiniband/hw/vmw_pvrdma/pvrdma_dev_api.h  |  586 ++++++++++
 drivers/infiniband/hw/vmw_pvrdma/pvrdma_doorbell.c |  127 ++
 drivers/infiniband/hw/vmw_pvrdma/pvrdma_main.c     | 1211 ++++++++++++++++++++
 drivers/infiniband/hw/vmw_pvrdma/pvrdma_misc.c     |  304 +++++
 drivers/infiniband/hw/vmw_pvrdma/pvrdma_mr.c       |  334 ++++++
 drivers/infiniband/hw/vmw_pvrdma/pvrdma_qp.c       |  972 ++++++++++++++++
 drivers/infiniband/hw/vmw_pvrdma/pvrdma_ring.h     |  131 +++
 drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.c    |  579 ++++++++++
 drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.h    |  436 +++++++
 16 files changed, 5710 insertions(+)
 create mode 100644 drivers/infiniband/hw/vmw_pvrdma/Kconfig
 create mode 100644 drivers/infiniband/hw/vmw_pvrdma/Makefile
 create mode 100644 drivers/infiniband/hw/vmw_pvrdma/pvrdma.h
 create mode 100644 drivers/infiniband/hw/vmw_pvrdma/pvrdma_cmd.c
 create mode 100644 drivers/infiniband/hw/vmw_pvrdma/pvrdma_cq.c
 create mode 100644 drivers/infiniband/hw/vmw_pvrdma/pvrdma_dev_api.h
 create mode 100644 drivers/infiniband/hw/vmw_pvrdma/pvrdma_doorbell.c
 create mode 100644 drivers/infiniband/hw/vmw_pvrdma/pvrdma_main.c
 create mode 100644 drivers/infiniband/hw/vmw_pvrdma/pvrdma_misc.c
 create mode 100644 drivers/infiniband/hw/vmw_pvrdma/pvrdma_mr.c
 create mode 100644 drivers/infiniband/hw/vmw_pvrdma/pvrdma_qp.c
 create mode 100644 drivers/infiniband/hw/vmw_pvrdma/pvrdma_ring.h
 create mode 100644 drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.c
 create mode 100644 drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.h

(limited to 'drivers/infiniband')

diff --git a/drivers/infiniband/Kconfig b/drivers/infiniband/Kconfig
index fb3fb89640e5..670917387eda 100644
--- a/drivers/infiniband/Kconfig
+++ b/drivers/infiniband/Kconfig
@@ -73,6 +73,7 @@ source "drivers/infiniband/hw/mlx4/Kconfig"
 source "drivers/infiniband/hw/mlx5/Kconfig"
 source "drivers/infiniband/hw/nes/Kconfig"
 source "drivers/infiniband/hw/ocrdma/Kconfig"
+source "drivers/infiniband/hw/vmw_pvrdma/Kconfig"
 source "drivers/infiniband/hw/usnic/Kconfig"
 source "drivers/infiniband/hw/hns/Kconfig"
 
diff --git a/drivers/infiniband/hw/Makefile b/drivers/infiniband/hw/Makefile
index e7a5ed9f6f3f..ed553de2ca12 100644
--- a/drivers/infiniband/hw/Makefile
+++ b/drivers/infiniband/hw/Makefile
@@ -7,6 +7,7 @@ obj-$(CONFIG_MLX4_INFINIBAND)		+= mlx4/
 obj-$(CONFIG_MLX5_INFINIBAND)		+= mlx5/
 obj-$(CONFIG_INFINIBAND_NES)		+= nes/
 obj-$(CONFIG_INFINIBAND_OCRDMA)		+= ocrdma/
+obj-$(CONFIG_INFINIBAND_VMWARE_PVRDMA)	+= vmw_pvrdma/
 obj-$(CONFIG_INFINIBAND_USNIC)		+= usnic/
 obj-$(CONFIG_INFINIBAND_HFI1)		+= hfi1/
 obj-$(CONFIG_INFINIBAND_HNS)		+= hns/
diff --git a/drivers/infiniband/hw/vmw_pvrdma/Kconfig b/drivers/infiniband/hw/vmw_pvrdma/Kconfig
new file mode 100644
index 000000000000..5a9790ac0ede
--- /dev/null
+++ b/drivers/infiniband/hw/vmw_pvrdma/Kconfig
@@ -0,0 +1,7 @@
+config INFINIBAND_VMWARE_PVRDMA
+	tristate "VMware Paravirtualized RDMA Driver"
+	depends on NETDEVICES && ETHERNET && PCI && INET && VMXNET3
+	---help---
+	  This driver provides low-level support for VMware Paravirtual
+	  RDMA adapter. It interacts with the VMXNet3 driver to provide
+	  Ethernet capabilities.
diff --git a/drivers/infiniband/hw/vmw_pvrdma/Makefile b/drivers/infiniband/hw/vmw_pvrdma/Makefile
new file mode 100644
index 000000000000..0194ed19f542
--- /dev/null
+++ b/drivers/infiniband/hw/vmw_pvrdma/Makefile
@@ -0,0 +1,3 @@
+obj-$(CONFIG_INFINIBAND_VMWARE_PVRDMA) += vmw_pvrdma.o
+
+vmw_pvrdma-y := pvrdma_cmd.o pvrdma_cq.o pvrdma_doorbell.o pvrdma_main.o pvrdma_misc.o pvrdma_mr.o pvrdma_qp.o pvrdma_verbs.o
diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma.h b/drivers/infiniband/hw/vmw_pvrdma/pvrdma.h
new file mode 100644
index 000000000000..71e1d55d69d6
--- /dev/null
+++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma.h
@@ -0,0 +1,474 @@
+/*
+ * Copyright (c) 2012-2016 VMware, Inc.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of EITHER the GNU General Public License
+ * version 2 as published by the Free Software Foundation or the BSD
+ * 2-Clause License. This program is distributed in the hope that it
+ * will be useful, but WITHOUT ANY WARRANTY; WITHOUT EVEN THE IMPLIED
+ * WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
+ * See the GNU General Public License version 2 for more details at
+ * http://www.gnu.org/licenses/old-licenses/gpl-2.0.en.html.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program available in the file COPYING in the main
+ * directory of this source tree.
+ *
+ * The BSD 2-Clause License
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+ * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
+ * COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT,
+ * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#ifndef __PVRDMA_H__
+#define __PVRDMA_H__
+
+#include <linux/compiler.h>
+#include <linux/interrupt.h>
+#include <linux/list.h>
+#include <linux/mutex.h>
+#include <linux/pci.h>
+#include <linux/semaphore.h>
+#include <linux/workqueue.h>
+#include <rdma/ib_umem.h>
+#include <rdma/ib_verbs.h>
+#include <rdma/vmw_pvrdma-abi.h>
+
+#include "pvrdma_ring.h"
+#include "pvrdma_dev_api.h"
+#include "pvrdma_verbs.h"
+
+/* NOT the same as BIT_MASK(). */
+#define PVRDMA_MASK(n) ((n << 1) - 1)
+
+/*
+ * VMware PVRDMA PCI device id.
+ */
+#define PCI_DEVICE_ID_VMWARE_PVRDMA	0x0820
+
+struct pvrdma_dev;
+
+struct pvrdma_page_dir {
+	dma_addr_t dir_dma;
+	u64 *dir;
+	int ntables;
+	u64 **tables;
+	u64 npages;
+	void **pages;
+};
+
+struct pvrdma_cq {
+	struct ib_cq ibcq;
+	int offset;
+	spinlock_t cq_lock; /* Poll lock. */
+	struct pvrdma_uar_map *uar;
+	struct ib_umem *umem;
+	struct pvrdma_ring_state *ring_state;
+	struct pvrdma_page_dir pdir;
+	u32 cq_handle;
+	bool is_kernel;
+	atomic_t refcnt;
+	wait_queue_head_t wait;
+};
+
+struct pvrdma_id_table {
+	u32 last;
+	u32 top;
+	u32 max;
+	u32 mask;
+	spinlock_t lock; /* Table lock. */
+	unsigned long *table;
+};
+
+struct pvrdma_uar_map {
+	unsigned long pfn;
+	void __iomem *map;
+	int index;
+};
+
+struct pvrdma_uar_table {
+	struct pvrdma_id_table tbl;
+	int size;
+};
+
+struct pvrdma_ucontext {
+	struct ib_ucontext ibucontext;
+	struct pvrdma_dev *dev;
+	struct pvrdma_uar_map uar;
+	u64 ctx_handle;
+};
+
+struct pvrdma_pd {
+	struct ib_pd ibpd;
+	u32 pdn;
+	u32 pd_handle;
+	int privileged;
+};
+
+struct pvrdma_mr {
+	u32 mr_handle;
+	u64 iova;
+	u64 size;
+};
+
+struct pvrdma_user_mr {
+	struct ib_mr ibmr;
+	struct ib_umem *umem;
+	struct pvrdma_mr mmr;
+	struct pvrdma_page_dir pdir;
+	u64 *pages;
+	u32 npages;
+	u32 max_pages;
+	u32 page_shift;
+};
+
+struct pvrdma_wq {
+	struct pvrdma_ring *ring;
+	spinlock_t lock; /* Work queue lock. */
+	int wqe_cnt;
+	int wqe_size;
+	int max_sg;
+	int offset;
+};
+
+struct pvrdma_ah {
+	struct ib_ah ibah;
+	struct pvrdma_av av;
+};
+
+struct pvrdma_qp {
+	struct ib_qp ibqp;
+	u32 qp_handle;
+	u32 qkey;
+	struct pvrdma_wq sq;
+	struct pvrdma_wq rq;
+	struct ib_umem *rumem;
+	struct ib_umem *sumem;
+	struct pvrdma_page_dir pdir;
+	int npages;
+	int npages_send;
+	int npages_recv;
+	u32 flags;
+	u8 port;
+	u8 state;
+	bool is_kernel;
+	struct mutex mutex; /* QP state mutex. */
+	atomic_t refcnt;
+	wait_queue_head_t wait;
+};
+
+struct pvrdma_dev {
+	/* PCI device-related information. */
+	struct ib_device ib_dev;
+	struct pci_dev *pdev;
+	void __iomem *regs;
+	struct pvrdma_device_shared_region *dsr; /* Shared region pointer */
+	dma_addr_t dsrbase; /* Shared region base address */
+	void *cmd_slot;
+	void *resp_slot;
+	unsigned long flags;
+	struct list_head device_link;
+
+	/* Locking and interrupt information. */
+	spinlock_t cmd_lock; /* Command lock. */
+	struct semaphore cmd_sema;
+	struct completion cmd_done;
+	struct {
+		enum pvrdma_intr_type type; /* Intr type */
+		struct msix_entry msix_entry[PVRDMA_MAX_INTERRUPTS];
+		irq_handler_t handler[PVRDMA_MAX_INTERRUPTS];
+		u8 enabled[PVRDMA_MAX_INTERRUPTS];
+		u8 size;
+	} intr;
+
+	/* RDMA-related device information. */
+	union ib_gid *sgid_tbl;
+	struct pvrdma_ring_state *async_ring_state;
+	struct pvrdma_page_dir async_pdir;
+	struct pvrdma_ring_state *cq_ring_state;
+	struct pvrdma_page_dir cq_pdir;
+	struct pvrdma_cq **cq_tbl;
+	spinlock_t cq_tbl_lock;
+	struct pvrdma_qp **qp_tbl;
+	spinlock_t qp_tbl_lock;
+	struct pvrdma_uar_table uar_table;
+	struct pvrdma_uar_map driver_uar;
+	__be64 sys_image_guid;
+	spinlock_t desc_lock; /* Device modification lock. */
+	u32 port_cap_mask;
+	struct mutex port_mutex; /* Port modification mutex. */
+	bool ib_active;
+	atomic_t num_qps;
+	atomic_t num_cqs;
+	atomic_t num_pds;
+	atomic_t num_ahs;
+
+	/* Network device information. */
+	struct net_device *netdev;
+	struct notifier_block nb_netdev;
+};
+
+struct pvrdma_netdevice_work {
+	struct work_struct work;
+	struct net_device *event_netdev;
+	unsigned long event;
+};
+
+static inline struct pvrdma_dev *to_vdev(struct ib_device *ibdev)
+{
+	return container_of(ibdev, struct pvrdma_dev, ib_dev);
+}
+
+static inline struct
+pvrdma_ucontext *to_vucontext(struct ib_ucontext *ibucontext)
+{
+	return container_of(ibucontext, struct pvrdma_ucontext, ibucontext);
+}
+
+static inline struct pvrdma_pd *to_vpd(struct ib_pd *ibpd)
+{
+	return container_of(ibpd, struct pvrdma_pd, ibpd);
+}
+
+static inline struct pvrdma_cq *to_vcq(struct ib_cq *ibcq)
+{
+	return container_of(ibcq, struct pvrdma_cq, ibcq);
+}
+
+static inline struct pvrdma_user_mr *to_vmr(struct ib_mr *ibmr)
+{
+	return container_of(ibmr, struct pvrdma_user_mr, ibmr);
+}
+
+static inline struct pvrdma_qp *to_vqp(struct ib_qp *ibqp)
+{
+	return container_of(ibqp, struct pvrdma_qp, ibqp);
+}
+
+static inline struct pvrdma_ah *to_vah(struct ib_ah *ibah)
+{
+	return container_of(ibah, struct pvrdma_ah, ibah);
+}
+
+static inline void pvrdma_write_reg(struct pvrdma_dev *dev, u32 reg, u32 val)
+{
+	writel(cpu_to_le32(val), dev->regs + reg);
+}
+
+static inline u32 pvrdma_read_reg(struct pvrdma_dev *dev, u32 reg)
+{
+	return le32_to_cpu(readl(dev->regs + reg));
+}
+
+static inline void pvrdma_write_uar_cq(struct pvrdma_dev *dev, u32 val)
+{
+	writel(cpu_to_le32(val), dev->driver_uar.map + PVRDMA_UAR_CQ_OFFSET);
+}
+
+static inline void pvrdma_write_uar_qp(struct pvrdma_dev *dev, u32 val)
+{
+	writel(cpu_to_le32(val), dev->driver_uar.map + PVRDMA_UAR_QP_OFFSET);
+}
+
+static inline void *pvrdma_page_dir_get_ptr(struct pvrdma_page_dir *pdir,
+					    u64 offset)
+{
+	return pdir->pages[offset / PAGE_SIZE] + (offset % PAGE_SIZE);
+}
+
+static inline enum pvrdma_mtu ib_mtu_to_pvrdma(enum ib_mtu mtu)
+{
+	return (enum pvrdma_mtu)mtu;
+}
+
+static inline enum ib_mtu pvrdma_mtu_to_ib(enum pvrdma_mtu mtu)
+{
+	return (enum ib_mtu)mtu;
+}
+
+static inline enum pvrdma_port_state ib_port_state_to_pvrdma(
+					enum ib_port_state state)
+{
+	return (enum pvrdma_port_state)state;
+}
+
+static inline enum ib_port_state pvrdma_port_state_to_ib(
+					enum pvrdma_port_state state)
+{
+	return (enum ib_port_state)state;
+}
+
+static inline int ib_port_cap_flags_to_pvrdma(int flags)
+{
+	return flags & PVRDMA_MASK(PVRDMA_PORT_CAP_FLAGS_MAX);
+}
+
+static inline int pvrdma_port_cap_flags_to_ib(int flags)
+{
+	return flags;
+}
+
+static inline enum pvrdma_port_width ib_port_width_to_pvrdma(
+					enum ib_port_width width)
+{
+	return (enum pvrdma_port_width)width;
+}
+
+static inline enum ib_port_width pvrdma_port_width_to_ib(
+					enum pvrdma_port_width width)
+{
+	return (enum ib_port_width)width;
+}
+
+static inline enum pvrdma_port_speed ib_port_speed_to_pvrdma(
+					enum ib_port_speed speed)
+{
+	return (enum pvrdma_port_speed)speed;
+}
+
+static inline enum ib_port_speed pvrdma_port_speed_to_ib(
+					enum pvrdma_port_speed speed)
+{
+	return (enum ib_port_speed)speed;
+}
+
+static inline int pvrdma_qp_attr_mask_to_ib(int attr_mask)
+{
+	return attr_mask;
+}
+
+static inline int ib_qp_attr_mask_to_pvrdma(int attr_mask)
+{
+	return attr_mask & PVRDMA_MASK(PVRDMA_QP_ATTR_MASK_MAX);
+}
+
+static inline enum pvrdma_mig_state ib_mig_state_to_pvrdma(
+					enum ib_mig_state state)
+{
+	return (enum pvrdma_mig_state)state;
+}
+
+static inline enum ib_mig_state pvrdma_mig_state_to_ib(
+					enum pvrdma_mig_state state)
+{
+	return (enum ib_mig_state)state;
+}
+
+static inline int ib_access_flags_to_pvrdma(int flags)
+{
+	return flags;
+}
+
+static inline int pvrdma_access_flags_to_ib(int flags)
+{
+	return flags & PVRDMA_MASK(PVRDMA_ACCESS_FLAGS_MAX);
+}
+
+static inline enum pvrdma_qp_type ib_qp_type_to_pvrdma(enum ib_qp_type type)
+{
+	return (enum pvrdma_qp_type)type;
+}
+
+static inline enum ib_qp_type pvrdma_qp_type_to_ib(enum pvrdma_qp_type type)
+{
+	return (enum ib_qp_type)type;
+}
+
+static inline enum pvrdma_qp_state ib_qp_state_to_pvrdma(enum ib_qp_state state)
+{
+	return (enum pvrdma_qp_state)state;
+}
+
+static inline enum ib_qp_state pvrdma_qp_state_to_ib(enum pvrdma_qp_state state)
+{
+	return (enum ib_qp_state)state;
+}
+
+static inline enum pvrdma_wr_opcode ib_wr_opcode_to_pvrdma(enum ib_wr_opcode op)
+{
+	return (enum pvrdma_wr_opcode)op;
+}
+
+static inline enum ib_wc_status pvrdma_wc_status_to_ib(
+					enum pvrdma_wc_status status)
+{
+	return (enum ib_wc_status)status;
+}
+
+static inline int pvrdma_wc_opcode_to_ib(int opcode)
+{
+	return opcode;
+}
+
+static inline int pvrdma_wc_flags_to_ib(int flags)
+{
+	return flags;
+}
+
+static inline int ib_send_flags_to_pvrdma(int flags)
+{
+	return flags & PVRDMA_MASK(PVRDMA_SEND_FLAGS_MAX);
+}
+
+void pvrdma_qp_cap_to_ib(struct ib_qp_cap *dst,
+			 const struct pvrdma_qp_cap *src);
+void ib_qp_cap_to_pvrdma(struct pvrdma_qp_cap *dst,
+			 const struct ib_qp_cap *src);
+void pvrdma_gid_to_ib(union ib_gid *dst, const union pvrdma_gid *src);
+void ib_gid_to_pvrdma(union pvrdma_gid *dst, const union ib_gid *src);
+void pvrdma_global_route_to_ib(struct ib_global_route *dst,
+			       const struct pvrdma_global_route *src);
+void ib_global_route_to_pvrdma(struct pvrdma_global_route *dst,
+			       const struct ib_global_route *src);
+void pvrdma_ah_attr_to_ib(struct ib_ah_attr *dst,
+			  const struct pvrdma_ah_attr *src);
+void ib_ah_attr_to_pvrdma(struct pvrdma_ah_attr *dst,
+			  const struct ib_ah_attr *src);
+
+int pvrdma_uar_table_init(struct pvrdma_dev *dev);
+void pvrdma_uar_table_cleanup(struct pvrdma_dev *dev);
+
+int pvrdma_uar_alloc(struct pvrdma_dev *dev, struct pvrdma_uar_map *uar);
+void pvrdma_uar_free(struct pvrdma_dev *dev, struct pvrdma_uar_map *uar);
+
+void _pvrdma_flush_cqe(struct pvrdma_qp *qp, struct pvrdma_cq *cq);
+
+int pvrdma_page_dir_init(struct pvrdma_dev *dev, struct pvrdma_page_dir *pdir,
+			 u64 npages, bool alloc_pages);
+void pvrdma_page_dir_cleanup(struct pvrdma_dev *dev,
+			     struct pvrdma_page_dir *pdir);
+int pvrdma_page_dir_insert_dma(struct pvrdma_page_dir *pdir, u64 idx,
+			       dma_addr_t daddr);
+int pvrdma_page_dir_insert_umem(struct pvrdma_page_dir *pdir,
+				struct ib_umem *umem, u64 offset);
+dma_addr_t pvrdma_page_dir_get_dma(struct pvrdma_page_dir *pdir, u64 idx);
+int pvrdma_page_dir_insert_page_list(struct pvrdma_page_dir *pdir,
+				     u64 *page_list, int num_pages);
+
+int pvrdma_cmd_post(struct pvrdma_dev *dev, union pvrdma_cmd_req *req,
+		    union pvrdma_cmd_resp *rsp, unsigned resp_code);
+
+#endif /* __PVRDMA_H__ */
diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_cmd.c b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_cmd.c
new file mode 100644
index 000000000000..4a78c537d8a1
--- /dev/null
+++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_cmd.c
@@ -0,0 +1,119 @@
+/*
+ * Copyright (c) 2012-2016 VMware, Inc.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of EITHER the GNU General Public License
+ * version 2 as published by the Free Software Foundation or the BSD
+ * 2-Clause License. This program is distributed in the hope that it
+ * will be useful, but WITHOUT ANY WARRANTY; WITHOUT EVEN THE IMPLIED
+ * WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
+ * See the GNU General Public License version 2 for more details at
+ * http://www.gnu.org/licenses/old-licenses/gpl-2.0.en.html.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program available in the file COPYING in the main
+ * directory of this source tree.
+ *
+ * The BSD 2-Clause License
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+ * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
+ * COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT,
+ * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <linux/list.h>
+
+#include "pvrdma.h"
+
+#define PVRDMA_CMD_TIMEOUT	10000 /* ms */
+
+static inline int pvrdma_cmd_recv(struct pvrdma_dev *dev,
+				  union pvrdma_cmd_resp *resp,
+				  unsigned resp_code)
+{
+	int err;
+
+	dev_dbg(&dev->pdev->dev, "receive response from device\n");
+
+	err = wait_for_completion_interruptible_timeout(&dev->cmd_done,
+			msecs_to_jiffies(PVRDMA_CMD_TIMEOUT));
+	if (err == 0 || err == -ERESTARTSYS) {
+		dev_warn(&dev->pdev->dev,
+			 "completion timeout or interrupted\n");
+		return -ETIMEDOUT;
+	}
+
+	spin_lock(&dev->cmd_lock);
+	memcpy(resp, dev->resp_slot, sizeof(*resp));
+	spin_unlock(&dev->cmd_lock);
+
+	if (resp->hdr.ack != resp_code) {
+		dev_warn(&dev->pdev->dev,
+			 "unknown response %#x expected %#x\n",
+			 resp->hdr.ack, resp_code);
+		return -EFAULT;
+	}
+
+	return 0;
+}
+
+int
+pvrdma_cmd_post(struct pvrdma_dev *dev, union pvrdma_cmd_req *req,
+		union pvrdma_cmd_resp *resp, unsigned resp_code)
+{
+	int err;
+
+	dev_dbg(&dev->pdev->dev, "post request to device\n");
+
+	/* Serializiation */
+	down(&dev->cmd_sema);
+
+	BUILD_BUG_ON(sizeof(union pvrdma_cmd_req) !=
+		     sizeof(struct pvrdma_cmd_modify_qp));
+
+	spin_lock(&dev->cmd_lock);
+	memcpy(dev->cmd_slot, req, sizeof(*req));
+	spin_unlock(&dev->cmd_lock);
+
+	init_completion(&dev->cmd_done);
+	pvrdma_write_reg(dev, PVRDMA_REG_REQUEST, 0);
+
+	/* Make sure the request is written before reading status. */
+	mb();
+
+	err = pvrdma_read_reg(dev, PVRDMA_REG_ERR);
+	if (err == 0) {
+		if (resp != NULL)
+			err = pvrdma_cmd_recv(dev, resp, resp_code);
+	} else {
+		dev_warn(&dev->pdev->dev,
+			 "failed to write request error reg: %d\n", err);
+		err = -EFAULT;
+	}
+
+	up(&dev->cmd_sema);
+
+	return err;
+}
diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_cq.c b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_cq.c
new file mode 100644
index 000000000000..e429ca5b16aa
--- /dev/null
+++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_cq.c
@@ -0,0 +1,425 @@
+/*
+ * Copyright (c) 2012-2016 VMware, Inc.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of EITHER the GNU General Public License
+ * version 2 as published by the Free Software Foundation or the BSD
+ * 2-Clause License. This program is distributed in the hope that it
+ * will be useful, but WITHOUT ANY WARRANTY; WITHOUT EVEN THE IMPLIED
+ * WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
+ * See the GNU General Public License version 2 for more details at
+ * http://www.gnu.org/licenses/old-licenses/gpl-2.0.en.html.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program available in the file COPYING in the main
+ * directory of this source tree.
+ *
+ * The BSD 2-Clause License
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+ * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
+ * COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT,
+ * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <asm/page.h>
+#include <linux/io.h>
+#include <linux/wait.h>
+#include <rdma/ib_addr.h>
+#include <rdma/ib_smi.h>
+#include <rdma/ib_user_verbs.h>
+
+#include "pvrdma.h"
+
+/**
+ * pvrdma_req_notify_cq - request notification for a completion queue
+ * @ibcq: the completion queue
+ * @notify_flags: notification flags
+ *
+ * @return: 0 for success.
+ */
+int pvrdma_req_notify_cq(struct ib_cq *ibcq,
+			 enum ib_cq_notify_flags notify_flags)
+{
+	struct pvrdma_dev *dev = to_vdev(ibcq->device);
+	struct pvrdma_cq *cq = to_vcq(ibcq);
+	u32 val = cq->cq_handle;
+
+	val |= (notify_flags & IB_CQ_SOLICITED_MASK) == IB_CQ_SOLICITED ?
+		PVRDMA_UAR_CQ_ARM_SOL : PVRDMA_UAR_CQ_ARM;
+
+	pvrdma_write_uar_cq(dev, val);
+
+	return 0;
+}
+
+/**
+ * pvrdma_create_cq - create completion queue
+ * @ibdev: the device
+ * @attr: completion queue attributes
+ * @context: user context
+ * @udata: user data
+ *
+ * @return: ib_cq completion queue pointer on success,
+ *          otherwise returns negative errno.
+ */
+struct ib_cq *pvrdma_create_cq(struct ib_device *ibdev,
+			       const struct ib_cq_init_attr *attr,
+			       struct ib_ucontext *context,
+			       struct ib_udata *udata)
+{
+	int entries = attr->cqe;
+	struct pvrdma_dev *dev = to_vdev(ibdev);
+	struct pvrdma_cq *cq;
+	int ret;
+	int npages;
+	unsigned long flags;
+	union pvrdma_cmd_req req;
+	union pvrdma_cmd_resp rsp;
+	struct pvrdma_cmd_create_cq *cmd = &req.create_cq;
+	struct pvrdma_cmd_create_cq_resp *resp = &rsp.create_cq_resp;
+	struct pvrdma_create_cq ucmd;
+
+	BUILD_BUG_ON(sizeof(struct pvrdma_cqe) != 64);
+
+	entries = roundup_pow_of_two(entries);
+	if (entries < 1 || entries > dev->dsr->caps.max_cqe)
+		return ERR_PTR(-EINVAL);
+
+	if (!atomic_add_unless(&dev->num_cqs, 1, dev->dsr->caps.max_cq))
+		return ERR_PTR(-ENOMEM);
+
+	cq = kzalloc(sizeof(*cq), GFP_KERNEL);
+	if (!cq) {
+		atomic_dec(&dev->num_cqs);
+		return ERR_PTR(-ENOMEM);
+	}
+
+	cq->ibcq.cqe = entries;
+
+	if (context) {
+		if (ib_copy_from_udata(&ucmd, udata, sizeof(ucmd))) {
+			ret = -EFAULT;
+			goto err_cq;
+		}
+
+		cq->umem = ib_umem_get(context, ucmd.buf_addr, ucmd.buf_size,
+				       IB_ACCESS_LOCAL_WRITE, 1);
+		if (IS_ERR(cq->umem)) {
+			ret = PTR_ERR(cq->umem);
+			goto err_cq;
+		}
+
+		npages = ib_umem_page_count(cq->umem);
+	} else {
+		cq->is_kernel = true;
+
+		/* One extra page for shared ring state */
+		npages = 1 + (entries * sizeof(struct pvrdma_cqe) +
+			      PAGE_SIZE - 1) / PAGE_SIZE;
+
+		/* Skip header page. */
+		cq->offset = PAGE_SIZE;
+	}
+
+	if (npages < 0 || npages > PVRDMA_PAGE_DIR_MAX_PAGES) {
+		dev_warn(&dev->pdev->dev,
+			 "overflow pages in completion queue\n");
+		ret = -EINVAL;
+		goto err_umem;
+	}
+
+	ret = pvrdma_page_dir_init(dev, &cq->pdir, npages, cq->is_kernel);
+	if (ret) {
+		dev_warn(&dev->pdev->dev,
+			 "could not allocate page directory\n");
+		goto err_umem;
+	}
+
+	/* Ring state is always the first page. Set in library for user cq. */
+	if (cq->is_kernel)
+		cq->ring_state = cq->pdir.pages[0];
+	else
+		pvrdma_page_dir_insert_umem(&cq->pdir, cq->umem, 0);
+
+	atomic_set(&cq->refcnt, 1);
+	init_waitqueue_head(&cq->wait);
+	spin_lock_init(&cq->cq_lock);
+
+	memset(cmd, 0, sizeof(*cmd));
+	cmd->hdr.cmd = PVRDMA_CMD_CREATE_CQ;
+	cmd->nchunks = npages;
+	cmd->ctx_handle = (context) ?
+		(u64)to_vucontext(context)->ctx_handle : 0;
+	cmd->cqe = entries;
+	cmd->pdir_dma = cq->pdir.dir_dma;
+	ret = pvrdma_cmd_post(dev, &req, &rsp, PVRDMA_CMD_CREATE_CQ_RESP);
+	if (ret < 0) {
+		dev_warn(&dev->pdev->dev,
+			 "could not create completion queue, error: %d\n", ret);
+		goto err_page_dir;
+	}
+
+	cq->ibcq.cqe = resp->cqe;
+	cq->cq_handle = resp->cq_handle;
+	spin_lock_irqsave(&dev->cq_tbl_lock, flags);
+	dev->cq_tbl[cq->cq_handle % dev->dsr->caps.max_cq] = cq;
+	spin_unlock_irqrestore(&dev->cq_tbl_lock, flags);
+
+	if (context) {
+		cq->uar = &(to_vucontext(context)->uar);
+
+		/* Copy udata back. */
+		if (ib_copy_to_udata(udata, &cq->cq_handle, sizeof(__u32))) {
+			dev_warn(&dev->pdev->dev,
+				 "failed to copy back udata\n");
+			pvrdma_destroy_cq(&cq->ibcq);
+			return ERR_PTR(-EINVAL);
+		}
+	}
+
+	return &cq->ibcq;
+
+err_page_dir:
+	pvrdma_page_dir_cleanup(dev, &cq->pdir);
+err_umem:
+	if (context)
+		ib_umem_release(cq->umem);
+err_cq:
+	atomic_dec(&dev->num_cqs);
+	kfree(cq);
+
+	return ERR_PTR(ret);
+}
+
+static void pvrdma_free_cq(struct pvrdma_dev *dev, struct pvrdma_cq *cq)
+{
+	atomic_dec(&cq->refcnt);
+	wait_event(cq->wait, !atomic_read(&cq->refcnt));
+
+	if (!cq->is_kernel)
+		ib_umem_release(cq->umem);
+
+	pvrdma_page_dir_cleanup(dev, &cq->pdir);
+	kfree(cq);
+}
+
+/**
+ * pvrdma_destroy_cq - destroy completion queue
+ * @cq: the completion queue to destroy.
+ *
+ * @return: 0 for success.
+ */
+int pvrdma_destroy_cq(struct ib_cq *cq)
+{
+	struct pvrdma_cq *vcq = to_vcq(cq);
+	union pvrdma_cmd_req req;
+	struct pvrdma_cmd_destroy_cq *cmd = &req.destroy_cq;
+	struct pvrdma_dev *dev = to_vdev(cq->device);
+	unsigned long flags;
+	int ret;
+
+	memset(cmd, 0, sizeof(*cmd));
+	cmd->hdr.cmd = PVRDMA_CMD_DESTROY_CQ;
+	cmd->cq_handle = vcq->cq_handle;
+
+	ret = pvrdma_cmd_post(dev, &req, NULL, 0);
+	if (ret < 0)
+		dev_warn(&dev->pdev->dev,
+			 "could not destroy completion queue, error: %d\n",
+			 ret);
+
+	/* free cq's resources */
+	spin_lock_irqsave(&dev->cq_tbl_lock, flags);
+	dev->cq_tbl[vcq->cq_handle] = NULL;
+	spin_unlock_irqrestore(&dev->cq_tbl_lock, flags);
+
+	pvrdma_free_cq(dev, vcq);
+	atomic_dec(&dev->num_cqs);
+
+	return ret;
+}
+
+/**
+ * pvrdma_modify_cq - modify the CQ moderation parameters
+ * @ibcq: the CQ to modify
+ * @cq_count: number of CQEs that will trigger an event
+ * @cq_period: max period of time in usec before triggering an event
+ *
+ * @return: -EOPNOTSUPP as CQ resize is not supported.
+ */
+int pvrdma_modify_cq(struct ib_cq *cq, u16 cq_count, u16 cq_period)
+{
+	return -EOPNOTSUPP;
+}
+
+static inline struct pvrdma_cqe *get_cqe(struct pvrdma_cq *cq, int i)
+{
+	return (struct pvrdma_cqe *)pvrdma_page_dir_get_ptr(
+					&cq->pdir,
+					cq->offset +
+					sizeof(struct pvrdma_cqe) * i);
+}
+
+void _pvrdma_flush_cqe(struct pvrdma_qp *qp, struct pvrdma_cq *cq)
+{
+	int head;
+	int has_data;
+
+	if (!cq->is_kernel)
+		return;
+
+	/* Lock held */
+	has_data = pvrdma_idx_ring_has_data(&cq->ring_state->rx,
+					    cq->ibcq.cqe, &head);
+	if (unlikely(has_data > 0)) {
+		int items;
+		int curr;
+		int tail = pvrdma_idx(&cq->ring_state->rx.prod_tail,
+				      cq->ibcq.cqe);
+		struct pvrdma_cqe *cqe;
+		struct pvrdma_cqe *curr_cqe;
+
+		items = (tail > head) ? (tail - head) :
+			(cq->ibcq.cqe - head + tail);
+		curr = --tail;
+		while (items-- > 0) {
+			if (curr < 0)
+				curr = cq->ibcq.cqe - 1;
+			if (tail < 0)
+				tail = cq->ibcq.cqe - 1;
+			curr_cqe = get_cqe(cq, curr);
+			if ((curr_cqe->qp & 0xFFFF) != qp->qp_handle) {
+				if (curr != tail) {
+					cqe = get_cqe(cq, tail);
+					*cqe = *curr_cqe;
+				}
+				tail--;
+			} else {
+				pvrdma_idx_ring_inc(
+					&cq->ring_state->rx.cons_head,
+					cq->ibcq.cqe);
+			}
+			curr--;
+		}
+	}
+}
+
+static int pvrdma_poll_one(struct pvrdma_cq *cq, struct pvrdma_qp **cur_qp,
+			   struct ib_wc *wc)
+{
+	struct pvrdma_dev *dev = to_vdev(cq->ibcq.device);
+	int has_data;
+	unsigned int head;
+	bool tried = false;
+	struct pvrdma_cqe *cqe;
+
+retry:
+	has_data = pvrdma_idx_ring_has_data(&cq->ring_state->rx,
+					    cq->ibcq.cqe, &head);
+	if (has_data == 0) {
+		if (tried)
+			return -EAGAIN;
+
+		pvrdma_write_uar_cq(dev, cq->cq_handle | PVRDMA_UAR_CQ_POLL);
+
+		tried = true;
+		goto retry;
+	} else if (has_data == PVRDMA_INVALID_IDX) {
+		dev_err(&dev->pdev->dev, "CQ ring state invalid\n");
+		return -EAGAIN;
+	}
+
+	cqe = get_cqe(cq, head);
+
+	/* Ensure cqe is valid. */
+	rmb();
+	if (dev->qp_tbl[cqe->qp & 0xffff])
+		*cur_qp = (struct pvrdma_qp *)dev->qp_tbl[cqe->qp & 0xffff];
+	else
+		return -EAGAIN;
+
+	wc->opcode = pvrdma_wc_opcode_to_ib(cqe->opcode);
+	wc->status = pvrdma_wc_status_to_ib(cqe->status);
+	wc->wr_id = cqe->wr_id;
+	wc->qp = &(*cur_qp)->ibqp;
+	wc->byte_len = cqe->byte_len;
+	wc->ex.imm_data = cqe->imm_data;
+	wc->src_qp = cqe->src_qp;
+	wc->wc_flags = pvrdma_wc_flags_to_ib(cqe->wc_flags);
+	wc->pkey_index = cqe->pkey_index;
+	wc->slid = cqe->slid;
+	wc->sl = cqe->sl;
+	wc->dlid_path_bits = cqe->dlid_path_bits;
+	wc->port_num = cqe->port_num;
+	wc->vendor_err = 0;
+
+	/* Update shared ring state */
+	pvrdma_idx_ring_inc(&cq->ring_state->rx.cons_head, cq->ibcq.cqe);
+
+	return 0;
+}
+
+/**
+ * pvrdma_poll_cq - poll for work completion queue entries
+ * @ibcq: completion queue
+ * @num_entries: the maximum number of entries
+ * @entry: pointer to work completion array
+ *
+ * @return: number of polled completion entries
+ */
+int pvrdma_poll_cq(struct ib_cq *ibcq, int num_entries, struct ib_wc *wc)
+{
+	struct pvrdma_cq *cq = to_vcq(ibcq);
+	struct pvrdma_qp *cur_qp = NULL;
+	unsigned long flags;
+	int npolled;
+
+	if (num_entries < 1 || wc == NULL)
+		return 0;
+
+	spin_lock_irqsave(&cq->cq_lock, flags);
+	for (npolled = 0; npolled < num_entries; ++npolled) {
+		if (pvrdma_poll_one(cq, &cur_qp, wc + npolled))
+			break;
+	}
+
+	spin_unlock_irqrestore(&cq->cq_lock, flags);
+
+	/* Ensure we do not return errors from poll_cq */
+	return npolled;
+}
+
+/**
+ * pvrdma_resize_cq - resize CQ
+ * @ibcq: the completion queue
+ * @entries: CQ entries
+ * @udata: user data
+ *
+ * @return: -EOPNOTSUPP as CQ resize is not supported.
+ */
+int pvrdma_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata)
+{
+	return -EOPNOTSUPP;
+}
diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_dev_api.h b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_dev_api.h
new file mode 100644
index 000000000000..c06768635d65
--- /dev/null
+++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_dev_api.h
@@ -0,0 +1,586 @@
+/*
+ * Copyright (c) 2012-2016 VMware, Inc.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of EITHER the GNU General Public License
+ * version 2 as published by the Free Software Foundation or the BSD
+ * 2-Clause License. This program is distributed in the hope that it
+ * will be useful, but WITHOUT ANY WARRANTY; WITHOUT EVEN THE IMPLIED
+ * WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
+ * See the GNU General Public License version 2 for more details at
+ * http://www.gnu.org/licenses/old-licenses/gpl-2.0.en.html.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program available in the file COPYING in the main
+ * directory of this source tree.
+ *
+ * The BSD 2-Clause License
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+ * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
+ * COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT,
+ * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#ifndef __PVRDMA_DEV_API_H__
+#define __PVRDMA_DEV_API_H__
+
+#include <linux/types.h>
+
+#include "pvrdma_verbs.h"
+
+#define PVRDMA_VERSION			17
+#define PVRDMA_BOARD_ID			1
+#define PVRDMA_REV_ID			1
+
+/*
+ * Masks and accessors for page directory, which is a two-level lookup:
+ * page directory -> page table -> page. Only one directory for now, but we
+ * could expand that easily. 9 bits for tables, 9 bits for pages, gives one
+ * gigabyte for memory regions and so forth.
+ */
+
+#define PVRDMA_PDIR_SHIFT		18
+#define PVRDMA_PTABLE_SHIFT		9
+#define PVRDMA_PAGE_DIR_DIR(x)		(((x) >> PVRDMA_PDIR_SHIFT) & 0x1)
+#define PVRDMA_PAGE_DIR_TABLE(x)	(((x) >> PVRDMA_PTABLE_SHIFT) & 0x1ff)
+#define PVRDMA_PAGE_DIR_PAGE(x)		((x) & 0x1ff)
+#define PVRDMA_PAGE_DIR_MAX_PAGES	(1 * 512 * 512)
+#define PVRDMA_MAX_FAST_REG_PAGES	128
+
+/*
+ * Max MSI-X vectors.
+ */
+
+#define PVRDMA_MAX_INTERRUPTS	3
+
+/* Register offsets within PCI resource on BAR1. */
+#define PVRDMA_REG_VERSION	0x00	/* R: Version of device. */
+#define PVRDMA_REG_DSRLOW	0x04	/* W: Device shared region low PA. */
+#define PVRDMA_REG_DSRHIGH	0x08	/* W: Device shared region high PA. */
+#define PVRDMA_REG_CTL		0x0c	/* W: PVRDMA_DEVICE_CTL */
+#define PVRDMA_REG_REQUEST	0x10	/* W: Indicate device request. */
+#define PVRDMA_REG_ERR		0x14	/* R: Device error. */
+#define PVRDMA_REG_ICR		0x18	/* R: Interrupt cause. */
+#define PVRDMA_REG_IMR		0x1c	/* R/W: Interrupt mask. */
+#define PVRDMA_REG_MACL		0x20	/* R/W: MAC address low. */
+#define PVRDMA_REG_MACH		0x24	/* R/W: MAC address high. */
+
+/* Object flags. */
+#define PVRDMA_CQ_FLAG_ARMED_SOL	BIT(0)	/* Armed for solicited-only. */
+#define PVRDMA_CQ_FLAG_ARMED		BIT(1)	/* Armed. */
+#define PVRDMA_MR_FLAG_DMA		BIT(0)	/* DMA region. */
+#define PVRDMA_MR_FLAG_FRMR		BIT(1)	/* Fast reg memory region. */
+
+/*
+ * Atomic operation capability (masked versions are extended atomic
+ * operations.
+ */
+
+#define PVRDMA_ATOMIC_OP_COMP_SWAP	BIT(0)	/* Compare and swap. */
+#define PVRDMA_ATOMIC_OP_FETCH_ADD	BIT(1)	/* Fetch and add. */
+#define PVRDMA_ATOMIC_OP_MASK_COMP_SWAP	BIT(2)	/* Masked compare and swap. */
+#define PVRDMA_ATOMIC_OP_MASK_FETCH_ADD	BIT(3)	/* Masked fetch and add. */
+
+/*
+ * Base Memory Management Extension flags to support Fast Reg Memory Regions
+ * and Fast Reg Work Requests. Each flag represents a verb operation and we
+ * must support all of them to qualify for the BMME device cap.
+ */
+
+#define PVRDMA_BMME_FLAG_LOCAL_INV	BIT(0)	/* Local Invalidate. */
+#define PVRDMA_BMME_FLAG_REMOTE_INV	BIT(1)	/* Remote Invalidate. */
+#define PVRDMA_BMME_FLAG_FAST_REG_WR	BIT(2)	/* Fast Reg Work Request. */
+
+/*
+ * GID types. The interpretation of the gid_types bit field in the device
+ * capabilities will depend on the device mode. For now, the device only
+ * supports RoCE as mode, so only the different GID types for RoCE are
+ * defined.
+ */
+
+#define PVRDMA_GID_TYPE_FLAG_ROCE_V1	BIT(0)
+#define PVRDMA_GID_TYPE_FLAG_ROCE_V2	BIT(1)
+
+enum pvrdma_pci_resource {
+	PVRDMA_PCI_RESOURCE_MSIX,	/* BAR0: MSI-X, MMIO. */
+	PVRDMA_PCI_RESOURCE_REG,	/* BAR1: Registers, MMIO. */
+	PVRDMA_PCI_RESOURCE_UAR,	/* BAR2: UAR pages, MMIO, 64-bit. */
+	PVRDMA_PCI_RESOURCE_LAST,	/* Last. */
+};
+
+enum pvrdma_device_ctl {
+	PVRDMA_DEVICE_CTL_ACTIVATE,	/* Activate device. */
+	PVRDMA_DEVICE_CTL_QUIESCE,	/* Quiesce device. */
+	PVRDMA_DEVICE_CTL_RESET,	/* Reset device. */
+};
+
+enum pvrdma_intr_vector {
+	PVRDMA_INTR_VECTOR_RESPONSE,	/* Command response. */
+	PVRDMA_INTR_VECTOR_ASYNC,	/* Async events. */
+	PVRDMA_INTR_VECTOR_CQ,		/* CQ notification. */
+	/* Additional CQ notification vectors. */
+};
+
+enum pvrdma_intr_cause {
+	PVRDMA_INTR_CAUSE_RESPONSE	= (1 << PVRDMA_INTR_VECTOR_RESPONSE),
+	PVRDMA_INTR_CAUSE_ASYNC		= (1 << PVRDMA_INTR_VECTOR_ASYNC),
+	PVRDMA_INTR_CAUSE_CQ		= (1 << PVRDMA_INTR_VECTOR_CQ),
+};
+
+enum pvrdma_intr_type {
+	PVRDMA_INTR_TYPE_INTX,		/* Legacy. */
+	PVRDMA_INTR_TYPE_MSI,		/* MSI. */
+	PVRDMA_INTR_TYPE_MSIX,		/* MSI-X. */
+};
+
+enum pvrdma_gos_bits {
+	PVRDMA_GOS_BITS_UNK,		/* Unknown. */
+	PVRDMA_GOS_BITS_32,		/* 32-bit. */
+	PVRDMA_GOS_BITS_64,		/* 64-bit. */
+};
+
+enum pvrdma_gos_type {
+	PVRDMA_GOS_TYPE_UNK,		/* Unknown. */
+	PVRDMA_GOS_TYPE_LINUX,		/* Linux. */
+};
+
+enum pvrdma_device_mode {
+	PVRDMA_DEVICE_MODE_ROCE,	/* RoCE. */
+	PVRDMA_DEVICE_MODE_IWARP,	/* iWarp. */
+	PVRDMA_DEVICE_MODE_IB,		/* InfiniBand. */
+};
+
+struct pvrdma_gos_info {
+	u32 gos_bits:2;			/* W: PVRDMA_GOS_BITS_ */
+	u32 gos_type:4;			/* W: PVRDMA_GOS_TYPE_ */
+	u32 gos_ver:16;			/* W: Guest OS version. */
+	u32 gos_misc:10;		/* W: Other. */
+	u32 pad;			/* Pad to 8-byte alignment. */
+};
+
+struct pvrdma_device_caps {
+	u64 fw_ver;				/* R: Query device. */
+	__be64 node_guid;
+	__be64 sys_image_guid;
+	u64 max_mr_size;
+	u64 page_size_cap;
+	u64 atomic_arg_sizes;			/* EX verbs. */
+	u32 ex_comp_mask;			/* EX verbs. */
+	u32 device_cap_flags2;			/* EX verbs. */
+	u32 max_fa_bit_boundary;		/* EX verbs. */
+	u32 log_max_atomic_inline_arg;		/* EX verbs. */
+	u32 vendor_id;
+	u32 vendor_part_id;
+	u32 hw_ver;
+	u32 max_qp;
+	u32 max_qp_wr;
+	u32 device_cap_flags;
+	u32 max_sge;
+	u32 max_sge_rd;
+	u32 max_cq;
+	u32 max_cqe;
+	u32 max_mr;
+	u32 max_pd;
+	u32 max_qp_rd_atom;
+	u32 max_ee_rd_atom;
+	u32 max_res_rd_atom;
+	u32 max_qp_init_rd_atom;
+	u32 max_ee_init_rd_atom;
+	u32 max_ee;
+	u32 max_rdd;
+	u32 max_mw;
+	u32 max_raw_ipv6_qp;
+	u32 max_raw_ethy_qp;
+	u32 max_mcast_grp;
+	u32 max_mcast_qp_attach;
+	u32 max_total_mcast_qp_attach;
+	u32 max_ah;
+	u32 max_fmr;
+	u32 max_map_per_fmr;
+	u32 max_srq;
+	u32 max_srq_wr;
+	u32 max_srq_sge;
+	u32 max_uar;
+	u32 gid_tbl_len;
+	u16 max_pkeys;
+	u8  local_ca_ack_delay;
+	u8  phys_port_cnt;
+	u8  mode;				/* PVRDMA_DEVICE_MODE_ */
+	u8  atomic_ops;				/* PVRDMA_ATOMIC_OP_* bits */
+	u8  bmme_flags;				/* FRWR Mem Mgmt Extensions */
+	u8  gid_types;				/* PVRDMA_GID_TYPE_FLAG_ */
+	u8  reserved[4];
+};
+
+struct pvrdma_ring_page_info {
+	u32 num_pages;				/* Num pages incl. header. */
+	u32 reserved;				/* Reserved. */
+	u64 pdir_dma;				/* Page directory PA. */
+};
+
+#pragma pack(push, 1)
+
+struct pvrdma_device_shared_region {
+	u32 driver_version;			/* W: Driver version. */
+	u32 pad;				/* Pad to 8-byte align. */
+	struct pvrdma_gos_info gos_info;	/* W: Guest OS information. */
+	u64 cmd_slot_dma;			/* W: Command slot address. */
+	u64 resp_slot_dma;			/* W: Response slot address. */
+	struct pvrdma_ring_page_info async_ring_pages;
+						/* W: Async ring page info. */
+	struct pvrdma_ring_page_info cq_ring_pages;
+						/* W: CQ ring page info. */
+	u32 uar_pfn;				/* W: UAR pageframe. */
+	u32 pad2;				/* Pad to 8-byte align. */
+	struct pvrdma_device_caps caps;		/* R: Device capabilities. */
+};
+
+#pragma pack(pop)
+
+/* Event types. Currently a 1:1 mapping with enum ib_event. */
+enum pvrdma_eqe_type {
+	PVRDMA_EVENT_CQ_ERR,
+	PVRDMA_EVENT_QP_FATAL,
+	PVRDMA_EVENT_QP_REQ_ERR,
+	PVRDMA_EVENT_QP_ACCESS_ERR,
+	PVRDMA_EVENT_COMM_EST,
+	PVRDMA_EVENT_SQ_DRAINED,
+	PVRDMA_EVENT_PATH_MIG,
+	PVRDMA_EVENT_PATH_MIG_ERR,
+	PVRDMA_EVENT_DEVICE_FATAL,
+	PVRDMA_EVENT_PORT_ACTIVE,
+	PVRDMA_EVENT_PORT_ERR,
+	PVRDMA_EVENT_LID_CHANGE,
+	PVRDMA_EVENT_PKEY_CHANGE,
+	PVRDMA_EVENT_SM_CHANGE,
+	PVRDMA_EVENT_SRQ_ERR,
+	PVRDMA_EVENT_SRQ_LIMIT_REACHED,
+	PVRDMA_EVENT_QP_LAST_WQE_REACHED,
+	PVRDMA_EVENT_CLIENT_REREGISTER,
+	PVRDMA_EVENT_GID_CHANGE,
+};
+
+/* Event queue element. */
+struct pvrdma_eqe {
+	u32 type;	/* Event type. */
+	u32 info;	/* Handle, other. */
+};
+
+/* CQ notification queue element. */
+struct pvrdma_cqne {
+	u32 info;	/* Handle */
+};
+
+enum {
+	PVRDMA_CMD_FIRST,
+	PVRDMA_CMD_QUERY_PORT = PVRDMA_CMD_FIRST,
+	PVRDMA_CMD_QUERY_PKEY,
+	PVRDMA_CMD_CREATE_PD,
+	PVRDMA_CMD_DESTROY_PD,
+	PVRDMA_CMD_CREATE_MR,
+	PVRDMA_CMD_DESTROY_MR,
+	PVRDMA_CMD_CREATE_CQ,
+	PVRDMA_CMD_RESIZE_CQ,
+	PVRDMA_CMD_DESTROY_CQ,
+	PVRDMA_CMD_CREATE_QP,
+	PVRDMA_CMD_MODIFY_QP,
+	PVRDMA_CMD_QUERY_QP,
+	PVRDMA_CMD_DESTROY_QP,
+	PVRDMA_CMD_CREATE_UC,
+	PVRDMA_CMD_DESTROY_UC,
+	PVRDMA_CMD_CREATE_BIND,
+	PVRDMA_CMD_DESTROY_BIND,
+	PVRDMA_CMD_MAX,
+};
+
+enum {
+	PVRDMA_CMD_FIRST_RESP = (1 << 31),
+	PVRDMA_CMD_QUERY_PORT_RESP = PVRDMA_CMD_FIRST_RESP,
+	PVRDMA_CMD_QUERY_PKEY_RESP,
+	PVRDMA_CMD_CREATE_PD_RESP,
+	PVRDMA_CMD_DESTROY_PD_RESP_NOOP,
+	PVRDMA_CMD_CREATE_MR_RESP,
+	PVRDMA_CMD_DESTROY_MR_RESP_NOOP,
+	PVRDMA_CMD_CREATE_CQ_RESP,
+	PVRDMA_CMD_RESIZE_CQ_RESP,
+	PVRDMA_CMD_DESTROY_CQ_RESP_NOOP,
+	PVRDMA_CMD_CREATE_QP_RESP,
+	PVRDMA_CMD_MODIFY_QP_RESP,
+	PVRDMA_CMD_QUERY_QP_RESP,
+	PVRDMA_CMD_DESTROY_QP_RESP,
+	PVRDMA_CMD_CREATE_UC_RESP,
+	PVRDMA_CMD_DESTROY_UC_RESP_NOOP,
+	PVRDMA_CMD_CREATE_BIND_RESP_NOOP,
+	PVRDMA_CMD_DESTROY_BIND_RESP_NOOP,
+	PVRDMA_CMD_MAX_RESP,
+};
+
+struct pvrdma_cmd_hdr {
+	u64 response;		/* Key for response lookup. */
+	u32 cmd;		/* PVRDMA_CMD_ */
+	u32 reserved;		/* Reserved. */
+};
+
+struct pvrdma_cmd_resp_hdr {
+	u64 response;		/* From cmd hdr. */
+	u32 ack;		/* PVRDMA_CMD_XXX_RESP */
+	u8 err;			/* Error. */
+	u8 reserved[3];		/* Reserved. */
+};
+
+struct pvrdma_cmd_query_port {
+	struct pvrdma_cmd_hdr hdr;
+	u8 port_num;
+	u8 reserved[7];
+};
+
+struct pvrdma_cmd_query_port_resp {
+	struct pvrdma_cmd_resp_hdr hdr;
+	struct pvrdma_port_attr attrs;
+};
+
+struct pvrdma_cmd_query_pkey {
+	struct pvrdma_cmd_hdr hdr;
+	u8 port_num;
+	u8 index;
+	u8 reserved[6];
+};
+
+struct pvrdma_cmd_query_pkey_resp {
+	struct pvrdma_cmd_resp_hdr hdr;
+	u16 pkey;
+	u8 reserved[6];
+};
+
+struct pvrdma_cmd_create_uc {
+	struct pvrdma_cmd_hdr hdr;
+	u32 pfn; /* UAR page frame number */
+	u8 reserved[4];
+};
+
+struct pvrdma_cmd_create_uc_resp {
+	struct pvrdma_cmd_resp_hdr hdr;
+	u32 ctx_handle;
+	u8 reserved[4];
+};
+
+struct pvrdma_cmd_destroy_uc {
+	struct pvrdma_cmd_hdr hdr;
+	u32 ctx_handle;
+	u8 reserved[4];
+};
+
+struct pvrdma_cmd_create_pd {
+	struct pvrdma_cmd_hdr hdr;
+	u32 ctx_handle;
+	u8 reserved[4];
+};
+
+struct pvrdma_cmd_create_pd_resp {
+	struct pvrdma_cmd_resp_hdr hdr;
+	u32 pd_handle;
+	u8 reserved[4];
+};
+
+struct pvrdma_cmd_destroy_pd {
+	struct pvrdma_cmd_hdr hdr;
+	u32 pd_handle;
+	u8 reserved[4];
+};
+
+struct pvrdma_cmd_create_mr {
+	struct pvrdma_cmd_hdr hdr;
+	u64 start;
+	u64 length;
+	u64 pdir_dma;
+	u32 pd_handle;
+	u32 access_flags;
+	u32 flags;
+	u32 nchunks;
+};
+
+struct pvrdma_cmd_create_mr_resp {
+	struct pvrdma_cmd_resp_hdr hdr;
+	u32 mr_handle;
+	u32 lkey;
+	u32 rkey;
+	u8 reserved[4];
+};
+
+struct pvrdma_cmd_destroy_mr {
+	struct pvrdma_cmd_hdr hdr;
+	u32 mr_handle;
+	u8 reserved[4];
+};
+
+struct pvrdma_cmd_create_cq {
+	struct pvrdma_cmd_hdr hdr;
+	u64 pdir_dma;
+	u32 ctx_handle;
+	u32 cqe;
+	u32 nchunks;
+	u8 reserved[4];
+};
+
+struct pvrdma_cmd_create_cq_resp {
+	struct pvrdma_cmd_resp_hdr hdr;
+	u32 cq_handle;
+	u32 cqe;
+};
+
+struct pvrdma_cmd_resize_cq {
+	struct pvrdma_cmd_hdr hdr;
+	u32 cq_handle;
+	u32 cqe;
+};
+
+struct pvrdma_cmd_resize_cq_resp {
+	struct pvrdma_cmd_resp_hdr hdr;
+	u32 cqe;
+	u8 reserved[4];
+};
+
+struct pvrdma_cmd_destroy_cq {
+	struct pvrdma_cmd_hdr hdr;
+	u32 cq_handle;
+	u8 reserved[4];
+};
+
+struct pvrdma_cmd_create_qp {
+	struct pvrdma_cmd_hdr hdr;
+	u64 pdir_dma;
+	u32 pd_handle;
+	u32 send_cq_handle;
+	u32 recv_cq_handle;
+	u32 srq_handle;
+	u32 max_send_wr;
+	u32 max_recv_wr;
+	u32 max_send_sge;
+	u32 max_recv_sge;
+	u32 max_inline_data;
+	u32 lkey;
+	u32 access_flags;
+	u16 total_chunks;
+	u16 send_chunks;
+	u16 max_atomic_arg;
+	u8 sq_sig_all;
+	u8 qp_type;
+	u8 is_srq;
+	u8 reserved[3];
+};
+
+struct pvrdma_cmd_create_qp_resp {
+	struct pvrdma_cmd_resp_hdr hdr;
+	u32 qpn;
+	u32 max_send_wr;
+	u32 max_recv_wr;
+	u32 max_send_sge;
+	u32 max_recv_sge;
+	u32 max_inline_data;
+};
+
+struct pvrdma_cmd_modify_qp {
+	struct pvrdma_cmd_hdr hdr;
+	u32 qp_handle;
+	u32 attr_mask;
+	struct pvrdma_qp_attr attrs;
+};
+
+struct pvrdma_cmd_query_qp {
+	struct pvrdma_cmd_hdr hdr;
+	u32 qp_handle;
+	u32 attr_mask;
+};
+
+struct pvrdma_cmd_query_qp_resp {
+	struct pvrdma_cmd_resp_hdr hdr;
+	struct pvrdma_qp_attr attrs;
+};
+
+struct pvrdma_cmd_destroy_qp {
+	struct pvrdma_cmd_hdr hdr;
+	u32 qp_handle;
+	u8 reserved[4];
+};
+
+struct pvrdma_cmd_destroy_qp_resp {
+	struct pvrdma_cmd_resp_hdr hdr;
+	u32 events_reported;
+	u8 reserved[4];
+};
+
+struct pvrdma_cmd_create_bind {
+	struct pvrdma_cmd_hdr hdr;
+	u32 mtu;
+	u32 vlan;
+	u32 index;
+	u8 new_gid[16];
+	u8 gid_type;
+	u8 reserved[3];
+};
+
+struct pvrdma_cmd_destroy_bind {
+	struct pvrdma_cmd_hdr hdr;
+	u32 index;
+	u8 dest_gid[16];
+	u8 reserved[4];
+};
+
+union pvrdma_cmd_req {
+	struct pvrdma_cmd_hdr hdr;
+	struct pvrdma_cmd_query_port query_port;
+	struct pvrdma_cmd_query_pkey query_pkey;
+	struct pvrdma_cmd_create_uc create_uc;
+	struct pvrdma_cmd_destroy_uc destroy_uc;
+	struct pvrdma_cmd_create_pd create_pd;
+	struct pvrdma_cmd_destroy_pd destroy_pd;
+	struct pvrdma_cmd_create_mr create_mr;
+	struct pvrdma_cmd_destroy_mr destroy_mr;
+	struct pvrdma_cmd_create_cq create_cq;
+	struct pvrdma_cmd_resize_cq resize_cq;
+	struct pvrdma_cmd_destroy_cq destroy_cq;
+	struct pvrdma_cmd_create_qp create_qp;
+	struct pvrdma_cmd_modify_qp modify_qp;
+	struct pvrdma_cmd_query_qp query_qp;
+	struct pvrdma_cmd_destroy_qp destroy_qp;
+	struct pvrdma_cmd_create_bind create_bind;
+	struct pvrdma_cmd_destroy_bind destroy_bind;
+};
+
+union pvrdma_cmd_resp {
+	struct pvrdma_cmd_resp_hdr hdr;
+	struct pvrdma_cmd_query_port_resp query_port_resp;
+	struct pvrdma_cmd_query_pkey_resp query_pkey_resp;
+	struct pvrdma_cmd_create_uc_resp create_uc_resp;
+	struct pvrdma_cmd_create_pd_resp create_pd_resp;
+	struct pvrdma_cmd_create_mr_resp create_mr_resp;
+	struct pvrdma_cmd_create_cq_resp create_cq_resp;
+	struct pvrdma_cmd_resize_cq_resp resize_cq_resp;
+	struct pvrdma_cmd_create_qp_resp create_qp_resp;
+	struct pvrdma_cmd_query_qp_resp query_qp_resp;
+	struct pvrdma_cmd_destroy_qp_resp destroy_qp_resp;
+};
+
+#endif /* __PVRDMA_DEV_API_H__ */
diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_doorbell.c b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_doorbell.c
new file mode 100644
index 000000000000..bf51357ea3aa
--- /dev/null
+++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_doorbell.c
@@ -0,0 +1,127 @@
+/*
+ * Copyright (c) 2012-2016 VMware, Inc.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of EITHER the GNU General Public License
+ * version 2 as published by the Free Software Foundation or the BSD
+ * 2-Clause License. This program is distributed in the hope that it
+ * will be useful, but WITHOUT ANY WARRANTY; WITHOUT EVEN THE IMPLIED
+ * WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
+ * See the GNU General Public License version 2 for more details at
+ * http://www.gnu.org/licenses/old-licenses/gpl-2.0.en.html.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program available in the file COPYING in the main
+ * directory of this source tree.
+ *
+ * The BSD 2-Clause License
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+ * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
+ * COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT,
+ * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <linux/bitmap.h>
+#include <linux/errno.h>
+#include <linux/slab.h>
+
+#include "pvrdma.h"
+
+int pvrdma_uar_table_init(struct pvrdma_dev *dev)
+{
+	u32 num = dev->dsr->caps.max_uar;
+	u32 mask = num - 1;
+	struct pvrdma_id_table *tbl = &dev->uar_table.tbl;
+
+	if (!is_power_of_2(num))
+		return -EINVAL;
+
+	tbl->last = 0;
+	tbl->top = 0;
+	tbl->max = num;
+	tbl->mask = mask;
+	spin_lock_init(&tbl->lock);
+	tbl->table = kcalloc(BITS_TO_LONGS(num), sizeof(long), GFP_KERNEL);
+	if (!tbl->table)
+		return -ENOMEM;
+
+	/* 0th UAR is taken by the device. */
+	set_bit(0, tbl->table);
+
+	return 0;
+}
+
+void pvrdma_uar_table_cleanup(struct pvrdma_dev *dev)
+{
+	struct pvrdma_id_table *tbl = &dev->uar_table.tbl;
+
+	kfree(tbl->table);
+}
+
+int pvrdma_uar_alloc(struct pvrdma_dev *dev, struct pvrdma_uar_map *uar)
+{
+	struct pvrdma_id_table *tbl;
+	unsigned long flags;
+	u32 obj;
+
+	tbl = &dev->uar_table.tbl;
+
+	spin_lock_irqsave(&tbl->lock, flags);
+	obj = find_next_zero_bit(tbl->table, tbl->max, tbl->last);
+	if (obj >= tbl->max) {
+		tbl->top = (tbl->top + tbl->max) & tbl->mask;
+		obj = find_first_zero_bit(tbl->table, tbl->max);
+	}
+
+	if (obj >= tbl->max) {
+		spin_unlock_irqrestore(&tbl->lock, flags);
+		return -ENOMEM;
+	}
+
+	set_bit(obj, tbl->table);
+	obj |= tbl->top;
+
+	spin_unlock_irqrestore(&tbl->lock, flags);
+
+	uar->index = obj;
+	uar->pfn = (pci_resource_start(dev->pdev, PVRDMA_PCI_RESOURCE_UAR) >>
+		    PAGE_SHIFT) + uar->index;
+
+	return 0;
+}
+
+void pvrdma_uar_free(struct pvrdma_dev *dev, struct pvrdma_uar_map *uar)
+{
+	struct pvrdma_id_table *tbl = &dev->uar_table.tbl;
+	unsigned long flags;
+	u32 obj;
+
+	obj = uar->index & (tbl->max - 1);
+	spin_lock_irqsave(&tbl->lock, flags);
+	clear_bit(obj, tbl->table);
+	tbl->last = min(tbl->last, obj);
+	tbl->top = (tbl->top + tbl->max) & tbl->mask;
+	spin_unlock_irqrestore(&tbl->lock, flags);
+}
diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_main.c b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_main.c
new file mode 100644
index 000000000000..231a1ce1f4be
--- /dev/null
+++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_main.c
@@ -0,0 +1,1211 @@
+/*
+ * Copyright (c) 2012-2016 VMware, Inc.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of EITHER the GNU General Public License
+ * version 2 as published by the Free Software Foundation or the BSD
+ * 2-Clause License. This program is distributed in the hope that it
+ * will be useful, but WITHOUT ANY WARRANTY; WITHOUT EVEN THE IMPLIED
+ * WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
+ * See the GNU General Public License version 2 for more details at
+ * http://www.gnu.org/licenses/old-licenses/gpl-2.0.en.html.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program available in the file COPYING in the main
+ * directory of this source tree.
+ *
+ * The BSD 2-Clause License
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+ * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
+ * COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT,
+ * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <linux/errno.h>
+#include <linux/inetdevice.h>
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <rdma/ib_addr.h>
+#include <rdma/ib_smi.h>
+#include <rdma/ib_user_verbs.h>
+#include <net/addrconf.h>
+
+#include "pvrdma.h"
+
+#define DRV_NAME	"vmw_pvrdma"
+#define DRV_VERSION	"1.0.0.0-k"
+
+static DEFINE_MUTEX(pvrdma_device_list_lock);
+static LIST_HEAD(pvrdma_device_list);
+static struct workqueue_struct *event_wq;
+
+static int pvrdma_add_gid(struct ib_device *ibdev,
+			  u8 port_num,
+			  unsigned int index,
+			  const union ib_gid *gid,
+			  const struct ib_gid_attr *attr,
+			  void **context);
+static int pvrdma_del_gid(struct ib_device *ibdev,
+			  u8 port_num,
+			  unsigned int index,
+			  void **context);
+
+
+static ssize_t show_hca(struct device *device, struct device_attribute *attr,
+			char *buf)
+{
+	return sprintf(buf, "VMW_PVRDMA-%s\n", DRV_VERSION);
+}
+
+static ssize_t show_rev(struct device *device, struct device_attribute *attr,
+			char *buf)
+{
+	return sprintf(buf, "%d\n", PVRDMA_REV_ID);
+}
+
+static ssize_t show_board(struct device *device, struct device_attribute *attr,
+			  char *buf)
+{
+	return sprintf(buf, "%d\n", PVRDMA_BOARD_ID);
+}
+
+static DEVICE_ATTR(hw_rev,   S_IRUGO, show_rev,	   NULL);
+static DEVICE_ATTR(hca_type, S_IRUGO, show_hca,	   NULL);
+static DEVICE_ATTR(board_id, S_IRUGO, show_board,  NULL);
+
+static struct device_attribute *pvrdma_class_attributes[] = {
+	&dev_attr_hw_rev,
+	&dev_attr_hca_type,
+	&dev_attr_board_id
+};
+
+static void pvrdma_get_fw_ver_str(struct ib_device *device, char *str,
+				  size_t str_len)
+{
+	struct pvrdma_dev *dev =
+		container_of(device, struct pvrdma_dev, ib_dev);
+	snprintf(str, str_len, "%d.%d.%d\n",
+		 (int) (dev->dsr->caps.fw_ver >> 32),
+		 (int) (dev->dsr->caps.fw_ver >> 16) & 0xffff,
+		 (int) dev->dsr->caps.fw_ver & 0xffff);
+}
+
+static int pvrdma_init_device(struct pvrdma_dev *dev)
+{
+	/*  Initialize some device related stuff */
+	spin_lock_init(&dev->cmd_lock);
+	sema_init(&dev->cmd_sema, 1);
+	atomic_set(&dev->num_qps, 0);
+	atomic_set(&dev->num_cqs, 0);
+	atomic_set(&dev->num_pds, 0);
+	atomic_set(&dev->num_ahs, 0);
+
+	return 0;
+}
+
+static int pvrdma_port_immutable(struct ib_device *ibdev, u8 port_num,
+				 struct ib_port_immutable *immutable)
+{
+	struct ib_port_attr attr;
+	int err;
+
+	err = pvrdma_query_port(ibdev, port_num, &attr);
+	if (err)
+		return err;
+
+	immutable->pkey_tbl_len = attr.pkey_tbl_len;
+	immutable->gid_tbl_len = attr.gid_tbl_len;
+	immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE;
+	immutable->max_mad_size = IB_MGMT_MAD_SIZE;
+	return 0;
+}
+
+static struct net_device *pvrdma_get_netdev(struct ib_device *ibdev,
+					    u8 port_num)
+{
+	struct net_device *netdev;
+	struct pvrdma_dev *dev = to_vdev(ibdev);
+
+	if (port_num != 1)
+		return NULL;
+
+	rcu_read_lock();
+	netdev = dev->netdev;
+	if (netdev)
+		dev_hold(netdev);
+	rcu_read_unlock();
+
+	return netdev;
+}
+
+static int pvrdma_register_device(struct pvrdma_dev *dev)
+{
+	int ret = -1;
+	int i = 0;
+
+	strlcpy(dev->ib_dev.name, "vmw_pvrdma%d", IB_DEVICE_NAME_MAX);
+	dev->ib_dev.node_guid = dev->dsr->caps.node_guid;
+	dev->sys_image_guid = dev->dsr->caps.sys_image_guid;
+	dev->flags = 0;
+	dev->ib_dev.owner = THIS_MODULE;
+	dev->ib_dev.num_comp_vectors = 1;
+	dev->ib_dev.dma_device = &dev->pdev->dev;
+	dev->ib_dev.uverbs_abi_ver = PVRDMA_UVERBS_ABI_VERSION;
+	dev->ib_dev.uverbs_cmd_mask =
+		(1ull << IB_USER_VERBS_CMD_GET_CONTEXT)		|
+		(1ull << IB_USER_VERBS_CMD_QUERY_DEVICE)	|
+		(1ull << IB_USER_VERBS_CMD_QUERY_PORT)		|
+		(1ull << IB_USER_VERBS_CMD_ALLOC_PD)		|
+		(1ull << IB_USER_VERBS_CMD_DEALLOC_PD)		|
+		(1ull << IB_USER_VERBS_CMD_REG_MR)		|
+		(1ull << IB_USER_VERBS_CMD_DEREG_MR)		|
+		(1ull << IB_USER_VERBS_CMD_CREATE_COMP_CHANNEL)	|
+		(1ull << IB_USER_VERBS_CMD_CREATE_CQ)		|
+		(1ull << IB_USER_VERBS_CMD_POLL_CQ)		|
+		(1ull << IB_USER_VERBS_CMD_REQ_NOTIFY_CQ)	|
+		(1ull << IB_USER_VERBS_CMD_DESTROY_CQ)		|
+		(1ull << IB_USER_VERBS_CMD_CREATE_QP)		|
+		(1ull << IB_USER_VERBS_CMD_MODIFY_QP)		|
+		(1ull << IB_USER_VERBS_CMD_QUERY_QP)		|
+		(1ull << IB_USER_VERBS_CMD_DESTROY_QP)		|
+		(1ull << IB_USER_VERBS_CMD_POST_SEND)		|
+		(1ull << IB_USER_VERBS_CMD_POST_RECV)		|
+		(1ull << IB_USER_VERBS_CMD_CREATE_AH)		|
+		(1ull << IB_USER_VERBS_CMD_DESTROY_AH);
+
+	dev->ib_dev.node_type = RDMA_NODE_IB_CA;
+	dev->ib_dev.phys_port_cnt = dev->dsr->caps.phys_port_cnt;
+
+	dev->ib_dev.query_device = pvrdma_query_device;
+	dev->ib_dev.query_port = pvrdma_query_port;
+	dev->ib_dev.query_gid = pvrdma_query_gid;
+	dev->ib_dev.query_pkey = pvrdma_query_pkey;
+	dev->ib_dev.modify_port	= pvrdma_modify_port;
+	dev->ib_dev.alloc_ucontext = pvrdma_alloc_ucontext;
+	dev->ib_dev.dealloc_ucontext = pvrdma_dealloc_ucontext;
+	dev->ib_dev.mmap = pvrdma_mmap;
+	dev->ib_dev.alloc_pd = pvrdma_alloc_pd;
+	dev->ib_dev.dealloc_pd = pvrdma_dealloc_pd;
+	dev->ib_dev.create_ah = pvrdma_create_ah;
+	dev->ib_dev.destroy_ah = pvrdma_destroy_ah;
+	dev->ib_dev.create_qp = pvrdma_create_qp;
+	dev->ib_dev.modify_qp = pvrdma_modify_qp;
+	dev->ib_dev.query_qp = pvrdma_query_qp;
+	dev->ib_dev.destroy_qp = pvrdma_destroy_qp;
+	dev->ib_dev.post_send = pvrdma_post_send;
+	dev->ib_dev.post_recv = pvrdma_post_recv;
+	dev->ib_dev.create_cq = pvrdma_create_cq;
+	dev->ib_dev.modify_cq = pvrdma_modify_cq;
+	dev->ib_dev.resize_cq = pvrdma_resize_cq;
+	dev->ib_dev.destroy_cq = pvrdma_destroy_cq;
+	dev->ib_dev.poll_cq = pvrdma_poll_cq;
+	dev->ib_dev.req_notify_cq = pvrdma_req_notify_cq;
+	dev->ib_dev.get_dma_mr = pvrdma_get_dma_mr;
+	dev->ib_dev.reg_user_mr	= pvrdma_reg_user_mr;
+	dev->ib_dev.dereg_mr = pvrdma_dereg_mr;
+	dev->ib_dev.alloc_mr = pvrdma_alloc_mr;
+	dev->ib_dev.map_mr_sg = pvrdma_map_mr_sg;
+	dev->ib_dev.add_gid = pvrdma_add_gid;
+	dev->ib_dev.del_gid = pvrdma_del_gid;
+	dev->ib_dev.get_netdev = pvrdma_get_netdev;
+	dev->ib_dev.get_port_immutable = pvrdma_port_immutable;
+	dev->ib_dev.get_link_layer = pvrdma_port_link_layer;
+	dev->ib_dev.get_dev_fw_str = pvrdma_get_fw_ver_str;
+
+	mutex_init(&dev->port_mutex);
+	spin_lock_init(&dev->desc_lock);
+
+	dev->cq_tbl = kcalloc(dev->dsr->caps.max_cq, sizeof(void *),
+			      GFP_KERNEL);
+	if (!dev->cq_tbl)
+		return ret;
+	spin_lock_init(&dev->cq_tbl_lock);
+
+	dev->qp_tbl = kcalloc(dev->dsr->caps.max_qp, sizeof(void *),
+			      GFP_KERNEL);
+	if (!dev->qp_tbl)
+		goto err_cq_free;
+	spin_lock_init(&dev->qp_tbl_lock);
+
+	ret = ib_register_device(&dev->ib_dev, NULL);
+	if (ret)
+		goto err_qp_free;
+
+	for (i = 0; i < ARRAY_SIZE(pvrdma_class_attributes); ++i) {
+		ret = device_create_file(&dev->ib_dev.dev,
+					 pvrdma_class_attributes[i]);
+		if (ret)
+			goto err_class;
+	}
+
+	dev->ib_active = true;
+
+	return 0;
+
+err_class:
+	ib_unregister_device(&dev->ib_dev);
+err_qp_free:
+	kfree(dev->qp_tbl);
+err_cq_free:
+	kfree(dev->cq_tbl);
+
+	return ret;
+}
+
+static irqreturn_t pvrdma_intr0_handler(int irq, void *dev_id)
+{
+	u32 icr = PVRDMA_INTR_CAUSE_RESPONSE;
+	struct pvrdma_dev *dev = dev_id;
+
+	dev_dbg(&dev->pdev->dev, "interrupt 0 (response) handler\n");
+
+	if (dev->intr.type != PVRDMA_INTR_TYPE_MSIX) {
+		/* Legacy intr */
+		icr = pvrdma_read_reg(dev, PVRDMA_REG_ICR);
+		if (icr == 0)
+			return IRQ_NONE;
+	}
+
+	if (icr == PVRDMA_INTR_CAUSE_RESPONSE)
+		complete(&dev->cmd_done);
+
+	return IRQ_HANDLED;
+}
+
+static void pvrdma_qp_event(struct pvrdma_dev *dev, u32 qpn, int type)
+{
+	struct pvrdma_qp *qp;
+	unsigned long flags;
+
+	spin_lock_irqsave(&dev->qp_tbl_lock, flags);
+	qp = dev->qp_tbl[qpn % dev->dsr->caps.max_qp];
+	if (qp)
+		atomic_inc(&qp->refcnt);
+	spin_unlock_irqrestore(&dev->qp_tbl_lock, flags);
+
+	if (qp && qp->ibqp.event_handler) {
+		struct ib_qp *ibqp = &qp->ibqp;
+		struct ib_event e;
+
+		e.device = ibqp->device;
+		e.element.qp = ibqp;
+		e.event = type; /* 1:1 mapping for now. */
+		ibqp->event_handler(&e, ibqp->qp_context);
+	}
+	if (qp) {
+		atomic_dec(&qp->refcnt);
+		if (atomic_read(&qp->refcnt) == 0)
+			wake_up(&qp->wait);
+	}
+}
+
+static void pvrdma_cq_event(struct pvrdma_dev *dev, u32 cqn, int type)
+{
+	struct pvrdma_cq *cq;
+	unsigned long flags;
+
+	spin_lock_irqsave(&dev->cq_tbl_lock, flags);
+	cq = dev->cq_tbl[cqn % dev->dsr->caps.max_cq];
+	if (cq)
+		atomic_inc(&cq->refcnt);
+	spin_unlock_irqrestore(&dev->cq_tbl_lock, flags);
+
+	if (cq && cq->ibcq.event_handler) {
+		struct ib_cq *ibcq = &cq->ibcq;
+		struct ib_event e;
+
+		e.device = ibcq->device;
+		e.element.cq = ibcq;
+		e.event = type; /* 1:1 mapping for now. */
+		ibcq->event_handler(&e, ibcq->cq_context);
+	}
+	if (cq) {
+		atomic_dec(&cq->refcnt);
+		if (atomic_read(&cq->refcnt) == 0)
+			wake_up(&cq->wait);
+	}
+}
+
+static void pvrdma_dispatch_event(struct pvrdma_dev *dev, int port,
+				  enum ib_event_type event)
+{
+	struct ib_event ib_event;
+
+	memset(&ib_event, 0, sizeof(ib_event));
+	ib_event.device = &dev->ib_dev;
+	ib_event.element.port_num = port;
+	ib_event.event = event;
+	ib_dispatch_event(&ib_event);
+}
+
+static void pvrdma_dev_event(struct pvrdma_dev *dev, u8 port, int type)
+{
+	if (port < 1 || port > dev->dsr->caps.phys_port_cnt) {
+		dev_warn(&dev->pdev->dev, "event on port %d\n", port);
+		return;
+	}
+
+	pvrdma_dispatch_event(dev, port, type);
+}
+
+static inline struct pvrdma_eqe *get_eqe(struct pvrdma_dev *dev, unsigned int i)
+{
+	return (struct pvrdma_eqe *)pvrdma_page_dir_get_ptr(
+					&dev->async_pdir,
+					PAGE_SIZE +
+					sizeof(struct pvrdma_eqe) * i);
+}
+
+static irqreturn_t pvrdma_intr1_handler(int irq, void *dev_id)
+{
+	struct pvrdma_dev *dev = dev_id;
+	struct pvrdma_ring *ring = &dev->async_ring_state->rx;
+	int ring_slots = (dev->dsr->async_ring_pages.num_pages - 1) *
+			 PAGE_SIZE / sizeof(struct pvrdma_eqe);
+	unsigned int head;
+
+	dev_dbg(&dev->pdev->dev, "interrupt 1 (async event) handler\n");
+
+	/*
+	 * Don't process events until the IB device is registered. Otherwise
+	 * we'll try to ib_dispatch_event() on an invalid device.
+	 */
+	if (!dev->ib_active)
+		return IRQ_HANDLED;
+
+	while (pvrdma_idx_ring_has_data(ring, ring_slots, &head) > 0) {
+		struct pvrdma_eqe *eqe;
+
+		eqe = get_eqe(dev, head);
+
+		switch (eqe->type) {
+		case PVRDMA_EVENT_QP_FATAL:
+		case PVRDMA_EVENT_QP_REQ_ERR:
+		case PVRDMA_EVENT_QP_ACCESS_ERR:
+		case PVRDMA_EVENT_COMM_EST:
+		case PVRDMA_EVENT_SQ_DRAINED:
+		case PVRDMA_EVENT_PATH_MIG:
+		case PVRDMA_EVENT_PATH_MIG_ERR:
+		case PVRDMA_EVENT_QP_LAST_WQE_REACHED:
+			pvrdma_qp_event(dev, eqe->info, eqe->type);
+			break;
+
+		case PVRDMA_EVENT_CQ_ERR:
+			pvrdma_cq_event(dev, eqe->info, eqe->type);
+			break;
+
+		case PVRDMA_EVENT_SRQ_ERR:
+		case PVRDMA_EVENT_SRQ_LIMIT_REACHED:
+			break;
+
+		case PVRDMA_EVENT_PORT_ACTIVE:
+		case PVRDMA_EVENT_PORT_ERR:
+		case PVRDMA_EVENT_LID_CHANGE:
+		case PVRDMA_EVENT_PKEY_CHANGE:
+		case PVRDMA_EVENT_SM_CHANGE:
+		case PVRDMA_EVENT_CLIENT_REREGISTER:
+		case PVRDMA_EVENT_GID_CHANGE:
+			pvrdma_dev_event(dev, eqe->info, eqe->type);
+			break;
+
+		case PVRDMA_EVENT_DEVICE_FATAL:
+			pvrdma_dev_event(dev, 1, eqe->type);
+			break;
+
+		default:
+			break;
+		}
+
+		pvrdma_idx_ring_inc(&ring->cons_head, ring_slots);
+	}
+
+	return IRQ_HANDLED;
+}
+
+static inline struct pvrdma_cqne *get_cqne(struct pvrdma_dev *dev,
+					   unsigned int i)
+{
+	return (struct pvrdma_cqne *)pvrdma_page_dir_get_ptr(
+					&dev->cq_pdir,
+					PAGE_SIZE +
+					sizeof(struct pvrdma_cqne) * i);
+}
+
+static irqreturn_t pvrdma_intrx_handler(int irq, void *dev_id)
+{
+	struct pvrdma_dev *dev = dev_id;
+	struct pvrdma_ring *ring = &dev->cq_ring_state->rx;
+	int ring_slots = (dev->dsr->cq_ring_pages.num_pages - 1) * PAGE_SIZE /
+			 sizeof(struct pvrdma_cqne);
+	unsigned int head;
+	unsigned long flags;
+
+	dev_dbg(&dev->pdev->dev, "interrupt x (completion) handler\n");
+
+	while (pvrdma_idx_ring_has_data(ring, ring_slots, &head) > 0) {
+		struct pvrdma_cqne *cqne;
+		struct pvrdma_cq *cq;
+
+		cqne = get_cqne(dev, head);
+		spin_lock_irqsave(&dev->cq_tbl_lock, flags);
+		cq = dev->cq_tbl[cqne->info % dev->dsr->caps.max_cq];
+		if (cq)
+			atomic_inc(&cq->refcnt);
+		spin_unlock_irqrestore(&dev->cq_tbl_lock, flags);
+
+		if (cq && cq->ibcq.comp_handler)
+			cq->ibcq.comp_handler(&cq->ibcq, cq->ibcq.cq_context);
+		if (cq) {
+			atomic_dec(&cq->refcnt);
+			if (atomic_read(&cq->refcnt))
+				wake_up(&cq->wait);
+		}
+		pvrdma_idx_ring_inc(&ring->cons_head, ring_slots);
+	}
+
+	return IRQ_HANDLED;
+}
+
+static void pvrdma_disable_msi_all(struct pvrdma_dev *dev)
+{
+	if (dev->intr.type == PVRDMA_INTR_TYPE_MSIX)
+		pci_disable_msix(dev->pdev);
+	else if (dev->intr.type == PVRDMA_INTR_TYPE_MSI)
+		pci_disable_msi(dev->pdev);
+}
+
+static void pvrdma_free_irq(struct pvrdma_dev *dev)
+{
+	int i;
+
+	dev_dbg(&dev->pdev->dev, "freeing interrupts\n");
+
+	if (dev->intr.type == PVRDMA_INTR_TYPE_MSIX) {
+		for (i = 0; i < dev->intr.size; i++) {
+			if (dev->intr.enabled[i]) {
+				free_irq(dev->intr.msix_entry[i].vector, dev);
+				dev->intr.enabled[i] = 0;
+			}
+		}
+	} else if (dev->intr.type == PVRDMA_INTR_TYPE_INTX ||
+		   dev->intr.type == PVRDMA_INTR_TYPE_MSI) {
+		free_irq(dev->pdev->irq, dev);
+	}
+}
+
+static void pvrdma_enable_intrs(struct pvrdma_dev *dev)
+{
+	dev_dbg(&dev->pdev->dev, "enable interrupts\n");
+	pvrdma_write_reg(dev, PVRDMA_REG_IMR, 0);
+}
+
+static void pvrdma_disable_intrs(struct pvrdma_dev *dev)
+{
+	dev_dbg(&dev->pdev->dev, "disable interrupts\n");
+	pvrdma_write_reg(dev, PVRDMA_REG_IMR, ~0);
+}
+
+static int pvrdma_enable_msix(struct pci_dev *pdev, struct pvrdma_dev *dev)
+{
+	int i;
+	int ret;
+
+	for (i = 0; i < PVRDMA_MAX_INTERRUPTS; i++) {
+		dev->intr.msix_entry[i].entry = i;
+		dev->intr.msix_entry[i].vector = i;
+
+		switch (i) {
+		case 0:
+			/* CMD ring handler */
+			dev->intr.handler[i] = pvrdma_intr0_handler;
+			break;
+		case 1:
+			/* Async event ring handler */
+			dev->intr.handler[i] = pvrdma_intr1_handler;
+			break;
+		default:
+			/* Completion queue handler */
+			dev->intr.handler[i] = pvrdma_intrx_handler;
+			break;
+		}
+	}
+
+	ret = pci_enable_msix(pdev, dev->intr.msix_entry,
+			      PVRDMA_MAX_INTERRUPTS);
+	if (!ret) {
+		dev->intr.type = PVRDMA_INTR_TYPE_MSIX;
+		dev->intr.size = PVRDMA_MAX_INTERRUPTS;
+	} else if (ret > 0) {
+		ret = pci_enable_msix(pdev, dev->intr.msix_entry, ret);
+		if (!ret) {
+			dev->intr.type = PVRDMA_INTR_TYPE_MSIX;
+			dev->intr.size = ret;
+		} else {
+			dev->intr.size = 0;
+		}
+	}
+
+	dev_dbg(&pdev->dev, "using interrupt type %d, size %d\n",
+		dev->intr.type, dev->intr.size);
+
+	return ret;
+}
+
+static int pvrdma_alloc_intrs(struct pvrdma_dev *dev)
+{
+	int ret = 0;
+	int i;
+
+	if (pci_find_capability(dev->pdev, PCI_CAP_ID_MSIX) &&
+	    pvrdma_enable_msix(dev->pdev, dev)) {
+		/* Try MSI */
+		ret = pci_enable_msi(dev->pdev);
+		if (!ret) {
+			dev->intr.type = PVRDMA_INTR_TYPE_MSI;
+		} else {
+			/* Legacy INTR */
+			dev->intr.type = PVRDMA_INTR_TYPE_INTX;
+		}
+	}
+
+	/* Request First IRQ */
+	switch (dev->intr.type) {
+	case PVRDMA_INTR_TYPE_INTX:
+	case PVRDMA_INTR_TYPE_MSI:
+		ret = request_irq(dev->pdev->irq, pvrdma_intr0_handler,
+				  IRQF_SHARED, DRV_NAME, dev);
+		if (ret) {
+			dev_err(&dev->pdev->dev,
+				"failed to request interrupt\n");
+			goto disable_msi;
+		}
+		break;
+	case PVRDMA_INTR_TYPE_MSIX:
+		ret = request_irq(dev->intr.msix_entry[0].vector,
+				  pvrdma_intr0_handler, 0, DRV_NAME, dev);
+		if (ret) {
+			dev_err(&dev->pdev->dev,
+				"failed to request interrupt 0\n");
+			goto disable_msi;
+		}
+		dev->intr.enabled[0] = 1;
+		break;
+	default:
+		/* Not reached */
+		break;
+	}
+
+	/* For MSIX: request intr for each vector */
+	if (dev->intr.size > 1) {
+		ret = request_irq(dev->intr.msix_entry[1].vector,
+				  pvrdma_intr1_handler, 0, DRV_NAME, dev);
+		if (ret) {
+			dev_err(&dev->pdev->dev,
+				"failed to request interrupt 1\n");
+			goto free_irq;
+		}
+		dev->intr.enabled[1] = 1;
+
+		for (i = 2; i < dev->intr.size; i++) {
+			ret = request_irq(dev->intr.msix_entry[i].vector,
+					  pvrdma_intrx_handler, 0,
+					  DRV_NAME, dev);
+			if (ret) {
+				dev_err(&dev->pdev->dev,
+					"failed to request interrupt %d\n", i);
+				goto free_irq;
+			}
+			dev->intr.enabled[i] = 1;
+		}
+	}
+
+	return 0;
+
+free_irq:
+	pvrdma_free_irq(dev);
+disable_msi:
+	pvrdma_disable_msi_all(dev);
+	return ret;
+}
+
+static void pvrdma_free_slots(struct pvrdma_dev *dev)
+{
+	struct pci_dev *pdev = dev->pdev;
+
+	if (dev->resp_slot)
+		dma_free_coherent(&pdev->dev, PAGE_SIZE, dev->resp_slot,
+				  dev->dsr->resp_slot_dma);
+	if (dev->cmd_slot)
+		dma_free_coherent(&pdev->dev, PAGE_SIZE, dev->cmd_slot,
+				  dev->dsr->cmd_slot_dma);
+}
+
+static int pvrdma_add_gid_at_index(struct pvrdma_dev *dev,
+				   const union ib_gid *gid,
+				   int index)
+{
+	int ret;
+	union pvrdma_cmd_req req;
+	struct pvrdma_cmd_create_bind *cmd_bind = &req.create_bind;
+
+	if (!dev->sgid_tbl) {
+		dev_warn(&dev->pdev->dev, "sgid table not initialized\n");
+		return -EINVAL;
+	}
+
+	memset(cmd_bind, 0, sizeof(*cmd_bind));
+	cmd_bind->hdr.cmd = PVRDMA_CMD_CREATE_BIND;
+	memcpy(cmd_bind->new_gid, gid->raw, 16);
+	cmd_bind->mtu = ib_mtu_enum_to_int(IB_MTU_1024);
+	cmd_bind->vlan = 0xfff;
+	cmd_bind->index = index;
+	cmd_bind->gid_type = PVRDMA_GID_TYPE_FLAG_ROCE_V1;
+
+	ret = pvrdma_cmd_post(dev, &req, NULL, 0);
+	if (ret < 0) {
+		dev_warn(&dev->pdev->dev,
+			 "could not create binding, error: %d\n", ret);
+		return -EFAULT;
+	}
+	memcpy(&dev->sgid_tbl[index], gid, sizeof(*gid));
+	return 0;
+}
+
+static int pvrdma_add_gid(struct ib_device *ibdev,
+			  u8 port_num,
+			  unsigned int index,
+			  const union ib_gid *gid,
+			  const struct ib_gid_attr *attr,
+			  void **context)
+{
+	struct pvrdma_dev *dev = to_vdev(ibdev);
+
+	return pvrdma_add_gid_at_index(dev, gid, index);
+}
+
+static int pvrdma_del_gid_at_index(struct pvrdma_dev *dev, int index)
+{
+	int ret;
+	union pvrdma_cmd_req req;
+	struct pvrdma_cmd_destroy_bind *cmd_dest = &req.destroy_bind;
+
+	/* Update sgid table. */
+	if (!dev->sgid_tbl) {
+		dev_warn(&dev->pdev->dev, "sgid table not initialized\n");
+		return -EINVAL;
+	}
+
+	memset(cmd_dest, 0, sizeof(*cmd_dest));
+	cmd_dest->hdr.cmd = PVRDMA_CMD_DESTROY_BIND;
+	memcpy(cmd_dest->dest_gid, &dev->sgid_tbl[index], 16);
+	cmd_dest->index = index;
+
+	ret = pvrdma_cmd_post(dev, &req, NULL, 0);
+	if (ret < 0) {
+		dev_warn(&dev->pdev->dev,
+			 "could not destroy binding, error: %d\n", ret);
+		return ret;
+	}
+	memset(&dev->sgid_tbl[index], 0, 16);
+	return 0;
+}
+
+static int pvrdma_del_gid(struct ib_device *ibdev,
+			  u8 port_num,
+			  unsigned int index,
+			  void **context)
+{
+	struct pvrdma_dev *dev = to_vdev(ibdev);
+
+	dev_dbg(&dev->pdev->dev, "removing gid at index %u from %s",
+		index, dev->netdev->name);
+
+	return pvrdma_del_gid_at_index(dev, index);
+}
+
+static void pvrdma_netdevice_event_handle(struct pvrdma_dev *dev,
+					  unsigned long event)
+{
+	switch (event) {
+	case NETDEV_REBOOT:
+	case NETDEV_DOWN:
+		pvrdma_dispatch_event(dev, 1, IB_EVENT_PORT_ERR);
+		break;
+	case NETDEV_UP:
+		pvrdma_dispatch_event(dev, 1, IB_EVENT_PORT_ACTIVE);
+		break;
+	default:
+		dev_dbg(&dev->pdev->dev, "ignore netdevice event %ld on %s\n",
+			event, dev->ib_dev.name);
+		break;
+	}
+}
+
+static void pvrdma_netdevice_event_work(struct work_struct *work)
+{
+	struct pvrdma_netdevice_work *netdev_work;
+	struct pvrdma_dev *dev;
+
+	netdev_work = container_of(work, struct pvrdma_netdevice_work, work);
+
+	mutex_lock(&pvrdma_device_list_lock);
+	list_for_each_entry(dev, &pvrdma_device_list, device_link) {
+		if (dev->netdev == netdev_work->event_netdev) {
+			pvrdma_netdevice_event_handle(dev, netdev_work->event);
+			break;
+		}
+	}
+	mutex_unlock(&pvrdma_device_list_lock);
+
+	kfree(netdev_work);
+}
+
+static int pvrdma_netdevice_event(struct notifier_block *this,
+				  unsigned long event, void *ptr)
+{
+	struct net_device *event_netdev = netdev_notifier_info_to_dev(ptr);
+	struct pvrdma_netdevice_work *netdev_work;
+
+	netdev_work = kmalloc(sizeof(*netdev_work), GFP_ATOMIC);
+	if (!netdev_work)
+		return NOTIFY_BAD;
+
+	INIT_WORK(&netdev_work->work, pvrdma_netdevice_event_work);
+	netdev_work->event_netdev = event_netdev;
+	netdev_work->event = event;
+	queue_work(event_wq, &netdev_work->work);
+
+	return NOTIFY_DONE;
+}
+
+static int pvrdma_pci_probe(struct pci_dev *pdev,
+			    const struct pci_device_id *id)
+{
+	struct pci_dev *pdev_net;
+	struct pvrdma_dev *dev;
+	int ret;
+	unsigned long start;
+	unsigned long len;
+	unsigned int version;
+	dma_addr_t slot_dma = 0;
+
+	dev_dbg(&pdev->dev, "initializing driver %s\n", pci_name(pdev));
+
+	/* Allocate zero-out device */
+	dev = (struct pvrdma_dev *)ib_alloc_device(sizeof(*dev));
+	if (!dev) {
+		dev_err(&pdev->dev, "failed to allocate IB device\n");
+		return -ENOMEM;
+	}
+
+	mutex_lock(&pvrdma_device_list_lock);
+	list_add(&dev->device_link, &pvrdma_device_list);
+	mutex_unlock(&pvrdma_device_list_lock);
+
+	ret = pvrdma_init_device(dev);
+	if (ret)
+		goto err_free_device;
+
+	dev->pdev = pdev;
+	pci_set_drvdata(pdev, dev);
+
+	ret = pci_enable_device(pdev);
+	if (ret) {
+		dev_err(&pdev->dev, "cannot enable PCI device\n");
+		goto err_free_device;
+	}
+
+	dev_dbg(&pdev->dev, "PCI resource flags BAR0 %#lx\n",
+		pci_resource_flags(pdev, 0));
+	dev_dbg(&pdev->dev, "PCI resource len %#llx\n",
+		(unsigned long long)pci_resource_len(pdev, 0));
+	dev_dbg(&pdev->dev, "PCI resource start %#llx\n",
+		(unsigned long long)pci_resource_start(pdev, 0));
+	dev_dbg(&pdev->dev, "PCI resource flags BAR1 %#lx\n",
+		pci_resource_flags(pdev, 1));
+	dev_dbg(&pdev->dev, "PCI resource len %#llx\n",
+		(unsigned long long)pci_resource_len(pdev, 1));
+	dev_dbg(&pdev->dev, "PCI resource start %#llx\n",
+		(unsigned long long)pci_resource_start(pdev, 1));
+
+	if (!(pci_resource_flags(pdev, 0) & IORESOURCE_MEM) ||
+	    !(pci_resource_flags(pdev, 1) & IORESOURCE_MEM)) {
+		dev_err(&pdev->dev, "PCI BAR region not MMIO\n");
+		ret = -ENOMEM;
+		goto err_free_device;
+	}
+
+	ret = pci_request_regions(pdev, DRV_NAME);
+	if (ret) {
+		dev_err(&pdev->dev, "cannot request PCI resources\n");
+		goto err_disable_pdev;
+	}
+
+	/* Enable 64-Bit DMA */
+	if (pci_set_dma_mask(pdev, DMA_BIT_MASK(64)) == 0) {
+		ret = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(64));
+		if (ret != 0) {
+			dev_err(&pdev->dev,
+				"pci_set_consistent_dma_mask failed\n");
+			goto err_free_resource;
+		}
+	} else {
+		ret = pci_set_dma_mask(pdev, DMA_BIT_MASK(32));
+		if (ret != 0) {
+			dev_err(&pdev->dev,
+				"pci_set_dma_mask failed\n");
+			goto err_free_resource;
+		}
+	}
+
+	pci_set_master(pdev);
+
+	/* Map register space */
+	start = pci_resource_start(dev->pdev, PVRDMA_PCI_RESOURCE_REG);
+	len = pci_resource_len(dev->pdev, PVRDMA_PCI_RESOURCE_REG);
+	dev->regs = ioremap(start, len);
+	if (!dev->regs) {
+		dev_err(&pdev->dev, "register mapping failed\n");
+		ret = -ENOMEM;
+		goto err_free_resource;
+	}
+
+	/* Setup per-device UAR. */
+	dev->driver_uar.index = 0;
+	dev->driver_uar.pfn =
+		pci_resource_start(dev->pdev, PVRDMA_PCI_RESOURCE_UAR) >>
+		PAGE_SHIFT;
+	dev->driver_uar.map =
+		ioremap(dev->driver_uar.pfn << PAGE_SHIFT, PAGE_SIZE);
+	if (!dev->driver_uar.map) {
+		dev_err(&pdev->dev, "failed to remap UAR pages\n");
+		ret = -ENOMEM;
+		goto err_unmap_regs;
+	}
+
+	version = pvrdma_read_reg(dev, PVRDMA_REG_VERSION);
+	dev_info(&pdev->dev, "device version %d, driver version %d\n",
+		 version, PVRDMA_VERSION);
+	if (version < PVRDMA_VERSION) {
+		dev_err(&pdev->dev, "incompatible device version\n");
+		goto err_uar_unmap;
+	}
+
+	dev->dsr = dma_alloc_coherent(&pdev->dev, sizeof(*dev->dsr),
+				      &dev->dsrbase, GFP_KERNEL);
+	if (!dev->dsr) {
+		dev_err(&pdev->dev, "failed to allocate shared region\n");
+		ret = -ENOMEM;
+		goto err_uar_unmap;
+	}
+
+	/* Setup the shared region */
+	memset(dev->dsr, 0, sizeof(*dev->dsr));
+	dev->dsr->driver_version = PVRDMA_VERSION;
+	dev->dsr->gos_info.gos_bits = sizeof(void *) == 4 ?
+		PVRDMA_GOS_BITS_32 :
+		PVRDMA_GOS_BITS_64;
+	dev->dsr->gos_info.gos_type = PVRDMA_GOS_TYPE_LINUX;
+	dev->dsr->gos_info.gos_ver = 1;
+	dev->dsr->uar_pfn = dev->driver_uar.pfn;
+
+	/* Command slot. */
+	dev->cmd_slot = dma_alloc_coherent(&pdev->dev, PAGE_SIZE,
+					   &slot_dma, GFP_KERNEL);
+	if (!dev->cmd_slot) {
+		ret = -ENOMEM;
+		goto err_free_dsr;
+	}
+
+	dev->dsr->cmd_slot_dma = (u64)slot_dma;
+
+	/* Response slot. */
+	dev->resp_slot = dma_alloc_coherent(&pdev->dev, PAGE_SIZE,
+					    &slot_dma, GFP_KERNEL);
+	if (!dev->resp_slot) {
+		ret = -ENOMEM;
+		goto err_free_slots;
+	}
+
+	dev->dsr->resp_slot_dma = (u64)slot_dma;
+
+	/* Async event ring */
+	dev->dsr->async_ring_pages.num_pages = 4;
+	ret = pvrdma_page_dir_init(dev, &dev->async_pdir,
+				   dev->dsr->async_ring_pages.num_pages, true);
+	if (ret)
+		goto err_free_slots;
+	dev->async_ring_state = dev->async_pdir.pages[0];
+	dev->dsr->async_ring_pages.pdir_dma = dev->async_pdir.dir_dma;
+
+	/* CQ notification ring */
+	dev->dsr->cq_ring_pages.num_pages = 4;
+	ret = pvrdma_page_dir_init(dev, &dev->cq_pdir,
+				   dev->dsr->cq_ring_pages.num_pages, true);
+	if (ret)
+		goto err_free_async_ring;
+	dev->cq_ring_state = dev->cq_pdir.pages[0];
+	dev->dsr->cq_ring_pages.pdir_dma = dev->cq_pdir.dir_dma;
+
+	/*
+	 * Write the PA of the shared region to the device. The writes must be
+	 * ordered such that the high bits are written last. When the writes
+	 * complete, the device will have filled out the capabilities.
+	 */
+
+	pvrdma_write_reg(dev, PVRDMA_REG_DSRLOW, (u32)dev->dsrbase);
+	pvrdma_write_reg(dev, PVRDMA_REG_DSRHIGH,
+			 (u32)((u64)(dev->dsrbase) >> 32));
+
+	/* Make sure the write is complete before reading status. */
+	mb();
+
+	/* Currently, the driver only supports RoCE mode. */
+	if (dev->dsr->caps.mode != PVRDMA_DEVICE_MODE_ROCE) {
+		dev_err(&pdev->dev, "unsupported transport %d\n",
+			dev->dsr->caps.mode);
+		ret = -EFAULT;
+		goto err_free_cq_ring;
+	}
+
+	/* Currently, the driver only supports RoCE V1. */
+	if (!(dev->dsr->caps.gid_types & PVRDMA_GID_TYPE_FLAG_ROCE_V1)) {
+		dev_err(&pdev->dev, "driver needs RoCE v1 support\n");
+		ret = -EFAULT;
+		goto err_free_cq_ring;
+	}
+
+	/* Paired vmxnet3 will have same bus, slot. But func will be 0 */
+	pdev_net = pci_get_slot(pdev->bus, PCI_DEVFN(PCI_SLOT(pdev->devfn), 0));
+	if (!pdev_net) {
+		dev_err(&pdev->dev, "failed to find paired net device\n");
+		ret = -ENODEV;
+		goto err_free_cq_ring;
+	}
+
+	if (pdev_net->vendor != PCI_VENDOR_ID_VMWARE ||
+	    pdev_net->device != PCI_DEVICE_ID_VMWARE_VMXNET3) {
+		dev_err(&pdev->dev, "failed to find paired vmxnet3 device\n");
+		pci_dev_put(pdev_net);
+		ret = -ENODEV;
+		goto err_free_cq_ring;
+	}
+
+	dev->netdev = pci_get_drvdata(pdev_net);
+	pci_dev_put(pdev_net);
+	if (!dev->netdev) {
+		dev_err(&pdev->dev, "failed to get vmxnet3 device\n");
+		ret = -ENODEV;
+		goto err_free_cq_ring;
+	}
+
+	dev_info(&pdev->dev, "paired device to %s\n", dev->netdev->name);
+
+	/* Interrupt setup */
+	ret = pvrdma_alloc_intrs(dev);
+	if (ret) {
+		dev_err(&pdev->dev, "failed to allocate interrupts\n");
+		ret = -ENOMEM;
+		goto err_netdevice;
+	}
+
+	/* Allocate UAR table. */
+	ret = pvrdma_uar_table_init(dev);
+	if (ret) {
+		dev_err(&pdev->dev, "failed to allocate UAR table\n");
+		ret = -ENOMEM;
+		goto err_free_intrs;
+	}
+
+	/* Allocate GID table */
+	dev->sgid_tbl = kcalloc(dev->dsr->caps.gid_tbl_len,
+				sizeof(union ib_gid), GFP_KERNEL);
+	if (!dev->sgid_tbl) {
+		ret = -ENOMEM;
+		goto err_free_uar_table;
+	}
+	dev_dbg(&pdev->dev, "gid table len %d\n", dev->dsr->caps.gid_tbl_len);
+
+	pvrdma_enable_intrs(dev);
+
+	/* Activate pvrdma device */
+	pvrdma_write_reg(dev, PVRDMA_REG_CTL, PVRDMA_DEVICE_CTL_ACTIVATE);
+
+	/* Make sure the write is complete before reading status. */
+	mb();
+
+	/* Check if device was successfully activated */
+	ret = pvrdma_read_reg(dev, PVRDMA_REG_ERR);
+	if (ret != 0) {
+		dev_err(&pdev->dev, "failed to activate device\n");
+		ret = -EFAULT;
+		goto err_disable_intr;
+	}
+
+	/* Register IB device */
+	ret = pvrdma_register_device(dev);
+	if (ret) {
+		dev_err(&pdev->dev, "failed to register IB device\n");
+		goto err_disable_intr;
+	}
+
+	dev->nb_netdev.notifier_call = pvrdma_netdevice_event;
+	ret = register_netdevice_notifier(&dev->nb_netdev);
+	if (ret) {
+		dev_err(&pdev->dev, "failed to register netdevice events\n");
+		goto err_unreg_ibdev;
+	}
+
+	dev_info(&pdev->dev, "attached to device\n");
+	return 0;
+
+err_unreg_ibdev:
+	ib_unregister_device(&dev->ib_dev);
+err_disable_intr:
+	pvrdma_disable_intrs(dev);
+	kfree(dev->sgid_tbl);
+err_free_uar_table:
+	pvrdma_uar_table_cleanup(dev);
+err_free_intrs:
+	pvrdma_free_irq(dev);
+	pvrdma_disable_msi_all(dev);
+err_netdevice:
+	unregister_netdevice_notifier(&dev->nb_netdev);
+err_free_cq_ring:
+	pvrdma_page_dir_cleanup(dev, &dev->cq_pdir);
+err_free_async_ring:
+	pvrdma_page_dir_cleanup(dev, &dev->async_pdir);
+err_free_slots:
+	pvrdma_free_slots(dev);
+err_free_dsr:
+	dma_free_coherent(&pdev->dev, sizeof(*dev->dsr), dev->dsr,
+			  dev->dsrbase);
+err_uar_unmap:
+	iounmap(dev->driver_uar.map);
+err_unmap_regs:
+	iounmap(dev->regs);
+err_free_resource:
+	pci_release_regions(pdev);
+err_disable_pdev:
+	pci_disable_device(pdev);
+	pci_set_drvdata(pdev, NULL);
+err_free_device:
+	mutex_lock(&pvrdma_device_list_lock);
+	list_del(&dev->device_link);
+	mutex_unlock(&pvrdma_device_list_lock);
+	ib_dealloc_device(&dev->ib_dev);
+	return ret;
+}
+
+static void pvrdma_pci_remove(struct pci_dev *pdev)
+{
+	struct pvrdma_dev *dev = pci_get_drvdata(pdev);
+
+	if (!dev)
+		return;
+
+	dev_info(&pdev->dev, "detaching from device\n");
+
+	unregister_netdevice_notifier(&dev->nb_netdev);
+	dev->nb_netdev.notifier_call = NULL;
+
+	flush_workqueue(event_wq);
+
+	/* Unregister ib device */
+	ib_unregister_device(&dev->ib_dev);
+
+	mutex_lock(&pvrdma_device_list_lock);
+	list_del(&dev->device_link);
+	mutex_unlock(&pvrdma_device_list_lock);
+
+	pvrdma_disable_intrs(dev);
+	pvrdma_free_irq(dev);
+	pvrdma_disable_msi_all(dev);
+
+	/* Deactivate pvrdma device */
+	pvrdma_write_reg(dev, PVRDMA_REG_CTL, PVRDMA_DEVICE_CTL_RESET);
+	pvrdma_page_dir_cleanup(dev, &dev->cq_pdir);
+	pvrdma_page_dir_cleanup(dev, &dev->async_pdir);
+	pvrdma_free_slots(dev);
+
+	iounmap(dev->regs);
+	kfree(dev->sgid_tbl);
+	kfree(dev->cq_tbl);
+	kfree(dev->qp_tbl);
+	pvrdma_uar_table_cleanup(dev);
+	iounmap(dev->driver_uar.map);
+
+	ib_dealloc_device(&dev->ib_dev);
+
+	/* Free pci resources */
+	pci_release_regions(pdev);
+	pci_disable_device(pdev);
+	pci_set_drvdata(pdev, NULL);
+}
+
+static struct pci_device_id pvrdma_pci_table[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_VMWARE, PCI_DEVICE_ID_VMWARE_PVRDMA), },
+	{ 0 },
+};
+
+MODULE_DEVICE_TABLE(pci, pvrdma_pci_table);
+
+static struct pci_driver pvrdma_driver = {
+	.name		= DRV_NAME,
+	.id_table	= pvrdma_pci_table,
+	.probe		= pvrdma_pci_probe,
+	.remove		= pvrdma_pci_remove,
+};
+
+static int __init pvrdma_init(void)
+{
+	int err;
+
+	event_wq = alloc_ordered_workqueue("pvrdma_event_wq", WQ_MEM_RECLAIM);
+	if (!event_wq)
+		return -ENOMEM;
+
+	err = pci_register_driver(&pvrdma_driver);
+	if (err)
+		destroy_workqueue(event_wq);
+
+	return err;
+}
+
+static void __exit pvrdma_cleanup(void)
+{
+	pci_unregister_driver(&pvrdma_driver);
+
+	destroy_workqueue(event_wq);
+}
+
+module_init(pvrdma_init);
+module_exit(pvrdma_cleanup);
+
+MODULE_AUTHOR("VMware, Inc");
+MODULE_DESCRIPTION("VMware Paravirtual RDMA driver");
+MODULE_VERSION(DRV_VERSION);
+MODULE_LICENSE("Dual BSD/GPL");
diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_misc.c b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_misc.c
new file mode 100644
index 000000000000..948b5ccd2a70
--- /dev/null
+++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_misc.c
@@ -0,0 +1,304 @@
+/*
+ * Copyright (c) 2012-2016 VMware, Inc.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of EITHER the GNU General Public License
+ * version 2 as published by the Free Software Foundation or the BSD
+ * 2-Clause License. This program is distributed in the hope that it
+ * will be useful, but WITHOUT ANY WARRANTY; WITHOUT EVEN THE IMPLIED
+ * WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
+ * See the GNU General Public License version 2 for more details at
+ * http://www.gnu.org/licenses/old-licenses/gpl-2.0.en.html.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program available in the file COPYING in the main
+ * directory of this source tree.
+ *
+ * The BSD 2-Clause License
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+ * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
+ * COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT,
+ * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <linux/errno.h>
+#include <linux/slab.h>
+#include <linux/bitmap.h>
+
+#include "pvrdma.h"
+
+int pvrdma_page_dir_init(struct pvrdma_dev *dev, struct pvrdma_page_dir *pdir,
+			 u64 npages, bool alloc_pages)
+{
+	u64 i;
+
+	if (npages > PVRDMA_PAGE_DIR_MAX_PAGES)
+		return -EINVAL;
+
+	memset(pdir, 0, sizeof(*pdir));
+
+	pdir->dir = dma_alloc_coherent(&dev->pdev->dev, PAGE_SIZE,
+				       &pdir->dir_dma, GFP_KERNEL);
+	if (!pdir->dir)
+		goto err;
+
+	pdir->ntables = PVRDMA_PAGE_DIR_TABLE(npages - 1) + 1;
+	pdir->tables = kcalloc(pdir->ntables, sizeof(*pdir->tables),
+			       GFP_KERNEL);
+	if (!pdir->tables)
+		goto err;
+
+	for (i = 0; i < pdir->ntables; i++) {
+		pdir->tables[i] = dma_alloc_coherent(&dev->pdev->dev, PAGE_SIZE,
+						(dma_addr_t *)&pdir->dir[i],
+						GFP_KERNEL);
+		if (!pdir->tables[i])
+			goto err;
+	}
+
+	pdir->npages = npages;
+
+	if (alloc_pages) {
+		pdir->pages = kcalloc(npages, sizeof(*pdir->pages),
+				      GFP_KERNEL);
+		if (!pdir->pages)
+			goto err;
+
+		for (i = 0; i < pdir->npages; i++) {
+			dma_addr_t page_dma;
+
+			pdir->pages[i] = dma_alloc_coherent(&dev->pdev->dev,
+							    PAGE_SIZE,
+							    &page_dma,
+							    GFP_KERNEL);
+			if (!pdir->pages[i])
+				goto err;
+
+			pvrdma_page_dir_insert_dma(pdir, i, page_dma);
+		}
+	}
+
+	return 0;
+
+err:
+	pvrdma_page_dir_cleanup(dev, pdir);
+
+	return -ENOMEM;
+}
+
+static u64 *pvrdma_page_dir_table(struct pvrdma_page_dir *pdir, u64 idx)
+{
+	return pdir->tables[PVRDMA_PAGE_DIR_TABLE(idx)];
+}
+
+dma_addr_t pvrdma_page_dir_get_dma(struct pvrdma_page_dir *pdir, u64 idx)
+{
+	return pvrdma_page_dir_table(pdir, idx)[PVRDMA_PAGE_DIR_PAGE(idx)];
+}
+
+static void pvrdma_page_dir_cleanup_pages(struct pvrdma_dev *dev,
+					  struct pvrdma_page_dir *pdir)
+{
+	if (pdir->pages) {
+		u64 i;
+
+		for (i = 0; i < pdir->npages && pdir->pages[i]; i++) {
+			dma_addr_t page_dma = pvrdma_page_dir_get_dma(pdir, i);
+
+			dma_free_coherent(&dev->pdev->dev, PAGE_SIZE,
+					  pdir->pages[i], page_dma);
+		}
+
+		kfree(pdir->pages);
+	}
+}
+
+static void pvrdma_page_dir_cleanup_tables(struct pvrdma_dev *dev,
+					   struct pvrdma_page_dir *pdir)
+{
+	if (pdir->tables) {
+		int i;
+
+		pvrdma_page_dir_cleanup_pages(dev, pdir);
+
+		for (i = 0; i < pdir->ntables; i++) {
+			u64 *table = pdir->tables[i];
+
+			if (table)
+				dma_free_coherent(&dev->pdev->dev, PAGE_SIZE,
+						  table, pdir->dir[i]);
+		}
+
+		kfree(pdir->tables);
+	}
+}
+
+void pvrdma_page_dir_cleanup(struct pvrdma_dev *dev,
+			     struct pvrdma_page_dir *pdir)
+{
+	if (pdir->dir) {
+		pvrdma_page_dir_cleanup_tables(dev, pdir);
+		dma_free_coherent(&dev->pdev->dev, PAGE_SIZE,
+				  pdir->dir, pdir->dir_dma);
+	}
+}
+
+int pvrdma_page_dir_insert_dma(struct pvrdma_page_dir *pdir, u64 idx,
+			       dma_addr_t daddr)
+{
+	u64 *table;
+
+	if (idx >= pdir->npages)
+		return -EINVAL;
+
+	table = pvrdma_page_dir_table(pdir, idx);
+	table[PVRDMA_PAGE_DIR_PAGE(idx)] = daddr;
+
+	return 0;
+}
+
+int pvrdma_page_dir_insert_umem(struct pvrdma_page_dir *pdir,
+				struct ib_umem *umem, u64 offset)
+{
+	u64 i = offset;
+	int j, entry;
+	int ret = 0, len = 0;
+	struct scatterlist *sg;
+
+	if (offset >= pdir->npages)
+		return -EINVAL;
+
+	for_each_sg(umem->sg_head.sgl, sg, umem->nmap, entry) {
+		len = sg_dma_len(sg) >> PAGE_SHIFT;
+		for (j = 0; j < len; j++) {
+			dma_addr_t addr = sg_dma_address(sg) +
+					  umem->page_size * j;
+
+			ret = pvrdma_page_dir_insert_dma(pdir, i, addr);
+			if (ret)
+				goto exit;
+
+			i++;
+		}
+	}
+
+exit:
+	return ret;
+}
+
+int pvrdma_page_dir_insert_page_list(struct pvrdma_page_dir *pdir,
+				     u64 *page_list,
+				     int num_pages)
+{
+	int i;
+	int ret;
+
+	if (num_pages > pdir->npages)
+		return -EINVAL;
+
+	for (i = 0; i < num_pages; i++) {
+		ret = pvrdma_page_dir_insert_dma(pdir, i, page_list[i]);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+void pvrdma_qp_cap_to_ib(struct ib_qp_cap *dst, const struct pvrdma_qp_cap *src)
+{
+	dst->max_send_wr = src->max_send_wr;
+	dst->max_recv_wr = src->max_recv_wr;
+	dst->max_send_sge = src->max_send_sge;
+	dst->max_recv_sge = src->max_recv_sge;
+	dst->max_inline_data = src->max_inline_data;
+}
+
+void ib_qp_cap_to_pvrdma(struct pvrdma_qp_cap *dst, const struct ib_qp_cap *src)
+{
+	dst->max_send_wr = src->max_send_wr;
+	dst->max_recv_wr = src->max_recv_wr;
+	dst->max_send_sge = src->max_send_sge;
+	dst->max_recv_sge = src->max_recv_sge;
+	dst->max_inline_data = src->max_inline_data;
+}
+
+void pvrdma_gid_to_ib(union ib_gid *dst, const union pvrdma_gid *src)
+{
+	BUILD_BUG_ON(sizeof(union pvrdma_gid) != sizeof(union ib_gid));
+	memcpy(dst, src, sizeof(*src));
+}
+
+void ib_gid_to_pvrdma(union pvrdma_gid *dst, const union ib_gid *src)
+{
+	BUILD_BUG_ON(sizeof(union pvrdma_gid) != sizeof(union ib_gid));
+	memcpy(dst, src, sizeof(*src));
+}
+
+void pvrdma_global_route_to_ib(struct ib_global_route *dst,
+			       const struct pvrdma_global_route *src)
+{
+	pvrdma_gid_to_ib(&dst->dgid, &src->dgid);
+	dst->flow_label = src->flow_label;
+	dst->sgid_index = src->sgid_index;
+	dst->hop_limit = src->hop_limit;
+	dst->traffic_class = src->traffic_class;
+}
+
+void ib_global_route_to_pvrdma(struct pvrdma_global_route *dst,
+			       const struct ib_global_route *src)
+{
+	ib_gid_to_pvrdma(&dst->dgid, &src->dgid);
+	dst->flow_label = src->flow_label;
+	dst->sgid_index = src->sgid_index;
+	dst->hop_limit = src->hop_limit;
+	dst->traffic_class = src->traffic_class;
+}
+
+void pvrdma_ah_attr_to_ib(struct ib_ah_attr *dst,
+			  const struct pvrdma_ah_attr *src)
+{
+	pvrdma_global_route_to_ib(&dst->grh, &src->grh);
+	dst->dlid = src->dlid;
+	dst->sl = src->sl;
+	dst->src_path_bits = src->src_path_bits;
+	dst->static_rate = src->static_rate;
+	dst->ah_flags = src->ah_flags;
+	dst->port_num = src->port_num;
+	memcpy(&dst->dmac, &src->dmac, sizeof(dst->dmac));
+}
+
+void ib_ah_attr_to_pvrdma(struct pvrdma_ah_attr *dst,
+			  const struct ib_ah_attr *src)
+{
+	ib_global_route_to_pvrdma(&dst->grh, &src->grh);
+	dst->dlid = src->dlid;
+	dst->sl = src->sl;
+	dst->src_path_bits = src->src_path_bits;
+	dst->static_rate = src->static_rate;
+	dst->ah_flags = src->ah_flags;
+	dst->port_num = src->port_num;
+	memcpy(&dst->dmac, &src->dmac, sizeof(dst->dmac));
+}
diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_mr.c b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_mr.c
new file mode 100644
index 000000000000..8519f3212e52
--- /dev/null
+++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_mr.c
@@ -0,0 +1,334 @@
+/*
+ * Copyright (c) 2012-2016 VMware, Inc.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of EITHER the GNU General Public License
+ * version 2 as published by the Free Software Foundation or the BSD
+ * 2-Clause License. This program is distributed in the hope that it
+ * will be useful, but WITHOUT ANY WARRANTY; WITHOUT EVEN THE IMPLIED
+ * WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
+ * See the GNU General Public License version 2 for more details at
+ * http://www.gnu.org/licenses/old-licenses/gpl-2.0.en.html.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program available in the file COPYING in the main
+ * directory of this source tree.
+ *
+ * The BSD 2-Clause License
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+ * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
+ * COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT,
+ * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <linux/list.h>
+#include <linux/slab.h>
+
+#include "pvrdma.h"
+
+/**
+ * pvrdma_get_dma_mr - get a DMA memory region
+ * @pd: protection domain
+ * @acc: access flags
+ *
+ * @return: ib_mr pointer on success, otherwise returns an errno.
+ */
+struct ib_mr *pvrdma_get_dma_mr(struct ib_pd *pd, int acc)
+{
+	struct pvrdma_dev *dev = to_vdev(pd->device);
+	struct pvrdma_user_mr *mr;
+	union pvrdma_cmd_req req;
+	union pvrdma_cmd_resp rsp;
+	struct pvrdma_cmd_create_mr *cmd = &req.create_mr;
+	struct pvrdma_cmd_create_mr_resp *resp = &rsp.create_mr_resp;
+	int ret;
+
+	/* Support only LOCAL_WRITE flag for DMA MRs */
+	if (acc & ~IB_ACCESS_LOCAL_WRITE) {
+		dev_warn(&dev->pdev->dev,
+			 "unsupported dma mr access flags %#x\n", acc);
+		return ERR_PTR(-EOPNOTSUPP);
+	}
+
+	mr = kzalloc(sizeof(*mr), GFP_KERNEL);
+	if (!mr)
+		return ERR_PTR(-ENOMEM);
+
+	memset(cmd, 0, sizeof(*cmd));
+	cmd->hdr.cmd = PVRDMA_CMD_CREATE_MR;
+	cmd->pd_handle = to_vpd(pd)->pd_handle;
+	cmd->access_flags = acc;
+	cmd->flags = PVRDMA_MR_FLAG_DMA;
+
+	ret = pvrdma_cmd_post(dev, &req, &rsp, PVRDMA_CMD_CREATE_MR_RESP);
+	if (ret < 0) {
+		dev_warn(&dev->pdev->dev,
+			 "could not get DMA mem region, error: %d\n", ret);
+		kfree(mr);
+		return ERR_PTR(ret);
+	}
+
+	mr->mmr.mr_handle = resp->mr_handle;
+	mr->ibmr.lkey = resp->lkey;
+	mr->ibmr.rkey = resp->rkey;
+
+	return &mr->ibmr;
+}
+
+/**
+ * pvrdma_reg_user_mr - register a userspace memory region
+ * @pd: protection domain
+ * @start: starting address
+ * @length: length of region
+ * @virt_addr: I/O virtual address
+ * @access_flags: access flags for memory region
+ * @udata: user data
+ *
+ * @return: ib_mr pointer on success, otherwise returns an errno.
+ */
+struct ib_mr *pvrdma_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
+				 u64 virt_addr, int access_flags,
+				 struct ib_udata *udata)
+{
+	struct pvrdma_dev *dev = to_vdev(pd->device);
+	struct pvrdma_user_mr *mr = NULL;
+	struct ib_umem *umem;
+	union pvrdma_cmd_req req;
+	union pvrdma_cmd_resp rsp;
+	struct pvrdma_cmd_create_mr *cmd = &req.create_mr;
+	struct pvrdma_cmd_create_mr_resp *resp = &rsp.create_mr_resp;
+	int nchunks;
+	int ret;
+	int entry;
+	struct scatterlist *sg;
+
+	if (length == 0 || length > dev->dsr->caps.max_mr_size) {
+		dev_warn(&dev->pdev->dev, "invalid mem region length\n");
+		return ERR_PTR(-EINVAL);
+	}
+
+	umem = ib_umem_get(pd->uobject->context, start,
+			   length, access_flags, 0);
+	if (IS_ERR(umem)) {
+		dev_warn(&dev->pdev->dev,
+			 "could not get umem for mem region\n");
+		return ERR_CAST(umem);
+	}
+
+	nchunks = 0;
+	for_each_sg(umem->sg_head.sgl, sg, umem->nmap, entry)
+		nchunks += sg_dma_len(sg) >> PAGE_SHIFT;
+
+	if (nchunks < 0 || nchunks > PVRDMA_PAGE_DIR_MAX_PAGES) {
+		dev_warn(&dev->pdev->dev, "overflow %d pages in mem region\n",
+			 nchunks);
+		ret = -EINVAL;
+		goto err_umem;
+	}
+
+	mr = kzalloc(sizeof(*mr), GFP_KERNEL);
+	if (!mr) {
+		ret = -ENOMEM;
+		goto err_umem;
+	}
+
+	mr->mmr.iova = virt_addr;
+	mr->mmr.size = length;
+	mr->umem = umem;
+
+	ret = pvrdma_page_dir_init(dev, &mr->pdir, nchunks, false);
+	if (ret) {
+		dev_warn(&dev->pdev->dev,
+			 "could not allocate page directory\n");
+		goto err_umem;
+	}
+
+	ret = pvrdma_page_dir_insert_umem(&mr->pdir, mr->umem, 0);
+	if (ret)
+		goto err_pdir;
+
+	memset(cmd, 0, sizeof(*cmd));
+	cmd->hdr.cmd = PVRDMA_CMD_CREATE_MR;
+	cmd->start = start;
+	cmd->length = length;
+	cmd->pd_handle = to_vpd(pd)->pd_handle;
+	cmd->access_flags = access_flags;
+	cmd->nchunks = nchunks;
+	cmd->pdir_dma = mr->pdir.dir_dma;
+
+	ret = pvrdma_cmd_post(dev, &req, &rsp, PVRDMA_CMD_CREATE_MR_RESP);
+	if (ret < 0) {
+		dev_warn(&dev->pdev->dev,
+			 "could not register mem region, error: %d\n", ret);
+		goto err_pdir;
+	}
+
+	mr->mmr.mr_handle = resp->mr_handle;
+	mr->ibmr.lkey = resp->lkey;
+	mr->ibmr.rkey = resp->rkey;
+
+	return &mr->ibmr;
+
+err_pdir:
+	pvrdma_page_dir_cleanup(dev, &mr->pdir);
+err_umem:
+	ib_umem_release(umem);
+	kfree(mr);
+
+	return ERR_PTR(ret);
+}
+
+/**
+ * pvrdma_alloc_mr - allocate a memory region
+ * @pd: protection domain
+ * @mr_type: type of memory region
+ * @max_num_sg: maximum number of pages
+ *
+ * @return: ib_mr pointer on success, otherwise returns an errno.
+ */
+struct ib_mr *pvrdma_alloc_mr(struct ib_pd *pd, enum ib_mr_type mr_type,
+			      u32 max_num_sg)
+{
+	struct pvrdma_dev *dev = to_vdev(pd->device);
+	struct pvrdma_user_mr *mr;
+	union pvrdma_cmd_req req;
+	union pvrdma_cmd_resp rsp;
+	struct pvrdma_cmd_create_mr *cmd = &req.create_mr;
+	struct pvrdma_cmd_create_mr_resp *resp = &rsp.create_mr_resp;
+	int size = max_num_sg * sizeof(u64);
+	int ret;
+
+	if (mr_type != IB_MR_TYPE_MEM_REG ||
+	    max_num_sg > PVRDMA_MAX_FAST_REG_PAGES)
+		return ERR_PTR(-EINVAL);
+
+	mr = kzalloc(sizeof(*mr), GFP_KERNEL);
+	if (!mr)
+		return ERR_PTR(-ENOMEM);
+
+	mr->pages = kzalloc(size, GFP_KERNEL);
+	if (!mr->pages) {
+		ret = -ENOMEM;
+		goto freemr;
+	}
+
+	ret = pvrdma_page_dir_init(dev, &mr->pdir, max_num_sg, false);
+	if (ret) {
+		dev_warn(&dev->pdev->dev,
+			 "failed to allocate page dir for mr\n");
+		ret = -ENOMEM;
+		goto freepages;
+	}
+
+	memset(cmd, 0, sizeof(*cmd));
+	cmd->hdr.cmd = PVRDMA_CMD_CREATE_MR;
+	cmd->pd_handle = to_vpd(pd)->pd_handle;
+	cmd->access_flags = 0;
+	cmd->flags = PVRDMA_MR_FLAG_FRMR;
+	cmd->nchunks = max_num_sg;
+
+	ret = pvrdma_cmd_post(dev, &req, &rsp, PVRDMA_CMD_CREATE_MR_RESP);
+	if (ret < 0) {
+		dev_warn(&dev->pdev->dev,
+			 "could not create FR mem region, error: %d\n", ret);
+		goto freepdir;
+	}
+
+	mr->max_pages = max_num_sg;
+	mr->mmr.mr_handle = resp->mr_handle;
+	mr->ibmr.lkey = resp->lkey;
+	mr->ibmr.rkey = resp->rkey;
+	mr->page_shift = PAGE_SHIFT;
+	mr->umem = NULL;
+
+	return &mr->ibmr;
+
+freepdir:
+	pvrdma_page_dir_cleanup(dev, &mr->pdir);
+freepages:
+	kfree(mr->pages);
+freemr:
+	kfree(mr);
+	return ERR_PTR(ret);
+}
+
+/**
+ * pvrdma_dereg_mr - deregister a memory region
+ * @ibmr: memory region
+ *
+ * @return: 0 on success.
+ */
+int pvrdma_dereg_mr(struct ib_mr *ibmr)
+{
+	struct pvrdma_user_mr *mr = to_vmr(ibmr);
+	struct pvrdma_dev *dev = to_vdev(ibmr->device);
+	union pvrdma_cmd_req req;
+	struct pvrdma_cmd_destroy_mr *cmd = &req.destroy_mr;
+	int ret;
+
+	memset(cmd, 0, sizeof(*cmd));
+	cmd->hdr.cmd = PVRDMA_CMD_DESTROY_MR;
+	cmd->mr_handle = mr->mmr.mr_handle;
+	ret = pvrdma_cmd_post(dev, &req, NULL, 0);
+	if (ret < 0)
+		dev_warn(&dev->pdev->dev,
+			 "could not deregister mem region, error: %d\n", ret);
+
+	pvrdma_page_dir_cleanup(dev, &mr->pdir);
+	if (mr->umem)
+		ib_umem_release(mr->umem);
+
+	kfree(mr->pages);
+	kfree(mr);
+
+	return 0;
+}
+
+static int pvrdma_set_page(struct ib_mr *ibmr, u64 addr)
+{
+	struct pvrdma_user_mr *mr = to_vmr(ibmr);
+
+	if (mr->npages == mr->max_pages)
+		return -ENOMEM;
+
+	mr->pages[mr->npages++] = addr;
+	return 0;
+}
+
+int pvrdma_map_mr_sg(struct ib_mr *ibmr, struct scatterlist *sg, int sg_nents,
+		     unsigned int *sg_offset)
+{
+	struct pvrdma_user_mr *mr = to_vmr(ibmr);
+	struct pvrdma_dev *dev = to_vdev(ibmr->device);
+	int ret;
+
+	mr->npages = 0;
+
+	ret = ib_sg_to_pages(ibmr, sg, sg_nents, sg_offset, pvrdma_set_page);
+	if (ret < 0)
+		dev_warn(&dev->pdev->dev, "could not map sg to pages\n");
+
+	return ret;
+}
diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_qp.c b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_qp.c
new file mode 100644
index 000000000000..c8c01e558125
--- /dev/null
+++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_qp.c
@@ -0,0 +1,972 @@
+/*
+ * Copyright (c) 2012-2016 VMware, Inc.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of EITHER the GNU General Public License
+ * version 2 as published by the Free Software Foundation or the BSD
+ * 2-Clause License. This program is distributed in the hope that it
+ * will be useful, but WITHOUT ANY WARRANTY; WITHOUT EVEN THE IMPLIED
+ * WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
+ * See the GNU General Public License version 2 for more details at
+ * http://www.gnu.org/licenses/old-licenses/gpl-2.0.en.html.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program available in the file COPYING in the main
+ * directory of this source tree.
+ *
+ * The BSD 2-Clause License
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+ * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
+ * COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT,
+ * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <asm/page.h>
+#include <linux/io.h>
+#include <linux/wait.h>
+#include <rdma/ib_addr.h>
+#include <rdma/ib_smi.h>
+#include <rdma/ib_user_verbs.h>
+
+#include "pvrdma.h"
+
+static inline void get_cqs(struct pvrdma_qp *qp, struct pvrdma_cq **send_cq,
+			   struct pvrdma_cq **recv_cq)
+{
+	*send_cq = to_vcq(qp->ibqp.send_cq);
+	*recv_cq = to_vcq(qp->ibqp.recv_cq);
+}
+
+static void pvrdma_lock_cqs(struct pvrdma_cq *scq, struct pvrdma_cq *rcq,
+			    unsigned long *scq_flags,
+			    unsigned long *rcq_flags)
+	__acquires(scq->cq_lock) __acquires(rcq->cq_lock)
+{
+	if (scq == rcq) {
+		spin_lock_irqsave(&scq->cq_lock, *scq_flags);
+		__acquire(rcq->cq_lock);
+	} else if (scq->cq_handle < rcq->cq_handle) {
+		spin_lock_irqsave(&scq->cq_lock, *scq_flags);
+		spin_lock_irqsave_nested(&rcq->cq_lock, *rcq_flags,
+					 SINGLE_DEPTH_NESTING);
+	} else {
+		spin_lock_irqsave(&rcq->cq_lock, *rcq_flags);
+		spin_lock_irqsave_nested(&scq->cq_lock, *scq_flags,
+					 SINGLE_DEPTH_NESTING);
+	}
+}
+
+static void pvrdma_unlock_cqs(struct pvrdma_cq *scq, struct pvrdma_cq *rcq,
+			      unsigned long *scq_flags,
+			      unsigned long *rcq_flags)
+	__releases(scq->cq_lock) __releases(rcq->cq_lock)
+{
+	if (scq == rcq) {
+		__release(rcq->cq_lock);
+		spin_unlock_irqrestore(&scq->cq_lock, *scq_flags);
+	} else if (scq->cq_handle < rcq->cq_handle) {
+		spin_unlock_irqrestore(&rcq->cq_lock, *rcq_flags);
+		spin_unlock_irqrestore(&scq->cq_lock, *scq_flags);
+	} else {
+		spin_unlock_irqrestore(&scq->cq_lock, *scq_flags);
+		spin_unlock_irqrestore(&rcq->cq_lock, *rcq_flags);
+	}
+}
+
+static void pvrdma_reset_qp(struct pvrdma_qp *qp)
+{
+	struct pvrdma_cq *scq, *rcq;
+	unsigned long scq_flags, rcq_flags;
+
+	/* Clean up cqes */
+	get_cqs(qp, &scq, &rcq);
+	pvrdma_lock_cqs(scq, rcq, &scq_flags, &rcq_flags);
+
+	_pvrdma_flush_cqe(qp, scq);
+	if (scq != rcq)
+		_pvrdma_flush_cqe(qp, rcq);
+
+	pvrdma_unlock_cqs(scq, rcq, &scq_flags, &rcq_flags);
+
+	/*
+	 * Reset queuepair. The checks are because usermode queuepairs won't
+	 * have kernel ringstates.
+	 */
+	if (qp->rq.ring) {
+		atomic_set(&qp->rq.ring->cons_head, 0);
+		atomic_set(&qp->rq.ring->prod_tail, 0);
+	}
+	if (qp->sq.ring) {
+		atomic_set(&qp->sq.ring->cons_head, 0);
+		atomic_set(&qp->sq.ring->prod_tail, 0);
+	}
+}
+
+static int pvrdma_set_rq_size(struct pvrdma_dev *dev,
+			      struct ib_qp_cap *req_cap,
+			      struct pvrdma_qp *qp)
+{
+	if (req_cap->max_recv_wr > dev->dsr->caps.max_qp_wr ||
+	    req_cap->max_recv_sge > dev->dsr->caps.max_sge) {
+		dev_warn(&dev->pdev->dev, "recv queue size invalid\n");
+		return -EINVAL;
+	}
+
+	qp->rq.wqe_cnt = roundup_pow_of_two(max(1U, req_cap->max_recv_wr));
+	qp->rq.max_sg = roundup_pow_of_two(max(1U, req_cap->max_recv_sge));
+
+	/* Write back */
+	req_cap->max_recv_wr = qp->rq.wqe_cnt;
+	req_cap->max_recv_sge = qp->rq.max_sg;
+
+	qp->rq.wqe_size = roundup_pow_of_two(sizeof(struct pvrdma_rq_wqe_hdr) +
+					     sizeof(struct pvrdma_sge) *
+					     qp->rq.max_sg);
+	qp->npages_recv = (qp->rq.wqe_cnt * qp->rq.wqe_size + PAGE_SIZE - 1) /
+			  PAGE_SIZE;
+
+	return 0;
+}
+
+static int pvrdma_set_sq_size(struct pvrdma_dev *dev, struct ib_qp_cap *req_cap,
+			      enum ib_qp_type type, struct pvrdma_qp *qp)
+{
+	if (req_cap->max_send_wr > dev->dsr->caps.max_qp_wr ||
+	    req_cap->max_send_sge > dev->dsr->caps.max_sge) {
+		dev_warn(&dev->pdev->dev, "send queue size invalid\n");
+		return -EINVAL;
+	}
+
+	qp->sq.wqe_cnt = roundup_pow_of_two(max(1U, req_cap->max_send_wr));
+	qp->sq.max_sg = roundup_pow_of_two(max(1U, req_cap->max_send_sge));
+
+	/* Write back */
+	req_cap->max_send_wr = qp->sq.wqe_cnt;
+	req_cap->max_send_sge = qp->sq.max_sg;
+
+	qp->sq.wqe_size = roundup_pow_of_two(sizeof(struct pvrdma_sq_wqe_hdr) +
+					     sizeof(struct pvrdma_sge) *
+					     qp->sq.max_sg);
+	/* Note: one extra page for the header. */
+	qp->npages_send = 1 + (qp->sq.wqe_cnt * qp->sq.wqe_size +
+			       PAGE_SIZE - 1) / PAGE_SIZE;
+
+	return 0;
+}
+
+/**
+ * pvrdma_create_qp - create queue pair
+ * @pd: protection domain
+ * @init_attr: queue pair attributes
+ * @udata: user data
+ *
+ * @return: the ib_qp pointer on success, otherwise returns an errno.
+ */
+struct ib_qp *pvrdma_create_qp(struct ib_pd *pd,
+			       struct ib_qp_init_attr *init_attr,
+			       struct ib_udata *udata)
+{
+	struct pvrdma_qp *qp = NULL;
+	struct pvrdma_dev *dev = to_vdev(pd->device);
+	union pvrdma_cmd_req req;
+	union pvrdma_cmd_resp rsp;
+	struct pvrdma_cmd_create_qp *cmd = &req.create_qp;
+	struct pvrdma_cmd_create_qp_resp *resp = &rsp.create_qp_resp;
+	struct pvrdma_create_qp ucmd;
+	unsigned long flags;
+	int ret;
+
+	if (init_attr->create_flags) {
+		dev_warn(&dev->pdev->dev,
+			 "invalid create queuepair flags %#x\n",
+			 init_attr->create_flags);
+		return ERR_PTR(-EINVAL);
+	}
+
+	if (init_attr->qp_type != IB_QPT_RC &&
+	    init_attr->qp_type != IB_QPT_UD &&
+	    init_attr->qp_type != IB_QPT_GSI) {
+		dev_warn(&dev->pdev->dev, "queuepair type %d not supported\n",
+			 init_attr->qp_type);
+		return ERR_PTR(-EINVAL);
+	}
+
+	if (!atomic_add_unless(&dev->num_qps, 1, dev->dsr->caps.max_qp))
+		return ERR_PTR(-ENOMEM);
+
+	switch (init_attr->qp_type) {
+	case IB_QPT_GSI:
+		if (init_attr->port_num == 0 ||
+		    init_attr->port_num > pd->device->phys_port_cnt ||
+		    udata) {
+			dev_warn(&dev->pdev->dev, "invalid queuepair attrs\n");
+			ret = -EINVAL;
+			goto err_qp;
+		}
+		/* fall through */
+	case IB_QPT_RC:
+	case IB_QPT_UD:
+		qp = kzalloc(sizeof(*qp), GFP_KERNEL);
+		if (!qp) {
+			ret = -ENOMEM;
+			goto err_qp;
+		}
+
+		spin_lock_init(&qp->sq.lock);
+		spin_lock_init(&qp->rq.lock);
+		mutex_init(&qp->mutex);
+		atomic_set(&qp->refcnt, 1);
+		init_waitqueue_head(&qp->wait);
+
+		qp->state = IB_QPS_RESET;
+
+		if (pd->uobject && udata) {
+			dev_dbg(&dev->pdev->dev,
+				"create queuepair from user space\n");
+
+			if (ib_copy_from_udata(&ucmd, udata, sizeof(ucmd))) {
+				ret = -EFAULT;
+				goto err_qp;
+			}
+
+			/* set qp->sq.wqe_cnt, shift, buf_size.. */
+			qp->rumem = ib_umem_get(pd->uobject->context,
+						ucmd.rbuf_addr,
+						ucmd.rbuf_size, 0, 0);
+			if (IS_ERR(qp->rumem)) {
+				ret = PTR_ERR(qp->rumem);
+				goto err_qp;
+			}
+
+			qp->sumem = ib_umem_get(pd->uobject->context,
+						ucmd.sbuf_addr,
+						ucmd.sbuf_size, 0, 0);
+			if (IS_ERR(qp->sumem)) {
+				ib_umem_release(qp->rumem);
+				ret = PTR_ERR(qp->sumem);
+				goto err_qp;
+			}
+
+			qp->npages_send = ib_umem_page_count(qp->sumem);
+			qp->npages_recv = ib_umem_page_count(qp->rumem);
+			qp->npages = qp->npages_send + qp->npages_recv;
+		} else {
+			qp->is_kernel = true;
+
+			ret = pvrdma_set_sq_size(to_vdev(pd->device),
+						 &init_attr->cap,
+						 init_attr->qp_type, qp);
+			if (ret)
+				goto err_qp;
+
+			ret = pvrdma_set_rq_size(to_vdev(pd->device),
+						 &init_attr->cap, qp);
+			if (ret)
+				goto err_qp;
+
+			qp->npages = qp->npages_send + qp->npages_recv;
+
+			/* Skip header page. */
+			qp->sq.offset = PAGE_SIZE;
+
+			/* Recv queue pages are after send pages. */
+			qp->rq.offset = qp->npages_send * PAGE_SIZE;
+		}
+
+		if (qp->npages < 0 || qp->npages > PVRDMA_PAGE_DIR_MAX_PAGES) {
+			dev_warn(&dev->pdev->dev,
+				 "overflow pages in queuepair\n");
+			ret = -EINVAL;
+			goto err_umem;
+		}
+
+		ret = pvrdma_page_dir_init(dev, &qp->pdir, qp->npages,
+					   qp->is_kernel);
+		if (ret) {
+			dev_warn(&dev->pdev->dev,
+				 "could not allocate page directory\n");
+			goto err_umem;
+		}
+
+		if (!qp->is_kernel) {
+			pvrdma_page_dir_insert_umem(&qp->pdir, qp->sumem, 0);
+			pvrdma_page_dir_insert_umem(&qp->pdir, qp->rumem,
+						    qp->npages_send);
+		} else {
+			/* Ring state is always the first page. */
+			qp->sq.ring = qp->pdir.pages[0];
+			qp->rq.ring = &qp->sq.ring[1];
+		}
+		break;
+	default:
+		ret = -EINVAL;
+		goto err_qp;
+	}
+
+	/* Not supported */
+	init_attr->cap.max_inline_data = 0;
+
+	memset(cmd, 0, sizeof(*cmd));
+	cmd->hdr.cmd = PVRDMA_CMD_CREATE_QP;
+	cmd->pd_handle = to_vpd(pd)->pd_handle;
+	cmd->send_cq_handle = to_vcq(init_attr->send_cq)->cq_handle;
+	cmd->recv_cq_handle = to_vcq(init_attr->recv_cq)->cq_handle;
+	cmd->max_send_wr = init_attr->cap.max_send_wr;
+	cmd->max_recv_wr = init_attr->cap.max_recv_wr;
+	cmd->max_send_sge = init_attr->cap.max_send_sge;
+	cmd->max_recv_sge = init_attr->cap.max_recv_sge;
+	cmd->max_inline_data = init_attr->cap.max_inline_data;
+	cmd->sq_sig_all = (init_attr->sq_sig_type == IB_SIGNAL_ALL_WR) ? 1 : 0;
+	cmd->qp_type = ib_qp_type_to_pvrdma(init_attr->qp_type);
+	cmd->access_flags = IB_ACCESS_LOCAL_WRITE;
+	cmd->total_chunks = qp->npages;
+	cmd->send_chunks = qp->npages_send - 1;
+	cmd->pdir_dma = qp->pdir.dir_dma;
+
+	dev_dbg(&dev->pdev->dev, "create queuepair with %d, %d, %d, %d\n",
+		cmd->max_send_wr, cmd->max_recv_wr, cmd->max_send_sge,
+		cmd->max_recv_sge);
+
+	ret = pvrdma_cmd_post(dev, &req, &rsp, PVRDMA_CMD_CREATE_QP_RESP);
+	if (ret < 0) {
+		dev_warn(&dev->pdev->dev,
+			 "could not create queuepair, error: %d\n", ret);
+		goto err_pdir;
+	}
+
+	/* max_send_wr/_recv_wr/_send_sge/_recv_sge/_inline_data */
+	qp->qp_handle = resp->qpn;
+	qp->port = init_attr->port_num;
+	qp->ibqp.qp_num = resp->qpn;
+	spin_lock_irqsave(&dev->qp_tbl_lock, flags);
+	dev->qp_tbl[qp->qp_handle % dev->dsr->caps.max_qp] = qp;
+	spin_unlock_irqrestore(&dev->qp_tbl_lock, flags);
+
+	return &qp->ibqp;
+
+err_pdir:
+	pvrdma_page_dir_cleanup(dev, &qp->pdir);
+err_umem:
+	if (pd->uobject && udata) {
+		if (qp->rumem)
+			ib_umem_release(qp->rumem);
+		if (qp->sumem)
+			ib_umem_release(qp->sumem);
+	}
+err_qp:
+	kfree(qp);
+	atomic_dec(&dev->num_qps);
+
+	return ERR_PTR(ret);
+}
+
+static void pvrdma_free_qp(struct pvrdma_qp *qp)
+{
+	struct pvrdma_dev *dev = to_vdev(qp->ibqp.device);
+	struct pvrdma_cq *scq;
+	struct pvrdma_cq *rcq;
+	unsigned long flags, scq_flags, rcq_flags;
+
+	/* In case cq is polling */
+	get_cqs(qp, &scq, &rcq);
+	pvrdma_lock_cqs(scq, rcq, &scq_flags, &rcq_flags);
+
+	_pvrdma_flush_cqe(qp, scq);
+	if (scq != rcq)
+		_pvrdma_flush_cqe(qp, rcq);
+
+	spin_lock_irqsave(&dev->qp_tbl_lock, flags);
+	dev->qp_tbl[qp->qp_handle] = NULL;
+	spin_unlock_irqrestore(&dev->qp_tbl_lock, flags);
+
+	pvrdma_unlock_cqs(scq, rcq, &scq_flags, &rcq_flags);
+
+	atomic_dec(&qp->refcnt);
+	wait_event(qp->wait, !atomic_read(&qp->refcnt));
+
+	pvrdma_page_dir_cleanup(dev, &qp->pdir);
+
+	kfree(qp);
+
+	atomic_dec(&dev->num_qps);
+}
+
+/**
+ * pvrdma_destroy_qp - destroy a queue pair
+ * @qp: the queue pair to destroy
+ *
+ * @return: 0 on success.
+ */
+int pvrdma_destroy_qp(struct ib_qp *qp)
+{
+	struct pvrdma_qp *vqp = to_vqp(qp);
+	union pvrdma_cmd_req req;
+	struct pvrdma_cmd_destroy_qp *cmd = &req.destroy_qp;
+	int ret;
+
+	memset(cmd, 0, sizeof(*cmd));
+	cmd->hdr.cmd = PVRDMA_CMD_DESTROY_QP;
+	cmd->qp_handle = vqp->qp_handle;
+
+	ret = pvrdma_cmd_post(to_vdev(qp->device), &req, NULL, 0);
+	if (ret < 0)
+		dev_warn(&to_vdev(qp->device)->pdev->dev,
+			 "destroy queuepair failed, error: %d\n", ret);
+
+	pvrdma_free_qp(vqp);
+
+	return 0;
+}
+
+/**
+ * pvrdma_modify_qp - modify queue pair attributes
+ * @ibqp: the queue pair
+ * @attr: the new queue pair's attributes
+ * @attr_mask: attributes mask
+ * @udata: user data
+ *
+ * @returns 0 on success, otherwise returns an errno.
+ */
+int pvrdma_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
+		     int attr_mask, struct ib_udata *udata)
+{
+	struct pvrdma_dev *dev = to_vdev(ibqp->device);
+	struct pvrdma_qp *qp = to_vqp(ibqp);
+	union pvrdma_cmd_req req;
+	union pvrdma_cmd_resp rsp;
+	struct pvrdma_cmd_modify_qp *cmd = &req.modify_qp;
+	int cur_state, next_state;
+	int ret;
+
+	/* Sanity checking. Should need lock here */
+	mutex_lock(&qp->mutex);
+	cur_state = (attr_mask & IB_QP_CUR_STATE) ? attr->cur_qp_state :
+		qp->state;
+	next_state = (attr_mask & IB_QP_STATE) ? attr->qp_state : cur_state;
+
+	if (!ib_modify_qp_is_ok(cur_state, next_state, ibqp->qp_type,
+				attr_mask, IB_LINK_LAYER_ETHERNET)) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	if (attr_mask & IB_QP_PORT) {
+		if (attr->port_num == 0 ||
+		    attr->port_num > ibqp->device->phys_port_cnt) {
+			ret = -EINVAL;
+			goto out;
+		}
+	}
+
+	if (attr_mask & IB_QP_MIN_RNR_TIMER) {
+		if (attr->min_rnr_timer > 31) {
+			ret = -EINVAL;
+			goto out;
+		}
+	}
+
+	if (attr_mask & IB_QP_PKEY_INDEX) {
+		if (attr->pkey_index >= dev->dsr->caps.max_pkeys) {
+			ret = -EINVAL;
+			goto out;
+		}
+	}
+
+	if (attr_mask & IB_QP_QKEY)
+		qp->qkey = attr->qkey;
+
+	if (cur_state == next_state && cur_state == IB_QPS_RESET) {
+		ret = 0;
+		goto out;
+	}
+
+	qp->state = next_state;
+	memset(cmd, 0, sizeof(*cmd));
+	cmd->hdr.cmd = PVRDMA_CMD_MODIFY_QP;
+	cmd->qp_handle = qp->qp_handle;
+	cmd->attr_mask = ib_qp_attr_mask_to_pvrdma(attr_mask);
+	cmd->attrs.qp_state = ib_qp_state_to_pvrdma(attr->qp_state);
+	cmd->attrs.cur_qp_state =
+		ib_qp_state_to_pvrdma(attr->cur_qp_state);
+	cmd->attrs.path_mtu = ib_mtu_to_pvrdma(attr->path_mtu);
+	cmd->attrs.path_mig_state =
+		ib_mig_state_to_pvrdma(attr->path_mig_state);
+	cmd->attrs.qkey = attr->qkey;
+	cmd->attrs.rq_psn = attr->rq_psn;
+	cmd->attrs.sq_psn = attr->sq_psn;
+	cmd->attrs.dest_qp_num = attr->dest_qp_num;
+	cmd->attrs.qp_access_flags =
+		ib_access_flags_to_pvrdma(attr->qp_access_flags);
+	cmd->attrs.pkey_index = attr->pkey_index;
+	cmd->attrs.alt_pkey_index = attr->alt_pkey_index;
+	cmd->attrs.en_sqd_async_notify = attr->en_sqd_async_notify;
+	cmd->attrs.sq_draining = attr->sq_draining;
+	cmd->attrs.max_rd_atomic = attr->max_rd_atomic;
+	cmd->attrs.max_dest_rd_atomic = attr->max_dest_rd_atomic;
+	cmd->attrs.min_rnr_timer = attr->min_rnr_timer;
+	cmd->attrs.port_num = attr->port_num;
+	cmd->attrs.timeout = attr->timeout;
+	cmd->attrs.retry_cnt = attr->retry_cnt;
+	cmd->attrs.rnr_retry = attr->rnr_retry;
+	cmd->attrs.alt_port_num = attr->alt_port_num;
+	cmd->attrs.alt_timeout = attr->alt_timeout;
+	ib_qp_cap_to_pvrdma(&cmd->attrs.cap, &attr->cap);
+	ib_ah_attr_to_pvrdma(&cmd->attrs.ah_attr, &attr->ah_attr);
+	ib_ah_attr_to_pvrdma(&cmd->attrs.alt_ah_attr, &attr->alt_ah_attr);
+
+	ret = pvrdma_cmd_post(dev, &req, &rsp, PVRDMA_CMD_MODIFY_QP_RESP);
+	if (ret < 0) {
+		dev_warn(&dev->pdev->dev,
+			 "could not modify queuepair, error: %d\n", ret);
+	} else if (rsp.hdr.err > 0) {
+		dev_warn(&dev->pdev->dev,
+			 "cannot modify queuepair, error: %d\n", rsp.hdr.err);
+		ret = -EINVAL;
+	}
+
+	if (ret == 0 && next_state == IB_QPS_RESET)
+		pvrdma_reset_qp(qp);
+
+out:
+	mutex_unlock(&qp->mutex);
+
+	return ret;
+}
+
+static inline void *get_sq_wqe(struct pvrdma_qp *qp, int n)
+{
+	return pvrdma_page_dir_get_ptr(&qp->pdir,
+				       qp->sq.offset + n * qp->sq.wqe_size);
+}
+
+static inline void *get_rq_wqe(struct pvrdma_qp *qp, int n)
+{
+	return pvrdma_page_dir_get_ptr(&qp->pdir,
+				       qp->rq.offset + n * qp->rq.wqe_size);
+}
+
+static int set_reg_seg(struct pvrdma_sq_wqe_hdr *wqe_hdr, struct ib_reg_wr *wr)
+{
+	struct pvrdma_user_mr *mr = to_vmr(wr->mr);
+
+	wqe_hdr->wr.fast_reg.iova_start = mr->ibmr.iova;
+	wqe_hdr->wr.fast_reg.pl_pdir_dma = mr->pdir.dir_dma;
+	wqe_hdr->wr.fast_reg.page_shift = mr->page_shift;
+	wqe_hdr->wr.fast_reg.page_list_len = mr->npages;
+	wqe_hdr->wr.fast_reg.length = mr->ibmr.length;
+	wqe_hdr->wr.fast_reg.access_flags = wr->access;
+	wqe_hdr->wr.fast_reg.rkey = wr->key;
+
+	return pvrdma_page_dir_insert_page_list(&mr->pdir, mr->pages,
+						mr->npages);
+}
+
+/**
+ * pvrdma_post_send - post send work request entries on a QP
+ * @ibqp: the QP
+ * @wr: work request list to post
+ * @bad_wr: the first bad WR returned
+ *
+ * @return: 0 on success, otherwise errno returned.
+ */
+int pvrdma_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
+		     struct ib_send_wr **bad_wr)
+{
+	struct pvrdma_qp *qp = to_vqp(ibqp);
+	struct pvrdma_dev *dev = to_vdev(ibqp->device);
+	unsigned long flags;
+	struct pvrdma_sq_wqe_hdr *wqe_hdr;
+	struct pvrdma_sge *sge;
+	int i, index;
+	int nreq;
+	int ret;
+
+	/*
+	 * In states lower than RTS, we can fail immediately. In other states,
+	 * just post and let the device figure it out.
+	 */
+	if (qp->state < IB_QPS_RTS) {
+		*bad_wr = wr;
+		return -EINVAL;
+	}
+
+	spin_lock_irqsave(&qp->sq.lock, flags);
+
+	index = pvrdma_idx(&qp->sq.ring->prod_tail, qp->sq.wqe_cnt);
+	for (nreq = 0; wr; nreq++, wr = wr->next) {
+		unsigned int tail;
+
+		if (unlikely(!pvrdma_idx_ring_has_space(
+				qp->sq.ring, qp->sq.wqe_cnt, &tail))) {
+			dev_warn_ratelimited(&dev->pdev->dev,
+					     "send queue is full\n");
+			*bad_wr = wr;
+			ret = -ENOMEM;
+			goto out;
+		}
+
+		if (unlikely(wr->num_sge > qp->sq.max_sg || wr->num_sge < 0)) {
+			dev_warn_ratelimited(&dev->pdev->dev,
+					     "send SGE overflow\n");
+			*bad_wr = wr;
+			ret = -EINVAL;
+			goto out;
+		}
+
+		if (unlikely(wr->opcode < 0)) {
+			dev_warn_ratelimited(&dev->pdev->dev,
+					     "invalid send opcode\n");
+			*bad_wr = wr;
+			ret = -EINVAL;
+			goto out;
+		}
+
+		/*
+		 * Only support UD, RC.
+		 * Need to check opcode table for thorough checking.
+		 * opcode		_UD	_UC	_RC
+		 * _SEND		x	x	x
+		 * _SEND_WITH_IMM	x	x	x
+		 * _RDMA_WRITE			x	x
+		 * _RDMA_WRITE_WITH_IMM		x	x
+		 * _LOCAL_INV			x	x
+		 * _SEND_WITH_INV		x	x
+		 * _RDMA_READ				x
+		 * _ATOMIC_CMP_AND_SWP			x
+		 * _ATOMIC_FETCH_AND_ADD		x
+		 * _MASK_ATOMIC_CMP_AND_SWP		x
+		 * _MASK_ATOMIC_FETCH_AND_ADD		x
+		 * _REG_MR				x
+		 *
+		 */
+		if (qp->ibqp.qp_type != IB_QPT_UD &&
+		    qp->ibqp.qp_type != IB_QPT_RC &&
+			wr->opcode != IB_WR_SEND) {
+			dev_warn_ratelimited(&dev->pdev->dev,
+					     "unsupported queuepair type\n");
+			*bad_wr = wr;
+			ret = -EINVAL;
+			goto out;
+		} else if (qp->ibqp.qp_type == IB_QPT_UD ||
+			   qp->ibqp.qp_type == IB_QPT_GSI) {
+			if (wr->opcode != IB_WR_SEND &&
+			    wr->opcode != IB_WR_SEND_WITH_IMM) {
+				dev_warn_ratelimited(&dev->pdev->dev,
+						     "invalid send opcode\n");
+				*bad_wr = wr;
+				ret = -EINVAL;
+				goto out;
+			}
+		}
+
+		wqe_hdr = (struct pvrdma_sq_wqe_hdr *)get_sq_wqe(qp, index);
+		memset(wqe_hdr, 0, sizeof(*wqe_hdr));
+		wqe_hdr->wr_id = wr->wr_id;
+		wqe_hdr->num_sge = wr->num_sge;
+		wqe_hdr->opcode = ib_wr_opcode_to_pvrdma(wr->opcode);
+		wqe_hdr->send_flags = ib_send_flags_to_pvrdma(wr->send_flags);
+		if (wr->opcode == IB_WR_SEND_WITH_IMM ||
+		    wr->opcode == IB_WR_RDMA_WRITE_WITH_IMM)
+			wqe_hdr->ex.imm_data = wr->ex.imm_data;
+
+		switch (qp->ibqp.qp_type) {
+		case IB_QPT_GSI:
+		case IB_QPT_UD:
+			if (unlikely(!ud_wr(wr)->ah)) {
+				dev_warn_ratelimited(&dev->pdev->dev,
+						     "invalid address handle\n");
+				*bad_wr = wr;
+				ret = -EINVAL;
+				goto out;
+			}
+
+			/*
+			 * Use qkey from qp context if high order bit set,
+			 * otherwise from work request.
+			 */
+			wqe_hdr->wr.ud.remote_qpn = ud_wr(wr)->remote_qpn;
+			wqe_hdr->wr.ud.remote_qkey =
+				ud_wr(wr)->remote_qkey & 0x80000000 ?
+				qp->qkey : ud_wr(wr)->remote_qkey;
+			wqe_hdr->wr.ud.av = to_vah(ud_wr(wr)->ah)->av;
+
+			break;
+		case IB_QPT_RC:
+			switch (wr->opcode) {
+			case IB_WR_RDMA_READ:
+			case IB_WR_RDMA_WRITE:
+			case IB_WR_RDMA_WRITE_WITH_IMM:
+				wqe_hdr->wr.rdma.remote_addr =
+					rdma_wr(wr)->remote_addr;
+				wqe_hdr->wr.rdma.rkey = rdma_wr(wr)->rkey;
+				break;
+			case IB_WR_LOCAL_INV:
+			case IB_WR_SEND_WITH_INV:
+				wqe_hdr->ex.invalidate_rkey =
+					wr->ex.invalidate_rkey;
+				break;
+			case IB_WR_ATOMIC_CMP_AND_SWP:
+			case IB_WR_ATOMIC_FETCH_AND_ADD:
+				wqe_hdr->wr.atomic.remote_addr =
+					atomic_wr(wr)->remote_addr;
+				wqe_hdr->wr.atomic.rkey = atomic_wr(wr)->rkey;
+				wqe_hdr->wr.atomic.compare_add =
+					atomic_wr(wr)->compare_add;
+				if (wr->opcode == IB_WR_ATOMIC_CMP_AND_SWP)
+					wqe_hdr->wr.atomic.swap =
+						atomic_wr(wr)->swap;
+				break;
+			case IB_WR_REG_MR:
+				ret = set_reg_seg(wqe_hdr, reg_wr(wr));
+				if (ret < 0) {
+					dev_warn_ratelimited(&dev->pdev->dev,
+							     "Failed to set fast register work request\n");
+					*bad_wr = wr;
+					goto out;
+				}
+				break;
+			default:
+				break;
+			}
+
+			break;
+		default:
+			dev_warn_ratelimited(&dev->pdev->dev,
+					     "invalid queuepair type\n");
+			ret = -EINVAL;
+			*bad_wr = wr;
+			goto out;
+		}
+
+		sge = (struct pvrdma_sge *)(wqe_hdr + 1);
+		for (i = 0; i < wr->num_sge; i++) {
+			/* Need to check wqe_size 0 or max size */
+			sge->addr = wr->sg_list[i].addr;
+			sge->length = wr->sg_list[i].length;
+			sge->lkey = wr->sg_list[i].lkey;
+			sge++;
+		}
+
+		/* Make sure wqe is written before index update */
+		smp_wmb();
+
+		index++;
+		if (unlikely(index >= qp->sq.wqe_cnt))
+			index = 0;
+		/* Update shared sq ring */
+		pvrdma_idx_ring_inc(&qp->sq.ring->prod_tail,
+				    qp->sq.wqe_cnt);
+	}
+
+	ret = 0;
+
+out:
+	spin_unlock_irqrestore(&qp->sq.lock, flags);
+
+	if (!ret)
+		pvrdma_write_uar_qp(dev, PVRDMA_UAR_QP_SEND | qp->qp_handle);
+
+	return ret;
+}
+
+/**
+ * pvrdma_post_receive - post receive work request entries on a QP
+ * @ibqp: the QP
+ * @wr: the work request list to post
+ * @bad_wr: the first bad WR returned
+ *
+ * @return: 0 on success, otherwise errno returned.
+ */
+int pvrdma_post_recv(struct ib_qp *ibqp, struct ib_recv_wr *wr,
+		     struct ib_recv_wr **bad_wr)
+{
+	struct pvrdma_dev *dev = to_vdev(ibqp->device);
+	unsigned long flags;
+	struct pvrdma_qp *qp = to_vqp(ibqp);
+	struct pvrdma_rq_wqe_hdr *wqe_hdr;
+	struct pvrdma_sge *sge;
+	int index, nreq;
+	int ret = 0;
+	int i;
+
+	/*
+	 * In the RESET state, we can fail immediately. For other states,
+	 * just post and let the device figure it out.
+	 */
+	if (qp->state == IB_QPS_RESET) {
+		*bad_wr = wr;
+		return -EINVAL;
+	}
+
+	spin_lock_irqsave(&qp->rq.lock, flags);
+
+	index = pvrdma_idx(&qp->rq.ring->prod_tail, qp->rq.wqe_cnt);
+	for (nreq = 0; wr; nreq++, wr = wr->next) {
+		unsigned int tail;
+
+		if (unlikely(wr->num_sge > qp->rq.max_sg ||
+			     wr->num_sge < 0)) {
+			ret = -EINVAL;
+			*bad_wr = wr;
+			dev_warn_ratelimited(&dev->pdev->dev,
+					     "recv SGE overflow\n");
+			goto out;
+		}
+
+		if (unlikely(!pvrdma_idx_ring_has_space(
+				qp->rq.ring, qp->rq.wqe_cnt, &tail))) {
+			ret = -ENOMEM;
+			*bad_wr = wr;
+			dev_warn_ratelimited(&dev->pdev->dev,
+					     "recv queue full\n");
+			goto out;
+		}
+
+		wqe_hdr = (struct pvrdma_rq_wqe_hdr *)get_rq_wqe(qp, index);
+		wqe_hdr->wr_id = wr->wr_id;
+		wqe_hdr->num_sge = wr->num_sge;
+		wqe_hdr->total_len = 0;
+
+		sge = (struct pvrdma_sge *)(wqe_hdr + 1);
+		for (i = 0; i < wr->num_sge; i++) {
+			sge->addr = wr->sg_list[i].addr;
+			sge->length = wr->sg_list[i].length;
+			sge->lkey = wr->sg_list[i].lkey;
+			sge++;
+		}
+
+		/* Make sure wqe is written before index update */
+		smp_wmb();
+
+		index++;
+		if (unlikely(index >= qp->rq.wqe_cnt))
+			index = 0;
+		/* Update shared rq ring */
+		pvrdma_idx_ring_inc(&qp->rq.ring->prod_tail,
+				    qp->rq.wqe_cnt);
+	}
+
+	spin_unlock_irqrestore(&qp->rq.lock, flags);
+
+	pvrdma_write_uar_qp(dev, PVRDMA_UAR_QP_RECV | qp->qp_handle);
+
+	return ret;
+
+out:
+	spin_unlock_irqrestore(&qp->rq.lock, flags);
+
+	return ret;
+}
+
+/**
+ * pvrdma_query_qp - query a queue pair's attributes
+ * @ibqp: the queue pair to query
+ * @attr: the queue pair's attributes
+ * @attr_mask: attributes mask
+ * @init_attr: initial queue pair attributes
+ *
+ * @returns 0 on success, otherwise returns an errno.
+ */
+int pvrdma_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
+		    int attr_mask, struct ib_qp_init_attr *init_attr)
+{
+	struct pvrdma_dev *dev = to_vdev(ibqp->device);
+	struct pvrdma_qp *qp = to_vqp(ibqp);
+	union pvrdma_cmd_req req;
+	union pvrdma_cmd_resp rsp;
+	struct pvrdma_cmd_query_qp *cmd = &req.query_qp;
+	struct pvrdma_cmd_query_qp_resp *resp = &rsp.query_qp_resp;
+	int ret = 0;
+
+	mutex_lock(&qp->mutex);
+
+	if (qp->state == IB_QPS_RESET) {
+		attr->qp_state = IB_QPS_RESET;
+		goto out;
+	}
+
+	memset(cmd, 0, sizeof(*cmd));
+	cmd->hdr.cmd = PVRDMA_CMD_QUERY_QP;
+	cmd->qp_handle = qp->qp_handle;
+	cmd->attr_mask = ib_qp_attr_mask_to_pvrdma(attr_mask);
+
+	ret = pvrdma_cmd_post(dev, &req, &rsp, PVRDMA_CMD_QUERY_QP_RESP);
+	if (ret < 0) {
+		dev_warn(&dev->pdev->dev,
+			 "could not query queuepair, error: %d\n", ret);
+		goto out;
+	}
+
+	attr->qp_state = pvrdma_qp_state_to_ib(resp->attrs.qp_state);
+	attr->cur_qp_state =
+		pvrdma_qp_state_to_ib(resp->attrs.cur_qp_state);
+	attr->path_mtu = pvrdma_mtu_to_ib(resp->attrs.path_mtu);
+	attr->path_mig_state =
+		pvrdma_mig_state_to_ib(resp->attrs.path_mig_state);
+	attr->qkey = resp->attrs.qkey;
+	attr->rq_psn = resp->attrs.rq_psn;
+	attr->sq_psn = resp->attrs.sq_psn;
+	attr->dest_qp_num = resp->attrs.dest_qp_num;
+	attr->qp_access_flags =
+		pvrdma_access_flags_to_ib(resp->attrs.qp_access_flags);
+	attr->pkey_index = resp->attrs.pkey_index;
+	attr->alt_pkey_index = resp->attrs.alt_pkey_index;
+	attr->en_sqd_async_notify = resp->attrs.en_sqd_async_notify;
+	attr->sq_draining = resp->attrs.sq_draining;
+	attr->max_rd_atomic = resp->attrs.max_rd_atomic;
+	attr->max_dest_rd_atomic = resp->attrs.max_dest_rd_atomic;
+	attr->min_rnr_timer = resp->attrs.min_rnr_timer;
+	attr->port_num = resp->attrs.port_num;
+	attr->timeout = resp->attrs.timeout;
+	attr->retry_cnt = resp->attrs.retry_cnt;
+	attr->rnr_retry = resp->attrs.rnr_retry;
+	attr->alt_port_num = resp->attrs.alt_port_num;
+	attr->alt_timeout = resp->attrs.alt_timeout;
+	pvrdma_qp_cap_to_ib(&attr->cap, &resp->attrs.cap);
+	pvrdma_ah_attr_to_ib(&attr->ah_attr, &resp->attrs.ah_attr);
+	pvrdma_ah_attr_to_ib(&attr->alt_ah_attr, &resp->attrs.alt_ah_attr);
+
+	qp->state = attr->qp_state;
+
+	ret = 0;
+
+out:
+	attr->cur_qp_state = attr->qp_state;
+
+	init_attr->event_handler = qp->ibqp.event_handler;
+	init_attr->qp_context = qp->ibqp.qp_context;
+	init_attr->send_cq = qp->ibqp.send_cq;
+	init_attr->recv_cq = qp->ibqp.recv_cq;
+	init_attr->srq = qp->ibqp.srq;
+	init_attr->xrcd = NULL;
+	init_attr->cap = attr->cap;
+	init_attr->sq_sig_type = 0;
+	init_attr->qp_type = qp->ibqp.qp_type;
+	init_attr->create_flags = 0;
+	init_attr->port_num = qp->port;
+
+	mutex_unlock(&qp->mutex);
+	return ret;
+}
diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_ring.h b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_ring.h
new file mode 100644
index 000000000000..ed9022a91a1d
--- /dev/null
+++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_ring.h
@@ -0,0 +1,131 @@
+/*
+ * Copyright (c) 2012-2016 VMware, Inc.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of EITHER the GNU General Public License
+ * version 2 as published by the Free Software Foundation or the BSD
+ * 2-Clause License. This program is distributed in the hope that it
+ * will be useful, but WITHOUT ANY WARRANTY; WITHOUT EVEN THE IMPLIED
+ * WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
+ * See the GNU General Public License version 2 for more details at
+ * http://www.gnu.org/licenses/old-licenses/gpl-2.0.en.html.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program available in the file COPYING in the main
+ * directory of this source tree.
+ *
+ * The BSD 2-Clause License
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+ * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
+ * COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT,
+ * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#ifndef __PVRDMA_RING_H__
+#define __PVRDMA_RING_H__
+
+#include <linux/types.h>
+
+#define PVRDMA_INVALID_IDX	-1	/* Invalid index. */
+
+struct pvrdma_ring {
+	atomic_t prod_tail;	/* Producer tail. */
+	atomic_t cons_head;	/* Consumer head. */
+};
+
+struct pvrdma_ring_state {
+	struct pvrdma_ring tx;	/* Tx ring. */
+	struct pvrdma_ring rx;	/* Rx ring. */
+};
+
+static inline int pvrdma_idx_valid(__u32 idx, __u32 max_elems)
+{
+	/* Generates fewer instructions than a less-than. */
+	return (idx & ~((max_elems << 1) - 1)) == 0;
+}
+
+static inline __s32 pvrdma_idx(atomic_t *var, __u32 max_elems)
+{
+	const unsigned int idx = atomic_read(var);
+
+	if (pvrdma_idx_valid(idx, max_elems))
+		return idx & (max_elems - 1);
+	return PVRDMA_INVALID_IDX;
+}
+
+static inline void pvrdma_idx_ring_inc(atomic_t *var, __u32 max_elems)
+{
+	__u32 idx = atomic_read(var) + 1;	/* Increment. */
+
+	idx &= (max_elems << 1) - 1;		/* Modulo size, flip gen. */
+	atomic_set(var, idx);
+}
+
+static inline __s32 pvrdma_idx_ring_has_space(const struct pvrdma_ring *r,
+					      __u32 max_elems, __u32 *out_tail)
+{
+	const __u32 tail = atomic_read(&r->prod_tail);
+	const __u32 head = atomic_read(&r->cons_head);
+
+	if (pvrdma_idx_valid(tail, max_elems) &&
+	    pvrdma_idx_valid(head, max_elems)) {
+		*out_tail = tail & (max_elems - 1);
+		return tail != (head ^ max_elems);
+	}
+	return PVRDMA_INVALID_IDX;
+}
+
+static inline __s32 pvrdma_idx_ring_has_data(const struct pvrdma_ring *r,
+					     __u32 max_elems, __u32 *out_head)
+{
+	const __u32 tail = atomic_read(&r->prod_tail);
+	const __u32 head = atomic_read(&r->cons_head);
+
+	if (pvrdma_idx_valid(tail, max_elems) &&
+	    pvrdma_idx_valid(head, max_elems)) {
+		*out_head = head & (max_elems - 1);
+		return tail != head;
+	}
+	return PVRDMA_INVALID_IDX;
+}
+
+static inline bool pvrdma_idx_ring_is_valid_idx(const struct pvrdma_ring *r,
+						__u32 max_elems, __u32 *idx)
+{
+	const __u32 tail = atomic_read(&r->prod_tail);
+	const __u32 head = atomic_read(&r->cons_head);
+
+	if (pvrdma_idx_valid(tail, max_elems) &&
+	    pvrdma_idx_valid(head, max_elems) &&
+	    pvrdma_idx_valid(*idx, max_elems)) {
+		if (tail > head && (*idx < tail && *idx >= head))
+			return true;
+		else if (head > tail && (*idx >= head || *idx < tail))
+			return true;
+	}
+	return false;
+}
+
+#endif /* __PVRDMA_RING_H__ */
diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.c b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.c
new file mode 100644
index 000000000000..54891370d18a
--- /dev/null
+++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.c
@@ -0,0 +1,579 @@
+/*
+ * Copyright (c) 2012-2016 VMware, Inc.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of EITHER the GNU General Public License
+ * version 2 as published by the Free Software Foundation or the BSD
+ * 2-Clause License. This program is distributed in the hope that it
+ * will be useful, but WITHOUT ANY WARRANTY; WITHOUT EVEN THE IMPLIED
+ * WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
+ * See the GNU General Public License version 2 for more details at
+ * http://www.gnu.org/licenses/old-licenses/gpl-2.0.en.html.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program available in the file COPYING in the main
+ * directory of this source tree.
+ *
+ * The BSD 2-Clause License
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+ * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
+ * COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT,
+ * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <asm/page.h>
+#include <linux/inet.h>
+#include <linux/io.h>
+#include <rdma/ib_addr.h>
+#include <rdma/ib_smi.h>
+#include <rdma/ib_user_verbs.h>
+#include <rdma/vmw_pvrdma-abi.h>
+
+#include "pvrdma.h"
+
+/**
+ * pvrdma_query_device - query device
+ * @ibdev: the device to query
+ * @props: the device properties
+ * @uhw: user data
+ *
+ * @return: 0 on success, otherwise negative errno
+ */
+int pvrdma_query_device(struct ib_device *ibdev,
+			struct ib_device_attr *props,
+			struct ib_udata *uhw)
+{
+	struct pvrdma_dev *dev = to_vdev(ibdev);
+
+	if (uhw->inlen || uhw->outlen)
+		return -EINVAL;
+
+	memset(props, 0, sizeof(*props));
+
+	props->fw_ver = dev->dsr->caps.fw_ver;
+	props->sys_image_guid = dev->dsr->caps.sys_image_guid;
+	props->max_mr_size = dev->dsr->caps.max_mr_size;
+	props->page_size_cap = dev->dsr->caps.page_size_cap;
+	props->vendor_id = dev->dsr->caps.vendor_id;
+	props->vendor_part_id = dev->pdev->device;
+	props->hw_ver = dev->dsr->caps.hw_ver;
+	props->max_qp = dev->dsr->caps.max_qp;
+	props->max_qp_wr = dev->dsr->caps.max_qp_wr;
+	props->device_cap_flags = dev->dsr->caps.device_cap_flags;
+	props->max_sge = dev->dsr->caps.max_sge;
+	props->max_cq = dev->dsr->caps.max_cq;
+	props->max_cqe = dev->dsr->caps.max_cqe;
+	props->max_mr = dev->dsr->caps.max_mr;
+	props->max_pd = dev->dsr->caps.max_pd;
+	props->max_qp_rd_atom = dev->dsr->caps.max_qp_rd_atom;
+	props->max_qp_init_rd_atom = dev->dsr->caps.max_qp_init_rd_atom;
+	props->atomic_cap =
+		dev->dsr->caps.atomic_ops &
+		(PVRDMA_ATOMIC_OP_COMP_SWAP | PVRDMA_ATOMIC_OP_FETCH_ADD) ?
+		IB_ATOMIC_HCA : IB_ATOMIC_NONE;
+	props->masked_atomic_cap = props->atomic_cap;
+	props->max_ah = dev->dsr->caps.max_ah;
+	props->max_pkeys = dev->dsr->caps.max_pkeys;
+	props->local_ca_ack_delay = dev->dsr->caps.local_ca_ack_delay;
+	if ((dev->dsr->caps.bmme_flags & PVRDMA_BMME_FLAG_LOCAL_INV) &&
+	    (dev->dsr->caps.bmme_flags & PVRDMA_BMME_FLAG_REMOTE_INV) &&
+	    (dev->dsr->caps.bmme_flags & PVRDMA_BMME_FLAG_FAST_REG_WR)) {
+		props->device_cap_flags |= IB_DEVICE_MEM_MGT_EXTENSIONS;
+	}
+
+	return 0;
+}
+
+/**
+ * pvrdma_query_port - query device port attributes
+ * @ibdev: the device to query
+ * @port: the port number
+ * @props: the device properties
+ *
+ * @return: 0 on success, otherwise negative errno
+ */
+int pvrdma_query_port(struct ib_device *ibdev, u8 port,
+		      struct ib_port_attr *props)
+{
+	struct pvrdma_dev *dev = to_vdev(ibdev);
+	union pvrdma_cmd_req req;
+	union pvrdma_cmd_resp rsp;
+	struct pvrdma_cmd_query_port *cmd = &req.query_port;
+	struct pvrdma_cmd_query_port_resp *resp = &rsp.query_port_resp;
+	int err;
+
+	memset(cmd, 0, sizeof(*cmd));
+	cmd->hdr.cmd = PVRDMA_CMD_QUERY_PORT;
+	cmd->port_num = port;
+
+	err = pvrdma_cmd_post(dev, &req, &rsp, PVRDMA_CMD_QUERY_PORT_RESP);
+	if (err < 0) {
+		dev_warn(&dev->pdev->dev,
+			 "could not query port, error: %d\n", err);
+		return err;
+	}
+
+	memset(props, 0, sizeof(*props));
+
+	props->state = pvrdma_port_state_to_ib(resp->attrs.state);
+	props->max_mtu = pvrdma_mtu_to_ib(resp->attrs.max_mtu);
+	props->active_mtu = pvrdma_mtu_to_ib(resp->attrs.active_mtu);
+	props->gid_tbl_len = resp->attrs.gid_tbl_len;
+	props->port_cap_flags =
+		pvrdma_port_cap_flags_to_ib(resp->attrs.port_cap_flags);
+	props->max_msg_sz = resp->attrs.max_msg_sz;
+	props->bad_pkey_cntr = resp->attrs.bad_pkey_cntr;
+	props->qkey_viol_cntr = resp->attrs.qkey_viol_cntr;
+	props->pkey_tbl_len = resp->attrs.pkey_tbl_len;
+	props->lid = resp->attrs.lid;
+	props->sm_lid = resp->attrs.sm_lid;
+	props->lmc = resp->attrs.lmc;
+	props->max_vl_num = resp->attrs.max_vl_num;
+	props->sm_sl = resp->attrs.sm_sl;
+	props->subnet_timeout = resp->attrs.subnet_timeout;
+	props->init_type_reply = resp->attrs.init_type_reply;
+	props->active_width = pvrdma_port_width_to_ib(resp->attrs.active_width);
+	props->active_speed = pvrdma_port_speed_to_ib(resp->attrs.active_speed);
+	props->phys_state = resp->attrs.phys_state;
+
+	return 0;
+}
+
+/**
+ * pvrdma_query_gid - query device gid
+ * @ibdev: the device to query
+ * @port: the port number
+ * @index: the index
+ * @gid: the device gid value
+ *
+ * @return: 0 on success, otherwise negative errno
+ */
+int pvrdma_query_gid(struct ib_device *ibdev, u8 port, int index,
+		     union ib_gid *gid)
+{
+	struct pvrdma_dev *dev = to_vdev(ibdev);
+
+	if (index >= dev->dsr->caps.gid_tbl_len)
+		return -EINVAL;
+
+	memcpy(gid, &dev->sgid_tbl[index], sizeof(union ib_gid));
+
+	return 0;
+}
+
+/**
+ * pvrdma_query_pkey - query device port's P_Key table
+ * @ibdev: the device to query
+ * @port: the port number
+ * @index: the index
+ * @pkey: the device P_Key value
+ *
+ * @return: 0 on success, otherwise negative errno
+ */
+int pvrdma_query_pkey(struct ib_device *ibdev, u8 port, u16 index,
+		      u16 *pkey)
+{
+	int err = 0;
+	union pvrdma_cmd_req req;
+	union pvrdma_cmd_resp rsp;
+	struct pvrdma_cmd_query_pkey *cmd = &req.query_pkey;
+
+	memset(cmd, 0, sizeof(*cmd));
+	cmd->hdr.cmd = PVRDMA_CMD_QUERY_PKEY;
+	cmd->port_num = port;
+	cmd->index = index;
+
+	err = pvrdma_cmd_post(to_vdev(ibdev), &req, &rsp,
+			      PVRDMA_CMD_QUERY_PKEY_RESP);
+	if (err < 0) {
+		dev_warn(&to_vdev(ibdev)->pdev->dev,
+			 "could not query pkey, error: %d\n", err);
+		return err;
+	}
+
+	*pkey = rsp.query_pkey_resp.pkey;
+
+	return 0;
+}
+
+enum rdma_link_layer pvrdma_port_link_layer(struct ib_device *ibdev,
+					    u8 port)
+{
+	return IB_LINK_LAYER_ETHERNET;
+}
+
+int pvrdma_modify_device(struct ib_device *ibdev, int mask,
+			 struct ib_device_modify *props)
+{
+	unsigned long flags;
+
+	if (mask & ~(IB_DEVICE_MODIFY_SYS_IMAGE_GUID |
+		     IB_DEVICE_MODIFY_NODE_DESC)) {
+		dev_warn(&to_vdev(ibdev)->pdev->dev,
+			 "unsupported device modify mask %#x\n", mask);
+		return -EOPNOTSUPP;
+	}
+
+	if (mask & IB_DEVICE_MODIFY_NODE_DESC) {
+		spin_lock_irqsave(&to_vdev(ibdev)->desc_lock, flags);
+		memcpy(ibdev->node_desc, props->node_desc, 64);
+		spin_unlock_irqrestore(&to_vdev(ibdev)->desc_lock, flags);
+	}
+
+	if (mask & IB_DEVICE_MODIFY_SYS_IMAGE_GUID) {
+		mutex_lock(&to_vdev(ibdev)->port_mutex);
+		to_vdev(ibdev)->sys_image_guid =
+			cpu_to_be64(props->sys_image_guid);
+		mutex_unlock(&to_vdev(ibdev)->port_mutex);
+	}
+
+	return 0;
+}
+
+/**
+ * pvrdma_modify_port - modify device port attributes
+ * @ibdev: the device to modify
+ * @port: the port number
+ * @mask: attributes to modify
+ * @props: the device properties
+ *
+ * @return: 0 on success, otherwise negative errno
+ */
+int pvrdma_modify_port(struct ib_device *ibdev, u8 port, int mask,
+		       struct ib_port_modify *props)
+{
+	struct ib_port_attr attr;
+	struct pvrdma_dev *vdev = to_vdev(ibdev);
+	int ret;
+
+	if (mask & ~IB_PORT_SHUTDOWN) {
+		dev_warn(&vdev->pdev->dev,
+			 "unsupported port modify mask %#x\n", mask);
+		return -EOPNOTSUPP;
+	}
+
+	mutex_lock(&vdev->port_mutex);
+	ret = pvrdma_query_port(ibdev, port, &attr);
+	if (ret)
+		goto out;
+
+	vdev->port_cap_mask |= props->set_port_cap_mask;
+	vdev->port_cap_mask &= ~props->clr_port_cap_mask;
+
+	if (mask & IB_PORT_SHUTDOWN)
+		vdev->ib_active = false;
+
+out:
+	mutex_unlock(&vdev->port_mutex);
+	return ret;
+}
+
+/**
+ * pvrdma_alloc_ucontext - allocate ucontext
+ * @ibdev: the IB device
+ * @udata: user data
+ *
+ * @return: the ib_ucontext pointer on success, otherwise errno.
+ */
+struct ib_ucontext *pvrdma_alloc_ucontext(struct ib_device *ibdev,
+					  struct ib_udata *udata)
+{
+	struct pvrdma_dev *vdev = to_vdev(ibdev);
+	struct pvrdma_ucontext *context;
+	union pvrdma_cmd_req req;
+	union pvrdma_cmd_resp rsp;
+	struct pvrdma_cmd_create_uc *cmd = &req.create_uc;
+	struct pvrdma_cmd_create_uc_resp *resp = &rsp.create_uc_resp;
+	struct pvrdma_alloc_ucontext_resp uresp;
+	int ret;
+	void *ptr;
+
+	if (!vdev->ib_active)
+		return ERR_PTR(-EAGAIN);
+
+	context = kmalloc(sizeof(*context), GFP_KERNEL);
+	if (!context)
+		return ERR_PTR(-ENOMEM);
+
+	context->dev = vdev;
+	ret = pvrdma_uar_alloc(vdev, &context->uar);
+	if (ret) {
+		kfree(context);
+		return ERR_PTR(-ENOMEM);
+	}
+
+	/* get ctx_handle from host */
+	memset(cmd, 0, sizeof(*cmd));
+	cmd->pfn = context->uar.pfn;
+	cmd->hdr.cmd = PVRDMA_CMD_CREATE_UC;
+	ret = pvrdma_cmd_post(vdev, &req, &rsp, PVRDMA_CMD_CREATE_UC_RESP);
+	if (ret < 0) {
+		dev_warn(&vdev->pdev->dev,
+			 "could not create ucontext, error: %d\n", ret);
+		ptr = ERR_PTR(ret);
+		goto err;
+	}
+
+	context->ctx_handle = resp->ctx_handle;
+
+	/* copy back to user */
+	uresp.qp_tab_size = vdev->dsr->caps.max_qp;
+	ret = ib_copy_to_udata(udata, &uresp, sizeof(uresp));
+	if (ret) {
+		pvrdma_uar_free(vdev, &context->uar);
+		context->ibucontext.device = ibdev;
+		pvrdma_dealloc_ucontext(&context->ibucontext);
+		return ERR_PTR(-EFAULT);
+	}
+
+	return &context->ibucontext;
+
+err:
+	pvrdma_uar_free(vdev, &context->uar);
+	kfree(context);
+	return ptr;
+}
+
+/**
+ * pvrdma_dealloc_ucontext - deallocate ucontext
+ * @ibcontext: the ucontext
+ *
+ * @return: 0 on success, otherwise errno.
+ */
+int pvrdma_dealloc_ucontext(struct ib_ucontext *ibcontext)
+{
+	struct pvrdma_ucontext *context = to_vucontext(ibcontext);
+	union pvrdma_cmd_req req;
+	struct pvrdma_cmd_destroy_uc *cmd = &req.destroy_uc;
+	int ret;
+
+	memset(cmd, 0, sizeof(*cmd));
+	cmd->hdr.cmd = PVRDMA_CMD_DESTROY_UC;
+	cmd->ctx_handle = context->ctx_handle;
+
+	ret = pvrdma_cmd_post(context->dev, &req, NULL, 0);
+	if (ret < 0)
+		dev_warn(&context->dev->pdev->dev,
+			 "destroy ucontext failed, error: %d\n", ret);
+
+	/* Free the UAR even if the device command failed */
+	pvrdma_uar_free(to_vdev(ibcontext->device), &context->uar);
+	kfree(context);
+
+	return ret;
+}
+
+/**
+ * pvrdma_mmap - create mmap region
+ * @ibcontext: the user context
+ * @vma: the VMA
+ *
+ * @return: 0 on success, otherwise errno.
+ */
+int pvrdma_mmap(struct ib_ucontext *ibcontext, struct vm_area_struct *vma)
+{
+	struct pvrdma_ucontext *context = to_vucontext(ibcontext);
+	unsigned long start = vma->vm_start;
+	unsigned long size = vma->vm_end - vma->vm_start;
+	unsigned long offset = vma->vm_pgoff << PAGE_SHIFT;
+
+	dev_dbg(&context->dev->pdev->dev, "create mmap region\n");
+
+	if ((size != PAGE_SIZE) || (offset & ~PAGE_MASK)) {
+		dev_warn(&context->dev->pdev->dev,
+			 "invalid params for mmap region\n");
+		return -EINVAL;
+	}
+
+	/* Map UAR to kernel space, VM_LOCKED? */
+	vma->vm_flags |= VM_DONTCOPY | VM_DONTEXPAND;
+	vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
+	if (io_remap_pfn_range(vma, start, context->uar.pfn, size,
+			       vma->vm_page_prot))
+		return -EAGAIN;
+
+	return 0;
+}
+
+/**
+ * pvrdma_alloc_pd - allocate protection domain
+ * @ibdev: the IB device
+ * @context: user context
+ * @udata: user data
+ *
+ * @return: the ib_pd protection domain pointer on success, otherwise errno.
+ */
+struct ib_pd *pvrdma_alloc_pd(struct ib_device *ibdev,
+			      struct ib_ucontext *context,
+			      struct ib_udata *udata)
+{
+	struct pvrdma_pd *pd;
+	struct pvrdma_dev *dev = to_vdev(ibdev);
+	union pvrdma_cmd_req req;
+	union pvrdma_cmd_resp rsp;
+	struct pvrdma_cmd_create_pd *cmd = &req.create_pd;
+	struct pvrdma_cmd_create_pd_resp *resp = &rsp.create_pd_resp;
+	int ret;
+	void *ptr;
+
+	/* Check allowed max pds */
+	if (!atomic_add_unless(&dev->num_pds, 1, dev->dsr->caps.max_pd))
+		return ERR_PTR(-ENOMEM);
+
+	pd = kmalloc(sizeof(*pd), GFP_KERNEL);
+	if (!pd) {
+		ptr = ERR_PTR(-ENOMEM);
+		goto err;
+	}
+
+	memset(cmd, 0, sizeof(*cmd));
+	cmd->hdr.cmd = PVRDMA_CMD_CREATE_PD;
+	cmd->ctx_handle = (context) ? to_vucontext(context)->ctx_handle : 0;
+	ret = pvrdma_cmd_post(dev, &req, &rsp, PVRDMA_CMD_CREATE_PD_RESP);
+	if (ret < 0) {
+		dev_warn(&dev->pdev->dev,
+			 "failed to allocate protection domain, error: %d\n",
+			 ret);
+		ptr = ERR_PTR(ret);
+		goto freepd;
+	}
+
+	pd->privileged = !context;
+	pd->pd_handle = resp->pd_handle;
+	pd->pdn = resp->pd_handle;
+
+	if (context) {
+		if (ib_copy_to_udata(udata, &pd->pdn, sizeof(__u32))) {
+			dev_warn(&dev->pdev->dev,
+				 "failed to copy back protection domain\n");
+			pvrdma_dealloc_pd(&pd->ibpd);
+			return ERR_PTR(-EFAULT);
+		}
+	}
+
+	/* u32 pd handle */
+	return &pd->ibpd;
+
+freepd:
+	kfree(pd);
+err:
+	atomic_dec(&dev->num_pds);
+	return ptr;
+}
+
+/**
+ * pvrdma_dealloc_pd - deallocate protection domain
+ * @pd: the protection domain to be released
+ *
+ * @return: 0 on success, otherwise errno.
+ */
+int pvrdma_dealloc_pd(struct ib_pd *pd)
+{
+	struct pvrdma_dev *dev = to_vdev(pd->device);
+	union pvrdma_cmd_req req;
+	struct pvrdma_cmd_destroy_pd *cmd = &req.destroy_pd;
+	int ret;
+
+	memset(cmd, 0, sizeof(*cmd));
+	cmd->hdr.cmd = PVRDMA_CMD_DESTROY_PD;
+	cmd->pd_handle = to_vpd(pd)->pd_handle;
+
+	ret = pvrdma_cmd_post(dev, &req, NULL, 0);
+	if (ret)
+		dev_warn(&dev->pdev->dev,
+			 "could not dealloc protection domain, error: %d\n",
+			 ret);
+
+	kfree(to_vpd(pd));
+	atomic_dec(&dev->num_pds);
+
+	return 0;
+}
+
+/**
+ * pvrdma_create_ah - create an address handle
+ * @pd: the protection domain
+ * @ah_attr: the attributes of the AH
+ * @udata: user data blob
+ *
+ * @return: the ib_ah pointer on success, otherwise errno.
+ */
+struct ib_ah *pvrdma_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr,
+			       struct ib_udata *udata)
+{
+	struct pvrdma_dev *dev = to_vdev(pd->device);
+	struct pvrdma_ah *ah;
+	enum rdma_link_layer ll;
+
+	if (!(ah_attr->ah_flags & IB_AH_GRH))
+		return ERR_PTR(-EINVAL);
+
+	ll = rdma_port_get_link_layer(pd->device, ah_attr->port_num);
+
+	if (ll != IB_LINK_LAYER_ETHERNET ||
+	    rdma_is_multicast_addr((struct in6_addr *)ah_attr->grh.dgid.raw))
+		return ERR_PTR(-EINVAL);
+
+	if (!atomic_add_unless(&dev->num_ahs, 1, dev->dsr->caps.max_ah))
+		return ERR_PTR(-ENOMEM);
+
+	ah = kzalloc(sizeof(*ah), GFP_KERNEL);
+	if (!ah) {
+		atomic_dec(&dev->num_ahs);
+		return ERR_PTR(-ENOMEM);
+	}
+
+	ah->av.port_pd = to_vpd(pd)->pd_handle | (ah_attr->port_num << 24);
+	ah->av.src_path_bits = ah_attr->src_path_bits;
+	ah->av.src_path_bits |= 0x80;
+	ah->av.gid_index = ah_attr->grh.sgid_index;
+	ah->av.hop_limit = ah_attr->grh.hop_limit;
+	ah->av.sl_tclass_flowlabel = (ah_attr->grh.traffic_class << 20) |
+				      ah_attr->grh.flow_label;
+	memcpy(ah->av.dgid, ah_attr->grh.dgid.raw, 16);
+	memcpy(ah->av.dmac, ah_attr->dmac, 6);
+
+	ah->ibah.device = pd->device;
+	ah->ibah.pd = pd;
+	ah->ibah.uobject = NULL;
+
+	return &ah->ibah;
+}
+
+/**
+ * pvrdma_destroy_ah - destroy an address handle
+ * @ah: the address handle to destroyed
+ *
+ * @return: 0 on success.
+ */
+int pvrdma_destroy_ah(struct ib_ah *ah)
+{
+	struct pvrdma_dev *dev = to_vdev(ah->device);
+
+	kfree(to_vah(ah));
+	atomic_dec(&dev->num_ahs);
+
+	return 0;
+}
diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.h b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.h
new file mode 100644
index 000000000000..bfbe96b56255
--- /dev/null
+++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.h
@@ -0,0 +1,436 @@
+/*
+ * Copyright (c) 2012-2016 VMware, Inc.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of EITHER the GNU General Public License
+ * version 2 as published by the Free Software Foundation or the BSD
+ * 2-Clause License. This program is distributed in the hope that it
+ * will be useful, but WITHOUT ANY WARRANTY; WITHOUT EVEN THE IMPLIED
+ * WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
+ * See the GNU General Public License version 2 for more details at
+ * http://www.gnu.org/licenses/old-licenses/gpl-2.0.en.html.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program available in the file COPYING in the main
+ * directory of this source tree.
+ *
+ * The BSD 2-Clause License
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+ * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
+ * COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT,
+ * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#ifndef __PVRDMA_VERBS_H__
+#define __PVRDMA_VERBS_H__
+
+#include <linux/types.h>
+
+union pvrdma_gid {
+	u8	raw[16];
+	struct {
+		__be64	subnet_prefix;
+		__be64	interface_id;
+	} global;
+};
+
+enum pvrdma_link_layer {
+	PVRDMA_LINK_LAYER_UNSPECIFIED,
+	PVRDMA_LINK_LAYER_INFINIBAND,
+	PVRDMA_LINK_LAYER_ETHERNET,
+};
+
+enum pvrdma_mtu {
+	PVRDMA_MTU_256  = 1,
+	PVRDMA_MTU_512  = 2,
+	PVRDMA_MTU_1024 = 3,
+	PVRDMA_MTU_2048 = 4,
+	PVRDMA_MTU_4096 = 5,
+};
+
+static inline int pvrdma_mtu_enum_to_int(enum pvrdma_mtu mtu)
+{
+	switch (mtu) {
+	case PVRDMA_MTU_256:	return  256;
+	case PVRDMA_MTU_512:	return  512;
+	case PVRDMA_MTU_1024:	return 1024;
+	case PVRDMA_MTU_2048:	return 2048;
+	case PVRDMA_MTU_4096:	return 4096;
+	default:		return   -1;
+	}
+}
+
+static inline enum pvrdma_mtu pvrdma_mtu_int_to_enum(int mtu)
+{
+	switch (mtu) {
+	case 256:	return PVRDMA_MTU_256;
+	case 512:	return PVRDMA_MTU_512;
+	case 1024:	return PVRDMA_MTU_1024;
+	case 2048:	return PVRDMA_MTU_2048;
+	case 4096:
+	default:	return PVRDMA_MTU_4096;
+	}
+}
+
+enum pvrdma_port_state {
+	PVRDMA_PORT_NOP			= 0,
+	PVRDMA_PORT_DOWN		= 1,
+	PVRDMA_PORT_INIT		= 2,
+	PVRDMA_PORT_ARMED		= 3,
+	PVRDMA_PORT_ACTIVE		= 4,
+	PVRDMA_PORT_ACTIVE_DEFER	= 5,
+};
+
+enum pvrdma_port_cap_flags {
+	PVRDMA_PORT_SM				= 1 <<  1,
+	PVRDMA_PORT_NOTICE_SUP			= 1 <<  2,
+	PVRDMA_PORT_TRAP_SUP			= 1 <<  3,
+	PVRDMA_PORT_OPT_IPD_SUP			= 1 <<  4,
+	PVRDMA_PORT_AUTO_MIGR_SUP		= 1 <<  5,
+	PVRDMA_PORT_SL_MAP_SUP			= 1 <<  6,
+	PVRDMA_PORT_MKEY_NVRAM			= 1 <<  7,
+	PVRDMA_PORT_PKEY_NVRAM			= 1 <<  8,
+	PVRDMA_PORT_LED_INFO_SUP		= 1 <<  9,
+	PVRDMA_PORT_SM_DISABLED			= 1 << 10,
+	PVRDMA_PORT_SYS_IMAGE_GUID_SUP		= 1 << 11,
+	PVRDMA_PORT_PKEY_SW_EXT_PORT_TRAP_SUP	= 1 << 12,
+	PVRDMA_PORT_EXTENDED_SPEEDS_SUP		= 1 << 14,
+	PVRDMA_PORT_CM_SUP			= 1 << 16,
+	PVRDMA_PORT_SNMP_TUNNEL_SUP		= 1 << 17,
+	PVRDMA_PORT_REINIT_SUP			= 1 << 18,
+	PVRDMA_PORT_DEVICE_MGMT_SUP		= 1 << 19,
+	PVRDMA_PORT_VENDOR_CLASS_SUP		= 1 << 20,
+	PVRDMA_PORT_DR_NOTICE_SUP		= 1 << 21,
+	PVRDMA_PORT_CAP_MASK_NOTICE_SUP		= 1 << 22,
+	PVRDMA_PORT_BOOT_MGMT_SUP		= 1 << 23,
+	PVRDMA_PORT_LINK_LATENCY_SUP		= 1 << 24,
+	PVRDMA_PORT_CLIENT_REG_SUP		= 1 << 25,
+	PVRDMA_PORT_IP_BASED_GIDS		= 1 << 26,
+	PVRDMA_PORT_CAP_FLAGS_MAX		= PVRDMA_PORT_IP_BASED_GIDS,
+};
+
+enum pvrdma_port_width {
+	PVRDMA_WIDTH_1X		= 1,
+	PVRDMA_WIDTH_4X		= 2,
+	PVRDMA_WIDTH_8X		= 4,
+	PVRDMA_WIDTH_12X	= 8,
+};
+
+static inline int pvrdma_width_enum_to_int(enum pvrdma_port_width width)
+{
+	switch (width) {
+	case PVRDMA_WIDTH_1X:	return  1;
+	case PVRDMA_WIDTH_4X:	return  4;
+	case PVRDMA_WIDTH_8X:	return  8;
+	case PVRDMA_WIDTH_12X:	return 12;
+	default:		return -1;
+	}
+}
+
+enum pvrdma_port_speed {
+	PVRDMA_SPEED_SDR	= 1,
+	PVRDMA_SPEED_DDR	= 2,
+	PVRDMA_SPEED_QDR	= 4,
+	PVRDMA_SPEED_FDR10	= 8,
+	PVRDMA_SPEED_FDR	= 16,
+	PVRDMA_SPEED_EDR	= 32,
+};
+
+struct pvrdma_port_attr {
+	enum pvrdma_port_state	state;
+	enum pvrdma_mtu		max_mtu;
+	enum pvrdma_mtu		active_mtu;
+	u32			gid_tbl_len;
+	u32			port_cap_flags;
+	u32			max_msg_sz;
+	u32			bad_pkey_cntr;
+	u32			qkey_viol_cntr;
+	u16			pkey_tbl_len;
+	u16			lid;
+	u16			sm_lid;
+	u8			lmc;
+	u8			max_vl_num;
+	u8			sm_sl;
+	u8			subnet_timeout;
+	u8			init_type_reply;
+	u8			active_width;
+	u8			active_speed;
+	u8			phys_state;
+	u8			reserved[2];
+};
+
+struct pvrdma_global_route {
+	union pvrdma_gid	dgid;
+	u32			flow_label;
+	u8			sgid_index;
+	u8			hop_limit;
+	u8			traffic_class;
+	u8			reserved;
+};
+
+struct pvrdma_grh {
+	__be32			version_tclass_flow;
+	__be16			paylen;
+	u8			next_hdr;
+	u8			hop_limit;
+	union pvrdma_gid	sgid;
+	union pvrdma_gid	dgid;
+};
+
+enum pvrdma_ah_flags {
+	PVRDMA_AH_GRH = 1,
+};
+
+enum pvrdma_rate {
+	PVRDMA_RATE_PORT_CURRENT	= 0,
+	PVRDMA_RATE_2_5_GBPS		= 2,
+	PVRDMA_RATE_5_GBPS		= 5,
+	PVRDMA_RATE_10_GBPS		= 3,
+	PVRDMA_RATE_20_GBPS		= 6,
+	PVRDMA_RATE_30_GBPS		= 4,
+	PVRDMA_RATE_40_GBPS		= 7,
+	PVRDMA_RATE_60_GBPS		= 8,
+	PVRDMA_RATE_80_GBPS		= 9,
+	PVRDMA_RATE_120_GBPS		= 10,
+	PVRDMA_RATE_14_GBPS		= 11,
+	PVRDMA_RATE_56_GBPS		= 12,
+	PVRDMA_RATE_112_GBPS		= 13,
+	PVRDMA_RATE_168_GBPS		= 14,
+	PVRDMA_RATE_25_GBPS		= 15,
+	PVRDMA_RATE_100_GBPS		= 16,
+	PVRDMA_RATE_200_GBPS		= 17,
+	PVRDMA_RATE_300_GBPS		= 18,
+};
+
+struct pvrdma_ah_attr {
+	struct pvrdma_global_route	grh;
+	u16				dlid;
+	u16				vlan_id;
+	u8				sl;
+	u8				src_path_bits;
+	u8				static_rate;
+	u8				ah_flags;
+	u8				port_num;
+	u8				dmac[6];
+	u8				reserved;
+};
+
+enum pvrdma_cq_notify_flags {
+	PVRDMA_CQ_SOLICITED		= 1 << 0,
+	PVRDMA_CQ_NEXT_COMP		= 1 << 1,
+	PVRDMA_CQ_SOLICITED_MASK	= PVRDMA_CQ_SOLICITED |
+					  PVRDMA_CQ_NEXT_COMP,
+	PVRDMA_CQ_REPORT_MISSED_EVENTS	= 1 << 2,
+};
+
+struct pvrdma_qp_cap {
+	u32	max_send_wr;
+	u32	max_recv_wr;
+	u32	max_send_sge;
+	u32	max_recv_sge;
+	u32	max_inline_data;
+	u32	reserved;
+};
+
+enum pvrdma_sig_type {
+	PVRDMA_SIGNAL_ALL_WR,
+	PVRDMA_SIGNAL_REQ_WR,
+};
+
+enum pvrdma_qp_type {
+	PVRDMA_QPT_SMI,
+	PVRDMA_QPT_GSI,
+	PVRDMA_QPT_RC,
+	PVRDMA_QPT_UC,
+	PVRDMA_QPT_UD,
+	PVRDMA_QPT_RAW_IPV6,
+	PVRDMA_QPT_RAW_ETHERTYPE,
+	PVRDMA_QPT_RAW_PACKET = 8,
+	PVRDMA_QPT_XRC_INI = 9,
+	PVRDMA_QPT_XRC_TGT,
+	PVRDMA_QPT_MAX,
+};
+
+enum pvrdma_qp_create_flags {
+	PVRDMA_QP_CREATE_IPOPVRDMA_UD_LSO		= 1 << 0,
+	PVRDMA_QP_CREATE_BLOCK_MULTICAST_LOOPBACK	= 1 << 1,
+};
+
+enum pvrdma_qp_attr_mask {
+	PVRDMA_QP_STATE			= 1 << 0,
+	PVRDMA_QP_CUR_STATE		= 1 << 1,
+	PVRDMA_QP_EN_SQD_ASYNC_NOTIFY	= 1 << 2,
+	PVRDMA_QP_ACCESS_FLAGS		= 1 << 3,
+	PVRDMA_QP_PKEY_INDEX		= 1 << 4,
+	PVRDMA_QP_PORT			= 1 << 5,
+	PVRDMA_QP_QKEY			= 1 << 6,
+	PVRDMA_QP_AV			= 1 << 7,
+	PVRDMA_QP_PATH_MTU		= 1 << 8,
+	PVRDMA_QP_TIMEOUT		= 1 << 9,
+	PVRDMA_QP_RETRY_CNT		= 1 << 10,
+	PVRDMA_QP_RNR_RETRY		= 1 << 11,
+	PVRDMA_QP_RQ_PSN		= 1 << 12,
+	PVRDMA_QP_MAX_QP_RD_ATOMIC	= 1 << 13,
+	PVRDMA_QP_ALT_PATH		= 1 << 14,
+	PVRDMA_QP_MIN_RNR_TIMER		= 1 << 15,
+	PVRDMA_QP_SQ_PSN		= 1 << 16,
+	PVRDMA_QP_MAX_DEST_RD_ATOMIC	= 1 << 17,
+	PVRDMA_QP_PATH_MIG_STATE	= 1 << 18,
+	PVRDMA_QP_CAP			= 1 << 19,
+	PVRDMA_QP_DEST_QPN		= 1 << 20,
+	PVRDMA_QP_ATTR_MASK_MAX		= PVRDMA_QP_DEST_QPN,
+};
+
+enum pvrdma_qp_state {
+	PVRDMA_QPS_RESET,
+	PVRDMA_QPS_INIT,
+	PVRDMA_QPS_RTR,
+	PVRDMA_QPS_RTS,
+	PVRDMA_QPS_SQD,
+	PVRDMA_QPS_SQE,
+	PVRDMA_QPS_ERR,
+};
+
+enum pvrdma_mig_state {
+	PVRDMA_MIG_MIGRATED,
+	PVRDMA_MIG_REARM,
+	PVRDMA_MIG_ARMED,
+};
+
+enum pvrdma_mw_type {
+	PVRDMA_MW_TYPE_1 = 1,
+	PVRDMA_MW_TYPE_2 = 2,
+};
+
+struct pvrdma_qp_attr {
+	enum pvrdma_qp_state	qp_state;
+	enum pvrdma_qp_state	cur_qp_state;
+	enum pvrdma_mtu		path_mtu;
+	enum pvrdma_mig_state	path_mig_state;
+	u32			qkey;
+	u32			rq_psn;
+	u32			sq_psn;
+	u32			dest_qp_num;
+	u32			qp_access_flags;
+	u16			pkey_index;
+	u16			alt_pkey_index;
+	u8			en_sqd_async_notify;
+	u8			sq_draining;
+	u8			max_rd_atomic;
+	u8			max_dest_rd_atomic;
+	u8			min_rnr_timer;
+	u8			port_num;
+	u8			timeout;
+	u8			retry_cnt;
+	u8			rnr_retry;
+	u8			alt_port_num;
+	u8			alt_timeout;
+	u8			reserved[5];
+	struct pvrdma_qp_cap	cap;
+	struct pvrdma_ah_attr	ah_attr;
+	struct pvrdma_ah_attr	alt_ah_attr;
+};
+
+enum pvrdma_send_flags {
+	PVRDMA_SEND_FENCE	= 1 << 0,
+	PVRDMA_SEND_SIGNALED	= 1 << 1,
+	PVRDMA_SEND_SOLICITED	= 1 << 2,
+	PVRDMA_SEND_INLINE	= 1 << 3,
+	PVRDMA_SEND_IP_CSUM	= 1 << 4,
+	PVRDMA_SEND_FLAGS_MAX	= PVRDMA_SEND_IP_CSUM,
+};
+
+enum pvrdma_access_flags {
+	PVRDMA_ACCESS_LOCAL_WRITE	= 1 << 0,
+	PVRDMA_ACCESS_REMOTE_WRITE	= 1 << 1,
+	PVRDMA_ACCESS_REMOTE_READ	= 1 << 2,
+	PVRDMA_ACCESS_REMOTE_ATOMIC	= 1 << 3,
+	PVRDMA_ACCESS_MW_BIND		= 1 << 4,
+	PVRDMA_ZERO_BASED		= 1 << 5,
+	PVRDMA_ACCESS_ON_DEMAND		= 1 << 6,
+	PVRDMA_ACCESS_FLAGS_MAX		= PVRDMA_ACCESS_ON_DEMAND,
+};
+
+int pvrdma_query_device(struct ib_device *ibdev,
+			struct ib_device_attr *props,
+			struct ib_udata *udata);
+int pvrdma_query_port(struct ib_device *ibdev, u8 port,
+		      struct ib_port_attr *props);
+int pvrdma_query_gid(struct ib_device *ibdev, u8 port,
+		     int index, union ib_gid *gid);
+int pvrdma_query_pkey(struct ib_device *ibdev, u8 port,
+		      u16 index, u16 *pkey);
+enum rdma_link_layer pvrdma_port_link_layer(struct ib_device *ibdev,
+					    u8 port);
+int pvrdma_modify_device(struct ib_device *ibdev, int mask,
+			 struct ib_device_modify *props);
+int pvrdma_modify_port(struct ib_device *ibdev, u8 port,
+		       int mask, struct ib_port_modify *props);
+int pvrdma_mmap(struct ib_ucontext *context, struct vm_area_struct *vma);
+struct ib_ucontext *pvrdma_alloc_ucontext(struct ib_device *ibdev,
+					  struct ib_udata *udata);
+int pvrdma_dealloc_ucontext(struct ib_ucontext *context);
+struct ib_pd *pvrdma_alloc_pd(struct ib_device *ibdev,
+			      struct ib_ucontext *context,
+			      struct ib_udata *udata);
+int pvrdma_dealloc_pd(struct ib_pd *ibpd);
+struct ib_mr *pvrdma_get_dma_mr(struct ib_pd *pd, int acc);
+struct ib_mr *pvrdma_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
+				 u64 virt_addr, int access_flags,
+				 struct ib_udata *udata);
+int pvrdma_dereg_mr(struct ib_mr *mr);
+struct ib_mr *pvrdma_alloc_mr(struct ib_pd *pd, enum ib_mr_type mr_type,
+			      u32 max_num_sg);
+int pvrdma_map_mr_sg(struct ib_mr *ibmr, struct scatterlist *sg,
+		     int sg_nents, unsigned int *sg_offset);
+int pvrdma_modify_cq(struct ib_cq *cq, u16 cq_count, u16 cq_period);
+int pvrdma_resize_cq(struct ib_cq *ibcq, int entries,
+		     struct ib_udata *udata);
+struct ib_cq *pvrdma_create_cq(struct ib_device *ibdev,
+			       const struct ib_cq_init_attr *attr,
+			       struct ib_ucontext *context,
+			       struct ib_udata *udata);
+int pvrdma_resize_cq(struct ib_cq *ibcq, int entries,
+		     struct ib_udata *udata);
+int pvrdma_destroy_cq(struct ib_cq *cq);
+int pvrdma_poll_cq(struct ib_cq *ibcq, int num_entries, struct ib_wc *wc);
+int pvrdma_req_notify_cq(struct ib_cq *cq, enum ib_cq_notify_flags flags);
+struct ib_ah *pvrdma_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr,
+			       struct ib_udata *udata);
+int pvrdma_destroy_ah(struct ib_ah *ah);
+struct ib_qp *pvrdma_create_qp(struct ib_pd *pd,
+			       struct ib_qp_init_attr *init_attr,
+			       struct ib_udata *udata);
+int pvrdma_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
+		     int attr_mask, struct ib_udata *udata);
+int pvrdma_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr,
+		    int qp_attr_mask, struct ib_qp_init_attr *qp_init_attr);
+int pvrdma_destroy_qp(struct ib_qp *qp);
+int pvrdma_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
+		     struct ib_send_wr **bad_wr);
+int pvrdma_post_recv(struct ib_qp *ibqp, struct ib_recv_wr *wr,
+		     struct ib_recv_wr **bad_wr);
+
+#endif /* __PVRDMA_VERBS_H__ */
-- 
cgit v1.2.1