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

Pull rdma fixes from Jason Gunthorpe:
 "Nothing particularly exciting, some small ODP regressions from the mmu
  notifier rework, another bunch of syzkaller fixes, and a bug fix for a
  botched syzkaller fix in the first rc pull request.

   - Fix busted syzkaller fix in 'get_new_pps' - this turned out to
     crash on certain HW configurations

   - Bug fixes for various missed things in error unwinds

   - Add a missing rcu_read_lock annotation in hfi/qib

   - Fix two ODP related regressions from the recent mmu notifier
     changes

   - Several more syzkaller bugs in siw, RDMA netlink, verbs and iwcm

   - Revert an old patch in CMA as it is now shown to not be allocating
     port numbers properly"

* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:
  RDMA/iwcm: Fix iwcm work deallocation
  RDMA/siw: Fix failure handling during device creation
  RDMA/nldev: Fix crash when set a QP to a new counter but QPN is missing
  RDMA/odp: Ensure the mm is still alive before creating an implicit child
  RDMA/core: Fix protection fault in ib_mr_pool_destroy
  IB/mlx5: Fix implicit ODP race
  IB/hfi1, qib: Ensure RCU is locked when accessing list
  RDMA/core: Fix pkey and port assignment in get_new_pps
  RMDA/cm: Fix missing ib_cm_destroy_id() in ib_cm_insert_listen()
  RDMA/rw: Fix error flow during RDMA context initialization
  RDMA/core: Fix use of logical OR in get_new_pps
  Revert "RDMA/cma: Simplify rdma_resolve_addr() error flow"

15 files changed, 95 insertions, 59 deletions

diff --git a/drivers/infiniband/core/cm.c b/drivers/infiniband/core/cm.c
index 68cc1b2d6824..15e99a888427 100644
--- a/drivers/infiniband/core/cm.c
+++ b/drivers/infiniband/core/cm.c
@@ -1191,6 +1191,7 @@ struct ib_cm_id *ib_cm_insert_listen(struct ib_device *device,
 			/* Sharing an ib_cm_id with different handlers is not
 			 * supported */
 			spin_unlock_irqrestore(&cm.lock, flags);
+			ib_destroy_cm_id(cm_id);
 			return ERR_PTR(-EINVAL);
 		}
 		refcount_inc(&cm_id_priv->refcount);
diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c
index 72f032160c4b..2dec3a02ab9f 100644
--- a/drivers/infiniband/core/cma.c
+++ b/drivers/infiniband/core/cma.c
@@ -3212,19 +3212,26 @@ int rdma_resolve_addr(struct rdma_cm_id *id, struct sockaddr *src_addr,
 	int ret;
 
 	id_priv = container_of(id, struct rdma_id_private, id);
+	memcpy(cma_dst_addr(id_priv), dst_addr, rdma_addr_size(dst_addr));
 	if (id_priv->state == RDMA_CM_IDLE) {
 		ret = cma_bind_addr(id, src_addr, dst_addr);
-		if (ret)
+		if (ret) {
+			memset(cma_dst_addr(id_priv), 0,
+			       rdma_addr_size(dst_addr));
 			return ret;
+		}
 	}
 
-	if (cma_family(id_priv) != dst_addr->sa_family)
+	if (cma_family(id_priv) != dst_addr->sa_family) {
+		memset(cma_dst_addr(id_priv), 0, rdma_addr_size(dst_addr));
 		return -EINVAL;
+	}
 
-	if (!cma_comp_exch(id_priv, RDMA_CM_ADDR_BOUND, RDMA_CM_ADDR_QUERY))
+	if (!cma_comp_exch(id_priv, RDMA_CM_ADDR_BOUND, RDMA_CM_ADDR_QUERY)) {
+		memset(cma_dst_addr(id_priv), 0, rdma_addr_size(dst_addr));
 		return -EINVAL;
+	}
 
-	memcpy(cma_dst_addr(id_priv), dst_addr, rdma_addr_size(dst_addr));
 	if (cma_any_addr(dst_addr)) {
 		ret = cma_resolve_loopback(id_priv);
 	} else {
diff --git a/drivers/infiniband/core/core_priv.h b/drivers/infiniband/core/core_priv.h
index b1457b3464d3..cf42acca4a3a 100644
--- a/drivers/infiniband/core/core_priv.h
+++ b/drivers/infiniband/core/core_priv.h
@@ -338,6 +338,20 @@ static inline struct ib_qp *_ib_create_qp(struct ib_device *dev,
 	qp->pd = pd;
 	qp->uobject = uobj;
 	qp->real_qp = qp;
+
+	qp->qp_type = attr->qp_type;
+	qp->rwq_ind_tbl = attr->rwq_ind_tbl;
+	qp->send_cq = attr->send_cq;
+	qp->recv_cq = attr->recv_cq;
+	qp->srq = attr->srq;
+	qp->rwq_ind_tbl = attr->rwq_ind_tbl;
+	qp->event_handler = attr->event_handler;
+
+	atomic_set(&qp->usecnt, 0);
+	spin_lock_init(&qp->mr_lock);
+	INIT_LIST_HEAD(&qp->rdma_mrs);
+	INIT_LIST_HEAD(&qp->sig_mrs);
+
 	/*
 	 * We don't track XRC QPs for now, because they don't have PD
 	 * and more importantly they are created internaly by driver,
diff --git a/drivers/infiniband/core/iwcm.c b/drivers/infiniband/core/iwcm.c
index ade71823370f..da8adadf4755 100644
--- a/drivers/infiniband/core/iwcm.c
+++ b/drivers/infiniband/core/iwcm.c
@@ -159,8 +159,10 @@ static void dealloc_work_entries(struct iwcm_id_private *cm_id_priv)
 {
 	struct list_head *e, *tmp;
 
-	list_for_each_safe(e, tmp, &cm_id_priv->work_free_list)
+	list_for_each_safe(e, tmp, &cm_id_priv->work_free_list) {
+		list_del(e);
 		kfree(list_entry(e, struct iwcm_work, free_list));
+	}
 }
 
 static int alloc_work_entries(struct iwcm_id_private *cm_id_priv, int count)
diff --git a/drivers/infiniband/core/nldev.c b/drivers/infiniband/core/nldev.c
index 37b433aa7306..e0b0a91da696 100644
--- a/drivers/infiniband/core/nldev.c
+++ b/drivers/infiniband/core/nldev.c
@@ -1757,6 +1757,8 @@ static int nldev_stat_set_doit(struct sk_buff *skb, struct nlmsghdr *nlh,
 		if (ret)
 			goto err_msg;
 	} else {
+		if (!tb[RDMA_NLDEV_ATTR_RES_LQPN])
+			goto err_msg;
 		qpn = nla_get_u32(tb[RDMA_NLDEV_ATTR_RES_LQPN]);
 		if (tb[RDMA_NLDEV_ATTR_STAT_COUNTER_ID]) {
 			cntn = nla_get_u32(tb[RDMA_NLDEV_ATTR_STAT_COUNTER_ID]);
diff --git a/drivers/infiniband/core/rw.c b/drivers/infiniband/core/rw.c
index 4fad732f9b3c..06e5b6787443 100644
--- a/drivers/infiniband/core/rw.c
+++ b/drivers/infiniband/core/rw.c
@@ -273,6 +273,23 @@ static int rdma_rw_init_single_wr(struct rdma_rw_ctx *ctx, struct ib_qp *qp,
 	return 1;
 }
 
+static void rdma_rw_unmap_sg(struct ib_device *dev, struct scatterlist *sg,
+			     u32 sg_cnt, enum dma_data_direction dir)
+{
+	if (is_pci_p2pdma_page(sg_page(sg)))
+		pci_p2pdma_unmap_sg(dev->dma_device, sg, sg_cnt, dir);
+	else
+		ib_dma_unmap_sg(dev, sg, sg_cnt, dir);
+}
+
+static int rdma_rw_map_sg(struct ib_device *dev, struct scatterlist *sg,
+			  u32 sg_cnt, enum dma_data_direction dir)
+{
+	if (is_pci_p2pdma_page(sg_page(sg)))
+		return pci_p2pdma_map_sg(dev->dma_device, sg, sg_cnt, dir);
+	return ib_dma_map_sg(dev, sg, sg_cnt, dir);
+}
+
 /**
  * rdma_rw_ctx_init - initialize a RDMA READ/WRITE context
  * @ctx:	context to initialize
@@ -295,11 +312,7 @@ int rdma_rw_ctx_init(struct rdma_rw_ctx *ctx, struct ib_qp *qp, u8 port_num,
 	struct ib_device *dev = qp->pd->device;
 	int ret;
 
-	if (is_pci_p2pdma_page(sg_page(sg)))
-		ret = pci_p2pdma_map_sg(dev->dma_device, sg, sg_cnt, dir);
-	else
-		ret = ib_dma_map_sg(dev, sg, sg_cnt, dir);
-
+	ret = rdma_rw_map_sg(dev, sg, sg_cnt, dir);
 	if (!ret)
 		return -ENOMEM;
 	sg_cnt = ret;
@@ -338,7 +351,7 @@ int rdma_rw_ctx_init(struct rdma_rw_ctx *ctx, struct ib_qp *qp, u8 port_num,
 	return ret;
 
 out_unmap_sg:
-	ib_dma_unmap_sg(dev, sg, sg_cnt, dir);
+	rdma_rw_unmap_sg(dev, sg, sg_cnt, dir);
 	return ret;
 }
 EXPORT_SYMBOL(rdma_rw_ctx_init);
@@ -588,11 +601,7 @@ void rdma_rw_ctx_destroy(struct rdma_rw_ctx *ctx, struct ib_qp *qp, u8 port_num,
 		break;
 	}
 
-	if (is_pci_p2pdma_page(sg_page(sg)))
-		pci_p2pdma_unmap_sg(qp->pd->device->dma_device, sg,
-				    sg_cnt, dir);
-	else
-		ib_dma_unmap_sg(qp->pd->device, sg, sg_cnt, dir);
+	rdma_rw_unmap_sg(qp->pd->device, sg, sg_cnt, dir);
 }
 EXPORT_SYMBOL(rdma_rw_ctx_destroy);
 
diff --git a/drivers/infiniband/core/security.c b/drivers/infiniband/core/security.c
index 2b4d80393bd0..2d5608315dc8 100644
--- a/drivers/infiniband/core/security.c
+++ b/drivers/infiniband/core/security.c
@@ -340,15 +340,19 @@ static struct ib_ports_pkeys *get_new_pps(const struct ib_qp *qp,
 		return NULL;
 
 	if (qp_attr_mask & IB_QP_PORT)
-		new_pps->main.port_num =
-			(qp_pps) ? qp_pps->main.port_num : qp_attr->port_num;
+		new_pps->main.port_num = qp_attr->port_num;
+	else if (qp_pps)
+		new_pps->main.port_num = qp_pps->main.port_num;
+
 	if (qp_attr_mask & IB_QP_PKEY_INDEX)
-		new_pps->main.pkey_index = (qp_pps) ? qp_pps->main.pkey_index :
-						      qp_attr->pkey_index;
+		new_pps->main.pkey_index = qp_attr->pkey_index;
+	else if (qp_pps)
+		new_pps->main.pkey_index = qp_pps->main.pkey_index;
+
 	if ((qp_attr_mask & IB_QP_PKEY_INDEX) && (qp_attr_mask & IB_QP_PORT))
 		new_pps->main.state = IB_PORT_PKEY_VALID;
 
-	if (!(qp_attr_mask & (IB_QP_PKEY_INDEX || IB_QP_PORT)) && qp_pps) {
+	if (!(qp_attr_mask & (IB_QP_PKEY_INDEX | IB_QP_PORT)) && qp_pps) {
 		new_pps->main.port_num = qp_pps->main.port_num;
 		new_pps->main.pkey_index = qp_pps->main.pkey_index;
 		if (qp_pps->main.state != IB_PORT_PKEY_NOT_VALID)
diff --git a/drivers/infiniband/core/umem_odp.c b/drivers/infiniband/core/umem_odp.c
index b8c657b28380..cd656ad4953b 100644
--- a/drivers/infiniband/core/umem_odp.c
+++ b/drivers/infiniband/core/umem_odp.c
@@ -181,14 +181,28 @@ ib_umem_odp_alloc_child(struct ib_umem_odp *root, unsigned long addr,
 	odp_data->page_shift = PAGE_SHIFT;
 	odp_data->notifier.ops = ops;
 
+	/*
+	 * A mmget must be held when registering a notifier, the owming_mm only
+	 * has a mm_grab at this point.
+	 */
+	if (!mmget_not_zero(umem->owning_mm)) {
+		ret = -EFAULT;
+		goto out_free;
+	}
+
 	odp_data->tgid = get_pid(root->tgid);
 	ret = ib_init_umem_odp(odp_data, ops);
-	if (ret) {
-		put_pid(odp_data->tgid);
-		kfree(odp_data);
-		return ERR_PTR(ret);
-	}
+	if (ret)
+		goto out_tgid;
+	mmput(umem->owning_mm);
 	return odp_data;
+
+out_tgid:
+	put_pid(odp_data->tgid);
+	mmput(umem->owning_mm);
+out_free:
+	kfree(odp_data);
+	return ERR_PTR(ret);
 }
 EXPORT_SYMBOL(ib_umem_odp_alloc_child);
 
diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c
index 025933752e1d..060b4ebbd2ba 100644
--- a/drivers/infiniband/core/uverbs_cmd.c
+++ b/drivers/infiniband/core/uverbs_cmd.c
@@ -1445,16 +1445,7 @@ static int create_qp(struct uverbs_attr_bundle *attrs,
 		if (ret)
 			goto err_cb;
 
-		qp->pd		  = pd;
-		qp->send_cq	  = attr.send_cq;
-		qp->recv_cq	  = attr.recv_cq;
-		qp->srq		  = attr.srq;
-		qp->rwq_ind_tbl	  = ind_tbl;
-		qp->event_handler = attr.event_handler;
-		qp->qp_type	  = attr.qp_type;
-		atomic_set(&qp->usecnt, 0);
 		atomic_inc(&pd->usecnt);
-		qp->port = 0;
 		if (attr.send_cq)
 			atomic_inc(&attr.send_cq->usecnt);
 		if (attr.recv_cq)
diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c
index 3ebae3b65c28..e62c9dfc7837 100644
--- a/drivers/infiniband/core/verbs.c
+++ b/drivers/infiniband/core/verbs.c
@@ -1185,16 +1185,6 @@ struct ib_qp *ib_create_qp_user(struct ib_pd *pd,
 	if (ret)
 		goto err;
 
-	qp->qp_type    = qp_init_attr->qp_type;
-	qp->rwq_ind_tbl = qp_init_attr->rwq_ind_tbl;
-
-	atomic_set(&qp->usecnt, 0);
-	qp->mrs_used = 0;
-	spin_lock_init(&qp->mr_lock);
-	INIT_LIST_HEAD(&qp->rdma_mrs);
-	INIT_LIST_HEAD(&qp->sig_mrs);
-	qp->port = 0;
-
 	if (qp_init_attr->qp_type == IB_QPT_XRC_TGT) {
 		struct ib_qp *xrc_qp =
 			create_xrc_qp_user(qp, qp_init_attr, udata);
diff --git a/drivers/infiniband/hw/hfi1/verbs.c b/drivers/infiniband/hw/hfi1/verbs.c
index 089e201d7550..2f6323ad9c59 100644
--- a/drivers/infiniband/hw/hfi1/verbs.c
+++ b/drivers/infiniband/hw/hfi1/verbs.c
@@ -515,10 +515,11 @@ static inline void hfi1_handle_packet(struct hfi1_packet *packet,
 				       opa_get_lid(packet->dlid, 9B));
 		if (!mcast)
 			goto drop;
+		rcu_read_lock();
 		list_for_each_entry_rcu(p, &mcast->qp_list, list) {
 			packet->qp = p->qp;
 			if (hfi1_do_pkey_check(packet))
-				goto drop;
+				goto unlock_drop;
 			spin_lock_irqsave(&packet->qp->r_lock, flags);
 			packet_handler = qp_ok(packet);
 			if (likely(packet_handler))
@@ -527,6 +528,7 @@ static inline void hfi1_handle_packet(struct hfi1_packet *packet,
 				ibp->rvp.n_pkt_drops++;
 			spin_unlock_irqrestore(&packet->qp->r_lock, flags);
 		}
+		rcu_read_unlock();
 		/*
 		 * Notify rvt_multicast_detach() if it is waiting for us
 		 * to finish.
diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h
index d9bffcc93587..bb78142bca5e 100644
--- a/drivers/infiniband/hw/mlx5/mlx5_ib.h
+++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h
@@ -636,6 +636,7 @@ struct mlx5_ib_mr {
 
 	/* For ODP and implicit */
 	atomic_t		num_deferred_work;
+	wait_queue_head_t       q_deferred_work;
 	struct xarray		implicit_children;
 	union {
 		struct rcu_head rcu;
diff --git a/drivers/infiniband/hw/mlx5/odp.c b/drivers/infiniband/hw/mlx5/odp.c
index 4216814ba871..bf50cd91f472 100644
--- a/drivers/infiniband/hw/mlx5/odp.c
+++ b/drivers/infiniband/hw/mlx5/odp.c
@@ -235,7 +235,8 @@ static void free_implicit_child_mr(struct mlx5_ib_mr *mr, bool need_imr_xlt)
 	mr->parent = NULL;
 	mlx5_mr_cache_free(mr->dev, mr);
 	ib_umem_odp_release(odp);
-	atomic_dec(&imr->num_deferred_work);
+	if (atomic_dec_and_test(&imr->num_deferred_work))
+		wake_up(&imr->q_deferred_work);
 }
 
 static void free_implicit_child_mr_work(struct work_struct *work)
@@ -554,6 +555,7 @@ struct mlx5_ib_mr *mlx5_ib_alloc_implicit_mr(struct mlx5_ib_pd *pd,
 	imr->umem = &umem_odp->umem;
 	imr->is_odp_implicit = true;
 	atomic_set(&imr->num_deferred_work, 0);
+	init_waitqueue_head(&imr->q_deferred_work);
 	xa_init(&imr->implicit_children);
 
 	err = mlx5_ib_update_xlt(imr, 0,
@@ -611,10 +613,7 @@ void mlx5_ib_free_implicit_mr(struct mlx5_ib_mr *imr)
 	 * under xa_lock while the child is in the xarray. Thus at this point
 	 * it is only decreasing, and all work holding it is now on the wq.
 	 */
-	if (atomic_read(&imr->num_deferred_work)) {
-		flush_workqueue(system_unbound_wq);
-		WARN_ON(atomic_read(&imr->num_deferred_work));
-	}
+	wait_event(imr->q_deferred_work, !atomic_read(&imr->num_deferred_work));
 
 	/*
 	 * Fence the imr before we destroy the children. This allows us to
@@ -645,10 +644,7 @@ void mlx5_ib_fence_odp_mr(struct mlx5_ib_mr *mr)
 	/* Wait for all running page-fault handlers to finish. */
 	synchronize_srcu(&mr->dev->odp_srcu);
 
-	if (atomic_read(&mr->num_deferred_work)) {
-		flush_workqueue(system_unbound_wq);
-		WARN_ON(atomic_read(&mr->num_deferred_work));
-	}
+	wait_event(mr->q_deferred_work, !atomic_read(&mr->num_deferred_work));
 
 	dma_fence_odp_mr(mr);
 }
@@ -1720,7 +1716,8 @@ static void destroy_prefetch_work(struct prefetch_mr_work *work)
 	u32 i;
 
 	for (i = 0; i < work->num_sge; ++i)
-		atomic_dec(&work->frags[i].mr->num_deferred_work);
+		if (atomic_dec_and_test(&work->frags[i].mr->num_deferred_work))
+			wake_up(&work->frags[i].mr->q_deferred_work);
 	kvfree(work);
 }
 
diff --git a/drivers/infiniband/hw/qib/qib_verbs.c b/drivers/infiniband/hw/qib/qib_verbs.c
index 33778d451b82..5ef93f8f17a1 100644
--- a/drivers/infiniband/hw/qib/qib_verbs.c
+++ b/drivers/infiniband/hw/qib/qib_verbs.c
@@ -329,8 +329,10 @@ void qib_ib_rcv(struct qib_ctxtdata *rcd, void *rhdr, void *data, u32 tlen)
 		if (mcast == NULL)
 			goto drop;
 		this_cpu_inc(ibp->pmastats->n_multicast_rcv);
+		rcu_read_lock();
 		list_for_each_entry_rcu(p, &mcast->qp_list, list)
 			qib_qp_rcv(rcd, hdr, 1, data, tlen, p->qp);
+		rcu_read_unlock();
 		/*
 		 * Notify rvt_multicast_detach() if it is waiting for us
 		 * to finish.
diff --git a/drivers/infiniband/sw/siw/siw_main.c b/drivers/infiniband/sw/siw/siw_main.c
index 96ed349c0939..5cd40fb9e20c 100644
--- a/drivers/infiniband/sw/siw/siw_main.c
+++ b/drivers/infiniband/sw/siw/siw_main.c
@@ -388,6 +388,9 @@ static struct siw_device *siw_device_create(struct net_device *netdev)
 		{ .max_segment_size = SZ_2G };
 	base_dev->num_comp_vectors = num_possible_cpus();
 
+	xa_init_flags(&sdev->qp_xa, XA_FLAGS_ALLOC1);
+	xa_init_flags(&sdev->mem_xa, XA_FLAGS_ALLOC1);
+
 	ib_set_device_ops(base_dev, &siw_device_ops);
 	rv = ib_device_set_netdev(base_dev, netdev, 1);
 	if (rv)
@@ -415,9 +418,6 @@ static struct siw_device *siw_device_create(struct net_device *netdev)
 	sdev->attrs.max_srq_wr = SIW_MAX_SRQ_WR;
 	sdev->attrs.max_srq_sge = SIW_MAX_SGE;
 
-	xa_init_flags(&sdev->qp_xa, XA_FLAGS_ALLOC1);
-	xa_init_flags(&sdev->mem_xa, XA_FLAGS_ALLOC1);
-
 	INIT_LIST_HEAD(&sdev->cep_list);
 	INIT_LIST_HEAD(&sdev->qp_list);