Merge branch 'stable-4.13' of git://git.infradead.org/users/pcmoore/selinux into next

drivers/infiniband/core/Makefile

@@ -10,7 +10,8 @@ obj-$(CONFIG_INFINIBAND_USER_ACCESS) +=	ib_uverbs.o ib_ucm.o \
 ib_core-y :=			packer.o ud_header.o verbs.o cq.o rw.o sysfs.o \
 				device.o fmr_pool.o cache.o netlink.o \
 				roce_gid_mgmt.o mr_pool.o addr.o sa_query.o \
-				multicast.o mad.o smi.o agent.o mad_rmpp.o
+				multicast.o mad.o smi.o agent.o mad_rmpp.o \
+				security.o
 ib_core-$(CONFIG_INFINIBAND_USER_MEM) += umem.o
 ib_core-$(CONFIG_INFINIBAND_ON_DEMAND_PAGING) += umem_odp.o umem_rbtree.o
 ib_core-$(CONFIG_CGROUP_RDMA) += cgroup.o

drivers/infiniband/core/cache.c

@@ -53,6 +53,7 @@ struct ib_update_work {
 	struct work_struct work;
 	struct ib_device  *device;
 	u8                 port_num;
+	bool		   enforce_security;
 };
 
 union ib_gid zgid;
@@ -911,6 +912,26 @@ int ib_get_cached_pkey(struct ib_device *device,
 }
 EXPORT_SYMBOL(ib_get_cached_pkey);
 
+int ib_get_cached_subnet_prefix(struct ib_device *device,
+				u8                port_num,
+				u64              *sn_pfx)
+{
+	unsigned long flags;
+	int p;
+
+	if (port_num < rdma_start_port(device) ||
+	    port_num > rdma_end_port(device))
+		return -EINVAL;
+
+	p = port_num - rdma_start_port(device);
+	read_lock_irqsave(&device->cache.lock, flags);
+	*sn_pfx = device->cache.ports[p].subnet_prefix;
+	read_unlock_irqrestore(&device->cache.lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL(ib_get_cached_subnet_prefix);
+
 int ib_find_cached_pkey(struct ib_device *device,
 			u8                port_num,
 			u16               pkey,
@@ -1022,7 +1043,8 @@ int ib_get_cached_port_state(struct ib_device   *device,
 EXPORT_SYMBOL(ib_get_cached_port_state);
 
 static void ib_cache_update(struct ib_device *device,
-			    u8                port)
+			    u8                port,
+			    bool	      enforce_security)
 {
 	struct ib_port_attr       *tprops = NULL;
 	struct ib_pkey_cache      *pkey_cache = NULL, *old_pkey_cache;
@@ -1108,8 +1130,15 @@ static void ib_cache_update(struct ib_device *device,
 	device->cache.ports[port - rdma_start_port(device)].port_state =
 		tprops->state;
 
+	device->cache.ports[port - rdma_start_port(device)].subnet_prefix =
+							tprops->subnet_prefix;
 	write_unlock_irq(&device->cache.lock);
 
+	if (enforce_security)
+		ib_security_cache_change(device,
+					 port,
+					 tprops->subnet_prefix);
+
 	kfree(gid_cache);
 	kfree(old_pkey_cache);
 	kfree(tprops);
@@ -1126,7 +1155,9 @@ static void ib_cache_task(struct work_struct *_work)
 	struct ib_update_work *work =
 		container_of(_work, struct ib_update_work, work);
 
-	ib_cache_update(work->device, work->port_num);
+	ib_cache_update(work->device,
+			work->port_num,
+			work->enforce_security);
 	kfree(work);
 }
 
@@ -1147,6 +1178,12 @@ static void ib_cache_event(struct ib_event_handler *handler,
 			INIT_WORK(&work->work, ib_cache_task);
 			work->device   = event->device;
 			work->port_num = event->element.port_num;
+			if (event->event == IB_EVENT_PKEY_CHANGE ||
+			    event->event == IB_EVENT_GID_CHANGE)
+				work->enforce_security = true;
+			else
+				work->enforce_security = false;
+
 			queue_work(ib_wq, &work->work);
 		}
 	}
@@ -1172,7 +1209,7 @@ int ib_cache_setup_one(struct ib_device *device)
 		goto out;
 
 	for (p = 0; p <= rdma_end_port(device) - rdma_start_port(device); ++p)
-		ib_cache_update(device, p + rdma_start_port(device));
+		ib_cache_update(device, p + rdma_start_port(device), true);
 
 	INIT_IB_EVENT_HANDLER(&device->cache.event_handler,
 			      device, ib_cache_event);

drivers/infiniband/core/core_priv.h

@@ -38,6 +38,16 @@
 #include <linux/cgroup_rdma.h>
 
 #include <rdma/ib_verbs.h>
+#include <rdma/ib_mad.h>
+#include "mad_priv.h"
+
+struct pkey_index_qp_list {
+	struct list_head    pkey_index_list;
+	u16                 pkey_index;
+	/* Lock to hold while iterating the qp_list. */
+	spinlock_t          qp_list_lock;
+	struct list_head    qp_list;
+};
 
 #if IS_ENABLED(CONFIG_INFINIBAND_ADDR_TRANS_CONFIGFS)
 int cma_configfs_init(void);
@@ -176,4 +186,109 @@ int ib_nl_handle_set_timeout(struct sk_buff *skb,
 int ib_nl_handle_ip_res_resp(struct sk_buff *skb,
 			     struct netlink_callback *cb);
 
+int ib_get_cached_subnet_prefix(struct ib_device *device,
+				u8                port_num,
+				u64              *sn_pfx);
+
+#ifdef CONFIG_SECURITY_INFINIBAND
+int ib_security_pkey_access(struct ib_device *dev,
+			    u8 port_num,
+			    u16 pkey_index,
+			    void *sec);
+
+void ib_security_destroy_port_pkey_list(struct ib_device *device);
+
+void ib_security_cache_change(struct ib_device *device,
+			      u8 port_num,
+			      u64 subnet_prefix);
+
+int ib_security_modify_qp(struct ib_qp *qp,
+			  struct ib_qp_attr *qp_attr,
+			  int qp_attr_mask,
+			  struct ib_udata *udata);
+
+int ib_create_qp_security(struct ib_qp *qp, struct ib_device *dev);
+void ib_destroy_qp_security_begin(struct ib_qp_security *sec);
+void ib_destroy_qp_security_abort(struct ib_qp_security *sec);
+void ib_destroy_qp_security_end(struct ib_qp_security *sec);
+int ib_open_shared_qp_security(struct ib_qp *qp, struct ib_device *dev);
+void ib_close_shared_qp_security(struct ib_qp_security *sec);
+int ib_mad_agent_security_setup(struct ib_mad_agent *agent,
+				enum ib_qp_type qp_type);
+void ib_mad_agent_security_cleanup(struct ib_mad_agent *agent);
+int ib_mad_enforce_security(struct ib_mad_agent_private *map, u16 pkey_index);
+#else
+static inline int ib_security_pkey_access(struct ib_device *dev,
+					  u8 port_num,
+					  u16 pkey_index,
+					  void *sec)
+{
+	return 0;
+}
+
+static inline void ib_security_destroy_port_pkey_list(struct ib_device *device)
+{
+}
+
+static inline void ib_security_cache_change(struct ib_device *device,
+					    u8 port_num,
+					    u64 subnet_prefix)
+{
+}
+
+static inline int ib_security_modify_qp(struct ib_qp *qp,
+					struct ib_qp_attr *qp_attr,
+					int qp_attr_mask,
+					struct ib_udata *udata)
+{
+	return qp->device->modify_qp(qp->real_qp,
+				     qp_attr,
+				     qp_attr_mask,
+				     udata);
+}
+
+static inline int ib_create_qp_security(struct ib_qp *qp,
+					struct ib_device *dev)
+{
+	return 0;
+}
+
+static inline void ib_destroy_qp_security_begin(struct ib_qp_security *sec)
+{
+}
+
+static inline void ib_destroy_qp_security_abort(struct ib_qp_security *sec)
+{
+}
+
+static inline void ib_destroy_qp_security_end(struct ib_qp_security *sec)
+{
+}
+
+static inline int ib_open_shared_qp_security(struct ib_qp *qp,
+					     struct ib_device *dev)
+{
+	return 0;
+}
+
+static inline void ib_close_shared_qp_security(struct ib_qp_security *sec)
+{
+}
+
+static inline int ib_mad_agent_security_setup(struct ib_mad_agent *agent,
+					      enum ib_qp_type qp_type)
+{
+	return 0;
+}
+
+static inline void ib_mad_agent_security_cleanup(struct ib_mad_agent *agent)
+{
+}
+
+static inline int ib_mad_enforce_security(struct ib_mad_agent_private *map,
+					  u16 pkey_index)
+{
+	return 0;
+}
+#endif
 #endif /* _CORE_PRIV_H */

drivers/infiniband/core/device.c

@@ -39,6 +39,8 @@
 #include <linux/init.h>
 #include <linux/mutex.h>
 #include <linux/netdevice.h>
+#include <linux/security.h>
+#include <linux/notifier.h>
 #include <rdma/rdma_netlink.h>
 #include <rdma/ib_addr.h>
 #include <rdma/ib_cache.h>
@@ -82,6 +84,14 @@ static LIST_HEAD(client_list);
 static DEFINE_MUTEX(device_mutex);
 static DECLARE_RWSEM(lists_rwsem);
 
+static int ib_security_change(struct notifier_block *nb, unsigned long event,
+			      void *lsm_data);
+static void ib_policy_change_task(struct work_struct *work);
+static DECLARE_WORK(ib_policy_change_work, ib_policy_change_task);
+
+static struct notifier_block ibdev_lsm_nb = {
+	.notifier_call = ib_security_change,
+};
 
 static int ib_device_check_mandatory(struct ib_device *device)
 {
@@ -325,6 +335,64 @@ void ib_get_device_fw_str(struct ib_device *dev, char *str, size_t str_len)
 }
 EXPORT_SYMBOL(ib_get_device_fw_str);
 
+static int setup_port_pkey_list(struct ib_device *device)
+{
+	int i;
+
+	/**
+	 * device->port_pkey_list is indexed directly by the port number,
+	 * Therefore it is declared as a 1 based array with potential empty
+	 * slots at the beginning.
+	 */
+	device->port_pkey_list = kcalloc(rdma_end_port(device) + 1,
+					 sizeof(*device->port_pkey_list),
+					 GFP_KERNEL);
+
+	if (!device->port_pkey_list)
+		return -ENOMEM;
+
+	for (i = 0; i < (rdma_end_port(device) + 1); i++) {
+		spin_lock_init(&device->port_pkey_list[i].list_lock);
+		INIT_LIST_HEAD(&device->port_pkey_list[i].pkey_list);
+	}
+
+	return 0;
+}
+
+static void ib_policy_change_task(struct work_struct *work)
+{
+	struct ib_device *dev;
+
+	down_read(&lists_rwsem);
+	list_for_each_entry(dev, &device_list, core_list) {
+		int i;
+
+		for (i = rdma_start_port(dev); i <= rdma_end_port(dev); i++) {
+			u64 sp;
+			int ret = ib_get_cached_subnet_prefix(dev,
+							      i,
+							      &sp);
+
+			WARN_ONCE(ret,
+				  "ib_get_cached_subnet_prefix err: %d, this should never happen here\n",
+				  ret);
+			ib_security_cache_change(dev, i, sp);
+		}
+	}
+	up_read(&lists_rwsem);
+}
+
+static int ib_security_change(struct notifier_block *nb, unsigned long event,
+			      void *lsm_data)
+{
+	if (event != LSM_POLICY_CHANGE)
+		return NOTIFY_DONE;
+
+	schedule_work(&ib_policy_change_work);
+
+	return NOTIFY_OK;
+}
+
 /**
  * ib_register_device - Register an IB device with IB core
  * @device:Device to register
@@ -385,6 +453,12 @@ int ib_register_device(struct ib_device *device,
 		goto out;
 	}
 
+	ret = setup_port_pkey_list(device);
+	if (ret) {
+		pr_warn("Couldn't create per port_pkey_list\n");
+		goto out;
+	}
+
 	ret = ib_cache_setup_one(device);
 	if (ret) {
 		pr_warn("Couldn't set up InfiniBand P_Key/GID cache\n");
@@ -468,6 +542,9 @@ void ib_unregister_device(struct ib_device *device)
 	ib_device_unregister_sysfs(device);
 	ib_cache_cleanup_one(device);
 
+	ib_security_destroy_port_pkey_list(device);
+	kfree(device->port_pkey_list);
+
 	down_write(&lists_rwsem);
 	spin_lock_irqsave(&device->client_data_lock, flags);
 	list_for_each_entry_safe(context, tmp, &device->client_data_list, list)
@@ -1082,10 +1159,18 @@ static int __init ib_core_init(void)
 		goto err_sa;
 	}
 
+	ret = register_lsm_notifier(&ibdev_lsm_nb);
+	if (ret) {
+		pr_warn("Couldn't register LSM notifier. ret %d\n", ret);
+		goto err_ibnl_clients;
+	}
+
 	ib_cache_setup();
 
 	return 0;
 
+err_ibnl_clients:
+	ib_remove_ibnl_clients();
 err_sa:
 	ib_sa_cleanup();
 err_mad:
@@ -1105,6 +1190,7 @@ err:
 
 static void __exit ib_core_cleanup(void)
 {
+	unregister_lsm_notifier(&ibdev_lsm_nb);
 	ib_cache_cleanup();
 	ib_remove_ibnl_clients();
 	ib_sa_cleanup();

drivers/infiniband/core/mad.c

@@ -40,9 +40,11 @@
 #include <linux/dma-mapping.h>
 #include <linux/slab.h>
 #include <linux/module.h>
+#include <linux/security.h>
 #include <rdma/ib_cache.h>
 
 #include "mad_priv.h"
+#include "core_priv.h"
 #include "mad_rmpp.h"
 #include "smi.h"
 #include "opa_smi.h"
@@ -369,6 +371,12 @@ struct ib_mad_agent *ib_register_mad_agent(struct ib_device *device,
 	atomic_set(&mad_agent_priv->refcount, 1);
 	init_completion(&mad_agent_priv->comp);
 
+	ret2 = ib_mad_agent_security_setup(&mad_agent_priv->agent, qp_type);
+	if (ret2) {
+		ret = ERR_PTR(ret2);
+		goto error4;
+	}
+
 	spin_lock_irqsave(&port_priv->reg_lock, flags);
 	mad_agent_priv->agent.hi_tid = ++ib_mad_client_id;
 
@@ -386,7 +394,7 @@ struct ib_mad_agent *ib_register_mad_agent(struct ib_device *device,
 				if (method) {
 					if (method_in_use(&method,
 							   mad_reg_req))
-						goto error4;
+						goto error5;
 				}
 			}
 			ret2 = add_nonoui_reg_req(mad_reg_req, mad_agent_priv,
@@ -402,14 +410,14 @@ struct ib_mad_agent *ib_register_mad_agent(struct ib_device *device,
 					if (is_vendor_method_in_use(
 							vendor_class,
 							mad_reg_req))
-						goto error4;
+						goto error5;
 				}
 			}
 			ret2 = add_oui_reg_req(mad_reg_req, mad_agent_priv);
 		}
 		if (ret2) {
 			ret = ERR_PTR(ret2);
-			goto error4;
+			goto error5;
 		}
 	}
 
@@ -418,9 +426,10 @@ struct ib_mad_agent *ib_register_mad_agent(struct ib_device *device,
 	spin_unlock_irqrestore(&port_priv->reg_lock, flags);
 
 	return &mad_agent_priv->agent;
-
-error4:
+error5:
 	spin_unlock_irqrestore(&port_priv->reg_lock, flags);
+	ib_mad_agent_security_cleanup(&mad_agent_priv->agent);
+error4:
 	kfree(reg_req);
 error3:
 	kfree(mad_agent_priv);
@@ -491,6 +500,7 @@ struct ib_mad_agent *ib_register_mad_snoop(struct ib_device *device,
 	struct ib_mad_agent *ret;
 	struct ib_mad_snoop_private *mad_snoop_priv;
 	int qpn;
+	int err;
 
 	/* Validate parameters */
 	if ((is_snooping_sends(mad_snoop_flags) && !snoop_handler) ||
@@ -525,17 +535,25 @@ struct ib_mad_agent *ib_register_mad_snoop(struct ib_device *device,
 	mad_snoop_priv->agent.port_num = port_num;
 	mad_snoop_priv->mad_snoop_flags = mad_snoop_flags;
 	init_completion(&mad_snoop_priv->comp);
+
+	err = ib_mad_agent_security_setup(&mad_snoop_priv->agent, qp_type);
+	if (err) {
+		ret = ERR_PTR(err);
+		goto error2;
+	}
+
 	mad_snoop_priv->snoop_index = register_snoop_agent(
 						&port_priv->qp_info[qpn],
 						mad_snoop_priv);
 	if (mad_snoop_priv->snoop_index < 0) {
 		ret = ERR_PTR(mad_snoop_priv->snoop_index);
-		goto error2;
+		goto error3;
 	}
 
 	atomic_set(&mad_snoop_priv->refcount, 1);
 	return &mad_snoop_priv->agent;
-
+error3:
+	ib_mad_agent_security_cleanup(&mad_snoop_priv->agent);
 error2:
 	kfree(mad_snoop_priv);
 error1:
@@ -581,6 +599,8 @@ static void unregister_mad_agent(struct ib_mad_agent_private *mad_agent_priv)
 	deref_mad_agent(mad_agent_priv);
 	wait_for_completion(&mad_agent_priv->comp);
 
+	ib_mad_agent_security_cleanup(&mad_agent_priv->agent);
+
 	kfree(mad_agent_priv->reg_req);
 	kfree(mad_agent_priv);
 }
@@ -599,6 +619,8 @@ static void unregister_mad_snoop(struct ib_mad_snoop_private *mad_snoop_priv)
 	deref_snoop_agent(mad_snoop_priv);
 	wait_for_completion(&mad_snoop_priv->comp);
 
+	ib_mad_agent_security_cleanup(&mad_snoop_priv->agent);
+
 	kfree(mad_snoop_priv);
 }
 
@@ -1215,12 +1237,16 @@ int ib_post_send_mad(struct ib_mad_send_buf *send_buf,
 
 	/* Walk list of send WRs and post each on send list */
 	for (; send_buf; send_buf = next_send_buf) {
-
 		mad_send_wr = container_of(send_buf,
 					   struct ib_mad_send_wr_private,
 					   send_buf);
 		mad_agent_priv = mad_send_wr->mad_agent_priv;
 
+		ret = ib_mad_enforce_security(mad_agent_priv,
+					      mad_send_wr->send_wr.pkey_index);
+		if (ret)
+			goto error;
+
 		if (!send_buf->mad_agent->send_handler ||
 		    (send_buf->timeout_ms &&
 		     !send_buf->mad_agent->recv_handler)) {
@@ -1946,6 +1972,14 @@ static void ib_mad_complete_recv(struct ib_mad_agent_private *mad_agent_priv,
 	struct ib_mad_send_wr_private *mad_send_wr;
 	struct ib_mad_send_wc mad_send_wc;
 	unsigned long flags;
+	int ret;
+
+	ret = ib_mad_enforce_security(mad_agent_priv,
+				      mad_recv_wc->wc->pkey_index);
+	if (ret) {
+		ib_free_recv_mad(mad_recv_wc);
+		deref_mad_agent(mad_agent_priv);
+	}
 
 	INIT_LIST_HEAD(&mad_recv_wc->rmpp_list);
 	list_add(&mad_recv_wc->recv_buf.list, &mad_recv_wc->rmpp_list);
@@ -2003,6 +2037,8 @@ static void ib_mad_complete_recv(struct ib_mad_agent_private *mad_agent_priv,
 						   mad_recv_wc);
 		deref_mad_agent(mad_agent_priv);
 	}
+
+	return;
 }
 
 static enum smi_action handle_ib_smi(const struct ib_mad_port_private *port_priv,

drivers/infiniband/core/uverbs_cmd.c

@@ -1508,6 +1508,10 @@ static int create_qp(struct ib_uverbs_file *file,
 	}
 
 	if (cmd->qp_type != IB_QPT_XRC_TGT) {
+		ret = ib_create_qp_security(qp, device);
+		if (ret)
+			goto err_cb;
+
 		qp->real_qp	  = qp;
 		qp->device	  = device;
 		qp->pd		  = pd;
@@ -2002,14 +2006,17 @@ static int modify_qp(struct ib_uverbs_file *file,
 			if (ret)
 				goto release_qp;
 		}
-		ret = qp->device->modify_qp(qp, attr,
+		ret = ib_security_modify_qp(qp,
+					    attr,
 					    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->base.attr_mask));
+		ret = ib_security_modify_qp(qp,
+					    attr,
+					    modify_qp_mask(qp->qp_type,
+							   cmd->base.attr_mask),
+					    NULL);
 	}
 
 release_qp:

drivers/infiniband/core/verbs.c

@@ -44,6 +44,7 @@
 #include <linux/in.h>
 #include <linux/in6.h>
 #include <net/addrconf.h>
+#include <linux/security.h>
 
 #include <rdma/ib_verbs.h>
 #include <rdma/ib_cache.h>
@@ -713,11 +714,19 @@ static struct ib_qp *__ib_open_qp(struct ib_qp *real_qp,
 {
 	struct ib_qp *qp;
 	unsigned long flags;
+	int err;
 
 	qp = kzalloc(sizeof *qp, GFP_KERNEL);
 	if (!qp)
 		return ERR_PTR(-ENOMEM);
 
+	qp->real_qp = real_qp;
+	err = ib_open_shared_qp_security(qp, real_qp->device);
+	if (err) {
+		kfree(qp);
+		return ERR_PTR(err);
+	}
+
 	qp->real_qp = real_qp;
 	atomic_inc(&real_qp->usecnt);
 	qp->device = real_qp->device;
@@ -804,6 +813,12 @@ struct ib_qp *ib_create_qp(struct ib_pd *pd,
 	if (IS_ERR(qp))
 		return qp;
 
+	ret = ib_create_qp_security(qp, device);
+	if (ret) {
+		ib_destroy_qp(qp);
+		return ERR_PTR(ret);
+	}
+
 	qp->device     = device;
 	qp->real_qp    = qp;
 	qp->uobject    = NULL;
@@ -1266,7 +1281,7 @@ int ib_modify_qp(struct ib_qp *qp,
 			return ret;
 	}
 
-	return qp->device->modify_qp(qp->real_qp, qp_attr, qp_attr_mask, NULL);
+	return ib_security_modify_qp(qp->real_qp, qp_attr, qp_attr_mask, NULL);
 }
 EXPORT_SYMBOL(ib_modify_qp);
 
@@ -1295,6 +1310,7 @@ int ib_close_qp(struct ib_qp *qp)
 	spin_unlock_irqrestore(&real_qp->device->event_handler_lock, flags);
 
 	atomic_dec(&real_qp->usecnt);
+	ib_close_shared_qp_security(qp->qp_sec);
 	kfree(qp);
 
 	return 0;
@@ -1335,6 +1351,7 @@ int ib_destroy_qp(struct ib_qp *qp)
 	struct ib_cq *scq, *rcq;
 	struct ib_srq *srq;
 	struct ib_rwq_ind_table *ind_tbl;
+	struct ib_qp_security *sec;
 	int ret;
 
 	WARN_ON_ONCE(qp->mrs_used > 0);
@@ -1350,6 +1367,9 @@ int ib_destroy_qp(struct ib_qp *qp)
 	rcq  = qp->recv_cq;
 	srq  = qp->srq;
 	ind_tbl = qp->rwq_ind_tbl;
+	sec  = qp->qp_sec;
+	if (sec)
+		ib_destroy_qp_security_begin(sec);
 
 	if (!qp->uobject)
 		rdma_rw_cleanup_mrs(qp);
@@ -1366,6 +1386,11 @@ int ib_destroy_qp(struct ib_qp *qp)
 			atomic_dec(&srq->usecnt);
 		if (ind_tbl)
 			atomic_dec(&ind_tbl->usecnt);
+		if (sec)
+			ib_destroy_qp_security_end(sec);
+	} else {
+		if (sec)
+			ib_destroy_qp_security_abort(sec);
 	}
 
 	return ret;

fs/nfs/super.c

@@ -2544,10 +2544,25 @@ EXPORT_SYMBOL_GPL(nfs_set_sb_security);
 int nfs_clone_sb_security(struct super_block *s, struct dentry *mntroot,
 			  struct nfs_mount_info *mount_info)
 {
+	int error;
+	unsigned long kflags = 0, kflags_out = 0;
+
 	/* clone any lsm security options from the parent to the new sb */
 	if (d_inode(mntroot)->i_op != NFS_SB(s)->nfs_client->rpc_ops->dir_inode_ops)
 		return -ESTALE;
-	return security_sb_clone_mnt_opts(mount_info->cloned->sb, s);
+
+	if (NFS_SB(s)->caps & NFS_CAP_SECURITY_LABEL)
+		kflags |= SECURITY_LSM_NATIVE_LABELS;
+
+	error = security_sb_clone_mnt_opts(mount_info->cloned->sb, s, kflags,
+			&kflags_out);
+	if (error)
+		return error;
+
+	if (NFS_SB(s)->caps & NFS_CAP_SECURITY_LABEL &&
+		!(kflags_out & SECURITY_LSM_NATIVE_LABELS))
+		NFS_SB(s)->caps &= ~NFS_CAP_SECURITY_LABEL;
+	return 0;
 }
 EXPORT_SYMBOL_GPL(nfs_clone_sb_security);
 

include/linux/lsm_audit.h

@@ -21,6 +21,7 @@
 #include <linux/path.h>
 #include <linux/key.h>
 #include <linux/skbuff.h>
+#include <rdma/ib_verbs.h>
 
 struct lsm_network_audit {
 	int netif;
@@ -45,6 +46,16 @@ struct lsm_ioctlop_audit {
 	u16 cmd;
 };
 
+struct lsm_ibpkey_audit {
+	u64	subnet_prefix;
+	u16	pkey;
+};
+
+struct lsm_ibendport_audit {
+	char	dev_name[IB_DEVICE_NAME_MAX];
+	u8	port;
+};
+
 /* Auxiliary data to use in generating the audit record. */
 struct common_audit_data {
 	char type;
@@ -60,6 +71,8 @@ struct common_audit_data {
 #define LSM_AUDIT_DATA_DENTRY	10
 #define LSM_AUDIT_DATA_IOCTL_OP	11
 #define LSM_AUDIT_DATA_FILE	12
+#define LSM_AUDIT_DATA_IBPKEY	13
+#define LSM_AUDIT_DATA_IBENDPORT 14
 	union 	{
 		struct path path;
 		struct dentry *dentry;
@@ -77,6 +90,8 @@ struct common_audit_data {
 		char *kmod_name;
 		struct lsm_ioctlop_audit *op;
 		struct file *file;
+		struct lsm_ibpkey_audit *ibpkey;
+		struct lsm_ibendport_audit *ibendport;
 	} u;
 	/* this union contains LSM specific data */
 	union {

include/linux/lsm_hooks.h

@@ -8,6 +8,7 @@
  * Copyright (C) 2001 Silicon Graphics, Inc. (Trust Technology Group)
  * Copyright (C) 2015 Intel Corporation.
  * Copyright (C) 2015 Casey Schaufler <casey@schaufler-ca.com>
+ * Copyright (C) 2016 Mellanox Techonologies
  *
  *	This program is free software; you can redistribute it and/or modify
  *	it under the terms of the GNU General Public License as published by
@@ -911,6 +912,26 @@
  *	associated with the TUN device's security structure.
  *	@security pointer to the TUN devices's security structure.
  *
+ * Security hooks for Infiniband
+ *
+ * @ib_pkey_access:
+ *	Check permission to access a pkey when modifing a QP.
+ *	@subnet_prefix the subnet prefix of the port being used.
+ *	@pkey the pkey to be accessed.
+ *	@sec pointer to a security structure.
+ * @ib_endport_manage_subnet:
+ *	Check permissions to send and receive SMPs on a end port.
+ *	@dev_name the IB device name (i.e. mlx4_0).
+ *	@port_num the port number.
+ *	@sec pointer to a security structure.
+ * @ib_alloc_security:
+ *	Allocate a security structure for Infiniband objects.
+ *	@sec pointer to a security structure pointer.
+ *	Returns 0 on success, non-zero on failure
+ * @ib_free_security:
+ *	Deallocate an Infiniband security structure.
+ *	@sec contains the security structure to be freed.
+ *
  * Security hooks for XFRM operations.
  *
  * @xfrm_policy_alloc_security:
@@ -1388,7 +1409,9 @@ union security_list_options {
 				unsigned long kern_flags,
 				unsigned long *set_kern_flags);
 	int (*sb_clone_mnt_opts)(const struct super_block *oldsb,
-					struct super_block *newsb);
+					struct super_block *newsb,
+					unsigned long kern_flags,
+					unsigned long *set_kern_flags);
 	int (*sb_parse_opts_str)(char *options, struct security_mnt_opts *opts);
 	int (*dentry_init_security)(struct dentry *dentry, int mode,
 					const struct qstr *name, void **ctx,
@@ -1620,6 +1643,14 @@ union security_list_options {
 	int (*tun_dev_open)(void *security);
 #endif	/* CONFIG_SECURITY_NETWORK */
 
+#ifdef CONFIG_SECURITY_INFINIBAND
+	int (*ib_pkey_access)(void *sec, u64 subnet_prefix, u16 pkey);
+	int (*ib_endport_manage_subnet)(void *sec, const char *dev_name,
+					u8 port_num);
+	int (*ib_alloc_security)(void **sec);
+	void (*ib_free_security)(void *sec);
+#endif	/* CONFIG_SECURITY_INFINIBAND */
+
 #ifdef CONFIG_SECURITY_NETWORK_XFRM
 	int (*xfrm_policy_alloc_security)(struct xfrm_sec_ctx **ctxp,
 					  struct xfrm_user_sec_ctx *sec_ctx,
@@ -1851,6 +1882,12 @@ struct security_hook_heads {
 	struct list_head tun_dev_attach;
 	struct list_head tun_dev_open;
 #endif	/* CONFIG_SECURITY_NETWORK */
+#ifdef CONFIG_SECURITY_INFINIBAND
+	struct list_head ib_pkey_access;
+	struct list_head ib_endport_manage_subnet;
+	struct list_head ib_alloc_security;
+	struct list_head ib_free_security;
+#endif	/* CONFIG_SECURITY_INFINIBAND */
 #ifdef CONFIG_SECURITY_NETWORK_XFRM
 	struct list_head xfrm_policy_alloc_security;
 	struct list_head xfrm_policy_clone_security;

include/linux/security.h

@@ -6,6 +6,7 @@
  * Copyright (C) 2001 Networks Associates Technology, Inc <ssmalley@nai.com>
  * Copyright (C) 2001 James Morris <jmorris@intercode.com.au>
  * Copyright (C) 2001 Silicon Graphics, Inc. (Trust Technology Group)
+ * Copyright (C) 2016 Mellanox Techonologies
  *
  *	This program is free software; you can redistribute it and/or modify
  *	it under the terms of the GNU General Public License as published by
@@ -68,6 +69,10 @@ struct audit_krule;
 struct user_namespace;
 struct timezone;
 
+enum lsm_event {
+	LSM_POLICY_CHANGE,
+};
+
 /* These functions are in security/commoncap.c */
 extern int cap_capable(const struct cred *cred, struct user_namespace *ns,
 		       int cap, int audit);
@@ -163,6 +168,10 @@ struct security_mnt_opts {
 	int num_mnt_opts;
 };
 
+int call_lsm_notifier(enum lsm_event event, void *data);
+int register_lsm_notifier(struct notifier_block *nb);
+int unregister_lsm_notifier(struct notifier_block *nb);
+
 static inline void security_init_mnt_opts(struct security_mnt_opts *opts)
 {
 	opts->mnt_opts = NULL;
@@ -240,7 +249,9 @@ int security_sb_set_mnt_opts(struct super_block *sb,
 				unsigned long kern_flags,
 				unsigned long *set_kern_flags);
 int security_sb_clone_mnt_opts(const struct super_block *oldsb,
-				struct super_block *newsb);
+				struct super_block *newsb,
+				unsigned long kern_flags,
+				unsigned long *set_kern_flags);
 int security_sb_parse_opts_str(char *options, struct security_mnt_opts *opts);
 int security_dentry_init_security(struct dentry *dentry, int mode,
 					const struct qstr *name, void **ctx,
@@ -381,6 +392,21 @@ int security_inode_getsecctx(struct inode *inode, void **ctx, u32 *ctxlen);
 struct security_mnt_opts {
 };
 
+static inline int call_lsm_notifier(enum lsm_event event, void *data)
+{
+	return 0;
+}
+
+static inline int register_lsm_notifier(struct notifier_block *nb)
+{
+	return 0;
+}
+
+static inline  int unregister_lsm_notifier(struct notifier_block *nb)
+{
+	return 0;
+}
+
 static inline void security_init_mnt_opts(struct security_mnt_opts *opts)
 {
 }
@@ -581,7 +607,9 @@ static inline int security_sb_set_mnt_opts(struct super_block *sb,
 }
 
 static inline int security_sb_clone_mnt_opts(const struct super_block *oldsb,
-					      struct super_block *newsb)
+					      struct super_block *newsb,
+					      unsigned long kern_flags,
+					      unsigned long *set_kern_flags)
 {
 	return 0;
 }
@@ -1406,6 +1434,32 @@ static inline int security_tun_dev_open(void *security)
 }
 #endif	/* CONFIG_SECURITY_NETWORK */
 
+#ifdef CONFIG_SECURITY_INFINIBAND
+int security_ib_pkey_access(void *sec, u64 subnet_prefix, u16 pkey);
+int security_ib_endport_manage_subnet(void *sec, const char *name, u8 port_num);
+int security_ib_alloc_security(void **sec);
+void security_ib_free_security(void *sec);
+#else	/* CONFIG_SECURITY_INFINIBAND */
+static inline int security_ib_pkey_access(void *sec, u64 subnet_prefix, u16 pkey)
+{
+	return 0;
+}
+
+static inline int security_ib_endport_manage_subnet(void *sec, const char *dev_name, u8 port_num)
+{
+	return 0;
+}
+
+static inline int security_ib_alloc_security(void **sec)
+{
+	return 0;
+}
+
+static inline void security_ib_free_security(void *sec)
+{
+}
+#endif	/* CONFIG_SECURITY_INFINIBAND */
+
 #ifdef CONFIG_SECURITY_NETWORK_XFRM
 
 int security_xfrm_policy_alloc(struct xfrm_sec_ctx **ctxp,

include/rdma/ib_mad.h

@@ -575,6 +575,10 @@ struct ib_mad_agent {
 	u32			flags;
 	u8			port_num;
 	u8			rmpp_version;
+	void			*security;
+	bool			smp_allowed;
+	bool			lsm_nb_reg;
+	struct notifier_block   lsm_nb;
 };
 
 /**

include/rdma/ib_verbs.h

@@ -1614,6 +1614,45 @@ struct ib_rwq_ind_table_init_attr {
 	struct ib_wq	**ind_tbl;
 };
 
+enum port_pkey_state {
+	IB_PORT_PKEY_NOT_VALID = 0,
+	IB_PORT_PKEY_VALID = 1,
+	IB_PORT_PKEY_LISTED = 2,
+};
+
+struct ib_qp_security;
+
+struct ib_port_pkey {
+	enum port_pkey_state	state;
+	u16			pkey_index;
+	u8			port_num;
+	struct list_head	qp_list;
+	struct list_head	to_error_list;
+	struct ib_qp_security  *sec;
+};
+
+struct ib_ports_pkeys {
+	struct ib_port_pkey	main;
+	struct ib_port_pkey	alt;
+};
+
+struct ib_qp_security {
+	struct ib_qp	       *qp;
+	struct ib_device       *dev;
+	/* Hold this mutex when changing port and pkey settings. */
+	struct mutex		mutex;
+	struct ib_ports_pkeys  *ports_pkeys;
+	/* A list of all open shared QP handles.  Required to enforce security
+	 * properly for all users of a shared QP.
+	 */
+	struct list_head        shared_qp_list;
+	void                   *security;
+	bool			destroying;
+	atomic_t		error_list_count;
+	struct completion	error_complete;
+	int			error_comps_pending;
+};
+
 /*
  * @max_write_sge: Maximum SGE elements per RDMA WRITE request.
  * @max_read_sge:  Maximum SGE elements per RDMA READ request.
@@ -1643,6 +1682,7 @@ struct ib_qp {
 	u32			max_read_sge;
 	enum ib_qp_type		qp_type;
 	struct ib_rwq_ind_table *rwq_ind_tbl;
+	struct ib_qp_security  *qp_sec;
 };
 
 struct ib_mr {
@@ -1891,6 +1931,7 @@ enum ib_mad_result {
 };
 
 struct ib_port_cache {
+	u64		      subnet_prefix;
 	struct ib_pkey_cache  *pkey;
 	struct ib_gid_table   *gid;
 	u8                     lmc;
@@ -1940,6 +1981,12 @@ struct rdma_netdev {
 			    union ib_gid *gid, u16 mlid);
 };
 
+struct ib_port_pkey_list {
+	/* Lock to hold while modifying the list. */
+	spinlock_t                    list_lock;
+	struct list_head              pkey_list;
+};
+
 struct ib_device {
 	/* Do not access @dma_device directly from ULP nor from HW drivers. */
 	struct device                *dma_device;
@@ -1963,6 +2010,8 @@ struct ib_device {
 
 	int			      num_comp_vectors;
 
+	struct ib_port_pkey_list     *port_pkey_list;
+
 	struct iw_cm_verbs	     *iwcm;
 
 	/**

security/Kconfig

@@ -54,6 +54,15 @@ config SECURITY_NETWORK
 	  implement socket and networking access controls.
 	  If you are unsure how to answer this question, answer N.
 
+config SECURITY_INFINIBAND
+	bool "Infiniband Security Hooks"
+	depends on SECURITY && INFINIBAND
+	help
+	  This enables the Infiniband security hooks.
+	  If enabled, a security module can use these hooks to
+	  implement Infiniband access controls.
+	  If you are unsure how to answer this question, answer N.
+
 config SECURITY_NETWORK_XFRM
 	bool "XFRM (IPSec) Networking Security Hooks"
 	depends on XFRM && SECURITY_NETWORK

security/lsm_audit.c

@@ -410,6 +410,22 @@ static void dump_common_audit_data(struct audit_buffer *ab,
 		audit_log_format(ab, " kmod=");
 		audit_log_untrustedstring(ab, a->u.kmod_name);
 		break;
+	case LSM_AUDIT_DATA_IBPKEY: {
+		struct in6_addr sbn_pfx;
+
+		memset(&sbn_pfx.s6_addr, 0,
+		       sizeof(sbn_pfx.s6_addr));
+		memcpy(&sbn_pfx.s6_addr, &a->u.ibpkey->subnet_prefix,
+		       sizeof(a->u.ibpkey->subnet_prefix));
+		audit_log_format(ab, " pkey=0x%x subnet_prefix=%pI6c",
+				 a->u.ibpkey->pkey, &sbn_pfx);
+		break;
+	}
+	case LSM_AUDIT_DATA_IBENDPORT:
+		audit_log_format(ab, " device=%s port_num=%u",
+				 a->u.ibendport->dev_name,
+				 a->u.ibendport->port);
+		break;
 	} /* switch (a->type) */
 }
 

security/security.c

@@ -4,6 +4,7 @@
  * Copyright (C) 2001 WireX Communications, Inc <chris@wirex.com>
  * Copyright (C) 2001-2002 Greg Kroah-Hartman <greg@kroah.com>
  * Copyright (C) 2001 Networks Associates Technology, Inc <ssmalley@nai.com>
+ * Copyright (C) 2016 Mellanox Technologies
  *
  *	This program is free software; you can redistribute it and/or modify
  *	it under the terms of the GNU General Public License as published by
@@ -34,6 +35,8 @@
 #define SECURITY_NAME_MAX	10
 
 struct security_hook_heads security_hook_heads __lsm_ro_after_init;
+static ATOMIC_NOTIFIER_HEAD(lsm_notifier_chain);
+
 char *lsm_names;
 /* Boot-time LSM user choice */
 static __initdata char chosen_lsm[SECURITY_NAME_MAX + 1] =
@@ -165,6 +168,24 @@ void __init security_add_hooks(struct security_hook_list *hooks, int count,
 		panic("%s - Cannot get early memory.\n", __func__);
 }
 
+int call_lsm_notifier(enum lsm_event event, void *data)
+{
+	return atomic_notifier_call_chain(&lsm_notifier_chain, event, data);
+}
+EXPORT_SYMBOL(call_lsm_notifier);
+
+int register_lsm_notifier(struct notifier_block *nb)
+{
+	return atomic_notifier_chain_register(&lsm_notifier_chain, nb);
+}
+EXPORT_SYMBOL(register_lsm_notifier);
+
+int unregister_lsm_notifier(struct notifier_block *nb)
+{
+	return atomic_notifier_chain_unregister(&lsm_notifier_chain, nb);
+}
+EXPORT_SYMBOL(unregister_lsm_notifier);
+
 /*
  * Hook list operation macros.
  *
@@ -399,9 +420,12 @@ int security_sb_set_mnt_opts(struct super_block *sb,
 EXPORT_SYMBOL(security_sb_set_mnt_opts);
 
 int security_sb_clone_mnt_opts(const struct super_block *oldsb,
-				struct super_block *newsb)
+				struct super_block *newsb,
+				unsigned long kern_flags,
+				unsigned long *set_kern_flags)
 {
-	return call_int_hook(sb_clone_mnt_opts, 0, oldsb, newsb);
+	return call_int_hook(sb_clone_mnt_opts, 0, oldsb, newsb,
+				kern_flags, set_kern_flags);
 }
 EXPORT_SYMBOL(security_sb_clone_mnt_opts);
 
@@ -1515,6 +1539,33 @@ EXPORT_SYMBOL(security_tun_dev_open);
 
 #endif	/* CONFIG_SECURITY_NETWORK */
 
+#ifdef CONFIG_SECURITY_INFINIBAND
+
+int security_ib_pkey_access(void *sec, u64 subnet_prefix, u16 pkey)
+{
+	return call_int_hook(ib_pkey_access, 0, sec, subnet_prefix, pkey);
+}
+EXPORT_SYMBOL(security_ib_pkey_access);
+
+int security_ib_endport_manage_subnet(void *sec, const char *dev_name, u8 port_num)
+{
+	return call_int_hook(ib_endport_manage_subnet, 0, sec, dev_name, port_num);
+}
+EXPORT_SYMBOL(security_ib_endport_manage_subnet);
+
+int security_ib_alloc_security(void **sec)
+{
+	return call_int_hook(ib_alloc_security, 0, sec);
+}
+EXPORT_SYMBOL(security_ib_alloc_security);
+
+void security_ib_free_security(void *sec)
+{
+	call_void_hook(ib_free_security, sec);
+}
+EXPORT_SYMBOL(security_ib_free_security);
+#endif	/* CONFIG_SECURITY_INFINIBAND */
+
 #ifdef CONFIG_SECURITY_NETWORK_XFRM
 
 int security_xfrm_policy_alloc(struct xfrm_sec_ctx **ctxp,

security/selinux/Makefile

@@ -5,7 +5,7 @@
 obj-$(CONFIG_SECURITY_SELINUX) := selinux.o
 
 selinux-y := avc.o hooks.o selinuxfs.o netlink.o nlmsgtab.o netif.o \
-	     netnode.o netport.o exports.o \
+	     netnode.o netport.o ibpkey.o exports.o \
 	     ss/ebitmap.o ss/hashtab.o ss/symtab.o ss/sidtab.o ss/avtab.o \
 	     ss/policydb.o ss/services.o ss/conditional.o ss/mls.o ss/status.o
 

security/selinux/hooks.c

@@ -17,6 +17,7 @@
  *	Paul Moore <paul@paul-moore.com>
  *  Copyright (C) 2007 Hitachi Software Engineering Co., Ltd.
  *		       Yuichi Nakamura <ynakam@hitachisoft.jp>
+ *  Copyright (C) 2016 Mellanox Technologies
  *
  *	This program is free software; you can redistribute it and/or modify
  *	it under the terms of the GNU General Public License version 2,
@@ -90,6 +91,7 @@
 #include "netif.h"
 #include "netnode.h"
 #include "netport.h"
+#include "ibpkey.h"
 #include "xfrm.h"
 #include "netlabel.h"
 #include "audit.h"
@@ -171,6 +173,16 @@ static int selinux_netcache_avc_callback(u32 event)
 	return 0;
 }
 
+static int selinux_lsm_notifier_avc_callback(u32 event)
+{
+	if (event == AVC_CALLBACK_RESET) {
+		sel_ib_pkey_flush();
+		call_lsm_notifier(LSM_POLICY_CHANGE, NULL);
+	}
+
+	return 0;
+}
+
 /*
  * initialise the security for the init task
  */
@@ -398,18 +410,6 @@ static void superblock_free_security(struct super_block *sb)
 	kfree(sbsec);
 }
 
-/* The file system's label must be initialized prior to use. */
-
-static const char *labeling_behaviors[7] = {
-	"uses xattr",
-	"uses transition SIDs",
-	"uses task SIDs",
-	"uses genfs_contexts",
-	"not configured for labeling",
-	"uses mountpoint labeling",
-	"uses native labeling",
-};
-
 static inline int inode_doinit(struct inode *inode)
 {
 	return inode_doinit_with_dentry(inode, NULL);
@@ -524,13 +524,17 @@ static int sb_finish_set_opts(struct super_block *sb)
 		}
 	}
 
-	if (sbsec->behavior > ARRAY_SIZE(labeling_behaviors))
-		printk(KERN_ERR "SELinux: initialized (dev %s, type %s), unknown behavior\n",
-		       sb->s_id, sb->s_type->name);
-
 	sbsec->flags |= SE_SBINITIALIZED;
+
+	/*
+	 * Explicitly set or clear SBLABEL_MNT.  It's not sufficient to simply
+	 * leave the flag untouched because sb_clone_mnt_opts might be handing
+	 * us a superblock that needs the flag to be cleared.
+	 */
 	if (selinux_is_sblabel_mnt(sb))
 		sbsec->flags |= SBLABEL_MNT;
+	else
+		sbsec->flags &= ~SBLABEL_MNT;
 
 	/* Initialize the root inode. */
 	rc = inode_doinit_with_dentry(root_inode, root);
@@ -809,6 +813,7 @@ static int selinux_set_mnt_opts(struct super_block *sb,
 		sbsec->flags |= SE_SBPROC | SE_SBGENFS;
 
 	if (!strcmp(sb->s_type->name, "debugfs") ||
+	    !strcmp(sb->s_type->name, "tracefs") ||
 	    !strcmp(sb->s_type->name, "sysfs") ||
 	    !strcmp(sb->s_type->name, "pstore"))
 		sbsec->flags |= SE_SBGENFS;
@@ -963,8 +968,11 @@ mismatch:
 }
 
 static int selinux_sb_clone_mnt_opts(const struct super_block *oldsb,
-					struct super_block *newsb)
+					struct super_block *newsb,
+					unsigned long kern_flags,
+					unsigned long *set_kern_flags)
 {
+	int rc = 0;
 	const struct superblock_security_struct *oldsbsec = oldsb->s_security;
 	struct superblock_security_struct *newsbsec = newsb->s_security;
 
@@ -979,6 +987,13 @@ static int selinux_sb_clone_mnt_opts(const struct super_block *oldsb,
 	if (!ss_initialized)
 		return 0;
 
+	/*
+	 * Specifying internal flags without providing a place to
+	 * place the results is not allowed.
+	 */
+	if (kern_flags && !set_kern_flags)
+		return -EINVAL;
+
 	/* how can we clone if the old one wasn't set up?? */
 	BUG_ON(!(oldsbsec->flags & SE_SBINITIALIZED));
 
@@ -994,6 +1009,18 @@ static int selinux_sb_clone_mnt_opts(const struct super_block *oldsb,
 	newsbsec->def_sid = oldsbsec->def_sid;
 	newsbsec->behavior = oldsbsec->behavior;
 
+	if (newsbsec->behavior == SECURITY_FS_USE_NATIVE &&
+		!(kern_flags & SECURITY_LSM_NATIVE_LABELS) && !set_context) {
+		rc = security_fs_use(newsb);
+		if (rc)
+			goto out;
+	}
+
+	if (kern_flags & SECURITY_LSM_NATIVE_LABELS && !set_context) {
+		newsbsec->behavior = SECURITY_FS_USE_NATIVE;
+		*set_kern_flags |= SECURITY_LSM_NATIVE_LABELS;
+	}
+
 	if (set_context) {
 		u32 sid = oldsbsec->mntpoint_sid;
 
@@ -1013,8 +1040,9 @@ static int selinux_sb_clone_mnt_opts(const struct super_block *oldsb,
 	}
 
 	sb_finish_set_opts(newsb);
+out:
 	mutex_unlock(&newsbsec->lock);
-	return 0;
+	return rc;
 }
 
 static int selinux_parse_opts_str(char *options,
@@ -2063,8 +2091,9 @@ static inline u32 file_to_av(struct file *file)
 static inline u32 open_file_to_av(struct file *file)
 {
 	u32 av = file_to_av(file);
+	struct inode *inode = file_inode(file);
 
-	if (selinux_policycap_openperm)
+	if (selinux_policycap_openperm && inode->i_sb->s_magic != SOCKFS_MAGIC)
 		av |= FILE__OPEN;
 
 	return av;
@@ -3059,6 +3088,7 @@ static int selinux_inode_permission(struct inode *inode, int mask)
 static int selinux_inode_setattr(struct dentry *dentry, struct iattr *iattr)
 {
 	const struct cred *cred = current_cred();
+	struct inode *inode = d_backing_inode(dentry);
 	unsigned int ia_valid = iattr->ia_valid;
 	__u32 av = FILE__WRITE;
 
@@ -3074,8 +3104,10 @@ static int selinux_inode_setattr(struct dentry *dentry, struct iattr *iattr)
 			ATTR_ATIME_SET | ATTR_MTIME_SET | ATTR_TIMES_SET))
 		return dentry_has_perm(cred, dentry, FILE__SETATTR);
 
-	if (selinux_policycap_openperm && (ia_valid & ATTR_SIZE)
-			&& !(ia_valid & ATTR_FILE))
+	if (selinux_policycap_openperm &&
+	    inode->i_sb->s_magic != SOCKFS_MAGIC &&
+	    (ia_valid & ATTR_SIZE) &&
+	    !(ia_valid & ATTR_FILE))
 		av |= FILE__OPEN;
 
 	return dentry_has_perm(cred, dentry, av);
@@ -3107,6 +3139,18 @@ static int selinux_inode_setotherxattr(struct dentry *dentry, const char *name)
 	return dentry_has_perm(cred, dentry, FILE__SETATTR);
 }
 
+static bool has_cap_mac_admin(bool audit)
+{
+	const struct cred *cred = current_cred();
+	int cap_audit = audit ? SECURITY_CAP_AUDIT : SECURITY_CAP_NOAUDIT;
+
+	if (cap_capable(cred, &init_user_ns, CAP_MAC_ADMIN, cap_audit))
+		return false;
+	if (cred_has_capability(cred, CAP_MAC_ADMIN, cap_audit, true))
+		return false;
+	return true;
+}
+
 static int selinux_inode_setxattr(struct dentry *dentry, const char *name,
 				  const void *value, size_t size, int flags)
 {
@@ -3138,7 +3182,7 @@ static int selinux_inode_setxattr(struct dentry *dentry, const char *name,
 
 	rc = security_context_to_sid(value, size, &newsid, GFP_KERNEL);
 	if (rc == -EINVAL) {
-		if (!capable(CAP_MAC_ADMIN)) {
+		if (!has_cap_mac_admin(true)) {
 			struct audit_buffer *ab;
 			size_t audit_size;
 			const char *str;
@@ -3264,13 +3308,8 @@ static int selinux_inode_getsecurity(struct inode *inode, const char *name, void
 	 * and lack of permission just means that we fall back to the
 	 * in-core context value, not a denial.
 	 */
-	error = cap_capable(current_cred(), &init_user_ns, CAP_MAC_ADMIN,
-			    SECURITY_CAP_NOAUDIT);
-	if (!error)
-		error = cred_has_capability(current_cred(), CAP_MAC_ADMIN,
-					    SECURITY_CAP_NOAUDIT, true);
 	isec = inode_security(inode);
-	if (!error)
+	if (has_cap_mac_admin(false))
 		error = security_sid_to_context_force(isec->sid, &context,
 						      &size);
 	else
@@ -3550,6 +3589,18 @@ static int selinux_mmap_addr(unsigned long addr)
 static int selinux_mmap_file(struct file *file, unsigned long reqprot,
 			     unsigned long prot, unsigned long flags)
 {
+	struct common_audit_data ad;
+	int rc;
+
+	if (file) {
+		ad.type = LSM_AUDIT_DATA_FILE;
+		ad.u.file = file;
+		rc = inode_has_perm(current_cred(), file_inode(file),
+				    FILE__MAP, &ad);
+		if (rc)
+			return rc;
+	}
+
 	if (selinux_checkreqprot)
 		prot = reqprot;
 
@@ -3710,7 +3761,8 @@ static int selinux_file_open(struct file *file, const struct cred *cred)
 
 /* task security operations */
 
-static int selinux_task_create(unsigned long clone_flags)
+static int selinux_task_alloc(struct task_struct *task,
+			      unsigned long clone_flags)
 {
 	u32 sid = current_sid();
 
@@ -5918,7 +5970,7 @@ static int selinux_setprocattr(const char *name, void *value, size_t size)
 		}
 		error = security_context_to_sid(value, size, &sid, GFP_KERNEL);
 		if (error == -EINVAL && !strcmp(name, "fscreate")) {
-			if (!capable(CAP_MAC_ADMIN)) {
+			if (!has_cap_mac_admin(true)) {
 				struct audit_buffer *ab;
 				size_t audit_size;
 
@@ -6128,7 +6180,70 @@ static int selinux_key_getsecurity(struct key *key, char **_buffer)
 	*_buffer = context;
 	return rc;
 }
+#endif
 
+#ifdef CONFIG_SECURITY_INFINIBAND
+static int selinux_ib_pkey_access(void *ib_sec, u64 subnet_prefix, u16 pkey_val)
+{
+	struct common_audit_data ad;
+	int err;
+	u32 sid = 0;
+	struct ib_security_struct *sec = ib_sec;
+	struct lsm_ibpkey_audit ibpkey;
+
+	err = sel_ib_pkey_sid(subnet_prefix, pkey_val, &sid);
+	if (err)
+		return err;
+
+	ad.type = LSM_AUDIT_DATA_IBPKEY;
+	ibpkey.subnet_prefix = subnet_prefix;
+	ibpkey.pkey = pkey_val;
+	ad.u.ibpkey = &ibpkey;
+	return avc_has_perm(sec->sid, sid,
+			    SECCLASS_INFINIBAND_PKEY,
+			    INFINIBAND_PKEY__ACCESS, &ad);
+}
+
+static int selinux_ib_endport_manage_subnet(void *ib_sec, const char *dev_name,
+					    u8 port_num)
+{
+	struct common_audit_data ad;
+	int err;
+	u32 sid = 0;
+	struct ib_security_struct *sec = ib_sec;
+	struct lsm_ibendport_audit ibendport;
+
+	err = security_ib_endport_sid(dev_name, port_num, &sid);
+
+	if (err)
+		return err;
+
+	ad.type = LSM_AUDIT_DATA_IBENDPORT;
+	strncpy(ibendport.dev_name, dev_name, sizeof(ibendport.dev_name));
+	ibendport.port = port_num;
+	ad.u.ibendport = &ibendport;
+	return avc_has_perm(sec->sid, sid,
+			    SECCLASS_INFINIBAND_ENDPORT,
+			    INFINIBAND_ENDPORT__MANAGE_SUBNET, &ad);
+}
+
+static int selinux_ib_alloc_security(void **ib_sec)
+{
+	struct ib_security_struct *sec;
+
+	sec = kzalloc(sizeof(*sec), GFP_KERNEL);
+	if (!sec)
+		return -ENOMEM;
+	sec->sid = current_sid();
+
+	*ib_sec = sec;
+	return 0;
+}
+
+static void selinux_ib_free_security(void *ib_sec)
+{
+	kfree(ib_sec);
+}
 #endif
 
 static struct security_hook_list selinux_hooks[] __lsm_ro_after_init = {
@@ -6213,7 +6328,7 @@ static struct security_hook_list selinux_hooks[] __lsm_ro_after_init = {
 
 	LSM_HOOK_INIT(file_open, selinux_file_open),
 
-	LSM_HOOK_INIT(task_create, selinux_task_create),
+	LSM_HOOK_INIT(task_alloc, selinux_task_alloc),
 	LSM_HOOK_INIT(cred_alloc_blank, selinux_cred_alloc_blank),
 	LSM_HOOK_INIT(cred_free, selinux_cred_free),
 	LSM_HOOK_INIT(cred_prepare, selinux_cred_prepare),
@@ -6315,7 +6430,13 @@ static struct security_hook_list selinux_hooks[] __lsm_ro_after_init = {
 	LSM_HOOK_INIT(tun_dev_attach_queue, selinux_tun_dev_attach_queue),
 	LSM_HOOK_INIT(tun_dev_attach, selinux_tun_dev_attach),
 	LSM_HOOK_INIT(tun_dev_open, selinux_tun_dev_open),
-
+#ifdef CONFIG_SECURITY_INFINIBAND
+	LSM_HOOK_INIT(ib_pkey_access, selinux_ib_pkey_access),
+	LSM_HOOK_INIT(ib_endport_manage_subnet,
+		      selinux_ib_endport_manage_subnet),
+	LSM_HOOK_INIT(ib_alloc_security, selinux_ib_alloc_security),
+	LSM_HOOK_INIT(ib_free_security, selinux_ib_free_security),
+#endif
 #ifdef CONFIG_SECURITY_NETWORK_XFRM
 	LSM_HOOK_INIT(xfrm_policy_alloc_security, selinux_xfrm_policy_alloc),
 	LSM_HOOK_INIT(xfrm_policy_clone_security, selinux_xfrm_policy_clone),
@@ -6379,6 +6500,9 @@ static __init int selinux_init(void)
 	if (avc_add_callback(selinux_netcache_avc_callback, AVC_CALLBACK_RESET))
 		panic("SELinux: Unable to register AVC netcache callback\n");
 
+	if (avc_add_callback(selinux_lsm_notifier_avc_callback, AVC_CALLBACK_RESET))
+		panic("SELinux: Unable to register AVC LSM notifier callback\n");
+
 	if (selinux_enforcing)
 		printk(KERN_DEBUG "SELinux:  Starting in enforcing mode\n");
 	else
@@ -6448,6 +6572,23 @@ static struct nf_hook_ops selinux_nf_ops[] = {
 #endif	/* IPV6 */
 };
 
+static int __net_init selinux_nf_register(struct net *net)
+{
+	return nf_register_net_hooks(net, selinux_nf_ops,
+				     ARRAY_SIZE(selinux_nf_ops));
+}
+
+static void __net_exit selinux_nf_unregister(struct net *net)
+{
+	nf_unregister_net_hooks(net, selinux_nf_ops,
+				ARRAY_SIZE(selinux_nf_ops));
+}
+
+static struct pernet_operations selinux_net_ops = {
+	.init = selinux_nf_register,
+	.exit = selinux_nf_unregister,
+};
+
 static int __init selinux_nf_ip_init(void)
 {
 	int err;
@@ -6457,13 +6598,12 @@ static int __init selinux_nf_ip_init(void)
 
 	printk(KERN_DEBUG "SELinux:  Registering netfilter hooks\n");
 
-	err = nf_register_hooks(selinux_nf_ops, ARRAY_SIZE(selinux_nf_ops));
+	err = register_pernet_subsys(&selinux_net_ops);
 	if (err)
-		panic("SELinux: nf_register_hooks: error %d\n", err);
+		panic("SELinux: register_pernet_subsys: error %d\n", err);
 
 	return 0;
 }
-
 __initcall(selinux_nf_ip_init);
 
 #ifdef CONFIG_SECURITY_SELINUX_DISABLE
@@ -6471,7 +6611,7 @@ static void selinux_nf_ip_exit(void)
 {
 	printk(KERN_DEBUG "SELinux:  Unregistering netfilter hooks\n");
 
-	nf_unregister_hooks(selinux_nf_ops, ARRAY_SIZE(selinux_nf_ops));
+	unregister_pernet_subsys(&selinux_net_ops);
 }
 #endif
 

security/selinux/ibpkey.c

@@ -0,0 +1,245 @@
+/*
+ * Pkey table
+ *
+ * SELinux must keep a mapping of Infinband PKEYs to labels/SIDs.  This
+ * mapping is maintained as part of the normal policy but a fast cache is
+ * needed to reduce the lookup overhead.
+ *
+ * This code is heavily based on the "netif" and "netport" concept originally
+ * developed by
+ * James Morris <jmorris@redhat.com> and
+ * Paul Moore <paul@paul-moore.com>
+ *   (see security/selinux/netif.c and security/selinux/netport.c for more
+ *   information)
+ *
+ */
+
+/*
+ * (c) Mellanox Technologies, 2016
+ *
+ * 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.
+ *
+ */
+
+#include <linux/types.h>
+#include <linux/rcupdate.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+
+#include "ibpkey.h"
+#include "objsec.h"
+
+#define SEL_PKEY_HASH_SIZE       256
+#define SEL_PKEY_HASH_BKT_LIMIT   16
+
+struct sel_ib_pkey_bkt {
+	int size;
+	struct list_head list;
+};
+
+struct sel_ib_pkey {
+	struct pkey_security_struct psec;
+	struct list_head list;
+	struct rcu_head rcu;
+};
+
+static LIST_HEAD(sel_ib_pkey_list);
+static DEFINE_SPINLOCK(sel_ib_pkey_lock);
+static struct sel_ib_pkey_bkt sel_ib_pkey_hash[SEL_PKEY_HASH_SIZE];
+
+/**
+ * sel_ib_pkey_hashfn - Hashing function for the pkey table
+ * @pkey: pkey number
+ *
+ * Description:
+ * This is the hashing function for the pkey table, it returns the bucket
+ * number for the given pkey.
+ *
+ */
+static unsigned int sel_ib_pkey_hashfn(u16 pkey)
+{
+	return (pkey & (SEL_PKEY_HASH_SIZE - 1));
+}
+
+/**
+ * sel_ib_pkey_find - Search for a pkey record
+ * @subnet_prefix: subnet_prefix
+ * @pkey_num: pkey_num
+ *
+ * Description:
+ * Search the pkey table and return the matching record.  If an entry
+ * can not be found in the table return NULL.
+ *
+ */
+static struct sel_ib_pkey *sel_ib_pkey_find(u64 subnet_prefix, u16 pkey_num)
+{
+	unsigned int idx;
+	struct sel_ib_pkey *pkey;
+
+	idx = sel_ib_pkey_hashfn(pkey_num);
+	list_for_each_entry_rcu(pkey, &sel_ib_pkey_hash[idx].list, list) {
+		if (pkey->psec.pkey == pkey_num &&
+		    pkey->psec.subnet_prefix == subnet_prefix)
+			return pkey;
+	}
+
+	return NULL;
+}
+
+/**
+ * sel_ib_pkey_insert - Insert a new pkey into the table
+ * @pkey: the new pkey record
+ *
+ * Description:
+ * Add a new pkey record to the hash table.
+ *
+ */
+static void sel_ib_pkey_insert(struct sel_ib_pkey *pkey)
+{
+	unsigned int idx;
+
+	/* we need to impose a limit on the growth of the hash table so check
+	 * this bucket to make sure it is within the specified bounds
+	 */
+	idx = sel_ib_pkey_hashfn(pkey->psec.pkey);
+	list_add_rcu(&pkey->list, &sel_ib_pkey_hash[idx].list);
+	if (sel_ib_pkey_hash[idx].size == SEL_PKEY_HASH_BKT_LIMIT) {
+		struct sel_ib_pkey *tail;
+
+		tail = list_entry(
+			rcu_dereference_protected(
+				sel_ib_pkey_hash[idx].list.prev,
+				lockdep_is_held(&sel_ib_pkey_lock)),
+			struct sel_ib_pkey, list);
+		list_del_rcu(&tail->list);
+		kfree_rcu(tail, rcu);
+	} else {
+		sel_ib_pkey_hash[idx].size++;
+	}
+}
+
+/**
+ * sel_ib_pkey_sid_slow - Lookup the SID of a pkey using the policy
+ * @subnet_prefix: subnet prefix
+ * @pkey_num: pkey number
+ * @sid: pkey SID
+ *
+ * Description:
+ * This function determines the SID of a pkey by querying the security
+ * policy.  The result is added to the pkey table to speedup future
+ * queries.  Returns zero on success, negative values on failure.
+ *
+ */
+static int sel_ib_pkey_sid_slow(u64 subnet_prefix, u16 pkey_num, u32 *sid)
+{
+	int ret;
+	struct sel_ib_pkey *pkey;
+	struct sel_ib_pkey *new = NULL;
+	unsigned long flags;
+
+	spin_lock_irqsave(&sel_ib_pkey_lock, flags);
+	pkey = sel_ib_pkey_find(subnet_prefix, pkey_num);
+	if (pkey) {
+		*sid = pkey->psec.sid;
+		spin_unlock_irqrestore(&sel_ib_pkey_lock, flags);
+		return 0;
+	}
+
+	ret = security_ib_pkey_sid(subnet_prefix, pkey_num, sid);
+	if (ret)
+		goto out;
+
+	/* If this memory allocation fails still return 0. The SID
+	 * is valid, it just won't be added to the cache.
+	 */
+	new = kzalloc(sizeof(*new), GFP_ATOMIC);
+	if (!new)
+		goto out;
+
+	new->psec.subnet_prefix = subnet_prefix;
+	new->psec.pkey = pkey_num;
+	new->psec.sid = *sid;
+	sel_ib_pkey_insert(new);
+
+out:
+	spin_unlock_irqrestore(&sel_ib_pkey_lock, flags);
+	return ret;
+}
+
+/**
+ * sel_ib_pkey_sid - Lookup the SID of a PKEY
+ * @subnet_prefix: subnet_prefix
+ * @pkey_num: pkey number
+ * @sid: pkey SID
+ *
+ * Description:
+ * This function determines the SID of a PKEY using the fastest method
+ * possible.  First the pkey table is queried, but if an entry can't be found
+ * then the policy is queried and the result is added to the table to speedup
+ * future queries.  Returns zero on success, negative values on failure.
+ *
+ */
+int sel_ib_pkey_sid(u64 subnet_prefix, u16 pkey_num, u32 *sid)
+{
+	struct sel_ib_pkey *pkey;
+
+	rcu_read_lock();
+	pkey = sel_ib_pkey_find(subnet_prefix, pkey_num);
+	if (pkey) {
+		*sid = pkey->psec.sid;
+		rcu_read_unlock();
+		return 0;
+	}
+	rcu_read_unlock();
+
+	return sel_ib_pkey_sid_slow(subnet_prefix, pkey_num, sid);
+}
+
+/**
+ * sel_ib_pkey_flush - Flush the entire pkey table
+ *
+ * Description:
+ * Remove all entries from the pkey table
+ *
+ */
+void sel_ib_pkey_flush(void)
+{
+	unsigned int idx;
+	struct sel_ib_pkey *pkey, *pkey_tmp;
+	unsigned long flags;
+
+	spin_lock_irqsave(&sel_ib_pkey_lock, flags);
+	for (idx = 0; idx < SEL_PKEY_HASH_SIZE; idx++) {
+		list_for_each_entry_safe(pkey, pkey_tmp,
+					 &sel_ib_pkey_hash[idx].list, list) {
+			list_del_rcu(&pkey->list);
+			kfree_rcu(pkey, rcu);
+		}
+		sel_ib_pkey_hash[idx].size = 0;
+	}
+	spin_unlock_irqrestore(&sel_ib_pkey_lock, flags);
+}
+
+static __init int sel_ib_pkey_init(void)
+{
+	int iter;
+
+	if (!selinux_enabled)
+		return 0;
+
+	for (iter = 0; iter < SEL_PKEY_HASH_SIZE; iter++) {
+		INIT_LIST_HEAD(&sel_ib_pkey_hash[iter].list);
+		sel_ib_pkey_hash[iter].size = 0;
+	}
+
+	return 0;
+}
+
+subsys_initcall(sel_ib_pkey_init);