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

Pull rdma updates from Jason Gunthorpe:
 "This has been a smaller cycle with many of the commits being smallish
  code fixes and improvements across the drivers.

   - Driver updates for bnxt_re, cxgb4, hfi1, hns, mlx5, nes, qedr, and
     rxe

   - Memory window support in hns

   - mlx5 user API 'flow mutate/steering' allows accessing the full
     packet mangling and matching machinery from user space

   - Support inter-working with verbs API calls in the 'devx' mlx5 user
     API, and provide options to use devx with less privilege

   - Modernize the use of syfs and the device interface to use attribute
     groups and cdev properly for uverbs, and clean up some of the core
     code's device list management

   - More progress on net namespaces for RDMA devices

   - Consolidate driver BAR mmapping support into core code helpers and
     rework how RDMA holds poitners to mm_struct for get_user_pages
     cases

   - First pass to use 'dev_name' instead of ib_device->name

   - Device renaming for RDMA devices"

* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma: (242 commits)
  IB/mlx5: Add support for extended atomic operations
  RDMA/core: Fix comment for hw stats init for port == 0
  RDMA/core: Refactor ib_register_device() function
  RDMA/core: Fix unwinding flow in case of error to register device
  ib_srp: Remove WARN_ON in srp_terminate_io()
  IB/mlx5: Allow scatter to CQE without global signaled WRs
  IB/mlx5: Verify that driver supports user flags
  IB/mlx5: Support scatter to CQE for DC transport type
  RDMA/drivers: Use core provided API for registering device attributes
  RDMA/core: Allow existing drivers to set one sysfs group per device
  IB/rxe: Remove unnecessary enum values
  RDMA/umad: Use kernel API to allocate umad indexes
  RDMA/uverbs: Use kernel API to allocate uverbs indexes
  RDMA/core: Increase total number of RDMA ports across all devices
  IB/mlx4: Add port and TID to MAD debug print
  IB/mlx4: Enable debug print of SMPs
  RDMA/core: Rename ports_parent to ports_kobj
  RDMA/core: Do not expose unsupported counters
  IB/mlx4: Refer to the device kobject instead of ports_parent
  RDMA/nldev: Allow IB device rename through RDMA netlink
  ...

13 files changed, 385 insertions, 146 deletions

diff --git a/include/rdma/ib_addr.h b/include/rdma/ib_addr.h
index 77c7908b7d73..2734c895c1bf 100644
--- a/include/rdma/ib_addr.h
+++ b/include/rdma/ib_addr.h
@@ -46,7 +46,6 @@
 #include <net/ip.h>
 #include <rdma/ib_verbs.h>
 #include <rdma/ib_pack.h>
-#include <net/ipv6.h>
 #include <net/net_namespace.h>
 
 /**
@@ -95,20 +94,18 @@ int rdma_translate_ip(const struct sockaddr *addr,
  * @timeout_ms: Amount of time to wait for the address resolution to complete.
  * @callback: Call invoked once address resolution has completed, timed out,
  *   or been canceled.  A status of 0 indicates success.
+ * @resolve_by_gid_attr:	Resolve the ip based on the GID attribute from
+ *				rdma_dev_addr.
  * @context: User-specified context associated with the call.
  */
 int rdma_resolve_ip(struct sockaddr *src_addr, const struct sockaddr *dst_addr,
-		    struct rdma_dev_addr *addr, int timeout_ms,
+		    struct rdma_dev_addr *addr, unsigned long timeout_ms,
 		    void (*callback)(int status, struct sockaddr *src_addr,
 				     struct rdma_dev_addr *addr, void *context),
-		    void *context);
+		    bool resolve_by_gid_attr, void *context);
 
 void rdma_addr_cancel(struct rdma_dev_addr *addr);
 
-void rdma_copy_addr(struct rdma_dev_addr *dev_addr,
-		    const struct net_device *dev,
-		    const unsigned char *dst_dev_addr);
-
 int rdma_addr_size(const struct sockaddr *addr);
 int rdma_addr_size_in6(struct sockaddr_in6 *addr);
 int rdma_addr_size_kss(struct __kernel_sockaddr_storage *addr);
diff --git a/include/rdma/ib_cm.h b/include/rdma/ib_cm.h
index c10f4b5ea8ab..49f4f75499b3 100644
--- a/include/rdma/ib_cm.h
+++ b/include/rdma/ib_cm.h
@@ -583,7 +583,7 @@ struct ib_cm_sidr_req_param {
 	struct sa_path_rec	*path;
 	const struct ib_gid_attr *sgid_attr;
 	__be64			service_id;
-	int			timeout_ms;
+	unsigned long		timeout_ms;
 	const void		*private_data;
 	u8			private_data_len;
 	u8			max_cm_retries;
diff --git a/include/rdma/ib_sa.h b/include/rdma/ib_sa.h
index b6ddf2a1b9d8..19520979b84c 100644
--- a/include/rdma/ib_sa.h
+++ b/include/rdma/ib_sa.h
@@ -449,28 +449,23 @@ struct ib_sa_query;
 
 void ib_sa_cancel_query(int id, struct ib_sa_query *query);
 
-int ib_sa_path_rec_get(struct ib_sa_client *client,
-		       struct ib_device *device, u8 port_num,
-		       struct sa_path_rec *rec,
-		       ib_sa_comp_mask comp_mask,
-		       int timeout_ms, gfp_t gfp_mask,
-		       void (*callback)(int status,
-					struct sa_path_rec *resp,
+int ib_sa_path_rec_get(struct ib_sa_client *client, struct ib_device *device,
+		       u8 port_num, struct sa_path_rec *rec,
+		       ib_sa_comp_mask comp_mask, unsigned long timeout_ms,
+		       gfp_t gfp_mask,
+		       void (*callback)(int status, struct sa_path_rec *resp,
 					void *context),
-		       void *context,
-		       struct ib_sa_query **query);
+		       void *context, struct ib_sa_query **query);
 
 int ib_sa_service_rec_query(struct ib_sa_client *client,
-			 struct ib_device *device, u8 port_num,
-			 u8 method,
-			 struct ib_sa_service_rec *rec,
-			 ib_sa_comp_mask comp_mask,
-			 int timeout_ms, gfp_t gfp_mask,
-			 void (*callback)(int status,
-					  struct ib_sa_service_rec *resp,
-					  void *context),
-			 void *context,
-			 struct ib_sa_query **sa_query);
+			    struct ib_device *device, u8 port_num, u8 method,
+			    struct ib_sa_service_rec *rec,
+			    ib_sa_comp_mask comp_mask, unsigned long timeout_ms,
+			    gfp_t gfp_mask,
+			    void (*callback)(int status,
+					     struct ib_sa_service_rec *resp,
+					     void *context),
+			    void *context, struct ib_sa_query **sa_query);
 
 struct ib_sa_multicast {
 	struct ib_sa_mcmember_rec rec;
@@ -573,12 +568,11 @@ int ib_sa_guid_info_rec_query(struct ib_sa_client *client,
 			      struct ib_device *device, u8 port_num,
 			      struct ib_sa_guidinfo_rec *rec,
 			      ib_sa_comp_mask comp_mask, u8 method,
-			      int timeout_ms, gfp_t gfp_mask,
+			      unsigned long timeout_ms, gfp_t gfp_mask,
 			      void (*callback)(int status,
 					       struct ib_sa_guidinfo_rec *resp,
 					       void *context),
-			      void *context,
-			      struct ib_sa_query **sa_query);
+			      void *context, struct ib_sa_query **sa_query);
 
 bool ib_sa_sendonly_fullmem_support(struct ib_sa_client *client,
 				    struct ib_device *device,
diff --git a/include/rdma/ib_umem.h b/include/rdma/ib_umem.h
index a1fd63871d17..5d3755ec5afa 100644
--- a/include/rdma/ib_umem.h
+++ b/include/rdma/ib_umem.h
@@ -42,15 +42,14 @@ struct ib_umem_odp;
 
 struct ib_umem {
 	struct ib_ucontext     *context;
+	struct mm_struct       *owning_mm;
 	size_t			length;
 	unsigned long		address;
 	int			page_shift;
-	int                     writable;
-	int                     hugetlb;
+	u32 writable : 1;
+	u32 hugetlb : 1;
+	u32 is_odp : 1;
 	struct work_struct	work;
-	struct mm_struct       *mm;
-	unsigned long		diff;
-	struct ib_umem_odp     *odp_data;
 	struct sg_table sg_head;
 	int             nmap;
 	int             npages;
diff --git a/include/rdma/ib_umem_odp.h b/include/rdma/ib_umem_odp.h
index 381cdf5a9bd1..0b1446fe2fab 100644
--- a/include/rdma/ib_umem_odp.h
+++ b/include/rdma/ib_umem_odp.h
@@ -43,6 +43,9 @@ struct umem_odp_node {
 };
 
 struct ib_umem_odp {
+	struct ib_umem umem;
+	struct ib_ucontext_per_mm *per_mm;
+
 	/*
 	 * An array of the pages included in the on-demand paging umem.
 	 * Indices of pages that are currently not mapped into the device will
@@ -64,16 +67,9 @@ struct ib_umem_odp {
 	struct mutex		umem_mutex;
 	void			*private; /* for the HW driver to use. */
 
-	/* When false, use the notifier counter in the ucontext struct. */
-	bool mn_counters_active;
 	int notifiers_seq;
 	int notifiers_count;
 
-	/* A linked list of umems that don't have private mmu notifier
-	 * counters yet. */
-	struct list_head no_private_counters;
-	struct ib_umem		*umem;
-
 	/* Tree tracking */
 	struct umem_odp_node	interval_tree;
 
@@ -82,15 +78,34 @@ struct ib_umem_odp {
 	struct work_struct	work;
 };
 
+static inline struct ib_umem_odp *to_ib_umem_odp(struct ib_umem *umem)
+{
+	return container_of(umem, struct ib_umem_odp, umem);
+}
+
 #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
 
-int ib_umem_odp_get(struct ib_ucontext *context, struct ib_umem *umem,
-		    int access);
-struct ib_umem *ib_alloc_odp_umem(struct ib_ucontext *context,
-				  unsigned long addr,
-				  size_t size);
+struct ib_ucontext_per_mm {
+	struct ib_ucontext *context;
+	struct mm_struct *mm;
+	struct pid *tgid;
+	bool active;
+
+	struct rb_root_cached umem_tree;
+	/* Protects umem_tree */
+	struct rw_semaphore umem_rwsem;
 
-void ib_umem_odp_release(struct ib_umem *umem);
+	struct mmu_notifier mn;
+	unsigned int odp_mrs_count;
+
+	struct list_head ucontext_list;
+	struct rcu_head rcu;
+};
+
+int ib_umem_odp_get(struct ib_umem_odp *umem_odp, int access);
+struct ib_umem_odp *ib_alloc_odp_umem(struct ib_ucontext_per_mm *per_mm,
+				      unsigned long addr, size_t size);
+void ib_umem_odp_release(struct ib_umem_odp *umem_odp);
 
 /*
  * The lower 2 bits of the DMA address signal the R/W permissions for
@@ -105,13 +120,14 @@ void ib_umem_odp_release(struct ib_umem *umem);
 
 #define ODP_DMA_ADDR_MASK (~(ODP_READ_ALLOWED_BIT | ODP_WRITE_ALLOWED_BIT))
 
-int ib_umem_odp_map_dma_pages(struct ib_umem *umem, u64 start_offset, u64 bcnt,
-			      u64 access_mask, unsigned long current_seq);
+int ib_umem_odp_map_dma_pages(struct ib_umem_odp *umem_odp, u64 start_offset,
+			      u64 bcnt, u64 access_mask,
+			      unsigned long current_seq);
 
-void ib_umem_odp_unmap_dma_pages(struct ib_umem *umem, u64 start_offset,
+void ib_umem_odp_unmap_dma_pages(struct ib_umem_odp *umem_odp, u64 start_offset,
 				 u64 bound);
 
-typedef int (*umem_call_back)(struct ib_umem *item, u64 start, u64 end,
+typedef int (*umem_call_back)(struct ib_umem_odp *item, u64 start, u64 end,
 			      void *cookie);
 /*
  * Call the callback on each ib_umem in the range. Returns the logical or of
@@ -129,46 +145,37 @@ int rbt_ib_umem_for_each_in_range(struct rb_root_cached *root,
 struct ib_umem_odp *rbt_ib_umem_lookup(struct rb_root_cached *root,
 				       u64 addr, u64 length);
 
-static inline int ib_umem_mmu_notifier_retry(struct ib_umem *item,
+static inline int ib_umem_mmu_notifier_retry(struct ib_umem_odp *umem_odp,
 					     unsigned long mmu_seq)
 {
 	/*
 	 * This code is strongly based on the KVM code from
 	 * mmu_notifier_retry. Should be called with
-	 * the relevant locks taken (item->odp_data->umem_mutex
+	 * the relevant locks taken (umem_odp->umem_mutex
 	 * and the ucontext umem_mutex semaphore locked for read).
 	 */
 
-	/* Do not allow page faults while the new ib_umem hasn't seen a state
-	 * with zero notifiers yet, and doesn't have its own valid set of
-	 * private counters. */
-	if (!item->odp_data->mn_counters_active)
-		return 1;
-
-	if (unlikely(item->odp_data->notifiers_count))
+	if (unlikely(umem_odp->notifiers_count))
 		return 1;
-	if (item->odp_data->notifiers_seq != mmu_seq)
+	if (umem_odp->notifiers_seq != mmu_seq)
 		return 1;
 	return 0;
 }
 
 #else /* CONFIG_INFINIBAND_ON_DEMAND_PAGING */
 
-static inline int ib_umem_odp_get(struct ib_ucontext *context,
-				  struct ib_umem *umem,
-				  int access)
+static inline int ib_umem_odp_get(struct ib_umem_odp *umem_odp, int access)
 {
 	return -EINVAL;
 }
 
-static inline struct ib_umem *ib_alloc_odp_umem(struct ib_ucontext *context,
-						unsigned long addr,
-						size_t size)
+static inline struct ib_umem_odp *
+ib_alloc_odp_umem(struct ib_ucontext *context, unsigned long addr, size_t size)
 {
 	return ERR_PTR(-EINVAL);
 }
 
-static inline void ib_umem_odp_release(struct ib_umem *umem) {}
+static inline void ib_umem_odp_release(struct ib_umem_odp *umem_odp) {}
 
 #endif /* CONFIG_INFINIBAND_ON_DEMAND_PAGING */
 
diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h
index 0ed5d913a492..9c0c2132a2d6 100644
--- a/include/rdma/ib_verbs.h
+++ b/include/rdma/ib_verbs.h
@@ -69,8 +69,11 @@
 
 #define IB_FW_VERSION_NAME_MAX	ETHTOOL_FWVERS_LEN
 
+struct ib_umem_odp;
+
 extern struct workqueue_struct *ib_wq;
 extern struct workqueue_struct *ib_comp_wq;
+extern struct workqueue_struct *ib_comp_unbound_wq;
 
 union ib_gid {
 	u8	raw[16];
@@ -1137,7 +1140,9 @@ enum ib_qp_create_flags {
  */
 
 struct ib_qp_init_attr {
+	/* Consumer's event_handler callback must not block */
 	void                  (*event_handler)(struct ib_event *, void *);
+
 	void		       *qp_context;
 	struct ib_cq	       *send_cq;
 	struct ib_cq	       *recv_cq;
@@ -1146,7 +1151,7 @@ struct ib_qp_init_attr {
 	struct ib_qp_cap	cap;
 	enum ib_sig_type	sq_sig_type;
 	enum ib_qp_type		qp_type;
-	enum ib_qp_create_flags	create_flags;
+	u32			create_flags;
 
 	/*
 	 * Only needed for special QP types, or when using the RW API.
@@ -1278,21 +1283,27 @@ struct ib_qp_attr {
 };
 
 enum ib_wr_opcode {
-	IB_WR_RDMA_WRITE,
-	IB_WR_RDMA_WRITE_WITH_IMM,
-	IB_WR_SEND,
-	IB_WR_SEND_WITH_IMM,
-	IB_WR_RDMA_READ,
-	IB_WR_ATOMIC_CMP_AND_SWP,
-	IB_WR_ATOMIC_FETCH_AND_ADD,
-	IB_WR_LSO,
-	IB_WR_SEND_WITH_INV,
-	IB_WR_RDMA_READ_WITH_INV,
-	IB_WR_LOCAL_INV,
-	IB_WR_REG_MR,
-	IB_WR_MASKED_ATOMIC_CMP_AND_SWP,
-	IB_WR_MASKED_ATOMIC_FETCH_AND_ADD,
+	/* These are shared with userspace */
+	IB_WR_RDMA_WRITE = IB_UVERBS_WR_RDMA_WRITE,
+	IB_WR_RDMA_WRITE_WITH_IMM = IB_UVERBS_WR_RDMA_WRITE_WITH_IMM,
+	IB_WR_SEND = IB_UVERBS_WR_SEND,
+	IB_WR_SEND_WITH_IMM = IB_UVERBS_WR_SEND_WITH_IMM,
+	IB_WR_RDMA_READ = IB_UVERBS_WR_RDMA_READ,
+	IB_WR_ATOMIC_CMP_AND_SWP = IB_UVERBS_WR_ATOMIC_CMP_AND_SWP,
+	IB_WR_ATOMIC_FETCH_AND_ADD = IB_UVERBS_WR_ATOMIC_FETCH_AND_ADD,
+	IB_WR_LSO = IB_UVERBS_WR_TSO,
+	IB_WR_SEND_WITH_INV = IB_UVERBS_WR_SEND_WITH_INV,
+	IB_WR_RDMA_READ_WITH_INV = IB_UVERBS_WR_RDMA_READ_WITH_INV,
+	IB_WR_LOCAL_INV = IB_UVERBS_WR_LOCAL_INV,
+	IB_WR_MASKED_ATOMIC_CMP_AND_SWP =
+		IB_UVERBS_WR_MASKED_ATOMIC_CMP_AND_SWP,
+	IB_WR_MASKED_ATOMIC_FETCH_AND_ADD =
+		IB_UVERBS_WR_MASKED_ATOMIC_FETCH_AND_ADD,
+
+	/* These are kernel only and can not be issued by userspace */
+	IB_WR_REG_MR = 0x20,
 	IB_WR_REG_SIG_MR,
+
 	/* reserve values for low level drivers' internal use.
 	 * These values will not be used at all in the ib core layer.
 	 */
@@ -1485,26 +1496,15 @@ struct ib_ucontext {
 	 * it is set when we are closing the file descriptor and indicates
 	 * that mm_sem may be locked.
 	 */
-	int			closing;
+	bool closing;
 
 	bool cleanup_retryable;
 
-	struct pid             *tgid;
 #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
-	struct rb_root_cached   umem_tree;
-	/*
-	 * Protects .umem_rbroot and tree, as well as odp_mrs_count and
-	 * mmu notifiers registration.
-	 */
-	struct rw_semaphore	umem_rwsem;
-	void (*invalidate_range)(struct ib_umem *umem,
+	void (*invalidate_range)(struct ib_umem_odp *umem_odp,
 				 unsigned long start, unsigned long end);
-
-	struct mmu_notifier	mn;
-	atomic_t		notifier_count;
-	/* A list of umems that don't have private mmu notifier counters yet. */
-	struct list_head	no_private_counters;
-	int                     odp_mrs_count;
+	struct mutex per_mm_list_lock;
+	struct list_head per_mm_list;
 #endif
 
 	struct ib_rdmacg_object	cg_obj;
@@ -1570,9 +1570,10 @@ struct ib_ah {
 typedef void (*ib_comp_handler)(struct ib_cq *cq, void *cq_context);
 
 enum ib_poll_context {
-	IB_POLL_DIRECT,		/* caller context, no hw completions */
-	IB_POLL_SOFTIRQ,	/* poll from softirq context */
-	IB_POLL_WORKQUEUE,	/* poll from workqueue */
+	IB_POLL_DIRECT,		   /* caller context, no hw completions */
+	IB_POLL_SOFTIRQ,	   /* poll from softirq context */
+	IB_POLL_WORKQUEUE,	   /* poll from workqueue */
+	IB_POLL_UNBOUND_WORKQUEUE, /* poll from unbound workqueue */
 };
 
 struct ib_cq {
@@ -1589,6 +1590,7 @@ struct ib_cq {
 		struct irq_poll		iop;
 		struct work_struct	work;
 	};
+	struct workqueue_struct *comp_wq;
 	/*
 	 * Implementation details of the RDMA core, don't use in drivers:
 	 */
@@ -2263,10 +2265,11 @@ struct ib_device {
 	struct list_head              event_handler_list;
 	spinlock_t                    event_handler_lock;
 
-	spinlock_t                    client_data_lock;
+	rwlock_t			client_data_lock;
 	struct list_head              core_list;
 	/* Access to the client_data_list is protected by the client_data_lock
-	 * spinlock and the lists_rwsem read-write semaphore */
+	 * rwlock and the lists_rwsem read-write semaphore
+	 */
 	struct list_head              client_data_list;
 
 	struct ib_cache               cache;
@@ -2550,7 +2553,13 @@ struct ib_device {
 
 	struct module               *owner;
 	struct device                dev;
-	struct kobject               *ports_parent;
+	/* First group for device attributes,
+	 * Second group for driver provided attributes (optional).
+	 * It is NULL terminated array.
+	 */
+	const struct attribute_group	*groups[3];
+
+	struct kobject			*ports_kobj;
 	struct list_head             port_list;
 
 	enum {
@@ -2633,9 +2642,9 @@ void ib_dealloc_device(struct ib_device *device);
 
 void ib_get_device_fw_str(struct ib_device *device, char *str);
 
-int ib_register_device(struct ib_device *device,
-		       int (*port_callback)(struct ib_device *,
-					    u8, struct kobject *));
+int ib_register_device(struct ib_device *device, const char *name,
+		       int (*port_callback)(struct ib_device *, u8,
+					    struct kobject *));
 void ib_unregister_device(struct ib_device *device);
 
 int ib_register_client   (struct ib_client *client);
@@ -2645,6 +2654,28 @@ void *ib_get_client_data(struct ib_device *device, struct ib_client *client);
 void  ib_set_client_data(struct ib_device *device, struct ib_client *client,
 			 void *data);
 
+#if IS_ENABLED(CONFIG_INFINIBAND_USER_ACCESS)
+int rdma_user_mmap_io(struct ib_ucontext *ucontext, struct vm_area_struct *vma,
+		      unsigned long pfn, unsigned long size, pgprot_t prot);
+int rdma_user_mmap_page(struct ib_ucontext *ucontext,
+			struct vm_area_struct *vma, struct page *page,
+			unsigned long size);
+#else
+static inline int rdma_user_mmap_io(struct ib_ucontext *ucontext,
+				    struct vm_area_struct *vma,
+				    unsigned long pfn, unsigned long size,
+				    pgprot_t prot)
+{
+	return -EINVAL;
+}
+static inline int rdma_user_mmap_page(struct ib_ucontext *ucontext,
+				struct vm_area_struct *vma, struct page *page,
+				unsigned long size)
+{
+	return -EINVAL;
+}
+#endif
+
 static inline int ib_copy_from_udata(void *dest, struct ib_udata *udata, size_t len)
 {
 	return copy_from_user(dest, udata->inbuf, len) ? -EFAULT : 0;
@@ -2728,7 +2759,6 @@ static inline int ib_destroy_usecnt(atomic_t *usecnt,
  * @next_state: Next QP state
  * @type: QP type
  * @mask: Mask of supplied QP attributes
- * @ll : link layer of port
  *
  * This function is a helper function that a low-level driver's
  * modify_qp method can use to validate the consumer's input.  It
@@ -2737,8 +2767,7 @@ static inline int ib_destroy_usecnt(atomic_t *usecnt,
  * and that the attribute mask supplied is allowed for the transition.
  */
 bool ib_modify_qp_is_ok(enum ib_qp_state cur_state, enum ib_qp_state next_state,
-			enum ib_qp_type type, enum ib_qp_attr_mask mask,
-			enum rdma_link_layer ll);
+			enum ib_qp_type type, enum ib_qp_attr_mask mask);
 
 void ib_register_event_handler(struct ib_event_handler *event_handler);
 void ib_unregister_event_handler(struct ib_event_handler *event_handler);
@@ -4167,20 +4196,6 @@ ib_get_vector_affinity(struct ib_device *device, int comp_vector)
 
 }
 
-static inline void ib_set_flow(struct ib_uobject *uobj, struct ib_flow *ibflow,
-			       struct ib_qp *qp, struct ib_device *device)
-{
-	uobj->object = ibflow;
-	ibflow->uobject = uobj;
-
-	if (qp) {
-		atomic_inc(&qp->usecnt);
-		ibflow->qp = qp;
-	}
-
-	ibflow->device = device;
-}
-
 /**
  * rdma_roce_rescan_device - Rescan all of the network devices in the system
  * and add their gids, as needed, to the relevant RoCE devices.
@@ -4205,4 +4220,26 @@ int rdma_init_netdev(struct ib_device *device, u8 port_num,
 		     void (*setup)(struct net_device *),
 		     struct net_device *netdev);
 
+/**
+ * rdma_set_device_sysfs_group - Set device attributes group to have
+ *				 driver specific sysfs entries at
+ *				 for infiniband class.
+ *
+ * @device:	device pointer for which attributes to be created
+ * @group:	Pointer to group which should be added when device
+ *		is registered with sysfs.
+ * rdma_set_device_sysfs_group() allows existing drivers to expose one
+ * group per device to have sysfs attributes.
+ *
+ * NOTE: New drivers should not make use of this API; instead new device
+ * parameter should be exposed via netlink command. This API and mechanism
+ * exist only for existing drivers.
+ */
+static inline void
+rdma_set_device_sysfs_group(struct ib_device *dev,
+			    const struct attribute_group *group)
+{
+	dev->groups[1] = group;
+}
+
 #endif /* IB_VERBS_H */
diff --git a/include/rdma/rdma_cm.h b/include/rdma/rdma_cm.h
index 5d71a7f51a9f..60987a5903b7 100644
--- a/include/rdma/rdma_cm.h
+++ b/include/rdma/rdma_cm.h
@@ -152,7 +152,11 @@ struct rdma_cm_id *__rdma_create_id(struct net *net,
  * @ps: RDMA port space.
  * @qp_type: type of queue pair associated with the id.
  *
- * The id holds a reference on the network namespace until it is destroyed.
+ * Returns a new rdma_cm_id. The id holds a reference on the network
+ * namespace until it is destroyed.
+ *
+ * The event handler callback serializes on the id's mutex and is
+ * allowed to sleep.
  */
 #define rdma_create_id(net, event_handler, context, ps, qp_type) \
 	__rdma_create_id((net), (event_handler), (context), (ps), (qp_type), \
@@ -192,7 +196,8 @@ int rdma_bind_addr(struct rdma_cm_id *id, struct sockaddr *addr);
  * @timeout_ms: Time to wait for resolution to complete.
  */
 int rdma_resolve_addr(struct rdma_cm_id *id, struct sockaddr *src_addr,
-		      const struct sockaddr *dst_addr, int timeout_ms);
+		      const struct sockaddr *dst_addr,
+		      unsigned long timeout_ms);
 
 /**
  * rdma_resolve_route - Resolve the RDMA address bound to the RDMA identifier
@@ -202,7 +207,7 @@ int rdma_resolve_addr(struct rdma_cm_id *id, struct sockaddr *src_addr,
  * Users must have first called rdma_resolve_addr to resolve a dst_addr
  * into an RDMA address before calling this routine.
  */
-int rdma_resolve_route(struct rdma_cm_id *id, int timeout_ms);
+int rdma_resolve_route(struct rdma_cm_id *id, unsigned long timeout_ms);
 
 /**
  * rdma_create_qp - Allocate a QP and associate it with the specified RDMA
diff --git a/include/rdma/rdma_netlink.h b/include/rdma/rdma_netlink.h
index c369703fcd69..70218e6b5187 100644
--- a/include/rdma/rdma_netlink.h
+++ b/include/rdma/rdma_netlink.h
@@ -96,7 +96,7 @@ int rdma_nl_multicast(struct sk_buff *skb, unsigned int group, gfp_t flags);
 /**
  * Check if there are any listeners to the netlink group
  * @group: the netlink group ID
- * Returns 0 on success or a negative for no listeners.
+ * Returns true on success or false if no listeners.
  */
-int rdma_nl_chk_listeners(unsigned int group);
+bool rdma_nl_chk_listeners(unsigned int group);
 #endif /* _RDMA_NETLINK_H */
diff --git a/include/rdma/rdma_vt.h b/include/rdma/rdma_vt.h
index e79229a0cf01..3584d0816fcd 100644
--- a/include/rdma/rdma_vt.h
+++ b/include/rdma/rdma_vt.h
@@ -149,6 +149,10 @@ struct rvt_ibport {
 
 #define RVT_CQN_MAX 16 /* maximum length of cq name */
 
+#define RVT_SGE_COPY_MEMCPY	0
+#define RVT_SGE_COPY_CACHELESS	1
+#define RVT_SGE_COPY_ADAPTIVE	2
+
 /*
  * Things that are driver specific, module parameters in hfi1 and qib
  */
@@ -161,6 +165,9 @@ struct rvt_driver_params {
 	 */
 	unsigned int lkey_table_size;
 	unsigned int qp_table_size;
+	unsigned int sge_copy_mode;
+	unsigned int wss_threshold;
+	unsigned int wss_clean_period;
 	int qpn_start;
 	int qpn_inc;
 	int qpn_res_start;
@@ -193,6 +200,19 @@ struct rvt_ah {
 	u8 log_pmtu;
 };
 
+/* memory working set size */
+struct rvt_wss {
+	unsigned long *entries;
+	atomic_t total_count;
+	atomic_t clean_counter;
+	atomic_t clean_entry;
+
+	int threshold;
+	int num_entries;
+	long pages_mask;
+	unsigned int clean_period;
+};
+
 struct rvt_dev_info;
 struct rvt_swqe;
 struct rvt_driver_provided {
@@ -211,11 +231,18 @@ struct rvt_driver_provided {
 	 * version requires the s_lock not to be held. The other assumes the
 	 * s_lock is held.
 	 */
-	void (*schedule_send)(struct rvt_qp *qp);
-	void (*schedule_send_no_lock)(struct rvt_qp *qp);
+	bool (*schedule_send)(struct rvt_qp *qp);
+	bool (*schedule_send_no_lock)(struct rvt_qp *qp);
 
-	/* Driver specific work request checking */
-	int (*check_send_wqe)(struct rvt_qp *qp, struct rvt_swqe *wqe);
+	/*
+	 * Driver specific work request setup and checking.
+	 * This function is allowed to perform any setup, checks, or
+	 * adjustments required to the SWQE in order to be usable by
+	 * underlying protocols. This includes private data structure
+	 * allocations.
+	 */
+	int (*setup_wqe)(struct rvt_qp *qp, struct rvt_swqe *wqe,
+			 bool *call_send);
 
 	/*
 	 * Sometimes rdmavt needs to kick the driver's send progress. That is
@@ -371,6 +398,9 @@ struct rvt_dev_info {
 	/* post send table */
 	const struct rvt_operation_params *post_parms;
 
+	/* opcode translation table */
+	const enum ib_wc_opcode *wc_opcode;
+
 	/* Driver specific helper functions */
 	struct rvt_driver_provided driver_f;
 
@@ -411,6 +441,8 @@ struct rvt_dev_info {
 	u32 n_mcast_grps_allocated; /* number of mcast groups allocated */
 	spinlock_t n_mcast_grps_lock;
 
+	/* Memory Working Set Size */
+	struct rvt_wss *wss;
 };
 
 /**
@@ -423,7 +455,14 @@ static inline void rvt_set_ibdev_name(struct rvt_dev_info *rdi,
 				      const char *fmt, const char *name,
 				      const int unit)
 {
-	snprintf(rdi->ibdev.name, sizeof(rdi->ibdev.name), fmt, name, unit);
+	/*
+	 * FIXME: rvt and its users want to touch the ibdev before
+	 * registration and have things like the name work. We don't have the
+	 * infrastructure in the core to support this directly today, hack it
+	 * to work by setting the name manually here.
+	 */
+	dev_set_name(&rdi->ibdev.dev, fmt, name, unit);
+	strlcpy(rdi->ibdev.name, dev_name(&rdi->ibdev.dev), IB_DEVICE_NAME_MAX);
 }
 
 /**
@@ -434,7 +473,7 @@ static inline void rvt_set_ibdev_name(struct rvt_dev_info *rdi,
  */
 static inline const char *rvt_get_ibdev_name(const struct rvt_dev_info *rdi)
 {
-	return rdi->ibdev.name;
+	return dev_name(&rdi->ibdev.dev);
 }
 
 static inline struct rvt_pd *ibpd_to_rvtpd(struct ib_pd *ibpd)
diff --git a/include/rdma/rdmavt_qp.h b/include/rdma/rdmavt_qp.h
index 927f6d5b6d0f..cbafb1878669 100644
--- a/include/rdma/rdmavt_qp.h
+++ b/include/rdma/rdmavt_qp.h
@@ -678,6 +678,13 @@ void rvt_del_timers_sync(struct rvt_qp *qp);
 void rvt_stop_rc_timers(struct rvt_qp *qp);
 void rvt_add_retry_timer(struct rvt_qp *qp);
 
+void rvt_copy_sge(struct rvt_qp *qp, struct rvt_sge_state *ss,
+		  void *data, u32 length,
+		  bool release, bool copy_last);
+void rvt_send_complete(struct rvt_qp *qp, struct rvt_swqe *wqe,
+		       enum ib_wc_status status);
+void rvt_ruc_loopback(struct rvt_qp *qp);
+
 /**
  * struct rvt_qp_iter - the iterator for QPs
  * @qp - the current QP
diff --git a/include/rdma/restrack.h b/include/rdma/restrack.h
index 9654d33edd98..2638fa7cd702 100644
--- a/include/rdma/restrack.h
+++ b/include/rdma/restrack.h
@@ -173,16 +173,10 @@ int rdma_restrack_put(struct rdma_restrack_entry *res);
 /**
  * rdma_restrack_set_task() - set the task for this resource
  * @res:  resource entry
- * @task: task struct
+ * @caller: kernel name, the current task will be used if the caller is NULL.
  */
-static inline void rdma_restrack_set_task(struct rdma_restrack_entry *res,
-					  struct task_struct *task)
-{
-	if (res->task)
-		put_task_struct(res->task);
-	get_task_struct(task);
-	res->task = task;
-}
+void rdma_restrack_set_task(struct rdma_restrack_entry *res,
+			    const char *caller);
 
 /*
  * Helper functions for rdma drivers when filling out
diff --git a/include/rdma/uverbs_ioctl.h b/include/rdma/uverbs_ioctl.h
index 9e997c3c2f04..84d3d15f1f38 100644
--- a/include/rdma/uverbs_ioctl.h
+++ b/include/rdma/uverbs_ioctl.h
@@ -52,6 +52,7 @@ enum uverbs_attr_type {
 	UVERBS_ATTR_TYPE_IDR,
 	UVERBS_ATTR_TYPE_FD,
 	UVERBS_ATTR_TYPE_ENUM_IN,
+	UVERBS_ATTR_TYPE_IDRS_ARRAY,
 };
 
 enum uverbs_obj_access {
@@ -101,7 +102,7 @@ struct uverbs_attr_spec {
 		} enum_def;
 	} u;
 
-	/* This weird split of the enum lets us remove some padding */
+	/* This weird split lets us remove some padding */
 	union {
 		struct {
 			/*
@@ -111,6 +112,17 @@ struct uverbs_attr_spec {
 			 */
 			const struct uverbs_attr_spec *ids;
 		} enum_def;
+
+		struct {
+			/*
+			 * higher bits mean the namespace and lower bits mean
+			 * the type id within the namespace.
+			 */
+			u16				obj_type;
+			u16				min_len;
+			u16				max_len;
+			u8				access;
+		} objs_arr;
 	} u2;
 };
 
@@ -251,6 +263,11 @@ static inline __attribute_const__ u32 uapi_bkey_attr(u32 attr_key)
 	return attr_key - 1;
 }
 
+static inline __attribute_const__ u32 uapi_bkey_to_key_attr(u32 attr_bkey)
+{
+	return attr_bkey + 1;
+}
+
 /*
  * =======================================
  *	Verbs definitions
@@ -323,6 +340,27 @@ struct uverbs_object_tree_def {
 #define UA_MANDATORY .mandatory = 1
 #define UA_OPTIONAL .mandatory = 0
 
+/*
+ * min_len must be bigger than 0 and _max_len must be smaller than 4095.  Only
+ * READ\WRITE accesses are supported.
+ */
+#define UVERBS_ATTR_IDRS_ARR(_attr_id, _idr_type, _access, _min_len, _max_len, \
+			     ...)                                              \
+	(&(const struct uverbs_attr_def){                                      \
+		.id = (_attr_id) +                                             \
+		      BUILD_BUG_ON_ZERO((_min_len) == 0 ||                     \
+					(_max_len) >                           \
+						PAGE_SIZE / sizeof(void *) ||  \
+					(_min_len) > (_max_len) ||             \
+					(_access) == UVERBS_ACCESS_NEW ||      \
+					(_access) == UVERBS_ACCESS_DESTROY),   \
+		.attr = { .type = UVERBS_ATTR_TYPE_IDRS_ARRAY,                 \
+			  .u2.objs_arr.obj_type = _idr_type,                   \
+			  .u2.objs_arr.access = _access,                       \
+			  .u2.objs_arr.min_len = _min_len,                     \
+			  .u2.objs_arr.max_len = _max_len,                     \
+			  __VA_ARGS__ } })
+
 #define UVERBS_ATTR_IDR(_attr_id, _idr_type, _access, ...)                     \
 	(&(const struct uverbs_attr_def){                                      \
 		.id = _attr_id,                                                \
@@ -365,6 +403,15 @@ struct uverbs_object_tree_def {
 			  __VA_ARGS__ },                                       \
 	})
 
+/* An input value that is a member in the enum _enum_type. */
+#define UVERBS_ATTR_CONST_IN(_attr_id, _enum_type, ...)                        \
+	UVERBS_ATTR_PTR_IN(                                                    \
+		_attr_id,                                                      \
+		UVERBS_ATTR_SIZE(                                              \
+			sizeof(u64) + BUILD_BUG_ON_ZERO(!sizeof(_enum_type)),  \
+			sizeof(u64)),                                          \
+		__VA_ARGS__)
+
 /*
  * An input value that is a bitwise combination of values of _enum_type.
  * This permits the flag value to be passed as either a u32 or u64, it must
@@ -431,10 +478,16 @@ struct uverbs_obj_attr {
 	const struct uverbs_api_attr	*attr_elm;
 };
 
+struct uverbs_objs_arr_attr {
+	struct ib_uobject **uobjects;
+	u16 len;
+};
+
 struct uverbs_attr {
 	union {
 		struct uverbs_ptr_attr	ptr_attr;
 		struct uverbs_obj_attr	obj_attr;
+		struct uverbs_objs_arr_attr objs_arr_attr;
 	};
 };
 
@@ -507,6 +560,31 @@ uverbs_attr_get_len(const struct uverbs_attr_bundle *attrs_bundle, u16 idx)
 	return attr->ptr_attr.len;
 }
 
+/**
+ * uverbs_attr_get_uobjs_arr() - Provides array's properties for attribute for
+ * UVERBS_ATTR_TYPE_IDRS_ARRAY.
+ * @arr: Returned pointer to array of pointers for uobjects or NULL if
+ *       the attribute isn't provided.
+ *
+ * Return: The array length or 0 if no attribute was provided.
+ */
+static inline int uverbs_attr_get_uobjs_arr(
+	const struct uverbs_attr_bundle *attrs_bundle, u16 attr_idx,
+	struct ib_uobject ***arr)
+{
+	const struct uverbs_attr *attr =
+			uverbs_attr_get(attrs_bundle, attr_idx);
+
+	if (IS_ERR(attr)) {
+		*arr = NULL;
+		return 0;
+	}
+
+	*arr = attr->objs_arr_attr.uobjects;
+
+	return attr->objs_arr_attr.len;
+}
+
 static inline bool uverbs_attr_ptr_is_inline(const struct uverbs_attr *attr)
 {
 	return attr->ptr_attr.len <= sizeof(attr->ptr_attr.data);
@@ -603,6 +681,9 @@ static inline __malloc void *uverbs_zalloc(struct uverbs_attr_bundle *bundle,
 {
 	return _uverbs_alloc(bundle, size, GFP_KERNEL | __GFP_ZERO);
 }
+int _uverbs_get_const(s64 *to, const struct uverbs_attr_bundle *attrs_bundle,
+		      size_t idx, s64 lower_bound, u64 upper_bound,
+		      s64 *def_val);
 #else
 static inline int
 uverbs_get_flags64(u64 *to, const struct uverbs_attr_bundle *attrs_bundle,
@@ -631,6 +712,34 @@ static inline __malloc void *uverbs_zalloc(struct uverbs_attr_bundle *bundle,
 {
 	return ERR_PTR(-EINVAL);
 }
+static inline int
+_uverbs_get_const(s64 *to, const struct uverbs_attr_bundle *attrs_bundle,
+		  size_t idx, s64 lower_bound, u64 upper_bound,
+		  s64 *def_val)
+{
+	return -EINVAL;
+}
 #endif
 
+#define uverbs_get_const(_to, _attrs_bundle, _idx)                             \
+	({                                                                     \
+		s64 _val;                                                      \
+		int _ret = _uverbs_get_const(&_val, _attrs_bundle, _idx,       \
+					     type_min(typeof(*_to)),           \
+					     type_max(typeof(*_to)), NULL);    \
+		(*_to) = _val;                                                 \
+		_ret;                                                          \
+	})
+
+#define uverbs_get_const_default(_to, _attrs_bundle, _idx, _default)           \
+	({                                                                     \
+		s64 _val;                                                      \
+		s64 _def_val = _default;                                       \
+		int _ret =                                                     \
+			_uverbs_get_const(&_val, _attrs_bundle, _idx,          \
+					  type_min(typeof(*_to)),              \
+					  type_max(typeof(*_to)), &_def_val);  \
+		(*_to) = _val;                                                 \
+		_ret;                                                          \
+	})
 #endif
diff --git a/include/rdma/uverbs_std_types.h b/include/rdma/uverbs_std_types.h
index 3b00231cc084..3db2802fbc68 100644
--- a/include/rdma/uverbs_std_types.h
+++ b/include/rdma/uverbs_std_types.h
@@ -140,5 +140,56 @@ __uobj_alloc(const struct uverbs_api_object *obj, struct ib_uverbs_file *ufile,
 #define uobj_alloc(_type, _ufile, _ib_dev)                                     \
 	__uobj_alloc(uobj_get_type(_ufile, _type), _ufile, _ib_dev)
 
+static inline void uverbs_flow_action_fill_action(struct ib_flow_action *action,
+						  struct ib_uobject *uobj,
+						  struct ib_device *ib_dev,
+						  enum ib_flow_action_type type)
+{
+	atomic_set(&action->usecnt, 0);
+	action->device = ib_dev;
+	action->type = type;
+	action->uobject = uobj;
+	uobj->object = action;
+}
+
+struct ib_uflow_resources {
+	size_t			max;
+	size_t			num;
+	size_t			collection_num;
+	size_t			counters_num;
+	struct ib_counters	**counters;
+	struct ib_flow_action	**collection;
+};
+
+struct ib_uflow_object {
+	struct ib_uobject		uobject;
+	struct ib_uflow_resources	*resources;
+};
+
+struct ib_uflow_resources *flow_resources_alloc(size_t num_specs);
+void flow_resources_add(struct ib_uflow_resources *uflow_res,
+			enum ib_flow_spec_type type,
+			void *ibobj);
+void ib_uverbs_flow_resources_free(struct ib_uflow_resources *uflow_res);
+
+static inline void ib_set_flow(struct ib_uobject *uobj, struct ib_flow *ibflow,
+			       struct ib_qp *qp, struct ib_device *device,
+			       struct ib_uflow_resources *uflow_res)
+{
+	struct ib_uflow_object *uflow;
+
+	uobj->object = ibflow;
+	ibflow->uobject = uobj;
+
+	if (qp) {
+		atomic_inc(&qp->usecnt);
+		ibflow->qp = qp;
+	}
+
+	ibflow->device = device;
+	uflow = container_of(uobj, typeof(*uflow), uobject);
+	uflow->resources = uflow_res;
+}
+
 #endif