Index: sys/dev/cxgbe/iw_cxgbe/cm.c =================================================================== --- sys/dev/cxgbe/iw_cxgbe/cm.c +++ sys/dev/cxgbe/iw_cxgbe/cm.c @@ -2009,7 +2009,6 @@ int err; struct c4iw_ep *ep = to_ep(cm_id); CTR2(KTR_IW_CXGBE, "%s:crcB %p", __func__, ep); - int abort = 0; if ((state_read(&ep->com) == DEAD) || (state_read(&ep->com) != MPA_REQ_RCVD)) { @@ -2020,18 +2019,15 @@ } set_bit(ULP_REJECT, &ep->com.history); - if (mpa_rev == 0) { - - CTR2(KTR_IW_CXGBE, "%s:crc2 %p", __func__, ep); - abort = 1; - } - else { - - CTR2(KTR_IW_CXGBE, "%s:crc3 %p", __func__, ep); - abort = send_mpa_reject(ep, pdata, pdata_len); - } + CTR3(KTR_IW_CXGBE, "%s:crc2 mpa_rev %d ep %p", __func__, mpa_rev, ep); + if (mpa_rev != 0) + err = send_mpa_reject(ep, pdata, pdata_len); stop_ep_timer(ep); - err = c4iw_ep_disconnect(ep, abort != 0, GFP_KERNEL); + /* FIXME: Due to design limitation, only ABORTIVE close is performed + * here. But should do NORMAL close when MPA_REJ is sent successfully. + */ + err = c4iw_ep_disconnect(ep, 1, GFP_KERNEL); + soclose(ep->com.so); c4iw_put_ep(&ep->com); CTR3(KTR_IW_CXGBE, "%s:crc4 %p, err: %d", __func__, ep, err); return 0; Index: sys/dev/cxgbe/iw_cxgbe/provider.c =================================================================== --- sys/dev/cxgbe/iw_cxgbe/provider.c +++ sys/dev/cxgbe/iw_cxgbe/provider.c @@ -388,6 +388,24 @@ return 0; } +static int c4iw_port_immutable(struct ib_device *ibdev, u8 port_num, + struct ib_port_immutable *immutable) +{ + struct ib_port_attr attr; + int err; + + immutable->core_cap_flags = RDMA_CORE_PORT_IWARP; + + err = ib_query_port(ibdev, port_num, &attr); + if (err) + return err; + + immutable->pkey_tbl_len = attr.pkey_tbl_len; + immutable->gid_tbl_len = attr.gid_tbl_len; + + return 0; +} + /* * Returns -errno on error. */ @@ -471,6 +489,7 @@ ibdev->post_send = c4iw_post_send; ibdev->post_recv = c4iw_post_receive; ibdev->uverbs_abi_ver = C4IW_UVERBS_ABI_VERSION; + ibdev->get_port_immutable = c4iw_port_immutable; iwcm = kmalloc(sizeof(*iwcm), GFP_KERNEL); if (iwcm == NULL) Index: sys/dev/mlx4/mlx4_ib/mlx4_ib_main.c =================================================================== --- sys/dev/mlx4/mlx4_ib/mlx4_ib_main.c +++ sys/dev/mlx4/mlx4_ib/mlx4_ib_main.c @@ -2229,6 +2229,38 @@ return -1; } +static int mlx4_port_immutable(struct ib_device *ibdev, u8 port_num, + struct ib_port_immutable *immutable) +{ + struct ib_port_attr attr; + struct mlx4_ib_dev *mdev = to_mdev(ibdev); + int err; + + if (mlx4_ib_port_link_layer(ibdev, port_num) == IB_LINK_LAYER_INFINIBAND) { + immutable->core_cap_flags = RDMA_CORE_PORT_IBA_IB; + immutable->max_mad_size = IB_MGMT_MAD_SIZE; + } else { + if (mdev->dev->caps.flags & MLX4_DEV_CAP_FLAG_IBOE) + immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE; + if (mdev->dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_ROCEV2) + immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE | + RDMA_CORE_PORT_IBA_ROCE_UDP_ENCAP; + immutable->core_cap_flags |= RDMA_CORE_PORT_RAW_PACKET; + if (immutable->core_cap_flags & (RDMA_CORE_PORT_IBA_ROCE | + RDMA_CORE_PORT_IBA_ROCE_UDP_ENCAP)) + immutable->max_mad_size = IB_MGMT_MAD_SIZE; + } + + err = ib_query_port(ibdev, port_num, &attr); + if (err) + return err; + + immutable->pkey_tbl_len = attr.pkey_tbl_len; + immutable->gid_tbl_len = attr.gid_tbl_len; + + return 0; +} + static void *mlx4_ib_add(struct mlx4_dev *dev) { struct mlx4_ib_dev *ibdev; @@ -2360,6 +2392,7 @@ ibdev->ib_dev.attach_mcast = mlx4_ib_mcg_attach; ibdev->ib_dev.detach_mcast = mlx4_ib_mcg_detach; ibdev->ib_dev.process_mad = mlx4_ib_process_mad; + ibdev->ib_dev.get_port_immutable = mlx4_port_immutable; ibdev->ib_dev.get_netdev = mlx4_ib_get_netdev; ibdev->ib_dev.ioctl = mlx4_ib_ioctl; ibdev->ib_dev.query_values = mlx4_ib_query_values; Index: sys/dev/mlx5/mlx5_ib/mlx5_ib_main.c =================================================================== --- sys/dev/mlx5/mlx5_ib/mlx5_ib_main.c +++ sys/dev/mlx5/mlx5_ib/mlx5_ib_main.c @@ -1670,6 +1670,57 @@ mlx5_ib_dealloc_pd(devr->p0); } +static u32 get_core_cap_flags(struct ib_device *ibdev) +{ + struct mlx5_ib_dev *dev = to_mdev(ibdev); + enum rdma_link_layer ll = mlx5_ib_port_link_layer(ibdev, 1); + u8 l3_type_cap = MLX5_CAP_ROCE(dev->mdev, l3_type); + u8 roce_version_cap = MLX5_CAP_ROCE(dev->mdev, roce_version); + u32 ret = 0; + + if (ll == IB_LINK_LAYER_INFINIBAND) + return RDMA_CORE_PORT_IBA_IB; + + ret = RDMA_CORE_PORT_RAW_PACKET; + + if (!(l3_type_cap & MLX5_ROCE_L3_TYPE_IPV4_CAP)) + return ret; + + if (!(l3_type_cap & MLX5_ROCE_L3_TYPE_IPV6_CAP)) + return ret; + + if (roce_version_cap & MLX5_ROCE_VERSION_1_CAP) + ret |= RDMA_CORE_PORT_IBA_ROCE; + + if (roce_version_cap & MLX5_ROCE_VERSION_2_CAP) + ret |= RDMA_CORE_PORT_IBA_ROCE_UDP_ENCAP; + + return ret; +} + +static int mlx5_port_immutable(struct ib_device *ibdev, u8 port_num, + struct ib_port_immutable *immutable) +{ + struct ib_port_attr attr; + struct mlx5_ib_dev *dev = to_mdev(ibdev); + enum rdma_link_layer ll = mlx5_ib_port_link_layer(ibdev, port_num); + int err; + + immutable->core_cap_flags = get_core_cap_flags(ibdev); + + err = ib_query_port(ibdev, port_num, &attr); + if (err) + return err; + + immutable->pkey_tbl_len = attr.pkey_tbl_len; + immutable->gid_tbl_len = attr.gid_tbl_len; + immutable->core_cap_flags = get_core_cap_flags(ibdev); + if ((ll == IB_LINK_LAYER_INFINIBAND) || MLX5_CAP_GEN(dev->mdev, roce)) + immutable->max_mad_size = IB_MGMT_MAD_SIZE; + + return 0; +} + static void enable_dc_tracer(struct mlx5_ib_dev *dev) { struct device *device = dev->ib_dev.dma_device; @@ -2115,6 +2166,7 @@ dev->ib_dev.attach_mcast = mlx5_ib_mcg_attach; dev->ib_dev.detach_mcast = mlx5_ib_mcg_detach; dev->ib_dev.process_mad = mlx5_ib_process_mad; + dev->ib_dev.get_port_immutable = mlx5_port_immutable; dev->ib_dev.alloc_fast_reg_mr = mlx5_ib_alloc_fast_reg_mr; dev->ib_dev.alloc_fast_reg_page_list = mlx5_ib_alloc_fast_reg_page_list; dev->ib_dev.free_fast_reg_page_list = mlx5_ib_free_fast_reg_page_list; Index: sys/dev/mlx5/mlx5_ib/mlx5_ib_qp.c =================================================================== --- sys/dev/mlx5/mlx5_ib/mlx5_ib_qp.c +++ sys/dev/mlx5/mlx5_ib/mlx5_ib_qp.c @@ -1576,7 +1576,7 @@ int gid_type; if ((ll == IB_LINK_LAYER_ETHERNET) || (ah->ah_flags & IB_AH_GRH)) { - int len = dev->ib_dev.gid_tbl_len[port - 1]; + int len = dev->mdev->port_caps[port - 1].gid_table_len; if (ah->grh.sgid_index >= len) { printf("mlx5_ib: ERR: ""sgid_index (%u) too large. max is %d\n", ah->grh.sgid_index, len - 1); return -EINVAL; Index: sys/ofed/drivers/infiniband/core/cma.c =================================================================== --- sys/ofed/drivers/infiniband/core/cma.c +++ sys/ofed/drivers/infiniband/core/cma.c @@ -75,11 +75,6 @@ module_param_named(def_prec2sl, def_prec2sl, int, 0644); MODULE_PARM_DESC(def_prec2sl, "Default value for SL priority with RoCE. Valid values 0 - 7"); -static int unify_tcp_port_space = 1; -module_param(unify_tcp_port_space, int, 0644); -MODULE_PARM_DESC(unify_tcp_port_space, "Unify the host TCP and RDMA port " - "space allocation (default=1)"); - static int debug_level = 0; #define cma_pr(level, priv, format, arg...) \ printk(level "CMA: %p: %s: " format, ((struct rdma_id_priv *) priv) , __func__, ## arg) @@ -198,6 +193,7 @@ /* cache for mc record params */ struct ib_sa_mcmember_rec rec; int is_valid_rec; + int unify_ps_tcp; }; struct cma_multicast { @@ -822,29 +818,24 @@ int ret = 0; id_priv = container_of(id, struct rdma_id_private, id); - switch (rdma_node_get_transport(id_priv->id.device->node_type)) { - case RDMA_TRANSPORT_IB: + if (rdma_cap_ib_cm(id->device, id->port_num)) { if (!id_priv->cm_id.ib || (id_priv->id.qp_type == IB_QPT_UD)) ret = cma_ib_init_qp_attr(id_priv, qp_attr, qp_attr_mask); else ret = ib_cm_init_qp_attr(id_priv->cm_id.ib, qp_attr, qp_attr_mask); + if (qp_attr->qp_state == IB_QPS_RTR) qp_attr->rq_psn = id_priv->seq_num; - break; - case RDMA_TRANSPORT_IWARP: - case RDMA_TRANSPORT_SCIF: + } else if (rdma_cap_iw_cm(id->device, id->port_num)) { if (!id_priv->cm_id.iw) { qp_attr->qp_access_flags = 0; *qp_attr_mask = IB_QP_STATE | IB_QP_ACCESS_FLAGS; } else ret = iw_cm_init_qp_attr(id_priv->cm_id.iw, qp_attr, qp_attr_mask); - break; - default: + } else ret = -ENOSYS; - break; - } return ret; } @@ -1105,6 +1096,10 @@ if (id_priv->internal_id) cma_deref_id(id_priv->id.context); + if (id_priv->sock != NULL && !id_priv->internal_id && + !id_priv->unify_ps_tcp) + sock_release(id_priv->sock); + kfree(id_priv->id.route.path_rec); kfree(id_priv); } @@ -1128,8 +1123,7 @@ mutex_unlock(&id_priv->handler_mutex); if (id_priv->cma_dev) { - switch (rdma_node_get_transport(id_priv->id.device->node_type)) { - case RDMA_TRANSPORT_IB: + if (rdma_cap_ib_cm(id_priv->id.device, 1)) { spin_lock_irqsave(&id_priv->cm_lock, flags); if (id_priv->cm_id.ib && !IS_ERR(id_priv->cm_id.ib)) { ib = id_priv->cm_id.ib; @@ -1138,14 +1132,9 @@ ib_destroy_cm_id(ib); } else spin_unlock_irqrestore(&id_priv->cm_lock, flags); - break; - case RDMA_TRANSPORT_IWARP: - case RDMA_TRANSPORT_SCIF: + } else if (rdma_cap_iw_cm(id_priv->id.device, 1)) { if (id_priv->cm_id.iw) iw_destroy_cm_id(id_priv->cm_id.iw); - break; - default: - break; } cma_leave_mc_groups(id_priv); cma_release_dev(id_priv); @@ -1671,6 +1660,11 @@ rdma_destroy_id(new_cm_id); goto out; } + /* + * Update the socket pointer in the RDMA CM ID so that + * the socket gets properly freed. + */ + conn_id->sock = iw_event->so; ret = rdma_copy_addr(&conn_id->id.route.addr.dev_addr, dev, NULL); if (ret) { mutex_unlock(&conn_id->handler_mutex); @@ -2141,27 +2135,15 @@ return -EINVAL; atomic_inc(&id_priv->refcount); - switch (rdma_node_get_transport(id->device->node_type)) { - case RDMA_TRANSPORT_IB: - switch (rdma_port_get_link_layer(id->device, id->port_num)) { - case IB_LINK_LAYER_INFINIBAND: - ret = cma_resolve_ib_route(id_priv, timeout_ms); - break; - case IB_LINK_LAYER_ETHERNET: - ret = cma_resolve_iboe_route(id_priv); - break; - default: - ret = -ENOSYS; - } - break; - case RDMA_TRANSPORT_IWARP: - case RDMA_TRANSPORT_SCIF: + if (rdma_cap_ib_sa(id->device, id->port_num)) + ret = cma_resolve_ib_route(id_priv, timeout_ms); + else if (rdma_protocol_roce(id->device, id->port_num)) + ret = cma_resolve_iboe_route(id_priv); + else if (rdma_protocol_iwarp(id->device, id->port_num)) ret = cma_resolve_iw_route(id_priv, timeout_ms); - break; - default: + else ret = -ENOSYS; - break; - } + if (ret) goto err; @@ -2608,6 +2590,10 @@ (struct sockaddr *) &id_priv->id.route.addr.src_addr, ip_addr_size((struct sockaddr *) &id_priv->id.route.addr.src_addr)); #else + SOCK_LOCK(sock); + sock->so_options |= SO_REUSEADDR; + SOCK_UNLOCK(sock); + ret = -sobind(sock, (struct sockaddr *)&id_priv->id.route.addr.src_addr, curthread); @@ -2632,6 +2618,7 @@ static int cma_get_port(struct rdma_id_private *id_priv) { + struct cma_device *cma_dev; struct idr *ps; int ret; @@ -2641,7 +2628,18 @@ break; case RDMA_PS_TCP: ps = &tcp_ps; - if (unify_tcp_port_space) { + + mutex_lock(&lock); + /* check if there are any iWarp IB devices present */ + list_for_each_entry(cma_dev, &dev_list, list) { + if (rdma_protocol_iwarp(cma_dev->device, 1)) { + id_priv->unify_ps_tcp = 1; + break; + } + } + mutex_unlock(&lock); + + if (id_priv->unify_ps_tcp) { ret = cma_get_tcp_port(id_priv); if (ret) goto out; @@ -2713,19 +2711,15 @@ id_priv->backlog = backlog; if (id->device) { - switch (rdma_node_get_transport(id->device->node_type)) { - case RDMA_TRANSPORT_IB: + if (rdma_cap_ib_cm(id->device, 1)) { ret = cma_ib_listen(id_priv); if (ret) goto err; - break; - case RDMA_TRANSPORT_IWARP: - case RDMA_TRANSPORT_SCIF: + } else if (rdma_cap_iw_cm(id->device, 1)) { ret = cma_iw_listen(id_priv, backlog); if (ret) goto err; - break; - default: + } else { ret = -ENOSYS; goto err; } @@ -3094,21 +3088,15 @@ id_priv->srq = conn_param->srq; } - switch (rdma_node_get_transport(id->device->node_type)) { - case RDMA_TRANSPORT_IB: + if (rdma_cap_ib_cm(id->device, id->port_num)) { if (id->qp_type == IB_QPT_UD) ret = cma_resolve_ib_udp(id_priv, conn_param); else ret = cma_connect_ib(id_priv, conn_param); - break; - case RDMA_TRANSPORT_IWARP: - case RDMA_TRANSPORT_SCIF: + } else if (rdma_cap_iw_cm(id->device, id->port_num)) ret = cma_connect_iw(id_priv, conn_param); - break; - default: + else ret = -ENOSYS; - break; - } if (ret) goto err; @@ -3214,8 +3202,7 @@ id_priv->srq = conn_param->srq; } - switch (rdma_node_get_transport(id->device->node_type)) { - case RDMA_TRANSPORT_IB: + if (rdma_cap_ib_cm(id->device, id->port_num)) { if (id->qp_type == IB_QPT_UD) { if (conn_param) ret = cma_send_sidr_rep(id_priv, IB_SIDR_SUCCESS, @@ -3230,15 +3217,10 @@ else ret = cma_rep_recv(id_priv); } - break; - case RDMA_TRANSPORT_IWARP: - case RDMA_TRANSPORT_SCIF: + } else if (rdma_cap_iw_cm(id->device, id->port_num)) ret = cma_accept_iw(id_priv, conn_param); - break; - default: + else ret = -ENOSYS; - break; - } if (ret) goto reject; @@ -3282,8 +3264,7 @@ if (!id_priv->cm_id.ib) return -EINVAL; - switch (rdma_node_get_transport(id->device->node_type)) { - case RDMA_TRANSPORT_IB: + if (rdma_cap_ib_cm(id->device, id->port_num)) { if (id->qp_type == IB_QPT_UD) ret = cma_send_sidr_rep(id_priv, IB_SIDR_REJECT, private_data, private_data_len); @@ -3293,16 +3274,12 @@ IB_CM_REJ_CONSUMER_DEFINED, NULL, 0, private_data, private_data_len); } - break; - case RDMA_TRANSPORT_IWARP: - case RDMA_TRANSPORT_SCIF: + } else if (rdma_cap_iw_cm(id->device, id->port_num)) { ret = iw_cm_reject(id_priv->cm_id.iw, private_data, private_data_len); - break; - default: + } else ret = -ENOSYS; - break; - } + return ret; } EXPORT_SYMBOL(rdma_reject); @@ -3316,8 +3293,7 @@ if (!id_priv->cm_id.ib) return -EINVAL; - switch (rdma_node_get_transport(id->device->node_type)) { - case RDMA_TRANSPORT_IB: + if (rdma_cap_ib_cm(id->device, id->port_num)) { ret = cma_modify_qp_err(id_priv); if (ret) goto out; @@ -3327,15 +3303,11 @@ cma_dbg(id_priv, "sending DREP\n"); ib_send_cm_drep(id_priv->cm_id.ib, NULL, 0); } - break; - case RDMA_TRANSPORT_IWARP: - case RDMA_TRANSPORT_SCIF: + } else if (rdma_cap_iw_cm(id->device, id->port_num)) { ret = iw_cm_disconnect(id_priv->cm_id.iw, 0); - break; - default: + } else ret = -EINVAL; - break; - } + out: return ret; } Index: sys/ofed/drivers/infiniband/core/device.c =================================================================== --- sys/ofed/drivers/infiniband/core/device.c +++ sys/ofed/drivers/infiniband/core/device.c @@ -90,7 +90,8 @@ IB_MANDATORY_FUNC(poll_cq), IB_MANDATORY_FUNC(req_notify_cq), IB_MANDATORY_FUNC(get_dma_mr), - IB_MANDATORY_FUNC(dereg_mr) + IB_MANDATORY_FUNC(dereg_mr), + IB_MANDATORY_FUNC(get_port_immutable) }; int i; @@ -149,13 +150,13 @@ return 0; } -static int start_port(struct ib_device *device) +static int rdma_start_port(struct ib_device *device) { return (device->node_type == RDMA_NODE_IB_SWITCH) ? 0 : 1; } -static int end_port(struct ib_device *device) +static int rdma_end_port(struct ib_device *device) { return (device->node_type == RDMA_NODE_IB_SWITCH) ? 0 : device->phys_port_cnt; @@ -193,6 +194,7 @@ void ib_dealloc_device(struct ib_device *device) { if (device->reg_state == IB_DEV_UNINITIALIZED) { + kfree(device->port_immutable); kfree(device); return; } @@ -225,43 +227,42 @@ return 0; } -static int read_port_table_lengths(struct ib_device *device) +static int verify_immutable(const struct ib_device *dev, u8 port) { - struct ib_port_attr *tprops = NULL; - int num_ports, ret = -ENOMEM; - u8 port_index; - - tprops = kmalloc(sizeof *tprops, GFP_KERNEL); - if (!tprops) - goto out; - - num_ports = end_port(device) - start_port(device) + 1; + return WARN_ON(!rdma_cap_ib_mad(dev, port) && + rdma_max_mad_size(dev, port) != 0); +} - device->pkey_tbl_len = kmalloc(sizeof *device->pkey_tbl_len * num_ports, - GFP_KERNEL); - device->gid_tbl_len = kmalloc(sizeof *device->gid_tbl_len * num_ports, - GFP_KERNEL); - if (!device->pkey_tbl_len || !device->gid_tbl_len) - goto err; +static int read_port_immutable(struct ib_device *device) +{ + int ret; + u8 start_port = rdma_start_port(device); + u8 end_port = rdma_end_port(device); + u8 port; + + /** + * device->port_immutable is indexed directly by the port number to make + * access to this data as efficient as possible. + * + * Therefore port_immutable is declared as a 1 based array with + * potential empty slots at the beginning. + */ + device->port_immutable = kzalloc(sizeof(*device->port_immutable) + * (end_port + 1), + GFP_KERNEL); + if (!device->port_immutable) + return -ENOMEM; - for (port_index = 0; port_index < num_ports; ++port_index) { - ret = ib_query_port(device, port_index + start_port(device), - tprops); + for (port = start_port; port <= end_port; ++port) { + ret = device->get_port_immutable(device, port, + &device->port_immutable[port]); if (ret) - goto err; - device->pkey_tbl_len[port_index] = tprops->pkey_tbl_len; - device->gid_tbl_len[port_index] = tprops->gid_tbl_len; - } - - ret = 0; - goto out; + return ret; -err: - kfree(device->gid_tbl_len); - kfree(device->pkey_tbl_len); -out: - kfree(tprops); - return ret; + if (verify_immutable(device, port)) + return -EINVAL; + } + return 0; } /** @@ -297,10 +298,11 @@ spin_lock_init(&device->event_handler_lock); spin_lock_init(&device->client_data_lock); - ret = read_port_table_lengths(device); + + ret = read_port_immutable(device); if (ret) { - printk(KERN_WARNING "Couldn't create table lengths cache for device %s\n", - device->name); + printk(KERN_WARNING "Couldn't create per port immutable data %s\n", + device->name); goto out; } @@ -308,8 +310,7 @@ if (ret) { printk(KERN_WARNING "Couldn't register device %s with driver model\n", device->name); - kfree(device->gid_tbl_len); - kfree(device->pkey_tbl_len); + kfree(device->port_immutable); goto out; } @@ -351,9 +352,6 @@ list_del(&device->core_list); - kfree(device->gid_tbl_len); - kfree(device->pkey_tbl_len); - mutex_unlock(&device_mutex); ib_device_unregister_sysfs(device); @@ -578,7 +576,7 @@ u8 port_num, struct ib_port_attr *port_attr) { - if (port_num < start_port(device) || port_num > end_port(device)) + if (port_num < rdma_start_port(device) || port_num > rdma_end_port(device)) return -EINVAL; return device->query_port(device, port_num, port_attr); @@ -656,7 +654,7 @@ if (!device->modify_port) return -ENOSYS; - if (port_num < start_port(device) || port_num > end_port(device)) + if (port_num < rdma_start_port(device) || port_num > rdma_end_port(device)) return -EINVAL; return device->modify_port(device, port_num, port_modify_mask, @@ -679,8 +677,8 @@ union ib_gid tmp_gid; int ret, port, i; - for (port = start_port(device); port <= end_port(device); ++port) { - for (i = 0; i < device->gid_tbl_len[port - start_port(device)]; ++i) { + for (port = rdma_start_port(device); port <= rdma_end_port(device); ++port) { + for (i = 0; i < device->port_immutable[port].gid_tbl_len; ++i) { ret = ib_query_gid(device, port, i, &tmp_gid); if (ret) return ret; @@ -712,7 +710,7 @@ u16 tmp_pkey; int partial_ix = -1; - for (i = 0; i < device->pkey_tbl_len[port_num - start_port(device)]; ++i) { + for (i = 0; i < device->port_immutable[port_num].pkey_tbl_len; ++i) { ret = ib_query_pkey(device, port_num, i, &tmp_pkey); if (ret) return ret; Index: sys/ofed/drivers/infiniband/hw/mthca/mthca_provider.c =================================================================== --- sys/ofed/drivers/infiniband/hw/mthca/mthca_provider.c +++ sys/ofed/drivers/infiniband/hw/mthca/mthca_provider.c @@ -1297,6 +1297,25 @@ return err; } +static int mthca_port_immutable(struct ib_device *ibdev, u8 port_num, + struct ib_port_immutable *immutable) +{ + struct ib_port_attr attr; + int err; + + immutable->core_cap_flags = RDMA_CORE_PORT_IBA_IB; + + err = ib_query_port(ibdev, port_num, &attr); + if (err) + return err; + + immutable->pkey_tbl_len = attr.pkey_tbl_len; + immutable->gid_tbl_len = attr.gid_tbl_len; + immutable->max_mad_size = IB_MGMT_MAD_SIZE; + + return 0; +} + int mthca_register_device(struct mthca_dev *dev) { int ret; @@ -1376,6 +1395,7 @@ dev->ib_dev.reg_phys_mr = mthca_reg_phys_mr; dev->ib_dev.reg_user_mr = mthca_reg_user_mr; dev->ib_dev.dereg_mr = mthca_dereg_mr; + dev->ib_dev.get_port_immutable = mthca_port_immutable; if (dev->mthca_flags & MTHCA_FLAG_FMR) { dev->ib_dev.alloc_fmr = mthca_alloc_fmr; Index: sys/ofed/include/rdma/ib_mad.h =================================================================== --- sys/ofed/include/rdma/ib_mad.h +++ sys/ofed/include/rdma/ib_mad.h @@ -134,6 +134,7 @@ IB_MGMT_SA_DATA = 200, IB_MGMT_DEVICE_HDR = 64, IB_MGMT_DEVICE_DATA = 192, + IB_MGMT_MAD_SIZE = IB_MGMT_MAD_HDR + IB_MGMT_MAD_DATA, }; struct ib_mad_hdr { Index: sys/ofed/include/rdma/ib_verbs.h =================================================================== --- sys/ofed/include/rdma/ib_verbs.h +++ sys/ofed/include/rdma/ib_verbs.h @@ -348,6 +348,56 @@ struct iw_protocol_stats iw; }; +/* Define bits for the various functionality this port needs to be supported by + * the core. + */ +/* Management 0x00000FFF */ +#define RDMA_CORE_CAP_IB_MAD 0x00000001 +#define RDMA_CORE_CAP_IB_SMI 0x00000002 +#define RDMA_CORE_CAP_IB_CM 0x00000004 +#define RDMA_CORE_CAP_IW_CM 0x00000008 +#define RDMA_CORE_CAP_IB_SA 0x00000010 +#define RDMA_CORE_CAP_OPA_MAD 0x00000020 + +/* Address format 0x000FF000 */ +#define RDMA_CORE_CAP_AF_IB 0x00001000 +#define RDMA_CORE_CAP_ETH_AH 0x00002000 +#define RDMA_CORE_CAP_OPA_AH 0x00004000 + +/* Protocol 0xFFF00000 */ +#define RDMA_CORE_CAP_PROT_IB 0x00100000 +#define RDMA_CORE_CAP_PROT_ROCE 0x00200000 +#define RDMA_CORE_CAP_PROT_IWARP 0x00400000 +#define RDMA_CORE_CAP_PROT_ROCE_UDP_ENCAP 0x00800000 +#define RDMA_CORE_CAP_PROT_RAW_PACKET 0x01000000 +#define RDMA_CORE_CAP_PROT_USNIC 0x02000000 + +#define RDMA_CORE_PORT_IBA_IB (RDMA_CORE_CAP_PROT_IB \ + | RDMA_CORE_CAP_IB_MAD \ + | RDMA_CORE_CAP_IB_SMI \ + | RDMA_CORE_CAP_IB_CM \ + | RDMA_CORE_CAP_IB_SA \ + | RDMA_CORE_CAP_AF_IB) +#define RDMA_CORE_PORT_IBA_ROCE (RDMA_CORE_CAP_PROT_ROCE \ + | RDMA_CORE_CAP_IB_MAD \ + | RDMA_CORE_CAP_IB_CM \ + | RDMA_CORE_CAP_AF_IB \ + | RDMA_CORE_CAP_ETH_AH) +#define RDMA_CORE_PORT_IBA_ROCE_UDP_ENCAP \ + (RDMA_CORE_CAP_PROT_ROCE_UDP_ENCAP \ + | RDMA_CORE_CAP_IB_MAD \ + | RDMA_CORE_CAP_IB_CM \ + | RDMA_CORE_CAP_AF_IB \ + | RDMA_CORE_CAP_ETH_AH) +#define RDMA_CORE_PORT_IWARP (RDMA_CORE_CAP_PROT_IWARP \ + | RDMA_CORE_CAP_IW_CM) +#define RDMA_CORE_PORT_INTEL_OPA (RDMA_CORE_PORT_IBA_IB \ + | RDMA_CORE_CAP_OPA_MAD) + +#define RDMA_CORE_PORT_RAW_PACKET (RDMA_CORE_CAP_PROT_RAW_PACKET) + +#define RDMA_CORE_PORT_USNIC (RDMA_CORE_CAP_PROT_USNIC) + struct ib_port_attr { enum ib_port_state state; enum ib_mtu max_mtu; @@ -1623,6 +1673,14 @@ }; struct iw_cm_verbs; + +struct ib_port_immutable { + int pkey_tbl_len; + int gid_tbl_len; + u32 core_cap_flags; + u32 max_mad_size; +}; + struct ib_exp_device_attr; struct ib_exp_qp_init_attr; @@ -1639,8 +1697,10 @@ struct list_head client_data_list; struct ib_cache cache; - int *pkey_tbl_len; - int *gid_tbl_len; + /** + * port_immutable is indexed by port number + */ + struct ib_port_immutable *port_immutable; int num_comp_vectors; @@ -1845,6 +1905,15 @@ u32 cmd_n; spinlock_t cmd_perf_lock; + + /** + * The following mandatory functions are used only at device + * registration. Keep functions such as these at the end of this + * structure to avoid cache line misses when accessing struct ib_device + * in fast paths. + */ + int (*get_port_immutable)(struct ib_device *, u8, struct ib_port_immutable *); + /* * Experimental data and functions */ @@ -1928,6 +1997,252 @@ enum rdma_link_layer rdma_port_get_link_layer(struct ib_device *device, u8 port_num); +static inline bool rdma_protocol_ib(const struct ib_device *device, u8 port_num) +{ + return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_PROT_IB; +} + +static inline bool rdma_protocol_roce(const struct ib_device *device, u8 port_num) +{ + return device->port_immutable[port_num].core_cap_flags & + (RDMA_CORE_CAP_PROT_ROCE | RDMA_CORE_CAP_PROT_ROCE_UDP_ENCAP); +} + +static inline bool rdma_protocol_roce_udp_encap(const struct ib_device *device, u8 port_num) +{ + return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_PROT_ROCE_UDP_ENCAP; +} + +static inline bool rdma_protocol_roce_eth_encap(const struct ib_device *device, u8 port_num) +{ + return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_PROT_ROCE; +} + +static inline bool rdma_protocol_iwarp(const struct ib_device *device, u8 port_num) +{ + return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_PROT_IWARP; +} + +static inline bool rdma_ib_or_roce(const struct ib_device *device, u8 port_num) +{ + return rdma_protocol_ib(device, port_num) || + rdma_protocol_roce(device, port_num); +} + +/** + * rdma_cap_ib_mad - Check if the port of a device supports Infiniband + * Management Datagrams. + * @device: Device to check + * @port_num: Port number to check + * + * Management Datagrams (MAD) are a required part of the InfiniBand + * specification and are supported on all InfiniBand devices. A slightly + * extended version are also supported on OPA interfaces. + * + * Return: true if the port supports sending/receiving of MAD packets. + */ +static inline bool rdma_cap_ib_mad(const struct ib_device *device, u8 port_num) +{ + return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_IB_MAD; +} + +/** + * rdma_cap_opa_mad - Check if the port of device provides support for OPA + * Management Datagrams. + * @device: Device to check + * @port_num: Port number to check + * + * Intel OmniPath devices extend and/or replace the InfiniBand Management + * datagrams with their own versions. These OPA MADs share many but not all of + * the characteristics of InfiniBand MADs. + * + * OPA MADs differ in the following ways: + * + * 1) MADs are variable size up to 2K + * IBTA defined MADs remain fixed at 256 bytes + * 2) OPA SMPs must carry valid PKeys + * 3) OPA SMP packets are a different format + * + * Return: true if the port supports OPA MAD packet formats. + */ +static inline bool rdma_cap_opa_mad(struct ib_device *device, u8 port_num) +{ + return (device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_OPA_MAD) + == RDMA_CORE_CAP_OPA_MAD; +} + +/** + * rdma_cap_ib_smi - Check if the port of a device provides an Infiniband + * Subnet Management Agent (SMA) on the Subnet Management Interface (SMI). + * @device: Device to check + * @port_num: Port number to check + * + * Each InfiniBand node is required to provide a Subnet Management Agent + * that the subnet manager can access. Prior to the fabric being fully + * configured by the subnet manager, the SMA is accessed via a well known + * interface called the Subnet Management Interface (SMI). This interface + * uses directed route packets to communicate with the SM to get around the + * chicken and egg problem of the SM needing to know what's on the fabric + * in order to configure the fabric, and needing to configure the fabric in + * order to send packets to the devices on the fabric. These directed + * route packets do not need the fabric fully configured in order to reach + * their destination. The SMI is the only method allowed to send + * directed route packets on an InfiniBand fabric. + * + * Return: true if the port provides an SMI. + */ +static inline bool rdma_cap_ib_smi(const struct ib_device *device, u8 port_num) +{ + return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_IB_SMI; +} + +/** + * rdma_cap_ib_cm - Check if the port of device has the capability Infiniband + * Communication Manager. + * @device: Device to check + * @port_num: Port number to check + * + * The InfiniBand Communication Manager is one of many pre-defined General + * Service Agents (GSA) that are accessed via the General Service + * Interface (GSI). It's role is to facilitate establishment of connections + * between nodes as well as other management related tasks for established + * connections. + * + * Return: true if the port supports an IB CM (this does not guarantee that + * a CM is actually running however). + */ +static inline bool rdma_cap_ib_cm(const struct ib_device *device, u8 port_num) +{ + return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_IB_CM; +} + +/** + * rdma_cap_iw_cm - Check if the port of device has the capability IWARP + * Communication Manager. + * @device: Device to check + * @port_num: Port number to check + * + * Similar to above, but specific to iWARP connections which have a different + * managment protocol than InfiniBand. + * + * Return: true if the port supports an iWARP CM (this does not guarantee that + * a CM is actually running however). + */ +static inline bool rdma_cap_iw_cm(const struct ib_device *device, u8 port_num) +{ + return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_IW_CM; +} + +/** + * rdma_cap_ib_sa - Check if the port of device has the capability Infiniband + * Subnet Administration. + * @device: Device to check + * @port_num: Port number to check + * + * An InfiniBand Subnet Administration (SA) service is a pre-defined General + * Service Agent (GSA) provided by the Subnet Manager (SM). On InfiniBand + * fabrics, devices should resolve routes to other hosts by contacting the + * SA to query the proper route. + * + * Return: true if the port should act as a client to the fabric Subnet + * Administration interface. This does not imply that the SA service is + * running locally. + */ +static inline bool rdma_cap_ib_sa(const struct ib_device *device, u8 port_num) +{ + return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_IB_SA; +} + +/** + * rdma_cap_ib_mcast - Check if the port of device has the capability Infiniband + * Multicast. + * @device: Device to check + * @port_num: Port number to check + * + * InfiniBand multicast registration is more complex than normal IPv4 or + * IPv6 multicast registration. Each Host Channel Adapter must register + * with the Subnet Manager when it wishes to join a multicast group. It + * should do so only once regardless of how many queue pairs it subscribes + * to this group. And it should leave the group only after all queue pairs + * attached to the group have been detached. + * + * Return: true if the port must undertake the additional adminstrative + * overhead of registering/unregistering with the SM and tracking of the + * total number of queue pairs attached to the multicast group. + */ +static inline bool rdma_cap_ib_mcast(const struct ib_device *device, u8 port_num) +{ + return rdma_cap_ib_sa(device, port_num); +} + +/** + * rdma_cap_af_ib - Check if the port of device has the capability + * Native Infiniband Address. + * @device: Device to check + * @port_num: Port number to check + * + * InfiniBand addressing uses a port's GUID + Subnet Prefix to make a default + * GID. RoCE uses a different mechanism, but still generates a GID via + * a prescribed mechanism and port specific data. + * + * Return: true if the port uses a GID address to identify devices on the + * network. + */ +static inline bool rdma_cap_af_ib(const struct ib_device *device, u8 port_num) +{ + return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_AF_IB; +} + +/** + * rdma_cap_eth_ah - Check if the port of device has the capability + * Ethernet Address Handle. + * @device: Device to check + * @port_num: Port number to check + * + * RoCE is InfiniBand over Ethernet, and it uses a well defined technique + * to fabricate GIDs over Ethernet/IP specific addresses native to the + * port. Normally, packet headers are generated by the sending host + * adapter, but when sending connectionless datagrams, we must manually + * inject the proper headers for the fabric we are communicating over. + * + * Return: true if we are running as a RoCE port and must force the + * addition of a Global Route Header built from our Ethernet Address + * Handle into our header list for connectionless packets. + */ +static inline bool rdma_cap_eth_ah(const struct ib_device *device, u8 port_num) +{ + return device->port_immutable[port_num].core_cap_flags & RDMA_CORE_CAP_ETH_AH; +} + +/** + * rdma_max_mad_size - Return the max MAD size required by this RDMA Port. + * + * @device: Device + * @port_num: Port number + * + * This MAD size includes the MAD headers and MAD payload. No other headers + * are included. + * + * Return the max MAD size required by the Port. Will return 0 if the port + * does not support MADs + */ +static inline size_t rdma_max_mad_size(const struct ib_device *device, u8 port_num) +{ + return device->port_immutable[port_num].max_mad_size; +} + +/* + * Check if the device supports READ W/ INVALIDATE. + */ +static inline bool rdma_cap_read_inv(struct ib_device *dev, u32 port_num) +{ + /* + * iWarp drivers must support READ W/ INVALIDATE. No other protocol + * has support for it yet. + */ + return rdma_protocol_iwarp(dev, port_num); +} + int ib_query_gid(struct ib_device *device, u8 port_num, int index, union ib_gid *gid);