diff options
Diffstat (limited to 'drivers/infiniband')
110 files changed, 5015 insertions, 2205 deletions
diff --git a/drivers/infiniband/core/addr.c b/drivers/infiniband/core/addr.c index f4e8185bccd3..a5b4cf030c11 100644 --- a/drivers/infiniband/core/addr.c +++ b/drivers/infiniband/core/addr.c @@ -243,8 +243,7 @@ void rdma_copy_addr(struct rdma_dev_addr *dev_addr, EXPORT_SYMBOL(rdma_copy_addr); int rdma_translate_ip(const struct sockaddr *addr, - struct rdma_dev_addr *dev_addr, - u16 *vlan_id) + struct rdma_dev_addr *dev_addr) { struct net_device *dev; @@ -266,9 +265,6 @@ int rdma_translate_ip(const struct sockaddr *addr, return -EADDRNOTAVAIL; rdma_copy_addr(dev_addr, dev, NULL); - dev_addr->bound_dev_if = dev->ifindex; - if (vlan_id) - *vlan_id = rdma_vlan_dev_vlan_id(dev); dev_put(dev); break; #if IS_ENABLED(CONFIG_IPV6) @@ -279,9 +275,6 @@ int rdma_translate_ip(const struct sockaddr *addr, &((const struct sockaddr_in6 *)addr)->sin6_addr, dev, 1)) { rdma_copy_addr(dev_addr, dev, NULL); - dev_addr->bound_dev_if = dev->ifindex; - if (vlan_id) - *vlan_id = rdma_vlan_dev_vlan_id(dev); break; } } @@ -481,7 +474,7 @@ static int addr_resolve_neigh(struct dst_entry *dst, if (dst->dev->flags & IFF_LOOPBACK) { int ret; - ret = rdma_translate_ip(dst_in, addr, NULL); + ret = rdma_translate_ip(dst_in, addr); if (!ret) memcpy(addr->dst_dev_addr, addr->src_dev_addr, MAX_ADDR_LEN); @@ -558,7 +551,7 @@ static int addr_resolve(struct sockaddr *src_in, } if (ndev->flags & IFF_LOOPBACK) { - ret = rdma_translate_ip(dst_in, addr, NULL); + ret = rdma_translate_ip(dst_in, addr); /* * Put the loopback device and get the translated * device instead. @@ -744,7 +737,6 @@ void rdma_addr_cancel(struct rdma_dev_addr *addr) EXPORT_SYMBOL(rdma_addr_cancel); struct resolve_cb_context { - struct rdma_dev_addr *addr; struct completion comp; int status; }; @@ -752,39 +744,31 @@ struct resolve_cb_context { static void resolve_cb(int status, struct sockaddr *src_addr, struct rdma_dev_addr *addr, void *context) { - if (!status) - memcpy(((struct resolve_cb_context *)context)->addr, - addr, sizeof(struct rdma_dev_addr)); ((struct resolve_cb_context *)context)->status = status; complete(&((struct resolve_cb_context *)context)->comp); } int rdma_addr_find_l2_eth_by_grh(const union ib_gid *sgid, const union ib_gid *dgid, - u8 *dmac, u16 *vlan_id, int *if_index, + u8 *dmac, const struct net_device *ndev, int *hoplimit) { - int ret = 0; struct rdma_dev_addr dev_addr; struct resolve_cb_context ctx; - struct net_device *dev; - union { struct sockaddr _sockaddr; struct sockaddr_in _sockaddr_in; struct sockaddr_in6 _sockaddr_in6; } sgid_addr, dgid_addr; - + int ret; rdma_gid2ip(&sgid_addr._sockaddr, sgid); rdma_gid2ip(&dgid_addr._sockaddr, dgid); memset(&dev_addr, 0, sizeof(dev_addr)); - if (if_index) - dev_addr.bound_dev_if = *if_index; + dev_addr.bound_dev_if = ndev->ifindex; dev_addr.net = &init_net; - ctx.addr = &dev_addr; init_completion(&ctx.comp); ret = rdma_resolve_ip(&self, &sgid_addr._sockaddr, &dgid_addr._sockaddr, &dev_addr, 1000, resolve_cb, &ctx); @@ -798,42 +782,9 @@ int rdma_addr_find_l2_eth_by_grh(const union ib_gid *sgid, return ret; memcpy(dmac, dev_addr.dst_dev_addr, ETH_ALEN); - dev = dev_get_by_index(&init_net, dev_addr.bound_dev_if); - if (!dev) - return -ENODEV; - if (if_index) - *if_index = dev_addr.bound_dev_if; - if (vlan_id) - *vlan_id = rdma_vlan_dev_vlan_id(dev); - if (hoplimit) - *hoplimit = dev_addr.hoplimit; - dev_put(dev); - return ret; -} -EXPORT_SYMBOL(rdma_addr_find_l2_eth_by_grh); - -int rdma_addr_find_smac_by_sgid(union ib_gid *sgid, u8 *smac, u16 *vlan_id) -{ - int ret = 0; - struct rdma_dev_addr dev_addr; - union { - struct sockaddr _sockaddr; - struct sockaddr_in _sockaddr_in; - struct sockaddr_in6 _sockaddr_in6; - } gid_addr; - - rdma_gid2ip(&gid_addr._sockaddr, sgid); - - memset(&dev_addr, 0, sizeof(dev_addr)); - dev_addr.net = &init_net; - ret = rdma_translate_ip(&gid_addr._sockaddr, &dev_addr, vlan_id); - if (ret) - return ret; - - memcpy(smac, dev_addr.src_dev_addr, ETH_ALEN); - return ret; + *hoplimit = dev_addr.hoplimit; + return 0; } -EXPORT_SYMBOL(rdma_addr_find_smac_by_sgid); static int netevent_callback(struct notifier_block *self, unsigned long event, void *ctx) diff --git a/drivers/infiniband/core/cache.c b/drivers/infiniband/core/cache.c index 77515638c55c..e9a409d7f4e2 100644 --- a/drivers/infiniband/core/cache.c +++ b/drivers/infiniband/core/cache.c @@ -573,27 +573,24 @@ static int ib_cache_gid_find_by_filter(struct ib_device *ib_dev, struct ib_gid_attr attr; if (table->data_vec[i].props & GID_TABLE_ENTRY_INVALID) - goto next; + continue; if (memcmp(gid, &table->data_vec[i].gid, sizeof(*gid))) - goto next; + continue; memcpy(&attr, &table->data_vec[i].attr, sizeof(attr)); - if (filter(gid, &attr, context)) + if (filter(gid, &attr, context)) { found = true; - -next: - if (found) + if (index) + *index = i; break; + } } read_unlock_irqrestore(&table->rwlock, flags); if (!found) return -ENOENT; - - if (index) - *index = i; return 0; } @@ -824,12 +821,7 @@ static int gid_table_setup_one(struct ib_device *ib_dev) if (err) return err; - err = roce_rescan_device(ib_dev); - - if (err) { - gid_table_cleanup_one(ib_dev); - gid_table_release_one(ib_dev); - } + rdma_roce_rescan_device(ib_dev); return err; } @@ -883,7 +875,6 @@ int ib_find_gid_by_filter(struct ib_device *device, port_num, filter, context, index); } -EXPORT_SYMBOL(ib_find_gid_by_filter); int ib_get_cached_pkey(struct ib_device *device, u8 port_num, diff --git a/drivers/infiniband/core/cm.c b/drivers/infiniband/core/cm.c index f6b159d79977..b8f8d3128a53 100644 --- a/drivers/infiniband/core/cm.c +++ b/drivers/infiniband/core/cm.c @@ -452,13 +452,14 @@ static void cm_set_private_data(struct cm_id_private *cm_id_priv, cm_id_priv->private_data_len = private_data_len; } -static void cm_init_av_for_response(struct cm_port *port, struct ib_wc *wc, - struct ib_grh *grh, struct cm_av *av) +static int cm_init_av_for_response(struct cm_port *port, struct ib_wc *wc, + struct ib_grh *grh, struct cm_av *av) { av->port = port; av->pkey_index = wc->pkey_index; - ib_init_ah_from_wc(port->cm_dev->ib_device, port->port_num, wc, - grh, &av->ah_attr); + return ib_init_ah_attr_from_wc(port->cm_dev->ib_device, + port->port_num, wc, + grh, &av->ah_attr); } static int cm_init_av_by_path(struct sa_path_rec *path, struct cm_av *av, @@ -494,8 +495,11 @@ static int cm_init_av_by_path(struct sa_path_rec *path, struct cm_av *av, return ret; av->port = port; - ib_init_ah_from_path(cm_dev->ib_device, port->port_num, path, - &av->ah_attr); + ret = ib_init_ah_attr_from_path(cm_dev->ib_device, port->port_num, path, + &av->ah_attr); + if (ret) + return ret; + av->timeout = path->packet_life_time + 1; spin_lock_irqsave(&cm.lock, flags); @@ -1560,6 +1564,37 @@ static u16 cm_get_bth_pkey(struct cm_work *work) return pkey; } +/** + * Convert OPA SGID to IB SGID + * ULPs (such as IPoIB) do not understand OPA GIDs and will + * reject them as the local_gid will not match the sgid. Therefore, + * change the pathrec's SGID to an IB SGID. + * + * @work: Work completion + * @path: Path record + */ +static void cm_opa_to_ib_sgid(struct cm_work *work, + struct sa_path_rec *path) +{ + struct ib_device *dev = work->port->cm_dev->ib_device; + struct ib_gid_attr gid_attr; + u8 port_num = work->port->port_num; + + if (rdma_cap_opa_ah(dev, port_num) && + (ib_is_opa_gid(&path->sgid))) { + union ib_gid sgid; + + if (ib_get_cached_gid(dev, port_num, 0, + &sgid, &gid_attr)) { + dev_warn(&dev->dev, + "Error updating sgid in CM request\n"); + return; + } + + path->sgid = sgid; + } +} + static void cm_format_req_event(struct cm_work *work, struct cm_id_private *cm_id_priv, struct ib_cm_id *listen_id) @@ -1573,10 +1608,13 @@ static void cm_format_req_event(struct cm_work *work, param->bth_pkey = cm_get_bth_pkey(work); param->port = cm_id_priv->av.port->port_num; param->primary_path = &work->path[0]; - if (cm_req_has_alt_path(req_msg)) + cm_opa_to_ib_sgid(work, param->primary_path); + if (cm_req_has_alt_path(req_msg)) { param->alternate_path = &work->path[1]; - else + cm_opa_to_ib_sgid(work, param->alternate_path); + } else { param->alternate_path = NULL; + } param->remote_ca_guid = req_msg->local_ca_guid; param->remote_qkey = be32_to_cpu(req_msg->local_qkey); param->remote_qpn = be32_to_cpu(cm_req_get_local_qpn(req_msg)); @@ -1826,9 +1864,11 @@ static int cm_req_handler(struct cm_work *work) cm_id_priv = container_of(cm_id, struct cm_id_private, id); cm_id_priv->id.remote_id = req_msg->local_comm_id; - cm_init_av_for_response(work->port, work->mad_recv_wc->wc, - work->mad_recv_wc->recv_buf.grh, - &cm_id_priv->av); + ret = cm_init_av_for_response(work->port, work->mad_recv_wc->wc, + work->mad_recv_wc->recv_buf.grh, + &cm_id_priv->av); + if (ret) + goto destroy; cm_id_priv->timewait_info = cm_create_timewait_info(cm_id_priv-> id.local_id); if (IS_ERR(cm_id_priv->timewait_info)) { @@ -1841,9 +1881,10 @@ static int cm_req_handler(struct cm_work *work) listen_cm_id_priv = cm_match_req(work, cm_id_priv); if (!listen_cm_id_priv) { + pr_debug("%s: local_id %d, no listen_cm_id_priv\n", __func__, + be32_to_cpu(cm_id->local_id)); ret = -EINVAL; - kfree(cm_id_priv->timewait_info); - goto destroy; + goto free_timeinfo; } cm_id_priv->id.cm_handler = listen_cm_id_priv->id.cm_handler; @@ -1861,56 +1902,52 @@ static int cm_req_handler(struct cm_work *work) work->port->port_num, grh->sgid_index, &gid, &gid_attr); - if (!ret) { - if (gid_attr.ndev) { - work->path[0].rec_type = - sa_conv_gid_to_pathrec_type(gid_attr.gid_type); - sa_path_set_ifindex(&work->path[0], - gid_attr.ndev->ifindex); - sa_path_set_ndev(&work->path[0], - dev_net(gid_attr.ndev)); + if (ret) { + if (gid_attr.ndev) dev_put(gid_attr.ndev); - } else { - cm_path_set_rec_type(work->port->cm_dev->ib_device, - work->port->port_num, - &work->path[0], - &req_msg->primary_local_gid); - } - if (cm_req_has_alt_path(req_msg)) - work->path[1].rec_type = work->path[0].rec_type; - cm_format_paths_from_req(req_msg, &work->path[0], - &work->path[1]); - if (cm_id_priv->av.ah_attr.type == RDMA_AH_ATTR_TYPE_ROCE) - sa_path_set_dmac(&work->path[0], - cm_id_priv->av.ah_attr.roce.dmac); - work->path[0].hop_limit = grh->hop_limit; - ret = cm_init_av_by_path(&work->path[0], &cm_id_priv->av, - cm_id_priv); + ib_send_cm_rej(cm_id, IB_CM_REJ_UNSUPPORTED, NULL, 0, NULL, 0); + goto rejected; } + + if (gid_attr.ndev) { + work->path[0].rec_type = + sa_conv_gid_to_pathrec_type(gid_attr.gid_type); + sa_path_set_ifindex(&work->path[0], + gid_attr.ndev->ifindex); + sa_path_set_ndev(&work->path[0], + dev_net(gid_attr.ndev)); + dev_put(gid_attr.ndev); + } else { + cm_path_set_rec_type(work->port->cm_dev->ib_device, + work->port->port_num, + &work->path[0], + &req_msg->primary_local_gid); + } + if (cm_req_has_alt_path(req_msg)) + work->path[1].rec_type = work->path[0].rec_type; + cm_format_paths_from_req(req_msg, &work->path[0], + &work->path[1]); + if (cm_id_priv->av.ah_attr.type == RDMA_AH_ATTR_TYPE_ROCE) + sa_path_set_dmac(&work->path[0], + cm_id_priv->av.ah_attr.roce.dmac); + work->path[0].hop_limit = grh->hop_limit; + ret = cm_init_av_by_path(&work->path[0], &cm_id_priv->av, + cm_id_priv); if (ret) { - int err = ib_get_cached_gid(work->port->cm_dev->ib_device, - work->port->port_num, 0, - &work->path[0].sgid, - &gid_attr); - if (!err && gid_attr.ndev) { - work->path[0].rec_type = - sa_conv_gid_to_pathrec_type(gid_attr.gid_type); - sa_path_set_ifindex(&work->path[0], - gid_attr.ndev->ifindex); - sa_path_set_ndev(&work->path[0], - dev_net(gid_attr.ndev)); - dev_put(gid_attr.ndev); - } else { - cm_path_set_rec_type(work->port->cm_dev->ib_device, - work->port->port_num, - &work->path[0], - &req_msg->primary_local_gid); - } - if (cm_req_has_alt_path(req_msg)) - work->path[1].rec_type = work->path[0].rec_type; - ib_send_cm_rej(cm_id, IB_CM_REJ_INVALID_GID, - &work->path[0].sgid, sizeof work->path[0].sgid, - NULL, 0); + int err; + + err = ib_get_cached_gid(work->port->cm_dev->ib_device, + work->port->port_num, 0, + &work->path[0].sgid, + NULL); + if (err) + ib_send_cm_rej(cm_id, IB_CM_REJ_INVALID_GID, + NULL, 0, NULL, 0); + else + ib_send_cm_rej(cm_id, IB_CM_REJ_INVALID_GID, + &work->path[0].sgid, + sizeof(work->path[0].sgid), + NULL, 0); goto rejected; } if (cm_req_has_alt_path(req_msg)) { @@ -1919,7 +1956,7 @@ static int cm_req_handler(struct cm_work *work) if (ret) { ib_send_cm_rej(cm_id, IB_CM_REJ_INVALID_ALT_GID, &work->path[0].sgid, - sizeof work->path[0].sgid, NULL, 0); + sizeof(work->path[0].sgid), NULL, 0); goto rejected; } } @@ -1945,6 +1982,8 @@ static int cm_req_handler(struct cm_work *work) rejected: atomic_dec(&cm_id_priv->refcount); cm_deref_id(listen_cm_id_priv); +free_timeinfo: + kfree(cm_id_priv->timewait_info); destroy: ib_destroy_cm_id(cm_id); return ret; @@ -1997,6 +2036,8 @@ int ib_send_cm_rep(struct ib_cm_id *cm_id, spin_lock_irqsave(&cm_id_priv->lock, flags); if (cm_id->state != IB_CM_REQ_RCVD && cm_id->state != IB_CM_MRA_REQ_SENT) { + pr_debug("%s: local_comm_id %d, cm_id->state: %d\n", __func__, + be32_to_cpu(cm_id_priv->id.local_id), cm_id->state); ret = -EINVAL; goto out; } @@ -2063,6 +2104,8 @@ int ib_send_cm_rtu(struct ib_cm_id *cm_id, spin_lock_irqsave(&cm_id_priv->lock, flags); if (cm_id->state != IB_CM_REP_RCVD && cm_id->state != IB_CM_MRA_REP_SENT) { + pr_debug("%s: local_id %d, cm_id->state %d\n", __func__, + be32_to_cpu(cm_id->local_id), cm_id->state); ret = -EINVAL; goto error; } @@ -2170,6 +2213,8 @@ static int cm_rep_handler(struct cm_work *work) cm_id_priv = cm_acquire_id(rep_msg->remote_comm_id, 0); if (!cm_id_priv) { cm_dup_rep_handler(work); + pr_debug("%s: remote_comm_id %d, no cm_id_priv\n", __func__, + be32_to_cpu(rep_msg->remote_comm_id)); return -EINVAL; } @@ -2183,6 +2228,10 @@ static int cm_rep_handler(struct cm_work *work) default: spin_unlock_irq(&cm_id_priv->lock); ret = -EINVAL; + pr_debug("%s: cm_id_priv->id.state: %d, local_comm_id %d, remote_comm_id %d\n", + __func__, cm_id_priv->id.state, + be32_to_cpu(rep_msg->local_comm_id), + be32_to_cpu(rep_msg->remote_comm_id)); goto error; } @@ -2196,6 +2245,8 @@ static int cm_rep_handler(struct cm_work *work) spin_unlock(&cm.lock); spin_unlock_irq(&cm_id_priv->lock); ret = -EINVAL; + pr_debug("%s: Failed to insert remote id %d\n", __func__, + be32_to_cpu(rep_msg->remote_comm_id)); goto error; } /* Check for a stale connection. */ @@ -2213,6 +2264,10 @@ static int cm_rep_handler(struct cm_work *work) IB_CM_REJ_STALE_CONN, CM_MSG_RESPONSE_REP, NULL, 0); ret = -EINVAL; + pr_debug("%s: Stale connection. local_comm_id %d, remote_comm_id %d\n", + __func__, be32_to_cpu(rep_msg->local_comm_id), + be32_to_cpu(rep_msg->remote_comm_id)); + if (cur_cm_id_priv) { cm_id = &cur_cm_id_priv->id; ib_send_cm_dreq(cm_id, NULL, 0); @@ -2359,6 +2414,8 @@ int ib_send_cm_dreq(struct ib_cm_id *cm_id, cm_id_priv = container_of(cm_id, struct cm_id_private, id); spin_lock_irqsave(&cm_id_priv->lock, flags); if (cm_id->state != IB_CM_ESTABLISHED) { + pr_debug("%s: local_id %d, cm_id->state: %d\n", __func__, + be32_to_cpu(cm_id->local_id), cm_id->state); ret = -EINVAL; goto out; } @@ -2428,6 +2485,8 @@ int ib_send_cm_drep(struct ib_cm_id *cm_id, if (cm_id->state != IB_CM_DREQ_RCVD) { spin_unlock_irqrestore(&cm_id_priv->lock, flags); kfree(data); + pr_debug("%s: local_id %d, cm_idcm_id->state(%d) != IB_CM_DREQ_RCVD\n", + __func__, be32_to_cpu(cm_id->local_id), cm_id->state); return -EINVAL; } @@ -2493,6 +2552,9 @@ static int cm_dreq_handler(struct cm_work *work) atomic_long_inc(&work->port->counter_group[CM_RECV_DUPLICATES]. counter[CM_DREQ_COUNTER]); cm_issue_drep(work->port, work->mad_recv_wc); + pr_debug("%s: no cm_id_priv, local_comm_id %d, remote_comm_id %d\n", + __func__, be32_to_cpu(dreq_msg->local_comm_id), + be32_to_cpu(dreq_msg->remote_comm_id)); return -EINVAL; } @@ -2535,6 +2597,9 @@ static int cm_dreq_handler(struct cm_work *work) counter[CM_DREQ_COUNTER]); goto unlock; default: + pr_debug("%s: local_id %d, cm_id_priv->id.state: %d\n", + __func__, be32_to_cpu(cm_id_priv->id.local_id), + cm_id_priv->id.state); goto unlock; } cm_id_priv->id.state = IB_CM_DREQ_RCVD; @@ -2638,6 +2703,8 @@ int ib_send_cm_rej(struct ib_cm_id *cm_id, cm_enter_timewait(cm_id_priv); break; default: + pr_debug("%s: local_id %d, cm_id->state: %d\n", __func__, + be32_to_cpu(cm_id_priv->id.local_id), cm_id->state); ret = -EINVAL; goto out; } @@ -2748,6 +2815,9 @@ static int cm_rej_handler(struct cm_work *work) /* fall through */ default: spin_unlock_irq(&cm_id_priv->lock); + pr_debug("%s: local_id %d, cm_id_priv->id.state: %d\n", + __func__, be32_to_cpu(cm_id_priv->id.local_id), + cm_id_priv->id.state); ret = -EINVAL; goto out; } @@ -2811,6 +2881,9 @@ int ib_send_cm_mra(struct ib_cm_id *cm_id, } /* fall through */ default: + pr_debug("%s: local_id %d, cm_id_priv->id.state: %d\n", + __func__, be32_to_cpu(cm_id_priv->id.local_id), + cm_id_priv->id.state); ret = -EINVAL; goto error1; } @@ -2912,6 +2985,9 @@ static int cm_mra_handler(struct cm_work *work) counter[CM_MRA_COUNTER]); /* fall through */ default: + pr_debug("%s local_id %d, cm_id_priv->id.state: %d\n", + __func__, be32_to_cpu(cm_id_priv->id.local_id), + cm_id_priv->id.state); goto out; } @@ -3085,6 +3161,12 @@ static int cm_lap_handler(struct cm_work *work) if (!cm_id_priv) return -EINVAL; + ret = cm_init_av_for_response(work->port, work->mad_recv_wc->wc, + work->mad_recv_wc->recv_buf.grh, + &cm_id_priv->av); + if (ret) + goto deref; + param = &work->cm_event.param.lap_rcvd; memset(&work->path[0], 0, sizeof(work->path[1])); cm_path_set_rec_type(work->port->cm_dev->ib_device, @@ -3131,9 +3213,6 @@ static int cm_lap_handler(struct cm_work *work) cm_id_priv->id.lap_state = IB_CM_LAP_RCVD; cm_id_priv->tid = lap_msg->hdr.tid; - cm_init_av_for_response(work->port, work->mad_recv_wc->wc, - work->mad_recv_wc->recv_buf.grh, - &cm_id_priv->av); cm_init_av_by_path(param->alternate_path, &cm_id_priv->alt_av, cm_id_priv); ret = atomic_inc_and_test(&cm_id_priv->work_count); @@ -3386,6 +3465,7 @@ static int cm_sidr_req_handler(struct cm_work *work) struct cm_id_private *cm_id_priv, *cur_cm_id_priv; struct cm_sidr_req_msg *sidr_req_msg; struct ib_wc *wc; + int ret; cm_id = ib_create_cm_id(work->port->cm_dev->ib_device, NULL, NULL); if (IS_ERR(cm_id)) @@ -3398,9 +3478,12 @@ static int cm_sidr_req_handler(struct cm_work *work) wc = work->mad_recv_wc->wc; cm_id_priv->av.dgid.global.subnet_prefix = cpu_to_be64(wc->slid); cm_id_priv->av.dgid.global.interface_id = 0; - cm_init_av_for_response(work->port, work->mad_recv_wc->wc, - work->mad_recv_wc->recv_buf.grh, - &cm_id_priv->av); + ret = cm_init_av_for_response(work->port, work->mad_recv_wc->wc, + work->mad_recv_wc->recv_buf.grh, + &cm_id_priv->av); + if (ret) + goto out; + cm_id_priv->id.remote_id = sidr_req_msg->request_id; cm_id_priv->tid = sidr_req_msg->hdr.tid; atomic_inc(&cm_id_priv->work_count); @@ -3692,6 +3775,7 @@ static void cm_work_handler(struct work_struct *_work) ret = cm_timewait_handler(work); break; default: + pr_debug("cm_event.event: 0x%x\n", work->cm_event.event); ret = -EINVAL; break; } @@ -3727,6 +3811,8 @@ static int cm_establish(struct ib_cm_id *cm_id) ret = -EISCONN; break; default: + pr_debug("%s: local_id %d, cm_id->state: %d\n", __func__, + be32_to_cpu(cm_id->local_id), cm_id->state); ret = -EINVAL; break; } @@ -3924,6 +4010,9 @@ static int cm_init_qp_init_attr(struct cm_id_private *cm_id_priv, ret = 0; break; default: + pr_debug("%s: local_id %d, cm_id_priv->id.state: %d\n", + __func__, be32_to_cpu(cm_id_priv->id.local_id), + cm_id_priv->id.state); ret = -EINVAL; break; } @@ -3971,6 +4060,9 @@ static int cm_init_qp_rtr_attr(struct cm_id_private *cm_id_priv, ret = 0; break; default: + pr_debug("%s: local_id %d, cm_id_priv->id.state: %d\n", + __func__, be32_to_cpu(cm_id_priv->id.local_id), + cm_id_priv->id.state); ret = -EINVAL; break; } @@ -4030,6 +4122,9 @@ static int cm_init_qp_rts_attr(struct cm_id_private *cm_id_priv, ret = 0; break; default: + pr_debug("%s: local_id %d, cm_id_priv->id.state: %d\n", + __func__, be32_to_cpu(cm_id_priv->id.local_id), + cm_id_priv->id.state); ret = -EINVAL; break; } diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index 6294a7001d33..65c55f79444a 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -601,7 +601,7 @@ static int cma_translate_addr(struct sockaddr *addr, struct rdma_dev_addr *dev_a int ret; if (addr->sa_family != AF_IB) { - ret = rdma_translate_ip(addr, dev_addr, NULL); + ret = rdma_translate_ip(addr, dev_addr); } else { cma_translate_ib((struct sockaddr_ib *) addr, dev_addr); ret = 0; @@ -2132,7 +2132,7 @@ static int iw_conn_req_handler(struct iw_cm_id *cm_id, mutex_lock_nested(&conn_id->handler_mutex, SINGLE_DEPTH_NESTING); conn_id->state = RDMA_CM_CONNECT; - ret = rdma_translate_ip(laddr, &conn_id->id.route.addr.dev_addr, NULL); + ret = rdma_translate_ip(laddr, &conn_id->id.route.addr.dev_addr); if (ret) { mutex_unlock(&conn_id->handler_mutex); rdma_destroy_id(new_cm_id); @@ -2414,6 +2414,26 @@ out: kfree(work); } +static void cma_init_resolve_route_work(struct cma_work *work, + struct rdma_id_private *id_priv) +{ + work->id = id_priv; + INIT_WORK(&work->work, cma_work_handler); + work->old_state = RDMA_CM_ROUTE_QUERY; + work->new_state = RDMA_CM_ROUTE_RESOLVED; + work->event.event = RDMA_CM_EVENT_ROUTE_RESOLVED; +} + +static void cma_init_resolve_addr_work(struct cma_work *work, + struct rdma_id_private *id_priv) +{ + work->id = id_priv; + INIT_WORK(&work->work, cma_work_handler); + work->old_state = RDMA_CM_ADDR_QUERY; + work->new_state = RDMA_CM_ADDR_RESOLVED; + work->event.event = RDMA_CM_EVENT_ADDR_RESOLVED; +} + static int cma_resolve_ib_route(struct rdma_id_private *id_priv, int timeout_ms) { struct rdma_route *route = &id_priv->id.route; @@ -2424,11 +2444,7 @@ static int cma_resolve_ib_route(struct rdma_id_private *id_priv, int timeout_ms) if (!work) return -ENOMEM; - work->id = id_priv; - INIT_WORK(&work->work, cma_work_handler); - work->old_state = RDMA_CM_ROUTE_QUERY; - work->new_state = RDMA_CM_ROUTE_RESOLVED; - work->event.event = RDMA_CM_EVENT_ROUTE_RESOLVED; + cma_init_resolve_route_work(work, id_priv); route->path_rec = kmalloc(sizeof *route->path_rec, GFP_KERNEL); if (!route->path_rec) { @@ -2483,11 +2499,7 @@ static int cma_resolve_iw_route(struct rdma_id_private *id_priv, int timeout_ms) if (!work) return -ENOMEM; - work->id = id_priv; - INIT_WORK(&work->work, cma_work_handler); - work->old_state = RDMA_CM_ROUTE_QUERY; - work->new_state = RDMA_CM_ROUTE_RESOLVED; - work->event.event = RDMA_CM_EVENT_ROUTE_RESOLVED; + cma_init_resolve_route_work(work, id_priv); queue_work(cma_wq, &work->work); return 0; } @@ -2528,8 +2540,10 @@ static int cma_resolve_iboe_route(struct rdma_id_private *id_priv) struct rdma_addr *addr = &route->addr; struct cma_work *work; int ret; - struct net_device *ndev = NULL; - enum ib_gid_type gid_type = IB_GID_TYPE_IB; + struct net_device *ndev; + enum ib_gid_type gid_type = IB_GID_TYPE_ROCE; + unsigned long supported_gids; + u8 default_roce_tos = id_priv->cma_dev->default_roce_tos[id_priv->id.port_num - rdma_start_port(id_priv->cma_dev->device)]; u8 tos = id_priv->tos_set ? id_priv->tos : default_roce_tos; @@ -2539,9 +2553,6 @@ static int cma_resolve_iboe_route(struct rdma_id_private *id_priv) if (!work) return -ENOMEM; - work->id = id_priv; - INIT_WORK(&work->work, cma_work_handler); - route->path_rec = kzalloc(sizeof *route->path_rec, GFP_KERNEL); if (!route->path_rec) { ret = -ENOMEM; @@ -2550,30 +2561,29 @@ static int cma_resolve_iboe_route(struct rdma_id_private *id_priv) route->num_paths = 1; - if (addr->dev_addr.bound_dev_if) { - unsigned long supported_gids; - - ndev = dev_get_by_index(&init_net, addr->dev_addr.bound_dev_if); - if (!ndev) { - ret = -ENODEV; - goto err2; - } - - supported_gids = roce_gid_type_mask_support(id_priv->id.device, - id_priv->id.port_num); - gid_type = cma_route_gid_type(addr->dev_addr.network, - supported_gids, - id_priv->gid_type); - route->path_rec->rec_type = - sa_conv_gid_to_pathrec_type(gid_type); - sa_path_set_ndev(route->path_rec, &init_net); - sa_path_set_ifindex(route->path_rec, ndev->ifindex); + if (!addr->dev_addr.bound_dev_if) { + ret = -ENODEV; + goto err2; } + + ndev = dev_get_by_index(&init_net, addr->dev_addr.bound_dev_if); if (!ndev) { ret = -ENODEV; goto err2; } + supported_gids = roce_gid_type_mask_support(id_priv->id.device, + id_priv->id.port_num); + gid_type = cma_route_gid_type(addr->dev_addr.network, + supported_gids, + id_priv->gid_type); + /* Use the hint from IP Stack to select GID Type */ + if (gid_type < ib_network_to_gid_type(addr->dev_addr.network)) + gid_type = ib_network_to_gid_type(addr->dev_addr.network); + route->path_rec->rec_type = sa_conv_gid_to_pathrec_type(gid_type); + + sa_path_set_ndev(route->path_rec, &init_net); + sa_path_set_ifindex(route->path_rec, ndev->ifindex); sa_path_set_dmac(route->path_rec, addr->dev_addr.dst_dev_addr); rdma_ip2gid((struct sockaddr *)&id_priv->id.route.addr.src_addr, @@ -2581,11 +2591,6 @@ static int cma_resolve_iboe_route(struct rdma_id_private *id_priv) rdma_ip2gid((struct sockaddr *)&id_priv->id.route.addr.dst_addr, &route->path_rec->dgid); - /* Use the hint from IP Stack to select GID Type */ - if (gid_type < ib_network_to_gid_type(addr->dev_addr.network)) - gid_type = ib_network_to_gid_type(addr->dev_addr.network); - route->path_rec->rec_type = sa_conv_gid_to_pathrec_type(gid_type); - if (((struct sockaddr *)&id_priv->id.route.addr.dst_addr)->sa_family != AF_IB) /* TODO: get the hoplimit from the inet/inet6 device */ route->path_rec->hop_limit = addr->dev_addr.hoplimit; @@ -2607,11 +2612,7 @@ static int cma_resolve_iboe_route(struct rdma_id_private *id_priv) goto err2; } - work->old_state = RDMA_CM_ROUTE_QUERY; - work->new_state = RDMA_CM_ROUTE_RESOLVED; - work->event.event = RDMA_CM_EVENT_ROUTE_RESOLVED; - work->event.status = 0; - + cma_init_resolve_route_work(work, id_priv); queue_work(cma_wq, &work->work); return 0; @@ -2791,11 +2792,7 @@ static int cma_resolve_loopback(struct rdma_id_private *id_priv) rdma_addr_get_sgid(&id_priv->id.route.addr.dev_addr, &gid); rdma_addr_set_dgid(&id_priv->id.route.addr.dev_addr, &gid); - work->id = id_priv; - INIT_WORK(&work->work, cma_work_handler); - work->old_state = RDMA_CM_ADDR_QUERY; - work->new_state = RDMA_CM_ADDR_RESOLVED; - work->event.event = RDMA_CM_EVENT_ADDR_RESOLVED; + cma_init_resolve_addr_work(work, id_priv); queue_work(cma_wq, &work->work); return 0; err: @@ -2821,11 +2818,7 @@ static int cma_resolve_ib_addr(struct rdma_id_private *id_priv) rdma_addr_set_dgid(&id_priv->id.route.addr.dev_addr, (union ib_gid *) &(((struct sockaddr_ib *) &id_priv->id.route.addr.dst_addr)->sib_addr)); - work->id = id_priv; - INIT_WORK(&work->work, cma_work_handler); - work->old_state = RDMA_CM_ADDR_QUERY; - work->new_state = RDMA_CM_ADDR_RESOLVED; - work->event.event = RDMA_CM_EVENT_ADDR_RESOLVED; + cma_init_resolve_addr_work(work, id_priv); queue_work(cma_wq, &work->work); return 0; err: @@ -3404,9 +3397,10 @@ static int cma_sidr_rep_handler(struct ib_cm_id *cm_id, event.status = ret; break; } - ib_init_ah_from_path(id_priv->id.device, id_priv->id.port_num, - id_priv->id.route.path_rec, - &event.param.ud.ah_attr); + ib_init_ah_attr_from_path(id_priv->id.device, + id_priv->id.port_num, + id_priv->id.route.path_rec, + &event.param.ud.ah_attr); event.param.ud.qp_num = rep->qpn; event.param.ud.qkey = rep->qkey; event.event = RDMA_CM_EVENT_ESTABLISHED; @@ -4010,8 +4004,10 @@ static void cma_iboe_set_mgid(struct sockaddr *addr, union ib_gid *mgid, } else if (addr->sa_family == AF_INET6) { memcpy(mgid, &sin6->sin6_addr, sizeof *mgid); } else { - mgid->raw[0] = (gid_type == IB_GID_TYPE_IB) ? 0xff : 0; - mgid->raw[1] = (gid_type == IB_GID_TYPE_IB) ? 0x0e : 0; + mgid->raw[0] = + (gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP) ? 0 : 0xff; + mgid->raw[1] = + (gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP) ? 0 : 0x0e; mgid->raw[2] = 0; mgid->raw[3] = 0; mgid->raw[4] = 0; @@ -4432,7 +4428,7 @@ static int cma_get_id_stats(struct sk_buff *skb, struct netlink_callback *cb) RDMA_NL_RDMA_CM_ATTR_SRC_ADDR)) goto out; if (ibnl_put_attr(skb, nlh, - rdma_addr_size(cma_src_addr(id_priv)), + rdma_addr_size(cma_dst_addr(id_priv)), cma_dst_addr(id_priv), RDMA_NL_RDMA_CM_ATTR_DST_ADDR)) goto out; @@ -4444,6 +4440,7 @@ static int cma_get_id_stats(struct sk_buff *skb, struct netlink_callback *cb) id_stats->qp_type = id->qp_type; i_id++; + nlmsg_end(skb, nlh); } cb->args[1] = 0; diff --git a/drivers/infiniband/core/core_priv.h b/drivers/infiniband/core/core_priv.h index 66f0268f37a6..74091bdca245 100644 --- a/drivers/infiniband/core/core_priv.h +++ b/drivers/infiniband/core/core_priv.h @@ -137,7 +137,6 @@ int ib_cache_gid_del_all_netdev_gids(struct ib_device *ib_dev, u8 port, int roce_gid_mgmt_init(void); void roce_gid_mgmt_cleanup(void); -int roce_rescan_device(struct ib_device *ib_dev); unsigned long roce_gid_type_mask_support(struct ib_device *ib_dev, u8 port); int ib_cache_setup_one(struct ib_device *device); @@ -191,13 +190,6 @@ void ib_sa_cleanup(void); int rdma_nl_init(void); void rdma_nl_exit(void); -/** - * 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. - */ -int ibnl_chk_listeners(unsigned int group); - int ib_nl_handle_resolve_resp(struct sk_buff *skb, struct nlmsghdr *nlh, struct netlink_ext_ack *extack); @@ -213,11 +205,6 @@ int ib_get_cached_subnet_prefix(struct ib_device *device, 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, @@ -240,14 +227,6 @@ int ib_mad_agent_security_setup(struct ib_mad_agent *agent, 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) { } diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c index 465520627e4b..2826e06311a5 100644 --- a/drivers/infiniband/core/device.c +++ b/drivers/infiniband/core/device.c @@ -288,7 +288,7 @@ void ib_dealloc_device(struct ib_device *device) { WARN_ON(device->reg_state != IB_DEV_UNREGISTERED && device->reg_state != IB_DEV_UNINITIALIZED); - kobject_put(&device->dev.kobj); + put_device(&device->dev); } EXPORT_SYMBOL(ib_dealloc_device); @@ -462,7 +462,6 @@ int ib_register_device(struct ib_device *device, struct ib_udata uhw = {.outlen = 0, .inlen = 0}; struct device *parent = device->dev.parent; - WARN_ON_ONCE(!parent); WARN_ON_ONCE(device->dma_device); if (device->dev.dma_ops) { /* @@ -471,16 +470,25 @@ int ib_register_device(struct ib_device *device, * into device->dev. */ device->dma_device = &device->dev; - if (!device->dev.dma_mask) - device->dev.dma_mask = parent->dma_mask; - if (!device->dev.coherent_dma_mask) - device->dev.coherent_dma_mask = - parent->coherent_dma_mask; + if (!device->dev.dma_mask) { + if (parent) + device->dev.dma_mask = parent->dma_mask; + else + WARN_ON_ONCE(true); + } + if (!device->dev.coherent_dma_mask) { + if (parent) + device->dev.coherent_dma_mask = + parent->coherent_dma_mask; + else + WARN_ON_ONCE(true); + } } else { /* * The caller did not provide custom DMA operations. Use the * DMA mapping operations of the parent device. */ + WARN_ON_ONCE(!parent); device->dma_device = parent; } @@ -1033,32 +1041,22 @@ EXPORT_SYMBOL(ib_modify_port); /** * ib_find_gid - Returns the port number and GID table index where - * a specified GID value occurs. + * a specified GID value occurs. Its searches only for IB link layer. * @device: The device to query. * @gid: The GID value to search for. - * @gid_type: Type of GID. * @ndev: The ndev related to the GID to search for. * @port_num: The port number of the device where the GID value was found. * @index: The index into the GID table where the GID was found. This * parameter may be NULL. */ int ib_find_gid(struct ib_device *device, union ib_gid *gid, - enum ib_gid_type gid_type, struct net_device *ndev, - u8 *port_num, u16 *index) + struct net_device *ndev, u8 *port_num, u16 *index) { union ib_gid tmp_gid; int ret, port, i; for (port = rdma_start_port(device); port <= rdma_end_port(device); ++port) { - if (rdma_cap_roce_gid_table(device, port)) { - if (!ib_find_cached_gid_by_port(device, gid, gid_type, port, - ndev, index)) { - *port_num = port; - return 0; - } - } - - if (gid_type != IB_GID_TYPE_IB) + if (rdma_cap_roce_gid_table(device, port)) continue; for (i = 0; i < device->port_immutable[port].gid_tbl_len; ++i) { diff --git a/drivers/infiniband/core/iwpm_util.c b/drivers/infiniband/core/iwpm_util.c index 3c4faadb8cdd..81528f64061a 100644 --- a/drivers/infiniband/core/iwpm_util.c +++ b/drivers/infiniband/core/iwpm_util.c @@ -654,6 +654,7 @@ int iwpm_send_mapinfo(u8 nl_client, int iwpm_pid) } skb_num++; spin_lock_irqsave(&iwpm_mapinfo_lock, flags); + ret = -EINVAL; for (i = 0; i < IWPM_MAPINFO_HASH_SIZE; i++) { hlist_for_each_entry(map_info, &iwpm_hash_bucket[i], hlist_node) { diff --git a/drivers/infiniband/core/mad.c b/drivers/infiniband/core/mad.c index cb91245e9163..c50596f7f98a 100644 --- a/drivers/infiniband/core/mad.c +++ b/drivers/infiniband/core/mad.c @@ -49,7 +49,6 @@ #include "smi.h" #include "opa_smi.h" #include "agent.h" -#include "core_priv.h" static int mad_sendq_size = IB_MAD_QP_SEND_SIZE; static int mad_recvq_size = IB_MAD_QP_RECV_SIZE; diff --git a/drivers/infiniband/core/netlink.c b/drivers/infiniband/core/netlink.c index 1fb72c356e36..3ccaae18ad75 100644 --- a/drivers/infiniband/core/netlink.c +++ b/drivers/infiniband/core/netlink.c @@ -41,8 +41,6 @@ #include <linux/module.h> #include "core_priv.h" -#include "core_priv.h" - static DEFINE_MUTEX(rdma_nl_mutex); static struct sock *nls; static struct { @@ -83,15 +81,13 @@ static bool is_nl_valid(unsigned int type, unsigned int op) if (!is_nl_msg_valid(type, op)) return false; - cb_table = rdma_nl_types[type].cb_table; -#ifdef CONFIG_MODULES - if (!cb_table) { + if (!rdma_nl_types[type].cb_table) { mutex_unlock(&rdma_nl_mutex); request_module("rdma-netlink-subsys-%d", type); mutex_lock(&rdma_nl_mutex); - cb_table = rdma_nl_types[type].cb_table; } -#endif + + cb_table = rdma_nl_types[type].cb_table; if (!cb_table || (!cb_table[op].dump && !cb_table[op].doit)) return false; diff --git a/drivers/infiniband/core/nldev.c b/drivers/infiniband/core/nldev.c index 0dcd1aa6f683..5d790c507c7e 100644 --- a/drivers/infiniband/core/nldev.c +++ b/drivers/infiniband/core/nldev.c @@ -54,14 +54,23 @@ static const struct nla_policy nldev_policy[RDMA_NLDEV_ATTR_MAX] = { [RDMA_NLDEV_ATTR_DEV_NODE_TYPE] = { .type = NLA_U8 }, }; -static int fill_dev_info(struct sk_buff *msg, struct ib_device *device) +static int fill_nldev_handle(struct sk_buff *msg, struct ib_device *device) { - char fw[IB_FW_VERSION_NAME_MAX]; - if (nla_put_u32(msg, RDMA_NLDEV_ATTR_DEV_INDEX, device->index)) return -EMSGSIZE; if (nla_put_string(msg, RDMA_NLDEV_ATTR_DEV_NAME, device->name)) return -EMSGSIZE; + + return 0; +} + +static int fill_dev_info(struct sk_buff *msg, struct ib_device *device) +{ + char fw[IB_FW_VERSION_NAME_MAX]; + + if (fill_nldev_handle(msg, device)) + return -EMSGSIZE; + if (nla_put_u32(msg, RDMA_NLDEV_ATTR_PORT_INDEX, rdma_end_port(device))) return -EMSGSIZE; @@ -92,10 +101,9 @@ static int fill_port_info(struct sk_buff *msg, struct ib_port_attr attr; int ret; - if (nla_put_u32(msg, RDMA_NLDEV_ATTR_DEV_INDEX, device->index)) - return -EMSGSIZE; - if (nla_put_string(msg, RDMA_NLDEV_ATTR_DEV_NAME, device->name)) + if (fill_nldev_handle(msg, device)) return -EMSGSIZE; + if (nla_put_u32(msg, RDMA_NLDEV_ATTR_PORT_INDEX, port)) return -EMSGSIZE; diff --git a/drivers/infiniband/core/roce_gid_mgmt.c b/drivers/infiniband/core/roce_gid_mgmt.c index 90e3889b7fbe..5a52ec77940a 100644 --- a/drivers/infiniband/core/roce_gid_mgmt.c +++ b/drivers/infiniband/core/roce_gid_mgmt.c @@ -410,15 +410,18 @@ static void enum_all_gids_of_dev_cb(struct ib_device *ib_dev, rtnl_unlock(); } -/* This function will rescan all of the network devices in the system - * and add their gids, as needed, to the relevant RoCE devices. */ -int roce_rescan_device(struct ib_device *ib_dev) +/** + * rdma_roce_rescan_device - Rescan all of the network devices in the system + * and add their gids, as needed, to the relevant RoCE devices. + * + * @device: the rdma device + */ +void rdma_roce_rescan_device(struct ib_device *ib_dev) { ib_enum_roce_netdev(ib_dev, pass_all_filter, NULL, enum_all_gids_of_dev_cb, NULL); - - return 0; } +EXPORT_SYMBOL(rdma_roce_rescan_device); static void callback_for_addr_gid_device_scan(struct ib_device *device, u8 port, diff --git a/drivers/infiniband/core/sa_query.c b/drivers/infiniband/core/sa_query.c index ab5e1024fea9..8cf15d4a8ac4 100644 --- a/drivers/infiniband/core/sa_query.c +++ b/drivers/infiniband/core/sa_query.c @@ -1227,9 +1227,9 @@ static u8 get_src_path_mask(struct ib_device *device, u8 port_num) return src_path_mask; } -int ib_init_ah_from_path(struct ib_device *device, u8 port_num, - struct sa_path_rec *rec, - struct rdma_ah_attr *ah_attr) +int ib_init_ah_attr_from_path(struct ib_device *device, u8 port_num, + struct sa_path_rec *rec, + struct rdma_ah_attr *ah_attr) { int ret; u16 gid_index; @@ -1341,10 +1341,11 @@ int ib_init_ah_from_path(struct ib_device *device, u8 port_num, return 0; } -EXPORT_SYMBOL(ib_init_ah_from_path); +EXPORT_SYMBOL(ib_init_ah_attr_from_path); static int alloc_mad(struct ib_sa_query *query, gfp_t gfp_mask) { + struct rdma_ah_attr ah_attr; unsigned long flags; spin_lock_irqsave(&query->port->ah_lock, flags); @@ -1356,6 +1357,15 @@ static int alloc_mad(struct ib_sa_query *query, gfp_t gfp_mask) query->sm_ah = query->port->sm_ah; spin_unlock_irqrestore(&query->port->ah_lock, flags); + /* + * Always check if sm_ah has valid dlid assigned, + * before querying for class port info + */ + if ((rdma_query_ah(query->sm_ah->ah, &ah_attr) < 0) || + !rdma_is_valid_unicast_lid(&ah_attr)) { + kref_put(&query->sm_ah->ref, free_sm_ah); + return -EAGAIN; + } query->mad_buf = ib_create_send_mad(query->port->agent, 1, query->sm_ah->pkey_index, 0, IB_MGMT_SA_HDR, IB_MGMT_SA_DATA, diff --git a/drivers/infiniband/core/security.c b/drivers/infiniband/core/security.c index 59b2f96d986a..b61dda6b04fc 100644 --- a/drivers/infiniband/core/security.c +++ b/drivers/infiniband/core/security.c @@ -653,12 +653,11 @@ int ib_security_modify_qp(struct ib_qp *qp, } return ret; } -EXPORT_SYMBOL(ib_security_modify_qp); -int ib_security_pkey_access(struct ib_device *dev, - u8 port_num, - u16 pkey_index, - void *sec) +static int ib_security_pkey_access(struct ib_device *dev, + u8 port_num, + u16 pkey_index, + void *sec) { u64 subnet_prefix; u16 pkey; @@ -678,7 +677,6 @@ int ib_security_pkey_access(struct ib_device *dev, return security_ib_pkey_access(sec, subnet_prefix, pkey); } -EXPORT_SYMBOL(ib_security_pkey_access); static int ib_mad_agent_security_change(struct notifier_block *nb, unsigned long event, diff --git a/drivers/infiniband/core/sysfs.c b/drivers/infiniband/core/sysfs.c index e30d86fa1855..8ae1308eecc7 100644 --- a/drivers/infiniband/core/sysfs.c +++ b/drivers/infiniband/core/sysfs.c @@ -1276,7 +1276,6 @@ int ib_device_register_sysfs(struct ib_device *device, int ret; int i; - WARN_ON_ONCE(!device->dev.parent); ret = dev_set_name(class_dev, "%s", device->name); if (ret) return ret; diff --git a/drivers/infiniband/core/umem.c b/drivers/infiniband/core/umem.c index 21e60b1e2ff4..9a4e899d94b3 100644 --- a/drivers/infiniband/core/umem.c +++ b/drivers/infiniband/core/umem.c @@ -191,7 +191,7 @@ struct ib_umem *ib_umem_get(struct ib_ucontext *context, unsigned long addr, sg_list_start = umem->sg_head.sgl; while (npages) { - ret = get_user_pages(cur_base, + ret = get_user_pages_longterm(cur_base, min_t(unsigned long, npages, PAGE_SIZE / sizeof (struct page *)), gup_flags, page_list, vma_list); @@ -352,7 +352,7 @@ int ib_umem_copy_from(void *dst, struct ib_umem *umem, size_t offset, return -EINVAL; } - ret = sg_pcopy_to_buffer(umem->sg_head.sgl, umem->nmap, dst, length, + ret = sg_pcopy_to_buffer(umem->sg_head.sgl, umem->npages, dst, length, offset + ib_umem_offset(umem)); if (ret < 0) diff --git a/drivers/infiniband/core/user_mad.c b/drivers/infiniband/core/user_mad.c index 4b64dd02e090..d4de187b1064 100644 --- a/drivers/infiniband/core/user_mad.c +++ b/drivers/infiniband/core/user_mad.c @@ -233,8 +233,7 @@ static void recv_handler(struct ib_mad_agent *agent, * On OPA devices it is okay to lose the upper 16 bits of LID as this * information is obtained elsewhere. Mask off the upper 16 bits. */ - if (agent->device->port_immutable[agent->port_num].core_cap_flags & - RDMA_CORE_PORT_INTEL_OPA) + if (rdma_cap_opa_mad(agent->device, agent->port_num)) packet->mad.hdr.lid = ib_lid_be16(0xFFFF & mad_recv_wc->wc->slid); else @@ -246,10 +245,14 @@ static void recv_handler(struct ib_mad_agent *agent, if (packet->mad.hdr.grh_present) { struct rdma_ah_attr ah_attr; const struct ib_global_route *grh; + int ret; - ib_init_ah_from_wc(agent->device, agent->port_num, - mad_recv_wc->wc, mad_recv_wc->recv_buf.grh, - &ah_attr); + ret = ib_init_ah_attr_from_wc(agent->device, agent->port_num, + mad_recv_wc->wc, + mad_recv_wc->recv_buf.grh, + &ah_attr); + if (ret) + goto err2; grh = rdma_ah_read_grh(&ah_attr); packet->mad.hdr.gid_index = grh->sgid_index; diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c index 840b24096690..c216d98bb816 100644 --- a/drivers/infiniband/core/uverbs_cmd.c +++ b/drivers/infiniband/core/uverbs_cmd.c @@ -1145,10 +1145,7 @@ int ib_uverbs_ex_create_cq(struct ib_uverbs_file *file, min(ucore->inlen, sizeof(cmd)), ib_uverbs_ex_create_cq_cb, NULL); - if (IS_ERR(obj)) - return PTR_ERR(obj); - - return 0; + return PTR_ERR_OR_ZERO(obj); } ssize_t ib_uverbs_resize_cq(struct ib_uverbs_file *file, diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c index e36d27ed4daa..a1bcdacb7f40 100644 --- a/drivers/infiniband/core/verbs.c +++ b/drivers/infiniband/core/verbs.c @@ -124,16 +124,24 @@ EXPORT_SYMBOL(ib_wc_status_msg); __attribute_const__ int ib_rate_to_mult(enum ib_rate rate) { switch (rate) { - case IB_RATE_2_5_GBPS: return 1; - case IB_RATE_5_GBPS: return 2; - case IB_RATE_10_GBPS: return 4; - case IB_RATE_20_GBPS: return 8; - case IB_RATE_30_GBPS: return 12; - case IB_RATE_40_GBPS: return 16; - case IB_RATE_60_GBPS: return 24; - case IB_RATE_80_GBPS: return 32; - case IB_RATE_120_GBPS: return 48; - default: return -1; + case IB_RATE_2_5_GBPS: return 1; + case IB_RATE_5_GBPS: return 2; + case IB_RATE_10_GBPS: return 4; + case IB_RATE_20_GBPS: return 8; + case IB_RATE_30_GBPS: return 12; + case IB_RATE_40_GBPS: return 16; + case IB_RATE_60_GBPS: return 24; + case IB_RATE_80_GBPS: return 32; + case IB_RATE_120_GBPS: return 48; + case IB_RATE_14_GBPS: return 6; + case IB_RATE_56_GBPS: return 22; + case IB_RATE_112_GBPS: return 45; + case IB_RATE_168_GBPS: return 67; + case IB_RATE_25_GBPS: return 10; + case IB_RATE_100_GBPS: return 40; + case IB_RATE_200_GBPS: return 80; + case IB_RATE_300_GBPS: return 120; + default: return -1; } } EXPORT_SYMBOL(ib_rate_to_mult); @@ -141,16 +149,24 @@ EXPORT_SYMBOL(ib_rate_to_mult); __attribute_const__ enum ib_rate mult_to_ib_rate(int mult) { switch (mult) { - case 1: return IB_RATE_2_5_GBPS; - case 2: return IB_RATE_5_GBPS; - case 4: return IB_RATE_10_GBPS; - case 8: return IB_RATE_20_GBPS; - case 12: return IB_RATE_30_GBPS; - case 16: return IB_RATE_40_GBPS; - case 24: return IB_RATE_60_GBPS; - case 32: return IB_RATE_80_GBPS; - case 48: return IB_RATE_120_GBPS; - default: return IB_RATE_PORT_CURRENT; + case 1: return IB_RATE_2_5_GBPS; + case 2: return IB_RATE_5_GBPS; + case 4: return IB_RATE_10_GBPS; + case 8: return IB_RATE_20_GBPS; + case 12: return IB_RATE_30_GBPS; + case 16: return IB_RATE_40_GBPS; + case 24: return IB_RATE_60_GBPS; + case 32: return IB_RATE_80_GBPS; + case 48: return IB_RATE_120_GBPS; + case 6: return IB_RATE_14_GBPS; + case 22: return IB_RATE_56_GBPS; + case 45: return IB_RATE_112_GBPS; + case 67: return IB_RATE_168_GBPS; + case 10: return IB_RATE_25_GBPS; + case 40: return IB_RATE_100_GBPS; + case 80: return IB_RATE_200_GBPS; + case 120: return IB_RATE_300_GBPS; + default: return IB_RATE_PORT_CURRENT; } } EXPORT_SYMBOL(mult_to_ib_rate); @@ -421,8 +437,7 @@ static bool find_gid_index(const union ib_gid *gid, const struct ib_gid_attr *gid_attr, void *context) { - struct find_gid_index_context *ctx = - (struct find_gid_index_context *)context; + struct find_gid_index_context *ctx = context; if (ctx->gid_type != gid_attr->gid_type) return false; @@ -481,8 +496,53 @@ int ib_get_gids_from_rdma_hdr(const union rdma_network_hdr *hdr, } EXPORT_SYMBOL(ib_get_gids_from_rdma_hdr); +/* Resolve destination mac address and hop limit for unicast destination + * GID entry, considering the source GID entry as well. + * ah_attribute must have have valid port_num, sgid_index. + */ +static int ib_resolve_unicast_gid_dmac(struct ib_device *device, + struct rdma_ah_attr *ah_attr) +{ + struct ib_gid_attr sgid_attr; + struct ib_global_route *grh; + int hop_limit = 0xff; + union ib_gid sgid; + int ret; + + grh = rdma_ah_retrieve_grh(ah_attr); + + ret = ib_query_gid(device, + rdma_ah_get_port_num(ah_attr), + grh->sgid_index, + &sgid, &sgid_attr); + if (ret || !sgid_attr.ndev) { + if (!ret) + ret = -ENXIO; + return ret; + } + + /* If destination is link local and source GID is RoCEv1, + * IP stack is not used. + */ + if (rdma_link_local_addr((struct in6_addr *)grh->dgid.raw) && + sgid_attr.gid_type == IB_GID_TYPE_ROCE) { + rdma_get_ll_mac((struct in6_addr *)grh->dgid.raw, + ah_attr->roce.dmac); + goto done; + } + + ret = rdma_addr_find_l2_eth_by_grh(&sgid, &grh->dgid, + ah_attr->roce.dmac, + sgid_attr.ndev, &hop_limit); +done: + dev_put(sgid_attr.ndev); + + grh->hop_limit = hop_limit; + return ret; +} + /* - * This function creates ah from the incoming packet. + * This function initializes address handle attributes from the incoming packet. * Incoming packet has dgid of the receiver node on which this code is * getting executed and, sgid contains the GID of the sender. * @@ -490,13 +550,10 @@ EXPORT_SYMBOL(ib_get_gids_from_rdma_hdr); * as sgid and, sgid is used as dgid because sgid contains destinations * GID whom to respond to. * - * This is why when calling rdma_addr_find_l2_eth_by_grh() function, the - * position of arguments dgid and sgid do not match the order of the - * parameters. */ -int ib_init_ah_from_wc(struct ib_device *device, u8 port_num, - const struct ib_wc *wc, const struct ib_grh *grh, - struct rdma_ah_attr *ah_attr) +int ib_init_ah_attr_from_wc(struct ib_device *device, u8 port_num, + const struct ib_wc *wc, const struct ib_grh *grh, + struct rdma_ah_attr *ah_attr) { u32 flow_class; u16 gid_index; @@ -523,57 +580,33 @@ int ib_init_ah_from_wc(struct ib_device *device, u8 port_num, if (ret) return ret; + rdma_ah_set_sl(ah_attr, wc->sl); + rdma_ah_set_port_num(ah_attr, port_num); + if (rdma_protocol_roce(device, port_num)) { - int if_index = 0; u16 vlan_id = wc->wc_flags & IB_WC_WITH_VLAN ? wc->vlan_id : 0xffff; - struct net_device *idev; - struct net_device *resolved_dev; if (!(wc->wc_flags & IB_WC_GRH)) return -EPROTOTYPE; - if (!device->get_netdev) - return -EOPNOTSUPP; - - idev = device->get_netdev(device, port_num); - if (!idev) - return -ENODEV; - - ret = rdma_addr_find_l2_eth_by_grh(&dgid, &sgid, - ah_attr->roce.dmac, - wc->wc_flags & IB_WC_WITH_VLAN ? - NULL : &vlan_id, - &if_index, &hoplimit); - if (ret) { - dev_put(idev); - return ret; - } - - resolved_dev = dev_get_by_index(&init_net, if_index); - rcu_read_lock(); - if (resolved_dev != idev && !rdma_is_upper_dev_rcu(idev, - resolved_dev)) - ret = -EHOSTUNREACH; - rcu_read_unlock(); - dev_put(idev); - dev_put(resolved_dev); - if (ret) - return ret; - - ret = get_sgid_index_from_eth(device, port_num, vlan_id, - &dgid, gid_type, &gid_index); + ret = get_sgid_index_from_eth(device, port_num, + vlan_id, &dgid, + gid_type, &gid_index); if (ret) return ret; - } - rdma_ah_set_dlid(ah_attr, wc->slid); - rdma_ah_set_sl(ah_attr, wc->sl); - rdma_ah_set_path_bits(ah_attr, wc->dlid_path_bits); - rdma_ah_set_port_num(ah_attr, port_num); + flow_class = be32_to_cpu(grh->version_tclass_flow); + rdma_ah_set_grh(ah_attr, &sgid, + flow_class & 0xFFFFF, + (u8)gid_index, hoplimit, + (flow_class >> 20) & 0xFF); + return ib_resolve_unicast_gid_dmac(device, ah_attr); + } else { + rdma_ah_set_dlid(ah_attr, wc->slid); + rdma_ah_set_path_bits(ah_attr, wc->dlid_path_bits); - if (wc->wc_flags & IB_WC_GRH) { - if (!rdma_cap_eth_ah(device, port_num)) { + if (wc->wc_flags & IB_WC_GRH) { if (dgid.global.interface_id != cpu_to_be64(IB_SA_WELL_KNOWN_GUID)) { ret = ib_find_cached_gid_by_port(device, &dgid, IB_GID_TYPE_IB, @@ -584,18 +617,17 @@ int ib_init_ah_from_wc(struct ib_device *device, u8 port_num, } else { gid_index = 0; } - } - - flow_class = be32_to_cpu(grh->version_tclass_flow); - rdma_ah_set_grh(ah_attr, &sgid, - flow_class & 0xFFFFF, - (u8)gid_index, hoplimit, - (flow_class >> 20) & 0xFF); + flow_class = be32_to_cpu(grh->version_tclass_flow); + rdma_ah_set_grh(ah_attr, &sgid, + flow_class & 0xFFFFF, + (u8)gid_index, hoplimit, + (flow_class >> 20) & 0xFF); + } + return 0; } - return 0; } -EXPORT_SYMBOL(ib_init_ah_from_wc); +EXPORT_SYMBOL(ib_init_ah_attr_from_wc); struct ib_ah *ib_create_ah_from_wc(struct ib_pd *pd, const struct ib_wc *wc, const struct ib_grh *grh, u8 port_num) @@ -603,7 +635,7 @@ struct ib_ah *ib_create_ah_from_wc(struct ib_pd *pd, const struct ib_wc *wc, struct rdma_ah_attr ah_attr; int ret; - ret = ib_init_ah_from_wc(pd->device, port_num, wc, grh, &ah_attr); + ret = ib_init_ah_attr_from_wc(pd->device, port_num, wc, grh, &ah_attr); if (ret) return ERR_PTR(ret); @@ -1274,11 +1306,6 @@ static int ib_resolve_eth_dmac(struct ib_device *device, grh = rdma_ah_retrieve_grh(ah_attr); - if (rdma_link_local_addr((struct in6_addr *)grh->dgid.raw)) { - rdma_get_ll_mac((struct in6_addr *)grh->dgid.raw, - ah_attr->roce.dmac); - return 0; - } if (rdma_is_multicast_addr((struct in6_addr *)ah_attr->grh.dgid.raw)) { if (ipv6_addr_v4mapped((struct in6_addr *)ah_attr->grh.dgid.raw)) { __be32 addr = 0; @@ -1290,34 +1317,8 @@ static int ib_resolve_eth_dmac(struct ib_device *device, (char *)ah_attr->roce.dmac); } } else { - union ib_gid sgid; - struct ib_gid_attr sgid_attr; - int ifindex; - int hop_limit; - - ret = ib_query_gid(device, - rdma_ah_get_port_num(ah_attr), - grh->sgid_index, - &sgid, &sgid_attr); - - if (ret || !sgid_attr.ndev) { - if (!ret) - ret = -ENXIO; - goto out; - } - - ifindex = sgid_attr.ndev->ifindex; - - ret = - rdma_addr_find_l2_eth_by_grh(&sgid, &grh->dgid, - ah_attr->roce.dmac, - NULL, &ifindex, &hop_limit); - - dev_put(sgid_attr.ndev); - - grh->hop_limit = hop_limit; + ret = ib_resolve_unicast_gid_dmac(device, ah_attr); } -out: return ret; } @@ -1335,6 +1336,7 @@ out: int ib_modify_qp_with_udata(struct ib_qp *qp, struct ib_qp_attr *attr, int attr_mask, struct ib_udata *udata) { + u8 port = attr_mask & IB_QP_PORT ? attr->port_num : qp->port; int ret; if (attr_mask & IB_QP_AV) { @@ -1342,6 +1344,21 @@ int ib_modify_qp_with_udata(struct ib_qp *qp, struct ib_qp_attr *attr, if (ret) return ret; } + + if (rdma_ib_or_roce(qp->device, port)) { + if (attr_mask & IB_QP_RQ_PSN && attr->rq_psn & ~0xffffff) { + pr_warn("%s: %s rq_psn overflow, masking to 24 bits\n", + __func__, qp->device->name); + attr->rq_psn &= 0xffffff; + } + + if (attr_mask & IB_QP_SQ_PSN && attr->sq_psn & ~0xffffff) { + pr_warn("%s: %s sq_psn overflow, masking to 24 bits\n", + __func__, qp->device->name); + attr->sq_psn &= 0xffffff; + } + } + ret = ib_security_modify_qp(qp, attr, attr_mask, udata); if (!ret && (attr_mask & IB_QP_PORT)) qp->port = attr->port_num; diff --git a/drivers/infiniband/hw/bnxt_re/main.c b/drivers/infiniband/hw/bnxt_re/main.c index aafc19aa5de1..bf268bf1f496 100644 --- a/drivers/infiniband/hw/bnxt_re/main.c +++ b/drivers/infiniband/hw/bnxt_re/main.c @@ -417,7 +417,7 @@ static struct bnxt_en_dev *bnxt_re_dev_probe(struct net_device *netdev) return ERR_PTR(-EINVAL); if (!(en_dev->flags & BNXT_EN_FLAG_ROCE_CAP)) { - dev_dbg(&pdev->dev, + dev_info(&pdev->dev, "%s: probe error: RoCE is not supported on this device", ROCE_DRV_MODULE_NAME); return ERR_PTR(-ENODEV); diff --git a/drivers/infiniband/hw/bnxt_re/qplib_fp.c b/drivers/infiniband/hw/bnxt_re/qplib_fp.c index 61764f7aa79b..eb7195c20b88 100644 --- a/drivers/infiniband/hw/bnxt_re/qplib_fp.c +++ b/drivers/infiniband/hw/bnxt_re/qplib_fp.c @@ -410,7 +410,6 @@ int bnxt_qplib_enable_nq(struct pci_dev *pdev, struct bnxt_qplib_nq *nq, if (rc) { dev_err(&nq->pdev->dev, "Failed to request IRQ for NQ: %#x", rc); - bnxt_qplib_disable_nq(nq); goto fail; } diff --git a/drivers/infiniband/hw/bnxt_re/qplib_res.c b/drivers/infiniband/hw/bnxt_re/qplib_res.c index 4e101704e801..ad37d54affcc 100644 --- a/drivers/infiniband/hw/bnxt_re/qplib_res.c +++ b/drivers/infiniband/hw/bnxt_re/qplib_res.c @@ -104,13 +104,12 @@ static int __alloc_pbl(struct pci_dev *pdev, struct bnxt_qplib_pbl *pbl, if (!sghead) { for (i = 0; i < pages; i++) { - pbl->pg_arr[i] = dma_alloc_coherent(&pdev->dev, - pbl->pg_size, - &pbl->pg_map_arr[i], - GFP_KERNEL); + pbl->pg_arr[i] = dma_zalloc_coherent(&pdev->dev, + pbl->pg_size, + &pbl->pg_map_arr[i], + GFP_KERNEL); if (!pbl->pg_arr[i]) goto fail; - memset(pbl->pg_arr[i], 0, pbl->pg_size); pbl->pg_count++; } } else { diff --git a/drivers/infiniband/hw/cxgb4/cm.c b/drivers/infiniband/hw/cxgb4/cm.c index 21db3b48a617..4cf17c650c36 100644 --- a/drivers/infiniband/hw/cxgb4/cm.c +++ b/drivers/infiniband/hw/cxgb4/cm.c @@ -257,8 +257,8 @@ static void set_emss(struct c4iw_ep *ep, u16 opt) if (ep->emss < 128) ep->emss = 128; if (ep->emss & 7) - pr_warn("Warning: misaligned mtu idx %u mss %u emss=%u\n", - TCPOPT_MSS_G(opt), ep->mss, ep->emss); + pr_debug("Warning: misaligned mtu idx %u mss %u emss=%u\n", + TCPOPT_MSS_G(opt), ep->mss, ep->emss); pr_debug("mss_idx %u mss %u emss=%u\n", TCPOPT_MSS_G(opt), ep->mss, ep->emss); } @@ -2733,9 +2733,8 @@ static int peer_abort(struct c4iw_dev *dev, struct sk_buff *skb) return 0; if (cxgb_is_neg_adv(req->status)) { - pr_warn("%s Negative advice on abort- tid %u status %d (%s)\n", - __func__, ep->hwtid, req->status, - neg_adv_str(req->status)); + pr_debug("Negative advice on abort- tid %u status %d (%s)\n", + ep->hwtid, req->status, neg_adv_str(req->status)); ep->stats.abort_neg_adv++; mutex_lock(&dev->rdev.stats.lock); dev->rdev.stats.neg_adv++; @@ -3567,8 +3566,8 @@ int c4iw_ep_disconnect(struct c4iw_ep *ep, int abrupt, gfp_t gfp) case MORIBUND: case ABORTING: case DEAD: - pr_info("%s ignoring disconnect ep %p state %u\n", - __func__, ep, ep->com.state); + pr_debug("ignoring disconnect ep %p state %u\n", + ep, ep->com.state); break; default: WARN_ONCE(1, "Bad endpoint state %u\n", ep->com.state); @@ -4097,9 +4096,15 @@ static void process_work(struct work_struct *work) dev = *((struct c4iw_dev **) (skb->cb + sizeof(void *))); opcode = rpl->ot.opcode; - ret = work_handlers[opcode](dev, skb); - if (!ret) + if (opcode >= ARRAY_SIZE(work_handlers) || + !work_handlers[opcode]) { + pr_err("No handler for opcode 0x%x.\n", opcode); kfree_skb(skb); + } else { + ret = work_handlers[opcode](dev, skb); + if (!ret) + kfree_skb(skb); + } process_timedout_eps(); } } @@ -4201,8 +4206,8 @@ static int peer_abort_intr(struct c4iw_dev *dev, struct sk_buff *skb) return 0; } if (cxgb_is_neg_adv(req->status)) { - pr_warn("%s Negative advice on abort- tid %u status %d (%s)\n", - __func__, ep->hwtid, req->status, + pr_debug("Negative advice on abort- tid %u status %d (%s)\n", + ep->hwtid, req->status, neg_adv_str(req->status)); goto out; } diff --git a/drivers/infiniband/hw/cxgb4/device.c b/drivers/infiniband/hw/cxgb4/device.c index af77d128d242..7a9d0de89d6a 100644 --- a/drivers/infiniband/hw/cxgb4/device.c +++ b/drivers/infiniband/hw/cxgb4/device.c @@ -66,7 +66,7 @@ MODULE_PARM_DESC(c4iw_wr_log_size_order, static LIST_HEAD(uld_ctx_list); static DEFINE_MUTEX(dev_mutex); -struct workqueue_struct *reg_workq; +static struct workqueue_struct *reg_workq; #define DB_FC_RESUME_SIZE 64 #define DB_FC_RESUME_DELAY 1 @@ -108,19 +108,19 @@ void c4iw_log_wr_stats(struct t4_wq *wq, struct t4_cqe *cqe) idx = (atomic_inc_return(&wq->rdev->wr_log_idx) - 1) & (wq->rdev->wr_log_size - 1); le.poll_sge_ts = cxgb4_read_sge_timestamp(wq->rdev->lldi.ports[0]); - getnstimeofday(&le.poll_host_ts); + le.poll_host_time = ktime_get(); le.valid = 1; le.cqe_sge_ts = CQE_TS(cqe); if (SQ_TYPE(cqe)) { le.qid = wq->sq.qid; le.opcode = CQE_OPCODE(cqe); - le.post_host_ts = wq->sq.sw_sq[wq->sq.cidx].host_ts; + le.post_host_time = wq->sq.sw_sq[wq->sq.cidx].host_time; le.post_sge_ts = wq->sq.sw_sq[wq->sq.cidx].sge_ts; le.wr_id = CQE_WRID_SQ_IDX(cqe); } else { le.qid = wq->rq.qid; le.opcode = FW_RI_RECEIVE; - le.post_host_ts = wq->rq.sw_rq[wq->rq.cidx].host_ts; + le.post_host_time = wq->rq.sw_rq[wq->rq.cidx].host_time; le.post_sge_ts = wq->rq.sw_rq[wq->rq.cidx].sge_ts; le.wr_id = CQE_WRID_MSN(cqe); } @@ -130,9 +130,9 @@ void c4iw_log_wr_stats(struct t4_wq *wq, struct t4_cqe *cqe) static int wr_log_show(struct seq_file *seq, void *v) { struct c4iw_dev *dev = seq->private; - struct timespec prev_ts = {0, 0}; + ktime_t prev_time; struct wr_log_entry *lep; - int prev_ts_set = 0; + int prev_time_set = 0; int idx, end; #define ts2ns(ts) div64_u64((ts) * dev->rdev.lldi.cclk_ps, 1000) @@ -145,33 +145,29 @@ static int wr_log_show(struct seq_file *seq, void *v) lep = &dev->rdev.wr_log[idx]; while (idx != end) { if (lep->valid) { - if (!prev_ts_set) { - prev_ts_set = 1; - prev_ts = lep->poll_host_ts; + if (!prev_time_set) { + prev_time_set = 1; + prev_time = lep->poll_host_time; } - seq_printf(seq, "%04u: sec %lu nsec %lu qid %u opcode " - "%u %s 0x%x host_wr_delta sec %lu nsec %lu " + seq_printf(seq, "%04u: nsec %llu qid %u opcode " + "%u %s 0x%x host_wr_delta nsec %llu " "post_sge_ts 0x%llx cqe_sge_ts 0x%llx " "poll_sge_ts 0x%llx post_poll_delta_ns %llu " "cqe_poll_delta_ns %llu\n", idx, - timespec_sub(lep->poll_host_ts, - prev_ts).tv_sec, - timespec_sub(lep->poll_host_ts, - prev_ts).tv_nsec, + ktime_to_ns(ktime_sub(lep->poll_host_time, + prev_time)), lep->qid, lep->opcode, lep->opcode == FW_RI_RECEIVE ? "msn" : "wrid", lep->wr_id, - timespec_sub(lep->poll_host_ts, - lep->post_host_ts).tv_sec, - timespec_sub(lep->poll_host_ts, - lep->post_host_ts).tv_nsec, + ktime_to_ns(ktime_sub(lep->poll_host_time, + lep->post_host_time)), lep->post_sge_ts, lep->cqe_sge_ts, lep->poll_sge_ts, ts2ns(lep->poll_sge_ts - lep->post_sge_ts), ts2ns(lep->poll_sge_ts - lep->cqe_sge_ts)); - prev_ts = lep->poll_host_ts; + prev_time = lep->poll_host_time; } idx++; if (idx > (dev->rdev.wr_log_size - 1)) diff --git a/drivers/infiniband/hw/cxgb4/ev.c b/drivers/infiniband/hw/cxgb4/ev.c index a252d5c40ae3..3e9d8b277ab9 100644 --- a/drivers/infiniband/hw/cxgb4/ev.c +++ b/drivers/infiniband/hw/cxgb4/ev.c @@ -236,7 +236,7 @@ int c4iw_ev_handler(struct c4iw_dev *dev, u32 qid) if (atomic_dec_and_test(&chp->refcnt)) wake_up(&chp->wait); } else { - pr_warn("%s unknown cqid 0x%x\n", __func__, qid); + pr_debug("unknown cqid 0x%x\n", qid); spin_unlock_irqrestore(&dev->lock, flag); } return 0; diff --git a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h index 65dd3726ca02..cc929002c05e 100644 --- a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h +++ b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h @@ -153,8 +153,8 @@ struct c4iw_hw_queue { }; struct wr_log_entry { - struct timespec post_host_ts; - struct timespec poll_host_ts; + ktime_t post_host_time; + ktime_t poll_host_time; u64 post_sge_ts; u64 cqe_sge_ts; u64 poll_sge_ts; diff --git a/drivers/infiniband/hw/cxgb4/qp.c b/drivers/infiniband/hw/cxgb4/qp.c index d5c92fc520d6..de77b6027d69 100644 --- a/drivers/infiniband/hw/cxgb4/qp.c +++ b/drivers/infiniband/hw/cxgb4/qp.c @@ -1042,7 +1042,7 @@ int c4iw_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, if (c4iw_wr_log) { swsqe->sge_ts = cxgb4_read_sge_timestamp( qhp->rhp->rdev.lldi.ports[0]); - getnstimeofday(&swsqe->host_ts); + swsqe->host_time = ktime_get(); } init_wr_hdr(wqe, qhp->wq.sq.pidx, fw_opcode, fw_flags, len16); @@ -1117,8 +1117,8 @@ int c4iw_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *wr, qhp->wq.rq.sw_rq[qhp->wq.rq.pidx].sge_ts = cxgb4_read_sge_timestamp( qhp->rhp->rdev.lldi.ports[0]); - getnstimeofday( - &qhp->wq.rq.sw_rq[qhp->wq.rq.pidx].host_ts); + qhp->wq.rq.sw_rq[qhp->wq.rq.pidx].host_time = + ktime_get(); } wqe->recv.opcode = FW_RI_RECV_WR; diff --git a/drivers/infiniband/hw/cxgb4/t4.h b/drivers/infiniband/hw/cxgb4/t4.h index 79e8ee12c391..8369c7c8de83 100644 --- a/drivers/infiniband/hw/cxgb4/t4.h +++ b/drivers/infiniband/hw/cxgb4/t4.h @@ -277,7 +277,7 @@ struct t4_swsqe { int signaled; u16 idx; int flushed; - struct timespec host_ts; + ktime_t host_time; u64 sge_ts; }; @@ -318,7 +318,7 @@ struct t4_sq { struct t4_swrqe { u64 wr_id; - struct timespec host_ts; + ktime_t host_time; u64 sge_ts; }; diff --git a/drivers/infiniband/hw/hfi1/chip.c b/drivers/infiniband/hw/hfi1/chip.c index 4f057e8ffe50..6660f920f42e 100644 --- a/drivers/infiniband/hw/hfi1/chip.c +++ b/drivers/infiniband/hw/hfi1/chip.c @@ -6518,11 +6518,12 @@ static void _dc_start(struct hfi1_devdata *dd) if (!dd->dc_shutdown) return; - /* - * Take the 8051 out of reset, wait until 8051 is ready, and set host - * version bit. - */ - release_and_wait_ready_8051_firmware(dd); + /* Take the 8051 out of reset */ + write_csr(dd, DC_DC8051_CFG_RST, 0ull); + /* Wait until 8051 is ready */ + if (wait_fm_ready(dd, TIMEOUT_8051_START)) + dd_dev_err(dd, "%s: timeout starting 8051 firmware\n", + __func__); /* Take away reset for LCB and RX FPE (set in lcb_shutdown). */ write_csr(dd, DCC_CFG_RESET, 0x10); @@ -8564,23 +8565,27 @@ int write_lcb_csr(struct hfi1_devdata *dd, u32 addr, u64 data) } /* - * If the 8051 is in reset mode (dd->dc_shutdown == 1), this function - * will still continue executing. - * * Returns: * < 0 = Linux error, not able to get access * > 0 = 8051 command RETURN_CODE */ -static int _do_8051_command(struct hfi1_devdata *dd, u32 type, u64 in_data, - u64 *out_data) +static int do_8051_command(struct hfi1_devdata *dd, u32 type, u64 in_data, + u64 *out_data) { u64 reg, completed; int return_code; unsigned long timeout; - lockdep_assert_held(&dd->dc8051_lock); hfi1_cdbg(DC8051, "type %d, data 0x%012llx", type, in_data); + mutex_lock(&dd->dc8051_lock); + + /* We can't send any commands to the 8051 if it's in reset */ + if (dd->dc_shutdown) { + return_code = -ENODEV; + goto fail; + } + /* * If an 8051 host command timed out previously, then the 8051 is * stuck. @@ -8681,29 +8686,6 @@ static int _do_8051_command(struct hfi1_devdata *dd, u32 type, u64 in_data, write_csr(dd, DC_DC8051_CFG_HOST_CMD_0, 0); fail: - return return_code; -} - -/* - * Returns: - * < 0 = Linux error, not able to get access - * > 0 = 8051 command RETURN_CODE - */ -static int do_8051_command(struct hfi1_devdata *dd, u32 type, u64 in_data, - u64 *out_data) -{ - int return_code; - - mutex_lock(&dd->dc8051_lock); - /* We can't send any commands to the 8051 if it's in reset */ - if (dd->dc_shutdown) { - return_code = -ENODEV; - goto fail; - } - - return_code = _do_8051_command(dd, type, in_data, out_data); - -fail: mutex_unlock(&dd->dc8051_lock); return return_code; } @@ -8713,17 +8695,16 @@ static int set_physical_link_state(struct hfi1_devdata *dd, u64 state) return do_8051_command(dd, HCMD_CHANGE_PHY_STATE, state, NULL); } -static int _load_8051_config(struct hfi1_devdata *dd, u8 field_id, - u8 lane_id, u32 config_data) +int load_8051_config(struct hfi1_devdata *dd, u8 field_id, + u8 lane_id, u32 config_data) { u64 data; int ret; - lockdep_assert_held(&dd->dc8051_lock); data = (u64)field_id << LOAD_DATA_FIELD_ID_SHIFT | (u64)lane_id << LOAD_DATA_LANE_ID_SHIFT | (u64)config_data << LOAD_DATA_DATA_SHIFT; - ret = _do_8051_command(dd, HCMD_LOAD_CONFIG_DATA, data, NULL); + ret = do_8051_command(dd, HCMD_LOAD_CONFIG_DATA, data, NULL); if (ret != HCMD_SUCCESS) { dd_dev_err(dd, "load 8051 config: field id %d, lane %d, err %d\n", @@ -8732,18 +8713,6 @@ static int _load_8051_config(struct hfi1_devdata *dd, u8 field_id, return ret; } -int load_8051_config(struct hfi1_devdata *dd, u8 field_id, - u8 lane_id, u32 config_data) -{ - int return_code; - - mutex_lock(&dd->dc8051_lock); - return_code = _load_8051_config(dd, field_id, lane_id, config_data); - mutex_unlock(&dd->dc8051_lock); - - return return_code; -} - /* * Read the 8051 firmware "registers". Use the RAM directly. Always * set the result, even on error. @@ -8859,14 +8828,13 @@ int write_host_interface_version(struct hfi1_devdata *dd, u8 version) u32 frame; u32 mask; - lockdep_assert_held(&dd->dc8051_lock); mask = (HOST_INTERFACE_VERSION_MASK << HOST_INTERFACE_VERSION_SHIFT); read_8051_config(dd, RESERVED_REGISTERS, GENERAL_CONFIG, &frame); /* Clear, then set field */ frame &= ~mask; frame |= ((u32)version << HOST_INTERFACE_VERSION_SHIFT); - return _load_8051_config(dd, RESERVED_REGISTERS, GENERAL_CONFIG, - frame); + return load_8051_config(dd, RESERVED_REGISTERS, GENERAL_CONFIG, + frame); } void read_misc_status(struct hfi1_devdata *dd, u8 *ver_major, u8 *ver_minor, @@ -9270,6 +9238,14 @@ static int set_local_link_attributes(struct hfi1_pportdata *ppd) if (ret != HCMD_SUCCESS) goto set_local_link_attributes_fail; + ret = write_host_interface_version(dd, HOST_INTERFACE_VERSION); + if (ret != HCMD_SUCCESS) { + dd_dev_err(dd, + "Failed to set host interface version, return 0x%x\n", + ret); + goto set_local_link_attributes_fail; + } + /* * DC supports continuous updates. */ @@ -14944,9 +14920,8 @@ struct hfi1_devdata *hfi1_init_dd(struct pci_dev *pdev, if (num_vls < HFI1_MIN_VLS_SUPPORTED || num_vls > HFI1_MAX_VLS_SUPPORTED) { - hfi1_early_err(&pdev->dev, - "Invalid num_vls %u, using %u VLs\n", - num_vls, HFI1_MAX_VLS_SUPPORTED); + dd_dev_err(dd, "Invalid num_vls %u, using %u VLs\n", + num_vls, HFI1_MAX_VLS_SUPPORTED); num_vls = HFI1_MAX_VLS_SUPPORTED; } ppd->vls_supported = num_vls; diff --git a/drivers/infiniband/hw/hfi1/chip.h b/drivers/infiniband/hw/hfi1/chip.h index 133e313feca4..21fca8ec5076 100644 --- a/drivers/infiniband/hw/hfi1/chip.h +++ b/drivers/infiniband/hw/hfi1/chip.h @@ -508,6 +508,7 @@ #define DOWN_REMOTE_REASON_SHIFT 16 #define DOWN_REMOTE_REASON_MASK 0xff +#define HOST_INTERFACE_VERSION 1 #define HOST_INTERFACE_VERSION_SHIFT 16 #define HOST_INTERFACE_VERSION_MASK 0xff @@ -713,7 +714,6 @@ void read_misc_status(struct hfi1_devdata *dd, u8 *ver_major, u8 *ver_minor, u8 *ver_patch); int write_host_interface_version(struct hfi1_devdata *dd, u8 version); void read_guid(struct hfi1_devdata *dd); -int release_and_wait_ready_8051_firmware(struct hfi1_devdata *dd); int wait_fm_ready(struct hfi1_devdata *dd, u32 mstimeout); void set_link_down_reason(struct hfi1_pportdata *ppd, u8 lcl_reason, u8 neigh_reason, u8 rem_reason); diff --git a/drivers/infiniband/hw/hfi1/driver.c b/drivers/infiniband/hw/hfi1/driver.c index 4f65ac671044..067b29f35f21 100644 --- a/drivers/infiniband/hw/hfi1/driver.c +++ b/drivers/infiniband/hw/hfi1/driver.c @@ -159,22 +159,6 @@ static int hfi1_caps_get(char *buffer, const struct kernel_param *kp) return scnprintf(buffer, PAGE_SIZE, "0x%lx", cap_mask); } -const char *get_unit_name(int unit) -{ - static char iname[16]; - - snprintf(iname, sizeof(iname), DRIVER_NAME "_%u", unit); - return iname; -} - -const char *get_card_name(struct rvt_dev_info *rdi) -{ - struct hfi1_ibdev *ibdev = container_of(rdi, struct hfi1_ibdev, rdi); - struct hfi1_devdata *dd = container_of(ibdev, - struct hfi1_devdata, verbs_dev); - return get_unit_name(dd->unit); -} - struct pci_dev *get_pci_dev(struct rvt_dev_info *rdi) { struct hfi1_ibdev *ibdev = container_of(rdi, struct hfi1_ibdev, rdi); diff --git a/drivers/infiniband/hw/hfi1/firmware.c b/drivers/infiniband/hw/hfi1/firmware.c index 98868df78a7e..2b57ba70ddd6 100644 --- a/drivers/infiniband/hw/hfi1/firmware.c +++ b/drivers/infiniband/hw/hfi1/firmware.c @@ -68,7 +68,6 @@ #define ALT_FW_FABRIC_NAME "hfi1_fabric_d.fw" #define ALT_FW_SBUS_NAME "hfi1_sbus_d.fw" #define ALT_FW_PCIE_NAME "hfi1_pcie_d.fw" -#define HOST_INTERFACE_VERSION 1 MODULE_FIRMWARE(DEFAULT_FW_8051_NAME_ASIC); MODULE_FIRMWARE(DEFAULT_FW_FABRIC_NAME); @@ -976,46 +975,6 @@ int wait_fm_ready(struct hfi1_devdata *dd, u32 mstimeout) } /* - * Clear all reset bits, releasing the 8051. - * Wait for firmware to be ready to accept host requests. - * Then, set host version bit. - * - * This function executes even if the 8051 is in reset mode when - * dd->dc_shutdown == 1. - * - * Expects dd->dc8051_lock to be held. - */ -int release_and_wait_ready_8051_firmware(struct hfi1_devdata *dd) -{ - int ret; - - lockdep_assert_held(&dd->dc8051_lock); - /* clear all reset bits, releasing the 8051 */ - write_csr(dd, DC_DC8051_CFG_RST, 0ull); - - /* - * Wait for firmware to be ready to accept host - * requests. - */ - ret = wait_fm_ready(dd, TIMEOUT_8051_START); - if (ret) { - dd_dev_err(dd, "8051 start timeout, current FW state 0x%x\n", - get_firmware_state(dd)); - return ret; - } - - ret = write_host_interface_version(dd, HOST_INTERFACE_VERSION); - if (ret != HCMD_SUCCESS) { - dd_dev_err(dd, - "Failed to set host interface version, return 0x%x\n", - ret); - return -EIO; - } - - return 0; -} - -/* * Load the 8051 firmware. */ static int load_8051_firmware(struct hfi1_devdata *dd, @@ -1080,22 +1039,31 @@ static int load_8051_firmware(struct hfi1_devdata *dd, if (ret) return ret; + /* clear all reset bits, releasing the 8051 */ + write_csr(dd, DC_DC8051_CFG_RST, 0ull); + /* - * Clear all reset bits, releasing the 8051. * DC reset step 5. Wait for firmware to be ready to accept host * requests. - * Then, set host version bit. */ - mutex_lock(&dd->dc8051_lock); - ret = release_and_wait_ready_8051_firmware(dd); - mutex_unlock(&dd->dc8051_lock); - if (ret) - return ret; + ret = wait_fm_ready(dd, TIMEOUT_8051_START); + if (ret) { /* timed out */ + dd_dev_err(dd, "8051 start timeout, current state 0x%x\n", + get_firmware_state(dd)); + return -ETIMEDOUT; + } read_misc_status(dd, &ver_major, &ver_minor, &ver_patch); dd_dev_info(dd, "8051 firmware version %d.%d.%d\n", (int)ver_major, (int)ver_minor, (int)ver_patch); dd->dc8051_ver = dc8051_ver(ver_major, ver_minor, ver_patch); + ret = write_host_interface_version(dd, HOST_INTERFACE_VERSION); + if (ret != HCMD_SUCCESS) { + dd_dev_err(dd, + "Failed to set host interface version, return 0x%x\n", + ret); + return -EIO; + } return 0; } diff --git a/drivers/infiniband/hw/hfi1/hfi.h b/drivers/infiniband/hw/hfi1/hfi.h index 8ce9118d4a7f..b42c22292597 100644 --- a/drivers/infiniband/hw/hfi1/hfi.h +++ b/drivers/infiniband/hw/hfi1/hfi.h @@ -1623,7 +1623,7 @@ static int ingress_pkey_table_search(struct hfi1_pportdata *ppd, u16 pkey) * the 'error info' for this failure. */ static void ingress_pkey_table_fail(struct hfi1_pportdata *ppd, u16 pkey, - u16 slid) + u32 slid) { struct hfi1_devdata *dd = ppd->dd; @@ -1971,8 +1971,6 @@ int get_platform_config_field(struct hfi1_devdata *dd, table_type, int table_index, int field_index, u32 *data, u32 len); -const char *get_unit_name(int unit); -const char *get_card_name(struct rvt_dev_info *rdi); struct pci_dev *get_pci_dev(struct rvt_dev_info *rdi); /* @@ -2122,39 +2120,42 @@ static inline u64 hfi1_pkt_base_sdma_integrity(struct hfi1_devdata *dd) #define dd_dev_emerg(dd, fmt, ...) \ dev_emerg(&(dd)->pcidev->dev, "%s: " fmt, \ - get_unit_name((dd)->unit), ##__VA_ARGS__) + rvt_get_ibdev_name(&(dd)->verbs_dev.rdi), ##__VA_ARGS__) #define dd_dev_err(dd, fmt, ...) \ dev_err(&(dd)->pcidev->dev, "%s: " fmt, \ - get_unit_name((dd)->unit), ##__VA_ARGS__) + rvt_get_ibdev_name(&(dd)->verbs_dev.rdi), ##__VA_ARGS__) #define dd_dev_err_ratelimited(dd, fmt, ...) \ dev_err_ratelimited(&(dd)->pcidev->dev, "%s: " fmt, \ - get_unit_name((dd)->unit), ##__VA_ARGS__) + rvt_get_ibdev_name(&(dd)->verbs_dev.rdi), \ + ##__VA_ARGS__) #define dd_dev_warn(dd, fmt, ...) \ dev_warn(&(dd)->pcidev->dev, "%s: " fmt, \ - get_unit_name((dd)->unit), ##__VA_ARGS__) + rvt_get_ibdev_name(&(dd)->verbs_dev.rdi), ##__VA_ARGS__) #define dd_dev_warn_ratelimited(dd, fmt, ...) \ dev_warn_ratelimited(&(dd)->pcidev->dev, "%s: " fmt, \ - get_unit_name((dd)->unit), ##__VA_ARGS__) + rvt_get_ibdev_name(&(dd)->verbs_dev.rdi), \ + ##__VA_ARGS__) #define dd_dev_info(dd, fmt, ...) \ dev_info(&(dd)->pcidev->dev, "%s: " fmt, \ - get_unit_name((dd)->unit), ##__VA_ARGS__) + rvt_get_ibdev_name(&(dd)->verbs_dev.rdi), ##__VA_ARGS__) #define dd_dev_info_ratelimited(dd, fmt, ...) \ dev_info_ratelimited(&(dd)->pcidev->dev, "%s: " fmt, \ - get_unit_name((dd)->unit), ##__VA_ARGS__) + rvt_get_ibdev_name(&(dd)->verbs_dev.rdi), \ + ##__VA_ARGS__) #define dd_dev_dbg(dd, fmt, ...) \ dev_dbg(&(dd)->pcidev->dev, "%s: " fmt, \ - get_unit_name((dd)->unit), ##__VA_ARGS__) + rvt_get_ibdev_name(&(dd)->verbs_dev.rdi), ##__VA_ARGS__) #define hfi1_dev_porterr(dd, port, fmt, ...) \ dev_err(&(dd)->pcidev->dev, "%s: port %u: " fmt, \ - get_unit_name((dd)->unit), (port), ##__VA_ARGS__) + rvt_get_ibdev_name(&(dd)->verbs_dev.rdi), (port), ##__VA_ARGS__) /* * this is used for formatting hw error messages... diff --git a/drivers/infiniband/hw/hfi1/init.c b/drivers/infiniband/hw/hfi1/init.c index 8e3b3e7d829a..9b128268fb28 100644 --- a/drivers/infiniband/hw/hfi1/init.c +++ b/drivers/infiniband/hw/hfi1/init.c @@ -1272,6 +1272,8 @@ struct hfi1_devdata *hfi1_alloc_devdata(struct pci_dev *pdev, size_t extra) "Could not allocate unit ID: error %d\n", -ret); goto bail; } + rvt_set_ibdev_name(&dd->verbs_dev.rdi, "%s_%d", class_name(), dd->unit); + /* * Initialize all locks for the device. This needs to be as early as * possible so locks are usable. diff --git a/drivers/infiniband/hw/hfi1/mad.c b/drivers/infiniband/hw/hfi1/mad.c index cf8dba34fe30..34547a48a445 100644 --- a/drivers/infiniband/hw/hfi1/mad.c +++ b/drivers/infiniband/hw/hfi1/mad.c @@ -4348,11 +4348,7 @@ static int opa_local_smp_check(struct hfi1_ibport *ibp, */ if (pkey == LIM_MGMT_P_KEY || pkey == FULL_MGMT_P_KEY) return 0; - /* - * On OPA devices it is okay to lose the upper 16 bits of LID as this - * information is obtained elsewhere. Mask off the upper 16 bits. - */ - ingress_pkey_table_fail(ppd, pkey, ib_lid_cpu16(0xFFFF & in_wc->slid)); + ingress_pkey_table_fail(ppd, pkey, in_wc->slid); return 1; } diff --git a/drivers/infiniband/hw/hfi1/qp.c b/drivers/infiniband/hw/hfi1/qp.c index 4b01ccd895b4..5507910e8b8a 100644 --- a/drivers/infiniband/hw/hfi1/qp.c +++ b/drivers/infiniband/hw/hfi1/qp.c @@ -556,6 +556,8 @@ void qp_iter_print(struct seq_file *s, struct rvt_qp_iter *iter) struct sdma_engine *sde; struct send_context *send_context; struct rvt_ack_entry *e = NULL; + struct rvt_srq *srq = qp->ibqp.srq ? + ibsrq_to_rvtsrq(qp->ibqp.srq) : NULL; sde = qp_to_sdma_engine(qp, priv->s_sc); wqe = rvt_get_swqe_ptr(qp, qp->s_last); @@ -563,7 +565,7 @@ void qp_iter_print(struct seq_file *s, struct rvt_qp_iter *iter) if (qp->s_ack_queue) e = &qp->s_ack_queue[qp->s_tail_ack_queue]; seq_printf(s, - "N %d %s QP %x R %u %s %u %u %u f=%x %u %u %u %u %u %u SPSN %x %x %x %x %x RPSN %x S(%u %u %u %u %u %u %u) R(%u %u %u) RQP %x LID %x SL %u MTU %u %u %u %u %u SDE %p,%u SC %p,%u SCQ %u %u PID %d OS %x %x E %x %x %x\n", + "N %d %s QP %x R %u %s %u %u %u f=%x %u %u %u %u %u %u SPSN %x %x %x %x %x RPSN %x S(%u %u %u %u %u %u %u) R(%u %u %u) RQP %x LID %x SL %u MTU %u %u %u %u %u SDE %p,%u SC %p,%u SCQ %u %u PID %d OS %x %x E %x %x %x RNR %d %s %d\n", iter->n, qp_idle(qp) ? "I" : "B", qp->ibqp.qp_num, @@ -610,7 +612,11 @@ void qp_iter_print(struct seq_file *s, struct rvt_qp_iter *iter) /* ack queue information */ e ? e->opcode : 0, e ? e->psn : 0, - e ? e->lpsn : 0); + e ? e->lpsn : 0, + qp->r_min_rnr_timer, + srq ? "SRQ" : "RQ", + srq ? srq->rq.size : qp->r_rq.size + ); } void *qp_priv_alloc(struct rvt_dev_info *rdi, struct rvt_qp *qp) diff --git a/drivers/infiniband/hw/hfi1/rc.c b/drivers/infiniband/hw/hfi1/rc.c index af5f7936f7e5..68d5c3cce2eb 100644 --- a/drivers/infiniband/hw/hfi1/rc.c +++ b/drivers/infiniband/hw/hfi1/rc.c @@ -843,11 +843,11 @@ static inline void hfi1_make_rc_ack_16B(struct rvt_qp *qp, /* Convert dwords to flits */ len = (*hwords + *nwords) >> 1; - hfi1_make_16b_hdr(hdr, - ppd->lid | rdma_ah_get_path_bits(&qp->remote_ah_attr), + hfi1_make_16b_hdr(hdr, ppd->lid | + (rdma_ah_get_path_bits(&qp->remote_ah_attr) & + ((1 << ppd->lmc) - 1)), opa_get_lid(rdma_ah_get_dlid(&qp->remote_ah_attr), - 16B), - len, pkey, becn, 0, l4, sc5); + 16B), len, pkey, becn, 0, l4, sc5); bth0 = pkey | (OP(ACKNOWLEDGE) << 24); bth0 |= extra_bytes << 20; diff --git a/drivers/infiniband/hw/hfi1/verbs.c b/drivers/infiniband/hw/hfi1/verbs.c index a38785e224cc..b8776a362a91 100644 --- a/drivers/infiniband/hw/hfi1/verbs.c +++ b/drivers/infiniband/hw/hfi1/verbs.c @@ -1486,7 +1486,7 @@ static int query_port(struct rvt_dev_info *rdi, u8 port_num, props->max_mtu = mtu_to_enum((!valid_ib_mtu(hfi1_max_mtu) ? 4096 : hfi1_max_mtu), IB_MTU_4096); props->active_mtu = !valid_ib_mtu(ppd->ibmtu) ? props->max_mtu : - mtu_to_enum(ppd->ibmtu, IB_MTU_2048); + mtu_to_enum(ppd->ibmtu, IB_MTU_4096); /* * sm_lid of 0xFFFF needs special handling so that it can @@ -1844,7 +1844,6 @@ int hfi1_register_ib_device(struct hfi1_devdata *dd) struct hfi1_ibport *ibp = &ppd->ibport_data; unsigned i; int ret; - size_t lcpysz = IB_DEVICE_NAME_MAX; for (i = 0; i < dd->num_pports; i++) init_ibport(ppd + i); @@ -1872,8 +1871,6 @@ int hfi1_register_ib_device(struct hfi1_devdata *dd) */ if (!ib_hfi1_sys_image_guid) ib_hfi1_sys_image_guid = ibdev->node_guid; - lcpysz = strlcpy(ibdev->name, class_name(), lcpysz); - strlcpy(ibdev->name + lcpysz, "_%d", IB_DEVICE_NAME_MAX - lcpysz); ibdev->owner = THIS_MODULE; ibdev->phys_port_cnt = dd->num_pports; ibdev->dev.parent = &dd->pcidev->dev; @@ -1893,7 +1890,6 @@ int hfi1_register_ib_device(struct hfi1_devdata *dd) * Fill in rvt info object. */ dd->verbs_dev.rdi.driver_f.port_callback = hfi1_create_port_files; - dd->verbs_dev.rdi.driver_f.get_card_name = get_card_name; dd->verbs_dev.rdi.driver_f.get_pci_dev = get_pci_dev; dd->verbs_dev.rdi.driver_f.check_ah = hfi1_check_ah; dd->verbs_dev.rdi.driver_f.notify_new_ah = hfi1_notify_new_ah; diff --git a/drivers/infiniband/hw/hns/Makefile b/drivers/infiniband/hw/hns/Makefile index ff426a625e13..97bf2cd1cacb 100644 --- a/drivers/infiniband/hw/hns/Makefile +++ b/drivers/infiniband/hw/hns/Makefile @@ -5,7 +5,7 @@ ccflags-y := -Idrivers/net/ethernet/hisilicon/hns3 obj-$(CONFIG_INFINIBAND_HNS) += hns-roce.o -hns-roce-objs := hns_roce_main.o hns_roce_cmd.o hns_roce_eq.o hns_roce_pd.o \ +hns-roce-objs := hns_roce_main.o hns_roce_cmd.o hns_roce_pd.o \ hns_roce_ah.o hns_roce_hem.o hns_roce_mr.o hns_roce_qp.o \ hns_roce_cq.o hns_roce_alloc.o obj-$(CONFIG_INFINIBAND_HNS_HIP06) += hns-roce-hw-v1.o diff --git a/drivers/infiniband/hw/hns/hns_roce_cmd.c b/drivers/infiniband/hw/hns/hns_roce_cmd.c index 1085cb249bc1..9ebe839d8b24 100644 --- a/drivers/infiniband/hw/hns/hns_roce_cmd.c +++ b/drivers/infiniband/hw/hns/hns_roce_cmd.c @@ -103,6 +103,7 @@ void hns_roce_cmd_event(struct hns_roce_dev *hr_dev, u16 token, u8 status, context->out_param = out_param; complete(&context->done); } +EXPORT_SYMBOL_GPL(hns_roce_cmd_event); /* this should be called with "use_events" */ static int __hns_roce_cmd_mbox_wait(struct hns_roce_dev *hr_dev, u64 in_param, diff --git a/drivers/infiniband/hw/hns/hns_roce_cmd.h b/drivers/infiniband/hw/hns/hns_roce_cmd.h index b1c94223c28b..9549ae51a0dd 100644 --- a/drivers/infiniband/hw/hns/hns_roce_cmd.h +++ b/drivers/infiniband/hw/hns/hns_roce_cmd.h @@ -88,6 +88,16 @@ enum { HNS_ROCE_CMD_DESTROY_SRQC_BT0 = 0x38, HNS_ROCE_CMD_DESTROY_SRQC_BT1 = 0x39, HNS_ROCE_CMD_DESTROY_SRQC_BT2 = 0x3a, + + /* EQC commands */ + HNS_ROCE_CMD_CREATE_AEQC = 0x80, + HNS_ROCE_CMD_MODIFY_AEQC = 0x81, + HNS_ROCE_CMD_QUERY_AEQC = 0x82, + HNS_ROCE_CMD_DESTROY_AEQC = 0x83, + HNS_ROCE_CMD_CREATE_CEQC = 0x90, + HNS_ROCE_CMD_MODIFY_CEQC = 0x91, + HNS_ROCE_CMD_QUERY_CEQC = 0x92, + HNS_ROCE_CMD_DESTROY_CEQC = 0x93, }; enum { diff --git a/drivers/infiniband/hw/hns/hns_roce_common.h b/drivers/infiniband/hw/hns/hns_roce_common.h index 7ecb7a4147a8..dd67fafd0c40 100644 --- a/drivers/infiniband/hw/hns/hns_roce_common.h +++ b/drivers/infiniband/hw/hns/hns_roce_common.h @@ -376,6 +376,12 @@ #define ROCEE_RX_CMQ_TAIL_REG 0x07024 #define ROCEE_RX_CMQ_HEAD_REG 0x07028 +#define ROCEE_VF_MB_CFG0_REG 0x40 +#define ROCEE_VF_MB_STATUS_REG 0x58 + +#define ROCEE_VF_EQ_DB_CFG0_REG 0x238 +#define ROCEE_VF_EQ_DB_CFG1_REG 0x23C + #define ROCEE_VF_SMAC_CFG0_REG 0x12000 #define ROCEE_VF_SMAC_CFG1_REG 0x12004 @@ -385,4 +391,9 @@ #define ROCEE_VF_SGID_CFG3_REG 0x1000c #define ROCEE_VF_SGID_CFG4_REG 0x10010 +#define ROCEE_VF_ABN_INT_CFG_REG 0x13000 +#define ROCEE_VF_ABN_INT_ST_REG 0x13004 +#define ROCEE_VF_ABN_INT_EN_REG 0x13008 +#define ROCEE_VF_EVENT_INT_EN_REG 0x1300c + #endif /* _HNS_ROCE_COMMON_H */ diff --git a/drivers/infiniband/hw/hns/hns_roce_cq.c b/drivers/infiniband/hw/hns/hns_roce_cq.c index 2111b57a3489..bccc9b54c9ce 100644 --- a/drivers/infiniband/hw/hns/hns_roce_cq.c +++ b/drivers/infiniband/hw/hns/hns_roce_cq.c @@ -196,15 +196,14 @@ void hns_roce_free_cq(struct hns_roce_dev *hr_dev, struct hns_roce_cq *hr_cq) if (ret) dev_err(dev, "HW2SW_CQ failed (%d) for CQN %06lx\n", ret, hr_cq->cqn); - if (hr_dev->eq_table.eq) { - /* Waiting interrupt process procedure carried out */ - synchronize_irq(hr_dev->eq_table.eq[hr_cq->vector].irq); - - /* wait for all interrupt processed */ - if (atomic_dec_and_test(&hr_cq->refcount)) - complete(&hr_cq->free); - wait_for_completion(&hr_cq->free); - } + + /* Waiting interrupt process procedure carried out */ + synchronize_irq(hr_dev->eq_table.eq[hr_cq->vector].irq); + + /* wait for all interrupt processed */ + if (atomic_dec_and_test(&hr_cq->refcount)) + complete(&hr_cq->free); + wait_for_completion(&hr_cq->free); spin_lock_irq(&cq_table->lock); radix_tree_delete(&cq_table->tree, hr_cq->cqn); @@ -460,6 +459,7 @@ void hns_roce_cq_completion(struct hns_roce_dev *hr_dev, u32 cqn) ++cq->arm_sn; cq->comp(cq); } +EXPORT_SYMBOL_GPL(hns_roce_cq_completion); void hns_roce_cq_event(struct hns_roce_dev *hr_dev, u32 cqn, int event_type) { @@ -482,6 +482,7 @@ void hns_roce_cq_event(struct hns_roce_dev *hr_dev, u32 cqn, int event_type) if (atomic_dec_and_test(&cq->refcount)) complete(&cq->free); } +EXPORT_SYMBOL_GPL(hns_roce_cq_event); int hns_roce_init_cq_table(struct hns_roce_dev *hr_dev) { diff --git a/drivers/infiniband/hw/hns/hns_roce_device.h b/drivers/infiniband/hw/hns/hns_roce_device.h index b154ce40cded..4afa070b20fd 100644 --- a/drivers/infiniband/hw/hns/hns_roce_device.h +++ b/drivers/infiniband/hw/hns/hns_roce_device.h @@ -62,12 +62,16 @@ #define HNS_ROCE_CQE_WCMD_EMPTY_BIT 0x2 #define HNS_ROCE_MIN_CQE_CNT 16 -#define HNS_ROCE_MAX_IRQ_NUM 34 +#define HNS_ROCE_MAX_IRQ_NUM 128 -#define HNS_ROCE_COMP_VEC_NUM 32 +#define EQ_ENABLE 1 +#define EQ_DISABLE 0 -#define HNS_ROCE_AEQE_VEC_NUM 1 -#define HNS_ROCE_AEQE_OF_VEC_NUM 1 +#define HNS_ROCE_CEQ 0 +#define HNS_ROCE_AEQ 1 + +#define HNS_ROCE_CEQ_ENTRY_SIZE 0x4 +#define HNS_ROCE_AEQ_ENTRY_SIZE 0x10 /* 4G/4K = 1M */ #define HNS_ROCE_SL_SHIFT 28 @@ -130,6 +134,7 @@ enum hns_roce_event { HNS_ROCE_EVENT_TYPE_DB_OVERFLOW = 0x12, HNS_ROCE_EVENT_TYPE_MB = 0x13, HNS_ROCE_EVENT_TYPE_CEQ_OVERFLOW = 0x14, + HNS_ROCE_EVENT_TYPE_FLR = 0x15, }; /* Local Work Queue Catastrophic Error,SUBTYPE 0x5 */ @@ -173,6 +178,7 @@ enum { enum { HNS_ROCE_CAP_FLAG_REREG_MR = BIT(0), HNS_ROCE_CAP_FLAG_ROCE_V1_V2 = BIT(1), + HNS_ROCE_CAP_FLAG_RQ_INLINE = BIT(2) }; enum hns_roce_mtt_type { @@ -441,6 +447,21 @@ struct hns_roce_cmd_mailbox { struct hns_roce_dev; +struct hns_roce_rinl_sge { + void *addr; + u32 len; +}; + +struct hns_roce_rinl_wqe { + struct hns_roce_rinl_sge *sg_list; + u32 sge_cnt; +}; + +struct hns_roce_rinl_buf { + struct hns_roce_rinl_wqe *wqe_list; + u32 wqe_cnt; +}; + struct hns_roce_qp { struct ib_qp ibqp; struct hns_roce_buf hr_buf; @@ -462,6 +483,7 @@ struct hns_roce_qp { u8 resp_depth; u8 state; u32 access_flags; + u32 atomic_rd_en; u32 pkey_index; void (*event)(struct hns_roce_qp *, enum hns_roce_event); @@ -472,6 +494,8 @@ struct hns_roce_qp { struct hns_roce_sge sge; u32 next_sge; + + struct hns_roce_rinl_buf rq_inl_buf; }; struct hns_roce_sqp { @@ -485,6 +509,45 @@ struct hns_roce_ib_iboe { u8 phy_port[HNS_ROCE_MAX_PORTS]; }; +enum { + HNS_ROCE_EQ_STAT_INVALID = 0, + HNS_ROCE_EQ_STAT_VALID = 2, +}; + +struct hns_roce_ceqe { + u32 comp; +}; + +struct hns_roce_aeqe { + u32 asyn; + union { + struct { + u32 qp; + u32 rsv0; + u32 rsv1; + } qp_event; + + struct { + u32 cq; + u32 rsv0; + u32 rsv1; + } cq_event; + + struct { + u32 ceqe; + u32 rsv0; + u32 rsv1; + } ce_event; + + struct { + __le64 out_param; + __le16 token; + u8 status; + u8 rsv0; + } __packed cmd; + } event; +}; + struct hns_roce_eq { struct hns_roce_dev *hr_dev; void __iomem *doorbell; @@ -498,11 +561,31 @@ struct hns_roce_eq { int log_page_size; int cons_index; struct hns_roce_buf_list *buf_list; + int over_ignore; + int coalesce; + int arm_st; + u64 eqe_ba; + int eqe_ba_pg_sz; + int eqe_buf_pg_sz; + int hop_num; + u64 *bt_l0; /* Base address table for L0 */ + u64 **bt_l1; /* Base address table for L1 */ + u64 **buf; + dma_addr_t l0_dma; + dma_addr_t *l1_dma; + dma_addr_t *buf_dma; + u32 l0_last_num; /* L0 last chunk num */ + u32 l1_last_num; /* L1 last chunk num */ + int eq_max_cnt; + int eq_period; + int shift; + dma_addr_t cur_eqe_ba; + dma_addr_t nxt_eqe_ba; }; struct hns_roce_eq_table { struct hns_roce_eq *eq; - void __iomem **eqc_base; + void __iomem **eqc_base; /* only for hw v1 */ }; struct hns_roce_caps { @@ -528,7 +611,7 @@ struct hns_roce_caps { u32 min_wqes; int reserved_cqs; int num_aeq_vectors; /* 1 */ - int num_comp_vectors; /* 32 ceq */ + int num_comp_vectors; int num_other_vectors; int num_mtpts; u32 num_mtt_segs; @@ -550,7 +633,7 @@ struct hns_roce_caps { u32 pbl_buf_pg_sz; u32 pbl_hop_num; int aeqe_depth; - int ceqe_depth[HNS_ROCE_COMP_VEC_NUM]; + int ceqe_depth; enum ib_mtu max_mtu; u32 qpc_bt_num; u32 srqc_bt_num; @@ -574,6 +657,9 @@ struct hns_roce_caps { u32 cqe_ba_pg_sz; u32 cqe_buf_pg_sz; u32 cqe_hop_num; + u32 eqe_ba_pg_sz; + u32 eqe_buf_pg_sz; + u32 eqe_hop_num; u32 chunk_sz; /* chunk size in non multihop mode*/ u64 flags; }; @@ -623,6 +709,8 @@ struct hns_roce_hw { int (*dereg_mr)(struct hns_roce_dev *hr_dev, struct hns_roce_mr *mr); int (*destroy_cq)(struct ib_cq *ibcq); int (*modify_cq)(struct ib_cq *cq, u16 cq_count, u16 cq_period); + int (*init_eq)(struct hns_roce_dev *hr_dev); + void (*cleanup_eq)(struct hns_roce_dev *hr_dev); }; struct hns_roce_dev { diff --git a/drivers/infiniband/hw/hns/hns_roce_eq.c b/drivers/infiniband/hw/hns/hns_roce_eq.c deleted file mode 100644 index d184431e2bf5..000000000000 --- a/drivers/infiniband/hw/hns/hns_roce_eq.c +++ /dev/null @@ -1,759 +0,0 @@ -/* - * Copyright (c) 2016 Hisilicon Limited. - * - * This software is available to you under a choice of one of two - * licenses. You may choose to be licensed under the terms of the GNU - * General Public License (GPL) Version 2, available from the file - * COPYING in the main directory of this source tree, or the - * OpenIB.org BSD license below: - * - * Redistribution and use in source and binary forms, with or - * without modification, are permitted provided that the following - * conditions are met: - * - * - Redistributions of source code must retain the above - * copyright notice, this list of conditions and the following - * disclaimer. - * - * - Redistributions in binary form must reproduce the above - * copyright notice, this list of conditions and the following - * disclaimer in the documentation and/or other materials - * provided with the distribution. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, - * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF - * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND - * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS - * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN - * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN - * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE - * SOFTWARE. - */ - -#include <linux/platform_device.h> -#include <linux/interrupt.h> -#include "hns_roce_common.h" -#include "hns_roce_device.h" -#include "hns_roce_eq.h" - -static void eq_set_cons_index(struct hns_roce_eq *eq, int req_not) -{ - roce_raw_write((eq->cons_index & CONS_INDEX_MASK) | - (req_not << eq->log_entries), eq->doorbell); - /* Memory barrier */ - mb(); -} - -static struct hns_roce_aeqe *get_aeqe(struct hns_roce_eq *eq, u32 entry) -{ - unsigned long off = (entry & (eq->entries - 1)) * - HNS_ROCE_AEQ_ENTRY_SIZE; - - return (struct hns_roce_aeqe *)((u8 *) - (eq->buf_list[off / HNS_ROCE_BA_SIZE].buf) + - off % HNS_ROCE_BA_SIZE); -} - -static struct hns_roce_aeqe *next_aeqe_sw(struct hns_roce_eq *eq) -{ - struct hns_roce_aeqe *aeqe = get_aeqe(eq, eq->cons_index); - - return (roce_get_bit(aeqe->asyn, HNS_ROCE_AEQE_U32_4_OWNER_S) ^ - !!(eq->cons_index & eq->entries)) ? aeqe : NULL; -} - -static void hns_roce_wq_catas_err_handle(struct hns_roce_dev *hr_dev, - struct hns_roce_aeqe *aeqe, int qpn) -{ - struct device *dev = &hr_dev->pdev->dev; - - dev_warn(dev, "Local Work Queue Catastrophic Error.\n"); - switch (roce_get_field(aeqe->asyn, HNS_ROCE_AEQE_U32_4_EVENT_SUB_TYPE_M, - HNS_ROCE_AEQE_U32_4_EVENT_SUB_TYPE_S)) { - case HNS_ROCE_LWQCE_QPC_ERROR: - dev_warn(dev, "QP %d, QPC error.\n", qpn); - break; - case HNS_ROCE_LWQCE_MTU_ERROR: - dev_warn(dev, "QP %d, MTU error.\n", qpn); - break; - case HNS_ROCE_LWQCE_WQE_BA_ADDR_ERROR: - dev_warn(dev, "QP %d, WQE BA addr error.\n", qpn); - break; - case HNS_ROCE_LWQCE_WQE_ADDR_ERROR: - dev_warn(dev, "QP %d, WQE addr error.\n", qpn); - break; - case HNS_ROCE_LWQCE_SQ_WQE_SHIFT_ERROR: - dev_warn(dev, "QP %d, WQE shift error\n", qpn); - break; - case HNS_ROCE_LWQCE_SL_ERROR: - dev_warn(dev, "QP %d, SL error.\n", qpn); - break; - case HNS_ROCE_LWQCE_PORT_ERROR: - dev_warn(dev, "QP %d, port error.\n", qpn); - break; - default: - break; - } -} - -static void hns_roce_local_wq_access_err_handle(struct hns_roce_dev *hr_dev, - struct hns_roce_aeqe *aeqe, - int qpn) -{ - struct device *dev = &hr_dev->pdev->dev; - - dev_warn(dev, "Local Access Violation Work Queue Error.\n"); - switch (roce_get_field(aeqe->asyn, HNS_ROCE_AEQE_U32_4_EVENT_SUB_TYPE_M, - HNS_ROCE_AEQE_U32_4_EVENT_SUB_TYPE_S)) { - case HNS_ROCE_LAVWQE_R_KEY_VIOLATION: - dev_warn(dev, "QP %d, R_key violation.\n", qpn); - break; - case HNS_ROCE_LAVWQE_LENGTH_ERROR: - dev_warn(dev, "QP %d, length error.\n", qpn); - break; - case HNS_ROCE_LAVWQE_VA_ERROR: - dev_warn(dev, "QP %d, VA error.\n", qpn); - break; - case HNS_ROCE_LAVWQE_PD_ERROR: - dev_err(dev, "QP %d, PD error.\n", qpn); - break; - case HNS_ROCE_LAVWQE_RW_ACC_ERROR: - dev_warn(dev, "QP %d, rw acc error.\n", qpn); - break; - case HNS_ROCE_LAVWQE_KEY_STATE_ERROR: - dev_warn(dev, "QP %d, key state error.\n", qpn); - break; - case HNS_ROCE_LAVWQE_MR_OPERATION_ERROR: - dev_warn(dev, "QP %d, MR operation error.\n", qpn); - break; - default: - break; - } -} - -static void hns_roce_qp_err_handle(struct hns_roce_dev *hr_dev, - struct hns_roce_aeqe *aeqe, - int event_type) -{ - struct device *dev = &hr_dev->pdev->dev; - int phy_port; - int qpn; - - qpn = roce_get_field(aeqe->event.qp_event.qp, - HNS_ROCE_AEQE_EVENT_QP_EVENT_QP_QPN_M, - HNS_ROCE_AEQE_EVENT_QP_EVENT_QP_QPN_S); - phy_port = roce_get_field(aeqe->event.qp_event.qp, - HNS_ROCE_AEQE_EVENT_QP_EVENT_PORT_NUM_M, - HNS_ROCE_AEQE_EVENT_QP_EVENT_PORT_NUM_S); - if (qpn <= 1) - qpn = HNS_ROCE_MAX_PORTS * qpn + phy_port; - - switch (event_type) { - case HNS_ROCE_EVENT_TYPE_INV_REQ_LOCAL_WQ_ERROR: - dev_warn(dev, "Invalid Req Local Work Queue Error.\n" - "QP %d, phy_port %d.\n", qpn, phy_port); - break; - case HNS_ROCE_EVENT_TYPE_WQ_CATAS_ERROR: - hns_roce_wq_catas_err_handle(hr_dev, aeqe, qpn); - break; - case HNS_ROCE_EVENT_TYPE_LOCAL_WQ_ACCESS_ERROR: - hns_roce_local_wq_access_err_handle(hr_dev, aeqe, qpn); - break; - default: - break; - } - - hns_roce_qp_event(hr_dev, qpn, event_type); -} - -static void hns_roce_cq_err_handle(struct hns_roce_dev *hr_dev, - struct hns_roce_aeqe *aeqe, - int event_type) -{ - struct device *dev = &hr_dev->pdev->dev; - u32 cqn; - - cqn = le32_to_cpu(roce_get_field(aeqe->event.cq_event.cq, - HNS_ROCE_AEQE_EVENT_CQ_EVENT_CQ_CQN_M, - HNS_ROCE_AEQE_EVENT_CQ_EVENT_CQ_CQN_S)); - - switch (event_type) { - case HNS_ROCE_EVENT_TYPE_CQ_ACCESS_ERROR: - dev_warn(dev, "CQ 0x%x access err.\n", cqn); - break; - case HNS_ROCE_EVENT_TYPE_CQ_OVERFLOW: - dev_warn(dev, "CQ 0x%x overflow\n", cqn); - break; - case HNS_ROCE_EVENT_TYPE_CQ_ID_INVALID: - dev_warn(dev, "CQ 0x%x ID invalid.\n", cqn); - break; - default: - break; - } - - hns_roce_cq_event(hr_dev, cqn, event_type); -} - -static void hns_roce_db_overflow_handle(struct hns_roce_dev *hr_dev, - struct hns_roce_aeqe *aeqe) -{ - struct device *dev = &hr_dev->pdev->dev; - - switch (roce_get_field(aeqe->asyn, HNS_ROCE_AEQE_U32_4_EVENT_SUB_TYPE_M, - HNS_ROCE_AEQE_U32_4_EVENT_SUB_TYPE_S)) { - case HNS_ROCE_DB_SUBTYPE_SDB_OVF: - dev_warn(dev, "SDB overflow.\n"); - break; - case HNS_ROCE_DB_SUBTYPE_SDB_ALM_OVF: - dev_warn(dev, "SDB almost overflow.\n"); - break; - case HNS_ROCE_DB_SUBTYPE_SDB_ALM_EMP: - dev_warn(dev, "SDB almost empty.\n"); - break; - case HNS_ROCE_DB_SUBTYPE_ODB_OVF: - dev_warn(dev, "ODB overflow.\n"); - break; - case HNS_ROCE_DB_SUBTYPE_ODB_ALM_OVF: - dev_warn(dev, "ODB almost overflow.\n"); - break; - case HNS_ROCE_DB_SUBTYPE_ODB_ALM_EMP: - dev_warn(dev, "SDB almost empty.\n"); - break; - default: - break; - } -} - -static int hns_roce_aeq_int(struct hns_roce_dev *hr_dev, struct hns_roce_eq *eq) -{ - struct device *dev = &hr_dev->pdev->dev; - struct hns_roce_aeqe *aeqe; - int aeqes_found = 0; - int event_type; - - while ((aeqe = next_aeqe_sw(eq))) { - dev_dbg(dev, "aeqe = %p, aeqe->asyn.event_type = 0x%lx\n", aeqe, - roce_get_field(aeqe->asyn, - HNS_ROCE_AEQE_U32_4_EVENT_TYPE_M, - HNS_ROCE_AEQE_U32_4_EVENT_TYPE_S)); - /* Memory barrier */ - rmb(); - - event_type = roce_get_field(aeqe->asyn, - HNS_ROCE_AEQE_U32_4_EVENT_TYPE_M, - HNS_ROCE_AEQE_U32_4_EVENT_TYPE_S); - switch (event_type) { - case HNS_ROCE_EVENT_TYPE_PATH_MIG: - dev_warn(dev, "PATH MIG not supported\n"); - break; - case HNS_ROCE_EVENT_TYPE_COMM_EST: - dev_warn(dev, "COMMUNICATION established\n"); - break; - case HNS_ROCE_EVENT_TYPE_SQ_DRAINED: - dev_warn(dev, "SQ DRAINED not supported\n"); - break; - case HNS_ROCE_EVENT_TYPE_PATH_MIG_FAILED: - dev_warn(dev, "PATH MIG failed\n"); - break; - case HNS_ROCE_EVENT_TYPE_INV_REQ_LOCAL_WQ_ERROR: - case HNS_ROCE_EVENT_TYPE_WQ_CATAS_ERROR: - case HNS_ROCE_EVENT_TYPE_LOCAL_WQ_ACCESS_ERROR: - hns_roce_qp_err_handle(hr_dev, aeqe, event_type); - break; - case HNS_ROCE_EVENT_TYPE_SRQ_LIMIT_REACH: - case HNS_ROCE_EVENT_TYPE_SRQ_CATAS_ERROR: - case HNS_ROCE_EVENT_TYPE_SRQ_LAST_WQE_REACH: - dev_warn(dev, "SRQ not support!\n"); - break; - case HNS_ROCE_EVENT_TYPE_CQ_ACCESS_ERROR: - case HNS_ROCE_EVENT_TYPE_CQ_OVERFLOW: - case HNS_ROCE_EVENT_TYPE_CQ_ID_INVALID: - hns_roce_cq_err_handle(hr_dev, aeqe, event_type); - break; - case HNS_ROCE_EVENT_TYPE_PORT_CHANGE: - dev_warn(dev, "port change.\n"); - break; - case HNS_ROCE_EVENT_TYPE_MB: - hns_roce_cmd_event(hr_dev, - le16_to_cpu(aeqe->event.cmd.token), - aeqe->event.cmd.status, - le64_to_cpu(aeqe->event.cmd.out_param - )); - break; - case HNS_ROCE_EVENT_TYPE_DB_OVERFLOW: - hns_roce_db_overflow_handle(hr_dev, aeqe); - break; - case HNS_ROCE_EVENT_TYPE_CEQ_OVERFLOW: - dev_warn(dev, "CEQ 0x%lx overflow.\n", - roce_get_field(aeqe->event.ce_event.ceqe, - HNS_ROCE_AEQE_EVENT_CE_EVENT_CEQE_CEQN_M, - HNS_ROCE_AEQE_EVENT_CE_EVENT_CEQE_CEQN_S)); - break; - default: - dev_warn(dev, "Unhandled event %d on EQ %d at index %u\n", - event_type, eq->eqn, eq->cons_index); - break; - } - - eq->cons_index++; - aeqes_found = 1; - - if (eq->cons_index > 2 * hr_dev->caps.aeqe_depth - 1) { - dev_warn(dev, "cons_index overflow, set back to zero\n" - ); - eq->cons_index = 0; - } - } - - eq_set_cons_index(eq, 0); - - return aeqes_found; -} - -static struct hns_roce_ceqe *get_ceqe(struct hns_roce_eq *eq, u32 entry) -{ - unsigned long off = (entry & (eq->entries - 1)) * - HNS_ROCE_CEQ_ENTRY_SIZE; - - return (struct hns_roce_ceqe *)((u8 *) - (eq->buf_list[off / HNS_ROCE_BA_SIZE].buf) + - off % HNS_ROCE_BA_SIZE); -} - -static struct hns_roce_ceqe *next_ceqe_sw(struct hns_roce_eq *eq) -{ - struct hns_roce_ceqe *ceqe = get_ceqe(eq, eq->cons_index); - - return (!!(roce_get_bit(ceqe->ceqe.comp, - HNS_ROCE_CEQE_CEQE_COMP_OWNER_S))) ^ - (!!(eq->cons_index & eq->entries)) ? ceqe : NULL; -} - -static int hns_roce_ceq_int(struct hns_roce_dev *hr_dev, struct hns_roce_eq *eq) -{ - struct hns_roce_ceqe *ceqe; - int ceqes_found = 0; - u32 cqn; - - while ((ceqe = next_ceqe_sw(eq))) { - /* Memory barrier */ - rmb(); - cqn = roce_get_field(ceqe->ceqe.comp, - HNS_ROCE_CEQE_CEQE_COMP_CQN_M, - HNS_ROCE_CEQE_CEQE_COMP_CQN_S); - hns_roce_cq_completion(hr_dev, cqn); - - ++eq->cons_index; - ceqes_found = 1; - - if (eq->cons_index > 2 * hr_dev->caps.ceqe_depth[eq->eqn] - 1) { - dev_warn(&eq->hr_dev->pdev->dev, - "cons_index overflow, set back to zero\n"); - eq->cons_index = 0; - } - } - - eq_set_cons_index(eq, 0); - - return ceqes_found; -} - -static int hns_roce_aeq_ovf_int(struct hns_roce_dev *hr_dev, - struct hns_roce_eq *eq) -{ - struct device *dev = &eq->hr_dev->pdev->dev; - int eqovf_found = 0; - u32 caepaemask_val; - u32 cealmovf_val; - u32 caepaest_val; - u32 aeshift_val; - u32 ceshift_val; - u32 cemask_val; - int i = 0; - - /** - * AEQ overflow ECC mult bit err CEQ overflow alarm - * must clear interrupt, mask irq, clear irq, cancel mask operation - */ - aeshift_val = roce_read(hr_dev, ROCEE_CAEP_AEQC_AEQE_SHIFT_REG); - - if (roce_get_bit(aeshift_val, - ROCEE_CAEP_AEQC_AEQE_SHIFT_CAEP_AEQ_ALM_OVF_INT_ST_S) == 1) { - dev_warn(dev, "AEQ overflow!\n"); - - /* Set mask */ - caepaemask_val = roce_read(hr_dev, ROCEE_CAEP_AE_MASK_REG); - roce_set_bit(caepaemask_val, - ROCEE_CAEP_AE_MASK_CAEP_AEQ_ALM_OVF_MASK_S, - HNS_ROCE_INT_MASK_ENABLE); - roce_write(hr_dev, ROCEE_CAEP_AE_MASK_REG, caepaemask_val); - - /* Clear int state(INT_WC : write 1 clear) */ - caepaest_val = roce_read(hr_dev, ROCEE_CAEP_AE_ST_REG); - roce_set_bit(caepaest_val, - ROCEE_CAEP_AE_ST_CAEP_AEQ_ALM_OVF_S, 1); - roce_write(hr_dev, ROCEE_CAEP_AE_ST_REG, caepaest_val); - - /* Clear mask */ - caepaemask_val = roce_read(hr_dev, ROCEE_CAEP_AE_MASK_REG); - roce_set_bit(caepaemask_val, - ROCEE_CAEP_AE_MASK_CAEP_AEQ_ALM_OVF_MASK_S, - HNS_ROCE_INT_MASK_DISABLE); - roce_write(hr_dev, ROCEE_CAEP_AE_MASK_REG, caepaemask_val); - } - - /* CEQ almost overflow */ - for (i = 0; i < hr_dev->caps.num_comp_vectors; i++) { - ceshift_val = roce_read(hr_dev, ROCEE_CAEP_CEQC_SHIFT_0_REG + - i * CEQ_REG_OFFSET); - - if (roce_get_bit(ceshift_val, - ROCEE_CAEP_CEQC_SHIFT_CAEP_CEQ_ALM_OVF_INT_ST_S) == 1) { - dev_warn(dev, "CEQ[%d] almost overflow!\n", i); - eqovf_found++; - - /* Set mask */ - cemask_val = roce_read(hr_dev, - ROCEE_CAEP_CE_IRQ_MASK_0_REG + - i * CEQ_REG_OFFSET); - roce_set_bit(cemask_val, - ROCEE_CAEP_CE_IRQ_MASK_CAEP_CEQ_ALM_OVF_MASK_S, - HNS_ROCE_INT_MASK_ENABLE); - roce_write(hr_dev, ROCEE_CAEP_CE_IRQ_MASK_0_REG + - i * CEQ_REG_OFFSET, cemask_val); - - /* Clear int state(INT_WC : write 1 clear) */ - cealmovf_val = roce_read(hr_dev, - ROCEE_CAEP_CEQ_ALM_OVF_0_REG + - i * CEQ_REG_OFFSET); - roce_set_bit(cealmovf_val, - ROCEE_CAEP_CEQ_ALM_OVF_CAEP_CEQ_ALM_OVF_S, - 1); - roce_write(hr_dev, ROCEE_CAEP_CEQ_ALM_OVF_0_REG + - i * CEQ_REG_OFFSET, cealmovf_val); - - /* Clear mask */ - cemask_val = roce_read(hr_dev, - ROCEE_CAEP_CE_IRQ_MASK_0_REG + - i * CEQ_REG_OFFSET); - roce_set_bit(cemask_val, - ROCEE_CAEP_CE_IRQ_MASK_CAEP_CEQ_ALM_OVF_MASK_S, - HNS_ROCE_INT_MASK_DISABLE); - roce_write(hr_dev, ROCEE_CAEP_CE_IRQ_MASK_0_REG + - i * CEQ_REG_OFFSET, cemask_val); - } - } - - /* ECC multi-bit error alarm */ - dev_warn(dev, "ECC UCERR ALARM: 0x%x, 0x%x, 0x%x\n", - roce_read(hr_dev, ROCEE_ECC_UCERR_ALM0_REG), - roce_read(hr_dev, ROCEE_ECC_UCERR_ALM1_REG), - roce_read(hr_dev, ROCEE_ECC_UCERR_ALM2_REG)); - - dev_warn(dev, "ECC CERR ALARM: 0x%x, 0x%x, 0x%x\n", - roce_read(hr_dev, ROCEE_ECC_CERR_ALM0_REG), - roce_read(hr_dev, ROCEE_ECC_CERR_ALM1_REG), - roce_read(hr_dev, ROCEE_ECC_CERR_ALM2_REG)); - - return eqovf_found; -} - -static int hns_roce_eq_int(struct hns_roce_dev *hr_dev, struct hns_roce_eq *eq) -{ - int eqes_found = 0; - - if (likely(eq->type_flag == HNS_ROCE_CEQ)) - /* CEQ irq routine, CEQ is pulse irq, not clear */ - eqes_found = hns_roce_ceq_int(hr_dev, eq); - else if (likely(eq->type_flag == HNS_ROCE_AEQ)) - /* AEQ irq routine, AEQ is pulse irq, not clear */ - eqes_found = hns_roce_aeq_int(hr_dev, eq); - else - /* AEQ queue overflow irq */ - eqes_found = hns_roce_aeq_ovf_int(hr_dev, eq); - - return eqes_found; -} - -static irqreturn_t hns_roce_msi_x_interrupt(int irq, void *eq_ptr) -{ - int int_work = 0; - struct hns_roce_eq *eq = eq_ptr; - struct hns_roce_dev *hr_dev = eq->hr_dev; - - int_work = hns_roce_eq_int(hr_dev, eq); - - return IRQ_RETVAL(int_work); -} - -static void hns_roce_enable_eq(struct hns_roce_dev *hr_dev, int eq_num, - int enable_flag) -{ - void __iomem *eqc = hr_dev->eq_table.eqc_base[eq_num]; - u32 val; - - val = readl(eqc); - - if (enable_flag) - roce_set_field(val, - ROCEE_CAEP_AEQC_AEQE_SHIFT_CAEP_AEQC_STATE_M, - ROCEE_CAEP_AEQC_AEQE_SHIFT_CAEP_AEQC_STATE_S, - HNS_ROCE_EQ_STAT_VALID); - else - roce_set_field(val, - ROCEE_CAEP_AEQC_AEQE_SHIFT_CAEP_AEQC_STATE_M, - ROCEE_CAEP_AEQC_AEQE_SHIFT_CAEP_AEQC_STATE_S, - HNS_ROCE_EQ_STAT_INVALID); - writel(val, eqc); -} - -static int hns_roce_create_eq(struct hns_roce_dev *hr_dev, - struct hns_roce_eq *eq) -{ - void __iomem *eqc = hr_dev->eq_table.eqc_base[eq->eqn]; - struct device *dev = &hr_dev->pdev->dev; - dma_addr_t tmp_dma_addr; - u32 eqconsindx_val = 0; - u32 eqcuridx_val = 0; - u32 eqshift_val = 0; - int num_bas = 0; - int ret; - int i; - - num_bas = (PAGE_ALIGN(eq->entries * eq->eqe_size) + - HNS_ROCE_BA_SIZE - 1) / HNS_ROCE_BA_SIZE; - - if ((eq->entries * eq->eqe_size) > HNS_ROCE_BA_SIZE) { - dev_err(dev, "[error]eq buf %d gt ba size(%d) need bas=%d\n", - (eq->entries * eq->eqe_size), HNS_ROCE_BA_SIZE, - num_bas); - return -EINVAL; - } - - eq->buf_list = kcalloc(num_bas, sizeof(*eq->buf_list), GFP_KERNEL); - if (!eq->buf_list) - return -ENOMEM; - - for (i = 0; i < num_bas; ++i) { - eq->buf_list[i].buf = dma_alloc_coherent(dev, HNS_ROCE_BA_SIZE, - &tmp_dma_addr, - GFP_KERNEL); - if (!eq->buf_list[i].buf) { - ret = -ENOMEM; - goto err_out_free_pages; - } - - eq->buf_list[i].map = tmp_dma_addr; - memset(eq->buf_list[i].buf, 0, HNS_ROCE_BA_SIZE); - } - eq->cons_index = 0; - roce_set_field(eqshift_val, - ROCEE_CAEP_AEQC_AEQE_SHIFT_CAEP_AEQC_STATE_M, - ROCEE_CAEP_AEQC_AEQE_SHIFT_CAEP_AEQC_STATE_S, - HNS_ROCE_EQ_STAT_INVALID); - roce_set_field(eqshift_val, - ROCEE_CAEP_AEQC_AEQE_SHIFT_CAEP_AEQC_AEQE_SHIFT_M, - ROCEE_CAEP_AEQC_AEQE_SHIFT_CAEP_AEQC_AEQE_SHIFT_S, - eq->log_entries); - writel(eqshift_val, eqc); - - /* Configure eq extended address 12~44bit */ - writel((u32)(eq->buf_list[0].map >> 12), eqc + 4); - - /* - * Configure eq extended address 45~49 bit. - * 44 = 32 + 12, When evaluating addr to hardware, shift 12 because of - * using 4K page, and shift more 32 because of - * caculating the high 32 bit value evaluated to hardware. - */ - roce_set_field(eqcuridx_val, ROCEE_CAEP_AEQE_CUR_IDX_CAEP_AEQ_BT_H_M, - ROCEE_CAEP_AEQE_CUR_IDX_CAEP_AEQ_BT_H_S, - eq->buf_list[0].map >> 44); - roce_set_field(eqcuridx_val, - ROCEE_CAEP_AEQE_CUR_IDX_CAEP_AEQE_CUR_IDX_M, - ROCEE_CAEP_AEQE_CUR_IDX_CAEP_AEQE_CUR_IDX_S, 0); - writel(eqcuridx_val, eqc + 8); - - /* Configure eq consumer index */ - roce_set_field(eqconsindx_val, - ROCEE_CAEP_AEQE_CONS_IDX_CAEP_AEQE_CONS_IDX_M, - ROCEE_CAEP_AEQE_CONS_IDX_CAEP_AEQE_CONS_IDX_S, 0); - writel(eqconsindx_val, eqc + 0xc); - - return 0; - -err_out_free_pages: - for (i = i - 1; i >= 0; i--) - dma_free_coherent(dev, HNS_ROCE_BA_SIZE, eq->buf_list[i].buf, - eq->buf_list[i].map); - - kfree(eq->buf_list); - return ret; -} - -static void hns_roce_free_eq(struct hns_roce_dev *hr_dev, - struct hns_roce_eq *eq) -{ - int i = 0; - int npages = (PAGE_ALIGN(eq->eqe_size * eq->entries) + - HNS_ROCE_BA_SIZE - 1) / HNS_ROCE_BA_SIZE; - - if (!eq->buf_list) - return; - - for (i = 0; i < npages; ++i) - dma_free_coherent(&hr_dev->pdev->dev, HNS_ROCE_BA_SIZE, - eq->buf_list[i].buf, eq->buf_list[i].map); - - kfree(eq->buf_list); -} - -static void hns_roce_int_mask_en(struct hns_roce_dev *hr_dev) -{ - int i = 0; - u32 aemask_val; - int masken = 0; - - /* AEQ INT */ - aemask_val = roce_read(hr_dev, ROCEE_CAEP_AE_MASK_REG); - roce_set_bit(aemask_val, ROCEE_CAEP_AE_MASK_CAEP_AEQ_ALM_OVF_MASK_S, - masken); - roce_set_bit(aemask_val, ROCEE_CAEP_AE_MASK_CAEP_AE_IRQ_MASK_S, masken); - roce_write(hr_dev, ROCEE_CAEP_AE_MASK_REG, aemask_val); - - /* CEQ INT */ - for (i = 0; i < hr_dev->caps.num_comp_vectors; i++) { - /* IRQ mask */ - roce_write(hr_dev, ROCEE_CAEP_CE_IRQ_MASK_0_REG + - i * CEQ_REG_OFFSET, masken); - } -} - -static void hns_roce_ce_int_default_cfg(struct hns_roce_dev *hr_dev) -{ - /* Configure ce int interval */ - roce_write(hr_dev, ROCEE_CAEP_CE_INTERVAL_CFG_REG, - HNS_ROCE_CEQ_DEFAULT_INTERVAL); - - /* Configure ce int burst num */ - roce_write(hr_dev, ROCEE_CAEP_CE_BURST_NUM_CFG_REG, - HNS_ROCE_CEQ_DEFAULT_BURST_NUM); -} - -int hns_roce_init_eq_table(struct hns_roce_dev *hr_dev) -{ - struct hns_roce_eq_table *eq_table = &hr_dev->eq_table; - struct device *dev = &hr_dev->pdev->dev; - struct hns_roce_eq *eq = NULL; - int eq_num = 0; - int ret = 0; - int i = 0; - int j = 0; - - eq_num = hr_dev->caps.num_comp_vectors + hr_dev->caps.num_aeq_vectors; - eq_table->eq = kcalloc(eq_num, sizeof(*eq_table->eq), GFP_KERNEL); - if (!eq_table->eq) - return -ENOMEM; - - eq_table->eqc_base = kcalloc(eq_num, sizeof(*eq_table->eqc_base), - GFP_KERNEL); - if (!eq_table->eqc_base) { - ret = -ENOMEM; - goto err_eqc_base_alloc_fail; - } - - for (i = 0; i < eq_num; i++) { - eq = &eq_table->eq[i]; - eq->hr_dev = hr_dev; - eq->eqn = i; - eq->irq = hr_dev->irq[i]; - eq->log_page_size = PAGE_SHIFT; - - if (i < hr_dev->caps.num_comp_vectors) { - /* CEQ */ - eq_table->eqc_base[i] = hr_dev->reg_base + - ROCEE_CAEP_CEQC_SHIFT_0_REG + - HNS_ROCE_CEQC_REG_OFFSET * i; - eq->type_flag = HNS_ROCE_CEQ; - eq->doorbell = hr_dev->reg_base + - ROCEE_CAEP_CEQC_CONS_IDX_0_REG + - HNS_ROCE_CEQC_REG_OFFSET * i; - eq->entries = hr_dev->caps.ceqe_depth[i]; - eq->log_entries = ilog2(eq->entries); - eq->eqe_size = sizeof(struct hns_roce_ceqe); - } else { - /* AEQ */ - eq_table->eqc_base[i] = hr_dev->reg_base + - ROCEE_CAEP_AEQC_AEQE_SHIFT_REG; - eq->type_flag = HNS_ROCE_AEQ; - eq->doorbell = hr_dev->reg_base + - ROCEE_CAEP_AEQE_CONS_IDX_REG; - eq->entries = hr_dev->caps.aeqe_depth; - eq->log_entries = ilog2(eq->entries); - eq->eqe_size = sizeof(struct hns_roce_aeqe); - } - } - - /* Disable irq */ - hns_roce_int_mask_en(hr_dev); - - /* Configure CE irq interval and burst num */ - hns_roce_ce_int_default_cfg(hr_dev); - - for (i = 0; i < eq_num; i++) { - ret = hns_roce_create_eq(hr_dev, &eq_table->eq[i]); - if (ret) { - dev_err(dev, "eq create failed\n"); - goto err_create_eq_fail; - } - } - - for (j = 0; j < eq_num; j++) { - ret = request_irq(eq_table->eq[j].irq, hns_roce_msi_x_interrupt, - 0, hr_dev->irq_names[j], eq_table->eq + j); - if (ret) { - dev_err(dev, "request irq error!\n"); - goto err_request_irq_fail; - } - } - - for (i = 0; i < eq_num; i++) - hns_roce_enable_eq(hr_dev, i, EQ_ENABLE); - - return 0; - -err_request_irq_fail: - for (j = j - 1; j >= 0; j--) - free_irq(eq_table->eq[j].irq, eq_table->eq + j); - -err_create_eq_fail: - for (i = i - 1; i >= 0; i--) - hns_roce_free_eq(hr_dev, &eq_table->eq[i]); - - kfree(eq_table->eqc_base); - -err_eqc_base_alloc_fail: - kfree(eq_table->eq); - - return ret; -} - -void hns_roce_cleanup_eq_table(struct hns_roce_dev *hr_dev) -{ - int i; - int eq_num; - struct hns_roce_eq_table *eq_table = &hr_dev->eq_table; - - eq_num = hr_dev->caps.num_comp_vectors + hr_dev->caps.num_aeq_vectors; - for (i = 0; i < eq_num; i++) { - /* Disable EQ */ - hns_roce_enable_eq(hr_dev, i, EQ_DISABLE); - - free_irq(eq_table->eq[i].irq, eq_table->eq + i); - - hns_roce_free_eq(hr_dev, &eq_table->eq[i]); - } - - kfree(eq_table->eqc_base); - kfree(eq_table->eq); -} diff --git a/drivers/infiniband/hw/hns/hns_roce_eq.h b/drivers/infiniband/hw/hns/hns_roce_eq.h deleted file mode 100644 index c6d212d12e03..000000000000 --- a/drivers/infiniband/hw/hns/hns_roce_eq.h +++ /dev/null @@ -1,134 +0,0 @@ -/* - * Copyright (c) 2016 Hisilicon Limited. - * - * This software is available to you under a choice of one of two - * licenses. You may choose to be licensed under the terms of the GNU - * General Public License (GPL) Version 2, available from the file - * COPYING in the main directory of this source tree, or the - * OpenIB.org BSD license below: - * - * Redistribution and use in source and binary forms, with or - * without modification, are permitted provided that the following - * conditions are met: - * - * - Redistributions of source code must retain the above - * copyright notice, this list of conditions and the following - * disclaimer. - * - * - Redistributions in binary form must reproduce the above - * copyright notice, this list of conditions and the following - * disclaimer in the documentation and/or other materials - * provided with the distribution. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, - * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF - * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND - * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS - * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN - * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN - * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE - * SOFTWARE. - */ - -#ifndef _HNS_ROCE_EQ_H -#define _HNS_ROCE_EQ_H - -#define HNS_ROCE_CEQ 1 -#define HNS_ROCE_AEQ 2 - -#define HNS_ROCE_CEQ_ENTRY_SIZE 0x4 -#define HNS_ROCE_AEQ_ENTRY_SIZE 0x10 -#define HNS_ROCE_CEQC_REG_OFFSET 0x18 - -#define HNS_ROCE_CEQ_DEFAULT_INTERVAL 0x10 -#define HNS_ROCE_CEQ_DEFAULT_BURST_NUM 0x10 - -#define HNS_ROCE_INT_MASK_DISABLE 0 -#define HNS_ROCE_INT_MASK_ENABLE 1 - -#define EQ_ENABLE 1 -#define EQ_DISABLE 0 -#define CONS_INDEX_MASK 0xffff - -#define CEQ_REG_OFFSET 0x18 - -enum { - HNS_ROCE_EQ_STAT_INVALID = 0, - HNS_ROCE_EQ_STAT_VALID = 2, -}; - -struct hns_roce_aeqe { - u32 asyn; - union { - struct { - u32 qp; - u32 rsv0; - u32 rsv1; - } qp_event; - - struct { - u32 cq; - u32 rsv0; - u32 rsv1; - } cq_event; - - struct { - u32 port; - u32 rsv0; - u32 rsv1; - } port_event; - - struct { - u32 ceqe; - u32 rsv0; - u32 rsv1; - } ce_event; - - struct { - __le64 out_param; - __le16 token; - u8 status; - u8 rsv0; - } __packed cmd; - } event; -}; - -#define HNS_ROCE_AEQE_U32_4_EVENT_TYPE_S 16 -#define HNS_ROCE_AEQE_U32_4_EVENT_TYPE_M \ - (((1UL << 8) - 1) << HNS_ROCE_AEQE_U32_4_EVENT_TYPE_S) - -#define HNS_ROCE_AEQE_U32_4_EVENT_SUB_TYPE_S 24 -#define HNS_ROCE_AEQE_U32_4_EVENT_SUB_TYPE_M \ - (((1UL << 7) - 1) << HNS_ROCE_AEQE_U32_4_EVENT_SUB_TYPE_S) - -#define HNS_ROCE_AEQE_U32_4_OWNER_S 31 - -#define HNS_ROCE_AEQE_EVENT_QP_EVENT_QP_QPN_S 0 -#define HNS_ROCE_AEQE_EVENT_QP_EVENT_QP_QPN_M \ - (((1UL << 24) - 1) << HNS_ROCE_AEQE_EVENT_QP_EVENT_QP_QPN_S) - -#define HNS_ROCE_AEQE_EVENT_QP_EVENT_PORT_NUM_S 25 -#define HNS_ROCE_AEQE_EVENT_QP_EVENT_PORT_NUM_M \ - (((1UL << 3) - 1) << HNS_ROCE_AEQE_EVENT_QP_EVENT_PORT_NUM_S) - -#define HNS_ROCE_AEQE_EVENT_CQ_EVENT_CQ_CQN_S 0 -#define HNS_ROCE_AEQE_EVENT_CQ_EVENT_CQ_CQN_M \ - (((1UL << 16) - 1) << HNS_ROCE_AEQE_EVENT_CQ_EVENT_CQ_CQN_S) - -#define HNS_ROCE_AEQE_EVENT_CE_EVENT_CEQE_CEQN_S 0 -#define HNS_ROCE_AEQE_EVENT_CE_EVENT_CEQE_CEQN_M \ - (((1UL << 5) - 1) << HNS_ROCE_AEQE_EVENT_CE_EVENT_CEQE_CEQN_S) - -struct hns_roce_ceqe { - union { - int comp; - } ceqe; -}; - -#define HNS_ROCE_CEQE_CEQE_COMP_OWNER_S 0 - -#define HNS_ROCE_CEQE_CEQE_COMP_CQN_S 16 -#define HNS_ROCE_CEQE_CEQE_COMP_CQN_M \ - (((1UL << 16) - 1) << HNS_ROCE_CEQE_CEQE_COMP_CQN_S) - -#endif /* _HNS_ROCE_EQ_H */ diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c index af27168faf0f..939355ede14a 100644 --- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c +++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c @@ -33,6 +33,7 @@ #include <linux/platform_device.h> #include <linux/acpi.h> #include <linux/etherdevice.h> +#include <linux/interrupt.h> #include <linux/of.h> #include <linux/of_platform.h> #include <rdma/ib_umem.h> @@ -774,7 +775,7 @@ static int hns_roce_v1_rsv_lp_qp(struct hns_roce_dev *hr_dev) goto create_lp_qp_failed; } - ret = hr_dev->hw->modify_qp(&hr_qp->ibqp, &attr, attr_mask, + ret = hr_dev->hw->modify_qp(&hr_qp->ibqp, &attr, IB_QP_DEST_QPN, IB_QPS_INIT, IB_QPS_RTR); if (ret) { dev_err(dev, "modify qp failed(%d)!\n", ret); @@ -1492,9 +1493,9 @@ static int hns_roce_v1_profile(struct hns_roce_dev *hr_dev) caps->max_sq_inline = HNS_ROCE_V1_INLINE_SIZE; caps->num_uars = HNS_ROCE_V1_UAR_NUM; caps->phy_num_uars = HNS_ROCE_V1_PHY_UAR_NUM; - caps->num_aeq_vectors = HNS_ROCE_AEQE_VEC_NUM; - caps->num_comp_vectors = HNS_ROCE_COMP_VEC_NUM; - caps->num_other_vectors = HNS_ROCE_AEQE_OF_VEC_NUM; + caps->num_aeq_vectors = HNS_ROCE_V1_AEQE_VEC_NUM; + caps->num_comp_vectors = HNS_ROCE_V1_COMP_VEC_NUM; + caps->num_other_vectors = HNS_ROCE_V1_ABNORMAL_VEC_NUM; caps->num_mtpts = HNS_ROCE_V1_MAX_MTPT_NUM; caps->num_mtt_segs = HNS_ROCE_V1_MAX_MTT_SEGS; caps->num_pds = HNS_ROCE_V1_MAX_PD_NUM; @@ -1529,10 +1530,8 @@ static int hns_roce_v1_profile(struct hns_roce_dev *hr_dev) caps->num_ports + 1; } - for (i = 0; i < caps->num_comp_vectors; i++) - caps->ceqe_depth[i] = HNS_ROCE_V1_NUM_COMP_EQE; - - caps->aeqe_depth = HNS_ROCE_V1_NUM_ASYNC_EQE; + caps->ceqe_depth = HNS_ROCE_V1_COMP_EQE_NUM; + caps->aeqe_depth = HNS_ROCE_V1_ASYNC_EQE_NUM; caps->local_ca_ack_delay = le32_to_cpu(roce_read(hr_dev, ROCEE_ACK_DELAY_REG)); caps->max_mtu = IB_MTU_2048; @@ -3960,6 +3959,732 @@ static int hns_roce_v1_destroy_cq(struct ib_cq *ibcq) return ret; } +static void set_eq_cons_index_v1(struct hns_roce_eq *eq, int req_not) +{ + roce_raw_write((eq->cons_index & HNS_ROCE_V1_CONS_IDX_M) | + (req_not << eq->log_entries), eq->doorbell); +} + +static void hns_roce_v1_wq_catas_err_handle(struct hns_roce_dev *hr_dev, + struct hns_roce_aeqe *aeqe, int qpn) +{ + struct device *dev = &hr_dev->pdev->dev; + + dev_warn(dev, "Local Work Queue Catastrophic Error.\n"); + switch (roce_get_field(aeqe->asyn, HNS_ROCE_AEQE_U32_4_EVENT_SUB_TYPE_M, + HNS_ROCE_AEQE_U32_4_EVENT_SUB_TYPE_S)) { + case HNS_ROCE_LWQCE_QPC_ERROR: + dev_warn(dev, "QP %d, QPC error.\n", qpn); + break; + case HNS_ROCE_LWQCE_MTU_ERROR: + dev_warn(dev, "QP %d, MTU error.\n", qpn); + break; + case HNS_ROCE_LWQCE_WQE_BA_ADDR_ERROR: + dev_warn(dev, "QP %d, WQE BA addr error.\n", qpn); + break; + case HNS_ROCE_LWQCE_WQE_ADDR_ERROR: + dev_warn(dev, "QP %d, WQE addr error.\n", qpn); + break; + case HNS_ROCE_LWQCE_SQ_WQE_SHIFT_ERROR: + dev_warn(dev, "QP %d, WQE shift error\n", qpn); + break; + case HNS_ROCE_LWQCE_SL_ERROR: + dev_warn(dev, "QP %d, SL error.\n", qpn); + break; + case HNS_ROCE_LWQCE_PORT_ERROR: + dev_warn(dev, "QP %d, port error.\n", qpn); + break; + default: + break; + } +} + +static void hns_roce_v1_local_wq_access_err_handle(struct hns_roce_dev *hr_dev, + struct hns_roce_aeqe *aeqe, + int qpn) +{ + struct device *dev = &hr_dev->pdev->dev; + + dev_warn(dev, "Local Access Violation Work Queue Error.\n"); + switch (roce_get_field(aeqe->asyn, HNS_ROCE_AEQE_U32_4_EVENT_SUB_TYPE_M, + HNS_ROCE_AEQE_U32_4_EVENT_SUB_TYPE_S)) { + case HNS_ROCE_LAVWQE_R_KEY_VIOLATION: + dev_warn(dev, "QP %d, R_key violation.\n", qpn); + break; + case HNS_ROCE_LAVWQE_LENGTH_ERROR: + dev_warn(dev, "QP %d, length error.\n", qpn); + break; + case HNS_ROCE_LAVWQE_VA_ERROR: + dev_warn(dev, "QP %d, VA error.\n", qpn); + break; + case HNS_ROCE_LAVWQE_PD_ERROR: + dev_err(dev, "QP %d, PD error.\n", qpn); + break; + case HNS_ROCE_LAVWQE_RW_ACC_ERROR: + dev_warn(dev, "QP %d, rw acc error.\n", qpn); + break; + case HNS_ROCE_LAVWQE_KEY_STATE_ERROR: + dev_warn(dev, "QP %d, key state error.\n", qpn); + break; + case HNS_ROCE_LAVWQE_MR_OPERATION_ERROR: + dev_warn(dev, "QP %d, MR operation error.\n", qpn); + break; + default: + break; + } +} + +static void hns_roce_v1_qp_err_handle(struct hns_roce_dev *hr_dev, + struct hns_roce_aeqe *aeqe, + int event_type) +{ + struct device *dev = &hr_dev->pdev->dev; + int phy_port; + int qpn; + + qpn = roce_get_field(aeqe->event.qp_event.qp, + HNS_ROCE_AEQE_EVENT_QP_EVENT_QP_QPN_M, + HNS_ROCE_AEQE_EVENT_QP_EVENT_QP_QPN_S); + phy_port = roce_get_field(aeqe->event.qp_event.qp, + HNS_ROCE_AEQE_EVENT_QP_EVENT_PORT_NUM_M, + HNS_ROCE_AEQE_EVENT_QP_EVENT_PORT_NUM_S); + if (qpn <= 1) + qpn = HNS_ROCE_MAX_PORTS * qpn + phy_port; + + switch (event_type) { + case HNS_ROCE_EVENT_TYPE_INV_REQ_LOCAL_WQ_ERROR: + dev_warn(dev, "Invalid Req Local Work Queue Error.\n" + "QP %d, phy_port %d.\n", qpn, phy_port); + break; + case HNS_ROCE_EVENT_TYPE_WQ_CATAS_ERROR: + hns_roce_v1_wq_catas_err_handle(hr_dev, aeqe, qpn); + break; + case HNS_ROCE_EVENT_TYPE_LOCAL_WQ_ACCESS_ERROR: + hns_roce_v1_local_wq_access_err_handle(hr_dev, aeqe, qpn); + break; + default: + break; + } + + hns_roce_qp_event(hr_dev, qpn, event_type); +} + +static void hns_roce_v1_cq_err_handle(struct hns_roce_dev *hr_dev, + struct hns_roce_aeqe *aeqe, + int event_type) +{ + struct device *dev = &hr_dev->pdev->dev; + u32 cqn; + + cqn = le32_to_cpu(roce_get_field(aeqe->event.cq_event.cq, + HNS_ROCE_AEQE_EVENT_CQ_EVENT_CQ_CQN_M, + HNS_ROCE_AEQE_EVENT_CQ_EVENT_CQ_CQN_S)); + + switch (event_type) { + case HNS_ROCE_EVENT_TYPE_CQ_ACCESS_ERROR: + dev_warn(dev, "CQ 0x%x access err.\n", cqn); + break; + case HNS_ROCE_EVENT_TYPE_CQ_OVERFLOW: + dev_warn(dev, "CQ 0x%x overflow\n", cqn); + break; + case HNS_ROCE_EVENT_TYPE_CQ_ID_INVALID: + dev_warn(dev, "CQ 0x%x ID invalid.\n", cqn); + break; + default: + break; + } + + hns_roce_cq_event(hr_dev, cqn, event_type); +} + +static void hns_roce_v1_db_overflow_handle(struct hns_roce_dev *hr_dev, + struct hns_roce_aeqe *aeqe) +{ + struct device *dev = &hr_dev->pdev->dev; + + switch (roce_get_field(aeqe->asyn, HNS_ROCE_AEQE_U32_4_EVENT_SUB_TYPE_M, + HNS_ROCE_AEQE_U32_4_EVENT_SUB_TYPE_S)) { + case HNS_ROCE_DB_SUBTYPE_SDB_OVF: + dev_warn(dev, "SDB overflow.\n"); + break; + case HNS_ROCE_DB_SUBTYPE_SDB_ALM_OVF: + dev_warn(dev, "SDB almost overflow.\n"); + break; + case HNS_ROCE_DB_SUBTYPE_SDB_ALM_EMP: + dev_warn(dev, "SDB almost empty.\n"); + break; + case HNS_ROCE_DB_SUBTYPE_ODB_OVF: + dev_warn(dev, "ODB overflow.\n"); + break; + case HNS_ROCE_DB_SUBTYPE_ODB_ALM_OVF: + dev_warn(dev, "ODB almost overflow.\n"); + break; + case HNS_ROCE_DB_SUBTYPE_ODB_ALM_EMP: + dev_warn(dev, "SDB almost empty.\n"); + break; + default: + break; + } +} + +static struct hns_roce_aeqe *get_aeqe_v1(struct hns_roce_eq *eq, u32 entry) +{ + unsigned long off = (entry & (eq->entries - 1)) * + HNS_ROCE_AEQ_ENTRY_SIZE; + + return (struct hns_roce_aeqe *)((u8 *) + (eq->buf_list[off / HNS_ROCE_BA_SIZE].buf) + + off % HNS_ROCE_BA_SIZE); +} + +static struct hns_roce_aeqe *next_aeqe_sw_v1(struct hns_roce_eq *eq) +{ + struct hns_roce_aeqe *aeqe = get_aeqe_v1(eq, eq->cons_index); + + return (roce_get_bit(aeqe->asyn, HNS_ROCE_AEQE_U32_4_OWNER_S) ^ + !!(eq->cons_index & eq->entries)) ? aeqe : NULL; +} + +static int hns_roce_v1_aeq_int(struct hns_roce_dev *hr_dev, + struct hns_roce_eq *eq) +{ + struct device *dev = &hr_dev->pdev->dev; + struct hns_roce_aeqe *aeqe; + int aeqes_found = 0; + int event_type; + + while ((aeqe = next_aeqe_sw_v1(eq))) { + + /* Make sure we read the AEQ entry after we have checked the + * ownership bit + */ + dma_rmb(); + + dev_dbg(dev, "aeqe = %p, aeqe->asyn.event_type = 0x%lx\n", aeqe, + roce_get_field(aeqe->asyn, + HNS_ROCE_AEQE_U32_4_EVENT_TYPE_M, + HNS_ROCE_AEQE_U32_4_EVENT_TYPE_S)); + event_type = roce_get_field(aeqe->asyn, + HNS_ROCE_AEQE_U32_4_EVENT_TYPE_M, + HNS_ROCE_AEQE_U32_4_EVENT_TYPE_S); + switch (event_type) { + case HNS_ROCE_EVENT_TYPE_PATH_MIG: + dev_warn(dev, "PATH MIG not supported\n"); + break; + case HNS_ROCE_EVENT_TYPE_COMM_EST: + dev_warn(dev, "COMMUNICATION established\n"); + break; + case HNS_ROCE_EVENT_TYPE_SQ_DRAINED: + dev_warn(dev, "SQ DRAINED not supported\n"); + break; + case HNS_ROCE_EVENT_TYPE_PATH_MIG_FAILED: + dev_warn(dev, "PATH MIG failed\n"); + break; + case HNS_ROCE_EVENT_TYPE_INV_REQ_LOCAL_WQ_ERROR: + case HNS_ROCE_EVENT_TYPE_WQ_CATAS_ERROR: + case HNS_ROCE_EVENT_TYPE_LOCAL_WQ_ACCESS_ERROR: + hns_roce_v1_qp_err_handle(hr_dev, aeqe, event_type); + break; + case HNS_ROCE_EVENT_TYPE_SRQ_LIMIT_REACH: + case HNS_ROCE_EVENT_TYPE_SRQ_CATAS_ERROR: + case HNS_ROCE_EVENT_TYPE_SRQ_LAST_WQE_REACH: + dev_warn(dev, "SRQ not support!\n"); + break; + case HNS_ROCE_EVENT_TYPE_CQ_ACCESS_ERROR: + case HNS_ROCE_EVENT_TYPE_CQ_OVERFLOW: + case HNS_ROCE_EVENT_TYPE_CQ_ID_INVALID: + hns_roce_v1_cq_err_handle(hr_dev, aeqe, event_type); + break; + case HNS_ROCE_EVENT_TYPE_PORT_CHANGE: + dev_warn(dev, "port change.\n"); + break; + case HNS_ROCE_EVENT_TYPE_MB: + hns_roce_cmd_event(hr_dev, + le16_to_cpu(aeqe->event.cmd.token), + aeqe->event.cmd.status, + le64_to_cpu(aeqe->event.cmd.out_param + )); + break; + case HNS_ROCE_EVENT_TYPE_DB_OVERFLOW: + hns_roce_v1_db_overflow_handle(hr_dev, aeqe); + break; + case HNS_ROCE_EVENT_TYPE_CEQ_OVERFLOW: + dev_warn(dev, "CEQ 0x%lx overflow.\n", + roce_get_field(aeqe->event.ce_event.ceqe, + HNS_ROCE_AEQE_EVENT_CE_EVENT_CEQE_CEQN_M, + HNS_ROCE_AEQE_EVENT_CE_EVENT_CEQE_CEQN_S)); + break; + default: + dev_warn(dev, "Unhandled event %d on EQ %d at idx %u.\n", + event_type, eq->eqn, eq->cons_index); + break; + } + + eq->cons_index++; + aeqes_found = 1; + + if (eq->cons_index > 2 * hr_dev->caps.aeqe_depth - 1) { + dev_warn(dev, "cons_index overflow, set back to 0.\n"); + eq->cons_index = 0; + } + } + + set_eq_cons_index_v1(eq, 0); + + return aeqes_found; +} + +static struct hns_roce_ceqe *get_ceqe_v1(struct hns_roce_eq *eq, u32 entry) +{ + unsigned long off = (entry & (eq->entries - 1)) * + HNS_ROCE_CEQ_ENTRY_SIZE; + + return (struct hns_roce_ceqe *)((u8 *) + (eq->buf_list[off / HNS_ROCE_BA_SIZE].buf) + + off % HNS_ROCE_BA_SIZE); +} + +static struct hns_roce_ceqe *next_ceqe_sw_v1(struct hns_roce_eq *eq) +{ + struct hns_roce_ceqe *ceqe = get_ceqe_v1(eq, eq->cons_index); + + return (!!(roce_get_bit(ceqe->comp, + HNS_ROCE_CEQE_CEQE_COMP_OWNER_S))) ^ + (!!(eq->cons_index & eq->entries)) ? ceqe : NULL; +} + +static int hns_roce_v1_ceq_int(struct hns_roce_dev *hr_dev, + struct hns_roce_eq *eq) +{ + struct hns_roce_ceqe *ceqe; + int ceqes_found = 0; + u32 cqn; + + while ((ceqe = next_ceqe_sw_v1(eq))) { + + /* Make sure we read CEQ entry after we have checked the + * ownership bit + */ + dma_rmb(); + + cqn = roce_get_field(ceqe->comp, + HNS_ROCE_CEQE_CEQE_COMP_CQN_M, + HNS_ROCE_CEQE_CEQE_COMP_CQN_S); + hns_roce_cq_completion(hr_dev, cqn); + + ++eq->cons_index; + ceqes_found = 1; + + if (eq->cons_index > 2 * hr_dev->caps.ceqe_depth - 1) { + dev_warn(&eq->hr_dev->pdev->dev, + "cons_index overflow, set back to 0.\n"); + eq->cons_index = 0; + } + } + + set_eq_cons_index_v1(eq, 0); + + return ceqes_found; +} + +static irqreturn_t hns_roce_v1_msix_interrupt_eq(int irq, void *eq_ptr) +{ + struct hns_roce_eq *eq = eq_ptr; + struct hns_roce_dev *hr_dev = eq->hr_dev; + int int_work = 0; + + if (eq->type_flag == HNS_ROCE_CEQ) + /* CEQ irq routine, CEQ is pulse irq, not clear */ + int_work = hns_roce_v1_ceq_int(hr_dev, eq); + else + /* AEQ irq routine, AEQ is pulse irq, not clear */ + int_work = hns_roce_v1_aeq_int(hr_dev, eq); + + return IRQ_RETVAL(int_work); +} + +static irqreturn_t hns_roce_v1_msix_interrupt_abn(int irq, void *dev_id) +{ + struct hns_roce_dev *hr_dev = dev_id; + struct device *dev = &hr_dev->pdev->dev; + int int_work = 0; + u32 caepaemask_val; + u32 cealmovf_val; + u32 caepaest_val; + u32 aeshift_val; + u32 ceshift_val; + u32 cemask_val; + int i; + + /* + * Abnormal interrupt: + * AEQ overflow, ECC multi-bit err, CEQ overflow must clear + * interrupt, mask irq, clear irq, cancel mask operation + */ + aeshift_val = roce_read(hr_dev, ROCEE_CAEP_AEQC_AEQE_SHIFT_REG); + + /* AEQE overflow */ + if (roce_get_bit(aeshift_val, + ROCEE_CAEP_AEQC_AEQE_SHIFT_CAEP_AEQ_ALM_OVF_INT_ST_S) == 1) { + dev_warn(dev, "AEQ overflow!\n"); + + /* Set mask */ + caepaemask_val = roce_read(hr_dev, ROCEE_CAEP_AE_MASK_REG); + roce_set_bit(caepaemask_val, + ROCEE_CAEP_AE_MASK_CAEP_AEQ_ALM_OVF_MASK_S, + HNS_ROCE_INT_MASK_ENABLE); + roce_write(hr_dev, ROCEE_CAEP_AE_MASK_REG, caepaemask_val); + + /* Clear int state(INT_WC : write 1 clear) */ + caepaest_val = roce_read(hr_dev, ROCEE_CAEP_AE_ST_REG); + roce_set_bit(caepaest_val, + ROCEE_CAEP_AE_ST_CAEP_AEQ_ALM_OVF_S, 1); + roce_write(hr_dev, ROCEE_CAEP_AE_ST_REG, caepaest_val); + + /* Clear mask */ + caepaemask_val = roce_read(hr_dev, ROCEE_CAEP_AE_MASK_REG); + roce_set_bit(caepaemask_val, + ROCEE_CAEP_AE_MASK_CAEP_AEQ_ALM_OVF_MASK_S, + HNS_ROCE_INT_MASK_DISABLE); + roce_write(hr_dev, ROCEE_CAEP_AE_MASK_REG, caepaemask_val); + } + + /* CEQ almost overflow */ + for (i = 0; i < hr_dev->caps.num_comp_vectors; i++) { + ceshift_val = roce_read(hr_dev, ROCEE_CAEP_CEQC_SHIFT_0_REG + + i * CEQ_REG_OFFSET); + + if (roce_get_bit(ceshift_val, + ROCEE_CAEP_CEQC_SHIFT_CAEP_CEQ_ALM_OVF_INT_ST_S) == 1) { + dev_warn(dev, "CEQ[%d] almost overflow!\n", i); + int_work++; + + /* Set mask */ + cemask_val = roce_read(hr_dev, + ROCEE_CAEP_CE_IRQ_MASK_0_REG + + i * CEQ_REG_OFFSET); + roce_set_bit(cemask_val, + ROCEE_CAEP_CE_IRQ_MASK_CAEP_CEQ_ALM_OVF_MASK_S, + HNS_ROCE_INT_MASK_ENABLE); + roce_write(hr_dev, ROCEE_CAEP_CE_IRQ_MASK_0_REG + + i * CEQ_REG_OFFSET, cemask_val); + + /* Clear int state(INT_WC : write 1 clear) */ + cealmovf_val = roce_read(hr_dev, + ROCEE_CAEP_CEQ_ALM_OVF_0_REG + + i * CEQ_REG_OFFSET); + roce_set_bit(cealmovf_val, + ROCEE_CAEP_CEQ_ALM_OVF_CAEP_CEQ_ALM_OVF_S, + 1); + roce_write(hr_dev, ROCEE_CAEP_CEQ_ALM_OVF_0_REG + + i * CEQ_REG_OFFSET, cealmovf_val); + + /* Clear mask */ + cemask_val = roce_read(hr_dev, + ROCEE_CAEP_CE_IRQ_MASK_0_REG + + i * CEQ_REG_OFFSET); + roce_set_bit(cemask_val, + ROCEE_CAEP_CE_IRQ_MASK_CAEP_CEQ_ALM_OVF_MASK_S, + HNS_ROCE_INT_MASK_DISABLE); + roce_write(hr_dev, ROCEE_CAEP_CE_IRQ_MASK_0_REG + + i * CEQ_REG_OFFSET, cemask_val); + } + } + + /* ECC multi-bit error alarm */ + dev_warn(dev, "ECC UCERR ALARM: 0x%x, 0x%x, 0x%x\n", + roce_read(hr_dev, ROCEE_ECC_UCERR_ALM0_REG), + roce_read(hr_dev, ROCEE_ECC_UCERR_ALM1_REG), + roce_read(hr_dev, ROCEE_ECC_UCERR_ALM2_REG)); + + dev_warn(dev, "ECC CERR ALARM: 0x%x, 0x%x, 0x%x\n", + roce_read(hr_dev, ROCEE_ECC_CERR_ALM0_REG), + roce_read(hr_dev, ROCEE_ECC_CERR_ALM1_REG), + roce_read(hr_dev, ROCEE_ECC_CERR_ALM2_REG)); + + return IRQ_RETVAL(int_work); +} + +static void hns_roce_v1_int_mask_enable(struct hns_roce_dev *hr_dev) +{ + u32 aemask_val; + int masken = 0; + int i; + + /* AEQ INT */ + aemask_val = roce_read(hr_dev, ROCEE_CAEP_AE_MASK_REG); + roce_set_bit(aemask_val, ROCEE_CAEP_AE_MASK_CAEP_AEQ_ALM_OVF_MASK_S, + masken); + roce_set_bit(aemask_val, ROCEE_CAEP_AE_MASK_CAEP_AE_IRQ_MASK_S, masken); + roce_write(hr_dev, ROCEE_CAEP_AE_MASK_REG, aemask_val); + + /* CEQ INT */ + for (i = 0; i < hr_dev->caps.num_comp_vectors; i++) { + /* IRQ mask */ + roce_write(hr_dev, ROCEE_CAEP_CE_IRQ_MASK_0_REG + + i * CEQ_REG_OFFSET, masken); + } +} + +static void hns_roce_v1_free_eq(struct hns_roce_dev *hr_dev, + struct hns_roce_eq *eq) +{ + int npages = (PAGE_ALIGN(eq->eqe_size * eq->entries) + + HNS_ROCE_BA_SIZE - 1) / HNS_ROCE_BA_SIZE; + int i; + + if (!eq->buf_list) + return; + + for (i = 0; i < npages; ++i) + dma_free_coherent(&hr_dev->pdev->dev, HNS_ROCE_BA_SIZE, + eq->buf_list[i].buf, eq->buf_list[i].map); + + kfree(eq->buf_list); +} + +static void hns_roce_v1_enable_eq(struct hns_roce_dev *hr_dev, int eq_num, + int enable_flag) +{ + void __iomem *eqc = hr_dev->eq_table.eqc_base[eq_num]; + u32 val; + + val = readl(eqc); + + if (enable_flag) + roce_set_field(val, + ROCEE_CAEP_AEQC_AEQE_SHIFT_CAEP_AEQC_STATE_M, + ROCEE_CAEP_AEQC_AEQE_SHIFT_CAEP_AEQC_STATE_S, + HNS_ROCE_EQ_STAT_VALID); + else + roce_set_field(val, + ROCEE_CAEP_AEQC_AEQE_SHIFT_CAEP_AEQC_STATE_M, + ROCEE_CAEP_AEQC_AEQE_SHIFT_CAEP_AEQC_STATE_S, + HNS_ROCE_EQ_STAT_INVALID); + writel(val, eqc); +} + +static int hns_roce_v1_create_eq(struct hns_roce_dev *hr_dev, + struct hns_roce_eq *eq) +{ + void __iomem *eqc = hr_dev->eq_table.eqc_base[eq->eqn]; + struct device *dev = &hr_dev->pdev->dev; + dma_addr_t tmp_dma_addr; + u32 eqconsindx_val = 0; + u32 eqcuridx_val = 0; + u32 eqshift_val = 0; + int num_bas; + int ret; + int i; + + num_bas = (PAGE_ALIGN(eq->entries * eq->eqe_size) + + HNS_ROCE_BA_SIZE - 1) / HNS_ROCE_BA_SIZE; + + if ((eq->entries * eq->eqe_size) > HNS_ROCE_BA_SIZE) { + dev_err(dev, "[error]eq buf %d gt ba size(%d) need bas=%d\n", + (eq->entries * eq->eqe_size), HNS_ROCE_BA_SIZE, + num_bas); + return -EINVAL; + } + + eq->buf_list = kcalloc(num_bas, sizeof(*eq->buf_list), GFP_KERNEL); + if (!eq->buf_list) + return -ENOMEM; + + for (i = 0; i < num_bas; ++i) { + eq->buf_list[i].buf = dma_alloc_coherent(dev, HNS_ROCE_BA_SIZE, + &tmp_dma_addr, + GFP_KERNEL); + if (!eq->buf_list[i].buf) { + ret = -ENOMEM; + goto err_out_free_pages; + } + + eq->buf_list[i].map = tmp_dma_addr; + memset(eq->buf_list[i].buf, 0, HNS_ROCE_BA_SIZE); + } + eq->cons_index = 0; + roce_set_field(eqshift_val, + ROCEE_CAEP_AEQC_AEQE_SHIFT_CAEP_AEQC_STATE_M, + ROCEE_CAEP_AEQC_AEQE_SHIFT_CAEP_AEQC_STATE_S, + HNS_ROCE_EQ_STAT_INVALID); + roce_set_field(eqshift_val, + ROCEE_CAEP_AEQC_AEQE_SHIFT_CAEP_AEQC_AEQE_SHIFT_M, + ROCEE_CAEP_AEQC_AEQE_SHIFT_CAEP_AEQC_AEQE_SHIFT_S, + eq->log_entries); + writel(eqshift_val, eqc); + + /* Configure eq extended address 12~44bit */ + writel((u32)(eq->buf_list[0].map >> 12), eqc + 4); + + /* + * Configure eq extended address 45~49 bit. + * 44 = 32 + 12, When evaluating addr to hardware, shift 12 because of + * using 4K page, and shift more 32 because of + * caculating the high 32 bit value evaluated to hardware. + */ + roce_set_field(eqcuridx_val, ROCEE_CAEP_AEQE_CUR_IDX_CAEP_AEQ_BT_H_M, + ROCEE_CAEP_AEQE_CUR_IDX_CAEP_AEQ_BT_H_S, + eq->buf_list[0].map >> 44); + roce_set_field(eqcuridx_val, + ROCEE_CAEP_AEQE_CUR_IDX_CAEP_AEQE_CUR_IDX_M, + ROCEE_CAEP_AEQE_CUR_IDX_CAEP_AEQE_CUR_IDX_S, 0); + writel(eqcuridx_val, eqc + 8); + + /* Configure eq consumer index */ + roce_set_field(eqconsindx_val, + ROCEE_CAEP_AEQE_CONS_IDX_CAEP_AEQE_CONS_IDX_M, + ROCEE_CAEP_AEQE_CONS_IDX_CAEP_AEQE_CONS_IDX_S, 0); + writel(eqconsindx_val, eqc + 0xc); + + return 0; + +err_out_free_pages: + for (i -= 1; i >= 0; i--) + dma_free_coherent(dev, HNS_ROCE_BA_SIZE, eq->buf_list[i].buf, + eq->buf_list[i].map); + + kfree(eq->buf_list); + return ret; +} + +static int hns_roce_v1_init_eq_table(struct hns_roce_dev *hr_dev) +{ + struct hns_roce_eq_table *eq_table = &hr_dev->eq_table; + struct device *dev = &hr_dev->pdev->dev; + struct hns_roce_eq *eq; + int irq_num; + int eq_num; + int ret; + int i, j; + + eq_num = hr_dev->caps.num_comp_vectors + hr_dev->caps.num_aeq_vectors; + irq_num = eq_num + hr_dev->caps.num_other_vectors; + + eq_table->eq = kcalloc(eq_num, sizeof(*eq_table->eq), GFP_KERNEL); + if (!eq_table->eq) + return -ENOMEM; + + eq_table->eqc_base = kcalloc(eq_num, sizeof(*eq_table->eqc_base), + GFP_KERNEL); + if (!eq_table->eqc_base) { + ret = -ENOMEM; + goto err_eqc_base_alloc_fail; + } + + for (i = 0; i < eq_num; i++) { + eq = &eq_table->eq[i]; + eq->hr_dev = hr_dev; + eq->eqn = i; + eq->irq = hr_dev->irq[i]; + eq->log_page_size = PAGE_SHIFT; + + if (i < hr_dev->caps.num_comp_vectors) { + /* CEQ */ + eq_table->eqc_base[i] = hr_dev->reg_base + + ROCEE_CAEP_CEQC_SHIFT_0_REG + + CEQ_REG_OFFSET * i; + eq->type_flag = HNS_ROCE_CEQ; + eq->doorbell = hr_dev->reg_base + + ROCEE_CAEP_CEQC_CONS_IDX_0_REG + + CEQ_REG_OFFSET * i; + eq->entries = hr_dev->caps.ceqe_depth; + eq->log_entries = ilog2(eq->entries); + eq->eqe_size = HNS_ROCE_CEQ_ENTRY_SIZE; + } else { + /* AEQ */ + eq_table->eqc_base[i] = hr_dev->reg_base + + ROCEE_CAEP_AEQC_AEQE_SHIFT_REG; + eq->type_flag = HNS_ROCE_AEQ; + eq->doorbell = hr_dev->reg_base + + ROCEE_CAEP_AEQE_CONS_IDX_REG; + eq->entries = hr_dev->caps.aeqe_depth; + eq->log_entries = ilog2(eq->entries); + eq->eqe_size = HNS_ROCE_AEQ_ENTRY_SIZE; + } + } + + /* Disable irq */ + hns_roce_v1_int_mask_enable(hr_dev); + + /* Configure ce int interval */ + roce_write(hr_dev, ROCEE_CAEP_CE_INTERVAL_CFG_REG, + HNS_ROCE_CEQ_DEFAULT_INTERVAL); + + /* Configure ce int burst num */ + roce_write(hr_dev, ROCEE_CAEP_CE_BURST_NUM_CFG_REG, + HNS_ROCE_CEQ_DEFAULT_BURST_NUM); + + for (i = 0; i < eq_num; i++) { + ret = hns_roce_v1_create_eq(hr_dev, &eq_table->eq[i]); + if (ret) { + dev_err(dev, "eq create failed\n"); + goto err_create_eq_fail; + } + } + + for (j = 0; j < irq_num; j++) { + if (j < eq_num) + ret = request_irq(hr_dev->irq[j], + hns_roce_v1_msix_interrupt_eq, 0, + hr_dev->irq_names[j], + &eq_table->eq[j]); + else + ret = request_irq(hr_dev->irq[j], + hns_roce_v1_msix_interrupt_abn, 0, + hr_dev->irq_names[j], hr_dev); + + if (ret) { + dev_err(dev, "request irq error!\n"); + goto err_request_irq_fail; + } + } + + for (i = 0; i < eq_num; i++) + hns_roce_v1_enable_eq(hr_dev, i, EQ_ENABLE); + + return 0; + +err_request_irq_fail: + for (j -= 1; j >= 0; j--) + free_irq(hr_dev->irq[j], &eq_table->eq[j]); + +err_create_eq_fail: + for (i -= 1; i >= 0; i--) + hns_roce_v1_free_eq(hr_dev, &eq_table->eq[i]); + + kfree(eq_table->eqc_base); + +err_eqc_base_alloc_fail: + kfree(eq_table->eq); + + return ret; +} + +static void hns_roce_v1_cleanup_eq_table(struct hns_roce_dev *hr_dev) +{ + struct hns_roce_eq_table *eq_table = &hr_dev->eq_table; + int irq_num; + int eq_num; + int i; + + eq_num = hr_dev->caps.num_comp_vectors + hr_dev->caps.num_aeq_vectors; + irq_num = eq_num + hr_dev->caps.num_other_vectors; + for (i = 0; i < eq_num; i++) { + /* Disable EQ */ + hns_roce_v1_enable_eq(hr_dev, i, EQ_DISABLE); + + free_irq(hr_dev->irq[i], &eq_table->eq[i]); + + hns_roce_v1_free_eq(hr_dev, &eq_table->eq[i]); + } + for (i = eq_num; i < irq_num; i++) + free_irq(hr_dev->irq[i], hr_dev); + + kfree(eq_table->eqc_base); + kfree(eq_table->eq); +} + static const struct hns_roce_hw hns_roce_hw_v1 = { .reset = hns_roce_v1_reset, .hw_profile = hns_roce_v1_profile, @@ -3983,6 +4708,8 @@ static const struct hns_roce_hw hns_roce_hw_v1 = { .poll_cq = hns_roce_v1_poll_cq, .dereg_mr = hns_roce_v1_dereg_mr, .destroy_cq = hns_roce_v1_destroy_cq, + .init_eq = hns_roce_v1_init_eq_table, + .cleanup_eq = hns_roce_v1_cleanup_eq_table, }; static const struct of_device_id hns_roce_of_match[] = { @@ -4132,14 +4859,14 @@ static int hns_roce_get_cfg(struct hns_roce_dev *hr_dev) /* read the interrupt names from the DT or ACPI */ ret = device_property_read_string_array(dev, "interrupt-names", hr_dev->irq_names, - HNS_ROCE_MAX_IRQ_NUM); + HNS_ROCE_V1_MAX_IRQ_NUM); if (ret < 0) { dev_err(dev, "couldn't get interrupt names from DT or ACPI!\n"); return ret; } /* fetch the interrupt numbers */ - for (i = 0; i < HNS_ROCE_MAX_IRQ_NUM; i++) { + for (i = 0; i < HNS_ROCE_V1_MAX_IRQ_NUM; i++) { hr_dev->irq[i] = platform_get_irq(hr_dev->pdev, i); if (hr_dev->irq[i] <= 0) { dev_err(dev, "platform get of irq[=%d] failed!\n", i); diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.h b/drivers/infiniband/hw/hns/hns_roce_hw_v1.h index 21a07ef0afc9..b44ddd239060 100644 --- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.h +++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.h @@ -60,8 +60,13 @@ #define HNS_ROCE_V1_GID_NUM 16 #define HNS_ROCE_V1_RESV_QP 8 -#define HNS_ROCE_V1_NUM_COMP_EQE 0x8000 -#define HNS_ROCE_V1_NUM_ASYNC_EQE 0x400 +#define HNS_ROCE_V1_MAX_IRQ_NUM 34 +#define HNS_ROCE_V1_COMP_VEC_NUM 32 +#define HNS_ROCE_V1_AEQE_VEC_NUM 1 +#define HNS_ROCE_V1_ABNORMAL_VEC_NUM 1 + +#define HNS_ROCE_V1_COMP_EQE_NUM 0x8000 +#define HNS_ROCE_V1_ASYNC_EQE_NUM 0x400 #define HNS_ROCE_V1_QPC_ENTRY_SIZE 256 #define HNS_ROCE_V1_IRRL_ENTRY_SIZE 8 @@ -159,6 +164,41 @@ #define SDB_INV_CNT_OFFSET 8 #define SDB_ST_CMP_VAL 8 +#define HNS_ROCE_CEQ_DEFAULT_INTERVAL 0x10 +#define HNS_ROCE_CEQ_DEFAULT_BURST_NUM 0x10 + +#define HNS_ROCE_INT_MASK_DISABLE 0 +#define HNS_ROCE_INT_MASK_ENABLE 1 + +#define CEQ_REG_OFFSET 0x18 + +#define HNS_ROCE_CEQE_CEQE_COMP_OWNER_S 0 + +#define HNS_ROCE_V1_CONS_IDX_M GENMASK(15, 0) + +#define HNS_ROCE_CEQE_CEQE_COMP_CQN_S 16 +#define HNS_ROCE_CEQE_CEQE_COMP_CQN_M GENMASK(31, 16) + +#define HNS_ROCE_AEQE_U32_4_EVENT_TYPE_S 16 +#define HNS_ROCE_AEQE_U32_4_EVENT_TYPE_M GENMASK(23, 16) + +#define HNS_ROCE_AEQE_U32_4_EVENT_SUB_TYPE_S 24 +#define HNS_ROCE_AEQE_U32_4_EVENT_SUB_TYPE_M GENMASK(30, 24) + +#define HNS_ROCE_AEQE_U32_4_OWNER_S 31 + +#define HNS_ROCE_AEQE_EVENT_QP_EVENT_QP_QPN_S 0 +#define HNS_ROCE_AEQE_EVENT_QP_EVENT_QP_QPN_M GENMASK(23, 0) + +#define HNS_ROCE_AEQE_EVENT_QP_EVENT_PORT_NUM_S 25 +#define HNS_ROCE_AEQE_EVENT_QP_EVENT_PORT_NUM_M GENMASK(27, 25) + +#define HNS_ROCE_AEQE_EVENT_CQ_EVENT_CQ_CQN_S 0 +#define HNS_ROCE_AEQE_EVENT_CQ_EVENT_CQ_CQN_M GENMASK(15, 0) + +#define HNS_ROCE_AEQE_EVENT_CE_EVENT_CEQE_CEQN_S 0 +#define HNS_ROCE_AEQE_EVENT_CE_EVENT_CEQE_CEQN_M GENMASK(4, 0) + struct hns_roce_cq_context { u32 cqc_byte_4; u32 cq_bt_l; diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c index 8e18445714a9..2ca35e341d09 100644 --- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c +++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c @@ -76,7 +76,8 @@ static int hns_roce_v2_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, return -EOPNOTSUPP; } - if (unlikely(qp->state != IB_QPS_RTS && qp->state != IB_QPS_SQD)) { + if (unlikely(qp->state == IB_QPS_RESET || qp->state == IB_QPS_INIT || + qp->state == IB_QPS_RTR)) { dev_err(dev, "Post WQE fail, QP state %d err!\n", qp->state); *bad_wr = wr; return -EINVAL; @@ -230,26 +231,37 @@ static int hns_roce_v2_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, V2_RC_SEND_WQE_BYTE_4_INLINE_S, 1); } else { if (wr->num_sge <= 2) { - for (i = 0; i < wr->num_sge; i++) - set_data_seg_v2(dseg + i, - wr->sg_list + i); + for (i = 0; i < wr->num_sge; i++) { + if (likely(wr->sg_list[i].length)) { + set_data_seg_v2(dseg, + wr->sg_list + i); + dseg++; + } + } } else { roce_set_field(rc_sq_wqe->byte_20, V2_RC_SEND_WQE_BYTE_20_MSG_START_SGE_IDX_M, V2_RC_SEND_WQE_BYTE_20_MSG_START_SGE_IDX_S, sge_ind & (qp->sge.sge_cnt - 1)); - for (i = 0; i < 2; i++) - set_data_seg_v2(dseg + i, - wr->sg_list + i); + for (i = 0; i < 2; i++) { + if (likely(wr->sg_list[i].length)) { + set_data_seg_v2(dseg, + wr->sg_list + i); + dseg++; + } + } dseg = get_send_extend_sge(qp, sge_ind & (qp->sge.sge_cnt - 1)); for (i = 0; i < wr->num_sge - 2; i++) { - set_data_seg_v2(dseg + i, - wr->sg_list + 2 + i); - sge_ind++; + if (likely(wr->sg_list[i + 2].length)) { + set_data_seg_v2(dseg, + wr->sg_list + 2 + i); + dseg++; + sge_ind++; + } } } @@ -299,6 +311,7 @@ static int hns_roce_v2_post_recv(struct ib_qp *ibqp, struct ib_recv_wr *wr, struct hns_roce_dev *hr_dev = to_hr_dev(ibqp->device); struct hns_roce_qp *hr_qp = to_hr_qp(ibqp); struct hns_roce_v2_wqe_data_seg *dseg; + struct hns_roce_rinl_sge *sge_list; struct device *dev = hr_dev->dev; struct hns_roce_v2_db rq_db; unsigned long flags; @@ -347,6 +360,14 @@ static int hns_roce_v2_post_recv(struct ib_qp *ibqp, struct ib_recv_wr *wr, dseg[i].addr = 0; } + /* rq support inline data */ + sge_list = hr_qp->rq_inl_buf.wqe_list[ind].sg_list; + hr_qp->rq_inl_buf.wqe_list[ind].sge_cnt = (u32)wr->num_sge; + for (i = 0; i < wr->num_sge; i++) { + sge_list[i].addr = (void *)(u64)wr->sg_list[i].addr; + sge_list[i].len = wr->sg_list[i].length; + } + hr_qp->rq.wrid[ind] = wr->wr_id; ind = (ind + 1) & (hr_qp->rq.wqe_cnt - 1); @@ -908,9 +929,9 @@ static int hns_roce_v2_profile(struct hns_roce_dev *hr_dev) caps->max_sq_inline = HNS_ROCE_V2_MAX_SQ_INLINE; caps->num_uars = HNS_ROCE_V2_UAR_NUM; caps->phy_num_uars = HNS_ROCE_V2_PHY_UAR_NUM; - caps->num_aeq_vectors = 1; - caps->num_comp_vectors = 63; - caps->num_other_vectors = 0; + caps->num_aeq_vectors = HNS_ROCE_V2_AEQE_VEC_NUM; + caps->num_comp_vectors = HNS_ROCE_V2_COMP_VEC_NUM; + caps->num_other_vectors = HNS_ROCE_V2_ABNORMAL_VEC_NUM; caps->num_mtpts = HNS_ROCE_V2_MAX_MTPT_NUM; caps->num_mtt_segs = HNS_ROCE_V2_MAX_MTT_SEGS; caps->num_cqe_segs = HNS_ROCE_V2_MAX_CQE_SEGS; @@ -955,12 +976,18 @@ static int hns_roce_v2_profile(struct hns_roce_dev *hr_dev) caps->cqe_ba_pg_sz = 0; caps->cqe_buf_pg_sz = 0; caps->cqe_hop_num = HNS_ROCE_CQE_HOP_NUM; + caps->eqe_ba_pg_sz = 0; + caps->eqe_buf_pg_sz = 0; + caps->eqe_hop_num = HNS_ROCE_EQE_HOP_NUM; caps->chunk_sz = HNS_ROCE_V2_TABLE_CHUNK_SIZE; caps->flags = HNS_ROCE_CAP_FLAG_REREG_MR | - HNS_ROCE_CAP_FLAG_ROCE_V1_V2; + HNS_ROCE_CAP_FLAG_ROCE_V1_V2 | + HNS_ROCE_CAP_FLAG_RQ_INLINE; caps->pkey_table_len[0] = 1; caps->gid_table_len[0] = HNS_ROCE_V2_GID_INDEX_NUM; + caps->ceqe_depth = HNS_ROCE_V2_COMP_EQE_NUM; + caps->aeqe_depth = HNS_ROCE_V2_ASYNC_EQE_NUM; caps->local_ca_ack_delay = 0; caps->max_mtu = IB_MTU_4096; @@ -1382,6 +1409,8 @@ static void hns_roce_v2_write_cqc(struct hns_roce_dev *hr_dev, roce_set_field(cq_context->byte_4_pg_ceqn, V2_CQC_BYTE_4_CQ_ST_M, V2_CQC_BYTE_4_CQ_ST_S, V2_CQ_STATE_VALID); + roce_set_field(cq_context->byte_4_pg_ceqn, V2_CQC_BYTE_4_ARM_ST_M, + V2_CQC_BYTE_4_ARM_ST_S, REG_NXT_CEQE); roce_set_field(cq_context->byte_4_pg_ceqn, V2_CQC_BYTE_4_SHIFT_M, V2_CQC_BYTE_4_SHIFT_S, ilog2((unsigned int)nent)); roce_set_field(cq_context->byte_4_pg_ceqn, V2_CQC_BYTE_4_CEQN_M, @@ -1422,6 +1451,15 @@ static void hns_roce_v2_write_cqc(struct hns_roce_dev *hr_dev, roce_set_field(cq_context->byte_40_cqe_ba, V2_CQC_BYTE_40_CQE_BA_M, V2_CQC_BYTE_40_CQE_BA_S, (dma_handle >> (32 + 3))); + + roce_set_field(cq_context->byte_56_cqe_period_maxcnt, + V2_CQC_BYTE_56_CQ_MAX_CNT_M, + V2_CQC_BYTE_56_CQ_MAX_CNT_S, + HNS_ROCE_V2_CQ_DEFAULT_BURST_NUM); + roce_set_field(cq_context->byte_56_cqe_period_maxcnt, + V2_CQC_BYTE_56_CQ_PERIOD_M, + V2_CQC_BYTE_56_CQ_PERIOD_S, + HNS_ROCE_V2_CQ_DEFAULT_INTERVAL); } static int hns_roce_v2_req_notify_cq(struct ib_cq *ibcq, @@ -1457,6 +1495,40 @@ static int hns_roce_v2_req_notify_cq(struct ib_cq *ibcq, return 0; } +static int hns_roce_handle_recv_inl_wqe(struct hns_roce_v2_cqe *cqe, + struct hns_roce_qp **cur_qp, + struct ib_wc *wc) +{ + struct hns_roce_rinl_sge *sge_list; + u32 wr_num, wr_cnt, sge_num; + u32 sge_cnt, data_len, size; + void *wqe_buf; + + wr_num = roce_get_field(cqe->byte_4, V2_CQE_BYTE_4_WQE_INDX_M, + V2_CQE_BYTE_4_WQE_INDX_S) & 0xffff; + wr_cnt = wr_num & ((*cur_qp)->rq.wqe_cnt - 1); + + sge_list = (*cur_qp)->rq_inl_buf.wqe_list[wr_cnt].sg_list; + sge_num = (*cur_qp)->rq_inl_buf.wqe_list[wr_cnt].sge_cnt; + wqe_buf = get_recv_wqe(*cur_qp, wr_cnt); + data_len = wc->byte_len; + + for (sge_cnt = 0; (sge_cnt < sge_num) && (data_len); sge_cnt++) { + size = min(sge_list[sge_cnt].len, data_len); + memcpy((void *)sge_list[sge_cnt].addr, wqe_buf, size); + + data_len -= size; + wqe_buf += size; + } + + if (data_len) { + wc->status = IB_WC_LOC_LEN_ERR; + return -EAGAIN; + } + + return 0; +} + static int hns_roce_v2_poll_one(struct hns_roce_cq *hr_cq, struct hns_roce_qp **cur_qp, struct ib_wc *wc) { @@ -1469,6 +1541,7 @@ static int hns_roce_v2_poll_one(struct hns_roce_cq *hr_cq, u32 opcode; u32 status; int qpn; + int ret; /* Find cqe according to consumer index */ cqe = next_cqe_sw_v2(hr_cq); @@ -1657,6 +1730,17 @@ static int hns_roce_v2_poll_one(struct hns_roce_cq *hr_cq, break; } + if ((wc->qp->qp_type == IB_QPT_RC || + wc->qp->qp_type == IB_QPT_UC) && + (opcode == HNS_ROCE_V2_OPCODE_SEND || + opcode == HNS_ROCE_V2_OPCODE_SEND_WITH_IMM || + opcode == HNS_ROCE_V2_OPCODE_SEND_WITH_INV) && + (roce_get_bit(cqe->byte_4, V2_CQE_BYTE_4_RQ_INLINE_S))) { + ret = hns_roce_handle_recv_inl_wqe(cqe, cur_qp, wc); + if (ret) + return -EAGAIN; + } + /* Update tail pointer, record wr_id */ wq = &(*cur_qp)->rq; wc->wr_id = wq->wrid[wq->tail & (wq->wqe_cnt - 1)]; @@ -1859,6 +1943,36 @@ static int hns_roce_v2_qp_modify(struct hns_roce_dev *hr_dev, return ret; } +static void set_access_flags(struct hns_roce_qp *hr_qp, + struct hns_roce_v2_qp_context *context, + struct hns_roce_v2_qp_context *qpc_mask, + const struct ib_qp_attr *attr, int attr_mask) +{ + u8 dest_rd_atomic; + u32 access_flags; + + dest_rd_atomic = !!(attr_mask & IB_QP_MAX_DEST_RD_ATOMIC) ? + attr->max_dest_rd_atomic : hr_qp->resp_depth; + + access_flags = !!(attr_mask & IB_QP_ACCESS_FLAGS) ? + attr->qp_access_flags : hr_qp->atomic_rd_en; + + if (!dest_rd_atomic) + access_flags &= IB_ACCESS_REMOTE_WRITE; + + roce_set_bit(context->byte_76_srqn_op_en, V2_QPC_BYTE_76_RRE_S, + !!(access_flags & IB_ACCESS_REMOTE_READ)); + roce_set_bit(qpc_mask->byte_76_srqn_op_en, V2_QPC_BYTE_76_RRE_S, 0); + + roce_set_bit(context->byte_76_srqn_op_en, V2_QPC_BYTE_76_RWE_S, + !!(access_flags & IB_ACCESS_REMOTE_WRITE)); + roce_set_bit(qpc_mask->byte_76_srqn_op_en, V2_QPC_BYTE_76_RWE_S, 0); + + roce_set_bit(context->byte_76_srqn_op_en, V2_QPC_BYTE_76_ATE_S, + !!(access_flags & IB_ACCESS_REMOTE_ATOMIC)); + roce_set_bit(qpc_mask->byte_76_srqn_op_en, V2_QPC_BYTE_76_ATE_S, 0); +} + static void modify_qp_reset_to_init(struct ib_qp *ibqp, const struct ib_qp_attr *attr, struct hns_roce_v2_qp_context *context, @@ -1944,18 +2058,7 @@ static void modify_qp_reset_to_init(struct ib_qp *ibqp, roce_set_bit(qpc_mask->byte_28_at_fl, V2_QPC_BYTE_28_CNP_TX_FLAG_S, 0); roce_set_bit(qpc_mask->byte_28_at_fl, V2_QPC_BYTE_28_CE_FLAG_S, 0); - roce_set_bit(context->byte_76_srqn_op_en, V2_QPC_BYTE_76_RRE_S, - !!(attr->qp_access_flags & IB_ACCESS_REMOTE_READ)); - roce_set_bit(qpc_mask->byte_76_srqn_op_en, V2_QPC_BYTE_76_RRE_S, 0); - - roce_set_bit(context->byte_76_srqn_op_en, V2_QPC_BYTE_76_RWE_S, - !!(attr->qp_access_flags & IB_ACCESS_REMOTE_WRITE)); - roce_set_bit(qpc_mask->byte_76_srqn_op_en, V2_QPC_BYTE_76_RWE_S, 0); - - roce_set_bit(context->byte_76_srqn_op_en, V2_QPC_BYTE_76_ATE_S, - !!(attr->qp_access_flags & IB_ACCESS_REMOTE_ATOMIC)); - roce_set_bit(qpc_mask->byte_76_srqn_op_en, V2_QPC_BYTE_76_ATE_S, 0); - + roce_set_bit(context->byte_76_srqn_op_en, V2_QPC_BYTE_76_RQIE_S, 1); roce_set_bit(qpc_mask->byte_76_srqn_op_en, V2_QPC_BYTE_76_RQIE_S, 0); roce_set_field(context->byte_80_rnr_rx_cqn, V2_QPC_BYTE_80_RX_CQN_M, @@ -2463,11 +2566,14 @@ static int modify_qp_init_to_rtr(struct ib_qp *ibqp, roce_set_bit(qpc_mask->byte_28_at_fl, V2_QPC_BYTE_28_LBI_S, 0); } - roce_set_field(context->byte_140_raq, V2_QPC_BYTE_140_RR_MAX_M, - V2_QPC_BYTE_140_RR_MAX_S, - ilog2((unsigned int)attr->max_dest_rd_atomic)); - roce_set_field(qpc_mask->byte_140_raq, V2_QPC_BYTE_140_RR_MAX_M, - V2_QPC_BYTE_140_RR_MAX_S, 0); + if ((attr_mask & IB_QP_MAX_DEST_RD_ATOMIC) && + attr->max_dest_rd_atomic) { + roce_set_field(context->byte_140_raq, V2_QPC_BYTE_140_RR_MAX_M, + V2_QPC_BYTE_140_RR_MAX_S, + fls(attr->max_dest_rd_atomic - 1)); + roce_set_field(qpc_mask->byte_140_raq, V2_QPC_BYTE_140_RR_MAX_M, + V2_QPC_BYTE_140_RR_MAX_S, 0); + } roce_set_field(context->byte_56_dqpn_err, V2_QPC_BYTE_56_DQPN_M, V2_QPC_BYTE_56_DQPN_S, attr->dest_qp_num); @@ -2557,12 +2663,6 @@ static int modify_qp_init_to_rtr(struct ib_qp *ibqp, V2_QPC_BYTE_168_LP_SGEN_INI_M, V2_QPC_BYTE_168_LP_SGEN_INI_S, 0); - roce_set_field(context->byte_208_irrl, V2_QPC_BYTE_208_SR_MAX_M, - V2_QPC_BYTE_208_SR_MAX_S, - ilog2((unsigned int)attr->max_rd_atomic)); - roce_set_field(qpc_mask->byte_208_irrl, V2_QPC_BYTE_208_SR_MAX_M, - V2_QPC_BYTE_208_SR_MAX_S, 0); - roce_set_field(context->byte_28_at_fl, V2_QPC_BYTE_28_SL_M, V2_QPC_BYTE_28_SL_S, rdma_ah_get_sl(&attr->ah_attr)); roce_set_field(qpc_mask->byte_28_at_fl, V2_QPC_BYTE_28_SL_M, @@ -2766,6 +2866,14 @@ static int modify_qp_rtr_to_rts(struct ib_qp *ibqp, roce_set_field(qpc_mask->byte_196_sq_psn, V2_QPC_BYTE_196_SQ_MAX_PSN_M, V2_QPC_BYTE_196_SQ_MAX_PSN_S, 0); + if ((attr_mask & IB_QP_MAX_QP_RD_ATOMIC) && attr->max_rd_atomic) { + roce_set_field(context->byte_208_irrl, V2_QPC_BYTE_208_SR_MAX_M, + V2_QPC_BYTE_208_SR_MAX_S, + fls(attr->max_rd_atomic - 1)); + roce_set_field(qpc_mask->byte_208_irrl, + V2_QPC_BYTE_208_SR_MAX_M, + V2_QPC_BYTE_208_SR_MAX_S, 0); + } return 0; } @@ -2829,6 +2937,9 @@ static int hns_roce_v2_modify_qp(struct ib_qp *ibqp, goto out; } + if (attr_mask & (IB_QP_ACCESS_FLAGS | IB_QP_MAX_DEST_RD_ATOMIC)) + set_access_flags(hr_qp, context, qpc_mask, attr, attr_mask); + /* Every status migrate must change state */ roce_set_field(context->byte_60_qpst_mapid, V2_QPC_BYTE_60_QP_ST_M, V2_QPC_BYTE_60_QP_ST_S, new_state); @@ -2845,6 +2956,9 @@ static int hns_roce_v2_modify_qp(struct ib_qp *ibqp, hr_qp->state = new_state; + if (attr_mask & IB_QP_ACCESS_FLAGS) + hr_qp->atomic_rd_en = attr->qp_access_flags; + if (attr_mask & IB_QP_MAX_DEST_RD_ATOMIC) hr_qp->resp_depth = attr->max_dest_rd_atomic; if (attr_mask & IB_QP_PORT) { @@ -3098,6 +3212,11 @@ static int hns_roce_v2_destroy_qp_common(struct hns_roce_dev *hr_dev, hns_roce_buf_free(hr_dev, hr_qp->buff_size, &hr_qp->hr_buf); } + if (hr_dev->caps.flags & HNS_ROCE_CAP_FLAG_RQ_INLINE) { + kfree(hr_qp->rq_inl_buf.wqe_list[0].sg_list); + kfree(hr_qp->rq_inl_buf.wqe_list); + } + return 0; } @@ -3162,6 +3281,1146 @@ static int hns_roce_v2_modify_cq(struct ib_cq *cq, u16 cq_count, u16 cq_period) return ret; } +static void set_eq_cons_index_v2(struct hns_roce_eq *eq) +{ + u32 doorbell[2]; + + doorbell[0] = 0; + doorbell[1] = 0; + + if (eq->type_flag == HNS_ROCE_AEQ) { + roce_set_field(doorbell[0], HNS_ROCE_V2_EQ_DB_CMD_M, + HNS_ROCE_V2_EQ_DB_CMD_S, + eq->arm_st == HNS_ROCE_V2_EQ_ALWAYS_ARMED ? + HNS_ROCE_EQ_DB_CMD_AEQ : + HNS_ROCE_EQ_DB_CMD_AEQ_ARMED); + } else { + roce_set_field(doorbell[0], HNS_ROCE_V2_EQ_DB_TAG_M, + HNS_ROCE_V2_EQ_DB_TAG_S, eq->eqn); + + roce_set_field(doorbell[0], HNS_ROCE_V2_EQ_DB_CMD_M, + HNS_ROCE_V2_EQ_DB_CMD_S, + eq->arm_st == HNS_ROCE_V2_EQ_ALWAYS_ARMED ? + HNS_ROCE_EQ_DB_CMD_CEQ : + HNS_ROCE_EQ_DB_CMD_CEQ_ARMED); + } + + roce_set_field(doorbell[1], HNS_ROCE_V2_EQ_DB_PARA_M, + HNS_ROCE_V2_EQ_DB_PARA_S, + (eq->cons_index & HNS_ROCE_V2_CONS_IDX_M)); + + hns_roce_write64_k(doorbell, eq->doorbell); +} + +static void hns_roce_v2_wq_catas_err_handle(struct hns_roce_dev *hr_dev, + struct hns_roce_aeqe *aeqe, + u32 qpn) +{ + struct device *dev = hr_dev->dev; + int sub_type; + + dev_warn(dev, "Local work queue catastrophic error.\n"); + sub_type = roce_get_field(aeqe->asyn, HNS_ROCE_V2_AEQE_SUB_TYPE_M, + HNS_ROCE_V2_AEQE_SUB_TYPE_S); + switch (sub_type) { + case HNS_ROCE_LWQCE_QPC_ERROR: + dev_warn(dev, "QP %d, QPC error.\n", qpn); + break; + case HNS_ROCE_LWQCE_MTU_ERROR: + dev_warn(dev, "QP %d, MTU error.\n", qpn); + break; + case HNS_ROCE_LWQCE_WQE_BA_ADDR_ERROR: + dev_warn(dev, "QP %d, WQE BA addr error.\n", qpn); + break; + case HNS_ROCE_LWQCE_WQE_ADDR_ERROR: + dev_warn(dev, "QP %d, WQE addr error.\n", qpn); + break; + case HNS_ROCE_LWQCE_SQ_WQE_SHIFT_ERROR: + dev_warn(dev, "QP %d, WQE shift error.\n", qpn); + break; + default: + dev_err(dev, "Unhandled sub_event type %d.\n", sub_type); + break; + } +} + +static void hns_roce_v2_local_wq_access_err_handle(struct hns_roce_dev *hr_dev, + struct hns_roce_aeqe *aeqe, u32 qpn) +{ + struct device *dev = hr_dev->dev; + int sub_type; + + dev_warn(dev, "Local access violation work queue error.\n"); + sub_type = roce_get_field(aeqe->asyn, HNS_ROCE_V2_AEQE_SUB_TYPE_M, + HNS_ROCE_V2_AEQE_SUB_TYPE_S); + switch (sub_type) { + case HNS_ROCE_LAVWQE_R_KEY_VIOLATION: + dev_warn(dev, "QP %d, R_key violation.\n", qpn); + break; + case HNS_ROCE_LAVWQE_LENGTH_ERROR: + dev_warn(dev, "QP %d, length error.\n", qpn); + break; + case HNS_ROCE_LAVWQE_VA_ERROR: + dev_warn(dev, "QP %d, VA error.\n", qpn); + break; + case HNS_ROCE_LAVWQE_PD_ERROR: + dev_err(dev, "QP %d, PD error.\n", qpn); + break; + case HNS_ROCE_LAVWQE_RW_ACC_ERROR: + dev_warn(dev, "QP %d, rw acc error.\n", qpn); + break; + case HNS_ROCE_LAVWQE_KEY_STATE_ERROR: + dev_warn(dev, "QP %d, key state error.\n", qpn); + break; + case HNS_ROCE_LAVWQE_MR_OPERATION_ERROR: + dev_warn(dev, "QP %d, MR operation error.\n", qpn); + break; + default: + dev_err(dev, "Unhandled sub_event type %d.\n", sub_type); + break; + } +} + +static void hns_roce_v2_qp_err_handle(struct hns_roce_dev *hr_dev, + struct hns_roce_aeqe *aeqe, + int event_type) +{ + struct device *dev = hr_dev->dev; + u32 qpn; + + qpn = roce_get_field(aeqe->event.qp_event.qp, + HNS_ROCE_V2_AEQE_EVENT_QUEUE_NUM_M, + HNS_ROCE_V2_AEQE_EVENT_QUEUE_NUM_S); + + switch (event_type) { + case HNS_ROCE_EVENT_TYPE_COMM_EST: + dev_warn(dev, "Communication established.\n"); + break; + case HNS_ROCE_EVENT_TYPE_SQ_DRAINED: + dev_warn(dev, "Send queue drained.\n"); + break; + case HNS_ROCE_EVENT_TYPE_WQ_CATAS_ERROR: + hns_roce_v2_wq_catas_err_handle(hr_dev, aeqe, qpn); + break; + case HNS_ROCE_EVENT_TYPE_INV_REQ_LOCAL_WQ_ERROR: + dev_warn(dev, "Invalid request local work queue error.\n"); + break; + case HNS_ROCE_EVENT_TYPE_LOCAL_WQ_ACCESS_ERROR: + hns_roce_v2_local_wq_access_err_handle(hr_dev, aeqe, qpn); + break; + default: + break; + } + + hns_roce_qp_event(hr_dev, qpn, event_type); +} + +static void hns_roce_v2_cq_err_handle(struct hns_roce_dev *hr_dev, + struct hns_roce_aeqe *aeqe, + int event_type) +{ + struct device *dev = hr_dev->dev; + u32 cqn; + + cqn = roce_get_field(aeqe->event.cq_event.cq, + HNS_ROCE_V2_AEQE_EVENT_QUEUE_NUM_M, + HNS_ROCE_V2_AEQE_EVENT_QUEUE_NUM_S); + + switch (event_type) { + case HNS_ROCE_EVENT_TYPE_CQ_ACCESS_ERROR: + dev_warn(dev, "CQ 0x%x access err.\n", cqn); + break; + case HNS_ROCE_EVENT_TYPE_CQ_OVERFLOW: + dev_warn(dev, "CQ 0x%x overflow\n", cqn); + break; + default: + break; + } + + hns_roce_cq_event(hr_dev, cqn, event_type); +} + +static struct hns_roce_aeqe *get_aeqe_v2(struct hns_roce_eq *eq, u32 entry) +{ + u32 buf_chk_sz; + unsigned long off; + + buf_chk_sz = 1 << (eq->eqe_buf_pg_sz + PAGE_SHIFT); + off = (entry & (eq->entries - 1)) * HNS_ROCE_AEQ_ENTRY_SIZE; + + return (struct hns_roce_aeqe *)((char *)(eq->buf_list->buf) + + off % buf_chk_sz); +} + +static struct hns_roce_aeqe *mhop_get_aeqe(struct hns_roce_eq *eq, u32 entry) +{ + u32 buf_chk_sz; + unsigned long off; + + buf_chk_sz = 1 << (eq->eqe_buf_pg_sz + PAGE_SHIFT); + + off = (entry & (eq->entries - 1)) * HNS_ROCE_AEQ_ENTRY_SIZE; + + if (eq->hop_num == HNS_ROCE_HOP_NUM_0) + return (struct hns_roce_aeqe *)((u8 *)(eq->bt_l0) + + off % buf_chk_sz); + else + return (struct hns_roce_aeqe *)((u8 *) + (eq->buf[off / buf_chk_sz]) + off % buf_chk_sz); +} + +static struct hns_roce_aeqe *next_aeqe_sw_v2(struct hns_roce_eq *eq) +{ + struct hns_roce_aeqe *aeqe; + + if (!eq->hop_num) + aeqe = get_aeqe_v2(eq, eq->cons_index); + else + aeqe = mhop_get_aeqe(eq, eq->cons_index); + + return (roce_get_bit(aeqe->asyn, HNS_ROCE_V2_AEQ_AEQE_OWNER_S) ^ + !!(eq->cons_index & eq->entries)) ? aeqe : NULL; +} + +static int hns_roce_v2_aeq_int(struct hns_roce_dev *hr_dev, + struct hns_roce_eq *eq) +{ + struct device *dev = hr_dev->dev; + struct hns_roce_aeqe *aeqe; + int aeqe_found = 0; + int event_type; + + while ((aeqe = next_aeqe_sw_v2(eq))) { + + /* Make sure we read AEQ entry after we have checked the + * ownership bit + */ + dma_rmb(); + + event_type = roce_get_field(aeqe->asyn, + HNS_ROCE_V2_AEQE_EVENT_TYPE_M, + HNS_ROCE_V2_AEQE_EVENT_TYPE_S); + + switch (event_type) { + case HNS_ROCE_EVENT_TYPE_PATH_MIG: + dev_warn(dev, "Path migrated succeeded.\n"); + break; + case HNS_ROCE_EVENT_TYPE_PATH_MIG_FAILED: + dev_warn(dev, "Path migration failed.\n"); + break; + case HNS_ROCE_EVENT_TYPE_COMM_EST: + case HNS_ROCE_EVENT_TYPE_SQ_DRAINED: + case HNS_ROCE_EVENT_TYPE_WQ_CATAS_ERROR: + case HNS_ROCE_EVENT_TYPE_INV_REQ_LOCAL_WQ_ERROR: + case HNS_ROCE_EVENT_TYPE_LOCAL_WQ_ACCESS_ERROR: + hns_roce_v2_qp_err_handle(hr_dev, aeqe, event_type); + break; + case HNS_ROCE_EVENT_TYPE_SRQ_LIMIT_REACH: + case HNS_ROCE_EVENT_TYPE_SRQ_LAST_WQE_REACH: + case HNS_ROCE_EVENT_TYPE_SRQ_CATAS_ERROR: + dev_warn(dev, "SRQ not support.\n"); + break; + case HNS_ROCE_EVENT_TYPE_CQ_ACCESS_ERROR: + case HNS_ROCE_EVENT_TYPE_CQ_OVERFLOW: + hns_roce_v2_cq_err_handle(hr_dev, aeqe, event_type); + break; + case HNS_ROCE_EVENT_TYPE_DB_OVERFLOW: + dev_warn(dev, "DB overflow.\n"); + break; + case HNS_ROCE_EVENT_TYPE_MB: + hns_roce_cmd_event(hr_dev, + le16_to_cpu(aeqe->event.cmd.token), + aeqe->event.cmd.status, + le64_to_cpu(aeqe->event.cmd.out_param)); + break; + case HNS_ROCE_EVENT_TYPE_CEQ_OVERFLOW: + dev_warn(dev, "CEQ overflow.\n"); + break; + case HNS_ROCE_EVENT_TYPE_FLR: + dev_warn(dev, "Function level reset.\n"); + break; + default: + dev_err(dev, "Unhandled event %d on EQ %d at idx %u.\n", + event_type, eq->eqn, eq->cons_index); + break; + }; + + ++eq->cons_index; + aeqe_found = 1; + + if (eq->cons_index > (2 * eq->entries - 1)) { + dev_warn(dev, "cons_index overflow, set back to 0.\n"); + eq->cons_index = 0; + } + } + + set_eq_cons_index_v2(eq); + return aeqe_found; +} + +static struct hns_roce_ceqe *get_ceqe_v2(struct hns_roce_eq *eq, u32 entry) +{ + u32 buf_chk_sz; + unsigned long off; + + buf_chk_sz = 1 << (eq->eqe_buf_pg_sz + PAGE_SHIFT); + off = (entry & (eq->entries - 1)) * HNS_ROCE_CEQ_ENTRY_SIZE; + + return (struct hns_roce_ceqe *)((char *)(eq->buf_list->buf) + + off % buf_chk_sz); +} + +static struct hns_roce_ceqe *mhop_get_ceqe(struct hns_roce_eq *eq, u32 entry) +{ + u32 buf_chk_sz; + unsigned long off; + + buf_chk_sz = 1 << (eq->eqe_buf_pg_sz + PAGE_SHIFT); + + off = (entry & (eq->entries - 1)) * HNS_ROCE_CEQ_ENTRY_SIZE; + + if (eq->hop_num == HNS_ROCE_HOP_NUM_0) + return (struct hns_roce_ceqe *)((u8 *)(eq->bt_l0) + + off % buf_chk_sz); + else + return (struct hns_roce_ceqe *)((u8 *)(eq->buf[off / + buf_chk_sz]) + off % buf_chk_sz); +} + +static struct hns_roce_ceqe *next_ceqe_sw_v2(struct hns_roce_eq *eq) +{ + struct hns_roce_ceqe *ceqe; + + if (!eq->hop_num) + ceqe = get_ceqe_v2(eq, eq->cons_index); + else + ceqe = mhop_get_ceqe(eq, eq->cons_index); + + return (!!(roce_get_bit(ceqe->comp, HNS_ROCE_V2_CEQ_CEQE_OWNER_S))) ^ + (!!(eq->cons_index & eq->entries)) ? ceqe : NULL; +} + +static int hns_roce_v2_ceq_int(struct hns_roce_dev *hr_dev, + struct hns_roce_eq *eq) +{ + struct device *dev = hr_dev->dev; + struct hns_roce_ceqe *ceqe; + int ceqe_found = 0; + u32 cqn; + + while ((ceqe = next_ceqe_sw_v2(eq))) { + + /* Make sure we read CEQ entry after we have checked the + * ownership bit + */ + dma_rmb(); + + cqn = roce_get_field(ceqe->comp, + HNS_ROCE_V2_CEQE_COMP_CQN_M, + HNS_ROCE_V2_CEQE_COMP_CQN_S); + + hns_roce_cq_completion(hr_dev, cqn); + + ++eq->cons_index; + ceqe_found = 1; + + if (eq->cons_index > (2 * eq->entries - 1)) { + dev_warn(dev, "cons_index overflow, set back to 0.\n"); + eq->cons_index = 0; + } + } + + set_eq_cons_index_v2(eq); + + return ceqe_found; +} + +static irqreturn_t hns_roce_v2_msix_interrupt_eq(int irq, void *eq_ptr) +{ + struct hns_roce_eq *eq = eq_ptr; + struct hns_roce_dev *hr_dev = eq->hr_dev; + int int_work = 0; + + if (eq->type_flag == HNS_ROCE_CEQ) + /* Completion event interrupt */ + int_work = hns_roce_v2_ceq_int(hr_dev, eq); + else + /* Asychronous event interrupt */ + int_work = hns_roce_v2_aeq_int(hr_dev, eq); + + return IRQ_RETVAL(int_work); +} + +static irqreturn_t hns_roce_v2_msix_interrupt_abn(int irq, void *dev_id) +{ + struct hns_roce_dev *hr_dev = dev_id; + struct device *dev = hr_dev->dev; + int int_work = 0; + u32 int_st; + u32 int_en; + + /* Abnormal interrupt */ + int_st = roce_read(hr_dev, ROCEE_VF_ABN_INT_ST_REG); + int_en = roce_read(hr_dev, ROCEE_VF_ABN_INT_EN_REG); + + if (roce_get_bit(int_st, HNS_ROCE_V2_VF_INT_ST_AEQ_OVERFLOW_S)) { + dev_err(dev, "AEQ overflow!\n"); + + roce_set_bit(int_st, HNS_ROCE_V2_VF_INT_ST_AEQ_OVERFLOW_S, 1); + roce_write(hr_dev, ROCEE_VF_ABN_INT_ST_REG, int_st); + + roce_set_bit(int_en, HNS_ROCE_V2_VF_ABN_INT_EN_S, 1); + roce_write(hr_dev, ROCEE_VF_ABN_INT_EN_REG, int_en); + + int_work = 1; + } else if (roce_get_bit(int_st, HNS_ROCE_V2_VF_INT_ST_BUS_ERR_S)) { + dev_err(dev, "BUS ERR!\n"); + + roce_set_bit(int_st, HNS_ROCE_V2_VF_INT_ST_BUS_ERR_S, 1); + roce_write(hr_dev, ROCEE_VF_ABN_INT_ST_REG, int_st); + + roce_set_bit(int_en, HNS_ROCE_V2_VF_ABN_INT_EN_S, 1); + roce_write(hr_dev, ROCEE_VF_ABN_INT_EN_REG, int_en); + + int_work = 1; + } else if (roce_get_bit(int_st, HNS_ROCE_V2_VF_INT_ST_OTHER_ERR_S)) { + dev_err(dev, "OTHER ERR!\n"); + + roce_set_bit(int_st, HNS_ROCE_V2_VF_INT_ST_OTHER_ERR_S, 1); + roce_write(hr_dev, ROCEE_VF_ABN_INT_ST_REG, int_st); + + roce_set_bit(int_en, HNS_ROCE_V2_VF_ABN_INT_EN_S, 1); + roce_write(hr_dev, ROCEE_VF_ABN_INT_EN_REG, int_en); + + int_work = 1; + } else + dev_err(dev, "There is no abnormal irq found!\n"); + + return IRQ_RETVAL(int_work); +} + +static void hns_roce_v2_int_mask_enable(struct hns_roce_dev *hr_dev, + int eq_num, int enable_flag) +{ + int i; + + if (enable_flag == EQ_ENABLE) { + for (i = 0; i < eq_num; i++) + roce_write(hr_dev, ROCEE_VF_EVENT_INT_EN_REG + + i * EQ_REG_OFFSET, + HNS_ROCE_V2_VF_EVENT_INT_EN_M); + + roce_write(hr_dev, ROCEE_VF_ABN_INT_EN_REG, + HNS_ROCE_V2_VF_ABN_INT_EN_M); + roce_write(hr_dev, ROCEE_VF_ABN_INT_CFG_REG, + HNS_ROCE_V2_VF_ABN_INT_CFG_M); + } else { + for (i = 0; i < eq_num; i++) + roce_write(hr_dev, ROCEE_VF_EVENT_INT_EN_REG + + i * EQ_REG_OFFSET, + HNS_ROCE_V2_VF_EVENT_INT_EN_M & 0x0); + + roce_write(hr_dev, ROCEE_VF_ABN_INT_EN_REG, + HNS_ROCE_V2_VF_ABN_INT_EN_M & 0x0); + roce_write(hr_dev, ROCEE_VF_ABN_INT_CFG_REG, + HNS_ROCE_V2_VF_ABN_INT_CFG_M & 0x0); + } +} + +static void hns_roce_v2_destroy_eqc(struct hns_roce_dev *hr_dev, int eqn) +{ + struct device *dev = hr_dev->dev; + int ret; + + if (eqn < hr_dev->caps.num_comp_vectors) + ret = hns_roce_cmd_mbox(hr_dev, 0, 0, eqn & HNS_ROCE_V2_EQN_M, + 0, HNS_ROCE_CMD_DESTROY_CEQC, + HNS_ROCE_CMD_TIMEOUT_MSECS); + else + ret = hns_roce_cmd_mbox(hr_dev, 0, 0, eqn & HNS_ROCE_V2_EQN_M, + 0, HNS_ROCE_CMD_DESTROY_AEQC, + HNS_ROCE_CMD_TIMEOUT_MSECS); + if (ret) + dev_err(dev, "[mailbox cmd] destroy eqc(%d) failed.\n", eqn); +} + +static void hns_roce_mhop_free_eq(struct hns_roce_dev *hr_dev, + struct hns_roce_eq *eq) +{ + struct device *dev = hr_dev->dev; + u64 idx; + u64 size; + u32 buf_chk_sz; + u32 bt_chk_sz; + u32 mhop_num; + int eqe_alloc; + int ba_num; + int i = 0; + int j = 0; + + mhop_num = hr_dev->caps.eqe_hop_num; + buf_chk_sz = 1 << (hr_dev->caps.eqe_buf_pg_sz + PAGE_SHIFT); + bt_chk_sz = 1 << (hr_dev->caps.eqe_ba_pg_sz + PAGE_SHIFT); + ba_num = (PAGE_ALIGN(eq->entries * eq->eqe_size) + buf_chk_sz - 1) / + buf_chk_sz; + + /* hop_num = 0 */ + if (mhop_num == HNS_ROCE_HOP_NUM_0) { + dma_free_coherent(dev, (unsigned int)(eq->entries * + eq->eqe_size), eq->bt_l0, eq->l0_dma); + return; + } + + /* hop_num = 1 or hop = 2 */ + dma_free_coherent(dev, bt_chk_sz, eq->bt_l0, eq->l0_dma); + if (mhop_num == 1) { + for (i = 0; i < eq->l0_last_num; i++) { + if (i == eq->l0_last_num - 1) { + eqe_alloc = i * (buf_chk_sz / eq->eqe_size); + size = (eq->entries - eqe_alloc) * eq->eqe_size; + dma_free_coherent(dev, size, eq->buf[i], + eq->buf_dma[i]); + break; + } + dma_free_coherent(dev, buf_chk_sz, eq->buf[i], + eq->buf_dma[i]); + } + } else if (mhop_num == 2) { + for (i = 0; i < eq->l0_last_num; i++) { + dma_free_coherent(dev, bt_chk_sz, eq->bt_l1[i], + eq->l1_dma[i]); + + for (j = 0; j < bt_chk_sz / 8; j++) { + idx = i * (bt_chk_sz / 8) + j; + if ((i == eq->l0_last_num - 1) + && j == eq->l1_last_num - 1) { + eqe_alloc = (buf_chk_sz / eq->eqe_size) + * idx; + size = (eq->entries - eqe_alloc) + * eq->eqe_size; + dma_free_coherent(dev, size, + eq->buf[idx], + eq->buf_dma[idx]); + break; + } + dma_free_coherent(dev, buf_chk_sz, eq->buf[idx], + eq->buf_dma[idx]); + } + } + } + kfree(eq->buf_dma); + kfree(eq->buf); + kfree(eq->l1_dma); + kfree(eq->bt_l1); + eq->buf_dma = NULL; + eq->buf = NULL; + eq->l1_dma = NULL; + eq->bt_l1 = NULL; +} + +static void hns_roce_v2_free_eq(struct hns_roce_dev *hr_dev, + struct hns_roce_eq *eq) +{ + u32 buf_chk_sz; + + buf_chk_sz = 1 << (eq->eqe_buf_pg_sz + PAGE_SHIFT); + + if (hr_dev->caps.eqe_hop_num) { + hns_roce_mhop_free_eq(hr_dev, eq); + return; + } + + if (eq->buf_list) + dma_free_coherent(hr_dev->dev, buf_chk_sz, + eq->buf_list->buf, eq->buf_list->map); +} + +static void hns_roce_config_eqc(struct hns_roce_dev *hr_dev, + struct hns_roce_eq *eq, + void *mb_buf) +{ + struct hns_roce_eq_context *eqc; + + eqc = mb_buf; + memset(eqc, 0, sizeof(struct hns_roce_eq_context)); + + /* init eqc */ + eq->doorbell = hr_dev->reg_base + ROCEE_VF_EQ_DB_CFG0_REG; + eq->hop_num = hr_dev->caps.eqe_hop_num; + eq->cons_index = 0; + eq->over_ignore = HNS_ROCE_V2_EQ_OVER_IGNORE_0; + eq->coalesce = HNS_ROCE_V2_EQ_COALESCE_0; + eq->arm_st = HNS_ROCE_V2_EQ_ALWAYS_ARMED; + eq->eqe_ba_pg_sz = hr_dev->caps.eqe_ba_pg_sz; + eq->eqe_buf_pg_sz = hr_dev->caps.eqe_buf_pg_sz; + eq->shift = ilog2((unsigned int)eq->entries); + + if (!eq->hop_num) + eq->eqe_ba = eq->buf_list->map; + else + eq->eqe_ba = eq->l0_dma; + + /* set eqc state */ + roce_set_field(eqc->byte_4, + HNS_ROCE_EQC_EQ_ST_M, + HNS_ROCE_EQC_EQ_ST_S, + HNS_ROCE_V2_EQ_STATE_VALID); + + /* set eqe hop num */ + roce_set_field(eqc->byte_4, + HNS_ROCE_EQC_HOP_NUM_M, + HNS_ROCE_EQC_HOP_NUM_S, eq->hop_num); + + /* set eqc over_ignore */ + roce_set_field(eqc->byte_4, + HNS_ROCE_EQC_OVER_IGNORE_M, + HNS_ROCE_EQC_OVER_IGNORE_S, eq->over_ignore); + + /* set eqc coalesce */ + roce_set_field(eqc->byte_4, + HNS_ROCE_EQC_COALESCE_M, + HNS_ROCE_EQC_COALESCE_S, eq->coalesce); + + /* set eqc arm_state */ + roce_set_field(eqc->byte_4, + HNS_ROCE_EQC_ARM_ST_M, + HNS_ROCE_EQC_ARM_ST_S, eq->arm_st); + + /* set eqn */ + roce_set_field(eqc->byte_4, + HNS_ROCE_EQC_EQN_M, + HNS_ROCE_EQC_EQN_S, eq->eqn); + + /* set eqe_cnt */ + roce_set_field(eqc->byte_4, + HNS_ROCE_EQC_EQE_CNT_M, + HNS_ROCE_EQC_EQE_CNT_S, + HNS_ROCE_EQ_INIT_EQE_CNT); + + /* set eqe_ba_pg_sz */ + roce_set_field(eqc->byte_8, + HNS_ROCE_EQC_BA_PG_SZ_M, + HNS_ROCE_EQC_BA_PG_SZ_S, eq->eqe_ba_pg_sz); + + /* set eqe_buf_pg_sz */ + roce_set_field(eqc->byte_8, + HNS_ROCE_EQC_BUF_PG_SZ_M, + HNS_ROCE_EQC_BUF_PG_SZ_S, eq->eqe_buf_pg_sz); + + /* set eq_producer_idx */ + roce_set_field(eqc->byte_8, + HNS_ROCE_EQC_PROD_INDX_M, + HNS_ROCE_EQC_PROD_INDX_S, + HNS_ROCE_EQ_INIT_PROD_IDX); + + /* set eq_max_cnt */ + roce_set_field(eqc->byte_12, + HNS_ROCE_EQC_MAX_CNT_M, + HNS_ROCE_EQC_MAX_CNT_S, eq->eq_max_cnt); + + /* set eq_period */ + roce_set_field(eqc->byte_12, + HNS_ROCE_EQC_PERIOD_M, + HNS_ROCE_EQC_PERIOD_S, eq->eq_period); + + /* set eqe_report_timer */ + roce_set_field(eqc->eqe_report_timer, + HNS_ROCE_EQC_REPORT_TIMER_M, + HNS_ROCE_EQC_REPORT_TIMER_S, + HNS_ROCE_EQ_INIT_REPORT_TIMER); + + /* set eqe_ba [34:3] */ + roce_set_field(eqc->eqe_ba0, + HNS_ROCE_EQC_EQE_BA_L_M, + HNS_ROCE_EQC_EQE_BA_L_S, eq->eqe_ba >> 3); + + /* set eqe_ba [64:35] */ + roce_set_field(eqc->eqe_ba1, + HNS_ROCE_EQC_EQE_BA_H_M, + HNS_ROCE_EQC_EQE_BA_H_S, eq->eqe_ba >> 35); + + /* set eq shift */ + roce_set_field(eqc->byte_28, + HNS_ROCE_EQC_SHIFT_M, + HNS_ROCE_EQC_SHIFT_S, eq->shift); + + /* set eq MSI_IDX */ + roce_set_field(eqc->byte_28, + HNS_ROCE_EQC_MSI_INDX_M, + HNS_ROCE_EQC_MSI_INDX_S, + HNS_ROCE_EQ_INIT_MSI_IDX); + + /* set cur_eqe_ba [27:12] */ + roce_set_field(eqc->byte_28, + HNS_ROCE_EQC_CUR_EQE_BA_L_M, + HNS_ROCE_EQC_CUR_EQE_BA_L_S, eq->cur_eqe_ba >> 12); + + /* set cur_eqe_ba [59:28] */ + roce_set_field(eqc->byte_32, + HNS_ROCE_EQC_CUR_EQE_BA_M_M, + HNS_ROCE_EQC_CUR_EQE_BA_M_S, eq->cur_eqe_ba >> 28); + + /* set cur_eqe_ba [63:60] */ + roce_set_field(eqc->byte_36, + HNS_ROCE_EQC_CUR_EQE_BA_H_M, + HNS_ROCE_EQC_CUR_EQE_BA_H_S, eq->cur_eqe_ba >> 60); + + /* set eq consumer idx */ + roce_set_field(eqc->byte_36, + HNS_ROCE_EQC_CONS_INDX_M, + HNS_ROCE_EQC_CONS_INDX_S, + HNS_ROCE_EQ_INIT_CONS_IDX); + + /* set nex_eqe_ba[43:12] */ + roce_set_field(eqc->nxt_eqe_ba0, + HNS_ROCE_EQC_NXT_EQE_BA_L_M, + HNS_ROCE_EQC_NXT_EQE_BA_L_S, eq->nxt_eqe_ba >> 12); + + /* set nex_eqe_ba[63:44] */ + roce_set_field(eqc->nxt_eqe_ba1, + HNS_ROCE_EQC_NXT_EQE_BA_H_M, + HNS_ROCE_EQC_NXT_EQE_BA_H_S, eq->nxt_eqe_ba >> 44); +} + +static int hns_roce_mhop_alloc_eq(struct hns_roce_dev *hr_dev, + struct hns_roce_eq *eq) +{ + struct device *dev = hr_dev->dev; + int eq_alloc_done = 0; + int eq_buf_cnt = 0; + int eqe_alloc; + u32 buf_chk_sz; + u32 bt_chk_sz; + u32 mhop_num; + u64 size; + u64 idx; + int ba_num; + int bt_num; + int record_i; + int record_j; + int i = 0; + int j = 0; + + mhop_num = hr_dev->caps.eqe_hop_num; + buf_chk_sz = 1 << (hr_dev->caps.eqe_buf_pg_sz + PAGE_SHIFT); + bt_chk_sz = 1 << (hr_dev->caps.eqe_ba_pg_sz + PAGE_SHIFT); + + ba_num = (PAGE_ALIGN(eq->entries * eq->eqe_size) + buf_chk_sz - 1) + / buf_chk_sz; + bt_num = (ba_num + bt_chk_sz / 8 - 1) / (bt_chk_sz / 8); + + /* hop_num = 0 */ + if (mhop_num == HNS_ROCE_HOP_NUM_0) { + if (eq->entries > buf_chk_sz / eq->eqe_size) { + dev_err(dev, "eq entries %d is larger than buf_pg_sz!", + eq->entries); + return -EINVAL; + } + eq->bt_l0 = dma_alloc_coherent(dev, eq->entries * eq->eqe_size, + &(eq->l0_dma), GFP_KERNEL); + if (!eq->bt_l0) + return -ENOMEM; + + eq->cur_eqe_ba = eq->l0_dma; + eq->nxt_eqe_ba = 0; + + memset(eq->bt_l0, 0, eq->entries * eq->eqe_size); + + return 0; + } + + eq->buf_dma = kcalloc(ba_num, sizeof(*eq->buf_dma), GFP_KERNEL); + if (!eq->buf_dma) + return -ENOMEM; + eq->buf = kcalloc(ba_num, sizeof(*eq->buf), GFP_KERNEL); + if (!eq->buf) + goto err_kcalloc_buf; + + if (mhop_num == 2) { + eq->l1_dma = kcalloc(bt_num, sizeof(*eq->l1_dma), GFP_KERNEL); + if (!eq->l1_dma) + goto err_kcalloc_l1_dma; + + eq->bt_l1 = kcalloc(bt_num, sizeof(*eq->bt_l1), GFP_KERNEL); + if (!eq->bt_l1) + goto err_kcalloc_bt_l1; + } + + /* alloc L0 BT */ + eq->bt_l0 = dma_alloc_coherent(dev, bt_chk_sz, &eq->l0_dma, GFP_KERNEL); + if (!eq->bt_l0) + goto err_dma_alloc_l0; + + if (mhop_num == 1) { + if (ba_num > (bt_chk_sz / 8)) + dev_err(dev, "ba_num %d is too large for 1 hop\n", + ba_num); + + /* alloc buf */ + for (i = 0; i < bt_chk_sz / 8; i++) { + if (eq_buf_cnt + 1 < ba_num) { + size = buf_chk_sz; + } else { + eqe_alloc = i * (buf_chk_sz / eq->eqe_size); + size = (eq->entries - eqe_alloc) * eq->eqe_size; + } + eq->buf[i] = dma_alloc_coherent(dev, size, + &(eq->buf_dma[i]), + GFP_KERNEL); + if (!eq->buf[i]) + goto err_dma_alloc_buf; + + memset(eq->buf[i], 0, size); + *(eq->bt_l0 + i) = eq->buf_dma[i]; + + eq_buf_cnt++; + if (eq_buf_cnt >= ba_num) + break; + } + eq->cur_eqe_ba = eq->buf_dma[0]; + eq->nxt_eqe_ba = eq->buf_dma[1]; + + } else if (mhop_num == 2) { + /* alloc L1 BT and buf */ + for (i = 0; i < bt_chk_sz / 8; i++) { + eq->bt_l1[i] = dma_alloc_coherent(dev, bt_chk_sz, + &(eq->l1_dma[i]), + GFP_KERNEL); + if (!eq->bt_l1[i]) + goto err_dma_alloc_l1; + *(eq->bt_l0 + i) = eq->l1_dma[i]; + + for (j = 0; j < bt_chk_sz / 8; j++) { + idx = i * bt_chk_sz / 8 + j; + if (eq_buf_cnt + 1 < ba_num) { + size = buf_chk_sz; + } else { + eqe_alloc = (buf_chk_sz / eq->eqe_size) + * idx; + size = (eq->entries - eqe_alloc) + * eq->eqe_size; + } + eq->buf[idx] = dma_alloc_coherent(dev, size, + &(eq->buf_dma[idx]), + GFP_KERNEL); + if (!eq->buf[idx]) + goto err_dma_alloc_buf; + + memset(eq->buf[idx], 0, size); + *(eq->bt_l1[i] + j) = eq->buf_dma[idx]; + + eq_buf_cnt++; + if (eq_buf_cnt >= ba_num) { + eq_alloc_done = 1; + break; + } + } + + if (eq_alloc_done) + break; + } + eq->cur_eqe_ba = eq->buf_dma[0]; + eq->nxt_eqe_ba = eq->buf_dma[1]; + } + + eq->l0_last_num = i + 1; + if (mhop_num == 2) + eq->l1_last_num = j + 1; + + return 0; + +err_dma_alloc_l1: + dma_free_coherent(dev, bt_chk_sz, eq->bt_l0, eq->l0_dma); + eq->bt_l0 = NULL; + eq->l0_dma = 0; + for (i -= 1; i >= 0; i--) { + dma_free_coherent(dev, bt_chk_sz, eq->bt_l1[i], + eq->l1_dma[i]); + + for (j = 0; j < bt_chk_sz / 8; j++) { + idx = i * bt_chk_sz / 8 + j; + dma_free_coherent(dev, buf_chk_sz, eq->buf[idx], + eq->buf_dma[idx]); + } + } + goto err_dma_alloc_l0; + +err_dma_alloc_buf: + dma_free_coherent(dev, bt_chk_sz, eq->bt_l0, eq->l0_dma); + eq->bt_l0 = NULL; + eq->l0_dma = 0; + + if (mhop_num == 1) + for (i -= i; i >= 0; i--) + dma_free_coherent(dev, buf_chk_sz, eq->buf[i], + eq->buf_dma[i]); + else if (mhop_num == 2) { + record_i = i; + record_j = j; + for (; i >= 0; i--) { + dma_free_coherent(dev, bt_chk_sz, eq->bt_l1[i], + eq->l1_dma[i]); + + for (j = 0; j < bt_chk_sz / 8; j++) { + if (i == record_i && j >= record_j) + break; + + idx = i * bt_chk_sz / 8 + j; + dma_free_coherent(dev, buf_chk_sz, + eq->buf[idx], + eq->buf_dma[idx]); + } + } + } + +err_dma_alloc_l0: + kfree(eq->bt_l1); + eq->bt_l1 = NULL; + +err_kcalloc_bt_l1: + kfree(eq->l1_dma); + eq->l1_dma = NULL; + +err_kcalloc_l1_dma: + kfree(eq->buf); + eq->buf = NULL; + +err_kcalloc_buf: + kfree(eq->buf_dma); + eq->buf_dma = NULL; + + return -ENOMEM; +} + +static int hns_roce_v2_create_eq(struct hns_roce_dev *hr_dev, + struct hns_roce_eq *eq, + unsigned int eq_cmd) +{ + struct device *dev = hr_dev->dev; + struct hns_roce_cmd_mailbox *mailbox; + u32 buf_chk_sz = 0; + int ret; + + /* Allocate mailbox memory */ + mailbox = hns_roce_alloc_cmd_mailbox(hr_dev); + if (IS_ERR(mailbox)) + return PTR_ERR(mailbox); + + if (!hr_dev->caps.eqe_hop_num) { + buf_chk_sz = 1 << (hr_dev->caps.eqe_buf_pg_sz + PAGE_SHIFT); + + eq->buf_list = kzalloc(sizeof(struct hns_roce_buf_list), + GFP_KERNEL); + if (!eq->buf_list) { + ret = -ENOMEM; + goto free_cmd_mbox; + } + + eq->buf_list->buf = dma_alloc_coherent(dev, buf_chk_sz, + &(eq->buf_list->map), + GFP_KERNEL); + if (!eq->buf_list->buf) { + ret = -ENOMEM; + goto err_alloc_buf; + } + + memset(eq->buf_list->buf, 0, buf_chk_sz); + } else { + ret = hns_roce_mhop_alloc_eq(hr_dev, eq); + if (ret) { + ret = -ENOMEM; + goto free_cmd_mbox; + } + } + + hns_roce_config_eqc(hr_dev, eq, mailbox->buf); + + ret = hns_roce_cmd_mbox(hr_dev, mailbox->dma, 0, eq->eqn, 0, + eq_cmd, HNS_ROCE_CMD_TIMEOUT_MSECS); + if (ret) { + dev_err(dev, "[mailbox cmd] creat eqc failed.\n"); + goto err_cmd_mbox; + } + + hns_roce_free_cmd_mailbox(hr_dev, mailbox); + + return 0; + +err_cmd_mbox: + if (!hr_dev->caps.eqe_hop_num) + dma_free_coherent(dev, buf_chk_sz, eq->buf_list->buf, + eq->buf_list->map); + else { + hns_roce_mhop_free_eq(hr_dev, eq); + goto free_cmd_mbox; + } + +err_alloc_buf: + kfree(eq->buf_list); + +free_cmd_mbox: + hns_roce_free_cmd_mailbox(hr_dev, mailbox); + + return ret; +} + +static int hns_roce_v2_init_eq_table(struct hns_roce_dev *hr_dev) +{ + struct hns_roce_eq_table *eq_table = &hr_dev->eq_table; + struct device *dev = hr_dev->dev; + struct hns_roce_eq *eq; + unsigned int eq_cmd; + int irq_num; + int eq_num; + int other_num; + int comp_num; + int aeq_num; + int i, j, k; + int ret; + + other_num = hr_dev->caps.num_other_vectors; + comp_num = hr_dev->caps.num_comp_vectors; + aeq_num = hr_dev->caps.num_aeq_vectors; + + eq_num = comp_num + aeq_num; + irq_num = eq_num + other_num; + + eq_table->eq = kcalloc(eq_num, sizeof(*eq_table->eq), GFP_KERNEL); + if (!eq_table->eq) + return -ENOMEM; + + for (i = 0; i < irq_num; i++) { + hr_dev->irq_names[i] = kzalloc(HNS_ROCE_INT_NAME_LEN, + GFP_KERNEL); + if (!hr_dev->irq_names[i]) { + ret = -ENOMEM; + goto err_failed_kzalloc; + } + } + + /* create eq */ + for (j = 0; j < eq_num; j++) { + eq = &eq_table->eq[j]; + eq->hr_dev = hr_dev; + eq->eqn = j; + if (j < comp_num) { + /* CEQ */ + eq_cmd = HNS_ROCE_CMD_CREATE_CEQC; + eq->type_flag = HNS_ROCE_CEQ; + eq->entries = hr_dev->caps.ceqe_depth; + eq->eqe_size = HNS_ROCE_CEQ_ENTRY_SIZE; + eq->irq = hr_dev->irq[j + other_num + aeq_num]; + eq->eq_max_cnt = HNS_ROCE_CEQ_DEFAULT_BURST_NUM; + eq->eq_period = HNS_ROCE_CEQ_DEFAULT_INTERVAL; + } else { + /* AEQ */ + eq_cmd = HNS_ROCE_CMD_CREATE_AEQC; + eq->type_flag = HNS_ROCE_AEQ; + eq->entries = hr_dev->caps.aeqe_depth; + eq->eqe_size = HNS_ROCE_AEQ_ENTRY_SIZE; + eq->irq = hr_dev->irq[j - comp_num + other_num]; + eq->eq_max_cnt = HNS_ROCE_AEQ_DEFAULT_BURST_NUM; + eq->eq_period = HNS_ROCE_AEQ_DEFAULT_INTERVAL; + } + + ret = hns_roce_v2_create_eq(hr_dev, eq, eq_cmd); + if (ret) { + dev_err(dev, "eq create failed.\n"); + goto err_create_eq_fail; + } + } + + /* enable irq */ + hns_roce_v2_int_mask_enable(hr_dev, eq_num, EQ_ENABLE); + + /* irq contains: abnormal + AEQ + CEQ*/ + for (k = 0; k < irq_num; k++) + if (k < other_num) + snprintf((char *)hr_dev->irq_names[k], + HNS_ROCE_INT_NAME_LEN, "hns-abn-%d", k); + else if (k < (other_num + aeq_num)) + snprintf((char *)hr_dev->irq_names[k], + HNS_ROCE_INT_NAME_LEN, "hns-aeq-%d", + k - other_num); + else + snprintf((char *)hr_dev->irq_names[k], + HNS_ROCE_INT_NAME_LEN, "hns-ceq-%d", + k - other_num - aeq_num); + + for (k = 0; k < irq_num; k++) { + if (k < other_num) + ret = request_irq(hr_dev->irq[k], + hns_roce_v2_msix_interrupt_abn, + 0, hr_dev->irq_names[k], hr_dev); + + else if (k < (other_num + comp_num)) + ret = request_irq(eq_table->eq[k - other_num].irq, + hns_roce_v2_msix_interrupt_eq, + 0, hr_dev->irq_names[k + aeq_num], + &eq_table->eq[k - other_num]); + else + ret = request_irq(eq_table->eq[k - other_num].irq, + hns_roce_v2_msix_interrupt_eq, + 0, hr_dev->irq_names[k - comp_num], + &eq_table->eq[k - other_num]); + if (ret) { + dev_err(dev, "Request irq error!\n"); + goto err_request_irq_fail; + } + } + + return 0; + +err_request_irq_fail: + for (k -= 1; k >= 0; k--) + if (k < other_num) + free_irq(hr_dev->irq[k], hr_dev); + else + free_irq(eq_table->eq[k - other_num].irq, + &eq_table->eq[k - other_num]); + +err_create_eq_fail: + for (j -= 1; j >= 0; j--) + hns_roce_v2_free_eq(hr_dev, &eq_table->eq[j]); + +err_failed_kzalloc: + for (i -= 1; i >= 0; i--) + kfree(hr_dev->irq_names[i]); + kfree(eq_table->eq); + + return ret; +} + +static void hns_roce_v2_cleanup_eq_table(struct hns_roce_dev *hr_dev) +{ + struct hns_roce_eq_table *eq_table = &hr_dev->eq_table; + int irq_num; + int eq_num; + int i; + + eq_num = hr_dev->caps.num_comp_vectors + hr_dev->caps.num_aeq_vectors; + irq_num = eq_num + hr_dev->caps.num_other_vectors; + + /* Disable irq */ + hns_roce_v2_int_mask_enable(hr_dev, eq_num, EQ_DISABLE); + + for (i = 0; i < hr_dev->caps.num_other_vectors; i++) + free_irq(hr_dev->irq[i], hr_dev); + + for (i = 0; i < eq_num; i++) { + hns_roce_v2_destroy_eqc(hr_dev, i); + + free_irq(eq_table->eq[i].irq, &eq_table->eq[i]); + + hns_roce_v2_free_eq(hr_dev, &eq_table->eq[i]); + } + + for (i = 0; i < irq_num; i++) + kfree(hr_dev->irq_names[i]); + + kfree(eq_table->eq); +} + static const struct hns_roce_hw hns_roce_hw_v2 = { .cmq_init = hns_roce_v2_cmq_init, .cmq_exit = hns_roce_v2_cmq_exit, @@ -3183,6 +4442,8 @@ static const struct hns_roce_hw hns_roce_hw_v2 = { .post_recv = hns_roce_v2_post_recv, .req_notify_cq = hns_roce_v2_req_notify_cq, .poll_cq = hns_roce_v2_poll_cq, + .init_eq = hns_roce_v2_init_eq_table, + .cleanup_eq = hns_roce_v2_cleanup_eq_table, }; static const struct pci_device_id hns_roce_hw_v2_pci_tbl[] = { @@ -3197,6 +4458,7 @@ static int hns_roce_hw_v2_get_cfg(struct hns_roce_dev *hr_dev, struct hnae3_handle *handle) { const struct pci_device_id *id; + int i; id = pci_match_id(hns_roce_hw_v2_pci_tbl, hr_dev->pci_dev); if (!id) { @@ -3214,8 +4476,12 @@ static int hns_roce_hw_v2_get_cfg(struct hns_roce_dev *hr_dev, hr_dev->iboe.netdevs[0] = handle->rinfo.netdev; hr_dev->iboe.phy_port[0] = 0; + for (i = 0; i < HNS_ROCE_V2_MAX_IRQ_NUM; i++) + hr_dev->irq[i] = pci_irq_vector(handle->pdev, + i + handle->rinfo.base_vector); + /* cmd issue mode: 0 is poll, 1 is event */ - hr_dev->cmd_mod = 0; + hr_dev->cmd_mod = 1; hr_dev->loop_idc = 0; return 0; diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.h b/drivers/infiniband/hw/hns/hns_roce_hw_v2.h index 04b7a51b8efb..463edab9b719 100644 --- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.h +++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.h @@ -53,6 +53,10 @@ #define HNS_ROCE_V2_MAX_SQ_INLINE 0x20 #define HNS_ROCE_V2_UAR_NUM 256 #define HNS_ROCE_V2_PHY_UAR_NUM 1 +#define HNS_ROCE_V2_MAX_IRQ_NUM 65 +#define HNS_ROCE_V2_COMP_VEC_NUM 63 +#define HNS_ROCE_V2_AEQE_VEC_NUM 1 +#define HNS_ROCE_V2_ABNORMAL_VEC_NUM 1 #define HNS_ROCE_V2_MAX_MTPT_NUM 0x8000 #define HNS_ROCE_V2_MAX_MTT_SEGS 0x1000000 #define HNS_ROCE_V2_MAX_CQE_SEGS 0x1000000 @@ -78,6 +82,8 @@ #define HNS_ROCE_MTT_HOP_NUM 1 #define HNS_ROCE_CQE_HOP_NUM 1 #define HNS_ROCE_PBL_HOP_NUM 2 +#define HNS_ROCE_EQE_HOP_NUM 2 + #define HNS_ROCE_V2_GID_INDEX_NUM 256 #define HNS_ROCE_V2_TABLE_CHUNK_SIZE (1 << 18) @@ -105,6 +111,12 @@ (step_idx == 1 && hop_num == 1) || \ (step_idx == 2 && hop_num == 2)) +enum { + NO_ARMED = 0x0, + REG_NXT_CEQE = 0x2, + REG_NXT_SE_CEQE = 0x3 +}; + #define V2_CQ_DB_REQ_NOT_SOL 0 #define V2_CQ_DB_REQ_NOT 1 @@ -229,6 +241,9 @@ struct hns_roce_v2_cq_context { u32 cqe_report_timer; u32 byte_64_se_cqe_idx; }; +#define HNS_ROCE_V2_CQ_DEFAULT_BURST_NUM 0x0 +#define HNS_ROCE_V2_CQ_DEFAULT_INTERVAL 0x0 + #define V2_CQC_BYTE_4_CQ_ST_S 0 #define V2_CQC_BYTE_4_CQ_ST_M GENMASK(1, 0) @@ -1129,9 +1144,6 @@ struct hns_roce_cmq_desc { u32 data[6]; }; -#define ROCEE_VF_MB_CFG0_REG 0x40 -#define ROCEE_VF_MB_STATUS_REG 0x58 - #define HNS_ROCE_V2_GO_BIT_TIMEOUT_MSECS 10000 #define HNS_ROCE_HW_RUN_BIT_SHIFT 31 @@ -1174,4 +1186,178 @@ struct hns_roce_v2_priv { struct hns_roce_v2_cmq cmq; }; +struct hns_roce_eq_context { + u32 byte_4; + u32 byte_8; + u32 byte_12; + u32 eqe_report_timer; + u32 eqe_ba0; + u32 eqe_ba1; + u32 byte_28; + u32 byte_32; + u32 byte_36; + u32 nxt_eqe_ba0; + u32 nxt_eqe_ba1; + u32 rsv[5]; +}; + +#define HNS_ROCE_AEQ_DEFAULT_BURST_NUM 0x0 +#define HNS_ROCE_AEQ_DEFAULT_INTERVAL 0x0 +#define HNS_ROCE_CEQ_DEFAULT_BURST_NUM 0x0 +#define HNS_ROCE_CEQ_DEFAULT_INTERVAL 0x0 + +#define HNS_ROCE_V2_EQ_STATE_INVALID 0 +#define HNS_ROCE_V2_EQ_STATE_VALID 1 +#define HNS_ROCE_V2_EQ_STATE_OVERFLOW 2 +#define HNS_ROCE_V2_EQ_STATE_FAILURE 3 + +#define HNS_ROCE_V2_EQ_OVER_IGNORE_0 0 +#define HNS_ROCE_V2_EQ_OVER_IGNORE_1 1 + +#define HNS_ROCE_V2_EQ_COALESCE_0 0 +#define HNS_ROCE_V2_EQ_COALESCE_1 1 + +#define HNS_ROCE_V2_EQ_FIRED 0 +#define HNS_ROCE_V2_EQ_ARMED 1 +#define HNS_ROCE_V2_EQ_ALWAYS_ARMED 3 + +#define HNS_ROCE_EQ_INIT_EQE_CNT 0 +#define HNS_ROCE_EQ_INIT_PROD_IDX 0 +#define HNS_ROCE_EQ_INIT_REPORT_TIMER 0 +#define HNS_ROCE_EQ_INIT_MSI_IDX 0 +#define HNS_ROCE_EQ_INIT_CONS_IDX 0 +#define HNS_ROCE_EQ_INIT_NXT_EQE_BA 0 + +#define HNS_ROCE_V2_CEQ_CEQE_OWNER_S 31 +#define HNS_ROCE_V2_AEQ_AEQE_OWNER_S 31 + +#define HNS_ROCE_V2_COMP_EQE_NUM 0x1000 +#define HNS_ROCE_V2_ASYNC_EQE_NUM 0x1000 + +#define HNS_ROCE_V2_VF_INT_ST_AEQ_OVERFLOW_S 0 +#define HNS_ROCE_V2_VF_INT_ST_BUS_ERR_S 1 +#define HNS_ROCE_V2_VF_INT_ST_OTHER_ERR_S 2 + +#define HNS_ROCE_EQ_DB_CMD_AEQ 0x0 +#define HNS_ROCE_EQ_DB_CMD_AEQ_ARMED 0x1 +#define HNS_ROCE_EQ_DB_CMD_CEQ 0x2 +#define HNS_ROCE_EQ_DB_CMD_CEQ_ARMED 0x3 + +#define EQ_ENABLE 1 +#define EQ_DISABLE 0 + +#define EQ_REG_OFFSET 0x4 + +#define HNS_ROCE_INT_NAME_LEN 32 +#define HNS_ROCE_V2_EQN_M GENMASK(23, 0) + +#define HNS_ROCE_V2_CONS_IDX_M GENMASK(23, 0) + +#define HNS_ROCE_V2_VF_ABN_INT_EN_S 0 +#define HNS_ROCE_V2_VF_ABN_INT_EN_M GENMASK(0, 0) +#define HNS_ROCE_V2_VF_ABN_INT_ST_M GENMASK(2, 0) +#define HNS_ROCE_V2_VF_ABN_INT_CFG_M GENMASK(2, 0) +#define HNS_ROCE_V2_VF_EVENT_INT_EN_M GENMASK(0, 0) + +/* WORD0 */ +#define HNS_ROCE_EQC_EQ_ST_S 0 +#define HNS_ROCE_EQC_EQ_ST_M GENMASK(1, 0) + +#define HNS_ROCE_EQC_HOP_NUM_S 2 +#define HNS_ROCE_EQC_HOP_NUM_M GENMASK(3, 2) + +#define HNS_ROCE_EQC_OVER_IGNORE_S 4 +#define HNS_ROCE_EQC_OVER_IGNORE_M GENMASK(4, 4) + +#define HNS_ROCE_EQC_COALESCE_S 5 +#define HNS_ROCE_EQC_COALESCE_M GENMASK(5, 5) + +#define HNS_ROCE_EQC_ARM_ST_S 6 +#define HNS_ROCE_EQC_ARM_ST_M GENMASK(7, 6) + +#define HNS_ROCE_EQC_EQN_S 8 +#define HNS_ROCE_EQC_EQN_M GENMASK(15, 8) + +#define HNS_ROCE_EQC_EQE_CNT_S 16 +#define HNS_ROCE_EQC_EQE_CNT_M GENMASK(31, 16) + +/* WORD1 */ +#define HNS_ROCE_EQC_BA_PG_SZ_S 0 +#define HNS_ROCE_EQC_BA_PG_SZ_M GENMASK(3, 0) + +#define HNS_ROCE_EQC_BUF_PG_SZ_S 4 +#define HNS_ROCE_EQC_BUF_PG_SZ_M GENMASK(7, 4) + +#define HNS_ROCE_EQC_PROD_INDX_S 8 +#define HNS_ROCE_EQC_PROD_INDX_M GENMASK(31, 8) + +/* WORD2 */ +#define HNS_ROCE_EQC_MAX_CNT_S 0 +#define HNS_ROCE_EQC_MAX_CNT_M GENMASK(15, 0) + +#define HNS_ROCE_EQC_PERIOD_S 16 +#define HNS_ROCE_EQC_PERIOD_M GENMASK(31, 16) + +/* WORD3 */ +#define HNS_ROCE_EQC_REPORT_TIMER_S 0 +#define HNS_ROCE_EQC_REPORT_TIMER_M GENMASK(31, 0) + +/* WORD4 */ +#define HNS_ROCE_EQC_EQE_BA_L_S 0 +#define HNS_ROCE_EQC_EQE_BA_L_M GENMASK(31, 0) + +/* WORD5 */ +#define HNS_ROCE_EQC_EQE_BA_H_S 0 +#define HNS_ROCE_EQC_EQE_BA_H_M GENMASK(28, 0) + +/* WORD6 */ +#define HNS_ROCE_EQC_SHIFT_S 0 +#define HNS_ROCE_EQC_SHIFT_M GENMASK(7, 0) + +#define HNS_ROCE_EQC_MSI_INDX_S 8 +#define HNS_ROCE_EQC_MSI_INDX_M GENMASK(15, 8) + +#define HNS_ROCE_EQC_CUR_EQE_BA_L_S 16 +#define HNS_ROCE_EQC_CUR_EQE_BA_L_M GENMASK(31, 16) + +/* WORD7 */ +#define HNS_ROCE_EQC_CUR_EQE_BA_M_S 0 +#define HNS_ROCE_EQC_CUR_EQE_BA_M_M GENMASK(31, 0) + +/* WORD8 */ +#define HNS_ROCE_EQC_CUR_EQE_BA_H_S 0 +#define HNS_ROCE_EQC_CUR_EQE_BA_H_M GENMASK(3, 0) + +#define HNS_ROCE_EQC_CONS_INDX_S 8 +#define HNS_ROCE_EQC_CONS_INDX_M GENMASK(31, 8) + +/* WORD9 */ +#define HNS_ROCE_EQC_NXT_EQE_BA_L_S 0 +#define HNS_ROCE_EQC_NXT_EQE_BA_L_M GENMASK(31, 0) + +/* WORD10 */ +#define HNS_ROCE_EQC_NXT_EQE_BA_H_S 0 +#define HNS_ROCE_EQC_NXT_EQE_BA_H_M GENMASK(19, 0) + +#define HNS_ROCE_V2_CEQE_COMP_CQN_S 0 +#define HNS_ROCE_V2_CEQE_COMP_CQN_M GENMASK(23, 0) + +#define HNS_ROCE_V2_AEQE_EVENT_TYPE_S 0 +#define HNS_ROCE_V2_AEQE_EVENT_TYPE_M GENMASK(7, 0) + +#define HNS_ROCE_V2_AEQE_SUB_TYPE_S 8 +#define HNS_ROCE_V2_AEQE_SUB_TYPE_M GENMASK(15, 8) + +#define HNS_ROCE_V2_EQ_DB_CMD_S 16 +#define HNS_ROCE_V2_EQ_DB_CMD_M GENMASK(17, 16) + +#define HNS_ROCE_V2_EQ_DB_TAG_S 0 +#define HNS_ROCE_V2_EQ_DB_TAG_M GENMASK(7, 0) + +#define HNS_ROCE_V2_EQ_DB_PARA_S 0 +#define HNS_ROCE_V2_EQ_DB_PARA_M GENMASK(23, 0) + +#define HNS_ROCE_V2_AEQE_EVENT_QUEUE_NUM_S 0 +#define HNS_ROCE_V2_AEQE_EVENT_QUEUE_NUM_M GENMASK(23, 0) + #endif diff --git a/drivers/infiniband/hw/hns/hns_roce_main.c b/drivers/infiniband/hw/hns/hns_roce_main.c index cf02ac2d3596..aa0c242ddc50 100644 --- a/drivers/infiniband/hw/hns/hns_roce_main.c +++ b/drivers/infiniband/hw/hns/hns_roce_main.c @@ -748,12 +748,10 @@ int hns_roce_init(struct hns_roce_dev *hr_dev) goto error_failed_cmd_init; } - if (hr_dev->cmd_mod) { - ret = hns_roce_init_eq_table(hr_dev); - if (ret) { - dev_err(dev, "eq init failed!\n"); - goto error_failed_eq_table; - } + ret = hr_dev->hw->init_eq(hr_dev); + if (ret) { + dev_err(dev, "eq init failed!\n"); + goto error_failed_eq_table; } if (hr_dev->cmd_mod) { @@ -805,8 +803,7 @@ error_failed_init_hem: hns_roce_cmd_use_polling(hr_dev); error_failed_use_event: - if (hr_dev->cmd_mod) - hns_roce_cleanup_eq_table(hr_dev); + hr_dev->hw->cleanup_eq(hr_dev); error_failed_eq_table: hns_roce_cmd_cleanup(hr_dev); @@ -837,8 +834,7 @@ void hns_roce_exit(struct hns_roce_dev *hr_dev) if (hr_dev->cmd_mod) hns_roce_cmd_use_polling(hr_dev); - if (hr_dev->cmd_mod) - hns_roce_cleanup_eq_table(hr_dev); + hr_dev->hw->cleanup_eq(hr_dev); hns_roce_cmd_cleanup(hr_dev); if (hr_dev->hw->cmq_exit) hr_dev->hw->cmq_exit(hr_dev); diff --git a/drivers/infiniband/hw/hns/hns_roce_qp.c b/drivers/infiniband/hw/hns/hns_roce_qp.c index 49586ec8126a..351fa3119420 100644 --- a/drivers/infiniband/hw/hns/hns_roce_qp.c +++ b/drivers/infiniband/hw/hns/hns_roce_qp.c @@ -65,6 +65,7 @@ void hns_roce_qp_event(struct hns_roce_dev *hr_dev, u32 qpn, int event_type) if (atomic_dec_and_test(&qp->refcount)) complete(&qp->free); } +EXPORT_SYMBOL_GPL(hns_roce_qp_event); static void hns_roce_ib_qp_event(struct hns_roce_qp *hr_qp, enum hns_roce_event type) @@ -493,6 +494,7 @@ static int hns_roce_create_qp_common(struct hns_roce_dev *hr_dev, int ret = 0; u32 page_shift; u32 npages; + int i; mutex_init(&hr_qp->mutex); spin_lock_init(&hr_qp->sq.lock); @@ -512,18 +514,48 @@ static int hns_roce_create_qp_common(struct hns_roce_dev *hr_dev, goto err_out; } + if (hr_dev->caps.flags & HNS_ROCE_CAP_FLAG_RQ_INLINE) { + /* allocate recv inline buf */ + hr_qp->rq_inl_buf.wqe_list = kcalloc(hr_qp->rq.wqe_cnt, + sizeof(struct hns_roce_rinl_wqe), + GFP_KERNEL); + if (!hr_qp->rq_inl_buf.wqe_list) { + ret = -ENOMEM; + goto err_out; + } + + hr_qp->rq_inl_buf.wqe_cnt = hr_qp->rq.wqe_cnt; + + /* Firstly, allocate a list of sge space buffer */ + hr_qp->rq_inl_buf.wqe_list[0].sg_list = + kcalloc(hr_qp->rq_inl_buf.wqe_cnt, + init_attr->cap.max_recv_sge * + sizeof(struct hns_roce_rinl_sge), + GFP_KERNEL); + if (!hr_qp->rq_inl_buf.wqe_list[0].sg_list) { + ret = -ENOMEM; + goto err_wqe_list; + } + + for (i = 1; i < hr_qp->rq_inl_buf.wqe_cnt; i++) + /* Secondly, reallocate the buffer */ + hr_qp->rq_inl_buf.wqe_list[i].sg_list = + &hr_qp->rq_inl_buf.wqe_list[0].sg_list[i * + init_attr->cap.max_recv_sge]; + } + if (ib_pd->uobject) { if (ib_copy_from_udata(&ucmd, udata, sizeof(ucmd))) { dev_err(dev, "ib_copy_from_udata error for create qp\n"); ret = -EFAULT; - goto err_out; + goto err_rq_sge_list; } ret = hns_roce_set_user_sq_size(hr_dev, &init_attr->cap, hr_qp, &ucmd); if (ret) { dev_err(dev, "hns_roce_set_user_sq_size error for create qp\n"); - goto err_out; + goto err_rq_sge_list; } hr_qp->umem = ib_umem_get(ib_pd->uobject->context, @@ -532,7 +564,7 @@ static int hns_roce_create_qp_common(struct hns_roce_dev *hr_dev, if (IS_ERR(hr_qp->umem)) { dev_err(dev, "ib_umem_get error for create qp\n"); ret = PTR_ERR(hr_qp->umem); - goto err_out; + goto err_rq_sge_list; } hr_qp->mtt.mtt_type = MTT_TYPE_WQE; @@ -566,13 +598,13 @@ static int hns_roce_create_qp_common(struct hns_roce_dev *hr_dev, IB_QP_CREATE_BLOCK_MULTICAST_LOOPBACK) { dev_err(dev, "init_attr->create_flags error!\n"); ret = -EINVAL; - goto err_out; + goto err_rq_sge_list; } if (init_attr->create_flags & IB_QP_CREATE_IPOIB_UD_LSO) { dev_err(dev, "init_attr->create_flags error!\n"); ret = -EINVAL; - goto err_out; + goto err_rq_sge_list; } /* Set SQ size */ @@ -580,7 +612,7 @@ static int hns_roce_create_qp_common(struct hns_roce_dev *hr_dev, hr_qp); if (ret) { dev_err(dev, "hns_roce_set_kernel_sq_size error!\n"); - goto err_out; + goto err_rq_sge_list; } /* QP doorbell register address */ @@ -596,7 +628,7 @@ static int hns_roce_create_qp_common(struct hns_roce_dev *hr_dev, &hr_qp->hr_buf, page_shift)) { dev_err(dev, "hns_roce_buf_alloc error!\n"); ret = -ENOMEM; - goto err_out; + goto err_rq_sge_list; } hr_qp->mtt.mtt_type = MTT_TYPE_WQE; @@ -678,6 +710,14 @@ err_buf: else hns_roce_buf_free(hr_dev, hr_qp->buff_size, &hr_qp->hr_buf); +err_rq_sge_list: + if (hr_dev->caps.flags & HNS_ROCE_CAP_FLAG_RQ_INLINE) + kfree(hr_qp->rq_inl_buf.wqe_list[0].sg_list); + +err_wqe_list: + if (hr_dev->caps.flags & HNS_ROCE_CAP_FLAG_RQ_INLINE) + kfree(hr_qp->rq_inl_buf.wqe_list); + err_out: return ret; } diff --git a/drivers/infiniband/hw/i40iw/Kconfig b/drivers/infiniband/hw/i40iw/Kconfig index f6d20ba88c03..2962979c06e9 100644 --- a/drivers/infiniband/hw/i40iw/Kconfig +++ b/drivers/infiniband/hw/i40iw/Kconfig @@ -5,4 +5,3 @@ config INFINIBAND_I40IW select GENERIC_ALLOCATOR ---help--- Intel(R) Ethernet X722 iWARP Driver - INET && I40IW && INFINIBAND && I40E diff --git a/drivers/infiniband/hw/i40iw/i40iw.h b/drivers/infiniband/hw/i40iw/i40iw.h index 4ae9131b6350..bcddd7061fc0 100644 --- a/drivers/infiniband/hw/i40iw/i40iw.h +++ b/drivers/infiniband/hw/i40iw/i40iw.h @@ -587,5 +587,8 @@ int i40iw_inet6addr_event(struct notifier_block *notifier, int i40iw_net_event(struct notifier_block *notifier, unsigned long event, void *ptr); +int i40iw_netdevice_event(struct notifier_block *notifier, + unsigned long event, + void *ptr); #endif diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.c b/drivers/infiniband/hw/i40iw/i40iw_cm.c index 77870f9e1736..abf4cd897849 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_cm.c +++ b/drivers/infiniband/hw/i40iw/i40iw_cm.c @@ -92,14 +92,9 @@ void i40iw_free_sqbuf(struct i40iw_sc_vsi *vsi, void *bufp) static u8 i40iw_derive_hw_ird_setting(u16 cm_ird) { u8 encoded_ird_size; - u8 pof2_cm_ird = 1; - - /* round-off to next powerof2 */ - while (pof2_cm_ird < cm_ird) - pof2_cm_ird *= 2; /* ird_size field is encoded in qp_ctx */ - switch (pof2_cm_ird) { + switch (cm_ird ? roundup_pow_of_two(cm_ird) : 0) { case I40IW_HW_IRD_SETTING_64: encoded_ird_size = 3; break; @@ -125,13 +120,16 @@ static u8 i40iw_derive_hw_ird_setting(u16 cm_ird) * @conn_ird: connection IRD * @conn_ord: connection ORD */ -static void i40iw_record_ird_ord(struct i40iw_cm_node *cm_node, u16 conn_ird, u16 conn_ord) +static void i40iw_record_ird_ord(struct i40iw_cm_node *cm_node, u32 conn_ird, + u32 conn_ord) { if (conn_ird > I40IW_MAX_IRD_SIZE) conn_ird = I40IW_MAX_IRD_SIZE; if (conn_ord > I40IW_MAX_ORD_SIZE) conn_ord = I40IW_MAX_ORD_SIZE; + else if (!conn_ord && cm_node->send_rdma0_op == SEND_RDMA_READ_ZERO) + conn_ord = 1; cm_node->ird_size = conn_ird; cm_node->ord_size = conn_ord; @@ -2878,15 +2876,13 @@ static struct i40iw_cm_listener *i40iw_make_listen_node( * i40iw_create_cm_node - make a connection node with params * @cm_core: cm's core * @iwdev: iwarp device structure - * @private_data_len: len to provate data for mpa request - * @private_data: pointer to private data for connection + * @conn_param: upper layer connection parameters * @cm_info: quad info for connection */ static struct i40iw_cm_node *i40iw_create_cm_node( struct i40iw_cm_core *cm_core, struct i40iw_device *iwdev, - u16 private_data_len, - void *private_data, + struct iw_cm_conn_param *conn_param, struct i40iw_cm_info *cm_info) { struct i40iw_cm_node *cm_node; @@ -2894,6 +2890,9 @@ static struct i40iw_cm_node *i40iw_create_cm_node( struct i40iw_cm_node *loopback_remotenode; struct i40iw_cm_info loopback_cm_info; + u16 private_data_len = conn_param->private_data_len; + const void *private_data = conn_param->private_data; + /* create a CM connection node */ cm_node = i40iw_make_cm_node(cm_core, iwdev, cm_info, NULL); if (!cm_node) @@ -2902,6 +2901,8 @@ static struct i40iw_cm_node *i40iw_create_cm_node( cm_node->tcp_cntxt.client = 1; cm_node->tcp_cntxt.rcv_wscale = I40IW_CM_DEFAULT_RCV_WND_SCALE; + i40iw_record_ird_ord(cm_node, conn_param->ird, conn_param->ord); + if (!memcmp(cm_info->loc_addr, cm_info->rem_addr, sizeof(cm_info->loc_addr))) { loopback_remotelistener = i40iw_find_listener( cm_core, @@ -2935,6 +2936,10 @@ static struct i40iw_cm_node *i40iw_create_cm_node( private_data_len); loopback_remotenode->pdata.size = private_data_len; + if (loopback_remotenode->ord_size > cm_node->ird_size) + loopback_remotenode->ord_size = + cm_node->ird_size; + cm_node->state = I40IW_CM_STATE_OFFLOADED; cm_node->tcp_cntxt.rcv_nxt = loopback_remotenode->tcp_cntxt.loc_seq_num; @@ -3691,7 +3696,7 @@ int i40iw_accept(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) cm_node->qhash_set = false; i40iw_modify_qp(&iwqp->ibqp, &attr, IB_QP_STATE, NULL); - cm_node->accelerated = 1; + cm_node->accelerated = true; status = i40iw_send_cm_event(cm_node, cm_id, IW_CM_EVENT_ESTABLISHED, 0); if (status) @@ -3815,9 +3820,7 @@ int i40iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) __func__, cm_id->tos, cm_info.user_pri); cm_id->add_ref(cm_id); cm_node = i40iw_create_cm_node(&iwdev->cm_core, iwdev, - conn_param->private_data_len, - (void *)conn_param->private_data, - &cm_info); + conn_param, &cm_info); if (IS_ERR(cm_node)) { ret = PTR_ERR(cm_node); @@ -3849,11 +3852,6 @@ int i40iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) } cm_node->apbvt_set = true; - i40iw_record_ird_ord(cm_node, (u16)conn_param->ird, (u16)conn_param->ord); - if (cm_node->send_rdma0_op == SEND_RDMA_READ_ZERO && - !cm_node->ord_size) - cm_node->ord_size = 1; - iwqp->cm_node = cm_node; cm_node->iwqp = iwqp; iwqp->cm_id = cm_id; @@ -4058,7 +4056,7 @@ static void i40iw_cm_event_connected(struct i40iw_cm_event *event) cm_node->qhash_set = false; i40iw_modify_qp(&iwqp->ibqp, &attr, IB_QP_STATE, NULL); - cm_node->accelerated = 1; + cm_node->accelerated = true; status = i40iw_send_cm_event(cm_node, cm_id, IW_CM_EVENT_CONNECT_REPLY, 0); if (status) @@ -4242,10 +4240,16 @@ set_qhash: } /** - * i40iw_cm_disconnect_all - disconnect all connected qp's + * i40iw_cm_teardown_connections - teardown QPs * @iwdev: device pointer + * @ipaddr: Pointer to IPv4 or IPv6 address + * @ipv4: flag indicating IPv4 when true + * @disconnect_all: flag indicating disconnect all QPs + * teardown QPs where source or destination addr matches ip addr */ -void i40iw_cm_disconnect_all(struct i40iw_device *iwdev) +void i40iw_cm_teardown_connections(struct i40iw_device *iwdev, u32 *ipaddr, + struct i40iw_cm_info *nfo, + bool disconnect_all) { struct i40iw_cm_core *cm_core = &iwdev->cm_core; struct list_head *list_core_temp; @@ -4259,8 +4263,13 @@ void i40iw_cm_disconnect_all(struct i40iw_device *iwdev) spin_lock_irqsave(&cm_core->ht_lock, flags); list_for_each_safe(list_node, list_core_temp, &cm_core->connected_nodes) { cm_node = container_of(list_node, struct i40iw_cm_node, list); - atomic_inc(&cm_node->ref_count); - list_add(&cm_node->connected_entry, &connected_list); + if (disconnect_all || + (nfo->vlan_id == cm_node->vlan_id && + (!memcmp(cm_node->loc_addr, ipaddr, nfo->ipv4 ? 4 : 16) || + !memcmp(cm_node->rem_addr, ipaddr, nfo->ipv4 ? 4 : 16)))) { + atomic_inc(&cm_node->ref_count); + list_add(&cm_node->connected_entry, &connected_list); + } } spin_unlock_irqrestore(&cm_core->ht_lock, flags); @@ -4294,6 +4303,9 @@ void i40iw_if_notify(struct i40iw_device *iwdev, struct net_device *netdev, enum i40iw_quad_hash_manage_type op = ifup ? I40IW_QHASH_MANAGE_TYPE_ADD : I40IW_QHASH_MANAGE_TYPE_DELETE; + nfo.vlan_id = vlan_id; + nfo.ipv4 = ipv4; + /* Disable or enable qhash for listeners */ spin_lock_irqsave(&cm_core->listen_list_lock, flags); list_for_each_entry(listen_node, &cm_core->listen_nodes, list) { @@ -4303,8 +4315,6 @@ void i40iw_if_notify(struct i40iw_device *iwdev, struct net_device *netdev, memcpy(nfo.loc_addr, listen_node->loc_addr, sizeof(nfo.loc_addr)); nfo.loc_port = listen_node->loc_port; - nfo.ipv4 = listen_node->ipv4; - nfo.vlan_id = listen_node->vlan_id; nfo.user_pri = listen_node->user_pri; if (!list_empty(&listen_node->child_listen_list)) { i40iw_qhash_ctrl(iwdev, @@ -4326,7 +4336,7 @@ void i40iw_if_notify(struct i40iw_device *iwdev, struct net_device *netdev, } spin_unlock_irqrestore(&cm_core->listen_list_lock, flags); - /* disconnect any connected qp's on ifdown */ + /* teardown connected qp's on ifdown */ if (!ifup) - i40iw_cm_disconnect_all(iwdev); + i40iw_cm_teardown_connections(iwdev, ipaddr, &nfo, false); } diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.h b/drivers/infiniband/hw/i40iw/i40iw_cm.h index 0d5840d2c4fc..cf60c451e071 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_cm.h +++ b/drivers/infiniband/hw/i40iw/i40iw_cm.h @@ -276,8 +276,6 @@ struct i40iw_cm_tcp_context { u32 mss; u8 snd_wscale; u8 rcv_wscale; - - struct timeval sent_ts; }; enum i40iw_cm_listener_state { @@ -337,7 +335,7 @@ struct i40iw_cm_node { u16 mpav2_ird_ord; struct iw_cm_id *cm_id; struct list_head list; - int accelerated; + bool accelerated; struct i40iw_cm_listener *listener; int apbvt_set; int accept_pend; @@ -455,5 +453,7 @@ int i40iw_arp_table(struct i40iw_device *iwdev, void i40iw_if_notify(struct i40iw_device *iwdev, struct net_device *netdev, u32 *ipaddr, bool ipv4, bool ifup); -void i40iw_cm_disconnect_all(struct i40iw_device *iwdev); +void i40iw_cm_teardown_connections(struct i40iw_device *iwdev, u32 *ipaddr, + struct i40iw_cm_info *nfo, + bool disconnect_all); #endif /* I40IW_CM_H */ diff --git a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c index da9821a10e0d..caa958177631 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c +++ b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c @@ -3872,7 +3872,6 @@ enum i40iw_status_code i40iw_config_fpm_values(struct i40iw_sc_dev *dev, u32 qp_ struct i40iw_virt_mem virt_mem; u32 i, mem_size; u32 qpwantedoriginal, qpwanted, mrwanted, pblewanted; - u32 powerof2; u64 sd_needed; u32 loop_count = 0; @@ -3928,8 +3927,10 @@ enum i40iw_status_code i40iw_config_fpm_values(struct i40iw_sc_dev *dev, u32 qp_ hmc_info->hmc_obj[I40IW_HMC_IW_APBVT_ENTRY].cnt = 1; hmc_info->hmc_obj[I40IW_HMC_IW_MR].cnt = mrwanted; - hmc_info->hmc_obj[I40IW_HMC_IW_XF].cnt = I40IW_MAX_WQ_ENTRIES * qpwanted; - hmc_info->hmc_obj[I40IW_HMC_IW_Q1].cnt = 4 * I40IW_MAX_IRD_SIZE * qpwanted; + hmc_info->hmc_obj[I40IW_HMC_IW_XF].cnt = + roundup_pow_of_two(I40IW_MAX_WQ_ENTRIES * qpwanted); + hmc_info->hmc_obj[I40IW_HMC_IW_Q1].cnt = + roundup_pow_of_two(2 * I40IW_MAX_IRD_SIZE * qpwanted); hmc_info->hmc_obj[I40IW_HMC_IW_XFFL].cnt = hmc_info->hmc_obj[I40IW_HMC_IW_XF].cnt / hmc_fpm_misc->xf_block_size; hmc_info->hmc_obj[I40IW_HMC_IW_Q1FL].cnt = @@ -3945,16 +3946,10 @@ enum i40iw_status_code i40iw_config_fpm_values(struct i40iw_sc_dev *dev, u32 qp_ if ((loop_count > 1000) || ((!(loop_count % 10)) && (qpwanted > qpwantedoriginal * 2 / 3))) { - if (qpwanted > FPM_MULTIPLIER) { - qpwanted -= FPM_MULTIPLIER; - powerof2 = 1; - while (powerof2 < qpwanted) - powerof2 *= 2; - powerof2 /= 2; - qpwanted = powerof2; - } else { - qpwanted /= 2; - } + if (qpwanted > FPM_MULTIPLIER) + qpwanted = roundup_pow_of_two(qpwanted - + FPM_MULTIPLIER); + qpwanted >>= 1; } if (mrwanted > FPM_MULTIPLIER * 10) mrwanted -= FPM_MULTIPLIER * 10; diff --git a/drivers/infiniband/hw/i40iw/i40iw_d.h b/drivers/infiniband/hw/i40iw/i40iw_d.h index 029083cb81d5..4b65e4140bd7 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_d.h +++ b/drivers/infiniband/hw/i40iw/i40iw_d.h @@ -97,6 +97,7 @@ #define RDMA_OPCODE_MASK 0x0f #define RDMA_READ_REQ_OPCODE 1 #define Q2_BAD_FRAME_OFFSET 72 +#define Q2_FPSN_OFFSET 64 #define CQE_MAJOR_DRV 0x8000 #define I40IW_TERM_SENT 0x01 diff --git a/drivers/infiniband/hw/i40iw/i40iw_hw.c b/drivers/infiniband/hw/i40iw/i40iw_hw.c index e96bdafbcbb3..61540e14e4b9 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_hw.c +++ b/drivers/infiniband/hw/i40iw/i40iw_hw.c @@ -385,6 +385,8 @@ void i40iw_process_aeq(struct i40iw_device *iwdev) iwcq->ibcq.event_handler(&ibevent, iwcq->ibcq.cq_context); } break; + case I40IW_AE_LLP_DOUBT_REACHABILITY: + break; case I40IW_AE_PRIV_OPERATION_DENIED: case I40IW_AE_STAG_ZERO_INVALID: case I40IW_AE_IB_RREQ_AND_Q1_FULL: @@ -403,7 +405,6 @@ void i40iw_process_aeq(struct i40iw_device *iwdev) case I40IW_AE_LLP_SEGMENT_TOO_SMALL: case I40IW_AE_LLP_SYN_RECEIVED: case I40IW_AE_LLP_TOO_MANY_RETRIES: - case I40IW_AE_LLP_DOUBT_REACHABILITY: case I40IW_AE_LCE_QP_CATASTROPHIC: case I40IW_AE_LCE_FUNCTION_CATASTROPHIC: case I40IW_AE_LCE_CQ_CATASTROPHIC: diff --git a/drivers/infiniband/hw/i40iw/i40iw_main.c b/drivers/infiniband/hw/i40iw/i40iw_main.c index e824296713e2..b08862978de8 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_main.c +++ b/drivers/infiniband/hw/i40iw/i40iw_main.c @@ -99,6 +99,10 @@ static struct notifier_block i40iw_net_notifier = { .notifier_call = i40iw_net_event }; +static struct notifier_block i40iw_netdevice_notifier = { + .notifier_call = i40iw_netdevice_event +}; + /** * i40iw_find_i40e_handler - find a handler given a client info * @ldev: pointer to a client info @@ -483,6 +487,7 @@ static enum i40iw_status_code i40iw_create_hmc_objs(struct i40iw_device *iwdev, for (i = 0; i < IW_HMC_OBJ_TYPE_NUM; i++) { info.rsrc_type = iw_hmc_obj_types[i]; info.count = dev->hmc_info->hmc_obj[info.rsrc_type].cnt; + info.add_sd_cnt = 0; status = i40iw_create_hmc_obj_type(dev, &info); if (status) { i40iw_pr_err("create obj type %d status = %d\n", @@ -607,7 +612,7 @@ static enum i40iw_status_code i40iw_create_cqp(struct i40iw_device *iwdev) INIT_LIST_HEAD(&cqp->cqp_avail_reqs); INIT_LIST_HEAD(&cqp->cqp_pending_reqs); /* init the waitq of the cqp_requests and add them to the list */ - for (i = 0; i < I40IW_CQP_SW_SQSIZE_2048; i++) { + for (i = 0; i < sqsize; i++) { init_waitqueue_head(&cqp->cqp_requests[i].waitq); list_add_tail(&cqp->cqp_requests[i].list, &cqp->cqp_avail_reqs); } @@ -1285,7 +1290,7 @@ static void i40iw_wait_pe_ready(struct i40iw_hw *hw) __LINE__, statuscpu2); if ((statuscpu0 == 0x80) && (statuscpu1 == 0x80) && (statuscpu2 == 0x80)) break; /* SUCCESS */ - mdelay(1000); + msleep(1000); retrycount++; } while (retrycount < 14); i40iw_wr32(hw, 0xb4040, 0x4C104C5); @@ -1393,6 +1398,7 @@ static void i40iw_register_notifiers(void) register_inetaddr_notifier(&i40iw_inetaddr_notifier); register_inet6addr_notifier(&i40iw_inetaddr6_notifier); register_netevent_notifier(&i40iw_net_notifier); + register_netdevice_notifier(&i40iw_netdevice_notifier); } /** @@ -1404,6 +1410,7 @@ static void i40iw_unregister_notifiers(void) unregister_netevent_notifier(&i40iw_net_notifier); unregister_inetaddr_notifier(&i40iw_inetaddr_notifier); unregister_inet6addr_notifier(&i40iw_inetaddr6_notifier); + unregister_netdevice_notifier(&i40iw_netdevice_notifier); } /** @@ -1793,7 +1800,7 @@ static void i40iw_close(struct i40e_info *ldev, struct i40e_client *client, bool if (reset) iwdev->reset = true; - i40iw_cm_disconnect_all(iwdev); + i40iw_cm_teardown_connections(iwdev, NULL, NULL, true); destroy_workqueue(iwdev->virtchnl_wq); i40iw_deinit_device(iwdev); } diff --git a/drivers/infiniband/hw/i40iw/i40iw_puda.c b/drivers/infiniband/hw/i40iw/i40iw_puda.c index 796a815b53fd..f64b6700f43f 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_puda.c +++ b/drivers/infiniband/hw/i40iw/i40iw_puda.c @@ -1378,7 +1378,7 @@ static void i40iw_ieq_handle_exception(struct i40iw_puda_rsrc *ieq, u32 *hw_host_ctx = (u32 *)qp->hw_host_ctx; u32 rcv_wnd = hw_host_ctx[23]; /* first partial seq # in q2 */ - u32 fps = qp->q2_buf[16]; + u32 fps = *(u32 *)(qp->q2_buf + Q2_FPSN_OFFSET); struct list_head *rxlist = &pfpdu->rxlist; struct list_head *plist; diff --git a/drivers/infiniband/hw/i40iw/i40iw_uk.c b/drivers/infiniband/hw/i40iw/i40iw_uk.c index 3ec5389a81a1..8afa5a67a86b 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_uk.c +++ b/drivers/infiniband/hw/i40iw/i40iw_uk.c @@ -894,20 +894,6 @@ exit: } /** - * i40iw_qp_roundup - return round up QP WQ depth - * @wqdepth: WQ depth in quantas to round up - */ -static int i40iw_qp_round_up(u32 wqdepth) -{ - int scount = 1; - - for (wqdepth--; scount <= 16; scount *= 2) - wqdepth |= wqdepth >> scount; - - return ++wqdepth; -} - -/** * i40iw_get_wqe_shift - get shift count for maximum wqe size * @sge: Maximum Scatter Gather Elements wqe * @inline_data: Maximum inline data size @@ -934,7 +920,7 @@ void i40iw_get_wqe_shift(u32 sge, u32 inline_data, u8 *shift) */ enum i40iw_status_code i40iw_get_sqdepth(u32 sq_size, u8 shift, u32 *sqdepth) { - *sqdepth = i40iw_qp_round_up((sq_size << shift) + I40IW_SQ_RSVD); + *sqdepth = roundup_pow_of_two((sq_size << shift) + I40IW_SQ_RSVD); if (*sqdepth < (I40IW_QP_SW_MIN_WQSIZE << shift)) *sqdepth = I40IW_QP_SW_MIN_WQSIZE << shift; @@ -953,7 +939,7 @@ enum i40iw_status_code i40iw_get_sqdepth(u32 sq_size, u8 shift, u32 *sqdepth) */ enum i40iw_status_code i40iw_get_rqdepth(u32 rq_size, u8 shift, u32 *rqdepth) { - *rqdepth = i40iw_qp_round_up((rq_size << shift) + I40IW_RQ_RSVD); + *rqdepth = roundup_pow_of_two((rq_size << shift) + I40IW_RQ_RSVD); if (*rqdepth < (I40IW_QP_SW_MIN_WQSIZE << shift)) *rqdepth = I40IW_QP_SW_MIN_WQSIZE << shift; diff --git a/drivers/infiniband/hw/i40iw/i40iw_user.h b/drivers/infiniband/hw/i40iw/i40iw_user.h index e73efc59a0ab..5467c6fdad03 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_user.h +++ b/drivers/infiniband/hw/i40iw/i40iw_user.h @@ -72,7 +72,7 @@ enum i40iw_device_capabilities_const { I40IW_MAX_SQ_PAYLOAD_SIZE = 2145386496, I40IW_MAX_INLINE_DATA_SIZE = 48, I40IW_MAX_PUSHMODE_INLINE_DATA_SIZE = 48, - I40IW_MAX_IRD_SIZE = 63, + I40IW_MAX_IRD_SIZE = 64, I40IW_MAX_ORD_SIZE = 127, I40IW_MAX_WQ_ENTRIES = 2048, I40IW_Q2_BUFFER_SIZE = (248 + 100), diff --git a/drivers/infiniband/hw/i40iw/i40iw_utils.c b/drivers/infiniband/hw/i40iw/i40iw_utils.c index 8845dba7c438..ddc1056b0b4e 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_utils.c +++ b/drivers/infiniband/hw/i40iw/i40iw_utils.c @@ -137,7 +137,7 @@ inline u32 i40iw_rd32(struct i40iw_hw *hw, u32 reg) } /** - * i40iw_inetaddr_event - system notifier for netdev events + * i40iw_inetaddr_event - system notifier for ipv4 addr events * @notfier: not used * @event: event for notifier * @ptr: if address @@ -200,7 +200,7 @@ int i40iw_inetaddr_event(struct notifier_block *notifier, } /** - * i40iw_inet6addr_event - system notifier for ipv6 netdev events + * i40iw_inet6addr_event - system notifier for ipv6 addr events * @notfier: not used * @event: event for notifier * @ptr: if address @@ -252,7 +252,7 @@ int i40iw_inet6addr_event(struct notifier_block *notifier, } /** - * i40iw_net_event - system notifier for net events + * i40iw_net_event - system notifier for netevents * @notfier: not used * @event: event for notifier * @ptr: neighbor @@ -297,6 +297,50 @@ int i40iw_net_event(struct notifier_block *notifier, unsigned long event, void * } /** + * i40iw_netdevice_event - system notifier for netdev events + * @notfier: not used + * @event: event for notifier + * @ptr: netdev + */ +int i40iw_netdevice_event(struct notifier_block *notifier, + unsigned long event, + void *ptr) +{ + struct net_device *event_netdev; + struct net_device *netdev; + struct i40iw_device *iwdev; + struct i40iw_handler *hdl; + + event_netdev = netdev_notifier_info_to_dev(ptr); + + hdl = i40iw_find_netdev(event_netdev); + if (!hdl) + return NOTIFY_DONE; + + iwdev = &hdl->device; + if (iwdev->init_state < RDMA_DEV_REGISTERED || iwdev->closing) + return NOTIFY_DONE; + + netdev = iwdev->ldev->netdev; + if (netdev != event_netdev) + return NOTIFY_DONE; + + iwdev->iw_status = 1; + + switch (event) { + case NETDEV_DOWN: + iwdev->iw_status = 0; + /* Fall through */ + case NETDEV_UP: + i40iw_port_ibevent(iwdev); + break; + default: + break; + } + return NOTIFY_DONE; +} + +/** * i40iw_get_cqp_request - get cqp struct * @cqp: device cqp ptr * @wait: cqp to be used in wait mode diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c index 8c8a16791a3f..5695ce53fddb 100644 --- a/drivers/infiniband/hw/mlx4/main.c +++ b/drivers/infiniband/hw/mlx4/main.c @@ -589,6 +589,7 @@ static int mlx4_ib_query_device(struct ib_device *ibdev, if (props->rss_caps.supported_qpts) { resp.rss_caps.rx_hash_function = MLX4_IB_RX_HASH_FUNC_TOEPLITZ; + resp.rss_caps.rx_hash_fields_mask = MLX4_IB_RX_HASH_SRC_IPV4 | MLX4_IB_RX_HASH_DST_IPV4 | @@ -598,6 +599,11 @@ static int mlx4_ib_query_device(struct ib_device *ibdev, MLX4_IB_RX_HASH_DST_PORT_TCP | MLX4_IB_RX_HASH_SRC_PORT_UDP | MLX4_IB_RX_HASH_DST_PORT_UDP; + + if (dev->dev->caps.tunnel_offload_mode == + MLX4_TUNNEL_OFFLOAD_MODE_VXLAN) + resp.rss_caps.rx_hash_fields_mask |= + MLX4_IB_RX_HASH_INNER; } } diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c index caf490ab24c8..f045491f2c14 100644 --- a/drivers/infiniband/hw/mlx4/qp.c +++ b/drivers/infiniband/hw/mlx4/qp.c @@ -734,10 +734,24 @@ static int set_qp_rss(struct mlx4_ib_dev *dev, struct mlx4_ib_rss *rss_ctx, return (-EOPNOTSUPP); } + if (ucmd->rx_hash_fields_mask & MLX4_IB_RX_HASH_INNER) { + if (dev->dev->caps.tunnel_offload_mode == + MLX4_TUNNEL_OFFLOAD_MODE_VXLAN) { + /* + * Hash according to inner headers if exist, otherwise + * according to outer headers. + */ + rss_ctx->flags |= MLX4_RSS_BY_INNER_HEADERS_IPONLY; + } else { + pr_debug("RSS Hash for inner headers isn't supported\n"); + return (-EOPNOTSUPP); + } + } + return 0; } -static int create_qp_rss(struct mlx4_ib_dev *dev, struct ib_pd *ibpd, +static int create_qp_rss(struct mlx4_ib_dev *dev, struct ib_qp_init_attr *init_attr, struct mlx4_ib_create_qp_rss *ucmd, struct mlx4_ib_qp *qp) @@ -860,7 +874,7 @@ static struct ib_qp *_mlx4_ib_create_qp_rss(struct ib_pd *pd, qp->pri.vid = 0xFFFF; qp->alt.vid = 0xFFFF; - err = create_qp_rss(to_mdev(pd->device), pd, init_attr, &ucmd, qp); + err = create_qp_rss(to_mdev(pd->device), init_attr, &ucmd, qp); if (err) { kfree(qp); return ERR_PTR(err); @@ -1836,6 +1850,8 @@ static int _mlx4_set_path(struct mlx4_ib_dev *dev, mlx4_ib_gid_index_to_real_index(dev, port, grh->sgid_index); + if (real_sgid_index < 0) + return real_sgid_index; if (real_sgid_index >= dev->dev->caps.gid_table_len[port]) { pr_err("sgid_index (%u) too large. max is %d\n", real_sgid_index, dev->dev->caps.gid_table_len[port] - 1); diff --git a/drivers/infiniband/hw/mlx5/cong.c b/drivers/infiniband/hw/mlx5/cong.c index 2d32b519bb61..985fa2637390 100644 --- a/drivers/infiniband/hw/mlx5/cong.c +++ b/drivers/infiniband/hw/mlx5/cong.c @@ -247,21 +247,30 @@ static void mlx5_ib_set_cc_param_mask_val(void *field, int offset, } } -static int mlx5_ib_get_cc_params(struct mlx5_ib_dev *dev, int offset, u32 *var) +static int mlx5_ib_get_cc_params(struct mlx5_ib_dev *dev, u8 port_num, + int offset, u32 *var) { int outlen = MLX5_ST_SZ_BYTES(query_cong_params_out); void *out; void *field; int err; enum mlx5_ib_cong_node_type node; + struct mlx5_core_dev *mdev; + + /* Takes a 1-based port number */ + mdev = mlx5_ib_get_native_port_mdev(dev, port_num + 1, NULL); + if (!mdev) + return -ENODEV; out = kvzalloc(outlen, GFP_KERNEL); - if (!out) - return -ENOMEM; + if (!out) { + err = -ENOMEM; + goto alloc_err; + } node = mlx5_ib_param_to_node(offset); - err = mlx5_cmd_query_cong_params(dev->mdev, node, out, outlen); + err = mlx5_cmd_query_cong_params(mdev, node, out, outlen); if (err) goto free; @@ -270,21 +279,32 @@ static int mlx5_ib_get_cc_params(struct mlx5_ib_dev *dev, int offset, u32 *var) free: kvfree(out); +alloc_err: + mlx5_ib_put_native_port_mdev(dev, port_num + 1); return err; } -static int mlx5_ib_set_cc_params(struct mlx5_ib_dev *dev, int offset, u32 var) +static int mlx5_ib_set_cc_params(struct mlx5_ib_dev *dev, u8 port_num, + int offset, u32 var) { int inlen = MLX5_ST_SZ_BYTES(modify_cong_params_in); void *in; void *field; enum mlx5_ib_cong_node_type node; + struct mlx5_core_dev *mdev; u32 attr_mask = 0; int err; + /* Takes a 1-based port number */ + mdev = mlx5_ib_get_native_port_mdev(dev, port_num + 1, NULL); + if (!mdev) + return -ENODEV; + in = kvzalloc(inlen, GFP_KERNEL); - if (!in) - return -ENOMEM; + if (!in) { + err = -ENOMEM; + goto alloc_err; + } MLX5_SET(modify_cong_params_in, in, opcode, MLX5_CMD_OP_MODIFY_CONG_PARAMS); @@ -299,8 +319,10 @@ static int mlx5_ib_set_cc_params(struct mlx5_ib_dev *dev, int offset, u32 var) MLX5_SET(field_select_r_roce_rp, field, field_select_r_roce_rp, attr_mask); - err = mlx5_cmd_modify_cong_params(dev->mdev, in, inlen); + err = mlx5_cmd_modify_cong_params(mdev, in, inlen); kvfree(in); +alloc_err: + mlx5_ib_put_native_port_mdev(dev, port_num + 1); return err; } @@ -324,7 +346,7 @@ static ssize_t set_param(struct file *filp, const char __user *buf, if (kstrtou32(lbuf, 0, &var)) return -EINVAL; - ret = mlx5_ib_set_cc_params(param->dev, offset, var); + ret = mlx5_ib_set_cc_params(param->dev, param->port_num, offset, var); return ret ? ret : count; } @@ -340,7 +362,7 @@ static ssize_t get_param(struct file *filp, char __user *buf, size_t count, if (*pos) return 0; - ret = mlx5_ib_get_cc_params(param->dev, offset, &var); + ret = mlx5_ib_get_cc_params(param->dev, param->port_num, offset, &var); if (ret) return ret; @@ -362,44 +384,51 @@ static const struct file_operations dbg_cc_fops = { .read = get_param, }; -void mlx5_ib_cleanup_cong_debugfs(struct mlx5_ib_dev *dev) +void mlx5_ib_cleanup_cong_debugfs(struct mlx5_ib_dev *dev, u8 port_num) { if (!mlx5_debugfs_root || - !dev->dbg_cc_params || - !dev->dbg_cc_params->root) + !dev->port[port_num].dbg_cc_params || + !dev->port[port_num].dbg_cc_params->root) return; - debugfs_remove_recursive(dev->dbg_cc_params->root); - kfree(dev->dbg_cc_params); - dev->dbg_cc_params = NULL; + debugfs_remove_recursive(dev->port[port_num].dbg_cc_params->root); + kfree(dev->port[port_num].dbg_cc_params); + dev->port[port_num].dbg_cc_params = NULL; } -int mlx5_ib_init_cong_debugfs(struct mlx5_ib_dev *dev) +int mlx5_ib_init_cong_debugfs(struct mlx5_ib_dev *dev, u8 port_num) { struct mlx5_ib_dbg_cc_params *dbg_cc_params; + struct mlx5_core_dev *mdev; int i; if (!mlx5_debugfs_root) goto out; - if (!MLX5_CAP_GEN(dev->mdev, cc_query_allowed) || - !MLX5_CAP_GEN(dev->mdev, cc_modify_allowed)) + /* Takes a 1-based port number */ + mdev = mlx5_ib_get_native_port_mdev(dev, port_num + 1, NULL); + if (!mdev) goto out; + if (!MLX5_CAP_GEN(mdev, cc_query_allowed) || + !MLX5_CAP_GEN(mdev, cc_modify_allowed)) + goto put_mdev; + dbg_cc_params = kzalloc(sizeof(*dbg_cc_params), GFP_KERNEL); if (!dbg_cc_params) - goto out; + goto err; - dev->dbg_cc_params = dbg_cc_params; + dev->port[port_num].dbg_cc_params = dbg_cc_params; dbg_cc_params->root = debugfs_create_dir("cc_params", - dev->mdev->priv.dbg_root); + mdev->priv.dbg_root); if (!dbg_cc_params->root) goto err; for (i = 0; i < MLX5_IB_DBG_CC_MAX; i++) { dbg_cc_params->params[i].offset = i; dbg_cc_params->params[i].dev = dev; + dbg_cc_params->params[i].port_num = port_num; dbg_cc_params->params[i].dentry = debugfs_create_file(mlx5_ib_dbg_cc_name[i], 0600, dbg_cc_params->root, @@ -408,11 +437,17 @@ int mlx5_ib_init_cong_debugfs(struct mlx5_ib_dev *dev) if (!dbg_cc_params->params[i].dentry) goto err; } -out: return 0; + +put_mdev: + mlx5_ib_put_native_port_mdev(dev, port_num + 1); +out: + return 0; err: mlx5_ib_warn(dev, "cong debugfs failure\n"); - mlx5_ib_cleanup_cong_debugfs(dev); + mlx5_ib_cleanup_cong_debugfs(dev, port_num); + mlx5_ib_put_native_port_mdev(dev, port_num + 1); + /* * We don't want to fail driver if debugfs failed to initialize, * so we are not forwarding error to the user. diff --git a/drivers/infiniband/hw/mlx5/mad.c b/drivers/infiniband/hw/mlx5/mad.c index 1003b0133a49..32a9e9228b13 100644 --- a/drivers/infiniband/hw/mlx5/mad.c +++ b/drivers/infiniband/hw/mlx5/mad.c @@ -197,10 +197,9 @@ static void pma_cnt_assign(struct ib_pma_portcounters *pma_cnt, vl_15_dropped); } -static int process_pma_cmd(struct ib_device *ibdev, u8 port_num, +static int process_pma_cmd(struct mlx5_core_dev *mdev, u8 port_num, const struct ib_mad *in_mad, struct ib_mad *out_mad) { - struct mlx5_ib_dev *dev = to_mdev(ibdev); int err; void *out_cnt; @@ -222,7 +221,7 @@ static int process_pma_cmd(struct ib_device *ibdev, u8 port_num, if (!out_cnt) return IB_MAD_RESULT_FAILURE; - err = mlx5_core_query_vport_counter(dev->mdev, 0, 0, + err = mlx5_core_query_vport_counter(mdev, 0, 0, port_num, out_cnt, sz); if (!err) pma_cnt_ext_assign(pma_cnt_ext, out_cnt); @@ -235,7 +234,7 @@ static int process_pma_cmd(struct ib_device *ibdev, u8 port_num, if (!out_cnt) return IB_MAD_RESULT_FAILURE; - err = mlx5_core_query_ib_ppcnt(dev->mdev, port_num, + err = mlx5_core_query_ib_ppcnt(mdev, port_num, out_cnt, sz); if (!err) pma_cnt_assign(pma_cnt, out_cnt); @@ -255,9 +254,11 @@ int mlx5_ib_process_mad(struct ib_device *ibdev, int mad_flags, u8 port_num, u16 *out_mad_pkey_index) { struct mlx5_ib_dev *dev = to_mdev(ibdev); - struct mlx5_core_dev *mdev = dev->mdev; const struct ib_mad *in_mad = (const struct ib_mad *)in; struct ib_mad *out_mad = (struct ib_mad *)out; + struct mlx5_core_dev *mdev; + u8 mdev_port_num; + int ret; if (WARN_ON_ONCE(in_mad_size != sizeof(*in_mad) || *out_mad_size != sizeof(*out_mad))) @@ -265,14 +266,20 @@ int mlx5_ib_process_mad(struct ib_device *ibdev, int mad_flags, u8 port_num, memset(out_mad->data, 0, sizeof(out_mad->data)); + mdev = mlx5_ib_get_native_port_mdev(dev, port_num, &mdev_port_num); + if (!mdev) + return IB_MAD_RESULT_FAILURE; + if (MLX5_CAP_GEN(mdev, vport_counters) && in_mad->mad_hdr.mgmt_class == IB_MGMT_CLASS_PERF_MGMT && in_mad->mad_hdr.method == IB_MGMT_METHOD_GET) { - return process_pma_cmd(ibdev, port_num, in_mad, out_mad); + ret = process_pma_cmd(mdev, mdev_port_num, in_mad, out_mad); } else { - return process_mad(ibdev, mad_flags, port_num, in_wc, in_grh, + ret = process_mad(ibdev, mad_flags, port_num, in_wc, in_grh, in_mad, out_mad); } + mlx5_ib_put_native_port_mdev(dev, port_num); + return ret; } int mlx5_query_ext_port_caps(struct mlx5_ib_dev *dev, u8 port) @@ -519,7 +526,7 @@ int mlx5_query_mad_ifc_port(struct ib_device *ibdev, u8 port, int ext_active_speed; int err = -ENOMEM; - if (port < 1 || port > MLX5_CAP_GEN(mdev, num_ports)) { + if (port < 1 || port > dev->num_ports) { mlx5_ib_warn(dev, "invalid port number %d\n", port); return -EINVAL; } diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c index 8ac50de2b242..5d6fba986fa5 100644 --- a/drivers/infiniband/hw/mlx5/main.c +++ b/drivers/infiniband/hw/mlx5/main.c @@ -50,16 +50,14 @@ #include <rdma/ib_cache.h> #include <linux/mlx5/port.h> #include <linux/mlx5/vport.h> +#include <linux/mlx5/fs.h> #include <linux/list.h> #include <rdma/ib_smi.h> #include <rdma/ib_umem.h> #include <linux/in.h> #include <linux/etherdevice.h> -#include <linux/mlx5/fs.h> -#include <linux/mlx5/vport.h> #include "mlx5_ib.h" #include "cmd.h" -#include <linux/mlx5/vport.h> #define DRIVER_NAME "mlx5_ib" #define DRIVER_VERSION "5.0-0" @@ -72,10 +70,36 @@ static char mlx5_version[] = DRIVER_NAME ": Mellanox Connect-IB Infiniband driver v" DRIVER_VERSION "\n"; +struct mlx5_ib_event_work { + struct work_struct work; + struct mlx5_core_dev *dev; + void *context; + enum mlx5_dev_event event; + unsigned long param; +}; + enum { MLX5_ATOMIC_SIZE_QP_8BYTES = 1 << 3, }; +static struct workqueue_struct *mlx5_ib_event_wq; +static LIST_HEAD(mlx5_ib_unaffiliated_port_list); +static LIST_HEAD(mlx5_ib_dev_list); +/* + * This mutex should be held when accessing either of the above lists + */ +static DEFINE_MUTEX(mlx5_ib_multiport_mutex); + +struct mlx5_ib_dev *mlx5_ib_get_ibdev_from_mpi(struct mlx5_ib_multiport_info *mpi) +{ + struct mlx5_ib_dev *dev; + + mutex_lock(&mlx5_ib_multiport_mutex); + dev = mpi->ibdev; + mutex_unlock(&mlx5_ib_multiport_mutex); + return dev; +} + static enum rdma_link_layer mlx5_port_type_cap_to_rdma_ll(int port_type_cap) { @@ -115,24 +139,32 @@ static int get_port_state(struct ib_device *ibdev, static int mlx5_netdev_event(struct notifier_block *this, unsigned long event, void *ptr) { + struct mlx5_roce *roce = container_of(this, struct mlx5_roce, nb); struct net_device *ndev = netdev_notifier_info_to_dev(ptr); - struct mlx5_ib_dev *ibdev = container_of(this, struct mlx5_ib_dev, - roce.nb); + u8 port_num = roce->native_port_num; + struct mlx5_core_dev *mdev; + struct mlx5_ib_dev *ibdev; + + ibdev = roce->dev; + mdev = mlx5_ib_get_native_port_mdev(ibdev, port_num, NULL); + if (!mdev) + return NOTIFY_DONE; switch (event) { case NETDEV_REGISTER: case NETDEV_UNREGISTER: - write_lock(&ibdev->roce.netdev_lock); - if (ndev->dev.parent == &ibdev->mdev->pdev->dev) - ibdev->roce.netdev = (event == NETDEV_UNREGISTER) ? - NULL : ndev; - write_unlock(&ibdev->roce.netdev_lock); + write_lock(&roce->netdev_lock); + + if (ndev->dev.parent == &mdev->pdev->dev) + roce->netdev = (event == NETDEV_UNREGISTER) ? + NULL : ndev; + write_unlock(&roce->netdev_lock); break; case NETDEV_CHANGE: case NETDEV_UP: case NETDEV_DOWN: { - struct net_device *lag_ndev = mlx5_lag_get_roce_netdev(ibdev->mdev); + struct net_device *lag_ndev = mlx5_lag_get_roce_netdev(mdev); struct net_device *upper = NULL; if (lag_ndev) { @@ -140,27 +172,28 @@ static int mlx5_netdev_event(struct notifier_block *this, dev_put(lag_ndev); } - if ((upper == ndev || (!upper && ndev == ibdev->roce.netdev)) + if ((upper == ndev || (!upper && ndev == roce->netdev)) && ibdev->ib_active) { struct ib_event ibev = { }; enum ib_port_state port_state; - if (get_port_state(&ibdev->ib_dev, 1, &port_state)) - return NOTIFY_DONE; + if (get_port_state(&ibdev->ib_dev, port_num, + &port_state)) + goto done; - if (ibdev->roce.last_port_state == port_state) - return NOTIFY_DONE; + if (roce->last_port_state == port_state) + goto done; - ibdev->roce.last_port_state = port_state; + roce->last_port_state = port_state; ibev.device = &ibdev->ib_dev; if (port_state == IB_PORT_DOWN) ibev.event = IB_EVENT_PORT_ERR; else if (port_state == IB_PORT_ACTIVE) ibev.event = IB_EVENT_PORT_ACTIVE; else - return NOTIFY_DONE; + goto done; - ibev.element.port_num = 1; + ibev.element.port_num = port_num; ib_dispatch_event(&ibev); } break; @@ -169,7 +202,8 @@ static int mlx5_netdev_event(struct notifier_block *this, default: break; } - +done: + mlx5_ib_put_native_port_mdev(ibdev, port_num); return NOTIFY_DONE; } @@ -178,22 +212,88 @@ static struct net_device *mlx5_ib_get_netdev(struct ib_device *device, { struct mlx5_ib_dev *ibdev = to_mdev(device); struct net_device *ndev; + struct mlx5_core_dev *mdev; + + mdev = mlx5_ib_get_native_port_mdev(ibdev, port_num, NULL); + if (!mdev) + return NULL; - ndev = mlx5_lag_get_roce_netdev(ibdev->mdev); + ndev = mlx5_lag_get_roce_netdev(mdev); if (ndev) - return ndev; + goto out; /* Ensure ndev does not disappear before we invoke dev_hold() */ - read_lock(&ibdev->roce.netdev_lock); - ndev = ibdev->roce.netdev; + read_lock(&ibdev->roce[port_num - 1].netdev_lock); + ndev = ibdev->roce[port_num - 1].netdev; if (ndev) dev_hold(ndev); - read_unlock(&ibdev->roce.netdev_lock); + read_unlock(&ibdev->roce[port_num - 1].netdev_lock); +out: + mlx5_ib_put_native_port_mdev(ibdev, port_num); return ndev; } +struct mlx5_core_dev *mlx5_ib_get_native_port_mdev(struct mlx5_ib_dev *ibdev, + u8 ib_port_num, + u8 *native_port_num) +{ + enum rdma_link_layer ll = mlx5_ib_port_link_layer(&ibdev->ib_dev, + ib_port_num); + struct mlx5_core_dev *mdev = NULL; + struct mlx5_ib_multiport_info *mpi; + struct mlx5_ib_port *port; + + if (native_port_num) + *native_port_num = 1; + + if (!mlx5_core_mp_enabled(ibdev->mdev) || ll != IB_LINK_LAYER_ETHERNET) + return ibdev->mdev; + + port = &ibdev->port[ib_port_num - 1]; + if (!port) + return NULL; + + spin_lock(&port->mp.mpi_lock); + mpi = ibdev->port[ib_port_num - 1].mp.mpi; + if (mpi && !mpi->unaffiliate) { + mdev = mpi->mdev; + /* If it's the master no need to refcount, it'll exist + * as long as the ib_dev exists. + */ + if (!mpi->is_master) + mpi->mdev_refcnt++; + } + spin_unlock(&port->mp.mpi_lock); + + return mdev; +} + +void mlx5_ib_put_native_port_mdev(struct mlx5_ib_dev *ibdev, u8 port_num) +{ + enum rdma_link_layer ll = mlx5_ib_port_link_layer(&ibdev->ib_dev, + port_num); + struct mlx5_ib_multiport_info *mpi; + struct mlx5_ib_port *port; + + if (!mlx5_core_mp_enabled(ibdev->mdev) || ll != IB_LINK_LAYER_ETHERNET) + return; + + port = &ibdev->port[port_num - 1]; + + spin_lock(&port->mp.mpi_lock); + mpi = ibdev->port[port_num - 1].mp.mpi; + if (mpi->is_master) + goto out; + + mpi->mdev_refcnt--; + if (mpi->unaffiliate) + complete(&mpi->unref_comp); +out: + spin_unlock(&port->mp.mpi_lock); +} + static int translate_eth_proto_oper(u32 eth_proto_oper, u8 *active_speed, u8 *active_width) { @@ -259,16 +359,30 @@ static int mlx5_query_port_roce(struct ib_device *device, u8 port_num, struct mlx5_core_dev *mdev = dev->mdev; struct net_device *ndev, *upper; enum ib_mtu ndev_ib_mtu; + bool put_mdev = true; u16 qkey_viol_cntr; u32 eth_prot_oper; + u8 mdev_port_num; int err; + mdev = mlx5_ib_get_native_port_mdev(dev, port_num, &mdev_port_num); + if (!mdev) { + /* This means the port isn't affiliated yet. Get the + * info for the master port instead. + */ + put_mdev = false; + mdev = dev->mdev; + mdev_port_num = 1; + port_num = 1; + } + /* Possible bad flows are checked before filling out props so in case * of an error it will still be zeroed out. */ - err = mlx5_query_port_eth_proto_oper(mdev, ð_prot_oper, port_num); + err = mlx5_query_port_eth_proto_oper(mdev, ð_prot_oper, + mdev_port_num); if (err) - return err; + goto out; translate_eth_proto_oper(eth_prot_oper, &props->active_speed, &props->active_width); @@ -284,12 +398,16 @@ static int mlx5_query_port_roce(struct ib_device *device, u8 port_num, props->state = IB_PORT_DOWN; props->phys_state = 3; - mlx5_query_nic_vport_qkey_viol_cntr(dev->mdev, &qkey_viol_cntr); + mlx5_query_nic_vport_qkey_viol_cntr(mdev, &qkey_viol_cntr); props->qkey_viol_cntr = qkey_viol_cntr; + /* If this is a stub query for an unaffiliated port stop here */ + if (!put_mdev) + goto out; + ndev = mlx5_ib_get_netdev(device, port_num); if (!ndev) - return 0; + goto out; if (mlx5_lag_is_active(dev->mdev)) { rcu_read_lock(); @@ -312,7 +430,10 @@ static int mlx5_query_port_roce(struct ib_device *device, u8 port_num, dev_put(ndev); props->active_mtu = min(props->max_mtu, ndev_ib_mtu); - return 0; +out: + if (put_mdev) + mlx5_ib_put_native_port_mdev(dev, port_num); + return err; } static int set_roce_addr(struct mlx5_ib_dev *dev, u8 port_num, @@ -354,7 +475,7 @@ static int set_roce_addr(struct mlx5_ib_dev *dev, u8 port_num, return mlx5_core_roce_gid_set(dev->mdev, index, roce_version, roce_l3_type, gid->raw, mac, vlan, - vlan_id); + vlan_id, port_num); } static int mlx5_ib_add_gid(struct ib_device *device, u8 port_num, @@ -438,11 +559,11 @@ static int mlx5_get_vport_access_method(struct ib_device *ibdev) } static void get_atomic_caps(struct mlx5_ib_dev *dev, + u8 atomic_size_qp, struct ib_device_attr *props) { u8 tmp; u8 atomic_operations = MLX5_CAP_ATOMIC(dev->mdev, atomic_operations); - u8 atomic_size_qp = MLX5_CAP_ATOMIC(dev->mdev, atomic_size_qp); u8 atomic_req_8B_endianness_mode = MLX5_CAP_ATOMIC(dev->mdev, atomic_req_8B_endianness_mode); @@ -459,6 +580,29 @@ static void get_atomic_caps(struct mlx5_ib_dev *dev, } } +static void get_atomic_caps_qp(struct mlx5_ib_dev *dev, + struct ib_device_attr *props) +{ + u8 atomic_size_qp = MLX5_CAP_ATOMIC(dev->mdev, atomic_size_qp); + + get_atomic_caps(dev, atomic_size_qp, props); +} + +static void get_atomic_caps_dc(struct mlx5_ib_dev *dev, + struct ib_device_attr *props) +{ + u8 atomic_size_qp = MLX5_CAP_ATOMIC(dev->mdev, atomic_size_dc); + + get_atomic_caps(dev, atomic_size_qp, props); +} + +bool mlx5_ib_dc_atomic_is_supported(struct mlx5_ib_dev *dev) +{ + struct ib_device_attr props = {}; + + get_atomic_caps_dc(dev, &props); + return (props.atomic_cap == IB_ATOMIC_HCA) ? true : false; +} static int mlx5_query_system_image_guid(struct ib_device *ibdev, __be64 *sys_image_guid) { @@ -587,6 +731,7 @@ static int mlx5_ib_query_device(struct ib_device *ibdev, int max_rq_sg; int max_sq_sg; u64 min_page_size = 1ull << MLX5_CAP_GEN(mdev, log_pg_sz); + bool raw_support = !mlx5_core_mp_enabled(mdev); struct mlx5_ib_query_device_resp resp = {}; size_t resp_len; u64 max_tso; @@ -650,7 +795,7 @@ static int mlx5_ib_query_device(struct ib_device *ibdev, if (MLX5_CAP_GEN(mdev, block_lb_mc)) props->device_cap_flags |= IB_DEVICE_BLOCK_MULTICAST_LOOPBACK; - if (MLX5_CAP_GEN(dev->mdev, eth_net_offloads)) { + if (MLX5_CAP_GEN(dev->mdev, eth_net_offloads) && raw_support) { if (MLX5_CAP_ETH(mdev, csum_cap)) { /* Legacy bit to support old userspace libraries */ props->device_cap_flags |= IB_DEVICE_RAW_IP_CSUM; @@ -682,7 +827,8 @@ static int mlx5_ib_query_device(struct ib_device *ibdev, MLX5_RX_HASH_SRC_PORT_TCP | MLX5_RX_HASH_DST_PORT_TCP | MLX5_RX_HASH_SRC_PORT_UDP | - MLX5_RX_HASH_DST_PORT_UDP; + MLX5_RX_HASH_DST_PORT_UDP | + MLX5_RX_HASH_INNER; resp.response_length += sizeof(resp.rss_caps); } } else { @@ -698,7 +844,8 @@ static int mlx5_ib_query_device(struct ib_device *ibdev, } if (MLX5_CAP_GEN(dev->mdev, rq_delay_drop) && - MLX5_CAP_GEN(dev->mdev, general_notification_event)) + MLX5_CAP_GEN(dev->mdev, general_notification_event) && + raw_support) props->raw_packet_caps |= IB_RAW_PACKET_CAP_DELAY_DROP; if (MLX5_CAP_GEN(mdev, ipoib_enhanced_offloads) && @@ -706,7 +853,8 @@ static int mlx5_ib_query_device(struct ib_device *ibdev, props->device_cap_flags |= IB_DEVICE_UD_IP_CSUM; if (MLX5_CAP_GEN(dev->mdev, eth_net_offloads) && - MLX5_CAP_ETH(dev->mdev, scatter_fcs)) { + MLX5_CAP_ETH(dev->mdev, scatter_fcs) && + raw_support) { /* Legacy bit to support old userspace libraries */ props->device_cap_flags |= IB_DEVICE_RAW_SCATTER_FCS; props->raw_packet_caps |= IB_RAW_PACKET_CAP_SCATTER_FCS; @@ -746,7 +894,7 @@ static int mlx5_ib_query_device(struct ib_device *ibdev, props->max_srq_sge = max_rq_sg - 1; props->max_fast_reg_page_list_len = 1 << MLX5_CAP_GEN(mdev, log_max_klm_list_size); - get_atomic_caps(dev, props); + get_atomic_caps_qp(dev, props); props->masked_atomic_cap = IB_ATOMIC_NONE; props->max_mcast_grp = 1 << MLX5_CAP_GEN(mdev, log_max_mcg); props->max_mcast_qp_attach = MLX5_CAP_GEN(mdev, max_qp_mcg); @@ -770,7 +918,7 @@ static int mlx5_ib_query_device(struct ib_device *ibdev, props->device_cap_flags |= IB_DEVICE_VIRTUAL_FUNCTION; if (mlx5_ib_port_link_layer(ibdev, 1) == - IB_LINK_LAYER_ETHERNET) { + IB_LINK_LAYER_ETHERNET && raw_support) { props->rss_caps.max_rwq_indirection_tables = 1 << MLX5_CAP_GEN(dev->mdev, log_max_rqt); props->rss_caps.max_rwq_indirection_table_size = @@ -807,7 +955,8 @@ static int mlx5_ib_query_device(struct ib_device *ibdev, resp.response_length += sizeof(resp.cqe_comp_caps); } - if (field_avail(typeof(resp), packet_pacing_caps, uhw->outlen)) { + if (field_avail(typeof(resp), packet_pacing_caps, uhw->outlen) && + raw_support) { if (MLX5_CAP_QOS(mdev, packet_pacing) && MLX5_CAP_GEN(mdev, qos)) { resp.packet_pacing_caps.qp_rate_limit_max = @@ -866,7 +1015,8 @@ static int mlx5_ib_query_device(struct ib_device *ibdev, } } - if (field_avail(typeof(resp), striding_rq_caps, uhw->outlen)) { + if (field_avail(typeof(resp), striding_rq_caps, uhw->outlen) && + raw_support) { resp.response_length += sizeof(resp.striding_rq_caps); if (MLX5_CAP_GEN(mdev, striding_rq)) { resp.striding_rq_caps.min_single_stride_log_num_of_bytes = @@ -1097,7 +1247,22 @@ int mlx5_ib_query_port(struct ib_device *ibdev, u8 port, } if (!ret && props) { - count = mlx5_core_reserved_gids_count(to_mdev(ibdev)->mdev); + struct mlx5_ib_dev *dev = to_mdev(ibdev); + struct mlx5_core_dev *mdev; + bool put_mdev = true; + + mdev = mlx5_ib_get_native_port_mdev(dev, port, NULL); + if (!mdev) { + /* If the port isn't affiliated yet query the master. + * The master and slave will have the same values. + */ + mdev = dev->mdev; + port = 1; + put_mdev = false; + } + count = mlx5_core_reserved_gids_count(mdev); + if (put_mdev) + mlx5_ib_put_native_port_mdev(dev, port); props->gid_tbl_len -= count; } return ret; @@ -1122,20 +1287,43 @@ static int mlx5_ib_query_gid(struct ib_device *ibdev, u8 port, int index, } -static int mlx5_ib_query_pkey(struct ib_device *ibdev, u8 port, u16 index, - u16 *pkey) +static int mlx5_query_hca_nic_pkey(struct ib_device *ibdev, u8 port, + u16 index, u16 *pkey) { struct mlx5_ib_dev *dev = to_mdev(ibdev); - struct mlx5_core_dev *mdev = dev->mdev; + struct mlx5_core_dev *mdev; + bool put_mdev = true; + u8 mdev_port_num; + int err; + mdev = mlx5_ib_get_native_port_mdev(dev, port, &mdev_port_num); + if (!mdev) { + /* The port isn't affiliated yet, get the PKey from the master + * port. For RoCE the PKey tables will be the same. + */ + put_mdev = false; + mdev = dev->mdev; + mdev_port_num = 1; + } + + err = mlx5_query_hca_vport_pkey(mdev, 0, mdev_port_num, 0, + index, pkey); + if (put_mdev) + mlx5_ib_put_native_port_mdev(dev, port); + + return err; +} + +static int mlx5_ib_query_pkey(struct ib_device *ibdev, u8 port, u16 index, + u16 *pkey) +{ switch (mlx5_get_vport_access_method(ibdev)) { case MLX5_VPORT_ACCESS_METHOD_MAD: return mlx5_query_mad_ifc_pkey(ibdev, port, index, pkey); case MLX5_VPORT_ACCESS_METHOD_HCA: case MLX5_VPORT_ACCESS_METHOD_NIC: - return mlx5_query_hca_vport_pkey(mdev, 0, port, 0, index, - pkey); + return mlx5_query_hca_nic_pkey(ibdev, port, index, pkey); default: return -EINVAL; } @@ -1174,23 +1362,32 @@ static int set_port_caps_atomic(struct mlx5_ib_dev *dev, u8 port_num, u32 mask, u32 value) { struct mlx5_hca_vport_context ctx = {}; + struct mlx5_core_dev *mdev; + u8 mdev_port_num; int err; - err = mlx5_query_hca_vport_context(dev->mdev, 0, - port_num, 0, &ctx); + mdev = mlx5_ib_get_native_port_mdev(dev, port_num, &mdev_port_num); + if (!mdev) + return -ENODEV; + + err = mlx5_query_hca_vport_context(mdev, 0, mdev_port_num, 0, &ctx); if (err) - return err; + goto out; if (~ctx.cap_mask1_perm & mask) { mlx5_ib_warn(dev, "trying to change bitmask 0x%X but change supported 0x%X\n", mask, ctx.cap_mask1_perm); - return -EINVAL; + err = -EINVAL; + goto out; } ctx.cap_mask1 = value; ctx.cap_mask1_perm = mask; - err = mlx5_core_modify_hca_vport_context(dev->mdev, 0, - port_num, 0, &ctx); + err = mlx5_core_modify_hca_vport_context(mdev, 0, mdev_port_num, + 0, &ctx); + +out: + mlx5_ib_put_native_port_mdev(dev, port_num); return err; } @@ -1241,9 +1438,18 @@ static void print_lib_caps(struct mlx5_ib_dev *dev, u64 caps) caps & MLX5_LIB_CAP_4K_UAR ? "y" : "n"); } +static u16 calc_dynamic_bfregs(int uars_per_sys_page) +{ + /* Large page with non 4k uar support might limit the dynamic size */ + if (uars_per_sys_page == 1 && PAGE_SIZE > 4096) + return MLX5_MIN_DYN_BFREGS; + + return MLX5_MAX_DYN_BFREGS; +} + static int calc_total_bfregs(struct mlx5_ib_dev *dev, bool lib_uar_4k, struct mlx5_ib_alloc_ucontext_req_v2 *req, - u32 *num_sys_pages) + struct mlx5_bfreg_info *bfregi) { int uars_per_sys_page; int bfregs_per_sys_page; @@ -1260,16 +1466,21 @@ static int calc_total_bfregs(struct mlx5_ib_dev *dev, bool lib_uar_4k, uars_per_sys_page = get_uars_per_sys_page(dev, lib_uar_4k); bfregs_per_sys_page = uars_per_sys_page * MLX5_NON_FP_BFREGS_PER_UAR; + /* This holds the required static allocation asked by the user */ req->total_num_bfregs = ALIGN(req->total_num_bfregs, bfregs_per_sys_page); - *num_sys_pages = req->total_num_bfregs / bfregs_per_sys_page; - if (req->num_low_latency_bfregs > req->total_num_bfregs - 1) return -EINVAL; - mlx5_ib_dbg(dev, "uar_4k: fw support %s, lib support %s, user requested %d bfregs, allocated %d, using %d sys pages\n", + bfregi->num_static_sys_pages = req->total_num_bfregs / bfregs_per_sys_page; + bfregi->num_dyn_bfregs = ALIGN(calc_dynamic_bfregs(uars_per_sys_page), bfregs_per_sys_page); + bfregi->total_num_bfregs = req->total_num_bfregs + bfregi->num_dyn_bfregs; + bfregi->num_sys_pages = bfregi->total_num_bfregs / bfregs_per_sys_page; + + mlx5_ib_dbg(dev, "uar_4k: fw support %s, lib support %s, user requested %d bfregs, allocated %d, total bfregs %d, using %d sys pages\n", MLX5_CAP_GEN(dev->mdev, uar_4k) ? "yes" : "no", lib_uar_4k ? "yes" : "no", ref_bfregs, - req->total_num_bfregs, *num_sys_pages); + req->total_num_bfregs, bfregi->total_num_bfregs, + bfregi->num_sys_pages); return 0; } @@ -1281,13 +1492,17 @@ static int allocate_uars(struct mlx5_ib_dev *dev, struct mlx5_ib_ucontext *conte int i; bfregi = &context->bfregi; - for (i = 0; i < bfregi->num_sys_pages; i++) { + for (i = 0; i < bfregi->num_static_sys_pages; i++) { err = mlx5_cmd_alloc_uar(dev->mdev, &bfregi->sys_pages[i]); if (err) goto error; mlx5_ib_dbg(dev, "allocated uar %d\n", bfregi->sys_pages[i]); } + + for (i = bfregi->num_static_sys_pages; i < bfregi->num_sys_pages; i++) + bfregi->sys_pages[i] = MLX5_IB_INVALID_UAR_INDEX; + return 0; error: @@ -1306,12 +1521,16 @@ static int deallocate_uars(struct mlx5_ib_dev *dev, struct mlx5_ib_ucontext *con bfregi = &context->bfregi; for (i = 0; i < bfregi->num_sys_pages; i++) { - err = mlx5_cmd_free_uar(dev->mdev, bfregi->sys_pages[i]); - if (err) { - mlx5_ib_warn(dev, "failed to free uar %d\n", i); - return err; + if (i < bfregi->num_static_sys_pages || + bfregi->sys_pages[i] != MLX5_IB_INVALID_UAR_INDEX) { + err = mlx5_cmd_free_uar(dev->mdev, bfregi->sys_pages[i]); + if (err) { + mlx5_ib_warn(dev, "failed to free uar %d, err=%d\n", i, err); + return err; + } } } + return 0; } @@ -1420,13 +1639,13 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev, bfregi = &context->bfregi; /* updates req->total_num_bfregs */ - err = calc_total_bfregs(dev, lib_uar_4k, &req, &bfregi->num_sys_pages); + err = calc_total_bfregs(dev, lib_uar_4k, &req, bfregi); if (err) goto out_ctx; mutex_init(&bfregi->lock); bfregi->lib_uar_4k = lib_uar_4k; - bfregi->count = kcalloc(req.total_num_bfregs, sizeof(*bfregi->count), + bfregi->count = kcalloc(bfregi->total_num_bfregs, sizeof(*bfregi->count), GFP_KERNEL); if (!bfregi->count) { err = -ENOMEM; @@ -1468,7 +1687,7 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev, mutex_init(&context->db_page_mutex); resp.tot_bfregs = req.total_num_bfregs; - resp.num_ports = MLX5_CAP_GEN(dev->mdev, num_ports); + resp.num_ports = dev->num_ports; if (field_avail(typeof(resp), cqe_version, udata->outlen)) resp.response_length += sizeof(resp.cqe_version); @@ -1510,6 +1729,11 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev, if (field_avail(typeof(resp), num_uars_per_page, udata->outlen)) resp.response_length += sizeof(resp.num_uars_per_page); + if (field_avail(typeof(resp), num_dyn_bfregs, udata->outlen)) { + resp.num_dyn_bfregs = bfregi->num_dyn_bfregs; + resp.response_length += sizeof(resp.num_dyn_bfregs); + } + err = ib_copy_to_udata(udata, &resp, resp.response_length); if (err) goto out_td; @@ -1564,15 +1788,13 @@ static int mlx5_ib_dealloc_ucontext(struct ib_ucontext *ibcontext) } static phys_addr_t uar_index2pfn(struct mlx5_ib_dev *dev, - struct mlx5_bfreg_info *bfregi, - int idx) + int uar_idx) { int fw_uars_per_page; fw_uars_per_page = MLX5_CAP_GEN(dev->mdev, uar_4k) ? MLX5_UARS_IN_PAGE : 1; - return (pci_resource_start(dev->mdev->pdev, 0) >> PAGE_SHIFT) + - bfregi->sys_pages[idx] / fw_uars_per_page; + return (pci_resource_start(dev->mdev->pdev, 0) >> PAGE_SHIFT) + uar_idx / fw_uars_per_page; } static int get_command(unsigned long offset) @@ -1590,6 +1812,12 @@ static int get_index(unsigned long offset) return get_arg(offset); } +/* Index resides in an extra byte to enable larger values than 255 */ +static int get_extended_index(unsigned long offset) +{ + return get_arg(offset) | ((offset >> 16) & 0xff) << 8; +} + static void mlx5_ib_vma_open(struct vm_area_struct *area) { /* vma_open is called when a new VMA is created on top of our VMA. This @@ -1740,21 +1968,29 @@ static int uar_mmap(struct mlx5_ib_dev *dev, enum mlx5_ib_mmap_cmd cmd, unsigned long idx; phys_addr_t pfn, pa; pgprot_t prot; - int uars_per_page; + u32 bfreg_dyn_idx = 0; + u32 uar_index; + int dyn_uar = (cmd == MLX5_IB_MMAP_ALLOC_WC); + int max_valid_idx = dyn_uar ? bfregi->num_sys_pages : + bfregi->num_static_sys_pages; if (vma->vm_end - vma->vm_start != PAGE_SIZE) return -EINVAL; - uars_per_page = get_uars_per_sys_page(dev, bfregi->lib_uar_4k); - idx = get_index(vma->vm_pgoff); - if (idx % uars_per_page || - idx * uars_per_page >= bfregi->num_sys_pages) { - mlx5_ib_warn(dev, "invalid uar index %lu\n", idx); + if (dyn_uar) + idx = get_extended_index(vma->vm_pgoff) + bfregi->num_static_sys_pages; + else + idx = get_index(vma->vm_pgoff); + + if (idx >= max_valid_idx) { + mlx5_ib_warn(dev, "invalid uar index %lu, max=%d\n", + idx, max_valid_idx); return -EINVAL; } switch (cmd) { case MLX5_IB_MMAP_WC_PAGE: + case MLX5_IB_MMAP_ALLOC_WC: /* Some architectures don't support WC memory */ #if defined(CONFIG_X86) if (!pat_enabled()) @@ -1774,7 +2010,40 @@ static int uar_mmap(struct mlx5_ib_dev *dev, enum mlx5_ib_mmap_cmd cmd, return -EINVAL; } - pfn = uar_index2pfn(dev, bfregi, idx); + if (dyn_uar) { + int uars_per_page; + + uars_per_page = get_uars_per_sys_page(dev, bfregi->lib_uar_4k); + bfreg_dyn_idx = idx * (uars_per_page * MLX5_NON_FP_BFREGS_PER_UAR); + if (bfreg_dyn_idx >= bfregi->total_num_bfregs) { + mlx5_ib_warn(dev, "invalid bfreg_dyn_idx %u, max=%u\n", + bfreg_dyn_idx, bfregi->total_num_bfregs); + return -EINVAL; + } + + mutex_lock(&bfregi->lock); + /* Fail if uar already allocated, first bfreg index of each + * page holds its count. + */ + if (bfregi->count[bfreg_dyn_idx]) { + mlx5_ib_warn(dev, "wrong offset, idx %lu is busy, bfregn=%u\n", idx, bfreg_dyn_idx); + mutex_unlock(&bfregi->lock); + return -EINVAL; + } + + bfregi->count[bfreg_dyn_idx]++; + mutex_unlock(&bfregi->lock); + + err = mlx5_cmd_alloc_uar(dev->mdev, &uar_index); + if (err) { + mlx5_ib_warn(dev, "UAR alloc failed\n"); + goto free_bfreg; + } + } else { + uar_index = bfregi->sys_pages[idx]; + } + + pfn = uar_index2pfn(dev, uar_index); mlx5_ib_dbg(dev, "uar idx 0x%lx, pfn %pa\n", idx, &pfn); vma->vm_page_prot = prot; @@ -1783,14 +2052,32 @@ static int uar_mmap(struct mlx5_ib_dev *dev, enum mlx5_ib_mmap_cmd cmd, if (err) { mlx5_ib_err(dev, "io_remap_pfn_range failed with error=%d, vm_start=0x%lx, pfn=%pa, mmap_cmd=%s\n", err, vma->vm_start, &pfn, mmap_cmd2str(cmd)); - return -EAGAIN; + err = -EAGAIN; + goto err; } pa = pfn << PAGE_SHIFT; mlx5_ib_dbg(dev, "mapped %s at 0x%lx, PA %pa\n", mmap_cmd2str(cmd), vma->vm_start, &pa); - return mlx5_ib_set_vma_data(vma, context); + err = mlx5_ib_set_vma_data(vma, context); + if (err) + goto err; + + if (dyn_uar) + bfregi->sys_pages[idx] = uar_index; + return 0; + +err: + if (!dyn_uar) + return err; + + mlx5_cmd_free_uar(dev->mdev, idx); + +free_bfreg: + mlx5_ib_free_bfreg(dev, bfregi, bfreg_dyn_idx); + + return err; } static int mlx5_ib_mmap(struct ib_ucontext *ibcontext, struct vm_area_struct *vma) @@ -1805,6 +2092,7 @@ static int mlx5_ib_mmap(struct ib_ucontext *ibcontext, struct vm_area_struct *vm case MLX5_IB_MMAP_WC_PAGE: case MLX5_IB_MMAP_NC_PAGE: case MLX5_IB_MMAP_REGULAR_PAGE: + case MLX5_IB_MMAP_ALLOC_WC: return uar_mmap(dev, command, vma, context); case MLX5_IB_MMAP_GET_CONTIGUOUS_PAGES: @@ -2661,7 +2949,7 @@ static struct ib_flow *mlx5_ib_create_flow(struct ib_qp *qp, return ERR_PTR(-ENOMEM); if (domain != IB_FLOW_DOMAIN_USER || - flow_attr->port > MLX5_CAP_GEN(dev->mdev, num_ports) || + flow_attr->port > dev->num_ports || (flow_attr->flags & ~IB_FLOW_ATTR_FLAGS_DONT_TRAP)) return ERR_PTR(-EINVAL); @@ -2926,15 +3214,24 @@ static void delay_drop_handler(struct work_struct *work) mutex_unlock(&delay_drop->lock); } -static void mlx5_ib_event(struct mlx5_core_dev *dev, void *context, - enum mlx5_dev_event event, unsigned long param) +static void mlx5_ib_handle_event(struct work_struct *_work) { - struct mlx5_ib_dev *ibdev = (struct mlx5_ib_dev *)context; + struct mlx5_ib_event_work *work = + container_of(_work, struct mlx5_ib_event_work, work); + struct mlx5_ib_dev *ibdev; struct ib_event ibev; bool fatal = false; u8 port = 0; - switch (event) { + if (mlx5_core_is_mp_slave(work->dev)) { + ibdev = mlx5_ib_get_ibdev_from_mpi(work->context); + if (!ibdev) + goto out; + } else { + ibdev = work->context; + } + + switch (work->event) { case MLX5_DEV_EVENT_SYS_ERROR: ibev.event = IB_EVENT_DEVICE_FATAL; mlx5_ib_handle_internal_error(ibdev); @@ -2944,39 +3241,39 @@ static void mlx5_ib_event(struct mlx5_core_dev *dev, void *context, case MLX5_DEV_EVENT_PORT_UP: case MLX5_DEV_EVENT_PORT_DOWN: case MLX5_DEV_EVENT_PORT_INITIALIZED: - port = (u8)param; + port = (u8)work->param; /* In RoCE, port up/down events are handled in * mlx5_netdev_event(). */ if (mlx5_ib_port_link_layer(&ibdev->ib_dev, port) == IB_LINK_LAYER_ETHERNET) - return; + goto out; - ibev.event = (event == MLX5_DEV_EVENT_PORT_UP) ? + ibev.event = (work->event == MLX5_DEV_EVENT_PORT_UP) ? IB_EVENT_PORT_ACTIVE : IB_EVENT_PORT_ERR; break; case MLX5_DEV_EVENT_LID_CHANGE: ibev.event = IB_EVENT_LID_CHANGE; - port = (u8)param; + port = (u8)work->param; break; case MLX5_DEV_EVENT_PKEY_CHANGE: ibev.event = IB_EVENT_PKEY_CHANGE; - port = (u8)param; + port = (u8)work->param; schedule_work(&ibdev->devr.ports[port - 1].pkey_change_work); break; case MLX5_DEV_EVENT_GUID_CHANGE: ibev.event = IB_EVENT_GID_CHANGE; - port = (u8)param; + port = (u8)work->param; break; case MLX5_DEV_EVENT_CLIENT_REREG: ibev.event = IB_EVENT_CLIENT_REREGISTER; - port = (u8)param; + port = (u8)work->param; break; case MLX5_DEV_EVENT_DELAY_DROP_TIMEOUT: schedule_work(&ibdev->delay_drop.delay_drop_work); @@ -2998,9 +3295,29 @@ static void mlx5_ib_event(struct mlx5_core_dev *dev, void *context, if (fatal) ibdev->ib_active = false; - out: - return; + kfree(work); +} + +static void mlx5_ib_event(struct mlx5_core_dev *dev, void *context, + enum mlx5_dev_event event, unsigned long param) +{ + struct mlx5_ib_event_work *work; + + work = kmalloc(sizeof(*work), GFP_ATOMIC); + if (work) { + INIT_WORK(&work->work, mlx5_ib_handle_event); + work->dev = dev; + work->param = param; + work->context = context; + work->event = event; + + queue_work(mlx5_ib_event_wq, &work->work); + return; + } + + dev_warn(&dev->pdev->dev, "%s: mlx5_dev_event: %d, with param: %lu dropped, couldn't allocate memory.\n", + __func__, event, param); } static int set_has_smi_cap(struct mlx5_ib_dev *dev) @@ -3009,7 +3326,7 @@ static int set_has_smi_cap(struct mlx5_ib_dev *dev) int err; int port; - for (port = 1; port <= MLX5_CAP_GEN(dev->mdev, num_ports); port++) { + for (port = 1; port <= dev->num_ports; port++) { dev->mdev->port_caps[port - 1].has_smi = false; if (MLX5_CAP_GEN(dev->mdev, port_type) == MLX5_CAP_PORT_TYPE_IB) { @@ -3036,16 +3353,15 @@ static void get_ext_port_caps(struct mlx5_ib_dev *dev) { int port; - for (port = 1; port <= MLX5_CAP_GEN(dev->mdev, num_ports); port++) + for (port = 1; port <= dev->num_ports; port++) mlx5_query_ext_port_caps(dev, port); } -static int get_port_caps(struct mlx5_ib_dev *dev) +static int get_port_caps(struct mlx5_ib_dev *dev, u8 port) { struct ib_device_attr *dprops = NULL; struct ib_port_attr *pprops = NULL; int err = -ENOMEM; - int port; struct ib_udata uhw = {.inlen = 0, .outlen = 0}; pprops = kmalloc(sizeof(*pprops), GFP_KERNEL); @@ -3066,22 +3382,21 @@ static int get_port_caps(struct mlx5_ib_dev *dev) goto out; } - for (port = 1; port <= MLX5_CAP_GEN(dev->mdev, num_ports); port++) { - memset(pprops, 0, sizeof(*pprops)); - err = mlx5_ib_query_port(&dev->ib_dev, port, pprops); - if (err) { - mlx5_ib_warn(dev, "query_port %d failed %d\n", - port, err); - break; - } - dev->mdev->port_caps[port - 1].pkey_table_len = - dprops->max_pkeys; - dev->mdev->port_caps[port - 1].gid_table_len = - pprops->gid_tbl_len; - mlx5_ib_dbg(dev, "pkey_table_len %d, gid_table_len %d\n", - dprops->max_pkeys, pprops->gid_tbl_len); + memset(pprops, 0, sizeof(*pprops)); + err = mlx5_ib_query_port(&dev->ib_dev, port, pprops); + if (err) { + mlx5_ib_warn(dev, "query_port %d failed %d\n", + port, err); + goto out; } + dev->mdev->port_caps[port - 1].pkey_table_len = + dprops->max_pkeys; + dev->mdev->port_caps[port - 1].gid_table_len = + pprops->gid_tbl_len; + mlx5_ib_dbg(dev, "port %d: pkey_table_len %d, gid_table_len %d\n", + port, dprops->max_pkeys, pprops->gid_tbl_len); + out: kfree(pprops); kfree(dprops); @@ -3371,12 +3686,14 @@ static u32 get_core_cap_flags(struct ib_device *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); + bool raw_support = !mlx5_core_mp_enabled(dev->mdev); u32 ret = 0; if (ll == IB_LINK_LAYER_INFINIBAND) return RDMA_CORE_PORT_IBA_IB; - ret = RDMA_CORE_PORT_RAW_PACKET; + if (raw_support) + ret = RDMA_CORE_PORT_RAW_PACKET; if (!(l3_type_cap & MLX5_ROCE_L3_TYPE_IPV4_CAP)) return ret; @@ -3466,33 +3783,33 @@ static void mlx5_eth_lag_cleanup(struct mlx5_ib_dev *dev) } } -static int mlx5_add_netdev_notifier(struct mlx5_ib_dev *dev) +static int mlx5_add_netdev_notifier(struct mlx5_ib_dev *dev, u8 port_num) { int err; - dev->roce.nb.notifier_call = mlx5_netdev_event; - err = register_netdevice_notifier(&dev->roce.nb); + dev->roce[port_num].nb.notifier_call = mlx5_netdev_event; + err = register_netdevice_notifier(&dev->roce[port_num].nb); if (err) { - dev->roce.nb.notifier_call = NULL; + dev->roce[port_num].nb.notifier_call = NULL; return err; } return 0; } -static void mlx5_remove_netdev_notifier(struct mlx5_ib_dev *dev) +static void mlx5_remove_netdev_notifier(struct mlx5_ib_dev *dev, u8 port_num) { - if (dev->roce.nb.notifier_call) { - unregister_netdevice_notifier(&dev->roce.nb); - dev->roce.nb.notifier_call = NULL; + if (dev->roce[port_num].nb.notifier_call) { + unregister_netdevice_notifier(&dev->roce[port_num].nb); + dev->roce[port_num].nb.notifier_call = NULL; } } -static int mlx5_enable_eth(struct mlx5_ib_dev *dev) +static int mlx5_enable_eth(struct mlx5_ib_dev *dev, u8 port_num) { int err; - err = mlx5_add_netdev_notifier(dev); + err = mlx5_add_netdev_notifier(dev, port_num); if (err) return err; @@ -3513,7 +3830,7 @@ err_disable_roce: mlx5_nic_vport_disable_roce(dev->mdev); err_unregister_netdevice_notifier: - mlx5_remove_netdev_notifier(dev); + mlx5_remove_netdev_notifier(dev, port_num); return err; } @@ -3575,11 +3892,12 @@ static const struct mlx5_ib_counter extended_err_cnts[] = { static void mlx5_ib_dealloc_counters(struct mlx5_ib_dev *dev) { - unsigned int i; + int i; for (i = 0; i < dev->num_ports; i++) { - mlx5_core_dealloc_q_counter(dev->mdev, - dev->port[i].cnts.set_id); + if (dev->port[i].cnts.set_id) + mlx5_core_dealloc_q_counter(dev->mdev, + dev->port[i].cnts.set_id); kfree(dev->port[i].cnts.names); kfree(dev->port[i].cnts.offsets); } @@ -3621,6 +3939,7 @@ static int __mlx5_ib_alloc_counters(struct mlx5_ib_dev *dev, err_names: kfree(cnts->names); + cnts->names = NULL; return -ENOMEM; } @@ -3667,37 +3986,33 @@ static void mlx5_ib_fill_counters(struct mlx5_ib_dev *dev, static int mlx5_ib_alloc_counters(struct mlx5_ib_dev *dev) { + int err = 0; int i; - int ret; for (i = 0; i < dev->num_ports; i++) { - struct mlx5_ib_port *port = &dev->port[i]; + err = __mlx5_ib_alloc_counters(dev, &dev->port[i].cnts); + if (err) + goto err_alloc; + + mlx5_ib_fill_counters(dev, dev->port[i].cnts.names, + dev->port[i].cnts.offsets); - ret = mlx5_core_alloc_q_counter(dev->mdev, - &port->cnts.set_id); - if (ret) { + err = mlx5_core_alloc_q_counter(dev->mdev, + &dev->port[i].cnts.set_id); + if (err) { mlx5_ib_warn(dev, "couldn't allocate queue counter for port %d, err %d\n", - i + 1, ret); - goto dealloc_counters; + i + 1, err); + goto err_alloc; } - - ret = __mlx5_ib_alloc_counters(dev, &port->cnts); - if (ret) - goto dealloc_counters; - - mlx5_ib_fill_counters(dev, port->cnts.names, - port->cnts.offsets); + dev->port[i].cnts.set_id_valid = true; } return 0; -dealloc_counters: - while (--i >= 0) - mlx5_core_dealloc_q_counter(dev->mdev, - dev->port[i].cnts.set_id); - - return ret; +err_alloc: + mlx5_ib_dealloc_counters(dev); + return err; } static struct rdma_hw_stats *mlx5_ib_alloc_hw_stats(struct ib_device *ibdev, @@ -3716,7 +4031,7 @@ static struct rdma_hw_stats *mlx5_ib_alloc_hw_stats(struct ib_device *ibdev, RDMA_HW_STATS_DEFAULT_LIFESPAN); } -static int mlx5_ib_query_q_counters(struct mlx5_ib_dev *dev, +static int mlx5_ib_query_q_counters(struct mlx5_core_dev *mdev, struct mlx5_ib_port *port, struct rdma_hw_stats *stats) { @@ -3729,7 +4044,7 @@ static int mlx5_ib_query_q_counters(struct mlx5_ib_dev *dev, if (!out) return -ENOMEM; - ret = mlx5_core_query_q_counter(dev->mdev, + ret = mlx5_core_query_q_counter(mdev, port->cnts.set_id, 0, out, outlen); if (ret) @@ -3751,28 +4066,43 @@ static int mlx5_ib_get_hw_stats(struct ib_device *ibdev, { struct mlx5_ib_dev *dev = to_mdev(ibdev); struct mlx5_ib_port *port = &dev->port[port_num - 1]; + struct mlx5_core_dev *mdev; int ret, num_counters; + u8 mdev_port_num; if (!stats) return -EINVAL; - ret = mlx5_ib_query_q_counters(dev, port, stats); + num_counters = port->cnts.num_q_counters + port->cnts.num_cong_counters; + + /* q_counters are per IB device, query the master mdev */ + ret = mlx5_ib_query_q_counters(dev->mdev, port, stats); if (ret) return ret; - num_counters = port->cnts.num_q_counters; if (MLX5_CAP_GEN(dev->mdev, cc_query_allowed)) { + mdev = mlx5_ib_get_native_port_mdev(dev, port_num, + &mdev_port_num); + if (!mdev) { + /* If port is not affiliated yet, its in down state + * which doesn't have any counters yet, so it would be + * zero. So no need to read from the HCA. + */ + goto done; + } ret = mlx5_lag_query_cong_counters(dev->mdev, stats->value + port->cnts.num_q_counters, port->cnts.num_cong_counters, port->cnts.offsets + port->cnts.num_q_counters); + + mlx5_ib_put_native_port_mdev(dev, port_num); if (ret) return ret; - num_counters += port->cnts.num_cong_counters; } +done: return num_counters; } @@ -3934,36 +4264,250 @@ mlx5_ib_get_vector_affinity(struct ib_device *ibdev, int comp_vector) return mlx5_get_vector_affinity(dev->mdev, comp_vector); } -static void *mlx5_ib_add(struct mlx5_core_dev *mdev) +/* The mlx5_ib_multiport_mutex should be held when calling this function */ +static void mlx5_ib_unbind_slave_port(struct mlx5_ib_dev *ibdev, + struct mlx5_ib_multiport_info *mpi) { - struct mlx5_ib_dev *dev; - enum rdma_link_layer ll; - int port_type_cap; - const char *name; + u8 port_num = mlx5_core_native_port_num(mpi->mdev) - 1; + struct mlx5_ib_port *port = &ibdev->port[port_num]; + int comps; int err; int i; - port_type_cap = MLX5_CAP_GEN(mdev, port_type); - ll = mlx5_port_type_cap_to_rdma_ll(port_type_cap); + mlx5_ib_cleanup_cong_debugfs(ibdev, port_num); - printk_once(KERN_INFO "%s", mlx5_version); + spin_lock(&port->mp.mpi_lock); + if (!mpi->ibdev) { + spin_unlock(&port->mp.mpi_lock); + return; + } + mpi->ibdev = NULL; - dev = (struct mlx5_ib_dev *)ib_alloc_device(sizeof(*dev)); - if (!dev) - return NULL; + spin_unlock(&port->mp.mpi_lock); + mlx5_remove_netdev_notifier(ibdev, port_num); + spin_lock(&port->mp.mpi_lock); - dev->mdev = mdev; + comps = mpi->mdev_refcnt; + if (comps) { + mpi->unaffiliate = true; + init_completion(&mpi->unref_comp); + spin_unlock(&port->mp.mpi_lock); + + for (i = 0; i < comps; i++) + wait_for_completion(&mpi->unref_comp); + + spin_lock(&port->mp.mpi_lock); + mpi->unaffiliate = false; + } + + port->mp.mpi = NULL; + + list_add_tail(&mpi->list, &mlx5_ib_unaffiliated_port_list); + + spin_unlock(&port->mp.mpi_lock); + + err = mlx5_nic_vport_unaffiliate_multiport(mpi->mdev); + + mlx5_ib_dbg(ibdev, "unaffiliated port %d\n", port_num + 1); + /* Log an error, still needed to cleanup the pointers and add + * it back to the list. + */ + if (err) + mlx5_ib_err(ibdev, "Failed to unaffiliate port %u\n", + port_num + 1); + + ibdev->roce[port_num].last_port_state = IB_PORT_DOWN; +} + +/* The mlx5_ib_multiport_mutex should be held when calling this function */ +static bool mlx5_ib_bind_slave_port(struct mlx5_ib_dev *ibdev, + struct mlx5_ib_multiport_info *mpi) +{ + u8 port_num = mlx5_core_native_port_num(mpi->mdev) - 1; + int err; + + spin_lock(&ibdev->port[port_num].mp.mpi_lock); + if (ibdev->port[port_num].mp.mpi) { + mlx5_ib_warn(ibdev, "port %d already affiliated.\n", + port_num + 1); + spin_unlock(&ibdev->port[port_num].mp.mpi_lock); + return false; + } + + ibdev->port[port_num].mp.mpi = mpi; + mpi->ibdev = ibdev; + spin_unlock(&ibdev->port[port_num].mp.mpi_lock); + + err = mlx5_nic_vport_affiliate_multiport(ibdev->mdev, mpi->mdev); + if (err) + goto unbind; + + err = get_port_caps(ibdev, mlx5_core_native_port_num(mpi->mdev)); + if (err) + goto unbind; + + err = mlx5_add_netdev_notifier(ibdev, port_num); + if (err) { + mlx5_ib_err(ibdev, "failed adding netdev notifier for port %u\n", + port_num + 1); + goto unbind; + } + + err = mlx5_ib_init_cong_debugfs(ibdev, port_num); + if (err) + goto unbind; + + return true; + +unbind: + mlx5_ib_unbind_slave_port(ibdev, mpi); + return false; +} + +static int mlx5_ib_init_multiport_master(struct mlx5_ib_dev *dev) +{ + int port_num = mlx5_core_native_port_num(dev->mdev) - 1; + enum rdma_link_layer ll = mlx5_ib_port_link_layer(&dev->ib_dev, + port_num + 1); + struct mlx5_ib_multiport_info *mpi; + int err; + int i; + + if (!mlx5_core_is_mp_master(dev->mdev) || ll != IB_LINK_LAYER_ETHERNET) + return 0; + + err = mlx5_query_nic_vport_system_image_guid(dev->mdev, + &dev->sys_image_guid); + if (err) + return err; + + err = mlx5_nic_vport_enable_roce(dev->mdev); + if (err) + return err; + + mutex_lock(&mlx5_ib_multiport_mutex); + for (i = 0; i < dev->num_ports; i++) { + bool bound = false; + + /* build a stub multiport info struct for the native port. */ + if (i == port_num) { + mpi = kzalloc(sizeof(*mpi), GFP_KERNEL); + if (!mpi) { + mutex_unlock(&mlx5_ib_multiport_mutex); + mlx5_nic_vport_disable_roce(dev->mdev); + return -ENOMEM; + } + + mpi->is_master = true; + mpi->mdev = dev->mdev; + mpi->sys_image_guid = dev->sys_image_guid; + dev->port[i].mp.mpi = mpi; + mpi->ibdev = dev; + mpi = NULL; + continue; + } + + list_for_each_entry(mpi, &mlx5_ib_unaffiliated_port_list, + list) { + if (dev->sys_image_guid == mpi->sys_image_guid && + (mlx5_core_native_port_num(mpi->mdev) - 1) == i) { + bound = mlx5_ib_bind_slave_port(dev, mpi); + } + + if (bound) { + dev_dbg(&mpi->mdev->pdev->dev, "removing port from unaffiliated list.\n"); + mlx5_ib_dbg(dev, "port %d bound\n", i + 1); + list_del(&mpi->list); + break; + } + } + if (!bound) { + get_port_caps(dev, i + 1); + mlx5_ib_dbg(dev, "no free port found for port %d\n", + i + 1); + } + } + + list_add_tail(&dev->ib_dev_list, &mlx5_ib_dev_list); + mutex_unlock(&mlx5_ib_multiport_mutex); + return err; +} + +static void mlx5_ib_cleanup_multiport_master(struct mlx5_ib_dev *dev) +{ + int port_num = mlx5_core_native_port_num(dev->mdev) - 1; + enum rdma_link_layer ll = mlx5_ib_port_link_layer(&dev->ib_dev, + port_num + 1); + int i; + + if (!mlx5_core_is_mp_master(dev->mdev) || ll != IB_LINK_LAYER_ETHERNET) + return; + + mutex_lock(&mlx5_ib_multiport_mutex); + for (i = 0; i < dev->num_ports; i++) { + if (dev->port[i].mp.mpi) { + /* Destroy the native port stub */ + if (i == port_num) { + kfree(dev->port[i].mp.mpi); + dev->port[i].mp.mpi = NULL; + } else { + mlx5_ib_dbg(dev, "unbinding port_num: %d\n", i + 1); + mlx5_ib_unbind_slave_port(dev, dev->port[i].mp.mpi); + } + } + } + + mlx5_ib_dbg(dev, "removing from devlist\n"); + list_del(&dev->ib_dev_list); + mutex_unlock(&mlx5_ib_multiport_mutex); - dev->port = kcalloc(MLX5_CAP_GEN(mdev, num_ports), sizeof(*dev->port), + mlx5_nic_vport_disable_roce(dev->mdev); +} + +static void mlx5_ib_stage_init_cleanup(struct mlx5_ib_dev *dev) +{ + mlx5_ib_cleanup_multiport_master(dev); +#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING + cleanup_srcu_struct(&dev->mr_srcu); +#endif + kfree(dev->port); +} + +static int mlx5_ib_stage_init_init(struct mlx5_ib_dev *dev) +{ + struct mlx5_core_dev *mdev = dev->mdev; + const char *name; + int err; + int i; + + dev->port = kcalloc(dev->num_ports, sizeof(*dev->port), GFP_KERNEL); if (!dev->port) - goto err_dealloc; + return -ENOMEM; - rwlock_init(&dev->roce.netdev_lock); - err = get_port_caps(dev); + for (i = 0; i < dev->num_ports; i++) { + spin_lock_init(&dev->port[i].mp.mpi_lock); + rwlock_init(&dev->roce[i].netdev_lock); + } + + err = mlx5_ib_init_multiport_master(dev); if (err) goto err_free_port; + if (!mlx5_core_mp_enabled(mdev)) { + int i; + + for (i = 1; i <= dev->num_ports; i++) { + err = get_port_caps(dev, i); + if (err) + break; + } + } else { + err = get_port_caps(dev, mlx5_core_native_port_num(mdev)); + } + if (err) + goto err_mp; + if (mlx5_use_mad_ifc(dev)) get_ext_port_caps(dev); @@ -3976,12 +4520,37 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev) dev->ib_dev.owner = THIS_MODULE; dev->ib_dev.node_type = RDMA_NODE_IB_CA; dev->ib_dev.local_dma_lkey = 0 /* not supported for now */; - dev->num_ports = MLX5_CAP_GEN(mdev, num_ports); - dev->ib_dev.phys_port_cnt = dev->num_ports; + dev->ib_dev.phys_port_cnt = dev->num_ports; dev->ib_dev.num_comp_vectors = dev->mdev->priv.eq_table.num_comp_vectors; dev->ib_dev.dev.parent = &mdev->pdev->dev; + mutex_init(&dev->flow_db.lock); + mutex_init(&dev->cap_mask_mutex); + INIT_LIST_HEAD(&dev->qp_list); + spin_lock_init(&dev->reset_flow_resource_lock); + +#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING + err = init_srcu_struct(&dev->mr_srcu); + if (err) + goto err_free_port; +#endif + + return 0; +err_mp: + mlx5_ib_cleanup_multiport_master(dev); + +err_free_port: + kfree(dev->port); + + return -ENOMEM; +} + +static int mlx5_ib_stage_caps_init(struct mlx5_ib_dev *dev) +{ + struct mlx5_core_dev *mdev = dev->mdev; + int err; + dev->ib_dev.uverbs_abi_ver = MLX5_IB_UVERBS_ABI_VERSION; dev->ib_dev.uverbs_cmd_mask = (1ull << IB_USER_VERBS_CMD_GET_CONTEXT) | @@ -4020,8 +4589,6 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev) dev->ib_dev.query_device = mlx5_ib_query_device; dev->ib_dev.query_port = mlx5_ib_query_port; dev->ib_dev.get_link_layer = mlx5_ib_port_link_layer; - if (ll == IB_LINK_LAYER_ETHERNET) - dev->ib_dev.get_netdev = mlx5_ib_get_netdev; dev->ib_dev.query_gid = mlx5_ib_query_gid; dev->ib_dev.add_gid = mlx5_ib_add_gid; dev->ib_dev.del_gid = mlx5_ib_del_gid; @@ -4078,8 +4645,6 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev) dev->ib_dev.disassociate_ucontext = mlx5_ib_disassociate_ucontext; - mlx5_ib_internal_fill_odp_caps(dev); - dev->umr_fence = mlx5_get_umr_fence(MLX5_CAP_GEN(mdev, umr_fence)); if (MLX5_CAP_GEN(mdev, imaicl)) { @@ -4090,11 +4655,6 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev) (1ull << IB_USER_VERBS_CMD_DEALLOC_MW); } - if (MLX5_CAP_GEN(dev->mdev, max_qp_cnt)) { - dev->ib_dev.get_hw_stats = mlx5_ib_get_hw_stats; - dev->ib_dev.alloc_hw_stats = mlx5_ib_alloc_hw_stats; - } - if (MLX5_CAP_GEN(mdev, xrc)) { dev->ib_dev.alloc_xrcd = mlx5_ib_alloc_xrcd; dev->ib_dev.dealloc_xrcd = mlx5_ib_dealloc_xrcd; @@ -4109,8 +4669,38 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev) (1ull << IB_USER_VERBS_EX_CMD_CREATE_FLOW) | (1ull << IB_USER_VERBS_EX_CMD_DESTROY_FLOW); - if (mlx5_ib_port_link_layer(&dev->ib_dev, 1) == - IB_LINK_LAYER_ETHERNET) { + err = init_node_data(dev); + if (err) + return err; + + if ((MLX5_CAP_GEN(dev->mdev, port_type) == MLX5_CAP_PORT_TYPE_ETH) && + MLX5_CAP_GEN(dev->mdev, disable_local_lb)) + mutex_init(&dev->lb_mutex); + + return 0; +} + +static int mlx5_ib_stage_roce_init(struct mlx5_ib_dev *dev) +{ + struct mlx5_core_dev *mdev = dev->mdev; + enum rdma_link_layer ll; + int port_type_cap; + u8 port_num; + int err; + int i; + + port_num = mlx5_core_native_port_num(dev->mdev) - 1; + port_type_cap = MLX5_CAP_GEN(mdev, port_type); + ll = mlx5_port_type_cap_to_rdma_ll(port_type_cap); + + if (ll == IB_LINK_LAYER_ETHERNET) { + for (i = 0; i < dev->num_ports; i++) { + dev->roce[i].dev = dev; + dev->roce[i].native_port_num = i + 1; + dev->roce[i].last_port_state = IB_PORT_DOWN; + } + + dev->ib_dev.get_netdev = mlx5_ib_get_netdev; dev->ib_dev.create_wq = mlx5_ib_create_wq; dev->ib_dev.modify_wq = mlx5_ib_modify_wq; dev->ib_dev.destroy_wq = mlx5_ib_destroy_wq; @@ -4122,142 +4712,329 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev) (1ull << IB_USER_VERBS_EX_CMD_DESTROY_WQ) | (1ull << IB_USER_VERBS_EX_CMD_CREATE_RWQ_IND_TBL) | (1ull << IB_USER_VERBS_EX_CMD_DESTROY_RWQ_IND_TBL); + err = mlx5_enable_eth(dev, port_num); + if (err) + return err; } - err = init_node_data(dev); - if (err) - goto err_free_port; - mutex_init(&dev->flow_db.lock); - mutex_init(&dev->cap_mask_mutex); - INIT_LIST_HEAD(&dev->qp_list); - spin_lock_init(&dev->reset_flow_resource_lock); + return 0; +} + +static void mlx5_ib_stage_roce_cleanup(struct mlx5_ib_dev *dev) +{ + struct mlx5_core_dev *mdev = dev->mdev; + enum rdma_link_layer ll; + int port_type_cap; + u8 port_num; + + port_num = mlx5_core_native_port_num(dev->mdev) - 1; + port_type_cap = MLX5_CAP_GEN(mdev, port_type); + ll = mlx5_port_type_cap_to_rdma_ll(port_type_cap); if (ll == IB_LINK_LAYER_ETHERNET) { - err = mlx5_enable_eth(dev); - if (err) - goto err_free_port; - dev->roce.last_port_state = IB_PORT_DOWN; + mlx5_disable_eth(dev); + mlx5_remove_netdev_notifier(dev, port_num); } +} - err = create_dev_resources(&dev->devr); - if (err) - goto err_disable_eth; +static int mlx5_ib_stage_dev_res_init(struct mlx5_ib_dev *dev) +{ + return create_dev_resources(&dev->devr); +} - err = mlx5_ib_odp_init_one(dev); - if (err) - goto err_rsrc; +static void mlx5_ib_stage_dev_res_cleanup(struct mlx5_ib_dev *dev) +{ + destroy_dev_resources(&dev->devr); +} + +static int mlx5_ib_stage_odp_init(struct mlx5_ib_dev *dev) +{ + mlx5_ib_internal_fill_odp_caps(dev); + return mlx5_ib_odp_init_one(dev); +} + +static int mlx5_ib_stage_counters_init(struct mlx5_ib_dev *dev) +{ if (MLX5_CAP_GEN(dev->mdev, max_qp_cnt)) { - err = mlx5_ib_alloc_counters(dev); - if (err) - goto err_odp; + dev->ib_dev.get_hw_stats = mlx5_ib_get_hw_stats; + dev->ib_dev.alloc_hw_stats = mlx5_ib_alloc_hw_stats; + + return mlx5_ib_alloc_counters(dev); } - err = mlx5_ib_init_cong_debugfs(dev); - if (err) - goto err_cnt; + return 0; +} + +static void mlx5_ib_stage_counters_cleanup(struct mlx5_ib_dev *dev) +{ + if (MLX5_CAP_GEN(dev->mdev, max_qp_cnt)) + mlx5_ib_dealloc_counters(dev); +} + +static int mlx5_ib_stage_cong_debugfs_init(struct mlx5_ib_dev *dev) +{ + return mlx5_ib_init_cong_debugfs(dev, + mlx5_core_native_port_num(dev->mdev) - 1); +} +static void mlx5_ib_stage_cong_debugfs_cleanup(struct mlx5_ib_dev *dev) +{ + mlx5_ib_cleanup_cong_debugfs(dev, + mlx5_core_native_port_num(dev->mdev) - 1); +} + +static int mlx5_ib_stage_uar_init(struct mlx5_ib_dev *dev) +{ dev->mdev->priv.uar = mlx5_get_uars_page(dev->mdev); if (!dev->mdev->priv.uar) - goto err_cong; + return -ENOMEM; + return 0; +} + +static void mlx5_ib_stage_uar_cleanup(struct mlx5_ib_dev *dev) +{ + mlx5_put_uars_page(dev->mdev, dev->mdev->priv.uar); +} + +static int mlx5_ib_stage_bfrag_init(struct mlx5_ib_dev *dev) +{ + int err; err = mlx5_alloc_bfreg(dev->mdev, &dev->bfreg, false, false); if (err) - goto err_uar_page; + return err; err = mlx5_alloc_bfreg(dev->mdev, &dev->fp_bfreg, false, true); if (err) - goto err_bfreg; + mlx5_free_bfreg(dev->mdev, &dev->fp_bfreg); - err = ib_register_device(&dev->ib_dev, NULL); - if (err) - goto err_fp_bfreg; + return err; +} - err = create_umr_res(dev); - if (err) - goto err_dev; +static void mlx5_ib_stage_bfrag_cleanup(struct mlx5_ib_dev *dev) +{ + mlx5_free_bfreg(dev->mdev, &dev->fp_bfreg); + mlx5_free_bfreg(dev->mdev, &dev->bfreg); +} + +static int mlx5_ib_stage_ib_reg_init(struct mlx5_ib_dev *dev) +{ + return ib_register_device(&dev->ib_dev, NULL); +} + +static void mlx5_ib_stage_ib_reg_cleanup(struct mlx5_ib_dev *dev) +{ + ib_unregister_device(&dev->ib_dev); +} + +static int mlx5_ib_stage_umr_res_init(struct mlx5_ib_dev *dev) +{ + return create_umr_res(dev); +} +static void mlx5_ib_stage_umr_res_cleanup(struct mlx5_ib_dev *dev) +{ + destroy_umrc_res(dev); +} + +static int mlx5_ib_stage_delay_drop_init(struct mlx5_ib_dev *dev) +{ init_delay_drop(dev); + return 0; +} + +static void mlx5_ib_stage_delay_drop_cleanup(struct mlx5_ib_dev *dev) +{ + cancel_delay_drop(dev); +} + +static int mlx5_ib_stage_class_attr_init(struct mlx5_ib_dev *dev) +{ + int err; + int i; + for (i = 0; i < ARRAY_SIZE(mlx5_class_attributes); i++) { err = device_create_file(&dev->ib_dev.dev, mlx5_class_attributes[i]); if (err) - goto err_delay_drop; + return err; } - if ((MLX5_CAP_GEN(mdev, port_type) == MLX5_CAP_PORT_TYPE_ETH) && - MLX5_CAP_GEN(mdev, disable_local_lb)) - mutex_init(&dev->lb_mutex); + return 0; +} + +static void __mlx5_ib_remove(struct mlx5_ib_dev *dev, + const struct mlx5_ib_profile *profile, + int stage) +{ + /* Number of stages to cleanup */ + while (stage) { + stage--; + if (profile->stage[stage].cleanup) + profile->stage[stage].cleanup(dev); + } + + ib_dealloc_device((struct ib_device *)dev); +} + +static void *mlx5_ib_add_slave_port(struct mlx5_core_dev *mdev, u8 port_num); + +static void *__mlx5_ib_add(struct mlx5_core_dev *mdev, + const struct mlx5_ib_profile *profile) +{ + struct mlx5_ib_dev *dev; + int err; + int i; + + printk_once(KERN_INFO "%s", mlx5_version); + dev = (struct mlx5_ib_dev *)ib_alloc_device(sizeof(*dev)); + if (!dev) + return NULL; + + dev->mdev = mdev; + dev->num_ports = max(MLX5_CAP_GEN(mdev, num_ports), + MLX5_CAP_GEN(mdev, num_vhca_ports)); + + for (i = 0; i < MLX5_IB_STAGE_MAX; i++) { + if (profile->stage[i].init) { + err = profile->stage[i].init(dev); + if (err) + goto err_out; + } + } + + dev->profile = profile; dev->ib_active = true; return dev; -err_delay_drop: - cancel_delay_drop(dev); - destroy_umrc_res(dev); +err_out: + __mlx5_ib_remove(dev, profile, i); -err_dev: - ib_unregister_device(&dev->ib_dev); + return NULL; +} -err_fp_bfreg: - mlx5_free_bfreg(dev->mdev, &dev->fp_bfreg); +static const struct mlx5_ib_profile pf_profile = { + STAGE_CREATE(MLX5_IB_STAGE_INIT, + mlx5_ib_stage_init_init, + mlx5_ib_stage_init_cleanup), + STAGE_CREATE(MLX5_IB_STAGE_CAPS, + mlx5_ib_stage_caps_init, + NULL), + STAGE_CREATE(MLX5_IB_STAGE_ROCE, + mlx5_ib_stage_roce_init, + mlx5_ib_stage_roce_cleanup), + STAGE_CREATE(MLX5_IB_STAGE_DEVICE_RESOURCES, + mlx5_ib_stage_dev_res_init, + mlx5_ib_stage_dev_res_cleanup), + STAGE_CREATE(MLX5_IB_STAGE_ODP, + mlx5_ib_stage_odp_init, + NULL), + STAGE_CREATE(MLX5_IB_STAGE_COUNTERS, + mlx5_ib_stage_counters_init, + mlx5_ib_stage_counters_cleanup), + STAGE_CREATE(MLX5_IB_STAGE_CONG_DEBUGFS, + mlx5_ib_stage_cong_debugfs_init, + mlx5_ib_stage_cong_debugfs_cleanup), + STAGE_CREATE(MLX5_IB_STAGE_UAR, + mlx5_ib_stage_uar_init, + mlx5_ib_stage_uar_cleanup), + STAGE_CREATE(MLX5_IB_STAGE_BFREG, + mlx5_ib_stage_bfrag_init, + mlx5_ib_stage_bfrag_cleanup), + STAGE_CREATE(MLX5_IB_STAGE_IB_REG, + mlx5_ib_stage_ib_reg_init, + mlx5_ib_stage_ib_reg_cleanup), + STAGE_CREATE(MLX5_IB_STAGE_UMR_RESOURCES, + mlx5_ib_stage_umr_res_init, + mlx5_ib_stage_umr_res_cleanup), + STAGE_CREATE(MLX5_IB_STAGE_DELAY_DROP, + mlx5_ib_stage_delay_drop_init, + mlx5_ib_stage_delay_drop_cleanup), + STAGE_CREATE(MLX5_IB_STAGE_CLASS_ATTR, + mlx5_ib_stage_class_attr_init, + NULL), +}; -err_bfreg: - mlx5_free_bfreg(dev->mdev, &dev->bfreg); +static void *mlx5_ib_add_slave_port(struct mlx5_core_dev *mdev, u8 port_num) +{ + struct mlx5_ib_multiport_info *mpi; + struct mlx5_ib_dev *dev; + bool bound = false; + int err; -err_uar_page: - mlx5_put_uars_page(dev->mdev, dev->mdev->priv.uar); + mpi = kzalloc(sizeof(*mpi), GFP_KERNEL); + if (!mpi) + return NULL; -err_cong: - mlx5_ib_cleanup_cong_debugfs(dev); -err_cnt: - if (MLX5_CAP_GEN(dev->mdev, max_qp_cnt)) - mlx5_ib_dealloc_counters(dev); + mpi->mdev = mdev; -err_odp: - mlx5_ib_odp_remove_one(dev); + err = mlx5_query_nic_vport_system_image_guid(mdev, + &mpi->sys_image_guid); + if (err) { + kfree(mpi); + return NULL; + } -err_rsrc: - destroy_dev_resources(&dev->devr); + mutex_lock(&mlx5_ib_multiport_mutex); + list_for_each_entry(dev, &mlx5_ib_dev_list, ib_dev_list) { + if (dev->sys_image_guid == mpi->sys_image_guid) + bound = mlx5_ib_bind_slave_port(dev, mpi); -err_disable_eth: - if (ll == IB_LINK_LAYER_ETHERNET) { - mlx5_disable_eth(dev); - mlx5_remove_netdev_notifier(dev); + if (bound) { + rdma_roce_rescan_device(&dev->ib_dev); + break; + } } -err_free_port: - kfree(dev->port); + if (!bound) { + list_add_tail(&mpi->list, &mlx5_ib_unaffiliated_port_list); + dev_dbg(&mdev->pdev->dev, "no suitable IB device found to bind to, added to unaffiliated list.\n"); + } else { + mlx5_ib_dbg(dev, "bound port %u\n", port_num + 1); + } + mutex_unlock(&mlx5_ib_multiport_mutex); -err_dealloc: - ib_dealloc_device((struct ib_device *)dev); + return mpi; +} - return NULL; +static void *mlx5_ib_add(struct mlx5_core_dev *mdev) +{ + enum rdma_link_layer ll; + int port_type_cap; + + port_type_cap = MLX5_CAP_GEN(mdev, port_type); + ll = mlx5_port_type_cap_to_rdma_ll(port_type_cap); + + if (mlx5_core_is_mp_slave(mdev) && ll == IB_LINK_LAYER_ETHERNET) { + u8 port_num = mlx5_core_native_port_num(mdev) - 1; + + return mlx5_ib_add_slave_port(mdev, port_num); + } + + return __mlx5_ib_add(mdev, &pf_profile); } static void mlx5_ib_remove(struct mlx5_core_dev *mdev, void *context) { - struct mlx5_ib_dev *dev = context; - enum rdma_link_layer ll = mlx5_ib_port_link_layer(&dev->ib_dev, 1); + struct mlx5_ib_multiport_info *mpi; + struct mlx5_ib_dev *dev; - cancel_delay_drop(dev); - mlx5_remove_netdev_notifier(dev); - ib_unregister_device(&dev->ib_dev); - mlx5_free_bfreg(dev->mdev, &dev->fp_bfreg); - mlx5_free_bfreg(dev->mdev, &dev->bfreg); - mlx5_put_uars_page(dev->mdev, mdev->priv.uar); - mlx5_ib_cleanup_cong_debugfs(dev); - if (MLX5_CAP_GEN(dev->mdev, max_qp_cnt)) - mlx5_ib_dealloc_counters(dev); - destroy_umrc_res(dev); - mlx5_ib_odp_remove_one(dev); - destroy_dev_resources(&dev->devr); - if (ll == IB_LINK_LAYER_ETHERNET) - mlx5_disable_eth(dev); - kfree(dev->port); - ib_dealloc_device(&dev->ib_dev); + if (mlx5_core_is_mp_slave(mdev)) { + mpi = context; + mutex_lock(&mlx5_ib_multiport_mutex); + if (mpi->ibdev) + mlx5_ib_unbind_slave_port(mpi->ibdev, mpi); + list_del(&mpi->list); + mutex_unlock(&mlx5_ib_multiport_mutex); + return; + } + + dev = context; + __mlx5_ib_remove(dev, dev->profile, MLX5_IB_STAGE_MAX); } static struct mlx5_interface mlx5_ib_interface = { @@ -4274,6 +5051,10 @@ static int __init mlx5_ib_init(void) { int err; + mlx5_ib_event_wq = alloc_ordered_workqueue("mlx5_ib_event_wq", 0); + if (!mlx5_ib_event_wq) + return -ENOMEM; + mlx5_ib_odp_init(); err = mlx5_register_interface(&mlx5_ib_interface); @@ -4284,6 +5065,7 @@ static int __init mlx5_ib_init(void) static void __exit mlx5_ib_cleanup(void) { mlx5_unregister_interface(&mlx5_ib_interface); + destroy_workqueue(mlx5_ib_event_wq); } module_init(mlx5_ib_init); diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h index 2c5f3533bbc9..51228dfcfbe7 100644 --- a/drivers/infiniband/hw/mlx5/mlx5_ib.h +++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h @@ -77,6 +77,7 @@ enum mlx5_ib_mmap_cmd { MLX5_IB_MMAP_NC_PAGE = 3, /* 5 is chosen in order to be compatible with old versions of libmlx5 */ MLX5_IB_MMAP_CORE_CLOCK = 5, + MLX5_IB_MMAP_ALLOC_WC = 6, }; enum { @@ -112,6 +113,11 @@ enum { MLX5_TM_MAX_SGE = 1, }; +enum { + MLX5_IB_INVALID_UAR_INDEX = BIT(31), + MLX5_IB_INVALID_BFREG = BIT(31), +}; + struct mlx5_ib_vma_private_data { struct list_head list; struct vm_area_struct *vma; @@ -200,6 +206,8 @@ struct mlx5_ib_flow_db { * creates the actual hardware QP. */ #define MLX5_IB_QPT_HW_GSI IB_QPT_RESERVED2 +#define MLX5_IB_QPT_DCI IB_QPT_RESERVED3 +#define MLX5_IB_QPT_DCT IB_QPT_RESERVED4 #define MLX5_IB_WR_UMR IB_WR_RESERVED1 #define MLX5_IB_UMR_OCTOWORD 16 @@ -360,12 +368,18 @@ struct mlx5_bf { struct mlx5_sq_bfreg *bfreg; }; +struct mlx5_ib_dct { + struct mlx5_core_dct mdct; + u32 *in; +}; + struct mlx5_ib_qp { struct ib_qp ibqp; union { struct mlx5_ib_qp_trans trans_qp; struct mlx5_ib_raw_packet_qp raw_packet_qp; struct mlx5_ib_rss_qp rss_qp; + struct mlx5_ib_dct dct; }; struct mlx5_buf buf; @@ -404,6 +418,8 @@ struct mlx5_ib_qp { u32 rate_limit; u32 underlay_qpn; bool tunnel_offload_en; + /* storage for qp sub type when core qp type is IB_QPT_DRIVER */ + enum ib_qp_type qp_sub_type; }; struct mlx5_ib_cq_buf { @@ -636,10 +652,21 @@ struct mlx5_ib_counters { u32 num_q_counters; u32 num_cong_counters; u16 set_id; + bool set_id_valid; +}; + +struct mlx5_ib_multiport_info; + +struct mlx5_ib_multiport { + struct mlx5_ib_multiport_info *mpi; + /* To be held when accessing the multiport info */ + spinlock_t mpi_lock; }; struct mlx5_ib_port { struct mlx5_ib_counters cnts; + struct mlx5_ib_multiport mp; + struct mlx5_ib_dbg_cc_params *dbg_cc_params; }; struct mlx5_roce { @@ -651,12 +678,15 @@ struct mlx5_roce { struct notifier_block nb; atomic_t next_port; enum ib_port_state last_port_state; + struct mlx5_ib_dev *dev; + u8 native_port_num; }; struct mlx5_ib_dbg_param { int offset; struct mlx5_ib_dev *dev; struct dentry *dentry; + u8 port_num; }; enum mlx5_ib_dbg_cc_types { @@ -709,10 +739,50 @@ struct mlx5_ib_delay_drop { struct mlx5_ib_dbg_delay_drop *dbg; }; +enum mlx5_ib_stages { + MLX5_IB_STAGE_INIT, + MLX5_IB_STAGE_CAPS, + MLX5_IB_STAGE_ROCE, + MLX5_IB_STAGE_DEVICE_RESOURCES, + MLX5_IB_STAGE_ODP, + MLX5_IB_STAGE_COUNTERS, + MLX5_IB_STAGE_CONG_DEBUGFS, + MLX5_IB_STAGE_UAR, + MLX5_IB_STAGE_BFREG, + MLX5_IB_STAGE_IB_REG, + MLX5_IB_STAGE_UMR_RESOURCES, + MLX5_IB_STAGE_DELAY_DROP, + MLX5_IB_STAGE_CLASS_ATTR, + MLX5_IB_STAGE_MAX, +}; + +struct mlx5_ib_stage { + int (*init)(struct mlx5_ib_dev *dev); + void (*cleanup)(struct mlx5_ib_dev *dev); +}; + +#define STAGE_CREATE(_stage, _init, _cleanup) \ + .stage[_stage] = {.init = _init, .cleanup = _cleanup} + +struct mlx5_ib_profile { + struct mlx5_ib_stage stage[MLX5_IB_STAGE_MAX]; +}; + +struct mlx5_ib_multiport_info { + struct list_head list; + struct mlx5_ib_dev *ibdev; + struct mlx5_core_dev *mdev; + struct completion unref_comp; + u64 sys_image_guid; + u32 mdev_refcnt; + bool is_master; + bool unaffiliate; +}; + struct mlx5_ib_dev { struct ib_device ib_dev; struct mlx5_core_dev *mdev; - struct mlx5_roce roce; + struct mlx5_roce roce[MLX5_MAX_PORTS]; int num_ports; /* serialize update of capability mask */ @@ -746,12 +816,14 @@ struct mlx5_ib_dev { struct mlx5_sq_bfreg bfreg; struct mlx5_sq_bfreg fp_bfreg; struct mlx5_ib_delay_drop delay_drop; - struct mlx5_ib_dbg_cc_params *dbg_cc_params; + const struct mlx5_ib_profile *profile; /* protect the user_td */ struct mutex lb_mutex; u32 user_td; u8 umr_fence; + struct list_head ib_dev_list; + u64 sys_image_guid; }; static inline struct mlx5_ib_cq *to_mibcq(struct mlx5_core_cq *mcq) @@ -956,13 +1028,14 @@ struct ib_rwq_ind_table *mlx5_ib_create_rwq_ind_table(struct ib_device *device, struct ib_rwq_ind_table_init_attr *init_attr, struct ib_udata *udata); int mlx5_ib_destroy_rwq_ind_table(struct ib_rwq_ind_table *wq_ind_table); +bool mlx5_ib_dc_atomic_is_supported(struct mlx5_ib_dev *dev); + #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING void mlx5_ib_internal_fill_odp_caps(struct mlx5_ib_dev *dev); void mlx5_ib_pfault(struct mlx5_core_dev *mdev, void *context, struct mlx5_pagefault *pfault); int mlx5_ib_odp_init_one(struct mlx5_ib_dev *ibdev); -void mlx5_ib_odp_remove_one(struct mlx5_ib_dev *ibdev); int __init mlx5_ib_odp_init(void); void mlx5_ib_odp_cleanup(void); void mlx5_ib_invalidate_range(struct ib_umem *umem, unsigned long start, @@ -977,7 +1050,6 @@ static inline void mlx5_ib_internal_fill_odp_caps(struct mlx5_ib_dev *dev) } static inline int mlx5_ib_odp_init_one(struct mlx5_ib_dev *ibdev) { return 0; } -static inline void mlx5_ib_odp_remove_one(struct mlx5_ib_dev *ibdev) {} static inline int mlx5_ib_odp_init(void) { return 0; } static inline void mlx5_ib_odp_cleanup(void) {} static inline void mlx5_odp_init_mr_cache_entry(struct mlx5_cache_ent *ent) {} @@ -1001,8 +1073,8 @@ __be16 mlx5_get_roce_udp_sport(struct mlx5_ib_dev *dev, u8 port_num, int mlx5_get_roce_gid_type(struct mlx5_ib_dev *dev, u8 port_num, int index, enum ib_gid_type *gid_type); -void mlx5_ib_cleanup_cong_debugfs(struct mlx5_ib_dev *dev); -int mlx5_ib_init_cong_debugfs(struct mlx5_ib_dev *dev); +void mlx5_ib_cleanup_cong_debugfs(struct mlx5_ib_dev *dev, u8 port_num); +int mlx5_ib_init_cong_debugfs(struct mlx5_ib_dev *dev, u8 port_num); /* GSI QP helper functions */ struct ib_qp *mlx5_ib_gsi_create_qp(struct ib_pd *pd, @@ -1021,6 +1093,15 @@ void mlx5_ib_gsi_pkey_change(struct mlx5_ib_gsi_qp *gsi); int mlx5_ib_generate_wc(struct ib_cq *ibcq, struct ib_wc *wc); +void mlx5_ib_free_bfreg(struct mlx5_ib_dev *dev, struct mlx5_bfreg_info *bfregi, + int bfregn); +struct mlx5_ib_dev *mlx5_ib_get_ibdev_from_mpi(struct mlx5_ib_multiport_info *mpi); +struct mlx5_core_dev *mlx5_ib_get_native_port_mdev(struct mlx5_ib_dev *dev, + u8 ib_port_num, + u8 *native_port_num); +void mlx5_ib_put_native_port_mdev(struct mlx5_ib_dev *dev, + u8 port_num); + static inline void init_query_mad(struct ib_smp *mad) { mad->base_version = 1; @@ -1113,10 +1194,10 @@ static inline int get_uars_per_sys_page(struct mlx5_ib_dev *dev, bool lib_suppor MLX5_UARS_IN_PAGE : 1; } -static inline int get_num_uars(struct mlx5_ib_dev *dev, - struct mlx5_bfreg_info *bfregi) +static inline int get_num_static_uars(struct mlx5_ib_dev *dev, + struct mlx5_bfreg_info *bfregi) { - return get_uars_per_sys_page(dev, bfregi->lib_uar_4k) * bfregi->num_sys_pages; + return get_uars_per_sys_page(dev, bfregi->lib_uar_4k) * bfregi->num_static_sys_pages; } #endif /* MLX5_IB_H */ diff --git a/drivers/infiniband/hw/mlx5/mr.c b/drivers/infiniband/hw/mlx5/mr.c index d109fe8290a7..556e015678de 100644 --- a/drivers/infiniband/hw/mlx5/mr.c +++ b/drivers/infiniband/hw/mlx5/mr.c @@ -1206,6 +1206,9 @@ struct ib_mr *mlx5_ib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, int err; bool use_umr = true; + if (!IS_ENABLED(CONFIG_INFINIBAND_USER_MEM)) + return ERR_PTR(-EINVAL); + mlx5_ib_dbg(dev, "start 0x%llx, virt_addr 0x%llx, length 0x%llx, access_flags 0x%x\n", start, virt_addr, length, access_flags); diff --git a/drivers/infiniband/hw/mlx5/odp.c b/drivers/infiniband/hw/mlx5/odp.c index e2197bdda89c..f1a87a690a4c 100644 --- a/drivers/infiniband/hw/mlx5/odp.c +++ b/drivers/infiniband/hw/mlx5/odp.c @@ -1207,10 +1207,6 @@ int mlx5_ib_odp_init_one(struct mlx5_ib_dev *dev) { int ret; - ret = init_srcu_struct(&dev->mr_srcu); - if (ret) - return ret; - if (dev->odp_caps.general_caps & IB_ODP_SUPPORT_IMPLICIT) { ret = mlx5_cmd_null_mkey(dev->mdev, &dev->null_mkey); if (ret) { @@ -1222,11 +1218,6 @@ int mlx5_ib_odp_init_one(struct mlx5_ib_dev *dev) return 0; } -void mlx5_ib_odp_remove_one(struct mlx5_ib_dev *dev) -{ - cleanup_srcu_struct(&dev->mr_srcu); -} - int mlx5_ib_odp_init(void) { mlx5_imr_ksm_entries = BIT_ULL(get_order(TASK_SIZE) - diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c index 31ad28853efa..ae36db3d0deb 100644 --- a/drivers/infiniband/hw/mlx5/qp.c +++ b/drivers/infiniband/hw/mlx5/qp.c @@ -493,7 +493,7 @@ enum { static int max_bfregs(struct mlx5_ib_dev *dev, struct mlx5_bfreg_info *bfregi) { - return get_num_uars(dev, bfregi) * MLX5_NON_FP_BFREGS_PER_UAR; + return get_num_static_uars(dev, bfregi) * MLX5_NON_FP_BFREGS_PER_UAR; } static int num_med_bfreg(struct mlx5_ib_dev *dev, @@ -581,7 +581,7 @@ static int alloc_bfreg(struct mlx5_ib_dev *dev, return bfregn; } -static void free_bfreg(struct mlx5_ib_dev *dev, struct mlx5_bfreg_info *bfregi, int bfregn) +void mlx5_ib_free_bfreg(struct mlx5_ib_dev *dev, struct mlx5_bfreg_info *bfregi, int bfregn) { mutex_lock(&bfregi->lock); bfregi->count[bfregn]--; @@ -613,6 +613,7 @@ static int to_mlx5_st(enum ib_qp_type type) case IB_QPT_XRC_TGT: return MLX5_QP_ST_XRC; case IB_QPT_SMI: return MLX5_QP_ST_QP0; case MLX5_IB_QPT_HW_GSI: return MLX5_QP_ST_QP1; + case MLX5_IB_QPT_DCI: return MLX5_QP_ST_DCI; case IB_QPT_RAW_IPV6: return MLX5_QP_ST_RAW_IPV6; case IB_QPT_RAW_PACKET: case IB_QPT_RAW_ETHERTYPE: return MLX5_QP_ST_RAW_ETHERTYPE; @@ -627,7 +628,8 @@ static void mlx5_ib_unlock_cqs(struct mlx5_ib_cq *send_cq, struct mlx5_ib_cq *recv_cq); static int bfregn_to_uar_index(struct mlx5_ib_dev *dev, - struct mlx5_bfreg_info *bfregi, int bfregn) + struct mlx5_bfreg_info *bfregi, int bfregn, + bool dyn_bfreg) { int bfregs_per_sys_page; int index_of_sys_page; @@ -637,8 +639,16 @@ static int bfregn_to_uar_index(struct mlx5_ib_dev *dev, MLX5_NON_FP_BFREGS_PER_UAR; index_of_sys_page = bfregn / bfregs_per_sys_page; - offset = bfregn % bfregs_per_sys_page / MLX5_NON_FP_BFREGS_PER_UAR; + if (dyn_bfreg) { + index_of_sys_page += bfregi->num_static_sys_pages; + if (bfregn > bfregi->num_dyn_bfregs || + bfregi->sys_pages[index_of_sys_page] == MLX5_IB_INVALID_UAR_INDEX) { + mlx5_ib_dbg(dev, "Invalid dynamic uar index\n"); + return -EINVAL; + } + } + offset = bfregn % bfregs_per_sys_page / MLX5_NON_FP_BFREGS_PER_UAR; return bfregi->sys_pages[index_of_sys_page] + offset; } @@ -764,7 +774,7 @@ static int create_user_qp(struct mlx5_ib_dev *dev, struct ib_pd *pd, struct mlx5_ib_create_qp ucmd; struct mlx5_ib_ubuffer *ubuffer = &base->ubuffer; int page_shift = 0; - int uar_index; + int uar_index = 0; int npages; u32 offset = 0; int bfregn; @@ -780,12 +790,20 @@ static int create_user_qp(struct mlx5_ib_dev *dev, struct ib_pd *pd, } context = to_mucontext(pd->uobject->context); - /* - * TBD: should come from the verbs when we have the API - */ - if (qp->flags & MLX5_IB_QP_CROSS_CHANNEL) + if (ucmd.flags & MLX5_QP_FLAG_BFREG_INDEX) { + uar_index = bfregn_to_uar_index(dev, &context->bfregi, + ucmd.bfreg_index, true); + if (uar_index < 0) + return uar_index; + + bfregn = MLX5_IB_INVALID_BFREG; + } else if (qp->flags & MLX5_IB_QP_CROSS_CHANNEL) { + /* + * TBD: should come from the verbs when we have the API + */ /* In CROSS_CHANNEL CQ and QP must use the same UAR */ bfregn = MLX5_CROSS_CHANNEL_BFREG; + } else { bfregn = alloc_bfreg(dev, &context->bfregi, MLX5_IB_LATENCY_CLASS_HIGH); if (bfregn < 0) { @@ -804,8 +822,10 @@ static int create_user_qp(struct mlx5_ib_dev *dev, struct ib_pd *pd, } } - uar_index = bfregn_to_uar_index(dev, &context->bfregi, bfregn); mlx5_ib_dbg(dev, "bfregn 0x%x, uar_index 0x%x\n", bfregn, uar_index); + if (bfregn != MLX5_IB_INVALID_BFREG) + uar_index = bfregn_to_uar_index(dev, &context->bfregi, bfregn, + false); qp->rq.offset = 0; qp->sq.wqe_shift = ilog2(MLX5_SEND_WQE_BB); @@ -845,7 +865,10 @@ static int create_user_qp(struct mlx5_ib_dev *dev, struct ib_pd *pd, MLX5_SET(qpc, qpc, page_offset, offset); MLX5_SET(qpc, qpc, uar_page, uar_index); - resp->bfreg_index = adjust_bfregn(dev, &context->bfregi, bfregn); + if (bfregn != MLX5_IB_INVALID_BFREG) + resp->bfreg_index = adjust_bfregn(dev, &context->bfregi, bfregn); + else + resp->bfreg_index = MLX5_IB_INVALID_BFREG; qp->bfregn = bfregn; err = mlx5_ib_db_map_user(context, ucmd.db_addr, &qp->db); @@ -874,7 +897,8 @@ err_umem: ib_umem_release(ubuffer->umem); err_bfreg: - free_bfreg(dev, &context->bfregi, bfregn); + if (bfregn != MLX5_IB_INVALID_BFREG) + mlx5_ib_free_bfreg(dev, &context->bfregi, bfregn); return err; } @@ -887,7 +911,13 @@ static void destroy_qp_user(struct mlx5_ib_dev *dev, struct ib_pd *pd, mlx5_ib_db_unmap_user(context, &qp->db); if (base->ubuffer.umem) ib_umem_release(base->ubuffer.umem); - free_bfreg(dev, &context->bfregi, qp->bfregn); + + /* + * Free only the BFREGs which are handled by the kernel. + * BFREGs of UARs allocated dynamically are handled by user. + */ + if (qp->bfregn != MLX5_IB_INVALID_BFREG) + mlx5_ib_free_bfreg(dev, &context->bfregi, qp->bfregn); } static int create_kernel_qp(struct mlx5_ib_dev *dev, @@ -1015,6 +1045,7 @@ static void destroy_qp_kernel(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp) static u32 get_rx_type(struct mlx5_ib_qp *qp, struct ib_qp_init_attr *attr) { if (attr->srq || (attr->qp_type == IB_QPT_XRC_TGT) || + (attr->qp_type == MLX5_IB_QPT_DCI) || (attr->qp_type == IB_QPT_XRC_INI)) return MLX5_SRQ_RQ; else if (!qp->has_rq) @@ -2086,20 +2117,108 @@ static const char *ib_qp_type_str(enum ib_qp_type type) return "IB_QPT_RAW_PACKET"; case MLX5_IB_QPT_REG_UMR: return "MLX5_IB_QPT_REG_UMR"; + case IB_QPT_DRIVER: + return "IB_QPT_DRIVER"; case IB_QPT_MAX: default: return "Invalid QP type"; } } +static struct ib_qp *mlx5_ib_create_dct(struct ib_pd *pd, + struct ib_qp_init_attr *attr, + struct mlx5_ib_create_qp *ucmd) +{ + struct mlx5_ib_dev *dev; + struct mlx5_ib_qp *qp; + int err = 0; + u32 uidx = MLX5_IB_DEFAULT_UIDX; + void *dctc; + + if (!attr->srq || !attr->recv_cq) + return ERR_PTR(-EINVAL); + + dev = to_mdev(pd->device); + + err = get_qp_user_index(to_mucontext(pd->uobject->context), + ucmd, sizeof(*ucmd), &uidx); + if (err) + return ERR_PTR(err); + + qp = kzalloc(sizeof(*qp), GFP_KERNEL); + if (!qp) + return ERR_PTR(-ENOMEM); + + qp->dct.in = kzalloc(MLX5_ST_SZ_BYTES(create_dct_in), GFP_KERNEL); + if (!qp->dct.in) { + err = -ENOMEM; + goto err_free; + } + + dctc = MLX5_ADDR_OF(create_dct_in, qp->dct.in, dct_context_entry); + qp->qp_sub_type = MLX5_IB_QPT_DCT; + MLX5_SET(dctc, dctc, pd, to_mpd(pd)->pdn); + MLX5_SET(dctc, dctc, srqn_xrqn, to_msrq(attr->srq)->msrq.srqn); + MLX5_SET(dctc, dctc, cqn, to_mcq(attr->recv_cq)->mcq.cqn); + MLX5_SET64(dctc, dctc, dc_access_key, ucmd->access_key); + MLX5_SET(dctc, dctc, user_index, uidx); + + qp->state = IB_QPS_RESET; + + return &qp->ibqp; +err_free: + kfree(qp); + return ERR_PTR(err); +} + +static int set_mlx_qp_type(struct mlx5_ib_dev *dev, + struct ib_qp_init_attr *init_attr, + struct mlx5_ib_create_qp *ucmd, + struct ib_udata *udata) +{ + enum { MLX_QP_FLAGS = MLX5_QP_FLAG_TYPE_DCT | MLX5_QP_FLAG_TYPE_DCI }; + int err; + + if (!udata) + return -EINVAL; + + if (udata->inlen < sizeof(*ucmd)) { + mlx5_ib_dbg(dev, "create_qp user command is smaller than expected\n"); + return -EINVAL; + } + err = ib_copy_from_udata(ucmd, udata, sizeof(*ucmd)); + if (err) + return err; + + if ((ucmd->flags & MLX_QP_FLAGS) == MLX5_QP_FLAG_TYPE_DCI) { + init_attr->qp_type = MLX5_IB_QPT_DCI; + } else { + if ((ucmd->flags & MLX_QP_FLAGS) == MLX5_QP_FLAG_TYPE_DCT) { + init_attr->qp_type = MLX5_IB_QPT_DCT; + } else { + mlx5_ib_dbg(dev, "Invalid QP flags\n"); + return -EINVAL; + } + } + + if (!MLX5_CAP_GEN(dev->mdev, dct)) { + mlx5_ib_dbg(dev, "DC transport is not supported\n"); + return -EOPNOTSUPP; + } + + return 0; +} + struct ib_qp *mlx5_ib_create_qp(struct ib_pd *pd, - struct ib_qp_init_attr *init_attr, + struct ib_qp_init_attr *verbs_init_attr, struct ib_udata *udata) { struct mlx5_ib_dev *dev; struct mlx5_ib_qp *qp; u16 xrcdn = 0; int err; + struct ib_qp_init_attr mlx_init_attr; + struct ib_qp_init_attr *init_attr = verbs_init_attr; if (pd) { dev = to_mdev(pd->device); @@ -2124,6 +2243,26 @@ struct ib_qp *mlx5_ib_create_qp(struct ib_pd *pd, dev = to_mdev(to_mxrcd(init_attr->xrcd)->ibxrcd.device); } + if (init_attr->qp_type == IB_QPT_DRIVER) { + struct mlx5_ib_create_qp ucmd; + + init_attr = &mlx_init_attr; + memcpy(init_attr, verbs_init_attr, sizeof(*verbs_init_attr)); + err = set_mlx_qp_type(dev, init_attr, &ucmd, udata); + if (err) + return ERR_PTR(err); + + if (init_attr->qp_type == MLX5_IB_QPT_DCI) { + if (init_attr->cap.max_recv_wr || + init_attr->cap.max_recv_sge) { + mlx5_ib_dbg(dev, "DCI QP requires zero size receive queue\n"); + return ERR_PTR(-EINVAL); + } + } else { + return mlx5_ib_create_dct(pd, init_attr, &ucmd); + } + } + switch (init_attr->qp_type) { case IB_QPT_XRC_TGT: case IB_QPT_XRC_INI: @@ -2145,6 +2284,7 @@ struct ib_qp *mlx5_ib_create_qp(struct ib_pd *pd, case IB_QPT_SMI: case MLX5_IB_QPT_HW_GSI: case MLX5_IB_QPT_REG_UMR: + case MLX5_IB_QPT_DCI: qp = kzalloc(sizeof(*qp), GFP_KERNEL); if (!qp) return ERR_PTR(-ENOMEM); @@ -2185,9 +2325,31 @@ struct ib_qp *mlx5_ib_create_qp(struct ib_pd *pd, return ERR_PTR(-EINVAL); } + if (verbs_init_attr->qp_type == IB_QPT_DRIVER) + qp->qp_sub_type = init_attr->qp_type; + return &qp->ibqp; } +static int mlx5_ib_destroy_dct(struct mlx5_ib_qp *mqp) +{ + struct mlx5_ib_dev *dev = to_mdev(mqp->ibqp.device); + + if (mqp->state == IB_QPS_RTR) { + int err; + + err = mlx5_core_destroy_dct(dev->mdev, &mqp->dct.mdct); + if (err) { + mlx5_ib_warn(dev, "failed to destroy DCT %d\n", err); + return err; + } + } + + kfree(mqp->dct.in); + kfree(mqp); + return 0; +} + int mlx5_ib_destroy_qp(struct ib_qp *qp) { struct mlx5_ib_dev *dev = to_mdev(qp->device); @@ -2196,6 +2358,9 @@ int mlx5_ib_destroy_qp(struct ib_qp *qp) if (unlikely(qp->qp_type == IB_QPT_GSI)) return mlx5_ib_gsi_destroy_qp(qp); + if (mqp->qp_sub_type == MLX5_IB_QPT_DCT) + return mlx5_ib_destroy_dct(mqp); + destroy_qp_common(dev, mqp); kfree(mqp); @@ -2763,7 +2928,8 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp, if (!context) return -ENOMEM; - err = to_mlx5_st(ibqp->qp_type); + err = to_mlx5_st(ibqp->qp_type == IB_QPT_DRIVER ? + qp->qp_sub_type : ibqp->qp_type); if (err < 0) { mlx5_ib_dbg(dev, "unsupported qp type %d\n", ibqp->qp_type); goto out; @@ -2796,8 +2962,9 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp, (ibqp->qp_type == IB_QPT_XRC_INI) || (ibqp->qp_type == IB_QPT_XRC_TGT)) { if (mlx5_lag_is_active(dev->mdev)) { + u8 p = mlx5_core_native_port_num(dev->mdev); tx_affinity = (unsigned int)atomic_add_return(1, - &dev->roce.next_port) % + &dev->roce[p].next_port) % MLX5_MAX_PORTS + 1; context->flags |= cpu_to_be32(tx_affinity << 24); } @@ -2922,7 +3089,8 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp, mlx5_cur = to_mlx5_state(cur_state); mlx5_new = to_mlx5_state(new_state); - mlx5_st = to_mlx5_st(ibqp->qp_type); + mlx5_st = to_mlx5_st(ibqp->qp_type == IB_QPT_DRIVER ? + qp->qp_sub_type : ibqp->qp_type); if (mlx5_st < 0) goto out; @@ -2994,6 +3162,139 @@ out: return err; } +static inline bool is_valid_mask(int mask, int req, int opt) +{ + if ((mask & req) != req) + return false; + + if (mask & ~(req | opt)) + return false; + + return true; +} + +/* check valid transition for driver QP types + * for now the only QP type that this function supports is DCI + */ +static bool modify_dci_qp_is_ok(enum ib_qp_state cur_state, enum ib_qp_state new_state, + enum ib_qp_attr_mask attr_mask) +{ + int req = IB_QP_STATE; + int opt = 0; + + if (cur_state == IB_QPS_RESET && new_state == IB_QPS_INIT) { + req |= IB_QP_PKEY_INDEX | IB_QP_PORT; + return is_valid_mask(attr_mask, req, opt); + } else if (cur_state == IB_QPS_INIT && new_state == IB_QPS_INIT) { + opt = IB_QP_PKEY_INDEX | IB_QP_PORT; + return is_valid_mask(attr_mask, req, opt); + } else if (cur_state == IB_QPS_INIT && new_state == IB_QPS_RTR) { + req |= IB_QP_PATH_MTU; + opt = IB_QP_PKEY_INDEX; + return is_valid_mask(attr_mask, req, opt); + } else if (cur_state == IB_QPS_RTR && new_state == IB_QPS_RTS) { + req |= IB_QP_TIMEOUT | IB_QP_RETRY_CNT | IB_QP_RNR_RETRY | + IB_QP_MAX_QP_RD_ATOMIC | IB_QP_SQ_PSN; + opt = IB_QP_MIN_RNR_TIMER; + return is_valid_mask(attr_mask, req, opt); + } else if (cur_state == IB_QPS_RTS && new_state == IB_QPS_RTS) { + opt = IB_QP_MIN_RNR_TIMER; + return is_valid_mask(attr_mask, req, opt); + } else if (cur_state != IB_QPS_RESET && new_state == IB_QPS_ERR) { + return is_valid_mask(attr_mask, req, opt); + } + return false; +} + +/* mlx5_ib_modify_dct: modify a DCT QP + * valid transitions are: + * RESET to INIT: must set access_flags, pkey_index and port + * INIT to RTR : must set min_rnr_timer, tclass, flow_label, + * mtu, gid_index and hop_limit + * Other transitions and attributes are illegal + */ +static int mlx5_ib_modify_dct(struct ib_qp *ibqp, struct ib_qp_attr *attr, + int attr_mask, struct ib_udata *udata) +{ + struct mlx5_ib_qp *qp = to_mqp(ibqp); + struct mlx5_ib_dev *dev = to_mdev(ibqp->device); + enum ib_qp_state cur_state, new_state; + int err = 0; + int required = IB_QP_STATE; + void *dctc; + + if (!(attr_mask & IB_QP_STATE)) + return -EINVAL; + + cur_state = qp->state; + new_state = attr->qp_state; + + dctc = MLX5_ADDR_OF(create_dct_in, qp->dct.in, dct_context_entry); + if (cur_state == IB_QPS_RESET && new_state == IB_QPS_INIT) { + required |= IB_QP_ACCESS_FLAGS | IB_QP_PKEY_INDEX | IB_QP_PORT; + if (!is_valid_mask(attr_mask, required, 0)) + return -EINVAL; + + if (attr->port_num == 0 || + attr->port_num > MLX5_CAP_GEN(dev->mdev, num_ports)) { + mlx5_ib_dbg(dev, "invalid port number %d. number of ports is %d\n", + attr->port_num, dev->num_ports); + return -EINVAL; + } + if (attr->qp_access_flags & IB_ACCESS_REMOTE_READ) + MLX5_SET(dctc, dctc, rre, 1); + if (attr->qp_access_flags & IB_ACCESS_REMOTE_WRITE) + MLX5_SET(dctc, dctc, rwe, 1); + if (attr->qp_access_flags & IB_ACCESS_REMOTE_ATOMIC) { + if (!mlx5_ib_dc_atomic_is_supported(dev)) + return -EOPNOTSUPP; + MLX5_SET(dctc, dctc, rae, 1); + MLX5_SET(dctc, dctc, atomic_mode, MLX5_ATOMIC_MODE_DCT_CX); + } + MLX5_SET(dctc, dctc, pkey_index, attr->pkey_index); + MLX5_SET(dctc, dctc, port, attr->port_num); + MLX5_SET(dctc, dctc, counter_set_id, dev->port[attr->port_num - 1].cnts.set_id); + + } else if (cur_state == IB_QPS_INIT && new_state == IB_QPS_RTR) { + struct mlx5_ib_modify_qp_resp resp = {}; + u32 min_resp_len = offsetof(typeof(resp), dctn) + + sizeof(resp.dctn); + + if (udata->outlen < min_resp_len) + return -EINVAL; + resp.response_length = min_resp_len; + + required |= IB_QP_MIN_RNR_TIMER | IB_QP_AV | IB_QP_PATH_MTU; + if (!is_valid_mask(attr_mask, required, 0)) + return -EINVAL; + MLX5_SET(dctc, dctc, min_rnr_nak, attr->min_rnr_timer); + MLX5_SET(dctc, dctc, tclass, attr->ah_attr.grh.traffic_class); + MLX5_SET(dctc, dctc, flow_label, attr->ah_attr.grh.flow_label); + MLX5_SET(dctc, dctc, mtu, attr->path_mtu); + MLX5_SET(dctc, dctc, my_addr_index, attr->ah_attr.grh.sgid_index); + MLX5_SET(dctc, dctc, hop_limit, attr->ah_attr.grh.hop_limit); + + err = mlx5_core_create_dct(dev->mdev, &qp->dct.mdct, qp->dct.in, + MLX5_ST_SZ_BYTES(create_dct_in)); + if (err) + return err; + resp.dctn = qp->dct.mdct.mqp.qpn; + err = ib_copy_to_udata(udata, &resp, resp.response_length); + if (err) { + mlx5_core_destroy_dct(dev->mdev, &qp->dct.mdct); + return err; + } + } else { + mlx5_ib_warn(dev, "Modify DCT: Invalid transition from %d to %d\n", cur_state, new_state); + return -EINVAL; + } + if (err) + qp->state = IB_QPS_ERR; + else + qp->state = new_state; + return err; +} + int mlx5_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask, struct ib_udata *udata) { @@ -3011,8 +3312,14 @@ int mlx5_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, if (unlikely(ibqp->qp_type == IB_QPT_GSI)) return mlx5_ib_gsi_modify_qp(ibqp, attr, attr_mask); - qp_type = (unlikely(ibqp->qp_type == MLX5_IB_QPT_HW_GSI)) ? - IB_QPT_GSI : ibqp->qp_type; + if (ibqp->qp_type == IB_QPT_DRIVER) + qp_type = qp->qp_sub_type; + else + qp_type = (unlikely(ibqp->qp_type == MLX5_IB_QPT_HW_GSI)) ? + IB_QPT_GSI : ibqp->qp_type; + + if (qp_type == MLX5_IB_QPT_DCT) + return mlx5_ib_modify_dct(ibqp, attr, attr_mask, udata); mutex_lock(&qp->mutex); @@ -3031,15 +3338,21 @@ int mlx5_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, goto out; } } else if (qp_type != MLX5_IB_QPT_REG_UMR && - !ib_modify_qp_is_ok(cur_state, new_state, qp_type, attr_mask, ll)) { + qp_type != MLX5_IB_QPT_DCI && + !ib_modify_qp_is_ok(cur_state, new_state, qp_type, attr_mask, ll)) { mlx5_ib_dbg(dev, "invalid QP state transition from %d to %d, qp_type %d, attr_mask 0x%x\n", cur_state, new_state, ibqp->qp_type, attr_mask); goto out; + } else if (qp_type == MLX5_IB_QPT_DCI && + !modify_dci_qp_is_ok(cur_state, new_state, attr_mask)) { + mlx5_ib_dbg(dev, "invalid QP state transition from %d to %d, qp_type %d, attr_mask 0x%x\n", + cur_state, new_state, qp_type, attr_mask); + goto out; } if ((attr_mask & IB_QP_PORT) && (attr->port_num == 0 || - attr->port_num > MLX5_CAP_GEN(dev->mdev, num_ports))) { + attr->port_num > dev->num_ports)) { mlx5_ib_dbg(dev, "invalid port number %d. number of ports is %d\n", attr->port_num, dev->num_ports); goto out; @@ -4358,14 +4671,13 @@ static void to_rdma_ah_attr(struct mlx5_ib_dev *ibdev, struct rdma_ah_attr *ah_attr, struct mlx5_qp_path *path) { - struct mlx5_core_dev *dev = ibdev->mdev; memset(ah_attr, 0, sizeof(*ah_attr)); ah_attr->type = rdma_ah_find_type(&ibdev->ib_dev, path->port); rdma_ah_set_port_num(ah_attr, path->port); if (rdma_ah_get_port_num(ah_attr) == 0 || - rdma_ah_get_port_num(ah_attr) > MLX5_CAP_GEN(dev, num_ports)) + rdma_ah_get_port_num(ah_attr) > ibdev->num_ports) return; rdma_ah_set_port_num(ah_attr, path->port); @@ -4578,6 +4890,71 @@ out: return err; } +static int mlx5_ib_dct_query_qp(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *mqp, + struct ib_qp_attr *qp_attr, int qp_attr_mask, + struct ib_qp_init_attr *qp_init_attr) +{ + struct mlx5_core_dct *dct = &mqp->dct.mdct; + u32 *out; + u32 access_flags = 0; + int outlen = MLX5_ST_SZ_BYTES(query_dct_out); + void *dctc; + int err; + int supported_mask = IB_QP_STATE | + IB_QP_ACCESS_FLAGS | + IB_QP_PORT | + IB_QP_MIN_RNR_TIMER | + IB_QP_AV | + IB_QP_PATH_MTU | + IB_QP_PKEY_INDEX; + + if (qp_attr_mask & ~supported_mask) + return -EINVAL; + if (mqp->state != IB_QPS_RTR) + return -EINVAL; + + out = kzalloc(outlen, GFP_KERNEL); + if (!out) + return -ENOMEM; + + err = mlx5_core_dct_query(dev->mdev, dct, out, outlen); + if (err) + goto out; + + dctc = MLX5_ADDR_OF(query_dct_out, out, dct_context_entry); + + if (qp_attr_mask & IB_QP_STATE) + qp_attr->qp_state = IB_QPS_RTR; + + if (qp_attr_mask & IB_QP_ACCESS_FLAGS) { + if (MLX5_GET(dctc, dctc, rre)) + access_flags |= IB_ACCESS_REMOTE_READ; + if (MLX5_GET(dctc, dctc, rwe)) + access_flags |= IB_ACCESS_REMOTE_WRITE; + if (MLX5_GET(dctc, dctc, rae)) + access_flags |= IB_ACCESS_REMOTE_ATOMIC; + qp_attr->qp_access_flags = access_flags; + } + + if (qp_attr_mask & IB_QP_PORT) + qp_attr->port_num = MLX5_GET(dctc, dctc, port); + if (qp_attr_mask & IB_QP_MIN_RNR_TIMER) + qp_attr->min_rnr_timer = MLX5_GET(dctc, dctc, min_rnr_nak); + if (qp_attr_mask & IB_QP_AV) { + qp_attr->ah_attr.grh.traffic_class = MLX5_GET(dctc, dctc, tclass); + qp_attr->ah_attr.grh.flow_label = MLX5_GET(dctc, dctc, flow_label); + qp_attr->ah_attr.grh.sgid_index = MLX5_GET(dctc, dctc, my_addr_index); + qp_attr->ah_attr.grh.hop_limit = MLX5_GET(dctc, dctc, hop_limit); + } + if (qp_attr_mask & IB_QP_PATH_MTU) + qp_attr->path_mtu = MLX5_GET(dctc, dctc, mtu); + if (qp_attr_mask & IB_QP_PKEY_INDEX) + qp_attr->pkey_index = MLX5_GET(dctc, dctc, pkey_index); +out: + kfree(out); + return err; +} + int mlx5_ib_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr, int qp_attr_mask, struct ib_qp_init_attr *qp_init_attr) { @@ -4597,6 +4974,10 @@ int mlx5_ib_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr, memset(qp_init_attr, 0, sizeof(*qp_init_attr)); memset(qp_attr, 0, sizeof(*qp_attr)); + if (unlikely(qp->qp_sub_type == MLX5_IB_QPT_DCT)) + return mlx5_ib_dct_query_qp(dev, qp, qp_attr, + qp_attr_mask, qp_init_attr); + mutex_lock(&qp->mutex); if (qp->ibqp.qp_type == IB_QPT_RAW_PACKET || diff --git a/drivers/infiniband/hw/mthca/mthca_memfree.c b/drivers/infiniband/hw/mthca/mthca_memfree.c index c6fe89d79248..2a41ed74add8 100644 --- a/drivers/infiniband/hw/mthca/mthca_memfree.c +++ b/drivers/infiniband/hw/mthca/mthca_memfree.c @@ -623,13 +623,12 @@ int mthca_alloc_db(struct mthca_dev *dev, enum mthca_db_type type, page = dev->db_tab->page + end; alloc: - page->db_rec = dma_alloc_coherent(&dev->pdev->dev, MTHCA_ICM_PAGE_SIZE, - &page->mapping, GFP_KERNEL); + page->db_rec = dma_zalloc_coherent(&dev->pdev->dev, MTHCA_ICM_PAGE_SIZE, + &page->mapping, GFP_KERNEL); if (!page->db_rec) { ret = -ENOMEM; goto out; } - memset(page->db_rec, 0, MTHCA_ICM_PAGE_SIZE); ret = mthca_MAP_ICM_page(dev, page->mapping, mthca_uarc_virt(dev, &dev->driver_uar, i)); diff --git a/drivers/infiniband/hw/nes/nes_cm.c b/drivers/infiniband/hw/nes/nes_cm.c index c56ca2a74df5..6cdfbf8c5674 100644 --- a/drivers/infiniband/hw/nes/nes_cm.c +++ b/drivers/infiniband/hw/nes/nes_cm.c @@ -1365,7 +1365,7 @@ static int mini_cm_del_listen(struct nes_cm_core *cm_core, static inline int mini_cm_accelerated(struct nes_cm_core *cm_core, struct nes_cm_node *cm_node) { - cm_node->accelerated = 1; + cm_node->accelerated = true; if (cm_node->accept_pend) { BUG_ON(!cm_node->listener); diff --git a/drivers/infiniband/hw/nes/nes_cm.h b/drivers/infiniband/hw/nes/nes_cm.h index d827d03e3941..b9cc02b4e8d5 100644 --- a/drivers/infiniband/hw/nes/nes_cm.h +++ b/drivers/infiniband/hw/nes/nes_cm.h @@ -279,7 +279,6 @@ struct nes_cm_tcp_context { u8 rcv_wscale; struct nes_cm_tsa_context tsa_cntxt; - struct timeval sent_ts; }; @@ -341,7 +340,7 @@ struct nes_cm_node { u16 mpa_frame_size; struct iw_cm_id *cm_id; struct list_head list; - int accelerated; + bool accelerated; struct nes_cm_listener *listener; enum nes_cm_conn_type conn_type; struct nes_vnic *nesvnic; diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_hw.c b/drivers/infiniband/hw/ocrdma/ocrdma_hw.c index 0ba695a88b62..9904918589a4 100644 --- a/drivers/infiniband/hw/ocrdma/ocrdma_hw.c +++ b/drivers/infiniband/hw/ocrdma/ocrdma_hw.c @@ -380,11 +380,10 @@ static int ocrdma_alloc_q(struct ocrdma_dev *dev, q->len = len; q->entry_size = entry_size; q->size = len * entry_size; - q->va = dma_alloc_coherent(&dev->nic_info.pdev->dev, q->size, - &q->dma, GFP_KERNEL); + q->va = dma_zalloc_coherent(&dev->nic_info.pdev->dev, q->size, + &q->dma, GFP_KERNEL); if (!q->va) return -ENOMEM; - memset(q->va, 0, q->size); return 0; } @@ -1819,12 +1818,11 @@ int ocrdma_mbx_create_cq(struct ocrdma_dev *dev, struct ocrdma_cq *cq, return -ENOMEM; ocrdma_init_mch(&cmd->cmd.req, OCRDMA_CMD_CREATE_CQ, OCRDMA_SUBSYS_COMMON, sizeof(*cmd)); - cq->va = dma_alloc_coherent(&pdev->dev, cq->len, &cq->pa, GFP_KERNEL); + cq->va = dma_zalloc_coherent(&pdev->dev, cq->len, &cq->pa, GFP_KERNEL); if (!cq->va) { status = -ENOMEM; goto mem_err; } - memset(cq->va, 0, cq->len); page_size = cq->len / hw_pages; cmd->cmd.pgsz_pgcnt = (page_size / OCRDMA_MIN_Q_PAGE_SIZE) << OCRDMA_CREATE_CQ_PAGE_SIZE_SHIFT; @@ -2212,10 +2210,9 @@ static int ocrdma_set_create_qp_sq_cmd(struct ocrdma_create_qp_req *cmd, qp->sq.max_cnt = max_wqe_allocated; len = (hw_pages * hw_page_size); - qp->sq.va = dma_alloc_coherent(&pdev->dev, len, &pa, GFP_KERNEL); + qp->sq.va = dma_zalloc_coherent(&pdev->dev, len, &pa, GFP_KERNEL); if (!qp->sq.va) return -EINVAL; - memset(qp->sq.va, 0, len); qp->sq.len = len; qp->sq.pa = pa; qp->sq.entry_size = dev->attr.wqe_size; @@ -2263,10 +2260,9 @@ static int ocrdma_set_create_qp_rq_cmd(struct ocrdma_create_qp_req *cmd, qp->rq.max_cnt = max_rqe_allocated; len = (hw_pages * hw_page_size); - qp->rq.va = dma_alloc_coherent(&pdev->dev, len, &pa, GFP_KERNEL); + qp->rq.va = dma_zalloc_coherent(&pdev->dev, len, &pa, GFP_KERNEL); if (!qp->rq.va) return -ENOMEM; - memset(qp->rq.va, 0, len); qp->rq.pa = pa; qp->rq.len = len; qp->rq.entry_size = dev->attr.rqe_size; @@ -2320,11 +2316,10 @@ static int ocrdma_set_create_qp_ird_cmd(struct ocrdma_create_qp_req *cmd, if (dev->attr.ird == 0) return 0; - qp->ird_q_va = dma_alloc_coherent(&pdev->dev, ird_q_len, - &pa, GFP_KERNEL); + qp->ird_q_va = dma_zalloc_coherent(&pdev->dev, ird_q_len, &pa, + GFP_KERNEL); if (!qp->ird_q_va) return -ENOMEM; - memset(qp->ird_q_va, 0, ird_q_len); ocrdma_build_q_pages(&cmd->ird_addr[0], dev->attr.num_ird_pages, pa, ird_page_size); for (; i < ird_q_len / dev->attr.rqe_size; i++) { diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_stats.c b/drivers/infiniband/hw/ocrdma/ocrdma_stats.c index e528d7acb7f6..24d20a4aa262 100644 --- a/drivers/infiniband/hw/ocrdma/ocrdma_stats.c +++ b/drivers/infiniband/hw/ocrdma/ocrdma_stats.c @@ -73,15 +73,13 @@ bool ocrdma_alloc_stats_resources(struct ocrdma_dev *dev) mem->size = max_t(u32, sizeof(struct ocrdma_rdma_stats_req), sizeof(struct ocrdma_rdma_stats_resp)); - mem->va = dma_alloc_coherent(&dev->nic_info.pdev->dev, mem->size, - &mem->pa, GFP_KERNEL); + mem->va = dma_zalloc_coherent(&dev->nic_info.pdev->dev, mem->size, + &mem->pa, GFP_KERNEL); if (!mem->va) { pr_err("%s: stats mbox allocation failed\n", __func__); return false; } - memset(mem->va, 0, mem->size); - /* Alloc debugfs mem */ mem->debugfs_mem = kzalloc(OCRDMA_MAX_DBGFS_MEM, GFP_KERNEL); if (!mem->debugfs_mem) @@ -834,7 +832,7 @@ void ocrdma_add_port_stats(struct ocrdma_dev *dev) dev->reset_stats.type = OCRDMA_RESET_STATS; dev->reset_stats.dev = dev; - if (!debugfs_create_file("reset_stats", S_IRUSR, dev->dir, + if (!debugfs_create_file("reset_stats", 0200, dev->dir, &dev->reset_stats, &ocrdma_dbg_ops)) goto err; diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c b/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c index 7866fd8051f6..8009bdad4e5b 100644 --- a/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c +++ b/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c @@ -461,7 +461,7 @@ retry: static inline int is_ucontext_pd(struct ocrdma_ucontext *uctx, struct ocrdma_pd *pd) { - return (uctx->cntxt_pd == pd ? true : false); + return (uctx->cntxt_pd == pd); } static int _ocrdma_dealloc_pd(struct ocrdma_dev *dev, @@ -550,13 +550,12 @@ struct ib_ucontext *ocrdma_alloc_ucontext(struct ib_device *ibdev, INIT_LIST_HEAD(&ctx->mm_head); mutex_init(&ctx->mm_list_lock); - ctx->ah_tbl.va = dma_alloc_coherent(&pdev->dev, map_len, - &ctx->ah_tbl.pa, GFP_KERNEL); + ctx->ah_tbl.va = dma_zalloc_coherent(&pdev->dev, map_len, + &ctx->ah_tbl.pa, GFP_KERNEL); if (!ctx->ah_tbl.va) { kfree(ctx); return ERR_PTR(-ENOMEM); } - memset(ctx->ah_tbl.va, 0, map_len); ctx->ah_tbl.len = map_len; memset(&resp, 0, sizeof(resp)); @@ -885,13 +884,12 @@ static int ocrdma_build_pbl_tbl(struct ocrdma_dev *dev, struct ocrdma_hw_mr *mr) return -ENOMEM; for (i = 0; i < mr->num_pbls; i++) { - va = dma_alloc_coherent(&pdev->dev, dma_len, &pa, GFP_KERNEL); + va = dma_zalloc_coherent(&pdev->dev, dma_len, &pa, GFP_KERNEL); if (!va) { ocrdma_free_mr_pbl_tbl(dev, mr); status = -ENOMEM; break; } - memset(va, 0, dma_len); mr->pbl_table[i].va = va; mr->pbl_table[i].pa = pa; } diff --git a/drivers/infiniband/hw/qedr/verbs.c b/drivers/infiniband/hw/qedr/verbs.c index b26aa88dab48..3b9c89848d66 100644 --- a/drivers/infiniband/hw/qedr/verbs.c +++ b/drivers/infiniband/hw/qedr/verbs.c @@ -604,12 +604,11 @@ static struct qedr_pbl *qedr_alloc_pbl_tbl(struct qedr_dev *dev, return ERR_PTR(-ENOMEM); for (i = 0; i < pbl_info->num_pbls; i++) { - va = dma_alloc_coherent(&pdev->dev, pbl_info->pbl_size, - &pa, flags); + va = dma_zalloc_coherent(&pdev->dev, pbl_info->pbl_size, + &pa, flags); if (!va) goto err; - memset(va, 0, pbl_info->pbl_size); pbl_table[i].va = va; pbl_table[i].pa = pa; } diff --git a/drivers/infiniband/hw/qib/qib.h b/drivers/infiniband/hw/qib/qib.h index 092ed8103842..0235f76bbc72 100644 --- a/drivers/infiniband/hw/qib/qib.h +++ b/drivers/infiniband/hw/qib/qib.h @@ -1428,8 +1428,6 @@ u64 qib_sps_ints(void); */ dma_addr_t qib_map_page(struct pci_dev *, struct page *, unsigned long, size_t, int); -const char *qib_get_unit_name(int unit); -const char *qib_get_card_name(struct rvt_dev_info *rdi); struct pci_dev *qib_get_pci_dev(struct rvt_dev_info *rdi); /* @@ -1488,15 +1486,15 @@ extern struct mutex qib_mutex; #define qib_dev_err(dd, fmt, ...) \ dev_err(&(dd)->pcidev->dev, "%s: " fmt, \ - qib_get_unit_name((dd)->unit), ##__VA_ARGS__) + rvt_get_ibdev_name(&(dd)->verbs_dev.rdi), ##__VA_ARGS__) #define qib_dev_warn(dd, fmt, ...) \ dev_warn(&(dd)->pcidev->dev, "%s: " fmt, \ - qib_get_unit_name((dd)->unit), ##__VA_ARGS__) + rvt_get_ibdev_name(&(dd)->verbs_dev.rdi), ##__VA_ARGS__) #define qib_dev_porterr(dd, port, fmt, ...) \ dev_err(&(dd)->pcidev->dev, "%s: IB%u:%u " fmt, \ - qib_get_unit_name((dd)->unit), (dd)->unit, (port), \ + rvt_get_ibdev_name(&(dd)->verbs_dev.rdi), (dd)->unit, (port), \ ##__VA_ARGS__) #define qib_devinfo(pcidev, fmt, ...) \ diff --git a/drivers/infiniband/hw/qib/qib_driver.c b/drivers/infiniband/hw/qib/qib_driver.c index 33d3335385e8..3117cc5f2a9a 100644 --- a/drivers/infiniband/hw/qib/qib_driver.c +++ b/drivers/infiniband/hw/qib/qib_driver.c @@ -81,22 +81,6 @@ MODULE_DESCRIPTION("Intel IB driver"); struct qlogic_ib_stats qib_stats; -const char *qib_get_unit_name(int unit) -{ - static char iname[16]; - - snprintf(iname, sizeof(iname), "infinipath%u", unit); - return iname; -} - -const char *qib_get_card_name(struct rvt_dev_info *rdi) -{ - struct qib_ibdev *ibdev = container_of(rdi, struct qib_ibdev, rdi); - struct qib_devdata *dd = container_of(ibdev, - struct qib_devdata, verbs_dev); - return qib_get_unit_name(dd->unit); -} - struct pci_dev *qib_get_pci_dev(struct rvt_dev_info *rdi) { struct qib_ibdev *ibdev = container_of(rdi, struct qib_ibdev, rdi); diff --git a/drivers/infiniband/hw/qib/qib_eeprom.c b/drivers/infiniband/hw/qib/qib_eeprom.c index 33a2e74c8495..5838b3bf34b9 100644 --- a/drivers/infiniband/hw/qib/qib_eeprom.c +++ b/drivers/infiniband/hw/qib/qib_eeprom.c @@ -163,8 +163,7 @@ void qib_get_eeprom_info(struct qib_devdata *dd) if (bguid[6] == 0xff) { if (bguid[5] == 0xff) { qib_dev_err(dd, - "Can't set %s GUID from base, wraps to OUI!\n", - qib_get_unit_name(t)); + "Can't set GUID from base, wraps to OUI!\n"); dd->base_guid = 0; goto bail; } diff --git a/drivers/infiniband/hw/qib/qib_file_ops.c b/drivers/infiniband/hw/qib/qib_file_ops.c index 2d6a191afec0..a9a48b393323 100644 --- a/drivers/infiniband/hw/qib/qib_file_ops.c +++ b/drivers/infiniband/hw/qib/qib_file_ops.c @@ -568,20 +568,16 @@ done: static int qib_set_part_key(struct qib_ctxtdata *rcd, u16 key) { struct qib_pportdata *ppd = rcd->ppd; - int i, any = 0, pidx = -1; + int i, pidx = -1; + bool any = false; u16 lkey = key & 0x7FFF; - int ret; - if (lkey == (QIB_DEFAULT_P_KEY & 0x7FFF)) { + if (lkey == (QIB_DEFAULT_P_KEY & 0x7FFF)) /* nothing to do; this key always valid */ - ret = 0; - goto bail; - } + return 0; - if (!lkey) { - ret = -EINVAL; - goto bail; - } + if (!lkey) + return -EINVAL; /* * Set the full membership bit, because it has to be @@ -594,18 +590,14 @@ static int qib_set_part_key(struct qib_ctxtdata *rcd, u16 key) for (i = 0; i < ARRAY_SIZE(rcd->pkeys); i++) { if (!rcd->pkeys[i] && pidx == -1) pidx = i; - if (rcd->pkeys[i] == key) { - ret = -EEXIST; - goto bail; - } + if (rcd->pkeys[i] == key) + return -EEXIST; } - if (pidx == -1) { - ret = -EBUSY; - goto bail; - } - for (any = i = 0; i < ARRAY_SIZE(ppd->pkeys); i++) { + if (pidx == -1) + return -EBUSY; + for (i = 0; i < ARRAY_SIZE(ppd->pkeys); i++) { if (!ppd->pkeys[i]) { - any++; + any = true; continue; } if (ppd->pkeys[i] == key) { @@ -613,44 +605,34 @@ static int qib_set_part_key(struct qib_ctxtdata *rcd, u16 key) if (atomic_inc_return(pkrefs) > 1) { rcd->pkeys[pidx] = key; - ret = 0; - goto bail; - } else { - /* - * lost race, decrement count, catch below - */ - atomic_dec(pkrefs); - any++; + return 0; } + /* + * lost race, decrement count, catch below + */ + atomic_dec(pkrefs); + any = true; } - if ((ppd->pkeys[i] & 0x7FFF) == lkey) { + if ((ppd->pkeys[i] & 0x7FFF) == lkey) /* * It makes no sense to have both the limited and * full membership PKEY set at the same time since * the unlimited one will disable the limited one. */ - ret = -EEXIST; - goto bail; - } - } - if (!any) { - ret = -EBUSY; - goto bail; + return -EEXIST; } - for (any = i = 0; i < ARRAY_SIZE(ppd->pkeys); i++) { + if (!any) + return -EBUSY; + for (i = 0; i < ARRAY_SIZE(ppd->pkeys); i++) { if (!ppd->pkeys[i] && atomic_inc_return(&ppd->pkeyrefs[i]) == 1) { rcd->pkeys[pidx] = key; ppd->pkeys[i] = key; (void) ppd->dd->f_set_ib_cfg(ppd, QIB_IB_CFG_PKEYS, 0); - ret = 0; - goto bail; + return 0; } } - ret = -EBUSY; - -bail: - return ret; + return -EBUSY; } /** diff --git a/drivers/infiniband/hw/qib/qib_init.c b/drivers/infiniband/hw/qib/qib_init.c index 85dfbba427f6..3990f386aa32 100644 --- a/drivers/infiniband/hw/qib/qib_init.c +++ b/drivers/infiniband/hw/qib/qib_init.c @@ -1119,6 +1119,8 @@ struct qib_devdata *qib_alloc_devdata(struct pci_dev *pdev, size_t extra) "Could not allocate unit ID: error %d\n", -ret); goto bail; } + rvt_set_ibdev_name(&dd->verbs_dev.rdi, "%s%d", "qib", dd->unit); + dd->int_counter = alloc_percpu(u64); if (!dd->int_counter) { ret = -ENOMEM; diff --git a/drivers/infiniband/hw/qib/qib_rc.c b/drivers/infiniband/hw/qib/qib_rc.c index 8f5754fb8579..e4a9ba1dd9ba 100644 --- a/drivers/infiniband/hw/qib/qib_rc.c +++ b/drivers/infiniband/hw/qib/qib_rc.c @@ -434,13 +434,13 @@ no_flow_control: qp->s_state = OP(COMPARE_SWAP); put_ib_ateth_swap(wqe->atomic_wr.swap, &ohdr->u.atomic_eth); - put_ib_ateth_swap(wqe->atomic_wr.compare_add, - &ohdr->u.atomic_eth); + put_ib_ateth_compare(wqe->atomic_wr.compare_add, + &ohdr->u.atomic_eth); } else { qp->s_state = OP(FETCH_ADD); put_ib_ateth_swap(wqe->atomic_wr.compare_add, &ohdr->u.atomic_eth); - put_ib_ateth_swap(0, &ohdr->u.atomic_eth); + put_ib_ateth_compare(0, &ohdr->u.atomic_eth); } put_ib_ateth_vaddr(wqe->atomic_wr.remote_addr, &ohdr->u.atomic_eth); diff --git a/drivers/infiniband/hw/qib/qib_verbs.c b/drivers/infiniband/hw/qib/qib_verbs.c index c55000501582..fabee760407e 100644 --- a/drivers/infiniband/hw/qib/qib_verbs.c +++ b/drivers/infiniband/hw/qib/qib_verbs.c @@ -1571,7 +1571,6 @@ int qib_register_ib_device(struct qib_devdata *dd) if (!ib_qib_sys_image_guid) ib_qib_sys_image_guid = ppd->guid; - strlcpy(ibdev->name, "qib%d", IB_DEVICE_NAME_MAX); ibdev->owner = THIS_MODULE; ibdev->node_guid = ppd->guid; ibdev->phys_port_cnt = dd->num_pports; @@ -1586,7 +1585,6 @@ int qib_register_ib_device(struct qib_devdata *dd) * Fill in rvt info object. */ dd->verbs_dev.rdi.driver_f.port_callback = qib_create_port_files; - dd->verbs_dev.rdi.driver_f.get_card_name = qib_get_card_name; dd->verbs_dev.rdi.driver_f.get_pci_dev = qib_get_pci_dev; dd->verbs_dev.rdi.driver_f.check_ah = qib_check_ah; dd->verbs_dev.rdi.driver_f.check_send_wqe = qib_check_send_wqe; diff --git a/drivers/infiniband/hw/usnic/usnic_ib_sysfs.c b/drivers/infiniband/hw/usnic/usnic_ib_sysfs.c index 685ef2293cb8..4210ca14014d 100644 --- a/drivers/infiniband/hw/usnic/usnic_ib_sysfs.c +++ b/drivers/infiniband/hw/usnic/usnic_ib_sysfs.c @@ -45,7 +45,6 @@ #include "usnic_ib_verbs.h" #include "usnic_ib_sysfs.h" #include "usnic_log.h" -#include "usnic_ib_sysfs.h" static ssize_t usnic_ib_show_board(struct device *device, struct device_attribute *attr, diff --git a/drivers/infiniband/hw/usnic/usnic_ib_verbs.c b/drivers/infiniband/hw/usnic/usnic_ib_verbs.c index aa2456a4f9bd..a688a5669168 100644 --- a/drivers/infiniband/hw/usnic/usnic_ib_verbs.c +++ b/drivers/infiniband/hw/usnic/usnic_ib_verbs.c @@ -47,7 +47,6 @@ #include "usnic_log.h" #include "usnic_uiom.h" #include "usnic_transport.h" -#include "usnic_ib_verbs.h" #define USNIC_DEFAULT_TRANSPORT USNIC_TRANSPORT_ROCE_CUSTOM diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma.h b/drivers/infiniband/hw/vmw_pvrdma/pvrdma.h index 4f7bd3b6a315..44cb1cfba417 100644 --- a/drivers/infiniband/hw/vmw_pvrdma/pvrdma.h +++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma.h @@ -93,7 +93,7 @@ struct pvrdma_cq { struct pvrdma_page_dir pdir; u32 cq_handle; bool is_kernel; - atomic_t refcnt; + refcount_t refcnt; struct completion free; }; @@ -196,7 +196,7 @@ struct pvrdma_qp { u8 state; bool is_kernel; struct mutex mutex; /* QP state mutex. */ - atomic_t refcnt; + refcount_t refcnt; struct completion free; }; diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_cq.c b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_cq.c index e529622cefad..faa9478c14a6 100644 --- a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_cq.c +++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_cq.c @@ -132,8 +132,9 @@ struct ib_cq *pvrdma_create_cq(struct ib_device *ibdev, } cq->ibcq.cqe = entries; + cq->is_kernel = !context; - if (context) { + if (!cq->is_kernel) { if (ib_copy_from_udata(&ucmd, udata, sizeof(ucmd))) { ret = -EFAULT; goto err_cq; @@ -148,8 +149,6 @@ struct ib_cq *pvrdma_create_cq(struct ib_device *ibdev, npages = ib_umem_page_count(cq->umem); } else { - cq->is_kernel = true; - /* One extra page for shared ring state */ npages = 1 + (entries * sizeof(struct pvrdma_cqe) + PAGE_SIZE - 1) / PAGE_SIZE; @@ -178,7 +177,7 @@ struct ib_cq *pvrdma_create_cq(struct ib_device *ibdev, else pvrdma_page_dir_insert_umem(&cq->pdir, cq->umem, 0); - atomic_set(&cq->refcnt, 1); + refcount_set(&cq->refcnt, 1); init_completion(&cq->free); spin_lock_init(&cq->cq_lock); @@ -202,7 +201,7 @@ struct ib_cq *pvrdma_create_cq(struct ib_device *ibdev, dev->cq_tbl[cq->cq_handle % dev->dsr->caps.max_cq] = cq; spin_unlock_irqrestore(&dev->cq_tbl_lock, flags); - if (context) { + if (!cq->is_kernel) { cq->uar = &(to_vucontext(context)->uar); /* Copy udata back. */ @@ -219,7 +218,7 @@ struct ib_cq *pvrdma_create_cq(struct ib_device *ibdev, err_page_dir: pvrdma_page_dir_cleanup(dev, &cq->pdir); err_umem: - if (context) + if (!cq->is_kernel) ib_umem_release(cq->umem); err_cq: atomic_dec(&dev->num_cqs); @@ -230,7 +229,7 @@ err_cq: static void pvrdma_free_cq(struct pvrdma_dev *dev, struct pvrdma_cq *cq) { - if (atomic_dec_and_test(&cq->refcnt)) + if (refcount_dec_and_test(&cq->refcnt)) complete(&cq->free); wait_for_completion(&cq->free); diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_main.c b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_main.c index e92681878c93..d650a9fcde24 100644 --- a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_main.c +++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_main.c @@ -243,13 +243,13 @@ static int pvrdma_register_device(struct pvrdma_dev *dev) mutex_init(&dev->port_mutex); spin_lock_init(&dev->desc_lock); - dev->cq_tbl = kcalloc(dev->dsr->caps.max_cq, sizeof(void *), + dev->cq_tbl = kcalloc(dev->dsr->caps.max_cq, sizeof(struct pvrdma_cq *), GFP_KERNEL); if (!dev->cq_tbl) return ret; spin_lock_init(&dev->cq_tbl_lock); - dev->qp_tbl = kcalloc(dev->dsr->caps.max_qp, sizeof(void *), + dev->qp_tbl = kcalloc(dev->dsr->caps.max_qp, sizeof(struct pvrdma_qp *), GFP_KERNEL); if (!dev->qp_tbl) goto err_cq_free; @@ -333,7 +333,7 @@ static void pvrdma_qp_event(struct pvrdma_dev *dev, u32 qpn, int type) spin_lock_irqsave(&dev->qp_tbl_lock, flags); qp = dev->qp_tbl[qpn % dev->dsr->caps.max_qp]; if (qp) - atomic_inc(&qp->refcnt); + refcount_inc(&qp->refcnt); spin_unlock_irqrestore(&dev->qp_tbl_lock, flags); if (qp && qp->ibqp.event_handler) { @@ -346,7 +346,7 @@ static void pvrdma_qp_event(struct pvrdma_dev *dev, u32 qpn, int type) ibqp->event_handler(&e, ibqp->qp_context); } if (qp) { - if (atomic_dec_and_test(&qp->refcnt)) + if (refcount_dec_and_test(&qp->refcnt)) complete(&qp->free); } } @@ -359,7 +359,7 @@ static void pvrdma_cq_event(struct pvrdma_dev *dev, u32 cqn, int type) spin_lock_irqsave(&dev->cq_tbl_lock, flags); cq = dev->cq_tbl[cqn % dev->dsr->caps.max_cq]; if (cq) - atomic_inc(&cq->refcnt); + refcount_inc(&cq->refcnt); spin_unlock_irqrestore(&dev->cq_tbl_lock, flags); if (cq && cq->ibcq.event_handler) { @@ -372,7 +372,7 @@ static void pvrdma_cq_event(struct pvrdma_dev *dev, u32 cqn, int type) ibcq->event_handler(&e, ibcq->cq_context); } if (cq) { - if (atomic_dec_and_test(&cq->refcnt)) + if (refcount_dec_and_test(&cq->refcnt)) complete(&cq->free); } } @@ -531,13 +531,13 @@ static irqreturn_t pvrdma_intrx_handler(int irq, void *dev_id) spin_lock_irqsave(&dev->cq_tbl_lock, flags); cq = dev->cq_tbl[cqne->info % dev->dsr->caps.max_cq]; if (cq) - atomic_inc(&cq->refcnt); + refcount_inc(&cq->refcnt); spin_unlock_irqrestore(&dev->cq_tbl_lock, flags); if (cq && cq->ibcq.comp_handler) cq->ibcq.comp_handler(&cq->ibcq, cq->ibcq.cq_context); if (cq) { - if (atomic_dec_and_test(&cq->refcnt)) + if (refcount_dec_and_test(&cq->refcnt)) complete(&cq->free); } pvrdma_idx_ring_inc(&ring->cons_head, ring_slots); @@ -882,8 +882,8 @@ static int pvrdma_pci_probe(struct pci_dev *pdev, dev_info(&pdev->dev, "device version %d, driver version %d\n", dev->dsr_version, PVRDMA_VERSION); - dev->dsr = dma_alloc_coherent(&pdev->dev, sizeof(*dev->dsr), - &dev->dsrbase, GFP_KERNEL); + dev->dsr = dma_zalloc_coherent(&pdev->dev, sizeof(*dev->dsr), + &dev->dsrbase, GFP_KERNEL); if (!dev->dsr) { dev_err(&pdev->dev, "failed to allocate shared region\n"); ret = -ENOMEM; @@ -891,7 +891,6 @@ static int pvrdma_pci_probe(struct pci_dev *pdev, } /* Setup the shared region */ - memset(dev->dsr, 0, sizeof(*dev->dsr)); dev->dsr->driver_version = PVRDMA_VERSION; dev->dsr->gos_info.gos_bits = sizeof(void *) == 4 ? PVRDMA_GOS_BITS_32 : diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_mr.c b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_mr.c index 8519f3212e52..fa96fa4fb829 100644 --- a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_mr.c +++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_mr.c @@ -119,10 +119,7 @@ struct ib_mr *pvrdma_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, union pvrdma_cmd_resp rsp; struct pvrdma_cmd_create_mr *cmd = &req.create_mr; struct pvrdma_cmd_create_mr_resp *resp = &rsp.create_mr_resp; - int nchunks; int ret; - int entry; - struct scatterlist *sg; if (length == 0 || length > dev->dsr->caps.max_mr_size) { dev_warn(&dev->pdev->dev, "invalid mem region length\n"); @@ -137,13 +134,9 @@ struct ib_mr *pvrdma_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, return ERR_CAST(umem); } - nchunks = 0; - for_each_sg(umem->sg_head.sgl, sg, umem->nmap, entry) - nchunks += sg_dma_len(sg) >> PAGE_SHIFT; - - if (nchunks < 0 || nchunks > PVRDMA_PAGE_DIR_MAX_PAGES) { + if (umem->npages < 0 || umem->npages > PVRDMA_PAGE_DIR_MAX_PAGES) { dev_warn(&dev->pdev->dev, "overflow %d pages in mem region\n", - nchunks); + umem->npages); ret = -EINVAL; goto err_umem; } @@ -158,7 +151,7 @@ struct ib_mr *pvrdma_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, mr->mmr.size = length; mr->umem = umem; - ret = pvrdma_page_dir_init(dev, &mr->pdir, nchunks, false); + ret = pvrdma_page_dir_init(dev, &mr->pdir, umem->npages, false); if (ret) { dev_warn(&dev->pdev->dev, "could not allocate page directory\n"); @@ -175,7 +168,7 @@ struct ib_mr *pvrdma_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, cmd->length = length; cmd->pd_handle = to_vpd(pd)->pd_handle; cmd->access_flags = access_flags; - cmd->nchunks = nchunks; + cmd->nchunks = umem->npages; cmd->pdir_dma = mr->pdir.dir_dma; ret = pvrdma_cmd_post(dev, &req, &rsp, PVRDMA_CMD_CREATE_MR_RESP); diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_qp.c b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_qp.c index 4059308e1454..7bf518bdbf21 100644 --- a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_qp.c +++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_qp.c @@ -245,12 +245,13 @@ struct ib_qp *pvrdma_create_qp(struct ib_pd *pd, spin_lock_init(&qp->sq.lock); spin_lock_init(&qp->rq.lock); mutex_init(&qp->mutex); - atomic_set(&qp->refcnt, 1); + refcount_set(&qp->refcnt, 1); init_completion(&qp->free); qp->state = IB_QPS_RESET; + qp->is_kernel = !(pd->uobject && udata); - if (pd->uobject && udata) { + if (!qp->is_kernel) { dev_dbg(&dev->pdev->dev, "create queuepair from user space\n"); @@ -291,8 +292,6 @@ struct ib_qp *pvrdma_create_qp(struct ib_pd *pd, qp->npages_recv = 0; qp->npages = qp->npages_send + qp->npages_recv; } else { - qp->is_kernel = true; - ret = pvrdma_set_sq_size(to_vdev(pd->device), &init_attr->cap, qp); if (ret) @@ -394,7 +393,7 @@ struct ib_qp *pvrdma_create_qp(struct ib_pd *pd, err_pdir: pvrdma_page_dir_cleanup(dev, &qp->pdir); err_umem: - if (pd->uobject && udata) { + if (!qp->is_kernel) { if (qp->rumem) ib_umem_release(qp->rumem); if (qp->sumem) @@ -428,7 +427,7 @@ static void pvrdma_free_qp(struct pvrdma_qp *qp) pvrdma_unlock_cqs(scq, rcq, &scq_flags, &rcq_flags); - if (atomic_dec_and_test(&qp->refcnt)) + if (refcount_dec_and_test(&qp->refcnt)) complete(&qp->free); wait_for_completion(&qp->free); diff --git a/drivers/infiniband/sw/rdmavt/cq.c b/drivers/infiniband/sw/rdmavt/cq.c index 97d71e49c092..88fa4d44ab5f 100644 --- a/drivers/infiniband/sw/rdmavt/cq.c +++ b/drivers/infiniband/sw/rdmavt/cq.c @@ -198,7 +198,7 @@ struct ib_cq *rvt_create_cq(struct ib_device *ibdev, return ERR_PTR(-EINVAL); /* Allocate the completion queue structure. */ - cq = kzalloc(sizeof(*cq), GFP_KERNEL); + cq = kzalloc_node(sizeof(*cq), GFP_KERNEL, rdi->dparms.node); if (!cq) return ERR_PTR(-ENOMEM); @@ -214,7 +214,9 @@ struct ib_cq *rvt_create_cq(struct ib_device *ibdev, sz += sizeof(struct ib_uverbs_wc) * (entries + 1); else sz += sizeof(struct ib_wc) * (entries + 1); - wc = vmalloc_user(sz); + wc = udata ? + vmalloc_user(sz) : + vzalloc_node(sz, rdi->dparms.node); if (!wc) { ret = ERR_PTR(-ENOMEM); goto bail_cq; @@ -369,7 +371,9 @@ int rvt_resize_cq(struct ib_cq *ibcq, int cqe, struct ib_udata *udata) sz += sizeof(struct ib_uverbs_wc) * (cqe + 1); else sz += sizeof(struct ib_wc) * (cqe + 1); - wc = vmalloc_user(sz); + wc = udata ? + vmalloc_user(sz) : + vzalloc_node(sz, rdi->dparms.node); if (!wc) return -ENOMEM; diff --git a/drivers/infiniband/sw/rdmavt/qp.c b/drivers/infiniband/sw/rdmavt/qp.c index 9177df60742a..0bbf20597056 100644 --- a/drivers/infiniband/sw/rdmavt/qp.c +++ b/drivers/infiniband/sw/rdmavt/qp.c @@ -2075,6 +2075,7 @@ void rvt_add_rnr_timer(struct rvt_qp *qp, u32 aeth) lockdep_assert_held(&qp->s_lock); qp->s_flags |= RVT_S_WAIT_RNR; to = rvt_aeth_to_usec(aeth); + trace_rvt_rnrnak_add(qp, to); hrtimer_start(&qp->s_rnr_timer, ns_to_ktime(1000 * to), HRTIMER_MODE_REL); } @@ -2104,17 +2105,14 @@ EXPORT_SYMBOL(rvt_stop_rc_timers); * stop an rnr timer and return if the timer * had been pending. */ -static int rvt_stop_rnr_timer(struct rvt_qp *qp) +static void rvt_stop_rnr_timer(struct rvt_qp *qp) { - int rval = 0; - lockdep_assert_held(&qp->s_lock); /* Remove QP from rnr timer */ if (qp->s_flags & RVT_S_WAIT_RNR) { qp->s_flags &= ~RVT_S_WAIT_RNR; - rval = hrtimer_try_to_cancel(&qp->s_rnr_timer); + trace_rvt_rnrnak_stop(qp, 0); } - return rval; } /** @@ -2167,6 +2165,7 @@ enum hrtimer_restart rvt_rc_rnr_retry(struct hrtimer *t) spin_lock_irqsave(&qp->s_lock, flags); rvt_stop_rnr_timer(qp); + trace_rvt_rnrnak_timeout(qp, 0); rdi->driver_f.schedule_send(qp); spin_unlock_irqrestore(&qp->s_lock, flags); return HRTIMER_NORESTART; diff --git a/drivers/infiniband/sw/rdmavt/srq.c b/drivers/infiniband/sw/rdmavt/srq.c index f7c48e9023de..3707952b4364 100644 --- a/drivers/infiniband/sw/rdmavt/srq.c +++ b/drivers/infiniband/sw/rdmavt/srq.c @@ -90,7 +90,7 @@ struct ib_srq *rvt_create_srq(struct ib_pd *ibpd, srq_init_attr->attr.max_wr > dev->dparms.props.max_srq_wr) return ERR_PTR(-EINVAL); - srq = kmalloc(sizeof(*srq), GFP_KERNEL); + srq = kzalloc_node(sizeof(*srq), GFP_KERNEL, dev->dparms.node); if (!srq) return ERR_PTR(-ENOMEM); @@ -101,7 +101,10 @@ struct ib_srq *rvt_create_srq(struct ib_pd *ibpd, srq->rq.max_sge = srq_init_attr->attr.max_sge; sz = sizeof(struct ib_sge) * srq->rq.max_sge + sizeof(struct rvt_rwqe); - srq->rq.wq = vmalloc_user(sizeof(struct rvt_rwq) + srq->rq.size * sz); + srq->rq.wq = udata ? + vmalloc_user(sizeof(struct rvt_rwq) + srq->rq.size * sz) : + vzalloc_node(sizeof(struct rvt_rwq) + srq->rq.size * sz, + dev->dparms.node); if (!srq->rq.wq) { ret = ERR_PTR(-ENOMEM); goto bail_srq; @@ -129,16 +132,12 @@ struct ib_srq *rvt_create_srq(struct ib_pd *ibpd, ret = ERR_PTR(err); goto bail_ip; } - } else { - srq->ip = NULL; } /* * ib_create_srq() will initialize srq->ibsrq. */ spin_lock_init(&srq->rq.lock); - srq->rq.wq->head = 0; - srq->rq.wq->tail = 0; srq->limit = srq_init_attr->attr.srq_limit; spin_lock(&dev->n_srqs_lock); @@ -200,7 +199,10 @@ int rvt_modify_srq(struct ib_srq *ibsrq, struct ib_srq_attr *attr, sz = sizeof(struct rvt_rwqe) + srq->rq.max_sge * sizeof(struct ib_sge); size = attr->max_wr + 1; - wq = vmalloc_user(sizeof(struct rvt_rwq) + size * sz); + wq = udata ? + vmalloc_user(sizeof(struct rvt_rwq) + size * sz) : + vzalloc_node(sizeof(struct rvt_rwq) + size * sz, + dev->dparms.node); if (!wq) return -ENOMEM; diff --git a/drivers/infiniband/sw/rdmavt/trace.h b/drivers/infiniband/sw/rdmavt/trace.h index bb4b1e710f22..36ddbd291ee0 100644 --- a/drivers/infiniband/sw/rdmavt/trace.h +++ b/drivers/infiniband/sw/rdmavt/trace.h @@ -45,8 +45,8 @@ * */ -#define RDI_DEV_ENTRY(rdi) __string(dev, rdi->driver_f.get_card_name(rdi)) -#define RDI_DEV_ASSIGN(rdi) __assign_str(dev, rdi->driver_f.get_card_name(rdi)) +#define RDI_DEV_ENTRY(rdi) __string(dev, rvt_get_ibdev_name(rdi)) +#define RDI_DEV_ASSIGN(rdi) __assign_str(dev, rvt_get_ibdev_name(rdi)) #include "trace_rvt.h" #include "trace_qp.h" diff --git a/drivers/infiniband/sw/rdmavt/trace_qp.h b/drivers/infiniband/sw/rdmavt/trace_qp.h index 4c77a3119bda..efc9d814b032 100644 --- a/drivers/infiniband/sw/rdmavt/trace_qp.h +++ b/drivers/infiniband/sw/rdmavt/trace_qp.h @@ -85,6 +85,48 @@ DEFINE_EVENT(rvt_qphash_template, rvt_qpremove, TP_PROTO(struct rvt_qp *qp, u32 bucket), TP_ARGS(qp, bucket)); +DECLARE_EVENT_CLASS( + rvt_rnrnak_template, + TP_PROTO(struct rvt_qp *qp, u32 to), + TP_ARGS(qp, to), + TP_STRUCT__entry( + RDI_DEV_ENTRY(ib_to_rvt(qp->ibqp.device)) + __field(u32, qpn) + __field(void *, hrtimer) + __field(u32, s_flags) + __field(u32, to) + ), + TP_fast_assign( + RDI_DEV_ASSIGN(ib_to_rvt(qp->ibqp.device)) + __entry->qpn = qp->ibqp.qp_num; + __entry->hrtimer = &qp->s_rnr_timer; + __entry->s_flags = qp->s_flags; + __entry->to = to; + ), + TP_printk( + "[%s] qpn 0x%x hrtimer 0x%p s_flags 0x%x timeout %u us", + __get_str(dev), + __entry->qpn, + __entry->hrtimer, + __entry->s_flags, + __entry->to + ) +); + +DEFINE_EVENT( + rvt_rnrnak_template, rvt_rnrnak_add, + TP_PROTO(struct rvt_qp *qp, u32 to), + TP_ARGS(qp, to)); + +DEFINE_EVENT( + rvt_rnrnak_template, rvt_rnrnak_timeout, + TP_PROTO(struct rvt_qp *qp, u32 to), + TP_ARGS(qp, to)); + +DEFINE_EVENT( + rvt_rnrnak_template, rvt_rnrnak_stop, + TP_PROTO(struct rvt_qp *qp, u32 to), + TP_ARGS(qp, to)); #endif /* __RVT_TRACE_QP_H */ diff --git a/drivers/infiniband/sw/rdmavt/vt.c b/drivers/infiniband/sw/rdmavt/vt.c index 64bdd442078a..088fb2d6d919 100644 --- a/drivers/infiniband/sw/rdmavt/vt.c +++ b/drivers/infiniband/sw/rdmavt/vt.c @@ -413,7 +413,6 @@ static noinline int check_support(struct rvt_dev_info *rdi, int verb) * required for rdmavt to function. */ if ((!rdi->driver_f.port_callback) || - (!rdi->driver_f.get_card_name) || (!rdi->driver_f.get_pci_dev)) return -EINVAL; break; diff --git a/drivers/infiniband/sw/rdmavt/vt.h b/drivers/infiniband/sw/rdmavt/vt.h index f363505312be..8823b2e7aac6 100644 --- a/drivers/infiniband/sw/rdmavt/vt.h +++ b/drivers/infiniband/sw/rdmavt/vt.h @@ -63,19 +63,19 @@ #define rvt_pr_info(rdi, fmt, ...) \ __rvt_pr_info(rdi->driver_f.get_pci_dev(rdi), \ - rdi->driver_f.get_card_name(rdi), \ + rvt_get_ibdev_name(rdi), \ fmt, \ ##__VA_ARGS__) #define rvt_pr_warn(rdi, fmt, ...) \ __rvt_pr_warn(rdi->driver_f.get_pci_dev(rdi), \ - rdi->driver_f.get_card_name(rdi), \ + rvt_get_ibdev_name(rdi), \ fmt, \ ##__VA_ARGS__) #define rvt_pr_err(rdi, fmt, ...) \ __rvt_pr_err(rdi->driver_f.get_pci_dev(rdi), \ - rdi->driver_f.get_card_name(rdi), \ + rvt_get_ibdev_name(rdi), \ fmt, \ ##__VA_ARGS__) diff --git a/drivers/infiniband/sw/rxe/rxe.c b/drivers/infiniband/sw/rxe/rxe.c index 8c3d30b3092d..b7debb6f2eac 100644 --- a/drivers/infiniband/sw/rxe/rxe.c +++ b/drivers/infiniband/sw/rxe/rxe.c @@ -77,12 +77,6 @@ void rxe_release(struct kref *kref) ib_dealloc_device(&rxe->ib_dev); } -void rxe_dev_put(struct rxe_dev *rxe) -{ - kref_put(&rxe->ref_cnt, rxe_release); -} -EXPORT_SYMBOL_GPL(rxe_dev_put); - /* initialize rxe device parameters */ static int rxe_init_device_param(struct rxe_dev *rxe) { diff --git a/drivers/infiniband/sw/rxe/rxe.h b/drivers/infiniband/sw/rxe/rxe.h index 6447d736d5a4..7d232611303f 100644 --- a/drivers/infiniband/sw/rxe/rxe.h +++ b/drivers/infiniband/sw/rxe/rxe.h @@ -57,6 +57,7 @@ #include "rxe_hdr.h" #include "rxe_param.h" #include "rxe_verbs.h" +#include "rxe_loc.h" #define RXE_UVERBS_ABI_VERSION (1) @@ -95,7 +96,10 @@ void rxe_remove_all(void); int rxe_rcv(struct sk_buff *skb); -void rxe_dev_put(struct rxe_dev *rxe); +static inline void rxe_dev_put(struct rxe_dev *rxe) +{ + kref_put(&rxe->ref_cnt, rxe_release); +} struct rxe_dev *net_to_rxe(struct net_device *ndev); struct rxe_dev *get_rxe_by_name(const char *name); diff --git a/drivers/infiniband/sw/rxe/rxe_recv.c b/drivers/infiniband/sw/rxe/rxe_recv.c index fb8c83e055e1..4c3f899241d4 100644 --- a/drivers/infiniband/sw/rxe/rxe_recv.c +++ b/drivers/infiniband/sw/rxe/rxe_recv.c @@ -336,7 +336,6 @@ static int rxe_match_dgid(struct rxe_dev *rxe, struct sk_buff *skb) { union ib_gid dgid; union ib_gid *pdgid; - u16 index; if (skb->protocol == htons(ETH_P_IP)) { ipv6_addr_set_v4mapped(ip_hdr(skb)->daddr, @@ -348,7 +347,7 @@ static int rxe_match_dgid(struct rxe_dev *rxe, struct sk_buff *skb) return ib_find_cached_gid_by_port(&rxe->ib_dev, pdgid, IB_GID_TYPE_ROCE_UDP_ENCAP, - 1, rxe->ndev, &index); + 1, rxe->ndev, NULL); } /* rxe_rcv is called from the interface driver */ diff --git a/drivers/infiniband/ulp/ipoib/ipoib_cm.c b/drivers/infiniband/ulp/ipoib/ipoib_cm.c index 2c13123bfd69..dfbb8fdda5f6 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_cm.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_cm.c @@ -766,12 +766,14 @@ void ipoib_cm_send(struct net_device *dev, struct sk_buff *skb, struct ipoib_cm_ skb_orphan(skb); skb_dst_drop(skb); - if (netif_queue_stopped(dev)) - if (ib_req_notify_cq(priv->send_cq, IB_CQ_NEXT_COMP | - IB_CQ_REPORT_MISSED_EVENTS)) { + if (netif_queue_stopped(dev)) { + rc = ib_req_notify_cq(priv->send_cq, IB_CQ_NEXT_COMP | + IB_CQ_REPORT_MISSED_EVENTS); + if (unlikely(rc < 0)) ipoib_warn(priv, "IPoIB/CM:request notify on send CQ failed\n"); + else if (rc) napi_schedule(&priv->send_napi); - } + } rc = post_send(priv, tx, tx->tx_head & (ipoib_sendq_size - 1), tx_req); if (unlikely(rc)) { @@ -876,7 +878,7 @@ int ipoib_cm_dev_open(struct net_device *dev) priv->cm.id = ib_create_cm_id(priv->ca, ipoib_cm_rx_handler, dev); if (IS_ERR(priv->cm.id)) { - printk(KERN_WARNING "%s: failed to create CM ID\n", priv->ca->name); + pr_warn("%s: failed to create CM ID\n", priv->ca->name); ret = PTR_ERR(priv->cm.id); goto err_cm; } @@ -884,8 +886,8 @@ int ipoib_cm_dev_open(struct net_device *dev) ret = ib_cm_listen(priv->cm.id, cpu_to_be64(IPOIB_CM_IETF_ID | priv->qp->qp_num), 0); if (ret) { - printk(KERN_WARNING "%s: failed to listen on ID 0x%llx\n", priv->ca->name, - IPOIB_CM_IETF_ID | priv->qp->qp_num); + pr_warn("%s: failed to listen on ID 0x%llx\n", priv->ca->name, + IPOIB_CM_IETF_ID | priv->qp->qp_num); goto err_listen; } @@ -1563,7 +1565,7 @@ static void ipoib_cm_create_srq(struct net_device *dev, int max_sge) priv->cm.srq = ib_create_srq(priv->pd, &srq_init_attr); if (IS_ERR(priv->cm.srq)) { if (PTR_ERR(priv->cm.srq) != -ENOSYS) - printk(KERN_WARNING "%s: failed to allocate SRQ, error %ld\n", + pr_warn("%s: failed to allocate SRQ, error %ld\n", priv->ca->name, PTR_ERR(priv->cm.srq)); priv->cm.srq = NULL; return; diff --git a/drivers/infiniband/ulp/ipoib/ipoib_ib.c b/drivers/infiniband/ulp/ipoib/ipoib_ib.c index e6151a29c412..10384ea50bed 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_ib.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_ib.c @@ -644,7 +644,7 @@ int ipoib_send(struct net_device *dev, struct sk_buff *skb, if (netif_queue_stopped(dev)) if (ib_req_notify_cq(priv->send_cq, IB_CQ_NEXT_COMP | - IB_CQ_REPORT_MISSED_EVENTS)) + IB_CQ_REPORT_MISSED_EVENTS) < 0) ipoib_warn(priv, "request notify on send CQ failed\n"); rc = post_send(priv, priv->tx_head & (ipoib_sendq_size - 1), @@ -1085,8 +1085,7 @@ static bool ipoib_dev_addr_changed_valid(struct ipoib_dev_priv *priv) netif_addr_unlock_bh(priv->dev); - err = ib_find_gid(priv->ca, &search_gid, IB_GID_TYPE_IB, - priv->dev, &port, &index); + err = ib_find_gid(priv->ca, &search_gid, priv->dev, &port, &index); netif_addr_lock_bh(priv->dev); diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c index 8880351df179..5930c7d9a8fb 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_main.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c @@ -768,13 +768,30 @@ static void path_rec_completion(int status, if (!status) { struct rdma_ah_attr av; - if (!ib_init_ah_from_path(priv->ca, priv->port, pathrec, &av)) + if (!ib_init_ah_attr_from_path(priv->ca, priv->port, + pathrec, &av)) ah = ipoib_create_ah(dev, priv->pd, &av); } spin_lock_irqsave(&priv->lock, flags); if (!IS_ERR_OR_NULL(ah)) { + /* + * pathrec.dgid is used as the database key from the LLADDR, + * it must remain unchanged even if the SA returns a different + * GID to use in the AH. + */ + if (memcmp(pathrec->dgid.raw, path->pathrec.dgid.raw, + sizeof(union ib_gid))) { + ipoib_dbg( + priv, + "%s got PathRec for gid %pI6 while asked for %pI6\n", + dev->name, pathrec->dgid.raw, + path->pathrec.dgid.raw); + memcpy(pathrec->dgid.raw, path->pathrec.dgid.raw, + sizeof(union ib_gid)); + } + path->pathrec = *pathrec; old_ah = path->ah; @@ -840,6 +857,23 @@ static void path_rec_completion(int status, } } +static void init_path_rec(struct ipoib_dev_priv *priv, struct ipoib_path *path, + void *gid) +{ + path->dev = priv->dev; + + if (rdma_cap_opa_ah(priv->ca, priv->port)) + path->pathrec.rec_type = SA_PATH_REC_TYPE_OPA; + else + path->pathrec.rec_type = SA_PATH_REC_TYPE_IB; + + memcpy(path->pathrec.dgid.raw, gid, sizeof(union ib_gid)); + path->pathrec.sgid = priv->local_gid; + path->pathrec.pkey = cpu_to_be16(priv->pkey); + path->pathrec.numb_path = 1; + path->pathrec.traffic_class = priv->broadcast->mcmember.traffic_class; +} + static struct ipoib_path *path_rec_create(struct net_device *dev, void *gid) { struct ipoib_dev_priv *priv = ipoib_priv(dev); @@ -852,21 +886,11 @@ static struct ipoib_path *path_rec_create(struct net_device *dev, void *gid) if (!path) return NULL; - path->dev = dev; - skb_queue_head_init(&path->queue); INIT_LIST_HEAD(&path->neigh_list); - if (rdma_cap_opa_ah(priv->ca, priv->port)) - path->pathrec.rec_type = SA_PATH_REC_TYPE_OPA; - else - path->pathrec.rec_type = SA_PATH_REC_TYPE_IB; - memcpy(path->pathrec.dgid.raw, gid, sizeof (union ib_gid)); - path->pathrec.sgid = priv->local_gid; - path->pathrec.pkey = cpu_to_be16(priv->pkey); - path->pathrec.numb_path = 1; - path->pathrec.traffic_class = priv->broadcast->mcmember.traffic_class; + init_path_rec(priv, path, gid); return path; } @@ -1005,6 +1029,10 @@ static void unicast_arp_send(struct sk_buff *skb, struct net_device *dev, spin_lock_irqsave(&priv->lock, flags); + /* no broadcast means that all paths are (going to be) not valid */ + if (!priv->broadcast) + goto drop_and_unlock; + path = __path_find(dev, phdr->hwaddr + 4); if (!path || !path->valid) { int new_path = 0; @@ -1014,6 +1042,10 @@ static void unicast_arp_send(struct sk_buff *skb, struct net_device *dev, new_path = 1; } if (path) { + if (!new_path) + /* make sure there is no changes in the existing path record */ + init_path_rec(priv, path, phdr->hwaddr + 4); + if (skb_queue_len(&path->queue) < IPOIB_MAX_PATH_REC_QUEUE) { push_pseudo_header(skb, phdr->hwaddr); __skb_queue_tail(&path->queue, skb); @@ -1030,8 +1062,7 @@ static void unicast_arp_send(struct sk_buff *skb, struct net_device *dev, } else __path_add(dev, path); } else { - ++dev->stats.tx_dropped; - dev_kfree_skb_any(skb); + goto drop_and_unlock; } spin_unlock_irqrestore(&priv->lock, flags); @@ -1051,11 +1082,16 @@ static void unicast_arp_send(struct sk_buff *skb, struct net_device *dev, push_pseudo_header(skb, phdr->hwaddr); __skb_queue_tail(&path->queue, skb); } else { - ++dev->stats.tx_dropped; - dev_kfree_skb_any(skb); + goto drop_and_unlock; } spin_unlock_irqrestore(&priv->lock, flags); + return; + +drop_and_unlock: + ++dev->stats.tx_dropped; + dev_kfree_skb_any(skb); + spin_unlock_irqrestore(&priv->lock, flags); } static int ipoib_start_xmit(struct sk_buff *skb, struct net_device *dev) @@ -1674,8 +1710,8 @@ static int ipoib_dev_init_default(struct net_device *dev) priv->tx_ring = vzalloc(ipoib_sendq_size * sizeof *priv->tx_ring); if (!priv->tx_ring) { - printk(KERN_WARNING "%s: failed to allocate TX ring (%d entries)\n", - priv->ca->name, ipoib_sendq_size); + pr_warn("%s: failed to allocate TX ring (%d entries)\n", + priv->ca->name, ipoib_sendq_size); goto out_rx_ring_cleanup; } @@ -2207,16 +2243,17 @@ static struct net_device *ipoib_add_port(const char *format, int result = -ENOMEM; priv = ipoib_intf_alloc(hca, port, format); - if (!priv) + if (!priv) { + pr_warn("%s, %d: ipoib_intf_alloc failed\n", hca->name, port); goto alloc_mem_failed; + } SET_NETDEV_DEV(priv->dev, hca->dev.parent); priv->dev->dev_id = port - 1; result = ib_query_port(hca, port, &attr); if (result) { - printk(KERN_WARNING "%s: ib_query_port %d failed\n", - hca->name, port); + pr_warn("%s: ib_query_port %d failed\n", hca->name, port); goto device_init_failed; } @@ -2231,8 +2268,8 @@ static struct net_device *ipoib_add_port(const char *format, result = ib_query_pkey(hca, port, 0, &priv->pkey); if (result) { - printk(KERN_WARNING "%s: ib_query_pkey port %d failed (ret = %d)\n", - hca->name, port, result); + pr_warn("%s: ib_query_pkey port %d failed (ret = %d)\n", + hca->name, port, result); goto device_init_failed; } @@ -2249,8 +2286,8 @@ static struct net_device *ipoib_add_port(const char *format, result = ib_query_gid(hca, port, 0, &priv->local_gid, NULL); if (result) { - printk(KERN_WARNING "%s: ib_query_gid port %d failed (ret = %d)\n", - hca->name, port, result); + pr_warn("%s: ib_query_gid port %d failed (ret = %d)\n", + hca->name, port, result); goto device_init_failed; } @@ -2260,8 +2297,8 @@ static struct net_device *ipoib_add_port(const char *format, result = ipoib_dev_init(priv->dev, hca, port); if (result) { - printk(KERN_WARNING "%s: failed to initialize port %d (ret = %d)\n", - hca->name, port, result); + pr_warn("%s: failed to initialize port %d (ret = %d)\n", + hca->name, port, result); goto device_init_failed; } @@ -2271,8 +2308,8 @@ static struct net_device *ipoib_add_port(const char *format, result = register_netdev(priv->dev); if (result) { - printk(KERN_WARNING "%s: couldn't register ipoib port %d; error %d\n", - hca->name, port, result); + pr_warn("%s: couldn't register ipoib port %d; error %d\n", + hca->name, port, result); goto register_failed; } @@ -2337,8 +2374,7 @@ static void ipoib_add_one(struct ib_device *device) } if (!count) { - pr_err("Failed to init port, removing it\n"); - ipoib_remove_one(device, dev_list); + kfree(dev_list); return; } diff --git a/drivers/infiniband/ulp/ipoib/ipoib_verbs.c b/drivers/infiniband/ulp/ipoib/ipoib_verbs.c index a1ed25422b72..984a88096f39 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_verbs.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_verbs.c @@ -178,7 +178,7 @@ int ipoib_transport_dev_init(struct net_device *dev, struct ib_device *ca) priv->recv_cq = ib_create_cq(priv->ca, ipoib_ib_rx_completion, NULL, priv, &cq_attr); if (IS_ERR(priv->recv_cq)) { - printk(KERN_WARNING "%s: failed to create receive CQ\n", ca->name); + pr_warn("%s: failed to create receive CQ\n", ca->name); goto out_cm_dev_cleanup; } @@ -187,7 +187,7 @@ int ipoib_transport_dev_init(struct net_device *dev, struct ib_device *ca) priv->send_cq = ib_create_cq(priv->ca, ipoib_ib_tx_completion, NULL, priv, &cq_attr); if (IS_ERR(priv->send_cq)) { - printk(KERN_WARNING "%s: failed to create send CQ\n", ca->name); + pr_warn("%s: failed to create send CQ\n", ca->name); goto out_free_recv_cq; } @@ -208,7 +208,7 @@ int ipoib_transport_dev_init(struct net_device *dev, struct ib_device *ca) priv->qp = ib_create_qp(priv->pd, &init_attr); if (IS_ERR(priv->qp)) { - printk(KERN_WARNING "%s: failed to create QP\n", ca->name); + pr_warn("%s: failed to create QP\n", ca->name); goto out_free_send_cq; } diff --git a/drivers/infiniband/ulp/isert/ib_isert.c b/drivers/infiniband/ulp/isert/ib_isert.c index 720dfb3a1ac2..fd55163801a3 100644 --- a/drivers/infiniband/ulp/isert/ib_isert.c +++ b/drivers/infiniband/ulp/isert/ib_isert.c @@ -2123,6 +2123,9 @@ isert_rdma_rw_ctx_post(struct isert_cmd *cmd, struct isert_conn *conn, u32 rkey, offset; int ret; + if (cmd->ctx_init_done) + goto rdma_ctx_post; + if (dir == DMA_FROM_DEVICE) { addr = cmd->write_va; rkey = cmd->write_stag; @@ -2150,11 +2153,15 @@ isert_rdma_rw_ctx_post(struct isert_cmd *cmd, struct isert_conn *conn, se_cmd->t_data_sg, se_cmd->t_data_nents, offset, addr, rkey, dir); } + if (ret < 0) { isert_err("Cmd: %p failed to prepare RDMA res\n", cmd); return ret; } + cmd->ctx_init_done = true; + +rdma_ctx_post: ret = rdma_rw_ctx_post(&cmd->rw, conn->qp, port_num, cqe, chain_wr); if (ret < 0) isert_err("Cmd: %p failed to post RDMA res\n", cmd); diff --git a/drivers/infiniband/ulp/isert/ib_isert.h b/drivers/infiniband/ulp/isert/ib_isert.h index d6fd248320ae..3b296bac4f60 100644 --- a/drivers/infiniband/ulp/isert/ib_isert.h +++ b/drivers/infiniband/ulp/isert/ib_isert.h @@ -126,6 +126,7 @@ struct isert_cmd { struct rdma_rw_ctx rw; struct work_struct comp_work; struct scatterlist sg; + bool ctx_init_done; }; static inline struct isert_cmd *tx_desc_to_cmd(struct iser_tx_desc *desc) diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c index 972d4b3c5223..62d88212c1b0 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.c +++ b/drivers/infiniband/ulp/srp/ib_srp.c @@ -3110,7 +3110,6 @@ static int srp_parse_options(const char *buf, struct srp_target_port *target) { char *options, *sep_opt; char *p; - char dgid[3]; substring_t args[MAX_OPT_ARGS]; int opt_mask = 0; int token; @@ -3162,16 +3161,10 @@ static int srp_parse_options(const char *buf, struct srp_target_port *target) goto out; } - for (i = 0; i < 16; ++i) { - strlcpy(dgid, p + i * 2, sizeof(dgid)); - if (sscanf(dgid, "%hhx", - &target->orig_dgid.raw[i]) < 1) { - ret = -EINVAL; - kfree(p); - goto out; - } - } + ret = hex2bin(target->orig_dgid.raw, p, 16); kfree(p); + if (ret < 0) + goto out; break; case SRP_OPT_PKEY: |