diff options
Diffstat (limited to 'drivers/infiniband')
46 files changed, 501 insertions, 444 deletions
diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index 19f1730a4f24..a68d0ccf67a4 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -4724,10 +4724,14 @@ static int __init cma_init(void) if (ret) goto err; - cma_configfs_init(); + ret = cma_configfs_init(); + if (ret) + goto err_ib; return 0; +err_ib: + ib_unregister_client(&cma_client); err: unregister_netdevice_notifier(&cma_nb); ib_sa_unregister_client(&sa_client); diff --git a/drivers/infiniband/core/core_priv.h b/drivers/infiniband/core/core_priv.h index 888d89ce81df..beee7b7e0d9a 100644 --- a/drivers/infiniband/core/core_priv.h +++ b/drivers/infiniband/core/core_priv.h @@ -302,7 +302,9 @@ static inline struct ib_qp *_ib_create_qp(struct ib_device *dev, struct ib_udata *udata, struct ib_uobject *uobj) { + enum ib_qp_type qp_type = attr->qp_type; struct ib_qp *qp; + bool is_xrc; if (!dev->ops.create_qp) return ERR_PTR(-EOPNOTSUPP); @@ -320,7 +322,8 @@ static inline struct ib_qp *_ib_create_qp(struct ib_device *dev, * and more importantly they are created internaly by driver, * see mlx5 create_dev_resources() as an example. */ - if (attr->qp_type < IB_QPT_XRC_INI) { + is_xrc = qp_type == IB_QPT_XRC_INI || qp_type == IB_QPT_XRC_TGT; + if ((qp_type < IB_QPT_MAX && !is_xrc) || qp_type == IB_QPT_DRIVER) { qp->res.type = RDMA_RESTRACK_QP; if (uobj) rdma_restrack_uadd(&qp->res); diff --git a/drivers/infiniband/core/counters.c b/drivers/infiniband/core/counters.c index 01faef7bc061..af8c85d18e62 100644 --- a/drivers/infiniband/core/counters.c +++ b/drivers/infiniband/core/counters.c @@ -38,6 +38,9 @@ int rdma_counter_set_auto_mode(struct ib_device *dev, u8 port, int ret; port_counter = &dev->port_data[port].port_counter; + if (!port_counter->hstats) + return -EOPNOTSUPP; + mutex_lock(&port_counter->lock); if (on) { ret = __counter_set_mode(&port_counter->mode, @@ -146,13 +149,11 @@ static bool auto_mode_match(struct ib_qp *qp, struct rdma_counter *counter, struct auto_mode_param *param = &counter->mode.param; bool match = true; - if (rdma_is_kernel_res(&counter->res) != rdma_is_kernel_res(&qp->res)) + if (!rdma_is_visible_in_pid_ns(&qp->res)) return false; - /* Ensure that counter belong to right PID */ - if (!rdma_is_kernel_res(&counter->res) && - !rdma_is_kernel_res(&qp->res) && - (task_pid_vnr(counter->res.task) != current->pid)) + /* Ensure that counter belongs to the right PID */ + if (task_pid_nr(counter->res.task) != task_pid_nr(qp->res.task)) return false; if (auto_mask & RDMA_COUNTER_MASK_QP_TYPE) @@ -393,6 +394,9 @@ u64 rdma_counter_get_hwstat_value(struct ib_device *dev, u8 port, u32 index) u64 sum; port_counter = &dev->port_data[port].port_counter; + if (!port_counter->hstats) + return 0; + sum = get_running_counters_hwstat_sum(dev, port, index); sum += port_counter->hstats->value[index]; @@ -418,7 +422,7 @@ static struct ib_qp *rdma_counter_get_qp(struct ib_device *dev, u32 qp_num) return qp; err: - rdma_restrack_put(&qp->res); + rdma_restrack_put(res); return NULL; } @@ -506,6 +510,9 @@ int rdma_counter_bind_qpn_alloc(struct ib_device *dev, u8 port, if (!rdma_is_port_valid(dev, port)) return -EINVAL; + if (!dev->port_data[port].port_counter.hstats) + return -EOPNOTSUPP; + qp = rdma_counter_get_qp(dev, qp_num); if (!qp) return -ENOENT; @@ -594,7 +601,7 @@ void rdma_counter_init(struct ib_device *dev) struct rdma_port_counter *port_counter; u32 port; - if (!dev->ops.alloc_hw_stats || !dev->port_data) + if (!dev->port_data) return; rdma_for_each_port(dev, port) { @@ -602,6 +609,9 @@ void rdma_counter_init(struct ib_device *dev) port_counter->mode.mode = RDMA_COUNTER_MODE_NONE; mutex_init(&port_counter->lock); + if (!dev->ops.alloc_hw_stats) + continue; + port_counter->hstats = dev->ops.alloc_hw_stats(dev, port); if (!port_counter->hstats) goto fail; @@ -624,9 +634,6 @@ void rdma_counter_release(struct ib_device *dev) struct rdma_port_counter *port_counter; u32 port; - if (!dev->ops.alloc_hw_stats) - return; - rdma_for_each_port(dev, port) { port_counter = &dev->port_data[port].port_counter; kfree(port_counter->hstats); diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c index 9773145dee09..ea8661a00651 100644 --- a/drivers/infiniband/core/device.c +++ b/drivers/infiniband/core/device.c @@ -94,11 +94,17 @@ static DEFINE_XARRAY_FLAGS(devices, XA_FLAGS_ALLOC); static DECLARE_RWSEM(devices_rwsem); #define DEVICE_REGISTERED XA_MARK_1 -static LIST_HEAD(client_list); +static u32 highest_client_id; #define CLIENT_REGISTERED XA_MARK_1 static DEFINE_XARRAY_FLAGS(clients, XA_FLAGS_ALLOC); static DECLARE_RWSEM(clients_rwsem); +static void ib_client_put(struct ib_client *client) +{ + if (refcount_dec_and_test(&client->uses)) + complete(&client->uses_zero); +} + /* * If client_data is registered then the corresponding client must also still * be registered. @@ -661,6 +667,14 @@ static int add_client_context(struct ib_device *device, down_write(&device->client_data_rwsem); /* + * So long as the client is registered hold both the client and device + * unregistration locks. + */ + if (!refcount_inc_not_zero(&client->uses)) + goto out_unlock; + refcount_inc(&device->refcount); + + /* * Another caller to add_client_context got here first and has already * completely initialized context. */ @@ -683,6 +697,9 @@ static int add_client_context(struct ib_device *device, return 0; out: + ib_device_put(device); + ib_client_put(client); +out_unlock: up_write(&device->client_data_rwsem); return ret; } @@ -702,7 +719,7 @@ static void remove_client_context(struct ib_device *device, client_data = xa_load(&device->client_data, client_id); xa_clear_mark(&device->client_data, client_id, CLIENT_DATA_REGISTERED); client = xa_load(&clients, client_id); - downgrade_write(&device->client_data_rwsem); + up_write(&device->client_data_rwsem); /* * Notice we cannot be holding any exclusive locks when calling the @@ -712,17 +729,13 @@ static void remove_client_context(struct ib_device *device, * * For this reason clients and drivers should not call the * unregistration functions will holdling any locks. - * - * It tempting to drop the client_data_rwsem too, but this is required - * to ensure that unregister_client does not return until all clients - * are completely unregistered, which is required to avoid module - * unloading races. */ if (client->remove) client->remove(device, client_data); xa_erase(&device->client_data, client_id); - up_read(&device->client_data_rwsem); + ib_device_put(device); + ib_client_put(client); } static int alloc_port_data(struct ib_device *device) @@ -1224,7 +1237,7 @@ static int setup_device(struct ib_device *device) static void disable_device(struct ib_device *device) { - struct ib_client *client; + u32 cid; WARN_ON(!refcount_read(&device->refcount)); @@ -1232,10 +1245,19 @@ static void disable_device(struct ib_device *device) xa_clear_mark(&devices, device->index, DEVICE_REGISTERED); up_write(&devices_rwsem); + /* + * Remove clients in LIFO order, see assign_client_id. This could be + * more efficient if xarray learns to reverse iterate. Since no new + * clients can be added to this ib_device past this point we only need + * the maximum possible client_id value here. + */ down_read(&clients_rwsem); - list_for_each_entry_reverse(client, &client_list, list) - remove_client_context(device, client->client_id); + cid = highest_client_id; up_read(&clients_rwsem); + while (cid) { + cid--; + remove_client_context(device, cid); + } /* Pairs with refcount_set in enable_device */ ib_device_put(device); @@ -1662,30 +1684,31 @@ static int assign_client_id(struct ib_client *client) /* * The add/remove callbacks must be called in FIFO/LIFO order. To * achieve this we assign client_ids so they are sorted in - * registration order, and retain a linked list we can reverse iterate - * to get the LIFO order. The extra linked list can go away if xarray - * learns to reverse iterate. + * registration order. */ - if (list_empty(&client_list)) { - client->client_id = 0; - } else { - struct ib_client *last; - - last = list_last_entry(&client_list, struct ib_client, list); - client->client_id = last->client_id + 1; - } + client->client_id = highest_client_id; ret = xa_insert(&clients, client->client_id, client, GFP_KERNEL); if (ret) goto out; + highest_client_id++; xa_set_mark(&clients, client->client_id, CLIENT_REGISTERED); - list_add_tail(&client->list, &client_list); out: up_write(&clients_rwsem); return ret; } +static void remove_client_id(struct ib_client *client) +{ + down_write(&clients_rwsem); + xa_erase(&clients, client->client_id); + for (; highest_client_id; highest_client_id--) + if (xa_load(&clients, highest_client_id - 1)) + break; + up_write(&clients_rwsem); +} + /** * ib_register_client - Register an IB client * @client:Client to register @@ -1705,6 +1728,8 @@ int ib_register_client(struct ib_client *client) unsigned long index; int ret; + refcount_set(&client->uses, 1); + init_completion(&client->uses_zero); ret = assign_client_id(client); if (ret) return ret; @@ -1740,21 +1765,30 @@ void ib_unregister_client(struct ib_client *client) unsigned long index; down_write(&clients_rwsem); + ib_client_put(client); xa_clear_mark(&clients, client->client_id, CLIENT_REGISTERED); up_write(&clients_rwsem); - /* - * Every device still known must be serialized to make sure we are - * done with the client callbacks before we return. - */ - down_read(&devices_rwsem); - xa_for_each (&devices, index, device) + + /* We do not want to have locks while calling client->remove() */ + rcu_read_lock(); + xa_for_each (&devices, index, device) { + if (!ib_device_try_get(device)) + continue; + rcu_read_unlock(); + remove_client_context(device, client->client_id); - up_read(&devices_rwsem); - down_write(&clients_rwsem); - list_del(&client->list); - xa_erase(&clients, client->client_id); - up_write(&clients_rwsem); + ib_device_put(device); + rcu_read_lock(); + } + rcu_read_unlock(); + + /* + * remove_client_context() is not a fence, it can return even though a + * removal is ongoing. Wait until all removals are completed. + */ + wait_for_completion(&client->uses_zero); + remove_client_id(client); } EXPORT_SYMBOL(ib_unregister_client); diff --git a/drivers/infiniband/core/mad.c b/drivers/infiniband/core/mad.c index cc99479b2c09..9947d16edef2 100644 --- a/drivers/infiniband/core/mad.c +++ b/drivers/infiniband/core/mad.c @@ -3224,18 +3224,18 @@ static int ib_mad_port_open(struct ib_device *device, if (has_smi) cq_size *= 2; + port_priv->pd = ib_alloc_pd(device, 0); + if (IS_ERR(port_priv->pd)) { + dev_err(&device->dev, "Couldn't create ib_mad PD\n"); + ret = PTR_ERR(port_priv->pd); + goto error3; + } + port_priv->cq = ib_alloc_cq(port_priv->device, port_priv, cq_size, 0, IB_POLL_UNBOUND_WORKQUEUE); if (IS_ERR(port_priv->cq)) { dev_err(&device->dev, "Couldn't create ib_mad CQ\n"); ret = PTR_ERR(port_priv->cq); - goto error3; - } - - port_priv->pd = ib_alloc_pd(device, 0); - if (IS_ERR(port_priv->pd)) { - dev_err(&device->dev, "Couldn't create ib_mad PD\n"); - ret = PTR_ERR(port_priv->pd); goto error4; } @@ -3278,11 +3278,11 @@ error8: error7: destroy_mad_qp(&port_priv->qp_info[0]); error6: - ib_dealloc_pd(port_priv->pd); -error4: ib_free_cq(port_priv->cq); cleanup_recv_queue(&port_priv->qp_info[1]); cleanup_recv_queue(&port_priv->qp_info[0]); +error4: + ib_dealloc_pd(port_priv->pd); error3: kfree(port_priv); @@ -3312,8 +3312,8 @@ static int ib_mad_port_close(struct ib_device *device, int port_num) destroy_workqueue(port_priv->wq); destroy_mad_qp(&port_priv->qp_info[1]); destroy_mad_qp(&port_priv->qp_info[0]); - ib_dealloc_pd(port_priv->pd); ib_free_cq(port_priv->cq); + ib_dealloc_pd(port_priv->pd); cleanup_recv_queue(&port_priv->qp_info[1]); cleanup_recv_queue(&port_priv->qp_info[0]); /* XXX: Handle deallocation of MAD registration tables */ diff --git a/drivers/infiniband/core/nldev.c b/drivers/infiniband/core/nldev.c index 783e465e7c41..020c26976558 100644 --- a/drivers/infiniband/core/nldev.c +++ b/drivers/infiniband/core/nldev.c @@ -382,8 +382,7 @@ static int fill_res_info(struct sk_buff *msg, struct ib_device *device) for (i = 0; i < RDMA_RESTRACK_MAX; i++) { if (!names[i]) continue; - curr = rdma_restrack_count(device, i, - task_active_pid_ns(current)); + curr = rdma_restrack_count(device, i); ret = fill_res_info_entry(msg, names[i], curr); if (ret) goto err; @@ -1952,12 +1951,16 @@ static int stat_get_doit_qp(struct sk_buff *skb, struct nlmsghdr *nlh, if (fill_nldev_handle(msg, device) || nla_put_u32(msg, RDMA_NLDEV_ATTR_PORT_INDEX, port) || - nla_put_u32(msg, RDMA_NLDEV_ATTR_STAT_MODE, mode)) + nla_put_u32(msg, RDMA_NLDEV_ATTR_STAT_MODE, mode)) { + ret = -EMSGSIZE; goto err_msg; + } if ((mode == RDMA_COUNTER_MODE_AUTO) && - nla_put_u32(msg, RDMA_NLDEV_ATTR_STAT_AUTO_MODE_MASK, mask)) + nla_put_u32(msg, RDMA_NLDEV_ATTR_STAT_AUTO_MODE_MASK, mask)) { + ret = -EMSGSIZE; goto err_msg; + } nlmsg_end(msg, nlh); ib_device_put(device); diff --git a/drivers/infiniband/core/restrack.c b/drivers/infiniband/core/restrack.c index bddff426ee0f..a07665f7ef8c 100644 --- a/drivers/infiniband/core/restrack.c +++ b/drivers/infiniband/core/restrack.c @@ -107,10 +107,8 @@ void rdma_restrack_clean(struct ib_device *dev) * rdma_restrack_count() - the current usage of specific object * @dev: IB device * @type: actual type of object to operate - * @ns: PID namespace */ -int rdma_restrack_count(struct ib_device *dev, enum rdma_restrack_type type, - struct pid_namespace *ns) +int rdma_restrack_count(struct ib_device *dev, enum rdma_restrack_type type) { struct rdma_restrack_root *rt = &dev->res[type]; struct rdma_restrack_entry *e; @@ -119,10 +117,9 @@ int rdma_restrack_count(struct ib_device *dev, enum rdma_restrack_type type, xa_lock(&rt->xa); xas_for_each(&xas, e, U32_MAX) { - if (ns == &init_pid_ns || - (!rdma_is_kernel_res(e) && - ns == task_active_pid_ns(e->task))) - cnt++; + if (!rdma_is_visible_in_pid_ns(e)) + continue; + cnt++; } xa_unlock(&rt->xa); return cnt; @@ -360,5 +357,7 @@ bool rdma_is_visible_in_pid_ns(struct rdma_restrack_entry *res) */ if (rdma_is_kernel_res(res)) return task_active_pid_ns(current) == &init_pid_ns; - return task_active_pid_ns(current) == task_active_pid_ns(res->task); + + /* PID 0 means that resource is not found in current namespace */ + return task_pid_vnr(res->task); } diff --git a/drivers/infiniband/core/umem.c b/drivers/infiniband/core/umem.c index 08da840ed7ee..56553668256f 100644 --- a/drivers/infiniband/core/umem.c +++ b/drivers/infiniband/core/umem.c @@ -379,14 +379,9 @@ EXPORT_SYMBOL(ib_umem_release); int ib_umem_page_count(struct ib_umem *umem) { - int i; - int n; + int i, n = 0; struct scatterlist *sg; - if (umem->is_odp) - return ib_umem_num_pages(umem); - - n = 0; for_each_sg(umem->sg_head.sgl, sg, umem->nmap, i) n += sg_dma_len(sg) >> PAGE_SHIFT; diff --git a/drivers/infiniband/core/umem_odp.c b/drivers/infiniband/core/umem_odp.c index 2a75c6f8d827..c0e15db34680 100644 --- a/drivers/infiniband/core/umem_odp.c +++ b/drivers/infiniband/core/umem_odp.c @@ -112,10 +112,6 @@ static int ib_umem_notifier_release_trampoline(struct ib_umem_odp *umem_odp, * prevent any further fault handling on this MR. */ ib_umem_notifier_start_account(umem_odp); - umem_odp->dying = 1; - /* Make sure that the fact the umem is dying is out before we release - * all pending page faults. */ - smp_wmb(); complete_all(&umem_odp->notifier_completion); umem_odp->umem.context->invalidate_range( umem_odp, ib_umem_start(umem_odp), ib_umem_end(umem_odp)); diff --git a/drivers/infiniband/core/user_mad.c b/drivers/infiniband/core/user_mad.c index 9f8a48016b41..ffdeaf6e0b68 100644 --- a/drivers/infiniband/core/user_mad.c +++ b/drivers/infiniband/core/user_mad.c @@ -49,6 +49,7 @@ #include <linux/sched.h> #include <linux/semaphore.h> #include <linux/slab.h> +#include <linux/nospec.h> #include <linux/uaccess.h> @@ -884,11 +885,14 @@ static int ib_umad_unreg_agent(struct ib_umad_file *file, u32 __user *arg) if (get_user(id, arg)) return -EFAULT; + if (id >= IB_UMAD_MAX_AGENTS) + return -EINVAL; mutex_lock(&file->port->file_mutex); mutex_lock(&file->mutex); - if (id >= IB_UMAD_MAX_AGENTS || !__get_agent(file, id)) { + id = array_index_nospec(id, IB_UMAD_MAX_AGENTS); + if (!__get_agent(file, id)) { ret = -EINVAL; goto out; } diff --git a/drivers/infiniband/hw/bnxt_re/ib_verbs.c b/drivers/infiniband/hw/bnxt_re/ib_verbs.c index a91653aabf38..098ab883733e 100644 --- a/drivers/infiniband/hw/bnxt_re/ib_verbs.c +++ b/drivers/infiniband/hw/bnxt_re/ib_verbs.c @@ -308,6 +308,7 @@ int bnxt_re_del_gid(const struct ib_gid_attr *attr, void **context) struct bnxt_re_dev *rdev = to_bnxt_re_dev(attr->device, ibdev); struct bnxt_qplib_sgid_tbl *sgid_tbl = &rdev->qplib_res.sgid_tbl; struct bnxt_qplib_gid *gid_to_del; + u16 vlan_id = 0xFFFF; /* Delete the entry from the hardware */ ctx = *context; @@ -317,7 +318,8 @@ int bnxt_re_del_gid(const struct ib_gid_attr *attr, void **context) if (sgid_tbl && sgid_tbl->active) { if (ctx->idx >= sgid_tbl->max) return -EINVAL; - gid_to_del = &sgid_tbl->tbl[ctx->idx]; + gid_to_del = &sgid_tbl->tbl[ctx->idx].gid; + vlan_id = sgid_tbl->tbl[ctx->idx].vlan_id; /* DEL_GID is called in WQ context(netdevice_event_work_handler) * or via the ib_unregister_device path. In the former case QP1 * may not be destroyed yet, in which case just return as FW @@ -335,7 +337,8 @@ int bnxt_re_del_gid(const struct ib_gid_attr *attr, void **context) } ctx->refcnt--; if (!ctx->refcnt) { - rc = bnxt_qplib_del_sgid(sgid_tbl, gid_to_del, true); + rc = bnxt_qplib_del_sgid(sgid_tbl, gid_to_del, + vlan_id, true); if (rc) { dev_err(rdev_to_dev(rdev), "Failed to remove GID: %#x", rc); diff --git a/drivers/infiniband/hw/bnxt_re/qplib_rcfw.c b/drivers/infiniband/hw/bnxt_re/qplib_rcfw.c index 48b04d2f175f..60c8f76aab33 100644 --- a/drivers/infiniband/hw/bnxt_re/qplib_rcfw.c +++ b/drivers/infiniband/hw/bnxt_re/qplib_rcfw.c @@ -136,6 +136,13 @@ static int __send_message(struct bnxt_qplib_rcfw *rcfw, struct cmdq_base *req, spin_unlock_irqrestore(&cmdq->lock, flags); return -EBUSY; } + + size = req->cmd_size; + /* change the cmd_size to the number of 16byte cmdq unit. + * req->cmd_size is modified here + */ + bnxt_qplib_set_cmd_slots(req); + memset(resp, 0, sizeof(*resp)); crsqe->resp = (struct creq_qp_event *)resp; crsqe->resp->cookie = req->cookie; @@ -150,7 +157,6 @@ static int __send_message(struct bnxt_qplib_rcfw *rcfw, struct cmdq_base *req, cmdq_ptr = (struct bnxt_qplib_cmdqe **)cmdq->pbl_ptr; preq = (u8 *)req; - size = req->cmd_size * BNXT_QPLIB_CMDQE_UNITS; do { /* Locate the next cmdq slot */ sw_prod = HWQ_CMP(cmdq->prod, cmdq); diff --git a/drivers/infiniband/hw/bnxt_re/qplib_rcfw.h b/drivers/infiniband/hw/bnxt_re/qplib_rcfw.h index 2138533bb642..dfeadc192e17 100644 --- a/drivers/infiniband/hw/bnxt_re/qplib_rcfw.h +++ b/drivers/infiniband/hw/bnxt_re/qplib_rcfw.h @@ -55,9 +55,7 @@ do { \ memset(&(req), 0, sizeof((req))); \ (req).opcode = CMDQ_BASE_OPCODE_##CMD; \ - (req).cmd_size = (sizeof((req)) + \ - BNXT_QPLIB_CMDQE_UNITS - 1) / \ - BNXT_QPLIB_CMDQE_UNITS; \ + (req).cmd_size = sizeof((req)); \ (req).flags = cpu_to_le16(cmd_flags); \ } while (0) @@ -95,6 +93,13 @@ static inline u32 bnxt_qplib_cmdqe_cnt_per_pg(u32 depth) BNXT_QPLIB_CMDQE_UNITS); } +/* Set the cmd_size to a factor of CMDQE unit */ +static inline void bnxt_qplib_set_cmd_slots(struct cmdq_base *req) +{ + req->cmd_size = (req->cmd_size + BNXT_QPLIB_CMDQE_UNITS - 1) / + BNXT_QPLIB_CMDQE_UNITS; +} + #define MAX_CMDQ_IDX(depth) ((depth) - 1) static inline u32 bnxt_qplib_max_cmdq_idx_per_pg(u32 depth) diff --git a/drivers/infiniband/hw/bnxt_re/qplib_res.c b/drivers/infiniband/hw/bnxt_re/qplib_res.c index 37928b1111df..bdbde8e22420 100644 --- a/drivers/infiniband/hw/bnxt_re/qplib_res.c +++ b/drivers/infiniband/hw/bnxt_re/qplib_res.c @@ -488,7 +488,7 @@ static int bnxt_qplib_alloc_sgid_tbl(struct bnxt_qplib_res *res, struct bnxt_qplib_sgid_tbl *sgid_tbl, u16 max) { - sgid_tbl->tbl = kcalloc(max, sizeof(struct bnxt_qplib_gid), GFP_KERNEL); + sgid_tbl->tbl = kcalloc(max, sizeof(*sgid_tbl->tbl), GFP_KERNEL); if (!sgid_tbl->tbl) return -ENOMEM; @@ -526,9 +526,10 @@ static void bnxt_qplib_cleanup_sgid_tbl(struct bnxt_qplib_res *res, for (i = 0; i < sgid_tbl->max; i++) { if (memcmp(&sgid_tbl->tbl[i], &bnxt_qplib_gid_zero, sizeof(bnxt_qplib_gid_zero))) - bnxt_qplib_del_sgid(sgid_tbl, &sgid_tbl->tbl[i], true); + bnxt_qplib_del_sgid(sgid_tbl, &sgid_tbl->tbl[i].gid, + sgid_tbl->tbl[i].vlan_id, true); } - memset(sgid_tbl->tbl, 0, sizeof(struct bnxt_qplib_gid) * sgid_tbl->max); + memset(sgid_tbl->tbl, 0, sizeof(*sgid_tbl->tbl) * sgid_tbl->max); memset(sgid_tbl->hw_id, -1, sizeof(u16) * sgid_tbl->max); memset(sgid_tbl->vlan, 0, sizeof(u8) * sgid_tbl->max); sgid_tbl->active = 0; @@ -537,7 +538,11 @@ static void bnxt_qplib_cleanup_sgid_tbl(struct bnxt_qplib_res *res, static void bnxt_qplib_init_sgid_tbl(struct bnxt_qplib_sgid_tbl *sgid_tbl, struct net_device *netdev) { - memset(sgid_tbl->tbl, 0, sizeof(struct bnxt_qplib_gid) * sgid_tbl->max); + u32 i; + + for (i = 0; i < sgid_tbl->max; i++) + sgid_tbl->tbl[i].vlan_id = 0xffff; + memset(sgid_tbl->hw_id, -1, sizeof(u16) * sgid_tbl->max); } diff --git a/drivers/infiniband/hw/bnxt_re/qplib_res.h b/drivers/infiniband/hw/bnxt_re/qplib_res.h index 30c42c92fac7..fbda11a7ab1a 100644 --- a/drivers/infiniband/hw/bnxt_re/qplib_res.h +++ b/drivers/infiniband/hw/bnxt_re/qplib_res.h @@ -111,7 +111,7 @@ struct bnxt_qplib_pd_tbl { }; struct bnxt_qplib_sgid_tbl { - struct bnxt_qplib_gid *tbl; + struct bnxt_qplib_gid_info *tbl; u16 *hw_id; u16 max; u16 active; diff --git a/drivers/infiniband/hw/bnxt_re/qplib_sp.c b/drivers/infiniband/hw/bnxt_re/qplib_sp.c index 48793d3512ac..40296b97d21e 100644 --- a/drivers/infiniband/hw/bnxt_re/qplib_sp.c +++ b/drivers/infiniband/hw/bnxt_re/qplib_sp.c @@ -213,12 +213,12 @@ int bnxt_qplib_get_sgid(struct bnxt_qplib_res *res, index, sgid_tbl->max); return -EINVAL; } - memcpy(gid, &sgid_tbl->tbl[index], sizeof(*gid)); + memcpy(gid, &sgid_tbl->tbl[index].gid, sizeof(*gid)); return 0; } int bnxt_qplib_del_sgid(struct bnxt_qplib_sgid_tbl *sgid_tbl, - struct bnxt_qplib_gid *gid, bool update) + struct bnxt_qplib_gid *gid, u16 vlan_id, bool update) { struct bnxt_qplib_res *res = to_bnxt_qplib(sgid_tbl, struct bnxt_qplib_res, @@ -236,7 +236,8 @@ int bnxt_qplib_del_sgid(struct bnxt_qplib_sgid_tbl *sgid_tbl, return -ENOMEM; } for (index = 0; index < sgid_tbl->max; index++) { - if (!memcmp(&sgid_tbl->tbl[index], gid, sizeof(*gid))) + if (!memcmp(&sgid_tbl->tbl[index].gid, gid, sizeof(*gid)) && + vlan_id == sgid_tbl->tbl[index].vlan_id) break; } if (index == sgid_tbl->max) { @@ -262,8 +263,9 @@ int bnxt_qplib_del_sgid(struct bnxt_qplib_sgid_tbl *sgid_tbl, if (rc) return rc; } - memcpy(&sgid_tbl->tbl[index], &bnxt_qplib_gid_zero, + memcpy(&sgid_tbl->tbl[index].gid, &bnxt_qplib_gid_zero, sizeof(bnxt_qplib_gid_zero)); + sgid_tbl->tbl[index].vlan_id = 0xFFFF; sgid_tbl->vlan[index] = 0; sgid_tbl->active--; dev_dbg(&res->pdev->dev, @@ -296,7 +298,8 @@ int bnxt_qplib_add_sgid(struct bnxt_qplib_sgid_tbl *sgid_tbl, } free_idx = sgid_tbl->max; for (i = 0; i < sgid_tbl->max; i++) { - if (!memcmp(&sgid_tbl->tbl[i], gid, sizeof(*gid))) { + if (!memcmp(&sgid_tbl->tbl[i], gid, sizeof(*gid)) && + sgid_tbl->tbl[i].vlan_id == vlan_id) { dev_dbg(&res->pdev->dev, "SGID entry already exist in entry %d!\n", i); *index = i; @@ -351,6 +354,7 @@ int bnxt_qplib_add_sgid(struct bnxt_qplib_sgid_tbl *sgid_tbl, } /* Add GID to the sgid_tbl */ memcpy(&sgid_tbl->tbl[free_idx], gid, sizeof(*gid)); + sgid_tbl->tbl[free_idx].vlan_id = vlan_id; sgid_tbl->active++; if (vlan_id != 0xFFFF) sgid_tbl->vlan[free_idx] = 1; diff --git a/drivers/infiniband/hw/bnxt_re/qplib_sp.h b/drivers/infiniband/hw/bnxt_re/qplib_sp.h index 0ec3b12b0bcd..13d9432d5ce2 100644 --- a/drivers/infiniband/hw/bnxt_re/qplib_sp.h +++ b/drivers/infiniband/hw/bnxt_re/qplib_sp.h @@ -84,6 +84,11 @@ struct bnxt_qplib_gid { u8 data[16]; }; +struct bnxt_qplib_gid_info { + struct bnxt_qplib_gid gid; + u16 vlan_id; +}; + struct bnxt_qplib_ah { struct bnxt_qplib_gid dgid; struct bnxt_qplib_pd *pd; @@ -221,7 +226,7 @@ int bnxt_qplib_get_sgid(struct bnxt_qplib_res *res, struct bnxt_qplib_sgid_tbl *sgid_tbl, int index, struct bnxt_qplib_gid *gid); int bnxt_qplib_del_sgid(struct bnxt_qplib_sgid_tbl *sgid_tbl, - struct bnxt_qplib_gid *gid, bool update); + struct bnxt_qplib_gid *gid, u16 vlan_id, bool update); int bnxt_qplib_add_sgid(struct bnxt_qplib_sgid_tbl *sgid_tbl, struct bnxt_qplib_gid *gid, u8 *mac, u16 vlan_id, bool update, u32 *index); diff --git a/drivers/infiniband/hw/hfi1/chip.c b/drivers/infiniband/hw/hfi1/chip.c index d5b643a1d9fd..67052dc3100c 100644 --- a/drivers/infiniband/hw/hfi1/chip.c +++ b/drivers/infiniband/hw/hfi1/chip.c @@ -14452,7 +14452,7 @@ void hfi1_deinit_vnic_rsm(struct hfi1_devdata *dd) clear_rcvctrl(dd, RCV_CTRL_RCV_RSM_ENABLE_SMASK); } -static void init_rxe(struct hfi1_devdata *dd) +static int init_rxe(struct hfi1_devdata *dd) { struct rsm_map_table *rmt; u64 val; @@ -14461,6 +14461,9 @@ static void init_rxe(struct hfi1_devdata *dd) write_csr(dd, RCV_ERR_MASK, ~0ull); rmt = alloc_rsm_map_table(dd); + if (!rmt) + return -ENOMEM; + /* set up QOS, including the QPN map table */ init_qos(dd, rmt); init_fecn_handling(dd, rmt); @@ -14487,6 +14490,7 @@ static void init_rxe(struct hfi1_devdata *dd) val |= ((4ull & RCV_BYPASS_HDR_SIZE_MASK) << RCV_BYPASS_HDR_SIZE_SHIFT); write_csr(dd, RCV_BYPASS, val); + return 0; } static void init_other(struct hfi1_devdata *dd) @@ -15024,7 +15028,10 @@ int hfi1_init_dd(struct hfi1_devdata *dd) goto bail_cleanup; /* set initial RXE CSRs */ - init_rxe(dd); + ret = init_rxe(dd); + if (ret) + goto bail_cleanup; + /* set initial TXE CSRs */ init_txe(dd); /* set initial non-RXE, non-TXE CSRs */ diff --git a/drivers/infiniband/hw/hfi1/fault.c b/drivers/infiniband/hw/hfi1/fault.c index 93613e5def9b..986c12153e62 100644 --- a/drivers/infiniband/hw/hfi1/fault.c +++ b/drivers/infiniband/hw/hfi1/fault.c @@ -141,12 +141,14 @@ static ssize_t fault_opcodes_write(struct file *file, const char __user *buf, if (!data) return -ENOMEM; copy = min(len, datalen - 1); - if (copy_from_user(data, buf, copy)) - return -EFAULT; + if (copy_from_user(data, buf, copy)) { + ret = -EFAULT; + goto free_data; + } ret = debugfs_file_get(file->f_path.dentry); if (unlikely(ret)) - return ret; + goto free_data; ptr = data; token = ptr; for (ptr = data; *ptr; ptr = end + 1, token = ptr) { @@ -195,6 +197,7 @@ static ssize_t fault_opcodes_write(struct file *file, const char __user *buf, ret = len; debugfs_file_put(file->f_path.dentry); +free_data: kfree(data); return ret; } @@ -214,7 +217,7 @@ static ssize_t fault_opcodes_read(struct file *file, char __user *buf, return -ENOMEM; ret = debugfs_file_get(file->f_path.dentry); if (unlikely(ret)) - return ret; + goto free_data; bit = find_first_bit(fault->opcodes, bitsize); while (bit < bitsize) { zero = find_next_zero_bit(fault->opcodes, bitsize, bit); @@ -232,6 +235,7 @@ static ssize_t fault_opcodes_read(struct file *file, char __user *buf, data[size - 1] = '\n'; data[size] = '\0'; ret = simple_read_from_buffer(buf, len, pos, data, size); +free_data: kfree(data); return ret; } diff --git a/drivers/infiniband/hw/hfi1/rc.c b/drivers/infiniband/hw/hfi1/rc.c index 0477c14633ab..024a7c2b6124 100644 --- a/drivers/infiniband/hw/hfi1/rc.c +++ b/drivers/infiniband/hw/hfi1/rc.c @@ -1835,7 +1835,6 @@ void hfi1_rc_send_complete(struct rvt_qp *qp, struct hfi1_opa_header *opah) cmp_psn(qp->s_sending_psn, qp->s_sending_hpsn) <= 0) break; trdma_clean_swqe(qp, wqe); - rvt_qp_wqe_unreserve(qp, wqe); trace_hfi1_qp_send_completion(qp, wqe, qp->s_last); rvt_qp_complete_swqe(qp, wqe, @@ -1882,7 +1881,6 @@ struct rvt_swqe *do_rc_completion(struct rvt_qp *qp, if (cmp_psn(wqe->lpsn, qp->s_sending_psn) < 0 || cmp_psn(qp->s_sending_psn, qp->s_sending_hpsn) > 0) { trdma_clean_swqe(qp, wqe); - rvt_qp_wqe_unreserve(qp, wqe); trace_hfi1_qp_send_completion(qp, wqe, qp->s_last); rvt_qp_complete_swqe(qp, wqe, diff --git a/drivers/infiniband/hw/hfi1/tid_rdma.c b/drivers/infiniband/hw/hfi1/tid_rdma.c index 92acccaaaa86..6141f4edc6bf 100644 --- a/drivers/infiniband/hw/hfi1/tid_rdma.c +++ b/drivers/infiniband/hw/hfi1/tid_rdma.c @@ -1620,6 +1620,7 @@ static int hfi1_kern_exp_rcv_alloc_flows(struct tid_rdma_request *req, flows[i].req = req; flows[i].npagesets = 0; flows[i].pagesets[0].mapped = 0; + flows[i].resync_npkts = 0; } req->flows = flows; return 0; @@ -1673,34 +1674,6 @@ static struct tid_rdma_flow *find_flow_ib(struct tid_rdma_request *req, return NULL; } -static struct tid_rdma_flow * -__find_flow_ranged(struct tid_rdma_request *req, u16 head, u16 tail, - u32 psn, u16 *fidx) -{ - for ( ; CIRC_CNT(head, tail, MAX_FLOWS); - tail = CIRC_NEXT(tail, MAX_FLOWS)) { - struct tid_rdma_flow *flow = &req->flows[tail]; - u32 spsn, lpsn; - - spsn = full_flow_psn(flow, flow->flow_state.spsn); - lpsn = full_flow_psn(flow, flow->flow_state.lpsn); - - if (cmp_psn(psn, spsn) >= 0 && cmp_psn(psn, lpsn) <= 0) { - if (fidx) - *fidx = tail; - return flow; - } - } - return NULL; -} - -static struct tid_rdma_flow *find_flow(struct tid_rdma_request *req, - u32 psn, u16 *fidx) -{ - return __find_flow_ranged(req, req->setup_head, req->clear_tail, psn, - fidx); -} - /* TID RDMA READ functions */ u32 hfi1_build_tid_rdma_read_packet(struct rvt_swqe *wqe, struct ib_other_headers *ohdr, u32 *bth1, @@ -2601,18 +2574,9 @@ void hfi1_kern_read_tid_flow_free(struct rvt_qp *qp) hfi1_kern_clear_hw_flow(priv->rcd, qp); } -static bool tid_rdma_tid_err(struct hfi1_ctxtdata *rcd, - struct hfi1_packet *packet, u8 rcv_type, - u8 opcode) +static bool tid_rdma_tid_err(struct hfi1_packet *packet, u8 rcv_type) { struct rvt_qp *qp = packet->qp; - struct hfi1_qp_priv *qpriv = qp->priv; - u32 ipsn; - struct ib_other_headers *ohdr = packet->ohdr; - struct rvt_ack_entry *e; - struct tid_rdma_request *req; - struct rvt_dev_info *rdi = ib_to_rvt(qp->ibqp.device); - u32 i; if (rcv_type >= RHF_RCV_TYPE_IB) goto done; @@ -2629,41 +2593,9 @@ static bool tid_rdma_tid_err(struct hfi1_ctxtdata *rcd, if (rcv_type == RHF_RCV_TYPE_EAGER) { hfi1_restart_rc(qp, qp->s_last_psn + 1, 1); hfi1_schedule_send(qp); - goto done_unlock; - } - - /* - * For TID READ response, error out QP after freeing the tid - * resources. - */ - if (opcode == TID_OP(READ_RESP)) { - ipsn = mask_psn(be32_to_cpu(ohdr->u.tid_rdma.r_rsp.verbs_psn)); - if (cmp_psn(ipsn, qp->s_last_psn) > 0 && - cmp_psn(ipsn, qp->s_psn) < 0) { - hfi1_kern_read_tid_flow_free(qp); - spin_unlock(&qp->s_lock); - rvt_rc_error(qp, IB_WC_LOC_QP_OP_ERR); - goto done; - } - goto done_unlock; } - /* - * Error out the qp for TID RDMA WRITE - */ - hfi1_kern_clear_hw_flow(qpriv->rcd, qp); - for (i = 0; i < rvt_max_atomic(rdi); i++) { - e = &qp->s_ack_queue[i]; - if (e->opcode == TID_OP(WRITE_REQ)) { - req = ack_to_tid_req(e); - hfi1_kern_exp_rcv_clear_all(req); - } - } - spin_unlock(&qp->s_lock); - rvt_rc_error(qp, IB_WC_LOC_LEN_ERR); - goto done; - -done_unlock: + /* Since no payload is delivered, just drop the packet */ spin_unlock(&qp->s_lock); done: return true; @@ -2714,12 +2646,12 @@ static bool handle_read_kdeth_eflags(struct hfi1_ctxtdata *rcd, u32 fpsn; lockdep_assert_held(&qp->r_lock); + spin_lock(&qp->s_lock); /* If the psn is out of valid range, drop the packet */ if (cmp_psn(ibpsn, qp->s_last_psn) < 0 || cmp_psn(ibpsn, qp->s_psn) > 0) - return ret; + goto s_unlock; - spin_lock(&qp->s_lock); /* * Note that NAKs implicitly ACK outstanding SEND and RDMA write * requests and implicitly NAK RDMA read and atomic requests issued @@ -2767,9 +2699,12 @@ static bool handle_read_kdeth_eflags(struct hfi1_ctxtdata *rcd, wqe = do_rc_completion(qp, wqe, ibp); if (qp->s_acked == qp->s_tail) - break; + goto s_unlock; } + if (qp->s_acked == qp->s_tail) + goto s_unlock; + /* Handle the eflags for the request */ if (wqe->wr.opcode != IB_WR_TID_RDMA_READ) goto s_unlock; @@ -2788,19 +2723,7 @@ static bool handle_read_kdeth_eflags(struct hfi1_ctxtdata *rcd, * to prevent continuous Flow Sequence errors for any * packets that could be still in the fabric. */ - flow = find_flow(req, psn, NULL); - if (!flow) { - /* - * We can't find the IB PSN matching the - * received KDETH PSN. The only thing we can - * do at this point is report the error to - * the QP. - */ - hfi1_kern_read_tid_flow_free(qp); - spin_unlock(&qp->s_lock); - rvt_rc_error(qp, IB_WC_LOC_QP_OP_ERR); - return ret; - } + flow = &req->flows[req->clear_tail]; if (priv->s_flags & HFI1_R_TID_SW_PSN) { diff = cmp_psn(psn, flow->flow_state.r_next_psn); @@ -2961,7 +2884,7 @@ bool hfi1_handle_kdeth_eflags(struct hfi1_ctxtdata *rcd, if (lnh == HFI1_LRH_GRH) goto r_unlock; - if (tid_rdma_tid_err(rcd, packet, rcv_type, opcode)) + if (tid_rdma_tid_err(packet, rcv_type)) goto r_unlock; } @@ -2981,8 +2904,15 @@ bool hfi1_handle_kdeth_eflags(struct hfi1_ctxtdata *rcd, */ spin_lock(&qp->s_lock); qpriv = qp->priv; + if (qpriv->r_tid_tail == HFI1_QP_WQE_INVALID || + qpriv->r_tid_tail == qpriv->r_tid_head) + goto unlock; e = &qp->s_ack_queue[qpriv->r_tid_tail]; + if (e->opcode != TID_OP(WRITE_REQ)) + goto unlock; req = ack_to_tid_req(e); + if (req->comp_seg == req->cur_seg) + goto unlock; flow = &req->flows[req->clear_tail]; trace_hfi1_eflags_err_write(qp, rcv_type, rte, psn); trace_hfi1_rsp_handle_kdeth_eflags(qp, psn); @@ -4548,7 +4478,7 @@ void hfi1_rc_rcv_tid_rdma_ack(struct hfi1_packet *packet) struct rvt_swqe *wqe; struct tid_rdma_request *req; struct tid_rdma_flow *flow; - u32 aeth, psn, req_psn, ack_psn, resync_psn, ack_kpsn; + u32 aeth, psn, req_psn, ack_psn, flpsn, resync_psn, ack_kpsn; unsigned long flags; u16 fidx; @@ -4577,6 +4507,9 @@ void hfi1_rc_rcv_tid_rdma_ack(struct hfi1_packet *packet) ack_kpsn--; } + if (unlikely(qp->s_acked == qp->s_tail)) + goto ack_op_err; + wqe = rvt_get_swqe_ptr(qp, qp->s_acked); if (wqe->wr.opcode != IB_WR_TID_RDMA_WRITE) @@ -4589,7 +4522,8 @@ void hfi1_rc_rcv_tid_rdma_ack(struct hfi1_packet *packet) trace_hfi1_tid_flow_rcv_tid_ack(qp, req->acked_tail, flow); /* Drop stale ACK/NAK */ - if (cmp_psn(psn, full_flow_psn(flow, flow->flow_state.spsn)) < 0) + if (cmp_psn(psn, full_flow_psn(flow, flow->flow_state.spsn)) < 0 || + cmp_psn(req_psn, flow->flow_state.resp_ib_psn) < 0) goto ack_op_err; while (cmp_psn(ack_kpsn, @@ -4751,7 +4685,12 @@ done: switch ((aeth >> IB_AETH_CREDIT_SHIFT) & IB_AETH_CREDIT_MASK) { case 0: /* PSN sequence error */ + if (!req->flows) + break; flow = &req->flows[req->acked_tail]; + flpsn = full_flow_psn(flow, flow->flow_state.lpsn); + if (cmp_psn(psn, flpsn) > 0) + break; trace_hfi1_tid_flow_rcv_tid_ack(qp, req->acked_tail, flow); req->r_ack_psn = mask_psn(be32_to_cpu(ohdr->bth[2])); diff --git a/drivers/infiniband/hw/hfi1/verbs.c b/drivers/infiniband/hw/hfi1/verbs.c index c4b243f50c76..646f61545ed6 100644 --- a/drivers/infiniband/hw/hfi1/verbs.c +++ b/drivers/infiniband/hw/hfi1/verbs.c @@ -54,6 +54,7 @@ #include <linux/mm.h> #include <linux/vmalloc.h> #include <rdma/opa_addr.h> +#include <linux/nospec.h> #include "hfi.h" #include "common.h" @@ -1536,6 +1537,7 @@ static int hfi1_check_ah(struct ib_device *ibdev, struct rdma_ah_attr *ah_attr) sl = rdma_ah_get_sl(ah_attr); if (sl >= ARRAY_SIZE(ibp->sl_to_sc)) return -EINVAL; + sl = array_index_nospec(sl, ARRAY_SIZE(ibp->sl_to_sc)); sc5 = ibp->sl_to_sc[sl]; if (sc_to_vlt(dd, sc5) > num_vls && sc_to_vlt(dd, sc5) != 0xf) diff --git a/drivers/infiniband/hw/hns/Kconfig b/drivers/infiniband/hw/hns/Kconfig index 8bf847bcd8d3..54782197c717 100644 --- a/drivers/infiniband/hw/hns/Kconfig +++ b/drivers/infiniband/hw/hns/Kconfig @@ -1,6 +1,6 @@ # SPDX-License-Identifier: GPL-2.0-only config INFINIBAND_HNS - tristate "HNS RoCE Driver" + bool "HNS RoCE Driver" depends on NET_VENDOR_HISILICON depends on ARM64 || (COMPILE_TEST && 64BIT) ---help--- @@ -11,7 +11,7 @@ config INFINIBAND_HNS To compile HIP06 or HIP08 driver as module, choose M here. config INFINIBAND_HNS_HIP06 - bool "Hisilicon Hip06 Family RoCE support" + tristate "Hisilicon Hip06 Family RoCE support" depends on INFINIBAND_HNS && HNS && HNS_DSAF && HNS_ENET ---help--- RoCE driver support for Hisilicon RoCE engine in Hisilicon Hip06 and @@ -21,7 +21,7 @@ config INFINIBAND_HNS_HIP06 module will be called hns-roce-hw-v1 config INFINIBAND_HNS_HIP08 - bool "Hisilicon Hip08 Family RoCE support" + tristate "Hisilicon Hip08 Family RoCE support" depends on INFINIBAND_HNS && PCI && HNS3 ---help--- RoCE driver support for Hisilicon RoCE engine in Hisilicon Hip08 SoC. diff --git a/drivers/infiniband/hw/hns/Makefile b/drivers/infiniband/hw/hns/Makefile index e105945b94a1..449a2d81319d 100644 --- a/drivers/infiniband/hw/hns/Makefile +++ b/drivers/infiniband/hw/hns/Makefile @@ -9,12 +9,8 @@ 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 hns_roce_db.o hns_roce_srq.o hns_roce_restrack.o -ifdef CONFIG_INFINIBAND_HNS_HIP06 hns-roce-hw-v1-objs := hns_roce_hw_v1.o $(hns-roce-objs) -obj-$(CONFIG_INFINIBAND_HNS) += hns-roce-hw-v1.o -endif +obj-$(CONFIG_INFINIBAND_HNS_HIP06) += hns-roce-hw-v1.o -ifdef CONFIG_INFINIBAND_HNS_HIP08 hns-roce-hw-v2-objs := hns_roce_hw_v2.o hns_roce_hw_v2_dfx.o $(hns-roce-objs) -obj-$(CONFIG_INFINIBAND_HNS) += hns-roce-hw-v2.o -endif +obj-$(CONFIG_INFINIBAND_HNS_HIP08) += hns-roce-hw-v2.o diff --git a/drivers/infiniband/hw/hns/hns_roce_db.c b/drivers/infiniband/hw/hns/hns_roce_db.c index 627aa46ef683..c00714c2f16a 100644 --- a/drivers/infiniband/hw/hns/hns_roce_db.c +++ b/drivers/infiniband/hw/hns/hns_roce_db.c @@ -12,13 +12,15 @@ int hns_roce_db_map_user(struct hns_roce_ucontext *context, struct ib_udata *udata, unsigned long virt, struct hns_roce_db *db) { + unsigned long page_addr = virt & PAGE_MASK; struct hns_roce_user_db_page *page; + unsigned int offset; int ret = 0; mutex_lock(&context->page_mutex); list_for_each_entry(page, &context->page_list, list) - if (page->user_virt == (virt & PAGE_MASK)) + if (page->user_virt == page_addr) goto found; page = kmalloc(sizeof(*page), GFP_KERNEL); @@ -28,8 +30,8 @@ int hns_roce_db_map_user(struct hns_roce_ucontext *context, } refcount_set(&page->refcount, 1); - page->user_virt = (virt & PAGE_MASK); - page->umem = ib_umem_get(udata, virt & PAGE_MASK, PAGE_SIZE, 0, 0); + page->user_virt = page_addr; + page->umem = ib_umem_get(udata, page_addr, PAGE_SIZE, 0, 0); if (IS_ERR(page->umem)) { ret = PTR_ERR(page->umem); kfree(page); @@ -39,10 +41,9 @@ int hns_roce_db_map_user(struct hns_roce_ucontext *context, list_add(&page->list, &context->page_list); found: - db->dma = sg_dma_address(page->umem->sg_head.sgl) + - (virt & ~PAGE_MASK); - page->umem->sg_head.sgl->offset = virt & ~PAGE_MASK; - db->virt_addr = sg_virt(page->umem->sg_head.sgl); + offset = virt - page_addr; + db->dma = sg_dma_address(page->umem->sg_head.sgl) + offset; + db->virt_addr = sg_virt(page->umem->sg_head.sgl) + offset; db->u.user_page = page; refcount_inc(&page->refcount); diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c index 81e6dedb1e02..c07e387a07a3 100644 --- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c +++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c @@ -750,8 +750,10 @@ static int hns_roce_v1_rsv_lp_qp(struct hns_roce_dev *hr_dev) atomic_set(&free_mr->mr_free_cq->ib_cq.usecnt, 0); pd = rdma_zalloc_drv_obj(ibdev, ib_pd); - if (!pd) + if (!pd) { + ret = -ENOMEM; goto alloc_mem_failed; + } pd->device = ibdev; ret = hns_roce_alloc_pd(pd, NULL); diff --git a/drivers/infiniband/hw/mlx4/mad.c b/drivers/infiniband/hw/mlx4/mad.c index 68c951491a08..57079110af9b 100644 --- a/drivers/infiniband/hw/mlx4/mad.c +++ b/drivers/infiniband/hw/mlx4/mad.c @@ -1677,8 +1677,6 @@ tx_err: tx_buf_size, DMA_TO_DEVICE); kfree(tun_qp->tx_ring[i].buf.addr); } - kfree(tun_qp->tx_ring); - tun_qp->tx_ring = NULL; i = MLX4_NUM_TUNNEL_BUFS; err: while (i > 0) { @@ -1687,6 +1685,8 @@ err: rx_buf_size, DMA_FROM_DEVICE); kfree(tun_qp->ring[i].addr); } + kfree(tun_qp->tx_ring); + tun_qp->tx_ring = NULL; kfree(tun_qp->ring); tun_qp->ring = NULL; return -ENOMEM; diff --git a/drivers/infiniband/hw/mlx5/devx.c b/drivers/infiniband/hw/mlx5/devx.c index ec4370f99381..af5bbb35c058 100644 --- a/drivers/infiniband/hw/mlx5/devx.c +++ b/drivers/infiniband/hw/mlx5/devx.c @@ -2026,7 +2026,7 @@ static int UVERBS_HANDLER(MLX5_IB_METHOD_DEVX_SUBSCRIBE_EVENT)( event_sub->eventfd = eventfd_ctx_fdget(redirect_fd); - if (IS_ERR(event_sub)) { + if (IS_ERR(event_sub->eventfd)) { err = PTR_ERR(event_sub->eventfd); event_sub->eventfd = NULL; goto err; @@ -2644,12 +2644,13 @@ static int devx_async_event_close(struct inode *inode, struct file *filp) struct devx_async_event_file *ev_file = filp->private_data; struct devx_event_subscription *event_sub, *event_sub_tmp; struct devx_async_event_data *entry, *tmp; + struct mlx5_ib_dev *dev = ev_file->dev; - mutex_lock(&ev_file->dev->devx_event_table.event_xa_lock); + mutex_lock(&dev->devx_event_table.event_xa_lock); /* delete the subscriptions which are related to this FD */ list_for_each_entry_safe(event_sub, event_sub_tmp, &ev_file->subscribed_events_list, file_list) { - devx_cleanup_subscription(ev_file->dev, event_sub); + devx_cleanup_subscription(dev, event_sub); if (event_sub->eventfd) eventfd_ctx_put(event_sub->eventfd); @@ -2658,7 +2659,7 @@ static int devx_async_event_close(struct inode *inode, struct file *filp) kfree_rcu(event_sub, rcu); } - mutex_unlock(&ev_file->dev->devx_event_table.event_xa_lock); + mutex_unlock(&dev->devx_event_table.event_xa_lock); /* free the pending events allocation */ if (!ev_file->omit_data) { @@ -2670,7 +2671,7 @@ static int devx_async_event_close(struct inode *inode, struct file *filp) } uverbs_close_fd(filp); - put_device(&ev_file->dev->ib_dev.dev); + put_device(&dev->ib_dev.dev); return 0; } diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c index c2a5780cb394..0569bcab02d4 100644 --- a/drivers/infiniband/hw/mlx5/main.c +++ b/drivers/infiniband/hw/mlx5/main.c @@ -1023,7 +1023,7 @@ static int mlx5_ib_query_device(struct ib_device *ibdev, props->timestamp_mask = 0x7FFFFFFFFFFFFFFFULL; if (IS_ENABLED(CONFIG_INFINIBAND_ON_DEMAND_PAGING)) { - if (MLX5_CAP_GEN(mdev, pg)) + if (dev->odp_caps.general_caps & IB_ODP_SUPPORT) props->device_cap_flags |= IB_DEVICE_ON_DEMAND_PAGING; props->odp_caps = dev->odp_caps; } @@ -5802,13 +5802,12 @@ static void mlx5_ib_unbind_slave_port(struct mlx5_ib_dev *ibdev, return; } - if (mpi->mdev_events.notifier_call) - mlx5_notifier_unregister(mpi->mdev, &mpi->mdev_events); - mpi->mdev_events.notifier_call = NULL; - mpi->ibdev = NULL; spin_unlock(&port->mp.mpi_lock); + if (mpi->mdev_events.notifier_call) + mlx5_notifier_unregister(mpi->mdev, &mpi->mdev_events); + mpi->mdev_events.notifier_call = NULL; mlx5_remove_netdev_notifier(ibdev, port_num); spin_lock(&port->mp.mpi_lock); @@ -6140,6 +6139,8 @@ static int mlx5_ib_stage_init_init(struct mlx5_ib_dev *dev) dev->port[i].roce.last_port_state = IB_PORT_DOWN; } + mlx5_ib_internal_fill_odp_caps(dev); + err = mlx5_ib_init_multiport_master(dev); if (err) return err; @@ -6564,8 +6565,6 @@ static void mlx5_ib_stage_dev_res_cleanup(struct mlx5_ib_dev *dev) 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); } diff --git a/drivers/infiniband/hw/mlx5/mem.c b/drivers/infiniband/hw/mlx5/mem.c index fe1a76d8531c..a40e0abf2338 100644 --- a/drivers/infiniband/hw/mlx5/mem.c +++ b/drivers/infiniband/hw/mlx5/mem.c @@ -57,9 +57,10 @@ void mlx5_ib_cont_pages(struct ib_umem *umem, u64 addr, int entry; if (umem->is_odp) { - unsigned int page_shift = to_ib_umem_odp(umem)->page_shift; + struct ib_umem_odp *odp = to_ib_umem_odp(umem); + unsigned int page_shift = odp->page_shift; - *ncont = ib_umem_page_count(umem); + *ncont = ib_umem_odp_num_pages(odp); *count = *ncont << (page_shift - PAGE_SHIFT); *shift = page_shift; if (order) diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h index c482f19958b3..9ae587b74b12 100644 --- a/drivers/infiniband/hw/mlx5/mlx5_ib.h +++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h @@ -481,6 +481,7 @@ struct mlx5_umr_wr { u64 length; int access_flags; u32 mkey; + u8 ignore_free_state:1; }; static inline const struct mlx5_umr_wr *umr_wr(const struct ib_send_wr *wr) @@ -1474,4 +1475,18 @@ int bfregn_to_uar_index(struct mlx5_ib_dev *dev, bool dyn_bfreg); int mlx5_ib_qp_set_counter(struct ib_qp *qp, struct rdma_counter *counter); + +static inline bool mlx5_ib_can_use_umr(struct mlx5_ib_dev *dev, + bool do_modify_atomic) +{ + if (MLX5_CAP_GEN(dev->mdev, umr_modify_entity_size_disabled)) + return false; + + if (do_modify_atomic && + MLX5_CAP_GEN(dev->mdev, atomic) && + MLX5_CAP_GEN(dev->mdev, umr_modify_atomic_disabled)) + return false; + + return true; +} #endif /* MLX5_IB_H */ diff --git a/drivers/infiniband/hw/mlx5/mr.c b/drivers/infiniband/hw/mlx5/mr.c index 20ece6e0b2fc..3401f5f6792e 100644 --- a/drivers/infiniband/hw/mlx5/mr.c +++ b/drivers/infiniband/hw/mlx5/mr.c @@ -51,22 +51,12 @@ static void clean_mr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr); static void dereg_mr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr); static int mr_cache_max_order(struct mlx5_ib_dev *dev); static int unreg_umr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr); -static bool umr_can_modify_entity_size(struct mlx5_ib_dev *dev) -{ - return !MLX5_CAP_GEN(dev->mdev, umr_modify_entity_size_disabled); -} static bool umr_can_use_indirect_mkey(struct mlx5_ib_dev *dev) { return !MLX5_CAP_GEN(dev->mdev, umr_indirect_mkey_disabled); } -static bool use_umr(struct mlx5_ib_dev *dev, int order) -{ - return order <= mr_cache_max_order(dev) && - umr_can_modify_entity_size(dev); -} - static int destroy_mkey(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr) { int err = mlx5_core_destroy_mkey(dev->mdev, &mr->mmkey); @@ -545,13 +535,16 @@ void mlx5_mr_cache_free(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr) return; c = order2idx(dev, mr->order); - if (c < 0 || c >= MAX_MR_CACHE_ENTRIES) { - mlx5_ib_warn(dev, "order %d, cache index %d\n", mr->order, c); - return; - } + WARN_ON(c < 0 || c >= MAX_MR_CACHE_ENTRIES); - if (unreg_umr(dev, mr)) + if (unreg_umr(dev, mr)) { + mr->allocated_from_cache = false; + destroy_mkey(dev, mr); + ent = &cache->ent[c]; + if (ent->cur < ent->limit) + queue_work(cache->wq, &ent->work); return; + } ent = &cache->ent[c]; spin_lock_irq(&ent->lock); @@ -1268,7 +1261,7 @@ struct ib_mr *mlx5_ib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, { struct mlx5_ib_dev *dev = to_mdev(pd->device); struct mlx5_ib_mr *mr = NULL; - bool populate_mtts = false; + bool use_umr; struct ib_umem *umem; int page_shift; int npages; @@ -1300,29 +1293,28 @@ struct ib_mr *mlx5_ib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, if (err < 0) return ERR_PTR(err); - if (use_umr(dev, order)) { + use_umr = mlx5_ib_can_use_umr(dev, true); + + if (order <= mr_cache_max_order(dev) && use_umr) { mr = alloc_mr_from_cache(pd, umem, virt_addr, length, ncont, page_shift, order, access_flags); if (PTR_ERR(mr) == -EAGAIN) { mlx5_ib_dbg(dev, "cache empty for order %d\n", order); mr = NULL; } - populate_mtts = false; } else if (!MLX5_CAP_GEN(dev->mdev, umr_extended_translation_offset)) { if (access_flags & IB_ACCESS_ON_DEMAND) { err = -EINVAL; pr_err("Got MR registration for ODP MR > 512MB, not supported for Connect-IB\n"); goto error; } - populate_mtts = true; + use_umr = false; } if (!mr) { - if (!umr_can_modify_entity_size(dev)) - populate_mtts = true; mutex_lock(&dev->slow_path_mutex); mr = reg_create(NULL, pd, virt_addr, length, umem, ncont, - page_shift, access_flags, populate_mtts); + page_shift, access_flags, !use_umr); mutex_unlock(&dev->slow_path_mutex); } @@ -1338,7 +1330,7 @@ struct ib_mr *mlx5_ib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, update_odp_mr(mr); - if (!populate_mtts) { + if (use_umr) { int update_xlt_flags = MLX5_IB_UPD_XLT_ENABLE; if (access_flags & IB_ACCESS_ON_DEMAND) @@ -1373,9 +1365,11 @@ static int unreg_umr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr) return 0; umrwr.wr.send_flags = MLX5_IB_SEND_UMR_DISABLE_MR | - MLX5_IB_SEND_UMR_FAIL_IF_FREE; + MLX5_IB_SEND_UMR_UPDATE_PD_ACCESS; umrwr.wr.opcode = MLX5_IB_WR_UMR; + umrwr.pd = dev->umrc.pd; umrwr.mkey = mr->mmkey.key; + umrwr.ignore_free_state = 1; return mlx5_ib_post_send_wait(dev, &umrwr); } @@ -1452,7 +1446,8 @@ int mlx5_ib_rereg_user_mr(struct ib_mr *ib_mr, int flags, u64 start, goto err; } - if (flags & IB_MR_REREG_TRANS && !use_umr_mtt_update(mr, addr, len)) { + if (!mlx5_ib_can_use_umr(dev, true) || + (flags & IB_MR_REREG_TRANS && !use_umr_mtt_update(mr, addr, len))) { /* * UMR can't be used - MKey needs to be replaced. */ @@ -1577,10 +1572,10 @@ static void clean_mr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr) mr->sig = NULL; } - mlx5_free_priv_descs(mr); - - if (!allocated_from_cache) + if (!allocated_from_cache) { destroy_mkey(dev, mr); + mlx5_free_priv_descs(mr); + } } static void dereg_mr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr) diff --git a/drivers/infiniband/hw/mlx5/odp.c b/drivers/infiniband/hw/mlx5/odp.c index 5b642d81e617..0a59912a4cef 100644 --- a/drivers/infiniband/hw/mlx5/odp.c +++ b/drivers/infiniband/hw/mlx5/odp.c @@ -246,7 +246,7 @@ void mlx5_ib_invalidate_range(struct ib_umem_odp *umem_odp, unsigned long start, * overwrite the same MTTs. Concurent invalidations might race us, * but they will write 0s as well, so no difference in the end result. */ - + mutex_lock(&umem_odp->umem_mutex); for (addr = start; addr < end; addr += BIT(umem_odp->page_shift)) { idx = (addr - ib_umem_start(umem_odp)) >> umem_odp->page_shift; /* @@ -278,6 +278,7 @@ void mlx5_ib_invalidate_range(struct ib_umem_odp *umem_odp, unsigned long start, idx - blk_start_idx + 1, 0, MLX5_IB_UPD_XLT_ZAP | MLX5_IB_UPD_XLT_ATOMIC); + mutex_unlock(&umem_odp->umem_mutex); /* * We are now sure that the device will not access the * memory. We can safely unmap it, and mark it as dirty if @@ -300,7 +301,8 @@ void mlx5_ib_internal_fill_odp_caps(struct mlx5_ib_dev *dev) memset(caps, 0, sizeof(*caps)); - if (!MLX5_CAP_GEN(dev->mdev, pg)) + if (!MLX5_CAP_GEN(dev->mdev, pg) || + !mlx5_ib_can_use_umr(dev, true)) return; caps->general_caps = IB_ODP_SUPPORT; @@ -354,7 +356,8 @@ void mlx5_ib_internal_fill_odp_caps(struct mlx5_ib_dev *dev) if (MLX5_CAP_GEN(dev->mdev, fixed_buffer_size) && MLX5_CAP_GEN(dev->mdev, null_mkey) && - MLX5_CAP_GEN(dev->mdev, umr_extended_translation_offset)) + MLX5_CAP_GEN(dev->mdev, umr_extended_translation_offset) && + !MLX5_CAP_GEN(dev->mdev, umr_indirect_mkey_disabled)) caps->general_caps |= IB_ODP_SUPPORT_IMPLICIT; return; @@ -578,7 +581,6 @@ static int pagefault_mr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr, u32 flags) { int npages = 0, current_seq, page_shift, ret, np; - bool implicit = false; struct ib_umem_odp *odp_mr = to_ib_umem_odp(mr->umem); bool downgrade = flags & MLX5_PF_FLAGS_DOWNGRADE; bool prefetch = flags & MLX5_PF_FLAGS_PREFETCH; @@ -593,7 +595,6 @@ static int pagefault_mr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr, if (IS_ERR(odp)) return PTR_ERR(odp); mr = odp->private; - implicit = true; } else { odp = odp_mr; } @@ -681,19 +682,15 @@ next_mr: out: if (ret == -EAGAIN) { - if (implicit || !odp->dying) { - unsigned long timeout = - msecs_to_jiffies(MMU_NOTIFIER_TIMEOUT); - - if (!wait_for_completion_timeout( - &odp->notifier_completion, - timeout)) { - mlx5_ib_warn(dev, "timeout waiting for mmu notifier. seq %d against %d. notifiers_count=%d\n", - current_seq, odp->notifiers_seq, odp->notifiers_count); - } - } else { - /* The MR is being killed, kill the QP as well. */ - ret = -EFAULT; + unsigned long timeout = msecs_to_jiffies(MMU_NOTIFIER_TIMEOUT); + + if (!wait_for_completion_timeout(&odp->notifier_completion, + timeout)) { + mlx5_ib_warn( + dev, + "timeout waiting for mmu notifier. seq %d against %d. notifiers_count=%d\n", + current_seq, odp->notifiers_seq, + odp->notifiers_count); } } @@ -1627,8 +1624,10 @@ int mlx5_ib_odp_init_one(struct mlx5_ib_dev *dev) { int ret = 0; - if (dev->odp_caps.general_caps & IB_ODP_SUPPORT) - ib_set_device_ops(&dev->ib_dev, &mlx5_ib_dev_odp_ops); + if (!(dev->odp_caps.general_caps & IB_ODP_SUPPORT)) + return ret; + + ib_set_device_ops(&dev->ib_dev, &mlx5_ib_dev_odp_ops); if (dev->odp_caps.general_caps & IB_ODP_SUPPORT_IMPLICIT) { ret = mlx5_cmd_null_mkey(dev->mdev, &dev->null_mkey); @@ -1638,9 +1637,6 @@ int mlx5_ib_odp_init_one(struct mlx5_ib_dev *dev) } } - if (!MLX5_CAP_GEN(dev->mdev, pg)) - return ret; - ret = mlx5_ib_create_pf_eq(dev, &dev->odp_pf_eq); return ret; @@ -1648,7 +1644,7 @@ int mlx5_ib_odp_init_one(struct mlx5_ib_dev *dev) void mlx5_ib_odp_cleanup_one(struct mlx5_ib_dev *dev) { - if (!MLX5_CAP_GEN(dev->mdev, pg)) + if (!(dev->odp_caps.general_caps & IB_ODP_SUPPORT)) return; mlx5_ib_destroy_pf_eq(dev, &dev->odp_pf_eq); @@ -1771,7 +1767,7 @@ static void mlx5_ib_prefetch_mr_work(struct work_struct *work) num_pending_prefetch_dec(to_mdev(w->pd->device), w->sg_list, w->num_sge, 0); - kfree(w); + kvfree(w); } int mlx5_ib_advise_mr_prefetch(struct ib_pd *pd, @@ -1813,7 +1809,7 @@ int mlx5_ib_advise_mr_prefetch(struct ib_pd *pd, if (valid_req) queue_work(system_unbound_wq, &work->work); else - kfree(work); + kvfree(work); srcu_read_unlock(&dev->mr_srcu, srcu_key); diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c index 2a97619ed603..72869ff4a334 100644 --- a/drivers/infiniband/hw/mlx5/qp.c +++ b/drivers/infiniband/hw/mlx5/qp.c @@ -1713,7 +1713,6 @@ static int create_rss_raw_qp_tir(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp, } MLX5_SET(tirc, tirc, rx_hash_fn, MLX5_RX_HASH_FN_TOEPLITZ); - MLX5_SET(tirc, tirc, rx_hash_symmetric, 1); memcpy(rss_key, ucmd.rx_hash_key, len); break; } @@ -4163,7 +4162,7 @@ static u64 get_xlt_octo(u64 bytes) MLX5_IB_UMR_OCTOWORD; } -static __be64 frwr_mkey_mask(void) +static __be64 frwr_mkey_mask(bool atomic) { u64 result; @@ -4176,10 +4175,12 @@ static __be64 frwr_mkey_mask(void) MLX5_MKEY_MASK_LW | MLX5_MKEY_MASK_RR | MLX5_MKEY_MASK_RW | - MLX5_MKEY_MASK_A | MLX5_MKEY_MASK_SMALL_FENCE | MLX5_MKEY_MASK_FREE; + if (atomic) + result |= MLX5_MKEY_MASK_A; + return cpu_to_be64(result); } @@ -4205,7 +4206,7 @@ static __be64 sig_mkey_mask(void) } static void set_reg_umr_seg(struct mlx5_wqe_umr_ctrl_seg *umr, - struct mlx5_ib_mr *mr, u8 flags) + struct mlx5_ib_mr *mr, u8 flags, bool atomic) { int size = (mr->ndescs + mr->meta_ndescs) * mr->desc_size; @@ -4213,7 +4214,7 @@ static void set_reg_umr_seg(struct mlx5_wqe_umr_ctrl_seg *umr, umr->flags = flags; umr->xlt_octowords = cpu_to_be16(get_xlt_octo(size)); - umr->mkey_mask = frwr_mkey_mask(); + umr->mkey_mask = frwr_mkey_mask(atomic); } static void set_linv_umr_seg(struct mlx5_wqe_umr_ctrl_seg *umr) @@ -4295,10 +4296,14 @@ static int set_reg_umr_segment(struct mlx5_ib_dev *dev, memset(umr, 0, sizeof(*umr)); - if (wr->send_flags & MLX5_IB_SEND_UMR_FAIL_IF_FREE) - umr->flags = MLX5_UMR_CHECK_FREE; /* fail if free */ - else - umr->flags = MLX5_UMR_CHECK_NOT_FREE; /* fail if not free */ + if (!umrwr->ignore_free_state) { + if (wr->send_flags & MLX5_IB_SEND_UMR_FAIL_IF_FREE) + /* fail if free */ + umr->flags = MLX5_UMR_CHECK_FREE; + else + /* fail if not free */ + umr->flags = MLX5_UMR_CHECK_NOT_FREE; + } umr->xlt_octowords = cpu_to_be16(get_xlt_octo(umrwr->xlt_size)); if (wr->send_flags & MLX5_IB_SEND_UMR_UPDATE_XLT) { @@ -4808,10 +4813,22 @@ static int set_reg_wr(struct mlx5_ib_qp *qp, { struct mlx5_ib_mr *mr = to_mmr(wr->mr); struct mlx5_ib_pd *pd = to_mpd(qp->ibqp.pd); + struct mlx5_ib_dev *dev = to_mdev(pd->ibpd.device); int mr_list_size = (mr->ndescs + mr->meta_ndescs) * mr->desc_size; bool umr_inline = mr_list_size <= MLX5_IB_SQ_UMR_INLINE_THRESHOLD; + bool atomic = wr->access & IB_ACCESS_REMOTE_ATOMIC; u8 flags = 0; + if (!mlx5_ib_can_use_umr(dev, atomic)) { + mlx5_ib_warn(to_mdev(qp->ibqp.device), + "Fast update of %s for MR is disabled\n", + (MLX5_CAP_GEN(dev->mdev, + umr_modify_entity_size_disabled)) ? + "entity size" : + "atomic access"); + return -EINVAL; + } + if (unlikely(wr->wr.send_flags & IB_SEND_INLINE)) { mlx5_ib_warn(to_mdev(qp->ibqp.device), "Invalid IB_SEND_INLINE send flag\n"); @@ -4823,7 +4840,7 @@ static int set_reg_wr(struct mlx5_ib_qp *qp, if (umr_inline) flags |= MLX5_UMR_INLINE; - set_reg_umr_seg(*seg, mr, flags); + set_reg_umr_seg(*seg, mr, flags, atomic); *seg += sizeof(struct mlx5_wqe_umr_ctrl_seg); *size += sizeof(struct mlx5_wqe_umr_ctrl_seg) / 16; handle_post_send_edge(&qp->sq, seg, *size, cur_edge); diff --git a/drivers/infiniband/hw/qedr/main.c b/drivers/infiniband/hw/qedr/main.c index 533157a2a3be..f97b3d65b30c 100644 --- a/drivers/infiniband/hw/qedr/main.c +++ b/drivers/infiniband/hw/qedr/main.c @@ -125,14 +125,20 @@ static ssize_t hw_rev_show(struct device *device, struct device_attribute *attr, struct qedr_dev *dev = rdma_device_to_drv_device(device, struct qedr_dev, ibdev); - return scnprintf(buf, PAGE_SIZE, "0x%x\n", dev->pdev->vendor); + return scnprintf(buf, PAGE_SIZE, "0x%x\n", dev->attr.hw_ver); } static DEVICE_ATTR_RO(hw_rev); static ssize_t hca_type_show(struct device *device, struct device_attribute *attr, char *buf) { - return scnprintf(buf, PAGE_SIZE, "%s\n", "HCA_TYPE_TO_SET"); + struct qedr_dev *dev = + rdma_device_to_drv_device(device, struct qedr_dev, ibdev); + + return scnprintf(buf, PAGE_SIZE, "FastLinQ QL%x %s\n", + dev->pdev->device, + rdma_protocol_iwarp(&dev->ibdev, 1) ? + "iWARP" : "RoCE"); } static DEVICE_ATTR_RO(hca_type); diff --git a/drivers/infiniband/sw/siw/Kconfig b/drivers/infiniband/sw/siw/Kconfig index dace276aea14..b622fc62f2cd 100644 --- a/drivers/infiniband/sw/siw/Kconfig +++ b/drivers/infiniband/sw/siw/Kconfig @@ -1,6 +1,6 @@ config RDMA_SIW tristate "Software RDMA over TCP/IP (iWARP) driver" - depends on INET && INFINIBAND && LIBCRC32C && 64BIT + depends on INET && INFINIBAND && LIBCRC32C select DMA_VIRT_OPS help This driver implements the iWARP RDMA transport over diff --git a/drivers/infiniband/sw/siw/siw.h b/drivers/infiniband/sw/siw/siw.h index 03fd7b2f595f..dba4535494ab 100644 --- a/drivers/infiniband/sw/siw/siw.h +++ b/drivers/infiniband/sw/siw/siw.h @@ -138,9 +138,9 @@ struct siw_umem { }; struct siw_pble { - u64 addr; /* Address of assigned user buffer */ - u64 size; /* Size of this entry */ - u64 pbl_off; /* Total offset from start of PBL */ + dma_addr_t addr; /* Address of assigned buffer */ + unsigned int size; /* Size of this entry */ + unsigned long pbl_off; /* Total offset from start of PBL */ }; struct siw_pbl { @@ -214,7 +214,7 @@ struct siw_wqe { struct siw_cq { struct ib_cq base_cq; spinlock_t lock; - u64 *notify; + struct siw_cq_ctrl *notify; struct siw_cqe *queue; u32 cq_put; u32 cq_get; @@ -734,7 +734,7 @@ static inline void siw_crc_skb(struct siw_rx_stream *srx, unsigned int len) "MEM[0x%08x] %s: " fmt, mem->stag, __func__, ##__VA_ARGS__) #define siw_dbg_cep(cep, fmt, ...) \ - ibdev_dbg(&cep->sdev->base_dev, "CEP[0x%p] %s: " fmt, \ + ibdev_dbg(&cep->sdev->base_dev, "CEP[0x%pK] %s: " fmt, \ cep, __func__, ##__VA_ARGS__) void siw_cq_flush(struct siw_cq *cq); diff --git a/drivers/infiniband/sw/siw/siw_cm.c b/drivers/infiniband/sw/siw/siw_cm.c index a7cde98e73e8..1db5ad3d9580 100644 --- a/drivers/infiniband/sw/siw/siw_cm.c +++ b/drivers/infiniband/sw/siw/siw_cm.c @@ -220,13 +220,12 @@ static void siw_put_work(struct siw_cm_work *work) static void siw_cep_set_inuse(struct siw_cep *cep) { unsigned long flags; - int rv; retry: spin_lock_irqsave(&cep->lock, flags); if (cep->in_use) { spin_unlock_irqrestore(&cep->lock, flags); - rv = wait_event_interruptible(cep->waitq, !cep->in_use); + wait_event_interruptible(cep->waitq, !cep->in_use); if (signal_pending(current)) flush_signals(current); goto retry; @@ -356,8 +355,8 @@ static int siw_cm_upcall(struct siw_cep *cep, enum iw_cm_event_type reason, getname_local(cep->sock, &event.local_addr); getname_peer(cep->sock, &event.remote_addr); } - siw_dbg_cep(cep, "[QP %u]: id 0x%p, reason=%d, status=%d\n", - cep->qp ? qp_id(cep->qp) : -1, id, reason, status); + siw_dbg_cep(cep, "[QP %u]: reason=%d, status=%d\n", + cep->qp ? qp_id(cep->qp) : UINT_MAX, reason, status); return id->event_handler(id, &event); } @@ -948,8 +947,6 @@ static void siw_accept_newconn(struct siw_cep *cep) siw_cep_get(new_cep); new_s->sk->sk_user_data = new_cep; - siw_dbg_cep(cep, "listen socket 0x%p, new 0x%p\n", s, new_s); - if (siw_tcp_nagle == false) { int val = 1; @@ -1012,7 +1009,8 @@ static void siw_cm_work_handler(struct work_struct *w) cep = work->cep; siw_dbg_cep(cep, "[QP %u]: work type: %d, state %d\n", - cep->qp ? qp_id(cep->qp) : -1, work->type, cep->state); + cep->qp ? qp_id(cep->qp) : UINT_MAX, + work->type, cep->state); siw_cep_set_inuse(cep); @@ -1146,9 +1144,9 @@ static void siw_cm_work_handler(struct work_struct *w) } if (release_cep) { siw_dbg_cep(cep, - "release: timer=%s, QP[%u], id 0x%p\n", + "release: timer=%s, QP[%u]\n", cep->mpa_timer ? "y" : "n", - cep->qp ? qp_id(cep->qp) : -1, cep->cm_id); + cep->qp ? qp_id(cep->qp) : UINT_MAX); siw_cancel_mpatimer(cep); @@ -1212,8 +1210,8 @@ int siw_cm_queue_work(struct siw_cep *cep, enum siw_work_type type) else delay = MPAREP_TIMEOUT; } - siw_dbg_cep(cep, "[QP %u]: work type: %d, work 0x%p, timeout %lu\n", - cep->qp ? qp_id(cep->qp) : -1, type, work, delay); + siw_dbg_cep(cep, "[QP %u]: work type: %d, timeout %lu\n", + cep->qp ? qp_id(cep->qp) : -1, type, delay); queue_delayed_work(siw_cm_wq, &work->work, delay); @@ -1377,16 +1375,16 @@ int siw_connect(struct iw_cm_id *id, struct iw_cm_conn_param *params) } if (v4) siw_dbg_qp(qp, - "id 0x%p, pd_len %d, laddr %pI4 %d, raddr %pI4 %d\n", - id, pd_len, + "pd_len %d, laddr %pI4 %d, raddr %pI4 %d\n", + pd_len, &((struct sockaddr_in *)(laddr))->sin_addr, ntohs(((struct sockaddr_in *)(laddr))->sin_port), &((struct sockaddr_in *)(raddr))->sin_addr, ntohs(((struct sockaddr_in *)(raddr))->sin_port)); else siw_dbg_qp(qp, - "id 0x%p, pd_len %d, laddr %pI6 %d, raddr %pI6 %d\n", - id, pd_len, + "pd_len %d, laddr %pI6 %d, raddr %pI6 %d\n", + pd_len, &((struct sockaddr_in6 *)(laddr))->sin6_addr, ntohs(((struct sockaddr_in6 *)(laddr))->sin6_port), &((struct sockaddr_in6 *)(raddr))->sin6_addr, @@ -1509,14 +1507,13 @@ int siw_connect(struct iw_cm_id *id, struct iw_cm_conn_param *params) if (rv >= 0) { rv = siw_cm_queue_work(cep, SIW_CM_WORK_MPATIMEOUT); if (!rv) { - siw_dbg_cep(cep, "id 0x%p, [QP %u]: exit\n", id, - qp_id(qp)); + siw_dbg_cep(cep, "[QP %u]: exit\n", qp_id(qp)); siw_cep_set_free(cep); return 0; } } error: - siw_dbg_qp(qp, "failed: %d\n", rv); + siw_dbg(id->device, "failed: %d\n", rv); if (cep) { siw_socket_disassoc(s); @@ -1541,7 +1538,8 @@ error: } else if (s) { sock_release(s); } - siw_qp_put(qp); + if (qp) + siw_qp_put(qp); return rv; } @@ -1581,7 +1579,7 @@ int siw_accept(struct iw_cm_id *id, struct iw_cm_conn_param *params) siw_cancel_mpatimer(cep); if (cep->state != SIW_EPSTATE_RECVD_MPAREQ) { - siw_dbg_cep(cep, "id 0x%p: out of state\n", id); + siw_dbg_cep(cep, "out of state\n"); siw_cep_set_free(cep); siw_cep_put(cep); @@ -1602,7 +1600,7 @@ int siw_accept(struct iw_cm_id *id, struct iw_cm_conn_param *params) up_write(&qp->state_lock); goto error; } - siw_dbg_cep(cep, "id 0x%p\n", id); + siw_dbg_cep(cep, "[QP %d]\n", params->qpn); if (try_gso && cep->mpa.hdr.params.bits & MPA_RR_FLAG_GSO_EXP) { siw_dbg_cep(cep, "peer allows GSO on TX\n"); @@ -1612,8 +1610,8 @@ int siw_accept(struct iw_cm_id *id, struct iw_cm_conn_param *params) params->ird > sdev->attrs.max_ird) { siw_dbg_cep( cep, - "id 0x%p, [QP %u]: ord %d (max %d), ird %d (max %d)\n", - id, qp_id(qp), params->ord, sdev->attrs.max_ord, + "[QP %u]: ord %d (max %d), ird %d (max %d)\n", + qp_id(qp), params->ord, sdev->attrs.max_ord, params->ird, sdev->attrs.max_ird); rv = -EINVAL; up_write(&qp->state_lock); @@ -1625,8 +1623,8 @@ int siw_accept(struct iw_cm_id *id, struct iw_cm_conn_param *params) if (params->private_data_len > max_priv_data) { siw_dbg_cep( cep, - "id 0x%p, [QP %u]: private data length: %d (max %d)\n", - id, qp_id(qp), params->private_data_len, max_priv_data); + "[QP %u]: private data length: %d (max %d)\n", + qp_id(qp), params->private_data_len, max_priv_data); rv = -EINVAL; up_write(&qp->state_lock); goto error; @@ -1680,7 +1678,7 @@ int siw_accept(struct iw_cm_id *id, struct iw_cm_conn_param *params) qp_attrs.flags = SIW_MPA_CRC; qp_attrs.state = SIW_QP_STATE_RTS; - siw_dbg_cep(cep, "id 0x%p, [QP%u]: moving to rts\n", id, qp_id(qp)); + siw_dbg_cep(cep, "[QP%u]: moving to rts\n", qp_id(qp)); /* Associate QP with CEP */ siw_cep_get(cep); @@ -1701,8 +1699,8 @@ int siw_accept(struct iw_cm_id *id, struct iw_cm_conn_param *params) if (rv) goto error; - siw_dbg_cep(cep, "id 0x%p, [QP %u]: send mpa reply, %d byte pdata\n", - id, qp_id(qp), params->private_data_len); + siw_dbg_cep(cep, "[QP %u]: send mpa reply, %d byte pdata\n", + qp_id(qp), params->private_data_len); rv = siw_send_mpareqrep(cep, params->private_data, params->private_data_len); @@ -1760,14 +1758,14 @@ int siw_reject(struct iw_cm_id *id, const void *pdata, u8 pd_len) siw_cancel_mpatimer(cep); if (cep->state != SIW_EPSTATE_RECVD_MPAREQ) { - siw_dbg_cep(cep, "id 0x%p: out of state\n", id); + siw_dbg_cep(cep, "out of state\n"); siw_cep_set_free(cep); siw_cep_put(cep); /* put last reference */ return -ECONNRESET; } - siw_dbg_cep(cep, "id 0x%p, cep->state %d, pd_len %d\n", id, cep->state, + siw_dbg_cep(cep, "cep->state %d, pd_len %d\n", cep->state, pd_len); if (__mpa_rr_revision(cep->mpa.hdr.params.bits) >= MPA_REVISION_1) { @@ -1805,14 +1803,14 @@ static int siw_listen_address(struct iw_cm_id *id, int backlog, rv = kernel_setsockopt(s, SOL_SOCKET, SO_REUSEADDR, (char *)&s_val, sizeof(s_val)); if (rv) { - siw_dbg(id->device, "id 0x%p: setsockopt error: %d\n", id, rv); + siw_dbg(id->device, "setsockopt error: %d\n", rv); goto error; } rv = s->ops->bind(s, laddr, addr_family == AF_INET ? sizeof(struct sockaddr_in) : sizeof(struct sockaddr_in6)); if (rv) { - siw_dbg(id->device, "id 0x%p: socket bind error: %d\n", id, rv); + siw_dbg(id->device, "socket bind error: %d\n", rv); goto error; } cep = siw_cep_alloc(sdev); @@ -1825,13 +1823,13 @@ static int siw_listen_address(struct iw_cm_id *id, int backlog, rv = siw_cm_alloc_work(cep, backlog); if (rv) { siw_dbg(id->device, - "id 0x%p: alloc_work error %d, backlog %d\n", id, + "alloc_work error %d, backlog %d\n", rv, backlog); goto error; } rv = s->ops->listen(s, backlog); if (rv) { - siw_dbg(id->device, "id 0x%p: listen error %d\n", id, rv); + siw_dbg(id->device, "listen error %d\n", rv); goto error; } cep->cm_id = id; @@ -1915,8 +1913,7 @@ static void siw_drop_listeners(struct iw_cm_id *id) list_del(p); - siw_dbg_cep(cep, "id 0x%p: drop cep, state %d\n", id, - cep->state); + siw_dbg_cep(cep, "drop cep, state %d\n", cep->state); siw_cep_set_inuse(cep); @@ -1953,7 +1950,7 @@ int siw_create_listen(struct iw_cm_id *id, int backlog) struct net_device *dev = to_siw_dev(id->device)->netdev; int rv = 0, listeners = 0; - siw_dbg(id->device, "id 0x%p: backlog %d\n", id, backlog); + siw_dbg(id->device, "backlog %d\n", backlog); /* * For each attached address of the interface, create a @@ -1969,8 +1966,8 @@ int siw_create_listen(struct iw_cm_id *id, int backlog) s_raddr = (struct sockaddr_in *)&id->remote_addr; siw_dbg(id->device, - "id 0x%p: laddr %pI4:%d, raddr %pI4:%d\n", - id, &s_laddr.sin_addr, ntohs(s_laddr.sin_port), + "laddr %pI4:%d, raddr %pI4:%d\n", + &s_laddr.sin_addr, ntohs(s_laddr.sin_port), &s_raddr->sin_addr, ntohs(s_raddr->sin_port)); rtnl_lock(); @@ -1995,8 +1992,8 @@ int siw_create_listen(struct iw_cm_id *id, int backlog) *s_raddr = &to_sockaddr_in6(id->remote_addr); siw_dbg(id->device, - "id 0x%p: laddr %pI6:%d, raddr %pI6:%d\n", - id, &s_laddr->sin6_addr, ntohs(s_laddr->sin6_port), + "laddr %pI6:%d, raddr %pI6:%d\n", + &s_laddr->sin6_addr, ntohs(s_laddr->sin6_port), &s_raddr->sin6_addr, ntohs(s_raddr->sin6_port)); read_lock_bh(&in6_dev->lock); @@ -2029,17 +2026,15 @@ int siw_create_listen(struct iw_cm_id *id, int backlog) else if (!rv) rv = -EINVAL; - siw_dbg(id->device, "id 0x%p: %s\n", id, rv ? "FAIL" : "OK"); + siw_dbg(id->device, "%s\n", rv ? "FAIL" : "OK"); return rv; } int siw_destroy_listen(struct iw_cm_id *id) { - siw_dbg(id->device, "id 0x%p\n", id); - if (!id->provider_data) { - siw_dbg(id->device, "id 0x%p: no cep(s)\n", id); + siw_dbg(id->device, "no cep(s)\n"); return 0; } siw_drop_listeners(id); diff --git a/drivers/infiniband/sw/siw/siw_cq.c b/drivers/infiniband/sw/siw/siw_cq.c index e381ae9b7d62..d8db3bee9da7 100644 --- a/drivers/infiniband/sw/siw/siw_cq.c +++ b/drivers/infiniband/sw/siw/siw_cq.c @@ -71,9 +71,10 @@ int siw_reap_cqe(struct siw_cq *cq, struct ib_wc *wc) wc->wc_flags = IB_WC_WITH_INVALIDATE; } wc->qp = cqe->base_qp; - siw_dbg_cq(cq, "idx %u, type %d, flags %2x, id 0x%p\n", + siw_dbg_cq(cq, + "idx %u, type %d, flags %2x, id 0x%pK\n", cq->cq_get % cq->num_cqe, cqe->opcode, - cqe->flags, (void *)cqe->id); + cqe->flags, (void *)(uintptr_t)cqe->id); } WRITE_ONCE(cqe->flags, 0); cq->cq_get++; diff --git a/drivers/infiniband/sw/siw/siw_main.c b/drivers/infiniband/sw/siw/siw_main.c index f55c4e80aea4..05a92f997f60 100644 --- a/drivers/infiniband/sw/siw/siw_main.c +++ b/drivers/infiniband/sw/siw/siw_main.c @@ -160,10 +160,8 @@ static int siw_init_cpulist(void) out_err: siw_cpu_info.num_nodes = 0; - while (i) { + while (--i >= 0) kfree(siw_cpu_info.tx_valid_cpus[i]); - siw_cpu_info.tx_valid_cpus[i--] = NULL; - } kfree(siw_cpu_info.tx_valid_cpus); siw_cpu_info.tx_valid_cpus = NULL; @@ -612,6 +610,7 @@ static __init int siw_init_module(void) if (!siw_create_tx_threads()) { pr_info("siw: Could not start any TX thread\n"); + rv = -ENOMEM; goto out_error; } /* diff --git a/drivers/infiniband/sw/siw/siw_mem.c b/drivers/infiniband/sw/siw/siw_mem.c index 67171c82b0c4..87a56039f0ef 100644 --- a/drivers/infiniband/sw/siw/siw_mem.c +++ b/drivers/infiniband/sw/siw/siw_mem.c @@ -197,12 +197,12 @@ int siw_check_mem(struct ib_pd *pd, struct siw_mem *mem, u64 addr, */ if (addr < mem->va || addr + len > mem->va + mem->len) { siw_dbg_pd(pd, "MEM interval len %d\n", len); - siw_dbg_pd(pd, "[0x%016llx, 0x%016llx] out of bounds\n", - (unsigned long long)addr, - (unsigned long long)(addr + len)); - siw_dbg_pd(pd, "[0x%016llx, 0x%016llx] STag=0x%08x\n", - (unsigned long long)mem->va, - (unsigned long long)(mem->va + mem->len), + siw_dbg_pd(pd, "[0x%pK, 0x%pK] out of bounds\n", + (void *)(uintptr_t)addr, + (void *)(uintptr_t)(addr + len)); + siw_dbg_pd(pd, "[0x%pK, 0x%pK] STag=0x%08x\n", + (void *)(uintptr_t)mem->va, + (void *)(uintptr_t)(mem->va + mem->len), mem->stag); return -E_BASE_BOUNDS; @@ -330,7 +330,7 @@ out: * Optionally, provides remaining len within current element, and * current PBL index for later resume at same element. */ -u64 siw_pbl_get_buffer(struct siw_pbl *pbl, u64 off, int *len, int *idx) +dma_addr_t siw_pbl_get_buffer(struct siw_pbl *pbl, u64 off, int *len, int *idx) { int i = idx ? *idx : 0; diff --git a/drivers/infiniband/sw/siw/siw_mem.h b/drivers/infiniband/sw/siw/siw_mem.h index f43daf280891..db138c8423da 100644 --- a/drivers/infiniband/sw/siw/siw_mem.h +++ b/drivers/infiniband/sw/siw/siw_mem.h @@ -9,7 +9,7 @@ struct siw_umem *siw_umem_get(u64 start, u64 len, bool writable); void siw_umem_release(struct siw_umem *umem, bool dirty); struct siw_pbl *siw_pbl_alloc(u32 num_buf); -u64 siw_pbl_get_buffer(struct siw_pbl *pbl, u64 off, int *len, int *idx); +dma_addr_t siw_pbl_get_buffer(struct siw_pbl *pbl, u64 off, int *len, int *idx); struct siw_mem *siw_mem_id2obj(struct siw_device *sdev, int stag_index); int siw_mem_add(struct siw_device *sdev, struct siw_mem *m); int siw_invalidate_stag(struct ib_pd *pd, u32 stag); diff --git a/drivers/infiniband/sw/siw/siw_qp.c b/drivers/infiniband/sw/siw/siw_qp.c index 11383d9f95ef..430314c8abd9 100644 --- a/drivers/infiniband/sw/siw/siw_qp.c +++ b/drivers/infiniband/sw/siw/siw_qp.c @@ -220,12 +220,14 @@ static int siw_qp_enable_crc(struct siw_qp *qp) { struct siw_rx_stream *c_rx = &qp->rx_stream; struct siw_iwarp_tx *c_tx = &qp->tx_ctx; - int size = crypto_shash_descsize(siw_crypto_shash) + - sizeof(struct shash_desc); + int size; if (siw_crypto_shash == NULL) return -ENOENT; + size = crypto_shash_descsize(siw_crypto_shash) + + sizeof(struct shash_desc); + c_tx->mpa_crc_hd = kzalloc(size, GFP_KERNEL); c_rx->mpa_crc_hd = kzalloc(size, GFP_KERNEL); if (!c_tx->mpa_crc_hd || !c_rx->mpa_crc_hd) { @@ -947,7 +949,7 @@ skip_irq: rv = -EINVAL; goto out; } - wqe->sqe.sge[0].laddr = (u64)&wqe->sqe.sge[1]; + wqe->sqe.sge[0].laddr = (uintptr_t)&wqe->sqe.sge[1]; wqe->sqe.sge[0].lkey = 0; wqe->sqe.num_sge = 1; } @@ -1011,18 +1013,24 @@ out: */ static bool siw_cq_notify_now(struct siw_cq *cq, u32 flags) { - u64 cq_notify; + u32 cq_notify; if (!cq->base_cq.comp_handler) return false; - cq_notify = READ_ONCE(*cq->notify); + /* Read application shared notification state */ + cq_notify = READ_ONCE(cq->notify->flags); if ((cq_notify & SIW_NOTIFY_NEXT_COMPLETION) || ((cq_notify & SIW_NOTIFY_SOLICITED) && (flags & SIW_WQE_SOLICITED))) { - /* dis-arm CQ */ - smp_store_mb(*cq->notify, SIW_NOTIFY_NOT); + /* + * CQ notification is one-shot: Since the + * current CQE causes user notification, + * the CQ gets dis-aremd and must be re-aremd + * by the user for a new notification. + */ + WRITE_ONCE(cq->notify->flags, SIW_NOTIFY_NOT); return true; } diff --git a/drivers/infiniband/sw/siw/siw_qp_rx.c b/drivers/infiniband/sw/siw/siw_qp_rx.c index f87657a11657..c0a887240325 100644 --- a/drivers/infiniband/sw/siw/siw_qp_rx.c +++ b/drivers/infiniband/sw/siw/siw_qp_rx.c @@ -38,9 +38,10 @@ static int siw_rx_umem(struct siw_rx_stream *srx, struct siw_umem *umem, p = siw_get_upage(umem, dest_addr); if (unlikely(!p)) { - pr_warn("siw: %s: [QP %u]: bogus addr: %p, %p\n", + pr_warn("siw: %s: [QP %u]: bogus addr: %pK, %pK\n", __func__, qp_id(rx_qp(srx)), - (void *)dest_addr, (void *)umem->fp_addr); + (void *)(uintptr_t)dest_addr, + (void *)(uintptr_t)umem->fp_addr); /* siw internal error */ srx->skb_copied += copied; srx->skb_new -= copied; @@ -50,7 +51,7 @@ static int siw_rx_umem(struct siw_rx_stream *srx, struct siw_umem *umem, pg_off = dest_addr & ~PAGE_MASK; bytes = min(len, (int)PAGE_SIZE - pg_off); - siw_dbg_qp(rx_qp(srx), "page %p, bytes=%u\n", p, bytes); + siw_dbg_qp(rx_qp(srx), "page %pK, bytes=%u\n", p, bytes); dest = kmap_atomic(p); rv = skb_copy_bits(srx->skb, srx->skb_offset, dest + pg_off, @@ -104,11 +105,11 @@ static int siw_rx_kva(struct siw_rx_stream *srx, void *kva, int len) { int rv; - siw_dbg_qp(rx_qp(srx), "kva: 0x%p, len: %u\n", kva, len); + siw_dbg_qp(rx_qp(srx), "kva: 0x%pK, len: %u\n", kva, len); rv = skb_copy_bits(srx->skb, srx->skb_offset, kva, len); if (unlikely(rv)) { - pr_warn("siw: [QP %u]: %s, len %d, kva 0x%p, rv %d\n", + pr_warn("siw: [QP %u]: %s, len %d, kva 0x%pK, rv %d\n", qp_id(rx_qp(srx)), __func__, len, kva, rv); return rv; @@ -132,7 +133,7 @@ static int siw_rx_pbl(struct siw_rx_stream *srx, int *pbl_idx, while (len) { int bytes; - u64 buf_addr = + dma_addr_t buf_addr = siw_pbl_get_buffer(pbl, offset, &bytes, pbl_idx); if (!buf_addr) break; @@ -485,8 +486,8 @@ int siw_proc_send(struct siw_qp *qp) mem_p = *mem; if (mem_p->mem_obj == NULL) rv = siw_rx_kva(srx, - (void *)(sge->laddr + frx->sge_off), - sge_bytes); + (void *)(uintptr_t)(sge->laddr + frx->sge_off), + sge_bytes); else if (!mem_p->is_pbl) rv = siw_rx_umem(srx, mem_p->umem, sge->laddr + frx->sge_off, sge_bytes); @@ -598,8 +599,8 @@ int siw_proc_write(struct siw_qp *qp) if (mem->mem_obj == NULL) rv = siw_rx_kva(srx, - (void *)(srx->ddp_to + srx->fpdu_part_rcvd), - bytes); + (void *)(uintptr_t)(srx->ddp_to + srx->fpdu_part_rcvd), + bytes); else if (!mem->is_pbl) rv = siw_rx_umem(srx, mem->umem, srx->ddp_to + srx->fpdu_part_rcvd, bytes); @@ -841,8 +842,9 @@ int siw_proc_rresp(struct siw_qp *qp) bytes = min(srx->fpdu_part_rem, srx->skb_new); if (mem_p->mem_obj == NULL) - rv = siw_rx_kva(srx, (void *)(sge->laddr + wqe->processed), - bytes); + rv = siw_rx_kva(srx, + (void *)(uintptr_t)(sge->laddr + wqe->processed), + bytes); else if (!mem_p->is_pbl) rv = siw_rx_umem(srx, mem_p->umem, sge->laddr + wqe->processed, bytes); diff --git a/drivers/infiniband/sw/siw/siw_qp_tx.c b/drivers/infiniband/sw/siw/siw_qp_tx.c index 43020d2040fc..438a2917a47c 100644 --- a/drivers/infiniband/sw/siw/siw_qp_tx.c +++ b/drivers/infiniband/sw/siw/siw_qp_tx.c @@ -26,7 +26,7 @@ static struct page *siw_get_pblpage(struct siw_mem *mem, u64 addr, int *idx) { struct siw_pbl *pbl = mem->pbl; u64 offset = addr - mem->va; - u64 paddr = siw_pbl_get_buffer(pbl, offset, NULL, idx); + dma_addr_t paddr = siw_pbl_get_buffer(pbl, offset, NULL, idx); if (paddr) return virt_to_page(paddr); @@ -37,7 +37,7 @@ static struct page *siw_get_pblpage(struct siw_mem *mem, u64 addr, int *idx) /* * Copy short payload at provided destination payload address */ -static int siw_try_1seg(struct siw_iwarp_tx *c_tx, u64 paddr) +static int siw_try_1seg(struct siw_iwarp_tx *c_tx, void *paddr) { struct siw_wqe *wqe = &c_tx->wqe_active; struct siw_sge *sge = &wqe->sqe.sge[0]; @@ -50,16 +50,16 @@ static int siw_try_1seg(struct siw_iwarp_tx *c_tx, u64 paddr) return 0; if (tx_flags(wqe) & SIW_WQE_INLINE) { - memcpy((void *)paddr, &wqe->sqe.sge[1], bytes); + memcpy(paddr, &wqe->sqe.sge[1], bytes); } else { struct siw_mem *mem = wqe->mem[0]; if (!mem->mem_obj) { /* Kernel client using kva */ - memcpy((void *)paddr, (void *)sge->laddr, bytes); + memcpy(paddr, + (const void *)(uintptr_t)sge->laddr, bytes); } else if (c_tx->in_syscall) { - if (copy_from_user((void *)paddr, - (const void __user *)sge->laddr, + if (copy_from_user(paddr, u64_to_user_ptr(sge->laddr), bytes)) return -EFAULT; } else { @@ -79,12 +79,12 @@ static int siw_try_1seg(struct siw_iwarp_tx *c_tx, u64 paddr) buffer = kmap_atomic(p); if (likely(PAGE_SIZE - off >= bytes)) { - memcpy((void *)paddr, buffer + off, bytes); + memcpy(paddr, buffer + off, bytes); kunmap_atomic(buffer); } else { unsigned long part = bytes - (PAGE_SIZE - off); - memcpy((void *)paddr, buffer + off, part); + memcpy(paddr, buffer + off, part); kunmap_atomic(buffer); if (!mem->is_pbl) @@ -98,7 +98,7 @@ static int siw_try_1seg(struct siw_iwarp_tx *c_tx, u64 paddr) return -EFAULT; buffer = kmap_atomic(p); - memcpy((void *)(paddr + part), buffer, + memcpy(paddr + part, buffer, bytes - part); kunmap_atomic(buffer); } @@ -166,7 +166,7 @@ static int siw_qp_prepare_tx(struct siw_iwarp_tx *c_tx) c_tx->ctrl_len = sizeof(struct iwarp_send); crc = (char *)&c_tx->pkt.send_pkt.crc; - data = siw_try_1seg(c_tx, (u64)crc); + data = siw_try_1seg(c_tx, crc); break; case SIW_OP_SEND_REMOTE_INV: @@ -189,7 +189,7 @@ static int siw_qp_prepare_tx(struct siw_iwarp_tx *c_tx) c_tx->ctrl_len = sizeof(struct iwarp_send_inv); crc = (char *)&c_tx->pkt.send_pkt.crc; - data = siw_try_1seg(c_tx, (u64)crc); + data = siw_try_1seg(c_tx, crc); break; case SIW_OP_WRITE: @@ -201,7 +201,7 @@ static int siw_qp_prepare_tx(struct siw_iwarp_tx *c_tx) c_tx->ctrl_len = sizeof(struct iwarp_rdma_write); crc = (char *)&c_tx->pkt.write_pkt.crc; - data = siw_try_1seg(c_tx, (u64)crc); + data = siw_try_1seg(c_tx, crc); break; case SIW_OP_READ_RESPONSE: @@ -216,7 +216,7 @@ static int siw_qp_prepare_tx(struct siw_iwarp_tx *c_tx) c_tx->ctrl_len = sizeof(struct iwarp_rdma_rresp); crc = (char *)&c_tx->pkt.write_pkt.crc; - data = siw_try_1seg(c_tx, (u64)crc); + data = siw_try_1seg(c_tx, crc); break; default: @@ -398,15 +398,13 @@ static int siw_0copy_tx(struct socket *s, struct page **page, #define MAX_TRAILER (MPA_CRC_SIZE + 4) -static void siw_unmap_pages(struct page **pages, int hdr_len, int num_maps) +static void siw_unmap_pages(struct page **pp, unsigned long kmap_mask) { - if (hdr_len) { - ++pages; - --num_maps; - } - while (num_maps-- > 0) { - kunmap(*pages); - pages++; + while (kmap_mask) { + if (kmap_mask & BIT(0)) + kunmap(*pp); + pp++; + kmap_mask >>= 1; } } @@ -437,6 +435,7 @@ static int siw_tx_hdt(struct siw_iwarp_tx *c_tx, struct socket *s) unsigned int data_len = c_tx->bytes_unsent, hdr_len = 0, trl_len = 0, sge_off = c_tx->sge_off, sge_idx = c_tx->sge_idx, pbl_idx = c_tx->pbl_idx; + unsigned long kmap_mask = 0L; if (c_tx->state == SIW_SEND_HDR) { if (c_tx->use_sendpage) { @@ -463,8 +462,7 @@ static int siw_tx_hdt(struct siw_iwarp_tx *c_tx, struct socket *s) if (!(tx_flags(wqe) & SIW_WQE_INLINE)) { mem = wqe->mem[sge_idx]; - if (!mem->mem_obj) - is_kva = 1; + is_kva = mem->mem_obj == NULL ? 1 : 0; } else { is_kva = 1; } @@ -473,7 +471,8 @@ static int siw_tx_hdt(struct siw_iwarp_tx *c_tx, struct socket *s) * tx from kernel virtual address: either inline data * or memory region with assigned kernel buffer */ - iov[seg].iov_base = (void *)(sge->laddr + sge_off); + iov[seg].iov_base = + (void *)(uintptr_t)(sge->laddr + sge_off); iov[seg].iov_len = sge_len; if (do_crc) @@ -500,12 +499,7 @@ static int siw_tx_hdt(struct siw_iwarp_tx *c_tx, struct socket *s) p = siw_get_upage(mem->umem, sge->laddr + sge_off); if (unlikely(!p)) { - if (hdr_len) - seg--; - if (!c_tx->use_sendpage && seg) { - siw_unmap_pages(page_array, - hdr_len, seg); - } + siw_unmap_pages(page_array, kmap_mask); wqe->processed -= c_tx->bytes_unsent; rv = -EFAULT; goto done_crc; @@ -515,6 +509,10 @@ static int siw_tx_hdt(struct siw_iwarp_tx *c_tx, struct socket *s) if (!c_tx->use_sendpage) { iov[seg].iov_base = kmap(p) + fp_off; iov[seg].iov_len = plen; + + /* Remember for later kunmap() */ + kmap_mask |= BIT(seg); + if (do_crc) crypto_shash_update( c_tx->mpa_crc_hd, @@ -526,13 +524,13 @@ static int siw_tx_hdt(struct siw_iwarp_tx *c_tx, struct socket *s) page_address(p) + fp_off, plen); } else { - u64 pa = ((sge->laddr + sge_off) & PAGE_MASK); + u64 va = sge->laddr + sge_off; - page_array[seg] = virt_to_page(pa); + page_array[seg] = virt_to_page(va & PAGE_MASK); if (do_crc) crypto_shash_update( c_tx->mpa_crc_hd, - (void *)(sge->laddr + sge_off), + (void *)(uintptr_t)va, plen); } @@ -543,10 +541,7 @@ static int siw_tx_hdt(struct siw_iwarp_tx *c_tx, struct socket *s) if (++seg > (int)MAX_ARRAY) { siw_dbg_qp(tx_qp(c_tx), "to many fragments\n"); - if (!is_kva && !c_tx->use_sendpage) { - siw_unmap_pages(page_array, hdr_len, - seg - 1); - } + siw_unmap_pages(page_array, kmap_mask); wqe->processed -= c_tx->bytes_unsent; rv = -EMSGSIZE; goto done_crc; @@ -597,8 +592,7 @@ sge_done: } else { rv = kernel_sendmsg(s, &msg, iov, seg + 1, hdr_len + data_len + trl_len); - if (!is_kva) - siw_unmap_pages(page_array, hdr_len, seg); + siw_unmap_pages(page_array, kmap_mask); } if (rv < (int)hdr_len) { /* Not even complete hdr pushed or negative rv */ @@ -829,7 +823,8 @@ static int siw_qp_sq_proc_tx(struct siw_qp *qp, struct siw_wqe *wqe) rv = -EINVAL; goto tx_error; } - wqe->sqe.sge[0].laddr = (u64)&wqe->sqe.sge[1]; + wqe->sqe.sge[0].laddr = + (u64)(uintptr_t)&wqe->sqe.sge[1]; } } wqe->wr_status = SIW_WR_INPROGRESS; @@ -924,7 +919,7 @@ tx_error: static int siw_fastreg_mr(struct ib_pd *pd, struct siw_sqe *sqe) { - struct ib_mr *base_mr = (struct ib_mr *)sqe->base_mr; + struct ib_mr *base_mr = (struct ib_mr *)(uintptr_t)sqe->base_mr; struct siw_device *sdev = to_siw_dev(pd->device); struct siw_mem *mem = siw_mem_id2obj(sdev, sqe->rkey >> 8); int rv = 0; @@ -954,8 +949,7 @@ static int siw_fastreg_mr(struct ib_pd *pd, struct siw_sqe *sqe) mem->stag = sqe->rkey; mem->perms = sqe->access; - siw_dbg_mem(mem, "STag now valid, MR va: 0x%016llx -> 0x%016llx\n", - mem->va, base_mr->iova); + siw_dbg_mem(mem, "STag 0x%08x now valid\n", sqe->rkey); mem->va = base_mr->iova; mem->stag_valid = 1; out: diff --git a/drivers/infiniband/sw/siw/siw_verbs.c b/drivers/infiniband/sw/siw/siw_verbs.c index 32dc79d0e898..da52c90e06d4 100644 --- a/drivers/infiniband/sw/siw/siw_verbs.c +++ b/drivers/infiniband/sw/siw/siw_verbs.c @@ -424,8 +424,7 @@ struct ib_qp *siw_create_qp(struct ib_pd *pd, */ qp->srq = to_siw_srq(attrs->srq); qp->attrs.rq_size = 0; - siw_dbg(base_dev, "QP [%u]: [SRQ 0x%p] attached\n", - qp->qp_num, qp->srq); + siw_dbg(base_dev, "QP [%u]: SRQ attached\n", qp->qp_num); } else if (num_rqe) { if (qp->kernel_verbs) qp->recvq = vzalloc(num_rqe * sizeof(struct siw_rqe)); @@ -610,7 +609,7 @@ int siw_destroy_qp(struct ib_qp *base_qp, struct ib_udata *udata) base_ucontext); struct siw_qp_attrs qp_attrs; - siw_dbg_qp(qp, "state %d, cep 0x%p\n", qp->attrs.state, qp->cep); + siw_dbg_qp(qp, "state %d\n", qp->attrs.state); /* * Mark QP as in process of destruction to prevent from @@ -662,7 +661,7 @@ static int siw_copy_inline_sgl(const struct ib_send_wr *core_wr, void *kbuf = &sqe->sge[1]; int num_sge = core_wr->num_sge, bytes = 0; - sqe->sge[0].laddr = (u64)kbuf; + sqe->sge[0].laddr = (uintptr_t)kbuf; sqe->sge[0].lkey = 0; while (num_sge--) { @@ -825,7 +824,7 @@ int siw_post_send(struct ib_qp *base_qp, const struct ib_send_wr *wr, break; case IB_WR_REG_MR: - sqe->base_mr = (uint64_t)reg_wr(wr)->mr; + sqe->base_mr = (uintptr_t)reg_wr(wr)->mr; sqe->rkey = reg_wr(wr)->key; sqe->access = reg_wr(wr)->access & IWARP_ACCESS_MASK; sqe->opcode = SIW_OP_REG_MR; @@ -842,8 +841,9 @@ int siw_post_send(struct ib_qp *base_qp, const struct ib_send_wr *wr, rv = -EINVAL; break; } - siw_dbg_qp(qp, "opcode %d, flags 0x%x, wr_id 0x%p\n", - sqe->opcode, sqe->flags, (void *)sqe->id); + siw_dbg_qp(qp, "opcode %d, flags 0x%x, wr_id 0x%pK\n", + sqe->opcode, sqe->flags, + (void *)(uintptr_t)sqe->id); if (unlikely(rv < 0)) break; @@ -1049,7 +1049,7 @@ int siw_create_cq(struct ib_cq *base_cq, const struct ib_cq_init_attr *attr, spin_lock_init(&cq->lock); - cq->notify = &((struct siw_cq_ctrl *)&cq->queue[size])->notify; + cq->notify = (struct siw_cq_ctrl *)&cq->queue[size]; if (udata) { struct siw_uresp_create_cq uresp = {}; @@ -1141,11 +1141,17 @@ int siw_req_notify_cq(struct ib_cq *base_cq, enum ib_cq_notify_flags flags) siw_dbg_cq(cq, "flags: 0x%02x\n", flags); if ((flags & IB_CQ_SOLICITED_MASK) == IB_CQ_SOLICITED) - /* CQ event for next solicited completion */ - smp_store_mb(*cq->notify, SIW_NOTIFY_SOLICITED); + /* + * Enable CQ event for next solicited completion. + * and make it visible to all associated producers. + */ + smp_store_mb(cq->notify->flags, SIW_NOTIFY_SOLICITED); else - /* CQ event for any signalled completion */ - smp_store_mb(*cq->notify, SIW_NOTIFY_ALL); + /* + * Enable CQ event for any signalled completion. + * and make it visible to all associated producers. + */ + smp_store_mb(cq->notify->flags, SIW_NOTIFY_ALL); if (flags & IB_CQ_REPORT_MISSED_EVENTS) return cq->cq_put - cq->cq_get; @@ -1199,8 +1205,8 @@ struct ib_mr *siw_reg_user_mr(struct ib_pd *pd, u64 start, u64 len, unsigned long mem_limit = rlimit(RLIMIT_MEMLOCK); int rv; - siw_dbg_pd(pd, "start: 0x%016llx, va: 0x%016llx, len: %llu\n", - (unsigned long long)start, (unsigned long long)rnic_va, + siw_dbg_pd(pd, "start: 0x%pK, va: 0x%pK, len: %llu\n", + (void *)(uintptr_t)start, (void *)(uintptr_t)rnic_va, (unsigned long long)len); if (atomic_inc_return(&sdev->num_mr) > SIW_MAX_MR) { @@ -1357,7 +1363,7 @@ int siw_map_mr_sg(struct ib_mr *base_mr, struct scatterlist *sl, int num_sle, struct siw_mem *mem = mr->mem; struct siw_pbl *pbl = mem->pbl; struct siw_pble *pble; - u64 pbl_size; + unsigned long pbl_size; int i, rv; if (!pbl) { @@ -1396,16 +1402,18 @@ int siw_map_mr_sg(struct ib_mr *base_mr, struct scatterlist *sl, int num_sle, pbl_size += sg_dma_len(slp); } siw_dbg_mem(mem, - "sge[%d], size %llu, addr 0x%016llx, total %llu\n", - i, pble->size, pble->addr, pbl_size); + "sge[%d], size %u, addr 0x%p, total %lu\n", + i, pble->size, (void *)(uintptr_t)pble->addr, + pbl_size); } rv = ib_sg_to_pages(base_mr, sl, num_sle, sg_off, siw_set_pbl_page); if (rv > 0) { mem->len = base_mr->length; mem->va = base_mr->iova; siw_dbg_mem(mem, - "%llu bytes, start 0x%016llx, %u SLE to %u entries\n", - mem->len, mem->va, num_sle, pbl->num_buf); + "%llu bytes, start 0x%pK, %u SLE to %u entries\n", + mem->len, (void *)(uintptr_t)mem->va, num_sle, + pbl->num_buf); } return rv; } @@ -1523,7 +1531,7 @@ int siw_create_srq(struct ib_srq *base_srq, } spin_lock_init(&srq->lock); - siw_dbg_pd(base_srq->pd, "[SRQ 0x%p]: success\n", srq); + siw_dbg_pd(base_srq->pd, "[SRQ]: success\n"); return 0; @@ -1644,8 +1652,7 @@ int siw_post_srq_recv(struct ib_srq *base_srq, const struct ib_recv_wr *wr, if (unlikely(!srq->kernel_verbs)) { siw_dbg_pd(base_srq->pd, - "[SRQ 0x%p]: no kernel post_recv for mapped srq\n", - srq); + "[SRQ]: no kernel post_recv for mapped srq\n"); rv = -EINVAL; goto out; } @@ -1667,8 +1674,7 @@ int siw_post_srq_recv(struct ib_srq *base_srq, const struct ib_recv_wr *wr, } if (unlikely(wr->num_sge > srq->max_sge)) { siw_dbg_pd(base_srq->pd, - "[SRQ 0x%p]: too many sge's: %d\n", srq, - wr->num_sge); + "[SRQ]: too many sge's: %d\n", wr->num_sge); rv = -EINVAL; break; } @@ -1687,7 +1693,7 @@ int siw_post_srq_recv(struct ib_srq *base_srq, const struct ib_recv_wr *wr, spin_unlock_irqrestore(&srq->lock, flags); out: if (unlikely(rv < 0)) { - siw_dbg_pd(base_srq->pd, "[SRQ 0x%p]: error %d\n", srq, rv); + siw_dbg_pd(base_srq->pd, "[SRQ]: error %d\n", rv); *bad_wr = wr; } return rv; |