diff options
Diffstat (limited to 'drivers/infiniband')
307 files changed, 33669 insertions, 11081 deletions
diff --git a/drivers/infiniband/Kconfig b/drivers/infiniband/Kconfig index 670917387eda..234fe01904e7 100644 --- a/drivers/infiniband/Kconfig +++ b/drivers/infiniband/Kconfig @@ -85,6 +85,7 @@ source "drivers/infiniband/ulp/srpt/Kconfig" source "drivers/infiniband/ulp/iser/Kconfig" source "drivers/infiniband/ulp/isert/Kconfig" +source "drivers/infiniband/ulp/opa_vnic/Kconfig" source "drivers/infiniband/sw/rdmavt/Kconfig" source "drivers/infiniband/sw/rxe/Kconfig" @@ -92,4 +93,6 @@ source "drivers/infiniband/hw/hfi1/Kconfig" source "drivers/infiniband/hw/qedr/Kconfig" +source "drivers/infiniband/hw/bnxt_re/Kconfig" + endif # INFINIBAND diff --git a/drivers/infiniband/core/Makefile b/drivers/infiniband/core/Makefile index edaae9f9853c..e3cdafff8ece 100644 --- a/drivers/infiniband/core/Makefile +++ b/drivers/infiniband/core/Makefile @@ -10,9 +10,11 @@ obj-$(CONFIG_INFINIBAND_USER_ACCESS) += ib_uverbs.o ib_ucm.o \ ib_core-y := packer.o ud_header.o verbs.o cq.o rw.o sysfs.o \ device.o fmr_pool.o cache.o netlink.o \ roce_gid_mgmt.o mr_pool.o addr.o sa_query.o \ - multicast.o mad.o smi.o agent.o mad_rmpp.o + multicast.o mad.o smi.o agent.o mad_rmpp.o \ + security.o ib_core-$(CONFIG_INFINIBAND_USER_MEM) += umem.o ib_core-$(CONFIG_INFINIBAND_ON_DEMAND_PAGING) += umem_odp.o umem_rbtree.o +ib_core-$(CONFIG_CGROUP_RDMA) += cgroup.o ib_cm-y := cm.o @@ -28,4 +30,5 @@ ib_umad-y := user_mad.o ib_ucm-y := ucm.o -ib_uverbs-y := uverbs_main.o uverbs_cmd.o uverbs_marshall.o +ib_uverbs-y := uverbs_main.o uverbs_cmd.o uverbs_marshall.o \ + rdma_core.o uverbs_std_types.o diff --git a/drivers/infiniband/core/addr.c b/drivers/infiniband/core/addr.c index 0f58f46dbad7..01236cef7bfb 100644 --- a/drivers/infiniband/core/addr.c +++ b/drivers/infiniband/core/addr.c @@ -88,7 +88,7 @@ static inline bool ib_nl_is_good_ip_resp(const struct nlmsghdr *nlh) return false; ret = nla_parse(tb, LS_NLA_TYPE_MAX - 1, nlmsg_data(nlh), - nlmsg_len(nlh), ib_nl_addr_policy); + nlmsg_len(nlh), ib_nl_addr_policy, NULL); if (ret) return false; @@ -179,8 +179,7 @@ static int ib_nl_ip_send_msg(struct rdma_dev_addr *dev_addr, } /* Construct the family header first */ - header = (struct rdma_ls_ip_resolve_header *) - skb_put(skb, NLMSG_ALIGN(sizeof(*header))); + header = skb_put(skb, NLMSG_ALIGN(sizeof(*header))); header->ifindex = dev_addr->bound_dev_if; nla_put(skb, attrtype, size, daddr); @@ -269,6 +268,7 @@ int rdma_translate_ip(const struct sockaddr *addr, return ret; ret = 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); @@ -281,6 +281,7 @@ int rdma_translate_ip(const struct sockaddr *addr, &((const struct sockaddr_in6 *)addr)->sin6_addr, dev, 1)) { ret = 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; @@ -406,10 +407,10 @@ static int addr4_resolve(struct sockaddr_in *src_in, fl4.saddr = src_ip; fl4.flowi4_oif = addr->bound_dev_if; rt = ip_route_output_key(addr->net, &fl4); - if (IS_ERR(rt)) { - ret = PTR_ERR(rt); - goto out; - } + ret = PTR_ERR_OR_ZERO(rt); + if (ret) + return ret; + src_in->sin_family = AF_INET; src_in->sin_addr.s_addr = fl4.saddr; @@ -424,8 +425,6 @@ static int addr4_resolve(struct sockaddr_in *src_in, *prt = rt; return 0; -out: - return ret; } #if IS_ENABLED(CONFIG_IPV6) @@ -444,17 +443,12 @@ static int addr6_resolve(struct sockaddr_in6 *src_in, fl6.saddr = src_in->sin6_addr; fl6.flowi6_oif = addr->bound_dev_if; - dst = ip6_route_output(addr->net, NULL, &fl6); - if ((ret = dst->error)) - goto put; + ret = ipv6_stub->ipv6_dst_lookup(addr->net, NULL, &dst, &fl6); + if (ret < 0) + return ret; rt = (struct rt6_info *)dst; - if (ipv6_addr_any(&fl6.saddr)) { - ret = ipv6_dev_get_saddr(addr->net, ip6_dst_idev(dst)->dev, - &fl6.daddr, 0, &fl6.saddr); - if (ret) - goto put; - + if (ipv6_addr_any(&src_in->sin6_addr)) { src_in->sin6_family = AF_INET6; src_in->sin6_addr = fl6.saddr; } @@ -471,9 +465,6 @@ static int addr6_resolve(struct sockaddr_in6 *src_in, *pdst = dst; return 0; -put: - dst_release(dst); - return ret; } #else static int addr6_resolve(struct sockaddr_in6 *src_in, @@ -518,6 +509,11 @@ static int addr_resolve(struct sockaddr *src_in, struct dst_entry *dst; int ret; + if (!addr->net) { + pr_warn_ratelimited("%s: missing namespace\n", __func__); + return -EINVAL; + } + if (src_in->sa_family == AF_INET) { struct rtable *rt = NULL; const struct sockaddr_in *dst_in4 = @@ -531,8 +527,12 @@ static int addr_resolve(struct sockaddr *src_in, if (resolve_neigh) ret = addr_resolve_neigh(&rt->dst, dst_in, addr, seq); - ndev = rt->dst.dev; - dev_hold(ndev); + if (addr->bound_dev_if) { + ndev = dev_get_by_index(addr->net, addr->bound_dev_if); + } else { + ndev = rt->dst.dev; + dev_hold(ndev); + } ip_rt_put(rt); } else { @@ -548,14 +548,27 @@ static int addr_resolve(struct sockaddr *src_in, if (resolve_neigh) ret = addr_resolve_neigh(dst, dst_in, addr, seq); - ndev = dst->dev; - dev_hold(ndev); + if (addr->bound_dev_if) { + ndev = dev_get_by_index(addr->net, addr->bound_dev_if); + } else { + ndev = dst->dev; + dev_hold(ndev); + } dst_release(dst); } - addr->bound_dev_if = ndev->ifindex; - addr->net = dev_net(ndev); + if (ndev->flags & IFF_LOOPBACK) { + ret = rdma_translate_ip(dst_in, addr, NULL); + /* + * Put the loopback device and get the translated + * device instead. + */ + dev_put(ndev); + ndev = dev_get_by_index(addr->net, addr->bound_dev_if); + } else { + addr->bound_dev_if = ndev->ifindex; + } dev_put(ndev); return ret; diff --git a/drivers/infiniband/core/agent.c b/drivers/infiniband/core/agent.c index 11dacd97a667..324ef85a13b6 100644 --- a/drivers/infiniband/core/agent.c +++ b/drivers/infiniband/core/agent.c @@ -137,13 +137,13 @@ void agent_send_response(const struct ib_mad_hdr *mad_hdr, const struct ib_grh * err2: ib_free_send_mad(send_buf); err1: - ib_destroy_ah(ah); + rdma_destroy_ah(ah); } static void agent_send_handler(struct ib_mad_agent *mad_agent, struct ib_mad_send_wc *mad_send_wc) { - ib_destroy_ah(mad_send_wc->send_buf->ah); + rdma_destroy_ah(mad_send_wc->send_buf->ah); ib_free_send_mad(mad_send_wc->send_buf); } diff --git a/drivers/infiniband/core/cache.c b/drivers/infiniband/core/cache.c index ae04826e82fc..efc94304dee3 100644 --- a/drivers/infiniband/core/cache.c +++ b/drivers/infiniband/core/cache.c @@ -53,6 +53,7 @@ struct ib_update_work { struct work_struct work; struct ib_device *device; u8 port_num; + bool enforce_security; }; union ib_gid zgid; @@ -314,14 +315,13 @@ static void make_default_gid(struct net_device *dev, union ib_gid *gid) int ib_cache_gid_add(struct ib_device *ib_dev, u8 port, union ib_gid *gid, struct ib_gid_attr *attr) { - struct ib_gid_table **ports_table = ib_dev->cache.gid_cache; struct ib_gid_table *table; int ix; int ret = 0; struct net_device *idev; int empty; - table = ports_table[port - rdma_start_port(ib_dev)]; + table = ib_dev->cache.ports[port - rdma_start_port(ib_dev)].gid; if (!memcmp(gid, &zgid, sizeof(*gid))) return -EINVAL; @@ -369,11 +369,10 @@ out_unlock: int ib_cache_gid_del(struct ib_device *ib_dev, u8 port, union ib_gid *gid, struct ib_gid_attr *attr) { - struct ib_gid_table **ports_table = ib_dev->cache.gid_cache; struct ib_gid_table *table; int ix; - table = ports_table[port - rdma_start_port(ib_dev)]; + table = ib_dev->cache.ports[port - rdma_start_port(ib_dev)].gid; mutex_lock(&table->lock); write_lock_irq(&table->rwlock); @@ -399,12 +398,11 @@ out_unlock: int ib_cache_gid_del_all_netdev_gids(struct ib_device *ib_dev, u8 port, struct net_device *ndev) { - struct ib_gid_table **ports_table = ib_dev->cache.gid_cache; struct ib_gid_table *table; int ix; bool deleted = false; - table = ports_table[port - rdma_start_port(ib_dev)]; + table = ib_dev->cache.ports[port - rdma_start_port(ib_dev)].gid; mutex_lock(&table->lock); write_lock_irq(&table->rwlock); @@ -428,10 +426,9 @@ int ib_cache_gid_del_all_netdev_gids(struct ib_device *ib_dev, u8 port, static int __ib_cache_gid_get(struct ib_device *ib_dev, u8 port, int index, union ib_gid *gid, struct ib_gid_attr *attr) { - struct ib_gid_table **ports_table = ib_dev->cache.gid_cache; struct ib_gid_table *table; - table = ports_table[port - rdma_start_port(ib_dev)]; + table = ib_dev->cache.ports[port - rdma_start_port(ib_dev)].gid; if (index < 0 || index >= table->sz) return -EINVAL; @@ -455,14 +452,13 @@ static int _ib_cache_gid_table_find(struct ib_device *ib_dev, unsigned long mask, u8 *port, u16 *index) { - struct ib_gid_table **ports_table = ib_dev->cache.gid_cache; struct ib_gid_table *table; u8 p; int local_index; unsigned long flags; for (p = 0; p < ib_dev->phys_port_cnt; p++) { - table = ports_table[p]; + table = ib_dev->cache.ports[p].gid; read_lock_irqsave(&table->rwlock, flags); local_index = find_gid(table, gid, val, false, mask, NULL); if (local_index >= 0) { @@ -503,18 +499,16 @@ int ib_find_cached_gid_by_port(struct ib_device *ib_dev, u16 *index) { int local_index; - struct ib_gid_table **ports_table = ib_dev->cache.gid_cache; struct ib_gid_table *table; unsigned long mask = GID_ATTR_FIND_MASK_GID | GID_ATTR_FIND_MASK_GID_TYPE; struct ib_gid_attr val = {.ndev = ndev, .gid_type = gid_type}; unsigned long flags; - if (port < rdma_start_port(ib_dev) || - port > rdma_end_port(ib_dev)) + if (!rdma_is_port_valid(ib_dev, port)) return -ENOENT; - table = ports_table[port - rdma_start_port(ib_dev)]; + table = ib_dev->cache.ports[port - rdma_start_port(ib_dev)].gid; if (ndev) mask |= GID_ATTR_FIND_MASK_NETDEV; @@ -562,21 +556,17 @@ static int ib_cache_gid_find_by_filter(struct ib_device *ib_dev, void *context, u16 *index) { - struct ib_gid_table **ports_table = ib_dev->cache.gid_cache; struct ib_gid_table *table; unsigned int i; unsigned long flags; bool found = false; - if (!ports_table) - return -EOPNOTSUPP; - if (port < rdma_start_port(ib_dev) || - port > rdma_end_port(ib_dev) || + if (!rdma_is_port_valid(ib_dev, port) || !rdma_protocol_roce(ib_dev, port)) return -EPROTONOSUPPORT; - table = ports_table[port - rdma_start_port(ib_dev)]; + table = ib_dev->cache.ports[port - rdma_start_port(ib_dev)].gid; read_lock_irqsave(&table->rwlock, flags); for (i = 0; i < table->sz; i++) { @@ -668,14 +658,13 @@ void ib_cache_gid_set_default_gid(struct ib_device *ib_dev, u8 port, unsigned long gid_type_mask, enum ib_cache_gid_default_mode mode) { - struct ib_gid_table **ports_table = ib_dev->cache.gid_cache; union ib_gid gid; struct ib_gid_attr gid_attr; struct ib_gid_attr zattr_type = zattr; struct ib_gid_table *table; unsigned int gid_type; - table = ports_table[port - rdma_start_port(ib_dev)]; + table = ib_dev->cache.ports[port - rdma_start_port(ib_dev)].gid; make_default_gid(ndev, &gid); memset(&gid_attr, 0, sizeof(gid_attr)); @@ -766,71 +755,64 @@ static int gid_table_reserve_default(struct ib_device *ib_dev, u8 port, static int _gid_table_setup_one(struct ib_device *ib_dev) { u8 port; - struct ib_gid_table **table; + struct ib_gid_table *table; int err = 0; - table = kcalloc(ib_dev->phys_port_cnt, sizeof(*table), GFP_KERNEL); - if (!table) - return -ENOMEM; - for (port = 0; port < ib_dev->phys_port_cnt; port++) { u8 rdma_port = port + rdma_start_port(ib_dev); - table[port] = + table = alloc_gid_table( ib_dev->port_immutable[rdma_port].gid_tbl_len); - if (!table[port]) { + if (!table) { err = -ENOMEM; goto rollback_table_setup; } err = gid_table_reserve_default(ib_dev, port + rdma_start_port(ib_dev), - table[port]); + table); if (err) goto rollback_table_setup; + ib_dev->cache.ports[port].gid = table; } - ib_dev->cache.gid_cache = table; return 0; rollback_table_setup: for (port = 0; port < ib_dev->phys_port_cnt; port++) { + table = ib_dev->cache.ports[port].gid; + cleanup_gid_table_port(ib_dev, port + rdma_start_port(ib_dev), - table[port]); - release_gid_table(table[port]); + table); + release_gid_table(table); } - kfree(table); return err; } static void gid_table_release_one(struct ib_device *ib_dev) { - struct ib_gid_table **table = ib_dev->cache.gid_cache; + struct ib_gid_table *table; u8 port; - if (!table) - return; - - for (port = 0; port < ib_dev->phys_port_cnt; port++) - release_gid_table(table[port]); - - kfree(table); - ib_dev->cache.gid_cache = NULL; + for (port = 0; port < ib_dev->phys_port_cnt; port++) { + table = ib_dev->cache.ports[port].gid; + release_gid_table(table); + ib_dev->cache.ports[port].gid = NULL; + } } static void gid_table_cleanup_one(struct ib_device *ib_dev) { - struct ib_gid_table **table = ib_dev->cache.gid_cache; + struct ib_gid_table *table; u8 port; - if (!table) - return; - - for (port = 0; port < ib_dev->phys_port_cnt; port++) + for (port = 0; port < ib_dev->phys_port_cnt; port++) { + table = ib_dev->cache.ports[port].gid; cleanup_gid_table_port(ib_dev, port + rdma_start_port(ib_dev), - table[port]); + table); + } } static int gid_table_setup_one(struct ib_device *ib_dev) @@ -860,12 +842,12 @@ int ib_get_cached_gid(struct ib_device *device, { int res; unsigned long flags; - struct ib_gid_table **ports_table = device->cache.gid_cache; - struct ib_gid_table *table = ports_table[port_num - rdma_start_port(device)]; + struct ib_gid_table *table; - if (port_num < rdma_start_port(device) || port_num > rdma_end_port(device)) + if (!rdma_is_port_valid(device, port_num)) return -EINVAL; + table = device->cache.ports[port_num - rdma_start_port(device)].gid; read_lock_irqsave(&table->rwlock, flags); res = __ib_cache_gid_get(device, port_num, index, gid, gid_attr); read_unlock_irqrestore(&table->rwlock, flags); @@ -912,12 +894,12 @@ int ib_get_cached_pkey(struct ib_device *device, unsigned long flags; int ret = 0; - if (port_num < rdma_start_port(device) || port_num > rdma_end_port(device)) + if (!rdma_is_port_valid(device, port_num)) return -EINVAL; read_lock_irqsave(&device->cache.lock, flags); - cache = device->cache.pkey_cache[port_num - rdma_start_port(device)]; + cache = device->cache.ports[port_num - rdma_start_port(device)].pkey; if (index < 0 || index >= cache->table_len) ret = -EINVAL; @@ -930,6 +912,26 @@ int ib_get_cached_pkey(struct ib_device *device, } EXPORT_SYMBOL(ib_get_cached_pkey); +int ib_get_cached_subnet_prefix(struct ib_device *device, + u8 port_num, + u64 *sn_pfx) +{ + unsigned long flags; + int p; + + if (port_num < rdma_start_port(device) || + port_num > rdma_end_port(device)) + return -EINVAL; + + p = port_num - rdma_start_port(device); + read_lock_irqsave(&device->cache.lock, flags); + *sn_pfx = device->cache.ports[p].subnet_prefix; + read_unlock_irqrestore(&device->cache.lock, flags); + + return 0; +} +EXPORT_SYMBOL(ib_get_cached_subnet_prefix); + int ib_find_cached_pkey(struct ib_device *device, u8 port_num, u16 pkey, @@ -941,12 +943,12 @@ int ib_find_cached_pkey(struct ib_device *device, int ret = -ENOENT; int partial_ix = -1; - if (port_num < rdma_start_port(device) || port_num > rdma_end_port(device)) + if (!rdma_is_port_valid(device, port_num)) return -EINVAL; read_lock_irqsave(&device->cache.lock, flags); - cache = device->cache.pkey_cache[port_num - rdma_start_port(device)]; + cache = device->cache.ports[port_num - rdma_start_port(device)].pkey; *index = -1; @@ -981,12 +983,12 @@ int ib_find_exact_cached_pkey(struct ib_device *device, int i; int ret = -ENOENT; - if (port_num < rdma_start_port(device) || port_num > rdma_end_port(device)) + if (!rdma_is_port_valid(device, port_num)) return -EINVAL; read_lock_irqsave(&device->cache.lock, flags); - cache = device->cache.pkey_cache[port_num - rdma_start_port(device)]; + cache = device->cache.ports[port_num - rdma_start_port(device)].pkey; *index = -1; @@ -1010,19 +1012,39 @@ int ib_get_cached_lmc(struct ib_device *device, unsigned long flags; int ret = 0; - if (port_num < rdma_start_port(device) || port_num > rdma_end_port(device)) + if (!rdma_is_port_valid(device, port_num)) return -EINVAL; read_lock_irqsave(&device->cache.lock, flags); - *lmc = device->cache.lmc_cache[port_num - rdma_start_port(device)]; + *lmc = device->cache.ports[port_num - rdma_start_port(device)].lmc; read_unlock_irqrestore(&device->cache.lock, flags); return ret; } EXPORT_SYMBOL(ib_get_cached_lmc); +int ib_get_cached_port_state(struct ib_device *device, + u8 port_num, + enum ib_port_state *port_state) +{ + unsigned long flags; + int ret = 0; + + if (port_num < rdma_start_port(device) || port_num > rdma_end_port(device)) + return -EINVAL; + + read_lock_irqsave(&device->cache.lock, flags); + *port_state = device->cache.ports[port_num + - rdma_start_port(device)].port_state; + read_unlock_irqrestore(&device->cache.lock, flags); + + return ret; +} +EXPORT_SYMBOL(ib_get_cached_port_state); + static void ib_cache_update(struct ib_device *device, - u8 port) + u8 port, + bool enforce_security) { struct ib_port_attr *tprops = NULL; struct ib_pkey_cache *pkey_cache = NULL, *old_pkey_cache; @@ -1033,14 +1055,13 @@ static void ib_cache_update(struct ib_device *device, int i; int ret; struct ib_gid_table *table; - struct ib_gid_table **ports_table = device->cache.gid_cache; bool use_roce_gid_table = rdma_cap_roce_gid_table(device, port); - if (port < rdma_start_port(device) || port > rdma_end_port(device)) + if (!rdma_is_port_valid(device, port)) return; - table = ports_table[port - rdma_start_port(device)]; + table = device->cache.ports[port - rdma_start_port(device)].gid; tprops = kmalloc(sizeof *tprops, GFP_KERNEL); if (!tprops) @@ -1092,9 +1113,10 @@ static void ib_cache_update(struct ib_device *device, write_lock_irq(&device->cache.lock); - old_pkey_cache = device->cache.pkey_cache[port - rdma_start_port(device)]; + old_pkey_cache = device->cache.ports[port - + rdma_start_port(device)].pkey; - device->cache.pkey_cache[port - rdma_start_port(device)] = pkey_cache; + device->cache.ports[port - rdma_start_port(device)].pkey = pkey_cache; if (!use_roce_gid_table) { write_lock(&table->rwlock); for (i = 0; i < gid_cache->table_len; i++) { @@ -1104,10 +1126,19 @@ static void ib_cache_update(struct ib_device *device, write_unlock(&table->rwlock); } - device->cache.lmc_cache[port - rdma_start_port(device)] = tprops->lmc; + device->cache.ports[port - rdma_start_port(device)].lmc = tprops->lmc; + device->cache.ports[port - rdma_start_port(device)].port_state = + tprops->state; + device->cache.ports[port - rdma_start_port(device)].subnet_prefix = + tprops->subnet_prefix; write_unlock_irq(&device->cache.lock); + if (enforce_security) + ib_security_cache_change(device, + port, + tprops->subnet_prefix); + kfree(gid_cache); kfree(old_pkey_cache); kfree(tprops); @@ -1124,7 +1155,9 @@ static void ib_cache_task(struct work_struct *_work) struct ib_update_work *work = container_of(_work, struct ib_update_work, work); - ib_cache_update(work->device, work->port_num); + ib_cache_update(work->device, + work->port_num, + work->enforce_security); kfree(work); } @@ -1145,6 +1178,12 @@ static void ib_cache_event(struct ib_event_handler *handler, INIT_WORK(&work->work, ib_cache_task); work->device = event->device; work->port_num = event->element.port_num; + if (event->event == IB_EVENT_PKEY_CHANGE || + event->event == IB_EVENT_GID_CHANGE) + work->enforce_security = true; + else + work->enforce_security = false; + queue_work(ib_wq, &work->work); } } @@ -1157,25 +1196,20 @@ int ib_cache_setup_one(struct ib_device *device) rwlock_init(&device->cache.lock); - device->cache.pkey_cache = - kzalloc(sizeof *device->cache.pkey_cache * + device->cache.ports = + kzalloc(sizeof(*device->cache.ports) * (rdma_end_port(device) - rdma_start_port(device) + 1), GFP_KERNEL); - device->cache.lmc_cache = kmalloc(sizeof *device->cache.lmc_cache * - (rdma_end_port(device) - - rdma_start_port(device) + 1), - GFP_KERNEL); - if (!device->cache.pkey_cache || - !device->cache.lmc_cache) { + if (!device->cache.ports) { err = -ENOMEM; - goto free; + goto out; } err = gid_table_setup_one(device); if (err) - goto free; + goto out; for (p = 0; p <= rdma_end_port(device) - rdma_start_port(device); ++p) - ib_cache_update(device, p + rdma_start_port(device)); + ib_cache_update(device, p + rdma_start_port(device), true); INIT_IB_EVENT_HANDLER(&device->cache.event_handler, device, ib_cache_event); @@ -1187,9 +1221,7 @@ int ib_cache_setup_one(struct ib_device *device) err: gid_table_cleanup_one(device); -free: - kfree(device->cache.pkey_cache); - kfree(device->cache.lmc_cache); +out: return err; } @@ -1203,14 +1235,11 @@ void ib_cache_release_one(struct ib_device *device) * all the device's resources when the cache could no * longer be accessed. */ - if (device->cache.pkey_cache) - for (p = 0; - p <= rdma_end_port(device) - rdma_start_port(device); ++p) - kfree(device->cache.pkey_cache[p]); + for (p = 0; p <= rdma_end_port(device) - rdma_start_port(device); ++p) + kfree(device->cache.ports[p].pkey); gid_table_release_one(device); - kfree(device->cache.pkey_cache); - kfree(device->cache.lmc_cache); + kfree(device->cache.ports); } void ib_cache_cleanup_one(struct ib_device *device) diff --git a/drivers/infiniband/core/cgroup.c b/drivers/infiniband/core/cgroup.c new file mode 100644 index 000000000000..126ac5f99db7 --- /dev/null +++ b/drivers/infiniband/core/cgroup.c @@ -0,0 +1,62 @@ +/* + * Copyright (C) 2016 Parav Pandit <pandit.parav@gmail.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + */ + +#include "core_priv.h" + +/** + * ib_device_register_rdmacg - register with rdma cgroup. + * @device: device to register to participate in resource + * accounting by rdma cgroup. + * + * Register with the rdma cgroup. Should be called before + * exposing rdma device to user space applications to avoid + * resource accounting leak. + * Returns 0 on success or otherwise failure code. + */ +int ib_device_register_rdmacg(struct ib_device *device) +{ + device->cg_device.name = device->name; + return rdmacg_register_device(&device->cg_device); +} + +/** + * ib_device_unregister_rdmacg - unregister with rdma cgroup. + * @device: device to unregister. + * + * Unregister with the rdma cgroup. Should be called after + * all the resources are deallocated, and after a stage when any + * other resource allocation by user application cannot be done + * for this device to avoid any leak in accounting. + */ +void ib_device_unregister_rdmacg(struct ib_device *device) +{ + rdmacg_unregister_device(&device->cg_device); +} + +int ib_rdmacg_try_charge(struct ib_rdmacg_object *cg_obj, + struct ib_device *device, + enum rdmacg_resource_type resource_index) +{ + return rdmacg_try_charge(&cg_obj->cg, &device->cg_device, + resource_index); +} +EXPORT_SYMBOL(ib_rdmacg_try_charge); + +void ib_rdmacg_uncharge(struct ib_rdmacg_object *cg_obj, + struct ib_device *device, + enum rdmacg_resource_type resource_index) +{ + rdmacg_uncharge(cg_obj->cg, &device->cg_device, + resource_index); +} +EXPORT_SYMBOL(ib_rdmacg_uncharge); diff --git a/drivers/infiniband/core/cm.c b/drivers/infiniband/core/cm.c index cf1edfa1cbac..2b4d613a3474 100644 --- a/drivers/infiniband/core/cm.c +++ b/drivers/infiniband/core/cm.c @@ -228,7 +228,7 @@ struct cm_device { struct cm_av { struct cm_port *port; union ib_gid dgid; - struct ib_ah_attr ah_attr; + struct rdma_ah_attr ah_attr; u16 pkey_index; u8 timeout; }; @@ -241,7 +241,7 @@ struct cm_work { __be32 local_id; /* Established / timewait */ __be32 remote_id; struct ib_cm_event cm_event; - struct ib_sa_path_rec path[0]; + struct sa_path_rec path[0]; }; struct cm_timewait_info { @@ -343,7 +343,7 @@ static int cm_alloc_msg(struct cm_id_private *cm_id_priv, ret = -ENODEV; goto out; } - ah = ib_create_ah(mad_agent->qp->pd, &av->ah_attr); + ah = rdma_create_ah(mad_agent->qp->pd, &av->ah_attr); if (IS_ERR(ah)) { ret = PTR_ERR(ah); goto out; @@ -355,7 +355,7 @@ static int cm_alloc_msg(struct cm_id_private *cm_id_priv, GFP_ATOMIC, IB_MGMT_BASE_VERSION); if (IS_ERR(m)) { - ib_destroy_ah(ah); + rdma_destroy_ah(ah); ret = PTR_ERR(m); goto out; } @@ -390,7 +390,7 @@ static int cm_alloc_response_msg(struct cm_port *port, GFP_ATOMIC, IB_MGMT_BASE_VERSION); if (IS_ERR(m)) { - ib_destroy_ah(ah); + rdma_destroy_ah(ah); return PTR_ERR(m); } m->ah = ah; @@ -400,7 +400,7 @@ static int cm_alloc_response_msg(struct cm_port *port, static void cm_free_msg(struct ib_mad_send_buf *msg) { - ib_destroy_ah(msg->ah); + rdma_destroy_ah(msg->ah); if (msg->context[0]) cm_deref_id(msg->context[0]); ib_free_send_mad(msg); @@ -440,7 +440,7 @@ static void cm_init_av_for_response(struct cm_port *port, struct ib_wc *wc, grh, &av->ah_attr); } -static int cm_init_av_by_path(struct ib_sa_path_rec *path, struct cm_av *av, +static int cm_init_av_by_path(struct sa_path_rec *path, struct cm_av *av, struct cm_id_private *cm_id_priv) { struct cm_device *cm_dev; @@ -453,7 +453,8 @@ static int cm_init_av_by_path(struct ib_sa_path_rec *path, struct cm_av *av, read_lock_irqsave(&cm.device_lock, flags); list_for_each_entry(cm_dev, &cm.device_list, list) { if (!ib_find_cached_gid(cm_dev->ib_device, &path->sgid, - path->gid_type, ndev, &p, NULL)) { + sa_conv_pathrec_to_gid_type(path), + ndev, &p, NULL)) { port = cm_dev->port[p-1]; break; } @@ -1172,8 +1173,8 @@ static void cm_format_req(struct cm_req_msg *req_msg, struct cm_id_private *cm_id_priv, struct ib_cm_req_param *param) { - struct ib_sa_path_rec *pri_path = param->primary_path; - struct ib_sa_path_rec *alt_path = param->alternate_path; + struct sa_path_rec *pri_path = param->primary_path; + struct sa_path_rec *alt_path = param->alternate_path; cm_format_mad_hdr(&req_msg->hdr, CM_REQ_ATTR_ID, cm_form_tid(cm_id_priv, CM_MSG_SEQUENCE_REQ)); @@ -1202,8 +1203,10 @@ static void cm_format_req(struct cm_req_msg *req_msg, } if (pri_path->hop_limit <= 1) { - req_msg->primary_local_lid = pri_path->slid; - req_msg->primary_remote_lid = pri_path->dlid; + req_msg->primary_local_lid = + htons(ntohl(sa_path_get_slid(pri_path))); + req_msg->primary_remote_lid = + htons(ntohl(sa_path_get_dlid(pri_path))); } else { /* Work-around until there's a way to obtain remote LID info */ req_msg->primary_local_lid = IB_LID_PERMISSIVE; @@ -1223,8 +1226,10 @@ static void cm_format_req(struct cm_req_msg *req_msg, if (alt_path) { if (alt_path->hop_limit <= 1) { - req_msg->alt_local_lid = alt_path->slid; - req_msg->alt_remote_lid = alt_path->dlid; + req_msg->alt_local_lid = + htons(ntohl(sa_path_get_slid(alt_path))); + req_msg->alt_remote_lid = + htons(ntohl(sa_path_get_dlid(alt_path))); } else { req_msg->alt_local_lid = IB_LID_PERMISSIVE; req_msg->alt_remote_lid = IB_LID_PERMISSIVE; @@ -1401,14 +1406,15 @@ static inline int cm_is_active_peer(__be64 local_ca_guid, __be64 remote_ca_guid, } static void cm_format_paths_from_req(struct cm_req_msg *req_msg, - struct ib_sa_path_rec *primary_path, - struct ib_sa_path_rec *alt_path) + struct sa_path_rec *primary_path, + struct sa_path_rec *alt_path) { - memset(primary_path, 0, sizeof *primary_path); primary_path->dgid = req_msg->primary_local_gid; primary_path->sgid = req_msg->primary_remote_gid; - primary_path->dlid = req_msg->primary_local_lid; - primary_path->slid = req_msg->primary_remote_lid; + sa_path_set_dlid(primary_path, + htonl(ntohs(req_msg->primary_local_lid))); + sa_path_set_slid(primary_path, + htonl(ntohs(req_msg->primary_remote_lid))); primary_path->flow_label = cm_req_get_primary_flow_label(req_msg); primary_path->hop_limit = req_msg->primary_hop_limit; primary_path->traffic_class = req_msg->primary_traffic_class; @@ -1426,11 +1432,12 @@ static void cm_format_paths_from_req(struct cm_req_msg *req_msg, primary_path->service_id = req_msg->service_id; if (req_msg->alt_local_lid) { - memset(alt_path, 0, sizeof *alt_path); alt_path->dgid = req_msg->alt_local_gid; alt_path->sgid = req_msg->alt_remote_gid; - alt_path->dlid = req_msg->alt_local_lid; - alt_path->slid = req_msg->alt_remote_lid; + sa_path_set_dlid(alt_path, + htonl(ntohs(req_msg->alt_local_lid))); + sa_path_set_slid(alt_path, + htonl(ntohs(req_msg->alt_remote_lid))); alt_path->flow_label = cm_req_get_alt_flow_label(req_msg); alt_path->hop_limit = req_msg->alt_hop_limit; alt_path->traffic_class = req_msg->alt_traffic_class; @@ -1722,6 +1729,7 @@ static int cm_req_handler(struct cm_work *work) struct cm_req_msg *req_msg; union ib_gid gid; struct ib_gid_attr gid_attr; + const struct ib_global_route *grh; int ret; req_msg = (struct cm_req_msg *)work->mad_recv_wc->recv_buf.mad; @@ -1758,21 +1766,34 @@ static int cm_req_handler(struct cm_work *work) cm_id_priv->id.service_mask = ~cpu_to_be64(0); cm_process_routed_req(req_msg, work->mad_recv_wc->wc); - cm_format_paths_from_req(req_msg, &work->path[0], &work->path[1]); - memcpy(work->path[0].dmac, cm_id_priv->av.ah_attr.dmac, ETH_ALEN); - work->path[0].hop_limit = cm_id_priv->av.ah_attr.grh.hop_limit; + memset(&work->path[0], 0, sizeof(work->path[0])); + memset(&work->path[1], 0, sizeof(work->path[1])); + grh = rdma_ah_read_grh(&cm_id_priv->av.ah_attr); ret = ib_get_cached_gid(work->port->cm_dev->ib_device, work->port->port_num, - cm_id_priv->av.ah_attr.grh.sgid_index, + grh->sgid_index, &gid, &gid_attr); if (!ret) { if (gid_attr.ndev) { - work->path[0].ifindex = gid_attr.ndev->ifindex; - work->path[0].net = dev_net(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 { + work->path[0].rec_type = SA_PATH_REC_TYPE_IB; } - work->path[0].gid_type = gid_attr.gid_type; + if (req_msg->alt_local_lid) + 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); } @@ -1782,11 +1803,18 @@ static int cm_req_handler(struct cm_work *work) &work->path[0].sgid, &gid_attr); if (!err && gid_attr.ndev) { - work->path[0].ifindex = gid_attr.ndev->ifindex; - work->path[0].net = dev_net(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 { + work->path[0].rec_type = SA_PATH_REC_TYPE_IB; } - work->path[0].gid_type = gid_attr.gid_type; + if (req_msg->alt_local_lid) + 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); @@ -2811,7 +2839,7 @@ out: static void cm_format_lap(struct cm_lap_msg *lap_msg, struct cm_id_private *cm_id_priv, - struct ib_sa_path_rec *alternate_path, + struct sa_path_rec *alternate_path, const void *private_data, u8 private_data_len) { @@ -2822,8 +2850,10 @@ static void cm_format_lap(struct cm_lap_msg *lap_msg, cm_lap_set_remote_qpn(lap_msg, cm_id_priv->remote_qpn); /* todo: need remote CM response timeout */ cm_lap_set_remote_resp_timeout(lap_msg, 0x1F); - lap_msg->alt_local_lid = alternate_path->slid; - lap_msg->alt_remote_lid = alternate_path->dlid; + lap_msg->alt_local_lid = + htons(ntohl(sa_path_get_slid(alternate_path))); + lap_msg->alt_remote_lid = + htons(ntohl(sa_path_get_dlid(alternate_path))); lap_msg->alt_local_gid = alternate_path->sgid; lap_msg->alt_remote_gid = alternate_path->dgid; cm_lap_set_flow_label(lap_msg, alternate_path->flow_label); @@ -2841,7 +2871,7 @@ static void cm_format_lap(struct cm_lap_msg *lap_msg, } int ib_send_cm_lap(struct ib_cm_id *cm_id, - struct ib_sa_path_rec *alternate_path, + struct sa_path_rec *alternate_path, const void *private_data, u8 private_data_len) { @@ -2895,14 +2925,15 @@ out: spin_unlock_irqrestore(&cm_id_priv->lock, flags); EXPORT_SYMBOL(ib_send_cm_lap); static void cm_format_path_from_lap(struct cm_id_private *cm_id_priv, - struct ib_sa_path_rec *path, + struct sa_path_rec *path, struct cm_lap_msg *lap_msg) { memset(path, 0, sizeof *path); + path->rec_type = SA_PATH_REC_TYPE_IB; path->dgid = lap_msg->alt_local_gid; path->sgid = lap_msg->alt_remote_gid; - path->dlid = lap_msg->alt_local_lid; - path->slid = lap_msg->alt_remote_lid; + sa_path_set_dlid(path, htonl(ntohs(lap_msg->alt_local_lid))); + sa_path_set_slid(path, htonl(ntohs(lap_msg->alt_remote_lid))); path->flow_label = cm_lap_get_flow_label(lap_msg); path->hop_limit = lap_msg->alt_hop_limit; path->traffic_class = cm_lap_get_traffic_class(lap_msg); @@ -3409,6 +3440,8 @@ static void cm_process_send_error(struct ib_mad_send_buf *msg, if (msg != cm_id_priv->msg || state != cm_id_priv->id.state) goto discard; + pr_debug_ratelimited("CM: failed sending MAD in state %d. (%s)\n", + state, ib_wc_status_msg(wc_status)); switch (state) { case IB_CM_REQ_SENT: case IB_CM_MRA_REQ_RCVD: @@ -3706,7 +3739,7 @@ static void cm_recv_handler(struct ib_mad_agent *mad_agent, atomic_long_inc(&port->counter_group[CM_RECV]. counter[attr_id - CM_ATTR_ID_OFFSET]); - work = kmalloc(sizeof *work + sizeof(struct ib_sa_path_rec) * paths, + work = kmalloc(sizeof(*work) + sizeof(struct sa_path_rec) * paths, GFP_KERNEL); if (!work) { ib_free_recv_mad(mad_recv_wc); @@ -3798,7 +3831,7 @@ static int cm_init_qp_rtr_attr(struct cm_id_private *cm_id_priv, cm_id_priv->responder_resources; qp_attr->min_rnr_timer = 0; } - if (cm_id_priv->alt_av.ah_attr.dlid) { + if (rdma_ah_get_dlid(&cm_id_priv->alt_av.ah_attr)) { *qp_attr_mask |= IB_QP_ALT_PATH; qp_attr->alt_port_num = cm_id_priv->alt_av.port->port_num; qp_attr->alt_pkey_index = cm_id_priv->alt_av.pkey_index; @@ -3852,7 +3885,7 @@ static int cm_init_qp_rts_attr(struct cm_id_private *cm_id_priv, default: break; } - if (cm_id_priv->alt_av.ah_attr.dlid) { + if (rdma_ah_get_dlid(&cm_id_priv->alt_av.ah_attr)) { *qp_attr_mask |= IB_QP_PATH_MIG_STATE; qp_attr->path_mig_state = IB_MIG_REARM; } diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index 3e70a9c5d79d..0eb393237ba2 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -198,6 +198,7 @@ struct cma_device { atomic_t refcount; struct list_head id_list; enum ib_gid_type *default_gid_type; + u8 *default_roce_tos; }; struct rdma_bind_list { @@ -269,8 +270,7 @@ struct cma_device *cma_enum_devices_by_ibdev(cma_device_filter filter, int cma_get_default_gid_type(struct cma_device *cma_dev, unsigned int port) { - if (port < rdma_start_port(cma_dev->device) || - port > rdma_end_port(cma_dev->device)) + if (!rdma_is_port_valid(cma_dev->device, port)) return -EINVAL; return cma_dev->default_gid_type[port - rdma_start_port(cma_dev->device)]; @@ -282,8 +282,7 @@ int cma_set_default_gid_type(struct cma_device *cma_dev, { unsigned long supported_gids; - if (port < rdma_start_port(cma_dev->device) || - port > rdma_end_port(cma_dev->device)) + if (!rdma_is_port_valid(cma_dev->device, port)) return -EINVAL; supported_gids = roce_gid_type_mask_support(cma_dev->device, port); @@ -297,6 +296,25 @@ int cma_set_default_gid_type(struct cma_device *cma_dev, return 0; } +int cma_get_default_roce_tos(struct cma_device *cma_dev, unsigned int port) +{ + if (!rdma_is_port_valid(cma_dev->device, port)) + return -EINVAL; + + return cma_dev->default_roce_tos[port - rdma_start_port(cma_dev->device)]; +} + +int cma_set_default_roce_tos(struct cma_device *cma_dev, unsigned int port, + u8 default_roce_tos) +{ + if (!rdma_is_port_valid(cma_dev->device, port)) + return -EINVAL; + + cma_dev->default_roce_tos[port - rdma_start_port(cma_dev->device)] = + default_roce_tos; + + return 0; +} struct ib_device *cma_get_ib_dev(struct cma_device *cma_dev) { return cma_dev->device; @@ -343,6 +361,7 @@ struct rdma_id_private { u32 options; u8 srq; u8 tos; + bool tos_set; u8 reuseaddr; u8 afonly; enum ib_gid_type gid_type; @@ -604,22 +623,11 @@ static inline int cma_validate_port(struct ib_device *device, u8 port, if ((dev_type != ARPHRD_INFINIBAND) && rdma_protocol_ib(device, port)) return ret; - if (dev_type == ARPHRD_ETHER && rdma_protocol_roce(device, port)) { + if (dev_type == ARPHRD_ETHER && rdma_protocol_roce(device, port)) ndev = dev_get_by_index(&init_net, bound_if_index); - if (ndev && ndev->flags & IFF_LOOPBACK) { - pr_info("detected loopback device\n"); - dev_put(ndev); - - if (!device->get_netdev) - return -EOPNOTSUPP; - - ndev = device->get_netdev(device, port); - if (!ndev) - return -ENODEV; - } - } else { + else gid_type = IB_GID_TYPE_IB; - } + ret = ib_find_cached_gid_by_port(device, gid, gid_type, port, ndev, NULL); @@ -709,6 +717,7 @@ static int cma_resolve_ib_dev(struct rdma_id_private *id_priv) union ib_gid gid, sgid, *dgid; u16 pkey, index; u8 p; + enum ib_port_state port_state; int i; cma_dev = NULL; @@ -724,6 +733,8 @@ static int cma_resolve_ib_dev(struct rdma_id_private *id_priv) if (ib_find_cached_pkey(cur_dev->device, p, pkey, &index)) continue; + if (ib_get_cached_port_state(cur_dev->device, p, &port_state)) + continue; for (i = 0; !ib_get_cached_gid(cur_dev->device, p, i, &gid, NULL); i++) { @@ -735,7 +746,8 @@ static int cma_resolve_ib_dev(struct rdma_id_private *id_priv) } if (!cma_dev && (gid.global.subnet_prefix == - dgid->global.subnet_prefix)) { + dgid->global.subnet_prefix) && + port_state == IB_PORT_ACTIVE) { cma_dev = cur_dev; sgid = gid; id_priv->id.port_num = p; @@ -778,6 +790,7 @@ struct rdma_cm_id *rdma_create_id(struct net *net, id_priv->id.event_handler = event_handler; id_priv->id.ps = ps; id_priv->id.qp_type = qp_type; + id_priv->tos_set = false; spin_lock_init(&id_priv->lock); mutex_init(&id_priv->qp_mutex); init_completion(&id_priv->comp); @@ -905,7 +918,8 @@ static int cma_modify_qp_rtr(struct rdma_id_private *id_priv, goto out; ret = ib_query_gid(id_priv->id.device, id_priv->id.port_num, - qp_attr.ah_attr.grh.sgid_index, &sgid, NULL); + rdma_ah_read_grh(&qp_attr.ah_attr)->sgid_index, + &sgid, NULL); if (ret) goto out; @@ -1019,6 +1033,8 @@ int rdma_init_qp_attr(struct rdma_cm_id *id, struct ib_qp_attr *qp_attr, } else ret = iw_cm_init_qp_attr(id_priv->cm_id.iw, qp_attr, qp_attr_mask); + qp_attr->port_num = id_priv->id.port_num; + *qp_attr_mask |= IB_QP_PORT; } else ret = -ENOSYS; @@ -1103,7 +1119,7 @@ static inline int cma_any_port(struct sockaddr *addr) static void cma_save_ib_info(struct sockaddr *src_addr, struct sockaddr *dst_addr, struct rdma_cm_id *listen_id, - struct ib_sa_path_rec *path) + struct sa_path_rec *path) { struct sockaddr_ib *listen_ib, *ib; @@ -1249,7 +1265,7 @@ static int cma_save_req_info(const struct ib_cm_event *ib_event, memcpy(&req->local_gid, &req_param->primary_path->sgid, sizeof(req->local_gid)); req->has_gid = true; - req->service_id = req_param->primary_path->service_id; + req->service_id = req_param->primary_path->service_id; req->pkey = be16_to_cpu(req_param->primary_path->pkey); if (req->pkey != req_param->bth_pkey) pr_warn_ratelimited("RDMA CMA: got different BTH P_Key (0x%x) and primary path P_Key (0x%x)\n" @@ -1689,6 +1705,7 @@ static int cma_rep_recv(struct rdma_id_private *id_priv) return 0; reject: + pr_debug_ratelimited("RDMA CM: CONNECT_ERROR: failed to handle reply. status %d\n", ret); cma_modify_qp_err(id_priv); ib_send_cm_rej(id_priv->cm_id.ib, IB_CM_REJ_CONSUMER_DEFINED, NULL, 0, NULL, 0); @@ -1730,6 +1747,9 @@ static int cma_ib_handler(struct ib_cm_id *cm_id, struct ib_cm_event *ib_event) event.status = -ETIMEDOUT; break; case IB_CM_REP_RECEIVED: + if (cma_comp(id_priv, RDMA_CM_CONNECT) && + (id_priv->id.qp_type != IB_QPT_UD)) + ib_send_cm_mra(cm_id, CMA_CM_MRA_SETTING, NULL, 0); if (id_priv->id.qp) { event.status = cma_rep_recv(id_priv); event.event = event.status ? RDMA_CM_EVENT_CONNECT_ERROR : @@ -1760,6 +1780,8 @@ static int cma_ib_handler(struct ib_cm_id *cm_id, struct ib_cm_event *ib_event) /* ignore event */ goto out; case IB_CM_REJ_RECEIVED: + pr_debug_ratelimited("RDMA CM: REJECTED: %s\n", rdma_reject_msg(&id_priv->id, + ib_event->param.rej_rcvd.reason)); cma_modify_qp_err(id_priv); event.status = ib_event->param.rej_rcvd.reason; event.event = RDMA_CM_EVENT_REJECTED; @@ -1794,8 +1816,9 @@ static struct rdma_id_private *cma_new_conn_id(struct rdma_cm_id *listen_id, struct rdma_cm_id *id; struct rdma_route *rt; const sa_family_t ss_family = listen_id->route.addr.src_addr.ss_family; + struct sa_path_rec *path = ib_event->param.req_rcvd.primary_path; const __be64 service_id = - ib_event->param.req_rcvd.primary_path->service_id; + ib_event->param.req_rcvd.primary_path->service_id; int ret; id = rdma_create_id(listen_id->route.addr.dev_addr.net, @@ -1817,7 +1840,7 @@ static struct rdma_id_private *cma_new_conn_id(struct rdma_cm_id *listen_id, if (!rt->path_rec) goto err; - rt->path_rec[0] = *ib_event->param.req_rcvd.primary_path; + rt->path_rec[0] = *path; if (rt->num_paths == 2) rt->path_rec[1] = *ib_event->param.req_rcvd.alternate_path; @@ -2266,10 +2289,11 @@ void rdma_set_service_type(struct rdma_cm_id *id, int tos) id_priv = container_of(id, struct rdma_id_private, id); id_priv->tos = (u8) tos; + id_priv->tos_set = true; } EXPORT_SYMBOL(rdma_set_service_type); -static void cma_query_handler(int status, struct ib_sa_path_rec *path_rec, +static void cma_query_handler(int status, struct sa_path_rec *path_rec, void *context) { struct cma_work *work = context; @@ -2285,6 +2309,8 @@ static void cma_query_handler(int status, struct ib_sa_path_rec *path_rec, work->new_state = RDMA_CM_ADDR_RESOLVED; work->event.event = RDMA_CM_EVENT_ROUTE_ERROR; work->event.status = status; + pr_debug_ratelimited("RDMA CM: ROUTE_ERROR: failed to query path. status %d\n", + status); } queue_work(cma_wq, &work->work); @@ -2294,18 +2320,24 @@ static int cma_query_ib_route(struct rdma_id_private *id_priv, int timeout_ms, struct cma_work *work) { struct rdma_dev_addr *dev_addr = &id_priv->id.route.addr.dev_addr; - struct ib_sa_path_rec path_rec; + struct sa_path_rec path_rec; ib_sa_comp_mask comp_mask; struct sockaddr_in6 *sin6; struct sockaddr_ib *sib; memset(&path_rec, 0, sizeof path_rec); + + if (rdma_cap_opa_ah(id_priv->id.device, id_priv->id.port_num)) + path_rec.rec_type = SA_PATH_REC_TYPE_OPA; + else + path_rec.rec_type = SA_PATH_REC_TYPE_IB; rdma_addr_get_sgid(dev_addr, &path_rec.sgid); rdma_addr_get_dgid(dev_addr, &path_rec.dgid); path_rec.pkey = cpu_to_be16(ib_addr_get_pkey(dev_addr)); path_rec.numb_path = 1; path_rec.reversible = 1; - path_rec.service_id = rdma_get_service_id(&id_priv->id, cma_dst_addr(id_priv)); + path_rec.service_id = rdma_get_service_id(&id_priv->id, + cma_dst_addr(id_priv)); comp_mask = IB_SA_PATH_REC_DGID | IB_SA_PATH_REC_SGID | IB_SA_PATH_REC_PKEY | IB_SA_PATH_REC_NUMB_PATH | @@ -2419,7 +2451,7 @@ err1: } int rdma_set_ib_paths(struct rdma_cm_id *id, - struct ib_sa_path_rec *path_rec, int num_paths) + struct sa_path_rec *path_rec, int num_paths) { struct rdma_id_private *id_priv; int ret; @@ -2467,14 +2499,12 @@ static int iboe_tos_to_sl(struct net_device *ndev, int tos) struct net_device *dev; prio = rt_tos2priority(tos); - dev = ndev->priv_flags & IFF_802_1Q_VLAN ? - vlan_dev_real_dev(ndev) : ndev; - + dev = is_vlan_dev(ndev) ? vlan_dev_real_dev(ndev) : ndev; if (dev->num_tc) return netdev_get_prio_tc_map(dev, prio); #if IS_ENABLED(CONFIG_VLAN_8021Q) - if (ndev->priv_flags & IFF_802_1Q_VLAN) + if (is_vlan_dev(ndev)) return (vlan_dev_get_egress_qos_mask(ndev, prio) & VLAN_PRIO_MASK) >> VLAN_PRIO_SHIFT; #endif @@ -2500,6 +2530,10 @@ static int cma_resolve_iboe_route(struct rdma_id_private *id_priv) struct cma_work *work; int ret; struct net_device *ndev = NULL; + enum ib_gid_type gid_type = IB_GID_TYPE_IB; + 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; work = kzalloc(sizeof *work, GFP_KERNEL); @@ -2526,36 +2560,22 @@ static int cma_resolve_iboe_route(struct rdma_id_private *id_priv) goto err2; } - if (ndev->flags & IFF_LOOPBACK) { - dev_put(ndev); - if (!id_priv->id.device->get_netdev) { - ret = -EOPNOTSUPP; - goto err2; - } - - ndev = id_priv->id.device->get_netdev(id_priv->id.device, - id_priv->id.port_num); - if (!ndev) { - ret = -ENODEV; - goto err2; - } - } - - route->path_rec->net = &init_net; - route->path_rec->ifindex = ndev->ifindex; supported_gids = roce_gid_type_mask_support(id_priv->id.device, id_priv->id.port_num); - route->path_rec->gid_type = - cma_route_gid_type(addr->dev_addr.network, - supported_gids, - id_priv->gid_type); + 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 (!ndev) { ret = -ENODEV; goto err2; } - memcpy(route->path_rec->dmac, addr->dev_addr.dst_dev_addr, ETH_ALEN); + sa_path_set_dmac(route->path_rec, addr->dev_addr.dst_dev_addr); rdma_ip2gid((struct sockaddr *)&id_priv->id.route.addr.src_addr, &route->path_rec->sgid); @@ -2563,8 +2583,10 @@ static int cma_resolve_iboe_route(struct rdma_id_private *id_priv) &route->path_rec->dgid); /* Use the hint from IP Stack to select GID Type */ - if (route->path_rec->gid_type < ib_network_to_gid_type(addr->dev_addr.network)) - route->path_rec->gid_type = ib_network_to_gid_type(addr->dev_addr.network); + 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; @@ -2573,7 +2595,8 @@ static int cma_resolve_iboe_route(struct rdma_id_private *id_priv) route->path_rec->reversible = 1; route->path_rec->pkey = cpu_to_be16(0xffff); route->path_rec->mtu_selector = IB_SA_EQ; - route->path_rec->sl = iboe_tos_to_sl(ndev, id_priv->tos); + route->path_rec->sl = iboe_tos_to_sl(ndev, tos); + route->path_rec->traffic_class = tos; route->path_rec->mtu = iboe_get_mtu(ndev->mtu); route->path_rec->rate_selector = IB_SA_EQ; route->path_rec->rate = iboe_get_rate(ndev); @@ -2652,8 +2675,8 @@ static void cma_set_loopback(struct sockaddr *addr) static int cma_bind_loopback(struct rdma_id_private *id_priv) { struct cma_device *cma_dev, *cur_dev; - struct ib_port_attr port_attr; union ib_gid gid; + enum ib_port_state port_state; u16 pkey; int ret; u8 p; @@ -2669,8 +2692,8 @@ static int cma_bind_loopback(struct rdma_id_private *id_priv) cma_dev = cur_dev; for (p = 1; p <= cur_dev->device->phys_port_cnt; ++p) { - if (!ib_query_port(cur_dev->device, p, &port_attr) && - port_attr.state == IB_PORT_ACTIVE) { + if (!ib_get_cached_port_state(cur_dev->device, p, &port_state) && + port_state == IB_PORT_ACTIVE) { cma_dev = cur_dev; goto port_found; } @@ -2720,8 +2743,14 @@ static void addr_handler(int status, struct sockaddr *src_addr, goto out; memcpy(cma_src_addr(id_priv), src_addr, rdma_addr_size(src_addr)); - if (!status && !id_priv->cma_dev) + if (!status && !id_priv->cma_dev) { status = cma_acquire_dev(id_priv, NULL); + if (status) + pr_debug_ratelimited("RDMA CM: ADDR_ERROR: failed to acquire device. status %d\n", + status); + } else { + pr_debug_ratelimited("RDMA CM: ADDR_ERROR: failed to resolve IP. status %d\n", status); + } if (status) { if (!cma_comp_exch(id_priv, RDMA_CM_ADDR_RESOLVED, @@ -2833,20 +2862,26 @@ int rdma_resolve_addr(struct rdma_cm_id *id, struct sockaddr *src_addr, int ret; id_priv = container_of(id, struct rdma_id_private, id); + memcpy(cma_dst_addr(id_priv), dst_addr, rdma_addr_size(dst_addr)); if (id_priv->state == RDMA_CM_IDLE) { ret = cma_bind_addr(id, src_addr, dst_addr); - if (ret) + if (ret) { + memset(cma_dst_addr(id_priv), 0, rdma_addr_size(dst_addr)); return ret; + } } - if (cma_family(id_priv) != dst_addr->sa_family) + if (cma_family(id_priv) != dst_addr->sa_family) { + memset(cma_dst_addr(id_priv), 0, rdma_addr_size(dst_addr)); return -EINVAL; + } - if (!cma_comp_exch(id_priv, RDMA_CM_ADDR_BOUND, RDMA_CM_ADDR_QUERY)) + if (!cma_comp_exch(id_priv, RDMA_CM_ADDR_BOUND, RDMA_CM_ADDR_QUERY)) { + memset(cma_dst_addr(id_priv), 0, rdma_addr_size(dst_addr)); return -EINVAL; + } atomic_inc(&id_priv->refcount); - memcpy(cma_dst_addr(id_priv), dst_addr, rdma_addr_size(dst_addr)); if (cma_any_addr(dst_addr)) { ret = cma_resolve_loopback(id_priv); } else { @@ -2962,6 +2997,43 @@ err: return ret == -ENOSPC ? -EADDRNOTAVAIL : ret; } +static int cma_port_is_unique(struct rdma_bind_list *bind_list, + struct rdma_id_private *id_priv) +{ + struct rdma_id_private *cur_id; + struct sockaddr *daddr = cma_dst_addr(id_priv); + struct sockaddr *saddr = cma_src_addr(id_priv); + __be16 dport = cma_port(daddr); + + hlist_for_each_entry(cur_id, &bind_list->owners, node) { + struct sockaddr *cur_daddr = cma_dst_addr(cur_id); + struct sockaddr *cur_saddr = cma_src_addr(cur_id); + __be16 cur_dport = cma_port(cur_daddr); + + if (id_priv == cur_id) + continue; + + /* different dest port -> unique */ + if (!cma_any_port(cur_daddr) && + (dport != cur_dport)) + continue; + + /* different src address -> unique */ + if (!cma_any_addr(saddr) && + !cma_any_addr(cur_saddr) && + cma_addr_cmp(saddr, cur_saddr)) + continue; + + /* different dst address -> unique */ + if (!cma_any_addr(cur_daddr) && + cma_addr_cmp(daddr, cur_daddr)) + continue; + + return -EADDRNOTAVAIL; + } + return 0; +} + static int cma_alloc_any_port(enum rdma_port_space ps, struct rdma_id_private *id_priv) { @@ -2974,9 +3046,19 @@ static int cma_alloc_any_port(enum rdma_port_space ps, remaining = (high - low) + 1; rover = prandom_u32() % remaining + low; retry: - if (last_used_port != rover && - !cma_ps_find(net, ps, (unsigned short)rover)) { - int ret = cma_alloc_port(ps, id_priv, rover); + if (last_used_port != rover) { + struct rdma_bind_list *bind_list; + int ret; + + bind_list = cma_ps_find(net, ps, (unsigned short)rover); + + if (!bind_list) { + ret = cma_alloc_port(ps, id_priv, rover); + } else { + ret = cma_port_is_unique(bind_list, id_priv); + if (!ret) + cma_bind_port(bind_list, id_priv); + } /* * Remember previously used port number in order to avoid * re-using same port immediately after it is closed. @@ -3205,6 +3287,7 @@ int rdma_bind_addr(struct rdma_cm_id *id, struct sockaddr *addr) { struct rdma_id_private *id_priv; int ret; + struct sockaddr *daddr; if (addr->sa_family != AF_INET && addr->sa_family != AF_INET6 && addr->sa_family != AF_IB) @@ -3244,6 +3327,9 @@ int rdma_bind_addr(struct rdma_cm_id *id, struct sockaddr *addr) if (ret) goto err2; + daddr = cma_dst_addr(id_priv); + daddr->sa_family = addr->sa_family; + return 0; err2: if (id_priv->cma_dev) @@ -3308,10 +3394,13 @@ static int cma_sidr_rep_handler(struct ib_cm_id *cm_id, if (rep->status != IB_SIDR_SUCCESS) { event.event = RDMA_CM_EVENT_UNREACHABLE; event.status = ib_event->param.sidr_rep_rcvd.status; + pr_debug_ratelimited("RDMA CM: UNREACHABLE: bad SIDR reply. status %d\n", + event.status); break; } ret = cma_set_qkey(id_priv, rep->qkey); if (ret) { + pr_debug_ratelimited("RDMA CM: ADDR_ERROR: failed to set qkey. status %d\n", ret); event.event = RDMA_CM_EVENT_ADDR_ERROR; event.status = ret; break; @@ -3583,6 +3672,9 @@ static int cma_accept_iw(struct rdma_id_private *id_priv, struct iw_cm_conn_param iw_param; int ret; + if (!conn_param) + return -EINVAL; + ret = cma_modify_qp_rtr(id_priv, conn_param); if (ret) return ret; @@ -3760,10 +3852,17 @@ static int cma_ib_mc_handler(int status, struct ib_sa_multicast *multicast) if (!status) status = cma_set_qkey(id_priv, be32_to_cpu(multicast->rec.qkey)); + else + pr_debug_ratelimited("RDMA CM: MULTICAST_ERROR: failed to join multicast. status %d\n", + status); mutex_lock(&id_priv->qp_mutex); - if (!status && id_priv->id.qp) + if (!status && id_priv->id.qp) { status = ib_attach_mcast(id_priv->id.qp, &multicast->rec.mgid, be16_to_cpu(multicast->rec.mlid)); + if (status) + pr_debug_ratelimited("RDMA CM: MULTICAST_ERROR: failed to attach QP. status %d\n", + status); + } mutex_unlock(&id_priv->qp_mutex); memset(&event, 0, sizeof event); @@ -3833,63 +3932,10 @@ static void cma_set_mgid(struct rdma_id_private *id_priv, } } -static void cma_query_sa_classport_info_cb(int status, - struct ib_class_port_info *rec, - void *context) -{ - struct class_port_info_context *cb_ctx = context; - - WARN_ON(!context); - - if (status || !rec) { - pr_debug("RDMA CM: %s port %u failed query ClassPortInfo status: %d\n", - cb_ctx->device->name, cb_ctx->port_num, status); - goto out; - } - - memcpy(cb_ctx->class_port_info, rec, sizeof(struct ib_class_port_info)); - -out: - complete(&cb_ctx->done); -} - -static int cma_query_sa_classport_info(struct ib_device *device, u8 port_num, - struct ib_class_port_info *class_port_info) -{ - struct class_port_info_context *cb_ctx; - int ret; - - cb_ctx = kmalloc(sizeof(*cb_ctx), GFP_KERNEL); - if (!cb_ctx) - return -ENOMEM; - - cb_ctx->device = device; - cb_ctx->class_port_info = class_port_info; - cb_ctx->port_num = port_num; - init_completion(&cb_ctx->done); - - ret = ib_sa_classport_info_rec_query(&sa_client, device, port_num, - CMA_QUERY_CLASSPORT_INFO_TIMEOUT, - GFP_KERNEL, cma_query_sa_classport_info_cb, - cb_ctx, &cb_ctx->sa_query); - if (ret < 0) { - pr_err("RDMA CM: %s port %u failed to send ClassPortInfo query, ret: %d\n", - device->name, port_num, ret); - goto out; - } - - wait_for_completion(&cb_ctx->done); - -out: - kfree(cb_ctx); - return ret; -} - static int cma_join_ib_multicast(struct rdma_id_private *id_priv, struct cma_multicast *mc) { struct ib_sa_mcmember_rec rec; - struct ib_class_port_info class_port_info; struct rdma_dev_addr *dev_addr = &id_priv->id.route.addr.dev_addr; ib_sa_comp_mask comp_mask; int ret; @@ -3910,21 +3956,14 @@ static int cma_join_ib_multicast(struct rdma_id_private *id_priv, rec.pkey = cpu_to_be16(ib_addr_get_pkey(dev_addr)); rec.join_state = mc->join_state; - if (rec.join_state == BIT(SENDONLY_FULLMEMBER_JOIN)) { - ret = cma_query_sa_classport_info(id_priv->id.device, - id_priv->id.port_num, - &class_port_info); - - if (ret) - return ret; - - if (!(ib_get_cpi_capmask2(&class_port_info) & - IB_SA_CAP_MASK2_SENDONLY_FULL_MEM_SUPPORT)) { - pr_warn("RDMA CM: %s port %u Unable to multicast join\n" - "RDMA CM: SM doesn't support Send Only Full Member option\n", - id_priv->id.device->name, id_priv->id.port_num); - return -EOPNOTSUPP; - } + if ((rec.join_state == BIT(SENDONLY_FULLMEMBER_JOIN)) && + (!ib_sa_sendonly_fullmem_support(&sa_client, + id_priv->id.device, + id_priv->id.port_num))) { + pr_warn("RDMA CM: %s port %u Unable to multicast join\n" + "RDMA CM: SM doesn't support Send Only Full Member option\n", + id_priv->id.device->name, id_priv->id.port_num); + return -EOPNOTSUPP; } comp_mask = IB_SA_MCMEMBER_REC_MGID | IB_SA_MCMEMBER_REC_PORT_GID | @@ -4229,15 +4268,21 @@ static void cma_add_one(struct ib_device *device) cma_dev->default_gid_type = kcalloc(device->phys_port_cnt, sizeof(*cma_dev->default_gid_type), GFP_KERNEL); - if (!cma_dev->default_gid_type) { - kfree(cma_dev); - return; - } + if (!cma_dev->default_gid_type) + goto free_cma_dev; + + cma_dev->default_roce_tos = kcalloc(device->phys_port_cnt, + sizeof(*cma_dev->default_roce_tos), + GFP_KERNEL); + if (!cma_dev->default_roce_tos) + goto free_gid_type; + for (i = rdma_start_port(device); i <= rdma_end_port(device); i++) { supported_gids = roce_gid_type_mask_support(device, i); WARN_ON(!supported_gids); cma_dev->default_gid_type[i - rdma_start_port(device)] = find_first_bit(&supported_gids, BITS_PER_LONG); + cma_dev->default_roce_tos[i - rdma_start_port(device)] = 0; } init_completion(&cma_dev->comp); @@ -4250,6 +4295,16 @@ static void cma_add_one(struct ib_device *device) list_for_each_entry(id_priv, &listen_any_list, list) cma_listen_on_dev(id_priv, cma_dev); mutex_unlock(&lock); + + return; + +free_gid_type: + kfree(cma_dev->default_gid_type); + +free_cma_dev: + kfree(cma_dev); + + return; } static int cma_remove_id_dev(struct rdma_id_private *id_priv) @@ -4318,6 +4373,7 @@ static void cma_remove_one(struct ib_device *device, void *client_data) mutex_unlock(&lock); cma_process_remove(cma_dev); + kfree(cma_dev->default_roce_tos); kfree(cma_dev->default_gid_type); kfree(cma_dev); } diff --git a/drivers/infiniband/core/cma_configfs.c b/drivers/infiniband/core/cma_configfs.c index 41573df1d9fc..54076a3e8007 100644 --- a/drivers/infiniband/core/cma_configfs.c +++ b/drivers/infiniband/core/cma_configfs.c @@ -139,8 +139,50 @@ static ssize_t default_roce_mode_store(struct config_item *item, CONFIGFS_ATTR(, default_roce_mode); +static ssize_t default_roce_tos_show(struct config_item *item, char *buf) +{ + struct cma_device *cma_dev; + struct cma_dev_port_group *group; + ssize_t ret; + u8 tos; + + ret = cma_configfs_params_get(item, &cma_dev, &group); + if (ret) + return ret; + + tos = cma_get_default_roce_tos(cma_dev, group->port_num); + cma_configfs_params_put(cma_dev); + + return sprintf(buf, "%u\n", tos); +} + +static ssize_t default_roce_tos_store(struct config_item *item, + const char *buf, size_t count) +{ + struct cma_device *cma_dev; + struct cma_dev_port_group *group; + ssize_t ret; + u8 tos; + + ret = kstrtou8(buf, 0, &tos); + if (ret) + return ret; + + ret = cma_configfs_params_get(item, &cma_dev, &group); + if (ret) + return ret; + + ret = cma_set_default_roce_tos(cma_dev, group->port_num, tos); + cma_configfs_params_put(cma_dev); + + return ret ? ret : strnlen(buf, count); +} + +CONFIGFS_ATTR(, default_roce_tos); + static struct configfs_attribute *cma_configfs_attributes[] = { &attr_default_roce_mode, + &attr_default_roce_tos, NULL, }; diff --git a/drivers/infiniband/core/core_priv.h b/drivers/infiniband/core/core_priv.h index d29372624f3a..11ae67514e13 100644 --- a/drivers/infiniband/core/core_priv.h +++ b/drivers/infiniband/core/core_priv.h @@ -35,8 +35,19 @@ #include <linux/list.h> #include <linux/spinlock.h> +#include <linux/cgroup_rdma.h> #include <rdma/ib_verbs.h> +#include <rdma/ib_mad.h> +#include "mad_priv.h" + +struct pkey_index_qp_list { + struct list_head pkey_index_list; + u16 pkey_index; + /* Lock to hold while iterating the qp_list. */ + spinlock_t qp_list_lock; + struct list_head qp_list; +}; #if IS_ENABLED(CONFIG_INFINIBAND_ADDR_TRANS_CONFIGFS) int cma_configfs_init(void); @@ -62,6 +73,9 @@ int cma_get_default_gid_type(struct cma_device *cma_dev, int cma_set_default_gid_type(struct cma_device *cma_dev, unsigned int port, enum ib_gid_type default_gid_type); +int cma_get_default_roce_tos(struct cma_device *cma_dev, unsigned int port); +int cma_set_default_roce_tos(struct cma_device *a_dev, unsigned int port, + u8 default_roce_tos); struct ib_device *cma_get_ib_dev(struct cma_device *cma_dev); int ib_device_register_sysfs(struct ib_device *device, @@ -121,6 +135,35 @@ int ib_cache_setup_one(struct ib_device *device); void ib_cache_cleanup_one(struct ib_device *device); void ib_cache_release_one(struct ib_device *device); +#ifdef CONFIG_CGROUP_RDMA +int ib_device_register_rdmacg(struct ib_device *device); +void ib_device_unregister_rdmacg(struct ib_device *device); + +int ib_rdmacg_try_charge(struct ib_rdmacg_object *cg_obj, + struct ib_device *device, + enum rdmacg_resource_type resource_index); + +void ib_rdmacg_uncharge(struct ib_rdmacg_object *cg_obj, + struct ib_device *device, + enum rdmacg_resource_type resource_index); +#else +static inline int ib_device_register_rdmacg(struct ib_device *device) +{ return 0; } + +static inline void ib_device_unregister_rdmacg(struct ib_device *device) +{ } + +static inline int ib_rdmacg_try_charge(struct ib_rdmacg_object *cg_obj, + struct ib_device *device, + enum rdmacg_resource_type resource_index) +{ return 0; } + +static inline void ib_rdmacg_uncharge(struct ib_rdmacg_object *cg_obj, + struct ib_device *device, + enum rdmacg_resource_type resource_index) +{ } +#endif + static inline bool rdma_is_upper_dev_rcu(struct net_device *dev, struct net_device *upper) { @@ -136,6 +179,16 @@ void ib_mad_cleanup(void); int ib_sa_init(void); void ib_sa_cleanup(void); +int ibnl_init(void); +void ibnl_cleanup(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 netlink_callback *cb); int ib_nl_handle_set_timeout(struct sk_buff *skb, @@ -143,4 +196,109 @@ int ib_nl_handle_set_timeout(struct sk_buff *skb, int ib_nl_handle_ip_res_resp(struct sk_buff *skb, struct netlink_callback *cb); +int ib_get_cached_subnet_prefix(struct ib_device *device, + u8 port_num, + u64 *sn_pfx); + +#ifdef CONFIG_SECURITY_INFINIBAND +int ib_security_pkey_access(struct ib_device *dev, + u8 port_num, + u16 pkey_index, + void *sec); + +void ib_security_destroy_port_pkey_list(struct ib_device *device); + +void ib_security_cache_change(struct ib_device *device, + u8 port_num, + u64 subnet_prefix); + +int ib_security_modify_qp(struct ib_qp *qp, + struct ib_qp_attr *qp_attr, + int qp_attr_mask, + struct ib_udata *udata); + +int ib_create_qp_security(struct ib_qp *qp, struct ib_device *dev); +void ib_destroy_qp_security_begin(struct ib_qp_security *sec); +void ib_destroy_qp_security_abort(struct ib_qp_security *sec); +void ib_destroy_qp_security_end(struct ib_qp_security *sec); +int ib_open_shared_qp_security(struct ib_qp *qp, struct ib_device *dev); +void ib_close_shared_qp_security(struct ib_qp_security *sec); +int ib_mad_agent_security_setup(struct ib_mad_agent *agent, + enum ib_qp_type qp_type); +void ib_mad_agent_security_cleanup(struct ib_mad_agent *agent); +int ib_mad_enforce_security(struct ib_mad_agent_private *map, u16 pkey_index); +#else +static inline int ib_security_pkey_access(struct ib_device *dev, + u8 port_num, + u16 pkey_index, + void *sec) +{ + return 0; +} + +static inline void ib_security_destroy_port_pkey_list(struct ib_device *device) +{ +} + +static inline void ib_security_cache_change(struct ib_device *device, + u8 port_num, + u64 subnet_prefix) +{ +} + +static inline int ib_security_modify_qp(struct ib_qp *qp, + struct ib_qp_attr *qp_attr, + int qp_attr_mask, + struct ib_udata *udata) +{ + return qp->device->modify_qp(qp->real_qp, + qp_attr, + qp_attr_mask, + udata); +} + +static inline int ib_create_qp_security(struct ib_qp *qp, + struct ib_device *dev) +{ + return 0; +} + +static inline void ib_destroy_qp_security_begin(struct ib_qp_security *sec) +{ +} + +static inline void ib_destroy_qp_security_abort(struct ib_qp_security *sec) +{ +} + +static inline void ib_destroy_qp_security_end(struct ib_qp_security *sec) +{ +} + +static inline int ib_open_shared_qp_security(struct ib_qp *qp, + struct ib_device *dev) +{ + return 0; +} + +static inline void ib_close_shared_qp_security(struct ib_qp_security *sec) +{ +} + +static inline int ib_mad_agent_security_setup(struct ib_mad_agent *agent, + enum ib_qp_type qp_type) +{ + return 0; +} + +static inline void ib_mad_agent_security_cleanup(struct ib_mad_agent *agent) +{ +} + +static inline int ib_mad_enforce_security(struct ib_mad_agent_private *map, + u16 pkey_index) +{ + return 0; +} +#endif #endif /* _CORE_PRIV_H */ diff --git a/drivers/infiniband/core/cq.c b/drivers/infiniband/core/cq.c index a754fc727de5..f2ae75fa3128 100644 --- a/drivers/infiniband/core/cq.c +++ b/drivers/infiniband/core/cq.c @@ -29,7 +29,13 @@ static int __ib_process_cq(struct ib_cq *cq, int budget) { int i, n, completed = 0; - while ((n = ib_poll_cq(cq, IB_POLL_BATCH, cq->wc)) > 0) { + /* + * budget might be (-1) if the caller does not + * want to bound this call, thus we need unsigned + * minimum here. + */ + while ((n = ib_poll_cq(cq, min_t(u32, IB_POLL_BATCH, + budget - completed), cq->wc)) > 0) { for (i = 0; i < n; i++) { struct ib_wc *wc = &cq->wc[i]; @@ -58,8 +64,8 @@ static int __ib_process_cq(struct ib_cq *cq, int budget) * %IB_POLL_DIRECT CQ. It does not offload CQ processing to a different * context and does not ask for completion interrupts from the HCA. * - * Note: for compatibility reasons -1 can be passed in %budget for unlimited - * polling. Do not use this feature in new code, it will be removed soon. + * Note: do not pass -1 as %budget unless it is guaranteed that the number + * of completions that will be processed is small. */ int ib_process_cq_direct(struct ib_cq *cq, int budget) { @@ -120,7 +126,7 @@ static void ib_cq_completion_workqueue(struct ib_cq *cq, void *private) * * This is the proper interface to allocate a CQ for in-kernel users. A * CQ allocated with this interface will automatically be polled from the - * specified context. The ULP needs must use wr->wr_cqe instead of wr->wr_id + * specified context. The ULP must use wr->wr_cqe instead of wr->wr_id * to use this CQ abstraction. */ struct ib_cq *ib_alloc_cq(struct ib_device *dev, void *private, @@ -196,7 +202,7 @@ void ib_free_cq(struct ib_cq *cq) irq_poll_disable(&cq->iop); break; case IB_POLL_WORKQUEUE: - flush_work(&cq->work); + cancel_work_sync(&cq->work); break; default: WARN_ON_ONCE(1); diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c index 571974cd3919..a5dfab6adf49 100644 --- a/drivers/infiniband/core/device.c +++ b/drivers/infiniband/core/device.c @@ -39,6 +39,8 @@ #include <linux/init.h> #include <linux/mutex.h> #include <linux/netdevice.h> +#include <linux/security.h> +#include <linux/notifier.h> #include <rdma/rdma_netlink.h> #include <rdma/ib_addr.h> #include <rdma/ib_cache.h> @@ -82,6 +84,14 @@ static LIST_HEAD(client_list); static DEFINE_MUTEX(device_mutex); static DECLARE_RWSEM(lists_rwsem); +static int ib_security_change(struct notifier_block *nb, unsigned long event, + void *lsm_data); +static void ib_policy_change_task(struct work_struct *work); +static DECLARE_WORK(ib_policy_change_work, ib_policy_change_task); + +static struct notifier_block ibdev_lsm_nb = { + .notifier_call = ib_security_change, +}; static int ib_device_check_mandatory(struct ib_device *device) { @@ -172,8 +182,16 @@ static void ib_device_release(struct device *device) { struct ib_device *dev = container_of(device, struct ib_device, dev); - ib_cache_release_one(dev); - kfree(dev->port_immutable); + WARN_ON(dev->reg_state == IB_DEV_REGISTERED); + if (dev->reg_state == IB_DEV_UNREGISTERED) { + /* + * In IB_DEV_UNINITIALIZED state, cache or port table + * is not even created. Free cache and port table only when + * device reaches UNREGISTERED state. + */ + ib_cache_release_one(dev); + kfree(dev->port_immutable); + } kfree(dev); } @@ -317,6 +335,65 @@ void ib_get_device_fw_str(struct ib_device *dev, char *str, size_t str_len) } EXPORT_SYMBOL(ib_get_device_fw_str); +static int setup_port_pkey_list(struct ib_device *device) +{ + int i; + + /** + * device->port_pkey_list is indexed directly by the port number, + * Therefore it is declared as a 1 based array with potential empty + * slots at the beginning. + */ + device->port_pkey_list = kcalloc(rdma_end_port(device) + 1, + sizeof(*device->port_pkey_list), + GFP_KERNEL); + + if (!device->port_pkey_list) + return -ENOMEM; + + for (i = 0; i < (rdma_end_port(device) + 1); i++) { + spin_lock_init(&device->port_pkey_list[i].list_lock); + INIT_LIST_HEAD(&device->port_pkey_list[i].pkey_list); + } + + return 0; +} + +static void ib_policy_change_task(struct work_struct *work) +{ + struct ib_device *dev; + + down_read(&lists_rwsem); + list_for_each_entry(dev, &device_list, core_list) { + int i; + + for (i = rdma_start_port(dev); i <= rdma_end_port(dev); i++) { + u64 sp; + int ret = ib_get_cached_subnet_prefix(dev, + i, + &sp); + + WARN_ONCE(ret, + "ib_get_cached_subnet_prefix err: %d, this should never happen here\n", + ret); + if (!ret) + ib_security_cache_change(dev, i, sp); + } + } + up_read(&lists_rwsem); +} + +static int ib_security_change(struct notifier_block *nb, unsigned long event, + void *lsm_data) +{ + if (event != LSM_POLICY_CHANGE) + return NOTIFY_DONE; + + schedule_work(&ib_policy_change_work); + + return NOTIFY_OK; +} + /** * ib_register_device - Register an IB device with IB core * @device:Device to register @@ -333,6 +410,29 @@ int ib_register_device(struct ib_device *device, int ret; struct ib_client *client; 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) { + /* + * The caller provided custom DMA operations. Copy the + * DMA-related fields that are used by e.g. dma_alloc_coherent() + * 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; + } else { + /* + * The caller did not provide custom DMA operations. Use the + * DMA mapping operations of the parent device. + */ + device->dma_device = parent; + } mutex_lock(&device_mutex); @@ -354,26 +454,36 @@ int ib_register_device(struct ib_device *device, goto out; } + ret = setup_port_pkey_list(device); + if (ret) { + pr_warn("Couldn't create per port_pkey_list\n"); + goto out; + } + ret = ib_cache_setup_one(device); if (ret) { pr_warn("Couldn't set up InfiniBand P_Key/GID cache\n"); - goto out; + goto port_cleanup; + } + + ret = ib_device_register_rdmacg(device); + if (ret) { + pr_warn("Couldn't register device with rdma cgroup\n"); + goto cache_cleanup; } memset(&device->attrs, 0, sizeof(device->attrs)); ret = device->query_device(device, &device->attrs, &uhw); if (ret) { pr_warn("Couldn't query the device attributes\n"); - ib_cache_cleanup_one(device); - goto out; + goto cache_cleanup; } ret = ib_device_register_sysfs(device, port_callback); if (ret) { pr_warn("Couldn't register device %s with driver model\n", device->name); - ib_cache_cleanup_one(device); - goto out; + goto cache_cleanup; } device->reg_state = IB_DEV_REGISTERED; @@ -385,6 +495,14 @@ int ib_register_device(struct ib_device *device, down_write(&lists_rwsem); list_add_tail(&device->core_list, &device_list); up_write(&lists_rwsem); + mutex_unlock(&device_mutex); + return 0; + +cache_cleanup: + ib_cache_cleanup_one(device); + ib_cache_release_one(device); +port_cleanup: + kfree(device->port_immutable); out: mutex_unlock(&device_mutex); return ret; @@ -421,9 +539,13 @@ void ib_unregister_device(struct ib_device *device) mutex_unlock(&device_mutex); + ib_device_unregister_rdmacg(device); ib_device_unregister_sysfs(device); ib_cache_cleanup_one(device); + ib_security_destroy_port_pkey_list(device); + kfree(device->port_pkey_list); + down_write(&lists_rwsem); spin_lock_irqsave(&device->client_data_lock, flags); list_for_each_entry_safe(context, tmp, &device->client_data_list, list) @@ -659,7 +781,7 @@ int ib_query_port(struct ib_device *device, union ib_gid gid; int err; - if (port_num < rdma_start_port(device) || port_num > rdma_end_port(device)) + if (!rdma_is_port_valid(device, port_num)) return -EINVAL; memset(port_attr, 0, sizeof(*port_attr)); @@ -825,7 +947,7 @@ int ib_modify_port(struct ib_device *device, if (!device->modify_port) return -ENOSYS; - if (port_num < rdma_start_port(device) || port_num > rdma_end_port(device)) + if (!rdma_is_port_valid(device, port_num)) return -EINVAL; return device->modify_port(device, port_num, port_modify_mask, @@ -996,8 +1118,7 @@ static int __init ib_core_init(void) return -ENOMEM; ib_comp_wq = alloc_workqueue("ib-comp-wq", - WQ_UNBOUND | WQ_HIGHPRI | WQ_MEM_RECLAIM, - WQ_UNBOUND_MAX_ACTIVE); + WQ_HIGHPRI | WQ_MEM_RECLAIM | WQ_SYSFS, 0); if (!ib_comp_wq) { ret = -ENOMEM; goto err; @@ -1039,10 +1160,18 @@ static int __init ib_core_init(void) goto err_sa; } + ret = register_lsm_notifier(&ibdev_lsm_nb); + if (ret) { + pr_warn("Couldn't register LSM notifier. ret %d\n", ret); + goto err_ibnl_clients; + } + ib_cache_setup(); return 0; +err_ibnl_clients: + ib_remove_ibnl_clients(); err_sa: ib_sa_cleanup(); err_mad: @@ -1062,6 +1191,7 @@ err: static void __exit ib_core_cleanup(void) { + unregister_lsm_notifier(&ibdev_lsm_nb); ib_cache_cleanup(); ib_remove_ibnl_clients(); ib_sa_cleanup(); diff --git a/drivers/infiniband/core/fmr_pool.c b/drivers/infiniband/core/fmr_pool.c index cdfad5f26212..84d2615b5d4b 100644 --- a/drivers/infiniband/core/fmr_pool.c +++ b/drivers/infiniband/core/fmr_pool.c @@ -96,7 +96,8 @@ struct ib_fmr_pool { void * arg); void *flush_arg; - struct task_struct *thread; + struct kthread_worker *worker; + struct kthread_work work; atomic_t req_ser; atomic_t flush_ser; @@ -174,29 +175,19 @@ static void ib_fmr_batch_release(struct ib_fmr_pool *pool) spin_unlock_irq(&pool->pool_lock); } -static int ib_fmr_cleanup_thread(void *pool_ptr) +static void ib_fmr_cleanup_func(struct kthread_work *work) { - struct ib_fmr_pool *pool = pool_ptr; + struct ib_fmr_pool *pool = container_of(work, struct ib_fmr_pool, work); - do { - if (atomic_read(&pool->flush_ser) - atomic_read(&pool->req_ser) < 0) { - ib_fmr_batch_release(pool); - - atomic_inc(&pool->flush_ser); - wake_up_interruptible(&pool->force_wait); - - if (pool->flush_function) - pool->flush_function(pool, pool->flush_arg); - } + ib_fmr_batch_release(pool); + atomic_inc(&pool->flush_ser); + wake_up_interruptible(&pool->force_wait); - set_current_state(TASK_INTERRUPTIBLE); - if (atomic_read(&pool->flush_ser) - atomic_read(&pool->req_ser) >= 0 && - !kthread_should_stop()) - schedule(); - __set_current_state(TASK_RUNNING); - } while (!kthread_should_stop()); + if (pool->flush_function) + pool->flush_function(pool, pool->flush_arg); - return 0; + if (atomic_read(&pool->flush_ser) - atomic_read(&pool->req_ser) < 0) + kthread_queue_work(pool->worker, &pool->work); } /** @@ -265,15 +256,13 @@ struct ib_fmr_pool *ib_create_fmr_pool(struct ib_pd *pd, atomic_set(&pool->flush_ser, 0); init_waitqueue_head(&pool->force_wait); - pool->thread = kthread_run(ib_fmr_cleanup_thread, - pool, - "ib_fmr(%s)", - device->name); - if (IS_ERR(pool->thread)) { - pr_warn(PFX "couldn't start cleanup thread\n"); - ret = PTR_ERR(pool->thread); + pool->worker = kthread_create_worker(0, "ib_fmr(%s)", device->name); + if (IS_ERR(pool->worker)) { + pr_warn(PFX "couldn't start cleanup kthread worker\n"); + ret = PTR_ERR(pool->worker); goto out_free_pool; } + kthread_init_work(&pool->work, ib_fmr_cleanup_func); { struct ib_pool_fmr *fmr; @@ -338,7 +327,7 @@ void ib_destroy_fmr_pool(struct ib_fmr_pool *pool) LIST_HEAD(fmr_list); int i; - kthread_stop(pool->thread); + kthread_destroy_worker(pool->worker); ib_fmr_batch_release(pool); i = 0; @@ -388,7 +377,7 @@ int ib_flush_fmr_pool(struct ib_fmr_pool *pool) spin_unlock_irq(&pool->pool_lock); serial = atomic_inc_return(&pool->req_ser); - wake_up_process(pool->thread); + kthread_queue_work(pool->worker, &pool->work); if (wait_event_interruptible(pool->force_wait, atomic_read(&pool->flush_ser) - serial >= 0)) @@ -502,7 +491,7 @@ int ib_fmr_pool_unmap(struct ib_pool_fmr *fmr) list_add_tail(&fmr->list, &pool->dirty_list); if (++pool->dirty_len >= pool->dirty_watermark) { atomic_inc(&pool->req_ser); - wake_up_process(pool->thread); + kthread_queue_work(pool->worker, &pool->work); } } } diff --git a/drivers/infiniband/core/iwpm_util.c b/drivers/infiniband/core/iwpm_util.c index 3ef51a96bbf1..f13870e69ccd 100644 --- a/drivers/infiniband/core/iwpm_util.c +++ b/drivers/infiniband/core/iwpm_util.c @@ -472,12 +472,14 @@ int iwpm_parse_nlmsg(struct netlink_callback *cb, int policy_max, int ret; const char *err_str = ""; - ret = nlmsg_validate(cb->nlh, nlh_len, policy_max-1, nlmsg_policy); + ret = nlmsg_validate(cb->nlh, nlh_len, policy_max - 1, nlmsg_policy, + NULL); if (ret) { err_str = "Invalid attribute"; goto parse_nlmsg_error; } - ret = nlmsg_parse(cb->nlh, nlh_len, nltb, policy_max-1, nlmsg_policy); + ret = nlmsg_parse(cb->nlh, nlh_len, nltb, policy_max - 1, + nlmsg_policy, NULL); if (ret) { err_str = "Unable to parse the nlmsg"; goto parse_nlmsg_error; diff --git a/drivers/infiniband/core/mad.c b/drivers/infiniband/core/mad.c index a009f7132c73..f8f53bb90837 100644 --- a/drivers/infiniband/core/mad.c +++ b/drivers/infiniband/core/mad.c @@ -40,9 +40,11 @@ #include <linux/dma-mapping.h> #include <linux/slab.h> #include <linux/module.h> +#include <linux/security.h> #include <rdma/ib_cache.h> #include "mad_priv.h" +#include "core_priv.h" #include "mad_rmpp.h" #include "smi.h" #include "opa_smi.h" @@ -316,7 +318,9 @@ struct ib_mad_agent *ib_register_mad_agent(struct ib_device *device, /* Validate device and port */ port_priv = ib_get_mad_port(device, port_num); if (!port_priv) { - dev_notice(&device->dev, "ib_register_mad_agent: Invalid port\n"); + dev_notice(&device->dev, + "ib_register_mad_agent: Invalid port %d\n", + port_num); ret = ERR_PTR(-ENODEV); goto error1; } @@ -367,6 +371,12 @@ struct ib_mad_agent *ib_register_mad_agent(struct ib_device *device, atomic_set(&mad_agent_priv->refcount, 1); init_completion(&mad_agent_priv->comp); + ret2 = ib_mad_agent_security_setup(&mad_agent_priv->agent, qp_type); + if (ret2) { + ret = ERR_PTR(ret2); + goto error4; + } + spin_lock_irqsave(&port_priv->reg_lock, flags); mad_agent_priv->agent.hi_tid = ++ib_mad_client_id; @@ -384,7 +394,7 @@ struct ib_mad_agent *ib_register_mad_agent(struct ib_device *device, if (method) { if (method_in_use(&method, mad_reg_req)) - goto error4; + goto error5; } } ret2 = add_nonoui_reg_req(mad_reg_req, mad_agent_priv, @@ -400,14 +410,14 @@ struct ib_mad_agent *ib_register_mad_agent(struct ib_device *device, if (is_vendor_method_in_use( vendor_class, mad_reg_req)) - goto error4; + goto error5; } } ret2 = add_oui_reg_req(mad_reg_req, mad_agent_priv); } if (ret2) { ret = ERR_PTR(ret2); - goto error4; + goto error5; } } @@ -416,9 +426,10 @@ struct ib_mad_agent *ib_register_mad_agent(struct ib_device *device, spin_unlock_irqrestore(&port_priv->reg_lock, flags); return &mad_agent_priv->agent; - -error4: +error5: spin_unlock_irqrestore(&port_priv->reg_lock, flags); + ib_mad_agent_security_cleanup(&mad_agent_priv->agent); +error4: kfree(reg_req); error3: kfree(mad_agent_priv); @@ -489,6 +500,7 @@ struct ib_mad_agent *ib_register_mad_snoop(struct ib_device *device, struct ib_mad_agent *ret; struct ib_mad_snoop_private *mad_snoop_priv; int qpn; + int err; /* Validate parameters */ if ((is_snooping_sends(mad_snoop_flags) && !snoop_handler) || @@ -523,17 +535,25 @@ struct ib_mad_agent *ib_register_mad_snoop(struct ib_device *device, mad_snoop_priv->agent.port_num = port_num; mad_snoop_priv->mad_snoop_flags = mad_snoop_flags; init_completion(&mad_snoop_priv->comp); + + err = ib_mad_agent_security_setup(&mad_snoop_priv->agent, qp_type); + if (err) { + ret = ERR_PTR(err); + goto error2; + } + mad_snoop_priv->snoop_index = register_snoop_agent( &port_priv->qp_info[qpn], mad_snoop_priv); if (mad_snoop_priv->snoop_index < 0) { ret = ERR_PTR(mad_snoop_priv->snoop_index); - goto error2; + goto error3; } atomic_set(&mad_snoop_priv->refcount, 1); return &mad_snoop_priv->agent; - +error3: + ib_mad_agent_security_cleanup(&mad_snoop_priv->agent); error2: kfree(mad_snoop_priv); error1: @@ -579,6 +599,8 @@ static void unregister_mad_agent(struct ib_mad_agent_private *mad_agent_priv) deref_mad_agent(mad_agent_priv); wait_for_completion(&mad_agent_priv->comp); + ib_mad_agent_security_cleanup(&mad_agent_priv->agent); + kfree(mad_agent_priv->reg_req); kfree(mad_agent_priv); } @@ -597,13 +619,15 @@ static void unregister_mad_snoop(struct ib_mad_snoop_private *mad_snoop_priv) deref_snoop_agent(mad_snoop_priv); wait_for_completion(&mad_snoop_priv->comp); + ib_mad_agent_security_cleanup(&mad_snoop_priv->agent); + kfree(mad_snoop_priv); } /* * ib_unregister_mad_agent - Unregisters a client from using MAD services */ -int ib_unregister_mad_agent(struct ib_mad_agent *mad_agent) +void ib_unregister_mad_agent(struct ib_mad_agent *mad_agent) { struct ib_mad_agent_private *mad_agent_priv; struct ib_mad_snoop_private *mad_snoop_priv; @@ -620,7 +644,6 @@ int ib_unregister_mad_agent(struct ib_mad_agent *mad_agent) agent); unregister_mad_snoop(mad_snoop_priv); } - return 0; } EXPORT_SYMBOL(ib_unregister_mad_agent); @@ -1214,12 +1237,16 @@ int ib_post_send_mad(struct ib_mad_send_buf *send_buf, /* Walk list of send WRs and post each on send list */ for (; send_buf; send_buf = next_send_buf) { - mad_send_wr = container_of(send_buf, struct ib_mad_send_wr_private, send_buf); mad_agent_priv = mad_send_wr->mad_agent_priv; + ret = ib_mad_enforce_security(mad_agent_priv, + mad_send_wr->send_wr.pkey_index); + if (ret) + goto error; + if (!send_buf->mad_agent->send_handler || (send_buf->timeout_ms && !send_buf->mad_agent->recv_handler)) { @@ -1832,12 +1859,13 @@ static inline int rcv_has_same_gid(const struct ib_mad_agent_private *mad_agent_ const struct ib_mad_send_wr_private *wr, const struct ib_mad_recv_wc *rwc ) { - struct ib_ah_attr attr; + struct rdma_ah_attr attr; u8 send_resp, rcv_resp; union ib_gid sgid; struct ib_device *device = mad_agent_priv->agent.device; u8 port_num = mad_agent_priv->agent.port_num; u8 lmc; + bool has_grh; send_resp = ib_response_mad((struct ib_mad_hdr *)wr->send_buf.mad); rcv_resp = ib_response_mad(&rwc->recv_buf.mad->mad_hdr); @@ -1846,36 +1874,40 @@ static inline int rcv_has_same_gid(const struct ib_mad_agent_private *mad_agent_ /* both requests, or both responses. GIDs different */ return 0; - if (ib_query_ah(wr->send_buf.ah, &attr)) + if (rdma_query_ah(wr->send_buf.ah, &attr)) /* Assume not equal, to avoid false positives. */ return 0; - if (!!(attr.ah_flags & IB_AH_GRH) != - !!(rwc->wc->wc_flags & IB_WC_GRH)) + has_grh = !!(rdma_ah_get_ah_flags(&attr) & IB_AH_GRH); + if (has_grh != !!(rwc->wc->wc_flags & IB_WC_GRH)) /* one has GID, other does not. Assume different */ return 0; if (!send_resp && rcv_resp) { /* is request/response. */ - if (!(attr.ah_flags & IB_AH_GRH)) { + if (!has_grh) { if (ib_get_cached_lmc(device, port_num, &lmc)) return 0; - return (!lmc || !((attr.src_path_bits ^ + return (!lmc || !((rdma_ah_get_path_bits(&attr) ^ rwc->wc->dlid_path_bits) & ((1 << lmc) - 1))); } else { + const struct ib_global_route *grh = + rdma_ah_read_grh(&attr); + if (ib_get_cached_gid(device, port_num, - attr.grh.sgid_index, &sgid, NULL)) + grh->sgid_index, &sgid, NULL)) return 0; return !memcmp(sgid.raw, rwc->recv_buf.grh->dgid.raw, 16); } } - if (!(attr.ah_flags & IB_AH_GRH)) - return attr.dlid == rwc->wc->slid; + if (!has_grh) + return rdma_ah_get_dlid(&attr) == rwc->wc->slid; else - return !memcmp(attr.grh.dgid.raw, rwc->recv_buf.grh->sgid.raw, + return !memcmp(rdma_ah_read_grh(&attr)->dgid.raw, + rwc->recv_buf.grh->sgid.raw, 16); } @@ -1940,6 +1972,14 @@ static void ib_mad_complete_recv(struct ib_mad_agent_private *mad_agent_priv, struct ib_mad_send_wr_private *mad_send_wr; struct ib_mad_send_wc mad_send_wc; unsigned long flags; + int ret; + + ret = ib_mad_enforce_security(mad_agent_priv, + mad_recv_wc->wc->pkey_index); + if (ret) { + ib_free_recv_mad(mad_recv_wc); + deref_mad_agent(mad_agent_priv); + } INIT_LIST_HEAD(&mad_recv_wc->rmpp_list); list_add(&mad_recv_wc->recv_buf.list, &mad_recv_wc->rmpp_list); @@ -1997,6 +2037,8 @@ static void ib_mad_complete_recv(struct ib_mad_agent_private *mad_agent_priv, mad_recv_wc); deref_mad_agent(mad_agent_priv); } + + return; } static enum smi_action handle_ib_smi(const struct ib_mad_port_private *port_priv, diff --git a/drivers/infiniband/core/mad_rmpp.c b/drivers/infiniband/core/mad_rmpp.c index 382941b46e43..0d3cca0a8890 100644 --- a/drivers/infiniband/core/mad_rmpp.c +++ b/drivers/infiniband/core/mad_rmpp.c @@ -81,7 +81,7 @@ static void destroy_rmpp_recv(struct mad_rmpp_recv *rmpp_recv) { deref_rmpp_recv(rmpp_recv); wait_for_completion(&rmpp_recv->comp); - ib_destroy_ah(rmpp_recv->ah); + rdma_destroy_ah(rmpp_recv->ah); kfree(rmpp_recv); } @@ -171,7 +171,7 @@ static struct ib_mad_send_buf *alloc_response_msg(struct ib_mad_agent *agent, hdr_len, 0, GFP_KERNEL, IB_MGMT_BASE_VERSION); if (IS_ERR(msg)) - ib_destroy_ah(ah); + rdma_destroy_ah(ah); else { msg->ah = ah; msg->context[0] = ah; @@ -201,7 +201,7 @@ static void ack_ds_ack(struct ib_mad_agent_private *agent, ret = ib_post_send_mad(msg, NULL); if (ret) { - ib_destroy_ah(msg->ah); + rdma_destroy_ah(msg->ah); ib_free_send_mad(msg); } } @@ -209,7 +209,7 @@ static void ack_ds_ack(struct ib_mad_agent_private *agent, void ib_rmpp_send_handler(struct ib_mad_send_wc *mad_send_wc) { if (mad_send_wc->send_buf->context[0] == mad_send_wc->send_buf->ah) - ib_destroy_ah(mad_send_wc->send_buf->ah); + rdma_destroy_ah(mad_send_wc->send_buf->ah); ib_free_send_mad(mad_send_wc->send_buf); } @@ -237,7 +237,7 @@ static void nack_recv(struct ib_mad_agent_private *agent, ret = ib_post_send_mad(msg, NULL); if (ret) { - ib_destroy_ah(msg->ah); + rdma_destroy_ah(msg->ah); ib_free_send_mad(msg); } } @@ -852,7 +852,7 @@ static int init_newwin(struct ib_mad_send_wr_private *mad_send_wr) struct ib_mad_agent_private *agent = mad_send_wr->mad_agent_priv; struct ib_mad_hdr *mad_hdr = mad_send_wr->send_buf.mad; struct mad_rmpp_recv *rmpp_recv; - struct ib_ah_attr ah_attr; + struct rdma_ah_attr ah_attr; unsigned long flags; int newwin = 1; @@ -867,10 +867,10 @@ static int init_newwin(struct ib_mad_send_wr_private *mad_send_wr) (rmpp_recv->method & IB_MGMT_METHOD_RESP)) continue; - if (ib_query_ah(mad_send_wr->send_buf.ah, &ah_attr)) + if (rdma_query_ah(mad_send_wr->send_buf.ah, &ah_attr)) continue; - if (rmpp_recv->slid == ah_attr.dlid) { + if (rmpp_recv->slid == rdma_ah_get_dlid(&ah_attr)) { newwin = rmpp_recv->repwin; break; } diff --git a/drivers/infiniband/core/multicast.c b/drivers/infiniband/core/multicast.c index 322cb67b07a9..45f2f095f793 100644 --- a/drivers/infiniband/core/multicast.c +++ b/drivers/infiniband/core/multicast.c @@ -720,7 +720,7 @@ int ib_init_ah_from_mcmember(struct ib_device *device, u8 port_num, struct ib_sa_mcmember_rec *rec, struct net_device *ndev, enum ib_gid_type gid_type, - struct ib_ah_attr *ah_attr) + struct rdma_ah_attr *ah_attr) { int ret; u16 gid_index; @@ -743,19 +743,18 @@ int ib_init_ah_from_mcmember(struct ib_device *device, u8 port_num, return ret; memset(ah_attr, 0, sizeof *ah_attr); - ah_attr->dlid = be16_to_cpu(rec->mlid); - ah_attr->sl = rec->sl; - ah_attr->port_num = port_num; - ah_attr->static_rate = rec->rate; - - ah_attr->ah_flags = IB_AH_GRH; - ah_attr->grh.dgid = rec->mgid; - - ah_attr->grh.sgid_index = (u8) gid_index; - ah_attr->grh.flow_label = be32_to_cpu(rec->flow_label); - ah_attr->grh.hop_limit = rec->hop_limit; - ah_attr->grh.traffic_class = rec->traffic_class; - + ah_attr->type = rdma_ah_find_type(device, port_num); + + rdma_ah_set_dlid(ah_attr, be16_to_cpu(rec->mlid)); + rdma_ah_set_sl(ah_attr, rec->sl); + rdma_ah_set_port_num(ah_attr, port_num); + rdma_ah_set_static_rate(ah_attr, rec->rate); + + rdma_ah_set_grh(ah_attr, &rec->mgid, + be32_to_cpu(rec->flow_label), + (u8)gid_index, + rec->hop_limit, + rec->traffic_class); return 0; } EXPORT_SYMBOL(ib_init_ah_from_mcmember); diff --git a/drivers/infiniband/core/netlink.c b/drivers/infiniband/core/netlink.c index 10469b0088b5..94931c474d41 100644 --- a/drivers/infiniband/core/netlink.c +++ b/drivers/infiniband/core/netlink.c @@ -37,6 +37,7 @@ #include <net/net_namespace.h> #include <net/sock.h> #include <rdma/rdma_netlink.h> +#include "core_priv.h" struct ibnl_client { struct list_head list; @@ -55,7 +56,6 @@ int ibnl_chk_listeners(unsigned int group) return -1; return 0; } -EXPORT_SYMBOL(ibnl_chk_listeners); int ibnl_add_client(int index, int nops, const struct ibnl_client_cbs cb_table[]) @@ -146,7 +146,8 @@ nla_put_failure: } EXPORT_SYMBOL(ibnl_put_attr); -static int ibnl_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh) +static int ibnl_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh, + struct netlink_ext_ack *extack) { struct ibnl_client *client; int type = nlh->nlmsg_type; @@ -209,7 +210,7 @@ static void ibnl_rcv_reply_skb(struct sk_buff *skb) if (nlh->nlmsg_flags & NLM_F_REQUEST) return; - ibnl_rcv_msg(skb, nlh); + ibnl_rcv_msg(skb, nlh, NULL); msglen = NLMSG_ALIGN(nlh->nlmsg_len); if (msglen > skb->len) diff --git a/drivers/infiniband/core/rdma_core.c b/drivers/infiniband/core/rdma_core.c new file mode 100644 index 000000000000..41c31a2bf093 --- /dev/null +++ b/drivers/infiniband/core/rdma_core.c @@ -0,0 +1,627 @@ +/* + * Copyright (c) 2016, Mellanox Technologies inc. All rights reserved. + * + * 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/file.h> +#include <linux/anon_inodes.h> +#include <rdma/ib_verbs.h> +#include <rdma/uverbs_types.h> +#include <linux/rcupdate.h> +#include "uverbs.h" +#include "core_priv.h" +#include "rdma_core.h" + +void uverbs_uobject_get(struct ib_uobject *uobject) +{ + kref_get(&uobject->ref); +} + +static void uverbs_uobject_free(struct kref *ref) +{ + struct ib_uobject *uobj = + container_of(ref, struct ib_uobject, ref); + + if (uobj->type->type_class->needs_kfree_rcu) + kfree_rcu(uobj, rcu); + else + kfree(uobj); +} + +void uverbs_uobject_put(struct ib_uobject *uobject) +{ + kref_put(&uobject->ref, uverbs_uobject_free); +} + +static int uverbs_try_lock_object(struct ib_uobject *uobj, bool exclusive) +{ + /* + * When a shared access is required, we use a positive counter. Each + * shared access request checks that the value != -1 and increment it. + * Exclusive access is required for operations like write or destroy. + * In exclusive access mode, we check that the counter is zero (nobody + * claimed this object) and we set it to -1. Releasing a shared access + * lock is done simply by decreasing the counter. As for exclusive + * access locks, since only a single one of them is is allowed + * concurrently, setting the counter to zero is enough for releasing + * this lock. + */ + if (!exclusive) + return __atomic_add_unless(&uobj->usecnt, 1, -1) == -1 ? + -EBUSY : 0; + + /* lock is either WRITE or DESTROY - should be exclusive */ + return atomic_cmpxchg(&uobj->usecnt, 0, -1) == 0 ? 0 : -EBUSY; +} + +static struct ib_uobject *alloc_uobj(struct ib_ucontext *context, + const struct uverbs_obj_type *type) +{ + struct ib_uobject *uobj = kzalloc(type->obj_size, GFP_KERNEL); + + if (!uobj) + return ERR_PTR(-ENOMEM); + /* + * user_handle should be filled by the handler, + * The object is added to the list in the commit stage. + */ + uobj->context = context; + uobj->type = type; + atomic_set(&uobj->usecnt, 0); + kref_init(&uobj->ref); + + return uobj; +} + +static int idr_add_uobj(struct ib_uobject *uobj) +{ + int ret; + + idr_preload(GFP_KERNEL); + spin_lock(&uobj->context->ufile->idr_lock); + + /* + * We start with allocating an idr pointing to NULL. This represents an + * object which isn't initialized yet. We'll replace it later on with + * the real object once we commit. + */ + ret = idr_alloc(&uobj->context->ufile->idr, NULL, 0, + min_t(unsigned long, U32_MAX - 1, INT_MAX), GFP_NOWAIT); + if (ret >= 0) + uobj->id = ret; + + spin_unlock(&uobj->context->ufile->idr_lock); + idr_preload_end(); + + return ret < 0 ? ret : 0; +} + +/* + * It only removes it from the uobjects list, uverbs_uobject_put() is still + * required. + */ +static void uverbs_idr_remove_uobj(struct ib_uobject *uobj) +{ + spin_lock(&uobj->context->ufile->idr_lock); + idr_remove(&uobj->context->ufile->idr, uobj->id); + spin_unlock(&uobj->context->ufile->idr_lock); +} + +/* Returns the ib_uobject or an error. The caller should check for IS_ERR. */ +static struct ib_uobject *lookup_get_idr_uobject(const struct uverbs_obj_type *type, + struct ib_ucontext *ucontext, + int id, bool exclusive) +{ + struct ib_uobject *uobj; + + rcu_read_lock(); + /* object won't be released as we're protected in rcu */ + uobj = idr_find(&ucontext->ufile->idr, id); + if (!uobj) { + uobj = ERR_PTR(-ENOENT); + goto free; + } + + uverbs_uobject_get(uobj); +free: + rcu_read_unlock(); + return uobj; +} + +static struct ib_uobject *lookup_get_fd_uobject(const struct uverbs_obj_type *type, + struct ib_ucontext *ucontext, + int id, bool exclusive) +{ + struct file *f; + struct ib_uobject *uobject; + const struct uverbs_obj_fd_type *fd_type = + container_of(type, struct uverbs_obj_fd_type, type); + + if (exclusive) + return ERR_PTR(-EOPNOTSUPP); + + f = fget(id); + if (!f) + return ERR_PTR(-EBADF); + + uobject = f->private_data; + /* + * fget(id) ensures we are not currently running uverbs_close_fd, + * and the caller is expected to ensure that uverbs_close_fd is never + * done while a call top lookup is possible. + */ + if (f->f_op != fd_type->fops) { + fput(f); + return ERR_PTR(-EBADF); + } + + uverbs_uobject_get(uobject); + return uobject; +} + +struct ib_uobject *rdma_lookup_get_uobject(const struct uverbs_obj_type *type, + struct ib_ucontext *ucontext, + int id, bool exclusive) +{ + struct ib_uobject *uobj; + int ret; + + uobj = type->type_class->lookup_get(type, ucontext, id, exclusive); + if (IS_ERR(uobj)) + return uobj; + + if (uobj->type != type) { + ret = -EINVAL; + goto free; + } + + ret = uverbs_try_lock_object(uobj, exclusive); + if (ret) { + WARN(ucontext->cleanup_reason, + "ib_uverbs: Trying to lookup_get while cleanup context\n"); + goto free; + } + + return uobj; +free: + uobj->type->type_class->lookup_put(uobj, exclusive); + uverbs_uobject_put(uobj); + return ERR_PTR(ret); +} + +static struct ib_uobject *alloc_begin_idr_uobject(const struct uverbs_obj_type *type, + struct ib_ucontext *ucontext) +{ + int ret; + struct ib_uobject *uobj; + + uobj = alloc_uobj(ucontext, type); + if (IS_ERR(uobj)) + return uobj; + + ret = idr_add_uobj(uobj); + if (ret) + goto uobj_put; + + ret = ib_rdmacg_try_charge(&uobj->cg_obj, ucontext->device, + RDMACG_RESOURCE_HCA_OBJECT); + if (ret) + goto idr_remove; + + return uobj; + +idr_remove: + uverbs_idr_remove_uobj(uobj); +uobj_put: + uverbs_uobject_put(uobj); + return ERR_PTR(ret); +} + +static struct ib_uobject *alloc_begin_fd_uobject(const struct uverbs_obj_type *type, + struct ib_ucontext *ucontext) +{ + const struct uverbs_obj_fd_type *fd_type = + container_of(type, struct uverbs_obj_fd_type, type); + int new_fd; + struct ib_uobject *uobj; + struct ib_uobject_file *uobj_file; + struct file *filp; + + new_fd = get_unused_fd_flags(O_CLOEXEC); + if (new_fd < 0) + return ERR_PTR(new_fd); + + uobj = alloc_uobj(ucontext, type); + if (IS_ERR(uobj)) { + put_unused_fd(new_fd); + return uobj; + } + + uobj_file = container_of(uobj, struct ib_uobject_file, uobj); + filp = anon_inode_getfile(fd_type->name, + fd_type->fops, + uobj_file, + fd_type->flags); + if (IS_ERR(filp)) { + put_unused_fd(new_fd); + uverbs_uobject_put(uobj); + return (void *)filp; + } + + uobj_file->uobj.id = new_fd; + uobj_file->uobj.object = filp; + uobj_file->ufile = ucontext->ufile; + INIT_LIST_HEAD(&uobj->list); + kref_get(&uobj_file->ufile->ref); + + return uobj; +} + +struct ib_uobject *rdma_alloc_begin_uobject(const struct uverbs_obj_type *type, + struct ib_ucontext *ucontext) +{ + return type->type_class->alloc_begin(type, ucontext); +} + +static void uverbs_uobject_add(struct ib_uobject *uobject) +{ + mutex_lock(&uobject->context->uobjects_lock); + list_add(&uobject->list, &uobject->context->uobjects); + mutex_unlock(&uobject->context->uobjects_lock); +} + +static int __must_check remove_commit_idr_uobject(struct ib_uobject *uobj, + enum rdma_remove_reason why) +{ + const struct uverbs_obj_idr_type *idr_type = + container_of(uobj->type, struct uverbs_obj_idr_type, + type); + int ret = idr_type->destroy_object(uobj, why); + + /* + * We can only fail gracefully if the user requested to destroy the + * object. In the rest of the cases, just remove whatever you can. + */ + if (why == RDMA_REMOVE_DESTROY && ret) + return ret; + + ib_rdmacg_uncharge(&uobj->cg_obj, uobj->context->device, + RDMACG_RESOURCE_HCA_OBJECT); + uverbs_idr_remove_uobj(uobj); + + return ret; +} + +static void alloc_abort_fd_uobject(struct ib_uobject *uobj) +{ + struct ib_uobject_file *uobj_file = + container_of(uobj, struct ib_uobject_file, uobj); + struct file *filp = uobj->object; + int id = uobj_file->uobj.id; + + /* Unsuccessful NEW */ + fput(filp); + put_unused_fd(id); +} + +static int __must_check remove_commit_fd_uobject(struct ib_uobject *uobj, + enum rdma_remove_reason why) +{ + const struct uverbs_obj_fd_type *fd_type = + container_of(uobj->type, struct uverbs_obj_fd_type, type); + struct ib_uobject_file *uobj_file = + container_of(uobj, struct ib_uobject_file, uobj); + int ret = fd_type->context_closed(uobj_file, why); + + if (why == RDMA_REMOVE_DESTROY && ret) + return ret; + + if (why == RDMA_REMOVE_DURING_CLEANUP) { + alloc_abort_fd_uobject(uobj); + return ret; + } + + uobj_file->uobj.context = NULL; + return ret; +} + +static void lockdep_check(struct ib_uobject *uobj, bool exclusive) +{ +#ifdef CONFIG_LOCKDEP + if (exclusive) + WARN_ON(atomic_read(&uobj->usecnt) > 0); + else + WARN_ON(atomic_read(&uobj->usecnt) == -1); +#endif +} + +static int __must_check _rdma_remove_commit_uobject(struct ib_uobject *uobj, + enum rdma_remove_reason why) +{ + int ret; + struct ib_ucontext *ucontext = uobj->context; + + ret = uobj->type->type_class->remove_commit(uobj, why); + if (ret && why == RDMA_REMOVE_DESTROY) { + /* We couldn't remove the object, so just unlock the uobject */ + atomic_set(&uobj->usecnt, 0); + uobj->type->type_class->lookup_put(uobj, true); + } else { + mutex_lock(&ucontext->uobjects_lock); + list_del(&uobj->list); + mutex_unlock(&ucontext->uobjects_lock); + /* put the ref we took when we created the object */ + uverbs_uobject_put(uobj); + } + + return ret; +} + +/* This is called only for user requested DESTROY reasons */ +int __must_check rdma_remove_commit_uobject(struct ib_uobject *uobj) +{ + int ret; + struct ib_ucontext *ucontext = uobj->context; + + /* put the ref count we took at lookup_get */ + uverbs_uobject_put(uobj); + /* Cleanup is running. Calling this should have been impossible */ + if (!down_read_trylock(&ucontext->cleanup_rwsem)) { + WARN(true, "ib_uverbs: Cleanup is running while removing an uobject\n"); + return 0; + } + lockdep_check(uobj, true); + ret = _rdma_remove_commit_uobject(uobj, RDMA_REMOVE_DESTROY); + + up_read(&ucontext->cleanup_rwsem); + return ret; +} + +static void alloc_commit_idr_uobject(struct ib_uobject *uobj) +{ + uverbs_uobject_add(uobj); + spin_lock(&uobj->context->ufile->idr_lock); + /* + * We already allocated this IDR with a NULL object, so + * this shouldn't fail. + */ + WARN_ON(idr_replace(&uobj->context->ufile->idr, + uobj, uobj->id)); + spin_unlock(&uobj->context->ufile->idr_lock); +} + +static void alloc_commit_fd_uobject(struct ib_uobject *uobj) +{ + struct ib_uobject_file *uobj_file = + container_of(uobj, struct ib_uobject_file, uobj); + + uverbs_uobject_add(&uobj_file->uobj); + fd_install(uobj_file->uobj.id, uobj->object); + /* This shouldn't be used anymore. Use the file object instead */ + uobj_file->uobj.id = 0; + /* Get another reference as we export this to the fops */ + uverbs_uobject_get(&uobj_file->uobj); +} + +int rdma_alloc_commit_uobject(struct ib_uobject *uobj) +{ + /* Cleanup is running. Calling this should have been impossible */ + if (!down_read_trylock(&uobj->context->cleanup_rwsem)) { + int ret; + + WARN(true, "ib_uverbs: Cleanup is running while allocating an uobject\n"); + ret = uobj->type->type_class->remove_commit(uobj, + RDMA_REMOVE_DURING_CLEANUP); + if (ret) + pr_warn("ib_uverbs: cleanup of idr object %d failed\n", + uobj->id); + return ret; + } + + uobj->type->type_class->alloc_commit(uobj); + up_read(&uobj->context->cleanup_rwsem); + + return 0; +} + +static void alloc_abort_idr_uobject(struct ib_uobject *uobj) +{ + uverbs_idr_remove_uobj(uobj); + ib_rdmacg_uncharge(&uobj->cg_obj, uobj->context->device, + RDMACG_RESOURCE_HCA_OBJECT); + uverbs_uobject_put(uobj); +} + +void rdma_alloc_abort_uobject(struct ib_uobject *uobj) +{ + uobj->type->type_class->alloc_abort(uobj); +} + +static void lookup_put_idr_uobject(struct ib_uobject *uobj, bool exclusive) +{ +} + +static void lookup_put_fd_uobject(struct ib_uobject *uobj, bool exclusive) +{ + struct file *filp = uobj->object; + + WARN_ON(exclusive); + /* This indirectly calls uverbs_close_fd and free the object */ + fput(filp); +} + +void rdma_lookup_put_uobject(struct ib_uobject *uobj, bool exclusive) +{ + lockdep_check(uobj, exclusive); + uobj->type->type_class->lookup_put(uobj, exclusive); + /* + * In order to unlock an object, either decrease its usecnt for + * read access or zero it in case of exclusive access. See + * uverbs_try_lock_object for locking schema information. + */ + if (!exclusive) + atomic_dec(&uobj->usecnt); + else + atomic_set(&uobj->usecnt, 0); + + uverbs_uobject_put(uobj); +} + +const struct uverbs_obj_type_class uverbs_idr_class = { + .alloc_begin = alloc_begin_idr_uobject, + .lookup_get = lookup_get_idr_uobject, + .alloc_commit = alloc_commit_idr_uobject, + .alloc_abort = alloc_abort_idr_uobject, + .lookup_put = lookup_put_idr_uobject, + .remove_commit = remove_commit_idr_uobject, + /* + * When we destroy an object, we first just lock it for WRITE and + * actually DESTROY it in the finalize stage. So, the problematic + * scenario is when we just started the finalize stage of the + * destruction (nothing was executed yet). Now, the other thread + * fetched the object for READ access, but it didn't lock it yet. + * The DESTROY thread continues and starts destroying the object. + * When the other thread continue - without the RCU, it would + * access freed memory. However, the rcu_read_lock delays the free + * until the rcu_read_lock of the READ operation quits. Since the + * exclusive lock of the object is still taken by the DESTROY flow, the + * READ operation will get -EBUSY and it'll just bail out. + */ + .needs_kfree_rcu = true, +}; + +static void _uverbs_close_fd(struct ib_uobject_file *uobj_file) +{ + struct ib_ucontext *ucontext; + struct ib_uverbs_file *ufile = uobj_file->ufile; + int ret; + + mutex_lock(&uobj_file->ufile->cleanup_mutex); + + /* uobject was either already cleaned up or is cleaned up right now anyway */ + if (!uobj_file->uobj.context || + !down_read_trylock(&uobj_file->uobj.context->cleanup_rwsem)) + goto unlock; + + ucontext = uobj_file->uobj.context; + ret = _rdma_remove_commit_uobject(&uobj_file->uobj, RDMA_REMOVE_CLOSE); + up_read(&ucontext->cleanup_rwsem); + if (ret) + pr_warn("uverbs: unable to clean up uobject file in uverbs_close_fd.\n"); +unlock: + mutex_unlock(&ufile->cleanup_mutex); +} + +void uverbs_close_fd(struct file *f) +{ + struct ib_uobject_file *uobj_file = f->private_data; + struct kref *uverbs_file_ref = &uobj_file->ufile->ref; + + _uverbs_close_fd(uobj_file); + uverbs_uobject_put(&uobj_file->uobj); + kref_put(uverbs_file_ref, ib_uverbs_release_file); +} + +void uverbs_cleanup_ucontext(struct ib_ucontext *ucontext, bool device_removed) +{ + enum rdma_remove_reason reason = device_removed ? + RDMA_REMOVE_DRIVER_REMOVE : RDMA_REMOVE_CLOSE; + unsigned int cur_order = 0; + + ucontext->cleanup_reason = reason; + /* + * Waits for all remove_commit and alloc_commit to finish. Logically, We + * want to hold this forever as the context is going to be destroyed, + * but we'll release it since it causes a "held lock freed" BUG message. + */ + down_write(&ucontext->cleanup_rwsem); + + while (!list_empty(&ucontext->uobjects)) { + struct ib_uobject *obj, *next_obj; + unsigned int next_order = UINT_MAX; + + /* + * This shouldn't run while executing other commands on this + * context. Thus, the only thing we should take care of is + * releasing a FD while traversing this list. The FD could be + * closed and released from the _release fop of this FD. + * In order to mitigate this, we add a lock. + * We take and release the lock per order traversal in order + * to let other threads (which might still use the FDs) chance + * to run. + */ + mutex_lock(&ucontext->uobjects_lock); + list_for_each_entry_safe(obj, next_obj, &ucontext->uobjects, + list) { + if (obj->type->destroy_order == cur_order) { + int ret; + + /* + * if we hit this WARN_ON, that means we are + * racing with a lookup_get. + */ + WARN_ON(uverbs_try_lock_object(obj, true)); + ret = obj->type->type_class->remove_commit(obj, + reason); + list_del(&obj->list); + if (ret) + pr_warn("ib_uverbs: failed to remove uobject id %d order %u\n", + obj->id, cur_order); + /* put the ref we took when we created the object */ + uverbs_uobject_put(obj); + } else { + next_order = min(next_order, + obj->type->destroy_order); + } + } + mutex_unlock(&ucontext->uobjects_lock); + cur_order = next_order; + } + up_write(&ucontext->cleanup_rwsem); +} + +void uverbs_initialize_ucontext(struct ib_ucontext *ucontext) +{ + ucontext->cleanup_reason = 0; + mutex_init(&ucontext->uobjects_lock); + INIT_LIST_HEAD(&ucontext->uobjects); + init_rwsem(&ucontext->cleanup_rwsem); +} + +const struct uverbs_obj_type_class uverbs_fd_class = { + .alloc_begin = alloc_begin_fd_uobject, + .lookup_get = lookup_get_fd_uobject, + .alloc_commit = alloc_commit_fd_uobject, + .alloc_abort = alloc_abort_fd_uobject, + .lookup_put = lookup_put_fd_uobject, + .remove_commit = remove_commit_fd_uobject, + .needs_kfree_rcu = false, +}; + diff --git a/drivers/infiniband/core/rdma_core.h b/drivers/infiniband/core/rdma_core.h new file mode 100644 index 000000000000..1b82e7ff7fe8 --- /dev/null +++ b/drivers/infiniband/core/rdma_core.h @@ -0,0 +1,78 @@ +/* + * Copyright (c) 2005 Topspin Communications. All rights reserved. + * Copyright (c) 2005, 2006 Cisco Systems. All rights reserved. + * Copyright (c) 2005-2017 Mellanox Technologies. All rights reserved. + * Copyright (c) 2005 Voltaire, Inc. All rights reserved. + * Copyright (c) 2005 PathScale, Inc. All rights reserved. + * + * 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 RDMA_CORE_H +#define RDMA_CORE_H + +#include <linux/idr.h> +#include <rdma/uverbs_types.h> +#include <rdma/ib_verbs.h> +#include <linux/mutex.h> + +/* + * These functions initialize the context and cleanups its uobjects. + * The context has a list of objects which is protected by a mutex + * on the context. initialize_ucontext should be called when we create + * a context. + * cleanup_ucontext removes all uobjects from the context and puts them. + */ +void uverbs_cleanup_ucontext(struct ib_ucontext *ucontext, bool device_removed); +void uverbs_initialize_ucontext(struct ib_ucontext *ucontext); + +/* + * uverbs_uobject_get is called in order to increase the reference count on + * an uobject. This is useful when a handler wants to keep the uobject's memory + * alive, regardless if this uobject is still alive in the context's objects + * repository. Objects are put via uverbs_uobject_put. + */ +void uverbs_uobject_get(struct ib_uobject *uobject); + +/* + * In order to indicate we no longer needs this uobject, uverbs_uobject_put + * is called. When the reference count is decreased, the uobject is freed. + * For example, this is used when attaching a completion channel to a CQ. + */ +void uverbs_uobject_put(struct ib_uobject *uobject); + +/* Indicate this fd is no longer used by this consumer, but its memory isn't + * necessarily released yet. When the last reference is put, we release the + * memory. After this call is executed, calling uverbs_uobject_get isn't + * allowed. + * This must be called from the release file_operations of the file! + */ +void uverbs_close_fd(struct file *f); + +#endif /* RDMA_CORE_H */ diff --git a/drivers/infiniband/core/roce_gid_mgmt.c b/drivers/infiniband/core/roce_gid_mgmt.c index 0621f4455732..94a9eefb3cfc 100644 --- a/drivers/infiniband/core/roce_gid_mgmt.c +++ b/drivers/infiniband/core/roce_gid_mgmt.c @@ -42,6 +42,8 @@ #include <rdma/ib_cache.h> #include <rdma/ib_addr.h> +static struct workqueue_struct *gid_cache_wq; + enum gid_op_type { GID_DEL = 0, GID_ADD @@ -144,7 +146,6 @@ static enum bonding_slave_state is_eth_active_slave_of_bonding_rcu(struct net_de static int is_eth_port_of_netdev(struct ib_device *ib_dev, u8 port, struct net_device *rdma_ndev, void *cookie) { - struct net_device *event_ndev = (struct net_device *)cookie; struct net_device *real_dev; int res; @@ -152,11 +153,11 @@ static int is_eth_port_of_netdev(struct ib_device *ib_dev, u8 port, return 0; rcu_read_lock(); - real_dev = rdma_vlan_dev_real_dev(event_ndev); + real_dev = rdma_vlan_dev_real_dev(cookie); if (!real_dev) - real_dev = event_ndev; + real_dev = cookie; - res = ((rdma_is_upper_dev_rcu(rdma_ndev, event_ndev) && + res = ((rdma_is_upper_dev_rcu(rdma_ndev, cookie) && (is_eth_active_slave_of_bonding_rcu(rdma_ndev, real_dev) & REQUIRED_BOND_STATES)) || real_dev == rdma_ndev); @@ -192,17 +193,16 @@ static int pass_all_filter(struct ib_device *ib_dev, u8 port, static int upper_device_filter(struct ib_device *ib_dev, u8 port, struct net_device *rdma_ndev, void *cookie) { - struct net_device *event_ndev = (struct net_device *)cookie; int res; if (!rdma_ndev) return 0; - if (rdma_ndev == event_ndev) + if (rdma_ndev == cookie) return 1; rcu_read_lock(); - res = rdma_is_upper_dev_rcu(rdma_ndev, event_ndev); + res = rdma_is_upper_dev_rcu(rdma_ndev, cookie); rcu_read_unlock(); return res; @@ -379,18 +379,14 @@ static void _add_netdev_ips(struct ib_device *ib_dev, u8 port, static void add_netdev_ips(struct ib_device *ib_dev, u8 port, struct net_device *rdma_ndev, void *cookie) { - struct net_device *event_ndev = (struct net_device *)cookie; - - enum_netdev_default_gids(ib_dev, port, event_ndev, rdma_ndev); - _add_netdev_ips(ib_dev, port, event_ndev); + enum_netdev_default_gids(ib_dev, port, cookie, rdma_ndev); + _add_netdev_ips(ib_dev, port, cookie); } static void del_netdev_ips(struct ib_device *ib_dev, u8 port, struct net_device *rdma_ndev, void *cookie) { - struct net_device *event_ndev = (struct net_device *)cookie; - - ib_cache_gid_del_all_netdev_gids(ib_dev, port, event_ndev); + ib_cache_gid_del_all_netdev_gids(ib_dev, port, cookie); } static void enum_all_gids_of_dev_cb(struct ib_device *ib_dev, @@ -460,7 +456,7 @@ static void handle_netdev_upper(struct ib_device *ib_dev, u8 port, u8 port, struct net_device *ndev)) { - struct net_device *ndev = (struct net_device *)cookie; + struct net_device *ndev = cookie; struct upper_list *upper_iter; struct upper_list *upper_temp; LIST_HEAD(upper_list); @@ -519,9 +515,7 @@ static void del_netdev_default_ips_join(struct ib_device *ib_dev, u8 port, static void del_netdev_default_ips(struct ib_device *ib_dev, u8 port, struct net_device *rdma_ndev, void *cookie) { - struct net_device *event_ndev = (struct net_device *)cookie; - - bond_delete_netdev_default_gids(ib_dev, port, event_ndev, rdma_ndev); + bond_delete_netdev_default_gids(ib_dev, port, cookie, rdma_ndev); } /* The following functions operate on all IB devices. netdevice_event and @@ -568,7 +562,7 @@ static int netdevice_queue_work(struct netdev_event_work_cmd *cmds, } INIT_WORK(&ndev_work->work, netdevice_event_work_handler); - queue_work(ib_wq, &ndev_work->work); + queue_work(gid_cache_wq, &ndev_work->work); return NOTIFY_DONE; } @@ -701,7 +695,7 @@ static int addr_event(struct notifier_block *this, unsigned long event, dev_hold(ndev); work->gid_attr.ndev = ndev; - queue_work(ib_wq, &work->work); + queue_work(gid_cache_wq, &work->work); return NOTIFY_DONE; } @@ -748,6 +742,10 @@ static struct notifier_block nb_inet6addr = { int __init roce_gid_mgmt_init(void) { + gid_cache_wq = alloc_ordered_workqueue("gid-cache-wq", 0); + if (!gid_cache_wq) + return -ENOMEM; + register_inetaddr_notifier(&nb_inetaddr); if (IS_ENABLED(CONFIG_IPV6)) register_inet6addr_notifier(&nb_inet6addr); @@ -772,4 +770,5 @@ void __exit roce_gid_mgmt_cleanup(void) * ib-core is removed, all physical devices have been removed, * so no issue with remaining hardware contexts. */ + destroy_workqueue(gid_cache_wq); } diff --git a/drivers/infiniband/core/sa_query.c b/drivers/infiniband/core/sa_query.c index 81b742ca1639..70fa4cabe48e 100644 --- a/drivers/infiniband/core/sa_query.c +++ b/drivers/infiniband/core/sa_query.c @@ -56,6 +56,8 @@ #define IB_SA_LOCAL_SVC_TIMEOUT_MIN 100 #define IB_SA_LOCAL_SVC_TIMEOUT_DEFAULT 2000 #define IB_SA_LOCAL_SVC_TIMEOUT_MAX 200000 +#define IB_SA_CPI_MAX_RETRY_CNT 3 +#define IB_SA_CPI_RETRY_WAIT 1000 /*msecs */ static int sa_local_svc_timeout_ms = IB_SA_LOCAL_SVC_TIMEOUT_DEFAULT; struct ib_sa_sm_ah { @@ -65,9 +67,23 @@ struct ib_sa_sm_ah { u8 src_path_mask; }; +enum rdma_class_port_info_type { + RDMA_CLASS_PORT_INFO_IB, + RDMA_CLASS_PORT_INFO_OPA +}; + +struct rdma_class_port_info { + enum rdma_class_port_info_type type; + union { + struct ib_class_port_info ib; + struct opa_class_port_info opa; + }; +}; + struct ib_sa_classport_cache { bool valid; - struct ib_class_port_info data; + int retry_cnt; + struct rdma_class_port_info data; }; struct ib_sa_port { @@ -75,6 +91,7 @@ struct ib_sa_port { struct ib_sa_sm_ah *sm_ah; struct work_struct update_task; struct ib_sa_classport_cache classport_info; + struct delayed_work ib_cpi_work; spinlock_t classport_lock; /* protects class port info set */ spinlock_t ah_lock; u8 port_num; @@ -103,6 +120,7 @@ struct ib_sa_query { #define IB_SA_ENABLE_LOCAL_SERVICE 0x00000001 #define IB_SA_CANCEL 0x00000002 +#define IB_SA_QUERY_OPA 0x00000004 struct ib_sa_service_query { void (*callback)(int, struct ib_sa_service_rec *, void *); @@ -111,9 +129,10 @@ struct ib_sa_service_query { }; struct ib_sa_path_query { - void (*callback)(int, struct ib_sa_path_rec *, void *); + void (*callback)(int, struct sa_path_rec *, void *); void *context; struct ib_sa_query sa_query; + struct sa_path_rec *conv_pr; }; struct ib_sa_guidinfo_query { @@ -123,7 +142,7 @@ struct ib_sa_guidinfo_query { }; struct ib_sa_classport_info_query { - void (*callback)(int, struct ib_class_port_info *, void *); + void (*callback)(void *); void *context; struct ib_sa_query sa_query; }; @@ -170,8 +189,8 @@ static DEFINE_SPINLOCK(tid_lock); static u32 tid; #define PATH_REC_FIELD(field) \ - .struct_offset_bytes = offsetof(struct ib_sa_path_rec, field), \ - .struct_size_bytes = sizeof ((struct ib_sa_path_rec *) 0)->field, \ + .struct_offset_bytes = offsetof(struct sa_path_rec, field), \ + .struct_size_bytes = sizeof((struct sa_path_rec *)0)->field, \ .field_name = "sa_path_rec:" #field static const struct ib_field path_rec_table[] = { @@ -187,15 +206,15 @@ static const struct ib_field path_rec_table[] = { .offset_words = 6, .offset_bits = 0, .size_bits = 128 }, - { PATH_REC_FIELD(dlid), + { PATH_REC_FIELD(ib.dlid), .offset_words = 10, .offset_bits = 0, .size_bits = 16 }, - { PATH_REC_FIELD(slid), + { PATH_REC_FIELD(ib.slid), .offset_words = 10, .offset_bits = 16, .size_bits = 16 }, - { PATH_REC_FIELD(raw_traffic), + { PATH_REC_FIELD(ib.raw_traffic), .offset_words = 11, .offset_bits = 0, .size_bits = 1 }, @@ -269,6 +288,136 @@ static const struct ib_field path_rec_table[] = { .size_bits = 48 }, }; +#define OPA_PATH_REC_FIELD(field) \ + .struct_offset_bytes = \ + offsetof(struct sa_path_rec, field), \ + .struct_size_bytes = \ + sizeof((struct sa_path_rec *)0)->field, \ + .field_name = "sa_path_rec:" #field + +static const struct ib_field opa_path_rec_table[] = { + { OPA_PATH_REC_FIELD(service_id), + .offset_words = 0, + .offset_bits = 0, + .size_bits = 64 }, + { OPA_PATH_REC_FIELD(dgid), + .offset_words = 2, + .offset_bits = 0, + .size_bits = 128 }, + { OPA_PATH_REC_FIELD(sgid), + .offset_words = 6, + .offset_bits = 0, + .size_bits = 128 }, + { OPA_PATH_REC_FIELD(opa.dlid), + .offset_words = 10, + .offset_bits = 0, + .size_bits = 32 }, + { OPA_PATH_REC_FIELD(opa.slid), + .offset_words = 11, + .offset_bits = 0, + .size_bits = 32 }, + { OPA_PATH_REC_FIELD(opa.raw_traffic), + .offset_words = 12, + .offset_bits = 0, + .size_bits = 1 }, + { RESERVED, + .offset_words = 12, + .offset_bits = 1, + .size_bits = 3 }, + { OPA_PATH_REC_FIELD(flow_label), + .offset_words = 12, + .offset_bits = 4, + .size_bits = 20 }, + { OPA_PATH_REC_FIELD(hop_limit), + .offset_words = 12, + .offset_bits = 24, + .size_bits = 8 }, + { OPA_PATH_REC_FIELD(traffic_class), + .offset_words = 13, + .offset_bits = 0, + .size_bits = 8 }, + { OPA_PATH_REC_FIELD(reversible), + .offset_words = 13, + .offset_bits = 8, + .size_bits = 1 }, + { OPA_PATH_REC_FIELD(numb_path), + .offset_words = 13, + .offset_bits = 9, + .size_bits = 7 }, + { OPA_PATH_REC_FIELD(pkey), + .offset_words = 13, + .offset_bits = 16, + .size_bits = 16 }, + { OPA_PATH_REC_FIELD(opa.l2_8B), + .offset_words = 14, + .offset_bits = 0, + .size_bits = 1 }, + { OPA_PATH_REC_FIELD(opa.l2_10B), + .offset_words = 14, + .offset_bits = 1, + .size_bits = 1 }, + { OPA_PATH_REC_FIELD(opa.l2_9B), + .offset_words = 14, + .offset_bits = 2, + .size_bits = 1 }, + { OPA_PATH_REC_FIELD(opa.l2_16B), + .offset_words = 14, + .offset_bits = 3, + .size_bits = 1 }, + { RESERVED, + .offset_words = 14, + .offset_bits = 4, + .size_bits = 2 }, + { OPA_PATH_REC_FIELD(opa.qos_type), + .offset_words = 14, + .offset_bits = 6, + .size_bits = 2 }, + { OPA_PATH_REC_FIELD(opa.qos_priority), + .offset_words = 14, + .offset_bits = 8, + .size_bits = 8 }, + { RESERVED, + .offset_words = 14, + .offset_bits = 16, + .size_bits = 3 }, + { OPA_PATH_REC_FIELD(sl), + .offset_words = 14, + .offset_bits = 19, + .size_bits = 5 }, + { RESERVED, + .offset_words = 14, + .offset_bits = 24, + .size_bits = 8 }, + { OPA_PATH_REC_FIELD(mtu_selector), + .offset_words = 15, + .offset_bits = 0, + .size_bits = 2 }, + { OPA_PATH_REC_FIELD(mtu), + .offset_words = 15, + .offset_bits = 2, + .size_bits = 6 }, + { OPA_PATH_REC_FIELD(rate_selector), + .offset_words = 15, + .offset_bits = 8, + .size_bits = 2 }, + { OPA_PATH_REC_FIELD(rate), + .offset_words = 15, + .offset_bits = 10, + .size_bits = 6 }, + { OPA_PATH_REC_FIELD(packet_life_time_selector), + .offset_words = 15, + .offset_bits = 16, + .size_bits = 2 }, + { OPA_PATH_REC_FIELD(packet_life_time), + .offset_words = 15, + .offset_bits = 18, + .size_bits = 6 }, + { OPA_PATH_REC_FIELD(preference), + .offset_words = 15, + .offset_bits = 24, + .size_bits = 8 }, +}; + #define MCMEMBER_REC_FIELD(field) \ .struct_offset_bytes = offsetof(struct ib_sa_mcmember_rec, field), \ .struct_size_bytes = sizeof ((struct ib_sa_mcmember_rec *) 0)->field, \ @@ -406,7 +555,7 @@ static const struct ib_field service_rec_table[] = { .struct_size_bytes = sizeof((struct ib_class_port_info *)0)->field, \ .field_name = "ib_class_port_info:" #field -static const struct ib_field classport_info_rec_table[] = { +static const struct ib_field ib_classport_info_rec_table[] = { { CLASSPORTINFO_REC_FIELD(base_version), .offset_words = 0, .offset_bits = 0, @@ -477,6 +626,88 @@ static const struct ib_field classport_info_rec_table[] = { .size_bits = 32 }, }; +#define OPA_CLASSPORTINFO_REC_FIELD(field) \ + .struct_offset_bytes =\ + offsetof(struct opa_class_port_info, field), \ + .struct_size_bytes = \ + sizeof((struct opa_class_port_info *)0)->field, \ + .field_name = "opa_class_port_info:" #field + +static const struct ib_field opa_classport_info_rec_table[] = { + { OPA_CLASSPORTINFO_REC_FIELD(base_version), + .offset_words = 0, + .offset_bits = 0, + .size_bits = 8 }, + { OPA_CLASSPORTINFO_REC_FIELD(class_version), + .offset_words = 0, + .offset_bits = 8, + .size_bits = 8 }, + { OPA_CLASSPORTINFO_REC_FIELD(cap_mask), + .offset_words = 0, + .offset_bits = 16, + .size_bits = 16 }, + { OPA_CLASSPORTINFO_REC_FIELD(cap_mask2_resp_time), + .offset_words = 1, + .offset_bits = 0, + .size_bits = 32 }, + { OPA_CLASSPORTINFO_REC_FIELD(redirect_gid), + .offset_words = 2, + .offset_bits = 0, + .size_bits = 128 }, + { OPA_CLASSPORTINFO_REC_FIELD(redirect_tc_fl), + .offset_words = 6, + .offset_bits = 0, + .size_bits = 32 }, + { OPA_CLASSPORTINFO_REC_FIELD(redirect_lid), + .offset_words = 7, + .offset_bits = 0, + .size_bits = 32 }, + { OPA_CLASSPORTINFO_REC_FIELD(redirect_sl_qp), + .offset_words = 8, + .offset_bits = 0, + .size_bits = 32 }, + { OPA_CLASSPORTINFO_REC_FIELD(redirect_qkey), + .offset_words = 9, + .offset_bits = 0, + .size_bits = 32 }, + { OPA_CLASSPORTINFO_REC_FIELD(trap_gid), + .offset_words = 10, + .offset_bits = 0, + .size_bits = 128 }, + { OPA_CLASSPORTINFO_REC_FIELD(trap_tc_fl), + .offset_words = 14, + .offset_bits = 0, + .size_bits = 32 }, + { OPA_CLASSPORTINFO_REC_FIELD(trap_lid), + .offset_words = 15, + .offset_bits = 0, + .size_bits = 32 }, + { OPA_CLASSPORTINFO_REC_FIELD(trap_hl_qp), + .offset_words = 16, + .offset_bits = 0, + .size_bits = 32 }, + { OPA_CLASSPORTINFO_REC_FIELD(trap_qkey), + .offset_words = 17, + .offset_bits = 0, + .size_bits = 32 }, + { OPA_CLASSPORTINFO_REC_FIELD(trap_pkey), + .offset_words = 18, + .offset_bits = 0, + .size_bits = 16 }, + { OPA_CLASSPORTINFO_REC_FIELD(redirect_pkey), + .offset_words = 18, + .offset_bits = 16, + .size_bits = 16 }, + { OPA_CLASSPORTINFO_REC_FIELD(trap_sl_rsvd), + .offset_words = 19, + .offset_bits = 0, + .size_bits = 8 }, + { RESERVED, + .offset_words = 19, + .offset_bits = 8, + .size_bits = 24 }, +}; + #define GUIDINFO_REC_FIELD(field) \ .struct_offset_bytes = offsetof(struct ib_sa_guidinfo_rec, field), \ .struct_size_bytes = sizeof((struct ib_sa_guidinfo_rec *) 0)->field, \ @@ -518,7 +749,7 @@ static inline int ib_sa_query_cancelled(struct ib_sa_query *query) static void ib_nl_set_path_rec_attrs(struct sk_buff *skb, struct ib_sa_query *query) { - struct ib_sa_path_rec *sa_rec = query->mad_buf->context[1]; + struct sa_path_rec *sa_rec = query->mad_buf->context[1]; struct ib_sa_mad *mad = query->mad_buf->mad; ib_sa_comp_mask comp_mask = mad->sa_hdr.comp_mask; u16 val16; @@ -528,8 +759,7 @@ static void ib_nl_set_path_rec_attrs(struct sk_buff *skb, query->mad_buf->context[1] = NULL; /* Construct the family header first */ - header = (struct rdma_ls_resolve_header *) - skb_put(skb, NLMSG_ALIGN(sizeof(*header))); + header = skb_put(skb, NLMSG_ALIGN(sizeof(*header))); memcpy(header->device_name, query->port->agent->device->name, LS_DEVICE_NAME_MAX); header->port_num = query->port->port_num; @@ -808,7 +1038,7 @@ int ib_nl_handle_set_timeout(struct sk_buff *skb, return -EPERM; ret = nla_parse(tb, LS_NLA_TYPE_MAX - 1, nlmsg_data(nlh), - nlmsg_len(nlh), ib_nl_policy); + nlmsg_len(nlh), ib_nl_policy, NULL); attr = (const struct nlattr *)tb[LS_NLA_TYPE_TIMEOUT]; if (ret || !attr) goto settimeout_out; @@ -860,7 +1090,7 @@ static inline int ib_nl_is_good_resolve_resp(const struct nlmsghdr *nlh) return 0; ret = nla_parse(tb, LS_NLA_TYPE_MAX - 1, nlmsg_data(nlh), - nlmsg_len(nlh), ib_nl_policy); + nlmsg_len(nlh), ib_nl_policy, NULL); if (ret) return 0; @@ -927,96 +1157,10 @@ static void free_sm_ah(struct kref *kref) { struct ib_sa_sm_ah *sm_ah = container_of(kref, struct ib_sa_sm_ah, ref); - ib_destroy_ah(sm_ah->ah); + rdma_destroy_ah(sm_ah->ah); kfree(sm_ah); } -static void update_sm_ah(struct work_struct *work) -{ - struct ib_sa_port *port = - container_of(work, struct ib_sa_port, update_task); - struct ib_sa_sm_ah *new_ah; - struct ib_port_attr port_attr; - struct ib_ah_attr ah_attr; - - if (ib_query_port(port->agent->device, port->port_num, &port_attr)) { - pr_warn("Couldn't query port\n"); - return; - } - - new_ah = kmalloc(sizeof *new_ah, GFP_KERNEL); - if (!new_ah) { - return; - } - - kref_init(&new_ah->ref); - new_ah->src_path_mask = (1 << port_attr.lmc) - 1; - - new_ah->pkey_index = 0; - if (ib_find_pkey(port->agent->device, port->port_num, - IB_DEFAULT_PKEY_FULL, &new_ah->pkey_index)) - pr_err("Couldn't find index for default PKey\n"); - - memset(&ah_attr, 0, sizeof ah_attr); - ah_attr.dlid = port_attr.sm_lid; - ah_attr.sl = port_attr.sm_sl; - ah_attr.port_num = port->port_num; - if (port_attr.grh_required) { - ah_attr.ah_flags = IB_AH_GRH; - ah_attr.grh.dgid.global.subnet_prefix = cpu_to_be64(port_attr.subnet_prefix); - ah_attr.grh.dgid.global.interface_id = cpu_to_be64(IB_SA_WELL_KNOWN_GUID); - } - - new_ah->ah = ib_create_ah(port->agent->qp->pd, &ah_attr); - if (IS_ERR(new_ah->ah)) { - pr_warn("Couldn't create new SM AH\n"); - kfree(new_ah); - return; - } - - spin_lock_irq(&port->ah_lock); - if (port->sm_ah) - kref_put(&port->sm_ah->ref, free_sm_ah); - port->sm_ah = new_ah; - spin_unlock_irq(&port->ah_lock); - -} - -static void ib_sa_event(struct ib_event_handler *handler, struct ib_event *event) -{ - if (event->event == IB_EVENT_PORT_ERR || - event->event == IB_EVENT_PORT_ACTIVE || - event->event == IB_EVENT_LID_CHANGE || - event->event == IB_EVENT_PKEY_CHANGE || - event->event == IB_EVENT_SM_CHANGE || - event->event == IB_EVENT_CLIENT_REREGISTER) { - unsigned long flags; - struct ib_sa_device *sa_dev = - container_of(handler, typeof(*sa_dev), event_handler); - struct ib_sa_port *port = - &sa_dev->port[event->element.port_num - sa_dev->start_port]; - - if (!rdma_cap_ib_sa(handler->device, port->port_num)) - return; - - spin_lock_irqsave(&port->ah_lock, flags); - if (port->sm_ah) - kref_put(&port->sm_ah->ref, free_sm_ah); - port->sm_ah = NULL; - spin_unlock_irqrestore(&port->ah_lock, flags); - - if (event->event == IB_EVENT_SM_CHANGE || - event->event == IB_EVENT_CLIENT_REREGISTER || - event->event == IB_EVENT_LID_CHANGE) { - spin_lock_irqsave(&port->classport_lock, flags); - port->classport_info.valid = false; - spin_unlock_irqrestore(&port->classport_lock, flags); - } - queue_work(ib_wq, &sa_dev->port[event->element.port_num - - sa_dev->start_port].update_task); - } -} - void ib_sa_register_client(struct ib_sa_client *client) { atomic_set(&client->users, 1); @@ -1085,7 +1229,8 @@ static u8 get_src_path_mask(struct ib_device *device, u8 port_num) } int ib_init_ah_from_path(struct ib_device *device, u8 port_num, - struct ib_sa_path_rec *rec, struct ib_ah_attr *ah_attr) + struct sa_path_rec *rec, + struct rdma_ah_attr *ah_attr) { int ret; u16 gid_index; @@ -1093,21 +1238,26 @@ int ib_init_ah_from_path(struct ib_device *device, u8 port_num, struct net_device *ndev = NULL; memset(ah_attr, 0, sizeof *ah_attr); - ah_attr->dlid = be16_to_cpu(rec->dlid); - ah_attr->sl = rec->sl; - ah_attr->src_path_bits = be16_to_cpu(rec->slid) & - get_src_path_mask(device, port_num); - ah_attr->port_num = port_num; - ah_attr->static_rate = rec->rate; - + ah_attr->type = rdma_ah_find_type(device, port_num); + + rdma_ah_set_dlid(ah_attr, be32_to_cpu(sa_path_get_dlid(rec))); + rdma_ah_set_sl(ah_attr, rec->sl); + rdma_ah_set_path_bits(ah_attr, be32_to_cpu(sa_path_get_slid(rec)) & + get_src_path_mask(device, port_num)); + rdma_ah_set_port_num(ah_attr, port_num); + rdma_ah_set_static_rate(ah_attr, rec->rate); use_roce = rdma_cap_eth_ah(device, port_num); if (use_roce) { struct net_device *idev; struct net_device *resolved_dev; - struct rdma_dev_addr dev_addr = {.bound_dev_if = rec->ifindex, - .net = rec->net ? rec->net : - &init_net}; + struct rdma_dev_addr dev_addr = { + .bound_dev_if = ((sa_path_get_ifindex(rec) >= 0) ? + sa_path_get_ifindex(rec) : 0), + .net = sa_path_get_ndev(rec) ? + sa_path_get_ndev(rec) : + &init_net + }; union { struct sockaddr _sockaddr; struct sockaddr_in _sockaddr_in; @@ -1128,7 +1278,7 @@ int ib_init_ah_from_path(struct ib_device *device, u8 port_num, if ((dev_addr.network == RDMA_NETWORK_IPV4 || dev_addr.network == RDMA_NETWORK_IPV6) && - rec->gid_type != IB_GID_TYPE_ROCE_UDP_ENCAP) + rec->rec_type != SA_PATH_REC_TYPE_ROCE_V2) return -EINVAL; idev = device->get_netdev(device, port_num); @@ -1159,28 +1309,31 @@ int ib_init_ah_from_path(struct ib_device *device, u8 port_num, } if (rec->hop_limit > 0 || use_roce) { - ah_attr->ah_flags = IB_AH_GRH; - ah_attr->grh.dgid = rec->dgid; + enum ib_gid_type type = sa_conv_pathrec_to_gid_type(rec); - ret = ib_find_cached_gid_by_port(device, &rec->sgid, - rec->gid_type, port_num, ndev, - &gid_index); + ret = ib_find_cached_gid_by_port(device, &rec->sgid, type, + port_num, ndev, &gid_index); if (ret) { if (ndev) dev_put(ndev); return ret; } - ah_attr->grh.sgid_index = gid_index; - ah_attr->grh.flow_label = be32_to_cpu(rec->flow_label); - ah_attr->grh.hop_limit = rec->hop_limit; - ah_attr->grh.traffic_class = rec->traffic_class; + rdma_ah_set_grh(ah_attr, &rec->dgid, + be32_to_cpu(rec->flow_label), + gid_index, rec->hop_limit, + rec->traffic_class); if (ndev) dev_put(ndev); } - if (use_roce) - memcpy(ah_attr->dmac, rec->dmac, ETH_ALEN); + if (use_roce) { + u8 *dmac = sa_path_get_dmac(rec); + + if (!dmac) + return -EINVAL; + memcpy(ah_attr->roce.dmac, dmac, ETH_ALEN); + } return 0; } @@ -1203,7 +1356,9 @@ static int alloc_mad(struct ib_sa_query *query, gfp_t gfp_mask) query->sm_ah->pkey_index, 0, IB_MGMT_SA_HDR, IB_MGMT_SA_DATA, gfp_mask, - IB_MGMT_BASE_VERSION); + ((query->flags & IB_SA_QUERY_OPA) ? + OPA_MGMT_BASE_VERSION : + IB_MGMT_BASE_VERSION)); if (IS_ERR(query->mad_buf)) { kref_put(&query->sm_ah->ref, free_sm_ah); return -ENOMEM; @@ -1220,16 +1375,21 @@ static void free_mad(struct ib_sa_query *query) kref_put(&query->sm_ah->ref, free_sm_ah); } -static void init_mad(struct ib_sa_mad *mad, struct ib_mad_agent *agent) +static void init_mad(struct ib_sa_query *query, struct ib_mad_agent *agent) { + struct ib_sa_mad *mad = query->mad_buf->mad; unsigned long flags; memset(mad, 0, sizeof *mad); - mad->mad_hdr.base_version = IB_MGMT_BASE_VERSION; + if (query->flags & IB_SA_QUERY_OPA) { + mad->mad_hdr.base_version = OPA_MGMT_BASE_VERSION; + mad->mad_hdr.class_version = OPA_SA_CLASS_VERSION; + } else { + mad->mad_hdr.base_version = IB_MGMT_BASE_VERSION; + mad->mad_hdr.class_version = IB_SA_CLASS_VERSION; + } mad->mad_hdr.mgmt_class = IB_MGMT_CLASS_SUBN_ADM; - mad->mad_hdr.class_version = IB_SA_CLASS_VERSION; - spin_lock_irqsave(&tid_lock, flags); mad->mad_hdr.tid = cpu_to_be64(((u64) agent->hi_tid) << 32 | tid++); @@ -1258,7 +1418,8 @@ static int send_mad(struct ib_sa_query *query, int timeout_ms, gfp_t gfp_mask) query->mad_buf->context[0] = query; query->id = id; - if (query->flags & IB_SA_ENABLE_LOCAL_SERVICE) { + if ((query->flags & IB_SA_ENABLE_LOCAL_SERVICE) && + (!(query->flags & IB_SA_QUERY_OPA))) { if (!ibnl_chk_listeners(RDMA_NL_GROUP_LS)) { if (!ib_nl_make_request(query, gfp_mask)) return id; @@ -1281,18 +1442,75 @@ static int send_mad(struct ib_sa_query *query, int timeout_ms, gfp_t gfp_mask) return ret ? ret : id; } -void ib_sa_unpack_path(void *attribute, struct ib_sa_path_rec *rec) +void ib_sa_unpack_path(void *attribute, struct sa_path_rec *rec) { ib_unpack(path_rec_table, ARRAY_SIZE(path_rec_table), attribute, rec); } EXPORT_SYMBOL(ib_sa_unpack_path); -void ib_sa_pack_path(struct ib_sa_path_rec *rec, void *attribute) +void ib_sa_pack_path(struct sa_path_rec *rec, void *attribute) { ib_pack(path_rec_table, ARRAY_SIZE(path_rec_table), rec, attribute); } EXPORT_SYMBOL(ib_sa_pack_path); +static bool ib_sa_opa_pathrecord_support(struct ib_sa_client *client, + struct ib_device *device, + u8 port_num) +{ + struct ib_sa_device *sa_dev = ib_get_client_data(device, &sa_client); + struct ib_sa_port *port; + unsigned long flags; + bool ret = false; + + if (!sa_dev) + return ret; + + port = &sa_dev->port[port_num - sa_dev->start_port]; + spin_lock_irqsave(&port->classport_lock, flags); + if (!port->classport_info.valid) + goto ret; + + if (port->classport_info.data.type == RDMA_CLASS_PORT_INFO_OPA) + ret = opa_get_cpi_capmask2(&port->classport_info.data.opa) & + OPA_CLASS_PORT_INFO_PR_SUPPORT; +ret: + spin_unlock_irqrestore(&port->classport_lock, flags); + return ret; +} + +enum opa_pr_supported { + PR_NOT_SUPPORTED, + PR_OPA_SUPPORTED, + PR_IB_SUPPORTED +}; + +/** + * Check if current PR query can be an OPA query. + * Retuns PR_NOT_SUPPORTED if a path record query is not + * possible, PR_OPA_SUPPORTED if an OPA path record query + * is possible and PR_IB_SUPPORTED if an IB path record + * query is possible. + */ +static int opa_pr_query_possible(struct ib_sa_client *client, + struct ib_device *device, + u8 port_num, + struct sa_path_rec *rec) +{ + struct ib_port_attr port_attr; + + if (ib_query_port(device, port_num, &port_attr)) + return PR_NOT_SUPPORTED; + + if (ib_sa_opa_pathrecord_support(client, device, port_num)) + return PR_OPA_SUPPORTED; + + if (port_attr.lid >= be16_to_cpu(IB_MULTICAST_LID_BASE)) + return PR_NOT_SUPPORTED; + else + return PR_IB_SUPPORTED; +} + static void ib_sa_path_rec_callback(struct ib_sa_query *sa_query, int status, struct ib_sa_mad *mad) @@ -1301,22 +1519,44 @@ static void ib_sa_path_rec_callback(struct ib_sa_query *sa_query, container_of(sa_query, struct ib_sa_path_query, sa_query); if (mad) { - struct ib_sa_path_rec rec; - - ib_unpack(path_rec_table, ARRAY_SIZE(path_rec_table), - mad->data, &rec); - rec.net = NULL; - rec.ifindex = 0; - rec.gid_type = IB_GID_TYPE_IB; - eth_zero_addr(rec.dmac); - query->callback(status, &rec, query->context); + struct sa_path_rec rec; + + if (sa_query->flags & IB_SA_QUERY_OPA) { + ib_unpack(opa_path_rec_table, + ARRAY_SIZE(opa_path_rec_table), + mad->data, &rec); + rec.rec_type = SA_PATH_REC_TYPE_OPA; + query->callback(status, &rec, query->context); + } else { + ib_unpack(path_rec_table, + ARRAY_SIZE(path_rec_table), + mad->data, &rec); + rec.rec_type = SA_PATH_REC_TYPE_IB; + sa_path_set_ndev(&rec, NULL); + sa_path_set_ifindex(&rec, 0); + sa_path_set_dmac_zero(&rec); + + if (query->conv_pr) { + struct sa_path_rec opa; + + memset(&opa, 0, sizeof(struct sa_path_rec)); + sa_convert_path_ib_to_opa(&opa, &rec); + query->callback(status, &opa, query->context); + } else { + query->callback(status, &rec, query->context); + } + } } else query->callback(status, NULL, query->context); } static void ib_sa_path_rec_release(struct ib_sa_query *sa_query) { - kfree(container_of(sa_query, struct ib_sa_path_query, sa_query)); + struct ib_sa_path_query *query = + container_of(sa_query, struct ib_sa_path_query, sa_query); + + kfree(query->conv_pr); + kfree(query); } /** @@ -1346,11 +1586,11 @@ static void ib_sa_path_rec_release(struct ib_sa_query *sa_query) */ int ib_sa_path_rec_get(struct ib_sa_client *client, struct ib_device *device, u8 port_num, - struct ib_sa_path_rec *rec, + struct sa_path_rec *rec, ib_sa_comp_mask comp_mask, int timeout_ms, gfp_t gfp_mask, void (*callback)(int status, - struct ib_sa_path_rec *resp, + struct sa_path_rec *resp, void *context), void *context, struct ib_sa_query **sa_query) @@ -1360,11 +1600,16 @@ int ib_sa_path_rec_get(struct ib_sa_client *client, struct ib_sa_port *port; struct ib_mad_agent *agent; struct ib_sa_mad *mad; + enum opa_pr_supported status; int ret; if (!sa_dev) return -ENODEV; + if ((rec->rec_type != SA_PATH_REC_TYPE_IB) && + (rec->rec_type != SA_PATH_REC_TYPE_OPA)) + return -EINVAL; + port = &sa_dev->port[port_num - sa_dev->start_port]; agent = port->agent; @@ -1373,9 +1618,26 @@ int ib_sa_path_rec_get(struct ib_sa_client *client, return -ENOMEM; query->sa_query.port = port; + if (rec->rec_type == SA_PATH_REC_TYPE_OPA) { + status = opa_pr_query_possible(client, device, port_num, rec); + if (status == PR_NOT_SUPPORTED) { + ret = -EINVAL; + goto err1; + } else if (status == PR_OPA_SUPPORTED) { + query->sa_query.flags |= IB_SA_QUERY_OPA; + } else { + query->conv_pr = + kmalloc(sizeof(*query->conv_pr), gfp_mask); + if (!query->conv_pr) { + ret = -ENOMEM; + goto err1; + } + } + } + ret = alloc_mad(&query->sa_query, gfp_mask); if (ret) - goto err1; + goto err2; ib_sa_client_get(client); query->sa_query.client = client; @@ -1383,7 +1645,7 @@ int ib_sa_path_rec_get(struct ib_sa_client *client, query->context = context; mad = query->sa_query.mad_buf->mad; - init_mad(mad, agent); + init_mad(&query->sa_query, agent); query->sa_query.callback = callback ? ib_sa_path_rec_callback : NULL; query->sa_query.release = ib_sa_path_rec_release; @@ -1391,24 +1653,36 @@ int ib_sa_path_rec_get(struct ib_sa_client *client, mad->mad_hdr.attr_id = cpu_to_be16(IB_SA_ATTR_PATH_REC); mad->sa_hdr.comp_mask = comp_mask; - ib_pack(path_rec_table, ARRAY_SIZE(path_rec_table), rec, mad->data); + if (query->sa_query.flags & IB_SA_QUERY_OPA) { + ib_pack(opa_path_rec_table, ARRAY_SIZE(opa_path_rec_table), + rec, mad->data); + } else if (query->conv_pr) { + sa_convert_path_opa_to_ib(query->conv_pr, rec); + ib_pack(path_rec_table, ARRAY_SIZE(path_rec_table), + query->conv_pr, mad->data); + } else { + ib_pack(path_rec_table, ARRAY_SIZE(path_rec_table), + rec, mad->data); + } *sa_query = &query->sa_query; query->sa_query.flags |= IB_SA_ENABLE_LOCAL_SERVICE; - query->sa_query.mad_buf->context[1] = rec; + query->sa_query.mad_buf->context[1] = (query->conv_pr) ? + query->conv_pr : rec; ret = send_mad(&query->sa_query, timeout_ms, gfp_mask); if (ret < 0) - goto err2; + goto err3; return ret; -err2: +err3: *sa_query = NULL; ib_sa_client_put(query->sa_query.client); free_mad(&query->sa_query); - +err2: + kfree(query->conv_pr); err1: kfree(query); return ret; @@ -1508,7 +1782,7 @@ int ib_sa_service_rec_query(struct ib_sa_client *client, query->context = context; mad = query->sa_query.mad_buf->mad; - init_mad(mad, agent); + init_mad(&query->sa_query, agent); query->sa_query.callback = callback ? ib_sa_service_rec_callback : NULL; query->sa_query.release = ib_sa_service_rec_release; @@ -1600,7 +1874,7 @@ int ib_sa_mcmember_rec_query(struct ib_sa_client *client, query->context = context; mad = query->sa_query.mad_buf->mad; - init_mad(mad, agent); + init_mad(&query->sa_query, agent); query->sa_query.callback = callback ? ib_sa_mcmember_rec_callback : NULL; query->sa_query.release = ib_sa_mcmember_rec_release; @@ -1697,7 +1971,7 @@ int ib_sa_guid_info_rec_query(struct ib_sa_client *client, query->context = context; mad = query->sa_query.mad_buf->mad; - init_mad(mad, agent); + init_mad(&query->sa_query, agent); query->sa_query.callback = callback ? ib_sa_guidinfo_rec_callback : NULL; query->sa_query.release = ib_sa_guidinfo_rec_release; @@ -1728,7 +2002,42 @@ err1: } EXPORT_SYMBOL(ib_sa_guid_info_rec_query); -/* Support get SA ClassPortInfo */ +bool ib_sa_sendonly_fullmem_support(struct ib_sa_client *client, + struct ib_device *device, + u8 port_num) +{ + struct ib_sa_device *sa_dev = ib_get_client_data(device, &sa_client); + struct ib_sa_port *port; + bool ret = false; + unsigned long flags; + + if (!sa_dev) + return ret; + + port = &sa_dev->port[port_num - sa_dev->start_port]; + + spin_lock_irqsave(&port->classport_lock, flags); + if ((port->classport_info.valid) && + (port->classport_info.data.type == RDMA_CLASS_PORT_INFO_IB)) + ret = ib_get_cpi_capmask2(&port->classport_info.data.ib) + & IB_SA_CAP_MASK2_SENDONLY_FULL_MEM_SUPPORT; + spin_unlock_irqrestore(&port->classport_lock, flags); + return ret; +} +EXPORT_SYMBOL(ib_sa_sendonly_fullmem_support); + +struct ib_classport_info_context { + struct completion done; + struct ib_sa_query *sa_query; +}; + +static void ib_classportinfo_cb(void *context) +{ + struct ib_classport_info_context *cb_ctx = context; + + complete(&cb_ctx->done); +} + static void ib_sa_classport_info_rec_callback(struct ib_sa_query *sa_query, int status, struct ib_sa_mad *mad) @@ -1736,91 +2045,91 @@ static void ib_sa_classport_info_rec_callback(struct ib_sa_query *sa_query, unsigned long flags; struct ib_sa_classport_info_query *query = container_of(sa_query, struct ib_sa_classport_info_query, sa_query); + struct ib_sa_classport_cache *info = &sa_query->port->classport_info; if (mad) { - struct ib_class_port_info rec; + if (sa_query->flags & IB_SA_QUERY_OPA) { + struct opa_class_port_info rec; - ib_unpack(classport_info_rec_table, - ARRAY_SIZE(classport_info_rec_table), - mad->data, &rec); + ib_unpack(opa_classport_info_rec_table, + ARRAY_SIZE(opa_classport_info_rec_table), + mad->data, &rec); - spin_lock_irqsave(&sa_query->port->classport_lock, flags); - if (!status && !sa_query->port->classport_info.valid) { - memcpy(&sa_query->port->classport_info.data, &rec, - sizeof(sa_query->port->classport_info.data)); + spin_lock_irqsave(&sa_query->port->classport_lock, + flags); + if (!status && !info->valid) { + memcpy(&info->data.opa, &rec, + sizeof(info->data.opa)); - sa_query->port->classport_info.valid = true; - } - spin_unlock_irqrestore(&sa_query->port->classport_lock, flags); + info->valid = true; + info->data.type = RDMA_CLASS_PORT_INFO_OPA; + } + spin_unlock_irqrestore(&sa_query->port->classport_lock, + flags); - query->callback(status, &rec, query->context); - } else { - query->callback(status, NULL, query->context); + } else { + struct ib_class_port_info rec; + + ib_unpack(ib_classport_info_rec_table, + ARRAY_SIZE(ib_classport_info_rec_table), + mad->data, &rec); + + spin_lock_irqsave(&sa_query->port->classport_lock, + flags); + if (!status && !info->valid) { + memcpy(&info->data.ib, &rec, + sizeof(info->data.ib)); + + info->valid = true; + info->data.type = RDMA_CLASS_PORT_INFO_IB; + } + spin_unlock_irqrestore(&sa_query->port->classport_lock, + flags); + } } + query->callback(query->context); } -static void ib_sa_portclass_info_rec_release(struct ib_sa_query *sa_query) +static void ib_sa_classport_info_rec_release(struct ib_sa_query *sa_query) { kfree(container_of(sa_query, struct ib_sa_classport_info_query, sa_query)); } -int ib_sa_classport_info_rec_query(struct ib_sa_client *client, - struct ib_device *device, u8 port_num, - int timeout_ms, gfp_t gfp_mask, - void (*callback)(int status, - struct ib_class_port_info *resp, - void *context), - void *context, - struct ib_sa_query **sa_query) +static int ib_sa_classport_info_rec_query(struct ib_sa_port *port, + int timeout_ms, + void (*callback)(void *context), + void *context, + struct ib_sa_query **sa_query) { - struct ib_sa_classport_info_query *query; - struct ib_sa_device *sa_dev = ib_get_client_data(device, &sa_client); - struct ib_sa_port *port; struct ib_mad_agent *agent; + struct ib_sa_classport_info_query *query; struct ib_sa_mad *mad; - struct ib_class_port_info cached_class_port_info; + gfp_t gfp_mask = GFP_KERNEL; int ret; - unsigned long flags; - if (!sa_dev) - return -ENODEV; - - port = &sa_dev->port[port_num - sa_dev->start_port]; agent = port->agent; - /* Use cached ClassPortInfo attribute if valid instead of sending mad */ - spin_lock_irqsave(&port->classport_lock, flags); - if (port->classport_info.valid && callback) { - memcpy(&cached_class_port_info, &port->classport_info.data, - sizeof(cached_class_port_info)); - spin_unlock_irqrestore(&port->classport_lock, flags); - callback(0, &cached_class_port_info, context); - return 0; - } - spin_unlock_irqrestore(&port->classport_lock, flags); - query = kzalloc(sizeof(*query), gfp_mask); if (!query) return -ENOMEM; query->sa_query.port = port; + query->sa_query.flags |= rdma_cap_opa_ah(port->agent->device, + port->port_num) ? + IB_SA_QUERY_OPA : 0; ret = alloc_mad(&query->sa_query, gfp_mask); if (ret) - goto err1; + goto err_free; - ib_sa_client_get(client); - query->sa_query.client = client; - query->callback = callback; - query->context = context; + query->callback = callback; + query->context = context; mad = query->sa_query.mad_buf->mad; - init_mad(mad, agent); + init_mad(&query->sa_query, agent); - query->sa_query.callback = callback ? ib_sa_classport_info_rec_callback : NULL; - - query->sa_query.release = ib_sa_portclass_info_rec_release; - /* support GET only */ + query->sa_query.callback = ib_sa_classport_info_rec_callback; + query->sa_query.release = ib_sa_classport_info_rec_release; mad->mad_hdr.method = IB_MGMT_METHOD_GET; mad->mad_hdr.attr_id = cpu_to_be16(IB_SA_ATTR_CLASS_PORTINFO); mad->sa_hdr.comp_mask = 0; @@ -1828,20 +2137,71 @@ int ib_sa_classport_info_rec_query(struct ib_sa_client *client, ret = send_mad(&query->sa_query, timeout_ms, gfp_mask); if (ret < 0) - goto err2; + goto err_free_mad; return ret; -err2: +err_free_mad: *sa_query = NULL; - ib_sa_client_put(query->sa_query.client); free_mad(&query->sa_query); -err1: +err_free: kfree(query); return ret; } -EXPORT_SYMBOL(ib_sa_classport_info_rec_query); + +static void update_ib_cpi(struct work_struct *work) +{ + struct ib_sa_port *port = + container_of(work, struct ib_sa_port, ib_cpi_work.work); + struct ib_classport_info_context *cb_context; + unsigned long flags; + int ret; + + /* If the classport info is valid, nothing + * to do here. + */ + spin_lock_irqsave(&port->classport_lock, flags); + if (port->classport_info.valid) { + spin_unlock_irqrestore(&port->classport_lock, flags); + return; + } + spin_unlock_irqrestore(&port->classport_lock, flags); + + cb_context = kmalloc(sizeof(*cb_context), GFP_KERNEL); + if (!cb_context) + goto err_nomem; + + init_completion(&cb_context->done); + + ret = ib_sa_classport_info_rec_query(port, 3000, + ib_classportinfo_cb, cb_context, + &cb_context->sa_query); + if (ret < 0) + goto free_cb_err; + wait_for_completion(&cb_context->done); +free_cb_err: + kfree(cb_context); + spin_lock_irqsave(&port->classport_lock, flags); + + /* If the classport info is still not valid, the query should have + * failed for some reason. Retry issuing the query + */ + if (!port->classport_info.valid) { + port->classport_info.retry_cnt++; + if (port->classport_info.retry_cnt <= + IB_SA_CPI_MAX_RETRY_CNT) { + unsigned long delay = + msecs_to_jiffies(IB_SA_CPI_RETRY_WAIT); + + queue_delayed_work(ib_wq, &port->ib_cpi_work, delay); + } + } + spin_unlock_irqrestore(&port->classport_lock, flags); + +err_nomem: + return; +} static void send_handler(struct ib_mad_agent *agent, struct ib_mad_send_wc *mad_send_wc) @@ -1870,7 +2230,8 @@ static void send_handler(struct ib_mad_agent *agent, spin_unlock_irqrestore(&idr_lock, flags); free_mad(query); - ib_sa_client_put(query->client); + if (query->client) + ib_sa_client_put(query->client); query->release(query); } @@ -1897,6 +2258,102 @@ static void recv_handler(struct ib_mad_agent *mad_agent, ib_free_recv_mad(mad_recv_wc); } +static void update_sm_ah(struct work_struct *work) +{ + struct ib_sa_port *port = + container_of(work, struct ib_sa_port, update_task); + struct ib_sa_sm_ah *new_ah; + struct ib_port_attr port_attr; + struct rdma_ah_attr ah_attr; + + if (ib_query_port(port->agent->device, port->port_num, &port_attr)) { + pr_warn("Couldn't query port\n"); + return; + } + + new_ah = kmalloc(sizeof(*new_ah), GFP_KERNEL); + if (!new_ah) + return; + + kref_init(&new_ah->ref); + new_ah->src_path_mask = (1 << port_attr.lmc) - 1; + + new_ah->pkey_index = 0; + if (ib_find_pkey(port->agent->device, port->port_num, + IB_DEFAULT_PKEY_FULL, &new_ah->pkey_index)) + pr_err("Couldn't find index for default PKey\n"); + + memset(&ah_attr, 0, sizeof(ah_attr)); + ah_attr.type = rdma_ah_find_type(port->agent->device, + port->port_num); + rdma_ah_set_dlid(&ah_attr, port_attr.sm_lid); + rdma_ah_set_sl(&ah_attr, port_attr.sm_sl); + rdma_ah_set_port_num(&ah_attr, port->port_num); + if (port_attr.grh_required) { + rdma_ah_set_ah_flags(&ah_attr, IB_AH_GRH); + + rdma_ah_set_subnet_prefix(&ah_attr, + cpu_to_be64(port_attr.subnet_prefix)); + rdma_ah_set_interface_id(&ah_attr, + cpu_to_be64(IB_SA_WELL_KNOWN_GUID)); + } + + new_ah->ah = rdma_create_ah(port->agent->qp->pd, &ah_attr); + if (IS_ERR(new_ah->ah)) { + pr_warn("Couldn't create new SM AH\n"); + kfree(new_ah); + return; + } + + spin_lock_irq(&port->ah_lock); + if (port->sm_ah) + kref_put(&port->sm_ah->ref, free_sm_ah); + port->sm_ah = new_ah; + spin_unlock_irq(&port->ah_lock); +} + +static void ib_sa_event(struct ib_event_handler *handler, + struct ib_event *event) +{ + if (event->event == IB_EVENT_PORT_ERR || + event->event == IB_EVENT_PORT_ACTIVE || + event->event == IB_EVENT_LID_CHANGE || + event->event == IB_EVENT_PKEY_CHANGE || + event->event == IB_EVENT_SM_CHANGE || + event->event == IB_EVENT_CLIENT_REREGISTER) { + unsigned long flags; + struct ib_sa_device *sa_dev = + container_of(handler, typeof(*sa_dev), event_handler); + u8 port_num = event->element.port_num - sa_dev->start_port; + struct ib_sa_port *port = &sa_dev->port[port_num]; + + if (!rdma_cap_ib_sa(handler->device, port->port_num)) + return; + + spin_lock_irqsave(&port->ah_lock, flags); + if (port->sm_ah) + kref_put(&port->sm_ah->ref, free_sm_ah); + port->sm_ah = NULL; + spin_unlock_irqrestore(&port->ah_lock, flags); + + if (event->event == IB_EVENT_SM_CHANGE || + event->event == IB_EVENT_CLIENT_REREGISTER || + event->event == IB_EVENT_LID_CHANGE || + event->event == IB_EVENT_PORT_ACTIVE) { + unsigned long delay = + msecs_to_jiffies(IB_SA_CPI_RETRY_WAIT); + + spin_lock_irqsave(&port->classport_lock, flags); + port->classport_info.valid = false; + port->classport_info.retry_cnt = 0; + spin_unlock_irqrestore(&port->classport_lock, flags); + queue_delayed_work(ib_wq, + &port->ib_cpi_work, delay); + } + queue_work(ib_wq, &sa_dev->port[port_num].update_task); + } +} + static void ib_sa_add_one(struct ib_device *device) { struct ib_sa_device *sa_dev; @@ -1934,6 +2391,8 @@ static void ib_sa_add_one(struct ib_device *device) goto err; INIT_WORK(&sa_dev->port[i].update_task, update_sm_ah); + INIT_DELAYED_WORK(&sa_dev->port[i].ib_cpi_work, + update_ib_cpi); count++; } @@ -1980,11 +2439,11 @@ static void ib_sa_remove_one(struct ib_device *device, void *client_data) return; ib_unregister_event_handler(&sa_dev->event_handler); - flush_workqueue(ib_wq); for (i = 0; i <= sa_dev->end_port - sa_dev->start_port; ++i) { if (rdma_cap_ib_sa(device, i + 1)) { + cancel_delayed_work_sync(&sa_dev->port[i].ib_cpi_work); ib_unregister_mad_agent(sa_dev->port[i].agent); if (sa_dev->port[i].sm_ah) kref_put(&sa_dev->port[i].sm_ah->ref, free_sm_ah); diff --git a/drivers/infiniband/core/security.c b/drivers/infiniband/core/security.c new file mode 100644 index 000000000000..70ad19c4c73e --- /dev/null +++ b/drivers/infiniband/core/security.c @@ -0,0 +1,709 @@ +/* + * Copyright (c) 2016 Mellanox Technologies Ltd. All rights reserved. + * + * 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. + */ + +#ifdef CONFIG_SECURITY_INFINIBAND + +#include <linux/security.h> +#include <linux/completion.h> +#include <linux/list.h> + +#include <rdma/ib_verbs.h> +#include <rdma/ib_cache.h> +#include "core_priv.h" +#include "mad_priv.h" + +static struct pkey_index_qp_list *get_pkey_idx_qp_list(struct ib_port_pkey *pp) +{ + struct pkey_index_qp_list *pkey = NULL; + struct pkey_index_qp_list *tmp_pkey; + struct ib_device *dev = pp->sec->dev; + + spin_lock(&dev->port_pkey_list[pp->port_num].list_lock); + list_for_each_entry(tmp_pkey, + &dev->port_pkey_list[pp->port_num].pkey_list, + pkey_index_list) { + if (tmp_pkey->pkey_index == pp->pkey_index) { + pkey = tmp_pkey; + break; + } + } + spin_unlock(&dev->port_pkey_list[pp->port_num].list_lock); + return pkey; +} + +static int get_pkey_and_subnet_prefix(struct ib_port_pkey *pp, + u16 *pkey, + u64 *subnet_prefix) +{ + struct ib_device *dev = pp->sec->dev; + int ret; + + ret = ib_get_cached_pkey(dev, pp->port_num, pp->pkey_index, pkey); + if (ret) + return ret; + + ret = ib_get_cached_subnet_prefix(dev, pp->port_num, subnet_prefix); + + return ret; +} + +static int enforce_qp_pkey_security(u16 pkey, + u64 subnet_prefix, + struct ib_qp_security *qp_sec) +{ + struct ib_qp_security *shared_qp_sec; + int ret; + + ret = security_ib_pkey_access(qp_sec->security, subnet_prefix, pkey); + if (ret) + return ret; + + if (qp_sec->qp == qp_sec->qp->real_qp) { + list_for_each_entry(shared_qp_sec, + &qp_sec->shared_qp_list, + shared_qp_list) { + ret = security_ib_pkey_access(shared_qp_sec->security, + subnet_prefix, + pkey); + if (ret) + return ret; + } + } + return 0; +} + +/* The caller of this function must hold the QP security + * mutex of the QP of the security structure in *pps. + * + * It takes separate ports_pkeys and security structure + * because in some cases the pps will be for a new settings + * or the pps will be for the real QP and security structure + * will be for a shared QP. + */ +static int check_qp_port_pkey_settings(struct ib_ports_pkeys *pps, + struct ib_qp_security *sec) +{ + u64 subnet_prefix; + u16 pkey; + int ret = 0; + + if (!pps) + return 0; + + if (pps->main.state != IB_PORT_PKEY_NOT_VALID) { + ret = get_pkey_and_subnet_prefix(&pps->main, + &pkey, + &subnet_prefix); + if (ret) + return ret; + + ret = enforce_qp_pkey_security(pkey, + subnet_prefix, + sec); + if (ret) + return ret; + } + + if (pps->alt.state != IB_PORT_PKEY_NOT_VALID) { + ret = get_pkey_and_subnet_prefix(&pps->alt, + &pkey, + &subnet_prefix); + if (ret) + return ret; + + ret = enforce_qp_pkey_security(pkey, + subnet_prefix, + sec); + } + + return ret; +} + +/* The caller of this function must hold the QP security + * mutex. + */ +static void qp_to_error(struct ib_qp_security *sec) +{ + struct ib_qp_security *shared_qp_sec; + struct ib_qp_attr attr = { + .qp_state = IB_QPS_ERR + }; + struct ib_event event = { + .event = IB_EVENT_QP_FATAL + }; + + /* If the QP is in the process of being destroyed + * the qp pointer in the security structure is + * undefined. It cannot be modified now. + */ + if (sec->destroying) + return; + + ib_modify_qp(sec->qp, + &attr, + IB_QP_STATE); + + if (sec->qp->event_handler && sec->qp->qp_context) { + event.element.qp = sec->qp; + sec->qp->event_handler(&event, + sec->qp->qp_context); + } + + list_for_each_entry(shared_qp_sec, + &sec->shared_qp_list, + shared_qp_list) { + struct ib_qp *qp = shared_qp_sec->qp; + + if (qp->event_handler && qp->qp_context) { + event.element.qp = qp; + event.device = qp->device; + qp->event_handler(&event, + qp->qp_context); + } + } +} + +static inline void check_pkey_qps(struct pkey_index_qp_list *pkey, + struct ib_device *device, + u8 port_num, + u64 subnet_prefix) +{ + struct ib_port_pkey *pp, *tmp_pp; + bool comp; + LIST_HEAD(to_error_list); + u16 pkey_val; + + if (!ib_get_cached_pkey(device, + port_num, + pkey->pkey_index, + &pkey_val)) { + spin_lock(&pkey->qp_list_lock); + list_for_each_entry(pp, &pkey->qp_list, qp_list) { + if (atomic_read(&pp->sec->error_list_count)) + continue; + + if (enforce_qp_pkey_security(pkey_val, + subnet_prefix, + pp->sec)) { + atomic_inc(&pp->sec->error_list_count); + list_add(&pp->to_error_list, + &to_error_list); + } + } + spin_unlock(&pkey->qp_list_lock); + } + + list_for_each_entry_safe(pp, + tmp_pp, + &to_error_list, + to_error_list) { + mutex_lock(&pp->sec->mutex); + qp_to_error(pp->sec); + list_del(&pp->to_error_list); + atomic_dec(&pp->sec->error_list_count); + comp = pp->sec->destroying; + mutex_unlock(&pp->sec->mutex); + + if (comp) + complete(&pp->sec->error_complete); + } +} + +/* The caller of this function must hold the QP security + * mutex. + */ +static int port_pkey_list_insert(struct ib_port_pkey *pp) +{ + struct pkey_index_qp_list *tmp_pkey; + struct pkey_index_qp_list *pkey; + struct ib_device *dev; + u8 port_num = pp->port_num; + int ret = 0; + + if (pp->state != IB_PORT_PKEY_VALID) + return 0; + + dev = pp->sec->dev; + + pkey = get_pkey_idx_qp_list(pp); + + if (!pkey) { + bool found = false; + + pkey = kzalloc(sizeof(*pkey), GFP_KERNEL); + if (!pkey) + return -ENOMEM; + + spin_lock(&dev->port_pkey_list[port_num].list_lock); + /* Check for the PKey again. A racing process may + * have created it. + */ + list_for_each_entry(tmp_pkey, + &dev->port_pkey_list[port_num].pkey_list, + pkey_index_list) { + if (tmp_pkey->pkey_index == pp->pkey_index) { + kfree(pkey); + pkey = tmp_pkey; + found = true; + break; + } + } + + if (!found) { + pkey->pkey_index = pp->pkey_index; + spin_lock_init(&pkey->qp_list_lock); + INIT_LIST_HEAD(&pkey->qp_list); + list_add(&pkey->pkey_index_list, + &dev->port_pkey_list[port_num].pkey_list); + } + spin_unlock(&dev->port_pkey_list[port_num].list_lock); + } + + spin_lock(&pkey->qp_list_lock); + list_add(&pp->qp_list, &pkey->qp_list); + spin_unlock(&pkey->qp_list_lock); + + pp->state = IB_PORT_PKEY_LISTED; + + return ret; +} + +/* The caller of this function must hold the QP security + * mutex. + */ +static void port_pkey_list_remove(struct ib_port_pkey *pp) +{ + struct pkey_index_qp_list *pkey; + + if (pp->state != IB_PORT_PKEY_LISTED) + return; + + pkey = get_pkey_idx_qp_list(pp); + + spin_lock(&pkey->qp_list_lock); + list_del(&pp->qp_list); + spin_unlock(&pkey->qp_list_lock); + + /* The setting may still be valid, i.e. after + * a destroy has failed for example. + */ + pp->state = IB_PORT_PKEY_VALID; +} + +static void destroy_qp_security(struct ib_qp_security *sec) +{ + security_ib_free_security(sec->security); + kfree(sec->ports_pkeys); + kfree(sec); +} + +/* The caller of this function must hold the QP security + * mutex. + */ +static struct ib_ports_pkeys *get_new_pps(const struct ib_qp *qp, + const struct ib_qp_attr *qp_attr, + int qp_attr_mask) +{ + struct ib_ports_pkeys *new_pps; + struct ib_ports_pkeys *qp_pps = qp->qp_sec->ports_pkeys; + + new_pps = kzalloc(sizeof(*new_pps), GFP_KERNEL); + if (!new_pps) + return NULL; + + if (qp_attr_mask & (IB_QP_PKEY_INDEX | IB_QP_PORT)) { + if (!qp_pps) { + new_pps->main.port_num = qp_attr->port_num; + new_pps->main.pkey_index = qp_attr->pkey_index; + } else { + new_pps->main.port_num = (qp_attr_mask & IB_QP_PORT) ? + qp_attr->port_num : + qp_pps->main.port_num; + + new_pps->main.pkey_index = + (qp_attr_mask & IB_QP_PKEY_INDEX) ? + qp_attr->pkey_index : + qp_pps->main.pkey_index; + } + new_pps->main.state = IB_PORT_PKEY_VALID; + } else if (qp_pps) { + new_pps->main.port_num = qp_pps->main.port_num; + new_pps->main.pkey_index = qp_pps->main.pkey_index; + if (qp_pps->main.state != IB_PORT_PKEY_NOT_VALID) + new_pps->main.state = IB_PORT_PKEY_VALID; + } + + if (qp_attr_mask & IB_QP_ALT_PATH) { + new_pps->alt.port_num = qp_attr->alt_port_num; + new_pps->alt.pkey_index = qp_attr->alt_pkey_index; + new_pps->alt.state = IB_PORT_PKEY_VALID; + } else if (qp_pps) { + new_pps->alt.port_num = qp_pps->alt.port_num; + new_pps->alt.pkey_index = qp_pps->alt.pkey_index; + if (qp_pps->alt.state != IB_PORT_PKEY_NOT_VALID) + new_pps->alt.state = IB_PORT_PKEY_VALID; + } + + new_pps->main.sec = qp->qp_sec; + new_pps->alt.sec = qp->qp_sec; + return new_pps; +} + +int ib_open_shared_qp_security(struct ib_qp *qp, struct ib_device *dev) +{ + struct ib_qp *real_qp = qp->real_qp; + int ret; + + ret = ib_create_qp_security(qp, dev); + + if (ret) + return ret; + + mutex_lock(&real_qp->qp_sec->mutex); + ret = check_qp_port_pkey_settings(real_qp->qp_sec->ports_pkeys, + qp->qp_sec); + + if (ret) + goto ret; + + if (qp != real_qp) + list_add(&qp->qp_sec->shared_qp_list, + &real_qp->qp_sec->shared_qp_list); +ret: + mutex_unlock(&real_qp->qp_sec->mutex); + if (ret) + destroy_qp_security(qp->qp_sec); + + return ret; +} + +void ib_close_shared_qp_security(struct ib_qp_security *sec) +{ + struct ib_qp *real_qp = sec->qp->real_qp; + + mutex_lock(&real_qp->qp_sec->mutex); + list_del(&sec->shared_qp_list); + mutex_unlock(&real_qp->qp_sec->mutex); + + destroy_qp_security(sec); +} + +int ib_create_qp_security(struct ib_qp *qp, struct ib_device *dev) +{ + int ret; + + qp->qp_sec = kzalloc(sizeof(*qp->qp_sec), GFP_KERNEL); + if (!qp->qp_sec) + return -ENOMEM; + + qp->qp_sec->qp = qp; + qp->qp_sec->dev = dev; + mutex_init(&qp->qp_sec->mutex); + INIT_LIST_HEAD(&qp->qp_sec->shared_qp_list); + atomic_set(&qp->qp_sec->error_list_count, 0); + init_completion(&qp->qp_sec->error_complete); + ret = security_ib_alloc_security(&qp->qp_sec->security); + if (ret) + kfree(qp->qp_sec); + + return ret; +} +EXPORT_SYMBOL(ib_create_qp_security); + +void ib_destroy_qp_security_begin(struct ib_qp_security *sec) +{ + mutex_lock(&sec->mutex); + + /* Remove the QP from the lists so it won't get added to + * a to_error_list during the destroy process. + */ + if (sec->ports_pkeys) { + port_pkey_list_remove(&sec->ports_pkeys->main); + port_pkey_list_remove(&sec->ports_pkeys->alt); + } + + /* If the QP is already in one or more of those lists + * the destroying flag will ensure the to error flow + * doesn't operate on an undefined QP. + */ + sec->destroying = true; + + /* Record the error list count to know how many completions + * to wait for. + */ + sec->error_comps_pending = atomic_read(&sec->error_list_count); + + mutex_unlock(&sec->mutex); +} + +void ib_destroy_qp_security_abort(struct ib_qp_security *sec) +{ + int ret; + int i; + + /* If a concurrent cache update is in progress this + * QP security could be marked for an error state + * transition. Wait for this to complete. + */ + for (i = 0; i < sec->error_comps_pending; i++) + wait_for_completion(&sec->error_complete); + + mutex_lock(&sec->mutex); + sec->destroying = false; + + /* Restore the position in the lists and verify + * access is still allowed in case a cache update + * occurred while attempting to destroy. + * + * Because these setting were listed already + * and removed during ib_destroy_qp_security_begin + * we know the pkey_index_qp_list for the PKey + * already exists so port_pkey_list_insert won't fail. + */ + if (sec->ports_pkeys) { + port_pkey_list_insert(&sec->ports_pkeys->main); + port_pkey_list_insert(&sec->ports_pkeys->alt); + } + + ret = check_qp_port_pkey_settings(sec->ports_pkeys, sec); + if (ret) + qp_to_error(sec); + + mutex_unlock(&sec->mutex); +} + +void ib_destroy_qp_security_end(struct ib_qp_security *sec) +{ + int i; + + /* If a concurrent cache update is occurring we must + * wait until this QP security structure is processed + * in the QP to error flow before destroying it because + * the to_error_list is in use. + */ + for (i = 0; i < sec->error_comps_pending; i++) + wait_for_completion(&sec->error_complete); + + destroy_qp_security(sec); +} + +void ib_security_cache_change(struct ib_device *device, + u8 port_num, + u64 subnet_prefix) +{ + struct pkey_index_qp_list *pkey; + + list_for_each_entry(pkey, + &device->port_pkey_list[port_num].pkey_list, + pkey_index_list) { + check_pkey_qps(pkey, + device, + port_num, + subnet_prefix); + } +} + +void ib_security_destroy_port_pkey_list(struct ib_device *device) +{ + struct pkey_index_qp_list *pkey, *tmp_pkey; + int i; + + for (i = rdma_start_port(device); i <= rdma_end_port(device); i++) { + spin_lock(&device->port_pkey_list[i].list_lock); + list_for_each_entry_safe(pkey, + tmp_pkey, + &device->port_pkey_list[i].pkey_list, + pkey_index_list) { + list_del(&pkey->pkey_index_list); + kfree(pkey); + } + spin_unlock(&device->port_pkey_list[i].list_lock); + } +} + +int ib_security_modify_qp(struct ib_qp *qp, + struct ib_qp_attr *qp_attr, + int qp_attr_mask, + struct ib_udata *udata) +{ + int ret = 0; + struct ib_ports_pkeys *tmp_pps; + struct ib_ports_pkeys *new_pps; + bool special_qp = (qp->qp_type == IB_QPT_SMI || + qp->qp_type == IB_QPT_GSI || + qp->qp_type >= IB_QPT_RESERVED1); + bool pps_change = ((qp_attr_mask & (IB_QP_PKEY_INDEX | IB_QP_PORT)) || + (qp_attr_mask & IB_QP_ALT_PATH)); + + if (pps_change && !special_qp) { + mutex_lock(&qp->qp_sec->mutex); + new_pps = get_new_pps(qp, + qp_attr, + qp_attr_mask); + + /* Add this QP to the lists for the new port + * and pkey settings before checking for permission + * in case there is a concurrent cache update + * occurring. Walking the list for a cache change + * doesn't acquire the security mutex unless it's + * sending the QP to error. + */ + ret = port_pkey_list_insert(&new_pps->main); + + if (!ret) + ret = port_pkey_list_insert(&new_pps->alt); + + if (!ret) + ret = check_qp_port_pkey_settings(new_pps, + qp->qp_sec); + } + + if (!ret) + ret = qp->device->modify_qp(qp->real_qp, + qp_attr, + qp_attr_mask, + udata); + + if (pps_change && !special_qp) { + /* Clean up the lists and free the appropriate + * ports_pkeys structure. + */ + if (ret) { + tmp_pps = new_pps; + } else { + tmp_pps = qp->qp_sec->ports_pkeys; + qp->qp_sec->ports_pkeys = new_pps; + } + + if (tmp_pps) { + port_pkey_list_remove(&tmp_pps->main); + port_pkey_list_remove(&tmp_pps->alt); + } + kfree(tmp_pps); + mutex_unlock(&qp->qp_sec->mutex); + } + 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) +{ + u64 subnet_prefix; + u16 pkey; + int ret; + + ret = ib_get_cached_pkey(dev, port_num, pkey_index, &pkey); + if (ret) + return ret; + + ret = ib_get_cached_subnet_prefix(dev, port_num, &subnet_prefix); + + if (ret) + return ret; + + 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, + void *data) +{ + struct ib_mad_agent *ag = container_of(nb, struct ib_mad_agent, lsm_nb); + + if (event != LSM_POLICY_CHANGE) + return NOTIFY_DONE; + + ag->smp_allowed = !security_ib_endport_manage_subnet(ag->security, + ag->device->name, + ag->port_num); + + return NOTIFY_OK; +} + +int ib_mad_agent_security_setup(struct ib_mad_agent *agent, + enum ib_qp_type qp_type) +{ + int ret; + + ret = security_ib_alloc_security(&agent->security); + if (ret) + return ret; + + if (qp_type != IB_QPT_SMI) + return 0; + + ret = security_ib_endport_manage_subnet(agent->security, + agent->device->name, + agent->port_num); + if (ret) + return ret; + + agent->lsm_nb.notifier_call = ib_mad_agent_security_change; + ret = register_lsm_notifier(&agent->lsm_nb); + if (ret) + return ret; + + agent->smp_allowed = true; + agent->lsm_nb_reg = true; + return 0; +} + +void ib_mad_agent_security_cleanup(struct ib_mad_agent *agent) +{ + security_ib_free_security(agent->security); + if (agent->lsm_nb_reg) + unregister_lsm_notifier(&agent->lsm_nb); +} + +int ib_mad_enforce_security(struct ib_mad_agent_private *map, u16 pkey_index) +{ + int ret; + + if (map->agent.qp->qp_type == IB_QPT_SMI && !map->agent.smp_allowed) + return -EACCES; + + ret = ib_security_pkey_access(map->agent.device, + map->agent.port_num, + pkey_index, + map->agent.security); + + if (ret) + return ret; + + return 0; +} + +#endif /* CONFIG_SECURITY_INFINIBAND */ diff --git a/drivers/infiniband/core/sysfs.c b/drivers/infiniband/core/sysfs.c index c1fb545e8d78..7ebe1ef23652 100644 --- a/drivers/infiniband/core/sysfs.c +++ b/drivers/infiniband/core/sysfs.c @@ -253,6 +253,10 @@ static ssize_t rate_show(struct ib_port *p, struct port_attribute *unused, speed = " EDR"; rate = 250; break; + case IB_SPEED_HDR: + speed = " HDR"; + rate = 500; + break; case IB_SPEED_SDR: default: /* default to SDR for invalid rates */ rate = 25; @@ -1258,7 +1262,7 @@ int ib_device_register_sysfs(struct ib_device *device, int ret; int i; - device->dev.parent = device->dma_device; + WARN_ON_ONCE(!device->dev.parent); ret = dev_set_name(class_dev, "%s", device->name); if (ret) return ret; @@ -1301,7 +1305,7 @@ err_put: free_port_list_attributes(device); err_unregister: - device_unregister(class_dev); + device_del(class_dev); err: return ret; diff --git a/drivers/infiniband/core/ucm.c b/drivers/infiniband/core/ucm.c index e0a995b85a2d..112099c86a19 100644 --- a/drivers/infiniband/core/ucm.c +++ b/drivers/infiniband/core/ucm.c @@ -702,10 +702,10 @@ static int ib_ucm_alloc_data(const void **dest, u64 src, u32 len) return 0; } -static int ib_ucm_path_get(struct ib_sa_path_rec **path, u64 src) +static int ib_ucm_path_get(struct sa_path_rec **path, u64 src) { struct ib_user_path_rec upath; - struct ib_sa_path_rec *sa_path; + struct sa_path_rec *sa_path; *path = NULL; @@ -962,7 +962,7 @@ static ssize_t ib_ucm_send_lap(struct ib_ucm_file *file, int in_len, int out_len) { struct ib_ucm_context *ctx; - struct ib_sa_path_rec *path = NULL; + struct sa_path_rec *path = NULL; struct ib_ucm_lap cmd; const void *data = NULL; int result; @@ -1205,12 +1205,15 @@ static void ib_ucm_release_dev(struct device *dev) struct ib_ucm_device *ucm_dev; ucm_dev = container_of(dev, struct ib_ucm_device, dev); - cdev_del(&ucm_dev->cdev); + kfree(ucm_dev); +} + +static void ib_ucm_free_dev(struct ib_ucm_device *ucm_dev) +{ if (ucm_dev->devnum < IB_UCM_MAX_DEVICES) clear_bit(ucm_dev->devnum, dev_map); else clear_bit(ucm_dev->devnum - IB_UCM_MAX_DEVICES, overflow_map); - kfree(ucm_dev); } static const struct file_operations ucm_fops = { @@ -1266,7 +1269,9 @@ static void ib_ucm_add_one(struct ib_device *device) if (!ucm_dev) return; + device_initialize(&ucm_dev->dev); ucm_dev->ib_dev = device; + ucm_dev->dev.release = ib_ucm_release_dev; devnum = find_first_zero_bit(dev_map, IB_UCM_MAX_DEVICES); if (devnum >= IB_UCM_MAX_DEVICES) { @@ -1286,16 +1291,14 @@ static void ib_ucm_add_one(struct ib_device *device) cdev_init(&ucm_dev->cdev, &ucm_fops); ucm_dev->cdev.owner = THIS_MODULE; kobject_set_name(&ucm_dev->cdev.kobj, "ucm%d", ucm_dev->devnum); - if (cdev_add(&ucm_dev->cdev, base, 1)) - goto err; ucm_dev->dev.class = &cm_class; - ucm_dev->dev.parent = device->dma_device; - ucm_dev->dev.devt = ucm_dev->cdev.dev; - ucm_dev->dev.release = ib_ucm_release_dev; + ucm_dev->dev.parent = device->dev.parent; + ucm_dev->dev.devt = base; + dev_set_name(&ucm_dev->dev, "ucm%d", ucm_dev->devnum); - if (device_register(&ucm_dev->dev)) - goto err_cdev; + if (cdev_device_add(&ucm_dev->cdev, &ucm_dev->dev)) + goto err_devnum; if (device_create_file(&ucm_dev->dev, &dev_attr_ibdev)) goto err_dev; @@ -1304,15 +1307,11 @@ static void ib_ucm_add_one(struct ib_device *device) return; err_dev: - device_unregister(&ucm_dev->dev); -err_cdev: - cdev_del(&ucm_dev->cdev); - if (ucm_dev->devnum < IB_UCM_MAX_DEVICES) - clear_bit(devnum, dev_map); - else - clear_bit(devnum, overflow_map); + cdev_device_del(&ucm_dev->cdev, &ucm_dev->dev); +err_devnum: + ib_ucm_free_dev(ucm_dev); err: - kfree(ucm_dev); + put_device(&ucm_dev->dev); return; } @@ -1323,7 +1322,9 @@ static void ib_ucm_remove_one(struct ib_device *device, void *client_data) if (!ucm_dev) return; - device_unregister(&ucm_dev->dev); + cdev_device_del(&ucm_dev->cdev, &ucm_dev->dev); + ib_ucm_free_dev(ucm_dev); + put_device(&ucm_dev->dev); } static CLASS_ATTR_STRING(abi_version, S_IRUGO, diff --git a/drivers/infiniband/core/ucma.c b/drivers/infiniband/core/ucma.c index e12f8faf8c23..276f0ef835bd 100644 --- a/drivers/infiniband/core/ucma.c +++ b/drivers/infiniband/core/ucma.c @@ -898,11 +898,18 @@ static ssize_t ucma_query_path(struct ucma_context *ctx, for (i = 0, out_len -= sizeof(*resp); i < resp->num_paths && out_len > sizeof(struct ib_path_rec_data); i++, out_len -= sizeof(struct ib_path_rec_data)) { + struct sa_path_rec *rec = &ctx->cm_id->route.path_rec[i]; resp->path_data[i].flags = IB_PATH_GMP | IB_PATH_PRIMARY | IB_PATH_BIDIRECTIONAL; - ib_sa_pack_path(&ctx->cm_id->route.path_rec[i], - &resp->path_data[i].path_rec); + if (rec->rec_type == SA_PATH_REC_TYPE_IB) { + ib_sa_pack_path(rec, &resp->path_data[i].path_rec); + } else { + struct sa_path_rec ib; + + sa_convert_path_opa_to_ib(&ib, rec); + ib_sa_pack_path(&ib, &resp->path_data[i].path_rec); + } } if (copy_to_user(response, resp, @@ -1197,7 +1204,7 @@ static int ucma_set_option_id(struct ucma_context *ctx, int optname, static int ucma_set_ib_path(struct ucma_context *ctx, struct ib_path_rec_data *path_data, size_t optlen) { - struct ib_sa_path_rec sa_path; + struct sa_path_rec sa_path; struct rdma_cm_event event; int ret; @@ -1215,8 +1222,17 @@ static int ucma_set_ib_path(struct ucma_context *ctx, memset(&sa_path, 0, sizeof(sa_path)); + sa_path.rec_type = SA_PATH_REC_TYPE_IB; ib_sa_unpack_path(path_data->path_rec, &sa_path); - ret = rdma_set_ib_paths(ctx->cm_id, &sa_path, 1); + + if (rdma_cap_opa_ah(ctx->cm_id->device, ctx->cm_id->port_num)) { + struct sa_path_rec opa; + + sa_convert_path_ib_to_opa(&opa, &sa_path); + ret = rdma_set_ib_paths(ctx->cm_id, &opa, 1); + } else { + ret = rdma_set_ib_paths(ctx->cm_id, &sa_path, 1); + } if (ret) return ret; diff --git a/drivers/infiniband/core/umem.c b/drivers/infiniband/core/umem.c index 4609b921f899..21e60b1e2ff4 100644 --- a/drivers/infiniband/core/umem.c +++ b/drivers/infiniband/core/umem.c @@ -34,7 +34,8 @@ #include <linux/mm.h> #include <linux/dma-mapping.h> -#include <linux/sched.h> +#include <linux/sched/signal.h> +#include <linux/sched/mm.h> #include <linux/export.h> #include <linux/hugetlb.h> #include <linux/slab.h> @@ -57,7 +58,7 @@ static void __ib_umem_release(struct ib_device *dev, struct ib_umem *umem, int d for_each_sg(umem->sg_head.sgl, sg, umem->npages, i) { page = sg_page(sg); - if (umem->writable && dirty) + if (!PageDirty(page) && umem->writable && dirty) set_page_dirty_lock(page); put_page(page); } @@ -99,9 +100,6 @@ struct ib_umem *ib_umem_get(struct ib_ucontext *context, unsigned long addr, if (dmasync) dma_attrs |= DMA_ATTR_WRITE_BARRIER; - if (!size) - return ERR_PTR(-EINVAL); - /* * If the combination of the addr and size requested for this memory * region causes an integer overflow, return error. @@ -117,11 +115,11 @@ struct ib_umem *ib_umem_get(struct ib_ucontext *context, unsigned long addr, if (!umem) return ERR_PTR(-ENOMEM); - umem->context = context; - umem->length = size; - umem->address = addr; - umem->page_size = PAGE_SIZE; - umem->pid = get_task_pid(current, PIDTYPE_PID); + umem->context = context; + umem->length = size; + umem->address = addr; + umem->page_shift = PAGE_SHIFT; + umem->pid = get_task_pid(current, PIDTYPE_PID); /* * We ask for writable memory if any of the following * access flags are set. "Local write" and "remote write" @@ -135,7 +133,7 @@ struct ib_umem *ib_umem_get(struct ib_ucontext *context, unsigned long addr, if (access & IB_ACCESS_ON_DEMAND) { put_pid(umem->pid); - ret = ib_umem_odp_get(context, umem); + ret = ib_umem_odp_get(context, umem, access); if (ret) { kfree(umem); return ERR_PTR(ret); @@ -317,7 +315,6 @@ EXPORT_SYMBOL(ib_umem_release); int ib_umem_page_count(struct ib_umem *umem) { - int shift; int i; int n; struct scatterlist *sg; @@ -325,11 +322,9 @@ int ib_umem_page_count(struct ib_umem *umem) if (umem->odp_data) return ib_umem_num_pages(umem); - shift = ilog2(umem->page_size); - n = 0; for_each_sg(umem->sg_head.sgl, sg, umem->nmap, i) - n += sg_dma_len(sg) >> shift; + n += sg_dma_len(sg) >> umem->page_shift; return n; } diff --git a/drivers/infiniband/core/umem_odp.c b/drivers/infiniband/core/umem_odp.c index 6b079a31dced..8c4ec564e495 100644 --- a/drivers/infiniband/core/umem_odp.c +++ b/drivers/infiniband/core/umem_odp.c @@ -32,10 +32,13 @@ #include <linux/types.h> #include <linux/sched.h> +#include <linux/sched/mm.h> +#include <linux/sched/task.h> #include <linux/pid.h> #include <linux/slab.h> #include <linux/export.h> #include <linux/vmalloc.h> +#include <linux/hugetlb.h> #include <rdma/ib_verbs.h> #include <rdma/ib_umem.h> @@ -239,7 +242,73 @@ static const struct mmu_notifier_ops ib_umem_notifiers = { .invalidate_range_end = ib_umem_notifier_invalidate_range_end, }; -int ib_umem_odp_get(struct ib_ucontext *context, struct ib_umem *umem) +struct ib_umem *ib_alloc_odp_umem(struct ib_ucontext *context, + unsigned long addr, + size_t size) +{ + struct ib_umem *umem; + struct ib_umem_odp *odp_data; + int pages = size >> PAGE_SHIFT; + int ret; + + umem = kzalloc(sizeof(*umem), GFP_KERNEL); + if (!umem) + return ERR_PTR(-ENOMEM); + + umem->context = context; + umem->length = size; + umem->address = addr; + umem->page_shift = PAGE_SHIFT; + umem->writable = 1; + + odp_data = kzalloc(sizeof(*odp_data), GFP_KERNEL); + if (!odp_data) { + ret = -ENOMEM; + goto out_umem; + } + odp_data->umem = umem; + + mutex_init(&odp_data->umem_mutex); + init_completion(&odp_data->notifier_completion); + + odp_data->page_list = vzalloc(pages * sizeof(*odp_data->page_list)); + if (!odp_data->page_list) { + ret = -ENOMEM; + goto out_odp_data; + } + + odp_data->dma_list = vzalloc(pages * sizeof(*odp_data->dma_list)); + if (!odp_data->dma_list) { + ret = -ENOMEM; + goto out_page_list; + } + + down_write(&context->umem_rwsem); + context->odp_mrs_count++; + rbt_ib_umem_insert(&odp_data->interval_tree, &context->umem_tree); + if (likely(!atomic_read(&context->notifier_count))) + odp_data->mn_counters_active = true; + else + list_add(&odp_data->no_private_counters, + &context->no_private_counters); + up_write(&context->umem_rwsem); + + umem->odp_data = odp_data; + + return umem; + +out_page_list: + vfree(odp_data->page_list); +out_odp_data: + kfree(odp_data); +out_umem: + kfree(umem); + return ERR_PTR(ret); +} +EXPORT_SYMBOL(ib_alloc_odp_umem); + +int ib_umem_odp_get(struct ib_ucontext *context, struct ib_umem *umem, + int access) { int ret_val; struct pid *our_pid; @@ -248,6 +317,24 @@ int ib_umem_odp_get(struct ib_ucontext *context, struct ib_umem *umem) if (!mm) return -EINVAL; + if (access & IB_ACCESS_HUGETLB) { + struct vm_area_struct *vma; + struct hstate *h; + + down_read(&mm->mmap_sem); + vma = find_vma(mm, ib_umem_start(umem)); + if (!vma || !is_vm_hugetlb_page(vma)) { + up_read(&mm->mmap_sem); + return -EINVAL; + } + h = hstate_vma(vma); + umem->page_shift = huge_page_shift(h); + up_read(&mm->mmap_sem); + umem->hugetlb = 1; + } else { + umem->hugetlb = 0; + } + /* Prevent creating ODP MRs in child processes */ rcu_read_lock(); our_pid = get_task_pid(current->group_leader, PIDTYPE_PID); @@ -258,7 +345,6 @@ int ib_umem_odp_get(struct ib_ucontext *context, struct ib_umem *umem) goto out_mm; } - umem->hugetlb = 0; umem->odp_data = kzalloc(sizeof(*umem->odp_data), GFP_KERNEL); if (!umem->odp_data) { ret_val = -ENOMEM; @@ -270,18 +356,20 @@ int ib_umem_odp_get(struct ib_ucontext *context, struct ib_umem *umem) init_completion(&umem->odp_data->notifier_completion); - umem->odp_data->page_list = vzalloc(ib_umem_num_pages(umem) * + if (ib_umem_num_pages(umem)) { + umem->odp_data->page_list = vzalloc(ib_umem_num_pages(umem) * sizeof(*umem->odp_data->page_list)); - if (!umem->odp_data->page_list) { - ret_val = -ENOMEM; - goto out_odp_data; - } + if (!umem->odp_data->page_list) { + ret_val = -ENOMEM; + goto out_odp_data; + } - umem->odp_data->dma_list = vzalloc(ib_umem_num_pages(umem) * + umem->odp_data->dma_list = vzalloc(ib_umem_num_pages(umem) * sizeof(*umem->odp_data->dma_list)); - if (!umem->odp_data->dma_list) { - ret_val = -ENOMEM; - goto out_page_list; + if (!umem->odp_data->dma_list) { + ret_val = -ENOMEM; + goto out_page_list; + } } /* @@ -435,7 +523,6 @@ out: static int ib_umem_odp_map_dma_single_page( struct ib_umem *umem, int page_index, - u64 base_virt_addr, struct page *page, u64 access_mask, unsigned long current_seq) @@ -458,7 +545,7 @@ static int ib_umem_odp_map_dma_single_page( if (!(umem->odp_data->dma_list[page_index])) { dma_addr = ib_dma_map_page(dev, page, - 0, PAGE_SIZE, + 0, BIT(umem->page_shift), DMA_BIDIRECTIONAL); if (ib_dma_mapping_error(dev, dma_addr)) { ret = -EFAULT; @@ -466,6 +553,7 @@ static int ib_umem_odp_map_dma_single_page( } umem->odp_data->dma_list[page_index] = dma_addr | access_mask; umem->odp_data->page_list[page_index] = page; + umem->npages++; stored_page = 1; } else if (umem->odp_data->page_list[page_index] == page) { umem->odp_data->dma_list[page_index] |= access_mask; @@ -485,8 +573,9 @@ out: if (remove_existing_mapping && umem->context->invalidate_range) { invalidate_page_trampoline( umem, - base_virt_addr + (page_index * PAGE_SIZE), - base_virt_addr + ((page_index+1)*PAGE_SIZE), + ib_umem_start(umem) + (page_index >> umem->page_shift), + ib_umem_start(umem) + ((page_index + 1) >> + umem->page_shift), NULL); ret = -EAGAIN; } @@ -505,7 +594,8 @@ out: * for failure. * An -EAGAIN error code is returned when a concurrent mmu notifier prevents * the function from completing its task. - * + * An -ENOENT error code indicates that userspace process is being terminated + * and mm was already destroyed. * @umem: the umem to map and pin * @user_virt: the address from which we need to map. * @bcnt: the minimal number of bytes to pin and map. The mapping might be @@ -524,10 +614,10 @@ int ib_umem_odp_map_dma_pages(struct ib_umem *umem, u64 user_virt, u64 bcnt, struct task_struct *owning_process = NULL; struct mm_struct *owning_mm = NULL; struct page **local_page_list = NULL; - u64 off; - int j, k, ret = 0, start_idx, npages = 0; - u64 base_virt_addr; + u64 page_mask, off; + int j, k, ret = 0, start_idx, npages = 0, page_shift; unsigned int flags = 0; + phys_addr_t p = 0; if (access_mask == 0) return -EINVAL; @@ -540,9 +630,10 @@ int ib_umem_odp_map_dma_pages(struct ib_umem *umem, u64 user_virt, u64 bcnt, if (!local_page_list) return -ENOMEM; - off = user_virt & (~PAGE_MASK); - user_virt = user_virt & PAGE_MASK; - base_virt_addr = user_virt; + page_shift = umem->page_shift; + page_mask = ~(BIT(page_shift) - 1); + off = user_virt & (~page_mask); + user_virt = user_virt & page_mask; bcnt += off; /* Charge for the first page offset as well. */ owning_process = get_pid_task(umem->context->tgid, PIDTYPE_PID); @@ -553,20 +644,20 @@ int ib_umem_odp_map_dma_pages(struct ib_umem *umem, u64 user_virt, u64 bcnt, owning_mm = get_task_mm(owning_process); if (owning_mm == NULL) { - ret = -EINVAL; + ret = -ENOENT; goto out_put_task; } if (access_mask & ODP_WRITE_ALLOWED_BIT) flags |= FOLL_WRITE; - start_idx = (user_virt - ib_umem_start(umem)) >> PAGE_SHIFT; + start_idx = (user_virt - ib_umem_start(umem)) >> page_shift; k = start_idx; while (bcnt > 0) { - const size_t gup_num_pages = - min_t(size_t, ALIGN(bcnt, PAGE_SIZE) / PAGE_SIZE, - PAGE_SIZE / sizeof(struct page *)); + const size_t gup_num_pages = min_t(size_t, + (bcnt + BIT(page_shift) - 1) >> page_shift, + PAGE_SIZE / sizeof(struct page *)); down_read(&owning_mm->mmap_sem); /* @@ -585,14 +676,25 @@ int ib_umem_odp_map_dma_pages(struct ib_umem *umem, u64 user_virt, u64 bcnt, break; bcnt -= min_t(size_t, npages << PAGE_SHIFT, bcnt); - user_virt += npages << PAGE_SHIFT; mutex_lock(&umem->odp_data->umem_mutex); - for (j = 0; j < npages; ++j) { + for (j = 0; j < npages; j++, user_virt += PAGE_SIZE) { + if (user_virt & ~page_mask) { + p += PAGE_SIZE; + if (page_to_phys(local_page_list[j]) != p) { + ret = -EFAULT; + break; + } + put_page(local_page_list[j]); + continue; + } + ret = ib_umem_odp_map_dma_single_page( - umem, k, base_virt_addr, local_page_list[j], - access_mask, current_seq); + umem, k, local_page_list[j], + access_mask, current_seq); if (ret < 0) break; + + p = page_to_phys(local_page_list[j]); k++; } mutex_unlock(&umem->odp_data->umem_mutex); @@ -636,8 +738,8 @@ void ib_umem_odp_unmap_dma_pages(struct ib_umem *umem, u64 virt, * invalidations, so we must make sure we free each page only * once. */ mutex_lock(&umem->odp_data->umem_mutex); - for (addr = virt; addr < bound; addr += (u64)umem->page_size) { - idx = (addr - ib_umem_start(umem)) / PAGE_SIZE; + for (addr = virt; addr < bound; addr += BIT(umem->page_shift)) { + idx = (addr - ib_umem_start(umem)) >> umem->page_shift; if (umem->odp_data->page_list[idx]) { struct page *page = umem->odp_data->page_list[idx]; dma_addr_t dma = umem->odp_data->dma_list[idx]; @@ -665,6 +767,7 @@ void ib_umem_odp_unmap_dma_pages(struct ib_umem *umem, u64 virt, put_page(page); umem->odp_data->page_list[idx] = NULL; umem->odp_data->dma_list[idx] = 0; + umem->npages--; } } mutex_unlock(&umem->odp_data->umem_mutex); diff --git a/drivers/infiniband/core/umem_rbtree.c b/drivers/infiniband/core/umem_rbtree.c index 727d788448f5..d176597b4d78 100644 --- a/drivers/infiniband/core/umem_rbtree.c +++ b/drivers/infiniband/core/umem_rbtree.c @@ -78,17 +78,32 @@ int rbt_ib_umem_for_each_in_range(struct rb_root *root, void *cookie) { int ret_val = 0; - struct umem_odp_node *node; + struct umem_odp_node *node, *next; struct ib_umem_odp *umem; if (unlikely(start == last)) return ret_val; - for (node = rbt_ib_umem_iter_first(root, start, last - 1); node; - node = rbt_ib_umem_iter_next(node, start, last - 1)) { + for (node = rbt_ib_umem_iter_first(root, start, last - 1); + node; node = next) { + next = rbt_ib_umem_iter_next(node, start, last - 1); umem = container_of(node, struct ib_umem_odp, interval_tree); ret_val = cb(umem->umem, start, last, cookie) || ret_val; } return ret_val; } +EXPORT_SYMBOL(rbt_ib_umem_for_each_in_range); + +struct ib_umem_odp *rbt_ib_umem_lookup(struct rb_root *root, + u64 addr, u64 length) +{ + struct umem_odp_node *node; + + node = rbt_ib_umem_iter_first(root, addr, addr + length - 1); + if (node) + return container_of(node, struct ib_umem_odp, interval_tree); + return NULL; + +} +EXPORT_SYMBOL(rbt_ib_umem_lookup); diff --git a/drivers/infiniband/core/user_mad.c b/drivers/infiniband/core/user_mad.c index 249b403b43a4..36a6f5c8914c 100644 --- a/drivers/infiniband/core/user_mad.c +++ b/drivers/infiniband/core/user_mad.c @@ -197,7 +197,7 @@ static void send_handler(struct ib_mad_agent *agent, struct ib_umad_packet *packet = send_wc->send_buf->context[0]; dequeue_send(file, packet); - ib_destroy_ah(packet->msg->ah); + rdma_destroy_ah(packet->msg->ah); ib_free_send_mad(packet->msg); if (send_wc->status == IB_WC_RESP_TIMEOUT_ERR) { @@ -235,17 +235,19 @@ static void recv_handler(struct ib_mad_agent *agent, packet->mad.hdr.pkey_index = mad_recv_wc->wc->pkey_index; packet->mad.hdr.grh_present = !!(mad_recv_wc->wc->wc_flags & IB_WC_GRH); if (packet->mad.hdr.grh_present) { - struct ib_ah_attr ah_attr; + struct rdma_ah_attr ah_attr; + const struct ib_global_route *grh; ib_init_ah_from_wc(agent->device, agent->port_num, mad_recv_wc->wc, mad_recv_wc->recv_buf.grh, &ah_attr); - packet->mad.hdr.gid_index = ah_attr.grh.sgid_index; - packet->mad.hdr.hop_limit = ah_attr.grh.hop_limit; - packet->mad.hdr.traffic_class = ah_attr.grh.traffic_class; - memcpy(packet->mad.hdr.gid, &ah_attr.grh.dgid, 16); - packet->mad.hdr.flow_label = cpu_to_be32(ah_attr.grh.flow_label); + grh = rdma_ah_read_grh(&ah_attr); + packet->mad.hdr.gid_index = grh->sgid_index; + packet->mad.hdr.hop_limit = grh->hop_limit; + packet->mad.hdr.traffic_class = grh->traffic_class; + memcpy(packet->mad.hdr.gid, &grh->dgid, 16); + packet->mad.hdr.flow_label = cpu_to_be32(grh->flow_label); } if (queue_packet(file, agent, packet)) @@ -449,7 +451,7 @@ static ssize_t ib_umad_write(struct file *filp, const char __user *buf, struct ib_umad_file *file = filp->private_data; struct ib_umad_packet *packet; struct ib_mad_agent *agent; - struct ib_ah_attr ah_attr; + struct rdma_ah_attr ah_attr; struct ib_ah *ah; struct ib_rmpp_mad *rmpp_mad; __be64 *tid; @@ -489,20 +491,22 @@ static ssize_t ib_umad_write(struct file *filp, const char __user *buf, } memset(&ah_attr, 0, sizeof ah_attr); - ah_attr.dlid = be16_to_cpu(packet->mad.hdr.lid); - ah_attr.sl = packet->mad.hdr.sl; - ah_attr.src_path_bits = packet->mad.hdr.path_bits; - ah_attr.port_num = file->port->port_num; + ah_attr.type = rdma_ah_find_type(file->port->ib_dev, + file->port->port_num); + rdma_ah_set_dlid(&ah_attr, be16_to_cpu(packet->mad.hdr.lid)); + rdma_ah_set_sl(&ah_attr, packet->mad.hdr.sl); + rdma_ah_set_path_bits(&ah_attr, packet->mad.hdr.path_bits); + rdma_ah_set_port_num(&ah_attr, file->port->port_num); if (packet->mad.hdr.grh_present) { - ah_attr.ah_flags = IB_AH_GRH; - memcpy(ah_attr.grh.dgid.raw, packet->mad.hdr.gid, 16); - ah_attr.grh.sgid_index = packet->mad.hdr.gid_index; - ah_attr.grh.flow_label = be32_to_cpu(packet->mad.hdr.flow_label); - ah_attr.grh.hop_limit = packet->mad.hdr.hop_limit; - ah_attr.grh.traffic_class = packet->mad.hdr.traffic_class; + rdma_ah_set_grh(&ah_attr, NULL, + be32_to_cpu(packet->mad.hdr.flow_label), + packet->mad.hdr.gid_index, + packet->mad.hdr.hop_limit, + packet->mad.hdr.traffic_class); + rdma_ah_set_dgid_raw(&ah_attr, packet->mad.hdr.gid); } - ah = ib_create_ah(agent->qp->pd, &ah_attr); + ah = rdma_create_ah(agent->qp->pd, &ah_attr); if (IS_ERR(ah)) { ret = PTR_ERR(ah); goto err_up; @@ -596,7 +600,7 @@ err_send: err_msg: ib_free_send_mad(packet->msg); err_ah: - ib_destroy_ah(ah); + rdma_destroy_ah(ah); err_up: mutex_unlock(&file->mutex); err: @@ -1183,12 +1187,12 @@ static int ib_umad_init_port(struct ib_device *device, int port_num, cdev_init(&port->cdev, &umad_fops); port->cdev.owner = THIS_MODULE; - port->cdev.kobj.parent = &umad_dev->kobj; + cdev_set_parent(&port->cdev, &umad_dev->kobj); kobject_set_name(&port->cdev.kobj, "umad%d", port->dev_num); if (cdev_add(&port->cdev, base, 1)) goto err_cdev; - port->dev = device_create(umad_class, device->dma_device, + port->dev = device_create(umad_class, device->dev.parent, port->cdev.dev, port, "umad%d", port->dev_num); if (IS_ERR(port->dev)) @@ -1202,12 +1206,12 @@ static int ib_umad_init_port(struct ib_device *device, int port_num, base += IB_UMAD_MAX_PORTS; cdev_init(&port->sm_cdev, &umad_sm_fops); port->sm_cdev.owner = THIS_MODULE; - port->sm_cdev.kobj.parent = &umad_dev->kobj; + cdev_set_parent(&port->sm_cdev, &umad_dev->kobj); kobject_set_name(&port->sm_cdev.kobj, "issm%d", port->dev_num); if (cdev_add(&port->sm_cdev, base, 1)) goto err_sm_cdev; - port->sm_dev = device_create(umad_class, device->dma_device, + port->sm_dev = device_create(umad_class, device->dev.parent, port->sm_cdev.dev, port, "issm%d", port->dev_num); if (IS_ERR(port->sm_dev)) diff --git a/drivers/infiniband/core/uverbs.h b/drivers/infiniband/core/uverbs.h index 455034ac994e..64d494a64daf 100644 --- a/drivers/infiniband/core/uverbs.h +++ b/drivers/infiniband/core/uverbs.h @@ -76,12 +76,13 @@ * an asynchronous event queue file is created and released when the * event file is closed. * - * struct ib_uverbs_event_file: One reference is held by the VFS and - * released when the file is closed. For asynchronous event files, - * another reference is held by the corresponding main context file - * and released when that file is closed. For completion event files, - * a reference is taken when a CQ is created that uses the file, and - * released when the CQ is destroyed. + * struct ib_uverbs_event_queue: Base structure for + * struct ib_uverbs_async_event_file and struct ib_uverbs_completion_event_file. + * One reference is held by the VFS and released when the file is closed. + * For asynchronous event files, another reference is held by the corresponding + * main context file and released when that file is closed. For completion + * event files, a reference is taken when a CQ is created that uses the file, + * and released when the CQ is destroyed. */ struct ib_uverbs_device { @@ -101,18 +102,26 @@ struct ib_uverbs_device { struct list_head uverbs_events_file_list; }; -struct ib_uverbs_event_file { - struct kref ref; - int is_async; - struct ib_uverbs_file *uverbs_file; +struct ib_uverbs_event_queue { spinlock_t lock; int is_closed; wait_queue_head_t poll_wait; struct fasync_struct *async_queue; struct list_head event_list; +}; + +struct ib_uverbs_async_event_file { + struct ib_uverbs_event_queue ev_queue; + struct ib_uverbs_file *uverbs_file; + struct kref ref; struct list_head list; }; +struct ib_uverbs_completion_event_file { + struct ib_uobject_file uobj_file; + struct ib_uverbs_event_queue ev_queue; +}; + struct ib_uverbs_file { struct kref ref; struct mutex mutex; @@ -120,9 +129,13 @@ struct ib_uverbs_file { struct ib_uverbs_device *device; struct ib_ucontext *ucontext; struct ib_event_handler event_handler; - struct ib_uverbs_event_file *async_file; + struct ib_uverbs_async_event_file *async_file; struct list_head list; int is_closed; + + struct idr idr; + /* spinlock protects write access to idr */ + spinlock_t idr_lock; }; struct ib_uverbs_event { @@ -159,6 +172,8 @@ struct ib_usrq_object { struct ib_uqp_object { struct ib_uevent_object uevent; + /* lock for mcast list */ + struct mutex mcast_lock; struct list_head mcast_list; struct ib_uxrcd_object *uxrcd; }; @@ -176,32 +191,18 @@ struct ib_ucq_object { u32 async_events_reported; }; -extern spinlock_t ib_uverbs_idr_lock; -extern struct idr ib_uverbs_pd_idr; -extern struct idr ib_uverbs_mr_idr; -extern struct idr ib_uverbs_mw_idr; -extern struct idr ib_uverbs_ah_idr; -extern struct idr ib_uverbs_cq_idr; -extern struct idr ib_uverbs_qp_idr; -extern struct idr ib_uverbs_srq_idr; -extern struct idr ib_uverbs_xrcd_idr; -extern struct idr ib_uverbs_rule_idr; -extern struct idr ib_uverbs_wq_idr; -extern struct idr ib_uverbs_rwq_ind_tbl_idr; - -void idr_remove_uobj(struct idr *idp, struct ib_uobject *uobj); - -struct file *ib_uverbs_alloc_event_file(struct ib_uverbs_file *uverbs_file, - struct ib_device *ib_dev, - int is_async); +extern const struct file_operations uverbs_event_fops; +void ib_uverbs_init_event_queue(struct ib_uverbs_event_queue *ev_queue); +struct file *ib_uverbs_alloc_async_event_file(struct ib_uverbs_file *uverbs_file, + struct ib_device *ib_dev); void ib_uverbs_free_async_event_file(struct ib_uverbs_file *uverbs_file); -struct ib_uverbs_event_file *ib_uverbs_lookup_comp_file(int fd); void ib_uverbs_release_ucq(struct ib_uverbs_file *file, - struct ib_uverbs_event_file *ev_file, + struct ib_uverbs_completion_event_file *ev_file, struct ib_ucq_object *uobj); void ib_uverbs_release_uevent(struct ib_uverbs_file *file, struct ib_uevent_object *uobj); +void ib_uverbs_release_file(struct kref *ref); void ib_uverbs_comp_handler(struct ib_cq *cq, void *cq_context); void ib_uverbs_cq_event_handler(struct ib_event *event, void *context_ptr); @@ -210,9 +211,12 @@ void ib_uverbs_wq_event_handler(struct ib_event *event, void *context_ptr); void ib_uverbs_srq_event_handler(struct ib_event *event, void *context_ptr); void ib_uverbs_event_handler(struct ib_event_handler *handler, struct ib_event *event); -void ib_uverbs_dealloc_xrcd(struct ib_uverbs_device *dev, struct ib_xrcd *xrcd); +int ib_uverbs_dealloc_xrcd(struct ib_uverbs_device *dev, struct ib_xrcd *xrcd, + enum rdma_remove_reason why); int uverbs_dealloc_mw(struct ib_mw *mw); +void ib_uverbs_detach_umcast(struct ib_qp *qp, + struct ib_uqp_object *uobj); struct ib_uverbs_flow_spec { union { @@ -228,6 +232,8 @@ struct ib_uverbs_flow_spec { struct ib_uverbs_flow_spec_ipv4 ipv4; struct ib_uverbs_flow_spec_tcp_udp tcp_udp; struct ib_uverbs_flow_spec_ipv6 ipv6; + struct ib_uverbs_flow_spec_action_tag flow_tag; + struct ib_uverbs_flow_spec_action_drop drop; }; }; diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c index 700782203483..2c98533a0203 100644 --- a/drivers/infiniband/core/uverbs_cmd.c +++ b/drivers/infiniband/core/uverbs_cmd.c @@ -40,270 +40,29 @@ #include <linux/uaccess.h> +#include <rdma/uverbs_types.h> +#include <rdma/uverbs_std_types.h> +#include "rdma_core.h" + #include "uverbs.h" #include "core_priv.h" -struct uverbs_lock_class { - struct lock_class_key key; - char name[16]; -}; - -static struct uverbs_lock_class pd_lock_class = { .name = "PD-uobj" }; -static struct uverbs_lock_class mr_lock_class = { .name = "MR-uobj" }; -static struct uverbs_lock_class mw_lock_class = { .name = "MW-uobj" }; -static struct uverbs_lock_class cq_lock_class = { .name = "CQ-uobj" }; -static struct uverbs_lock_class qp_lock_class = { .name = "QP-uobj" }; -static struct uverbs_lock_class ah_lock_class = { .name = "AH-uobj" }; -static struct uverbs_lock_class srq_lock_class = { .name = "SRQ-uobj" }; -static struct uverbs_lock_class xrcd_lock_class = { .name = "XRCD-uobj" }; -static struct uverbs_lock_class rule_lock_class = { .name = "RULE-uobj" }; -static struct uverbs_lock_class wq_lock_class = { .name = "WQ-uobj" }; -static struct uverbs_lock_class rwq_ind_table_lock_class = { .name = "IND_TBL-uobj" }; - -/* - * The ib_uobject locking scheme is as follows: - * - * - ib_uverbs_idr_lock protects the uverbs idrs themselves, so it - * needs to be held during all idr write operations. When an object is - * looked up, a reference must be taken on the object's kref before - * dropping this lock. For read operations, the rcu_read_lock() - * and rcu_write_lock() but similarly the kref reference is grabbed - * before the rcu_read_unlock(). - * - * - Each object also has an rwsem. This rwsem must be held for - * reading while an operation that uses the object is performed. - * For example, while registering an MR, the associated PD's - * uobject.mutex must be held for reading. The rwsem must be held - * for writing while initializing or destroying an object. - * - * - In addition, each object has a "live" flag. If this flag is not - * set, then lookups of the object will fail even if it is found in - * the idr. This handles a reader that blocks and does not acquire - * the rwsem until after the object is destroyed. The destroy - * operation will set the live flag to 0 and then drop the rwsem; - * this will allow the reader to acquire the rwsem, see that the - * live flag is 0, and then drop the rwsem and its reference to - * object. The underlying storage will not be freed until the last - * reference to the object is dropped. - */ - -static void init_uobj(struct ib_uobject *uobj, u64 user_handle, - struct ib_ucontext *context, struct uverbs_lock_class *c) -{ - uobj->user_handle = user_handle; - uobj->context = context; - kref_init(&uobj->ref); - init_rwsem(&uobj->mutex); - lockdep_set_class_and_name(&uobj->mutex, &c->key, c->name); - uobj->live = 0; -} - -static void release_uobj(struct kref *kref) -{ - kfree_rcu(container_of(kref, struct ib_uobject, ref), rcu); -} - -static void put_uobj(struct ib_uobject *uobj) -{ - kref_put(&uobj->ref, release_uobj); -} - -static void put_uobj_read(struct ib_uobject *uobj) -{ - up_read(&uobj->mutex); - put_uobj(uobj); -} - -static void put_uobj_write(struct ib_uobject *uobj) -{ - up_write(&uobj->mutex); - put_uobj(uobj); -} - -static int idr_add_uobj(struct idr *idr, struct ib_uobject *uobj) -{ - int ret; - - idr_preload(GFP_KERNEL); - spin_lock(&ib_uverbs_idr_lock); - - ret = idr_alloc(idr, uobj, 0, 0, GFP_NOWAIT); - if (ret >= 0) - uobj->id = ret; - - spin_unlock(&ib_uverbs_idr_lock); - idr_preload_end(); - - return ret < 0 ? ret : 0; -} - -void idr_remove_uobj(struct idr *idr, struct ib_uobject *uobj) +static struct ib_uverbs_completion_event_file * +ib_uverbs_lookup_comp_file(int fd, struct ib_ucontext *context) { - spin_lock(&ib_uverbs_idr_lock); - idr_remove(idr, uobj->id); - spin_unlock(&ib_uverbs_idr_lock); -} + struct ib_uobject *uobj = uobj_get_read(uobj_get_type(comp_channel), + fd, context); + struct ib_uobject_file *uobj_file; -static struct ib_uobject *__idr_get_uobj(struct idr *idr, int id, - struct ib_ucontext *context) -{ - struct ib_uobject *uobj; + if (IS_ERR(uobj)) + return (void *)uobj; - rcu_read_lock(); - uobj = idr_find(idr, id); - if (uobj) { - if (uobj->context == context) - kref_get(&uobj->ref); - else - uobj = NULL; - } - rcu_read_unlock(); + uverbs_uobject_get(uobj); + uobj_put_read(uobj); - return uobj; -} - -static struct ib_uobject *idr_read_uobj(struct idr *idr, int id, - struct ib_ucontext *context, int nested) -{ - struct ib_uobject *uobj; - - uobj = __idr_get_uobj(idr, id, context); - if (!uobj) - return NULL; - - if (nested) - down_read_nested(&uobj->mutex, SINGLE_DEPTH_NESTING); - else - down_read(&uobj->mutex); - if (!uobj->live) { - put_uobj_read(uobj); - return NULL; - } - - return uobj; -} - -static struct ib_uobject *idr_write_uobj(struct idr *idr, int id, - struct ib_ucontext *context) -{ - struct ib_uobject *uobj; - - uobj = __idr_get_uobj(idr, id, context); - if (!uobj) - return NULL; - - down_write(&uobj->mutex); - if (!uobj->live) { - put_uobj_write(uobj); - return NULL; - } - - return uobj; -} - -static void *idr_read_obj(struct idr *idr, int id, struct ib_ucontext *context, - int nested) -{ - struct ib_uobject *uobj; - - uobj = idr_read_uobj(idr, id, context, nested); - return uobj ? uobj->object : NULL; -} - -static struct ib_pd *idr_read_pd(int pd_handle, struct ib_ucontext *context) -{ - return idr_read_obj(&ib_uverbs_pd_idr, pd_handle, context, 0); -} - -static void put_pd_read(struct ib_pd *pd) -{ - put_uobj_read(pd->uobject); -} - -static struct ib_cq *idr_read_cq(int cq_handle, struct ib_ucontext *context, int nested) -{ - return idr_read_obj(&ib_uverbs_cq_idr, cq_handle, context, nested); -} - -static void put_cq_read(struct ib_cq *cq) -{ - put_uobj_read(cq->uobject); -} - -static struct ib_ah *idr_read_ah(int ah_handle, struct ib_ucontext *context) -{ - return idr_read_obj(&ib_uverbs_ah_idr, ah_handle, context, 0); -} - -static void put_ah_read(struct ib_ah *ah) -{ - put_uobj_read(ah->uobject); -} - -static struct ib_qp *idr_read_qp(int qp_handle, struct ib_ucontext *context) -{ - return idr_read_obj(&ib_uverbs_qp_idr, qp_handle, context, 0); -} - -static struct ib_wq *idr_read_wq(int wq_handle, struct ib_ucontext *context) -{ - return idr_read_obj(&ib_uverbs_wq_idr, wq_handle, context, 0); -} - -static void put_wq_read(struct ib_wq *wq) -{ - put_uobj_read(wq->uobject); -} - -static struct ib_rwq_ind_table *idr_read_rwq_indirection_table(int ind_table_handle, - struct ib_ucontext *context) -{ - return idr_read_obj(&ib_uverbs_rwq_ind_tbl_idr, ind_table_handle, context, 0); -} - -static void put_rwq_indirection_table_read(struct ib_rwq_ind_table *ind_table) -{ - put_uobj_read(ind_table->uobject); -} - -static struct ib_qp *idr_write_qp(int qp_handle, struct ib_ucontext *context) -{ - struct ib_uobject *uobj; - - uobj = idr_write_uobj(&ib_uverbs_qp_idr, qp_handle, context); - return uobj ? uobj->object : NULL; -} - -static void put_qp_read(struct ib_qp *qp) -{ - put_uobj_read(qp->uobject); -} - -static void put_qp_write(struct ib_qp *qp) -{ - put_uobj_write(qp->uobject); -} - -static struct ib_srq *idr_read_srq(int srq_handle, struct ib_ucontext *context) -{ - return idr_read_obj(&ib_uverbs_srq_idr, srq_handle, context, 0); -} - -static void put_srq_read(struct ib_srq *srq) -{ - put_uobj_read(srq->uobject); -} - -static struct ib_xrcd *idr_read_xrcd(int xrcd_handle, struct ib_ucontext *context, - struct ib_uobject **uobj) -{ - *uobj = idr_read_uobj(&ib_uverbs_xrcd_idr, xrcd_handle, context, 0); - return *uobj ? (*uobj)->object : NULL; -} - -static void put_xrcd_read(struct ib_uobject *uobj) -{ - put_uobj_read(uobj); + uobj_file = container_of(uobj, struct ib_uobject_file, uobj); + return container_of(uobj_file, struct ib_uverbs_completion_event_file, + uobj_file); } ssize_t ib_uverbs_get_context(struct ib_uverbs_file *file, @@ -316,6 +75,7 @@ ssize_t ib_uverbs_get_context(struct ib_uverbs_file *file, struct ib_udata udata; struct ib_ucontext *ucontext; struct file *filp; + struct ib_rdmacg_object cg_obj; int ret; if (out_len < sizeof resp) @@ -335,24 +95,22 @@ ssize_t ib_uverbs_get_context(struct ib_uverbs_file *file, (unsigned long) cmd.response + sizeof resp, in_len - sizeof cmd, out_len - sizeof resp); + ret = ib_rdmacg_try_charge(&cg_obj, ib_dev, RDMACG_RESOURCE_HCA_HANDLE); + if (ret) + goto err; + ucontext = ib_dev->alloc_ucontext(ib_dev, &udata); if (IS_ERR(ucontext)) { ret = PTR_ERR(ucontext); - goto err; + goto err_alloc; } ucontext->device = ib_dev; - INIT_LIST_HEAD(&ucontext->pd_list); - INIT_LIST_HEAD(&ucontext->mr_list); - INIT_LIST_HEAD(&ucontext->mw_list); - INIT_LIST_HEAD(&ucontext->cq_list); - INIT_LIST_HEAD(&ucontext->qp_list); - INIT_LIST_HEAD(&ucontext->srq_list); - INIT_LIST_HEAD(&ucontext->ah_list); - INIT_LIST_HEAD(&ucontext->wq_list); - INIT_LIST_HEAD(&ucontext->rwq_ind_tbl_list); - INIT_LIST_HEAD(&ucontext->xrcd_list); - INIT_LIST_HEAD(&ucontext->rule_list); + ucontext->cg_obj = cg_obj; + /* ufile is required when some objects are released */ + ucontext->ufile = file; + uverbs_initialize_ucontext(ucontext); + rcu_read_lock(); ucontext->tgid = get_task_pid(current->group_leader, PIDTYPE_PID); rcu_read_unlock(); @@ -376,7 +134,7 @@ ssize_t ib_uverbs_get_context(struct ib_uverbs_file *file, goto err_free; resp.async_fd = ret; - filp = ib_uverbs_alloc_event_file(file, ib_dev, 1); + filp = ib_uverbs_alloc_async_event_file(file, ib_dev); if (IS_ERR(filp)) { ret = PTR_ERR(filp); goto err_fd; @@ -407,6 +165,9 @@ err_free: put_pid(ucontext->tgid); ib_dev->dealloc_ucontext(ucontext); +err_alloc: + ib_rdmacg_uncharge(&cg_obj, ib_dev, RDMACG_RESOURCE_HCA_HANDLE); + err: mutex_unlock(&file->mutex); return ret; @@ -556,12 +317,9 @@ ssize_t ib_uverbs_alloc_pd(struct ib_uverbs_file *file, (unsigned long) cmd.response + sizeof resp, in_len - sizeof cmd, out_len - sizeof resp); - uobj = kmalloc(sizeof *uobj, GFP_KERNEL); - if (!uobj) - return -ENOMEM; - - init_uobj(uobj, 0, file->ucontext, &pd_lock_class); - down_write(&uobj->mutex); + uobj = uobj_alloc(uobj_get_type(pd), file->ucontext); + if (IS_ERR(uobj)) + return PTR_ERR(uobj); pd = ib_dev->alloc_pd(ib_dev, file->ucontext, &udata); if (IS_ERR(pd)) { @@ -575,10 +333,6 @@ ssize_t ib_uverbs_alloc_pd(struct ib_uverbs_file *file, atomic_set(&pd->usecnt, 0); uobj->object = pd; - ret = idr_add_uobj(&ib_uverbs_pd_idr, uobj); - if (ret) - goto err_idr; - memset(&resp, 0, sizeof resp); resp.pd_handle = uobj->id; @@ -588,24 +342,15 @@ ssize_t ib_uverbs_alloc_pd(struct ib_uverbs_file *file, goto err_copy; } - mutex_lock(&file->mutex); - list_add_tail(&uobj->list, &file->ucontext->pd_list); - mutex_unlock(&file->mutex); - - uobj->live = 1; - - up_write(&uobj->mutex); + uobj_alloc_commit(uobj); return in_len; err_copy: - idr_remove_uobj(&ib_uverbs_pd_idr, uobj); - -err_idr: ib_dealloc_pd(pd); err: - put_uobj_write(uobj); + uobj_alloc_abort(uobj); return ret; } @@ -616,43 +361,19 @@ ssize_t ib_uverbs_dealloc_pd(struct ib_uverbs_file *file, { struct ib_uverbs_dealloc_pd cmd; struct ib_uobject *uobj; - struct ib_pd *pd; int ret; if (copy_from_user(&cmd, buf, sizeof cmd)) return -EFAULT; - uobj = idr_write_uobj(&ib_uverbs_pd_idr, cmd.pd_handle, file->ucontext); - if (!uobj) - return -EINVAL; - pd = uobj->object; - - if (atomic_read(&pd->usecnt)) { - ret = -EBUSY; - goto err_put; - } - - ret = pd->device->dealloc_pd(uobj->object); - WARN_ONCE(ret, "Infiniband HW driver failed dealloc_pd"); - if (ret) - goto err_put; - - uobj->live = 0; - put_uobj_write(uobj); + uobj = uobj_get_write(uobj_get_type(pd), cmd.pd_handle, + file->ucontext); + if (IS_ERR(uobj)) + return PTR_ERR(uobj); - idr_remove_uobj(&ib_uverbs_pd_idr, uobj); + ret = uobj_remove_commit(uobj); - mutex_lock(&file->mutex); - list_del(&uobj->list); - mutex_unlock(&file->mutex); - - put_uobj(uobj); - - return in_len; - -err_put: - put_uobj_write(uobj); - return ret; + return ret ?: in_len; } struct xrcd_table_entry { @@ -789,16 +510,13 @@ ssize_t ib_uverbs_open_xrcd(struct ib_uverbs_file *file, } } - obj = kmalloc(sizeof *obj, GFP_KERNEL); - if (!obj) { - ret = -ENOMEM; + obj = (struct ib_uxrcd_object *)uobj_alloc(uobj_get_type(xrcd), + file->ucontext); + if (IS_ERR(obj)) { + ret = PTR_ERR(obj); goto err_tree_mutex_unlock; } - init_uobj(&obj->uobject, 0, file->ucontext, &xrcd_lock_class); - - down_write(&obj->uobject.mutex); - if (!xrcd) { xrcd = ib_dev->alloc_xrcd(ib_dev, file->ucontext, &udata); if (IS_ERR(xrcd)) { @@ -816,10 +534,6 @@ ssize_t ib_uverbs_open_xrcd(struct ib_uverbs_file *file, atomic_set(&obj->refcnt, 0); obj->uobject.object = xrcd; - ret = idr_add_uobj(&ib_uverbs_xrcd_idr, &obj->uobject); - if (ret) - goto err_idr; - memset(&resp, 0, sizeof resp); resp.xrcd_handle = obj->uobject.id; @@ -828,7 +542,7 @@ ssize_t ib_uverbs_open_xrcd(struct ib_uverbs_file *file, /* create new inode/xrcd table entry */ ret = xrcd_table_insert(file->device, inode, xrcd); if (ret) - goto err_insert_xrcd; + goto err_dealloc_xrcd; } atomic_inc(&xrcd->usecnt); } @@ -842,12 +556,7 @@ ssize_t ib_uverbs_open_xrcd(struct ib_uverbs_file *file, if (f.file) fdput(f); - mutex_lock(&file->mutex); - list_add_tail(&obj->uobject.list, &file->ucontext->xrcd_list); - mutex_unlock(&file->mutex); - - obj->uobject.live = 1; - up_write(&obj->uobject.mutex); + uobj_alloc_commit(&obj->uobject); mutex_unlock(&file->device->xrcd_tree_mutex); return in_len; @@ -859,14 +568,11 @@ err_copy: atomic_dec(&xrcd->usecnt); } -err_insert_xrcd: - idr_remove_uobj(&ib_uverbs_xrcd_idr, &obj->uobject); - -err_idr: +err_dealloc_xrcd: ib_dealloc_xrcd(xrcd); err: - put_uobj_write(&obj->uobject); + uobj_alloc_abort(&obj->uobject); err_tree_mutex_unlock: if (f.file) @@ -884,75 +590,41 @@ ssize_t ib_uverbs_close_xrcd(struct ib_uverbs_file *file, { struct ib_uverbs_close_xrcd cmd; struct ib_uobject *uobj; - struct ib_xrcd *xrcd = NULL; - struct inode *inode = NULL; - struct ib_uxrcd_object *obj; - int live; int ret = 0; if (copy_from_user(&cmd, buf, sizeof cmd)) return -EFAULT; - mutex_lock(&file->device->xrcd_tree_mutex); - uobj = idr_write_uobj(&ib_uverbs_xrcd_idr, cmd.xrcd_handle, file->ucontext); - if (!uobj) { - ret = -EINVAL; - goto out; - } - - xrcd = uobj->object; - inode = xrcd->inode; - obj = container_of(uobj, struct ib_uxrcd_object, uobject); - if (atomic_read(&obj->refcnt)) { - put_uobj_write(uobj); - ret = -EBUSY; - goto out; - } - - if (!inode || atomic_dec_and_test(&xrcd->usecnt)) { - ret = ib_dealloc_xrcd(uobj->object); - if (!ret) - uobj->live = 0; + uobj = uobj_get_write(uobj_get_type(xrcd), cmd.xrcd_handle, + file->ucontext); + if (IS_ERR(uobj)) { + mutex_unlock(&file->device->xrcd_tree_mutex); + return PTR_ERR(uobj); } - live = uobj->live; - if (inode && ret) - atomic_inc(&xrcd->usecnt); - - put_uobj_write(uobj); - - if (ret) - goto out; - - if (inode && !live) - xrcd_table_delete(file->device, inode); - - idr_remove_uobj(&ib_uverbs_xrcd_idr, uobj); - mutex_lock(&file->mutex); - list_del(&uobj->list); - mutex_unlock(&file->mutex); - - put_uobj(uobj); - ret = in_len; - -out: - mutex_unlock(&file->device->xrcd_tree_mutex); - return ret; + ret = uobj_remove_commit(uobj); + return ret ?: in_len; } -void ib_uverbs_dealloc_xrcd(struct ib_uverbs_device *dev, - struct ib_xrcd *xrcd) +int ib_uverbs_dealloc_xrcd(struct ib_uverbs_device *dev, + struct ib_xrcd *xrcd, + enum rdma_remove_reason why) { struct inode *inode; + int ret; inode = xrcd->inode; if (inode && !atomic_dec_and_test(&xrcd->usecnt)) - return; + return 0; - ib_dealloc_xrcd(xrcd); + ret = ib_dealloc_xrcd(xrcd); - if (inode) + if (why == RDMA_REMOVE_DESTROY && ret) + atomic_inc(&xrcd->usecnt); + else if (inode) xrcd_table_delete(dev, inode); + + return ret; } ssize_t ib_uverbs_reg_mr(struct ib_uverbs_file *file, @@ -985,14 +657,11 @@ ssize_t ib_uverbs_reg_mr(struct ib_uverbs_file *file, if (ret) return ret; - uobj = kmalloc(sizeof *uobj, GFP_KERNEL); - if (!uobj) - return -ENOMEM; - - init_uobj(uobj, 0, file->ucontext, &mr_lock_class); - down_write(&uobj->mutex); + uobj = uobj_alloc(uobj_get_type(mr), file->ucontext); + if (IS_ERR(uobj)) + return PTR_ERR(uobj); - pd = idr_read_pd(cmd.pd_handle, file->ucontext); + pd = uobj_get_obj_read(pd, cmd.pd_handle, file->ucontext); if (!pd) { ret = -EINVAL; goto err_free; @@ -1020,9 +689,6 @@ ssize_t ib_uverbs_reg_mr(struct ib_uverbs_file *file, atomic_inc(&pd->usecnt); uobj->object = mr; - ret = idr_add_uobj(&ib_uverbs_mr_idr, uobj); - if (ret) - goto err_unreg; memset(&resp, 0, sizeof resp); resp.lkey = mr->lkey; @@ -1035,29 +701,20 @@ ssize_t ib_uverbs_reg_mr(struct ib_uverbs_file *file, goto err_copy; } - put_pd_read(pd); - - mutex_lock(&file->mutex); - list_add_tail(&uobj->list, &file->ucontext->mr_list); - mutex_unlock(&file->mutex); - - uobj->live = 1; + uobj_put_obj_read(pd); - up_write(&uobj->mutex); + uobj_alloc_commit(uobj); return in_len; err_copy: - idr_remove_uobj(&ib_uverbs_mr_idr, uobj); - -err_unreg: ib_dereg_mr(mr); err_put: - put_pd_read(pd); + uobj_put_obj_read(pd); err_free: - put_uobj_write(uobj); + uobj_alloc_abort(uobj); return ret; } @@ -1093,11 +750,10 @@ ssize_t ib_uverbs_rereg_mr(struct ib_uverbs_file *file, (cmd.start & ~PAGE_MASK) != (cmd.hca_va & ~PAGE_MASK))) return -EINVAL; - uobj = idr_write_uobj(&ib_uverbs_mr_idr, cmd.mr_handle, - file->ucontext); - - if (!uobj) - return -EINVAL; + uobj = uobj_get_write(uobj_get_type(mr), cmd.mr_handle, + file->ucontext); + if (IS_ERR(uobj)) + return PTR_ERR(uobj); mr = uobj->object; @@ -1108,7 +764,7 @@ ssize_t ib_uverbs_rereg_mr(struct ib_uverbs_file *file, } if (cmd.flags & IB_MR_REREG_PD) { - pd = idr_read_pd(cmd.pd_handle, file->ucontext); + pd = uobj_get_obj_read(pd, cmd.pd_handle, file->ucontext); if (!pd) { ret = -EINVAL; goto put_uobjs; @@ -1141,11 +797,10 @@ ssize_t ib_uverbs_rereg_mr(struct ib_uverbs_file *file, put_uobj_pd: if (cmd.flags & IB_MR_REREG_PD) - put_pd_read(pd); + uobj_put_obj_read(pd); put_uobjs: - - put_uobj_write(mr->uobject); + uobj_put_write(uobj); return ret; } @@ -1156,37 +811,20 @@ ssize_t ib_uverbs_dereg_mr(struct ib_uverbs_file *file, int out_len) { struct ib_uverbs_dereg_mr cmd; - struct ib_mr *mr; struct ib_uobject *uobj; int ret = -EINVAL; if (copy_from_user(&cmd, buf, sizeof cmd)) return -EFAULT; - uobj = idr_write_uobj(&ib_uverbs_mr_idr, cmd.mr_handle, file->ucontext); - if (!uobj) - return -EINVAL; + uobj = uobj_get_write(uobj_get_type(mr), cmd.mr_handle, + file->ucontext); + if (IS_ERR(uobj)) + return PTR_ERR(uobj); - mr = uobj->object; + ret = uobj_remove_commit(uobj); - ret = ib_dereg_mr(mr); - if (!ret) - uobj->live = 0; - - put_uobj_write(uobj); - - if (ret) - return ret; - - idr_remove_uobj(&ib_uverbs_mr_idr, uobj); - - mutex_lock(&file->mutex); - list_del(&uobj->list); - mutex_unlock(&file->mutex); - - put_uobj(uobj); - - return in_len; + return ret ?: in_len; } ssize_t ib_uverbs_alloc_mw(struct ib_uverbs_file *file, @@ -1208,14 +846,11 @@ ssize_t ib_uverbs_alloc_mw(struct ib_uverbs_file *file, if (copy_from_user(&cmd, buf, sizeof(cmd))) return -EFAULT; - uobj = kmalloc(sizeof(*uobj), GFP_KERNEL); - if (!uobj) - return -ENOMEM; + uobj = uobj_alloc(uobj_get_type(mw), file->ucontext); + if (IS_ERR(uobj)) + return PTR_ERR(uobj); - init_uobj(uobj, 0, file->ucontext, &mw_lock_class); - down_write(&uobj->mutex); - - pd = idr_read_pd(cmd.pd_handle, file->ucontext); + pd = uobj_get_obj_read(pd, cmd.pd_handle, file->ucontext); if (!pd) { ret = -EINVAL; goto err_free; @@ -1238,9 +873,6 @@ ssize_t ib_uverbs_alloc_mw(struct ib_uverbs_file *file, atomic_inc(&pd->usecnt); uobj->object = mw; - ret = idr_add_uobj(&ib_uverbs_mw_idr, uobj); - if (ret) - goto err_unalloc; memset(&resp, 0, sizeof(resp)); resp.rkey = mw->rkey; @@ -1252,29 +884,17 @@ ssize_t ib_uverbs_alloc_mw(struct ib_uverbs_file *file, goto err_copy; } - put_pd_read(pd); - - mutex_lock(&file->mutex); - list_add_tail(&uobj->list, &file->ucontext->mw_list); - mutex_unlock(&file->mutex); - - uobj->live = 1; - - up_write(&uobj->mutex); + uobj_put_obj_read(pd); + uobj_alloc_commit(uobj); return in_len; err_copy: - idr_remove_uobj(&ib_uverbs_mw_idr, uobj); - -err_unalloc: uverbs_dealloc_mw(mw); - err_put: - put_pd_read(pd); - + uobj_put_obj_read(pd); err_free: - put_uobj_write(uobj); + uobj_alloc_abort(uobj); return ret; } @@ -1284,37 +904,19 @@ ssize_t ib_uverbs_dealloc_mw(struct ib_uverbs_file *file, int out_len) { struct ib_uverbs_dealloc_mw cmd; - struct ib_mw *mw; struct ib_uobject *uobj; int ret = -EINVAL; if (copy_from_user(&cmd, buf, sizeof(cmd))) return -EFAULT; - uobj = idr_write_uobj(&ib_uverbs_mw_idr, cmd.mw_handle, file->ucontext); - if (!uobj) - return -EINVAL; - - mw = uobj->object; + uobj = uobj_get_write(uobj_get_type(mw), cmd.mw_handle, + file->ucontext); + if (IS_ERR(uobj)) + return PTR_ERR(uobj); - ret = uverbs_dealloc_mw(mw); - if (!ret) - uobj->live = 0; - - put_uobj_write(uobj); - - if (ret) - return ret; - - idr_remove_uobj(&ib_uverbs_mw_idr, uobj); - - mutex_lock(&file->mutex); - list_del(&uobj->list); - mutex_unlock(&file->mutex); - - put_uobj(uobj); - - return in_len; + ret = uobj_remove_commit(uobj); + return ret ?: in_len; } ssize_t ib_uverbs_create_comp_channel(struct ib_uverbs_file *file, @@ -1324,8 +926,8 @@ ssize_t ib_uverbs_create_comp_channel(struct ib_uverbs_file *file, { struct ib_uverbs_create_comp_channel cmd; struct ib_uverbs_create_comp_channel_resp resp; - struct file *filp; - int ret; + struct ib_uobject *uobj; + struct ib_uverbs_completion_event_file *ev_file; if (out_len < sizeof resp) return -ENOSPC; @@ -1333,25 +935,23 @@ ssize_t ib_uverbs_create_comp_channel(struct ib_uverbs_file *file, if (copy_from_user(&cmd, buf, sizeof cmd)) return -EFAULT; - ret = get_unused_fd_flags(O_CLOEXEC); - if (ret < 0) - return ret; - resp.fd = ret; + uobj = uobj_alloc(uobj_get_type(comp_channel), file->ucontext); + if (IS_ERR(uobj)) + return PTR_ERR(uobj); - filp = ib_uverbs_alloc_event_file(file, ib_dev, 0); - if (IS_ERR(filp)) { - put_unused_fd(resp.fd); - return PTR_ERR(filp); - } + resp.fd = uobj->id; + + ev_file = container_of(uobj, struct ib_uverbs_completion_event_file, + uobj_file.uobj); + ib_uverbs_init_event_queue(&ev_file->ev_queue); if (copy_to_user((void __user *) (unsigned long) cmd.response, &resp, sizeof resp)) { - put_unused_fd(resp.fd); - fput(filp); + uobj_alloc_abort(uobj); return -EFAULT; } - fd_install(resp.fd, filp); + uobj_alloc_commit(uobj); return in_len; } @@ -1369,7 +969,7 @@ static struct ib_ucq_object *create_cq(struct ib_uverbs_file *file, void *context) { struct ib_ucq_object *obj; - struct ib_uverbs_event_file *ev_file = NULL; + struct ib_uverbs_completion_event_file *ev_file = NULL; struct ib_cq *cq; int ret; struct ib_uverbs_ex_create_cq_resp resp; @@ -1378,21 +978,21 @@ static struct ib_ucq_object *create_cq(struct ib_uverbs_file *file, if (cmd->comp_vector >= file->device->num_comp_vectors) return ERR_PTR(-EINVAL); - obj = kmalloc(sizeof *obj, GFP_KERNEL); - if (!obj) - return ERR_PTR(-ENOMEM); - - init_uobj(&obj->uobject, cmd->user_handle, file->ucontext, &cq_lock_class); - down_write(&obj->uobject.mutex); + obj = (struct ib_ucq_object *)uobj_alloc(uobj_get_type(cq), + file->ucontext); + if (IS_ERR(obj)) + return obj; if (cmd->comp_channel >= 0) { - ev_file = ib_uverbs_lookup_comp_file(cmd->comp_channel); - if (!ev_file) { - ret = -EINVAL; + ev_file = ib_uverbs_lookup_comp_file(cmd->comp_channel, + file->ucontext); + if (IS_ERR(ev_file)) { + ret = PTR_ERR(ev_file); goto err; } } + obj->uobject.user_handle = cmd->user_handle; obj->uverbs_file = file; obj->comp_events_reported = 0; obj->async_events_reported = 0; @@ -1405,8 +1005,7 @@ static struct ib_ucq_object *create_cq(struct ib_uverbs_file *file, if (cmd_sz > offsetof(typeof(*cmd), flags) + sizeof(cmd->flags)) attr.flags = cmd->flags; - cq = ib_dev->create_cq(ib_dev, &attr, - file->ucontext, uhw); + cq = ib_dev->create_cq(ib_dev, &attr, file->ucontext, uhw); if (IS_ERR(cq)) { ret = PTR_ERR(cq); goto err_file; @@ -1416,14 +1015,10 @@ static struct ib_ucq_object *create_cq(struct ib_uverbs_file *file, cq->uobject = &obj->uobject; cq->comp_handler = ib_uverbs_comp_handler; cq->event_handler = ib_uverbs_cq_event_handler; - cq->cq_context = ev_file; + cq->cq_context = &ev_file->ev_queue; atomic_set(&cq->usecnt, 0); obj->uobject.object = cq; - ret = idr_add_uobj(&ib_uverbs_cq_idr, &obj->uobject); - if (ret) - goto err_free; - memset(&resp, 0, sizeof resp); resp.base.cq_handle = obj->uobject.id; resp.base.cqe = cq->cqe; @@ -1435,20 +1030,11 @@ static struct ib_ucq_object *create_cq(struct ib_uverbs_file *file, if (ret) goto err_cb; - mutex_lock(&file->mutex); - list_add_tail(&obj->uobject.list, &file->ucontext->cq_list); - mutex_unlock(&file->mutex); - - obj->uobject.live = 1; - - up_write(&obj->uobject.mutex); + uobj_alloc_commit(&obj->uobject); return obj; err_cb: - idr_remove_uobj(&ib_uverbs_cq_idr, &obj->uobject); - -err_free: ib_destroy_cq(cq); err_file: @@ -1456,7 +1042,7 @@ err_file: ib_uverbs_release_ucq(file, ev_file, obj); err: - put_uobj_write(&obj->uobject); + uobj_alloc_abort(&obj->uobject); return ERR_PTR(ret); } @@ -1579,7 +1165,7 @@ ssize_t ib_uverbs_resize_cq(struct ib_uverbs_file *file, (unsigned long) cmd.response + sizeof resp, in_len - sizeof cmd, out_len - sizeof resp); - cq = idr_read_cq(cmd.cq_handle, file->ucontext, 0); + cq = uobj_get_obj_read(cq, cmd.cq_handle, file->ucontext); if (!cq) return -EINVAL; @@ -1594,7 +1180,7 @@ ssize_t ib_uverbs_resize_cq(struct ib_uverbs_file *file, ret = -EFAULT; out: - put_cq_read(cq); + uobj_put_obj_read(cq); return ret ? ret : in_len; } @@ -1641,7 +1227,7 @@ ssize_t ib_uverbs_poll_cq(struct ib_uverbs_file *file, if (copy_from_user(&cmd, buf, sizeof cmd)) return -EFAULT; - cq = idr_read_cq(cmd.cq_handle, file->ucontext, 0); + cq = uobj_get_obj_read(cq, cmd.cq_handle, file->ucontext); if (!cq) return -EINVAL; @@ -1673,7 +1259,7 @@ ssize_t ib_uverbs_poll_cq(struct ib_uverbs_file *file, ret = in_len; out_put: - put_cq_read(cq); + uobj_put_obj_read(cq); return ret; } @@ -1688,14 +1274,14 @@ ssize_t ib_uverbs_req_notify_cq(struct ib_uverbs_file *file, if (copy_from_user(&cmd, buf, sizeof cmd)) return -EFAULT; - cq = idr_read_cq(cmd.cq_handle, file->ucontext, 0); + cq = uobj_get_obj_read(cq, cmd.cq_handle, file->ucontext); if (!cq) return -EINVAL; ib_req_notify_cq(cq, cmd.solicited_only ? IB_CQ_SOLICITED : IB_CQ_NEXT_COMP); - put_cq_read(cq); + uobj_put_obj_read(cq); return in_len; } @@ -1710,42 +1296,36 @@ ssize_t ib_uverbs_destroy_cq(struct ib_uverbs_file *file, struct ib_uobject *uobj; struct ib_cq *cq; struct ib_ucq_object *obj; - struct ib_uverbs_event_file *ev_file; int ret = -EINVAL; if (copy_from_user(&cmd, buf, sizeof cmd)) return -EFAULT; - uobj = idr_write_uobj(&ib_uverbs_cq_idr, cmd.cq_handle, file->ucontext); - if (!uobj) - return -EINVAL; + uobj = uobj_get_write(uobj_get_type(cq), cmd.cq_handle, + file->ucontext); + if (IS_ERR(uobj)) + return PTR_ERR(uobj); + + /* + * Make sure we don't free the memory in remove_commit as we still + * needs the uobject memory to create the response. + */ + uverbs_uobject_get(uobj); cq = uobj->object; - ev_file = cq->cq_context; obj = container_of(cq->uobject, struct ib_ucq_object, uobject); - ret = ib_destroy_cq(cq); - if (!ret) - uobj->live = 0; - - put_uobj_write(uobj); + memset(&resp, 0, sizeof(resp)); - if (ret) + ret = uobj_remove_commit(uobj); + if (ret) { + uverbs_uobject_put(uobj); return ret; + } - idr_remove_uobj(&ib_uverbs_cq_idr, uobj); - - mutex_lock(&file->mutex); - list_del(&uobj->list); - mutex_unlock(&file->mutex); - - ib_uverbs_release_ucq(file, ev_file, obj); - - memset(&resp, 0, sizeof resp); resp.comp_events_reported = obj->comp_events_reported; resp.async_events_reported = obj->async_events_reported; - put_uobj(uobj); - + uverbs_uobject_put(uobj); if (copy_to_user((void __user *) (unsigned long) cmd.response, &resp, sizeof resp)) return -EFAULT; @@ -1767,7 +1347,7 @@ static int create_qp(struct ib_uverbs_file *file, struct ib_device *device; struct ib_pd *pd = NULL; struct ib_xrcd *xrcd = NULL; - struct ib_uobject *uninitialized_var(xrcd_uobj); + struct ib_uobject *xrcd_uobj = ERR_PTR(-ENOENT); struct ib_cq *scq = NULL, *rcq = NULL; struct ib_srq *srq = NULL; struct ib_qp *qp; @@ -1781,18 +1361,20 @@ static int create_qp(struct ib_uverbs_file *file, if (cmd->qp_type == IB_QPT_RAW_PACKET && !capable(CAP_NET_RAW)) return -EPERM; - obj = kzalloc(sizeof *obj, GFP_KERNEL); - if (!obj) - return -ENOMEM; + obj = (struct ib_uqp_object *)uobj_alloc(uobj_get_type(qp), + file->ucontext); + if (IS_ERR(obj)) + return PTR_ERR(obj); + obj->uxrcd = NULL; + obj->uevent.uobject.user_handle = cmd->user_handle; + mutex_init(&obj->mcast_lock); - init_uobj(&obj->uevent.uobject, cmd->user_handle, file->ucontext, - &qp_lock_class); - down_write(&obj->uevent.uobject.mutex); if (cmd_sz >= offsetof(typeof(*cmd), rwq_ind_tbl_handle) + sizeof(cmd->rwq_ind_tbl_handle) && (cmd->comp_mask & IB_UVERBS_CREATE_QP_MASK_IND_TABLE)) { - ind_tbl = idr_read_rwq_indirection_table(cmd->rwq_ind_tbl_handle, - file->ucontext); + ind_tbl = uobj_get_obj_read(rwq_ind_table, + cmd->rwq_ind_tbl_handle, + file->ucontext); if (!ind_tbl) { ret = -EINVAL; goto err_put; @@ -1816,8 +1398,15 @@ static int create_qp(struct ib_uverbs_file *file, has_sq = false; if (cmd->qp_type == IB_QPT_XRC_TGT) { - xrcd = idr_read_xrcd(cmd->pd_handle, file->ucontext, - &xrcd_uobj); + xrcd_uobj = uobj_get_read(uobj_get_type(xrcd), cmd->pd_handle, + file->ucontext); + + if (IS_ERR(xrcd_uobj)) { + ret = -EINVAL; + goto err_put; + } + + xrcd = (struct ib_xrcd *)xrcd_uobj->object; if (!xrcd) { ret = -EINVAL; goto err_put; @@ -1829,8 +1418,8 @@ static int create_qp(struct ib_uverbs_file *file, cmd->max_recv_sge = 0; } else { if (cmd->is_srq) { - srq = idr_read_srq(cmd->srq_handle, - file->ucontext); + srq = uobj_get_obj_read(srq, cmd->srq_handle, + file->ucontext); if (!srq || srq->srq_type != IB_SRQT_BASIC) { ret = -EINVAL; goto err_put; @@ -1839,8 +1428,8 @@ static int create_qp(struct ib_uverbs_file *file, if (!ind_tbl) { if (cmd->recv_cq_handle != cmd->send_cq_handle) { - rcq = idr_read_cq(cmd->recv_cq_handle, - file->ucontext, 0); + rcq = uobj_get_obj_read(cq, cmd->recv_cq_handle, + file->ucontext); if (!rcq) { ret = -EINVAL; goto err_put; @@ -1850,10 +1439,11 @@ static int create_qp(struct ib_uverbs_file *file, } if (has_sq) - scq = idr_read_cq(cmd->send_cq_handle, file->ucontext, !!rcq); + scq = uobj_get_obj_read(cq, cmd->send_cq_handle, + file->ucontext); if (!ind_tbl) rcq = rcq ?: scq; - pd = idr_read_pd(cmd->pd_handle, file->ucontext); + pd = uobj_get_obj_read(pd, cmd->pd_handle, file->ucontext); if (!pd || (!scq && has_sq)) { ret = -EINVAL; goto err_put; @@ -1891,7 +1481,8 @@ static int create_qp(struct ib_uverbs_file *file, IB_QP_CREATE_CROSS_CHANNEL | IB_QP_CREATE_MANAGED_SEND | IB_QP_CREATE_MANAGED_RECV | - IB_QP_CREATE_SCATTER_FCS)) { + IB_QP_CREATE_SCATTER_FCS | + IB_QP_CREATE_CVLAN_STRIPPING)) { ret = -EINVAL; goto err_put; } @@ -1915,6 +1506,10 @@ static int create_qp(struct ib_uverbs_file *file, } if (cmd->qp_type != IB_QPT_XRC_TGT) { + ret = ib_create_qp_security(qp, device); + if (ret) + goto err_cb; + qp->real_qp = qp; qp->device = device; qp->pd = pd; @@ -1939,9 +1534,6 @@ static int create_qp(struct ib_uverbs_file *file, qp->uobject = &obj->uevent.uobject; obj->uevent.uobject.object = qp; - ret = idr_add_uobj(&ib_uverbs_qp_idr, &obj->uevent.uobject); - if (ret) - goto err_destroy; memset(&resp, 0, sizeof resp); resp.base.qpn = qp->qp_num; @@ -1963,50 +1555,41 @@ static int create_qp(struct ib_uverbs_file *file, obj->uxrcd = container_of(xrcd_uobj, struct ib_uxrcd_object, uobject); atomic_inc(&obj->uxrcd->refcnt); - put_xrcd_read(xrcd_uobj); + uobj_put_read(xrcd_uobj); } if (pd) - put_pd_read(pd); + uobj_put_obj_read(pd); if (scq) - put_cq_read(scq); + uobj_put_obj_read(scq); if (rcq && rcq != scq) - put_cq_read(rcq); + uobj_put_obj_read(rcq); if (srq) - put_srq_read(srq); + uobj_put_obj_read(srq); if (ind_tbl) - put_rwq_indirection_table_read(ind_tbl); + uobj_put_obj_read(ind_tbl); - mutex_lock(&file->mutex); - list_add_tail(&obj->uevent.uobject.list, &file->ucontext->qp_list); - mutex_unlock(&file->mutex); - - obj->uevent.uobject.live = 1; - - up_write(&obj->uevent.uobject.mutex); + uobj_alloc_commit(&obj->uevent.uobject); return 0; err_cb: - idr_remove_uobj(&ib_uverbs_qp_idr, &obj->uevent.uobject); - -err_destroy: ib_destroy_qp(qp); err_put: - if (xrcd) - put_xrcd_read(xrcd_uobj); + if (!IS_ERR(xrcd_uobj)) + uobj_put_read(xrcd_uobj); if (pd) - put_pd_read(pd); + uobj_put_obj_read(pd); if (scq) - put_cq_read(scq); + uobj_put_obj_read(scq); if (rcq && rcq != scq) - put_cq_read(rcq); + uobj_put_obj_read(rcq); if (srq) - put_srq_read(srq); + uobj_put_obj_read(srq); if (ind_tbl) - put_rwq_indirection_table_read(ind_tbl); + uobj_put_obj_read(ind_tbl); - put_uobj_write(&obj->uevent.uobject); + uobj_alloc_abort(&obj->uevent.uobject); return ret; } @@ -2142,17 +1725,22 @@ ssize_t ib_uverbs_open_qp(struct ib_uverbs_file *file, (unsigned long) cmd.response + sizeof resp, in_len - sizeof cmd, out_len - sizeof resp); - obj = kmalloc(sizeof *obj, GFP_KERNEL); - if (!obj) - return -ENOMEM; + obj = (struct ib_uqp_object *)uobj_alloc(uobj_get_type(qp), + file->ucontext); + if (IS_ERR(obj)) + return PTR_ERR(obj); - init_uobj(&obj->uevent.uobject, cmd.user_handle, file->ucontext, &qp_lock_class); - down_write(&obj->uevent.uobject.mutex); + xrcd_uobj = uobj_get_read(uobj_get_type(xrcd), cmd.pd_handle, + file->ucontext); + if (IS_ERR(xrcd_uobj)) { + ret = -EINVAL; + goto err_put; + } - xrcd = idr_read_xrcd(cmd.pd_handle, file->ucontext, &xrcd_uobj); + xrcd = (struct ib_xrcd *)xrcd_uobj->object; if (!xrcd) { ret = -EINVAL; - goto err_put; + goto err_xrcd; } attr.event_handler = ib_uverbs_qp_event_handler; @@ -2167,15 +1755,11 @@ ssize_t ib_uverbs_open_qp(struct ib_uverbs_file *file, qp = ib_open_qp(xrcd, &attr); if (IS_ERR(qp)) { ret = PTR_ERR(qp); - goto err_put; + goto err_xrcd; } - qp->uobject = &obj->uevent.uobject; - obj->uevent.uobject.object = qp; - ret = idr_add_uobj(&ib_uverbs_qp_idr, &obj->uevent.uobject); - if (ret) - goto err_destroy; + obj->uevent.uobject.user_handle = cmd.user_handle; memset(&resp, 0, sizeof resp); resp.qpn = qp->qp_num; @@ -2184,32 +1768,25 @@ ssize_t ib_uverbs_open_qp(struct ib_uverbs_file *file, if (copy_to_user((void __user *) (unsigned long) cmd.response, &resp, sizeof resp)) { ret = -EFAULT; - goto err_remove; + goto err_destroy; } obj->uxrcd = container_of(xrcd_uobj, struct ib_uxrcd_object, uobject); atomic_inc(&obj->uxrcd->refcnt); - put_xrcd_read(xrcd_uobj); - - mutex_lock(&file->mutex); - list_add_tail(&obj->uevent.uobject.list, &file->ucontext->qp_list); - mutex_unlock(&file->mutex); + qp->uobject = &obj->uevent.uobject; + uobj_put_read(xrcd_uobj); - obj->uevent.uobject.live = 1; - up_write(&obj->uevent.uobject.mutex); + uobj_alloc_commit(&obj->uevent.uobject); return in_len; -err_remove: - idr_remove_uobj(&ib_uverbs_qp_idr, &obj->uevent.uobject); - err_destroy: ib_destroy_qp(qp); - +err_xrcd: + uobj_put_read(xrcd_uobj); err_put: - put_xrcd_read(xrcd_uobj); - put_uobj_write(&obj->uevent.uobject); + uobj_alloc_abort(&obj->uevent.uobject); return ret; } @@ -2223,6 +1800,7 @@ ssize_t ib_uverbs_query_qp(struct ib_uverbs_file *file, struct ib_qp *qp; struct ib_qp_attr *attr; struct ib_qp_init_attr *init_attr; + const struct ib_global_route *grh; int ret; if (copy_from_user(&cmd, buf, sizeof cmd)) @@ -2235,7 +1813,7 @@ ssize_t ib_uverbs_query_qp(struct ib_uverbs_file *file, goto out; } - qp = idr_read_qp(cmd.qp_handle, file->ucontext); + qp = uobj_get_obj_read(qp, cmd.qp_handle, file->ucontext); if (!qp) { ret = -EINVAL; goto out; @@ -2243,7 +1821,7 @@ ssize_t ib_uverbs_query_qp(struct ib_uverbs_file *file, ret = ib_query_qp(qp, attr, cmd.attr_mask, init_attr); - put_qp_read(qp); + uobj_put_obj_read(qp); if (ret) goto out; @@ -2272,29 +1850,39 @@ ssize_t ib_uverbs_query_qp(struct ib_uverbs_file *file, resp.alt_port_num = attr->alt_port_num; resp.alt_timeout = attr->alt_timeout; - memcpy(resp.dest.dgid, attr->ah_attr.grh.dgid.raw, 16); - resp.dest.flow_label = attr->ah_attr.grh.flow_label; - resp.dest.sgid_index = attr->ah_attr.grh.sgid_index; - resp.dest.hop_limit = attr->ah_attr.grh.hop_limit; - resp.dest.traffic_class = attr->ah_attr.grh.traffic_class; - resp.dest.dlid = attr->ah_attr.dlid; - resp.dest.sl = attr->ah_attr.sl; - resp.dest.src_path_bits = attr->ah_attr.src_path_bits; - resp.dest.static_rate = attr->ah_attr.static_rate; - resp.dest.is_global = !!(attr->ah_attr.ah_flags & IB_AH_GRH); - resp.dest.port_num = attr->ah_attr.port_num; - - memcpy(resp.alt_dest.dgid, attr->alt_ah_attr.grh.dgid.raw, 16); - resp.alt_dest.flow_label = attr->alt_ah_attr.grh.flow_label; - resp.alt_dest.sgid_index = attr->alt_ah_attr.grh.sgid_index; - resp.alt_dest.hop_limit = attr->alt_ah_attr.grh.hop_limit; - resp.alt_dest.traffic_class = attr->alt_ah_attr.grh.traffic_class; - resp.alt_dest.dlid = attr->alt_ah_attr.dlid; - resp.alt_dest.sl = attr->alt_ah_attr.sl; - resp.alt_dest.src_path_bits = attr->alt_ah_attr.src_path_bits; - resp.alt_dest.static_rate = attr->alt_ah_attr.static_rate; - resp.alt_dest.is_global = !!(attr->alt_ah_attr.ah_flags & IB_AH_GRH); - resp.alt_dest.port_num = attr->alt_ah_attr.port_num; + resp.dest.dlid = rdma_ah_get_dlid(&attr->ah_attr); + resp.dest.sl = rdma_ah_get_sl(&attr->ah_attr); + resp.dest.src_path_bits = rdma_ah_get_path_bits(&attr->ah_attr); + resp.dest.static_rate = rdma_ah_get_static_rate(&attr->ah_attr); + resp.dest.is_global = !!(rdma_ah_get_ah_flags(&attr->ah_attr) & + IB_AH_GRH); + if (resp.dest.is_global) { + grh = rdma_ah_read_grh(&attr->ah_attr); + memcpy(resp.dest.dgid, grh->dgid.raw, 16); + resp.dest.flow_label = grh->flow_label; + resp.dest.sgid_index = grh->sgid_index; + resp.dest.hop_limit = grh->hop_limit; + resp.dest.traffic_class = grh->traffic_class; + } + resp.dest.port_num = rdma_ah_get_port_num(&attr->ah_attr); + + resp.alt_dest.dlid = rdma_ah_get_dlid(&attr->alt_ah_attr); + resp.alt_dest.sl = rdma_ah_get_sl(&attr->alt_ah_attr); + resp.alt_dest.src_path_bits = rdma_ah_get_path_bits(&attr->alt_ah_attr); + resp.alt_dest.static_rate + = rdma_ah_get_static_rate(&attr->alt_ah_attr); + resp.alt_dest.is_global + = !!(rdma_ah_get_ah_flags(&attr->alt_ah_attr) & + IB_AH_GRH); + if (resp.alt_dest.is_global) { + grh = rdma_ah_read_grh(&attr->alt_ah_attr); + memcpy(resp.alt_dest.dgid, grh->dgid.raw, 16); + resp.alt_dest.flow_label = grh->flow_label; + resp.alt_dest.sgid_index = grh->sgid_index; + resp.alt_dest.hop_limit = grh->hop_limit; + resp.alt_dest.traffic_class = grh->traffic_class; + } + resp.alt_dest.port_num = rdma_ah_get_port_num(&attr->alt_ah_attr); resp.max_send_wr = init_attr->cap.max_send_wr; resp.max_recv_wr = init_attr->cap.max_recv_wr; @@ -2339,12 +1927,18 @@ static int modify_qp(struct ib_uverbs_file *file, if (!attr) return -ENOMEM; - qp = idr_read_qp(cmd->base.qp_handle, file->ucontext); + qp = uobj_get_obj_read(qp, cmd->base.qp_handle, file->ucontext); if (!qp) { ret = -EINVAL; goto out; } + if ((cmd->base.attr_mask & IB_QP_PORT) && + !rdma_is_port_valid(qp->device, cmd->base.port_num)) { + ret = -EINVAL; + goto release_qp; + } + attr->qp_state = cmd->base.qp_state; attr->cur_qp_state = cmd->base.cur_qp_state; attr->path_mtu = cmd->base.path_mtu; @@ -2368,51 +1962,55 @@ static int modify_qp(struct ib_uverbs_file *file, attr->alt_timeout = cmd->base.alt_timeout; attr->rate_limit = cmd->rate_limit; - memcpy(attr->ah_attr.grh.dgid.raw, cmd->base.dest.dgid, 16); - attr->ah_attr.grh.flow_label = cmd->base.dest.flow_label; - attr->ah_attr.grh.sgid_index = cmd->base.dest.sgid_index; - attr->ah_attr.grh.hop_limit = cmd->base.dest.hop_limit; - attr->ah_attr.grh.traffic_class = cmd->base.dest.traffic_class; - attr->ah_attr.dlid = cmd->base.dest.dlid; - attr->ah_attr.sl = cmd->base.dest.sl; - attr->ah_attr.src_path_bits = cmd->base.dest.src_path_bits; - attr->ah_attr.static_rate = cmd->base.dest.static_rate; - attr->ah_attr.ah_flags = cmd->base.dest.is_global ? - IB_AH_GRH : 0; - attr->ah_attr.port_num = cmd->base.dest.port_num; - - memcpy(attr->alt_ah_attr.grh.dgid.raw, cmd->base.alt_dest.dgid, 16); - attr->alt_ah_attr.grh.flow_label = cmd->base.alt_dest.flow_label; - attr->alt_ah_attr.grh.sgid_index = cmd->base.alt_dest.sgid_index; - attr->alt_ah_attr.grh.hop_limit = cmd->base.alt_dest.hop_limit; - attr->alt_ah_attr.grh.traffic_class = cmd->base.alt_dest.traffic_class; - attr->alt_ah_attr.dlid = cmd->base.alt_dest.dlid; - attr->alt_ah_attr.sl = cmd->base.alt_dest.sl; - attr->alt_ah_attr.src_path_bits = cmd->base.alt_dest.src_path_bits; - attr->alt_ah_attr.static_rate = cmd->base.alt_dest.static_rate; - attr->alt_ah_attr.ah_flags = cmd->base.alt_dest.is_global ? - IB_AH_GRH : 0; - attr->alt_ah_attr.port_num = cmd->base.alt_dest.port_num; - - if (qp->real_qp == qp) { - if (cmd->base.attr_mask & IB_QP_AV) { - ret = ib_resolve_eth_dmac(qp->device, &attr->ah_attr); - if (ret) - goto release_qp; - } - ret = qp->device->modify_qp(qp, attr, - modify_qp_mask(qp->qp_type, - cmd->base.attr_mask), - udata); + attr->ah_attr.type = rdma_ah_find_type(qp->device, + cmd->base.dest.port_num); + if (cmd->base.dest.is_global) { + rdma_ah_set_grh(&attr->ah_attr, NULL, + cmd->base.dest.flow_label, + cmd->base.dest.sgid_index, + cmd->base.dest.hop_limit, + cmd->base.dest.traffic_class); + rdma_ah_set_dgid_raw(&attr->ah_attr, cmd->base.dest.dgid); + } else { + rdma_ah_set_ah_flags(&attr->ah_attr, 0); + } + rdma_ah_set_dlid(&attr->ah_attr, cmd->base.dest.dlid); + rdma_ah_set_sl(&attr->ah_attr, cmd->base.dest.sl); + rdma_ah_set_path_bits(&attr->ah_attr, cmd->base.dest.src_path_bits); + rdma_ah_set_static_rate(&attr->ah_attr, cmd->base.dest.static_rate); + rdma_ah_set_port_num(&attr->ah_attr, + cmd->base.dest.port_num); + + attr->alt_ah_attr.type = rdma_ah_find_type(qp->device, + cmd->base.dest.port_num); + if (cmd->base.alt_dest.is_global) { + rdma_ah_set_grh(&attr->alt_ah_attr, NULL, + cmd->base.alt_dest.flow_label, + cmd->base.alt_dest.sgid_index, + cmd->base.alt_dest.hop_limit, + cmd->base.alt_dest.traffic_class); + rdma_ah_set_dgid_raw(&attr->alt_ah_attr, + cmd->base.alt_dest.dgid); } else { - ret = ib_modify_qp(qp, attr, - modify_qp_mask(qp->qp_type, - cmd->base.attr_mask)); + rdma_ah_set_ah_flags(&attr->alt_ah_attr, 0); } -release_qp: - put_qp_read(qp); + rdma_ah_set_dlid(&attr->alt_ah_attr, cmd->base.alt_dest.dlid); + rdma_ah_set_sl(&attr->alt_ah_attr, cmd->base.alt_dest.sl); + rdma_ah_set_path_bits(&attr->alt_ah_attr, + cmd->base.alt_dest.src_path_bits); + rdma_ah_set_static_rate(&attr->alt_ah_attr, + cmd->base.alt_dest.static_rate); + rdma_ah_set_port_num(&attr->alt_ah_attr, + cmd->base.alt_dest.port_num); + + ret = ib_modify_qp_with_udata(qp, attr, + modify_qp_mask(qp->qp_type, + cmd->base.attr_mask), + udata); +release_qp: + uobj_put_obj_read(qp); out: kfree(attr); @@ -2489,7 +2087,6 @@ ssize_t ib_uverbs_destroy_qp(struct ib_uverbs_file *file, struct ib_uverbs_destroy_qp cmd; struct ib_uverbs_destroy_qp_resp resp; struct ib_uobject *uobj; - struct ib_qp *qp; struct ib_uqp_object *obj; int ret = -EINVAL; @@ -2498,40 +2095,26 @@ ssize_t ib_uverbs_destroy_qp(struct ib_uverbs_file *file, memset(&resp, 0, sizeof resp); - uobj = idr_write_uobj(&ib_uverbs_qp_idr, cmd.qp_handle, file->ucontext); - if (!uobj) - return -EINVAL; - qp = uobj->object; - obj = container_of(uobj, struct ib_uqp_object, uevent.uobject); - - if (!list_empty(&obj->mcast_list)) { - put_uobj_write(uobj); - return -EBUSY; - } - - ret = ib_destroy_qp(qp); - if (!ret) - uobj->live = 0; + uobj = uobj_get_write(uobj_get_type(qp), cmd.qp_handle, + file->ucontext); + if (IS_ERR(uobj)) + return PTR_ERR(uobj); - put_uobj_write(uobj); + obj = container_of(uobj, struct ib_uqp_object, uevent.uobject); + /* + * Make sure we don't free the memory in remove_commit as we still + * needs the uobject memory to create the response. + */ + uverbs_uobject_get(uobj); - if (ret) + ret = uobj_remove_commit(uobj); + if (ret) { + uverbs_uobject_put(uobj); return ret; - - if (obj->uxrcd) - atomic_dec(&obj->uxrcd->refcnt); - - idr_remove_uobj(&ib_uverbs_qp_idr, uobj); - - mutex_lock(&file->mutex); - list_del(&uobj->list); - mutex_unlock(&file->mutex); - - ib_uverbs_release_uevent(file, &obj->uevent); + } resp.events_reported = obj->uevent.events_reported; - - put_uobj(uobj); + uverbs_uobject_put(uobj); if (copy_to_user((void __user *) (unsigned long) cmd.response, &resp, sizeof resp)) @@ -2542,9 +2125,13 @@ ssize_t ib_uverbs_destroy_qp(struct ib_uverbs_file *file, static void *alloc_wr(size_t wr_size, __u32 num_sge) { + if (num_sge >= (U32_MAX - ALIGN(wr_size, sizeof (struct ib_sge))) / + sizeof (struct ib_sge)) + return NULL; + return kmalloc(ALIGN(wr_size, sizeof (struct ib_sge)) + num_sge * sizeof (struct ib_sge), GFP_KERNEL); -}; +} ssize_t ib_uverbs_post_send(struct ib_uverbs_file *file, struct ib_device *ib_dev, @@ -2575,7 +2162,7 @@ ssize_t ib_uverbs_post_send(struct ib_uverbs_file *file, if (!user_wr) return -ENOMEM; - qp = idr_read_qp(cmd.qp_handle, file->ucontext); + qp = uobj_get_obj_read(qp, cmd.qp_handle, file->ucontext); if (!qp) goto out; @@ -2611,7 +2198,8 @@ ssize_t ib_uverbs_post_send(struct ib_uverbs_file *file, goto out_put; } - ud->ah = idr_read_ah(user_wr->wr.ud.ah, file->ucontext); + ud->ah = uobj_get_obj_read(ah, user_wr->wr.ud.ah, + file->ucontext); if (!ud->ah) { kfree(ud); ret = -EINVAL; @@ -2718,11 +2306,11 @@ ssize_t ib_uverbs_post_send(struct ib_uverbs_file *file, ret = -EFAULT; out_put: - put_qp_read(qp); + uobj_put_obj_read(qp); while (wr) { if (is_ud && ud_wr(wr)->ah) - put_ah_read(ud_wr(wr)->ah); + uobj_put_obj_read(ud_wr(wr)->ah); next = wr->next; kfree(wr); wr = next; @@ -2771,6 +2359,13 @@ static struct ib_recv_wr *ib_uverbs_unmarshall_recv(const char __user *buf, goto err; } + if (user_wr->num_sge >= + (U32_MAX - ALIGN(sizeof *next, sizeof (struct ib_sge))) / + sizeof (struct ib_sge)) { + ret = -EINVAL; + goto err; + } + next = kmalloc(ALIGN(sizeof *next, sizeof (struct ib_sge)) + user_wr->num_sge * sizeof (struct ib_sge), GFP_KERNEL); @@ -2839,21 +2434,21 @@ ssize_t ib_uverbs_post_recv(struct ib_uverbs_file *file, if (IS_ERR(wr)) return PTR_ERR(wr); - qp = idr_read_qp(cmd.qp_handle, file->ucontext); + qp = uobj_get_obj_read(qp, cmd.qp_handle, file->ucontext); if (!qp) goto out; resp.bad_wr = 0; ret = qp->device->post_recv(qp->real_qp, wr, &bad_wr); - put_qp_read(qp); - - if (ret) + uobj_put_obj_read(qp); + if (ret) { for (next = wr; next; next = next->next) { ++resp.bad_wr; if (next == bad_wr) break; } + } if (copy_to_user((void __user *) (unsigned long) cmd.response, &resp, sizeof resp)) @@ -2889,14 +2484,14 @@ ssize_t ib_uverbs_post_srq_recv(struct ib_uverbs_file *file, if (IS_ERR(wr)) return PTR_ERR(wr); - srq = idr_read_srq(cmd.srq_handle, file->ucontext); + srq = uobj_get_obj_read(srq, cmd.srq_handle, file->ucontext); if (!srq) goto out; resp.bad_wr = 0; ret = srq->device->post_srq_recv(srq, wr, &bad_wr); - put_srq_read(srq); + uobj_put_obj_read(srq); if (ret) for (next = wr; next; next = next->next) { @@ -2929,9 +2524,10 @@ ssize_t ib_uverbs_create_ah(struct ib_uverbs_file *file, struct ib_uobject *uobj; struct ib_pd *pd; struct ib_ah *ah; - struct ib_ah_attr attr; + struct rdma_ah_attr attr; int ret; struct ib_udata udata; + u8 *dmac; if (out_len < sizeof resp) return -ENOSPC; @@ -2939,35 +2535,42 @@ ssize_t ib_uverbs_create_ah(struct ib_uverbs_file *file, if (copy_from_user(&cmd, buf, sizeof cmd)) return -EFAULT; + if (!rdma_is_port_valid(ib_dev, cmd.attr.port_num)) + return -EINVAL; + INIT_UDATA(&udata, buf + sizeof(cmd), (unsigned long)cmd.response + sizeof(resp), in_len - sizeof(cmd), out_len - sizeof(resp)); - uobj = kmalloc(sizeof *uobj, GFP_KERNEL); - if (!uobj) - return -ENOMEM; - - init_uobj(uobj, cmd.user_handle, file->ucontext, &ah_lock_class); - down_write(&uobj->mutex); + uobj = uobj_alloc(uobj_get_type(ah), file->ucontext); + if (IS_ERR(uobj)) + return PTR_ERR(uobj); - pd = idr_read_pd(cmd.pd_handle, file->ucontext); + pd = uobj_get_obj_read(pd, cmd.pd_handle, file->ucontext); if (!pd) { ret = -EINVAL; goto err; } - attr.dlid = cmd.attr.dlid; - attr.sl = cmd.attr.sl; - attr.src_path_bits = cmd.attr.src_path_bits; - attr.static_rate = cmd.attr.static_rate; - attr.ah_flags = cmd.attr.is_global ? IB_AH_GRH : 0; - attr.port_num = cmd.attr.port_num; - attr.grh.flow_label = cmd.attr.grh.flow_label; - attr.grh.sgid_index = cmd.attr.grh.sgid_index; - attr.grh.hop_limit = cmd.attr.grh.hop_limit; - attr.grh.traffic_class = cmd.attr.grh.traffic_class; - memset(&attr.dmac, 0, sizeof(attr.dmac)); - memcpy(attr.grh.dgid.raw, cmd.attr.grh.dgid, 16); + attr.type = rdma_ah_find_type(ib_dev, cmd.attr.port_num); + rdma_ah_set_dlid(&attr, cmd.attr.dlid); + rdma_ah_set_sl(&attr, cmd.attr.sl); + rdma_ah_set_path_bits(&attr, cmd.attr.src_path_bits); + rdma_ah_set_static_rate(&attr, cmd.attr.static_rate); + rdma_ah_set_port_num(&attr, cmd.attr.port_num); + + if (cmd.attr.is_global) { + rdma_ah_set_grh(&attr, NULL, cmd.attr.grh.flow_label, + cmd.attr.grh.sgid_index, + cmd.attr.grh.hop_limit, + cmd.attr.grh.traffic_class); + rdma_ah_set_dgid_raw(&attr, cmd.attr.grh.dgid); + } else { + rdma_ah_set_ah_flags(&attr, 0); + } + dmac = rdma_ah_retrieve_dmac(&attr); + if (dmac) + memset(dmac, 0, ETH_ALEN); ah = pd->device->create_ah(pd, &attr, &udata); @@ -2980,12 +2583,9 @@ ssize_t ib_uverbs_create_ah(struct ib_uverbs_file *file, ah->pd = pd; atomic_inc(&pd->usecnt); ah->uobject = uobj; + uobj->user_handle = cmd.user_handle; uobj->object = ah; - ret = idr_add_uobj(&ib_uverbs_ah_idr, uobj); - if (ret) - goto err_destroy; - resp.ah_handle = uobj->id; if (copy_to_user((void __user *) (unsigned long) cmd.response, @@ -2994,29 +2594,19 @@ ssize_t ib_uverbs_create_ah(struct ib_uverbs_file *file, goto err_copy; } - put_pd_read(pd); - - mutex_lock(&file->mutex); - list_add_tail(&uobj->list, &file->ucontext->ah_list); - mutex_unlock(&file->mutex); - - uobj->live = 1; - - up_write(&uobj->mutex); + uobj_put_obj_read(pd); + uobj_alloc_commit(uobj); return in_len; err_copy: - idr_remove_uobj(&ib_uverbs_ah_idr, uobj); - -err_destroy: - ib_destroy_ah(ah); + rdma_destroy_ah(ah); err_put: - put_pd_read(pd); + uobj_put_obj_read(pd); err: - put_uobj_write(uobj); + uobj_alloc_abort(uobj); return ret; } @@ -3025,36 +2615,19 @@ ssize_t ib_uverbs_destroy_ah(struct ib_uverbs_file *file, const char __user *buf, int in_len, int out_len) { struct ib_uverbs_destroy_ah cmd; - struct ib_ah *ah; struct ib_uobject *uobj; int ret; if (copy_from_user(&cmd, buf, sizeof cmd)) return -EFAULT; - uobj = idr_write_uobj(&ib_uverbs_ah_idr, cmd.ah_handle, file->ucontext); - if (!uobj) - return -EINVAL; - ah = uobj->object; - - ret = ib_destroy_ah(ah); - if (!ret) - uobj->live = 0; - - put_uobj_write(uobj); - - if (ret) - return ret; - - idr_remove_uobj(&ib_uverbs_ah_idr, uobj); + uobj = uobj_get_write(uobj_get_type(ah), cmd.ah_handle, + file->ucontext); + if (IS_ERR(uobj)) + return PTR_ERR(uobj); - mutex_lock(&file->mutex); - list_del(&uobj->list); - mutex_unlock(&file->mutex); - - put_uobj(uobj); - - return in_len; + ret = uobj_remove_commit(uobj); + return ret ?: in_len; } ssize_t ib_uverbs_attach_mcast(struct ib_uverbs_file *file, @@ -3071,12 +2644,13 @@ ssize_t ib_uverbs_attach_mcast(struct ib_uverbs_file *file, if (copy_from_user(&cmd, buf, sizeof cmd)) return -EFAULT; - qp = idr_write_qp(cmd.qp_handle, file->ucontext); + qp = uobj_get_obj_read(qp, cmd.qp_handle, file->ucontext); if (!qp) return -EINVAL; obj = container_of(qp->uobject, struct ib_uqp_object, uevent.uobject); + mutex_lock(&obj->mcast_lock); list_for_each_entry(mcast, &obj->mcast_list, list) if (cmd.mlid == mcast->lid && !memcmp(cmd.gid, mcast->gid.raw, sizeof mcast->gid.raw)) { @@ -3100,7 +2674,8 @@ ssize_t ib_uverbs_attach_mcast(struct ib_uverbs_file *file, kfree(mcast); out_put: - put_qp_write(qp); + mutex_unlock(&obj->mcast_lock); + uobj_put_obj_read(qp); return ret ? ret : in_len; } @@ -3115,34 +2690,66 @@ ssize_t ib_uverbs_detach_mcast(struct ib_uverbs_file *file, struct ib_qp *qp; struct ib_uverbs_mcast_entry *mcast; int ret = -EINVAL; + bool found = false; if (copy_from_user(&cmd, buf, sizeof cmd)) return -EFAULT; - qp = idr_write_qp(cmd.qp_handle, file->ucontext); + qp = uobj_get_obj_read(qp, cmd.qp_handle, file->ucontext); if (!qp) return -EINVAL; - ret = ib_detach_mcast(qp, (union ib_gid *) cmd.gid, cmd.mlid); - if (ret) - goto out_put; - obj = container_of(qp->uobject, struct ib_uqp_object, uevent.uobject); + mutex_lock(&obj->mcast_lock); list_for_each_entry(mcast, &obj->mcast_list, list) if (cmd.mlid == mcast->lid && !memcmp(cmd.gid, mcast->gid.raw, sizeof mcast->gid.raw)) { list_del(&mcast->list); kfree(mcast); + found = true; break; } -out_put: - put_qp_write(qp); + if (!found) { + ret = -EINVAL; + goto out_put; + } + ret = ib_detach_mcast(qp, (union ib_gid *)cmd.gid, cmd.mlid); + +out_put: + mutex_unlock(&obj->mcast_lock); + uobj_put_obj_read(qp); return ret ? ret : in_len; } +static int kern_spec_to_ib_spec_action(struct ib_uverbs_flow_spec *kern_spec, + union ib_flow_spec *ib_spec) +{ + ib_spec->type = kern_spec->type; + switch (ib_spec->type) { + case IB_FLOW_SPEC_ACTION_TAG: + if (kern_spec->flow_tag.size != + sizeof(struct ib_uverbs_flow_spec_action_tag)) + return -EINVAL; + + ib_spec->flow_tag.size = sizeof(struct ib_flow_spec_action_tag); + ib_spec->flow_tag.tag_id = kern_spec->flow_tag.tag_id; + break; + case IB_FLOW_SPEC_ACTION_DROP: + if (kern_spec->drop.size != + sizeof(struct ib_uverbs_flow_spec_action_drop)) + return -EINVAL; + + ib_spec->drop.size = sizeof(struct ib_flow_spec_action_drop); + break; + default: + return -EINVAL; + } + return 0; +} + static size_t kern_spec_filter_sz(struct ib_uverbs_flow_spec_hdr *spec) { /* Returns user space filter size, includes padding */ @@ -3167,8 +2774,8 @@ static ssize_t spec_filter_size(void *kern_spec_filter, u16 kern_filter_size, return kern_filter_size; } -static int kern_spec_to_ib_spec(struct ib_uverbs_flow_spec *kern_spec, - union ib_flow_spec *ib_spec) +static int kern_spec_to_ib_spec_filter(struct ib_uverbs_flow_spec *kern_spec, + union ib_flow_spec *ib_spec) { ssize_t actual_filter_sz; ssize_t kern_filter_sz; @@ -3263,6 +2870,18 @@ static int kern_spec_to_ib_spec(struct ib_uverbs_flow_spec *kern_spec, return 0; } +static int kern_spec_to_ib_spec(struct ib_uverbs_flow_spec *kern_spec, + union ib_flow_spec *ib_spec) +{ + if (kern_spec->reserved) + return -EINVAL; + + if (kern_spec->type >= IB_FLOW_SPEC_ACTION_TAG) + return kern_spec_to_ib_spec_action(kern_spec, ib_spec); + else + return kern_spec_to_ib_spec_filter(kern_spec, ib_spec); +} + int ib_uverbs_ex_create_wq(struct ib_uverbs_file *file, struct ib_device *ib_dev, struct ib_udata *ucore, @@ -3300,20 +2919,18 @@ int ib_uverbs_ex_create_wq(struct ib_uverbs_file *file, if (cmd.comp_mask) return -EOPNOTSUPP; - obj = kmalloc(sizeof(*obj), GFP_KERNEL); - if (!obj) - return -ENOMEM; + obj = (struct ib_uwq_object *)uobj_alloc(uobj_get_type(wq), + file->ucontext); + if (IS_ERR(obj)) + return PTR_ERR(obj); - init_uobj(&obj->uevent.uobject, cmd.user_handle, file->ucontext, - &wq_lock_class); - down_write(&obj->uevent.uobject.mutex); - pd = idr_read_pd(cmd.pd_handle, file->ucontext); + pd = uobj_get_obj_read(pd, cmd.pd_handle, file->ucontext); if (!pd) { err = -EINVAL; goto err_uobj; } - cq = idr_read_cq(cmd.cq_handle, file->ucontext, 0); + cq = uobj_get_obj_read(cq, cmd.cq_handle, file->ucontext); if (!cq) { err = -EINVAL; goto err_put_pd; @@ -3325,6 +2942,9 @@ int ib_uverbs_ex_create_wq(struct ib_uverbs_file *file, wq_init_attr.wq_context = file; wq_init_attr.wq_type = cmd.wq_type; wq_init_attr.event_handler = ib_uverbs_wq_event_handler; + if (ucore->inlen >= (offsetof(typeof(cmd), create_flags) + + sizeof(cmd.create_flags))) + wq_init_attr.create_flags = cmd.create_flags; obj->uevent.events_reported = 0; INIT_LIST_HEAD(&obj->uevent.event_list); wq = pd->device->create_wq(pd, &wq_init_attr, uhw); @@ -3345,9 +2965,6 @@ int ib_uverbs_ex_create_wq(struct ib_uverbs_file *file, atomic_inc(&cq->usecnt); wq->uobject = &obj->uevent.uobject; obj->uevent.uobject.object = wq; - err = idr_add_uobj(&ib_uverbs_wq_idr, &obj->uevent.uobject); - if (err) - goto destroy_wq; memset(&resp, 0, sizeof(resp)); resp.wq_handle = obj->uevent.uobject.id; @@ -3360,27 +2977,19 @@ int ib_uverbs_ex_create_wq(struct ib_uverbs_file *file, if (err) goto err_copy; - put_pd_read(pd); - put_cq_read(cq); - - mutex_lock(&file->mutex); - list_add_tail(&obj->uevent.uobject.list, &file->ucontext->wq_list); - mutex_unlock(&file->mutex); - - obj->uevent.uobject.live = 1; - up_write(&obj->uevent.uobject.mutex); + uobj_put_obj_read(pd); + uobj_put_obj_read(cq); + uobj_alloc_commit(&obj->uevent.uobject); return 0; err_copy: - idr_remove_uobj(&ib_uverbs_wq_idr, &obj->uevent.uobject); -destroy_wq: ib_destroy_wq(wq); err_put_cq: - put_cq_read(cq); + uobj_put_obj_read(cq); err_put_pd: - put_pd_read(pd); + uobj_put_obj_read(pd); err_uobj: - put_uobj_write(&obj->uevent.uobject); + uobj_alloc_abort(&obj->uevent.uobject); return err; } @@ -3392,7 +3001,6 @@ int ib_uverbs_ex_destroy_wq(struct ib_uverbs_file *file, { struct ib_uverbs_ex_destroy_wq cmd = {}; struct ib_uverbs_ex_destroy_wq_resp resp = {}; - struct ib_wq *wq; struct ib_uobject *uobj; struct ib_uwq_object *obj; size_t required_cmd_sz; @@ -3421,36 +3029,25 @@ int ib_uverbs_ex_destroy_wq(struct ib_uverbs_file *file, return -EOPNOTSUPP; resp.response_length = required_resp_len; - uobj = idr_write_uobj(&ib_uverbs_wq_idr, cmd.wq_handle, - file->ucontext); - if (!uobj) - return -EINVAL; + uobj = uobj_get_write(uobj_get_type(wq), cmd.wq_handle, + file->ucontext); + if (IS_ERR(uobj)) + return PTR_ERR(uobj); - wq = uobj->object; obj = container_of(uobj, struct ib_uwq_object, uevent.uobject); - ret = ib_destroy_wq(wq); - if (!ret) - uobj->live = 0; - - put_uobj_write(uobj); - if (ret) - return ret; - - idr_remove_uobj(&ib_uverbs_wq_idr, uobj); - - mutex_lock(&file->mutex); - list_del(&uobj->list); - mutex_unlock(&file->mutex); + /* + * Make sure we don't free the memory in remove_commit as we still + * needs the uobject memory to create the response. + */ + uverbs_uobject_get(uobj); - ib_uverbs_release_uevent(file, &obj->uevent); + ret = uobj_remove_commit(uobj); resp.events_reported = obj->uevent.events_reported; - put_uobj(uobj); - - ret = ib_copy_to_udata(ucore, &resp, resp.response_length); + uverbs_uobject_put(uobj); if (ret) return ret; - return 0; + return ib_copy_to_udata(ucore, &resp, resp.response_length); } int ib_uverbs_ex_modify_wq(struct ib_uverbs_file *file, @@ -3480,17 +3077,21 @@ int ib_uverbs_ex_modify_wq(struct ib_uverbs_file *file, if (!cmd.attr_mask) return -EINVAL; - if (cmd.attr_mask > (IB_WQ_STATE | IB_WQ_CUR_STATE)) + if (cmd.attr_mask > (IB_WQ_STATE | IB_WQ_CUR_STATE | IB_WQ_FLAGS)) return -EINVAL; - wq = idr_read_wq(cmd.wq_handle, file->ucontext); + wq = uobj_get_obj_read(wq, cmd.wq_handle, file->ucontext); if (!wq) return -EINVAL; wq_attr.curr_wq_state = cmd.curr_wq_state; wq_attr.wq_state = cmd.wq_state; + if (cmd.attr_mask & IB_WQ_FLAGS) { + wq_attr.flags = cmd.flags; + wq_attr.flags_mask = cmd.flags_mask; + } ret = wq->device->modify_wq(wq, &wq_attr, cmd.attr_mask, uhw); - put_wq_read(wq); + uobj_put_obj_read(wq); return ret; } @@ -3568,7 +3169,8 @@ int ib_uverbs_ex_create_rwq_ind_table(struct ib_uverbs_file *file, for (num_read_wqs = 0; num_read_wqs < num_wq_handles; num_read_wqs++) { - wq = idr_read_wq(wqs_handles[num_read_wqs], file->ucontext); + wq = uobj_get_obj_read(wq, wqs_handles[num_read_wqs], + file->ucontext); if (!wq) { err = -EINVAL; goto put_wqs; @@ -3577,14 +3179,12 @@ int ib_uverbs_ex_create_rwq_ind_table(struct ib_uverbs_file *file, wqs[num_read_wqs] = wq; } - uobj = kmalloc(sizeof(*uobj), GFP_KERNEL); - if (!uobj) { - err = -ENOMEM; + uobj = uobj_alloc(uobj_get_type(rwq_ind_table), file->ucontext); + if (IS_ERR(uobj)) { + err = PTR_ERR(uobj); goto put_wqs; } - init_uobj(uobj, 0, file->ucontext, &rwq_ind_table_lock_class); - down_write(&uobj->mutex); init_attr.log_ind_tbl_size = cmd.log_ind_tbl_size; init_attr.ind_tbl = wqs; rwq_ind_tbl = ib_dev->create_rwq_ind_table(ib_dev, &init_attr, uhw); @@ -3604,10 +3204,6 @@ int ib_uverbs_ex_create_rwq_ind_table(struct ib_uverbs_file *file, for (i = 0; i < num_wq_handles; i++) atomic_inc(&wqs[i]->usecnt); - err = idr_add_uobj(&ib_uverbs_rwq_ind_tbl_idr, uobj); - if (err) - goto destroy_ind_tbl; - resp.ind_tbl_handle = uobj->id; resp.ind_tbl_num = rwq_ind_tbl->ind_tbl_num; resp.response_length = required_resp_len; @@ -3620,26 +3216,18 @@ int ib_uverbs_ex_create_rwq_ind_table(struct ib_uverbs_file *file, kfree(wqs_handles); for (j = 0; j < num_read_wqs; j++) - put_wq_read(wqs[j]); - - mutex_lock(&file->mutex); - list_add_tail(&uobj->list, &file->ucontext->rwq_ind_tbl_list); - mutex_unlock(&file->mutex); + uobj_put_obj_read(wqs[j]); - uobj->live = 1; - - up_write(&uobj->mutex); + uobj_alloc_commit(uobj); return 0; err_copy: - idr_remove_uobj(&ib_uverbs_rwq_ind_tbl_idr, uobj); -destroy_ind_tbl: ib_destroy_rwq_ind_table(rwq_ind_tbl); err_uobj: - put_uobj_write(uobj); + uobj_alloc_abort(uobj); put_wqs: for (j = 0; j < num_read_wqs; j++) - put_wq_read(wqs[j]); + uobj_put_obj_read(wqs[j]); err_free: kfree(wqs_handles); kfree(wqs); @@ -3652,10 +3240,8 @@ int ib_uverbs_ex_destroy_rwq_ind_table(struct ib_uverbs_file *file, struct ib_udata *uhw) { struct ib_uverbs_ex_destroy_rwq_ind_table cmd = {}; - struct ib_rwq_ind_table *rwq_ind_tbl; struct ib_uobject *uobj; int ret; - struct ib_wq **ind_tbl; size_t required_cmd_sz; required_cmd_sz = offsetof(typeof(cmd), ind_tbl_handle) + sizeof(cmd.ind_tbl_handle); @@ -3675,31 +3261,12 @@ int ib_uverbs_ex_destroy_rwq_ind_table(struct ib_uverbs_file *file, if (cmd.comp_mask) return -EOPNOTSUPP; - uobj = idr_write_uobj(&ib_uverbs_rwq_ind_tbl_idr, cmd.ind_tbl_handle, - file->ucontext); - if (!uobj) - return -EINVAL; - rwq_ind_tbl = uobj->object; - ind_tbl = rwq_ind_tbl->ind_tbl; - - ret = ib_destroy_rwq_ind_table(rwq_ind_tbl); - if (!ret) - uobj->live = 0; - - put_uobj_write(uobj); + uobj = uobj_get_write(uobj_get_type(rwq_ind_table), cmd.ind_tbl_handle, + file->ucontext); + if (IS_ERR(uobj)) + return PTR_ERR(uobj); - if (ret) - return ret; - - idr_remove_uobj(&ib_uverbs_rwq_ind_tbl_idr, uobj); - - mutex_lock(&file->mutex); - list_del(&uobj->list); - mutex_unlock(&file->mutex); - - put_uobj(uobj); - kfree(ind_tbl); - return ret; + return uobj_remove_commit(uobj); } int ib_uverbs_ex_create_flow(struct ib_uverbs_file *file, @@ -3773,15 +3340,13 @@ int ib_uverbs_ex_create_flow(struct ib_uverbs_file *file, kern_flow_attr = &cmd.flow_attr; } - uobj = kmalloc(sizeof(*uobj), GFP_KERNEL); - if (!uobj) { - err = -ENOMEM; + uobj = uobj_alloc(uobj_get_type(flow), file->ucontext); + if (IS_ERR(uobj)) { + err = PTR_ERR(uobj); goto err_free_attr; } - init_uobj(uobj, 0, file->ucontext, &rule_lock_class); - down_write(&uobj->mutex); - qp = idr_read_qp(cmd.qp_handle, file->ucontext); + qp = uobj_get_obj_read(qp, cmd.qp_handle, file->ucontext); if (!qp) { err = -EINVAL; goto err_uobj; @@ -3830,10 +3395,6 @@ int ib_uverbs_ex_create_flow(struct ib_uverbs_file *file, flow_id->uobject = uobj; uobj->object = flow_id; - err = idr_add_uobj(&ib_uverbs_rule_idr, uobj); - if (err) - goto destroy_flow; - memset(&resp, 0, sizeof(resp)); resp.flow_handle = uobj->id; @@ -3842,28 +3403,20 @@ int ib_uverbs_ex_create_flow(struct ib_uverbs_file *file, if (err) goto err_copy; - put_qp_read(qp); - mutex_lock(&file->mutex); - list_add_tail(&uobj->list, &file->ucontext->rule_list); - mutex_unlock(&file->mutex); - - uobj->live = 1; - - up_write(&uobj->mutex); + uobj_put_obj_read(qp); + uobj_alloc_commit(uobj); kfree(flow_attr); if (cmd.flow_attr.num_of_specs) kfree(kern_flow_attr); return 0; err_copy: - idr_remove_uobj(&ib_uverbs_rule_idr, uobj); -destroy_flow: ib_destroy_flow(flow_id); err_free: kfree(flow_attr); err_put: - put_qp_read(qp); + uobj_put_obj_read(qp); err_uobj: - put_uobj_write(uobj); + uobj_alloc_abort(uobj); err_free_attr: if (cmd.flow_attr.num_of_specs) kfree(kern_flow_attr); @@ -3876,7 +3429,6 @@ int ib_uverbs_ex_destroy_flow(struct ib_uverbs_file *file, struct ib_udata *uhw) { struct ib_uverbs_destroy_flow cmd; - struct ib_flow *flow_id; struct ib_uobject *uobj; int ret; @@ -3890,26 +3442,12 @@ int ib_uverbs_ex_destroy_flow(struct ib_uverbs_file *file, if (cmd.comp_mask) return -EINVAL; - uobj = idr_write_uobj(&ib_uverbs_rule_idr, cmd.flow_handle, - file->ucontext); - if (!uobj) - return -EINVAL; - flow_id = uobj->object; - - ret = ib_destroy_flow(flow_id); - if (!ret) - uobj->live = 0; - - put_uobj_write(uobj); - - idr_remove_uobj(&ib_uverbs_rule_idr, uobj); - - mutex_lock(&file->mutex); - list_del(&uobj->list); - mutex_unlock(&file->mutex); - - put_uobj(uobj); + uobj = uobj_get_write(uobj_get_type(flow), cmd.flow_handle, + file->ucontext); + if (IS_ERR(uobj)) + return PTR_ERR(uobj); + ret = uobj_remove_commit(uobj); return ret; } @@ -3926,31 +3464,37 @@ static int __uverbs_create_xsrq(struct ib_uverbs_file *file, struct ib_srq_init_attr attr; int ret; - obj = kmalloc(sizeof *obj, GFP_KERNEL); - if (!obj) - return -ENOMEM; - - init_uobj(&obj->uevent.uobject, cmd->user_handle, file->ucontext, &srq_lock_class); - down_write(&obj->uevent.uobject.mutex); + obj = (struct ib_usrq_object *)uobj_alloc(uobj_get_type(srq), + file->ucontext); + if (IS_ERR(obj)) + return PTR_ERR(obj); if (cmd->srq_type == IB_SRQT_XRC) { - attr.ext.xrc.xrcd = idr_read_xrcd(cmd->xrcd_handle, file->ucontext, &xrcd_uobj); - if (!attr.ext.xrc.xrcd) { + xrcd_uobj = uobj_get_read(uobj_get_type(xrcd), cmd->xrcd_handle, + file->ucontext); + if (IS_ERR(xrcd_uobj)) { ret = -EINVAL; goto err; } + attr.ext.xrc.xrcd = (struct ib_xrcd *)xrcd_uobj->object; + if (!attr.ext.xrc.xrcd) { + ret = -EINVAL; + goto err_put_xrcd; + } + obj->uxrcd = container_of(xrcd_uobj, struct ib_uxrcd_object, uobject); atomic_inc(&obj->uxrcd->refcnt); - attr.ext.xrc.cq = idr_read_cq(cmd->cq_handle, file->ucontext, 0); + attr.ext.xrc.cq = uobj_get_obj_read(cq, cmd->cq_handle, + file->ucontext); if (!attr.ext.xrc.cq) { ret = -EINVAL; goto err_put_xrcd; } } - pd = idr_read_pd(cmd->pd_handle, file->ucontext); + pd = uobj_get_obj_read(pd, cmd->pd_handle, file->ucontext); if (!pd) { ret = -EINVAL; goto err_put_cq; @@ -3990,9 +3534,7 @@ static int __uverbs_create_xsrq(struct ib_uverbs_file *file, atomic_set(&srq->usecnt, 0); obj->uevent.uobject.object = srq; - ret = idr_add_uobj(&ib_uverbs_srq_idr, &obj->uevent.uobject); - if (ret) - goto err_destroy; + obj->uevent.uobject.user_handle = cmd->user_handle; memset(&resp, 0, sizeof resp); resp.srq_handle = obj->uevent.uobject.id; @@ -4008,42 +3550,32 @@ static int __uverbs_create_xsrq(struct ib_uverbs_file *file, } if (cmd->srq_type == IB_SRQT_XRC) { - put_uobj_read(xrcd_uobj); - put_cq_read(attr.ext.xrc.cq); + uobj_put_read(xrcd_uobj); + uobj_put_obj_read(attr.ext.xrc.cq); } - put_pd_read(pd); - - mutex_lock(&file->mutex); - list_add_tail(&obj->uevent.uobject.list, &file->ucontext->srq_list); - mutex_unlock(&file->mutex); - - obj->uevent.uobject.live = 1; - - up_write(&obj->uevent.uobject.mutex); + uobj_put_obj_read(pd); + uobj_alloc_commit(&obj->uevent.uobject); return 0; err_copy: - idr_remove_uobj(&ib_uverbs_srq_idr, &obj->uevent.uobject); - -err_destroy: ib_destroy_srq(srq); err_put: - put_pd_read(pd); + uobj_put_obj_read(pd); err_put_cq: if (cmd->srq_type == IB_SRQT_XRC) - put_cq_read(attr.ext.xrc.cq); + uobj_put_obj_read(attr.ext.xrc.cq); err_put_xrcd: if (cmd->srq_type == IB_SRQT_XRC) { atomic_dec(&obj->uxrcd->refcnt); - put_uobj_read(xrcd_uobj); + uobj_put_read(xrcd_uobj); } err: - put_uobj_write(&obj->uevent.uobject); + uobj_alloc_abort(&obj->uevent.uobject); return ret; } @@ -4128,7 +3660,7 @@ ssize_t ib_uverbs_modify_srq(struct ib_uverbs_file *file, INIT_UDATA(&udata, buf + sizeof cmd, NULL, in_len - sizeof cmd, out_len); - srq = idr_read_srq(cmd.srq_handle, file->ucontext); + srq = uobj_get_obj_read(srq, cmd.srq_handle, file->ucontext); if (!srq) return -EINVAL; @@ -4137,7 +3669,7 @@ ssize_t ib_uverbs_modify_srq(struct ib_uverbs_file *file, ret = srq->device->modify_srq(srq, &attr, cmd.attr_mask, &udata); - put_srq_read(srq); + uobj_put_obj_read(srq); return ret ? ret : in_len; } @@ -4159,13 +3691,13 @@ ssize_t ib_uverbs_query_srq(struct ib_uverbs_file *file, if (copy_from_user(&cmd, buf, sizeof cmd)) return -EFAULT; - srq = idr_read_srq(cmd.srq_handle, file->ucontext); + srq = uobj_get_obj_read(srq, cmd.srq_handle, file->ucontext); if (!srq) return -EINVAL; ret = ib_query_srq(srq, &attr); - put_srq_read(srq); + uobj_put_obj_read(srq); if (ret) return ret; @@ -4191,54 +3723,38 @@ ssize_t ib_uverbs_destroy_srq(struct ib_uverbs_file *file, struct ib_uverbs_destroy_srq cmd; struct ib_uverbs_destroy_srq_resp resp; struct ib_uobject *uobj; - struct ib_srq *srq; struct ib_uevent_object *obj; int ret = -EINVAL; - struct ib_usrq_object *us; - enum ib_srq_type srq_type; if (copy_from_user(&cmd, buf, sizeof cmd)) return -EFAULT; - uobj = idr_write_uobj(&ib_uverbs_srq_idr, cmd.srq_handle, file->ucontext); - if (!uobj) - return -EINVAL; - srq = uobj->object; - obj = container_of(uobj, struct ib_uevent_object, uobject); - srq_type = srq->srq_type; + uobj = uobj_get_write(uobj_get_type(srq), cmd.srq_handle, + file->ucontext); + if (IS_ERR(uobj)) + return PTR_ERR(uobj); - ret = ib_destroy_srq(srq); - if (!ret) - uobj->live = 0; + obj = container_of(uobj, struct ib_uevent_object, uobject); + /* + * Make sure we don't free the memory in remove_commit as we still + * needs the uobject memory to create the response. + */ + uverbs_uobject_get(uobj); - put_uobj_write(uobj); + memset(&resp, 0, sizeof(resp)); - if (ret) + ret = uobj_remove_commit(uobj); + if (ret) { + uverbs_uobject_put(uobj); return ret; - - if (srq_type == IB_SRQT_XRC) { - us = container_of(obj, struct ib_usrq_object, uevent); - atomic_dec(&us->uxrcd->refcnt); } - - idr_remove_uobj(&ib_uverbs_srq_idr, uobj); - - mutex_lock(&file->mutex); - list_del(&uobj->list); - mutex_unlock(&file->mutex); - - ib_uverbs_release_uevent(file, obj); - - memset(&resp, 0, sizeof resp); resp.events_reported = obj->events_reported; + uverbs_uobject_put(uobj); + if (copy_to_user((void __user *)(unsigned long)cmd.response, + &resp, sizeof(resp))) + return -EFAULT; - put_uobj(uobj); - - if (copy_to_user((void __user *) (unsigned long) cmd.response, - &resp, sizeof resp)) - ret = -EFAULT; - - return ret ? ret : in_len; + return in_len; } int ib_uverbs_ex_query_device(struct ib_uverbs_file *file, @@ -4323,6 +3839,12 @@ int ib_uverbs_ex_query_device(struct ib_uverbs_file *file, resp.max_wq_type_rq = attr.max_wq_type_rq; resp.response_length += sizeof(resp.max_wq_type_rq); + + if (ucore->outlen < resp.response_length + sizeof(resp.raw_packet_caps)) + goto end; + + resp.raw_packet_caps = attr.raw_packet_caps; + resp.response_length += sizeof(resp.raw_packet_caps); end: err = ib_copy_to_udata(ucore, &resp, resp.response_length); return err; diff --git a/drivers/infiniband/core/uverbs_main.c b/drivers/infiniband/core/uverbs_main.c index b3f95d453fba..3d2609608f58 100644 --- a/drivers/infiniband/core/uverbs_main.c +++ b/drivers/infiniband/core/uverbs_main.c @@ -51,6 +51,8 @@ #include <rdma/ib.h> #include "uverbs.h" +#include "core_priv.h" +#include "rdma_core.h" MODULE_AUTHOR("Roland Dreier"); MODULE_DESCRIPTION("InfiniBand userspace verbs access"); @@ -66,19 +68,6 @@ enum { static struct class *uverbs_class; -DEFINE_SPINLOCK(ib_uverbs_idr_lock); -DEFINE_IDR(ib_uverbs_pd_idr); -DEFINE_IDR(ib_uverbs_mr_idr); -DEFINE_IDR(ib_uverbs_mw_idr); -DEFINE_IDR(ib_uverbs_ah_idr); -DEFINE_IDR(ib_uverbs_cq_idr); -DEFINE_IDR(ib_uverbs_qp_idr); -DEFINE_IDR(ib_uverbs_srq_idr); -DEFINE_IDR(ib_uverbs_xrcd_idr); -DEFINE_IDR(ib_uverbs_rule_idr); -DEFINE_IDR(ib_uverbs_wq_idr); -DEFINE_IDR(ib_uverbs_rwq_ind_tbl_idr); - static DEFINE_SPINLOCK(map_lock); static DECLARE_BITMAP(dev_map, IB_UVERBS_MAX_DEVICES); @@ -167,37 +156,37 @@ static struct kobj_type ib_uverbs_dev_ktype = { .release = ib_uverbs_release_dev, }; -static void ib_uverbs_release_event_file(struct kref *ref) +static void ib_uverbs_release_async_event_file(struct kref *ref) { - struct ib_uverbs_event_file *file = - container_of(ref, struct ib_uverbs_event_file, ref); + struct ib_uverbs_async_event_file *file = + container_of(ref, struct ib_uverbs_async_event_file, ref); kfree(file); } void ib_uverbs_release_ucq(struct ib_uverbs_file *file, - struct ib_uverbs_event_file *ev_file, + struct ib_uverbs_completion_event_file *ev_file, struct ib_ucq_object *uobj) { struct ib_uverbs_event *evt, *tmp; if (ev_file) { - spin_lock_irq(&ev_file->lock); + spin_lock_irq(&ev_file->ev_queue.lock); list_for_each_entry_safe(evt, tmp, &uobj->comp_list, obj_list) { list_del(&evt->list); kfree(evt); } - spin_unlock_irq(&ev_file->lock); + spin_unlock_irq(&ev_file->ev_queue.lock); - kref_put(&ev_file->ref, ib_uverbs_release_event_file); + uverbs_uobject_put(&ev_file->uobj_file.uobj); } - spin_lock_irq(&file->async_file->lock); + spin_lock_irq(&file->async_file->ev_queue.lock); list_for_each_entry_safe(evt, tmp, &uobj->async_list, obj_list) { list_del(&evt->list); kfree(evt); } - spin_unlock_irq(&file->async_file->lock); + spin_unlock_irq(&file->async_file->ev_queue.lock); } void ib_uverbs_release_uevent(struct ib_uverbs_file *file, @@ -205,16 +194,16 @@ void ib_uverbs_release_uevent(struct ib_uverbs_file *file, { struct ib_uverbs_event *evt, *tmp; - spin_lock_irq(&file->async_file->lock); + spin_lock_irq(&file->async_file->ev_queue.lock); list_for_each_entry_safe(evt, tmp, &uobj->event_list, obj_list) { list_del(&evt->list); kfree(evt); } - spin_unlock_irq(&file->async_file->lock); + spin_unlock_irq(&file->async_file->ev_queue.lock); } -static void ib_uverbs_detach_umcast(struct ib_qp *qp, - struct ib_uqp_object *uobj) +void ib_uverbs_detach_umcast(struct ib_qp *qp, + struct ib_uqp_object *uobj) { struct ib_uverbs_mcast_entry *mcast, *tmp; @@ -226,124 +215,16 @@ static void ib_uverbs_detach_umcast(struct ib_qp *qp, } static int ib_uverbs_cleanup_ucontext(struct ib_uverbs_file *file, - struct ib_ucontext *context) + struct ib_ucontext *context, + bool device_removed) { - struct ib_uobject *uobj, *tmp; - context->closing = 1; - - list_for_each_entry_safe(uobj, tmp, &context->ah_list, list) { - struct ib_ah *ah = uobj->object; - - idr_remove_uobj(&ib_uverbs_ah_idr, uobj); - ib_destroy_ah(ah); - kfree(uobj); - } - - /* Remove MWs before QPs, in order to support type 2A MWs. */ - list_for_each_entry_safe(uobj, tmp, &context->mw_list, list) { - struct ib_mw *mw = uobj->object; - - idr_remove_uobj(&ib_uverbs_mw_idr, uobj); - uverbs_dealloc_mw(mw); - kfree(uobj); - } - - list_for_each_entry_safe(uobj, tmp, &context->rule_list, list) { - struct ib_flow *flow_id = uobj->object; - - idr_remove_uobj(&ib_uverbs_rule_idr, uobj); - ib_destroy_flow(flow_id); - kfree(uobj); - } - - list_for_each_entry_safe(uobj, tmp, &context->qp_list, list) { - struct ib_qp *qp = uobj->object; - struct ib_uqp_object *uqp = - container_of(uobj, struct ib_uqp_object, uevent.uobject); - - idr_remove_uobj(&ib_uverbs_qp_idr, uobj); - if (qp == qp->real_qp) - ib_uverbs_detach_umcast(qp, uqp); - ib_destroy_qp(qp); - ib_uverbs_release_uevent(file, &uqp->uevent); - kfree(uqp); - } - - list_for_each_entry_safe(uobj, tmp, &context->rwq_ind_tbl_list, list) { - struct ib_rwq_ind_table *rwq_ind_tbl = uobj->object; - struct ib_wq **ind_tbl = rwq_ind_tbl->ind_tbl; - - idr_remove_uobj(&ib_uverbs_rwq_ind_tbl_idr, uobj); - ib_destroy_rwq_ind_table(rwq_ind_tbl); - kfree(ind_tbl); - kfree(uobj); - } - - list_for_each_entry_safe(uobj, tmp, &context->wq_list, list) { - struct ib_wq *wq = uobj->object; - struct ib_uwq_object *uwq = - container_of(uobj, struct ib_uwq_object, uevent.uobject); - - idr_remove_uobj(&ib_uverbs_wq_idr, uobj); - ib_destroy_wq(wq); - ib_uverbs_release_uevent(file, &uwq->uevent); - kfree(uwq); - } - - list_for_each_entry_safe(uobj, tmp, &context->srq_list, list) { - struct ib_srq *srq = uobj->object; - struct ib_uevent_object *uevent = - container_of(uobj, struct ib_uevent_object, uobject); - - idr_remove_uobj(&ib_uverbs_srq_idr, uobj); - ib_destroy_srq(srq); - ib_uverbs_release_uevent(file, uevent); - kfree(uevent); - } - - list_for_each_entry_safe(uobj, tmp, &context->cq_list, list) { - struct ib_cq *cq = uobj->object; - struct ib_uverbs_event_file *ev_file = cq->cq_context; - struct ib_ucq_object *ucq = - container_of(uobj, struct ib_ucq_object, uobject); - - idr_remove_uobj(&ib_uverbs_cq_idr, uobj); - ib_destroy_cq(cq); - ib_uverbs_release_ucq(file, ev_file, ucq); - kfree(ucq); - } - - list_for_each_entry_safe(uobj, tmp, &context->mr_list, list) { - struct ib_mr *mr = uobj->object; - - idr_remove_uobj(&ib_uverbs_mr_idr, uobj); - ib_dereg_mr(mr); - kfree(uobj); - } - - mutex_lock(&file->device->xrcd_tree_mutex); - list_for_each_entry_safe(uobj, tmp, &context->xrcd_list, list) { - struct ib_xrcd *xrcd = uobj->object; - struct ib_uxrcd_object *uxrcd = - container_of(uobj, struct ib_uxrcd_object, uobject); - - idr_remove_uobj(&ib_uverbs_xrcd_idr, uobj); - ib_uverbs_dealloc_xrcd(file->device, xrcd); - kfree(uxrcd); - } - mutex_unlock(&file->device->xrcd_tree_mutex); - - list_for_each_entry_safe(uobj, tmp, &context->pd_list, list) { - struct ib_pd *pd = uobj->object; - - idr_remove_uobj(&ib_uverbs_pd_idr, uobj); - ib_dealloc_pd(pd); - kfree(uobj); - } - + uverbs_cleanup_ucontext(context, device_removed); put_pid(context->tgid); + ib_rdmacg_uncharge(&context->cg_obj, context->device, + RDMACG_RESOURCE_HCA_HANDLE); + return context->device->dealloc_ucontext(context); } @@ -352,7 +233,7 @@ static void ib_uverbs_comp_dev(struct ib_uverbs_device *dev) complete(&dev->comp); } -static void ib_uverbs_release_file(struct kref *ref) +void ib_uverbs_release_file(struct kref *ref) { struct ib_uverbs_file *file = container_of(ref, struct ib_uverbs_file, ref); @@ -372,58 +253,54 @@ static void ib_uverbs_release_file(struct kref *ref) kfree(file); } -static ssize_t ib_uverbs_event_read(struct file *filp, char __user *buf, - size_t count, loff_t *pos) +static ssize_t ib_uverbs_event_read(struct ib_uverbs_event_queue *ev_queue, + struct ib_uverbs_file *uverbs_file, + struct file *filp, char __user *buf, + size_t count, loff_t *pos, + size_t eventsz) { - struct ib_uverbs_event_file *file = filp->private_data; struct ib_uverbs_event *event; - int eventsz; int ret = 0; - spin_lock_irq(&file->lock); + spin_lock_irq(&ev_queue->lock); - while (list_empty(&file->event_list)) { - spin_unlock_irq(&file->lock); + while (list_empty(&ev_queue->event_list)) { + spin_unlock_irq(&ev_queue->lock); if (filp->f_flags & O_NONBLOCK) return -EAGAIN; - if (wait_event_interruptible(file->poll_wait, - (!list_empty(&file->event_list) || + if (wait_event_interruptible(ev_queue->poll_wait, + (!list_empty(&ev_queue->event_list) || /* The barriers built into wait_event_interruptible() * and wake_up() guarentee this will see the null set * without using RCU */ - !file->uverbs_file->device->ib_dev))) + !uverbs_file->device->ib_dev))) return -ERESTARTSYS; /* If device was disassociated and no event exists set an error */ - if (list_empty(&file->event_list) && - !file->uverbs_file->device->ib_dev) + if (list_empty(&ev_queue->event_list) && + !uverbs_file->device->ib_dev) return -EIO; - spin_lock_irq(&file->lock); + spin_lock_irq(&ev_queue->lock); } - event = list_entry(file->event_list.next, struct ib_uverbs_event, list); - - if (file->is_async) - eventsz = sizeof (struct ib_uverbs_async_event_desc); - else - eventsz = sizeof (struct ib_uverbs_comp_event_desc); + event = list_entry(ev_queue->event_list.next, struct ib_uverbs_event, list); if (eventsz > count) { ret = -EINVAL; event = NULL; } else { - list_del(file->event_list.next); + list_del(ev_queue->event_list.next); if (event->counter) { ++(*event->counter); list_del(&event->obj_list); } } - spin_unlock_irq(&file->lock); + spin_unlock_irq(&ev_queue->lock); if (event) { if (copy_to_user(buf, event, eventsz)) @@ -437,87 +314,158 @@ static ssize_t ib_uverbs_event_read(struct file *filp, char __user *buf, return ret; } -static unsigned int ib_uverbs_event_poll(struct file *filp, +static ssize_t ib_uverbs_async_event_read(struct file *filp, char __user *buf, + size_t count, loff_t *pos) +{ + struct ib_uverbs_async_event_file *file = filp->private_data; + + return ib_uverbs_event_read(&file->ev_queue, file->uverbs_file, filp, + buf, count, pos, + sizeof(struct ib_uverbs_async_event_desc)); +} + +static ssize_t ib_uverbs_comp_event_read(struct file *filp, char __user *buf, + size_t count, loff_t *pos) +{ + struct ib_uverbs_completion_event_file *comp_ev_file = + filp->private_data; + + return ib_uverbs_event_read(&comp_ev_file->ev_queue, + comp_ev_file->uobj_file.ufile, filp, + buf, count, pos, + sizeof(struct ib_uverbs_comp_event_desc)); +} + +static unsigned int ib_uverbs_event_poll(struct ib_uverbs_event_queue *ev_queue, + struct file *filp, struct poll_table_struct *wait) { unsigned int pollflags = 0; - struct ib_uverbs_event_file *file = filp->private_data; - poll_wait(filp, &file->poll_wait, wait); + poll_wait(filp, &ev_queue->poll_wait, wait); - spin_lock_irq(&file->lock); - if (!list_empty(&file->event_list)) + spin_lock_irq(&ev_queue->lock); + if (!list_empty(&ev_queue->event_list)) pollflags = POLLIN | POLLRDNORM; - spin_unlock_irq(&file->lock); + spin_unlock_irq(&ev_queue->lock); return pollflags; } -static int ib_uverbs_event_fasync(int fd, struct file *filp, int on) +static unsigned int ib_uverbs_async_event_poll(struct file *filp, + struct poll_table_struct *wait) +{ + return ib_uverbs_event_poll(filp->private_data, filp, wait); +} + +static unsigned int ib_uverbs_comp_event_poll(struct file *filp, + struct poll_table_struct *wait) { - struct ib_uverbs_event_file *file = filp->private_data; + struct ib_uverbs_completion_event_file *comp_ev_file = + filp->private_data; - return fasync_helper(fd, filp, on, &file->async_queue); + return ib_uverbs_event_poll(&comp_ev_file->ev_queue, filp, wait); } -static int ib_uverbs_event_close(struct inode *inode, struct file *filp) +static int ib_uverbs_async_event_fasync(int fd, struct file *filp, int on) { - struct ib_uverbs_event_file *file = filp->private_data; + struct ib_uverbs_event_queue *ev_queue = filp->private_data; + + return fasync_helper(fd, filp, on, &ev_queue->async_queue); +} + +static int ib_uverbs_comp_event_fasync(int fd, struct file *filp, int on) +{ + struct ib_uverbs_completion_event_file *comp_ev_file = + filp->private_data; + + return fasync_helper(fd, filp, on, &comp_ev_file->ev_queue.async_queue); +} + +static int ib_uverbs_async_event_close(struct inode *inode, struct file *filp) +{ + struct ib_uverbs_async_event_file *file = filp->private_data; + struct ib_uverbs_file *uverbs_file = file->uverbs_file; struct ib_uverbs_event *entry, *tmp; int closed_already = 0; - mutex_lock(&file->uverbs_file->device->lists_mutex); - spin_lock_irq(&file->lock); - closed_already = file->is_closed; - file->is_closed = 1; - list_for_each_entry_safe(entry, tmp, &file->event_list, list) { + mutex_lock(&uverbs_file->device->lists_mutex); + spin_lock_irq(&file->ev_queue.lock); + closed_already = file->ev_queue.is_closed; + file->ev_queue.is_closed = 1; + list_for_each_entry_safe(entry, tmp, &file->ev_queue.event_list, list) { if (entry->counter) list_del(&entry->obj_list); kfree(entry); } - spin_unlock_irq(&file->lock); + spin_unlock_irq(&file->ev_queue.lock); if (!closed_already) { list_del(&file->list); - if (file->is_async) - ib_unregister_event_handler(&file->uverbs_file-> - event_handler); + ib_unregister_event_handler(&uverbs_file->event_handler); } - mutex_unlock(&file->uverbs_file->device->lists_mutex); + mutex_unlock(&uverbs_file->device->lists_mutex); - kref_put(&file->uverbs_file->ref, ib_uverbs_release_file); - kref_put(&file->ref, ib_uverbs_release_event_file); + kref_put(&uverbs_file->ref, ib_uverbs_release_file); + kref_put(&file->ref, ib_uverbs_release_async_event_file); return 0; } -static const struct file_operations uverbs_event_fops = { +static int ib_uverbs_comp_event_close(struct inode *inode, struct file *filp) +{ + struct ib_uverbs_completion_event_file *file = filp->private_data; + struct ib_uverbs_event *entry, *tmp; + + spin_lock_irq(&file->ev_queue.lock); + list_for_each_entry_safe(entry, tmp, &file->ev_queue.event_list, list) { + if (entry->counter) + list_del(&entry->obj_list); + kfree(entry); + } + spin_unlock_irq(&file->ev_queue.lock); + + uverbs_close_fd(filp); + + return 0; +} + +const struct file_operations uverbs_event_fops = { .owner = THIS_MODULE, - .read = ib_uverbs_event_read, - .poll = ib_uverbs_event_poll, - .release = ib_uverbs_event_close, - .fasync = ib_uverbs_event_fasync, + .read = ib_uverbs_comp_event_read, + .poll = ib_uverbs_comp_event_poll, + .release = ib_uverbs_comp_event_close, + .fasync = ib_uverbs_comp_event_fasync, + .llseek = no_llseek, +}; + +static const struct file_operations uverbs_async_event_fops = { + .owner = THIS_MODULE, + .read = ib_uverbs_async_event_read, + .poll = ib_uverbs_async_event_poll, + .release = ib_uverbs_async_event_close, + .fasync = ib_uverbs_async_event_fasync, .llseek = no_llseek, }; void ib_uverbs_comp_handler(struct ib_cq *cq, void *cq_context) { - struct ib_uverbs_event_file *file = cq_context; + struct ib_uverbs_event_queue *ev_queue = cq_context; struct ib_ucq_object *uobj; struct ib_uverbs_event *entry; unsigned long flags; - if (!file) + if (!ev_queue) return; - spin_lock_irqsave(&file->lock, flags); - if (file->is_closed) { - spin_unlock_irqrestore(&file->lock, flags); + spin_lock_irqsave(&ev_queue->lock, flags); + if (ev_queue->is_closed) { + spin_unlock_irqrestore(&ev_queue->lock, flags); return; } entry = kmalloc(sizeof *entry, GFP_ATOMIC); if (!entry) { - spin_unlock_irqrestore(&file->lock, flags); + spin_unlock_irqrestore(&ev_queue->lock, flags); return; } @@ -526,12 +474,12 @@ void ib_uverbs_comp_handler(struct ib_cq *cq, void *cq_context) entry->desc.comp.cq_handle = cq->uobject->user_handle; entry->counter = &uobj->comp_events_reported; - list_add_tail(&entry->list, &file->event_list); + list_add_tail(&entry->list, &ev_queue->event_list); list_add_tail(&entry->obj_list, &uobj->comp_list); - spin_unlock_irqrestore(&file->lock, flags); + spin_unlock_irqrestore(&ev_queue->lock, flags); - wake_up_interruptible(&file->poll_wait); - kill_fasync(&file->async_queue, SIGIO, POLL_IN); + wake_up_interruptible(&ev_queue->poll_wait); + kill_fasync(&ev_queue->async_queue, SIGIO, POLL_IN); } static void ib_uverbs_async_handler(struct ib_uverbs_file *file, @@ -542,15 +490,15 @@ static void ib_uverbs_async_handler(struct ib_uverbs_file *file, struct ib_uverbs_event *entry; unsigned long flags; - spin_lock_irqsave(&file->async_file->lock, flags); - if (file->async_file->is_closed) { - spin_unlock_irqrestore(&file->async_file->lock, flags); + spin_lock_irqsave(&file->async_file->ev_queue.lock, flags); + if (file->async_file->ev_queue.is_closed) { + spin_unlock_irqrestore(&file->async_file->ev_queue.lock, flags); return; } entry = kmalloc(sizeof *entry, GFP_ATOMIC); if (!entry) { - spin_unlock_irqrestore(&file->async_file->lock, flags); + spin_unlock_irqrestore(&file->async_file->ev_queue.lock, flags); return; } @@ -559,13 +507,13 @@ static void ib_uverbs_async_handler(struct ib_uverbs_file *file, entry->desc.async.reserved = 0; entry->counter = counter; - list_add_tail(&entry->list, &file->async_file->event_list); + list_add_tail(&entry->list, &file->async_file->ev_queue.event_list); if (obj_list) list_add_tail(&entry->obj_list, obj_list); - spin_unlock_irqrestore(&file->async_file->lock, flags); + spin_unlock_irqrestore(&file->async_file->ev_queue.lock, flags); - wake_up_interruptible(&file->async_file->poll_wait); - kill_fasync(&file->async_file->async_queue, SIGIO, POLL_IN); + wake_up_interruptible(&file->async_file->ev_queue.poll_wait); + kill_fasync(&file->async_file->ev_queue.async_queue, SIGIO, POLL_IN); } void ib_uverbs_cq_event_handler(struct ib_event *event, void *context_ptr) @@ -583,7 +531,7 @@ void ib_uverbs_qp_event_handler(struct ib_event *event, void *context_ptr) struct ib_uevent_object *uobj; /* for XRC target qp's, check that qp is live */ - if (!event->element.qp->uobject || !event->element.qp->uobject->live) + if (!event->element.qp->uobject) return; uobj = container_of(event->element.qp->uobject, @@ -628,15 +576,23 @@ void ib_uverbs_event_handler(struct ib_event_handler *handler, void ib_uverbs_free_async_event_file(struct ib_uverbs_file *file) { - kref_put(&file->async_file->ref, ib_uverbs_release_event_file); + kref_put(&file->async_file->ref, ib_uverbs_release_async_event_file); file->async_file = NULL; } -struct file *ib_uverbs_alloc_event_file(struct ib_uverbs_file *uverbs_file, - struct ib_device *ib_dev, - int is_async) +void ib_uverbs_init_event_queue(struct ib_uverbs_event_queue *ev_queue) +{ + spin_lock_init(&ev_queue->lock); + INIT_LIST_HEAD(&ev_queue->event_list); + init_waitqueue_head(&ev_queue->poll_wait); + ev_queue->is_closed = 0; + ev_queue->async_queue = NULL; +} + +struct file *ib_uverbs_alloc_async_event_file(struct ib_uverbs_file *uverbs_file, + struct ib_device *ib_dev) { - struct ib_uverbs_event_file *ev_file; + struct ib_uverbs_async_event_file *ev_file; struct file *filp; int ret; @@ -644,16 +600,11 @@ struct file *ib_uverbs_alloc_event_file(struct ib_uverbs_file *uverbs_file, if (!ev_file) return ERR_PTR(-ENOMEM); - kref_init(&ev_file->ref); - spin_lock_init(&ev_file->lock); - INIT_LIST_HEAD(&ev_file->event_list); - init_waitqueue_head(&ev_file->poll_wait); + ib_uverbs_init_event_queue(&ev_file->ev_queue); ev_file->uverbs_file = uverbs_file; kref_get(&ev_file->uverbs_file->ref); - ev_file->async_queue = NULL; - ev_file->is_closed = 0; - - filp = anon_inode_getfile("[infinibandevent]", &uverbs_event_fops, + kref_init(&ev_file->ref); + filp = anon_inode_getfile("[infinibandevent]", &uverbs_async_event_fops, ev_file, O_RDONLY); if (IS_ERR(filp)) goto err_put_refs; @@ -663,64 +614,33 @@ struct file *ib_uverbs_alloc_event_file(struct ib_uverbs_file *uverbs_file, &uverbs_file->device->uverbs_events_file_list); mutex_unlock(&uverbs_file->device->lists_mutex); - if (is_async) { - WARN_ON(uverbs_file->async_file); - uverbs_file->async_file = ev_file; - kref_get(&uverbs_file->async_file->ref); - INIT_IB_EVENT_HANDLER(&uverbs_file->event_handler, - ib_dev, - ib_uverbs_event_handler); - ret = ib_register_event_handler(&uverbs_file->event_handler); - if (ret) - goto err_put_file; - - /* At that point async file stuff was fully set */ - ev_file->is_async = 1; - } + WARN_ON(uverbs_file->async_file); + uverbs_file->async_file = ev_file; + kref_get(&uverbs_file->async_file->ref); + INIT_IB_EVENT_HANDLER(&uverbs_file->event_handler, + ib_dev, + ib_uverbs_event_handler); + ret = ib_register_event_handler(&uverbs_file->event_handler); + if (ret) + goto err_put_file; + + /* At that point async file stuff was fully set */ return filp; err_put_file: fput(filp); - kref_put(&uverbs_file->async_file->ref, ib_uverbs_release_event_file); + kref_put(&uverbs_file->async_file->ref, + ib_uverbs_release_async_event_file); uverbs_file->async_file = NULL; return ERR_PTR(ret); err_put_refs: kref_put(&ev_file->uverbs_file->ref, ib_uverbs_release_file); - kref_put(&ev_file->ref, ib_uverbs_release_event_file); + kref_put(&ev_file->ref, ib_uverbs_release_async_event_file); return filp; } -/* - * Look up a completion event file by FD. If lookup is successful, - * takes a ref to the event file struct that it returns; if - * unsuccessful, returns NULL. - */ -struct ib_uverbs_event_file *ib_uverbs_lookup_comp_file(int fd) -{ - struct ib_uverbs_event_file *ev_file = NULL; - struct fd f = fdget(fd); - - if (!f.file) - return NULL; - - if (f.file->f_op != &uverbs_event_fops) - goto out; - - ev_file = f.file->private_data; - if (ev_file->is_async) { - ev_file = NULL; - goto out; - } - - kref_get(&ev_file->ref); - -out: - fdput(f); - return ev_file; -} - static int verify_command_mask(struct ib_device *ib_dev, __u32 command) { u64 mask; @@ -966,6 +886,8 @@ static int ib_uverbs_open(struct inode *inode, struct file *filp) } file->device = dev; + spin_lock_init(&file->idr_lock); + idr_init(&file->idr); file->ucontext = NULL; file->async_file = NULL; kref_init(&file->ref); @@ -999,10 +921,11 @@ static int ib_uverbs_close(struct inode *inode, struct file *filp) mutex_lock(&file->cleanup_mutex); if (file->ucontext) { - ib_uverbs_cleanup_ucontext(file, file->ucontext); + ib_uverbs_cleanup_ucontext(file, file->ucontext, false); file->ucontext = NULL; } mutex_unlock(&file->cleanup_mutex); + idr_destroy(&file->idr); mutex_lock(&file->device->lists_mutex); if (!file->is_closed) { @@ -1012,7 +935,8 @@ static int ib_uverbs_close(struct inode *inode, struct file *filp) mutex_unlock(&file->device->lists_mutex); if (file->async_file) - kref_put(&file->async_file->ref, ib_uverbs_release_event_file); + kref_put(&file->async_file->ref, + ib_uverbs_release_async_event_file); kref_put(&file->ref, ib_uverbs_release_file); kobject_put(&dev->kobj); @@ -1169,12 +1093,12 @@ static void ib_uverbs_add_one(struct ib_device *device) cdev_init(&uverbs_dev->cdev, NULL); uverbs_dev->cdev.owner = THIS_MODULE; uverbs_dev->cdev.ops = device->mmap ? &uverbs_mmap_fops : &uverbs_fops; - uverbs_dev->cdev.kobj.parent = &uverbs_dev->kobj; + cdev_set_parent(&uverbs_dev->cdev, &uverbs_dev->kobj); kobject_set_name(&uverbs_dev->cdev.kobj, "uverbs%d", uverbs_dev->devnum); if (cdev_add(&uverbs_dev->cdev, base, 1)) goto err_cdev; - uverbs_dev->dev = device_create(uverbs_class, device->dma_device, + uverbs_dev->dev = device_create(uverbs_class, device->dev.parent, uverbs_dev->cdev.dev, uverbs_dev, "uverbs%d", uverbs_dev->devnum); if (IS_ERR(uverbs_dev->dev)) @@ -1211,7 +1135,7 @@ static void ib_uverbs_free_hw_resources(struct ib_uverbs_device *uverbs_dev, struct ib_device *ib_dev) { struct ib_uverbs_file *file; - struct ib_uverbs_event_file *event_file; + struct ib_uverbs_async_event_file *event_file; struct ib_event event; /* Pending running commands to terminate */ @@ -1248,7 +1172,9 @@ static void ib_uverbs_free_hw_resources(struct ib_uverbs_device *uverbs_dev, * (e.g mmput). */ ib_dev->disassociate_ucontext(ucontext); - ib_uverbs_cleanup_ucontext(file, ucontext); + mutex_lock(&file->cleanup_mutex); + ib_uverbs_cleanup_ucontext(file, ucontext, true); + mutex_unlock(&file->cleanup_mutex); } mutex_lock(&uverbs_dev->lists_mutex); @@ -1258,21 +1184,20 @@ static void ib_uverbs_free_hw_resources(struct ib_uverbs_device *uverbs_dev, while (!list_empty(&uverbs_dev->uverbs_events_file_list)) { event_file = list_first_entry(&uverbs_dev-> uverbs_events_file_list, - struct ib_uverbs_event_file, + struct ib_uverbs_async_event_file, list); - spin_lock_irq(&event_file->lock); - event_file->is_closed = 1; - spin_unlock_irq(&event_file->lock); + spin_lock_irq(&event_file->ev_queue.lock); + event_file->ev_queue.is_closed = 1; + spin_unlock_irq(&event_file->ev_queue.lock); list_del(&event_file->list); - if (event_file->is_async) { - ib_unregister_event_handler(&event_file->uverbs_file-> - event_handler); - event_file->uverbs_file->event_handler.device = NULL; - } + ib_unregister_event_handler( + &event_file->uverbs_file->event_handler); + event_file->uverbs_file->event_handler.device = + NULL; - wake_up_interruptible(&event_file->poll_wait); - kill_fasync(&event_file->async_queue, SIGIO, POLL_IN); + wake_up_interruptible(&event_file->ev_queue.poll_wait); + kill_fasync(&event_file->ev_queue.async_queue, SIGIO, POLL_IN); } mutex_unlock(&uverbs_dev->lists_mutex); } @@ -1376,13 +1301,6 @@ static void __exit ib_uverbs_cleanup(void) unregister_chrdev_region(IB_UVERBS_BASE_DEV, IB_UVERBS_MAX_DEVICES); if (overflow_maj) unregister_chrdev_region(overflow_maj, IB_UVERBS_MAX_DEVICES); - idr_destroy(&ib_uverbs_pd_idr); - idr_destroy(&ib_uverbs_mr_idr); - idr_destroy(&ib_uverbs_mw_idr); - idr_destroy(&ib_uverbs_ah_idr); - idr_destroy(&ib_uverbs_cq_idr); - idr_destroy(&ib_uverbs_qp_idr); - idr_destroy(&ib_uverbs_srq_idr); } module_init(ib_uverbs_init); diff --git a/drivers/infiniband/core/uverbs_marshall.c b/drivers/infiniband/core/uverbs_marshall.c index af020f80d50f..94fd989c9060 100644 --- a/drivers/infiniband/core/uverbs_marshall.c +++ b/drivers/infiniband/core/uverbs_marshall.c @@ -34,20 +34,25 @@ #include <rdma/ib_marshall.h> void ib_copy_ah_attr_to_user(struct ib_uverbs_ah_attr *dst, - struct ib_ah_attr *src) + struct rdma_ah_attr *src) { - memcpy(dst->grh.dgid, src->grh.dgid.raw, sizeof src->grh.dgid); - dst->grh.flow_label = src->grh.flow_label; - dst->grh.sgid_index = src->grh.sgid_index; - dst->grh.hop_limit = src->grh.hop_limit; - dst->grh.traffic_class = src->grh.traffic_class; memset(&dst->grh.reserved, 0, sizeof(dst->grh.reserved)); - dst->dlid = src->dlid; - dst->sl = src->sl; - dst->src_path_bits = src->src_path_bits; - dst->static_rate = src->static_rate; - dst->is_global = src->ah_flags & IB_AH_GRH ? 1 : 0; - dst->port_num = src->port_num; + dst->dlid = rdma_ah_get_dlid(src); + dst->sl = rdma_ah_get_sl(src); + dst->src_path_bits = rdma_ah_get_path_bits(src); + dst->static_rate = rdma_ah_get_static_rate(src); + dst->is_global = rdma_ah_get_ah_flags(src) & + IB_AH_GRH ? 1 : 0; + if (dst->is_global) { + const struct ib_global_route *grh = rdma_ah_read_grh(src); + + memcpy(dst->grh.dgid, grh->dgid.raw, sizeof(grh->dgid)); + dst->grh.flow_label = grh->flow_label; + dst->grh.sgid_index = grh->sgid_index; + dst->grh.hop_limit = grh->hop_limit; + dst->grh.traffic_class = grh->traffic_class; + } + dst->port_num = rdma_ah_get_port_num(src); dst->reserved = 0; } EXPORT_SYMBOL(ib_copy_ah_attr_to_user); @@ -91,15 +96,15 @@ void ib_copy_qp_attr_to_user(struct ib_uverbs_qp_attr *dst, } EXPORT_SYMBOL(ib_copy_qp_attr_to_user); -void ib_copy_path_rec_to_user(struct ib_user_path_rec *dst, - struct ib_sa_path_rec *src) +static void __ib_copy_path_rec_to_user(struct ib_user_path_rec *dst, + struct sa_path_rec *src) { - memcpy(dst->dgid, src->dgid.raw, sizeof src->dgid); - memcpy(dst->sgid, src->sgid.raw, sizeof src->sgid); + memcpy(dst->dgid, src->dgid.raw, sizeof(src->dgid)); + memcpy(dst->sgid, src->sgid.raw, sizeof(src->sgid)); - dst->dlid = src->dlid; - dst->slid = src->slid; - dst->raw_traffic = src->raw_traffic; + dst->dlid = htons(ntohl(sa_path_get_dlid(src))); + dst->slid = htons(ntohl(sa_path_get_slid(src))); + dst->raw_traffic = sa_path_get_raw_traffic(src); dst->flow_label = src->flow_label; dst->hop_limit = src->hop_limit; dst->traffic_class = src->traffic_class; @@ -115,17 +120,43 @@ void ib_copy_path_rec_to_user(struct ib_user_path_rec *dst, dst->preference = src->preference; dst->packet_life_time_selector = src->packet_life_time_selector; } + +void ib_copy_path_rec_to_user(struct ib_user_path_rec *dst, + struct sa_path_rec *src) +{ + struct sa_path_rec rec; + + if (src->rec_type == SA_PATH_REC_TYPE_OPA) { + sa_convert_path_opa_to_ib(&rec, src); + __ib_copy_path_rec_to_user(dst, &rec); + return; + } + __ib_copy_path_rec_to_user(dst, src); +} EXPORT_SYMBOL(ib_copy_path_rec_to_user); -void ib_copy_path_rec_from_user(struct ib_sa_path_rec *dst, +void ib_copy_path_rec_from_user(struct sa_path_rec *dst, struct ib_user_path_rec *src) { + __be32 slid, dlid; + + memset(dst, 0, sizeof(*dst)); + if ((ib_is_opa_gid((union ib_gid *)src->sgid)) || + (ib_is_opa_gid((union ib_gid *)src->dgid))) { + dst->rec_type = SA_PATH_REC_TYPE_OPA; + slid = htonl(opa_get_lid_from_gid((union ib_gid *)src->sgid)); + dlid = htonl(opa_get_lid_from_gid((union ib_gid *)src->dgid)); + } else { + dst->rec_type = SA_PATH_REC_TYPE_IB; + slid = htonl(ntohs(src->slid)); + dlid = htonl(ntohs(src->dlid)); + } memcpy(dst->dgid.raw, src->dgid, sizeof dst->dgid); memcpy(dst->sgid.raw, src->sgid, sizeof dst->sgid); - dst->dlid = src->dlid; - dst->slid = src->slid; - dst->raw_traffic = src->raw_traffic; + sa_path_set_dlid(dst, dlid); + sa_path_set_slid(dst, slid); + sa_path_set_raw_traffic(dst, src->raw_traffic); dst->flow_label = src->flow_label; dst->hop_limit = src->hop_limit; dst->traffic_class = src->traffic_class; @@ -141,9 +172,9 @@ void ib_copy_path_rec_from_user(struct ib_sa_path_rec *dst, dst->preference = src->preference; dst->packet_life_time_selector = src->packet_life_time_selector; - memset(dst->dmac, 0, sizeof(dst->dmac)); - dst->net = NULL; - dst->ifindex = 0; - dst->gid_type = IB_GID_TYPE_IB; + /* TODO: No need to set this */ + sa_path_set_dmac_zero(dst); + sa_path_set_ndev(dst, NULL); + sa_path_set_ifindex(dst, 0); } EXPORT_SYMBOL(ib_copy_path_rec_from_user); diff --git a/drivers/infiniband/core/uverbs_std_types.c b/drivers/infiniband/core/uverbs_std_types.c new file mode 100644 index 000000000000..ef293379f37a --- /dev/null +++ b/drivers/infiniband/core/uverbs_std_types.c @@ -0,0 +1,275 @@ +/* + * Copyright (c) 2017, Mellanox Technologies inc. All rights reserved. + * + * 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 <rdma/uverbs_std_types.h> +#include <rdma/ib_user_verbs.h> +#include <rdma/ib_verbs.h> +#include <linux/bug.h> +#include <linux/file.h> +#include "rdma_core.h" +#include "uverbs.h" + +static int uverbs_free_ah(struct ib_uobject *uobject, + enum rdma_remove_reason why) +{ + return rdma_destroy_ah((struct ib_ah *)uobject->object); +} + +static int uverbs_free_flow(struct ib_uobject *uobject, + enum rdma_remove_reason why) +{ + return ib_destroy_flow((struct ib_flow *)uobject->object); +} + +static int uverbs_free_mw(struct ib_uobject *uobject, + enum rdma_remove_reason why) +{ + return uverbs_dealloc_mw((struct ib_mw *)uobject->object); +} + +static int uverbs_free_qp(struct ib_uobject *uobject, + enum rdma_remove_reason why) +{ + struct ib_qp *qp = uobject->object; + struct ib_uqp_object *uqp = + container_of(uobject, struct ib_uqp_object, uevent.uobject); + int ret; + + if (why == RDMA_REMOVE_DESTROY) { + if (!list_empty(&uqp->mcast_list)) + return -EBUSY; + } else if (qp == qp->real_qp) { + ib_uverbs_detach_umcast(qp, uqp); + } + + ret = ib_destroy_qp(qp); + if (ret && why == RDMA_REMOVE_DESTROY) + return ret; + + if (uqp->uxrcd) + atomic_dec(&uqp->uxrcd->refcnt); + + ib_uverbs_release_uevent(uobject->context->ufile, &uqp->uevent); + return ret; +} + +static int uverbs_free_rwq_ind_tbl(struct ib_uobject *uobject, + enum rdma_remove_reason why) +{ + struct ib_rwq_ind_table *rwq_ind_tbl = uobject->object; + struct ib_wq **ind_tbl = rwq_ind_tbl->ind_tbl; + int ret; + + ret = ib_destroy_rwq_ind_table(rwq_ind_tbl); + if (!ret || why != RDMA_REMOVE_DESTROY) + kfree(ind_tbl); + return ret; +} + +static int uverbs_free_wq(struct ib_uobject *uobject, + enum rdma_remove_reason why) +{ + struct ib_wq *wq = uobject->object; + struct ib_uwq_object *uwq = + container_of(uobject, struct ib_uwq_object, uevent.uobject); + int ret; + + ret = ib_destroy_wq(wq); + if (!ret || why != RDMA_REMOVE_DESTROY) + ib_uverbs_release_uevent(uobject->context->ufile, &uwq->uevent); + return ret; +} + +static int uverbs_free_srq(struct ib_uobject *uobject, + enum rdma_remove_reason why) +{ + struct ib_srq *srq = uobject->object; + struct ib_uevent_object *uevent = + container_of(uobject, struct ib_uevent_object, uobject); + enum ib_srq_type srq_type = srq->srq_type; + int ret; + + ret = ib_destroy_srq(srq); + + if (ret && why == RDMA_REMOVE_DESTROY) + return ret; + + if (srq_type == IB_SRQT_XRC) { + struct ib_usrq_object *us = + container_of(uevent, struct ib_usrq_object, uevent); + + atomic_dec(&us->uxrcd->refcnt); + } + + ib_uverbs_release_uevent(uobject->context->ufile, uevent); + return ret; +} + +static int uverbs_free_cq(struct ib_uobject *uobject, + enum rdma_remove_reason why) +{ + struct ib_cq *cq = uobject->object; + struct ib_uverbs_event_queue *ev_queue = cq->cq_context; + struct ib_ucq_object *ucq = + container_of(uobject, struct ib_ucq_object, uobject); + int ret; + + ret = ib_destroy_cq(cq); + if (!ret || why != RDMA_REMOVE_DESTROY) + ib_uverbs_release_ucq(uobject->context->ufile, ev_queue ? + container_of(ev_queue, + struct ib_uverbs_completion_event_file, + ev_queue) : NULL, + ucq); + return ret; +} + +static int uverbs_free_mr(struct ib_uobject *uobject, + enum rdma_remove_reason why) +{ + return ib_dereg_mr((struct ib_mr *)uobject->object); +} + +static int uverbs_free_xrcd(struct ib_uobject *uobject, + enum rdma_remove_reason why) +{ + struct ib_xrcd *xrcd = uobject->object; + struct ib_uxrcd_object *uxrcd = + container_of(uobject, struct ib_uxrcd_object, uobject); + int ret; + + mutex_lock(&uobject->context->ufile->device->xrcd_tree_mutex); + if (why == RDMA_REMOVE_DESTROY && atomic_read(&uxrcd->refcnt)) + ret = -EBUSY; + else + ret = ib_uverbs_dealloc_xrcd(uobject->context->ufile->device, + xrcd, why); + mutex_unlock(&uobject->context->ufile->device->xrcd_tree_mutex); + + return ret; +} + +static int uverbs_free_pd(struct ib_uobject *uobject, + enum rdma_remove_reason why) +{ + struct ib_pd *pd = uobject->object; + + if (why == RDMA_REMOVE_DESTROY && atomic_read(&pd->usecnt)) + return -EBUSY; + + ib_dealloc_pd((struct ib_pd *)uobject->object); + return 0; +} + +static int uverbs_hot_unplug_completion_event_file(struct ib_uobject_file *uobj_file, + enum rdma_remove_reason why) +{ + struct ib_uverbs_completion_event_file *comp_event_file = + container_of(uobj_file, struct ib_uverbs_completion_event_file, + uobj_file); + struct ib_uverbs_event_queue *event_queue = &comp_event_file->ev_queue; + + spin_lock_irq(&event_queue->lock); + event_queue->is_closed = 1; + spin_unlock_irq(&event_queue->lock); + + if (why == RDMA_REMOVE_DRIVER_REMOVE) { + wake_up_interruptible(&event_queue->poll_wait); + kill_fasync(&event_queue->async_queue, SIGIO, POLL_IN); + } + return 0; +}; + +const struct uverbs_obj_fd_type uverbs_type_attrs_comp_channel = { + .type = UVERBS_TYPE_ALLOC_FD(sizeof(struct ib_uverbs_completion_event_file), 0), + .context_closed = uverbs_hot_unplug_completion_event_file, + .fops = &uverbs_event_fops, + .name = "[infinibandevent]", + .flags = O_RDONLY, +}; + +const struct uverbs_obj_idr_type uverbs_type_attrs_cq = { + .type = UVERBS_TYPE_ALLOC_IDR_SZ(sizeof(struct ib_ucq_object), 0), + .destroy_object = uverbs_free_cq, +}; + +const struct uverbs_obj_idr_type uverbs_type_attrs_qp = { + .type = UVERBS_TYPE_ALLOC_IDR_SZ(sizeof(struct ib_uqp_object), 0), + .destroy_object = uverbs_free_qp, +}; + +const struct uverbs_obj_idr_type uverbs_type_attrs_mw = { + .type = UVERBS_TYPE_ALLOC_IDR(0), + .destroy_object = uverbs_free_mw, +}; + +const struct uverbs_obj_idr_type uverbs_type_attrs_mr = { + /* 1 is used in order to free the MR after all the MWs */ + .type = UVERBS_TYPE_ALLOC_IDR(1), + .destroy_object = uverbs_free_mr, +}; + +const struct uverbs_obj_idr_type uverbs_type_attrs_srq = { + .type = UVERBS_TYPE_ALLOC_IDR_SZ(sizeof(struct ib_usrq_object), 0), + .destroy_object = uverbs_free_srq, +}; + +const struct uverbs_obj_idr_type uverbs_type_attrs_ah = { + .type = UVERBS_TYPE_ALLOC_IDR(0), + .destroy_object = uverbs_free_ah, +}; + +const struct uverbs_obj_idr_type uverbs_type_attrs_flow = { + .type = UVERBS_TYPE_ALLOC_IDR(0), + .destroy_object = uverbs_free_flow, +}; + +const struct uverbs_obj_idr_type uverbs_type_attrs_wq = { + .type = UVERBS_TYPE_ALLOC_IDR_SZ(sizeof(struct ib_uwq_object), 0), + .destroy_object = uverbs_free_wq, +}; + +const struct uverbs_obj_idr_type uverbs_type_attrs_rwq_ind_table = { + .type = UVERBS_TYPE_ALLOC_IDR(0), + .destroy_object = uverbs_free_rwq_ind_tbl, +}; + +const struct uverbs_obj_idr_type uverbs_type_attrs_xrcd = { + .type = UVERBS_TYPE_ALLOC_IDR_SZ(sizeof(struct ib_uxrcd_object), 0), + .destroy_object = uverbs_free_xrcd, +}; + +const struct uverbs_obj_idr_type uverbs_type_attrs_pd = { + /* 2 is used in order to free the PD after MRs */ + .type = UVERBS_TYPE_ALLOC_IDR(2), + .destroy_object = uverbs_free_pd, +}; diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c index 71580cc28c9e..fb98ed67d5bc 100644 --- a/drivers/infiniband/core/verbs.c +++ b/drivers/infiniband/core/verbs.c @@ -44,6 +44,7 @@ #include <linux/in.h> #include <linux/in6.h> #include <net/addrconf.h> +#include <linux/security.h> #include <rdma/ib_verbs.h> #include <rdma/ib_cache.h> @@ -311,7 +312,7 @@ EXPORT_SYMBOL(ib_dealloc_pd); /* Address handles */ -struct ib_ah *ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr) +struct ib_ah *rdma_create_ah(struct ib_pd *pd, struct rdma_ah_attr *ah_attr) { struct ib_ah *ah; @@ -321,12 +322,13 @@ struct ib_ah *ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr) ah->device = pd->device; ah->pd = pd; ah->uobject = NULL; + ah->type = ah_attr->type; atomic_inc(&pd->usecnt); } return ah; } -EXPORT_SYMBOL(ib_create_ah); +EXPORT_SYMBOL(rdma_create_ah); int ib_get_rdma_header_version(const union rdma_network_hdr *hdr) { @@ -450,9 +452,22 @@ int ib_get_gids_from_rdma_hdr(const union rdma_network_hdr *hdr, } EXPORT_SYMBOL(ib_get_gids_from_rdma_hdr); +/* + * This function creates ah 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. + * + * When resolving mac address of destination, the arrived dgid is used + * 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 ib_ah_attr *ah_attr) + struct rdma_ah_attr *ah_attr) { u32 flow_class; u16 gid_index; @@ -464,6 +479,7 @@ int ib_init_ah_from_wc(struct ib_device *device, u8 port_num, union ib_gid sgid; memset(ah_attr, 0, sizeof *ah_attr); + ah_attr->type = rdma_ah_find_type(device, port_num); if (rdma_cap_eth_ah(device, port_num)) { if (wc->wc_flags & IB_WC_WITH_NETWORK_HDR_TYPE) net_type = wc->network_hdr_type; @@ -494,7 +510,7 @@ int ib_init_ah_from_wc(struct ib_device *device, u8 port_num, return -ENODEV; ret = rdma_addr_find_l2_eth_by_grh(&dgid, &sgid, - ah_attr->dmac, + ah_attr->roce.dmac, wc->wc_flags & IB_WC_WITH_VLAN ? NULL : &vlan_id, &if_index, &hoplimit); @@ -504,11 +520,6 @@ int ib_init_ah_from_wc(struct ib_device *device, u8 port_num, } resolved_dev = dev_get_by_index(&init_net, if_index); - if (resolved_dev->flags & IFF_LOOPBACK) { - dev_put(resolved_dev); - resolved_dev = idev; - dev_hold(resolved_dev); - } rcu_read_lock(); if (resolved_dev != idev && !rdma_is_upper_dev_rcu(idev, resolved_dev)) @@ -525,15 +536,12 @@ int ib_init_ah_from_wc(struct ib_device *device, u8 port_num, return ret; } - ah_attr->dlid = wc->slid; - ah_attr->sl = wc->sl; - ah_attr->src_path_bits = wc->dlid_path_bits; - ah_attr->port_num = port_num; + 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); if (wc->wc_flags & IB_WC_GRH) { - ah_attr->ah_flags = IB_AH_GRH; - ah_attr->grh.dgid = sgid; - if (!rdma_cap_eth_ah(device, port_num)) { if (dgid.global.interface_id != cpu_to_be64(IB_SA_WELL_KNOWN_GUID)) { ret = ib_find_cached_gid_by_port(device, &dgid, @@ -547,11 +555,12 @@ int ib_init_ah_from_wc(struct ib_device *device, u8 port_num, } } - ah_attr->grh.sgid_index = (u8) gid_index; flow_class = be32_to_cpu(grh->version_tclass_flow); - ah_attr->grh.flow_label = flow_class & 0xFFFFF; - ah_attr->grh.hop_limit = hoplimit; - ah_attr->grh.traffic_class = (flow_class >> 20) & 0xFF; + rdma_ah_set_grh(ah_attr, &sgid, + flow_class & 0xFFFFF, + (u8)gid_index, hoplimit, + (flow_class >> 20) & 0xFF); + } return 0; } @@ -560,34 +569,37 @@ EXPORT_SYMBOL(ib_init_ah_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) { - struct ib_ah_attr ah_attr; + struct rdma_ah_attr ah_attr; int ret; ret = ib_init_ah_from_wc(pd->device, port_num, wc, grh, &ah_attr); if (ret) return ERR_PTR(ret); - return ib_create_ah(pd, &ah_attr); + return rdma_create_ah(pd, &ah_attr); } EXPORT_SYMBOL(ib_create_ah_from_wc); -int ib_modify_ah(struct ib_ah *ah, struct ib_ah_attr *ah_attr) +int rdma_modify_ah(struct ib_ah *ah, struct rdma_ah_attr *ah_attr) { + if (ah->type != ah_attr->type) + return -EINVAL; + return ah->device->modify_ah ? ah->device->modify_ah(ah, ah_attr) : -ENOSYS; } -EXPORT_SYMBOL(ib_modify_ah); +EXPORT_SYMBOL(rdma_modify_ah); -int ib_query_ah(struct ib_ah *ah, struct ib_ah_attr *ah_attr) +int rdma_query_ah(struct ib_ah *ah, struct rdma_ah_attr *ah_attr) { return ah->device->query_ah ? ah->device->query_ah(ah, ah_attr) : -ENOSYS; } -EXPORT_SYMBOL(ib_query_ah); +EXPORT_SYMBOL(rdma_query_ah); -int ib_destroy_ah(struct ib_ah *ah) +int rdma_destroy_ah(struct ib_ah *ah) { struct ib_pd *pd; int ret; @@ -599,7 +611,7 @@ int ib_destroy_ah(struct ib_ah *ah) return ret; } -EXPORT_SYMBOL(ib_destroy_ah); +EXPORT_SYMBOL(rdma_destroy_ah); /* Shared receive queues */ @@ -710,12 +722,20 @@ static struct ib_qp *__ib_open_qp(struct ib_qp *real_qp, { struct ib_qp *qp; unsigned long flags; + int err; qp = kzalloc(sizeof *qp, GFP_KERNEL); if (!qp) return ERR_PTR(-ENOMEM); qp->real_qp = real_qp; + err = ib_open_shared_qp_security(qp, real_qp->device); + if (err) { + kfree(qp); + return ERR_PTR(err); + } + + qp->real_qp = real_qp; atomic_inc(&real_qp->usecnt); qp->device = real_qp->device; qp->event_handler = event_handler; @@ -801,6 +821,12 @@ struct ib_qp *ib_create_qp(struct ib_pd *pd, if (IS_ERR(qp)) return qp; + ret = ib_create_qp_security(qp, device); + if (ret) { + ib_destroy_qp(qp); + return ERR_PTR(ret); + } + qp->device = device; qp->real_qp = qp; qp->uobject = NULL; @@ -869,6 +895,7 @@ static const struct { } qp_state_table[IB_QPS_ERR + 1][IB_QPS_ERR + 1] = { [IB_QPS_RESET] = { [IB_QPS_RESET] = { .valid = 1 }, + [IB_QPS_ERR] = { .valid = 1 }, [IB_QPS_INIT] = { .valid = 1, .req_param = { @@ -1201,20 +1228,22 @@ int ib_modify_qp_is_ok(enum ib_qp_state cur_state, enum ib_qp_state next_state, EXPORT_SYMBOL(ib_modify_qp_is_ok); int ib_resolve_eth_dmac(struct ib_device *device, - struct ib_ah_attr *ah_attr) + struct rdma_ah_attr *ah_attr) { int ret = 0; + struct ib_global_route *grh; - if (ah_attr->port_num < rdma_start_port(device) || - ah_attr->port_num > rdma_end_port(device)) + if (!rdma_is_port_valid(device, rdma_ah_get_port_num(ah_attr))) return -EINVAL; - if (!rdma_cap_eth_ah(device, ah_attr->port_num)) + if (ah_attr->type != RDMA_AH_ATTR_TYPE_ROCE) return 0; - if (rdma_link_local_addr((struct in6_addr *)ah_attr->grh.dgid.raw)) { - rdma_get_ll_mac((struct in6_addr *)ah_attr->grh.dgid.raw, - ah_attr->dmac); + 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); } else { union ib_gid sgid; struct ib_gid_attr sgid_attr; @@ -1222,8 +1251,8 @@ int ib_resolve_eth_dmac(struct ib_device *device, int hop_limit; ret = ib_query_gid(device, - ah_attr->port_num, - ah_attr->grh.sgid_index, + rdma_ah_get_port_num(ah_attr), + grh->sgid_index, &sgid, &sgid_attr); if (ret || !sgid_attr.ndev) { @@ -1234,34 +1263,50 @@ int ib_resolve_eth_dmac(struct ib_device *device, ifindex = sgid_attr.ndev->ifindex; - ret = rdma_addr_find_l2_eth_by_grh(&sgid, - &ah_attr->grh.dgid, - ah_attr->dmac, - NULL, &ifindex, &hop_limit); + ret = + rdma_addr_find_l2_eth_by_grh(&sgid, &grh->dgid, + ah_attr->roce.dmac, + NULL, &ifindex, &hop_limit); dev_put(sgid_attr.ndev); - ah_attr->grh.hop_limit = hop_limit; + grh->hop_limit = hop_limit; } out: return ret; } EXPORT_SYMBOL(ib_resolve_eth_dmac); -int ib_modify_qp(struct ib_qp *qp, - struct ib_qp_attr *qp_attr, - int qp_attr_mask) +/** + * ib_modify_qp_with_udata - Modifies the attributes for the specified QP. + * @qp: The QP to modify. + * @attr: On input, specifies the QP attributes to modify. On output, + * the current values of selected QP attributes are returned. + * @attr_mask: A bit-mask used to specify which attributes of the QP + * are being modified. + * @udata: pointer to user's input output buffer information + * are being modified. + * It returns 0 on success and returns appropriate error code on error. + */ +int ib_modify_qp_with_udata(struct ib_qp *qp, struct ib_qp_attr *attr, + int attr_mask, struct ib_udata *udata) { + int ret; - if (qp_attr_mask & IB_QP_AV) { - int ret; - - ret = ib_resolve_eth_dmac(qp->device, &qp_attr->ah_attr); + if (attr_mask & IB_QP_AV) { + ret = ib_resolve_eth_dmac(qp->device, &attr->ah_attr); if (ret) return ret; } + return ib_security_modify_qp(qp, attr, attr_mask, udata); +} +EXPORT_SYMBOL(ib_modify_qp_with_udata); - return qp->device->modify_qp(qp->real_qp, qp_attr, qp_attr_mask, NULL); +int ib_modify_qp(struct ib_qp *qp, + struct ib_qp_attr *qp_attr, + int qp_attr_mask) +{ + return ib_modify_qp_with_udata(qp, qp_attr, qp_attr_mask, NULL); } EXPORT_SYMBOL(ib_modify_qp); @@ -1290,6 +1335,7 @@ int ib_close_qp(struct ib_qp *qp) spin_unlock_irqrestore(&real_qp->device->event_handler_lock, flags); atomic_dec(&real_qp->usecnt); + ib_close_shared_qp_security(qp->qp_sec); kfree(qp); return 0; @@ -1330,6 +1376,7 @@ int ib_destroy_qp(struct ib_qp *qp) struct ib_cq *scq, *rcq; struct ib_srq *srq; struct ib_rwq_ind_table *ind_tbl; + struct ib_qp_security *sec; int ret; WARN_ON_ONCE(qp->mrs_used > 0); @@ -1345,6 +1392,9 @@ int ib_destroy_qp(struct ib_qp *qp) rcq = qp->recv_cq; srq = qp->srq; ind_tbl = qp->rwq_ind_tbl; + sec = qp->qp_sec; + if (sec) + ib_destroy_qp_security_begin(sec); if (!qp->uobject) rdma_rw_cleanup_mrs(qp); @@ -1361,6 +1411,11 @@ int ib_destroy_qp(struct ib_qp *qp) atomic_dec(&srq->usecnt); if (ind_tbl) atomic_dec(&ind_tbl->usecnt); + if (sec) + ib_destroy_qp_security_end(sec); + } else { + if (sec) + ib_destroy_qp_security_abort(sec); } return ret; @@ -1520,7 +1575,9 @@ int ib_attach_mcast(struct ib_qp *qp, union ib_gid *gid, u16 lid) if (!qp->device->attach_mcast) return -ENOSYS; - if (gid->raw[0] != 0xff || qp->qp_type != IB_QPT_UD) + if (gid->raw[0] != 0xff || qp->qp_type != IB_QPT_UD || + lid < be16_to_cpu(IB_MULTICAST_LID_BASE) || + lid == be16_to_cpu(IB_LID_PERMISSIVE)) return -EINVAL; ret = qp->device->attach_mcast(qp, gid, lid); @@ -1536,7 +1593,9 @@ int ib_detach_mcast(struct ib_qp *qp, union ib_gid *gid, u16 lid) if (!qp->device->detach_mcast) return -ENOSYS; - if (gid->raw[0] != 0xff || qp->qp_type != IB_QPT_UD) + if (gid->raw[0] != 0xff || qp->qp_type != IB_QPT_UD || + lid < be16_to_cpu(IB_MULTICAST_LID_BASE) || + lid == be16_to_cpu(IB_LID_PERMISSIVE)) return -EINVAL; ret = qp->device->detach_mcast(qp, gid, lid); @@ -1949,17 +2008,12 @@ static void ib_drain_qp_done(struct ib_cq *cq, struct ib_wc *wc) */ static void __ib_drain_sq(struct ib_qp *qp) { + struct ib_cq *cq = qp->send_cq; struct ib_qp_attr attr = { .qp_state = IB_QPS_ERR }; struct ib_drain_cqe sdrain; struct ib_send_wr swr = {}, *bad_swr; int ret; - if (qp->send_cq->poll_ctx == IB_POLL_DIRECT) { - WARN_ONCE(qp->send_cq->poll_ctx == IB_POLL_DIRECT, - "IB_POLL_DIRECT poll_ctx not supported for drain\n"); - return; - } - swr.wr_cqe = &sdrain.cqe; sdrain.cqe.done = ib_drain_qp_done; init_completion(&sdrain.done); @@ -1976,7 +2030,11 @@ static void __ib_drain_sq(struct ib_qp *qp) return; } - wait_for_completion(&sdrain.done); + if (cq->poll_ctx == IB_POLL_DIRECT) + while (wait_for_completion_timeout(&sdrain.done, HZ / 10) <= 0) + ib_process_cq_direct(cq, -1); + else + wait_for_completion(&sdrain.done); } /* @@ -1984,17 +2042,12 @@ static void __ib_drain_sq(struct ib_qp *qp) */ static void __ib_drain_rq(struct ib_qp *qp) { + struct ib_cq *cq = qp->recv_cq; struct ib_qp_attr attr = { .qp_state = IB_QPS_ERR }; struct ib_drain_cqe rdrain; struct ib_recv_wr rwr = {}, *bad_rwr; int ret; - if (qp->recv_cq->poll_ctx == IB_POLL_DIRECT) { - WARN_ONCE(qp->recv_cq->poll_ctx == IB_POLL_DIRECT, - "IB_POLL_DIRECT poll_ctx not supported for drain\n"); - return; - } - rwr.wr_cqe = &rdrain.cqe; rdrain.cqe.done = ib_drain_qp_done; init_completion(&rdrain.done); @@ -2011,7 +2064,11 @@ static void __ib_drain_rq(struct ib_qp *qp) return; } - wait_for_completion(&rdrain.done); + if (cq->poll_ctx == IB_POLL_DIRECT) + while (wait_for_completion_timeout(&rdrain.done, HZ / 10) <= 0) + ib_process_cq_direct(cq, -1); + else + wait_for_completion(&rdrain.done); } /** @@ -2028,8 +2085,7 @@ static void __ib_drain_rq(struct ib_qp *qp) * ensure there is room in the CQ and SQ for the drain work request and * completion. * - * allocate the CQ using ib_alloc_cq() and the CQ poll context cannot be - * IB_POLL_DIRECT. + * allocate the CQ using ib_alloc_cq(). * * ensure that there are no other contexts that are posting WRs concurrently. * Otherwise the drain is not guaranteed. @@ -2057,8 +2113,7 @@ EXPORT_SYMBOL(ib_drain_sq); * ensure there is room in the CQ and RQ for the drain work request and * completion. * - * allocate the CQ using ib_alloc_cq() and the CQ poll context cannot be - * IB_POLL_DIRECT. + * allocate the CQ using ib_alloc_cq(). * * ensure that there are no other contexts that are posting WRs concurrently. * Otherwise the drain is not guaranteed. @@ -2082,8 +2137,7 @@ EXPORT_SYMBOL(ib_drain_rq); * ensure there is room in the CQ(s), SQ, and RQ for drain work requests * and completions. * - * allocate the CQs using ib_alloc_cq() and the CQ poll context cannot be - * IB_POLL_DIRECT. + * allocate the CQs using ib_alloc_cq(). * * ensure that there are no other contexts that are posting WRs concurrently. * Otherwise the drain is not guaranteed. diff --git a/drivers/infiniband/hw/Makefile b/drivers/infiniband/hw/Makefile index ed553de2ca12..34c93abf0fe0 100644 --- a/drivers/infiniband/hw/Makefile +++ b/drivers/infiniband/hw/Makefile @@ -12,3 +12,4 @@ obj-$(CONFIG_INFINIBAND_USNIC) += usnic/ obj-$(CONFIG_INFINIBAND_HFI1) += hfi1/ obj-$(CONFIG_INFINIBAND_HNS) += hns/ obj-$(CONFIG_INFINIBAND_QEDR) += qedr/ +obj-$(CONFIG_INFINIBAND_BNXT_RE) += bnxt_re/ diff --git a/drivers/infiniband/hw/bnxt_re/Kconfig b/drivers/infiniband/hw/bnxt_re/Kconfig new file mode 100644 index 000000000000..19982a4a9bba --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/Kconfig @@ -0,0 +1,9 @@ +config INFINIBAND_BNXT_RE + tristate "Broadcom Netxtreme HCA support" + depends on ETHERNET && NETDEVICES && PCI && INET && DCB + select NET_VENDOR_BROADCOM + select BNXT + ---help--- + This driver supports Broadcom NetXtreme-E 10/25/40/50 gigabit + RoCE HCAs. To compile this driver as a module, choose M here: + the module will be called bnxt_re. diff --git a/drivers/infiniband/hw/bnxt_re/Makefile b/drivers/infiniband/hw/bnxt_re/Makefile new file mode 100644 index 000000000000..036f84efbc73 --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/Makefile @@ -0,0 +1,6 @@ + +ccflags-y := -Idrivers/net/ethernet/broadcom/bnxt +obj-$(CONFIG_INFINIBAND_BNXT_RE) += bnxt_re.o +bnxt_re-y := main.o ib_verbs.o \ + qplib_res.o qplib_rcfw.o \ + qplib_sp.o qplib_fp.o diff --git a/drivers/infiniband/hw/bnxt_re/bnxt_re.h b/drivers/infiniband/hw/bnxt_re/bnxt_re.h new file mode 100644 index 000000000000..85527532c49d --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/bnxt_re.h @@ -0,0 +1,159 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * 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 + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. 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. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: Slow Path Operators (header) + * + */ + +#ifndef __BNXT_RE_H__ +#define __BNXT_RE_H__ +#define ROCE_DRV_MODULE_NAME "bnxt_re" +#define ROCE_DRV_MODULE_VERSION "1.0.0" + +#define BNXT_RE_DESC "Broadcom NetXtreme-C/E RoCE Driver" + +#define BNXT_RE_PAGE_SIZE_4K BIT(12) +#define BNXT_RE_PAGE_SIZE_8K BIT(13) +#define BNXT_RE_PAGE_SIZE_64K BIT(16) +#define BNXT_RE_PAGE_SIZE_2M BIT(21) +#define BNXT_RE_PAGE_SIZE_8M BIT(23) +#define BNXT_RE_PAGE_SIZE_1G BIT(30) + +#define BNXT_RE_MAX_MR_SIZE BIT(30) + +#define BNXT_RE_MAX_QPC_COUNT (64 * 1024) +#define BNXT_RE_MAX_MRW_COUNT (64 * 1024) +#define BNXT_RE_MAX_SRQC_COUNT (64 * 1024) +#define BNXT_RE_MAX_CQ_COUNT (64 * 1024) + +#define BNXT_RE_UD_QP_HW_STALL 0x400000 + +#define BNXT_RE_RQ_WQE_THRESHOLD 32 + +/* + * Setting the default ack delay value to 16, which means + * the default timeout is approx. 260ms(4 usec * 2 ^(timeout)) + */ + +#define BNXT_RE_DEFAULT_ACK_DELAY 16 + +struct bnxt_re_work { + struct work_struct work; + unsigned long event; + struct bnxt_re_dev *rdev; + struct net_device *vlan_dev; +}; + +struct bnxt_re_sqp_entries { + struct bnxt_qplib_sge sge; + u64 wrid; + /* For storing the actual qp1 cqe */ + struct bnxt_qplib_cqe cqe; + struct bnxt_re_qp *qp1_qp; +}; + +#define BNXT_RE_MIN_MSIX 2 +#define BNXT_RE_MAX_MSIX 16 +#define BNXT_RE_AEQ_IDX 0 +#define BNXT_RE_NQ_IDX 1 + +struct bnxt_re_dev { + struct ib_device ibdev; + struct list_head list; + unsigned long flags; +#define BNXT_RE_FLAG_NETDEV_REGISTERED 0 +#define BNXT_RE_FLAG_IBDEV_REGISTERED 1 +#define BNXT_RE_FLAG_GOT_MSIX 2 +#define BNXT_RE_FLAG_RCFW_CHANNEL_EN 8 +#define BNXT_RE_FLAG_QOS_WORK_REG 16 + struct net_device *netdev; + unsigned int version, major, minor; + struct bnxt_en_dev *en_dev; + struct bnxt_msix_entry msix_entries[BNXT_RE_MAX_MSIX]; + int num_msix; + + int id; + + struct delayed_work worker; + u8 cur_prio_map; + + /* FP Notification Queue (CQ & SRQ) */ + struct tasklet_struct nq_task; + + /* RCFW Channel */ + struct bnxt_qplib_rcfw rcfw; + + /* NQ */ + struct bnxt_qplib_nq nq; + + /* Device Resources */ + struct bnxt_qplib_dev_attr dev_attr; + struct bnxt_qplib_ctx qplib_ctx; + struct bnxt_qplib_res qplib_res; + struct bnxt_qplib_dpi dpi_privileged; + + atomic_t qp_count; + struct mutex qp_lock; /* protect qp list */ + struct list_head qp_list; + + atomic_t cq_count; + atomic_t srq_count; + atomic_t mr_count; + atomic_t mw_count; + /* Max of 2 lossless traffic class supported per port */ + u16 cosq[2]; + + /* QP for for handling QP1 packets */ + u32 sqp_id; + struct bnxt_re_qp *qp1_sqp; + struct bnxt_re_ah *sqp_ah; + struct bnxt_re_sqp_entries sqp_tbl[1024]; +}; + +#define to_bnxt_re_dev(ptr, member) \ + container_of((ptr), struct bnxt_re_dev, member) + +#define BNXT_RE_ROCE_V1_PACKET 0 +#define BNXT_RE_ROCEV2_IPV4_PACKET 2 +#define BNXT_RE_ROCEV2_IPV6_PACKET 3 + +static inline struct device *rdev_to_dev(struct bnxt_re_dev *rdev) +{ + if (rdev) + return &rdev->ibdev.dev; + return NULL; +} + +#endif diff --git a/drivers/infiniband/hw/bnxt_re/ib_verbs.c b/drivers/infiniband/hw/bnxt_re/ib_verbs.c new file mode 100644 index 000000000000..f0e01b3ac711 --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/ib_verbs.c @@ -0,0 +1,3452 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * 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 + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. 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. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: IB Verbs interpreter + */ + +#include <linux/interrupt.h> +#include <linux/types.h> +#include <linux/pci.h> +#include <linux/netdevice.h> +#include <linux/if_ether.h> + +#include <rdma/ib_verbs.h> +#include <rdma/ib_user_verbs.h> +#include <rdma/ib_umem.h> +#include <rdma/ib_addr.h> +#include <rdma/ib_mad.h> +#include <rdma/ib_cache.h> + +#include "bnxt_ulp.h" + +#include "roce_hsi.h" +#include "qplib_res.h" +#include "qplib_sp.h" +#include "qplib_fp.h" +#include "qplib_rcfw.h" + +#include "bnxt_re.h" +#include "ib_verbs.h" +#include <rdma/bnxt_re-abi.h> + +static int __from_ib_access_flags(int iflags) +{ + int qflags = 0; + + if (iflags & IB_ACCESS_LOCAL_WRITE) + qflags |= BNXT_QPLIB_ACCESS_LOCAL_WRITE; + if (iflags & IB_ACCESS_REMOTE_READ) + qflags |= BNXT_QPLIB_ACCESS_REMOTE_READ; + if (iflags & IB_ACCESS_REMOTE_WRITE) + qflags |= BNXT_QPLIB_ACCESS_REMOTE_WRITE; + if (iflags & IB_ACCESS_REMOTE_ATOMIC) + qflags |= BNXT_QPLIB_ACCESS_REMOTE_ATOMIC; + if (iflags & IB_ACCESS_MW_BIND) + qflags |= BNXT_QPLIB_ACCESS_MW_BIND; + if (iflags & IB_ZERO_BASED) + qflags |= BNXT_QPLIB_ACCESS_ZERO_BASED; + if (iflags & IB_ACCESS_ON_DEMAND) + qflags |= BNXT_QPLIB_ACCESS_ON_DEMAND; + return qflags; +}; + +static enum ib_access_flags __to_ib_access_flags(int qflags) +{ + enum ib_access_flags iflags = 0; + + if (qflags & BNXT_QPLIB_ACCESS_LOCAL_WRITE) + iflags |= IB_ACCESS_LOCAL_WRITE; + if (qflags & BNXT_QPLIB_ACCESS_REMOTE_WRITE) + iflags |= IB_ACCESS_REMOTE_WRITE; + if (qflags & BNXT_QPLIB_ACCESS_REMOTE_READ) + iflags |= IB_ACCESS_REMOTE_READ; + if (qflags & BNXT_QPLIB_ACCESS_REMOTE_ATOMIC) + iflags |= IB_ACCESS_REMOTE_ATOMIC; + if (qflags & BNXT_QPLIB_ACCESS_MW_BIND) + iflags |= IB_ACCESS_MW_BIND; + if (qflags & BNXT_QPLIB_ACCESS_ZERO_BASED) + iflags |= IB_ZERO_BASED; + if (qflags & BNXT_QPLIB_ACCESS_ON_DEMAND) + iflags |= IB_ACCESS_ON_DEMAND; + return iflags; +}; + +static int bnxt_re_build_sgl(struct ib_sge *ib_sg_list, + struct bnxt_qplib_sge *sg_list, int num) +{ + int i, total = 0; + + for (i = 0; i < num; i++) { + sg_list[i].addr = ib_sg_list[i].addr; + sg_list[i].lkey = ib_sg_list[i].lkey; + sg_list[i].size = ib_sg_list[i].length; + total += sg_list[i].size; + } + return total; +} + +/* Device */ +struct net_device *bnxt_re_get_netdev(struct ib_device *ibdev, u8 port_num) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); + struct net_device *netdev = NULL; + + rcu_read_lock(); + if (rdev) + netdev = rdev->netdev; + if (netdev) + dev_hold(netdev); + + rcu_read_unlock(); + return netdev; +} + +int bnxt_re_query_device(struct ib_device *ibdev, + struct ib_device_attr *ib_attr, + struct ib_udata *udata) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); + struct bnxt_qplib_dev_attr *dev_attr = &rdev->dev_attr; + + memset(ib_attr, 0, sizeof(*ib_attr)); + + ib_attr->fw_ver = (u64)(unsigned long)(dev_attr->fw_ver); + bnxt_qplib_get_guid(rdev->netdev->dev_addr, + (u8 *)&ib_attr->sys_image_guid); + ib_attr->max_mr_size = BNXT_RE_MAX_MR_SIZE; + ib_attr->page_size_cap = BNXT_RE_PAGE_SIZE_4K; + + ib_attr->vendor_id = rdev->en_dev->pdev->vendor; + ib_attr->vendor_part_id = rdev->en_dev->pdev->device; + ib_attr->hw_ver = rdev->en_dev->pdev->subsystem_device; + ib_attr->max_qp = dev_attr->max_qp; + ib_attr->max_qp_wr = dev_attr->max_qp_wqes; + ib_attr->device_cap_flags = + IB_DEVICE_CURR_QP_STATE_MOD + | IB_DEVICE_RC_RNR_NAK_GEN + | IB_DEVICE_SHUTDOWN_PORT + | IB_DEVICE_SYS_IMAGE_GUID + | IB_DEVICE_LOCAL_DMA_LKEY + | IB_DEVICE_RESIZE_MAX_WR + | IB_DEVICE_PORT_ACTIVE_EVENT + | IB_DEVICE_N_NOTIFY_CQ + | IB_DEVICE_MEM_WINDOW + | IB_DEVICE_MEM_WINDOW_TYPE_2B + | IB_DEVICE_MEM_MGT_EXTENSIONS; + ib_attr->max_sge = dev_attr->max_qp_sges; + ib_attr->max_sge_rd = dev_attr->max_qp_sges; + ib_attr->max_cq = dev_attr->max_cq; + ib_attr->max_cqe = dev_attr->max_cq_wqes; + ib_attr->max_mr = dev_attr->max_mr; + ib_attr->max_pd = dev_attr->max_pd; + ib_attr->max_qp_rd_atom = dev_attr->max_qp_rd_atom; + ib_attr->max_qp_init_rd_atom = dev_attr->max_qp_init_rd_atom; + if (dev_attr->is_atomic) { + ib_attr->atomic_cap = IB_ATOMIC_HCA; + ib_attr->masked_atomic_cap = IB_ATOMIC_HCA; + } + + ib_attr->max_ee_rd_atom = 0; + ib_attr->max_res_rd_atom = 0; + ib_attr->max_ee_init_rd_atom = 0; + ib_attr->max_ee = 0; + ib_attr->max_rdd = 0; + ib_attr->max_mw = dev_attr->max_mw; + ib_attr->max_raw_ipv6_qp = 0; + ib_attr->max_raw_ethy_qp = dev_attr->max_raw_ethy_qp; + ib_attr->max_mcast_grp = 0; + ib_attr->max_mcast_qp_attach = 0; + ib_attr->max_total_mcast_qp_attach = 0; + ib_attr->max_ah = dev_attr->max_ah; + + ib_attr->max_fmr = 0; + ib_attr->max_map_per_fmr = 0; + + ib_attr->max_srq = dev_attr->max_srq; + ib_attr->max_srq_wr = dev_attr->max_srq_wqes; + ib_attr->max_srq_sge = dev_attr->max_srq_sges; + + ib_attr->max_fast_reg_page_list_len = MAX_PBL_LVL_1_PGS; + + ib_attr->max_pkeys = 1; + ib_attr->local_ca_ack_delay = BNXT_RE_DEFAULT_ACK_DELAY; + return 0; +} + +int bnxt_re_modify_device(struct ib_device *ibdev, + int device_modify_mask, + struct ib_device_modify *device_modify) +{ + switch (device_modify_mask) { + case IB_DEVICE_MODIFY_SYS_IMAGE_GUID: + /* Modify the GUID requires the modification of the GID table */ + /* GUID should be made as READ-ONLY */ + break; + case IB_DEVICE_MODIFY_NODE_DESC: + /* Node Desc should be made as READ-ONLY */ + break; + default: + break; + } + return 0; +} + +static void __to_ib_speed_width(struct net_device *netdev, u8 *speed, u8 *width) +{ + struct ethtool_link_ksettings lksettings; + u32 espeed; + + if (netdev->ethtool_ops && netdev->ethtool_ops->get_link_ksettings) { + memset(&lksettings, 0, sizeof(lksettings)); + rtnl_lock(); + netdev->ethtool_ops->get_link_ksettings(netdev, &lksettings); + rtnl_unlock(); + espeed = lksettings.base.speed; + } else { + espeed = SPEED_UNKNOWN; + } + switch (espeed) { + case SPEED_1000: + *speed = IB_SPEED_SDR; + *width = IB_WIDTH_1X; + break; + case SPEED_10000: + *speed = IB_SPEED_QDR; + *width = IB_WIDTH_1X; + break; + case SPEED_20000: + *speed = IB_SPEED_DDR; + *width = IB_WIDTH_4X; + break; + case SPEED_25000: + *speed = IB_SPEED_EDR; + *width = IB_WIDTH_1X; + break; + case SPEED_40000: + *speed = IB_SPEED_QDR; + *width = IB_WIDTH_4X; + break; + case SPEED_50000: + break; + default: + *speed = IB_SPEED_SDR; + *width = IB_WIDTH_1X; + break; + } +} + +/* Port */ +int bnxt_re_query_port(struct ib_device *ibdev, u8 port_num, + struct ib_port_attr *port_attr) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); + struct bnxt_qplib_dev_attr *dev_attr = &rdev->dev_attr; + + memset(port_attr, 0, sizeof(*port_attr)); + + if (netif_running(rdev->netdev) && netif_carrier_ok(rdev->netdev)) { + port_attr->state = IB_PORT_ACTIVE; + port_attr->phys_state = 5; + } else { + port_attr->state = IB_PORT_DOWN; + port_attr->phys_state = 3; + } + port_attr->max_mtu = IB_MTU_4096; + port_attr->active_mtu = iboe_get_mtu(rdev->netdev->mtu); + port_attr->gid_tbl_len = dev_attr->max_sgid; + port_attr->port_cap_flags = IB_PORT_CM_SUP | IB_PORT_REINIT_SUP | + IB_PORT_DEVICE_MGMT_SUP | + IB_PORT_VENDOR_CLASS_SUP | + IB_PORT_IP_BASED_GIDS; + + /* Max MSG size set to 2G for now */ + port_attr->max_msg_sz = 0x80000000; + port_attr->bad_pkey_cntr = 0; + port_attr->qkey_viol_cntr = 0; + port_attr->pkey_tbl_len = dev_attr->max_pkey; + port_attr->lid = 0; + port_attr->sm_lid = 0; + port_attr->lmc = 0; + port_attr->max_vl_num = 4; + port_attr->sm_sl = 0; + port_attr->subnet_timeout = 0; + port_attr->init_type_reply = 0; + /* call the underlying netdev's ethtool hooks to query speed settings + * for which we acquire rtnl_lock _only_ if it's registered with + * IB stack to avoid race in the NETDEV_UNREG path + */ + if (test_bit(BNXT_RE_FLAG_IBDEV_REGISTERED, &rdev->flags)) + __to_ib_speed_width(rdev->netdev, &port_attr->active_speed, + &port_attr->active_width); + return 0; +} + +int bnxt_re_modify_port(struct ib_device *ibdev, u8 port_num, + int port_modify_mask, + struct ib_port_modify *port_modify) +{ + switch (port_modify_mask) { + case IB_PORT_SHUTDOWN: + break; + case IB_PORT_INIT_TYPE: + break; + case IB_PORT_RESET_QKEY_CNTR: + break; + default: + break; + } + return 0; +} + +int bnxt_re_get_port_immutable(struct ib_device *ibdev, u8 port_num, + struct ib_port_immutable *immutable) +{ + struct ib_port_attr port_attr; + + if (bnxt_re_query_port(ibdev, port_num, &port_attr)) + return -EINVAL; + + immutable->pkey_tbl_len = port_attr.pkey_tbl_len; + immutable->gid_tbl_len = port_attr.gid_tbl_len; + immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE; + immutable->core_cap_flags |= RDMA_CORE_CAP_PROT_ROCE_UDP_ENCAP; + immutable->max_mad_size = IB_MGMT_MAD_SIZE; + return 0; +} + +int bnxt_re_query_pkey(struct ib_device *ibdev, u8 port_num, + u16 index, u16 *pkey) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); + + /* Ignore port_num */ + + memset(pkey, 0, sizeof(*pkey)); + return bnxt_qplib_get_pkey(&rdev->qplib_res, + &rdev->qplib_res.pkey_tbl, index, pkey); +} + +int bnxt_re_query_gid(struct ib_device *ibdev, u8 port_num, + int index, union ib_gid *gid) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); + int rc = 0; + + /* Ignore port_num */ + memset(gid, 0, sizeof(*gid)); + rc = bnxt_qplib_get_sgid(&rdev->qplib_res, + &rdev->qplib_res.sgid_tbl, index, + (struct bnxt_qplib_gid *)gid); + return rc; +} + +int bnxt_re_del_gid(struct ib_device *ibdev, u8 port_num, + unsigned int index, void **context) +{ + int rc = 0; + struct bnxt_re_gid_ctx *ctx, **ctx_tbl; + struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); + struct bnxt_qplib_sgid_tbl *sgid_tbl = &rdev->qplib_res.sgid_tbl; + + /* Delete the entry from the hardware */ + ctx = *context; + if (!ctx) + return -EINVAL; + + if (sgid_tbl && sgid_tbl->active) { + if (ctx->idx >= sgid_tbl->max) + return -EINVAL; + ctx->refcnt--; + if (!ctx->refcnt) { + rc = bnxt_qplib_del_sgid(sgid_tbl, + &sgid_tbl->tbl[ctx->idx], + true); + if (rc) { + dev_err(rdev_to_dev(rdev), + "Failed to remove GID: %#x", rc); + } else { + ctx_tbl = sgid_tbl->ctx; + ctx_tbl[ctx->idx] = NULL; + kfree(ctx); + } + } + } else { + return -EINVAL; + } + return rc; +} + +int bnxt_re_add_gid(struct ib_device *ibdev, u8 port_num, + unsigned int index, const union ib_gid *gid, + const struct ib_gid_attr *attr, void **context) +{ + int rc; + u32 tbl_idx = 0; + u16 vlan_id = 0xFFFF; + struct bnxt_re_gid_ctx *ctx, **ctx_tbl; + struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); + struct bnxt_qplib_sgid_tbl *sgid_tbl = &rdev->qplib_res.sgid_tbl; + + if ((attr->ndev) && is_vlan_dev(attr->ndev)) + vlan_id = vlan_dev_vlan_id(attr->ndev); + + rc = bnxt_qplib_add_sgid(sgid_tbl, (struct bnxt_qplib_gid *)gid, + rdev->qplib_res.netdev->dev_addr, + vlan_id, true, &tbl_idx); + if (rc == -EALREADY) { + ctx_tbl = sgid_tbl->ctx; + ctx_tbl[tbl_idx]->refcnt++; + *context = ctx_tbl[tbl_idx]; + return 0; + } + + if (rc < 0) { + dev_err(rdev_to_dev(rdev), "Failed to add GID: %#x", rc); + return rc; + } + + ctx = kmalloc(sizeof(*ctx), GFP_KERNEL); + if (!ctx) + return -ENOMEM; + ctx_tbl = sgid_tbl->ctx; + ctx->idx = tbl_idx; + ctx->refcnt = 1; + ctx_tbl[tbl_idx] = ctx; + + return rc; +} + +enum rdma_link_layer bnxt_re_get_link_layer(struct ib_device *ibdev, + u8 port_num) +{ + return IB_LINK_LAYER_ETHERNET; +} + +#define BNXT_RE_FENCE_PBL_SIZE DIV_ROUND_UP(BNXT_RE_FENCE_BYTES, PAGE_SIZE) + +static void bnxt_re_create_fence_wqe(struct bnxt_re_pd *pd) +{ + struct bnxt_re_fence_data *fence = &pd->fence; + struct ib_mr *ib_mr = &fence->mr->ib_mr; + struct bnxt_qplib_swqe *wqe = &fence->bind_wqe; + + memset(wqe, 0, sizeof(*wqe)); + wqe->type = BNXT_QPLIB_SWQE_TYPE_BIND_MW; + wqe->wr_id = BNXT_QPLIB_FENCE_WRID; + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SIGNAL_COMP; + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_UC_FENCE; + wqe->bind.zero_based = false; + wqe->bind.parent_l_key = ib_mr->lkey; + wqe->bind.va = (u64)(unsigned long)fence->va; + wqe->bind.length = fence->size; + wqe->bind.access_cntl = __from_ib_access_flags(IB_ACCESS_REMOTE_READ); + wqe->bind.mw_type = SQ_BIND_MW_TYPE_TYPE1; + + /* Save the initial rkey in fence structure for now; + * wqe->bind.r_key will be set at (re)bind time. + */ + fence->bind_rkey = ib_inc_rkey(fence->mw->rkey); +} + +static int bnxt_re_bind_fence_mw(struct bnxt_qplib_qp *qplib_qp) +{ + struct bnxt_re_qp *qp = container_of(qplib_qp, struct bnxt_re_qp, + qplib_qp); + struct ib_pd *ib_pd = qp->ib_qp.pd; + struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); + struct bnxt_re_fence_data *fence = &pd->fence; + struct bnxt_qplib_swqe *fence_wqe = &fence->bind_wqe; + struct bnxt_qplib_swqe wqe; + int rc; + + memcpy(&wqe, fence_wqe, sizeof(wqe)); + wqe.bind.r_key = fence->bind_rkey; + fence->bind_rkey = ib_inc_rkey(fence->bind_rkey); + + dev_dbg(rdev_to_dev(qp->rdev), + "Posting bind fence-WQE: rkey: %#x QP: %d PD: %p\n", + wqe.bind.r_key, qp->qplib_qp.id, pd); + rc = bnxt_qplib_post_send(&qp->qplib_qp, &wqe); + if (rc) { + dev_err(rdev_to_dev(qp->rdev), "Failed to bind fence-WQE\n"); + return rc; + } + bnxt_qplib_post_send_db(&qp->qplib_qp); + + return rc; +} + +static void bnxt_re_destroy_fence_mr(struct bnxt_re_pd *pd) +{ + struct bnxt_re_fence_data *fence = &pd->fence; + struct bnxt_re_dev *rdev = pd->rdev; + struct device *dev = &rdev->en_dev->pdev->dev; + struct bnxt_re_mr *mr = fence->mr; + + if (fence->mw) { + bnxt_re_dealloc_mw(fence->mw); + fence->mw = NULL; + } + if (mr) { + if (mr->ib_mr.rkey) + bnxt_qplib_dereg_mrw(&rdev->qplib_res, &mr->qplib_mr, + true); + if (mr->ib_mr.lkey) + bnxt_qplib_free_mrw(&rdev->qplib_res, &mr->qplib_mr); + kfree(mr); + fence->mr = NULL; + } + if (fence->dma_addr) { + dma_unmap_single(dev, fence->dma_addr, BNXT_RE_FENCE_BYTES, + DMA_BIDIRECTIONAL); + fence->dma_addr = 0; + } +} + +static int bnxt_re_create_fence_mr(struct bnxt_re_pd *pd) +{ + int mr_access_flags = IB_ACCESS_LOCAL_WRITE | IB_ACCESS_MW_BIND; + struct bnxt_re_fence_data *fence = &pd->fence; + struct bnxt_re_dev *rdev = pd->rdev; + struct device *dev = &rdev->en_dev->pdev->dev; + struct bnxt_re_mr *mr = NULL; + dma_addr_t dma_addr = 0; + struct ib_mw *mw; + u64 pbl_tbl; + int rc; + + dma_addr = dma_map_single(dev, fence->va, BNXT_RE_FENCE_BYTES, + DMA_BIDIRECTIONAL); + rc = dma_mapping_error(dev, dma_addr); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to dma-map fence-MR-mem\n"); + rc = -EIO; + fence->dma_addr = 0; + goto fail; + } + fence->dma_addr = dma_addr; + + /* Allocate a MR */ + mr = kzalloc(sizeof(*mr), GFP_KERNEL); + if (!mr) { + rc = -ENOMEM; + goto fail; + } + fence->mr = mr; + mr->rdev = rdev; + mr->qplib_mr.pd = &pd->qplib_pd; + mr->qplib_mr.type = CMDQ_ALLOCATE_MRW_MRW_FLAGS_PMR; + mr->qplib_mr.flags = __from_ib_access_flags(mr_access_flags); + rc = bnxt_qplib_alloc_mrw(&rdev->qplib_res, &mr->qplib_mr); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to alloc fence-HW-MR\n"); + goto fail; + } + + /* Register MR */ + mr->ib_mr.lkey = mr->qplib_mr.lkey; + mr->qplib_mr.va = (u64)(unsigned long)fence->va; + mr->qplib_mr.total_size = BNXT_RE_FENCE_BYTES; + pbl_tbl = dma_addr; + rc = bnxt_qplib_reg_mr(&rdev->qplib_res, &mr->qplib_mr, &pbl_tbl, + BNXT_RE_FENCE_PBL_SIZE, false); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to register fence-MR\n"); + goto fail; + } + mr->ib_mr.rkey = mr->qplib_mr.rkey; + + /* Create a fence MW only for kernel consumers */ + mw = bnxt_re_alloc_mw(&pd->ib_pd, IB_MW_TYPE_1, NULL); + if (IS_ERR(mw)) { + dev_err(rdev_to_dev(rdev), + "Failed to create fence-MW for PD: %p\n", pd); + rc = PTR_ERR(mw); + goto fail; + } + fence->mw = mw; + + bnxt_re_create_fence_wqe(pd); + return 0; + +fail: + bnxt_re_destroy_fence_mr(pd); + return rc; +} + +/* Protection Domains */ +int bnxt_re_dealloc_pd(struct ib_pd *ib_pd) +{ + struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); + struct bnxt_re_dev *rdev = pd->rdev; + int rc; + + bnxt_re_destroy_fence_mr(pd); + + if (pd->qplib_pd.id) { + rc = bnxt_qplib_dealloc_pd(&rdev->qplib_res, + &rdev->qplib_res.pd_tbl, + &pd->qplib_pd); + if (rc) + dev_err(rdev_to_dev(rdev), "Failed to deallocate HW PD"); + } + + kfree(pd); + return 0; +} + +struct ib_pd *bnxt_re_alloc_pd(struct ib_device *ibdev, + struct ib_ucontext *ucontext, + struct ib_udata *udata) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); + struct bnxt_re_ucontext *ucntx = container_of(ucontext, + struct bnxt_re_ucontext, + ib_uctx); + struct bnxt_re_pd *pd; + int rc; + + pd = kzalloc(sizeof(*pd), GFP_KERNEL); + if (!pd) + return ERR_PTR(-ENOMEM); + + pd->rdev = rdev; + if (bnxt_qplib_alloc_pd(&rdev->qplib_res.pd_tbl, &pd->qplib_pd)) { + dev_err(rdev_to_dev(rdev), "Failed to allocate HW PD"); + rc = -ENOMEM; + goto fail; + } + + if (udata) { + struct bnxt_re_pd_resp resp; + + if (!ucntx->dpi.dbr) { + /* Allocate DPI in alloc_pd to avoid failing of + * ibv_devinfo and family of application when DPIs + * are depleted. + */ + if (bnxt_qplib_alloc_dpi(&rdev->qplib_res.dpi_tbl, + &ucntx->dpi, ucntx)) { + rc = -ENOMEM; + goto dbfail; + } + } + + resp.pdid = pd->qplib_pd.id; + /* Still allow mapping this DBR to the new user PD. */ + resp.dpi = ucntx->dpi.dpi; + resp.dbr = (u64)ucntx->dpi.umdbr; + + rc = ib_copy_to_udata(udata, &resp, sizeof(resp)); + if (rc) { + dev_err(rdev_to_dev(rdev), + "Failed to copy user response\n"); + goto dbfail; + } + } + + if (!udata) + if (bnxt_re_create_fence_mr(pd)) + dev_warn(rdev_to_dev(rdev), + "Failed to create Fence-MR\n"); + return &pd->ib_pd; +dbfail: + (void)bnxt_qplib_dealloc_pd(&rdev->qplib_res, &rdev->qplib_res.pd_tbl, + &pd->qplib_pd); +fail: + kfree(pd); + return ERR_PTR(rc); +} + +/* Address Handles */ +int bnxt_re_destroy_ah(struct ib_ah *ib_ah) +{ + struct bnxt_re_ah *ah = container_of(ib_ah, struct bnxt_re_ah, ib_ah); + struct bnxt_re_dev *rdev = ah->rdev; + int rc; + + rc = bnxt_qplib_destroy_ah(&rdev->qplib_res, &ah->qplib_ah); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to destroy HW AH"); + return rc; + } + kfree(ah); + return 0; +} + +struct ib_ah *bnxt_re_create_ah(struct ib_pd *ib_pd, + struct rdma_ah_attr *ah_attr, + struct ib_udata *udata) +{ + struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); + struct bnxt_re_dev *rdev = pd->rdev; + struct bnxt_re_ah *ah; + const struct ib_global_route *grh = rdma_ah_read_grh(ah_attr); + int rc; + u16 vlan_tag; + u8 nw_type; + + struct ib_gid_attr sgid_attr; + + if (!(rdma_ah_get_ah_flags(ah_attr) & IB_AH_GRH)) { + dev_err(rdev_to_dev(rdev), "Failed to alloc AH: GRH not set"); + return ERR_PTR(-EINVAL); + } + ah = kzalloc(sizeof(*ah), GFP_ATOMIC); + if (!ah) + return ERR_PTR(-ENOMEM); + + ah->rdev = rdev; + ah->qplib_ah.pd = &pd->qplib_pd; + + /* Supply the configuration for the HW */ + memcpy(ah->qplib_ah.dgid.data, grh->dgid.raw, + sizeof(union ib_gid)); + /* + * If RoCE V2 is enabled, stack will have two entries for + * each GID entry. Avoiding this duplicte entry in HW. Dividing + * the GID index by 2 for RoCE V2 + */ + ah->qplib_ah.sgid_index = grh->sgid_index / 2; + ah->qplib_ah.host_sgid_index = grh->sgid_index; + ah->qplib_ah.traffic_class = grh->traffic_class; + ah->qplib_ah.flow_label = grh->flow_label; + ah->qplib_ah.hop_limit = grh->hop_limit; + ah->qplib_ah.sl = rdma_ah_get_sl(ah_attr); + if (ib_pd->uobject && + !rdma_is_multicast_addr((struct in6_addr *) + grh->dgid.raw) && + !rdma_link_local_addr((struct in6_addr *) + grh->dgid.raw)) { + union ib_gid sgid; + + rc = ib_get_cached_gid(&rdev->ibdev, 1, + grh->sgid_index, &sgid, + &sgid_attr); + if (rc) { + dev_err(rdev_to_dev(rdev), + "Failed to query gid at index %d", + grh->sgid_index); + goto fail; + } + if (sgid_attr.ndev) { + if (is_vlan_dev(sgid_attr.ndev)) + vlan_tag = vlan_dev_vlan_id(sgid_attr.ndev); + dev_put(sgid_attr.ndev); + } + /* Get network header type for this GID */ + nw_type = ib_gid_to_network_type(sgid_attr.gid_type, &sgid); + switch (nw_type) { + case RDMA_NETWORK_IPV4: + ah->qplib_ah.nw_type = CMDQ_CREATE_AH_TYPE_V2IPV4; + break; + case RDMA_NETWORK_IPV6: + ah->qplib_ah.nw_type = CMDQ_CREATE_AH_TYPE_V2IPV6; + break; + default: + ah->qplib_ah.nw_type = CMDQ_CREATE_AH_TYPE_V1; + break; + } + rc = rdma_addr_find_l2_eth_by_grh(&sgid, &grh->dgid, + ah_attr->roce.dmac, &vlan_tag, + &sgid_attr.ndev->ifindex, + NULL); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to get dmac\n"); + goto fail; + } + } + + memcpy(ah->qplib_ah.dmac, ah_attr->roce.dmac, ETH_ALEN); + rc = bnxt_qplib_create_ah(&rdev->qplib_res, &ah->qplib_ah); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to allocate HW AH"); + goto fail; + } + + /* Write AVID to shared page. */ + if (ib_pd->uobject) { + struct ib_ucontext *ib_uctx = ib_pd->uobject->context; + struct bnxt_re_ucontext *uctx; + unsigned long flag; + u32 *wrptr; + + uctx = container_of(ib_uctx, struct bnxt_re_ucontext, ib_uctx); + spin_lock_irqsave(&uctx->sh_lock, flag); + wrptr = (u32 *)(uctx->shpg + BNXT_RE_AVID_OFFT); + *wrptr = ah->qplib_ah.id; + wmb(); /* make sure cache is updated. */ + spin_unlock_irqrestore(&uctx->sh_lock, flag); + } + + return &ah->ib_ah; + +fail: + kfree(ah); + return ERR_PTR(rc); +} + +int bnxt_re_modify_ah(struct ib_ah *ib_ah, struct rdma_ah_attr *ah_attr) +{ + return 0; +} + +int bnxt_re_query_ah(struct ib_ah *ib_ah, struct rdma_ah_attr *ah_attr) +{ + struct bnxt_re_ah *ah = container_of(ib_ah, struct bnxt_re_ah, ib_ah); + + ah_attr->type = ib_ah->type; + rdma_ah_set_sl(ah_attr, ah->qplib_ah.sl); + memcpy(ah_attr->roce.dmac, ah->qplib_ah.dmac, ETH_ALEN); + rdma_ah_set_grh(ah_attr, NULL, 0, + ah->qplib_ah.host_sgid_index, + 0, ah->qplib_ah.traffic_class); + rdma_ah_set_dgid_raw(ah_attr, ah->qplib_ah.dgid.data); + rdma_ah_set_port_num(ah_attr, 1); + rdma_ah_set_static_rate(ah_attr, 0); + return 0; +} + +/* Queue Pairs */ +int bnxt_re_destroy_qp(struct ib_qp *ib_qp) +{ + struct bnxt_re_qp *qp = container_of(ib_qp, struct bnxt_re_qp, ib_qp); + struct bnxt_re_dev *rdev = qp->rdev; + int rc; + + rc = bnxt_qplib_destroy_qp(&rdev->qplib_res, &qp->qplib_qp); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to destroy HW QP"); + return rc; + } + if (ib_qp->qp_type == IB_QPT_GSI && rdev->qp1_sqp) { + rc = bnxt_qplib_destroy_ah(&rdev->qplib_res, + &rdev->sqp_ah->qplib_ah); + if (rc) { + dev_err(rdev_to_dev(rdev), + "Failed to destroy HW AH for shadow QP"); + return rc; + } + + rc = bnxt_qplib_destroy_qp(&rdev->qplib_res, + &rdev->qp1_sqp->qplib_qp); + if (rc) { + dev_err(rdev_to_dev(rdev), + "Failed to destroy Shadow QP"); + return rc; + } + mutex_lock(&rdev->qp_lock); + list_del(&rdev->qp1_sqp->list); + atomic_dec(&rdev->qp_count); + mutex_unlock(&rdev->qp_lock); + + kfree(rdev->sqp_ah); + kfree(rdev->qp1_sqp); + } + + if (!IS_ERR_OR_NULL(qp->rumem)) + ib_umem_release(qp->rumem); + if (!IS_ERR_OR_NULL(qp->sumem)) + ib_umem_release(qp->sumem); + + mutex_lock(&rdev->qp_lock); + list_del(&qp->list); + atomic_dec(&rdev->qp_count); + mutex_unlock(&rdev->qp_lock); + kfree(qp); + return 0; +} + +static u8 __from_ib_qp_type(enum ib_qp_type type) +{ + switch (type) { + case IB_QPT_GSI: + return CMDQ_CREATE_QP1_TYPE_GSI; + case IB_QPT_RC: + return CMDQ_CREATE_QP_TYPE_RC; + case IB_QPT_UD: + return CMDQ_CREATE_QP_TYPE_UD; + default: + return IB_QPT_MAX; + } +} + +static int bnxt_re_init_user_qp(struct bnxt_re_dev *rdev, struct bnxt_re_pd *pd, + struct bnxt_re_qp *qp, struct ib_udata *udata) +{ + struct bnxt_re_qp_req ureq; + struct bnxt_qplib_qp *qplib_qp = &qp->qplib_qp; + struct ib_umem *umem; + int bytes = 0; + struct ib_ucontext *context = pd->ib_pd.uobject->context; + struct bnxt_re_ucontext *cntx = container_of(context, + struct bnxt_re_ucontext, + ib_uctx); + if (ib_copy_from_udata(&ureq, udata, sizeof(ureq))) + return -EFAULT; + + bytes = (qplib_qp->sq.max_wqe * BNXT_QPLIB_MAX_SQE_ENTRY_SIZE); + /* Consider mapping PSN search memory only for RC QPs. */ + if (qplib_qp->type == CMDQ_CREATE_QP_TYPE_RC) + bytes += (qplib_qp->sq.max_wqe * sizeof(struct sq_psn_search)); + bytes = PAGE_ALIGN(bytes); + umem = ib_umem_get(context, ureq.qpsva, bytes, + IB_ACCESS_LOCAL_WRITE, 1); + if (IS_ERR(umem)) + return PTR_ERR(umem); + + qp->sumem = umem; + qplib_qp->sq.sglist = umem->sg_head.sgl; + qplib_qp->sq.nmap = umem->nmap; + qplib_qp->qp_handle = ureq.qp_handle; + + if (!qp->qplib_qp.srq) { + bytes = (qplib_qp->rq.max_wqe * BNXT_QPLIB_MAX_RQE_ENTRY_SIZE); + bytes = PAGE_ALIGN(bytes); + umem = ib_umem_get(context, ureq.qprva, bytes, + IB_ACCESS_LOCAL_WRITE, 1); + if (IS_ERR(umem)) + goto rqfail; + qp->rumem = umem; + qplib_qp->rq.sglist = umem->sg_head.sgl; + qplib_qp->rq.nmap = umem->nmap; + } + + qplib_qp->dpi = &cntx->dpi; + return 0; +rqfail: + ib_umem_release(qp->sumem); + qp->sumem = NULL; + qplib_qp->sq.sglist = NULL; + qplib_qp->sq.nmap = 0; + + return PTR_ERR(umem); +} + +static struct bnxt_re_ah *bnxt_re_create_shadow_qp_ah + (struct bnxt_re_pd *pd, + struct bnxt_qplib_res *qp1_res, + struct bnxt_qplib_qp *qp1_qp) +{ + struct bnxt_re_dev *rdev = pd->rdev; + struct bnxt_re_ah *ah; + union ib_gid sgid; + int rc; + + ah = kzalloc(sizeof(*ah), GFP_KERNEL); + if (!ah) + return NULL; + + memset(ah, 0, sizeof(*ah)); + ah->rdev = rdev; + ah->qplib_ah.pd = &pd->qplib_pd; + + rc = bnxt_re_query_gid(&rdev->ibdev, 1, 0, &sgid); + if (rc) + goto fail; + + /* supply the dgid data same as sgid */ + memcpy(ah->qplib_ah.dgid.data, &sgid.raw, + sizeof(union ib_gid)); + ah->qplib_ah.sgid_index = 0; + + ah->qplib_ah.traffic_class = 0; + ah->qplib_ah.flow_label = 0; + ah->qplib_ah.hop_limit = 1; + ah->qplib_ah.sl = 0; + /* Have DMAC same as SMAC */ + ether_addr_copy(ah->qplib_ah.dmac, rdev->netdev->dev_addr); + + rc = bnxt_qplib_create_ah(&rdev->qplib_res, &ah->qplib_ah); + if (rc) { + dev_err(rdev_to_dev(rdev), + "Failed to allocate HW AH for Shadow QP"); + goto fail; + } + + return ah; + +fail: + kfree(ah); + return NULL; +} + +static struct bnxt_re_qp *bnxt_re_create_shadow_qp + (struct bnxt_re_pd *pd, + struct bnxt_qplib_res *qp1_res, + struct bnxt_qplib_qp *qp1_qp) +{ + struct bnxt_re_dev *rdev = pd->rdev; + struct bnxt_re_qp *qp; + int rc; + + qp = kzalloc(sizeof(*qp), GFP_KERNEL); + if (!qp) + return NULL; + + memset(qp, 0, sizeof(*qp)); + qp->rdev = rdev; + + /* Initialize the shadow QP structure from the QP1 values */ + ether_addr_copy(qp->qplib_qp.smac, rdev->netdev->dev_addr); + + qp->qplib_qp.pd = &pd->qplib_pd; + qp->qplib_qp.qp_handle = (u64)(unsigned long)(&qp->qplib_qp); + qp->qplib_qp.type = IB_QPT_UD; + + qp->qplib_qp.max_inline_data = 0; + qp->qplib_qp.sig_type = true; + + /* Shadow QP SQ depth should be same as QP1 RQ depth */ + qp->qplib_qp.sq.max_wqe = qp1_qp->rq.max_wqe; + qp->qplib_qp.sq.max_sge = 2; + /* Q full delta can be 1 since it is internal QP */ + qp->qplib_qp.sq.q_full_delta = 1; + + qp->qplib_qp.scq = qp1_qp->scq; + qp->qplib_qp.rcq = qp1_qp->rcq; + + qp->qplib_qp.rq.max_wqe = qp1_qp->rq.max_wqe; + qp->qplib_qp.rq.max_sge = qp1_qp->rq.max_sge; + /* Q full delta can be 1 since it is internal QP */ + qp->qplib_qp.rq.q_full_delta = 1; + + qp->qplib_qp.mtu = qp1_qp->mtu; + + qp->qplib_qp.sq_hdr_buf_size = 0; + qp->qplib_qp.rq_hdr_buf_size = BNXT_QPLIB_MAX_GRH_HDR_SIZE_IPV6; + qp->qplib_qp.dpi = &rdev->dpi_privileged; + + rc = bnxt_qplib_create_qp(qp1_res, &qp->qplib_qp); + if (rc) + goto fail; + + rdev->sqp_id = qp->qplib_qp.id; + + spin_lock_init(&qp->sq_lock); + INIT_LIST_HEAD(&qp->list); + mutex_lock(&rdev->qp_lock); + list_add_tail(&qp->list, &rdev->qp_list); + atomic_inc(&rdev->qp_count); + mutex_unlock(&rdev->qp_lock); + return qp; +fail: + kfree(qp); + return NULL; +} + +struct ib_qp *bnxt_re_create_qp(struct ib_pd *ib_pd, + struct ib_qp_init_attr *qp_init_attr, + struct ib_udata *udata) +{ + struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); + struct bnxt_re_dev *rdev = pd->rdev; + struct bnxt_qplib_dev_attr *dev_attr = &rdev->dev_attr; + struct bnxt_re_qp *qp; + struct bnxt_re_cq *cq; + int rc, entries; + + if ((qp_init_attr->cap.max_send_wr > dev_attr->max_qp_wqes) || + (qp_init_attr->cap.max_recv_wr > dev_attr->max_qp_wqes) || + (qp_init_attr->cap.max_send_sge > dev_attr->max_qp_sges) || + (qp_init_attr->cap.max_recv_sge > dev_attr->max_qp_sges) || + (qp_init_attr->cap.max_inline_data > dev_attr->max_inline_data)) + return ERR_PTR(-EINVAL); + + qp = kzalloc(sizeof(*qp), GFP_KERNEL); + if (!qp) + return ERR_PTR(-ENOMEM); + + qp->rdev = rdev; + ether_addr_copy(qp->qplib_qp.smac, rdev->netdev->dev_addr); + qp->qplib_qp.pd = &pd->qplib_pd; + qp->qplib_qp.qp_handle = (u64)(unsigned long)(&qp->qplib_qp); + qp->qplib_qp.type = __from_ib_qp_type(qp_init_attr->qp_type); + if (qp->qplib_qp.type == IB_QPT_MAX) { + dev_err(rdev_to_dev(rdev), "QP type 0x%x not supported", + qp->qplib_qp.type); + rc = -EINVAL; + goto fail; + } + qp->qplib_qp.max_inline_data = qp_init_attr->cap.max_inline_data; + qp->qplib_qp.sig_type = ((qp_init_attr->sq_sig_type == + IB_SIGNAL_ALL_WR) ? true : false); + + qp->qplib_qp.sq.max_sge = qp_init_attr->cap.max_send_sge; + if (qp->qplib_qp.sq.max_sge > dev_attr->max_qp_sges) + qp->qplib_qp.sq.max_sge = dev_attr->max_qp_sges; + + if (qp_init_attr->send_cq) { + cq = container_of(qp_init_attr->send_cq, struct bnxt_re_cq, + ib_cq); + if (!cq) { + dev_err(rdev_to_dev(rdev), "Send CQ not found"); + rc = -EINVAL; + goto fail; + } + qp->qplib_qp.scq = &cq->qplib_cq; + } + + if (qp_init_attr->recv_cq) { + cq = container_of(qp_init_attr->recv_cq, struct bnxt_re_cq, + ib_cq); + if (!cq) { + dev_err(rdev_to_dev(rdev), "Receive CQ not found"); + rc = -EINVAL; + goto fail; + } + qp->qplib_qp.rcq = &cq->qplib_cq; + } + + if (qp_init_attr->srq) { + dev_err(rdev_to_dev(rdev), "SRQ not supported"); + rc = -ENOTSUPP; + goto fail; + } else { + /* Allocate 1 more than what's provided so posting max doesn't + * mean empty + */ + entries = roundup_pow_of_two(qp_init_attr->cap.max_recv_wr + 1); + qp->qplib_qp.rq.max_wqe = min_t(u32, entries, + dev_attr->max_qp_wqes + 1); + + qp->qplib_qp.rq.q_full_delta = qp->qplib_qp.rq.max_wqe - + qp_init_attr->cap.max_recv_wr; + + qp->qplib_qp.rq.max_sge = qp_init_attr->cap.max_recv_sge; + if (qp->qplib_qp.rq.max_sge > dev_attr->max_qp_sges) + qp->qplib_qp.rq.max_sge = dev_attr->max_qp_sges; + } + + qp->qplib_qp.mtu = ib_mtu_enum_to_int(iboe_get_mtu(rdev->netdev->mtu)); + + if (qp_init_attr->qp_type == IB_QPT_GSI) { + /* Allocate 1 more than what's provided */ + entries = roundup_pow_of_two(qp_init_attr->cap.max_send_wr + 1); + qp->qplib_qp.sq.max_wqe = min_t(u32, entries, + dev_attr->max_qp_wqes + 1); + qp->qplib_qp.sq.q_full_delta = qp->qplib_qp.sq.max_wqe - + qp_init_attr->cap.max_send_wr; + qp->qplib_qp.rq.max_sge = dev_attr->max_qp_sges; + if (qp->qplib_qp.rq.max_sge > dev_attr->max_qp_sges) + qp->qplib_qp.rq.max_sge = dev_attr->max_qp_sges; + qp->qplib_qp.sq.max_sge++; + if (qp->qplib_qp.sq.max_sge > dev_attr->max_qp_sges) + qp->qplib_qp.sq.max_sge = dev_attr->max_qp_sges; + + qp->qplib_qp.rq_hdr_buf_size = + BNXT_QPLIB_MAX_QP1_RQ_HDR_SIZE_V2; + + qp->qplib_qp.sq_hdr_buf_size = + BNXT_QPLIB_MAX_QP1_SQ_HDR_SIZE_V2; + qp->qplib_qp.dpi = &rdev->dpi_privileged; + rc = bnxt_qplib_create_qp1(&rdev->qplib_res, &qp->qplib_qp); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to create HW QP1"); + goto fail; + } + /* Create a shadow QP to handle the QP1 traffic */ + rdev->qp1_sqp = bnxt_re_create_shadow_qp(pd, &rdev->qplib_res, + &qp->qplib_qp); + if (!rdev->qp1_sqp) { + rc = -EINVAL; + dev_err(rdev_to_dev(rdev), + "Failed to create Shadow QP for QP1"); + goto qp_destroy; + } + rdev->sqp_ah = bnxt_re_create_shadow_qp_ah(pd, &rdev->qplib_res, + &qp->qplib_qp); + if (!rdev->sqp_ah) { + bnxt_qplib_destroy_qp(&rdev->qplib_res, + &rdev->qp1_sqp->qplib_qp); + rc = -EINVAL; + dev_err(rdev_to_dev(rdev), + "Failed to create AH entry for ShadowQP"); + goto qp_destroy; + } + + } else { + /* Allocate 128 + 1 more than what's provided */ + entries = roundup_pow_of_two(qp_init_attr->cap.max_send_wr + + BNXT_QPLIB_RESERVED_QP_WRS + 1); + qp->qplib_qp.sq.max_wqe = min_t(u32, entries, + dev_attr->max_qp_wqes + + BNXT_QPLIB_RESERVED_QP_WRS + 1); + qp->qplib_qp.sq.q_full_delta = BNXT_QPLIB_RESERVED_QP_WRS + 1; + + /* + * Reserving one slot for Phantom WQE. Application can + * post one extra entry in this case. But allowing this to avoid + * unexpected Queue full condition + */ + + qp->qplib_qp.sq.q_full_delta -= 1; + + qp->qplib_qp.max_rd_atomic = dev_attr->max_qp_rd_atom; + qp->qplib_qp.max_dest_rd_atomic = dev_attr->max_qp_init_rd_atom; + if (udata) { + rc = bnxt_re_init_user_qp(rdev, pd, qp, udata); + if (rc) + goto fail; + } else { + qp->qplib_qp.dpi = &rdev->dpi_privileged; + } + + rc = bnxt_qplib_create_qp(&rdev->qplib_res, &qp->qplib_qp); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to create HW QP"); + goto fail; + } + } + + qp->ib_qp.qp_num = qp->qplib_qp.id; + spin_lock_init(&qp->sq_lock); + spin_lock_init(&qp->rq_lock); + + if (udata) { + struct bnxt_re_qp_resp resp; + + resp.qpid = qp->ib_qp.qp_num; + resp.rsvd = 0; + rc = ib_copy_to_udata(udata, &resp, sizeof(resp)); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to copy QP udata"); + goto qp_destroy; + } + } + INIT_LIST_HEAD(&qp->list); + mutex_lock(&rdev->qp_lock); + list_add_tail(&qp->list, &rdev->qp_list); + atomic_inc(&rdev->qp_count); + mutex_unlock(&rdev->qp_lock); + + return &qp->ib_qp; +qp_destroy: + bnxt_qplib_destroy_qp(&rdev->qplib_res, &qp->qplib_qp); +fail: + kfree(qp); + return ERR_PTR(rc); +} + +static u8 __from_ib_qp_state(enum ib_qp_state state) +{ + switch (state) { + case IB_QPS_RESET: + return CMDQ_MODIFY_QP_NEW_STATE_RESET; + case IB_QPS_INIT: + return CMDQ_MODIFY_QP_NEW_STATE_INIT; + case IB_QPS_RTR: + return CMDQ_MODIFY_QP_NEW_STATE_RTR; + case IB_QPS_RTS: + return CMDQ_MODIFY_QP_NEW_STATE_RTS; + case IB_QPS_SQD: + return CMDQ_MODIFY_QP_NEW_STATE_SQD; + case IB_QPS_SQE: + return CMDQ_MODIFY_QP_NEW_STATE_SQE; + case IB_QPS_ERR: + default: + return CMDQ_MODIFY_QP_NEW_STATE_ERR; + } +} + +static enum ib_qp_state __to_ib_qp_state(u8 state) +{ + switch (state) { + case CMDQ_MODIFY_QP_NEW_STATE_RESET: + return IB_QPS_RESET; + case CMDQ_MODIFY_QP_NEW_STATE_INIT: + return IB_QPS_INIT; + case CMDQ_MODIFY_QP_NEW_STATE_RTR: + return IB_QPS_RTR; + case CMDQ_MODIFY_QP_NEW_STATE_RTS: + return IB_QPS_RTS; + case CMDQ_MODIFY_QP_NEW_STATE_SQD: + return IB_QPS_SQD; + case CMDQ_MODIFY_QP_NEW_STATE_SQE: + return IB_QPS_SQE; + case CMDQ_MODIFY_QP_NEW_STATE_ERR: + default: + return IB_QPS_ERR; + } +} + +static u32 __from_ib_mtu(enum ib_mtu mtu) +{ + switch (mtu) { + case IB_MTU_256: + return CMDQ_MODIFY_QP_PATH_MTU_MTU_256; + case IB_MTU_512: + return CMDQ_MODIFY_QP_PATH_MTU_MTU_512; + case IB_MTU_1024: + return CMDQ_MODIFY_QP_PATH_MTU_MTU_1024; + case IB_MTU_2048: + return CMDQ_MODIFY_QP_PATH_MTU_MTU_2048; + case IB_MTU_4096: + return CMDQ_MODIFY_QP_PATH_MTU_MTU_4096; + default: + return CMDQ_MODIFY_QP_PATH_MTU_MTU_2048; + } +} + +static enum ib_mtu __to_ib_mtu(u32 mtu) +{ + switch (mtu & CREQ_QUERY_QP_RESP_SB_PATH_MTU_MASK) { + case CMDQ_MODIFY_QP_PATH_MTU_MTU_256: + return IB_MTU_256; + case CMDQ_MODIFY_QP_PATH_MTU_MTU_512: + return IB_MTU_512; + case CMDQ_MODIFY_QP_PATH_MTU_MTU_1024: + return IB_MTU_1024; + case CMDQ_MODIFY_QP_PATH_MTU_MTU_2048: + return IB_MTU_2048; + case CMDQ_MODIFY_QP_PATH_MTU_MTU_4096: + return IB_MTU_4096; + default: + return IB_MTU_2048; + } +} + +static int bnxt_re_modify_shadow_qp(struct bnxt_re_dev *rdev, + struct bnxt_re_qp *qp1_qp, + int qp_attr_mask) +{ + struct bnxt_re_qp *qp = rdev->qp1_sqp; + int rc = 0; + + if (qp_attr_mask & IB_QP_STATE) { + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_STATE; + qp->qplib_qp.state = qp1_qp->qplib_qp.state; + } + if (qp_attr_mask & IB_QP_PKEY_INDEX) { + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_PKEY; + qp->qplib_qp.pkey_index = qp1_qp->qplib_qp.pkey_index; + } + + if (qp_attr_mask & IB_QP_QKEY) { + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_QKEY; + /* Using a Random QKEY */ + qp->qplib_qp.qkey = 0x81818181; + } + if (qp_attr_mask & IB_QP_SQ_PSN) { + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_SQ_PSN; + qp->qplib_qp.sq.psn = qp1_qp->qplib_qp.sq.psn; + } + + rc = bnxt_qplib_modify_qp(&rdev->qplib_res, &qp->qplib_qp); + if (rc) + dev_err(rdev_to_dev(rdev), + "Failed to modify Shadow QP for QP1"); + return rc; +} + +int bnxt_re_modify_qp(struct ib_qp *ib_qp, struct ib_qp_attr *qp_attr, + int qp_attr_mask, struct ib_udata *udata) +{ + struct bnxt_re_qp *qp = container_of(ib_qp, struct bnxt_re_qp, ib_qp); + struct bnxt_re_dev *rdev = qp->rdev; + struct bnxt_qplib_dev_attr *dev_attr = &rdev->dev_attr; + enum ib_qp_state curr_qp_state, new_qp_state; + int rc, entries; + int status; + union ib_gid sgid; + struct ib_gid_attr sgid_attr; + u8 nw_type; + + qp->qplib_qp.modify_flags = 0; + if (qp_attr_mask & IB_QP_STATE) { + curr_qp_state = __to_ib_qp_state(qp->qplib_qp.cur_qp_state); + new_qp_state = qp_attr->qp_state; + if (!ib_modify_qp_is_ok(curr_qp_state, new_qp_state, + ib_qp->qp_type, qp_attr_mask, + IB_LINK_LAYER_ETHERNET)) { + dev_err(rdev_to_dev(rdev), + "Invalid attribute mask: %#x specified ", + qp_attr_mask); + dev_err(rdev_to_dev(rdev), + "for qpn: %#x type: %#x", + ib_qp->qp_num, ib_qp->qp_type); + dev_err(rdev_to_dev(rdev), + "curr_qp_state=0x%x, new_qp_state=0x%x\n", + curr_qp_state, new_qp_state); + return -EINVAL; + } + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_STATE; + qp->qplib_qp.state = __from_ib_qp_state(qp_attr->qp_state); + } + if (qp_attr_mask & IB_QP_EN_SQD_ASYNC_NOTIFY) { + qp->qplib_qp.modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_EN_SQD_ASYNC_NOTIFY; + qp->qplib_qp.en_sqd_async_notify = true; + } + if (qp_attr_mask & IB_QP_ACCESS_FLAGS) { + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_ACCESS; + qp->qplib_qp.access = + __from_ib_access_flags(qp_attr->qp_access_flags); + /* LOCAL_WRITE access must be set to allow RC receive */ + qp->qplib_qp.access |= BNXT_QPLIB_ACCESS_LOCAL_WRITE; + } + if (qp_attr_mask & IB_QP_PKEY_INDEX) { + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_PKEY; + qp->qplib_qp.pkey_index = qp_attr->pkey_index; + } + if (qp_attr_mask & IB_QP_QKEY) { + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_QKEY; + qp->qplib_qp.qkey = qp_attr->qkey; + } + if (qp_attr_mask & IB_QP_AV) { + const struct ib_global_route *grh = + rdma_ah_read_grh(&qp_attr->ah_attr); + + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_DGID | + CMDQ_MODIFY_QP_MODIFY_MASK_FLOW_LABEL | + CMDQ_MODIFY_QP_MODIFY_MASK_SGID_INDEX | + CMDQ_MODIFY_QP_MODIFY_MASK_HOP_LIMIT | + CMDQ_MODIFY_QP_MODIFY_MASK_TRAFFIC_CLASS | + CMDQ_MODIFY_QP_MODIFY_MASK_DEST_MAC | + CMDQ_MODIFY_QP_MODIFY_MASK_VLAN_ID; + memcpy(qp->qplib_qp.ah.dgid.data, grh->dgid.raw, + sizeof(qp->qplib_qp.ah.dgid.data)); + qp->qplib_qp.ah.flow_label = grh->flow_label; + /* If RoCE V2 is enabled, stack will have two entries for + * each GID entry. Avoiding this duplicte entry in HW. Dividing + * the GID index by 2 for RoCE V2 + */ + qp->qplib_qp.ah.sgid_index = grh->sgid_index / 2; + qp->qplib_qp.ah.host_sgid_index = grh->sgid_index; + qp->qplib_qp.ah.hop_limit = grh->hop_limit; + qp->qplib_qp.ah.traffic_class = grh->traffic_class; + qp->qplib_qp.ah.sl = rdma_ah_get_sl(&qp_attr->ah_attr); + ether_addr_copy(qp->qplib_qp.ah.dmac, + qp_attr->ah_attr.roce.dmac); + + status = ib_get_cached_gid(&rdev->ibdev, 1, + grh->sgid_index, + &sgid, &sgid_attr); + if (!status && sgid_attr.ndev) { + memcpy(qp->qplib_qp.smac, sgid_attr.ndev->dev_addr, + ETH_ALEN); + dev_put(sgid_attr.ndev); + nw_type = ib_gid_to_network_type(sgid_attr.gid_type, + &sgid); + switch (nw_type) { + case RDMA_NETWORK_IPV4: + qp->qplib_qp.nw_type = + CMDQ_MODIFY_QP_NETWORK_TYPE_ROCEV2_IPV4; + break; + case RDMA_NETWORK_IPV6: + qp->qplib_qp.nw_type = + CMDQ_MODIFY_QP_NETWORK_TYPE_ROCEV2_IPV6; + break; + default: + qp->qplib_qp.nw_type = + CMDQ_MODIFY_QP_NETWORK_TYPE_ROCEV1; + break; + } + } + } + + if (qp_attr_mask & IB_QP_PATH_MTU) { + qp->qplib_qp.modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_PATH_MTU; + qp->qplib_qp.path_mtu = __from_ib_mtu(qp_attr->path_mtu); + } else if (qp_attr->qp_state == IB_QPS_RTR) { + qp->qplib_qp.modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_PATH_MTU; + qp->qplib_qp.path_mtu = + __from_ib_mtu(iboe_get_mtu(rdev->netdev->mtu)); + } + + if (qp_attr_mask & IB_QP_TIMEOUT) { + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_TIMEOUT; + qp->qplib_qp.timeout = qp_attr->timeout; + } + if (qp_attr_mask & IB_QP_RETRY_CNT) { + qp->qplib_qp.modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_RETRY_CNT; + qp->qplib_qp.retry_cnt = qp_attr->retry_cnt; + } + if (qp_attr_mask & IB_QP_RNR_RETRY) { + qp->qplib_qp.modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_RNR_RETRY; + qp->qplib_qp.rnr_retry = qp_attr->rnr_retry; + } + if (qp_attr_mask & IB_QP_MIN_RNR_TIMER) { + qp->qplib_qp.modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_MIN_RNR_TIMER; + qp->qplib_qp.min_rnr_timer = qp_attr->min_rnr_timer; + } + if (qp_attr_mask & IB_QP_RQ_PSN) { + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_RQ_PSN; + qp->qplib_qp.rq.psn = qp_attr->rq_psn; + } + if (qp_attr_mask & IB_QP_MAX_QP_RD_ATOMIC) { + qp->qplib_qp.modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_MAX_RD_ATOMIC; + /* Cap the max_rd_atomic to device max */ + qp->qplib_qp.max_rd_atomic = min_t(u32, qp_attr->max_rd_atomic, + dev_attr->max_qp_rd_atom); + } + if (qp_attr_mask & IB_QP_SQ_PSN) { + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_SQ_PSN; + qp->qplib_qp.sq.psn = qp_attr->sq_psn; + } + if (qp_attr_mask & IB_QP_MAX_DEST_RD_ATOMIC) { + if (qp_attr->max_dest_rd_atomic > + dev_attr->max_qp_init_rd_atom) { + dev_err(rdev_to_dev(rdev), + "max_dest_rd_atomic requested%d is > dev_max%d", + qp_attr->max_dest_rd_atomic, + dev_attr->max_qp_init_rd_atom); + return -EINVAL; + } + + qp->qplib_qp.modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_MAX_DEST_RD_ATOMIC; + qp->qplib_qp.max_dest_rd_atomic = qp_attr->max_dest_rd_atomic; + } + if (qp_attr_mask & IB_QP_CAP) { + qp->qplib_qp.modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_SQ_SIZE | + CMDQ_MODIFY_QP_MODIFY_MASK_RQ_SIZE | + CMDQ_MODIFY_QP_MODIFY_MASK_SQ_SGE | + CMDQ_MODIFY_QP_MODIFY_MASK_RQ_SGE | + CMDQ_MODIFY_QP_MODIFY_MASK_MAX_INLINE_DATA; + if ((qp_attr->cap.max_send_wr >= dev_attr->max_qp_wqes) || + (qp_attr->cap.max_recv_wr >= dev_attr->max_qp_wqes) || + (qp_attr->cap.max_send_sge >= dev_attr->max_qp_sges) || + (qp_attr->cap.max_recv_sge >= dev_attr->max_qp_sges) || + (qp_attr->cap.max_inline_data >= + dev_attr->max_inline_data)) { + dev_err(rdev_to_dev(rdev), + "Create QP failed - max exceeded"); + return -EINVAL; + } + entries = roundup_pow_of_two(qp_attr->cap.max_send_wr); + qp->qplib_qp.sq.max_wqe = min_t(u32, entries, + dev_attr->max_qp_wqes + 1); + qp->qplib_qp.sq.q_full_delta = qp->qplib_qp.sq.max_wqe - + qp_attr->cap.max_send_wr; + /* + * Reserving one slot for Phantom WQE. Some application can + * post one extra entry in this case. Allowing this to avoid + * unexpected Queue full condition + */ + qp->qplib_qp.sq.q_full_delta -= 1; + qp->qplib_qp.sq.max_sge = qp_attr->cap.max_send_sge; + if (qp->qplib_qp.rq.max_wqe) { + entries = roundup_pow_of_two(qp_attr->cap.max_recv_wr); + qp->qplib_qp.rq.max_wqe = + min_t(u32, entries, dev_attr->max_qp_wqes + 1); + qp->qplib_qp.rq.q_full_delta = qp->qplib_qp.rq.max_wqe - + qp_attr->cap.max_recv_wr; + qp->qplib_qp.rq.max_sge = qp_attr->cap.max_recv_sge; + } else { + /* SRQ was used prior, just ignore the RQ caps */ + } + } + if (qp_attr_mask & IB_QP_DEST_QPN) { + qp->qplib_qp.modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_DEST_QP_ID; + qp->qplib_qp.dest_qpn = qp_attr->dest_qp_num; + } + rc = bnxt_qplib_modify_qp(&rdev->qplib_res, &qp->qplib_qp); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to modify HW QP"); + return rc; + } + if (ib_qp->qp_type == IB_QPT_GSI && rdev->qp1_sqp) + rc = bnxt_re_modify_shadow_qp(rdev, qp, qp_attr_mask); + return rc; +} + +int bnxt_re_query_qp(struct ib_qp *ib_qp, struct ib_qp_attr *qp_attr, + int qp_attr_mask, struct ib_qp_init_attr *qp_init_attr) +{ + struct bnxt_re_qp *qp = container_of(ib_qp, struct bnxt_re_qp, ib_qp); + struct bnxt_re_dev *rdev = qp->rdev; + struct bnxt_qplib_qp qplib_qp; + int rc; + + memset(&qplib_qp, 0, sizeof(struct bnxt_qplib_qp)); + qplib_qp.id = qp->qplib_qp.id; + qplib_qp.ah.host_sgid_index = qp->qplib_qp.ah.host_sgid_index; + + rc = bnxt_qplib_query_qp(&rdev->qplib_res, &qplib_qp); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to query HW QP"); + return rc; + } + qp_attr->qp_state = __to_ib_qp_state(qplib_qp.state); + qp_attr->en_sqd_async_notify = qplib_qp.en_sqd_async_notify ? 1 : 0; + qp_attr->qp_access_flags = __to_ib_access_flags(qplib_qp.access); + qp_attr->pkey_index = qplib_qp.pkey_index; + qp_attr->qkey = qplib_qp.qkey; + qp_attr->ah_attr.type = RDMA_AH_ATTR_TYPE_ROCE; + rdma_ah_set_grh(&qp_attr->ah_attr, NULL, qplib_qp.ah.flow_label, + qplib_qp.ah.host_sgid_index, + qplib_qp.ah.hop_limit, + qplib_qp.ah.traffic_class); + rdma_ah_set_dgid_raw(&qp_attr->ah_attr, qplib_qp.ah.dgid.data); + rdma_ah_set_sl(&qp_attr->ah_attr, qplib_qp.ah.sl); + ether_addr_copy(qp_attr->ah_attr.roce.dmac, qplib_qp.ah.dmac); + qp_attr->path_mtu = __to_ib_mtu(qplib_qp.path_mtu); + qp_attr->timeout = qplib_qp.timeout; + qp_attr->retry_cnt = qplib_qp.retry_cnt; + qp_attr->rnr_retry = qplib_qp.rnr_retry; + qp_attr->min_rnr_timer = qplib_qp.min_rnr_timer; + qp_attr->rq_psn = qplib_qp.rq.psn; + qp_attr->max_rd_atomic = qplib_qp.max_rd_atomic; + qp_attr->sq_psn = qplib_qp.sq.psn; + qp_attr->max_dest_rd_atomic = qplib_qp.max_dest_rd_atomic; + qp_init_attr->sq_sig_type = qplib_qp.sig_type ? IB_SIGNAL_ALL_WR : + IB_SIGNAL_REQ_WR; + qp_attr->dest_qp_num = qplib_qp.dest_qpn; + + qp_attr->cap.max_send_wr = qp->qplib_qp.sq.max_wqe; + qp_attr->cap.max_send_sge = qp->qplib_qp.sq.max_sge; + qp_attr->cap.max_recv_wr = qp->qplib_qp.rq.max_wqe; + qp_attr->cap.max_recv_sge = qp->qplib_qp.rq.max_sge; + qp_attr->cap.max_inline_data = qp->qplib_qp.max_inline_data; + qp_init_attr->cap = qp_attr->cap; + + return 0; +} + +/* Routine for sending QP1 packets for RoCE V1 an V2 + */ +static int bnxt_re_build_qp1_send_v2(struct bnxt_re_qp *qp, + struct ib_send_wr *wr, + struct bnxt_qplib_swqe *wqe, + int payload_size) +{ + struct ib_device *ibdev = &qp->rdev->ibdev; + struct bnxt_re_ah *ah = container_of(ud_wr(wr)->ah, struct bnxt_re_ah, + ib_ah); + struct bnxt_qplib_ah *qplib_ah = &ah->qplib_ah; + struct bnxt_qplib_sge sge; + union ib_gid sgid; + u8 nw_type; + u16 ether_type; + struct ib_gid_attr sgid_attr; + union ib_gid dgid; + bool is_eth = false; + bool is_vlan = false; + bool is_grh = false; + bool is_udp = false; + u8 ip_version = 0; + u16 vlan_id = 0xFFFF; + void *buf; + int i, rc = 0, size; + + memset(&qp->qp1_hdr, 0, sizeof(qp->qp1_hdr)); + + rc = ib_get_cached_gid(ibdev, 1, + qplib_ah->host_sgid_index, &sgid, + &sgid_attr); + if (rc) { + dev_err(rdev_to_dev(qp->rdev), + "Failed to query gid at index %d", + qplib_ah->host_sgid_index); + return rc; + } + if (sgid_attr.ndev) { + if (is_vlan_dev(sgid_attr.ndev)) + vlan_id = vlan_dev_vlan_id(sgid_attr.ndev); + dev_put(sgid_attr.ndev); + } + /* Get network header type for this GID */ + nw_type = ib_gid_to_network_type(sgid_attr.gid_type, &sgid); + switch (nw_type) { + case RDMA_NETWORK_IPV4: + nw_type = BNXT_RE_ROCEV2_IPV4_PACKET; + break; + case RDMA_NETWORK_IPV6: + nw_type = BNXT_RE_ROCEV2_IPV6_PACKET; + break; + default: + nw_type = BNXT_RE_ROCE_V1_PACKET; + break; + } + memcpy(&dgid.raw, &qplib_ah->dgid, 16); + is_udp = sgid_attr.gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP; + if (is_udp) { + if (ipv6_addr_v4mapped((struct in6_addr *)&sgid)) { + ip_version = 4; + ether_type = ETH_P_IP; + } else { + ip_version = 6; + ether_type = ETH_P_IPV6; + } + is_grh = false; + } else { + ether_type = ETH_P_IBOE; + is_grh = true; + } + + is_eth = true; + is_vlan = (vlan_id && (vlan_id < 0x1000)) ? true : false; + + ib_ud_header_init(payload_size, !is_eth, is_eth, is_vlan, is_grh, + ip_version, is_udp, 0, &qp->qp1_hdr); + + /* ETH */ + ether_addr_copy(qp->qp1_hdr.eth.dmac_h, ah->qplib_ah.dmac); + ether_addr_copy(qp->qp1_hdr.eth.smac_h, qp->qplib_qp.smac); + + /* For vlan, check the sgid for vlan existence */ + + if (!is_vlan) { + qp->qp1_hdr.eth.type = cpu_to_be16(ether_type); + } else { + qp->qp1_hdr.vlan.type = cpu_to_be16(ether_type); + qp->qp1_hdr.vlan.tag = cpu_to_be16(vlan_id); + } + + if (is_grh || (ip_version == 6)) { + memcpy(qp->qp1_hdr.grh.source_gid.raw, sgid.raw, sizeof(sgid)); + memcpy(qp->qp1_hdr.grh.destination_gid.raw, qplib_ah->dgid.data, + sizeof(sgid)); + qp->qp1_hdr.grh.hop_limit = qplib_ah->hop_limit; + } + + if (ip_version == 4) { + qp->qp1_hdr.ip4.tos = 0; + qp->qp1_hdr.ip4.id = 0; + qp->qp1_hdr.ip4.frag_off = htons(IP_DF); + qp->qp1_hdr.ip4.ttl = qplib_ah->hop_limit; + + memcpy(&qp->qp1_hdr.ip4.saddr, sgid.raw + 12, 4); + memcpy(&qp->qp1_hdr.ip4.daddr, qplib_ah->dgid.data + 12, 4); + qp->qp1_hdr.ip4.check = ib_ud_ip4_csum(&qp->qp1_hdr); + } + + if (is_udp) { + qp->qp1_hdr.udp.dport = htons(ROCE_V2_UDP_DPORT); + qp->qp1_hdr.udp.sport = htons(0x8CD1); + qp->qp1_hdr.udp.csum = 0; + } + + /* BTH */ + if (wr->opcode == IB_WR_SEND_WITH_IMM) { + qp->qp1_hdr.bth.opcode = IB_OPCODE_UD_SEND_ONLY_WITH_IMMEDIATE; + qp->qp1_hdr.immediate_present = 1; + } else { + qp->qp1_hdr.bth.opcode = IB_OPCODE_UD_SEND_ONLY; + } + if (wr->send_flags & IB_SEND_SOLICITED) + qp->qp1_hdr.bth.solicited_event = 1; + /* pad_count */ + qp->qp1_hdr.bth.pad_count = (4 - payload_size) & 3; + + /* P_key for QP1 is for all members */ + qp->qp1_hdr.bth.pkey = cpu_to_be16(0xFFFF); + qp->qp1_hdr.bth.destination_qpn = IB_QP1; + qp->qp1_hdr.bth.ack_req = 0; + qp->send_psn++; + qp->send_psn &= BTH_PSN_MASK; + qp->qp1_hdr.bth.psn = cpu_to_be32(qp->send_psn); + /* DETH */ + /* Use the priviledged Q_Key for QP1 */ + qp->qp1_hdr.deth.qkey = cpu_to_be32(IB_QP1_QKEY); + qp->qp1_hdr.deth.source_qpn = IB_QP1; + + /* Pack the QP1 to the transmit buffer */ + buf = bnxt_qplib_get_qp1_sq_buf(&qp->qplib_qp, &sge); + if (buf) { + size = ib_ud_header_pack(&qp->qp1_hdr, buf); + for (i = wqe->num_sge; i; i--) { + wqe->sg_list[i].addr = wqe->sg_list[i - 1].addr; + wqe->sg_list[i].lkey = wqe->sg_list[i - 1].lkey; + wqe->sg_list[i].size = wqe->sg_list[i - 1].size; + } + + /* + * Max Header buf size for IPV6 RoCE V2 is 86, + * which is same as the QP1 SQ header buffer. + * Header buf size for IPV4 RoCE V2 can be 66. + * ETH(14) + VLAN(4)+ IP(20) + UDP (8) + BTH(20). + * Subtract 20 bytes from QP1 SQ header buf size + */ + if (is_udp && ip_version == 4) + sge.size -= 20; + /* + * Max Header buf size for RoCE V1 is 78. + * ETH(14) + VLAN(4) + GRH(40) + BTH(20). + * Subtract 8 bytes from QP1 SQ header buf size + */ + if (!is_udp) + sge.size -= 8; + + /* Subtract 4 bytes for non vlan packets */ + if (!is_vlan) + sge.size -= 4; + + wqe->sg_list[0].addr = sge.addr; + wqe->sg_list[0].lkey = sge.lkey; + wqe->sg_list[0].size = sge.size; + wqe->num_sge++; + + } else { + dev_err(rdev_to_dev(qp->rdev), "QP1 buffer is empty!"); + rc = -ENOMEM; + } + return rc; +} + +/* For the MAD layer, it only provides the recv SGE the size of + * ib_grh + MAD datagram. No Ethernet headers, Ethertype, BTH, DETH, + * nor RoCE iCRC. The Cu+ solution must provide buffer for the entire + * receive packet (334 bytes) with no VLAN and then copy the GRH + * and the MAD datagram out to the provided SGE. + */ +static int bnxt_re_build_qp1_shadow_qp_recv(struct bnxt_re_qp *qp, + struct ib_recv_wr *wr, + struct bnxt_qplib_swqe *wqe, + int payload_size) +{ + struct bnxt_qplib_sge ref, sge; + u32 rq_prod_index; + struct bnxt_re_sqp_entries *sqp_entry; + + rq_prod_index = bnxt_qplib_get_rq_prod_index(&qp->qplib_qp); + + if (!bnxt_qplib_get_qp1_rq_buf(&qp->qplib_qp, &sge)) + return -ENOMEM; + + /* Create 1 SGE to receive the entire + * ethernet packet + */ + /* Save the reference from ULP */ + ref.addr = wqe->sg_list[0].addr; + ref.lkey = wqe->sg_list[0].lkey; + ref.size = wqe->sg_list[0].size; + + sqp_entry = &qp->rdev->sqp_tbl[rq_prod_index]; + + /* SGE 1 */ + wqe->sg_list[0].addr = sge.addr; + wqe->sg_list[0].lkey = sge.lkey; + wqe->sg_list[0].size = BNXT_QPLIB_MAX_QP1_RQ_HDR_SIZE_V2; + sge.size -= wqe->sg_list[0].size; + + sqp_entry->sge.addr = ref.addr; + sqp_entry->sge.lkey = ref.lkey; + sqp_entry->sge.size = ref.size; + /* Store the wrid for reporting completion */ + sqp_entry->wrid = wqe->wr_id; + /* change the wqe->wrid to table index */ + wqe->wr_id = rq_prod_index; + return 0; +} + +static int is_ud_qp(struct bnxt_re_qp *qp) +{ + return qp->qplib_qp.type == CMDQ_CREATE_QP_TYPE_UD; +} + +static int bnxt_re_build_send_wqe(struct bnxt_re_qp *qp, + struct ib_send_wr *wr, + struct bnxt_qplib_swqe *wqe) +{ + struct bnxt_re_ah *ah = NULL; + + if (is_ud_qp(qp)) { + ah = container_of(ud_wr(wr)->ah, struct bnxt_re_ah, ib_ah); + wqe->send.q_key = ud_wr(wr)->remote_qkey; + wqe->send.dst_qp = ud_wr(wr)->remote_qpn; + wqe->send.avid = ah->qplib_ah.id; + } + switch (wr->opcode) { + case IB_WR_SEND: + wqe->type = BNXT_QPLIB_SWQE_TYPE_SEND; + break; + case IB_WR_SEND_WITH_IMM: + wqe->type = BNXT_QPLIB_SWQE_TYPE_SEND_WITH_IMM; + wqe->send.imm_data = wr->ex.imm_data; + break; + case IB_WR_SEND_WITH_INV: + wqe->type = BNXT_QPLIB_SWQE_TYPE_SEND_WITH_INV; + wqe->send.inv_key = wr->ex.invalidate_rkey; + break; + default: + return -EINVAL; + } + if (wr->send_flags & IB_SEND_SIGNALED) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SIGNAL_COMP; + if (wr->send_flags & IB_SEND_FENCE) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_UC_FENCE; + if (wr->send_flags & IB_SEND_SOLICITED) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SOLICIT_EVENT; + if (wr->send_flags & IB_SEND_INLINE) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_INLINE; + + return 0; +} + +static int bnxt_re_build_rdma_wqe(struct ib_send_wr *wr, + struct bnxt_qplib_swqe *wqe) +{ + switch (wr->opcode) { + case IB_WR_RDMA_WRITE: + wqe->type = BNXT_QPLIB_SWQE_TYPE_RDMA_WRITE; + break; + case IB_WR_RDMA_WRITE_WITH_IMM: + wqe->type = BNXT_QPLIB_SWQE_TYPE_RDMA_WRITE_WITH_IMM; + wqe->rdma.imm_data = wr->ex.imm_data; + break; + case IB_WR_RDMA_READ: + wqe->type = BNXT_QPLIB_SWQE_TYPE_RDMA_READ; + wqe->rdma.inv_key = wr->ex.invalidate_rkey; + break; + default: + return -EINVAL; + } + wqe->rdma.remote_va = rdma_wr(wr)->remote_addr; + wqe->rdma.r_key = rdma_wr(wr)->rkey; + if (wr->send_flags & IB_SEND_SIGNALED) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SIGNAL_COMP; + if (wr->send_flags & IB_SEND_FENCE) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_UC_FENCE; + if (wr->send_flags & IB_SEND_SOLICITED) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SOLICIT_EVENT; + if (wr->send_flags & IB_SEND_INLINE) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_INLINE; + + return 0; +} + +static int bnxt_re_build_atomic_wqe(struct ib_send_wr *wr, + struct bnxt_qplib_swqe *wqe) +{ + switch (wr->opcode) { + case IB_WR_ATOMIC_CMP_AND_SWP: + wqe->type = BNXT_QPLIB_SWQE_TYPE_ATOMIC_CMP_AND_SWP; + wqe->atomic.swap_data = atomic_wr(wr)->swap; + break; + case IB_WR_ATOMIC_FETCH_AND_ADD: + wqe->type = BNXT_QPLIB_SWQE_TYPE_ATOMIC_FETCH_AND_ADD; + wqe->atomic.cmp_data = atomic_wr(wr)->compare_add; + break; + default: + return -EINVAL; + } + wqe->atomic.remote_va = atomic_wr(wr)->remote_addr; + wqe->atomic.r_key = atomic_wr(wr)->rkey; + if (wr->send_flags & IB_SEND_SIGNALED) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SIGNAL_COMP; + if (wr->send_flags & IB_SEND_FENCE) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_UC_FENCE; + if (wr->send_flags & IB_SEND_SOLICITED) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SOLICIT_EVENT; + return 0; +} + +static int bnxt_re_build_inv_wqe(struct ib_send_wr *wr, + struct bnxt_qplib_swqe *wqe) +{ + wqe->type = BNXT_QPLIB_SWQE_TYPE_LOCAL_INV; + wqe->local_inv.inv_l_key = wr->ex.invalidate_rkey; + + if (wr->send_flags & IB_SEND_SIGNALED) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SIGNAL_COMP; + if (wr->send_flags & IB_SEND_FENCE) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_UC_FENCE; + if (wr->send_flags & IB_SEND_SOLICITED) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SOLICIT_EVENT; + + return 0; +} + +static int bnxt_re_build_reg_wqe(struct ib_reg_wr *wr, + struct bnxt_qplib_swqe *wqe) +{ + struct bnxt_re_mr *mr = container_of(wr->mr, struct bnxt_re_mr, ib_mr); + struct bnxt_qplib_frpl *qplib_frpl = &mr->qplib_frpl; + int access = wr->access; + + wqe->frmr.pbl_ptr = (__le64 *)qplib_frpl->hwq.pbl_ptr[0]; + wqe->frmr.pbl_dma_ptr = qplib_frpl->hwq.pbl_dma_ptr[0]; + wqe->frmr.page_list = mr->pages; + wqe->frmr.page_list_len = mr->npages; + wqe->frmr.levels = qplib_frpl->hwq.level + 1; + wqe->type = BNXT_QPLIB_SWQE_TYPE_REG_MR; + + if (wr->wr.send_flags & IB_SEND_FENCE) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_UC_FENCE; + if (wr->wr.send_flags & IB_SEND_SIGNALED) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SIGNAL_COMP; + + if (access & IB_ACCESS_LOCAL_WRITE) + wqe->frmr.access_cntl |= SQ_FR_PMR_ACCESS_CNTL_LOCAL_WRITE; + if (access & IB_ACCESS_REMOTE_READ) + wqe->frmr.access_cntl |= SQ_FR_PMR_ACCESS_CNTL_REMOTE_READ; + if (access & IB_ACCESS_REMOTE_WRITE) + wqe->frmr.access_cntl |= SQ_FR_PMR_ACCESS_CNTL_REMOTE_WRITE; + if (access & IB_ACCESS_REMOTE_ATOMIC) + wqe->frmr.access_cntl |= SQ_FR_PMR_ACCESS_CNTL_REMOTE_ATOMIC; + if (access & IB_ACCESS_MW_BIND) + wqe->frmr.access_cntl |= SQ_FR_PMR_ACCESS_CNTL_WINDOW_BIND; + + wqe->frmr.l_key = wr->key; + wqe->frmr.length = wr->mr->length; + wqe->frmr.pbl_pg_sz_log = (wr->mr->page_size >> PAGE_SHIFT_4K) - 1; + wqe->frmr.va = wr->mr->iova; + return 0; +} + +static int bnxt_re_copy_inline_data(struct bnxt_re_dev *rdev, + struct ib_send_wr *wr, + struct bnxt_qplib_swqe *wqe) +{ + /* Copy the inline data to the data field */ + u8 *in_data; + u32 i, sge_len; + void *sge_addr; + + in_data = wqe->inline_data; + for (i = 0; i < wr->num_sge; i++) { + sge_addr = (void *)(unsigned long) + wr->sg_list[i].addr; + sge_len = wr->sg_list[i].length; + + if ((sge_len + wqe->inline_len) > + BNXT_QPLIB_SWQE_MAX_INLINE_LENGTH) { + dev_err(rdev_to_dev(rdev), + "Inline data size requested > supported value"); + return -EINVAL; + } + sge_len = wr->sg_list[i].length; + + memcpy(in_data, sge_addr, sge_len); + in_data += wr->sg_list[i].length; + wqe->inline_len += wr->sg_list[i].length; + } + return wqe->inline_len; +} + +static int bnxt_re_copy_wr_payload(struct bnxt_re_dev *rdev, + struct ib_send_wr *wr, + struct bnxt_qplib_swqe *wqe) +{ + int payload_sz = 0; + + if (wr->send_flags & IB_SEND_INLINE) + payload_sz = bnxt_re_copy_inline_data(rdev, wr, wqe); + else + payload_sz = bnxt_re_build_sgl(wr->sg_list, wqe->sg_list, + wqe->num_sge); + + return payload_sz; +} + +static void bnxt_ud_qp_hw_stall_workaround(struct bnxt_re_qp *qp) +{ + if ((qp->ib_qp.qp_type == IB_QPT_UD || + qp->ib_qp.qp_type == IB_QPT_GSI || + qp->ib_qp.qp_type == IB_QPT_RAW_ETHERTYPE) && + qp->qplib_qp.wqe_cnt == BNXT_RE_UD_QP_HW_STALL) { + int qp_attr_mask; + struct ib_qp_attr qp_attr; + + qp_attr_mask = IB_QP_STATE; + qp_attr.qp_state = IB_QPS_RTS; + bnxt_re_modify_qp(&qp->ib_qp, &qp_attr, qp_attr_mask, NULL); + qp->qplib_qp.wqe_cnt = 0; + } +} + +static int bnxt_re_post_send_shadow_qp(struct bnxt_re_dev *rdev, + struct bnxt_re_qp *qp, + struct ib_send_wr *wr) +{ + struct bnxt_qplib_swqe wqe; + int rc = 0, payload_sz = 0; + unsigned long flags; + + spin_lock_irqsave(&qp->sq_lock, flags); + memset(&wqe, 0, sizeof(wqe)); + while (wr) { + /* House keeping */ + memset(&wqe, 0, sizeof(wqe)); + + /* Common */ + wqe.num_sge = wr->num_sge; + if (wr->num_sge > qp->qplib_qp.sq.max_sge) { + dev_err(rdev_to_dev(rdev), + "Limit exceeded for Send SGEs"); + rc = -EINVAL; + goto bad; + } + + payload_sz = bnxt_re_copy_wr_payload(qp->rdev, wr, &wqe); + if (payload_sz < 0) { + rc = -EINVAL; + goto bad; + } + wqe.wr_id = wr->wr_id; + + wqe.type = BNXT_QPLIB_SWQE_TYPE_SEND; + + rc = bnxt_re_build_send_wqe(qp, wr, &wqe); + if (!rc) + rc = bnxt_qplib_post_send(&qp->qplib_qp, &wqe); +bad: + if (rc) { + dev_err(rdev_to_dev(rdev), + "Post send failed opcode = %#x rc = %d", + wr->opcode, rc); + break; + } + wr = wr->next; + } + bnxt_qplib_post_send_db(&qp->qplib_qp); + bnxt_ud_qp_hw_stall_workaround(qp); + spin_unlock_irqrestore(&qp->sq_lock, flags); + return rc; +} + +int bnxt_re_post_send(struct ib_qp *ib_qp, struct ib_send_wr *wr, + struct ib_send_wr **bad_wr) +{ + struct bnxt_re_qp *qp = container_of(ib_qp, struct bnxt_re_qp, ib_qp); + struct bnxt_qplib_swqe wqe; + int rc = 0, payload_sz = 0; + unsigned long flags; + + spin_lock_irqsave(&qp->sq_lock, flags); + while (wr) { + /* House keeping */ + memset(&wqe, 0, sizeof(wqe)); + + /* Common */ + wqe.num_sge = wr->num_sge; + if (wr->num_sge > qp->qplib_qp.sq.max_sge) { + dev_err(rdev_to_dev(qp->rdev), + "Limit exceeded for Send SGEs"); + rc = -EINVAL; + goto bad; + } + + payload_sz = bnxt_re_copy_wr_payload(qp->rdev, wr, &wqe); + if (payload_sz < 0) { + rc = -EINVAL; + goto bad; + } + wqe.wr_id = wr->wr_id; + + switch (wr->opcode) { + case IB_WR_SEND: + case IB_WR_SEND_WITH_IMM: + if (ib_qp->qp_type == IB_QPT_GSI) { + rc = bnxt_re_build_qp1_send_v2(qp, wr, &wqe, + payload_sz); + if (rc) + goto bad; + wqe.rawqp1.lflags |= + SQ_SEND_RAWETH_QP1_LFLAGS_ROCE_CRC; + } + switch (wr->send_flags) { + case IB_SEND_IP_CSUM: + wqe.rawqp1.lflags |= + SQ_SEND_RAWETH_QP1_LFLAGS_IP_CHKSUM; + break; + default: + break; + } + /* Fall thru to build the wqe */ + case IB_WR_SEND_WITH_INV: + rc = bnxt_re_build_send_wqe(qp, wr, &wqe); + break; + case IB_WR_RDMA_WRITE: + case IB_WR_RDMA_WRITE_WITH_IMM: + case IB_WR_RDMA_READ: + rc = bnxt_re_build_rdma_wqe(wr, &wqe); + break; + case IB_WR_ATOMIC_CMP_AND_SWP: + case IB_WR_ATOMIC_FETCH_AND_ADD: + rc = bnxt_re_build_atomic_wqe(wr, &wqe); + break; + case IB_WR_RDMA_READ_WITH_INV: + dev_err(rdev_to_dev(qp->rdev), + "RDMA Read with Invalidate is not supported"); + rc = -EINVAL; + goto bad; + case IB_WR_LOCAL_INV: + rc = bnxt_re_build_inv_wqe(wr, &wqe); + break; + case IB_WR_REG_MR: + rc = bnxt_re_build_reg_wqe(reg_wr(wr), &wqe); + break; + default: + /* Unsupported WRs */ + dev_err(rdev_to_dev(qp->rdev), + "WR (%#x) is not supported", wr->opcode); + rc = -EINVAL; + goto bad; + } + if (!rc) + rc = bnxt_qplib_post_send(&qp->qplib_qp, &wqe); +bad: + if (rc) { + dev_err(rdev_to_dev(qp->rdev), + "post_send failed op:%#x qps = %#x rc = %d\n", + wr->opcode, qp->qplib_qp.state, rc); + *bad_wr = wr; + break; + } + wr = wr->next; + } + bnxt_qplib_post_send_db(&qp->qplib_qp); + bnxt_ud_qp_hw_stall_workaround(qp); + spin_unlock_irqrestore(&qp->sq_lock, flags); + + return rc; +} + +static int bnxt_re_post_recv_shadow_qp(struct bnxt_re_dev *rdev, + struct bnxt_re_qp *qp, + struct ib_recv_wr *wr) +{ + struct bnxt_qplib_swqe wqe; + int rc = 0, payload_sz = 0; + + memset(&wqe, 0, sizeof(wqe)); + while (wr) { + /* House keeping */ + memset(&wqe, 0, sizeof(wqe)); + + /* Common */ + wqe.num_sge = wr->num_sge; + if (wr->num_sge > qp->qplib_qp.rq.max_sge) { + dev_err(rdev_to_dev(rdev), + "Limit exceeded for Receive SGEs"); + rc = -EINVAL; + break; + } + payload_sz = bnxt_re_build_sgl(wr->sg_list, wqe.sg_list, + wr->num_sge); + wqe.wr_id = wr->wr_id; + wqe.type = BNXT_QPLIB_SWQE_TYPE_RECV; + + rc = bnxt_qplib_post_recv(&qp->qplib_qp, &wqe); + if (rc) + break; + + wr = wr->next; + } + if (!rc) + bnxt_qplib_post_recv_db(&qp->qplib_qp); + return rc; +} + +int bnxt_re_post_recv(struct ib_qp *ib_qp, struct ib_recv_wr *wr, + struct ib_recv_wr **bad_wr) +{ + struct bnxt_re_qp *qp = container_of(ib_qp, struct bnxt_re_qp, ib_qp); + struct bnxt_qplib_swqe wqe; + int rc = 0, payload_sz = 0; + unsigned long flags; + u32 count = 0; + + spin_lock_irqsave(&qp->rq_lock, flags); + while (wr) { + /* House keeping */ + memset(&wqe, 0, sizeof(wqe)); + + /* Common */ + wqe.num_sge = wr->num_sge; + if (wr->num_sge > qp->qplib_qp.rq.max_sge) { + dev_err(rdev_to_dev(qp->rdev), + "Limit exceeded for Receive SGEs"); + rc = -EINVAL; + *bad_wr = wr; + break; + } + + payload_sz = bnxt_re_build_sgl(wr->sg_list, wqe.sg_list, + wr->num_sge); + wqe.wr_id = wr->wr_id; + wqe.type = BNXT_QPLIB_SWQE_TYPE_RECV; + + if (ib_qp->qp_type == IB_QPT_GSI) + rc = bnxt_re_build_qp1_shadow_qp_recv(qp, wr, &wqe, + payload_sz); + if (!rc) + rc = bnxt_qplib_post_recv(&qp->qplib_qp, &wqe); + if (rc) { + *bad_wr = wr; + break; + } + + /* Ring DB if the RQEs posted reaches a threshold value */ + if (++count >= BNXT_RE_RQ_WQE_THRESHOLD) { + bnxt_qplib_post_recv_db(&qp->qplib_qp); + count = 0; + } + + wr = wr->next; + } + + if (count) + bnxt_qplib_post_recv_db(&qp->qplib_qp); + + spin_unlock_irqrestore(&qp->rq_lock, flags); + + return rc; +} + +/* Completion Queues */ +int bnxt_re_destroy_cq(struct ib_cq *ib_cq) +{ + struct bnxt_re_cq *cq = container_of(ib_cq, struct bnxt_re_cq, ib_cq); + struct bnxt_re_dev *rdev = cq->rdev; + int rc; + + rc = bnxt_qplib_destroy_cq(&rdev->qplib_res, &cq->qplib_cq); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to destroy HW CQ"); + return rc; + } + if (!IS_ERR_OR_NULL(cq->umem)) + ib_umem_release(cq->umem); + + if (cq) { + kfree(cq->cql); + kfree(cq); + } + atomic_dec(&rdev->cq_count); + rdev->nq.budget--; + return 0; +} + +struct ib_cq *bnxt_re_create_cq(struct ib_device *ibdev, + const struct ib_cq_init_attr *attr, + struct ib_ucontext *context, + struct ib_udata *udata) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); + struct bnxt_qplib_dev_attr *dev_attr = &rdev->dev_attr; + struct bnxt_re_cq *cq = NULL; + int rc, entries; + int cqe = attr->cqe; + + /* Validate CQ fields */ + if (cqe < 1 || cqe > dev_attr->max_cq_wqes) { + dev_err(rdev_to_dev(rdev), "Failed to create CQ -max exceeded"); + return ERR_PTR(-EINVAL); + } + cq = kzalloc(sizeof(*cq), GFP_KERNEL); + if (!cq) + return ERR_PTR(-ENOMEM); + + cq->rdev = rdev; + cq->qplib_cq.cq_handle = (u64)(unsigned long)(&cq->qplib_cq); + + entries = roundup_pow_of_two(cqe + 1); + if (entries > dev_attr->max_cq_wqes + 1) + entries = dev_attr->max_cq_wqes + 1; + + if (context) { + struct bnxt_re_cq_req req; + struct bnxt_re_ucontext *uctx = container_of + (context, + struct bnxt_re_ucontext, + ib_uctx); + if (ib_copy_from_udata(&req, udata, sizeof(req))) { + rc = -EFAULT; + goto fail; + } + + cq->umem = ib_umem_get(context, req.cq_va, + entries * sizeof(struct cq_base), + IB_ACCESS_LOCAL_WRITE, 1); + if (IS_ERR(cq->umem)) { + rc = PTR_ERR(cq->umem); + goto fail; + } + cq->qplib_cq.sghead = cq->umem->sg_head.sgl; + cq->qplib_cq.nmap = cq->umem->nmap; + cq->qplib_cq.dpi = &uctx->dpi; + } else { + cq->max_cql = min_t(u32, entries, MAX_CQL_PER_POLL); + cq->cql = kcalloc(cq->max_cql, sizeof(struct bnxt_qplib_cqe), + GFP_KERNEL); + if (!cq->cql) { + rc = -ENOMEM; + goto fail; + } + + cq->qplib_cq.dpi = &rdev->dpi_privileged; + cq->qplib_cq.sghead = NULL; + cq->qplib_cq.nmap = 0; + } + cq->qplib_cq.max_wqe = entries; + cq->qplib_cq.cnq_hw_ring_id = rdev->nq.ring_id; + + rc = bnxt_qplib_create_cq(&rdev->qplib_res, &cq->qplib_cq); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to create HW CQ"); + goto fail; + } + + cq->ib_cq.cqe = entries; + cq->cq_period = cq->qplib_cq.period; + rdev->nq.budget++; + + atomic_inc(&rdev->cq_count); + + if (context) { + struct bnxt_re_cq_resp resp; + + resp.cqid = cq->qplib_cq.id; + resp.tail = cq->qplib_cq.hwq.cons; + resp.phase = cq->qplib_cq.period; + resp.rsvd = 0; + rc = ib_copy_to_udata(udata, &resp, sizeof(resp)); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to copy CQ udata"); + bnxt_qplib_destroy_cq(&rdev->qplib_res, &cq->qplib_cq); + goto c2fail; + } + } + + return &cq->ib_cq; + +c2fail: + if (context) + ib_umem_release(cq->umem); +fail: + kfree(cq->cql); + kfree(cq); + return ERR_PTR(rc); +} + +static u8 __req_to_ib_wc_status(u8 qstatus) +{ + switch (qstatus) { + case CQ_REQ_STATUS_OK: + return IB_WC_SUCCESS; + case CQ_REQ_STATUS_BAD_RESPONSE_ERR: + return IB_WC_BAD_RESP_ERR; + case CQ_REQ_STATUS_LOCAL_LENGTH_ERR: + return IB_WC_LOC_LEN_ERR; + case CQ_REQ_STATUS_LOCAL_QP_OPERATION_ERR: + return IB_WC_LOC_QP_OP_ERR; + case CQ_REQ_STATUS_LOCAL_PROTECTION_ERR: + return IB_WC_LOC_PROT_ERR; + case CQ_REQ_STATUS_MEMORY_MGT_OPERATION_ERR: + return IB_WC_GENERAL_ERR; + case CQ_REQ_STATUS_REMOTE_INVALID_REQUEST_ERR: + return IB_WC_REM_INV_REQ_ERR; + case CQ_REQ_STATUS_REMOTE_ACCESS_ERR: + return IB_WC_REM_ACCESS_ERR; + case CQ_REQ_STATUS_REMOTE_OPERATION_ERR: + return IB_WC_REM_OP_ERR; + case CQ_REQ_STATUS_RNR_NAK_RETRY_CNT_ERR: + return IB_WC_RNR_RETRY_EXC_ERR; + case CQ_REQ_STATUS_TRANSPORT_RETRY_CNT_ERR: + return IB_WC_RETRY_EXC_ERR; + case CQ_REQ_STATUS_WORK_REQUEST_FLUSHED_ERR: + return IB_WC_WR_FLUSH_ERR; + default: + return IB_WC_GENERAL_ERR; + } + return 0; +} + +static u8 __rawqp1_to_ib_wc_status(u8 qstatus) +{ + switch (qstatus) { + case CQ_RES_RAWETH_QP1_STATUS_OK: + return IB_WC_SUCCESS; + case CQ_RES_RAWETH_QP1_STATUS_LOCAL_ACCESS_ERROR: + return IB_WC_LOC_ACCESS_ERR; + case CQ_RES_RAWETH_QP1_STATUS_HW_LOCAL_LENGTH_ERR: + return IB_WC_LOC_LEN_ERR; + case CQ_RES_RAWETH_QP1_STATUS_LOCAL_PROTECTION_ERR: + return IB_WC_LOC_PROT_ERR; + case CQ_RES_RAWETH_QP1_STATUS_LOCAL_QP_OPERATION_ERR: + return IB_WC_LOC_QP_OP_ERR; + case CQ_RES_RAWETH_QP1_STATUS_MEMORY_MGT_OPERATION_ERR: + return IB_WC_GENERAL_ERR; + case CQ_RES_RAWETH_QP1_STATUS_WORK_REQUEST_FLUSHED_ERR: + return IB_WC_WR_FLUSH_ERR; + case CQ_RES_RAWETH_QP1_STATUS_HW_FLUSH_ERR: + return IB_WC_WR_FLUSH_ERR; + default: + return IB_WC_GENERAL_ERR; + } +} + +static u8 __rc_to_ib_wc_status(u8 qstatus) +{ + switch (qstatus) { + case CQ_RES_RC_STATUS_OK: + return IB_WC_SUCCESS; + case CQ_RES_RC_STATUS_LOCAL_ACCESS_ERROR: + return IB_WC_LOC_ACCESS_ERR; + case CQ_RES_RC_STATUS_LOCAL_LENGTH_ERR: + return IB_WC_LOC_LEN_ERR; + case CQ_RES_RC_STATUS_LOCAL_PROTECTION_ERR: + return IB_WC_LOC_PROT_ERR; + case CQ_RES_RC_STATUS_LOCAL_QP_OPERATION_ERR: + return IB_WC_LOC_QP_OP_ERR; + case CQ_RES_RC_STATUS_MEMORY_MGT_OPERATION_ERR: + return IB_WC_GENERAL_ERR; + case CQ_RES_RC_STATUS_REMOTE_INVALID_REQUEST_ERR: + return IB_WC_REM_INV_REQ_ERR; + case CQ_RES_RC_STATUS_WORK_REQUEST_FLUSHED_ERR: + return IB_WC_WR_FLUSH_ERR; + case CQ_RES_RC_STATUS_HW_FLUSH_ERR: + return IB_WC_WR_FLUSH_ERR; + default: + return IB_WC_GENERAL_ERR; + } +} + +static void bnxt_re_process_req_wc(struct ib_wc *wc, struct bnxt_qplib_cqe *cqe) +{ + switch (cqe->type) { + case BNXT_QPLIB_SWQE_TYPE_SEND: + wc->opcode = IB_WC_SEND; + break; + case BNXT_QPLIB_SWQE_TYPE_SEND_WITH_IMM: + wc->opcode = IB_WC_SEND; + wc->wc_flags |= IB_WC_WITH_IMM; + break; + case BNXT_QPLIB_SWQE_TYPE_SEND_WITH_INV: + wc->opcode = IB_WC_SEND; + wc->wc_flags |= IB_WC_WITH_INVALIDATE; + break; + case BNXT_QPLIB_SWQE_TYPE_RDMA_WRITE: + wc->opcode = IB_WC_RDMA_WRITE; + break; + case BNXT_QPLIB_SWQE_TYPE_RDMA_WRITE_WITH_IMM: + wc->opcode = IB_WC_RDMA_WRITE; + wc->wc_flags |= IB_WC_WITH_IMM; + break; + case BNXT_QPLIB_SWQE_TYPE_RDMA_READ: + wc->opcode = IB_WC_RDMA_READ; + break; + case BNXT_QPLIB_SWQE_TYPE_ATOMIC_CMP_AND_SWP: + wc->opcode = IB_WC_COMP_SWAP; + break; + case BNXT_QPLIB_SWQE_TYPE_ATOMIC_FETCH_AND_ADD: + wc->opcode = IB_WC_FETCH_ADD; + break; + case BNXT_QPLIB_SWQE_TYPE_LOCAL_INV: + wc->opcode = IB_WC_LOCAL_INV; + break; + case BNXT_QPLIB_SWQE_TYPE_REG_MR: + wc->opcode = IB_WC_REG_MR; + break; + default: + wc->opcode = IB_WC_SEND; + break; + } + + wc->status = __req_to_ib_wc_status(cqe->status); +} + +static int bnxt_re_check_packet_type(u16 raweth_qp1_flags, + u16 raweth_qp1_flags2) +{ + bool is_udp = false, is_ipv6 = false, is_ipv4 = false; + + /* raweth_qp1_flags Bit 9-6 indicates itype */ + if ((raweth_qp1_flags & CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_ROCE) + != CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_ROCE) + return -1; + + if (raweth_qp1_flags2 & + CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_IP_CS_CALC && + raweth_qp1_flags2 & + CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_L4_CS_CALC) { + is_udp = true; + /* raweth_qp1_flags2 Bit 8 indicates ip_type. 0-v4 1 - v6 */ + (raweth_qp1_flags2 & + CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_IP_TYPE) ? + (is_ipv6 = true) : (is_ipv4 = true); + return ((is_ipv6) ? + BNXT_RE_ROCEV2_IPV6_PACKET : + BNXT_RE_ROCEV2_IPV4_PACKET); + } else { + return BNXT_RE_ROCE_V1_PACKET; + } +} + +static int bnxt_re_to_ib_nw_type(int nw_type) +{ + u8 nw_hdr_type = 0xFF; + + switch (nw_type) { + case BNXT_RE_ROCE_V1_PACKET: + nw_hdr_type = RDMA_NETWORK_ROCE_V1; + break; + case BNXT_RE_ROCEV2_IPV4_PACKET: + nw_hdr_type = RDMA_NETWORK_IPV4; + break; + case BNXT_RE_ROCEV2_IPV6_PACKET: + nw_hdr_type = RDMA_NETWORK_IPV6; + break; + } + return nw_hdr_type; +} + +static bool bnxt_re_is_loopback_packet(struct bnxt_re_dev *rdev, + void *rq_hdr_buf) +{ + u8 *tmp_buf = NULL; + struct ethhdr *eth_hdr; + u16 eth_type; + bool rc = false; + + tmp_buf = (u8 *)rq_hdr_buf; + /* + * If dest mac is not same as I/F mac, this could be a + * loopback address or multicast address, check whether + * it is a loopback packet + */ + if (!ether_addr_equal(tmp_buf, rdev->netdev->dev_addr)) { + tmp_buf += 4; + /* Check the ether type */ + eth_hdr = (struct ethhdr *)tmp_buf; + eth_type = ntohs(eth_hdr->h_proto); + switch (eth_type) { + case ETH_P_IBOE: + rc = true; + break; + case ETH_P_IP: + case ETH_P_IPV6: { + u32 len; + struct udphdr *udp_hdr; + + len = (eth_type == ETH_P_IP ? sizeof(struct iphdr) : + sizeof(struct ipv6hdr)); + tmp_buf += sizeof(struct ethhdr) + len; + udp_hdr = (struct udphdr *)tmp_buf; + if (ntohs(udp_hdr->dest) == + ROCE_V2_UDP_DPORT) + rc = true; + break; + } + default: + break; + } + } + + return rc; +} + +static int bnxt_re_process_raw_qp_pkt_rx(struct bnxt_re_qp *qp1_qp, + struct bnxt_qplib_cqe *cqe) +{ + struct bnxt_re_dev *rdev = qp1_qp->rdev; + struct bnxt_re_sqp_entries *sqp_entry = NULL; + struct bnxt_re_qp *qp = rdev->qp1_sqp; + struct ib_send_wr *swr; + struct ib_ud_wr udwr; + struct ib_recv_wr rwr; + int pkt_type = 0; + u32 tbl_idx; + void *rq_hdr_buf; + dma_addr_t rq_hdr_buf_map; + dma_addr_t shrq_hdr_buf_map; + u32 offset = 0; + u32 skip_bytes = 0; + struct ib_sge s_sge[2]; + struct ib_sge r_sge[2]; + int rc; + + memset(&udwr, 0, sizeof(udwr)); + memset(&rwr, 0, sizeof(rwr)); + memset(&s_sge, 0, sizeof(s_sge)); + memset(&r_sge, 0, sizeof(r_sge)); + + swr = &udwr.wr; + tbl_idx = cqe->wr_id; + + rq_hdr_buf = qp1_qp->qplib_qp.rq_hdr_buf + + (tbl_idx * qp1_qp->qplib_qp.rq_hdr_buf_size); + rq_hdr_buf_map = bnxt_qplib_get_qp_buf_from_index(&qp1_qp->qplib_qp, + tbl_idx); + + /* Shadow QP header buffer */ + shrq_hdr_buf_map = bnxt_qplib_get_qp_buf_from_index(&qp->qplib_qp, + tbl_idx); + sqp_entry = &rdev->sqp_tbl[tbl_idx]; + + /* Store this cqe */ + memcpy(&sqp_entry->cqe, cqe, sizeof(struct bnxt_qplib_cqe)); + sqp_entry->qp1_qp = qp1_qp; + + /* Find packet type from the cqe */ + + pkt_type = bnxt_re_check_packet_type(cqe->raweth_qp1_flags, + cqe->raweth_qp1_flags2); + if (pkt_type < 0) { + dev_err(rdev_to_dev(rdev), "Invalid packet\n"); + return -EINVAL; + } + + /* Adjust the offset for the user buffer and post in the rq */ + + if (pkt_type == BNXT_RE_ROCEV2_IPV4_PACKET) + offset = 20; + + /* + * QP1 loopback packet has 4 bytes of internal header before + * ether header. Skip these four bytes. + */ + if (bnxt_re_is_loopback_packet(rdev, rq_hdr_buf)) + skip_bytes = 4; + + /* First send SGE . Skip the ether header*/ + s_sge[0].addr = rq_hdr_buf_map + BNXT_QPLIB_MAX_QP1_RQ_ETH_HDR_SIZE + + skip_bytes; + s_sge[0].lkey = 0xFFFFFFFF; + s_sge[0].length = offset ? BNXT_QPLIB_MAX_GRH_HDR_SIZE_IPV4 : + BNXT_QPLIB_MAX_GRH_HDR_SIZE_IPV6; + + /* Second Send SGE */ + s_sge[1].addr = s_sge[0].addr + s_sge[0].length + + BNXT_QPLIB_MAX_QP1_RQ_BDETH_HDR_SIZE; + if (pkt_type != BNXT_RE_ROCE_V1_PACKET) + s_sge[1].addr += 8; + s_sge[1].lkey = 0xFFFFFFFF; + s_sge[1].length = 256; + + /* First recv SGE */ + + r_sge[0].addr = shrq_hdr_buf_map; + r_sge[0].lkey = 0xFFFFFFFF; + r_sge[0].length = 40; + + r_sge[1].addr = sqp_entry->sge.addr + offset; + r_sge[1].lkey = sqp_entry->sge.lkey; + r_sge[1].length = BNXT_QPLIB_MAX_GRH_HDR_SIZE_IPV6 + 256 - offset; + + /* Create receive work request */ + rwr.num_sge = 2; + rwr.sg_list = r_sge; + rwr.wr_id = tbl_idx; + rwr.next = NULL; + + rc = bnxt_re_post_recv_shadow_qp(rdev, qp, &rwr); + if (rc) { + dev_err(rdev_to_dev(rdev), + "Failed to post Rx buffers to shadow QP"); + return -ENOMEM; + } + + swr->num_sge = 2; + swr->sg_list = s_sge; + swr->wr_id = tbl_idx; + swr->opcode = IB_WR_SEND; + swr->next = NULL; + + udwr.ah = &rdev->sqp_ah->ib_ah; + udwr.remote_qpn = rdev->qp1_sqp->qplib_qp.id; + udwr.remote_qkey = rdev->qp1_sqp->qplib_qp.qkey; + + /* post data received in the send queue */ + rc = bnxt_re_post_send_shadow_qp(rdev, qp, swr); + + return 0; +} + +static void bnxt_re_process_res_rawqp1_wc(struct ib_wc *wc, + struct bnxt_qplib_cqe *cqe) +{ + wc->opcode = IB_WC_RECV; + wc->status = __rawqp1_to_ib_wc_status(cqe->status); + wc->wc_flags |= IB_WC_GRH; +} + +static void bnxt_re_process_res_rc_wc(struct ib_wc *wc, + struct bnxt_qplib_cqe *cqe) +{ + wc->opcode = IB_WC_RECV; + wc->status = __rc_to_ib_wc_status(cqe->status); + + if (cqe->flags & CQ_RES_RC_FLAGS_IMM) + wc->wc_flags |= IB_WC_WITH_IMM; + if (cqe->flags & CQ_RES_RC_FLAGS_INV) + wc->wc_flags |= IB_WC_WITH_INVALIDATE; + if ((cqe->flags & (CQ_RES_RC_FLAGS_RDMA | CQ_RES_RC_FLAGS_IMM)) == + (CQ_RES_RC_FLAGS_RDMA | CQ_RES_RC_FLAGS_IMM)) + wc->opcode = IB_WC_RECV_RDMA_WITH_IMM; +} + +static void bnxt_re_process_res_shadow_qp_wc(struct bnxt_re_qp *qp, + struct ib_wc *wc, + struct bnxt_qplib_cqe *cqe) +{ + u32 tbl_idx; + struct bnxt_re_dev *rdev = qp->rdev; + struct bnxt_re_qp *qp1_qp = NULL; + struct bnxt_qplib_cqe *orig_cqe = NULL; + struct bnxt_re_sqp_entries *sqp_entry = NULL; + int nw_type; + + tbl_idx = cqe->wr_id; + + sqp_entry = &rdev->sqp_tbl[tbl_idx]; + qp1_qp = sqp_entry->qp1_qp; + orig_cqe = &sqp_entry->cqe; + + wc->wr_id = sqp_entry->wrid; + wc->byte_len = orig_cqe->length; + wc->qp = &qp1_qp->ib_qp; + + wc->ex.imm_data = orig_cqe->immdata; + wc->src_qp = orig_cqe->src_qp; + memcpy(wc->smac, orig_cqe->smac, ETH_ALEN); + wc->port_num = 1; + wc->vendor_err = orig_cqe->status; + + wc->opcode = IB_WC_RECV; + wc->status = __rawqp1_to_ib_wc_status(orig_cqe->status); + wc->wc_flags |= IB_WC_GRH; + + nw_type = bnxt_re_check_packet_type(orig_cqe->raweth_qp1_flags, + orig_cqe->raweth_qp1_flags2); + if (nw_type >= 0) { + wc->network_hdr_type = bnxt_re_to_ib_nw_type(nw_type); + wc->wc_flags |= IB_WC_WITH_NETWORK_HDR_TYPE; + } +} + +static void bnxt_re_process_res_ud_wc(struct ib_wc *wc, + struct bnxt_qplib_cqe *cqe) +{ + wc->opcode = IB_WC_RECV; + wc->status = __rc_to_ib_wc_status(cqe->status); + + if (cqe->flags & CQ_RES_RC_FLAGS_IMM) + wc->wc_flags |= IB_WC_WITH_IMM; + if (cqe->flags & CQ_RES_RC_FLAGS_INV) + wc->wc_flags |= IB_WC_WITH_INVALIDATE; + if ((cqe->flags & (CQ_RES_RC_FLAGS_RDMA | CQ_RES_RC_FLAGS_IMM)) == + (CQ_RES_RC_FLAGS_RDMA | CQ_RES_RC_FLAGS_IMM)) + wc->opcode = IB_WC_RECV_RDMA_WITH_IMM; +} + +static int send_phantom_wqe(struct bnxt_re_qp *qp) +{ + struct bnxt_qplib_qp *lib_qp = &qp->qplib_qp; + unsigned long flags; + int rc = 0; + + spin_lock_irqsave(&qp->sq_lock, flags); + + rc = bnxt_re_bind_fence_mw(lib_qp); + if (!rc) { + lib_qp->sq.phantom_wqe_cnt++; + dev_dbg(&lib_qp->sq.hwq.pdev->dev, + "qp %#x sq->prod %#x sw_prod %#x phantom_wqe_cnt %d\n", + lib_qp->id, lib_qp->sq.hwq.prod, + HWQ_CMP(lib_qp->sq.hwq.prod, &lib_qp->sq.hwq), + lib_qp->sq.phantom_wqe_cnt); + } + + spin_unlock_irqrestore(&qp->sq_lock, flags); + return rc; +} + +int bnxt_re_poll_cq(struct ib_cq *ib_cq, int num_entries, struct ib_wc *wc) +{ + struct bnxt_re_cq *cq = container_of(ib_cq, struct bnxt_re_cq, ib_cq); + struct bnxt_re_qp *qp; + struct bnxt_qplib_cqe *cqe; + int i, ncqe, budget; + struct bnxt_qplib_q *sq; + struct bnxt_qplib_qp *lib_qp; + u32 tbl_idx; + struct bnxt_re_sqp_entries *sqp_entry = NULL; + unsigned long flags; + + spin_lock_irqsave(&cq->cq_lock, flags); + budget = min_t(u32, num_entries, cq->max_cql); + num_entries = budget; + if (!cq->cql) { + dev_err(rdev_to_dev(cq->rdev), "POLL CQ : no CQL to use"); + goto exit; + } + cqe = &cq->cql[0]; + while (budget) { + lib_qp = NULL; + ncqe = bnxt_qplib_poll_cq(&cq->qplib_cq, cqe, budget, &lib_qp); + if (lib_qp) { + sq = &lib_qp->sq; + if (sq->send_phantom) { + qp = container_of(lib_qp, + struct bnxt_re_qp, qplib_qp); + if (send_phantom_wqe(qp) == -ENOMEM) + dev_err(rdev_to_dev(cq->rdev), + "Phantom failed! Scheduled to send again\n"); + else + sq->send_phantom = false; + } + } + + if (!ncqe) + break; + + for (i = 0; i < ncqe; i++, cqe++) { + /* Transcribe each qplib_wqe back to ib_wc */ + memset(wc, 0, sizeof(*wc)); + + wc->wr_id = cqe->wr_id; + wc->byte_len = cqe->length; + qp = container_of + ((struct bnxt_qplib_qp *) + (unsigned long)(cqe->qp_handle), + struct bnxt_re_qp, qplib_qp); + if (!qp) { + dev_err(rdev_to_dev(cq->rdev), + "POLL CQ : bad QP handle"); + continue; + } + wc->qp = &qp->ib_qp; + wc->ex.imm_data = cqe->immdata; + wc->src_qp = cqe->src_qp; + memcpy(wc->smac, cqe->smac, ETH_ALEN); + wc->port_num = 1; + wc->vendor_err = cqe->status; + + switch (cqe->opcode) { + case CQ_BASE_CQE_TYPE_REQ: + if (qp->qplib_qp.id == + qp->rdev->qp1_sqp->qplib_qp.id) { + /* Handle this completion with + * the stored completion + */ + memset(wc, 0, sizeof(*wc)); + continue; + } + bnxt_re_process_req_wc(wc, cqe); + break; + case CQ_BASE_CQE_TYPE_RES_RAWETH_QP1: + if (!cqe->status) { + int rc = 0; + + rc = bnxt_re_process_raw_qp_pkt_rx + (qp, cqe); + if (!rc) { + memset(wc, 0, sizeof(*wc)); + continue; + } + cqe->status = -1; + } + /* Errors need not be looped back. + * But change the wr_id to the one + * stored in the table + */ + tbl_idx = cqe->wr_id; + sqp_entry = &cq->rdev->sqp_tbl[tbl_idx]; + wc->wr_id = sqp_entry->wrid; + bnxt_re_process_res_rawqp1_wc(wc, cqe); + break; + case CQ_BASE_CQE_TYPE_RES_RC: + bnxt_re_process_res_rc_wc(wc, cqe); + break; + case CQ_BASE_CQE_TYPE_RES_UD: + if (qp->qplib_qp.id == + qp->rdev->qp1_sqp->qplib_qp.id) { + /* Handle this completion with + * the stored completion + */ + if (cqe->status) { + continue; + } else { + bnxt_re_process_res_shadow_qp_wc + (qp, wc, cqe); + break; + } + } + bnxt_re_process_res_ud_wc(wc, cqe); + break; + default: + dev_err(rdev_to_dev(cq->rdev), + "POLL CQ : type 0x%x not handled", + cqe->opcode); + continue; + } + wc++; + budget--; + } + } +exit: + spin_unlock_irqrestore(&cq->cq_lock, flags); + return num_entries - budget; +} + +int bnxt_re_req_notify_cq(struct ib_cq *ib_cq, + enum ib_cq_notify_flags ib_cqn_flags) +{ + struct bnxt_re_cq *cq = container_of(ib_cq, struct bnxt_re_cq, ib_cq); + int type = 0; + + /* Trigger on the very next completion */ + if (ib_cqn_flags & IB_CQ_NEXT_COMP) + type = DBR_DBR_TYPE_CQ_ARMALL; + /* Trigger on the next solicited completion */ + else if (ib_cqn_flags & IB_CQ_SOLICITED) + type = DBR_DBR_TYPE_CQ_ARMSE; + + /* Poll to see if there are missed events */ + if ((ib_cqn_flags & IB_CQ_REPORT_MISSED_EVENTS) && + !(bnxt_qplib_is_cq_empty(&cq->qplib_cq))) + return 1; + + bnxt_qplib_req_notify_cq(&cq->qplib_cq, type); + + return 0; +} + +/* Memory Regions */ +struct ib_mr *bnxt_re_get_dma_mr(struct ib_pd *ib_pd, int mr_access_flags) +{ + struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); + struct bnxt_re_dev *rdev = pd->rdev; + struct bnxt_re_mr *mr; + u64 pbl = 0; + int rc; + + mr = kzalloc(sizeof(*mr), GFP_KERNEL); + if (!mr) + return ERR_PTR(-ENOMEM); + + mr->rdev = rdev; + mr->qplib_mr.pd = &pd->qplib_pd; + mr->qplib_mr.flags = __from_ib_access_flags(mr_access_flags); + mr->qplib_mr.type = CMDQ_ALLOCATE_MRW_MRW_FLAGS_PMR; + + /* Allocate and register 0 as the address */ + rc = bnxt_qplib_alloc_mrw(&rdev->qplib_res, &mr->qplib_mr); + if (rc) + goto fail; + + mr->qplib_mr.hwq.level = PBL_LVL_MAX; + mr->qplib_mr.total_size = -1; /* Infinte length */ + rc = bnxt_qplib_reg_mr(&rdev->qplib_res, &mr->qplib_mr, &pbl, 0, false); + if (rc) + goto fail_mr; + + mr->ib_mr.lkey = mr->qplib_mr.lkey; + if (mr_access_flags & (IB_ACCESS_REMOTE_WRITE | IB_ACCESS_REMOTE_READ | + IB_ACCESS_REMOTE_ATOMIC)) + mr->ib_mr.rkey = mr->ib_mr.lkey; + atomic_inc(&rdev->mr_count); + + return &mr->ib_mr; + +fail_mr: + bnxt_qplib_free_mrw(&rdev->qplib_res, &mr->qplib_mr); +fail: + kfree(mr); + return ERR_PTR(rc); +} + +int bnxt_re_dereg_mr(struct ib_mr *ib_mr) +{ + struct bnxt_re_mr *mr = container_of(ib_mr, struct bnxt_re_mr, ib_mr); + struct bnxt_re_dev *rdev = mr->rdev; + int rc; + + rc = bnxt_qplib_free_mrw(&rdev->qplib_res, &mr->qplib_mr); + if (rc) { + dev_err(rdev_to_dev(rdev), "Dereg MR failed: %#x\n", rc); + return rc; + } + + if (mr->npages && mr->pages) { + rc = bnxt_qplib_free_fast_reg_page_list(&rdev->qplib_res, + &mr->qplib_frpl); + kfree(mr->pages); + mr->npages = 0; + mr->pages = NULL; + } + if (!IS_ERR_OR_NULL(mr->ib_umem)) + ib_umem_release(mr->ib_umem); + + kfree(mr); + atomic_dec(&rdev->mr_count); + return rc; +} + +static int bnxt_re_set_page(struct ib_mr *ib_mr, u64 addr) +{ + struct bnxt_re_mr *mr = container_of(ib_mr, struct bnxt_re_mr, ib_mr); + + if (unlikely(mr->npages == mr->qplib_frpl.max_pg_ptrs)) + return -ENOMEM; + + mr->pages[mr->npages++] = addr; + return 0; +} + +int bnxt_re_map_mr_sg(struct ib_mr *ib_mr, struct scatterlist *sg, int sg_nents, + unsigned int *sg_offset) +{ + struct bnxt_re_mr *mr = container_of(ib_mr, struct bnxt_re_mr, ib_mr); + + mr->npages = 0; + return ib_sg_to_pages(ib_mr, sg, sg_nents, sg_offset, bnxt_re_set_page); +} + +struct ib_mr *bnxt_re_alloc_mr(struct ib_pd *ib_pd, enum ib_mr_type type, + u32 max_num_sg) +{ + struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); + struct bnxt_re_dev *rdev = pd->rdev; + struct bnxt_re_mr *mr = NULL; + int rc; + + if (type != IB_MR_TYPE_MEM_REG) { + dev_dbg(rdev_to_dev(rdev), "MR type 0x%x not supported", type); + return ERR_PTR(-EINVAL); + } + if (max_num_sg > MAX_PBL_LVL_1_PGS) + return ERR_PTR(-EINVAL); + + mr = kzalloc(sizeof(*mr), GFP_KERNEL); + if (!mr) + return ERR_PTR(-ENOMEM); + + mr->rdev = rdev; + mr->qplib_mr.pd = &pd->qplib_pd; + mr->qplib_mr.flags = BNXT_QPLIB_FR_PMR; + mr->qplib_mr.type = CMDQ_ALLOCATE_MRW_MRW_FLAGS_PMR; + + rc = bnxt_qplib_alloc_mrw(&rdev->qplib_res, &mr->qplib_mr); + if (rc) + goto fail; + + mr->ib_mr.lkey = mr->qplib_mr.lkey; + mr->ib_mr.rkey = mr->ib_mr.lkey; + + mr->pages = kcalloc(max_num_sg, sizeof(u64), GFP_KERNEL); + if (!mr->pages) { + rc = -ENOMEM; + goto fail; + } + rc = bnxt_qplib_alloc_fast_reg_page_list(&rdev->qplib_res, + &mr->qplib_frpl, max_num_sg); + if (rc) { + dev_err(rdev_to_dev(rdev), + "Failed to allocate HW FR page list"); + goto fail_mr; + } + + atomic_inc(&rdev->mr_count); + return &mr->ib_mr; + +fail_mr: + bnxt_qplib_free_mrw(&rdev->qplib_res, &mr->qplib_mr); +fail: + kfree(mr->pages); + kfree(mr); + return ERR_PTR(rc); +} + +struct ib_mw *bnxt_re_alloc_mw(struct ib_pd *ib_pd, enum ib_mw_type type, + struct ib_udata *udata) +{ + struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); + struct bnxt_re_dev *rdev = pd->rdev; + struct bnxt_re_mw *mw; + int rc; + + mw = kzalloc(sizeof(*mw), GFP_KERNEL); + if (!mw) + return ERR_PTR(-ENOMEM); + mw->rdev = rdev; + mw->qplib_mw.pd = &pd->qplib_pd; + + mw->qplib_mw.type = (type == IB_MW_TYPE_1 ? + CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE1 : + CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE2B); + rc = bnxt_qplib_alloc_mrw(&rdev->qplib_res, &mw->qplib_mw); + if (rc) { + dev_err(rdev_to_dev(rdev), "Allocate MW failed!"); + goto fail; + } + mw->ib_mw.rkey = mw->qplib_mw.rkey; + + atomic_inc(&rdev->mw_count); + return &mw->ib_mw; + +fail: + kfree(mw); + return ERR_PTR(rc); +} + +int bnxt_re_dealloc_mw(struct ib_mw *ib_mw) +{ + struct bnxt_re_mw *mw = container_of(ib_mw, struct bnxt_re_mw, ib_mw); + struct bnxt_re_dev *rdev = mw->rdev; + int rc; + + rc = bnxt_qplib_free_mrw(&rdev->qplib_res, &mw->qplib_mw); + if (rc) { + dev_err(rdev_to_dev(rdev), "Free MW failed: %#x\n", rc); + return rc; + } + + kfree(mw); + atomic_dec(&rdev->mw_count); + return rc; +} + +/* uverbs */ +struct ib_mr *bnxt_re_reg_user_mr(struct ib_pd *ib_pd, u64 start, u64 length, + u64 virt_addr, int mr_access_flags, + struct ib_udata *udata) +{ + struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); + struct bnxt_re_dev *rdev = pd->rdev; + struct bnxt_re_mr *mr; + struct ib_umem *umem; + u64 *pbl_tbl, *pbl_tbl_orig; + int i, umem_pgs, pages, rc; + struct scatterlist *sg; + int entry; + + if (length > BNXT_RE_MAX_MR_SIZE) { + dev_err(rdev_to_dev(rdev), "MR Size: %lld > Max supported:%ld\n", + length, BNXT_RE_MAX_MR_SIZE); + return ERR_PTR(-ENOMEM); + } + + mr = kzalloc(sizeof(*mr), GFP_KERNEL); + if (!mr) + return ERR_PTR(-ENOMEM); + + mr->rdev = rdev; + mr->qplib_mr.pd = &pd->qplib_pd; + mr->qplib_mr.flags = __from_ib_access_flags(mr_access_flags); + mr->qplib_mr.type = CMDQ_ALLOCATE_MRW_MRW_FLAGS_MR; + + umem = ib_umem_get(ib_pd->uobject->context, start, length, + mr_access_flags, 0); + if (IS_ERR(umem)) { + dev_err(rdev_to_dev(rdev), "Failed to get umem"); + rc = -EFAULT; + goto free_mr; + } + mr->ib_umem = umem; + + rc = bnxt_qplib_alloc_mrw(&rdev->qplib_res, &mr->qplib_mr); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to allocate MR"); + goto release_umem; + } + /* The fixed portion of the rkey is the same as the lkey */ + mr->ib_mr.rkey = mr->qplib_mr.rkey; + + mr->qplib_mr.va = virt_addr; + umem_pgs = ib_umem_page_count(umem); + if (!umem_pgs) { + dev_err(rdev_to_dev(rdev), "umem is invalid!"); + rc = -EINVAL; + goto free_mrw; + } + mr->qplib_mr.total_size = length; + + pbl_tbl = kcalloc(umem_pgs, sizeof(u64 *), GFP_KERNEL); + if (!pbl_tbl) { + rc = -EINVAL; + goto free_mrw; + } + pbl_tbl_orig = pbl_tbl; + + if (umem->hugetlb) { + dev_err(rdev_to_dev(rdev), "umem hugetlb not supported!"); + rc = -EFAULT; + goto fail; + } + + if (umem->page_shift != PAGE_SHIFT) { + dev_err(rdev_to_dev(rdev), "umem page shift unsupported!"); + rc = -EFAULT; + goto fail; + } + /* Map umem buf ptrs to the PBL */ + for_each_sg(umem->sg_head.sgl, sg, umem->nmap, entry) { + pages = sg_dma_len(sg) >> umem->page_shift; + for (i = 0; i < pages; i++, pbl_tbl++) + *pbl_tbl = sg_dma_address(sg) + (i << umem->page_shift); + } + rc = bnxt_qplib_reg_mr(&rdev->qplib_res, &mr->qplib_mr, pbl_tbl_orig, + umem_pgs, false); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to register user MR"); + goto fail; + } + + kfree(pbl_tbl_orig); + + mr->ib_mr.lkey = mr->qplib_mr.lkey; + mr->ib_mr.rkey = mr->qplib_mr.lkey; + atomic_inc(&rdev->mr_count); + + return &mr->ib_mr; +fail: + kfree(pbl_tbl_orig); +free_mrw: + bnxt_qplib_free_mrw(&rdev->qplib_res, &mr->qplib_mr); +release_umem: + ib_umem_release(umem); +free_mr: + kfree(mr); + return ERR_PTR(rc); +} + +struct ib_ucontext *bnxt_re_alloc_ucontext(struct ib_device *ibdev, + struct ib_udata *udata) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); + struct bnxt_re_uctx_resp resp; + struct bnxt_re_ucontext *uctx; + struct bnxt_qplib_dev_attr *dev_attr = &rdev->dev_attr; + int rc; + + dev_dbg(rdev_to_dev(rdev), "ABI version requested %d", + ibdev->uverbs_abi_ver); + + if (ibdev->uverbs_abi_ver != BNXT_RE_ABI_VERSION) { + dev_dbg(rdev_to_dev(rdev), " is different from the device %d ", + BNXT_RE_ABI_VERSION); + return ERR_PTR(-EPERM); + } + + uctx = kzalloc(sizeof(*uctx), GFP_KERNEL); + if (!uctx) + return ERR_PTR(-ENOMEM); + + uctx->rdev = rdev; + + uctx->shpg = (void *)__get_free_page(GFP_KERNEL); + if (!uctx->shpg) { + rc = -ENOMEM; + goto fail; + } + spin_lock_init(&uctx->sh_lock); + + resp.dev_id = rdev->en_dev->pdev->devfn; /*Temp, Use idr_alloc instead*/ + resp.max_qp = rdev->qplib_ctx.qpc_count; + resp.pg_size = PAGE_SIZE; + resp.cqe_sz = sizeof(struct cq_base); + resp.max_cqd = dev_attr->max_cq_wqes; + resp.rsvd = 0; + + rc = ib_copy_to_udata(udata, &resp, sizeof(resp)); + if (rc) { + dev_err(rdev_to_dev(rdev), "Failed to copy user context"); + rc = -EFAULT; + goto cfail; + } + + return &uctx->ib_uctx; +cfail: + free_page((unsigned long)uctx->shpg); + uctx->shpg = NULL; +fail: + kfree(uctx); + return ERR_PTR(rc); +} + +int bnxt_re_dealloc_ucontext(struct ib_ucontext *ib_uctx) +{ + struct bnxt_re_ucontext *uctx = container_of(ib_uctx, + struct bnxt_re_ucontext, + ib_uctx); + + struct bnxt_re_dev *rdev = uctx->rdev; + int rc = 0; + + if (uctx->shpg) + free_page((unsigned long)uctx->shpg); + + if (uctx->dpi.dbr) { + /* Free DPI only if this is the first PD allocated by the + * application and mark the context dpi as NULL + */ + rc = bnxt_qplib_dealloc_dpi(&rdev->qplib_res, + &rdev->qplib_res.dpi_tbl, + &uctx->dpi); + if (rc) + dev_err(rdev_to_dev(rdev), "Deallocte HW DPI failed!"); + /* Don't fail, continue*/ + uctx->dpi.dbr = NULL; + } + + kfree(uctx); + return 0; +} + +/* Helper function to mmap the virtual memory from user app */ +int bnxt_re_mmap(struct ib_ucontext *ib_uctx, struct vm_area_struct *vma) +{ + struct bnxt_re_ucontext *uctx = container_of(ib_uctx, + struct bnxt_re_ucontext, + ib_uctx); + struct bnxt_re_dev *rdev = uctx->rdev; + u64 pfn; + + if (vma->vm_end - vma->vm_start != PAGE_SIZE) + return -EINVAL; + + if (vma->vm_pgoff) { + vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot); + if (io_remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff, + PAGE_SIZE, vma->vm_page_prot)) { + dev_err(rdev_to_dev(rdev), "Failed to map DPI"); + return -EAGAIN; + } + } else { + pfn = virt_to_phys(uctx->shpg) >> PAGE_SHIFT; + if (remap_pfn_range(vma, vma->vm_start, + pfn, PAGE_SIZE, vma->vm_page_prot)) { + dev_err(rdev_to_dev(rdev), + "Failed to map shared page"); + return -EAGAIN; + } + } + + return 0; +} diff --git a/drivers/infiniband/hw/bnxt_re/ib_verbs.h b/drivers/infiniband/hw/bnxt_re/ib_verbs.h new file mode 100644 index 000000000000..a0bb7e33d7ca --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/ib_verbs.h @@ -0,0 +1,206 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * 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 + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. 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. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: IB Verbs interpreter (header) + */ + +#ifndef __BNXT_RE_IB_VERBS_H__ +#define __BNXT_RE_IB_VERBS_H__ + +struct bnxt_re_gid_ctx { + u32 idx; + u32 refcnt; +}; + +#define BNXT_RE_FENCE_BYTES 64 +struct bnxt_re_fence_data { + u32 size; + u8 va[BNXT_RE_FENCE_BYTES]; + dma_addr_t dma_addr; + struct bnxt_re_mr *mr; + struct ib_mw *mw; + struct bnxt_qplib_swqe bind_wqe; + u32 bind_rkey; +}; + +struct bnxt_re_pd { + struct bnxt_re_dev *rdev; + struct ib_pd ib_pd; + struct bnxt_qplib_pd qplib_pd; + struct bnxt_re_fence_data fence; +}; + +struct bnxt_re_ah { + struct bnxt_re_dev *rdev; + struct ib_ah ib_ah; + struct bnxt_qplib_ah qplib_ah; +}; + +struct bnxt_re_qp { + struct list_head list; + struct bnxt_re_dev *rdev; + struct ib_qp ib_qp; + spinlock_t sq_lock; /* protect sq */ + spinlock_t rq_lock; /* protect rq */ + struct bnxt_qplib_qp qplib_qp; + struct ib_umem *sumem; + struct ib_umem *rumem; + /* QP1 */ + u32 send_psn; + struct ib_ud_header qp1_hdr; +}; + +struct bnxt_re_cq { + struct bnxt_re_dev *rdev; + spinlock_t cq_lock; /* protect cq */ + u16 cq_count; + u16 cq_period; + struct ib_cq ib_cq; + struct bnxt_qplib_cq qplib_cq; + struct bnxt_qplib_cqe *cql; +#define MAX_CQL_PER_POLL 1024 + u32 max_cql; + struct ib_umem *umem; +}; + +struct bnxt_re_mr { + struct bnxt_re_dev *rdev; + struct ib_mr ib_mr; + struct ib_umem *ib_umem; + struct bnxt_qplib_mrw qplib_mr; + u32 npages; + u64 *pages; + struct bnxt_qplib_frpl qplib_frpl; +}; + +struct bnxt_re_frpl { + struct bnxt_re_dev *rdev; + struct bnxt_qplib_frpl qplib_frpl; + u64 *page_list; +}; + +struct bnxt_re_fmr { + struct bnxt_re_dev *rdev; + struct ib_fmr ib_fmr; + struct bnxt_qplib_mrw qplib_fmr; +}; + +struct bnxt_re_mw { + struct bnxt_re_dev *rdev; + struct ib_mw ib_mw; + struct bnxt_qplib_mrw qplib_mw; +}; + +struct bnxt_re_ucontext { + struct bnxt_re_dev *rdev; + struct ib_ucontext ib_uctx; + struct bnxt_qplib_dpi dpi; + void *shpg; + spinlock_t sh_lock; /* protect shpg */ +}; + +struct net_device *bnxt_re_get_netdev(struct ib_device *ibdev, u8 port_num); + +int bnxt_re_query_device(struct ib_device *ibdev, + struct ib_device_attr *ib_attr, + struct ib_udata *udata); +int bnxt_re_modify_device(struct ib_device *ibdev, + int device_modify_mask, + struct ib_device_modify *device_modify); +int bnxt_re_query_port(struct ib_device *ibdev, u8 port_num, + struct ib_port_attr *port_attr); +int bnxt_re_modify_port(struct ib_device *ibdev, u8 port_num, + int port_modify_mask, + struct ib_port_modify *port_modify); +int bnxt_re_get_port_immutable(struct ib_device *ibdev, u8 port_num, + struct ib_port_immutable *immutable); +int bnxt_re_query_pkey(struct ib_device *ibdev, u8 port_num, + u16 index, u16 *pkey); +int bnxt_re_del_gid(struct ib_device *ibdev, u8 port_num, + unsigned int index, void **context); +int bnxt_re_add_gid(struct ib_device *ibdev, u8 port_num, + unsigned int index, const union ib_gid *gid, + const struct ib_gid_attr *attr, void **context); +int bnxt_re_query_gid(struct ib_device *ibdev, u8 port_num, + int index, union ib_gid *gid); +enum rdma_link_layer bnxt_re_get_link_layer(struct ib_device *ibdev, + u8 port_num); +struct ib_pd *bnxt_re_alloc_pd(struct ib_device *ibdev, + struct ib_ucontext *context, + struct ib_udata *udata); +int bnxt_re_dealloc_pd(struct ib_pd *pd); +struct ib_ah *bnxt_re_create_ah(struct ib_pd *pd, + struct rdma_ah_attr *ah_attr, + struct ib_udata *udata); +int bnxt_re_modify_ah(struct ib_ah *ah, struct rdma_ah_attr *ah_attr); +int bnxt_re_query_ah(struct ib_ah *ah, struct rdma_ah_attr *ah_attr); +int bnxt_re_destroy_ah(struct ib_ah *ah); +struct ib_qp *bnxt_re_create_qp(struct ib_pd *pd, + struct ib_qp_init_attr *qp_init_attr, + struct ib_udata *udata); +int bnxt_re_modify_qp(struct ib_qp *qp, struct ib_qp_attr *qp_attr, + int qp_attr_mask, struct ib_udata *udata); +int bnxt_re_query_qp(struct ib_qp *qp, struct ib_qp_attr *qp_attr, + int qp_attr_mask, struct ib_qp_init_attr *qp_init_attr); +int bnxt_re_destroy_qp(struct ib_qp *qp); +int bnxt_re_post_send(struct ib_qp *qp, struct ib_send_wr *send_wr, + struct ib_send_wr **bad_send_wr); +int bnxt_re_post_recv(struct ib_qp *qp, struct ib_recv_wr *recv_wr, + struct ib_recv_wr **bad_recv_wr); +struct ib_cq *bnxt_re_create_cq(struct ib_device *ibdev, + const struct ib_cq_init_attr *attr, + struct ib_ucontext *context, + struct ib_udata *udata); +int bnxt_re_destroy_cq(struct ib_cq *cq); +int bnxt_re_poll_cq(struct ib_cq *cq, int num_entries, struct ib_wc *wc); +int bnxt_re_req_notify_cq(struct ib_cq *cq, enum ib_cq_notify_flags flags); +struct ib_mr *bnxt_re_get_dma_mr(struct ib_pd *pd, int mr_access_flags); + +int bnxt_re_map_mr_sg(struct ib_mr *ib_mr, struct scatterlist *sg, int sg_nents, + unsigned int *sg_offset); +struct ib_mr *bnxt_re_alloc_mr(struct ib_pd *ib_pd, enum ib_mr_type mr_type, + u32 max_num_sg); +int bnxt_re_dereg_mr(struct ib_mr *mr); +struct ib_mw *bnxt_re_alloc_mw(struct ib_pd *ib_pd, enum ib_mw_type type, + struct ib_udata *udata); +int bnxt_re_dealloc_mw(struct ib_mw *mw); +struct ib_mr *bnxt_re_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, + u64 virt_addr, int mr_access_flags, + struct ib_udata *udata); +struct ib_ucontext *bnxt_re_alloc_ucontext(struct ib_device *ibdev, + struct ib_udata *udata); +int bnxt_re_dealloc_ucontext(struct ib_ucontext *context); +int bnxt_re_mmap(struct ib_ucontext *context, struct vm_area_struct *vma); +#endif /* __BNXT_RE_IB_VERBS_H__ */ diff --git a/drivers/infiniband/hw/bnxt_re/main.c b/drivers/infiniband/hw/bnxt_re/main.c new file mode 100644 index 000000000000..ceae2d92fb08 --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/main.c @@ -0,0 +1,1312 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * 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 + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. 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. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: Main component of the bnxt_re driver + */ + +#include <linux/module.h> +#include <linux/netdevice.h> +#include <linux/ethtool.h> +#include <linux/mutex.h> +#include <linux/list.h> +#include <linux/rculist.h> +#include <linux/spinlock.h> +#include <linux/pci.h> +#include <net/dcbnl.h> +#include <net/ipv6.h> +#include <net/addrconf.h> +#include <linux/if_ether.h> + +#include <rdma/ib_verbs.h> +#include <rdma/ib_user_verbs.h> +#include <rdma/ib_umem.h> +#include <rdma/ib_addr.h> + +#include "bnxt_ulp.h" +#include "roce_hsi.h" +#include "qplib_res.h" +#include "qplib_sp.h" +#include "qplib_fp.h" +#include "qplib_rcfw.h" +#include "bnxt_re.h" +#include "ib_verbs.h" +#include <rdma/bnxt_re-abi.h> +#include "bnxt.h" +static char version[] = + BNXT_RE_DESC " v" ROCE_DRV_MODULE_VERSION "\n"; + +MODULE_AUTHOR("Eddie Wai <eddie.wai@broadcom.com>"); +MODULE_DESCRIPTION(BNXT_RE_DESC " Driver"); +MODULE_LICENSE("Dual BSD/GPL"); +MODULE_VERSION(ROCE_DRV_MODULE_VERSION); + +/* globals */ +static struct list_head bnxt_re_dev_list = LIST_HEAD_INIT(bnxt_re_dev_list); +/* Mutex to protect the list of bnxt_re devices added */ +static DEFINE_MUTEX(bnxt_re_dev_lock); +static struct workqueue_struct *bnxt_re_wq; + +/* for handling bnxt_en callbacks later */ +static void bnxt_re_stop(void *p) +{ +} + +static void bnxt_re_start(void *p) +{ +} + +static void bnxt_re_sriov_config(void *p, int num_vfs) +{ +} + +static struct bnxt_ulp_ops bnxt_re_ulp_ops = { + .ulp_async_notifier = NULL, + .ulp_stop = bnxt_re_stop, + .ulp_start = bnxt_re_start, + .ulp_sriov_config = bnxt_re_sriov_config +}; + +/* RoCE -> Net driver */ + +/* Driver registration routines used to let the networking driver (bnxt_en) + * to know that the RoCE driver is now installed + */ +static int bnxt_re_unregister_netdev(struct bnxt_re_dev *rdev, bool lock_wait) +{ + struct bnxt_en_dev *en_dev; + int rc; + + if (!rdev) + return -EINVAL; + + en_dev = rdev->en_dev; + /* Acquire rtnl lock if it is not invokded from netdev event */ + if (lock_wait) + rtnl_lock(); + + rc = en_dev->en_ops->bnxt_unregister_device(rdev->en_dev, + BNXT_ROCE_ULP); + if (lock_wait) + rtnl_unlock(); + return rc; +} + +static int bnxt_re_register_netdev(struct bnxt_re_dev *rdev) +{ + struct bnxt_en_dev *en_dev; + int rc = 0; + + if (!rdev) + return -EINVAL; + + en_dev = rdev->en_dev; + + rtnl_lock(); + rc = en_dev->en_ops->bnxt_register_device(en_dev, BNXT_ROCE_ULP, + &bnxt_re_ulp_ops, rdev); + rtnl_unlock(); + return rc; +} + +static int bnxt_re_free_msix(struct bnxt_re_dev *rdev, bool lock_wait) +{ + struct bnxt_en_dev *en_dev; + int rc; + + if (!rdev) + return -EINVAL; + + en_dev = rdev->en_dev; + + if (lock_wait) + rtnl_lock(); + + rc = en_dev->en_ops->bnxt_free_msix(rdev->en_dev, BNXT_ROCE_ULP); + + if (lock_wait) + rtnl_unlock(); + return rc; +} + +static int bnxt_re_request_msix(struct bnxt_re_dev *rdev) +{ + int rc = 0, num_msix_want = BNXT_RE_MIN_MSIX, num_msix_got; + struct bnxt_en_dev *en_dev; + + if (!rdev) + return -EINVAL; + + en_dev = rdev->en_dev; + + rtnl_lock(); + num_msix_got = en_dev->en_ops->bnxt_request_msix(en_dev, BNXT_ROCE_ULP, + rdev->msix_entries, + num_msix_want); + if (num_msix_got < BNXT_RE_MIN_MSIX) { + rc = -EINVAL; + goto done; + } + if (num_msix_got != num_msix_want) { + dev_warn(rdev_to_dev(rdev), + "Requested %d MSI-X vectors, got %d\n", + num_msix_want, num_msix_got); + } + rdev->num_msix = num_msix_got; +done: + rtnl_unlock(); + return rc; +} + +static void bnxt_re_init_hwrm_hdr(struct bnxt_re_dev *rdev, struct input *hdr, + u16 opcd, u16 crid, u16 trid) +{ + hdr->req_type = cpu_to_le16(opcd); + hdr->cmpl_ring = cpu_to_le16(crid); + hdr->target_id = cpu_to_le16(trid); +} + +static void bnxt_re_fill_fw_msg(struct bnxt_fw_msg *fw_msg, void *msg, + int msg_len, void *resp, int resp_max_len, + int timeout) +{ + fw_msg->msg = msg; + fw_msg->msg_len = msg_len; + fw_msg->resp = resp; + fw_msg->resp_max_len = resp_max_len; + fw_msg->timeout = timeout; +} + +static int bnxt_re_net_ring_free(struct bnxt_re_dev *rdev, u16 fw_ring_id, + bool lock_wait) +{ + struct bnxt_en_dev *en_dev = rdev->en_dev; + struct hwrm_ring_free_input req = {0}; + struct hwrm_ring_free_output resp; + struct bnxt_fw_msg fw_msg; + bool do_unlock = false; + int rc = -EINVAL; + + if (!en_dev) + return rc; + + memset(&fw_msg, 0, sizeof(fw_msg)); + if (lock_wait) { + rtnl_lock(); + do_unlock = true; + } + + bnxt_re_init_hwrm_hdr(rdev, (void *)&req, HWRM_RING_FREE, -1, -1); + req.ring_type = RING_ALLOC_REQ_RING_TYPE_L2_CMPL; + req.ring_id = cpu_to_le16(fw_ring_id); + bnxt_re_fill_fw_msg(&fw_msg, (void *)&req, sizeof(req), (void *)&resp, + sizeof(resp), DFLT_HWRM_CMD_TIMEOUT); + rc = en_dev->en_ops->bnxt_send_fw_msg(en_dev, BNXT_ROCE_ULP, &fw_msg); + if (rc) + dev_err(rdev_to_dev(rdev), + "Failed to free HW ring:%d :%#x", req.ring_id, rc); + if (do_unlock) + rtnl_unlock(); + return rc; +} + +static int bnxt_re_net_ring_alloc(struct bnxt_re_dev *rdev, dma_addr_t *dma_arr, + int pages, int type, u32 ring_mask, + u32 map_index, u16 *fw_ring_id) +{ + struct bnxt_en_dev *en_dev = rdev->en_dev; + struct hwrm_ring_alloc_input req = {0}; + struct hwrm_ring_alloc_output resp; + struct bnxt_fw_msg fw_msg; + int rc = -EINVAL; + + if (!en_dev) + return rc; + + memset(&fw_msg, 0, sizeof(fw_msg)); + rtnl_lock(); + bnxt_re_init_hwrm_hdr(rdev, (void *)&req, HWRM_RING_ALLOC, -1, -1); + req.enables = 0; + req.page_tbl_addr = cpu_to_le64(dma_arr[0]); + if (pages > 1) { + /* Page size is in log2 units */ + req.page_size = BNXT_PAGE_SHIFT; + req.page_tbl_depth = 1; + } + req.fbo = 0; + /* Association of ring index with doorbell index and MSIX number */ + req.logical_id = cpu_to_le16(map_index); + req.length = cpu_to_le32(ring_mask + 1); + req.ring_type = RING_ALLOC_REQ_RING_TYPE_L2_CMPL; + req.int_mode = RING_ALLOC_REQ_INT_MODE_MSIX; + bnxt_re_fill_fw_msg(&fw_msg, (void *)&req, sizeof(req), (void *)&resp, + sizeof(resp), DFLT_HWRM_CMD_TIMEOUT); + rc = en_dev->en_ops->bnxt_send_fw_msg(en_dev, BNXT_ROCE_ULP, &fw_msg); + if (!rc) + *fw_ring_id = le16_to_cpu(resp.ring_id); + + rtnl_unlock(); + return rc; +} + +static int bnxt_re_net_stats_ctx_free(struct bnxt_re_dev *rdev, + u32 fw_stats_ctx_id, bool lock_wait) +{ + struct bnxt_en_dev *en_dev = rdev->en_dev; + struct hwrm_stat_ctx_free_input req = {0}; + struct bnxt_fw_msg fw_msg; + bool do_unlock = false; + int rc = -EINVAL; + + if (!en_dev) + return rc; + + memset(&fw_msg, 0, sizeof(fw_msg)); + if (lock_wait) { + rtnl_lock(); + do_unlock = true; + } + + bnxt_re_init_hwrm_hdr(rdev, (void *)&req, HWRM_STAT_CTX_FREE, -1, -1); + req.stat_ctx_id = cpu_to_le32(fw_stats_ctx_id); + bnxt_re_fill_fw_msg(&fw_msg, (void *)&req, sizeof(req), (void *)&req, + sizeof(req), DFLT_HWRM_CMD_TIMEOUT); + rc = en_dev->en_ops->bnxt_send_fw_msg(en_dev, BNXT_ROCE_ULP, &fw_msg); + if (rc) + dev_err(rdev_to_dev(rdev), + "Failed to free HW stats context %#x", rc); + + if (do_unlock) + rtnl_unlock(); + return rc; +} + +static int bnxt_re_net_stats_ctx_alloc(struct bnxt_re_dev *rdev, + dma_addr_t dma_map, + u32 *fw_stats_ctx_id) +{ + struct hwrm_stat_ctx_alloc_output resp = {0}; + struct hwrm_stat_ctx_alloc_input req = {0}; + struct bnxt_en_dev *en_dev = rdev->en_dev; + struct bnxt_fw_msg fw_msg; + int rc = -EINVAL; + + *fw_stats_ctx_id = INVALID_STATS_CTX_ID; + + if (!en_dev) + return rc; + + memset(&fw_msg, 0, sizeof(fw_msg)); + rtnl_lock(); + + bnxt_re_init_hwrm_hdr(rdev, (void *)&req, HWRM_STAT_CTX_ALLOC, -1, -1); + req.update_period_ms = cpu_to_le32(1000); + req.stats_dma_addr = cpu_to_le64(dma_map); + req.stat_ctx_flags = STAT_CTX_ALLOC_REQ_STAT_CTX_FLAGS_ROCE; + bnxt_re_fill_fw_msg(&fw_msg, (void *)&req, sizeof(req), (void *)&resp, + sizeof(resp), DFLT_HWRM_CMD_TIMEOUT); + rc = en_dev->en_ops->bnxt_send_fw_msg(en_dev, BNXT_ROCE_ULP, &fw_msg); + if (!rc) + *fw_stats_ctx_id = le32_to_cpu(resp.stat_ctx_id); + + rtnl_unlock(); + return rc; +} + +/* Device */ + +static bool is_bnxt_re_dev(struct net_device *netdev) +{ + struct ethtool_drvinfo drvinfo; + + if (netdev->ethtool_ops && netdev->ethtool_ops->get_drvinfo) { + memset(&drvinfo, 0, sizeof(drvinfo)); + netdev->ethtool_ops->get_drvinfo(netdev, &drvinfo); + + if (strcmp(drvinfo.driver, "bnxt_en")) + return false; + return true; + } + return false; +} + +static struct bnxt_re_dev *bnxt_re_from_netdev(struct net_device *netdev) +{ + struct bnxt_re_dev *rdev; + + rcu_read_lock(); + list_for_each_entry_rcu(rdev, &bnxt_re_dev_list, list) { + if (rdev->netdev == netdev) { + rcu_read_unlock(); + return rdev; + } + } + rcu_read_unlock(); + return NULL; +} + +static void bnxt_re_dev_unprobe(struct net_device *netdev, + struct bnxt_en_dev *en_dev) +{ + dev_put(netdev); + module_put(en_dev->pdev->driver->driver.owner); +} + +static struct bnxt_en_dev *bnxt_re_dev_probe(struct net_device *netdev) +{ + struct bnxt *bp = netdev_priv(netdev); + struct bnxt_en_dev *en_dev; + struct pci_dev *pdev; + + /* Call bnxt_en's RoCE probe via indirect API */ + if (!bp->ulp_probe) + return ERR_PTR(-EINVAL); + + en_dev = bp->ulp_probe(netdev); + if (IS_ERR(en_dev)) + return en_dev; + + pdev = en_dev->pdev; + if (!pdev) + return ERR_PTR(-EINVAL); + + if (!(en_dev->flags & BNXT_EN_FLAG_ROCE_CAP)) { + dev_dbg(&pdev->dev, + "%s: probe error: RoCE is not supported on this device", + ROCE_DRV_MODULE_NAME); + return ERR_PTR(-ENODEV); + } + + /* Bump net device reference count */ + if (!try_module_get(pdev->driver->driver.owner)) + return ERR_PTR(-ENODEV); + + dev_hold(netdev); + + return en_dev; +} + +static void bnxt_re_unregister_ib(struct bnxt_re_dev *rdev) +{ + ib_unregister_device(&rdev->ibdev); +} + +static int bnxt_re_register_ib(struct bnxt_re_dev *rdev) +{ + struct ib_device *ibdev = &rdev->ibdev; + + /* ib device init */ + ibdev->owner = THIS_MODULE; + ibdev->node_type = RDMA_NODE_IB_CA; + strlcpy(ibdev->name, "bnxt_re%d", IB_DEVICE_NAME_MAX); + strlcpy(ibdev->node_desc, BNXT_RE_DESC " HCA", + strlen(BNXT_RE_DESC) + 5); + ibdev->phys_port_cnt = 1; + + bnxt_qplib_get_guid(rdev->netdev->dev_addr, (u8 *)&ibdev->node_guid); + + ibdev->num_comp_vectors = 1; + ibdev->dev.parent = &rdev->en_dev->pdev->dev; + ibdev->local_dma_lkey = BNXT_QPLIB_RSVD_LKEY; + + /* User space */ + ibdev->uverbs_abi_ver = BNXT_RE_ABI_VERSION; + ibdev->uverbs_cmd_mask = + (1ull << IB_USER_VERBS_CMD_GET_CONTEXT) | + (1ull << IB_USER_VERBS_CMD_QUERY_DEVICE) | + (1ull << IB_USER_VERBS_CMD_QUERY_PORT) | + (1ull << IB_USER_VERBS_CMD_ALLOC_PD) | + (1ull << IB_USER_VERBS_CMD_DEALLOC_PD) | + (1ull << IB_USER_VERBS_CMD_REG_MR) | + (1ull << IB_USER_VERBS_CMD_REREG_MR) | + (1ull << IB_USER_VERBS_CMD_DEREG_MR) | + (1ull << IB_USER_VERBS_CMD_CREATE_COMP_CHANNEL) | + (1ull << IB_USER_VERBS_CMD_CREATE_CQ) | + (1ull << IB_USER_VERBS_CMD_RESIZE_CQ) | + (1ull << IB_USER_VERBS_CMD_DESTROY_CQ) | + (1ull << IB_USER_VERBS_CMD_CREATE_QP) | + (1ull << IB_USER_VERBS_CMD_MODIFY_QP) | + (1ull << IB_USER_VERBS_CMD_QUERY_QP) | + (1ull << IB_USER_VERBS_CMD_DESTROY_QP) | + (1ull << IB_USER_VERBS_CMD_CREATE_SRQ) | + (1ull << IB_USER_VERBS_CMD_MODIFY_SRQ) | + (1ull << IB_USER_VERBS_CMD_QUERY_SRQ) | + (1ull << IB_USER_VERBS_CMD_DESTROY_SRQ) | + (1ull << IB_USER_VERBS_CMD_CREATE_AH) | + (1ull << IB_USER_VERBS_CMD_MODIFY_AH) | + (1ull << IB_USER_VERBS_CMD_QUERY_AH) | + (1ull << IB_USER_VERBS_CMD_DESTROY_AH); + /* POLL_CQ and REQ_NOTIFY_CQ is directly handled in libbnxt_re */ + + /* Kernel verbs */ + ibdev->query_device = bnxt_re_query_device; + ibdev->modify_device = bnxt_re_modify_device; + + ibdev->query_port = bnxt_re_query_port; + ibdev->modify_port = bnxt_re_modify_port; + ibdev->get_port_immutable = bnxt_re_get_port_immutable; + ibdev->query_pkey = bnxt_re_query_pkey; + ibdev->query_gid = bnxt_re_query_gid; + ibdev->get_netdev = bnxt_re_get_netdev; + ibdev->add_gid = bnxt_re_add_gid; + ibdev->del_gid = bnxt_re_del_gid; + ibdev->get_link_layer = bnxt_re_get_link_layer; + + ibdev->alloc_pd = bnxt_re_alloc_pd; + ibdev->dealloc_pd = bnxt_re_dealloc_pd; + + ibdev->create_ah = bnxt_re_create_ah; + ibdev->modify_ah = bnxt_re_modify_ah; + ibdev->query_ah = bnxt_re_query_ah; + ibdev->destroy_ah = bnxt_re_destroy_ah; + + ibdev->create_qp = bnxt_re_create_qp; + ibdev->modify_qp = bnxt_re_modify_qp; + ibdev->query_qp = bnxt_re_query_qp; + ibdev->destroy_qp = bnxt_re_destroy_qp; + + ibdev->post_send = bnxt_re_post_send; + ibdev->post_recv = bnxt_re_post_recv; + + ibdev->create_cq = bnxt_re_create_cq; + ibdev->destroy_cq = bnxt_re_destroy_cq; + ibdev->poll_cq = bnxt_re_poll_cq; + ibdev->req_notify_cq = bnxt_re_req_notify_cq; + + ibdev->get_dma_mr = bnxt_re_get_dma_mr; + ibdev->dereg_mr = bnxt_re_dereg_mr; + ibdev->alloc_mr = bnxt_re_alloc_mr; + ibdev->map_mr_sg = bnxt_re_map_mr_sg; + + ibdev->reg_user_mr = bnxt_re_reg_user_mr; + ibdev->alloc_ucontext = bnxt_re_alloc_ucontext; + ibdev->dealloc_ucontext = bnxt_re_dealloc_ucontext; + ibdev->mmap = bnxt_re_mmap; + + return ib_register_device(ibdev, NULL); +} + +static ssize_t show_rev(struct device *device, struct device_attribute *attr, + char *buf) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(device, ibdev.dev); + + return scnprintf(buf, PAGE_SIZE, "0x%x\n", rdev->en_dev->pdev->vendor); +} + +static ssize_t show_fw_ver(struct device *device, struct device_attribute *attr, + char *buf) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(device, ibdev.dev); + + return scnprintf(buf, PAGE_SIZE, "%s\n", rdev->dev_attr.fw_ver); +} + +static ssize_t show_hca(struct device *device, struct device_attribute *attr, + char *buf) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(device, ibdev.dev); + + return scnprintf(buf, PAGE_SIZE, "%s\n", rdev->ibdev.node_desc); +} + +static DEVICE_ATTR(hw_rev, 0444, show_rev, NULL); +static DEVICE_ATTR(fw_rev, 0444, show_fw_ver, NULL); +static DEVICE_ATTR(hca_type, 0444, show_hca, NULL); + +static struct device_attribute *bnxt_re_attributes[] = { + &dev_attr_hw_rev, + &dev_attr_fw_rev, + &dev_attr_hca_type +}; + +static void bnxt_re_dev_remove(struct bnxt_re_dev *rdev) +{ + dev_put(rdev->netdev); + rdev->netdev = NULL; + + mutex_lock(&bnxt_re_dev_lock); + list_del_rcu(&rdev->list); + mutex_unlock(&bnxt_re_dev_lock); + + synchronize_rcu(); + flush_workqueue(bnxt_re_wq); + + ib_dealloc_device(&rdev->ibdev); + /* rdev is gone */ +} + +static struct bnxt_re_dev *bnxt_re_dev_add(struct net_device *netdev, + struct bnxt_en_dev *en_dev) +{ + struct bnxt_re_dev *rdev; + + /* Allocate bnxt_re_dev instance here */ + rdev = (struct bnxt_re_dev *)ib_alloc_device(sizeof(*rdev)); + if (!rdev) { + dev_err(NULL, "%s: bnxt_re_dev allocation failure!", + ROCE_DRV_MODULE_NAME); + return NULL; + } + /* Default values */ + rdev->netdev = netdev; + dev_hold(rdev->netdev); + rdev->en_dev = en_dev; + rdev->id = rdev->en_dev->pdev->devfn; + INIT_LIST_HEAD(&rdev->qp_list); + mutex_init(&rdev->qp_lock); + atomic_set(&rdev->qp_count, 0); + atomic_set(&rdev->cq_count, 0); + atomic_set(&rdev->srq_count, 0); + atomic_set(&rdev->mr_count, 0); + atomic_set(&rdev->mw_count, 0); + rdev->cosq[0] = 0xFFFF; + rdev->cosq[1] = 0xFFFF; + + mutex_lock(&bnxt_re_dev_lock); + list_add_tail_rcu(&rdev->list, &bnxt_re_dev_list); + mutex_unlock(&bnxt_re_dev_lock); + return rdev; +} + +static int bnxt_re_aeq_handler(struct bnxt_qplib_rcfw *rcfw, + struct creq_func_event *aeqe) +{ + switch (aeqe->event) { + case CREQ_FUNC_EVENT_EVENT_TX_WQE_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_TX_DATA_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_RX_WQE_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_RX_DATA_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_CQ_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_TQM_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_CFCQ_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_CFCS_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_CFCC_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_CFCM_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_TIM_ERROR: + break; + default: + return -EINVAL; + } + return 0; +} + +static int bnxt_re_cqn_handler(struct bnxt_qplib_nq *nq, + struct bnxt_qplib_cq *handle) +{ + struct bnxt_re_cq *cq = container_of(handle, struct bnxt_re_cq, + qplib_cq); + + if (!cq) { + dev_err(NULL, "%s: CQ is NULL, CQN not handled", + ROCE_DRV_MODULE_NAME); + return -EINVAL; + } + if (cq->ib_cq.comp_handler) { + /* Lock comp_handler? */ + (*cq->ib_cq.comp_handler)(&cq->ib_cq, cq->ib_cq.cq_context); + } + + return 0; +} + +static void bnxt_re_cleanup_res(struct bnxt_re_dev *rdev) +{ + if (rdev->nq.hwq.max_elements) + bnxt_qplib_disable_nq(&rdev->nq); + + if (rdev->qplib_res.rcfw) + bnxt_qplib_cleanup_res(&rdev->qplib_res); +} + +static int bnxt_re_init_res(struct bnxt_re_dev *rdev) +{ + int rc = 0; + + bnxt_qplib_init_res(&rdev->qplib_res); + + if (rdev->msix_entries[BNXT_RE_NQ_IDX].vector <= 0) + return -EINVAL; + + rc = bnxt_qplib_enable_nq(rdev->en_dev->pdev, &rdev->nq, + rdev->msix_entries[BNXT_RE_NQ_IDX].vector, + rdev->msix_entries[BNXT_RE_NQ_IDX].db_offset, + &bnxt_re_cqn_handler, + NULL); + + if (rc) + dev_err(rdev_to_dev(rdev), "Failed to enable NQ: %#x", rc); + + return rc; +} + +static void bnxt_re_free_res(struct bnxt_re_dev *rdev, bool lock_wait) +{ + if (rdev->nq.hwq.max_elements) { + bnxt_re_net_ring_free(rdev, rdev->nq.ring_id, lock_wait); + bnxt_qplib_free_nq(&rdev->nq); + } + if (rdev->qplib_res.dpi_tbl.max) { + bnxt_qplib_dealloc_dpi(&rdev->qplib_res, + &rdev->qplib_res.dpi_tbl, + &rdev->dpi_privileged); + } + if (rdev->qplib_res.rcfw) { + bnxt_qplib_free_res(&rdev->qplib_res); + rdev->qplib_res.rcfw = NULL; + } +} + +static int bnxt_re_alloc_res(struct bnxt_re_dev *rdev) +{ + int rc = 0; + + /* Configure and allocate resources for qplib */ + rdev->qplib_res.rcfw = &rdev->rcfw; + rc = bnxt_qplib_get_dev_attr(&rdev->rcfw, &rdev->dev_attr); + if (rc) + goto fail; + + rc = bnxt_qplib_alloc_res(&rdev->qplib_res, rdev->en_dev->pdev, + rdev->netdev, &rdev->dev_attr); + if (rc) + goto fail; + + rc = bnxt_qplib_alloc_dpi(&rdev->qplib_res.dpi_tbl, + &rdev->dpi_privileged, + rdev); + if (rc) + goto fail; + + rdev->nq.hwq.max_elements = BNXT_RE_MAX_CQ_COUNT + + BNXT_RE_MAX_SRQC_COUNT + 2; + rc = bnxt_qplib_alloc_nq(rdev->en_dev->pdev, &rdev->nq); + if (rc) { + dev_err(rdev_to_dev(rdev), + "Failed to allocate NQ memory: %#x", rc); + goto fail; + } + rc = bnxt_re_net_ring_alloc + (rdev, rdev->nq.hwq.pbl[PBL_LVL_0].pg_map_arr, + rdev->nq.hwq.pbl[rdev->nq.hwq.level].pg_count, + HWRM_RING_ALLOC_CMPL, BNXT_QPLIB_NQE_MAX_CNT - 1, + rdev->msix_entries[BNXT_RE_NQ_IDX].ring_idx, + &rdev->nq.ring_id); + if (rc) { + dev_err(rdev_to_dev(rdev), + "Failed to allocate NQ ring: %#x", rc); + goto free_nq; + } + return 0; +free_nq: + bnxt_qplib_free_nq(&rdev->nq); +fail: + rdev->qplib_res.rcfw = NULL; + return rc; +} + +static void bnxt_re_dispatch_event(struct ib_device *ibdev, struct ib_qp *qp, + u8 port_num, enum ib_event_type event) +{ + struct ib_event ib_event; + + ib_event.device = ibdev; + if (qp) + ib_event.element.qp = qp; + else + ib_event.element.port_num = port_num; + ib_event.event = event; + ib_dispatch_event(&ib_event); +} + +#define HWRM_QUEUE_PRI2COS_QCFG_INPUT_FLAGS_IVLAN 0x02 +static int bnxt_re_query_hwrm_pri2cos(struct bnxt_re_dev *rdev, u8 dir, + u64 *cid_map) +{ + struct hwrm_queue_pri2cos_qcfg_input req = {0}; + struct bnxt *bp = netdev_priv(rdev->netdev); + struct hwrm_queue_pri2cos_qcfg_output resp; + struct bnxt_en_dev *en_dev = rdev->en_dev; + struct bnxt_fw_msg fw_msg; + u32 flags = 0; + u8 *qcfgmap, *tmp_map; + int rc = 0, i; + + if (!cid_map) + return -EINVAL; + + memset(&fw_msg, 0, sizeof(fw_msg)); + bnxt_re_init_hwrm_hdr(rdev, (void *)&req, + HWRM_QUEUE_PRI2COS_QCFG, -1, -1); + flags |= (dir & 0x01); + flags |= HWRM_QUEUE_PRI2COS_QCFG_INPUT_FLAGS_IVLAN; + req.flags = cpu_to_le32(flags); + req.port_id = bp->pf.port_id; + + bnxt_re_fill_fw_msg(&fw_msg, (void *)&req, sizeof(req), (void *)&resp, + sizeof(resp), DFLT_HWRM_CMD_TIMEOUT); + rc = en_dev->en_ops->bnxt_send_fw_msg(en_dev, BNXT_ROCE_ULP, &fw_msg); + if (rc) + return rc; + + if (resp.queue_cfg_info) { + dev_warn(rdev_to_dev(rdev), + "Asymmetric cos queue configuration detected"); + dev_warn(rdev_to_dev(rdev), + " on device, QoS may not be fully functional\n"); + } + qcfgmap = &resp.pri0_cos_queue_id; + tmp_map = (u8 *)cid_map; + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) + tmp_map[i] = qcfgmap[i]; + + return rc; +} + +static bool bnxt_re_is_qp1_or_shadow_qp(struct bnxt_re_dev *rdev, + struct bnxt_re_qp *qp) +{ + return (qp->ib_qp.qp_type == IB_QPT_GSI) || (qp == rdev->qp1_sqp); +} + +static void bnxt_re_dev_stop(struct bnxt_re_dev *rdev) +{ + int mask = IB_QP_STATE; + struct ib_qp_attr qp_attr; + struct bnxt_re_qp *qp; + + qp_attr.qp_state = IB_QPS_ERR; + mutex_lock(&rdev->qp_lock); + list_for_each_entry(qp, &rdev->qp_list, list) { + /* Modify the state of all QPs except QP1/Shadow QP */ + if (!bnxt_re_is_qp1_or_shadow_qp(rdev, qp)) { + if (qp->qplib_qp.state != + CMDQ_MODIFY_QP_NEW_STATE_RESET && + qp->qplib_qp.state != + CMDQ_MODIFY_QP_NEW_STATE_ERR) { + bnxt_re_dispatch_event(&rdev->ibdev, &qp->ib_qp, + 1, IB_EVENT_QP_FATAL); + bnxt_re_modify_qp(&qp->ib_qp, &qp_attr, mask, + NULL); + } + } + } + mutex_unlock(&rdev->qp_lock); +} + +static u32 bnxt_re_get_priority_mask(struct bnxt_re_dev *rdev) +{ + u32 prio_map = 0, tmp_map = 0; + struct net_device *netdev; + struct dcb_app app; + + netdev = rdev->netdev; + + memset(&app, 0, sizeof(app)); + app.selector = IEEE_8021QAZ_APP_SEL_ETHERTYPE; + app.protocol = ETH_P_IBOE; + tmp_map = dcb_ieee_getapp_mask(netdev, &app); + prio_map = tmp_map; + + app.selector = IEEE_8021QAZ_APP_SEL_DGRAM; + app.protocol = ROCE_V2_UDP_DPORT; + tmp_map = dcb_ieee_getapp_mask(netdev, &app); + prio_map |= tmp_map; + + if (!prio_map) + prio_map = -EFAULT; + return prio_map; +} + +static void bnxt_re_parse_cid_map(u8 prio_map, u8 *cid_map, u16 *cosq) +{ + u16 prio; + u8 id; + + for (prio = 0, id = 0; prio < 8; prio++) { + if (prio_map & (1 << prio)) { + cosq[id] = cid_map[prio]; + id++; + if (id == 2) /* Max 2 tcs supported */ + break; + } + } +} + +static int bnxt_re_setup_qos(struct bnxt_re_dev *rdev) +{ + u8 prio_map = 0; + u64 cid_map; + int rc; + + /* Get priority for roce */ + rc = bnxt_re_get_priority_mask(rdev); + if (rc < 0) + return rc; + prio_map = (u8)rc; + + if (prio_map == rdev->cur_prio_map) + return 0; + rdev->cur_prio_map = prio_map; + /* Get cosq id for this priority */ + rc = bnxt_re_query_hwrm_pri2cos(rdev, 0, &cid_map); + if (rc) { + dev_warn(rdev_to_dev(rdev), "no cos for p_mask %x\n", prio_map); + return rc; + } + /* Parse CoS IDs for app priority */ + bnxt_re_parse_cid_map(prio_map, (u8 *)&cid_map, rdev->cosq); + + /* Config BONO. */ + rc = bnxt_qplib_map_tc2cos(&rdev->qplib_res, rdev->cosq); + if (rc) { + dev_warn(rdev_to_dev(rdev), "no tc for cos{%x, %x}\n", + rdev->cosq[0], rdev->cosq[1]); + return rc; + } + + return 0; +} + +static void bnxt_re_ib_unreg(struct bnxt_re_dev *rdev, bool lock_wait) +{ + int i, rc; + + if (test_and_clear_bit(BNXT_RE_FLAG_IBDEV_REGISTERED, &rdev->flags)) { + for (i = 0; i < ARRAY_SIZE(bnxt_re_attributes); i++) + device_remove_file(&rdev->ibdev.dev, + bnxt_re_attributes[i]); + /* Cleanup ib dev */ + bnxt_re_unregister_ib(rdev); + } + if (test_and_clear_bit(BNXT_RE_FLAG_QOS_WORK_REG, &rdev->flags)) + cancel_delayed_work(&rdev->worker); + + bnxt_re_cleanup_res(rdev); + bnxt_re_free_res(rdev, lock_wait); + + if (test_and_clear_bit(BNXT_RE_FLAG_RCFW_CHANNEL_EN, &rdev->flags)) { + rc = bnxt_qplib_deinit_rcfw(&rdev->rcfw); + if (rc) + dev_warn(rdev_to_dev(rdev), + "Failed to deinitialize RCFW: %#x", rc); + bnxt_re_net_stats_ctx_free(rdev, rdev->qplib_ctx.stats.fw_id, + lock_wait); + bnxt_qplib_free_ctx(rdev->en_dev->pdev, &rdev->qplib_ctx); + bnxt_qplib_disable_rcfw_channel(&rdev->rcfw); + bnxt_re_net_ring_free(rdev, rdev->rcfw.creq_ring_id, lock_wait); + bnxt_qplib_free_rcfw_channel(&rdev->rcfw); + } + if (test_and_clear_bit(BNXT_RE_FLAG_GOT_MSIX, &rdev->flags)) { + rc = bnxt_re_free_msix(rdev, lock_wait); + if (rc) + dev_warn(rdev_to_dev(rdev), + "Failed to free MSI-X vectors: %#x", rc); + } + if (test_and_clear_bit(BNXT_RE_FLAG_NETDEV_REGISTERED, &rdev->flags)) { + rc = bnxt_re_unregister_netdev(rdev, lock_wait); + if (rc) + dev_warn(rdev_to_dev(rdev), + "Failed to unregister with netdev: %#x", rc); + } +} + +static void bnxt_re_set_resource_limits(struct bnxt_re_dev *rdev) +{ + u32 i; + + rdev->qplib_ctx.qpc_count = BNXT_RE_MAX_QPC_COUNT; + rdev->qplib_ctx.mrw_count = BNXT_RE_MAX_MRW_COUNT; + rdev->qplib_ctx.srqc_count = BNXT_RE_MAX_SRQC_COUNT; + rdev->qplib_ctx.cq_count = BNXT_RE_MAX_CQ_COUNT; + for (i = 0; i < MAX_TQM_ALLOC_REQ; i++) + rdev->qplib_ctx.tqm_count[i] = + rdev->dev_attr.tqm_alloc_reqs[i]; +} + +/* worker thread for polling periodic events. Now used for QoS programming*/ +static void bnxt_re_worker(struct work_struct *work) +{ + struct bnxt_re_dev *rdev = container_of(work, struct bnxt_re_dev, + worker.work); + + bnxt_re_setup_qos(rdev); + schedule_delayed_work(&rdev->worker, msecs_to_jiffies(30000)); +} + +static int bnxt_re_ib_reg(struct bnxt_re_dev *rdev) +{ + int i, j, rc; + + /* Registered a new RoCE device instance to netdev */ + rc = bnxt_re_register_netdev(rdev); + if (rc) { + pr_err("Failed to register with netedev: %#x\n", rc); + return -EINVAL; + } + set_bit(BNXT_RE_FLAG_NETDEV_REGISTERED, &rdev->flags); + + rc = bnxt_re_request_msix(rdev); + if (rc) { + pr_err("Failed to get MSI-X vectors: %#x\n", rc); + rc = -EINVAL; + goto fail; + } + set_bit(BNXT_RE_FLAG_GOT_MSIX, &rdev->flags); + + /* Establish RCFW Communication Channel to initialize the context + * memory for the function and all child VFs + */ + rc = bnxt_qplib_alloc_rcfw_channel(rdev->en_dev->pdev, &rdev->rcfw); + if (rc) + goto fail; + + rc = bnxt_re_net_ring_alloc + (rdev, rdev->rcfw.creq.pbl[PBL_LVL_0].pg_map_arr, + rdev->rcfw.creq.pbl[rdev->rcfw.creq.level].pg_count, + HWRM_RING_ALLOC_CMPL, BNXT_QPLIB_CREQE_MAX_CNT - 1, + rdev->msix_entries[BNXT_RE_AEQ_IDX].ring_idx, + &rdev->rcfw.creq_ring_id); + if (rc) { + pr_err("Failed to allocate CREQ: %#x\n", rc); + goto free_rcfw; + } + rc = bnxt_qplib_enable_rcfw_channel + (rdev->en_dev->pdev, &rdev->rcfw, + rdev->msix_entries[BNXT_RE_AEQ_IDX].vector, + rdev->msix_entries[BNXT_RE_AEQ_IDX].db_offset, + 0, &bnxt_re_aeq_handler); + if (rc) { + pr_err("Failed to enable RCFW channel: %#x\n", rc); + goto free_ring; + } + + rc = bnxt_qplib_get_dev_attr(&rdev->rcfw, &rdev->dev_attr); + if (rc) + goto disable_rcfw; + bnxt_re_set_resource_limits(rdev); + + rc = bnxt_qplib_alloc_ctx(rdev->en_dev->pdev, &rdev->qplib_ctx, 0); + if (rc) { + pr_err("Failed to allocate QPLIB context: %#x\n", rc); + goto disable_rcfw; + } + rc = bnxt_re_net_stats_ctx_alloc(rdev, + rdev->qplib_ctx.stats.dma_map, + &rdev->qplib_ctx.stats.fw_id); + if (rc) { + pr_err("Failed to allocate stats context: %#x\n", rc); + goto free_ctx; + } + + rc = bnxt_qplib_init_rcfw(&rdev->rcfw, &rdev->qplib_ctx, 0); + if (rc) { + pr_err("Failed to initialize RCFW: %#x\n", rc); + goto free_sctx; + } + set_bit(BNXT_RE_FLAG_RCFW_CHANNEL_EN, &rdev->flags); + + /* Resources based on the 'new' device caps */ + rc = bnxt_re_alloc_res(rdev); + if (rc) { + pr_err("Failed to allocate resources: %#x\n", rc); + goto fail; + } + rc = bnxt_re_init_res(rdev); + if (rc) { + pr_err("Failed to initialize resources: %#x\n", rc); + goto fail; + } + + rc = bnxt_re_setup_qos(rdev); + if (rc) + pr_info("RoCE priority not yet configured\n"); + + INIT_DELAYED_WORK(&rdev->worker, bnxt_re_worker); + set_bit(BNXT_RE_FLAG_QOS_WORK_REG, &rdev->flags); + schedule_delayed_work(&rdev->worker, msecs_to_jiffies(30000)); + + /* Register ib dev */ + rc = bnxt_re_register_ib(rdev); + if (rc) { + pr_err("Failed to register with IB: %#x\n", rc); + goto fail; + } + dev_info(rdev_to_dev(rdev), "Device registered successfully"); + for (i = 0; i < ARRAY_SIZE(bnxt_re_attributes); i++) { + rc = device_create_file(&rdev->ibdev.dev, + bnxt_re_attributes[i]); + if (rc) { + dev_err(rdev_to_dev(rdev), + "Failed to create IB sysfs: %#x", rc); + /* Must clean up all created device files */ + for (j = 0; j < i; j++) + device_remove_file(&rdev->ibdev.dev, + bnxt_re_attributes[j]); + bnxt_re_unregister_ib(rdev); + goto fail; + } + } + set_bit(BNXT_RE_FLAG_IBDEV_REGISTERED, &rdev->flags); + bnxt_re_dispatch_event(&rdev->ibdev, NULL, 1, IB_EVENT_PORT_ACTIVE); + bnxt_re_dispatch_event(&rdev->ibdev, NULL, 1, IB_EVENT_GID_CHANGE); + + return 0; +free_sctx: + bnxt_re_net_stats_ctx_free(rdev, rdev->qplib_ctx.stats.fw_id, true); +free_ctx: + bnxt_qplib_free_ctx(rdev->en_dev->pdev, &rdev->qplib_ctx); +disable_rcfw: + bnxt_qplib_disable_rcfw_channel(&rdev->rcfw); +free_ring: + bnxt_re_net_ring_free(rdev, rdev->rcfw.creq_ring_id, true); +free_rcfw: + bnxt_qplib_free_rcfw_channel(&rdev->rcfw); +fail: + bnxt_re_ib_unreg(rdev, true); + return rc; +} + +static void bnxt_re_dev_unreg(struct bnxt_re_dev *rdev) +{ + struct bnxt_en_dev *en_dev = rdev->en_dev; + struct net_device *netdev = rdev->netdev; + + bnxt_re_dev_remove(rdev); + + if (netdev) + bnxt_re_dev_unprobe(netdev, en_dev); +} + +static int bnxt_re_dev_reg(struct bnxt_re_dev **rdev, struct net_device *netdev) +{ + struct bnxt_en_dev *en_dev; + int rc = 0; + + if (!is_bnxt_re_dev(netdev)) + return -ENODEV; + + en_dev = bnxt_re_dev_probe(netdev); + if (IS_ERR(en_dev)) { + if (en_dev != ERR_PTR(-ENODEV)) + pr_err("%s: Failed to probe\n", ROCE_DRV_MODULE_NAME); + rc = PTR_ERR(en_dev); + goto exit; + } + *rdev = bnxt_re_dev_add(netdev, en_dev); + if (!*rdev) { + rc = -ENOMEM; + bnxt_re_dev_unprobe(netdev, en_dev); + goto exit; + } +exit: + return rc; +} + +static void bnxt_re_remove_one(struct bnxt_re_dev *rdev) +{ + pci_dev_put(rdev->en_dev->pdev); +} + +/* Handle all deferred netevents tasks */ +static void bnxt_re_task(struct work_struct *work) +{ + struct bnxt_re_work *re_work; + struct bnxt_re_dev *rdev; + int rc = 0; + + re_work = container_of(work, struct bnxt_re_work, work); + rdev = re_work->rdev; + + if (re_work->event != NETDEV_REGISTER && + !test_bit(BNXT_RE_FLAG_IBDEV_REGISTERED, &rdev->flags)) + return; + + switch (re_work->event) { + case NETDEV_REGISTER: + rc = bnxt_re_ib_reg(rdev); + if (rc) + dev_err(rdev_to_dev(rdev), + "Failed to register with IB: %#x", rc); + break; + case NETDEV_UP: + bnxt_re_dispatch_event(&rdev->ibdev, NULL, 1, + IB_EVENT_PORT_ACTIVE); + break; + case NETDEV_DOWN: + bnxt_re_dev_stop(rdev); + break; + case NETDEV_CHANGE: + if (!netif_carrier_ok(rdev->netdev)) + bnxt_re_dev_stop(rdev); + else if (netif_carrier_ok(rdev->netdev)) + bnxt_re_dispatch_event(&rdev->ibdev, NULL, 1, + IB_EVENT_PORT_ACTIVE); + break; + default: + break; + } + kfree(re_work); +} + +static void bnxt_re_init_one(struct bnxt_re_dev *rdev) +{ + pci_dev_get(rdev->en_dev->pdev); +} + +/* + * "Notifier chain callback can be invoked for the same chain from + * different CPUs at the same time". + * + * For cases when the netdev is already present, our call to the + * register_netdevice_notifier() will actually get the rtnl_lock() + * before sending NETDEV_REGISTER and (if up) NETDEV_UP + * events. + * + * But for cases when the netdev is not already present, the notifier + * chain is subjected to be invoked from different CPUs simultaneously. + * + * This is protected by the netdev_mutex. + */ +static int bnxt_re_netdev_event(struct notifier_block *notifier, + unsigned long event, void *ptr) +{ + struct net_device *real_dev, *netdev = netdev_notifier_info_to_dev(ptr); + struct bnxt_re_work *re_work; + struct bnxt_re_dev *rdev; + int rc = 0; + bool sch_work = false; + + real_dev = rdma_vlan_dev_real_dev(netdev); + if (!real_dev) + real_dev = netdev; + + rdev = bnxt_re_from_netdev(real_dev); + if (!rdev && event != NETDEV_REGISTER) + goto exit; + if (real_dev != netdev) + goto exit; + + switch (event) { + case NETDEV_REGISTER: + if (rdev) + break; + rc = bnxt_re_dev_reg(&rdev, real_dev); + if (rc == -ENODEV) + break; + if (rc) { + pr_err("Failed to register with the device %s: %#x\n", + real_dev->name, rc); + break; + } + bnxt_re_init_one(rdev); + sch_work = true; + break; + + case NETDEV_UNREGISTER: + bnxt_re_ib_unreg(rdev, false); + bnxt_re_remove_one(rdev); + bnxt_re_dev_unreg(rdev); + break; + + default: + sch_work = true; + break; + } + if (sch_work) { + /* Allocate for the deferred task */ + re_work = kzalloc(sizeof(*re_work), GFP_ATOMIC); + if (re_work) { + re_work->rdev = rdev; + re_work->event = event; + re_work->vlan_dev = (real_dev == netdev ? + NULL : netdev); + INIT_WORK(&re_work->work, bnxt_re_task); + queue_work(bnxt_re_wq, &re_work->work); + } + } + +exit: + return NOTIFY_DONE; +} + +static struct notifier_block bnxt_re_netdev_notifier = { + .notifier_call = bnxt_re_netdev_event +}; + +static int __init bnxt_re_mod_init(void) +{ + int rc = 0; + + pr_info("%s: %s", ROCE_DRV_MODULE_NAME, version); + + bnxt_re_wq = create_singlethread_workqueue("bnxt_re"); + if (!bnxt_re_wq) + return -ENOMEM; + + INIT_LIST_HEAD(&bnxt_re_dev_list); + + rc = register_netdevice_notifier(&bnxt_re_netdev_notifier); + if (rc) { + pr_err("%s: Cannot register to netdevice_notifier", + ROCE_DRV_MODULE_NAME); + goto err_netdev; + } + return 0; + +err_netdev: + destroy_workqueue(bnxt_re_wq); + + return rc; +} + +static void __exit bnxt_re_mod_exit(void) +{ + unregister_netdevice_notifier(&bnxt_re_netdev_notifier); + if (bnxt_re_wq) + destroy_workqueue(bnxt_re_wq); +} + +module_init(bnxt_re_mod_init); +module_exit(bnxt_re_mod_exit); diff --git a/drivers/infiniband/hw/bnxt_re/qplib_fp.c b/drivers/infiniband/hw/bnxt_re/qplib_fp.c new file mode 100644 index 000000000000..9af1514e5944 --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/qplib_fp.c @@ -0,0 +1,2210 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * 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 + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. 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. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: Fast Path Operators + */ + +#include <linux/interrupt.h> +#include <linux/spinlock.h> +#include <linux/sched.h> +#include <linux/slab.h> +#include <linux/pci.h> +#include <linux/prefetch.h> + +#include "roce_hsi.h" + +#include "qplib_res.h" +#include "qplib_rcfw.h" +#include "qplib_sp.h" +#include "qplib_fp.h" + +static void bnxt_qplib_arm_cq_enable(struct bnxt_qplib_cq *cq); + +static void bnxt_qplib_free_qp_hdr_buf(struct bnxt_qplib_res *res, + struct bnxt_qplib_qp *qp) +{ + struct bnxt_qplib_q *rq = &qp->rq; + struct bnxt_qplib_q *sq = &qp->sq; + + if (qp->rq_hdr_buf) + dma_free_coherent(&res->pdev->dev, + rq->hwq.max_elements * qp->rq_hdr_buf_size, + qp->rq_hdr_buf, qp->rq_hdr_buf_map); + if (qp->sq_hdr_buf) + dma_free_coherent(&res->pdev->dev, + sq->hwq.max_elements * qp->sq_hdr_buf_size, + qp->sq_hdr_buf, qp->sq_hdr_buf_map); + qp->rq_hdr_buf = NULL; + qp->sq_hdr_buf = NULL; + qp->rq_hdr_buf_map = 0; + qp->sq_hdr_buf_map = 0; + qp->sq_hdr_buf_size = 0; + qp->rq_hdr_buf_size = 0; +} + +static int bnxt_qplib_alloc_qp_hdr_buf(struct bnxt_qplib_res *res, + struct bnxt_qplib_qp *qp) +{ + struct bnxt_qplib_q *rq = &qp->rq; + struct bnxt_qplib_q *sq = &qp->rq; + int rc = 0; + + if (qp->sq_hdr_buf_size && sq->hwq.max_elements) { + qp->sq_hdr_buf = dma_alloc_coherent(&res->pdev->dev, + sq->hwq.max_elements * + qp->sq_hdr_buf_size, + &qp->sq_hdr_buf_map, GFP_KERNEL); + if (!qp->sq_hdr_buf) { + rc = -ENOMEM; + dev_err(&res->pdev->dev, + "QPLIB: Failed to create sq_hdr_buf"); + goto fail; + } + } + + if (qp->rq_hdr_buf_size && rq->hwq.max_elements) { + qp->rq_hdr_buf = dma_alloc_coherent(&res->pdev->dev, + rq->hwq.max_elements * + qp->rq_hdr_buf_size, + &qp->rq_hdr_buf_map, + GFP_KERNEL); + if (!qp->rq_hdr_buf) { + rc = -ENOMEM; + dev_err(&res->pdev->dev, + "QPLIB: Failed to create rq_hdr_buf"); + goto fail; + } + } + return 0; + +fail: + bnxt_qplib_free_qp_hdr_buf(res, qp); + return rc; +} + +static void bnxt_qplib_service_nq(unsigned long data) +{ + struct bnxt_qplib_nq *nq = (struct bnxt_qplib_nq *)data; + struct bnxt_qplib_hwq *hwq = &nq->hwq; + struct nq_base *nqe, **nq_ptr; + int num_cqne_processed = 0; + u32 sw_cons, raw_cons; + u16 type; + int budget = nq->budget; + u64 q_handle; + + /* Service the NQ until empty */ + raw_cons = hwq->cons; + while (budget--) { + sw_cons = HWQ_CMP(raw_cons, hwq); + nq_ptr = (struct nq_base **)hwq->pbl_ptr; + nqe = &nq_ptr[NQE_PG(sw_cons)][NQE_IDX(sw_cons)]; + if (!NQE_CMP_VALID(nqe, raw_cons, hwq->max_elements)) + break; + + type = le16_to_cpu(nqe->info10_type) & NQ_BASE_TYPE_MASK; + switch (type) { + case NQ_BASE_TYPE_CQ_NOTIFICATION: + { + struct nq_cn *nqcne = (struct nq_cn *)nqe; + + q_handle = le32_to_cpu(nqcne->cq_handle_low); + q_handle |= (u64)le32_to_cpu(nqcne->cq_handle_high) + << 32; + bnxt_qplib_arm_cq_enable((struct bnxt_qplib_cq *) + ((unsigned long)q_handle)); + if (!nq->cqn_handler(nq, (struct bnxt_qplib_cq *) + ((unsigned long)q_handle))) + num_cqne_processed++; + else + dev_warn(&nq->pdev->dev, + "QPLIB: cqn - type 0x%x not handled", + type); + break; + } + case NQ_BASE_TYPE_DBQ_EVENT: + break; + default: + dev_warn(&nq->pdev->dev, + "QPLIB: nqe with type = 0x%x not handled", + type); + break; + } + raw_cons++; + } + if (hwq->cons != raw_cons) { + hwq->cons = raw_cons; + NQ_DB_REARM(nq->bar_reg_iomem, hwq->cons, hwq->max_elements); + } +} + +static irqreturn_t bnxt_qplib_nq_irq(int irq, void *dev_instance) +{ + struct bnxt_qplib_nq *nq = dev_instance; + struct bnxt_qplib_hwq *hwq = &nq->hwq; + struct nq_base **nq_ptr; + u32 sw_cons; + + /* Prefetch the NQ element */ + sw_cons = HWQ_CMP(hwq->cons, hwq); + nq_ptr = (struct nq_base **)nq->hwq.pbl_ptr; + prefetch(&nq_ptr[NQE_PG(sw_cons)][NQE_IDX(sw_cons)]); + + /* Fan out to CPU affinitized kthreads? */ + tasklet_schedule(&nq->worker); + + return IRQ_HANDLED; +} + +void bnxt_qplib_disable_nq(struct bnxt_qplib_nq *nq) +{ + /* Make sure the HW is stopped! */ + synchronize_irq(nq->vector); + tasklet_disable(&nq->worker); + tasklet_kill(&nq->worker); + + if (nq->requested) { + free_irq(nq->vector, nq); + nq->requested = false; + } + if (nq->bar_reg_iomem) + iounmap(nq->bar_reg_iomem); + nq->bar_reg_iomem = NULL; + + nq->cqn_handler = NULL; + nq->srqn_handler = NULL; + nq->vector = 0; +} + +int bnxt_qplib_enable_nq(struct pci_dev *pdev, struct bnxt_qplib_nq *nq, + int msix_vector, int bar_reg_offset, + int (*cqn_handler)(struct bnxt_qplib_nq *nq, + struct bnxt_qplib_cq *), + int (*srqn_handler)(struct bnxt_qplib_nq *nq, + void *, u8 event)) +{ + resource_size_t nq_base; + int rc; + + nq->pdev = pdev; + nq->vector = msix_vector; + + nq->cqn_handler = cqn_handler; + + nq->srqn_handler = srqn_handler; + + tasklet_init(&nq->worker, bnxt_qplib_service_nq, (unsigned long)nq); + + nq->requested = false; + rc = request_irq(nq->vector, bnxt_qplib_nq_irq, 0, "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; + } + nq->requested = true; + nq->bar_reg = NQ_CONS_PCI_BAR_REGION; + nq->bar_reg_off = bar_reg_offset; + nq_base = pci_resource_start(pdev, nq->bar_reg); + if (!nq_base) { + rc = -ENOMEM; + goto fail; + } + nq->bar_reg_iomem = ioremap_nocache(nq_base + nq->bar_reg_off, 4); + if (!nq->bar_reg_iomem) { + rc = -ENOMEM; + goto fail; + } + NQ_DB_REARM(nq->bar_reg_iomem, nq->hwq.cons, nq->hwq.max_elements); + + return 0; +fail: + bnxt_qplib_disable_nq(nq); + return rc; +} + +void bnxt_qplib_free_nq(struct bnxt_qplib_nq *nq) +{ + if (nq->hwq.max_elements) + bnxt_qplib_free_hwq(nq->pdev, &nq->hwq); +} + +int bnxt_qplib_alloc_nq(struct pci_dev *pdev, struct bnxt_qplib_nq *nq) +{ + nq->pdev = pdev; + if (!nq->hwq.max_elements || + nq->hwq.max_elements > BNXT_QPLIB_NQE_MAX_CNT) + nq->hwq.max_elements = BNXT_QPLIB_NQE_MAX_CNT; + + if (bnxt_qplib_alloc_init_hwq(nq->pdev, &nq->hwq, NULL, 0, + &nq->hwq.max_elements, + BNXT_QPLIB_MAX_NQE_ENTRY_SIZE, 0, + PAGE_SIZE, HWQ_TYPE_L2_CMPL)) + return -ENOMEM; + + nq->budget = 8; + return 0; +} + +/* QP */ +int bnxt_qplib_create_qp1(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_create_qp1 req; + struct creq_create_qp1_resp resp; + struct bnxt_qplib_pbl *pbl; + struct bnxt_qplib_q *sq = &qp->sq; + struct bnxt_qplib_q *rq = &qp->rq; + int rc; + u16 cmd_flags = 0; + u32 qp_flags = 0; + + RCFW_CMD_PREP(req, CREATE_QP1, cmd_flags); + + /* General */ + req.type = qp->type; + req.dpi = cpu_to_le32(qp->dpi->dpi); + req.qp_handle = cpu_to_le64(qp->qp_handle); + + /* SQ */ + sq->hwq.max_elements = sq->max_wqe; + rc = bnxt_qplib_alloc_init_hwq(res->pdev, &sq->hwq, NULL, 0, + &sq->hwq.max_elements, + BNXT_QPLIB_MAX_SQE_ENTRY_SIZE, 0, + PAGE_SIZE, HWQ_TYPE_QUEUE); + if (rc) + goto exit; + + sq->swq = kcalloc(sq->hwq.max_elements, sizeof(*sq->swq), GFP_KERNEL); + if (!sq->swq) { + rc = -ENOMEM; + goto fail_sq; + } + pbl = &sq->hwq.pbl[PBL_LVL_0]; + req.sq_pbl = cpu_to_le64(pbl->pg_map_arr[0]); + req.sq_pg_size_sq_lvl = + ((sq->hwq.level & CMDQ_CREATE_QP1_SQ_LVL_MASK) + << CMDQ_CREATE_QP1_SQ_LVL_SFT) | + (pbl->pg_size == ROCE_PG_SIZE_4K ? + CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_4K : + pbl->pg_size == ROCE_PG_SIZE_8K ? + CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_8K : + pbl->pg_size == ROCE_PG_SIZE_64K ? + CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_64K : + pbl->pg_size == ROCE_PG_SIZE_2M ? + CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_2M : + pbl->pg_size == ROCE_PG_SIZE_8M ? + CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_8M : + pbl->pg_size == ROCE_PG_SIZE_1G ? + CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_1G : + CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_4K); + + if (qp->scq) + req.scq_cid = cpu_to_le32(qp->scq->id); + + qp_flags |= CMDQ_CREATE_QP1_QP_FLAGS_RESERVED_LKEY_ENABLE; + + /* RQ */ + if (rq->max_wqe) { + rq->hwq.max_elements = qp->rq.max_wqe; + rc = bnxt_qplib_alloc_init_hwq(res->pdev, &rq->hwq, NULL, 0, + &rq->hwq.max_elements, + BNXT_QPLIB_MAX_RQE_ENTRY_SIZE, 0, + PAGE_SIZE, HWQ_TYPE_QUEUE); + if (rc) + goto fail_sq; + + rq->swq = kcalloc(rq->hwq.max_elements, sizeof(*rq->swq), + GFP_KERNEL); + if (!rq->swq) { + rc = -ENOMEM; + goto fail_rq; + } + pbl = &rq->hwq.pbl[PBL_LVL_0]; + req.rq_pbl = cpu_to_le64(pbl->pg_map_arr[0]); + req.rq_pg_size_rq_lvl = + ((rq->hwq.level & CMDQ_CREATE_QP1_RQ_LVL_MASK) << + CMDQ_CREATE_QP1_RQ_LVL_SFT) | + (pbl->pg_size == ROCE_PG_SIZE_4K ? + CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_4K : + pbl->pg_size == ROCE_PG_SIZE_8K ? + CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_8K : + pbl->pg_size == ROCE_PG_SIZE_64K ? + CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_64K : + pbl->pg_size == ROCE_PG_SIZE_2M ? + CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_2M : + pbl->pg_size == ROCE_PG_SIZE_8M ? + CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_8M : + pbl->pg_size == ROCE_PG_SIZE_1G ? + CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_1G : + CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_4K); + if (qp->rcq) + req.rcq_cid = cpu_to_le32(qp->rcq->id); + } + + /* Header buffer - allow hdr_buf pass in */ + rc = bnxt_qplib_alloc_qp_hdr_buf(res, qp); + if (rc) { + rc = -ENOMEM; + goto fail; + } + req.qp_flags = cpu_to_le32(qp_flags); + req.sq_size = cpu_to_le32(sq->hwq.max_elements); + req.rq_size = cpu_to_le32(rq->hwq.max_elements); + + req.sq_fwo_sq_sge = + cpu_to_le16((sq->max_sge & CMDQ_CREATE_QP1_SQ_SGE_MASK) << + CMDQ_CREATE_QP1_SQ_SGE_SFT); + req.rq_fwo_rq_sge = + cpu_to_le16((rq->max_sge & CMDQ_CREATE_QP1_RQ_SGE_MASK) << + CMDQ_CREATE_QP1_RQ_SGE_SFT); + + req.pd_id = cpu_to_le32(qp->pd->id); + + rc = bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + (void *)&resp, NULL, 0); + if (rc) + goto fail; + + qp->id = le32_to_cpu(resp.xid); + qp->cur_qp_state = CMDQ_MODIFY_QP_NEW_STATE_RESET; + sq->flush_in_progress = false; + rq->flush_in_progress = false; + + return 0; + +fail: + bnxt_qplib_free_qp_hdr_buf(res, qp); +fail_rq: + bnxt_qplib_free_hwq(res->pdev, &rq->hwq); + kfree(rq->swq); +fail_sq: + bnxt_qplib_free_hwq(res->pdev, &sq->hwq); + kfree(sq->swq); +exit: + return rc; +} + +int bnxt_qplib_create_qp(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct sq_send *hw_sq_send_hdr, **hw_sq_send_ptr; + struct cmdq_create_qp req; + struct creq_create_qp_resp resp; + struct bnxt_qplib_pbl *pbl; + struct sq_psn_search **psn_search_ptr; + unsigned long int psn_search, poff = 0; + struct bnxt_qplib_q *sq = &qp->sq; + struct bnxt_qplib_q *rq = &qp->rq; + struct bnxt_qplib_hwq *xrrq; + int i, rc, req_size, psn_sz; + u16 cmd_flags = 0, max_ssge; + u32 sw_prod, qp_flags = 0; + + RCFW_CMD_PREP(req, CREATE_QP, cmd_flags); + + /* General */ + req.type = qp->type; + req.dpi = cpu_to_le32(qp->dpi->dpi); + req.qp_handle = cpu_to_le64(qp->qp_handle); + + /* SQ */ + psn_sz = (qp->type == CMDQ_CREATE_QP_TYPE_RC) ? + sizeof(struct sq_psn_search) : 0; + sq->hwq.max_elements = sq->max_wqe; + rc = bnxt_qplib_alloc_init_hwq(res->pdev, &sq->hwq, sq->sglist, + sq->nmap, &sq->hwq.max_elements, + BNXT_QPLIB_MAX_SQE_ENTRY_SIZE, + psn_sz, + PAGE_SIZE, HWQ_TYPE_QUEUE); + if (rc) + goto exit; + + sq->swq = kcalloc(sq->hwq.max_elements, sizeof(*sq->swq), GFP_KERNEL); + if (!sq->swq) { + rc = -ENOMEM; + goto fail_sq; + } + hw_sq_send_ptr = (struct sq_send **)sq->hwq.pbl_ptr; + if (psn_sz) { + psn_search_ptr = (struct sq_psn_search **) + &hw_sq_send_ptr[get_sqe_pg + (sq->hwq.max_elements)]; + psn_search = (unsigned long int) + &hw_sq_send_ptr[get_sqe_pg(sq->hwq.max_elements)] + [get_sqe_idx(sq->hwq.max_elements)]; + if (psn_search & ~PAGE_MASK) { + /* If the psn_search does not start on a page boundary, + * then calculate the offset + */ + poff = (psn_search & ~PAGE_MASK) / + BNXT_QPLIB_MAX_PSNE_ENTRY_SIZE; + } + for (i = 0; i < sq->hwq.max_elements; i++) + sq->swq[i].psn_search = + &psn_search_ptr[get_psne_pg(i + poff)] + [get_psne_idx(i + poff)]; + } + pbl = &sq->hwq.pbl[PBL_LVL_0]; + req.sq_pbl = cpu_to_le64(pbl->pg_map_arr[0]); + req.sq_pg_size_sq_lvl = + ((sq->hwq.level & CMDQ_CREATE_QP_SQ_LVL_MASK) + << CMDQ_CREATE_QP_SQ_LVL_SFT) | + (pbl->pg_size == ROCE_PG_SIZE_4K ? + CMDQ_CREATE_QP_SQ_PG_SIZE_PG_4K : + pbl->pg_size == ROCE_PG_SIZE_8K ? + CMDQ_CREATE_QP_SQ_PG_SIZE_PG_8K : + pbl->pg_size == ROCE_PG_SIZE_64K ? + CMDQ_CREATE_QP_SQ_PG_SIZE_PG_64K : + pbl->pg_size == ROCE_PG_SIZE_2M ? + CMDQ_CREATE_QP_SQ_PG_SIZE_PG_2M : + pbl->pg_size == ROCE_PG_SIZE_8M ? + CMDQ_CREATE_QP_SQ_PG_SIZE_PG_8M : + pbl->pg_size == ROCE_PG_SIZE_1G ? + CMDQ_CREATE_QP_SQ_PG_SIZE_PG_1G : + CMDQ_CREATE_QP_SQ_PG_SIZE_PG_4K); + + /* initialize all SQ WQEs to LOCAL_INVALID (sq prep for hw fetch) */ + hw_sq_send_ptr = (struct sq_send **)sq->hwq.pbl_ptr; + for (sw_prod = 0; sw_prod < sq->hwq.max_elements; sw_prod++) { + hw_sq_send_hdr = &hw_sq_send_ptr[get_sqe_pg(sw_prod)] + [get_sqe_idx(sw_prod)]; + hw_sq_send_hdr->wqe_type = SQ_BASE_WQE_TYPE_LOCAL_INVALID; + } + + if (qp->scq) + req.scq_cid = cpu_to_le32(qp->scq->id); + + qp_flags |= CMDQ_CREATE_QP_QP_FLAGS_RESERVED_LKEY_ENABLE; + qp_flags |= CMDQ_CREATE_QP_QP_FLAGS_FR_PMR_ENABLED; + if (qp->sig_type) + qp_flags |= CMDQ_CREATE_QP_QP_FLAGS_FORCE_COMPLETION; + + /* RQ */ + if (rq->max_wqe) { + rq->hwq.max_elements = rq->max_wqe; + rc = bnxt_qplib_alloc_init_hwq(res->pdev, &rq->hwq, rq->sglist, + rq->nmap, &rq->hwq.max_elements, + BNXT_QPLIB_MAX_RQE_ENTRY_SIZE, 0, + PAGE_SIZE, HWQ_TYPE_QUEUE); + if (rc) + goto fail_sq; + + rq->swq = kcalloc(rq->hwq.max_elements, sizeof(*rq->swq), + GFP_KERNEL); + if (!rq->swq) { + rc = -ENOMEM; + goto fail_rq; + } + pbl = &rq->hwq.pbl[PBL_LVL_0]; + req.rq_pbl = cpu_to_le64(pbl->pg_map_arr[0]); + req.rq_pg_size_rq_lvl = + ((rq->hwq.level & CMDQ_CREATE_QP_RQ_LVL_MASK) << + CMDQ_CREATE_QP_RQ_LVL_SFT) | + (pbl->pg_size == ROCE_PG_SIZE_4K ? + CMDQ_CREATE_QP_RQ_PG_SIZE_PG_4K : + pbl->pg_size == ROCE_PG_SIZE_8K ? + CMDQ_CREATE_QP_RQ_PG_SIZE_PG_8K : + pbl->pg_size == ROCE_PG_SIZE_64K ? + CMDQ_CREATE_QP_RQ_PG_SIZE_PG_64K : + pbl->pg_size == ROCE_PG_SIZE_2M ? + CMDQ_CREATE_QP_RQ_PG_SIZE_PG_2M : + pbl->pg_size == ROCE_PG_SIZE_8M ? + CMDQ_CREATE_QP_RQ_PG_SIZE_PG_8M : + pbl->pg_size == ROCE_PG_SIZE_1G ? + CMDQ_CREATE_QP_RQ_PG_SIZE_PG_1G : + CMDQ_CREATE_QP_RQ_PG_SIZE_PG_4K); + } + + if (qp->rcq) + req.rcq_cid = cpu_to_le32(qp->rcq->id); + req.qp_flags = cpu_to_le32(qp_flags); + req.sq_size = cpu_to_le32(sq->hwq.max_elements); + req.rq_size = cpu_to_le32(rq->hwq.max_elements); + qp->sq_hdr_buf = NULL; + qp->rq_hdr_buf = NULL; + + rc = bnxt_qplib_alloc_qp_hdr_buf(res, qp); + if (rc) + goto fail_rq; + + /* CTRL-22434: Irrespective of the requested SGE count on the SQ + * always create the QP with max send sges possible if the requested + * inline size is greater than 0. + */ + max_ssge = qp->max_inline_data ? 6 : sq->max_sge; + req.sq_fwo_sq_sge = cpu_to_le16( + ((max_ssge & CMDQ_CREATE_QP_SQ_SGE_MASK) + << CMDQ_CREATE_QP_SQ_SGE_SFT) | 0); + req.rq_fwo_rq_sge = cpu_to_le16( + ((rq->max_sge & CMDQ_CREATE_QP_RQ_SGE_MASK) + << CMDQ_CREATE_QP_RQ_SGE_SFT) | 0); + /* ORRQ and IRRQ */ + if (psn_sz) { + xrrq = &qp->orrq; + xrrq->max_elements = + ORD_LIMIT_TO_ORRQ_SLOTS(qp->max_rd_atomic); + req_size = xrrq->max_elements * + BNXT_QPLIB_MAX_ORRQE_ENTRY_SIZE + PAGE_SIZE - 1; + req_size &= ~(PAGE_SIZE - 1); + rc = bnxt_qplib_alloc_init_hwq(res->pdev, xrrq, NULL, 0, + &xrrq->max_elements, + BNXT_QPLIB_MAX_ORRQE_ENTRY_SIZE, + 0, req_size, HWQ_TYPE_CTX); + if (rc) + goto fail_buf_free; + pbl = &xrrq->pbl[PBL_LVL_0]; + req.orrq_addr = cpu_to_le64(pbl->pg_map_arr[0]); + + xrrq = &qp->irrq; + xrrq->max_elements = IRD_LIMIT_TO_IRRQ_SLOTS( + qp->max_dest_rd_atomic); + req_size = xrrq->max_elements * + BNXT_QPLIB_MAX_IRRQE_ENTRY_SIZE + PAGE_SIZE - 1; + req_size &= ~(PAGE_SIZE - 1); + + rc = bnxt_qplib_alloc_init_hwq(res->pdev, xrrq, NULL, 0, + &xrrq->max_elements, + BNXT_QPLIB_MAX_IRRQE_ENTRY_SIZE, + 0, req_size, HWQ_TYPE_CTX); + if (rc) + goto fail_orrq; + + pbl = &xrrq->pbl[PBL_LVL_0]; + req.irrq_addr = cpu_to_le64(pbl->pg_map_arr[0]); + } + req.pd_id = cpu_to_le32(qp->pd->id); + + rc = bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + (void *)&resp, NULL, 0); + if (rc) + goto fail; + + qp->id = le32_to_cpu(resp.xid); + qp->cur_qp_state = CMDQ_MODIFY_QP_NEW_STATE_RESET; + sq->flush_in_progress = false; + rq->flush_in_progress = false; + + return 0; + +fail: + if (qp->irrq.max_elements) + bnxt_qplib_free_hwq(res->pdev, &qp->irrq); +fail_orrq: + if (qp->orrq.max_elements) + bnxt_qplib_free_hwq(res->pdev, &qp->orrq); +fail_buf_free: + bnxt_qplib_free_qp_hdr_buf(res, qp); +fail_rq: + bnxt_qplib_free_hwq(res->pdev, &rq->hwq); + kfree(rq->swq); +fail_sq: + bnxt_qplib_free_hwq(res->pdev, &sq->hwq); + kfree(sq->swq); +exit: + return rc; +} + +static void __modify_flags_from_init_state(struct bnxt_qplib_qp *qp) +{ + switch (qp->state) { + case CMDQ_MODIFY_QP_NEW_STATE_RTR: + /* INIT->RTR, configure the path_mtu to the default + * 2048 if not being requested + */ + if (!(qp->modify_flags & + CMDQ_MODIFY_QP_MODIFY_MASK_PATH_MTU)) { + qp->modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_PATH_MTU; + qp->path_mtu = + CMDQ_MODIFY_QP_PATH_MTU_MTU_2048; + } + qp->modify_flags &= + ~CMDQ_MODIFY_QP_MODIFY_MASK_VLAN_ID; + /* Bono FW require the max_dest_rd_atomic to be >= 1 */ + if (qp->max_dest_rd_atomic < 1) + qp->max_dest_rd_atomic = 1; + qp->modify_flags &= ~CMDQ_MODIFY_QP_MODIFY_MASK_SRC_MAC; + /* Bono FW 20.6.5 requires SGID_INDEX configuration */ + if (!(qp->modify_flags & + CMDQ_MODIFY_QP_MODIFY_MASK_SGID_INDEX)) { + qp->modify_flags |= + CMDQ_MODIFY_QP_MODIFY_MASK_SGID_INDEX; + qp->ah.sgid_index = 0; + } + break; + default: + break; + } +} + +static void __modify_flags_from_rtr_state(struct bnxt_qplib_qp *qp) +{ + switch (qp->state) { + case CMDQ_MODIFY_QP_NEW_STATE_RTS: + /* Bono FW requires the max_rd_atomic to be >= 1 */ + if (qp->max_rd_atomic < 1) + qp->max_rd_atomic = 1; + /* Bono FW does not allow PKEY_INDEX, + * DGID, FLOW_LABEL, SGID_INDEX, HOP_LIMIT, + * TRAFFIC_CLASS, DEST_MAC, PATH_MTU, RQ_PSN, + * MIN_RNR_TIMER, MAX_DEST_RD_ATOMIC, DEST_QP_ID + * modification + */ + qp->modify_flags &= + ~(CMDQ_MODIFY_QP_MODIFY_MASK_PKEY | + CMDQ_MODIFY_QP_MODIFY_MASK_DGID | + CMDQ_MODIFY_QP_MODIFY_MASK_FLOW_LABEL | + CMDQ_MODIFY_QP_MODIFY_MASK_SGID_INDEX | + CMDQ_MODIFY_QP_MODIFY_MASK_HOP_LIMIT | + CMDQ_MODIFY_QP_MODIFY_MASK_TRAFFIC_CLASS | + CMDQ_MODIFY_QP_MODIFY_MASK_DEST_MAC | + CMDQ_MODIFY_QP_MODIFY_MASK_PATH_MTU | + CMDQ_MODIFY_QP_MODIFY_MASK_RQ_PSN | + CMDQ_MODIFY_QP_MODIFY_MASK_MIN_RNR_TIMER | + CMDQ_MODIFY_QP_MODIFY_MASK_MAX_DEST_RD_ATOMIC | + CMDQ_MODIFY_QP_MODIFY_MASK_DEST_QP_ID); + break; + default: + break; + } +} + +static void __filter_modify_flags(struct bnxt_qplib_qp *qp) +{ + switch (qp->cur_qp_state) { + case CMDQ_MODIFY_QP_NEW_STATE_RESET: + break; + case CMDQ_MODIFY_QP_NEW_STATE_INIT: + __modify_flags_from_init_state(qp); + break; + case CMDQ_MODIFY_QP_NEW_STATE_RTR: + __modify_flags_from_rtr_state(qp); + break; + case CMDQ_MODIFY_QP_NEW_STATE_RTS: + break; + case CMDQ_MODIFY_QP_NEW_STATE_SQD: + break; + case CMDQ_MODIFY_QP_NEW_STATE_SQE: + break; + case CMDQ_MODIFY_QP_NEW_STATE_ERR: + break; + default: + break; + } +} + +int bnxt_qplib_modify_qp(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_modify_qp req; + struct creq_modify_qp_resp resp; + u16 cmd_flags = 0, pkey; + u32 temp32[4]; + u32 bmask; + int rc; + + RCFW_CMD_PREP(req, MODIFY_QP, cmd_flags); + + /* Filter out the qp_attr_mask based on the state->new transition */ + __filter_modify_flags(qp); + bmask = qp->modify_flags; + req.modify_mask = cpu_to_le32(qp->modify_flags); + req.qp_cid = cpu_to_le32(qp->id); + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_STATE) { + req.network_type_en_sqd_async_notify_new_state = + (qp->state & CMDQ_MODIFY_QP_NEW_STATE_MASK) | + (qp->en_sqd_async_notify ? + CMDQ_MODIFY_QP_EN_SQD_ASYNC_NOTIFY : 0); + } + req.network_type_en_sqd_async_notify_new_state |= qp->nw_type; + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_ACCESS) + req.access = qp->access; + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_PKEY) { + if (!bnxt_qplib_get_pkey(res, &res->pkey_tbl, + qp->pkey_index, &pkey)) + req.pkey = cpu_to_le16(pkey); + } + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_QKEY) + req.qkey = cpu_to_le32(qp->qkey); + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_DGID) { + memcpy(temp32, qp->ah.dgid.data, sizeof(struct bnxt_qplib_gid)); + req.dgid[0] = cpu_to_le32(temp32[0]); + req.dgid[1] = cpu_to_le32(temp32[1]); + req.dgid[2] = cpu_to_le32(temp32[2]); + req.dgid[3] = cpu_to_le32(temp32[3]); + } + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_FLOW_LABEL) + req.flow_label = cpu_to_le32(qp->ah.flow_label); + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_SGID_INDEX) + req.sgid_index = cpu_to_le16(res->sgid_tbl.hw_id + [qp->ah.sgid_index]); + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_HOP_LIMIT) + req.hop_limit = qp->ah.hop_limit; + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_TRAFFIC_CLASS) + req.traffic_class = qp->ah.traffic_class; + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_DEST_MAC) + memcpy(req.dest_mac, qp->ah.dmac, 6); + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_PATH_MTU) + req.path_mtu = qp->path_mtu; + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_TIMEOUT) + req.timeout = qp->timeout; + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_RETRY_CNT) + req.retry_cnt = qp->retry_cnt; + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_RNR_RETRY) + req.rnr_retry = qp->rnr_retry; + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_MIN_RNR_TIMER) + req.min_rnr_timer = qp->min_rnr_timer; + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_RQ_PSN) + req.rq_psn = cpu_to_le32(qp->rq.psn); + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_SQ_PSN) + req.sq_psn = cpu_to_le32(qp->sq.psn); + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_MAX_RD_ATOMIC) + req.max_rd_atomic = + ORD_LIMIT_TO_ORRQ_SLOTS(qp->max_rd_atomic); + + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_MAX_DEST_RD_ATOMIC) + req.max_dest_rd_atomic = + IRD_LIMIT_TO_IRRQ_SLOTS(qp->max_dest_rd_atomic); + + req.sq_size = cpu_to_le32(qp->sq.hwq.max_elements); + req.rq_size = cpu_to_le32(qp->rq.hwq.max_elements); + req.sq_sge = cpu_to_le16(qp->sq.max_sge); + req.rq_sge = cpu_to_le16(qp->rq.max_sge); + req.max_inline_data = cpu_to_le32(qp->max_inline_data); + if (bmask & CMDQ_MODIFY_QP_MODIFY_MASK_DEST_QP_ID) + req.dest_qp_id = cpu_to_le32(qp->dest_qpn); + + req.vlan_pcp_vlan_dei_vlan_id = cpu_to_le16(qp->vlan_id); + + rc = bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + (void *)&resp, NULL, 0); + if (rc) + return rc; + qp->cur_qp_state = qp->state; + return 0; +} + +int bnxt_qplib_query_qp(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_query_qp req; + struct creq_query_qp_resp resp; + struct bnxt_qplib_rcfw_sbuf *sbuf; + struct creq_query_qp_resp_sb *sb; + u16 cmd_flags = 0; + u32 temp32[4]; + int i, rc = 0; + + RCFW_CMD_PREP(req, QUERY_QP, cmd_flags); + + sbuf = bnxt_qplib_rcfw_alloc_sbuf(rcfw, sizeof(*sb)); + if (!sbuf) + return -ENOMEM; + sb = sbuf->sb; + + req.qp_cid = cpu_to_le32(qp->id); + req.resp_size = sizeof(*sb) / BNXT_QPLIB_CMDQE_UNITS; + rc = bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, (void *)&resp, + (void *)sbuf, 0); + if (rc) + goto bail; + /* Extract the context from the side buffer */ + qp->state = sb->en_sqd_async_notify_state & + CREQ_QUERY_QP_RESP_SB_STATE_MASK; + qp->en_sqd_async_notify = sb->en_sqd_async_notify_state & + CREQ_QUERY_QP_RESP_SB_EN_SQD_ASYNC_NOTIFY ? + true : false; + qp->access = sb->access; + qp->pkey_index = le16_to_cpu(sb->pkey); + qp->qkey = le32_to_cpu(sb->qkey); + + temp32[0] = le32_to_cpu(sb->dgid[0]); + temp32[1] = le32_to_cpu(sb->dgid[1]); + temp32[2] = le32_to_cpu(sb->dgid[2]); + temp32[3] = le32_to_cpu(sb->dgid[3]); + memcpy(qp->ah.dgid.data, temp32, sizeof(qp->ah.dgid.data)); + + qp->ah.flow_label = le32_to_cpu(sb->flow_label); + + qp->ah.sgid_index = 0; + for (i = 0; i < res->sgid_tbl.max; i++) { + if (res->sgid_tbl.hw_id[i] == le16_to_cpu(sb->sgid_index)) { + qp->ah.sgid_index = i; + break; + } + } + if (i == res->sgid_tbl.max) + dev_warn(&res->pdev->dev, "QPLIB: SGID not found??"); + + qp->ah.hop_limit = sb->hop_limit; + qp->ah.traffic_class = sb->traffic_class; + memcpy(qp->ah.dmac, sb->dest_mac, 6); + qp->ah.vlan_id = (le16_to_cpu(sb->path_mtu_dest_vlan_id) & + CREQ_QUERY_QP_RESP_SB_VLAN_ID_MASK) >> + CREQ_QUERY_QP_RESP_SB_VLAN_ID_SFT; + qp->path_mtu = (le16_to_cpu(sb->path_mtu_dest_vlan_id) & + CREQ_QUERY_QP_RESP_SB_PATH_MTU_MASK) >> + CREQ_QUERY_QP_RESP_SB_PATH_MTU_SFT; + qp->timeout = sb->timeout; + qp->retry_cnt = sb->retry_cnt; + qp->rnr_retry = sb->rnr_retry; + qp->min_rnr_timer = sb->min_rnr_timer; + qp->rq.psn = le32_to_cpu(sb->rq_psn); + qp->max_rd_atomic = ORRQ_SLOTS_TO_ORD_LIMIT(sb->max_rd_atomic); + qp->sq.psn = le32_to_cpu(sb->sq_psn); + qp->max_dest_rd_atomic = + IRRQ_SLOTS_TO_IRD_LIMIT(sb->max_dest_rd_atomic); + qp->sq.max_wqe = qp->sq.hwq.max_elements; + qp->rq.max_wqe = qp->rq.hwq.max_elements; + qp->sq.max_sge = le16_to_cpu(sb->sq_sge); + qp->rq.max_sge = le16_to_cpu(sb->rq_sge); + qp->max_inline_data = le32_to_cpu(sb->max_inline_data); + qp->dest_qpn = le32_to_cpu(sb->dest_qp_id); + memcpy(qp->smac, sb->src_mac, 6); + qp->vlan_id = le16_to_cpu(sb->vlan_pcp_vlan_dei_vlan_id); +bail: + bnxt_qplib_rcfw_free_sbuf(rcfw, sbuf); + return rc; +} + +static void __clean_cq(struct bnxt_qplib_cq *cq, u64 qp) +{ + struct bnxt_qplib_hwq *cq_hwq = &cq->hwq; + struct cq_base *hw_cqe, **hw_cqe_ptr; + int i; + + for (i = 0; i < cq_hwq->max_elements; i++) { + hw_cqe_ptr = (struct cq_base **)cq_hwq->pbl_ptr; + hw_cqe = &hw_cqe_ptr[CQE_PG(i)][CQE_IDX(i)]; + if (!CQE_CMP_VALID(hw_cqe, i, cq_hwq->max_elements)) + continue; + switch (hw_cqe->cqe_type_toggle & CQ_BASE_CQE_TYPE_MASK) { + case CQ_BASE_CQE_TYPE_REQ: + case CQ_BASE_CQE_TYPE_TERMINAL: + { + struct cq_req *cqe = (struct cq_req *)hw_cqe; + + if (qp == le64_to_cpu(cqe->qp_handle)) + cqe->qp_handle = 0; + break; + } + case CQ_BASE_CQE_TYPE_RES_RC: + case CQ_BASE_CQE_TYPE_RES_UD: + case CQ_BASE_CQE_TYPE_RES_RAWETH_QP1: + { + struct cq_res_rc *cqe = (struct cq_res_rc *)hw_cqe; + + if (qp == le64_to_cpu(cqe->qp_handle)) + cqe->qp_handle = 0; + break; + } + default: + break; + } + } +} + +int bnxt_qplib_destroy_qp(struct bnxt_qplib_res *res, + struct bnxt_qplib_qp *qp) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_destroy_qp req; + struct creq_destroy_qp_resp resp; + unsigned long flags; + u16 cmd_flags = 0; + int rc; + + RCFW_CMD_PREP(req, DESTROY_QP, cmd_flags); + + req.qp_cid = cpu_to_le32(qp->id); + rc = bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + (void *)&resp, NULL, 0); + if (rc) + return rc; + + /* Must walk the associated CQs to nullified the QP ptr */ + spin_lock_irqsave(&qp->scq->hwq.lock, flags); + + __clean_cq(qp->scq, (u64)(unsigned long)qp); + + if (qp->rcq && qp->rcq != qp->scq) { + spin_lock(&qp->rcq->hwq.lock); + __clean_cq(qp->rcq, (u64)(unsigned long)qp); + spin_unlock(&qp->rcq->hwq.lock); + } + + spin_unlock_irqrestore(&qp->scq->hwq.lock, flags); + + bnxt_qplib_free_qp_hdr_buf(res, qp); + bnxt_qplib_free_hwq(res->pdev, &qp->sq.hwq); + kfree(qp->sq.swq); + + bnxt_qplib_free_hwq(res->pdev, &qp->rq.hwq); + kfree(qp->rq.swq); + + if (qp->irrq.max_elements) + bnxt_qplib_free_hwq(res->pdev, &qp->irrq); + if (qp->orrq.max_elements) + bnxt_qplib_free_hwq(res->pdev, &qp->orrq); + + return 0; +} + +void *bnxt_qplib_get_qp1_sq_buf(struct bnxt_qplib_qp *qp, + struct bnxt_qplib_sge *sge) +{ + struct bnxt_qplib_q *sq = &qp->sq; + u32 sw_prod; + + memset(sge, 0, sizeof(*sge)); + + if (qp->sq_hdr_buf) { + sw_prod = HWQ_CMP(sq->hwq.prod, &sq->hwq); + sge->addr = (dma_addr_t)(qp->sq_hdr_buf_map + + sw_prod * qp->sq_hdr_buf_size); + sge->lkey = 0xFFFFFFFF; + sge->size = qp->sq_hdr_buf_size; + return qp->sq_hdr_buf + sw_prod * sge->size; + } + return NULL; +} + +u32 bnxt_qplib_get_rq_prod_index(struct bnxt_qplib_qp *qp) +{ + struct bnxt_qplib_q *rq = &qp->rq; + + return HWQ_CMP(rq->hwq.prod, &rq->hwq); +} + +dma_addr_t bnxt_qplib_get_qp_buf_from_index(struct bnxt_qplib_qp *qp, u32 index) +{ + return (qp->rq_hdr_buf_map + index * qp->rq_hdr_buf_size); +} + +void *bnxt_qplib_get_qp1_rq_buf(struct bnxt_qplib_qp *qp, + struct bnxt_qplib_sge *sge) +{ + struct bnxt_qplib_q *rq = &qp->rq; + u32 sw_prod; + + memset(sge, 0, sizeof(*sge)); + + if (qp->rq_hdr_buf) { + sw_prod = HWQ_CMP(rq->hwq.prod, &rq->hwq); + sge->addr = (dma_addr_t)(qp->rq_hdr_buf_map + + sw_prod * qp->rq_hdr_buf_size); + sge->lkey = 0xFFFFFFFF; + sge->size = qp->rq_hdr_buf_size; + return qp->rq_hdr_buf + sw_prod * sge->size; + } + return NULL; +} + +void bnxt_qplib_post_send_db(struct bnxt_qplib_qp *qp) +{ + struct bnxt_qplib_q *sq = &qp->sq; + struct dbr_dbr db_msg = { 0 }; + u32 sw_prod; + + sw_prod = HWQ_CMP(sq->hwq.prod, &sq->hwq); + + db_msg.index = cpu_to_le32((sw_prod << DBR_DBR_INDEX_SFT) & + DBR_DBR_INDEX_MASK); + db_msg.type_xid = + cpu_to_le32(((qp->id << DBR_DBR_XID_SFT) & DBR_DBR_XID_MASK) | + DBR_DBR_TYPE_SQ); + /* Flush all the WQE writes to HW */ + wmb(); + __iowrite64_copy(qp->dpi->dbr, &db_msg, sizeof(db_msg) / sizeof(u64)); +} + +int bnxt_qplib_post_send(struct bnxt_qplib_qp *qp, + struct bnxt_qplib_swqe *wqe) +{ + struct bnxt_qplib_q *sq = &qp->sq; + struct bnxt_qplib_swq *swq; + struct sq_send *hw_sq_send_hdr, **hw_sq_send_ptr; + struct sq_sge *hw_sge; + u32 sw_prod; + u8 wqe_size16; + int i, rc = 0, data_len = 0, pkt_num = 0; + __le32 temp32; + + if (qp->state != CMDQ_MODIFY_QP_NEW_STATE_RTS) { + rc = -EINVAL; + goto done; + } + + if (bnxt_qplib_queue_full(sq)) { + dev_err(&sq->hwq.pdev->dev, + "QPLIB: prod = %#x cons = %#x qdepth = %#x delta = %#x", + sq->hwq.prod, sq->hwq.cons, sq->hwq.max_elements, + sq->q_full_delta); + rc = -ENOMEM; + goto done; + } + sw_prod = HWQ_CMP(sq->hwq.prod, &sq->hwq); + swq = &sq->swq[sw_prod]; + swq->wr_id = wqe->wr_id; + swq->type = wqe->type; + swq->flags = wqe->flags; + if (qp->sig_type) + swq->flags |= SQ_SEND_FLAGS_SIGNAL_COMP; + swq->start_psn = sq->psn & BTH_PSN_MASK; + + hw_sq_send_ptr = (struct sq_send **)sq->hwq.pbl_ptr; + hw_sq_send_hdr = &hw_sq_send_ptr[get_sqe_pg(sw_prod)] + [get_sqe_idx(sw_prod)]; + + memset(hw_sq_send_hdr, 0, BNXT_QPLIB_MAX_SQE_ENTRY_SIZE); + + if (wqe->flags & BNXT_QPLIB_SWQE_FLAGS_INLINE) { + /* Copy the inline data */ + if (wqe->inline_len > BNXT_QPLIB_SWQE_MAX_INLINE_LENGTH) { + dev_warn(&sq->hwq.pdev->dev, + "QPLIB: Inline data length > 96 detected"); + data_len = BNXT_QPLIB_SWQE_MAX_INLINE_LENGTH; + } else { + data_len = wqe->inline_len; + } + memcpy(hw_sq_send_hdr->data, wqe->inline_data, data_len); + wqe_size16 = (data_len + 15) >> 4; + } else { + for (i = 0, hw_sge = (struct sq_sge *)hw_sq_send_hdr->data; + i < wqe->num_sge; i++, hw_sge++) { + hw_sge->va_or_pa = cpu_to_le64(wqe->sg_list[i].addr); + hw_sge->l_key = cpu_to_le32(wqe->sg_list[i].lkey); + hw_sge->size = cpu_to_le32(wqe->sg_list[i].size); + data_len += wqe->sg_list[i].size; + } + /* Each SGE entry = 1 WQE size16 */ + wqe_size16 = wqe->num_sge; + /* HW requires wqe size has room for atleast one SGE even if + * none was supplied by ULP + */ + if (!wqe->num_sge) + wqe_size16++; + } + + /* Specifics */ + switch (wqe->type) { + case BNXT_QPLIB_SWQE_TYPE_SEND: + if (qp->type == CMDQ_CREATE_QP1_TYPE_GSI) { + /* Assemble info for Raw Ethertype QPs */ + struct sq_send_raweth_qp1 *sqe = + (struct sq_send_raweth_qp1 *)hw_sq_send_hdr; + + sqe->wqe_type = wqe->type; + sqe->flags = wqe->flags; + sqe->wqe_size = wqe_size16 + + ((offsetof(typeof(*sqe), data) + 15) >> 4); + sqe->cfa_action = cpu_to_le16(wqe->rawqp1.cfa_action); + sqe->lflags = cpu_to_le16(wqe->rawqp1.lflags); + sqe->length = cpu_to_le32(data_len); + sqe->cfa_meta = cpu_to_le32((wqe->rawqp1.cfa_meta & + SQ_SEND_RAWETH_QP1_CFA_META_VLAN_VID_MASK) << + SQ_SEND_RAWETH_QP1_CFA_META_VLAN_VID_SFT); + + break; + } + /* else, just fall thru */ + case BNXT_QPLIB_SWQE_TYPE_SEND_WITH_IMM: + case BNXT_QPLIB_SWQE_TYPE_SEND_WITH_INV: + { + struct sq_send *sqe = (struct sq_send *)hw_sq_send_hdr; + + sqe->wqe_type = wqe->type; + sqe->flags = wqe->flags; + sqe->wqe_size = wqe_size16 + + ((offsetof(typeof(*sqe), data) + 15) >> 4); + sqe->inv_key_or_imm_data = cpu_to_le32( + wqe->send.inv_key); + if (qp->type == CMDQ_CREATE_QP_TYPE_UD) { + sqe->q_key = cpu_to_le32(wqe->send.q_key); + sqe->dst_qp = cpu_to_le32( + wqe->send.dst_qp & SQ_SEND_DST_QP_MASK); + sqe->length = cpu_to_le32(data_len); + sqe->avid = cpu_to_le32(wqe->send.avid & + SQ_SEND_AVID_MASK); + sq->psn = (sq->psn + 1) & BTH_PSN_MASK; + } else { + sqe->length = cpu_to_le32(data_len); + sqe->dst_qp = 0; + sqe->avid = 0; + if (qp->mtu) + pkt_num = (data_len + qp->mtu - 1) / qp->mtu; + if (!pkt_num) + pkt_num = 1; + sq->psn = (sq->psn + pkt_num) & BTH_PSN_MASK; + } + break; + } + case BNXT_QPLIB_SWQE_TYPE_RDMA_WRITE: + case BNXT_QPLIB_SWQE_TYPE_RDMA_WRITE_WITH_IMM: + case BNXT_QPLIB_SWQE_TYPE_RDMA_READ: + { + struct sq_rdma *sqe = (struct sq_rdma *)hw_sq_send_hdr; + + sqe->wqe_type = wqe->type; + sqe->flags = wqe->flags; + sqe->wqe_size = wqe_size16 + + ((offsetof(typeof(*sqe), data) + 15) >> 4); + sqe->imm_data = cpu_to_le32(wqe->rdma.inv_key); + sqe->length = cpu_to_le32((u32)data_len); + sqe->remote_va = cpu_to_le64(wqe->rdma.remote_va); + sqe->remote_key = cpu_to_le32(wqe->rdma.r_key); + if (qp->mtu) + pkt_num = (data_len + qp->mtu - 1) / qp->mtu; + if (!pkt_num) + pkt_num = 1; + sq->psn = (sq->psn + pkt_num) & BTH_PSN_MASK; + break; + } + case BNXT_QPLIB_SWQE_TYPE_ATOMIC_CMP_AND_SWP: + case BNXT_QPLIB_SWQE_TYPE_ATOMIC_FETCH_AND_ADD: + { + struct sq_atomic *sqe = (struct sq_atomic *)hw_sq_send_hdr; + + sqe->wqe_type = wqe->type; + sqe->flags = wqe->flags; + sqe->remote_key = cpu_to_le32(wqe->atomic.r_key); + sqe->remote_va = cpu_to_le64(wqe->atomic.remote_va); + sqe->swap_data = cpu_to_le64(wqe->atomic.swap_data); + sqe->cmp_data = cpu_to_le64(wqe->atomic.cmp_data); + if (qp->mtu) + pkt_num = (data_len + qp->mtu - 1) / qp->mtu; + if (!pkt_num) + pkt_num = 1; + sq->psn = (sq->psn + pkt_num) & BTH_PSN_MASK; + break; + } + case BNXT_QPLIB_SWQE_TYPE_LOCAL_INV: + { + struct sq_localinvalidate *sqe = + (struct sq_localinvalidate *)hw_sq_send_hdr; + + sqe->wqe_type = wqe->type; + sqe->flags = wqe->flags; + sqe->inv_l_key = cpu_to_le32(wqe->local_inv.inv_l_key); + + break; + } + case BNXT_QPLIB_SWQE_TYPE_FAST_REG_MR: + { + struct sq_fr_pmr *sqe = (struct sq_fr_pmr *)hw_sq_send_hdr; + + sqe->wqe_type = wqe->type; + sqe->flags = wqe->flags; + sqe->access_cntl = wqe->frmr.access_cntl | + SQ_FR_PMR_ACCESS_CNTL_LOCAL_WRITE; + sqe->zero_based_page_size_log = + (wqe->frmr.pg_sz_log & SQ_FR_PMR_PAGE_SIZE_LOG_MASK) << + SQ_FR_PMR_PAGE_SIZE_LOG_SFT | + (wqe->frmr.zero_based ? SQ_FR_PMR_ZERO_BASED : 0); + sqe->l_key = cpu_to_le32(wqe->frmr.l_key); + temp32 = cpu_to_le32(wqe->frmr.length); + memcpy(sqe->length, &temp32, sizeof(wqe->frmr.length)); + sqe->numlevels_pbl_page_size_log = + ((wqe->frmr.pbl_pg_sz_log << + SQ_FR_PMR_PBL_PAGE_SIZE_LOG_SFT) & + SQ_FR_PMR_PBL_PAGE_SIZE_LOG_MASK) | + ((wqe->frmr.levels << SQ_FR_PMR_NUMLEVELS_SFT) & + SQ_FR_PMR_NUMLEVELS_MASK); + + for (i = 0; i < wqe->frmr.page_list_len; i++) + wqe->frmr.pbl_ptr[i] = cpu_to_le64( + wqe->frmr.page_list[i] | + PTU_PTE_VALID); + sqe->pblptr = cpu_to_le64(wqe->frmr.pbl_dma_ptr); + sqe->va = cpu_to_le64(wqe->frmr.va); + + break; + } + case BNXT_QPLIB_SWQE_TYPE_BIND_MW: + { + struct sq_bind *sqe = (struct sq_bind *)hw_sq_send_hdr; + + sqe->wqe_type = wqe->type; + sqe->flags = wqe->flags; + sqe->access_cntl = wqe->bind.access_cntl; + sqe->mw_type_zero_based = wqe->bind.mw_type | + (wqe->bind.zero_based ? SQ_BIND_ZERO_BASED : 0); + sqe->parent_l_key = cpu_to_le32(wqe->bind.parent_l_key); + sqe->l_key = cpu_to_le32(wqe->bind.r_key); + sqe->va = cpu_to_le64(wqe->bind.va); + temp32 = cpu_to_le32(wqe->bind.length); + memcpy(&sqe->length, &temp32, sizeof(wqe->bind.length)); + break; + } + default: + /* Bad wqe, return error */ + rc = -EINVAL; + goto done; + } + swq->next_psn = sq->psn & BTH_PSN_MASK; + if (swq->psn_search) { + swq->psn_search->opcode_start_psn = cpu_to_le32( + ((swq->start_psn << SQ_PSN_SEARCH_START_PSN_SFT) & + SQ_PSN_SEARCH_START_PSN_MASK) | + ((wqe->type << SQ_PSN_SEARCH_OPCODE_SFT) & + SQ_PSN_SEARCH_OPCODE_MASK)); + swq->psn_search->flags_next_psn = cpu_to_le32( + ((swq->next_psn << SQ_PSN_SEARCH_NEXT_PSN_SFT) & + SQ_PSN_SEARCH_NEXT_PSN_MASK)); + } + + sq->hwq.prod++; + + qp->wqe_cnt++; + +done: + return rc; +} + +void bnxt_qplib_post_recv_db(struct bnxt_qplib_qp *qp) +{ + struct bnxt_qplib_q *rq = &qp->rq; + struct dbr_dbr db_msg = { 0 }; + u32 sw_prod; + + sw_prod = HWQ_CMP(rq->hwq.prod, &rq->hwq); + db_msg.index = cpu_to_le32((sw_prod << DBR_DBR_INDEX_SFT) & + DBR_DBR_INDEX_MASK); + db_msg.type_xid = + cpu_to_le32(((qp->id << DBR_DBR_XID_SFT) & DBR_DBR_XID_MASK) | + DBR_DBR_TYPE_RQ); + + /* Flush the writes to HW Rx WQE before the ringing Rx DB */ + wmb(); + __iowrite64_copy(qp->dpi->dbr, &db_msg, sizeof(db_msg) / sizeof(u64)); +} + +int bnxt_qplib_post_recv(struct bnxt_qplib_qp *qp, + struct bnxt_qplib_swqe *wqe) +{ + struct bnxt_qplib_q *rq = &qp->rq; + struct rq_wqe *rqe, **rqe_ptr; + struct sq_sge *hw_sge; + u32 sw_prod; + int i, rc = 0; + + if (qp->state == CMDQ_MODIFY_QP_NEW_STATE_ERR) { + dev_err(&rq->hwq.pdev->dev, + "QPLIB: FP: QP (0x%x) is in the 0x%x state", + qp->id, qp->state); + rc = -EINVAL; + goto done; + } + if (bnxt_qplib_queue_full(rq)) { + dev_err(&rq->hwq.pdev->dev, + "QPLIB: FP: QP (0x%x) RQ is full!", qp->id); + rc = -EINVAL; + goto done; + } + sw_prod = HWQ_CMP(rq->hwq.prod, &rq->hwq); + rq->swq[sw_prod].wr_id = wqe->wr_id; + + rqe_ptr = (struct rq_wqe **)rq->hwq.pbl_ptr; + rqe = &rqe_ptr[RQE_PG(sw_prod)][RQE_IDX(sw_prod)]; + + memset(rqe, 0, BNXT_QPLIB_MAX_RQE_ENTRY_SIZE); + + /* Calculate wqe_size16 and data_len */ + for (i = 0, hw_sge = (struct sq_sge *)rqe->data; + i < wqe->num_sge; i++, hw_sge++) { + hw_sge->va_or_pa = cpu_to_le64(wqe->sg_list[i].addr); + hw_sge->l_key = cpu_to_le32(wqe->sg_list[i].lkey); + hw_sge->size = cpu_to_le32(wqe->sg_list[i].size); + } + rqe->wqe_type = wqe->type; + rqe->flags = wqe->flags; + rqe->wqe_size = wqe->num_sge + + ((offsetof(typeof(*rqe), data) + 15) >> 4); + /* HW requires wqe size has room for atleast one SGE even if none + * was supplied by ULP + */ + if (!wqe->num_sge) + rqe->wqe_size++; + + /* Supply the rqe->wr_id index to the wr_id_tbl for now */ + rqe->wr_id[0] = cpu_to_le32(sw_prod); + + rq->hwq.prod++; +done: + return rc; +} + +/* CQ */ + +/* Spinlock must be held */ +static void bnxt_qplib_arm_cq_enable(struct bnxt_qplib_cq *cq) +{ + struct dbr_dbr db_msg = { 0 }; + + db_msg.type_xid = + cpu_to_le32(((cq->id << DBR_DBR_XID_SFT) & DBR_DBR_XID_MASK) | + DBR_DBR_TYPE_CQ_ARMENA); + /* Flush memory writes before enabling the CQ */ + wmb(); + __iowrite64_copy(cq->dbr_base, &db_msg, sizeof(db_msg) / sizeof(u64)); +} + +static void bnxt_qplib_arm_cq(struct bnxt_qplib_cq *cq, u32 arm_type) +{ + struct bnxt_qplib_hwq *cq_hwq = &cq->hwq; + struct dbr_dbr db_msg = { 0 }; + u32 sw_cons; + + /* Ring DB */ + sw_cons = HWQ_CMP(cq_hwq->cons, cq_hwq); + db_msg.index = cpu_to_le32((sw_cons << DBR_DBR_INDEX_SFT) & + DBR_DBR_INDEX_MASK); + db_msg.type_xid = + cpu_to_le32(((cq->id << DBR_DBR_XID_SFT) & DBR_DBR_XID_MASK) | + arm_type); + /* flush memory writes before arming the CQ */ + wmb(); + __iowrite64_copy(cq->dpi->dbr, &db_msg, sizeof(db_msg) / sizeof(u64)); +} + +int bnxt_qplib_create_cq(struct bnxt_qplib_res *res, struct bnxt_qplib_cq *cq) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_create_cq req; + struct creq_create_cq_resp resp; + struct bnxt_qplib_pbl *pbl; + u16 cmd_flags = 0; + int rc; + + cq->hwq.max_elements = cq->max_wqe; + rc = bnxt_qplib_alloc_init_hwq(res->pdev, &cq->hwq, cq->sghead, + cq->nmap, &cq->hwq.max_elements, + BNXT_QPLIB_MAX_CQE_ENTRY_SIZE, 0, + PAGE_SIZE, HWQ_TYPE_QUEUE); + if (rc) + goto exit; + + RCFW_CMD_PREP(req, CREATE_CQ, cmd_flags); + + if (!cq->dpi) { + dev_err(&rcfw->pdev->dev, + "QPLIB: FP: CREATE_CQ failed due to NULL DPI"); + return -EINVAL; + } + req.dpi = cpu_to_le32(cq->dpi->dpi); + req.cq_handle = cpu_to_le64(cq->cq_handle); + + req.cq_size = cpu_to_le32(cq->hwq.max_elements); + pbl = &cq->hwq.pbl[PBL_LVL_0]; + req.pg_size_lvl = cpu_to_le32( + ((cq->hwq.level & CMDQ_CREATE_CQ_LVL_MASK) << + CMDQ_CREATE_CQ_LVL_SFT) | + (pbl->pg_size == ROCE_PG_SIZE_4K ? CMDQ_CREATE_CQ_PG_SIZE_PG_4K : + pbl->pg_size == ROCE_PG_SIZE_8K ? CMDQ_CREATE_CQ_PG_SIZE_PG_8K : + pbl->pg_size == ROCE_PG_SIZE_64K ? CMDQ_CREATE_CQ_PG_SIZE_PG_64K : + pbl->pg_size == ROCE_PG_SIZE_2M ? CMDQ_CREATE_CQ_PG_SIZE_PG_2M : + pbl->pg_size == ROCE_PG_SIZE_8M ? CMDQ_CREATE_CQ_PG_SIZE_PG_8M : + pbl->pg_size == ROCE_PG_SIZE_1G ? CMDQ_CREATE_CQ_PG_SIZE_PG_1G : + CMDQ_CREATE_CQ_PG_SIZE_PG_4K)); + + req.pbl = cpu_to_le64(pbl->pg_map_arr[0]); + + req.cq_fco_cnq_id = cpu_to_le32( + (cq->cnq_hw_ring_id & CMDQ_CREATE_CQ_CNQ_ID_MASK) << + CMDQ_CREATE_CQ_CNQ_ID_SFT); + + rc = bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + (void *)&resp, NULL, 0); + if (rc) + goto fail; + + cq->id = le32_to_cpu(resp.xid); + cq->dbr_base = res->dpi_tbl.dbr_bar_reg_iomem; + cq->period = BNXT_QPLIB_QUEUE_START_PERIOD; + init_waitqueue_head(&cq->waitq); + + bnxt_qplib_arm_cq_enable(cq); + return 0; + +fail: + bnxt_qplib_free_hwq(res->pdev, &cq->hwq); +exit: + return rc; +} + +int bnxt_qplib_destroy_cq(struct bnxt_qplib_res *res, struct bnxt_qplib_cq *cq) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_destroy_cq req; + struct creq_destroy_cq_resp resp; + u16 cmd_flags = 0; + int rc; + + RCFW_CMD_PREP(req, DESTROY_CQ, cmd_flags); + + req.cq_cid = cpu_to_le32(cq->id); + rc = bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + (void *)&resp, NULL, 0); + if (rc) + return rc; + bnxt_qplib_free_hwq(res->pdev, &cq->hwq); + return 0; +} + +static int __flush_sq(struct bnxt_qplib_q *sq, struct bnxt_qplib_qp *qp, + struct bnxt_qplib_cqe **pcqe, int *budget) +{ + u32 sw_prod, sw_cons; + struct bnxt_qplib_cqe *cqe; + int rc = 0; + + /* Now complete all outstanding SQEs with FLUSHED_ERR */ + sw_prod = HWQ_CMP(sq->hwq.prod, &sq->hwq); + cqe = *pcqe; + while (*budget) { + sw_cons = HWQ_CMP(sq->hwq.cons, &sq->hwq); + if (sw_cons == sw_prod) { + sq->flush_in_progress = false; + break; + } + memset(cqe, 0, sizeof(*cqe)); + cqe->status = CQ_REQ_STATUS_WORK_REQUEST_FLUSHED_ERR; + cqe->opcode = CQ_BASE_CQE_TYPE_REQ; + cqe->qp_handle = (u64)(unsigned long)qp; + cqe->wr_id = sq->swq[sw_cons].wr_id; + cqe->src_qp = qp->id; + cqe->type = sq->swq[sw_cons].type; + cqe++; + (*budget)--; + sq->hwq.cons++; + } + *pcqe = cqe; + if (!(*budget) && HWQ_CMP(sq->hwq.cons, &sq->hwq) != sw_prod) + /* Out of budget */ + rc = -EAGAIN; + + return rc; +} + +static int __flush_rq(struct bnxt_qplib_q *rq, struct bnxt_qplib_qp *qp, + int opcode, struct bnxt_qplib_cqe **pcqe, int *budget) +{ + struct bnxt_qplib_cqe *cqe; + u32 sw_prod, sw_cons; + int rc = 0; + + /* Flush the rest of the RQ */ + sw_prod = HWQ_CMP(rq->hwq.prod, &rq->hwq); + cqe = *pcqe; + while (*budget) { + sw_cons = HWQ_CMP(rq->hwq.cons, &rq->hwq); + if (sw_cons == sw_prod) + break; + memset(cqe, 0, sizeof(*cqe)); + cqe->status = + CQ_RES_RC_STATUS_WORK_REQUEST_FLUSHED_ERR; + cqe->opcode = opcode; + cqe->qp_handle = (unsigned long)qp; + cqe->wr_id = rq->swq[sw_cons].wr_id; + cqe++; + (*budget)--; + rq->hwq.cons++; + } + *pcqe = cqe; + if (!*budget && HWQ_CMP(rq->hwq.cons, &rq->hwq) != sw_prod) + /* Out of budget */ + rc = -EAGAIN; + + return rc; +} + +/* Note: SQE is valid from sw_sq_cons up to cqe_sq_cons (exclusive) + * CQE is track from sw_cq_cons to max_element but valid only if VALID=1 + */ +static int do_wa9060(struct bnxt_qplib_qp *qp, struct bnxt_qplib_cq *cq, + u32 cq_cons, u32 sw_sq_cons, u32 cqe_sq_cons) +{ + struct bnxt_qplib_q *sq = &qp->sq; + struct bnxt_qplib_swq *swq; + u32 peek_sw_cq_cons, peek_raw_cq_cons, peek_sq_cons_idx; + struct cq_base *peek_hwcqe, **peek_hw_cqe_ptr; + struct cq_req *peek_req_hwcqe; + struct bnxt_qplib_qp *peek_qp; + struct bnxt_qplib_q *peek_sq; + int i, rc = 0; + + /* Normal mode */ + /* Check for the psn_search marking before completing */ + swq = &sq->swq[sw_sq_cons]; + if (swq->psn_search && + le32_to_cpu(swq->psn_search->flags_next_psn) & 0x80000000) { + /* Unmark */ + swq->psn_search->flags_next_psn = cpu_to_le32 + (le32_to_cpu(swq->psn_search->flags_next_psn) + & ~0x80000000); + dev_dbg(&cq->hwq.pdev->dev, + "FP: Process Req cq_cons=0x%x qp=0x%x sq cons sw=0x%x cqe=0x%x marked!\n", + cq_cons, qp->id, sw_sq_cons, cqe_sq_cons); + sq->condition = true; + sq->send_phantom = true; + + /* TODO: Only ARM if the previous SQE is ARMALL */ + bnxt_qplib_arm_cq(cq, DBR_DBR_TYPE_CQ_ARMALL); + + rc = -EAGAIN; + goto out; + } + if (sq->condition) { + /* Peek at the completions */ + peek_raw_cq_cons = cq->hwq.cons; + peek_sw_cq_cons = cq_cons; + i = cq->hwq.max_elements; + while (i--) { + peek_sw_cq_cons = HWQ_CMP((peek_sw_cq_cons), &cq->hwq); + peek_hw_cqe_ptr = (struct cq_base **)cq->hwq.pbl_ptr; + peek_hwcqe = &peek_hw_cqe_ptr[CQE_PG(peek_sw_cq_cons)] + [CQE_IDX(peek_sw_cq_cons)]; + /* If the next hwcqe is VALID */ + if (CQE_CMP_VALID(peek_hwcqe, peek_raw_cq_cons, + cq->hwq.max_elements)) { + /* If the next hwcqe is a REQ */ + if ((peek_hwcqe->cqe_type_toggle & + CQ_BASE_CQE_TYPE_MASK) == + CQ_BASE_CQE_TYPE_REQ) { + peek_req_hwcqe = (struct cq_req *) + peek_hwcqe; + peek_qp = (struct bnxt_qplib_qp *) + ((unsigned long) + le64_to_cpu + (peek_req_hwcqe->qp_handle)); + peek_sq = &peek_qp->sq; + peek_sq_cons_idx = HWQ_CMP(le16_to_cpu( + peek_req_hwcqe->sq_cons_idx) - 1 + , &sq->hwq); + /* If the hwcqe's sq's wr_id matches */ + if (peek_sq == sq && + sq->swq[peek_sq_cons_idx].wr_id == + BNXT_QPLIB_FENCE_WRID) { + /* + * Unbreak only if the phantom + * comes back + */ + dev_dbg(&cq->hwq.pdev->dev, + "FP:Got Phantom CQE"); + sq->condition = false; + sq->single = true; + rc = 0; + goto out; + } + } + /* Valid but not the phantom, so keep looping */ + } else { + /* Not valid yet, just exit and wait */ + rc = -EINVAL; + goto out; + } + peek_sw_cq_cons++; + peek_raw_cq_cons++; + } + dev_err(&cq->hwq.pdev->dev, + "Should not have come here! cq_cons=0x%x qp=0x%x sq cons sw=0x%x hw=0x%x", + cq_cons, qp->id, sw_sq_cons, cqe_sq_cons); + rc = -EINVAL; + } +out: + return rc; +} + +static int bnxt_qplib_cq_process_req(struct bnxt_qplib_cq *cq, + struct cq_req *hwcqe, + struct bnxt_qplib_cqe **pcqe, int *budget, + u32 cq_cons, struct bnxt_qplib_qp **lib_qp) +{ + struct bnxt_qplib_qp *qp; + struct bnxt_qplib_q *sq; + struct bnxt_qplib_cqe *cqe; + u32 sw_sq_cons, cqe_sq_cons; + struct bnxt_qplib_swq *swq; + int rc = 0; + + qp = (struct bnxt_qplib_qp *)((unsigned long) + le64_to_cpu(hwcqe->qp_handle)); + if (!qp) { + dev_err(&cq->hwq.pdev->dev, + "QPLIB: FP: Process Req qp is NULL"); + return -EINVAL; + } + sq = &qp->sq; + + cqe_sq_cons = HWQ_CMP(le16_to_cpu(hwcqe->sq_cons_idx), &sq->hwq); + if (cqe_sq_cons > sq->hwq.max_elements) { + dev_err(&cq->hwq.pdev->dev, + "QPLIB: FP: CQ Process req reported "); + dev_err(&cq->hwq.pdev->dev, + "QPLIB: sq_cons_idx 0x%x which exceeded max 0x%x", + cqe_sq_cons, sq->hwq.max_elements); + return -EINVAL; + } + /* If we were in the middle of flushing the SQ, continue */ + if (sq->flush_in_progress) + goto flush; + + /* Require to walk the sq's swq to fabricate CQEs for all previously + * signaled SWQEs due to CQE aggregation from the current sq cons + * to the cqe_sq_cons + */ + cqe = *pcqe; + while (*budget) { + sw_sq_cons = HWQ_CMP(sq->hwq.cons, &sq->hwq); + if (sw_sq_cons == cqe_sq_cons) + /* Done */ + break; + + swq = &sq->swq[sw_sq_cons]; + memset(cqe, 0, sizeof(*cqe)); + cqe->opcode = CQ_BASE_CQE_TYPE_REQ; + cqe->qp_handle = (u64)(unsigned long)qp; + cqe->src_qp = qp->id; + cqe->wr_id = swq->wr_id; + if (cqe->wr_id == BNXT_QPLIB_FENCE_WRID) + goto skip; + cqe->type = swq->type; + + /* For the last CQE, check for status. For errors, regardless + * of the request being signaled or not, it must complete with + * the hwcqe error status + */ + if (HWQ_CMP((sw_sq_cons + 1), &sq->hwq) == cqe_sq_cons && + hwcqe->status != CQ_REQ_STATUS_OK) { + cqe->status = hwcqe->status; + dev_err(&cq->hwq.pdev->dev, + "QPLIB: FP: CQ Processed Req "); + dev_err(&cq->hwq.pdev->dev, + "QPLIB: wr_id[%d] = 0x%llx with status 0x%x", + sw_sq_cons, cqe->wr_id, cqe->status); + cqe++; + (*budget)--; + sq->flush_in_progress = true; + /* Must block new posting of SQ and RQ */ + qp->state = CMDQ_MODIFY_QP_NEW_STATE_ERR; + sq->condition = false; + sq->single = false; + } else { + if (swq->flags & SQ_SEND_FLAGS_SIGNAL_COMP) { + /* Before we complete, do WA 9060 */ + if (do_wa9060(qp, cq, cq_cons, sw_sq_cons, + cqe_sq_cons)) { + *lib_qp = qp; + goto out; + } + cqe->status = CQ_REQ_STATUS_OK; + cqe++; + (*budget)--; + } + } +skip: + sq->hwq.cons++; + if (sq->single) + break; + } +out: + *pcqe = cqe; + if (HWQ_CMP(sq->hwq.cons, &sq->hwq) != cqe_sq_cons) { + /* Out of budget */ + rc = -EAGAIN; + goto done; + } + /* + * Back to normal completion mode only after it has completed all of + * the WC for this CQE + */ + sq->single = false; + if (!sq->flush_in_progress) + goto done; +flush: + /* Require to walk the sq's swq to fabricate CQEs for all + * previously posted SWQEs due to the error CQE received + */ + rc = __flush_sq(sq, qp, pcqe, budget); + if (!rc) + sq->flush_in_progress = false; +done: + return rc; +} + +static int bnxt_qplib_cq_process_res_rc(struct bnxt_qplib_cq *cq, + struct cq_res_rc *hwcqe, + struct bnxt_qplib_cqe **pcqe, + int *budget) +{ + struct bnxt_qplib_qp *qp; + struct bnxt_qplib_q *rq; + struct bnxt_qplib_cqe *cqe; + u32 wr_id_idx; + int rc = 0; + + qp = (struct bnxt_qplib_qp *)((unsigned long) + le64_to_cpu(hwcqe->qp_handle)); + if (!qp) { + dev_err(&cq->hwq.pdev->dev, "QPLIB: process_cq RC qp is NULL"); + return -EINVAL; + } + cqe = *pcqe; + cqe->opcode = hwcqe->cqe_type_toggle & CQ_BASE_CQE_TYPE_MASK; + cqe->length = le32_to_cpu(hwcqe->length); + cqe->invrkey = le32_to_cpu(hwcqe->imm_data_or_inv_r_key); + cqe->mr_handle = le64_to_cpu(hwcqe->mr_handle); + cqe->flags = le16_to_cpu(hwcqe->flags); + cqe->status = hwcqe->status; + cqe->qp_handle = (u64)(unsigned long)qp; + + wr_id_idx = le32_to_cpu(hwcqe->srq_or_rq_wr_id) & + CQ_RES_RC_SRQ_OR_RQ_WR_ID_MASK; + rq = &qp->rq; + if (wr_id_idx > rq->hwq.max_elements) { + dev_err(&cq->hwq.pdev->dev, "QPLIB: FP: CQ Process RC "); + dev_err(&cq->hwq.pdev->dev, + "QPLIB: wr_id idx 0x%x exceeded RQ max 0x%x", + wr_id_idx, rq->hwq.max_elements); + return -EINVAL; + } + if (rq->flush_in_progress) + goto flush_rq; + + cqe->wr_id = rq->swq[wr_id_idx].wr_id; + cqe++; + (*budget)--; + rq->hwq.cons++; + *pcqe = cqe; + + if (hwcqe->status != CQ_RES_RC_STATUS_OK) { + rq->flush_in_progress = true; +flush_rq: + rc = __flush_rq(rq, qp, CQ_BASE_CQE_TYPE_RES_RC, pcqe, budget); + if (!rc) + rq->flush_in_progress = false; + } + return rc; +} + +static int bnxt_qplib_cq_process_res_ud(struct bnxt_qplib_cq *cq, + struct cq_res_ud *hwcqe, + struct bnxt_qplib_cqe **pcqe, + int *budget) +{ + struct bnxt_qplib_qp *qp; + struct bnxt_qplib_q *rq; + struct bnxt_qplib_cqe *cqe; + u32 wr_id_idx; + int rc = 0; + + qp = (struct bnxt_qplib_qp *)((unsigned long) + le64_to_cpu(hwcqe->qp_handle)); + if (!qp) { + dev_err(&cq->hwq.pdev->dev, "QPLIB: process_cq UD qp is NULL"); + return -EINVAL; + } + cqe = *pcqe; + cqe->opcode = hwcqe->cqe_type_toggle & CQ_BASE_CQE_TYPE_MASK; + cqe->length = le32_to_cpu(hwcqe->length); + cqe->invrkey = le32_to_cpu(hwcqe->imm_data); + cqe->flags = le16_to_cpu(hwcqe->flags); + cqe->status = hwcqe->status; + cqe->qp_handle = (u64)(unsigned long)qp; + memcpy(cqe->smac, hwcqe->src_mac, 6); + wr_id_idx = le32_to_cpu(hwcqe->src_qp_high_srq_or_rq_wr_id) + & CQ_RES_UD_SRQ_OR_RQ_WR_ID_MASK; + cqe->src_qp = le16_to_cpu(hwcqe->src_qp_low) | + ((le32_to_cpu( + hwcqe->src_qp_high_srq_or_rq_wr_id) & + CQ_RES_UD_SRC_QP_HIGH_MASK) >> 8); + + rq = &qp->rq; + if (wr_id_idx > rq->hwq.max_elements) { + dev_err(&cq->hwq.pdev->dev, "QPLIB: FP: CQ Process UD "); + dev_err(&cq->hwq.pdev->dev, + "QPLIB: wr_id idx %#x exceeded RQ max %#x", + wr_id_idx, rq->hwq.max_elements); + return -EINVAL; + } + if (rq->flush_in_progress) + goto flush_rq; + + cqe->wr_id = rq->swq[wr_id_idx].wr_id; + cqe++; + (*budget)--; + rq->hwq.cons++; + *pcqe = cqe; + + if (hwcqe->status != CQ_RES_RC_STATUS_OK) { + rq->flush_in_progress = true; +flush_rq: + rc = __flush_rq(rq, qp, CQ_BASE_CQE_TYPE_RES_UD, pcqe, budget); + if (!rc) + rq->flush_in_progress = false; + } + return rc; +} + +bool bnxt_qplib_is_cq_empty(struct bnxt_qplib_cq *cq) +{ + struct cq_base *hw_cqe, **hw_cqe_ptr; + unsigned long flags; + u32 sw_cons, raw_cons; + bool rc = true; + + spin_lock_irqsave(&cq->hwq.lock, flags); + raw_cons = cq->hwq.cons; + sw_cons = HWQ_CMP(raw_cons, &cq->hwq); + hw_cqe_ptr = (struct cq_base **)cq->hwq.pbl_ptr; + hw_cqe = &hw_cqe_ptr[CQE_PG(sw_cons)][CQE_IDX(sw_cons)]; + + /* Check for Valid bit. If the CQE is valid, return false */ + rc = !CQE_CMP_VALID(hw_cqe, raw_cons, cq->hwq.max_elements); + spin_unlock_irqrestore(&cq->hwq.lock, flags); + return rc; +} + +static int bnxt_qplib_cq_process_res_raweth_qp1(struct bnxt_qplib_cq *cq, + struct cq_res_raweth_qp1 *hwcqe, + struct bnxt_qplib_cqe **pcqe, + int *budget) +{ + struct bnxt_qplib_qp *qp; + struct bnxt_qplib_q *rq; + struct bnxt_qplib_cqe *cqe; + u32 wr_id_idx; + int rc = 0; + + qp = (struct bnxt_qplib_qp *)((unsigned long) + le64_to_cpu(hwcqe->qp_handle)); + if (!qp) { + dev_err(&cq->hwq.pdev->dev, + "QPLIB: process_cq Raw/QP1 qp is NULL"); + return -EINVAL; + } + cqe = *pcqe; + cqe->opcode = hwcqe->cqe_type_toggle & CQ_BASE_CQE_TYPE_MASK; + cqe->flags = le16_to_cpu(hwcqe->flags); + cqe->qp_handle = (u64)(unsigned long)qp; + + wr_id_idx = + le32_to_cpu(hwcqe->raweth_qp1_payload_offset_srq_or_rq_wr_id) + & CQ_RES_RAWETH_QP1_SRQ_OR_RQ_WR_ID_MASK; + cqe->src_qp = qp->id; + if (qp->id == 1 && !cqe->length) { + /* Add workaround for the length misdetection */ + cqe->length = 296; + } else { + cqe->length = le16_to_cpu(hwcqe->length); + } + cqe->pkey_index = qp->pkey_index; + memcpy(cqe->smac, qp->smac, 6); + + cqe->raweth_qp1_flags = le16_to_cpu(hwcqe->raweth_qp1_flags); + cqe->raweth_qp1_flags2 = le32_to_cpu(hwcqe->raweth_qp1_flags2); + + rq = &qp->rq; + if (wr_id_idx > rq->hwq.max_elements) { + dev_err(&cq->hwq.pdev->dev, "QPLIB: FP: CQ Process Raw/QP1 RQ wr_id "); + dev_err(&cq->hwq.pdev->dev, "QPLIB: ix 0x%x exceeded RQ max 0x%x", + wr_id_idx, rq->hwq.max_elements); + return -EINVAL; + } + if (rq->flush_in_progress) + goto flush_rq; + + cqe->wr_id = rq->swq[wr_id_idx].wr_id; + cqe++; + (*budget)--; + rq->hwq.cons++; + *pcqe = cqe; + + if (hwcqe->status != CQ_RES_RC_STATUS_OK) { + rq->flush_in_progress = true; +flush_rq: + rc = __flush_rq(rq, qp, CQ_BASE_CQE_TYPE_RES_RAWETH_QP1, pcqe, + budget); + if (!rc) + rq->flush_in_progress = false; + } + return rc; +} + +static int bnxt_qplib_cq_process_terminal(struct bnxt_qplib_cq *cq, + struct cq_terminal *hwcqe, + struct bnxt_qplib_cqe **pcqe, + int *budget) +{ + struct bnxt_qplib_qp *qp; + struct bnxt_qplib_q *sq, *rq; + struct bnxt_qplib_cqe *cqe; + u32 sw_cons = 0, cqe_cons; + int rc = 0; + u8 opcode = 0; + + /* Check the Status */ + if (hwcqe->status != CQ_TERMINAL_STATUS_OK) + dev_warn(&cq->hwq.pdev->dev, + "QPLIB: FP: CQ Process Terminal Error status = 0x%x", + hwcqe->status); + + qp = (struct bnxt_qplib_qp *)((unsigned long) + le64_to_cpu(hwcqe->qp_handle)); + if (!qp) { + dev_err(&cq->hwq.pdev->dev, + "QPLIB: FP: CQ Process terminal qp is NULL"); + return -EINVAL; + } + /* Must block new posting of SQ and RQ */ + qp->state = CMDQ_MODIFY_QP_NEW_STATE_ERR; + + sq = &qp->sq; + rq = &qp->rq; + + cqe_cons = le16_to_cpu(hwcqe->sq_cons_idx); + if (cqe_cons == 0xFFFF) + goto do_rq; + + if (cqe_cons > sq->hwq.max_elements) { + dev_err(&cq->hwq.pdev->dev, + "QPLIB: FP: CQ Process terminal reported "); + dev_err(&cq->hwq.pdev->dev, + "QPLIB: sq_cons_idx 0x%x which exceeded max 0x%x", + cqe_cons, sq->hwq.max_elements); + goto do_rq; + } + /* If we were in the middle of flushing, continue */ + if (sq->flush_in_progress) + goto flush_sq; + + /* Terminal CQE can also include aggregated successful CQEs prior. + * So we must complete all CQEs from the current sq's cons to the + * cq_cons with status OK + */ + cqe = *pcqe; + while (*budget) { + sw_cons = HWQ_CMP(sq->hwq.cons, &sq->hwq); + if (sw_cons == cqe_cons) + break; + if (sq->swq[sw_cons].flags & SQ_SEND_FLAGS_SIGNAL_COMP) { + memset(cqe, 0, sizeof(*cqe)); + cqe->status = CQ_REQ_STATUS_OK; + cqe->opcode = CQ_BASE_CQE_TYPE_REQ; + cqe->qp_handle = (u64)(unsigned long)qp; + cqe->src_qp = qp->id; + cqe->wr_id = sq->swq[sw_cons].wr_id; + cqe->type = sq->swq[sw_cons].type; + cqe++; + (*budget)--; + } + sq->hwq.cons++; + } + *pcqe = cqe; + if (!(*budget) && sw_cons != cqe_cons) { + /* Out of budget */ + rc = -EAGAIN; + goto sq_done; + } + sq->flush_in_progress = true; +flush_sq: + rc = __flush_sq(sq, qp, pcqe, budget); + if (!rc) + sq->flush_in_progress = false; +sq_done: + if (rc) + return rc; +do_rq: + cqe_cons = le16_to_cpu(hwcqe->rq_cons_idx); + if (cqe_cons == 0xFFFF) { + goto done; + } else if (cqe_cons > rq->hwq.max_elements) { + dev_err(&cq->hwq.pdev->dev, + "QPLIB: FP: CQ Processed terminal "); + dev_err(&cq->hwq.pdev->dev, + "QPLIB: reported rq_cons_idx 0x%x exceeds max 0x%x", + cqe_cons, rq->hwq.max_elements); + goto done; + } + /* Terminal CQE requires all posted RQEs to complete with FLUSHED_ERR + * from the current rq->cons to the rq->prod regardless what the + * rq->cons the terminal CQE indicates + */ + rq->flush_in_progress = true; + switch (qp->type) { + case CMDQ_CREATE_QP1_TYPE_GSI: + opcode = CQ_BASE_CQE_TYPE_RES_RAWETH_QP1; + break; + case CMDQ_CREATE_QP_TYPE_RC: + opcode = CQ_BASE_CQE_TYPE_RES_RC; + break; + case CMDQ_CREATE_QP_TYPE_UD: + opcode = CQ_BASE_CQE_TYPE_RES_UD; + break; + } + + rc = __flush_rq(rq, qp, opcode, pcqe, budget); + if (!rc) + rq->flush_in_progress = false; +done: + return rc; +} + +static int bnxt_qplib_cq_process_cutoff(struct bnxt_qplib_cq *cq, + struct cq_cutoff *hwcqe) +{ + /* Check the Status */ + if (hwcqe->status != CQ_CUTOFF_STATUS_OK) { + dev_err(&cq->hwq.pdev->dev, + "QPLIB: FP: CQ Process Cutoff Error status = 0x%x", + hwcqe->status); + return -EINVAL; + } + clear_bit(CQ_FLAGS_RESIZE_IN_PROG, &cq->flags); + wake_up_interruptible(&cq->waitq); + + return 0; +} + +int bnxt_qplib_poll_cq(struct bnxt_qplib_cq *cq, struct bnxt_qplib_cqe *cqe, + int num_cqes, struct bnxt_qplib_qp **lib_qp) +{ + struct cq_base *hw_cqe, **hw_cqe_ptr; + unsigned long flags; + u32 sw_cons, raw_cons; + int budget, rc = 0; + + spin_lock_irqsave(&cq->hwq.lock, flags); + raw_cons = cq->hwq.cons; + budget = num_cqes; + + while (budget) { + sw_cons = HWQ_CMP(raw_cons, &cq->hwq); + hw_cqe_ptr = (struct cq_base **)cq->hwq.pbl_ptr; + hw_cqe = &hw_cqe_ptr[CQE_PG(sw_cons)][CQE_IDX(sw_cons)]; + + /* Check for Valid bit */ + if (!CQE_CMP_VALID(hw_cqe, raw_cons, cq->hwq.max_elements)) + break; + + /* From the device's respective CQE format to qplib_wc*/ + switch (hw_cqe->cqe_type_toggle & CQ_BASE_CQE_TYPE_MASK) { + case CQ_BASE_CQE_TYPE_REQ: + rc = bnxt_qplib_cq_process_req(cq, + (struct cq_req *)hw_cqe, + &cqe, &budget, + sw_cons, lib_qp); + break; + case CQ_BASE_CQE_TYPE_RES_RC: + rc = bnxt_qplib_cq_process_res_rc(cq, + (struct cq_res_rc *) + hw_cqe, &cqe, + &budget); + break; + case CQ_BASE_CQE_TYPE_RES_UD: + rc = bnxt_qplib_cq_process_res_ud + (cq, (struct cq_res_ud *)hw_cqe, &cqe, + &budget); + break; + case CQ_BASE_CQE_TYPE_RES_RAWETH_QP1: + rc = bnxt_qplib_cq_process_res_raweth_qp1 + (cq, (struct cq_res_raweth_qp1 *) + hw_cqe, &cqe, &budget); + break; + case CQ_BASE_CQE_TYPE_TERMINAL: + rc = bnxt_qplib_cq_process_terminal + (cq, (struct cq_terminal *)hw_cqe, + &cqe, &budget); + break; + case CQ_BASE_CQE_TYPE_CUT_OFF: + bnxt_qplib_cq_process_cutoff + (cq, (struct cq_cutoff *)hw_cqe); + /* Done processing this CQ */ + goto exit; + default: + dev_err(&cq->hwq.pdev->dev, + "QPLIB: process_cq unknown type 0x%lx", + hw_cqe->cqe_type_toggle & + CQ_BASE_CQE_TYPE_MASK); + rc = -EINVAL; + break; + } + if (rc < 0) { + if (rc == -EAGAIN) + break; + /* Error while processing the CQE, just skip to the + * next one + */ + dev_err(&cq->hwq.pdev->dev, + "QPLIB: process_cqe error rc = 0x%x", rc); + } + raw_cons++; + } + if (cq->hwq.cons != raw_cons) { + cq->hwq.cons = raw_cons; + bnxt_qplib_arm_cq(cq, DBR_DBR_TYPE_CQ); + } +exit: + spin_unlock_irqrestore(&cq->hwq.lock, flags); + return num_cqes - budget; +} + +void bnxt_qplib_req_notify_cq(struct bnxt_qplib_cq *cq, u32 arm_type) +{ + unsigned long flags; + + spin_lock_irqsave(&cq->hwq.lock, flags); + if (arm_type) + bnxt_qplib_arm_cq(cq, arm_type); + + spin_unlock_irqrestore(&cq->hwq.lock, flags); +} diff --git a/drivers/infiniband/hw/bnxt_re/qplib_fp.h b/drivers/infiniband/hw/bnxt_re/qplib_fp.h new file mode 100644 index 000000000000..19176e06c98a --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/qplib_fp.h @@ -0,0 +1,456 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * 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 + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. 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. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: Fast Path Operators (header) + */ + +#ifndef __BNXT_QPLIB_FP_H__ +#define __BNXT_QPLIB_FP_H__ + +struct bnxt_qplib_sge { + u64 addr; + u32 lkey; + u32 size; +}; + +#define BNXT_QPLIB_MAX_SQE_ENTRY_SIZE sizeof(struct sq_send) + +#define SQE_CNT_PER_PG (PAGE_SIZE / BNXT_QPLIB_MAX_SQE_ENTRY_SIZE) +#define SQE_MAX_IDX_PER_PG (SQE_CNT_PER_PG - 1) + +static inline u32 get_sqe_pg(u32 val) +{ + return ((val & ~SQE_MAX_IDX_PER_PG) / SQE_CNT_PER_PG); +} + +static inline u32 get_sqe_idx(u32 val) +{ + return (val & SQE_MAX_IDX_PER_PG); +} + +#define BNXT_QPLIB_MAX_PSNE_ENTRY_SIZE sizeof(struct sq_psn_search) + +#define PSNE_CNT_PER_PG (PAGE_SIZE / BNXT_QPLIB_MAX_PSNE_ENTRY_SIZE) +#define PSNE_MAX_IDX_PER_PG (PSNE_CNT_PER_PG - 1) + +static inline u32 get_psne_pg(u32 val) +{ + return ((val & ~PSNE_MAX_IDX_PER_PG) / PSNE_CNT_PER_PG); +} + +static inline u32 get_psne_idx(u32 val) +{ + return (val & PSNE_MAX_IDX_PER_PG); +} + +#define BNXT_QPLIB_QP_MAX_SGL 6 + +struct bnxt_qplib_swq { + u64 wr_id; + u8 type; + u8 flags; + u32 start_psn; + u32 next_psn; + struct sq_psn_search *psn_search; +}; + +struct bnxt_qplib_swqe { + /* General */ +#define BNXT_QPLIB_FENCE_WRID 0x46454E43 /* "FENC" */ + u64 wr_id; + u8 reqs_type; + u8 type; +#define BNXT_QPLIB_SWQE_TYPE_SEND 0 +#define BNXT_QPLIB_SWQE_TYPE_SEND_WITH_IMM 1 +#define BNXT_QPLIB_SWQE_TYPE_SEND_WITH_INV 2 +#define BNXT_QPLIB_SWQE_TYPE_RDMA_WRITE 4 +#define BNXT_QPLIB_SWQE_TYPE_RDMA_WRITE_WITH_IMM 5 +#define BNXT_QPLIB_SWQE_TYPE_RDMA_READ 6 +#define BNXT_QPLIB_SWQE_TYPE_ATOMIC_CMP_AND_SWP 8 +#define BNXT_QPLIB_SWQE_TYPE_ATOMIC_FETCH_AND_ADD 11 +#define BNXT_QPLIB_SWQE_TYPE_LOCAL_INV 12 +#define BNXT_QPLIB_SWQE_TYPE_FAST_REG_MR 13 +#define BNXT_QPLIB_SWQE_TYPE_REG_MR 13 +#define BNXT_QPLIB_SWQE_TYPE_BIND_MW 14 +#define BNXT_QPLIB_SWQE_TYPE_RECV 128 +#define BNXT_QPLIB_SWQE_TYPE_RECV_RDMA_IMM 129 + u8 flags; +#define BNXT_QPLIB_SWQE_FLAGS_SIGNAL_COMP BIT(0) +#define BNXT_QPLIB_SWQE_FLAGS_RD_ATOMIC_FENCE BIT(1) +#define BNXT_QPLIB_SWQE_FLAGS_UC_FENCE BIT(2) +#define BNXT_QPLIB_SWQE_FLAGS_SOLICIT_EVENT BIT(3) +#define BNXT_QPLIB_SWQE_FLAGS_INLINE BIT(4) + struct bnxt_qplib_sge sg_list[BNXT_QPLIB_QP_MAX_SGL]; + int num_sge; + /* Max inline data is 96 bytes */ + u32 inline_len; +#define BNXT_QPLIB_SWQE_MAX_INLINE_LENGTH 96 + u8 inline_data[BNXT_QPLIB_SWQE_MAX_INLINE_LENGTH]; + + union { + /* Send, with imm, inval key */ + struct { + union { + __be32 imm_data; + u32 inv_key; + }; + u32 q_key; + u32 dst_qp; + u16 avid; + } send; + + /* Send Raw Ethernet and QP1 */ + struct { + u16 lflags; + u16 cfa_action; + u32 cfa_meta; + } rawqp1; + + /* RDMA write, with imm, read */ + struct { + union { + __be32 imm_data; + u32 inv_key; + }; + u64 remote_va; + u32 r_key; + } rdma; + + /* Atomic cmp/swap, fetch/add */ + struct { + u64 remote_va; + u32 r_key; + u64 swap_data; + u64 cmp_data; + } atomic; + + /* Local Invalidate */ + struct { + u32 inv_l_key; + } local_inv; + + /* FR-PMR */ + struct { + u8 access_cntl; + u8 pg_sz_log; + bool zero_based; + u32 l_key; + u32 length; + u8 pbl_pg_sz_log; +#define BNXT_QPLIB_SWQE_PAGE_SIZE_4K 0 +#define BNXT_QPLIB_SWQE_PAGE_SIZE_8K 1 +#define BNXT_QPLIB_SWQE_PAGE_SIZE_64K 4 +#define BNXT_QPLIB_SWQE_PAGE_SIZE_256K 6 +#define BNXT_QPLIB_SWQE_PAGE_SIZE_1M 8 +#define BNXT_QPLIB_SWQE_PAGE_SIZE_2M 9 +#define BNXT_QPLIB_SWQE_PAGE_SIZE_4M 10 +#define BNXT_QPLIB_SWQE_PAGE_SIZE_1G 18 + u8 levels; +#define PAGE_SHIFT_4K 12 + __le64 *pbl_ptr; + dma_addr_t pbl_dma_ptr; + u64 *page_list; + u16 page_list_len; + u64 va; + } frmr; + + /* Bind */ + struct { + u8 access_cntl; +#define BNXT_QPLIB_BIND_SWQE_ACCESS_LOCAL_WRITE BIT(0) +#define BNXT_QPLIB_BIND_SWQE_ACCESS_REMOTE_READ BIT(1) +#define BNXT_QPLIB_BIND_SWQE_ACCESS_REMOTE_WRITE BIT(2) +#define BNXT_QPLIB_BIND_SWQE_ACCESS_REMOTE_ATOMIC BIT(3) +#define BNXT_QPLIB_BIND_SWQE_ACCESS_WINDOW_BIND BIT(4) + bool zero_based; + u8 mw_type; + u32 parent_l_key; + u32 r_key; + u64 va; + u32 length; + } bind; + }; +}; + +#define BNXT_QPLIB_MAX_RQE_ENTRY_SIZE sizeof(struct rq_wqe) + +#define RQE_CNT_PER_PG (PAGE_SIZE / BNXT_QPLIB_MAX_RQE_ENTRY_SIZE) +#define RQE_MAX_IDX_PER_PG (RQE_CNT_PER_PG - 1) +#define RQE_PG(x) (((x) & ~RQE_MAX_IDX_PER_PG) / RQE_CNT_PER_PG) +#define RQE_IDX(x) ((x) & RQE_MAX_IDX_PER_PG) + +struct bnxt_qplib_q { + struct bnxt_qplib_hwq hwq; + struct bnxt_qplib_swq *swq; + struct scatterlist *sglist; + u32 nmap; + u32 max_wqe; + u16 q_full_delta; + u16 max_sge; + u32 psn; + bool flush_in_progress; + bool condition; + bool single; + bool send_phantom; + u32 phantom_wqe_cnt; + u32 phantom_cqe_cnt; + u32 next_cq_cons; +}; + +struct bnxt_qplib_qp { + struct bnxt_qplib_pd *pd; + struct bnxt_qplib_dpi *dpi; + u64 qp_handle; + u32 id; + u8 type; + u8 sig_type; + u32 modify_flags; + u8 state; + u8 cur_qp_state; + u32 max_inline_data; + u32 mtu; + u8 path_mtu; + bool en_sqd_async_notify; + u16 pkey_index; + u32 qkey; + u32 dest_qp_id; + u8 access; + u8 timeout; + u8 retry_cnt; + u8 rnr_retry; + u64 wqe_cnt; + u32 min_rnr_timer; + u32 max_rd_atomic; + u32 max_dest_rd_atomic; + u32 dest_qpn; + u8 smac[6]; + u16 vlan_id; + u8 nw_type; + struct bnxt_qplib_ah ah; + +#define BTH_PSN_MASK ((1 << 24) - 1) + /* SQ */ + struct bnxt_qplib_q sq; + /* RQ */ + struct bnxt_qplib_q rq; + /* SRQ */ + struct bnxt_qplib_srq *srq; + /* CQ */ + struct bnxt_qplib_cq *scq; + struct bnxt_qplib_cq *rcq; + /* IRRQ and ORRQ */ + struct bnxt_qplib_hwq irrq; + struct bnxt_qplib_hwq orrq; + /* Header buffer for QP1 */ + int sq_hdr_buf_size; + int rq_hdr_buf_size; +/* + * Buffer space for ETH(14), IP or GRH(40), UDP header(8) + * and ib_bth + ib_deth (20). + * Max required is 82 when RoCE V2 is enabled + */ +#define BNXT_QPLIB_MAX_QP1_SQ_HDR_SIZE_V2 86 + /* Ethernet header = 14 */ + /* ib_grh = 40 (provided by MAD) */ + /* ib_bth + ib_deth = 20 */ + /* MAD = 256 (provided by MAD) */ + /* iCRC = 4 */ +#define BNXT_QPLIB_MAX_QP1_RQ_ETH_HDR_SIZE 14 +#define BNXT_QPLIB_MAX_QP1_RQ_HDR_SIZE_V2 512 +#define BNXT_QPLIB_MAX_GRH_HDR_SIZE_IPV4 20 +#define BNXT_QPLIB_MAX_GRH_HDR_SIZE_IPV6 40 +#define BNXT_QPLIB_MAX_QP1_RQ_BDETH_HDR_SIZE 20 + void *sq_hdr_buf; + dma_addr_t sq_hdr_buf_map; + void *rq_hdr_buf; + dma_addr_t rq_hdr_buf_map; +}; + +#define BNXT_QPLIB_MAX_CQE_ENTRY_SIZE sizeof(struct cq_base) + +#define CQE_CNT_PER_PG (PAGE_SIZE / BNXT_QPLIB_MAX_CQE_ENTRY_SIZE) +#define CQE_MAX_IDX_PER_PG (CQE_CNT_PER_PG - 1) +#define CQE_PG(x) (((x) & ~CQE_MAX_IDX_PER_PG) / CQE_CNT_PER_PG) +#define CQE_IDX(x) ((x) & CQE_MAX_IDX_PER_PG) + +#define ROCE_CQE_CMP_V 0 +#define CQE_CMP_VALID(hdr, raw_cons, cp_bit) \ + (!!((hdr)->cqe_type_toggle & CQ_BASE_TOGGLE) == \ + !((raw_cons) & (cp_bit))) + +static inline bool bnxt_qplib_queue_full(struct bnxt_qplib_q *qplib_q) +{ + return HWQ_CMP((qplib_q->hwq.prod + qplib_q->q_full_delta), + &qplib_q->hwq) == HWQ_CMP(qplib_q->hwq.cons, + &qplib_q->hwq); +} + +struct bnxt_qplib_cqe { + u8 status; + u8 type; + u8 opcode; + u32 length; + u64 wr_id; + union { + __be32 immdata; + u32 invrkey; + }; + u64 qp_handle; + u64 mr_handle; + u16 flags; + u8 smac[6]; + u32 src_qp; + u16 raweth_qp1_flags; + u16 raweth_qp1_errors; + u16 raweth_qp1_cfa_code; + u32 raweth_qp1_flags2; + u32 raweth_qp1_metadata; + u8 raweth_qp1_payload_offset; + u16 pkey_index; +}; + +#define BNXT_QPLIB_QUEUE_START_PERIOD 0x01 +struct bnxt_qplib_cq { + struct bnxt_qplib_dpi *dpi; + void __iomem *dbr_base; + u32 max_wqe; + u32 id; + u16 count; + u16 period; + struct bnxt_qplib_hwq hwq; + u32 cnq_hw_ring_id; + bool resize_in_progress; + struct scatterlist *sghead; + u32 nmap; + u64 cq_handle; + +#define CQ_RESIZE_WAIT_TIME_MS 500 + unsigned long flags; +#define CQ_FLAGS_RESIZE_IN_PROG 1 + wait_queue_head_t waitq; +}; + +#define BNXT_QPLIB_MAX_IRRQE_ENTRY_SIZE sizeof(struct xrrq_irrq) +#define BNXT_QPLIB_MAX_ORRQE_ENTRY_SIZE sizeof(struct xrrq_orrq) +#define IRD_LIMIT_TO_IRRQ_SLOTS(x) (2 * (x) + 2) +#define IRRQ_SLOTS_TO_IRD_LIMIT(s) (((s) >> 1) - 1) +#define ORD_LIMIT_TO_ORRQ_SLOTS(x) ((x) + 1) +#define ORRQ_SLOTS_TO_ORD_LIMIT(s) ((s) - 1) + +#define BNXT_QPLIB_MAX_NQE_ENTRY_SIZE sizeof(struct nq_base) + +#define NQE_CNT_PER_PG (PAGE_SIZE / BNXT_QPLIB_MAX_NQE_ENTRY_SIZE) +#define NQE_MAX_IDX_PER_PG (NQE_CNT_PER_PG - 1) +#define NQE_PG(x) (((x) & ~NQE_MAX_IDX_PER_PG) / NQE_CNT_PER_PG) +#define NQE_IDX(x) ((x) & NQE_MAX_IDX_PER_PG) + +#define NQE_CMP_VALID(hdr, raw_cons, cp_bit) \ + (!!(le32_to_cpu((hdr)->info63_v[0]) & NQ_BASE_V) == \ + !((raw_cons) & (cp_bit))) + +#define BNXT_QPLIB_NQE_MAX_CNT (128 * 1024) + +#define NQ_CONS_PCI_BAR_REGION 2 +#define NQ_DB_KEY_CP (0x2 << CMPL_DOORBELL_KEY_SFT) +#define NQ_DB_IDX_VALID CMPL_DOORBELL_IDX_VALID +#define NQ_DB_IRQ_DIS CMPL_DOORBELL_MASK +#define NQ_DB_CP_FLAGS_REARM (NQ_DB_KEY_CP | \ + NQ_DB_IDX_VALID) +#define NQ_DB_CP_FLAGS (NQ_DB_KEY_CP | \ + NQ_DB_IDX_VALID | \ + NQ_DB_IRQ_DIS) +#define NQ_DB_REARM(db, raw_cons, cp_bit) \ + writel(NQ_DB_CP_FLAGS_REARM | ((raw_cons) & ((cp_bit) - 1)), db) +#define NQ_DB(db, raw_cons, cp_bit) \ + writel(NQ_DB_CP_FLAGS | ((raw_cons) & ((cp_bit) - 1)), db) + +struct bnxt_qplib_nq { + struct pci_dev *pdev; + + int vector; + int budget; + bool requested; + struct tasklet_struct worker; + struct bnxt_qplib_hwq hwq; + + u16 bar_reg; + u16 bar_reg_off; + u16 ring_id; + void __iomem *bar_reg_iomem; + + int (*cqn_handler) + (struct bnxt_qplib_nq *nq, + struct bnxt_qplib_cq *cq); + int (*srqn_handler) + (struct bnxt_qplib_nq *nq, + void *srq, + u8 event); +}; + +void bnxt_qplib_disable_nq(struct bnxt_qplib_nq *nq); +int bnxt_qplib_enable_nq(struct pci_dev *pdev, struct bnxt_qplib_nq *nq, + int msix_vector, int bar_reg_offset, + int (*cqn_handler)(struct bnxt_qplib_nq *nq, + struct bnxt_qplib_cq *cq), + int (*srqn_handler)(struct bnxt_qplib_nq *nq, + void *srq, + u8 event)); +int bnxt_qplib_create_qp1(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp); +int bnxt_qplib_create_qp(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp); +int bnxt_qplib_modify_qp(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp); +int bnxt_qplib_query_qp(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp); +int bnxt_qplib_destroy_qp(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp); +void *bnxt_qplib_get_qp1_sq_buf(struct bnxt_qplib_qp *qp, + struct bnxt_qplib_sge *sge); +void *bnxt_qplib_get_qp1_rq_buf(struct bnxt_qplib_qp *qp, + struct bnxt_qplib_sge *sge); +u32 bnxt_qplib_get_rq_prod_index(struct bnxt_qplib_qp *qp); +dma_addr_t bnxt_qplib_get_qp_buf_from_index(struct bnxt_qplib_qp *qp, + u32 index); +void bnxt_qplib_post_send_db(struct bnxt_qplib_qp *qp); +int bnxt_qplib_post_send(struct bnxt_qplib_qp *qp, + struct bnxt_qplib_swqe *wqe); +void bnxt_qplib_post_recv_db(struct bnxt_qplib_qp *qp); +int bnxt_qplib_post_recv(struct bnxt_qplib_qp *qp, + struct bnxt_qplib_swqe *wqe); +int bnxt_qplib_create_cq(struct bnxt_qplib_res *res, struct bnxt_qplib_cq *cq); +int bnxt_qplib_destroy_cq(struct bnxt_qplib_res *res, struct bnxt_qplib_cq *cq); +int bnxt_qplib_poll_cq(struct bnxt_qplib_cq *cq, struct bnxt_qplib_cqe *cqe, + int num, struct bnxt_qplib_qp **qp); +bool bnxt_qplib_is_cq_empty(struct bnxt_qplib_cq *cq); +void bnxt_qplib_req_notify_cq(struct bnxt_qplib_cq *cq, u32 arm_type); +void bnxt_qplib_free_nq(struct bnxt_qplib_nq *nq); +int bnxt_qplib_alloc_nq(struct pci_dev *pdev, struct bnxt_qplib_nq *nq); +#endif /* __BNXT_QPLIB_FP_H__ */ diff --git a/drivers/infiniband/hw/bnxt_re/qplib_rcfw.c b/drivers/infiniband/hw/bnxt_re/qplib_rcfw.c new file mode 100644 index 000000000000..16e42754dbec --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/qplib_rcfw.c @@ -0,0 +1,708 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * 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 + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. 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. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: RDMA Controller HW interface + */ +#include <linux/interrupt.h> +#include <linux/spinlock.h> +#include <linux/pci.h> +#include <linux/prefetch.h> +#include <linux/delay.h> + +#include "roce_hsi.h" +#include "qplib_res.h" +#include "qplib_rcfw.h" +static void bnxt_qplib_service_creq(unsigned long data); + +/* Hardware communication channel */ +static int __wait_for_resp(struct bnxt_qplib_rcfw *rcfw, u16 cookie) +{ + u16 cbit; + int rc; + + cbit = cookie % RCFW_MAX_OUTSTANDING_CMD; + rc = wait_event_timeout(rcfw->waitq, + !test_bit(cbit, rcfw->cmdq_bitmap), + msecs_to_jiffies(RCFW_CMD_WAIT_TIME_MS)); + return rc ? 0 : -ETIMEDOUT; +}; + +static int __block_for_resp(struct bnxt_qplib_rcfw *rcfw, u16 cookie) +{ + u32 count = RCFW_BLOCKED_CMD_WAIT_COUNT; + u16 cbit; + + cbit = cookie % RCFW_MAX_OUTSTANDING_CMD; + if (!test_bit(cbit, rcfw->cmdq_bitmap)) + goto done; + do { + mdelay(1); /* 1m sec */ + bnxt_qplib_service_creq((unsigned long)rcfw); + } while (test_bit(cbit, rcfw->cmdq_bitmap) && --count); +done: + return count ? 0 : -ETIMEDOUT; +}; + +static int __send_message(struct bnxt_qplib_rcfw *rcfw, struct cmdq_base *req, + struct creq_base *resp, void *sb, u8 is_block) +{ + struct bnxt_qplib_cmdqe *cmdqe, **cmdq_ptr; + struct bnxt_qplib_hwq *cmdq = &rcfw->cmdq; + struct bnxt_qplib_crsq *crsqe; + u32 sw_prod, cmdq_prod; + unsigned long flags; + u32 size, opcode; + u16 cookie, cbit; + int pg, idx; + u8 *preq; + + opcode = req->opcode; + if (!test_bit(FIRMWARE_INITIALIZED_FLAG, &rcfw->flags) && + (opcode != CMDQ_BASE_OPCODE_QUERY_FUNC && + opcode != CMDQ_BASE_OPCODE_INITIALIZE_FW)) { + dev_err(&rcfw->pdev->dev, + "QPLIB: RCFW not initialized, reject opcode 0x%x", + opcode); + return -EINVAL; + } + + if (test_bit(FIRMWARE_INITIALIZED_FLAG, &rcfw->flags) && + opcode == CMDQ_BASE_OPCODE_INITIALIZE_FW) { + dev_err(&rcfw->pdev->dev, "QPLIB: RCFW already initialized!"); + return -EINVAL; + } + + /* Cmdq are in 16-byte units, each request can consume 1 or more + * cmdqe + */ + spin_lock_irqsave(&cmdq->lock, flags); + if (req->cmd_size >= HWQ_FREE_SLOTS(cmdq)) { + dev_err(&rcfw->pdev->dev, "QPLIB: RCFW: CMDQ is full!"); + spin_unlock_irqrestore(&cmdq->lock, flags); + return -EAGAIN; + } + + + cookie = rcfw->seq_num & RCFW_MAX_COOKIE_VALUE; + cbit = cookie % RCFW_MAX_OUTSTANDING_CMD; + if (is_block) + cookie |= RCFW_CMD_IS_BLOCKING; + + set_bit(cbit, rcfw->cmdq_bitmap); + req->cookie = cpu_to_le16(cookie); + crsqe = &rcfw->crsqe_tbl[cbit]; + if (crsqe->resp) { + spin_unlock_irqrestore(&cmdq->lock, flags); + return -EBUSY; + } + memset(resp, 0, sizeof(*resp)); + crsqe->resp = (struct creq_qp_event *)resp; + crsqe->resp->cookie = req->cookie; + crsqe->req_size = req->cmd_size; + if (req->resp_size && sb) { + struct bnxt_qplib_rcfw_sbuf *sbuf = sb; + + req->resp_addr = cpu_to_le64(sbuf->dma_addr); + req->resp_size = (sbuf->size + BNXT_QPLIB_CMDQE_UNITS - 1) / + BNXT_QPLIB_CMDQE_UNITS; + } + + cmdq_ptr = (struct bnxt_qplib_cmdqe **)cmdq->pbl_ptr; + preq = (u8 *)req; + size = req->cmd_size * BNXT_QPLIB_CMDQE_UNITS; + do { + pg = 0; + idx = 0; + + /* Locate the next cmdq slot */ + sw_prod = HWQ_CMP(cmdq->prod, cmdq); + cmdqe = &cmdq_ptr[get_cmdq_pg(sw_prod)][get_cmdq_idx(sw_prod)]; + if (!cmdqe) { + dev_err(&rcfw->pdev->dev, + "QPLIB: RCFW request failed with no cmdqe!"); + goto done; + } + /* Copy a segment of the req cmd to the cmdq */ + memset(cmdqe, 0, sizeof(*cmdqe)); + memcpy(cmdqe, preq, min_t(u32, size, sizeof(*cmdqe))); + preq += min_t(u32, size, sizeof(*cmdqe)); + size -= min_t(u32, size, sizeof(*cmdqe)); + cmdq->prod++; + rcfw->seq_num++; + } while (size > 0); + + rcfw->seq_num++; + + cmdq_prod = cmdq->prod; + if (rcfw->flags & FIRMWARE_FIRST_FLAG) { + /* The very first doorbell write + * is required to set this flag + * which prompts the FW to reset + * its internal pointers + */ + cmdq_prod |= FIRMWARE_FIRST_FLAG; + rcfw->flags &= ~FIRMWARE_FIRST_FLAG; + } + + /* ring CMDQ DB */ + wmb(); + writel(cmdq_prod, rcfw->cmdq_bar_reg_iomem + + rcfw->cmdq_bar_reg_prod_off); + writel(RCFW_CMDQ_TRIG_VAL, rcfw->cmdq_bar_reg_iomem + + rcfw->cmdq_bar_reg_trig_off); +done: + spin_unlock_irqrestore(&cmdq->lock, flags); + /* Return the CREQ response pointer */ + return 0; +} + +int bnxt_qplib_rcfw_send_message(struct bnxt_qplib_rcfw *rcfw, + struct cmdq_base *req, + struct creq_base *resp, + void *sb, u8 is_block) +{ + struct creq_qp_event *evnt = (struct creq_qp_event *)resp; + u16 cookie; + u8 opcode, retry_cnt = 0xFF; + int rc = 0; + + do { + opcode = req->opcode; + rc = __send_message(rcfw, req, resp, sb, is_block); + cookie = le16_to_cpu(req->cookie) & RCFW_MAX_COOKIE_VALUE; + if (!rc) + break; + + if (!retry_cnt || (rc != -EAGAIN && rc != -EBUSY)) { + /* send failed */ + dev_err(&rcfw->pdev->dev, "QPLIB: cmdq[%#x]=%#x send failed", + cookie, opcode); + return rc; + } + is_block ? mdelay(1) : usleep_range(500, 1000); + + } while (retry_cnt--); + + if (is_block) + rc = __block_for_resp(rcfw, cookie); + else + rc = __wait_for_resp(rcfw, cookie); + if (rc) { + /* timed out */ + dev_err(&rcfw->pdev->dev, "QPLIB: cmdq[%#x]=%#x timedout (%d)msec", + cookie, opcode, RCFW_CMD_WAIT_TIME_MS); + return rc; + } + + if (evnt->status) { + /* failed with status */ + dev_err(&rcfw->pdev->dev, "QPLIB: cmdq[%#x]=%#x status %#x", + cookie, opcode, evnt->status); + rc = -EFAULT; + } + + return rc; +} +/* Completions */ +static int bnxt_qplib_process_func_event(struct bnxt_qplib_rcfw *rcfw, + struct creq_func_event *func_event) +{ + switch (func_event->event) { + case CREQ_FUNC_EVENT_EVENT_TX_WQE_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_TX_DATA_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_RX_WQE_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_RX_DATA_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_CQ_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_TQM_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_CFCQ_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_CFCS_ERROR: + /* SRQ ctx error, call srq_handler?? + * But there's no SRQ handle! + */ + break; + case CREQ_FUNC_EVENT_EVENT_CFCC_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_CFCM_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_TIM_ERROR: + break; + case CREQ_FUNC_EVENT_EVENT_VF_COMM_REQUEST: + break; + case CREQ_FUNC_EVENT_EVENT_RESOURCE_EXHAUSTED: + break; + default: + return -EINVAL; + } + return 0; +} + +static int bnxt_qplib_process_qp_event(struct bnxt_qplib_rcfw *rcfw, + struct creq_qp_event *qp_event) +{ + struct bnxt_qplib_hwq *cmdq = &rcfw->cmdq; + struct bnxt_qplib_crsq *crsqe; + unsigned long flags; + u16 cbit, blocked = 0; + u16 cookie; + __le16 mcookie; + + switch (qp_event->event) { + case CREQ_QP_EVENT_EVENT_QP_ERROR_NOTIFICATION: + dev_dbg(&rcfw->pdev->dev, + "QPLIB: Received QP error notification"); + break; + default: + /* Command Response */ + spin_lock_irqsave(&cmdq->lock, flags); + cookie = le16_to_cpu(qp_event->cookie); + mcookie = qp_event->cookie; + blocked = cookie & RCFW_CMD_IS_BLOCKING; + cookie &= RCFW_MAX_COOKIE_VALUE; + cbit = cookie % RCFW_MAX_OUTSTANDING_CMD; + crsqe = &rcfw->crsqe_tbl[cbit]; + if (crsqe->resp && + crsqe->resp->cookie == mcookie) { + memcpy(crsqe->resp, qp_event, sizeof(*qp_event)); + crsqe->resp = NULL; + } else { + dev_err(&rcfw->pdev->dev, + "QPLIB: CMD %s resp->cookie = %#x, evnt->cookie = %#x", + crsqe->resp ? "mismatch" : "collision", + crsqe->resp ? crsqe->resp->cookie : 0, mcookie); + } + if (!test_and_clear_bit(cbit, rcfw->cmdq_bitmap)) + dev_warn(&rcfw->pdev->dev, + "QPLIB: CMD bit %d was not requested", cbit); + cmdq->cons += crsqe->req_size; + crsqe->req_size = 0; + + if (!blocked) + wake_up(&rcfw->waitq); + spin_unlock_irqrestore(&cmdq->lock, flags); + } + return 0; +} + +/* SP - CREQ Completion handlers */ +static void bnxt_qplib_service_creq(unsigned long data) +{ + struct bnxt_qplib_rcfw *rcfw = (struct bnxt_qplib_rcfw *)data; + struct bnxt_qplib_hwq *creq = &rcfw->creq; + struct creq_base *creqe, **creq_ptr; + u32 sw_cons, raw_cons; + unsigned long flags; + u32 type, budget = CREQ_ENTRY_POLL_BUDGET; + + /* Service the CREQ until budget is over */ + spin_lock_irqsave(&creq->lock, flags); + raw_cons = creq->cons; + while (budget > 0) { + sw_cons = HWQ_CMP(raw_cons, creq); + creq_ptr = (struct creq_base **)creq->pbl_ptr; + creqe = &creq_ptr[get_creq_pg(sw_cons)][get_creq_idx(sw_cons)]; + if (!CREQ_CMP_VALID(creqe, raw_cons, creq->max_elements)) + break; + + type = creqe->type & CREQ_BASE_TYPE_MASK; + switch (type) { + case CREQ_BASE_TYPE_QP_EVENT: + bnxt_qplib_process_qp_event + (rcfw, (struct creq_qp_event *)creqe); + rcfw->creq_qp_event_processed++; + break; + case CREQ_BASE_TYPE_FUNC_EVENT: + if (!bnxt_qplib_process_func_event + (rcfw, (struct creq_func_event *)creqe)) + rcfw->creq_func_event_processed++; + else + dev_warn + (&rcfw->pdev->dev, "QPLIB:aeqe:%#x Not handled", + type); + break; + default: + dev_warn(&rcfw->pdev->dev, "QPLIB: creqe with "); + dev_warn(&rcfw->pdev->dev, + "QPLIB: op_event = 0x%x not handled", type); + break; + } + raw_cons++; + budget--; + } + + if (creq->cons != raw_cons) { + creq->cons = raw_cons; + CREQ_DB_REARM(rcfw->creq_bar_reg_iomem, raw_cons, + creq->max_elements); + } + spin_unlock_irqrestore(&creq->lock, flags); +} + +static irqreturn_t bnxt_qplib_creq_irq(int irq, void *dev_instance) +{ + struct bnxt_qplib_rcfw *rcfw = dev_instance; + struct bnxt_qplib_hwq *creq = &rcfw->creq; + struct creq_base **creq_ptr; + u32 sw_cons; + + /* Prefetch the CREQ element */ + sw_cons = HWQ_CMP(creq->cons, creq); + creq_ptr = (struct creq_base **)rcfw->creq.pbl_ptr; + prefetch(&creq_ptr[get_creq_pg(sw_cons)][get_creq_idx(sw_cons)]); + + tasklet_schedule(&rcfw->worker); + + return IRQ_HANDLED; +} + +/* RCFW */ +int bnxt_qplib_deinit_rcfw(struct bnxt_qplib_rcfw *rcfw) +{ + struct cmdq_deinitialize_fw req; + struct creq_deinitialize_fw_resp resp; + u16 cmd_flags = 0; + int rc; + + RCFW_CMD_PREP(req, DEINITIALIZE_FW, cmd_flags); + rc = bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, (void *)&resp, + NULL, 0); + if (rc) + return rc; + + clear_bit(FIRMWARE_INITIALIZED_FLAG, &rcfw->flags); + return 0; +} + +static int __get_pbl_pg_idx(struct bnxt_qplib_pbl *pbl) +{ + return (pbl->pg_size == ROCE_PG_SIZE_4K ? + CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_4K : + pbl->pg_size == ROCE_PG_SIZE_8K ? + CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_8K : + pbl->pg_size == ROCE_PG_SIZE_64K ? + CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_64K : + pbl->pg_size == ROCE_PG_SIZE_2M ? + CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_2M : + pbl->pg_size == ROCE_PG_SIZE_8M ? + CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_8M : + pbl->pg_size == ROCE_PG_SIZE_1G ? + CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_1G : + CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_4K); +} + +int bnxt_qplib_init_rcfw(struct bnxt_qplib_rcfw *rcfw, + struct bnxt_qplib_ctx *ctx, int is_virtfn) +{ + struct cmdq_initialize_fw req; + struct creq_initialize_fw_resp resp; + u16 cmd_flags = 0, level; + int rc; + + RCFW_CMD_PREP(req, INITIALIZE_FW, cmd_flags); + + /* + * VFs need not setup the HW context area, PF + * shall setup this area for VF. Skipping the + * HW programming + */ + if (is_virtfn) + goto skip_ctx_setup; + + level = ctx->qpc_tbl.level; + req.qpc_pg_size_qpc_lvl = (level << CMDQ_INITIALIZE_FW_QPC_LVL_SFT) | + __get_pbl_pg_idx(&ctx->qpc_tbl.pbl[level]); + level = ctx->mrw_tbl.level; + req.mrw_pg_size_mrw_lvl = (level << CMDQ_INITIALIZE_FW_MRW_LVL_SFT) | + __get_pbl_pg_idx(&ctx->mrw_tbl.pbl[level]); + level = ctx->srqc_tbl.level; + req.srq_pg_size_srq_lvl = (level << CMDQ_INITIALIZE_FW_SRQ_LVL_SFT) | + __get_pbl_pg_idx(&ctx->srqc_tbl.pbl[level]); + level = ctx->cq_tbl.level; + req.cq_pg_size_cq_lvl = (level << CMDQ_INITIALIZE_FW_CQ_LVL_SFT) | + __get_pbl_pg_idx(&ctx->cq_tbl.pbl[level]); + level = ctx->srqc_tbl.level; + req.srq_pg_size_srq_lvl = (level << CMDQ_INITIALIZE_FW_SRQ_LVL_SFT) | + __get_pbl_pg_idx(&ctx->srqc_tbl.pbl[level]); + level = ctx->cq_tbl.level; + req.cq_pg_size_cq_lvl = (level << CMDQ_INITIALIZE_FW_CQ_LVL_SFT) | + __get_pbl_pg_idx(&ctx->cq_tbl.pbl[level]); + level = ctx->tim_tbl.level; + req.tim_pg_size_tim_lvl = (level << CMDQ_INITIALIZE_FW_TIM_LVL_SFT) | + __get_pbl_pg_idx(&ctx->tim_tbl.pbl[level]); + level = ctx->tqm_pde_level; + req.tqm_pg_size_tqm_lvl = (level << CMDQ_INITIALIZE_FW_TQM_LVL_SFT) | + __get_pbl_pg_idx(&ctx->tqm_pde.pbl[level]); + + req.qpc_page_dir = + cpu_to_le64(ctx->qpc_tbl.pbl[PBL_LVL_0].pg_map_arr[0]); + req.mrw_page_dir = + cpu_to_le64(ctx->mrw_tbl.pbl[PBL_LVL_0].pg_map_arr[0]); + req.srq_page_dir = + cpu_to_le64(ctx->srqc_tbl.pbl[PBL_LVL_0].pg_map_arr[0]); + req.cq_page_dir = + cpu_to_le64(ctx->cq_tbl.pbl[PBL_LVL_0].pg_map_arr[0]); + req.tim_page_dir = + cpu_to_le64(ctx->tim_tbl.pbl[PBL_LVL_0].pg_map_arr[0]); + req.tqm_page_dir = + cpu_to_le64(ctx->tqm_pde.pbl[PBL_LVL_0].pg_map_arr[0]); + + req.number_of_qp = cpu_to_le32(ctx->qpc_tbl.max_elements); + req.number_of_mrw = cpu_to_le32(ctx->mrw_tbl.max_elements); + req.number_of_srq = cpu_to_le32(ctx->srqc_tbl.max_elements); + req.number_of_cq = cpu_to_le32(ctx->cq_tbl.max_elements); + + req.max_qp_per_vf = cpu_to_le32(ctx->vf_res.max_qp_per_vf); + req.max_mrw_per_vf = cpu_to_le32(ctx->vf_res.max_mrw_per_vf); + req.max_srq_per_vf = cpu_to_le32(ctx->vf_res.max_srq_per_vf); + req.max_cq_per_vf = cpu_to_le32(ctx->vf_res.max_cq_per_vf); + req.max_gid_per_vf = cpu_to_le32(ctx->vf_res.max_gid_per_vf); + +skip_ctx_setup: + req.stat_ctx_id = cpu_to_le32(ctx->stats.fw_id); + rc = bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, (void *)&resp, + NULL, 0); + if (rc) + return rc; + set_bit(FIRMWARE_INITIALIZED_FLAG, &rcfw->flags); + return 0; +} + +void bnxt_qplib_free_rcfw_channel(struct bnxt_qplib_rcfw *rcfw) +{ + kfree(rcfw->crsqe_tbl); + bnxt_qplib_free_hwq(rcfw->pdev, &rcfw->cmdq); + bnxt_qplib_free_hwq(rcfw->pdev, &rcfw->creq); + rcfw->pdev = NULL; +} + +int bnxt_qplib_alloc_rcfw_channel(struct pci_dev *pdev, + struct bnxt_qplib_rcfw *rcfw) +{ + rcfw->pdev = pdev; + rcfw->creq.max_elements = BNXT_QPLIB_CREQE_MAX_CNT; + if (bnxt_qplib_alloc_init_hwq(rcfw->pdev, &rcfw->creq, NULL, 0, + &rcfw->creq.max_elements, + BNXT_QPLIB_CREQE_UNITS, 0, PAGE_SIZE, + HWQ_TYPE_L2_CMPL)) { + dev_err(&rcfw->pdev->dev, + "QPLIB: HW channel CREQ allocation failed"); + goto fail; + } + rcfw->cmdq.max_elements = BNXT_QPLIB_CMDQE_MAX_CNT; + if (bnxt_qplib_alloc_init_hwq(rcfw->pdev, &rcfw->cmdq, NULL, 0, + &rcfw->cmdq.max_elements, + BNXT_QPLIB_CMDQE_UNITS, 0, PAGE_SIZE, + HWQ_TYPE_CTX)) { + dev_err(&rcfw->pdev->dev, + "QPLIB: HW channel CMDQ allocation failed"); + goto fail; + } + + rcfw->crsqe_tbl = kcalloc(rcfw->cmdq.max_elements, + sizeof(*rcfw->crsqe_tbl), GFP_KERNEL); + if (!rcfw->crsqe_tbl) + goto fail; + + return 0; + +fail: + bnxt_qplib_free_rcfw_channel(rcfw); + return -ENOMEM; +} + +void bnxt_qplib_disable_rcfw_channel(struct bnxt_qplib_rcfw *rcfw) +{ + unsigned long indx; + + /* Make sure the HW channel is stopped! */ + synchronize_irq(rcfw->vector); + tasklet_disable(&rcfw->worker); + tasklet_kill(&rcfw->worker); + + if (rcfw->requested) { + free_irq(rcfw->vector, rcfw); + rcfw->requested = false; + } + if (rcfw->cmdq_bar_reg_iomem) + iounmap(rcfw->cmdq_bar_reg_iomem); + rcfw->cmdq_bar_reg_iomem = NULL; + + if (rcfw->creq_bar_reg_iomem) + iounmap(rcfw->creq_bar_reg_iomem); + rcfw->creq_bar_reg_iomem = NULL; + + indx = find_first_bit(rcfw->cmdq_bitmap, rcfw->bmap_size); + if (indx != rcfw->bmap_size) + dev_err(&rcfw->pdev->dev, + "QPLIB: disabling RCFW with pending cmd-bit %lx", indx); + kfree(rcfw->cmdq_bitmap); + rcfw->bmap_size = 0; + + rcfw->aeq_handler = NULL; + rcfw->vector = 0; +} + +int bnxt_qplib_enable_rcfw_channel(struct pci_dev *pdev, + struct bnxt_qplib_rcfw *rcfw, + int msix_vector, + int cp_bar_reg_off, int virt_fn, + int (*aeq_handler)(struct bnxt_qplib_rcfw *, + struct creq_func_event *)) +{ + resource_size_t res_base; + struct cmdq_init init; + u16 bmap_size; + int rc; + + /* General */ + rcfw->seq_num = 0; + rcfw->flags = FIRMWARE_FIRST_FLAG; + bmap_size = BITS_TO_LONGS(RCFW_MAX_OUTSTANDING_CMD * + sizeof(unsigned long)); + rcfw->cmdq_bitmap = kzalloc(bmap_size, GFP_KERNEL); + if (!rcfw->cmdq_bitmap) + return -ENOMEM; + rcfw->bmap_size = bmap_size; + + /* CMDQ */ + rcfw->cmdq_bar_reg = RCFW_COMM_PCI_BAR_REGION; + res_base = pci_resource_start(pdev, rcfw->cmdq_bar_reg); + if (!res_base) + return -ENOMEM; + + rcfw->cmdq_bar_reg_iomem = ioremap_nocache(res_base + + RCFW_COMM_BASE_OFFSET, + RCFW_COMM_SIZE); + if (!rcfw->cmdq_bar_reg_iomem) { + dev_err(&rcfw->pdev->dev, + "QPLIB: CMDQ BAR region %d mapping failed", + rcfw->cmdq_bar_reg); + return -ENOMEM; + } + + rcfw->cmdq_bar_reg_prod_off = virt_fn ? RCFW_VF_COMM_PROD_OFFSET : + RCFW_PF_COMM_PROD_OFFSET; + + rcfw->cmdq_bar_reg_trig_off = RCFW_COMM_TRIG_OFFSET; + + /* CREQ */ + rcfw->creq_bar_reg = RCFW_COMM_CONS_PCI_BAR_REGION; + res_base = pci_resource_start(pdev, rcfw->creq_bar_reg); + if (!res_base) + dev_err(&rcfw->pdev->dev, + "QPLIB: CREQ BAR region %d resc start is 0!", + rcfw->creq_bar_reg); + rcfw->creq_bar_reg_iomem = ioremap_nocache(res_base + cp_bar_reg_off, + 4); + if (!rcfw->creq_bar_reg_iomem) { + dev_err(&rcfw->pdev->dev, + "QPLIB: CREQ BAR region %d mapping failed", + rcfw->creq_bar_reg); + return -ENOMEM; + } + rcfw->creq_qp_event_processed = 0; + rcfw->creq_func_event_processed = 0; + + rcfw->vector = msix_vector; + if (aeq_handler) + rcfw->aeq_handler = aeq_handler; + + tasklet_init(&rcfw->worker, bnxt_qplib_service_creq, + (unsigned long)rcfw); + + rcfw->requested = false; + rc = request_irq(rcfw->vector, bnxt_qplib_creq_irq, 0, + "bnxt_qplib_creq", rcfw); + if (rc) { + dev_err(&rcfw->pdev->dev, + "QPLIB: Failed to request IRQ for CREQ rc = 0x%x", rc); + bnxt_qplib_disable_rcfw_channel(rcfw); + return rc; + } + rcfw->requested = true; + + init_waitqueue_head(&rcfw->waitq); + + CREQ_DB_REARM(rcfw->creq_bar_reg_iomem, 0, rcfw->creq.max_elements); + + init.cmdq_pbl = cpu_to_le64(rcfw->cmdq.pbl[PBL_LVL_0].pg_map_arr[0]); + init.cmdq_size_cmdq_lvl = cpu_to_le16( + ((BNXT_QPLIB_CMDQE_MAX_CNT << CMDQ_INIT_CMDQ_SIZE_SFT) & + CMDQ_INIT_CMDQ_SIZE_MASK) | + ((rcfw->cmdq.level << CMDQ_INIT_CMDQ_LVL_SFT) & + CMDQ_INIT_CMDQ_LVL_MASK)); + init.creq_ring_id = cpu_to_le16(rcfw->creq_ring_id); + + /* Write to the Bono mailbox register */ + __iowrite32_copy(rcfw->cmdq_bar_reg_iomem, &init, sizeof(init) / 4); + return 0; +} + +struct bnxt_qplib_rcfw_sbuf *bnxt_qplib_rcfw_alloc_sbuf( + struct bnxt_qplib_rcfw *rcfw, + u32 size) +{ + struct bnxt_qplib_rcfw_sbuf *sbuf; + + sbuf = kzalloc(sizeof(*sbuf), GFP_ATOMIC); + if (!sbuf) + return NULL; + + sbuf->size = size; + sbuf->sb = dma_zalloc_coherent(&rcfw->pdev->dev, sbuf->size, + &sbuf->dma_addr, GFP_ATOMIC); + if (!sbuf->sb) + goto bail; + + return sbuf; +bail: + kfree(sbuf); + return NULL; +} + +void bnxt_qplib_rcfw_free_sbuf(struct bnxt_qplib_rcfw *rcfw, + struct bnxt_qplib_rcfw_sbuf *sbuf) +{ + if (sbuf->sb) + dma_free_coherent(&rcfw->pdev->dev, sbuf->size, + sbuf->sb, sbuf->dma_addr); + kfree(sbuf); +} diff --git a/drivers/infiniband/hw/bnxt_re/qplib_rcfw.h b/drivers/infiniband/hw/bnxt_re/qplib_rcfw.h new file mode 100644 index 000000000000..09ce121770cd --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/qplib_rcfw.h @@ -0,0 +1,210 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * 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 + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. 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. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: RDMA Controller HW interface (header) + */ + +#ifndef __BNXT_QPLIB_RCFW_H__ +#define __BNXT_QPLIB_RCFW_H__ + +#define RCFW_CMDQ_TRIG_VAL 1 +#define RCFW_COMM_PCI_BAR_REGION 0 +#define RCFW_COMM_CONS_PCI_BAR_REGION 2 +#define RCFW_COMM_BASE_OFFSET 0x600 +#define RCFW_PF_COMM_PROD_OFFSET 0xc +#define RCFW_VF_COMM_PROD_OFFSET 0xc +#define RCFW_COMM_TRIG_OFFSET 0x100 +#define RCFW_COMM_SIZE 0x104 + +#define RCFW_DBR_PCI_BAR_REGION 2 + +#define RCFW_CMD_PREP(req, CMD, cmd_flags) \ + 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).flags = cpu_to_le16(cmd_flags); \ + } while (0) + +#define RCFW_CMD_WAIT_TIME_MS 20000 /* 20 Seconds timeout */ + +/* CMDQ elements */ +#define BNXT_QPLIB_CMDQE_MAX_CNT 256 +#define BNXT_QPLIB_CMDQE_UNITS sizeof(struct bnxt_qplib_cmdqe) +#define BNXT_QPLIB_CMDQE_CNT_PER_PG (PAGE_SIZE / BNXT_QPLIB_CMDQE_UNITS) + +#define MAX_CMDQ_IDX (BNXT_QPLIB_CMDQE_MAX_CNT - 1) +#define MAX_CMDQ_IDX_PER_PG (BNXT_QPLIB_CMDQE_CNT_PER_PG - 1) + +#define RCFW_MAX_OUTSTANDING_CMD BNXT_QPLIB_CMDQE_MAX_CNT +#define RCFW_MAX_COOKIE_VALUE 0x7FFF +#define RCFW_CMD_IS_BLOCKING 0x8000 +#define RCFW_BLOCKED_CMD_WAIT_COUNT 0x4E20 + +/* Cmdq contains a fix number of a 16-Byte slots */ +struct bnxt_qplib_cmdqe { + u8 data[16]; +}; + +static inline u32 get_cmdq_pg(u32 val) +{ + return (val & ~MAX_CMDQ_IDX_PER_PG) / BNXT_QPLIB_CMDQE_CNT_PER_PG; +} + +static inline u32 get_cmdq_idx(u32 val) +{ + return val & MAX_CMDQ_IDX_PER_PG; +} + +/* Crsq buf is 1024-Byte */ +struct bnxt_qplib_crsbe { + u8 data[1024]; +}; + +/* CREQ */ +/* Allocate 1 per QP for async error notification for now */ +#define BNXT_QPLIB_CREQE_MAX_CNT (64 * 1024) +#define BNXT_QPLIB_CREQE_UNITS 16 /* 16-Bytes per prod unit */ +#define BNXT_QPLIB_CREQE_CNT_PER_PG (PAGE_SIZE / BNXT_QPLIB_CREQE_UNITS) + +#define MAX_CREQ_IDX (BNXT_QPLIB_CREQE_MAX_CNT - 1) +#define MAX_CREQ_IDX_PER_PG (BNXT_QPLIB_CREQE_CNT_PER_PG - 1) + +static inline u32 get_creq_pg(u32 val) +{ + return (val & ~MAX_CREQ_IDX_PER_PG) / BNXT_QPLIB_CREQE_CNT_PER_PG; +} + +static inline u32 get_creq_idx(u32 val) +{ + return val & MAX_CREQ_IDX_PER_PG; +} + +#define BNXT_QPLIB_CREQE_PER_PG (PAGE_SIZE / sizeof(struct creq_base)) + +#define CREQ_CMP_VALID(hdr, raw_cons, cp_bit) \ + (!!((hdr)->v & CREQ_BASE_V) == \ + !((raw_cons) & (cp_bit))) + +#define CREQ_DB_KEY_CP (0x2 << CMPL_DOORBELL_KEY_SFT) +#define CREQ_DB_IDX_VALID CMPL_DOORBELL_IDX_VALID +#define CREQ_DB_IRQ_DIS CMPL_DOORBELL_MASK +#define CREQ_DB_CP_FLAGS_REARM (CREQ_DB_KEY_CP | \ + CREQ_DB_IDX_VALID) +#define CREQ_DB_CP_FLAGS (CREQ_DB_KEY_CP | \ + CREQ_DB_IDX_VALID | \ + CREQ_DB_IRQ_DIS) +#define CREQ_DB_REARM(db, raw_cons, cp_bit) \ + writel(CREQ_DB_CP_FLAGS_REARM | ((raw_cons) & ((cp_bit) - 1)), db) +#define CREQ_DB(db, raw_cons, cp_bit) \ + writel(CREQ_DB_CP_FLAGS | ((raw_cons) & ((cp_bit) - 1)), db) + +#define CREQ_ENTRY_POLL_BUDGET 0x100 + +/* HWQ */ + +struct bnxt_qplib_crsq { + struct creq_qp_event *resp; + u32 req_size; +}; + +struct bnxt_qplib_rcfw_sbuf { + void *sb; + dma_addr_t dma_addr; + u32 size; +}; + +/* RCFW Communication Channels */ +struct bnxt_qplib_rcfw { + struct pci_dev *pdev; + int vector; + struct tasklet_struct worker; + bool requested; + unsigned long *cmdq_bitmap; + u32 bmap_size; + unsigned long flags; +#define FIRMWARE_INITIALIZED_FLAG 1 +#define FIRMWARE_FIRST_FLAG BIT(31) + wait_queue_head_t waitq; + int (*aeq_handler)(struct bnxt_qplib_rcfw *, + struct creq_func_event *); + u32 seq_num; + + /* Bar region info */ + void __iomem *cmdq_bar_reg_iomem; + u16 cmdq_bar_reg; + u16 cmdq_bar_reg_prod_off; + u16 cmdq_bar_reg_trig_off; + u16 creq_ring_id; + u16 creq_bar_reg; + void __iomem *creq_bar_reg_iomem; + + /* Cmd-Resp and Async Event notification queue */ + struct bnxt_qplib_hwq creq; + u64 creq_qp_event_processed; + u64 creq_func_event_processed; + + /* Actual Cmd and Resp Queues */ + struct bnxt_qplib_hwq cmdq; + struct bnxt_qplib_crsq *crsqe_tbl; +}; + +void bnxt_qplib_free_rcfw_channel(struct bnxt_qplib_rcfw *rcfw); +int bnxt_qplib_alloc_rcfw_channel(struct pci_dev *pdev, + struct bnxt_qplib_rcfw *rcfw); +void bnxt_qplib_disable_rcfw_channel(struct bnxt_qplib_rcfw *rcfw); +int bnxt_qplib_enable_rcfw_channel(struct pci_dev *pdev, + struct bnxt_qplib_rcfw *rcfw, + int msix_vector, + int cp_bar_reg_off, int virt_fn, + int (*aeq_handler) + (struct bnxt_qplib_rcfw *, + struct creq_func_event *)); + +struct bnxt_qplib_rcfw_sbuf *bnxt_qplib_rcfw_alloc_sbuf( + struct bnxt_qplib_rcfw *rcfw, + u32 size); +void bnxt_qplib_rcfw_free_sbuf(struct bnxt_qplib_rcfw *rcfw, + struct bnxt_qplib_rcfw_sbuf *sbuf); +int bnxt_qplib_rcfw_send_message(struct bnxt_qplib_rcfw *rcfw, + struct cmdq_base *req, struct creq_base *resp, + void *sbuf, u8 is_block); + +int bnxt_qplib_deinit_rcfw(struct bnxt_qplib_rcfw *rcfw); +int bnxt_qplib_init_rcfw(struct bnxt_qplib_rcfw *rcfw, + struct bnxt_qplib_ctx *ctx, int is_virtfn); +#endif /* __BNXT_QPLIB_RCFW_H__ */ diff --git a/drivers/infiniband/hw/bnxt_re/qplib_res.c b/drivers/infiniband/hw/bnxt_re/qplib_res.c new file mode 100644 index 000000000000..62447b3badec --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/qplib_res.c @@ -0,0 +1,825 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * 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 + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. 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. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: QPLib resource manager + */ + +#include <linux/spinlock.h> +#include <linux/pci.h> +#include <linux/interrupt.h> +#include <linux/inetdevice.h> +#include <linux/dma-mapping.h> +#include <linux/if_vlan.h> +#include "roce_hsi.h" +#include "qplib_res.h" +#include "qplib_sp.h" +#include "qplib_rcfw.h" + +static void bnxt_qplib_free_stats_ctx(struct pci_dev *pdev, + struct bnxt_qplib_stats *stats); +static int bnxt_qplib_alloc_stats_ctx(struct pci_dev *pdev, + struct bnxt_qplib_stats *stats); + +/* PBL */ +static void __free_pbl(struct pci_dev *pdev, struct bnxt_qplib_pbl *pbl, + bool is_umem) +{ + int i; + + if (!is_umem) { + for (i = 0; i < pbl->pg_count; i++) { + if (pbl->pg_arr[i]) + dma_free_coherent(&pdev->dev, pbl->pg_size, + (void *)((unsigned long) + pbl->pg_arr[i] & + PAGE_MASK), + pbl->pg_map_arr[i]); + else + dev_warn(&pdev->dev, + "QPLIB: PBL free pg_arr[%d] empty?!", + i); + pbl->pg_arr[i] = NULL; + } + } + kfree(pbl->pg_arr); + pbl->pg_arr = NULL; + kfree(pbl->pg_map_arr); + pbl->pg_map_arr = NULL; + pbl->pg_count = 0; + pbl->pg_size = 0; +} + +static int __alloc_pbl(struct pci_dev *pdev, struct bnxt_qplib_pbl *pbl, + struct scatterlist *sghead, u32 pages, u32 pg_size) +{ + struct scatterlist *sg; + bool is_umem = false; + int i; + + /* page ptr arrays */ + pbl->pg_arr = kcalloc(pages, sizeof(void *), GFP_KERNEL); + if (!pbl->pg_arr) + return -ENOMEM; + + pbl->pg_map_arr = kcalloc(pages, sizeof(dma_addr_t), GFP_KERNEL); + if (!pbl->pg_map_arr) { + kfree(pbl->pg_arr); + pbl->pg_arr = NULL; + return -ENOMEM; + } + pbl->pg_count = 0; + pbl->pg_size = pg_size; + + 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); + if (!pbl->pg_arr[i]) + goto fail; + memset(pbl->pg_arr[i], 0, pbl->pg_size); + pbl->pg_count++; + } + } else { + i = 0; + is_umem = true; + for_each_sg(sghead, sg, pages, i) { + pbl->pg_map_arr[i] = sg_dma_address(sg); + pbl->pg_arr[i] = sg_virt(sg); + if (!pbl->pg_arr[i]) + goto fail; + + pbl->pg_count++; + } + } + + return 0; + +fail: + __free_pbl(pdev, pbl, is_umem); + return -ENOMEM; +} + +/* HWQ */ +void bnxt_qplib_free_hwq(struct pci_dev *pdev, struct bnxt_qplib_hwq *hwq) +{ + int i; + + if (!hwq->max_elements) + return; + if (hwq->level >= PBL_LVL_MAX) + return; + + for (i = 0; i < hwq->level + 1; i++) { + if (i == hwq->level) + __free_pbl(pdev, &hwq->pbl[i], hwq->is_user); + else + __free_pbl(pdev, &hwq->pbl[i], false); + } + + hwq->level = PBL_LVL_MAX; + hwq->max_elements = 0; + hwq->element_size = 0; + hwq->prod = 0; + hwq->cons = 0; + hwq->cp_bit = 0; +} + +/* All HWQs are power of 2 in size */ +int bnxt_qplib_alloc_init_hwq(struct pci_dev *pdev, struct bnxt_qplib_hwq *hwq, + struct scatterlist *sghead, int nmap, + u32 *elements, u32 element_size, u32 aux, + u32 pg_size, enum bnxt_qplib_hwq_type hwq_type) +{ + u32 pages, slots, size, aux_pages = 0, aux_size = 0; + dma_addr_t *src_phys_ptr, **dst_virt_ptr; + int i, rc; + + hwq->level = PBL_LVL_MAX; + + slots = roundup_pow_of_two(*elements); + if (aux) { + aux_size = roundup_pow_of_two(aux); + aux_pages = (slots * aux_size) / pg_size; + if ((slots * aux_size) % pg_size) + aux_pages++; + } + size = roundup_pow_of_two(element_size); + + if (!sghead) { + hwq->is_user = false; + pages = (slots * size) / pg_size + aux_pages; + if ((slots * size) % pg_size) + pages++; + if (!pages) + return -EINVAL; + } else { + hwq->is_user = true; + pages = nmap; + } + + /* Alloc the 1st memory block; can be a PDL/PTL/PBL */ + if (sghead && (pages == MAX_PBL_LVL_0_PGS)) + rc = __alloc_pbl(pdev, &hwq->pbl[PBL_LVL_0], sghead, + pages, pg_size); + else + rc = __alloc_pbl(pdev, &hwq->pbl[PBL_LVL_0], NULL, 1, pg_size); + if (rc) + goto fail; + + hwq->level = PBL_LVL_0; + + if (pages > MAX_PBL_LVL_0_PGS) { + if (pages > MAX_PBL_LVL_1_PGS) { + /* 2 levels of indirection */ + rc = __alloc_pbl(pdev, &hwq->pbl[PBL_LVL_1], NULL, + MAX_PBL_LVL_1_PGS_FOR_LVL_2, pg_size); + if (rc) + goto fail; + /* Fill in lvl0 PBL */ + dst_virt_ptr = + (dma_addr_t **)hwq->pbl[PBL_LVL_0].pg_arr; + src_phys_ptr = hwq->pbl[PBL_LVL_1].pg_map_arr; + for (i = 0; i < hwq->pbl[PBL_LVL_1].pg_count; i++) + dst_virt_ptr[PTR_PG(i)][PTR_IDX(i)] = + src_phys_ptr[i] | PTU_PDE_VALID; + hwq->level = PBL_LVL_1; + + rc = __alloc_pbl(pdev, &hwq->pbl[PBL_LVL_2], sghead, + pages, pg_size); + if (rc) + goto fail; + + /* Fill in lvl1 PBL */ + dst_virt_ptr = + (dma_addr_t **)hwq->pbl[PBL_LVL_1].pg_arr; + src_phys_ptr = hwq->pbl[PBL_LVL_2].pg_map_arr; + for (i = 0; i < hwq->pbl[PBL_LVL_2].pg_count; i++) { + dst_virt_ptr[PTR_PG(i)][PTR_IDX(i)] = + src_phys_ptr[i] | PTU_PTE_VALID; + } + if (hwq_type == HWQ_TYPE_QUEUE) { + /* Find the last pg of the size */ + i = hwq->pbl[PBL_LVL_2].pg_count; + dst_virt_ptr[PTR_PG(i - 1)][PTR_IDX(i - 1)] |= + PTU_PTE_LAST; + if (i > 1) + dst_virt_ptr[PTR_PG(i - 2)] + [PTR_IDX(i - 2)] |= + PTU_PTE_NEXT_TO_LAST; + } + hwq->level = PBL_LVL_2; + } else { + u32 flag = hwq_type == HWQ_TYPE_L2_CMPL ? 0 : + PTU_PTE_VALID; + + /* 1 level of indirection */ + rc = __alloc_pbl(pdev, &hwq->pbl[PBL_LVL_1], sghead, + pages, pg_size); + if (rc) + goto fail; + /* Fill in lvl0 PBL */ + dst_virt_ptr = + (dma_addr_t **)hwq->pbl[PBL_LVL_0].pg_arr; + src_phys_ptr = hwq->pbl[PBL_LVL_1].pg_map_arr; + for (i = 0; i < hwq->pbl[PBL_LVL_1].pg_count; i++) { + dst_virt_ptr[PTR_PG(i)][PTR_IDX(i)] = + src_phys_ptr[i] | flag; + } + if (hwq_type == HWQ_TYPE_QUEUE) { + /* Find the last pg of the size */ + i = hwq->pbl[PBL_LVL_1].pg_count; + dst_virt_ptr[PTR_PG(i - 1)][PTR_IDX(i - 1)] |= + PTU_PTE_LAST; + if (i > 1) + dst_virt_ptr[PTR_PG(i - 2)] + [PTR_IDX(i - 2)] |= + PTU_PTE_NEXT_TO_LAST; + } + hwq->level = PBL_LVL_1; + } + } + hwq->pdev = pdev; + spin_lock_init(&hwq->lock); + hwq->prod = 0; + hwq->cons = 0; + *elements = hwq->max_elements = slots; + hwq->element_size = size; + + /* For direct access to the elements */ + hwq->pbl_ptr = hwq->pbl[hwq->level].pg_arr; + hwq->pbl_dma_ptr = hwq->pbl[hwq->level].pg_map_arr; + + return 0; + +fail: + bnxt_qplib_free_hwq(pdev, hwq); + return -ENOMEM; +} + +/* Context Tables */ +void bnxt_qplib_free_ctx(struct pci_dev *pdev, + struct bnxt_qplib_ctx *ctx) +{ + int i; + + bnxt_qplib_free_hwq(pdev, &ctx->qpc_tbl); + bnxt_qplib_free_hwq(pdev, &ctx->mrw_tbl); + bnxt_qplib_free_hwq(pdev, &ctx->srqc_tbl); + bnxt_qplib_free_hwq(pdev, &ctx->cq_tbl); + bnxt_qplib_free_hwq(pdev, &ctx->tim_tbl); + for (i = 0; i < MAX_TQM_ALLOC_REQ; i++) + bnxt_qplib_free_hwq(pdev, &ctx->tqm_tbl[i]); + bnxt_qplib_free_hwq(pdev, &ctx->tqm_pde); + bnxt_qplib_free_stats_ctx(pdev, &ctx->stats); +} + +/* + * Routine: bnxt_qplib_alloc_ctx + * Description: + * Context tables are memories which are used by the chip fw. + * The 6 tables defined are: + * QPC ctx - holds QP states + * MRW ctx - holds memory region and window + * SRQ ctx - holds shared RQ states + * CQ ctx - holds completion queue states + * TQM ctx - holds Tx Queue Manager context + * TIM ctx - holds timer context + * Depending on the size of the tbl requested, either a 1 Page Buffer List + * or a 1-to-2-stage indirection Page Directory List + 1 PBL is used + * instead. + * Table might be employed as follows: + * For 0 < ctx size <= 1 PAGE, 0 level of ind is used + * For 1 PAGE < ctx size <= 512 entries size, 1 level of ind is used + * For 512 < ctx size <= MAX, 2 levels of ind is used + * Returns: + * 0 if success, else -ERRORS + */ +int bnxt_qplib_alloc_ctx(struct pci_dev *pdev, + struct bnxt_qplib_ctx *ctx, + bool virt_fn) +{ + int i, j, k, rc = 0; + int fnz_idx = -1; + __le64 **pbl_ptr; + + if (virt_fn) + goto stats_alloc; + + /* QPC Tables */ + ctx->qpc_tbl.max_elements = ctx->qpc_count; + rc = bnxt_qplib_alloc_init_hwq(pdev, &ctx->qpc_tbl, NULL, 0, + &ctx->qpc_tbl.max_elements, + BNXT_QPLIB_MAX_QP_CTX_ENTRY_SIZE, 0, + PAGE_SIZE, HWQ_TYPE_CTX); + if (rc) + goto fail; + + /* MRW Tables */ + ctx->mrw_tbl.max_elements = ctx->mrw_count; + rc = bnxt_qplib_alloc_init_hwq(pdev, &ctx->mrw_tbl, NULL, 0, + &ctx->mrw_tbl.max_elements, + BNXT_QPLIB_MAX_MRW_CTX_ENTRY_SIZE, 0, + PAGE_SIZE, HWQ_TYPE_CTX); + if (rc) + goto fail; + + /* SRQ Tables */ + ctx->srqc_tbl.max_elements = ctx->srqc_count; + rc = bnxt_qplib_alloc_init_hwq(pdev, &ctx->srqc_tbl, NULL, 0, + &ctx->srqc_tbl.max_elements, + BNXT_QPLIB_MAX_SRQ_CTX_ENTRY_SIZE, 0, + PAGE_SIZE, HWQ_TYPE_CTX); + if (rc) + goto fail; + + /* CQ Tables */ + ctx->cq_tbl.max_elements = ctx->cq_count; + rc = bnxt_qplib_alloc_init_hwq(pdev, &ctx->cq_tbl, NULL, 0, + &ctx->cq_tbl.max_elements, + BNXT_QPLIB_MAX_CQ_CTX_ENTRY_SIZE, 0, + PAGE_SIZE, HWQ_TYPE_CTX); + if (rc) + goto fail; + + /* TQM Buffer */ + ctx->tqm_pde.max_elements = 512; + rc = bnxt_qplib_alloc_init_hwq(pdev, &ctx->tqm_pde, NULL, 0, + &ctx->tqm_pde.max_elements, sizeof(u64), + 0, PAGE_SIZE, HWQ_TYPE_CTX); + if (rc) + goto fail; + + for (i = 0; i < MAX_TQM_ALLOC_REQ; i++) { + if (!ctx->tqm_count[i]) + continue; + ctx->tqm_tbl[i].max_elements = ctx->qpc_count * + ctx->tqm_count[i]; + rc = bnxt_qplib_alloc_init_hwq(pdev, &ctx->tqm_tbl[i], NULL, 0, + &ctx->tqm_tbl[i].max_elements, 1, + 0, PAGE_SIZE, HWQ_TYPE_CTX); + if (rc) + goto fail; + } + pbl_ptr = (__le64 **)ctx->tqm_pde.pbl_ptr; + for (i = 0, j = 0; i < MAX_TQM_ALLOC_REQ; + i++, j += MAX_TQM_ALLOC_BLK_SIZE) { + if (!ctx->tqm_tbl[i].max_elements) + continue; + if (fnz_idx == -1) + fnz_idx = i; + switch (ctx->tqm_tbl[i].level) { + case PBL_LVL_2: + for (k = 0; k < ctx->tqm_tbl[i].pbl[PBL_LVL_1].pg_count; + k++) + pbl_ptr[PTR_PG(j + k)][PTR_IDX(j + k)] = + cpu_to_le64( + ctx->tqm_tbl[i].pbl[PBL_LVL_1].pg_map_arr[k] + | PTU_PTE_VALID); + break; + case PBL_LVL_1: + case PBL_LVL_0: + default: + pbl_ptr[PTR_PG(j)][PTR_IDX(j)] = cpu_to_le64( + ctx->tqm_tbl[i].pbl[PBL_LVL_0].pg_map_arr[0] | + PTU_PTE_VALID); + break; + } + } + if (fnz_idx == -1) + fnz_idx = 0; + ctx->tqm_pde_level = ctx->tqm_tbl[fnz_idx].level == PBL_LVL_2 ? + PBL_LVL_2 : ctx->tqm_tbl[fnz_idx].level + 1; + + /* TIM Buffer */ + ctx->tim_tbl.max_elements = ctx->qpc_count * 16; + rc = bnxt_qplib_alloc_init_hwq(pdev, &ctx->tim_tbl, NULL, 0, + &ctx->tim_tbl.max_elements, 1, + 0, PAGE_SIZE, HWQ_TYPE_CTX); + if (rc) + goto fail; + +stats_alloc: + /* Stats */ + rc = bnxt_qplib_alloc_stats_ctx(pdev, &ctx->stats); + if (rc) + goto fail; + + return 0; + +fail: + bnxt_qplib_free_ctx(pdev, ctx); + return rc; +} + +/* GUID */ +void bnxt_qplib_get_guid(u8 *dev_addr, u8 *guid) +{ + u8 mac[ETH_ALEN]; + + /* MAC-48 to EUI-64 mapping */ + memcpy(mac, dev_addr, ETH_ALEN); + guid[0] = mac[0] ^ 2; + guid[1] = mac[1]; + guid[2] = mac[2]; + guid[3] = 0xff; + guid[4] = 0xfe; + guid[5] = mac[3]; + guid[6] = mac[4]; + guid[7] = mac[5]; +} + +static void bnxt_qplib_free_sgid_tbl(struct bnxt_qplib_res *res, + struct bnxt_qplib_sgid_tbl *sgid_tbl) +{ + kfree(sgid_tbl->tbl); + kfree(sgid_tbl->hw_id); + kfree(sgid_tbl->ctx); + sgid_tbl->tbl = NULL; + sgid_tbl->hw_id = NULL; + sgid_tbl->ctx = NULL; + sgid_tbl->max = 0; + sgid_tbl->active = 0; +} + +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); + if (!sgid_tbl->tbl) + return -ENOMEM; + + sgid_tbl->hw_id = kcalloc(max, sizeof(u16), GFP_KERNEL); + if (!sgid_tbl->hw_id) + goto out_free1; + + sgid_tbl->ctx = kcalloc(max, sizeof(void *), GFP_KERNEL); + if (!sgid_tbl->ctx) + goto out_free2; + + sgid_tbl->max = max; + return 0; +out_free2: + kfree(sgid_tbl->hw_id); + sgid_tbl->hw_id = NULL; +out_free1: + kfree(sgid_tbl->tbl); + sgid_tbl->tbl = NULL; + return -ENOMEM; +}; + +static void bnxt_qplib_cleanup_sgid_tbl(struct bnxt_qplib_res *res, + struct bnxt_qplib_sgid_tbl *sgid_tbl) +{ + int i; + + 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); + } + memset(sgid_tbl->tbl, 0, sizeof(struct bnxt_qplib_gid) * sgid_tbl->max); + memset(sgid_tbl->hw_id, -1, sizeof(u16) * sgid_tbl->max); + sgid_tbl->active = 0; +} + +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); + memset(sgid_tbl->hw_id, -1, sizeof(u16) * sgid_tbl->max); +} + +static void bnxt_qplib_free_pkey_tbl(struct bnxt_qplib_res *res, + struct bnxt_qplib_pkey_tbl *pkey_tbl) +{ + if (!pkey_tbl->tbl) + dev_dbg(&res->pdev->dev, "QPLIB: PKEY tbl not present"); + else + kfree(pkey_tbl->tbl); + + pkey_tbl->tbl = NULL; + pkey_tbl->max = 0; + pkey_tbl->active = 0; +} + +static int bnxt_qplib_alloc_pkey_tbl(struct bnxt_qplib_res *res, + struct bnxt_qplib_pkey_tbl *pkey_tbl, + u16 max) +{ + pkey_tbl->tbl = kcalloc(max, sizeof(u16), GFP_KERNEL); + if (!pkey_tbl->tbl) + return -ENOMEM; + + pkey_tbl->max = max; + return 0; +}; + +/* PDs */ +int bnxt_qplib_alloc_pd(struct bnxt_qplib_pd_tbl *pdt, struct bnxt_qplib_pd *pd) +{ + u32 bit_num; + + bit_num = find_first_bit(pdt->tbl, pdt->max); + if (bit_num == pdt->max) + return -ENOMEM; + + /* Found unused PD */ + clear_bit(bit_num, pdt->tbl); + pd->id = bit_num; + return 0; +} + +int bnxt_qplib_dealloc_pd(struct bnxt_qplib_res *res, + struct bnxt_qplib_pd_tbl *pdt, + struct bnxt_qplib_pd *pd) +{ + if (test_and_set_bit(pd->id, pdt->tbl)) { + dev_warn(&res->pdev->dev, "Freeing an unused PD? pdn = %d", + pd->id); + return -EINVAL; + } + pd->id = 0; + return 0; +} + +static void bnxt_qplib_free_pd_tbl(struct bnxt_qplib_pd_tbl *pdt) +{ + kfree(pdt->tbl); + pdt->tbl = NULL; + pdt->max = 0; +} + +static int bnxt_qplib_alloc_pd_tbl(struct bnxt_qplib_res *res, + struct bnxt_qplib_pd_tbl *pdt, + u32 max) +{ + u32 bytes; + + bytes = max >> 3; + if (!bytes) + bytes = 1; + pdt->tbl = kmalloc(bytes, GFP_KERNEL); + if (!pdt->tbl) + return -ENOMEM; + + pdt->max = max; + memset((u8 *)pdt->tbl, 0xFF, bytes); + + return 0; +} + +/* DPIs */ +int bnxt_qplib_alloc_dpi(struct bnxt_qplib_dpi_tbl *dpit, + struct bnxt_qplib_dpi *dpi, + void *app) +{ + u32 bit_num; + + bit_num = find_first_bit(dpit->tbl, dpit->max); + if (bit_num == dpit->max) + return -ENOMEM; + + /* Found unused DPI */ + clear_bit(bit_num, dpit->tbl); + dpit->app_tbl[bit_num] = app; + + dpi->dpi = bit_num; + dpi->dbr = dpit->dbr_bar_reg_iomem + (bit_num * PAGE_SIZE); + dpi->umdbr = dpit->unmapped_dbr + (bit_num * PAGE_SIZE); + + return 0; +} + +int bnxt_qplib_dealloc_dpi(struct bnxt_qplib_res *res, + struct bnxt_qplib_dpi_tbl *dpit, + struct bnxt_qplib_dpi *dpi) +{ + if (dpi->dpi >= dpit->max) { + dev_warn(&res->pdev->dev, "Invalid DPI? dpi = %d", dpi->dpi); + return -EINVAL; + } + if (test_and_set_bit(dpi->dpi, dpit->tbl)) { + dev_warn(&res->pdev->dev, "Freeing an unused DPI? dpi = %d", + dpi->dpi); + return -EINVAL; + } + if (dpit->app_tbl) + dpit->app_tbl[dpi->dpi] = NULL; + memset(dpi, 0, sizeof(*dpi)); + + return 0; +} + +static void bnxt_qplib_free_dpi_tbl(struct bnxt_qplib_res *res, + struct bnxt_qplib_dpi_tbl *dpit) +{ + kfree(dpit->tbl); + kfree(dpit->app_tbl); + if (dpit->dbr_bar_reg_iomem) + pci_iounmap(res->pdev, dpit->dbr_bar_reg_iomem); + memset(dpit, 0, sizeof(*dpit)); +} + +static int bnxt_qplib_alloc_dpi_tbl(struct bnxt_qplib_res *res, + struct bnxt_qplib_dpi_tbl *dpit, + u32 dbr_offset) +{ + u32 dbr_bar_reg = RCFW_DBR_PCI_BAR_REGION; + resource_size_t bar_reg_base; + u32 dbr_len, bytes; + + if (dpit->dbr_bar_reg_iomem) { + dev_err(&res->pdev->dev, + "QPLIB: DBR BAR region %d already mapped", dbr_bar_reg); + return -EALREADY; + } + + bar_reg_base = pci_resource_start(res->pdev, dbr_bar_reg); + if (!bar_reg_base) { + dev_err(&res->pdev->dev, + "QPLIB: BAR region %d resc start failed", dbr_bar_reg); + return -ENOMEM; + } + + dbr_len = pci_resource_len(res->pdev, dbr_bar_reg) - dbr_offset; + if (!dbr_len || ((dbr_len & (PAGE_SIZE - 1)) != 0)) { + dev_err(&res->pdev->dev, "QPLIB: Invalid DBR length %d", + dbr_len); + return -ENOMEM; + } + + dpit->dbr_bar_reg_iomem = ioremap_nocache(bar_reg_base + dbr_offset, + dbr_len); + if (!dpit->dbr_bar_reg_iomem) { + dev_err(&res->pdev->dev, + "QPLIB: FP: DBR BAR region %d mapping failed", + dbr_bar_reg); + return -ENOMEM; + } + + dpit->unmapped_dbr = bar_reg_base + dbr_offset; + dpit->max = dbr_len / PAGE_SIZE; + + dpit->app_tbl = kcalloc(dpit->max, sizeof(void *), GFP_KERNEL); + if (!dpit->app_tbl) { + pci_iounmap(res->pdev, dpit->dbr_bar_reg_iomem); + dev_err(&res->pdev->dev, + "QPLIB: DPI app tbl allocation failed"); + return -ENOMEM; + } + + bytes = dpit->max >> 3; + if (!bytes) + bytes = 1; + + dpit->tbl = kmalloc(bytes, GFP_KERNEL); + if (!dpit->tbl) { + pci_iounmap(res->pdev, dpit->dbr_bar_reg_iomem); + kfree(dpit->app_tbl); + dpit->app_tbl = NULL; + dev_err(&res->pdev->dev, + "QPLIB: DPI tbl allocation failed for size = %d", + bytes); + return -ENOMEM; + } + + memset((u8 *)dpit->tbl, 0xFF, bytes); + + return 0; +} + +/* PKEYs */ +static void bnxt_qplib_cleanup_pkey_tbl(struct bnxt_qplib_pkey_tbl *pkey_tbl) +{ + memset(pkey_tbl->tbl, 0, sizeof(u16) * pkey_tbl->max); + pkey_tbl->active = 0; +} + +static void bnxt_qplib_init_pkey_tbl(struct bnxt_qplib_res *res, + struct bnxt_qplib_pkey_tbl *pkey_tbl) +{ + u16 pkey = 0xFFFF; + + memset(pkey_tbl->tbl, 0, sizeof(u16) * pkey_tbl->max); + + /* pkey default = 0xFFFF */ + bnxt_qplib_add_pkey(res, pkey_tbl, &pkey, false); +} + +/* Stats */ +static void bnxt_qplib_free_stats_ctx(struct pci_dev *pdev, + struct bnxt_qplib_stats *stats) +{ + if (stats->dma) { + dma_free_coherent(&pdev->dev, stats->size, + stats->dma, stats->dma_map); + } + memset(stats, 0, sizeof(*stats)); + stats->fw_id = -1; +} + +static int bnxt_qplib_alloc_stats_ctx(struct pci_dev *pdev, + struct bnxt_qplib_stats *stats) +{ + memset(stats, 0, sizeof(*stats)); + stats->fw_id = -1; + stats->size = sizeof(struct ctx_hw_stats); + stats->dma = dma_alloc_coherent(&pdev->dev, stats->size, + &stats->dma_map, GFP_KERNEL); + if (!stats->dma) { + dev_err(&pdev->dev, "QPLIB: Stats DMA allocation failed"); + return -ENOMEM; + } + return 0; +} + +void bnxt_qplib_cleanup_res(struct bnxt_qplib_res *res) +{ + bnxt_qplib_cleanup_pkey_tbl(&res->pkey_tbl); + bnxt_qplib_cleanup_sgid_tbl(res, &res->sgid_tbl); +} + +int bnxt_qplib_init_res(struct bnxt_qplib_res *res) +{ + bnxt_qplib_init_sgid_tbl(&res->sgid_tbl, res->netdev); + bnxt_qplib_init_pkey_tbl(res, &res->pkey_tbl); + + return 0; +} + +void bnxt_qplib_free_res(struct bnxt_qplib_res *res) +{ + bnxt_qplib_free_pkey_tbl(res, &res->pkey_tbl); + bnxt_qplib_free_sgid_tbl(res, &res->sgid_tbl); + bnxt_qplib_free_pd_tbl(&res->pd_tbl); + bnxt_qplib_free_dpi_tbl(res, &res->dpi_tbl); + + res->netdev = NULL; + res->pdev = NULL; +} + +int bnxt_qplib_alloc_res(struct bnxt_qplib_res *res, struct pci_dev *pdev, + struct net_device *netdev, + struct bnxt_qplib_dev_attr *dev_attr) +{ + int rc = 0; + + res->pdev = pdev; + res->netdev = netdev; + + rc = bnxt_qplib_alloc_sgid_tbl(res, &res->sgid_tbl, dev_attr->max_sgid); + if (rc) + goto fail; + + rc = bnxt_qplib_alloc_pkey_tbl(res, &res->pkey_tbl, dev_attr->max_pkey); + if (rc) + goto fail; + + rc = bnxt_qplib_alloc_pd_tbl(res, &res->pd_tbl, dev_attr->max_pd); + if (rc) + goto fail; + + rc = bnxt_qplib_alloc_dpi_tbl(res, &res->dpi_tbl, dev_attr->l2_db_size); + if (rc) + goto fail; + + return 0; +fail: + bnxt_qplib_free_res(res); + return rc; +} diff --git a/drivers/infiniband/hw/bnxt_re/qplib_res.h b/drivers/infiniband/hw/bnxt_re/qplib_res.h new file mode 100644 index 000000000000..2e4855509719 --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/qplib_res.h @@ -0,0 +1,227 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * 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 + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. 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. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: QPLib resource manager (header) + */ + +#ifndef __BNXT_QPLIB_RES_H__ +#define __BNXT_QPLIB_RES_H__ + +extern const struct bnxt_qplib_gid bnxt_qplib_gid_zero; + +#define PTR_CNT_PER_PG (PAGE_SIZE / sizeof(void *)) +#define PTR_MAX_IDX_PER_PG (PTR_CNT_PER_PG - 1) +#define PTR_PG(x) (((x) & ~PTR_MAX_IDX_PER_PG) / PTR_CNT_PER_PG) +#define PTR_IDX(x) ((x) & PTR_MAX_IDX_PER_PG) + +#define HWQ_CMP(idx, hwq) ((idx) & ((hwq)->max_elements - 1)) + +#define HWQ_FREE_SLOTS(hwq) (hwq->max_elements - \ + ((HWQ_CMP(hwq->prod, hwq)\ + - HWQ_CMP(hwq->cons, hwq))\ + & (hwq->max_elements - 1))) +enum bnxt_qplib_hwq_type { + HWQ_TYPE_CTX, + HWQ_TYPE_QUEUE, + HWQ_TYPE_L2_CMPL +}; + +#define MAX_PBL_LVL_0_PGS 1 +#define MAX_PBL_LVL_1_PGS 512 +#define MAX_PBL_LVL_1_PGS_SHIFT 9 +#define MAX_PBL_LVL_1_PGS_FOR_LVL_2 256 +#define MAX_PBL_LVL_2_PGS (256 * 512) + +enum bnxt_qplib_pbl_lvl { + PBL_LVL_0, + PBL_LVL_1, + PBL_LVL_2, + PBL_LVL_MAX +}; + +#define ROCE_PG_SIZE_4K (4 * 1024) +#define ROCE_PG_SIZE_8K (8 * 1024) +#define ROCE_PG_SIZE_64K (64 * 1024) +#define ROCE_PG_SIZE_2M (2 * 1024 * 1024) +#define ROCE_PG_SIZE_8M (8 * 1024 * 1024) +#define ROCE_PG_SIZE_1G (1024 * 1024 * 1024) + +struct bnxt_qplib_pbl { + u32 pg_count; + u32 pg_size; + void **pg_arr; + dma_addr_t *pg_map_arr; +}; + +struct bnxt_qplib_hwq { + struct pci_dev *pdev; + /* lock to protect qplib_hwq */ + spinlock_t lock; + struct bnxt_qplib_pbl pbl[PBL_LVL_MAX]; + enum bnxt_qplib_pbl_lvl level; /* 0, 1, or 2 */ + /* ptr for easy access to the PBL entries */ + void **pbl_ptr; + /* ptr for easy access to the dma_addr */ + dma_addr_t *pbl_dma_ptr; + u32 max_elements; + u16 element_size; /* Size of each entry */ + + u32 prod; /* raw */ + u32 cons; /* raw */ + u8 cp_bit; + u8 is_user; +}; + +/* Tables */ +struct bnxt_qplib_pd_tbl { + unsigned long *tbl; + u32 max; +}; + +struct bnxt_qplib_sgid_tbl { + struct bnxt_qplib_gid *tbl; + u16 *hw_id; + u16 max; + u16 active; + void *ctx; +}; + +struct bnxt_qplib_pkey_tbl { + u16 *tbl; + u16 max; + u16 active; +}; + +struct bnxt_qplib_dpi { + u32 dpi; + void __iomem *dbr; + u64 umdbr; +}; + +struct bnxt_qplib_dpi_tbl { + void **app_tbl; + unsigned long *tbl; + u16 max; + void __iomem *dbr_bar_reg_iomem; + u64 unmapped_dbr; +}; + +struct bnxt_qplib_stats { + dma_addr_t dma_map; + void *dma; + u32 size; + u32 fw_id; +}; + +struct bnxt_qplib_vf_res { + u32 max_qp_per_vf; + u32 max_mrw_per_vf; + u32 max_srq_per_vf; + u32 max_cq_per_vf; + u32 max_gid_per_vf; +}; + +#define BNXT_QPLIB_MAX_QP_CTX_ENTRY_SIZE 448 +#define BNXT_QPLIB_MAX_SRQ_CTX_ENTRY_SIZE 64 +#define BNXT_QPLIB_MAX_CQ_CTX_ENTRY_SIZE 64 +#define BNXT_QPLIB_MAX_MRW_CTX_ENTRY_SIZE 128 + +struct bnxt_qplib_ctx { + u32 qpc_count; + struct bnxt_qplib_hwq qpc_tbl; + u32 mrw_count; + struct bnxt_qplib_hwq mrw_tbl; + u32 srqc_count; + struct bnxt_qplib_hwq srqc_tbl; + u32 cq_count; + struct bnxt_qplib_hwq cq_tbl; + struct bnxt_qplib_hwq tim_tbl; +#define MAX_TQM_ALLOC_REQ 32 +#define MAX_TQM_ALLOC_BLK_SIZE 8 + u8 tqm_count[MAX_TQM_ALLOC_REQ]; + struct bnxt_qplib_hwq tqm_pde; + u32 tqm_pde_level; + struct bnxt_qplib_hwq tqm_tbl[MAX_TQM_ALLOC_REQ]; + struct bnxt_qplib_stats stats; + struct bnxt_qplib_vf_res vf_res; +}; + +struct bnxt_qplib_res { + struct pci_dev *pdev; + struct net_device *netdev; + + struct bnxt_qplib_rcfw *rcfw; + + struct bnxt_qplib_pd_tbl pd_tbl; + struct bnxt_qplib_sgid_tbl sgid_tbl; + struct bnxt_qplib_pkey_tbl pkey_tbl; + struct bnxt_qplib_dpi_tbl dpi_tbl; +}; + +#define to_bnxt_qplib(ptr, type, member) \ + container_of(ptr, type, member) + +struct bnxt_qplib_pd; +struct bnxt_qplib_dev_attr; + +void bnxt_qplib_free_hwq(struct pci_dev *pdev, struct bnxt_qplib_hwq *hwq); +int bnxt_qplib_alloc_init_hwq(struct pci_dev *pdev, struct bnxt_qplib_hwq *hwq, + struct scatterlist *sl, int nmap, u32 *elements, + u32 elements_per_page, u32 aux, u32 pg_size, + enum bnxt_qplib_hwq_type hwq_type); +void bnxt_qplib_get_guid(u8 *dev_addr, u8 *guid); +int bnxt_qplib_alloc_pd(struct bnxt_qplib_pd_tbl *pd_tbl, + struct bnxt_qplib_pd *pd); +int bnxt_qplib_dealloc_pd(struct bnxt_qplib_res *res, + struct bnxt_qplib_pd_tbl *pd_tbl, + struct bnxt_qplib_pd *pd); +int bnxt_qplib_alloc_dpi(struct bnxt_qplib_dpi_tbl *dpit, + struct bnxt_qplib_dpi *dpi, + void *app); +int bnxt_qplib_dealloc_dpi(struct bnxt_qplib_res *res, + struct bnxt_qplib_dpi_tbl *dpi_tbl, + struct bnxt_qplib_dpi *dpi); +void bnxt_qplib_cleanup_res(struct bnxt_qplib_res *res); +int bnxt_qplib_init_res(struct bnxt_qplib_res *res); +void bnxt_qplib_free_res(struct bnxt_qplib_res *res); +int bnxt_qplib_alloc_res(struct bnxt_qplib_res *res, struct pci_dev *pdev, + struct net_device *netdev, + struct bnxt_qplib_dev_attr *dev_attr); +void bnxt_qplib_free_ctx(struct pci_dev *pdev, + struct bnxt_qplib_ctx *ctx); +int bnxt_qplib_alloc_ctx(struct pci_dev *pdev, + struct bnxt_qplib_ctx *ctx, + bool virt_fn); +#endif /* __BNXT_QPLIB_RES_H__ */ diff --git a/drivers/infiniband/hw/bnxt_re/qplib_sp.c b/drivers/infiniband/hw/bnxt_re/qplib_sp.c new file mode 100644 index 000000000000..ef91ab786dd4 --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/qplib_sp.c @@ -0,0 +1,687 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * 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 + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. 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. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: Slow Path Operators + */ + +#include <linux/interrupt.h> +#include <linux/spinlock.h> +#include <linux/sched.h> +#include <linux/pci.h> + +#include "roce_hsi.h" + +#include "qplib_res.h" +#include "qplib_rcfw.h" +#include "qplib_sp.h" + +const struct bnxt_qplib_gid bnxt_qplib_gid_zero = {{ 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0 } }; + +/* Device */ + +static bool bnxt_qplib_is_atomic_cap(struct bnxt_qplib_rcfw *rcfw) +{ + int rc; + u16 pcie_ctl2; + + rc = pcie_capability_read_word(rcfw->pdev, PCI_EXP_DEVCTL2, + &pcie_ctl2); + if (rc) + return false; + return !!(pcie_ctl2 & PCI_EXP_DEVCTL2_ATOMIC_REQ); +} + +int bnxt_qplib_get_dev_attr(struct bnxt_qplib_rcfw *rcfw, + struct bnxt_qplib_dev_attr *attr) +{ + struct cmdq_query_func req; + struct creq_query_func_resp resp; + struct bnxt_qplib_rcfw_sbuf *sbuf; + struct creq_query_func_resp_sb *sb; + u16 cmd_flags = 0; + u32 temp; + u8 *tqm_alloc; + int i, rc = 0; + + RCFW_CMD_PREP(req, QUERY_FUNC, cmd_flags); + + sbuf = bnxt_qplib_rcfw_alloc_sbuf(rcfw, sizeof(*sb)); + if (!sbuf) { + dev_err(&rcfw->pdev->dev, + "QPLIB: SP: QUERY_FUNC alloc side buffer failed"); + return -ENOMEM; + } + + sb = sbuf->sb; + req.resp_size = sizeof(*sb) / BNXT_QPLIB_CMDQE_UNITS; + rc = bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, (void *)&resp, + (void *)sbuf, 0); + if (rc) + goto bail; + + /* Extract the context from the side buffer */ + attr->max_qp = le32_to_cpu(sb->max_qp); + /* max_qp value reported by FW for PF doesn't include the QP1 for PF */ + attr->max_qp += 1; + attr->max_qp_rd_atom = + sb->max_qp_rd_atom > BNXT_QPLIB_MAX_OUT_RD_ATOM ? + BNXT_QPLIB_MAX_OUT_RD_ATOM : sb->max_qp_rd_atom; + attr->max_qp_init_rd_atom = + sb->max_qp_init_rd_atom > BNXT_QPLIB_MAX_OUT_RD_ATOM ? + BNXT_QPLIB_MAX_OUT_RD_ATOM : sb->max_qp_init_rd_atom; + attr->max_qp_wqes = le16_to_cpu(sb->max_qp_wr); + /* + * 128 WQEs needs to be reserved for the HW (8916). Prevent + * reporting the max number + */ + attr->max_qp_wqes -= BNXT_QPLIB_RESERVED_QP_WRS; + attr->max_qp_sges = sb->max_sge; + attr->max_cq = le32_to_cpu(sb->max_cq); + attr->max_cq_wqes = le32_to_cpu(sb->max_cqe); + attr->max_cq_sges = attr->max_qp_sges; + attr->max_mr = le32_to_cpu(sb->max_mr); + attr->max_mw = le32_to_cpu(sb->max_mw); + + attr->max_mr_size = le64_to_cpu(sb->max_mr_size); + attr->max_pd = 64 * 1024; + attr->max_raw_ethy_qp = le32_to_cpu(sb->max_raw_eth_qp); + attr->max_ah = le32_to_cpu(sb->max_ah); + + attr->max_fmr = le32_to_cpu(sb->max_fmr); + attr->max_map_per_fmr = sb->max_map_per_fmr; + + attr->max_srq = le16_to_cpu(sb->max_srq); + attr->max_srq_wqes = le32_to_cpu(sb->max_srq_wr) - 1; + attr->max_srq_sges = sb->max_srq_sge; + /* Bono only reports 1 PKEY for now, but it can support > 1 */ + attr->max_pkey = le32_to_cpu(sb->max_pkeys); + + attr->max_inline_data = le32_to_cpu(sb->max_inline_data); + attr->l2_db_size = (sb->l2_db_space_size + 1) * PAGE_SIZE; + attr->max_sgid = le32_to_cpu(sb->max_gid); + + strlcpy(attr->fw_ver, "20.6.28.0", sizeof(attr->fw_ver)); + + for (i = 0; i < MAX_TQM_ALLOC_REQ / 4; i++) { + temp = le32_to_cpu(sb->tqm_alloc_reqs[i]); + tqm_alloc = (u8 *)&temp; + attr->tqm_alloc_reqs[i * 4] = *tqm_alloc; + attr->tqm_alloc_reqs[i * 4 + 1] = *(++tqm_alloc); + attr->tqm_alloc_reqs[i * 4 + 2] = *(++tqm_alloc); + attr->tqm_alloc_reqs[i * 4 + 3] = *(++tqm_alloc); + } + + attr->is_atomic = bnxt_qplib_is_atomic_cap(rcfw); +bail: + bnxt_qplib_rcfw_free_sbuf(rcfw, sbuf); + return rc; +} + +/* SGID */ +int bnxt_qplib_get_sgid(struct bnxt_qplib_res *res, + struct bnxt_qplib_sgid_tbl *sgid_tbl, int index, + struct bnxt_qplib_gid *gid) +{ + if (index > sgid_tbl->max) { + dev_err(&res->pdev->dev, + "QPLIB: Index %d exceeded SGID table max (%d)", + index, sgid_tbl->max); + return -EINVAL; + } + memcpy(gid, &sgid_tbl->tbl[index], 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_res *res = to_bnxt_qplib(sgid_tbl, + struct bnxt_qplib_res, + sgid_tbl); + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + int index; + + if (!sgid_tbl) { + dev_err(&res->pdev->dev, "QPLIB: SGID table not allocated"); + return -EINVAL; + } + /* Do we need a sgid_lock here? */ + if (!sgid_tbl->active) { + dev_err(&res->pdev->dev, + "QPLIB: SGID table has no active entries"); + return -ENOMEM; + } + for (index = 0; index < sgid_tbl->max; index++) { + if (!memcmp(&sgid_tbl->tbl[index], gid, sizeof(*gid))) + break; + } + if (index == sgid_tbl->max) { + dev_warn(&res->pdev->dev, "GID not found in the SGID table"); + return 0; + } + /* Remove GID from the SGID table */ + if (update) { + struct cmdq_delete_gid req; + struct creq_delete_gid_resp resp; + u16 cmd_flags = 0; + int rc; + + RCFW_CMD_PREP(req, DELETE_GID, cmd_flags); + if (sgid_tbl->hw_id[index] == 0xFFFF) { + dev_err(&res->pdev->dev, + "QPLIB: GID entry contains an invalid HW id"); + return -EINVAL; + } + req.gid_index = cpu_to_le16(sgid_tbl->hw_id[index]); + rc = bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + (void *)&resp, NULL, 0); + if (rc) + return rc; + } + memcpy(&sgid_tbl->tbl[index], &bnxt_qplib_gid_zero, + sizeof(bnxt_qplib_gid_zero)); + sgid_tbl->active--; + dev_dbg(&res->pdev->dev, + "QPLIB: SGID deleted hw_id[0x%x] = 0x%x active = 0x%x", + index, sgid_tbl->hw_id[index], sgid_tbl->active); + sgid_tbl->hw_id[index] = (u16)-1; + + /* unlock */ + return 0; +} + +int bnxt_qplib_add_sgid(struct bnxt_qplib_sgid_tbl *sgid_tbl, + struct bnxt_qplib_gid *gid, u8 *smac, u16 vlan_id, + bool update, u32 *index) +{ + struct bnxt_qplib_res *res = to_bnxt_qplib(sgid_tbl, + struct bnxt_qplib_res, + sgid_tbl); + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + int i, free_idx; + + if (!sgid_tbl) { + dev_err(&res->pdev->dev, "QPLIB: SGID table not allocated"); + return -EINVAL; + } + /* Do we need a sgid_lock here? */ + if (sgid_tbl->active == sgid_tbl->max) { + dev_err(&res->pdev->dev, "QPLIB: SGID table is full"); + return -ENOMEM; + } + free_idx = sgid_tbl->max; + for (i = 0; i < sgid_tbl->max; i++) { + if (!memcmp(&sgid_tbl->tbl[i], gid, sizeof(*gid))) { + dev_dbg(&res->pdev->dev, + "QPLIB: SGID entry already exist in entry %d!", + i); + *index = i; + return -EALREADY; + } else if (!memcmp(&sgid_tbl->tbl[i], &bnxt_qplib_gid_zero, + sizeof(bnxt_qplib_gid_zero)) && + free_idx == sgid_tbl->max) { + free_idx = i; + } + } + if (free_idx == sgid_tbl->max) { + dev_err(&res->pdev->dev, + "QPLIB: SGID table is FULL but count is not MAX??"); + return -ENOMEM; + } + if (update) { + struct cmdq_add_gid req; + struct creq_add_gid_resp resp; + u16 cmd_flags = 0; + u32 temp32[4]; + u16 temp16[3]; + int rc; + + RCFW_CMD_PREP(req, ADD_GID, cmd_flags); + + memcpy(temp32, gid->data, sizeof(struct bnxt_qplib_gid)); + req.gid[0] = cpu_to_be32(temp32[3]); + req.gid[1] = cpu_to_be32(temp32[2]); + req.gid[2] = cpu_to_be32(temp32[1]); + req.gid[3] = cpu_to_be32(temp32[0]); + if (vlan_id != 0xFFFF) + req.vlan = cpu_to_le16((vlan_id & + CMDQ_ADD_GID_VLAN_VLAN_ID_MASK) | + CMDQ_ADD_GID_VLAN_TPID_TPID_8100 | + CMDQ_ADD_GID_VLAN_VLAN_EN); + + /* MAC in network format */ + memcpy(temp16, smac, 6); + req.src_mac[0] = cpu_to_be16(temp16[0]); + req.src_mac[1] = cpu_to_be16(temp16[1]); + req.src_mac[2] = cpu_to_be16(temp16[2]); + + rc = bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + (void *)&resp, NULL, 0); + if (rc) + return rc; + sgid_tbl->hw_id[free_idx] = le32_to_cpu(resp.xid); + } + /* Add GID to the sgid_tbl */ + memcpy(&sgid_tbl->tbl[free_idx], gid, sizeof(*gid)); + sgid_tbl->active++; + dev_dbg(&res->pdev->dev, + "QPLIB: SGID added hw_id[0x%x] = 0x%x active = 0x%x", + free_idx, sgid_tbl->hw_id[free_idx], sgid_tbl->active); + + *index = free_idx; + /* unlock */ + return 0; +} + +/* pkeys */ +int bnxt_qplib_get_pkey(struct bnxt_qplib_res *res, + struct bnxt_qplib_pkey_tbl *pkey_tbl, u16 index, + u16 *pkey) +{ + if (index == 0xFFFF) { + *pkey = 0xFFFF; + return 0; + } + if (index > pkey_tbl->max) { + dev_err(&res->pdev->dev, + "QPLIB: Index %d exceeded PKEY table max (%d)", + index, pkey_tbl->max); + return -EINVAL; + } + memcpy(pkey, &pkey_tbl->tbl[index], sizeof(*pkey)); + return 0; +} + +int bnxt_qplib_del_pkey(struct bnxt_qplib_res *res, + struct bnxt_qplib_pkey_tbl *pkey_tbl, u16 *pkey, + bool update) +{ + int i, rc = 0; + + if (!pkey_tbl) { + dev_err(&res->pdev->dev, "QPLIB: PKEY table not allocated"); + return -EINVAL; + } + + /* Do we need a pkey_lock here? */ + if (!pkey_tbl->active) { + dev_err(&res->pdev->dev, + "QPLIB: PKEY table has no active entries"); + return -ENOMEM; + } + for (i = 0; i < pkey_tbl->max; i++) { + if (!memcmp(&pkey_tbl->tbl[i], pkey, sizeof(*pkey))) + break; + } + if (i == pkey_tbl->max) { + dev_err(&res->pdev->dev, + "QPLIB: PKEY 0x%04x not found in the pkey table", + *pkey); + return -ENOMEM; + } + memset(&pkey_tbl->tbl[i], 0, sizeof(*pkey)); + pkey_tbl->active--; + + /* unlock */ + return rc; +} + +int bnxt_qplib_add_pkey(struct bnxt_qplib_res *res, + struct bnxt_qplib_pkey_tbl *pkey_tbl, u16 *pkey, + bool update) +{ + int i, free_idx, rc = 0; + + if (!pkey_tbl) { + dev_err(&res->pdev->dev, "QPLIB: PKEY table not allocated"); + return -EINVAL; + } + + /* Do we need a pkey_lock here? */ + if (pkey_tbl->active == pkey_tbl->max) { + dev_err(&res->pdev->dev, "QPLIB: PKEY table is full"); + return -ENOMEM; + } + free_idx = pkey_tbl->max; + for (i = 0; i < pkey_tbl->max; i++) { + if (!memcmp(&pkey_tbl->tbl[i], pkey, sizeof(*pkey))) + return -EALREADY; + else if (!pkey_tbl->tbl[i] && free_idx == pkey_tbl->max) + free_idx = i; + } + if (free_idx == pkey_tbl->max) { + dev_err(&res->pdev->dev, + "QPLIB: PKEY table is FULL but count is not MAX??"); + return -ENOMEM; + } + /* Add PKEY to the pkey_tbl */ + memcpy(&pkey_tbl->tbl[free_idx], pkey, sizeof(*pkey)); + pkey_tbl->active++; + + /* unlock */ + return rc; +} + +/* AH */ +int bnxt_qplib_create_ah(struct bnxt_qplib_res *res, struct bnxt_qplib_ah *ah) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_create_ah req; + struct creq_create_ah_resp resp; + u16 cmd_flags = 0; + u32 temp32[4]; + u16 temp16[3]; + int rc; + + RCFW_CMD_PREP(req, CREATE_AH, cmd_flags); + + memcpy(temp32, ah->dgid.data, sizeof(struct bnxt_qplib_gid)); + req.dgid[0] = cpu_to_le32(temp32[0]); + req.dgid[1] = cpu_to_le32(temp32[1]); + req.dgid[2] = cpu_to_le32(temp32[2]); + req.dgid[3] = cpu_to_le32(temp32[3]); + + req.type = ah->nw_type; + req.hop_limit = ah->hop_limit; + req.sgid_index = cpu_to_le16(res->sgid_tbl.hw_id[ah->sgid_index]); + req.dest_vlan_id_flow_label = cpu_to_le32((ah->flow_label & + CMDQ_CREATE_AH_FLOW_LABEL_MASK) | + CMDQ_CREATE_AH_DEST_VLAN_ID_MASK); + req.pd_id = cpu_to_le32(ah->pd->id); + req.traffic_class = ah->traffic_class; + + /* MAC in network format */ + memcpy(temp16, ah->dmac, 6); + req.dest_mac[0] = cpu_to_le16(temp16[0]); + req.dest_mac[1] = cpu_to_le16(temp16[1]); + req.dest_mac[2] = cpu_to_le16(temp16[2]); + + rc = bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, (void *)&resp, + NULL, 1); + if (rc) + return rc; + + ah->id = le32_to_cpu(resp.xid); + return 0; +} + +int bnxt_qplib_destroy_ah(struct bnxt_qplib_res *res, struct bnxt_qplib_ah *ah) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_destroy_ah req; + struct creq_destroy_ah_resp resp; + u16 cmd_flags = 0; + int rc; + + /* Clean up the AH table in the device */ + RCFW_CMD_PREP(req, DESTROY_AH, cmd_flags); + + req.ah_cid = cpu_to_le32(ah->id); + + rc = bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, (void *)&resp, + NULL, 1); + if (rc) + return rc; + return 0; +} + +/* MRW */ +int bnxt_qplib_free_mrw(struct bnxt_qplib_res *res, struct bnxt_qplib_mrw *mrw) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_deallocate_key req; + struct creq_deallocate_key_resp resp; + u16 cmd_flags = 0; + int rc; + + if (mrw->lkey == 0xFFFFFFFF) { + dev_info(&res->pdev->dev, + "QPLIB: SP: Free a reserved lkey MRW"); + return 0; + } + + RCFW_CMD_PREP(req, DEALLOCATE_KEY, cmd_flags); + + req.mrw_flags = mrw->type; + + if ((mrw->type == CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE1) || + (mrw->type == CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE2A) || + (mrw->type == CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE2B)) + req.key = cpu_to_le32(mrw->rkey); + else + req.key = cpu_to_le32(mrw->lkey); + + rc = bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, (void *)&resp, + NULL, 0); + if (rc) + return rc; + + /* Free the qplib's MRW memory */ + if (mrw->hwq.max_elements) + bnxt_qplib_free_hwq(res->pdev, &mrw->hwq); + + return 0; +} + +int bnxt_qplib_alloc_mrw(struct bnxt_qplib_res *res, struct bnxt_qplib_mrw *mrw) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_allocate_mrw req; + struct creq_allocate_mrw_resp resp; + u16 cmd_flags = 0; + unsigned long tmp; + int rc; + + RCFW_CMD_PREP(req, ALLOCATE_MRW, cmd_flags); + + req.pd_id = cpu_to_le32(mrw->pd->id); + req.mrw_flags = mrw->type; + if ((mrw->type == CMDQ_ALLOCATE_MRW_MRW_FLAGS_PMR && + mrw->flags & BNXT_QPLIB_FR_PMR) || + mrw->type == CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE2A || + mrw->type == CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE2B) + req.access = CMDQ_ALLOCATE_MRW_ACCESS_CONSUMER_OWNED_KEY; + tmp = (unsigned long)mrw; + req.mrw_handle = cpu_to_le64(tmp); + + rc = bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + (void *)&resp, NULL, 0); + if (rc) + return rc; + + if ((mrw->type == CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE1) || + (mrw->type == CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE2A) || + (mrw->type == CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE2B)) + mrw->rkey = le32_to_cpu(resp.xid); + else + mrw->lkey = le32_to_cpu(resp.xid); + return 0; +} + +int bnxt_qplib_dereg_mrw(struct bnxt_qplib_res *res, struct bnxt_qplib_mrw *mrw, + bool block) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_deregister_mr req; + struct creq_deregister_mr_resp resp; + u16 cmd_flags = 0; + int rc; + + RCFW_CMD_PREP(req, DEREGISTER_MR, cmd_flags); + + req.lkey = cpu_to_le32(mrw->lkey); + rc = bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + (void *)&resp, NULL, block); + if (rc) + return rc; + + /* Free the qplib's MR memory */ + if (mrw->hwq.max_elements) { + mrw->va = 0; + mrw->total_size = 0; + bnxt_qplib_free_hwq(res->pdev, &mrw->hwq); + } + + return 0; +} + +int bnxt_qplib_reg_mr(struct bnxt_qplib_res *res, struct bnxt_qplib_mrw *mr, + u64 *pbl_tbl, int num_pbls, bool block) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_register_mr req; + struct creq_register_mr_resp resp; + u16 cmd_flags = 0, level; + int pg_ptrs, pages, i, rc; + dma_addr_t **pbl_ptr; + u32 pg_size; + + if (num_pbls) { + pg_ptrs = roundup_pow_of_two(num_pbls); + pages = pg_ptrs >> MAX_PBL_LVL_1_PGS_SHIFT; + if (!pages) + pages++; + + if (pages > MAX_PBL_LVL_1_PGS) { + dev_err(&res->pdev->dev, "QPLIB: SP: Reg MR pages "); + dev_err(&res->pdev->dev, + "requested (0x%x) exceeded max (0x%x)", + pages, MAX_PBL_LVL_1_PGS); + return -ENOMEM; + } + /* Free the hwq if it already exist, must be a rereg */ + if (mr->hwq.max_elements) + bnxt_qplib_free_hwq(res->pdev, &mr->hwq); + + mr->hwq.max_elements = pages; + rc = bnxt_qplib_alloc_init_hwq(res->pdev, &mr->hwq, NULL, 0, + &mr->hwq.max_elements, + PAGE_SIZE, 0, PAGE_SIZE, + HWQ_TYPE_CTX); + if (rc) { + dev_err(&res->pdev->dev, + "SP: Reg MR memory allocation failed"); + return -ENOMEM; + } + /* Write to the hwq */ + pbl_ptr = (dma_addr_t **)mr->hwq.pbl_ptr; + for (i = 0; i < num_pbls; i++) + pbl_ptr[PTR_PG(i)][PTR_IDX(i)] = + (pbl_tbl[i] & PAGE_MASK) | PTU_PTE_VALID; + } + + RCFW_CMD_PREP(req, REGISTER_MR, cmd_flags); + + /* Configure the request */ + if (mr->hwq.level == PBL_LVL_MAX) { + level = 0; + req.pbl = 0; + pg_size = PAGE_SIZE; + } else { + level = mr->hwq.level + 1; + req.pbl = cpu_to_le64(mr->hwq.pbl[PBL_LVL_0].pg_map_arr[0]); + pg_size = mr->hwq.pbl[PBL_LVL_0].pg_size; + } + req.log2_pg_size_lvl = (level << CMDQ_REGISTER_MR_LVL_SFT) | + ((ilog2(pg_size) << + CMDQ_REGISTER_MR_LOG2_PG_SIZE_SFT) & + CMDQ_REGISTER_MR_LOG2_PG_SIZE_MASK); + req.access = (mr->flags & 0xFFFF); + req.va = cpu_to_le64(mr->va); + req.key = cpu_to_le32(mr->lkey); + req.mr_size = cpu_to_le64(mr->total_size); + + rc = bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + (void *)&resp, NULL, block); + if (rc) + goto fail; + + return 0; + +fail: + if (mr->hwq.max_elements) + bnxt_qplib_free_hwq(res->pdev, &mr->hwq); + return rc; +} + +int bnxt_qplib_alloc_fast_reg_page_list(struct bnxt_qplib_res *res, + struct bnxt_qplib_frpl *frpl, + int max_pg_ptrs) +{ + int pg_ptrs, pages, rc; + + /* Re-calculate the max to fit the HWQ allocation model */ + pg_ptrs = roundup_pow_of_two(max_pg_ptrs); + pages = pg_ptrs >> MAX_PBL_LVL_1_PGS_SHIFT; + if (!pages) + pages++; + + if (pages > MAX_PBL_LVL_1_PGS) + return -ENOMEM; + + frpl->hwq.max_elements = pages; + rc = bnxt_qplib_alloc_init_hwq(res->pdev, &frpl->hwq, NULL, 0, + &frpl->hwq.max_elements, PAGE_SIZE, 0, + PAGE_SIZE, HWQ_TYPE_CTX); + if (!rc) + frpl->max_pg_ptrs = pg_ptrs; + + return rc; +} + +int bnxt_qplib_free_fast_reg_page_list(struct bnxt_qplib_res *res, + struct bnxt_qplib_frpl *frpl) +{ + bnxt_qplib_free_hwq(res->pdev, &frpl->hwq); + return 0; +} + +int bnxt_qplib_map_tc2cos(struct bnxt_qplib_res *res, u16 *cids) +{ + struct bnxt_qplib_rcfw *rcfw = res->rcfw; + struct cmdq_map_tc_to_cos req; + struct creq_map_tc_to_cos_resp resp; + u16 cmd_flags = 0; + int rc = 0; + + RCFW_CMD_PREP(req, MAP_TC_TO_COS, cmd_flags); + req.cos0 = cpu_to_le16(cids[0]); + req.cos1 = cpu_to_le16(cids[1]); + + rc = bnxt_qplib_rcfw_send_message(rcfw, (void *)&req, + (void *)&resp, NULL, 0); + return 0; +} diff --git a/drivers/infiniband/hw/bnxt_re/qplib_sp.h b/drivers/infiniband/hw/bnxt_re/qplib_sp.h new file mode 100644 index 000000000000..2ce7e2a32cf0 --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/qplib_sp.h @@ -0,0 +1,165 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * 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 + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. 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. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: Slow Path Operators (header) + * + */ + +#ifndef __BNXT_QPLIB_SP_H__ +#define __BNXT_QPLIB_SP_H__ + +#define BNXT_QPLIB_RESERVED_QP_WRS 128 + +#define PCI_EXP_DEVCTL2_ATOMIC_REQ 0x0040 + +struct bnxt_qplib_dev_attr { + char fw_ver[32]; + u16 max_sgid; + u16 max_mrw; + u32 max_qp; +#define BNXT_QPLIB_MAX_OUT_RD_ATOM 126 + u32 max_qp_rd_atom; + u32 max_qp_init_rd_atom; + u32 max_qp_wqes; + u32 max_qp_sges; + u32 max_cq; + u32 max_cq_wqes; + u32 max_cq_sges; + u32 max_mr; + u64 max_mr_size; + u32 max_pd; + u32 max_mw; + u32 max_raw_ethy_qp; + u32 max_ah; + u32 max_fmr; + u32 max_map_per_fmr; + u32 max_srq; + u32 max_srq_wqes; + u32 max_srq_sges; + u32 max_pkey; + u32 max_inline_data; + u32 l2_db_size; + u8 tqm_alloc_reqs[MAX_TQM_ALLOC_REQ]; + bool is_atomic; +}; + +struct bnxt_qplib_pd { + u32 id; +}; + +struct bnxt_qplib_gid { + u8 data[16]; +}; + +struct bnxt_qplib_ah { + struct bnxt_qplib_gid dgid; + struct bnxt_qplib_pd *pd; + u32 id; + u8 sgid_index; + /* For Query AH if the hw table and SW table are differnt */ + u8 host_sgid_index; + u8 traffic_class; + u32 flow_label; + u8 hop_limit; + u8 sl; + u8 dmac[6]; + u16 vlan_id; + u8 nw_type; +}; + +struct bnxt_qplib_mrw { + struct bnxt_qplib_pd *pd; + int type; + u32 flags; +#define BNXT_QPLIB_FR_PMR 0x80000000 + u32 lkey; + u32 rkey; +#define BNXT_QPLIB_RSVD_LKEY 0xFFFFFFFF + u64 va; + u64 total_size; + u32 npages; + u64 mr_handle; + struct bnxt_qplib_hwq hwq; +}; + +struct bnxt_qplib_frpl { + int max_pg_ptrs; + struct bnxt_qplib_hwq hwq; +}; + +#define BNXT_QPLIB_ACCESS_LOCAL_WRITE BIT(0) +#define BNXT_QPLIB_ACCESS_REMOTE_READ BIT(1) +#define BNXT_QPLIB_ACCESS_REMOTE_WRITE BIT(2) +#define BNXT_QPLIB_ACCESS_REMOTE_ATOMIC BIT(3) +#define BNXT_QPLIB_ACCESS_MW_BIND BIT(4) +#define BNXT_QPLIB_ACCESS_ZERO_BASED BIT(5) +#define BNXT_QPLIB_ACCESS_ON_DEMAND BIT(6) + +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); +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); +int bnxt_qplib_get_pkey(struct bnxt_qplib_res *res, + struct bnxt_qplib_pkey_tbl *pkey_tbl, u16 index, + u16 *pkey); +int bnxt_qplib_del_pkey(struct bnxt_qplib_res *res, + struct bnxt_qplib_pkey_tbl *pkey_tbl, u16 *pkey, + bool update); +int bnxt_qplib_add_pkey(struct bnxt_qplib_res *res, + struct bnxt_qplib_pkey_tbl *pkey_tbl, u16 *pkey, + bool update); +int bnxt_qplib_get_dev_attr(struct bnxt_qplib_rcfw *rcfw, + struct bnxt_qplib_dev_attr *attr); +int bnxt_qplib_create_ah(struct bnxt_qplib_res *res, struct bnxt_qplib_ah *ah); +int bnxt_qplib_destroy_ah(struct bnxt_qplib_res *res, struct bnxt_qplib_ah *ah); +int bnxt_qplib_alloc_mrw(struct bnxt_qplib_res *res, + struct bnxt_qplib_mrw *mrw); +int bnxt_qplib_dereg_mrw(struct bnxt_qplib_res *res, struct bnxt_qplib_mrw *mrw, + bool block); +int bnxt_qplib_reg_mr(struct bnxt_qplib_res *res, struct bnxt_qplib_mrw *mr, + u64 *pbl_tbl, int num_pbls, bool block); +int bnxt_qplib_free_mrw(struct bnxt_qplib_res *res, struct bnxt_qplib_mrw *mr); +int bnxt_qplib_alloc_fast_reg_mr(struct bnxt_qplib_res *res, + struct bnxt_qplib_mrw *mr, int max); +int bnxt_qplib_alloc_fast_reg_page_list(struct bnxt_qplib_res *res, + struct bnxt_qplib_frpl *frpl, int max); +int bnxt_qplib_free_fast_reg_page_list(struct bnxt_qplib_res *res, + struct bnxt_qplib_frpl *frpl); +int bnxt_qplib_map_tc2cos(struct bnxt_qplib_res *res, u16 *cids); +#endif /* __BNXT_QPLIB_SP_H__*/ diff --git a/drivers/infiniband/hw/bnxt_re/roce_hsi.h b/drivers/infiniband/hw/bnxt_re/roce_hsi.h new file mode 100644 index 000000000000..fc23477ac52f --- /dev/null +++ b/drivers/infiniband/hw/bnxt_re/roce_hsi.h @@ -0,0 +1,2821 @@ +/* + * Broadcom NetXtreme-E RoCE driver. + * + * Copyright (c) 2016 - 2017, Broadcom. All rights reserved. The term + * Broadcom refers to Broadcom Limited and/or its subsidiaries. + * + * 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 + * BSD license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. 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. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN + * IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * Description: RoCE HSI File - Autogenerated + */ + +#ifndef __BNXT_RE_HSI_H__ +#define __BNXT_RE_HSI_H__ + +/* include bnxt_hsi.h from bnxt_en driver */ +#include "bnxt_hsi.h" + +/* CMP Door Bell Format (4 bytes) */ +struct cmpl_doorbell { + __le32 key_mask_valid_idx; + #define CMPL_DOORBELL_IDX_MASK 0xffffffUL + #define CMPL_DOORBELL_IDX_SFT 0 + #define CMPL_DOORBELL_RESERVED_MASK 0x3000000UL + #define CMPL_DOORBELL_RESERVED_SFT 24 + #define CMPL_DOORBELL_IDX_VALID 0x4000000UL + #define CMPL_DOORBELL_MASK 0x8000000UL + #define CMPL_DOORBELL_KEY_MASK 0xf0000000UL + #define CMPL_DOORBELL_KEY_SFT 28 + #define CMPL_DOORBELL_KEY_CMPL (0x2UL << 28) +}; + +/* Status Door Bell Format (4 bytes) */ +struct status_doorbell { + __le32 key_idx; + #define STATUS_DOORBELL_IDX_MASK 0xffffffUL + #define STATUS_DOORBELL_IDX_SFT 0 + #define STATUS_DOORBELL_RESERVED_MASK 0xf000000UL + #define STATUS_DOORBELL_RESERVED_SFT 24 + #define STATUS_DOORBELL_KEY_MASK 0xf0000000UL + #define STATUS_DOORBELL_KEY_SFT 28 + #define STATUS_DOORBELL_KEY_STAT (0x3UL << 28) +}; + +/* RoCE Host Structures */ + +/* Doorbell Structures */ +/* 64b Doorbell Format (8 bytes) */ +struct dbr_dbr { + __le32 index; + #define DBR_DBR_INDEX_MASK 0xfffffUL + #define DBR_DBR_INDEX_SFT 0 + #define DBR_DBR_RESERVED12_MASK 0xfff00000UL + #define DBR_DBR_RESERVED12_SFT 20 + __le32 type_xid; + #define DBR_DBR_XID_MASK 0xfffffUL + #define DBR_DBR_XID_SFT 0 + #define DBR_DBR_RESERVED8_MASK 0xff00000UL + #define DBR_DBR_RESERVED8_SFT 20 + #define DBR_DBR_TYPE_MASK 0xf0000000UL + #define DBR_DBR_TYPE_SFT 28 + #define DBR_DBR_TYPE_SQ (0x0UL << 28) + #define DBR_DBR_TYPE_RQ (0x1UL << 28) + #define DBR_DBR_TYPE_SRQ (0x2UL << 28) + #define DBR_DBR_TYPE_SRQ_ARM (0x3UL << 28) + #define DBR_DBR_TYPE_CQ (0x4UL << 28) + #define DBR_DBR_TYPE_CQ_ARMSE (0x5UL << 28) + #define DBR_DBR_TYPE_CQ_ARMALL (0x6UL << 28) + #define DBR_DBR_TYPE_CQ_ARMENA (0x7UL << 28) + #define DBR_DBR_TYPE_SRQ_ARMENA (0x8UL << 28) + #define DBR_DBR_TYPE_CQ_CUTOFF_ACK (0x9UL << 28) + #define DBR_DBR_TYPE_NULL (0xfUL << 28) +}; + +/* 32b Doorbell Format (4 bytes) */ +struct dbr_dbr32 { + __le32 type_abs_incr_xid; + #define DBR_DBR32_XID_MASK 0xfffffUL + #define DBR_DBR32_XID_SFT 0 + #define DBR_DBR32_RESERVED4_MASK 0xf00000UL + #define DBR_DBR32_RESERVED4_SFT 20 + #define DBR_DBR32_INCR_MASK 0xf000000UL + #define DBR_DBR32_INCR_SFT 24 + #define DBR_DBR32_ABS 0x10000000UL + #define DBR_DBR32_TYPE_MASK 0xe0000000UL + #define DBR_DBR32_TYPE_SFT 29 + #define DBR_DBR32_TYPE_SQ (0x0UL << 29) +}; + +/* SQ WQE Structures */ +/* Base SQ WQE (8 bytes) */ +struct sq_base { + u8 wqe_type; + #define SQ_BASE_WQE_TYPE_SEND 0x0UL + #define SQ_BASE_WQE_TYPE_SEND_W_IMMEAD 0x1UL + #define SQ_BASE_WQE_TYPE_SEND_W_INVALID 0x2UL + #define SQ_BASE_WQE_TYPE_WRITE_WQE 0x4UL + #define SQ_BASE_WQE_TYPE_WRITE_W_IMMEAD 0x5UL + #define SQ_BASE_WQE_TYPE_READ_WQE 0x6UL + #define SQ_BASE_WQE_TYPE_ATOMIC_CS 0x8UL + #define SQ_BASE_WQE_TYPE_ATOMIC_FA 0xbUL + #define SQ_BASE_WQE_TYPE_LOCAL_INVALID 0xcUL + #define SQ_BASE_WQE_TYPE_FR_PMR 0xdUL + #define SQ_BASE_WQE_TYPE_BIND 0xeUL + u8 unused_0[7]; +}; + +/* WQE SGE (16 bytes) */ +struct sq_sge { + __le64 va_or_pa; + __le32 l_key; + __le32 size; +}; + +/* PSN Search Structure (8 bytes) */ +struct sq_psn_search { + __le32 opcode_start_psn; + #define SQ_PSN_SEARCH_START_PSN_MASK 0xffffffUL + #define SQ_PSN_SEARCH_START_PSN_SFT 0 + #define SQ_PSN_SEARCH_OPCODE_MASK 0xff000000UL + #define SQ_PSN_SEARCH_OPCODE_SFT 24 + __le32 flags_next_psn; + #define SQ_PSN_SEARCH_NEXT_PSN_MASK 0xffffffUL + #define SQ_PSN_SEARCH_NEXT_PSN_SFT 0 + #define SQ_PSN_SEARCH_FLAGS_MASK 0xff000000UL + #define SQ_PSN_SEARCH_FLAGS_SFT 24 +}; + +/* Send SQ WQE (40 bytes) */ +struct sq_send { + u8 wqe_type; + #define SQ_SEND_WQE_TYPE_SEND 0x0UL + #define SQ_SEND_WQE_TYPE_SEND_W_IMMEAD 0x1UL + #define SQ_SEND_WQE_TYPE_SEND_W_INVALID 0x2UL + u8 flags; + #define SQ_SEND_FLAGS_SIGNAL_COMP 0x1UL + #define SQ_SEND_FLAGS_RD_OR_ATOMIC_FENCE 0x2UL + #define SQ_SEND_FLAGS_UC_FENCE 0x4UL + #define SQ_SEND_FLAGS_SE 0x8UL + #define SQ_SEND_FLAGS_INLINE 0x10UL + u8 wqe_size; + u8 reserved8_1; + __le32 inv_key_or_imm_data; + __le32 length; + __le32 q_key; + __le32 dst_qp; + #define SQ_SEND_DST_QP_MASK 0xffffffUL + #define SQ_SEND_DST_QP_SFT 0 + #define SQ_SEND_RESERVED8_2_MASK 0xff000000UL + #define SQ_SEND_RESERVED8_2_SFT 24 + __le32 avid; + #define SQ_SEND_AVID_MASK 0xfffffUL + #define SQ_SEND_AVID_SFT 0 + #define SQ_SEND_RESERVED_AVID_MASK 0xfff00000UL + #define SQ_SEND_RESERVED_AVID_SFT 20 + __le64 reserved64; + __le32 data[24]; +}; + +/* Send Raw Ethernet and QP1 SQ WQE (40 bytes) */ +struct sq_send_raweth_qp1 { + u8 wqe_type; + #define SQ_SEND_RAWETH_QP1_WQE_TYPE_SEND 0x0UL + u8 flags; + #define SQ_SEND_RAWETH_QP1_FLAGS_SIGNAL_COMP 0x1UL + #define SQ_SEND_RAWETH_QP1_FLAGS_RD_OR_ATOMIC_FENCE 0x2UL + #define SQ_SEND_RAWETH_QP1_FLAGS_UC_FENCE 0x4UL + #define SQ_SEND_RAWETH_QP1_FLAGS_SE 0x8UL + #define SQ_SEND_RAWETH_QP1_FLAGS_INLINE 0x10UL + u8 wqe_size; + u8 reserved8; + __le16 lflags; + #define SQ_SEND_RAWETH_QP1_LFLAGS_TCP_UDP_CHKSUM 0x1UL + #define SQ_SEND_RAWETH_QP1_LFLAGS_IP_CHKSUM 0x2UL + #define SQ_SEND_RAWETH_QP1_LFLAGS_NOCRC 0x4UL + #define SQ_SEND_RAWETH_QP1_LFLAGS_STAMP 0x8UL + #define SQ_SEND_RAWETH_QP1_LFLAGS_T_IP_CHKSUM 0x10UL + #define SQ_SEND_RAWETH_QP1_LFLAGS_RESERVED1_1 0x20UL + #define SQ_SEND_RAWETH_QP1_LFLAGS_RESERVED1_2 0x40UL + #define SQ_SEND_RAWETH_QP1_LFLAGS_RESERVED1_3 0x80UL + #define SQ_SEND_RAWETH_QP1_LFLAGS_ROCE_CRC 0x100UL + #define SQ_SEND_RAWETH_QP1_LFLAGS_FCOE_CRC 0x200UL + __le16 cfa_action; + __le32 length; + __le32 reserved32_1; + __le32 cfa_meta; + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_VID_MASK 0xfffUL + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_VID_SFT 0 + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_DE 0x1000UL + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_PRI_MASK 0xe000UL + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_PRI_SFT 13 + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_TPID_MASK 0x70000UL + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_TPID_SFT 16 + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_TPID_TPID88A8 (0x0UL << 16) + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_TPID_TPID8100 (0x1UL << 16) + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_TPID_TPID9100 (0x2UL << 16) + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_TPID_TPID9200 (0x3UL << 16) + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_TPID_TPID9300 (0x4UL << 16) + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_TPID_TPIDCFG (0x5UL << 16) + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_TPID_LAST \ + SQ_SEND_RAWETH_QP1_CFA_META_VLAN_TPID_TPIDCFG + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_RESERVED_MASK 0xff80000UL + #define SQ_SEND_RAWETH_QP1_CFA_META_VLAN_RESERVED_SFT 19 + #define SQ_SEND_RAWETH_QP1_CFA_META_KEY_MASK 0xf0000000UL + #define SQ_SEND_RAWETH_QP1_CFA_META_KEY_SFT 28 + #define SQ_SEND_RAWETH_QP1_CFA_META_KEY_NONE (0x0UL << 28) + #define SQ_SEND_RAWETH_QP1_CFA_META_KEY_VLAN_TAG (0x1UL << 28) + #define SQ_SEND_RAWETH_QP1_CFA_META_KEY_LAST \ + SQ_SEND_RAWETH_QP1_CFA_META_KEY_VLAN_TAG + __le32 reserved32_2; + __le64 reserved64; + __le32 data[24]; +}; + +/* RDMA SQ WQE (40 bytes) */ +struct sq_rdma { + u8 wqe_type; + #define SQ_RDMA_WQE_TYPE_WRITE_WQE 0x4UL + #define SQ_RDMA_WQE_TYPE_WRITE_W_IMMEAD 0x5UL + #define SQ_RDMA_WQE_TYPE_READ_WQE 0x6UL + u8 flags; + #define SQ_RDMA_FLAGS_SIGNAL_COMP 0x1UL + #define SQ_RDMA_FLAGS_RD_OR_ATOMIC_FENCE 0x2UL + #define SQ_RDMA_FLAGS_UC_FENCE 0x4UL + #define SQ_RDMA_FLAGS_SE 0x8UL + #define SQ_RDMA_FLAGS_INLINE 0x10UL + u8 wqe_size; + u8 reserved8; + __le32 imm_data; + __le32 length; + __le32 reserved32_1; + __le64 remote_va; + __le32 remote_key; + __le32 reserved32_2; + __le32 data[24]; +}; + +/* Atomic SQ WQE (40 bytes) */ +struct sq_atomic { + u8 wqe_type; + #define SQ_ATOMIC_WQE_TYPE_ATOMIC_CS 0x8UL + #define SQ_ATOMIC_WQE_TYPE_ATOMIC_FA 0xbUL + u8 flags; + #define SQ_ATOMIC_FLAGS_SIGNAL_COMP 0x1UL + #define SQ_ATOMIC_FLAGS_RD_OR_ATOMIC_FENCE 0x2UL + #define SQ_ATOMIC_FLAGS_UC_FENCE 0x4UL + #define SQ_ATOMIC_FLAGS_SE 0x8UL + #define SQ_ATOMIC_FLAGS_INLINE 0x10UL + __le16 reserved16; + __le32 remote_key; + __le64 remote_va; + __le64 swap_data; + __le64 cmp_data; + __le32 data[24]; +}; + +/* Local Invalidate SQ WQE (40 bytes) */ +struct sq_localinvalidate { + u8 wqe_type; + #define SQ_LOCALINVALIDATE_WQE_TYPE_LOCAL_INVALID 0xcUL + u8 flags; + #define SQ_LOCALINVALIDATE_FLAGS_SIGNAL_COMP 0x1UL + #define SQ_LOCALINVALIDATE_FLAGS_RD_OR_ATOMIC_FENCE 0x2UL + #define SQ_LOCALINVALIDATE_FLAGS_UC_FENCE 0x4UL + #define SQ_LOCALINVALIDATE_FLAGS_SE 0x8UL + #define SQ_LOCALINVALIDATE_FLAGS_INLINE 0x10UL + __le16 reserved16; + __le32 inv_l_key; + __le64 reserved64; + __le32 reserved128[4]; + __le32 data[24]; +}; + +/* FR-PMR SQ WQE (40 bytes) */ +struct sq_fr_pmr { + u8 wqe_type; + #define SQ_FR_PMR_WQE_TYPE_FR_PMR 0xdUL + u8 flags; + #define SQ_FR_PMR_FLAGS_SIGNAL_COMP 0x1UL + #define SQ_FR_PMR_FLAGS_RD_OR_ATOMIC_FENCE 0x2UL + #define SQ_FR_PMR_FLAGS_UC_FENCE 0x4UL + #define SQ_FR_PMR_FLAGS_SE 0x8UL + #define SQ_FR_PMR_FLAGS_INLINE 0x10UL + u8 access_cntl; + #define SQ_FR_PMR_ACCESS_CNTL_LOCAL_WRITE 0x1UL + #define SQ_FR_PMR_ACCESS_CNTL_REMOTE_READ 0x2UL + #define SQ_FR_PMR_ACCESS_CNTL_REMOTE_WRITE 0x4UL + #define SQ_FR_PMR_ACCESS_CNTL_REMOTE_ATOMIC 0x8UL + #define SQ_FR_PMR_ACCESS_CNTL_WINDOW_BIND 0x10UL + u8 zero_based_page_size_log; + #define SQ_FR_PMR_PAGE_SIZE_LOG_MASK 0x1fUL + #define SQ_FR_PMR_PAGE_SIZE_LOG_SFT 0 + #define SQ_FR_PMR_PAGE_SIZE_LOG_PGSZ_4K 0x0UL + #define SQ_FR_PMR_PAGE_SIZE_LOG_PGSZ_8K 0x1UL + #define SQ_FR_PMR_PAGE_SIZE_LOG_PGSZ_64K 0x4UL + #define SQ_FR_PMR_PAGE_SIZE_LOG_PGSZ_256K 0x6UL + #define SQ_FR_PMR_PAGE_SIZE_LOG_PGSZ_1M 0x8UL + #define SQ_FR_PMR_PAGE_SIZE_LOG_PGSZ_2M 0x9UL + #define SQ_FR_PMR_PAGE_SIZE_LOG_PGSZ_4M 0xaUL + #define SQ_FR_PMR_PAGE_SIZE_LOG_PGSZ_1G 0x12UL + #define SQ_FR_PMR_ZERO_BASED 0x20UL + #define SQ_FR_PMR_RESERVED2_MASK 0xc0UL + #define SQ_FR_PMR_RESERVED2_SFT 6 + __le32 l_key; + u8 length[5]; + u8 reserved8_1; + u8 reserved8_2; + u8 numlevels_pbl_page_size_log; + #define SQ_FR_PMR_PBL_PAGE_SIZE_LOG_MASK 0x1fUL + #define SQ_FR_PMR_PBL_PAGE_SIZE_LOG_SFT 0 + #define SQ_FR_PMR_PBL_PAGE_SIZE_LOG_PGSZ_4K 0x0UL + #define SQ_FR_PMR_PBL_PAGE_SIZE_LOG_PGSZ_8K 0x1UL + #define SQ_FR_PMR_PBL_PAGE_SIZE_LOG_PGSZ_64K 0x4UL + #define SQ_FR_PMR_PBL_PAGE_SIZE_LOG_PGSZ_256K 0x6UL + #define SQ_FR_PMR_PBL_PAGE_SIZE_LOG_PGSZ_1M 0x8UL + #define SQ_FR_PMR_PBL_PAGE_SIZE_LOG_PGSZ_2M 0x9UL + #define SQ_FR_PMR_PBL_PAGE_SIZE_LOG_PGSZ_4M 0xaUL + #define SQ_FR_PMR_PBL_PAGE_SIZE_LOG_PGSZ_1G 0x12UL + #define SQ_FR_PMR_RESERVED1 0x20UL + #define SQ_FR_PMR_NUMLEVELS_MASK 0xc0UL + #define SQ_FR_PMR_NUMLEVELS_SFT 6 + #define SQ_FR_PMR_NUMLEVELS_PHYSICAL (0x0UL << 6) + #define SQ_FR_PMR_NUMLEVELS_LAYER1 (0x1UL << 6) + #define SQ_FR_PMR_NUMLEVELS_LAYER2 (0x2UL << 6) + __le64 pblptr; + __le64 va; + __le32 data[24]; +}; + +/* Bind SQ WQE (40 bytes) */ +struct sq_bind { + u8 wqe_type; + #define SQ_BIND_WQE_TYPE_BIND 0xeUL + u8 flags; + #define SQ_BIND_FLAGS_SIGNAL_COMP 0x1UL + #define SQ_BIND_FLAGS_RD_OR_ATOMIC_FENCE 0x2UL + #define SQ_BIND_FLAGS_UC_FENCE 0x4UL + #define SQ_BIND_FLAGS_SE 0x8UL + #define SQ_BIND_FLAGS_INLINE 0x10UL + u8 access_cntl; + #define SQ_BIND_ACCESS_CNTL_LOCAL_WRITE 0x1UL + #define SQ_BIND_ACCESS_CNTL_REMOTE_READ 0x2UL + #define SQ_BIND_ACCESS_CNTL_REMOTE_WRITE 0x4UL + #define SQ_BIND_ACCESS_CNTL_REMOTE_ATOMIC 0x8UL + #define SQ_BIND_ACCESS_CNTL_WINDOW_BIND 0x10UL + u8 reserved8_1; + u8 mw_type_zero_based; + #define SQ_BIND_ZERO_BASED 0x1UL + #define SQ_BIND_MW_TYPE 0x2UL + #define SQ_BIND_MW_TYPE_TYPE1 (0x0UL << 1) + #define SQ_BIND_MW_TYPE_TYPE2 (0x1UL << 1) + #define SQ_BIND_RESERVED6_MASK 0xfcUL + #define SQ_BIND_RESERVED6_SFT 2 + u8 reserved8_2; + __le16 reserved16; + __le32 parent_l_key; + __le32 l_key; + __le64 va; + u8 length[5]; + u8 data_reserved24[99]; + #define SQ_BIND_RESERVED24_MASK 0xffffff00UL + #define SQ_BIND_RESERVED24_SFT 8 + #define SQ_BIND_DATA_MASK 0xffffffffUL + #define SQ_BIND_DATA_SFT 0 +}; + +/* RQ/SRQ WQE Structures */ +/* RQ/SRQ WQE (40 bytes) */ +struct rq_wqe { + u8 wqe_type; + #define RQ_WQE_WQE_TYPE_RCV 0x80UL + u8 flags; + u8 wqe_size; + u8 reserved8; + __le32 reserved32; + __le32 wr_id[2]; + #define RQ_WQE_WR_ID_MASK 0xfffffUL + #define RQ_WQE_WR_ID_SFT 0 + #define RQ_WQE_RESERVED44_MASK 0xfff00000UL + #define RQ_WQE_RESERVED44_SFT 20 + __le32 reserved128[4]; + __le32 data[24]; +}; + +/* CQ CQE Structures */ +/* Base CQE (32 bytes) */ +struct cq_base { + __le64 reserved64_1; + __le64 reserved64_2; + __le64 reserved64_3; + u8 cqe_type_toggle; + #define CQ_BASE_TOGGLE 0x1UL + #define CQ_BASE_CQE_TYPE_MASK 0x1eUL + #define CQ_BASE_CQE_TYPE_SFT 1 + #define CQ_BASE_CQE_TYPE_REQ (0x0UL << 1) + #define CQ_BASE_CQE_TYPE_RES_RC (0x1UL << 1) + #define CQ_BASE_CQE_TYPE_RES_UD (0x2UL << 1) + #define CQ_BASE_CQE_TYPE_RES_RAWETH_QP1 (0x3UL << 1) + #define CQ_BASE_CQE_TYPE_TERMINAL (0xeUL << 1) + #define CQ_BASE_CQE_TYPE_CUT_OFF (0xfUL << 1) + #define CQ_BASE_RESERVED3_MASK 0xe0UL + #define CQ_BASE_RESERVED3_SFT 5 + u8 status; + __le16 reserved16; + __le32 reserved32; +}; + +/* Requester CQ CQE (32 bytes) */ +struct cq_req { + __le64 qp_handle; + __le16 sq_cons_idx; + __le16 reserved16_1; + __le32 reserved32_2; + __le64 reserved64; + u8 cqe_type_toggle; + #define CQ_REQ_TOGGLE 0x1UL + #define CQ_REQ_CQE_TYPE_MASK 0x1eUL + #define CQ_REQ_CQE_TYPE_SFT 1 + #define CQ_REQ_CQE_TYPE_REQ (0x0UL << 1) + #define CQ_REQ_RESERVED3_MASK 0xe0UL + #define CQ_REQ_RESERVED3_SFT 5 + u8 status; + #define CQ_REQ_STATUS_OK 0x0UL + #define CQ_REQ_STATUS_BAD_RESPONSE_ERR 0x1UL + #define CQ_REQ_STATUS_LOCAL_LENGTH_ERR 0x2UL + #define CQ_REQ_STATUS_LOCAL_QP_OPERATION_ERR 0x3UL + #define CQ_REQ_STATUS_LOCAL_PROTECTION_ERR 0x4UL + #define CQ_REQ_STATUS_MEMORY_MGT_OPERATION_ERR 0x5UL + #define CQ_REQ_STATUS_REMOTE_INVALID_REQUEST_ERR 0x6UL + #define CQ_REQ_STATUS_REMOTE_ACCESS_ERR 0x7UL + #define CQ_REQ_STATUS_REMOTE_OPERATION_ERR 0x8UL + #define CQ_REQ_STATUS_RNR_NAK_RETRY_CNT_ERR 0x9UL + #define CQ_REQ_STATUS_TRANSPORT_RETRY_CNT_ERR 0xaUL + #define CQ_REQ_STATUS_WORK_REQUEST_FLUSHED_ERR 0xbUL + __le16 reserved16_2; + __le32 reserved32_1; +}; + +/* Responder RC CQE (32 bytes) */ +struct cq_res_rc { + __le32 length; + __le32 imm_data_or_inv_r_key; + __le64 qp_handle; + __le64 mr_handle; + u8 cqe_type_toggle; + #define CQ_RES_RC_TOGGLE 0x1UL + #define CQ_RES_RC_CQE_TYPE_MASK 0x1eUL + #define CQ_RES_RC_CQE_TYPE_SFT 1 + #define CQ_RES_RC_CQE_TYPE_RES_RC (0x1UL << 1) + #define CQ_RES_RC_RESERVED3_MASK 0xe0UL + #define CQ_RES_RC_RESERVED3_SFT 5 + u8 status; + #define CQ_RES_RC_STATUS_OK 0x0UL + #define CQ_RES_RC_STATUS_LOCAL_ACCESS_ERROR 0x1UL + #define CQ_RES_RC_STATUS_LOCAL_LENGTH_ERR 0x2UL + #define CQ_RES_RC_STATUS_LOCAL_PROTECTION_ERR 0x3UL + #define CQ_RES_RC_STATUS_LOCAL_QP_OPERATION_ERR 0x4UL + #define CQ_RES_RC_STATUS_MEMORY_MGT_OPERATION_ERR 0x5UL + #define CQ_RES_RC_STATUS_REMOTE_INVALID_REQUEST_ERR 0x6UL + #define CQ_RES_RC_STATUS_WORK_REQUEST_FLUSHED_ERR 0x7UL + #define CQ_RES_RC_STATUS_HW_FLUSH_ERR 0x8UL + __le16 flags; + #define CQ_RES_RC_FLAGS_SRQ 0x1UL + #define CQ_RES_RC_FLAGS_SRQ_RQ (0x0UL << 0) + #define CQ_RES_RC_FLAGS_SRQ_SRQ (0x1UL << 0) + #define CQ_RES_RC_FLAGS_SRQ_LAST CQ_RES_RC_FLAGS_SRQ_SRQ + #define CQ_RES_RC_FLAGS_IMM 0x2UL + #define CQ_RES_RC_FLAGS_INV 0x4UL + #define CQ_RES_RC_FLAGS_RDMA 0x8UL + #define CQ_RES_RC_FLAGS_RDMA_SEND (0x0UL << 3) + #define CQ_RES_RC_FLAGS_RDMA_RDMA_WRITE (0x1UL << 3) + #define CQ_RES_RC_FLAGS_RDMA_LAST CQ_RES_RC_FLAGS_RDMA_RDMA_WRITE + __le32 srq_or_rq_wr_id; + #define CQ_RES_RC_SRQ_OR_RQ_WR_ID_MASK 0xfffffUL + #define CQ_RES_RC_SRQ_OR_RQ_WR_ID_SFT 0 + #define CQ_RES_RC_RESERVED12_MASK 0xfff00000UL + #define CQ_RES_RC_RESERVED12_SFT 20 +}; + +/* Responder UD CQE (32 bytes) */ +struct cq_res_ud { + __le32 length; + #define CQ_RES_UD_LENGTH_MASK 0x3fffUL + #define CQ_RES_UD_LENGTH_SFT 0 + #define CQ_RES_UD_RESERVED18_MASK 0xffffc000UL + #define CQ_RES_UD_RESERVED18_SFT 14 + __le32 imm_data; + __le64 qp_handle; + __le16 src_mac[3]; + __le16 src_qp_low; + u8 cqe_type_toggle; + #define CQ_RES_UD_TOGGLE 0x1UL + #define CQ_RES_UD_CQE_TYPE_MASK 0x1eUL + #define CQ_RES_UD_CQE_TYPE_SFT 1 + #define CQ_RES_UD_CQE_TYPE_RES_UD (0x2UL << 1) + #define CQ_RES_UD_RESERVED3_MASK 0xe0UL + #define CQ_RES_UD_RESERVED3_SFT 5 + u8 status; + #define CQ_RES_UD_STATUS_OK 0x0UL + #define CQ_RES_UD_STATUS_LOCAL_ACCESS_ERROR 0x1UL + #define CQ_RES_UD_STATUS_HW_LOCAL_LENGTH_ERR 0x2UL + #define CQ_RES_UD_STATUS_LOCAL_PROTECTION_ERR 0x3UL + #define CQ_RES_UD_STATUS_LOCAL_QP_OPERATION_ERR 0x4UL + #define CQ_RES_UD_STATUS_MEMORY_MGT_OPERATION_ERR 0x5UL + #define CQ_RES_UD_STATUS_WORK_REQUEST_FLUSHED_ERR 0x7UL + #define CQ_RES_UD_STATUS_HW_FLUSH_ERR 0x8UL + __le16 flags; + #define CQ_RES_UD_FLAGS_SRQ 0x1UL + #define CQ_RES_UD_FLAGS_SRQ_RQ (0x0UL << 0) + #define CQ_RES_UD_FLAGS_SRQ_SRQ (0x1UL << 0) + #define CQ_RES_UD_FLAGS_SRQ_LAST CQ_RES_UD_FLAGS_SRQ_SRQ + #define CQ_RES_UD_FLAGS_IMM 0x2UL + #define CQ_RES_UD_FLAGS_ROCE_IP_VER_MASK 0xcUL + #define CQ_RES_UD_FLAGS_ROCE_IP_VER_SFT 2 + #define CQ_RES_UD_FLAGS_ROCE_IP_VER_V1 (0x0UL << 2) + #define CQ_RES_UD_FLAGS_ROCE_IP_VER_V2IPV4 (0x2UL << 2) + #define CQ_RES_UD_FLAGS_ROCE_IP_VER_V2IPV6 (0x3UL << 2) + #define CQ_RES_UD_FLAGS_ROCE_IP_VER_LAST \ + CQ_RES_UD_FLAGS_ROCE_IP_VER_V2IPV6 + __le32 src_qp_high_srq_or_rq_wr_id; + #define CQ_RES_UD_SRQ_OR_RQ_WR_ID_MASK 0xfffffUL + #define CQ_RES_UD_SRQ_OR_RQ_WR_ID_SFT 0 + #define CQ_RES_UD_RESERVED4_MASK 0xf00000UL + #define CQ_RES_UD_RESERVED4_SFT 20 + #define CQ_RES_UD_SRC_QP_HIGH_MASK 0xff000000UL + #define CQ_RES_UD_SRC_QP_HIGH_SFT 24 +}; + +/* Responder RawEth and QP1 CQE (32 bytes) */ +struct cq_res_raweth_qp1 { + __le16 length; + #define CQ_RES_RAWETH_QP1_LENGTH_MASK 0x3fffUL + #define CQ_RES_RAWETH_QP1_LENGTH_SFT 0 + #define CQ_RES_RAWETH_QP1_RESERVED2_MASK 0xc000UL + #define CQ_RES_RAWETH_QP1_RESERVED2_SFT 14 + __le16 raweth_qp1_flags; + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ERROR 0x1UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_RESERVED5_1_MASK 0x3eUL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_RESERVED5_1_SFT 1 + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_MASK 0x3c0UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_SFT 6 + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_NOT_KNOWN (0x0UL << 6) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_IP (0x1UL << 6) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_TCP (0x2UL << 6) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_UDP (0x3UL << 6) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_FCOE (0x4UL << 6) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_ROCE (0x5UL << 6) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_ICMP (0x7UL << 6) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_PTP_WO_TIMESTAMP \ + (0x8UL << 6) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_PTP_W_TIMESTAMP \ + (0x9UL << 6) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_LAST \ + CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_ITYPE_PTP_W_TIMESTAMP + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_MASK 0x3ffUL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS_SFT 0 + #define CQ_RES_RAWETH_QP1_RESERVED6_MASK 0xfc00UL + #define CQ_RES_RAWETH_QP1_RESERVED6_SFT 10 + __le16 raweth_qp1_errors; + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_RESERVED4_MASK 0xfUL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_RESERVED4_SFT 0 + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_IP_CS_ERROR 0x10UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_L4_CS_ERROR 0x20UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_IP_CS_ERROR 0x40UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_L4_CS_ERROR 0x80UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_CRC_ERROR 0x100UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_PKT_ERROR_MASK 0xe00UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_PKT_ERROR_SFT 9 + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_PKT_ERROR_NO_ERROR \ + (0x0UL << 9) + #define \ + CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_PKT_ERROR_T_L3_BAD_VERSION \ + (0x1UL << 9) + #define \ + CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_PKT_ERROR_T_L3_BAD_HDR_LEN \ + (0x2UL << 9) + #define \ + CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_PKT_ERROR_TUNNEL_TOTAL_ERROR \ + (0x3UL << 9) + #define \ + CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_PKT_ERROR_T_IP_TOTAL_ERROR \ + (0x4UL << 9) + #define \ + CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_PKT_ERROR_T_UDP_TOTAL_ERROR \ + (0x5UL << 9) + #define \ + CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_PKT_ERROR_T_L3_BAD_TTL \ + (0x6UL << 9) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_PKT_ERROR_LAST \ + CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_T_PKT_ERROR_T_L3_BAD_TTL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_MASK 0xf000UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_SFT 12 + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_NO_ERROR \ + (0x0UL << 12) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_L3_BAD_VERSION \ + (0x1UL << 12) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_L3_BAD_HDR_LEN \ + (0x2UL << 12) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_L3_BAD_TTL \ + (0x3UL << 12) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_IP_TOTAL_ERROR \ + (0x4UL << 12) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_UDP_TOTAL_ERROR \ + (0x5UL << 12) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_L4_BAD_HDR_LEN \ + (0x6UL << 12) + #define \ + CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_L4_BAD_HDR_LEN_TOO_SMALL\ + (0x7UL << 12) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_L4_BAD_OPT_LEN \ + (0x8UL << 12) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_LAST \ + CQ_RES_RAWETH_QP1_RAWETH_QP1_ERRORS_PKT_ERROR_L4_BAD_OPT_LEN + __le16 raweth_qp1_cfa_code; + __le64 qp_handle; + __le32 raweth_qp1_flags2; + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_IP_CS_CALC 0x1UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_L4_CS_CALC 0x2UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_T_IP_CS_CALC 0x4UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_T_L4_CS_CALC 0x8UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_META_FORMAT_MASK 0xf0UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_META_FORMAT_SFT 4 + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_META_FORMAT_NONE \ + (0x0UL << 4) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_META_FORMAT_VLAN \ + (0x1UL << 4) + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_META_FORMAT_LAST\ + CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_META_FORMAT_VLAN + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_FLAGS2_IP_TYPE 0x100UL + __le32 raweth_qp1_metadata; + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_METADATA_VID_MASK 0xfffUL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_METADATA_VID_SFT 0 + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_METADATA_DE 0x1000UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_METADATA_PRI_MASK 0xe000UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_METADATA_PRI_SFT 13 + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_METADATA_TPID_MASK 0xffff0000UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_METADATA_TPID_SFT 16 + u8 cqe_type_toggle; + #define CQ_RES_RAWETH_QP1_TOGGLE 0x1UL + #define CQ_RES_RAWETH_QP1_CQE_TYPE_MASK 0x1eUL + #define CQ_RES_RAWETH_QP1_CQE_TYPE_SFT 1 + #define CQ_RES_RAWETH_QP1_CQE_TYPE_RES_RAWETH_QP1 (0x3UL << 1) + #define CQ_RES_RAWETH_QP1_RESERVED3_MASK 0xe0UL + #define CQ_RES_RAWETH_QP1_RESERVED3_SFT 5 + u8 status; + #define CQ_RES_RAWETH_QP1_STATUS_OK 0x0UL + #define CQ_RES_RAWETH_QP1_STATUS_LOCAL_ACCESS_ERROR 0x1UL + #define CQ_RES_RAWETH_QP1_STATUS_HW_LOCAL_LENGTH_ERR 0x2UL + #define CQ_RES_RAWETH_QP1_STATUS_LOCAL_PROTECTION_ERR 0x3UL + #define CQ_RES_RAWETH_QP1_STATUS_LOCAL_QP_OPERATION_ERR 0x4UL + #define CQ_RES_RAWETH_QP1_STATUS_MEMORY_MGT_OPERATION_ERR 0x5UL + #define CQ_RES_RAWETH_QP1_STATUS_WORK_REQUEST_FLUSHED_ERR 0x7UL + #define CQ_RES_RAWETH_QP1_STATUS_HW_FLUSH_ERR 0x8UL + __le16 flags; + #define CQ_RES_RAWETH_QP1_FLAGS_SRQ 0x1UL + #define CQ_RES_RAWETH_QP1_FLAGS_SRQ_RQ 0x0UL + #define CQ_RES_RAWETH_QP1_FLAGS_SRQ_SRQ 0x1UL + #define CQ_RES_RAWETH_QP1_FLAGS_SRQ_LAST \ + CQ_RES_RAWETH_QP1_FLAGS_SRQ_SRQ + __le32 raweth_qp1_payload_offset_srq_or_rq_wr_id; + #define CQ_RES_RAWETH_QP1_SRQ_OR_RQ_WR_ID_MASK 0xfffffUL + #define CQ_RES_RAWETH_QP1_SRQ_OR_RQ_WR_ID_SFT 0 + #define CQ_RES_RAWETH_QP1_RESERVED4_MASK 0xf00000UL + #define CQ_RES_RAWETH_QP1_RESERVED4_SFT 20 + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_PAYLOAD_OFFSET_MASK 0xff000000UL + #define CQ_RES_RAWETH_QP1_RAWETH_QP1_PAYLOAD_OFFSET_SFT 24 +}; + +/* Terminal CQE (32 bytes) */ +struct cq_terminal { + __le64 qp_handle; + __le16 sq_cons_idx; + __le16 rq_cons_idx; + __le32 reserved32_1; + __le64 reserved64_3; + u8 cqe_type_toggle; + #define CQ_TERMINAL_TOGGLE 0x1UL + #define CQ_TERMINAL_CQE_TYPE_MASK 0x1eUL + #define CQ_TERMINAL_CQE_TYPE_SFT 1 + #define CQ_TERMINAL_CQE_TYPE_TERMINAL (0xeUL << 1) + #define CQ_TERMINAL_RESERVED3_MASK 0xe0UL + #define CQ_TERMINAL_RESERVED3_SFT 5 + u8 status; + #define CQ_TERMINAL_STATUS_OK 0x0UL + __le16 reserved16; + __le32 reserved32_2; +}; + +/* Cutoff CQE (32 bytes) */ +struct cq_cutoff { + __le64 reserved64_1; + __le64 reserved64_2; + __le64 reserved64_3; + u8 cqe_type_toggle; + #define CQ_CUTOFF_TOGGLE 0x1UL + #define CQ_CUTOFF_CQE_TYPE_MASK 0x1eUL + #define CQ_CUTOFF_CQE_TYPE_SFT 1 + #define CQ_CUTOFF_CQE_TYPE_CUT_OFF (0xfUL << 1) + #define CQ_CUTOFF_RESERVED3_MASK 0xe0UL + #define CQ_CUTOFF_RESERVED3_SFT 5 + u8 status; + #define CQ_CUTOFF_STATUS_OK 0x0UL + __le16 reserved16; + __le32 reserved32; +}; + +/* Notification Queue (NQ) Structures */ +/* Base NQ Record (16 bytes) */ +struct nq_base { + __le16 info10_type; + #define NQ_BASE_TYPE_MASK 0x3fUL + #define NQ_BASE_TYPE_SFT 0 + #define NQ_BASE_TYPE_CQ_NOTIFICATION 0x30UL + #define NQ_BASE_TYPE_SRQ_EVENT 0x32UL + #define NQ_BASE_TYPE_DBQ_EVENT 0x34UL + #define NQ_BASE_TYPE_QP_EVENT 0x38UL + #define NQ_BASE_TYPE_FUNC_EVENT 0x3aUL + #define NQ_BASE_INFO10_MASK 0xffc0UL + #define NQ_BASE_INFO10_SFT 6 + __le16 info16; + __le32 info32; + __le32 info63_v[2]; + #define NQ_BASE_V 0x1UL + #define NQ_BASE_INFO63_MASK 0xfffffffeUL + #define NQ_BASE_INFO63_SFT 1 +}; + +/* Completion Queue Notification (16 bytes) */ +struct nq_cn { + __le16 type; + #define NQ_CN_TYPE_MASK 0x3fUL + #define NQ_CN_TYPE_SFT 0 + #define NQ_CN_TYPE_CQ_NOTIFICATION 0x30UL + #define NQ_CN_RESERVED9_MASK 0xffc0UL + #define NQ_CN_RESERVED9_SFT 6 + __le16 reserved16; + __le32 cq_handle_low; + __le32 v; + #define NQ_CN_V 0x1UL + #define NQ_CN_RESERVED31_MASK 0xfffffffeUL + #define NQ_CN_RESERVED31_SFT 1 + __le32 cq_handle_high; +}; + +/* SRQ Event Notification (16 bytes) */ +struct nq_srq_event { + u8 type; + #define NQ_SRQ_EVENT_TYPE_MASK 0x3fUL + #define NQ_SRQ_EVENT_TYPE_SFT 0 + #define NQ_SRQ_EVENT_TYPE_SRQ_EVENT 0x32UL + #define NQ_SRQ_EVENT_RESERVED1_MASK 0xc0UL + #define NQ_SRQ_EVENT_RESERVED1_SFT 6 + u8 event; + #define NQ_SRQ_EVENT_EVENT_SRQ_THRESHOLD_EVENT 0x1UL + __le16 reserved16; + __le32 srq_handle_low; + __le32 v; + #define NQ_SRQ_EVENT_V 0x1UL + #define NQ_SRQ_EVENT_RESERVED31_MASK 0xfffffffeUL + #define NQ_SRQ_EVENT_RESERVED31_SFT 1 + __le32 srq_handle_high; +}; + +/* DBQ Async Event Notification (16 bytes) */ +struct nq_dbq_event { + u8 type; + #define NQ_DBQ_EVENT_TYPE_MASK 0x3fUL + #define NQ_DBQ_EVENT_TYPE_SFT 0 + #define NQ_DBQ_EVENT_TYPE_DBQ_EVENT 0x34UL + #define NQ_DBQ_EVENT_RESERVED1_MASK 0xc0UL + #define NQ_DBQ_EVENT_RESERVED1_SFT 6 + u8 event; + #define NQ_DBQ_EVENT_EVENT_DBQ_THRESHOLD_EVENT 0x1UL + __le16 db_pfid; + #define NQ_DBQ_EVENT_DB_PFID_MASK 0xfUL + #define NQ_DBQ_EVENT_DB_PFID_SFT 0 + #define NQ_DBQ_EVENT_RESERVED12_MASK 0xfff0UL + #define NQ_DBQ_EVENT_RESERVED12_SFT 4 + __le32 db_dpi; + #define NQ_DBQ_EVENT_DB_DPI_MASK 0xfffffUL + #define NQ_DBQ_EVENT_DB_DPI_SFT 0 + #define NQ_DBQ_EVENT_RESERVED12_2_MASK 0xfff00000UL + #define NQ_DBQ_EVENT_RESERVED12_2_SFT 20 + __le32 v; + #define NQ_DBQ_EVENT_V 0x1UL + #define NQ_DBQ_EVENT_RESERVED32_MASK 0xfffffffeUL + #define NQ_DBQ_EVENT_RESERVED32_SFT 1 + __le32 db_type_db_xid; + #define NQ_DBQ_EVENT_DB_XID_MASK 0xfffffUL + #define NQ_DBQ_EVENT_DB_XID_SFT 0 + #define NQ_DBQ_EVENT_RESERVED8_MASK 0xff00000UL + #define NQ_DBQ_EVENT_RESERVED8_SFT 20 + #define NQ_DBQ_EVENT_DB_TYPE_MASK 0xf0000000UL + #define NQ_DBQ_EVENT_DB_TYPE_SFT 28 +}; + +/* Read Request/Response Queue Structures */ +/* Input Read Request Queue (IRRQ) Message (32 bytes) */ +struct xrrq_irrq { + __le16 credits_type; + #define XRRQ_IRRQ_TYPE 0x1UL + #define XRRQ_IRRQ_TYPE_READ_REQ 0x0UL + #define XRRQ_IRRQ_TYPE_ATOMIC_REQ 0x1UL + #define XRRQ_IRRQ_RESERVED10_MASK 0x7feUL + #define XRRQ_IRRQ_RESERVED10_SFT 1 + #define XRRQ_IRRQ_CREDITS_MASK 0xf800UL + #define XRRQ_IRRQ_CREDITS_SFT 11 + __le16 reserved16; + __le32 reserved32; + __le32 psn; + #define XRRQ_IRRQ_PSN_MASK 0xffffffUL + #define XRRQ_IRRQ_PSN_SFT 0 + #define XRRQ_IRRQ_RESERVED8_1_MASK 0xff000000UL + #define XRRQ_IRRQ_RESERVED8_1_SFT 24 + __le32 msn; + #define XRRQ_IRRQ_MSN_MASK 0xffffffUL + #define XRRQ_IRRQ_MSN_SFT 0 + #define XRRQ_IRRQ_RESERVED8_2_MASK 0xff000000UL + #define XRRQ_IRRQ_RESERVED8_2_SFT 24 + __le64 va_or_atomic_result; + __le32 rdma_r_key; + __le32 length; +}; + +/* Output Read Request Queue (ORRQ) Message (32 bytes) */ +struct xrrq_orrq { + __le16 num_sges_type; + #define XRRQ_ORRQ_TYPE 0x1UL + #define XRRQ_ORRQ_TYPE_READ_REQ 0x0UL + #define XRRQ_ORRQ_TYPE_ATOMIC_REQ 0x1UL + #define XRRQ_ORRQ_RESERVED10_MASK 0x7feUL + #define XRRQ_ORRQ_RESERVED10_SFT 1 + #define XRRQ_ORRQ_NUM_SGES_MASK 0xf800UL + #define XRRQ_ORRQ_NUM_SGES_SFT 11 + __le16 reserved16; + __le32 length; + __le32 psn; + #define XRRQ_ORRQ_PSN_MASK 0xffffffUL + #define XRRQ_ORRQ_PSN_SFT 0 + #define XRRQ_ORRQ_RESERVED8_1_MASK 0xff000000UL + #define XRRQ_ORRQ_RESERVED8_1_SFT 24 + __le32 end_psn; + #define XRRQ_ORRQ_END_PSN_MASK 0xffffffUL + #define XRRQ_ORRQ_END_PSN_SFT 0 + #define XRRQ_ORRQ_RESERVED8_2_MASK 0xff000000UL + #define XRRQ_ORRQ_RESERVED8_2_SFT 24 + __le64 first_sge_phy_or_sing_sge_va; + __le32 single_sge_l_key; + __le32 single_sge_size; +}; + +/* Page Buffer List Memory Structures (PBL) */ +/* Page Table Entry (PTE) (8 bytes) */ +struct ptu_pte { + __le32 page_next_to_last_last_valid[2]; + #define PTU_PTE_VALID 0x1UL + #define PTU_PTE_LAST 0x2UL + #define PTU_PTE_NEXT_TO_LAST 0x4UL + #define PTU_PTE_PAGE_MASK 0xfffff000UL + #define PTU_PTE_PAGE_SFT 12 +}; + +/* Page Directory Entry (PDE) (8 bytes) */ +struct ptu_pde { + __le32 page_valid[2]; + #define PTU_PDE_VALID 0x1UL + #define PTU_PDE_PAGE_MASK 0xfffff000UL + #define PTU_PDE_PAGE_SFT 12 +}; + +/* RoCE Fastpath Host Structures */ +/* Command Queue (CMDQ) Interface */ +/* Init CMDQ (16 bytes) */ +struct cmdq_init { + __le64 cmdq_pbl; + __le16 cmdq_size_cmdq_lvl; + #define CMDQ_INIT_CMDQ_LVL_MASK 0x3UL + #define CMDQ_INIT_CMDQ_LVL_SFT 0 + #define CMDQ_INIT_CMDQ_SIZE_MASK 0xfffcUL + #define CMDQ_INIT_CMDQ_SIZE_SFT 2 + __le16 creq_ring_id; + __le32 prod_idx; +}; + +/* Update CMDQ producer index (16 bytes) */ +struct cmdq_update { + __le64 reserved64; + __le32 reserved32; + __le32 prod_idx; +}; + +/* CMDQ common header structure (16 bytes) */ +struct cmdq_base { + u8 opcode; + #define CMDQ_BASE_OPCODE_CREATE_QP 0x1UL + #define CMDQ_BASE_OPCODE_DESTROY_QP 0x2UL + #define CMDQ_BASE_OPCODE_MODIFY_QP 0x3UL + #define CMDQ_BASE_OPCODE_QUERY_QP 0x4UL + #define CMDQ_BASE_OPCODE_CREATE_SRQ 0x5UL + #define CMDQ_BASE_OPCODE_DESTROY_SRQ 0x6UL + #define CMDQ_BASE_OPCODE_QUERY_SRQ 0x8UL + #define CMDQ_BASE_OPCODE_CREATE_CQ 0x9UL + #define CMDQ_BASE_OPCODE_DESTROY_CQ 0xaUL + #define CMDQ_BASE_OPCODE_RESIZE_CQ 0xcUL + #define CMDQ_BASE_OPCODE_ALLOCATE_MRW 0xdUL + #define CMDQ_BASE_OPCODE_DEALLOCATE_KEY 0xeUL + #define CMDQ_BASE_OPCODE_REGISTER_MR 0xfUL + #define CMDQ_BASE_OPCODE_DEREGISTER_MR 0x10UL + #define CMDQ_BASE_OPCODE_ADD_GID 0x11UL + #define CMDQ_BASE_OPCODE_DELETE_GID 0x12UL + #define CMDQ_BASE_OPCODE_MODIFY_GID 0x17UL + #define CMDQ_BASE_OPCODE_QUERY_GID 0x18UL + #define CMDQ_BASE_OPCODE_CREATE_QP1 0x13UL + #define CMDQ_BASE_OPCODE_DESTROY_QP1 0x14UL + #define CMDQ_BASE_OPCODE_CREATE_AH 0x15UL + #define CMDQ_BASE_OPCODE_DESTROY_AH 0x16UL + #define CMDQ_BASE_OPCODE_INITIALIZE_FW 0x80UL + #define CMDQ_BASE_OPCODE_DEINITIALIZE_FW 0x81UL + #define CMDQ_BASE_OPCODE_STOP_FUNC 0x82UL + #define CMDQ_BASE_OPCODE_QUERY_FUNC 0x83UL + #define CMDQ_BASE_OPCODE_SET_FUNC_RESOURCES 0x84UL + #define CMDQ_BASE_OPCODE_READ_CONTEXT 0x85UL + #define CMDQ_BASE_OPCODE_VF_BACKCHANNEL_REQUEST 0x86UL + #define CMDQ_BASE_OPCODE_READ_VF_MEMORY 0x87UL + #define CMDQ_BASE_OPCODE_COMPLETE_VF_REQUEST 0x88UL + #define CMDQ_BASE_OPCODE_EXTEND_CONTEXT_ARRRAY 0x89UL + #define CMDQ_BASE_OPCODE_MAP_TC_TO_COS 0x8aUL + #define CMDQ_BASE_OPCODE_QUERY_VERSION 0x8bUL + #define CMDQ_BASE_OPCODE_MODIFY_CC 0x8cUL + #define CMDQ_BASE_OPCODE_QUERY_CC 0x8dUL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; +}; + +/* Create QP command (96 bytes) */ +struct cmdq_create_qp { + u8 opcode; + #define CMDQ_CREATE_QP_OPCODE_CREATE_QP 0x1UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le64 qp_handle; + __le32 qp_flags; + #define CMDQ_CREATE_QP_QP_FLAGS_SRQ_USED 0x1UL + #define CMDQ_CREATE_QP_QP_FLAGS_FORCE_COMPLETION 0x2UL + #define CMDQ_CREATE_QP_QP_FLAGS_RESERVED_LKEY_ENABLE 0x4UL + #define CMDQ_CREATE_QP_QP_FLAGS_FR_PMR_ENABLED 0x8UL + u8 type; + #define CMDQ_CREATE_QP_TYPE_RC 0x2UL + #define CMDQ_CREATE_QP_TYPE_UD 0x4UL + #define CMDQ_CREATE_QP_TYPE_RAW_ETHERTYPE 0x6UL + u8 sq_pg_size_sq_lvl; + #define CMDQ_CREATE_QP_SQ_LVL_MASK 0xfUL + #define CMDQ_CREATE_QP_SQ_LVL_SFT 0 + #define CMDQ_CREATE_QP_SQ_LVL_LVL_0 0x0UL + #define CMDQ_CREATE_QP_SQ_LVL_LVL_1 0x1UL + #define CMDQ_CREATE_QP_SQ_LVL_LVL_2 0x2UL + #define CMDQ_CREATE_QP_SQ_PG_SIZE_MASK 0xf0UL + #define CMDQ_CREATE_QP_SQ_PG_SIZE_SFT 4 + #define CMDQ_CREATE_QP_SQ_PG_SIZE_PG_4K (0x0UL << 4) + #define CMDQ_CREATE_QP_SQ_PG_SIZE_PG_8K (0x1UL << 4) + #define CMDQ_CREATE_QP_SQ_PG_SIZE_PG_64K (0x2UL << 4) + #define CMDQ_CREATE_QP_SQ_PG_SIZE_PG_2M (0x3UL << 4) + #define CMDQ_CREATE_QP_SQ_PG_SIZE_PG_8M (0x4UL << 4) + #define CMDQ_CREATE_QP_SQ_PG_SIZE_PG_1G (0x5UL << 4) + u8 rq_pg_size_rq_lvl; + #define CMDQ_CREATE_QP_RQ_LVL_MASK 0xfUL + #define CMDQ_CREATE_QP_RQ_LVL_SFT 0 + #define CMDQ_CREATE_QP_RQ_LVL_LVL_0 0x0UL + #define CMDQ_CREATE_QP_RQ_LVL_LVL_1 0x1UL + #define CMDQ_CREATE_QP_RQ_LVL_LVL_2 0x2UL + #define CMDQ_CREATE_QP_RQ_PG_SIZE_MASK 0xf0UL + #define CMDQ_CREATE_QP_RQ_PG_SIZE_SFT 4 + #define CMDQ_CREATE_QP_RQ_PG_SIZE_PG_4K (0x0UL << 4) + #define CMDQ_CREATE_QP_RQ_PG_SIZE_PG_8K (0x1UL << 4) + #define CMDQ_CREATE_QP_RQ_PG_SIZE_PG_64K (0x2UL << 4) + #define CMDQ_CREATE_QP_RQ_PG_SIZE_PG_2M (0x3UL << 4) + #define CMDQ_CREATE_QP_RQ_PG_SIZE_PG_8M (0x4UL << 4) + #define CMDQ_CREATE_QP_RQ_PG_SIZE_PG_1G (0x5UL << 4) + u8 unused_0; + __le32 dpi; + __le32 sq_size; + __le32 rq_size; + __le16 sq_fwo_sq_sge; + #define CMDQ_CREATE_QP_SQ_SGE_MASK 0xfUL + #define CMDQ_CREATE_QP_SQ_SGE_SFT 0 + #define CMDQ_CREATE_QP_SQ_FWO_MASK 0xfff0UL + #define CMDQ_CREATE_QP_SQ_FWO_SFT 4 + __le16 rq_fwo_rq_sge; + #define CMDQ_CREATE_QP_RQ_SGE_MASK 0xfUL + #define CMDQ_CREATE_QP_RQ_SGE_SFT 0 + #define CMDQ_CREATE_QP_RQ_FWO_MASK 0xfff0UL + #define CMDQ_CREATE_QP_RQ_FWO_SFT 4 + __le32 scq_cid; + __le32 rcq_cid; + __le32 srq_cid; + __le32 pd_id; + __le64 sq_pbl; + __le64 rq_pbl; + __le64 irrq_addr; + __le64 orrq_addr; +}; + +/* Destroy QP command (24 bytes) */ +struct cmdq_destroy_qp { + u8 opcode; + #define CMDQ_DESTROY_QP_OPCODE_DESTROY_QP 0x2UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 qp_cid; + __le32 unused_0; +}; + +/* Modify QP command (112 bytes) */ +struct cmdq_modify_qp { + u8 opcode; + #define CMDQ_MODIFY_QP_OPCODE_MODIFY_QP 0x3UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 modify_mask; + #define CMDQ_MODIFY_QP_MODIFY_MASK_STATE 0x1UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_EN_SQD_ASYNC_NOTIFY 0x2UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_ACCESS 0x4UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_PKEY 0x8UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_QKEY 0x10UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_DGID 0x20UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_FLOW_LABEL 0x40UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_SGID_INDEX 0x80UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_HOP_LIMIT 0x100UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_TRAFFIC_CLASS 0x200UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_DEST_MAC 0x400UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_PATH_MTU 0x1000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_TIMEOUT 0x2000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_RETRY_CNT 0x4000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_RNR_RETRY 0x8000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_RQ_PSN 0x10000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_MAX_RD_ATOMIC 0x20000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_MIN_RNR_TIMER 0x40000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_SQ_PSN 0x80000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_MAX_DEST_RD_ATOMIC 0x100000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_SQ_SIZE 0x200000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_RQ_SIZE 0x400000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_SQ_SGE 0x800000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_RQ_SGE 0x1000000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_MAX_INLINE_DATA 0x2000000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_DEST_QP_ID 0x4000000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_SRC_MAC 0x8000000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_VLAN_ID 0x10000000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_ENABLE_CC 0x20000000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_TOS_ECN 0x40000000UL + #define CMDQ_MODIFY_QP_MODIFY_MASK_TOS_DSCP 0x80000000UL + __le32 qp_cid; + u8 network_type_en_sqd_async_notify_new_state; + #define CMDQ_MODIFY_QP_NEW_STATE_MASK 0xfUL + #define CMDQ_MODIFY_QP_NEW_STATE_SFT 0 + #define CMDQ_MODIFY_QP_NEW_STATE_RESET 0x0UL + #define CMDQ_MODIFY_QP_NEW_STATE_INIT 0x1UL + #define CMDQ_MODIFY_QP_NEW_STATE_RTR 0x2UL + #define CMDQ_MODIFY_QP_NEW_STATE_RTS 0x3UL + #define CMDQ_MODIFY_QP_NEW_STATE_SQD 0x4UL + #define CMDQ_MODIFY_QP_NEW_STATE_SQE 0x5UL + #define CMDQ_MODIFY_QP_NEW_STATE_ERR 0x6UL + #define CMDQ_MODIFY_QP_EN_SQD_ASYNC_NOTIFY 0x10UL + #define CMDQ_MODIFY_QP_NETWORK_TYPE_MASK 0xc0UL + #define CMDQ_MODIFY_QP_NETWORK_TYPE_SFT 6 + #define CMDQ_MODIFY_QP_NETWORK_TYPE_ROCEV1 (0x0UL << 6) + #define CMDQ_MODIFY_QP_NETWORK_TYPE_ROCEV2_IPV4 (0x2UL << 6) + #define CMDQ_MODIFY_QP_NETWORK_TYPE_ROCEV2_IPV6 (0x3UL << 6) + u8 access; + #define CMDQ_MODIFY_QP_ACCESS_LOCAL_WRITE 0x1UL + #define CMDQ_MODIFY_QP_ACCESS_REMOTE_WRITE 0x2UL + #define CMDQ_MODIFY_QP_ACCESS_REMOTE_READ 0x4UL + #define CMDQ_MODIFY_QP_ACCESS_REMOTE_ATOMIC 0x8UL + __le16 pkey; + __le32 qkey; + __le32 dgid[4]; + __le32 flow_label; + __le16 sgid_index; + u8 hop_limit; + u8 traffic_class; + __le16 dest_mac[3]; + u8 tos_dscp_tos_ecn; + #define CMDQ_MODIFY_QP_TOS_ECN_MASK 0x3UL + #define CMDQ_MODIFY_QP_TOS_ECN_SFT 0 + #define CMDQ_MODIFY_QP_TOS_DSCP_MASK 0xfcUL + #define CMDQ_MODIFY_QP_TOS_DSCP_SFT 2 + u8 path_mtu; + #define CMDQ_MODIFY_QP_PATH_MTU_MASK 0xf0UL + #define CMDQ_MODIFY_QP_PATH_MTU_SFT 4 + #define CMDQ_MODIFY_QP_PATH_MTU_MTU_256 (0x0UL << 4) + #define CMDQ_MODIFY_QP_PATH_MTU_MTU_512 (0x1UL << 4) + #define CMDQ_MODIFY_QP_PATH_MTU_MTU_1024 (0x2UL << 4) + #define CMDQ_MODIFY_QP_PATH_MTU_MTU_2048 (0x3UL << 4) + #define CMDQ_MODIFY_QP_PATH_MTU_MTU_4096 (0x4UL << 4) + #define CMDQ_MODIFY_QP_PATH_MTU_MTU_8192 (0x5UL << 4) + u8 timeout; + u8 retry_cnt; + u8 rnr_retry; + u8 min_rnr_timer; + __le32 rq_psn; + __le32 sq_psn; + u8 max_rd_atomic; + u8 max_dest_rd_atomic; + __le16 enable_cc; + #define CMDQ_MODIFY_QP_ENABLE_CC 0x1UL + __le32 sq_size; + __le32 rq_size; + __le16 sq_sge; + __le16 rq_sge; + __le32 max_inline_data; + __le32 dest_qp_id; + __le32 unused_3; + __le16 src_mac[3]; + __le16 vlan_pcp_vlan_dei_vlan_id; + #define CMDQ_MODIFY_QP_VLAN_ID_MASK 0xfffUL + #define CMDQ_MODIFY_QP_VLAN_ID_SFT 0 + #define CMDQ_MODIFY_QP_VLAN_DEI 0x1000UL + #define CMDQ_MODIFY_QP_VLAN_PCP_MASK 0xe000UL + #define CMDQ_MODIFY_QP_VLAN_PCP_SFT 13 +}; + +/* Query QP command (24 bytes) */ +struct cmdq_query_qp { + u8 opcode; + #define CMDQ_QUERY_QP_OPCODE_QUERY_QP 0x4UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 qp_cid; + __le32 unused_0; +}; + +/* Create SRQ command (48 bytes) */ +struct cmdq_create_srq { + u8 opcode; + #define CMDQ_CREATE_SRQ_OPCODE_CREATE_SRQ 0x5UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le64 srq_handle; + __le16 pg_size_lvl; + #define CMDQ_CREATE_SRQ_LVL_MASK 0x3UL + #define CMDQ_CREATE_SRQ_LVL_SFT 0 + #define CMDQ_CREATE_SRQ_LVL_LVL_0 0x0UL + #define CMDQ_CREATE_SRQ_LVL_LVL_1 0x1UL + #define CMDQ_CREATE_SRQ_LVL_LVL_2 0x2UL + #define CMDQ_CREATE_SRQ_PG_SIZE_MASK 0x1cUL + #define CMDQ_CREATE_SRQ_PG_SIZE_SFT 2 + #define CMDQ_CREATE_SRQ_PG_SIZE_PG_4K (0x0UL << 2) + #define CMDQ_CREATE_SRQ_PG_SIZE_PG_8K (0x1UL << 2) + #define CMDQ_CREATE_SRQ_PG_SIZE_PG_64K (0x2UL << 2) + #define CMDQ_CREATE_SRQ_PG_SIZE_PG_2M (0x3UL << 2) + #define CMDQ_CREATE_SRQ_PG_SIZE_PG_8M (0x4UL << 2) + #define CMDQ_CREATE_SRQ_PG_SIZE_PG_1G (0x5UL << 2) + __le16 eventq_id; + #define CMDQ_CREATE_SRQ_EVENTQ_ID_MASK 0xfffUL + #define CMDQ_CREATE_SRQ_EVENTQ_ID_SFT 0 + __le16 srq_size; + __le16 srq_fwo; + __le32 dpi; + __le32 pd_id; + __le64 pbl; +}; + +/* Destroy SRQ command (24 bytes) */ +struct cmdq_destroy_srq { + u8 opcode; + #define CMDQ_DESTROY_SRQ_OPCODE_DESTROY_SRQ 0x6UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 srq_cid; + __le32 unused_0; +}; + +/* Query SRQ command (24 bytes) */ +struct cmdq_query_srq { + u8 opcode; + #define CMDQ_QUERY_SRQ_OPCODE_QUERY_SRQ 0x8UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 srq_cid; + __le32 unused_0; +}; + +/* Create CQ command (48 bytes) */ +struct cmdq_create_cq { + u8 opcode; + #define CMDQ_CREATE_CQ_OPCODE_CREATE_CQ 0x9UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le64 cq_handle; + __le32 pg_size_lvl; + #define CMDQ_CREATE_CQ_LVL_MASK 0x3UL + #define CMDQ_CREATE_CQ_LVL_SFT 0 + #define CMDQ_CREATE_CQ_LVL_LVL_0 0x0UL + #define CMDQ_CREATE_CQ_LVL_LVL_1 0x1UL + #define CMDQ_CREATE_CQ_LVL_LVL_2 0x2UL + #define CMDQ_CREATE_CQ_PG_SIZE_MASK 0x1cUL + #define CMDQ_CREATE_CQ_PG_SIZE_SFT 2 + #define CMDQ_CREATE_CQ_PG_SIZE_PG_4K (0x0UL << 2) + #define CMDQ_CREATE_CQ_PG_SIZE_PG_8K (0x1UL << 2) + #define CMDQ_CREATE_CQ_PG_SIZE_PG_64K (0x2UL << 2) + #define CMDQ_CREATE_CQ_PG_SIZE_PG_2M (0x3UL << 2) + #define CMDQ_CREATE_CQ_PG_SIZE_PG_8M (0x4UL << 2) + #define CMDQ_CREATE_CQ_PG_SIZE_PG_1G (0x5UL << 2) + __le32 cq_fco_cnq_id; + #define CMDQ_CREATE_CQ_CNQ_ID_MASK 0xfffUL + #define CMDQ_CREATE_CQ_CNQ_ID_SFT 0 + #define CMDQ_CREATE_CQ_CQ_FCO_MASK 0xfffff000UL + #define CMDQ_CREATE_CQ_CQ_FCO_SFT 12 + __le32 dpi; + __le32 cq_size; + __le64 pbl; +}; + +/* Destroy CQ command (24 bytes) */ +struct cmdq_destroy_cq { + u8 opcode; + #define CMDQ_DESTROY_CQ_OPCODE_DESTROY_CQ 0xaUL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 cq_cid; + __le32 unused_0; +}; + +/* Resize CQ command (40 bytes) */ +struct cmdq_resize_cq { + u8 opcode; + #define CMDQ_RESIZE_CQ_OPCODE_RESIZE_CQ 0xcUL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 cq_cid; + __le32 new_cq_size_pg_size_lvl; + #define CMDQ_RESIZE_CQ_LVL_MASK 0x3UL + #define CMDQ_RESIZE_CQ_LVL_SFT 0 + #define CMDQ_RESIZE_CQ_LVL_LVL_0 0x0UL + #define CMDQ_RESIZE_CQ_LVL_LVL_1 0x1UL + #define CMDQ_RESIZE_CQ_LVL_LVL_2 0x2UL + #define CMDQ_RESIZE_CQ_PG_SIZE_MASK 0x1cUL + #define CMDQ_RESIZE_CQ_PG_SIZE_SFT 2 + #define CMDQ_RESIZE_CQ_PG_SIZE_PG_4K (0x0UL << 2) + #define CMDQ_RESIZE_CQ_PG_SIZE_PG_8K (0x1UL << 2) + #define CMDQ_RESIZE_CQ_PG_SIZE_PG_64K (0x2UL << 2) + #define CMDQ_RESIZE_CQ_PG_SIZE_PG_2M (0x3UL << 2) + #define CMDQ_RESIZE_CQ_PG_SIZE_PG_8M (0x4UL << 2) + #define CMDQ_RESIZE_CQ_PG_SIZE_PG_1G (0x5UL << 2) + #define CMDQ_RESIZE_CQ_NEW_CQ_SIZE_MASK 0x1fffe0UL + #define CMDQ_RESIZE_CQ_NEW_CQ_SIZE_SFT 5 + __le64 new_pbl; + __le32 new_cq_fco; + __le32 unused_2; +}; + +/* Allocate MRW command (32 bytes) */ +struct cmdq_allocate_mrw { + u8 opcode; + #define CMDQ_ALLOCATE_MRW_OPCODE_ALLOCATE_MRW 0xdUL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le64 mrw_handle; + u8 mrw_flags; + #define CMDQ_ALLOCATE_MRW_MRW_FLAGS_MASK 0xfUL + #define CMDQ_ALLOCATE_MRW_MRW_FLAGS_SFT 0 + #define CMDQ_ALLOCATE_MRW_MRW_FLAGS_MR 0x0UL + #define CMDQ_ALLOCATE_MRW_MRW_FLAGS_PMR 0x1UL + #define CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE1 0x2UL + #define CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE2A 0x3UL + #define CMDQ_ALLOCATE_MRW_MRW_FLAGS_MW_TYPE2B 0x4UL + u8 access; + #define CMDQ_ALLOCATE_MRW_ACCESS_RESERVED_MASK 0x1fUL + #define CMDQ_ALLOCATE_MRW_ACCESS_RESERVED_SFT 0 + #define CMDQ_ALLOCATE_MRW_ACCESS_CONSUMER_OWNED_KEY 0x20UL + __le16 unused_1; + __le32 pd_id; +}; + +/* De-allocate key command (24 bytes) */ +struct cmdq_deallocate_key { + u8 opcode; + #define CMDQ_DEALLOCATE_KEY_OPCODE_DEALLOCATE_KEY 0xeUL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + u8 mrw_flags; + #define CMDQ_DEALLOCATE_KEY_MRW_FLAGS_MASK 0xfUL + #define CMDQ_DEALLOCATE_KEY_MRW_FLAGS_SFT 0 + #define CMDQ_DEALLOCATE_KEY_MRW_FLAGS_MR 0x0UL + #define CMDQ_DEALLOCATE_KEY_MRW_FLAGS_PMR 0x1UL + #define CMDQ_DEALLOCATE_KEY_MRW_FLAGS_MW_TYPE1 0x2UL + #define CMDQ_DEALLOCATE_KEY_MRW_FLAGS_MW_TYPE2A 0x3UL + #define CMDQ_DEALLOCATE_KEY_MRW_FLAGS_MW_TYPE2B 0x4UL + u8 unused_1[3]; + __le32 key; +}; + +/* Register MR command (48 bytes) */ +struct cmdq_register_mr { + u8 opcode; + #define CMDQ_REGISTER_MR_OPCODE_REGISTER_MR 0xfUL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + u8 log2_pg_size_lvl; + #define CMDQ_REGISTER_MR_LVL_MASK 0x3UL + #define CMDQ_REGISTER_MR_LVL_SFT 0 + #define CMDQ_REGISTER_MR_LVL_LVL_0 0x0UL + #define CMDQ_REGISTER_MR_LVL_LVL_1 0x1UL + #define CMDQ_REGISTER_MR_LVL_LVL_2 0x2UL + #define CMDQ_REGISTER_MR_LOG2_PG_SIZE_MASK 0x7cUL + #define CMDQ_REGISTER_MR_LOG2_PG_SIZE_SFT 2 + u8 access; + #define CMDQ_REGISTER_MR_ACCESS_LOCAL_WRITE 0x1UL + #define CMDQ_REGISTER_MR_ACCESS_REMOTE_READ 0x2UL + #define CMDQ_REGISTER_MR_ACCESS_REMOTE_WRITE 0x4UL + #define CMDQ_REGISTER_MR_ACCESS_REMOTE_ATOMIC 0x8UL + #define CMDQ_REGISTER_MR_ACCESS_MW_BIND 0x10UL + #define CMDQ_REGISTER_MR_ACCESS_ZERO_BASED 0x20UL + __le16 unused_1; + __le32 key; + __le64 pbl; + __le64 va; + __le64 mr_size; +}; + +/* Deregister MR command (24 bytes) */ +struct cmdq_deregister_mr { + u8 opcode; + #define CMDQ_DEREGISTER_MR_OPCODE_DEREGISTER_MR 0x10UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 lkey; + __le32 unused_0; +}; + +/* Add GID command (48 bytes) */ +struct cmdq_add_gid { + u8 opcode; + #define CMDQ_ADD_GID_OPCODE_ADD_GID 0x11UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __be32 gid[4]; + __be16 src_mac[3]; + __le16 vlan; + #define CMDQ_ADD_GID_VLAN_VLAN_ID_MASK 0xfffUL + #define CMDQ_ADD_GID_VLAN_VLAN_ID_SFT 0 + #define CMDQ_ADD_GID_VLAN_TPID_MASK 0x7000UL + #define CMDQ_ADD_GID_VLAN_TPID_SFT 12 + #define CMDQ_ADD_GID_VLAN_TPID_TPID_88A8 (0x0UL << 12) + #define CMDQ_ADD_GID_VLAN_TPID_TPID_8100 (0x1UL << 12) + #define CMDQ_ADD_GID_VLAN_TPID_TPID_9100 (0x2UL << 12) + #define CMDQ_ADD_GID_VLAN_TPID_TPID_9200 (0x3UL << 12) + #define CMDQ_ADD_GID_VLAN_TPID_TPID_9300 (0x4UL << 12) + #define CMDQ_ADD_GID_VLAN_TPID_TPID_CFG1 (0x5UL << 12) + #define CMDQ_ADD_GID_VLAN_TPID_TPID_CFG2 (0x6UL << 12) + #define CMDQ_ADD_GID_VLAN_TPID_TPID_CFG3 (0x7UL << 12) + #define CMDQ_ADD_GID_VLAN_TPID_LAST CMDQ_ADD_GID_VLAN_TPID_TPID_CFG3 + #define CMDQ_ADD_GID_VLAN_VLAN_EN 0x8000UL + __le16 ipid; + __le16 stats_ctx; + #define CMDQ_ADD_GID_STATS_CTX_STATS_CTX_ID_MASK 0x7fffUL + #define CMDQ_ADD_GID_STATS_CTX_STATS_CTX_ID_SFT 0 + #define CMDQ_ADD_GID_STATS_CTX_STATS_CTX_VALID 0x8000UL + __le32 unused_0; +}; + +/* Delete GID command (24 bytes) */ +struct cmdq_delete_gid { + u8 opcode; + #define CMDQ_DELETE_GID_OPCODE_DELETE_GID 0x12UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le16 gid_index; + __le16 unused_0; + __le32 unused_1; +}; + +/* Modify GID command (48 bytes) */ +struct cmdq_modify_gid { + u8 opcode; + #define CMDQ_MODIFY_GID_OPCODE_MODIFY_GID 0x17UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 gid[4]; + __le16 src_mac[3]; + __le16 vlan; + #define CMDQ_MODIFY_GID_VLAN_VLAN_ID_MASK 0xfffUL + #define CMDQ_MODIFY_GID_VLAN_VLAN_ID_SFT 0 + #define CMDQ_MODIFY_GID_VLAN_TPID_MASK 0x7000UL + #define CMDQ_MODIFY_GID_VLAN_TPID_SFT 12 + #define CMDQ_MODIFY_GID_VLAN_TPID_TPID_88A8 (0x0UL << 12) + #define CMDQ_MODIFY_GID_VLAN_TPID_TPID_8100 (0x1UL << 12) + #define CMDQ_MODIFY_GID_VLAN_TPID_TPID_9100 (0x2UL << 12) + #define CMDQ_MODIFY_GID_VLAN_TPID_TPID_9200 (0x3UL << 12) + #define CMDQ_MODIFY_GID_VLAN_TPID_TPID_9300 (0x4UL << 12) + #define CMDQ_MODIFY_GID_VLAN_TPID_TPID_CFG1 (0x5UL << 12) + #define CMDQ_MODIFY_GID_VLAN_TPID_TPID_CFG2 (0x6UL << 12) + #define CMDQ_MODIFY_GID_VLAN_TPID_TPID_CFG3 (0x7UL << 12) + #define CMDQ_MODIFY_GID_VLAN_TPID_LAST \ + CMDQ_MODIFY_GID_VLAN_TPID_TPID_CFG3 + #define CMDQ_MODIFY_GID_VLAN_VLAN_EN 0x8000UL + __le16 ipid; + __le16 gid_index; + __le16 stats_ctx; + #define CMDQ_MODIFY_GID_STATS_CTX_STATS_CTX_ID_MASK 0x7fffUL + #define CMDQ_MODIFY_GID_STATS_CTX_STATS_CTX_ID_SFT 0 + #define CMDQ_MODIFY_GID_STATS_CTX_STATS_CTX_VALID 0x8000UL + __le16 unused_0; +}; + +/* Query GID command (24 bytes) */ +struct cmdq_query_gid { + u8 opcode; + #define CMDQ_QUERY_GID_OPCODE_QUERY_GID 0x18UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le16 gid_index; + __le16 unused_0; + __le32 unused_1; +}; + +/* Create QP1 command (80 bytes) */ +struct cmdq_create_qp1 { + u8 opcode; + #define CMDQ_CREATE_QP1_OPCODE_CREATE_QP1 0x13UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le64 qp_handle; + __le32 qp_flags; + #define CMDQ_CREATE_QP1_QP_FLAGS_SRQ_USED 0x1UL + #define CMDQ_CREATE_QP1_QP_FLAGS_FORCE_COMPLETION 0x2UL + #define CMDQ_CREATE_QP1_QP_FLAGS_RESERVED_LKEY_ENABLE 0x4UL + u8 type; + #define CMDQ_CREATE_QP1_TYPE_GSI 0x1UL + u8 sq_pg_size_sq_lvl; + #define CMDQ_CREATE_QP1_SQ_LVL_MASK 0xfUL + #define CMDQ_CREATE_QP1_SQ_LVL_SFT 0 + #define CMDQ_CREATE_QP1_SQ_LVL_LVL_0 0x0UL + #define CMDQ_CREATE_QP1_SQ_LVL_LVL_1 0x1UL + #define CMDQ_CREATE_QP1_SQ_LVL_LVL_2 0x2UL + #define CMDQ_CREATE_QP1_SQ_PG_SIZE_MASK 0xf0UL + #define CMDQ_CREATE_QP1_SQ_PG_SIZE_SFT 4 + #define CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_4K (0x0UL << 4) + #define CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_8K (0x1UL << 4) + #define CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_64K (0x2UL << 4) + #define CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_2M (0x3UL << 4) + #define CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_8M (0x4UL << 4) + #define CMDQ_CREATE_QP1_SQ_PG_SIZE_PG_1G (0x5UL << 4) + u8 rq_pg_size_rq_lvl; + #define CMDQ_CREATE_QP1_RQ_LVL_MASK 0xfUL + #define CMDQ_CREATE_QP1_RQ_LVL_SFT 0 + #define CMDQ_CREATE_QP1_RQ_LVL_LVL_0 0x0UL + #define CMDQ_CREATE_QP1_RQ_LVL_LVL_1 0x1UL + #define CMDQ_CREATE_QP1_RQ_LVL_LVL_2 0x2UL + #define CMDQ_CREATE_QP1_RQ_PG_SIZE_MASK 0xf0UL + #define CMDQ_CREATE_QP1_RQ_PG_SIZE_SFT 4 + #define CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_4K (0x0UL << 4) + #define CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_8K (0x1UL << 4) + #define CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_64K (0x2UL << 4) + #define CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_2M (0x3UL << 4) + #define CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_8M (0x4UL << 4) + #define CMDQ_CREATE_QP1_RQ_PG_SIZE_PG_1G (0x5UL << 4) + u8 unused_0; + __le32 dpi; + __le32 sq_size; + __le32 rq_size; + __le16 sq_fwo_sq_sge; + #define CMDQ_CREATE_QP1_SQ_SGE_MASK 0xfUL + #define CMDQ_CREATE_QP1_SQ_SGE_SFT 0 + #define CMDQ_CREATE_QP1_SQ_FWO_MASK 0xfff0UL + #define CMDQ_CREATE_QP1_SQ_FWO_SFT 4 + __le16 rq_fwo_rq_sge; + #define CMDQ_CREATE_QP1_RQ_SGE_MASK 0xfUL + #define CMDQ_CREATE_QP1_RQ_SGE_SFT 0 + #define CMDQ_CREATE_QP1_RQ_FWO_MASK 0xfff0UL + #define CMDQ_CREATE_QP1_RQ_FWO_SFT 4 + __le32 scq_cid; + __le32 rcq_cid; + __le32 srq_cid; + __le32 pd_id; + __le64 sq_pbl; + __le64 rq_pbl; +}; + +/* Destroy QP1 command (24 bytes) */ +struct cmdq_destroy_qp1 { + u8 opcode; + #define CMDQ_DESTROY_QP1_OPCODE_DESTROY_QP1 0x14UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 qp1_cid; + __le32 unused_0; +}; + +/* Create AH command (64 bytes) */ +struct cmdq_create_ah { + u8 opcode; + #define CMDQ_CREATE_AH_OPCODE_CREATE_AH 0x15UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le64 ah_handle; + __le32 dgid[4]; + u8 type; + #define CMDQ_CREATE_AH_TYPE_V1 0x0UL + #define CMDQ_CREATE_AH_TYPE_V2IPV4 0x2UL + #define CMDQ_CREATE_AH_TYPE_V2IPV6 0x3UL + u8 hop_limit; + __le16 sgid_index; + __le32 dest_vlan_id_flow_label; + #define CMDQ_CREATE_AH_FLOW_LABEL_MASK 0xfffffUL + #define CMDQ_CREATE_AH_FLOW_LABEL_SFT 0 + #define CMDQ_CREATE_AH_DEST_VLAN_ID_MASK 0xfff00000UL + #define CMDQ_CREATE_AH_DEST_VLAN_ID_SFT 20 + __le32 pd_id; + __le32 unused_0; + __le16 dest_mac[3]; + u8 traffic_class; + u8 unused_1; +}; + +/* Destroy AH command (24 bytes) */ +struct cmdq_destroy_ah { + u8 opcode; + #define CMDQ_DESTROY_AH_OPCODE_DESTROY_AH 0x16UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 ah_cid; + __le32 unused_0; +}; + +/* Initialize Firmware command (112 bytes) */ +struct cmdq_initialize_fw { + u8 opcode; + #define CMDQ_INITIALIZE_FW_OPCODE_INITIALIZE_FW 0x80UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + u8 qpc_pg_size_qpc_lvl; + #define CMDQ_INITIALIZE_FW_QPC_LVL_MASK 0xfUL + #define CMDQ_INITIALIZE_FW_QPC_LVL_SFT 0 + #define CMDQ_INITIALIZE_FW_QPC_LVL_LVL_0 0x0UL + #define CMDQ_INITIALIZE_FW_QPC_LVL_LVL_1 0x1UL + #define CMDQ_INITIALIZE_FW_QPC_LVL_LVL_2 0x2UL + #define CMDQ_INITIALIZE_FW_QPC_PG_SIZE_MASK 0xf0UL + #define CMDQ_INITIALIZE_FW_QPC_PG_SIZE_SFT 4 + #define CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_4K (0x0UL << 4) + #define CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_8K (0x1UL << 4) + #define CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_64K (0x2UL << 4) + #define CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_2M (0x3UL << 4) + #define CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_8M (0x4UL << 4) + #define CMDQ_INITIALIZE_FW_QPC_PG_SIZE_PG_1G (0x5UL << 4) + u8 mrw_pg_size_mrw_lvl; + #define CMDQ_INITIALIZE_FW_MRW_LVL_MASK 0xfUL + #define CMDQ_INITIALIZE_FW_MRW_LVL_SFT 0 + #define CMDQ_INITIALIZE_FW_MRW_LVL_LVL_0 0x0UL + #define CMDQ_INITIALIZE_FW_MRW_LVL_LVL_1 0x1UL + #define CMDQ_INITIALIZE_FW_MRW_LVL_LVL_2 0x2UL + #define CMDQ_INITIALIZE_FW_MRW_PG_SIZE_MASK 0xf0UL + #define CMDQ_INITIALIZE_FW_MRW_PG_SIZE_SFT 4 + #define CMDQ_INITIALIZE_FW_MRW_PG_SIZE_PG_4K (0x0UL << 4) + #define CMDQ_INITIALIZE_FW_MRW_PG_SIZE_PG_8K (0x1UL << 4) + #define CMDQ_INITIALIZE_FW_MRW_PG_SIZE_PG_64K (0x2UL << 4) + #define CMDQ_INITIALIZE_FW_MRW_PG_SIZE_PG_2M (0x3UL << 4) + #define CMDQ_INITIALIZE_FW_MRW_PG_SIZE_PG_8M (0x4UL << 4) + #define CMDQ_INITIALIZE_FW_MRW_PG_SIZE_PG_1G (0x5UL << 4) + u8 srq_pg_size_srq_lvl; + #define CMDQ_INITIALIZE_FW_SRQ_LVL_MASK 0xfUL + #define CMDQ_INITIALIZE_FW_SRQ_LVL_SFT 0 + #define CMDQ_INITIALIZE_FW_SRQ_LVL_LVL_0 0x0UL + #define CMDQ_INITIALIZE_FW_SRQ_LVL_LVL_1 0x1UL + #define CMDQ_INITIALIZE_FW_SRQ_LVL_LVL_2 0x2UL + #define CMDQ_INITIALIZE_FW_SRQ_PG_SIZE_MASK 0xf0UL + #define CMDQ_INITIALIZE_FW_SRQ_PG_SIZE_SFT 4 + #define CMDQ_INITIALIZE_FW_SRQ_PG_SIZE_PG_4K (0x0UL << 4) + #define CMDQ_INITIALIZE_FW_SRQ_PG_SIZE_PG_8K (0x1UL << 4) + #define CMDQ_INITIALIZE_FW_SRQ_PG_SIZE_PG_64K (0x2UL << 4) + #define CMDQ_INITIALIZE_FW_SRQ_PG_SIZE_PG_2M (0x3UL << 4) + #define CMDQ_INITIALIZE_FW_SRQ_PG_SIZE_PG_8M (0x4UL << 4) + #define CMDQ_INITIALIZE_FW_SRQ_PG_SIZE_PG_1G (0x5UL << 4) + u8 cq_pg_size_cq_lvl; + #define CMDQ_INITIALIZE_FW_CQ_LVL_MASK 0xfUL + #define CMDQ_INITIALIZE_FW_CQ_LVL_SFT 0 + #define CMDQ_INITIALIZE_FW_CQ_LVL_LVL_0 0x0UL + #define CMDQ_INITIALIZE_FW_CQ_LVL_LVL_1 0x1UL + #define CMDQ_INITIALIZE_FW_CQ_LVL_LVL_2 0x2UL + #define CMDQ_INITIALIZE_FW_CQ_PG_SIZE_MASK 0xf0UL + #define CMDQ_INITIALIZE_FW_CQ_PG_SIZE_SFT 4 + #define CMDQ_INITIALIZE_FW_CQ_PG_SIZE_PG_4K (0x0UL << 4) + #define CMDQ_INITIALIZE_FW_CQ_PG_SIZE_PG_8K (0x1UL << 4) + #define CMDQ_INITIALIZE_FW_CQ_PG_SIZE_PG_64K (0x2UL << 4) + #define CMDQ_INITIALIZE_FW_CQ_PG_SIZE_PG_2M (0x3UL << 4) + #define CMDQ_INITIALIZE_FW_CQ_PG_SIZE_PG_8M (0x4UL << 4) + #define CMDQ_INITIALIZE_FW_CQ_PG_SIZE_PG_1G (0x5UL << 4) + u8 tqm_pg_size_tqm_lvl; + #define CMDQ_INITIALIZE_FW_TQM_LVL_MASK 0xfUL + #define CMDQ_INITIALIZE_FW_TQM_LVL_SFT 0 + #define CMDQ_INITIALIZE_FW_TQM_LVL_LVL_0 0x0UL + #define CMDQ_INITIALIZE_FW_TQM_LVL_LVL_1 0x1UL + #define CMDQ_INITIALIZE_FW_TQM_LVL_LVL_2 0x2UL + #define CMDQ_INITIALIZE_FW_TQM_PG_SIZE_MASK 0xf0UL + #define CMDQ_INITIALIZE_FW_TQM_PG_SIZE_SFT 4 + #define CMDQ_INITIALIZE_FW_TQM_PG_SIZE_PG_4K (0x0UL << 4) + #define CMDQ_INITIALIZE_FW_TQM_PG_SIZE_PG_8K (0x1UL << 4) + #define CMDQ_INITIALIZE_FW_TQM_PG_SIZE_PG_64K (0x2UL << 4) + #define CMDQ_INITIALIZE_FW_TQM_PG_SIZE_PG_2M (0x3UL << 4) + #define CMDQ_INITIALIZE_FW_TQM_PG_SIZE_PG_8M (0x4UL << 4) + #define CMDQ_INITIALIZE_FW_TQM_PG_SIZE_PG_1G (0x5UL << 4) + u8 tim_pg_size_tim_lvl; + #define CMDQ_INITIALIZE_FW_TIM_LVL_MASK 0xfUL + #define CMDQ_INITIALIZE_FW_TIM_LVL_SFT 0 + #define CMDQ_INITIALIZE_FW_TIM_LVL_LVL_0 0x0UL + #define CMDQ_INITIALIZE_FW_TIM_LVL_LVL_1 0x1UL + #define CMDQ_INITIALIZE_FW_TIM_LVL_LVL_2 0x2UL + #define CMDQ_INITIALIZE_FW_TIM_PG_SIZE_MASK 0xf0UL + #define CMDQ_INITIALIZE_FW_TIM_PG_SIZE_SFT 4 + #define CMDQ_INITIALIZE_FW_TIM_PG_SIZE_PG_4K (0x0UL << 4) + #define CMDQ_INITIALIZE_FW_TIM_PG_SIZE_PG_8K (0x1UL << 4) + #define CMDQ_INITIALIZE_FW_TIM_PG_SIZE_PG_64K (0x2UL << 4) + #define CMDQ_INITIALIZE_FW_TIM_PG_SIZE_PG_2M (0x3UL << 4) + #define CMDQ_INITIALIZE_FW_TIM_PG_SIZE_PG_8M (0x4UL << 4) + #define CMDQ_INITIALIZE_FW_TIM_PG_SIZE_PG_1G (0x5UL << 4) + __le16 reserved16; + __le64 qpc_page_dir; + __le64 mrw_page_dir; + __le64 srq_page_dir; + __le64 cq_page_dir; + __le64 tqm_page_dir; + __le64 tim_page_dir; + __le32 number_of_qp; + __le32 number_of_mrw; + __le32 number_of_srq; + __le32 number_of_cq; + __le32 max_qp_per_vf; + __le32 max_mrw_per_vf; + __le32 max_srq_per_vf; + __le32 max_cq_per_vf; + __le32 max_gid_per_vf; + __le32 stat_ctx_id; +}; + +/* De-initialize Firmware command (16 bytes) */ +struct cmdq_deinitialize_fw { + u8 opcode; + #define CMDQ_DEINITIALIZE_FW_OPCODE_DEINITIALIZE_FW 0x81UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; +}; + +/* Stop function command (16 bytes) */ +struct cmdq_stop_func { + u8 opcode; + #define CMDQ_STOP_FUNC_OPCODE_STOP_FUNC 0x82UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; +}; + +/* Query function command (16 bytes) */ +struct cmdq_query_func { + u8 opcode; + #define CMDQ_QUERY_FUNC_OPCODE_QUERY_FUNC 0x83UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; +}; + +/* Set function resources command (16 bytes) */ +struct cmdq_set_func_resources { + u8 opcode; + #define CMDQ_SET_FUNC_RESOURCES_OPCODE_SET_FUNC_RESOURCES 0x84UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; +}; + +/* Read hardware resource context command (24 bytes) */ +struct cmdq_read_context { + u8 opcode; + #define CMDQ_READ_CONTEXT_OPCODE_READ_CONTEXT 0x85UL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le32 type_xid; + #define CMDQ_READ_CONTEXT_XID_MASK 0xffffffUL + #define CMDQ_READ_CONTEXT_XID_SFT 0 + #define CMDQ_READ_CONTEXT_TYPE_MASK 0xff000000UL + #define CMDQ_READ_CONTEXT_TYPE_SFT 24 + #define CMDQ_READ_CONTEXT_TYPE_QPC (0x0UL << 24) + #define CMDQ_READ_CONTEXT_TYPE_CQ (0x1UL << 24) + #define CMDQ_READ_CONTEXT_TYPE_MRW (0x2UL << 24) + #define CMDQ_READ_CONTEXT_TYPE_SRQ (0x3UL << 24) + __le32 unused_0; +}; + +/* Map TC to COS. Can only be issued from a PF (24 bytes) */ +struct cmdq_map_tc_to_cos { + u8 opcode; + #define CMDQ_MAP_TC_TO_COS_OPCODE_MAP_TC_TO_COS 0x8aUL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; + __le16 cos0; + #define CMDQ_MAP_TC_TO_COS_COS0_NO_CHANGE 0xffffUL + __le16 cos1; + #define CMDQ_MAP_TC_TO_COS_COS1_DISABLE 0x8000UL + #define CMDQ_MAP_TC_TO_COS_COS1_NO_CHANGE 0xffffUL + __le32 unused_0; +}; + +/* Query version command (16 bytes) */ +struct cmdq_query_version { + u8 opcode; + #define CMDQ_QUERY_VERSION_OPCODE_QUERY_VERSION 0x8bUL + u8 cmd_size; + __le16 flags; + __le16 cookie; + u8 resp_size; + u8 reserved8; + __le64 resp_addr; +}; + +/* Command-Response Event Queue (CREQ) Structures */ +/* Base CREQ Record (16 bytes) */ +struct creq_base { + u8 type; + #define CREQ_BASE_TYPE_MASK 0x3fUL + #define CREQ_BASE_TYPE_SFT 0 + #define CREQ_BASE_TYPE_QP_EVENT 0x38UL + #define CREQ_BASE_TYPE_FUNC_EVENT 0x3aUL + #define CREQ_BASE_RESERVED2_MASK 0xc0UL + #define CREQ_BASE_RESERVED2_SFT 6 + u8 reserved56[7]; + u8 v; + #define CREQ_BASE_V 0x1UL + #define CREQ_BASE_RESERVED7_MASK 0xfeUL + #define CREQ_BASE_RESERVED7_SFT 1 + u8 event; + __le16 reserved48[3]; +}; + +/* RoCE Function Async Event Notification (16 bytes) */ +struct creq_func_event { + u8 type; + #define CREQ_FUNC_EVENT_TYPE_MASK 0x3fUL + #define CREQ_FUNC_EVENT_TYPE_SFT 0 + #define CREQ_FUNC_EVENT_TYPE_FUNC_EVENT 0x3aUL + #define CREQ_FUNC_EVENT_RESERVED2_MASK 0xc0UL + #define CREQ_FUNC_EVENT_RESERVED2_SFT 6 + u8 reserved56[7]; + u8 v; + #define CREQ_FUNC_EVENT_V 0x1UL + #define CREQ_FUNC_EVENT_RESERVED7_MASK 0xfeUL + #define CREQ_FUNC_EVENT_RESERVED7_SFT 1 + u8 event; + #define CREQ_FUNC_EVENT_EVENT_TX_WQE_ERROR 0x1UL + #define CREQ_FUNC_EVENT_EVENT_TX_DATA_ERROR 0x2UL + #define CREQ_FUNC_EVENT_EVENT_RX_WQE_ERROR 0x3UL + #define CREQ_FUNC_EVENT_EVENT_RX_DATA_ERROR 0x4UL + #define CREQ_FUNC_EVENT_EVENT_CQ_ERROR 0x5UL + #define CREQ_FUNC_EVENT_EVENT_TQM_ERROR 0x6UL + #define CREQ_FUNC_EVENT_EVENT_CFCQ_ERROR 0x7UL + #define CREQ_FUNC_EVENT_EVENT_CFCS_ERROR 0x8UL + #define CREQ_FUNC_EVENT_EVENT_CFCC_ERROR 0x9UL + #define CREQ_FUNC_EVENT_EVENT_CFCM_ERROR 0xaUL + #define CREQ_FUNC_EVENT_EVENT_TIM_ERROR 0xbUL + #define CREQ_FUNC_EVENT_EVENT_VF_COMM_REQUEST 0x80UL + #define CREQ_FUNC_EVENT_EVENT_RESOURCE_EXHAUSTED 0x81UL + __le16 reserved48[3]; +}; + +/* RoCE Slowpath Command Completion (16 bytes) */ +struct creq_qp_event { + u8 type; + #define CREQ_QP_EVENT_TYPE_MASK 0x3fUL + #define CREQ_QP_EVENT_TYPE_SFT 0 + #define CREQ_QP_EVENT_TYPE_QP_EVENT 0x38UL + #define CREQ_QP_EVENT_RESERVED2_MASK 0xc0UL + #define CREQ_QP_EVENT_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 reserved32; + u8 v; + #define CREQ_QP_EVENT_V 0x1UL + #define CREQ_QP_EVENT_RESERVED7_MASK 0xfeUL + #define CREQ_QP_EVENT_RESERVED7_SFT 1 + u8 event; + #define CREQ_QP_EVENT_EVENT_CREATE_QP 0x1UL + #define CREQ_QP_EVENT_EVENT_DESTROY_QP 0x2UL + #define CREQ_QP_EVENT_EVENT_MODIFY_QP 0x3UL + #define CREQ_QP_EVENT_EVENT_QUERY_QP 0x4UL + #define CREQ_QP_EVENT_EVENT_CREATE_SRQ 0x5UL + #define CREQ_QP_EVENT_EVENT_DESTROY_SRQ 0x6UL + #define CREQ_QP_EVENT_EVENT_QUERY_SRQ 0x8UL + #define CREQ_QP_EVENT_EVENT_CREATE_CQ 0x9UL + #define CREQ_QP_EVENT_EVENT_DESTROY_CQ 0xaUL + #define CREQ_QP_EVENT_EVENT_RESIZE_CQ 0xcUL + #define CREQ_QP_EVENT_EVENT_ALLOCATE_MRW 0xdUL + #define CREQ_QP_EVENT_EVENT_DEALLOCATE_KEY 0xeUL + #define CREQ_QP_EVENT_EVENT_REGISTER_MR 0xfUL + #define CREQ_QP_EVENT_EVENT_DEREGISTER_MR 0x10UL + #define CREQ_QP_EVENT_EVENT_ADD_GID 0x11UL + #define CREQ_QP_EVENT_EVENT_DELETE_GID 0x12UL + #define CREQ_QP_EVENT_EVENT_MODIFY_GID 0x17UL + #define CREQ_QP_EVENT_EVENT_QUERY_GID 0x18UL + #define CREQ_QP_EVENT_EVENT_CREATE_QP1 0x13UL + #define CREQ_QP_EVENT_EVENT_DESTROY_QP1 0x14UL + #define CREQ_QP_EVENT_EVENT_CREATE_AH 0x15UL + #define CREQ_QP_EVENT_EVENT_DESTROY_AH 0x16UL + #define CREQ_QP_EVENT_EVENT_INITIALIZE_FW 0x80UL + #define CREQ_QP_EVENT_EVENT_DEINITIALIZE_FW 0x81UL + #define CREQ_QP_EVENT_EVENT_STOP_FUNC 0x82UL + #define CREQ_QP_EVENT_EVENT_QUERY_FUNC 0x83UL + #define CREQ_QP_EVENT_EVENT_SET_FUNC_RESOURCES 0x84UL + #define CREQ_QP_EVENT_EVENT_MAP_TC_TO_COS 0x8aUL + #define CREQ_QP_EVENT_EVENT_QUERY_VERSION 0x8bUL + #define CREQ_QP_EVENT_EVENT_MODIFY_CC 0x8cUL + #define CREQ_QP_EVENT_EVENT_QUERY_CC 0x8dUL + #define CREQ_QP_EVENT_EVENT_QP_ERROR_NOTIFICATION 0xc0UL + __le16 reserved48[3]; +}; + +/* Create QP command response (16 bytes) */ +struct creq_create_qp_resp { + u8 type; + #define CREQ_CREATE_QP_RESP_TYPE_MASK 0x3fUL + #define CREQ_CREATE_QP_RESP_TYPE_SFT 0 + #define CREQ_CREATE_QP_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_CREATE_QP_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_CREATE_QP_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_CREATE_QP_RESP_V 0x1UL + #define CREQ_CREATE_QP_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_CREATE_QP_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_CREATE_QP_RESP_EVENT_CREATE_QP 0x1UL + __le16 reserved48[3]; +}; + +/* Destroy QP command response (16 bytes) */ +struct creq_destroy_qp_resp { + u8 type; + #define CREQ_DESTROY_QP_RESP_TYPE_MASK 0x3fUL + #define CREQ_DESTROY_QP_RESP_TYPE_SFT 0 + #define CREQ_DESTROY_QP_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_DESTROY_QP_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_DESTROY_QP_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_DESTROY_QP_RESP_V 0x1UL + #define CREQ_DESTROY_QP_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_DESTROY_QP_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_DESTROY_QP_RESP_EVENT_DESTROY_QP 0x2UL + __le16 reserved48[3]; +}; + +/* Modify QP command response (16 bytes) */ +struct creq_modify_qp_resp { + u8 type; + #define CREQ_MODIFY_QP_RESP_TYPE_MASK 0x3fUL + #define CREQ_MODIFY_QP_RESP_TYPE_SFT 0 + #define CREQ_MODIFY_QP_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_MODIFY_QP_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_MODIFY_QP_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_MODIFY_QP_RESP_V 0x1UL + #define CREQ_MODIFY_QP_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_MODIFY_QP_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_MODIFY_QP_RESP_EVENT_MODIFY_QP 0x3UL + __le16 reserved48[3]; +}; + +/* Query QP command response (16 bytes) */ +struct creq_query_qp_resp { + u8 type; + #define CREQ_QUERY_QP_RESP_TYPE_MASK 0x3fUL + #define CREQ_QUERY_QP_RESP_TYPE_SFT 0 + #define CREQ_QUERY_QP_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_QUERY_QP_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_QUERY_QP_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 size; + u8 v; + #define CREQ_QUERY_QP_RESP_V 0x1UL + #define CREQ_QUERY_QP_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_QUERY_QP_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_QUERY_QP_RESP_EVENT_QUERY_QP 0x4UL + __le16 reserved48[3]; +}; + +/* Query QP command response side buffer structure (104 bytes) */ +struct creq_query_qp_resp_sb { + u8 opcode; + #define CREQ_QUERY_QP_RESP_SB_OPCODE_QUERY_QP 0x4UL + u8 status; + __le16 cookie; + __le16 flags; + u8 resp_size; + u8 reserved8; + __le32 xid; + u8 en_sqd_async_notify_state; + #define CREQ_QUERY_QP_RESP_SB_STATE_MASK 0xfUL + #define CREQ_QUERY_QP_RESP_SB_STATE_SFT 0 + #define CREQ_QUERY_QP_RESP_SB_STATE_RESET 0x0UL + #define CREQ_QUERY_QP_RESP_SB_STATE_INIT 0x1UL + #define CREQ_QUERY_QP_RESP_SB_STATE_RTR 0x2UL + #define CREQ_QUERY_QP_RESP_SB_STATE_RTS 0x3UL + #define CREQ_QUERY_QP_RESP_SB_STATE_SQD 0x4UL + #define CREQ_QUERY_QP_RESP_SB_STATE_SQE 0x5UL + #define CREQ_QUERY_QP_RESP_SB_STATE_ERR 0x6UL + #define CREQ_QUERY_QP_RESP_SB_EN_SQD_ASYNC_NOTIFY 0x10UL + u8 access; + #define CREQ_QUERY_QP_RESP_SB_ACCESS_LOCAL_WRITE 0x1UL + #define CREQ_QUERY_QP_RESP_SB_ACCESS_REMOTE_WRITE 0x2UL + #define CREQ_QUERY_QP_RESP_SB_ACCESS_REMOTE_READ 0x4UL + #define CREQ_QUERY_QP_RESP_SB_ACCESS_REMOTE_ATOMIC 0x8UL + __le16 pkey; + __le32 qkey; + __le32 reserved32; + __le32 dgid[4]; + __le32 flow_label; + __le16 sgid_index; + u8 hop_limit; + u8 traffic_class; + __le16 dest_mac[3]; + __le16 path_mtu_dest_vlan_id; + #define CREQ_QUERY_QP_RESP_SB_DEST_VLAN_ID_MASK 0xfffUL + #define CREQ_QUERY_QP_RESP_SB_DEST_VLAN_ID_SFT 0 + #define CREQ_QUERY_QP_RESP_SB_PATH_MTU_MASK 0xf000UL + #define CREQ_QUERY_QP_RESP_SB_PATH_MTU_SFT 12 + #define CREQ_QUERY_QP_RESP_SB_PATH_MTU_MTU_256 (0x0UL << 12) + #define CREQ_QUERY_QP_RESP_SB_PATH_MTU_MTU_512 (0x1UL << 12) + #define CREQ_QUERY_QP_RESP_SB_PATH_MTU_MTU_1024 (0x2UL << 12) + #define CREQ_QUERY_QP_RESP_SB_PATH_MTU_MTU_2048 (0x3UL << 12) + #define CREQ_QUERY_QP_RESP_SB_PATH_MTU_MTU_4096 (0x4UL << 12) + #define CREQ_QUERY_QP_RESP_SB_PATH_MTU_MTU_8192 (0x5UL << 12) + u8 timeout; + u8 retry_cnt; + u8 rnr_retry; + u8 min_rnr_timer; + __le32 rq_psn; + __le32 sq_psn; + u8 max_rd_atomic; + u8 max_dest_rd_atomic; + u8 tos_dscp_tos_ecn; + #define CREQ_QUERY_QP_RESP_SB_TOS_ECN_MASK 0x3UL + #define CREQ_QUERY_QP_RESP_SB_TOS_ECN_SFT 0 + #define CREQ_QUERY_QP_RESP_SB_TOS_DSCP_MASK 0xfcUL + #define CREQ_QUERY_QP_RESP_SB_TOS_DSCP_SFT 2 + u8 enable_cc; + #define CREQ_QUERY_QP_RESP_SB_ENABLE_CC 0x1UL + #define CREQ_QUERY_QP_RESP_SB_RESERVED7_MASK 0xfeUL + #define CREQ_QUERY_QP_RESP_SB_RESERVED7_SFT 1 + __le32 sq_size; + __le32 rq_size; + __le16 sq_sge; + __le16 rq_sge; + __le32 max_inline_data; + __le32 dest_qp_id; + __le32 unused_1; + __le16 src_mac[3]; + __le16 vlan_pcp_vlan_dei_vlan_id; + #define CREQ_QUERY_QP_RESP_SB_VLAN_ID_MASK 0xfffUL + #define CREQ_QUERY_QP_RESP_SB_VLAN_ID_SFT 0 + #define CREQ_QUERY_QP_RESP_SB_VLAN_DEI 0x1000UL + #define CREQ_QUERY_QP_RESP_SB_VLAN_PCP_MASK 0xe000UL + #define CREQ_QUERY_QP_RESP_SB_VLAN_PCP_SFT 13 +}; + +/* Create SRQ command response (16 bytes) */ +struct creq_create_srq_resp { + u8 type; + #define CREQ_CREATE_SRQ_RESP_TYPE_MASK 0x3fUL + #define CREQ_CREATE_SRQ_RESP_TYPE_SFT 0 + #define CREQ_CREATE_SRQ_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_CREATE_SRQ_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_CREATE_SRQ_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_CREATE_SRQ_RESP_V 0x1UL + #define CREQ_CREATE_SRQ_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_CREATE_SRQ_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_CREATE_SRQ_RESP_EVENT_CREATE_SRQ 0x5UL + __le16 reserved48[3]; +}; + +/* Destroy SRQ command response (16 bytes) */ +struct creq_destroy_srq_resp { + u8 type; + #define CREQ_DESTROY_SRQ_RESP_TYPE_MASK 0x3fUL + #define CREQ_DESTROY_SRQ_RESP_TYPE_SFT 0 + #define CREQ_DESTROY_SRQ_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_DESTROY_SRQ_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_DESTROY_SRQ_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_DESTROY_SRQ_RESP_V 0x1UL + #define CREQ_DESTROY_SRQ_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_DESTROY_SRQ_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_DESTROY_SRQ_RESP_EVENT_DESTROY_SRQ 0x6UL + __le16 enable_for_arm[3]; + #define CREQ_DESTROY_SRQ_RESP_ENABLE_FOR_ARM_MASK 0x30000UL + #define CREQ_DESTROY_SRQ_RESP_ENABLE_FOR_ARM_SFT 16 + #define CREQ_DESTROY_SRQ_RESP_RESERVED46_MASK 0xfffc0000UL + #define CREQ_DESTROY_SRQ_RESP_RESERVED46_SFT 18 +}; + +/* Query SRQ command response (16 bytes) */ +struct creq_query_srq_resp { + u8 type; + #define CREQ_QUERY_SRQ_RESP_TYPE_MASK 0x3fUL + #define CREQ_QUERY_SRQ_RESP_TYPE_SFT 0 + #define CREQ_QUERY_SRQ_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_QUERY_SRQ_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_QUERY_SRQ_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 size; + u8 v; + #define CREQ_QUERY_SRQ_RESP_V 0x1UL + #define CREQ_QUERY_SRQ_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_QUERY_SRQ_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_QUERY_SRQ_RESP_EVENT_QUERY_SRQ 0x8UL + __le16 reserved48[3]; +}; + +/* Query SRQ command response side buffer structure (24 bytes) */ +struct creq_query_srq_resp_sb { + u8 opcode; + #define CREQ_QUERY_SRQ_RESP_SB_OPCODE_QUERY_SRQ 0x8UL + u8 status; + __le16 cookie; + __le16 flags; + u8 resp_size; + u8 reserved8; + __le32 xid; + __le16 srq_limit; + __le16 reserved16; + __le32 data[4]; +}; + +/* Create CQ command Response (16 bytes) */ +struct creq_create_cq_resp { + u8 type; + #define CREQ_CREATE_CQ_RESP_TYPE_MASK 0x3fUL + #define CREQ_CREATE_CQ_RESP_TYPE_SFT 0 + #define CREQ_CREATE_CQ_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_CREATE_CQ_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_CREATE_CQ_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_CREATE_CQ_RESP_V 0x1UL + #define CREQ_CREATE_CQ_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_CREATE_CQ_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_CREATE_CQ_RESP_EVENT_CREATE_CQ 0x9UL + __le16 reserved48[3]; +}; + +/* Destroy CQ command response (16 bytes) */ +struct creq_destroy_cq_resp { + u8 type; + #define CREQ_DESTROY_CQ_RESP_TYPE_MASK 0x3fUL + #define CREQ_DESTROY_CQ_RESP_TYPE_SFT 0 + #define CREQ_DESTROY_CQ_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_DESTROY_CQ_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_DESTROY_CQ_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_DESTROY_CQ_RESP_V 0x1UL + #define CREQ_DESTROY_CQ_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_DESTROY_CQ_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_DESTROY_CQ_RESP_EVENT_DESTROY_CQ 0xaUL + __le16 cq_arm_lvl; + #define CREQ_DESTROY_CQ_RESP_CQ_ARM_LVL_MASK 0x3UL + #define CREQ_DESTROY_CQ_RESP_CQ_ARM_LVL_SFT 0 + #define CREQ_DESTROY_CQ_RESP_RESERVED14_MASK 0xfffcUL + #define CREQ_DESTROY_CQ_RESP_RESERVED14_SFT 2 + __le16 total_cnq_events; + __le16 reserved16; +}; + +/* Resize CQ command response (16 bytes) */ +struct creq_resize_cq_resp { + u8 type; + #define CREQ_RESIZE_CQ_RESP_TYPE_MASK 0x3fUL + #define CREQ_RESIZE_CQ_RESP_TYPE_SFT 0 + #define CREQ_RESIZE_CQ_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_RESIZE_CQ_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_RESIZE_CQ_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_RESIZE_CQ_RESP_V 0x1UL + #define CREQ_RESIZE_CQ_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_RESIZE_CQ_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_RESIZE_CQ_RESP_EVENT_RESIZE_CQ 0xcUL + __le16 reserved48[3]; +}; + +/* Allocate MRW command response (16 bytes) */ +struct creq_allocate_mrw_resp { + u8 type; + #define CREQ_ALLOCATE_MRW_RESP_TYPE_MASK 0x3fUL + #define CREQ_ALLOCATE_MRW_RESP_TYPE_SFT 0 + #define CREQ_ALLOCATE_MRW_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_ALLOCATE_MRW_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_ALLOCATE_MRW_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_ALLOCATE_MRW_RESP_V 0x1UL + #define CREQ_ALLOCATE_MRW_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_ALLOCATE_MRW_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_ALLOCATE_MRW_RESP_EVENT_ALLOCATE_MRW 0xdUL + __le16 reserved48[3]; +}; + +/* De-allocate key command response (16 bytes) */ +struct creq_deallocate_key_resp { + u8 type; + #define CREQ_DEALLOCATE_KEY_RESP_TYPE_MASK 0x3fUL + #define CREQ_DEALLOCATE_KEY_RESP_TYPE_SFT 0 + #define CREQ_DEALLOCATE_KEY_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_DEALLOCATE_KEY_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_DEALLOCATE_KEY_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_DEALLOCATE_KEY_RESP_V 0x1UL + #define CREQ_DEALLOCATE_KEY_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_DEALLOCATE_KEY_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_DEALLOCATE_KEY_RESP_EVENT_DEALLOCATE_KEY 0xeUL + __le16 reserved16; + __le32 bound_window_info; +}; + +/* Register MR command response (16 bytes) */ +struct creq_register_mr_resp { + u8 type; + #define CREQ_REGISTER_MR_RESP_TYPE_MASK 0x3fUL + #define CREQ_REGISTER_MR_RESP_TYPE_SFT 0 + #define CREQ_REGISTER_MR_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_REGISTER_MR_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_REGISTER_MR_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_REGISTER_MR_RESP_V 0x1UL + #define CREQ_REGISTER_MR_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_REGISTER_MR_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_REGISTER_MR_RESP_EVENT_REGISTER_MR 0xfUL + __le16 reserved48[3]; +}; + +/* Deregister MR command response (16 bytes) */ +struct creq_deregister_mr_resp { + u8 type; + #define CREQ_DEREGISTER_MR_RESP_TYPE_MASK 0x3fUL + #define CREQ_DEREGISTER_MR_RESP_TYPE_SFT 0 + #define CREQ_DEREGISTER_MR_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_DEREGISTER_MR_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_DEREGISTER_MR_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_DEREGISTER_MR_RESP_V 0x1UL + #define CREQ_DEREGISTER_MR_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_DEREGISTER_MR_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_DEREGISTER_MR_RESP_EVENT_DEREGISTER_MR 0x10UL + __le16 reserved16; + __le32 bound_windows; +}; + +/* Add GID command response (16 bytes) */ +struct creq_add_gid_resp { + u8 type; + #define CREQ_ADD_GID_RESP_TYPE_MASK 0x3fUL + #define CREQ_ADD_GID_RESP_TYPE_SFT 0 + #define CREQ_ADD_GID_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_ADD_GID_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_ADD_GID_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_ADD_GID_RESP_V 0x1UL + #define CREQ_ADD_GID_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_ADD_GID_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_ADD_GID_RESP_EVENT_ADD_GID 0x11UL + __le16 reserved48[3]; +}; + +/* Delete GID command response (16 bytes) */ +struct creq_delete_gid_resp { + u8 type; + #define CREQ_DELETE_GID_RESP_TYPE_MASK 0x3fUL + #define CREQ_DELETE_GID_RESP_TYPE_SFT 0 + #define CREQ_DELETE_GID_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_DELETE_GID_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_DELETE_GID_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_DELETE_GID_RESP_V 0x1UL + #define CREQ_DELETE_GID_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_DELETE_GID_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_DELETE_GID_RESP_EVENT_DELETE_GID 0x12UL + __le16 reserved48[3]; +}; + +/* Modify GID command response (16 bytes) */ +struct creq_modify_gid_resp { + u8 type; + #define CREQ_MODIFY_GID_RESP_TYPE_MASK 0x3fUL + #define CREQ_MODIFY_GID_RESP_TYPE_SFT 0 + #define CREQ_MODIFY_GID_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_MODIFY_GID_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_MODIFY_GID_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_MODIFY_GID_RESP_V 0x1UL + #define CREQ_MODIFY_GID_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_MODIFY_GID_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_MODIFY_GID_RESP_EVENT_ADD_GID 0x11UL + __le16 reserved48[3]; +}; + +/* Query GID command response (16 bytes) */ +struct creq_query_gid_resp { + u8 type; + #define CREQ_QUERY_GID_RESP_TYPE_MASK 0x3fUL + #define CREQ_QUERY_GID_RESP_TYPE_SFT 0 + #define CREQ_QUERY_GID_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_QUERY_GID_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_QUERY_GID_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 size; + u8 v; + #define CREQ_QUERY_GID_RESP_V 0x1UL + #define CREQ_QUERY_GID_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_QUERY_GID_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_QUERY_GID_RESP_EVENT_QUERY_GID 0x18UL + __le16 reserved48[3]; +}; + +/* Query GID command response side buffer structure (40 bytes) */ +struct creq_query_gid_resp_sb { + u8 opcode; + #define CREQ_QUERY_GID_RESP_SB_OPCODE_QUERY_GID 0x18UL + u8 status; + __le16 cookie; + __le16 flags; + u8 resp_size; + u8 reserved8; + __le32 gid[4]; + __le16 src_mac[3]; + __le16 vlan; + #define CREQ_QUERY_GID_RESP_SB_VLAN_VLAN_ID_MASK 0xfffUL + #define CREQ_QUERY_GID_RESP_SB_VLAN_VLAN_ID_SFT 0 + #define CREQ_QUERY_GID_RESP_SB_VLAN_TPID_MASK 0x7000UL + #define CREQ_QUERY_GID_RESP_SB_VLAN_TPID_SFT 12 + #define CREQ_QUERY_GID_RESP_SB_VLAN_TPID_TPID_88A8 (0x0UL << 12) + #define CREQ_QUERY_GID_RESP_SB_VLAN_TPID_TPID_8100 (0x1UL << 12) + #define CREQ_QUERY_GID_RESP_SB_VLAN_TPID_TPID_9100 (0x2UL << 12) + #define CREQ_QUERY_GID_RESP_SB_VLAN_TPID_TPID_9200 (0x3UL << 12) + #define CREQ_QUERY_GID_RESP_SB_VLAN_TPID_TPID_9300 (0x4UL << 12) + #define CREQ_QUERY_GID_RESP_SB_VLAN_TPID_TPID_CFG1 (0x5UL << 12) + #define CREQ_QUERY_GID_RESP_SB_VLAN_TPID_TPID_CFG2 (0x6UL << 12) + #define CREQ_QUERY_GID_RESP_SB_VLAN_TPID_TPID_CFG3 (0x7UL << 12) + #define CREQ_QUERY_GID_RESP_SB_VLAN_TPID_LAST \ + CREQ_QUERY_GID_RESP_SB_VLAN_TPID_TPID_CFG3 + #define CREQ_QUERY_GID_RESP_SB_VLAN_VLAN_EN 0x8000UL + __le16 ipid; + __le16 gid_index; + __le32 unused_0; +}; + +/* Create QP1 command response (16 bytes) */ +struct creq_create_qp1_resp { + u8 type; + #define CREQ_CREATE_QP1_RESP_TYPE_MASK 0x3fUL + #define CREQ_CREATE_QP1_RESP_TYPE_SFT 0 + #define CREQ_CREATE_QP1_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_CREATE_QP1_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_CREATE_QP1_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_CREATE_QP1_RESP_V 0x1UL + #define CREQ_CREATE_QP1_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_CREATE_QP1_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_CREATE_QP1_RESP_EVENT_CREATE_QP1 0x13UL + __le16 reserved48[3]; +}; + +/* Destroy QP1 command response (16 bytes) */ +struct creq_destroy_qp1_resp { + u8 type; + #define CREQ_DESTROY_QP1_RESP_TYPE_MASK 0x3fUL + #define CREQ_DESTROY_QP1_RESP_TYPE_SFT 0 + #define CREQ_DESTROY_QP1_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_DESTROY_QP1_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_DESTROY_QP1_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_DESTROY_QP1_RESP_V 0x1UL + #define CREQ_DESTROY_QP1_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_DESTROY_QP1_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_DESTROY_QP1_RESP_EVENT_DESTROY_QP1 0x14UL + __le16 reserved48[3]; +}; + +/* Create AH command response (16 bytes) */ +struct creq_create_ah_resp { + u8 type; + #define CREQ_CREATE_AH_RESP_TYPE_MASK 0x3fUL + #define CREQ_CREATE_AH_RESP_TYPE_SFT 0 + #define CREQ_CREATE_AH_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_CREATE_AH_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_CREATE_AH_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_CREATE_AH_RESP_V 0x1UL + #define CREQ_CREATE_AH_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_CREATE_AH_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_CREATE_AH_RESP_EVENT_CREATE_AH 0x15UL + __le16 reserved48[3]; +}; + +/* Destroy AH command response (16 bytes) */ +struct creq_destroy_ah_resp { + u8 type; + #define CREQ_DESTROY_AH_RESP_TYPE_MASK 0x3fUL + #define CREQ_DESTROY_AH_RESP_TYPE_SFT 0 + #define CREQ_DESTROY_AH_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_DESTROY_AH_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_DESTROY_AH_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 xid; + u8 v; + #define CREQ_DESTROY_AH_RESP_V 0x1UL + #define CREQ_DESTROY_AH_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_DESTROY_AH_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_DESTROY_AH_RESP_EVENT_DESTROY_AH 0x16UL + __le16 reserved48[3]; +}; + +/* Initialize Firmware command response (16 bytes) */ +struct creq_initialize_fw_resp { + u8 type; + #define CREQ_INITIALIZE_FW_RESP_TYPE_MASK 0x3fUL + #define CREQ_INITIALIZE_FW_RESP_TYPE_SFT 0 + #define CREQ_INITIALIZE_FW_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_INITIALIZE_FW_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_INITIALIZE_FW_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 reserved32; + u8 v; + #define CREQ_INITIALIZE_FW_RESP_V 0x1UL + #define CREQ_INITIALIZE_FW_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_INITIALIZE_FW_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_INITIALIZE_FW_RESP_EVENT_INITIALIZE_FW 0x80UL + __le16 reserved48[3]; +}; + +/* De-initialize Firmware command response (16 bytes) */ +struct creq_deinitialize_fw_resp { + u8 type; + #define CREQ_DEINITIALIZE_FW_RESP_TYPE_MASK 0x3fUL + #define CREQ_DEINITIALIZE_FW_RESP_TYPE_SFT 0 + #define CREQ_DEINITIALIZE_FW_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_DEINITIALIZE_FW_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_DEINITIALIZE_FW_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 reserved32; + u8 v; + #define CREQ_DEINITIALIZE_FW_RESP_V 0x1UL + #define CREQ_DEINITIALIZE_FW_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_DEINITIALIZE_FW_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_DEINITIALIZE_FW_RESP_EVENT_DEINITIALIZE_FW 0x81UL + __le16 reserved48[3]; +}; + +/* Stop function command response (16 bytes) */ +struct creq_stop_func_resp { + u8 type; + #define CREQ_STOP_FUNC_RESP_TYPE_MASK 0x3fUL + #define CREQ_STOP_FUNC_RESP_TYPE_SFT 0 + #define CREQ_STOP_FUNC_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_STOP_FUNC_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_STOP_FUNC_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 reserved32; + u8 v; + #define CREQ_STOP_FUNC_RESP_V 0x1UL + #define CREQ_STOP_FUNC_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_STOP_FUNC_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_STOP_FUNC_RESP_EVENT_STOP_FUNC 0x82UL + __le16 reserved48[3]; +}; + +/* Query function command response (16 bytes) */ +struct creq_query_func_resp { + u8 type; + #define CREQ_QUERY_FUNC_RESP_TYPE_MASK 0x3fUL + #define CREQ_QUERY_FUNC_RESP_TYPE_SFT 0 + #define CREQ_QUERY_FUNC_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_QUERY_FUNC_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_QUERY_FUNC_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 size; + u8 v; + #define CREQ_QUERY_FUNC_RESP_V 0x1UL + #define CREQ_QUERY_FUNC_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_QUERY_FUNC_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_QUERY_FUNC_RESP_EVENT_QUERY_FUNC 0x83UL + __le16 reserved48[3]; +}; + +/* Query function command response side buffer structure (88 bytes) */ +struct creq_query_func_resp_sb { + u8 opcode; + #define CREQ_QUERY_FUNC_RESP_SB_OPCODE_QUERY_FUNC 0x83UL + u8 status; + __le16 cookie; + __le16 flags; + u8 resp_size; + u8 reserved8; + __le64 max_mr_size; + __le32 max_qp; + __le16 max_qp_wr; + __le16 dev_cap_flags; + #define CREQ_QUERY_FUNC_RESP_SB_DEV_CAP_FLAGS_RESIZE_QP 0x1UL + __le32 max_cq; + __le32 max_cqe; + __le32 max_pd; + u8 max_sge; + u8 max_srq_sge; + u8 max_qp_rd_atom; + u8 max_qp_init_rd_atom; + __le32 max_mr; + __le32 max_mw; + __le32 max_raw_eth_qp; + __le32 max_ah; + __le32 max_fmr; + __le32 max_srq_wr; + __le32 max_pkeys; + __le32 max_inline_data; + u8 max_map_per_fmr; + u8 l2_db_space_size; + __le16 max_srq; + __le32 max_gid; + __le32 tqm_alloc_reqs[8]; +}; + +/* Set resources command response (16 bytes) */ +struct creq_set_func_resources_resp { + u8 type; + #define CREQ_SET_FUNC_RESOURCES_RESP_TYPE_MASK 0x3fUL + #define CREQ_SET_FUNC_RESOURCES_RESP_TYPE_SFT 0 + #define CREQ_SET_FUNC_RESOURCES_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_SET_FUNC_RESOURCES_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_SET_FUNC_RESOURCES_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 reserved32; + u8 v; + #define CREQ_SET_FUNC_RESOURCES_RESP_V 0x1UL + #define CREQ_SET_FUNC_RESOURCES_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_SET_FUNC_RESOURCES_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_SET_FUNC_RESOURCES_RESP_EVENT_SET_FUNC_RESOURCES 0x84UL + __le16 reserved48[3]; +}; + +/* Map TC to COS response (16 bytes) */ +struct creq_map_tc_to_cos_resp { + u8 type; + #define CREQ_MAP_TC_TO_COS_RESP_TYPE_MASK 0x3fUL + #define CREQ_MAP_TC_TO_COS_RESP_TYPE_SFT 0 + #define CREQ_MAP_TC_TO_COS_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_MAP_TC_TO_COS_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_MAP_TC_TO_COS_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 reserved32; + u8 v; + #define CREQ_MAP_TC_TO_COS_RESP_V 0x1UL + #define CREQ_MAP_TC_TO_COS_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_MAP_TC_TO_COS_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_MAP_TC_TO_COS_RESP_EVENT_MAP_TC_TO_COS 0x8aUL + __le16 reserved48[3]; +}; + +/* Query version response (16 bytes) */ +struct creq_query_version_resp { + u8 type; + #define CREQ_QUERY_VERSION_RESP_TYPE_MASK 0x3fUL + #define CREQ_QUERY_VERSION_RESP_TYPE_SFT 0 + #define CREQ_QUERY_VERSION_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_QUERY_VERSION_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_QUERY_VERSION_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + u8 fw_maj; + u8 fw_minor; + u8 fw_bld; + u8 fw_rsvd; + u8 v; + #define CREQ_QUERY_VERSION_RESP_V 0x1UL + #define CREQ_QUERY_VERSION_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_QUERY_VERSION_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_QUERY_VERSION_RESP_EVENT_QUERY_VERSION 0x8bUL + __le16 reserved16; + u8 intf_maj; + u8 intf_minor; + u8 intf_bld; + u8 intf_rsvd; +}; + +/* Modify congestion control command response (16 bytes) */ +struct creq_modify_cc_resp { + u8 type; + #define CREQ_MODIFY_CC_RESP_TYPE_MASK 0x3fUL + #define CREQ_MODIFY_CC_RESP_TYPE_SFT 0 + #define CREQ_MODIFY_CC_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_MODIFY_CC_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_MODIFY_CC_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 reserved32; + u8 v; + #define CREQ_MODIFY_CC_RESP_V 0x1UL + #define CREQ_MODIFY_CC_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_MODIFY_CC_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_MODIFY_CC_RESP_EVENT_MODIFY_CC 0x8cUL + __le16 reserved48[3]; +}; + +/* Query congestion control command response (16 bytes) */ +struct creq_query_cc_resp { + u8 type; + #define CREQ_QUERY_CC_RESP_TYPE_MASK 0x3fUL + #define CREQ_QUERY_CC_RESP_TYPE_SFT 0 + #define CREQ_QUERY_CC_RESP_TYPE_QP_EVENT 0x38UL + #define CREQ_QUERY_CC_RESP_RESERVED2_MASK 0xc0UL + #define CREQ_QUERY_CC_RESP_RESERVED2_SFT 6 + u8 status; + __le16 cookie; + __le32 size; + u8 v; + #define CREQ_QUERY_CC_RESP_V 0x1UL + #define CREQ_QUERY_CC_RESP_RESERVED7_MASK 0xfeUL + #define CREQ_QUERY_CC_RESP_RESERVED7_SFT 1 + u8 event; + #define CREQ_QUERY_CC_RESP_EVENT_QUERY_CC 0x8dUL + __le16 reserved48[3]; +}; + +/* Query congestion control command response side buffer structure (32 bytes) */ +struct creq_query_cc_resp_sb { + u8 opcode; + #define CREQ_QUERY_CC_RESP_SB_OPCODE_QUERY_CC 0x8dUL + u8 status; + __le16 cookie; + __le16 flags; + u8 resp_size; + u8 reserved8; + u8 enable_cc; + #define CREQ_QUERY_CC_RESP_SB_ENABLE_CC 0x1UL + u8 g; + #define CREQ_QUERY_CC_RESP_SB_G_MASK 0x7UL + #define CREQ_QUERY_CC_RESP_SB_G_SFT 0 + u8 num_phases_per_state; + __le16 init_cr; + u8 unused_2; + __le16 unused_3; + u8 unused_4; + __le16 init_tr; + u8 tos_dscp_tos_ecn; + #define CREQ_QUERY_CC_RESP_SB_TOS_ECN_MASK 0x3UL + #define CREQ_QUERY_CC_RESP_SB_TOS_ECN_SFT 0 + #define CREQ_QUERY_CC_RESP_SB_TOS_DSCP_MASK 0xfcUL + #define CREQ_QUERY_CC_RESP_SB_TOS_DSCP_SFT 2 + __le64 reserved64; + __le64 reserved64_1; +}; + +/* QP error notification event (16 bytes) */ +struct creq_qp_error_notification { + u8 type; + #define CREQ_QP_ERROR_NOTIFICATION_TYPE_MASK 0x3fUL + #define CREQ_QP_ERROR_NOTIFICATION_TYPE_SFT 0 + #define CREQ_QP_ERROR_NOTIFICATION_TYPE_QP_EVENT 0x38UL + #define CREQ_QP_ERROR_NOTIFICATION_RESERVED2_MASK 0xc0UL + #define CREQ_QP_ERROR_NOTIFICATION_RESERVED2_SFT 6 + u8 status; + u8 req_slow_path_state; + u8 req_err_state_reason; + __le32 xid; + u8 v; + #define CREQ_QP_ERROR_NOTIFICATION_V 0x1UL + #define CREQ_QP_ERROR_NOTIFICATION_RESERVED7_MASK 0xfeUL + #define CREQ_QP_ERROR_NOTIFICATION_RESERVED7_SFT 1 + u8 event; + #define CREQ_QP_ERROR_NOTIFICATION_EVENT_QP_ERROR_NOTIFICATION 0xc0UL + u8 res_slow_path_state; + u8 res_err_state_reason; + __le16 sq_cons_idx; + __le16 rq_cons_idx; +}; + +/* RoCE Slowpath HSI Specification 1.6.0 */ +#define ROCE_SP_HSI_VERSION_MAJOR 1 +#define ROCE_SP_HSI_VERSION_MINOR 6 +#define ROCE_SP_HSI_VERSION_UPDATE 0 + +#define ROCE_SP_HSI_VERSION_STR "1.6.0" +/* + * Following is the signature for ROCE_SP_HSI message field that indicates not + * applicable (All F's). Need to cast it the size of the field if needed. + */ +#define ROCE_SP_HSI_NA_SIGNATURE ((__le32)(-1)) +#endif /* __BNXT_RE_HSI_H__ */ diff --git a/drivers/infiniband/hw/cxgb3/cxio_dbg.c b/drivers/infiniband/hw/cxgb3/cxio_dbg.c index 445e89e5e7cf..97dbe728520a 100644 --- a/drivers/infiniband/hw/cxgb3/cxio_dbg.c +++ b/drivers/infiniband/hw/cxgb3/cxio_dbg.c @@ -51,17 +51,18 @@ void cxio_dump_tpt(struct cxio_rdev *rdev, u32 stag) m->mem_id = MEM_PMRX; m->addr = (stag>>8) * 32 + rdev->rnic_info.tpt_base; m->len = size; - PDBG("%s TPT addr 0x%x len %d\n", __func__, m->addr, m->len); + pr_debug("%s TPT addr 0x%x len %d\n", __func__, m->addr, m->len); rc = rdev->t3cdev_p->ctl(rdev->t3cdev_p, RDMA_GET_MEM, m); if (rc) { - PDBG("%s toectl returned error %d\n", __func__, rc); + pr_debug("%s toectl returned error %d\n", __func__, rc); kfree(m); return; } data = (u64 *)m->buf; while (size > 0) { - PDBG("TPT %08x: %016llx\n", m->addr, (unsigned long long) *data); + pr_debug("TPT %08x: %016llx\n", + m->addr, (unsigned long long)*data); size -= 8; data++; m->addr += 8; @@ -87,18 +88,19 @@ void cxio_dump_pbl(struct cxio_rdev *rdev, u32 pbl_addr, uint len, u8 shift) m->mem_id = MEM_PMRX; m->addr = pbl_addr; m->len = size; - PDBG("%s PBL addr 0x%x len %d depth %d\n", - __func__, m->addr, m->len, npages); + pr_debug("%s PBL addr 0x%x len %d depth %d\n", + __func__, m->addr, m->len, npages); rc = rdev->t3cdev_p->ctl(rdev->t3cdev_p, RDMA_GET_MEM, m); if (rc) { - PDBG("%s toectl returned error %d\n", __func__, rc); + pr_debug("%s toectl returned error %d\n", __func__, rc); kfree(m); return; } data = (u64 *)m->buf; while (size > 0) { - PDBG("PBL %08x: %016llx\n", m->addr, (unsigned long long) *data); + pr_debug("PBL %08x: %016llx\n", + m->addr, (unsigned long long)*data); size -= 8; data++; m->addr += 8; @@ -114,8 +116,8 @@ void cxio_dump_wqe(union t3_wr *wqe) if (size == 0) size = 8; while (size > 0) { - PDBG("WQE %p: %016llx\n", data, - (unsigned long long) be64_to_cpu(*data)); + pr_debug("WQE %p: %016llx\n", + data, (unsigned long long)be64_to_cpu(*data)); size--; data++; } @@ -127,8 +129,8 @@ void cxio_dump_wce(struct t3_cqe *wce) int size = sizeof(*wce); while (size > 0) { - PDBG("WCE %p: %016llx\n", data, - (unsigned long long) be64_to_cpu(*data)); + pr_debug("WCE %p: %016llx\n", + data, (unsigned long long)be64_to_cpu(*data)); size -= 8; data++; } @@ -148,17 +150,18 @@ void cxio_dump_rqt(struct cxio_rdev *rdev, u32 hwtid, int nents) m->mem_id = MEM_PMRX; m->addr = ((hwtid)<<10) + rdev->rnic_info.rqt_base; m->len = size; - PDBG("%s RQT addr 0x%x len %d\n", __func__, m->addr, m->len); + pr_debug("%s RQT addr 0x%x len %d\n", __func__, m->addr, m->len); rc = rdev->t3cdev_p->ctl(rdev->t3cdev_p, RDMA_GET_MEM, m); if (rc) { - PDBG("%s toectl returned error %d\n", __func__, rc); + pr_debug("%s toectl returned error %d\n", __func__, rc); kfree(m); return; } data = (u64 *)m->buf; while (size > 0) { - PDBG("RQT %08x: %016llx\n", m->addr, (unsigned long long) *data); + pr_debug("RQT %08x: %016llx\n", + m->addr, (unsigned long long)*data); size -= 8; data++; m->addr += 8; @@ -180,10 +183,10 @@ void cxio_dump_tcb(struct cxio_rdev *rdev, u32 hwtid) m->mem_id = MEM_CM; m->addr = hwtid * size; m->len = size; - PDBG("%s TCB %d len %d\n", __func__, m->addr, m->len); + pr_debug("%s TCB %d len %d\n", __func__, m->addr, m->len); rc = rdev->t3cdev_p->ctl(rdev->t3cdev_p, RDMA_GET_MEM, m); if (rc) { - PDBG("%s toectl returned error %d\n", __func__, rc); + pr_debug("%s toectl returned error %d\n", __func__, rc); kfree(m); return; } diff --git a/drivers/infiniband/hw/cxgb3/cxio_hal.c b/drivers/infiniband/hw/cxgb3/cxio_hal.c index ada2e5009c86..3eff6541bd6f 100644 --- a/drivers/infiniband/hw/cxgb3/cxio_hal.c +++ b/drivers/infiniband/hw/cxgb3/cxio_hal.c @@ -110,8 +110,7 @@ int cxio_hal_cq_op(struct cxio_rdev *rdev_p, struct t3_cq *cq, while (!CQ_VLD_ENTRY(rptr, cq->size_log2, cqe)) { udelay(1); if (i++ > 1000000) { - printk(KERN_ERR "%s: stalled rnic\n", - rdev_p->dev_name); + pr_err("%s: stalled rnic\n", rdev_p->dev_name); return -EIO; } } @@ -140,11 +139,10 @@ static int cxio_hal_clear_qp_ctx(struct cxio_rdev *rdev_p, u32 qpid) struct t3_modify_qp_wr *wqe; struct sk_buff *skb = alloc_skb(sizeof(*wqe), GFP_KERNEL); if (!skb) { - PDBG("%s alloc_skb failed\n", __func__); + pr_debug("%s alloc_skb failed\n", __func__); return -ENOMEM; } - wqe = (struct t3_modify_qp_wr *) skb_put(skb, sizeof(*wqe)); - memset(wqe, 0, sizeof(*wqe)); + wqe = skb_put_zero(skb, sizeof(*wqe)); build_fw_riwrh((struct fw_riwrh *) wqe, T3_WR_QP_MOD, T3_COMPLETION_FLAG | T3_NOTIFY_FLAG, 0, qpid, 7, T3_SOPEOP); @@ -230,7 +228,7 @@ static u32 get_qpid(struct cxio_rdev *rdev_p, struct cxio_ucontext *uctx) } out: mutex_unlock(&uctx->lock); - PDBG("%s qpid 0x%x\n", __func__, qpid); + pr_debug("%s qpid 0x%x\n", __func__, qpid); return qpid; } @@ -242,7 +240,7 @@ static void put_qpid(struct cxio_rdev *rdev_p, u32 qpid, entry = kmalloc(sizeof *entry, GFP_KERNEL); if (!entry) return; - PDBG("%s qpid 0x%x\n", __func__, qpid); + pr_debug("%s qpid 0x%x\n", __func__, qpid); entry->qpid = qpid; mutex_lock(&uctx->lock); list_add_tail(&entry->entry, &uctx->qpids); @@ -306,8 +304,8 @@ int cxio_create_qp(struct cxio_rdev *rdev_p, u32 kernel_domain, wq->udb = (u64)rdev_p->rnic_info.udbell_physbase + (wq->qpid << rdev_p->qpshift); wq->rdev = rdev_p; - PDBG("%s qpid 0x%x doorbell 0x%p udb 0x%llx\n", __func__, - wq->qpid, wq->doorbell, (unsigned long long) wq->udb); + pr_debug("%s qpid 0x%x doorbell 0x%p udb 0x%llx\n", + __func__, wq->qpid, wq->doorbell, (unsigned long long)wq->udb); return 0; err4: kfree(wq->sq); @@ -351,8 +349,8 @@ static void insert_recv_cqe(struct t3_wq *wq, struct t3_cq *cq) { struct t3_cqe cqe; - PDBG("%s wq %p cq %p sw_rptr 0x%x sw_wptr 0x%x\n", __func__, - wq, cq, cq->sw_rptr, cq->sw_wptr); + pr_debug("%s wq %p cq %p sw_rptr 0x%x sw_wptr 0x%x\n", __func__, + wq, cq, cq->sw_rptr, cq->sw_wptr); memset(&cqe, 0, sizeof(cqe)); cqe.header = cpu_to_be32(V_CQE_STATUS(TPT_ERR_SWFLUSH) | V_CQE_OPCODE(T3_SEND) | @@ -370,11 +368,11 @@ int cxio_flush_rq(struct t3_wq *wq, struct t3_cq *cq, int count) u32 ptr; int flushed = 0; - PDBG("%s wq %p cq %p\n", __func__, wq, cq); + pr_debug("%s wq %p cq %p\n", __func__, wq, cq); /* flush RQ */ - PDBG("%s rq_rptr %u rq_wptr %u skip count %u\n", __func__, - wq->rq_rptr, wq->rq_wptr, count); + pr_debug("%s rq_rptr %u rq_wptr %u skip count %u\n", __func__, + wq->rq_rptr, wq->rq_wptr, count); ptr = wq->rq_rptr + count; while (ptr++ != wq->rq_wptr) { insert_recv_cqe(wq, cq); @@ -388,8 +386,8 @@ static void insert_sq_cqe(struct t3_wq *wq, struct t3_cq *cq, { struct t3_cqe cqe; - PDBG("%s wq %p cq %p sw_rptr 0x%x sw_wptr 0x%x\n", __func__, - wq, cq, cq->sw_rptr, cq->sw_wptr); + pr_debug("%s wq %p cq %p sw_rptr 0x%x sw_wptr 0x%x\n", __func__, + wq, cq, cq->sw_rptr, cq->sw_wptr); memset(&cqe, 0, sizeof(cqe)); cqe.header = cpu_to_be32(V_CQE_STATUS(TPT_ERR_SWFLUSH) | V_CQE_OPCODE(sqp->opcode) | @@ -429,11 +427,11 @@ void cxio_flush_hw_cq(struct t3_cq *cq) { struct t3_cqe *cqe, *swcqe; - PDBG("%s cq %p cqid 0x%x\n", __func__, cq, cq->cqid); + pr_debug("%s cq %p cqid 0x%x\n", __func__, cq, cq->cqid); cqe = cxio_next_hw_cqe(cq); while (cqe) { - PDBG("%s flushing hwcq rptr 0x%x to swcq wptr 0x%x\n", - __func__, cq->rptr, cq->sw_wptr); + pr_debug("%s flushing hwcq rptr 0x%x to swcq wptr 0x%x\n", + __func__, cq->rptr, cq->sw_wptr); swcqe = cq->sw_queue + Q_PTR2IDX(cq->sw_wptr, cq->size_log2); *swcqe = *cqe; swcqe->header |= cpu_to_be32(V_CQE_SWCQE(1)); @@ -476,7 +474,7 @@ void cxio_count_scqes(struct t3_cq *cq, struct t3_wq *wq, int *count) (*count)++; ptr++; } - PDBG("%s cq %p count %d\n", __func__, cq, *count); + pr_debug("%s cq %p count %d\n", __func__, cq, *count); } void cxio_count_rcqes(struct t3_cq *cq, struct t3_wq *wq, int *count) @@ -485,7 +483,7 @@ void cxio_count_rcqes(struct t3_cq *cq, struct t3_wq *wq, int *count) u32 ptr; *count = 0; - PDBG("%s count zero %d\n", __func__, *count); + pr_debug("%s count zero %d\n", __func__, *count); ptr = cq->sw_rptr; while (!Q_EMPTY(ptr, cq->sw_wptr)) { cqe = cq->sw_queue + (Q_PTR2IDX(ptr, cq->size_log2)); @@ -494,7 +492,7 @@ void cxio_count_rcqes(struct t3_cq *cq, struct t3_wq *wq, int *count) (*count)++; ptr++; } - PDBG("%s cq %p count %d\n", __func__, cq, *count); + pr_debug("%s cq %p count %d\n", __func__, cq, *count); } static int cxio_hal_init_ctrl_cq(struct cxio_rdev *rdev_p) @@ -521,12 +519,12 @@ static int cxio_hal_init_ctrl_qp(struct cxio_rdev *rdev_p) skb = alloc_skb(sizeof(*wqe), GFP_KERNEL); if (!skb) { - PDBG("%s alloc_skb failed\n", __func__); + pr_debug("%s alloc_skb failed\n", __func__); return -ENOMEM; } err = cxio_hal_init_ctrl_cq(rdev_p); if (err) { - PDBG("%s err %d initializing ctrl_cq\n", __func__, err); + pr_debug("%s err %d initializing ctrl_cq\n", __func__, err); goto err; } rdev_p->ctrl_qp.workq = dma_alloc_coherent( @@ -536,7 +534,7 @@ static int cxio_hal_init_ctrl_qp(struct cxio_rdev *rdev_p) &(rdev_p->ctrl_qp.dma_addr), GFP_KERNEL); if (!rdev_p->ctrl_qp.workq) { - PDBG("%s dma_alloc_coherent failed\n", __func__); + pr_debug("%s dma_alloc_coherent failed\n", __func__); err = -ENOMEM; goto err; } @@ -562,8 +560,7 @@ static int cxio_hal_init_ctrl_qp(struct cxio_rdev *rdev_p) ctx1 |= ((u64) (V_EC_BASE_HI((u32) base_addr & 0xf) | V_EC_RESPQ(0) | V_EC_TYPE(0) | V_EC_GEN(1) | V_EC_UP_TOKEN(T3_CTL_QP_TID) | F_EC_VALID)) << 32; - wqe = (struct t3_modify_qp_wr *) skb_put(skb, sizeof(*wqe)); - memset(wqe, 0, sizeof(*wqe)); + wqe = skb_put_zero(skb, sizeof(*wqe)); build_fw_riwrh((struct fw_riwrh *) wqe, T3_WR_QP_MOD, 0, 0, T3_CTL_QP_TID, 7, T3_SOPEOP); wqe->flags = cpu_to_be32(MODQP_WRITE_EC); @@ -571,9 +568,9 @@ static int cxio_hal_init_ctrl_qp(struct cxio_rdev *rdev_p) wqe->sge_cmd = cpu_to_be64(sge_cmd); wqe->ctx1 = cpu_to_be64(ctx1); wqe->ctx0 = cpu_to_be64(ctx0); - PDBG("CtrlQP dma_addr 0x%llx workq %p size %d\n", - (unsigned long long) rdev_p->ctrl_qp.dma_addr, - rdev_p->ctrl_qp.workq, 1 << T3_CTRL_QP_SIZE_LOG2); + pr_debug("CtrlQP dma_addr 0x%llx workq %p size %d\n", + (unsigned long long)rdev_p->ctrl_qp.dma_addr, + rdev_p->ctrl_qp.workq, 1 << T3_CTRL_QP_SIZE_LOG2); skb->priority = CPL_PRIORITY_CONTROL; return iwch_cxgb3_ofld_send(rdev_p->t3cdev_p, skb); err: @@ -605,26 +602,26 @@ static int cxio_hal_ctrl_qp_write_mem(struct cxio_rdev *rdev_p, u32 addr, u64 utx_cmd; addr &= 0x7FFFFFF; nr_wqe = len % 96 ? len / 96 + 1 : len / 96; /* 96B max per WQE */ - PDBG("%s wptr 0x%x rptr 0x%x len %d, nr_wqe %d data %p addr 0x%0x\n", - __func__, rdev_p->ctrl_qp.wptr, rdev_p->ctrl_qp.rptr, len, - nr_wqe, data, addr); + pr_debug("%s wptr 0x%x rptr 0x%x len %d, nr_wqe %d data %p addr 0x%0x\n", + __func__, rdev_p->ctrl_qp.wptr, rdev_p->ctrl_qp.rptr, len, + nr_wqe, data, addr); utx_len = 3; /* in 32B unit */ for (i = 0; i < nr_wqe; i++) { if (Q_FULL(rdev_p->ctrl_qp.rptr, rdev_p->ctrl_qp.wptr, T3_CTRL_QP_SIZE_LOG2)) { - PDBG("%s ctrl_qp full wtpr 0x%0x rptr 0x%0x, " - "wait for more space i %d\n", __func__, - rdev_p->ctrl_qp.wptr, rdev_p->ctrl_qp.rptr, i); + pr_debug("%s ctrl_qp full wtpr 0x%0x rptr 0x%0x, wait for more space i %d\n", + __func__, + rdev_p->ctrl_qp.wptr, rdev_p->ctrl_qp.rptr, i); if (wait_event_interruptible(rdev_p->ctrl_qp.waitq, !Q_FULL(rdev_p->ctrl_qp.rptr, rdev_p->ctrl_qp.wptr, T3_CTRL_QP_SIZE_LOG2))) { - PDBG("%s ctrl_qp workq interrupted\n", - __func__); + pr_debug("%s ctrl_qp workq interrupted\n", + __func__); return -ERESTARTSYS; } - PDBG("%s ctrl_qp wakeup, continue posting work request " - "i %d\n", __func__, i); + pr_debug("%s ctrl_qp wakeup, continue posting work request i %d\n", + __func__, i); } wqe = (__be64 *)(rdev_p->ctrl_qp.workq + (rdev_p->ctrl_qp.wptr % (1 << T3_CTRL_QP_SIZE_LOG2))); @@ -645,7 +642,7 @@ static int cxio_hal_ctrl_qp_write_mem(struct cxio_rdev *rdev_p, u32 addr, if ((i != 0) && (i % (((1 << T3_CTRL_QP_SIZE_LOG2)) >> 1) == 0)) { flag = T3_COMPLETION_FLAG; - PDBG("%s force completion at i %d\n", __func__, i); + pr_debug("%s force completion at i %d\n", __func__, i); } /* build the utx mem command */ @@ -717,8 +714,8 @@ static int __cxio_tpt_op(struct cxio_rdev *rdev_p, u32 reset_tpt_entry, return -ENOMEM; *stag = (stag_idx << 8) | ((*stag) & 0xFF); } - PDBG("%s stag_state 0x%0x type 0x%0x pdid 0x%0x, stag_idx 0x%x\n", - __func__, stag_state, type, pdid, stag_idx); + pr_debug("%s stag_state 0x%0x type 0x%0x pdid 0x%0x, stag_idx 0x%x\n", + __func__, stag_state, type, pdid, stag_idx); mutex_lock(&rdev_p->ctrl_qp.lock); @@ -767,9 +764,9 @@ int cxio_write_pbl(struct cxio_rdev *rdev_p, __be64 *pbl, u32 wptr; int err; - PDBG("%s *pdb_addr 0x%x, pbl_base 0x%x, pbl_size %d\n", - __func__, pbl_addr, rdev_p->rnic_info.pbl_base, - pbl_size); + pr_debug("%s *pdb_addr 0x%x, pbl_base 0x%x, pbl_size %d\n", + __func__, pbl_addr, rdev_p->rnic_info.pbl_base, + pbl_size); mutex_lock(&rdev_p->ctrl_qp.lock); err = cxio_hal_ctrl_qp_write_mem(rdev_p, pbl_addr >> 5, pbl_size << 3, @@ -837,8 +834,8 @@ int cxio_rdma_init(struct cxio_rdev *rdev_p, struct t3_rdma_init_attr *attr) struct sk_buff *skb = alloc_skb(sizeof(*wqe), GFP_ATOMIC); if (!skb) return -ENOMEM; - PDBG("%s rdev_p %p\n", __func__, rdev_p); - wqe = (struct t3_rdma_init_wr *) __skb_put(skb, sizeof(*wqe)); + pr_debug("%s rdev_p %p\n", __func__, rdev_p); + wqe = __skb_put(skb, sizeof(*wqe)); wqe->wrh.op_seop_flags = cpu_to_be32(V_FW_RIWR_OP(T3_WR_INIT)); wqe->wrh.gen_tid_len = cpu_to_be32(V_FW_RIWR_TID(attr->tid) | V_FW_RIWR_LEN(sizeof(*wqe) >> 3)); @@ -880,22 +877,20 @@ static int cxio_hal_ev_handler(struct t3cdev *t3cdev_p, struct sk_buff *skb) static int cnt; struct cxio_rdev *rdev_p = NULL; struct respQ_msg_t *rsp_msg = (struct respQ_msg_t *) skb->data; - PDBG("%d: %s cq_id 0x%x cq_ptr 0x%x genbit %0x overflow %0x an %0x" - " se %0x notify %0x cqbranch %0x creditth %0x\n", - cnt, __func__, RSPQ_CQID(rsp_msg), RSPQ_CQPTR(rsp_msg), - RSPQ_GENBIT(rsp_msg), RSPQ_OVERFLOW(rsp_msg), RSPQ_AN(rsp_msg), - RSPQ_SE(rsp_msg), RSPQ_NOTIFY(rsp_msg), RSPQ_CQBRANCH(rsp_msg), - RSPQ_CREDIT_THRESH(rsp_msg)); - PDBG("CQE: QPID 0x%0x genbit %0x type 0x%0x status 0x%0x opcode %d " - "len 0x%0x wrid_hi_stag 0x%x wrid_low_msn 0x%x\n", - CQE_QPID(rsp_msg->cqe), CQE_GENBIT(rsp_msg->cqe), - CQE_TYPE(rsp_msg->cqe), CQE_STATUS(rsp_msg->cqe), - CQE_OPCODE(rsp_msg->cqe), CQE_LEN(rsp_msg->cqe), - CQE_WRID_HI(rsp_msg->cqe), CQE_WRID_LOW(rsp_msg->cqe)); + pr_debug("%d: %s cq_id 0x%x cq_ptr 0x%x genbit %0x overflow %0x an %0x se %0x notify %0x cqbranch %0x creditth %0x\n", + cnt, __func__, RSPQ_CQID(rsp_msg), RSPQ_CQPTR(rsp_msg), + RSPQ_GENBIT(rsp_msg), RSPQ_OVERFLOW(rsp_msg), RSPQ_AN(rsp_msg), + RSPQ_SE(rsp_msg), RSPQ_NOTIFY(rsp_msg), RSPQ_CQBRANCH(rsp_msg), + RSPQ_CREDIT_THRESH(rsp_msg)); + pr_debug("CQE: QPID 0x%0x genbit %0x type 0x%0x status 0x%0x opcode %d len 0x%0x wrid_hi_stag 0x%x wrid_low_msn 0x%x\n", + CQE_QPID(rsp_msg->cqe), CQE_GENBIT(rsp_msg->cqe), + CQE_TYPE(rsp_msg->cqe), CQE_STATUS(rsp_msg->cqe), + CQE_OPCODE(rsp_msg->cqe), CQE_LEN(rsp_msg->cqe), + CQE_WRID_HI(rsp_msg->cqe), CQE_WRID_LOW(rsp_msg->cqe)); rdev_p = (struct cxio_rdev *)t3cdev_p->ulp; if (!rdev_p) { - PDBG("%s called by t3cdev %p with null ulp\n", __func__, - t3cdev_p); + pr_debug("%s called by t3cdev %p with null ulp\n", __func__, + t3cdev_p); return 0; } if (CQE_QPID(rsp_msg->cqe) == T3_CTRL_QP_ID) { @@ -934,13 +929,13 @@ int cxio_rdev_open(struct cxio_rdev *rdev_p) strncpy(rdev_p->dev_name, rdev_p->t3cdev_p->name, T3_MAX_DEV_NAME_LEN); } else { - PDBG("%s t3cdev_p or dev_name must be set\n", __func__); + pr_debug("%s t3cdev_p or dev_name must be set\n", __func__); return -EINVAL; } list_add_tail(&rdev_p->entry, &rdev_list); - PDBG("%s opening rnic dev %s\n", __func__, rdev_p->dev_name); + pr_debug("%s opening rnic dev %s\n", __func__, rdev_p->dev_name); memset(&rdev_p->ctrl_qp, 0, sizeof(rdev_p->ctrl_qp)); if (!rdev_p->t3cdev_p) rdev_p->t3cdev_p = dev2t3cdev(netdev_p); @@ -949,13 +944,12 @@ int cxio_rdev_open(struct cxio_rdev *rdev_p) err = rdev_p->t3cdev_p->ctl(rdev_p->t3cdev_p, GET_EMBEDDED_INFO, &(rdev_p->fw_info)); if (err) { - printk(KERN_ERR "%s t3cdev_p(%p)->ctl returned error %d.\n", - __func__, rdev_p->t3cdev_p, err); + pr_err("%s t3cdev_p(%p)->ctl returned error %d\n", + __func__, rdev_p->t3cdev_p, err); goto err1; } if (G_FW_VERSION_MAJOR(rdev_p->fw_info.fw_vers) != CXIO_FW_MAJ) { - printk(KERN_ERR MOD "fatal firmware version mismatch: " - "need version %u but adapter has version %u\n", + pr_err("fatal firmware version mismatch: need version %u but adapter has version %u\n", CXIO_FW_MAJ, G_FW_VERSION_MAJOR(rdev_p->fw_info.fw_vers)); err = -EINVAL; @@ -965,15 +959,15 @@ int cxio_rdev_open(struct cxio_rdev *rdev_p) err = rdev_p->t3cdev_p->ctl(rdev_p->t3cdev_p, RDMA_GET_PARAMS, &(rdev_p->rnic_info)); if (err) { - printk(KERN_ERR "%s t3cdev_p(%p)->ctl returned error %d.\n", - __func__, rdev_p->t3cdev_p, err); + pr_err("%s t3cdev_p(%p)->ctl returned error %d\n", + __func__, rdev_p->t3cdev_p, err); goto err1; } err = rdev_p->t3cdev_p->ctl(rdev_p->t3cdev_p, GET_PORTS, &(rdev_p->port_info)); if (err) { - printk(KERN_ERR "%s t3cdev_p(%p)->ctl returned error %d.\n", - __func__, rdev_p->t3cdev_p, err); + pr_err("%s t3cdev_p(%p)->ctl returned error %d\n", + __func__, rdev_p->t3cdev_p, err); goto err1; } @@ -988,42 +982,39 @@ int cxio_rdev_open(struct cxio_rdev *rdev_p) PAGE_SHIFT)); rdev_p->qpnr = rdev_p->rnic_info.udbell_len >> PAGE_SHIFT; rdev_p->qpmask = (65536 >> ilog2(rdev_p->qpnr)) - 1; - PDBG("%s rnic %s info: tpt_base 0x%0x tpt_top 0x%0x num stags %d " - "pbl_base 0x%0x pbl_top 0x%0x rqt_base 0x%0x, rqt_top 0x%0x\n", - __func__, rdev_p->dev_name, rdev_p->rnic_info.tpt_base, - rdev_p->rnic_info.tpt_top, cxio_num_stags(rdev_p), - rdev_p->rnic_info.pbl_base, - rdev_p->rnic_info.pbl_top, rdev_p->rnic_info.rqt_base, - rdev_p->rnic_info.rqt_top); - PDBG("udbell_len 0x%0x udbell_physbase 0x%lx kdb_addr %p qpshift %lu " - "qpnr %d qpmask 0x%x\n", - rdev_p->rnic_info.udbell_len, - rdev_p->rnic_info.udbell_physbase, rdev_p->rnic_info.kdb_addr, - rdev_p->qpshift, rdev_p->qpnr, rdev_p->qpmask); + pr_debug("%s rnic %s info: tpt_base 0x%0x tpt_top 0x%0x num stags %d pbl_base 0x%0x pbl_top 0x%0x rqt_base 0x%0x, rqt_top 0x%0x\n", + __func__, rdev_p->dev_name, rdev_p->rnic_info.tpt_base, + rdev_p->rnic_info.tpt_top, cxio_num_stags(rdev_p), + rdev_p->rnic_info.pbl_base, + rdev_p->rnic_info.pbl_top, rdev_p->rnic_info.rqt_base, + rdev_p->rnic_info.rqt_top); + pr_debug("udbell_len 0x%0x udbell_physbase 0x%lx kdb_addr %p qpshift %lu qpnr %d qpmask 0x%x\n", + rdev_p->rnic_info.udbell_len, + rdev_p->rnic_info.udbell_physbase, rdev_p->rnic_info.kdb_addr, + rdev_p->qpshift, rdev_p->qpnr, rdev_p->qpmask); err = cxio_hal_init_ctrl_qp(rdev_p); if (err) { - printk(KERN_ERR "%s error %d initializing ctrl_qp.\n", - __func__, err); + pr_err("%s error %d initializing ctrl_qp\n", __func__, err); goto err1; } err = cxio_hal_init_resource(rdev_p, cxio_num_stags(rdev_p), 0, 0, T3_MAX_NUM_QP, T3_MAX_NUM_CQ, T3_MAX_NUM_PD); if (err) { - printk(KERN_ERR "%s error %d initializing hal resources.\n", + pr_err("%s error %d initializing hal resources\n", __func__, err); goto err2; } err = cxio_hal_pblpool_create(rdev_p); if (err) { - printk(KERN_ERR "%s error %d initializing pbl mem pool.\n", + pr_err("%s error %d initializing pbl mem pool\n", __func__, err); goto err3; } err = cxio_hal_rqtpool_create(rdev_p); if (err) { - printk(KERN_ERR "%s error %d initializing rqt mem pool.\n", + pr_err("%s error %d initializing rqt mem pool\n", __func__, err); goto err4; } @@ -1086,9 +1077,9 @@ static void flush_completed_wrs(struct t3_wq *wq, struct t3_cq *cq) /* * Insert this completed cqe into the swcq. */ - PDBG("%s moving cqe into swcq sq idx %ld cq idx %ld\n", - __func__, Q_PTR2IDX(ptr, wq->sq_size_log2), - Q_PTR2IDX(cq->sw_wptr, cq->size_log2)); + pr_debug("%s moving cqe into swcq sq idx %ld cq idx %ld\n", + __func__, Q_PTR2IDX(ptr, wq->sq_size_log2), + Q_PTR2IDX(cq->sw_wptr, cq->size_log2)); sqp->cqe.header |= htonl(V_CQE_SWCQE(1)); *(cq->sw_queue + Q_PTR2IDX(cq->sw_wptr, cq->size_log2)) = sqp->cqe; @@ -1154,12 +1145,11 @@ int cxio_poll_cq(struct t3_wq *wq, struct t3_cq *cq, struct t3_cqe *cqe, *credit = 0; hw_cqe = cxio_next_cqe(cq); - PDBG("%s CQE OOO %d qpid 0x%0x genbit %d type %d status 0x%0x" - " opcode 0x%0x len 0x%0x wrid_hi_stag 0x%x wrid_low_msn 0x%x\n", - __func__, CQE_OOO(*hw_cqe), CQE_QPID(*hw_cqe), - CQE_GENBIT(*hw_cqe), CQE_TYPE(*hw_cqe), CQE_STATUS(*hw_cqe), - CQE_OPCODE(*hw_cqe), CQE_LEN(*hw_cqe), CQE_WRID_HI(*hw_cqe), - CQE_WRID_LOW(*hw_cqe)); + pr_debug("%s CQE OOO %d qpid 0x%0x genbit %d type %d status 0x%0x opcode 0x%0x len 0x%0x wrid_hi_stag 0x%x wrid_low_msn 0x%x\n", + __func__, CQE_OOO(*hw_cqe), CQE_QPID(*hw_cqe), + CQE_GENBIT(*hw_cqe), CQE_TYPE(*hw_cqe), CQE_STATUS(*hw_cqe), + CQE_OPCODE(*hw_cqe), CQE_LEN(*hw_cqe), CQE_WRID_HI(*hw_cqe), + CQE_WRID_LOW(*hw_cqe)); /* * skip cqe's not affiliated with a QP. @@ -1278,9 +1268,10 @@ int cxio_poll_cq(struct t3_wq *wq, struct t3_cq *cq, struct t3_cqe *cqe, if (!SW_CQE(*hw_cqe) && (CQE_WRID_SQ_WPTR(*hw_cqe) != wq->sq_rptr)) { struct t3_swsq *sqp; - PDBG("%s out of order completion going in swsq at idx %ld\n", - __func__, - Q_PTR2IDX(CQE_WRID_SQ_WPTR(*hw_cqe), wq->sq_size_log2)); + pr_debug("%s out of order completion going in swsq at idx %ld\n", + __func__, + Q_PTR2IDX(CQE_WRID_SQ_WPTR(*hw_cqe), + wq->sq_size_log2)); sqp = wq->sq + Q_PTR2IDX(CQE_WRID_SQ_WPTR(*hw_cqe), wq->sq_size_log2); sqp->cqe = *hw_cqe; @@ -1298,13 +1289,13 @@ proc_cqe: */ if (SQ_TYPE(*hw_cqe)) { wq->sq_rptr = CQE_WRID_SQ_WPTR(*hw_cqe); - PDBG("%s completing sq idx %ld\n", __func__, - Q_PTR2IDX(wq->sq_rptr, wq->sq_size_log2)); + pr_debug("%s completing sq idx %ld\n", __func__, + Q_PTR2IDX(wq->sq_rptr, wq->sq_size_log2)); *cookie = wq->sq[Q_PTR2IDX(wq->sq_rptr, wq->sq_size_log2)].wr_id; wq->sq_rptr++; } else { - PDBG("%s completing rq idx %ld\n", __func__, - Q_PTR2IDX(wq->rq_rptr, wq->rq_size_log2)); + pr_debug("%s completing rq idx %ld\n", __func__, + Q_PTR2IDX(wq->rq_rptr, wq->rq_size_log2)); *cookie = wq->rq[Q_PTR2IDX(wq->rq_rptr, wq->rq_size_log2)].wr_id; if (wq->rq[Q_PTR2IDX(wq->rq_rptr, wq->rq_size_log2)].pbl_addr) cxio_hal_pblpool_free(wq->rdev, @@ -1322,12 +1313,12 @@ flush_wq: skip_cqe: if (SW_CQE(*hw_cqe)) { - PDBG("%s cq %p cqid 0x%x skip sw cqe sw_rptr 0x%x\n", - __func__, cq, cq->cqid, cq->sw_rptr); + pr_debug("%s cq %p cqid 0x%x skip sw cqe sw_rptr 0x%x\n", + __func__, cq, cq->cqid, cq->sw_rptr); ++cq->sw_rptr; } else { - PDBG("%s cq %p cqid 0x%x skip hw cqe rptr 0x%x\n", - __func__, cq, cq->cqid, cq->rptr); + pr_debug("%s cq %p cqid 0x%x skip hw cqe rptr 0x%x\n", + __func__, cq, cq->cqid, cq->rptr); ++cq->rptr; /* diff --git a/drivers/infiniband/hw/cxgb3/cxio_hal.h b/drivers/infiniband/hw/cxgb3/cxio_hal.h index 78fbe9ffe7f0..7e70c5492262 100644 --- a/drivers/infiniband/hw/cxgb3/cxio_hal.h +++ b/drivers/infiniband/hw/cxgb3/cxio_hal.h @@ -196,8 +196,11 @@ int cxio_poll_cq(struct t3_wq *wq, struct t3_cq *cq, struct t3_cqe *cqe, u8 *cqe_flushed, u64 *cookie, u32 *credit); int iwch_cxgb3_ofld_send(struct t3cdev *tdev, struct sk_buff *skb); -#define MOD "iw_cxgb3: " -#define PDBG(fmt, args...) pr_debug(MOD fmt, ## args) +#ifdef pr_fmt +#undef pr_fmt +#endif + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt #ifdef DEBUG void cxio_dump_tpt(struct cxio_rdev *rev, u32 stag); diff --git a/drivers/infiniband/hw/cxgb3/cxio_resource.c b/drivers/infiniband/hw/cxgb3/cxio_resource.c index c40088ecf9f3..c6e7bc4420b6 100644 --- a/drivers/infiniband/hw/cxgb3/cxio_resource.c +++ b/drivers/infiniband/hw/cxgb3/cxio_resource.c @@ -209,13 +209,13 @@ u32 cxio_hal_get_qpid(struct cxio_hal_resource *rscp) { u32 qpid = cxio_hal_get_resource(&rscp->qpid_fifo, &rscp->qpid_fifo_lock); - PDBG("%s qpid 0x%x\n", __func__, qpid); + pr_debug("%s qpid 0x%x\n", __func__, qpid); return qpid; } void cxio_hal_put_qpid(struct cxio_hal_resource *rscp, u32 qpid) { - PDBG("%s qpid 0x%x\n", __func__, qpid); + pr_debug("%s qpid 0x%x\n", __func__, qpid); cxio_hal_put_resource(&rscp->qpid_fifo, &rscp->qpid_fifo_lock, qpid); } @@ -257,13 +257,13 @@ void cxio_hal_destroy_resource(struct cxio_hal_resource *rscp) u32 cxio_hal_pblpool_alloc(struct cxio_rdev *rdev_p, int size) { unsigned long addr = gen_pool_alloc(rdev_p->pbl_pool, size); - PDBG("%s addr 0x%x size %d\n", __func__, (u32)addr, size); + pr_debug("%s addr 0x%x size %d\n", __func__, (u32)addr, size); return (u32)addr; } void cxio_hal_pblpool_free(struct cxio_rdev *rdev_p, u32 addr, int size) { - PDBG("%s addr 0x%x size %d\n", __func__, addr, size); + pr_debug("%s addr 0x%x size %d\n", __func__, addr, size); gen_pool_free(rdev_p->pbl_pool, (unsigned long)addr, size); } @@ -282,17 +282,18 @@ int cxio_hal_pblpool_create(struct cxio_rdev *rdev_p) pbl_chunk = min(rdev_p->rnic_info.pbl_top - pbl_start + 1, pbl_chunk); if (gen_pool_add(rdev_p->pbl_pool, pbl_start, pbl_chunk, -1)) { - PDBG("%s failed to add PBL chunk (%x/%x)\n", - __func__, pbl_start, pbl_chunk); + pr_debug("%s failed to add PBL chunk (%x/%x)\n", + __func__, pbl_start, pbl_chunk); if (pbl_chunk <= 1024 << MIN_PBL_SHIFT) { - printk(KERN_WARNING MOD "%s: Failed to add all PBL chunks (%x/%x)\n", - __func__, pbl_start, rdev_p->rnic_info.pbl_top - pbl_start); + pr_warn("%s: Failed to add all PBL chunks (%x/%x)\n", + __func__, pbl_start, + rdev_p->rnic_info.pbl_top - pbl_start); return 0; } pbl_chunk >>= 1; } else { - PDBG("%s added PBL chunk (%x/%x)\n", - __func__, pbl_start, pbl_chunk); + pr_debug("%s added PBL chunk (%x/%x)\n", + __func__, pbl_start, pbl_chunk); pbl_start += pbl_chunk; } } @@ -315,13 +316,13 @@ void cxio_hal_pblpool_destroy(struct cxio_rdev *rdev_p) u32 cxio_hal_rqtpool_alloc(struct cxio_rdev *rdev_p, int size) { unsigned long addr = gen_pool_alloc(rdev_p->rqt_pool, size << 6); - PDBG("%s addr 0x%x size %d\n", __func__, (u32)addr, size << 6); + pr_debug("%s addr 0x%x size %d\n", __func__, (u32)addr, size << 6); return (u32)addr; } void cxio_hal_rqtpool_free(struct cxio_rdev *rdev_p, u32 addr, int size) { - PDBG("%s addr 0x%x size %d\n", __func__, addr, size << 6); + pr_debug("%s addr 0x%x size %d\n", __func__, addr, size << 6); gen_pool_free(rdev_p->rqt_pool, (unsigned long)addr, size << 6); } diff --git a/drivers/infiniband/hw/cxgb3/iwch.c b/drivers/infiniband/hw/cxgb3/iwch.c index b3e11329801d..47b2ce2ef203 100644 --- a/drivers/infiniband/hw/cxgb3/iwch.c +++ b/drivers/infiniband/hw/cxgb3/iwch.c @@ -105,7 +105,7 @@ static void iwch_db_drop_task(struct work_struct *work) static void rnic_init(struct iwch_dev *rnicp) { - PDBG("%s iwch_dev %p\n", __func__, rnicp); + pr_debug("%s iwch_dev %p\n", __func__, rnicp); idr_init(&rnicp->cqidr); idr_init(&rnicp->qpidr); idr_init(&rnicp->mmidr); @@ -145,12 +145,11 @@ static void open_rnic_dev(struct t3cdev *tdev) { struct iwch_dev *rnicp; - PDBG("%s t3cdev %p\n", __func__, tdev); - printk_once(KERN_INFO MOD "Chelsio T3 RDMA Driver - version %s\n", - DRV_VERSION); + pr_debug("%s t3cdev %p\n", __func__, tdev); + pr_info_once("Chelsio T3 RDMA Driver - version %s\n", DRV_VERSION); rnicp = (struct iwch_dev *)ib_alloc_device(sizeof(*rnicp)); if (!rnicp) { - printk(KERN_ERR MOD "Cannot allocate ib device\n"); + pr_err("Cannot allocate ib device\n"); return; } rnicp->rdev.ulp = rnicp; @@ -160,7 +159,7 @@ static void open_rnic_dev(struct t3cdev *tdev) if (cxio_rdev_open(&rnicp->rdev)) { mutex_unlock(&dev_mutex); - printk(KERN_ERR MOD "Unable to open CXIO rdev\n"); + pr_err("Unable to open CXIO rdev\n"); ib_dealloc_device(&rnicp->ibdev); return; } @@ -171,18 +170,18 @@ static void open_rnic_dev(struct t3cdev *tdev) mutex_unlock(&dev_mutex); if (iwch_register_device(rnicp)) { - printk(KERN_ERR MOD "Unable to register device\n"); + pr_err("Unable to register device\n"); close_rnic_dev(tdev); } - printk(KERN_INFO MOD "Initialized device %s\n", - pci_name(rnicp->rdev.rnic_info.pdev)); + pr_info("Initialized device %s\n", + pci_name(rnicp->rdev.rnic_info.pdev)); return; } static void close_rnic_dev(struct t3cdev *tdev) { struct iwch_dev *dev, *tmp; - PDBG("%s t3cdev %p\n", __func__, tdev); + pr_debug("%s t3cdev %p\n", __func__, tdev); mutex_lock(&dev_mutex); list_for_each_entry_safe(dev, tmp, &dev_list, entry) { if (dev->rdev.t3cdev_p == tdev) { diff --git a/drivers/infiniband/hw/cxgb3/iwch_cm.c b/drivers/infiniband/hw/cxgb3/iwch_cm.c index 65ee64400deb..86975370a4c0 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_cm.c +++ b/drivers/infiniband/hw/cxgb3/iwch_cm.c @@ -112,9 +112,9 @@ static void connect_reply_upcall(struct iwch_ep *ep, int status); static void start_ep_timer(struct iwch_ep *ep) { - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); if (timer_pending(&ep->timer)) { - PDBG("%s stopped / restarted timer ep %p\n", __func__, ep); + pr_debug("%s stopped / restarted timer ep %p\n", __func__, ep); del_timer_sync(&ep->timer); } else get_ep(&ep->com); @@ -126,7 +126,7 @@ static void start_ep_timer(struct iwch_ep *ep) static void stop_ep_timer(struct iwch_ep *ep) { - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); if (!timer_pending(&ep->timer)) { WARN(1, "%s timer stopped when its not running! ep %p state %u\n", __func__, ep, ep->com.state); @@ -175,7 +175,7 @@ static void release_tid(struct t3cdev *tdev, u32 hwtid, struct sk_buff *skb) skb = get_skb(skb, sizeof *req, GFP_KERNEL); if (!skb) return; - req = (struct cpl_tid_release *) skb_put(skb, sizeof(*req)); + req = skb_put(skb, sizeof(*req)); req->wr.wr_hi = htonl(V_WR_OP(FW_WROPCODE_FORWARD)); OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_TID_RELEASE, hwtid)); skb->priority = CPL_PRIORITY_SETUP; @@ -190,7 +190,7 @@ int iwch_quiesce_tid(struct iwch_ep *ep) if (!skb) return -ENOMEM; - req = (struct cpl_set_tcb_field *) skb_put(skb, sizeof(*req)); + req = skb_put(skb, sizeof(*req)); req->wr.wr_hi = htonl(V_WR_OP(FW_WROPCODE_FORWARD)); req->wr.wr_lo = htonl(V_WR_TID(ep->hwtid)); OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_SET_TCB_FIELD, ep->hwtid)); @@ -211,7 +211,7 @@ int iwch_resume_tid(struct iwch_ep *ep) if (!skb) return -ENOMEM; - req = (struct cpl_set_tcb_field *) skb_put(skb, sizeof(*req)); + req = skb_put(skb, sizeof(*req)); req->wr.wr_hi = htonl(V_WR_OP(FW_WROPCODE_FORWARD)); req->wr.wr_lo = htonl(V_WR_TID(ep->hwtid)); OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_SET_TCB_FIELD, ep->hwtid)); @@ -227,13 +227,13 @@ int iwch_resume_tid(struct iwch_ep *ep) static void set_emss(struct iwch_ep *ep, u16 opt) { - PDBG("%s ep %p opt %u\n", __func__, ep, opt); + pr_debug("%s ep %p opt %u\n", __func__, ep, opt); ep->emss = T3C_DATA(ep->com.tdev)->mtus[G_TCPOPT_MSS(opt)] - 40; if (G_TCPOPT_TSTAMP(opt)) ep->emss -= 12; if (ep->emss < 128) ep->emss = 128; - PDBG("emss=%d\n", ep->emss); + pr_debug("emss=%d\n", ep->emss); } static enum iwch_ep_state state_read(struct iwch_ep_common *epc) @@ -257,7 +257,7 @@ static void state_set(struct iwch_ep_common *epc, enum iwch_ep_state new) unsigned long flags; spin_lock_irqsave(&epc->lock, flags); - PDBG("%s - %s -> %s\n", __func__, states[epc->state], states[new]); + pr_debug("%s - %s -> %s\n", __func__, states[epc->state], states[new]); __state_set(epc, new); spin_unlock_irqrestore(&epc->lock, flags); return; @@ -273,7 +273,7 @@ static void *alloc_ep(int size, gfp_t gfp) spin_lock_init(&epc->lock); init_waitqueue_head(&epc->waitq); } - PDBG("%s alloc ep %p\n", __func__, epc); + pr_debug("%s alloc ep %p\n", __func__, epc); return epc; } @@ -282,7 +282,8 @@ void __free_ep(struct kref *kref) struct iwch_ep *ep; ep = container_of(container_of(kref, struct iwch_ep_common, kref), struct iwch_ep, com); - PDBG("%s ep %p state %s\n", __func__, ep, states[state_read(&ep->com)]); + pr_debug("%s ep %p state %s\n", + __func__, ep, states[state_read(&ep->com)]); if (test_bit(RELEASE_RESOURCES, &ep->com.flags)) { cxgb3_remove_tid(ep->com.tdev, (void *)ep, ep->hwtid); dst_release(ep->dst); @@ -293,7 +294,7 @@ void __free_ep(struct kref *kref) static void release_ep_resources(struct iwch_ep *ep) { - PDBG("%s ep %p tid %d\n", __func__, ep, ep->hwtid); + pr_debug("%s ep %p tid %d\n", __func__, ep, ep->hwtid); set_bit(RELEASE_RESOURCES, &ep->com.flags); put_ep(&ep->com); } @@ -358,7 +359,7 @@ static unsigned int find_best_mtu(const struct t3c_data *d, unsigned short mtu) static void arp_failure_discard(struct t3cdev *dev, struct sk_buff *skb) { - PDBG("%s t3cdev %p\n", __func__, dev); + pr_debug("%s t3cdev %p\n", __func__, dev); kfree_skb(skb); } @@ -367,7 +368,7 @@ static void arp_failure_discard(struct t3cdev *dev, struct sk_buff *skb) */ static void act_open_req_arp_failure(struct t3cdev *dev, struct sk_buff *skb) { - printk(KERN_ERR MOD "ARP failure during connect\n"); + pr_err("ARP failure during connect\n"); kfree_skb(skb); } @@ -379,7 +380,7 @@ static void abort_arp_failure(struct t3cdev *dev, struct sk_buff *skb) { struct cpl_abort_req *req = cplhdr(skb); - PDBG("%s t3cdev %p\n", __func__, dev); + pr_debug("%s t3cdev %p\n", __func__, dev); req->cmd = CPL_ABORT_NO_RST; iwch_cxgb3_ofld_send(dev, skb); } @@ -389,15 +390,15 @@ static int send_halfclose(struct iwch_ep *ep, gfp_t gfp) struct cpl_close_con_req *req; struct sk_buff *skb; - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); skb = get_skb(NULL, sizeof(*req), gfp); if (!skb) { - printk(KERN_ERR MOD "%s - failed to alloc skb\n", __func__); + pr_err("%s - failed to alloc skb\n", __func__); return -ENOMEM; } skb->priority = CPL_PRIORITY_DATA; set_arp_failure_handler(skb, arp_failure_discard); - req = (struct cpl_close_con_req *) skb_put(skb, sizeof(*req)); + req = skb_put(skb, sizeof(*req)); req->wr.wr_hi = htonl(V_WR_OP(FW_WROPCODE_OFLD_CLOSE_CON)); req->wr.wr_lo = htonl(V_WR_TID(ep->hwtid)); OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_CLOSE_CON_REQ, ep->hwtid)); @@ -408,17 +409,15 @@ static int send_abort(struct iwch_ep *ep, struct sk_buff *skb, gfp_t gfp) { struct cpl_abort_req *req; - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); skb = get_skb(skb, sizeof(*req), gfp); if (!skb) { - printk(KERN_ERR MOD "%s - failed to alloc skb.\n", - __func__); + pr_err("%s - failed to alloc skb\n", __func__); return -ENOMEM; } skb->priority = CPL_PRIORITY_DATA; set_arp_failure_handler(skb, abort_arp_failure); - req = (struct cpl_abort_req *) skb_put(skb, sizeof(*req)); - memset(req, 0, sizeof(*req)); + req = skb_put_zero(skb, sizeof(*req)); req->wr.wr_hi = htonl(V_WR_OP(FW_WROPCODE_OFLD_HOST_ABORT_CON_REQ)); req->wr.wr_lo = htonl(V_WR_TID(ep->hwtid)); OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_ABORT_REQ, ep->hwtid)); @@ -434,12 +433,11 @@ static int send_connect(struct iwch_ep *ep) unsigned int mtu_idx; int wscale; - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); skb = get_skb(NULL, sizeof(*req), GFP_KERNEL); if (!skb) { - printk(KERN_ERR MOD "%s - failed to alloc skb.\n", - __func__); + pr_err("%s - failed to alloc skb\n", __func__); return -ENOMEM; } mtu_idx = find_best_mtu(T3C_DATA(ep->com.tdev), dst_mtu(ep->dst)); @@ -457,7 +455,7 @@ static int send_connect(struct iwch_ep *ep) skb->priority = CPL_PRIORITY_SETUP; set_arp_failure_handler(skb, act_open_req_arp_failure); - req = (struct cpl_act_open_req *) skb_put(skb, sizeof(*req)); + req = skb_put(skb, sizeof(*req)); req->wr.wr_hi = htonl(V_WR_OP(FW_WROPCODE_FORWARD)); OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_ACT_OPEN_REQ, ep->atid)); req->local_port = ep->com.local_addr.sin_port; @@ -478,7 +476,7 @@ static void send_mpa_req(struct iwch_ep *ep, struct sk_buff *skb) struct mpa_message *mpa; int len; - PDBG("%s ep %p pd_len %d\n", __func__, ep, ep->plen); + pr_debug("%s ep %p pd_len %d\n", __func__, ep, ep->plen); BUG_ON(skb_cloned(skb)); @@ -515,7 +513,7 @@ static void send_mpa_req(struct iwch_ep *ep, struct sk_buff *skb) set_arp_failure_handler(skb, arp_failure_discard); skb_reset_transport_header(skb); len = skb->len; - req = (struct tx_data_wr *) skb_push(skb, sizeof(*req)); + req = skb_push(skb, sizeof(*req)); req->wr_hi = htonl(V_WR_OP(FW_WROPCODE_OFLD_TX_DATA)|F_WR_COMPL); req->wr_lo = htonl(V_WR_TID(ep->hwtid)); req->len = htonl(len); @@ -538,17 +536,17 @@ static int send_mpa_reject(struct iwch_ep *ep, const void *pdata, u8 plen) struct mpa_message *mpa; struct sk_buff *skb; - PDBG("%s ep %p plen %d\n", __func__, ep, plen); + pr_debug("%s ep %p plen %d\n", __func__, ep, plen); mpalen = sizeof(*mpa) + plen; skb = get_skb(NULL, mpalen + sizeof(*req), GFP_KERNEL); if (!skb) { - printk(KERN_ERR MOD "%s - cannot alloc skb!\n", __func__); + pr_err("%s - cannot alloc skb!\n", __func__); return -ENOMEM; } skb_reserve(skb, sizeof(*req)); - mpa = (struct mpa_message *) skb_put(skb, mpalen); + mpa = skb_put(skb, mpalen); memset(mpa, 0, sizeof(*mpa)); memcpy(mpa->key, MPA_KEY_REP, sizeof(mpa->key)); mpa->flags = MPA_REJECT; @@ -566,7 +564,7 @@ static int send_mpa_reject(struct iwch_ep *ep, const void *pdata, u8 plen) skb->priority = CPL_PRIORITY_DATA; set_arp_failure_handler(skb, arp_failure_discard); skb_reset_transport_header(skb); - req = (struct tx_data_wr *) skb_push(skb, sizeof(*req)); + req = skb_push(skb, sizeof(*req)); req->wr_hi = htonl(V_WR_OP(FW_WROPCODE_OFLD_TX_DATA)|F_WR_COMPL); req->wr_lo = htonl(V_WR_TID(ep->hwtid)); req->len = htonl(mpalen); @@ -587,18 +585,18 @@ static int send_mpa_reply(struct iwch_ep *ep, const void *pdata, u8 plen) int len; struct sk_buff *skb; - PDBG("%s ep %p plen %d\n", __func__, ep, plen); + pr_debug("%s ep %p plen %d\n", __func__, ep, plen); mpalen = sizeof(*mpa) + plen; skb = get_skb(NULL, mpalen + sizeof(*req), GFP_KERNEL); if (!skb) { - printk(KERN_ERR MOD "%s - cannot alloc skb!\n", __func__); + pr_err("%s - cannot alloc skb!\n", __func__); return -ENOMEM; } skb->priority = CPL_PRIORITY_DATA; skb_reserve(skb, sizeof(*req)); - mpa = (struct mpa_message *) skb_put(skb, mpalen); + mpa = skb_put(skb, mpalen); memset(mpa, 0, sizeof(*mpa)); memcpy(mpa->key, MPA_KEY_REP, sizeof(mpa->key)); mpa->flags = (ep->mpa_attr.crc_enabled ? MPA_CRC : 0) | @@ -617,7 +615,7 @@ static int send_mpa_reply(struct iwch_ep *ep, const void *pdata, u8 plen) set_arp_failure_handler(skb, arp_failure_discard); skb_reset_transport_header(skb); len = skb->len; - req = (struct tx_data_wr *) skb_push(skb, sizeof(*req)); + req = skb_push(skb, sizeof(*req)); req->wr_hi = htonl(V_WR_OP(FW_WROPCODE_OFLD_TX_DATA)|F_WR_COMPL); req->wr_lo = htonl(V_WR_TID(ep->hwtid)); req->len = htonl(len); @@ -636,7 +634,7 @@ static int act_establish(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) struct cpl_act_establish *req = cplhdr(skb); unsigned int tid = GET_TID(req); - PDBG("%s ep %p tid %d\n", __func__, ep, tid); + pr_debug("%s ep %p tid %d\n", __func__, ep, tid); dst_confirm(ep->dst); @@ -660,7 +658,7 @@ static int act_establish(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) static void abort_connection(struct iwch_ep *ep, struct sk_buff *skb, gfp_t gfp) { - PDBG("%s ep %p\n", __FILE__, ep); + pr_debug("%s ep %p\n", __FILE__, ep); state_set(&ep->com, ABORTING); send_abort(ep, skb, gfp); } @@ -669,12 +667,12 @@ static void close_complete_upcall(struct iwch_ep *ep) { struct iw_cm_event event; - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); memset(&event, 0, sizeof(event)); event.event = IW_CM_EVENT_CLOSE; if (ep->com.cm_id) { - PDBG("close complete delivered ep %p cm_id %p tid %d\n", - ep, ep->com.cm_id, ep->hwtid); + pr_debug("close complete delivered ep %p cm_id %p tid %d\n", + ep, ep->com.cm_id, ep->hwtid); ep->com.cm_id->event_handler(ep->com.cm_id, &event); ep->com.cm_id->rem_ref(ep->com.cm_id); ep->com.cm_id = NULL; @@ -686,12 +684,12 @@ static void peer_close_upcall(struct iwch_ep *ep) { struct iw_cm_event event; - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); memset(&event, 0, sizeof(event)); event.event = IW_CM_EVENT_DISCONNECT; if (ep->com.cm_id) { - PDBG("peer close delivered ep %p cm_id %p tid %d\n", - ep, ep->com.cm_id, ep->hwtid); + pr_debug("peer close delivered ep %p cm_id %p tid %d\n", + ep, ep->com.cm_id, ep->hwtid); ep->com.cm_id->event_handler(ep->com.cm_id, &event); } } @@ -700,13 +698,13 @@ static void peer_abort_upcall(struct iwch_ep *ep) { struct iw_cm_event event; - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); memset(&event, 0, sizeof(event)); event.event = IW_CM_EVENT_CLOSE; event.status = -ECONNRESET; if (ep->com.cm_id) { - PDBG("abort delivered ep %p cm_id %p tid %d\n", ep, - ep->com.cm_id, ep->hwtid); + pr_debug("abort delivered ep %p cm_id %p tid %d\n", ep, + ep->com.cm_id, ep->hwtid); ep->com.cm_id->event_handler(ep->com.cm_id, &event); ep->com.cm_id->rem_ref(ep->com.cm_id); ep->com.cm_id = NULL; @@ -718,7 +716,7 @@ static void connect_reply_upcall(struct iwch_ep *ep, int status) { struct iw_cm_event event; - PDBG("%s ep %p status %d\n", __func__, ep, status); + pr_debug("%s ep %p status %d\n", __func__, ep, status); memset(&event, 0, sizeof(event)); event.event = IW_CM_EVENT_CONNECT_REPLY; event.status = status; @@ -732,8 +730,8 @@ static void connect_reply_upcall(struct iwch_ep *ep, int status) event.private_data = ep->mpa_pkt + sizeof(struct mpa_message); } if (ep->com.cm_id) { - PDBG("%s ep %p tid %d status %d\n", __func__, ep, - ep->hwtid, status); + pr_debug("%s ep %p tid %d status %d\n", __func__, ep, + ep->hwtid, status); ep->com.cm_id->event_handler(ep->com.cm_id, &event); } if (status < 0) { @@ -747,7 +745,7 @@ static void connect_request_upcall(struct iwch_ep *ep) { struct iw_cm_event event; - PDBG("%s ep %p tid %d\n", __func__, ep, ep->hwtid); + pr_debug("%s ep %p tid %d\n", __func__, ep, ep->hwtid); memset(&event, 0, sizeof(event)); event.event = IW_CM_EVENT_CONNECT_REQUEST; memcpy(&event.local_addr, &ep->com.local_addr, @@ -776,7 +774,7 @@ static void established_upcall(struct iwch_ep *ep) { struct iw_cm_event event; - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); memset(&event, 0, sizeof(event)); event.event = IW_CM_EVENT_ESTABLISHED; /* @@ -785,7 +783,7 @@ static void established_upcall(struct iwch_ep *ep) */ event.ird = event.ord = 8; if (ep->com.cm_id) { - PDBG("%s ep %p tid %d\n", __func__, ep, ep->hwtid); + pr_debug("%s ep %p tid %d\n", __func__, ep, ep->hwtid); ep->com.cm_id->event_handler(ep->com.cm_id, &event); } } @@ -795,14 +793,14 @@ static int update_rx_credits(struct iwch_ep *ep, u32 credits) struct cpl_rx_data_ack *req; struct sk_buff *skb; - PDBG("%s ep %p credits %u\n", __func__, ep, credits); + pr_debug("%s ep %p credits %u\n", __func__, ep, credits); skb = get_skb(NULL, sizeof(*req), GFP_KERNEL); if (!skb) { - printk(KERN_ERR MOD "update_rx_credits - cannot alloc skb!\n"); + pr_err("update_rx_credits - cannot alloc skb!\n"); return 0; } - req = (struct cpl_rx_data_ack *) skb_put(skb, sizeof(*req)); + req = skb_put(skb, sizeof(*req)); req->wr.wr_hi = htonl(V_WR_OP(FW_WROPCODE_FORWARD)); OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_RX_DATA_ACK, ep->hwtid)); req->credit_dack = htonl(V_RX_CREDITS(credits) | V_RX_FORCE_ACK(1)); @@ -819,7 +817,7 @@ static void process_mpa_reply(struct iwch_ep *ep, struct sk_buff *skb) enum iwch_qp_attr_mask mask; int err; - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); /* * Stop mpa timer. If it expired, then the state has @@ -906,10 +904,10 @@ static void process_mpa_reply(struct iwch_ep *ep, struct sk_buff *skb) ep->mpa_attr.recv_marker_enabled = markers_enabled; ep->mpa_attr.xmit_marker_enabled = mpa->flags & MPA_MARKERS ? 1 : 0; ep->mpa_attr.version = mpa_rev; - PDBG("%s - crc_enabled=%d, recv_marker_enabled=%d, " - "xmit_marker_enabled=%d, version=%d\n", __func__, - ep->mpa_attr.crc_enabled, ep->mpa_attr.recv_marker_enabled, - ep->mpa_attr.xmit_marker_enabled, ep->mpa_attr.version); + pr_debug("%s - crc_enabled=%d, recv_marker_enabled=%d, xmit_marker_enabled=%d, version=%d\n", + __func__, + ep->mpa_attr.crc_enabled, ep->mpa_attr.recv_marker_enabled, + ep->mpa_attr.xmit_marker_enabled, ep->mpa_attr.version); attrs.mpa_attr = ep->mpa_attr; attrs.max_ird = ep->ird; @@ -944,7 +942,7 @@ static void process_mpa_request(struct iwch_ep *ep, struct sk_buff *skb) struct mpa_message *mpa; u16 plen; - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); /* * Stop mpa timer. If it expired, then the state has @@ -964,7 +962,7 @@ static void process_mpa_request(struct iwch_ep *ep, struct sk_buff *skb) return; } - PDBG("%s enter (%s line %u)\n", __func__, __FILE__, __LINE__); + pr_debug("%s enter (%s line %u)\n", __func__, __FILE__, __LINE__); /* * Copy the new data into our accumulation buffer. @@ -979,7 +977,7 @@ static void process_mpa_request(struct iwch_ep *ep, struct sk_buff *skb) */ if (ep->mpa_pkt_len < sizeof(*mpa)) return; - PDBG("%s enter (%s line %u)\n", __func__, __FILE__, __LINE__); + pr_debug("%s enter (%s line %u)\n", __func__, __FILE__, __LINE__); mpa = (struct mpa_message *) ep->mpa_pkt; /* @@ -1029,10 +1027,10 @@ static void process_mpa_request(struct iwch_ep *ep, struct sk_buff *skb) ep->mpa_attr.recv_marker_enabled = markers_enabled; ep->mpa_attr.xmit_marker_enabled = mpa->flags & MPA_MARKERS ? 1 : 0; ep->mpa_attr.version = mpa_rev; - PDBG("%s - crc_enabled=%d, recv_marker_enabled=%d, " - "xmit_marker_enabled=%d, version=%d\n", __func__, - ep->mpa_attr.crc_enabled, ep->mpa_attr.recv_marker_enabled, - ep->mpa_attr.xmit_marker_enabled, ep->mpa_attr.version); + pr_debug("%s - crc_enabled=%d, recv_marker_enabled=%d, xmit_marker_enabled=%d, version=%d\n", + __func__, + ep->mpa_attr.crc_enabled, ep->mpa_attr.recv_marker_enabled, + ep->mpa_attr.xmit_marker_enabled, ep->mpa_attr.version); state_set(&ep->com, MPA_REQ_RCVD); @@ -1047,7 +1045,7 @@ static int rx_data(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) struct cpl_rx_data *hdr = cplhdr(skb); unsigned int dlen = ntohs(hdr->len); - PDBG("%s ep %p dlen %u\n", __func__, ep, dlen); + pr_debug("%s ep %p dlen %u\n", __func__, ep, dlen); skb_pull(skb, sizeof(*hdr)); skb_trim(skb, dlen); @@ -1065,8 +1063,7 @@ static int rx_data(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) case MPA_REP_SENT: break; default: - printk(KERN_ERR MOD "%s Unexpected streaming data." - " ep %p state %d tid %d\n", + pr_err("%s Unexpected streaming data. ep %p state %d tid %d\n", __func__, ep, state_read(&ep->com), ep->hwtid); /* @@ -1095,11 +1092,11 @@ static int tx_ack(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) unsigned long flags; int post_zb = 0; - PDBG("%s ep %p credits %u\n", __func__, ep, credits); + pr_debug("%s ep %p credits %u\n", __func__, ep, credits); if (credits == 0) { - PDBG("%s 0 credit ack ep %p state %u\n", - __func__, ep, state_read(&ep->com)); + pr_debug("%s 0 credit ack ep %p state %u\n", + __func__, ep, state_read(&ep->com)); return CPL_RET_BUF_DONE; } @@ -1107,24 +1104,24 @@ static int tx_ack(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) BUG_ON(credits != 1); dst_confirm(ep->dst); if (!ep->mpa_skb) { - PDBG("%s rdma_init wr_ack ep %p state %u\n", - __func__, ep, ep->com.state); + pr_debug("%s rdma_init wr_ack ep %p state %u\n", + __func__, ep, ep->com.state); if (ep->mpa_attr.initiator) { - PDBG("%s initiator ep %p state %u\n", - __func__, ep, ep->com.state); + pr_debug("%s initiator ep %p state %u\n", + __func__, ep, ep->com.state); if (peer2peer && ep->com.state == FPDU_MODE) post_zb = 1; } else { - PDBG("%s responder ep %p state %u\n", - __func__, ep, ep->com.state); + pr_debug("%s responder ep %p state %u\n", + __func__, ep, ep->com.state); if (ep->com.state == MPA_REQ_RCVD) { ep->com.rpl_done = 1; wake_up(&ep->com.waitq); } } } else { - PDBG("%s lsm ack ep %p state %u freeing skb\n", - __func__, ep, ep->com.state); + pr_debug("%s lsm ack ep %p state %u freeing skb\n", + __func__, ep, ep->com.state); kfree_skb(ep->mpa_skb); ep->mpa_skb = NULL; } @@ -1140,7 +1137,7 @@ static int abort_rpl(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) unsigned long flags; int release = 0; - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); BUG_ON(!ep); /* @@ -1159,8 +1156,7 @@ static int abort_rpl(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) release = 1; break; default: - printk(KERN_ERR "%s ep %p state %d\n", - __func__, ep, ep->com.state); + pr_err("%s ep %p state %d\n", __func__, ep, ep->com.state); break; } spin_unlock_irqrestore(&ep->com.lock, flags); @@ -1184,8 +1180,8 @@ static int act_open_rpl(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) struct iwch_ep *ep = ctx; struct cpl_act_open_rpl *rpl = cplhdr(skb); - PDBG("%s ep %p status %u errno %d\n", __func__, ep, rpl->status, - status2errno(rpl->status)); + pr_debug("%s ep %p status %u errno %d\n", __func__, ep, rpl->status, + status2errno(rpl->status)); connect_reply_upcall(ep, status2errno(rpl->status)); state_set(&ep->com, DEAD); if (ep->com.tdev->type != T3A && act_open_has_tid(rpl->status)) @@ -1202,14 +1198,14 @@ static int listen_start(struct iwch_listen_ep *ep) struct sk_buff *skb; struct cpl_pass_open_req *req; - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); skb = get_skb(NULL, sizeof(*req), GFP_KERNEL); if (!skb) { - printk(KERN_ERR MOD "t3c_listen_start failed to alloc skb!\n"); + pr_err("t3c_listen_start failed to alloc skb!\n"); return -ENOMEM; } - req = (struct cpl_pass_open_req *) skb_put(skb, sizeof(*req)); + req = skb_put(skb, sizeof(*req)); req->wr.wr_hi = htonl(V_WR_OP(FW_WROPCODE_FORWARD)); OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_PASS_OPEN_REQ, ep->stid)); req->local_port = ep->com.local_addr.sin_port; @@ -1230,8 +1226,8 @@ static int pass_open_rpl(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) struct iwch_listen_ep *ep = ctx; struct cpl_pass_open_rpl *rpl = cplhdr(skb); - PDBG("%s ep %p status %d error %d\n", __func__, ep, - rpl->status, status2errno(rpl->status)); + pr_debug("%s ep %p status %d error %d\n", __func__, ep, + rpl->status, status2errno(rpl->status)); ep->com.rpl_err = status2errno(rpl->status); ep->com.rpl_done = 1; wake_up(&ep->com.waitq); @@ -1244,13 +1240,13 @@ static int listen_stop(struct iwch_listen_ep *ep) struct sk_buff *skb; struct cpl_close_listserv_req *req; - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); skb = get_skb(NULL, sizeof(*req), GFP_KERNEL); if (!skb) { - printk(KERN_ERR MOD "%s - failed to alloc skb\n", __func__); + pr_err("%s - failed to alloc skb\n", __func__); return -ENOMEM; } - req = (struct cpl_close_listserv_req *) skb_put(skb, sizeof(*req)); + req = skb_put(skb, sizeof(*req)); req->wr.wr_hi = htonl(V_WR_OP(FW_WROPCODE_FORWARD)); req->cpu_idx = 0; OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_CLOSE_LISTSRV_REQ, ep->stid)); @@ -1264,7 +1260,7 @@ static int close_listsrv_rpl(struct t3cdev *tdev, struct sk_buff *skb, struct iwch_listen_ep *ep = ctx; struct cpl_close_listserv_rpl *rpl = cplhdr(skb); - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); ep->com.rpl_err = status2errno(rpl->status); ep->com.rpl_done = 1; wake_up(&ep->com.waitq); @@ -1278,7 +1274,7 @@ static void accept_cr(struct iwch_ep *ep, __be32 peer_ip, struct sk_buff *skb) u32 opt0h, opt0l, opt2; int wscale; - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); BUG_ON(skb_cloned(skb)); skb_trim(skb, sizeof(*rpl)); skb_get(skb); @@ -1312,8 +1308,8 @@ static void accept_cr(struct iwch_ep *ep, __be32 peer_ip, struct sk_buff *skb) static void reject_cr(struct t3cdev *tdev, u32 hwtid, __be32 peer_ip, struct sk_buff *skb) { - PDBG("%s t3cdev %p tid %u peer_ip %x\n", __func__, tdev, hwtid, - peer_ip); + pr_debug("%s t3cdev %p tid %u peer_ip %x\n", __func__, tdev, hwtid, + peer_ip); BUG_ON(skb_cloned(skb)); skb_trim(skb, sizeof(struct cpl_tid_release)); skb_get(skb); @@ -1347,11 +1343,10 @@ static int pass_accept_req(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) struct rtable *rt; struct iff_mac tim; - PDBG("%s parent ep %p tid %u\n", __func__, parent_ep, hwtid); + pr_debug("%s parent ep %p tid %u\n", __func__, parent_ep, hwtid); if (state_read(&parent_ep->com) != LISTEN) { - printk(KERN_ERR "%s - listening ep not in LISTEN\n", - __func__); + pr_err("%s - listening ep not in LISTEN\n", __func__); goto reject; } @@ -1361,8 +1356,7 @@ static int pass_accept_req(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) tim.mac_addr = req->dst_mac; tim.vlan_tag = ntohs(req->vlan_tag); if (tdev->ctl(tdev, GET_IFF_FROM_MAC, &tim) < 0 || !tim.dev) { - printk(KERN_ERR "%s bad dst mac %pM\n", - __func__, req->dst_mac); + pr_err("%s bad dst mac %pM\n", __func__, req->dst_mac); goto reject; } @@ -1373,22 +1367,19 @@ static int pass_accept_req(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) req->local_port, req->peer_port, G_PASS_OPEN_TOS(ntohl(req->tos_tid))); if (!rt) { - printk(KERN_ERR MOD "%s - failed to find dst entry!\n", - __func__); + pr_err("%s - failed to find dst entry!\n", __func__); goto reject; } dst = &rt->dst; l2t = t3_l2t_get(tdev, dst, NULL, &req->peer_ip); if (!l2t) { - printk(KERN_ERR MOD "%s - failed to allocate l2t entry!\n", - __func__); + pr_err("%s - failed to allocate l2t entry!\n", __func__); dst_release(dst); goto reject; } child_ep = alloc_ep(sizeof(*child_ep), GFP_KERNEL); if (!child_ep) { - printk(KERN_ERR MOD "%s - failed to allocate ep entry!\n", - __func__); + pr_err("%s - failed to allocate ep entry!\n", __func__); l2t_release(tdev, l2t); dst_release(dst); goto reject; @@ -1423,7 +1414,7 @@ static int pass_establish(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) struct iwch_ep *ep = ctx; struct cpl_pass_establish *req = cplhdr(skb); - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); ep->snd_seq = ntohl(req->snd_isn); ep->rcv_seq = ntohl(req->rcv_isn); @@ -1444,7 +1435,7 @@ static int peer_close(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) int disconnect = 1; int release = 0; - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); dst_confirm(ep->dst); spin_lock_irqsave(&ep->com.lock, flags); @@ -1467,14 +1458,14 @@ static int peer_close(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) __state_set(&ep->com, CLOSING); ep->com.rpl_done = 1; ep->com.rpl_err = -ECONNRESET; - PDBG("waking up ep %p\n", ep); + pr_debug("waking up ep %p\n", ep); wake_up(&ep->com.waitq); break; case MPA_REP_SENT: __state_set(&ep->com, CLOSING); ep->com.rpl_done = 1; ep->com.rpl_err = -ECONNRESET; - PDBG("waking up ep %p\n", ep); + pr_debug("waking up ep %p\n", ep); wake_up(&ep->com.waitq); break; case FPDU_MODE: @@ -1539,8 +1530,8 @@ static int peer_abort(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) unsigned long flags; if (is_neg_adv_abort(req->status)) { - PDBG("%s neg_adv_abort ep %p tid %d\n", __func__, ep, - ep->hwtid); + pr_debug("%s neg_adv_abort ep %p tid %d\n", __func__, ep, + ep->hwtid); t3_l2t_send_event(ep->com.tdev, ep->l2t); return CPL_RET_BUF_DONE; } @@ -1554,7 +1545,7 @@ static int peer_abort(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) } spin_lock_irqsave(&ep->com.lock, flags); - PDBG("%s ep %p state %u\n", __func__, ep, ep->com.state); + pr_debug("%s ep %p state %u\n", __func__, ep, ep->com.state); switch (ep->com.state) { case CONNECTING: break; @@ -1568,7 +1559,7 @@ static int peer_abort(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) case MPA_REP_SENT: ep->com.rpl_done = 1; ep->com.rpl_err = -ECONNRESET; - PDBG("waking up ep %p\n", ep); + pr_debug("waking up ep %p\n", ep); wake_up(&ep->com.waitq); break; case MPA_REQ_RCVD: @@ -1581,7 +1572,7 @@ static int peer_abort(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) */ ep->com.rpl_done = 1; ep->com.rpl_err = -ECONNRESET; - PDBG("waking up ep %p\n", ep); + pr_debug("waking up ep %p\n", ep); wake_up(&ep->com.waitq); break; case MORIBUND: @@ -1595,16 +1586,14 @@ static int peer_abort(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) ep->com.qp, IWCH_QP_ATTR_NEXT_STATE, &attrs, 1); if (ret) - printk(KERN_ERR MOD - "%s - qp <- error failed!\n", - __func__); + pr_err("%s - qp <- error failed!\n", __func__); } peer_abort_upcall(ep); break; case ABORTING: break; case DEAD: - PDBG("%s PEER_ABORT IN DEAD STATE!!!!\n", __func__); + pr_debug("%s PEER_ABORT IN DEAD STATE!!!!\n", __func__); spin_unlock_irqrestore(&ep->com.lock, flags); return CPL_RET_BUF_DONE; default: @@ -1620,13 +1609,12 @@ static int peer_abort(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) rpl_skb = get_skb(skb, sizeof(*rpl), GFP_KERNEL); if (!rpl_skb) { - printk(KERN_ERR MOD "%s - cannot allocate skb!\n", - __func__); + pr_err("%s - cannot allocate skb!\n", __func__); release = 1; goto out; } rpl_skb->priority = CPL_PRIORITY_DATA; - rpl = (struct cpl_abort_rpl *) skb_put(rpl_skb, sizeof(*rpl)); + rpl = skb_put(rpl_skb, sizeof(*rpl)); rpl->wr.wr_hi = htonl(V_WR_OP(FW_WROPCODE_OFLD_HOST_ABORT_CON_RPL)); rpl->wr.wr_lo = htonl(V_WR_TID(ep->hwtid)); OPCODE_TID(rpl) = htonl(MK_OPCODE_TID(CPL_ABORT_RPL, ep->hwtid)); @@ -1645,7 +1633,7 @@ static int close_con_rpl(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) unsigned long flags; int release = 0; - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); BUG_ON(!ep); /* The cm_id may be null if we failed to connect */ @@ -1699,9 +1687,9 @@ static int terminate(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) if (state_read(&ep->com) != FPDU_MODE) return CPL_RET_BUF_DONE; - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); skb_pull(skb, sizeof(struct cpl_rdma_terminate)); - PDBG("%s saving %d bytes of term msg\n", __func__, skb->len); + pr_debug("%s saving %d bytes of term msg\n", __func__, skb->len); skb_copy_from_linear_data(skb, ep->com.qp->attr.terminate_buffer, skb->len); ep->com.qp->attr.terminate_msg_len = skb->len; @@ -1714,12 +1702,12 @@ static int ec_status(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) struct cpl_rdma_ec_status *rep = cplhdr(skb); struct iwch_ep *ep = ctx; - PDBG("%s ep %p tid %u status %d\n", __func__, ep, ep->hwtid, - rep->status); + pr_debug("%s ep %p tid %u status %d\n", __func__, ep, ep->hwtid, + rep->status); if (rep->status) { struct iwch_qp_attributes attrs; - printk(KERN_ERR MOD "%s BAD CLOSE - Aborting tid %u\n", + pr_err("%s BAD CLOSE - Aborting tid %u\n", __func__, ep->hwtid); stop_ep_timer(ep); attrs.next_state = IWCH_QP_STATE_ERROR; @@ -1739,8 +1727,8 @@ static void ep_timeout(unsigned long arg) int abort = 1; spin_lock_irqsave(&ep->com.lock, flags); - PDBG("%s ep %p tid %u state %d\n", __func__, ep, ep->hwtid, - ep->com.state); + pr_debug("%s ep %p tid %u state %d\n", __func__, ep, ep->hwtid, + ep->com.state); switch (ep->com.state) { case MPA_REQ_SENT: __state_set(&ep->com, ABORTING); @@ -1774,7 +1762,7 @@ int iwch_reject_cr(struct iw_cm_id *cm_id, const void *pdata, u8 pdata_len) { int err; struct iwch_ep *ep = to_ep(cm_id); - PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); + pr_debug("%s ep %p tid %u\n", __func__, ep, ep->hwtid); if (state_read(&ep->com) == DEAD) { put_ep(&ep->com); @@ -1800,7 +1788,7 @@ int iwch_accept_cr(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) struct iwch_dev *h = to_iwch_dev(cm_id->device); struct iwch_qp *qp = get_qhp(h, conn_param->qpn); - PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); + pr_debug("%s ep %p tid %u\n", __func__, ep, ep->hwtid); if (state_read(&ep->com) == DEAD) { err = -ECONNRESET; goto err; @@ -1826,7 +1814,7 @@ int iwch_accept_cr(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) if (peer2peer && ep->ird == 0) ep->ird = 1; - PDBG("%s %d ird %d ord %d\n", __func__, __LINE__, ep->ird, ep->ord); + pr_debug("%s %d ird %d ord %d\n", __func__, __LINE__, ep->ird, ep->ord); /* bind QP to EP and move to RTS */ attrs.mpa_attr = ep->mpa_attr; @@ -1907,7 +1895,7 @@ int iwch_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) ep = alloc_ep(sizeof(*ep), GFP_KERNEL); if (!ep) { - printk(KERN_ERR MOD "%s - cannot alloc ep.\n", __func__); + pr_err("%s - cannot alloc ep\n", __func__); err = -ENOMEM; goto out; } @@ -1928,15 +1916,15 @@ int iwch_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) ep->com.cm_id = cm_id; ep->com.qp = get_qhp(h, conn_param->qpn); BUG_ON(!ep->com.qp); - PDBG("%s qpn 0x%x qp %p cm_id %p\n", __func__, conn_param->qpn, - ep->com.qp, cm_id); + pr_debug("%s qpn 0x%x qp %p cm_id %p\n", __func__, conn_param->qpn, + ep->com.qp, cm_id); /* * Allocate an active TID to initiate a TCP connection. */ ep->atid = cxgb3_alloc_atid(h->rdev.t3cdev_p, &t3c_client, ep); if (ep->atid == -1) { - printk(KERN_ERR MOD "%s - cannot alloc atid.\n", __func__); + pr_err("%s - cannot alloc atid\n", __func__); err = -ENOMEM; goto fail2; } @@ -1946,7 +1934,7 @@ int iwch_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) raddr->sin_addr.s_addr, laddr->sin_port, raddr->sin_port, IPTOS_LOWDELAY); if (!rt) { - printk(KERN_ERR MOD "%s - cannot find route.\n", __func__); + pr_err("%s - cannot find route\n", __func__); err = -EHOSTUNREACH; goto fail3; } @@ -1954,7 +1942,7 @@ int iwch_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) ep->l2t = t3_l2t_get(ep->com.tdev, ep->dst, NULL, &raddr->sin_addr.s_addr); if (!ep->l2t) { - printk(KERN_ERR MOD "%s - cannot alloc l2e.\n", __func__); + pr_err("%s - cannot alloc l2e\n", __func__); err = -ENOMEM; goto fail4; } @@ -1999,11 +1987,11 @@ int iwch_create_listen(struct iw_cm_id *cm_id, int backlog) ep = alloc_ep(sizeof(*ep), GFP_KERNEL); if (!ep) { - printk(KERN_ERR MOD "%s - cannot alloc ep.\n", __func__); + pr_err("%s - cannot alloc ep\n", __func__); err = -ENOMEM; goto fail1; } - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); ep->com.tdev = h->rdev.t3cdev_p; cm_id->add_ref(cm_id); ep->com.cm_id = cm_id; @@ -2016,7 +2004,7 @@ int iwch_create_listen(struct iw_cm_id *cm_id, int backlog) */ ep->stid = cxgb3_alloc_stid(h->rdev.t3cdev_p, &t3c_client, ep); if (ep->stid == -1) { - printk(KERN_ERR MOD "%s - cannot alloc atid.\n", __func__); + pr_err("%s - cannot alloc atid\n", __func__); err = -ENOMEM; goto fail2; } @@ -2048,7 +2036,7 @@ int iwch_destroy_listen(struct iw_cm_id *cm_id) int err; struct iwch_listen_ep *ep = to_listen_ep(cm_id); - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); might_sleep(); state_set(&ep->com, DEAD); @@ -2077,8 +2065,8 @@ int iwch_ep_disconnect(struct iwch_ep *ep, int abrupt, gfp_t gfp) spin_lock_irqsave(&ep->com.lock, flags); - PDBG("%s ep %p state %s, abrupt %d\n", __func__, ep, - states[ep->com.state], abrupt); + pr_debug("%s ep %p state %s, abrupt %d\n", __func__, ep, + states[ep->com.state], abrupt); tdev = (struct t3cdev *)ep->com.tdev; rdev = (struct cxio_rdev *)tdev->ulp; @@ -2115,8 +2103,8 @@ int iwch_ep_disconnect(struct iwch_ep *ep, int abrupt, gfp_t gfp) case MORIBUND: case ABORTING: case DEAD: - PDBG("%s ignoring disconnect ep %p state %u\n", - __func__, ep, ep->com.state); + pr_debug("%s ignoring disconnect ep %p state %u\n", + __func__, ep, ep->com.state); break; default: BUG(); @@ -2145,8 +2133,8 @@ int iwch_ep_redirect(void *ctx, struct dst_entry *old, struct dst_entry *new, if (ep->dst != old) return 0; - PDBG("%s ep %p redirect to dst %p l2t %p\n", __func__, ep, new, - l2t); + pr_debug("%s ep %p redirect to dst %p l2t %p\n", __func__, ep, new, + l2t); dst_hold(new); l2t_release(ep->com.tdev, ep->l2t); ep->l2t = l2t; @@ -2225,8 +2213,8 @@ static int set_tcb_rpl(struct t3cdev *tdev, struct sk_buff *skb, void *ctx) struct cpl_set_tcb_rpl *rpl = cplhdr(skb); if (rpl->status != CPL_ERR_NONE) { - printk(KERN_ERR MOD "Unexpected SET_TCB_RPL status %u " - "for tid %u\n", rpl->status, GET_TID(rpl)); + pr_err("Unexpected SET_TCB_RPL status %u for tid %u\n", + rpl->status, GET_TID(rpl)); } return CPL_RET_BUF_DONE; } diff --git a/drivers/infiniband/hw/cxgb3/iwch_cm.h b/drivers/infiniband/hw/cxgb3/iwch_cm.h index b9efadfffb4f..cc7fe644d260 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_cm.h +++ b/drivers/infiniband/hw/cxgb3/iwch_cm.h @@ -53,17 +53,17 @@ #define MPA_MARKERS 0x80 #define MPA_FLAGS_MASK 0xE0 -#define put_ep(ep) { \ - PDBG("put_ep (via %s:%u) ep %p refcnt %d\n", __func__, __LINE__, \ - ep, atomic_read(&((ep)->kref.refcount))); \ - WARN_ON(atomic_read(&((ep)->kref.refcount)) < 1); \ - kref_put(&((ep)->kref), __free_ep); \ +#define put_ep(ep) { \ + pr_debug("put_ep (via %s:%u) ep %p refcnt %d\n", \ + __func__, __LINE__, ep, kref_read(&((ep)->kref))); \ + WARN_ON(kref_read(&((ep)->kref)) < 1); \ + kref_put(&((ep)->kref), __free_ep); \ } -#define get_ep(ep) { \ - PDBG("get_ep (via %s:%u) ep %p, refcnt %d\n", __func__, __LINE__, \ - ep, atomic_read(&((ep)->kref.refcount))); \ - kref_get(&((ep)->kref)); \ +#define get_ep(ep) { \ + pr_debug("get_ep (via %s:%u) ep %p, refcnt %d\n", \ + __func__, __LINE__, ep, kref_read(&((ep)->kref))); \ + kref_get(&((ep)->kref)); \ } struct mpa_message { diff --git a/drivers/infiniband/hw/cxgb3/iwch_cq.c b/drivers/infiniband/hw/cxgb3/iwch_cq.c index 97fbfd2c298e..dd5348e48806 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_cq.c +++ b/drivers/infiniband/hw/cxgb3/iwch_cq.c @@ -67,8 +67,8 @@ static int iwch_poll_cq_one(struct iwch_dev *rhp, struct iwch_cq *chp, ret = cxio_poll_cq(wq, &(chp->cq), &cqe, &cqe_flushed, &cookie, &credit); if (t3a_device(chp->rhp) && credit) { - PDBG("%s updating %d cq credits on id %d\n", __func__, - credit, chp->cq.cqid); + pr_debug("%s updating %d cq credits on id %d\n", __func__, + credit, chp->cq.cqid); cxio_hal_cq_op(&rhp->rdev, &chp->cq, CQ_CREDIT_UPDATE, credit); } @@ -83,11 +83,11 @@ static int iwch_poll_cq_one(struct iwch_dev *rhp, struct iwch_cq *chp, wc->vendor_err = CQE_STATUS(cqe); wc->wc_flags = 0; - PDBG("%s qpid 0x%x type %d opcode %d status 0x%x wrid hi 0x%x " - "lo 0x%x cookie 0x%llx\n", __func__, - CQE_QPID(cqe), CQE_TYPE(cqe), - CQE_OPCODE(cqe), CQE_STATUS(cqe), CQE_WRID_HI(cqe), - CQE_WRID_LOW(cqe), (unsigned long long) cookie); + pr_debug("%s qpid 0x%x type %d opcode %d status 0x%x wrid hi 0x%x lo 0x%x cookie 0x%llx\n", + __func__, + CQE_QPID(cqe), CQE_TYPE(cqe), + CQE_OPCODE(cqe), CQE_STATUS(cqe), CQE_WRID_HI(cqe), + CQE_WRID_LOW(cqe), (unsigned long long)cookie); if (CQE_TYPE(cqe) == 0) { if (!CQE_STATUS(cqe)) @@ -122,8 +122,7 @@ static int iwch_poll_cq_one(struct iwch_dev *rhp, struct iwch_cq *chp, wc->opcode = IB_WC_REG_MR; break; default: - printk(KERN_ERR MOD "Unexpected opcode %d " - "in the CQE received for QPID=0x%0x\n", + pr_err("Unexpected opcode %d in the CQE received for QPID=0x%0x\n", CQE_OPCODE(cqe), CQE_QPID(cqe)); ret = -EINVAL; goto out; @@ -177,8 +176,8 @@ static int iwch_poll_cq_one(struct iwch_dev *rhp, struct iwch_cq *chp, wc->status = IB_WC_WR_FLUSH_ERR; break; default: - printk(KERN_ERR MOD "Unexpected cqe_status 0x%x for " - "QPID=0x%0x\n", CQE_STATUS(cqe), CQE_QPID(cqe)); + pr_err("Unexpected cqe_status 0x%x for QPID=0x%0x\n", + CQE_STATUS(cqe), CQE_QPID(cqe)); ret = -EINVAL; } } diff --git a/drivers/infiniband/hw/cxgb3/iwch_ev.c b/drivers/infiniband/hw/cxgb3/iwch_ev.c index abcc9e76962b..4a0c82a8fb60 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_ev.c +++ b/drivers/infiniband/hw/cxgb3/iwch_ev.c @@ -52,7 +52,7 @@ static void post_qp_event(struct iwch_dev *rnicp, struct iwch_cq *chp, qhp = get_qhp(rnicp, CQE_QPID(rsp_msg->cqe)); if (!qhp) { - printk(KERN_ERR "%s unaffiliated error 0x%x qpid 0x%x\n", + pr_err("%s unaffiliated error 0x%x qpid 0x%x\n", __func__, CQE_STATUS(rsp_msg->cqe), CQE_QPID(rsp_msg->cqe)); spin_unlock(&rnicp->lock); @@ -61,15 +61,16 @@ static void post_qp_event(struct iwch_dev *rnicp, struct iwch_cq *chp, if ((qhp->attr.state == IWCH_QP_STATE_ERROR) || (qhp->attr.state == IWCH_QP_STATE_TERMINATE)) { - PDBG("%s AE received after RTS - " - "qp state %d qpid 0x%x status 0x%x\n", __func__, - qhp->attr.state, qhp->wq.qpid, CQE_STATUS(rsp_msg->cqe)); + pr_debug("%s AE received after RTS - qp state %d qpid 0x%x status 0x%x\n", + __func__, + qhp->attr.state, qhp->wq.qpid, + CQE_STATUS(rsp_msg->cqe)); spin_unlock(&rnicp->lock); return; } - printk(KERN_ERR "%s - AE qpid 0x%x opcode %d status 0x%x " - "type %d wrid.hi 0x%x wrid.lo 0x%x \n", __func__, + pr_err("%s - AE qpid 0x%x opcode %d status 0x%x type %d wrid.hi 0x%x wrid.lo 0x%x\n", + __func__, CQE_QPID(rsp_msg->cqe), CQE_OPCODE(rsp_msg->cqe), CQE_STATUS(rsp_msg->cqe), CQE_TYPE(rsp_msg->cqe), CQE_WRID_HI(rsp_msg->cqe), CQE_WRID_LOW(rsp_msg->cqe)); @@ -117,8 +118,7 @@ void iwch_ev_dispatch(struct cxio_rdev *rdev_p, struct sk_buff *skb) chp = get_chp(rnicp, cqid); qhp = get_qhp(rnicp, CQE_QPID(rsp_msg->cqe)); if (!chp || !qhp) { - printk(KERN_ERR MOD "BAD AE cqid 0x%x qpid 0x%x opcode %d " - "status 0x%x type %d wrid.hi 0x%x wrid.lo 0x%x \n", + pr_err("BAD AE cqid 0x%x qpid 0x%x opcode %d status 0x%x type %d wrid.hi 0x%x wrid.lo 0x%x\n", cqid, CQE_QPID(rsp_msg->cqe), CQE_OPCODE(rsp_msg->cqe), CQE_STATUS(rsp_msg->cqe), CQE_TYPE(rsp_msg->cqe), CQE_WRID_HI(rsp_msg->cqe), @@ -137,12 +137,12 @@ void iwch_ev_dispatch(struct cxio_rdev *rdev_p, struct sk_buff *skb) if ((CQE_OPCODE(rsp_msg->cqe) == T3_TERMINATE) && (CQE_STATUS(rsp_msg->cqe) == 0)) { if (SQ_TYPE(rsp_msg->cqe)) { - PDBG("%s QPID 0x%x ep %p disconnecting\n", - __func__, qhp->wq.qpid, qhp->ep); + pr_debug("%s QPID 0x%x ep %p disconnecting\n", + __func__, qhp->wq.qpid, qhp->ep); iwch_ep_disconnect(qhp->ep, 0, GFP_ATOMIC); } else { - PDBG("%s post REQ_ERR AE QPID 0x%x\n", __func__, - qhp->wq.qpid); + pr_debug("%s post REQ_ERR AE QPID 0x%x\n", __func__, + qhp->wq.qpid); post_qp_event(rnicp, chp, rsp_msg, IB_EVENT_QP_REQ_ERR, 0); iwch_ep_disconnect(qhp->ep, 0, GFP_ATOMIC); @@ -218,7 +218,7 @@ void iwch_ev_dispatch(struct cxio_rdev *rdev_p, struct sk_buff *skb) break; default: - printk(KERN_ERR MOD "Unknown T3 status 0x%x QPID 0x%x\n", + pr_err("Unknown T3 status 0x%x QPID 0x%x\n", CQE_STATUS(rsp_msg->cqe), qhp->wq.qpid); post_qp_event(rnicp, chp, rsp_msg, IB_EVENT_QP_FATAL, 1); break; diff --git a/drivers/infiniband/hw/cxgb3/iwch_mem.c b/drivers/infiniband/hw/cxgb3/iwch_mem.c index 1d04c872c9d5..12886b1b4b10 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_mem.c +++ b/drivers/infiniband/hw/cxgb3/iwch_mem.c @@ -48,7 +48,7 @@ static int iwch_finish_mem_reg(struct iwch_mr *mhp, u32 stag) mhp->attr.stag = stag; mmid = stag >> 8; mhp->ibmr.rkey = mhp->ibmr.lkey = stag; - PDBG("%s mmid 0x%x mhp %p\n", __func__, mmid, mhp); + pr_debug("%s mmid 0x%x mhp %p\n", __func__, mmid, mhp); return insert_handle(mhp->rhp, &mhp->rhp->mmidr, mhp, mmid); } diff --git a/drivers/infiniband/hw/cxgb3/iwch_provider.c b/drivers/infiniband/hw/cxgb3/iwch_provider.c index 6262dc035f3c..0cd0c1fa27d4 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_provider.c +++ b/drivers/infiniband/hw/cxgb3/iwch_provider.c @@ -37,7 +37,7 @@ #include <linux/delay.h> #include <linux/errno.h> #include <linux/list.h> -#include <linux/sched.h> +#include <linux/sched/mm.h> #include <linux/spinlock.h> #include <linux/ethtool.h> #include <linux/rtnetlink.h> @@ -62,7 +62,7 @@ #include "common.h" static struct ib_ah *iwch_ah_create(struct ib_pd *pd, - struct ib_ah_attr *ah_attr, + struct rdma_ah_attr *ah_attr, struct ib_udata *udata) { return ERR_PTR(-ENOSYS); @@ -103,7 +103,7 @@ static int iwch_dealloc_ucontext(struct ib_ucontext *context) struct iwch_ucontext *ucontext = to_iwch_ucontext(context); struct iwch_mm_entry *mm, *tmp; - PDBG("%s context %p\n", __func__, context); + pr_debug("%s context %p\n", __func__, context); list_for_each_entry_safe(mm, tmp, &ucontext->mmaps, entry) kfree(mm); cxio_release_ucontext(&rhp->rdev, &ucontext->uctx); @@ -117,7 +117,7 @@ static struct ib_ucontext *iwch_alloc_ucontext(struct ib_device *ibdev, struct iwch_ucontext *context; struct iwch_dev *rhp = to_iwch_dev(ibdev); - PDBG("%s ibdev %p\n", __func__, ibdev); + pr_debug("%s ibdev %p\n", __func__, ibdev); context = kzalloc(sizeof(*context), GFP_KERNEL); if (!context) return ERR_PTR(-ENOMEM); @@ -131,7 +131,7 @@ static int iwch_destroy_cq(struct ib_cq *ib_cq) { struct iwch_cq *chp; - PDBG("%s ib_cq %p\n", __func__, ib_cq); + pr_debug("%s ib_cq %p\n", __func__, ib_cq); chp = to_iwch_cq(ib_cq); remove_handle(chp->rhp, &chp->rhp->cqidr, chp->cq.cqid); @@ -157,7 +157,7 @@ static struct ib_cq *iwch_create_cq(struct ib_device *ibdev, static int warned; size_t resplen; - PDBG("%s ib_dev %p entries %d\n", __func__, ibdev, entries); + pr_debug("%s ib_dev %p entries %d\n", __func__, ibdev, entries); if (attr->flags) return ERR_PTR(-EINVAL); @@ -227,8 +227,7 @@ static struct ib_cq *iwch_create_cq(struct ib_device *ibdev, mm->addr = virt_to_phys(chp->cq.queue); if (udata->outlen < sizeof uresp) { if (!warned++) - printk(KERN_WARNING MOD "Warning - " - "downlevel libcxgb3 (non-fatal).\n"); + pr_warn("Warning - downlevel libcxgb3 (non-fatal)\n"); mm->len = PAGE_ALIGN((1UL << uresp.size_log2) * sizeof(struct t3_cqe)); resplen = sizeof(struct iwch_create_cq_resp_v0); @@ -246,9 +245,9 @@ static struct ib_cq *iwch_create_cq(struct ib_device *ibdev, } insert_mmap(ucontext, mm); } - PDBG("created cqid 0x%0x chp %p size 0x%0x, dma_addr 0x%0llx\n", - chp->cq.cqid, chp, (1 << chp->cq.size_log2), - (unsigned long long) chp->cq.dma_addr); + pr_debug("created cqid 0x%0x chp %p size 0x%0x, dma_addr 0x%0llx\n", + chp->cq.cqid, chp, (1 << chp->cq.size_log2), + (unsigned long long)chp->cq.dma_addr); return &chp->ibcq; } @@ -259,7 +258,7 @@ static int iwch_resize_cq(struct ib_cq *cq, int cqe, struct ib_udata *udata) struct t3_cq oldcq, newcq; int ret; - PDBG("%s ib_cq %p cqe %d\n", __func__, cq, cqe); + pr_debug("%s ib_cq %p cqe %d\n", __func__, cq, cqe); /* We don't downsize... */ if (cqe <= cq->cqe) @@ -306,8 +305,7 @@ static int iwch_resize_cq(struct ib_cq *cq, int cqe, struct ib_udata *udata) oldcq.cqid = newcq.cqid; ret = cxio_destroy_cq(&chp->rhp->rdev, &oldcq); if (ret) { - printk(KERN_ERR MOD "%s - cxio_destroy_cq failed %d\n", - __func__, ret); + pr_err("%s - cxio_destroy_cq failed %d\n", __func__, ret); } /* add user hooks here */ @@ -342,12 +340,11 @@ static int iwch_arm_cq(struct ib_cq *ibcq, enum ib_cq_notify_flags flags) chp->cq.rptr = rptr; } else spin_lock_irqsave(&chp->lock, flag); - PDBG("%s rptr 0x%x\n", __func__, chp->cq.rptr); + pr_debug("%s rptr 0x%x\n", __func__, chp->cq.rptr); err = cxio_hal_cq_op(&rhp->rdev, &chp->cq, cq_op, 0); spin_unlock_irqrestore(&chp->lock, flag); if (err < 0) - printk(KERN_ERR MOD "Error %d rearming CQID 0x%x\n", err, - chp->cq.cqid); + pr_err("Error %d rearming CQID 0x%x\n", err, chp->cq.cqid); if (err > 0 && !(flags & IB_CQ_REPORT_MISSED_EVENTS)) err = 0; return err; @@ -363,8 +360,8 @@ static int iwch_mmap(struct ib_ucontext *context, struct vm_area_struct *vma) struct iwch_ucontext *ucontext; u64 addr; - PDBG("%s pgoff 0x%lx key 0x%x len %d\n", __func__, vma->vm_pgoff, - key, len); + pr_debug("%s pgoff 0x%lx key 0x%x len %d\n", __func__, vma->vm_pgoff, + key, len); if (vma->vm_start & (PAGE_SIZE-1)) { return -EINVAL; @@ -416,7 +413,7 @@ static int iwch_deallocate_pd(struct ib_pd *pd) php = to_iwch_pd(pd); rhp = php->rhp; - PDBG("%s ibpd %p pdid 0x%x\n", __func__, pd, php->pdid); + pr_debug("%s ibpd %p pdid 0x%x\n", __func__, pd, php->pdid); cxio_hal_put_pdid(rhp->rdev.rscp, php->pdid); kfree(php); return 0; @@ -430,7 +427,7 @@ static struct ib_pd *iwch_allocate_pd(struct ib_device *ibdev, u32 pdid; struct iwch_dev *rhp; - PDBG("%s ibdev %p\n", __func__, ibdev); + pr_debug("%s ibdev %p\n", __func__, ibdev); rhp = (struct iwch_dev *) ibdev; pdid = cxio_hal_get_pdid(rhp->rdev.rscp); if (!pdid) @@ -448,7 +445,7 @@ static struct ib_pd *iwch_allocate_pd(struct ib_device *ibdev, return ERR_PTR(-EFAULT); } } - PDBG("%s pdid 0x%0x ptr 0x%p\n", __func__, pdid, php); + pr_debug("%s pdid 0x%0x ptr 0x%p\n", __func__, pdid, php); return &php->ibpd; } @@ -458,7 +455,7 @@ static int iwch_dereg_mr(struct ib_mr *ib_mr) struct iwch_mr *mhp; u32 mmid; - PDBG("%s ib_mr %p\n", __func__, ib_mr); + pr_debug("%s ib_mr %p\n", __func__, ib_mr); mhp = to_iwch_mr(ib_mr); kfree(mhp->pages); @@ -472,7 +469,7 @@ static int iwch_dereg_mr(struct ib_mr *ib_mr) kfree((void *) (unsigned long) mhp->kva); if (mhp->umem) ib_umem_release(mhp->umem); - PDBG("%s mmid 0x%x ptr %p\n", __func__, mmid, mhp); + pr_debug("%s mmid 0x%x ptr %p\n", __func__, mmid, mhp); kfree(mhp); return 0; } @@ -487,13 +484,13 @@ static struct ib_mr *iwch_get_dma_mr(struct ib_pd *pd, int acc) __be64 *page_list; int shift = 26, npages, ret, i; - PDBG("%s ib_pd %p\n", __func__, pd); + pr_debug("%s ib_pd %p\n", __func__, pd); /* * T3 only supports 32 bits of size. */ if (sizeof(phys_addr_t) > 4) { - pr_warn_once(MOD "Cannot support dma_mrs on this platform.\n"); + pr_warn_once("Cannot support dma_mrs on this platform\n"); return ERR_PTR(-ENOTSUPP); } @@ -518,8 +515,8 @@ static struct ib_mr *iwch_get_dma_mr(struct ib_pd *pd, int acc) for (i = 0; i < npages; i++) page_list[i] = cpu_to_be64((u64)i << shift); - PDBG("%s mask 0x%llx shift %d len %lld pbl_size %d\n", - __func__, mask, shift, total_size, npages); + pr_debug("%s mask 0x%llx shift %d len %lld pbl_size %d\n", + __func__, mask, shift, total_size, npages); ret = iwch_alloc_pbl(mhp, npages); if (ret) { @@ -567,7 +564,7 @@ static struct ib_mr *iwch_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, struct iwch_mr *mhp; struct iwch_reg_user_mr_resp uresp; struct scatterlist *sg; - PDBG("%s ib_pd %p\n", __func__, pd); + pr_debug("%s ib_pd %p\n", __func__, pd); php = to_iwch_pd(pd); rhp = php->rhp; @@ -584,7 +581,7 @@ static struct ib_mr *iwch_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, return ERR_PTR(err); } - shift = ffs(mhp->umem->page_size) - 1; + shift = mhp->umem->page_shift; n = mhp->umem->nmap; @@ -604,7 +601,7 @@ static struct ib_mr *iwch_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, len = sg_dma_len(sg) >> shift; for (k = 0; k < len; ++k) { pages[i++] = cpu_to_be64(sg_dma_address(sg) + - mhp->umem->page_size * k); + (k << shift)); if (i == PAGE_SIZE / sizeof *pages) { err = iwch_write_pbl(mhp, pages, i, n); if (err) @@ -637,8 +634,8 @@ pbl_done: if (udata && !t3a_device(rhp)) { uresp.pbl_addr = (mhp->attr.pbl_addr - rhp->rdev.rnic_info.pbl_base) >> 3; - PDBG("%s user resp pbl_addr 0x%x\n", __func__, - uresp.pbl_addr); + pr_debug("%s user resp pbl_addr 0x%x\n", __func__, + uresp.pbl_addr); if (ib_copy_to_udata(udata, &uresp, sizeof (uresp))) { iwch_dereg_mr(&mhp->ibmr); @@ -692,7 +689,7 @@ static struct ib_mw *iwch_alloc_mw(struct ib_pd *pd, enum ib_mw_type type, kfree(mhp); return ERR_PTR(-ENOMEM); } - PDBG("%s mmid 0x%x mhp %p stag 0x%x\n", __func__, mmid, mhp, stag); + pr_debug("%s mmid 0x%x mhp %p stag 0x%x\n", __func__, mmid, mhp, stag); return &(mhp->ibmw); } @@ -707,7 +704,7 @@ static int iwch_dealloc_mw(struct ib_mw *mw) mmid = (mw->rkey) >> 8; cxio_deallocate_window(&rhp->rdev, mhp->attr.stag); remove_handle(rhp, &rhp->mmidr, mmid); - PDBG("%s ib_mw %p mmid 0x%x ptr %p\n", __func__, mw, mmid, mhp); + pr_debug("%s ib_mw %p mmid 0x%x ptr %p\n", __func__, mw, mmid, mhp); kfree(mhp); return 0; } @@ -721,7 +718,7 @@ static struct ib_mr *iwch_alloc_mr(struct ib_pd *pd, struct iwch_mr *mhp; u32 mmid; u32 stag = 0; - int ret = 0; + int ret = -ENOMEM; if (mr_type != IB_MR_TYPE_MEM_REG || max_num_sg > T3_MAX_FASTREG_DEPTH) @@ -734,10 +731,8 @@ static struct ib_mr *iwch_alloc_mr(struct ib_pd *pd, goto err; mhp->pages = kcalloc(max_num_sg, sizeof(u64), GFP_KERNEL); - if (!mhp->pages) { - ret = -ENOMEM; + if (!mhp->pages) goto pl_err; - } mhp->rhp = rhp; ret = iwch_alloc_pbl(mhp, max_num_sg); @@ -754,10 +749,11 @@ static struct ib_mr *iwch_alloc_mr(struct ib_pd *pd, mhp->attr.state = 1; mmid = (stag) >> 8; mhp->ibmr.rkey = mhp->ibmr.lkey = stag; - if (insert_handle(rhp, &rhp->mmidr, mhp, mmid)) + ret = insert_handle(rhp, &rhp->mmidr, mhp, mmid); + if (ret) goto err3; - PDBG("%s mmid 0x%x mhp %p stag 0x%x\n", __func__, mmid, mhp, stag); + pr_debug("%s mmid 0x%x mhp %p stag 0x%x\n", __func__, mmid, mhp, stag); return &(mhp->ibmr); err3: cxio_dereg_mem(&rhp->rdev, stag, mhp->attr.pbl_size, @@ -818,8 +814,8 @@ static int iwch_destroy_qp(struct ib_qp *ib_qp) cxio_destroy_qp(&rhp->rdev, &qhp->wq, ucontext ? &ucontext->uctx : &rhp->rdev.uctx); - PDBG("%s ib_qp %p qpid 0x%0x qhp %p\n", __func__, - ib_qp, qhp->wq.qpid, qhp); + pr_debug("%s ib_qp %p qpid 0x%0x qhp %p\n", __func__, + ib_qp, qhp->wq.qpid, qhp); kfree(qhp); return 0; } @@ -837,7 +833,7 @@ static struct ib_qp *iwch_create_qp(struct ib_pd *pd, int wqsize, sqsize, rqsize; struct iwch_ucontext *ucontext; - PDBG("%s ib_pd %p\n", __func__, pd); + pr_debug("%s ib_pd %p\n", __func__, pd); if (attrs->qp_type != IB_QPT_RC) return ERR_PTR(-EINVAL); php = to_iwch_pd(pd); @@ -878,8 +874,8 @@ static struct ib_qp *iwch_create_qp(struct ib_pd *pd, if (!ucontext && wqsize < (rqsize + (2 * sqsize))) wqsize = roundup_pow_of_two(rqsize + roundup_pow_of_two(attrs->cap.max_send_wr * 2)); - PDBG("%s wqsize %d sqsize %d rqsize %d\n", __func__, - wqsize, sqsize, rqsize); + pr_debug("%s wqsize %d sqsize %d rqsize %d\n", __func__, + wqsize, sqsize, rqsize); qhp = kzalloc(sizeof(*qhp), GFP_KERNEL); if (!qhp) return ERR_PTR(-ENOMEM); @@ -974,11 +970,10 @@ static struct ib_qp *iwch_create_qp(struct ib_pd *pd, } qhp->ibqp.qp_num = qhp->wq.qpid; init_timer(&(qhp->timer)); - PDBG("%s sq_num_entries %d, rq_num_entries %d " - "qpid 0x%0x qhp %p dma_addr 0x%llx size %d rq_addr 0x%x\n", - __func__, qhp->attr.sq_num_entries, qhp->attr.rq_num_entries, - qhp->wq.qpid, qhp, (unsigned long long) qhp->wq.dma_addr, - 1 << qhp->wq.size_log2, qhp->wq.rq_addr); + pr_debug("%s sq_num_entries %d, rq_num_entries %d qpid 0x%0x qhp %p dma_addr 0x%llx size %d rq_addr 0x%x\n", + __func__, qhp->attr.sq_num_entries, qhp->attr.rq_num_entries, + qhp->wq.qpid, qhp, (unsigned long long)qhp->wq.dma_addr, + 1 << qhp->wq.size_log2, qhp->wq.rq_addr); return &qhp->ibqp; } @@ -990,7 +985,7 @@ static int iwch_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, enum iwch_qp_attr_mask mask = 0; struct iwch_qp_attributes attrs; - PDBG("%s ib_qp %p\n", __func__, ibqp); + pr_debug("%s ib_qp %p\n", __func__, ibqp); /* iwarp does not support the RTR state */ if ((attr_mask & IB_QP_STATE) && (attr->qp_state == IB_QPS_RTR)) @@ -1023,20 +1018,20 @@ static int iwch_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, void iwch_qp_add_ref(struct ib_qp *qp) { - PDBG("%s ib_qp %p\n", __func__, qp); + pr_debug("%s ib_qp %p\n", __func__, qp); atomic_inc(&(to_iwch_qp(qp)->refcnt)); } void iwch_qp_rem_ref(struct ib_qp *qp) { - PDBG("%s ib_qp %p\n", __func__, qp); + pr_debug("%s ib_qp %p\n", __func__, qp); if (atomic_dec_and_test(&(to_iwch_qp(qp)->refcnt))) wake_up(&(to_iwch_qp(qp)->wait)); } static struct ib_qp *iwch_get_qp(struct ib_device *dev, int qpn) { - PDBG("%s ib_dev %p qpn 0x%x\n", __func__, dev, qpn); + pr_debug("%s ib_dev %p qpn 0x%x\n", __func__, dev, qpn); return (struct ib_qp *)get_qhp(to_iwch_dev(dev), qpn); } @@ -1044,7 +1039,7 @@ static struct ib_qp *iwch_get_qp(struct ib_device *dev, int qpn) static int iwch_query_pkey(struct ib_device *ibdev, u8 port, u16 index, u16 * pkey) { - PDBG("%s ibdev %p\n", __func__, ibdev); + pr_debug("%s ibdev %p\n", __func__, ibdev); *pkey = 0; return 0; } @@ -1054,8 +1049,8 @@ static int iwch_query_gid(struct ib_device *ibdev, u8 port, { struct iwch_dev *dev; - PDBG("%s ibdev %p, port %d, index %d, gid %p\n", - __func__, ibdev, port, index, gid); + pr_debug("%s ibdev %p, port %d, index %d, gid %p\n", + __func__, ibdev, port, index, gid); dev = to_iwch_dev(ibdev); BUG_ON(port == 0 || port > 2); memset(&(gid->raw[0]), 0, sizeof(gid->raw)); @@ -1090,7 +1085,7 @@ static int iwch_query_device(struct ib_device *ibdev, struct ib_device_attr *pro struct iwch_dev *dev; - PDBG("%s ibdev %p\n", __func__, ibdev); + pr_debug("%s ibdev %p\n", __func__, ibdev); if (uhw->inlen || uhw->outlen) return -EINVAL; @@ -1128,12 +1123,12 @@ static int iwch_query_port(struct ib_device *ibdev, struct net_device *netdev; struct in_device *inetdev; - PDBG("%s ibdev %p\n", __func__, ibdev); + pr_debug("%s ibdev %p\n", __func__, ibdev); dev = to_iwch_dev(ibdev); netdev = dev->rdev.port_info.lldevs[port-1]; - memset(props, 0, sizeof(struct ib_port_attr)); + /* props being zeroed by the caller, avoid zeroing it here */ props->max_mtu = IB_MTU_4096; props->active_mtu = ib_mtu_int_to_enum(netdev->mtu); @@ -1171,7 +1166,7 @@ static ssize_t show_rev(struct device *dev, struct device_attribute *attr, { struct iwch_dev *iwch_dev = container_of(dev, struct iwch_dev, ibdev.dev); - PDBG("%s dev 0x%p\n", __func__, dev); + pr_debug("%s dev 0x%p\n", __func__, dev); return sprintf(buf, "%d\n", iwch_dev->rdev.t3cdev_p->type); } @@ -1183,7 +1178,7 @@ static ssize_t show_hca(struct device *dev, struct device_attribute *attr, struct ethtool_drvinfo info; struct net_device *lldev = iwch_dev->rdev.t3cdev_p->lldev; - PDBG("%s dev 0x%p\n", __func__, dev); + pr_debug("%s dev 0x%p\n", __func__, dev); lldev->ethtool_ops->get_drvinfo(lldev, &info); return sprintf(buf, "%s\n", info.driver); } @@ -1193,7 +1188,7 @@ static ssize_t show_board(struct device *dev, struct device_attribute *attr, { struct iwch_dev *iwch_dev = container_of(dev, struct iwch_dev, ibdev.dev); - PDBG("%s dev 0x%p\n", __func__, dev); + pr_debug("%s dev 0x%p\n", __func__, dev); return sprintf(buf, "%x.%x\n", iwch_dev->rdev.rnic_info.pdev->vendor, iwch_dev->rdev.rnic_info.pdev->device); } @@ -1278,7 +1273,7 @@ static int iwch_get_mib(struct ib_device *ibdev, struct rdma_hw_stats *stats, if (port != 0 || !stats) return -ENOSYS; - PDBG("%s ibdev %p\n", __func__, ibdev); + pr_debug("%s ibdev %p\n", __func__, ibdev); dev = to_iwch_dev(ibdev); ret = dev->rdev.t3cdev_p->ctl(dev->rdev.t3cdev_p, RDMA_GET_MIB, &m); if (ret) @@ -1329,13 +1324,14 @@ static int iwch_port_immutable(struct ib_device *ibdev, u8 port_num, struct ib_port_attr attr; int err; - err = iwch_query_port(ibdev, port_num, &attr); + immutable->core_cap_flags = RDMA_CORE_PORT_IWARP; + + err = ib_query_port(ibdev, port_num, &attr); if (err) return err; immutable->pkey_tbl_len = attr.pkey_tbl_len; immutable->gid_tbl_len = attr.gid_tbl_len; - immutable->core_cap_flags = RDMA_CORE_PORT_IWARP; return 0; } @@ -1347,7 +1343,7 @@ static void get_dev_fw_ver_str(struct ib_device *ibdev, char *str, struct ethtool_drvinfo info; struct net_device *lldev = iwch_dev->rdev.t3cdev_p->lldev; - PDBG("%s dev 0x%p\n", __func__, iwch_dev); + pr_debug("%s dev 0x%p\n", __func__, iwch_dev); lldev->ethtool_ops->get_drvinfo(lldev, &info); snprintf(str, str_len, "%s", info.fw_version); } @@ -1357,7 +1353,7 @@ int iwch_register_device(struct iwch_dev *dev) int ret; int i; - PDBG("%s iwch_dev %p\n", __func__, dev); + pr_debug("%s iwch_dev %p\n", __func__, dev); strlcpy(dev->ibdev.name, "cxgb3_%d", IB_DEVICE_NAME_MAX); memset(&dev->ibdev.node_guid, 0, sizeof(dev->ibdev.node_guid)); memcpy(&dev->ibdev.node_guid, dev->rdev.t3cdev_p->lldev->dev_addr, 6); @@ -1392,7 +1388,7 @@ int iwch_register_device(struct iwch_dev *dev) memcpy(dev->ibdev.node_desc, IWCH_NODE_DESC, sizeof(IWCH_NODE_DESC)); dev->ibdev.phys_port_cnt = dev->rdev.port_info.nports; dev->ibdev.num_comp_vectors = 1; - dev->ibdev.dma_device = &(dev->rdev.rnic_info.pdev->dev); + dev->ibdev.dev.parent = &dev->rdev.rnic_info.pdev->dev; dev->ibdev.query_device = iwch_query_device; dev->ibdev.query_port = iwch_query_port; dev->ibdev.query_pkey = iwch_query_pkey; @@ -1468,7 +1464,7 @@ void iwch_unregister_device(struct iwch_dev *dev) { int i; - PDBG("%s iwch_dev %p\n", __func__, dev); + pr_debug("%s iwch_dev %p\n", __func__, dev); for (i = 0; i < ARRAY_SIZE(iwch_class_attributes); ++i) device_remove_file(&dev->ibdev.dev, iwch_class_attributes[i]); diff --git a/drivers/infiniband/hw/cxgb3/iwch_provider.h b/drivers/infiniband/hw/cxgb3/iwch_provider.h index 252c464a09f6..9e216edec4c0 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_provider.h +++ b/drivers/infiniband/hw/cxgb3/iwch_provider.h @@ -217,8 +217,9 @@ static inline struct iwch_mm_entry *remove_mmap(struct iwch_ucontext *ucontext, if (mm->key == key && mm->len == len) { list_del_init(&mm->entry); spin_unlock(&ucontext->mmap_lock); - PDBG("%s key 0x%x addr 0x%llx len %d\n", __func__, - key, (unsigned long long) mm->addr, mm->len); + pr_debug("%s key 0x%x addr 0x%llx len %d\n", + __func__, key, + (unsigned long long)mm->addr, mm->len); return mm; } } @@ -230,8 +231,8 @@ static inline void insert_mmap(struct iwch_ucontext *ucontext, struct iwch_mm_entry *mm) { spin_lock(&ucontext->mmap_lock); - PDBG("%s key 0x%x addr 0x%llx len %d\n", __func__, - mm->key, (unsigned long long) mm->addr, mm->len); + pr_debug("%s key 0x%x addr 0x%llx len %d\n", + __func__, mm->key, (unsigned long long)mm->addr, mm->len); list_add_tail(&mm->entry, &ucontext->mmaps); spin_unlock(&ucontext->mmap_lock); } diff --git a/drivers/infiniband/hw/cxgb3/iwch_qp.c b/drivers/infiniband/hw/cxgb3/iwch_qp.c index d939980a708f..7f633da0185d 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_qp.c +++ b/drivers/infiniband/hw/cxgb3/iwch_qp.c @@ -208,30 +208,30 @@ static int iwch_sgl2pbl_map(struct iwch_dev *rhp, struct ib_sge *sg_list, mhp = get_mhp(rhp, (sg_list[i].lkey) >> 8); if (!mhp) { - PDBG("%s %d\n", __func__, __LINE__); + pr_debug("%s %d\n", __func__, __LINE__); return -EIO; } if (!mhp->attr.state) { - PDBG("%s %d\n", __func__, __LINE__); + pr_debug("%s %d\n", __func__, __LINE__); return -EIO; } if (mhp->attr.zbva) { - PDBG("%s %d\n", __func__, __LINE__); + pr_debug("%s %d\n", __func__, __LINE__); return -EIO; } if (sg_list[i].addr < mhp->attr.va_fbo) { - PDBG("%s %d\n", __func__, __LINE__); + pr_debug("%s %d\n", __func__, __LINE__); return -EINVAL; } if (sg_list[i].addr + ((u64) sg_list[i].length) < sg_list[i].addr) { - PDBG("%s %d\n", __func__, __LINE__); + pr_debug("%s %d\n", __func__, __LINE__); return -EINVAL; } if (sg_list[i].addr + ((u64) sg_list[i].length) > mhp->attr.va_fbo + ((u64) mhp->attr.len)) { - PDBG("%s %d\n", __func__, __LINE__); + pr_debug("%s %d\n", __func__, __LINE__); return -EINVAL; } offset = sg_list[i].addr - mhp->attr.va_fbo; @@ -427,8 +427,8 @@ int iwch_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, err = build_inv_stag(wqe, wr, &t3_wr_flit_cnt); break; default: - PDBG("%s post of type=%d TBD!\n", __func__, - wr->opcode); + pr_debug("%s post of type=%d TBD!\n", __func__, + wr->opcode); err = -EINVAL; } if (err) @@ -444,10 +444,10 @@ int iwch_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, Q_GENBIT(qhp->wq.wptr, qhp->wq.size_log2), 0, t3_wr_flit_cnt, (wr_cnt == 1) ? T3_SOPEOP : T3_SOP); - PDBG("%s cookie 0x%llx wq idx 0x%x swsq idx %ld opcode %d\n", - __func__, (unsigned long long) wr->wr_id, idx, - Q_PTR2IDX(qhp->wq.sq_wptr, qhp->wq.sq_size_log2), - sqp->opcode); + pr_debug("%s cookie 0x%llx wq idx 0x%x swsq idx %ld opcode %d\n", + __func__, (unsigned long long)wr->wr_id, idx, + Q_PTR2IDX(qhp->wq.sq_wptr, qhp->wq.sq_size_log2), + sqp->opcode); wr = wr->next; num_wrs--; qhp->wq.wptr += wr_cnt; @@ -508,9 +508,9 @@ int iwch_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *wr, build_fw_riwrh((void *) wqe, T3_WR_RCV, T3_COMPLETION_FLAG, Q_GENBIT(qhp->wq.wptr, qhp->wq.size_log2), 0, sizeof(struct t3_receive_wr) >> 3, T3_SOPEOP); - PDBG("%s cookie 0x%llx idx 0x%x rq_wptr 0x%x rw_rptr 0x%x " - "wqe %p \n", __func__, (unsigned long long) wr->wr_id, - idx, qhp->wq.rq_wptr, qhp->wq.rq_rptr, wqe); + pr_debug("%s cookie 0x%llx idx 0x%x rq_wptr 0x%x rw_rptr 0x%x wqe %p\n", + __func__, (unsigned long long)wr->wr_id, + idx, qhp->wq.rq_wptr, qhp->wq.rq_rptr, wqe); ++(qhp->wq.rq_wptr); ++(qhp->wq.wptr); wr = wr->next; @@ -664,14 +664,13 @@ int iwch_post_zb_read(struct iwch_ep *ep) struct sk_buff *skb; u8 flit_cnt = sizeof(struct t3_rdma_read_wr) >> 3; - PDBG("%s enter\n", __func__); + pr_debug("%s enter\n", __func__); skb = alloc_skb(40, GFP_KERNEL); if (!skb) { - printk(KERN_ERR "%s cannot send zb_read!!\n", __func__); + pr_err("%s cannot send zb_read!!\n", __func__); return -ENOMEM; } - wqe = (union t3_wr *)skb_put(skb, sizeof(struct t3_rdma_read_wr)); - memset(wqe, 0, sizeof(struct t3_rdma_read_wr)); + wqe = skb_put_zero(skb, sizeof(struct t3_rdma_read_wr)); wqe->read.rdmaop = T3_READ_REQ; wqe->read.reserved[0] = 0; wqe->read.reserved[1] = 0; @@ -696,14 +695,13 @@ int iwch_post_terminate(struct iwch_qp *qhp, struct respQ_msg_t *rsp_msg) struct terminate_message *term; struct sk_buff *skb; - PDBG("%s %d\n", __func__, __LINE__); + pr_debug("%s %d\n", __func__, __LINE__); skb = alloc_skb(40, GFP_ATOMIC); if (!skb) { - printk(KERN_ERR "%s cannot send TERMINATE!\n", __func__); + pr_err("%s cannot send TERMINATE!\n", __func__); return -ENOMEM; } - wqe = (union t3_wr *)skb_put(skb, 40); - memset(wqe, 0, 40); + wqe = skb_put_zero(skb, 40); wqe->send.rdmaop = T3_TERMINATE; /* immediate data length */ @@ -729,7 +727,7 @@ static void __flush_qp(struct iwch_qp *qhp, struct iwch_cq *rchp, int flushed; - PDBG("%s qhp %p rchp %p schp %p\n", __func__, qhp, rchp, schp); + pr_debug("%s qhp %p rchp %p schp %p\n", __func__, qhp, rchp, schp); /* take a ref on the qhp since we must release the lock */ atomic_inc(&qhp->refcnt); spin_unlock(&qhp->lock); @@ -807,7 +805,7 @@ u16 iwch_rqes_posted(struct iwch_qp *qhp) count++; wqe++; } - PDBG("%s qhp %p count %u\n", __func__, qhp, count); + pr_debug("%s qhp %p count %u\n", __func__, qhp, count); return count; } @@ -854,12 +852,12 @@ static int rdma_init(struct iwch_dev *rhp, struct iwch_qp *qhp, } else init_attr.rtr_type = 0; init_attr.irs = qhp->ep->rcv_seq; - PDBG("%s init_attr.rq_addr 0x%x init_attr.rq_size = %d " - "flags 0x%x qpcaps 0x%x\n", __func__, - init_attr.rq_addr, init_attr.rq_size, - init_attr.flags, init_attr.qpcaps); + pr_debug("%s init_attr.rq_addr 0x%x init_attr.rq_size = %d flags 0x%x qpcaps 0x%x\n", + __func__, + init_attr.rq_addr, init_attr.rq_size, + init_attr.flags, init_attr.qpcaps); ret = cxio_rdma_init(&rhp->rdev, &init_attr); - PDBG("%s ret %d\n", __func__, ret); + pr_debug("%s ret %d\n", __func__, ret); return ret; } @@ -877,9 +875,9 @@ int iwch_modify_qp(struct iwch_dev *rhp, struct iwch_qp *qhp, int free = 0; struct iwch_ep *ep = NULL; - PDBG("%s qhp %p qpid 0x%x ep %p state %d -> %d\n", __func__, - qhp, qhp->wq.qpid, qhp->ep, qhp->attr.state, - (mask & IWCH_QP_ATTR_NEXT_STATE) ? attrs->next_state : -1); + pr_debug("%s qhp %p qpid 0x%x ep %p state %d -> %d\n", __func__, + qhp, qhp->wq.qpid, qhp->ep, qhp->attr.state, + (mask & IWCH_QP_ATTR_NEXT_STATE) ? attrs->next_state : -1); spin_lock_irqsave(&qhp->lock, flag); @@ -961,7 +959,7 @@ int iwch_modify_qp(struct iwch_dev *rhp, struct iwch_qp *qhp, case IWCH_QP_STATE_RTS: switch (attrs->next_state) { case IWCH_QP_STATE_CLOSING: - BUG_ON(atomic_read(&qhp->ep->com.kref.refcount) < 2); + BUG_ON(kref_read(&qhp->ep->com.kref) < 2); qhp->attr.state = IWCH_QP_STATE_CLOSING; if (!internal) { abort=0; @@ -1034,16 +1032,15 @@ int iwch_modify_qp(struct iwch_dev *rhp, struct iwch_qp *qhp, goto err; break; default: - printk(KERN_ERR "%s in a bad state %d\n", - __func__, qhp->attr.state); + pr_err("%s in a bad state %d\n", __func__, qhp->attr.state); ret = -EINVAL; goto err; break; } goto out; err: - PDBG("%s disassociating ep %p qpid 0x%x\n", __func__, qhp->ep, - qhp->wq.qpid); + pr_debug("%s disassociating ep %p qpid 0x%x\n", __func__, qhp->ep, + qhp->wq.qpid); /* disassociate the LLP connection */ qhp->attr.llp_stream_handle = NULL; @@ -1077,6 +1074,6 @@ out: if (free) put_ep(&ep->com); - PDBG("%s exit state %d\n", __func__, qhp->attr.state); + pr_debug("%s exit state %d\n", __func__, qhp->attr.state); return ret; } diff --git a/drivers/infiniband/hw/cxgb4/cm.c b/drivers/infiniband/hw/cxgb4/cm.c index 9398143d7c5e..e49b34c3b136 100644 --- a/drivers/infiniband/hw/cxgb4/cm.c +++ b/drivers/infiniband/hw/cxgb4/cm.c @@ -101,7 +101,7 @@ MODULE_PARM_DESC(enable_tcp_window_scaling, int c4iw_debug; module_param(c4iw_debug, int, 0644); -MODULE_PARM_DESC(c4iw_debug, "Enable debug logging (default=0)"); +MODULE_PARM_DESC(c4iw_debug, "obsolete"); static int peer2peer = 1; module_param(peer2peer, int, 0644); @@ -180,7 +180,7 @@ static void ref_qp(struct c4iw_ep *ep) static void start_ep_timer(struct c4iw_ep *ep) { - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); if (timer_pending(&ep->timer)) { pr_err("%s timer already started! ep %p\n", __func__, ep); @@ -196,7 +196,7 @@ static void start_ep_timer(struct c4iw_ep *ep) static int stop_ep_timer(struct c4iw_ep *ep) { - PDBG("%s ep %p stopping\n", __func__, ep); + pr_debug("%s ep %p stopping\n", __func__, ep); del_timer_sync(&ep->timer); if (!test_and_set_bit(TIMEOUT, &ep->com.flags)) { c4iw_put_ep(&ep->com); @@ -212,7 +212,7 @@ static int c4iw_l2t_send(struct c4iw_rdev *rdev, struct sk_buff *skb, if (c4iw_fatal_error(rdev)) { kfree_skb(skb); - PDBG("%s - device in error state - dropping\n", __func__); + pr_debug("%s - device in error state - dropping\n", __func__); return -EIO; } error = cxgb4_l2t_send(rdev->lldi.ports[0], skb, l2e); @@ -229,7 +229,7 @@ int c4iw_ofld_send(struct c4iw_rdev *rdev, struct sk_buff *skb) if (c4iw_fatal_error(rdev)) { kfree_skb(skb); - PDBG("%s - device in error state - dropping\n", __func__); + pr_debug("%s - device in error state - dropping\n", __func__); return -EIO; } error = cxgb4_ofld_send(rdev->lldi.ports[0], skb); @@ -263,10 +263,10 @@ static void set_emss(struct c4iw_ep *ep, u16 opt) if (ep->emss < 128) ep->emss = 128; if (ep->emss & 7) - PDBG("Warning: misaligned mtu idx %u mss %u emss=%u\n", - TCPOPT_MSS_G(opt), ep->mss, ep->emss); - PDBG("%s mss_idx %u mss %u emss=%u\n", __func__, 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("%s mss_idx %u mss %u emss=%u\n", __func__, TCPOPT_MSS_G(opt), + ep->mss, ep->emss); } static enum c4iw_ep_state state_read(struct c4iw_ep_common *epc) @@ -287,7 +287,7 @@ static void __state_set(struct c4iw_ep_common *epc, enum c4iw_ep_state new) static void state_set(struct c4iw_ep_common *epc, enum c4iw_ep_state new) { mutex_lock(&epc->mutex); - PDBG("%s - %s -> %s\n", __func__, states[epc->state], states[new]); + pr_debug("%s - %s -> %s\n", __func__, states[epc->state], states[new]); __state_set(epc, new); mutex_unlock(&epc->mutex); return; @@ -322,7 +322,7 @@ static void *alloc_ep(int size, gfp_t gfp) mutex_init(&epc->mutex); c4iw_init_wr_wait(&epc->wr_wait); } - PDBG("%s alloc ep %p\n", __func__, epc); + pr_debug("%s alloc ep %p\n", __func__, epc); return epc; } @@ -384,7 +384,7 @@ void _c4iw_free_ep(struct kref *kref) struct c4iw_ep *ep; ep = container_of(kref, struct c4iw_ep, com.kref); - PDBG("%s ep %p state %s\n", __func__, ep, states[ep->com.state]); + pr_debug("%s ep %p state %s\n", __func__, ep, states[ep->com.state]); if (test_bit(QP_REFERENCED, &ep->com.flags)) deref_qp(ep); if (test_bit(RELEASE_RESOURCES, &ep->com.flags)) { @@ -398,7 +398,8 @@ void _c4iw_free_ep(struct kref *kref) (const u32 *)&sin6->sin6_addr.s6_addr, 1); } - cxgb4_remove_tid(ep->com.dev->rdev.lldi.tids, 0, ep->hwtid); + cxgb4_remove_tid(ep->com.dev->rdev.lldi.tids, 0, ep->hwtid, + ep->com.local_addr.ss_family); dst_release(ep->dst); cxgb4_l2t_release(ep->l2t); if (ep->mpa_skb) @@ -467,7 +468,7 @@ static struct net_device *get_real_dev(struct net_device *egress_dev) static void arp_failure_discard(void *handle, struct sk_buff *skb) { - pr_err(MOD "ARP failure\n"); + pr_err("ARP failure\n"); kfree_skb(skb); } @@ -488,6 +489,7 @@ static int _put_ep_safe(struct c4iw_dev *dev, struct sk_buff *skb) ep = *((struct c4iw_ep **)(skb->cb + 2 * sizeof(void *))); release_ep_resources(ep); + kfree_skb(skb); return 0; } @@ -498,6 +500,7 @@ static int _put_pass_ep_safe(struct c4iw_dev *dev, struct sk_buff *skb) ep = *((struct c4iw_ep **)(skb->cb + 2 * sizeof(void *))); c4iw_put_ep(&ep->parent_ep->com); release_ep_resources(ep); + kfree_skb(skb); return 0; } @@ -528,7 +531,7 @@ static void pass_accept_rpl_arp_failure(void *handle, struct sk_buff *skb) { struct c4iw_ep *ep = handle; - pr_err(MOD "ARP failure during accept - tid %u -dropping connection\n", + pr_err("ARP failure during accept - tid %u - dropping connection\n", ep->hwtid); __state_set(&ep->com, DEAD); @@ -542,7 +545,7 @@ static void act_open_req_arp_failure(void *handle, struct sk_buff *skb) { struct c4iw_ep *ep = handle; - printk(KERN_ERR MOD "ARP failure during connect\n"); + pr_err("ARP failure during connect\n"); connect_reply_upcall(ep, -EHOSTUNREACH); __state_set(&ep->com, DEAD); if (ep->com.remote_addr.ss_family == AF_INET6) { @@ -567,13 +570,15 @@ static void abort_arp_failure(void *handle, struct sk_buff *skb) struct c4iw_rdev *rdev = &ep->com.dev->rdev; struct cpl_abort_req *req = cplhdr(skb); - PDBG("%s rdev %p\n", __func__, rdev); + pr_debug("%s rdev %p\n", __func__, rdev); req->cmd = CPL_ABORT_NO_RST; + skb_get(skb); ret = c4iw_ofld_send(rdev, skb); if (ret) { __state_set(&ep->com, DEAD); queue_arp_failure_cpl(ep, skb, FAKE_CPL_PUT_EP_SAFE); - } + } else + kfree_skb(skb); } static int send_flowc(struct c4iw_ep *ep) @@ -592,7 +597,7 @@ static int send_flowc(struct c4iw_ep *ep) else nparams = 9; - flowc = (struct fw_flowc_wr *)__skb_put(skb, FLOWC_LEN); + flowc = __skb_put(skb, FLOWC_LEN); flowc->op_to_nparams = cpu_to_be32(FW_WR_OP_V(FW_FLOWC_WR) | FW_FLOWC_WR_NPARAMS_V(nparams)); @@ -642,7 +647,7 @@ static int send_halfclose(struct c4iw_ep *ep) struct sk_buff *skb = skb_dequeue(&ep->com.ep_skb_list); u32 wrlen = roundup(sizeof(struct cpl_close_con_req), 16); - PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); + pr_debug("%s ep %p tid %u\n", __func__, ep, ep->hwtid); if (WARN_ON(!skb)) return -ENOMEM; @@ -657,7 +662,7 @@ static int send_abort(struct c4iw_ep *ep) u32 wrlen = roundup(sizeof(struct cpl_abort_req), 16); struct sk_buff *req_skb = skb_dequeue(&ep->com.ep_skb_list); - PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); + pr_debug("%s ep %p tid %u\n", __func__, ep, ep->hwtid); if (WARN_ON(!req_skb)) return -ENOMEM; @@ -692,6 +697,10 @@ static int send_connect(struct c4iw_ep *ep) int ret; enum chip_type adapter_type = ep->com.dev->rdev.lldi.adapter_type; u32 isn = (prandom_u32() & ~7UL) - 1; + struct net_device *netdev; + u64 params; + + netdev = ep->com.dev->rdev.lldi.ports[0]; switch (CHELSIO_CHIP_VERSION(adapter_type)) { case CHELSIO_T4: @@ -716,12 +725,11 @@ static int send_connect(struct c4iw_ep *ep) roundup(sizev4, 16) : roundup(sizev6, 16); - PDBG("%s ep %p atid %u\n", __func__, ep, ep->atid); + pr_debug("%s ep %p atid %u\n", __func__, ep, ep->atid); skb = get_skb(NULL, wrlen, GFP_KERNEL); if (!skb) { - printk(KERN_ERR MOD "%s - failed to alloc skb.\n", - __func__); + pr_err("%s - failed to alloc skb\n", __func__); return -ENOMEM; } set_wr_txq(skb, CPL_PRIORITY_SETUP, ep->ctrlq_idx); @@ -768,6 +776,8 @@ static int send_connect(struct c4iw_ep *ep) opt2 |= T5_ISS_F; } + params = cxgb4_select_ntuple(netdev, ep->l2t); + if (ep->com.remote_addr.ss_family == AF_INET6) cxgb4_clip_get(ep->com.dev->rdev.lldi.ports[0], (const u32 *)&la6->sin6_addr.s6_addr, 1); @@ -777,18 +787,16 @@ static int send_connect(struct c4iw_ep *ep) if (ep->com.remote_addr.ss_family == AF_INET) { switch (CHELSIO_CHIP_VERSION(adapter_type)) { case CHELSIO_T4: - req = (struct cpl_act_open_req *)skb_put(skb, wrlen); + req = skb_put(skb, wrlen); INIT_TP_WR(req, 0); break; case CHELSIO_T5: - t5req = (struct cpl_t5_act_open_req *)skb_put(skb, - wrlen); + t5req = skb_put(skb, wrlen); INIT_TP_WR(t5req, 0); req = (struct cpl_act_open_req *)t5req; break; case CHELSIO_T6: - t6req = (struct cpl_t6_act_open_req *)skb_put(skb, - wrlen); + t6req = skb_put(skb, wrlen); INIT_TP_WR(t6req, 0); req = (struct cpl_act_open_req *)t6req; t5req = (struct cpl_t5_act_open_req *)t6req; @@ -809,34 +817,36 @@ static int send_connect(struct c4iw_ep *ep) req->opt0 = cpu_to_be64(opt0); if (is_t4(ep->com.dev->rdev.lldi.adapter_type)) { - req->params = cpu_to_be32(cxgb4_select_ntuple( - ep->com.dev->rdev.lldi.ports[0], - ep->l2t)); + req->params = cpu_to_be32(params); req->opt2 = cpu_to_be32(opt2); } else { - t5req->params = cpu_to_be64(FILTER_TUPLE_V( - cxgb4_select_ntuple( - ep->com.dev->rdev.lldi.ports[0], - ep->l2t))); - t5req->rsvd = cpu_to_be32(isn); - PDBG("%s snd_isn %u\n", __func__, t5req->rsvd); - t5req->opt2 = cpu_to_be32(opt2); + if (is_t5(ep->com.dev->rdev.lldi.adapter_type)) { + t5req->params = + cpu_to_be64(FILTER_TUPLE_V(params)); + t5req->rsvd = cpu_to_be32(isn); + pr_debug("%s snd_isn %u\n", __func__, t5req->rsvd); + t5req->opt2 = cpu_to_be32(opt2); + } else { + t6req->params = + cpu_to_be64(FILTER_TUPLE_V(params)); + t6req->rsvd = cpu_to_be32(isn); + pr_debug("%s snd_isn %u\n", __func__, t6req->rsvd); + t6req->opt2 = cpu_to_be32(opt2); + } } } else { switch (CHELSIO_CHIP_VERSION(adapter_type)) { case CHELSIO_T4: - req6 = (struct cpl_act_open_req6 *)skb_put(skb, wrlen); + req6 = skb_put(skb, wrlen); INIT_TP_WR(req6, 0); break; case CHELSIO_T5: - t5req6 = (struct cpl_t5_act_open_req6 *)skb_put(skb, - wrlen); + t5req6 = skb_put(skb, wrlen); INIT_TP_WR(t5req6, 0); req6 = (struct cpl_act_open_req6 *)t5req6; break; case CHELSIO_T6: - t6req6 = (struct cpl_t6_act_open_req6 *)skb_put(skb, - wrlen); + t6req6 = skb_put(skb, wrlen); INIT_TP_WR(t6req6, 0); req6 = (struct cpl_act_open_req6 *)t6req6; t5req6 = (struct cpl_t5_act_open_req6 *)t6req6; @@ -859,18 +869,24 @@ static int send_connect(struct c4iw_ep *ep) req6->opt0 = cpu_to_be64(opt0); if (is_t4(ep->com.dev->rdev.lldi.adapter_type)) { - req6->params = cpu_to_be32(cxgb4_select_ntuple( - ep->com.dev->rdev.lldi.ports[0], - ep->l2t)); + req6->params = cpu_to_be32(cxgb4_select_ntuple(netdev, + ep->l2t)); req6->opt2 = cpu_to_be32(opt2); } else { - t5req6->params = cpu_to_be64(FILTER_TUPLE_V( - cxgb4_select_ntuple( - ep->com.dev->rdev.lldi.ports[0], - ep->l2t))); - t5req6->rsvd = cpu_to_be32(isn); - PDBG("%s snd_isn %u\n", __func__, t5req6->rsvd); - t5req6->opt2 = cpu_to_be32(opt2); + if (is_t5(ep->com.dev->rdev.lldi.adapter_type)) { + t5req6->params = + cpu_to_be64(FILTER_TUPLE_V(params)); + t5req6->rsvd = cpu_to_be32(isn); + pr_debug("%s snd_isn %u\n", __func__, t5req6->rsvd); + t5req6->opt2 = cpu_to_be32(opt2); + } else { + t6req6->params = + cpu_to_be64(FILTER_TUPLE_V(params)); + t6req6->rsvd = cpu_to_be32(isn); + pr_debug("%s snd_isn %u\n", __func__, t6req6->rsvd); + t6req6->opt2 = cpu_to_be32(opt2); + } + } } @@ -891,7 +907,8 @@ static int send_mpa_req(struct c4iw_ep *ep, struct sk_buff *skb, struct mpa_message *mpa; struct mpa_v2_conn_params mpa_v2_params; - PDBG("%s ep %p tid %u pd_len %d\n", __func__, ep, ep->hwtid, ep->plen); + pr_debug("%s ep %p tid %u pd_len %d\n", + __func__, ep, ep->hwtid, ep->plen); BUG_ON(skb_cloned(skb)); @@ -906,8 +923,7 @@ static int send_mpa_req(struct c4iw_ep *ep, struct sk_buff *skb, } set_wr_txq(skb, CPL_PRIORITY_DATA, ep->txq_idx); - req = (struct fw_ofld_tx_data_wr *)skb_put(skb, wrlen); - memset(req, 0, wrlen); + req = skb_put_zero(skb, wrlen); req->op_to_immdlen = cpu_to_be32( FW_WR_OP_V(FW_OFLD_TX_DATA_WR) | FW_WR_COMPL_F | @@ -945,8 +961,8 @@ static int send_mpa_req(struct c4iw_ep *ep, struct sk_buff *skb, if (mpa_rev_to_use == 2) { mpa->private_data_size = htons(ntohs(mpa->private_data_size) + sizeof (struct mpa_v2_conn_params)); - PDBG("%s initiator ird %u ord %u\n", __func__, ep->ird, - ep->ord); + pr_debug("%s initiator ird %u ord %u\n", __func__, ep->ird, + ep->ord); mpa_v2_params.ird = htons((u16)ep->ird); mpa_v2_params.ord = htons((u16)ep->ord); @@ -998,7 +1014,8 @@ static int send_mpa_reject(struct c4iw_ep *ep, const void *pdata, u8 plen) struct sk_buff *skb; struct mpa_v2_conn_params mpa_v2_params; - PDBG("%s ep %p tid %u pd_len %d\n", __func__, ep, ep->hwtid, ep->plen); + pr_debug("%s ep %p tid %u pd_len %d\n", + __func__, ep, ep->hwtid, ep->plen); mpalen = sizeof(*mpa) + plen; if (ep->mpa_attr.version == 2 && ep->mpa_attr.enhanced_rdma_conn) @@ -1007,13 +1024,12 @@ static int send_mpa_reject(struct c4iw_ep *ep, const void *pdata, u8 plen) skb = get_skb(NULL, wrlen, GFP_KERNEL); if (!skb) { - printk(KERN_ERR MOD "%s - cannot alloc skb!\n", __func__); + pr_err("%s - cannot alloc skb!\n", __func__); return -ENOMEM; } set_wr_txq(skb, CPL_PRIORITY_DATA, ep->txq_idx); - req = (struct fw_ofld_tx_data_wr *)skb_put(skb, wrlen); - memset(req, 0, wrlen); + req = skb_put_zero(skb, wrlen); req->op_to_immdlen = cpu_to_be32( FW_WR_OP_V(FW_OFLD_TX_DATA_WR) | FW_WR_COMPL_F | @@ -1078,7 +1094,8 @@ static int send_mpa_reply(struct c4iw_ep *ep, const void *pdata, u8 plen) struct sk_buff *skb; struct mpa_v2_conn_params mpa_v2_params; - PDBG("%s ep %p tid %u pd_len %d\n", __func__, ep, ep->hwtid, ep->plen); + pr_debug("%s ep %p tid %u pd_len %d\n", + __func__, ep, ep->hwtid, ep->plen); mpalen = sizeof(*mpa) + plen; if (ep->mpa_attr.version == 2 && ep->mpa_attr.enhanced_rdma_conn) @@ -1087,13 +1104,12 @@ static int send_mpa_reply(struct c4iw_ep *ep, const void *pdata, u8 plen) skb = get_skb(NULL, wrlen, GFP_KERNEL); if (!skb) { - printk(KERN_ERR MOD "%s - cannot alloc skb!\n", __func__); + pr_err("%s - cannot alloc skb!\n", __func__); return -ENOMEM; } set_wr_txq(skb, CPL_PRIORITY_DATA, ep->txq_idx); - req = (struct fw_ofld_tx_data_wr *) skb_put(skb, wrlen); - memset(req, 0, wrlen); + req = skb_put_zero(skb, wrlen); req->op_to_immdlen = cpu_to_be32( FW_WR_OP_V(FW_OFLD_TX_DATA_WR) | FW_WR_COMPL_F | @@ -1169,15 +1185,15 @@ static int act_establish(struct c4iw_dev *dev, struct sk_buff *skb) ep = lookup_atid(t, atid); - PDBG("%s ep %p tid %u snd_isn %u rcv_isn %u\n", __func__, ep, tid, - be32_to_cpu(req->snd_isn), be32_to_cpu(req->rcv_isn)); + pr_debug("%s ep %p tid %u snd_isn %u rcv_isn %u\n", __func__, ep, tid, + be32_to_cpu(req->snd_isn), be32_to_cpu(req->rcv_isn)); mutex_lock(&ep->com.mutex); dst_confirm(ep->dst); /* setup the hwtid for this connection */ ep->hwtid = tid; - cxgb4_insert_tid(t, ep, tid); + cxgb4_insert_tid(t, ep, tid, ep->com.local_addr.ss_family); insert_ep_tid(ep); ep->snd_seq = be32_to_cpu(req->snd_isn); @@ -1213,13 +1229,13 @@ static void close_complete_upcall(struct c4iw_ep *ep, int status) { struct iw_cm_event event; - PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); + pr_debug("%s ep %p tid %u\n", __func__, ep, ep->hwtid); memset(&event, 0, sizeof(event)); event.event = IW_CM_EVENT_CLOSE; event.status = status; if (ep->com.cm_id) { - PDBG("close complete delivered ep %p cm_id %p tid %u\n", - ep, ep->com.cm_id, ep->hwtid); + pr_debug("close complete delivered ep %p cm_id %p tid %u\n", + ep, ep->com.cm_id, ep->hwtid); ep->com.cm_id->event_handler(ep->com.cm_id, &event); deref_cm_id(&ep->com); set_bit(CLOSE_UPCALL, &ep->com.history); @@ -1230,12 +1246,12 @@ static void peer_close_upcall(struct c4iw_ep *ep) { struct iw_cm_event event; - PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); + pr_debug("%s ep %p tid %u\n", __func__, ep, ep->hwtid); memset(&event, 0, sizeof(event)); event.event = IW_CM_EVENT_DISCONNECT; if (ep->com.cm_id) { - PDBG("peer close delivered ep %p cm_id %p tid %u\n", - ep, ep->com.cm_id, ep->hwtid); + pr_debug("peer close delivered ep %p cm_id %p tid %u\n", + ep, ep->com.cm_id, ep->hwtid); ep->com.cm_id->event_handler(ep->com.cm_id, &event); set_bit(DISCONN_UPCALL, &ep->com.history); } @@ -1245,13 +1261,13 @@ static void peer_abort_upcall(struct c4iw_ep *ep) { struct iw_cm_event event; - PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); + pr_debug("%s ep %p tid %u\n", __func__, ep, ep->hwtid); memset(&event, 0, sizeof(event)); event.event = IW_CM_EVENT_CLOSE; event.status = -ECONNRESET; if (ep->com.cm_id) { - PDBG("abort delivered ep %p cm_id %p tid %u\n", ep, - ep->com.cm_id, ep->hwtid); + pr_debug("abort delivered ep %p cm_id %p tid %u\n", ep, + ep->com.cm_id, ep->hwtid); ep->com.cm_id->event_handler(ep->com.cm_id, &event); deref_cm_id(&ep->com); set_bit(ABORT_UPCALL, &ep->com.history); @@ -1262,7 +1278,8 @@ static void connect_reply_upcall(struct c4iw_ep *ep, int status) { struct iw_cm_event event; - PDBG("%s ep %p tid %u status %d\n", __func__, ep, ep->hwtid, status); + pr_debug("%s ep %p tid %u status %d\n", + __func__, ep, ep->hwtid, status); memset(&event, 0, sizeof(event)); event.event = IW_CM_EVENT_CONNECT_REPLY; event.status = status; @@ -1291,8 +1308,8 @@ static void connect_reply_upcall(struct c4iw_ep *ep, int status) } } - PDBG("%s ep %p tid %u status %d\n", __func__, ep, - ep->hwtid, status); + pr_debug("%s ep %p tid %u status %d\n", __func__, ep, + ep->hwtid, status); set_bit(CONN_RPL_UPCALL, &ep->com.history); ep->com.cm_id->event_handler(ep->com.cm_id, &event); @@ -1305,7 +1322,7 @@ static int connect_request_upcall(struct c4iw_ep *ep) struct iw_cm_event event; int ret; - PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); + pr_debug("%s ep %p tid %u\n", __func__, ep, ep->hwtid); memset(&event, 0, sizeof(event)); event.event = IW_CM_EVENT_CONNECT_REQUEST; memcpy(&event.local_addr, &ep->com.local_addr, @@ -1342,13 +1359,13 @@ static void established_upcall(struct c4iw_ep *ep) { struct iw_cm_event event; - PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); + pr_debug("%s ep %p tid %u\n", __func__, ep, ep->hwtid); memset(&event, 0, sizeof(event)); event.event = IW_CM_EVENT_ESTABLISHED; event.ird = ep->ord; event.ord = ep->ird; if (ep->com.cm_id) { - PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); + pr_debug("%s ep %p tid %u\n", __func__, ep, ep->hwtid); ep->com.cm_id->event_handler(ep->com.cm_id, &event); set_bit(ESTAB_UPCALL, &ep->com.history); } @@ -1360,10 +1377,11 @@ static int update_rx_credits(struct c4iw_ep *ep, u32 credits) u32 wrlen = roundup(sizeof(struct cpl_rx_data_ack), 16); u32 credit_dack; - PDBG("%s ep %p tid %u credits %u\n", __func__, ep, ep->hwtid, credits); + pr_debug("%s ep %p tid %u credits %u\n", + __func__, ep, ep->hwtid, credits); skb = get_skb(NULL, wrlen, GFP_KERNEL); if (!skb) { - printk(KERN_ERR MOD "update_rx_credits - cannot alloc skb!\n"); + pr_err("update_rx_credits - cannot alloc skb!\n"); return 0; } @@ -1411,7 +1429,7 @@ static int process_mpa_reply(struct c4iw_ep *ep, struct sk_buff *skb) int err; int disconnect = 0; - PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); + pr_debug("%s ep %p tid %u\n", __func__, ep, ep->hwtid); /* * If we get more than the supported amount of private data @@ -1438,8 +1456,8 @@ static int process_mpa_reply(struct c4iw_ep *ep, struct sk_buff *skb) /* Validate MPA header. */ if (mpa->revision > mpa_rev) { - printk(KERN_ERR MOD "%s MPA version mismatch. Local = %d," - " Received = %d\n", __func__, mpa_rev, mpa->revision); + pr_err("%s MPA version mismatch. Local = %d, Received = %d\n", + __func__, mpa_rev, mpa->revision); err = -EPROTO; goto err_stop_timer; } @@ -1509,8 +1527,9 @@ static int process_mpa_reply(struct c4iw_ep *ep, struct sk_buff *skb) MPA_V2_IRD_ORD_MASK; resp_ord = ntohs(mpa_v2_params->ord) & MPA_V2_IRD_ORD_MASK; - PDBG("%s responder ird %u ord %u ep ird %u ord %u\n", - __func__, resp_ird, resp_ord, ep->ird, ep->ord); + pr_debug("%s responder ird %u ord %u ep ird %u ord %u\n", + __func__, + resp_ird, resp_ord, ep->ird, ep->ord); /* * This is a double-check. Ideally, below checks are @@ -1554,12 +1573,11 @@ static int process_mpa_reply(struct c4iw_ep *ep, struct sk_buff *skb) if (peer2peer) ep->mpa_attr.p2p_type = p2p_type; - PDBG("%s - crc_enabled=%d, recv_marker_enabled=%d, " - "xmit_marker_enabled=%d, version=%d p2p_type=%d local-p2p_type = " - "%d\n", __func__, ep->mpa_attr.crc_enabled, - ep->mpa_attr.recv_marker_enabled, - ep->mpa_attr.xmit_marker_enabled, ep->mpa_attr.version, - ep->mpa_attr.p2p_type, p2p_type); + pr_debug("%s - crc_enabled=%d, recv_marker_enabled=%d, xmit_marker_enabled=%d, version=%d p2p_type=%d local-p2p_type = %d\n", + __func__, ep->mpa_attr.crc_enabled, + ep->mpa_attr.recv_marker_enabled, + ep->mpa_attr.xmit_marker_enabled, ep->mpa_attr.version, + ep->mpa_attr.p2p_type, p2p_type); /* * If responder's RTR does not match with that of initiator, assign @@ -1594,7 +1612,7 @@ static int process_mpa_reply(struct c4iw_ep *ep, struct sk_buff *skb) * supports, generate TERM message */ if (rtr_mismatch) { - printk(KERN_ERR "%s: RTR mismatch, sending TERM\n", __func__); + pr_err("%s: RTR mismatch, sending TERM\n", __func__); attrs.layer_etype = LAYER_MPA | DDP_LLP; attrs.ecode = MPA_NOMATCH_RTR; attrs.next_state = C4IW_QP_STATE_TERMINATE; @@ -1613,8 +1631,7 @@ static int process_mpa_reply(struct c4iw_ep *ep, struct sk_buff *skb) * initiator ORD. */ if (insuff_ird) { - printk(KERN_ERR "%s: Insufficient IRD, sending TERM\n", - __func__); + pr_err("%s: Insufficient IRD, sending TERM\n", __func__); attrs.layer_etype = LAYER_MPA | DDP_LLP; attrs.ecode = MPA_INSUFF_IRD; attrs.next_state = C4IW_QP_STATE_TERMINATE; @@ -1653,7 +1670,7 @@ static int process_mpa_request(struct c4iw_ep *ep, struct sk_buff *skb) struct mpa_v2_conn_params *mpa_v2_params; u16 plen; - PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); + pr_debug("%s ep %p tid %u\n", __func__, ep, ep->hwtid); /* * If we get more than the supported amount of private data @@ -1662,7 +1679,7 @@ static int process_mpa_request(struct c4iw_ep *ep, struct sk_buff *skb) if (ep->mpa_pkt_len + skb->len > sizeof(ep->mpa_pkt)) goto err_stop_timer; - PDBG("%s enter (%s line %u)\n", __func__, __FILE__, __LINE__); + pr_debug("%s enter (%s line %u)\n", __func__, __FILE__, __LINE__); /* * Copy the new data into our accumulation buffer. @@ -1678,15 +1695,15 @@ static int process_mpa_request(struct c4iw_ep *ep, struct sk_buff *skb) if (ep->mpa_pkt_len < sizeof(*mpa)) return 0; - PDBG("%s enter (%s line %u)\n", __func__, __FILE__, __LINE__); + pr_debug("%s enter (%s line %u)\n", __func__, __FILE__, __LINE__); mpa = (struct mpa_message *) ep->mpa_pkt; /* * Validate MPA Header. */ if (mpa->revision > mpa_rev) { - printk(KERN_ERR MOD "%s MPA version mismatch. Local = %d," - " Received = %d\n", __func__, mpa_rev, mpa->revision); + pr_err("%s MPA version mismatch. Local = %d, Received = %d\n", + __func__, mpa_rev, mpa->revision); goto err_stop_timer; } @@ -1741,8 +1758,8 @@ static int process_mpa_request(struct c4iw_ep *ep, struct sk_buff *skb) MPA_V2_IRD_ORD_MASK; ep->ord = min_t(u32, ep->ord, cur_max_read_depth(ep->com.dev)); - PDBG("%s initiator ird %u ord %u\n", __func__, ep->ird, - ep->ord); + pr_debug("%s initiator ird %u ord %u\n", + __func__, ep->ird, ep->ord); if (ntohs(mpa_v2_params->ird) & MPA_V2_PEER2PEER_MODEL) if (peer2peer) { if (ntohs(mpa_v2_params->ord) & @@ -1759,11 +1776,11 @@ static int process_mpa_request(struct c4iw_ep *ep, struct sk_buff *skb) if (peer2peer) ep->mpa_attr.p2p_type = p2p_type; - PDBG("%s - crc_enabled=%d, recv_marker_enabled=%d, " - "xmit_marker_enabled=%d, version=%d p2p_type=%d\n", __func__, - ep->mpa_attr.crc_enabled, ep->mpa_attr.recv_marker_enabled, - ep->mpa_attr.xmit_marker_enabled, ep->mpa_attr.version, - ep->mpa_attr.p2p_type); + pr_debug("%s - crc_enabled=%d, recv_marker_enabled=%d, xmit_marker_enabled=%d, version=%d p2p_type=%d\n", + __func__, + ep->mpa_attr.crc_enabled, ep->mpa_attr.recv_marker_enabled, + ep->mpa_attr.xmit_marker_enabled, ep->mpa_attr.version, + ep->mpa_attr.p2p_type); __state_set(&ep->com, MPA_REQ_RCVD); @@ -1799,7 +1816,7 @@ static int rx_data(struct c4iw_dev *dev, struct sk_buff *skb) ep = get_ep_from_tid(dev, tid); if (!ep) return 0; - PDBG("%s ep %p tid %u dlen %u\n", __func__, ep, ep->hwtid, dlen); + pr_debug("%s ep %p tid %u dlen %u\n", __func__, ep, ep->hwtid, dlen); skb_pull(skb, sizeof(*hdr)); skb_trim(skb, dlen); mutex_lock(&ep->com.mutex); @@ -1850,10 +1867,10 @@ static int abort_rpl(struct c4iw_dev *dev, struct sk_buff *skb) ep = get_ep_from_tid(dev, tid); if (!ep) { - printk(KERN_WARNING MOD "Abort rpl to freed endpoint\n"); + pr_warn("Abort rpl to freed endpoint\n"); return 0; } - PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); + pr_debug("%s ep %p tid %u\n", __func__, ep, ep->hwtid); mutex_lock(&ep->com.mutex); switch (ep->com.state) { case ABORTING: @@ -1862,8 +1879,7 @@ static int abort_rpl(struct c4iw_dev *dev, struct sk_buff *skb) release = 1; break; default: - printk(KERN_ERR "%s ep %p state %d\n", - __func__, ep, ep->com.state); + pr_err("%s ep %p state %d\n", __func__, ep, ep->com.state); break; } mutex_unlock(&ep->com.mutex); @@ -1884,8 +1900,7 @@ static int send_fw_act_open_req(struct c4iw_ep *ep, unsigned int atid) int win; skb = get_skb(NULL, sizeof(*req), GFP_KERNEL); - req = (struct fw_ofld_connection_wr *)__skb_put(skb, sizeof(*req)); - memset(req, 0, sizeof(*req)); + req = __skb_put_zero(skb, sizeof(*req)); req->op_compl = htonl(WR_OP_V(FW_OFLD_CONNECTION_WR)); req->len16_pkd = htonl(FW_WR_LEN16_V(DIV_ROUND_UP(sizeof(*req), 16))); req->le.filter = cpu_to_be32(cxgb4_select_ntuple( @@ -1979,7 +1994,8 @@ static void set_tcp_window(struct c4iw_ep *ep, struct port_info *pi) { ep->snd_win = snd_win; ep->rcv_win = rcv_win; - PDBG("%s snd_win %d rcv_win %d\n", __func__, ep->snd_win, ep->rcv_win); + pr_debug("%s snd_win %d rcv_win %d\n", + __func__, ep->snd_win, ep->rcv_win); } #define ACT_OPEN_RETRY_COUNT 2 @@ -2084,7 +2100,7 @@ static int c4iw_reconnect(struct c4iw_ep *ep) int iptype; __u8 *ra; - PDBG("%s qp %p cm_id %p\n", __func__, ep->com.qp, ep->com.cm_id); + pr_debug("%s qp %p cm_id %p\n", __func__, ep->com.qp, ep->com.cm_id); init_timer(&ep->timer); c4iw_init_wr_wait(&ep->com.wr_wait); @@ -2108,7 +2124,7 @@ static int c4iw_reconnect(struct c4iw_ep *ep) */ ep->atid = cxgb4_alloc_atid(ep->com.dev->rdev.lldi.tids, ep); if (ep->atid == -1) { - pr_err("%s - cannot alloc atid.\n", __func__); + pr_err("%s - cannot alloc atid\n", __func__); err = -ENOMEM; goto fail2; } @@ -2135,7 +2151,7 @@ static int c4iw_reconnect(struct c4iw_ep *ep) ra = (__u8 *)&raddr6->sin6_addr; } if (!ep->dst) { - pr_err("%s - cannot find route.\n", __func__); + pr_err("%s - cannot find route\n", __func__); err = -EHOSTUNREACH; goto fail3; } @@ -2143,13 +2159,13 @@ static int c4iw_reconnect(struct c4iw_ep *ep) ep->com.dev->rdev.lldi.adapter_type, ep->com.cm_id->tos); if (err) { - pr_err("%s - cannot alloc l2e.\n", __func__); + pr_err("%s - cannot alloc l2e\n", __func__); goto fail4; } - PDBG("%s txq_idx %u tx_chan %u smac_idx %u rss_qid %u l2t_idx %u\n", - __func__, ep->txq_idx, ep->tx_chan, ep->smac_idx, ep->rss_qid, - ep->l2t->idx); + pr_debug("%s txq_idx %u tx_chan %u smac_idx %u rss_qid %u l2t_idx %u\n", + __func__, ep->txq_idx, ep->tx_chan, ep->smac_idx, ep->rss_qid, + ep->l2t->idx); state_set(&ep->com, CONNECTING); ep->tos = ep->com.cm_id->tos; @@ -2199,12 +2215,12 @@ static int act_open_rpl(struct c4iw_dev *dev, struct sk_buff *skb) la6 = (struct sockaddr_in6 *)&ep->com.local_addr; ra6 = (struct sockaddr_in6 *)&ep->com.remote_addr; - PDBG("%s ep %p atid %u status %u errno %d\n", __func__, ep, atid, - status, status2errno(status)); + pr_debug("%s ep %p atid %u status %u errno %d\n", __func__, ep, atid, + status, status2errno(status)); if (cxgb_is_neg_adv(status)) { - PDBG("%s Connection problems for atid %u status %u (%s)\n", - __func__, atid, status, neg_adv_str(status)); + pr_debug("%s Connection problems for atid %u status %u (%s)\n", + __func__, atid, status, neg_adv_str(status)); ep->stats.connect_neg_adv++; mutex_lock(&dev->rdev.stats.lock); dev->rdev.stats.neg_adv++; @@ -2281,7 +2297,8 @@ fail: (const u32 *)&sin6->sin6_addr.s6_addr, 1); } if (status && act_open_has_tid(status)) - cxgb4_remove_tid(ep->com.dev->rdev.lldi.tids, 0, GET_TID(rpl)); + cxgb4_remove_tid(ep->com.dev->rdev.lldi.tids, 0, GET_TID(rpl), + ep->com.local_addr.ss_family); remove_handle(ep->com.dev, &ep->com.dev->atid_idr, atid); cxgb4_free_atid(t, atid); @@ -2299,11 +2316,11 @@ static int pass_open_rpl(struct c4iw_dev *dev, struct sk_buff *skb) struct c4iw_listen_ep *ep = get_ep_from_stid(dev, stid); if (!ep) { - PDBG("%s stid %d lookup failure!\n", __func__, stid); + pr_debug("%s stid %d lookup failure!\n", __func__, stid); goto out; } - PDBG("%s ep %p status %d error %d\n", __func__, ep, - rpl->status, status2errno(rpl->status)); + pr_debug("%s ep %p status %d error %d\n", __func__, ep, + rpl->status, status2errno(rpl->status)); c4iw_wake_up(&ep->com.wr_wait, status2errno(rpl->status)); c4iw_put_ep(&ep->com); out: @@ -2316,7 +2333,7 @@ static int close_listsrv_rpl(struct c4iw_dev *dev, struct sk_buff *skb) unsigned int stid = GET_TID(rpl); struct c4iw_listen_ep *ep = get_ep_from_stid(dev, stid); - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); c4iw_wake_up(&ep->com.wr_wait, status2errno(rpl->status)); c4iw_put_ep(&ep->com); return 0; @@ -2334,7 +2351,7 @@ static int accept_cr(struct c4iw_ep *ep, struct sk_buff *skb, int win; enum chip_type adapter_type = ep->com.dev->rdev.lldi.adapter_type; - PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); + pr_debug("%s ep %p tid %u\n", __func__, ep, ep->hwtid); BUG_ON(skb_cloned(skb)); skb_get(skb); @@ -2405,7 +2422,7 @@ static int accept_cr(struct c4iw_ep *ep, struct sk_buff *skb, if (peer2peer) isn += 4; rpl5->iss = cpu_to_be32(isn); - PDBG("%s iss %u\n", __func__, be32_to_cpu(rpl5->iss)); + pr_debug("%s iss %u\n", __func__, be32_to_cpu(rpl5->iss)); } rpl->opt0 = cpu_to_be64(opt0); @@ -2418,7 +2435,7 @@ static int accept_cr(struct c4iw_ep *ep, struct sk_buff *skb, static void reject_cr(struct c4iw_dev *dev, u32 hwtid, struct sk_buff *skb) { - PDBG("%s c4iw_dev %p tid %u\n", __func__, dev, hwtid); + pr_debug("%s c4iw_dev %p tid %u\n", __func__, dev, hwtid); BUG_ON(skb_cloned(skb)); skb_trim(skb, sizeof(struct cpl_tid_release)); release_tid(&dev->rdev, hwtid, skb); @@ -2444,12 +2461,13 @@ static int pass_accept_req(struct c4iw_dev *dev, struct sk_buff *skb) parent_ep = (struct c4iw_ep *)get_ep_from_stid(dev, stid); if (!parent_ep) { - PDBG("%s connect request on invalid stid %d\n", __func__, stid); + pr_debug("%s connect request on invalid stid %d\n", + __func__, stid); goto reject; } if (state_read(&parent_ep->com) != LISTEN) { - PDBG("%s - listening ep not in LISTEN\n", __func__); + pr_debug("%s - listening ep not in LISTEN\n", __func__); goto reject; } @@ -2458,18 +2476,18 @@ static int pass_accept_req(struct c4iw_dev *dev, struct sk_buff *skb) /* Find output route */ if (iptype == 4) { - PDBG("%s parent ep %p hwtid %u laddr %pI4 raddr %pI4 lport %d rport %d peer_mss %d\n" - , __func__, parent_ep, hwtid, - local_ip, peer_ip, ntohs(local_port), - ntohs(peer_port), peer_mss); + pr_debug("%s parent ep %p hwtid %u laddr %pI4 raddr %pI4 lport %d rport %d peer_mss %d\n" + , __func__, parent_ep, hwtid, + local_ip, peer_ip, ntohs(local_port), + ntohs(peer_port), peer_mss); dst = cxgb_find_route(&dev->rdev.lldi, get_real_dev, *(__be32 *)local_ip, *(__be32 *)peer_ip, local_port, peer_port, tos); } else { - PDBG("%s parent ep %p hwtid %u laddr %pI6 raddr %pI6 lport %d rport %d peer_mss %d\n" - , __func__, parent_ep, hwtid, - local_ip, peer_ip, ntohs(local_port), - ntohs(peer_port), peer_mss); + pr_debug("%s parent ep %p hwtid %u laddr %pI6 raddr %pI6 lport %d rport %d peer_mss %d\n" + , __func__, parent_ep, hwtid, + local_ip, peer_ip, ntohs(local_port), + ntohs(peer_port), peer_mss); dst = cxgb_find_route6(&dev->rdev.lldi, get_real_dev, local_ip, peer_ip, local_port, peer_port, PASS_OPEN_TOS_G(ntohl(req->tos_stid)), @@ -2477,15 +2495,13 @@ static int pass_accept_req(struct c4iw_dev *dev, struct sk_buff *skb) &parent_ep->com.local_addr)->sin6_scope_id); } if (!dst) { - printk(KERN_ERR MOD "%s - failed to find dst entry!\n", - __func__); + pr_err("%s - failed to find dst entry!\n", __func__); goto reject; } child_ep = alloc_ep(sizeof(*child_ep), GFP_KERNEL); if (!child_ep) { - printk(KERN_ERR MOD "%s - failed to allocate ep entry!\n", - __func__); + pr_err("%s - failed to allocate ep entry!\n", __func__); dst_release(dst); goto reject; } @@ -2493,14 +2509,14 @@ static int pass_accept_req(struct c4iw_dev *dev, struct sk_buff *skb) err = import_ep(child_ep, iptype, peer_ip, dst, dev, false, parent_ep->com.dev->rdev.lldi.adapter_type, tos); if (err) { - printk(KERN_ERR MOD "%s - failed to allocate l2t entry!\n", - __func__); + pr_err("%s - failed to allocate l2t entry!\n", __func__); dst_release(dst); kfree(child_ep); goto reject; } - hdrs = sizeof(struct iphdr) + sizeof(struct tcphdr) + + hdrs = ((iptype == 4) ? sizeof(struct iphdr) : sizeof(struct ipv6hdr)) + + sizeof(struct tcphdr) + ((enable_tcp_timestamps && req->tcpopt.tstamp) ? 12 : 0); if (peer_mss && child_ep->mtu > (peer_mss + hdrs)) child_ep->mtu = peer_mss + hdrs; @@ -2517,18 +2533,18 @@ static int pass_accept_req(struct c4iw_dev *dev, struct sk_buff *skb) struct sockaddr_in *sin = (struct sockaddr_in *) &child_ep->com.local_addr; - sin->sin_family = PF_INET; + sin->sin_family = AF_INET; sin->sin_port = local_port; sin->sin_addr.s_addr = *(__be32 *)local_ip; sin = (struct sockaddr_in *)&child_ep->com.local_addr; - sin->sin_family = PF_INET; + sin->sin_family = AF_INET; sin->sin_port = ((struct sockaddr_in *) &parent_ep->com.local_addr)->sin_port; sin->sin_addr.s_addr = *(__be32 *)local_ip; sin = (struct sockaddr_in *)&child_ep->com.remote_addr; - sin->sin_family = PF_INET; + sin->sin_family = AF_INET; sin->sin_port = peer_port; sin->sin_addr.s_addr = *(__be32 *)peer_ip; } else { @@ -2555,11 +2571,12 @@ static int pass_accept_req(struct c4iw_dev *dev, struct sk_buff *skb) child_ep->dst = dst; child_ep->hwtid = hwtid; - PDBG("%s tx_chan %u smac_idx %u rss_qid %u\n", __func__, - child_ep->tx_chan, child_ep->smac_idx, child_ep->rss_qid); + pr_debug("%s tx_chan %u smac_idx %u rss_qid %u\n", __func__, + child_ep->tx_chan, child_ep->smac_idx, child_ep->rss_qid); init_timer(&child_ep->timer); - cxgb4_insert_tid(t, child_ep, hwtid); + cxgb4_insert_tid(t, child_ep, hwtid, + child_ep->com.local_addr.ss_family); insert_ep_tid(child_ep); if (accept_cr(child_ep, skb, req)) { c4iw_put_ep(&parent_ep->com); @@ -2591,12 +2608,12 @@ static int pass_establish(struct c4iw_dev *dev, struct sk_buff *skb) int ret; ep = get_ep_from_tid(dev, tid); - PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); + pr_debug("%s ep %p tid %u\n", __func__, ep, ep->hwtid); ep->snd_seq = be32_to_cpu(req->snd_isn); ep->rcv_seq = be32_to_cpu(req->rcv_isn); - PDBG("%s ep %p hwtid %u tcp_opt 0x%02x\n", __func__, ep, tid, - ntohs(req->tcp_opt)); + pr_debug("%s ep %p hwtid %u tcp_opt 0x%02x\n", __func__, ep, tid, + ntohs(req->tcp_opt)); set_emss(ep, ntohs(req->tcp_opt)); @@ -2628,7 +2645,7 @@ static int peer_close(struct c4iw_dev *dev, struct sk_buff *skb) if (!ep) return 0; - PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); + pr_debug("%s ep %p tid %u\n", __func__, ep, ep->hwtid); dst_confirm(ep->dst); set_bit(PEER_CLOSE, &ep->com.history); @@ -2650,12 +2667,12 @@ static int peer_close(struct c4iw_dev *dev, struct sk_buff *skb) * in rdma connection migration (see c4iw_accept_cr()). */ __state_set(&ep->com, CLOSING); - PDBG("waking up ep %p tid %u\n", ep, ep->hwtid); + pr_debug("waking up ep %p tid %u\n", ep, ep->hwtid); c4iw_wake_up(&ep->com.wr_wait, -ECONNRESET); break; case MPA_REP_SENT: __state_set(&ep->com, CLOSING); - PDBG("waking up ep %p tid %u\n", ep, ep->hwtid); + pr_debug("waking up ep %p tid %u\n", ep, ep->hwtid); c4iw_wake_up(&ep->com.wr_wait, -ECONNRESET); break; case FPDU_MODE: @@ -2719,17 +2736,17 @@ static int peer_abort(struct c4iw_dev *dev, struct sk_buff *skb) return 0; if (cxgb_is_neg_adv(req->status)) { - PDBG("%s Negative advice on abort- tid %u status %d (%s)\n", - __func__, ep->hwtid, req->status, - neg_adv_str(req->status)); + pr_debug("%s Negative advice on abort- tid %u status %d (%s)\n", + __func__, 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++; mutex_unlock(&dev->rdev.stats.lock); goto deref_ep; } - PDBG("%s ep %p tid %u state %u\n", __func__, ep, ep->hwtid, - ep->com.state); + pr_debug("%s ep %p tid %u state %u\n", __func__, ep, ep->hwtid, + ep->com.state); set_bit(PEER_ABORT, &ep->com.history); /* @@ -2761,8 +2778,8 @@ static int peer_abort(struct c4iw_dev *dev, struct sk_buff *skb) * do some housekeeping so as to re-initiate the * connection */ - PDBG("%s: mpa_rev=%d. Retrying with mpav1\n", __func__, - mpa_rev); + pr_debug("%s: mpa_rev=%d. Retrying with mpav1\n", + __func__, mpa_rev); ep->retry_with_mpa_v1 = 1; } break; @@ -2781,16 +2798,14 @@ static int peer_abort(struct c4iw_dev *dev, struct sk_buff *skb) ep->com.qp, C4IW_QP_ATTR_NEXT_STATE, &attrs, 1); if (ret) - printk(KERN_ERR MOD - "%s - qp <- error failed!\n", - __func__); + pr_err("%s - qp <- error failed!\n", __func__); } peer_abort_upcall(ep); break; case ABORTING: break; case DEAD: - PDBG("%s PEER_ABORT IN DEAD STATE!!!!\n", __func__); + pr_debug("%s PEER_ABORT IN DEAD STATE!!!!\n", __func__); mutex_unlock(&ep->com.mutex); goto deref_ep; default: @@ -2829,7 +2844,8 @@ out: 1); } remove_handle(ep->com.dev, &ep->com.dev->hwtid_idr, ep->hwtid); - cxgb4_remove_tid(ep->com.dev->rdev.lldi.tids, 0, ep->hwtid); + cxgb4_remove_tid(ep->com.dev->rdev.lldi.tids, 0, ep->hwtid, + ep->com.local_addr.ss_family); dst_release(ep->dst); cxgb4_l2t_release(ep->l2t); c4iw_reconnect(ep); @@ -2854,7 +2870,7 @@ static int close_con_rpl(struct c4iw_dev *dev, struct sk_buff *skb) if (!ep) return 0; - PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); + pr_debug("%s ep %p tid %u\n", __func__, ep, ep->hwtid); BUG_ON(!ep); /* The cm_id may be null if we failed to connect */ @@ -2902,13 +2918,13 @@ static int terminate(struct c4iw_dev *dev, struct sk_buff *skb) BUG_ON(!ep); if (ep && ep->com.qp) { - printk(KERN_WARNING MOD "TERM received tid %u qpid %u\n", tid, - ep->com.qp->wq.sq.qid); + pr_warn("TERM received tid %u qpid %u\n", + tid, ep->com.qp->wq.sq.qid); attrs.next_state = C4IW_QP_STATE_TERMINATE; c4iw_modify_qp(ep->com.qp->rhp, ep->com.qp, C4IW_QP_ATTR_NEXT_STATE, &attrs, 1); } else - printk(KERN_WARNING MOD "TERM received tid %u no ep/qp\n", tid); + pr_warn("TERM received tid %u no ep/qp\n", tid); c4iw_put_ep(&ep->com); return 0; @@ -2930,18 +2946,19 @@ static int fw4_ack(struct c4iw_dev *dev, struct sk_buff *skb) ep = get_ep_from_tid(dev, tid); if (!ep) return 0; - PDBG("%s ep %p tid %u credits %u\n", __func__, ep, ep->hwtid, credits); + pr_debug("%s ep %p tid %u credits %u\n", + __func__, ep, ep->hwtid, credits); if (credits == 0) { - PDBG("%s 0 credit ack ep %p tid %u state %u\n", - __func__, ep, ep->hwtid, state_read(&ep->com)); + pr_debug("%s 0 credit ack ep %p tid %u state %u\n", + __func__, ep, ep->hwtid, state_read(&ep->com)); goto out; } dst_confirm(ep->dst); if (ep->mpa_skb) { - PDBG("%s last streaming msg ack ep %p tid %u state %u " - "initiator %u freeing skb\n", __func__, ep, ep->hwtid, - state_read(&ep->com), ep->mpa_attr.initiator ? 1 : 0); + pr_debug("%s last streaming msg ack ep %p tid %u state %u initiator %u freeing skb\n", + __func__, ep, ep->hwtid, + state_read(&ep->com), ep->mpa_attr.initiator ? 1 : 0); mutex_lock(&ep->com.mutex); kfree_skb(ep->mpa_skb); ep->mpa_skb = NULL; @@ -2959,7 +2976,7 @@ int c4iw_reject_cr(struct iw_cm_id *cm_id, const void *pdata, u8 pdata_len) int abort; struct c4iw_ep *ep = to_ep(cm_id); - PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); + pr_debug("%s ep %p tid %u\n", __func__, ep, ep->hwtid); mutex_lock(&ep->com.mutex); if (ep->com.state != MPA_REQ_RCVD) { @@ -2990,7 +3007,7 @@ int c4iw_accept_cr(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) struct c4iw_qp *qp = get_qhp(h, conn_param->qpn); int abort = 0; - PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); + pr_debug("%s ep %p tid %u\n", __func__, ep, ep->hwtid); mutex_lock(&ep->com.mutex); if (ep->com.state != MPA_REQ_RCVD) { @@ -3043,7 +3060,7 @@ int c4iw_accept_cr(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) ep->ird = 1; } - PDBG("%s %d ird %d ord %d\n", __func__, __LINE__, ep->ird, ep->ord); + pr_debug("%s %d ird %d ord %d\n", __func__, __LINE__, ep->ird, ep->ord); ep->com.cm_id = cm_id; ref_cm_id(&ep->com); @@ -3172,7 +3189,7 @@ int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) } ep = alloc_ep(sizeof(*ep), GFP_KERNEL); if (!ep) { - printk(KERN_ERR MOD "%s - cannot alloc ep.\n", __func__); + pr_err("%s - cannot alloc ep\n", __func__); err = -ENOMEM; goto out; } @@ -3199,20 +3216,20 @@ int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) ep->com.dev = dev; ep->com.qp = get_qhp(dev, conn_param->qpn); if (!ep->com.qp) { - PDBG("%s qpn 0x%x not found!\n", __func__, conn_param->qpn); + pr_debug("%s qpn 0x%x not found!\n", __func__, conn_param->qpn); err = -EINVAL; goto fail2; } ref_qp(ep); - PDBG("%s qpn 0x%x qp %p cm_id %p\n", __func__, conn_param->qpn, - ep->com.qp, cm_id); + pr_debug("%s qpn 0x%x qp %p cm_id %p\n", __func__, conn_param->qpn, + ep->com.qp, cm_id); /* * Allocate an active TID to initiate a TCP connection. */ ep->atid = cxgb4_alloc_atid(dev->rdev.lldi.tids, ep); if (ep->atid == -1) { - printk(KERN_ERR MOD "%s - cannot alloc atid.\n", __func__); + pr_err("%s - cannot alloc atid\n", __func__); err = -ENOMEM; goto fail2; } @@ -3242,9 +3259,9 @@ int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) } /* find a route */ - PDBG("%s saddr %pI4 sport 0x%x raddr %pI4 rport 0x%x\n", - __func__, &laddr->sin_addr, ntohs(laddr->sin_port), - ra, ntohs(raddr->sin_port)); + pr_debug("%s saddr %pI4 sport 0x%x raddr %pI4 rport 0x%x\n", + __func__, &laddr->sin_addr, ntohs(laddr->sin_port), + ra, ntohs(raddr->sin_port)); ep->dst = cxgb_find_route(&dev->rdev.lldi, get_real_dev, laddr->sin_addr.s_addr, raddr->sin_addr.s_addr, @@ -3264,10 +3281,10 @@ int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) } /* find a route */ - PDBG("%s saddr %pI6 sport 0x%x raddr %pI6 rport 0x%x\n", - __func__, laddr6->sin6_addr.s6_addr, - ntohs(laddr6->sin6_port), - raddr6->sin6_addr.s6_addr, ntohs(raddr6->sin6_port)); + pr_debug("%s saddr %pI6 sport 0x%x raddr %pI6 rport 0x%x\n", + __func__, laddr6->sin6_addr.s6_addr, + ntohs(laddr6->sin6_port), + raddr6->sin6_addr.s6_addr, ntohs(raddr6->sin6_port)); ep->dst = cxgb_find_route6(&dev->rdev.lldi, get_real_dev, laddr6->sin6_addr.s6_addr, raddr6->sin6_addr.s6_addr, @@ -3276,7 +3293,7 @@ int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) raddr6->sin6_scope_id); } if (!ep->dst) { - printk(KERN_ERR MOD "%s - cannot find route.\n", __func__); + pr_err("%s - cannot find route\n", __func__); err = -EHOSTUNREACH; goto fail3; } @@ -3284,13 +3301,13 @@ int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) err = import_ep(ep, iptype, ra, ep->dst, ep->com.dev, true, ep->com.dev->rdev.lldi.adapter_type, cm_id->tos); if (err) { - printk(KERN_ERR MOD "%s - cannot alloc l2e.\n", __func__); + pr_err("%s - cannot alloc l2e\n", __func__); goto fail4; } - PDBG("%s txq_idx %u tx_chan %u smac_idx %u rss_qid %u l2t_idx %u\n", - __func__, ep->txq_idx, ep->tx_chan, ep->smac_idx, ep->rss_qid, - ep->l2t->idx); + pr_debug("%s txq_idx %u tx_chan %u smac_idx %u rss_qid %u l2t_idx %u\n", + __func__, ep->txq_idx, ep->tx_chan, ep->smac_idx, ep->rss_qid, + ep->l2t->idx); state_set(&ep->com, CONNECTING); ep->tos = cm_id->tos; @@ -3398,12 +3415,12 @@ int c4iw_create_listen(struct iw_cm_id *cm_id, int backlog) ep = alloc_ep(sizeof(*ep), GFP_KERNEL); if (!ep) { - printk(KERN_ERR MOD "%s - cannot alloc ep.\n", __func__); + pr_err("%s - cannot alloc ep\n", __func__); err = -ENOMEM; goto fail1; } skb_queue_head_init(&ep->com.ep_skb_list); - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); ep->com.cm_id = cm_id; ref_cm_id(&ep->com); ep->com.dev = dev; @@ -3423,7 +3440,7 @@ int c4iw_create_listen(struct iw_cm_id *cm_id, int backlog) cm_id->m_local_addr.ss_family, ep); if (ep->stid == -1) { - printk(KERN_ERR MOD "%s - cannot alloc stid.\n", __func__); + pr_err("%s - cannot alloc stid\n", __func__); err = -ENOMEM; goto fail2; } @@ -3457,7 +3474,7 @@ int c4iw_destroy_listen(struct iw_cm_id *cm_id) int err; struct c4iw_listen_ep *ep = to_listen_ep(cm_id); - PDBG("%s ep %p\n", __func__, ep); + pr_debug("%s ep %p\n", __func__, ep); might_sleep(); state_set(&ep->com, DEAD); @@ -3498,8 +3515,8 @@ int c4iw_ep_disconnect(struct c4iw_ep *ep, int abrupt, gfp_t gfp) mutex_lock(&ep->com.mutex); - PDBG("%s ep %p state %s, abrupt %d\n", __func__, ep, - states[ep->com.state], abrupt); + pr_debug("%s ep %p state %s, abrupt %d\n", __func__, ep, + states[ep->com.state], abrupt); /* * Ref the ep here in case we have fatal errors causing the @@ -3552,8 +3569,8 @@ int c4iw_ep_disconnect(struct c4iw_ep *ep, int abrupt, gfp_t gfp) case MORIBUND: case ABORTING: case DEAD: - PDBG("%s ignoring disconnect ep %p state %u\n", - __func__, ep, ep->com.state); + pr_debug("%s ignoring disconnect ep %p state %u\n", + __func__, ep, ep->com.state); break; default: BUG(); @@ -3584,8 +3601,7 @@ int c4iw_ep_disconnect(struct c4iw_ep *ep, int abrupt, gfp_t gfp) C4IW_QP_ATTR_NEXT_STATE, &attrs, 1); if (ret) - pr_err(MOD - "%s - qp <- error failed!\n", + pr_err("%s - qp <- error failed!\n", __func__); } fatal = 1; @@ -3658,7 +3674,7 @@ static void passive_ofld_conn_reply(struct c4iw_dev *dev, struct sk_buff *skb, rpl_skb = (struct sk_buff *)(unsigned long)req->cookie; BUG_ON(!rpl_skb); if (req->retval) { - PDBG("%s passive open failure %d\n", __func__, req->retval); + pr_debug("%s passive open failure %d\n", __func__, req->retval); mutex_lock(&dev->rdev.stats.lock); dev->rdev.stats.pas_ofld_conn_fails++; mutex_unlock(&dev->rdev.stats.lock); @@ -3732,9 +3748,9 @@ static void build_cpl_pass_accept_req(struct sk_buff *skb, int stid , u8 tos) */ memset(&tmp_opt, 0, sizeof(tmp_opt)); tcp_clear_options(&tmp_opt); - tcp_parse_options(skb, &tmp_opt, 0, NULL); + tcp_parse_options(&init_net, skb, &tmp_opt, 0, NULL); - req = (struct cpl_pass_accept_req *)__skb_push(skb, sizeof(*req)); + req = __skb_push(skb, sizeof(*req)); memset(req, 0, sizeof(*req)); req->l2info = cpu_to_be16(SYN_INTF_V(intf) | SYN_MAC_IDX_V(RX_MACIDX_G( @@ -3784,8 +3800,9 @@ static void send_fw_pass_open_req(struct c4iw_dev *dev, struct sk_buff *skb, int ret; req_skb = alloc_skb(sizeof(struct fw_ofld_connection_wr), GFP_KERNEL); - req = (struct fw_ofld_connection_wr *)__skb_put(req_skb, sizeof(*req)); - memset(req, 0, sizeof(*req)); + if (!req_skb) + return; + req = __skb_put_zero(req_skb, sizeof(*req)); req->op_compl = htonl(WR_OP_V(FW_OFLD_CONNECTION_WR) | FW_WR_COMPL_F); req->len16_pkd = htonl(FW_WR_LEN16_V(DIV_ROUND_UP(sizeof(*req), 16))); req->le.version_cpl = htonl(FW_OFLD_CONNECTION_WR_CPL_F); @@ -3874,7 +3891,8 @@ static int rx_pkt(struct c4iw_dev *dev, struct sk_buff *skb) lep = (struct c4iw_ep *)get_ep_from_stid(dev, stid); if (!lep) { - PDBG("%s connect request on invalid stid %d\n", __func__, stid); + pr_debug("%s connect request on invalid stid %d\n", + __func__, stid); goto reject; } @@ -3911,9 +3929,9 @@ static int rx_pkt(struct c4iw_dev *dev, struct sk_buff *skb) skb_set_transport_header(skb, (void *)tcph - (void *)rss); skb_get(skb); - PDBG("%s lip 0x%x lport %u pip 0x%x pport %u tos %d\n", __func__, - ntohl(iph->daddr), ntohs(tcph->dest), ntohl(iph->saddr), - ntohs(tcph->source), iph->tos); + pr_debug("%s lip 0x%x lport %u pip 0x%x pport %u tos %d\n", __func__, + ntohl(iph->daddr), ntohs(tcph->dest), ntohl(iph->saddr), + ntohs(tcph->source), iph->tos); dst = cxgb_find_route(&dev->rdev.lldi, get_real_dev, iph->daddr, iph->saddr, tcph->dest, @@ -4010,8 +4028,8 @@ static void process_timeout(struct c4iw_ep *ep) int abort = 1; mutex_lock(&ep->com.mutex); - PDBG("%s ep %p tid %u state %d\n", __func__, ep, ep->hwtid, - ep->com.state); + pr_debug("%s ep %p tid %u state %d\n", __func__, ep, ep->hwtid, + ep->com.state); set_bit(TIMEDOUT, &ep->com.history); switch (ep->com.state) { case MPA_REQ_SENT: @@ -4141,8 +4159,8 @@ static int set_tcb_rpl(struct c4iw_dev *dev, struct sk_buff *skb) struct cpl_set_tcb_rpl *rpl = cplhdr(skb); if (rpl->status != CPL_ERR_NONE) { - printk(KERN_ERR MOD "Unexpected SET_TCB_RPL status %u " - "for tid %u\n", rpl->status, GET_TID(rpl)); + pr_err("Unexpected SET_TCB_RPL status %u for tid %u\n", + rpl->status, GET_TID(rpl)); } kfree_skb(skb); return 0; @@ -4154,13 +4172,13 @@ static int fw6_msg(struct c4iw_dev *dev, struct sk_buff *skb) struct c4iw_wr_wait *wr_waitp; int ret; - PDBG("%s type %u\n", __func__, rpl->type); + pr_debug("%s type %u\n", __func__, rpl->type); switch (rpl->type) { case FW6_TYPE_WR_RPL: ret = (int)((be64_to_cpu(rpl->data[0]) >> 8) & 0xff); wr_waitp = (struct c4iw_wr_wait *)(__force unsigned long) rpl->data[1]; - PDBG("%s wr_waitp %p ret %u\n", __func__, wr_waitp, ret); + pr_debug("%s wr_waitp %p ret %u\n", __func__, wr_waitp, ret); if (wr_waitp) c4iw_wake_up(wr_waitp, ret ? -ret : 0); kfree_skb(skb); @@ -4170,8 +4188,8 @@ static int fw6_msg(struct c4iw_dev *dev, struct sk_buff *skb) sched(dev, skb); break; default: - printk(KERN_ERR MOD "%s unexpected fw6 msg type %u\n", __func__, - rpl->type); + pr_err("%s unexpected fw6 msg type %u\n", + __func__, rpl->type); kfree_skb(skb); break; } @@ -4187,19 +4205,18 @@ static int peer_abort_intr(struct c4iw_dev *dev, struct sk_buff *skb) ep = get_ep_from_tid(dev, tid); /* This EP will be dereferenced in peer_abort() */ if (!ep) { - printk(KERN_WARNING MOD - "Abort on non-existent endpoint, tid %d\n", tid); + pr_warn("Abort on non-existent endpoint, tid %d\n", tid); kfree_skb(skb); return 0; } if (cxgb_is_neg_adv(req->status)) { - PDBG("%s Negative advice on abort- tid %u status %d (%s)\n", - __func__, ep->hwtid, req->status, - neg_adv_str(req->status)); + pr_debug("%s Negative advice on abort- tid %u status %d (%s)\n", + __func__, ep->hwtid, req->status, + neg_adv_str(req->status)); goto out; } - PDBG("%s ep %p tid %u state %u\n", __func__, ep, ep->hwtid, - ep->com.state); + pr_debug("%s ep %p tid %u state %u\n", __func__, ep, ep->hwtid, + ep->com.state); c4iw_wake_up(&ep->com.wr_wait, -ECONNRESET); out: diff --git a/drivers/infiniband/hw/cxgb4/cq.c b/drivers/infiniband/hw/cxgb4/cq.c index bec82a600d77..be07da1997e6 100644 --- a/drivers/infiniband/hw/cxgb4/cq.c +++ b/drivers/infiniband/hw/cxgb4/cq.c @@ -44,8 +44,7 @@ static int destroy_cq(struct c4iw_rdev *rdev, struct t4_cq *cq, wr_len = sizeof *res_wr + sizeof *res; set_wr_txq(skb, CPL_PRIORITY_CONTROL, 0); - res_wr = (struct fw_ri_res_wr *)__skb_put(skb, wr_len); - memset(res_wr, 0, wr_len); + res_wr = __skb_put_zero(skb, wr_len); res_wr->op_nres = cpu_to_be32( FW_WR_OP_V(FW_RI_RES_WR) | FW_RI_RES_WR_NRES_V(1) | @@ -114,8 +113,7 @@ static int create_cq(struct c4iw_rdev *rdev, struct t4_cq *cq, } set_wr_txq(skb, CPL_PRIORITY_CONTROL, 0); - res_wr = (struct fw_ri_res_wr *)__skb_put(skb, wr_len); - memset(res_wr, 0, wr_len); + res_wr = __skb_put_zero(skb, wr_len); res_wr->op_nres = cpu_to_be32( FW_WR_OP_V(FW_RI_RES_WR) | FW_RI_RES_WR_NRES_V(1) | @@ -146,7 +144,7 @@ static int create_cq(struct c4iw_rdev *rdev, struct t4_cq *cq, ret = c4iw_ofld_send(rdev, skb); if (ret) goto err4; - PDBG("%s wait_event wr_wait %p\n", __func__, &wr_wait); + pr_debug("%s wait_event wr_wait %p\n", __func__, &wr_wait); ret = c4iw_wait_for_reply(rdev, &wr_wait, 0, 0, __func__); if (ret) goto err4; @@ -159,7 +157,7 @@ static int create_cq(struct c4iw_rdev *rdev, struct t4_cq *cq, &cq->bar2_qid, user ? &cq->bar2_pa : NULL); if (user && !cq->bar2_pa) { - pr_warn(MOD "%s: cqid %u not in BAR2 range.\n", + pr_warn("%s: cqid %u not in BAR2 range\n", pci_name(rdev->lldi.pdev), cq->cqid); ret = -EINVAL; goto err4; @@ -180,8 +178,8 @@ static void insert_recv_cqe(struct t4_wq *wq, struct t4_cq *cq) { struct t4_cqe cqe; - PDBG("%s wq %p cq %p sw_cidx %u sw_pidx %u\n", __func__, - wq, cq, cq->sw_cidx, cq->sw_pidx); + pr_debug("%s wq %p cq %p sw_cidx %u sw_pidx %u\n", __func__, + wq, cq, cq->sw_cidx, cq->sw_pidx); memset(&cqe, 0, sizeof(cqe)); cqe.header = cpu_to_be32(CQE_STATUS_V(T4_ERR_SWFLUSH) | CQE_OPCODE_V(FW_RI_SEND) | @@ -199,8 +197,8 @@ int c4iw_flush_rq(struct t4_wq *wq, struct t4_cq *cq, int count) int in_use = wq->rq.in_use - count; BUG_ON(in_use < 0); - PDBG("%s wq %p cq %p rq.in_use %u skip count %u\n", __func__, - wq, cq, wq->rq.in_use, count); + pr_debug("%s wq %p cq %p rq.in_use %u skip count %u\n", __func__, + wq, cq, wq->rq.in_use, count); while (in_use--) { insert_recv_cqe(wq, cq); flushed++; @@ -213,8 +211,8 @@ static void insert_sq_cqe(struct t4_wq *wq, struct t4_cq *cq, { struct t4_cqe cqe; - PDBG("%s wq %p cq %p sw_cidx %u sw_pidx %u\n", __func__, - wq, cq, cq->sw_cidx, cq->sw_pidx); + pr_debug("%s wq %p cq %p sw_cidx %u sw_pidx %u\n", __func__, + wq, cq, cq->sw_cidx, cq->sw_pidx); memset(&cqe, 0, sizeof(cqe)); cqe.header = cpu_to_be32(CQE_STATUS_V(T4_ERR_SWFLUSH) | CQE_OPCODE_V(swcqe->opcode) | @@ -283,8 +281,8 @@ static void flush_completed_wrs(struct t4_wq *wq, struct t4_cq *cq) /* * Insert this completed cqe into the swcq. */ - PDBG("%s moving cqe into swcq sq idx %u cq idx %u\n", - __func__, cidx, cq->sw_pidx); + pr_debug("%s moving cqe into swcq sq idx %u cq idx %u\n", + __func__, cidx, cq->sw_pidx); swsqe->cqe.header |= htonl(CQE_SWCQE_V(1)); cq->sw_queue[cq->sw_pidx] = swsqe->cqe; t4_swcq_produce(cq); @@ -339,7 +337,7 @@ void c4iw_flush_hw_cq(struct c4iw_cq *chp) struct t4_swsqe *swsqe; int ret; - PDBG("%s cqid 0x%x\n", __func__, chp->cq.cqid); + pr_debug("%s cqid 0x%x\n", __func__, chp->cq.cqid); ret = t4_next_hw_cqe(&chp->cq, &hw_cqe); /* @@ -432,7 +430,7 @@ void c4iw_count_rcqes(struct t4_cq *cq, struct t4_wq *wq, int *count) u32 ptr; *count = 0; - PDBG("%s count zero %d\n", __func__, *count); + pr_debug("%s count zero %d\n", __func__, *count); ptr = cq->sw_cidx; while (ptr != cq->sw_pidx) { cqe = &cq->sw_queue[ptr]; @@ -442,7 +440,7 @@ void c4iw_count_rcqes(struct t4_cq *cq, struct t4_wq *wq, int *count) if (++ptr == cq->size) ptr = 0; } - PDBG("%s cq %p count %d\n", __func__, cq, *count); + pr_debug("%s cq %p count %d\n", __func__, cq, *count); } /* @@ -473,12 +471,11 @@ static int poll_cq(struct t4_wq *wq, struct t4_cq *cq, struct t4_cqe *cqe, if (ret) return ret; - PDBG("%s CQE OVF %u qpid 0x%0x genbit %u type %u status 0x%0x" - " opcode 0x%0x len 0x%0x wrid_hi_stag 0x%x wrid_low_msn 0x%x\n", - __func__, CQE_OVFBIT(hw_cqe), CQE_QPID(hw_cqe), - CQE_GENBIT(hw_cqe), CQE_TYPE(hw_cqe), CQE_STATUS(hw_cqe), - CQE_OPCODE(hw_cqe), CQE_LEN(hw_cqe), CQE_WRID_HI(hw_cqe), - CQE_WRID_LOW(hw_cqe)); + pr_debug("%s CQE OVF %u qpid 0x%0x genbit %u type %u status 0x%0x opcode 0x%0x len 0x%0x wrid_hi_stag 0x%x wrid_low_msn 0x%x\n", + __func__, CQE_OVFBIT(hw_cqe), CQE_QPID(hw_cqe), + CQE_GENBIT(hw_cqe), CQE_TYPE(hw_cqe), CQE_STATUS(hw_cqe), + CQE_OPCODE(hw_cqe), CQE_LEN(hw_cqe), CQE_WRID_HI(hw_cqe), + CQE_WRID_LOW(hw_cqe)); /* * skip cqe's not affiliated with a QP. @@ -606,8 +603,8 @@ static int poll_cq(struct t4_wq *wq, struct t4_cq *cq, struct t4_cqe *cqe, if (!SW_CQE(hw_cqe) && (CQE_WRID_SQ_IDX(hw_cqe) != wq->sq.cidx)) { struct t4_swsqe *swsqe; - PDBG("%s out of order completion going in sw_sq at idx %u\n", - __func__, CQE_WRID_SQ_IDX(hw_cqe)); + pr_debug("%s out of order completion going in sw_sq at idx %u\n", + __func__, CQE_WRID_SQ_IDX(hw_cqe)); swsqe = &wq->sq.sw_sq[CQE_WRID_SQ_IDX(hw_cqe)]; swsqe->cqe = *hw_cqe; swsqe->complete = 1; @@ -641,13 +638,13 @@ proc_cqe: BUG_ON(wq->sq.in_use <= 0 && wq->sq.in_use >= wq->sq.size); wq->sq.cidx = (uint16_t)idx; - PDBG("%s completing sq idx %u\n", __func__, wq->sq.cidx); + pr_debug("%s completing sq idx %u\n", __func__, wq->sq.cidx); *cookie = wq->sq.sw_sq[wq->sq.cidx].wr_id; if (c4iw_wr_log) c4iw_log_wr_stats(wq, hw_cqe); t4_sq_consume(wq); } else { - PDBG("%s completing rq idx %u\n", __func__, wq->rq.cidx); + pr_debug("%s completing rq idx %u\n", __func__, wq->rq.cidx); *cookie = wq->rq.sw_rq[wq->rq.cidx].wr_id; BUG_ON(t4_rq_empty(wq)); if (c4iw_wr_log) @@ -664,12 +661,12 @@ flush_wq: skip_cqe: if (SW_CQE(hw_cqe)) { - PDBG("%s cq %p cqid 0x%x skip sw cqe cidx %u\n", - __func__, cq, cq->cqid, cq->sw_cidx); + pr_debug("%s cq %p cqid 0x%x skip sw cqe cidx %u\n", + __func__, cq, cq->cqid, cq->sw_cidx); t4_swcq_consume(cq); } else { - PDBG("%s cq %p cqid 0x%x skip hw cqe cidx %u\n", - __func__, cq, cq->cqid, cq->cidx); + pr_debug("%s cq %p cqid 0x%x skip hw cqe cidx %u\n", + __func__, cq, cq->cqid, cq->cidx); t4_hwcq_consume(cq); } return ret; @@ -715,10 +712,12 @@ static int c4iw_poll_cq_one(struct c4iw_cq *chp, struct ib_wc *wc) wc->vendor_err = CQE_STATUS(&cqe); wc->wc_flags = 0; - PDBG("%s qpid 0x%x type %d opcode %d status 0x%x len %u wrid hi 0x%x " - "lo 0x%x cookie 0x%llx\n", __func__, CQE_QPID(&cqe), - CQE_TYPE(&cqe), CQE_OPCODE(&cqe), CQE_STATUS(&cqe), CQE_LEN(&cqe), - CQE_WRID_HI(&cqe), CQE_WRID_LOW(&cqe), (unsigned long long)cookie); + pr_debug("%s qpid 0x%x type %d opcode %d status 0x%x len %u wrid hi 0x%x lo 0x%x cookie 0x%llx\n", + __func__, CQE_QPID(&cqe), + CQE_TYPE(&cqe), CQE_OPCODE(&cqe), + CQE_STATUS(&cqe), CQE_LEN(&cqe), + CQE_WRID_HI(&cqe), CQE_WRID_LOW(&cqe), + (unsigned long long)cookie); if (CQE_TYPE(&cqe) == 0) { if (!CQE_STATUS(&cqe)) @@ -766,8 +765,7 @@ static int c4iw_poll_cq_one(struct c4iw_cq *chp, struct ib_wc *wc) wc->opcode = IB_WC_SEND; break; default: - printk(KERN_ERR MOD "Unexpected opcode %d " - "in the CQE received for QPID=0x%0x\n", + pr_err("Unexpected opcode %d in the CQE received for QPID=0x%0x\n", CQE_OPCODE(&cqe), CQE_QPID(&cqe)); ret = -EINVAL; goto out; @@ -822,8 +820,7 @@ static int c4iw_poll_cq_one(struct c4iw_cq *chp, struct ib_wc *wc) wc->status = IB_WC_WR_FLUSH_ERR; break; default: - printk(KERN_ERR MOD - "Unexpected cqe_status 0x%x for QPID=0x%0x\n", + pr_err("Unexpected cqe_status 0x%x for QPID=0x%0x\n", CQE_STATUS(&cqe), CQE_QPID(&cqe)); wc->status = IB_WC_FATAL_ERR; } @@ -860,7 +857,7 @@ int c4iw_destroy_cq(struct ib_cq *ib_cq) struct c4iw_cq *chp; struct c4iw_ucontext *ucontext; - PDBG("%s ib_cq %p\n", __func__, ib_cq); + pr_debug("%s ib_cq %p\n", __func__, ib_cq); chp = to_c4iw_cq(ib_cq); remove_handle(chp->rhp, &chp->rhp->cqidr, chp->cq.cqid); @@ -892,7 +889,7 @@ struct ib_cq *c4iw_create_cq(struct ib_device *ibdev, size_t memsize, hwentries; struct c4iw_mm_entry *mm, *mm2; - PDBG("%s ib_dev %p entries %d\n", __func__, ibdev, entries); + pr_debug("%s ib_dev %p entries %d\n", __func__, ibdev, entries); if (attr->flags) return ERR_PTR(-EINVAL); @@ -966,6 +963,7 @@ struct ib_cq *c4iw_create_cq(struct ib_device *ibdev, goto err3; if (ucontext) { + ret = -ENOMEM; mm = kmalloc(sizeof *mm, GFP_KERNEL); if (!mm) goto err4; @@ -998,9 +996,9 @@ struct ib_cq *c4iw_create_cq(struct ib_device *ibdev, mm2->len = PAGE_SIZE; insert_mmap(ucontext, mm2); } - PDBG("%s cqid 0x%0x chp %p size %u memsize %zu, dma_addr 0x%0llx\n", - __func__, chp->cq.cqid, chp, chp->cq.size, - chp->cq.memsize, (unsigned long long) chp->cq.dma_addr); + pr_debug("%s cqid 0x%0x chp %p size %u memsize %zu, dma_addr 0x%0llx\n", + __func__, chp->cq.cqid, chp, chp->cq.size, + chp->cq.memsize, (unsigned long long)chp->cq.dma_addr); return &chp->ibcq; err6: kfree(mm2); diff --git a/drivers/infiniband/hw/cxgb4/device.c b/drivers/infiniband/hw/cxgb4/device.c index 40c0e7b9fc6e..ae0b79aeea2e 100644 --- a/drivers/infiniband/hw/cxgb4/device.c +++ b/drivers/infiniband/hw/cxgb4/device.c @@ -214,6 +214,52 @@ static const struct file_operations wr_log_debugfs_fops = { .write = wr_log_clear, }; +static struct sockaddr_in zero_sin = { + .sin_family = AF_INET, +}; + +static struct sockaddr_in6 zero_sin6 = { + .sin6_family = AF_INET6, +}; + +static void set_ep_sin_addrs(struct c4iw_ep *ep, + struct sockaddr_in **lsin, + struct sockaddr_in **rsin, + struct sockaddr_in **m_lsin, + struct sockaddr_in **m_rsin) +{ + struct iw_cm_id *id = ep->com.cm_id; + + *lsin = (struct sockaddr_in *)&ep->com.local_addr; + *rsin = (struct sockaddr_in *)&ep->com.remote_addr; + if (id) { + *m_lsin = (struct sockaddr_in *)&id->m_local_addr; + *m_rsin = (struct sockaddr_in *)&id->m_remote_addr; + } else { + *m_lsin = &zero_sin; + *m_rsin = &zero_sin; + } +} + +static void set_ep_sin6_addrs(struct c4iw_ep *ep, + struct sockaddr_in6 **lsin6, + struct sockaddr_in6 **rsin6, + struct sockaddr_in6 **m_lsin6, + struct sockaddr_in6 **m_rsin6) +{ + struct iw_cm_id *id = ep->com.cm_id; + + *lsin6 = (struct sockaddr_in6 *)&ep->com.local_addr; + *rsin6 = (struct sockaddr_in6 *)&ep->com.remote_addr; + if (id) { + *m_lsin6 = (struct sockaddr_in6 *)&id->m_local_addr; + *m_rsin6 = (struct sockaddr_in6 *)&id->m_remote_addr; + } else { + *m_lsin6 = &zero_sin6; + *m_rsin6 = &zero_sin6; + } +} + static int dump_qp(int id, void *p, void *data) { struct c4iw_qp *qp = p; @@ -229,16 +275,15 @@ static int dump_qp(int id, void *p, void *data) return 1; if (qp->ep) { - if (qp->ep->com.local_addr.ss_family == AF_INET) { - struct sockaddr_in *lsin = (struct sockaddr_in *) - &qp->ep->com.cm_id->local_addr; - struct sockaddr_in *rsin = (struct sockaddr_in *) - &qp->ep->com.cm_id->remote_addr; - struct sockaddr_in *mapped_lsin = (struct sockaddr_in *) - &qp->ep->com.cm_id->m_local_addr; - struct sockaddr_in *mapped_rsin = (struct sockaddr_in *) - &qp->ep->com.cm_id->m_remote_addr; + struct c4iw_ep *ep = qp->ep; + + if (ep->com.local_addr.ss_family == AF_INET) { + struct sockaddr_in *lsin; + struct sockaddr_in *rsin; + struct sockaddr_in *m_lsin; + struct sockaddr_in *m_rsin; + set_ep_sin_addrs(ep, &lsin, &rsin, &m_lsin, &m_rsin); cc = snprintf(qpd->buf + qpd->pos, space, "rc qp sq id %u rq id %u state %u " "onchip %u ep tid %u state %u " @@ -246,23 +291,19 @@ static int dump_qp(int id, void *p, void *data) qp->wq.sq.qid, qp->wq.rq.qid, (int)qp->attr.state, qp->wq.sq.flags & T4_SQ_ONCHIP, - qp->ep->hwtid, (int)qp->ep->com.state, + ep->hwtid, (int)ep->com.state, &lsin->sin_addr, ntohs(lsin->sin_port), - ntohs(mapped_lsin->sin_port), + ntohs(m_lsin->sin_port), &rsin->sin_addr, ntohs(rsin->sin_port), - ntohs(mapped_rsin->sin_port)); + ntohs(m_rsin->sin_port)); } else { - struct sockaddr_in6 *lsin6 = (struct sockaddr_in6 *) - &qp->ep->com.cm_id->local_addr; - struct sockaddr_in6 *rsin6 = (struct sockaddr_in6 *) - &qp->ep->com.cm_id->remote_addr; - struct sockaddr_in6 *mapped_lsin6 = - (struct sockaddr_in6 *) - &qp->ep->com.cm_id->m_local_addr; - struct sockaddr_in6 *mapped_rsin6 = - (struct sockaddr_in6 *) - &qp->ep->com.cm_id->m_remote_addr; + struct sockaddr_in6 *lsin6; + struct sockaddr_in6 *rsin6; + struct sockaddr_in6 *m_lsin6; + struct sockaddr_in6 *m_rsin6; + set_ep_sin6_addrs(ep, &lsin6, &rsin6, &m_lsin6, + &m_rsin6); cc = snprintf(qpd->buf + qpd->pos, space, "rc qp sq id %u rq id %u state %u " "onchip %u ep tid %u state %u " @@ -270,13 +311,13 @@ static int dump_qp(int id, void *p, void *data) qp->wq.sq.qid, qp->wq.rq.qid, (int)qp->attr.state, qp->wq.sq.flags & T4_SQ_ONCHIP, - qp->ep->hwtid, (int)qp->ep->com.state, + ep->hwtid, (int)ep->com.state, &lsin6->sin6_addr, ntohs(lsin6->sin6_port), - ntohs(mapped_lsin6->sin6_port), + ntohs(m_lsin6->sin6_port), &rsin6->sin6_addr, ntohs(rsin6->sin6_port), - ntohs(mapped_rsin6->sin6_port)); + ntohs(m_rsin6->sin6_port)); } } else cc = snprintf(qpd->buf + qpd->pos, space, @@ -293,7 +334,7 @@ static int qp_release(struct inode *inode, struct file *file) { struct c4iw_debugfs_data *qpd = file->private_data; if (!qpd) { - printk(KERN_INFO "%s null qpd?\n", __func__); + pr_info("%s null qpd?\n", __func__); return 0; } vfree(qpd->buf); @@ -381,7 +422,7 @@ static int stag_release(struct inode *inode, struct file *file) { struct c4iw_debugfs_data *stagd = file->private_data; if (!stagd) { - printk(KERN_INFO "%s null stagd?\n", __func__); + pr_info("%s null stagd?\n", __func__); return 0; } vfree(stagd->buf); @@ -533,15 +574,12 @@ static int dump_ep(int id, void *p, void *data) return 1; if (ep->com.local_addr.ss_family == AF_INET) { - struct sockaddr_in *lsin = (struct sockaddr_in *) - &ep->com.cm_id->local_addr; - struct sockaddr_in *rsin = (struct sockaddr_in *) - &ep->com.cm_id->remote_addr; - struct sockaddr_in *mapped_lsin = (struct sockaddr_in *) - &ep->com.cm_id->m_local_addr; - struct sockaddr_in *mapped_rsin = (struct sockaddr_in *) - &ep->com.cm_id->m_remote_addr; + struct sockaddr_in *lsin; + struct sockaddr_in *rsin; + struct sockaddr_in *m_lsin; + struct sockaddr_in *m_rsin; + set_ep_sin_addrs(ep, &lsin, &rsin, &m_lsin, &m_rsin); cc = snprintf(epd->buf + epd->pos, space, "ep %p cm_id %p qp %p state %d flags 0x%lx " "history 0x%lx hwtid %d atid %d " @@ -553,19 +591,16 @@ static int dump_ep(int id, void *p, void *data) ep->stats.connect_neg_adv, ep->stats.abort_neg_adv, &lsin->sin_addr, ntohs(lsin->sin_port), - ntohs(mapped_lsin->sin_port), + ntohs(m_lsin->sin_port), &rsin->sin_addr, ntohs(rsin->sin_port), - ntohs(mapped_rsin->sin_port)); + ntohs(m_rsin->sin_port)); } else { - struct sockaddr_in6 *lsin6 = (struct sockaddr_in6 *) - &ep->com.cm_id->local_addr; - struct sockaddr_in6 *rsin6 = (struct sockaddr_in6 *) - &ep->com.cm_id->remote_addr; - struct sockaddr_in6 *mapped_lsin6 = (struct sockaddr_in6 *) - &ep->com.cm_id->m_local_addr; - struct sockaddr_in6 *mapped_rsin6 = (struct sockaddr_in6 *) - &ep->com.cm_id->m_remote_addr; + struct sockaddr_in6 *lsin6; + struct sockaddr_in6 *rsin6; + struct sockaddr_in6 *m_lsin6; + struct sockaddr_in6 *m_rsin6; + set_ep_sin6_addrs(ep, &lsin6, &rsin6, &m_lsin6, &m_rsin6); cc = snprintf(epd->buf + epd->pos, space, "ep %p cm_id %p qp %p state %d flags 0x%lx " "history 0x%lx hwtid %d atid %d " @@ -577,9 +612,9 @@ static int dump_ep(int id, void *p, void *data) ep->stats.connect_neg_adv, ep->stats.abort_neg_adv, &lsin6->sin6_addr, ntohs(lsin6->sin6_port), - ntohs(mapped_lsin6->sin6_port), + ntohs(m_lsin6->sin6_port), &rsin6->sin6_addr, ntohs(rsin6->sin6_port), - ntohs(mapped_rsin6->sin6_port)); + ntohs(m_rsin6->sin6_port)); } if (cc < space) epd->pos += cc; @@ -600,7 +635,7 @@ static int dump_listen_ep(int id, void *p, void *data) if (ep->com.local_addr.ss_family == AF_INET) { struct sockaddr_in *lsin = (struct sockaddr_in *) &ep->com.cm_id->local_addr; - struct sockaddr_in *mapped_lsin = (struct sockaddr_in *) + struct sockaddr_in *m_lsin = (struct sockaddr_in *) &ep->com.cm_id->m_local_addr; cc = snprintf(epd->buf + epd->pos, space, @@ -609,11 +644,11 @@ static int dump_listen_ep(int id, void *p, void *data) ep, ep->com.cm_id, (int)ep->com.state, ep->com.flags, ep->stid, ep->backlog, &lsin->sin_addr, ntohs(lsin->sin_port), - ntohs(mapped_lsin->sin_port)); + ntohs(m_lsin->sin_port)); } else { struct sockaddr_in6 *lsin6 = (struct sockaddr_in6 *) &ep->com.cm_id->local_addr; - struct sockaddr_in6 *mapped_lsin6 = (struct sockaddr_in6 *) + struct sockaddr_in6 *m_lsin6 = (struct sockaddr_in6 *) &ep->com.cm_id->m_local_addr; cc = snprintf(epd->buf + epd->pos, space, @@ -622,7 +657,7 @@ static int dump_listen_ep(int id, void *p, void *data) ep, ep->com.cm_id, (int)ep->com.state, ep->com.flags, ep->stid, ep->backlog, &lsin6->sin6_addr, ntohs(lsin6->sin6_port), - ntohs(mapped_lsin6->sin6_port)); + ntohs(m_lsin6->sin6_port)); } if (cc < space) epd->pos += cc; @@ -732,7 +767,7 @@ void c4iw_release_dev_ucontext(struct c4iw_rdev *rdev, kfree(entry); } - list_for_each_safe(pos, nxt, &uctx->qpids) { + list_for_each_safe(pos, nxt, &uctx->cqids) { entry = list_entry(pos, struct c4iw_qid_list, entry); list_del_init(&entry->entry); kfree(entry); @@ -761,15 +796,14 @@ static int c4iw_rdev_open(struct c4iw_rdev *rdev) * cqid and qpid range must match for now. */ if (rdev->lldi.udb_density != rdev->lldi.ucq_density) { - pr_err(MOD "%s: unsupported udb/ucq densities %u/%u\n", + pr_err("%s: unsupported udb/ucq densities %u/%u\n", pci_name(rdev->lldi.pdev), rdev->lldi.udb_density, rdev->lldi.ucq_density); return -EINVAL; } if (rdev->lldi.vr->qp.start != rdev->lldi.vr->cq.start || rdev->lldi.vr->qp.size != rdev->lldi.vr->cq.size) { - pr_err(MOD "%s: unsupported qp and cq id ranges " - "qp start %u size %u cq start %u size %u\n", + pr_err("%s: unsupported qp and cq id ranges qp start %u size %u cq start %u size %u\n", pci_name(rdev->lldi.pdev), rdev->lldi.vr->qp.start, rdev->lldi.vr->qp.size, rdev->lldi.vr->cq.size, rdev->lldi.vr->cq.size); @@ -778,23 +812,20 @@ static int c4iw_rdev_open(struct c4iw_rdev *rdev) rdev->qpmask = rdev->lldi.udb_density - 1; rdev->cqmask = rdev->lldi.ucq_density - 1; - PDBG("%s dev %s stag start 0x%0x size 0x%0x num stags %d " - "pbl start 0x%0x size 0x%0x rq start 0x%0x size 0x%0x " - "qp qid start %u size %u cq qid start %u size %u\n", - __func__, pci_name(rdev->lldi.pdev), rdev->lldi.vr->stag.start, - rdev->lldi.vr->stag.size, c4iw_num_stags(rdev), - rdev->lldi.vr->pbl.start, - rdev->lldi.vr->pbl.size, rdev->lldi.vr->rq.start, - rdev->lldi.vr->rq.size, - rdev->lldi.vr->qp.start, - rdev->lldi.vr->qp.size, - rdev->lldi.vr->cq.start, - rdev->lldi.vr->cq.size); - PDBG("udb %pR db_reg %p gts_reg %p " - "qpmask 0x%x cqmask 0x%x\n", - &rdev->lldi.pdev->resource[2], - rdev->lldi.db_reg, rdev->lldi.gts_reg, - rdev->qpmask, rdev->cqmask); + pr_debug("%s dev %s stag start 0x%0x size 0x%0x num stags %d pbl start 0x%0x size 0x%0x rq start 0x%0x size 0x%0x qp qid start %u size %u cq qid start %u size %u\n", + __func__, pci_name(rdev->lldi.pdev), rdev->lldi.vr->stag.start, + rdev->lldi.vr->stag.size, c4iw_num_stags(rdev), + rdev->lldi.vr->pbl.start, + rdev->lldi.vr->pbl.size, rdev->lldi.vr->rq.start, + rdev->lldi.vr->rq.size, + rdev->lldi.vr->qp.start, + rdev->lldi.vr->qp.size, + rdev->lldi.vr->cq.start, + rdev->lldi.vr->cq.size); + pr_debug("udb %pR db_reg %p gts_reg %p qpmask 0x%x cqmask 0x%x\n", + &rdev->lldi.pdev->resource[2], + rdev->lldi.db_reg, rdev->lldi.gts_reg, + rdev->qpmask, rdev->cqmask); if (c4iw_num_stags(rdev) == 0) return -EINVAL; @@ -808,22 +839,22 @@ static int c4iw_rdev_open(struct c4iw_rdev *rdev) err = c4iw_init_resource(rdev, c4iw_num_stags(rdev), T4_MAX_NUM_PD); if (err) { - printk(KERN_ERR MOD "error %d initializing resources\n", err); + pr_err("error %d initializing resources\n", err); return err; } err = c4iw_pblpool_create(rdev); if (err) { - printk(KERN_ERR MOD "error %d initializing pbl pool\n", err); + pr_err("error %d initializing pbl pool\n", err); goto destroy_resource; } err = c4iw_rqtpool_create(rdev); if (err) { - printk(KERN_ERR MOD "error %d initializing rqt pool\n", err); + pr_err("error %d initializing rqt pool\n", err); goto destroy_pblpool; } err = c4iw_ocqp_pool_create(rdev); if (err) { - printk(KERN_ERR MOD "error %d initializing ocqp pool\n", err); + pr_err("error %d initializing ocqp pool\n", err); goto destroy_rqtpool; } rdev->status_page = (struct t4_dev_status_page *) @@ -849,13 +880,15 @@ static int c4iw_rdev_open(struct c4iw_rdev *rdev) rdev->free_workq = create_singlethread_workqueue("iw_cxgb4_free"); if (!rdev->free_workq) { err = -ENOMEM; - goto err_free_status_page; + goto err_free_status_page_and_wr_log; } rdev->status_page->db_off = 0; return 0; -err_free_status_page: +err_free_status_page_and_wr_log: + if (c4iw_wr_log && rdev->wr_log) + kfree(rdev->wr_log); free_page((unsigned long)rdev->status_page); destroy_ocqp_pool: c4iw_ocqp_pool_destroy(rdev); @@ -872,9 +905,11 @@ static void c4iw_rdev_close(struct c4iw_rdev *rdev) { destroy_workqueue(rdev->free_workq); kfree(rdev->wr_log); + c4iw_release_dev_ucontext(rdev, &rdev->uctx); free_page((unsigned long)rdev->status_page); c4iw_pblpool_destroy(rdev); c4iw_rqtpool_destroy(rdev); + c4iw_ocqp_pool_destroy(rdev); c4iw_destroy_resource(&rdev->resource); } @@ -901,7 +936,7 @@ static void c4iw_dealloc(struct uld_ctx *ctx) static void c4iw_remove(struct uld_ctx *ctx) { - PDBG("%s c4iw_dev %p\n", __func__, ctx->dev); + pr_debug("%s c4iw_dev %p\n", __func__, ctx->dev); c4iw_unregister_device(ctx->dev); c4iw_dealloc(ctx); } @@ -919,28 +954,28 @@ static struct c4iw_dev *c4iw_alloc(const struct cxgb4_lld_info *infop) int ret; if (!rdma_supported(infop)) { - printk(KERN_INFO MOD "%s: RDMA not supported on this device.\n", - pci_name(infop->pdev)); + pr_info("%s: RDMA not supported on this device\n", + pci_name(infop->pdev)); return ERR_PTR(-ENOSYS); } if (!ocqp_supported(infop)) - pr_info("%s: On-Chip Queues not supported on this device.\n", + pr_info("%s: On-Chip Queues not supported on this device\n", pci_name(infop->pdev)); devp = (struct c4iw_dev *)ib_alloc_device(sizeof(*devp)); if (!devp) { - printk(KERN_ERR MOD "Cannot allocate ib device\n"); + pr_err("Cannot allocate ib device\n"); return ERR_PTR(-ENOMEM); } devp->rdev.lldi = *infop; /* init various hw-queue params based on lld info */ - PDBG("%s: Ing. padding boundary is %d, egrsstatuspagesize = %d\n", - __func__, devp->rdev.lldi.sge_ingpadboundary, - devp->rdev.lldi.sge_egrstatuspagesize); + pr_debug("%s: Ing. padding boundary is %d, egrsstatuspagesize = %d\n", + __func__, devp->rdev.lldi.sge_ingpadboundary, + devp->rdev.lldi.sge_egrstatuspagesize); devp->rdev.hw_queue.t4_eq_status_entries = - devp->rdev.lldi.sge_ingpadboundary > 64 ? 2 : 1; + devp->rdev.lldi.sge_egrstatuspagesize / 64; devp->rdev.hw_queue.t4_max_eq_size = 65520; devp->rdev.hw_queue.t4_max_iq_size = 65520; devp->rdev.hw_queue.t4_max_rq_size = 8192 - @@ -965,7 +1000,7 @@ static struct c4iw_dev *c4iw_alloc(const struct cxgb4_lld_info *infop) devp->rdev.bar2_kva = ioremap_wc(devp->rdev.bar2_pa, pci_resource_len(devp->rdev.lldi.pdev, 2)); if (!devp->rdev.bar2_kva) { - pr_err(MOD "Unable to ioremap BAR2\n"); + pr_err("Unable to ioremap BAR2\n"); ib_dealloc_device(&devp->ibdev); return ERR_PTR(-EINVAL); } @@ -977,20 +1012,19 @@ static struct c4iw_dev *c4iw_alloc(const struct cxgb4_lld_info *infop) devp->rdev.oc_mw_kva = ioremap_wc(devp->rdev.oc_mw_pa, devp->rdev.lldi.vr->ocq.size); if (!devp->rdev.oc_mw_kva) { - pr_err(MOD "Unable to ioremap onchip mem\n"); + pr_err("Unable to ioremap onchip mem\n"); ib_dealloc_device(&devp->ibdev); return ERR_PTR(-EINVAL); } } - PDBG(KERN_INFO MOD "ocq memory: " - "hw_start 0x%x size %u mw_pa 0x%lx mw_kva %p\n", - devp->rdev.lldi.vr->ocq.start, devp->rdev.lldi.vr->ocq.size, - devp->rdev.oc_mw_pa, devp->rdev.oc_mw_kva); + pr_debug("ocq memory: hw_start 0x%x size %u mw_pa 0x%lx mw_kva %p\n", + devp->rdev.lldi.vr->ocq.start, devp->rdev.lldi.vr->ocq.size, + devp->rdev.oc_mw_pa, devp->rdev.oc_mw_kva); ret = c4iw_rdev_open(&devp->rdev); if (ret) { - printk(KERN_ERR MOD "Unable to open CXIO rdev err %d\n", ret); + pr_err("Unable to open CXIO rdev err %d\n", ret); ib_dealloc_device(&devp->ibdev); return ERR_PTR(ret); } @@ -1036,17 +1070,17 @@ static void *c4iw_uld_add(const struct cxgb4_lld_info *infop) } ctx->lldi = *infop; - PDBG("%s found device %s nchan %u nrxq %u ntxq %u nports %u\n", - __func__, pci_name(ctx->lldi.pdev), - ctx->lldi.nchan, ctx->lldi.nrxq, - ctx->lldi.ntxq, ctx->lldi.nports); + pr_debug("%s found device %s nchan %u nrxq %u ntxq %u nports %u\n", + __func__, pci_name(ctx->lldi.pdev), + ctx->lldi.nchan, ctx->lldi.nrxq, + ctx->lldi.ntxq, ctx->lldi.nports); mutex_lock(&dev_mutex); list_add_tail(&ctx->entry, &uld_ctx_list); mutex_unlock(&dev_mutex); for (i = 0; i < ctx->lldi.nrxq; i++) - PDBG("rxqid[%u] %u\n", i, ctx->lldi.rxq_ids[i]); + pr_debug("rxqid[%u] %u\n", i, ctx->lldi.rxq_ids[i]); out: return ctx; } @@ -1103,8 +1137,7 @@ static inline int recv_rx_pkt(struct c4iw_dev *dev, const struct pkt_gl *gl, goto out; if (c4iw_handlers[opcode] == NULL) { - pr_info("%s no handler opcode 0x%x...\n", __func__, - opcode); + pr_info("%s no handler opcode 0x%x...\n", __func__, opcode); kfree_skb(skb); goto out; } @@ -1141,13 +1174,11 @@ static int c4iw_uld_rx_handler(void *handle, const __be64 *rsp, if (recv_rx_pkt(dev, gl, rsp)) return 0; - pr_info("%s: unexpected FL contents at %p, " \ - "RSS %#llx, FL %#llx, len %u\n", - pci_name(ctx->lldi.pdev), gl->va, - (unsigned long long)be64_to_cpu(*rsp), - (unsigned long long)be64_to_cpu( - *(__force __be64 *)gl->va), - gl->tot_len); + pr_info("%s: unexpected FL contents at %p, RSS %#llx, FL %#llx, len %u\n", + pci_name(ctx->lldi.pdev), gl->va, + be64_to_cpu(*rsp), + be64_to_cpu(*(__force __be64 *)gl->va), + gl->tot_len); return 0; } else { @@ -1160,8 +1191,7 @@ static int c4iw_uld_rx_handler(void *handle, const __be64 *rsp, if (c4iw_handlers[opcode]) { c4iw_handlers[opcode](dev, skb); } else { - pr_info("%s no handler opcode 0x%x...\n", __func__, - opcode); + pr_info("%s no handler opcode 0x%x...\n", __func__, opcode); kfree_skb(skb); } @@ -1174,17 +1204,16 @@ static int c4iw_uld_state_change(void *handle, enum cxgb4_state new_state) { struct uld_ctx *ctx = handle; - PDBG("%s new_state %u\n", __func__, new_state); + pr_debug("%s new_state %u\n", __func__, new_state); switch (new_state) { case CXGB4_STATE_UP: - printk(KERN_INFO MOD "%s: Up\n", pci_name(ctx->lldi.pdev)); + pr_info("%s: Up\n", pci_name(ctx->lldi.pdev)); if (!ctx->dev) { int ret; ctx->dev = c4iw_alloc(&ctx->lldi); if (IS_ERR(ctx->dev)) { - printk(KERN_ERR MOD - "%s: initialization failed: %ld\n", + pr_err("%s: initialization failed: %ld\n", pci_name(ctx->lldi.pdev), PTR_ERR(ctx->dev)); ctx->dev = NULL; @@ -1192,22 +1221,19 @@ static int c4iw_uld_state_change(void *handle, enum cxgb4_state new_state) } ret = c4iw_register_device(ctx->dev); if (ret) { - printk(KERN_ERR MOD - "%s: RDMA registration failed: %d\n", + pr_err("%s: RDMA registration failed: %d\n", pci_name(ctx->lldi.pdev), ret); c4iw_dealloc(ctx); } } break; case CXGB4_STATE_DOWN: - printk(KERN_INFO MOD "%s: Down\n", - pci_name(ctx->lldi.pdev)); + pr_info("%s: Down\n", pci_name(ctx->lldi.pdev)); if (ctx->dev) c4iw_remove(ctx); break; case CXGB4_STATE_START_RECOVERY: - printk(KERN_INFO MOD "%s: Fatal Error\n", - pci_name(ctx->lldi.pdev)); + pr_info("%s: Fatal Error\n", pci_name(ctx->lldi.pdev)); if (ctx->dev) { struct ib_event event; @@ -1220,8 +1246,7 @@ static int c4iw_uld_state_change(void *handle, enum cxgb4_state new_state) } break; case CXGB4_STATE_DETACH: - printk(KERN_INFO MOD "%s: Detach\n", - pci_name(ctx->lldi.pdev)); + pr_info("%s: Detach\n", pci_name(ctx->lldi.pdev)); if (ctx->dev) c4iw_remove(ctx); break; @@ -1371,9 +1396,7 @@ static void recover_lost_dbs(struct uld_ctx *ctx, struct qp_list *qp_list) t4_sq_host_wq_pidx(&qp->wq), t4_sq_wq_size(&qp->wq)); if (ret) { - pr_err(MOD "%s: Fatal error - " - "DB overflow recovery failed - " - "error syncing SQ qid %u\n", + pr_err("%s: Fatal error - DB overflow recovery failed - error syncing SQ qid %u\n", pci_name(ctx->lldi.pdev), qp->wq.sq.qid); spin_unlock(&qp->lock); spin_unlock_irq(&qp->rhp->lock); @@ -1387,9 +1410,7 @@ static void recover_lost_dbs(struct uld_ctx *ctx, struct qp_list *qp_list) t4_rq_wq_size(&qp->wq)); if (ret) { - pr_err(MOD "%s: Fatal error - " - "DB overflow recovery failed - " - "error syncing RQ qid %u\n", + pr_err("%s: Fatal error - DB overflow recovery failed - error syncing RQ qid %u\n", pci_name(ctx->lldi.pdev), qp->wq.rq.qid); spin_unlock(&qp->lock); spin_unlock_irq(&qp->rhp->lock); @@ -1420,7 +1441,7 @@ static void recover_queues(struct uld_ctx *ctx) /* flush the SGE contexts */ ret = cxgb4_flush_eq_cache(ctx->dev->rdev.lldi.ports[0]); if (ret) { - printk(KERN_ERR MOD "%s: Fatal error - DB overflow recovery failed\n", + pr_err("%s: Fatal error - DB overflow recovery failed\n", pci_name(ctx->lldi.pdev)); return; } @@ -1478,8 +1499,8 @@ static int c4iw_uld_control(void *handle, enum cxgb4_control control, ...) mutex_unlock(&ctx->dev->rdev.stats.lock); break; default: - printk(KERN_WARNING MOD "%s: unknown control cmd %u\n", - pci_name(ctx->lldi.pdev), control); + pr_warn("%s: unknown control cmd %u\n", + pci_name(ctx->lldi.pdev), control); break; } return 0; @@ -1508,8 +1529,7 @@ static int __init c4iw_init_module(void) c4iw_debugfs_root = debugfs_create_dir(DRV_NAME, NULL); if (!c4iw_debugfs_root) - printk(KERN_WARNING MOD - "could not create debugfs entry, continuing\n"); + pr_warn("could not create debugfs entry, continuing\n"); cxgb4_register_uld(CXGB4_ULD_RDMA, &c4iw_uld_info); diff --git a/drivers/infiniband/hw/cxgb4/ev.c b/drivers/infiniband/hw/cxgb4/ev.c index bdfac2ccb704..8f963df0bffc 100644 --- a/drivers/infiniband/hw/cxgb4/ev.c +++ b/drivers/infiniband/hw/cxgb4/ev.c @@ -47,17 +47,16 @@ static void print_tpte(struct c4iw_dev *dev, u32 stag) "%s cxgb4_read_tpte err %d\n", __func__, ret); return; } - PDBG("stag idx 0x%x valid %d key 0x%x state %d pdid %d " - "perm 0x%x ps %d len 0x%llx va 0x%llx\n", - stag & 0xffffff00, - FW_RI_TPTE_VALID_G(ntohl(tpte.valid_to_pdid)), - FW_RI_TPTE_STAGKEY_G(ntohl(tpte.valid_to_pdid)), - FW_RI_TPTE_STAGSTATE_G(ntohl(tpte.valid_to_pdid)), - FW_RI_TPTE_PDID_G(ntohl(tpte.valid_to_pdid)), - FW_RI_TPTE_PERM_G(ntohl(tpte.locread_to_qpid)), - FW_RI_TPTE_PS_G(ntohl(tpte.locread_to_qpid)), - ((u64)ntohl(tpte.len_hi) << 32) | ntohl(tpte.len_lo), - ((u64)ntohl(tpte.va_hi) << 32) | ntohl(tpte.va_lo_fbo)); + pr_debug("stag idx 0x%x valid %d key 0x%x state %d pdid %d perm 0x%x ps %d len 0x%llx va 0x%llx\n", + stag & 0xffffff00, + FW_RI_TPTE_VALID_G(ntohl(tpte.valid_to_pdid)), + FW_RI_TPTE_STAGKEY_G(ntohl(tpte.valid_to_pdid)), + FW_RI_TPTE_STAGSTATE_G(ntohl(tpte.valid_to_pdid)), + FW_RI_TPTE_PDID_G(ntohl(tpte.valid_to_pdid)), + FW_RI_TPTE_PERM_G(ntohl(tpte.locread_to_qpid)), + FW_RI_TPTE_PS_G(ntohl(tpte.locread_to_qpid)), + ((u64)ntohl(tpte.len_hi) << 32) | ntohl(tpte.len_lo), + ((u64)ntohl(tpte.va_hi) << 32) | ntohl(tpte.va_lo_fbo)); } static void dump_err_cqe(struct c4iw_dev *dev, struct t4_cqe *err_cqe) @@ -71,9 +70,9 @@ static void dump_err_cqe(struct c4iw_dev *dev, struct t4_cqe *err_cqe) CQE_STATUS(err_cqe), CQE_TYPE(err_cqe), ntohl(err_cqe->len), CQE_WRID_HI(err_cqe), CQE_WRID_LOW(err_cqe)); - PDBG("%016llx %016llx %016llx %016llx\n", - be64_to_cpu(p[0]), be64_to_cpu(p[1]), be64_to_cpu(p[2]), - be64_to_cpu(p[3])); + pr_debug("%016llx %016llx %016llx %016llx\n", + be64_to_cpu(p[0]), be64_to_cpu(p[1]), be64_to_cpu(p[2]), + be64_to_cpu(p[3])); /* * Ingress WRITE and READ_RESP errors provide @@ -124,8 +123,7 @@ void c4iw_ev_dispatch(struct c4iw_dev *dev, struct t4_cqe *err_cqe) spin_lock_irq(&dev->lock); qhp = get_qhp(dev, CQE_QPID(err_cqe)); if (!qhp) { - printk(KERN_ERR MOD "BAD AE qpid 0x%x opcode %d " - "status 0x%x type %d wrid.hi 0x%x wrid.lo 0x%x\n", + pr_err("BAD AE qpid 0x%x opcode %d status 0x%x type %d wrid.hi 0x%x wrid.lo 0x%x\n", CQE_QPID(err_cqe), CQE_OPCODE(err_cqe), CQE_STATUS(err_cqe), CQE_TYPE(err_cqe), CQE_WRID_HI(err_cqe), @@ -140,8 +138,7 @@ void c4iw_ev_dispatch(struct c4iw_dev *dev, struct t4_cqe *err_cqe) cqid = qhp->attr.rcq; chp = get_chp(dev, cqid); if (!chp) { - printk(KERN_ERR MOD "BAD AE cqid 0x%x qpid 0x%x opcode %d " - "status 0x%x type %d wrid.hi 0x%x wrid.lo 0x%x\n", + pr_err("BAD AE cqid 0x%x qpid 0x%x opcode %d status 0x%x type %d wrid.hi 0x%x wrid.lo 0x%x\n", cqid, CQE_QPID(err_cqe), CQE_OPCODE(err_cqe), CQE_STATUS(err_cqe), CQE_TYPE(err_cqe), CQE_WRID_HI(err_cqe), @@ -165,7 +162,7 @@ void c4iw_ev_dispatch(struct c4iw_dev *dev, struct t4_cqe *err_cqe) /* Completion Events */ case T4_ERR_SUCCESS: - printk(KERN_ERR MOD "AE with status 0!\n"); + pr_err("AE with status 0!\n"); break; case T4_ERR_STAG: @@ -207,7 +204,7 @@ void c4iw_ev_dispatch(struct c4iw_dev *dev, struct t4_cqe *err_cqe) break; default: - printk(KERN_ERR MOD "Unknown T4 status 0x%x QPID 0x%x\n", + pr_err("Unknown T4 status 0x%x QPID 0x%x\n", CQE_STATUS(err_cqe), qhp->wq.sq.qid); post_qp_event(dev, chp, qhp, err_cqe, IB_EVENT_QP_FATAL); break; @@ -237,7 +234,7 @@ int c4iw_ev_handler(struct c4iw_dev *dev, u32 qid) if (atomic_dec_and_test(&chp->refcnt)) wake_up(&chp->wait); } else { - PDBG("%s unknown cqid 0x%x\n", __func__, qid); + pr_debug("%s unknown cqid 0x%x\n", __func__, 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 8cd4d054a87e..819a30635d53 100644 --- a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h +++ b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h @@ -37,7 +37,7 @@ #include <linux/idr.h> #include <linux/completion.h> #include <linux/netdevice.h> -#include <linux/sched.h> +#include <linux/sched/mm.h> #include <linux/pci.h> #include <linux/dma-mapping.h> #include <linux/inet.h> @@ -64,12 +64,11 @@ #define DRV_NAME "iw_cxgb4" #define MOD DRV_NAME ":" -extern int c4iw_debug; -#define PDBG(fmt, args...) \ -do { \ - if (c4iw_debug) \ - printk(MOD fmt, ## args); \ -} while (0) +#ifdef pr_fmt +#undef pr_fmt +#endif + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt #include "t4.h" @@ -231,15 +230,15 @@ static inline int c4iw_wait_for_reply(struct c4iw_rdev *rdev, ret = wait_for_completion_timeout(&wr_waitp->completion, C4IW_WR_TO); if (!ret) { - PDBG("%s - Device %s not responding (disabling device) - tid %u qpid %u\n", - func, pci_name(rdev->lldi.pdev), hwtid, qpid); + pr_debug("%s - Device %s not responding (disabling device) - tid %u qpid %u\n", + func, pci_name(rdev->lldi.pdev), hwtid, qpid); rdev->flags |= T4_FATAL_ERROR; wr_waitp->ret = -EIO; } out: if (wr_waitp->ret) - PDBG("%s: FW reply %d tid %u qpid %u\n", - pci_name(rdev->lldi.pdev), wr_waitp->ret, hwtid, qpid); + pr_debug("%s: FW reply %d tid %u qpid %u\n", + pci_name(rdev->lldi.pdev), wr_waitp->ret, hwtid, qpid); return wr_waitp->ret; } @@ -538,8 +537,9 @@ static inline struct c4iw_mm_entry *remove_mmap(struct c4iw_ucontext *ucontext, if (mm->key == key && mm->len == len) { list_del_init(&mm->entry); spin_unlock(&ucontext->mmap_lock); - PDBG("%s key 0x%x addr 0x%llx len %d\n", __func__, - key, (unsigned long long) mm->addr, mm->len); + pr_debug("%s key 0x%x addr 0x%llx len %d\n", + __func__, key, + (unsigned long long)mm->addr, mm->len); return mm; } } @@ -551,8 +551,8 @@ static inline void insert_mmap(struct c4iw_ucontext *ucontext, struct c4iw_mm_entry *mm) { spin_lock(&ucontext->mmap_lock); - PDBG("%s key 0x%x addr 0x%llx len %d\n", __func__, - mm->key, (unsigned long long) mm->addr, mm->len); + pr_debug("%s key 0x%x addr 0x%llx len %d\n", + __func__, mm->key, (unsigned long long)mm->addr, mm->len); list_add_tail(&mm->entry, &ucontext->mmaps); spin_unlock(&ucontext->mmap_lock); } @@ -670,17 +670,19 @@ enum c4iw_mmid_state { #define MPA_V2_RDMA_READ_RTR 0x4000 #define MPA_V2_IRD_ORD_MASK 0x3FFF -#define c4iw_put_ep(ep) { \ - PDBG("put_ep (via %s:%u) ep %p refcnt %d\n", __func__, __LINE__, \ - ep, atomic_read(&((ep)->kref.refcount))); \ - WARN_ON(atomic_read(&((ep)->kref.refcount)) < 1); \ - kref_put(&((ep)->kref), _c4iw_free_ep); \ +#define c4iw_put_ep(ep) { \ + pr_debug("put_ep (via %s:%u) ep %p refcnt %d\n", \ + __func__, __LINE__, \ + ep, kref_read(&((ep)->kref))); \ + WARN_ON(kref_read(&((ep)->kref)) < 1); \ + kref_put(&((ep)->kref), _c4iw_free_ep); \ } -#define c4iw_get_ep(ep) { \ - PDBG("get_ep (via %s:%u) ep %p, refcnt %d\n", __func__, __LINE__, \ - ep, atomic_read(&((ep)->kref.refcount))); \ - kref_get(&((ep)->kref)); \ +#define c4iw_get_ep(ep) { \ + pr_debug("get_ep (via %s:%u) ep %p, refcnt %d\n", \ + __func__, __LINE__, \ + ep, kref_read(&((ep)->kref))); \ + kref_get(&((ep)->kref)); \ } void _c4iw_free_ep(struct kref *kref); diff --git a/drivers/infiniband/hw/cxgb4/mem.c b/drivers/infiniband/hw/cxgb4/mem.c index 410408f886c1..5332f06b99ba 100644 --- a/drivers/infiniband/hw/cxgb4/mem.c +++ b/drivers/infiniband/hw/cxgb4/mem.c @@ -38,9 +38,9 @@ #include "iw_cxgb4.h" -int use_dsgl = 0; +int use_dsgl = 1; module_param(use_dsgl, int, 0644); -MODULE_PARM_DESC(use_dsgl, "Use DSGL for PBL/FastReg (default=0)"); +MODULE_PARM_DESC(use_dsgl, "Use DSGL for PBL/FastReg (default=1) (DEPRECATED)"); #define T4_ULPTX_MIN_IO 32 #define C4IW_MAX_INLINE_SIZE 96 @@ -81,8 +81,7 @@ static int _c4iw_write_mem_dma_aligned(struct c4iw_rdev *rdev, u32 addr, } set_wr_txq(skb, CPL_PRIORITY_CONTROL, 0); - req = (struct ulp_mem_io *)__skb_put(skb, wr_len); - memset(req, 0, wr_len); + req = __skb_put_zero(skb, wr_len); INIT_ULPTX_WR(req, wr_len, 0, 0); req->wr.wr_hi = cpu_to_be32(FW_WR_OP_V(FW_ULPTX_WR) | (wait ? FW_WR_COMPL_F : 0)); @@ -125,7 +124,7 @@ static int _c4iw_write_mem_inline(struct c4iw_rdev *rdev, u32 addr, u32 len, cmd |= cpu_to_be32(T5_ULP_MEMIO_IMM_F); addr &= 0x7FFFFFF; - PDBG("%s addr 0x%x len %u\n", __func__, addr, len); + pr_debug("%s addr 0x%x len %u\n", __func__, addr, len); num_wqe = DIV_ROUND_UP(len, C4IW_MAX_INLINE_SIZE); c4iw_init_wr_wait(&wr_wait); for (i = 0; i < num_wqe; i++) { @@ -142,8 +141,7 @@ static int _c4iw_write_mem_inline(struct c4iw_rdev *rdev, u32 addr, u32 len, } set_wr_txq(skb, CPL_PRIORITY_CONTROL, 0); - req = (struct ulp_mem_io *)__skb_put(skb, wr_len); - memset(req, 0, wr_len); + req = __skb_put_zero(skb, wr_len); INIT_ULPTX_WR(req, wr_len, 0, 0); if (i == (num_wqe-1)) { @@ -231,13 +229,11 @@ out: static int write_adapter_mem(struct c4iw_rdev *rdev, u32 addr, u32 len, void *data, struct sk_buff *skb) { - if (is_t5(rdev->lldi.adapter_type) && use_dsgl) { + if (rdev->lldi.ulptx_memwrite_dsgl && use_dsgl) { if (len > inline_threshold) { if (_c4iw_write_mem_dma(rdev, addr, len, data, skb)) { - printk_ratelimited(KERN_WARNING - "%s: dma map" - " failure (non fatal)\n", - pci_name(rdev->lldi.pdev)); + pr_warn_ratelimited("%s: dma map failure (non fatal)\n", + pci_name(rdev->lldi.pdev)); return _c4iw_write_mem_inline(rdev, addr, len, data, skb); } else { @@ -289,8 +285,8 @@ static int write_tpt_entry(struct c4iw_rdev *rdev, u32 reset_tpt_entry, mutex_unlock(&rdev->stats.lock); *stag = (stag_idx << 8) | (atomic_inc_return(&key) & 0xff); } - PDBG("%s stag_state 0x%0x type 0x%0x pdid 0x%0x, stag_idx 0x%x\n", - __func__, stag_state, type, pdid, stag_idx); + pr_debug("%s stag_state 0x%0x type 0x%0x pdid 0x%0x, stag_idx 0x%x\n", + __func__, stag_state, type, pdid, stag_idx); /* write TPT entry */ if (reset_tpt_entry) @@ -331,9 +327,9 @@ static int write_pbl(struct c4iw_rdev *rdev, __be64 *pbl, { int err; - PDBG("%s *pdb_addr 0x%x, pbl_base 0x%x, pbl_size %d\n", - __func__, pbl_addr, rdev->lldi.vr->pbl.start, - pbl_size); + pr_debug("%s *pdb_addr 0x%x, pbl_base 0x%x, pbl_size %d\n", + __func__, pbl_addr, rdev->lldi.vr->pbl.start, + pbl_size); err = write_adapter_mem(rdev, pbl_addr >> 5, pbl_size << 3, pbl, NULL); return err; @@ -376,7 +372,7 @@ static int finish_mem_reg(struct c4iw_mr *mhp, u32 stag) mhp->attr.stag = stag; mmid = stag >> 8; mhp->ibmr.rkey = mhp->ibmr.lkey = stag; - PDBG("%s mmid 0x%x mhp %p\n", __func__, mmid, mhp); + pr_debug("%s mmid 0x%x mhp %p\n", __func__, mmid, mhp); return insert_handle(mhp->rhp, &mhp->rhp->mmidr, mhp, mmid); } @@ -426,7 +422,7 @@ struct ib_mr *c4iw_get_dma_mr(struct ib_pd *pd, int acc) int ret; u32 stag = T4_STAG_UNSET; - PDBG("%s ib_pd %p\n", __func__, pd); + pr_debug("%s ib_pd %p\n", __func__, pd); php = to_c4iw_pd(pd); rhp = php->rhp; @@ -483,7 +479,7 @@ struct ib_mr *c4iw_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, struct c4iw_pd *php; struct c4iw_mr *mhp; - PDBG("%s ib_pd %p\n", __func__, pd); + pr_debug("%s ib_pd %p\n", __func__, pd); if (length == ~0ULL) return ERR_PTR(-EINVAL); @@ -517,7 +513,7 @@ struct ib_mr *c4iw_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, return ERR_PTR(err); } - shift = ffs(mhp->umem->page_size) - 1; + shift = mhp->umem->page_shift; n = mhp->umem->nmap; err = alloc_pbl(mhp, n); @@ -536,7 +532,7 @@ struct ib_mr *c4iw_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, len = sg_dma_len(sg) >> shift; for (k = 0; k < len; ++k) { pages[i++] = cpu_to_be64(sg_dma_address(sg) + - mhp->umem->page_size * k); + (k << shift)); if (i == PAGE_SIZE / sizeof *pages) { err = write_pbl(&mhp->rhp->rdev, pages, @@ -620,7 +616,7 @@ struct ib_mw *c4iw_alloc_mw(struct ib_pd *pd, enum ib_mw_type type, ret = -ENOMEM; goto dealloc_win; } - PDBG("%s mmid 0x%x mhp %p stag 0x%x\n", __func__, mmid, mhp, stag); + pr_debug("%s mmid 0x%x mhp %p stag 0x%x\n", __func__, mmid, mhp, stag); return &(mhp->ibmw); dealloc_win: @@ -645,7 +641,7 @@ int c4iw_dealloc_mw(struct ib_mw *mw) deallocate_window(&rhp->rdev, mhp->attr.stag, mhp->dereg_skb); kfree_skb(mhp->dereg_skb); kfree(mhp); - PDBG("%s ib_mw %p mmid 0x%x ptr %p\n", __func__, mw, mmid, mhp); + pr_debug("%s ib_mw %p mmid 0x%x ptr %p\n", __func__, mw, mmid, mhp); return 0; } @@ -703,7 +699,7 @@ struct ib_mr *c4iw_alloc_mr(struct ib_pd *pd, goto err3; } - PDBG("%s mmid 0x%x mhp %p stag 0x%x\n", __func__, mmid, mhp, stag); + pr_debug("%s mmid 0x%x mhp %p stag 0x%x\n", __func__, mmid, mhp, stag); return &(mhp->ibmr); err3: dereg_mem(&rhp->rdev, stag, mhp->attr.pbl_size, @@ -748,7 +744,7 @@ int c4iw_dereg_mr(struct ib_mr *ib_mr) struct c4iw_mr *mhp; u32 mmid; - PDBG("%s ib_mr %p\n", __func__, ib_mr); + pr_debug("%s ib_mr %p\n", __func__, ib_mr); mhp = to_c4iw_mr(ib_mr); rhp = mhp->rhp; @@ -766,7 +762,7 @@ int c4iw_dereg_mr(struct ib_mr *ib_mr) kfree((void *) (unsigned long) mhp->kva); if (mhp->umem) ib_umem_release(mhp->umem); - PDBG("%s mmid 0x%x ptr %p\n", __func__, mmid, mhp); + pr_debug("%s mmid 0x%x ptr %p\n", __func__, mmid, mhp); kfree(mhp); return 0; } diff --git a/drivers/infiniband/hw/cxgb4/provider.c b/drivers/infiniband/hw/cxgb4/provider.c index 3345e1c312f7..0771e9a4d061 100644 --- a/drivers/infiniband/hw/cxgb4/provider.c +++ b/drivers/infiniband/hw/cxgb4/provider.c @@ -59,7 +59,7 @@ module_param(fastreg_support, int, 0644); MODULE_PARM_DESC(fastreg_support, "Advertise fastreg support (default=1)"); static struct ib_ah *c4iw_ah_create(struct ib_pd *pd, - struct ib_ah_attr *ah_attr, + struct rdma_ah_attr *ah_attr, struct ib_udata *udata) { @@ -102,7 +102,7 @@ void _c4iw_free_ucontext(struct kref *kref) ucontext = container_of(kref, struct c4iw_ucontext, kref); rhp = to_c4iw_dev(ucontext->ibucontext.device); - PDBG("%s ucontext %p\n", __func__, ucontext); + pr_debug("%s ucontext %p\n", __func__, ucontext); list_for_each_entry_safe(mm, tmp, &ucontext->mmaps, entry) kfree(mm); c4iw_release_dev_ucontext(&rhp->rdev, &ucontext->uctx); @@ -113,7 +113,7 @@ static int c4iw_dealloc_ucontext(struct ib_ucontext *context) { struct c4iw_ucontext *ucontext = to_c4iw_ucontext(context); - PDBG("%s context %p\n", __func__, context); + pr_debug("%s context %p\n", __func__, context); c4iw_put_ucontext(ucontext); return 0; } @@ -123,12 +123,11 @@ static struct ib_ucontext *c4iw_alloc_ucontext(struct ib_device *ibdev, { struct c4iw_ucontext *context; struct c4iw_dev *rhp = to_c4iw_dev(ibdev); - static int warned; struct c4iw_alloc_ucontext_resp uresp; int ret = 0; struct c4iw_mm_entry *mm = NULL; - PDBG("%s ibdev %p\n", __func__, ibdev); + pr_debug("%s ibdev %p\n", __func__, ibdev); context = kzalloc(sizeof(*context), GFP_KERNEL); if (!context) { ret = -ENOMEM; @@ -141,8 +140,7 @@ static struct ib_ucontext *c4iw_alloc_ucontext(struct ib_device *ibdev, kref_init(&context->kref); if (udata->outlen < sizeof(uresp) - sizeof(uresp.reserved)) { - if (!warned++) - pr_err(MOD "Warning - downlevel libcxgb4 (non-fatal), device status page disabled."); + pr_err_once("Warning - downlevel libcxgb4 (non-fatal), device status page disabled\n"); rhp->rdev.flags |= T4_STATUS_PAGE_DISABLED; } else { mm = kmalloc(sizeof(*mm), GFP_KERNEL); @@ -187,8 +185,8 @@ static int c4iw_mmap(struct ib_ucontext *context, struct vm_area_struct *vma) struct c4iw_ucontext *ucontext; u64 addr; - PDBG("%s pgoff 0x%lx key 0x%x len %d\n", __func__, vma->vm_pgoff, - key, len); + pr_debug("%s pgoff 0x%lx key 0x%x len %d\n", __func__, vma->vm_pgoff, + key, len); if (vma->vm_start & (PAGE_SIZE-1)) return -EINVAL; @@ -253,7 +251,7 @@ static int c4iw_deallocate_pd(struct ib_pd *pd) php = to_c4iw_pd(pd); rhp = php->rhp; - PDBG("%s ibpd %p pdid 0x%x\n", __func__, pd, php->pdid); + pr_debug("%s ibpd %p pdid 0x%x\n", __func__, pd, php->pdid); c4iw_put_resource(&rhp->rdev.resource.pdid_table, php->pdid); mutex_lock(&rhp->rdev.stats.lock); rhp->rdev.stats.pd.cur--; @@ -270,7 +268,7 @@ static struct ib_pd *c4iw_allocate_pd(struct ib_device *ibdev, u32 pdid; struct c4iw_dev *rhp; - PDBG("%s ibdev %p\n", __func__, ibdev); + pr_debug("%s ibdev %p\n", __func__, ibdev); rhp = (struct c4iw_dev *) ibdev; pdid = c4iw_get_resource(&rhp->rdev.resource.pdid_table); if (!pdid) @@ -293,14 +291,14 @@ static struct ib_pd *c4iw_allocate_pd(struct ib_device *ibdev, if (rhp->rdev.stats.pd.cur > rhp->rdev.stats.pd.max) rhp->rdev.stats.pd.max = rhp->rdev.stats.pd.cur; mutex_unlock(&rhp->rdev.stats.lock); - PDBG("%s pdid 0x%0x ptr 0x%p\n", __func__, pdid, php); + pr_debug("%s pdid 0x%0x ptr 0x%p\n", __func__, pdid, php); return &php->ibpd; } static int c4iw_query_pkey(struct ib_device *ibdev, u8 port, u16 index, u16 *pkey) { - PDBG("%s ibdev %p\n", __func__, ibdev); + pr_debug("%s ibdev %p\n", __func__, ibdev); *pkey = 0; return 0; } @@ -310,8 +308,8 @@ static int c4iw_query_gid(struct ib_device *ibdev, u8 port, int index, { struct c4iw_dev *dev; - PDBG("%s ibdev %p, port %d, index %d, gid %p\n", - __func__, ibdev, port, index, gid); + pr_debug("%s ibdev %p, port %d, index %d, gid %p\n", + __func__, ibdev, port, index, gid); dev = to_c4iw_dev(ibdev); BUG_ON(port == 0); memset(&(gid->raw[0]), 0, sizeof(gid->raw)); @@ -325,7 +323,7 @@ static int c4iw_query_device(struct ib_device *ibdev, struct ib_device_attr *pro struct c4iw_dev *dev; - PDBG("%s ibdev %p\n", __func__, ibdev); + pr_debug("%s ibdev %p\n", __func__, ibdev); if (uhw->inlen || uhw->outlen) return -EINVAL; @@ -366,12 +364,11 @@ static int c4iw_query_port(struct ib_device *ibdev, u8 port, struct net_device *netdev; struct in_device *inetdev; - PDBG("%s ibdev %p\n", __func__, ibdev); + pr_debug("%s ibdev %p\n", __func__, ibdev); dev = to_c4iw_dev(ibdev); netdev = dev->rdev.lldi.ports[port-1]; - - memset(props, 0, sizeof(struct ib_port_attr)); + /* props being zeroed by the caller, avoid zeroing it here */ props->max_mtu = IB_MTU_4096; props->active_mtu = ib_mtu_int_to_enum(netdev->mtu); @@ -409,7 +406,7 @@ static ssize_t show_rev(struct device *dev, struct device_attribute *attr, { struct c4iw_dev *c4iw_dev = container_of(dev, struct c4iw_dev, ibdev.dev); - PDBG("%s dev 0x%p\n", __func__, dev); + pr_debug("%s dev 0x%p\n", __func__, dev); return sprintf(buf, "%d\n", CHELSIO_CHIP_RELEASE(c4iw_dev->rdev.lldi.adapter_type)); } @@ -422,7 +419,7 @@ static ssize_t show_hca(struct device *dev, struct device_attribute *attr, struct ethtool_drvinfo info; struct net_device *lldev = c4iw_dev->rdev.lldi.ports[0]; - PDBG("%s dev 0x%p\n", __func__, dev); + pr_debug("%s dev 0x%p\n", __func__, dev); lldev->ethtool_ops->get_drvinfo(lldev, &info); return sprintf(buf, "%s\n", info.driver); } @@ -432,7 +429,7 @@ static ssize_t show_board(struct device *dev, struct device_attribute *attr, { struct c4iw_dev *c4iw_dev = container_of(dev, struct c4iw_dev, ibdev.dev); - PDBG("%s dev 0x%p\n", __func__, dev); + pr_debug("%s dev 0x%p\n", __func__, dev); return sprintf(buf, "%x.%x\n", c4iw_dev->rdev.lldi.pdev->vendor, c4iw_dev->rdev.lldi.pdev->device); } @@ -508,13 +505,14 @@ static int c4iw_port_immutable(struct ib_device *ibdev, u8 port_num, struct ib_port_attr attr; int err; - err = c4iw_query_port(ibdev, port_num, &attr); + immutable->core_cap_flags = RDMA_CORE_PORT_IWARP; + + err = ib_query_port(ibdev, port_num, &attr); if (err) return err; immutable->pkey_tbl_len = attr.pkey_tbl_len; immutable->gid_tbl_len = attr.gid_tbl_len; - immutable->core_cap_flags = RDMA_CORE_PORT_IWARP; return 0; } @@ -524,7 +522,7 @@ static void get_dev_fw_str(struct ib_device *dev, char *str, { struct c4iw_dev *c4iw_dev = container_of(dev, struct c4iw_dev, ibdev); - PDBG("%s dev 0x%p\n", __func__, dev); + pr_debug("%s dev 0x%p\n", __func__, dev); snprintf(str, str_len, "%u.%u.%u.%u", FW_HDR_FW_VER_MAJOR_G(c4iw_dev->rdev.lldi.fw_vers), @@ -538,7 +536,7 @@ int c4iw_register_device(struct c4iw_dev *dev) int ret; int i; - PDBG("%s c4iw_dev %p\n", __func__, dev); + pr_debug("%s c4iw_dev %p\n", __func__, dev); BUG_ON(!dev->rdev.lldi.ports[0]); strlcpy(dev->ibdev.name, "cxgb4_%d", IB_DEVICE_NAME_MAX); memset(&dev->ibdev.node_guid, 0, sizeof(dev->ibdev.node_guid)); @@ -572,7 +570,7 @@ int c4iw_register_device(struct c4iw_dev *dev) memcpy(dev->ibdev.node_desc, C4IW_NODE_DESC, sizeof(C4IW_NODE_DESC)); dev->ibdev.phys_port_cnt = dev->rdev.lldi.nports; dev->ibdev.num_comp_vectors = dev->rdev.lldi.nciq; - dev->ibdev.dma_device = &(dev->rdev.lldi.pdev->dev); + dev->ibdev.dev.parent = &dev->rdev.lldi.pdev->dev; dev->ibdev.query_device = c4iw_query_device; dev->ibdev.query_port = c4iw_query_port; dev->ibdev.query_pkey = c4iw_query_pkey; @@ -648,7 +646,7 @@ void c4iw_unregister_device(struct c4iw_dev *dev) { int i; - PDBG("%s c4iw_dev %p\n", __func__, dev); + pr_debug("%s c4iw_dev %p\n", __func__, dev); for (i = 0; i < ARRAY_SIZE(c4iw_class_attributes); ++i) device_remove_file(&dev->ibdev.dev, c4iw_class_attributes[i]); diff --git a/drivers/infiniband/hw/cxgb4/qp.c b/drivers/infiniband/hw/cxgb4/qp.c index 04c1c382dedb..cb7fc0d35d1d 100644 --- a/drivers/infiniband/hw/cxgb4/qp.c +++ b/drivers/infiniband/hw/cxgb4/qp.c @@ -254,11 +254,11 @@ static int create_qp(struct c4iw_rdev *rdev, struct t4_wq *wq, ret = -ENOMEM; goto free_sq; } - PDBG("%s sq base va 0x%p pa 0x%llx rq base va 0x%p pa 0x%llx\n", - __func__, wq->sq.queue, - (unsigned long long)virt_to_phys(wq->sq.queue), - wq->rq.queue, - (unsigned long long)virt_to_phys(wq->rq.queue)); + pr_debug("%s sq base va 0x%p pa 0x%llx rq base va 0x%p pa 0x%llx\n", + __func__, wq->sq.queue, + (unsigned long long)virt_to_phys(wq->sq.queue), + wq->rq.queue, + (unsigned long long)virt_to_phys(wq->rq.queue)); memset(wq->rq.queue, 0, wq->rq.memsize); dma_unmap_addr_set(&wq->rq, mapping, wq->rq.dma_addr); @@ -275,7 +275,7 @@ static int create_qp(struct c4iw_rdev *rdev, struct t4_wq *wq, * User mode must have bar2 access. */ if (user && (!wq->sq.bar2_pa || !wq->rq.bar2_pa)) { - pr_warn(MOD "%s: sqid %u or rqid %u not in BAR2 range.\n", + pr_warn("%s: sqid %u or rqid %u not in BAR2 range\n", pci_name(rdev->lldi.pdev), wq->sq.qid, wq->rq.qid); goto free_dma; } @@ -293,8 +293,7 @@ static int create_qp(struct c4iw_rdev *rdev, struct t4_wq *wq, } set_wr_txq(skb, CPL_PRIORITY_CONTROL, 0); - res_wr = (struct fw_ri_res_wr *)__skb_put(skb, wr_len); - memset(res_wr, 0, wr_len); + res_wr = __skb_put_zero(skb, wr_len); res_wr->op_nres = cpu_to_be32( FW_WR_OP_V(FW_RI_RES_WR) | FW_RI_RES_WR_NRES_V(2) | @@ -362,9 +361,9 @@ static int create_qp(struct c4iw_rdev *rdev, struct t4_wq *wq, if (ret) goto free_dma; - PDBG("%s sqid 0x%x rqid 0x%x kdb 0x%p sq_bar2_addr %p rq_bar2_addr %p\n", - __func__, wq->sq.qid, wq->rq.qid, wq->db, - wq->sq.bar2_va, wq->rq.bar2_va); + pr_debug("%s sqid 0x%x rqid 0x%x kdb 0x%p sq_bar2_addr %p rq_bar2_addr %p\n", + __func__, wq->sq.qid, wq->rq.qid, wq->db, + wq->sq.bar2_va, wq->rq.bar2_va); return 0; free_dma: @@ -570,7 +569,7 @@ static int build_rdma_read(union t4_wr *wqe, struct ib_send_wr *wr, u8 *len16) { if (wr->num_sge > 1) return -EINVAL; - if (wr->num_sge) { + if (wr->num_sge && wr->sg_list[0].length) { wqe->read.stag_src = cpu_to_be32(rdma_wr(wr)->rkey); wqe->read.to_src_hi = cpu_to_be32((u32)(rdma_wr(wr)->remote_addr >> 32)); @@ -725,7 +724,7 @@ static void free_qp_work(struct work_struct *work) ucontext = qhp->ucontext; rhp = qhp->rhp; - PDBG("%s qhp %p ucontext %p\n", __func__, qhp, ucontext); + pr_debug("%s qhp %p ucontext %p\n", __func__, qhp, ucontext); destroy_qp(&rhp->rdev, &qhp->wq, ucontext ? &ucontext->uctx : &rhp->rdev.uctx); @@ -739,19 +738,19 @@ static void queue_qp_free(struct kref *kref) struct c4iw_qp *qhp; qhp = container_of(kref, struct c4iw_qp, kref); - PDBG("%s qhp %p\n", __func__, qhp); + pr_debug("%s qhp %p\n", __func__, qhp); queue_work(qhp->rhp->rdev.free_workq, &qhp->free_work); } void c4iw_qp_add_ref(struct ib_qp *qp) { - PDBG("%s ib_qp %p\n", __func__, qp); + pr_debug("%s ib_qp %p\n", __func__, qp); kref_get(&to_c4iw_qp(qp)->kref); } void c4iw_qp_rem_ref(struct ib_qp *qp) { - PDBG("%s ib_qp %p\n", __func__, qp); + pr_debug("%s ib_qp %p\n", __func__, qp); kref_put(&to_c4iw_qp(qp)->kref, queue_qp_free); } @@ -959,8 +958,8 @@ int c4iw_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, c4iw_invalidate_mr(qhp->rhp, wr->ex.invalidate_rkey); break; default: - PDBG("%s post of type=%d TBD!\n", __func__, - wr->opcode); + pr_debug("%s post of type=%d TBD!\n", __func__, + wr->opcode); err = -EINVAL; } if (err) { @@ -981,9 +980,10 @@ int c4iw_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, init_wr_hdr(wqe, qhp->wq.sq.pidx, fw_opcode, fw_flags, len16); - PDBG("%s cookie 0x%llx pidx 0x%x opcode 0x%x read_len %u\n", - __func__, (unsigned long long)wr->wr_id, qhp->wq.sq.pidx, - swsqe->opcode, swsqe->read_len); + pr_debug("%s cookie 0x%llx pidx 0x%x opcode 0x%x read_len %u\n", + __func__, + (unsigned long long)wr->wr_id, qhp->wq.sq.pidx, + swsqe->opcode, swsqe->read_len); wr = wr->next; num_wrs--; t4_sq_produce(&qhp->wq, len16); @@ -1057,8 +1057,9 @@ int c4iw_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *wr, wqe->recv.r2[1] = 0; wqe->recv.r2[2] = 0; wqe->recv.len16 = len16; - PDBG("%s cookie 0x%llx pidx %u\n", __func__, - (unsigned long long) wr->wr_id, qhp->wq.rq.pidx); + pr_debug("%s cookie 0x%llx pidx %u\n", + __func__, + (unsigned long long)wr->wr_id, qhp->wq.rq.pidx); t4_rq_produce(&qhp->wq, len16); idx += DIV_ROUND_UP(len16*16, T4_EQ_ENTRY_SIZE); wr = wr->next; @@ -1217,8 +1218,8 @@ static void post_terminate(struct c4iw_qp *qhp, struct t4_cqe *err_cqe, struct sk_buff *skb; struct terminate_message *term; - PDBG("%s qhp %p qid 0x%x tid %u\n", __func__, qhp, qhp->wq.sq.qid, - qhp->ep->hwtid); + pr_debug("%s qhp %p qid 0x%x tid %u\n", __func__, qhp, qhp->wq.sq.qid, + qhp->ep->hwtid); skb = skb_dequeue(&qhp->ep->com.ep_skb_list); if (WARN_ON(!skb)) @@ -1226,7 +1227,7 @@ static void post_terminate(struct c4iw_qp *qhp, struct t4_cqe *err_cqe, set_wr_txq(skb, CPL_PRIORITY_DATA, qhp->ep->txq_idx); - wqe = (struct fw_ri_wr *)__skb_put(skb, sizeof(*wqe)); + wqe = __skb_put(skb, sizeof(*wqe)); memset(wqe, 0, sizeof *wqe); wqe->op_compl = cpu_to_be32(FW_WR_OP_V(FW_RI_INIT_WR)); wqe->flowid_len16 = cpu_to_be32( @@ -1254,7 +1255,7 @@ static void __flush_qp(struct c4iw_qp *qhp, struct c4iw_cq *rchp, int rq_flushed, sq_flushed; unsigned long flag; - PDBG("%s qhp %p rchp %p schp %p\n", __func__, qhp, rchp, schp); + pr_debug("%s qhp %p rchp %p schp %p\n", __func__, qhp, rchp, schp); /* locking hierarchy: cq lock first, then qp lock. */ spin_lock_irqsave(&rchp->lock, flag); @@ -1339,8 +1340,8 @@ static int rdma_fini(struct c4iw_dev *rhp, struct c4iw_qp *qhp, int ret; struct sk_buff *skb; - PDBG("%s qhp %p qid 0x%x tid %u\n", __func__, qhp, qhp->wq.sq.qid, - ep->hwtid); + pr_debug("%s qhp %p qid 0x%x tid %u\n", __func__, qhp, qhp->wq.sq.qid, + ep->hwtid); skb = skb_dequeue(&ep->com.ep_skb_list); if (WARN_ON(!skb)) @@ -1348,7 +1349,7 @@ static int rdma_fini(struct c4iw_dev *rhp, struct c4iw_qp *qhp, set_wr_txq(skb, CPL_PRIORITY_DATA, ep->txq_idx); - wqe = (struct fw_ri_wr *)__skb_put(skb, sizeof(*wqe)); + wqe = __skb_put(skb, sizeof(*wqe)); memset(wqe, 0, sizeof *wqe); wqe->op_compl = cpu_to_be32( FW_WR_OP_V(FW_RI_INIT_WR) | @@ -1366,13 +1367,13 @@ static int rdma_fini(struct c4iw_dev *rhp, struct c4iw_qp *qhp, ret = c4iw_wait_for_reply(&rhp->rdev, &ep->com.wr_wait, qhp->ep->hwtid, qhp->wq.sq.qid, __func__); out: - PDBG("%s ret %d\n", __func__, ret); + pr_debug("%s ret %d\n", __func__, ret); return ret; } static void build_rtr_msg(u8 p2p_type, struct fw_ri_init *init) { - PDBG("%s p2p_type = %d\n", __func__, p2p_type); + pr_debug("%s p2p_type = %d\n", __func__, p2p_type); memset(&init->u, 0, sizeof init->u); switch (p2p_type) { case FW_RI_INIT_P2PTYPE_RDMA_WRITE: @@ -1401,8 +1402,8 @@ static int rdma_init(struct c4iw_dev *rhp, struct c4iw_qp *qhp) int ret; struct sk_buff *skb; - PDBG("%s qhp %p qid 0x%x tid %u ird %u ord %u\n", __func__, qhp, - qhp->wq.sq.qid, qhp->ep->hwtid, qhp->ep->ird, qhp->ep->ord); + pr_debug("%s qhp %p qid 0x%x tid %u ird %u ord %u\n", __func__, qhp, + qhp->wq.sq.qid, qhp->ep->hwtid, qhp->ep->ird, qhp->ep->ord); skb = alloc_skb(sizeof *wqe, GFP_KERNEL); if (!skb) { @@ -1417,7 +1418,7 @@ static int rdma_init(struct c4iw_dev *rhp, struct c4iw_qp *qhp) } set_wr_txq(skb, CPL_PRIORITY_DATA, qhp->ep->txq_idx); - wqe = (struct fw_ri_wr *)__skb_put(skb, sizeof(*wqe)); + wqe = __skb_put(skb, sizeof(*wqe)); memset(wqe, 0, sizeof *wqe); wqe->op_compl = cpu_to_be32( FW_WR_OP_V(FW_RI_INIT_WR) | @@ -1474,7 +1475,7 @@ static int rdma_init(struct c4iw_dev *rhp, struct c4iw_qp *qhp) err1: free_ird(rhp, qhp->attr.max_ird); out: - PDBG("%s ret %d\n", __func__, ret); + pr_debug("%s ret %d\n", __func__, ret); return ret; } @@ -1491,9 +1492,10 @@ int c4iw_modify_qp(struct c4iw_dev *rhp, struct c4iw_qp *qhp, int free = 0; struct c4iw_ep *ep = NULL; - PDBG("%s qhp %p sqid 0x%x rqid 0x%x ep %p state %d -> %d\n", __func__, - qhp, qhp->wq.sq.qid, qhp->wq.rq.qid, qhp->ep, qhp->attr.state, - (mask & C4IW_QP_ATTR_NEXT_STATE) ? attrs->next_state : -1); + pr_debug("%s qhp %p sqid 0x%x rqid 0x%x ep %p state %d -> %d\n", + __func__, + qhp, qhp->wq.sq.qid, qhp->wq.rq.qid, qhp->ep, qhp->attr.state, + (mask & C4IW_QP_ATTR_NEXT_STATE) ? attrs->next_state : -1); mutex_lock(&qhp->mutex); @@ -1580,7 +1582,7 @@ int c4iw_modify_qp(struct c4iw_dev *rhp, struct c4iw_qp *qhp, case C4IW_QP_STATE_RTS: switch (attrs->next_state) { case C4IW_QP_STATE_CLOSING: - BUG_ON(atomic_read(&qhp->ep->com.kref.refcount) < 2); + BUG_ON(kref_read(&qhp->ep->com.kref) < 2); t4_set_wq_in_error(&qhp->wq); set_state(qhp, C4IW_QP_STATE_CLOSING); ep = qhp->ep; @@ -1671,16 +1673,15 @@ int c4iw_modify_qp(struct c4iw_dev *rhp, struct c4iw_qp *qhp, goto err; break; default: - printk(KERN_ERR "%s in a bad state %d\n", - __func__, qhp->attr.state); + pr_err("%s in a bad state %d\n", __func__, qhp->attr.state); ret = -EINVAL; goto err; break; } goto out; err: - PDBG("%s disassociating ep %p qpid 0x%x\n", __func__, qhp->ep, - qhp->wq.sq.qid); + pr_debug("%s disassociating ep %p qpid 0x%x\n", __func__, qhp->ep, + qhp->wq.sq.qid); /* disassociate the LLP connection */ qhp->attr.llp_stream_handle = NULL; @@ -1716,7 +1717,7 @@ out: */ if (free) c4iw_put_ep(&ep->com); - PDBG("%s exit state %d\n", __func__, qhp->attr.state); + pr_debug("%s exit state %d\n", __func__, qhp->attr.state); return ret; } @@ -1746,7 +1747,7 @@ int c4iw_destroy_qp(struct ib_qp *ib_qp) c4iw_qp_rem_ref(ib_qp); - PDBG("%s ib_qp %p qpid 0x%0x\n", __func__, ib_qp, qhp->wq.sq.qid); + pr_debug("%s ib_qp %p qpid 0x%0x\n", __func__, ib_qp, qhp->wq.sq.qid); return 0; } @@ -1765,7 +1766,7 @@ struct ib_qp *c4iw_create_qp(struct ib_pd *pd, struct ib_qp_init_attr *attrs, struct c4iw_mm_entry *sq_key_mm, *rq_key_mm = NULL, *sq_db_key_mm; struct c4iw_mm_entry *rq_db_key_mm = NULL, *ma_sync_key_mm = NULL; - PDBG("%s ib_pd %p\n", __func__, pd); + pr_debug("%s ib_pd %p\n", __func__, pd); if (attrs->qp_type != IB_QPT_RC) return ERR_PTR(-EINVAL); @@ -1936,11 +1937,11 @@ struct ib_qp *c4iw_create_qp(struct ib_pd *pd, struct ib_qp_init_attr *attrs, qhp->ibqp.qp_num = qhp->wq.sq.qid; init_timer(&(qhp->timer)); INIT_LIST_HEAD(&qhp->db_fc_entry); - PDBG("%s sq id %u size %u memsize %zu num_entries %u " - "rq id %u size %u memsize %zu num_entries %u\n", __func__, - qhp->wq.sq.qid, qhp->wq.sq.size, qhp->wq.sq.memsize, - attrs->cap.max_send_wr, qhp->wq.rq.qid, qhp->wq.rq.size, - qhp->wq.rq.memsize, attrs->cap.max_recv_wr); + pr_debug("%s sq id %u size %u memsize %zu num_entries %u rq id %u size %u memsize %zu num_entries %u\n", + __func__, + qhp->wq.sq.qid, qhp->wq.sq.size, qhp->wq.sq.memsize, + attrs->cap.max_send_wr, qhp->wq.rq.qid, qhp->wq.rq.size, + qhp->wq.rq.memsize, attrs->cap.max_recv_wr); return &qhp->ibqp; err8: kfree(ma_sync_key_mm); @@ -1970,7 +1971,7 @@ int c4iw_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, enum c4iw_qp_attr_mask mask = 0; struct c4iw_qp_attributes attrs; - PDBG("%s ib_qp %p\n", __func__, ibqp); + pr_debug("%s ib_qp %p\n", __func__, ibqp); /* iwarp does not support the RTR state */ if ((attr_mask & IB_QP_STATE) && (attr->qp_state == IB_QPS_RTR)) @@ -2016,7 +2017,7 @@ int c4iw_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, struct ib_qp *c4iw_get_qp(struct ib_device *dev, int qpn) { - PDBG("%s ib_dev %p qpn 0x%x\n", __func__, dev, qpn); + pr_debug("%s ib_dev %p qpn 0x%x\n", __func__, dev, qpn); return (struct ib_qp *)get_qhp(to_c4iw_dev(dev), qpn); } diff --git a/drivers/infiniband/hw/cxgb4/resource.c b/drivers/infiniband/hw/cxgb4/resource.c index 67df71a7012e..8ff0cbe5cb16 100644 --- a/drivers/infiniband/hw/cxgb4/resource.c +++ b/drivers/infiniband/hw/cxgb4/resource.c @@ -90,7 +90,7 @@ u32 c4iw_get_resource(struct c4iw_id_table *id_table) void c4iw_put_resource(struct c4iw_id_table *id_table, u32 entry) { - PDBG("%s entry 0x%x\n", __func__, entry); + pr_debug("%s entry 0x%x\n", __func__, entry); c4iw_id_free(id_table, entry); } @@ -141,7 +141,7 @@ u32 c4iw_get_cqid(struct c4iw_rdev *rdev, struct c4iw_dev_ucontext *uctx) } out: mutex_unlock(&uctx->lock); - PDBG("%s qid 0x%x\n", __func__, qid); + pr_debug("%s qid 0x%x\n", __func__, qid); mutex_lock(&rdev->stats.lock); if (rdev->stats.qid.cur > rdev->stats.qid.max) rdev->stats.qid.max = rdev->stats.qid.cur; @@ -157,7 +157,7 @@ void c4iw_put_cqid(struct c4iw_rdev *rdev, u32 qid, entry = kmalloc(sizeof *entry, GFP_KERNEL); if (!entry) return; - PDBG("%s qid 0x%x\n", __func__, qid); + pr_debug("%s qid 0x%x\n", __func__, qid); entry->qid = qid; mutex_lock(&uctx->lock); list_add_tail(&entry->entry, &uctx->cqids); @@ -215,7 +215,7 @@ u32 c4iw_get_qpid(struct c4iw_rdev *rdev, struct c4iw_dev_ucontext *uctx) } out: mutex_unlock(&uctx->lock); - PDBG("%s qid 0x%x\n", __func__, qid); + pr_debug("%s qid 0x%x\n", __func__, qid); mutex_lock(&rdev->stats.lock); if (rdev->stats.qid.cur > rdev->stats.qid.max) rdev->stats.qid.max = rdev->stats.qid.cur; @@ -231,7 +231,7 @@ void c4iw_put_qpid(struct c4iw_rdev *rdev, u32 qid, entry = kmalloc(sizeof *entry, GFP_KERNEL); if (!entry) return; - PDBG("%s qid 0x%x\n", __func__, qid); + pr_debug("%s qid 0x%x\n", __func__, qid); entry->qid = qid; mutex_lock(&uctx->lock); list_add_tail(&entry->entry, &uctx->qpids); @@ -254,7 +254,7 @@ void c4iw_destroy_resource(struct c4iw_resource *rscp) u32 c4iw_pblpool_alloc(struct c4iw_rdev *rdev, int size) { unsigned long addr = gen_pool_alloc(rdev->pbl_pool, size); - PDBG("%s addr 0x%x size %d\n", __func__, (u32)addr, size); + pr_debug("%s addr 0x%x size %d\n", __func__, (u32)addr, size); mutex_lock(&rdev->stats.lock); if (addr) { rdev->stats.pbl.cur += roundup(size, 1 << MIN_PBL_SHIFT); @@ -268,7 +268,7 @@ u32 c4iw_pblpool_alloc(struct c4iw_rdev *rdev, int size) void c4iw_pblpool_free(struct c4iw_rdev *rdev, u32 addr, int size) { - PDBG("%s addr 0x%x size %d\n", __func__, addr, size); + pr_debug("%s addr 0x%x size %d\n", __func__, addr, size); mutex_lock(&rdev->stats.lock); rdev->stats.pbl.cur -= roundup(size, 1 << MIN_PBL_SHIFT); mutex_unlock(&rdev->stats.lock); @@ -290,19 +290,17 @@ int c4iw_pblpool_create(struct c4iw_rdev *rdev) while (pbl_start < pbl_top) { pbl_chunk = min(pbl_top - pbl_start + 1, pbl_chunk); if (gen_pool_add(rdev->pbl_pool, pbl_start, pbl_chunk, -1)) { - PDBG("%s failed to add PBL chunk (%x/%x)\n", - __func__, pbl_start, pbl_chunk); + pr_debug("%s failed to add PBL chunk (%x/%x)\n", + __func__, pbl_start, pbl_chunk); if (pbl_chunk <= 1024 << MIN_PBL_SHIFT) { - printk(KERN_WARNING MOD - "Failed to add all PBL chunks (%x/%x)\n", - pbl_start, - pbl_top - pbl_start); + pr_warn("Failed to add all PBL chunks (%x/%x)\n", + pbl_start, pbl_top - pbl_start); return 0; } pbl_chunk >>= 1; } else { - PDBG("%s added PBL chunk (%x/%x)\n", - __func__, pbl_start, pbl_chunk); + pr_debug("%s added PBL chunk (%x/%x)\n", + __func__, pbl_start, pbl_chunk); pbl_start += pbl_chunk; } } @@ -324,9 +322,9 @@ void c4iw_pblpool_destroy(struct c4iw_rdev *rdev) u32 c4iw_rqtpool_alloc(struct c4iw_rdev *rdev, int size) { unsigned long addr = gen_pool_alloc(rdev->rqt_pool, size << 6); - PDBG("%s addr 0x%x size %d\n", __func__, (u32)addr, size << 6); + pr_debug("%s addr 0x%x size %d\n", __func__, (u32)addr, size << 6); if (!addr) - pr_warn_ratelimited(MOD "%s: Out of RQT memory\n", + pr_warn_ratelimited("%s: Out of RQT memory\n", pci_name(rdev->lldi.pdev)); mutex_lock(&rdev->stats.lock); if (addr) { @@ -341,7 +339,7 @@ u32 c4iw_rqtpool_alloc(struct c4iw_rdev *rdev, int size) void c4iw_rqtpool_free(struct c4iw_rdev *rdev, u32 addr, int size) { - PDBG("%s addr 0x%x size %d\n", __func__, addr, size << 6); + pr_debug("%s addr 0x%x size %d\n", __func__, addr, size << 6); mutex_lock(&rdev->stats.lock); rdev->stats.rqt.cur -= roundup(size << 6, 1 << MIN_RQT_SHIFT); mutex_unlock(&rdev->stats.lock); @@ -363,18 +361,17 @@ int c4iw_rqtpool_create(struct c4iw_rdev *rdev) while (rqt_start < rqt_top) { rqt_chunk = min(rqt_top - rqt_start + 1, rqt_chunk); if (gen_pool_add(rdev->rqt_pool, rqt_start, rqt_chunk, -1)) { - PDBG("%s failed to add RQT chunk (%x/%x)\n", - __func__, rqt_start, rqt_chunk); + pr_debug("%s failed to add RQT chunk (%x/%x)\n", + __func__, rqt_start, rqt_chunk); if (rqt_chunk <= 1024 << MIN_RQT_SHIFT) { - printk(KERN_WARNING MOD - "Failed to add all RQT chunks (%x/%x)\n", - rqt_start, rqt_top - rqt_start); + pr_warn("Failed to add all RQT chunks (%x/%x)\n", + rqt_start, rqt_top - rqt_start); return 0; } rqt_chunk >>= 1; } else { - PDBG("%s added RQT chunk (%x/%x)\n", - __func__, rqt_start, rqt_chunk); + pr_debug("%s added RQT chunk (%x/%x)\n", + __func__, rqt_start, rqt_chunk); rqt_start += rqt_chunk; } } @@ -394,7 +391,7 @@ void c4iw_rqtpool_destroy(struct c4iw_rdev *rdev) u32 c4iw_ocqp_pool_alloc(struct c4iw_rdev *rdev, int size) { unsigned long addr = gen_pool_alloc(rdev->ocqp_pool, size); - PDBG("%s addr 0x%x size %d\n", __func__, (u32)addr, size); + pr_debug("%s addr 0x%x size %d\n", __func__, (u32)addr, size); if (addr) { mutex_lock(&rdev->stats.lock); rdev->stats.ocqp.cur += roundup(size, 1 << MIN_OCQP_SHIFT); @@ -407,7 +404,7 @@ u32 c4iw_ocqp_pool_alloc(struct c4iw_rdev *rdev, int size) void c4iw_ocqp_pool_free(struct c4iw_rdev *rdev, u32 addr, int size) { - PDBG("%s addr 0x%x size %d\n", __func__, addr, size); + pr_debug("%s addr 0x%x size %d\n", __func__, addr, size); mutex_lock(&rdev->stats.lock); rdev->stats.ocqp.cur -= roundup(size, 1 << MIN_OCQP_SHIFT); mutex_unlock(&rdev->stats.lock); @@ -429,18 +426,17 @@ int c4iw_ocqp_pool_create(struct c4iw_rdev *rdev) while (start < top) { chunk = min(top - start + 1, chunk); if (gen_pool_add(rdev->ocqp_pool, start, chunk, -1)) { - PDBG("%s failed to add OCQP chunk (%x/%x)\n", - __func__, start, chunk); + pr_debug("%s failed to add OCQP chunk (%x/%x)\n", + __func__, start, chunk); if (chunk <= 1024 << MIN_OCQP_SHIFT) { - printk(KERN_WARNING MOD - "Failed to add all OCQP chunks (%x/%x)\n", - start, top - start); + pr_warn("Failed to add all OCQP chunks (%x/%x)\n", + start, top - start); return 0; } chunk >>= 1; } else { - PDBG("%s added OCQP chunk (%x/%x)\n", - __func__, start, chunk); + pr_debug("%s added OCQP chunk (%x/%x)\n", + __func__, start, chunk); start += chunk; } } diff --git a/drivers/infiniband/hw/cxgb4/t4.h b/drivers/infiniband/hw/cxgb4/t4.h index 640d22148a3e..e765c00303cd 100644 --- a/drivers/infiniband/hw/cxgb4/t4.h +++ b/drivers/infiniband/hw/cxgb4/t4.h @@ -466,14 +466,14 @@ static inline void t4_ring_sq_db(struct t4_wq *wq, u16 inc, union t4_wr *wqe) wmb(); if (wq->sq.bar2_va) { if (inc == 1 && wq->sq.bar2_qid == 0 && wqe) { - PDBG("%s: WC wq->sq.pidx = %d\n", - __func__, wq->sq.pidx); + pr_debug("%s: WC wq->sq.pidx = %d\n", + __func__, wq->sq.pidx); pio_copy((u64 __iomem *) (wq->sq.bar2_va + SGE_UDB_WCDOORBELL), (u64 *)wqe); } else { - PDBG("%s: DB wq->sq.pidx = %d\n", - __func__, wq->sq.pidx); + pr_debug("%s: DB wq->sq.pidx = %d\n", + __func__, wq->sq.pidx); writel(PIDX_T5_V(inc) | QID_V(wq->sq.bar2_qid), wq->sq.bar2_va + SGE_UDB_KDOORBELL); } @@ -493,14 +493,14 @@ static inline void t4_ring_rq_db(struct t4_wq *wq, u16 inc, wmb(); if (wq->rq.bar2_va) { if (inc == 1 && wq->rq.bar2_qid == 0 && wqe) { - PDBG("%s: WC wq->rq.pidx = %d\n", - __func__, wq->rq.pidx); + pr_debug("%s: WC wq->rq.pidx = %d\n", + __func__, wq->rq.pidx); pio_copy((u64 __iomem *) (wq->rq.bar2_va + SGE_UDB_WCDOORBELL), (void *)wqe); } else { - PDBG("%s: DB wq->rq.pidx = %d\n", - __func__, wq->rq.pidx); + pr_debug("%s: DB wq->rq.pidx = %d\n", + __func__, wq->rq.pidx); writel(PIDX_T5_V(inc) | QID_V(wq->rq.bar2_qid), wq->rq.bar2_va + SGE_UDB_KDOORBELL); } @@ -601,7 +601,8 @@ static inline void t4_swcq_produce(struct t4_cq *cq) { cq->sw_in_use++; if (cq->sw_in_use == cq->size) { - PDBG("%s cxgb4 sw cq overflow cqid %u\n", __func__, cq->cqid); + pr_debug("%s cxgb4 sw cq overflow cqid %u\n", + __func__, cq->cqid); cq->error = 1; BUG_ON(1); } @@ -656,7 +657,7 @@ static inline int t4_next_hw_cqe(struct t4_cq *cq, struct t4_cqe **cqe) if (cq->queue[prev_cidx].bits_type_ts != cq->bits_type_ts) { ret = -EOVERFLOW; cq->error = 1; - printk(KERN_ERR MOD "cq overflow cqid %u\n", cq->cqid); + pr_err("cq overflow cqid %u\n", cq->cqid); BUG_ON(1); } else if (t4_valid_cqe(cq, &cq->queue[cq->cidx])) { @@ -672,7 +673,8 @@ static inline int t4_next_hw_cqe(struct t4_cq *cq, struct t4_cqe **cqe) static inline struct t4_cqe *t4_next_sw_cqe(struct t4_cq *cq) { if (cq->sw_in_use == cq->size) { - PDBG("%s cxgb4 sw cq overflow cqid %u\n", __func__, cq->cqid); + pr_debug("%s cxgb4 sw cq overflow cqid %u\n", + __func__, cq->cqid); cq->error = 1; BUG_ON(1); return NULL; diff --git a/drivers/infiniband/hw/hfi1/Makefile b/drivers/infiniband/hw/hfi1/Makefile index 0cf97a09b64b..88085f65432e 100644 --- a/drivers/infiniband/hw/hfi1/Makefile +++ b/drivers/infiniband/hw/hfi1/Makefile @@ -12,7 +12,7 @@ hfi1-y := affinity.o chip.o device.o driver.o efivar.o \ init.o intr.o mad.o mmu_rb.o pcie.o pio.o pio_copy.o platform.o \ qp.o qsfp.o rc.o ruc.o sdma.o sysfs.o trace.o \ uc.o ud.o user_exp_rcv.o user_pages.o user_sdma.o verbs.o \ - verbs_txreq.o + verbs_txreq.o vnic_main.o vnic_sdma.o hfi1-$(CONFIG_DEBUG_FS) += debugfs.o CFLAGS_trace.o = -I$(src) diff --git a/drivers/infiniband/hw/hfi1/affinity.c b/drivers/infiniband/hw/hfi1/affinity.c index 7a3d906b3671..e2cd2cd3b28a 100644 --- a/drivers/infiniband/hw/hfi1/affinity.c +++ b/drivers/infiniband/hw/hfi1/affinity.c @@ -576,7 +576,7 @@ int hfi1_get_proc_affinity(int node) struct hfi1_affinity_node *entry; cpumask_var_t diff, hw_thread_mask, available_mask, intrs_mask; const struct cpumask *node_mask, - *proc_mask = tsk_cpus_allowed(current); + *proc_mask = ¤t->cpus_allowed; struct hfi1_affinity_node_list *affinity = &node_affinity; struct cpu_mask_set *set = &affinity->proc; diff --git a/drivers/infiniband/hw/hfi1/aspm.h b/drivers/infiniband/hw/hfi1/aspm.h index 0d58fe3b49b5..794e6814a531 100644 --- a/drivers/infiniband/hw/hfi1/aspm.h +++ b/drivers/infiniband/hw/hfi1/aspm.h @@ -1,5 +1,5 @@ /* - * Copyright(c) 2015, 2016 Intel Corporation. + * Copyright(c) 2015-2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -229,14 +229,17 @@ static inline void aspm_ctx_timer_function(unsigned long data) spin_unlock_irqrestore(&rcd->aspm_lock, flags); } -/* Disable interrupt processing for verbs contexts when PSM contexts are open */ +/* + * Disable interrupt processing for verbs contexts when PSM or VNIC contexts + * are open. + */ static inline void aspm_disable_all(struct hfi1_devdata *dd) { struct hfi1_ctxtdata *rcd; unsigned long flags; unsigned i; - for (i = 0; i < dd->first_user_ctxt; i++) { + for (i = 0; i < dd->first_dyn_alloc_ctxt; i++) { rcd = dd->rcd[i]; del_timer_sync(&rcd->aspm_timer); spin_lock_irqsave(&rcd->aspm_lock, flags); @@ -260,7 +263,7 @@ static inline void aspm_enable_all(struct hfi1_devdata *dd) if (aspm_mode != ASPM_MODE_DYNAMIC) return; - for (i = 0; i < dd->first_user_ctxt; i++) { + for (i = 0; i < dd->first_dyn_alloc_ctxt; i++) { rcd = dd->rcd[i]; spin_lock_irqsave(&rcd->aspm_lock, flags); rcd->aspm_intr_enable = true; @@ -276,7 +279,7 @@ static inline void aspm_ctx_init(struct hfi1_ctxtdata *rcd) (unsigned long)rcd); rcd->aspm_intr_supported = rcd->dd->aspm_supported && aspm_mode == ASPM_MODE_DYNAMIC && - rcd->ctxt < rcd->dd->first_user_ctxt; + rcd->ctxt < rcd->dd->first_dyn_alloc_ctxt; } static inline void aspm_init(struct hfi1_devdata *dd) @@ -286,7 +289,7 @@ static inline void aspm_init(struct hfi1_devdata *dd) spin_lock_init(&dd->aspm_lock); dd->aspm_supported = aspm_hw_l1_supported(dd); - for (i = 0; i < dd->first_user_ctxt; i++) + for (i = 0; i < dd->first_dyn_alloc_ctxt; i++) aspm_ctx_init(dd->rcd[i]); /* Start with ASPM disabled */ diff --git a/drivers/infiniband/hw/hfi1/chip.c b/drivers/infiniband/hw/hfi1/chip.c index ef72bc2a9e1d..94b54850ec75 100644 --- a/drivers/infiniband/hw/hfi1/chip.c +++ b/drivers/infiniband/hw/hfi1/chip.c @@ -1,5 +1,5 @@ /* - * Copyright(c) 2015, 2016 Intel Corporation. + * Copyright(c) 2015 - 2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -64,6 +64,7 @@ #include "platform.h" #include "aspm.h" #include "affinity.h" +#include "debugfs.h" #define NUM_IB_PORTS 1 @@ -125,9 +126,16 @@ struct flag_table { #define DEFAULT_KRCVQS 2 #define MIN_KERNEL_KCTXTS 2 #define FIRST_KERNEL_KCTXT 1 -/* sizes for both the QP and RSM map tables */ -#define NUM_MAP_ENTRIES 256 -#define NUM_MAP_REGS 32 + +/* + * RSM instance allocation + * 0 - Verbs + * 1 - User Fecn Handling + * 2 - Vnic + */ +#define RSM_INS_VERBS 0 +#define RSM_INS_FECN 1 +#define RSM_INS_VNIC 2 /* Bit offset into the GUID which carries HFI id information */ #define GUID_HFI_INDEX_SHIFT 39 @@ -138,8 +146,7 @@ struct flag_table { #define is_emulator_p(dd) ((((dd)->irev) & 0xf) == 3) #define is_emulator_s(dd) ((((dd)->irev) & 0xf) == 4) -/* RSM fields */ - +/* RSM fields for Verbs */ /* packet type */ #define IB_PACKET_TYPE 2ull #define QW_SHIFT 6ull @@ -169,6 +176,28 @@ struct flag_table { /* QPN[m+n:1] QW 1, OFFSET 1 */ #define QPN_SELECT_OFFSET ((1ull << QW_SHIFT) | (1ull)) +/* RSM fields for Vnic */ +/* L2_TYPE: QW 0, OFFSET 61 - for match */ +#define L2_TYPE_QW 0ull +#define L2_TYPE_BIT_OFFSET 61ull +#define L2_TYPE_OFFSET(off) ((L2_TYPE_QW << QW_SHIFT) | (off)) +#define L2_TYPE_MATCH_OFFSET L2_TYPE_OFFSET(L2_TYPE_BIT_OFFSET) +#define L2_TYPE_MASK 3ull +#define L2_16B_VALUE 2ull + +/* L4_TYPE QW 1, OFFSET 0 - for match */ +#define L4_TYPE_QW 1ull +#define L4_TYPE_BIT_OFFSET 0ull +#define L4_TYPE_OFFSET(off) ((L4_TYPE_QW << QW_SHIFT) | (off)) +#define L4_TYPE_MATCH_OFFSET L4_TYPE_OFFSET(L4_TYPE_BIT_OFFSET) +#define L4_16B_TYPE_MASK 0xFFull +#define L4_16B_ETH_VALUE 0x78ull + +/* 16B VESWID - for select */ +#define L4_16B_HDR_VESWID_OFFSET ((2 << QW_SHIFT) | (16ull)) +/* 16B ENTROPY - for select */ +#define L2_16B_ENTROPY_OFFSET ((1 << QW_SHIFT) | (32ull)) + /* defines to build power on SC2VL table */ #define SC2VL_VAL( \ num, \ @@ -1026,7 +1055,7 @@ static void handle_pio_err(struct hfi1_devdata *dd, u32 unused, u64 reg); static void handle_sdma_err(struct hfi1_devdata *dd, u32 unused, u64 reg); static void handle_egress_err(struct hfi1_devdata *dd, u32 unused, u64 reg); static void handle_txe_err(struct hfi1_devdata *dd, u32 unused, u64 reg); -static void set_partition_keys(struct hfi1_pportdata *); +static void set_partition_keys(struct hfi1_pportdata *ppd); static const char *link_state_name(u32 state); static const char *link_state_reason_name(struct hfi1_pportdata *ppd, u32 state); @@ -1039,12 +1068,14 @@ static int wait_logical_linkstate(struct hfi1_pportdata *ppd, u32 state, int msecs); static void read_planned_down_reason_code(struct hfi1_devdata *dd, u8 *pdrrc); static void read_link_down_reason(struct hfi1_devdata *dd, u8 *ldr); -static void handle_temp_err(struct hfi1_devdata *); -static void dc_shutdown(struct hfi1_devdata *); -static void dc_start(struct hfi1_devdata *); +static void handle_temp_err(struct hfi1_devdata *dd); +static void dc_shutdown(struct hfi1_devdata *dd); +static void dc_start(struct hfi1_devdata *dd); static int qos_rmt_entries(struct hfi1_devdata *dd, unsigned int *mp, unsigned int *np); static void clear_full_mgmt_pkey(struct hfi1_pportdata *ppd); +static int wait_link_transfer_active(struct hfi1_devdata *dd, int wait_ms); +static void clear_rsm_rule(struct hfi1_devdata *dd, u8 rule_index); /* * Error interrupt table entry. This is used as input to the interrupt @@ -6281,25 +6312,38 @@ static void handle_8051_request(struct hfi1_pportdata *ppd) } } -static void write_global_credit(struct hfi1_devdata *dd, - u8 vau, u16 total, u16 shared) +/* + * Set up allocation unit vaulue. + */ +void set_up_vau(struct hfi1_devdata *dd, u8 vau) { - write_csr(dd, SEND_CM_GLOBAL_CREDIT, - ((u64)total << - SEND_CM_GLOBAL_CREDIT_TOTAL_CREDIT_LIMIT_SHIFT) | - ((u64)shared << - SEND_CM_GLOBAL_CREDIT_SHARED_LIMIT_SHIFT) | - ((u64)vau << SEND_CM_GLOBAL_CREDIT_AU_SHIFT)); + u64 reg = read_csr(dd, SEND_CM_GLOBAL_CREDIT); + + /* do not modify other values in the register */ + reg &= ~SEND_CM_GLOBAL_CREDIT_AU_SMASK; + reg |= (u64)vau << SEND_CM_GLOBAL_CREDIT_AU_SHIFT; + write_csr(dd, SEND_CM_GLOBAL_CREDIT, reg); } /* * Set up initial VL15 credits of the remote. Assumes the rest of - * the CM credit registers are zero from a previous global or credit reset . + * the CM credit registers are zero from a previous global or credit reset. + * Shared limit for VL15 will always be 0. */ -void set_up_vl15(struct hfi1_devdata *dd, u8 vau, u16 vl15buf) +void set_up_vl15(struct hfi1_devdata *dd, u16 vl15buf) { - /* leave shared count at zero for both global and VL15 */ - write_global_credit(dd, vau, vl15buf, 0); + u64 reg = read_csr(dd, SEND_CM_GLOBAL_CREDIT); + + /* set initial values for total and shared credit limit */ + reg &= ~(SEND_CM_GLOBAL_CREDIT_TOTAL_CREDIT_LIMIT_SMASK | + SEND_CM_GLOBAL_CREDIT_SHARED_LIMIT_SMASK); + + /* + * Set total limit to be equal to VL15 credits. + * Leave shared limit at 0. + */ + reg |= (u64)vl15buf << SEND_CM_GLOBAL_CREDIT_TOTAL_CREDIT_LIMIT_SHIFT; + write_csr(dd, SEND_CM_GLOBAL_CREDIT, reg); write_csr(dd, SEND_CM_CREDIT_VL15, (u64)vl15buf << SEND_CM_CREDIT_VL15_DEDICATED_LIMIT_VL_SHIFT); @@ -6317,9 +6361,11 @@ void reset_link_credits(struct hfi1_devdata *dd) for (i = 0; i < TXE_NUM_DATA_VL; i++) write_csr(dd, SEND_CM_CREDIT_VL + (8 * i), 0); write_csr(dd, SEND_CM_CREDIT_VL15, 0); - write_global_credit(dd, 0, 0, 0); + write_csr(dd, SEND_CM_GLOBAL_CREDIT, 0); /* reset the CM block */ pio_send_control(dd, PSC_CM_RESET); + /* reset cached value */ + dd->vl15buf_cached = 0; } /* convert a vCU to a CU */ @@ -6379,18 +6425,17 @@ static void lcb_shutdown(struct hfi1_devdata *dd, int abort) * * The expectation is that the caller of this routine would have taken * care of properly transitioning the link into the correct state. + * NOTE: the caller needs to acquire the dd->dc8051_lock lock + * before calling this function. */ -static void dc_shutdown(struct hfi1_devdata *dd) +static void _dc_shutdown(struct hfi1_devdata *dd) { - unsigned long flags; + lockdep_assert_held(&dd->dc8051_lock); - spin_lock_irqsave(&dd->dc8051_lock, flags); - if (dd->dc_shutdown) { - spin_unlock_irqrestore(&dd->dc8051_lock, flags); + if (dd->dc_shutdown) return; - } + dd->dc_shutdown = 1; - spin_unlock_irqrestore(&dd->dc8051_lock, flags); /* Shutdown the LCB */ lcb_shutdown(dd, 1); /* @@ -6401,35 +6446,45 @@ static void dc_shutdown(struct hfi1_devdata *dd) write_csr(dd, DC_DC8051_CFG_RST, 0x1); } +static void dc_shutdown(struct hfi1_devdata *dd) +{ + mutex_lock(&dd->dc8051_lock); + _dc_shutdown(dd); + mutex_unlock(&dd->dc8051_lock); +} + /* * Calling this after the DC has been brought out of reset should not * do any damage. + * NOTE: the caller needs to acquire the dd->dc8051_lock lock + * before calling this function. */ -static void dc_start(struct hfi1_devdata *dd) +static void _dc_start(struct hfi1_devdata *dd) { - unsigned long flags; - int ret; + lockdep_assert_held(&dd->dc8051_lock); - spin_lock_irqsave(&dd->dc8051_lock, flags); if (!dd->dc_shutdown) - goto done; - spin_unlock_irqrestore(&dd->dc8051_lock, flags); + return; + /* Take the 8051 out of reset */ write_csr(dd, DC_DC8051_CFG_RST, 0ull); /* Wait until 8051 is ready */ - ret = wait_fm_ready(dd, TIMEOUT_8051_START); - if (ret) { + 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); /* lcb_shutdown() with abort=1 does not restore these */ write_csr(dd, DC_LCB_ERR_EN, dd->lcb_err_en); - spin_lock_irqsave(&dd->dc8051_lock, flags); dd->dc_shutdown = 0; -done: - spin_unlock_irqrestore(&dd->dc8051_lock, flags); +} + +static void dc_start(struct hfi1_devdata *dd) +{ + mutex_lock(&dd->dc8051_lock); + _dc_start(dd); + mutex_unlock(&dd->dc8051_lock); } /* @@ -6701,7 +6756,13 @@ static void rxe_kernel_unfreeze(struct hfi1_devdata *dd) int i; /* enable all kernel contexts */ - for (i = 0; i < dd->n_krcv_queues; i++) { + for (i = 0; i < dd->num_rcv_contexts; i++) { + struct hfi1_ctxtdata *rcd = dd->rcd[i]; + + /* Ensure all non-user contexts(including vnic) are enabled */ + if (!rcd || !rcd->sc || (rcd->sc->type == SC_USER)) + continue; + rcvmask = HFI1_RCVCTRL_CTXT_ENB; /* HFI1_RCVCTRL_TAILUPD_[ENB|DIS] needs to be set explicitly */ rcvmask |= HFI1_CAP_KGET_MASK(dd->rcd[i]->flags, DMA_RTAIL) ? @@ -6793,24 +6854,35 @@ void handle_link_up(struct work_struct *work) { struct hfi1_pportdata *ppd = container_of(work, struct hfi1_pportdata, link_up_work); + struct hfi1_devdata *dd = ppd->dd; + set_link_state(ppd, HLS_UP_INIT); /* cache the read of DC_LCB_STS_ROUND_TRIP_LTP_CNT */ - read_ltp_rtt(ppd->dd); + read_ltp_rtt(dd); /* * OPA specifies that certain counters are cleared on a transition * to link up, so do that. */ - clear_linkup_counters(ppd->dd); + clear_linkup_counters(dd); /* * And (re)set link up default values. */ set_linkup_defaults(ppd); + /* + * Set VL15 credits. Use cached value from verify cap interrupt. + * In case of quick linkup or simulator, vl15 value will be set by + * handle_linkup_change. VerifyCap interrupt handler will not be + * called in those scenarios. + */ + if (!(quick_linkup || dd->icode == ICODE_FUNCTIONAL_SIMULATOR)) + set_up_vl15(dd, dd->vl15buf_cached); + /* enforce link speed enabled */ if ((ppd->link_speed_active & ppd->link_speed_enabled) == 0) { /* oops - current speed is not enabled, bounce */ - dd_dev_err(ppd->dd, + dd_dev_err(dd, "Link speed active 0x%x is outside enabled 0x%x, downing link\n", ppd->link_speed_active, ppd->link_speed_enabled); set_link_down_reason(ppd, OPA_LINKDOWN_REASON_SPEED_POLICY, 0, @@ -7077,7 +7149,7 @@ static void add_full_mgmt_pkey(struct hfi1_pportdata *ppd) { struct hfi1_devdata *dd = ppd->dd; - /* Sanity check - ppd->pkeys[2] should be 0, or already initalized */ + /* Sanity check - ppd->pkeys[2] should be 0, or already initialized */ if (!((ppd->pkeys[2] == 0) || (ppd->pkeys[2] == FULL_MGMT_P_KEY))) dd_dev_warn(dd, "%s pkey[2] already set to 0x%x, resetting it to 0x%x\n", __func__, ppd->pkeys[2], FULL_MGMT_P_KEY); @@ -7165,7 +7237,7 @@ static void get_link_widths(struct hfi1_devdata *dd, u16 *tx_width, * set the max_rate field in handle_verify_cap until v0.19. */ if ((dd->icode == ICODE_RTL_SILICON) && - (dd->dc8051_ver < dc8051_ver(0, 19))) { + (dd->dc8051_ver < dc8051_ver(0, 19, 0))) { /* max_rate: 0 = 12.5G, 1 = 25G */ switch (max_rate) { case 0: @@ -7277,15 +7349,6 @@ void handle_verify_cap(struct work_struct *work) lcb_shutdown(dd, 0); adjust_lcb_for_fpga_serdes(dd); - /* - * These are now valid: - * remote VerifyCap fields in the general LNI config - * CSR DC8051_STS_REMOTE_GUID - * CSR DC8051_STS_REMOTE_NODE_TYPE - * CSR DC8051_STS_REMOTE_FM_SECURITY - * CSR DC8051_STS_REMOTE_PORT_NO - */ - read_vc_remote_phy(dd, &power_management, &continious); read_vc_remote_fabric(dd, &vau, &z, &vcu, &vl15buf, &partner_supported_crc); @@ -7320,7 +7383,14 @@ void handle_verify_cap(struct work_struct *work) */ if (vau == 0) vau = 1; - set_up_vl15(dd, vau, vl15buf); + set_up_vau(dd, vau); + + /* + * Set VL15 credits to 0 in global credit register. Cache remote VL15 + * credits value and wait for link-up interrupt ot set it. + */ + set_up_vl15(dd, 0); + dd->vl15buf_cached = vl15buf; /* set up the LCB CRC mode */ crc_mask = ppd->port_crc_mode_enabled & partner_supported_crc; @@ -7350,7 +7420,7 @@ void handle_verify_cap(struct work_struct *work) } ppd->link_speed_active = 0; /* invalid value */ - if (dd->dc8051_ver < dc8051_ver(0, 20)) { + if (dd->dc8051_ver < dc8051_ver(0, 20, 0)) { /* remote_tx_rate: 0 = 12.5G, 1 = 25G */ switch (remote_tx_rate) { case 0: @@ -7416,20 +7486,6 @@ void handle_verify_cap(struct work_struct *work) write_csr(dd, DC_LCB_ERR_EN, 0); /* mask LCB errors */ set_8051_lcb_access(dd); - ppd->neighbor_guid = - read_csr(dd, DC_DC8051_STS_REMOTE_GUID); - ppd->neighbor_port_number = read_csr(dd, DC_DC8051_STS_REMOTE_PORT_NO) & - DC_DC8051_STS_REMOTE_PORT_NO_VAL_SMASK; - ppd->neighbor_type = - read_csr(dd, DC_DC8051_STS_REMOTE_NODE_TYPE) & - DC_DC8051_STS_REMOTE_NODE_TYPE_VAL_MASK; - ppd->neighbor_fm_security = - read_csr(dd, DC_DC8051_STS_REMOTE_FM_SECURITY) & - DC_DC8051_STS_LOCAL_FM_SECURITY_DISABLED_MASK; - dd_dev_info(dd, - "Neighbor Guid: %llx Neighbor type %d MgmtAllowed %d FM security bypass %d\n", - ppd->neighbor_guid, ppd->neighbor_type, - ppd->mgmt_allowed, ppd->neighbor_fm_security); if (ppd->mgmt_allowed) add_full_mgmt_pkey(ppd); @@ -7827,7 +7883,8 @@ static void handle_dcc_err(struct hfi1_devdata *dd, u32 unused, u64 reg) } /* just report this */ - dd_dev_info(dd, "DCC Error: fmconfig error: %s\n", extra); + dd_dev_info_ratelimited(dd, "DCC Error: fmconfig error: %s\n", + extra); reg &= ~DCC_ERR_FLG_FMCONFIG_ERR_SMASK; } @@ -7878,34 +7935,38 @@ static void handle_dcc_err(struct hfi1_devdata *dd, u32 unused, u64 reg) } /* just report this */ - dd_dev_info(dd, "DCC Error: PortRcv error: %s\n", extra); - dd_dev_info(dd, " hdr0 0x%llx, hdr1 0x%llx\n", - hdr0, hdr1); + dd_dev_info_ratelimited(dd, "DCC Error: PortRcv error: %s\n" + " hdr0 0x%llx, hdr1 0x%llx\n", + extra, hdr0, hdr1); reg &= ~DCC_ERR_FLG_RCVPORT_ERR_SMASK; } if (reg & DCC_ERR_FLG_EN_CSR_ACCESS_BLOCKED_UC_SMASK) { /* informative only */ - dd_dev_info(dd, "8051 access to LCB blocked\n"); + dd_dev_info_ratelimited(dd, "8051 access to LCB blocked\n"); reg &= ~DCC_ERR_FLG_EN_CSR_ACCESS_BLOCKED_UC_SMASK; } if (reg & DCC_ERR_FLG_EN_CSR_ACCESS_BLOCKED_HOST_SMASK) { /* informative only */ - dd_dev_info(dd, "host access to LCB blocked\n"); + dd_dev_info_ratelimited(dd, "host access to LCB blocked\n"); reg &= ~DCC_ERR_FLG_EN_CSR_ACCESS_BLOCKED_HOST_SMASK; } + if (unlikely(hfi1_dbg_fault_suppress_err(&dd->verbs_dev))) + reg &= ~DCC_ERR_FLG_LATE_EBP_ERR_SMASK; + /* report any remaining errors */ if (reg) - dd_dev_info(dd, "DCC Error: %s\n", - dcc_err_string(buf, sizeof(buf), reg)); + dd_dev_info_ratelimited(dd, "DCC Error: %s\n", + dcc_err_string(buf, sizeof(buf), reg)); if (lcl_reason == 0) lcl_reason = OPA_LINKDOWN_REASON_UNKNOWN; if (do_bounce) { - dd_dev_info(dd, "%s: PortErrorAction bounce\n", __func__); + dd_dev_info_ratelimited(dd, "%s: PortErrorAction bounce\n", + __func__); set_link_down_reason(ppd, lcl_reason, 0, lcl_reason); queue_work(ppd->hfi1_wq, &ppd->link_bounce_work); } @@ -7993,7 +8054,9 @@ static void is_rcv_avail_int(struct hfi1_devdata *dd, unsigned int source) if (likely(source < dd->num_rcv_contexts)) { rcd = dd->rcd[source]; if (rcd) { - if (source < dd->first_user_ctxt) + /* Check for non-user contexts, including vnic */ + if ((source < dd->first_dyn_alloc_ctxt) || + (rcd->sc && (rcd->sc->type == SC_KERNEL))) rcd->do_interrupt(rcd, 0); else handle_user_interrupt(rcd); @@ -8021,7 +8084,8 @@ static void is_rcv_urgent_int(struct hfi1_devdata *dd, unsigned int source) rcd = dd->rcd[source]; if (rcd) { /* only pay attention to user urgent interrupts */ - if (source >= dd->first_user_ctxt) + if ((source >= dd->first_dyn_alloc_ctxt) && + (!rcd->sc || (rcd->sc->type == SC_USER))) handle_user_interrupt(rcd); return; /* OK */ } @@ -8154,10 +8218,10 @@ static irqreturn_t sdma_interrupt(int irq, void *data) /* handle the interrupt(s) */ sdma_engine_interrupt(sde, status); - } else + } else { dd_dev_err(dd, "SDMA engine %u interrupt, but no status bits set\n", sde->this_idx); - + } return IRQ_HANDLED; } @@ -8342,6 +8406,52 @@ static int read_lcb_via_8051(struct hfi1_devdata *dd, u32 addr, u64 *data) } /* + * Provide a cache for some of the LCB registers in case the LCB is + * unavailable. + * (The LCB is unavailable in certain link states, for example.) + */ +struct lcb_datum { + u32 off; + u64 val; +}; + +static struct lcb_datum lcb_cache[] = { + { DC_LCB_ERR_INFO_RX_REPLAY_CNT, 0}, + { DC_LCB_ERR_INFO_SEQ_CRC_CNT, 0 }, + { DC_LCB_ERR_INFO_REINIT_FROM_PEER_CNT, 0 }, +}; + +static void update_lcb_cache(struct hfi1_devdata *dd) +{ + int i; + int ret; + u64 val; + + for (i = 0; i < ARRAY_SIZE(lcb_cache); i++) { + ret = read_lcb_csr(dd, lcb_cache[i].off, &val); + + /* Update if we get good data */ + if (likely(ret != -EBUSY)) + lcb_cache[i].val = val; + } +} + +static int read_lcb_cache(u32 off, u64 *val) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(lcb_cache); i++) { + if (lcb_cache[i].off == off) { + *val = lcb_cache[i].val; + return 0; + } + } + + pr_warn("%s bad offset 0x%x\n", __func__, off); + return -1; +} + +/* * Read an LCB CSR. Access may not be in host control, so check. * Return 0 on success, -EBUSY on failure. */ @@ -8352,9 +8462,13 @@ int read_lcb_csr(struct hfi1_devdata *dd, u32 addr, u64 *data) /* if up, go through the 8051 for the value */ if (ppd->host_link_state & HLS_UP) return read_lcb_via_8051(dd, addr, data); - /* if going up or down, no access */ - if (ppd->host_link_state & (HLS_GOING_UP | HLS_GOING_OFFLINE)) - return -EBUSY; + /* if going up or down, check the cache, otherwise, no access */ + if (ppd->host_link_state & (HLS_GOING_UP | HLS_GOING_OFFLINE)) { + if (read_lcb_cache(addr, data)) + return -EBUSY; + return 0; + } + /* otherwise, host has access */ *data = read_csr(dd, addr); return 0; @@ -8369,7 +8483,7 @@ static int write_lcb_via_8051(struct hfi1_devdata *dd, u32 addr, u64 data) int ret; if (dd->icode == ICODE_FUNCTIONAL_SIMULATOR || - (dd->dc8051_ver < dc8051_ver(0, 20))) { + (dd->dc8051_ver < dc8051_ver(0, 20, 0))) { if (acquire_lcb_access(dd, 0) == 0) { write_csr(dd, addr, data); release_lcb_access(dd, 0); @@ -8418,16 +8532,11 @@ static int do_8051_command( { u64 reg, completed; int return_code; - unsigned long flags; unsigned long timeout; hfi1_cdbg(DC8051, "type %d, data 0x%012llx", type, in_data); - /* - * Alternative to holding the lock for a long time: - * - keep busy wait - have other users bounce off - */ - spin_lock_irqsave(&dd->dc8051_lock, flags); + mutex_lock(&dd->dc8051_lock); /* We can't send any commands to the 8051 if it's in reset */ if (dd->dc_shutdown) { @@ -8453,10 +8562,8 @@ static int do_8051_command( return_code = -ENXIO; goto fail; } - spin_unlock_irqrestore(&dd->dc8051_lock, flags); - dc_shutdown(dd); - dc_start(dd); - spin_lock_irqsave(&dd->dc8051_lock, flags); + _dc_shutdown(dd); + _dc_start(dd); } /* @@ -8537,8 +8644,7 @@ static int do_8051_command( write_csr(dd, DC_DC8051_CFG_HOST_CMD_0, 0); fail: - spin_unlock_irqrestore(&dd->dc8051_lock, flags); - + mutex_unlock(&dd->dc8051_lock); return return_code; } @@ -8675,13 +8781,20 @@ static void read_remote_device_id(struct hfi1_devdata *dd, u16 *device_id, & REMOTE_DEVICE_REV_MASK; } -void read_misc_status(struct hfi1_devdata *dd, u8 *ver_a, u8 *ver_b) +void read_misc_status(struct hfi1_devdata *dd, u8 *ver_major, u8 *ver_minor, + u8 *ver_patch) { u32 frame; read_8051_config(dd, MISC_STATUS, GENERAL_CONFIG, &frame); - *ver_a = (frame >> STS_FM_VERSION_A_SHIFT) & STS_FM_VERSION_A_MASK; - *ver_b = (frame >> STS_FM_VERSION_B_SHIFT) & STS_FM_VERSION_B_MASK; + *ver_major = (frame >> STS_FM_VERSION_MAJOR_SHIFT) & + STS_FM_VERSION_MAJOR_MASK; + *ver_minor = (frame >> STS_FM_VERSION_MINOR_SHIFT) & + STS_FM_VERSION_MINOR_MASK; + + read_8051_config(dd, VERSION_PATCH, GENERAL_CONFIG, &frame); + *ver_patch = (frame >> STS_FM_VERSION_PATCH_SHIFT) & + STS_FM_VERSION_PATCH_MASK; } static void read_vc_remote_phy(struct hfi1_devdata *dd, u8 *power_management, @@ -8889,8 +9002,6 @@ int send_idle_sma(struct hfi1_devdata *dd, u64 message) */ static int do_quick_linkup(struct hfi1_devdata *dd) { - u64 reg; - unsigned long timeout; int ret; lcb_shutdown(dd, 0); @@ -8913,19 +9024,9 @@ static int do_quick_linkup(struct hfi1_devdata *dd) write_csr(dd, DC_LCB_CFG_RUN, 1ull << DC_LCB_CFG_RUN_EN_SHIFT); - /* watch LCB_STS_LINK_TRANSFER_ACTIVE */ - timeout = jiffies + msecs_to_jiffies(10); - while (1) { - reg = read_csr(dd, DC_LCB_STS_LINK_TRANSFER_ACTIVE); - if (reg) - break; - if (time_after(jiffies, timeout)) { - dd_dev_err(dd, - "timeout waiting for LINK_TRANSFER_ACTIVE\n"); - return -ETIMEDOUT; - } - udelay(2); - } + ret = wait_link_transfer_active(dd, 10); + if (ret) + return ret; write_csr(dd, DC_LCB_CFG_ALLOW_LINK_UP, 1ull << DC_LCB_CFG_ALLOW_LINK_UP_VAL_SHIFT); @@ -9089,7 +9190,7 @@ static int set_local_link_attributes(struct hfi1_pportdata *ppd) if (ret) goto set_local_link_attributes_fail; - if (dd->dc8051_ver < dc8051_ver(0, 20)) { + if (dd->dc8051_ver < dc8051_ver(0, 20, 0)) { /* set the tx rate to the fastest enabled */ if (ppd->link_speed_enabled & OPA_LINK_SPEED_25G) ppd->local_tx_rate = 1; @@ -9272,7 +9373,7 @@ static int handle_qsfp_error_conditions(struct hfi1_pportdata *ppd, if ((qsfp_interrupt_status[0] & QSFP_HIGH_TEMP_ALARM) || (qsfp_interrupt_status[0] & QSFP_HIGH_TEMP_WARNING)) - dd_dev_info(dd, "%s: QSFP cable on fire\n", + dd_dev_info(dd, "%s: QSFP cable temperature too high\n", __func__); if ((qsfp_interrupt_status[0] & QSFP_LOW_TEMP_ALARM) || @@ -9492,8 +9593,11 @@ static int test_qsfp_read(struct hfi1_pportdata *ppd) int ret; u8 status; - /* report success if not a QSFP */ - if (ppd->port_type != PORT_TYPE_QSFP) + /* + * Report success if not a QSFP or, if it is a QSFP, but the cable is + * not present + */ + if (ppd->port_type != PORT_TYPE_QSFP || !qsfp_mod_present(ppd)) return 0; /* read byte 2, the status byte */ @@ -10080,6 +10184,64 @@ static void check_lni_states(struct hfi1_pportdata *ppd) decode_state_complete(ppd, last_remote_state, "received"); } +/* wait for wait_ms for LINK_TRANSFER_ACTIVE to go to 1 */ +static int wait_link_transfer_active(struct hfi1_devdata *dd, int wait_ms) +{ + u64 reg; + unsigned long timeout; + + /* watch LCB_STS_LINK_TRANSFER_ACTIVE */ + timeout = jiffies + msecs_to_jiffies(wait_ms); + while (1) { + reg = read_csr(dd, DC_LCB_STS_LINK_TRANSFER_ACTIVE); + if (reg) + break; + if (time_after(jiffies, timeout)) { + dd_dev_err(dd, + "timeout waiting for LINK_TRANSFER_ACTIVE\n"); + return -ETIMEDOUT; + } + udelay(2); + } + return 0; +} + +/* called when the logical link state is not down as it should be */ +static void force_logical_link_state_down(struct hfi1_pportdata *ppd) +{ + struct hfi1_devdata *dd = ppd->dd; + + /* + * Bring link up in LCB loopback + */ + write_csr(dd, DC_LCB_CFG_TX_FIFOS_RESET, 1); + write_csr(dd, DC_LCB_CFG_IGNORE_LOST_RCLK, + DC_LCB_CFG_IGNORE_LOST_RCLK_EN_SMASK); + + write_csr(dd, DC_LCB_CFG_LANE_WIDTH, 0); + write_csr(dd, DC_LCB_CFG_REINIT_AS_SLAVE, 0); + write_csr(dd, DC_LCB_CFG_CNT_FOR_SKIP_STALL, 0x110); + write_csr(dd, DC_LCB_CFG_LOOPBACK, 0x2); + + write_csr(dd, DC_LCB_CFG_TX_FIFOS_RESET, 0); + (void)read_csr(dd, DC_LCB_CFG_TX_FIFOS_RESET); + udelay(3); + write_csr(dd, DC_LCB_CFG_ALLOW_LINK_UP, 1); + write_csr(dd, DC_LCB_CFG_RUN, 1ull << DC_LCB_CFG_RUN_EN_SHIFT); + + wait_link_transfer_active(dd, 100); + + /* + * Bring the link down again. + */ + write_csr(dd, DC_LCB_CFG_TX_FIFOS_RESET, 1); + write_csr(dd, DC_LCB_CFG_ALLOW_LINK_UP, 0); + write_csr(dd, DC_LCB_CFG_IGNORE_LOST_RCLK, 0); + + /* call again to adjust ppd->statusp, if needed */ + get_logical_state(ppd); +} + /* * Helper for set_link_state(). Do not call except from that routine. * Expects ppd->hls_mutex to be held. @@ -10096,13 +10258,15 @@ static int goto_offline(struct hfi1_pportdata *ppd, u8 rem_reason) int do_transition; int do_wait; + update_lcb_cache(dd); + previous_state = ppd->host_link_state; ppd->host_link_state = HLS_GOING_OFFLINE; pstate = read_physical_state(dd); if (pstate == PLS_OFFLINE) { do_transition = 0; /* in right state */ do_wait = 0; /* ...no need to wait */ - } else if ((pstate & 0xff) == PLS_OFFLINE) { + } else if ((pstate & 0xf0) == PLS_OFFLINE) { do_transition = 0; /* in an offline transient state */ do_wait = 1; /* ...wait for it to settle */ } else { @@ -10133,15 +10297,18 @@ static int goto_offline(struct hfi1_pportdata *ppd, u8 rem_reason) return ret; } - /* make sure the logical state is also down */ - wait_logical_linkstate(ppd, IB_PORT_DOWN, 1000); - /* * Now in charge of LCB - must be after the physical state is * offline.quiet and before host_link_state is changed. */ set_host_lcb_access(dd); write_csr(dd, DC_LCB_ERR_EN, ~0ull); /* watch LCB errors */ + + /* make sure the logical state is also down */ + ret = wait_logical_linkstate(ppd, IB_PORT_DOWN, 1000); + if (ret) + force_logical_link_state_down(ppd); + ppd->host_link_state = HLS_LINK_COOLDOWN; /* LCB access allowed */ if (ppd->port_type == PORT_TYPE_QSFP && @@ -10378,11 +10545,8 @@ int set_link_state(struct hfi1_pportdata *ppd, u32 state) goto unexpected; } - ppd->host_link_state = HLS_UP_INIT; ret = wait_logical_linkstate(ppd, IB_PORT_INIT, 1000); if (ret) { - /* logical state didn't change, stay at going_up */ - ppd->host_link_state = HLS_GOING_UP; dd_dev_err(dd, "%s: logical state did not change to INIT\n", __func__); @@ -10396,6 +10560,7 @@ int set_link_state(struct hfi1_pportdata *ppd, u32 state) add_rcvctrl(dd, RCV_CTRL_RCV_PORT_ENABLE_SMASK); handle_linkup_change(dd, 1); + ppd->host_link_state = HLS_UP_INIT; } break; case HLS_UP_ARMED: @@ -10508,16 +10673,18 @@ int set_link_state(struct hfi1_pportdata *ppd, u32 state) ppd->remote_link_down_reason = 0; } - ret1 = set_physical_link_state(dd, PLS_DISABLED); - if (ret1 != HCMD_SUCCESS) { - dd_dev_err(dd, - "Failed to transition to Disabled link state, return 0x%x\n", - ret1); - ret = -EINVAL; - break; + if (!dd->dc_shutdown) { + ret1 = set_physical_link_state(dd, PLS_DISABLED); + if (ret1 != HCMD_SUCCESS) { + dd_dev_err(dd, + "Failed to transition to Disabled link state, return 0x%x\n", + ret1); + ret = -EINVAL; + break; + } + dc_shutdown(dd); } ppd->host_link_state = HLS_DN_DISABLE; - dc_shutdown(dd); break; case HLS_DN_OFFLINE: if (ppd->host_link_state == HLS_DN_DISABLE) @@ -11849,6 +12016,10 @@ static void free_cntrs(struct hfi1_devdata *dd) dd->scntrs = NULL; kfree(dd->cntrnames); dd->cntrnames = NULL; + if (dd->update_cntr_wq) { + destroy_workqueue(dd->update_cntr_wq); + dd->update_cntr_wq = NULL; + } } static u64 read_dev_port_cntr(struct hfi1_devdata *dd, struct cntr_entry *entry, @@ -12004,7 +12175,7 @@ u64 write_port_cntr(struct hfi1_pportdata *ppd, int index, int vl, u64 data) return write_dev_port_cntr(ppd->dd, entry, sval, ppd, vl, data); } -static void update_synth_timer(unsigned long opaque) +static void do_update_synth_timer(struct work_struct *work) { u64 cur_tx; u64 cur_rx; @@ -12013,8 +12184,8 @@ static void update_synth_timer(unsigned long opaque) int i, j, vl; struct hfi1_pportdata *ppd; struct cntr_entry *entry; - - struct hfi1_devdata *dd = (struct hfi1_devdata *)opaque; + struct hfi1_devdata *dd = container_of(work, struct hfi1_devdata, + update_cntr_work); /* * Rather than keep beating on the CSRs pick a minimal set that we can @@ -12097,7 +12268,13 @@ static void update_synth_timer(unsigned long opaque) } else { hfi1_cdbg(CNTR, "[%d] No update necessary", dd->unit); } +} + +static void update_synth_timer(unsigned long opaque) +{ + struct hfi1_devdata *dd = (struct hfi1_devdata *)opaque; + queue_work(dd->update_cntr_wq, &dd->update_cntr_work); mod_timer(&dd->synth_stats_timer, jiffies + HZ * SYNTH_CNT_TIME); } @@ -12333,6 +12510,13 @@ static int init_cntrs(struct hfi1_devdata *dd) if (init_cpu_counters(dd)) goto bail; + dd->update_cntr_wq = alloc_ordered_workqueue("hfi1_update_cntr_%d", + WQ_MEM_RECLAIM, dd->unit); + if (!dd->update_cntr_wq) + goto bail; + + INIT_WORK(&dd->update_cntr_work, do_update_synth_timer); + mod_timer(&dd->synth_stats_timer, jiffies + HZ * SYNTH_CNT_TIME); return 0; bail: @@ -12511,7 +12695,7 @@ u8 hfi1_ibphys_portstate(struct hfi1_pportdata *ppd) #define SET_STATIC_RATE_CONTROL_SMASK(r) \ (r |= SEND_CTXT_CHECK_ENABLE_DISALLOW_PBC_STATIC_RATE_CONTROL_SMASK) -int hfi1_init_ctxt(struct send_context *sc) +void hfi1_init_ctxt(struct send_context *sc) { if (sc) { struct hfi1_devdata *dd = sc->dd; @@ -12528,7 +12712,6 @@ int hfi1_init_ctxt(struct send_context *sc) write_kctxt_csr(dd, sc->hw_context, SEND_CTXT_CHECK_ENABLE, reg); } - return 0; } int hfi1_tempsense_rd(struct hfi1_devdata *dd, struct hfi1_temp *temp) @@ -12664,7 +12847,12 @@ static void remap_intr(struct hfi1_devdata *dd, int isrc, int msix_intr) /* clear from the handled mask of the general interrupt */ m = isrc / 64; n = isrc % 64; - dd->gi_mask[m] &= ~((u64)1 << n); + if (likely(m < CCE_NUM_INT_CSRS)) { + dd->gi_mask[m] &= ~((u64)1 << n); + } else { + dd_dev_err(dd, "remap interrupt err\n"); + return; + } /* direct the chip source to the given MSI-X interrupt */ m = isrc / 8; @@ -12722,7 +12910,10 @@ static int request_msix_irqs(struct hfi1_devdata *dd) first_sdma = last_general; last_sdma = first_sdma + dd->num_sdma; first_rx = last_sdma; - last_rx = first_rx + dd->n_krcv_queues; + last_rx = first_rx + dd->n_krcv_queues + HFI1_NUM_VNIC_CTXT; + + /* VNIC MSIx interrupts get mapped when VNIC contexts are created */ + dd->first_dyn_msix_idx = first_rx + dd->n_krcv_queues; /* * Sanity check - the code expects all SDMA chip source @@ -12736,7 +12927,7 @@ static int request_msix_irqs(struct hfi1_devdata *dd) const char *err_info; irq_handler_t handler; irq_handler_t thread = NULL; - void *arg; + void *arg = NULL; int idx; struct hfi1_ctxtdata *rcd = NULL; struct sdma_engine *sde = NULL; @@ -12763,24 +12954,25 @@ static int request_msix_irqs(struct hfi1_devdata *dd) } else if (first_rx <= i && i < last_rx) { idx = i - first_rx; rcd = dd->rcd[idx]; - /* no interrupt if no rcd */ - if (!rcd) - continue; - /* - * Set the interrupt register and mask for this - * context's interrupt. - */ - rcd->ireg = (IS_RCVAVAIL_START + idx) / 64; - rcd->imask = ((u64)1) << - ((IS_RCVAVAIL_START + idx) % 64); - handler = receive_context_interrupt; - thread = receive_context_thread; - arg = rcd; - snprintf(me->name, sizeof(me->name), - DRIVER_NAME "_%d kctxt%d", dd->unit, idx); - err_info = "receive context"; - remap_intr(dd, IS_RCVAVAIL_START + idx, i); - me->type = IRQ_RCVCTXT; + if (rcd) { + /* + * Set the interrupt register and mask for this + * context's interrupt. + */ + rcd->ireg = (IS_RCVAVAIL_START + idx) / 64; + rcd->imask = ((u64)1) << + ((IS_RCVAVAIL_START + idx) % 64); + handler = receive_context_interrupt; + thread = receive_context_thread; + arg = rcd; + snprintf(me->name, sizeof(me->name), + DRIVER_NAME "_%d kctxt%d", + dd->unit, idx); + err_info = "receive context"; + remap_intr(dd, IS_RCVAVAIL_START + idx, i); + me->type = IRQ_RCVCTXT; + rcd->msix_intr = i; + } } else { /* not in our expected range - complain, then * ignore it @@ -12818,6 +13010,84 @@ static int request_msix_irqs(struct hfi1_devdata *dd) return ret; } +void hfi1_vnic_synchronize_irq(struct hfi1_devdata *dd) +{ + int i; + + if (!dd->num_msix_entries) { + synchronize_irq(dd->pcidev->irq); + return; + } + + for (i = 0; i < dd->vnic.num_ctxt; i++) { + struct hfi1_ctxtdata *rcd = dd->vnic.ctxt[i]; + struct hfi1_msix_entry *me = &dd->msix_entries[rcd->msix_intr]; + + synchronize_irq(me->msix.vector); + } +} + +void hfi1_reset_vnic_msix_info(struct hfi1_ctxtdata *rcd) +{ + struct hfi1_devdata *dd = rcd->dd; + struct hfi1_msix_entry *me = &dd->msix_entries[rcd->msix_intr]; + + if (!me->arg) /* => no irq, no affinity */ + return; + + hfi1_put_irq_affinity(dd, me); + free_irq(me->msix.vector, me->arg); + + me->arg = NULL; +} + +void hfi1_set_vnic_msix_info(struct hfi1_ctxtdata *rcd) +{ + struct hfi1_devdata *dd = rcd->dd; + struct hfi1_msix_entry *me; + int idx = rcd->ctxt; + void *arg = rcd; + int ret; + + rcd->msix_intr = dd->vnic.msix_idx++; + me = &dd->msix_entries[rcd->msix_intr]; + + /* + * Set the interrupt register and mask for this + * context's interrupt. + */ + rcd->ireg = (IS_RCVAVAIL_START + idx) / 64; + rcd->imask = ((u64)1) << + ((IS_RCVAVAIL_START + idx) % 64); + + snprintf(me->name, sizeof(me->name), + DRIVER_NAME "_%d kctxt%d", dd->unit, idx); + me->name[sizeof(me->name) - 1] = 0; + me->type = IRQ_RCVCTXT; + + remap_intr(dd, IS_RCVAVAIL_START + idx, rcd->msix_intr); + + ret = request_threaded_irq(me->msix.vector, receive_context_interrupt, + receive_context_thread, 0, me->name, arg); + if (ret) { + dd_dev_err(dd, "vnic irq request (vector %d, idx %d) fail %d\n", + me->msix.vector, idx, ret); + return; + } + /* + * assign arg after request_irq call, so it will be + * cleaned up + */ + me->arg = arg; + + ret = hfi1_get_irq_affinity(dd, me); + if (ret) { + dd_dev_err(dd, + "unable to pin IRQ %d\n", ret); + free_irq(me->msix.vector, me->arg); + } +} + /* * Set the general handler to accept all interrupts, remap all * chip interrupts back to MSI-X 0. @@ -12849,7 +13119,7 @@ static int set_up_interrupts(struct hfi1_devdata *dd) * N interrupts - one per used SDMA engine * M interrupt - one per kernel receive context */ - total = 1 + dd->num_sdma + dd->n_krcv_queues; + total = 1 + dd->num_sdma + dd->n_krcv_queues + HFI1_NUM_VNIC_CTXT; entries = kcalloc(total, sizeof(*entries), GFP_KERNEL); if (!entries) { @@ -12914,7 +13184,8 @@ fail: * * num_rcv_contexts - number of contexts being used * n_krcv_queues - number of kernel contexts - * first_user_ctxt - first non-kernel context in array of contexts + * first_dyn_alloc_ctxt - first dynamically allocated context + * in array of contexts * freectxts - number of free user contexts * num_send_contexts - number of PIO send contexts being used */ @@ -12991,10 +13262,14 @@ static int set_up_context_variables(struct hfi1_devdata *dd) total_contexts = num_kernel_contexts + num_user_contexts; } - /* the first N are kernel contexts, the rest are user contexts */ + /* Accommodate VNIC contexts */ + if ((total_contexts + HFI1_NUM_VNIC_CTXT) <= dd->chip_rcv_contexts) + total_contexts += HFI1_NUM_VNIC_CTXT; + + /* the first N are kernel contexts, the rest are user/vnic contexts */ dd->num_rcv_contexts = total_contexts; dd->n_krcv_queues = num_kernel_contexts; - dd->first_user_ctxt = num_kernel_contexts; + dd->first_dyn_alloc_ctxt = num_kernel_contexts; dd->num_user_contexts = num_user_contexts; dd->freectxts = num_user_contexts; dd_dev_info(dd, @@ -13450,11 +13725,8 @@ static void reset_rxe_csrs(struct hfi1_devdata *dd) write_csr(dd, RCV_COUNTER_ARRAY32 + (8 * i), 0); for (i = 0; i < RXE_NUM_64_BIT_COUNTERS; i++) write_csr(dd, RCV_COUNTER_ARRAY64 + (8 * i), 0); - for (i = 0; i < RXE_NUM_RSM_INSTANCES; i++) { - write_csr(dd, RCV_RSM_CFG + (8 * i), 0); - write_csr(dd, RCV_RSM_SELECT + (8 * i), 0); - write_csr(dd, RCV_RSM_MATCH + (8 * i), 0); - } + for (i = 0; i < RXE_NUM_RSM_INSTANCES; i++) + clear_rsm_rule(dd, i); for (i = 0; i < 32; i++) write_csr(dd, RCV_RSM_MAP_TABLE + (8 * i), 0); @@ -13606,14 +13878,14 @@ static void init_chip(struct hfi1_devdata *dd) dd_dev_info(dd, "Resetting CSRs with FLR\n"); /* do the FLR, the DC reset will remain */ - hfi1_pcie_flr(dd); + pcie_flr(dd->pcidev); /* restore command and BARs */ restore_pci_variables(dd); if (is_ax(dd)) { dd_dev_info(dd, "Resetting CSRs with FLR\n"); - hfi1_pcie_flr(dd); + pcie_flr(dd->pcidev); restore_pci_variables(dd); } } else { @@ -13813,6 +14085,16 @@ static void add_rsm_rule(struct hfi1_devdata *dd, u8 rule_index, (u64)rrd->value2 << RCV_RSM_MATCH_VALUE2_SHIFT); } +/* + * Clear a receive side mapping rule. + */ +static void clear_rsm_rule(struct hfi1_devdata *dd, u8 rule_index) +{ + write_csr(dd, RCV_RSM_CFG + (8 * rule_index), 0); + write_csr(dd, RCV_RSM_SELECT + (8 * rule_index), 0); + write_csr(dd, RCV_RSM_MATCH + (8 * rule_index), 0); +} + /* return the number of RSM map table entries that will be used for QOS */ static int qos_rmt_entries(struct hfi1_devdata *dd, unsigned int *mp, unsigned int *np) @@ -13928,7 +14210,7 @@ static void init_qos(struct hfi1_devdata *dd, struct rsm_map_table *rmt) rrd.value2 = LRH_SC_VALUE; /* add rule 0 */ - add_rsm_rule(dd, 0, &rrd); + add_rsm_rule(dd, RSM_INS_VERBS, &rrd); /* mark RSM map entries as used */ rmt->used += rmt_entries; @@ -13958,7 +14240,7 @@ static void init_user_fecn_handling(struct hfi1_devdata *dd, /* * RSM will extract the destination context as an index into the * map table. The destination contexts are a sequential block - * in the range first_user_ctxt...num_rcv_contexts-1 (inclusive). + * in the range first_dyn_alloc_ctxt...num_rcv_contexts-1 (inclusive). * Map entries are accessed as offset + extracted value. Adjust * the added offset so this sequence can be placed anywhere in * the table - as long as the entries themselves do not wrap. @@ -13966,9 +14248,9 @@ static void init_user_fecn_handling(struct hfi1_devdata *dd, * start with that to allow for a "negative" offset. */ offset = (u8)(NUM_MAP_ENTRIES + (int)rmt->used - - (int)dd->first_user_ctxt); + (int)dd->first_dyn_alloc_ctxt); - for (i = dd->first_user_ctxt, idx = rmt->used; + for (i = dd->first_dyn_alloc_ctxt, idx = rmt->used; i < dd->num_rcv_contexts; i++, idx++) { /* replace with identity mapping */ regoff = (idx % 8) * 8; @@ -14002,11 +14284,84 @@ static void init_user_fecn_handling(struct hfi1_devdata *dd, rrd.value2 = 1; /* add rule 1 */ - add_rsm_rule(dd, 1, &rrd); + add_rsm_rule(dd, RSM_INS_FECN, &rrd); rmt->used += dd->num_user_contexts; } +/* Initialize RSM for VNIC */ +void hfi1_init_vnic_rsm(struct hfi1_devdata *dd) +{ + u8 i, j; + u8 ctx_id = 0; + u64 reg; + u32 regoff; + struct rsm_rule_data rrd; + + if (hfi1_vnic_is_rsm_full(dd, NUM_VNIC_MAP_ENTRIES)) { + dd_dev_err(dd, "Vnic RSM disabled, rmt entries used = %d\n", + dd->vnic.rmt_start); + return; + } + + dev_dbg(&(dd)->pcidev->dev, "Vnic rsm start = %d, end %d\n", + dd->vnic.rmt_start, + dd->vnic.rmt_start + NUM_VNIC_MAP_ENTRIES); + + /* Update RSM mapping table, 32 regs, 256 entries - 1 ctx per byte */ + regoff = RCV_RSM_MAP_TABLE + (dd->vnic.rmt_start / 8) * 8; + reg = read_csr(dd, regoff); + for (i = 0; i < NUM_VNIC_MAP_ENTRIES; i++) { + /* Update map register with vnic context */ + j = (dd->vnic.rmt_start + i) % 8; + reg &= ~(0xffllu << (j * 8)); + reg |= (u64)dd->vnic.ctxt[ctx_id++]->ctxt << (j * 8); + /* Wrap up vnic ctx index */ + ctx_id %= dd->vnic.num_ctxt; + /* Write back map register */ + if (j == 7 || ((i + 1) == NUM_VNIC_MAP_ENTRIES)) { + dev_dbg(&(dd)->pcidev->dev, + "Vnic rsm map reg[%d] =0x%llx\n", + regoff - RCV_RSM_MAP_TABLE, reg); + + write_csr(dd, regoff, reg); + regoff += 8; + if (i < (NUM_VNIC_MAP_ENTRIES - 1)) + reg = read_csr(dd, regoff); + } + } + + /* Add rule for vnic */ + rrd.offset = dd->vnic.rmt_start; + rrd.pkt_type = 4; + /* Match 16B packets */ + rrd.field1_off = L2_TYPE_MATCH_OFFSET; + rrd.mask1 = L2_TYPE_MASK; + rrd.value1 = L2_16B_VALUE; + /* Match ETH L4 packets */ + rrd.field2_off = L4_TYPE_MATCH_OFFSET; + rrd.mask2 = L4_16B_TYPE_MASK; + rrd.value2 = L4_16B_ETH_VALUE; + /* Calc context from veswid and entropy */ + rrd.index1_off = L4_16B_HDR_VESWID_OFFSET; + rrd.index1_width = ilog2(NUM_VNIC_MAP_ENTRIES); + rrd.index2_off = L2_16B_ENTROPY_OFFSET; + rrd.index2_width = ilog2(NUM_VNIC_MAP_ENTRIES); + add_rsm_rule(dd, RSM_INS_VNIC, &rrd); + + /* Enable RSM if not already enabled */ + add_rcvctrl(dd, RCV_CTRL_RCV_RSM_ENABLE_SMASK); +} + +void hfi1_deinit_vnic_rsm(struct hfi1_devdata *dd) +{ + clear_rsm_rule(dd, RSM_INS_VNIC); + + /* Disable RSM if used only by vnic */ + if (dd->vnic.rmt_start == 0) + clear_rcvctrl(dd, RCV_CTRL_RCV_RSM_ENABLE_SMASK); +} + static void init_rxe(struct hfi1_devdata *dd) { struct rsm_map_table *rmt; @@ -14019,6 +14374,8 @@ static void init_rxe(struct hfi1_devdata *dd) init_qos(dd, rmt); init_user_fecn_handling(dd, rmt); complete_rsm_map_table(dd, rmt); + /* record number of used rsm map entries for vnic */ + dd->vnic.rmt_start = rmt->used; kfree(rmt); /* @@ -14208,30 +14565,24 @@ done: return ret; } -int hfi1_clear_ctxt_pkey(struct hfi1_devdata *dd, unsigned ctxt) +int hfi1_clear_ctxt_pkey(struct hfi1_devdata *dd, struct hfi1_ctxtdata *ctxt) { - struct hfi1_ctxtdata *rcd; - unsigned sctxt; - int ret = 0; + u8 hw_ctxt; u64 reg; - if (ctxt < dd->num_rcv_contexts) { - rcd = dd->rcd[ctxt]; - } else { - ret = -EINVAL; - goto done; - } - if (!rcd || !rcd->sc) { - ret = -EINVAL; - goto done; - } - sctxt = rcd->sc->hw_context; - reg = read_kctxt_csr(dd, sctxt, SEND_CTXT_CHECK_ENABLE); + if (!ctxt || !ctxt->sc) + return -EINVAL; + + if (ctxt->ctxt >= dd->num_rcv_contexts) + return -EINVAL; + + hw_ctxt = ctxt->sc->hw_context; + reg = read_kctxt_csr(dd, hw_ctxt, SEND_CTXT_CHECK_ENABLE); reg &= ~SEND_CTXT_CHECK_ENABLE_CHECK_PARTITION_KEY_SMASK; - write_kctxt_csr(dd, sctxt, SEND_CTXT_CHECK_ENABLE, reg); - write_kctxt_csr(dd, sctxt, SEND_CTXT_CHECK_PARTITION_KEY, 0); -done: - return ret; + write_kctxt_csr(dd, hw_ctxt, SEND_CTXT_CHECK_ENABLE, reg); + write_kctxt_csr(dd, hw_ctxt, SEND_CTXT_CHECK_PARTITION_KEY, 0); + + return 0; } /* diff --git a/drivers/infiniband/hw/hfi1/chip.h b/drivers/infiniband/hw/hfi1/chip.h index 043fd21dc5f3..cbe455d9ab8b 100644 --- a/drivers/infiniband/hw/hfi1/chip.h +++ b/drivers/infiniband/hw/hfi1/chip.h @@ -1,7 +1,7 @@ #ifndef _CHIP_H #define _CHIP_H /* - * Copyright(c) 2015, 2016 Intel Corporation. + * Copyright(c) 2015 - 2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -394,7 +394,8 @@ #define LAST_REMOTE_STATE_COMPLETE 0x13 #define LINK_QUALITY_INFO 0x14 #define REMOTE_DEVICE_ID 0x15 -#define LINK_DOWN_REASON 0x16 +#define LINK_DOWN_REASON 0x16 /* first byte of offset 0x16 */ +#define VERSION_PATCH 0x16 /* last byte of offset 0x16 */ /* 8051 lane specific register field IDs */ #define TX_EQ_SETTINGS 0x00 @@ -524,10 +525,12 @@ enum { #define SUPPORTED_CRCS (CAP_CRC_14B | CAP_CRC_48B) /* misc status version fields */ -#define STS_FM_VERSION_A_SHIFT 16 -#define STS_FM_VERSION_A_MASK 0xff -#define STS_FM_VERSION_B_SHIFT 24 -#define STS_FM_VERSION_B_MASK 0xff +#define STS_FM_VERSION_MINOR_SHIFT 16 +#define STS_FM_VERSION_MINOR_MASK 0xff +#define STS_FM_VERSION_MAJOR_SHIFT 24 +#define STS_FM_VERSION_MAJOR_MASK 0xff +#define STS_FM_VERSION_PATCH_SHIFT 24 +#define STS_FM_VERSION_PATCH_MASK 0xff /* LCB_CFG_CRC_MODE TX_VAL and RX_VAL CRC mode values */ #define LCB_CRC_16B 0x0 /* 16b CRC */ @@ -633,7 +636,8 @@ static inline void write_uctxt_csr(struct hfi1_devdata *dd, int ctxt, write_csr(dd, offset0 + (0x1000 * ctxt), value); } -u64 create_pbc(struct hfi1_pportdata *ppd, u64, int, u32, u32); +u64 create_pbc(struct hfi1_pportdata *ppd, u64 flags, int srate_mbs, u32 vl, + u32 dw_len); /* firmware.c */ #define SBUS_MASTER_BROADCAST 0xfd @@ -698,7 +702,8 @@ void fabric_serdes_reset(struct hfi1_devdata *dd); int read_8051_data(struct hfi1_devdata *dd, u32 addr, u32 len, u64 *result); /* chip.c */ -void read_misc_status(struct hfi1_devdata *dd, u8 *ver_a, u8 *ver_b); +void read_misc_status(struct hfi1_devdata *dd, u8 *ver_major, u8 *ver_minor, + u8 *ver_patch); void read_guid(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, @@ -724,7 +729,8 @@ int bringup_serdes(struct hfi1_pportdata *ppd); void set_intr_state(struct hfi1_devdata *dd, u32 enable); void apply_link_downgrade_policy(struct hfi1_pportdata *ppd, int refresh_widths); -void update_usrhead(struct hfi1_ctxtdata *, u32, u32, u32, u32, u32); +void update_usrhead(struct hfi1_ctxtdata *rcd, u32 hd, u32 updegr, u32 egrhd, + u32 intr_adjust, u32 npkts); int stop_drain_data_vls(struct hfi1_devdata *dd); int open_fill_data_vls(struct hfi1_devdata *dd); u32 ns_to_cclock(struct hfi1_devdata *dd, u32 ns); @@ -1343,7 +1349,7 @@ void hfi1_start_cleanup(struct hfi1_devdata *dd); void hfi1_clear_tids(struct hfi1_ctxtdata *rcd); struct ib_header *hfi1_get_msgheader( struct hfi1_devdata *dd, __le32 *rhf_addr); -int hfi1_init_ctxt(struct send_context *sc); +void hfi1_init_ctxt(struct send_context *sc); void hfi1_put_tid(struct hfi1_devdata *dd, u32 index, u32 type, unsigned long pa, u16 order); void hfi1_quiet_serdes(struct hfi1_pportdata *ppd); @@ -1356,8 +1362,10 @@ int hfi1_set_ib_cfg(struct hfi1_pportdata *ppd, int which, u32 val); int hfi1_set_ctxt_jkey(struct hfi1_devdata *dd, unsigned ctxt, u16 jkey); int hfi1_clear_ctxt_jkey(struct hfi1_devdata *dd, unsigned ctxt); int hfi1_set_ctxt_pkey(struct hfi1_devdata *dd, unsigned ctxt, u16 pkey); -int hfi1_clear_ctxt_pkey(struct hfi1_devdata *dd, unsigned ctxt); +int hfi1_clear_ctxt_pkey(struct hfi1_devdata *dd, struct hfi1_ctxtdata *ctxt); void hfi1_read_link_quality(struct hfi1_devdata *dd, u8 *link_quality); +void hfi1_init_vnic_rsm(struct hfi1_devdata *dd); +void hfi1_deinit_vnic_rsm(struct hfi1_devdata *dd); /* * Interrupt source table. diff --git a/drivers/infiniband/hw/hfi1/chip_registers.h b/drivers/infiniband/hw/hfi1/chip_registers.h index 5bfa839d1c48..793514f1d15f 100644 --- a/drivers/infiniband/hw/hfi1/chip_registers.h +++ b/drivers/infiniband/hw/hfi1/chip_registers.h @@ -839,7 +839,9 @@ #define SEND_CM_CTRL_FORCE_CREDIT_MODE_SMASK 0x8ull #define SEND_CM_CTRL_RESETCSR 0x0000000000000020ull #define SEND_CM_GLOBAL_CREDIT (TXE + 0x000000000508) +#define SEND_CM_GLOBAL_CREDIT_AU_MASK 0x7ull #define SEND_CM_GLOBAL_CREDIT_AU_SHIFT 16 +#define SEND_CM_GLOBAL_CREDIT_AU_SMASK 0x70000ull #define SEND_CM_GLOBAL_CREDIT_RESETCSR 0x0000094000030000ull #define SEND_CM_GLOBAL_CREDIT_SHARED_LIMIT_MASK 0xFFFFull #define SEND_CM_GLOBAL_CREDIT_SHARED_LIMIT_SHIFT 0 diff --git a/drivers/infiniband/hw/hfi1/common.h b/drivers/infiniband/hw/hfi1/common.h index da7be21bedb4..995d62c7f9a7 100644 --- a/drivers/infiniband/hw/hfi1/common.h +++ b/drivers/infiniband/hw/hfi1/common.h @@ -331,16 +331,15 @@ struct diag_pkt { #define FULL_MGMT_P_KEY 0xFFFF #define DEFAULT_P_KEY LIM_MGMT_P_KEY -#define HFI1_AETH_CREDIT_SHIFT 24 -#define HFI1_AETH_CREDIT_MASK 0x1F -#define HFI1_AETH_CREDIT_INVAL 0x1F -#define HFI1_MSN_MASK 0xFFFFFF -#define HFI1_FECN_SHIFT 31 -#define HFI1_FECN_MASK 1 -#define HFI1_FECN_SMASK BIT(HFI1_FECN_SHIFT) -#define HFI1_BECN_SHIFT 30 -#define HFI1_BECN_MASK 1 -#define HFI1_BECN_SMASK BIT(HFI1_BECN_SHIFT) + +/** + * 0xF8 - 4 bits of multicast range and 1 bit for collective range + * Example: For 24 bit LID space, + * Multicast range: 0xF00000 to 0xF7FFFF + * Collective range: 0xF80000 to 0xFFFFFE + */ +#define HFI1_MCAST_NR 0x4 /* Number of top bits set */ +#define HFI1_COLLECTIVE_NR 0x1 /* Number of bits after MCAST_NR */ #define HFI1_PSM_IOC_BASE_SEQ 0x0 diff --git a/drivers/infiniband/hw/hfi1/debugfs.c b/drivers/infiniband/hw/hfi1/debugfs.c index 8725f4c086cf..e9fa3c293e42 100644 --- a/drivers/infiniband/hw/hfi1/debugfs.c +++ b/drivers/infiniband/hw/hfi1/debugfs.c @@ -1,6 +1,6 @@ #ifdef CONFIG_DEBUG_FS /* - * Copyright(c) 2015, 2016 Intel Corporation. + * Copyright(c) 2015-2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -50,8 +50,13 @@ #include <linux/kernel.h> #include <linux/export.h> #include <linux/module.h> +#include <linux/string.h> +#include <linux/types.h> +#include <linux/ratelimit.h> +#include <linux/fault-inject.h> #include "hfi.h" +#include "trace.h" #include "debugfs.h" #include "device.h" #include "qp.h" @@ -169,7 +174,7 @@ static int _opcode_stats_seq_show(struct seq_file *s, void *v) struct hfi1_ibdev *ibd = (struct hfi1_ibdev *)s->private; struct hfi1_devdata *dd = dd_from_dev(ibd); - for (j = 0; j < dd->first_user_ctxt; j++) { + for (j = 0; j < dd->first_dyn_alloc_ctxt; j++) { if (!dd->rcd[j]) continue; n_packets += dd->rcd[j]->opstats->stats[i].n_packets; @@ -195,7 +200,7 @@ static void *_ctx_stats_seq_start(struct seq_file *s, loff_t *pos) if (!*pos) return SEQ_START_TOKEN; - if (*pos >= dd->first_user_ctxt) + if (*pos >= dd->first_dyn_alloc_ctxt) return NULL; return pos; } @@ -209,7 +214,7 @@ static void *_ctx_stats_seq_next(struct seq_file *s, void *v, loff_t *pos) return pos; ++*pos; - if (*pos >= dd->first_user_ctxt) + if (*pos >= dd->first_dyn_alloc_ctxt) return NULL; return pos; } @@ -503,18 +508,11 @@ static ssize_t asic_flags_write(struct file *file, const char __user *buf, ppd = private2ppd(file); dd = ppd->dd; - buff = kmalloc(count + 1, GFP_KERNEL); - if (!buff) - return -ENOMEM; - - ret = copy_from_user(buff, buf, count); - if (ret > 0) { - ret = -EFAULT; - goto do_free; - } - /* zero terminate and read the expected integer */ - buff[count] = 0; + buff = memdup_user_nul(buf, count); + if (IS_ERR(buff)) + return PTR_ERR(buff); + ret = kstrtoull(buff, 0, &value); if (ret) goto do_free; @@ -692,15 +690,9 @@ static ssize_t __i2c_debugfs_write(struct file *file, const char __user *buf, if (i2c_addr == 0) return -EINVAL; - buff = kmalloc(count, GFP_KERNEL); - if (!buff) - return -ENOMEM; - - ret = copy_from_user(buff, buf, count); - if (ret > 0) { - ret = -EFAULT; - goto _free; - } + buff = memdup_user(buf, count); + if (IS_ERR(buff)) + return PTR_ERR(buff); total_written = i2c_write(ppd, target, i2c_addr, offset, buff, count); if (total_written < 0) { @@ -805,15 +797,10 @@ static ssize_t __qsfp_debugfs_write(struct file *file, const char __user *buf, ppd = private2ppd(file); - buff = kmalloc(count, GFP_KERNEL); - if (!buff) - return -ENOMEM; + buff = memdup_user(buf, count); + if (IS_ERR(buff)) + return PTR_ERR(buff); - ret = copy_from_user(buff, buf, count); - if (ret > 0) { - ret = -EFAULT; - goto _free; - } total_written = qsfp_write(ppd, target, *ppos, buff, count); if (total_written < 0) { ret = total_written; @@ -1080,6 +1067,222 @@ DEBUGFS_SEQ_FILE_OPS(sdma_cpu_list); DEBUGFS_SEQ_FILE_OPEN(sdma_cpu_list) DEBUGFS_FILE_OPS(sdma_cpu_list); +#ifdef CONFIG_FAULT_INJECTION +static void *_fault_stats_seq_start(struct seq_file *s, loff_t *pos) +{ + struct hfi1_opcode_stats_perctx *opstats; + + if (*pos >= ARRAY_SIZE(opstats->stats)) + return NULL; + return pos; +} + +static void *_fault_stats_seq_next(struct seq_file *s, void *v, loff_t *pos) +{ + struct hfi1_opcode_stats_perctx *opstats; + + ++*pos; + if (*pos >= ARRAY_SIZE(opstats->stats)) + return NULL; + return pos; +} + +static void _fault_stats_seq_stop(struct seq_file *s, void *v) +{ +} + +static int _fault_stats_seq_show(struct seq_file *s, void *v) +{ + loff_t *spos = v; + loff_t i = *spos, j; + u64 n_packets = 0, n_bytes = 0; + struct hfi1_ibdev *ibd = (struct hfi1_ibdev *)s->private; + struct hfi1_devdata *dd = dd_from_dev(ibd); + + for (j = 0; j < dd->first_dyn_alloc_ctxt; j++) { + if (!dd->rcd[j]) + continue; + n_packets += dd->rcd[j]->opstats->stats[i].n_packets; + n_bytes += dd->rcd[j]->opstats->stats[i].n_bytes; + } + if (!n_packets && !n_bytes) + return SEQ_SKIP; + if (!ibd->fault_opcode->n_rxfaults[i] && + !ibd->fault_opcode->n_txfaults[i]) + return SEQ_SKIP; + seq_printf(s, "%02llx %llu/%llu (faults rx:%llu faults: tx:%llu)\n", i, + (unsigned long long)n_packets, + (unsigned long long)n_bytes, + (unsigned long long)ibd->fault_opcode->n_rxfaults[i], + (unsigned long long)ibd->fault_opcode->n_txfaults[i]); + return 0; +} + +DEBUGFS_SEQ_FILE_OPS(fault_stats); +DEBUGFS_SEQ_FILE_OPEN(fault_stats); +DEBUGFS_FILE_OPS(fault_stats); + +static void fault_exit_opcode_debugfs(struct hfi1_ibdev *ibd) +{ + debugfs_remove_recursive(ibd->fault_opcode->dir); + kfree(ibd->fault_opcode); + ibd->fault_opcode = NULL; +} + +static int fault_init_opcode_debugfs(struct hfi1_ibdev *ibd) +{ + struct dentry *parent = ibd->hfi1_ibdev_dbg; + + ibd->fault_opcode = kzalloc(sizeof(*ibd->fault_opcode), GFP_KERNEL); + if (!ibd->fault_opcode) + return -ENOMEM; + + ibd->fault_opcode->attr.interval = 1; + ibd->fault_opcode->attr.require_end = ULONG_MAX; + ibd->fault_opcode->attr.stacktrace_depth = 32; + ibd->fault_opcode->attr.dname = NULL; + ibd->fault_opcode->attr.verbose = 0; + ibd->fault_opcode->fault_by_opcode = false; + ibd->fault_opcode->opcode = 0; + ibd->fault_opcode->mask = 0xff; + + ibd->fault_opcode->dir = + fault_create_debugfs_attr("fault_opcode", + parent, + &ibd->fault_opcode->attr); + if (IS_ERR(ibd->fault_opcode->dir)) { + kfree(ibd->fault_opcode); + return -ENOENT; + } + + DEBUGFS_SEQ_FILE_CREATE(fault_stats, ibd->fault_opcode->dir, ibd); + if (!debugfs_create_bool("fault_by_opcode", 0600, + ibd->fault_opcode->dir, + &ibd->fault_opcode->fault_by_opcode)) + goto fail; + if (!debugfs_create_x8("opcode", 0600, ibd->fault_opcode->dir, + &ibd->fault_opcode->opcode)) + goto fail; + if (!debugfs_create_x8("mask", 0600, ibd->fault_opcode->dir, + &ibd->fault_opcode->mask)) + goto fail; + + return 0; +fail: + fault_exit_opcode_debugfs(ibd); + return -ENOMEM; +} + +static void fault_exit_packet_debugfs(struct hfi1_ibdev *ibd) +{ + debugfs_remove_recursive(ibd->fault_packet->dir); + kfree(ibd->fault_packet); + ibd->fault_packet = NULL; +} + +static int fault_init_packet_debugfs(struct hfi1_ibdev *ibd) +{ + struct dentry *parent = ibd->hfi1_ibdev_dbg; + + ibd->fault_packet = kzalloc(sizeof(*ibd->fault_packet), GFP_KERNEL); + if (!ibd->fault_packet) + return -ENOMEM; + + ibd->fault_packet->attr.interval = 1; + ibd->fault_packet->attr.require_end = ULONG_MAX; + ibd->fault_packet->attr.stacktrace_depth = 32; + ibd->fault_packet->attr.dname = NULL; + ibd->fault_packet->attr.verbose = 0; + ibd->fault_packet->fault_by_packet = false; + + ibd->fault_packet->dir = + fault_create_debugfs_attr("fault_packet", + parent, + &ibd->fault_opcode->attr); + if (IS_ERR(ibd->fault_packet->dir)) { + kfree(ibd->fault_packet); + return -ENOENT; + } + + if (!debugfs_create_bool("fault_by_packet", 0600, + ibd->fault_packet->dir, + &ibd->fault_packet->fault_by_packet)) + goto fail; + if (!debugfs_create_u64("fault_stats", 0400, + ibd->fault_packet->dir, + &ibd->fault_packet->n_faults)) + goto fail; + + return 0; +fail: + fault_exit_packet_debugfs(ibd); + return -ENOMEM; +} + +static void fault_exit_debugfs(struct hfi1_ibdev *ibd) +{ + fault_exit_opcode_debugfs(ibd); + fault_exit_packet_debugfs(ibd); +} + +static int fault_init_debugfs(struct hfi1_ibdev *ibd) +{ + int ret = 0; + + ret = fault_init_opcode_debugfs(ibd); + if (ret) + return ret; + + ret = fault_init_packet_debugfs(ibd); + if (ret) + fault_exit_opcode_debugfs(ibd); + + return ret; +} + +bool hfi1_dbg_fault_suppress_err(struct hfi1_ibdev *ibd) +{ + return ibd->fault_suppress_err; +} + +bool hfi1_dbg_fault_opcode(struct rvt_qp *qp, u32 opcode, bool rx) +{ + bool ret = false; + struct hfi1_ibdev *ibd = to_idev(qp->ibqp.device); + + if (!ibd->fault_opcode || !ibd->fault_opcode->fault_by_opcode) + return false; + if (ibd->fault_opcode->opcode != (opcode & ibd->fault_opcode->mask)) + return false; + ret = should_fail(&ibd->fault_opcode->attr, 1); + if (ret) { + trace_hfi1_fault_opcode(qp, opcode); + if (rx) + ibd->fault_opcode->n_rxfaults[opcode]++; + else + ibd->fault_opcode->n_txfaults[opcode]++; + } + return ret; +} + +bool hfi1_dbg_fault_packet(struct hfi1_packet *packet) +{ + struct rvt_dev_info *rdi = &packet->rcd->ppd->dd->verbs_dev.rdi; + struct hfi1_ibdev *ibd = dev_from_rdi(rdi); + bool ret = false; + + if (!ibd->fault_packet || !ibd->fault_packet->fault_by_packet) + return false; + + ret = should_fail(&ibd->fault_packet->attr, 1); + if (ret) { + ++ibd->fault_packet->n_faults; + trace_hfi1_fault_packet(packet); + } + return ret; +} +#endif + void hfi1_dbg_ibdev_init(struct hfi1_ibdev *ibd) { char name[sizeof("port0counters") + 1]; @@ -1129,12 +1332,22 @@ void hfi1_dbg_ibdev_init(struct hfi1_ibdev *ibd) !port_cntr_ops[i].ops.write ? S_IRUGO : S_IRUGO | S_IWUSR); } + +#ifdef CONFIG_FAULT_INJECTION + debugfs_create_bool("fault_suppress_err", 0600, + ibd->hfi1_ibdev_dbg, + &ibd->fault_suppress_err); + fault_init_debugfs(ibd); +#endif } void hfi1_dbg_ibdev_exit(struct hfi1_ibdev *ibd) { if (!hfi1_dbg_root) goto out; +#ifdef CONFIG_FAULT_INJECTION + fault_exit_debugfs(ibd); +#endif debugfs_remove(ibd->hfi1_ibdev_link); debugfs_remove_recursive(ibd->hfi1_ibdev_dbg); out: diff --git a/drivers/infiniband/hw/hfi1/debugfs.h b/drivers/infiniband/hw/hfi1/debugfs.h index b6fb6814f1b8..38c38a98156d 100644 --- a/drivers/infiniband/hw/hfi1/debugfs.h +++ b/drivers/infiniband/hw/hfi1/debugfs.h @@ -53,23 +53,79 @@ void hfi1_dbg_ibdev_init(struct hfi1_ibdev *ibd); void hfi1_dbg_ibdev_exit(struct hfi1_ibdev *ibd); void hfi1_dbg_init(void); void hfi1_dbg_exit(void); + +#ifdef CONFIG_FAULT_INJECTION +#include <linux/fault-inject.h> +struct fault_opcode { + struct fault_attr attr; + struct dentry *dir; + bool fault_by_opcode; + u64 n_rxfaults[256]; + u64 n_txfaults[256]; + u8 opcode; + u8 mask; +}; + +struct fault_packet { + struct fault_attr attr; + struct dentry *dir; + bool fault_by_packet; + u64 n_faults; +}; + +bool hfi1_dbg_fault_opcode(struct rvt_qp *qp, u32 opcode, bool rx); +bool hfi1_dbg_fault_packet(struct hfi1_packet *packet); +bool hfi1_dbg_fault_suppress_err(struct hfi1_ibdev *ibd); +#else +static inline bool hfi1_dbg_fault_packet(struct hfi1_packet *packet) +{ + return false; +} + +static inline bool hfi1_dbg_fault_opcode(struct rvt_qp *qp, + u32 opcode, bool rx) +{ + return false; +} + +static inline bool hfi1_dbg_fault_suppress_err(struct hfi1_ibdev *ibd) +{ + return false; +} +#endif + #else static inline void hfi1_dbg_ibdev_init(struct hfi1_ibdev *ibd) { } -void hfi1_dbg_ibdev_exit(struct hfi1_ibdev *ibd) +static inline void hfi1_dbg_ibdev_exit(struct hfi1_ibdev *ibd) +{ +} + +static inline void hfi1_dbg_init(void) { } -void hfi1_dbg_init(void) +static inline void hfi1_dbg_exit(void) { } -void hfi1_dbg_exit(void) +static inline bool hfi1_dbg_fault_packet(struct hfi1_packet *packet) { + return false; } +static inline bool hfi1_dbg_fault_opcode(struct rvt_qp *qp, + u32 opcode, bool rx) +{ + return false; +} + +static inline bool hfi1_dbg_fault_suppress_err(struct hfi1_ibdev *ibd) +{ + return false; +} #endif #endif /* _HFI1_DEBUGFS_H */ diff --git a/drivers/infiniband/hw/hfi1/device.c b/drivers/infiniband/hw/hfi1/device.c index bf64b5a7bfd7..bbb6069dec2a 100644 --- a/drivers/infiniband/hw/hfi1/device.c +++ b/drivers/infiniband/hw/hfi1/device.c @@ -69,7 +69,7 @@ int hfi1_cdev_init(int minor, const char *name, cdev_init(cdev, fops); cdev->owner = THIS_MODULE; - cdev->kobj.parent = parent; + cdev_set_parent(cdev, parent); kobject_set_name(&cdev->kobj, name); ret = cdev_add(cdev, dev, 1); diff --git a/drivers/infiniband/hw/hfi1/dma.c b/drivers/infiniband/hw/hfi1/dma.c deleted file mode 100644 index 7e8dab892848..000000000000 --- a/drivers/infiniband/hw/hfi1/dma.c +++ /dev/null @@ -1,183 +0,0 @@ -/* - * Copyright(c) 2015, 2016 Intel Corporation. - * - * This file is provided under a dual BSD/GPLv2 license. When using or - * redistributing this file, you may do so under either license. - * - * GPL LICENSE SUMMARY - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of version 2 of the GNU General Public License as - * published by the Free Software Foundation. - * - * This program is distributed in the hope that it will be useful, but - * WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * General Public License for more details. - * - * BSD LICENSE - * - * 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. - * - Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * - */ -#include <linux/types.h> -#include <linux/scatterlist.h> - -#include "verbs.h" - -#define BAD_DMA_ADDRESS ((u64)0) - -/* - * The following functions implement driver specific replacements - * for the ib_dma_*() functions. - * - * These functions return kernel virtual addresses instead of - * device bus addresses since the driver uses the CPU to copy - * data instead of using hardware DMA. - */ - -static int hfi1_mapping_error(struct ib_device *dev, u64 dma_addr) -{ - return dma_addr == BAD_DMA_ADDRESS; -} - -static u64 hfi1_dma_map_single(struct ib_device *dev, void *cpu_addr, - size_t size, enum dma_data_direction direction) -{ - if (WARN_ON(!valid_dma_direction(direction))) - return BAD_DMA_ADDRESS; - - return (u64)cpu_addr; -} - -static void hfi1_dma_unmap_single(struct ib_device *dev, u64 addr, size_t size, - enum dma_data_direction direction) -{ - /* This is a stub, nothing to be done here */ -} - -static u64 hfi1_dma_map_page(struct ib_device *dev, struct page *page, - unsigned long offset, size_t size, - enum dma_data_direction direction) -{ - u64 addr; - - if (WARN_ON(!valid_dma_direction(direction))) - return BAD_DMA_ADDRESS; - - if (offset + size > PAGE_SIZE) - return BAD_DMA_ADDRESS; - - addr = (u64)page_address(page); - if (addr) - addr += offset; - - return addr; -} - -static void hfi1_dma_unmap_page(struct ib_device *dev, u64 addr, size_t size, - enum dma_data_direction direction) -{ - /* This is a stub, nothing to be done here */ -} - -static int hfi1_map_sg(struct ib_device *dev, struct scatterlist *sgl, - int nents, enum dma_data_direction direction) -{ - struct scatterlist *sg; - u64 addr; - int i; - int ret = nents; - - if (WARN_ON(!valid_dma_direction(direction))) - return BAD_DMA_ADDRESS; - - for_each_sg(sgl, sg, nents, i) { - addr = (u64)page_address(sg_page(sg)); - if (!addr) { - ret = 0; - break; - } - sg->dma_address = addr + sg->offset; -#ifdef CONFIG_NEED_SG_DMA_LENGTH - sg->dma_length = sg->length; -#endif - } - return ret; -} - -static void hfi1_unmap_sg(struct ib_device *dev, - struct scatterlist *sg, int nents, - enum dma_data_direction direction) -{ - /* This is a stub, nothing to be done here */ -} - -static void hfi1_sync_single_for_cpu(struct ib_device *dev, u64 addr, - size_t size, enum dma_data_direction dir) -{ -} - -static void hfi1_sync_single_for_device(struct ib_device *dev, u64 addr, - size_t size, - enum dma_data_direction dir) -{ -} - -static void *hfi1_dma_alloc_coherent(struct ib_device *dev, size_t size, - u64 *dma_handle, gfp_t flag) -{ - struct page *p; - void *addr = NULL; - - p = alloc_pages(flag, get_order(size)); - if (p) - addr = page_address(p); - if (dma_handle) - *dma_handle = (u64)addr; - return addr; -} - -static void hfi1_dma_free_coherent(struct ib_device *dev, size_t size, - void *cpu_addr, u64 dma_handle) -{ - free_pages((unsigned long)cpu_addr, get_order(size)); -} - -struct ib_dma_mapping_ops hfi1_dma_mapping_ops = { - .mapping_error = hfi1_mapping_error, - .map_single = hfi1_dma_map_single, - .unmap_single = hfi1_dma_unmap_single, - .map_page = hfi1_dma_map_page, - .unmap_page = hfi1_dma_unmap_page, - .map_sg = hfi1_map_sg, - .unmap_sg = hfi1_unmap_sg, - .sync_single_for_cpu = hfi1_sync_single_for_cpu, - .sync_single_for_device = hfi1_sync_single_for_device, - .alloc_coherent = hfi1_dma_alloc_coherent, - .free_coherent = hfi1_dma_free_coherent -}; diff --git a/drivers/infiniband/hw/hfi1/driver.c b/drivers/infiniband/hw/hfi1/driver.c index 4fbaee68012b..a50870e455a3 100644 --- a/drivers/infiniband/hw/hfi1/driver.c +++ b/drivers/infiniband/hw/hfi1/driver.c @@ -1,5 +1,5 @@ /* - * Copyright(c) 2015, 2016 Intel Corporation. + * Copyright(c) 2015-2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -59,6 +59,8 @@ #include "trace.h" #include "qp.h" #include "sdma.h" +#include "debugfs.h" +#include "vnic.h" #undef pr_fmt #define pr_fmt(fmt) DRIVER_NAME ": " fmt @@ -83,8 +85,8 @@ module_param_named(cu, hfi1_cu, uint, S_IRUGO); MODULE_PARM_DESC(cu, "Credit return units"); unsigned long hfi1_cap_mask = HFI1_CAP_MASK_DEFAULT; -static int hfi1_caps_set(const char *, const struct kernel_param *); -static int hfi1_caps_get(char *, const struct kernel_param *); +static int hfi1_caps_set(const char *val, const struct kernel_param *kp); +static int hfi1_caps_get(char *buffer, const struct kernel_param *kp); static const struct kernel_param_ops cap_ops = { .set = hfi1_caps_set, .get = hfi1_caps_get @@ -100,6 +102,11 @@ MODULE_VERSION(HFI1_DRIVER_VERSION); * MAX_PKT_RCV is the max # if packets processed per receive interrupt. */ #define MAX_PKT_RECV 64 +/* + * MAX_PKT_THREAD_RCV is the max # of packets processed before + * the qp_wait_list queue is flushed. + */ +#define MAX_PKT_RECV_THREAD (MAX_PKT_RECV * 4) #define EGR_HEAD_UPDATE_THRESHOLD 16 struct hfi1_ib_stats hfi1_stats; @@ -204,42 +211,6 @@ int hfi1_count_active_units(void) } /* - * Return count of all units, optionally return in arguments - * the number of usable (present) units, and the number of - * ports that are up. - */ -int hfi1_count_units(int *npresentp, int *nupp) -{ - int nunits = 0, npresent = 0, nup = 0; - struct hfi1_devdata *dd; - unsigned long flags; - int pidx; - struct hfi1_pportdata *ppd; - - spin_lock_irqsave(&hfi1_devs_lock, flags); - - list_for_each_entry(dd, &hfi1_dev_list, list) { - nunits++; - if ((dd->flags & HFI1_PRESENT) && dd->kregbase) - npresent++; - for (pidx = 0; pidx < dd->num_pports; ++pidx) { - ppd = dd->pport + pidx; - if (ppd->lid && ppd->linkup) - nup++; - } - } - - spin_unlock_irqrestore(&hfi1_devs_lock, flags); - - if (npresentp) - *npresentp = npresent; - if (nupp) - *nupp = nup; - - return nunits; -} - -/* * Get address of eager buffer from it's index (allocated in chunks, not * contiguous). */ @@ -259,7 +230,7 @@ static inline void *get_egrbuf(const struct hfi1_ctxtdata *rcd, u64 rhf, * allowed size ranges for the respective type and, optionally, * return the proper encoding. */ -inline int hfi1_rcvbuf_validate(u32 size, u8 type, u16 *encoded) +int hfi1_rcvbuf_validate(u32 size, u8 type, u16 *encoded) { if (unlikely(!PAGE_ALIGNED(size))) return 0; @@ -278,8 +249,8 @@ static void rcv_hdrerr(struct hfi1_ctxtdata *rcd, struct hfi1_pportdata *ppd, { struct ib_header *rhdr = packet->hdr; u32 rte = rhf_rcv_type_err(packet->rhf); - int lnh = be16_to_cpu(rhdr->lrh[0]) & 3; - struct hfi1_ibport *ibp = &ppd->ibport_data; + int lnh = ib_get_lnh(rhdr); + struct hfi1_ibport *ibp = rcd_to_iport(rcd); struct hfi1_devdata *dd = ppd->dd; struct rvt_dev_info *rdi = &dd->verbs_dev.rdi; @@ -290,7 +261,7 @@ static void rcv_hdrerr(struct hfi1_ctxtdata *rcd, struct hfi1_pportdata *ppd, /* For TIDERR and RC QPs preemptively schedule a NAK */ struct ib_other_headers *ohdr = NULL; u32 tlen = rhf_pkt_len(packet->rhf); /* in bytes */ - u16 lid = be16_to_cpu(rhdr->lrh[1]); + u16 lid = ib_get_dlid(rhdr); u32 qp_num; u32 rcv_flags = 0; @@ -391,7 +362,7 @@ static void rcv_hdrerr(struct hfi1_ctxtdata *rcd, struct hfi1_pportdata *ppd, u16 rlid; u8 svc_type, sl, sc5; - sc5 = hdr2sc(rhdr, packet->rhf); + sc5 = hfi1_9B_get_sc5(rhdr, packet->rhf); sl = ibp->sc_to_sl[sc5]; lqpn = be32_to_cpu(bth[1]) & RVT_QPN_MASK; @@ -409,7 +380,7 @@ static void rcv_hdrerr(struct hfi1_ctxtdata *rcd, struct hfi1_pportdata *ppd, svc_type = IB_CC_SVCTYPE_UD; break; case IB_QPT_UC: - rlid = be16_to_cpu(rhdr->lrh[3]); + rlid = ib_get_slid(rhdr); rqpn = qp->remote_qpn; svc_type = IB_CC_SVCTYPE_UC; break; @@ -455,7 +426,7 @@ void hfi1_process_ecn_slowpath(struct rvt_qp *qp, struct hfi1_packet *pkt, struct ib_other_headers *ohdr = pkt->ohdr; struct ib_grh *grh = NULL; u32 rqpn = 0, bth1; - u16 rlid, dlid = be16_to_cpu(hdr->lrh[1]); + u16 rlid, dlid = ib_get_dlid(hdr); u8 sc, svc_type; bool is_mcast = false; @@ -466,19 +437,19 @@ void hfi1_process_ecn_slowpath(struct rvt_qp *qp, struct hfi1_packet *pkt, case IB_QPT_SMI: case IB_QPT_GSI: case IB_QPT_UD: - rlid = be16_to_cpu(hdr->lrh[3]); + rlid = ib_get_slid(hdr); rqpn = be32_to_cpu(ohdr->u.ud.deth[1]) & RVT_QPN_MASK; svc_type = IB_CC_SVCTYPE_UD; is_mcast = (dlid > be16_to_cpu(IB_MULTICAST_LID_BASE)) && (dlid != be16_to_cpu(IB_LID_PERMISSIVE)); break; case IB_QPT_UC: - rlid = qp->remote_ah_attr.dlid; + rlid = rdma_ah_get_dlid(&qp->remote_ah_attr); rqpn = qp->remote_qpn; svc_type = IB_CC_SVCTYPE_UC; break; case IB_QPT_RC: - rlid = qp->remote_ah_attr.dlid; + rlid = rdma_ah_get_dlid(&qp->remote_ah_attr); rqpn = qp->remote_qpn; svc_type = IB_CC_SVCTYPE_RC; break; @@ -486,16 +457,16 @@ void hfi1_process_ecn_slowpath(struct rvt_qp *qp, struct hfi1_packet *pkt, return; } - sc = hdr2sc(hdr, pkt->rhf); + sc = hfi1_9B_get_sc5(hdr, pkt->rhf); bth1 = be32_to_cpu(ohdr->bth[1]); - if (do_cnp && (bth1 & HFI1_FECN_SMASK)) { + if (do_cnp && (bth1 & IB_FECN_SMASK)) { u16 pkey = (u16)be32_to_cpu(ohdr->bth[0]); return_cnp(ibp, qp, rqpn, pkey, dlid, rlid, sc, grh); } - if (!is_mcast && (bth1 & HFI1_BECN_SMASK)) { + if (!is_mcast && (bth1 & IB_BECN_SMASK)) { struct hfi1_pportdata *ppd = ppd_from_ibp(ibp); u32 lqpn = bth1 & RVT_QPN_MASK; u8 sl = ibp->sc_to_sl[sc]; @@ -594,7 +565,7 @@ static void __prescan_rxq(struct hfi1_packet *packet) while (1) { struct hfi1_devdata *dd = rcd->dd; - struct hfi1_ibport *ibp = &rcd->ppd->ibport_data; + struct hfi1_ibport *ibp = rcd_to_iport(rcd); __le32 *rhf_addr = (__le32 *)rcd->rcvhdrq + mdata.ps_head + dd->rhf_offset; struct rvt_qp *qp; @@ -616,8 +587,7 @@ static void __prescan_rxq(struct hfi1_packet *packet) packet->hdr = hfi1_get_msgheader(dd, rhf_addr); hdr = packet->hdr; - - lnh = be16_to_cpu(hdr->lrh[0]) & 3; + lnh = ib_get_lnh(hdr); if (lnh == HFI1_LRH_BTH) { packet->ohdr = &hdr->u.oth; @@ -629,7 +599,7 @@ static void __prescan_rxq(struct hfi1_packet *packet) } bth1 = be32_to_cpu(packet->ohdr->bth[1]); - is_ecn = !!(bth1 & (HFI1_FECN_SMASK | HFI1_BECN_SMASK)); + is_ecn = !!(bth1 & (IB_FECN_SMASK | IB_BECN_SMASK)); if (!is_ecn) goto next; @@ -647,31 +617,75 @@ static void __prescan_rxq(struct hfi1_packet *packet) rcu_read_unlock(); /* turn off BECN, FECN */ - bth1 &= ~(HFI1_FECN_SMASK | HFI1_BECN_SMASK); + bth1 &= ~(IB_FECN_SMASK | IB_BECN_SMASK); packet->ohdr->bth[1] = cpu_to_be32(bth1); next: update_ps_mdata(&mdata, rcd); } } -static inline int skip_rcv_packet(struct hfi1_packet *packet, int thread) +static void process_rcv_qp_work(struct hfi1_ctxtdata *rcd) +{ + struct rvt_qp *qp, *nqp; + + /* + * Iterate over all QPs waiting to respond. + * The list won't change since the IRQ is only run on one CPU. + */ + list_for_each_entry_safe(qp, nqp, &rcd->qp_wait_list, rspwait) { + list_del_init(&qp->rspwait); + if (qp->r_flags & RVT_R_RSP_NAK) { + qp->r_flags &= ~RVT_R_RSP_NAK; + hfi1_send_rc_ack(rcd, qp, 0); + } + if (qp->r_flags & RVT_R_RSP_SEND) { + unsigned long flags; + + qp->r_flags &= ~RVT_R_RSP_SEND; + spin_lock_irqsave(&qp->s_lock, flags); + if (ib_rvt_state_ops[qp->state] & + RVT_PROCESS_OR_FLUSH_SEND) + hfi1_schedule_send(qp); + spin_unlock_irqrestore(&qp->s_lock, flags); + } + rvt_put_qp(qp); + } +} + +static noinline int max_packet_exceeded(struct hfi1_packet *packet, int thread) +{ + if (thread) { + if ((packet->numpkt & (MAX_PKT_RECV_THREAD - 1)) == 0) + /* allow defered processing */ + process_rcv_qp_work(packet->rcd); + cond_resched(); + return RCV_PKT_OK; + } else { + this_cpu_inc(*packet->rcd->dd->rcv_limit); + return RCV_PKT_LIMIT; + } +} + +static inline int check_max_packet(struct hfi1_packet *packet, int thread) { int ret = RCV_PKT_OK; + if (unlikely((packet->numpkt & (MAX_PKT_RECV - 1)) == 0)) + ret = max_packet_exceeded(packet, thread); + return ret; +} + +static noinline int skip_rcv_packet(struct hfi1_packet *packet, int thread) +{ + int ret; + /* Set up for the next packet */ packet->rhqoff += packet->rsize; if (packet->rhqoff >= packet->maxcnt) packet->rhqoff = 0; packet->numpkt++; - if (unlikely((packet->numpkt & (MAX_PKT_RECV - 1)) == 0)) { - if (thread) { - cond_resched(); - } else { - ret = RCV_PKT_LIMIT; - this_cpu_inc(*packet->rcd->dd->rcv_limit); - } - } + ret = check_max_packet(packet, thread); packet->rhf_addr = (__le32 *)packet->rcd->rcvhdrq + packet->rhqoff + packet->rcd->dd->rhf_offset; @@ -682,7 +696,7 @@ static inline int skip_rcv_packet(struct hfi1_packet *packet, int thread) static inline int process_rcv_packet(struct hfi1_packet *packet, int thread) { - int ret = RCV_PKT_OK; + int ret; packet->hdr = hfi1_get_msgheader(packet->rcd->dd, packet->rhf_addr); @@ -723,14 +737,7 @@ static inline int process_rcv_packet(struct hfi1_packet *packet, int thread) if (packet->rhqoff >= packet->maxcnt) packet->rhqoff = 0; - if (unlikely((packet->numpkt & (MAX_PKT_RECV - 1)) == 0)) { - if (thread) { - cond_resched(); - } else { - ret = RCV_PKT_LIMIT; - this_cpu_inc(*packet->rcd->dd->rcv_limit); - } - } + ret = check_max_packet(packet, thread); packet->rhf_addr = (__le32 *)packet->rcd->rcvhdrq + packet->rhqoff + packet->rcd->dd->rhf_offset; @@ -767,38 +774,6 @@ static inline void finish_packet(struct hfi1_packet *packet) packet->etail, rcv_intr_dynamic, packet->numpkt); } -static inline void process_rcv_qp_work(struct hfi1_packet *packet) -{ - struct hfi1_ctxtdata *rcd; - struct rvt_qp *qp, *nqp; - - rcd = packet->rcd; - rcd->head = packet->rhqoff; - - /* - * Iterate over all QPs waiting to respond. - * The list won't change since the IRQ is only run on one CPU. - */ - list_for_each_entry_safe(qp, nqp, &rcd->qp_wait_list, rspwait) { - list_del_init(&qp->rspwait); - if (qp->r_flags & RVT_R_RSP_NAK) { - qp->r_flags &= ~RVT_R_RSP_NAK; - hfi1_send_rc_ack(rcd, qp, 0); - } - if (qp->r_flags & RVT_R_RSP_SEND) { - unsigned long flags; - - qp->r_flags &= ~RVT_R_RSP_SEND; - spin_lock_irqsave(&qp->s_lock, flags); - if (ib_rvt_state_ops[qp->state] & - RVT_PROCESS_OR_FLUSH_SEND) - hfi1_schedule_send(qp); - spin_unlock_irqrestore(&qp->s_lock, flags); - } - rvt_put_qp(qp); - } -} - /* * Handle receive interrupts when using the no dma rtail option. */ @@ -826,7 +801,8 @@ int handle_receive_interrupt_nodma_rtail(struct hfi1_ctxtdata *rcd, int thread) last = RCV_PKT_DONE; process_rcv_update(last, &packet); } - process_rcv_qp_work(&packet); + process_rcv_qp_work(rcd); + rcd->head = packet.rhqoff; bail: finish_packet(&packet); return last; @@ -854,26 +830,49 @@ int handle_receive_interrupt_dma_rtail(struct hfi1_ctxtdata *rcd, int thread) last = RCV_PKT_DONE; process_rcv_update(last, &packet); } - process_rcv_qp_work(&packet); + process_rcv_qp_work(rcd); + rcd->head = packet.rhqoff; bail: finish_packet(&packet); return last; } -static inline void set_all_nodma_rtail(struct hfi1_devdata *dd) +static inline void set_nodma_rtail(struct hfi1_devdata *dd, u8 ctxt) { int i; - for (i = HFI1_CTRL_CTXT + 1; i < dd->first_user_ctxt; i++) + /* + * For dynamically allocated kernel contexts (like vnic) switch + * interrupt handler only for that context. Otherwise, switch + * interrupt handler for all statically allocated kernel contexts. + */ + if (ctxt >= dd->first_dyn_alloc_ctxt) { + dd->rcd[ctxt]->do_interrupt = + &handle_receive_interrupt_nodma_rtail; + return; + } + + for (i = HFI1_CTRL_CTXT + 1; i < dd->first_dyn_alloc_ctxt; i++) dd->rcd[i]->do_interrupt = &handle_receive_interrupt_nodma_rtail; } -static inline void set_all_dma_rtail(struct hfi1_devdata *dd) +static inline void set_dma_rtail(struct hfi1_devdata *dd, u8 ctxt) { int i; - for (i = HFI1_CTRL_CTXT + 1; i < dd->first_user_ctxt; i++) + /* + * For dynamically allocated kernel contexts (like vnic) switch + * interrupt handler only for that context. Otherwise, switch + * interrupt handler for all statically allocated kernel contexts. + */ + if (ctxt >= dd->first_dyn_alloc_ctxt) { + dd->rcd[ctxt]->do_interrupt = + &handle_receive_interrupt_dma_rtail; + return; + } + + for (i = HFI1_CTRL_CTXT + 1; i < dd->first_dyn_alloc_ctxt; i++) dd->rcd[i]->do_interrupt = &handle_receive_interrupt_dma_rtail; } @@ -883,8 +882,13 @@ void set_all_slowpath(struct hfi1_devdata *dd) int i; /* HFI1_CTRL_CTXT must always use the slow path interrupt handler */ - for (i = HFI1_CTRL_CTXT + 1; i < dd->first_user_ctxt; i++) - dd->rcd[i]->do_interrupt = &handle_receive_interrupt; + for (i = HFI1_CTRL_CTXT + 1; i < dd->num_rcv_contexts; i++) { + struct hfi1_ctxtdata *rcd = dd->rcd[i]; + + if ((i < dd->first_dyn_alloc_ctxt) || + (rcd && rcd->sc && (rcd->sc->type == SC_KERNEL))) + rcd->do_interrupt = &handle_receive_interrupt; + } } static inline int set_armed_to_active(struct hfi1_ctxtdata *rcd, @@ -896,7 +900,8 @@ static inline int set_armed_to_active(struct hfi1_ctxtdata *rcd, packet->rhf_addr); u8 etype = rhf_rcv_type(packet->rhf); - if (etype == RHF_RCV_TYPE_IB && hdr2sc(hdr, packet->rhf) != 0xf) { + if (etype == RHF_RCV_TYPE_IB && + hfi1_9B_get_sc5(hdr, packet->rhf) != 0xf) { int hwstate = read_logical_state(dd); if (hwstate != LSTATE_ACTIVE) { @@ -994,7 +999,7 @@ int handle_receive_interrupt(struct hfi1_ctxtdata *rcd, int thread) last = RCV_PKT_DONE; if (needset) { dd_dev_info(dd, "Switching to NO_DMA_RTAIL\n"); - set_all_nodma_rtail(dd); + set_nodma_rtail(dd, rcd->ctxt); needset = 0; } } else { @@ -1016,7 +1021,7 @@ int handle_receive_interrupt(struct hfi1_ctxtdata *rcd, int thread) if (needset) { dd_dev_info(dd, "Switching to DMA_RTAIL\n"); - set_all_dma_rtail(dd); + set_dma_rtail(dd, rcd->ctxt); needset = 0; } } @@ -1024,7 +1029,8 @@ int handle_receive_interrupt(struct hfi1_ctxtdata *rcd, int thread) process_rcv_update(last, &packet); } - process_rcv_qp_work(&packet); + process_rcv_qp_work(rcd); + rcd->head = packet.rhqoff; bail: /* @@ -1064,10 +1070,10 @@ void receive_interrupt_work(struct work_struct *work) set_link_state(ppd, HLS_UP_ACTIVE); /* - * Interrupt all kernel contexts that could have had an - * interrupt during auto activation. + * Interrupt all statically allocated kernel contexts that could + * have had an interrupt during auto activation. */ - for (i = HFI1_CTRL_CTXT; i < dd->first_user_ctxt; i++) + for (i = HFI1_CTRL_CTXT; i < dd->first_dyn_alloc_ctxt; i++) force_recv_intr(dd->rcd[i]); } @@ -1281,8 +1287,9 @@ int hfi1_reset_device(int unit) spin_lock_irqsave(&dd->uctxt_lock, flags); if (dd->rcd) - for (i = dd->first_user_ctxt; i < dd->num_rcv_contexts; i++) { - if (!dd->rcd[i] || !dd->rcd[i]->cnt) + for (i = dd->first_dyn_alloc_ctxt; + i < dd->num_rcv_contexts; i++) { + if (!dd->rcd[i]) continue; spin_unlock_irqrestore(&dd->uctxt_lock, flags); ret = -EBUSY; @@ -1341,6 +1348,9 @@ void handle_eflags(struct hfi1_packet *packet) */ int process_receive_ib(struct hfi1_packet *packet) { + if (unlikely(hfi1_dbg_fault_packet(packet))) + return RHF_RCV_CONTINUE; + trace_hfi1_rcvhdr(packet->rcd->ppd->dd, packet->rcd->ctxt, rhf_err_flags(packet->rhf), @@ -1350,6 +1360,11 @@ int process_receive_ib(struct hfi1_packet *packet) packet->updegr, rhf_egr_index(packet->rhf)); + if (unlikely( + (hfi1_dbg_fault_suppress_err(&packet->rcd->dd->verbs_dev) && + (packet->rhf & RHF_DC_ERR)))) + return RHF_RCV_CONTINUE; + if (unlikely(rhf_err_flags(packet->rhf))) { handle_eflags(packet); return RHF_RCV_CONTINUE; @@ -1359,15 +1374,31 @@ int process_receive_ib(struct hfi1_packet *packet) return RHF_RCV_CONTINUE; } +static inline bool hfi1_is_vnic_packet(struct hfi1_packet *packet) +{ + /* Packet received in VNIC context via RSM */ + if (packet->rcd->is_vnic) + return true; + + if ((HFI1_GET_L2_TYPE(packet->ebuf) == OPA_VNIC_L2_TYPE) && + (HFI1_GET_L4_TYPE(packet->ebuf) == OPA_VNIC_L4_ETHR)) + return true; + + return false; +} + int process_receive_bypass(struct hfi1_packet *packet) { struct hfi1_devdata *dd = packet->rcd->dd; - if (unlikely(rhf_err_flags(packet->rhf))) + if (unlikely(rhf_err_flags(packet->rhf))) { handle_eflags(packet); + } else if (hfi1_is_vnic_packet(packet)) { + hfi1_vnic_bypass_rcv(packet); + return RHF_RCV_CONTINUE; + } - dd_dev_err(dd, - "Bypass packets are not supported in normal operation. Dropping\n"); + dd_dev_err(dd, "Unsupported bypass packet. Dropping\n"); incr_cntr64(&dd->sw_rcv_bypass_packet_errors); if (!(dd->err_info_rcvport.status_and_code & OPA_EI_STATUS_SMASK)) { u64 *flits = packet->ebuf; @@ -1385,6 +1416,12 @@ int process_receive_bypass(struct hfi1_packet *packet) int process_receive_error(struct hfi1_packet *packet) { + /* KHdrHCRCErr -- KDETH packet with a bad HCRC */ + if (unlikely( + hfi1_dbg_fault_suppress_err(&packet->rcd->dd->verbs_dev) && + rhf_rcv_type_err(packet->rhf) == 3)) + return RHF_RCV_CONTINUE; + handle_eflags(packet); if (unlikely(rhf_err_flags(packet->rhf))) @@ -1396,6 +1433,8 @@ int process_receive_error(struct hfi1_packet *packet) int kdeth_process_expected(struct hfi1_packet *packet) { + if (unlikely(hfi1_dbg_fault_packet(packet))) + return RHF_RCV_CONTINUE; if (unlikely(rhf_err_flags(packet->rhf))) handle_eflags(packet); @@ -1408,6 +1447,8 @@ int kdeth_process_eager(struct hfi1_packet *packet) { if (unlikely(rhf_err_flags(packet->rhf))) handle_eflags(packet); + if (unlikely(hfi1_dbg_fault_packet(packet))) + return RHF_RCV_CONTINUE; dd_dev_err(packet->rcd->dd, "Unhandled eager packet received. Dropping.\n"); diff --git a/drivers/infiniband/hw/hfi1/efivar.c b/drivers/infiniband/hw/hfi1/efivar.c index 106349fc1fb9..d106d23016ba 100644 --- a/drivers/infiniband/hw/hfi1/efivar.c +++ b/drivers/infiniband/hw/hfi1/efivar.c @@ -45,6 +45,7 @@ * */ +#include <linux/ctype.h> #include "efivar.h" /* GUID for HFI1 variables in EFI */ @@ -150,15 +151,32 @@ fail: int read_hfi1_efi_var(struct hfi1_devdata *dd, const char *kind, unsigned long *size, void **return_data) { + char prefix_name[64]; char name[64]; + int result; + int i; /* create a common prefix */ - snprintf(name, sizeof(name), "%04x:%02x:%02x.%x-%s", + snprintf(prefix_name, sizeof(prefix_name), "%04x:%02x:%02x.%x", pci_domain_nr(dd->pcidev->bus), dd->pcidev->bus->number, PCI_SLOT(dd->pcidev->devfn), - PCI_FUNC(dd->pcidev->devfn), - kind); + PCI_FUNC(dd->pcidev->devfn)); + snprintf(name, sizeof(name), "%s-%s", prefix_name, kind); + result = read_efi_var(name, size, return_data); + + /* + * If reading the lowercase EFI variable fail, read the uppercase + * variable. + */ + if (result) { + /* Converting to uppercase */ + for (i = 0; prefix_name[i]; i++) + if (isalpha(prefix_name[i])) + prefix_name[i] = toupper(prefix_name[i]); + snprintf(name, sizeof(name), "%s-%s", prefix_name, kind); + result = read_efi_var(name, size, return_data); + } - return read_efi_var(name, size, return_data); + return result; } diff --git a/drivers/infiniband/hw/hfi1/file_ops.c b/drivers/infiniband/hw/hfi1/file_ops.c index bd786b7bd30b..3158128d57e8 100644 --- a/drivers/infiniband/hw/hfi1/file_ops.c +++ b/drivers/infiniband/hw/hfi1/file_ops.c @@ -1,5 +1,5 @@ /* - * Copyright(c) 2015, 2016 Intel Corporation. + * Copyright(c) 2015-2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -48,6 +48,8 @@ #include <linux/cdev.h> #include <linux/vmalloc.h> #include <linux/io.h> +#include <linux/sched/mm.h> +#include <linux/bitmap.h> #include <rdma/ib.h> @@ -69,30 +71,37 @@ /* * File operation functions */ -static int hfi1_file_open(struct inode *, struct file *); -static int hfi1_file_close(struct inode *, struct file *); -static ssize_t hfi1_write_iter(struct kiocb *, struct iov_iter *); -static unsigned int hfi1_poll(struct file *, struct poll_table_struct *); -static int hfi1_file_mmap(struct file *, struct vm_area_struct *); - -static u64 kvirt_to_phys(void *); -static int assign_ctxt(struct file *, struct hfi1_user_info *); -static int init_subctxts(struct hfi1_ctxtdata *, const struct hfi1_user_info *); -static int user_init(struct file *); -static int get_ctxt_info(struct file *, void __user *, __u32); -static int get_base_info(struct file *, void __user *, __u32); -static int setup_ctxt(struct file *); -static int setup_subctxt(struct hfi1_ctxtdata *); -static int get_user_context(struct file *, struct hfi1_user_info *, int); -static int find_shared_ctxt(struct file *, const struct hfi1_user_info *); -static int allocate_ctxt(struct file *, struct hfi1_devdata *, - struct hfi1_user_info *); -static unsigned int poll_urgent(struct file *, struct poll_table_struct *); -static unsigned int poll_next(struct file *, struct poll_table_struct *); -static int user_event_ack(struct hfi1_ctxtdata *, int, unsigned long); -static int set_ctxt_pkey(struct hfi1_ctxtdata *, unsigned, u16); -static int manage_rcvq(struct hfi1_ctxtdata *, unsigned, int); -static int vma_fault(struct vm_area_struct *, struct vm_fault *); +static int hfi1_file_open(struct inode *inode, struct file *fp); +static int hfi1_file_close(struct inode *inode, struct file *fp); +static ssize_t hfi1_write_iter(struct kiocb *kiocb, struct iov_iter *from); +static unsigned int hfi1_poll(struct file *fp, struct poll_table_struct *pt); +static int hfi1_file_mmap(struct file *fp, struct vm_area_struct *vma); + +static u64 kvirt_to_phys(void *addr); +static int assign_ctxt(struct hfi1_filedata *fd, struct hfi1_user_info *uinfo); +static int init_subctxts(struct hfi1_ctxtdata *uctxt, + const struct hfi1_user_info *uinfo); +static int init_user_ctxt(struct hfi1_filedata *fd); +static void user_init(struct hfi1_ctxtdata *uctxt); +static int get_ctxt_info(struct hfi1_filedata *fd, void __user *ubase, + __u32 len); +static int get_base_info(struct hfi1_filedata *fd, void __user *ubase, + __u32 len); +static int setup_base_ctxt(struct hfi1_filedata *fd); +static int setup_subctxt(struct hfi1_ctxtdata *uctxt); + +static int find_sub_ctxt(struct hfi1_filedata *fd, + const struct hfi1_user_info *uinfo); +static int allocate_ctxt(struct hfi1_filedata *fd, struct hfi1_devdata *dd, + struct hfi1_user_info *uinfo); +static unsigned int poll_urgent(struct file *fp, struct poll_table_struct *pt); +static unsigned int poll_next(struct file *fp, struct poll_table_struct *pt); +static int user_event_ack(struct hfi1_ctxtdata *uctxt, u16 subctxt, + unsigned long events); +static int set_ctxt_pkey(struct hfi1_ctxtdata *uctxt, u16 subctxt, u16 pkey); +static int manage_rcvq(struct hfi1_ctxtdata *uctxt, u16 subctxt, + int start_stop); +static int vma_fault(struct vm_fault *vmf); static long hfi1_file_ioctl(struct file *fp, unsigned int cmd, unsigned long arg); @@ -172,6 +181,9 @@ static int hfi1_file_open(struct inode *inode, struct file *fp) struct hfi1_devdata, user_cdev); + if (!((dd->flags & HFI1_PRESENT) && dd->kregbase)) + return -EINVAL; + if (!atomic_inc_not_zero(&dd->user_refcount)) return -ENXIO; @@ -185,7 +197,8 @@ static int hfi1_file_open(struct inode *inode, struct file *fp) if (fd) { fd->rec_cpu_num = -1; /* no cpu affinity by default */ fd->mm = current->mm; - atomic_inc(&fd->mm->mm_count); + mmgrab(fd->mm); + fd->dd = dd; fp->private_data = fd; } else { fp->private_data = NULL; @@ -228,20 +241,14 @@ static long hfi1_file_ioctl(struct file *fp, unsigned int cmd, sizeof(uinfo))) return -EFAULT; - ret = assign_ctxt(fp, &uinfo); - if (ret < 0) - return ret; - ret = setup_ctxt(fp); - if (ret) - return ret; - ret = user_init(fp); + ret = assign_ctxt(fd, &uinfo); break; case HFI1_IOCTL_CTXT_INFO: - ret = get_ctxt_info(fp, (void __user *)(unsigned long)arg, + ret = get_ctxt_info(fd, (void __user *)(unsigned long)arg, sizeof(struct hfi1_ctxt_info)); break; case HFI1_IOCTL_USER_INFO: - ret = get_base_info(fp, (void __user *)(unsigned long)arg, + ret = get_base_info(fd, (void __user *)(unsigned long)arg, sizeof(struct hfi1_base_info)); break; case HFI1_IOCTL_CREDIT_UPD: @@ -255,7 +262,7 @@ static long hfi1_file_ioctl(struct file *fp, unsigned int cmd, sizeof(tinfo))) return -EFAULT; - ret = hfi1_user_exp_rcv_setup(fp, &tinfo); + ret = hfi1_user_exp_rcv_setup(fd, &tinfo); if (!ret) { /* * Copy the number of tidlist entries we used @@ -277,7 +284,7 @@ static long hfi1_file_ioctl(struct file *fp, unsigned int cmd, sizeof(tinfo))) return -EFAULT; - ret = hfi1_user_exp_rcv_clear(fp, &tinfo); + ret = hfi1_user_exp_rcv_clear(fd, &tinfo); if (ret) break; addr = arg + offsetof(struct hfi1_tid_info, tidcnt); @@ -292,7 +299,7 @@ static long hfi1_file_ioctl(struct file *fp, unsigned int cmd, sizeof(tinfo))) return -EFAULT; - ret = hfi1_user_exp_rcv_invalid(fp, &tinfo); + ret = hfi1_user_exp_rcv_invalid(fd, &tinfo); if (ret) break; addr = arg + offsetof(struct hfi1_tid_info, tidcnt); @@ -429,7 +436,7 @@ static ssize_t hfi1_write_iter(struct kiocb *kiocb, struct iov_iter *from) unsigned long count = 0; ret = hfi1_user_sdma_process_request( - kiocb->ki_filp, (struct iovec *)(from->iov + done), + fd, (struct iovec *)(from->iov + done), dim, &count); if (ret) { reqs = ret; @@ -585,8 +592,8 @@ static int hfi1_file_mmap(struct file *fp, struct vm_area_struct *vma) * knows where it's own bitmap is within the page. */ memaddr = (unsigned long)(dd->events + - ((uctxt->ctxt - dd->first_user_ctxt) * - HFI1_MAX_SHARED_CTXTS)) & PAGE_MASK; + ((uctxt->ctxt - dd->first_dyn_alloc_ctxt) * + HFI1_MAX_SHARED_CTXTS)) & PAGE_MASK; memlen = PAGE_SIZE; /* * v3.7 removes VM_RESERVED but the effect is kept by @@ -596,6 +603,10 @@ static int hfi1_file_mmap(struct file *fp, struct vm_area_struct *vma) vmf = 1; break; case STATUS: + if (flags & (unsigned long)(VM_WRITE | VM_EXEC)) { + ret = -EPERM; + goto done; + } memaddr = kvirt_to_phys((void *)dd->status); memlen = PAGE_SIZE; flags |= VM_IO | VM_DONTEXPAND; @@ -695,7 +706,7 @@ done: * Local (non-chip) user memory is not mapped right away but as it is * accessed by the user-level code. */ -static int vma_fault(struct vm_area_struct *vma, struct vm_fault *vmf) +static int vma_fault(struct vm_fault *vmf) { struct page *page; @@ -751,16 +762,19 @@ static int hfi1_file_close(struct inode *inode, struct file *fp) /* release the cpu */ hfi1_put_proc_affinity(fdata->rec_cpu_num); + /* clean up rcv side */ + hfi1_user_exp_rcv_free(fdata); + /* * Clear any left over, unhandled events so the next process that * gets this context doesn't get confused. */ - ev = dd->events + ((uctxt->ctxt - dd->first_user_ctxt) * + ev = dd->events + ((uctxt->ctxt - dd->first_dyn_alloc_ctxt) * HFI1_MAX_SHARED_CTXTS) + fdata->subctxt; *ev = 0; - if (--uctxt->cnt) { - uctxt->active_slaves &= ~(1 << fdata->subctxt); + __clear_bit(fdata->subctxt, uctxt->in_use_ctxts); + if (!bitmap_empty(uctxt->in_use_ctxts, HFI1_MAX_SHARED_CTXTS)) { mutex_unlock(&hfi1_mutex); goto done; } @@ -790,8 +804,8 @@ static int hfi1_file_close(struct inode *inode, struct file *fp) dd->rcd[uctxt->ctxt] = NULL; - hfi1_user_exp_rcv_free(fdata); - hfi1_clear_ctxt_pkey(dd, uctxt->ctxt); + hfi1_user_exp_rcv_grp_free(uctxt); + hfi1_clear_ctxt_pkey(dd, uctxt); uctxt->rcvwait_to = 0; uctxt->piowait_to = 0; @@ -831,121 +845,135 @@ static u64 kvirt_to_phys(void *addr) return paddr; } -static int assign_ctxt(struct file *fp, struct hfi1_user_info *uinfo) +static int assign_ctxt(struct hfi1_filedata *fd, struct hfi1_user_info *uinfo) { - int i_minor, ret = 0; + int ret; unsigned int swmajor, swminor; swmajor = uinfo->userversion >> 16; - if (swmajor != HFI1_USER_SWMAJOR) { - ret = -ENODEV; - goto done; - } + if (swmajor != HFI1_USER_SWMAJOR) + return -ENODEV; swminor = uinfo->userversion & 0xffff; mutex_lock(&hfi1_mutex); - /* First, lets check if we need to setup a shared context? */ + /* + * Get a sub context if necessary. + * ret < 0 error, 0 no context, 1 sub-context found + */ + ret = 0; if (uinfo->subctxt_cnt) { - struct hfi1_filedata *fd = fp->private_data; - - ret = find_shared_ctxt(fp, uinfo); - if (ret < 0) - goto done_unlock; - if (ret) { + ret = find_sub_ctxt(fd, uinfo); + if (ret > 0) fd->rec_cpu_num = hfi1_get_proc_affinity(fd->uctxt->numa_id); - } } /* - * We execute the following block if we couldn't find a - * shared context or if context sharing is not required. + * Allocate a base context if context sharing is not required or we + * couldn't find a sub context. */ - if (!ret) { - i_minor = iminor(file_inode(fp)) - HFI1_USER_MINOR_BASE; - ret = get_user_context(fp, uinfo, i_minor); - } -done_unlock: + if (!ret) + ret = allocate_ctxt(fd, fd->dd, uinfo); + mutex_unlock(&hfi1_mutex); -done: + + /* Depending on the context type, do the appropriate init */ + if (ret > 0) { + /* + * sub-context info can only be set up after the base + * context has been completed. + */ + ret = wait_event_interruptible(fd->uctxt->wait, !test_bit( + HFI1_CTXT_BASE_UNINIT, + &fd->uctxt->event_flags)); + if (test_bit(HFI1_CTXT_BASE_FAILED, &fd->uctxt->event_flags)) { + clear_bit(fd->subctxt, fd->uctxt->in_use_ctxts); + return -ENOMEM; + } + /* The only thing a sub context needs is the user_xxx stuff */ + if (!ret) + ret = init_user_ctxt(fd); + + if (ret) + clear_bit(fd->subctxt, fd->uctxt->in_use_ctxts); + } else if (!ret) { + ret = setup_base_ctxt(fd); + if (fd->uctxt->subctxt_cnt) { + /* If there is an error, set the failed bit. */ + if (ret) + set_bit(HFI1_CTXT_BASE_FAILED, + &fd->uctxt->event_flags); + /* + * Base context is done, notify anybody using a + * sub-context that is waiting for this completion + */ + clear_bit(HFI1_CTXT_BASE_UNINIT, + &fd->uctxt->event_flags); + wake_up(&fd->uctxt->wait); + } + } + return ret; } -static int get_user_context(struct file *fp, struct hfi1_user_info *uinfo, - int devno) +/* + * The hfi1_mutex must be held when this function is called. It is + * necessary to ensure serialized access to the bitmask in_use_ctxts. + */ +static int find_sub_ctxt(struct hfi1_filedata *fd, + const struct hfi1_user_info *uinfo) { - struct hfi1_devdata *dd = NULL; - int devmax, npresent, nup; + int i; + struct hfi1_devdata *dd = fd->dd; + u16 subctxt; - devmax = hfi1_count_units(&npresent, &nup); - if (!npresent) - return -ENXIO; + for (i = dd->first_dyn_alloc_ctxt; i < dd->num_rcv_contexts; i++) { + struct hfi1_ctxtdata *uctxt = dd->rcd[i]; - if (!nup) - return -ENETDOWN; + /* Skip ctxts which are not yet open */ + if (!uctxt || + bitmap_empty(uctxt->in_use_ctxts, + HFI1_MAX_SHARED_CTXTS)) + continue; - dd = hfi1_lookup(devno); - if (!dd) - return -ENODEV; - else if (!dd->freectxts) - return -EBUSY; + /* Skip dynamically allocted kernel contexts */ + if (uctxt->sc && (uctxt->sc->type == SC_KERNEL)) + continue; - return allocate_ctxt(fp, dd, uinfo); -} + /* Skip ctxt if it doesn't match the requested one */ + if (memcmp(uctxt->uuid, uinfo->uuid, + sizeof(uctxt->uuid)) || + uctxt->jkey != generate_jkey(current_uid()) || + uctxt->subctxt_id != uinfo->subctxt_id || + uctxt->subctxt_cnt != uinfo->subctxt_cnt) + continue; -static int find_shared_ctxt(struct file *fp, - const struct hfi1_user_info *uinfo) -{ - int devmax, ndev, i; - int ret = 0; - struct hfi1_filedata *fd = fp->private_data; + /* Verify the sharing process matches the master */ + if (uctxt->userversion != uinfo->userversion) + return -EINVAL; - devmax = hfi1_count_units(NULL, NULL); + /* Find an unused context */ + subctxt = find_first_zero_bit(uctxt->in_use_ctxts, + HFI1_MAX_SHARED_CTXTS); + if (subctxt >= uctxt->subctxt_cnt) + return -EBUSY; - for (ndev = 0; ndev < devmax; ndev++) { - struct hfi1_devdata *dd = hfi1_lookup(ndev); + fd->uctxt = uctxt; + fd->subctxt = subctxt; + __set_bit(fd->subctxt, uctxt->in_use_ctxts); - if (!(dd && (dd->flags & HFI1_PRESENT) && dd->kregbase)) - continue; - for (i = dd->first_user_ctxt; i < dd->num_rcv_contexts; i++) { - struct hfi1_ctxtdata *uctxt = dd->rcd[i]; - - /* Skip ctxts which are not yet open */ - if (!uctxt || !uctxt->cnt) - continue; - /* Skip ctxt if it doesn't match the requested one */ - if (memcmp(uctxt->uuid, uinfo->uuid, - sizeof(uctxt->uuid)) || - uctxt->jkey != generate_jkey(current_uid()) || - uctxt->subctxt_id != uinfo->subctxt_id || - uctxt->subctxt_cnt != uinfo->subctxt_cnt) - continue; - - /* Verify the sharing process matches the master */ - if (uctxt->userversion != uinfo->userversion || - uctxt->cnt >= uctxt->subctxt_cnt) { - ret = -EINVAL; - goto done; - } - fd->uctxt = uctxt; - fd->subctxt = uctxt->cnt++; - uctxt->active_slaves |= 1 << fd->subctxt; - ret = 1; - goto done; - } + return 1; } -done: - return ret; + return 0; } -static int allocate_ctxt(struct file *fp, struct hfi1_devdata *dd, +static int allocate_ctxt(struct hfi1_filedata *fd, struct hfi1_devdata *dd, struct hfi1_user_info *uinfo) { - struct hfi1_filedata *fd = fp->private_data; struct hfi1_ctxtdata *uctxt; - unsigned ctxt; + unsigned int ctxt; int ret, numa; if (dd->flags & HFI1_FROZEN) { @@ -959,7 +987,16 @@ static int allocate_ctxt(struct file *fp, struct hfi1_devdata *dd, return -EIO; } - for (ctxt = dd->first_user_ctxt; ctxt < dd->num_rcv_contexts; ctxt++) + /* + * This check is sort of redundant to the next EBUSY error. It would + * also indicate an inconsistancy in the driver if this value was + * zero, but there were still contexts available. + */ + if (!dd->freectxts) + return -EBUSY; + + for (ctxt = dd->first_dyn_alloc_ctxt; + ctxt < dd->num_rcv_contexts; ctxt++) if (!dd->rcd[ctxt]) break; @@ -1001,12 +1038,12 @@ static int allocate_ctxt(struct file *fp, struct hfi1_devdata *dd, goto ctxdata_free; /* - * Setup shared context resources if the user-level has requested - * shared contexts and this is the 'master' process. + * Setup sub context resources if the user-level has requested + * sub contexts. * This has to be done here so the rest of the sub-contexts find the * proper master. */ - if (uinfo->subctxt_cnt && !fd->subctxt) { + if (uinfo->subctxt_cnt) { ret = init_subctxts(uctxt, uinfo); /* * On error, we don't need to disable and de-allocate the @@ -1043,7 +1080,7 @@ ctxdata_free: static int init_subctxts(struct hfi1_ctxtdata *uctxt, const struct hfi1_user_info *uinfo) { - unsigned num_subctxts; + u16 num_subctxts; num_subctxts = uinfo->subctxt_cnt; if (num_subctxts > HFI1_MAX_SHARED_CTXTS) @@ -1051,9 +1088,8 @@ static int init_subctxts(struct hfi1_ctxtdata *uctxt, uctxt->subctxt_cnt = uinfo->subctxt_cnt; uctxt->subctxt_id = uinfo->subctxt_id; - uctxt->active_slaves = 1; uctxt->redirect_seq_cnt = 1; - set_bit(HFI1_CTXT_MASTER_UNINIT, &uctxt->event_flags); + set_bit(HFI1_CTXT_BASE_UNINIT, &uctxt->event_flags); return 0; } @@ -1061,13 +1097,12 @@ static int init_subctxts(struct hfi1_ctxtdata *uctxt, static int setup_subctxt(struct hfi1_ctxtdata *uctxt) { int ret = 0; - unsigned num_subctxts = uctxt->subctxt_cnt; + u16 num_subctxts = uctxt->subctxt_cnt; uctxt->subctxt_uregbase = vmalloc_user(PAGE_SIZE); - if (!uctxt->subctxt_uregbase) { - ret = -ENOMEM; - goto bail; - } + if (!uctxt->subctxt_uregbase) + return -ENOMEM; + /* We can take the size of the RcvHdr Queue from the master */ uctxt->subctxt_rcvhdr_base = vmalloc_user(uctxt->rcvhdrq_size * num_subctxts); @@ -1082,25 +1117,22 @@ static int setup_subctxt(struct hfi1_ctxtdata *uctxt) ret = -ENOMEM; goto bail_rhdr; } - goto bail; + + return 0; + bail_rhdr: vfree(uctxt->subctxt_rcvhdr_base); + uctxt->subctxt_rcvhdr_base = NULL; bail_ureg: vfree(uctxt->subctxt_uregbase); uctxt->subctxt_uregbase = NULL; -bail: + return ret; } -static int user_init(struct file *fp) +static void user_init(struct hfi1_ctxtdata *uctxt) { unsigned int rcvctrl_ops = 0; - struct hfi1_filedata *fd = fp->private_data; - struct hfi1_ctxtdata *uctxt = fd->uctxt; - - /* make sure that the context has already been setup */ - if (!test_bit(HFI1_CTXT_SETUP_DONE, &uctxt->event_flags)) - return -EFAULT; /* initialize poll variables... */ uctxt->urgent = 0; @@ -1148,20 +1180,12 @@ static int user_init(struct file *fp) else rcvctrl_ops |= HFI1_RCVCTRL_TAILUPD_DIS; hfi1_rcvctrl(uctxt->dd, rcvctrl_ops, uctxt->ctxt); - - /* Notify any waiting slaves */ - if (uctxt->subctxt_cnt) { - clear_bit(HFI1_CTXT_MASTER_UNINIT, &uctxt->event_flags); - wake_up(&uctxt->wait); - } - - return 0; } -static int get_ctxt_info(struct file *fp, void __user *ubase, __u32 len) +static int get_ctxt_info(struct hfi1_filedata *fd, void __user *ubase, + __u32 len) { struct hfi1_ctxt_info cinfo; - struct hfi1_filedata *fd = fp->private_data; struct hfi1_ctxtdata *uctxt = fd->uctxt; int ret = 0; @@ -1199,75 +1223,71 @@ static int get_ctxt_info(struct file *fp, void __user *ubase, __u32 len) return ret; } -static int setup_ctxt(struct file *fp) +static int init_user_ctxt(struct hfi1_filedata *fd) +{ + struct hfi1_ctxtdata *uctxt = fd->uctxt; + int ret; + + ret = hfi1_user_sdma_alloc_queues(uctxt, fd); + if (ret) + return ret; + + ret = hfi1_user_exp_rcv_init(fd); + + return ret; +} + +static int setup_base_ctxt(struct hfi1_filedata *fd) { - struct hfi1_filedata *fd = fp->private_data; struct hfi1_ctxtdata *uctxt = fd->uctxt; struct hfi1_devdata *dd = uctxt->dd; int ret = 0; - /* - * Context should be set up only once, including allocation and - * programming of eager buffers. This is done if context sharing - * is not requested or by the master process. - */ - if (!uctxt->subctxt_cnt || !fd->subctxt) { - ret = hfi1_init_ctxt(uctxt->sc); - if (ret) - goto done; + hfi1_init_ctxt(uctxt->sc); - /* Now allocate the RcvHdr queue and eager buffers. */ - ret = hfi1_create_rcvhdrq(dd, uctxt); - if (ret) - goto done; - ret = hfi1_setup_eagerbufs(uctxt); - if (ret) - goto done; - if (uctxt->subctxt_cnt && !fd->subctxt) { - ret = setup_subctxt(uctxt); - if (ret) - goto done; - } - } else { - ret = wait_event_interruptible(uctxt->wait, !test_bit( - HFI1_CTXT_MASTER_UNINIT, - &uctxt->event_flags)); - if (ret) - goto done; - } + /* Now allocate the RcvHdr queue and eager buffers. */ + ret = hfi1_create_rcvhdrq(dd, uctxt); + if (ret) + return ret; - ret = hfi1_user_sdma_alloc_queues(uctxt, fp); + ret = hfi1_setup_eagerbufs(uctxt); if (ret) - goto done; - /* - * Expected receive has to be setup for all processes (including - * shared contexts). However, it has to be done after the master - * context has been fully configured as it depends on the - * eager/expected split of the RcvArray entries. - * Setting it up here ensures that the subcontexts will be waiting - * (due to the above wait_event_interruptible() until the master - * is setup. - */ - ret = hfi1_user_exp_rcv_init(fp); + goto setup_failed; + + /* If sub-contexts are enabled, do the appropriate setup */ + if (uctxt->subctxt_cnt) + ret = setup_subctxt(uctxt); if (ret) - goto done; + goto setup_failed; - set_bit(HFI1_CTXT_SETUP_DONE, &uctxt->event_flags); -done: + ret = hfi1_user_exp_rcv_grp_init(fd); + if (ret) + goto setup_failed; + + ret = init_user_ctxt(fd); + if (ret) + goto setup_failed; + + user_init(uctxt); + + return 0; + +setup_failed: + hfi1_free_ctxtdata(dd, uctxt); return ret; } -static int get_base_info(struct file *fp, void __user *ubase, __u32 len) +static int get_base_info(struct hfi1_filedata *fd, void __user *ubase, + __u32 len) { struct hfi1_base_info binfo; - struct hfi1_filedata *fd = fp->private_data; struct hfi1_ctxtdata *uctxt = fd->uctxt; struct hfi1_devdata *dd = uctxt->dd; ssize_t sz; unsigned offset; int ret = 0; - trace_hfi1_uctxtdata(uctxt->dd, uctxt); + trace_hfi1_uctxtdata(uctxt->dd, uctxt, fd->subctxt); memset(&binfo, 0, sizeof(binfo)); binfo.hw_version = dd->revision; @@ -1305,7 +1325,7 @@ static int get_base_info(struct file *fp, void __user *ubase, __u32 len) */ binfo.user_regbase = HFI1_MMAP_TOKEN(UREGS, uctxt->ctxt, fd->subctxt, 0); - offset = offset_in_page((((uctxt->ctxt - dd->first_user_ctxt) * + offset = offset_in_page((((uctxt->ctxt - dd->first_dyn_alloc_ctxt) * HFI1_MAX_SHARED_CTXTS) + fd->subctxt) * sizeof(*dd->events)); binfo.events_bufbase = HFI1_MMAP_TOKEN(EVENTS, uctxt->ctxt, @@ -1399,12 +1419,12 @@ int hfi1_set_uevent_bits(struct hfi1_pportdata *ppd, const int evtbit) } spin_lock_irqsave(&dd->uctxt_lock, flags); - for (ctxt = dd->first_user_ctxt; ctxt < dd->num_rcv_contexts; + for (ctxt = dd->first_dyn_alloc_ctxt; ctxt < dd->num_rcv_contexts; ctxt++) { uctxt = dd->rcd[ctxt]; if (uctxt) { unsigned long *evs = dd->events + - (uctxt->ctxt - dd->first_user_ctxt) * + (uctxt->ctxt - dd->first_dyn_alloc_ctxt) * HFI1_MAX_SHARED_CTXTS; int i; /* @@ -1431,7 +1451,7 @@ done: * overflow conditions. start_stop==1 re-enables, to be used to * re-init the software copy of the head register */ -static int manage_rcvq(struct hfi1_ctxtdata *uctxt, unsigned subctxt, +static int manage_rcvq(struct hfi1_ctxtdata *uctxt, u16 subctxt, int start_stop) { struct hfi1_devdata *dd = uctxt->dd; @@ -1466,7 +1486,7 @@ bail: * User process then performs actions appropriate to bit having been * set, if desired, and checks again in future. */ -static int user_event_ack(struct hfi1_ctxtdata *uctxt, int subctxt, +static int user_event_ack(struct hfi1_ctxtdata *uctxt, u16 subctxt, unsigned long events) { int i; @@ -1476,7 +1496,7 @@ static int user_event_ack(struct hfi1_ctxtdata *uctxt, int subctxt, if (!dd->events) return 0; - evs = dd->events + ((uctxt->ctxt - dd->first_user_ctxt) * + evs = dd->events + ((uctxt->ctxt - dd->first_dyn_alloc_ctxt) * HFI1_MAX_SHARED_CTXTS) + subctxt; for (i = 0; i <= _HFI1_MAX_EVENT_BIT; i++) { @@ -1487,8 +1507,7 @@ static int user_event_ack(struct hfi1_ctxtdata *uctxt, int subctxt, return 0; } -static int set_ctxt_pkey(struct hfi1_ctxtdata *uctxt, unsigned subctxt, - u16 pkey) +static int set_ctxt_pkey(struct hfi1_ctxtdata *uctxt, u16 subctxt, u16 pkey) { int ret = -ENOENT, i, intable = 0; struct hfi1_pportdata *ppd = uctxt->ppd; diff --git a/drivers/infiniband/hw/hfi1/firmware.c b/drivers/infiniband/hw/hfi1/firmware.c index 0dd50cdb039a..4042c11b2742 100644 --- a/drivers/infiniband/hw/hfi1/firmware.c +++ b/drivers/infiniband/hw/hfi1/firmware.c @@ -1,5 +1,5 @@ /* - * Copyright(c) 2015, 2016 Intel Corporation. + * Copyright(c) 2015 - 2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -1004,7 +1004,9 @@ static int load_8051_firmware(struct hfi1_devdata *dd, { u64 reg; int ret; - u8 ver_a, ver_b; + u8 ver_major; + u8 ver_minor; + u8 ver_patch; /* * DC Reset sequence @@ -1073,10 +1075,10 @@ static int load_8051_firmware(struct hfi1_devdata *dd, return -ETIMEDOUT; } - read_misc_status(dd, &ver_a, &ver_b); - dd_dev_info(dd, "8051 firmware version %d.%d\n", - (int)ver_b, (int)ver_a); - dd->dc8051_ver = dc8051_ver(ver_b, ver_a); + 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); return 0; } diff --git a/drivers/infiniband/hw/hfi1/hfi.h b/drivers/infiniband/hw/hfi1/hfi.h index 751a0fb29fa5..414a04a481c2 100644 --- a/drivers/infiniband/hw/hfi1/hfi.h +++ b/drivers/infiniband/hw/hfi1/hfi.h @@ -1,7 +1,7 @@ #ifndef _HFI1_KERNEL_H #define _HFI1_KERNEL_H /* - * Copyright(c) 2015, 2016 Intel Corporation. + * Copyright(c) 2015-2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -54,6 +54,7 @@ #include <linux/list.h> #include <linux/scatterlist.h> #include <linux/slab.h> +#include <linux/idr.h> #include <linux/io.h> #include <linux/fs.h> #include <linux/completion.h> @@ -66,6 +67,7 @@ #include <linux/i2c-algo-bit.h> #include <rdma/ib_hdrs.h> #include <linux/rhashtable.h> +#include <linux/netdevice.h> #include <rdma/rdma_vt.h> #include "chip_registers.h" @@ -194,12 +196,6 @@ struct hfi1_ctxtdata { void *rcvhdrq; /* kernel virtual address where hdrqtail is updated */ volatile __le64 *rcvhdrtail_kvaddr; - /* - * Shared page for kernel to signal user processes that send buffers - * need disarming. The process should call HFI1_CMD_DISARM_BUFS - * or HFI1_CMD_ACK_EVENT with IPATH_EVENT_DISARM_BUFS set. - */ - unsigned long *user_event_mask; /* when waiting for rcv or pioavail */ wait_queue_head_t wait; /* rcvhdrq size (for freeing) */ @@ -222,13 +218,12 @@ struct hfi1_ctxtdata { * (ignoring forks, dup, etc. for now) */ int cnt; + /* Device context index */ + unsigned ctxt; /* - * how much space to leave at start of eager TID entries for - * protocol use, on each TID + * non-zero if ctxt can be shared, and defines the maximum number of + * sub-contexts for this device context. */ - /* instead of calculating it */ - unsigned ctxt; - /* non-zero if ctxt is being shared. */ u16 subctxt_cnt; /* non-zero if ctxt is being shared. */ u16 subctxt_id; @@ -278,16 +273,18 @@ struct hfi1_ctxtdata { struct hfi1_devdata *dd; /* so functions that need physical port can get it easily */ struct hfi1_pportdata *ppd; + /* associated msix interrupt */ + u32 msix_intr; /* A page of memory for rcvhdrhead, rcvegrhead, rcvegrtail * N */ void *subctxt_uregbase; /* An array of pages for the eager receive buffers * N */ void *subctxt_rcvegrbuf; /* An array of pages for the eager header queue entries * N */ void *subctxt_rcvhdr_base; + /* Bitmask of in use context(s) */ + DECLARE_BITMAP(in_use_ctxts, HFI1_MAX_SHARED_CTXTS); /* The version of the library which opened this ctxt */ u32 userversion; - /* Bitmask of active slaves */ - u32 active_slaves; /* Type of packets or conditions we want to poll for */ u16 poll_type; /* receive packet sequence counter */ @@ -337,6 +334,12 @@ struct hfi1_ctxtdata { * packets with the wrong interrupt handler. */ int (*do_interrupt)(struct hfi1_ctxtdata *rcd, int threaded); + + /* Indicates that this is vnic context */ + bool is_vnic; + + /* vnic queue index this context is mapped to */ + u8 vnic_q_idx; }; /* @@ -356,12 +359,11 @@ struct hfi1_packet { u64 rhf; u32 maxcnt; u32 rhqoff; - u32 hdrqtail; - int numpkt; u16 tlen; - u16 hlen; s16 etail; - u16 rsize; + u8 hlen; + u8 numpkt; + u8 rsize; u8 updegr; u8 rcv_flags; u8 etype; @@ -475,7 +477,7 @@ struct rvt_sge_state; #define HFI1_PART_ENFORCE_OUT 0x2 /* how often we check for synthetic counter wrap around */ -#define SYNTH_CNT_TIME 2 +#define SYNTH_CNT_TIME 3 /* Counter flags */ #define CNTR_NORMAL 0x0 /* Normal counters, just read register */ @@ -809,6 +811,32 @@ struct hfi1_asic_data { struct hfi1_i2c_bus *i2c_bus1; }; +/* sizes for both the QP and RSM map tables */ +#define NUM_MAP_ENTRIES 256 +#define NUM_MAP_REGS 32 + +/* + * Number of VNIC contexts used. Ensure it is less than or equal to + * max queues supported by VNIC (HFI1_VNIC_MAX_QUEUE). + */ +#define HFI1_NUM_VNIC_CTXT 8 + +/* Number of VNIC RSM entries */ +#define NUM_VNIC_MAP_ENTRIES 8 + +/* Virtual NIC information */ +struct hfi1_vnic_data { + struct hfi1_ctxtdata *ctxt[HFI1_NUM_VNIC_CTXT]; + struct kmem_cache *txreq_cache; + u8 num_vports; + struct idr vesw_idr; + u8 rmt_start; + u8 num_ctxt; + u32 msix_idx; +}; + +struct hfi1_vnic_vport_info; + /* device data struct now contains only "general per-device" info. * fields related to a physical IB port are in a hfi1_pportdata struct. */ @@ -927,8 +955,9 @@ struct hfi1_devdata { spinlock_t rcvctrl_lock; /* protect changes to RcvCtrl */ /* around rcd and (user ctxts) ctxt_cnt use (intr vs free) */ spinlock_t uctxt_lock; /* rcd and user context changes */ - /* exclusive access to 8051 */ - spinlock_t dc8051_lock; + struct mutex dc8051_lock; /* exclusive access to 8051 */ + struct workqueue_struct *update_cntr_wq; + struct work_struct update_cntr_work; /* exclusive access to 8051 memory */ spinlock_t dc8051_memlock; int dc8051_timed_out; /* remember if the 8051 timed out */ @@ -1016,12 +1045,20 @@ struct hfi1_devdata { /* initial vl15 credits to use */ u16 vl15_init; + /* + * Cached value for vl15buf, read during verify cap interrupt. VL15 + * credits are to be kept at 0 and set when handling the link-up + * interrupt. This removes the possibility of receiving VL15 MAD + * packets before this HFI is ready. + */ + u16 vl15buf_cached; + /* Misc small ints */ u8 n_krcv_queues; u8 qos_shift; u16 irev; /* implementation revision */ - u16 dc8051_ver; /* 8051 firmware version */ + u32 dc8051_ver; /* 8051 firmware version */ spinlock_t hfi1_diag_trans_lock; /* protect diag observer ops */ struct platform_config platform_config; @@ -1032,6 +1069,7 @@ struct hfi1_devdata { /* MSI-X information */ struct hfi1_msix_entry *msix_entries; u32 num_msix_entries; + u32 first_dyn_msix_idx; /* INTx information */ u32 requested_intx_irq; /* did we request one? */ @@ -1116,6 +1154,9 @@ struct hfi1_devdata { send_routine process_dma_send; void (*pio_inline_send)(struct hfi1_devdata *dd, struct pio_buf *pbuf, u64 pbc, const void *from, size_t count); + int (*process_vnic_dma_send)(struct hfi1_devdata *dd, u8 q_idx, + struct hfi1_vnic_vport_info *vinfo, + struct sk_buff *skb, u64 pbc, u8 plen); /* hfi1_pportdata, points to array of (physical) port-specific * data structs, indexed by pidx (0..n-1) */ @@ -1127,8 +1168,8 @@ struct hfi1_devdata { u16 flags; /* Number of physical ports available */ u8 num_pports; - /* Lowest context number which can be used by user processes */ - u8 first_user_ctxt; + /* Lowest context number which can be used by user processes or VNIC */ + u8 first_dyn_alloc_ctxt; /* adding a new field here would make it part of this cacheline */ /* seqlock for sc2vl */ @@ -1168,15 +1209,24 @@ struct hfi1_devdata { bool eprom_available; /* true if EPROM is available for this device */ bool aspm_supported; /* Does HW support ASPM */ bool aspm_enabled; /* ASPM state: enabled/disabled */ - struct rhashtable sdma_rht; + struct rhashtable *sdma_rht; struct kobject kobj; + + /* vnic data */ + struct hfi1_vnic_data vnic; }; +static inline bool hfi1_vnic_is_rsm_full(struct hfi1_devdata *dd, int spare) +{ + return (dd->vnic.rmt_start + spare) > NUM_MAP_ENTRIES; +} + /* 8051 firmware version helper */ -#define dc8051_ver(a, b) ((a) << 8 | (b)) -#define dc8051_ver_maj(a) ((a & 0xff00) >> 8) -#define dc8051_ver_min(a) (a & 0x00ff) +#define dc8051_ver(a, b, c) ((a) << 16 | (b) << 8 | (c)) +#define dc8051_ver_maj(a) (((a) & 0xff0000) >> 16) +#define dc8051_ver_min(a) (((a) & 0x00ff00) >> 8) +#define dc8051_ver_patch(a) ((a) & 0x0000ff) /* f_put_tid types */ #define PT_EXPECTED 0 @@ -1189,10 +1239,11 @@ struct mmu_rb_handler; /* Private data for file operations */ struct hfi1_filedata { + struct hfi1_devdata *dd; struct hfi1_ctxtdata *uctxt; - unsigned subctxt; struct hfi1_user_sdma_comp_q *cq; struct hfi1_user_sdma_pkt_q *pq; + u16 subctxt; /* for cpu affinity; -1 if none */ int rec_cpu_num; u32 tid_n_pinned; @@ -1214,28 +1265,31 @@ struct hfi1_devdata *hfi1_lookup(int unit); extern u32 hfi1_cpulist_count; extern unsigned long *hfi1_cpulist; -int hfi1_init(struct hfi1_devdata *, int); -int hfi1_count_units(int *npresentp, int *nupp); +int hfi1_init(struct hfi1_devdata *dd, int reinit); int hfi1_count_active_units(void); -int hfi1_diag_add(struct hfi1_devdata *); -void hfi1_diag_remove(struct hfi1_devdata *); +int hfi1_diag_add(struct hfi1_devdata *dd); +void hfi1_diag_remove(struct hfi1_devdata *dd); void handle_linkup_change(struct hfi1_devdata *dd, u32 linkup); void handle_user_interrupt(struct hfi1_ctxtdata *rcd); -int hfi1_create_rcvhdrq(struct hfi1_devdata *, struct hfi1_ctxtdata *); -int hfi1_setup_eagerbufs(struct hfi1_ctxtdata *); +int hfi1_create_rcvhdrq(struct hfi1_devdata *dd, struct hfi1_ctxtdata *rcd); +int hfi1_setup_eagerbufs(struct hfi1_ctxtdata *rcd); int hfi1_create_ctxts(struct hfi1_devdata *dd); -struct hfi1_ctxtdata *hfi1_create_ctxtdata(struct hfi1_pportdata *, u32, int); -void hfi1_init_pportdata(struct pci_dev *, struct hfi1_pportdata *, - struct hfi1_devdata *, u8, u8); -void hfi1_free_ctxtdata(struct hfi1_devdata *, struct hfi1_ctxtdata *); - -int handle_receive_interrupt(struct hfi1_ctxtdata *, int); -int handle_receive_interrupt_nodma_rtail(struct hfi1_ctxtdata *, int); -int handle_receive_interrupt_dma_rtail(struct hfi1_ctxtdata *, int); +struct hfi1_ctxtdata *hfi1_create_ctxtdata(struct hfi1_pportdata *ppd, u32 ctxt, + int numa); +void hfi1_init_pportdata(struct pci_dev *pdev, struct hfi1_pportdata *ppd, + struct hfi1_devdata *dd, u8 hw_pidx, u8 port); +void hfi1_free_ctxtdata(struct hfi1_devdata *dd, struct hfi1_ctxtdata *rcd); + +int handle_receive_interrupt(struct hfi1_ctxtdata *rcd, int thread); +int handle_receive_interrupt_nodma_rtail(struct hfi1_ctxtdata *rcd, int thread); +int handle_receive_interrupt_dma_rtail(struct hfi1_ctxtdata *rcd, int thread); void set_all_slowpath(struct hfi1_devdata *dd); +void hfi1_vnic_synchronize_irq(struct hfi1_devdata *dd); +void hfi1_set_vnic_msix_info(struct hfi1_ctxtdata *rcd); +void hfi1_reset_vnic_msix_info(struct hfi1_ctxtdata *rcd); extern const struct pci_device_id hfi1_pci_tbl[]; @@ -1255,16 +1309,24 @@ int hfi1_reset_device(int); /* return the driver's idea of the logical OPA port state */ static inline u32 driver_lstate(struct hfi1_pportdata *ppd) { - return ppd->lstate; /* use the cached value */ + /* + * The driver does some processing from the time the logical + * link state is at INIT to the time the SM can be notified + * as such. Return IB_PORT_DOWN until the software state + * is ready. + */ + if (ppd->lstate == IB_PORT_INIT && !(ppd->host_link_state & HLS_UP)) + return IB_PORT_DOWN; + else + return ppd->lstate; } void receive_interrupt_work(struct work_struct *work); /* extract service channel from header and rhf */ -static inline int hdr2sc(struct ib_header *hdr, u64 rhf) +static inline int hfi1_9B_get_sc5(struct ib_header *hdr, u64 rhf) { - return ((be16_to_cpu(hdr->lrh[0]) >> 12) & 0xf) | - ((!!(rhf_dc_info(rhf))) << 4); + return ib_get_sc(hdr) | ((!!(rhf_dc_info(rhf))) << 4); } #define HFI1_JKEY_WIDTH 16 @@ -1520,7 +1582,7 @@ bad: u32 lrh_max_header_bytes(struct hfi1_devdata *dd); int mtu_to_enum(u32 mtu, int default_if_bad); -u16 enum_to_mtu(int); +u16 enum_to_mtu(int mtu); static inline int valid_ib_mtu(unsigned int mtu) { return mtu == 256 || mtu == 512 || @@ -1534,17 +1596,18 @@ static inline int valid_opa_max_mtu(unsigned int mtu) (valid_ib_mtu(mtu) || mtu == 8192 || mtu == 10240); } -int set_mtu(struct hfi1_pportdata *); +int set_mtu(struct hfi1_pportdata *ppd); -int hfi1_set_lid(struct hfi1_pportdata *, u32, u8); -void hfi1_disable_after_error(struct hfi1_devdata *); -int hfi1_set_uevent_bits(struct hfi1_pportdata *, const int); -int hfi1_rcvbuf_validate(u32, u8, u16 *); +int hfi1_set_lid(struct hfi1_pportdata *ppd, u32 lid, u8 lmc); +void hfi1_disable_after_error(struct hfi1_devdata *dd); +int hfi1_set_uevent_bits(struct hfi1_pportdata *ppd, const int evtbit); +int hfi1_rcvbuf_validate(u32 size, u8 type, u16 *encode); -int fm_get_table(struct hfi1_pportdata *, int, void *); -int fm_set_table(struct hfi1_pportdata *, int, void *); +int fm_get_table(struct hfi1_pportdata *ppd, int which, void *t); +int fm_set_table(struct hfi1_pportdata *ppd, int which, void *t); -void set_up_vl15(struct hfi1_devdata *dd, u8 vau, u16 vl15buf); +void set_up_vau(struct hfi1_devdata *dd, u8 vau); +void set_up_vl15(struct hfi1_devdata *dd, u16 vl15buf); void reset_link_credits(struct hfi1_devdata *dd); void assign_remote_cm_au_table(struct hfi1_devdata *dd, u8 vcu); @@ -1584,6 +1647,11 @@ static inline struct hfi1_ibport *to_iport(struct ib_device *ibdev, u8 port) return &dd->pport[pidx].ibport_data; } +static inline struct hfi1_ibport *rcd_to_iport(struct hfi1_ctxtdata *rcd) +{ + return &rcd->ppd->ibport_data; +} + void hfi1_process_ecn_slowpath(struct rvt_qp *qp, struct hfi1_packet *pkt, bool do_cnp); static inline bool process_ecn(struct rvt_qp *qp, struct hfi1_packet *pkt, @@ -1593,9 +1661,9 @@ static inline bool process_ecn(struct rvt_qp *qp, struct hfi1_packet *pkt, u32 bth1; bth1 = be32_to_cpu(ohdr->bth[1]); - if (unlikely(bth1 & (HFI1_BECN_SMASK | HFI1_FECN_SMASK))) { + if (unlikely(bth1 & (IB_BECN_SMASK | IB_FECN_SMASK))) { hfi1_process_ecn_slowpath(qp, pkt, do_cnp); - return bth1 & HFI1_FECN_SMASK; + return !!(bth1 & IB_FECN_SMASK); } return false; } @@ -1659,19 +1727,19 @@ struct cc_state *get_cc_state_protected(struct hfi1_pportdata *ppd) #define HFI1_PBC_LENGTH_MASK ((1 << 11) - 1) /* ctxt_flag bit offsets */ - /* context has been setup */ -#define HFI1_CTXT_SETUP_DONE 1 + /* base context has not finished initializing */ +#define HFI1_CTXT_BASE_UNINIT 1 + /* base context initaliation failed */ +#define HFI1_CTXT_BASE_FAILED 2 /* waiting for a packet to arrive */ -#define HFI1_CTXT_WAITING_RCV 2 - /* master has not finished initializing */ -#define HFI1_CTXT_MASTER_UNINIT 4 +#define HFI1_CTXT_WAITING_RCV 3 /* waiting for an urgent packet to arrive */ -#define HFI1_CTXT_WAITING_URG 5 +#define HFI1_CTXT_WAITING_URG 4 /* free up any allocated data at closes */ -struct hfi1_devdata *hfi1_init_dd(struct pci_dev *, - const struct pci_device_id *); -void hfi1_free_devdata(struct hfi1_devdata *); +struct hfi1_devdata *hfi1_init_dd(struct pci_dev *pdev, + const struct pci_device_id *ent); +void hfi1_free_devdata(struct hfi1_devdata *dd); struct hfi1_devdata *hfi1_alloc_devdata(struct pci_dev *pdev, size_t extra); /* LED beaconing functions */ @@ -1746,24 +1814,24 @@ static inline u32 get_rcvhdrtail(const struct hfi1_ctxtdata *rcd) extern const char ib_hfi1_version[]; -int hfi1_device_create(struct hfi1_devdata *); -void hfi1_device_remove(struct hfi1_devdata *); +int hfi1_device_create(struct hfi1_devdata *dd); +void hfi1_device_remove(struct hfi1_devdata *dd); int hfi1_create_port_files(struct ib_device *ibdev, u8 port_num, struct kobject *kobj); -int hfi1_verbs_register_sysfs(struct hfi1_devdata *); -void hfi1_verbs_unregister_sysfs(struct hfi1_devdata *); +int hfi1_verbs_register_sysfs(struct hfi1_devdata *dd); +void hfi1_verbs_unregister_sysfs(struct hfi1_devdata *dd); /* Hook for sysfs read of QSFP */ int qsfp_dump(struct hfi1_pportdata *ppd, char *buf, int len); -int hfi1_pcie_init(struct pci_dev *, const struct pci_device_id *); -void hfi1_pcie_cleanup(struct pci_dev *); -int hfi1_pcie_ddinit(struct hfi1_devdata *, struct pci_dev *); +int hfi1_pcie_init(struct pci_dev *pdev, const struct pci_device_id *ent); +void hfi1_pcie_cleanup(struct pci_dev *pdev); +int hfi1_pcie_ddinit(struct hfi1_devdata *dd, struct pci_dev *pdev); void hfi1_pcie_ddcleanup(struct hfi1_devdata *); -void hfi1_pcie_flr(struct hfi1_devdata *); -int pcie_speeds(struct hfi1_devdata *); -void request_msix(struct hfi1_devdata *, u32 *, struct hfi1_msix_entry *); -void hfi1_enable_intx(struct pci_dev *); +int pcie_speeds(struct hfi1_devdata *dd); +void request_msix(struct hfi1_devdata *dd, u32 *nent, + struct hfi1_msix_entry *entry); +void hfi1_enable_intx(struct pci_dev *pdev); void restore_pci_variables(struct hfi1_devdata *dd); int do_pcie_gen3_transition(struct hfi1_devdata *dd); int parse_platform_config(struct hfi1_devdata *dd); @@ -1793,8 +1861,6 @@ int kdeth_process_expected(struct hfi1_packet *packet); int kdeth_process_eager(struct hfi1_packet *packet); int process_receive_invalid(struct hfi1_packet *packet); -void update_sge(struct rvt_sge_state *ss, u32 length); - /* global module parameter variables */ extern unsigned int hfi1_max_mtu; extern unsigned int hfi1_cu; @@ -1940,6 +2006,10 @@ static inline u64 hfi1_pkt_base_sdma_integrity(struct hfi1_devdata *dd) dev_info(&(dd)->pcidev->dev, "%s: " fmt, \ get_unit_name((dd)->unit), ##__VA_ARGS__) +#define dd_dev_info_ratelimited(dd, fmt, ...) \ + dev_info_ratelimited(&(dd)->pcidev->dev, "%s: " fmt, \ + get_unit_name((dd)->unit), ##__VA_ARGS__) + #define dd_dev_dbg(dd, fmt, ...) \ dev_dbg(&(dd)->pcidev->dev, "%s: " fmt, \ get_unit_name((dd)->unit), ##__VA_ARGS__) diff --git a/drivers/infiniband/hw/hfi1/init.c b/drivers/infiniband/hw/hfi1/init.c index e3b5bc93bc70..4a11d4da4c92 100644 --- a/drivers/infiniband/hw/hfi1/init.c +++ b/drivers/infiniband/hw/hfi1/init.c @@ -1,5 +1,5 @@ /* - * Copyright(c) 2015, 2016 Intel Corporation. + * Copyright(c) 2015-2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -53,6 +53,7 @@ #include <linux/module.h> #include <linux/printk.h> #include <linux/hrtimer.h> +#include <linux/bitmap.h> #include <rdma/rdma_vt.h> #include "hfi.h" @@ -65,10 +66,12 @@ #include "verbs.h" #include "aspm.h" #include "affinity.h" +#include "vnic.h" #undef pr_fmt #define pr_fmt(fmt) DRIVER_NAME ": " fmt +#define HFI1_MAX_ACTIVE_WORKQUEUE_ENTRIES 5 /* * min buffers we want to have per context, after driver */ @@ -100,9 +103,9 @@ static unsigned hfi1_rcvarr_split = 25; module_param_named(rcvarr_split, hfi1_rcvarr_split, uint, S_IRUGO); MODULE_PARM_DESC(rcvarr_split, "Percent of context's RcvArray entries used for Eager buffers"); -static uint eager_buffer_size = (2 << 20); /* 2MB */ +static uint eager_buffer_size = (8 << 20); /* 8MB */ module_param(eager_buffer_size, uint, S_IRUGO); -MODULE_PARM_DESC(eager_buffer_size, "Size of the eager buffers, default: 2MB"); +MODULE_PARM_DESC(eager_buffer_size, "Size of the eager buffers, default: 8MB"); static uint rcvhdrcnt = 2048; /* 2x the max eager buffer count */ module_param_named(rcvhdrcnt, rcvhdrcnt, uint, S_IRUGO); @@ -116,7 +119,7 @@ unsigned int user_credit_return_threshold = 33; /* default is 33% */ module_param(user_credit_return_threshold, uint, S_IRUGO); MODULE_PARM_DESC(user_credit_return_threshold, "Credit return threshold for user send contexts, return when unreturned credits passes this many blocks (in percent of allocated blocks, 0 is off)"); -static inline u64 encode_rcv_header_entry_size(u16); +static inline u64 encode_rcv_header_entry_size(u16 size); static struct idr hfi1_unit_table; u32 hfi1_cpulist_count; @@ -139,7 +142,7 @@ int hfi1_create_ctxts(struct hfi1_devdata *dd) goto nomem; /* create one or more kernel contexts */ - for (i = 0; i < dd->first_user_ctxt; ++i) { + for (i = 0; i < dd->first_dyn_alloc_ctxt; ++i) { struct hfi1_pportdata *ppd; struct hfi1_ctxtdata *rcd; @@ -174,13 +177,7 @@ int hfi1_create_ctxts(struct hfi1_devdata *dd) goto nomem; } - ret = hfi1_init_ctxt(rcd->sc); - if (ret < 0) { - dd_dev_err(dd, - "Failed to setup kernel receive context, failing\n"); - ret = -EFAULT; - goto bail; - } + hfi1_init_ctxt(rcd->sc); } /* @@ -192,7 +189,7 @@ int hfi1_create_ctxts(struct hfi1_devdata *dd) return 0; nomem: ret = -ENOMEM; -bail: + if (dd->rcd) { for (i = 0; i < dd->num_rcv_contexts; ++i) hfi1_free_ctxtdata(dd, dd->rcd[i]); @@ -214,9 +211,9 @@ struct hfi1_ctxtdata *hfi1_create_ctxtdata(struct hfi1_pportdata *ppd, u32 ctxt, u32 base; if (dd->rcv_entries.nctxt_extra > - dd->num_rcv_contexts - dd->first_user_ctxt) + dd->num_rcv_contexts - dd->first_dyn_alloc_ctxt) kctxt_ngroups = (dd->rcv_entries.nctxt_extra - - (dd->num_rcv_contexts - dd->first_user_ctxt)); + (dd->num_rcv_contexts - dd->first_dyn_alloc_ctxt)); rcd = kzalloc_node(sizeof(*rcd), GFP_KERNEL, numa); if (rcd) { u32 rcvtids, max_entries; @@ -226,7 +223,7 @@ struct hfi1_ctxtdata *hfi1_create_ctxtdata(struct hfi1_pportdata *ppd, u32 ctxt, INIT_LIST_HEAD(&rcd->qp_wait_list); rcd->ppd = ppd; rcd->dd = dd; - rcd->cnt = 1; + __set_bit(0, rcd->in_use_ctxts); rcd->ctxt = ctxt; dd->rcd[ctxt] = rcd; rcd->numa_id = numa; @@ -238,27 +235,29 @@ struct hfi1_ctxtdata *hfi1_create_ctxtdata(struct hfi1_pportdata *ppd, u32 ctxt, * Calculate the context's RcvArray entry starting point. * We do this here because we have to take into account all * the RcvArray entries that previous context would have - * taken and we have to account for any extra groups - * assigned to the kernel or user contexts. + * taken and we have to account for any extra groups assigned + * to the static (kernel) or dynamic (vnic/user) contexts. */ - if (ctxt < dd->first_user_ctxt) { + if (ctxt < dd->first_dyn_alloc_ctxt) { if (ctxt < kctxt_ngroups) { base = ctxt * (dd->rcv_entries.ngroups + 1); rcd->rcv_array_groups++; - } else + } else { base = kctxt_ngroups + (ctxt * dd->rcv_entries.ngroups); + } } else { - u16 ct = ctxt - dd->first_user_ctxt; + u16 ct = ctxt - dd->first_dyn_alloc_ctxt; base = ((dd->n_krcv_queues * dd->rcv_entries.ngroups) + kctxt_ngroups); if (ct < dd->rcv_entries.nctxt_extra) { base += ct * (dd->rcv_entries.ngroups + 1); rcd->rcv_array_groups++; - } else + } else { base += dd->rcv_entries.nctxt_extra + (ct * dd->rcv_entries.ngroups); + } } rcd->eager_base = base * dd->rcv_entries.group_size; @@ -297,14 +296,15 @@ struct hfi1_ctxtdata *hfi1_create_ctxtdata(struct hfi1_pportdata *ppd, u32 ctxt, * The resulting value will be rounded down to the closest * multiple of dd->rcv_entries.group_size. */ - rcd->egrbufs.buffers = kcalloc(rcd->egrbufs.count, - sizeof(*rcd->egrbufs.buffers), - GFP_KERNEL); + rcd->egrbufs.buffers = kzalloc_node( + rcd->egrbufs.count * sizeof(*rcd->egrbufs.buffers), + GFP_KERNEL, numa); if (!rcd->egrbufs.buffers) goto bail; - rcd->egrbufs.rcvtids = kcalloc(rcd->egrbufs.count, - sizeof(*rcd->egrbufs.rcvtids), - GFP_KERNEL); + rcd->egrbufs.rcvtids = kzalloc_node( + rcd->egrbufs.count * + sizeof(*rcd->egrbufs.rcvtids), + GFP_KERNEL, numa); if (!rcd->egrbufs.rcvtids) goto bail; rcd->egrbufs.size = eager_buffer_size; @@ -321,9 +321,10 @@ struct hfi1_ctxtdata *hfi1_create_ctxtdata(struct hfi1_pportdata *ppd, u32 ctxt, } rcd->egrbufs.rcvtid_size = HFI1_MAX_EAGER_BUFFER_SIZE; - if (ctxt < dd->first_user_ctxt) { /* N/A for PSM contexts */ - rcd->opstats = kzalloc(sizeof(*rcd->opstats), - GFP_KERNEL); + /* Applicable only for statically created kernel contexts */ + if (ctxt < dd->first_dyn_alloc_ctxt) { + rcd->opstats = kzalloc_node(sizeof(*rcd->opstats), + GFP_KERNEL, numa); if (!rcd->opstats) goto bail; } @@ -481,6 +482,9 @@ void hfi1_init_pportdata(struct pci_dev *pdev, struct hfi1_pportdata *ppd, default_pkey_idx = 1; ppd->pkeys[default_pkey_idx] = DEFAULT_P_KEY; + ppd->part_enforce |= HFI1_PART_ENFORCE_IN; + ppd->part_enforce |= HFI1_PART_ENFORCE_OUT; + if (loopback) { hfi1_early_err(&pdev->dev, "Faking data partition 0x8001 in idx %u\n", @@ -584,7 +588,7 @@ static void enable_chip(struct hfi1_devdata *dd) * Enable kernel ctxts' receive and receive interrupt. * Other ctxts done as user opens and initializes them. */ - for (i = 0; i < dd->first_user_ctxt; ++i) { + for (i = 0; i < dd->first_dyn_alloc_ctxt; ++i) { rcvmask = HFI1_RCVCTRL_CTXT_ENB | HFI1_RCVCTRL_INTRAVAIL_ENB; rcvmask |= HFI1_CAP_KGET_MASK(dd->rcd[i]->flags, DMA_RTAIL) ? HFI1_RCVCTRL_TAILUPD_ENB : HFI1_RCVCTRL_TAILUPD_DIS; @@ -615,7 +619,7 @@ static int create_workqueues(struct hfi1_devdata *dd) alloc_workqueue( "hfi%d_%d", WQ_SYSFS | WQ_HIGHPRI | WQ_CPU_INTENSIVE, - dd->num_sdma, + HFI1_MAX_ACTIVE_WORKQUEUE_ENTRIES, dd->unit, pidx); if (!ppd->hfi1_wq) goto wq_error; @@ -678,6 +682,7 @@ int hfi1_init(struct hfi1_devdata *dd, int reinit) dd->process_pio_send = hfi1_verbs_send_pio; dd->process_dma_send = hfi1_verbs_send_dma; dd->pio_inline_send = pio_copy; + dd->process_vnic_dma_send = hfi1_vnic_send_dma; if (is_ax(dd)) { atomic_set(&dd->drop_packet, DROP_PACKET_ON); @@ -713,7 +718,7 @@ int hfi1_init(struct hfi1_devdata *dd, int reinit) } /* dd->rcd can be NULL if early initialization failed */ - for (i = 0; dd->rcd && i < dd->first_user_ctxt; ++i) { + for (i = 0; dd->rcd && i < dd->first_dyn_alloc_ctxt; ++i) { /* * Set up the (kernel) rcvhdr queue and egr TIDs. If doing * re-init, the simplest way to handle this is to free @@ -959,7 +964,6 @@ void hfi1_free_ctxtdata(struct hfi1_devdata *dd, struct hfi1_ctxtdata *rcd) kfree(rcd->egrbufs.buffers); sc_free(rcd->sc); - vfree(rcd->user_event_mask); vfree(rcd->subctxt_uregbase); vfree(rcd->subctxt_rcvegrbuf); vfree(rcd->subctxt_rcvhdr_base); @@ -1077,11 +1081,11 @@ struct hfi1_devdata *hfi1_alloc_devdata(struct pci_dev *pdev, size_t extra) spin_lock_init(&dd->uctxt_lock); spin_lock_init(&dd->hfi1_diag_trans_lock); spin_lock_init(&dd->sc_init_lock); - spin_lock_init(&dd->dc8051_lock); spin_lock_init(&dd->dc8051_memlock); seqlock_init(&dd->sc2vl_lock); spin_lock_init(&dd->sde_map_lock); spin_lock_init(&dd->pio_map_lock); + mutex_init(&dd->dc8051_lock); init_waitqueue_head(&dd->event_queue); dd->int_counter = alloc_percpu(u64); @@ -1424,6 +1428,16 @@ static int init_one(struct pci_dev *pdev, const struct pci_device_id *ent) /* First, lock the non-writable module parameters */ HFI1_CAP_LOCK(); + /* Validate dev ids */ + if (!(ent->device == PCI_DEVICE_ID_INTEL0 || + ent->device == PCI_DEVICE_ID_INTEL1)) { + hfi1_early_err(&pdev->dev, + "Failing on unknown Intel deviceid 0x%x\n", + ent->device); + ret = -ENODEV; + goto bail; + } + /* Validate some global module parameters */ ret = init_validate_rcvhdrcnt(&pdev->dev, rcvhdrcnt); if (ret) @@ -1469,15 +1483,6 @@ static int init_one(struct pci_dev *pdev, const struct pci_device_id *ent) if (ret) goto bail; - if (!(ent->device == PCI_DEVICE_ID_INTEL0 || - ent->device == PCI_DEVICE_ID_INTEL1)) { - hfi1_early_err(&pdev->dev, - "Failing on unknown Intel deviceid 0x%x\n", - ent->device); - ret = -ENODEV; - goto clean_bail; - } - /* * Do device-specific initialization, function table setup, dd * allocation, etc. @@ -1496,6 +1501,9 @@ static int init_one(struct pci_dev *pdev, const struct pci_device_id *ent) /* do the generic initialization */ initfail = hfi1_init(dd, 0); + /* setup vnic */ + hfi1_vnic_setup(dd); + ret = hfi1_register_ib_device(dd); /* @@ -1529,6 +1537,7 @@ static int init_one(struct pci_dev *pdev, const struct pci_device_id *ent) hfi1_device_remove(dd); if (!ret) hfi1_unregister_ib_device(dd); + hfi1_vnic_cleanup(dd); postinit_cleanup(dd); if (initfail) ret = initfail; @@ -1573,6 +1582,9 @@ static void remove_one(struct pci_dev *pdev) /* unregister from IB core */ hfi1_unregister_ib_device(dd); + /* cleanup vnic */ + hfi1_vnic_cleanup(dd); + /* * Disable the IB link, disable interrupts on the device, * clear dma engines, etc. @@ -1612,8 +1624,11 @@ int hfi1_create_rcvhdrq(struct hfi1_devdata *dd, struct hfi1_ctxtdata *rcd) amt = PAGE_ALIGN(rcd->rcvhdrq_cnt * rcd->rcvhdrqentsize * sizeof(u32)); - gfp_flags = (rcd->ctxt >= dd->first_user_ctxt) ? - GFP_USER : GFP_KERNEL; + if ((rcd->ctxt < dd->first_dyn_alloc_ctxt) || + (rcd->sc && (rcd->sc->type == SC_KERNEL))) + gfp_flags = GFP_KERNEL; + else + gfp_flags = GFP_USER; rcd->rcvhdrq = dma_zalloc_coherent( &dd->pcidev->dev, amt, &rcd->rcvhdrq_dma, gfp_flags | __GFP_COMP); @@ -1667,8 +1682,6 @@ bail_free: dd_dev_err(dd, "attempt to allocate 1 page for ctxt %u rcvhdrqtailaddr failed\n", rcd->ctxt); - vfree(rcd->user_event_mask); - rcd->user_event_mask = NULL; dma_free_coherent(&dd->pcidev->dev, amt, rcd->rcvhdrq, rcd->rcvhdrq_dma); rcd->rcvhdrq = NULL; @@ -1757,6 +1770,7 @@ int hfi1_setup_eagerbufs(struct hfi1_ctxtdata *rcd) !HFI1_CAP_KGET_MASK(rcd->flags, MULTI_PKT_EGR)) { dd_dev_err(dd, "ctxt%u: Failed to allocate eager buffers\n", rcd->ctxt); + ret = -ENOMEM; goto bail_rcvegrbuf_phys; } @@ -1834,7 +1848,7 @@ int hfi1_setup_eagerbufs(struct hfi1_ctxtdata *rcd) "ctxt%u: current Eager buffer size is invalid %u\n", rcd->ctxt, rcd->egrbufs.rcvtid_size); ret = -EINVAL; - goto bail; + goto bail_rcvegrbuf_phys; } for (idx = 0; idx < rcd->egrbufs.alloced; idx++) { @@ -1842,7 +1856,8 @@ int hfi1_setup_eagerbufs(struct hfi1_ctxtdata *rcd) rcd->egrbufs.rcvtids[idx].dma, order); cond_resched(); } - goto bail; + + return 0; bail_rcvegrbuf_phys: for (idx = 0; idx < rcd->egrbufs.alloced && @@ -1856,6 +1871,6 @@ bail_rcvegrbuf_phys: rcd->egrbufs.buffers[idx].dma = 0; rcd->egrbufs.buffers[idx].len = 0; } -bail: + return ret; } diff --git a/drivers/infiniband/hw/hfi1/intr.c b/drivers/infiniband/hw/hfi1/intr.c index 65348d16ab2f..04a5082d5ac5 100644 --- a/drivers/infiniband/hw/hfi1/intr.c +++ b/drivers/infiniband/hw/hfi1/intr.c @@ -47,6 +47,7 @@ #include <linux/pci.h> #include <linux/delay.h> +#include <linux/bitmap.h> #include "hfi.h" #include "common.h" @@ -129,21 +130,27 @@ void handle_linkup_change(struct hfi1_devdata *dd, u32 linkup) * the remote values. Both sides must be using the values. */ if (quick_linkup || dd->icode == ICODE_FUNCTIONAL_SIMULATOR) { - set_up_vl15(dd, dd->vau, dd->vl15_init); + set_up_vau(dd, dd->vau); + set_up_vl15(dd, dd->vl15_init); assign_remote_cm_au_table(dd, dd->vcu); - ppd->neighbor_guid = - read_csr(dd, DC_DC8051_STS_REMOTE_GUID); - ppd->neighbor_type = - read_csr(dd, DC_DC8051_STS_REMOTE_NODE_TYPE) & - DC_DC8051_STS_REMOTE_NODE_TYPE_VAL_MASK; - ppd->neighbor_port_number = - read_csr(dd, DC_DC8051_STS_REMOTE_PORT_NO) & - DC_DC8051_STS_REMOTE_PORT_NO_VAL_SMASK; - dd_dev_info(dd, "Neighbor GUID: %llx Neighbor type %d\n", - ppd->neighbor_guid, - ppd->neighbor_type); } + ppd->neighbor_guid = + read_csr(dd, DC_DC8051_STS_REMOTE_GUID); + ppd->neighbor_type = + read_csr(dd, DC_DC8051_STS_REMOTE_NODE_TYPE) & + DC_DC8051_STS_REMOTE_NODE_TYPE_VAL_MASK; + ppd->neighbor_port_number = + read_csr(dd, DC_DC8051_STS_REMOTE_PORT_NO) & + DC_DC8051_STS_REMOTE_PORT_NO_VAL_SMASK; + ppd->neighbor_fm_security = + read_csr(dd, DC_DC8051_STS_REMOTE_FM_SECURITY) & + DC_DC8051_STS_LOCAL_FM_SECURITY_DISABLED_MASK; + dd_dev_info(dd, + "Neighbor Guid %llx, Type %d, Port Num %d\n", + ppd->neighbor_guid, ppd->neighbor_type, + ppd->neighbor_port_number); + /* physical link went up */ ppd->linkup = 1; ppd->offline_disabled_reason = @@ -184,7 +191,7 @@ void handle_user_interrupt(struct hfi1_ctxtdata *rcd) unsigned long flags; spin_lock_irqsave(&dd->uctxt_lock, flags); - if (!rcd->cnt) + if (bitmap_empty(rcd->in_use_ctxts, HFI1_MAX_SHARED_CTXTS)) goto done; if (test_and_clear_bit(HFI1_CTXT_WAITING_RCV, &rcd->event_flags)) { diff --git a/drivers/infiniband/hw/hfi1/mad.c b/drivers/infiniband/hw/hfi1/mad.c index 6e595afca24c..5977673a52d4 100644 --- a/drivers/infiniband/hw/hfi1/mad.c +++ b/drivers/infiniband/hw/hfi1/mad.c @@ -1,5 +1,5 @@ /* - * Copyright(c) 2015, 2016 Intel Corporation. + * Copyright(c) 2015-2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -53,6 +53,7 @@ #include "mad.h" #include "trace.h" #include "qp.h" +#include "vnic.h" /* the reset value from the FM is supposed to be 0xffff, handle both */ #define OPA_LINK_WIDTH_RESET_OLD 0x0fff @@ -650,9 +651,11 @@ static int __subn_get_opa_portinfo(struct opa_smp *smp, u32 am, u8 *data, OPA_PI_MASK_PORT_ACTIVE_OPTOMIZE : 0); pi->port_packet_format.supported = - cpu_to_be16(OPA_PORT_PACKET_FORMAT_9B); + cpu_to_be16(OPA_PORT_PACKET_FORMAT_9B | + OPA_PORT_PACKET_FORMAT_16B); pi->port_packet_format.enabled = - cpu_to_be16(OPA_PORT_PACKET_FORMAT_9B); + cpu_to_be16(OPA_PORT_PACKET_FORMAT_9B | + OPA_PORT_PACKET_FORMAT_16B); /* flit_control.interleave is (OPA V1, version .76): * bits use @@ -701,7 +704,13 @@ static int __subn_get_opa_portinfo(struct opa_smp *smp, u32 am, u8 *data, buffer_units |= (dd->vl15_init << 11) & OPA_PI_MASK_BUF_UNIT_VL15_INIT; pi->buffer_units = cpu_to_be32(buffer_units); - pi->opa_cap_mask = cpu_to_be16(OPA_CAP_MASK3_IsSharedSpaceSupported); + pi->opa_cap_mask = cpu_to_be16(OPA_CAP_MASK3_IsSharedSpaceSupported | + OPA_CAP_MASK3_IsEthOnFabricSupported); + /* Driver does not support mcast/collective configuration */ + pi->opa_cap_mask &= + cpu_to_be16(~OPA_CAP_MASK3_IsAddrRangeConfigSupported); + pi->collectivemask_multicastmask = ((HFI1_COLLECTIVE_NR & 0x7) + << 3 | (HFI1_MCAST_NR & 0x7)); /* HFI supports a replay buffer 128 LTPs in size */ pi->replay_depth.buffer = 0x80; @@ -1146,16 +1155,6 @@ static int __subn_set_opa_portinfo(struct opa_smp *smp, u32 am, u8 *data, ppd->linkinit_reason = (pi->partenforce_filterraw & OPA_PI_MASK_LINKINIT_REASON); - /* enable/disable SW pkey checking as per FM control */ - if (pi->partenforce_filterraw & OPA_PI_MASK_PARTITION_ENFORCE_IN) - ppd->part_enforce |= HFI1_PART_ENFORCE_IN; - else - ppd->part_enforce &= ~HFI1_PART_ENFORCE_IN; - - if (pi->partenforce_filterraw & OPA_PI_MASK_PARTITION_ENFORCE_OUT) - ppd->part_enforce |= HFI1_PART_ENFORCE_OUT; - else - ppd->part_enforce &= ~HFI1_PART_ENFORCE_OUT; /* Must be a valid unicast LID address. */ if ((smlid == 0 && ls_old > IB_PORT_INIT) || @@ -1167,9 +1166,9 @@ static int __subn_set_opa_portinfo(struct opa_smp *smp, u32 am, u8 *data, spin_lock_irqsave(&ibp->rvp.lock, flags); if (ibp->rvp.sm_ah) { if (smlid != ibp->rvp.sm_lid) - ibp->rvp.sm_ah->attr.dlid = smlid; + rdma_ah_set_dlid(&ibp->rvp.sm_ah->attr, smlid); if (msl != ibp->rvp.sm_sl) - ibp->rvp.sm_ah->attr.sl = msl; + rdma_ah_set_sl(&ibp->rvp.sm_ah->attr, msl); } spin_unlock_irqrestore(&ibp->rvp.lock, flags); if (smlid != ibp->rvp.sm_lid) @@ -1465,25 +1464,15 @@ static int __subn_set_opa_pkeytable(struct opa_smp *smp, u32 am, u8 *data, return __subn_get_opa_pkeytable(smp, am, data, ibdev, port, resp_len); } -static int get_sc2vlt_tables(struct hfi1_devdata *dd, void *data) -{ - u64 *val = data; - - *val++ = read_csr(dd, SEND_SC2VLT0); - *val++ = read_csr(dd, SEND_SC2VLT1); - *val++ = read_csr(dd, SEND_SC2VLT2); - *val++ = read_csr(dd, SEND_SC2VLT3); - return 0; -} - #define ILLEGAL_VL 12 /* * filter_sc2vlt changes mappings to VL15 to ILLEGAL_VL (except * for SC15, which must map to VL15). If we don't remap things this * way it is possible for VL15 counters to increment when we try to * send on a SC which is mapped to an invalid VL. + * When getting the table convert ILLEGAL_VL back to VL15. */ -static void filter_sc2vlt(void *data) +static void filter_sc2vlt(void *data, bool set) { int i; u8 *pd = data; @@ -1491,8 +1480,14 @@ static void filter_sc2vlt(void *data) for (i = 0; i < OPA_MAX_SCS; i++) { if (i == 15) continue; - if ((pd[i] & 0x1f) == 0xf) - pd[i] = ILLEGAL_VL; + + if (set) { + if ((pd[i] & 0x1f) == 0xf) + pd[i] = ILLEGAL_VL; + } else { + if ((pd[i] & 0x1f) == ILLEGAL_VL) + pd[i] = 0xf; + } } } @@ -1500,7 +1495,7 @@ static int set_sc2vlt_tables(struct hfi1_devdata *dd, void *data) { u64 *val = data; - filter_sc2vlt(data); + filter_sc2vlt(data, true); write_csr(dd, SEND_SC2VLT0, *val++); write_csr(dd, SEND_SC2VLT1, *val++); @@ -1512,6 +1507,19 @@ static int set_sc2vlt_tables(struct hfi1_devdata *dd, void *data) return 0; } +static int get_sc2vlt_tables(struct hfi1_devdata *dd, void *data) +{ + u64 *val = (u64 *)data; + + *val++ = read_csr(dd, SEND_SC2VLT0); + *val++ = read_csr(dd, SEND_SC2VLT1); + *val++ = read_csr(dd, SEND_SC2VLT2); + *val++ = read_csr(dd, SEND_SC2VLT3); + + filter_sc2vlt((u64 *)data, false); + return 0; +} + static int __subn_get_opa_sl_to_sc(struct opa_smp *smp, u32 am, u8 *data, struct ib_device *ibdev, u8 port, u32 *resp_len) @@ -1986,31 +1994,6 @@ struct opa_pma_mad { u8 data[2024]; } __packed; -struct opa_class_port_info { - u8 base_version; - u8 class_version; - __be16 cap_mask; - __be32 cap_mask2_resp_time; - - u8 redirect_gid[16]; - __be32 redirect_tc_fl; - __be32 redirect_lid; - __be32 redirect_sl_qp; - __be32 redirect_qkey; - - u8 trap_gid[16]; - __be32 trap_tc_fl; - __be32 trap_lid; - __be32 trap_hl_qp; - __be32 trap_qkey; - - __be16 trap_pkey; - __be16 redirect_pkey; - - u8 trap_sl_rsvd; - u8 reserved[3]; -} __packed; - struct opa_port_status_req { __u8 port_num; __u8 reserved[3]; @@ -4406,7 +4389,7 @@ int hfi1_process_mad(struct ib_device *ibdev, int mad_flags, u8 port, switch (in_mad->base_version) { case OPA_MGMT_BASE_VERSION: if (unlikely(in_mad_size != sizeof(struct opa_mad))) { - dev_err(ibdev->dma_device, "invalid in_mad_size\n"); + dev_err(ibdev->dev.parent, "invalid in_mad_size\n"); return IB_MAD_RESULT_FAILURE; } return hfi1_process_opa_mad(ibdev, mad_flags, port, diff --git a/drivers/infiniband/hw/hfi1/pcie.c b/drivers/infiniband/hw/hfi1/pcie.c index 4ac8f330c5cb..6a9f6f9819e1 100644 --- a/drivers/infiniband/hw/hfi1/pcie.c +++ b/drivers/infiniband/hw/hfi1/pcie.c @@ -207,8 +207,8 @@ int hfi1_pcie_ddinit(struct hfi1_devdata *dd, struct pci_dev *pdev) /* * Save BARs and command to rewrite after device reset. */ - dd->pcibar0 = addr; - dd->pcibar1 = addr >> 32; + pci_read_config_dword(dd->pcidev, PCI_BASE_ADDRESS_0, &dd->pcibar0); + pci_read_config_dword(dd->pcidev, PCI_BASE_ADDRESS_1, &dd->pcibar1); pci_read_config_dword(dd->pcidev, PCI_ROM_ADDRESS, &dd->pci_rom); pci_read_config_word(dd->pcidev, PCI_COMMAND, &dd->pci_command); pcie_capability_read_word(dd->pcidev, PCI_EXP_DEVCTL, &dd->pcie_devctl); @@ -240,36 +240,6 @@ void hfi1_pcie_ddcleanup(struct hfi1_devdata *dd) iounmap(dd->piobase); } -/* - * Do a Function Level Reset (FLR) on the device. - * Based on static function drivers/pci/pci.c:pcie_flr(). - */ -void hfi1_pcie_flr(struct hfi1_devdata *dd) -{ - int i; - u16 status; - - /* no need to check for the capability - we know the device has it */ - - /* wait for Transaction Pending bit to clear, at most a few ms */ - for (i = 0; i < 4; i++) { - if (i) - msleep((1 << (i - 1)) * 100); - - pcie_capability_read_word(dd->pcidev, PCI_EXP_DEVSTA, &status); - if (!(status & PCI_EXP_DEVSTA_TRPND)) - goto clear; - } - - dd_dev_err(dd, "Transaction Pending bit is not clearing, proceeding with reset anyway\n"); - -clear: - pcie_capability_set_word(dd->pcidev, PCI_EXP_DEVCTL, - PCI_EXP_DEVCTL_BCR_FLR); - /* PCIe spec requires the function to be back within 100ms */ - msleep(100); -} - static void msix_setup(struct hfi1_devdata *dd, int pos, u32 *msixcnt, struct hfi1_msix_entry *hfi1_msix_entry) { @@ -583,7 +553,7 @@ pci_mmio_enabled(struct pci_dev *pdev) if (words == ~0ULL) ret = PCI_ERS_RESULT_NEED_RESET; dd_dev_info(dd, - "HFI1 mmio_enabled function called, read wordscntr %Lx, returning %d\n", + "HFI1 mmio_enabled function called, read wordscntr %llx, returning %d\n", words, ret); } return ret; @@ -598,15 +568,6 @@ pci_slot_reset(struct pci_dev *pdev) return PCI_ERS_RESULT_CAN_RECOVER; } -static pci_ers_result_t -pci_link_reset(struct pci_dev *pdev) -{ - struct hfi1_devdata *dd = pci_get_drvdata(pdev); - - dd_dev_info(dd, "HFI1 link_reset function called, ignored\n"); - return PCI_ERS_RESULT_CAN_RECOVER; -} - static void pci_resume(struct pci_dev *pdev) { @@ -625,7 +586,6 @@ pci_resume(struct pci_dev *pdev) const struct pci_error_handlers hfi1_pci_err_handler = { .error_detected = pci_error_detected, .mmio_enabled = pci_mmio_enabled, - .link_reset = pci_link_reset, .slot_reset = pci_slot_reset, .resume = pci_resume, }; @@ -673,12 +633,12 @@ MODULE_PARM_DESC(pcie_retry, "Driver will try this many times to reach requested #define UNSET_PSET 255 #define DEFAULT_DISCRETE_PSET 2 /* discrete HFI */ -#define DEFAULT_MCP_PSET 4 /* MCP HFI */ +#define DEFAULT_MCP_PSET 6 /* MCP HFI */ static uint pcie_pset = UNSET_PSET; module_param(pcie_pset, uint, S_IRUGO); MODULE_PARM_DESC(pcie_pset, "PCIe Eq Pset value to use, range is 0-10"); -static uint pcie_ctle = 1; /* discrete on, integrated off */ +static uint pcie_ctle = 3; /* discrete on, integrated on */ module_param(pcie_ctle, uint, S_IRUGO); MODULE_PARM_DESC(pcie_ctle, "PCIe static CTLE mode, bit 0 - discrete on/off, bit 1 - integrated on/off"); diff --git a/drivers/infiniband/hw/hfi1/pio.c b/drivers/infiniband/hw/hfi1/pio.c index 615be68e40b3..ed72b5aca139 100644 --- a/drivers/infiniband/hw/hfi1/pio.c +++ b/drivers/infiniband/hw/hfi1/pio.c @@ -1,5 +1,5 @@ /* - * Copyright(c) 2015, 2016 Intel Corporation. + * Copyright(c) 2015-2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -703,6 +703,7 @@ struct send_context *sc_alloc(struct hfi1_devdata *dd, int type, { struct send_context_info *sci; struct send_context *sc = NULL; + int req_type = type; dma_addr_t dma; unsigned long flags; u64 reg; @@ -729,6 +730,13 @@ struct send_context *sc_alloc(struct hfi1_devdata *dd, int type, return NULL; } + /* + * VNIC contexts are dynamically allocated. + * Hence, pick a user context for VNIC. + */ + if (type == SC_VNIC) + type = SC_USER; + spin_lock_irqsave(&dd->sc_lock, flags); ret = sc_hw_alloc(dd, type, &sw_index, &hw_context); if (ret) { @@ -738,6 +746,15 @@ struct send_context *sc_alloc(struct hfi1_devdata *dd, int type, return NULL; } + /* + * VNIC contexts are used by kernel driver. + * Hence, mark them as kernel contexts. + */ + if (req_type == SC_VNIC) { + dd->send_contexts[sw_index].type = SC_KERNEL; + type = SC_KERNEL; + } + sci = &dd->send_contexts[sw_index]; sci->sc = sc; diff --git a/drivers/infiniband/hw/hfi1/pio.h b/drivers/infiniband/hw/hfi1/pio.h index 867e5ffc3595..99ca5edb0b43 100644 --- a/drivers/infiniband/hw/hfi1/pio.h +++ b/drivers/infiniband/hw/hfi1/pio.h @@ -1,7 +1,7 @@ #ifndef _PIO_H #define _PIO_H /* - * Copyright(c) 2015, 2016 Intel Corporation. + * Copyright(c) 2015-2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -54,6 +54,12 @@ #define SC_USER 3 /* must be the last one: it may take all left */ #define SC_MAX 4 /* count of send context types */ +/* + * SC_VNIC types are allocated (dynamically) from the user context pool, + * (SC_USER) and used by kernel driver as kernel contexts (SC_KERNEL). + */ +#define SC_VNIC SC_MAX + /* invalid send context index */ #define INVALID_SCI 0xff @@ -195,7 +201,7 @@ struct sc_config_sizes { * | mask | --/ |--------------------| * |--------------------------| -/ | * | * | actual_vls (max 8) | -/ |--------------------| - * |--------------------------| --/ | ksc[n] -> sc n | + * |--------------------------| --/ | ksc[n-1] -> sc n | * | vls (max 8) | -/ +--------------------+ * |--------------------------| --/ * | map[0] |-/ @@ -208,21 +214,21 @@ struct sc_config_sizes { * |--------------------------| |--------------------| * | map[vls - 1] |- | * | * +--------------------------+ \- |--------------------| - * \- | ksc[m] -> sc m+n | + * \- | ksc[m-1] -> sc m+n | * \ +--------------------+ * \- * \ - * \- +--------------------+ - * \- | mask | - * \ |--------------------| - * \- | ksc[0] -> sc 1+m+n | - * \- |--------------------| - * >| ksc[1] -> sc 2+m+n | - * |--------------------| - * | * | - * |--------------------| - * | ksc[o] -> sc o+m+n | - * +--------------------+ + * \- +----------------------+ + * \- | mask | + * \ |----------------------| + * \- | ksc[0] -> sc 1+m+n | + * \- |----------------------| + * >| ksc[1] -> sc 2+m+n | + * |----------------------| + * | * | + * |----------------------| + * | ksc[o-1] -> sc o+m+n | + * +----------------------+ * */ diff --git a/drivers/infiniband/hw/hfi1/qp.c b/drivers/infiniband/hw/hfi1/qp.c index d752d6768a49..1a7af9f60c13 100644 --- a/drivers/infiniband/hw/hfi1/qp.c +++ b/drivers/infiniband/hw/hfi1/qp.c @@ -79,43 +79,6 @@ static inline unsigned mk_qpn(struct rvt_qpn_table *qpt, return (map - qpt->map) * RVT_BITS_PER_PAGE + off; } -/* - * Convert the AETH credit code into the number of credits. - */ -static const u16 credit_table[31] = { - 0, /* 0 */ - 1, /* 1 */ - 2, /* 2 */ - 3, /* 3 */ - 4, /* 4 */ - 6, /* 5 */ - 8, /* 6 */ - 12, /* 7 */ - 16, /* 8 */ - 24, /* 9 */ - 32, /* A */ - 48, /* B */ - 64, /* C */ - 96, /* D */ - 128, /* E */ - 192, /* F */ - 256, /* 10 */ - 384, /* 11 */ - 512, /* 12 */ - 768, /* 13 */ - 1024, /* 14 */ - 1536, /* 15 */ - 2048, /* 16 */ - 3072, /* 17 */ - 4096, /* 18 */ - 6144, /* 19 */ - 8192, /* 1A */ - 12288, /* 1B */ - 16384, /* 1C */ - 24576, /* 1D */ - 32768 /* 1E */ -}; - const struct rvt_operation_params hfi1_post_parms[RVT_OPERATION_MAX] = { [IB_WR_RDMA_WRITE] = { .length = sizeof(struct ib_rdma_wr), @@ -331,7 +294,7 @@ int hfi1_check_send_wqe(struct rvt_qp *qp, ah = ibah_to_rvtah(wqe->ud_wr.ah); if (wqe->length > (1 << ah->log_pmtu)) return -EINVAL; - if (ibp->sl_to_sc[ah->attr.sl] == 0xf) + if (ibp->sl_to_sc[rdma_ah_get_sl(&ah->attr)] == 0xf) return -EINVAL; default: break; @@ -340,68 +303,6 @@ int hfi1_check_send_wqe(struct rvt_qp *qp, } /** - * hfi1_compute_aeth - compute the AETH (syndrome + MSN) - * @qp: the queue pair to compute the AETH for - * - * Returns the AETH. - */ -__be32 hfi1_compute_aeth(struct rvt_qp *qp) -{ - u32 aeth = qp->r_msn & HFI1_MSN_MASK; - - if (qp->ibqp.srq) { - /* - * Shared receive queues don't generate credits. - * Set the credit field to the invalid value. - */ - aeth |= HFI1_AETH_CREDIT_INVAL << HFI1_AETH_CREDIT_SHIFT; - } else { - u32 min, max, x; - u32 credits; - struct rvt_rwq *wq = qp->r_rq.wq; - u32 head; - u32 tail; - - /* sanity check pointers before trusting them */ - head = wq->head; - if (head >= qp->r_rq.size) - head = 0; - tail = wq->tail; - if (tail >= qp->r_rq.size) - tail = 0; - /* - * Compute the number of credits available (RWQEs). - * There is a small chance that the pair of reads are - * not atomic, which is OK, since the fuzziness is - * resolved as further ACKs go out. - */ - credits = head - tail; - if ((int)credits < 0) - credits += qp->r_rq.size; - /* - * Binary search the credit table to find the code to - * use. - */ - min = 0; - max = 31; - for (;;) { - x = (min + max) / 2; - if (credit_table[x] == credits) - break; - if (credit_table[x] > credits) { - max = x; - } else { - if (min == x) - break; - min = x; - } - } - aeth |= x << HFI1_AETH_CREDIT_SHIFT; - } - return cpu_to_be32(aeth); -} - -/** * _hfi1_schedule_send - schedule progress * @qp: the QP * @@ -457,44 +358,6 @@ void hfi1_schedule_send(struct rvt_qp *qp) _hfi1_schedule_send(qp); } -/** - * hfi1_get_credit - handle credit in aeth - * @qp: the qp - * @aeth: the Acknowledge Extended Transport Header - * - * The QP s_lock should be held. - */ -void hfi1_get_credit(struct rvt_qp *qp, u32 aeth) -{ - u32 credit = (aeth >> HFI1_AETH_CREDIT_SHIFT) & HFI1_AETH_CREDIT_MASK; - - lockdep_assert_held(&qp->s_lock); - /* - * If the credit is invalid, we can send - * as many packets as we like. Otherwise, we have to - * honor the credit field. - */ - if (credit == HFI1_AETH_CREDIT_INVAL) { - if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT)) { - qp->s_flags |= RVT_S_UNLIMITED_CREDIT; - if (qp->s_flags & RVT_S_WAIT_SSN_CREDIT) { - qp->s_flags &= ~RVT_S_WAIT_SSN_CREDIT; - hfi1_schedule_send(qp); - } - } - } else if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT)) { - /* Compute new LSN (i.e., MSN + credit) */ - credit = (aeth + credit_table[credit]) & HFI1_MSN_MASK; - if (cmp_msn(credit, qp->s_lsn) > 0) { - qp->s_lsn = credit; - if (qp->s_flags & RVT_S_WAIT_SSN_CREDIT) { - qp->s_flags &= ~RVT_S_WAIT_SSN_CREDIT; - hfi1_schedule_send(qp); - } - } - } -} - void hfi1_qp_wakeup(struct rvt_qp *qp, u32 flag) { unsigned long flags; @@ -744,7 +607,7 @@ void qp_iter_print(struct seq_file *s, struct qp_iter *iter) wqe = rvt_get_swqe_ptr(qp, qp->s_last); send_context = qp_to_send_context(qp, priv->s_sc); seq_printf(s, - "N %d %s QP %x R %u %s %u %u %u f=%x %u %u %u %u %u %u PSN %x %x %x %x %x (%u %u %u %u %u %u %u) RQP %x LID %x SL %u MTU %u %u %u %u SDE %p,%u SC %p,%u SCQ %u %u PID %d\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 (%u %u %u %u %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\n", iter->n, qp_idle(qp) ? "I" : "B", qp->ibqp.qp_num, @@ -763,16 +626,18 @@ void qp_iter_print(struct seq_file *s, struct qp_iter *iter) qp->s_last_psn, qp->s_psn, qp->s_next_psn, qp->s_sending_psn, qp->s_sending_hpsn, + qp->r_psn, qp->s_last, qp->s_acked, qp->s_cur, qp->s_tail, qp->s_head, qp->s_size, qp->s_avail, qp->remote_qpn, - qp->remote_ah_attr.dlid, - qp->remote_ah_attr.sl, + rdma_ah_get_dlid(&qp->remote_ah_attr), + rdma_ah_get_sl(&qp->remote_ah_attr), qp->pmtu, qp->s_retry, qp->s_retry_cnt, qp->s_rnr_retry_cnt, + qp->s_rnr_retry, sde, sde ? sde->this_idx : 0, send_context, @@ -782,31 +647,17 @@ void qp_iter_print(struct seq_file *s, struct qp_iter *iter) qp->pid); } -void qp_comm_est(struct rvt_qp *qp) -{ - qp->r_flags |= RVT_R_COMM_EST; - if (qp->ibqp.event_handler) { - struct ib_event ev; - - ev.device = qp->ibqp.device; - ev.element.qp = &qp->ibqp; - ev.event = IB_EVENT_COMM_EST; - qp->ibqp.event_handler(&ev, qp->ibqp.qp_context); - } -} - -void *qp_priv_alloc(struct rvt_dev_info *rdi, struct rvt_qp *qp, - gfp_t gfp) +void *qp_priv_alloc(struct rvt_dev_info *rdi, struct rvt_qp *qp) { struct hfi1_qp_priv *priv; - priv = kzalloc_node(sizeof(*priv), gfp, rdi->dparms.node); + priv = kzalloc_node(sizeof(*priv), GFP_KERNEL, rdi->dparms.node); if (!priv) return ERR_PTR(-ENOMEM); priv->owner = qp; - priv->s_ahg = kzalloc_node(sizeof(*priv->s_ahg), gfp, + priv->s_ahg = kzalloc_node(sizeof(*priv->s_ahg), GFP_KERNEL, rdi->dparms.node); if (!priv->s_ahg) { kfree(priv); @@ -819,8 +670,6 @@ void *qp_priv_alloc(struct rvt_dev_info *rdi, struct rvt_qp *qp, iowait_sleep, iowait_wakeup, iowait_sdma_drained); - setup_timer(&priv->s_rnr_timer, hfi1_rc_rnr_retry, (unsigned long)qp); - qp->s_timer.function = hfi1_rc_timeout; return priv; } @@ -861,7 +710,6 @@ void flush_qp_waiters(struct rvt_qp *qp) { lockdep_assert_held(&qp->s_lock); flush_iowait(qp); - hfi1_stop_rc_timers(qp); } void stop_send_queue(struct rvt_qp *qp) @@ -869,7 +717,6 @@ void stop_send_queue(struct rvt_qp *qp) struct hfi1_qp_priv *priv = qp->priv; cancel_work_sync(&priv->s_iowait.iowork); - hfi1_del_timers_sync(qp); } void quiesce_qp(struct rvt_qp *qp) @@ -883,9 +730,7 @@ void quiesce_qp(struct rvt_qp *qp) void notify_qp_reset(struct rvt_qp *qp) { - struct hfi1_qp_priv *priv = qp->priv; - - priv->r_adefered = 0; + qp->r_adefered = 0; clear_ahg(qp); } @@ -900,7 +745,7 @@ void hfi1_migrate_qp(struct rvt_qp *qp) qp->s_mig_state = IB_MIG_MIGRATED; qp->remote_ah_attr = qp->alt_ah_attr; - qp->port_num = qp->alt_ah_attr.port_num; + qp->port_num = rdma_ah_get_port_num(&qp->alt_ah_attr); qp->s_pkey_index = qp->s_alt_pkey_index; qp->s_flags |= RVT_S_AHG_CLEAR; priv->s_sc = ah_to_sc(qp->ibqp.device, &qp->remote_ah_attr); @@ -930,7 +775,7 @@ u32 mtu_from_qp(struct rvt_dev_info *rdi, struct rvt_qp *qp, u32 pmtu) u8 sc, vl; ibp = &dd->pport[qp->port_num - 1].ibport_data; - sc = ibp->sl_to_sc[qp->remote_ah_attr.sl]; + sc = ibp->sl_to_sc[rdma_ah_get_sl(&qp->remote_ah_attr)]; vl = sc_to_vlt(dd, sc); mtu = verbs_mtu_enum_to_int(qp->ibqp.device, pmtu); @@ -961,17 +806,20 @@ int get_pmtu_from_attr(struct rvt_dev_info *rdi, struct rvt_qp *qp, void notify_error_qp(struct rvt_qp *qp) { - struct hfi1_ibdev *dev = to_idev(qp->ibqp.device); struct hfi1_qp_priv *priv = qp->priv; + seqlock_t *lock = priv->s_iowait.lock; - write_seqlock(&dev->iowait_lock); - if (!list_empty(&priv->s_iowait.list) && !(qp->s_flags & RVT_S_BUSY)) { - qp->s_flags &= ~RVT_S_ANY_WAIT_IO; - list_del_init(&priv->s_iowait.list); - priv->s_iowait.lock = NULL; - rvt_put_qp(qp); + if (lock) { + write_seqlock(lock); + if (!list_empty(&priv->s_iowait.list) && + !(qp->s_flags & RVT_S_BUSY)) { + qp->s_flags &= ~RVT_S_ANY_WAIT_IO; + list_del_init(&priv->s_iowait.list); + priv->s_iowait.lock = NULL; + rvt_put_qp(qp); + } + write_sequnlock(lock); } - write_sequnlock(&dev->iowait_lock); if (!(qp->s_flags & RVT_S_BUSY)) { qp->s_hdrwords = 0; @@ -1010,7 +858,7 @@ void hfi1_error_port_qps(struct hfi1_ibport *ibp, u8 sl) if (qp->port_num == ppd->port && (qp->ibqp.qp_type == IB_QPT_UC || qp->ibqp.qp_type == IB_QPT_RC) && - qp->remote_ah_attr.sl == sl && + rdma_ah_get_sl(&qp->remote_ah_attr) == sl && (ib_rvt_state_ops[qp->state] & RVT_POST_SEND_OK)) { spin_lock_irq(&qp->r_lock); diff --git a/drivers/infiniband/hw/hfi1/qp.h b/drivers/infiniband/hw/hfi1/qp.h index 587d84d65bb8..6fe542b6a927 100644 --- a/drivers/infiniband/hw/hfi1/qp.h +++ b/drivers/infiniband/hw/hfi1/qp.h @@ -71,14 +71,6 @@ static inline void clear_ahg(struct rvt_qp *qp) } /** - * hfi1_compute_aeth - compute the AETH (syndrome + MSN) - * @qp: the queue pair to compute the AETH for - * - * Returns the AETH. - */ -__be32 hfi1_compute_aeth(struct rvt_qp *qp); - -/** * hfi1_create_qp - create a queue pair for a device * @ibpd: the protection domain who's device we create the queue pair for * @init_attr: the attributes of the queue pair @@ -91,14 +83,6 @@ __be32 hfi1_compute_aeth(struct rvt_qp *qp); struct ib_qp *hfi1_create_qp(struct ib_pd *ibpd, struct ib_qp_init_attr *init_attr, struct ib_udata *udata); -/** - * hfi1_get_credit - flush the send work queue of a QP - * @qp: the qp who's send work queue to flush - * @aeth: the Acknowledge Extended Transport Header - * - * The QP s_lock should be held. - */ -void hfi1_get_credit(struct rvt_qp *qp, u32 aeth); /** * hfi1_qp_wakeup - wake up on the indicated event @@ -131,12 +115,6 @@ int qp_iter_next(struct qp_iter *iter); */ void qp_iter_print(struct seq_file *s, struct qp_iter *iter); -/** - * qp_comm_est - handle trap with QP established - * @qp: the QP - */ -void qp_comm_est(struct rvt_qp *qp); - void _hfi1_schedule_send(struct rvt_qp *qp); void hfi1_schedule_send(struct rvt_qp *qp); @@ -145,8 +123,7 @@ void hfi1_migrate_qp(struct rvt_qp *qp); /* * Functions provided by hfi1 driver for rdmavt to use */ -void *qp_priv_alloc(struct rvt_dev_info *rdi, struct rvt_qp *qp, - gfp_t gfp); +void *qp_priv_alloc(struct rvt_dev_info *rdi, struct rvt_qp *qp); void qp_priv_free(struct rvt_dev_info *rdi, struct rvt_qp *qp); unsigned free_all_qps(struct rvt_dev_info *rdi); void notify_qp_reset(struct rvt_qp *qp); diff --git a/drivers/infiniband/hw/hfi1/rc.c b/drivers/infiniband/hw/hfi1/rc.c index 809b26eb6d3c..1080778a1f7c 100644 --- a/drivers/infiniband/hw/hfi1/rc.c +++ b/drivers/infiniband/hw/hfi1/rc.c @@ -57,133 +57,6 @@ /* cut down ridiculously long IB macro names */ #define OP(x) RC_OP(x) -/** - * hfi1_add_retry_timer - add/start a retry timer - * @qp - the QP - * - * add a retry timer on the QP - */ -static inline void hfi1_add_retry_timer(struct rvt_qp *qp) -{ - struct ib_qp *ibqp = &qp->ibqp; - struct rvt_dev_info *rdi = ib_to_rvt(ibqp->device); - - lockdep_assert_held(&qp->s_lock); - qp->s_flags |= RVT_S_TIMER; - /* 4.096 usec. * (1 << qp->timeout) */ - qp->s_timer.expires = jiffies + qp->timeout_jiffies + - rdi->busy_jiffies; - add_timer(&qp->s_timer); -} - -/** - * hfi1_add_rnr_timer - add/start an rnr timer - * @qp - the QP - * @to - timeout in usecs - * - * add an rnr timer on the QP - */ -void hfi1_add_rnr_timer(struct rvt_qp *qp, u32 to) -{ - struct hfi1_qp_priv *priv = qp->priv; - - lockdep_assert_held(&qp->s_lock); - qp->s_flags |= RVT_S_WAIT_RNR; - priv->s_rnr_timer.expires = jiffies + usecs_to_jiffies(to); - add_timer(&priv->s_rnr_timer); -} - -/** - * hfi1_mod_retry_timer - mod a retry timer - * @qp - the QP - * - * Modify a potentially already running retry - * timer - */ -static inline void hfi1_mod_retry_timer(struct rvt_qp *qp) -{ - struct ib_qp *ibqp = &qp->ibqp; - struct rvt_dev_info *rdi = ib_to_rvt(ibqp->device); - - lockdep_assert_held(&qp->s_lock); - qp->s_flags |= RVT_S_TIMER; - /* 4.096 usec. * (1 << qp->timeout) */ - mod_timer(&qp->s_timer, jiffies + qp->timeout_jiffies + - rdi->busy_jiffies); -} - -/** - * hfi1_stop_retry_timer - stop a retry timer - * @qp - the QP - * - * stop a retry timer and return if the timer - * had been pending. - */ -static inline int hfi1_stop_retry_timer(struct rvt_qp *qp) -{ - int rval = 0; - - lockdep_assert_held(&qp->s_lock); - /* Remove QP from retry */ - if (qp->s_flags & RVT_S_TIMER) { - qp->s_flags &= ~RVT_S_TIMER; - rval = del_timer(&qp->s_timer); - } - return rval; -} - -/** - * hfi1_stop_rc_timers - stop all timers - * @qp - the QP - * - * stop any pending timers - */ -void hfi1_stop_rc_timers(struct rvt_qp *qp) -{ - struct hfi1_qp_priv *priv = qp->priv; - - lockdep_assert_held(&qp->s_lock); - /* Remove QP from all timers */ - if (qp->s_flags & (RVT_S_TIMER | RVT_S_WAIT_RNR)) { - qp->s_flags &= ~(RVT_S_TIMER | RVT_S_WAIT_RNR); - del_timer(&qp->s_timer); - del_timer(&priv->s_rnr_timer); - } -} - -/** - * hfi1_stop_rnr_timer - stop an rnr timer - * @qp - the QP - * - * stop an rnr timer and return if the timer - * had been pending. - */ -static inline int hfi1_stop_rnr_timer(struct rvt_qp *qp) -{ - int rval = 0; - struct hfi1_qp_priv *priv = qp->priv; - - 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 = del_timer(&priv->s_rnr_timer); - } - return rval; -} - -/** - * hfi1_del_timers_sync - wait for any timeout routines to exit - * @qp - the QP - */ -void hfi1_del_timers_sync(struct rvt_qp *qp) -{ - struct hfi1_qp_priv *priv = qp->priv; - - del_timer_sync(&qp->s_timer); - del_timer_sync(&priv->s_rnr_timer); -} - static u32 restart_sge(struct rvt_sge_state *ss, struct rvt_swqe *wqe, u32 psn, u32 pmtu) { @@ -194,7 +67,7 @@ static u32 restart_sge(struct rvt_sge_state *ss, struct rvt_swqe *wqe, ss->sg_list = wqe->sg_list + 1; ss->num_sge = wqe->wr.num_sge; ss->total_len = wqe->length; - hfi1_skip_sge(ss, len, 0); + rvt_skip_sge(ss, len, false); return wqe->length - len; } @@ -284,7 +157,7 @@ static int make_rc_ack(struct hfi1_ibdev *dev, struct rvt_qp *qp, qp->s_ack_state = OP(RDMA_READ_RESPONSE_ONLY); e->sent = 1; } - ohdr->u.aeth = hfi1_compute_aeth(qp); + ohdr->u.aeth = rvt_compute_aeth(qp); hwords++; qp->s_ack_rdma_psn = e->psn; bth2 = mask_psn(qp->s_ack_rdma_psn++); @@ -293,7 +166,7 @@ static int make_rc_ack(struct hfi1_ibdev *dev, struct rvt_qp *qp, ps->s_txreq->ss = NULL; len = 0; qp->s_ack_state = OP(ATOMIC_ACKNOWLEDGE); - ohdr->u.at.aeth = hfi1_compute_aeth(qp); + ohdr->u.at.aeth = rvt_compute_aeth(qp); ib_u64_put(e->atomic_data, &ohdr->u.at.atomic_ack_eth); hwords += sizeof(ohdr->u.at) / sizeof(u32); bth2 = mask_psn(e->psn); @@ -315,7 +188,7 @@ static int make_rc_ack(struct hfi1_ibdev *dev, struct rvt_qp *qp, len = pmtu; middle = HFI1_CAP_IS_KSET(SDMA_AHG); } else { - ohdr->u.aeth = hfi1_compute_aeth(qp); + ohdr->u.aeth = rvt_compute_aeth(qp); hwords++; qp->s_ack_state = OP(RDMA_READ_RESPONSE_LAST); e = &qp->s_ack_queue[qp->s_tail_ack_queue]; @@ -338,11 +211,11 @@ normal: ps->s_txreq->ss = NULL; if (qp->s_nak_state) ohdr->u.aeth = - cpu_to_be32((qp->r_msn & HFI1_MSN_MASK) | + cpu_to_be32((qp->r_msn & IB_MSN_MASK) | (qp->s_nak_state << - HFI1_AETH_CREDIT_SHIFT)); + IB_AETH_CREDIT_SHIFT)); else - ohdr->u.aeth = hfi1_compute_aeth(qp); + ohdr->u.aeth = rvt_compute_aeth(qp); hwords++; len = 0; bth0 = OP(ACKNOWLEDGE) << 24; @@ -401,7 +274,7 @@ int hfi1_make_rc_req(struct rvt_qp *qp, struct hfi1_pkt_state *ps) goto bail_no_tx; ohdr = &ps->s_txreq->phdr.hdr.u.oth; - if (qp->remote_ah_attr.ah_flags & IB_AH_GRH) + if (rdma_ah_get_ah_flags(&qp->remote_ah_attr) & IB_AH_GRH) ohdr = &ps->s_txreq->phdr.hdr.u.l.oth; /* Sending responses has higher priority over sending requests. */ @@ -414,7 +287,7 @@ int hfi1_make_rc_req(struct rvt_qp *qp, struct hfi1_pkt_state *ps) goto bail; /* We are in the error state, flush the work request. */ smp_read_barrier_depends(); /* see post_one_send() */ - if (qp->s_last == ACCESS_ONCE(qp->s_head)) + if (qp->s_last == READ_ONCE(qp->s_head)) goto bail; /* If DMAs are in progress, we can't flush immediately. */ if (iowait_sdma_pending(&priv->s_iowait)) { @@ -457,7 +330,8 @@ int hfi1_make_rc_req(struct rvt_qp *qp, struct hfi1_pkt_state *ps) newreq = 0; if (qp->s_cur == qp->s_tail) { /* Check if send work queue is empty. */ - if (qp->s_tail == qp->s_head) { + smp_read_barrier_depends(); /* see post_one_send() */ + if (qp->s_tail == READ_ONCE(qp->s_head)) { clear_ahg(qp); goto bail; } @@ -518,7 +392,7 @@ int hfi1_make_rc_req(struct rvt_qp *qp, struct hfi1_pkt_state *ps) case IB_WR_SEND_WITH_INV: /* If no credit, return. */ if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT) && - cmp_msn(wqe->ssn, qp->s_lsn + 1) > 0) { + rvt_cmp_msn(wqe->ssn, qp->s_lsn + 1) > 0) { qp->s_flags |= RVT_S_WAIT_SSN_CREDIT; goto bail; } @@ -555,7 +429,7 @@ int hfi1_make_rc_req(struct rvt_qp *qp, struct hfi1_pkt_state *ps) case IB_WR_RDMA_WRITE_WITH_IMM: /* If no credit, return. */ if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT) && - cmp_msn(wqe->ssn, qp->s_lsn + 1) > 0) { + rvt_cmp_msn(wqe->ssn, qp->s_lsn + 1) > 0) { qp->s_flags |= RVT_S_WAIT_SSN_CREDIT; goto bail; } @@ -840,7 +714,7 @@ bail_no_tx: void hfi1_send_rc_ack(struct hfi1_ctxtdata *rcd, struct rvt_qp *qp, int is_fecn) { - struct hfi1_ibport *ibp = to_iport(qp->ibqp.device, qp->port_num); + struct hfi1_ibport *ibp = rcd_to_iport(rcd); struct hfi1_pportdata *ppd = ppd_from_ibp(ibp); u64 pbc, pbc_flags = 0; u16 lrh0; @@ -854,6 +728,9 @@ void hfi1_send_rc_ack(struct hfi1_ctxtdata *rcd, struct rvt_qp *qp, struct ib_other_headers *ohdr; unsigned long flags; + /* clear the defer count */ + qp->r_adefered = 0; + /* Don't send ACK or NAK if a RDMA read or atomic is pending. */ if (qp->s_flags & RVT_S_RESP_PENDING) goto queue_ack; @@ -866,9 +743,10 @@ void hfi1_send_rc_ack(struct hfi1_ctxtdata *rcd, struct rvt_qp *qp, /* Construct the header */ /* header size in 32-bit words LRH+BTH+AETH = (8+12+4)/4 */ hwords = 6; - if (unlikely(qp->remote_ah_attr.ah_flags & IB_AH_GRH)) { + if (unlikely(rdma_ah_get_ah_flags(&qp->remote_ah_attr) & IB_AH_GRH)) { hwords += hfi1_make_grh(ibp, &hdr.u.l.grh, - &qp->remote_ah_attr.grh, hwords, 0); + rdma_ah_read_grh(&qp->remote_ah_attr), + hwords, 0); ohdr = &hdr.u.l.oth; lrh0 = HFI1_LRH_GRH; } else { @@ -880,22 +758,24 @@ void hfi1_send_rc_ack(struct hfi1_ctxtdata *rcd, struct rvt_qp *qp, if (qp->s_mig_state == IB_MIG_MIGRATED) bth0 |= IB_BTH_MIG_REQ; if (qp->r_nak_state) - ohdr->u.aeth = cpu_to_be32((qp->r_msn & HFI1_MSN_MASK) | + ohdr->u.aeth = cpu_to_be32((qp->r_msn & IB_MSN_MASK) | (qp->r_nak_state << - HFI1_AETH_CREDIT_SHIFT)); + IB_AETH_CREDIT_SHIFT)); else - ohdr->u.aeth = hfi1_compute_aeth(qp); - sc5 = ibp->sl_to_sc[qp->remote_ah_attr.sl]; + ohdr->u.aeth = rvt_compute_aeth(qp); + sc5 = ibp->sl_to_sc[rdma_ah_get_sl(&qp->remote_ah_attr)]; /* set PBC_DC_INFO bit (aka SC[4]) in pbc_flags */ pbc_flags |= ((!!(sc5 & 0x10)) << PBC_DC_INFO_SHIFT); - lrh0 |= (sc5 & 0xf) << 12 | (qp->remote_ah_attr.sl & 0xf) << 4; + lrh0 |= (sc5 & 0xf) << 12 | (rdma_ah_get_sl(&qp->remote_ah_attr) + & 0xf) << 4; hdr.lrh[0] = cpu_to_be16(lrh0); - hdr.lrh[1] = cpu_to_be16(qp->remote_ah_attr.dlid); + hdr.lrh[1] = cpu_to_be16(rdma_ah_get_dlid(&qp->remote_ah_attr)); hdr.lrh[2] = cpu_to_be16(hwords + SIZE_OF_CRC); - hdr.lrh[3] = cpu_to_be16(ppd->lid | qp->remote_ah_attr.src_path_bits); + hdr.lrh[3] = cpu_to_be16(ppd->lid | + rdma_ah_get_path_bits(&qp->remote_ah_attr)); ohdr->bth[0] = cpu_to_be32(bth0); ohdr->bth[1] = cpu_to_be32(qp->remote_qpn); - ohdr->bth[1] |= cpu_to_be32((!!is_fecn) << HFI1_BECN_SHIFT); + ohdr->bth[1] |= cpu_to_be32((!!is_fecn) << IB_BECN_SHIFT); ohdr->bth[2] = cpu_to_be32(mask_psn(qp->r_ack_psn)); /* Don't try to send ACKs if the link isn't ACTIVE */ @@ -1038,7 +918,7 @@ done: * Back up requester to resend the last un-ACKed request. * The QP r_lock and s_lock should be held and interrupts disabled. */ -static void restart_rc(struct rvt_qp *qp, u32 psn, int wait) +void hfi1_restart_rc(struct rvt_qp *qp, u32 psn, int wait) { struct rvt_swqe *wqe = rvt_get_swqe_ptr(qp, qp->s_acked); struct hfi1_ibport *ibp; @@ -1075,44 +955,6 @@ static void restart_rc(struct rvt_qp *qp, u32 psn, int wait) } /* - * This is called from s_timer for missing responses. - */ -void hfi1_rc_timeout(unsigned long arg) -{ - struct rvt_qp *qp = (struct rvt_qp *)arg; - struct hfi1_ibport *ibp; - unsigned long flags; - - spin_lock_irqsave(&qp->r_lock, flags); - spin_lock(&qp->s_lock); - if (qp->s_flags & RVT_S_TIMER) { - ibp = to_iport(qp->ibqp.device, qp->port_num); - ibp->rvp.n_rc_timeouts++; - qp->s_flags &= ~RVT_S_TIMER; - del_timer(&qp->s_timer); - trace_hfi1_timeout(qp, qp->s_last_psn + 1); - restart_rc(qp, qp->s_last_psn + 1, 1); - hfi1_schedule_send(qp); - } - spin_unlock(&qp->s_lock); - spin_unlock_irqrestore(&qp->r_lock, flags); -} - -/* - * This is called from s_timer for RNR timeouts. - */ -void hfi1_rc_rnr_retry(unsigned long arg) -{ - struct rvt_qp *qp = (struct rvt_qp *)arg; - unsigned long flags; - - spin_lock_irqsave(&qp->s_lock, flags); - hfi1_stop_rnr_timer(qp); - hfi1_schedule_send(qp); - spin_unlock_irqrestore(&qp->s_lock, flags); -} - -/* * Set qp->s_sending_psn to the next PSN after the given one. * This would be psn+1 except when RDMA reads are present. */ @@ -1150,16 +992,16 @@ void hfi1_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr) u32 psn; lockdep_assert_held(&qp->s_lock); - if (!(ib_rvt_state_ops[qp->state] & RVT_PROCESS_OR_FLUSH_SEND)) + if (!(ib_rvt_state_ops[qp->state] & RVT_SEND_OR_FLUSH_OR_RECV_OK)) return; /* Find out where the BTH is */ - if ((be16_to_cpu(hdr->lrh[0]) & 3) == HFI1_LRH_BTH) + if (ib_get_lnh(hdr) == HFI1_LRH_BTH) ohdr = &hdr->u.oth; else ohdr = &hdr->u.l.oth; - opcode = be32_to_cpu(ohdr->bth[0]) >> 24; + opcode = ib_bth_get_opcode(ohdr); if (opcode >= OP(RDMA_READ_RESPONSE_FIRST) && opcode <= OP(ATOMIC_ACKNOWLEDGE)) { WARN_ON(!qp->s_rdma_ack_cnt); @@ -1178,7 +1020,7 @@ void hfi1_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr) !(qp->s_flags & (RVT_S_TIMER | RVT_S_WAIT_RNR | RVT_S_WAIT_PSN)) && (ib_rvt_state_ops[qp->state] & RVT_PROCESS_RECV_OK)) - hfi1_add_retry_timer(qp); + rvt_add_retry_timer(qp); while (qp->s_last != qp->s_acked) { u32 s_last; @@ -1188,13 +1030,17 @@ void hfi1_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr) cmp_psn(qp->s_sending_psn, qp->s_sending_hpsn) <= 0) break; s_last = qp->s_last; + trace_hfi1_qp_send_completion(qp, wqe, s_last); if (++s_last >= qp->s_size) s_last = 0; qp->s_last = s_last; /* see post_send() */ barrier(); rvt_put_swqe(wqe); - rvt_qp_swqe_complete(qp, wqe, IB_WC_SUCCESS); + rvt_qp_swqe_complete(qp, + wqe, + ib_hfi1_wc_opcode[wqe->wr.opcode], + IB_WC_SUCCESS); } /* * If we were waiting for sends to complete before re-sending, @@ -1236,12 +1082,16 @@ static struct rvt_swqe *do_rc_completion(struct rvt_qp *qp, rvt_put_swqe(wqe); s_last = qp->s_last; + trace_hfi1_qp_send_completion(qp, wqe, s_last); if (++s_last >= qp->s_size) s_last = 0; qp->s_last = s_last; /* see post_send() */ barrier(); - rvt_qp_swqe_complete(qp, wqe, IB_WC_SUCCESS); + rvt_qp_swqe_complete(qp, + wqe, + ib_hfi1_wc_opcode[wqe->wr.opcode], + IB_WC_SUCCESS); } else { struct hfi1_pportdata *ppd = ppd_from_ibp(ibp); @@ -1252,10 +1102,11 @@ static struct rvt_swqe *do_rc_completion(struct rvt_qp *qp, */ if (ppd->dd->flags & HFI1_HAS_SEND_DMA) { struct sdma_engine *engine; + u8 sl = rdma_ah_get_sl(&qp->remote_ah_attr); u8 sc5; /* For now use sc to find engine */ - sc5 = ibp->sl_to_sc[qp->remote_ah_attr.sl]; + sc5 = ibp->sl_to_sc[sl]; engine = qp_to_sdma_engine(qp, sc5); sdma_engine_progress_schedule(engine); } @@ -1308,7 +1159,6 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, int ret = 0; u32 ack_psn; int diff; - unsigned long to; lockdep_assert_held(&qp->s_lock); /* @@ -1318,10 +1168,10 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, * request but will include an ACK'ed request(s). */ ack_psn = psn; - if (aeth >> 29) + if (aeth >> IB_AETH_NAK_SHIFT) ack_psn--; wqe = rvt_get_swqe_ptr(qp, qp->s_acked); - ibp = to_iport(qp->ibqp.device, qp->port_num); + ibp = rcd_to_iport(rcd); /* * The MSN might be for a later WQE than the PSN indicates so @@ -1357,7 +1207,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, /* Retry this request. */ if (!(qp->r_flags & RVT_R_RDMAR_SEQ)) { qp->r_flags |= RVT_R_RDMAR_SEQ; - restart_rc(qp, qp->s_last_psn + 1, 0); + hfi1_restart_rc(qp, qp->s_last_psn + 1, 0); if (list_empty(&qp->rspwait)) { qp->r_flags |= RVT_R_RSP_SEND; rvt_get_qp(qp); @@ -1398,7 +1248,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, break; } - switch (aeth >> 29) { + switch (aeth >> IB_AETH_NAK_SHIFT) { case 0: /* ACK */ this_cpu_inc(*ibp->rvp.rc_acks); if (qp->s_acked != qp->s_tail) { @@ -1406,7 +1256,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, * We are expecting more ACKs so * mod the retry timer. */ - hfi1_mod_retry_timer(qp); + rvt_mod_retry_timer(qp); /* * We can stop re-sending the earlier packets and * continue with the next packet the receiver wants. @@ -1415,7 +1265,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, reset_psn(qp, psn + 1); } else { /* No more acks - kill all timers */ - hfi1_stop_rc_timers(qp); + rvt_stop_rc_timers(qp); if (cmp_psn(qp->s_psn, psn) <= 0) { qp->s_state = OP(SEND_LAST); qp->s_psn = psn + 1; @@ -1425,7 +1275,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, qp->s_flags &= ~RVT_S_WAIT_ACK; hfi1_schedule_send(qp); } - hfi1_get_credit(qp, aeth); + rvt_get_credit(qp, aeth); qp->s_rnr_retry = qp->s_rnr_retry_cnt; qp->s_retry = qp->s_retry_cnt; update_last_psn(qp, psn); @@ -1452,11 +1302,8 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, reset_psn(qp, psn); qp->s_flags &= ~(RVT_S_WAIT_SSN_CREDIT | RVT_S_WAIT_ACK); - hfi1_stop_rc_timers(qp); - to = - ib_hfi1_rnr_table[(aeth >> HFI1_AETH_CREDIT_SHIFT) & - HFI1_AETH_CREDIT_MASK]; - hfi1_add_rnr_timer(qp, to); + rvt_stop_rc_timers(qp); + rvt_add_rnr_timer(qp, aeth); return 0; case 3: /* NAK */ @@ -1464,8 +1311,8 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, goto bail_stop; /* The last valid PSN is the previous PSN. */ update_last_psn(qp, psn - 1); - switch ((aeth >> HFI1_AETH_CREDIT_SHIFT) & - HFI1_AETH_CREDIT_MASK) { + switch ((aeth >> IB_AETH_CREDIT_SHIFT) & + IB_AETH_CREDIT_MASK) { case 0: /* PSN sequence error */ ibp->rvp.n_seq_naks++; /* @@ -1474,7 +1321,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, * RDMA READ response which terminates the RDMA * READ. */ - restart_rc(qp, psn, 0); + hfi1_restart_rc(qp, psn, 0); hfi1_schedule_send(qp); break; @@ -1513,7 +1360,7 @@ reserved: } /* cannot be reached */ bail_stop: - hfi1_stop_rc_timers(qp); + rvt_stop_rc_timers(qp); return ret; } @@ -1528,7 +1375,7 @@ static void rdma_seq_err(struct rvt_qp *qp, struct hfi1_ibport *ibp, u32 psn, lockdep_assert_held(&qp->s_lock); /* Remove QP from retry timer */ - hfi1_stop_rc_timers(qp); + rvt_stop_rc_timers(qp); wqe = rvt_get_swqe_ptr(qp, qp->s_acked); @@ -1542,7 +1389,7 @@ static void rdma_seq_err(struct rvt_qp *qp, struct hfi1_ibport *ibp, u32 psn, ibp->rvp.n_rdma_seq++; qp->r_flags |= RVT_R_RDMAR_SEQ; - restart_rc(qp, qp->s_last_psn + 1, 0); + hfi1_restart_rc(qp, qp->s_last_psn + 1, 0); if (list_empty(&qp->rspwait)) { qp->r_flags |= RVT_R_RSP_SEND; rvt_get_qp(qp); @@ -1586,7 +1433,7 @@ static void rc_rcv_resp(struct hfi1_ibport *ibp, /* Ignore invalid responses. */ smp_read_barrier_depends(); /* see post_one_send */ - if (cmp_psn(psn, ACCESS_ONCE(qp->s_next_psn)) >= 0) + if (cmp_psn(psn, READ_ONCE(qp->s_next_psn)) >= 0) goto ack_done; /* Ignore duplicate responses. */ @@ -1595,8 +1442,8 @@ static void rc_rcv_resp(struct hfi1_ibport *ibp, /* Update credits for "ghost" ACKs */ if (diff == 0 && opcode == OP(ACKNOWLEDGE)) { aeth = be32_to_cpu(ohdr->u.aeth); - if ((aeth >> 29) == 0) - hfi1_get_credit(qp, aeth); + if ((aeth >> IB_AETH_NAK_SHIFT) == 0) + rvt_get_credit(qp, aeth); } goto ack_done; } @@ -1656,8 +1503,7 @@ read_middle: * We got a response so update the timeout. * 4.096 usec. * (1 << qp->timeout) */ - qp->s_flags |= RVT_S_TIMER; - mod_timer(&qp->s_timer, jiffies + qp->timeout_jiffies); + rvt_mod_retry_timer(qp); if (qp->s_flags & RVT_S_WAIT_ACK) { qp->s_flags &= ~RVT_S_WAIT_ACK; hfi1_schedule_send(qp); @@ -1673,7 +1519,7 @@ read_middle: qp->s_rdma_read_len -= pmtu; update_last_psn(qp, psn); spin_unlock_irqrestore(&qp->s_lock, flags); - hfi1_copy_sge(&qp->s_rdma_read_sge, data, pmtu, 0, 0); + hfi1_copy_sge(&qp->s_rdma_read_sge, data, pmtu, false, false); goto bail; case OP(RDMA_READ_RESPONSE_ONLY): @@ -1681,7 +1527,7 @@ read_middle: if (!do_rc_ack(qp, aeth, psn, opcode, 0, rcd)) goto ack_done; /* Get the number of bytes the message was padded by. */ - pad = (be32_to_cpu(ohdr->bth[0]) >> 20) & 3; + pad = ib_bth_get_pad(ohdr); /* * Check that the data size is >= 0 && <= pmtu. * Remember to account for ICRC (4). @@ -1705,7 +1551,7 @@ read_middle: if (unlikely(wqe->wr.opcode != IB_WR_RDMA_READ)) goto ack_op_err; /* Get the number of bytes the message was padded by. */ - pad = (be32_to_cpu(ohdr->bth[0]) >> 20) & 3; + pad = ib_bth_get_pad(ohdr); /* * Check that the data size is >= 1 && <= pmtu. * Remember to account for ICRC (4). @@ -1717,7 +1563,7 @@ read_last: if (unlikely(tlen != qp->s_rdma_read_len)) goto ack_len_err; aeth = be32_to_cpu(ohdr->u.aeth); - hfi1_copy_sge(&qp->s_rdma_read_sge, data, tlen, 0, 0); + hfi1_copy_sge(&qp->s_rdma_read_sge, data, tlen, false, false); WARN_ON(qp->s_rdma_read_sge.num_sge); (void)do_rc_ack(qp, aeth, psn, OP(RDMA_READ_RESPONSE_LAST), 0, rcd); @@ -1757,9 +1603,7 @@ static inline void rc_defered_ack(struct hfi1_ctxtdata *rcd, static inline void rc_cancel_ack(struct rvt_qp *qp) { - struct hfi1_qp_priv *priv = qp->priv; - - priv->r_adefered = 0; + qp->r_adefered = 0; if (list_empty(&qp->rspwait)) return; list_del_init(&qp->rspwait); @@ -1786,7 +1630,7 @@ static noinline int rc_rcv_error(struct ib_other_headers *ohdr, void *data, struct rvt_qp *qp, u32 opcode, u32 psn, int diff, struct hfi1_ctxtdata *rcd) { - struct hfi1_ibport *ibp = to_iport(qp->ibqp.device, qp->port_num); + struct hfi1_ibport *ibp = rcd_to_iport(rcd); struct rvt_ack_entry *e; unsigned long flags; u8 i, prev; @@ -1961,25 +1805,6 @@ send_ack: return 0; } -void hfi1_rc_error(struct rvt_qp *qp, enum ib_wc_status err) -{ - unsigned long flags; - int lastwqe; - - spin_lock_irqsave(&qp->s_lock, flags); - lastwqe = rvt_error_qp(qp, err); - spin_unlock_irqrestore(&qp->s_lock, flags); - - if (lastwqe) { - struct ib_event ev; - - ev.device = qp->ibqp.device; - ev.element.qp = &qp->ibqp; - ev.event = IB_EVENT_QP_LAST_WQE_REACHED; - qp->ibqp.event_handler(&ev, qp->ibqp.qp_context); - } -} - static inline void update_ack_queue(struct rvt_qp *qp, unsigned n) { unsigned next; @@ -2095,7 +1920,7 @@ void hfi1_rc_rcv(struct hfi1_packet *packet) void *data = packet->ebuf; u32 tlen = packet->tlen; struct rvt_qp *qp = packet->qp; - struct hfi1_ibport *ibp = to_iport(qp->ibqp.device, qp->port_num); + struct hfi1_ibport *ibp = rcd_to_iport(rcd); struct ib_other_headers *ohdr = packet->ohdr; u32 bth0, opcode; u32 hdrsize = packet->hlen; @@ -2106,8 +1931,9 @@ void hfi1_rc_rcv(struct hfi1_packet *packet) int diff; struct ib_reth *reth; unsigned long flags; - int ret, is_fecn = 0; - int copy_last = 0; + int ret; + bool is_fecn = false; + bool copy_last = false; u32 rkey; lockdep_assert_held(&qp->r_lock); @@ -2118,7 +1944,7 @@ void hfi1_rc_rcv(struct hfi1_packet *packet) is_fecn = process_ecn(qp, packet, false); psn = be32_to_cpu(ohdr->bth[2]); - opcode = (bth0 >> 24) & 0xff; + opcode = ib_bth_get_opcode(ohdr); /* * Process responses (ACKs) before anything else. Note that the @@ -2180,7 +2006,7 @@ void hfi1_rc_rcv(struct hfi1_packet *packet) } if (qp->state == IB_QPS_RTR && !(qp->r_flags & RVT_R_COMM_EST)) - qp_comm_est(qp); + rvt_comm_est(qp); /* OK, process the packet. */ switch (opcode) { @@ -2201,7 +2027,7 @@ send_middle: qp->r_rcv_len += pmtu; if (unlikely(qp->r_rcv_len > qp->r_len)) goto nack_inv; - hfi1_copy_sge(&qp->r_sge, data, pmtu, 1, 0); + hfi1_copy_sge(&qp->r_sge, data, pmtu, true, false); break; case OP(RDMA_WRITE_LAST_WITH_IMMEDIATE): @@ -2241,7 +2067,7 @@ send_last_inv: wc.wc_flags = IB_WC_WITH_INVALIDATE; goto send_last; case OP(RDMA_WRITE_LAST): - copy_last = ibpd_to_rvtpd(qp->ibqp.pd)->user; + copy_last = rvt_is_user_qp(qp); /* fall through */ case OP(SEND_LAST): no_immediate_data: @@ -2249,7 +2075,7 @@ no_immediate_data: wc.ex.imm_data = 0; send_last: /* Get the number of bytes the message was padded by. */ - pad = (bth0 >> 20) & 3; + pad = ib_bth_get_pad(ohdr); /* Check for invalid length. */ /* LAST len should be >= 1 */ if (unlikely(tlen < (hdrsize + pad + 4))) @@ -2259,7 +2085,7 @@ send_last: wc.byte_len = tlen + qp->r_rcv_len; if (unlikely(wc.byte_len > qp->r_len)) goto nack_inv; - hfi1_copy_sge(&qp->r_sge, data, tlen, 1, copy_last); + hfi1_copy_sge(&qp->r_sge, data, tlen, true, copy_last); rvt_put_ss(&qp->r_sge); qp->r_msn++; if (!__test_and_clear_bit(RVT_R_WRID_VALID, &qp->r_aflags)) @@ -2273,7 +2099,7 @@ send_last: wc.opcode = IB_WC_RECV; wc.qp = &qp->ibqp; wc.src_qp = qp->remote_qpn; - wc.slid = qp->remote_ah_attr.dlid; + wc.slid = rdma_ah_get_dlid(&qp->remote_ah_attr); /* * It seems that IB mandates the presence of an SL in a * work completion only for the UD transport (see section @@ -2285,7 +2111,7 @@ send_last: * * See also OPA Vol. 1, section 9.7.6, and table 9-17. */ - wc.sl = qp->remote_ah_attr.sl; + wc.sl = rdma_ah_get_sl(&qp->remote_ah_attr); /* zero fields that are N/A */ wc.vendor_err = 0; wc.pkey_index = 0; @@ -2297,7 +2123,7 @@ send_last: break; case OP(RDMA_WRITE_ONLY): - copy_last = 1; + copy_last = rvt_is_user_qp(qp); /* fall through */ case OP(RDMA_WRITE_FIRST): case OP(RDMA_WRITE_ONLY_WITH_IMMEDIATE): @@ -2333,8 +2159,11 @@ send_last: ret = hfi1_rvt_get_rwqe(qp, 1); if (ret < 0) goto nack_op_err; - if (!ret) + if (!ret) { + /* peer will send again */ + rvt_put_ss(&qp->r_sge); goto rnr_nak; + } wc.ex.imm_data = ohdr->u.rc.imm_data; wc.wc_flags = IB_WC_WITH_IMM; goto send_last; @@ -2485,13 +2314,11 @@ send_last: qp->r_nak_state = 0; /* Send an ACK if requested or required. */ if (psn & IB_BTH_REQ_ACK) { - struct hfi1_qp_priv *priv = qp->priv; - if (packet->numpkt == 0) { rc_cancel_ack(qp); goto send_ack; } - if (priv->r_adefered >= HFI1_PSN_CREDIT) { + if (qp->r_adefered >= HFI1_PSN_CREDIT) { rc_cancel_ack(qp); goto send_ack; } @@ -2499,7 +2326,7 @@ send_last: rc_cancel_ack(qp); goto send_ack; } - priv->r_adefered++; + qp->r_adefered++; rc_defered_ack(rcd, qp); } return; @@ -2512,7 +2339,7 @@ rnr_nak: return; nack_op_err: - hfi1_rc_error(qp, IB_WC_LOC_QP_OP_ERR); + rvt_rc_error(qp, IB_WC_LOC_QP_OP_ERR); qp->r_nak_state = IB_NAK_REMOTE_OPERATIONAL_ERROR; qp->r_ack_psn = qp->r_psn; /* Queue NAK for later */ @@ -2522,7 +2349,7 @@ nack_op_err: nack_inv_unlck: spin_unlock_irqrestore(&qp->s_lock, flags); nack_inv: - hfi1_rc_error(qp, IB_WC_LOC_QP_OP_ERR); + rvt_rc_error(qp, IB_WC_LOC_QP_OP_ERR); qp->r_nak_state = IB_NAK_INVALID_REQUEST; qp->r_ack_psn = qp->r_psn; /* Queue NAK for later */ @@ -2532,7 +2359,7 @@ nack_inv: nack_acc_unlck: spin_unlock_irqrestore(&qp->s_lock, flags); nack_acc: - hfi1_rc_error(qp, IB_WC_LOC_PROT_ERR); + rvt_rc_error(qp, IB_WC_LOC_PROT_ERR); qp->r_nak_state = IB_NAK_REMOTE_ACCESS_ERROR; qp->r_ack_psn = qp->r_psn; send_ack: @@ -2547,7 +2374,7 @@ void hfi1_rc_hdrerr( { int has_grh = rcv_flags & HFI1_HAS_GRH; struct ib_other_headers *ohdr; - struct hfi1_ibport *ibp = to_iport(qp->ibqp.device, qp->port_num); + struct hfi1_ibport *ibp = rcd_to_iport(rcd); int diff; u32 opcode; u32 psn, bth0; @@ -2562,7 +2389,7 @@ void hfi1_rc_hdrerr( return; psn = be32_to_cpu(ohdr->bth[2]); - opcode = (bth0 >> 24) & 0xff; + opcode = ib_bth_get_opcode(ohdr); /* Only deal with RDMA Writes for now */ if (opcode < IB_OPCODE_RC_RDMA_READ_RESPONSE_FIRST) { diff --git a/drivers/infiniband/hw/hfi1/ruc.c b/drivers/infiniband/hw/hfi1/ruc.c index 717ed4b159d3..3a17daba28a9 100644 --- a/drivers/infiniband/hw/hfi1/ruc.c +++ b/drivers/infiniband/hw/hfi1/ruc.c @@ -1,5 +1,5 @@ /* - * Copyright(c) 2015, 2016 Intel Corporation. + * Copyright(c) 2015 - 2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -54,44 +54,6 @@ #include "trace.h" /* - * Convert the AETH RNR timeout code into the number of microseconds. - */ -const u32 ib_hfi1_rnr_table[32] = { - 655360, /* 00: 655.36 */ - 10, /* 01: .01 */ - 20, /* 02 .02 */ - 30, /* 03: .03 */ - 40, /* 04: .04 */ - 60, /* 05: .06 */ - 80, /* 06: .08 */ - 120, /* 07: .12 */ - 160, /* 08: .16 */ - 240, /* 09: .24 */ - 320, /* 0A: .32 */ - 480, /* 0B: .48 */ - 640, /* 0C: .64 */ - 960, /* 0D: .96 */ - 1280, /* 0E: 1.28 */ - 1920, /* 0F: 1.92 */ - 2560, /* 10: 2.56 */ - 3840, /* 11: 3.84 */ - 5120, /* 12: 5.12 */ - 7680, /* 13: 7.68 */ - 10240, /* 14: 10.24 */ - 15360, /* 15: 15.36 */ - 20480, /* 16: 20.48 */ - 30720, /* 17: 30.72 */ - 40960, /* 18: 40.96 */ - 61440, /* 19: 61.44 */ - 81920, /* 1A: 81.92 */ - 122880, /* 1B: 122.88 */ - 163840, /* 1C: 163.84 */ - 245760, /* 1D: 245.76 */ - 327680, /* 1E: 327.68 */ - 491520 /* 1F: 491.52 */ -}; - -/* * Validate a RWQE and fill in the SGE state. * Return 1 if OK. */ @@ -257,72 +219,84 @@ int hfi1_ruc_check_hdr(struct hfi1_ibport *ibp, struct ib_header *hdr, { __be64 guid; unsigned long flags; - u8 sc5 = ibp->sl_to_sc[qp->remote_ah_attr.sl]; + u8 sc5 = ibp->sl_to_sc[rdma_ah_get_sl(&qp->remote_ah_attr)]; if (qp->s_mig_state == IB_MIG_ARMED && (bth0 & IB_BTH_MIG_REQ)) { if (!has_grh) { - if (qp->alt_ah_attr.ah_flags & IB_AH_GRH) + if (rdma_ah_get_ah_flags(&qp->alt_ah_attr) & + IB_AH_GRH) goto err; } else { - if (!(qp->alt_ah_attr.ah_flags & IB_AH_GRH)) + const struct ib_global_route *grh; + + if (!(rdma_ah_get_ah_flags(&qp->alt_ah_attr) & + IB_AH_GRH)) goto err; - guid = get_sguid(ibp, qp->alt_ah_attr.grh.sgid_index); + grh = rdma_ah_read_grh(&qp->alt_ah_attr); + guid = get_sguid(ibp, grh->sgid_index); if (!gid_ok(&hdr->u.l.grh.dgid, ibp->rvp.gid_prefix, guid)) goto err; if (!gid_ok( &hdr->u.l.grh.sgid, - qp->alt_ah_attr.grh.dgid.global.subnet_prefix, - qp->alt_ah_attr.grh.dgid.global.interface_id)) + grh->dgid.global.subnet_prefix, + grh->dgid.global.interface_id)) goto err; } - if (unlikely(rcv_pkey_check(ppd_from_ibp(ibp), (u16)bth0, - sc5, be16_to_cpu(hdr->lrh[3])))) { + if (unlikely(rcv_pkey_check(ppd_from_ibp(ibp), (u16)bth0, sc5, + ib_get_slid(hdr)))) { hfi1_bad_pqkey(ibp, OPA_TRAP_BAD_P_KEY, (u16)bth0, - (be16_to_cpu(hdr->lrh[0]) >> 4) & 0xF, + ib_get_sl(hdr), 0, qp->ibqp.qp_num, - be16_to_cpu(hdr->lrh[3]), - be16_to_cpu(hdr->lrh[1])); + ib_get_slid(hdr), + ib_get_dlid(hdr)); goto err; } /* Validate the SLID. See Ch. 9.6.1.5 and 17.2.8 */ - if (be16_to_cpu(hdr->lrh[3]) != qp->alt_ah_attr.dlid || - ppd_from_ibp(ibp)->port != qp->alt_ah_attr.port_num) + if (ib_get_slid(hdr) != + rdma_ah_get_dlid(&qp->alt_ah_attr) || + ppd_from_ibp(ibp)->port != + rdma_ah_get_port_num(&qp->alt_ah_attr)) goto err; spin_lock_irqsave(&qp->s_lock, flags); hfi1_migrate_qp(qp); spin_unlock_irqrestore(&qp->s_lock, flags); } else { if (!has_grh) { - if (qp->remote_ah_attr.ah_flags & IB_AH_GRH) + if (rdma_ah_get_ah_flags(&qp->remote_ah_attr) & + IB_AH_GRH) goto err; } else { - if (!(qp->remote_ah_attr.ah_flags & IB_AH_GRH)) + const struct ib_global_route *grh; + + if (!(rdma_ah_get_ah_flags(&qp->remote_ah_attr) & + IB_AH_GRH)) goto err; - guid = get_sguid(ibp, - qp->remote_ah_attr.grh.sgid_index); + grh = rdma_ah_read_grh(&qp->remote_ah_attr); + guid = get_sguid(ibp, grh->sgid_index); if (!gid_ok(&hdr->u.l.grh.dgid, ibp->rvp.gid_prefix, guid)) goto err; if (!gid_ok( &hdr->u.l.grh.sgid, - qp->remote_ah_attr.grh.dgid.global.subnet_prefix, - qp->remote_ah_attr.grh.dgid.global.interface_id)) + grh->dgid.global.subnet_prefix, + grh->dgid.global.interface_id)) goto err; } - if (unlikely(rcv_pkey_check(ppd_from_ibp(ibp), (u16)bth0, - sc5, be16_to_cpu(hdr->lrh[3])))) { + if (unlikely(rcv_pkey_check(ppd_from_ibp(ibp), (u16)bth0, sc5, + ib_get_slid(hdr)))) { hfi1_bad_pqkey(ibp, OPA_TRAP_BAD_P_KEY, (u16)bth0, - (be16_to_cpu(hdr->lrh[0]) >> 4) & 0xF, + ib_get_sl(hdr), 0, qp->ibqp.qp_num, - be16_to_cpu(hdr->lrh[3]), - be16_to_cpu(hdr->lrh[1])); + ib_get_slid(hdr), + ib_get_dlid(hdr)); goto err; } /* Validate the SLID. See Ch. 9.6.1.5 */ - if (be16_to_cpu(hdr->lrh[3]) != qp->remote_ah_attr.dlid || + if (ib_get_slid(hdr) != + rdma_ah_get_dlid(&qp->remote_ah_attr) || ppd_from_ibp(ibp)->port != qp->port_num) goto err; if (qp->s_mig_state == IB_MIG_REARM && @@ -358,10 +332,9 @@ static void ruc_loopback(struct rvt_qp *sqp) u64 sdata; atomic64_t *maddr; enum ib_wc_status send_status; - int release; + bool release; int ret; - int copy_last = 0; - u32 to; + bool copy_last = false; int local_ops = 0; rcu_read_lock(); @@ -425,7 +398,7 @@ again: memset(&wc, 0, sizeof(wc)); send_status = IB_WC_SUCCESS; - release = 1; + release = true; sqp->s_sge.sge = wqe->sg_list[0]; sqp->s_sge.sg_list = wqe->sg_list + 1; sqp->s_sge.num_sge = wqe->wr.num_sge; @@ -476,7 +449,7 @@ send: /* skip copy_last set and qp_access_flags recheck */ goto do_write; case IB_WR_RDMA_WRITE: - copy_last = ibpd_to_rvtpd(qp->ibqp.pd)->user; + copy_last = rvt_is_user_qp(qp); if (unlikely(!(qp->qp_access_flags & IB_ACCESS_REMOTE_WRITE))) goto inv_err; do_write: @@ -500,7 +473,7 @@ do_write: wqe->rdma_wr.rkey, IB_ACCESS_REMOTE_READ))) goto acc_err; - release = 0; + release = false; sqp->s_sge.sg_list = NULL; sqp->s_sge.num_sge = 1; qp->r_sge.sge = wqe->sg_list[0]; @@ -581,8 +554,8 @@ do_write: wc.byte_len = wqe->length; wc.qp = &qp->ibqp; wc.src_qp = qp->remote_qpn; - wc.slid = qp->remote_ah_attr.dlid; - wc.sl = qp->remote_ah_attr.sl; + wc.slid = rdma_ah_get_dlid(&qp->remote_ah_attr); + wc.sl = rdma_ah_get_sl(&qp->remote_ah_attr); wc.port_num = 1; /* Signal completion event if the solicited bit is set. */ rvt_cq_enter(ibcq_to_rvtcq(qp->ibqp.recv_cq), &wc, @@ -618,8 +591,8 @@ rnr_nak: spin_lock_irqsave(&sqp->s_lock, flags); if (!(ib_rvt_state_ops[sqp->state] & RVT_PROCESS_RECV_OK)) goto clr_busy; - to = ib_hfi1_rnr_table[qp->r_min_rnr_timer]; - hfi1_add_rnr_timer(sqp, to); + rvt_add_rnr_timer(sqp, qp->r_min_rnr_timer << + IB_AETH_CREDIT_SHIFT); goto clr_busy; op_err: @@ -637,7 +610,7 @@ acc_err: wc.status = IB_WC_LOC_PROT_ERR; err: /* responder goes to error state */ - hfi1_rc_error(qp, wc.status); + rvt_rc_error(qp, wc.status); serr: spin_lock_irqsave(&sqp->s_lock, flags); @@ -676,7 +649,7 @@ done: * Return the size of the header in 32 bit words. */ u32 hfi1_make_grh(struct hfi1_ibport *ibp, struct ib_grh *hdr, - struct ib_global_route *grh, u32 hwords, u32 nwords) + const struct ib_global_route *grh, u32 hwords, u32 nwords) { hdr->version_tclass_flow = cpu_to_be32((IB_GRH_VERSION << IB_GRH_VERSION_SHIFT) | @@ -770,15 +743,17 @@ void hfi1_make_ruc_header(struct rvt_qp *qp, struct ib_other_headers *ohdr, extra_bytes = -ps->s_txreq->s_cur_size & 3; nwords = (ps->s_txreq->s_cur_size + extra_bytes) >> 2; lrh0 = HFI1_LRH_BTH; - if (unlikely(qp->remote_ah_attr.ah_flags & IB_AH_GRH)) { - qp->s_hdrwords += hfi1_make_grh(ibp, - &ps->s_txreq->phdr.hdr.u.l.grh, - &qp->remote_ah_attr.grh, - qp->s_hdrwords, nwords); + if (unlikely(rdma_ah_get_ah_flags(&qp->remote_ah_attr) & IB_AH_GRH)) { + qp->s_hdrwords += + hfi1_make_grh(ibp, + &ps->s_txreq->phdr.hdr.u.l.grh, + rdma_ah_read_grh(&qp->remote_ah_attr), + qp->s_hdrwords, nwords); lrh0 = HFI1_LRH_GRH; middle = 0; } - lrh0 |= (priv->s_sc & 0xf) << 12 | (qp->remote_ah_attr.sl & 0xf) << 4; + lrh0 |= (priv->s_sc & 0xf) << 12 | + (rdma_ah_get_sl(&qp->remote_ah_attr) & 0xf) << 4; /* * reset s_ahg/AHG fields * @@ -802,11 +777,13 @@ void hfi1_make_ruc_header(struct rvt_qp *qp, struct ib_other_headers *ohdr, else qp->s_flags &= ~RVT_S_AHG_VALID; ps->s_txreq->phdr.hdr.lrh[0] = cpu_to_be16(lrh0); - ps->s_txreq->phdr.hdr.lrh[1] = cpu_to_be16(qp->remote_ah_attr.dlid); + ps->s_txreq->phdr.hdr.lrh[1] = + cpu_to_be16(rdma_ah_get_dlid(&qp->remote_ah_attr)); ps->s_txreq->phdr.hdr.lrh[2] = cpu_to_be16(qp->s_hdrwords + nwords + SIZE_OF_CRC); - ps->s_txreq->phdr.hdr.lrh[3] = cpu_to_be16(ppd_from_ibp(ibp)->lid | - qp->remote_ah_attr.src_path_bits); + ps->s_txreq->phdr.hdr.lrh[3] = + cpu_to_be16(ppd_from_ibp(ibp)->lid | + rdma_ah_get_path_bits(&qp->remote_ah_attr)); bth0 |= hfi1_get_pkey(ibp, qp->s_pkey_index); bth0 |= extra_bytes << 20; ohdr->bth[0] = cpu_to_be32(bth0); @@ -814,7 +791,7 @@ void hfi1_make_ruc_header(struct rvt_qp *qp, struct ib_other_headers *ohdr, if (qp->s_flags & RVT_S_ECN) { qp->s_flags &= ~RVT_S_ECN; /* we recently received a FECN, so return a BECN */ - bth1 |= (HFI1_BECN_MASK << HFI1_BECN_SHIFT); + bth1 |= (IB_BECN_MASK << IB_BECN_SHIFT); } ohdr->bth[1] = cpu_to_be32(bth1); ohdr->bth[2] = cpu_to_be32(bth2); @@ -823,59 +800,102 @@ void hfi1_make_ruc_header(struct rvt_qp *qp, struct ib_other_headers *ohdr, /* when sending, force a reschedule every one of these periods */ #define SEND_RESCHED_TIMEOUT (5 * HZ) /* 5s in jiffies */ +/** + * schedule_send_yield - test for a yield required for QP send engine + * @timeout: Final time for timeout slice for jiffies + * @qp: a pointer to QP + * @ps: a pointer to a structure with commonly lookup values for + * the the send engine progress + * + * This routine checks if the time slice for the QP has expired + * for RC QPs, if so an additional work entry is queued. At this + * point, other QPs have an opportunity to be scheduled. It + * returns true if a yield is required, otherwise, false + * is returned. + */ +static bool schedule_send_yield(struct rvt_qp *qp, + struct hfi1_pkt_state *ps) +{ + if (unlikely(time_after(jiffies, ps->timeout))) { + if (!ps->in_thread || + workqueue_congested(ps->cpu, ps->ppd->hfi1_wq)) { + spin_lock_irqsave(&qp->s_lock, ps->flags); + qp->s_flags &= ~RVT_S_BUSY; + hfi1_schedule_send(qp); + spin_unlock_irqrestore(&qp->s_lock, ps->flags); + this_cpu_inc(*ps->ppd->dd->send_schedule); + trace_hfi1_rc_expired_time_slice(qp, true); + return true; + } + + cond_resched(); + this_cpu_inc(*ps->ppd->dd->send_schedule); + ps->timeout = jiffies + ps->timeout_int; + } + + trace_hfi1_rc_expired_time_slice(qp, false); + return false; +} + +void hfi1_do_send_from_rvt(struct rvt_qp *qp) +{ + hfi1_do_send(qp, false); +} + void _hfi1_do_send(struct work_struct *work) { struct iowait *wait = container_of(work, struct iowait, iowork); struct rvt_qp *qp = iowait_to_qp(wait); - hfi1_do_send(qp); + hfi1_do_send(qp, true); } /** * hfi1_do_send - perform a send on a QP * @work: contains a pointer to the QP + * @in_thread: true if in a workqueue thread * * Process entries in the send work queue until credit or queue is * exhausted. Only allow one CPU to send a packet per QP. * Otherwise, two threads could send packets out of order. */ -void hfi1_do_send(struct rvt_qp *qp) +void hfi1_do_send(struct rvt_qp *qp, bool in_thread) { struct hfi1_pkt_state ps; struct hfi1_qp_priv *priv = qp->priv; int (*make_req)(struct rvt_qp *qp, struct hfi1_pkt_state *ps); - unsigned long timeout; - unsigned long timeout_int; - int cpu; ps.dev = to_idev(qp->ibqp.device); ps.ibp = to_iport(qp->ibqp.device, qp->port_num); ps.ppd = ppd_from_ibp(ps.ibp); + ps.in_thread = in_thread; + + trace_hfi1_rc_do_send(qp, in_thread); switch (qp->ibqp.qp_type) { case IB_QPT_RC: - if (!loopback && ((qp->remote_ah_attr.dlid & ~((1 << ps.ppd->lmc - ) - 1)) == - ps.ppd->lid)) { + if (!loopback && ((rdma_ah_get_dlid(&qp->remote_ah_attr) & + ~((1 << ps.ppd->lmc) - 1)) == + ps.ppd->lid)) { ruc_loopback(qp); return; } make_req = hfi1_make_rc_req; - timeout_int = (qp->timeout_jiffies); + ps.timeout_int = qp->timeout_jiffies; break; case IB_QPT_UC: - if (!loopback && ((qp->remote_ah_attr.dlid & ~((1 << ps.ppd->lmc - ) - 1)) == - ps.ppd->lid)) { + if (!loopback && ((rdma_ah_get_dlid(&qp->remote_ah_attr) & + ~((1 << ps.ppd->lmc) - 1)) == + ps.ppd->lid)) { ruc_loopback(qp); return; } make_req = hfi1_make_uc_req; - timeout_int = SEND_RESCHED_TIMEOUT; + ps.timeout_int = SEND_RESCHED_TIMEOUT; break; default: make_req = hfi1_make_ud_req; - timeout_int = SEND_RESCHED_TIMEOUT; + ps.timeout_int = SEND_RESCHED_TIMEOUT; } spin_lock_irqsave(&qp->s_lock, ps.flags); @@ -888,9 +908,11 @@ void hfi1_do_send(struct rvt_qp *qp) qp->s_flags |= RVT_S_BUSY; - timeout = jiffies + (timeout_int) / 8; - cpu = priv->s_sde ? priv->s_sde->cpu : + ps.timeout_int = ps.timeout_int / 8; + ps.timeout = jiffies + ps.timeout_int; + ps.cpu = priv->s_sde ? priv->s_sde->cpu : cpumask_first(cpumask_of_node(ps.ppd->dd->node)); + /* insure a pre-built packet is handled */ ps.s_txreq = get_waiting_verbs_txreq(qp); do { @@ -906,28 +928,9 @@ void hfi1_do_send(struct rvt_qp *qp) /* Record that s_ahg is empty. */ qp->s_hdrwords = 0; /* allow other tasks to run */ - if (unlikely(time_after(jiffies, timeout))) { - if (workqueue_congested(cpu, - ps.ppd->hfi1_wq)) { - spin_lock_irqsave( - &qp->s_lock, - ps.flags); - qp->s_flags &= ~RVT_S_BUSY; - hfi1_schedule_send(qp); - spin_unlock_irqrestore( - &qp->s_lock, - ps.flags); - this_cpu_inc( - *ps.ppd->dd->send_schedule); - return; - } - if (!irqs_disabled()) { - cond_resched(); - this_cpu_inc( - *ps.ppd->dd->send_schedule); - } - timeout = jiffies + (timeout_int) / 8; - } + if (schedule_send_yield(qp, &ps)) + return; + spin_lock_irqsave(&qp->s_lock, ps.flags); } } while (make_req(qp, &ps)); @@ -948,8 +951,10 @@ void hfi1_send_complete(struct rvt_qp *qp, struct rvt_swqe *wqe, last = qp->s_last; old_last = last; + trace_hfi1_qp_send_completion(qp, wqe, last); if (++last >= qp->s_size) last = 0; + trace_hfi1_qp_send_completion(qp, wqe, last); qp->s_last = last; /* See post_send() */ barrier(); @@ -959,7 +964,10 @@ void hfi1_send_complete(struct rvt_qp *qp, struct rvt_swqe *wqe, qp->ibqp.qp_type == IB_QPT_GSI) atomic_dec(&ibah_to_rvtah(wqe->ud_wr.ah)->refcount); - rvt_qp_swqe_complete(qp, wqe, status); + rvt_qp_swqe_complete(qp, + wqe, + ib_hfi1_wc_opcode[wqe->wr.opcode], + status); if (qp->s_acked == old_last) qp->s_acked = last; diff --git a/drivers/infiniband/hw/hfi1/sdma.c b/drivers/infiniband/hw/hfi1/sdma.c index 1d81cac1fa6c..bfd0d5187e9b 100644 --- a/drivers/infiniband/hw/hfi1/sdma.c +++ b/drivers/infiniband/hw/hfi1/sdma.c @@ -856,7 +856,7 @@ struct sdma_engine *sdma_select_user_engine(struct hfi1_devdata *dd, { struct sdma_rht_node *rht_node; struct sdma_engine *sde = NULL; - const struct cpumask *current_mask = tsk_cpus_allowed(current); + const struct cpumask *current_mask = ¤t->cpus_allowed; unsigned long cpu_id; /* @@ -868,7 +868,7 @@ struct sdma_engine *sdma_select_user_engine(struct hfi1_devdata *dd, cpu_id = smp_processor_id(); rcu_read_lock(); - rht_node = rhashtable_lookup_fast(&dd->sdma_rht, &cpu_id, + rht_node = rhashtable_lookup_fast(dd->sdma_rht, &cpu_id, sdma_rht_params); if (rht_node && rht_node->map[vl]) { @@ -962,7 +962,12 @@ ssize_t sdma_set_cpu_to_sde_map(struct sdma_engine *sde, const char *buf, continue; } - rht_node = rhashtable_lookup_fast(&dd->sdma_rht, &cpu, + if (vl >= ARRAY_SIZE(rht_node->map)) { + ret = -EINVAL; + goto out; + } + + rht_node = rhashtable_lookup_fast(dd->sdma_rht, &cpu, sdma_rht_params); if (!rht_node) { rht_node = kzalloc(sizeof(*rht_node), GFP_KERNEL); @@ -982,7 +987,7 @@ ssize_t sdma_set_cpu_to_sde_map(struct sdma_engine *sde, const char *buf, rht_node->map[vl]->ctr = 1; rht_node->map[vl]->sde[0] = sde; - ret = rhashtable_insert_fast(&dd->sdma_rht, + ret = rhashtable_insert_fast(dd->sdma_rht, &rht_node->node, sdma_rht_params); if (ret) { @@ -1025,7 +1030,7 @@ ssize_t sdma_set_cpu_to_sde_map(struct sdma_engine *sde, const char *buf, if (cpumask_test_cpu(cpu, mask)) continue; - rht_node = rhashtable_lookup_fast(&dd->sdma_rht, &cpu, + rht_node = rhashtable_lookup_fast(dd->sdma_rht, &cpu, sdma_rht_params); if (rht_node) { bool empty = true; @@ -1049,7 +1054,7 @@ ssize_t sdma_set_cpu_to_sde_map(struct sdma_engine *sde, const char *buf, } if (empty) { - ret = rhashtable_remove_fast(&dd->sdma_rht, + ret = rhashtable_remove_fast(dd->sdma_rht, &rht_node->node, sdma_rht_params); WARN_ON(ret); @@ -1108,7 +1113,7 @@ void sdma_seqfile_dump_cpu_list(struct seq_file *s, struct sdma_rht_node *rht_node; int i, j; - rht_node = rhashtable_lookup_fast(&dd->sdma_rht, &cpuid, + rht_node = rhashtable_lookup_fast(dd->sdma_rht, &cpuid, sdma_rht_params); if (!rht_node) return; @@ -1322,6 +1327,12 @@ static void sdma_clean(struct hfi1_devdata *dd, size_t num_engines) synchronize_rcu(); kfree(dd->per_sdma); dd->per_sdma = NULL; + + if (dd->sdma_rht) { + rhashtable_free_and_destroy(dd->sdma_rht, sdma_rht_free, NULL); + kfree(dd->sdma_rht); + dd->sdma_rht = NULL; + } } /** @@ -1341,12 +1352,14 @@ int sdma_init(struct hfi1_devdata *dd, u8 port) { unsigned this_idx; struct sdma_engine *sde; + struct rhashtable *tmp_sdma_rht; u16 descq_cnt; void *curr_head; struct hfi1_pportdata *ppd = dd->pport + port; u32 per_sdma_credits; uint idle_cnt = sdma_idle_cnt; size_t num_engines = dd->chip_sdma_engines; + int ret = -ENOMEM; if (!HFI1_CAP_IS_KSET(SDMA)) { HFI1_CAP_CLEAR(SDMA_AHG); @@ -1378,7 +1391,7 @@ int sdma_init(struct hfi1_devdata *dd, u8 port) /* alloc memory for array of send engines */ dd->per_sdma = kcalloc(num_engines, sizeof(*dd->per_sdma), GFP_KERNEL); if (!dd->per_sdma) - return -ENOMEM; + return ret; idle_cnt = ns_to_cclock(dd, idle_cnt); if (!sdma_desct_intr) @@ -1507,18 +1520,27 @@ int sdma_init(struct hfi1_devdata *dd, u8 port) dd->flags |= HFI1_HAS_SEND_DMA; dd->flags |= idle_cnt ? HFI1_HAS_SDMA_TIMEOUT : 0; dd->num_sdma = num_engines; - if (sdma_map_init(dd, port, ppd->vls_operational, NULL)) + ret = sdma_map_init(dd, port, ppd->vls_operational, NULL); + if (ret < 0) goto bail; - if (rhashtable_init(&dd->sdma_rht, &sdma_rht_params)) + tmp_sdma_rht = kzalloc(sizeof(*tmp_sdma_rht), GFP_KERNEL); + if (!tmp_sdma_rht) { + ret = -ENOMEM; goto bail; + } + + ret = rhashtable_init(tmp_sdma_rht, &sdma_rht_params); + if (ret < 0) + goto bail; + dd->sdma_rht = tmp_sdma_rht; dd_dev_info(dd, "SDMA num_sdma: %u\n", dd->num_sdma); return 0; bail: sdma_clean(dd, num_engines); - return -ENOMEM; + return ret; } /** @@ -1604,7 +1626,6 @@ void sdma_exit(struct hfi1_devdata *dd) sdma_finalput(&sde->state); } sdma_clean(dd, dd->num_sdma); - rhashtable_free_and_destroy(&dd->sdma_rht, sdma_rht_free, NULL); } /* diff --git a/drivers/infiniband/hw/hfi1/sdma.h b/drivers/infiniband/hw/hfi1/sdma.h index 21f1e2834f37..64f10b8b5db8 100644 --- a/drivers/infiniband/hw/hfi1/sdma.h +++ b/drivers/infiniband/hw/hfi1/sdma.h @@ -966,34 +966,34 @@ void sdma_engine_interrupt(struct sdma_engine *sde, u64 status); * | mask | --/ |--------------------| * |--------------------------| -/ | * | * | actual_vls (max 8) | -/ |--------------------| - * |--------------------------| --/ | sde[n] -> eng n | + * |--------------------------| --/ | sde[n-1] -> eng n | * | vls (max 8) | -/ +--------------------+ * |--------------------------| --/ * | map[0] |-/ - * |--------------------------| +--------------------+ - * | map[1] |--- | mask | - * |--------------------------| \---- |--------------------| - * | * | \-- | sde[0] -> eng 1+n | - * | * | \---- |--------------------| - * | * | \->| sde[1] -> eng 2+n | - * |--------------------------| |--------------------| - * | map[vls - 1] |- | * | - * +--------------------------+ \- |--------------------| - * \- | sde[m] -> eng m+n | - * \ +--------------------+ + * |--------------------------| +---------------------+ + * | map[1] |--- | mask | + * |--------------------------| \---- |---------------------| + * | * | \-- | sde[0] -> eng 1+n | + * | * | \---- |---------------------| + * | * | \->| sde[1] -> eng 2+n | + * |--------------------------| |---------------------| + * | map[vls - 1] |- | * | + * +--------------------------+ \- |---------------------| + * \- | sde[m-1] -> eng m+n | + * \ +---------------------+ * \- * \ - * \- +--------------------+ - * \- | mask | - * \ |--------------------| - * \- | sde[0] -> eng 1+m+n| - * \- |--------------------| - * >| sde[1] -> eng 2+m+n| - * |--------------------| - * | * | - * |--------------------| - * | sde[o] -> eng o+m+n| - * +--------------------+ + * \- +----------------------+ + * \- | mask | + * \ |----------------------| + * \- | sde[0] -> eng 1+m+n | + * \- |----------------------| + * >| sde[1] -> eng 2+m+n | + * |----------------------| + * | * | + * |----------------------| + * | sde[o-1] -> eng o+m+n| + * +----------------------+ * */ diff --git a/drivers/infiniband/hw/hfi1/sysfs.c b/drivers/infiniband/hw/hfi1/sysfs.c index 919a5474e651..2f3bbcac1e34 100644 --- a/drivers/infiniband/hw/hfi1/sysfs.c +++ b/drivers/infiniband/hw/hfi1/sysfs.c @@ -1,5 +1,5 @@ /* - * Copyright(c) 2015, 2016 Intel Corporation. + * Copyright(c) 2015-2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -196,7 +196,8 @@ static const struct sysfs_ops port_cc_sysfs_ops = { }; static struct attribute *port_cc_default_attributes[] = { - &cc_prescan_attr.attr + &cc_prescan_attr.attr, + NULL }; static struct kobj_type port_cc_ktype = { @@ -542,7 +543,7 @@ static ssize_t show_nctxts(struct device *device, * give a more accurate picture of total contexts available. */ return scnprintf(buf, PAGE_SIZE, "%u\n", - min(dd->num_rcv_contexts - dd->first_user_ctxt, + min(dd->num_rcv_contexts - dd->first_dyn_alloc_ctxt, (u32)dd->sc_sizes[SC_USER].count)); } diff --git a/drivers/infiniband/hw/hfi1/trace.c b/drivers/infiniband/hw/hfi1/trace.c index 01f525cd985a..eafae487face 100644 --- a/drivers/infiniband/hw/hfi1/trace.c +++ b/drivers/infiniband/hw/hfi1/trace.c @@ -51,13 +51,12 @@ u8 ibhdr_exhdr_len(struct ib_header *hdr) { struct ib_other_headers *ohdr; u8 opcode; - u8 lnh = (u8)(be16_to_cpu(hdr->lrh[0]) & 3); - if (lnh == HFI1_LRH_BTH) + if (ib_get_lnh(hdr) == HFI1_LRH_BTH) ohdr = &hdr->u.oth; else ohdr = &hdr->u.l.oth; - opcode = be32_to_cpu(ohdr->bth[0]) >> 24; + opcode = ib_bth_get_opcode(ohdr); return hdr_len_by_opcode[opcode] == 0 ? 0 : hdr_len_by_opcode[opcode] - (12 + 8); } @@ -130,14 +129,14 @@ const char *parse_everbs_hdrs( case OP(RC, ACKNOWLEDGE): trace_seq_printf(p, AETH_PRN, be32_to_cpu(eh->aeth) >> 24, parse_syndrome(be32_to_cpu(eh->aeth) >> 24), - be32_to_cpu(eh->aeth) & HFI1_MSN_MASK); + be32_to_cpu(eh->aeth) & IB_MSN_MASK); break; /* aeth + atomicacketh */ case OP(RC, ATOMIC_ACKNOWLEDGE): trace_seq_printf(p, AETH_PRN " " ATOMICACKETH_PRN, be32_to_cpu(eh->at.aeth) >> 24, parse_syndrome(be32_to_cpu(eh->at.aeth) >> 24), - be32_to_cpu(eh->at.aeth) & HFI1_MSN_MASK, + be32_to_cpu(eh->at.aeth) & IB_MSN_MASK, ib_u64_get(&eh->at.atomic_ack_eth)); break; /* atomiceth */ diff --git a/drivers/infiniband/hw/hfi1/trace_ctxts.h b/drivers/infiniband/hw/hfi1/trace_ctxts.h index 26ae789e47cf..4eb4cc798035 100644 --- a/drivers/infiniband/hw/hfi1/trace_ctxts.h +++ b/drivers/infiniband/hw/hfi1/trace_ctxts.h @@ -57,12 +57,14 @@ #define UCTXT_FMT \ "cred:%u, credaddr:0x%llx, piobase:0x%p, rcvhdr_cnt:%u, " \ - "rcvbase:0x%llx, rcvegrc:%u, rcvegrb:0x%llx" + "rcvbase:0x%llx, rcvegrc:%u, rcvegrb:0x%llx, subctxt_cnt:%u" TRACE_EVENT(hfi1_uctxtdata, - TP_PROTO(struct hfi1_devdata *dd, struct hfi1_ctxtdata *uctxt), - TP_ARGS(dd, uctxt), + TP_PROTO(struct hfi1_devdata *dd, struct hfi1_ctxtdata *uctxt, + unsigned int subctxt), + TP_ARGS(dd, uctxt, subctxt), TP_STRUCT__entry(DD_DEV_ENTRY(dd) __field(unsigned int, ctxt) + __field(unsigned int, subctxt) __field(u32, credits) __field(u64, hw_free) __field(void __iomem *, piobase) @@ -70,9 +72,11 @@ TRACE_EVENT(hfi1_uctxtdata, __field(u64, rcvhdrq_dma) __field(u32, eager_cnt) __field(u64, rcvegr_dma) + __field(unsigned int, subctxt_cnt) ), TP_fast_assign(DD_DEV_ASSIGN(dd); __entry->ctxt = uctxt->ctxt; + __entry->subctxt = subctxt; __entry->credits = uctxt->sc->credits; __entry->hw_free = le64_to_cpu(*uctxt->sc->hw_free); __entry->piobase = uctxt->sc->base_addr; @@ -80,17 +84,20 @@ TRACE_EVENT(hfi1_uctxtdata, __entry->rcvhdrq_dma = uctxt->rcvhdrq_dma; __entry->eager_cnt = uctxt->egrbufs.alloced; __entry->rcvegr_dma = uctxt->egrbufs.rcvtids[0].dma; + __entry->subctxt_cnt = uctxt->subctxt_cnt; ), - TP_printk("[%s] ctxt %u " UCTXT_FMT, + TP_printk("[%s] ctxt %u:%u " UCTXT_FMT, __get_str(dev), __entry->ctxt, + __entry->subctxt, __entry->credits, __entry->hw_free, __entry->piobase, __entry->rcvhdrq_cnt, __entry->rcvhdrq_dma, __entry->eager_cnt, - __entry->rcvegr_dma + __entry->rcvegr_dma, + __entry->subctxt_cnt ) ); diff --git a/drivers/infiniband/hw/hfi1/trace_ibhdrs.h b/drivers/infiniband/hw/hfi1/trace_ibhdrs.h index 382fcda3a5f6..090f6b506953 100644 --- a/drivers/infiniband/hw/hfi1/trace_ibhdrs.h +++ b/drivers/infiniband/hw/hfi1/trace_ibhdrs.h @@ -139,11 +139,11 @@ DECLARE_EVENT_CLASS(hfi1_ibhdr_template, __entry->pkey = be32_to_cpu(ohdr->bth[0]) & 0xffff; __entry->f = - (be32_to_cpu(ohdr->bth[1]) >> HFI1_FECN_SHIFT) & - HFI1_FECN_MASK; + (be32_to_cpu(ohdr->bth[1]) >> IB_FECN_SHIFT) & + IB_FECN_MASK; __entry->b = - (be32_to_cpu(ohdr->bth[1]) >> HFI1_BECN_SHIFT) & - HFI1_BECN_MASK; + (be32_to_cpu(ohdr->bth[1]) >> IB_BECN_SHIFT) & + IB_BECN_MASK; __entry->qpn = be32_to_cpu(ohdr->bth[1]) & RVT_QPN_MASK; __entry->a = diff --git a/drivers/infiniband/hw/hfi1/trace_misc.h b/drivers/infiniband/hw/hfi1/trace_misc.h index d308454af7fd..deac77ddaeab 100644 --- a/drivers/infiniband/hw/hfi1/trace_misc.h +++ b/drivers/infiniband/hw/hfi1/trace_misc.h @@ -72,6 +72,54 @@ TRACE_EVENT(hfi1_interrupt, __entry->src) ); +#ifdef CONFIG_FAULT_INJECTION +TRACE_EVENT(hfi1_fault_opcode, + TP_PROTO(struct rvt_qp *qp, u8 opcode), + TP_ARGS(qp, opcode), + TP_STRUCT__entry(DD_DEV_ENTRY(dd_from_ibdev(qp->ibqp.device)) + __field(u32, qpn) + __field(u8, opcode) + ), + TP_fast_assign(DD_DEV_ASSIGN(dd_from_ibdev(qp->ibqp.device)) + __entry->qpn = qp->ibqp.qp_num; + __entry->opcode = opcode; + ), + TP_printk("[%s] qpn 0x%x opcode 0x%x", + __get_str(dev), __entry->qpn, __entry->opcode) +); + +TRACE_EVENT(hfi1_fault_packet, + TP_PROTO(struct hfi1_packet *packet), + TP_ARGS(packet), + TP_STRUCT__entry(DD_DEV_ENTRY(packet->rcd->ppd->dd) + __field(u64, eflags) + __field(u32, ctxt) + __field(u32, hlen) + __field(u32, tlen) + __field(u32, updegr) + __field(u32, etail) + ), + TP_fast_assign(DD_DEV_ASSIGN(packet->rcd->ppd->dd); + __entry->eflags = rhf_err_flags(packet->rhf); + __entry->ctxt = packet->rcd->ctxt; + __entry->hlen = packet->hlen; + __entry->tlen = packet->tlen; + __entry->updegr = packet->updegr; + __entry->etail = rhf_egr_index(packet->rhf); + ), + TP_printk( + "[%s] ctxt %d eflags 0x%llx hlen %d tlen %d updegr %d etail %d", + __get_str(dev), + __entry->ctxt, + __entry->eflags, + __entry->hlen, + __entry->tlen, + __entry->updegr, + __entry->etail + ) +); +#endif + #endif /* __HFI1_TRACE_MISC_H */ #undef TRACE_INCLUDE_PATH diff --git a/drivers/infiniband/hw/hfi1/trace_rc.h b/drivers/infiniband/hw/hfi1/trace_rc.h index 5ea5005f9f41..8ce476570462 100644 --- a/drivers/infiniband/hw/hfi1/trace_rc.h +++ b/drivers/infiniband/hw/hfi1/trace_rc.h @@ -1,5 +1,5 @@ /* -* Copyright(c) 2015, 2016 Intel Corporation. +* Copyright(c) 2015, 2016, 2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -104,11 +104,6 @@ DEFINE_EVENT(hfi1_rc_template, hfi1_ack, TP_ARGS(qp, psn) ); -DEFINE_EVENT(hfi1_rc_template, hfi1_timeout, - TP_PROTO(struct rvt_qp *qp, u32 psn), - TP_ARGS(qp, psn) -); - DEFINE_EVENT(hfi1_rc_template, hfi1_rcv_error, TP_PROTO(struct rvt_qp *qp, u32 psn), TP_ARGS(qp, psn) diff --git a/drivers/infiniband/hw/hfi1/trace_tx.h b/drivers/infiniband/hw/hfi1/trace_tx.h index 415d6be42c5d..c59809a7f121 100644 --- a/drivers/infiniband/hw/hfi1/trace_tx.h +++ b/drivers/infiniband/hw/hfi1/trace_tx.h @@ -633,6 +633,83 @@ DEFINE_EVENT(hfi1_bct_template, bct_get, TP_PROTO(struct hfi1_devdata *dd, struct buffer_control *bc), TP_ARGS(dd, bc)); +TRACE_EVENT( + hfi1_qp_send_completion, + TP_PROTO(struct rvt_qp *qp, struct rvt_swqe *wqe, u32 idx), + TP_ARGS(qp, wqe, idx), + TP_STRUCT__entry( + DD_DEV_ENTRY(dd_from_ibdev(qp->ibqp.device)) + __field(struct rvt_swqe *, wqe) + __field(u64, wr_id) + __field(u32, qpn) + __field(u32, qpt) + __field(u32, length) + __field(u32, idx) + __field(u32, ssn) + __field(enum ib_wr_opcode, opcode) + __field(int, send_flags) + ), + TP_fast_assign( + DD_DEV_ASSIGN(dd_from_ibdev(qp->ibqp.device)) + __entry->wqe = wqe; + __entry->wr_id = wqe->wr.wr_id; + __entry->qpn = qp->ibqp.qp_num; + __entry->qpt = qp->ibqp.qp_type; + __entry->length = wqe->length; + __entry->idx = idx; + __entry->ssn = wqe->ssn; + __entry->opcode = wqe->wr.opcode; + __entry->send_flags = wqe->wr.send_flags; + ), + TP_printk( + "[%s] qpn 0x%x qpt %u wqe %p idx %u wr_id %llx length %u ssn %u opcode %x send_flags %x", + __get_str(dev), + __entry->qpn, + __entry->qpt, + __entry->wqe, + __entry->idx, + __entry->wr_id, + __entry->length, + __entry->ssn, + __entry->opcode, + __entry->send_flags + ) +); + +DECLARE_EVENT_CLASS( + hfi1_do_send_template, + TP_PROTO(struct rvt_qp *qp, bool flag), + TP_ARGS(qp, flag), + TP_STRUCT__entry( + DD_DEV_ENTRY(dd_from_ibdev(qp->ibqp.device)) + __field(u32, qpn) + __field(bool, flag) + ), + TP_fast_assign( + DD_DEV_ASSIGN(dd_from_ibdev(qp->ibqp.device)) + __entry->qpn = qp->ibqp.qp_num; + __entry->flag = flag; + ), + TP_printk( + "[%s] qpn %x flag %d", + __get_str(dev), + __entry->qpn, + __entry->flag + ) +); + +DEFINE_EVENT( + hfi1_do_send_template, hfi1_rc_do_send, + TP_PROTO(struct rvt_qp *qp, bool flag), + TP_ARGS(qp, flag) +); + +DEFINE_EVENT( + hfi1_do_send_template, hfi1_rc_expired_time_slice, + TP_PROTO(struct rvt_qp *qp, bool flag), + TP_ARGS(qp, flag) +); + #endif /* __HFI1_TRACE_TX_H */ #undef TRACE_INCLUDE_PATH diff --git a/drivers/infiniband/hw/hfi1/uc.c b/drivers/infiniband/hw/hfi1/uc.c index b141a78ae38b..5da1e4546543 100644 --- a/drivers/infiniband/hw/hfi1/uc.c +++ b/drivers/infiniband/hw/hfi1/uc.c @@ -94,7 +94,7 @@ int hfi1_make_uc_req(struct rvt_qp *qp, struct hfi1_pkt_state *ps) } ohdr = &ps->s_txreq->phdr.hdr.u.oth; - if (qp->remote_ah_attr.ah_flags & IB_AH_GRH) + if (rdma_ah_get_ah_flags(&qp->remote_ah_attr) & IB_AH_GRH) ohdr = &ps->s_txreq->phdr.hdr.u.l.oth; /* Get the next send request. */ @@ -296,7 +296,7 @@ bail_no_tx: */ void hfi1_uc_rcv(struct hfi1_packet *packet) { - struct hfi1_ibport *ibp = &packet->rcd->ppd->ibport_data; + struct hfi1_ibport *ibp = rcd_to_iport(packet->rcd); struct ib_header *hdr = packet->hdr; u32 rcv_flags = packet->rcv_flags; void *data = packet->ebuf; @@ -320,7 +320,7 @@ void hfi1_uc_rcv(struct hfi1_packet *packet) process_ecn(qp, packet, true); psn = be32_to_cpu(ohdr->bth[2]); - opcode = (bth0 >> 24) & 0xff; + opcode = ib_bth_get_opcode(ohdr); /* Compare the PSN verses the expected PSN. */ if (unlikely(cmp_psn(psn, qp->r_psn) != 0)) { @@ -384,7 +384,7 @@ inv: } if (qp->state == IB_QPS_RTR && !(qp->r_flags & RVT_R_COMM_EST)) - qp_comm_est(qp); + rvt_comm_est(qp); /* OK, process the packet. */ switch (opcode) { @@ -419,7 +419,7 @@ send_first: qp->r_rcv_len += pmtu; if (unlikely(qp->r_rcv_len > qp->r_len)) goto rewind; - hfi1_copy_sge(&qp->r_sge, data, pmtu, 0, 0); + hfi1_copy_sge(&qp->r_sge, data, pmtu, false, false); break; case OP(SEND_LAST_WITH_IMMEDIATE): @@ -433,7 +433,7 @@ no_immediate_data: wc.wc_flags = 0; send_last: /* Get the number of bytes the message was padded by. */ - pad = (be32_to_cpu(ohdr->bth[0]) >> 20) & 3; + pad = ib_bth_get_pad(ohdr); /* Check for invalid length. */ /* LAST len should be >= 1 */ if (unlikely(tlen < (hdrsize + pad + 4))) @@ -444,14 +444,14 @@ send_last: if (unlikely(wc.byte_len > qp->r_len)) goto rewind; wc.opcode = IB_WC_RECV; - hfi1_copy_sge(&qp->r_sge, data, tlen, 0, 0); + hfi1_copy_sge(&qp->r_sge, data, tlen, false, false); rvt_put_ss(&qp->s_rdma_read_sge); last_imm: wc.wr_id = qp->r_wr_id; wc.status = IB_WC_SUCCESS; wc.qp = &qp->ibqp; wc.src_qp = qp->remote_qpn; - wc.slid = qp->remote_ah_attr.dlid; + wc.slid = rdma_ah_get_dlid(&qp->remote_ah_attr); /* * It seems that IB mandates the presence of an SL in a * work completion only for the UD transport (see section @@ -463,7 +463,7 @@ last_imm: * * See also OPA Vol. 1, section 9.7.6, and table 9-17. */ - wc.sl = qp->remote_ah_attr.sl; + wc.sl = rdma_ah_get_sl(&qp->remote_ah_attr); /* zero fields that are N/A */ wc.vendor_err = 0; wc.pkey_index = 0; @@ -519,7 +519,7 @@ rdma_first: qp->r_rcv_len += pmtu; if (unlikely(qp->r_rcv_len > qp->r_len)) goto drop; - hfi1_copy_sge(&qp->r_sge, data, pmtu, 1, 0); + hfi1_copy_sge(&qp->r_sge, data, pmtu, true, false); break; case OP(RDMA_WRITE_LAST_WITH_IMMEDIATE): @@ -528,7 +528,7 @@ rdma_last_imm: wc.wc_flags = IB_WC_WITH_IMM; /* Get the number of bytes the message was padded by. */ - pad = (be32_to_cpu(ohdr->bth[0]) >> 20) & 3; + pad = ib_bth_get_pad(ohdr); /* Check for invalid length. */ /* LAST len should be >= 1 */ if (unlikely(tlen < (hdrsize + pad + 4))) @@ -548,14 +548,14 @@ rdma_last_imm: } wc.byte_len = qp->r_len; wc.opcode = IB_WC_RECV_RDMA_WITH_IMM; - hfi1_copy_sge(&qp->r_sge, data, tlen, 1, 0); + hfi1_copy_sge(&qp->r_sge, data, tlen, true, false); rvt_put_ss(&qp->r_sge); goto last_imm; case OP(RDMA_WRITE_LAST): rdma_last: /* Get the number of bytes the message was padded by. */ - pad = (be32_to_cpu(ohdr->bth[0]) >> 20) & 3; + pad = ib_bth_get_pad(ohdr); /* Check for invalid length. */ /* LAST len should be >= 1 */ if (unlikely(tlen < (hdrsize + pad + 4))) @@ -564,7 +564,7 @@ rdma_last: tlen -= (hdrsize + pad + 4); if (unlikely(tlen + qp->r_rcv_len != qp->r_len)) goto drop; - hfi1_copy_sge(&qp->r_sge, data, tlen, 1, 0); + hfi1_copy_sge(&qp->r_sge, data, tlen, true, false); rvt_put_ss(&qp->r_sge); break; @@ -584,5 +584,5 @@ drop: return; op_err: - hfi1_rc_error(qp, IB_WC_LOC_QP_OP_ERR); + rvt_rc_error(qp, IB_WC_LOC_QP_OP_ERR); } diff --git a/drivers/infiniband/hw/hfi1/ud.c b/drivers/infiniband/hw/hfi1/ud.c index c071955c0272..6a4e95cefae5 100644 --- a/drivers/infiniband/hw/hfi1/ud.c +++ b/drivers/infiniband/hw/hfi1/ud.c @@ -68,7 +68,7 @@ static void ud_loopback(struct rvt_qp *sqp, struct rvt_swqe *swqe) struct hfi1_ibport *ibp = to_iport(sqp->ibqp.device, sqp->port_num); struct hfi1_pportdata *ppd; struct rvt_qp *qp; - struct ib_ah_attr *ah_attr; + struct rdma_ah_attr *ah_attr; unsigned long flags; struct rvt_sge_state ssge; struct rvt_sge *sge; @@ -103,17 +103,17 @@ static void ud_loopback(struct rvt_qp *sqp, struct rvt_swqe *swqe) if (qp->ibqp.qp_num > 1) { u16 pkey; u16 slid; - u8 sc5 = ibp->sl_to_sc[ah_attr->sl]; + u8 sc5 = ibp->sl_to_sc[rdma_ah_get_sl(ah_attr)]; pkey = hfi1_get_pkey(ibp, sqp->s_pkey_index); - slid = ppd->lid | (ah_attr->src_path_bits & + slid = ppd->lid | (rdma_ah_get_path_bits(ah_attr) & ((1 << ppd->lmc) - 1)); if (unlikely(ingress_pkey_check(ppd, pkey, sc5, qp->s_pkey_index, slid))) { hfi1_bad_pqkey(ibp, OPA_TRAP_BAD_P_KEY, pkey, - ah_attr->sl, + rdma_ah_get_sl(ah_attr), sqp->ibqp.qp_num, qp->ibqp.qp_num, - slid, ah_attr->dlid); + slid, rdma_ah_get_dlid(ah_attr)); goto drop; } } @@ -131,13 +131,13 @@ static void ud_loopback(struct rvt_qp *sqp, struct rvt_swqe *swqe) if (unlikely(qkey != qp->qkey)) { u16 lid; - lid = ppd->lid | (ah_attr->src_path_bits & + lid = ppd->lid | (rdma_ah_get_path_bits(ah_attr) & ((1 << ppd->lmc) - 1)); hfi1_bad_pqkey(ibp, OPA_TRAP_BAD_Q_KEY, qkey, - ah_attr->sl, + rdma_ah_get_sl(ah_attr), sqp->ibqp.qp_num, qp->ibqp.qp_num, lid, - ah_attr->dlid); + rdma_ah_get_dlid(ah_attr)); goto drop; } } @@ -167,7 +167,7 @@ static void ud_loopback(struct rvt_qp *sqp, struct rvt_swqe *swqe) ret = hfi1_rvt_get_rwqe(qp, 0); if (ret < 0) { - hfi1_rc_error(qp, IB_WC_LOC_QP_OP_ERR); + rvt_rc_error(qp, IB_WC_LOC_QP_OP_ERR); goto bail_unlock; } if (!ret) { @@ -183,16 +183,16 @@ static void ud_loopback(struct rvt_qp *sqp, struct rvt_swqe *swqe) goto bail_unlock; } - if (ah_attr->ah_flags & IB_AH_GRH) { + if (rdma_ah_get_ah_flags(ah_attr) & IB_AH_GRH) { struct ib_grh grh; - struct ib_global_route grd = ah_attr->grh; + const struct ib_global_route *grd = rdma_ah_read_grh(ah_attr); - hfi1_make_grh(ibp, &grh, &grd, 0, 0); + hfi1_make_grh(ibp, &grh, grd, 0, 0); hfi1_copy_sge(&qp->r_sge, &grh, - sizeof(grh), 1, 0); + sizeof(grh), true, false); wc.wc_flags |= IB_WC_GRH; } else { - hfi1_skip_sge(&qp->r_sge, sizeof(struct ib_grh), 1); + rvt_skip_sge(&qp->r_sge, sizeof(struct ib_grh), true); } ssge.sg_list = swqe->sg_list + 1; ssge.sge = *swqe->sg_list; @@ -206,7 +206,7 @@ static void ud_loopback(struct rvt_qp *sqp, struct rvt_swqe *swqe) if (len > sge->sge_length) len = sge->sge_length; WARN_ON_ONCE(len == 0); - hfi1_copy_sge(&qp->r_sge, sge->vaddr, len, 1, 0); + hfi1_copy_sge(&qp->r_sge, sge->vaddr, len, true, false); sge->vaddr += len; sge->length -= len; sge->sge_length -= len; @@ -243,12 +243,13 @@ static void ud_loopback(struct rvt_qp *sqp, struct rvt_swqe *swqe) } else { wc.pkey_index = 0; } - wc.slid = ppd->lid | (ah_attr->src_path_bits & ((1 << ppd->lmc) - 1)); + wc.slid = ppd->lid | (rdma_ah_get_path_bits(ah_attr) & + ((1 << ppd->lmc) - 1)); /* Check for loopback when the port lid is not set */ if (wc.slid == 0 && sqp->ibqp.qp_type == IB_QPT_GSI) wc.slid = be16_to_cpu(IB_LID_PERMISSIVE); - wc.sl = ah_attr->sl; - wc.dlid_path_bits = ah_attr->dlid & ((1 << ppd->lmc) - 1); + wc.sl = rdma_ah_get_sl(ah_attr); + wc.dlid_path_bits = rdma_ah_get_dlid(ah_attr) & ((1 << ppd->lmc) - 1); wc.port_num = qp->port_num; /* Signal completion event if the solicited bit is set. */ rvt_cq_enter(ibcq_to_rvtcq(qp->ibqp.recv_cq), &wc, @@ -272,7 +273,7 @@ int hfi1_make_ud_req(struct rvt_qp *qp, struct hfi1_pkt_state *ps) { struct hfi1_qp_priv *priv = qp->priv; struct ib_other_headers *ohdr; - struct ib_ah_attr *ah_attr; + struct rdma_ah_attr *ah_attr; struct hfi1_pportdata *ppd; struct hfi1_ibport *ibp; struct rvt_swqe *wqe; @@ -319,9 +320,9 @@ int hfi1_make_ud_req(struct rvt_qp *qp, struct hfi1_pkt_state *ps) ibp = to_iport(qp->ibqp.device, qp->port_num); ppd = ppd_from_ibp(ibp); ah_attr = &ibah_to_rvtah(wqe->ud_wr.ah)->attr; - if (ah_attr->dlid < be16_to_cpu(IB_MULTICAST_LID_BASE) || - ah_attr->dlid == be16_to_cpu(IB_LID_PERMISSIVE)) { - lid = ah_attr->dlid & ~((1 << ppd->lmc) - 1); + if (rdma_ah_get_dlid(ah_attr) < be16_to_cpu(IB_MULTICAST_LID_BASE) || + rdma_ah_get_dlid(ah_attr) == be16_to_cpu(IB_LID_PERMISSIVE)) { + lid = rdma_ah_get_dlid(ah_attr) & ~((1 << ppd->lmc) - 1); if (unlikely(!loopback && (lid == ppd->lid || (lid == be16_to_cpu(IB_LID_PERMISSIVE) && @@ -356,7 +357,7 @@ int hfi1_make_ud_req(struct rvt_qp *qp, struct hfi1_pkt_state *ps) qp->s_hdrwords = 7; ps->s_txreq->s_cur_size = wqe->length; ps->s_txreq->ss = &qp->s_sge; - qp->s_srate = ah_attr->static_rate; + qp->s_srate = rdma_ah_get_static_rate(ah_attr); qp->srate_mbps = ib_rate_to_mbps(qp->s_srate); qp->s_wqe = wqe; qp->s_sge.sge = wqe->sg_list[0]; @@ -364,11 +365,11 @@ int hfi1_make_ud_req(struct rvt_qp *qp, struct hfi1_pkt_state *ps) qp->s_sge.num_sge = wqe->wr.num_sge; qp->s_sge.total_len = wqe->length; - if (ah_attr->ah_flags & IB_AH_GRH) { + if (rdma_ah_get_ah_flags(ah_attr) & IB_AH_GRH) { /* Header size in 32-bit words. */ qp->s_hdrwords += hfi1_make_grh(ibp, &ps->s_txreq->phdr.hdr.u.l.grh, - &ah_attr->grh, + rdma_ah_read_grh(ah_attr), qp->s_hdrwords, nwords); lrh0 = HFI1_LRH_GRH; ohdr = &ps->s_txreq->phdr.hdr.u.l.oth; @@ -388,8 +389,8 @@ int hfi1_make_ud_req(struct rvt_qp *qp, struct hfi1_pkt_state *ps) } else { bth0 = IB_OPCODE_UD_SEND_ONLY << 24; } - sc5 = ibp->sl_to_sc[ah_attr->sl]; - lrh0 |= (ah_attr->sl & 0xf) << 4; + sc5 = ibp->sl_to_sc[rdma_ah_get_sl(ah_attr)]; + lrh0 |= (rdma_ah_get_sl(ah_attr) & 0xf) << 4; if (qp->ibqp.qp_type == IB_QPT_SMI) { lrh0 |= 0xF000; /* Set VL (see ch. 13.5.3.1) */ priv->s_sc = 0xf; @@ -402,15 +403,17 @@ int hfi1_make_ud_req(struct rvt_qp *qp, struct hfi1_pkt_state *ps) priv->s_sendcontext = qp_to_send_context(qp, priv->s_sc); ps->s_txreq->psc = priv->s_sendcontext; ps->s_txreq->phdr.hdr.lrh[0] = cpu_to_be16(lrh0); - ps->s_txreq->phdr.hdr.lrh[1] = cpu_to_be16(ah_attr->dlid); + ps->s_txreq->phdr.hdr.lrh[1] = + cpu_to_be16(rdma_ah_get_dlid(ah_attr)); ps->s_txreq->phdr.hdr.lrh[2] = cpu_to_be16(qp->s_hdrwords + nwords + SIZE_OF_CRC); - if (ah_attr->dlid == be16_to_cpu(IB_LID_PERMISSIVE)) { + if (rdma_ah_get_dlid(ah_attr) == be16_to_cpu(IB_LID_PERMISSIVE)) { ps->s_txreq->phdr.hdr.lrh[3] = IB_LID_PERMISSIVE; } else { lid = ppd->lid; if (lid) { - lid |= ah_attr->src_path_bits & ((1 << ppd->lmc) - 1); + lid |= rdma_ah_get_path_bits(ah_attr) & + ((1 << ppd->lmc) - 1); ps->s_txreq->phdr.hdr.lrh[3] = cpu_to_be16(lid); } else { ps->s_txreq->phdr.hdr.lrh[3] = IB_LID_PERMISSIVE; @@ -537,7 +540,7 @@ void return_cnp(struct hfi1_ibport *ibp, struct rvt_qp *qp, u32 remote_qpn, bth0 = pkey | (IB_OPCODE_CNP << 24); ohdr->bth[0] = cpu_to_be32(bth0); - ohdr->bth[1] = cpu_to_be32(remote_qpn | (1 << HFI1_BECN_SHIFT)); + ohdr->bth[1] = cpu_to_be32(remote_qpn | (1 << IB_BECN_SHIFT)); ohdr->bth[2] = 0; /* PSN 0 */ hdr.lrh[0] = cpu_to_be16(lrh0); @@ -672,7 +675,7 @@ void hfi1_ud_rcv(struct hfi1_packet *packet) u32 src_qp; u16 dlid, pkey; int mgmt_pkey_idx = -1; - struct hfi1_ibport *ibp = &packet->rcd->ppd->ibport_data; + struct hfi1_ibport *ibp = rcd_to_iport(packet->rcd); struct hfi1_pportdata *ppd = ppd_from_ibp(ibp); struct ib_header *hdr = packet->hdr; u32 rcv_flags = packet->rcv_flags; @@ -680,7 +683,7 @@ void hfi1_ud_rcv(struct hfi1_packet *packet) u32 tlen = packet->tlen; struct rvt_qp *qp = packet->qp; bool has_grh = rcv_flags & HFI1_HAS_GRH; - u8 sc5 = hdr2sc(hdr, packet->rhf); + u8 sc5 = hfi1_9B_get_sc5(hdr, packet->rhf); u32 bth1; u8 sl_from_sc, sl; u16 slid; @@ -688,18 +691,16 @@ void hfi1_ud_rcv(struct hfi1_packet *packet) qkey = be32_to_cpu(ohdr->u.ud.deth[0]); src_qp = be32_to_cpu(ohdr->u.ud.deth[1]) & RVT_QPN_MASK; - dlid = be16_to_cpu(hdr->lrh[1]); + dlid = ib_get_dlid(hdr); bth1 = be32_to_cpu(ohdr->bth[1]); - slid = be16_to_cpu(hdr->lrh[3]); - pkey = (u16)be32_to_cpu(ohdr->bth[0]); - sl = (be16_to_cpu(hdr->lrh[0]) >> 4) & 0xf; - extra_bytes = (be32_to_cpu(ohdr->bth[0]) >> 20) & 3; + slid = ib_get_slid(hdr); + pkey = ib_bth_get_pkey(ohdr); + opcode = ib_bth_get_opcode(ohdr); + sl = ib_get_sl(hdr); + extra_bytes = ib_bth_get_pad(ohdr); extra_bytes += (SIZE_OF_CRC << 2); sl_from_sc = ibp->sc_to_sl[sc5]; - opcode = be32_to_cpu(ohdr->bth[0]) >> 24; - opcode &= 0xff; - process_ecn(qp, packet, (opcode != IB_OPCODE_CNP)); /* * Get the number of bytes the message was padded by @@ -796,7 +797,7 @@ void hfi1_ud_rcv(struct hfi1_packet *packet) ret = hfi1_rvt_get_rwqe(qp, 0); if (ret < 0) { - hfi1_rc_error(qp, IB_WC_LOC_QP_OP_ERR); + rvt_rc_error(qp, IB_WC_LOC_QP_OP_ERR); return; } if (!ret) { @@ -812,13 +813,13 @@ void hfi1_ud_rcv(struct hfi1_packet *packet) } if (has_grh) { hfi1_copy_sge(&qp->r_sge, &hdr->u.l.grh, - sizeof(struct ib_grh), 1, 0); + sizeof(struct ib_grh), true, false); wc.wc_flags |= IB_WC_GRH; } else { - hfi1_skip_sge(&qp->r_sge, sizeof(struct ib_grh), 1); + rvt_skip_sge(&qp->r_sge, sizeof(struct ib_grh), true); } hfi1_copy_sge(&qp->r_sge, data, wc.byte_len - sizeof(struct ib_grh), - 1, 0); + true, false); rvt_put_ss(&qp->r_sge); if (!test_and_clear_bit(RVT_R_WRID_VALID, &qp->r_aflags)) return; diff --git a/drivers/infiniband/hw/hfi1/user_exp_rcv.c b/drivers/infiniband/hw/hfi1/user_exp_rcv.c index 64d26525435a..a8f0aa4722f6 100644 --- a/drivers/infiniband/hw/hfi1/user_exp_rcv.c +++ b/drivers/infiniband/hw/hfi1/user_exp_rcv.c @@ -1,5 +1,5 @@ /* - * Copyright(c) 2015, 2016 Intel Corporation. + * Copyright(c) 2015-2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -45,6 +45,7 @@ * */ #include <asm/page.h> +#include <linux/string.h> #include "user_exp_rcv.h" #include "trace.h" @@ -52,7 +53,7 @@ struct tid_group { struct list_head list; - unsigned base; + u32 base; u8 size; u8 used; u8 map; @@ -81,20 +82,25 @@ struct tid_pageset { (unsigned long)(len) - 1) & PAGE_MASK) - \ ((unsigned long)vaddr & PAGE_MASK)) >> PAGE_SHIFT)) -static void unlock_exp_tids(struct hfi1_ctxtdata *, struct exp_tid_set *, - struct hfi1_filedata *); -static u32 find_phys_blocks(struct page **, unsigned, struct tid_pageset *); -static int set_rcvarray_entry(struct file *, unsigned long, u32, - struct tid_group *, struct page **, unsigned); -static int tid_rb_insert(void *, struct mmu_rb_node *); +static void unlock_exp_tids(struct hfi1_ctxtdata *uctxt, + struct exp_tid_set *set, + struct hfi1_filedata *fd); +static u32 find_phys_blocks(struct page **pages, unsigned npages, + struct tid_pageset *list); +static int set_rcvarray_entry(struct hfi1_filedata *fd, unsigned long vaddr, + u32 rcventry, struct tid_group *grp, + struct page **pages, unsigned npages); +static int tid_rb_insert(void *arg, struct mmu_rb_node *node); static void cacheless_tid_rb_remove(struct hfi1_filedata *fdata, struct tid_rb_node *tnode); -static void tid_rb_remove(void *, struct mmu_rb_node *); -static int tid_rb_invalidate(void *, struct mmu_rb_node *); -static int program_rcvarray(struct file *, unsigned long, struct tid_group *, - struct tid_pageset *, unsigned, u16, struct page **, - u32 *, unsigned *, unsigned *); -static int unprogram_rcvarray(struct file *, u32, struct tid_group **); +static void tid_rb_remove(void *arg, struct mmu_rb_node *node); +static int tid_rb_invalidate(void *arg, struct mmu_rb_node *mnode); +static int program_rcvarray(struct hfi1_filedata *fd, unsigned long vaddr, + struct tid_group *grp, struct tid_pageset *sets, + unsigned start, u16 count, struct page **pages, + u32 *tidlist, unsigned *tididx, unsigned *pmapped); +static int unprogram_rcvarray(struct hfi1_filedata *fd, u32 tidinfo, + struct tid_group **grp); static void clear_tid_node(struct hfi1_filedata *fd, struct tid_rb_node *node); static struct mmu_rb_ops tid_rb_ops = { @@ -148,62 +154,72 @@ static inline void tid_group_move(struct tid_group *group, tid_group_add_tail(group, s2); } +int hfi1_user_exp_rcv_grp_init(struct hfi1_filedata *fd) +{ + struct hfi1_ctxtdata *uctxt = fd->uctxt; + struct hfi1_devdata *dd = fd->dd; + u32 tidbase; + u32 i; + struct tid_group *grp, *gptr; + + exp_tid_group_init(&uctxt->tid_group_list); + exp_tid_group_init(&uctxt->tid_used_list); + exp_tid_group_init(&uctxt->tid_full_list); + + tidbase = uctxt->expected_base; + for (i = 0; i < uctxt->expected_count / + dd->rcv_entries.group_size; i++) { + grp = kzalloc(sizeof(*grp), GFP_KERNEL); + if (!grp) + goto grp_failed; + + grp->size = dd->rcv_entries.group_size; + grp->base = tidbase; + tid_group_add_tail(grp, &uctxt->tid_group_list); + tidbase += dd->rcv_entries.group_size; + } + + return 0; + +grp_failed: + list_for_each_entry_safe(grp, gptr, &uctxt->tid_group_list.list, + list) { + list_del_init(&grp->list); + kfree(grp); + } + + return -ENOMEM; +} + /* * Initialize context and file private data needed for Expected * receive caching. This needs to be done after the context has * been configured with the eager/expected RcvEntry counts. */ -int hfi1_user_exp_rcv_init(struct file *fp) +int hfi1_user_exp_rcv_init(struct hfi1_filedata *fd) { - struct hfi1_filedata *fd = fp->private_data; struct hfi1_ctxtdata *uctxt = fd->uctxt; struct hfi1_devdata *dd = uctxt->dd; - unsigned tidbase; - int i, ret = 0; + int ret = 0; spin_lock_init(&fd->tid_lock); spin_lock_init(&fd->invalid_lock); - if (!uctxt->subctxt_cnt || !fd->subctxt) { - exp_tid_group_init(&uctxt->tid_group_list); - exp_tid_group_init(&uctxt->tid_used_list); - exp_tid_group_init(&uctxt->tid_full_list); - - tidbase = uctxt->expected_base; - for (i = 0; i < uctxt->expected_count / - dd->rcv_entries.group_size; i++) { - struct tid_group *grp; - - grp = kzalloc(sizeof(*grp), GFP_KERNEL); - if (!grp) { - /* - * If we fail here, the groups already - * allocated will be freed by the close - * call. - */ - ret = -ENOMEM; - goto done; - } - grp->size = dd->rcv_entries.group_size; - grp->base = tidbase; - tid_group_add_tail(grp, &uctxt->tid_group_list); - tidbase += dd->rcv_entries.group_size; - } - } - fd->entry_to_rb = kcalloc(uctxt->expected_count, - sizeof(struct rb_node *), - GFP_KERNEL); + sizeof(struct rb_node *), + GFP_KERNEL); if (!fd->entry_to_rb) return -ENOMEM; if (!HFI1_CAP_UGET_MASK(uctxt->flags, TID_UNMAP)) { fd->invalid_tid_idx = 0; - fd->invalid_tids = kzalloc(uctxt->expected_count * - sizeof(u32), GFP_KERNEL); + fd->invalid_tids = kcalloc(uctxt->expected_count, + sizeof(*fd->invalid_tids), + GFP_KERNEL); if (!fd->invalid_tids) { - ret = -ENOMEM; - goto done; + kfree(fd->entry_to_rb); + fd->entry_to_rb = NULL; + return -ENOMEM; } /* @@ -246,41 +262,44 @@ int hfi1_user_exp_rcv_init(struct file *fp) fd->tid_limit = uctxt->expected_count; } spin_unlock(&fd->tid_lock); -done: + return ret; } -int hfi1_user_exp_rcv_free(struct hfi1_filedata *fd) +void hfi1_user_exp_rcv_grp_free(struct hfi1_ctxtdata *uctxt) { - struct hfi1_ctxtdata *uctxt = fd->uctxt; struct tid_group *grp, *gptr; - if (!test_bit(HFI1_CTXT_SETUP_DONE, &uctxt->event_flags)) - return 0; + list_for_each_entry_safe(grp, gptr, &uctxt->tid_group_list.list, + list) { + list_del_init(&grp->list); + kfree(grp); + } + hfi1_clear_tids(uctxt); +} + +void hfi1_user_exp_rcv_free(struct hfi1_filedata *fd) +{ + struct hfi1_ctxtdata *uctxt = fd->uctxt; + /* * The notifier would have been removed when the process'es mm * was freed. */ - if (fd->handler) + if (fd->handler) { hfi1_mmu_rb_unregister(fd->handler); - - kfree(fd->invalid_tids); - - if (!uctxt->cnt) { + } else { if (!EXP_TID_SET_EMPTY(uctxt->tid_full_list)) unlock_exp_tids(uctxt, &uctxt->tid_full_list, fd); if (!EXP_TID_SET_EMPTY(uctxt->tid_used_list)) unlock_exp_tids(uctxt, &uctxt->tid_used_list, fd); - list_for_each_entry_safe(grp, gptr, &uctxt->tid_group_list.list, - list) { - list_del_init(&grp->list); - kfree(grp); - } - hfi1_clear_tids(uctxt); } + kfree(fd->invalid_tids); + fd->invalid_tids = NULL; + kfree(fd->entry_to_rb); - return 0; + fd->entry_to_rb = NULL; } /* @@ -349,10 +368,10 @@ static inline void rcv_array_wc_fill(struct hfi1_devdata *dd, u32 index) * can fit into the group. If the group becomes fully * used, move it to tid_full_list. */ -int hfi1_user_exp_rcv_setup(struct file *fp, struct hfi1_tid_info *tinfo) +int hfi1_user_exp_rcv_setup(struct hfi1_filedata *fd, + struct hfi1_tid_info *tinfo) { int ret = 0, need_group = 0, pinned; - struct hfi1_filedata *fd = fp->private_data; struct hfi1_ctxtdata *uctxt = fd->uctxt; struct hfi1_devdata *dd = uctxt->dd; unsigned npages, ngroups, pageidx = 0, pageset_count, npagesets, @@ -449,7 +468,7 @@ int hfi1_user_exp_rcv_setup(struct file *fp, struct hfi1_tid_info *tinfo) struct tid_group *grp = tid_group_pop(&uctxt->tid_group_list); - ret = program_rcvarray(fp, vaddr, grp, pagesets, + ret = program_rcvarray(fd, vaddr, grp, pagesets, pageidx, dd->rcv_entries.group_size, pages, tidlist, &tididx, &mapped); /* @@ -495,7 +514,7 @@ int hfi1_user_exp_rcv_setup(struct file *fp, struct hfi1_tid_info *tinfo) unsigned use = min_t(unsigned, pageset_count - pageidx, grp->size - grp->used); - ret = program_rcvarray(fp, vaddr, grp, pagesets, + ret = program_rcvarray(fd, vaddr, grp, pagesets, pageidx, use, pages, tidlist, &tididx, &mapped); if (ret < 0) { @@ -545,7 +564,7 @@ nomem: * everything done so far so we don't leak resources. */ tinfo->tidlist = (unsigned long)&tidlist; - hfi1_user_exp_rcv_clear(fp, tinfo); + hfi1_user_exp_rcv_clear(fd, tinfo); tinfo->tidlist = 0; ret = -EFAULT; goto bail; @@ -569,28 +588,25 @@ bail: return ret > 0 ? 0 : ret; } -int hfi1_user_exp_rcv_clear(struct file *fp, struct hfi1_tid_info *tinfo) +int hfi1_user_exp_rcv_clear(struct hfi1_filedata *fd, + struct hfi1_tid_info *tinfo) { int ret = 0; - struct hfi1_filedata *fd = fp->private_data; struct hfi1_ctxtdata *uctxt = fd->uctxt; u32 *tidinfo; unsigned tididx; - tidinfo = kcalloc(tinfo->tidcnt, sizeof(*tidinfo), GFP_KERNEL); - if (!tidinfo) - return -ENOMEM; + if (unlikely(tinfo->tidcnt > fd->tid_used)) + return -EINVAL; - if (copy_from_user(tidinfo, (void __user *)(unsigned long) - tinfo->tidlist, sizeof(tidinfo[0]) * - tinfo->tidcnt)) { - ret = -EFAULT; - goto done; - } + tidinfo = memdup_user((void __user *)(unsigned long)tinfo->tidlist, + sizeof(tidinfo[0]) * tinfo->tidcnt); + if (IS_ERR(tidinfo)) + return PTR_ERR(tidinfo); mutex_lock(&uctxt->exp_lock); for (tididx = 0; tididx < tinfo->tidcnt; tididx++) { - ret = unprogram_rcvarray(fp, tidinfo[tididx], NULL); + ret = unprogram_rcvarray(fd, tidinfo[tididx], NULL); if (ret) { hfi1_cdbg(TID, "Failed to unprogram rcv array %d", ret); @@ -602,17 +618,17 @@ int hfi1_user_exp_rcv_clear(struct file *fp, struct hfi1_tid_info *tinfo) spin_unlock(&fd->tid_lock); tinfo->tidcnt = tididx; mutex_unlock(&uctxt->exp_lock); -done: + kfree(tidinfo); return ret; } -int hfi1_user_exp_rcv_invalid(struct file *fp, struct hfi1_tid_info *tinfo) +int hfi1_user_exp_rcv_invalid(struct hfi1_filedata *fd, + struct hfi1_tid_info *tinfo) { - struct hfi1_filedata *fd = fp->private_data; struct hfi1_ctxtdata *uctxt = fd->uctxt; unsigned long *ev = uctxt->dd->events + - (((uctxt->ctxt - uctxt->dd->first_user_ctxt) * + (((uctxt->ctxt - uctxt->dd->first_dyn_alloc_ctxt) * HFI1_MAX_SHARED_CTXTS) + fd->subctxt); u32 *array; int ret = 0; @@ -724,7 +740,7 @@ static u32 find_phys_blocks(struct page **pages, unsigned npages, /** * program_rcvarray() - program an RcvArray group with receive buffers - * @fp: file pointer + * @fd: filedata pointer * @vaddr: starting user virtual address * @grp: RcvArray group * @sets: array of struct tid_pageset holding information on physically @@ -749,13 +765,12 @@ static u32 find_phys_blocks(struct page **pages, unsigned npages, * -ENOMEM or -EFAULT on error from set_rcvarray_entry(), or * number of RcvArray entries programmed. */ -static int program_rcvarray(struct file *fp, unsigned long vaddr, +static int program_rcvarray(struct hfi1_filedata *fd, unsigned long vaddr, struct tid_group *grp, struct tid_pageset *sets, unsigned start, u16 count, struct page **pages, u32 *tidlist, unsigned *tididx, unsigned *pmapped) { - struct hfi1_filedata *fd = fp->private_data; struct hfi1_ctxtdata *uctxt = fd->uctxt; struct hfi1_devdata *dd = uctxt->dd; u16 idx; @@ -796,7 +811,7 @@ static int program_rcvarray(struct file *fp, unsigned long vaddr, npages = sets[setidx].count; pageidx = sets[setidx].idx; - ret = set_rcvarray_entry(fp, vaddr + (pageidx * PAGE_SIZE), + ret = set_rcvarray_entry(fd, vaddr + (pageidx * PAGE_SIZE), rcventry, grp, pages + pageidx, npages); if (ret) @@ -818,12 +833,11 @@ static int program_rcvarray(struct file *fp, unsigned long vaddr, return idx; } -static int set_rcvarray_entry(struct file *fp, unsigned long vaddr, +static int set_rcvarray_entry(struct hfi1_filedata *fd, unsigned long vaddr, u32 rcventry, struct tid_group *grp, struct page **pages, unsigned npages) { int ret; - struct hfi1_filedata *fd = fp->private_data; struct hfi1_ctxtdata *uctxt = fd->uctxt; struct tid_rb_node *node; struct hfi1_devdata *dd = uctxt->dd; @@ -877,10 +891,9 @@ static int set_rcvarray_entry(struct file *fp, unsigned long vaddr, return 0; } -static int unprogram_rcvarray(struct file *fp, u32 tidinfo, +static int unprogram_rcvarray(struct hfi1_filedata *fd, u32 tidinfo, struct tid_group **grp) { - struct hfi1_filedata *fd = fp->private_data; struct hfi1_ctxtdata *uctxt = fd->uctxt; struct hfi1_devdata *dd = uctxt->dd; struct tid_rb_node *node; @@ -1016,8 +1029,8 @@ static int tid_rb_invalidate(void *arg, struct mmu_rb_node *mnode) * process in question. */ ev = uctxt->dd->events + - (((uctxt->ctxt - uctxt->dd->first_user_ctxt) * - HFI1_MAX_SHARED_CTXTS) + fdata->subctxt); + (((uctxt->ctxt - uctxt->dd->first_dyn_alloc_ctxt) * + HFI1_MAX_SHARED_CTXTS) + fdata->subctxt); set_bit(_HFI1_EVENT_TID_MMU_NOTIFY_BIT, ev); } fdata->invalid_tid_idx++; diff --git a/drivers/infiniband/hw/hfi1/user_exp_rcv.h b/drivers/infiniband/hw/hfi1/user_exp_rcv.h index 9bc8d9fba87e..5250c897298d 100644 --- a/drivers/infiniband/hw/hfi1/user_exp_rcv.h +++ b/drivers/infiniband/hw/hfi1/user_exp_rcv.h @@ -1,7 +1,7 @@ #ifndef _HFI1_USER_EXP_RCV_H #define _HFI1_USER_EXP_RCV_H /* - * Copyright(c) 2015, 2016 Intel Corporation. + * Copyright(c) 2015 - 2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -70,10 +70,15 @@ (tid) |= EXP_TID_SET(field, (value)); \ } while (0) -int hfi1_user_exp_rcv_init(struct file *); -int hfi1_user_exp_rcv_free(struct hfi1_filedata *); -int hfi1_user_exp_rcv_setup(struct file *, struct hfi1_tid_info *); -int hfi1_user_exp_rcv_clear(struct file *, struct hfi1_tid_info *); -int hfi1_user_exp_rcv_invalid(struct file *, struct hfi1_tid_info *); +void hfi1_user_exp_rcv_grp_free(struct hfi1_ctxtdata *uctxt); +int hfi1_user_exp_rcv_grp_init(struct hfi1_filedata *fd); +int hfi1_user_exp_rcv_init(struct hfi1_filedata *fd); +void hfi1_user_exp_rcv_free(struct hfi1_filedata *fd); +int hfi1_user_exp_rcv_setup(struct hfi1_filedata *fd, + struct hfi1_tid_info *tinfo); +int hfi1_user_exp_rcv_clear(struct hfi1_filedata *fd, + struct hfi1_tid_info *tinfo); +int hfi1_user_exp_rcv_invalid(struct hfi1_filedata *fd, + struct hfi1_tid_info *tinfo); #endif /* _HFI1_USER_EXP_RCV_H */ diff --git a/drivers/infiniband/hw/hfi1/user_pages.c b/drivers/infiniband/hw/hfi1/user_pages.c index 20f4ddcac3b0..e341e6dcc388 100644 --- a/drivers/infiniband/hw/hfi1/user_pages.c +++ b/drivers/infiniband/hw/hfi1/user_pages.c @@ -1,5 +1,5 @@ /* - * Copyright(c) 2015, 2016 Intel Corporation. + * Copyright(c) 2015-2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -46,7 +46,7 @@ */ #include <linux/mm.h> -#include <linux/sched.h> +#include <linux/sched/signal.h> #include <linux/device.h> #include <linux/module.h> @@ -73,7 +73,8 @@ bool hfi1_can_pin_pages(struct hfi1_devdata *dd, struct mm_struct *mm, { unsigned long ulimit = rlimit(RLIMIT_MEMLOCK), pinned, cache_limit, size = (cache_size * (1UL << 20)); /* convert to bytes */ - unsigned usr_ctxts = dd->num_rcv_contexts - dd->first_user_ctxt; + unsigned int usr_ctxts = + dd->num_rcv_contexts - dd->first_dyn_alloc_ctxt; bool can_lock = capable(CAP_IPC_LOCK); /* diff --git a/drivers/infiniband/hw/hfi1/user_sdma.c b/drivers/infiniband/hw/hfi1/user_sdma.c index 7d22f8ee98ef..d55339f5d73b 100644 --- a/drivers/infiniband/hw/hfi1/user_sdma.c +++ b/drivers/infiniband/hw/hfi1/user_sdma.c @@ -1,5 +1,5 @@ /* - * Copyright(c) 2015, 2016 Intel Corporation. + * Copyright(c) 2015 - 2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -60,6 +60,7 @@ #include <linux/mmu_context.h> #include <linux/module.h> #include <linux/vmalloc.h> +#include <linux/string.h> #include "hfi.h" #include "sdma.h" @@ -142,7 +143,9 @@ MODULE_PARM_DESC(sdma_comp_size, "Size of User SDMA completion ring. Default: 12 /* KDETH OM multipliers and switch over point */ #define KDETH_OM_SMALL 4 +#define KDETH_OM_SMALL_SHIFT 2 #define KDETH_OM_LARGE 64 +#define KDETH_OM_LARGE_SHIFT 6 #define KDETH_OM_MAX_SIZE (1 << ((KDETH_OM_LARGE / KDETH_OM_SMALL) + 1)) /* Tx request flag bits */ @@ -152,9 +155,8 @@ MODULE_PARM_DESC(sdma_comp_size, "Size of User SDMA completion ring. Default: 12 /* SDMA request flag bits */ #define SDMA_REQ_FOR_THREAD 1 #define SDMA_REQ_SEND_DONE 2 -#define SDMA_REQ_HAVE_AHG 3 -#define SDMA_REQ_HAS_ERROR 4 -#define SDMA_REQ_DONE_ERROR 5 +#define SDMA_REQ_HAS_ERROR 3 +#define SDMA_REQ_DONE_ERROR 4 #define SDMA_PKT_Q_INACTIVE BIT(0) #define SDMA_PKT_Q_ACTIVE BIT(1) @@ -213,7 +215,7 @@ struct user_sdma_request { * each request will need it's own engine pointer. */ struct sdma_engine *sde; - u8 ahg_idx; + s8 ahg_idx; u32 ahg[9]; /* * KDETH.Offset (Eager) field @@ -228,12 +230,6 @@ struct user_sdma_request { */ u32 tidoffset; /* - * KDETH.OM - * Remember this because the header template always sets it - * to 0. - */ - u8 omfactor; - /* * We copy the iovs for this request (based on * info.iovcnt). These are only the data vectors */ @@ -283,39 +279,43 @@ struct user_sdma_txreq { hfi1_cdbg(SDMA, "[%u:%u:%u] " fmt, (pq)->dd->unit, (pq)->ctxt, \ (pq)->subctxt, ##__VA_ARGS__) -static int user_sdma_send_pkts(struct user_sdma_request *, unsigned); -static int num_user_pages(const struct iovec *); -static void user_sdma_txreq_cb(struct sdma_txreq *, int); -static inline void pq_update(struct hfi1_user_sdma_pkt_q *); -static void user_sdma_free_request(struct user_sdma_request *, bool); -static int pin_vector_pages(struct user_sdma_request *, - struct user_sdma_iovec *); -static void unpin_vector_pages(struct mm_struct *, struct page **, unsigned, - unsigned); -static int check_header_template(struct user_sdma_request *, - struct hfi1_pkt_header *, u32, u32); -static int set_txreq_header(struct user_sdma_request *, - struct user_sdma_txreq *, u32); -static int set_txreq_header_ahg(struct user_sdma_request *, - struct user_sdma_txreq *, u32); -static inline void set_comp_state(struct hfi1_user_sdma_pkt_q *, - struct hfi1_user_sdma_comp_q *, - u16, enum hfi1_sdma_comp_state, int); -static inline u32 set_pkt_bth_psn(__be32, u8, u32); +static int user_sdma_send_pkts(struct user_sdma_request *req, + unsigned maxpkts); +static int num_user_pages(const struct iovec *iov); +static void user_sdma_txreq_cb(struct sdma_txreq *txreq, int status); +static inline void pq_update(struct hfi1_user_sdma_pkt_q *pq); +static void user_sdma_free_request(struct user_sdma_request *req, bool unpin); +static int pin_vector_pages(struct user_sdma_request *req, + struct user_sdma_iovec *iovec); +static void unpin_vector_pages(struct mm_struct *mm, struct page **pages, + unsigned start, unsigned npages); +static int check_header_template(struct user_sdma_request *req, + struct hfi1_pkt_header *hdr, u32 lrhlen, + u32 datalen); +static int set_txreq_header(struct user_sdma_request *req, + struct user_sdma_txreq *tx, u32 datalen); +static int set_txreq_header_ahg(struct user_sdma_request *req, + struct user_sdma_txreq *tx, u32 len); +static inline void set_comp_state(struct hfi1_user_sdma_pkt_q *pq, + struct hfi1_user_sdma_comp_q *cq, + u16 idx, enum hfi1_sdma_comp_state state, + int ret); +static inline u32 set_pkt_bth_psn(__be32 bthpsn, u8 expct, u32 frags); static inline u32 get_lrh_len(struct hfi1_pkt_header, u32 len); static int defer_packet_queue( - struct sdma_engine *, - struct iowait *, - struct sdma_txreq *, - unsigned seq); -static void activate_packet_queue(struct iowait *, int); -static bool sdma_rb_filter(struct mmu_rb_node *, unsigned long, unsigned long); -static int sdma_rb_insert(void *, struct mmu_rb_node *); + struct sdma_engine *sde, + struct iowait *wait, + struct sdma_txreq *txreq, + unsigned int seq); +static void activate_packet_queue(struct iowait *wait, int reason); +static bool sdma_rb_filter(struct mmu_rb_node *node, unsigned long addr, + unsigned long len); +static int sdma_rb_insert(void *arg, struct mmu_rb_node *mnode); static int sdma_rb_evict(void *arg, struct mmu_rb_node *mnode, void *arg2, bool *stop); -static void sdma_rb_remove(void *, struct mmu_rb_node *); -static int sdma_rb_invalidate(void *, struct mmu_rb_node *); +static void sdma_rb_remove(void *arg, struct mmu_rb_node *mnode); +static int sdma_rb_invalidate(void *arg, struct mmu_rb_node *mnode); static struct mmu_rb_ops sdma_rb_ops = { .filter = sdma_rb_filter, @@ -371,44 +371,27 @@ static void sdma_kmem_cache_ctor(void *obj) memset(tx, 0, sizeof(*tx)); } -int hfi1_user_sdma_alloc_queues(struct hfi1_ctxtdata *uctxt, struct file *fp) +int hfi1_user_sdma_alloc_queues(struct hfi1_ctxtdata *uctxt, + struct hfi1_filedata *fd) { - struct hfi1_filedata *fd; - int ret = 0; - unsigned memsize; + int ret = -ENOMEM; char buf[64]; struct hfi1_devdata *dd; struct hfi1_user_sdma_comp_q *cq; struct hfi1_user_sdma_pkt_q *pq; unsigned long flags; - if (!uctxt || !fp) { - ret = -EBADF; - goto done; - } + if (!uctxt || !fd) + return -EBADF; - fd = fp->private_data; - - if (!hfi1_sdma_comp_ring_size) { - ret = -EINVAL; - goto done; - } + if (!hfi1_sdma_comp_ring_size) + return -EINVAL; dd = uctxt->dd; pq = kzalloc(sizeof(*pq), GFP_KERNEL); if (!pq) - goto pq_nomem; - - memsize = sizeof(*pq->reqs) * hfi1_sdma_comp_ring_size; - pq->reqs = kzalloc(memsize, GFP_KERNEL); - if (!pq->reqs) - goto pq_reqs_nomem; - - memsize = BITS_TO_LONGS(hfi1_sdma_comp_ring_size) * sizeof(long); - pq->req_in_use = kzalloc(memsize, GFP_KERNEL); - if (!pq->req_in_use) - goto pq_reqs_no_in_use; + return -ENOMEM; INIT_LIST_HEAD(&pq->list); pq->dd = dd; @@ -424,10 +407,23 @@ int hfi1_user_sdma_alloc_queues(struct hfi1_ctxtdata *uctxt, struct file *fp) iowait_init(&pq->busy, 0, NULL, defer_packet_queue, activate_packet_queue, NULL); pq->reqidx = 0; + + pq->reqs = kcalloc(hfi1_sdma_comp_ring_size, + sizeof(*pq->reqs), + GFP_KERNEL); + if (!pq->reqs) + goto pq_reqs_nomem; + + pq->req_in_use = kcalloc(BITS_TO_LONGS(hfi1_sdma_comp_ring_size), + sizeof(*pq->req_in_use), + GFP_KERNEL); + if (!pq->req_in_use) + goto pq_reqs_no_in_use; + snprintf(buf, 64, "txreq-kmem-cache-%u-%u-%u", dd->unit, uctxt->ctxt, fd->subctxt); pq->txreq_cache = kmem_cache_create(buf, - sizeof(struct user_sdma_txreq), + sizeof(struct user_sdma_txreq), L1_CACHE_BYTES, SLAB_HWCACHE_ALIGN, sdma_kmem_cache_ctor); @@ -436,31 +432,36 @@ int hfi1_user_sdma_alloc_queues(struct hfi1_ctxtdata *uctxt, struct file *fp) uctxt->ctxt); goto pq_txreq_nomem; } - fd->pq = pq; + cq = kzalloc(sizeof(*cq), GFP_KERNEL); if (!cq) goto cq_nomem; - memsize = PAGE_ALIGN(sizeof(*cq->comps) * hfi1_sdma_comp_ring_size); - cq->comps = vmalloc_user(memsize); + cq->comps = vmalloc_user(PAGE_ALIGN(sizeof(*cq->comps) + * hfi1_sdma_comp_ring_size)); if (!cq->comps) goto cq_comps_nomem; cq->nentries = hfi1_sdma_comp_ring_size; - fd->cq = cq; ret = hfi1_mmu_rb_register(pq, pq->mm, &sdma_rb_ops, dd->pport->hfi1_wq, &pq->handler); if (ret) { dd_dev_err(dd, "Failed to register with MMU %d", ret); - goto done; + goto pq_mmu_fail; } + fd->pq = pq; + fd->cq = cq; + spin_lock_irqsave(&uctxt->sdma_qlock, flags); list_add(&pq->list, &uctxt->sdma_queues); spin_unlock_irqrestore(&uctxt->sdma_qlock, flags); - goto done; + return 0; + +pq_mmu_fail: + vfree(cq->comps); cq_comps_nomem: kfree(cq); cq_nomem: @@ -471,10 +472,7 @@ pq_reqs_no_in_use: kfree(pq->reqs); pq_reqs_nomem: kfree(pq); - fd->pq = NULL; -pq_nomem: - ret = -ENOMEM; -done: + return ret; } @@ -534,11 +532,11 @@ static u8 dlid_to_selector(u16 dlid) return mapping[hash]; } -int hfi1_user_sdma_process_request(struct file *fp, struct iovec *iovec, - unsigned long dim, unsigned long *count) +int hfi1_user_sdma_process_request(struct hfi1_filedata *fd, + struct iovec *iovec, unsigned long dim, + unsigned long *count) { int ret = 0, i; - struct hfi1_filedata *fd = fp->private_data; struct hfi1_ctxtdata *uctxt = fd->uctxt; struct hfi1_user_sdma_pkt_q *pq = fd->pq; struct hfi1_user_sdma_comp_q *cq = fd->cq; @@ -614,6 +612,7 @@ int hfi1_user_sdma_process_request(struct file *fp, struct iovec *iovec, req->pq = pq; req->cq = cq; req->status = -1; + req->ahg_idx = -1; INIT_LIST_HEAD(&req->txps); memcpy(&req->info, &info, sizeof(info)); @@ -703,7 +702,9 @@ int hfi1_user_sdma_process_request(struct file *fp, struct iovec *iovec, /* Save all the IO vector structures */ for (i = 0; i < req->data_iovs; i++) { INIT_LIST_HEAD(&req->iovs[i].list); - memcpy(&req->iovs[i].iov, iovec + idx++, sizeof(struct iovec)); + memcpy(&req->iovs[i].iov, + iovec + idx++, + sizeof(req->iovs[i].iov)); ret = pin_vector_pages(req, &req->iovs[i]); if (ret) { req->status = ret; @@ -725,30 +726,28 @@ int hfi1_user_sdma_process_request(struct file *fp, struct iovec *iovec, */ if (req_opcode(req->info.ctrl) == EXPECTED) { u16 ntids = iovec[idx].iov_len / sizeof(*req->tids); + u32 *tmp; if (!ntids || ntids > MAX_TID_PAIR_ENTRIES) { ret = -EINVAL; goto free_req; } - req->tids = kcalloc(ntids, sizeof(*req->tids), GFP_KERNEL); - if (!req->tids) { - ret = -ENOMEM; - goto free_req; - } + /* * We have to copy all of the tids because they may vary * in size and, therefore, the TID count might not be * equal to the pkt count. However, there is no way to * tell at this point. */ - ret = copy_from_user(req->tids, iovec[idx].iov_base, - ntids * sizeof(*req->tids)); - if (ret) { + tmp = memdup_user(iovec[idx].iov_base, + ntids * sizeof(*req->tids)); + if (IS_ERR(tmp)) { + ret = PTR_ERR(tmp); SDMA_DBG(req, "Failed to copy %d TIDs (%d)", ntids, ret); - ret = -EFAULT; goto free_req; } + req->tids = tmp; req->n_tids = ntids; idx++; } @@ -764,14 +763,8 @@ int hfi1_user_sdma_process_request(struct file *fp, struct iovec *iovec, } /* We don't need an AHG entry if the request contains only one packet */ - if (req->info.npkts > 1 && HFI1_CAP_IS_USET(SDMA_AHG)) { - int ahg = sdma_ahg_alloc(req->sde); - - if (likely(ahg >= 0)) { - req->ahg_idx = (u8)ahg; - set_bit(SDMA_REQ_HAVE_AHG, &req->flags); - } - } + if (req->info.npkts > 1 && HFI1_CAP_IS_USET(SDMA_AHG)) + req->ahg_idx = sdma_ahg_alloc(req->sde); set_comp_state(pq, cq, info.comp_idx, QUEUED, 0); atomic_inc(&pq->n_reqs); @@ -989,7 +982,7 @@ static int user_sdma_send_pkts(struct user_sdma_request *req, unsigned maxpkts) } } - if (test_bit(SDMA_REQ_HAVE_AHG, &req->flags)) { + if (req->ahg_idx >= 0) { if (!req->seqnum) { u16 pbclen = le16_to_cpu(req->hdr.pbc[0]); u32 lrhlen = get_lrh_len(req->hdr, @@ -1119,7 +1112,7 @@ dosend: * happen due to the sequential manner in which * descriptors are processed. */ - if (test_bit(SDMA_REQ_HAVE_AHG, &req->flags)) + if (req->ahg_idx >= 0) sdma_ahg_free(req->sde, req->ahg_idx); } return ret; @@ -1321,6 +1314,7 @@ static int set_txreq_header(struct user_sdma_request *req, { struct hfi1_user_sdma_pkt_q *pq = req->pq; struct hfi1_pkt_header *hdr = &tx->hdr; + u8 omfactor; /* KDETH.OM */ u16 pbclen; int ret; u32 tidval = 0, lrhlen = get_lrh_len(*hdr, pad_len(datalen)); @@ -1398,8 +1392,9 @@ static int set_txreq_header(struct user_sdma_request *req, } tidval = req->tids[req->tididx]; } - req->omfactor = EXP_TID_GET(tidval, LEN) * PAGE_SIZE >= - KDETH_OM_MAX_SIZE ? KDETH_OM_LARGE : KDETH_OM_SMALL; + omfactor = EXP_TID_GET(tidval, LEN) * PAGE_SIZE >= + KDETH_OM_MAX_SIZE ? KDETH_OM_LARGE_SHIFT : + KDETH_OM_SMALL_SHIFT; /* Set KDETH.TIDCtrl based on value for this TID. */ KDETH_SET(hdr->kdeth.ver_tid_offset, TIDCTRL, EXP_TID_GET(tidval, CTRL)); @@ -1414,12 +1409,12 @@ static int set_txreq_header(struct user_sdma_request *req, * transfer. */ SDMA_DBG(req, "TID offset %ubytes %uunits om%u", - req->tidoffset, req->tidoffset / req->omfactor, - req->omfactor != KDETH_OM_SMALL); + req->tidoffset, req->tidoffset >> omfactor, + omfactor != KDETH_OM_SMALL_SHIFT); KDETH_SET(hdr->kdeth.ver_tid_offset, OFFSET, - req->tidoffset / req->omfactor); + req->tidoffset >> omfactor); KDETH_SET(hdr->kdeth.ver_tid_offset, OM, - req->omfactor != KDETH_OM_SMALL); + omfactor != KDETH_OM_SMALL_SHIFT); } done: trace_hfi1_sdma_user_header(pq->dd, pq->ctxt, pq->subctxt, @@ -1431,6 +1426,7 @@ static int set_txreq_header_ahg(struct user_sdma_request *req, struct user_sdma_txreq *tx, u32 len) { int diff = 0; + u8 omfactor; /* KDETH.OM */ struct hfi1_user_sdma_pkt_q *pq = req->pq; struct hfi1_pkt_header *hdr = &req->hdr; u16 pbclen = le16_to_cpu(hdr->pbc[0]); @@ -1482,14 +1478,15 @@ static int set_txreq_header_ahg(struct user_sdma_request *req, } tidval = req->tids[req->tididx]; } - req->omfactor = ((EXP_TID_GET(tidval, LEN) * + omfactor = ((EXP_TID_GET(tidval, LEN) * PAGE_SIZE) >= - KDETH_OM_MAX_SIZE) ? KDETH_OM_LARGE : - KDETH_OM_SMALL; + KDETH_OM_MAX_SIZE) ? KDETH_OM_LARGE_SHIFT : + KDETH_OM_SMALL_SHIFT; /* KDETH.OM and KDETH.OFFSET (TID) */ AHG_HEADER_SET(req->ahg, diff, 7, 0, 16, - ((!!(req->omfactor - KDETH_OM_SMALL)) << 15 | - ((req->tidoffset / req->omfactor) & 0x7fff))); + ((!!(omfactor - KDETH_OM_SMALL_SHIFT)) << 15 | + ((req->tidoffset >> omfactor) + & 0x7fff))); /* KDETH.TIDCtrl, KDETH.TID, KDETH.Intr, KDETH.SH */ val = cpu_to_le16(((EXP_TID_GET(tidval, CTRL) & 0x3) << 10) | (EXP_TID_GET(tidval, IDX) & 0x3ff)); @@ -1616,9 +1613,10 @@ static inline void set_comp_state(struct hfi1_user_sdma_pkt_q *pq, { hfi1_cdbg(SDMA, "[%u:%u:%u:%u] Setting completion status %u %d", pq->dd->unit, pq->ctxt, pq->subctxt, idx, state, ret); - cq->comps[idx].status = state; if (state == ERROR) cq->comps[idx].errcode = -ret; + smp_wmb(); /* make sure errcode is visible first */ + cq->comps[idx].status = state; trace_hfi1_sdma_user_completion(pq->dd, pq->ctxt, pq->subctxt, idx, state, ret); } diff --git a/drivers/infiniband/hw/hfi1/user_sdma.h b/drivers/infiniband/hw/hfi1/user_sdma.h index 39001714f551..e5b10aefe212 100644 --- a/drivers/infiniband/hw/hfi1/user_sdma.h +++ b/drivers/infiniband/hw/hfi1/user_sdma.h @@ -1,5 +1,7 @@ +#ifndef _HFI1_USER_SDMA_H +#define _HFI1_USER_SDMA_H /* - * Copyright(c) 2015, 2016 Intel Corporation. + * Copyright(c) 2015 - 2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -56,7 +58,7 @@ extern uint extended_psn; struct hfi1_user_sdma_pkt_q { struct list_head list; unsigned ctxt; - unsigned subctxt; + u16 subctxt; u16 n_max_reqs; atomic_t n_reqs; u16 reqidx; @@ -78,7 +80,11 @@ struct hfi1_user_sdma_comp_q { struct hfi1_sdma_comp_entry *comps; }; -int hfi1_user_sdma_alloc_queues(struct hfi1_ctxtdata *, struct file *); -int hfi1_user_sdma_free_queues(struct hfi1_filedata *); -int hfi1_user_sdma_process_request(struct file *, struct iovec *, unsigned long, - unsigned long *); +int hfi1_user_sdma_alloc_queues(struct hfi1_ctxtdata *uctxt, + struct hfi1_filedata *fd); +int hfi1_user_sdma_free_queues(struct hfi1_filedata *fd); +int hfi1_user_sdma_process_request(struct hfi1_filedata *fd, + struct iovec *iovec, unsigned long dim, + unsigned long *count); + +#endif /* _HFI1_USER_SDMA_H */ diff --git a/drivers/infiniband/hw/hfi1/verbs.c b/drivers/infiniband/hw/hfi1/verbs.c index 95ed4d6da510..2d19f9bb434d 100644 --- a/drivers/infiniband/hw/hfi1/verbs.c +++ b/drivers/infiniband/hw/hfi1/verbs.c @@ -1,5 +1,5 @@ /* - * Copyright(c) 2015, 2016 Intel Corporation. + * Copyright(c) 2015 - 2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -60,6 +60,8 @@ #include "trace.h" #include "qp.h" #include "verbs_txreq.h" +#include "debugfs.h" +#include "vnic.h" static unsigned int hfi1_lkey_table_size = 16; module_param_named(lkey_table_size, hfi1_lkey_table_size, uint, @@ -291,12 +293,28 @@ static void wss_insert(void *address) /* * Is the working set larger than the threshold? */ -static inline int wss_exceeds_threshold(void) +static inline bool wss_exceeds_threshold(void) { return atomic_read(&wss.total_count) >= wss.threshold; } /* + * Translate ib_wr_opcode into ib_wc_opcode. + */ +const enum ib_wc_opcode ib_hfi1_wc_opcode[] = { + [IB_WR_RDMA_WRITE] = IB_WC_RDMA_WRITE, + [IB_WR_RDMA_WRITE_WITH_IMM] = IB_WC_RDMA_WRITE, + [IB_WR_SEND] = IB_WC_SEND, + [IB_WR_SEND_WITH_IMM] = IB_WC_SEND, + [IB_WR_RDMA_READ] = IB_WC_RDMA_READ, + [IB_WR_ATOMIC_CMP_AND_SWP] = IB_WC_COMP_SWAP, + [IB_WR_ATOMIC_FETCH_AND_ADD] = IB_WC_FETCH_ADD, + [IB_WR_SEND_WITH_INV] = IB_WC_SEND, + [IB_WR_LOCAL_INV] = IB_WC_LOCAL_INV, + [IB_WR_REG_MR] = IB_WC_REG_MR +}; + +/* * Length of header by opcode, 0 --> not supported */ const u8 hdr_len_by_opcode[256] = { @@ -419,18 +437,19 @@ __be64 ib_hfi1_sys_image_guid; * @ss: the SGE state * @data: the data to copy * @length: the length of the data + * @release: boolean to release MR * @copy_last: do a separate copy of the last 8 bytes */ void hfi1_copy_sge( struct rvt_sge_state *ss, void *data, u32 length, - int release, - int copy_last) + bool release, + bool copy_last) { struct rvt_sge *sge = &ss->sge; - int in_last = 0; int i; - int cacheless_copy = 0; + bool in_last = false; + bool cacheless_copy = false; if (sge_copy_mode == COPY_CACHELESS) { cacheless_copy = length >= PAGE_SIZE; @@ -454,19 +473,15 @@ void hfi1_copy_sge( if (length > 8) { length -= 8; } else { - copy_last = 0; - in_last = 1; + copy_last = false; + in_last = true; } } again: while (length) { - u32 len = sge->length; + u32 len = rvt_get_sge_length(sge, length); - if (len > length) - len = length; - if (len > sge->sge_length) - len = sge->sge_length; WARN_ON_ONCE(len == 0); if (unlikely(in_last)) { /* enforce byte transfer ordering */ @@ -477,77 +492,19 @@ again: } else { memcpy(sge->vaddr, data, len); } - sge->vaddr += len; - sge->length -= len; - sge->sge_length -= len; - if (sge->sge_length == 0) { - if (release) - rvt_put_mr(sge->mr); - if (--ss->num_sge) - *sge = *ss->sg_list++; - } else if (sge->length == 0 && sge->mr->lkey) { - if (++sge->n >= RVT_SEGSZ) { - if (++sge->m >= sge->mr->mapsz) - break; - sge->n = 0; - } - sge->vaddr = - sge->mr->map[sge->m]->segs[sge->n].vaddr; - sge->length = - sge->mr->map[sge->m]->segs[sge->n].length; - } + rvt_update_sge(ss, len, release); data += len; length -= len; } if (copy_last) { - copy_last = 0; - in_last = 1; + copy_last = false; + in_last = true; length = 8; goto again; } } -/** - * hfi1_skip_sge - skip over SGE memory - * @ss: the SGE state - * @length: the number of bytes to skip - */ -void hfi1_skip_sge(struct rvt_sge_state *ss, u32 length, int release) -{ - struct rvt_sge *sge = &ss->sge; - - while (length) { - u32 len = sge->length; - - if (len > length) - len = length; - if (len > sge->sge_length) - len = sge->sge_length; - WARN_ON_ONCE(len == 0); - sge->vaddr += len; - sge->length -= len; - sge->sge_length -= len; - if (sge->sge_length == 0) { - if (release) - rvt_put_mr(sge->mr); - if (--ss->num_sge) - *sge = *ss->sg_list++; - } else if (sge->length == 0 && sge->mr->lkey) { - if (++sge->n >= RVT_SEGSZ) { - if (++sge->m >= sge->mr->mapsz) - break; - sge->n = 0; - } - sge->vaddr = - sge->mr->map[sge->m]->segs[sge->n].vaddr; - sge->length = - sge->mr->map[sge->m]->segs[sge->n].length; - } - length -= len; - } -} - /* * Make sure the QP is ready and able to accept the given opcode. */ @@ -562,6 +519,35 @@ static inline opcode_handler qp_ok(int opcode, struct hfi1_packet *packet) return NULL; } +static u64 hfi1_fault_tx(struct rvt_qp *qp, u8 opcode, u64 pbc) +{ +#ifdef CONFIG_FAULT_INJECTION + if ((opcode & IB_OPCODE_MSP) == IB_OPCODE_MSP) + /* + * In order to drop non-IB traffic we + * set PbcInsertHrc to NONE (0x2). + * The packet will still be delivered + * to the receiving node but a + * KHdrHCRCErr (KDETH packet with a bad + * HCRC) will be triggered and the + * packet will not be delivered to the + * correct context. + */ + pbc |= (u64)PBC_IHCRC_NONE << PBC_INSERT_HCRC_SHIFT; + else + /* + * In order to drop regular verbs + * traffic we set the PbcTestEbp + * flag. The packet will still be + * delivered to the receiving node but + * a 'late ebp error' will be + * triggered and will be dropped. + */ + pbc |= PBC_TEST_EBP; +#endif + return pbc; +} + /** * hfi1_ib_rcv - process an incoming packet * @packet: data packet information @@ -576,7 +562,7 @@ void hfi1_ib_rcv(struct hfi1_packet *packet) struct ib_header *hdr = packet->hdr; u32 tlen = packet->tlen; struct hfi1_pportdata *ppd = rcd->ppd; - struct hfi1_ibport *ibp = &ppd->ibport_data; + struct hfi1_ibport *ibp = rcd_to_iport(rcd); struct rvt_dev_info *rdi = &ppd->dd->verbs_dev.rdi; opcode_handler packet_handler; unsigned long flags; @@ -586,7 +572,7 @@ void hfi1_ib_rcv(struct hfi1_packet *packet) u16 lid; /* Check for GRH */ - lnh = be16_to_cpu(hdr->lrh[0]) & 3; + lnh = ib_get_lnh(hdr); if (lnh == HFI1_LRH_BTH) { packet->ohdr = &hdr->u.oth; } else if (lnh == HFI1_LRH_GRH) { @@ -605,12 +591,12 @@ void hfi1_ib_rcv(struct hfi1_packet *packet) trace_input_ibhdr(rcd->dd, hdr); - opcode = (be32_to_cpu(packet->ohdr->bth[0]) >> 24); + opcode = ib_bth_get_opcode(packet->ohdr); inc_opstats(tlen, &rcd->opstats->stats[opcode]); /* Get the destination QP number. */ qp_num = be32_to_cpu(packet->ohdr->bth[1]) & RVT_QPN_MASK; - lid = be16_to_cpu(hdr->lrh[1]); + lid = ib_get_dlid(hdr); if (unlikely((lid >= be16_to_cpu(IB_MULTICAST_LID_BASE)) && (lid != be16_to_cpu(IB_LID_PERMISSIVE)))) { struct rvt_mcast *mcast; @@ -618,7 +604,7 @@ void hfi1_ib_rcv(struct hfi1_packet *packet) if (lnh != HFI1_LRH_GRH) goto drop; - mcast = rvt_mcast_find(&ibp->rvp, &hdr->u.l.grh.dgid); + mcast = rvt_mcast_find(&ibp->rvp, &hdr->u.l.grh.dgid, lid); if (!mcast) goto drop; list_for_each_entry_rcu(p, &mcast->qp_list, list) { @@ -644,6 +630,11 @@ void hfi1_ib_rcv(struct hfi1_packet *packet) rcu_read_unlock(); goto drop; } + if (unlikely(hfi1_dbg_fault_opcode(packet->qp, opcode, + true))) { + rcu_read_unlock(); + goto drop; + } spin_lock_irqsave(&packet->qp->r_lock, flags); packet_handler = qp_ok(opcode, packet); if (likely(packet_handler)) @@ -689,27 +680,6 @@ static void mem_timer(unsigned long data) hfi1_qp_wakeup(qp, RVT_S_WAIT_KMEM); } -void update_sge(struct rvt_sge_state *ss, u32 length) -{ - struct rvt_sge *sge = &ss->sge; - - sge->vaddr += length; - sge->length -= length; - sge->sge_length -= length; - if (sge->sge_length == 0) { - if (--ss->num_sge) - *sge = *ss->sg_list++; - } else if (sge->length == 0 && sge->mr->lkey) { - if (++sge->n >= RVT_SEGSZ) { - if (++sge->m >= sge->mr->mapsz) - return; - sge->n = 0; - } - sge->vaddr = sge->mr->map[sge->m]->segs[sge->n].vaddr; - sge->length = sge->mr->map[sge->m]->segs[sge->n].length; - } -} - /* * This is called with progress side lock held. */ @@ -798,7 +768,7 @@ static noinline int build_verbs_ulp_payload( len); if (ret) goto bail_txadd; - update_sge(ss, len); + rvt_update_sge(ss, len, false); length -= len; } return ret; @@ -863,7 +833,6 @@ static int build_verbs_tx_desc( if (ret) goto bail_txadd; } - /* add the ulp payload - if any. tx->ss can be NULL for acks */ if (tx->ss) ret = build_verbs_ulp_payload(sde, length, tx); @@ -882,7 +851,6 @@ int hfi1_verbs_send_dma(struct rvt_qp *qp, struct hfi1_pkt_state *ps, struct hfi1_ibdev *dev = ps->dev; struct hfi1_pportdata *ppd = ps->ppd; struct verbs_txreq *tx; - u64 pbc_flags = 0; u8 sc5 = priv->s_sc; int ret; @@ -891,12 +859,16 @@ int hfi1_verbs_send_dma(struct rvt_qp *qp, struct hfi1_pkt_state *ps, if (!sdma_txreq_built(&tx->txreq)) { if (likely(pbc == 0)) { u32 vl = sc_to_vlt(dd_from_ibdev(qp->ibqp.device), sc5); + u8 opcode = get_opcode(&tx->phdr.hdr); + /* No vl15 here */ /* set PBC_DC_INFO bit (aka SC[4]) in pbc_flags */ - pbc_flags |= (!!(sc5 & 0x10)) << PBC_DC_INFO_SHIFT; + pbc |= (!!(sc5 & 0x10)) << PBC_DC_INFO_SHIFT; + if (unlikely(hfi1_dbg_fault_opcode(qp, opcode, false))) + pbc = hfi1_fault_tx(qp, opcode, pbc); pbc = create_pbc(ppd, - pbc_flags, + pbc, qp->srate_mbps, vl, plen); @@ -999,7 +971,6 @@ int hfi1_verbs_send_pio(struct rvt_qp *qp, struct hfi1_pkt_state *ps, u32 plen = hdrwords + dwords + 2; /* includes pbc */ struct hfi1_pportdata *ppd = ps->ppd; u32 *hdr = (u32 *)&ps->s_txreq->phdr.hdr; - u64 pbc_flags = 0; u8 sc5; unsigned long flags = 0; struct send_context *sc; @@ -1024,9 +995,14 @@ int hfi1_verbs_send_pio(struct rvt_qp *qp, struct hfi1_pkt_state *ps, if (likely(pbc == 0)) { u8 vl = sc_to_vlt(dd_from_ibdev(qp->ibqp.device), sc5); + struct verbs_txreq *tx = ps->s_txreq; + u8 opcode = get_opcode(&tx->phdr.hdr); + /* set PBC_DC_INFO bit (aka SC[4]) in pbc_flags */ - pbc_flags |= (!!(sc5 & 0x10)) << PBC_DC_INFO_SHIFT; - pbc = create_pbc(ppd, pbc_flags, qp->srate_mbps, vl, plen); + pbc |= (!!(sc5 & 0x10)) << PBC_DC_INFO_SHIFT; + if (unlikely(hfi1_dbg_fault_opcode(qp, opcode, false))) + pbc = hfi1_fault_tx(qp, opcode, pbc); + pbc = create_pbc(ppd, pbc, qp->srate_mbps, vl, plen); } if (cb) iowait_pio_inc(&priv->s_iowait); @@ -1073,7 +1049,7 @@ int hfi1_verbs_send_pio(struct rvt_qp *qp, struct hfi1_pkt_state *ps, if (slen > len) slen = len; - update_sge(ss, slen); + rvt_update_sge(ss, slen, false); seg_pio_copy_mid(pbuf, addr, slen); len -= slen; } @@ -1255,7 +1231,7 @@ int hfi1_verbs_send(struct rvt_qp *qp, struct hfi1_pkt_state *ps) hdr = &ps->s_txreq->phdr.hdr; /* locate the pkey within the headers */ - lnh = be16_to_cpu(hdr->lrh[0]) & 3; + lnh = ib_get_lnh(hdr); if (lnh == HFI1_LRH_GRH) ohdr = &hdr->u.l.oth; else @@ -1302,17 +1278,20 @@ int hfi1_verbs_send(struct rvt_qp *qp, struct hfi1_pkt_state *ps) static void hfi1_fill_device_attr(struct hfi1_devdata *dd) { struct rvt_dev_info *rdi = &dd->verbs_dev.rdi; - u16 ver = dd->dc8051_ver; + u32 ver = dd->dc8051_ver; memset(&rdi->dparms.props, 0, sizeof(rdi->dparms.props)); - rdi->dparms.props.fw_ver = ((u64)(dc8051_ver_maj(ver)) << 16) | - (u64)dc8051_ver_min(ver); + rdi->dparms.props.fw_ver = ((u64)(dc8051_ver_maj(ver)) << 32) | + ((u64)(dc8051_ver_min(ver)) << 16) | + (u64)dc8051_ver_patch(ver); + rdi->dparms.props.device_cap_flags = IB_DEVICE_BAD_PKEY_CNTR | IB_DEVICE_BAD_QKEY_CNTR | IB_DEVICE_SHUTDOWN_PORT | IB_DEVICE_SYS_IMAGE_GUID | IB_DEVICE_RC_RNR_NAK_GEN | IB_DEVICE_PORT_ACTIVE_EVENT | IB_DEVICE_SRQ_RESIZE | - IB_DEVICE_MEM_MGT_EXTENSIONS; + IB_DEVICE_MEM_MGT_EXTENSIONS | + IB_DEVICE_RDMA_NETDEV_OPA_VNIC; rdi->dparms.props.page_size_cap = PAGE_SIZE; rdi->dparms.props.vendor_id = dd->oui1 << 16 | dd->oui2 << 8 | dd->oui3; rdi->dparms.props.vendor_part_id = dd->pcidev->device; @@ -1384,6 +1363,7 @@ static int query_port(struct rvt_dev_info *rdi, u8 port_num, struct hfi1_pportdata *ppd = &dd->pport[port_num - 1]; u16 lid = ppd->lid; + /* props being zeroed by the caller, avoid zeroing it here */ props->lid = lid ? lid : 0; props->lmc = ppd->lmc; /* OPA logical states match IB logical states */ @@ -1479,14 +1459,14 @@ static int hfi1_get_guid_be(struct rvt_dev_info *rdi, struct rvt_ibport *rvp, /* * convert ah port,sl to sc */ -u8 ah_to_sc(struct ib_device *ibdev, struct ib_ah_attr *ah) +u8 ah_to_sc(struct ib_device *ibdev, struct rdma_ah_attr *ah) { - struct hfi1_ibport *ibp = to_iport(ibdev, ah->port_num); + struct hfi1_ibport *ibp = to_iport(ibdev, rdma_ah_get_port_num(ah)); - return ibp->sl_to_sc[ah->sl]; + return ibp->sl_to_sc[rdma_ah_get_sl(ah)]; } -static int hfi1_check_ah(struct ib_device *ibdev, struct ib_ah_attr *ah_attr) +static int hfi1_check_ah(struct ib_device *ibdev, struct rdma_ah_attr *ah_attr) { struct hfi1_ibport *ibp; struct hfi1_pportdata *ppd; @@ -1494,9 +1474,9 @@ static int hfi1_check_ah(struct ib_device *ibdev, struct ib_ah_attr *ah_attr) u8 sc5; /* test the mapping for validity */ - ibp = to_iport(ibdev, ah_attr->port_num); + ibp = to_iport(ibdev, rdma_ah_get_port_num(ah_attr)); ppd = ppd_from_ibp(ibp); - sc5 = ibp->sl_to_sc[ah_attr->sl]; + sc5 = ibp->sl_to_sc[rdma_ah_get_sl(ah_attr)]; dd = dd_from_ppd(ppd); if (sc_to_vlt(dd, sc5) > num_vls && sc_to_vlt(dd, sc5) != 0xf) return -EINVAL; @@ -1504,7 +1484,7 @@ static int hfi1_check_ah(struct ib_device *ibdev, struct ib_ah_attr *ah_attr) } static void hfi1_notify_new_ah(struct ib_device *ibdev, - struct ib_ah_attr *ah_attr, + struct rdma_ah_attr *ah_attr, struct rvt_ah *ah) { struct hfi1_ibport *ibp; @@ -1517,9 +1497,9 @@ static void hfi1_notify_new_ah(struct ib_device *ibdev, * done being setup. We can however modify things which we need to set. */ - ibp = to_iport(ibdev, ah_attr->port_num); + ibp = to_iport(ibdev, rdma_ah_get_port_num(ah_attr)); ppd = ppd_from_ibp(ibp); - sc5 = ibp->sl_to_sc[ah->attr.sl]; + sc5 = ibp->sl_to_sc[rdma_ah_get_sl(&ah->attr)]; dd = dd_from_ppd(ppd); ah->vl = sc_to_vlt(dd, sc5); if (ah->vl < num_vls || ah->vl == 15) @@ -1528,17 +1508,21 @@ static void hfi1_notify_new_ah(struct ib_device *ibdev, struct ib_ah *hfi1_create_qp0_ah(struct hfi1_ibport *ibp, u16 dlid) { - struct ib_ah_attr attr; + struct rdma_ah_attr attr; struct ib_ah *ah = ERR_PTR(-EINVAL); struct rvt_qp *qp0; + struct hfi1_pportdata *ppd = ppd_from_ibp(ibp); + struct hfi1_devdata *dd = dd_from_ppd(ppd); + u8 port_num = ppd->port; memset(&attr, 0, sizeof(attr)); - attr.dlid = dlid; - attr.port_num = ppd_from_ibp(ibp)->port; + attr.type = rdma_ah_find_type(&dd->verbs_dev.rdi.ibdev, port_num); + rdma_ah_set_dlid(&attr, dlid); + rdma_ah_set_port_num(&attr, ppd_from_ibp(ibp)->port); rcu_read_lock(); qp0 = rcu_dereference(ibp->rvp.qp[0]); if (qp0) - ah = ib_create_ah(qp0->ibqp.pd, &attr); + ah = rdma_create_ah(qp0->ibqp.pd, &attr); rcu_read_unlock(); return ah; } @@ -1585,10 +1569,10 @@ static void hfi1_get_dev_fw_str(struct ib_device *ibdev, char *str, { struct rvt_dev_info *rdi = ib_to_rvt(ibdev); struct hfi1_ibdev *dev = dev_from_rdi(rdi); - u16 ver = dd_from_dev(dev)->dc8051_ver; + u32 ver = dd_from_dev(dev)->dc8051_ver; - snprintf(str, str_len, "%u.%u", dc8051_ver_maj(ver), - dc8051_ver_min(ver)); + snprintf(str, str_len, "%u.%u.%u", dc8051_ver_maj(ver), + dc8051_ver_min(ver), dc8051_ver_patch(ver)); } static const char * const driver_cntr_names[] = { @@ -1605,6 +1589,7 @@ static const char * const driver_cntr_names[] = { "DRIVER_EgrHdrFull" }; +static DEFINE_MUTEX(cntr_names_lock); /* protects the *_cntr_names bufers */ static const char **dev_cntr_names; static const char **port_cntr_names; static int num_driver_cntrs = ARRAY_SIZE(driver_cntr_names); @@ -1618,7 +1603,7 @@ static int cntr_names_initialized; * external strings. */ static int init_cntr_names(const char *names_in, - const int names_len, + const size_t names_len, int num_extra_names, int *num_cntrs, const char ***cntr_names) @@ -1659,6 +1644,7 @@ static struct rdma_hw_stats *alloc_hw_stats(struct ib_device *ibdev, { int i, err; + mutex_lock(&cntr_names_lock); if (!cntr_names_initialized) { struct hfi1_devdata *dd = dd_from_ibdev(ibdev); @@ -1667,8 +1653,10 @@ static struct rdma_hw_stats *alloc_hw_stats(struct ib_device *ibdev, num_driver_cntrs, &num_dev_cntrs, &dev_cntr_names); - if (err) + if (err) { + mutex_unlock(&cntr_names_lock); return NULL; + } for (i = 0; i < num_driver_cntrs; i++) dev_cntr_names[num_dev_cntrs + i] = @@ -1682,10 +1670,12 @@ static struct rdma_hw_stats *alloc_hw_stats(struct ib_device *ibdev, if (err) { kfree(dev_cntr_names); dev_cntr_names = NULL; + mutex_unlock(&cntr_names_lock); return NULL; } cntr_names_initialized = 1; } + mutex_unlock(&cntr_names_lock); if (!port_num) return rdma_alloc_hw_stats_struct( @@ -1784,10 +1774,11 @@ int hfi1_register_ib_device(struct hfi1_devdata *dd) strlcpy(ibdev->name + lcpysz, "_%d", IB_DEVICE_NAME_MAX - lcpysz); ibdev->owner = THIS_MODULE; ibdev->phys_port_cnt = dd->num_pports; - ibdev->dma_device = &dd->pcidev->dev; + ibdev->dev.parent = &dd->pcidev->dev; ibdev->modify_device = modify_device; ibdev->alloc_hw_stats = alloc_hw_stats; ibdev->get_hw_stats = get_hw_stats; + ibdev->alloc_rdma_netdev = hfi1_vnic_alloc_rn; /* keep process mad in the driver */ ibdev->process_mad = hfi1_process_mad; @@ -1832,7 +1823,7 @@ int hfi1_register_ib_device(struct hfi1_devdata *dd) dd->verbs_dev.rdi.driver_f.qp_priv_free = qp_priv_free; dd->verbs_dev.rdi.driver_f.free_all_qps = free_all_qps; dd->verbs_dev.rdi.driver_f.notify_qp_reset = notify_qp_reset; - dd->verbs_dev.rdi.driver_f.do_send = hfi1_do_send; + dd->verbs_dev.rdi.driver_f.do_send = hfi1_do_send_from_rvt; dd->verbs_dev.rdi.driver_f.schedule_send = hfi1_schedule_send; dd->verbs_dev.rdi.driver_f.schedule_send_no_lock = _hfi1_schedule_send; dd->verbs_dev.rdi.driver_f.get_pmtu_from_attr = get_pmtu_from_attr; @@ -1845,6 +1836,7 @@ int hfi1_register_ib_device(struct hfi1_devdata *dd) dd->verbs_dev.rdi.driver_f.mtu_to_path_mtu = mtu_to_path_mtu; dd->verbs_dev.rdi.driver_f.check_modify_qp = hfi1_check_modify_qp; dd->verbs_dev.rdi.driver_f.modify_qp = hfi1_modify_qp; + dd->verbs_dev.rdi.driver_f.notify_restart_rc = hfi1_restart_rc; dd->verbs_dev.rdi.driver_f.check_send_wqe = hfi1_check_send_wqe; /* completeion queue */ @@ -1903,14 +1895,18 @@ void hfi1_unregister_ib_device(struct hfi1_devdata *dd) del_timer_sync(&dev->mem_timer); verbs_txreq_exit(dev); + mutex_lock(&cntr_names_lock); kfree(dev_cntr_names); kfree(port_cntr_names); + dev_cntr_names = NULL; + port_cntr_names = NULL; cntr_names_initialized = 0; + mutex_unlock(&cntr_names_lock); } void hfi1_cnp_rcv(struct hfi1_packet *packet) { - struct hfi1_ibport *ibp = &packet->rcd->ppd->ibport_data; + struct hfi1_ibport *ibp = rcd_to_iport(packet->rcd); struct hfi1_pportdata *ppd = ppd_from_ibp(ibp); struct ib_header *hdr = packet->hdr; struct rvt_qp *qp = packet->qp; @@ -1920,12 +1916,12 @@ void hfi1_cnp_rcv(struct hfi1_packet *packet) switch (packet->qp->ibqp.qp_type) { case IB_QPT_UC: - rlid = qp->remote_ah_attr.dlid; + rlid = rdma_ah_get_dlid(&qp->remote_ah_attr); rqpn = qp->remote_qpn; svc_type = IB_CC_SVCTYPE_UC; break; case IB_QPT_RC: - rlid = qp->remote_ah_attr.dlid; + rlid = rdma_ah_get_dlid(&qp->remote_ah_attr); rqpn = qp->remote_qpn; svc_type = IB_CC_SVCTYPE_RC; break; @@ -1939,7 +1935,7 @@ void hfi1_cnp_rcv(struct hfi1_packet *packet) return; } - sc5 = hdr2sc(hdr, packet->rhf); + sc5 = hfi1_9B_get_sc5(hdr, packet->rhf); sl = ibp->sc_to_sl[sc5]; lqpn = qp->ibqp.qp_num; diff --git a/drivers/infiniband/hw/hfi1/verbs.h b/drivers/infiniband/hw/hfi1/verbs.h index e6b893010e6d..cd635d0c1d3b 100644 --- a/drivers/infiniband/hw/hfi1/verbs.h +++ b/drivers/infiniband/hw/hfi1/verbs.h @@ -1,5 +1,5 @@ /* - * Copyright(c) 2015, 2016 Intel Corporation. + * Copyright(c) 2015 - 2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -125,9 +125,7 @@ struct hfi1_qp_priv { struct sdma_engine *s_sde; /* current sde */ struct send_context *s_sendcontext; /* current sendcontext */ u8 s_sc; /* SC[0..4] for next packet */ - u8 r_adefered; /* number of acks defered */ struct iowait s_iowait; - struct timer_list s_rnr_timer; struct rvt_qp *owner; }; @@ -141,6 +139,10 @@ struct hfi1_pkt_state { struct hfi1_pportdata *ppd; struct verbs_txreq *s_txreq; unsigned long flags; + unsigned long timeout; + unsigned long timeout_int; + int cpu; + bool in_thread; }; #define HFI1_PSN_CREDIT 16 @@ -196,6 +198,11 @@ struct hfi1_ibdev { struct dentry *hfi1_ibdev_dbg; /* per HFI symlinks to above */ struct dentry *hfi1_ibdev_link; +#ifdef CONFIG_FAULT_INJECTION + struct fault_opcode *fault_opcode; + struct fault_packet *fault_packet; + bool fault_suppress_err; +#endif #endif }; @@ -260,15 +267,6 @@ int hfi1_process_mad(struct ib_device *ibdev, int mad_flags, u8 port, #define PSN_MODIFY_MASK 0xFFFFFF /* - * Compare the lower 24 bits of the msn values. - * Returns an integer <, ==, or > than zero. - */ -static inline int cmp_msn(u32 a, u32 b) -{ - return (((int)a) - ((int)b)) << 8; -} - -/* * Compare two PSNs * Returns an integer <, ==, or > than zero. */ @@ -299,9 +297,7 @@ void hfi1_put_txreq(struct verbs_txreq *tx); int hfi1_verbs_send(struct rvt_qp *qp, struct hfi1_pkt_state *ps); void hfi1_copy_sge(struct rvt_sge_state *ss, void *data, u32 length, - int release, int copy_last); - -void hfi1_skip_sge(struct rvt_sge_state *ss, u32 length, int release); + bool release, bool copy_last); void hfi1_cnp_rcv(struct hfi1_packet *packet); @@ -315,20 +311,12 @@ void hfi1_rc_hdrerr( u32 rcv_flags, struct rvt_qp *qp); -u8 ah_to_sc(struct ib_device *ibdev, struct ib_ah_attr *ah_attr); +u8 ah_to_sc(struct ib_device *ibdev, struct rdma_ah_attr *ah_attr); struct ib_ah *hfi1_create_qp0_ah(struct hfi1_ibport *ibp, u16 dlid); -void hfi1_rc_rnr_retry(unsigned long arg); -void hfi1_add_rnr_timer(struct rvt_qp *qp, u32 to); -void hfi1_rc_timeout(unsigned long arg); -void hfi1_del_timers_sync(struct rvt_qp *qp); -void hfi1_stop_rc_timers(struct rvt_qp *qp); - void hfi1_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr); -void hfi1_rc_error(struct rvt_qp *qp, enum ib_wc_status err); - void hfi1_ud_rcv(struct hfi1_packet *packet); int hfi1_lookup_pkey_idx(struct hfi1_ibport *ibp, u16 pkey); @@ -342,7 +330,7 @@ int hfi1_check_modify_qp(struct rvt_qp *qp, struct ib_qp_attr *attr, void hfi1_modify_qp(struct rvt_qp *qp, struct ib_qp_attr *attr, int attr_mask, struct ib_udata *udata); - +void hfi1_restart_rc(struct rvt_qp *qp, u32 psn, int wait); int hfi1_check_send_wqe(struct rvt_qp *qp, struct rvt_swqe *wqe); extern const u32 rc_only_opcode; @@ -362,7 +350,7 @@ int hfi1_ruc_check_hdr(struct hfi1_ibport *ibp, struct ib_header *hdr, int has_grh, struct rvt_qp *qp, u32 bth0); u32 hfi1_make_grh(struct hfi1_ibport *ibp, struct ib_grh *hdr, - struct ib_global_route *grh, u32 hwords, u32 nwords); + const struct ib_global_route *grh, u32 hwords, u32 nwords); void hfi1_make_ruc_header(struct rvt_qp *qp, struct ib_other_headers *ohdr, u32 bth0, u32 bth2, int middle, @@ -370,7 +358,9 @@ void hfi1_make_ruc_header(struct rvt_qp *qp, struct ib_other_headers *ohdr, void _hfi1_do_send(struct work_struct *work); -void hfi1_do_send(struct rvt_qp *qp); +void hfi1_do_send_from_rvt(struct rvt_qp *qp); + +void hfi1_do_send(struct rvt_qp *qp, bool in_thread); void hfi1_send_complete(struct rvt_qp *qp, struct rvt_swqe *wqe, enum ib_wc_status status); diff --git a/drivers/infiniband/hw/hfi1/vnic.h b/drivers/infiniband/hw/hfi1/vnic.h new file mode 100644 index 000000000000..4a621cde4abb --- /dev/null +++ b/drivers/infiniband/hw/hfi1/vnic.h @@ -0,0 +1,183 @@ +#ifndef _HFI1_VNIC_H +#define _HFI1_VNIC_H +/* + * Copyright(c) 2017 Intel Corporation. + * + * This file is provided under a dual BSD/GPLv2 license. When using or + * redistributing this file, you may do so under either license. + * + * GPL LICENSE SUMMARY + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * BSD LICENSE + * + * 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. + * - Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + */ + +#include <rdma/opa_vnic.h> +#include "hfi.h" +#include "sdma.h" + +#define HFI1_VNIC_MAX_TXQ 16 +#define HFI1_VNIC_MAX_PAD 12 + +/* L2 header definitions */ +#define HFI1_L2_TYPE_OFFSET 0x7 +#define HFI1_L2_TYPE_SHFT 0x5 +#define HFI1_L2_TYPE_MASK 0x3 + +#define HFI1_GET_L2_TYPE(hdr) \ + ((*((u8 *)(hdr) + HFI1_L2_TYPE_OFFSET) >> HFI1_L2_TYPE_SHFT) & \ + HFI1_L2_TYPE_MASK) + +/* L4 type definitions */ +#define HFI1_L4_TYPE_OFFSET 8 + +#define HFI1_GET_L4_TYPE(data) \ + (*((u8 *)(data) + HFI1_L4_TYPE_OFFSET)) + +/* L4 header definitions */ +#define HFI1_VNIC_L4_HDR_OFFSET OPA_VNIC_L2_HDR_LEN + +#define HFI1_VNIC_GET_L4_HDR(data) \ + (*((u16 *)((u8 *)(data) + HFI1_VNIC_L4_HDR_OFFSET))) + +#define HFI1_VNIC_GET_VESWID(data) \ + (HFI1_VNIC_GET_L4_HDR(data) & 0xFFF) + +/* Service class */ +#define HFI1_VNIC_SC_OFFSET_LOW 6 +#define HFI1_VNIC_SC_OFFSET_HI 7 +#define HFI1_VNIC_SC_SHIFT 4 + +#define HFI1_VNIC_MAX_QUEUE 16 + +/** + * struct hfi1_vnic_sdma - VNIC per Tx ring SDMA information + * @dd - device data pointer + * @sde - sdma engine + * @vinfo - vnic info pointer + * @wait - iowait structure + * @stx - sdma tx request + * @state - vnic Tx ring SDMA state + * @q_idx - vnic Tx queue index + */ +struct hfi1_vnic_sdma { + struct hfi1_devdata *dd; + struct sdma_engine *sde; + struct hfi1_vnic_vport_info *vinfo; + struct iowait wait; + struct sdma_txreq stx; + unsigned int state; + u8 q_idx; +}; + +/** + * struct hfi1_vnic_rx_queue - HFI1 VNIC receive queue + * @idx: queue index + * @vinfo: pointer to vport information + * @netdev: network device + * @napi: netdev napi structure + * @skbq: queue of received socket buffers + */ +struct hfi1_vnic_rx_queue { + u8 idx; + struct hfi1_vnic_vport_info *vinfo; + struct net_device *netdev; + struct napi_struct napi; + struct sk_buff_head skbq; +}; + +/** + * struct hfi1_vnic_vport_info - HFI1 VNIC virtual port information + * @dd: device data pointer + * @netdev: net device pointer + * @flags: state flags + * @lock: vport lock + * @num_tx_q: number of transmit queues + * @num_rx_q: number of receive queues + * @vesw_id: virtual switch id + * @rxq: Array of receive queues + * @stats: per queue stats + * @sdma: VNIC SDMA structure per TXQ + */ +struct hfi1_vnic_vport_info { + struct hfi1_devdata *dd; + struct net_device *netdev; + unsigned long flags; + + /* Lock used around state updates */ + struct mutex lock; + + u8 num_tx_q; + u8 num_rx_q; + u16 vesw_id; + struct hfi1_vnic_rx_queue rxq[HFI1_NUM_VNIC_CTXT]; + + struct opa_vnic_stats stats[HFI1_VNIC_MAX_QUEUE]; + struct hfi1_vnic_sdma sdma[HFI1_VNIC_MAX_TXQ]; +}; + +#define v_dbg(format, arg...) \ + netdev_dbg(vinfo->netdev, format, ## arg) +#define v_err(format, arg...) \ + netdev_err(vinfo->netdev, format, ## arg) +#define v_info(format, arg...) \ + netdev_info(vinfo->netdev, format, ## arg) + +/* vnic hfi1 internal functions */ +void hfi1_vnic_setup(struct hfi1_devdata *dd); +void hfi1_vnic_cleanup(struct hfi1_devdata *dd); +int hfi1_vnic_txreq_init(struct hfi1_devdata *dd); +void hfi1_vnic_txreq_deinit(struct hfi1_devdata *dd); + +void hfi1_vnic_bypass_rcv(struct hfi1_packet *packet); +void hfi1_vnic_sdma_init(struct hfi1_vnic_vport_info *vinfo); +bool hfi1_vnic_sdma_write_avail(struct hfi1_vnic_vport_info *vinfo, + u8 q_idx); + +/* vnic rdma netdev operations */ +struct net_device *hfi1_vnic_alloc_rn(struct ib_device *device, + u8 port_num, + enum rdma_netdev_t type, + const char *name, + unsigned char name_assign_type, + void (*setup)(struct net_device *)); +int hfi1_vnic_send_dma(struct hfi1_devdata *dd, u8 q_idx, + struct hfi1_vnic_vport_info *vinfo, + struct sk_buff *skb, u64 pbc, u8 plen); + +#endif /* _HFI1_VNIC_H */ diff --git a/drivers/infiniband/hw/hfi1/vnic_main.c b/drivers/infiniband/hw/hfi1/vnic_main.c new file mode 100644 index 000000000000..339f0cdd56d6 --- /dev/null +++ b/drivers/infiniband/hw/hfi1/vnic_main.c @@ -0,0 +1,904 @@ +/* + * Copyright(c) 2017 Intel Corporation. + * + * This file is provided under a dual BSD/GPLv2 license. When using or + * redistributing this file, you may do so under either license. + * + * GPL LICENSE SUMMARY + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * BSD LICENSE + * + * 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. + * - Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + */ + +/* + * This file contains HFI1 support for VNIC functionality + */ + +#include <linux/io.h> +#include <linux/if_vlan.h> + +#include "vnic.h" + +#define HFI_TX_TIMEOUT_MS 1000 + +#define HFI1_VNIC_RCV_Q_SIZE 1024 + +#define HFI1_VNIC_UP 0 + +static DEFINE_SPINLOCK(vport_cntr_lock); + +static int setup_vnic_ctxt(struct hfi1_devdata *dd, struct hfi1_ctxtdata *uctxt) +{ + unsigned int rcvctrl_ops = 0; + int ret; + + hfi1_init_ctxt(uctxt->sc); + + uctxt->do_interrupt = &handle_receive_interrupt; + + /* Now allocate the RcvHdr queue and eager buffers. */ + ret = hfi1_create_rcvhdrq(dd, uctxt); + if (ret) + goto done; + + ret = hfi1_setup_eagerbufs(uctxt); + if (ret) + goto done; + + if (uctxt->rcvhdrtail_kvaddr) + clear_rcvhdrtail(uctxt); + + rcvctrl_ops = HFI1_RCVCTRL_CTXT_ENB; + rcvctrl_ops |= HFI1_RCVCTRL_INTRAVAIL_ENB; + + if (!HFI1_CAP_KGET_MASK(uctxt->flags, MULTI_PKT_EGR)) + rcvctrl_ops |= HFI1_RCVCTRL_ONE_PKT_EGR_ENB; + if (HFI1_CAP_KGET_MASK(uctxt->flags, NODROP_EGR_FULL)) + rcvctrl_ops |= HFI1_RCVCTRL_NO_EGR_DROP_ENB; + if (HFI1_CAP_KGET_MASK(uctxt->flags, NODROP_RHQ_FULL)) + rcvctrl_ops |= HFI1_RCVCTRL_NO_RHQ_DROP_ENB; + if (HFI1_CAP_KGET_MASK(uctxt->flags, DMA_RTAIL)) + rcvctrl_ops |= HFI1_RCVCTRL_TAILUPD_ENB; + + hfi1_rcvctrl(uctxt->dd, rcvctrl_ops, uctxt->ctxt); + + uctxt->is_vnic = true; +done: + return ret; +} + +static int allocate_vnic_ctxt(struct hfi1_devdata *dd, + struct hfi1_ctxtdata **vnic_ctxt) +{ + struct hfi1_ctxtdata *uctxt; + unsigned int ctxt; + int ret; + + if (dd->flags & HFI1_FROZEN) + return -EIO; + + for (ctxt = dd->first_dyn_alloc_ctxt; + ctxt < dd->num_rcv_contexts; ctxt++) + if (!dd->rcd[ctxt]) + break; + + if (ctxt == dd->num_rcv_contexts) + return -EBUSY; + + uctxt = hfi1_create_ctxtdata(dd->pport, ctxt, dd->node); + if (!uctxt) { + dd_dev_err(dd, "Unable to create ctxtdata, failing open\n"); + return -ENOMEM; + } + + uctxt->flags = HFI1_CAP_KGET(MULTI_PKT_EGR) | + HFI1_CAP_KGET(NODROP_RHQ_FULL) | + HFI1_CAP_KGET(NODROP_EGR_FULL) | + HFI1_CAP_KGET(DMA_RTAIL); + uctxt->seq_cnt = 1; + + /* Allocate and enable a PIO send context */ + uctxt->sc = sc_alloc(dd, SC_VNIC, uctxt->rcvhdrqentsize, + uctxt->numa_id); + + ret = uctxt->sc ? 0 : -ENOMEM; + if (ret) + goto bail; + + dd_dev_dbg(dd, "allocated vnic send context %u(%u)\n", + uctxt->sc->sw_index, uctxt->sc->hw_context); + ret = sc_enable(uctxt->sc); + if (ret) + goto bail; + + if (dd->num_msix_entries) + hfi1_set_vnic_msix_info(uctxt); + + hfi1_stats.sps_ctxts++; + dd_dev_dbg(dd, "created vnic context %d\n", uctxt->ctxt); + *vnic_ctxt = uctxt; + + return ret; +bail: + /* + * hfi1_free_ctxtdata() also releases send_context + * structure if uctxt->sc is not null + */ + dd->rcd[uctxt->ctxt] = NULL; + hfi1_free_ctxtdata(dd, uctxt); + dd_dev_dbg(dd, "vnic allocation failed. rc %d\n", ret); + return ret; +} + +static void deallocate_vnic_ctxt(struct hfi1_devdata *dd, + struct hfi1_ctxtdata *uctxt) +{ + unsigned long flags; + + dd_dev_dbg(dd, "closing vnic context %d\n", uctxt->ctxt); + flush_wc(); + + if (dd->num_msix_entries) + hfi1_reset_vnic_msix_info(uctxt); + + spin_lock_irqsave(&dd->uctxt_lock, flags); + /* + * Disable receive context and interrupt available, reset all + * RcvCtxtCtrl bits to default values. + */ + hfi1_rcvctrl(dd, HFI1_RCVCTRL_CTXT_DIS | + HFI1_RCVCTRL_TIDFLOW_DIS | + HFI1_RCVCTRL_INTRAVAIL_DIS | + HFI1_RCVCTRL_ONE_PKT_EGR_DIS | + HFI1_RCVCTRL_NO_RHQ_DROP_DIS | + HFI1_RCVCTRL_NO_EGR_DROP_DIS, uctxt->ctxt); + /* + * VNIC contexts are allocated from user context pool. + * Release them back to user context pool. + * + * Reset context integrity checks to default. + * (writes to CSRs probably belong in chip.c) + */ + write_kctxt_csr(dd, uctxt->sc->hw_context, SEND_CTXT_CHECK_ENABLE, + hfi1_pkt_default_send_ctxt_mask(dd, SC_USER)); + sc_disable(uctxt->sc); + + dd->send_contexts[uctxt->sc->sw_index].type = SC_USER; + spin_unlock_irqrestore(&dd->uctxt_lock, flags); + + dd->rcd[uctxt->ctxt] = NULL; + uctxt->event_flags = 0; + + hfi1_clear_tids(uctxt); + hfi1_clear_ctxt_pkey(dd, uctxt); + + hfi1_stats.sps_ctxts--; + hfi1_free_ctxtdata(dd, uctxt); +} + +void hfi1_vnic_setup(struct hfi1_devdata *dd) +{ + idr_init(&dd->vnic.vesw_idr); +} + +void hfi1_vnic_cleanup(struct hfi1_devdata *dd) +{ + idr_destroy(&dd->vnic.vesw_idr); +} + +#define SUM_GRP_COUNTERS(stats, qstats, x_grp) do { \ + u64 *src64, *dst64; \ + for (src64 = &qstats->x_grp.unicast, \ + dst64 = &stats->x_grp.unicast; \ + dst64 <= &stats->x_grp.s_1519_max;) { \ + *dst64++ += *src64++; \ + } \ + } while (0) + +/* hfi1_vnic_update_stats - update statistics */ +static void hfi1_vnic_update_stats(struct hfi1_vnic_vport_info *vinfo, + struct opa_vnic_stats *stats) +{ + struct net_device *netdev = vinfo->netdev; + u8 i; + + /* add tx counters on different queues */ + for (i = 0; i < vinfo->num_tx_q; i++) { + struct opa_vnic_stats *qstats = &vinfo->stats[i]; + struct rtnl_link_stats64 *qnstats = &vinfo->stats[i].netstats; + + stats->netstats.tx_fifo_errors += qnstats->tx_fifo_errors; + stats->netstats.tx_carrier_errors += qnstats->tx_carrier_errors; + stats->tx_drop_state += qstats->tx_drop_state; + stats->tx_dlid_zero += qstats->tx_dlid_zero; + + SUM_GRP_COUNTERS(stats, qstats, tx_grp); + stats->netstats.tx_packets += qnstats->tx_packets; + stats->netstats.tx_bytes += qnstats->tx_bytes; + } + + /* add rx counters on different queues */ + for (i = 0; i < vinfo->num_rx_q; i++) { + struct opa_vnic_stats *qstats = &vinfo->stats[i]; + struct rtnl_link_stats64 *qnstats = &vinfo->stats[i].netstats; + + stats->netstats.rx_fifo_errors += qnstats->rx_fifo_errors; + stats->netstats.rx_nohandler += qnstats->rx_nohandler; + stats->rx_drop_state += qstats->rx_drop_state; + stats->rx_oversize += qstats->rx_oversize; + stats->rx_runt += qstats->rx_runt; + + SUM_GRP_COUNTERS(stats, qstats, rx_grp); + stats->netstats.rx_packets += qnstats->rx_packets; + stats->netstats.rx_bytes += qnstats->rx_bytes; + } + + stats->netstats.tx_errors = stats->netstats.tx_fifo_errors + + stats->netstats.tx_carrier_errors + + stats->tx_drop_state + stats->tx_dlid_zero; + stats->netstats.tx_dropped = stats->netstats.tx_errors; + + stats->netstats.rx_errors = stats->netstats.rx_fifo_errors + + stats->netstats.rx_nohandler + + stats->rx_drop_state + stats->rx_oversize + + stats->rx_runt; + stats->netstats.rx_dropped = stats->netstats.rx_errors; + + netdev->stats.tx_packets = stats->netstats.tx_packets; + netdev->stats.tx_bytes = stats->netstats.tx_bytes; + netdev->stats.tx_fifo_errors = stats->netstats.tx_fifo_errors; + netdev->stats.tx_carrier_errors = stats->netstats.tx_carrier_errors; + netdev->stats.tx_errors = stats->netstats.tx_errors; + netdev->stats.tx_dropped = stats->netstats.tx_dropped; + + netdev->stats.rx_packets = stats->netstats.rx_packets; + netdev->stats.rx_bytes = stats->netstats.rx_bytes; + netdev->stats.rx_fifo_errors = stats->netstats.rx_fifo_errors; + netdev->stats.multicast = stats->rx_grp.mcastbcast; + netdev->stats.rx_length_errors = stats->rx_oversize + stats->rx_runt; + netdev->stats.rx_errors = stats->netstats.rx_errors; + netdev->stats.rx_dropped = stats->netstats.rx_dropped; +} + +/* update_len_counters - update pkt's len histogram counters */ +static inline void update_len_counters(struct opa_vnic_grp_stats *grp, + int len) +{ + /* account for 4 byte FCS */ + if (len >= 1515) + grp->s_1519_max++; + else if (len >= 1020) + grp->s_1024_1518++; + else if (len >= 508) + grp->s_512_1023++; + else if (len >= 252) + grp->s_256_511++; + else if (len >= 124) + grp->s_128_255++; + else if (len >= 61) + grp->s_65_127++; + else + grp->s_64++; +} + +/* hfi1_vnic_update_tx_counters - update transmit counters */ +static void hfi1_vnic_update_tx_counters(struct hfi1_vnic_vport_info *vinfo, + u8 q_idx, struct sk_buff *skb, int err) +{ + struct ethhdr *mac_hdr = (struct ethhdr *)skb_mac_header(skb); + struct opa_vnic_stats *stats = &vinfo->stats[q_idx]; + struct opa_vnic_grp_stats *tx_grp = &stats->tx_grp; + u16 vlan_tci; + + stats->netstats.tx_packets++; + stats->netstats.tx_bytes += skb->len + ETH_FCS_LEN; + + update_len_counters(tx_grp, skb->len); + + /* rest of the counts are for good packets only */ + if (unlikely(err)) + return; + + if (is_multicast_ether_addr(mac_hdr->h_dest)) + tx_grp->mcastbcast++; + else + tx_grp->unicast++; + + if (!__vlan_get_tag(skb, &vlan_tci)) + tx_grp->vlan++; + else + tx_grp->untagged++; +} + +/* hfi1_vnic_update_rx_counters - update receive counters */ +static void hfi1_vnic_update_rx_counters(struct hfi1_vnic_vport_info *vinfo, + u8 q_idx, struct sk_buff *skb, int err) +{ + struct ethhdr *mac_hdr = (struct ethhdr *)skb->data; + struct opa_vnic_stats *stats = &vinfo->stats[q_idx]; + struct opa_vnic_grp_stats *rx_grp = &stats->rx_grp; + u16 vlan_tci; + + stats->netstats.rx_packets++; + stats->netstats.rx_bytes += skb->len + ETH_FCS_LEN; + + update_len_counters(rx_grp, skb->len); + + /* rest of the counts are for good packets only */ + if (unlikely(err)) + return; + + if (is_multicast_ether_addr(mac_hdr->h_dest)) + rx_grp->mcastbcast++; + else + rx_grp->unicast++; + + if (!__vlan_get_tag(skb, &vlan_tci)) + rx_grp->vlan++; + else + rx_grp->untagged++; +} + +/* This function is overloaded for opa_vnic specific implementation */ +static void hfi1_vnic_get_stats64(struct net_device *netdev, + struct rtnl_link_stats64 *stats) +{ + struct opa_vnic_stats *vstats = (struct opa_vnic_stats *)stats; + struct hfi1_vnic_vport_info *vinfo = opa_vnic_dev_priv(netdev); + + hfi1_vnic_update_stats(vinfo, vstats); +} + +static u64 create_bypass_pbc(u32 vl, u32 dw_len) +{ + u64 pbc; + + pbc = ((u64)PBC_IHCRC_NONE << PBC_INSERT_HCRC_SHIFT) + | PBC_INSERT_BYPASS_ICRC | PBC_CREDIT_RETURN + | PBC_PACKET_BYPASS + | ((vl & PBC_VL_MASK) << PBC_VL_SHIFT) + | (dw_len & PBC_LENGTH_DWS_MASK) << PBC_LENGTH_DWS_SHIFT; + + return pbc; +} + +/* hfi1_vnic_maybe_stop_tx - stop tx queue if required */ +static void hfi1_vnic_maybe_stop_tx(struct hfi1_vnic_vport_info *vinfo, + u8 q_idx) +{ + netif_stop_subqueue(vinfo->netdev, q_idx); + if (!hfi1_vnic_sdma_write_avail(vinfo, q_idx)) + return; + + netif_start_subqueue(vinfo->netdev, q_idx); +} + +static netdev_tx_t hfi1_netdev_start_xmit(struct sk_buff *skb, + struct net_device *netdev) +{ + struct hfi1_vnic_vport_info *vinfo = opa_vnic_dev_priv(netdev); + u8 pad_len, q_idx = skb->queue_mapping; + struct hfi1_devdata *dd = vinfo->dd; + struct opa_vnic_skb_mdata *mdata; + u32 pkt_len, total_len; + int err = -EINVAL; + u64 pbc; + + v_dbg("xmit: queue %d skb len %d\n", q_idx, skb->len); + if (unlikely(!netif_oper_up(netdev))) { + vinfo->stats[q_idx].tx_drop_state++; + goto tx_finish; + } + + /* take out meta data */ + mdata = (struct opa_vnic_skb_mdata *)skb->data; + skb_pull(skb, sizeof(*mdata)); + if (unlikely(mdata->flags & OPA_VNIC_SKB_MDATA_ENCAP_ERR)) { + vinfo->stats[q_idx].tx_dlid_zero++; + goto tx_finish; + } + + /* add tail padding (for 8 bytes size alignment) and icrc */ + pad_len = -(skb->len + OPA_VNIC_ICRC_TAIL_LEN) & 0x7; + pad_len += OPA_VNIC_ICRC_TAIL_LEN; + + /* + * pkt_len is how much data we have to write, includes header and data. + * total_len is length of the packet in Dwords plus the PBC should not + * include the CRC. + */ + pkt_len = (skb->len + pad_len) >> 2; + total_len = pkt_len + 2; /* PBC + packet */ + + pbc = create_bypass_pbc(mdata->vl, total_len); + + skb_get(skb); + v_dbg("pbc 0x%016llX len %d pad_len %d\n", pbc, skb->len, pad_len); + err = dd->process_vnic_dma_send(dd, q_idx, vinfo, skb, pbc, pad_len); + if (unlikely(err)) { + if (err == -ENOMEM) + vinfo->stats[q_idx].netstats.tx_fifo_errors++; + else if (err != -EBUSY) + vinfo->stats[q_idx].netstats.tx_carrier_errors++; + } + /* remove the header before updating tx counters */ + skb_pull(skb, OPA_VNIC_HDR_LEN); + + if (unlikely(err == -EBUSY)) { + hfi1_vnic_maybe_stop_tx(vinfo, q_idx); + dev_kfree_skb_any(skb); + return NETDEV_TX_BUSY; + } + +tx_finish: + /* update tx counters */ + hfi1_vnic_update_tx_counters(vinfo, q_idx, skb, err); + dev_kfree_skb_any(skb); + return NETDEV_TX_OK; +} + +static u16 hfi1_vnic_select_queue(struct net_device *netdev, + struct sk_buff *skb, + void *accel_priv, + select_queue_fallback_t fallback) +{ + struct hfi1_vnic_vport_info *vinfo = opa_vnic_dev_priv(netdev); + struct opa_vnic_skb_mdata *mdata; + struct sdma_engine *sde; + + mdata = (struct opa_vnic_skb_mdata *)skb->data; + sde = sdma_select_engine_vl(vinfo->dd, mdata->entropy, mdata->vl); + return sde->this_idx; +} + +/* hfi1_vnic_decap_skb - strip OPA header from the skb (ethernet) packet */ +static inline int hfi1_vnic_decap_skb(struct hfi1_vnic_rx_queue *rxq, + struct sk_buff *skb) +{ + struct hfi1_vnic_vport_info *vinfo = rxq->vinfo; + int max_len = vinfo->netdev->mtu + VLAN_ETH_HLEN; + int rc = -EFAULT; + + skb_pull(skb, OPA_VNIC_HDR_LEN); + + /* Validate Packet length */ + if (unlikely(skb->len > max_len)) + vinfo->stats[rxq->idx].rx_oversize++; + else if (unlikely(skb->len < ETH_ZLEN)) + vinfo->stats[rxq->idx].rx_runt++; + else + rc = 0; + return rc; +} + +static inline struct sk_buff *hfi1_vnic_get_skb(struct hfi1_vnic_rx_queue *rxq) +{ + unsigned char *pad_info; + struct sk_buff *skb; + + skb = skb_dequeue(&rxq->skbq); + if (unlikely(!skb)) + return NULL; + + /* remove tail padding and icrc */ + pad_info = skb->data + skb->len - 1; + skb_trim(skb, (skb->len - OPA_VNIC_ICRC_TAIL_LEN - + ((*pad_info) & 0x7))); + + return skb; +} + +/* hfi1_vnic_handle_rx - handle skb receive */ +static void hfi1_vnic_handle_rx(struct hfi1_vnic_rx_queue *rxq, + int *work_done, int work_to_do) +{ + struct hfi1_vnic_vport_info *vinfo = rxq->vinfo; + struct sk_buff *skb; + int rc; + + while (1) { + if (*work_done >= work_to_do) + break; + + skb = hfi1_vnic_get_skb(rxq); + if (unlikely(!skb)) + break; + + rc = hfi1_vnic_decap_skb(rxq, skb); + /* update rx counters */ + hfi1_vnic_update_rx_counters(vinfo, rxq->idx, skb, rc); + if (unlikely(rc)) { + dev_kfree_skb_any(skb); + continue; + } + + skb_checksum_none_assert(skb); + skb->protocol = eth_type_trans(skb, rxq->netdev); + + napi_gro_receive(&rxq->napi, skb); + (*work_done)++; + } +} + +/* hfi1_vnic_napi - napi receive polling callback function */ +static int hfi1_vnic_napi(struct napi_struct *napi, int budget) +{ + struct hfi1_vnic_rx_queue *rxq = container_of(napi, + struct hfi1_vnic_rx_queue, napi); + struct hfi1_vnic_vport_info *vinfo = rxq->vinfo; + int work_done = 0; + + v_dbg("napi %d budget %d\n", rxq->idx, budget); + hfi1_vnic_handle_rx(rxq, &work_done, budget); + + v_dbg("napi %d work_done %d\n", rxq->idx, work_done); + if (work_done < budget) + napi_complete(napi); + + return work_done; +} + +void hfi1_vnic_bypass_rcv(struct hfi1_packet *packet) +{ + struct hfi1_devdata *dd = packet->rcd->dd; + struct hfi1_vnic_vport_info *vinfo = NULL; + struct hfi1_vnic_rx_queue *rxq; + struct sk_buff *skb; + int l4_type, vesw_id = -1; + u8 q_idx; + + l4_type = HFI1_GET_L4_TYPE(packet->ebuf); + if (likely(l4_type == OPA_VNIC_L4_ETHR)) { + vesw_id = HFI1_VNIC_GET_VESWID(packet->ebuf); + vinfo = idr_find(&dd->vnic.vesw_idr, vesw_id); + + /* + * In case of invalid vesw id, count the error on + * the first available vport. + */ + if (unlikely(!vinfo)) { + struct hfi1_vnic_vport_info *vinfo_tmp; + int id_tmp = 0; + + vinfo_tmp = idr_get_next(&dd->vnic.vesw_idr, &id_tmp); + if (vinfo_tmp) { + spin_lock(&vport_cntr_lock); + vinfo_tmp->stats[0].netstats.rx_nohandler++; + spin_unlock(&vport_cntr_lock); + } + } + } + + if (unlikely(!vinfo)) { + dd_dev_warn(dd, "vnic rcv err: l4 %d vesw id %d ctx %d\n", + l4_type, vesw_id, packet->rcd->ctxt); + return; + } + + q_idx = packet->rcd->vnic_q_idx; + rxq = &vinfo->rxq[q_idx]; + if (unlikely(!netif_oper_up(vinfo->netdev))) { + vinfo->stats[q_idx].rx_drop_state++; + skb_queue_purge(&rxq->skbq); + return; + } + + if (unlikely(skb_queue_len(&rxq->skbq) > HFI1_VNIC_RCV_Q_SIZE)) { + vinfo->stats[q_idx].netstats.rx_fifo_errors++; + return; + } + + skb = netdev_alloc_skb(vinfo->netdev, packet->tlen); + if (unlikely(!skb)) { + vinfo->stats[q_idx].netstats.rx_fifo_errors++; + return; + } + + memcpy(skb->data, packet->ebuf, packet->tlen); + skb_put(skb, packet->tlen); + skb_queue_tail(&rxq->skbq, skb); + + if (napi_schedule_prep(&rxq->napi)) { + v_dbg("napi %d scheduling\n", q_idx); + __napi_schedule(&rxq->napi); + } +} + +static int hfi1_vnic_up(struct hfi1_vnic_vport_info *vinfo) +{ + struct hfi1_devdata *dd = vinfo->dd; + struct net_device *netdev = vinfo->netdev; + int i, rc; + + /* ensure virtual eth switch id is valid */ + if (!vinfo->vesw_id) + return -EINVAL; + + rc = idr_alloc(&dd->vnic.vesw_idr, vinfo, vinfo->vesw_id, + vinfo->vesw_id + 1, GFP_NOWAIT); + if (rc < 0) + return rc; + + for (i = 0; i < vinfo->num_rx_q; i++) { + struct hfi1_vnic_rx_queue *rxq = &vinfo->rxq[i]; + + skb_queue_head_init(&rxq->skbq); + napi_enable(&rxq->napi); + } + + netif_carrier_on(netdev); + netif_tx_start_all_queues(netdev); + set_bit(HFI1_VNIC_UP, &vinfo->flags); + + return 0; +} + +static void hfi1_vnic_down(struct hfi1_vnic_vport_info *vinfo) +{ + struct hfi1_devdata *dd = vinfo->dd; + u8 i; + + clear_bit(HFI1_VNIC_UP, &vinfo->flags); + netif_carrier_off(vinfo->netdev); + netif_tx_disable(vinfo->netdev); + idr_remove(&dd->vnic.vesw_idr, vinfo->vesw_id); + + /* ensure irqs see the change */ + hfi1_vnic_synchronize_irq(dd); + + /* remove unread skbs */ + for (i = 0; i < vinfo->num_rx_q; i++) { + struct hfi1_vnic_rx_queue *rxq = &vinfo->rxq[i]; + + napi_disable(&rxq->napi); + skb_queue_purge(&rxq->skbq); + } +} + +static int hfi1_netdev_open(struct net_device *netdev) +{ + struct hfi1_vnic_vport_info *vinfo = opa_vnic_dev_priv(netdev); + int rc; + + mutex_lock(&vinfo->lock); + rc = hfi1_vnic_up(vinfo); + mutex_unlock(&vinfo->lock); + return rc; +} + +static int hfi1_netdev_close(struct net_device *netdev) +{ + struct hfi1_vnic_vport_info *vinfo = opa_vnic_dev_priv(netdev); + + mutex_lock(&vinfo->lock); + if (test_bit(HFI1_VNIC_UP, &vinfo->flags)) + hfi1_vnic_down(vinfo); + mutex_unlock(&vinfo->lock); + return 0; +} + +static int hfi1_vnic_allot_ctxt(struct hfi1_devdata *dd, + struct hfi1_ctxtdata **vnic_ctxt) +{ + int rc; + + rc = allocate_vnic_ctxt(dd, vnic_ctxt); + if (rc) { + dd_dev_err(dd, "vnic ctxt alloc failed %d\n", rc); + return rc; + } + + rc = setup_vnic_ctxt(dd, *vnic_ctxt); + if (rc) { + dd_dev_err(dd, "vnic ctxt setup failed %d\n", rc); + deallocate_vnic_ctxt(dd, *vnic_ctxt); + *vnic_ctxt = NULL; + } + + return rc; +} + +static int hfi1_vnic_init(struct hfi1_vnic_vport_info *vinfo) +{ + struct hfi1_devdata *dd = vinfo->dd; + int i, rc = 0; + + mutex_lock(&hfi1_mutex); + if (!dd->vnic.num_vports) { + rc = hfi1_vnic_txreq_init(dd); + if (rc) + goto txreq_fail; + + dd->vnic.msix_idx = dd->first_dyn_msix_idx; + } + + for (i = dd->vnic.num_ctxt; i < vinfo->num_rx_q; i++) { + rc = hfi1_vnic_allot_ctxt(dd, &dd->vnic.ctxt[i]); + if (rc) + break; + dd->vnic.ctxt[i]->vnic_q_idx = i; + } + + if (i < vinfo->num_rx_q) { + /* + * If required amount of contexts is not + * allocated successfully then remaining contexts + * are released. + */ + while (i-- > dd->vnic.num_ctxt) { + deallocate_vnic_ctxt(dd, dd->vnic.ctxt[i]); + dd->vnic.ctxt[i] = NULL; + } + goto alloc_fail; + } + + if (dd->vnic.num_ctxt != i) { + dd->vnic.num_ctxt = i; + hfi1_init_vnic_rsm(dd); + } + + dd->vnic.num_vports++; + hfi1_vnic_sdma_init(vinfo); +alloc_fail: + if (!dd->vnic.num_vports) + hfi1_vnic_txreq_deinit(dd); +txreq_fail: + mutex_unlock(&hfi1_mutex); + return rc; +} + +static void hfi1_vnic_deinit(struct hfi1_vnic_vport_info *vinfo) +{ + struct hfi1_devdata *dd = vinfo->dd; + int i; + + mutex_lock(&hfi1_mutex); + if (--dd->vnic.num_vports == 0) { + for (i = 0; i < dd->vnic.num_ctxt; i++) { + deallocate_vnic_ctxt(dd, dd->vnic.ctxt[i]); + dd->vnic.ctxt[i] = NULL; + } + hfi1_deinit_vnic_rsm(dd); + dd->vnic.num_ctxt = 0; + hfi1_vnic_txreq_deinit(dd); + } + mutex_unlock(&hfi1_mutex); +} + +static void hfi1_vnic_set_vesw_id(struct net_device *netdev, int id) +{ + struct hfi1_vnic_vport_info *vinfo = opa_vnic_dev_priv(netdev); + bool reopen = false; + + /* + * If vesw_id is being changed, and if the vnic port is up, + * reset the vnic port to ensure new vesw_id gets picked up + */ + if (id != vinfo->vesw_id) { + mutex_lock(&vinfo->lock); + if (test_bit(HFI1_VNIC_UP, &vinfo->flags)) { + hfi1_vnic_down(vinfo); + reopen = true; + } + + vinfo->vesw_id = id; + if (reopen) + hfi1_vnic_up(vinfo); + + mutex_unlock(&vinfo->lock); + } +} + +/* netdev ops */ +static const struct net_device_ops hfi1_netdev_ops = { + .ndo_open = hfi1_netdev_open, + .ndo_stop = hfi1_netdev_close, + .ndo_start_xmit = hfi1_netdev_start_xmit, + .ndo_select_queue = hfi1_vnic_select_queue, + .ndo_get_stats64 = hfi1_vnic_get_stats64, +}; + +static void hfi1_vnic_free_rn(struct net_device *netdev) +{ + struct hfi1_vnic_vport_info *vinfo = opa_vnic_dev_priv(netdev); + + hfi1_vnic_deinit(vinfo); + mutex_destroy(&vinfo->lock); + free_netdev(netdev); +} + +struct net_device *hfi1_vnic_alloc_rn(struct ib_device *device, + u8 port_num, + enum rdma_netdev_t type, + const char *name, + unsigned char name_assign_type, + void (*setup)(struct net_device *)) +{ + struct hfi1_devdata *dd = dd_from_ibdev(device); + struct hfi1_vnic_vport_info *vinfo; + struct net_device *netdev; + struct rdma_netdev *rn; + int i, size, rc; + + if (!port_num || (port_num > dd->num_pports)) + return ERR_PTR(-EINVAL); + + if (type != RDMA_NETDEV_OPA_VNIC) + return ERR_PTR(-EOPNOTSUPP); + + size = sizeof(struct opa_vnic_rdma_netdev) + sizeof(*vinfo); + netdev = alloc_netdev_mqs(size, name, name_assign_type, setup, + dd->chip_sdma_engines, HFI1_NUM_VNIC_CTXT); + if (!netdev) + return ERR_PTR(-ENOMEM); + + rn = netdev_priv(netdev); + vinfo = opa_vnic_dev_priv(netdev); + vinfo->dd = dd; + vinfo->num_tx_q = dd->chip_sdma_engines; + vinfo->num_rx_q = HFI1_NUM_VNIC_CTXT; + vinfo->netdev = netdev; + rn->free_rdma_netdev = hfi1_vnic_free_rn; + rn->set_id = hfi1_vnic_set_vesw_id; + + netdev->features = NETIF_F_HIGHDMA | NETIF_F_SG; + netdev->hw_features = netdev->features; + netdev->vlan_features = netdev->features; + netdev->watchdog_timeo = msecs_to_jiffies(HFI_TX_TIMEOUT_MS); + netdev->netdev_ops = &hfi1_netdev_ops; + mutex_init(&vinfo->lock); + + for (i = 0; i < vinfo->num_rx_q; i++) { + struct hfi1_vnic_rx_queue *rxq = &vinfo->rxq[i]; + + rxq->idx = i; + rxq->vinfo = vinfo; + rxq->netdev = netdev; + netif_napi_add(netdev, &rxq->napi, hfi1_vnic_napi, 64); + } + + rc = hfi1_vnic_init(vinfo); + if (rc) + goto init_fail; + + return netdev; +init_fail: + mutex_destroy(&vinfo->lock); + free_netdev(netdev); + return ERR_PTR(rc); +} diff --git a/drivers/infiniband/hw/hfi1/vnic_sdma.c b/drivers/infiniband/hw/hfi1/vnic_sdma.c new file mode 100644 index 000000000000..51a817d3aa14 --- /dev/null +++ b/drivers/infiniband/hw/hfi1/vnic_sdma.c @@ -0,0 +1,323 @@ +/* + * Copyright(c) 2017 Intel Corporation. + * + * This file is provided under a dual BSD/GPLv2 license. When using or + * redistributing this file, you may do so under either license. + * + * GPL LICENSE SUMMARY + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * BSD LICENSE + * + * 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. + * - Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + */ + +/* + * This file contains HFI1 support for VNIC SDMA functionality + */ + +#include "sdma.h" +#include "vnic.h" + +#define HFI1_VNIC_SDMA_Q_ACTIVE BIT(0) +#define HFI1_VNIC_SDMA_Q_DEFERRED BIT(1) + +#define HFI1_VNIC_TXREQ_NAME_LEN 32 +#define HFI1_VNIC_SDMA_DESC_WTRMRK 64 +#define HFI1_VNIC_SDMA_RETRY_COUNT 1 + +/* + * struct vnic_txreq - VNIC transmit descriptor + * @txreq: sdma transmit request + * @sdma: vnic sdma pointer + * @skb: skb to send + * @pad: pad buffer + * @plen: pad length + * @pbc_val: pbc value + * @retry_count: tx retry count + */ +struct vnic_txreq { + struct sdma_txreq txreq; + struct hfi1_vnic_sdma *sdma; + + struct sk_buff *skb; + unsigned char pad[HFI1_VNIC_MAX_PAD]; + u16 plen; + __le64 pbc_val; + + u32 retry_count; +}; + +static void vnic_sdma_complete(struct sdma_txreq *txreq, + int status) +{ + struct vnic_txreq *tx = container_of(txreq, struct vnic_txreq, txreq); + struct hfi1_vnic_sdma *vnic_sdma = tx->sdma; + + sdma_txclean(vnic_sdma->dd, txreq); + dev_kfree_skb_any(tx->skb); + kmem_cache_free(vnic_sdma->dd->vnic.txreq_cache, tx); +} + +static noinline int build_vnic_ulp_payload(struct sdma_engine *sde, + struct vnic_txreq *tx) +{ + int i, ret = 0; + + ret = sdma_txadd_kvaddr( + sde->dd, + &tx->txreq, + tx->skb->data, + skb_headlen(tx->skb)); + if (unlikely(ret)) + goto bail_txadd; + + for (i = 0; i < skb_shinfo(tx->skb)->nr_frags; i++) { + struct skb_frag_struct *frag = &skb_shinfo(tx->skb)->frags[i]; + + /* combine physically continuous fragments later? */ + ret = sdma_txadd_page(sde->dd, + &tx->txreq, + skb_frag_page(frag), + frag->page_offset, + skb_frag_size(frag)); + if (unlikely(ret)) + goto bail_txadd; + } + + if (tx->plen) + ret = sdma_txadd_kvaddr(sde->dd, &tx->txreq, + tx->pad + HFI1_VNIC_MAX_PAD - tx->plen, + tx->plen); + +bail_txadd: + return ret; +} + +static int build_vnic_tx_desc(struct sdma_engine *sde, + struct vnic_txreq *tx, + u64 pbc) +{ + int ret = 0; + u16 hdrbytes = 2 << 2; /* PBC */ + + ret = sdma_txinit_ahg( + &tx->txreq, + 0, + hdrbytes + tx->skb->len + tx->plen, + 0, + 0, + NULL, + 0, + vnic_sdma_complete); + if (unlikely(ret)) + goto bail_txadd; + + /* add pbc */ + tx->pbc_val = cpu_to_le64(pbc); + ret = sdma_txadd_kvaddr( + sde->dd, + &tx->txreq, + &tx->pbc_val, + hdrbytes); + if (unlikely(ret)) + goto bail_txadd; + + /* add the ulp payload */ + ret = build_vnic_ulp_payload(sde, tx); +bail_txadd: + return ret; +} + +/* setup the last plen bypes of pad */ +static inline void hfi1_vnic_update_pad(unsigned char *pad, u8 plen) +{ + pad[HFI1_VNIC_MAX_PAD - 1] = plen - OPA_VNIC_ICRC_TAIL_LEN; +} + +int hfi1_vnic_send_dma(struct hfi1_devdata *dd, u8 q_idx, + struct hfi1_vnic_vport_info *vinfo, + struct sk_buff *skb, u64 pbc, u8 plen) +{ + struct hfi1_vnic_sdma *vnic_sdma = &vinfo->sdma[q_idx]; + struct sdma_engine *sde = vnic_sdma->sde; + struct vnic_txreq *tx; + int ret = -ECOMM; + + if (unlikely(READ_ONCE(vnic_sdma->state) != HFI1_VNIC_SDMA_Q_ACTIVE)) + goto tx_err; + + if (unlikely(!sde || !sdma_running(sde))) + goto tx_err; + + tx = kmem_cache_alloc(dd->vnic.txreq_cache, GFP_ATOMIC); + if (unlikely(!tx)) { + ret = -ENOMEM; + goto tx_err; + } + + tx->sdma = vnic_sdma; + tx->skb = skb; + hfi1_vnic_update_pad(tx->pad, plen); + tx->plen = plen; + ret = build_vnic_tx_desc(sde, tx, pbc); + if (unlikely(ret)) + goto free_desc; + tx->retry_count = 0; + + ret = sdma_send_txreq(sde, &vnic_sdma->wait, &tx->txreq); + /* When -ECOMM, sdma callback will be called with ABORT status */ + if (unlikely(ret && unlikely(ret != -ECOMM))) + goto free_desc; + + return ret; + +free_desc: + sdma_txclean(dd, &tx->txreq); + kmem_cache_free(dd->vnic.txreq_cache, tx); +tx_err: + if (ret != -EBUSY) + dev_kfree_skb_any(skb); + return ret; +} + +/* + * hfi1_vnic_sdma_sleep - vnic sdma sleep function + * + * This function gets called from sdma_send_txreq() when there are not enough + * sdma descriptors available to send the packet. It adds Tx queue's wait + * structure to sdma engine's dmawait list to be woken up when descriptors + * become available. + */ +static int hfi1_vnic_sdma_sleep(struct sdma_engine *sde, + struct iowait *wait, + struct sdma_txreq *txreq, + unsigned int seq) +{ + struct hfi1_vnic_sdma *vnic_sdma = + container_of(wait, struct hfi1_vnic_sdma, wait); + struct hfi1_ibdev *dev = &vnic_sdma->dd->verbs_dev; + struct vnic_txreq *tx = container_of(txreq, struct vnic_txreq, txreq); + + if (sdma_progress(sde, seq, txreq)) + if (tx->retry_count++ < HFI1_VNIC_SDMA_RETRY_COUNT) + return -EAGAIN; + + vnic_sdma->state = HFI1_VNIC_SDMA_Q_DEFERRED; + write_seqlock(&dev->iowait_lock); + if (list_empty(&vnic_sdma->wait.list)) + list_add_tail(&vnic_sdma->wait.list, &sde->dmawait); + write_sequnlock(&dev->iowait_lock); + return -EBUSY; +} + +/* + * hfi1_vnic_sdma_wakeup - vnic sdma wakeup function + * + * This function gets called when SDMA descriptors becomes available and Tx + * queue's wait structure was previously added to sdma engine's dmawait list. + * It notifies the upper driver about Tx queue wakeup. + */ +static void hfi1_vnic_sdma_wakeup(struct iowait *wait, int reason) +{ + struct hfi1_vnic_sdma *vnic_sdma = + container_of(wait, struct hfi1_vnic_sdma, wait); + struct hfi1_vnic_vport_info *vinfo = vnic_sdma->vinfo; + + vnic_sdma->state = HFI1_VNIC_SDMA_Q_ACTIVE; + if (__netif_subqueue_stopped(vinfo->netdev, vnic_sdma->q_idx)) + netif_wake_subqueue(vinfo->netdev, vnic_sdma->q_idx); +}; + +inline bool hfi1_vnic_sdma_write_avail(struct hfi1_vnic_vport_info *vinfo, + u8 q_idx) +{ + struct hfi1_vnic_sdma *vnic_sdma = &vinfo->sdma[q_idx]; + + return (READ_ONCE(vnic_sdma->state) == HFI1_VNIC_SDMA_Q_ACTIVE); +} + +void hfi1_vnic_sdma_init(struct hfi1_vnic_vport_info *vinfo) +{ + int i; + + for (i = 0; i < vinfo->num_tx_q; i++) { + struct hfi1_vnic_sdma *vnic_sdma = &vinfo->sdma[i]; + + iowait_init(&vnic_sdma->wait, 0, NULL, hfi1_vnic_sdma_sleep, + hfi1_vnic_sdma_wakeup, NULL); + vnic_sdma->sde = &vinfo->dd->per_sdma[i]; + vnic_sdma->dd = vinfo->dd; + vnic_sdma->vinfo = vinfo; + vnic_sdma->q_idx = i; + vnic_sdma->state = HFI1_VNIC_SDMA_Q_ACTIVE; + + /* Add a free descriptor watermark for wakeups */ + if (vnic_sdma->sde->descq_cnt > HFI1_VNIC_SDMA_DESC_WTRMRK) { + INIT_LIST_HEAD(&vnic_sdma->stx.list); + vnic_sdma->stx.num_desc = HFI1_VNIC_SDMA_DESC_WTRMRK; + list_add_tail(&vnic_sdma->stx.list, + &vnic_sdma->wait.tx_head); + } + } +} + +static void hfi1_vnic_txreq_kmem_cache_ctor(void *obj) +{ + struct vnic_txreq *tx = (struct vnic_txreq *)obj; + + memset(tx, 0, sizeof(*tx)); +} + +int hfi1_vnic_txreq_init(struct hfi1_devdata *dd) +{ + char buf[HFI1_VNIC_TXREQ_NAME_LEN]; + + snprintf(buf, sizeof(buf), "hfi1_%u_vnic_txreq_cache", dd->unit); + dd->vnic.txreq_cache = kmem_cache_create(buf, + sizeof(struct vnic_txreq), + 0, SLAB_HWCACHE_ALIGN, + hfi1_vnic_txreq_kmem_cache_ctor); + if (!dd->vnic.txreq_cache) + return -ENOMEM; + return 0; +} + +void hfi1_vnic_txreq_deinit(struct hfi1_devdata *dd) +{ + kmem_cache_destroy(dd->vnic.txreq_cache); + dd->vnic.txreq_cache = NULL; +} diff --git a/drivers/infiniband/hw/hns/hns_roce_ah.c b/drivers/infiniband/hw/hns/hns_roce_ah.c index 0ac294db3b29..f78a733a63ec 100644 --- a/drivers/infiniband/hw/hns/hns_roce_ah.c +++ b/drivers/infiniband/hw/hns/hns_roce_ah.c @@ -39,7 +39,8 @@ #define HNS_ROCE_VLAN_SL_BIT_MASK 7 #define HNS_ROCE_VLAN_SL_SHIFT 13 -struct ib_ah *hns_roce_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *ah_attr, +struct ib_ah *hns_roce_create_ah(struct ib_pd *ibpd, + struct rdma_ah_attr *ah_attr, struct ib_udata *udata) { struct hns_roce_dev *hr_dev = to_hr_dev(ibpd->device); @@ -48,6 +49,7 @@ struct ib_ah *hns_roce_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *ah_attr, struct hns_roce_ah *ah; u16 vlan_tag = 0xffff; struct in6_addr in6; + const struct ib_global_route *grh = rdma_ah_read_grh(ah_attr); union ib_gid sgid; int ret; @@ -56,15 +58,20 @@ struct ib_ah *hns_roce_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *ah_attr, return ERR_PTR(-ENOMEM); /* Get mac address */ - memcpy(&in6, ah_attr->grh.dgid.raw, sizeof(ah_attr->grh.dgid.raw)); - if (rdma_is_multicast_addr(&in6)) + memcpy(&in6, grh->dgid.raw, sizeof(grh->dgid.raw)); + if (rdma_is_multicast_addr(&in6)) { rdma_get_mcast_mac(&in6, ah->av.mac); - else - memcpy(ah->av.mac, ah_attr->dmac, sizeof(ah_attr->dmac)); + } else { + u8 *dmac = rdma_ah_retrieve_dmac(ah_attr); + + if (!dmac) + return ERR_PTR(-EINVAL); + memcpy(ah->av.mac, dmac, ETH_ALEN); + } /* Get source gid */ - ret = ib_get_cached_gid(ibpd->device, ah_attr->port_num, - ah_attr->grh.sgid_index, &sgid, &gid_attr); + ret = ib_get_cached_gid(ibpd->device, rdma_ah_get_port_num(ah_attr), + grh->sgid_index, &sgid, &gid_attr); if (ret) { dev_err(dev, "get sgid failed! ret = %d\n", ret); kfree(ah); @@ -78,45 +85,46 @@ struct ib_ah *hns_roce_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *ah_attr, } if (vlan_tag < 0x1000) - vlan_tag |= (ah_attr->sl & HNS_ROCE_VLAN_SL_BIT_MASK) << + vlan_tag |= (rdma_ah_get_sl(ah_attr) & + HNS_ROCE_VLAN_SL_BIT_MASK) << HNS_ROCE_VLAN_SL_SHIFT; - ah->av.port_pd = cpu_to_be32(to_hr_pd(ibpd)->pdn | (ah_attr->port_num << + ah->av.port_pd = cpu_to_be32(to_hr_pd(ibpd)->pdn | + (rdma_ah_get_port_num(ah_attr) << HNS_ROCE_PORT_NUM_SHIFT)); - ah->av.gid_index = ah_attr->grh.sgid_index; + ah->av.gid_index = grh->sgid_index; ah->av.vlan = cpu_to_le16(vlan_tag); dev_dbg(dev, "gid_index = 0x%x,vlan = 0x%x\n", ah->av.gid_index, ah->av.vlan); - if (ah_attr->static_rate) + if (rdma_ah_get_static_rate(ah_attr)) ah->av.stat_rate = IB_RATE_10_GBPS; - memcpy(ah->av.dgid, ah_attr->grh.dgid.raw, HNS_ROCE_GID_SIZE); - ah->av.sl_tclass_flowlabel = cpu_to_le32(ah_attr->sl << + memcpy(ah->av.dgid, grh->dgid.raw, HNS_ROCE_GID_SIZE); + ah->av.sl_tclass_flowlabel = cpu_to_le32(rdma_ah_get_sl(ah_attr) << HNS_ROCE_SL_SHIFT); return &ah->ibah; } -int hns_roce_query_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr) +int hns_roce_query_ah(struct ib_ah *ibah, struct rdma_ah_attr *ah_attr) { struct hns_roce_ah *ah = to_hr_ah(ibah); memset(ah_attr, 0, sizeof(*ah_attr)); - ah_attr->sl = le32_to_cpu(ah->av.sl_tclass_flowlabel) >> - HNS_ROCE_SL_SHIFT; - ah_attr->port_num = le32_to_cpu(ah->av.port_pd) >> - HNS_ROCE_PORT_NUM_SHIFT; - ah_attr->static_rate = ah->av.stat_rate; - ah_attr->ah_flags = IB_AH_GRH; - ah_attr->grh.traffic_class = le32_to_cpu(ah->av.sl_tclass_flowlabel) >> - HNS_ROCE_TCLASS_SHIFT; - ah_attr->grh.flow_label = le32_to_cpu(ah->av.sl_tclass_flowlabel) & - HNS_ROCE_FLOW_LABLE_MASK; - ah_attr->grh.hop_limit = ah->av.hop_limit; - ah_attr->grh.sgid_index = ah->av.gid_index; - memcpy(ah_attr->grh.dgid.raw, ah->av.dgid, HNS_ROCE_GID_SIZE); + rdma_ah_set_sl(ah_attr, (le32_to_cpu(ah->av.sl_tclass_flowlabel) >> + HNS_ROCE_SL_SHIFT)); + rdma_ah_set_port_num(ah_attr, (le32_to_cpu(ah->av.port_pd) >> + HNS_ROCE_PORT_NUM_SHIFT)); + rdma_ah_set_static_rate(ah_attr, ah->av.stat_rate); + rdma_ah_set_grh(ah_attr, NULL, + (le32_to_cpu(ah->av.sl_tclass_flowlabel) & + HNS_ROCE_FLOW_LABLE_MASK), ah->av.gid_index, + ah->av.hop_limit, + (le32_to_cpu(ah->av.sl_tclass_flowlabel) >> + HNS_ROCE_TCLASS_SHIFT)); + rdma_ah_set_dgid_raw(ah_attr, ah->av.dgid); return 0; } diff --git a/drivers/infiniband/hw/hns/hns_roce_cmd.c b/drivers/infiniband/hw/hns/hns_roce_cmd.c index 8c1f7a6f84d2..b94dcd823ad1 100644 --- a/drivers/infiniband/hw/hns/hns_roce_cmd.c +++ b/drivers/infiniband/hw/hns/hns_roce_cmd.c @@ -299,9 +299,9 @@ int hns_roce_cmd_use_events(struct hns_roce_dev *hr_dev) struct hns_roce_cmdq *hr_cmd = &hr_dev->cmd; int i; - hr_cmd->context = kmalloc(hr_cmd->max_cmds * - sizeof(struct hns_roce_cmd_context), - GFP_KERNEL); + hr_cmd->context = kmalloc_array(hr_cmd->max_cmds, + sizeof(*hr_cmd->context), + GFP_KERNEL); if (!hr_cmd->context) return -ENOMEM; diff --git a/drivers/infiniband/hw/hns/hns_roce_cq.c b/drivers/infiniband/hw/hns/hns_roce_cq.c index 589496c8fb9e..b89fd711019e 100644 --- a/drivers/infiniband/hw/hns/hns_roce_cq.c +++ b/drivers/infiniband/hw/hns/hns_roce_cq.c @@ -219,8 +219,7 @@ static int hns_roce_ib_get_cq_umem(struct hns_roce_dev *hr_dev, return PTR_ERR(*umem); ret = hns_roce_mtt_init(hr_dev, ib_umem_page_count(*umem), - ilog2((unsigned int)(*umem)->page_size), - &buf->hr_mtt); + (*umem)->page_shift, &buf->hr_mtt); if (ret) goto err_buf; diff --git a/drivers/infiniband/hw/hns/hns_roce_device.h b/drivers/infiniband/hw/hns/hns_roce_device.h index 1a6cb5d7a0dd..e493a61e14e1 100644 --- a/drivers/infiniband/hw/hns/hns_roce_device.h +++ b/drivers/infiniband/hw/hns/hns_roce_device.h @@ -687,9 +687,10 @@ void hns_roce_bitmap_free_range(struct hns_roce_bitmap *bitmap, unsigned long obj, int cnt, int rr); -struct ib_ah *hns_roce_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr, +struct ib_ah *hns_roce_create_ah(struct ib_pd *pd, + struct rdma_ah_attr *ah_attr, struct ib_udata *udata); -int hns_roce_query_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr); +int hns_roce_query_ah(struct ib_ah *ibah, struct rdma_ah_attr *ah_attr); int hns_roce_destroy_ah(struct ib_ah *ah); struct ib_pd *hns_roce_alloc_pd(struct ib_device *ib_dev, diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c index b8111b0c8877..23fad6d96944 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/of.h> #include <rdma/ib_umem.h> #include "hns_roce_common.h" #include "hns_roce_device.h" @@ -227,14 +228,14 @@ int hns_roce_v1_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, switch (wr->opcode) { case IB_WR_RDMA_READ: ps_opcode = HNS_ROCE_WQE_OPCODE_RDMA_READ; - set_raddr_seg(wqe, atomic_wr(wr)->remote_addr, - atomic_wr(wr)->rkey); + set_raddr_seg(wqe, rdma_wr(wr)->remote_addr, + rdma_wr(wr)->rkey); break; case IB_WR_RDMA_WRITE: case IB_WR_RDMA_WRITE_WITH_IMM: ps_opcode = HNS_ROCE_WQE_OPCODE_RDMA_WRITE; - set_raddr_seg(wqe, atomic_wr(wr)->remote_addr, - atomic_wr(wr)->rkey); + set_raddr_seg(wqe, rdma_wr(wr)->remote_addr, + rdma_wr(wr)->rkey); break; case IB_WR_SEND: case IB_WR_SEND_WITH_INV: @@ -657,11 +658,14 @@ static int hns_roce_v1_rsv_lp_qp(struct hns_roce_dev *hr_dev) struct hns_roce_qp *hr_qp; struct ib_cq *cq; struct ib_pd *pd; + union ib_gid dgid; u64 subnet_prefix; int attr_mask = 0; - int i; + int i, j; int ret; + u8 queue_en[HNS_ROCE_V1_RESV_QP] = { 0 }; u8 phy_port; + u8 port = 0; u8 sl; priv = (struct hns_roce_v1_priv *)hr_dev->hw->priv; @@ -707,15 +711,27 @@ static int hns_roce_v1_rsv_lp_qp(struct hns_roce_dev *hr_dev) attr.rnr_retry = 7; attr.timeout = 0x12; attr.path_mtu = IB_MTU_256; - attr.ah_attr.ah_flags = 1; - attr.ah_attr.static_rate = 3; - attr.ah_attr.grh.sgid_index = 0; - attr.ah_attr.grh.hop_limit = 1; - attr.ah_attr.grh.flow_label = 0; - attr.ah_attr.grh.traffic_class = 0; + attr.ah_attr.type = RDMA_AH_ATTR_TYPE_ROCE; + rdma_ah_set_grh(&attr.ah_attr, NULL, 0, 0, 1, 0); + rdma_ah_set_static_rate(&attr.ah_attr, 3); subnet_prefix = cpu_to_be64(0xfe80000000000000LL); for (i = 0; i < HNS_ROCE_V1_RESV_QP; i++) { + phy_port = (i >= HNS_ROCE_MAX_PORTS) ? (i - 2) : + (i % HNS_ROCE_MAX_PORTS); + sl = i / HNS_ROCE_MAX_PORTS; + + for (j = 0; j < caps->num_ports; j++) { + if (hr_dev->iboe.phy_port[j] == phy_port) { + queue_en[i] = 1; + port = j; + break; + } + } + + if (!queue_en[i]) + continue; + free_mr->mr_free_qp[i] = hns_roce_v1_create_lp_qp(hr_dev, pd); if (IS_ERR(free_mr->mr_free_qp[i])) { dev_err(dev, "Create loop qp failed!\n"); @@ -723,15 +739,7 @@ static int hns_roce_v1_rsv_lp_qp(struct hns_roce_dev *hr_dev) } hr_qp = free_mr->mr_free_qp[i]; - sl = i / caps->num_ports; - - if (caps->num_ports == HNS_ROCE_MAX_PORTS) - phy_port = (i >= HNS_ROCE_MAX_PORTS) ? (i - 2) : - (i % caps->num_ports); - else - phy_port = i % caps->num_ports; - - hr_qp->port = phy_port + 1; + hr_qp->port = port; hr_qp->phy_port = phy_port; hr_qp->ibqp.qp_type = IB_QPT_RC; hr_qp->ibqp.device = &hr_dev->ib_dev; @@ -741,25 +749,22 @@ static int hns_roce_v1_rsv_lp_qp(struct hns_roce_dev *hr_dev) hr_qp->ibqp.recv_cq = cq; hr_qp->ibqp.send_cq = cq; - attr.ah_attr.port_num = phy_port + 1; - attr.ah_attr.sl = sl; - attr.port_num = phy_port + 1; + rdma_ah_set_port_num(&attr.ah_attr, port + 1); + rdma_ah_set_sl(&attr.ah_attr, sl); + attr.port_num = port + 1; attr.dest_qp_num = hr_qp->qpn; - memcpy(attr.ah_attr.dmac, hr_dev->dev_addr[phy_port], + memcpy(rdma_ah_retrieve_dmac(&attr.ah_attr), + hr_dev->dev_addr[port], MAC_ADDR_OCTET_NUM); - memcpy(attr.ah_attr.grh.dgid.raw, - &subnet_prefix, sizeof(u64)); - memcpy(&attr.ah_attr.grh.dgid.raw[8], - hr_dev->dev_addr[phy_port], 3); - memcpy(&attr.ah_attr.grh.dgid.raw[13], - hr_dev->dev_addr[phy_port] + 3, 3); - attr.ah_attr.grh.dgid.raw[11] = 0xff; - attr.ah_attr.grh.dgid.raw[12] = 0xfe; - attr.ah_attr.grh.dgid.raw[8] ^= 2; - - attr_mask |= IB_QP_PORT; + memcpy(&dgid.raw, &subnet_prefix, sizeof(u64)); + memcpy(&dgid.raw[8], hr_dev->dev_addr[port], 3); + memcpy(&dgid.raw[13], hr_dev->dev_addr[port] + 3, 3); + dgid.raw[11] = 0xff; + dgid.raw[12] = 0xfe; + dgid.raw[8] ^= 2; + rdma_ah_set_dgid_raw(&attr.ah_attr, dgid.raw); ret = hr_dev->hw->modify_qp(&hr_qp->ibqp, &attr, attr_mask, IB_QPS_RESET, IB_QPS_INIT); @@ -816,6 +821,9 @@ static void hns_roce_v1_release_lp_qp(struct hns_roce_dev *hr_dev) for (i = 0; i < HNS_ROCE_V1_RESV_QP; i++) { hr_qp = free_mr->mr_free_qp[i]; + if (!hr_qp) + continue; + ret = hns_roce_v1_destroy_qp(&hr_qp->ibqp); if (ret) dev_err(dev, "Destroy qp %d for mr free failed(%d)!\n", @@ -967,7 +975,7 @@ static void hns_roce_v1_mr_free_work_fn(struct work_struct *work) msecs_to_jiffies(HNS_ROCE_V1_FREE_MR_TIMEOUT_MSECS) + jiffies; int i; int ret; - int ne; + int ne = 0; mr_work = container_of(work, struct hns_roce_mr_free_work, work); hr_mr = (struct hns_roce_mr *)mr_work->mr; @@ -980,6 +988,10 @@ static void hns_roce_v1_mr_free_work_fn(struct work_struct *work) for (i = 0; i < HNS_ROCE_V1_RESV_QP; i++) { hr_qp = free_mr->mr_free_qp[i]; + if (!hr_qp) + continue; + ne++; + ret = hns_roce_v1_send_lp_wqe(hr_qp); if (ret) { dev_err(dev, @@ -989,7 +1001,6 @@ static void hns_roce_v1_mr_free_work_fn(struct work_struct *work) } } - ne = HNS_ROCE_V1_RESV_QP; do { ret = hns_roce_v1_poll_cq(&mr_free_cq->ib_cq, ne, wc); if (ret < 0) { @@ -999,7 +1010,8 @@ static void hns_roce_v1_mr_free_work_fn(struct work_struct *work) goto free_work; } ne -= ret; - msleep(HNS_ROCE_V1_FREE_MR_WAIT_VALUE); + usleep_range(HNS_ROCE_V1_FREE_MR_WAIT_VALUE * 1000, + (1 + HNS_ROCE_V1_FREE_MR_WAIT_VALUE) * 1000); } while (ne && time_before_eq(jiffies, end)); if (ne != 0) @@ -1725,7 +1737,7 @@ int hns_roce_v1_write_mtpt(void *mb_buf, struct hns_roce_mr *mr, roce_set_field(mpt_entry->mpt_byte_64, MPT_BYTE_64_L_KEY_IDX_H_M, MPT_BYTE_64_L_KEY_IDX_H_S, mtpt_idx >> MTPT_IDX_SHIFT); - /* DMA momery regsiter */ + /* DMA memory register */ if (mr->type == MR_TYPE_DMA) return 0; @@ -1851,6 +1863,7 @@ void hns_roce_v1_cq_set_ci(struct hns_roce_cq *hr_cq, u32 cons_index) u32 doorbell[2]; doorbell[0] = cons_index & ((hr_cq->cq_depth << 1) - 1); + doorbell[1] = 0; roce_set_bit(doorbell[1], ROCEE_DB_OTHERS_H_ROCEE_DB_OTH_HW_SYNS_S, 1); roce_set_field(doorbell[1], ROCEE_DB_OTHERS_H_ROCEE_DB_OTH_CMD_M, ROCEE_DB_OTHERS_H_ROCEE_DB_OTH_CMD_S, 3); @@ -2184,7 +2197,7 @@ static int hns_roce_v1_poll_one(struct hns_roce_cq *hr_cq, } wc->wr_id = wq->wrid[wq->tail & (wq->wqe_cnt - 1)]; ++wq->tail; - } else { + } else { /* RQ conrespond to CQE */ wc->byte_len = le32_to_cpu(cqe->byte_cnt); opcode = roce_get_field(cqe->cqe_byte_4, @@ -2565,6 +2578,7 @@ static int hns_roce_v1_m_qp(struct ib_qp *ibqp, const struct ib_qp_attr *attr, struct hns_roce_qp *hr_qp = to_hr_qp(ibqp); struct device *dev = &hr_dev->pdev->dev; struct hns_roce_qp_context *context; + const struct ib_global_route *grh = rdma_ah_read_grh(&attr->ah_attr); dma_addr_t dma_handle_2 = 0; dma_addr_t dma_handle = 0; uint32_t doorbell[2] = {0}; @@ -2573,6 +2587,7 @@ static int hns_roce_v1_m_qp(struct ib_qp *ibqp, const struct ib_qp_attr *attr, int ret = -EINVAL; u64 *mtts = NULL; int port; + u8 port_num; u8 *dmac; u8 *smac; @@ -2739,7 +2754,7 @@ static int hns_roce_v1_m_qp(struct ib_qp *ibqp, const struct ib_qp_attr *attr, goto out; } - dmac = (u8 *)attr->ah_attr.dmac; + dmac = (u8 *)attr->ah_attr.roce.dmac; context->sq_rq_bt_l = (u32)(dma_handle); roce_set_field(context->qpc_bytes_24, @@ -2780,7 +2795,7 @@ static int hns_roce_v1_m_qp(struct ib_qp *ibqp, const struct ib_qp_attr *attr, roce_set_bit(context->qpc_bytes_32, QP_CONTEXT_QPC_BYTE_32_GLOBAL_HEADER_S, - attr->ah_attr.ah_flags); + rdma_ah_get_ah_flags(&attr->ah_attr)); roce_set_field(context->qpc_bytes_32, QP_CONTEXT_QPC_BYTES_32_RESPONDER_RESOURCES_M, QP_CONTEXT_QPC_BYTES_32_RESPONDER_RESOURCES_S, @@ -2792,12 +2807,13 @@ static int hns_roce_v1_m_qp(struct ib_qp *ibqp, const struct ib_qp_attr *attr, attr->dest_qp_num); /* Configure GID index */ + port_num = rdma_ah_get_port_num(&attr->ah_attr); roce_set_field(context->qpc_bytes_36, QP_CONTEXT_QPC_BYTES_36_SGID_INDEX_M, QP_CONTEXT_QPC_BYTES_36_SGID_INDEX_S, - hns_get_gid_index(hr_dev, - attr->ah_attr.port_num - 1, - attr->ah_attr.grh.sgid_index)); + hns_get_gid_index(hr_dev, + port_num - 1, + grh->sgid_index)); memcpy(&(context->dmac_l), dmac, 4); @@ -2808,26 +2824,26 @@ static int hns_roce_v1_m_qp(struct ib_qp *ibqp, const struct ib_qp_attr *attr, roce_set_field(context->qpc_bytes_44, QP_CONTEXT_QPC_BYTES_44_MAXIMUM_STATIC_RATE_M, QP_CONTEXT_QPC_BYTES_44_MAXIMUM_STATIC_RATE_S, - attr->ah_attr.static_rate); + rdma_ah_get_static_rate(&attr->ah_attr)); roce_set_field(context->qpc_bytes_44, QP_CONTEXT_QPC_BYTES_44_HOPLMT_M, QP_CONTEXT_QPC_BYTES_44_HOPLMT_S, - attr->ah_attr.grh.hop_limit); + grh->hop_limit); roce_set_field(context->qpc_bytes_48, QP_CONTEXT_QPC_BYTES_48_FLOWLABEL_M, QP_CONTEXT_QPC_BYTES_48_FLOWLABEL_S, - attr->ah_attr.grh.flow_label); + grh->flow_label); roce_set_field(context->qpc_bytes_48, QP_CONTEXT_QPC_BYTES_48_TCLASS_M, QP_CONTEXT_QPC_BYTES_48_TCLASS_S, - attr->ah_attr.grh.traffic_class); + grh->traffic_class); roce_set_field(context->qpc_bytes_48, QP_CONTEXT_QPC_BYTES_48_MTU_M, QP_CONTEXT_QPC_BYTES_48_MTU_S, attr->path_mtu); - memcpy(context->dgid, attr->ah_attr.grh.dgid.raw, - sizeof(attr->ah_attr.grh.dgid.raw)); + memcpy(context->dgid, grh->dgid.raw, + sizeof(grh->dgid.raw)); dev_dbg(dev, "dmac:%x :%lx\n", context->dmac_l, roce_get_field(context->qpc_bytes_44, @@ -2907,8 +2923,9 @@ static int hns_roce_v1_m_qp(struct ib_qp *ibqp, const struct ib_qp_attr *attr, hr_qp->phy_port); roce_set_field(context->qpc_bytes_156, QP_CONTEXT_QPC_BYTES_156_SL_M, - QP_CONTEXT_QPC_BYTES_156_SL_S, attr->ah_attr.sl); - hr_qp->sl = attr->ah_attr.sl; + QP_CONTEXT_QPC_BYTES_156_SL_S, + rdma_ah_get_sl(&attr->ah_attr)); + hr_qp->sl = rdma_ah_get_sl(&attr->ah_attr); } else if (cur_state == IB_QPS_RTR && new_state == IB_QPS_RTS) { /* If exist optional param, return error */ @@ -3019,8 +3036,9 @@ static int hns_roce_v1_m_qp(struct ib_qp *ibqp, const struct ib_qp_attr *attr, hr_qp->phy_port); roce_set_field(context->qpc_bytes_156, QP_CONTEXT_QPC_BYTES_156_SL_M, - QP_CONTEXT_QPC_BYTES_156_SL_S, attr->ah_attr.sl); - hr_qp->sl = attr->ah_attr.sl; + QP_CONTEXT_QPC_BYTES_156_SL_S, + rdma_ah_get_sl(&attr->ah_attr)); + hr_qp->sl = rdma_ah_get_sl(&attr->ah_attr); roce_set_field(context->qpc_bytes_156, QP_CONTEXT_QPC_BYTES_156_INITIATOR_DEPTH_M, QP_CONTEXT_QPC_BYTES_156_INITIATOR_DEPTH_S, @@ -3355,28 +3373,33 @@ static int hns_roce_v1_q_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr, if (hr_qp->ibqp.qp_type == IB_QPT_RC || hr_qp->ibqp.qp_type == IB_QPT_UC) { - qp_attr->ah_attr.sl = roce_get_field(context->qpc_bytes_156, - QP_CONTEXT_QPC_BYTES_156_SL_M, - QP_CONTEXT_QPC_BYTES_156_SL_S); - qp_attr->ah_attr.grh.flow_label = roce_get_field( - context->qpc_bytes_48, - QP_CONTEXT_QPC_BYTES_48_FLOWLABEL_M, - QP_CONTEXT_QPC_BYTES_48_FLOWLABEL_S); - qp_attr->ah_attr.grh.sgid_index = roce_get_field( - context->qpc_bytes_36, - QP_CONTEXT_QPC_BYTES_36_SGID_INDEX_M, - QP_CONTEXT_QPC_BYTES_36_SGID_INDEX_S); - qp_attr->ah_attr.grh.hop_limit = roce_get_field( - context->qpc_bytes_44, - QP_CONTEXT_QPC_BYTES_44_HOPLMT_M, - QP_CONTEXT_QPC_BYTES_44_HOPLMT_S); - qp_attr->ah_attr.grh.traffic_class = roce_get_field( - context->qpc_bytes_48, - QP_CONTEXT_QPC_BYTES_48_TCLASS_M, - QP_CONTEXT_QPC_BYTES_48_TCLASS_S); - - memcpy(qp_attr->ah_attr.grh.dgid.raw, context->dgid, - sizeof(qp_attr->ah_attr.grh.dgid.raw)); + struct ib_global_route *grh = + rdma_ah_retrieve_grh(&qp_attr->ah_attr); + + rdma_ah_set_sl(&qp_attr->ah_attr, + roce_get_field(context->qpc_bytes_156, + QP_CONTEXT_QPC_BYTES_156_SL_M, + QP_CONTEXT_QPC_BYTES_156_SL_S)); + rdma_ah_set_ah_flags(&qp_attr->ah_attr, IB_AH_GRH); + grh->flow_label = + roce_get_field(context->qpc_bytes_48, + QP_CONTEXT_QPC_BYTES_48_FLOWLABEL_M, + QP_CONTEXT_QPC_BYTES_48_FLOWLABEL_S); + grh->sgid_index = + roce_get_field(context->qpc_bytes_36, + QP_CONTEXT_QPC_BYTES_36_SGID_INDEX_M, + QP_CONTEXT_QPC_BYTES_36_SGID_INDEX_S); + grh->hop_limit = + roce_get_field(context->qpc_bytes_44, + QP_CONTEXT_QPC_BYTES_44_HOPLMT_M, + QP_CONTEXT_QPC_BYTES_44_HOPLMT_S); + grh->traffic_class = + roce_get_field(context->qpc_bytes_48, + QP_CONTEXT_QPC_BYTES_48_TCLASS_M, + QP_CONTEXT_QPC_BYTES_48_TCLASS_S); + + memcpy(grh->dgid.raw, context->dgid, + sizeof(grh->dgid.raw)); } qp_attr->pkey_index = roce_get_field(context->qpc_bytes_12, @@ -3526,10 +3549,12 @@ static int check_qp_db_process_status(struct hns_roce_dev *hr_dev, old_cnt = roce_get_field(old_send, ROCEE_SDB_SEND_PTR_SDB_SEND_PTR_M, ROCEE_SDB_SEND_PTR_SDB_SEND_PTR_S); - if (cur_cnt - old_cnt > SDB_ST_CMP_VAL) + if (cur_cnt - old_cnt > + SDB_ST_CMP_VAL) { success_flags = 1; - else { - send_ptr = roce_get_field(old_send, + } else { + send_ptr = + roce_get_field(old_send, ROCEE_SDB_SEND_PTR_SDB_SEND_PTR_M, ROCEE_SDB_SEND_PTR_SDB_SEND_PTR_S) + roce_get_field(sdb_retry_cnt, @@ -3634,6 +3659,7 @@ static void hns_roce_v1_destroy_qp_work_fn(struct work_struct *work) struct hns_roce_dev *hr_dev; struct hns_roce_qp *hr_qp; struct device *dev; + unsigned long qpn; int ret; qp_work_entry = container_of(work, struct hns_roce_qp_work, work); @@ -3641,8 +3667,9 @@ static void hns_roce_v1_destroy_qp_work_fn(struct work_struct *work) dev = &hr_dev->pdev->dev; priv = (struct hns_roce_v1_priv *)hr_dev->hw->priv; hr_qp = qp_work_entry->qp; + qpn = hr_qp->qpn; - dev_dbg(dev, "Schedule destroy QP(0x%lx) work.\n", hr_qp->qpn); + dev_dbg(dev, "Schedule destroy QP(0x%lx) work.\n", qpn); qp_work_entry->sche_cnt++; @@ -3653,7 +3680,7 @@ static void hns_roce_v1_destroy_qp_work_fn(struct work_struct *work) &qp_work_entry->db_wait_stage); if (ret) { dev_err(dev, "Check QP(0x%lx) db process status failed!\n", - hr_qp->qpn); + qpn); return; } @@ -3667,7 +3694,7 @@ static void hns_roce_v1_destroy_qp_work_fn(struct work_struct *work) ret = hns_roce_v1_modify_qp(&hr_qp->ibqp, NULL, 0, hr_qp->state, IB_QPS_RESET); if (ret) { - dev_err(dev, "Modify QP(0x%lx) to RST failed!\n", hr_qp->qpn); + dev_err(dev, "Modify QP(0x%lx) to RST failed!\n", qpn); return; } @@ -3676,14 +3703,14 @@ static void hns_roce_v1_destroy_qp_work_fn(struct work_struct *work) if (hr_qp->ibqp.qp_type == IB_QPT_RC) { /* RC QP, release QPN */ - hns_roce_release_range_qp(hr_dev, hr_qp->qpn, 1); + hns_roce_release_range_qp(hr_dev, qpn, 1); kfree(hr_qp); } else kfree(hr_to_hr_sqp(hr_qp)); kfree(qp_work_entry); - dev_dbg(dev, "Accomplished destroy QP(0x%lx) work.\n", hr_qp->qpn); + dev_dbg(dev, "Accomplished destroy QP(0x%lx) work.\n", qpn); } int hns_roce_v1_destroy_qp(struct ib_qp *ibqp) diff --git a/drivers/infiniband/hw/hns/hns_roce_main.c b/drivers/infiniband/hw/hns/hns_roce_main.c index 4953d9cb83a7..d9777b662eba 100644 --- a/drivers/infiniband/hw/hns/hns_roce_main.c +++ b/drivers/infiniband/hw/hns/hns_roce_main.c @@ -32,6 +32,7 @@ */ #include <linux/acpi.h> #include <linux/of_platform.h> +#include <linux/module.h> #include <rdma/ib_addr.h> #include <rdma/ib_smi.h> #include <rdma/ib_user_verbs.h> @@ -124,8 +125,6 @@ static int handle_en_event(struct hns_roce_dev *hr_dev, u8 port, return -ENODEV; } - spin_lock_bh(&hr_dev->iboe.lock); - switch (event) { case NETDEV_UP: case NETDEV_CHANGE: @@ -143,7 +142,6 @@ static int handle_en_event(struct hns_roce_dev *hr_dev, u8 port, break; } - spin_unlock_bh(&hr_dev->iboe.lock); return 0; } @@ -249,7 +247,7 @@ static int hns_roce_query_port(struct ib_device *ib_dev, u8 port_num, assert(port_num > 0); port = port_num - 1; - memset(props, 0, sizeof(*props)); + /* props being zeroed by the caller, avoid zeroing it here */ props->max_mtu = hr_dev->caps.max_mtu; props->gid_tbl_len = hr_dev->caps.gid_table_len[port]; @@ -400,14 +398,15 @@ static int hns_roce_port_immutable(struct ib_device *ib_dev, u8 port_num, struct ib_port_attr attr; int ret; - ret = hns_roce_query_port(ib_dev, port_num, &attr); + immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE; + + ret = ib_query_port(ib_dev, port_num, &attr); if (ret) return ret; immutable->pkey_tbl_len = attr.pkey_tbl_len; immutable->gid_tbl_len = attr.gid_tbl_len; - immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE; immutable->max_mad_size = IB_MGMT_MAD_SIZE; return 0; @@ -437,7 +436,7 @@ static int hns_roce_register_device(struct hns_roce_dev *hr_dev) ib_dev->owner = THIS_MODULE; ib_dev->node_type = RDMA_NODE_IB_CA; - ib_dev->dma_device = dev; + ib_dev->dev.parent = dev; ib_dev->phys_port_cnt = hr_dev->caps.num_ports; ib_dev->local_dma_lkey = hr_dev->caps.reserved_lkey; diff --git a/drivers/infiniband/hw/hns/hns_roce_mr.c b/drivers/infiniband/hw/hns/hns_roce_mr.c index 4139abee3b54..80fc01ffd8bd 100644 --- a/drivers/infiniband/hw/hns/hns_roce_mr.c +++ b/drivers/infiniband/hw/hns/hns_roce_mr.c @@ -127,11 +127,12 @@ static int hns_roce_buddy_init(struct hns_roce_buddy *buddy, int max_order) buddy->max_order = max_order; spin_lock_init(&buddy->lock); - - buddy->bits = kzalloc((buddy->max_order + 1) * sizeof(long *), - GFP_KERNEL); - buddy->num_free = kzalloc((buddy->max_order + 1) * sizeof(int *), - GFP_KERNEL); + buddy->bits = kcalloc(buddy->max_order + 1, + sizeof(*buddy->bits), + GFP_KERNEL); + buddy->num_free = kcalloc(buddy->max_order + 1, + sizeof(*buddy->num_free), + GFP_KERNEL); if (!buddy->bits || !buddy->num_free) goto err_out; @@ -204,7 +205,7 @@ int hns_roce_mtt_init(struct hns_roce_dev *hr_dev, int npages, int page_shift, return 0; } - /* Note: if page_shift is zero, FAST memory regsiter */ + /* Note: if page_shift is zero, FAST memory register */ mtt->page_shift = page_shift; /* Compute MTT entry necessary */ @@ -503,7 +504,8 @@ int hns_roce_ib_umem_write_mtt(struct hns_roce_dev *hr_dev, for_each_sg(umem->sg_head.sgl, sg, umem->nmap, entry) { len = sg_dma_len(sg) >> mtt->page_shift; for (k = 0; k < len; ++k) { - pages[i++] = sg_dma_address(sg) + umem->page_size * k; + pages[i++] = sg_dma_address(sg) + + (k << umem->page_shift); if (i == PAGE_SIZE / sizeof(u64)) { ret = hns_roce_write_mtt(hr_dev, mtt, n, i, pages); @@ -563,9 +565,9 @@ struct ib_mr *hns_roce_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, } n = ib_umem_page_count(mr->umem); - if (mr->umem->page_size != HNS_ROCE_HEM_PAGE_SIZE) { - dev_err(dev, "Just support 4K page size but is 0x%x now!\n", - mr->umem->page_size); + if (mr->umem->page_shift != HNS_ROCE_HEM_PAGE_SHIFT) { + dev_err(dev, "Just support 4K page size but is 0x%lx now!\n", + BIT(mr->umem->page_shift)); ret = -EINVAL; goto err_umem; } diff --git a/drivers/infiniband/hw/hns/hns_roce_qp.c b/drivers/infiniband/hw/hns/hns_roce_qp.c index f036f32f15d3..054c52699090 100644 --- a/drivers/infiniband/hw/hns/hns_roce_qp.c +++ b/drivers/infiniband/hw/hns/hns_roce_qp.c @@ -101,7 +101,7 @@ static void hns_roce_ib_qp_event(struct hns_roce_qp *hr_qp, event.event = IB_EVENT_QP_ACCESS_ERR; break; default: - dev_dbg(ibqp->device->dma_device, "roce_ib: Unexpected event type %d on QP %06lx\n", + dev_dbg(ibqp->device->dev.parent, "roce_ib: Unexpected event type %d on QP %06lx\n", type, hr_qp->qpn); return; } @@ -437,8 +437,7 @@ static int hns_roce_create_qp_common(struct hns_roce_dev *hr_dev, } ret = hns_roce_mtt_init(hr_dev, ib_umem_page_count(hr_qp->umem), - ilog2((unsigned int)hr_qp->umem->page_size), - &hr_qp->mtt); + hr_qp->umem->page_shift, &hr_qp->mtt); if (ret) { dev_err(dev, "hns_roce_mtt_init error for create qp\n"); goto err_buf; diff --git a/drivers/infiniband/hw/i40iw/i40iw.h b/drivers/infiniband/hw/i40iw/i40iw.h index da2eb5a281fa..9b1566468744 100644 --- a/drivers/infiniband/hw/i40iw/i40iw.h +++ b/drivers/infiniband/hw/i40iw/i40iw.h @@ -527,6 +527,7 @@ enum i40iw_status_code i40iw_add_mac_addr(struct i40iw_device *iwdev, int i40iw_modify_qp(struct ib_qp *, struct ib_qp_attr *, int, struct ib_udata *); void i40iw_cq_wq_destroy(struct i40iw_device *iwdev, struct i40iw_sc_cq *cq); +void i40iw_cleanup_pending_cqp_op(struct i40iw_device *iwdev); void i40iw_rem_pdusecount(struct i40iw_pd *iwpd, struct i40iw_device *iwdev); void i40iw_add_pdusecount(struct i40iw_pd *iwpd); void i40iw_rem_devusecount(struct i40iw_device *iwdev); diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.c b/drivers/infiniband/hw/i40iw/i40iw_cm.c index 95a0586a4da8..5a2fa743676c 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_cm.c +++ b/drivers/infiniband/hw/i40iw/i40iw_cm.c @@ -784,7 +784,6 @@ static void i40iw_build_mpa_v2(struct i40iw_cm_node *cm_node, } ctrl_ird |= IETF_PEER_TO_PEER; - ctrl_ird |= IETF_FLPDU_ZERO_LEN; switch (mpa_key) { case MPA_KEY_REQUEST: @@ -2446,8 +2445,8 @@ static void i40iw_handle_rcv_mpa(struct i40iw_cm_node *cm_node, } else { type = I40IW_CM_EVENT_CONNECTED; cm_node->state = I40IW_CM_STATE_OFFLOADED; - i40iw_send_ack(cm_node); } + i40iw_send_ack(cm_node); break; default: pr_err("%s wrong cm_node state =%d\n", __func__, cm_node->state); @@ -3184,9 +3183,8 @@ void i40iw_setup_cm_core(struct i40iw_device *iwdev) INIT_LIST_HEAD(&cm_core->connected_nodes); INIT_LIST_HEAD(&cm_core->listen_nodes); - init_timer(&cm_core->tcp_timer); - cm_core->tcp_timer.function = i40iw_cm_timer_tick; - cm_core->tcp_timer.data = (unsigned long)cm_core; + setup_timer(&cm_core->tcp_timer, i40iw_cm_timer_tick, + (unsigned long)cm_core); spin_lock_init(&cm_core->ht_lock); spin_lock_init(&cm_core->listen_list_lock); @@ -3489,7 +3487,8 @@ static void i40iw_cm_disconn_true(struct i40iw_qp *iwqp) if (((original_hw_tcp_state == I40IW_TCP_STATE_CLOSED) || (original_hw_tcp_state == I40IW_TCP_STATE_TIME_WAIT) || (last_ae == I40IW_AE_RDMAP_ROE_BAD_LLP_CLOSE) || - (last_ae == I40IW_AE_LLP_CONNECTION_RESET))) { + (last_ae == I40IW_AE_LLP_CONNECTION_RESET) || + iwdev->reset)) { issue_close = 1; iwqp->cm_id = NULL; if (!iwqp->flush_issued) { @@ -4267,6 +4266,8 @@ void i40iw_cm_disconnect_all(struct i40iw_device *iwdev) cm_node = container_of(list_node, struct i40iw_cm_node, connected_entry); attr.qp_state = IB_QPS_ERR; i40iw_modify_qp(&cm_node->iwqp->ibqp, &attr, IB_QP_STATE, NULL); + if (iwdev->reset) + i40iw_cm_disconn(cm_node->iwqp); i40iw_rem_ref_cm_node(cm_node); } } diff --git a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c index 98923a8cf86d..9ec1ae9a82c9 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c +++ b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c @@ -285,28 +285,20 @@ void i40iw_change_l2params(struct i40iw_sc_vsi *vsi, struct i40iw_l2params *l2pa struct i40iw_sc_dev *dev = vsi->dev; struct i40iw_sc_qp *qp = NULL; bool qs_handle_change = false; - bool mss_change = false; unsigned long flags; u16 qs_handle; int i; - if (vsi->mss != l2params->mss) { - mss_change = true; - vsi->mss = l2params->mss; - } + vsi->mss = l2params->mss; i40iw_fill_qos_list(l2params->qs_handle_list); for (i = 0; i < I40IW_MAX_USER_PRIORITY; i++) { qs_handle = l2params->qs_handle_list[i]; if (vsi->qos[i].qs_handle != qs_handle) qs_handle_change = true; - else if (!mss_change) - continue; /* no MSS nor qs handle change */ spin_lock_irqsave(&vsi->qos[i].lock, flags); qp = i40iw_get_qp(&vsi->qos[i].qplist, qp); while (qp) { - if (mss_change) - i40iw_qp_mss_modify(dev, qp); if (qs_handle_change) { qp->qs_handle = qs_handle; /* issue cqp suspend command */ @@ -450,6 +442,9 @@ static enum i40iw_status_code i40iw_sc_cqp_create(struct i40iw_sc_cqp *cqp, u32 cnt = 0, p1, p2, val = 0, err_code; enum i40iw_status_code ret_code; + *maj_err = 0; + *min_err = 0; + ret_code = i40iw_allocate_dma_mem(cqp->dev->hw, &cqp->sdbuf, 128, @@ -1975,6 +1970,8 @@ static enum i40iw_status_code i40iw_sc_ccq_destroy(struct i40iw_sc_cq *ccq, ret_code = i40iw_cqp_poll_registers(cqp, tail, 1000); } + cqp->process_cqp_sds = i40iw_update_sds_noccq; + return ret_code; } @@ -2392,7 +2389,6 @@ static enum i40iw_status_code i40iw_sc_qp_modify( set_64bit_val(wqe, 8, - LS_64(info->new_mss, I40IW_CQPSQ_QP_NEWMSS) | LS_64(term_len, I40IW_CQPSQ_QP_TERMLEN)); set_64bit_val(wqe, 16, qp->hw_host_ctx_pa); @@ -2407,7 +2403,6 @@ static enum i40iw_status_code i40iw_sc_qp_modify( LS_64(info->cq_num_valid, I40IW_CQPSQ_QP_CQNUMVALID) | LS_64(info->force_loopback, I40IW_CQPSQ_QP_FORCELOOPBACK) | LS_64(qp->qp_type, I40IW_CQPSQ_QP_QPTYPE) | - LS_64(info->mss_change, I40IW_CQPSQ_QP_MSSCHANGE) | LS_64(info->static_rsrc, I40IW_CQPSQ_QP_STATRSRC) | LS_64(info->remove_hash_idx, I40IW_CQPSQ_QP_REMOVEHASHENTRY) | LS_64(term_actions, I40IW_CQPSQ_QP_TERMACT) | @@ -4498,9 +4493,9 @@ void i40iw_sc_vsi_init(struct i40iw_sc_vsi *vsi, struct i40iw_vsi_init_info *inf i40iw_fill_qos_list(info->params->qs_handle_list); for (i = 0; i < I40IW_MAX_USER_PRIORITY; i++) { - vsi->qos[i].qs_handle = - info->params->qs_handle_list[i]; - i40iw_debug(vsi->dev, I40IW_DEBUG_DCB, "qset[%d]: %d\n", i, vsi->qos[i].qs_handle); + vsi->qos[i].qs_handle = info->params->qs_handle_list[i]; + i40iw_debug(vsi->dev, I40IW_DEBUG_DCB, "qset[%d]: %d\n", i, + vsi->qos[i].qs_handle); spin_lock_init(&vsi->qos[i].lock); INIT_LIST_HEAD(&vsi->qos[i].qplist); } @@ -4851,46 +4846,46 @@ void i40iw_vsi_stats_free(struct i40iw_sc_vsi *vsi) } static struct i40iw_cqp_ops iw_cqp_ops = { - i40iw_sc_cqp_init, - i40iw_sc_cqp_create, - i40iw_sc_cqp_post_sq, - i40iw_sc_cqp_get_next_send_wqe, - i40iw_sc_cqp_destroy, - i40iw_sc_poll_for_cqp_op_done + .cqp_init = i40iw_sc_cqp_init, + .cqp_create = i40iw_sc_cqp_create, + .cqp_post_sq = i40iw_sc_cqp_post_sq, + .cqp_get_next_send_wqe = i40iw_sc_cqp_get_next_send_wqe, + .cqp_destroy = i40iw_sc_cqp_destroy, + .poll_for_cqp_op_done = i40iw_sc_poll_for_cqp_op_done }; static struct i40iw_ccq_ops iw_ccq_ops = { - i40iw_sc_ccq_init, - i40iw_sc_ccq_create, - i40iw_sc_ccq_destroy, - i40iw_sc_ccq_create_done, - i40iw_sc_ccq_get_cqe_info, - i40iw_sc_ccq_arm + .ccq_init = i40iw_sc_ccq_init, + .ccq_create = i40iw_sc_ccq_create, + .ccq_destroy = i40iw_sc_ccq_destroy, + .ccq_create_done = i40iw_sc_ccq_create_done, + .ccq_get_cqe_info = i40iw_sc_ccq_get_cqe_info, + .ccq_arm = i40iw_sc_ccq_arm }; static struct i40iw_ceq_ops iw_ceq_ops = { - i40iw_sc_ceq_init, - i40iw_sc_ceq_create, - i40iw_sc_cceq_create_done, - i40iw_sc_cceq_destroy_done, - i40iw_sc_cceq_create, - i40iw_sc_ceq_destroy, - i40iw_sc_process_ceq + .ceq_init = i40iw_sc_ceq_init, + .ceq_create = i40iw_sc_ceq_create, + .cceq_create_done = i40iw_sc_cceq_create_done, + .cceq_destroy_done = i40iw_sc_cceq_destroy_done, + .cceq_create = i40iw_sc_cceq_create, + .ceq_destroy = i40iw_sc_ceq_destroy, + .process_ceq = i40iw_sc_process_ceq }; static struct i40iw_aeq_ops iw_aeq_ops = { - i40iw_sc_aeq_init, - i40iw_sc_aeq_create, - i40iw_sc_aeq_destroy, - i40iw_sc_get_next_aeqe, - i40iw_sc_repost_aeq_entries, - i40iw_sc_aeq_create_done, - i40iw_sc_aeq_destroy_done + .aeq_init = i40iw_sc_aeq_init, + .aeq_create = i40iw_sc_aeq_create, + .aeq_destroy = i40iw_sc_aeq_destroy, + .get_next_aeqe = i40iw_sc_get_next_aeqe, + .repost_aeq_entries = i40iw_sc_repost_aeq_entries, + .aeq_create_done = i40iw_sc_aeq_create_done, + .aeq_destroy_done = i40iw_sc_aeq_destroy_done }; /* iwarp pd ops */ static struct i40iw_pd_ops iw_pd_ops = { - i40iw_sc_pd_init, + .pd_init = i40iw_sc_pd_init, }; static struct i40iw_priv_qp_ops iw_priv_qp_ops = { @@ -4909,53 +4904,51 @@ static struct i40iw_priv_qp_ops iw_priv_qp_ops = { }; static struct i40iw_priv_cq_ops iw_priv_cq_ops = { - i40iw_sc_cq_init, - i40iw_sc_cq_create, - i40iw_sc_cq_destroy, - i40iw_sc_cq_modify, + .cq_init = i40iw_sc_cq_init, + .cq_create = i40iw_sc_cq_create, + .cq_destroy = i40iw_sc_cq_destroy, + .cq_modify = i40iw_sc_cq_modify, }; static struct i40iw_mr_ops iw_mr_ops = { - i40iw_sc_alloc_stag, - i40iw_sc_mr_reg_non_shared, - i40iw_sc_mr_reg_shared, - i40iw_sc_dealloc_stag, - i40iw_sc_query_stag, - i40iw_sc_mw_alloc + .alloc_stag = i40iw_sc_alloc_stag, + .mr_reg_non_shared = i40iw_sc_mr_reg_non_shared, + .mr_reg_shared = i40iw_sc_mr_reg_shared, + .dealloc_stag = i40iw_sc_dealloc_stag, + .query_stag = i40iw_sc_query_stag, + .mw_alloc = i40iw_sc_mw_alloc }; static struct i40iw_cqp_misc_ops iw_cqp_misc_ops = { - i40iw_sc_manage_push_page, - i40iw_sc_manage_hmc_pm_func_table, - i40iw_sc_set_hmc_resource_profile, - i40iw_sc_commit_fpm_values, - i40iw_sc_query_fpm_values, - i40iw_sc_static_hmc_pages_allocated, - i40iw_sc_add_arp_cache_entry, - i40iw_sc_del_arp_cache_entry, - i40iw_sc_query_arp_cache_entry, - i40iw_sc_manage_apbvt_entry, - i40iw_sc_manage_qhash_table_entry, - i40iw_sc_alloc_local_mac_ipaddr_entry, - i40iw_sc_add_local_mac_ipaddr_entry, - i40iw_sc_del_local_mac_ipaddr_entry, - i40iw_sc_cqp_nop, - i40iw_sc_commit_fpm_values_done, - i40iw_sc_query_fpm_values_done, - i40iw_sc_manage_hmc_pm_func_table_done, - i40iw_sc_suspend_qp, - i40iw_sc_resume_qp + .manage_push_page = i40iw_sc_manage_push_page, + .manage_hmc_pm_func_table = i40iw_sc_manage_hmc_pm_func_table, + .set_hmc_resource_profile = i40iw_sc_set_hmc_resource_profile, + .commit_fpm_values = i40iw_sc_commit_fpm_values, + .query_fpm_values = i40iw_sc_query_fpm_values, + .static_hmc_pages_allocated = i40iw_sc_static_hmc_pages_allocated, + .add_arp_cache_entry = i40iw_sc_add_arp_cache_entry, + .del_arp_cache_entry = i40iw_sc_del_arp_cache_entry, + .query_arp_cache_entry = i40iw_sc_query_arp_cache_entry, + .manage_apbvt_entry = i40iw_sc_manage_apbvt_entry, + .manage_qhash_table_entry = i40iw_sc_manage_qhash_table_entry, + .alloc_local_mac_ipaddr_table_entry = i40iw_sc_alloc_local_mac_ipaddr_entry, + .add_local_mac_ipaddr_entry = i40iw_sc_add_local_mac_ipaddr_entry, + .del_local_mac_ipaddr_entry = i40iw_sc_del_local_mac_ipaddr_entry, + .cqp_nop = i40iw_sc_cqp_nop, + .commit_fpm_values_done = i40iw_sc_commit_fpm_values_done, + .query_fpm_values_done = i40iw_sc_query_fpm_values_done, + .manage_hmc_pm_func_table_done = i40iw_sc_manage_hmc_pm_func_table_done, + .update_suspend_qp = i40iw_sc_suspend_qp, + .update_resume_qp = i40iw_sc_resume_qp }; static struct i40iw_hmc_ops iw_hmc_ops = { - i40iw_sc_init_iw_hmc, - i40iw_sc_parse_fpm_query_buf, - i40iw_sc_configure_iw_fpm, - i40iw_sc_parse_fpm_commit_buf, - i40iw_sc_create_hmc_obj, - i40iw_sc_del_hmc_obj, - NULL, - NULL + .init_iw_hmc = i40iw_sc_init_iw_hmc, + .parse_fpm_query_buf = i40iw_sc_parse_fpm_query_buf, + .configure_iw_fpm = i40iw_sc_configure_iw_fpm, + .parse_fpm_commit_buf = i40iw_sc_parse_fpm_commit_buf, + .create_hmc_object = i40iw_sc_create_hmc_obj, + .del_hmc_object = i40iw_sc_del_hmc_obj }; /** diff --git a/drivers/infiniband/hw/i40iw/i40iw_main.c b/drivers/infiniband/hw/i40iw/i40iw_main.c index 2728af3103ce..ae8463ff59a7 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_main.c +++ b/drivers/infiniband/hw/i40iw/i40iw_main.c @@ -243,6 +243,8 @@ static void i40iw_destroy_cqp(struct i40iw_device *iwdev, bool free_hwcqp) if (free_hwcqp) dev->cqp_ops->cqp_destroy(dev->cqp); + i40iw_cleanup_pending_cqp_op(iwdev); + i40iw_free_dma_mem(dev->hw, &cqp->sq); kfree(cqp->scratch_array); iwdev->cqp.scratch_array = NULL; @@ -274,13 +276,12 @@ static void i40iw_disable_irq(struct i40iw_sc_dev *dev, /** * i40iw_destroy_aeq - destroy aeq * @iwdev: iwarp device - * @reset: true if called before reset * * Issue a destroy aeq request and * free the resources associated with the aeq * The function is called during driver unload */ -static void i40iw_destroy_aeq(struct i40iw_device *iwdev, bool reset) +static void i40iw_destroy_aeq(struct i40iw_device *iwdev) { enum i40iw_status_code status = I40IW_ERR_NOT_READY; struct i40iw_sc_dev *dev = &iwdev->sc_dev; @@ -288,7 +289,7 @@ static void i40iw_destroy_aeq(struct i40iw_device *iwdev, bool reset) if (!iwdev->msix_shared) i40iw_disable_irq(dev, iwdev->iw_msixtbl, (void *)iwdev); - if (reset) + if (iwdev->reset) goto exit; if (!dev->aeq_ops->aeq_destroy(&aeq->sc_aeq, 0, 1)) @@ -304,19 +305,17 @@ exit: * i40iw_destroy_ceq - destroy ceq * @iwdev: iwarp device * @iwceq: ceq to be destroyed - * @reset: true if called before reset * * Issue a destroy ceq request and * free the resources associated with the ceq */ static void i40iw_destroy_ceq(struct i40iw_device *iwdev, - struct i40iw_ceq *iwceq, - bool reset) + struct i40iw_ceq *iwceq) { enum i40iw_status_code status; struct i40iw_sc_dev *dev = &iwdev->sc_dev; - if (reset) + if (iwdev->reset) goto exit; status = dev->ceq_ops->ceq_destroy(&iwceq->sc_ceq, 0, 1); @@ -335,12 +334,11 @@ exit: /** * i40iw_dele_ceqs - destroy all ceq's * @iwdev: iwarp device - * @reset: true if called before reset * * Go through all of the device ceq's and for each ceq * disable the ceq interrupt and destroy the ceq */ -static void i40iw_dele_ceqs(struct i40iw_device *iwdev, bool reset) +static void i40iw_dele_ceqs(struct i40iw_device *iwdev) { u32 i = 0; struct i40iw_sc_dev *dev = &iwdev->sc_dev; @@ -349,32 +347,31 @@ static void i40iw_dele_ceqs(struct i40iw_device *iwdev, bool reset) if (iwdev->msix_shared) { i40iw_disable_irq(dev, msix_vec, (void *)iwdev); - i40iw_destroy_ceq(iwdev, iwceq, reset); + i40iw_destroy_ceq(iwdev, iwceq); iwceq++; i++; } for (msix_vec++; i < iwdev->ceqs_count; i++, msix_vec++, iwceq++) { i40iw_disable_irq(dev, msix_vec, (void *)iwceq); - i40iw_destroy_ceq(iwdev, iwceq, reset); + i40iw_destroy_ceq(iwdev, iwceq); } } /** * i40iw_destroy_ccq - destroy control cq * @iwdev: iwarp device - * @reset: true if called before reset * * Issue destroy ccq request and * free the resources associated with the ccq */ -static void i40iw_destroy_ccq(struct i40iw_device *iwdev, bool reset) +static void i40iw_destroy_ccq(struct i40iw_device *iwdev) { struct i40iw_sc_dev *dev = &iwdev->sc_dev; struct i40iw_ccq *ccq = &iwdev->ccq; enum i40iw_status_code status = 0; - if (!reset) + if (!iwdev->reset) status = dev->ccq_ops->ccq_destroy(dev->ccq, 0, true); if (status) i40iw_pr_err("ccq destroy failed %d\n", status); @@ -810,7 +807,7 @@ static enum i40iw_status_code i40iw_setup_ceqs(struct i40iw_device *iwdev, iwceq->msix_idx = msix_vec->idx; status = i40iw_configure_ceq_vector(iwdev, iwceq, ceq_id, msix_vec); if (status) { - i40iw_destroy_ceq(iwdev, iwceq, false); + i40iw_destroy_ceq(iwdev, iwceq); break; } i40iw_enable_intr(&iwdev->sc_dev, msix_vec->idx); @@ -912,7 +909,7 @@ static enum i40iw_status_code i40iw_setup_aeq(struct i40iw_device *iwdev) status = i40iw_configure_aeq_vector(iwdev); if (status) { - i40iw_destroy_aeq(iwdev, false); + i40iw_destroy_aeq(iwdev); return status; } @@ -1319,13 +1316,13 @@ static enum i40iw_status_code i40iw_initialize_dev(struct i40iw_device *iwdev, status = i40iw_obj_aligned_mem(iwdev, &mem, I40IW_QUERY_FPM_BUF_SIZE, I40IW_FPM_QUERY_BUF_ALIGNMENT_MASK); if (status) - goto exit; + goto error; info.fpm_query_buf_pa = mem.pa; info.fpm_query_buf = mem.va; status = i40iw_obj_aligned_mem(iwdev, &mem, I40IW_COMMIT_FPM_BUF_SIZE, I40IW_FPM_COMMIT_BUF_ALIGNMENT_MASK); if (status) - goto exit; + goto error; info.fpm_commit_buf_pa = mem.pa; info.fpm_commit_buf = mem.va; info.hmc_fn_id = ldev->fid; @@ -1347,11 +1344,9 @@ static enum i40iw_status_code i40iw_initialize_dev(struct i40iw_device *iwdev, info.exception_lan_queue = 1; info.vchnl_send = i40iw_virtchnl_send; status = i40iw_device_init(&iwdev->sc_dev, &info); -exit: - if (status) { - kfree(iwdev->hmc_info_mem); - iwdev->hmc_info_mem = NULL; - } + + if (status) + goto error; memset(&vsi_info, 0, sizeof(vsi_info)); vsi_info.dev = &iwdev->sc_dev; vsi_info.back_vsi = (void *)iwdev; @@ -1362,11 +1357,19 @@ exit: memset(&stats_info, 0, sizeof(stats_info)); stats_info.fcn_id = ldev->fid; stats_info.pestat = kzalloc(sizeof(*stats_info.pestat), GFP_KERNEL); + if (!stats_info.pestat) { + status = I40IW_ERR_NO_MEMORY; + goto error; + } stats_info.stats_initialize = true; if (stats_info.pestat) i40iw_vsi_stats_init(&iwdev->vsi, &stats_info); } return status; +error: + kfree(iwdev->hmc_info_mem); + iwdev->hmc_info_mem = NULL; + return status; } /** @@ -1436,12 +1439,11 @@ static enum i40iw_status_code i40iw_save_msix_info(struct i40iw_device *iwdev, /** * i40iw_deinit_device - clean up the device resources * @iwdev: iwarp device - * @reset: true if called before reset * * Destroy the ib device interface, remove the mac ip entry and ipv4/ipv6 addresses, * destroy the device queues and free the pble and the hmc objects */ -static void i40iw_deinit_device(struct i40iw_device *iwdev, bool reset) +static void i40iw_deinit_device(struct i40iw_device *iwdev) { struct i40e_info *ldev = iwdev->ldev; @@ -1458,7 +1460,7 @@ static void i40iw_deinit_device(struct i40iw_device *iwdev, bool reset) i40iw_destroy_rdma_device(iwdev->iwibdev); /* fallthrough */ case IP_ADDR_REGISTERED: - if (!reset) + if (!iwdev->reset) i40iw_del_macip_entry(iwdev, (u8)iwdev->mac_ip_table_idx); /* fallthrough */ case INET_NOTIFIER: @@ -1468,26 +1470,26 @@ static void i40iw_deinit_device(struct i40iw_device *iwdev, bool reset) unregister_inet6addr_notifier(&i40iw_inetaddr6_notifier); } /* fallthrough */ + case PBLE_CHUNK_MEM: + i40iw_destroy_pble_pool(dev, iwdev->pble_rsrc); + /* fallthrough */ case CEQ_CREATED: - i40iw_dele_ceqs(iwdev, reset); + i40iw_dele_ceqs(iwdev); /* fallthrough */ case AEQ_CREATED: - i40iw_destroy_aeq(iwdev, reset); + i40iw_destroy_aeq(iwdev); /* fallthrough */ case IEQ_CREATED: - i40iw_puda_dele_resources(&iwdev->vsi, I40IW_PUDA_RSRC_TYPE_IEQ, reset); + i40iw_puda_dele_resources(&iwdev->vsi, I40IW_PUDA_RSRC_TYPE_IEQ, iwdev->reset); /* fallthrough */ case ILQ_CREATED: - i40iw_puda_dele_resources(&iwdev->vsi, I40IW_PUDA_RSRC_TYPE_ILQ, reset); + i40iw_puda_dele_resources(&iwdev->vsi, I40IW_PUDA_RSRC_TYPE_ILQ, iwdev->reset); /* fallthrough */ case CCQ_CREATED: - i40iw_destroy_ccq(iwdev, reset); - /* fallthrough */ - case PBLE_CHUNK_MEM: - i40iw_destroy_pble_pool(dev, iwdev->pble_rsrc); + i40iw_destroy_ccq(iwdev); /* fallthrough */ case HMC_OBJS_CREATED: - i40iw_del_hmc_objects(dev, dev->hmc_info, true, reset); + i40iw_del_hmc_objects(dev, dev->hmc_info, true, iwdev->reset); /* fallthrough */ case CQP_CREATED: i40iw_destroy_cqp(iwdev, true); @@ -1664,6 +1666,7 @@ static int i40iw_open(struct i40e_info *ldev, struct i40e_client *client) status = i40iw_hmc_init_pble(&iwdev->sc_dev, iwdev->pble_rsrc); if (status) break; + iwdev->init_state = PBLE_CHUNK_MEM; iwdev->virtchnl_wq = alloc_ordered_workqueue("iwvch", WQ_MEM_RECLAIM); i40iw_register_notifiers(); iwdev->init_state = INET_NOTIFIER; @@ -1687,7 +1690,7 @@ static int i40iw_open(struct i40e_info *ldev, struct i40e_client *client) } while (0); i40iw_pr_err("status = %d last completion = %d\n", status, iwdev->init_state); - i40iw_deinit_device(iwdev, false); + i40iw_deinit_device(iwdev); return -ERESTART; } @@ -1768,9 +1771,12 @@ static void i40iw_close(struct i40e_info *ldev, struct i40e_client *client, bool iwdev = &hdl->device; iwdev->closing = true; + if (reset) + iwdev->reset = true; + i40iw_cm_disconnect_all(iwdev); destroy_workqueue(iwdev->virtchnl_wq); - i40iw_deinit_device(iwdev, reset); + i40iw_deinit_device(iwdev); } /** @@ -1933,7 +1939,7 @@ static int i40iw_virtchnl_receive(struct i40e_info *ldev, bool i40iw_vf_clear_to_send(struct i40iw_sc_dev *dev) { struct i40iw_device *iwdev; - wait_queue_t wait; + wait_queue_entry_t wait; iwdev = dev->back_dev; diff --git a/drivers/infiniband/hw/i40iw/i40iw_osdep.h b/drivers/infiniband/hw/i40iw/i40iw_osdep.h index aa66c1c63dfa..f27be3e7830b 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_osdep.h +++ b/drivers/infiniband/hw/i40iw/i40iw_osdep.h @@ -199,7 +199,6 @@ void i40iw_cqp_spawn_worker(struct i40iw_sc_dev *dev, struct i40iw_virtchnl_work_info *work_info, u32 iw_vf_idx); void *i40iw_remove_head(struct list_head *list); void i40iw_qp_suspend_resume(struct i40iw_sc_dev *dev, struct i40iw_sc_qp *qp, bool suspend); -void i40iw_qp_mss_modify(struct i40iw_sc_dev *dev, struct i40iw_sc_qp *qp); void i40iw_term_modify_qp(struct i40iw_sc_qp *qp, u8 next_state, u8 term, u8 term_len); void i40iw_terminate_done(struct i40iw_sc_qp *qp, int timeout_occurred); diff --git a/drivers/infiniband/hw/i40iw/i40iw_puda.c b/drivers/infiniband/hw/i40iw/i40iw_puda.c index db41ab40da9c..71050c5d29a0 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_puda.c +++ b/drivers/infiniband/hw/i40iw/i40iw_puda.c @@ -408,6 +408,9 @@ enum i40iw_status_code i40iw_puda_send(struct i40iw_sc_qp *qp, set_64bit_val(wqe, 0, info->paddr); set_64bit_val(wqe, 8, LS_64(info->len, I40IWQPSQ_FRAG_LEN)); set_64bit_val(wqe, 16, header[0]); + + /* Ensure all data is written before writing valid bit */ + wmb(); set_64bit_val(wqe, 24, header[1]); i40iw_debug_buf(qp->dev, I40IW_DEBUG_PUDA, "PUDA SEND WQE", wqe, 32); @@ -1411,10 +1414,10 @@ static void i40iw_ieq_handle_exception(struct i40iw_puda_rsrc *ieq, if (!list_empty(rxlist)) { tmpbuf = (struct i40iw_puda_buf *)rxlist->next; - plist = &tmpbuf->list; while ((struct list_head *)tmpbuf != rxlist) { if ((int)(buf->seqnum - tmpbuf->seqnum) < 0) break; + plist = &tmpbuf->list; tmpbuf = (struct i40iw_puda_buf *)plist->next; } /* Insert buf before tmpbuf */ diff --git a/drivers/infiniband/hw/i40iw/i40iw_type.h b/drivers/infiniband/hw/i40iw/i40iw_type.h index 7b76259752b0..959ec81fba99 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_type.h +++ b/drivers/infiniband/hw/i40iw/i40iw_type.h @@ -541,7 +541,6 @@ struct i40iw_create_qp_info { struct i40iw_modify_qp_info { u64 rx_win0; u64 rx_win1; - u16 new_mss; u8 next_iwarp_state; u8 termlen; bool ord_valid; @@ -554,7 +553,6 @@ struct i40iw_modify_qp_info { bool dont_send_term; bool dont_send_fin; bool cached_var_valid; - bool mss_change; bool force_loopback; }; diff --git a/drivers/infiniband/hw/i40iw/i40iw_uk.c b/drivers/infiniband/hw/i40iw/i40iw_uk.c index 2800f796271c..b0d3a0e8a9b5 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_uk.c +++ b/drivers/infiniband/hw/i40iw/i40iw_uk.c @@ -913,29 +913,29 @@ enum i40iw_status_code i40iw_get_wqe_shift(u32 wqdepth, u32 sge, u32 inline_data } static struct i40iw_qp_uk_ops iw_qp_uk_ops = { - i40iw_qp_post_wr, - i40iw_qp_ring_push_db, - i40iw_rdma_write, - i40iw_rdma_read, - i40iw_send, - i40iw_inline_rdma_write, - i40iw_inline_send, - i40iw_stag_local_invalidate, - i40iw_mw_bind, - i40iw_post_receive, - i40iw_nop + .iw_qp_post_wr = i40iw_qp_post_wr, + .iw_qp_ring_push_db = i40iw_qp_ring_push_db, + .iw_rdma_write = i40iw_rdma_write, + .iw_rdma_read = i40iw_rdma_read, + .iw_send = i40iw_send, + .iw_inline_rdma_write = i40iw_inline_rdma_write, + .iw_inline_send = i40iw_inline_send, + .iw_stag_local_invalidate = i40iw_stag_local_invalidate, + .iw_mw_bind = i40iw_mw_bind, + .iw_post_receive = i40iw_post_receive, + .iw_post_nop = i40iw_nop }; static struct i40iw_cq_ops iw_cq_ops = { - i40iw_cq_request_notification, - i40iw_cq_poll_completion, - i40iw_cq_post_entries, - i40iw_clean_cq + .iw_cq_request_notification = i40iw_cq_request_notification, + .iw_cq_poll_completion = i40iw_cq_poll_completion, + .iw_cq_post_entries = i40iw_cq_post_entries, + .iw_cq_clean = i40iw_clean_cq }; static struct i40iw_device_uk_ops iw_device_uk_ops = { - i40iw_cq_uk_init, - i40iw_qp_uk_init, + .iwarp_cq_uk_init = i40iw_cq_uk_init, + .iwarp_qp_uk_init = i40iw_qp_uk_init, }; /** diff --git a/drivers/infiniband/hw/i40iw/i40iw_utils.c b/drivers/infiniband/hw/i40iw/i40iw_utils.c index 0f5d43d1f5fc..e311ec559f4e 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_utils.c +++ b/drivers/infiniband/hw/i40iw/i40iw_utils.c @@ -160,6 +160,9 @@ int i40iw_inetaddr_event(struct notifier_block *notifier, return NOTIFY_DONE; iwdev = &hdl->device; + if (iwdev->init_state < INET_NOTIFIER) + return NOTIFY_DONE; + netdev = iwdev->ldev->netdev; upper_dev = netdev_master_upper_dev_get(netdev); if (netdev != event_netdev) @@ -214,6 +217,9 @@ int i40iw_inet6addr_event(struct notifier_block *notifier, return NOTIFY_DONE; iwdev = &hdl->device; + if (iwdev->init_state < INET_NOTIFIER) + return NOTIFY_DONE; + netdev = iwdev->ldev->netdev; if (netdev != event_netdev) return NOTIFY_DONE; @@ -260,6 +266,8 @@ int i40iw_net_event(struct notifier_block *notifier, unsigned long event, void * if (!iwhdl) return NOTIFY_DONE; iwdev = &iwhdl->device; + if (iwdev->init_state < INET_NOTIFIER) + return NOTIFY_DONE; p = (__be32 *)neigh->primary_key; i40iw_copy_ip_ntohl(local_ipaddr, p); if (neigh->nud_state & NUD_VALID) { @@ -329,6 +337,7 @@ struct i40iw_cqp_request *i40iw_get_cqp_request(struct i40iw_cqp *cqp, bool wait */ void i40iw_free_cqp_request(struct i40iw_cqp *cqp, struct i40iw_cqp_request *cqp_request) { + struct i40iw_device *iwdev = container_of(cqp, struct i40iw_device, cqp); unsigned long flags; if (cqp_request->dynamic) { @@ -342,6 +351,7 @@ void i40iw_free_cqp_request(struct i40iw_cqp *cqp, struct i40iw_cqp_request *cqp list_add_tail(&cqp_request->list, &cqp->cqp_avail_reqs); spin_unlock_irqrestore(&cqp->req_lock, flags); } + wake_up(&iwdev->close_wq); } /** @@ -357,6 +367,56 @@ void i40iw_put_cqp_request(struct i40iw_cqp *cqp, } /** + * i40iw_free_pending_cqp_request -free pending cqp request objs + * @cqp: cqp ptr + * @cqp_request: to be put back in cqp list + */ +static void i40iw_free_pending_cqp_request(struct i40iw_cqp *cqp, + struct i40iw_cqp_request *cqp_request) +{ + struct i40iw_device *iwdev = container_of(cqp, struct i40iw_device, cqp); + + if (cqp_request->waiting) { + cqp_request->compl_info.error = true; + cqp_request->request_done = true; + wake_up(&cqp_request->waitq); + } + i40iw_put_cqp_request(cqp, cqp_request); + wait_event_timeout(iwdev->close_wq, + !atomic_read(&cqp_request->refcount), + 1000); +} + +/** + * i40iw_cleanup_pending_cqp_op - clean-up cqp with no completions + * @iwdev: iwarp device + */ +void i40iw_cleanup_pending_cqp_op(struct i40iw_device *iwdev) +{ + struct i40iw_sc_dev *dev = &iwdev->sc_dev; + struct i40iw_cqp *cqp = &iwdev->cqp; + struct i40iw_cqp_request *cqp_request = NULL; + struct cqp_commands_info *pcmdinfo = NULL; + u32 i, pending_work, wqe_idx; + + pending_work = I40IW_RING_WORK_AVAILABLE(cqp->sc_cqp.sq_ring); + wqe_idx = I40IW_RING_GETCURRENT_TAIL(cqp->sc_cqp.sq_ring); + for (i = 0; i < pending_work; i++) { + cqp_request = (struct i40iw_cqp_request *)(unsigned long)cqp->scratch_array[wqe_idx]; + if (cqp_request) + i40iw_free_pending_cqp_request(cqp, cqp_request); + wqe_idx = (wqe_idx + 1) % I40IW_RING_GETSIZE(cqp->sc_cqp.sq_ring); + } + + while (!list_empty(&dev->cqp_cmd_head)) { + pcmdinfo = (struct cqp_commands_info *)i40iw_remove_head(&dev->cqp_cmd_head); + cqp_request = container_of(pcmdinfo, struct i40iw_cqp_request, info); + if (cqp_request) + i40iw_free_pending_cqp_request(cqp, cqp_request); + } +} + +/** * i40iw_free_qp - callback after destroy cqp completes * @cqp_request: cqp request for destroy qp * @num: not used @@ -538,8 +598,12 @@ void i40iw_rem_ref(struct ib_qp *ibqp) cqp_info->in.u.qp_destroy.scratch = (uintptr_t)cqp_request; cqp_info->in.u.qp_destroy.remove_hash_idx = true; status = i40iw_handle_cqp_op(iwdev, cqp_request); - if (status) - i40iw_pr_err("CQP-OP Destroy QP fail"); + if (!status) + return; + + i40iw_rem_pdusecount(iwqp->iwpd, iwdev); + i40iw_free_qp_resources(iwdev, iwqp, qp_num); + i40iw_rem_devusecount(iwdev); } /** @@ -749,23 +813,6 @@ void i40iw_qp_suspend_resume(struct i40iw_sc_dev *dev, struct i40iw_sc_qp *qp, b } /** - * i40iw_qp_mss_modify - modify mss for qp - * @dev: hardware control device structure - * @qp: hardware control qp - */ -void i40iw_qp_mss_modify(struct i40iw_sc_dev *dev, struct i40iw_sc_qp *qp) -{ - struct i40iw_device *iwdev = (struct i40iw_device *)dev->back_dev; - struct i40iw_qp *iwqp = (struct i40iw_qp *)qp->back_qp; - struct i40iw_modify_qp_info info; - - memset(&info, 0, sizeof(info)); - info.mss_change = true; - info.new_mss = qp->vsi->mss; - i40iw_hw_modify_qp(iwdev, iwqp, &info, false); -} - -/** * i40iw_term_modify_qp - modify qp for term message * @qp: hardware control qp * @next_state: qp's next state @@ -836,10 +883,9 @@ void i40iw_terminate_start_timer(struct i40iw_sc_qp *qp) iwqp = (struct i40iw_qp *)qp->back_qp; i40iw_add_ref(&iwqp->ibqp); - init_timer(&iwqp->terminate_timer); - iwqp->terminate_timer.function = i40iw_terminate_timeout; + setup_timer(&iwqp->terminate_timer, i40iw_terminate_timeout, + (unsigned long)iwqp); iwqp->terminate_timer.expires = jiffies + HZ; - iwqp->terminate_timer.data = (unsigned long)iwqp; add_timer(&iwqp->terminate_timer); } @@ -1428,9 +1474,8 @@ void i40iw_hw_stats_start_timer(struct i40iw_sc_vsi *vsi) { struct i40iw_vsi_pestat *devstat = vsi->pestat; - init_timer(&devstat->stats_timer); - devstat->stats_timer.function = i40iw_hw_stats_timeout; - devstat->stats_timer.data = (unsigned long)vsi; + setup_timer(&devstat->stats_timer, i40iw_hw_stats_timeout, + (unsigned long)vsi); mod_timer(&devstat->stats_timer, jiffies + msecs_to_jiffies(STATS_TIMER_DELAY)); } diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c b/drivers/infiniband/hw/i40iw/i40iw_verbs.c index 4c000d60d5c6..02d871db7ca5 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c +++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c @@ -97,8 +97,7 @@ static int i40iw_query_port(struct ib_device *ibdev, struct i40iw_device *iwdev = to_iwdev(ibdev); struct net_device *netdev = iwdev->netdev; - memset(props, 0, sizeof(*props)); - + /* props being zeroed by the caller, avoid zeroing it here */ props->max_mtu = IB_MTU_4096; props->active_mtu = ib_mtu_int_to_enum(netdev->mtu); @@ -427,9 +426,13 @@ void i40iw_free_qp_resources(struct i40iw_device *iwdev, struct i40iw_qp *iwqp, u32 qp_num) { + struct i40iw_pbl *iwpbl = &iwqp->iwpbl; + i40iw_dealloc_push_page(iwdev, &iwqp->sc_qp); if (qp_num) i40iw_free_resource(iwdev, iwdev->allocated_qps, qp_num); + if (iwpbl->pbl_allocated) + i40iw_free_pble(iwdev->pble_rsrc, &iwpbl->pble_alloc); i40iw_free_dma_mem(iwdev->sc_dev.hw, &iwqp->q2_ctx_mem); i40iw_free_dma_mem(iwdev->sc_dev.hw, &iwqp->kqp.dma_mem); kfree(iwqp->kqp.wrid_mem); @@ -484,7 +487,7 @@ static int i40iw_setup_virt_qp(struct i40iw_device *iwdev, struct i40iw_qp *iwqp, struct i40iw_qp_init_info *init_info) { - struct i40iw_pbl *iwpbl = iwqp->iwpbl; + struct i40iw_pbl *iwpbl = &iwqp->iwpbl; struct i40iw_qp_mr *qpmr = &iwpbl->qp_mr; iwqp->page = qpmr->sq_page; @@ -689,19 +692,22 @@ static struct ib_qp *i40iw_create_qp(struct ib_pd *ibpd, ucontext = to_ucontext(ibpd->uobject->context); if (req.user_wqe_buffers) { + struct i40iw_pbl *iwpbl; + spin_lock_irqsave( &ucontext->qp_reg_mem_list_lock, flags); - iwqp->iwpbl = i40iw_get_pbl( + iwpbl = i40iw_get_pbl( (unsigned long)req.user_wqe_buffers, &ucontext->qp_reg_mem_list); spin_unlock_irqrestore( &ucontext->qp_reg_mem_list_lock, flags); - if (!iwqp->iwpbl) { + if (!iwpbl) { err_code = -ENODATA; i40iw_pr_err("no pbl info\n"); goto error; } + memcpy(&iwqp->iwpbl, iwpbl, sizeof(iwqp->iwpbl)); } } err_code = i40iw_setup_virt_qp(iwdev, iwqp, &init_info); @@ -1162,8 +1168,10 @@ static struct ib_cq *i40iw_create_cq(struct ib_device *ibdev, memset(&req, 0, sizeof(req)); iwcq->user_mode = true; ucontext = to_ucontext(context); - if (ib_copy_from_udata(&req, udata, sizeof(struct i40iw_create_cq_req))) + if (ib_copy_from_udata(&req, udata, sizeof(struct i40iw_create_cq_req))) { + err_code = -EFAULT; goto cq_free_resources; + } spin_lock_irqsave(&ucontext->cq_reg_mem_list_lock, flags); iwpbl = i40iw_get_pbl((unsigned long)req.user_cq_buffer, @@ -1346,7 +1354,7 @@ static void i40iw_copy_user_pgaddrs(struct i40iw_mr *iwmr, { struct ib_umem *region = iwmr->region; struct i40iw_pbl *iwpbl = &iwmr->iwpbl; - int chunk_pages, entry, pg_shift, i; + int chunk_pages, entry, i; struct i40iw_pble_alloc *palloc = &iwpbl->pble_alloc; struct i40iw_pble_info *pinfo; struct scatterlist *sg; @@ -1355,14 +1363,14 @@ static void i40iw_copy_user_pgaddrs(struct i40iw_mr *iwmr, pinfo = (level == I40IW_LEVEL_1) ? NULL : palloc->level2.leaf; - pg_shift = ffs(region->page_size) - 1; for_each_sg(region->sg_head.sgl, sg, region->nmap, entry) { - chunk_pages = sg_dma_len(sg) >> pg_shift; + chunk_pages = sg_dma_len(sg) >> region->page_shift; if ((iwmr->type == IW_MEMREG_TYPE_QP) && !iwpbl->qp_mr.sq_page) iwpbl->qp_mr.sq_page = sg_page(sg); for (i = 0; i < chunk_pages; i++) { - pg_addr = sg_dma_address(sg) + region->page_size * i; + pg_addr = sg_dma_address(sg) + + (i << region->page_shift); if ((entry + i) == 0) *pbl = cpu_to_le64(pg_addr & iwmr->page_msk); @@ -1848,7 +1856,7 @@ static struct ib_mr *i40iw_reg_user_mr(struct ib_pd *pd, iwmr->ibmr.device = pd->device; ucontext = to_ucontext(pd->uobject->context); - iwmr->page_size = region->page_size; + iwmr->page_size = PAGE_SIZE; iwmr->page_msk = PAGE_MASK; if (region->hugetlb && (req.reg_type == IW_MEMREG_TYPE_MEM)) @@ -2064,7 +2072,7 @@ static int i40iw_dereg_mr(struct ib_mr *ib_mr) ucontext = to_ucontext(ibpd->uobject->context); i40iw_del_memlist(iwmr, ucontext); } - if (iwpbl->pbl_allocated) + if (iwpbl->pbl_allocated && iwmr->type != IW_MEMREG_TYPE_QP) i40iw_free_pble(iwdev->pble_rsrc, palloc); kfree(iwmr); return 0; @@ -2497,14 +2505,15 @@ static int i40iw_port_immutable(struct ib_device *ibdev, u8 port_num, struct ib_port_attr attr; int err; - err = i40iw_query_port(ibdev, port_num, &attr); + immutable->core_cap_flags = RDMA_CORE_PORT_IWARP; + + err = ib_query_port(ibdev, port_num, &attr); if (err) return err; immutable->pkey_tbl_len = attr.pkey_tbl_len; immutable->gid_tbl_len = attr.gid_tbl_len; - immutable->core_cap_flags = RDMA_CORE_PORT_IWARP; return 0; } @@ -2696,7 +2705,7 @@ static int i40iw_query_pkey(struct ib_device *ibdev, * @ah_attr: address handle attributes */ static struct ib_ah *i40iw_create_ah(struct ib_pd *ibpd, - struct ib_ah_attr *attr, + struct rdma_ah_attr *attr, struct ib_udata *udata) { @@ -2758,7 +2767,6 @@ static struct i40iw_ib_device *i40iw_init_rdma_device(struct i40iw_device *iwdev (1ull << IB_USER_VERBS_CMD_POST_SEND); iwibdev->ibdev.phys_port_cnt = 1; iwibdev->ibdev.num_comp_vectors = iwdev->ceqs_count; - iwibdev->ibdev.dma_device = &pcidev->dev; iwibdev->ibdev.dev.parent = &pcidev->dev; iwibdev->ibdev.query_port = i40iw_query_port; iwibdev->ibdev.modify_port = i40iw_modify_port; diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.h b/drivers/infiniband/hw/i40iw/i40iw_verbs.h index 07c3fec77de6..9067443cd311 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_verbs.h +++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.h @@ -170,7 +170,7 @@ struct i40iw_qp { struct i40iw_qp_kmode kqp; struct i40iw_dma_mem host_ctx; struct timer_list terminate_timer; - struct i40iw_pbl *iwpbl; + struct i40iw_pbl iwpbl; struct i40iw_dma_mem q2_ctx_mem; struct i40iw_dma_mem ietf_mem; struct completion sq_drained; diff --git a/drivers/infiniband/hw/i40iw/i40iw_virtchnl.c b/drivers/infiniband/hw/i40iw/i40iw_virtchnl.c index f4d13683a403..48fd327f876b 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_virtchnl.c +++ b/drivers/infiniband/hw/i40iw/i40iw_virtchnl.c @@ -443,10 +443,7 @@ enum i40iw_status_code i40iw_vchnl_recv_pf(struct i40iw_sc_dev *dev, if (!dev->vchnl_up) return I40IW_ERR_NOT_READY; if (vchnl_msg->iw_op_code == I40IW_VCHNL_OP_GET_VER) { - if (vchnl_msg->iw_op_ver != I40IW_VCHNL_OP_GET_VER_V0) - vchnl_pf_send_get_ver_resp(dev, vf_id, vchnl_msg); - else - vchnl_pf_send_get_ver_resp(dev, vf_id, vchnl_msg); + vchnl_pf_send_get_ver_resp(dev, vf_id, vchnl_msg); return I40IW_SUCCESS; } for (iw_vf_idx = 0; iw_vf_idx < I40IW_MAX_PE_ENABLED_VF_COUNT; iw_vf_idx++) { diff --git a/drivers/infiniband/hw/mlx4/ah.c b/drivers/infiniband/hw/mlx4/ah.c index 077c33d2dc75..538c46a73248 100644 --- a/drivers/infiniband/hw/mlx4/ah.c +++ b/drivers/infiniband/hw/mlx4/ah.c @@ -40,36 +40,45 @@ #include "mlx4_ib.h" -static struct ib_ah *create_ib_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr, +static struct ib_ah *create_ib_ah(struct ib_pd *pd, + struct rdma_ah_attr *ah_attr, struct mlx4_ib_ah *ah) { struct mlx4_dev *dev = to_mdev(pd->device)->dev; - ah->av.ib.port_pd = cpu_to_be32(to_mpd(pd)->pdn | (ah_attr->port_num << 24)); - ah->av.ib.g_slid = ah_attr->src_path_bits; - ah->av.ib.sl_tclass_flowlabel = cpu_to_be32(ah_attr->sl << 28); - if (ah_attr->ah_flags & IB_AH_GRH) { + ah->av.ib.port_pd = cpu_to_be32(to_mpd(pd)->pdn | + (rdma_ah_get_port_num(ah_attr) << 24)); + ah->av.ib.g_slid = rdma_ah_get_path_bits(ah_attr); + ah->av.ib.sl_tclass_flowlabel = + cpu_to_be32(rdma_ah_get_sl(ah_attr) << 28); + if (rdma_ah_get_ah_flags(ah_attr) & IB_AH_GRH) { + const struct ib_global_route *grh = rdma_ah_read_grh(ah_attr); + ah->av.ib.g_slid |= 0x80; - ah->av.ib.gid_index = ah_attr->grh.sgid_index; - ah->av.ib.hop_limit = ah_attr->grh.hop_limit; + ah->av.ib.gid_index = grh->sgid_index; + ah->av.ib.hop_limit = grh->hop_limit; ah->av.ib.sl_tclass_flowlabel |= - cpu_to_be32((ah_attr->grh.traffic_class << 20) | - ah_attr->grh.flow_label); - memcpy(ah->av.ib.dgid, ah_attr->grh.dgid.raw, 16); + cpu_to_be32((grh->traffic_class << 20) | + grh->flow_label); + memcpy(ah->av.ib.dgid, grh->dgid.raw, 16); } - ah->av.ib.dlid = cpu_to_be16(ah_attr->dlid); - if (ah_attr->static_rate) { - ah->av.ib.stat_rate = ah_attr->static_rate + MLX4_STAT_RATE_OFFSET; - while (ah->av.ib.stat_rate > IB_RATE_2_5_GBPS + MLX4_STAT_RATE_OFFSET && - !(1 << ah->av.ib.stat_rate & dev->caps.stat_rate_support)) - --ah->av.ib.stat_rate; + ah->av.ib.dlid = cpu_to_be16(rdma_ah_get_dlid(ah_attr)); + if (rdma_ah_get_static_rate(ah_attr)) { + u8 static_rate = rdma_ah_get_static_rate(ah_attr) + + MLX4_STAT_RATE_OFFSET; + + while (static_rate > IB_RATE_2_5_GBPS + MLX4_STAT_RATE_OFFSET && + !(1 << static_rate & dev->caps.stat_rate_support)) + --static_rate; + ah->av.ib.stat_rate = static_rate; } return &ah->ibah; } -static struct ib_ah *create_iboe_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr, +static struct ib_ah *create_iboe_ah(struct ib_pd *pd, + struct rdma_ah_attr *ah_attr, struct mlx4_ib_ah *ah) { struct mlx4_ib_dev *ibdev = to_mdev(pd->device); @@ -79,17 +88,18 @@ static struct ib_ah *create_iboe_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr u16 vlan_tag = 0xffff; union ib_gid sgid; struct ib_gid_attr gid_attr; + const struct ib_global_route *grh = rdma_ah_read_grh(ah_attr); int ret; - memcpy(&in6, ah_attr->grh.dgid.raw, sizeof(in6)); + memcpy(&in6, grh->dgid.raw, sizeof(in6)); if (rdma_is_multicast_addr(&in6)) { is_mcast = 1; rdma_get_mcast_mac(&in6, ah->av.eth.mac); } else { - memcpy(ah->av.eth.mac, ah_attr->dmac, ETH_ALEN); + memcpy(ah->av.eth.mac, ah_attr->roce.dmac, ETH_ALEN); } - ret = ib_get_cached_gid(pd->device, ah_attr->port_num, - ah_attr->grh.sgid_index, &sgid, &gid_attr); + ret = ib_get_cached_gid(pd->device, rdma_ah_get_port_num(ah_attr), + grh->sgid_index, &sgid, &gid_attr); if (ret) return ERR_PTR(ret); eth_zero_addr(ah->av.eth.s_mac); @@ -100,36 +110,40 @@ static struct ib_ah *create_iboe_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr dev_put(gid_attr.ndev); } if (vlan_tag < 0x1000) - vlan_tag |= (ah_attr->sl & 7) << 13; - ah->av.eth.port_pd = cpu_to_be32(to_mpd(pd)->pdn | (ah_attr->port_num << 24)); - ret = mlx4_ib_gid_index_to_real_index(ibdev, ah_attr->port_num, ah_attr->grh.sgid_index); + vlan_tag |= (rdma_ah_get_sl(ah_attr) & 7) << 13; + ah->av.eth.port_pd = cpu_to_be32(to_mpd(pd)->pdn | + (rdma_ah_get_port_num(ah_attr) << 24)); + ret = mlx4_ib_gid_index_to_real_index(ibdev, + rdma_ah_get_port_num(ah_attr), + grh->sgid_index); if (ret < 0) return ERR_PTR(ret); ah->av.eth.gid_index = ret; ah->av.eth.vlan = cpu_to_be16(vlan_tag); - ah->av.eth.hop_limit = ah_attr->grh.hop_limit; - if (ah_attr->static_rate) { - ah->av.eth.stat_rate = ah_attr->static_rate + MLX4_STAT_RATE_OFFSET; + ah->av.eth.hop_limit = grh->hop_limit; + if (rdma_ah_get_static_rate(ah_attr)) { + ah->av.eth.stat_rate = rdma_ah_get_static_rate(ah_attr) + + MLX4_STAT_RATE_OFFSET; while (ah->av.eth.stat_rate > IB_RATE_2_5_GBPS + MLX4_STAT_RATE_OFFSET && !(1 << ah->av.eth.stat_rate & dev->caps.stat_rate_support)) --ah->av.eth.stat_rate; } ah->av.eth.sl_tclass_flowlabel |= - cpu_to_be32((ah_attr->grh.traffic_class << 20) | - ah_attr->grh.flow_label); + cpu_to_be32((grh->traffic_class << 20) | + grh->flow_label); /* * HW requires multicast LID so we just choose one. */ if (is_mcast) ah->av.ib.dlid = cpu_to_be16(0xc000); - memcpy(ah->av.eth.dgid, ah_attr->grh.dgid.raw, 16); - ah->av.eth.sl_tclass_flowlabel |= cpu_to_be32(ah_attr->sl << 29); - + memcpy(ah->av.eth.dgid, grh->dgid.raw, 16); + ah->av.eth.sl_tclass_flowlabel |= cpu_to_be32(rdma_ah_get_sl(ah_attr) + << 29); return &ah->ibah; } -struct ib_ah *mlx4_ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr, +struct ib_ah *mlx4_ib_create_ah(struct ib_pd *pd, struct rdma_ah_attr *ah_attr, struct ib_udata *udata) { @@ -140,8 +154,8 @@ struct ib_ah *mlx4_ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr, if (!ah) return ERR_PTR(-ENOMEM); - if (rdma_port_get_link_layer(pd->device, ah_attr->port_num) == IB_LINK_LAYER_ETHERNET) { - if (!(ah_attr->ah_flags & IB_AH_GRH)) { + if (ah_attr->type == RDMA_AH_ATTR_TYPE_ROCE) { + if (!(rdma_ah_get_ah_flags(ah_attr) & IB_AH_GRH)) { ret = ERR_PTR(-EINVAL); } else { /* @@ -163,34 +177,40 @@ struct ib_ah *mlx4_ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr, return create_ib_ah(pd, ah_attr, ah); /* never fails */ } -int mlx4_ib_query_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr) +int mlx4_ib_query_ah(struct ib_ah *ibah, struct rdma_ah_attr *ah_attr) { struct mlx4_ib_ah *ah = to_mah(ibah); - enum rdma_link_layer ll; + int port_num = be32_to_cpu(ah->av.ib.port_pd) >> 24; memset(ah_attr, 0, sizeof *ah_attr); - ah_attr->port_num = be32_to_cpu(ah->av.ib.port_pd) >> 24; - ll = rdma_port_get_link_layer(ibah->device, ah_attr->port_num); - if (ll == IB_LINK_LAYER_ETHERNET) - ah_attr->sl = be32_to_cpu(ah->av.eth.sl_tclass_flowlabel) >> 29; - else - ah_attr->sl = be32_to_cpu(ah->av.ib.sl_tclass_flowlabel) >> 28; - - ah_attr->dlid = ll == IB_LINK_LAYER_INFINIBAND ? be16_to_cpu(ah->av.ib.dlid) : 0; - if (ah->av.ib.stat_rate) - ah_attr->static_rate = ah->av.ib.stat_rate - MLX4_STAT_RATE_OFFSET; - ah_attr->src_path_bits = ah->av.ib.g_slid & 0x7F; + ah_attr->type = ibah->type; + if (ah_attr->type == RDMA_AH_ATTR_TYPE_ROCE) { + rdma_ah_set_dlid(ah_attr, 0); + rdma_ah_set_sl(ah_attr, + be32_to_cpu(ah->av.eth.sl_tclass_flowlabel) + >> 29); + } else { + rdma_ah_set_dlid(ah_attr, be16_to_cpu(ah->av.ib.dlid)); + rdma_ah_set_sl(ah_attr, + be32_to_cpu(ah->av.ib.sl_tclass_flowlabel) + >> 28); + } + + rdma_ah_set_port_num(ah_attr, port_num); + if (ah->av.ib.stat_rate) + rdma_ah_set_static_rate(ah_attr, + ah->av.ib.stat_rate - + MLX4_STAT_RATE_OFFSET); + rdma_ah_set_path_bits(ah_attr, ah->av.ib.g_slid & 0x7F); if (mlx4_ib_ah_grh_present(ah)) { - ah_attr->ah_flags = IB_AH_GRH; - - ah_attr->grh.traffic_class = - be32_to_cpu(ah->av.ib.sl_tclass_flowlabel) >> 20; - ah_attr->grh.flow_label = - be32_to_cpu(ah->av.ib.sl_tclass_flowlabel) & 0xfffff; - ah_attr->grh.hop_limit = ah->av.ib.hop_limit; - ah_attr->grh.sgid_index = ah->av.ib.gid_index; - memcpy(ah_attr->grh.dgid.raw, ah->av.ib.dgid, 16); + u32 tc_fl = be32_to_cpu(ah->av.ib.sl_tclass_flowlabel); + + rdma_ah_set_grh(ah_attr, NULL, + tc_fl & 0xfffff, ah->av.ib.gid_index, + ah->av.ib.hop_limit, + tc_fl >> 20); + rdma_ah_set_dgid_raw(ah_attr, ah->av.ib.dgid); } return 0; diff --git a/drivers/infiniband/hw/mlx4/alias_GUID.c b/drivers/infiniband/hw/mlx4/alias_GUID.c index 06020c54db20..ea24230ea0d4 100644 --- a/drivers/infiniband/hw/mlx4/alias_GUID.c +++ b/drivers/infiniband/hw/mlx4/alias_GUID.c @@ -499,6 +499,7 @@ static int set_guid_rec(struct ib_device *ibdev, struct list_head *head = &dev->sriov.alias_guid.ports_guid[port - 1].cb_list; + memset(&attr, 0, sizeof(attr)); err = __mlx4_ib_query_port(ibdev, port, &attr, 1); if (err) { pr_debug("mlx4_ib_query_port failed (err: %d), port: %d\n", diff --git a/drivers/infiniband/hw/mlx4/cm.c b/drivers/infiniband/hw/mlx4/cm.c index d64845335e87..fedaf8260105 100644 --- a/drivers/infiniband/hw/mlx4/cm.c +++ b/drivers/infiniband/hw/mlx4/cm.c @@ -279,14 +279,14 @@ id_map_alloc(struct ib_device *ibdev, int slave_id, u32 sl_cm_id) } static struct id_map_entry * -id_map_get(struct ib_device *ibdev, int *pv_cm_id, int sl_cm_id, int slave_id) +id_map_get(struct ib_device *ibdev, int *pv_cm_id, int slave_id, int sl_cm_id) { struct id_map_entry *ent; struct mlx4_ib_sriov *sriov = &to_mdev(ibdev)->sriov; spin_lock(&sriov->id_map_lock); if (*pv_cm_id == -1) { - ent = id_map_find_by_sl_id(ibdev, sl_cm_id, slave_id); + ent = id_map_find_by_sl_id(ibdev, slave_id, sl_cm_id); if (ent) *pv_cm_id = (int) ent->pv_cm_id; } else @@ -323,6 +323,9 @@ int mlx4_ib_multiplex_cm_handler(struct ib_device *ibdev, int port, int slave_id mad->mad_hdr.attr_id == CM_REP_ATTR_ID || mad->mad_hdr.attr_id == CM_SIDR_REQ_ATTR_ID) { sl_cm_id = get_local_comm_id(mad); + id = id_map_get(ibdev, &pv_cm_id, slave_id, sl_cm_id); + if (id) + goto cont; id = id_map_alloc(ibdev, slave_id, sl_cm_id); if (IS_ERR(id)) { mlx4_ib_warn(ibdev, "%s: id{slave: %d, sl_cm_id: 0x%x} Failed to id_map_alloc\n", @@ -343,6 +346,7 @@ int mlx4_ib_multiplex_cm_handler(struct ib_device *ibdev, int port, int slave_id return -EINVAL; } +cont: set_local_comm_id(mad, id->pv_cm_id); if (mad->mad_hdr.attr_id == CM_DREQ_ATTR_ID) @@ -414,7 +418,7 @@ void mlx4_ib_cm_paravirt_clean(struct mlx4_ib_dev *dev, int slave) struct rb_root *sl_id_map = &sriov->sl_id_map; struct list_head lh; struct rb_node *nd; - int need_flush = 1; + int need_flush = 0; struct id_map_entry *map, *tmp_map; /* cancel all delayed work queue entries */ INIT_LIST_HEAD(&lh); @@ -422,13 +426,13 @@ void mlx4_ib_cm_paravirt_clean(struct mlx4_ib_dev *dev, int slave) list_for_each_entry_safe(map, tmp_map, &dev->sriov.cm_list, list) { if (slave < 0 || slave == map->slave_id) { if (map->scheduled_delete) - need_flush &= !!cancel_delayed_work(&map->timeout); + need_flush |= !cancel_delayed_work(&map->timeout); } } spin_unlock(&sriov->id_map_lock); - if (!need_flush) + if (need_flush) flush_scheduled_work(); /* make sure all timers were flushed */ /* now, remove all leftover entries from databases*/ diff --git a/drivers/infiniband/hw/mlx4/cq.c b/drivers/infiniband/hw/mlx4/cq.c index 6a0fec357dae..ff931c580557 100644 --- a/drivers/infiniband/hw/mlx4/cq.c +++ b/drivers/infiniband/hw/mlx4/cq.c @@ -102,7 +102,7 @@ static int mlx4_ib_alloc_cq_buf(struct mlx4_ib_dev *dev, struct mlx4_ib_cq_buf * int err; err = mlx4_buf_alloc(dev->dev, nent * dev->dev->caps.cqe_size, - PAGE_SIZE * 2, &buf->buf, GFP_KERNEL); + PAGE_SIZE * 2, &buf->buf); if (err) goto out; @@ -113,7 +113,7 @@ static int mlx4_ib_alloc_cq_buf(struct mlx4_ib_dev *dev, struct mlx4_ib_cq_buf * if (err) goto err_buf; - err = mlx4_buf_write_mtt(dev->dev, &buf->mtt, &buf->buf, GFP_KERNEL); + err = mlx4_buf_write_mtt(dev->dev, &buf->mtt, &buf->buf); if (err) goto err_mtt; @@ -147,7 +147,7 @@ static int mlx4_ib_get_cq_umem(struct mlx4_ib_dev *dev, struct ib_ucontext *cont return PTR_ERR(*umem); err = mlx4_mtt_init(dev->dev, ib_umem_page_count(*umem), - ilog2((*umem)->page_size), &buf->mtt); + (*umem)->page_shift, &buf->mtt); if (err) goto err_buf; @@ -219,7 +219,7 @@ struct ib_cq *mlx4_ib_create_cq(struct ib_device *ibdev, uar = &to_mucontext(context)->uar; } else { - err = mlx4_db_alloc(dev->dev, &cq->db, 1, GFP_KERNEL); + err = mlx4_db_alloc(dev->dev, &cq->db, 1); if (err) goto err_cq; diff --git a/drivers/infiniband/hw/mlx4/mad.c b/drivers/infiniband/hw/mlx4/mad.c index db564ccc0f92..21d31cb1325f 100644 --- a/drivers/infiniband/hw/mlx4/mad.c +++ b/drivers/infiniband/hw/mlx4/mad.c @@ -189,25 +189,26 @@ int mlx4_MAD_IFC(struct mlx4_ib_dev *dev, int mad_ifc_flags, static void update_sm_ah(struct mlx4_ib_dev *dev, u8 port_num, u16 lid, u8 sl) { struct ib_ah *new_ah; - struct ib_ah_attr ah_attr; + struct rdma_ah_attr ah_attr; unsigned long flags; if (!dev->send_agent[port_num - 1][0]) return; memset(&ah_attr, 0, sizeof ah_attr); - ah_attr.dlid = lid; - ah_attr.sl = sl; - ah_attr.port_num = port_num; + ah_attr.type = rdma_ah_find_type(&dev->ib_dev, port_num); + rdma_ah_set_dlid(&ah_attr, lid); + rdma_ah_set_sl(&ah_attr, sl); + rdma_ah_set_port_num(&ah_attr, port_num); - new_ah = ib_create_ah(dev->send_agent[port_num - 1][0]->qp->pd, - &ah_attr); + new_ah = rdma_create_ah(dev->send_agent[port_num - 1][0]->qp->pd, + &ah_attr); if (IS_ERR(new_ah)) return; spin_lock_irqsave(&dev->sm_lock, flags); if (dev->sm_ah[port_num - 1]) - ib_destroy_ah(dev->sm_ah[port_num - 1]); + rdma_destroy_ah(dev->sm_ah[port_num - 1]); dev->sm_ah[port_num - 1] = new_ah; spin_unlock_irqrestore(&dev->sm_lock, flags); } @@ -509,7 +510,7 @@ int mlx4_ib_send_to_slave(struct mlx4_ib_dev *dev, int slave, u8 port, struct mlx4_ib_demux_pv_ctx *tun_ctx; struct mlx4_ib_demux_pv_qp *tun_qp; struct mlx4_rcv_tunnel_mad *tun_mad; - struct ib_ah_attr attr; + struct rdma_ah_attr attr; struct ib_ah *ah; struct ib_qp *src_qp = NULL; unsigned tun_tx_ix = 0; @@ -555,15 +556,18 @@ int mlx4_ib_send_to_slave(struct mlx4_ib_dev *dev, int slave, u8 port, /* create ah. Just need an empty one with the port num for the post send. * The driver will set the force loopback bit in post_send */ memset(&attr, 0, sizeof attr); - attr.port_num = port; + attr.type = rdma_ah_find_type(&dev->ib_dev, port); + + rdma_ah_set_port_num(&attr, port); if (is_eth) { union ib_gid sgid; + union ib_gid dgid; - if (get_gids_from_l3_hdr(grh, &sgid, &attr.grh.dgid)) + if (get_gids_from_l3_hdr(grh, &sgid, &dgid)) return -EINVAL; - attr.ah_flags = IB_AH_GRH; + rdma_ah_set_grh(&attr, &dgid, 0, 0, 0, 0); } - ah = ib_create_ah(tun_ctx->pd, &attr); + ah = rdma_create_ah(tun_ctx->pd, &attr); if (IS_ERR(ah)) return -ENOMEM; @@ -580,7 +584,7 @@ int mlx4_ib_send_to_slave(struct mlx4_ib_dev *dev, int slave, u8 port, tun_mad = (struct mlx4_rcv_tunnel_mad *) (tun_qp->tx_ring[tun_tx_ix].buf.addr); if (tun_qp->tx_ring[tun_tx_ix].ah) - ib_destroy_ah(tun_qp->tx_ring[tun_tx_ix].ah); + rdma_destroy_ah(tun_qp->tx_ring[tun_tx_ix].ah); tun_qp->tx_ring[tun_tx_ix].ah = ah; ib_dma_sync_single_for_cpu(&dev->ib_dev, tun_qp->tx_ring[tun_tx_ix].buf.map, @@ -653,7 +657,7 @@ int mlx4_ib_send_to_slave(struct mlx4_ib_dev *dev, int slave, u8 port, spin_unlock(&tun_qp->tx_lock); tun_qp->tx_ring[tun_tx_ix].ah = NULL; end: - ib_destroy_ah(ah); + rdma_destroy_ah(ah); return ret; } @@ -1018,7 +1022,7 @@ static void send_handler(struct ib_mad_agent *agent, struct ib_mad_send_wc *mad_send_wc) { if (mad_send_wc->send_buf->context[0]) - ib_destroy_ah(mad_send_wc->send_buf->context[0]); + rdma_destroy_ah(mad_send_wc->send_buf->context[0]); ib_free_send_mad(mad_send_wc->send_buf); } @@ -1073,7 +1077,7 @@ void mlx4_ib_mad_cleanup(struct mlx4_ib_dev *dev) } if (dev->sm_ah[p]) - ib_destroy_ah(dev->sm_ah[p]); + rdma_destroy_ah(dev->sm_ah[p]); } } @@ -1352,7 +1356,7 @@ static int is_proxy_qp0(struct mlx4_ib_dev *dev, int qpn, int slave) int mlx4_ib_send_to_wire(struct mlx4_ib_dev *dev, int slave, u8 port, enum ib_qp_type dest_qpt, u16 pkey_index, - u32 remote_qpn, u32 qkey, struct ib_ah_attr *attr, + u32 remote_qpn, u32 qkey, struct rdma_ah_attr *attr, u8 *s_mac, u16 vlan_id, struct ib_mad *mad) { struct ib_sge list; @@ -1363,6 +1367,7 @@ int mlx4_ib_send_to_wire(struct mlx4_ib_dev *dev, int slave, u8 port, struct mlx4_mad_snd_buf *sqp_mad; struct ib_ah *ah; struct ib_qp *send_qp = NULL; + struct ib_global_route *grh; unsigned wire_tx_ix = 0; int ret = 0; u16 wire_pkey_ix; @@ -1389,12 +1394,13 @@ int mlx4_ib_send_to_wire(struct mlx4_ib_dev *dev, int slave, u8 port, send_qp = sqp->qp; /* create ah */ - sgid_index = attr->grh.sgid_index; - attr->grh.sgid_index = 0; - ah = ib_create_ah(sqp_ctx->pd, attr); + grh = rdma_ah_retrieve_grh(attr); + sgid_index = grh->sgid_index; + grh->sgid_index = 0; + ah = rdma_create_ah(sqp_ctx->pd, attr); if (IS_ERR(ah)) return -ENOMEM; - attr->grh.sgid_index = sgid_index; + grh->sgid_index = sgid_index; to_mah(ah)->av.ib.gid_index = sgid_index; /* get rid of force-loopback bit */ to_mah(ah)->av.ib.port_pd &= cpu_to_be32(0x7FFFFFFF); @@ -1410,7 +1416,7 @@ int mlx4_ib_send_to_wire(struct mlx4_ib_dev *dev, int slave, u8 port, sqp_mad = (struct mlx4_mad_snd_buf *) (sqp->tx_ring[wire_tx_ix].buf.addr); if (sqp->tx_ring[wire_tx_ix].ah) - ib_destroy_ah(sqp->tx_ring[wire_tx_ix].ah); + rdma_destroy_ah(sqp->tx_ring[wire_tx_ix].ah); sqp->tx_ring[wire_tx_ix].ah = ah; ib_dma_sync_single_for_cpu(&dev->ib_dev, sqp->tx_ring[wire_tx_ix].buf.map, @@ -1442,7 +1448,7 @@ int mlx4_ib_send_to_wire(struct mlx4_ib_dev *dev, int slave, u8 port, if (s_mac) memcpy(to_mah(ah)->av.eth.s_mac, s_mac, 6); if (vlan_id < 0x1000) - vlan_id |= (attr->sl & 7) << 13; + vlan_id |= (rdma_ah_get_sl(attr) & 7) << 13; to_mah(ah)->av.eth.vlan = cpu_to_be16(vlan_id); @@ -1455,7 +1461,7 @@ int mlx4_ib_send_to_wire(struct mlx4_ib_dev *dev, int slave, u8 port, spin_unlock(&sqp->tx_lock); sqp->tx_ring[wire_tx_ix].ah = NULL; out: - ib_destroy_ah(ah); + rdma_destroy_ah(ah); return ret; } @@ -1467,12 +1473,13 @@ static int get_slave_base_gid_ix(struct mlx4_ib_dev *dev, int slave, int port) } static void fill_in_real_sgid_index(struct mlx4_ib_dev *dev, int slave, int port, - struct ib_ah_attr *ah_attr) + struct rdma_ah_attr *ah_attr) { + struct ib_global_route *grh = rdma_ah_retrieve_grh(ah_attr); if (rdma_port_get_link_layer(&dev->ib_dev, port) == IB_LINK_LAYER_INFINIBAND) - ah_attr->grh.sgid_index = slave; + grh->sgid_index = slave; else - ah_attr->grh.sgid_index += get_slave_base_gid_ix(dev, slave, port); + grh->sgid_index += get_slave_base_gid_ix(dev, slave, port); } static void mlx4_ib_multiplex_mad(struct mlx4_ib_demux_pv_ctx *ctx, struct ib_wc *wc) @@ -1482,11 +1489,13 @@ static void mlx4_ib_multiplex_mad(struct mlx4_ib_demux_pv_ctx *ctx, struct ib_wc int wr_ix = wc->wr_id & (MLX4_NUM_TUNNEL_BUFS - 1); struct mlx4_tunnel_mad *tunnel = tun_qp->ring[wr_ix].addr; struct mlx4_ib_ah ah; - struct ib_ah_attr ah_attr; + struct rdma_ah_attr ah_attr; u8 *slave_id; int slave; int port; u16 vlan_id; + u8 qos; + u8 *dmac; /* Get slave that sent this packet */ if (wc->src_qp < dev->dev->phys_caps.base_proxy_sqpn || @@ -1569,16 +1578,19 @@ static void mlx4_ib_multiplex_mad(struct mlx4_ib_demux_pv_ctx *ctx, struct ib_wc if (port < 0) return; ah.av.ib.port_pd = cpu_to_be32(port << 24 | (be32_to_cpu(ah.av.ib.port_pd) & 0xffffff)); + ah.ibah.type = rdma_ah_find_type(&dev->ib_dev, port); mlx4_ib_query_ah(&ah.ibah, &ah_attr); - if (ah_attr.ah_flags & IB_AH_GRH) + if (rdma_ah_get_ah_flags(&ah_attr) & IB_AH_GRH) fill_in_real_sgid_index(dev, slave, ctx->port, &ah_attr); - - memcpy(ah_attr.dmac, tunnel->hdr.mac, 6); + dmac = rdma_ah_retrieve_dmac(&ah_attr); + if (dmac) + memcpy(dmac, tunnel->hdr.mac, ETH_ALEN); vlan_id = be16_to_cpu(tunnel->hdr.vlan); /* if slave have default vlan use it */ - mlx4_get_slave_default_vlan(dev->dev, ctx->port, slave, - &vlan_id, &ah_attr.sl); + if (mlx4_get_slave_default_vlan(dev->dev, ctx->port, slave, + &vlan_id, &qos)) + rdma_ah_set_sl(&ah_attr, qos); mlx4_ib_send_to_wire(dev, slave, ctx->port, is_proxy_qp0(dev, wc->src_qp, slave) ? @@ -1714,7 +1726,7 @@ static void mlx4_ib_free_pv_qp_bufs(struct mlx4_ib_demux_pv_ctx *ctx, tx_buf_size, DMA_TO_DEVICE); kfree(tun_qp->tx_ring[i].buf.addr); if (tun_qp->tx_ring[i].ah) - ib_destroy_ah(tun_qp->tx_ring[i].ah); + rdma_destroy_ah(tun_qp->tx_ring[i].ah); } kfree(tun_qp->tx_ring); kfree(tun_qp->ring); @@ -1746,7 +1758,7 @@ static void mlx4_ib_tunnel_comp_worker(struct work_struct *work) pr_debug("received tunnel send completion:" "wrid=0x%llx, status=0x%x\n", wc.wr_id, wc.status); - ib_destroy_ah(tun_qp->tx_ring[wc.wr_id & + rdma_destroy_ah(tun_qp->tx_ring[wc.wr_id & (MLX4_NUM_TUNNEL_BUFS - 1)].ah); tun_qp->tx_ring[wc.wr_id & (MLX4_NUM_TUNNEL_BUFS - 1)].ah = NULL; @@ -1763,7 +1775,7 @@ static void mlx4_ib_tunnel_comp_worker(struct work_struct *work) " status = %d, wrid = 0x%llx\n", ctx->slave, wc.status, wc.wr_id); if (!MLX4_TUN_IS_RECV(wc.wr_id)) { - ib_destroy_ah(tun_qp->tx_ring[wc.wr_id & + rdma_destroy_ah(tun_qp->tx_ring[wc.wr_id & (MLX4_NUM_TUNNEL_BUFS - 1)].ah); tun_qp->tx_ring[wc.wr_id & (MLX4_NUM_TUNNEL_BUFS - 1)].ah = NULL; @@ -1900,7 +1912,7 @@ static void mlx4_ib_sqp_comp_worker(struct work_struct *work) if (wc.status == IB_WC_SUCCESS) { switch (wc.opcode) { case IB_WC_SEND: - ib_destroy_ah(sqp->tx_ring[wc.wr_id & + rdma_destroy_ah(sqp->tx_ring[wc.wr_id & (MLX4_NUM_TUNNEL_BUFS - 1)].ah); sqp->tx_ring[wc.wr_id & (MLX4_NUM_TUNNEL_BUFS - 1)].ah = NULL; @@ -1930,7 +1942,7 @@ static void mlx4_ib_sqp_comp_worker(struct work_struct *work) " status = %d, wrid = 0x%llx\n", ctx->slave, wc.status, wc.wr_id); if (!MLX4_TUN_IS_RECV(wc.wr_id)) { - ib_destroy_ah(sqp->tx_ring[wc.wr_id & + rdma_destroy_ah(sqp->tx_ring[wc.wr_id & (MLX4_NUM_TUNNEL_BUFS - 1)].ah); sqp->tx_ring[wc.wr_id & (MLX4_NUM_TUNNEL_BUFS - 1)].ah = NULL; diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c index 7031a8dd4d14..d1b43cbbfea7 100644 --- a/drivers/infiniband/hw/mlx4/main.c +++ b/drivers/infiniband/hw/mlx4/main.c @@ -39,6 +39,9 @@ #include <linux/inetdevice.h> #include <linux/rtnetlink.h> #include <linux/if_vlan.h> +#include <linux/sched/mm.h> +#include <linux/sched/task.h> + #include <net/ipv6.h> #include <net/addrconf.h> #include <net/devlink.h> @@ -58,8 +61,7 @@ #include <rdma/mlx4-abi.h> #define DRV_NAME MLX4_IB_DRV_NAME -#define DRV_VERSION "2.2-1" -#define DRV_RELDATE "Feb 2014" +#define DRV_VERSION "4.0-0" #define MLX4_IB_FLOW_MAX_PRIO 0xFFF #define MLX4_IB_FLOW_QPN_MASK 0xFFFFFF @@ -76,7 +78,7 @@ MODULE_PARM_DESC(sm_guid_assign, "Enable SM alias_GUID assignment if sm_guid_ass static const char mlx4_ib_version[] = DRV_NAME ": Mellanox ConnectX InfiniBand driver v" - DRV_VERSION " (" DRV_RELDATE ")\n"; + DRV_VERSION "\n"; static void do_slave_init(struct mlx4_ib_dev *ibdev, int slave, int do_init); @@ -678,7 +680,7 @@ static u8 state_to_phys_state(enum ib_port_state state) } static int eth_link_query_port(struct ib_device *ibdev, u8 port, - struct ib_port_attr *props, int netw_view) + struct ib_port_attr *props) { struct mlx4_ib_dev *mdev = to_mdev(ibdev); @@ -741,11 +743,11 @@ int __mlx4_ib_query_port(struct ib_device *ibdev, u8 port, { int err; - memset(props, 0, sizeof *props); + /* props being zeroed by the caller, avoid zeroing it here */ err = mlx4_ib_port_link_layer(ibdev, port) == IB_LINK_LAYER_INFINIBAND ? ib_link_query_port(ibdev, port, props, netw_view) : - eth_link_query_port(ibdev, port, props, netw_view); + eth_link_query_port(ibdev, port, props); return err; } @@ -1014,7 +1016,7 @@ static int mlx4_ib_modify_port(struct ib_device *ibdev, u8 port, int mask, mutex_lock(&mdev->cap_mask_mutex); - err = mlx4_ib_query_port(ibdev, port, &attr); + err = ib_query_port(ibdev, port, &attr); if (err) goto out; @@ -1153,7 +1155,7 @@ static void mlx4_ib_disassociate_ucontext(struct ib_ucontext *ibcontext) * call to mlx4_ib_vma_close. */ put_task_struct(owning_process); - msleep(1); + usleep_range(1000, 2000); owning_process = get_pid_task(ibcontext->tgid, PIDTYPE_PID); if (!owning_process || @@ -1170,7 +1172,7 @@ static void mlx4_ib_disassociate_ucontext(struct ib_ucontext *ibcontext) /* need to protect from a race on closing the vma as part of * mlx4_ib_vma_close(). */ - down_read(&owning_mm->mmap_sem); + down_write(&owning_mm->mmap_sem); for (i = 0; i < HW_BAR_COUNT; i++) { vma = context->hw_bar_info[i].vma; if (!vma) @@ -1184,11 +1186,13 @@ static void mlx4_ib_disassociate_ucontext(struct ib_ucontext *ibcontext) BUG_ON(1); } + context->hw_bar_info[i].vma->vm_flags &= + ~(VM_SHARED | VM_MAYSHARE); /* context going to be destroyed, should not access ops any more */ context->hw_bar_info[i].vma->vm_ops = NULL; } - up_read(&owning_mm->mmap_sem); + up_write(&owning_mm->mmap_sem); mmput(owning_mm); put_task_struct(owning_process); } @@ -2537,24 +2541,27 @@ static int mlx4_port_immutable(struct ib_device *ibdev, u8 port_num, struct mlx4_ib_dev *mdev = to_mdev(ibdev); int err; - err = mlx4_ib_query_port(ibdev, port_num, &attr); - if (err) - return err; - - immutable->pkey_tbl_len = attr.pkey_tbl_len; - immutable->gid_tbl_len = attr.gid_tbl_len; - if (mlx4_ib_port_link_layer(ibdev, port_num) == IB_LINK_LAYER_INFINIBAND) { immutable->core_cap_flags = RDMA_CORE_PORT_IBA_IB; + immutable->max_mad_size = IB_MGMT_MAD_SIZE; } else { if (mdev->dev->caps.flags & MLX4_DEV_CAP_FLAG_IBOE) immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE; if (mdev->dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_ROCE_V1_V2) immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE | RDMA_CORE_PORT_IBA_ROCE_UDP_ENCAP; + immutable->core_cap_flags |= RDMA_CORE_PORT_RAW_PACKET; + if (immutable->core_cap_flags & (RDMA_CORE_PORT_IBA_ROCE | + RDMA_CORE_PORT_IBA_ROCE_UDP_ENCAP)) + immutable->max_mad_size = IB_MGMT_MAD_SIZE; } - immutable->max_mad_size = IB_MGMT_MAD_SIZE; + err = ib_query_port(ibdev, port_num, &attr); + if (err) + return err; + + immutable->pkey_tbl_len = attr.pkey_tbl_len; + immutable->gid_tbl_len = attr.gid_tbl_len; return 0; } @@ -2625,7 +2632,7 @@ static void *mlx4_ib_add(struct mlx4_dev *dev) ibdev->ib_dev.phys_port_cnt = mlx4_is_bonded(dev) ? 1 : ibdev->num_ports; ibdev->ib_dev.num_comp_vectors = dev->caps.num_comp_vectors; - ibdev->ib_dev.dma_device = &dev->persist->pdev->dev; + ibdev->ib_dev.dev.parent = &dev->persist->pdev->dev; ibdev->ib_dev.get_netdev = mlx4_ib_get_netdev; ibdev->ib_dev.add_gid = mlx4_ib_add_gid; ibdev->ib_dev.del_gid = mlx4_ib_del_gid; @@ -2861,23 +2868,19 @@ static void *mlx4_ib_add(struct mlx4_dev *dev) if (mlx4_ib_init_sriov(ibdev)) goto err_mad; - if (dev->caps.flags & MLX4_DEV_CAP_FLAG_IBOE || - dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_ROCE_V1_V2) { - if (!iboe->nb.notifier_call) { - iboe->nb.notifier_call = mlx4_ib_netdev_event; - err = register_netdevice_notifier(&iboe->nb); - if (err) { - iboe->nb.notifier_call = NULL; - goto err_notif; - } - } - if (dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_ROCE_V1_V2) { - err = mlx4_config_roce_v2_port(dev, ROCE_V2_UDP_DPORT); - if (err) { - goto err_notif; - } + if (!iboe->nb.notifier_call) { + iboe->nb.notifier_call = mlx4_ib_netdev_event; + err = register_netdevice_notifier(&iboe->nb); + if (err) { + iboe->nb.notifier_call = NULL; + goto err_notif; } } + if (dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_ROCE_V1_V2) { + err = mlx4_config_roce_v2_port(dev, ROCE_V2_UDP_DPORT); + if (err) + goto err_notif; + } for (j = 0; j < ARRAY_SIZE(mlx4_class_attributes); ++j) { if (device_create_file(&ibdev->ib_dev.dev, @@ -2935,6 +2938,7 @@ err_counter: mlx4_ib_delete_counters_table(ibdev, &ibdev->counters_table[i]); err_map: + mlx4_ib_free_eqs(dev, ibdev); iounmap(ibdev->uar_map); err_uar: diff --git a/drivers/infiniband/hw/mlx4/mcg.c b/drivers/infiniband/hw/mlx4/mcg.c index e010fe459e67..b73f89700ef9 100644 --- a/drivers/infiniband/hw/mlx4/mcg.c +++ b/drivers/infiniband/hw/mlx4/mcg.c @@ -209,7 +209,7 @@ static struct mcast_group *mcast_insert(struct mlx4_ib_demux_ctx *ctx, static int send_mad_to_wire(struct mlx4_ib_demux_ctx *ctx, struct ib_mad *mad) { struct mlx4_ib_dev *dev = ctx->dev; - struct ib_ah_attr ah_attr; + struct rdma_ah_attr ah_attr; unsigned long flags; spin_lock_irqsave(&dev->sm_lock, flags); @@ -231,20 +231,20 @@ static int send_mad_to_slave(int slave, struct mlx4_ib_demux_ctx *ctx, struct mlx4_ib_dev *dev = ctx->dev; struct ib_mad_agent *agent = dev->send_agent[ctx->port - 1][1]; struct ib_wc wc; - struct ib_ah_attr ah_attr; + struct rdma_ah_attr ah_attr; /* Our agent might not yet be registered when mads start to arrive */ if (!agent) return -EAGAIN; - ib_query_ah(dev->sm_ah[ctx->port - 1], &ah_attr); + rdma_query_ah(dev->sm_ah[ctx->port - 1], &ah_attr); if (ib_find_cached_pkey(&dev->ib_dev, ctx->port, IB_DEFAULT_PKEY_FULL, &wc.pkey_index)) return -EINVAL; wc.sl = 0; wc.dlid_path_bits = 0; wc.port_num = ctx->port; - wc.slid = ah_attr.dlid; /* opensm lid */ + wc.slid = rdma_ah_get_dlid(&ah_attr); /* opensm lid */ wc.src_qp = 1; return mlx4_ib_send_to_slave(dev, slave, ctx->port, IB_QPT_GSI, &wc, NULL, mad); } @@ -1091,7 +1091,7 @@ static void _mlx4_ib_mcg_port_cleanup(struct mlx4_ib_demux_ctx *ctx, int destroy if (!count) break; - msleep(1); + usleep_range(1000, 2000); } while (time_after(end, jiffies)); flush_workqueue(ctx->mcg_wq); @@ -1102,7 +1102,8 @@ static void _mlx4_ib_mcg_port_cleanup(struct mlx4_ib_demux_ctx *ctx, int destroy while ((p = rb_first(&ctx->mcg_table)) != NULL) { group = rb_entry(p, struct mcast_group, node); if (atomic_read(&group->refcount)) - mcg_warn_group(group, "group refcount %d!!! (pointer %p)\n", atomic_read(&group->refcount), group); + mcg_debug_group(group, "group refcount %d!!! (pointer %p)\n", + atomic_read(&group->refcount), group); force_clean_group(group); } diff --git a/drivers/infiniband/hw/mlx4/mlx4_ib.h b/drivers/infiniband/hw/mlx4/mlx4_ib.h index 7f3d976d81ed..9db82e67e959 100644 --- a/drivers/infiniband/hw/mlx4/mlx4_ib.h +++ b/drivers/infiniband/hw/mlx4/mlx4_ib.h @@ -55,7 +55,7 @@ #define pr_fmt(fmt) "<" MLX4_IB_DRV_NAME "> %s: " fmt, __func__ #define mlx4_ib_warn(ibdev, format, arg...) \ - dev_warn((ibdev)->dma_device, MLX4_IB_DRV_NAME ": " format, ## arg) + dev_warn((ibdev)->dev.parent, MLX4_IB_DRV_NAME ": " format, ## arg) enum { MLX4_IB_SQ_MIN_WQE_SHIFT = 6, @@ -185,7 +185,6 @@ enum mlx4_ib_qp_flags { MLX4_IB_QP_LSO = IB_QP_CREATE_IPOIB_UD_LSO, MLX4_IB_QP_BLOCK_MULTICAST_LOOPBACK = IB_QP_CREATE_BLOCK_MULTICAST_LOOPBACK, MLX4_IB_QP_NETIF = IB_QP_CREATE_NETIF_QP, - MLX4_IB_QP_CREATE_USE_GFP_NOIO = IB_QP_CREATE_USE_GFP_NOIO, /* Mellanox specific flags start from IB_QP_CREATE_RESERVED_START */ MLX4_IB_ROCE_V2_GSI_QP = MLX4_IB_QP_CREATE_ROCE_V2_GSI, @@ -742,9 +741,9 @@ int mlx4_ib_arm_cq(struct ib_cq *cq, enum ib_cq_notify_flags flags); void __mlx4_ib_cq_clean(struct mlx4_ib_cq *cq, u32 qpn, struct mlx4_ib_srq *srq); void mlx4_ib_cq_clean(struct mlx4_ib_cq *cq, u32 qpn, struct mlx4_ib_srq *srq); -struct ib_ah *mlx4_ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr, +struct ib_ah *mlx4_ib_create_ah(struct ib_pd *pd, struct rdma_ah_attr *ah_attr, struct ib_udata *udata); -int mlx4_ib_query_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr); +int mlx4_ib_query_ah(struct ib_ah *ibah, struct rdma_ah_attr *ah_attr); int mlx4_ib_destroy_ah(struct ib_ah *ah); struct ib_srq *mlx4_ib_create_srq(struct ib_pd *pd, @@ -833,7 +832,7 @@ int mlx4_ib_send_to_slave(struct mlx4_ib_dev *dev, int slave, u8 port, int mlx4_ib_send_to_wire(struct mlx4_ib_dev *dev, int slave, u8 port, enum ib_qp_type dest_qpt, u16 pkey_index, u32 remote_qpn, - u32 qkey, struct ib_ah_attr *attr, u8 *s_mac, + u32 qkey, struct rdma_ah_attr *attr, u8 *s_mac, u16 vlan_id, struct ib_mad *mad); __be64 mlx4_ib_get_new_demux_tid(struct mlx4_ib_demux_ctx *ctx); diff --git a/drivers/infiniband/hw/mlx4/mr.c b/drivers/infiniband/hw/mlx4/mr.c index 5d73989d9771..e6f77f63da75 100644 --- a/drivers/infiniband/hw/mlx4/mr.c +++ b/drivers/infiniband/hw/mlx4/mr.c @@ -107,7 +107,7 @@ int mlx4_ib_umem_write_mtt(struct mlx4_ib_dev *dev, struct mlx4_mtt *mtt, len = sg_dma_len(sg) >> mtt->page_shift; for (k = 0; k < len; ++k) { pages[i++] = sg_dma_address(sg) + - umem->page_size * k; + (k << umem->page_shift); /* * Be friendly to mlx4_write_mtt() and * pass it chunks of appropriate size. @@ -155,7 +155,7 @@ struct ib_mr *mlx4_ib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, } n = ib_umem_page_count(mr->umem); - shift = ilog2(mr->umem->page_size); + shift = mr->umem->page_shift; err = mlx4_mr_alloc(dev->dev, to_mpd(pd)->pdn, virt_addr, length, convert_access(access_flags), n, shift, &mr->mmr); @@ -239,7 +239,7 @@ int mlx4_ib_rereg_user_mr(struct ib_mr *mr, int flags, goto release_mpt_entry; } n = ib_umem_page_count(mmr->umem); - shift = ilog2(mmr->umem->page_size); + shift = mmr->umem->page_shift; err = mlx4_mr_rereg_mem_write(dev->dev, &mmr->mmr, virt_addr, length, n, shift, @@ -292,10 +292,10 @@ mlx4_alloc_priv_pages(struct ib_device *device, if (!mr->pages) return -ENOMEM; - mr->page_map = dma_map_single(device->dma_device, mr->pages, + mr->page_map = dma_map_single(device->dev.parent, mr->pages, mr->page_map_size, DMA_TO_DEVICE); - if (dma_mapping_error(device->dma_device, mr->page_map)) { + if (dma_mapping_error(device->dev.parent, mr->page_map)) { ret = -ENOMEM; goto err; } @@ -313,7 +313,7 @@ mlx4_free_priv_pages(struct mlx4_ib_mr *mr) if (mr->pages) { struct ib_device *device = mr->ibmr.device; - dma_unmap_single(device->dma_device, mr->page_map, + dma_unmap_single(device->dev.parent, mr->page_map, mr->page_map_size, DMA_TO_DEVICE); free_page((unsigned long)mr->pages); mr->pages = NULL; diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c index c068add8838b..75c0e6c5dd56 100644 --- a/drivers/infiniband/hw/mlx4/qp.c +++ b/drivers/infiniband/hw/mlx4/qp.c @@ -76,10 +76,6 @@ enum { MLX4_IB_LSO_HEADER_SPARE = 128, }; -enum { - MLX4_IB_IBOE_ETHERTYPE = 0x8915 -}; - struct mlx4_ib_sqp { struct mlx4_ib_qp qp; int pkey_index; @@ -638,8 +634,8 @@ static void mlx4_ib_free_qp_counter(struct mlx4_ib_dev *dev, static int create_qp_common(struct mlx4_ib_dev *dev, struct ib_pd *pd, struct ib_qp_init_attr *init_attr, - struct ib_udata *udata, int sqpn, struct mlx4_ib_qp **caller_qp, - gfp_t gfp) + struct ib_udata *udata, int sqpn, + struct mlx4_ib_qp **caller_qp) { int qpn; int err; @@ -695,14 +691,14 @@ static int create_qp_common(struct mlx4_ib_dev *dev, struct ib_pd *pd, if (qp_type == MLX4_IB_QPT_SMI || qp_type == MLX4_IB_QPT_GSI || (qp_type & (MLX4_IB_QPT_PROXY_SMI | MLX4_IB_QPT_PROXY_SMI_OWNER | MLX4_IB_QPT_PROXY_GSI | MLX4_IB_QPT_TUN_SMI_OWNER))) { - sqp = kzalloc(sizeof (struct mlx4_ib_sqp), gfp); + sqp = kzalloc(sizeof(struct mlx4_ib_sqp), GFP_KERNEL); if (!sqp) return -ENOMEM; qp = &sqp->qp; qp->pri.vid = 0xFFFF; qp->alt.vid = 0xFFFF; } else { - qp = kzalloc(sizeof (struct mlx4_ib_qp), gfp); + qp = kzalloc(sizeof(struct mlx4_ib_qp), GFP_KERNEL); if (!qp) return -ENOMEM; qp->pri.vid = 0xFFFF; @@ -749,7 +745,7 @@ static int create_qp_common(struct mlx4_ib_dev *dev, struct ib_pd *pd, } err = mlx4_mtt_init(dev->dev, ib_umem_page_count(qp->umem), - ilog2(qp->umem->page_size), &qp->mtt); + qp->umem->page_shift, &qp->mtt); if (err) goto err_buf; @@ -784,7 +780,7 @@ static int create_qp_common(struct mlx4_ib_dev *dev, struct ib_pd *pd, goto err; if (qp_has_rq(init_attr)) { - err = mlx4_db_alloc(dev->dev, &qp->db, 0, gfp); + err = mlx4_db_alloc(dev->dev, &qp->db, 0); if (err) goto err; @@ -792,7 +788,7 @@ static int create_qp_common(struct mlx4_ib_dev *dev, struct ib_pd *pd, } if (mlx4_buf_alloc(dev->dev, qp->buf_size, qp->buf_size, - &qp->buf, gfp)) { + &qp->buf)) { memcpy(&init_attr->cap, &backup_cap, sizeof(backup_cap)); err = set_kernel_sq_size(dev, &init_attr->cap, qp_type, @@ -801,7 +797,7 @@ static int create_qp_common(struct mlx4_ib_dev *dev, struct ib_pd *pd, goto err_db; if (mlx4_buf_alloc(dev->dev, qp->buf_size, - PAGE_SIZE * 2, &qp->buf, gfp)) { + PAGE_SIZE * 2, &qp->buf)) { err = -ENOMEM; goto err_db; } @@ -812,20 +808,20 @@ static int create_qp_common(struct mlx4_ib_dev *dev, struct ib_pd *pd, if (err) goto err_buf; - err = mlx4_buf_write_mtt(dev->dev, &qp->mtt, &qp->buf, gfp); + err = mlx4_buf_write_mtt(dev->dev, &qp->mtt, &qp->buf); if (err) goto err_mtt; qp->sq.wrid = kmalloc_array(qp->sq.wqe_cnt, sizeof(u64), - gfp | __GFP_NOWARN); + GFP_KERNEL | __GFP_NOWARN); if (!qp->sq.wrid) qp->sq.wrid = __vmalloc(qp->sq.wqe_cnt * sizeof(u64), - gfp, PAGE_KERNEL); + GFP_KERNEL, PAGE_KERNEL); qp->rq.wrid = kmalloc_array(qp->rq.wqe_cnt, sizeof(u64), - gfp | __GFP_NOWARN); + GFP_KERNEL | __GFP_NOWARN); if (!qp->rq.wrid) qp->rq.wrid = __vmalloc(qp->rq.wqe_cnt * sizeof(u64), - gfp, PAGE_KERNEL); + GFP_KERNEL, PAGE_KERNEL); if (!qp->sq.wrid || !qp->rq.wrid) { err = -ENOMEM; goto err_wrid; @@ -863,7 +859,7 @@ static int create_qp_common(struct mlx4_ib_dev *dev, struct ib_pd *pd, if (init_attr->create_flags & IB_QP_CREATE_BLOCK_MULTICAST_LOOPBACK) qp->flags |= MLX4_IB_QP_BLOCK_MULTICAST_LOOPBACK; - err = mlx4_qp_alloc(dev->dev, qpn, &qp->mqp, gfp); + err = mlx4_qp_alloc(dev->dev, qpn, &qp->mqp); if (err) goto err_qpn; @@ -1131,10 +1127,7 @@ static struct ib_qp *_mlx4_ib_create_qp(struct ib_pd *pd, int err; int sup_u_create_flags = MLX4_IB_QP_BLOCK_MULTICAST_LOOPBACK; u16 xrcdn = 0; - gfp_t gfp; - gfp = (init_attr->create_flags & MLX4_IB_QP_CREATE_USE_GFP_NOIO) ? - GFP_NOIO : GFP_KERNEL; /* * We only support LSO, vendor flag1, and multicast loopback blocking, * and only for kernel UD QPs. @@ -1144,8 +1137,7 @@ static struct ib_qp *_mlx4_ib_create_qp(struct ib_pd *pd, MLX4_IB_SRIOV_TUNNEL_QP | MLX4_IB_SRIOV_SQP | MLX4_IB_QP_NETIF | - MLX4_IB_QP_CREATE_ROCE_V2_GSI | - MLX4_IB_QP_CREATE_USE_GFP_NOIO)) + MLX4_IB_QP_CREATE_ROCE_V2_GSI)) return ERR_PTR(-EINVAL); if (init_attr->create_flags & IB_QP_CREATE_NETIF_QP) { @@ -1158,7 +1150,6 @@ static struct ib_qp *_mlx4_ib_create_qp(struct ib_pd *pd, return ERR_PTR(-EINVAL); if ((init_attr->create_flags & ~(MLX4_IB_SRIOV_SQP | - MLX4_IB_QP_CREATE_USE_GFP_NOIO | MLX4_IB_QP_CREATE_ROCE_V2_GSI | MLX4_IB_QP_BLOCK_MULTICAST_LOOPBACK) && init_attr->qp_type != IB_QPT_UD) || @@ -1183,7 +1174,7 @@ static struct ib_qp *_mlx4_ib_create_qp(struct ib_pd *pd, case IB_QPT_RC: case IB_QPT_UC: case IB_QPT_RAW_PACKET: - qp = kzalloc(sizeof *qp, gfp); + qp = kzalloc(sizeof(*qp), GFP_KERNEL); if (!qp) return ERR_PTR(-ENOMEM); qp->pri.vid = 0xFFFF; @@ -1192,7 +1183,7 @@ static struct ib_qp *_mlx4_ib_create_qp(struct ib_pd *pd, case IB_QPT_UD: { err = create_qp_common(to_mdev(pd->device), pd, init_attr, - udata, 0, &qp, gfp); + udata, 0, &qp); if (err) { kfree(qp); return ERR_PTR(err); @@ -1221,8 +1212,7 @@ static struct ib_qp *_mlx4_ib_create_qp(struct ib_pd *pd, } err = create_qp_common(to_mdev(pd->device), pd, init_attr, udata, - sqpn, - &qp, gfp); + sqpn, &qp); if (err) return ERR_PTR(err); @@ -1387,31 +1377,31 @@ static void mlx4_set_sched(struct mlx4_qp_path *path, u8 port) path->sched_queue = (path->sched_queue & 0xbf) | ((port - 1) << 6); } -static int _mlx4_set_path(struct mlx4_ib_dev *dev, const struct ib_ah_attr *ah, +static int _mlx4_set_path(struct mlx4_ib_dev *dev, + const struct rdma_ah_attr *ah, u64 smac, u16 vlan_tag, struct mlx4_qp_path *path, struct mlx4_roce_smac_vlan_info *smac_info, u8 port) { - int is_eth = rdma_port_get_link_layer(&dev->ib_dev, port) == - IB_LINK_LAYER_ETHERNET; int vidx; int smac_index; int err; - - path->grh_mylmc = ah->src_path_bits & 0x7f; - path->rlid = cpu_to_be16(ah->dlid); - if (ah->static_rate) { - path->static_rate = ah->static_rate + MLX4_STAT_RATE_OFFSET; + path->grh_mylmc = rdma_ah_get_path_bits(ah) & 0x7f; + path->rlid = cpu_to_be16(rdma_ah_get_dlid(ah)); + if (rdma_ah_get_static_rate(ah)) { + path->static_rate = rdma_ah_get_static_rate(ah) + + MLX4_STAT_RATE_OFFSET; while (path->static_rate > IB_RATE_2_5_GBPS + MLX4_STAT_RATE_OFFSET && !(1 << path->static_rate & dev->dev->caps.stat_rate_support)) --path->static_rate; } else path->static_rate = 0; - if (ah->ah_flags & IB_AH_GRH) { - int real_sgid_index = mlx4_ib_gid_index_to_real_index(dev, - port, - ah->grh.sgid_index); + if (rdma_ah_get_ah_flags(ah) & IB_AH_GRH) { + const struct ib_global_route *grh = rdma_ah_read_grh(ah); + int real_sgid_index = + mlx4_ib_gid_index_to_real_index(dev, port, + grh->sgid_index); if (real_sgid_index >= dev->dev->caps.gid_table_len[port]) { pr_err("sgid_index (%u) too large. max is %d\n", @@ -1421,19 +1411,19 @@ static int _mlx4_set_path(struct mlx4_ib_dev *dev, const struct ib_ah_attr *ah, path->grh_mylmc |= 1 << 7; path->mgid_index = real_sgid_index; - path->hop_limit = ah->grh.hop_limit; + path->hop_limit = grh->hop_limit; path->tclass_flowlabel = - cpu_to_be32((ah->grh.traffic_class << 20) | - (ah->grh.flow_label)); - memcpy(path->rgid, ah->grh.dgid.raw, 16); + cpu_to_be32((grh->traffic_class << 20) | + (grh->flow_label)); + memcpy(path->rgid, grh->dgid.raw, 16); } - if (is_eth) { - if (!(ah->ah_flags & IB_AH_GRH)) + if (ah->type == RDMA_AH_ATTR_TYPE_ROCE) { + if (!(rdma_ah_get_ah_flags(ah) & IB_AH_GRH)) return -1; path->sched_queue = MLX4_IB_DEFAULT_SCHED_QUEUE | - ((port - 1) << 6) | ((ah->sl & 7) << 3); + ((port - 1) << 6) | ((rdma_ah_get_sl(ah) & 7) << 3); path->feup |= MLX4_FEUP_FORCE_ETH_UP; if (vlan_tag < 0x1000) { @@ -1492,14 +1482,13 @@ static int _mlx4_set_path(struct mlx4_ib_dev *dev, const struct ib_ah_attr *ah, } else { smac_index = smac_info->smac_index; } - - memcpy(path->dmac, ah->dmac, 6); + memcpy(path->dmac, ah->roce.dmac, 6); path->ackto = MLX4_IB_LINK_TYPE_ETH; /* put MAC table smac index for IBoE */ path->grh_mylmc = (u8) (smac_index) | 0x80; } else { path->sched_queue = MLX4_IB_DEFAULT_SCHED_QUEUE | - ((port - 1) << 6) | ((ah->sl & 0xf) << 2); + ((port - 1) << 6) | ((rdma_ah_get_sl(ah) & 0xf) << 2); } return 0; @@ -1771,11 +1760,13 @@ static int __mlx4_ib_modify_qp(struct ib_qp *ibqp, u16 vlan = 0xffff; u8 smac[ETH_ALEN]; int status = 0; - int is_eth = rdma_cap_eth_ah(&dev->ib_dev, port_num) && - attr->ah_attr.ah_flags & IB_AH_GRH; + int is_eth = + rdma_cap_eth_ah(&dev->ib_dev, port_num) && + rdma_ah_get_ah_flags(&attr->ah_attr) & IB_AH_GRH; - if (is_eth && attr->ah_attr.ah_flags & IB_AH_GRH) { - int index = attr->ah_attr.grh.sgid_index; + if (is_eth) { + int index = + rdma_ah_read_grh(&attr->ah_attr)->sgid_index; status = ib_get_cached_gid(ibqp->device, port_num, index, &gid, &gid_attr); @@ -2424,11 +2415,31 @@ static u8 sl_to_vl(struct mlx4_ib_dev *dev, u8 sl, int port_num) return vl; } +static int fill_gid_by_hw_index(struct mlx4_ib_dev *ibdev, u8 port_num, + int index, union ib_gid *gid, + enum ib_gid_type *gid_type) +{ + struct mlx4_ib_iboe *iboe = &ibdev->iboe; + struct mlx4_port_gid_table *port_gid_table; + unsigned long flags; + + port_gid_table = &iboe->gids[port_num - 1]; + spin_lock_irqsave(&iboe->lock, flags); + memcpy(gid, &port_gid_table->gids[index].gid, sizeof(*gid)); + *gid_type = port_gid_table->gids[index].gid_type; + spin_unlock_irqrestore(&iboe->lock, flags); + if (!memcmp(gid, &zgid, sizeof(*gid))) + return -ENOENT; + + return 0; +} + #define MLX4_ROCEV2_QP1_SPORT 0xC000 static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_ud_wr *wr, void *wqe, unsigned *mlx_seg_len) { struct ib_device *ib_dev = sqp->qp.ibqp.device; + struct mlx4_ib_dev *ibdev = to_mdev(ib_dev); struct mlx4_wqe_mlx_seg *mlx = wqe; struct mlx4_wqe_ctrl_seg *ctrl = wqe; struct mlx4_wqe_inline_seg *inl = wqe + sizeof *mlx; @@ -2454,8 +2465,7 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_ud_wr *wr, is_eth = rdma_port_get_link_layer(sqp->qp.ibqp.device, sqp->qp.port) == IB_LINK_LAYER_ETHERNET; is_grh = mlx4_ib_ah_grh_present(ah); if (is_eth) { - struct ib_gid_attr gid_attr; - + enum ib_gid_type gid_type; if (mlx4_is_mfunc(to_mdev(ib_dev)->dev)) { /* When multi-function is enabled, the ib_core gid * indexes don't necessarily match the hw ones, so @@ -2466,18 +2476,11 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_ud_wr *wr, if (err) return err; } else { - err = ib_get_cached_gid(ib_dev, - be32_to_cpu(ah->av.ib.port_pd) >> 24, - ah->av.ib.gid_index, &sgid, - &gid_attr); - if (!err) { - if (gid_attr.ndev) - dev_put(gid_attr.ndev); - if (!memcmp(&sgid, &zgid, sizeof(sgid))) - err = -ENOENT; - } + err = fill_gid_by_hw_index(ibdev, sqp->qp.port, + ah->av.ib.gid_index, + &sgid, &gid_type); if (!err) { - is_udp = gid_attr.gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP; + is_udp = gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP; if (is_udp) { if (ipv6_addr_v4mapped((struct in6_addr *)&sgid)) ip_version = 4; @@ -2588,7 +2591,7 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_ud_wr *wr, u16 ether_type; u16 pcp = (be32_to_cpu(ah->av.ib.sl_tclass_flowlabel) >> 29) << 13; - ether_type = (!is_udp) ? MLX4_IB_IBOE_ETHERTYPE : + ether_type = (!is_udp) ? ETH_P_IBOE: (ip_version == 4 ? ETH_P_IP : ETH_P_IPV6); mlx->sched_prio = cpu_to_be16(pcp); @@ -2955,21 +2958,17 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, if (sqp->roce_v2_gsi) { struct mlx4_ib_ah *ah = to_mah(ud_wr(wr)->ah); - struct ib_gid_attr gid_attr; + enum ib_gid_type gid_type; union ib_gid gid; - if (!ib_get_cached_gid(ibqp->device, - be32_to_cpu(ah->av.ib.port_pd) >> 24, - ah->av.ib.gid_index, &gid, - &gid_attr)) { - if (gid_attr.ndev) - dev_put(gid_attr.ndev); - qp = (gid_attr.gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP) ? - to_mqp(sqp->roce_v2_gsi) : qp; - } else { + if (!fill_gid_by_hw_index(mdev, sqp->qp.port, + ah->av.ib.gid_index, + &gid, &gid_type)) + qp = (gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP) ? + to_mqp(sqp->roce_v2_gsi) : qp; + else pr_err("Failed to get gid at index %d. RoCEv2 will not work properly\n", ah->av.ib.gid_index); - } } } @@ -3390,39 +3389,37 @@ static int to_ib_qp_access_flags(int mlx4_flags) return ib_flags; } -static void to_ib_ah_attr(struct mlx4_ib_dev *ibdev, struct ib_ah_attr *ib_ah_attr, - struct mlx4_qp_path *path) +static void to_rdma_ah_attr(struct mlx4_ib_dev *ibdev, + struct rdma_ah_attr *ah_attr, + struct mlx4_qp_path *path) { struct mlx4_dev *dev = ibdev->dev; - int is_eth; - - memset(ib_ah_attr, 0, sizeof *ib_ah_attr); - ib_ah_attr->port_num = path->sched_queue & 0x40 ? 2 : 1; + u8 port_num = path->sched_queue & 0x40 ? 2 : 1; - if (ib_ah_attr->port_num == 0 || ib_ah_attr->port_num > dev->caps.num_ports) + memset(ah_attr, 0, sizeof(*ah_attr)); + ah_attr->type = rdma_ah_find_type(&ibdev->ib_dev, port_num); + if (port_num == 0 || port_num > dev->caps.num_ports) return; - is_eth = rdma_port_get_link_layer(&ibdev->ib_dev, ib_ah_attr->port_num) == - IB_LINK_LAYER_ETHERNET; - if (is_eth) - ib_ah_attr->sl = ((path->sched_queue >> 3) & 0x7) | - ((path->sched_queue & 4) << 1); + if (ah_attr->type == RDMA_AH_ATTR_TYPE_ROCE) + rdma_ah_set_sl(ah_attr, ((path->sched_queue >> 3) & 0x7) | + ((path->sched_queue & 4) << 1)); else - ib_ah_attr->sl = (path->sched_queue >> 2) & 0xf; + rdma_ah_set_sl(ah_attr, (path->sched_queue >> 2) & 0xf); + rdma_ah_set_port_num(ah_attr, port_num); - ib_ah_attr->dlid = be16_to_cpu(path->rlid); - ib_ah_attr->src_path_bits = path->grh_mylmc & 0x7f; - ib_ah_attr->static_rate = path->static_rate ? path->static_rate - 5 : 0; - ib_ah_attr->ah_flags = (path->grh_mylmc & (1 << 7)) ? IB_AH_GRH : 0; - if (ib_ah_attr->ah_flags) { - ib_ah_attr->grh.sgid_index = path->mgid_index; - ib_ah_attr->grh.hop_limit = path->hop_limit; - ib_ah_attr->grh.traffic_class = - (be32_to_cpu(path->tclass_flowlabel) >> 20) & 0xff; - ib_ah_attr->grh.flow_label = - be32_to_cpu(path->tclass_flowlabel) & 0xfffff; - memcpy(ib_ah_attr->grh.dgid.raw, - path->rgid, sizeof ib_ah_attr->grh.dgid.raw); + rdma_ah_set_dlid(ah_attr, be16_to_cpu(path->rlid)); + rdma_ah_set_path_bits(ah_attr, path->grh_mylmc & 0x7f); + rdma_ah_set_static_rate(ah_attr, + path->static_rate ? path->static_rate - 5 : 0); + if (path->grh_mylmc & (1 << 7)) { + rdma_ah_set_grh(ah_attr, NULL, + be32_to_cpu(path->tclass_flowlabel) & 0xfffff, + path->mgid_index, + path->hop_limit, + (be32_to_cpu(path->tclass_flowlabel) + >> 20) & 0xff); + rdma_ah_set_dgid_raw(ah_attr, path->rgid); } } @@ -3463,10 +3460,11 @@ int mlx4_ib_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr, int qp_attr to_ib_qp_access_flags(be32_to_cpu(context.params2)); if (qp->ibqp.qp_type == IB_QPT_RC || qp->ibqp.qp_type == IB_QPT_UC) { - to_ib_ah_attr(dev, &qp_attr->ah_attr, &context.pri_path); - to_ib_ah_attr(dev, &qp_attr->alt_ah_attr, &context.alt_path); + to_rdma_ah_attr(dev, &qp_attr->ah_attr, &context.pri_path); + to_rdma_ah_attr(dev, &qp_attr->alt_ah_attr, &context.alt_path); qp_attr->alt_pkey_index = context.alt_path.pkey_index & 0x7f; - qp_attr->alt_port_num = qp_attr->alt_ah_attr.port_num; + qp_attr->alt_port_num = + rdma_ah_get_port_num(&qp_attr->alt_ah_attr); } qp_attr->pkey_index = context.pri_path.pkey_index & 0x7f; diff --git a/drivers/infiniband/hw/mlx4/srq.c b/drivers/infiniband/hw/mlx4/srq.c index 7dd3f267f06b..0facaf5f6d23 100644 --- a/drivers/infiniband/hw/mlx4/srq.c +++ b/drivers/infiniband/hw/mlx4/srq.c @@ -122,7 +122,7 @@ struct ib_srq *mlx4_ib_create_srq(struct ib_pd *pd, } err = mlx4_mtt_init(dev->dev, ib_umem_page_count(srq->umem), - ilog2(srq->umem->page_size), &srq->mtt); + srq->umem->page_shift, &srq->mtt); if (err) goto err_buf; @@ -135,14 +135,14 @@ struct ib_srq *mlx4_ib_create_srq(struct ib_pd *pd, if (err) goto err_mtt; } else { - err = mlx4_db_alloc(dev->dev, &srq->db, 0, GFP_KERNEL); + err = mlx4_db_alloc(dev->dev, &srq->db, 0); if (err) goto err_srq; *srq->db.db = 0; - if (mlx4_buf_alloc(dev->dev, buf_size, PAGE_SIZE * 2, &srq->buf, - GFP_KERNEL)) { + if (mlx4_buf_alloc(dev->dev, buf_size, PAGE_SIZE * 2, + &srq->buf)) { err = -ENOMEM; goto err_db; } @@ -167,7 +167,7 @@ struct ib_srq *mlx4_ib_create_srq(struct ib_pd *pd, if (err) goto err_buf; - err = mlx4_buf_write_mtt(dev->dev, &srq->mtt, &srq->buf, GFP_KERNEL); + err = mlx4_buf_write_mtt(dev->dev, &srq->mtt, &srq->buf); if (err) goto err_mtt; diff --git a/drivers/infiniband/hw/mlx4/sysfs.c b/drivers/infiniband/hw/mlx4/sysfs.c index 69fb5ba94d0f..0ba5ba7540c8 100644 --- a/drivers/infiniband/hw/mlx4/sysfs.c +++ b/drivers/infiniband/hw/mlx4/sysfs.c @@ -226,6 +226,7 @@ static int add_port_entries(struct mlx4_ib_dev *device, int port_num) int ret = 0 ; struct ib_port_attr attr; + memset(&attr, 0, sizeof(attr)); /* get the physical gid and pkey table sizes.*/ ret = __mlx4_ib_query_port(&device->ib_dev, port_num, &attr, 1); if (ret) diff --git a/drivers/infiniband/hw/mlx5/Makefile b/drivers/infiniband/hw/mlx5/Makefile index 7493a83acd28..90ad2adc752f 100644 --- a/drivers/infiniband/hw/mlx5/Makefile +++ b/drivers/infiniband/hw/mlx5/Makefile @@ -1,4 +1,4 @@ obj-$(CONFIG_MLX5_INFINIBAND) += mlx5_ib.o -mlx5_ib-y := main.o cq.o doorbell.o qp.o mem.o srq.o mr.o ah.o mad.o gsi.o ib_virt.o +mlx5_ib-y := main.o cq.o doorbell.o qp.o mem.o srq.o mr.o ah.o mad.o gsi.o ib_virt.o cmd.o mlx5_ib-$(CONFIG_INFINIBAND_ON_DEMAND_PAGING) += odp.o diff --git a/drivers/infiniband/hw/mlx5/ah.c b/drivers/infiniband/hw/mlx5/ah.c index d090e96f6f01..3363e29157f6 100644 --- a/drivers/infiniband/hw/mlx5/ah.c +++ b/drivers/infiniband/hw/mlx5/ah.c @@ -34,50 +34,51 @@ static struct ib_ah *create_ib_ah(struct mlx5_ib_dev *dev, struct mlx5_ib_ah *ah, - struct ib_ah_attr *ah_attr, - enum rdma_link_layer ll) + struct rdma_ah_attr *ah_attr) { - if (ah_attr->ah_flags & IB_AH_GRH) { - memcpy(ah->av.rgid, &ah_attr->grh.dgid, 16); - ah->av.grh_gid_fl = cpu_to_be32(ah_attr->grh.flow_label | + if (rdma_ah_get_ah_flags(ah_attr) & IB_AH_GRH) { + const struct ib_global_route *grh = rdma_ah_read_grh(ah_attr); + + memcpy(ah->av.rgid, &grh->dgid, 16); + ah->av.grh_gid_fl = cpu_to_be32(grh->flow_label | (1 << 30) | - ah_attr->grh.sgid_index << 20); - ah->av.hop_limit = ah_attr->grh.hop_limit; - ah->av.tclass = ah_attr->grh.traffic_class; + grh->sgid_index << 20); + ah->av.hop_limit = grh->hop_limit; + ah->av.tclass = grh->traffic_class; } - ah->av.stat_rate_sl = (ah_attr->static_rate << 4); + ah->av.stat_rate_sl = (rdma_ah_get_static_rate(ah_attr) << 4); - if (ll == IB_LINK_LAYER_ETHERNET) { - memcpy(ah->av.rmac, ah_attr->dmac, sizeof(ah_attr->dmac)); + if (ah_attr->type == RDMA_AH_ATTR_TYPE_ROCE) { + memcpy(ah->av.rmac, ah_attr->roce.dmac, + sizeof(ah_attr->roce.dmac)); ah->av.udp_sport = - mlx5_get_roce_udp_sport(dev, - ah_attr->port_num, - ah_attr->grh.sgid_index); - ah->av.stat_rate_sl |= (ah_attr->sl & 0x7) << 1; + mlx5_get_roce_udp_sport(dev, + rdma_ah_get_port_num(ah_attr), + rdma_ah_read_grh(ah_attr)->sgid_index); + ah->av.stat_rate_sl |= (rdma_ah_get_sl(ah_attr) & 0x7) << 1; } else { - ah->av.rlid = cpu_to_be16(ah_attr->dlid); - ah->av.fl_mlid = ah_attr->src_path_bits & 0x7f; - ah->av.stat_rate_sl |= (ah_attr->sl & 0xf); + ah->av.rlid = cpu_to_be16(rdma_ah_get_dlid(ah_attr)); + ah->av.fl_mlid = rdma_ah_get_path_bits(ah_attr) & 0x7f; + ah->av.stat_rate_sl |= (rdma_ah_get_sl(ah_attr) & 0xf); } return &ah->ibah; } -struct ib_ah *mlx5_ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr, +struct ib_ah *mlx5_ib_create_ah(struct ib_pd *pd, struct rdma_ah_attr *ah_attr, struct ib_udata *udata) { struct mlx5_ib_ah *ah; struct mlx5_ib_dev *dev = to_mdev(pd->device); - enum rdma_link_layer ll; - - ll = pd->device->get_link_layer(pd->device, ah_attr->port_num); + enum rdma_ah_attr_type ah_type = ah_attr->type; - if (ll == IB_LINK_LAYER_ETHERNET && !(ah_attr->ah_flags & IB_AH_GRH)) + if ((ah_type == RDMA_AH_ATTR_TYPE_ROCE) && + !(rdma_ah_get_ah_flags(ah_attr) & IB_AH_GRH)) return ERR_PTR(-EINVAL); - if (ll == IB_LINK_LAYER_ETHERNET && udata) { + if (ah_type == RDMA_AH_ATTR_TYPE_ROCE && udata) { int err; struct mlx5_ib_create_ah_resp resp = {}; u32 min_resp_len = offsetof(typeof(resp), dmac) + @@ -92,7 +93,7 @@ struct ib_ah *mlx5_ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr, if (err) return ERR_PTR(err); - memcpy(resp.dmac, ah_attr->dmac, ETH_ALEN); + memcpy(resp.dmac, ah_attr->roce.dmac, ETH_ALEN); err = ib_copy_to_udata(udata, &resp, resp.response_length); if (err) return ERR_PTR(err); @@ -102,28 +103,29 @@ struct ib_ah *mlx5_ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr, if (!ah) return ERR_PTR(-ENOMEM); - return create_ib_ah(dev, ah, ah_attr, ll); /* never fails */ + return create_ib_ah(dev, ah, ah_attr); /* never fails */ } -int mlx5_ib_query_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr) +int mlx5_ib_query_ah(struct ib_ah *ibah, struct rdma_ah_attr *ah_attr) { struct mlx5_ib_ah *ah = to_mah(ibah); u32 tmp; memset(ah_attr, 0, sizeof(*ah_attr)); + ah_attr->type = ibah->type; tmp = be32_to_cpu(ah->av.grh_gid_fl); if (tmp & (1 << 30)) { - ah_attr->ah_flags = IB_AH_GRH; - ah_attr->grh.sgid_index = (tmp >> 20) & 0xff; - ah_attr->grh.flow_label = tmp & 0xfffff; - memcpy(&ah_attr->grh.dgid, ah->av.rgid, 16); - ah_attr->grh.hop_limit = ah->av.hop_limit; - ah_attr->grh.traffic_class = ah->av.tclass; + rdma_ah_set_grh(ah_attr, NULL, + tmp & 0xfffff, + (tmp >> 20) & 0xff, + ah->av.hop_limit, + ah->av.tclass); + rdma_ah_set_dgid_raw(ah_attr, ah->av.rgid); } - ah_attr->dlid = be16_to_cpu(ah->av.rlid); - ah_attr->static_rate = ah->av.stat_rate_sl >> 4; - ah_attr->sl = ah->av.stat_rate_sl & 0xf; + rdma_ah_set_dlid(ah_attr, be16_to_cpu(ah->av.rlid)); + rdma_ah_set_static_rate(ah_attr, ah->av.stat_rate_sl >> 4); + rdma_ah_set_sl(ah_attr, ah->av.stat_rate_sl & 0xf); return 0; } diff --git a/drivers/infiniband/hw/mlx5/cmd.c b/drivers/infiniband/hw/mlx5/cmd.c new file mode 100644 index 000000000000..18d5e1db93ed --- /dev/null +++ b/drivers/infiniband/hw/mlx5/cmd.c @@ -0,0 +1,59 @@ +/* + * Copyright (c) 2017, Mellanox Technologies. All rights reserved. + * + * 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 "cmd.h" + +int mlx5_cmd_null_mkey(struct mlx5_core_dev *dev, u32 *null_mkey) +{ + u32 out[MLX5_ST_SZ_DW(query_special_contexts_out)] = {}; + u32 in[MLX5_ST_SZ_DW(query_special_contexts_in)] = {}; + int err; + + MLX5_SET(query_special_contexts_in, in, opcode, + MLX5_CMD_OP_QUERY_SPECIAL_CONTEXTS); + err = mlx5_cmd_exec(dev, in, sizeof(in), out, sizeof(out)); + if (!err) + *null_mkey = MLX5_GET(query_special_contexts_out, out, + null_mkey); + return err; +} + +int mlx5_cmd_query_cong_counter(struct mlx5_core_dev *dev, + bool reset, void *out, int out_size) +{ + u32 in[MLX5_ST_SZ_DW(query_cong_statistics_in)] = { }; + + MLX5_SET(query_cong_statistics_in, in, opcode, + MLX5_CMD_OP_QUERY_CONG_STATISTICS); + MLX5_SET(query_cong_statistics_in, in, clear, reset); + return mlx5_cmd_exec(dev, in, sizeof(in), out, out_size); +} diff --git a/drivers/infiniband/hw/qedr/qedr_hsi.h b/drivers/infiniband/hw/mlx5/cmd.h index 66d27521373f..fa09228193a6 100644 --- a/drivers/infiniband/hw/qedr/qedr_hsi.h +++ b/drivers/infiniband/hw/mlx5/cmd.h @@ -1,5 +1,5 @@ -/* QLogic qedr NIC Driver - * Copyright (c) 2015-2016 QLogic Corporation +/* + * Copyright (c) 2017, Mellanox Technologies. All rights reserved. * * 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 @@ -17,7 +17,7 @@ * * - 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 + * disclaimer in the documentation and/or other materials * provided with the distribution. * * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, @@ -29,28 +29,14 @@ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE * SOFTWARE. */ -#ifndef __QED_HSI_ROCE__ -#define __QED_HSI_ROCE__ -#include <linux/qed/common_hsi.h> -#include <linux/qed/roce_common.h> -#include "qedr_hsi_rdma.h" +#ifndef MLX5_IB_CMD_H +#define MLX5_IB_CMD_H -/* Affiliated asynchronous events / errors enumeration */ -enum roce_async_events_type { - ROCE_ASYNC_EVENT_NONE = 0, - ROCE_ASYNC_EVENT_COMM_EST = 1, - ROCE_ASYNC_EVENT_SQ_DRAINED, - ROCE_ASYNC_EVENT_SRQ_LIMIT, - ROCE_ASYNC_EVENT_LAST_WQE_REACHED, - ROCE_ASYNC_EVENT_CQ_ERR, - ROCE_ASYNC_EVENT_LOCAL_INVALID_REQUEST_ERR, - ROCE_ASYNC_EVENT_LOCAL_CATASTROPHIC_ERR, - ROCE_ASYNC_EVENT_LOCAL_ACCESS_ERR, - ROCE_ASYNC_EVENT_QP_CATASTROPHIC_ERR, - ROCE_ASYNC_EVENT_CQ_OVERFLOW_ERR, - ROCE_ASYNC_EVENT_SRQ_EMPTY, - MAX_ROCE_ASYNC_EVENTS_TYPE -}; +#include <linux/kernel.h> +#include <linux/mlx5/driver.h> -#endif /* __QED_HSI_ROCE__ */ +int mlx5_cmd_null_mkey(struct mlx5_core_dev *dev, u32 *null_mkey); +int mlx5_cmd_query_cong_counter(struct mlx5_core_dev *dev, + bool reset, void *out, int out_size); +#endif /* MLX5_IB_CMD_H */ diff --git a/drivers/infiniband/hw/mlx5/cq.c b/drivers/infiniband/hw/mlx5/cq.c index b3ef47c3ab73..a384d72ea3cd 100644 --- a/drivers/infiniband/hw/mlx5/cq.c +++ b/drivers/infiniband/hw/mlx5/cq.c @@ -172,6 +172,8 @@ static void handle_responder(struct ib_wc *wc, struct mlx5_cqe64 *cqe, struct mlx5_ib_srq *srq; struct mlx5_ib_wq *wq; u16 wqe_ctr; + u8 roce_packet_type; + bool vlan_present; u8 g; if (qp->ibqp.srq || qp->ibqp.xrcd) { @@ -223,7 +225,6 @@ static void handle_responder(struct ib_wc *wc, struct mlx5_cqe64 *cqe, break; } wc->slid = be16_to_cpu(cqe->slid); - wc->sl = (be32_to_cpu(cqe->flags_rqpn) >> 24) & 0xf; wc->src_qp = be32_to_cpu(cqe->flags_rqpn) & 0xffffff; wc->dlid_path_bits = cqe->ml_path; g = (be32_to_cpu(cqe->flags_rqpn) >> 28) & 3; @@ -237,10 +238,22 @@ static void handle_responder(struct ib_wc *wc, struct mlx5_cqe64 *cqe, wc->pkey_index = 0; } - if (ll != IB_LINK_LAYER_ETHERNET) + if (ll != IB_LINK_LAYER_ETHERNET) { + wc->sl = (be32_to_cpu(cqe->flags_rqpn) >> 24) & 0xf; return; + } + + vlan_present = cqe->l4_l3_hdr_type & 0x1; + roce_packet_type = (be32_to_cpu(cqe->flags_rqpn) >> 24) & 0x3; + if (vlan_present) { + wc->vlan_id = (be16_to_cpu(cqe->vlan_info)) & 0xfff; + wc->sl = (be16_to_cpu(cqe->vlan_info) >> 13) & 0x7; + wc->wc_flags |= IB_WC_WITH_VLAN; + } else { + wc->sl = 0; + } - switch (wc->sl & 0x3) { + switch (roce_packet_type) { case MLX5_CQE_ROCE_L3_HEADER_TYPE_GRH: wc->network_hdr_type = RDMA_NETWORK_IB; break; @@ -689,7 +702,7 @@ int mlx5_ib_arm_cq(struct ib_cq *ibcq, enum ib_cq_notify_flags flags) { struct mlx5_core_dev *mdev = to_mdev(ibcq->device)->mdev; struct mlx5_ib_cq *cq = to_mcq(ibcq); - void __iomem *uar_page = mdev->priv.uuari.uars[0].map; + void __iomem *uar_page = mdev->priv.uar->map; unsigned long irq_flags; int ret = 0; @@ -704,9 +717,7 @@ int mlx5_ib_arm_cq(struct ib_cq *ibcq, enum ib_cq_notify_flags flags) mlx5_cq_arm(&cq->mcq, (flags & IB_CQ_SOLICITED_MASK) == IB_CQ_SOLICITED ? MLX5_CQ_DB_REQ_NOT_SOL : MLX5_CQ_DB_REQ_NOT, - uar_page, - MLX5_GET_DOORBELL_LOCK(&mdev->priv.cq_uar_lock), - to_mcq(ibcq)->mcq.cons_index); + uar_page, to_mcq(ibcq)->mcq.cons_index); return ret; } @@ -777,7 +788,7 @@ static int create_cq_user(struct mlx5_ib_dev *dev, struct ib_udata *udata, *inlen = MLX5_ST_SZ_BYTES(create_cq_in) + MLX5_FLD_SZ_BYTES(create_cq_in, pas[0]) * ncont; - *cqb = mlx5_vzalloc(*inlen); + *cqb = kvzalloc(*inlen, GFP_KERNEL); if (!*cqb) { err = -ENOMEM; goto err_db; @@ -790,7 +801,7 @@ static int create_cq_user(struct mlx5_ib_dev *dev, struct ib_udata *udata, MLX5_SET(cqc, cqc, log_page_size, page_shift - MLX5_ADAPTER_PAGE_SHIFT); - *index = to_mucontext(context)->uuari.uars[0].index; + *index = to_mucontext(context)->bfregi.sys_pages[0]; if (ucmd.cqe_comp_en == 1) { if (unlikely((*cqe_size != 64) || @@ -820,7 +831,7 @@ static int create_cq_user(struct mlx5_ib_dev *dev, struct ib_udata *udata, return 0; err_cqb: - kfree(cqb); + kfree(*cqb); err_db: mlx5_ib_db_unmap_user(to_mucontext(context), &cq->db); @@ -873,7 +884,7 @@ static int create_cq_kernel(struct mlx5_ib_dev *dev, struct mlx5_ib_cq *cq, *inlen = MLX5_ST_SZ_BYTES(create_cq_in) + MLX5_FLD_SZ_BYTES(create_cq_in, pas[0]) * cq->buf.buf.npages; - *cqb = mlx5_vzalloc(*inlen); + *cqb = kvzalloc(*inlen, GFP_KERNEL); if (!*cqb) { err = -ENOMEM; goto err_buf; @@ -886,7 +897,7 @@ static int create_cq_kernel(struct mlx5_ib_dev *dev, struct mlx5_ib_cq *cq, MLX5_SET(cqc, cqc, log_page_size, cq->buf.buf.page_shift - MLX5_ADAPTER_PAGE_SHIFT); - *index = dev->mdev->priv.uuari.uars[0].index; + *index = dev->mdev->priv.uar->index; return 0; @@ -1303,7 +1314,7 @@ int mlx5_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata) inlen = MLX5_ST_SZ_BYTES(modify_cq_in) + MLX5_FLD_SZ_BYTES(modify_cq_in, pas[0]) * npas; - in = mlx5_vzalloc(inlen); + in = kvzalloc(inlen, GFP_KERNEL); if (!in) { err = -ENOMEM; goto ex_resize; diff --git a/drivers/infiniband/hw/mlx5/mad.c b/drivers/infiniband/hw/mlx5/mad.c index 39e58489dcc2..95db929bdc34 100644 --- a/drivers/infiniband/hw/mlx5/mad.c +++ b/drivers/infiniband/hw/mlx5/mad.c @@ -42,12 +42,24 @@ enum { MLX5_IB_VENDOR_CLASS2 = 0xa }; +static bool can_do_mad_ifc(struct mlx5_ib_dev *dev, u8 port_num, + struct ib_mad *in_mad) +{ + if (in_mad->mad_hdr.mgmt_class != IB_MGMT_CLASS_SUBN_LID_ROUTED && + in_mad->mad_hdr.mgmt_class != IB_MGMT_CLASS_SUBN_DIRECTED_ROUTE) + return true; + return dev->mdev->port_caps[port_num - 1].has_smi; +} + int mlx5_MAD_IFC(struct mlx5_ib_dev *dev, int ignore_mkey, int ignore_bkey, u8 port, const struct ib_wc *in_wc, const struct ib_grh *in_grh, const void *in_mad, void *response_mad) { u8 op_modifier = 0; + if (!can_do_mad_ifc(dev, port, (struct ib_mad *)in_mad)) + return -EPERM; + /* Key check traps can't be generated unless we have in_wc to * tell us where to send the trap. */ @@ -175,6 +187,8 @@ static void pma_cnt_assign(struct ib_pma_portcounters *pma_cnt, port_xmit_discards); MLX5_ASSIGN_PMA_CNTR(pma_cnt->port_xmit_constraint_errors, port_xmit_constraint_errors); + MLX5_ASSIGN_PMA_CNTR(pma_cnt->port_xmit_wait, + port_xmit_wait); MLX5_ASSIGN_PMA_CNTR(pma_cnt->port_rcv_constraint_errors, port_rcv_constraint_errors); MLX5_ASSIGN_PMA_CNTR(pma_cnt->link_overrun_errors, @@ -204,7 +218,7 @@ static int process_pma_cmd(struct ib_device *ibdev, u8 port_num, (struct ib_pma_portcounters_ext *)(out_mad->data + 40); int sz = MLX5_ST_SZ_BYTES(query_vport_counter_out); - out_cnt = mlx5_vzalloc(sz); + out_cnt = kvzalloc(sz, GFP_KERNEL); if (!out_cnt) return IB_MAD_RESULT_FAILURE; @@ -217,7 +231,7 @@ static int process_pma_cmd(struct ib_device *ibdev, u8 port_num, (struct ib_pma_portcounters *)(out_mad->data + 40); int sz = MLX5_ST_SZ_BYTES(ppcnt_reg); - out_cnt = mlx5_vzalloc(sz); + out_cnt = kvzalloc(sz, GFP_KERNEL); if (!out_cnt) return IB_MAD_RESULT_FAILURE; @@ -515,7 +529,7 @@ int mlx5_query_mad_ifc_port(struct ib_device *ibdev, u8 port, if (!in_mad || !out_mad) goto out; - memset(props, 0, sizeof(*props)); + /* props being zeroed by the caller, avoid zeroing it here */ init_query_mad(in_mad); in_mad->attr_id = IB_SMP_ATTR_PORT_INFO; diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c index d566f6738833..a7f2e60085c4 100644 --- a/drivers/infiniband/hw/mlx5/main.c +++ b/drivers/infiniband/hw/mlx5/main.c @@ -41,6 +41,8 @@ #include <asm/pat.h> #endif #include <linux/sched.h> +#include <linux/sched/mm.h> +#include <linux/sched/task.h> #include <linux/delay.h> #include <rdma/ib_user_verbs.h> #include <rdma/ib_addr.h> @@ -53,24 +55,21 @@ #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" #define DRIVER_NAME "mlx5_ib" -#define DRIVER_VERSION "2.2-1" -#define DRIVER_RELDATE "Feb 2014" +#define DRIVER_VERSION "5.0-0" MODULE_AUTHOR("Eli Cohen <eli@mellanox.com>"); MODULE_DESCRIPTION("Mellanox Connect-IB HCA IB driver"); MODULE_LICENSE("Dual BSD/GPL"); MODULE_VERSION(DRIVER_VERSION); -static int deprecated_prof_sel = 2; -module_param_named(prof_sel, deprecated_prof_sel, int, 0444); -MODULE_PARM_DESC(prof_sel, "profile selector. Deprecated here. Moved to module mlx5_core"); - static char mlx5_version[] = DRIVER_NAME ": Mellanox Connect-IB Infiniband driver v" - DRIVER_VERSION " (" DRIVER_RELDATE ")\n"; + DRIVER_VERSION "\n"; enum { MLX5_ATOMIC_SIZE_QP_8BYTES = 1 << 3, @@ -166,15 +165,84 @@ static struct net_device *mlx5_ib_get_netdev(struct ib_device *device, return ndev; } +static int translate_eth_proto_oper(u32 eth_proto_oper, u8 *active_speed, + u8 *active_width) +{ + switch (eth_proto_oper) { + case MLX5E_PROT_MASK(MLX5E_1000BASE_CX_SGMII): + case MLX5E_PROT_MASK(MLX5E_1000BASE_KX): + case MLX5E_PROT_MASK(MLX5E_100BASE_TX): + case MLX5E_PROT_MASK(MLX5E_1000BASE_T): + *active_width = IB_WIDTH_1X; + *active_speed = IB_SPEED_SDR; + break; + case MLX5E_PROT_MASK(MLX5E_10GBASE_T): + case MLX5E_PROT_MASK(MLX5E_10GBASE_CX4): + case MLX5E_PROT_MASK(MLX5E_10GBASE_KX4): + case MLX5E_PROT_MASK(MLX5E_10GBASE_KR): + case MLX5E_PROT_MASK(MLX5E_10GBASE_CR): + case MLX5E_PROT_MASK(MLX5E_10GBASE_SR): + case MLX5E_PROT_MASK(MLX5E_10GBASE_ER): + *active_width = IB_WIDTH_1X; + *active_speed = IB_SPEED_QDR; + break; + case MLX5E_PROT_MASK(MLX5E_25GBASE_CR): + case MLX5E_PROT_MASK(MLX5E_25GBASE_KR): + case MLX5E_PROT_MASK(MLX5E_25GBASE_SR): + *active_width = IB_WIDTH_1X; + *active_speed = IB_SPEED_EDR; + break; + case MLX5E_PROT_MASK(MLX5E_40GBASE_CR4): + case MLX5E_PROT_MASK(MLX5E_40GBASE_KR4): + case MLX5E_PROT_MASK(MLX5E_40GBASE_SR4): + case MLX5E_PROT_MASK(MLX5E_40GBASE_LR4): + *active_width = IB_WIDTH_4X; + *active_speed = IB_SPEED_QDR; + break; + case MLX5E_PROT_MASK(MLX5E_50GBASE_CR2): + case MLX5E_PROT_MASK(MLX5E_50GBASE_KR2): + case MLX5E_PROT_MASK(MLX5E_50GBASE_SR2): + *active_width = IB_WIDTH_1X; + *active_speed = IB_SPEED_HDR; + break; + case MLX5E_PROT_MASK(MLX5E_56GBASE_R4): + *active_width = IB_WIDTH_4X; + *active_speed = IB_SPEED_FDR; + break; + case MLX5E_PROT_MASK(MLX5E_100GBASE_CR4): + case MLX5E_PROT_MASK(MLX5E_100GBASE_SR4): + case MLX5E_PROT_MASK(MLX5E_100GBASE_KR4): + case MLX5E_PROT_MASK(MLX5E_100GBASE_LR4): + *active_width = IB_WIDTH_4X; + *active_speed = IB_SPEED_EDR; + break; + default: + return -EINVAL; + } + + return 0; +} + static int mlx5_query_port_roce(struct ib_device *device, u8 port_num, struct ib_port_attr *props) { struct mlx5_ib_dev *dev = to_mdev(device); + struct mlx5_core_dev *mdev = dev->mdev; struct net_device *ndev, *upper; enum ib_mtu ndev_ib_mtu; u16 qkey_viol_cntr; + u32 eth_prot_oper; + int err; - memset(props, 0, sizeof(*props)); + /* 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); + if (err) + return err; + + translate_eth_proto_oper(eth_prot_oper, &props->active_speed, + &props->active_width); props->port_cap_flags |= IB_PORT_CM_SUP; props->port_cap_flags |= IB_PORT_IP_BASED_GIDS; @@ -215,80 +283,49 @@ 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); - - props->active_width = IB_WIDTH_4X; /* TODO */ - props->active_speed = IB_SPEED_QDR; /* TODO */ - return 0; } -static void ib_gid_to_mlx5_roce_addr(const union ib_gid *gid, - const struct ib_gid_attr *attr, - void *mlx5_addr) +static int set_roce_addr(struct mlx5_ib_dev *dev, u8 port_num, + unsigned int index, const union ib_gid *gid, + const struct ib_gid_attr *attr) { -#define MLX5_SET_RA(p, f, v) MLX5_SET(roce_addr_layout, p, f, v) - char *mlx5_addr_l3_addr = MLX5_ADDR_OF(roce_addr_layout, mlx5_addr, - source_l3_address); - void *mlx5_addr_mac = MLX5_ADDR_OF(roce_addr_layout, mlx5_addr, - source_mac_47_32); - - if (!gid) - return; + enum ib_gid_type gid_type = IB_GID_TYPE_IB; + u8 roce_version = 0; + u8 roce_l3_type = 0; + bool vlan = false; + u8 mac[ETH_ALEN]; + u16 vlan_id = 0; - ether_addr_copy(mlx5_addr_mac, attr->ndev->dev_addr); + if (gid) { + gid_type = attr->gid_type; + ether_addr_copy(mac, attr->ndev->dev_addr); - if (is_vlan_dev(attr->ndev)) { - MLX5_SET_RA(mlx5_addr, vlan_valid, 1); - MLX5_SET_RA(mlx5_addr, vlan_id, vlan_dev_vlan_id(attr->ndev)); + if (is_vlan_dev(attr->ndev)) { + vlan = true; + vlan_id = vlan_dev_vlan_id(attr->ndev); + } } - switch (attr->gid_type) { + switch (gid_type) { case IB_GID_TYPE_IB: - MLX5_SET_RA(mlx5_addr, roce_version, MLX5_ROCE_VERSION_1); + roce_version = MLX5_ROCE_VERSION_1; break; case IB_GID_TYPE_ROCE_UDP_ENCAP: - MLX5_SET_RA(mlx5_addr, roce_version, MLX5_ROCE_VERSION_2); + roce_version = MLX5_ROCE_VERSION_2; + if (ipv6_addr_v4mapped((void *)gid)) + roce_l3_type = MLX5_ROCE_L3_TYPE_IPV4; + else + roce_l3_type = MLX5_ROCE_L3_TYPE_IPV6; break; default: - WARN_ON(true); - } - - if (attr->gid_type != IB_GID_TYPE_IB) { - if (ipv6_addr_v4mapped((void *)gid)) - MLX5_SET_RA(mlx5_addr, roce_l3_type, - MLX5_ROCE_L3_TYPE_IPV4); - else - MLX5_SET_RA(mlx5_addr, roce_l3_type, - MLX5_ROCE_L3_TYPE_IPV6); + mlx5_ib_warn(dev, "Unexpected GID type %u\n", gid_type); } - if ((attr->gid_type == IB_GID_TYPE_IB) || - !ipv6_addr_v4mapped((void *)gid)) - memcpy(mlx5_addr_l3_addr, gid, sizeof(*gid)); - else - memcpy(&mlx5_addr_l3_addr[12], &gid->raw[12], 4); -} - -static int set_roce_addr(struct ib_device *device, u8 port_num, - unsigned int index, - const union ib_gid *gid, - const struct ib_gid_attr *attr) -{ - struct mlx5_ib_dev *dev = to_mdev(device); - u32 in[MLX5_ST_SZ_DW(set_roce_address_in)] = {0}; - u32 out[MLX5_ST_SZ_DW(set_roce_address_out)] = {0}; - void *in_addr = MLX5_ADDR_OF(set_roce_address_in, in, roce_address); - enum rdma_link_layer ll = mlx5_ib_port_link_layer(device, port_num); - - if (ll != IB_LINK_LAYER_ETHERNET) - return -EINVAL; - - ib_gid_to_mlx5_roce_addr(gid, attr, in_addr); - - MLX5_SET(set_roce_address_in, in, roce_address_index, index); - MLX5_SET(set_roce_address_in, in, opcode, MLX5_CMD_OP_SET_ROCE_ADDRESS); - return mlx5_cmd_exec(dev->mdev, in, sizeof(in), out, sizeof(out)); + return mlx5_core_roce_gid_set(dev->mdev, index, roce_version, + roce_l3_type, gid->raw, mac, vlan, + vlan_id); } static int mlx5_ib_add_gid(struct ib_device *device, u8 port_num, @@ -296,13 +333,13 @@ static int mlx5_ib_add_gid(struct ib_device *device, u8 port_num, const struct ib_gid_attr *attr, __always_unused void **context) { - return set_roce_addr(device, port_num, index, gid, attr); + return set_roce_addr(to_mdev(device), port_num, index, gid, attr); } static int mlx5_ib_del_gid(struct ib_device *device, u8 port_num, unsigned int index, __always_unused void **context) { - return set_roce_addr(device, port_num, index, NULL, NULL); + return set_roce_addr(to_mdev(device), port_num, index, NULL, NULL); } __be16 mlx5_get_roce_udp_sport(struct mlx5_ib_dev *dev, u8 port_num, @@ -325,6 +362,27 @@ __be16 mlx5_get_roce_udp_sport(struct mlx5_ib_dev *dev, u8 port_num, return cpu_to_be16(MLX5_CAP_ROCE(dev->mdev, r_roce_min_src_udp_port)); } +int mlx5_get_roce_gid_type(struct mlx5_ib_dev *dev, u8 port_num, + int index, enum ib_gid_type *gid_type) +{ + struct ib_gid_attr attr; + union ib_gid gid; + int ret; + + ret = ib_get_cached_gid(&dev->ib_dev, port_num, index, &gid, &attr); + if (ret) + return ret; + + if (!attr.ndev) + return -ENODEV; + + dev_put(attr.ndev); + + *gid_type = attr.gid_type; + + return 0; +} + static int mlx5_use_mad_ifc(struct mlx5_ib_dev *dev) { if (MLX5_CAP_GEN(dev->mdev, port_type) == MLX5_CAP_PORT_TYPE_IB) @@ -357,7 +415,7 @@ static void get_atomic_caps(struct mlx5_ib_dev *dev, 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_endianess_mode); + MLX5_CAP_ATOMIC(dev->mdev, atomic_req_8B_endianness_mode); /* Check if HW supports 8 bytes standard atomic operations and capable * of host endianness respond @@ -564,8 +622,15 @@ static int mlx5_ib_query_device(struct ib_device *ibdev, props->device_cap_flags |= IB_DEVICE_BLOCK_MULTICAST_LOOPBACK; if (MLX5_CAP_GEN(dev->mdev, eth_net_offloads)) { - if (MLX5_CAP_ETH(mdev, csum_cap)) + if (MLX5_CAP_ETH(mdev, csum_cap)) { + /* Legacy bit to support old userspace libraries */ props->device_cap_flags |= IB_DEVICE_RAW_IP_CSUM; + props->raw_packet_caps |= IB_RAW_PACKET_CAP_IP_CSUM; + } + + if (MLX5_CAP_ETH(dev->mdev, vlan_cap)) + props->raw_packet_caps |= + IB_RAW_PACKET_CAP_CVLAN_STRIPPING; if (field_avail(typeof(resp), tso_caps, uhw->outlen)) { max_tso = MLX5_CAP_ETH(mdev, max_lso_cap); @@ -604,8 +669,11 @@ static int mlx5_ib_query_device(struct ib_device *ibdev, } if (MLX5_CAP_GEN(dev->mdev, eth_net_offloads) && - MLX5_CAP_ETH(dev->mdev, scatter_fcs)) + MLX5_CAP_ETH(dev->mdev, scatter_fcs)) { + /* 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; + } if (mlx5_get_flow_namespace(dev->mdev, MLX5_FLOW_NAMESPACE_BYPASS)) props->device_cap_flags |= IB_DEVICE_MANAGED_FLOW_STEERING; @@ -672,17 +740,6 @@ static int mlx5_ib_query_device(struct ib_device *ibdev, 1 << MLX5_CAP_GEN(dev->mdev, log_max_rq); } - if (field_avail(typeof(resp), mlx5_ib_support_multi_pkt_send_wqes, - uhw->outlen)) { - resp.mlx5_ib_support_multi_pkt_send_wqes = - MLX5_CAP_ETH(mdev, multi_pkt_send_wqe); - resp.response_length += - sizeof(resp.mlx5_ib_support_multi_pkt_send_wqes); - } - - if (field_avail(typeof(resp), reserved, uhw->outlen)) - resp.response_length += sizeof(resp.reserved); - if (field_avail(typeof(resp), cqe_comp_caps, uhw->outlen)) { resp.cqe_comp_caps.max_num = MLX5_CAP_GEN(dev->mdev, cqe_compression) ? @@ -706,6 +763,17 @@ static int mlx5_ib_query_device(struct ib_device *ibdev, resp.response_length += sizeof(resp.packet_pacing_caps); } + if (field_avail(typeof(resp), mlx5_ib_support_multi_pkt_send_wqes, + uhw->outlen)) { + resp.mlx5_ib_support_multi_pkt_send_wqes = + MLX5_CAP_ETH(mdev, multi_pkt_send_wqe); + resp.response_length += + sizeof(resp.mlx5_ib_support_multi_pkt_send_wqes); + } + + if (field_avail(typeof(resp), reserved, uhw->outlen)) + resp.response_length += sizeof(resp.reserved); + if (uhw->outlen) { err = ib_copy_to_udata(uhw, &resp, resp.response_length); @@ -830,7 +898,7 @@ static int mlx5_query_hca_port(struct ib_device *ibdev, u8 port, goto out; } - memset(props, 0, sizeof(*props)); + /* props being zeroed by the caller, avoid zeroing it here */ err = mlx5_query_hca_vport_context(mdev, 0, port, 0, rep); if (err) @@ -886,19 +954,31 @@ out: int mlx5_ib_query_port(struct ib_device *ibdev, u8 port, struct ib_port_attr *props) { + unsigned int count; + int ret; + switch (mlx5_get_vport_access_method(ibdev)) { case MLX5_VPORT_ACCESS_METHOD_MAD: - return mlx5_query_mad_ifc_port(ibdev, port, props); + ret = mlx5_query_mad_ifc_port(ibdev, port, props); + break; case MLX5_VPORT_ACCESS_METHOD_HCA: - return mlx5_query_hca_port(ibdev, port, props); + ret = mlx5_query_hca_port(ibdev, port, props); + break; case MLX5_VPORT_ACCESS_METHOD_NIC: - return mlx5_query_port_roce(ibdev, port, props); + ret = mlx5_query_port_roce(ibdev, port, props); + break; default: - return -EINVAL; + ret = -EINVAL; + } + + if (!ret && props) { + count = mlx5_core_reserved_gids_count(to_mdev(ibdev)->mdev); + props->gid_tbl_len -= count; } + return ret; } static int mlx5_ib_query_gid(struct ib_device *ibdev, u8 port, int index, @@ -968,6 +1048,31 @@ static int mlx5_ib_modify_device(struct ib_device *ibdev, int mask, return err; } +static int set_port_caps_atomic(struct mlx5_ib_dev *dev, u8 port_num, u32 mask, + u32 value) +{ + struct mlx5_hca_vport_context ctx = {}; + int err; + + err = mlx5_query_hca_vport_context(dev->mdev, 0, + port_num, 0, &ctx); + if (err) + return err; + + 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; + } + + ctx.cap_mask1 = value; + ctx.cap_mask1_perm = mask; + err = mlx5_core_modify_hca_vport_context(dev->mdev, 0, + port_num, 0, &ctx); + + return err; +} + static int mlx5_ib_modify_port(struct ib_device *ibdev, u8 port, int mask, struct ib_port_modify *props) { @@ -975,10 +1080,20 @@ static int mlx5_ib_modify_port(struct ib_device *ibdev, u8 port, int mask, struct ib_port_attr attr; u32 tmp; int err; + u32 change_mask; + u32 value; + bool is_ib = (mlx5_ib_port_link_layer(ibdev, port) == + IB_LINK_LAYER_INFINIBAND); + + if (MLX5_CAP_GEN(dev->mdev, ib_virt) && is_ib) { + change_mask = props->clr_port_cap_mask | props->set_port_cap_mask; + value = ~props->clr_port_cap_mask | props->set_port_cap_mask; + return set_port_caps_atomic(dev, port, change_mask, value); + } mutex_lock(&dev->cap_mask_mutex); - err = mlx5_ib_query_port(ibdev, port, &attr); + err = ib_query_port(ibdev, port, &attr); if (err) goto out; @@ -992,6 +1107,86 @@ out: return err; } +static void print_lib_caps(struct mlx5_ib_dev *dev, u64 caps) +{ + mlx5_ib_dbg(dev, "MLX5_LIB_CAP_4K_UAR = %s\n", + caps & MLX5_LIB_CAP_4K_UAR ? "y" : "n"); +} + +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) +{ + int uars_per_sys_page; + int bfregs_per_sys_page; + int ref_bfregs = req->total_num_bfregs; + + if (req->total_num_bfregs == 0) + return -EINVAL; + + BUILD_BUG_ON(MLX5_MAX_BFREGS % MLX5_NON_FP_BFREGS_IN_PAGE); + BUILD_BUG_ON(MLX5_MAX_BFREGS < MLX5_NON_FP_BFREGS_IN_PAGE); + + if (req->total_num_bfregs > MLX5_MAX_BFREGS) + return -ENOMEM; + + 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; + 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, alloated %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); + + return 0; +} + +static int allocate_uars(struct mlx5_ib_dev *dev, struct mlx5_ib_ucontext *context) +{ + struct mlx5_bfreg_info *bfregi; + int err; + int i; + + bfregi = &context->bfregi; + for (i = 0; i < bfregi->num_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]); + } + return 0; + +error: + for (--i; i >= 0; i--) + if (mlx5_cmd_free_uar(dev->mdev, bfregi->sys_pages[i])) + mlx5_ib_warn(dev, "failed to free uar %d\n", i); + + return err; +} + +static int deallocate_uars(struct mlx5_ib_dev *dev, struct mlx5_ib_ucontext *context) +{ + struct mlx5_bfreg_info *bfregi; + int err; + int i; + + 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; + } + } + return 0; +} + static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev, struct ib_udata *udata) { @@ -999,17 +1194,13 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev, struct mlx5_ib_alloc_ucontext_req_v2 req = {}; struct mlx5_ib_alloc_ucontext_resp resp = {}; struct mlx5_ib_ucontext *context; - struct mlx5_uuar_info *uuari; - struct mlx5_uar *uars; - int gross_uuars; - int num_uars; + struct mlx5_bfreg_info *bfregi; int ver; - int uuarn; int err; - int i; size_t reqlen; size_t min_req_v2 = offsetof(struct mlx5_ib_alloc_ucontext_req_v2, max_cqe_version); + bool lib_uar_4k; if (!dev->ib_active) return ERR_PTR(-EAGAIN); @@ -1032,27 +1223,14 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev, if (req.flags) return ERR_PTR(-EINVAL); - if (req.total_num_uuars > MLX5_MAX_UUARS) - return ERR_PTR(-ENOMEM); - - if (req.total_num_uuars == 0) - return ERR_PTR(-EINVAL); - if (req.comp_mask || req.reserved0 || req.reserved1 || req.reserved2) return ERR_PTR(-EOPNOTSUPP); - if (reqlen > sizeof(req) && - !ib_is_udata_cleared(udata, sizeof(req), - reqlen - sizeof(req))) - return ERR_PTR(-EOPNOTSUPP); - - req.total_num_uuars = ALIGN(req.total_num_uuars, - MLX5_NON_FP_BF_REGS_PER_PAGE); - if (req.num_low_latency_uuars > req.total_num_uuars - 1) + req.total_num_bfregs = ALIGN(req.total_num_bfregs, + MLX5_NON_FP_BFREGS_PER_UAR); + if (req.num_low_latency_bfregs > req.total_num_bfregs - 1) return ERR_PTR(-EINVAL); - num_uars = req.total_num_uuars / MLX5_NON_FP_BF_REGS_PER_PAGE; - gross_uuars = num_uars * MLX5_BF_REGS_PER_PAGE; resp.qp_tab_size = 1 << MLX5_CAP_GEN(dev->mdev, log_max_qp); if (mlx5_core_is_pf(dev->mdev) && MLX5_CAP_GEN(dev->mdev, bf)) resp.bf_reg_size = 1 << MLX5_CAP_GEN(dev->mdev, log_bf_reg_size); @@ -1065,6 +1243,10 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev, resp.cqe_version = min_t(__u8, (__u8)MLX5_CAP_GEN(dev->mdev, cqe_version), req.max_cqe_version); + resp.log_uar_size = MLX5_CAP_GEN(dev->mdev, uar_4k) ? + MLX5_ADAPTER_PAGE_SHIFT : PAGE_SHIFT; + resp.num_uars_per_page = MLX5_CAP_GEN(dev->mdev, uar_4k) ? + MLX5_CAP_GEN(dev->mdev, num_of_uars_per_page) : 1; resp.response_length = min(offsetof(typeof(resp), response_length) + sizeof(resp.response_length), udata->outlen); @@ -1072,58 +1254,58 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev, if (!context) return ERR_PTR(-ENOMEM); - uuari = &context->uuari; - mutex_init(&uuari->lock); - uars = kcalloc(num_uars, sizeof(*uars), GFP_KERNEL); - if (!uars) { - err = -ENOMEM; + lib_uar_4k = req.lib_caps & MLX5_LIB_CAP_4K_UAR; + bfregi = &context->bfregi; + + /* updates req->total_num_bfregs */ + err = calc_total_bfregs(dev, lib_uar_4k, &req, &bfregi->num_sys_pages); + if (err) goto out_ctx; - } - uuari->bitmap = kcalloc(BITS_TO_LONGS(gross_uuars), - sizeof(*uuari->bitmap), + mutex_init(&bfregi->lock); + bfregi->lib_uar_4k = lib_uar_4k; + bfregi->count = kcalloc(req.total_num_bfregs, sizeof(*bfregi->count), GFP_KERNEL); - if (!uuari->bitmap) { + if (!bfregi->count) { err = -ENOMEM; - goto out_uar_ctx; - } - /* - * clear all fast path uuars - */ - for (i = 0; i < gross_uuars; i++) { - uuarn = i & 3; - if (uuarn == 2 || uuarn == 3) - set_bit(i, uuari->bitmap); + goto out_ctx; } - uuari->count = kcalloc(gross_uuars, sizeof(*uuari->count), GFP_KERNEL); - if (!uuari->count) { + bfregi->sys_pages = kcalloc(bfregi->num_sys_pages, + sizeof(*bfregi->sys_pages), + GFP_KERNEL); + if (!bfregi->sys_pages) { err = -ENOMEM; - goto out_bitmap; + goto out_count; } - for (i = 0; i < num_uars; i++) { - err = mlx5_cmd_alloc_uar(dev->mdev, &uars[i].index); - if (err) - goto out_count; - } + err = allocate_uars(dev, context); + if (err) + goto out_sys_pages; #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING context->ibucontext.invalidate_range = &mlx5_ib_invalidate_range; #endif + context->upd_xlt_page = __get_free_page(GFP_KERNEL); + if (!context->upd_xlt_page) { + err = -ENOMEM; + goto out_uars; + } + mutex_init(&context->upd_xlt_page_mutex); + if (MLX5_CAP_GEN(dev->mdev, log_max_transport_domain)) { err = mlx5_core_alloc_transport_domain(dev->mdev, &context->tdn); if (err) - goto out_uars; + goto out_page; } INIT_LIST_HEAD(&context->vma_private_list); INIT_LIST_HEAD(&context->db_page_list); mutex_init(&context->db_page_mutex); - resp.tot_uuars = req.total_num_uuars; + resp.tot_bfregs = req.total_num_bfregs; resp.num_ports = MLX5_CAP_GEN(dev->mdev, num_ports); if (field_avail(typeof(resp), cqe_version, udata->outlen)) @@ -1135,32 +1317,46 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev, resp.response_length += sizeof(resp.cmds_supp_uhw); } + if (field_avail(typeof(resp), eth_min_inline, udata->outlen)) { + if (mlx5_ib_port_link_layer(ibdev, 1) == IB_LINK_LAYER_ETHERNET) { + mlx5_query_min_inline(dev->mdev, &resp.eth_min_inline); + resp.eth_min_inline++; + } + resp.response_length += sizeof(resp.eth_min_inline); + } + /* * We don't want to expose information from the PCI bar that is located * after 4096 bytes, so if the arch only supports larger pages, let's * pretend we don't support reading the HCA's core clock. This is also * forced by mmap function. */ - if (PAGE_SIZE <= 4096 && - field_avail(typeof(resp), hca_core_clock_offset, udata->outlen)) { - resp.comp_mask |= - MLX5_IB_ALLOC_UCONTEXT_RESP_MASK_CORE_CLOCK_OFFSET; - resp.hca_core_clock_offset = - offsetof(struct mlx5_init_seg, internal_timer_h) % - PAGE_SIZE; + if (field_avail(typeof(resp), hca_core_clock_offset, udata->outlen)) { + if (PAGE_SIZE <= 4096) { + resp.comp_mask |= + MLX5_IB_ALLOC_UCONTEXT_RESP_MASK_CORE_CLOCK_OFFSET; + resp.hca_core_clock_offset = + offsetof(struct mlx5_init_seg, internal_timer_h) % PAGE_SIZE; + } resp.response_length += sizeof(resp.hca_core_clock_offset) + sizeof(resp.reserved2); } + if (field_avail(typeof(resp), log_uar_size, udata->outlen)) + resp.response_length += sizeof(resp.log_uar_size); + + if (field_avail(typeof(resp), num_uars_per_page, udata->outlen)) + resp.response_length += sizeof(resp.num_uars_per_page); + err = ib_copy_to_udata(udata, &resp, resp.response_length); if (err) goto out_td; - uuari->ver = ver; - uuari->num_low_latency_uuars = req.num_low_latency_uuars; - uuari->uars = uars; - uuari->num_uars = num_uars; + bfregi->ver = ver; + bfregi->num_low_latency_bfregs = req.num_low_latency_bfregs; context->cqe_version = resp.cqe_version; + context->lib_caps = req.lib_caps; + print_lib_caps(dev, context->lib_caps); return &context->ibucontext; @@ -1168,20 +1364,21 @@ out_td: if (MLX5_CAP_GEN(dev->mdev, log_max_transport_domain)) mlx5_core_dealloc_transport_domain(dev->mdev, context->tdn); +out_page: + free_page(context->upd_xlt_page); + out_uars: - for (i--; i >= 0; i--) - mlx5_cmd_free_uar(dev->mdev, uars[i].index); -out_count: - kfree(uuari->count); + deallocate_uars(dev, context); -out_bitmap: - kfree(uuari->bitmap); +out_sys_pages: + kfree(bfregi->sys_pages); -out_uar_ctx: - kfree(uars); +out_count: + kfree(bfregi->count); out_ctx: kfree(context); + return ERR_PTR(err); } @@ -1189,28 +1386,31 @@ static int mlx5_ib_dealloc_ucontext(struct ib_ucontext *ibcontext) { struct mlx5_ib_ucontext *context = to_mucontext(ibcontext); struct mlx5_ib_dev *dev = to_mdev(ibcontext->device); - struct mlx5_uuar_info *uuari = &context->uuari; - int i; + struct mlx5_bfreg_info *bfregi; + bfregi = &context->bfregi; if (MLX5_CAP_GEN(dev->mdev, log_max_transport_domain)) mlx5_core_dealloc_transport_domain(dev->mdev, context->tdn); - for (i = 0; i < uuari->num_uars; i++) { - if (mlx5_cmd_free_uar(dev->mdev, uuari->uars[i].index)) - mlx5_ib_warn(dev, "failed to free UAR 0x%x\n", uuari->uars[i].index); - } - - kfree(uuari->count); - kfree(uuari->bitmap); - kfree(uuari->uars); + free_page(context->upd_xlt_page); + deallocate_uars(dev, context); + kfree(bfregi->sys_pages); + kfree(bfregi->count); kfree(context); return 0; } -static phys_addr_t uar_index2pfn(struct mlx5_ib_dev *dev, int index) +static phys_addr_t uar_index2pfn(struct mlx5_ib_dev *dev, + struct mlx5_bfreg_info *bfregi, + int idx) { - return (pci_resource_start(dev->mdev->pdev, 0) >> PAGE_SHIFT) + index; + 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; } static int get_command(unsigned long offset) @@ -1328,7 +1528,7 @@ static void mlx5_ib_disassociate_ucontext(struct ib_ucontext *ibcontext) /* need to protect from a race on closing the vma as part of * mlx5_ib_vma_close. */ - down_read(&owning_mm->mmap_sem); + down_write(&owning_mm->mmap_sem); list_for_each_entry_safe(vma_private, n, &context->vma_private_list, list) { vma = vma_private->vma; @@ -1338,11 +1538,12 @@ static void mlx5_ib_disassociate_ucontext(struct ib_ucontext *ibcontext) /* context going to be destroyed, should * not access ops any more. */ + vma->vm_flags &= ~(VM_SHARED | VM_MAYSHARE); vma->vm_ops = NULL; list_del(&vma_private->list); kfree(vma_private); } - up_read(&owning_mm->mmap_sem); + up_write(&owning_mm->mmap_sem); mmput(owning_mm); put_task_struct(owning_process); } @@ -1365,11 +1566,23 @@ static int uar_mmap(struct mlx5_ib_dev *dev, enum mlx5_ib_mmap_cmd cmd, struct vm_area_struct *vma, struct mlx5_ib_ucontext *context) { - struct mlx5_uuar_info *uuari = &context->uuari; + struct mlx5_bfreg_info *bfregi = &context->bfregi; int err; unsigned long idx; phys_addr_t pfn, pa; pgprot_t prot; + int uars_per_page; + + 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); + return -EINVAL; + } switch (cmd) { case MLX5_IB_MMAP_WC_PAGE: @@ -1392,14 +1605,7 @@ static int uar_mmap(struct mlx5_ib_dev *dev, enum mlx5_ib_mmap_cmd cmd, return -EINVAL; } - if (vma->vm_end - vma->vm_start != PAGE_SIZE) - return -EINVAL; - - idx = get_index(vma->vm_pgoff); - if (idx >= uuari->num_uars) - return -EINVAL; - - pfn = uar_index2pfn(dev, uuari->uars[idx].index); + pfn = uar_index2pfn(dev, bfregi, idx); mlx5_ib_dbg(dev, "uar idx 0x%lx, pfn %pa\n", idx, &pfn); vma->vm_page_prot = prot; @@ -1570,6 +1776,8 @@ static void set_tos(void *outer_c, void *outer_v, u8 mask, u8 val) #define LAST_IPV6_FIELD traffic_class #define LAST_TCP_UDP_FIELD src_port #define LAST_TUNNEL_FIELD tunnel_id +#define LAST_FLOW_TAG_FIELD tag_id +#define LAST_DROP_FIELD size /* Field is the last supported field */ #define FIELDS_NOT_SUPPORTED(filter, field)\ @@ -1579,8 +1787,11 @@ static void set_tos(void *outer_c, void *outer_v, u8 mask, u8 val) offsetof(typeof(filter), field) -\ sizeof(filter.field)) -static int parse_flow_attr(u32 *match_c, u32 *match_v, - const union ib_flow_spec *ib_spec) +#define IPV4_VERSION 4 +#define IPV6_VERSION 6 +static int parse_flow_attr(struct mlx5_core_dev *mdev, u32 *match_c, + u32 *match_v, const union ib_flow_spec *ib_spec, + u32 *tag_id, bool *is_drop) { void *misc_params_c = MLX5_ADDR_OF(fte_match_param, match_c, misc_parameters); @@ -1588,23 +1799,28 @@ static int parse_flow_attr(u32 *match_c, u32 *match_v, misc_parameters); void *headers_c; void *headers_v; + int match_ipv; if (ib_spec->type & IB_FLOW_SPEC_INNER) { headers_c = MLX5_ADDR_OF(fte_match_param, match_c, inner_headers); headers_v = MLX5_ADDR_OF(fte_match_param, match_v, inner_headers); + match_ipv = MLX5_CAP_FLOWTABLE_NIC_RX(mdev, + ft_field_support.inner_ip_version); } else { headers_c = MLX5_ADDR_OF(fte_match_param, match_c, outer_headers); headers_v = MLX5_ADDR_OF(fte_match_param, match_v, outer_headers); + match_ipv = MLX5_CAP_FLOWTABLE_NIC_RX(mdev, + ft_field_support.outer_ip_version); } switch (ib_spec->type & ~IB_FLOW_SPEC_INNER) { case IB_FLOW_SPEC_ETH: if (FIELDS_NOT_SUPPORTED(ib_spec->eth.mask, LAST_ETH_FIELD)) - return -ENOTSUPP; + return -EOPNOTSUPP; ether_addr_copy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, headers_c, dmac_47_16), @@ -1622,9 +1838,9 @@ static int parse_flow_attr(u32 *match_c, u32 *match_v, if (ib_spec->eth.mask.vlan_tag) { MLX5_SET(fte_match_set_lyr_2_4, headers_c, - vlan_tag, 1); + cvlan_tag, 1); MLX5_SET(fte_match_set_lyr_2_4, headers_v, - vlan_tag, 1); + cvlan_tag, 1); MLX5_SET(fte_match_set_lyr_2_4, headers_c, first_vid, ntohs(ib_spec->eth.mask.vlan_tag)); @@ -1652,12 +1868,19 @@ static int parse_flow_attr(u32 *match_c, u32 *match_v, break; case IB_FLOW_SPEC_IPV4: if (FIELDS_NOT_SUPPORTED(ib_spec->ipv4.mask, LAST_IPV4_FIELD)) - return -ENOTSUPP; + return -EOPNOTSUPP; - MLX5_SET(fte_match_set_lyr_2_4, headers_c, - ethertype, 0xffff); - MLX5_SET(fte_match_set_lyr_2_4, headers_v, - ethertype, ETH_P_IP); + if (match_ipv) { + MLX5_SET(fte_match_set_lyr_2_4, headers_c, + ip_version, 0xf); + MLX5_SET(fte_match_set_lyr_2_4, headers_v, + ip_version, IPV4_VERSION); + } else { + MLX5_SET(fte_match_set_lyr_2_4, headers_c, + ethertype, 0xffff); + MLX5_SET(fte_match_set_lyr_2_4, headers_v, + ethertype, ETH_P_IP); + } memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, headers_c, src_ipv4_src_ipv6.ipv4_layout.ipv4), @@ -1684,12 +1907,19 @@ static int parse_flow_attr(u32 *match_c, u32 *match_v, break; case IB_FLOW_SPEC_IPV6: if (FIELDS_NOT_SUPPORTED(ib_spec->ipv6.mask, LAST_IPV6_FIELD)) - return -ENOTSUPP; + return -EOPNOTSUPP; - MLX5_SET(fte_match_set_lyr_2_4, headers_c, - ethertype, 0xffff); - MLX5_SET(fte_match_set_lyr_2_4, headers_v, - ethertype, ETH_P_IPV6); + if (match_ipv) { + MLX5_SET(fte_match_set_lyr_2_4, headers_c, + ip_version, 0xf); + MLX5_SET(fte_match_set_lyr_2_4, headers_v, + ip_version, IPV6_VERSION); + } else { + MLX5_SET(fte_match_set_lyr_2_4, headers_c, + ethertype, 0xffff); + MLX5_SET(fte_match_set_lyr_2_4, headers_v, + ethertype, ETH_P_IPV6); + } memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, headers_c, src_ipv4_src_ipv6.ipv6_layout.ipv6), @@ -1725,7 +1955,7 @@ static int parse_flow_attr(u32 *match_c, u32 *match_v, case IB_FLOW_SPEC_TCP: if (FIELDS_NOT_SUPPORTED(ib_spec->tcp_udp.mask, LAST_TCP_UDP_FIELD)) - return -ENOTSUPP; + return -EOPNOTSUPP; MLX5_SET(fte_match_set_lyr_2_4, headers_c, ip_protocol, 0xff); @@ -1745,7 +1975,7 @@ static int parse_flow_attr(u32 *match_c, u32 *match_v, case IB_FLOW_SPEC_UDP: if (FIELDS_NOT_SUPPORTED(ib_spec->tcp_udp.mask, LAST_TCP_UDP_FIELD)) - return -ENOTSUPP; + return -EOPNOTSUPP; MLX5_SET(fte_match_set_lyr_2_4, headers_c, ip_protocol, 0xff); @@ -1765,13 +1995,28 @@ static int parse_flow_attr(u32 *match_c, u32 *match_v, case IB_FLOW_SPEC_VXLAN_TUNNEL: if (FIELDS_NOT_SUPPORTED(ib_spec->tunnel.mask, LAST_TUNNEL_FIELD)) - return -ENOTSUPP; + return -EOPNOTSUPP; MLX5_SET(fte_match_set_misc, misc_params_c, vxlan_vni, ntohl(ib_spec->tunnel.mask.tunnel_id)); MLX5_SET(fte_match_set_misc, misc_params_v, vxlan_vni, ntohl(ib_spec->tunnel.val.tunnel_id)); break; + case IB_FLOW_SPEC_ACTION_TAG: + if (FIELDS_NOT_SUPPORTED(ib_spec->flow_tag, + LAST_FLOW_TAG_FIELD)) + return -EOPNOTSUPP; + if (ib_spec->flow_tag.tag_id >= BIT(24)) + return -EINVAL; + + *tag_id = ib_spec->flow_tag.tag_id; + break; + case IB_FLOW_SPEC_ACTION_DROP: + if (FIELDS_NOT_SUPPORTED(ib_spec->drop, + LAST_DROP_FIELD)) + return -EOPNOTSUPP; + *is_drop = true; + break; default: return -EINVAL; } @@ -1802,26 +2047,60 @@ static bool flow_is_multicast_only(struct ib_flow_attr *ib_attr) is_multicast_ether_addr(eth_spec->val.dst_mac); } -static bool is_valid_attr(const struct ib_flow_attr *flow_attr) +static bool is_valid_ethertype(struct mlx5_core_dev *mdev, + const struct ib_flow_attr *flow_attr, + bool check_inner) { union ib_flow_spec *ib_spec = (union ib_flow_spec *)(flow_attr + 1); - bool has_ipv4_spec = false; - bool eth_type_ipv4 = true; + int match_ipv = check_inner ? + MLX5_CAP_FLOWTABLE_NIC_RX(mdev, + ft_field_support.inner_ip_version) : + MLX5_CAP_FLOWTABLE_NIC_RX(mdev, + ft_field_support.outer_ip_version); + int inner_bit = check_inner ? IB_FLOW_SPEC_INNER : 0; + bool ipv4_spec_valid, ipv6_spec_valid; + unsigned int ip_spec_type = 0; + bool has_ethertype = false; unsigned int spec_index; + bool mask_valid = true; + u16 eth_type = 0; + bool type_valid; /* Validate that ethertype is correct */ for (spec_index = 0; spec_index < flow_attr->num_of_specs; spec_index++) { - if (ib_spec->type == IB_FLOW_SPEC_ETH && + if ((ib_spec->type == (IB_FLOW_SPEC_ETH | inner_bit)) && ib_spec->eth.mask.ether_type) { - if (!((ib_spec->eth.mask.ether_type == htons(0xffff)) && - ib_spec->eth.val.ether_type == htons(ETH_P_IP))) - eth_type_ipv4 = false; - } else if (ib_spec->type == IB_FLOW_SPEC_IPV4) { - has_ipv4_spec = true; + mask_valid = (ib_spec->eth.mask.ether_type == + htons(0xffff)); + has_ethertype = true; + eth_type = ntohs(ib_spec->eth.val.ether_type); + } else if ((ib_spec->type == (IB_FLOW_SPEC_IPV4 | inner_bit)) || + (ib_spec->type == (IB_FLOW_SPEC_IPV6 | inner_bit))) { + ip_spec_type = ib_spec->type; } ib_spec = (void *)ib_spec + ib_spec->size; } - return !has_ipv4_spec || eth_type_ipv4; + + type_valid = (!has_ethertype) || (!ip_spec_type); + if (!type_valid && mask_valid) { + ipv4_spec_valid = (eth_type == ETH_P_IP) && + (ip_spec_type == (IB_FLOW_SPEC_IPV4 | inner_bit)); + ipv6_spec_valid = (eth_type == ETH_P_IPV6) && + (ip_spec_type == (IB_FLOW_SPEC_IPV6 | inner_bit)); + + type_valid = (ipv4_spec_valid) || (ipv6_spec_valid) || + (((eth_type == ETH_P_MPLS_UC) || + (eth_type == ETH_P_MPLS_MC)) && match_ipv); + } + + return type_valid; +} + +static bool is_valid_attr(struct mlx5_core_dev *mdev, + const struct ib_flow_attr *flow_attr) +{ + return is_valid_ethertype(mdev, flow_attr, false) && + is_valid_ethertype(mdev, flow_attr, true); } static void put_flow_table(struct mlx5_ib_dev *dev, @@ -1873,8 +2152,8 @@ enum flow_table_type { MLX5_IB_FT_TX }; -#define MLX5_FS_MAX_TYPES 10 -#define MLX5_FS_MAX_ENTRIES 32000UL +#define MLX5_FS_MAX_TYPES 6 +#define MLX5_FS_MAX_ENTRIES BIT(16) static struct mlx5_ib_flow_prio *get_flow_table(struct mlx5_ib_dev *dev, struct ib_flow_attr *flow_attr, enum flow_table_type ft_type) @@ -1883,11 +2162,14 @@ static struct mlx5_ib_flow_prio *get_flow_table(struct mlx5_ib_dev *dev, struct mlx5_flow_namespace *ns = NULL; struct mlx5_ib_flow_prio *prio; struct mlx5_flow_table *ft; + int max_table_size; int num_entries; int num_groups; int priority; int err = 0; + max_table_size = BIT(MLX5_CAP_FLOWTABLE_NIC_RX(dev->mdev, + log_max_ft_size)); if (flow_attr->type == IB_FLOW_ATTR_NORMAL) { if (flow_is_multicast_only(flow_attr) && !dont_trap) @@ -1926,6 +2208,9 @@ static struct mlx5_ib_flow_prio *get_flow_table(struct mlx5_ib_dev *dev, if (!ns) return ERR_PTR(-ENOTSUPP); + if (num_entries > max_table_size) + return ERR_PTR(-ENOMEM); + ft = prio->flow_table; if (!ft) { ft = mlx5_create_auto_grouped_flow_table(ns, priority, @@ -1953,14 +2238,18 @@ static struct mlx5_ib_flow_handler *create_flow_rule(struct mlx5_ib_dev *dev, struct mlx5_ib_flow_handler *handler; struct mlx5_flow_act flow_act = {0}; struct mlx5_flow_spec *spec; + struct mlx5_flow_destination *rule_dst = dst; const void *ib_flow = (const void *)flow_attr + sizeof(*flow_attr); unsigned int spec_index; + u32 flow_tag = MLX5_FS_DEFAULT_FLOW_TAG; + bool is_drop = false; int err = 0; + int dest_num = 1; - if (!is_valid_attr(flow_attr)) + if (!is_valid_attr(dev->mdev, flow_attr)) return ERR_PTR(-EINVAL); - spec = mlx5_vzalloc(sizeof(*spec)); + spec = kvzalloc(sizeof(*spec), GFP_KERNEL); handler = kzalloc(sizeof(*handler), GFP_KERNEL); if (!handler || !spec) { err = -ENOMEM; @@ -1970,8 +2259,9 @@ static struct mlx5_ib_flow_handler *create_flow_rule(struct mlx5_ib_dev *dev, INIT_LIST_HEAD(&handler->list); for (spec_index = 0; spec_index < flow_attr->num_of_specs; spec_index++) { - err = parse_flow_attr(spec->match_criteria, - spec->match_value, ib_flow); + err = parse_flow_attr(dev->mdev, spec->match_criteria, + spec->match_value, + ib_flow, &flow_tag, &is_drop); if (err < 0) goto free; @@ -1979,12 +2269,27 @@ static struct mlx5_ib_flow_handler *create_flow_rule(struct mlx5_ib_dev *dev, } spec->match_criteria_enable = get_match_criteria_enable(spec->match_criteria); - flow_act.action = dst ? MLX5_FLOW_CONTEXT_ACTION_FWD_DEST : - MLX5_FLOW_CONTEXT_ACTION_FWD_NEXT_PRIO; - flow_act.flow_tag = MLX5_FS_DEFAULT_FLOW_TAG; + if (is_drop) { + flow_act.action = MLX5_FLOW_CONTEXT_ACTION_DROP; + rule_dst = NULL; + dest_num = 0; + } else { + flow_act.action = dst ? MLX5_FLOW_CONTEXT_ACTION_FWD_DEST : + MLX5_FLOW_CONTEXT_ACTION_FWD_NEXT_PRIO; + } + + if (flow_tag != MLX5_FS_DEFAULT_FLOW_TAG && + (flow_attr->type == IB_FLOW_ATTR_ALL_DEFAULT || + flow_attr->type == IB_FLOW_ATTR_MC_DEFAULT)) { + mlx5_ib_warn(dev, "Flow tag %u and attribute type %x isn't allowed in leftovers\n", + flow_tag, flow_attr->type); + err = -EINVAL; + goto free; + } + flow_act.flow_tag = flow_tag; handler->rule = mlx5_add_flow_rules(ft, spec, &flow_act, - dst, 1); + rule_dst, dest_num); if (IS_ERR(handler->rule)) { err = PTR_ERR(handler->rule); @@ -2140,7 +2445,7 @@ static struct ib_flow *mlx5_ib_create_flow(struct ib_qp *qp, int err; if (flow_attr->priority > MLX5_IB_FLOW_LAST_PRIO) - return ERR_PTR(-ENOSPC); + return ERR_PTR(-ENOMEM); if (domain != IB_FLOW_DOMAIN_USER || flow_attr->port > MLX5_CAP_GEN(dev->mdev, num_ports) || @@ -2451,6 +2756,35 @@ static void mlx5_ib_event(struct mlx5_core_dev *dev, void *context, ibdev->ib_active = false; } +static int set_has_smi_cap(struct mlx5_ib_dev *dev) +{ + struct mlx5_hca_vport_context vport_ctx; + int err; + int port; + + for (port = 1; port <= MLX5_CAP_GEN(dev->mdev, 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) { + if (MLX5_CAP_GEN(dev->mdev, ib_virt)) { + err = mlx5_query_hca_vport_context(dev->mdev, 0, + port, 0, + &vport_ctx); + if (err) { + mlx5_ib_err(dev, "query_hca_vport_context for port=%d failed %d\n", + port, err); + return err; + } + dev->mdev->port_caps[port - 1].has_smi = + vport_ctx.has_smi; + } else { + dev->mdev->port_caps[port - 1].has_smi = true; + } + } + } + return 0; +} + static void get_ext_port_caps(struct mlx5_ib_dev *dev) { int port; @@ -2475,6 +2809,10 @@ static int get_port_caps(struct mlx5_ib_dev *dev) if (!dprops) goto out; + err = set_has_smi_cap(dev); + if (err) + goto out; + err = mlx5_ib_query_device(&dev->ib_dev, dprops, &uhw); if (err) { mlx5_ib_warn(dev, "query_device failed %d\n", err); @@ -2482,6 +2820,7 @@ static int get_port_caps(struct mlx5_ib_dev *dev) } 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", @@ -2626,6 +2965,18 @@ error_0: return ret; } +static u8 mlx5_get_umr_fence(u8 umr_fence_cap) +{ + switch (umr_fence_cap) { + case MLX5_CAP_UMR_FENCE_NONE: + return MLX5_FENCE_MODE_NONE; + case MLX5_CAP_UMR_FENCE_SMALL: + return MLX5_FENCE_MODE_INITIATOR_SMALL; + default: + return MLX5_FENCE_MODE_STRONG_ORDERING; + } +} + static int create_dev_resources(struct mlx5_ib_resources *devr) { struct ib_srq_init_attr attr; @@ -2776,11 +3127,13 @@ static u32 get_core_cap_flags(struct ib_device *ibdev) if (ll == IB_LINK_LAYER_INFINIBAND) return RDMA_CORE_PORT_IBA_IB; + ret = RDMA_CORE_PORT_RAW_PACKET; + if (!(l3_type_cap & MLX5_ROCE_L3_TYPE_IPV4_CAP)) - return 0; + return ret; if (!(l3_type_cap & MLX5_ROCE_L3_TYPE_IPV6_CAP)) - return 0; + return ret; if (roce_version_cap & MLX5_ROCE_VERSION_1_CAP) ret |= RDMA_CORE_PORT_IBA_ROCE; @@ -2799,7 +3152,9 @@ static int mlx5_port_immutable(struct ib_device *ibdev, u8 port_num, enum rdma_link_layer ll = mlx5_ib_port_link_layer(ibdev, port_num); int err; - err = mlx5_ib_query_port(ibdev, port_num, &attr); + immutable->core_cap_flags = get_core_cap_flags(ibdev); + + err = ib_query_port(ibdev, port_num, &attr); if (err) return err; @@ -2920,29 +3275,148 @@ static void mlx5_disable_eth(struct mlx5_ib_dev *dev) mlx5_nic_vport_disable_roce(dev->mdev); } -static void mlx5_ib_dealloc_q_counters(struct mlx5_ib_dev *dev) +struct mlx5_ib_counter { + const char *name; + size_t offset; +}; + +#define INIT_Q_COUNTER(_name) \ + { .name = #_name, .offset = MLX5_BYTE_OFF(query_q_counter_out, _name)} + +static const struct mlx5_ib_counter basic_q_cnts[] = { + INIT_Q_COUNTER(rx_write_requests), + INIT_Q_COUNTER(rx_read_requests), + INIT_Q_COUNTER(rx_atomic_requests), + INIT_Q_COUNTER(out_of_buffer), +}; + +static const struct mlx5_ib_counter out_of_seq_q_cnts[] = { + INIT_Q_COUNTER(out_of_sequence), +}; + +static const struct mlx5_ib_counter retrans_q_cnts[] = { + INIT_Q_COUNTER(duplicate_request), + INIT_Q_COUNTER(rnr_nak_retry_err), + INIT_Q_COUNTER(packet_seq_err), + INIT_Q_COUNTER(implied_nak_seq_err), + INIT_Q_COUNTER(local_ack_timeout_err), +}; + +#define INIT_CONG_COUNTER(_name) \ + { .name = #_name, .offset = \ + MLX5_BYTE_OFF(query_cong_statistics_out, _name ## _high)} + +static const struct mlx5_ib_counter cong_cnts[] = { + INIT_CONG_COUNTER(rp_cnp_ignored), + INIT_CONG_COUNTER(rp_cnp_handled), + INIT_CONG_COUNTER(np_ecn_marked_roce_packets), + INIT_CONG_COUNTER(np_cnp_sent), +}; + +static void mlx5_ib_dealloc_counters(struct mlx5_ib_dev *dev) { unsigned int i; - for (i = 0; i < dev->num_ports; i++) + for (i = 0; i < dev->num_ports; i++) { mlx5_core_dealloc_q_counter(dev->mdev, - dev->port[i].q_cnt_id); + dev->port[i].cnts.set_id); + kfree(dev->port[i].cnts.names); + kfree(dev->port[i].cnts.offsets); + } } -static int mlx5_ib_alloc_q_counters(struct mlx5_ib_dev *dev) +static int __mlx5_ib_alloc_counters(struct mlx5_ib_dev *dev, + struct mlx5_ib_counters *cnts) +{ + u32 num_counters; + + num_counters = ARRAY_SIZE(basic_q_cnts); + + if (MLX5_CAP_GEN(dev->mdev, out_of_seq_cnt)) + num_counters += ARRAY_SIZE(out_of_seq_q_cnts); + + if (MLX5_CAP_GEN(dev->mdev, retransmission_q_counters)) + num_counters += ARRAY_SIZE(retrans_q_cnts); + cnts->num_q_counters = num_counters; + + if (MLX5_CAP_GEN(dev->mdev, cc_query_allowed)) { + cnts->num_cong_counters = ARRAY_SIZE(cong_cnts); + num_counters += ARRAY_SIZE(cong_cnts); + } + + cnts->names = kcalloc(num_counters, sizeof(cnts->names), GFP_KERNEL); + if (!cnts->names) + return -ENOMEM; + + cnts->offsets = kcalloc(num_counters, + sizeof(cnts->offsets), GFP_KERNEL); + if (!cnts->offsets) + goto err_names; + + return 0; + +err_names: + kfree(cnts->names); + return -ENOMEM; +} + +static void mlx5_ib_fill_counters(struct mlx5_ib_dev *dev, + const char **names, + size_t *offsets) +{ + int i; + int j = 0; + + for (i = 0; i < ARRAY_SIZE(basic_q_cnts); i++, j++) { + names[j] = basic_q_cnts[i].name; + offsets[j] = basic_q_cnts[i].offset; + } + + if (MLX5_CAP_GEN(dev->mdev, out_of_seq_cnt)) { + for (i = 0; i < ARRAY_SIZE(out_of_seq_q_cnts); i++, j++) { + names[j] = out_of_seq_q_cnts[i].name; + offsets[j] = out_of_seq_q_cnts[i].offset; + } + } + + if (MLX5_CAP_GEN(dev->mdev, retransmission_q_counters)) { + for (i = 0; i < ARRAY_SIZE(retrans_q_cnts); i++, j++) { + names[j] = retrans_q_cnts[i].name; + offsets[j] = retrans_q_cnts[i].offset; + } + } + + if (MLX5_CAP_GEN(dev->mdev, cc_query_allowed)) { + for (i = 0; i < ARRAY_SIZE(cong_cnts); i++, j++) { + names[j] = cong_cnts[i].name; + offsets[j] = cong_cnts[i].offset; + } + } +} + +static int mlx5_ib_alloc_counters(struct mlx5_ib_dev *dev) { int i; int ret; for (i = 0; i < dev->num_ports; i++) { + struct mlx5_ib_port *port = &dev->port[i]; + ret = mlx5_core_alloc_q_counter(dev->mdev, - &dev->port[i].q_cnt_id); + &port->cnts.set_id); if (ret) { mlx5_ib_warn(dev, "couldn't allocate queue counter for port %d, err %d\n", i + 1, ret); goto dealloc_counters; } + + ret = __mlx5_ib_alloc_counters(dev, &port->cnts); + if (ret) + goto dealloc_counters; + + mlx5_ib_fill_counters(dev, port->cnts.names, + port->cnts.offsets); } return 0; @@ -2950,81 +3424,136 @@ static int mlx5_ib_alloc_q_counters(struct mlx5_ib_dev *dev) dealloc_counters: while (--i >= 0) mlx5_core_dealloc_q_counter(dev->mdev, - dev->port[i].q_cnt_id); + dev->port[i].cnts.set_id); return ret; } -static const char * const names[] = { - "rx_write_requests", - "rx_read_requests", - "rx_atomic_requests", - "out_of_buffer", - "out_of_sequence", - "duplicate_request", - "rnr_nak_retry_err", - "packet_seq_err", - "implied_nak_seq_err", - "local_ack_timeout_err", -}; - -static const size_t stats_offsets[] = { - MLX5_BYTE_OFF(query_q_counter_out, rx_write_requests), - MLX5_BYTE_OFF(query_q_counter_out, rx_read_requests), - MLX5_BYTE_OFF(query_q_counter_out, rx_atomic_requests), - MLX5_BYTE_OFF(query_q_counter_out, out_of_buffer), - MLX5_BYTE_OFF(query_q_counter_out, out_of_sequence), - MLX5_BYTE_OFF(query_q_counter_out, duplicate_request), - MLX5_BYTE_OFF(query_q_counter_out, rnr_nak_retry_err), - MLX5_BYTE_OFF(query_q_counter_out, packet_seq_err), - MLX5_BYTE_OFF(query_q_counter_out, implied_nak_seq_err), - MLX5_BYTE_OFF(query_q_counter_out, local_ack_timeout_err), -}; - static struct rdma_hw_stats *mlx5_ib_alloc_hw_stats(struct ib_device *ibdev, u8 port_num) { - BUILD_BUG_ON(ARRAY_SIZE(names) != ARRAY_SIZE(stats_offsets)); + struct mlx5_ib_dev *dev = to_mdev(ibdev); + struct mlx5_ib_port *port = &dev->port[port_num - 1]; /* We support only per port stats */ if (port_num == 0) return NULL; - return rdma_alloc_hw_stats_struct(names, ARRAY_SIZE(names), + return rdma_alloc_hw_stats_struct(port->cnts.names, + port->cnts.num_q_counters + + port->cnts.num_cong_counters, RDMA_HW_STATS_DEFAULT_LIFESPAN); } -static int mlx5_ib_get_hw_stats(struct ib_device *ibdev, - struct rdma_hw_stats *stats, - u8 port, int index) +static int mlx5_ib_query_q_counters(struct mlx5_ib_dev *dev, + struct mlx5_ib_port *port, + struct rdma_hw_stats *stats) { - struct mlx5_ib_dev *dev = to_mdev(ibdev); int outlen = MLX5_ST_SZ_BYTES(query_q_counter_out); void *out; __be32 val; - int ret; - int i; + int ret, i; - if (!port || !stats) - return -ENOSYS; - - out = mlx5_vzalloc(outlen); + out = kvzalloc(outlen, GFP_KERNEL); if (!out) return -ENOMEM; ret = mlx5_core_query_q_counter(dev->mdev, - dev->port[port - 1].q_cnt_id, 0, + port->cnts.set_id, 0, out, outlen); if (ret) goto free; - for (i = 0; i < ARRAY_SIZE(names); i++) { - val = *(__be32 *)(out + stats_offsets[i]); + for (i = 0; i < port->cnts.num_q_counters; i++) { + val = *(__be32 *)(out + port->cnts.offsets[i]); stats->value[i] = (u64)be32_to_cpu(val); } + free: kvfree(out); - return ARRAY_SIZE(names); + return ret; +} + +static int mlx5_ib_query_cong_counters(struct mlx5_ib_dev *dev, + struct mlx5_ib_port *port, + struct rdma_hw_stats *stats) +{ + int outlen = MLX5_ST_SZ_BYTES(query_cong_statistics_out); + void *out; + int ret, i; + int offset = port->cnts.num_q_counters; + + out = kvzalloc(outlen, GFP_KERNEL); + if (!out) + return -ENOMEM; + + ret = mlx5_cmd_query_cong_counter(dev->mdev, false, out, outlen); + if (ret) + goto free; + + for (i = 0; i < port->cnts.num_cong_counters; i++) { + stats->value[i + offset] = + be64_to_cpup((__be64 *)(out + + port->cnts.offsets[i + offset])); + } + +free: + kvfree(out); + return ret; +} + +static int mlx5_ib_get_hw_stats(struct ib_device *ibdev, + struct rdma_hw_stats *stats, + u8 port_num, int index) +{ + struct mlx5_ib_dev *dev = to_mdev(ibdev); + struct mlx5_ib_port *port = &dev->port[port_num - 1]; + int ret, num_counters; + + if (!stats) + return -EINVAL; + + ret = mlx5_ib_query_q_counters(dev, port, stats); + if (ret) + return ret; + num_counters = port->cnts.num_q_counters; + + if (MLX5_CAP_GEN(dev->mdev, cc_query_allowed)) { + ret = mlx5_ib_query_cong_counters(dev, port, stats); + if (ret) + return ret; + num_counters += port->cnts.num_cong_counters; + } + + return num_counters; +} + +static void mlx5_ib_free_rdma_netdev(struct net_device *netdev) +{ + return mlx5_rdma_netdev_free(netdev); +} + +static struct net_device* +mlx5_ib_alloc_rdma_netdev(struct ib_device *hca, + u8 port_num, + enum rdma_netdev_t type, + const char *name, + unsigned char name_assign_type, + void (*setup)(struct net_device *)) +{ + struct net_device *netdev; + struct rdma_netdev *rn; + + if (type != RDMA_NETDEV_IPOIB) + return ERR_PTR(-EOPNOTSUPP); + + netdev = mlx5_rdma_netdev_alloc(to_mdev(hca)->mdev, hca, + name, setup); + if (likely(!IS_ERR_OR_NULL(netdev))) { + rn = netdev_priv(netdev); + rn->free_rdma_netdev = mlx5_ib_free_rdma_netdev; + } + return netdev; } static void *mlx5_ib_add(struct mlx5_core_dev *mdev) @@ -3060,8 +3589,6 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev) if (mlx5_use_mad_ifc(dev)) get_ext_port_caps(dev); - MLX5_INIT_DOORBELL_LOCK(&dev->uar_lock); - if (!mlx5_lag_is_active(mdev)) name = "mlx5_%d"; else @@ -3075,7 +3602,7 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev) 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.dma_device = &mdev->pdev->dev; + dev->ib_dev.dev.parent = &mdev->pdev->dev; dev->ib_dev.uverbs_abi_ver = MLX5_IB_UVERBS_ABI_VERSION; dev->ib_dev.uverbs_cmd_mask = @@ -3159,6 +3686,9 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev) dev->ib_dev.check_mr_status = mlx5_ib_check_mr_status; dev->ib_dev.get_port_immutable = mlx5_port_immutable; dev->ib_dev.get_dev_fw_str = get_dev_fw_str; + if (MLX5_CAP_GEN(mdev, ipoib_enhanced_offloads)) + dev->ib_dev.alloc_rdma_netdev = mlx5_ib_alloc_rdma_netdev; + if (mlx5_core_is_pf(mdev)) { dev->ib_dev.get_vf_config = mlx5_ib_get_vf_config; dev->ib_dev.set_vf_link_state = mlx5_ib_set_vf_link_state; @@ -3170,6 +3700,8 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev) 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)) { dev->ib_dev.alloc_mw = mlx5_ib_alloc_mw; dev->ib_dev.dealloc_mw = mlx5_ib_dealloc_mw; @@ -3178,8 +3710,7 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev) (1ull << IB_USER_VERBS_CMD_DEALLOC_MW); } - if (MLX5_CAP_GEN(dev->mdev, out_of_seq_cnt) && - MLX5_CAP_GEN(dev->mdev, retransmission_q_counters)) { + 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; } @@ -3233,13 +3764,27 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev) if (err) goto err_rsrc; - err = mlx5_ib_alloc_q_counters(dev); + if (MLX5_CAP_GEN(dev->mdev, max_qp_cnt)) { + err = mlx5_ib_alloc_counters(dev); + if (err) + goto err_odp; + } + + dev->mdev->priv.uar = mlx5_get_uars_page(dev->mdev); + if (!dev->mdev->priv.uar) + goto err_cnt; + + err = mlx5_alloc_bfreg(dev->mdev, &dev->bfreg, false, false); + if (err) + goto err_uar_page; + + err = mlx5_alloc_bfreg(dev->mdev, &dev->fp_bfreg, false, true); if (err) - goto err_odp; + goto err_bfreg; err = ib_register_device(&dev->ib_dev, NULL); if (err) - goto err_q_cnt; + goto err_fp_bfreg; err = create_umr_res(dev); if (err) @@ -3262,8 +3807,18 @@ err_umrc: err_dev: ib_unregister_device(&dev->ib_dev); -err_q_cnt: - mlx5_ib_dealloc_q_counters(dev); +err_fp_bfreg: + mlx5_free_bfreg(dev->mdev, &dev->fp_bfreg); + +err_bfreg: + mlx5_free_bfreg(dev->mdev, &dev->bfreg); + +err_uar_page: + mlx5_put_uars_page(dev->mdev, dev->mdev->priv.uar); + +err_cnt: + if (MLX5_CAP_GEN(dev->mdev, max_qp_cnt)) + mlx5_ib_dealloc_counters(dev); err_odp: mlx5_ib_odp_remove_one(dev); @@ -3293,7 +3848,11 @@ static void mlx5_ib_remove(struct mlx5_core_dev *mdev, void *context) mlx5_remove_netdev_notifier(dev); ib_unregister_device(&dev->ib_dev); - mlx5_ib_dealloc_q_counters(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); + 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); @@ -3307,6 +3866,9 @@ static struct mlx5_interface mlx5_ib_interface = { .add = mlx5_ib_add, .remove = mlx5_ib_remove, .event = mlx5_ib_event, +#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING + .pfault = mlx5_ib_pfault, +#endif .protocol = MLX5_INTERFACE_PROTOCOL_IB, }; @@ -3314,28 +3876,16 @@ static int __init mlx5_ib_init(void) { int err; - if (deprecated_prof_sel != 2) - pr_warn("prof_sel is deprecated for mlx5_ib, set it for mlx5_core\n"); - - err = mlx5_ib_odp_init(); - if (err) - return err; + mlx5_ib_odp_init(); err = mlx5_register_interface(&mlx5_ib_interface); - if (err) - goto clean_odp; - - return err; -clean_odp: - mlx5_ib_odp_cleanup(); return err; } static void __exit mlx5_ib_cleanup(void) { mlx5_unregister_interface(&mlx5_ib_interface); - mlx5_ib_odp_cleanup(); } module_init(mlx5_ib_init); diff --git a/drivers/infiniband/hw/mlx5/mem.c b/drivers/infiniband/hw/mlx5/mem.c index 6851357c16f4..914f212e7ef6 100644 --- a/drivers/infiniband/hw/mlx5/mem.c +++ b/drivers/infiniband/hw/mlx5/mem.c @@ -59,15 +59,14 @@ void mlx5_ib_cont_pages(struct ib_umem *umem, u64 addr, u64 pfn; struct scatterlist *sg; int entry; - unsigned long page_shift = ilog2(umem->page_size); + unsigned long page_shift = umem->page_shift; - /* With ODP we must always match OS page size. */ if (umem->odp_data) { - *count = ib_umem_page_count(umem); - *shift = PAGE_SHIFT; - *ncont = *count; + *ncont = ib_umem_page_count(umem); + *count = *ncont << (page_shift - PAGE_SHIFT); + *shift = page_shift; if (order) - *order = ilog2(roundup_pow_of_two(*count)); + *order = ilog2(roundup_pow_of_two(*ncont)); return; } @@ -156,10 +155,10 @@ void __mlx5_ib_populate_pas(struct mlx5_ib_dev *dev, struct ib_umem *umem, int page_shift, size_t offset, size_t num_pages, __be64 *pas, int access_flags) { - unsigned long umem_page_shift = ilog2(umem->page_size); + unsigned long umem_page_shift = umem->page_shift; int shift = page_shift - umem_page_shift; int mask = (1 << shift) - 1; - int i, k; + int i, k, idx; u64 cur = 0; u64 base; int len; @@ -185,18 +184,36 @@ void __mlx5_ib_populate_pas(struct mlx5_ib_dev *dev, struct ib_umem *umem, for_each_sg(umem->sg_head.sgl, sg, umem->nmap, entry) { len = sg_dma_len(sg) >> umem_page_shift; base = sg_dma_address(sg); - for (k = 0; k < len; k++) { + + /* Skip elements below offset */ + if (i + len < offset << shift) { + i += len; + continue; + } + + /* Skip pages below offset */ + if (i < offset << shift) { + k = (offset << shift) - i; + i = offset << shift; + } else { + k = 0; + } + + for (; k < len; k++) { if (!(i & mask)) { cur = base + (k << umem_page_shift); cur |= access_flags; + idx = (i >> shift) - offset; - pas[i >> shift] = cpu_to_be64(cur); + pas[idx] = cpu_to_be64(cur); mlx5_ib_dbg(dev, "pas[%d] 0x%llx\n", - i >> shift, be64_to_cpu(pas[i >> shift])); - } else - mlx5_ib_dbg(dev, "=====> 0x%llx\n", - base + (k << umem_page_shift)); + i >> shift, be64_to_cpu(pas[idx])); + } i++; + + /* Stop after num_pages reached */ + if (i >> shift >= offset + num_pages) + return; } } } diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h index 6c6057eb60ea..bdcf25410c99 100644 --- a/drivers/infiniband/hw/mlx5/mlx5_ib.h +++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h @@ -90,7 +90,6 @@ enum mlx5_ib_latency_class { MLX5_IB_LATENCY_CLASS_LOW, MLX5_IB_LATENCY_CLASS_MEDIUM, MLX5_IB_LATENCY_CLASS_HIGH, - MLX5_IB_LATENCY_CLASS_FAST_PATH }; enum mlx5_ib_mad_ifc_flags { @@ -100,7 +99,7 @@ enum mlx5_ib_mad_ifc_flags { }; enum { - MLX5_CROSS_CHANNEL_UUAR = 0, + MLX5_CROSS_CHANNEL_BFREG = 0, }; enum { @@ -120,11 +119,16 @@ struct mlx5_ib_ucontext { /* protect doorbell record alloc/free */ struct mutex db_page_mutex; - struct mlx5_uuar_info uuari; + struct mlx5_bfreg_info bfregi; u8 cqe_version; /* Transport Domain number */ u32 tdn; struct list_head vma_private_list; + + unsigned long upd_xlt_page; + /* protect ODP/KSM */ + struct mutex upd_xlt_page_mutex; + u64 lib_caps; }; static inline struct mlx5_ib_ucontext *to_mucontext(struct ib_ucontext *ibucontext) @@ -174,13 +178,12 @@ struct mlx5_ib_flow_db { * enum ib_send_flags and enum ib_qp_type for low-level driver */ -#define MLX5_IB_SEND_UMR_UNREG IB_SEND_RESERVED_START -#define MLX5_IB_SEND_UMR_FAIL_IF_FREE (IB_SEND_RESERVED_START << 1) -#define MLX5_IB_SEND_UMR_UPDATE_MTT (IB_SEND_RESERVED_START << 2) - -#define MLX5_IB_SEND_UMR_UPDATE_TRANSLATION (IB_SEND_RESERVED_START << 3) -#define MLX5_IB_SEND_UMR_UPDATE_PD (IB_SEND_RESERVED_START << 4) -#define MLX5_IB_SEND_UMR_UPDATE_ACCESS IB_SEND_RESERVED_END +#define MLX5_IB_SEND_UMR_ENABLE_MR (IB_SEND_RESERVED_START << 0) +#define MLX5_IB_SEND_UMR_DISABLE_MR (IB_SEND_RESERVED_START << 1) +#define MLX5_IB_SEND_UMR_FAIL_IF_FREE (IB_SEND_RESERVED_START << 2) +#define MLX5_IB_SEND_UMR_UPDATE_XLT (IB_SEND_RESERVED_START << 3) +#define MLX5_IB_SEND_UMR_UPDATE_TRANSLATION (IB_SEND_RESERVED_START << 4) +#define MLX5_IB_SEND_UMR_UPDATE_PD_ACCESS IB_SEND_RESERVED_END #define MLX5_IB_QPT_REG_UMR IB_QPT_RESERVED1 /* @@ -190,6 +193,17 @@ struct mlx5_ib_flow_db { #define MLX5_IB_QPT_HW_GSI IB_QPT_RESERVED2 #define MLX5_IB_WR_UMR IB_WR_RESERVED1 +#define MLX5_IB_UMR_OCTOWORD 16 +#define MLX5_IB_UMR_XLT_ALIGNMENT 64 + +#define MLX5_IB_UPD_XLT_ZAP BIT(0) +#define MLX5_IB_UPD_XLT_ENABLE BIT(1) +#define MLX5_IB_UPD_XLT_ATOMIC BIT(2) +#define MLX5_IB_UPD_XLT_ADDR BIT(3) +#define MLX5_IB_UPD_XLT_PD BIT(4) +#define MLX5_IB_UPD_XLT_ACCESS BIT(5) +#define MLX5_IB_UPD_XLT_INDIRECT BIT(6) + /* Private QP creation flags to be passed in ib_qp_init_attr.create_flags. * * These flags are intended for internal use by the mlx5_ib driver, and they @@ -207,6 +221,10 @@ struct wr_list { u16 next; }; +enum mlx5_ib_rq_flags { + MLX5_IB_RQ_CVLAN_STRIPPING = 1 << 0, +}; + struct mlx5_ib_wq { u64 *wrid; u32 *wr_data; @@ -264,29 +282,6 @@ struct mlx5_ib_rwq_ind_table { u32 rqtn; }; -/* - * Connect-IB can trigger up to four concurrent pagefaults - * per-QP. - */ -enum mlx5_ib_pagefault_context { - MLX5_IB_PAGEFAULT_RESPONDER_READ, - MLX5_IB_PAGEFAULT_REQUESTOR_READ, - MLX5_IB_PAGEFAULT_RESPONDER_WRITE, - MLX5_IB_PAGEFAULT_REQUESTOR_WRITE, - MLX5_IB_PAGEFAULT_CONTEXTS -}; - -static inline enum mlx5_ib_pagefault_context - mlx5_ib_get_pagefault_context(struct mlx5_pagefault *pagefault) -{ - return pagefault->flags & (MLX5_PFAULT_REQUESTOR | MLX5_PFAULT_WRITE); -} - -struct mlx5_ib_pfault { - struct work_struct work; - struct mlx5_pagefault mpfault; -}; - struct mlx5_ib_ubuffer { struct ib_umem *umem; int buf_size; @@ -318,6 +313,7 @@ struct mlx5_ib_rq { struct mlx5_db *doorbell; u32 tirn; u8 state; + u32 flags; }; struct mlx5_ib_sq { @@ -334,6 +330,12 @@ struct mlx5_ib_raw_packet_qp { struct mlx5_ib_rq rq; }; +struct mlx5_bf { + int buf_size; + unsigned long offset; + struct mlx5_sq_bfreg *bfreg; +}; + struct mlx5_ib_qp { struct ib_qp ibqp; union { @@ -347,7 +349,7 @@ struct mlx5_ib_qp { struct mlx5_ib_wq rq; u8 sq_signal_bits; - u8 fm_cache; + u8 next_fence; struct mlx5_ib_wq sq; /* serialize qp state modifications @@ -359,33 +361,19 @@ struct mlx5_ib_qp { int wq_sig; int scat_cqe; int max_inline_data; - struct mlx5_bf *bf; + struct mlx5_bf bf; int has_rq; /* only for user space QPs. For kernel * we have it from the bf object */ - int uuarn; + int bfregn; int create_type; /* Store signature errors */ bool signature_en; -#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING - /* - * A flag that is true for QP's that are in a state that doesn't - * allow page faults, and shouldn't schedule any more faults. - */ - int disable_page_faults; - /* - * The disable_page_faults_lock protects a QP's disable_page_faults - * field, allowing for a thread to atomically check whether the QP - * allows page faults, and if so schedule a page fault. - */ - spinlock_t disable_page_faults_lock; - struct mlx5_ib_pfault pagefaults[MLX5_IB_PAGEFAULT_CONTEXTS]; -#endif struct list_head qps_list; struct list_head cq_recv_list; struct list_head cq_send_list; @@ -410,17 +398,16 @@ enum mlx5_ib_qp_flags { MLX5_IB_QP_SQPN_QP1 = 1 << 6, MLX5_IB_QP_CAP_SCATTER_FCS = 1 << 7, MLX5_IB_QP_RSS = 1 << 8, + MLX5_IB_QP_CVLAN_STRIPPING = 1 << 9, }; struct mlx5_umr_wr { struct ib_send_wr wr; - union { - u64 virt_addr; - u64 offset; - } target; + u64 virt_addr; + u64 offset; struct ib_pd *pd; unsigned int page_shift; - unsigned int npages; + unsigned int xlt_size; u64 length; int access_flags; u32 mkey; @@ -517,11 +504,16 @@ struct mlx5_ib_mr { int live; void *descs_alloc; int access_flags; /* Needed for rereg MR */ + + struct mlx5_ib_mr *parent; + atomic_t num_leaf_free; + wait_queue_head_t q_leaf_free; }; struct mlx5_ib_mw { struct ib_mw ibmw; struct mlx5_core_mkey mmkey; + int ndescs; }; struct mlx5_ib_umr_context { @@ -555,6 +547,10 @@ struct mlx5_cache_ent { struct dentry *dir; char name[4]; u32 order; + u32 xlt; + u32 access_mode; + u32 page; + u32 size; u32 cur; u32 miss; @@ -569,6 +565,7 @@ struct mlx5_cache_ent { struct work_struct work; struct delayed_work dwork; int pending; + struct completion compl; }; struct mlx5_mr_cache { @@ -599,8 +596,16 @@ struct mlx5_ib_resources { struct mutex mutex; }; +struct mlx5_ib_counters { + const char **names; + size_t *offsets; + u32 num_q_counters; + u32 num_cong_counters; + u16 set_id; +}; + struct mlx5_ib_port { - u16 q_cnt_id; + struct mlx5_ib_counters cnts; }; struct mlx5_roce { @@ -617,7 +622,6 @@ struct mlx5_ib_dev { struct ib_device ib_dev; struct mlx5_core_dev *mdev; struct mlx5_roce roce; - MLX5_DECLARE_DOORBELL_LOCK(uar_lock); int num_ports; /* serialize update of capability mask */ @@ -634,11 +638,13 @@ struct mlx5_ib_dev { int fill_delay; #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING struct ib_odp_caps odp_caps; + u64 odp_max_size; /* * Sleepable RCU that prevents destruction of MRs while they are still * being used by a page fault handler. */ struct srcu_struct mr_srcu; + u32 null_mkey; #endif struct mlx5_ib_flow_db flow_db; /* protect resources needed as part of reset flow */ @@ -646,6 +652,9 @@ struct mlx5_ib_dev { struct list_head qp_list; /* Array with num_ports elements */ struct mlx5_ib_port *port; + struct mlx5_sq_bfreg bfreg; + struct mlx5_sq_bfreg fp_bfreg; + u8 umr_fence; }; static inline struct mlx5_ib_cq *to_mibcq(struct mlx5_core_cq *mcq) @@ -723,16 +732,6 @@ static inline struct mlx5_ib_mw *to_mmw(struct ib_mw *ibmw) return container_of(ibmw, struct mlx5_ib_mw, ibmw); } -struct mlx5_ib_ah { - struct ib_ah ibah; - struct mlx5_av av; -}; - -static inline struct mlx5_ib_ah *to_mah(struct ib_ah *ibah) -{ - return container_of(ibah, struct mlx5_ib_ah, ibah); -} - int mlx5_ib_db_map_user(struct mlx5_ib_ucontext *context, unsigned long virt, struct mlx5_db *db); void mlx5_ib_db_unmap_user(struct mlx5_ib_ucontext *context, struct mlx5_db *db); @@ -742,9 +741,9 @@ void mlx5_ib_free_srq_wqe(struct mlx5_ib_srq *srq, int wqe_index); int mlx5_MAD_IFC(struct mlx5_ib_dev *dev, int ignore_mkey, int ignore_bkey, u8 port, const struct ib_wc *in_wc, const struct ib_grh *in_grh, const void *in_mad, void *response_mad); -struct ib_ah *mlx5_ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr, +struct ib_ah *mlx5_ib_create_ah(struct ib_pd *pd, struct rdma_ah_attr *ah_attr, struct ib_udata *udata); -int mlx5_ib_query_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr); +int mlx5_ib_query_ah(struct ib_ah *ibah, struct rdma_ah_attr *ah_attr); int mlx5_ib_destroy_ah(struct ib_ah *ah); struct ib_srq *mlx5_ib_create_srq(struct ib_pd *pd, struct ib_srq_init_attr *init_attr, @@ -787,8 +786,11 @@ struct ib_mr *mlx5_ib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, struct ib_mw *mlx5_ib_alloc_mw(struct ib_pd *pd, enum ib_mw_type type, struct ib_udata *udata); int mlx5_ib_dealloc_mw(struct ib_mw *mw); -int mlx5_ib_update_mtt(struct mlx5_ib_mr *mr, u64 start_page_index, - int npages, int zap); +int mlx5_ib_update_xlt(struct mlx5_ib_mr *mr, u64 idx, int npages, + int page_shift, int flags); +struct mlx5_ib_mr *mlx5_ib_alloc_implicit_mr(struct mlx5_ib_pd *pd, + int access_flags); +void mlx5_ib_free_implicit_mr(struct mlx5_ib_mr *mr); int mlx5_ib_rereg_user_mr(struct ib_mr *ib_mr, int flags, u64 start, u64 length, u64 virt_addr, int access_flags, struct ib_pd *pd, struct ib_udata *udata); @@ -842,7 +844,9 @@ void mlx5_ib_copy_pas(u64 *old, u64 *new, int step, int num); int mlx5_ib_get_cqe_size(struct mlx5_ib_dev *dev, struct ib_cq *ibcq); int mlx5_mr_cache_init(struct mlx5_ib_dev *dev); int mlx5_mr_cache_cleanup(struct mlx5_ib_dev *dev); -int mlx5_mr_ib_cont_pages(struct ib_umem *umem, u64 addr, int *count, int *shift); + +struct mlx5_ib_mr *mlx5_mr_cache_alloc(struct mlx5_ib_dev *dev, int entry); +void mlx5_mr_cache_free(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr); int mlx5_ib_check_mr_status(struct ib_mr *ibmr, u32 check_mask, struct ib_mr_status *mr_status); struct ib_wq *mlx5_ib_create_wq(struct ib_pd *pd, @@ -857,33 +861,32 @@ struct ib_rwq_ind_table *mlx5_ib_create_rwq_ind_table(struct ib_device *device, int mlx5_ib_destroy_rwq_ind_table(struct ib_rwq_ind_table *wq_ind_table); #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING -extern struct workqueue_struct *mlx5_ib_page_fault_wq; - void mlx5_ib_internal_fill_odp_caps(struct mlx5_ib_dev *dev); -void mlx5_ib_mr_pfault_handler(struct mlx5_ib_qp *qp, - struct mlx5_ib_pfault *pfault); -void mlx5_ib_odp_create_qp(struct mlx5_ib_qp *qp); +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_qp_disable_pagefaults(struct mlx5_ib_qp *qp); -void mlx5_ib_qp_enable_pagefaults(struct mlx5_ib_qp *qp); void mlx5_ib_invalidate_range(struct ib_umem *umem, unsigned long start, unsigned long end); +void mlx5_odp_init_mr_cache_entry(struct mlx5_cache_ent *ent); +void mlx5_odp_populate_klm(struct mlx5_klm *pklm, size_t offset, + size_t nentries, struct mlx5_ib_mr *mr, int flags); #else /* CONFIG_INFINIBAND_ON_DEMAND_PAGING */ static inline void mlx5_ib_internal_fill_odp_caps(struct mlx5_ib_dev *dev) { return; } -static inline void mlx5_ib_odp_create_qp(struct mlx5_ib_qp *qp) {} 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 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_ib_qp_disable_pagefaults(struct mlx5_ib_qp *qp) {} -static inline void mlx5_ib_qp_enable_pagefaults(struct mlx5_ib_qp *qp) {} +static inline void mlx5_ib_odp_cleanup(void) {} +static inline void mlx5_odp_init_mr_cache_entry(struct mlx5_cache_ent *ent) {} +static inline void mlx5_odp_populate_klm(struct mlx5_klm *pklm, size_t offset, + size_t nentries, struct mlx5_ib_mr *mr, + int flags) {} #endif /* CONFIG_INFINIBAND_ON_DEMAND_PAGING */ @@ -898,6 +901,8 @@ int mlx5_ib_set_vf_guid(struct ib_device *device, int vf, u8 port, __be16 mlx5_get_roce_udp_sport(struct mlx5_ib_dev *dev, u8 port_num, int index); +int mlx5_get_roce_gid_type(struct mlx5_ib_dev *dev, u8 port_num, + int index, enum ib_gid_type *gid_type); /* GSI QP helper functions */ struct ib_qp *mlx5_ib_gsi_create_qp(struct ib_pd *pd, @@ -1001,4 +1006,17 @@ static inline int get_srq_user_index(struct mlx5_ib_ucontext *ucontext, return verify_assign_uidx(cqe_version, ucmd->uidx, user_index); } + +static inline int get_uars_per_sys_page(struct mlx5_ib_dev *dev, bool lib_support) +{ + return lib_support && MLX5_CAP_GEN(dev->mdev, uar_4k) ? + MLX5_UARS_IN_PAGE : 1; +} + +static inline int get_num_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; +} + #endif /* MLX5_IB_H */ diff --git a/drivers/infiniband/hw/mlx5/mr.c b/drivers/infiniband/hw/mlx5/mr.c index 8f608debe141..2c40a2e989d2 100644 --- a/drivers/infiniband/hw/mlx5/mr.c +++ b/drivers/infiniband/hw/mlx5/mr.c @@ -46,14 +46,10 @@ enum { }; #define MLX5_UMR_ALIGN 2048 -#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING -static __be64 mlx5_ib_update_mtt_emergency_buffer[ - MLX5_UMR_MTT_MIN_CHUNK_SIZE/sizeof(__be64)] - __aligned(MLX5_UMR_ALIGN); -static DEFINE_MUTEX(mlx5_ib_update_mtt_emergency_buffer_mutex); -#endif static int clean_mr(struct mlx5_ib_mr *mr); +static int use_umr(struct mlx5_ib_dev *dev, int order); +static int unreg_umr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr); static int destroy_mkey(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr) { @@ -134,6 +130,7 @@ static void reg_mr_callback(int status, void *context) return; } + mr->mmkey.type = MLX5_MKEY_MR; spin_lock_irqsave(&dev->mdev->priv.mkey_lock, flags); key = dev->mdev->priv.mkey_key++; spin_unlock_irqrestore(&dev->mdev->priv.mkey_lock, flags); @@ -153,6 +150,9 @@ static void reg_mr_callback(int status, void *context) if (err) pr_err("Error inserting to mkey tree. 0x%x\n", -err); write_unlock_irqrestore(&table->lock, flags); + + if (!completion_done(&ent->compl)) + complete(&ent->compl); } static int add_keys(struct mlx5_ib_dev *dev, int c, int num) @@ -161,7 +161,6 @@ static int add_keys(struct mlx5_ib_dev *dev, int c, int num) struct mlx5_cache_ent *ent = &cache->ent[c]; int inlen = MLX5_ST_SZ_BYTES(create_mkey_in); struct mlx5_ib_mr *mr; - int npages = 1 << ent->order; void *mkc; u32 *in; int err = 0; @@ -189,11 +188,11 @@ static int add_keys(struct mlx5_ib_dev *dev, int c, int num) MLX5_SET(mkc, mkc, free, 1); MLX5_SET(mkc, mkc, umr_en, 1); - MLX5_SET(mkc, mkc, access_mode, MLX5_MKC_ACCESS_MODE_MTT); + MLX5_SET(mkc, mkc, access_mode, ent->access_mode); MLX5_SET(mkc, mkc, qpn, 0xffffff); - MLX5_SET(mkc, mkc, translations_octword_size, (npages + 1) / 2); - MLX5_SET(mkc, mkc, log_page_size, 12); + MLX5_SET(mkc, mkc, translations_octword_size, ent->xlt); + MLX5_SET(mkc, mkc, log_page_size, ent->page); spin_lock_irq(&ent->lock); ent->pending++; @@ -451,6 +450,42 @@ static void cache_work_func(struct work_struct *work) __cache_work_func(ent); } +struct mlx5_ib_mr *mlx5_mr_cache_alloc(struct mlx5_ib_dev *dev, int entry) +{ + struct mlx5_mr_cache *cache = &dev->cache; + struct mlx5_cache_ent *ent; + struct mlx5_ib_mr *mr; + int err; + + if (entry < 0 || entry >= MAX_MR_CACHE_ENTRIES) { + mlx5_ib_err(dev, "cache entry %d is out of range\n", entry); + return NULL; + } + + ent = &cache->ent[entry]; + while (1) { + spin_lock_irq(&ent->lock); + if (list_empty(&ent->head)) { + spin_unlock_irq(&ent->lock); + + err = add_keys(dev, entry, 1); + if (err && err != -EAGAIN) + return ERR_PTR(err); + + wait_for_completion(&ent->compl); + } else { + mr = list_first_entry(&ent->head, struct mlx5_ib_mr, + list); + list_del(&mr->list); + ent->cur--; + spin_unlock_irq(&ent->lock); + if (ent->cur < ent->limit) + queue_work(cache->wq, &ent->work); + return mr; + } + } +} + static struct mlx5_ib_mr *alloc_cached_mr(struct mlx5_ib_dev *dev, int order) { struct mlx5_mr_cache *cache = &dev->cache; @@ -460,12 +495,12 @@ static struct mlx5_ib_mr *alloc_cached_mr(struct mlx5_ib_dev *dev, int order) int i; c = order2idx(dev, order); - if (c < 0 || c >= MAX_MR_CACHE_ENTRIES) { + if (c < 0 || c > MAX_UMR_CACHE_ENTRY) { mlx5_ib_warn(dev, "order %d, cache index %d\n", order, c); return NULL; } - for (i = c; i < MAX_MR_CACHE_ENTRIES; i++) { + for (i = c; i < MAX_UMR_CACHE_ENTRY; i++) { ent = &cache->ent[i]; mlx5_ib_dbg(dev, "order %d, cache index %d\n", ent->order, i); @@ -492,7 +527,7 @@ static struct mlx5_ib_mr *alloc_cached_mr(struct mlx5_ib_dev *dev, int order) return mr; } -static void free_cached_mr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr) +void mlx5_mr_cache_free(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr) { struct mlx5_mr_cache *cache = &dev->cache; struct mlx5_cache_ent *ent; @@ -504,6 +539,10 @@ static void free_cached_mr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr) mlx5_ib_warn(dev, "order %d, cache index %d\n", mr->order, c); return; } + + if (unreg_umr(dev, mr)) + return; + ent = &cache->ent[c]; spin_lock_irq(&ent->lock); list_add_tail(&mr->list, &ent->head); @@ -543,6 +582,15 @@ static void clean_keys(struct mlx5_ib_dev *dev, int c) } } +static void mlx5_mr_cache_debugfs_cleanup(struct mlx5_ib_dev *dev) +{ + if (!mlx5_debugfs_root) + return; + + debugfs_remove_recursive(dev->cache.root); + dev->cache.root = NULL; +} + static int mlx5_mr_cache_debugfs_init(struct mlx5_ib_dev *dev) { struct mlx5_mr_cache *cache = &dev->cache; @@ -561,38 +609,34 @@ static int mlx5_mr_cache_debugfs_init(struct mlx5_ib_dev *dev) sprintf(ent->name, "%d", ent->order); ent->dir = debugfs_create_dir(ent->name, cache->root); if (!ent->dir) - return -ENOMEM; + goto err; ent->fsize = debugfs_create_file("size", 0600, ent->dir, ent, &size_fops); if (!ent->fsize) - return -ENOMEM; + goto err; ent->flimit = debugfs_create_file("limit", 0600, ent->dir, ent, &limit_fops); if (!ent->flimit) - return -ENOMEM; + goto err; ent->fcur = debugfs_create_u32("cur", 0400, ent->dir, &ent->cur); if (!ent->fcur) - return -ENOMEM; + goto err; ent->fmiss = debugfs_create_u32("miss", 0600, ent->dir, &ent->miss); if (!ent->fmiss) - return -ENOMEM; + goto err; } return 0; -} - -static void mlx5_mr_cache_debugfs_cleanup(struct mlx5_ib_dev *dev) -{ - if (!mlx5_debugfs_root) - return; +err: + mlx5_mr_cache_debugfs_cleanup(dev); - debugfs_remove_recursive(dev->cache.root); + return -ENOMEM; } static void delay_time_func(unsigned long ctx) @@ -606,7 +650,6 @@ int mlx5_mr_cache_init(struct mlx5_ib_dev *dev) { struct mlx5_mr_cache *cache = &dev->cache; struct mlx5_cache_ent *ent; - int limit; int err; int i; @@ -619,31 +662,46 @@ int mlx5_mr_cache_init(struct mlx5_ib_dev *dev) setup_timer(&dev->delay_timer, delay_time_func, (unsigned long)dev); for (i = 0; i < MAX_MR_CACHE_ENTRIES; i++) { - INIT_LIST_HEAD(&cache->ent[i].head); - spin_lock_init(&cache->ent[i].lock); - ent = &cache->ent[i]; INIT_LIST_HEAD(&ent->head); spin_lock_init(&ent->lock); ent->order = i + 2; ent->dev = dev; + ent->limit = 0; - if ((dev->mdev->profile->mask & MLX5_PROF_MASK_MR_CACHE) && - (mlx5_core_is_pf(dev->mdev))) - limit = dev->mdev->profile->mr_cache[i].limit; - else - limit = 0; - + init_completion(&ent->compl); INIT_WORK(&ent->work, cache_work_func); INIT_DELAYED_WORK(&ent->dwork, delayed_cache_work_func); - ent->limit = limit; queue_work(cache->wq, &ent->work); + + if (i > MAX_UMR_CACHE_ENTRY) { + mlx5_odp_init_mr_cache_entry(ent); + continue; + } + + if (!use_umr(dev, ent->order)) + continue; + + ent->page = PAGE_SHIFT; + ent->xlt = (1 << ent->order) * sizeof(struct mlx5_mtt) / + MLX5_IB_UMR_OCTOWORD; + ent->access_mode = MLX5_MKC_ACCESS_MODE_MTT; + if ((dev->mdev->profile->mask & MLX5_PROF_MASK_MR_CACHE) && + mlx5_core_is_pf(dev->mdev)) + ent->limit = dev->mdev->profile->mr_cache[i].limit; + else + ent->limit = 0; } err = mlx5_mr_cache_debugfs_init(dev); if (err) mlx5_ib_warn(dev, "cache debugfs failure\n"); + /* + * We don't want to fail driver if debugfs failed to initialize, + * so we are not forwarding error to the user. + */ + return 0; } @@ -732,6 +790,7 @@ struct ib_mr *mlx5_ib_get_dma_mr(struct ib_pd *pd, int acc) goto err_in; kfree(in); + mr->mmkey.type = MLX5_MKEY_MR; mr->ibmr.lkey = mr->mmkey.key; mr->ibmr.rkey = mr->mmkey.key; mr->umem = NULL; @@ -757,94 +816,13 @@ static int get_octo_len(u64 addr, u64 len, int page_size) return (npages + 1) / 2; } -static int use_umr(int order) +static int use_umr(struct mlx5_ib_dev *dev, int order) { + if (MLX5_CAP_GEN(dev->mdev, umr_extended_translation_offset)) + return order <= MAX_UMR_CACHE_ENTRY + 2; return order <= MLX5_MAX_UMR_SHIFT; } -static int dma_map_mr_pas(struct mlx5_ib_dev *dev, struct ib_umem *umem, - int npages, int page_shift, int *size, - __be64 **mr_pas, dma_addr_t *dma) -{ - __be64 *pas; - struct device *ddev = dev->ib_dev.dma_device; - - /* - * UMR copies MTTs in units of MLX5_UMR_MTT_ALIGNMENT bytes. - * To avoid copying garbage after the pas array, we allocate - * a little more. - */ - *size = ALIGN(sizeof(u64) * npages, MLX5_UMR_MTT_ALIGNMENT); - *mr_pas = kmalloc(*size + MLX5_UMR_ALIGN - 1, GFP_KERNEL); - if (!(*mr_pas)) - return -ENOMEM; - - pas = PTR_ALIGN(*mr_pas, MLX5_UMR_ALIGN); - mlx5_ib_populate_pas(dev, umem, page_shift, pas, MLX5_IB_MTT_PRESENT); - /* Clear padding after the actual pages. */ - memset(pas + npages, 0, *size - npages * sizeof(u64)); - - *dma = dma_map_single(ddev, pas, *size, DMA_TO_DEVICE); - if (dma_mapping_error(ddev, *dma)) { - kfree(*mr_pas); - return -ENOMEM; - } - - return 0; -} - -static void prep_umr_wqe_common(struct ib_pd *pd, struct ib_send_wr *wr, - struct ib_sge *sg, u64 dma, int n, u32 key, - int page_shift) -{ - struct mlx5_ib_dev *dev = to_mdev(pd->device); - struct mlx5_umr_wr *umrwr = umr_wr(wr); - - sg->addr = dma; - sg->length = ALIGN(sizeof(u64) * n, 64); - sg->lkey = dev->umrc.pd->local_dma_lkey; - - wr->next = NULL; - wr->sg_list = sg; - if (n) - wr->num_sge = 1; - else - wr->num_sge = 0; - - wr->opcode = MLX5_IB_WR_UMR; - - umrwr->npages = n; - umrwr->page_shift = page_shift; - umrwr->mkey = key; -} - -static void prep_umr_reg_wqe(struct ib_pd *pd, struct ib_send_wr *wr, - struct ib_sge *sg, u64 dma, int n, u32 key, - int page_shift, u64 virt_addr, u64 len, - int access_flags) -{ - struct mlx5_umr_wr *umrwr = umr_wr(wr); - - prep_umr_wqe_common(pd, wr, sg, dma, n, key, page_shift); - - wr->send_flags = 0; - - umrwr->target.virt_addr = virt_addr; - umrwr->length = len; - umrwr->access_flags = access_flags; - umrwr->pd = pd; -} - -static void prep_umr_unreg_wqe(struct mlx5_ib_dev *dev, - struct ib_send_wr *wr, u32 key) -{ - struct mlx5_umr_wr *umrwr = umr_wr(wr); - - wr->send_flags = MLX5_IB_SEND_UMR_UNREG | MLX5_IB_SEND_UMR_FAIL_IF_FREE; - wr->opcode = MLX5_IB_WR_UMR; - umrwr->mkey = key; -} - static int mr_umem_get(struct ib_pd *pd, u64 start, u64 length, int access_flags, struct ib_umem **umem, int *npages, int *page_shift, int *ncont, @@ -857,7 +835,7 @@ static int mr_umem_get(struct ib_pd *pd, u64 start, u64 length, access_flags, 0); err = PTR_ERR_OR_ZERO(*umem); if (err < 0) { - mlx5_ib_err(dev, "umem get failed (%ld)\n", PTR_ERR(umem)); + mlx5_ib_err(dev, "umem get failed (%d)\n", err); return err; } @@ -891,21 +869,39 @@ static inline void mlx5_ib_init_umr_context(struct mlx5_ib_umr_context *context) init_completion(&context->done); } +static int mlx5_ib_post_send_wait(struct mlx5_ib_dev *dev, + struct mlx5_umr_wr *umrwr) +{ + struct umr_common *umrc = &dev->umrc; + struct ib_send_wr *bad; + int err; + struct mlx5_ib_umr_context umr_context; + + mlx5_ib_init_umr_context(&umr_context); + umrwr->wr.wr_cqe = &umr_context.cqe; + + down(&umrc->sem); + err = ib_post_send(umrc->qp, &umrwr->wr, &bad); + if (err) { + mlx5_ib_warn(dev, "UMR post send failed, err %d\n", err); + } else { + wait_for_completion(&umr_context.done); + if (umr_context.status != IB_WC_SUCCESS) { + mlx5_ib_warn(dev, "reg umr failed (%u)\n", + umr_context.status); + err = -EFAULT; + } + } + up(&umrc->sem); + return err; +} + static struct mlx5_ib_mr *reg_umr(struct ib_pd *pd, struct ib_umem *umem, u64 virt_addr, u64 len, int npages, int page_shift, int order, int access_flags) { struct mlx5_ib_dev *dev = to_mdev(pd->device); - struct device *ddev = dev->ib_dev.dma_device; - struct umr_common *umrc = &dev->umrc; - struct mlx5_ib_umr_context umr_context; - struct mlx5_umr_wr umrwr = {}; - struct ib_send_wr *bad; struct mlx5_ib_mr *mr; - struct ib_sge sg; - int size; - __be64 *mr_pas; - dma_addr_t dma; int err = 0; int i; @@ -924,173 +920,181 @@ static struct mlx5_ib_mr *reg_umr(struct ib_pd *pd, struct ib_umem *umem, if (!mr) return ERR_PTR(-EAGAIN); - err = dma_map_mr_pas(dev, umem, npages, page_shift, &size, &mr_pas, - &dma); - if (err) - goto free_mr; - - mlx5_ib_init_umr_context(&umr_context); - - umrwr.wr.wr_cqe = &umr_context.cqe; - prep_umr_reg_wqe(pd, &umrwr.wr, &sg, dma, npages, mr->mmkey.key, - page_shift, virt_addr, len, access_flags); - - down(&umrc->sem); - err = ib_post_send(umrc->qp, &umrwr.wr, &bad); - if (err) { - mlx5_ib_warn(dev, "post send failed, err %d\n", err); - goto unmap_dma; - } else { - wait_for_completion(&umr_context.done); - if (umr_context.status != IB_WC_SUCCESS) { - mlx5_ib_warn(dev, "reg umr failed\n"); - err = -EFAULT; - } - } - + mr->ibmr.pd = pd; + mr->umem = umem; + mr->access_flags = access_flags; + mr->desc_size = sizeof(struct mlx5_mtt); mr->mmkey.iova = virt_addr; mr->mmkey.size = len; mr->mmkey.pd = to_mpd(pd)->pdn; - mr->live = 1; - -unmap_dma: - up(&umrc->sem); - dma_unmap_single(ddev, dma, size, DMA_TO_DEVICE); - - kfree(mr_pas); + err = mlx5_ib_update_xlt(mr, 0, npages, page_shift, + MLX5_IB_UPD_XLT_ENABLE); -free_mr: if (err) { - free_cached_mr(dev, mr); + mlx5_mr_cache_free(dev, mr); return ERR_PTR(err); } + mr->live = 1; + return mr; } -#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING -int mlx5_ib_update_mtt(struct mlx5_ib_mr *mr, u64 start_page_index, int npages, - int zap) +static inline int populate_xlt(struct mlx5_ib_mr *mr, int idx, int npages, + void *xlt, int page_shift, size_t size, + int flags) { struct mlx5_ib_dev *dev = mr->dev; - struct device *ddev = dev->ib_dev.dma_device; - struct umr_common *umrc = &dev->umrc; - struct mlx5_ib_umr_context umr_context; struct ib_umem *umem = mr->umem; + if (flags & MLX5_IB_UPD_XLT_INDIRECT) { + mlx5_odp_populate_klm(xlt, idx, npages, mr, flags); + return npages; + } + + npages = min_t(size_t, npages, ib_umem_num_pages(umem) - idx); + + if (!(flags & MLX5_IB_UPD_XLT_ZAP)) { + __mlx5_ib_populate_pas(dev, umem, page_shift, + idx, npages, xlt, + MLX5_IB_MTT_PRESENT); + /* Clear padding after the pages + * brought from the umem. + */ + memset(xlt + (npages * sizeof(struct mlx5_mtt)), 0, + size - npages * sizeof(struct mlx5_mtt)); + } + + return npages; +} + +#define MLX5_MAX_UMR_CHUNK ((1 << (MLX5_MAX_UMR_SHIFT + 4)) - \ + MLX5_UMR_MTT_ALIGNMENT) +#define MLX5_SPARE_UMR_CHUNK 0x10000 + +int mlx5_ib_update_xlt(struct mlx5_ib_mr *mr, u64 idx, int npages, + int page_shift, int flags) +{ + struct mlx5_ib_dev *dev = mr->dev; + struct device *ddev = dev->ib_dev.dev.parent; + struct mlx5_ib_ucontext *uctx = NULL; int size; - __be64 *pas; + void *xlt; dma_addr_t dma; - struct ib_send_wr *bad; struct mlx5_umr_wr wr; struct ib_sge sg; int err = 0; - const int page_index_alignment = MLX5_UMR_MTT_ALIGNMENT / sizeof(u64); - const int page_index_mask = page_index_alignment - 1; + int desc_size = (flags & MLX5_IB_UPD_XLT_INDIRECT) + ? sizeof(struct mlx5_klm) + : sizeof(struct mlx5_mtt); + const int page_align = MLX5_UMR_MTT_ALIGNMENT / desc_size; + const int page_mask = page_align - 1; size_t pages_mapped = 0; size_t pages_to_map = 0; size_t pages_iter = 0; - int use_emergency_buf = 0; + gfp_t gfp; /* UMR copies MTTs in units of MLX5_UMR_MTT_ALIGNMENT bytes, - * so we need to align the offset and length accordingly */ - if (start_page_index & page_index_mask) { - npages += start_page_index & page_index_mask; - start_page_index &= ~page_index_mask; + * so we need to align the offset and length accordingly + */ + if (idx & page_mask) { + npages += idx & page_mask; + idx &= ~page_mask; } - pages_to_map = ALIGN(npages, page_index_alignment); + gfp = flags & MLX5_IB_UPD_XLT_ATOMIC ? GFP_ATOMIC : GFP_KERNEL; + gfp |= __GFP_ZERO | __GFP_NOWARN; - if (start_page_index + pages_to_map > MLX5_MAX_UMR_PAGES) - return -EINVAL; + pages_to_map = ALIGN(npages, page_align); + size = desc_size * pages_to_map; + size = min_t(int, size, MLX5_MAX_UMR_CHUNK); - size = sizeof(u64) * pages_to_map; - size = min_t(int, PAGE_SIZE, size); - /* We allocate with GFP_ATOMIC to avoid recursion into page-reclaim - * code, when we are called from an invalidation. The pas buffer must - * be 2k-aligned for Connect-IB. */ - pas = (__be64 *)get_zeroed_page(GFP_ATOMIC); - if (!pas) { - mlx5_ib_warn(dev, "unable to allocate memory during MTT update, falling back to slower chunked mechanism.\n"); - pas = mlx5_ib_update_mtt_emergency_buffer; - size = MLX5_UMR_MTT_MIN_CHUNK_SIZE; - use_emergency_buf = 1; - mutex_lock(&mlx5_ib_update_mtt_emergency_buffer_mutex); - memset(pas, 0, size); + xlt = (void *)__get_free_pages(gfp, get_order(size)); + if (!xlt && size > MLX5_SPARE_UMR_CHUNK) { + mlx5_ib_dbg(dev, "Failed to allocate %d bytes of order %d. fallback to spare UMR allocation od %d bytes\n", + size, get_order(size), MLX5_SPARE_UMR_CHUNK); + + size = MLX5_SPARE_UMR_CHUNK; + xlt = (void *)__get_free_pages(gfp, get_order(size)); + } + + if (!xlt) { + uctx = to_mucontext(mr->ibmr.pd->uobject->context); + mlx5_ib_warn(dev, "Using XLT emergency buffer\n"); + size = PAGE_SIZE; + xlt = (void *)uctx->upd_xlt_page; + mutex_lock(&uctx->upd_xlt_page_mutex); + memset(xlt, 0, size); } - pages_iter = size / sizeof(u64); - dma = dma_map_single(ddev, pas, size, DMA_TO_DEVICE); + pages_iter = size / desc_size; + dma = dma_map_single(ddev, xlt, size, DMA_TO_DEVICE); if (dma_mapping_error(ddev, dma)) { - mlx5_ib_err(dev, "unable to map DMA during MTT update.\n"); + mlx5_ib_err(dev, "unable to map DMA during XLT update.\n"); err = -ENOMEM; - goto free_pas; + goto free_xlt; } + sg.addr = dma; + sg.lkey = dev->umrc.pd->local_dma_lkey; + + memset(&wr, 0, sizeof(wr)); + wr.wr.send_flags = MLX5_IB_SEND_UMR_UPDATE_XLT; + if (!(flags & MLX5_IB_UPD_XLT_ENABLE)) + wr.wr.send_flags |= MLX5_IB_SEND_UMR_FAIL_IF_FREE; + wr.wr.sg_list = &sg; + wr.wr.num_sge = 1; + wr.wr.opcode = MLX5_IB_WR_UMR; + + wr.pd = mr->ibmr.pd; + wr.mkey = mr->mmkey.key; + wr.length = mr->mmkey.size; + wr.virt_addr = mr->mmkey.iova; + wr.access_flags = mr->access_flags; + wr.page_shift = page_shift; + for (pages_mapped = 0; pages_mapped < pages_to_map && !err; - pages_mapped += pages_iter, start_page_index += pages_iter) { + pages_mapped += pages_iter, idx += pages_iter) { + npages = min_t(int, pages_iter, pages_to_map - pages_mapped); dma_sync_single_for_cpu(ddev, dma, size, DMA_TO_DEVICE); - - npages = min_t(size_t, - pages_iter, - ib_umem_num_pages(umem) - start_page_index); - - if (!zap) { - __mlx5_ib_populate_pas(dev, umem, PAGE_SHIFT, - start_page_index, npages, pas, - MLX5_IB_MTT_PRESENT); - /* Clear padding after the pages brought from the - * umem. */ - memset(pas + npages, 0, size - npages * sizeof(u64)); - } + npages = populate_xlt(mr, idx, npages, xlt, + page_shift, size, flags); dma_sync_single_for_device(ddev, dma, size, DMA_TO_DEVICE); - mlx5_ib_init_umr_context(&umr_context); - - memset(&wr, 0, sizeof(wr)); - wr.wr.wr_cqe = &umr_context.cqe; - - sg.addr = dma; - sg.length = ALIGN(npages * sizeof(u64), - MLX5_UMR_MTT_ALIGNMENT); - sg.lkey = dev->umrc.pd->local_dma_lkey; + sg.length = ALIGN(npages * desc_size, + MLX5_UMR_MTT_ALIGNMENT); + + if (pages_mapped + pages_iter >= pages_to_map) { + if (flags & MLX5_IB_UPD_XLT_ENABLE) + wr.wr.send_flags |= + MLX5_IB_SEND_UMR_ENABLE_MR | + MLX5_IB_SEND_UMR_UPDATE_PD_ACCESS | + MLX5_IB_SEND_UMR_UPDATE_TRANSLATION; + if (flags & MLX5_IB_UPD_XLT_PD || + flags & MLX5_IB_UPD_XLT_ACCESS) + wr.wr.send_flags |= + MLX5_IB_SEND_UMR_UPDATE_PD_ACCESS; + if (flags & MLX5_IB_UPD_XLT_ADDR) + wr.wr.send_flags |= + MLX5_IB_SEND_UMR_UPDATE_TRANSLATION; + } - wr.wr.send_flags = MLX5_IB_SEND_UMR_FAIL_IF_FREE | - MLX5_IB_SEND_UMR_UPDATE_MTT; - wr.wr.sg_list = &sg; - wr.wr.num_sge = 1; - wr.wr.opcode = MLX5_IB_WR_UMR; - wr.npages = sg.length / sizeof(u64); - wr.page_shift = PAGE_SHIFT; - wr.mkey = mr->mmkey.key; - wr.target.offset = start_page_index; + wr.offset = idx * desc_size; + wr.xlt_size = sg.length; - down(&umrc->sem); - err = ib_post_send(umrc->qp, &wr.wr, &bad); - if (err) { - mlx5_ib_err(dev, "UMR post send failed, err %d\n", err); - } else { - wait_for_completion(&umr_context.done); - if (umr_context.status != IB_WC_SUCCESS) { - mlx5_ib_err(dev, "UMR completion failed, code %d\n", - umr_context.status); - err = -EFAULT; - } - } - up(&umrc->sem); + err = mlx5_ib_post_send_wait(dev, &wr); } dma_unmap_single(ddev, dma, size, DMA_TO_DEVICE); -free_pas: - if (!use_emergency_buf) - free_page((unsigned long)pas); +free_xlt: + if (uctx) + mutex_unlock(&uctx->upd_xlt_page_mutex); else - mutex_unlock(&mlx5_ib_update_mtt_emergency_buffer_mutex); + free_pages((unsigned long)xlt, get_order(size)); return err; } -#endif /* * If ibmr is NULL it will be allocated by reg_create. @@ -1116,14 +1120,15 @@ static struct mlx5_ib_mr *reg_create(struct ib_mr *ibmr, struct ib_pd *pd, inlen = MLX5_ST_SZ_BYTES(create_mkey_in) + sizeof(*pas) * ((npages + 1) / 2) * 2; - in = mlx5_vzalloc(inlen); + in = kvzalloc(inlen, GFP_KERNEL); if (!in) { err = -ENOMEM; goto err_1; } pas = (__be64 *)MLX5_ADDR_OF(create_mkey_in, in, klm_pas_mtt); - mlx5_ib_populate_pas(dev, umem, page_shift, pas, - pg_cap ? MLX5_IB_MTT_PRESENT : 0); + if (!(access_flags & IB_ACCESS_ON_DEMAND)) + mlx5_ib_populate_pas(dev, umem, page_shift, pas, + pg_cap ? MLX5_IB_MTT_PRESENT : 0); /* The pg_access bit allows setting the access flags * in the page list submitted with the command. */ @@ -1153,6 +1158,8 @@ static struct mlx5_ib_mr *reg_create(struct ib_mr *ibmr, struct ib_pd *pd, mlx5_ib_warn(dev, "create mkey failed\n"); goto err_2; } + mr->mmkey.type = MLX5_MKEY_MR; + mr->desc_size = sizeof(struct mlx5_mtt); mr->umem = umem; mr->dev = dev; mr->live = 1; @@ -1198,20 +1205,33 @@ struct ib_mr *mlx5_ib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, 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); + +#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING + if (!start && length == U64_MAX) { + if (!(access_flags & IB_ACCESS_ON_DEMAND) || + !(dev->odp_caps.general_caps & IB_ODP_SUPPORT_IMPLICIT)) + return ERR_PTR(-EINVAL); + + mr = mlx5_ib_alloc_implicit_mr(to_mpd(pd), access_flags); + return &mr->ibmr; + } +#endif + err = mr_umem_get(pd, start, length, access_flags, &umem, &npages, &page_shift, &ncont, &order); if (err < 0) return ERR_PTR(err); - if (use_umr(order)) { + if (use_umr(dev, order)) { mr = reg_umr(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", order); mr = NULL; } - } else if (access_flags & IB_ACCESS_ON_DEMAND) { + } else if (access_flags & IB_ACCESS_ON_DEMAND && + !MLX5_CAP_GEN(dev->mdev, umr_extended_translation_offset)) { err = -EINVAL; pr_err("Got MR registration for ODP MR > 512MB, not supported for Connect-IB"); goto error; @@ -1248,106 +1268,39 @@ error: static int unreg_umr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr) { struct mlx5_core_dev *mdev = dev->mdev; - struct umr_common *umrc = &dev->umrc; - struct mlx5_ib_umr_context umr_context; struct mlx5_umr_wr umrwr = {}; - struct ib_send_wr *bad; - int err; if (mdev->state == MLX5_DEVICE_STATE_INTERNAL_ERROR) return 0; - mlx5_ib_init_umr_context(&umr_context); - - umrwr.wr.wr_cqe = &umr_context.cqe; - prep_umr_unreg_wqe(dev, &umrwr.wr, mr->mmkey.key); - - down(&umrc->sem); - err = ib_post_send(umrc->qp, &umrwr.wr, &bad); - if (err) { - up(&umrc->sem); - mlx5_ib_dbg(dev, "err %d\n", err); - goto error; - } else { - wait_for_completion(&umr_context.done); - up(&umrc->sem); - } - if (umr_context.status != IB_WC_SUCCESS) { - mlx5_ib_warn(dev, "unreg umr failed\n"); - err = -EFAULT; - goto error; - } - return 0; + umrwr.wr.send_flags = MLX5_IB_SEND_UMR_DISABLE_MR | + MLX5_IB_SEND_UMR_FAIL_IF_FREE; + umrwr.wr.opcode = MLX5_IB_WR_UMR; + umrwr.mkey = mr->mmkey.key; -error: - return err; + return mlx5_ib_post_send_wait(dev, &umrwr); } -static int rereg_umr(struct ib_pd *pd, struct mlx5_ib_mr *mr, u64 virt_addr, - u64 length, int npages, int page_shift, int order, +static int rereg_umr(struct ib_pd *pd, struct mlx5_ib_mr *mr, int access_flags, int flags) { struct mlx5_ib_dev *dev = to_mdev(pd->device); - struct device *ddev = dev->ib_dev.dma_device; - struct mlx5_ib_umr_context umr_context; - struct ib_send_wr *bad; struct mlx5_umr_wr umrwr = {}; - struct ib_sge sg; - struct umr_common *umrc = &dev->umrc; - dma_addr_t dma = 0; - __be64 *mr_pas = NULL; - int size; int err; - mlx5_ib_init_umr_context(&umr_context); - - umrwr.wr.wr_cqe = &umr_context.cqe; umrwr.wr.send_flags = MLX5_IB_SEND_UMR_FAIL_IF_FREE; - if (flags & IB_MR_REREG_TRANS) { - err = dma_map_mr_pas(dev, mr->umem, npages, page_shift, &size, - &mr_pas, &dma); - if (err) - return err; - - umrwr.target.virt_addr = virt_addr; - umrwr.length = length; - umrwr.wr.send_flags |= MLX5_IB_SEND_UMR_UPDATE_TRANSLATION; - } - - prep_umr_wqe_common(pd, &umrwr.wr, &sg, dma, npages, mr->mmkey.key, - page_shift); + umrwr.wr.opcode = MLX5_IB_WR_UMR; + umrwr.mkey = mr->mmkey.key; - if (flags & IB_MR_REREG_PD) { + if (flags & IB_MR_REREG_PD || flags & IB_MR_REREG_ACCESS) { umrwr.pd = pd; - umrwr.wr.send_flags |= MLX5_IB_SEND_UMR_UPDATE_PD; - } - - if (flags & IB_MR_REREG_ACCESS) { umrwr.access_flags = access_flags; - umrwr.wr.send_flags |= MLX5_IB_SEND_UMR_UPDATE_ACCESS; + umrwr.wr.send_flags |= MLX5_IB_SEND_UMR_UPDATE_PD_ACCESS; } - /* post send request to UMR QP */ - down(&umrc->sem); - err = ib_post_send(umrc->qp, &umrwr.wr, &bad); - - if (err) { - mlx5_ib_warn(dev, "post send failed, err %d\n", err); - } else { - wait_for_completion(&umr_context.done); - if (umr_context.status != IB_WC_SUCCESS) { - mlx5_ib_warn(dev, "reg umr failed (%u)\n", - umr_context.status); - err = -EFAULT; - } - } + err = mlx5_ib_post_send_wait(dev, &umrwr); - up(&umrc->sem); - if (flags & IB_MR_REREG_TRANS) { - dma_unmap_single(ddev, dma, size, DMA_TO_DEVICE); - kfree(mr_pas); - } return err; } @@ -1364,6 +1317,7 @@ int mlx5_ib_rereg_user_mr(struct ib_mr *ib_mr, int flags, u64 start, u64 addr = (flags & IB_MR_REREG_TRANS) ? virt_addr : mr->umem->address; u64 len = (flags & IB_MR_REREG_TRANS) ? length : mr->umem->length; int page_shift = 0; + int upd_flags = 0; int npages = 0; int ncont = 0; int order = 0; @@ -1372,6 +1326,8 @@ int mlx5_ib_rereg_user_mr(struct ib_mr *ib_mr, int flags, u64 start, 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); + atomic_sub(mr->npages, &dev->mdev->priv.reg_pages); + if (flags != IB_MR_REREG_PD) { /* * Replace umem. This needs to be done whether or not UMR is @@ -1382,7 +1338,7 @@ int mlx5_ib_rereg_user_mr(struct ib_mr *ib_mr, int flags, u64 start, err = mr_umem_get(pd, addr, len, access_flags, &mr->umem, &npages, &page_shift, &ncont, &order); if (err < 0) { - mr->umem = NULL; + clean_mr(mr); return err; } } @@ -1414,32 +1370,37 @@ int mlx5_ib_rereg_user_mr(struct ib_mr *ib_mr, int flags, u64 start, /* * Send a UMR WQE */ - err = rereg_umr(pd, mr, addr, len, npages, page_shift, - order, access_flags, flags); + mr->ibmr.pd = pd; + mr->access_flags = access_flags; + mr->mmkey.iova = addr; + mr->mmkey.size = len; + mr->mmkey.pd = to_mpd(pd)->pdn; + + if (flags & IB_MR_REREG_TRANS) { + upd_flags = MLX5_IB_UPD_XLT_ADDR; + if (flags & IB_MR_REREG_PD) + upd_flags |= MLX5_IB_UPD_XLT_PD; + if (flags & IB_MR_REREG_ACCESS) + upd_flags |= MLX5_IB_UPD_XLT_ACCESS; + err = mlx5_ib_update_xlt(mr, 0, npages, page_shift, + upd_flags); + } else { + err = rereg_umr(pd, mr, access_flags, flags); + } + if (err) { mlx5_ib_warn(dev, "Failed to rereg UMR\n"); + ib_umem_release(mr->umem); + clean_mr(mr); return err; } } - if (flags & IB_MR_REREG_PD) { - ib_mr->pd = pd; - mr->mmkey.pd = to_mpd(pd)->pdn; - } - - if (flags & IB_MR_REREG_ACCESS) - mr->access_flags = access_flags; + set_mr_fileds(dev, mr, npages, len, access_flags); - if (flags & IB_MR_REREG_TRANS) { - atomic_sub(mr->npages, &dev->mdev->priv.reg_pages); - set_mr_fileds(dev, mr, npages, len, access_flags); - mr->mmkey.iova = addr; - mr->mmkey.size = len; - } #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING update_odp_mr(mr); #endif - return 0; } @@ -1461,9 +1422,9 @@ mlx5_alloc_priv_descs(struct ib_device *device, mr->descs = PTR_ALIGN(mr->descs_alloc, MLX5_UMR_ALIGN); - mr->desc_map = dma_map_single(device->dma_device, mr->descs, + mr->desc_map = dma_map_single(device->dev.parent, mr->descs, size, DMA_TO_DEVICE); - if (dma_mapping_error(device->dma_device, mr->desc_map)) { + if (dma_mapping_error(device->dev.parent, mr->desc_map)) { ret = -ENOMEM; goto err; } @@ -1482,7 +1443,7 @@ mlx5_free_priv_descs(struct mlx5_ib_mr *mr) struct ib_device *device = mr->ibmr.device; int size = mr->max_descs * mr->desc_size; - dma_unmap_single(device->dma_device, mr->desc_map, + dma_unmap_single(device->dev.parent, mr->desc_map, size, DMA_TO_DEVICE); kfree(mr->descs_alloc); mr->descs = NULL; @@ -1518,12 +1479,7 @@ static int clean_mr(struct mlx5_ib_mr *mr) return err; } } else { - err = unreg_umr(dev, mr); - if (err) { - mlx5_ib_warn(dev, "failed unregister\n"); - return err; - } - free_cached_mr(dev, mr); + mlx5_mr_cache_free(dev, mr); } if (!umred) @@ -1546,8 +1502,11 @@ int mlx5_ib_dereg_mr(struct ib_mr *ibmr) /* Wait for all running page-fault handlers to finish. */ synchronize_srcu(&dev->mr_srcu); /* Destroy all page mappings */ - mlx5_ib_invalidate_range(umem, ib_umem_start(umem), - ib_umem_end(umem)); + if (umem->odp_data->page_list) + mlx5_ib_invalidate_range(umem, ib_umem_start(umem), + ib_umem_end(umem)); + else + mlx5_ib_free_implicit_mr(mr); /* * We kill the umem before the MR for ODP, * so that there will not be any invalidations in @@ -1603,11 +1562,11 @@ struct ib_mr *mlx5_ib_alloc_mr(struct ib_pd *pd, mr->access_mode = MLX5_MKC_ACCESS_MODE_MTT; MLX5_SET(mkc, mkc, log_page_size, PAGE_SHIFT); err = mlx5_alloc_priv_descs(pd->device, mr, - ndescs, sizeof(u64)); + ndescs, sizeof(struct mlx5_mtt)); if (err) goto err_free_in; - mr->desc_size = sizeof(u64); + mr->desc_size = sizeof(struct mlx5_mtt); mr->max_descs = ndescs; } else if (mr_type == IB_MR_TYPE_SG_GAPS) { mr->access_mode = MLX5_MKC_ACCESS_MODE_KLMS; @@ -1656,6 +1615,7 @@ struct ib_mr *mlx5_ib_alloc_mr(struct ib_pd *pd, if (err) goto err_destroy_psv; + mr->mmkey.type = MLX5_MKEY_MR; mr->ibmr.lkey = mr->mmkey.key; mr->ibmr.rkey = mr->mmkey.key; mr->umem = NULL; @@ -1736,7 +1696,9 @@ struct ib_mw *mlx5_ib_alloc_mw(struct ib_pd *pd, enum ib_mw_type type, if (err) goto free; + mw->mmkey.type = MLX5_MKEY_MW; mw->ibmw.rkey = mw->mmkey.key; + mw->ndescs = ndescs; resp.response_length = min(offsetof(typeof(resp), response_length) + sizeof(resp.response_length), udata->outlen); @@ -1827,12 +1789,12 @@ mlx5_ib_sg_to_klms(struct mlx5_ib_mr *mr, mr->ndescs = sg_nents; for_each_sg(sgl, sg, sg_nents, i) { - if (unlikely(i > mr->max_descs)) + if (unlikely(i >= mr->max_descs)) break; klms[i].va = cpu_to_be64(sg_dma_address(sg) + sg_offset); klms[i].bcount = cpu_to_be32(sg_dma_len(sg) - sg_offset); klms[i].key = cpu_to_be32(lkey); - mr->ibmr.length += sg_dma_len(sg); + mr->ibmr.length += sg_dma_len(sg) - sg_offset; sg_offset = 0; } diff --git a/drivers/infiniband/hw/mlx5/odp.c b/drivers/infiniband/hw/mlx5/odp.c index cacb631a7b0a..ae0746754008 100644 --- a/drivers/infiniband/hw/mlx5/odp.c +++ b/drivers/infiniband/hw/mlx5/odp.c @@ -34,6 +34,7 @@ #include <rdma/ib_umem_odp.h> #include "mlx5_ib.h" +#include "cmd.h" #define MAX_PREFETCH_LEN (4*1024*1024U) @@ -41,13 +42,139 @@ * a pagefault. */ #define MMU_NOTIFIER_TIMEOUT 1000 -struct workqueue_struct *mlx5_ib_page_fault_wq; +#define MLX5_IMR_MTT_BITS (30 - PAGE_SHIFT) +#define MLX5_IMR_MTT_SHIFT (MLX5_IMR_MTT_BITS + PAGE_SHIFT) +#define MLX5_IMR_MTT_ENTRIES BIT_ULL(MLX5_IMR_MTT_BITS) +#define MLX5_IMR_MTT_SIZE BIT_ULL(MLX5_IMR_MTT_SHIFT) +#define MLX5_IMR_MTT_MASK (~(MLX5_IMR_MTT_SIZE - 1)) + +#define MLX5_KSM_PAGE_SHIFT MLX5_IMR_MTT_SHIFT + +static u64 mlx5_imr_ksm_entries; + +static int check_parent(struct ib_umem_odp *odp, + struct mlx5_ib_mr *parent) +{ + struct mlx5_ib_mr *mr = odp->private; + + return mr && mr->parent == parent && !odp->dying; +} + +static struct ib_umem_odp *odp_next(struct ib_umem_odp *odp) +{ + struct mlx5_ib_mr *mr = odp->private, *parent = mr->parent; + struct ib_ucontext *ctx = odp->umem->context; + struct rb_node *rb; + + down_read(&ctx->umem_rwsem); + while (1) { + rb = rb_next(&odp->interval_tree.rb); + if (!rb) + goto not_found; + odp = rb_entry(rb, struct ib_umem_odp, interval_tree.rb); + if (check_parent(odp, parent)) + goto end; + } +not_found: + odp = NULL; +end: + up_read(&ctx->umem_rwsem); + return odp; +} + +static struct ib_umem_odp *odp_lookup(struct ib_ucontext *ctx, + u64 start, u64 length, + struct mlx5_ib_mr *parent) +{ + struct ib_umem_odp *odp; + struct rb_node *rb; + + down_read(&ctx->umem_rwsem); + odp = rbt_ib_umem_lookup(&ctx->umem_tree, start, length); + if (!odp) + goto end; + + while (1) { + if (check_parent(odp, parent)) + goto end; + rb = rb_next(&odp->interval_tree.rb); + if (!rb) + goto not_found; + odp = rb_entry(rb, struct ib_umem_odp, interval_tree.rb); + if (ib_umem_start(odp->umem) > start + length) + goto not_found; + } +not_found: + odp = NULL; +end: + up_read(&ctx->umem_rwsem); + return odp; +} + +void mlx5_odp_populate_klm(struct mlx5_klm *pklm, size_t offset, + size_t nentries, struct mlx5_ib_mr *mr, int flags) +{ + struct ib_pd *pd = mr->ibmr.pd; + struct ib_ucontext *ctx = pd->uobject->context; + struct mlx5_ib_dev *dev = to_mdev(pd->device); + struct ib_umem_odp *odp; + unsigned long va; + int i; + + if (flags & MLX5_IB_UPD_XLT_ZAP) { + for (i = 0; i < nentries; i++, pklm++) { + pklm->bcount = cpu_to_be32(MLX5_IMR_MTT_SIZE); + pklm->key = cpu_to_be32(dev->null_mkey); + pklm->va = 0; + } + return; + } + + odp = odp_lookup(ctx, offset * MLX5_IMR_MTT_SIZE, + nentries * MLX5_IMR_MTT_SIZE, mr); + + for (i = 0; i < nentries; i++, pklm++) { + pklm->bcount = cpu_to_be32(MLX5_IMR_MTT_SIZE); + va = (offset + i) * MLX5_IMR_MTT_SIZE; + if (odp && odp->umem->address == va) { + struct mlx5_ib_mr *mtt = odp->private; + + pklm->key = cpu_to_be32(mtt->ibmr.lkey); + odp = odp_next(odp); + } else { + pklm->key = cpu_to_be32(dev->null_mkey); + } + mlx5_ib_dbg(dev, "[%d] va %lx key %x\n", + i, va, be32_to_cpu(pklm->key)); + } +} + +static void mr_leaf_free_action(struct work_struct *work) +{ + struct ib_umem_odp *odp = container_of(work, struct ib_umem_odp, work); + int idx = ib_umem_start(odp->umem) >> MLX5_IMR_MTT_SHIFT; + struct mlx5_ib_mr *mr = odp->private, *imr = mr->parent; + + mr->parent = NULL; + synchronize_srcu(&mr->dev->mr_srcu); + + ib_umem_release(odp->umem); + if (imr->live) + mlx5_ib_update_xlt(imr, idx, 1, 0, + MLX5_IB_UPD_XLT_INDIRECT | + MLX5_IB_UPD_XLT_ATOMIC); + mlx5_mr_cache_free(mr->dev, mr); + + if (atomic_dec_and_test(&imr->num_leaf_free)) + wake_up(&imr->q_leaf_free); +} void mlx5_ib_invalidate_range(struct ib_umem *umem, unsigned long start, unsigned long end) { struct mlx5_ib_mr *mr; - const u64 umr_block_mask = (MLX5_UMR_MTT_ALIGNMENT / sizeof(u64)) - 1; + const u64 umr_block_mask = (MLX5_UMR_MTT_ALIGNMENT / + sizeof(struct mlx5_mtt)) - 1; u64 idx = 0, blk_start_idx = 0; int in_block = 0; u64 addr; @@ -72,8 +199,8 @@ void mlx5_ib_invalidate_range(struct ib_umem *umem, unsigned long start, * but they will write 0s as well, so no difference in the end result. */ - for (addr = start; addr < end; addr += (u64)umem->page_size) { - idx = (addr - ib_umem_start(umem)) / PAGE_SIZE; + for (addr = start; addr < end; addr += BIT(umem->page_shift)) { + idx = (addr - ib_umem_start(umem)) >> umem->page_shift; /* * Strive to write the MTTs in chunks, but avoid overwriting * non-existing MTTs. The huristic here can be improved to @@ -90,16 +217,19 @@ void mlx5_ib_invalidate_range(struct ib_umem *umem, unsigned long start, u64 umr_offset = idx & umr_block_mask; if (in_block && umr_offset == 0) { - mlx5_ib_update_mtt(mr, blk_start_idx, - idx - blk_start_idx, 1); + mlx5_ib_update_xlt(mr, blk_start_idx, + idx - blk_start_idx, 0, + MLX5_IB_UPD_XLT_ZAP | + MLX5_IB_UPD_XLT_ATOMIC); in_block = 0; } } } if (in_block) - mlx5_ib_update_mtt(mr, blk_start_idx, idx - blk_start_idx + 1, - 1); - + mlx5_ib_update_xlt(mr, blk_start_idx, + idx - blk_start_idx + 1, 0, + MLX5_IB_UPD_XLT_ZAP | + MLX5_IB_UPD_XLT_ATOMIC); /* * We are now sure that the device will not access the * memory. We can safely unmap it, and mark it as dirty if @@ -107,6 +237,13 @@ void mlx5_ib_invalidate_range(struct ib_umem *umem, unsigned long start, */ ib_umem_odp_unmap_dma_pages(umem, start, end); + + if (unlikely(!umem->npages && mr->parent && + !umem->odp_data->dying)) { + WRITE_ONCE(umem->odp_data->dying, 1); + atomic_inc(&mr->parent->num_leaf_free); + schedule_work(&umem->odp_data->work); + } } void mlx5_ib_internal_fill_odp_caps(struct mlx5_ib_dev *dev) @@ -120,6 +257,11 @@ void mlx5_ib_internal_fill_odp_caps(struct mlx5_ib_dev *dev) caps->general_caps = IB_ODP_SUPPORT; + if (MLX5_CAP_GEN(dev->mdev, umr_extended_translation_offset)) + dev->odp_max_size = U64_MAX; + else + dev->odp_max_size = BIT_ULL(MLX5_MAX_UMR_SHIFT + PAGE_SHIFT); + if (MLX5_CAP_ODP(dev->mdev, ud_odp_caps.send)) caps->per_transport_caps.ud_odp_caps |= IB_ODP_SUPPORT_SEND; @@ -135,159 +277,484 @@ void mlx5_ib_internal_fill_odp_caps(struct mlx5_ib_dev *dev) if (MLX5_CAP_ODP(dev->mdev, rc_odp_caps.read)) caps->per_transport_caps.rc_odp_caps |= IB_ODP_SUPPORT_READ; - return; -} + if (MLX5_CAP_ODP(dev->mdev, rc_odp_caps.atomic)) + caps->per_transport_caps.rc_odp_caps |= IB_ODP_SUPPORT_ATOMIC; -static struct mlx5_ib_mr *mlx5_ib_odp_find_mr_lkey(struct mlx5_ib_dev *dev, - u32 key) -{ - u32 base_key = mlx5_base_mkey(key); - struct mlx5_core_mkey *mmkey = __mlx5_mr_lookup(dev->mdev, base_key); - struct mlx5_ib_mr *mr = container_of(mmkey, struct mlx5_ib_mr, mmkey); + 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)) + caps->general_caps |= IB_ODP_SUPPORT_IMPLICIT; - if (!mmkey || mmkey->key != key || !mr->live) - return NULL; - - return container_of(mmkey, struct mlx5_ib_mr, mmkey); + return; } -static void mlx5_ib_page_fault_resume(struct mlx5_ib_qp *qp, - struct mlx5_ib_pfault *pfault, +static void mlx5_ib_page_fault_resume(struct mlx5_ib_dev *dev, + struct mlx5_pagefault *pfault, int error) { - struct mlx5_ib_dev *dev = to_mdev(qp->ibqp.pd->device); - u32 qpn = qp->trans_qp.base.mqp.qpn; + int wq_num = pfault->event_subtype == MLX5_PFAULT_SUBTYPE_WQE ? + pfault->wqe.wq_num : pfault->token; int ret = mlx5_core_page_fault_resume(dev->mdev, - qpn, - pfault->mpfault.flags, + pfault->token, + wq_num, + pfault->type, error); if (ret) - pr_err("Failed to resolve the page fault on QP 0x%x\n", qpn); + mlx5_ib_err(dev, "Failed to resolve the page fault on WQ 0x%x\n", + wq_num); } -/* - * Handle a single data segment in a page-fault WQE. - * - * Returns number of pages retrieved on success. The caller will continue to - * the next data segment. - * Can return the following error codes: - * -EAGAIN to designate a temporary error. The caller will abort handling the - * page fault and resolve it. - * -EFAULT when there's an error mapping the requested pages. The caller will - * abort the page fault handling and possibly move the QP to an error state. - * On other errors the QP should also be closed with an error. - */ -static int pagefault_single_data_segment(struct mlx5_ib_qp *qp, - struct mlx5_ib_pfault *pfault, - u32 key, u64 io_virt, size_t bcnt, - u32 *bytes_mapped) +static struct mlx5_ib_mr *implicit_mr_alloc(struct ib_pd *pd, + struct ib_umem *umem, + bool ksm, int access_flags) { - struct mlx5_ib_dev *mib_dev = to_mdev(qp->ibqp.pd->device); - int srcu_key; - unsigned int current_seq; - u64 start_idx; - int npages = 0, ret = 0; + struct mlx5_ib_dev *dev = to_mdev(pd->device); struct mlx5_ib_mr *mr; - u64 access_mask = ODP_READ_ALLOWED_BIT; + int err; + + mr = mlx5_mr_cache_alloc(dev, ksm ? MLX5_IMR_KSM_CACHE_ENTRY : + MLX5_IMR_MTT_CACHE_ENTRY); + + if (IS_ERR(mr)) + return mr; + + mr->ibmr.pd = pd; + + mr->dev = dev; + mr->access_flags = access_flags; + mr->mmkey.iova = 0; + mr->umem = umem; + + if (ksm) { + err = mlx5_ib_update_xlt(mr, 0, + mlx5_imr_ksm_entries, + MLX5_KSM_PAGE_SHIFT, + MLX5_IB_UPD_XLT_INDIRECT | + MLX5_IB_UPD_XLT_ZAP | + MLX5_IB_UPD_XLT_ENABLE); + + } else { + err = mlx5_ib_update_xlt(mr, 0, + MLX5_IMR_MTT_ENTRIES, + PAGE_SHIFT, + MLX5_IB_UPD_XLT_ZAP | + MLX5_IB_UPD_XLT_ENABLE | + MLX5_IB_UPD_XLT_ATOMIC); + } - srcu_key = srcu_read_lock(&mib_dev->mr_srcu); - mr = mlx5_ib_odp_find_mr_lkey(mib_dev, key); - /* - * If we didn't find the MR, it means the MR was closed while we were - * handling the ODP event. In this case we return -EFAULT so that the - * QP will be closed. - */ - if (!mr || !mr->ibmr.pd) { - pr_err("Failed to find relevant mr for lkey=0x%06x, probably the MR was destroyed\n", - key); - ret = -EFAULT; - goto srcu_unlock; + if (err) + goto fail; + + mr->ibmr.lkey = mr->mmkey.key; + mr->ibmr.rkey = mr->mmkey.key; + + mr->live = 1; + + mlx5_ib_dbg(dev, "key %x dev %p mr %p\n", + mr->mmkey.key, dev->mdev, mr); + + return mr; + +fail: + mlx5_ib_err(dev, "Failed to register MKEY %d\n", err); + mlx5_mr_cache_free(dev, mr); + + return ERR_PTR(err); +} + +static struct ib_umem_odp *implicit_mr_get_data(struct mlx5_ib_mr *mr, + u64 io_virt, size_t bcnt) +{ + struct ib_ucontext *ctx = mr->ibmr.pd->uobject->context; + struct mlx5_ib_dev *dev = to_mdev(mr->ibmr.pd->device); + struct ib_umem_odp *odp, *result = NULL; + u64 addr = io_virt & MLX5_IMR_MTT_MASK; + int nentries = 0, start_idx = 0, ret; + struct mlx5_ib_mr *mtt; + struct ib_umem *umem; + + mutex_lock(&mr->umem->odp_data->umem_mutex); + odp = odp_lookup(ctx, addr, 1, mr); + + mlx5_ib_dbg(dev, "io_virt:%llx bcnt:%zx addr:%llx odp:%p\n", + io_virt, bcnt, addr, odp); + +next_mr: + if (likely(odp)) { + if (nentries) + nentries++; + } else { + umem = ib_alloc_odp_umem(ctx, addr, MLX5_IMR_MTT_SIZE); + if (IS_ERR(umem)) { + mutex_unlock(&mr->umem->odp_data->umem_mutex); + return ERR_CAST(umem); + } + + mtt = implicit_mr_alloc(mr->ibmr.pd, umem, 0, mr->access_flags); + if (IS_ERR(mtt)) { + mutex_unlock(&mr->umem->odp_data->umem_mutex); + ib_umem_release(umem); + return ERR_CAST(mtt); + } + + odp = umem->odp_data; + odp->private = mtt; + mtt->umem = umem; + mtt->mmkey.iova = addr; + mtt->parent = mr; + INIT_WORK(&odp->work, mr_leaf_free_action); + + if (!nentries) + start_idx = addr >> MLX5_IMR_MTT_SHIFT; + nentries++; } - if (!mr->umem->odp_data) { - pr_debug("skipping non ODP MR (lkey=0x%06x) in page fault handler.\n", - key); - if (bytes_mapped) - *bytes_mapped += - (bcnt - pfault->mpfault.bytes_committed); - goto srcu_unlock; + + /* Return first odp if region not covered by single one */ + if (likely(!result)) + result = odp; + + addr += MLX5_IMR_MTT_SIZE; + if (unlikely(addr < io_virt + bcnt)) { + odp = odp_next(odp); + if (odp && odp->umem->address != addr) + odp = NULL; + goto next_mr; } - if (mr->ibmr.pd != qp->ibqp.pd) { - pr_err("Page-fault with different PDs for QP and MR.\n"); - ret = -EFAULT; - goto srcu_unlock; + + if (unlikely(nentries)) { + ret = mlx5_ib_update_xlt(mr, start_idx, nentries, 0, + MLX5_IB_UPD_XLT_INDIRECT | + MLX5_IB_UPD_XLT_ATOMIC); + if (ret) { + mlx5_ib_err(dev, "Failed to update PAS\n"); + result = ERR_PTR(ret); + } + } + + mutex_unlock(&mr->umem->odp_data->umem_mutex); + return result; +} + +struct mlx5_ib_mr *mlx5_ib_alloc_implicit_mr(struct mlx5_ib_pd *pd, + int access_flags) +{ + struct ib_ucontext *ctx = pd->ibpd.uobject->context; + struct mlx5_ib_mr *imr; + struct ib_umem *umem; + + umem = ib_umem_get(ctx, 0, 0, IB_ACCESS_ON_DEMAND, 0); + if (IS_ERR(umem)) + return ERR_CAST(umem); + + imr = implicit_mr_alloc(&pd->ibpd, umem, 1, access_flags); + if (IS_ERR(imr)) { + ib_umem_release(umem); + return ERR_CAST(imr); + } + + imr->umem = umem; + init_waitqueue_head(&imr->q_leaf_free); + atomic_set(&imr->num_leaf_free, 0); + + return imr; +} + +static int mr_leaf_free(struct ib_umem *umem, u64 start, + u64 end, void *cookie) +{ + struct mlx5_ib_mr *mr = umem->odp_data->private, *imr = cookie; + + if (mr->parent != imr) + return 0; + + ib_umem_odp_unmap_dma_pages(umem, + ib_umem_start(umem), + ib_umem_end(umem)); + + if (umem->odp_data->dying) + return 0; + + WRITE_ONCE(umem->odp_data->dying, 1); + atomic_inc(&imr->num_leaf_free); + schedule_work(&umem->odp_data->work); + + return 0; +} + +void mlx5_ib_free_implicit_mr(struct mlx5_ib_mr *imr) +{ + struct ib_ucontext *ctx = imr->ibmr.pd->uobject->context; + + down_read(&ctx->umem_rwsem); + rbt_ib_umem_for_each_in_range(&ctx->umem_tree, 0, ULLONG_MAX, + mr_leaf_free, imr); + up_read(&ctx->umem_rwsem); + + wait_event(imr->q_leaf_free, !atomic_read(&imr->num_leaf_free)); +} + +static int pagefault_mr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr, + u64 io_virt, size_t bcnt, u32 *bytes_mapped) +{ + u64 access_mask = ODP_READ_ALLOWED_BIT; + int npages = 0, page_shift, np; + u64 start_idx, page_mask; + struct ib_umem_odp *odp; + int current_seq; + size_t size; + int ret; + + if (!mr->umem->odp_data->page_list) { + odp = implicit_mr_get_data(mr, io_virt, bcnt); + + if (IS_ERR(odp)) + return PTR_ERR(odp); + mr = odp->private; + + } else { + odp = mr->umem->odp_data; } - current_seq = ACCESS_ONCE(mr->umem->odp_data->notifiers_seq); +next_mr: + size = min_t(size_t, bcnt, ib_umem_end(odp->umem) - io_virt); + + page_shift = mr->umem->page_shift; + page_mask = ~(BIT(page_shift) - 1); + start_idx = (io_virt - (mr->mmkey.iova & page_mask)) >> page_shift; + + if (mr->umem->writable) + access_mask |= ODP_WRITE_ALLOWED_BIT; + + current_seq = READ_ONCE(odp->notifiers_seq); /* * Ensure the sequence number is valid for some time before we call * gup. */ smp_rmb(); - /* - * Avoid branches - this code will perform correctly - * in all iterations (in iteration 2 and above, - * bytes_committed == 0). - */ - io_virt += pfault->mpfault.bytes_committed; - bcnt -= pfault->mpfault.bytes_committed; + ret = ib_umem_odp_map_dma_pages(mr->umem, io_virt, size, + access_mask, current_seq); - start_idx = (io_virt - (mr->mmkey.iova & PAGE_MASK)) >> PAGE_SHIFT; + if (ret < 0) + goto out; - if (mr->umem->writable) - access_mask |= ODP_WRITE_ALLOWED_BIT; - npages = ib_umem_odp_map_dma_pages(mr->umem, io_virt, bcnt, - access_mask, current_seq); - if (npages < 0) { - ret = npages; - goto srcu_unlock; + np = ret; + + mutex_lock(&odp->umem_mutex); + if (!ib_umem_mmu_notifier_retry(mr->umem, current_seq)) { + /* + * No need to check whether the MTTs really belong to + * this MR, since ib_umem_odp_map_dma_pages already + * checks this. + */ + ret = mlx5_ib_update_xlt(mr, start_idx, np, + page_shift, MLX5_IB_UPD_XLT_ATOMIC); + } else { + ret = -EAGAIN; } + mutex_unlock(&odp->umem_mutex); - if (npages > 0) { - mutex_lock(&mr->umem->odp_data->umem_mutex); - if (!ib_umem_mmu_notifier_retry(mr->umem, current_seq)) { - /* - * No need to check whether the MTTs really belong to - * this MR, since ib_umem_odp_map_dma_pages already - * checks this. - */ - ret = mlx5_ib_update_mtt(mr, start_idx, npages, 0); - } else { - ret = -EAGAIN; - } - mutex_unlock(&mr->umem->odp_data->umem_mutex); - if (ret < 0) { - if (ret != -EAGAIN) - pr_err("Failed to update mkey page tables\n"); - goto srcu_unlock; - } + if (ret < 0) { + if (ret != -EAGAIN) + mlx5_ib_err(dev, "Failed to update mkey page tables\n"); + goto out; + } + + if (bytes_mapped) { + u32 new_mappings = (np << page_shift) - + (io_virt - round_down(io_virt, 1 << page_shift)); + *bytes_mapped += min_t(u32, new_mappings, size); + } + + npages += np << (page_shift - PAGE_SHIFT); + bcnt -= size; - if (bytes_mapped) { - u32 new_mappings = npages * PAGE_SIZE - - (io_virt - round_down(io_virt, PAGE_SIZE)); - *bytes_mapped += min_t(u32, new_mappings, bcnt); + if (unlikely(bcnt)) { + struct ib_umem_odp *next; + + io_virt += size; + next = odp_next(odp); + if (unlikely(!next || next->umem->address != io_virt)) { + mlx5_ib_dbg(dev, "next implicit leaf removed at 0x%llx. got %p\n", + io_virt, next); + return -EAGAIN; } + odp = next; + mr = odp->private; + goto next_mr; } -srcu_unlock: + return npages; + +out: if (ret == -EAGAIN) { - if (!mr->umem->odp_data->dying) { - struct ib_umem_odp *odp_data = mr->umem->odp_data; + if (mr->parent || !odp->dying) { unsigned long timeout = msecs_to_jiffies(MMU_NOTIFIER_TIMEOUT); if (!wait_for_completion_timeout( - &odp_data->notifier_completion, + &odp->notifier_completion, timeout)) { - pr_warn("timeout waiting for mmu notifier completion\n"); + mlx5_ib_warn(dev, "timeout waiting for mmu notifier. seq %d against %d\n", + current_seq, odp->notifiers_seq); } } else { /* The MR is being killed, kill the QP as well. */ ret = -EFAULT; } } - srcu_read_unlock(&mib_dev->mr_srcu, srcu_key); - pfault->mpfault.bytes_committed = 0; + + return ret; +} + +struct pf_frame { + struct pf_frame *next; + u32 key; + u64 io_virt; + size_t bcnt; + int depth; +}; + +/* + * Handle a single data segment in a page-fault WQE or RDMA region. + * + * Returns number of OS pages retrieved on success. The caller may continue to + * the next data segment. + * Can return the following error codes: + * -EAGAIN to designate a temporary error. The caller will abort handling the + * page fault and resolve it. + * -EFAULT when there's an error mapping the requested pages. The caller will + * abort the page fault handling. + */ +static int pagefault_single_data_segment(struct mlx5_ib_dev *dev, + u32 key, u64 io_virt, size_t bcnt, + u32 *bytes_committed, + u32 *bytes_mapped) +{ + int npages = 0, srcu_key, ret, i, outlen, cur_outlen = 0, depth = 0; + struct pf_frame *head = NULL, *frame; + struct mlx5_core_mkey *mmkey; + struct mlx5_ib_mw *mw; + struct mlx5_ib_mr *mr; + struct mlx5_klm *pklm; + u32 *out = NULL; + size_t offset; + + srcu_key = srcu_read_lock(&dev->mr_srcu); + + io_virt += *bytes_committed; + bcnt -= *bytes_committed; + +next_mr: + mmkey = __mlx5_mr_lookup(dev->mdev, mlx5_base_mkey(key)); + if (!mmkey || mmkey->key != key) { + mlx5_ib_dbg(dev, "failed to find mkey %x\n", key); + ret = -EFAULT; + goto srcu_unlock; + } + + switch (mmkey->type) { + case MLX5_MKEY_MR: + mr = container_of(mmkey, struct mlx5_ib_mr, mmkey); + if (!mr->live || !mr->ibmr.pd) { + mlx5_ib_dbg(dev, "got dead MR\n"); + ret = -EFAULT; + goto srcu_unlock; + } + + ret = pagefault_mr(dev, mr, io_virt, bcnt, bytes_mapped); + if (ret < 0) + goto srcu_unlock; + + npages += ret; + ret = 0; + break; + + case MLX5_MKEY_MW: + mw = container_of(mmkey, struct mlx5_ib_mw, mmkey); + + if (depth >= MLX5_CAP_GEN(dev->mdev, max_indirection)) { + mlx5_ib_dbg(dev, "indirection level exceeded\n"); + ret = -EFAULT; + goto srcu_unlock; + } + + outlen = MLX5_ST_SZ_BYTES(query_mkey_out) + + sizeof(*pklm) * (mw->ndescs - 2); + + if (outlen > cur_outlen) { + kfree(out); + out = kzalloc(outlen, GFP_KERNEL); + if (!out) { + ret = -ENOMEM; + goto srcu_unlock; + } + cur_outlen = outlen; + } + + pklm = (struct mlx5_klm *)MLX5_ADDR_OF(query_mkey_out, out, + bsf0_klm0_pas_mtt0_1); + + ret = mlx5_core_query_mkey(dev->mdev, &mw->mmkey, out, outlen); + if (ret) + goto srcu_unlock; + + offset = io_virt - MLX5_GET64(query_mkey_out, out, + memory_key_mkey_entry.start_addr); + + for (i = 0; bcnt && i < mw->ndescs; i++, pklm++) { + if (offset >= be32_to_cpu(pklm->bcount)) { + offset -= be32_to_cpu(pklm->bcount); + continue; + } + + frame = kzalloc(sizeof(*frame), GFP_KERNEL); + if (!frame) { + ret = -ENOMEM; + goto srcu_unlock; + } + + frame->key = be32_to_cpu(pklm->key); + frame->io_virt = be64_to_cpu(pklm->va) + offset; + frame->bcnt = min_t(size_t, bcnt, + be32_to_cpu(pklm->bcount) - offset); + frame->depth = depth + 1; + frame->next = head; + head = frame; + + bcnt -= frame->bcnt; + } + break; + + default: + mlx5_ib_dbg(dev, "wrong mkey type %d\n", mmkey->type); + ret = -EFAULT; + goto srcu_unlock; + } + + if (head) { + frame = head; + head = frame->next; + + key = frame->key; + io_virt = frame->io_virt; + bcnt = frame->bcnt; + depth = frame->depth; + kfree(frame); + + goto next_mr; + } + +srcu_unlock: + while (head) { + frame = head; + head = frame->next; + kfree(frame); + } + kfree(out); + + srcu_read_unlock(&dev->mr_srcu, srcu_key); + *bytes_committed = 0; return ret ? ret : npages; } @@ -309,8 +776,9 @@ srcu_unlock: * Returns the number of pages loaded if positive, zero for an empty WQE, or a * negative error code. */ -static int pagefault_data_segments(struct mlx5_ib_qp *qp, - struct mlx5_ib_pfault *pfault, void *wqe, +static int pagefault_data_segments(struct mlx5_ib_dev *dev, + struct mlx5_pagefault *pfault, + struct mlx5_ib_qp *qp, void *wqe, void *wqe_end, u32 *bytes_mapped, u32 *total_wqe_bytes, int receive_queue) { @@ -354,22 +822,23 @@ static int pagefault_data_segments(struct mlx5_ib_qp *qp, if (!inline_segment && total_wqe_bytes) { *total_wqe_bytes += bcnt - min_t(size_t, bcnt, - pfault->mpfault.bytes_committed); + pfault->bytes_committed); } /* A zero length data segment designates a length of 2GB. */ if (bcnt == 0) bcnt = 1U << 31; - if (inline_segment || bcnt <= pfault->mpfault.bytes_committed) { - pfault->mpfault.bytes_committed -= + if (inline_segment || bcnt <= pfault->bytes_committed) { + pfault->bytes_committed -= min_t(size_t, bcnt, - pfault->mpfault.bytes_committed); + pfault->bytes_committed); continue; } - ret = pagefault_single_data_segment(qp, pfault, key, io_virt, - bcnt, bytes_mapped); + ret = pagefault_single_data_segment(dev, key, io_virt, bcnt, + &pfault->bytes_committed, + bytes_mapped); if (ret < 0) break; npages += ret; @@ -378,17 +847,29 @@ static int pagefault_data_segments(struct mlx5_ib_qp *qp, return ret < 0 ? ret : npages; } +static const u32 mlx5_ib_odp_opcode_cap[] = { + [MLX5_OPCODE_SEND] = IB_ODP_SUPPORT_SEND, + [MLX5_OPCODE_SEND_IMM] = IB_ODP_SUPPORT_SEND, + [MLX5_OPCODE_SEND_INVAL] = IB_ODP_SUPPORT_SEND, + [MLX5_OPCODE_RDMA_WRITE] = IB_ODP_SUPPORT_WRITE, + [MLX5_OPCODE_RDMA_WRITE_IMM] = IB_ODP_SUPPORT_WRITE, + [MLX5_OPCODE_RDMA_READ] = IB_ODP_SUPPORT_READ, + [MLX5_OPCODE_ATOMIC_CS] = IB_ODP_SUPPORT_ATOMIC, + [MLX5_OPCODE_ATOMIC_FA] = IB_ODP_SUPPORT_ATOMIC, +}; + /* * Parse initiator WQE. Advances the wqe pointer to point at the * scatter-gather list, and set wqe_end to the end of the WQE. */ static int mlx5_ib_mr_initiator_pfault_handler( - struct mlx5_ib_qp *qp, struct mlx5_ib_pfault *pfault, - void **wqe, void **wqe_end, int wqe_length) + struct mlx5_ib_dev *dev, struct mlx5_pagefault *pfault, + struct mlx5_ib_qp *qp, void **wqe, void **wqe_end, int wqe_length) { - struct mlx5_ib_dev *dev = to_mdev(qp->ibqp.pd->device); struct mlx5_wqe_ctrl_seg *ctrl = *wqe; - u16 wqe_index = pfault->mpfault.wqe.wqe_index; + u16 wqe_index = pfault->wqe.wqe_index; + u32 transport_caps; + struct mlx5_base_av *av; unsigned ds, opcode; #if defined(DEBUG) u32 ctrl_wqe_index, ctrl_qpn; @@ -434,53 +915,49 @@ static int mlx5_ib_mr_initiator_pfault_handler( opcode = be32_to_cpu(ctrl->opmod_idx_opcode) & MLX5_WQE_CTRL_OPCODE_MASK; + switch (qp->ibqp.qp_type) { case IB_QPT_RC: - switch (opcode) { - case MLX5_OPCODE_SEND: - case MLX5_OPCODE_SEND_IMM: - case MLX5_OPCODE_SEND_INVAL: - if (!(dev->odp_caps.per_transport_caps.rc_odp_caps & - IB_ODP_SUPPORT_SEND)) - goto invalid_transport_or_opcode; - break; - case MLX5_OPCODE_RDMA_WRITE: - case MLX5_OPCODE_RDMA_WRITE_IMM: - if (!(dev->odp_caps.per_transport_caps.rc_odp_caps & - IB_ODP_SUPPORT_WRITE)) - goto invalid_transport_or_opcode; - *wqe += sizeof(struct mlx5_wqe_raddr_seg); - break; - case MLX5_OPCODE_RDMA_READ: - if (!(dev->odp_caps.per_transport_caps.rc_odp_caps & - IB_ODP_SUPPORT_READ)) - goto invalid_transport_or_opcode; - *wqe += sizeof(struct mlx5_wqe_raddr_seg); - break; - default: - goto invalid_transport_or_opcode; - } + transport_caps = dev->odp_caps.per_transport_caps.rc_odp_caps; break; case IB_QPT_UD: - switch (opcode) { - case MLX5_OPCODE_SEND: - case MLX5_OPCODE_SEND_IMM: - if (!(dev->odp_caps.per_transport_caps.ud_odp_caps & - IB_ODP_SUPPORT_SEND)) - goto invalid_transport_or_opcode; - *wqe += sizeof(struct mlx5_wqe_datagram_seg); - break; - default: - goto invalid_transport_or_opcode; - } + transport_caps = dev->odp_caps.per_transport_caps.ud_odp_caps; break; default: -invalid_transport_or_opcode: - mlx5_ib_err(dev, "ODP fault on QP of an unsupported opcode or transport. transport: 0x%x opcode: 0x%x.\n", - qp->ibqp.qp_type, opcode); + mlx5_ib_err(dev, "ODP fault on QP of an unsupported transport 0x%x\n", + qp->ibqp.qp_type); + return -EFAULT; + } + + if (unlikely(opcode >= sizeof(mlx5_ib_odp_opcode_cap) / + sizeof(mlx5_ib_odp_opcode_cap[0]) || + !(transport_caps & mlx5_ib_odp_opcode_cap[opcode]))) { + mlx5_ib_err(dev, "ODP fault on QP of an unsupported opcode 0x%x\n", + opcode); return -EFAULT; } + if (qp->ibqp.qp_type != IB_QPT_RC) { + av = *wqe; + if (av->dqp_dct & be32_to_cpu(MLX5_WQE_AV_EXT)) + *wqe += sizeof(struct mlx5_av); + else + *wqe += sizeof(struct mlx5_base_av); + } + + switch (opcode) { + case MLX5_OPCODE_RDMA_WRITE: + case MLX5_OPCODE_RDMA_WRITE_IMM: + case MLX5_OPCODE_RDMA_READ: + *wqe += sizeof(struct mlx5_wqe_raddr_seg); + break; + case MLX5_OPCODE_ATOMIC_CS: + case MLX5_OPCODE_ATOMIC_FA: + *wqe += sizeof(struct mlx5_wqe_raddr_seg); + *wqe += sizeof(struct mlx5_wqe_atomic_seg); + break; + } + return 0; } @@ -489,10 +966,9 @@ invalid_transport_or_opcode: * scatter-gather list, and set wqe_end to the end of the WQE. */ static int mlx5_ib_mr_responder_pfault_handler( - struct mlx5_ib_qp *qp, struct mlx5_ib_pfault *pfault, - void **wqe, void **wqe_end, int wqe_length) + struct mlx5_ib_dev *dev, struct mlx5_pagefault *pfault, + struct mlx5_ib_qp *qp, void **wqe, void **wqe_end, int wqe_length) { - struct mlx5_ib_dev *dev = to_mdev(qp->ibqp.pd->device); struct mlx5_ib_wq *wq = &qp->rq; int wqe_size = 1 << wq->wqe_shift; @@ -529,70 +1005,80 @@ invalid_transport_or_opcode: return 0; } -static void mlx5_ib_mr_wqe_pfault_handler(struct mlx5_ib_qp *qp, - struct mlx5_ib_pfault *pfault) +static struct mlx5_ib_qp *mlx5_ib_odp_find_qp(struct mlx5_ib_dev *dev, + u32 wq_num) +{ + struct mlx5_core_qp *mqp = __mlx5_qp_lookup(dev->mdev, wq_num); + + if (!mqp) { + mlx5_ib_err(dev, "QPN 0x%6x not found\n", wq_num); + return NULL; + } + + return to_mibqp(mqp); +} + +static void mlx5_ib_mr_wqe_pfault_handler(struct mlx5_ib_dev *dev, + struct mlx5_pagefault *pfault) { - struct mlx5_ib_dev *dev = to_mdev(qp->ibqp.pd->device); int ret; void *wqe, *wqe_end; u32 bytes_mapped, total_wqe_bytes; char *buffer = NULL; - int resume_with_error = 0; - u16 wqe_index = pfault->mpfault.wqe.wqe_index; - int requestor = pfault->mpfault.flags & MLX5_PFAULT_REQUESTOR; - u32 qpn = qp->trans_qp.base.mqp.qpn; + int resume_with_error = 1; + u16 wqe_index = pfault->wqe.wqe_index; + int requestor = pfault->type & MLX5_PFAULT_REQUESTOR; + struct mlx5_ib_qp *qp; buffer = (char *)__get_free_page(GFP_KERNEL); if (!buffer) { mlx5_ib_err(dev, "Error allocating memory for IO page fault handling.\n"); - resume_with_error = 1; goto resolve_page_fault; } + qp = mlx5_ib_odp_find_qp(dev, pfault->wqe.wq_num); + if (!qp) + goto resolve_page_fault; + ret = mlx5_ib_read_user_wqe(qp, requestor, wqe_index, buffer, PAGE_SIZE, &qp->trans_qp.base); if (ret < 0) { - mlx5_ib_err(dev, "Failed reading a WQE following page fault, error=%x, wqe_index=%x, qpn=%x\n", - -ret, wqe_index, qpn); - resume_with_error = 1; + mlx5_ib_err(dev, "Failed reading a WQE following page fault, error=%d, wqe_index=%x, qpn=%x\n", + ret, wqe_index, pfault->token); goto resolve_page_fault; } wqe = buffer; if (requestor) - ret = mlx5_ib_mr_initiator_pfault_handler(qp, pfault, &wqe, + ret = mlx5_ib_mr_initiator_pfault_handler(dev, pfault, qp, &wqe, &wqe_end, ret); else - ret = mlx5_ib_mr_responder_pfault_handler(qp, pfault, &wqe, + ret = mlx5_ib_mr_responder_pfault_handler(dev, pfault, qp, &wqe, &wqe_end, ret); - if (ret < 0) { - resume_with_error = 1; + if (ret < 0) goto resolve_page_fault; - } if (wqe >= wqe_end) { mlx5_ib_err(dev, "ODP fault on invalid WQE.\n"); - resume_with_error = 1; goto resolve_page_fault; } - ret = pagefault_data_segments(qp, pfault, wqe, wqe_end, &bytes_mapped, - &total_wqe_bytes, !requestor); + ret = pagefault_data_segments(dev, pfault, qp, wqe, wqe_end, + &bytes_mapped, &total_wqe_bytes, + !requestor); if (ret == -EAGAIN) { + resume_with_error = 0; goto resolve_page_fault; } else if (ret < 0 || total_wqe_bytes > bytes_mapped) { - mlx5_ib_err(dev, "Error getting user pages for page fault. Error: 0x%x\n", - -ret); - resume_with_error = 1; goto resolve_page_fault; } + resume_with_error = 0; resolve_page_fault: - mlx5_ib_page_fault_resume(qp, pfault, resume_with_error); - mlx5_ib_dbg(dev, "PAGE FAULT completed. QP 0x%x resume_with_error=%d, flags: 0x%x\n", - qpn, resume_with_error, - pfault->mpfault.flags); - + mlx5_ib_page_fault_resume(dev, pfault, resume_with_error); + mlx5_ib_dbg(dev, "PAGE FAULT completed. QP 0x%x resume_with_error=%d, type: 0x%x\n", + pfault->wqe.wq_num, resume_with_error, + pfault->type); free_page((unsigned long)buffer); } @@ -602,15 +1088,14 @@ static int pages_in_range(u64 address, u32 length) (address & PAGE_MASK)) >> PAGE_SHIFT; } -static void mlx5_ib_mr_rdma_pfault_handler(struct mlx5_ib_qp *qp, - struct mlx5_ib_pfault *pfault) +static void mlx5_ib_mr_rdma_pfault_handler(struct mlx5_ib_dev *dev, + struct mlx5_pagefault *pfault) { - struct mlx5_pagefault *mpfault = &pfault->mpfault; u64 address; u32 length; - u32 prefetch_len = mpfault->bytes_committed; + u32 prefetch_len = pfault->bytes_committed; int prefetch_activated = 0; - u32 rkey = mpfault->rdma.r_key; + u32 rkey = pfault->rdma.r_key; int ret; /* The RDMA responder handler handles the page fault in two parts. @@ -619,38 +1104,40 @@ static void mlx5_ib_mr_rdma_pfault_handler(struct mlx5_ib_qp *qp, * prefetches more pages. The second operation cannot use the pfault * context and therefore uses the dummy_pfault context allocated on * the stack */ - struct mlx5_ib_pfault dummy_pfault = {}; + pfault->rdma.rdma_va += pfault->bytes_committed; + pfault->rdma.rdma_op_len -= min(pfault->bytes_committed, + pfault->rdma.rdma_op_len); + pfault->bytes_committed = 0; - dummy_pfault.mpfault.bytes_committed = 0; - - mpfault->rdma.rdma_va += mpfault->bytes_committed; - mpfault->rdma.rdma_op_len -= min(mpfault->bytes_committed, - mpfault->rdma.rdma_op_len); - mpfault->bytes_committed = 0; - - address = mpfault->rdma.rdma_va; - length = mpfault->rdma.rdma_op_len; + address = pfault->rdma.rdma_va; + length = pfault->rdma.rdma_op_len; /* For some operations, the hardware cannot tell the exact message * length, and in those cases it reports zero. Use prefetch * logic. */ if (length == 0) { prefetch_activated = 1; - length = mpfault->rdma.packet_size; + length = pfault->rdma.packet_size; prefetch_len = min(MAX_PREFETCH_LEN, prefetch_len); } - ret = pagefault_single_data_segment(qp, pfault, rkey, address, length, - NULL); + ret = pagefault_single_data_segment(dev, rkey, address, length, + &pfault->bytes_committed, NULL); if (ret == -EAGAIN) { /* We're racing with an invalidation, don't prefetch */ prefetch_activated = 0; } else if (ret < 0 || pages_in_range(address, length) > ret) { - mlx5_ib_page_fault_resume(qp, pfault, 1); + mlx5_ib_page_fault_resume(dev, pfault, 1); + if (ret != -ENOENT) + mlx5_ib_dbg(dev, "PAGE FAULT error %d. QP 0x%x, type: 0x%x\n", + ret, pfault->token, pfault->type); return; } - mlx5_ib_page_fault_resume(qp, pfault, 0); + mlx5_ib_page_fault_resume(dev, pfault, 0); + mlx5_ib_dbg(dev, "PAGE FAULT completed. QP 0x%x, type: 0x%x, prefetch_activated: %d\n", + pfault->token, pfault->type, + prefetch_activated); /* At this point, there might be a new pagefault already arriving in * the eq, switch to the dummy pagefault for the rest of the @@ -658,139 +1145,93 @@ static void mlx5_ib_mr_rdma_pfault_handler(struct mlx5_ib_qp *qp, * work-queue is being fenced. */ if (prefetch_activated) { - ret = pagefault_single_data_segment(qp, &dummy_pfault, rkey, - address, + u32 bytes_committed = 0; + + ret = pagefault_single_data_segment(dev, rkey, address, prefetch_len, - NULL); - if (ret < 0) { - pr_warn("Prefetch failed (ret = %d, prefetch_activated = %d) for QPN %d, address: 0x%.16llx, length = 0x%.16x\n", - ret, prefetch_activated, - qp->ibqp.qp_num, address, prefetch_len); + &bytes_committed, NULL); + if (ret < 0 && ret != -EAGAIN) { + mlx5_ib_dbg(dev, "Prefetch failed. ret: %d, QP 0x%x, address: 0x%.16llx, length = 0x%.16x\n", + ret, pfault->token, address, prefetch_len); } } } -void mlx5_ib_mr_pfault_handler(struct mlx5_ib_qp *qp, - struct mlx5_ib_pfault *pfault) +void mlx5_ib_pfault(struct mlx5_core_dev *mdev, void *context, + struct mlx5_pagefault *pfault) { - u8 event_subtype = pfault->mpfault.event_subtype; + struct mlx5_ib_dev *dev = context; + u8 event_subtype = pfault->event_subtype; switch (event_subtype) { case MLX5_PFAULT_SUBTYPE_WQE: - mlx5_ib_mr_wqe_pfault_handler(qp, pfault); + mlx5_ib_mr_wqe_pfault_handler(dev, pfault); break; case MLX5_PFAULT_SUBTYPE_RDMA: - mlx5_ib_mr_rdma_pfault_handler(qp, pfault); + mlx5_ib_mr_rdma_pfault_handler(dev, pfault); break; default: - pr_warn("Invalid page fault event subtype: 0x%x\n", - event_subtype); - mlx5_ib_page_fault_resume(qp, pfault, 1); - break; + mlx5_ib_err(dev, "Invalid page fault event subtype: 0x%x\n", + event_subtype); + mlx5_ib_page_fault_resume(dev, pfault, 1); } } -static void mlx5_ib_qp_pfault_action(struct work_struct *work) -{ - struct mlx5_ib_pfault *pfault = container_of(work, - struct mlx5_ib_pfault, - work); - enum mlx5_ib_pagefault_context context = - mlx5_ib_get_pagefault_context(&pfault->mpfault); - struct mlx5_ib_qp *qp = container_of(pfault, struct mlx5_ib_qp, - pagefaults[context]); - mlx5_ib_mr_pfault_handler(qp, pfault); -} - -void mlx5_ib_qp_disable_pagefaults(struct mlx5_ib_qp *qp) -{ - unsigned long flags; - - spin_lock_irqsave(&qp->disable_page_faults_lock, flags); - qp->disable_page_faults = 1; - spin_unlock_irqrestore(&qp->disable_page_faults_lock, flags); - - /* - * Note that at this point, we are guarenteed that no more - * work queue elements will be posted to the work queue with - * the QP we are closing. - */ - flush_workqueue(mlx5_ib_page_fault_wq); -} - -void mlx5_ib_qp_enable_pagefaults(struct mlx5_ib_qp *qp) -{ - unsigned long flags; - - spin_lock_irqsave(&qp->disable_page_faults_lock, flags); - qp->disable_page_faults = 0; - spin_unlock_irqrestore(&qp->disable_page_faults_lock, flags); -} - -static void mlx5_ib_pfault_handler(struct mlx5_core_qp *qp, - struct mlx5_pagefault *pfault) +void mlx5_odp_init_mr_cache_entry(struct mlx5_cache_ent *ent) { - /* - * Note that we will only get one fault event per QP per context - * (responder/initiator, read/write), until we resolve the page fault - * with the mlx5_ib_page_fault_resume command. Since this function is - * called from within the work element, there is no risk of missing - * events. - */ - struct mlx5_ib_qp *mibqp = to_mibqp(qp); - enum mlx5_ib_pagefault_context context = - mlx5_ib_get_pagefault_context(pfault); - struct mlx5_ib_pfault *qp_pfault = &mibqp->pagefaults[context]; - - qp_pfault->mpfault = *pfault; - - /* No need to stop interrupts here since we are in an interrupt */ - spin_lock(&mibqp->disable_page_faults_lock); - if (!mibqp->disable_page_faults) - queue_work(mlx5_ib_page_fault_wq, &qp_pfault->work); - spin_unlock(&mibqp->disable_page_faults_lock); -} - -void mlx5_ib_odp_create_qp(struct mlx5_ib_qp *qp) -{ - int i; - - qp->disable_page_faults = 1; - spin_lock_init(&qp->disable_page_faults_lock); + if (!(ent->dev->odp_caps.general_caps & IB_ODP_SUPPORT_IMPLICIT)) + return; - qp->trans_qp.base.mqp.pfault_handler = mlx5_ib_pfault_handler; + switch (ent->order - 2) { + case MLX5_IMR_MTT_CACHE_ENTRY: + ent->page = PAGE_SHIFT; + ent->xlt = MLX5_IMR_MTT_ENTRIES * + sizeof(struct mlx5_mtt) / + MLX5_IB_UMR_OCTOWORD; + ent->access_mode = MLX5_MKC_ACCESS_MODE_MTT; + ent->limit = 0; + break; - for (i = 0; i < MLX5_IB_PAGEFAULT_CONTEXTS; ++i) - INIT_WORK(&qp->pagefaults[i].work, mlx5_ib_qp_pfault_action); + case MLX5_IMR_KSM_CACHE_ENTRY: + ent->page = MLX5_KSM_PAGE_SHIFT; + ent->xlt = mlx5_imr_ksm_entries * + sizeof(struct mlx5_klm) / + MLX5_IB_UMR_OCTOWORD; + ent->access_mode = MLX5_MKC_ACCESS_MODE_KSM; + ent->limit = 0; + break; + } } -int mlx5_ib_odp_init_one(struct mlx5_ib_dev *ibdev) +int mlx5_ib_odp_init_one(struct mlx5_ib_dev *dev) { int ret; - ret = init_srcu_struct(&ibdev->mr_srcu); + 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) { + mlx5_ib_err(dev, "Error getting null_mkey %d\n", ret); + return ret; + } + } + return 0; } -void mlx5_ib_odp_remove_one(struct mlx5_ib_dev *ibdev) +void mlx5_ib_odp_remove_one(struct mlx5_ib_dev *dev) { - cleanup_srcu_struct(&ibdev->mr_srcu); + cleanup_srcu_struct(&dev->mr_srcu); } -int __init mlx5_ib_odp_init(void) +int mlx5_ib_odp_init(void) { - mlx5_ib_page_fault_wq = alloc_ordered_workqueue("mlx5_ib_page_faults", - WQ_MEM_RECLAIM); - if (!mlx5_ib_page_fault_wq) - return -ENOMEM; + mlx5_imr_ksm_entries = BIT_ULL(get_order(TASK_SIZE) - + MLX5_IMR_MTT_BITS); return 0; } -void mlx5_ib_odp_cleanup(void) -{ - destroy_workqueue(mlx5_ib_page_fault_wq); -} diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c index a1b3125f0a6e..0889ff367c86 100644 --- a/drivers/infiniband/hw/mlx5/qp.c +++ b/drivers/infiniband/hw/mlx5/qp.c @@ -475,60 +475,53 @@ static int qp_has_rq(struct ib_qp_init_attr *attr) return 1; } -static int first_med_uuar(void) +static int first_med_bfreg(void) { return 1; } -static int next_uuar(int n) -{ - n++; - - while (((n % 4) & 2)) - n++; +enum { + /* this is the first blue flame register in the array of bfregs assigned + * to a processes. Since we do not use it for blue flame but rather + * regular 64 bit doorbells, we do not need a lock for maintaiing + * "odd/even" order + */ + NUM_NON_BLUE_FLAME_BFREGS = 1, +}; - return n; +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; } -static int num_med_uuar(struct mlx5_uuar_info *uuari) +static int num_med_bfreg(struct mlx5_ib_dev *dev, + struct mlx5_bfreg_info *bfregi) { int n; - n = uuari->num_uars * MLX5_NON_FP_BF_REGS_PER_PAGE - - uuari->num_low_latency_uuars - 1; + n = max_bfregs(dev, bfregi) - bfregi->num_low_latency_bfregs - + NUM_NON_BLUE_FLAME_BFREGS; return n >= 0 ? n : 0; } -static int max_uuari(struct mlx5_uuar_info *uuari) -{ - return uuari->num_uars * 4; -} - -static int first_hi_uuar(struct mlx5_uuar_info *uuari) +static int first_hi_bfreg(struct mlx5_ib_dev *dev, + struct mlx5_bfreg_info *bfregi) { int med; - int i; - int t; - - med = num_med_uuar(uuari); - for (t = 0, i = first_med_uuar();; i = next_uuar(i)) { - t++; - if (t == med) - return next_uuar(i); - } - return 0; + med = num_med_bfreg(dev, bfregi); + return ++med; } -static int alloc_high_class_uuar(struct mlx5_uuar_info *uuari) +static int alloc_high_class_bfreg(struct mlx5_ib_dev *dev, + struct mlx5_bfreg_info *bfregi) { int i; - for (i = first_hi_uuar(uuari); i < max_uuari(uuari); i = next_uuar(i)) { - if (!test_bit(i, uuari->bitmap)) { - set_bit(i, uuari->bitmap); - uuari->count[i]++; + for (i = first_hi_bfreg(dev, bfregi); i < max_bfregs(dev, bfregi); i++) { + if (!bfregi->count[i]) { + bfregi->count[i]++; return i; } } @@ -536,87 +529,61 @@ static int alloc_high_class_uuar(struct mlx5_uuar_info *uuari) return -ENOMEM; } -static int alloc_med_class_uuar(struct mlx5_uuar_info *uuari) +static int alloc_med_class_bfreg(struct mlx5_ib_dev *dev, + struct mlx5_bfreg_info *bfregi) { - int minidx = first_med_uuar(); + int minidx = first_med_bfreg(); int i; - for (i = first_med_uuar(); i < first_hi_uuar(uuari); i = next_uuar(i)) { - if (uuari->count[i] < uuari->count[minidx]) + for (i = first_med_bfreg(); i < first_hi_bfreg(dev, bfregi); i++) { + if (bfregi->count[i] < bfregi->count[minidx]) minidx = i; + if (!bfregi->count[minidx]) + break; } - uuari->count[minidx]++; + bfregi->count[minidx]++; return minidx; } -static int alloc_uuar(struct mlx5_uuar_info *uuari, - enum mlx5_ib_latency_class lat) +static int alloc_bfreg(struct mlx5_ib_dev *dev, + struct mlx5_bfreg_info *bfregi, + enum mlx5_ib_latency_class lat) { - int uuarn = -EINVAL; + int bfregn = -EINVAL; - mutex_lock(&uuari->lock); + mutex_lock(&bfregi->lock); switch (lat) { case MLX5_IB_LATENCY_CLASS_LOW: - uuarn = 0; - uuari->count[uuarn]++; + BUILD_BUG_ON(NUM_NON_BLUE_FLAME_BFREGS != 1); + bfregn = 0; + bfregi->count[bfregn]++; break; case MLX5_IB_LATENCY_CLASS_MEDIUM: - if (uuari->ver < 2) - uuarn = -ENOMEM; + if (bfregi->ver < 2) + bfregn = -ENOMEM; else - uuarn = alloc_med_class_uuar(uuari); + bfregn = alloc_med_class_bfreg(dev, bfregi); break; case MLX5_IB_LATENCY_CLASS_HIGH: - if (uuari->ver < 2) - uuarn = -ENOMEM; + if (bfregi->ver < 2) + bfregn = -ENOMEM; else - uuarn = alloc_high_class_uuar(uuari); - break; - - case MLX5_IB_LATENCY_CLASS_FAST_PATH: - uuarn = 2; + bfregn = alloc_high_class_bfreg(dev, bfregi); break; } - mutex_unlock(&uuari->lock); - - return uuarn; -} + mutex_unlock(&bfregi->lock); -static void free_med_class_uuar(struct mlx5_uuar_info *uuari, int uuarn) -{ - clear_bit(uuarn, uuari->bitmap); - --uuari->count[uuarn]; + return bfregn; } -static void free_high_class_uuar(struct mlx5_uuar_info *uuari, int uuarn) +static void free_bfreg(struct mlx5_ib_dev *dev, struct mlx5_bfreg_info *bfregi, int bfregn) { - clear_bit(uuarn, uuari->bitmap); - --uuari->count[uuarn]; -} - -static void free_uuar(struct mlx5_uuar_info *uuari, int uuarn) -{ - int nuuars = uuari->num_uars * MLX5_BF_REGS_PER_PAGE; - int high_uuar = nuuars - uuari->num_low_latency_uuars; - - mutex_lock(&uuari->lock); - if (uuarn == 0) { - --uuari->count[uuarn]; - goto out; - } - - if (uuarn < high_uuar) { - free_med_class_uuar(uuari, uuarn); - goto out; - } - - free_high_class_uuar(uuari, uuarn); - -out: - mutex_unlock(&uuari->lock); + mutex_lock(&bfregi->lock); + bfregi->count[bfregn]--; + mutex_unlock(&bfregi->lock); } static enum mlx5_qp_state to_mlx5_state(enum ib_qp_state state) @@ -657,9 +624,20 @@ static void mlx5_ib_lock_cqs(struct mlx5_ib_cq *send_cq, static void mlx5_ib_unlock_cqs(struct mlx5_ib_cq *send_cq, struct mlx5_ib_cq *recv_cq); -static int uuarn_to_uar_index(struct mlx5_uuar_info *uuari, int uuarn) +static int bfregn_to_uar_index(struct mlx5_ib_dev *dev, + struct mlx5_bfreg_info *bfregi, int bfregn) { - return uuari->uars[uuarn / MLX5_BF_REGS_PER_PAGE].index; + int bfregs_per_sys_page; + int index_of_sys_page; + int offset; + + bfregs_per_sys_page = get_uars_per_sys_page(dev, bfregi->lib_uar_4k) * + 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; + + return bfregi->sys_pages[index_of_sys_page] + offset; } static int mlx5_ib_umem_get(struct mlx5_ib_dev *dev, @@ -762,6 +740,13 @@ err_umem: return err; } +static int adjust_bfregn(struct mlx5_ib_dev *dev, + struct mlx5_bfreg_info *bfregi, int bfregn) +{ + return bfregn / MLX5_NON_FP_BFREGS_PER_UAR * MLX5_BFREGS_PER_UAR + + bfregn % MLX5_NON_FP_BFREGS_PER_UAR; +} + static int create_user_qp(struct mlx5_ib_dev *dev, struct ib_pd *pd, struct mlx5_ib_qp *qp, struct ib_udata *udata, struct ib_qp_init_attr *attr, @@ -776,7 +761,7 @@ static int create_user_qp(struct mlx5_ib_dev *dev, struct ib_pd *pd, int uar_index; int npages; u32 offset = 0; - int uuarn; + int bfregn; int ncont = 0; __be64 *pas; void *qpc; @@ -794,27 +779,27 @@ static int create_user_qp(struct mlx5_ib_dev *dev, struct ib_pd *pd, */ if (qp->flags & MLX5_IB_QP_CROSS_CHANNEL) /* In CROSS_CHANNEL CQ and QP must use the same UAR */ - uuarn = MLX5_CROSS_CHANNEL_UUAR; + bfregn = MLX5_CROSS_CHANNEL_BFREG; else { - uuarn = alloc_uuar(&context->uuari, MLX5_IB_LATENCY_CLASS_HIGH); - if (uuarn < 0) { - mlx5_ib_dbg(dev, "failed to allocate low latency UUAR\n"); + bfregn = alloc_bfreg(dev, &context->bfregi, MLX5_IB_LATENCY_CLASS_HIGH); + if (bfregn < 0) { + mlx5_ib_dbg(dev, "failed to allocate low latency BFREG\n"); mlx5_ib_dbg(dev, "reverting to medium latency\n"); - uuarn = alloc_uuar(&context->uuari, MLX5_IB_LATENCY_CLASS_MEDIUM); - if (uuarn < 0) { - mlx5_ib_dbg(dev, "failed to allocate medium latency UUAR\n"); + bfregn = alloc_bfreg(dev, &context->bfregi, MLX5_IB_LATENCY_CLASS_MEDIUM); + if (bfregn < 0) { + mlx5_ib_dbg(dev, "failed to allocate medium latency BFREG\n"); mlx5_ib_dbg(dev, "reverting to high latency\n"); - uuarn = alloc_uuar(&context->uuari, MLX5_IB_LATENCY_CLASS_LOW); - if (uuarn < 0) { - mlx5_ib_warn(dev, "uuar allocation failed\n"); - return uuarn; + bfregn = alloc_bfreg(dev, &context->bfregi, MLX5_IB_LATENCY_CLASS_LOW); + if (bfregn < 0) { + mlx5_ib_warn(dev, "bfreg allocation failed\n"); + return bfregn; } } } } - uar_index = uuarn_to_uar_index(&context->uuari, uuarn); - mlx5_ib_dbg(dev, "uuarn 0x%x, uar_index 0x%x\n", uuarn, uar_index); + 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); qp->rq.offset = 0; qp->sq.wqe_shift = ilog2(MLX5_SEND_WQE_BB); @@ -822,7 +807,7 @@ static int create_user_qp(struct mlx5_ib_dev *dev, struct ib_pd *pd, err = set_user_buf_size(dev, qp, &ucmd, base, attr); if (err) - goto err_uuar; + goto err_bfreg; if (ucmd.buf_addr && ubuffer->buf_size) { ubuffer->buf_addr = ucmd.buf_addr; @@ -831,14 +816,14 @@ static int create_user_qp(struct mlx5_ib_dev *dev, struct ib_pd *pd, &ubuffer->umem, &npages, &page_shift, &ncont, &offset); if (err) - goto err_uuar; + goto err_bfreg; } else { ubuffer->umem = NULL; } *inlen = MLX5_ST_SZ_BYTES(create_qp_in) + MLX5_FLD_SZ_BYTES(create_qp_in, pas[0]) * ncont; - *in = mlx5_vzalloc(*inlen); + *in = kvzalloc(*inlen, GFP_KERNEL); if (!*in) { err = -ENOMEM; goto err_umem; @@ -854,8 +839,8 @@ 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->uuar_index = uuarn; - qp->uuarn = uuarn; + resp->bfreg_index = adjust_bfregn(dev, &context->bfregi, bfregn); + qp->bfregn = bfregn; err = mlx5_ib_db_map_user(context, ucmd.db_addr, &qp->db); if (err) { @@ -882,13 +867,13 @@ err_umem: if (ubuffer->umem) ib_umem_release(ubuffer->umem); -err_uuar: - free_uuar(&context->uuari, uuarn); +err_bfreg: + free_bfreg(dev, &context->bfregi, bfregn); return err; } -static void destroy_qp_user(struct ib_pd *pd, struct mlx5_ib_qp *qp, - struct mlx5_ib_qp_base *base) +static void destroy_qp_user(struct mlx5_ib_dev *dev, struct ib_pd *pd, + struct mlx5_ib_qp *qp, struct mlx5_ib_qp_base *base) { struct mlx5_ib_ucontext *context; @@ -896,7 +881,7 @@ static void destroy_qp_user(struct ib_pd *pd, struct mlx5_ib_qp *qp, mlx5_ib_db_unmap_user(context, &qp->db); if (base->ubuffer.umem) ib_umem_release(base->ubuffer.umem); - free_uuar(&context->uuari, qp->uuarn); + free_bfreg(dev, &context->bfregi, qp->bfregn); } static int create_kernel_qp(struct mlx5_ib_dev *dev, @@ -905,36 +890,32 @@ static int create_kernel_qp(struct mlx5_ib_dev *dev, u32 **in, int *inlen, struct mlx5_ib_qp_base *base) { - enum mlx5_ib_latency_class lc = MLX5_IB_LATENCY_CLASS_LOW; - struct mlx5_uuar_info *uuari; int uar_index; void *qpc; - int uuarn; int err; - uuari = &dev->mdev->priv.uuari; if (init_attr->create_flags & ~(IB_QP_CREATE_SIGNATURE_EN | IB_QP_CREATE_BLOCK_MULTICAST_LOOPBACK | IB_QP_CREATE_IPOIB_UD_LSO | + IB_QP_CREATE_NETIF_QP | mlx5_ib_create_qp_sqpn_qp1())) return -EINVAL; if (init_attr->qp_type == MLX5_IB_QPT_REG_UMR) - lc = MLX5_IB_LATENCY_CLASS_FAST_PATH; - - uuarn = alloc_uuar(uuari, lc); - if (uuarn < 0) { - mlx5_ib_dbg(dev, "\n"); - return -ENOMEM; - } + qp->bf.bfreg = &dev->fp_bfreg; + else + qp->bf.bfreg = &dev->bfreg; - qp->bf = &uuari->bfs[uuarn]; - uar_index = qp->bf->uar->index; + /* We need to divide by two since each register is comprised of + * two buffers of identical size, namely odd and even + */ + qp->bf.buf_size = (1 << MLX5_CAP_GEN(dev->mdev, log_bf_reg_size)) / 2; + uar_index = qp->bf.bfreg->index; err = calc_sq_size(dev, init_attr, qp); if (err < 0) { mlx5_ib_dbg(dev, "err %d\n", err); - goto err_uuar; + return err; } qp->rq.offset = 0; @@ -944,13 +925,13 @@ static int create_kernel_qp(struct mlx5_ib_dev *dev, err = mlx5_buf_alloc(dev->mdev, base->ubuffer.buf_size, &qp->buf); if (err) { mlx5_ib_dbg(dev, "err %d\n", err); - goto err_uuar; + return err; } qp->sq.qend = mlx5_get_send_wqe(qp, qp->sq.wqe_cnt); *inlen = MLX5_ST_SZ_BYTES(create_qp_in) + MLX5_FLD_SZ_BYTES(create_qp_in, pas[0]) * qp->buf.npages; - *in = mlx5_vzalloc(*inlen); + *in = kvzalloc(*inlen, GFP_KERNEL); if (!*in) { err = -ENOMEM; goto err_buf; @@ -994,34 +975,30 @@ static int create_kernel_qp(struct mlx5_ib_dev *dev, return 0; err_wrid: - mlx5_db_free(dev->mdev, &qp->db); kfree(qp->sq.wqe_head); kfree(qp->sq.w_list); kfree(qp->sq.wrid); kfree(qp->sq.wr_data); kfree(qp->rq.wrid); + mlx5_db_free(dev->mdev, &qp->db); err_free: kvfree(*in); err_buf: mlx5_buf_free(dev->mdev, &qp->buf); - -err_uuar: - free_uuar(&dev->mdev->priv.uuari, uuarn); return err; } static void destroy_qp_kernel(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp) { - mlx5_db_free(dev->mdev, &qp->db); kfree(qp->sq.wqe_head); kfree(qp->sq.w_list); kfree(qp->sq.wrid); kfree(qp->sq.wr_data); kfree(qp->rq.wrid); + mlx5_db_free(dev->mdev, &qp->db); mlx5_buf_free(dev->mdev, &qp->buf); - free_uuar(&dev->mdev->priv.uuari, qp->bf->uuarn); } static u32 get_rx_type(struct mlx5_ib_qp *qp, struct ib_qp_init_attr *attr) @@ -1083,7 +1060,7 @@ static int create_raw_packet_qp_sq(struct mlx5_ib_dev *dev, return err; inlen = MLX5_ST_SZ_BYTES(create_sq_in) + sizeof(u64) * ncont; - in = mlx5_vzalloc(inlen); + in = kvzalloc(inlen, GFP_KERNEL); if (!in) { err = -ENOMEM; goto err_umem; @@ -1163,12 +1140,13 @@ static int create_raw_packet_qp_rq(struct mlx5_ib_dev *dev, u32 rq_pas_size = get_rq_pas_size(qpc); inlen = MLX5_ST_SZ_BYTES(create_rq_in) + rq_pas_size; - in = mlx5_vzalloc(inlen); + in = kvzalloc(inlen, GFP_KERNEL); if (!in) return -ENOMEM; rqc = MLX5_ADDR_OF(create_rq_in, in, ctx); - MLX5_SET(rqc, rqc, vsd, 1); + if (!(rq->flags & MLX5_IB_RQ_CVLAN_STRIPPING)) + MLX5_SET(rqc, rqc, vsd, 1); MLX5_SET(rqc, rqc, mem_rq_type, MLX5_RQC_MEM_RQ_TYPE_MEMORY_RQ_INLINE); MLX5_SET(rqc, rqc, state, MLX5_RQC_STATE_RST); MLX5_SET(rqc, rqc, flush_in_error_en, 1); @@ -1215,7 +1193,7 @@ static int create_raw_packet_qp_tir(struct mlx5_ib_dev *dev, int err; inlen = MLX5_ST_SZ_BYTES(create_tir_in); - in = mlx5_vzalloc(inlen); + in = kvzalloc(inlen, GFP_KERNEL); if (!in) return -ENOMEM; @@ -1265,6 +1243,8 @@ static int create_raw_packet_qp(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp, if (qp->rq.wqe_cnt) { rq->base.container_mibqp = qp; + if (qp->flags & MLX5_IB_QP_CVLAN_STRIPPING) + rq->flags |= MLX5_IB_RQ_CVLAN_STRIPPING; err = create_raw_packet_qp_rq(dev, rq, in); if (err) goto err_destroy_sq; @@ -1353,7 +1333,7 @@ static int create_rss_raw_qp_tir(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp, if (init_attr->create_flags || init_attr->send_cq) return -EINVAL; - min_resp_len = offsetof(typeof(resp), uuar_index) + sizeof(resp.uuar_index); + min_resp_len = offsetof(typeof(resp), bfreg_index) + sizeof(resp.bfreg_index); if (udata->outlen < min_resp_len) return -EINVAL; @@ -1392,7 +1372,7 @@ static int create_rss_raw_qp_tir(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp, } inlen = MLX5_ST_SZ_BYTES(create_tir_in); - in = mlx5_vzalloc(inlen); + in = kvzalloc(inlen, GFP_KERNEL); if (!in) return -ENOMEM; @@ -1526,9 +1506,6 @@ static int create_qp_common(struct mlx5_ib_dev *dev, struct ib_pd *pd, &qp->raw_packet_qp.rq.base : &qp->trans_qp.base; - if (init_attr->qp_type != IB_QPT_RAW_PACKET) - mlx5_ib_odp_create_qp(qp); - mutex_init(&qp->mutex); spin_lock_init(&qp->sq.lock); spin_lock_init(&qp->rq.lock); @@ -1589,6 +1566,14 @@ static int create_qp_common(struct mlx5_ib_dev *dev, struct ib_pd *pd, if (init_attr->sq_sig_type == IB_SIGNAL_ALL_WR) qp->sq_signal_bits = MLX5_WQE_CTRL_CQ_UPDATE; + if (init_attr->create_flags & IB_QP_CREATE_CVLAN_STRIPPING) { + if (!(MLX5_CAP_GEN(dev->mdev, eth_net_offloads) && + MLX5_CAP_ETH(dev->mdev, vlan_cap)) || + (init_attr->qp_type != IB_QPT_RAW_PACKET)) + return -EOPNOTSUPP; + qp->flags |= MLX5_IB_QP_CVLAN_STRIPPING; + } + if (pd && pd->uobject) { if (ib_copy_from_udata(&ucmd, udata, sizeof(ucmd))) { mlx5_ib_dbg(dev, "copy failed\n"); @@ -1648,7 +1633,7 @@ static int create_qp_common(struct mlx5_ib_dev *dev, struct ib_pd *pd, if (err) return err; } else { - in = mlx5_vzalloc(inlen); + in = kvzalloc(inlen, GFP_KERNEL); if (!in) return -ENOMEM; @@ -1795,7 +1780,7 @@ static int create_qp_common(struct mlx5_ib_dev *dev, struct ib_pd *pd, err_create: if (qp->create_type == MLX5_QP_USER) - destroy_qp_user(pd, qp, base); + destroy_qp_user(dev, pd, qp, base); else if (qp->create_type == MLX5_QP_KERNEL) destroy_qp_kernel(dev, qp); @@ -1923,7 +1908,6 @@ static void destroy_qp_common(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp) if (qp->state != IB_QPS_RESET) { if (qp->ibqp.qp_type != IB_QPT_RAW_PACKET) { - mlx5_ib_qp_disable_pagefaults(qp); err = mlx5_core_qp_modify(dev->mdev, MLX5_CMD_OP_2RST_QP, 0, NULL, &base->mqp); @@ -1974,7 +1958,7 @@ static void destroy_qp_common(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp) if (qp->create_type == MLX5_QP_KERNEL) destroy_qp_kernel(dev, qp); else if (qp->create_type == MLX5_QP_USER) - destroy_qp_user(&get_pd(qp)->ibpd, qp, base); + destroy_qp_user(dev, &get_pd(qp)->ibpd, qp, base); } static const char *ib_qp_type_str(enum ib_qp_type type) @@ -2180,7 +2164,7 @@ static int modify_raw_packet_eth_prio(struct mlx5_core_dev *dev, int err; inlen = MLX5_ST_SZ_BYTES(modify_tis_in); - in = mlx5_vzalloc(inlen); + in = kvzalloc(inlen, GFP_KERNEL); if (!in) return -ENOMEM; @@ -2205,7 +2189,7 @@ static int modify_raw_packet_tx_affinity(struct mlx5_core_dev *dev, int err; inlen = MLX5_ST_SZ_BYTES(modify_tis_in); - in = mlx5_vzalloc(inlen); + in = kvzalloc(inlen, GFP_KERNEL); if (!in) return -ENOMEM; @@ -2222,56 +2206,65 @@ static int modify_raw_packet_tx_affinity(struct mlx5_core_dev *dev, } static int mlx5_set_path(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp, - const struct ib_ah_attr *ah, + const struct rdma_ah_attr *ah, struct mlx5_qp_path *path, u8 port, int attr_mask, u32 path_flags, const struct ib_qp_attr *attr, bool alt) { - enum rdma_link_layer ll = rdma_port_get_link_layer(&dev->ib_dev, port); + const struct ib_global_route *grh = rdma_ah_read_grh(ah); int err; + enum ib_gid_type gid_type; + u8 ah_flags = rdma_ah_get_ah_flags(ah); + u8 sl = rdma_ah_get_sl(ah); if (attr_mask & IB_QP_PKEY_INDEX) path->pkey_index = cpu_to_be16(alt ? attr->alt_pkey_index : attr->pkey_index); - if (ah->ah_flags & IB_AH_GRH) { - if (ah->grh.sgid_index >= + if (ah_flags & IB_AH_GRH) { + if (grh->sgid_index >= dev->mdev->port_caps[port - 1].gid_table_len) { pr_err("sgid_index (%u) too large. max is %d\n", - ah->grh.sgid_index, + grh->sgid_index, dev->mdev->port_caps[port - 1].gid_table_len); return -EINVAL; } } - if (ll == IB_LINK_LAYER_ETHERNET) { - if (!(ah->ah_flags & IB_AH_GRH)) + if (ah->type == RDMA_AH_ATTR_TYPE_ROCE) { + if (!(ah_flags & IB_AH_GRH)) return -EINVAL; - memcpy(path->rmac, ah->dmac, sizeof(ah->dmac)); + err = mlx5_get_roce_gid_type(dev, port, grh->sgid_index, + &gid_type); + if (err) + return err; + memcpy(path->rmac, ah->roce.dmac, sizeof(ah->roce.dmac)); path->udp_sport = mlx5_get_roce_udp_sport(dev, port, - ah->grh.sgid_index); - path->dci_cfi_prio_sl = (ah->sl & 0x7) << 4; + grh->sgid_index); + path->dci_cfi_prio_sl = (sl & 0x7) << 4; + if (gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP) + path->ecn_dscp = (grh->traffic_class >> 2) & 0x3f; } else { path->fl_free_ar = (path_flags & MLX5_PATH_FLAG_FL) ? 0x80 : 0; path->fl_free_ar |= (path_flags & MLX5_PATH_FLAG_FREE_AR) ? 0x40 : 0; - path->rlid = cpu_to_be16(ah->dlid); - path->grh_mlid = ah->src_path_bits & 0x7f; - if (ah->ah_flags & IB_AH_GRH) + path->rlid = cpu_to_be16(rdma_ah_get_dlid(ah)); + path->grh_mlid = rdma_ah_get_path_bits(ah) & 0x7f; + if (ah_flags & IB_AH_GRH) path->grh_mlid |= 1 << 7; - path->dci_cfi_prio_sl = ah->sl & 0xf; + path->dci_cfi_prio_sl = sl & 0xf; } - if (ah->ah_flags & IB_AH_GRH) { - path->mgid_index = ah->grh.sgid_index; - path->hop_limit = ah->grh.hop_limit; + if (ah_flags & IB_AH_GRH) { + path->mgid_index = grh->sgid_index; + path->hop_limit = grh->hop_limit; path->tclass_flowlabel = - cpu_to_be32((ah->grh.traffic_class << 20) | - (ah->grh.flow_label)); - memcpy(path->rgid, ah->grh.dgid.raw, 16); + cpu_to_be32((grh->traffic_class << 20) | + (grh->flow_label)); + memcpy(path->rgid, grh->dgid.raw, 16); } - err = ib_rate_to_mlx5(dev, ah->static_rate); + err = ib_rate_to_mlx5(dev, rdma_ah_get_static_rate(ah)); if (err < 0) return err; path->static_rate = err; @@ -2283,7 +2276,7 @@ static int mlx5_set_path(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp, if ((qp->ibqp.qp_type == IB_QPT_RAW_PACKET) && qp->sq.wqe_cnt) return modify_raw_packet_eth_prio(dev->mdev, &qp->raw_packet_qp.sq, - ah->sl & 0xf); + sl & 0xf); return 0; } @@ -2441,7 +2434,7 @@ static int modify_raw_packet_qp_rq(struct mlx5_ib_dev *dev, int err; inlen = MLX5_ST_SZ_BYTES(modify_rq_in); - in = mlx5_vzalloc(inlen); + in = kvzalloc(inlen, GFP_KERNEL); if (!in) return -ENOMEM; @@ -2453,7 +2446,7 @@ static int modify_raw_packet_qp_rq(struct mlx5_ib_dev *dev, if (raw_qp_param->set_mask & MLX5_RAW_QP_MOD_SET_RQ_Q_CTR_ID) { if (MLX5_CAP_GEN(dev->mdev, modify_rq_counter_set_id)) { MLX5_SET64(modify_rq_in, in, modify_bitmask, - MLX5_MODIFY_RQ_IN_MODIFY_BITMASK_MODIFY_RQ_COUNTER_SET_ID); + MLX5_MODIFY_RQ_IN_MODIFY_BITMASK_RQ_COUNTER_SET_ID); MLX5_SET(rqc, rqc, counter_set_id, raw_qp_param->rq_q_ctr_id); } else pr_info_once("%s: RAW PACKET QP counters are not supported on current FW\n", @@ -2486,7 +2479,7 @@ static int modify_raw_packet_qp_sq(struct mlx5_core_dev *dev, int err; inlen = MLX5_ST_SZ_BYTES(modify_sq_in); - in = mlx5_vzalloc(inlen); + in = kvzalloc(inlen, GFP_KERNEL); if (!in) return -ENOMEM; @@ -2808,7 +2801,7 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp, qp->port) - 1; mibport = &dev->port[port_num]; context->qp_counter_set_usr_page |= - cpu_to_be32((u32)(mibport->q_cnt_id) << 24); + cpu_to_be32((u32)(mibport->cnts.set_id) << 24); } if (!ibqp->uobject && cur_state == IB_QPS_RESET && new_state == IB_QPS_INIT) @@ -2823,16 +2816,6 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp, if (mlx5_st < 0) goto out; - /* If moving to a reset or error state, we must disable page faults on - * this QP and flush all current page faults. Otherwise a stale page - * fault may attempt to work on this QP after it is reset and moved - * again to RTS, and may cause the driver and the device to get out of - * sync. */ - if (cur_state != IB_QPS_RESET && cur_state != IB_QPS_ERR && - (new_state == IB_QPS_RESET || new_state == IB_QPS_ERR) && - (qp->ibqp.qp_type != IB_QPT_RAW_PACKET)) - mlx5_ib_qp_disable_pagefaults(qp); - if (mlx5_cur >= MLX5_QP_NUM_STATE || mlx5_new >= MLX5_QP_NUM_STATE || !optab[mlx5_cur][mlx5_new]) goto out; @@ -2846,7 +2829,7 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp, raw_qp_param.operation = op; if (cur_state == IB_QPS_RESET && new_state == IB_QPS_INIT) { - raw_qp_param.rq_q_ctr_id = mibport->q_cnt_id; + raw_qp_param.rq_q_ctr_id = mibport->cnts.set_id; raw_qp_param.set_mask |= MLX5_RAW_QP_MOD_SET_RQ_Q_CTR_ID; } @@ -2864,10 +2847,6 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp, if (err) goto out; - if (cur_state == IB_QPS_RESET && new_state == IB_QPS_INIT && - (qp->ibqp.qp_type != IB_QPT_RAW_PACKET)) - mlx5_ib_qp_enable_pagefaults(qp); - qp->state = new_state; if (attr_mask & IB_QP_ACCESS_FLAGS) @@ -3029,20 +3008,20 @@ static void *set_eth_seg(struct mlx5_wqe_eth_seg *eseg, if (wr->opcode == IB_WR_LSO) { struct ib_ud_wr *ud_wr = container_of(wr, struct ib_ud_wr, wr); - int size_of_inl_hdr_start = sizeof(eseg->inline_hdr_start); + int size_of_inl_hdr_start = sizeof(eseg->inline_hdr.start); u64 left, leftlen, copysz; void *pdata = ud_wr->header; left = ud_wr->hlen; eseg->mss = cpu_to_be16(ud_wr->mss); - eseg->inline_hdr_sz = cpu_to_be16(left); + eseg->inline_hdr.sz = cpu_to_be16(left); /* * check if there is space till the end of queue, if yes, * copy all in one shot, otherwise copy till the end of queue, * rollback and than the copy the left */ - leftlen = qend - (void *)eseg->inline_hdr_start; + leftlen = qend - (void *)eseg->inline_hdr.start; copysz = min_t(u64, leftlen, left); memcpy(seg - size_of_inl_hdr_start, pdata, copysz); @@ -3080,9 +3059,10 @@ static void set_data_ptr_seg(struct mlx5_wqe_data_seg *dseg, struct ib_sge *sg) dseg->addr = cpu_to_be64(sg->addr); } -static __be16 get_klm_octo(int npages) +static u64 get_xlt_octo(u64 bytes) { - return cpu_to_be16(ALIGN(npages, 8) / 2); + return ALIGN(bytes, MLX5_IB_UMR_XLT_ALIGNMENT) / + MLX5_IB_UMR_OCTOWORD; } static __be64 frwr_mkey_mask(void) @@ -3127,18 +3107,14 @@ static __be64 sig_mkey_mask(void) } static void set_reg_umr_seg(struct mlx5_wqe_umr_ctrl_seg *umr, - struct mlx5_ib_mr *mr) + struct mlx5_ib_mr *mr) { - int ndescs = mr->ndescs; + int size = mr->ndescs * mr->desc_size; memset(umr, 0, sizeof(*umr)); - if (mr->access_mode == MLX5_MKC_ACCESS_MODE_KLMS) - /* KLMs take twice the size of MTTs */ - ndescs *= 2; - umr->flags = MLX5_UMR_CHECK_NOT_FREE; - umr->klm_octowords = get_klm_octo(ndescs); + umr->xlt_octowords = cpu_to_be16(get_xlt_octo(size)); umr->mkey_mask = frwr_mkey_mask(); } @@ -3149,37 +3125,17 @@ static void set_linv_umr_seg(struct mlx5_wqe_umr_ctrl_seg *umr) umr->flags = MLX5_UMR_INLINE; } -static __be64 get_umr_reg_mr_mask(int atomic) +static __be64 get_umr_enable_mr_mask(void) { u64 result; - result = MLX5_MKEY_MASK_LEN | - MLX5_MKEY_MASK_PAGE_SIZE | - MLX5_MKEY_MASK_START_ADDR | - MLX5_MKEY_MASK_PD | - MLX5_MKEY_MASK_LR | - MLX5_MKEY_MASK_LW | - MLX5_MKEY_MASK_KEY | - MLX5_MKEY_MASK_RR | - MLX5_MKEY_MASK_RW | + result = MLX5_MKEY_MASK_KEY | MLX5_MKEY_MASK_FREE; - if (atomic) - result |= MLX5_MKEY_MASK_A; - return cpu_to_be64(result); } -static __be64 get_umr_unreg_mr_mask(void) -{ - u64 result; - - result = MLX5_MKEY_MASK_FREE; - - return cpu_to_be64(result); -} - -static __be64 get_umr_update_mtt_mask(void) +static __be64 get_umr_disable_mr_mask(void) { u64 result; @@ -3194,23 +3150,22 @@ static __be64 get_umr_update_translation_mask(void) result = MLX5_MKEY_MASK_LEN | MLX5_MKEY_MASK_PAGE_SIZE | - MLX5_MKEY_MASK_START_ADDR | - MLX5_MKEY_MASK_KEY | - MLX5_MKEY_MASK_FREE; + MLX5_MKEY_MASK_START_ADDR; return cpu_to_be64(result); } -static __be64 get_umr_update_access_mask(void) +static __be64 get_umr_update_access_mask(int atomic) { u64 result; - result = MLX5_MKEY_MASK_LW | + result = MLX5_MKEY_MASK_LR | + MLX5_MKEY_MASK_LW | MLX5_MKEY_MASK_RR | - MLX5_MKEY_MASK_RW | - MLX5_MKEY_MASK_A | - MLX5_MKEY_MASK_KEY | - MLX5_MKEY_MASK_FREE; + MLX5_MKEY_MASK_RW; + + if (atomic) + result |= MLX5_MKEY_MASK_A; return cpu_to_be64(result); } @@ -3219,9 +3174,7 @@ static __be64 get_umr_update_pd_mask(void) { u64 result; - result = MLX5_MKEY_MASK_PD | - MLX5_MKEY_MASK_KEY | - MLX5_MKEY_MASK_FREE; + result = MLX5_MKEY_MASK_PD; return cpu_to_be64(result); } @@ -3238,24 +3191,24 @@ static void set_reg_umr_segment(struct mlx5_wqe_umr_ctrl_seg *umr, else umr->flags = MLX5_UMR_CHECK_NOT_FREE; /* fail if not free */ - if (!(wr->send_flags & MLX5_IB_SEND_UMR_UNREG)) { - umr->klm_octowords = get_klm_octo(umrwr->npages); - if (wr->send_flags & MLX5_IB_SEND_UMR_UPDATE_MTT) { - umr->mkey_mask = get_umr_update_mtt_mask(); - umr->bsf_octowords = get_klm_octo(umrwr->target.offset); - umr->flags |= MLX5_UMR_TRANSLATION_OFFSET_EN; - } - if (wr->send_flags & MLX5_IB_SEND_UMR_UPDATE_TRANSLATION) - umr->mkey_mask |= get_umr_update_translation_mask(); - if (wr->send_flags & MLX5_IB_SEND_UMR_UPDATE_ACCESS) - umr->mkey_mask |= get_umr_update_access_mask(); - if (wr->send_flags & MLX5_IB_SEND_UMR_UPDATE_PD) - umr->mkey_mask |= get_umr_update_pd_mask(); - if (!umr->mkey_mask) - umr->mkey_mask = get_umr_reg_mr_mask(atomic); - } else { - umr->mkey_mask = get_umr_unreg_mr_mask(); + umr->xlt_octowords = cpu_to_be16(get_xlt_octo(umrwr->xlt_size)); + if (wr->send_flags & MLX5_IB_SEND_UMR_UPDATE_XLT) { + u64 offset = get_xlt_octo(umrwr->offset); + + umr->xlt_offset = cpu_to_be16(offset & 0xffff); + umr->xlt_offset_47_16 = cpu_to_be32(offset >> 16); + umr->flags |= MLX5_UMR_TRANSLATION_OFFSET_EN; + } + if (wr->send_flags & MLX5_IB_SEND_UMR_UPDATE_TRANSLATION) + umr->mkey_mask |= get_umr_update_translation_mask(); + if (wr->send_flags & MLX5_IB_SEND_UMR_UPDATE_PD_ACCESS) { + umr->mkey_mask |= get_umr_update_access_mask(atomic); + umr->mkey_mask |= get_umr_update_pd_mask(); } + if (wr->send_flags & MLX5_IB_SEND_UMR_ENABLE_MR) + umr->mkey_mask |= get_umr_enable_mr_mask(); + if (wr->send_flags & MLX5_IB_SEND_UMR_DISABLE_MR) + umr->mkey_mask |= get_umr_disable_mr_mask(); if (!wr->num_sge) umr->flags |= MLX5_UMR_INLINE; @@ -3303,17 +3256,17 @@ static void set_reg_mkey_segment(struct mlx5_mkey_seg *seg, struct ib_send_wr *w struct mlx5_umr_wr *umrwr = umr_wr(wr); memset(seg, 0, sizeof(*seg)); - if (wr->send_flags & MLX5_IB_SEND_UMR_UNREG) { + if (wr->send_flags & MLX5_IB_SEND_UMR_DISABLE_MR) seg->status = MLX5_MKEY_STATUS_FREE; - return; - } seg->flags = convert_access(umrwr->access_flags); - if (!(wr->send_flags & MLX5_IB_SEND_UMR_UPDATE_MTT)) { - if (umrwr->pd) - seg->flags_pd = cpu_to_be32(to_mpd(umrwr->pd)->pdn); - seg->start_addr = cpu_to_be64(umrwr->target.virt_addr); - } + if (umrwr->pd) + seg->flags_pd = cpu_to_be32(to_mpd(umrwr->pd)->pdn); + if (wr->send_flags & MLX5_IB_SEND_UMR_UPDATE_TRANSLATION && + !umrwr->length) + seg->flags_pd |= cpu_to_be32(MLX5_MKEY_LEN64); + + seg->start_addr = cpu_to_be64(umrwr->virt_addr); seg->len = cpu_to_be64(umrwr->length); seg->log2_page_size = umrwr->page_shift; seg->qpn_mkey7_0 = cpu_to_be32(0xffffff00 | @@ -3611,7 +3564,7 @@ static int set_sig_data_segment(struct ib_sig_handover_wr *wr, } static void set_sig_mkey_segment(struct mlx5_mkey_seg *seg, - struct ib_sig_handover_wr *wr, u32 nelements, + struct ib_sig_handover_wr *wr, u32 size, u32 length, u32 pdn) { struct ib_mr *sig_mr = wr->sig_mr; @@ -3626,17 +3579,17 @@ static void set_sig_mkey_segment(struct mlx5_mkey_seg *seg, seg->flags_pd = cpu_to_be32(MLX5_MKEY_REMOTE_INVAL | sigerr << 26 | MLX5_MKEY_BSF_EN | pdn); seg->len = cpu_to_be64(length); - seg->xlt_oct_size = cpu_to_be32(be16_to_cpu(get_klm_octo(nelements))); + seg->xlt_oct_size = cpu_to_be32(get_xlt_octo(size)); seg->bsfs_octo_size = cpu_to_be32(MLX5_MKEY_BSF_OCTO_SIZE); } static void set_sig_umr_segment(struct mlx5_wqe_umr_ctrl_seg *umr, - u32 nelements) + u32 size) { memset(umr, 0, sizeof(*umr)); umr->flags = MLX5_FLAGS_INLINE | MLX5_FLAGS_CHECK_FREE; - umr->klm_octowords = get_klm_octo(nelements); + umr->xlt_octowords = cpu_to_be16(get_xlt_octo(size)); umr->bsf_octowords = cpu_to_be16(MLX5_MKEY_BSF_OCTO_SIZE); umr->mkey_mask = sig_mkey_mask(); } @@ -3648,7 +3601,7 @@ static int set_sig_umr_wr(struct ib_send_wr *send_wr, struct mlx5_ib_qp *qp, struct ib_sig_handover_wr *wr = sig_handover_wr(send_wr); struct mlx5_ib_mr *sig_mr = to_mmr(wr->sig_mr); u32 pdn = get_pd(qp)->pdn; - u32 klm_oct_size; + u32 xlt_size; int region_len, ret; if (unlikely(wr->wr.num_sge != 1) || @@ -3670,15 +3623,15 @@ static int set_sig_umr_wr(struct ib_send_wr *send_wr, struct mlx5_ib_qp *qp, * then we use strided block format (3 octowords), * else we use single KLM (1 octoword) **/ - klm_oct_size = wr->prot ? 3 : 1; + xlt_size = wr->prot ? 0x30 : sizeof(struct mlx5_klm); - set_sig_umr_segment(*seg, klm_oct_size); + set_sig_umr_segment(*seg, xlt_size); *seg += sizeof(struct mlx5_wqe_umr_ctrl_seg); *size += sizeof(struct mlx5_wqe_umr_ctrl_seg) / 16; if (unlikely((*seg == qp->sq.qend))) *seg = mlx5_get_send_wqe(qp, 0); - set_sig_mkey_segment(*seg, wr, klm_oct_size, region_len, pdn); + set_sig_mkey_segment(*seg, wr, xlt_size, region_len, pdn); *seg += sizeof(struct mlx5_mkey_seg); *size += sizeof(struct mlx5_mkey_seg) / 16; if (unlikely((*seg == qp->sq.qend))) @@ -3708,8 +3661,9 @@ static int set_psv_wr(struct ib_sig_domain *domain, psv_seg->ref_tag = cpu_to_be32(domain->sig.dif.ref_tag); break; default: - pr_err("Bad signature type given.\n"); - return 1; + pr_err("Bad signature type (%d) is given.\n", + domain->sig_type); + return -EINVAL; } *seg += sizeof(*psv_seg); @@ -3784,42 +3738,6 @@ static void dump_wqe(struct mlx5_ib_qp *qp, int idx, int size_16) } } -static void mlx5_bf_copy(u64 __iomem *dst, u64 *src, - unsigned bytecnt, struct mlx5_ib_qp *qp) -{ - while (bytecnt > 0) { - __iowrite64_copy(dst++, src++, 8); - __iowrite64_copy(dst++, src++, 8); - __iowrite64_copy(dst++, src++, 8); - __iowrite64_copy(dst++, src++, 8); - __iowrite64_copy(dst++, src++, 8); - __iowrite64_copy(dst++, src++, 8); - __iowrite64_copy(dst++, src++, 8); - __iowrite64_copy(dst++, src++, 8); - bytecnt -= 64; - if (unlikely(src == qp->sq.qend)) - src = mlx5_get_send_wqe(qp, 0); - } -} - -static u8 get_fence(u8 fence, struct ib_send_wr *wr) -{ - if (unlikely(wr->opcode == IB_WR_LOCAL_INV && - wr->send_flags & IB_SEND_FENCE)) - return MLX5_FENCE_MODE_STRONG_ORDERING; - - if (unlikely(fence)) { - if (wr->send_flags & IB_SEND_FENCE) - return MLX5_FENCE_MODE_SMALL_AND_FENCE; - else - return fence; - } else if (unlikely(wr->send_flags & IB_SEND_FENCE)) { - return MLX5_FENCE_MODE_FENCE; - } - - return 0; -} - static int begin_wqe(struct mlx5_ib_qp *qp, void **seg, struct mlx5_wqe_ctrl_seg **ctrl, struct ib_send_wr *wr, unsigned *idx, @@ -3848,8 +3766,7 @@ static int begin_wqe(struct mlx5_ib_qp *qp, void **seg, static void finish_wqe(struct mlx5_ib_qp *qp, struct mlx5_wqe_ctrl_seg *ctrl, u8 size, unsigned idx, u64 wr_id, - int nreq, u8 fence, u8 next_fence, - u32 mlx5_opcode) + int nreq, u8 fence, u32 mlx5_opcode) { u8 opmod = 0; @@ -3857,7 +3774,6 @@ static void finish_wqe(struct mlx5_ib_qp *qp, mlx5_opcode | ((u32)opmod << 24)); ctrl->qpn_ds = cpu_to_be32(size | (qp->trans_qp.base.mqp.qpn << 8)); ctrl->fm_ce_se |= fence; - qp->fm_cache = next_fence; if (unlikely(qp->wq_sig)) ctrl->signature = wq_sig(ctrl); @@ -3897,7 +3813,7 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, return mlx5_ib_gsi_post_send(ibqp, wr, bad_wr); qp = to_mqp(ibqp); - bf = qp->bf; + bf = &qp->bf; qend = qp->sq.qend; spin_lock_irqsave(&qp->sq.lock, flags); @@ -3917,7 +3833,6 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, goto out; } - fence = qp->fm_cache; num_sge = wr->num_sge; if (unlikely(num_sge > qp->sq.max_gs)) { mlx5_ib_warn(dev, "\n"); @@ -3934,6 +3849,19 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, goto out; } + if (wr->opcode == IB_WR_LOCAL_INV || + wr->opcode == IB_WR_REG_MR) { + fence = dev->umr_fence; + next_fence = MLX5_FENCE_MODE_INITIATOR_SMALL; + } else if (wr->send_flags & IB_SEND_FENCE) { + if (qp->next_fence) + fence = MLX5_FENCE_MODE_SMALL_AND_FENCE; + else + fence = MLX5_FENCE_MODE_FENCE; + } else { + fence = qp->next_fence; + } + switch (ibqp->qp_type) { case IB_QPT_XRC_INI: xrc = seg; @@ -3960,7 +3888,6 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, goto out; case IB_WR_LOCAL_INV: - next_fence = MLX5_FENCE_MODE_INITIATOR_SMALL; qp->sq.wr_data[idx] = IB_WR_LOCAL_INV; ctrl->imm = cpu_to_be32(wr->ex.invalidate_rkey); set_linv_wr(qp, &seg, &size); @@ -3968,7 +3895,6 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, break; case IB_WR_REG_MR: - next_fence = MLX5_FENCE_MODE_INITIATOR_SMALL; qp->sq.wr_data[idx] = IB_WR_REG_MR; ctrl->imm = cpu_to_be32(reg_wr(wr)->key); err = set_reg_wr(qp, reg_wr(wr), &seg, &size); @@ -3991,9 +3917,8 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, goto out; } - finish_wqe(qp, ctrl, size, idx, wr->wr_id, - nreq, get_fence(fence, wr), - next_fence, MLX5_OPCODE_UMR); + finish_wqe(qp, ctrl, size, idx, wr->wr_id, nreq, + fence, MLX5_OPCODE_UMR); /* * SET_PSV WQEs are not signaled and solicited * on error @@ -4018,9 +3943,8 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, goto out; } - finish_wqe(qp, ctrl, size, idx, wr->wr_id, - nreq, get_fence(fence, wr), - next_fence, MLX5_OPCODE_SET_PSV); + finish_wqe(qp, ctrl, size, idx, wr->wr_id, nreq, + fence, MLX5_OPCODE_SET_PSV); err = begin_wqe(qp, &seg, &ctrl, wr, &idx, &size, nreq); if (err) { @@ -4030,7 +3954,6 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, goto out; } - next_fence = MLX5_FENCE_MODE_INITIATOR_SMALL; err = set_psv_wr(&sig_handover_wr(wr)->sig_attrs->wire, mr->sig->psv_wire.psv_idx, &seg, &size); @@ -4040,9 +3963,9 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, goto out; } - finish_wqe(qp, ctrl, size, idx, wr->wr_id, - nreq, get_fence(fence, wr), - next_fence, MLX5_OPCODE_SET_PSV); + finish_wqe(qp, ctrl, size, idx, wr->wr_id, nreq, + fence, MLX5_OPCODE_SET_PSV); + qp->next_fence = MLX5_FENCE_MODE_INITIATOR_SMALL; num_sge = 0; goto skip_psv; @@ -4067,6 +3990,12 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, break; case IB_QPT_SMI: + if (unlikely(!mdev->port_caps[qp->port - 1].has_smi)) { + mlx5_ib_warn(dev, "Send SMP MADs is not allowed\n"); + err = -EPERM; + *bad_wr = wr; + goto out; + } case MLX5_IB_QPT_HW_GSI: set_datagram_seg(seg, wr); seg += sizeof(struct mlx5_wqe_datagram_seg); @@ -4147,8 +4076,8 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, } } - finish_wqe(qp, ctrl, size, idx, wr->wr_id, nreq, - get_fence(fence, wr), next_fence, + qp->next_fence = next_fence; + finish_wqe(qp, ctrl, size, idx, wr->wr_id, nreq, fence, mlx5_ib_opcode[wr->opcode]); skip_psv: if (0) @@ -4170,28 +4099,13 @@ out: * we hit doorbell */ wmb(); - if (bf->need_lock) - spin_lock(&bf->lock); - else - __acquire(&bf->lock); - - /* TBD enable WC */ - if (0 && nreq == 1 && bf->uuarn && inl && size > 1 && size <= bf->buf_size / 16) { - mlx5_bf_copy(bf->reg + bf->offset, (u64 *)ctrl, ALIGN(size * 16, 64), qp); - /* wc_wmb(); */ - } else { - mlx5_write64((__be32 *)ctrl, bf->regreg + bf->offset, - MLX5_GET_DOORBELL_LOCK(&bf->lock32)); - /* Make sure doorbells don't leak out of SQ spinlock - * and reach the HCA out of order. - */ - mmiowb(); - } + /* currently we support only regular doorbells */ + mlx5_write64((__be32 *)ctrl, bf->bfreg->map + bf->offset, NULL); + /* Make sure doorbells don't leak out of SQ spinlock + * and reach the HCA out of order. + */ + mmiowb(); bf->offset ^= bf->buf_size; - if (bf->need_lock) - spin_unlock(&bf->lock); - else - __release(&bf->lock); } spin_unlock_irqrestore(&qp->sq.lock, flags); @@ -4324,33 +4238,36 @@ static int to_ib_qp_access_flags(int mlx5_flags) return ib_flags; } -static void to_ib_ah_attr(struct mlx5_ib_dev *ibdev, struct ib_ah_attr *ib_ah_attr, - struct mlx5_qp_path *path) +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(ib_ah_attr, 0, sizeof(*ib_ah_attr)); - ib_ah_attr->port_num = path->port; + memset(ah_attr, 0, sizeof(*ah_attr)); - if (ib_ah_attr->port_num == 0 || - ib_ah_attr->port_num > MLX5_CAP_GEN(dev, num_ports)) + 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)) return; - ib_ah_attr->sl = path->dci_cfi_prio_sl & 0xf; + rdma_ah_set_port_num(ah_attr, path->port); + rdma_ah_set_sl(ah_attr, path->dci_cfi_prio_sl & 0xf); - ib_ah_attr->dlid = be16_to_cpu(path->rlid); - ib_ah_attr->src_path_bits = path->grh_mlid & 0x7f; - ib_ah_attr->static_rate = path->static_rate ? path->static_rate - 5 : 0; - ib_ah_attr->ah_flags = (path->grh_mlid & (1 << 7)) ? IB_AH_GRH : 0; - if (ib_ah_attr->ah_flags) { - ib_ah_attr->grh.sgid_index = path->mgid_index; - ib_ah_attr->grh.hop_limit = path->hop_limit; - ib_ah_attr->grh.traffic_class = - (be32_to_cpu(path->tclass_flowlabel) >> 20) & 0xff; - ib_ah_attr->grh.flow_label = - be32_to_cpu(path->tclass_flowlabel) & 0xfffff; - memcpy(ib_ah_attr->grh.dgid.raw, - path->rgid, sizeof(ib_ah_attr->grh.dgid.raw)); + rdma_ah_set_dlid(ah_attr, be16_to_cpu(path->rlid)); + rdma_ah_set_path_bits(ah_attr, path->grh_mlid & 0x7f); + rdma_ah_set_static_rate(ah_attr, + path->static_rate ? path->static_rate - 5 : 0); + if (path->grh_mlid & (1 << 7)) { + u32 tc_fl = be32_to_cpu(path->tclass_flowlabel); + + rdma_ah_set_grh(ah_attr, NULL, + tc_fl & 0xfffff, + path->mgid_index, + path->hop_limit, + (tc_fl >> 20) & 0xff); + rdma_ah_set_dgid_raw(ah_attr, path->rgid); } } @@ -4364,7 +4281,7 @@ static int query_raw_packet_qp_sq_state(struct mlx5_ib_dev *dev, int err; inlen = MLX5_ST_SZ_BYTES(query_sq_out); - out = mlx5_vzalloc(inlen); + out = kvzalloc(inlen, GFP_KERNEL); if (!out) return -ENOMEM; @@ -4391,7 +4308,7 @@ static int query_raw_packet_qp_rq_state(struct mlx5_ib_dev *dev, int err; inlen = MLX5_ST_SZ_BYTES(query_rq_out); - out = mlx5_vzalloc(inlen); + out = kvzalloc(inlen, GFP_KERNEL); if (!out) return -ENOMEM; @@ -4515,11 +4432,12 @@ static int query_qp_attr(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp, to_ib_qp_access_flags(be32_to_cpu(context->params2)); if (qp->ibqp.qp_type == IB_QPT_RC || qp->ibqp.qp_type == IB_QPT_UC) { - to_ib_ah_attr(dev, &qp_attr->ah_attr, &context->pri_path); - to_ib_ah_attr(dev, &qp_attr->alt_ah_attr, &context->alt_path); + to_rdma_ah_attr(dev, &qp_attr->ah_attr, &context->pri_path); + to_rdma_ah_attr(dev, &qp_attr->alt_ah_attr, &context->alt_path); qp_attr->alt_pkey_index = be16_to_cpu(context->alt_path.pkey_index); - qp_attr->alt_port_num = qp_attr->alt_ah_attr.port_num; + qp_attr->alt_port_num = + rdma_ah_get_port_num(&qp_attr->alt_ah_attr); } qp_attr->pkey_index = be16_to_cpu(context->pri_path.pkey_index); @@ -4559,14 +4477,6 @@ int mlx5_ib_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr, return mlx5_ib_gsi_query_qp(ibqp, qp_attr, qp_attr_mask, qp_init_attr); -#ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING - /* - * Wait for any outstanding page faults, in case the user frees memory - * based upon this query's result. - */ - flush_workqueue(mlx5_ib_page_fault_wq); -#endif - mutex_lock(&qp->mutex); if (qp->ibqp.qp_type == IB_QPT_RAW_PACKET) { @@ -4691,6 +4601,7 @@ static int create_rq(struct mlx5_ib_rwq *rwq, struct ib_pd *pd, struct ib_wq_init_attr *init_attr) { struct mlx5_ib_dev *dev; + int has_net_offloads; __be64 *rq_pas0; void *in; void *rqc; @@ -4701,7 +4612,7 @@ static int create_rq(struct mlx5_ib_rwq *rwq, struct ib_pd *pd, dev = to_mdev(pd->device); inlen = MLX5_ST_SZ_BYTES(create_rq_in) + sizeof(u64) * rwq->rq_num_pas; - in = mlx5_vzalloc(inlen); + in = kvzalloc(inlen, GFP_KERNEL); if (!in) return -ENOMEM; @@ -4722,9 +4633,28 @@ static int create_rq(struct mlx5_ib_rwq *rwq, struct ib_pd *pd, MLX5_SET(wq, wq, log_wq_pg_sz, rwq->log_page_size); MLX5_SET(wq, wq, wq_signature, rwq->wq_sig); MLX5_SET64(wq, wq, dbr_addr, rwq->db.dma); + has_net_offloads = MLX5_CAP_GEN(dev->mdev, eth_net_offloads); + if (init_attr->create_flags & IB_WQ_FLAGS_CVLAN_STRIPPING) { + if (!(has_net_offloads && MLX5_CAP_ETH(dev->mdev, vlan_cap))) { + mlx5_ib_dbg(dev, "VLAN offloads are not supported\n"); + err = -EOPNOTSUPP; + goto out; + } + } else { + MLX5_SET(rqc, rqc, vsd, 1); + } + if (init_attr->create_flags & IB_WQ_FLAGS_SCATTER_FCS) { + if (!(has_net_offloads && MLX5_CAP_ETH(dev->mdev, scatter_fcs))) { + mlx5_ib_dbg(dev, "Scatter FCS is not supported\n"); + err = -EOPNOTSUPP; + goto out; + } + MLX5_SET(rqc, rqc, scatter_fcs, 1); + } rq_pas0 = (__be64 *)MLX5_ADDR_OF(wq, wq, pas); mlx5_ib_populate_pas(dev, rwq->umem, rwq->page_shift, rq_pas0, 0); err = mlx5_core_create_rq_tracked(dev->mdev, in, inlen, &rwq->core_qp); +out: kvfree(in); return err; } @@ -4912,7 +4842,7 @@ struct ib_rwq_ind_table *mlx5_ib_create_rwq_ind_table(struct ib_device *device, return ERR_PTR(-ENOMEM); inlen = MLX5_ST_SZ_BYTES(create_rqt_in) + sizeof(u32) * sz; - in = mlx5_vzalloc(inlen); + in = kvzalloc(inlen, GFP_KERNEL); if (!in) { err = -ENOMEM; goto err; @@ -4991,7 +4921,7 @@ int mlx5_ib_modify_wq(struct ib_wq *wq, struct ib_wq_attr *wq_attr, return -EOPNOTSUPP; inlen = MLX5_ST_SZ_BYTES(modify_rq_in); - in = mlx5_vzalloc(inlen); + in = kvzalloc(inlen, GFP_KERNEL); if (!in) return -ENOMEM; @@ -5008,10 +4938,38 @@ int mlx5_ib_modify_wq(struct ib_wq *wq, struct ib_wq_attr *wq_attr, MLX5_SET(modify_rq_in, in, rq_state, curr_wq_state); MLX5_SET(rqc, rqc, state, wq_state); + if (wq_attr_mask & IB_WQ_FLAGS) { + if (wq_attr->flags_mask & IB_WQ_FLAGS_CVLAN_STRIPPING) { + if (!(MLX5_CAP_GEN(dev->mdev, eth_net_offloads) && + MLX5_CAP_ETH(dev->mdev, vlan_cap))) { + mlx5_ib_dbg(dev, "VLAN offloads are not " + "supported\n"); + err = -EOPNOTSUPP; + goto out; + } + MLX5_SET64(modify_rq_in, in, modify_bitmask, + MLX5_MODIFY_RQ_IN_MODIFY_BITMASK_VSD); + MLX5_SET(rqc, rqc, vsd, + (wq_attr->flags & IB_WQ_FLAGS_CVLAN_STRIPPING) ? 0 : 1); + } + } + + if (curr_wq_state == IB_WQS_RESET && wq_state == IB_WQS_RDY) { + if (MLX5_CAP_GEN(dev->mdev, modify_rq_counter_set_id)) { + MLX5_SET64(modify_rq_in, in, modify_bitmask, + MLX5_MODIFY_RQ_IN_MODIFY_BITMASK_RQ_COUNTER_SET_ID); + MLX5_SET(rqc, rqc, counter_set_id, + dev->port->cnts.set_id); + } else + pr_info_once("%s: Receive WQ counters are not supported on current FW\n", + dev->ib_dev.name); + } + err = mlx5_core_modify_rq(dev->mdev, rwq->core_qp.qpn, in, inlen); - kvfree(in); if (!err) rwq->ibwq.state = (wq_state == MLX5_RQC_STATE_ERR) ? IB_WQS_ERR : wq_state; +out: + kvfree(in); return err; } diff --git a/drivers/infiniband/hw/mlx5/srq.c b/drivers/infiniband/hw/mlx5/srq.c index 6f4397ee1ed6..43707b101f47 100644 --- a/drivers/infiniband/hw/mlx5/srq.c +++ b/drivers/infiniband/hw/mlx5/srq.c @@ -127,7 +127,7 @@ static int create_srq_user(struct ib_pd *pd, struct mlx5_ib_srq *srq, goto err_umem; } - in->pas = mlx5_vzalloc(sizeof(*in->pas) * ncont); + in->pas = kvzalloc(sizeof(*in->pas) * ncont, GFP_KERNEL); if (!in->pas) { err = -ENOMEM; goto err_umem; @@ -165,8 +165,6 @@ static int create_srq_kernel(struct mlx5_ib_dev *dev, struct mlx5_ib_srq *srq, int err; int i; struct mlx5_wqe_srq_next_seg *next; - int page_shift; - int npages; err = mlx5_db_alloc(dev->mdev, &srq->db); if (err) { @@ -179,7 +177,6 @@ static int create_srq_kernel(struct mlx5_ib_dev *dev, struct mlx5_ib_srq *srq, err = -ENOMEM; goto err_db; } - page_shift = srq->buf.page_shift; srq->head = 0; srq->tail = srq->msrq.max - 1; @@ -191,10 +188,8 @@ static int create_srq_kernel(struct mlx5_ib_dev *dev, struct mlx5_ib_srq *srq, cpu_to_be16((i + 1) & (srq->msrq.max - 1)); } - npages = DIV_ROUND_UP(srq->buf.npages, 1 << (page_shift - PAGE_SHIFT)); - mlx5_ib_dbg(dev, "buf_size %d, page_shift %d, npages %d, calc npages %d\n", - buf_size, page_shift, srq->buf.npages, npages); - in->pas = mlx5_vzalloc(sizeof(*in->pas) * npages); + mlx5_ib_dbg(dev, "srq->buf.page_shift = %d\n", srq->buf.page_shift); + in->pas = kvzalloc(sizeof(*in->pas) * srq->buf.npages, GFP_KERNEL); if (!in->pas) { err = -ENOMEM; goto err_buf; @@ -208,7 +203,7 @@ static int create_srq_kernel(struct mlx5_ib_dev *dev, struct mlx5_ib_srq *srq, } srq->wq_sig = !!srq_signature; - in->log_page_size = page_shift - MLX5_ADAPTER_PAGE_SHIFT; + in->log_page_size = srq->buf.page_shift - MLX5_ADAPTER_PAGE_SHIFT; if (MLX5_CAP_GEN(dev->mdev, cqe_version) == MLX5_CQE_VERSION_V1 && in->type == IB_SRQT_XRC) in->user_index = MLX5_IB_DEFAULT_UIDX; diff --git a/drivers/infiniband/hw/mthca/mthca_av.c b/drivers/infiniband/hw/mthca/mthca_av.c index c9f0f364f484..2aec9908c40a 100644 --- a/drivers/infiniband/hw/mthca/mthca_av.c +++ b/drivers/infiniband/hw/mthca/mthca_av.c @@ -152,7 +152,7 @@ u8 mthca_get_rate(struct mthca_dev *dev, int static_rate, u8 port) int mthca_create_ah(struct mthca_dev *dev, struct mthca_pd *pd, - struct ib_ah_attr *ah_attr, + struct rdma_ah_attr *ah_attr, struct mthca_ah *ah) { u32 index = -1; @@ -196,21 +196,26 @@ on_hca_fail: ah->key = pd->ntmr.ibmr.lkey; - av->port_pd = cpu_to_be32(pd->pd_num | (ah_attr->port_num << 24)); - av->g_slid = ah_attr->src_path_bits; - av->dlid = cpu_to_be16(ah_attr->dlid); + av->port_pd = cpu_to_be32(pd->pd_num | + (rdma_ah_get_port_num(ah_attr) << 24)); + av->g_slid = rdma_ah_get_path_bits(ah_attr); + av->dlid = cpu_to_be16(rdma_ah_get_dlid(ah_attr)); av->msg_sr = (3 << 4) | /* 2K message */ - mthca_get_rate(dev, ah_attr->static_rate, ah_attr->port_num); - av->sl_tclass_flowlabel = cpu_to_be32(ah_attr->sl << 28); - if (ah_attr->ah_flags & IB_AH_GRH) { + mthca_get_rate(dev, rdma_ah_get_static_rate(ah_attr), + rdma_ah_get_port_num(ah_attr)); + av->sl_tclass_flowlabel = cpu_to_be32(rdma_ah_get_sl(ah_attr) << 28); + if (rdma_ah_get_ah_flags(ah_attr) & IB_AH_GRH) { + const struct ib_global_route *grh = rdma_ah_read_grh(ah_attr); + av->g_slid |= 0x80; - av->gid_index = (ah_attr->port_num - 1) * dev->limits.gid_table_len + - ah_attr->grh.sgid_index; - av->hop_limit = ah_attr->grh.hop_limit; + av->gid_index = (rdma_ah_get_port_num(ah_attr) - 1) * + dev->limits.gid_table_len + + grh->sgid_index; + av->hop_limit = grh->hop_limit; av->sl_tclass_flowlabel |= - cpu_to_be32((ah_attr->grh.traffic_class << 20) | - ah_attr->grh.flow_label); - memcpy(av->dgid, ah_attr->grh.dgid.raw, 16); + cpu_to_be32((grh->traffic_class << 20) | + grh->flow_label); + memcpy(av->dgid, grh->dgid.raw, 16); } else { /* Arbel workaround -- low byte of GID must be 2 */ av->dgid[3] = cpu_to_be32(2); @@ -287,33 +292,35 @@ int mthca_read_ah(struct mthca_dev *dev, struct mthca_ah *ah, return 0; } -int mthca_ah_query(struct ib_ah *ibah, struct ib_ah_attr *attr) +int mthca_ah_query(struct ib_ah *ibah, struct rdma_ah_attr *attr) { struct mthca_ah *ah = to_mah(ibah); struct mthca_dev *dev = to_mdev(ibah->device); + u8 port_num = be32_to_cpu(ah->av->port_pd) >> 24; /* Only implement for MAD and memfree ah for now. */ if (ah->type == MTHCA_AH_ON_HCA) return -ENOSYS; memset(attr, 0, sizeof *attr); - attr->dlid = be16_to_cpu(ah->av->dlid); - attr->sl = be32_to_cpu(ah->av->sl_tclass_flowlabel) >> 28; - attr->port_num = be32_to_cpu(ah->av->port_pd) >> 24; - attr->static_rate = mthca_rate_to_ib(dev, ah->av->msg_sr & 0x7, - attr->port_num); - attr->src_path_bits = ah->av->g_slid & 0x7F; - attr->ah_flags = mthca_ah_grh_present(ah) ? IB_AH_GRH : 0; - - if (attr->ah_flags) { - attr->grh.traffic_class = - be32_to_cpu(ah->av->sl_tclass_flowlabel) >> 20; - attr->grh.flow_label = - be32_to_cpu(ah->av->sl_tclass_flowlabel) & 0xfffff; - attr->grh.hop_limit = ah->av->hop_limit; - attr->grh.sgid_index = ah->av->gid_index & - (dev->limits.gid_table_len - 1); - memcpy(attr->grh.dgid.raw, ah->av->dgid, 16); + attr->type = ibah->type; + rdma_ah_set_dlid(attr, be16_to_cpu(ah->av->dlid)); + rdma_ah_set_sl(attr, be32_to_cpu(ah->av->sl_tclass_flowlabel) >> 28); + rdma_ah_set_port_num(attr, port_num); + rdma_ah_set_static_rate(attr, + mthca_rate_to_ib(dev, ah->av->msg_sr & 0x7, + port_num)); + rdma_ah_set_path_bits(attr, ah->av->g_slid & 0x7F); + if (mthca_ah_grh_present(ah)) { + u32 tc_fl = be32_to_cpu(ah->av->sl_tclass_flowlabel); + + rdma_ah_set_grh(attr, NULL, + tc_fl & 0xfffff, + ah->av->gid_index & + (dev->limits.gid_table_len - 1), + ah->av->hop_limit, + (tc_fl >> 20) & 0xff); + rdma_ah_set_dgid_raw(attr, ah->av->dgid); } return 0; diff --git a/drivers/infiniband/hw/mthca/mthca_cmd.c b/drivers/infiniband/hw/mthca/mthca_cmd.c index c7f49bbb0c72..9d83a53c0c67 100644 --- a/drivers/infiniband/hw/mthca/mthca_cmd.c +++ b/drivers/infiniband/hw/mthca/mthca_cmd.c @@ -367,12 +367,16 @@ static int mthca_cmd_poll(struct mthca_dev *dev, goto out; } - if (out_is_imm) + if (out_is_imm && out_param) { *out_param = (u64) be32_to_cpu((__force __be32) __raw_readl(dev->hcr + HCR_OUT_PARAM_OFFSET)) << 32 | (u64) be32_to_cpu((__force __be32) __raw_readl(dev->hcr + HCR_OUT_PARAM_OFFSET + 4)); + } else if (out_is_imm) { + err = -EINVAL; + goto out; + } status = be32_to_cpu((__force __be32) __raw_readl(dev->hcr + HCR_STATUS_OFFSET)) >> 24; if (status) { @@ -450,8 +454,12 @@ static int mthca_cmd_wait(struct mthca_dev *dev, err = mthca_status_to_errno(context->status); } - if (out_is_imm) + if (out_is_imm && out_param) { *out_param = context->out_param; + } else if (out_is_imm) { + err = -EINVAL; + goto out; + } out: spin_lock(&dev->cmd.context_lock); diff --git a/drivers/infiniband/hw/mthca/mthca_dev.h b/drivers/infiniband/hw/mthca/mthca_dev.h index 4393a022867b..ec7da9a474cd 100644 --- a/drivers/infiniband/hw/mthca/mthca_dev.h +++ b/drivers/infiniband/hw/mthca/mthca_dev.h @@ -560,12 +560,12 @@ int mthca_alloc_sqp(struct mthca_dev *dev, void mthca_free_qp(struct mthca_dev *dev, struct mthca_qp *qp); int mthca_create_ah(struct mthca_dev *dev, struct mthca_pd *pd, - struct ib_ah_attr *ah_attr, + struct rdma_ah_attr *ah_attr, struct mthca_ah *ah); int mthca_destroy_ah(struct mthca_dev *dev, struct mthca_ah *ah); int mthca_read_ah(struct mthca_dev *dev, struct mthca_ah *ah, struct ib_ud_header *header); -int mthca_ah_query(struct ib_ah *ibah, struct ib_ah_attr *attr); +int mthca_ah_query(struct ib_ah *ibah, struct rdma_ah_attr *attr); int mthca_ah_grh_present(struct mthca_ah *ah); u8 mthca_get_rate(struct mthca_dev *dev, int static_rate, u8 port); enum ib_rate mthca_rate_to_ib(struct mthca_dev *dev, u8 mthca_rate, u8 port); diff --git a/drivers/infiniband/hw/mthca/mthca_mad.c b/drivers/infiniband/hw/mthca/mthca_mad.c index 9139405c4810..7df3db71777a 100644 --- a/drivers/infiniband/hw/mthca/mthca_mad.c +++ b/drivers/infiniband/hw/mthca/mthca_mad.c @@ -75,25 +75,26 @@ static void update_sm_ah(struct mthca_dev *dev, u8 port_num, u16 lid, u8 sl) { struct ib_ah *new_ah; - struct ib_ah_attr ah_attr; + struct rdma_ah_attr ah_attr; unsigned long flags; if (!dev->send_agent[port_num - 1][0]) return; memset(&ah_attr, 0, sizeof ah_attr); - ah_attr.dlid = lid; - ah_attr.sl = sl; - ah_attr.port_num = port_num; + ah_attr.type = rdma_ah_find_type(&dev->ib_dev, port_num); + rdma_ah_set_dlid(&ah_attr, lid); + rdma_ah_set_sl(&ah_attr, sl); + rdma_ah_set_port_num(&ah_attr, port_num); - new_ah = ib_create_ah(dev->send_agent[port_num - 1][0]->qp->pd, - &ah_attr); + new_ah = rdma_create_ah(dev->send_agent[port_num - 1][0]->qp->pd, + &ah_attr); if (IS_ERR(new_ah)) return; spin_lock_irqsave(&dev->sm_lock, flags); if (dev->sm_ah[port_num - 1]) - ib_destroy_ah(dev->sm_ah[port_num - 1]); + rdma_destroy_ah(dev->sm_ah[port_num - 1]); dev->sm_ah[port_num - 1] = new_ah; spin_unlock_irqrestore(&dev->sm_lock, flags); } @@ -345,6 +346,6 @@ void mthca_free_agents(struct mthca_dev *dev) } if (dev->sm_ah[p]) - ib_destroy_ah(dev->sm_ah[p]); + rdma_destroy_ah(dev->sm_ah[p]); } } diff --git a/drivers/infiniband/hw/mthca/mthca_main.c b/drivers/infiniband/hw/mthca/mthca_main.c index ded76c101dde..c309e5c96383 100644 --- a/drivers/infiniband/hw/mthca/mthca_main.c +++ b/drivers/infiniband/hw/mthca/mthca_main.c @@ -851,20 +851,18 @@ err_uar_table_free: static int mthca_enable_msi_x(struct mthca_dev *mdev) { - struct msix_entry entries[3]; int err; - entries[0].entry = 0; - entries[1].entry = 1; - entries[2].entry = 2; - - err = pci_enable_msix_exact(mdev->pdev, entries, ARRAY_SIZE(entries)); - if (err) + err = pci_alloc_irq_vectors(mdev->pdev, 3, 3, PCI_IRQ_MSIX); + if (err < 0) return err; - mdev->eq_table.eq[MTHCA_EQ_COMP ].msi_x_vector = entries[0].vector; - mdev->eq_table.eq[MTHCA_EQ_ASYNC].msi_x_vector = entries[1].vector; - mdev->eq_table.eq[MTHCA_EQ_CMD ].msi_x_vector = entries[2].vector; + mdev->eq_table.eq[MTHCA_EQ_COMP ].msi_x_vector = + pci_irq_vector(mdev->pdev, 0); + mdev->eq_table.eq[MTHCA_EQ_ASYNC].msi_x_vector = + pci_irq_vector(mdev->pdev, 1); + mdev->eq_table.eq[MTHCA_EQ_CMD ].msi_x_vector = + pci_irq_vector(mdev->pdev, 2); return 0; } @@ -1018,7 +1016,7 @@ static int __mthca_init_one(struct pci_dev *pdev, int hca_type) err = mthca_setup_hca(mdev); if (err == -EBUSY && (mdev->mthca_flags & MTHCA_FLAG_MSI_X)) { if (mdev->mthca_flags & MTHCA_FLAG_MSI_X) - pci_disable_msix(pdev); + pci_free_irq_vectors(pdev); mdev->mthca_flags &= ~MTHCA_FLAG_MSI_X; err = mthca_setup_hca(mdev); @@ -1062,7 +1060,7 @@ err_cleanup: err_close: if (mdev->mthca_flags & MTHCA_FLAG_MSI_X) - pci_disable_msix(pdev); + pci_free_irq_vectors(pdev); mthca_close_hca(mdev); @@ -1113,7 +1111,7 @@ static void __mthca_remove_one(struct pci_dev *pdev) mthca_cmd_cleanup(mdev); if (mdev->mthca_flags & MTHCA_FLAG_MSI_X) - pci_disable_msix(pdev); + pci_free_irq_vectors(pdev); ib_dealloc_device(&mdev->ib_dev); pci_release_regions(pdev); diff --git a/drivers/infiniband/hw/mthca/mthca_provider.c b/drivers/infiniband/hw/mthca/mthca_provider.c index d31708742ba5..c197cd9b193f 100644 --- a/drivers/infiniband/hw/mthca/mthca_provider.c +++ b/drivers/infiniband/hw/mthca/mthca_provider.c @@ -146,7 +146,7 @@ static int mthca_query_port(struct ib_device *ibdev, if (!in_mad || !out_mad) goto out; - memset(props, 0, sizeof *props); + /* props being zeroed by the caller, avoid zeroing it here */ init_query_mad(in_mad); in_mad->attr_id = IB_SMP_ATTR_PORT_INFO; @@ -212,7 +212,7 @@ static int mthca_modify_port(struct ib_device *ibdev, if (mutex_lock_interruptible(&to_mdev(ibdev)->cap_mask_mutex)) return -ERESTARTSYS; - err = mthca_query_port(ibdev, port, &attr); + err = ib_query_port(ibdev, port, &attr); if (err) goto out; @@ -410,7 +410,7 @@ static int mthca_dealloc_pd(struct ib_pd *pd) } static struct ib_ah *mthca_ah_create(struct ib_pd *pd, - struct ib_ah_attr *ah_attr, + struct rdma_ah_attr *ah_attr, struct ib_udata *udata) { @@ -937,7 +937,7 @@ static struct ib_mr *mthca_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, goto err; } - shift = ffs(mr->umem->page_size) - 1; + shift = mr->umem->page_shift; n = mr->umem->nmap; mr->mtt = mthca_alloc_mtt(dev, n); @@ -959,8 +959,7 @@ static struct ib_mr *mthca_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, for_each_sg(mr->umem->sg_head.sgl, sg, mr->umem->nmap, entry) { len = sg_dma_len(sg) >> shift; for (k = 0; k < len; ++k) { - pages[i++] = sg_dma_address(sg) + - mr->umem->page_size * k; + pages[i++] = sg_dma_address(sg) + (k << shift); /* * Be friendly to write_mtt and pass it chunks * of appropriate size. @@ -1166,13 +1165,14 @@ static int mthca_port_immutable(struct ib_device *ibdev, u8 port_num, struct ib_port_attr attr; int err; - err = mthca_query_port(ibdev, port_num, &attr); + immutable->core_cap_flags = RDMA_CORE_PORT_IBA_IB; + + err = ib_query_port(ibdev, port_num, &attr); if (err) return err; immutable->pkey_tbl_len = attr.pkey_tbl_len; immutable->gid_tbl_len = attr.gid_tbl_len; - immutable->core_cap_flags = RDMA_CORE_PORT_IBA_IB; immutable->max_mad_size = IB_MGMT_MAD_SIZE; return 0; @@ -1223,7 +1223,7 @@ int mthca_register_device(struct mthca_dev *dev) dev->ib_dev.node_type = RDMA_NODE_IB_CA; dev->ib_dev.phys_port_cnt = dev->limits.num_ports; dev->ib_dev.num_comp_vectors = 1; - dev->ib_dev.dma_device = &dev->pdev->dev; + dev->ib_dev.dev.parent = &dev->pdev->dev; dev->ib_dev.query_device = mthca_query_device; dev->ib_dev.query_port = mthca_query_port; dev->ib_dev.modify_device = mthca_modify_device; diff --git a/drivers/infiniband/hw/mthca/mthca_qp.c b/drivers/infiniband/hw/mthca/mthca_qp.c index 96e5fb91fb48..d21960cd9a49 100644 --- a/drivers/infiniband/hw/mthca/mthca_qp.c +++ b/drivers/infiniband/hw/mthca/mthca_qp.c @@ -393,31 +393,36 @@ static int to_ib_qp_access_flags(int mthca_flags) return ib_flags; } -static void to_ib_ah_attr(struct mthca_dev *dev, struct ib_ah_attr *ib_ah_attr, - struct mthca_qp_path *path) +static void to_rdma_ah_attr(struct mthca_dev *dev, + struct rdma_ah_attr *ah_attr, + struct mthca_qp_path *path) { - memset(ib_ah_attr, 0, sizeof *ib_ah_attr); - ib_ah_attr->port_num = (be32_to_cpu(path->port_pkey) >> 24) & 0x3; + u8 port_num = (be32_to_cpu(path->port_pkey) >> 24) & 0x3; - if (ib_ah_attr->port_num == 0 || ib_ah_attr->port_num > dev->limits.num_ports) - return; + memset(ah_attr, 0, sizeof(*ah_attr)); - ib_ah_attr->dlid = be16_to_cpu(path->rlid); - ib_ah_attr->sl = be32_to_cpu(path->sl_tclass_flowlabel) >> 28; - ib_ah_attr->src_path_bits = path->g_mylmc & 0x7f; - ib_ah_attr->static_rate = mthca_rate_to_ib(dev, - path->static_rate & 0xf, - ib_ah_attr->port_num); - ib_ah_attr->ah_flags = (path->g_mylmc & (1 << 7)) ? IB_AH_GRH : 0; - if (ib_ah_attr->ah_flags) { - ib_ah_attr->grh.sgid_index = path->mgid_index & (dev->limits.gid_table_len - 1); - ib_ah_attr->grh.hop_limit = path->hop_limit; - ib_ah_attr->grh.traffic_class = - (be32_to_cpu(path->sl_tclass_flowlabel) >> 20) & 0xff; - ib_ah_attr->grh.flow_label = - be32_to_cpu(path->sl_tclass_flowlabel) & 0xfffff; - memcpy(ib_ah_attr->grh.dgid.raw, - path->rgid, sizeof ib_ah_attr->grh.dgid.raw); + if (port_num == 0 || port_num > dev->limits.num_ports) + return; + ah_attr->type = rdma_ah_find_type(&dev->ib_dev, port_num); + rdma_ah_set_port_num(ah_attr, port_num); + + rdma_ah_set_dlid(ah_attr, be16_to_cpu(path->rlid)); + rdma_ah_set_sl(ah_attr, be32_to_cpu(path->sl_tclass_flowlabel) >> 28); + rdma_ah_set_path_bits(ah_attr, path->g_mylmc & 0x7f); + rdma_ah_set_static_rate(ah_attr, + mthca_rate_to_ib(dev, + path->static_rate & 0xf, + port_num)); + if (path->g_mylmc & (1 << 7)) { + u32 tc_fl = be32_to_cpu(path->sl_tclass_flowlabel); + + rdma_ah_set_grh(ah_attr, NULL, + tc_fl & 0xfffff, + path->mgid_index & + (dev->limits.gid_table_len - 1), + path->hop_limit, + (tc_fl >> 20) & 0xff); + rdma_ah_set_dgid_raw(ah_attr, path->rgid); } } @@ -468,11 +473,12 @@ int mthca_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr, int qp_attr_m to_ib_qp_access_flags(be32_to_cpu(context->params2)); if (qp->transport == RC || qp->transport == UC) { - to_ib_ah_attr(dev, &qp_attr->ah_attr, &context->pri_path); - to_ib_ah_attr(dev, &qp_attr->alt_ah_attr, &context->alt_path); + to_rdma_ah_attr(dev, &qp_attr->ah_attr, &context->pri_path); + to_rdma_ah_attr(dev, &qp_attr->alt_ah_attr, &context->alt_path); qp_attr->alt_pkey_index = be32_to_cpu(context->alt_path.port_pkey) & 0x7f; - qp_attr->alt_port_num = qp_attr->alt_ah_attr.port_num; + qp_attr->alt_port_num = + rdma_ah_get_port_num(&qp_attr->alt_ah_attr); } qp_attr->pkey_index = be32_to_cpu(context->pri_path.port_pkey) & 0x7f; @@ -512,30 +518,36 @@ out: return err; } -static int mthca_path_set(struct mthca_dev *dev, const struct ib_ah_attr *ah, +static int mthca_path_set(struct mthca_dev *dev, const struct rdma_ah_attr *ah, struct mthca_qp_path *path, u8 port) { - path->g_mylmc = ah->src_path_bits & 0x7f; - path->rlid = cpu_to_be16(ah->dlid); - path->static_rate = mthca_get_rate(dev, ah->static_rate, port); + path->g_mylmc = rdma_ah_get_path_bits(ah) & 0x7f; + path->rlid = cpu_to_be16(rdma_ah_get_dlid(ah)); + path->static_rate = mthca_get_rate(dev, rdma_ah_get_static_rate(ah), + port); + + if (rdma_ah_get_ah_flags(ah) & IB_AH_GRH) { + const struct ib_global_route *grh = rdma_ah_read_grh(ah); - if (ah->ah_flags & IB_AH_GRH) { - if (ah->grh.sgid_index >= dev->limits.gid_table_len) { + if (grh->sgid_index >= dev->limits.gid_table_len) { mthca_dbg(dev, "sgid_index (%u) too large. max is %d\n", - ah->grh.sgid_index, dev->limits.gid_table_len-1); + grh->sgid_index, + dev->limits.gid_table_len - 1); return -1; } path->g_mylmc |= 1 << 7; - path->mgid_index = ah->grh.sgid_index; - path->hop_limit = ah->grh.hop_limit; + path->mgid_index = grh->sgid_index; + path->hop_limit = grh->hop_limit; path->sl_tclass_flowlabel = - cpu_to_be32((ah->sl << 28) | - (ah->grh.traffic_class << 20) | - (ah->grh.flow_label)); - memcpy(path->rgid, ah->grh.dgid.raw, 16); - } else - path->sl_tclass_flowlabel = cpu_to_be32(ah->sl << 28); + cpu_to_be32((rdma_ah_get_sl(ah) << 28) | + (grh->traffic_class << 20) | + (grh->flow_label)); + memcpy(path->rgid, grh->dgid.raw, 16); + } else { + path->sl_tclass_flowlabel = cpu_to_be32(rdma_ah_get_sl(ah) << + 28); + } return 0; } @@ -680,7 +692,7 @@ static int __mthca_modify_qp(struct ib_qp *ibqp, } if (mthca_path_set(dev, &attr->alt_ah_attr, &qp_context->alt_path, - attr->alt_ah_attr.port_num)) + rdma_ah_get_port_num(&attr->alt_ah_attr))) goto out_mailbox; qp_context->alt_path.port_pkey |= cpu_to_be32(attr->alt_pkey_index | diff --git a/drivers/infiniband/hw/nes/nes.c b/drivers/infiniband/hw/nes/nes.c index 5b9601014f0c..a30aa6527f7e 100644 --- a/drivers/infiniband/hw/nes/nes.c +++ b/drivers/infiniband/hw/nes/nes.c @@ -815,7 +815,7 @@ static struct pci_driver nes_pci_driver = { .remove = nes_remove, }; -static ssize_t nes_show_adapter(struct device_driver *ddp, char *buf) +static ssize_t adapter_show(struct device_driver *ddp, char *buf) { unsigned int devfn = 0xffffffff; unsigned char bus_number = 0xff; @@ -834,7 +834,7 @@ static ssize_t nes_show_adapter(struct device_driver *ddp, char *buf) return snprintf(buf, PAGE_SIZE, "%x:%x\n", bus_number, devfn); } -static ssize_t nes_store_adapter(struct device_driver *ddp, +static ssize_t adapter_store(struct device_driver *ddp, const char *buf, size_t count) { char *p = (char *)buf; @@ -843,7 +843,7 @@ static ssize_t nes_store_adapter(struct device_driver *ddp, return strnlen(buf, count); } -static ssize_t nes_show_ee_cmd(struct device_driver *ddp, char *buf) +static ssize_t eeprom_cmd_show(struct device_driver *ddp, char *buf) { u32 eeprom_cmd = 0xdead; u32 i = 0; @@ -859,7 +859,7 @@ static ssize_t nes_show_ee_cmd(struct device_driver *ddp, char *buf) return snprintf(buf, PAGE_SIZE, "0x%x\n", eeprom_cmd); } -static ssize_t nes_store_ee_cmd(struct device_driver *ddp, +static ssize_t eeprom_cmd_store(struct device_driver *ddp, const char *buf, size_t count) { char *p = (char *)buf; @@ -880,7 +880,7 @@ static ssize_t nes_store_ee_cmd(struct device_driver *ddp, return strnlen(buf, count); } -static ssize_t nes_show_ee_data(struct device_driver *ddp, char *buf) +static ssize_t eeprom_data_show(struct device_driver *ddp, char *buf) { u32 eeprom_data = 0xdead; u32 i = 0; @@ -897,7 +897,7 @@ static ssize_t nes_show_ee_data(struct device_driver *ddp, char *buf) return snprintf(buf, PAGE_SIZE, "0x%x\n", eeprom_data); } -static ssize_t nes_store_ee_data(struct device_driver *ddp, +static ssize_t eeprom_data_store(struct device_driver *ddp, const char *buf, size_t count) { char *p = (char *)buf; @@ -918,7 +918,7 @@ static ssize_t nes_store_ee_data(struct device_driver *ddp, return strnlen(buf, count); } -static ssize_t nes_show_flash_cmd(struct device_driver *ddp, char *buf) +static ssize_t flash_cmd_show(struct device_driver *ddp, char *buf) { u32 flash_cmd = 0xdead; u32 i = 0; @@ -935,7 +935,7 @@ static ssize_t nes_show_flash_cmd(struct device_driver *ddp, char *buf) return snprintf(buf, PAGE_SIZE, "0x%x\n", flash_cmd); } -static ssize_t nes_store_flash_cmd(struct device_driver *ddp, +static ssize_t flash_cmd_store(struct device_driver *ddp, const char *buf, size_t count) { char *p = (char *)buf; @@ -956,7 +956,7 @@ static ssize_t nes_store_flash_cmd(struct device_driver *ddp, return strnlen(buf, count); } -static ssize_t nes_show_flash_data(struct device_driver *ddp, char *buf) +static ssize_t flash_data_show(struct device_driver *ddp, char *buf) { u32 flash_data = 0xdead; u32 i = 0; @@ -973,7 +973,7 @@ static ssize_t nes_show_flash_data(struct device_driver *ddp, char *buf) return snprintf(buf, PAGE_SIZE, "0x%x\n", flash_data); } -static ssize_t nes_store_flash_data(struct device_driver *ddp, +static ssize_t flash_data_store(struct device_driver *ddp, const char *buf, size_t count) { char *p = (char *)buf; @@ -994,12 +994,12 @@ static ssize_t nes_store_flash_data(struct device_driver *ddp, return strnlen(buf, count); } -static ssize_t nes_show_nonidx_addr(struct device_driver *ddp, char *buf) +static ssize_t nonidx_addr_show(struct device_driver *ddp, char *buf) { return snprintf(buf, PAGE_SIZE, "0x%x\n", sysfs_nonidx_addr); } -static ssize_t nes_store_nonidx_addr(struct device_driver *ddp, +static ssize_t nonidx_addr_store(struct device_driver *ddp, const char *buf, size_t count) { char *p = (char *)buf; @@ -1010,7 +1010,7 @@ static ssize_t nes_store_nonidx_addr(struct device_driver *ddp, return strnlen(buf, count); } -static ssize_t nes_show_nonidx_data(struct device_driver *ddp, char *buf) +static ssize_t nonidx_data_show(struct device_driver *ddp, char *buf) { u32 nonidx_data = 0xdead; u32 i = 0; @@ -1027,7 +1027,7 @@ static ssize_t nes_show_nonidx_data(struct device_driver *ddp, char *buf) return snprintf(buf, PAGE_SIZE, "0x%x\n", nonidx_data); } -static ssize_t nes_store_nonidx_data(struct device_driver *ddp, +static ssize_t nonidx_data_store(struct device_driver *ddp, const char *buf, size_t count) { char *p = (char *)buf; @@ -1048,12 +1048,12 @@ static ssize_t nes_store_nonidx_data(struct device_driver *ddp, return strnlen(buf, count); } -static ssize_t nes_show_idx_addr(struct device_driver *ddp, char *buf) +static ssize_t idx_addr_show(struct device_driver *ddp, char *buf) { return snprintf(buf, PAGE_SIZE, "0x%x\n", sysfs_idx_addr); } -static ssize_t nes_store_idx_addr(struct device_driver *ddp, +static ssize_t idx_addr_store(struct device_driver *ddp, const char *buf, size_t count) { char *p = (char *)buf; @@ -1064,7 +1064,7 @@ static ssize_t nes_store_idx_addr(struct device_driver *ddp, return strnlen(buf, count); } -static ssize_t nes_show_idx_data(struct device_driver *ddp, char *buf) +static ssize_t idx_data_show(struct device_driver *ddp, char *buf) { u32 idx_data = 0xdead; u32 i = 0; @@ -1081,7 +1081,7 @@ static ssize_t nes_show_idx_data(struct device_driver *ddp, char *buf) return snprintf(buf, PAGE_SIZE, "0x%x\n", idx_data); } -static ssize_t nes_store_idx_data(struct device_driver *ddp, +static ssize_t idx_data_store(struct device_driver *ddp, const char *buf, size_t count) { char *p = (char *)buf; @@ -1102,11 +1102,7 @@ static ssize_t nes_store_idx_data(struct device_driver *ddp, return strnlen(buf, count); } - -/** - * nes_show_wqm_quanta - */ -static ssize_t nes_show_wqm_quanta(struct device_driver *ddp, char *buf) +static ssize_t wqm_quanta_show(struct device_driver *ddp, char *buf) { u32 wqm_quanta_value = 0xdead; u32 i = 0; @@ -1123,12 +1119,8 @@ static ssize_t nes_show_wqm_quanta(struct device_driver *ddp, char *buf) return snprintf(buf, PAGE_SIZE, "0x%X\n", wqm_quanta_value); } - -/** - * nes_store_wqm_quanta - */ -static ssize_t nes_store_wqm_quanta(struct device_driver *ddp, - const char *buf, size_t count) +static ssize_t wqm_quanta_store(struct device_driver *ddp, const char *buf, + size_t count) { unsigned long wqm_quanta_value; u32 wqm_config1; @@ -1153,26 +1145,16 @@ static ssize_t nes_store_wqm_quanta(struct device_driver *ddp, return strnlen(buf, count); } -static DRIVER_ATTR(adapter, S_IRUSR | S_IWUSR, - nes_show_adapter, nes_store_adapter); -static DRIVER_ATTR(eeprom_cmd, S_IRUSR | S_IWUSR, - nes_show_ee_cmd, nes_store_ee_cmd); -static DRIVER_ATTR(eeprom_data, S_IRUSR | S_IWUSR, - nes_show_ee_data, nes_store_ee_data); -static DRIVER_ATTR(flash_cmd, S_IRUSR | S_IWUSR, - nes_show_flash_cmd, nes_store_flash_cmd); -static DRIVER_ATTR(flash_data, S_IRUSR | S_IWUSR, - nes_show_flash_data, nes_store_flash_data); -static DRIVER_ATTR(nonidx_addr, S_IRUSR | S_IWUSR, - nes_show_nonidx_addr, nes_store_nonidx_addr); -static DRIVER_ATTR(nonidx_data, S_IRUSR | S_IWUSR, - nes_show_nonidx_data, nes_store_nonidx_data); -static DRIVER_ATTR(idx_addr, S_IRUSR | S_IWUSR, - nes_show_idx_addr, nes_store_idx_addr); -static DRIVER_ATTR(idx_data, S_IRUSR | S_IWUSR, - nes_show_idx_data, nes_store_idx_data); -static DRIVER_ATTR(wqm_quanta, S_IRUSR | S_IWUSR, - nes_show_wqm_quanta, nes_store_wqm_quanta); +static DRIVER_ATTR_RW(adapter); +static DRIVER_ATTR_RW(eeprom_cmd); +static DRIVER_ATTR_RW(eeprom_data); +static DRIVER_ATTR_RW(flash_cmd); +static DRIVER_ATTR_RW(flash_data); +static DRIVER_ATTR_RW(nonidx_addr); +static DRIVER_ATTR_RW(nonidx_data); +static DRIVER_ATTR_RW(idx_addr); +static DRIVER_ATTR_RW(idx_data); +static DRIVER_ATTR_RW(wqm_quanta); static int nes_create_driver_sysfs(struct pci_driver *drv) { diff --git a/drivers/infiniband/hw/nes/nes.h b/drivers/infiniband/hw/nes/nes.h index 85acd0843b50..3f9e56e8b379 100644 --- a/drivers/infiniband/hw/nes/nes.h +++ b/drivers/infiniband/hw/nes/nes.h @@ -36,6 +36,7 @@ #include <linux/netdevice.h> #include <linux/inetdevice.h> +#include <linux/interrupt.h> #include <linux/spinlock.h> #include <linux/kernel.h> #include <linux/delay.h> diff --git a/drivers/infiniband/hw/nes/nes_cm.c b/drivers/infiniband/hw/nes/nes_cm.c index 8e703479e7ce..de4025deaa4a 100644 --- a/drivers/infiniband/hw/nes/nes_cm.c +++ b/drivers/infiniband/hw/nes/nes_cm.c @@ -135,17 +135,17 @@ static void record_ird_ord(struct nes_cm_node *, u16, u16); /* instance of function pointers for client API */ /* set address of this instance to cm_core->cm_ops at cm_core alloc */ static const struct nes_cm_ops nes_cm_api = { - mini_cm_accelerated, - mini_cm_listen, - mini_cm_del_listen, - mini_cm_connect, - mini_cm_close, - mini_cm_accept, - mini_cm_reject, - mini_cm_recv_pkt, - mini_cm_dealloc_core, - mini_cm_get, - mini_cm_set + .accelerated = mini_cm_accelerated, + .listen = mini_cm_listen, + .stop_listener = mini_cm_del_listen, + .connect = mini_cm_connect, + .close = mini_cm_close, + .accept = mini_cm_accept, + .reject = mini_cm_reject, + .recv_pkt = mini_cm_recv_pkt, + .destroy_cm_core = mini_cm_dealloc_core, + .get = mini_cm_get, + .set = mini_cm_set }; static struct nes_cm_core *g_cm_core; @@ -610,7 +610,6 @@ static void build_mpa_v2(struct nes_cm_node *cm_node, ctrl_ord = cm_node->ord_size & IETF_NO_IRD_ORD; } ctrl_ird |= IETF_PEER_TO_PEER; - ctrl_ird |= IETF_FLPDU_ZERO_LEN; switch (mpa_key) { case MPA_KEY_REQUEST: @@ -743,7 +742,7 @@ int schedule_nes_timer(struct nes_cm_node *cm_node, struct sk_buff *skb, if (type == NES_TIMER_TYPE_SEND) { new_send->seq_num = ntohl(tcp_hdr(skb)->seq); - atomic_inc(&new_send->skb->users); + refcount_inc(&new_send->skb->users); spin_lock_irqsave(&cm_node->retrans_list_lock, flags); cm_node->send_entry = new_send; add_ref_cm_node(cm_node); @@ -925,7 +924,7 @@ static void nes_cm_timer_tick(unsigned long pass) flags); break; } - atomic_inc(&send_entry->skb->users); + refcount_inc(&send_entry->skb->users); cm_packets_retrans++; nes_debug(NES_DBG_CM, "Retransmitting send_entry %p " "for node %p, jiffies = %lu, time to send = " @@ -1826,7 +1825,7 @@ static void handle_rcv_mpa(struct nes_cm_node *cm_node, struct sk_buff *skb) type = NES_CM_EVENT_CONNECTED; cm_node->state = NES_CM_STATE_TSA; } - + send_ack(cm_node, NULL); break; default: WARN_ON(1); diff --git a/drivers/infiniband/hw/nes/nes_hw.c b/drivers/infiniband/hw/nes/nes_hw.c index 19acd13c6cb1..b0adf65e4bdb 100644 --- a/drivers/infiniband/hw/nes/nes_hw.c +++ b/drivers/infiniband/hw/nes/nes_hw.c @@ -551,7 +551,7 @@ struct nes_adapter *nes_init_adapter(struct nes_device *nesdev, u8 hw_rev) { if ((0x0F000100 == (pcs_control_status0 & 0x0F000100)) || (0x0F000100 == (pcs_control_status1 & 0x0F000100))) int_cnt++; - msleep(1); + usleep_range(1000, 2000); } if (int_cnt > 1) { spin_lock_irqsave(&nesadapter->phy_lock, flags); @@ -592,7 +592,7 @@ struct nes_adapter *nes_init_adapter(struct nes_device *nesdev, u8 hw_rev) { break; } } - msleep(1); + usleep_range(1000, 2000); } } } @@ -1849,9 +1849,8 @@ int nes_init_nic_qp(struct nes_device *nesdev, struct net_device *netdev) wqe_count -= counter; nes_write32(nesdev->regs+NES_WQE_ALLOC, (counter << 24) | nesvnic->nic.qp_id); } while (wqe_count); - init_timer(&nesvnic->rq_wqes_timer); - nesvnic->rq_wqes_timer.function = nes_rq_wqes_timeout; - nesvnic->rq_wqes_timer.data = (unsigned long)nesvnic; + setup_timer(&nesvnic->rq_wqes_timer, nes_rq_wqes_timeout, + (unsigned long)nesvnic); nes_debug(NES_DBG_INIT, "NAPI support Enabled\n"); if (nesdev->nesadapter->et_use_adaptive_rx_coalesce) { @@ -3055,7 +3054,7 @@ static void nes_cqp_ce_handler(struct nes_device *nesdev, struct nes_hw_cq *cq) memcpy(cqp_wqe, &cqp_request->cqp_wqe, sizeof(*cqp_wqe)); barrier(); - opcode = cqp_wqe->wqe_words[NES_CQP_WQE_OPCODE_IDX]; + opcode = le32_to_cpu(cqp_wqe->wqe_words[NES_CQP_WQE_OPCODE_IDX]); if ((opcode & NES_CQP_OPCODE_MASK) == NES_CQP_DOWNLOAD_SEGMENT) ctx_index = NES_CQP_WQE_DL_COMP_CTX_LOW_IDX; else diff --git a/drivers/infiniband/hw/nes/nes_mgt.c b/drivers/infiniband/hw/nes/nes_mgt.c index 33624f17c347..77226cf4ea02 100644 --- a/drivers/infiniband/hw/nes/nes_mgt.c +++ b/drivers/infiniband/hw/nes/nes_mgt.c @@ -1040,9 +1040,8 @@ int nes_init_mgt_qp(struct nes_device *nesdev, struct net_device *netdev, struct mgtvnic->mgt.rx_skb[counter] = skb; } - init_timer(&mgtvnic->rq_wqes_timer); - mgtvnic->rq_wqes_timer.function = nes_mgt_rq_wqes_timeout; - mgtvnic->rq_wqes_timer.data = (unsigned long)mgtvnic; + setup_timer(&mgtvnic->rq_wqes_timer, nes_mgt_rq_wqes_timeout, + (unsigned long)mgtvnic); wqe_count = NES_MGT_WQ_COUNT - 1; mgtvnic->mgt.rq_head = wqe_count; diff --git a/drivers/infiniband/hw/nes/nes_verbs.c b/drivers/infiniband/hw/nes/nes_verbs.c index 5a31f3c6a421..25dcd7573df9 100644 --- a/drivers/infiniband/hw/nes/nes_verbs.c +++ b/drivers/infiniband/hw/nes/nes_verbs.c @@ -475,7 +475,7 @@ static int nes_query_port(struct ib_device *ibdev, u8 port, struct ib_port_attr struct nes_vnic *nesvnic = to_nesvnic(ibdev); struct net_device *netdev = nesvnic->netdev; - memset(props, 0, sizeof(*props)); + /* props being zeroed by the caller, avoid zeroing it here */ props->max_mtu = IB_MTU_4096; props->active_mtu = ib_mtu_int_to_enum(netdev->mtu); @@ -761,7 +761,8 @@ static int nes_dealloc_pd(struct ib_pd *ibpd) /** * nes_create_ah */ -static struct ib_ah *nes_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr, +static struct ib_ah *nes_create_ah(struct ib_pd *pd, + struct rdma_ah_attr *ah_attr, struct ib_udata *udata) { return ERR_PTR(-ENOSYS); @@ -1308,9 +1309,8 @@ static struct ib_qp *nes_create_qp(struct ib_pd *ibpd, init_completion(&nesqp->rq_drained); nesqp->sig_all = (init_attr->sq_sig_type == IB_SIGNAL_ALL_WR); - init_timer(&nesqp->terminate_timer); - nesqp->terminate_timer.function = nes_terminate_timeout; - nesqp->terminate_timer.data = (unsigned long)nesqp; + setup_timer(&nesqp->terminate_timer, nes_terminate_timeout, + (unsigned long)nesqp); /* update the QP table */ nesdev->nesadapter->qp_table[nesqp->hwqp.qp_id-NES_FIRST_QPN] = nesqp; @@ -2165,9 +2165,9 @@ static struct ib_mr *nes_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, } nes_debug(NES_DBG_MR, "User base = 0x%lX, Virt base = 0x%lX, length = %u," - " offset = %u, page size = %u.\n", + " offset = %u, page size = %lu.\n", (unsigned long int)start, (unsigned long int)virt, (u32)length, - ib_umem_offset(region), region->page_size); + ib_umem_offset(region), BIT(region->page_shift)); skip_pages = ((u32)ib_umem_offset(region)) >> 12; @@ -3660,13 +3660,14 @@ static int nes_port_immutable(struct ib_device *ibdev, u8 port_num, struct ib_port_attr attr; int err; + immutable->core_cap_flags = RDMA_CORE_PORT_IWARP; + err = nes_query_port(ibdev, port_num, &attr); if (err) return err; immutable->pkey_tbl_len = attr.pkey_tbl_len; immutable->gid_tbl_len = attr.gid_tbl_len; - immutable->core_cap_flags = RDMA_CORE_PORT_IWARP; return 0; } @@ -3730,7 +3731,6 @@ struct nes_ib_device *nes_init_ofa_device(struct net_device *netdev) nesibdev->ibdev.phys_port_cnt = 1; nesibdev->ibdev.num_comp_vectors = 1; - nesibdev->ibdev.dma_device = &nesdev->pcidev->dev; nesibdev->ibdev.dev.parent = &nesdev->pcidev->dev; nesibdev->ibdev.query_device = nes_query_device; nesibdev->ibdev.query_port = nes_query_port; diff --git a/drivers/infiniband/hw/ocrdma/ocrdma.h b/drivers/infiniband/hw/ocrdma/ocrdma.h index 45bdfa0e3b2b..7baedc74e39d 100644 --- a/drivers/infiniband/hw/ocrdma/ocrdma.h +++ b/drivers/infiniband/hw/ocrdma/ocrdma.h @@ -527,17 +527,17 @@ static inline int is_cqe_wr_imm(struct ocrdma_cqe *cqe) } static inline int ocrdma_resolve_dmac(struct ocrdma_dev *dev, - struct ib_ah_attr *ah_attr, u8 *mac_addr) + struct rdma_ah_attr *ah_attr, u8 *mac_addr) { struct in6_addr in6; - memcpy(&in6, ah_attr->grh.dgid.raw, sizeof(in6)); + memcpy(&in6, rdma_ah_read_grh(ah_attr)->dgid.raw, sizeof(in6)); if (rdma_is_multicast_addr(&in6)) rdma_get_mcast_mac(&in6, mac_addr); else if (rdma_link_local_addr(&in6)) rdma_get_ll_mac(&in6, mac_addr); else - memcpy(mac_addr, ah_attr->dmac, ETH_ALEN); + memcpy(mac_addr, ah_attr->roce.dmac, ETH_ALEN); return 0; } diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_ah.c b/drivers/infiniband/hw/ocrdma/ocrdma_ah.c index 14d33b0f3950..d0249e463338 100644 --- a/drivers/infiniband/hw/ocrdma/ocrdma_ah.c +++ b/drivers/infiniband/hw/ocrdma/ocrdma_ah.c @@ -59,7 +59,7 @@ static u16 ocrdma_hdr_type_to_proto_num(int devid, u8 hdr_type) { switch (hdr_type) { case OCRDMA_L3_TYPE_IB_GRH: - return (u16)0x8915; + return (u16)ETH_P_IBOE; case OCRDMA_L3_TYPE_IPV4: return (u16)0x0800; case OCRDMA_L3_TYPE_IPV6: @@ -71,7 +71,7 @@ static u16 ocrdma_hdr_type_to_proto_num(int devid, u8 hdr_type) } static inline int set_av_attr(struct ocrdma_dev *dev, struct ocrdma_ah *ah, - struct ib_ah_attr *attr, union ib_gid *sgid, + struct rdma_ah_attr *attr, union ib_gid *sgid, int pdid, bool *isvlan, u16 vlan_tag) { int status; @@ -81,6 +81,7 @@ static inline int set_av_attr(struct ocrdma_dev *dev, struct ocrdma_ah *ah, u16 proto_num = 0; u8 nxthdr = 0x11; struct iphdr ipv4; + const struct ib_global_route *ib_grh; union { struct sockaddr _sockaddr; struct sockaddr_in _sockaddr_in; @@ -94,7 +95,7 @@ static inline int set_av_attr(struct ocrdma_dev *dev, struct ocrdma_ah *ah, proto_num = ocrdma_hdr_type_to_proto_num(dev->id, ah->hdr_type); if (!proto_num) return -EINVAL; - nxthdr = (proto_num == 0x8915) ? 0x1b : 0x11; + nxthdr = (proto_num == ETH_P_IBOE) ? 0x1b : 0x11; /* VLAN */ if (!vlan_tag || (vlan_tag > 0xFFF)) vlan_tag = dev->pvid; @@ -120,32 +121,33 @@ static inline int set_av_attr(struct ocrdma_dev *dev, struct ocrdma_ah *ah, status = ocrdma_resolve_dmac(dev, attr, ð.dmac[0]); if (status) return status; - ah->sgid_index = attr->grh.sgid_index; + ib_grh = rdma_ah_read_grh(attr); + ah->sgid_index = ib_grh->sgid_index; /* Eth HDR */ memcpy(&ah->av->eth_hdr, ð, eth_sz); if (ah->hdr_type == RDMA_NETWORK_IPV4) { *((__be16 *)&ipv4) = htons((4 << 12) | (5 << 8) | - attr->grh.traffic_class); + ib_grh->traffic_class); ipv4.id = cpu_to_be16(pdid); ipv4.frag_off = htons(IP_DF); ipv4.tot_len = htons(0); - ipv4.ttl = attr->grh.hop_limit; + ipv4.ttl = ib_grh->hop_limit; ipv4.protocol = nxthdr; rdma_gid2ip(&sgid_addr._sockaddr, sgid); ipv4.saddr = sgid_addr._sockaddr_in.sin_addr.s_addr; - rdma_gid2ip(&dgid_addr._sockaddr, &attr->grh.dgid); + rdma_gid2ip(&dgid_addr._sockaddr, &ib_grh->dgid); ipv4.daddr = dgid_addr._sockaddr_in.sin_addr.s_addr; memcpy((u8 *)ah->av + eth_sz, &ipv4, sizeof(struct iphdr)); } else { memcpy(&grh.sgid[0], sgid->raw, sizeof(union ib_gid)); grh.tclass_flow = cpu_to_be32((6 << 28) | - (attr->grh.traffic_class << 24) | - attr->grh.flow_label); - memcpy(&grh.dgid[0], attr->grh.dgid.raw, - sizeof(attr->grh.dgid.raw)); + (ib_grh->traffic_class << 24) | + ib_grh->flow_label); + memcpy(&grh.dgid[0], ib_grh->dgid.raw, + sizeof(ib_grh->dgid.raw)); grh.pdid_hoplimit = cpu_to_be32((pdid << 16) | (nxthdr << 8) | - attr->grh.hop_limit); + ib_grh->hop_limit); memcpy((u8 *)ah->av + eth_sz, &grh, sizeof(struct ocrdma_grh)); } if (*isvlan) @@ -154,7 +156,7 @@ static inline int set_av_attr(struct ocrdma_dev *dev, struct ocrdma_ah *ah, return status; } -struct ib_ah *ocrdma_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr, +struct ib_ah *ocrdma_create_ah(struct ib_pd *ibpd, struct rdma_ah_attr *attr, struct ib_udata *udata) { u32 *ahid_addr; @@ -165,11 +167,14 @@ struct ib_ah *ocrdma_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr, struct ib_gid_attr sgid_attr; struct ocrdma_pd *pd = get_ocrdma_pd(ibpd); struct ocrdma_dev *dev = get_ocrdma_dev(ibpd->device); + const struct ib_global_route *grh; union ib_gid sgid; - if (!(attr->ah_flags & IB_AH_GRH)) + if ((attr->type != RDMA_AH_ATTR_TYPE_ROCE) || + !(rdma_ah_get_ah_flags(attr) & IB_AH_GRH)) return ERR_PTR(-EINVAL); + grh = rdma_ah_read_grh(attr); if (atomic_cmpxchg(&dev->update_sl, 1, 0)) ocrdma_init_service_level(dev); @@ -181,7 +186,7 @@ struct ib_ah *ocrdma_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr, if (status) goto av_err; - status = ib_get_cached_gid(&dev->ibdev, 1, attr->grh.sgid_index, &sgid, + status = ib_get_cached_gid(&dev->ibdev, 1, grh->sgid_index, &sgid, &sgid_attr); if (status) { pr_err("%s(): Failed to query sgid, status = %d\n", @@ -197,10 +202,11 @@ struct ib_ah *ocrdma_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr, ah->hdr_type = ib_gid_to_network_type(sgid_attr.gid_type, &sgid); if ((pd->uctx) && - (!rdma_is_multicast_addr((struct in6_addr *)attr->grh.dgid.raw)) && - (!rdma_link_local_addr((struct in6_addr *)attr->grh.dgid.raw))) { - status = rdma_addr_find_l2_eth_by_grh(&sgid, &attr->grh.dgid, - attr->dmac, &vlan_tag, + (!rdma_is_multicast_addr((struct in6_addr *)grh->dgid.raw)) && + (!rdma_link_local_addr((struct in6_addr *)grh->dgid.raw))) { + status = rdma_addr_find_l2_eth_by_grh(&sgid, &grh->dgid, + attr->roce.dmac, + &vlan_tag, &sgid_attr.ndev->ifindex, NULL); if (status) { @@ -216,7 +222,7 @@ struct ib_ah *ocrdma_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr, /* if pd is for the user process, pass the ah_id to user space */ if ((pd->uctx) && (pd->uctx->ah_tbl.va)) { - ahid_addr = pd->uctx->ah_tbl.va + attr->dlid; + ahid_addr = pd->uctx->ah_tbl.va + rdma_ah_get_dlid(attr); *ahid_addr = 0; *ahid_addr |= ah->id & OCRDMA_AH_ID_MASK; if (ocrdma_is_udp_encap_supported(dev)) { @@ -248,30 +254,32 @@ int ocrdma_destroy_ah(struct ib_ah *ibah) return 0; } -int ocrdma_query_ah(struct ib_ah *ibah, struct ib_ah_attr *attr) +int ocrdma_query_ah(struct ib_ah *ibah, struct rdma_ah_attr *attr) { struct ocrdma_ah *ah = get_ocrdma_ah(ibah); struct ocrdma_av *av = ah->av; struct ocrdma_grh *grh; - attr->ah_flags |= IB_AH_GRH; + + attr->type = ibah->type; if (ah->av->valid & OCRDMA_AV_VALID) { grh = (struct ocrdma_grh *)((u8 *)ah->av + sizeof(struct ocrdma_eth_vlan)); - attr->sl = be16_to_cpu(av->eth_hdr.vlan_tag) >> 13; + rdma_ah_set_sl(attr, be16_to_cpu(av->eth_hdr.vlan_tag) >> 13); } else { grh = (struct ocrdma_grh *)((u8 *)ah->av + sizeof(struct ocrdma_eth_basic)); - attr->sl = 0; + rdma_ah_set_sl(attr, 0); } - memcpy(&attr->grh.dgid.raw[0], &grh->dgid[0], sizeof(grh->dgid)); - attr->grh.sgid_index = ah->sgid_index; - attr->grh.hop_limit = be32_to_cpu(grh->pdid_hoplimit) & 0xff; - attr->grh.traffic_class = be32_to_cpu(grh->tclass_flow) >> 24; - attr->grh.flow_label = be32_to_cpu(grh->tclass_flow) & 0x00ffffffff; + rdma_ah_set_grh(attr, NULL, + be32_to_cpu(grh->tclass_flow) & 0xffffffff, + ah->sgid_index, + be32_to_cpu(grh->pdid_hoplimit) & 0xff, + be32_to_cpu(grh->tclass_flow) >> 24); + rdma_ah_set_dgid_raw(attr, &grh->dgid[0]); return 0; } -int ocrdma_modify_ah(struct ib_ah *ibah, struct ib_ah_attr *attr) +int ocrdma_modify_ah(struct ib_ah *ibah, struct rdma_ah_attr *attr) { /* modify_ah is unsupported */ return -ENOSYS; diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_ah.h b/drivers/infiniband/hw/ocrdma/ocrdma_ah.h index 0704a24b17c8..1a65c47945aa 100644 --- a/drivers/infiniband/hw/ocrdma/ocrdma_ah.h +++ b/drivers/infiniband/hw/ocrdma/ocrdma_ah.h @@ -51,11 +51,11 @@ enum { OCRDMA_AH_L3_TYPE_SHIFT = 0x1D /* 29 bits */ }; -struct ib_ah *ocrdma_create_ah(struct ib_pd *, struct ib_ah_attr *, - struct ib_udata *); -int ocrdma_destroy_ah(struct ib_ah *); -int ocrdma_query_ah(struct ib_ah *, struct ib_ah_attr *); -int ocrdma_modify_ah(struct ib_ah *, struct ib_ah_attr *); +struct ib_ah *ocrdma_create_ah(struct ib_pd *pd, struct rdma_ah_attr *ah_attr, + struct ib_udata *udata); +int ocrdma_destroy_ah(struct ib_ah *ah); +int ocrdma_query_ah(struct ib_ah *ah, struct rdma_ah_attr *ah_attr); +int ocrdma_modify_ah(struct ib_ah *ah, struct rdma_ah_attr *ah_attr); int ocrdma_process_mad(struct ib_device *, int process_mad_flags, diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_hw.c b/drivers/infiniband/hw/ocrdma/ocrdma_hw.c index 9a305201545e..dcb5942f9fb5 100644 --- a/drivers/infiniband/hw/ocrdma/ocrdma_hw.c +++ b/drivers/infiniband/hw/ocrdma/ocrdma_hw.c @@ -44,6 +44,7 @@ #include <linux/interrupt.h> #include <linux/log2.h> #include <linux/dma-mapping.h> +#include <linux/if_ether.h> #include <rdma/ib_verbs.h> #include <rdma/ib_user_verbs.h> @@ -2498,7 +2499,7 @@ static int ocrdma_set_av_params(struct ocrdma_qp *qp, int attr_mask) { int status; - struct ib_ah_attr *ah_attr = &attrs->ah_attr; + struct rdma_ah_attr *ah_attr = &attrs->ah_attr; union ib_gid sgid, zgid; struct ib_gid_attr sgid_attr; u32 vlan_id = 0xFFFF; @@ -2509,25 +2510,28 @@ static int ocrdma_set_av_params(struct ocrdma_qp *qp, struct sockaddr_in6 _sockaddr_in6; } sgid_addr, dgid_addr; struct ocrdma_dev *dev = get_ocrdma_dev(qp->ibqp.device); + const struct ib_global_route *grh; - if ((ah_attr->ah_flags & IB_AH_GRH) == 0) + if ((rdma_ah_get_ah_flags(ah_attr) & IB_AH_GRH) == 0) return -EINVAL; + grh = rdma_ah_read_grh(ah_attr); if (atomic_cmpxchg(&dev->update_sl, 1, 0)) ocrdma_init_service_level(dev); cmd->params.tclass_sq_psn |= - (ah_attr->grh.traffic_class << OCRDMA_QP_PARAMS_TCLASS_SHIFT); + (grh->traffic_class << OCRDMA_QP_PARAMS_TCLASS_SHIFT); cmd->params.rnt_rc_sl_fl |= - (ah_attr->grh.flow_label & OCRDMA_QP_PARAMS_FLOW_LABEL_MASK); - cmd->params.rnt_rc_sl_fl |= (ah_attr->sl << OCRDMA_QP_PARAMS_SL_SHIFT); + (grh->flow_label & OCRDMA_QP_PARAMS_FLOW_LABEL_MASK); + cmd->params.rnt_rc_sl_fl |= (rdma_ah_get_sl(ah_attr) << + OCRDMA_QP_PARAMS_SL_SHIFT); cmd->params.hop_lmt_rq_psn |= - (ah_attr->grh.hop_limit << OCRDMA_QP_PARAMS_HOP_LMT_SHIFT); + (grh->hop_limit << OCRDMA_QP_PARAMS_HOP_LMT_SHIFT); cmd->flags |= OCRDMA_QP_PARA_FLOW_LBL_VALID; /* GIDs */ - memcpy(&cmd->params.dgid[0], &ah_attr->grh.dgid.raw[0], + memcpy(&cmd->params.dgid[0], &grh->dgid.raw[0], sizeof(cmd->params.dgid)); - status = ib_get_cached_gid(&dev->ibdev, 1, ah_attr->grh.sgid_index, + status = ib_get_cached_gid(&dev->ibdev, 1, grh->sgid_index, &sgid, &sgid_attr); if (!status && sgid_attr.ndev) { vlan_id = rdma_vlan_dev_vlan_id(sgid_attr.ndev); @@ -2539,7 +2543,7 @@ static int ocrdma_set_av_params(struct ocrdma_qp *qp, if (!memcmp(&sgid, &zgid, sizeof(zgid))) return -EINVAL; - qp->sgid_idx = ah_attr->grh.sgid_index; + qp->sgid_idx = grh->sgid_index; memcpy(&cmd->params.sgid[0], &sgid.raw[0], sizeof(cmd->params.sgid)); status = ocrdma_resolve_dmac(dev, ah_attr, &mac_addr[0]); if (status) @@ -2550,7 +2554,7 @@ static int ocrdma_set_av_params(struct ocrdma_qp *qp, hdr_type = ib_gid_to_network_type(sgid_attr.gid_type, &sgid); if (hdr_type == RDMA_NETWORK_IPV4) { rdma_gid2ip(&sgid_addr._sockaddr, &sgid); - rdma_gid2ip(&dgid_addr._sockaddr, &ah_attr->grh.dgid); + rdma_gid2ip(&dgid_addr._sockaddr, &grh->dgid); memcpy(&cmd->params.dgid[0], &dgid_addr._sockaddr_in.sin_addr.s_addr, 4); memcpy(&cmd->params.sgid[0], @@ -2984,7 +2988,7 @@ static int ocrdma_parse_dcbxcfg_rsp(struct ocrdma_dev *dev, int ptype, OCRDMA_APP_PARAM_APP_PROTO_MASK; if ( - valid && proto == OCRDMA_APP_PROTO_ROCE && + valid && proto == ETH_P_IBOE && proto_sel == OCRDMA_PROTO_SELECT_L2) { for (slindx = 0; slindx < OCRDMA_MAX_SERVICE_LEVEL_INDEX; slindx++) { diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_main.c b/drivers/infiniband/hw/ocrdma/ocrdma_main.c index 896071502739..57c9a2ad0260 100644 --- a/drivers/infiniband/hw/ocrdma/ocrdma_main.c +++ b/drivers/infiniband/hw/ocrdma/ocrdma_main.c @@ -93,15 +93,16 @@ static int ocrdma_port_immutable(struct ib_device *ibdev, u8 port_num, int err; dev = get_ocrdma_dev(ibdev); - err = ocrdma_query_port(ibdev, port_num, &attr); + immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE; + if (ocrdma_is_udp_encap_supported(dev)) + immutable->core_cap_flags |= RDMA_CORE_CAP_PROT_ROCE_UDP_ENCAP; + + err = ib_query_port(ibdev, port_num, &attr); if (err) return err; immutable->pkey_tbl_len = attr.pkey_tbl_len; immutable->gid_tbl_len = attr.gid_tbl_len; - immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE; - if (ocrdma_is_udp_encap_supported(dev)) - immutable->core_cap_flags |= RDMA_CORE_CAP_PROT_ROCE_UDP_ENCAP; immutable->max_mad_size = IB_MGMT_MAD_SIZE; return 0; @@ -198,7 +199,7 @@ static int ocrdma_register_device(struct ocrdma_dev *dev) dev->ibdev.alloc_ucontext = ocrdma_alloc_ucontext; dev->ibdev.dealloc_ucontext = ocrdma_dealloc_ucontext; dev->ibdev.mmap = ocrdma_mmap; - dev->ibdev.dma_device = &dev->nic_info.pdev->dev; + dev->ibdev.dev.parent = &dev->nic_info.pdev->dev; dev->ibdev.process_mad = ocrdma_process_mad; dev->ibdev.get_port_immutable = ocrdma_port_immutable; diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_sli.h b/drivers/infiniband/hw/ocrdma/ocrdma_sli.h index 37df4481bb8f..6ef89c226ad8 100644 --- a/drivers/infiniband/hw/ocrdma/ocrdma_sli.h +++ b/drivers/infiniband/hw/ocrdma/ocrdma_sli.h @@ -1901,7 +1901,6 @@ struct ocrdma_eth_vlan { u8 smac[6]; __be16 eth_type; __be16 vlan_tag; -#define OCRDMA_ROCE_ETH_TYPE 0x8915 __be16 roce_eth_type; } __packed; @@ -2179,10 +2178,6 @@ enum OCRDMA_DCBX_PARAM_TYPE { OCRDMA_PARAMETER_TYPE_PEER = 0x02 }; -enum OCRDMA_DCBX_APP_PROTO { - OCRDMA_APP_PROTO_ROCE = 0x8915 -}; - enum OCRDMA_DCBX_PROTO { OCRDMA_PROTO_SELECT_L2 = 0x00, OCRDMA_PROTO_SELECT_L4 = 0x01 diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_stats.c b/drivers/infiniband/hw/ocrdma/ocrdma_stats.c index f8e4b0a6486f..66056f9a9700 100644 --- a/drivers/infiniband/hw/ocrdma/ocrdma_stats.c +++ b/drivers/infiniband/hw/ocrdma/ocrdma_stats.c @@ -643,7 +643,7 @@ static ssize_t ocrdma_dbgfs_ops_write(struct file *filp, struct ocrdma_stats *pstats = filp->private_data; struct ocrdma_dev *dev = pstats->dev; - if (count > 32) + if (*ppos != 0 || count == 0 || count > sizeof(tmp_str)) goto err; if (copy_from_user(tmp_str, buffer, count)) diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c b/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c index 6af44f8db3d5..27d5e8d9f08d 100644 --- a/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c +++ b/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c @@ -210,6 +210,7 @@ int ocrdma_query_port(struct ib_device *ibdev, struct ocrdma_dev *dev; struct net_device *netdev; + /* props being zeroed by the caller, avoid zeroing it here */ dev = get_ocrdma_dev(ibdev); if (port > 1) { pr_err("%s(%d) invalid_port=0x%x\n", __func__, @@ -371,7 +372,7 @@ static int _ocrdma_pd_mgr_put_bitmap(struct ocrdma_dev *dev, u16 pd_id, return 0; } -static u8 ocrdma_put_pd_num(struct ocrdma_dev *dev, u16 pd_id, +static int ocrdma_put_pd_num(struct ocrdma_dev *dev, u16 pd_id, bool dpp_pool) { int status; @@ -743,7 +744,8 @@ err: if (is_uctx_pd) { ocrdma_release_ucontext_pd(uctx); } else { - status = _ocrdma_dealloc_pd(dev, pd); + if (_ocrdma_dealloc_pd(dev, pd)) + pr_err("%s: _ocrdma_dealloc_pd() failed\n", __func__); } exit: return ERR_PTR(status); @@ -913,21 +915,18 @@ static void build_user_pbes(struct ocrdma_dev *dev, struct ocrdma_mr *mr, pbe = (struct ocrdma_pbe *)pbl_tbl->va; pbe_cnt = 0; - shift = ilog2(umem->page_size); + shift = umem->page_shift; for_each_sg(umem->sg_head.sgl, sg, umem->nmap, entry) { pages = sg_dma_len(sg) >> shift; for (pg_cnt = 0; pg_cnt < pages; pg_cnt++) { /* store the page address in pbe */ pbe->pa_lo = - cpu_to_le32(sg_dma_address - (sg) + - (umem->page_size * pg_cnt)); + cpu_to_le32(sg_dma_address(sg) + + (pg_cnt << shift)); pbe->pa_hi = - cpu_to_le32(upper_32_bits - ((sg_dma_address - (sg) + - umem->page_size * pg_cnt))); + cpu_to_le32(upper_32_bits(sg_dma_address(sg) + + (pg_cnt << shift))); pbe_cnt += 1; total_num_pbes += 1; pbe++; @@ -977,7 +976,7 @@ struct ib_mr *ocrdma_reg_user_mr(struct ib_pd *ibpd, u64 start, u64 len, if (status) goto umem_err; - mr->hwmr.pbe_size = mr->umem->page_size; + mr->hwmr.pbe_size = BIT(mr->umem->page_shift); mr->hwmr.fbo = ib_umem_offset(mr->umem); mr->hwmr.va = usr_addr; mr->hwmr.len = len; @@ -1170,8 +1169,7 @@ int ocrdma_destroy_cq(struct ib_cq *ibcq) dev->cq_tbl[cq->id] = NULL; indx = ocrdma_get_eq_table_index(dev, cq->eqn); - if (indx == -EINVAL) - BUG(); + BUG_ON(indx == -EINVAL); eq = &dev->eq_tbl[indx]; irq = ocrdma_get_irq(dev, eq); @@ -1601,23 +1599,24 @@ int ocrdma_query_qp(struct ib_qp *ibqp, qp_attr->cap.max_recv_sge = qp->rq.max_sges; qp_attr->cap.max_inline_data = qp->max_inline_data; qp_init_attr->cap = qp_attr->cap; - memcpy(&qp_attr->ah_attr.grh.dgid, ¶ms.dgid[0], - sizeof(params.dgid)); - qp_attr->ah_attr.grh.flow_label = params.rnt_rc_sl_fl & - OCRDMA_QP_PARAMS_FLOW_LABEL_MASK; - qp_attr->ah_attr.grh.sgid_index = qp->sgid_idx; - qp_attr->ah_attr.grh.hop_limit = (params.hop_lmt_rq_psn & - OCRDMA_QP_PARAMS_HOP_LMT_MASK) >> - OCRDMA_QP_PARAMS_HOP_LMT_SHIFT; - qp_attr->ah_attr.grh.traffic_class = (params.tclass_sq_psn & - OCRDMA_QP_PARAMS_TCLASS_MASK) >> - OCRDMA_QP_PARAMS_TCLASS_SHIFT; - - qp_attr->ah_attr.ah_flags = IB_AH_GRH; - qp_attr->ah_attr.port_num = 1; - qp_attr->ah_attr.sl = (params.rnt_rc_sl_fl & - OCRDMA_QP_PARAMS_SL_MASK) >> - OCRDMA_QP_PARAMS_SL_SHIFT; + qp_attr->ah_attr.type = RDMA_AH_ATTR_TYPE_ROCE; + + rdma_ah_set_grh(&qp_attr->ah_attr, NULL, + params.rnt_rc_sl_fl & + OCRDMA_QP_PARAMS_FLOW_LABEL_MASK, + qp->sgid_idx, + (params.hop_lmt_rq_psn & + OCRDMA_QP_PARAMS_HOP_LMT_MASK) >> + OCRDMA_QP_PARAMS_HOP_LMT_SHIFT, + (params.tclass_sq_psn & + OCRDMA_QP_PARAMS_TCLASS_MASK) >> + OCRDMA_QP_PARAMS_TCLASS_SHIFT); + rdma_ah_set_dgid_raw(&qp_attr->ah_attr, ¶ms.dgid[0]); + + rdma_ah_set_port_num(&qp_attr->ah_attr, 1); + rdma_ah_set_sl(&qp_attr->ah_attr, (params.rnt_rc_sl_fl & + OCRDMA_QP_PARAMS_SL_MASK) >> + OCRDMA_QP_PARAMS_SL_SHIFT); qp_attr->timeout = (params.ack_to_rnr_rtc_dest_qpn & OCRDMA_QP_PARAMS_ACK_TIMEOUT_MASK) >> OCRDMA_QP_PARAMS_ACK_TIMEOUT_SHIFT; @@ -1630,8 +1629,8 @@ int ocrdma_query_qp(struct ib_qp *ibqp, qp_attr->min_rnr_timer = 0; qp_attr->pkey_index = 0; qp_attr->port_num = 1; - qp_attr->ah_attr.src_path_bits = 0; - qp_attr->ah_attr.static_rate = 0; + rdma_ah_set_path_bits(&qp_attr->ah_attr, 0); + rdma_ah_set_static_rate(&qp_attr->ah_attr, 0); qp_attr->alt_pkey_index = 0; qp_attr->alt_port_num = 0; qp_attr->alt_timeout = 0; @@ -1741,8 +1740,7 @@ static void ocrdma_discard_cqes(struct ocrdma_qp *qp, struct ocrdma_cq *cq) wqe_idx = (le32_to_cpu(cqe->rq.buftag_qpn) >> OCRDMA_CQE_BUFTAG_SHIFT) & qp->srq->rq.max_wqe_idx; - if (wqe_idx < 1) - BUG(); + BUG_ON(wqe_idx < 1); spin_lock_irqsave(&qp->srq->q_lock, flags); ocrdma_hwq_inc_tail(&qp->srq->rq); ocrdma_srq_toggle_bit(qp->srq, wqe_idx - 1); @@ -1904,6 +1902,7 @@ struct ib_srq *ocrdma_create_srq(struct ib_pd *ibpd, goto err; if (udata == NULL) { + status = -ENOMEM; srq->rqe_wr_id_tbl = kzalloc(sizeof(u64) * srq->rq.max_cnt, GFP_KERNEL); if (srq->rqe_wr_id_tbl == NULL) @@ -2388,15 +2387,13 @@ static int ocrdma_srq_get_idx(struct ocrdma_srq *srq) if (srq->idx_bit_fields[row]) { indx = ffs(srq->idx_bit_fields[row]); indx = (row * 32) + (indx - 1); - if (indx >= srq->rq.max_cnt) - BUG(); + BUG_ON(indx >= srq->rq.max_cnt); ocrdma_srq_toggle_bit(srq, indx); break; } } - if (row == srq->bit_fields_len) - BUG(); + BUG_ON(row == srq->bit_fields_len); return indx + 1; /* Use from index 1 */ } @@ -2754,8 +2751,7 @@ static void ocrdma_update_free_srq_cqe(struct ib_wc *ibwc, srq = get_ocrdma_srq(qp->ibqp.srq); wqe_idx = (le32_to_cpu(cqe->rq.buftag_qpn) >> OCRDMA_CQE_BUFTAG_SHIFT) & srq->rq.max_wqe_idx; - if (wqe_idx < 1) - BUG(); + BUG_ON(wqe_idx < 1); ibwc->wr_id = srq->rqe_wr_id_tbl[wqe_idx]; spin_lock_irqsave(&srq->q_lock, flags); diff --git a/drivers/infiniband/hw/qedr/main.c b/drivers/infiniband/hw/qedr/main.c index 3ac8aa5ef37d..b5851fd67d4f 100644 --- a/drivers/infiniband/hw/qedr/main.c +++ b/drivers/infiniband/hw/qedr/main.c @@ -35,8 +35,9 @@ #include <rdma/ib_user_verbs.h> #include <linux/netdevice.h> #include <linux/iommu.h> +#include <linux/pci.h> #include <net/addrconf.h> -#include <linux/qed/qede_roce.h> + #include <linux/qed/qed_chain.h> #include <linux/qed/qed_if.h> #include "qedr.h" @@ -170,7 +171,7 @@ static int qedr_register_device(struct qedr_dev *dev) dev->ibdev.get_port_immutable = qedr_port_immutable; dev->ibdev.get_netdev = qedr_get_netdev; - dev->ibdev.dma_device = &dev->pdev->dev; + dev->ibdev.dev.parent = &dev->pdev->dev; dev->ibdev.get_link_layer = qedr_link_layer; dev->ibdev.get_dev_fw_str = qedr_get_dev_fw_str; @@ -275,7 +276,7 @@ static int qedr_alloc_resources(struct qedr_dev *dev) QED_CHAIN_CNT_TYPE_U16, n_entries, sizeof(struct regpair *), - &cnq->pbl); + &cnq->pbl, NULL); if (rc) goto err4; @@ -340,43 +341,58 @@ static void qedr_remove_sysfiles(struct qedr_dev *dev) static void qedr_pci_set_atomic(struct qedr_dev *dev, struct pci_dev *pdev) { struct pci_dev *bridge; - u32 val; - - dev->atomic_cap = IB_ATOMIC_NONE; + u32 ctl2, cap2; + u16 flags; + int rc; bridge = pdev->bus->self; if (!bridge) - return; - - /* Check whether we are connected directly or via a switch */ - while (bridge && bridge->bus->parent) { - DP_DEBUG(dev, QEDR_MSG_INIT, - "Device is not connected directly to root. bridge->bus->number=%d primary=%d\n", - bridge->bus->number, bridge->bus->primary); - /* Need to check Atomic Op Routing Supported all the way to - * root complex. - */ - pcie_capability_read_dword(bridge, PCI_EXP_DEVCAP2, &val); - if (!(val & PCI_EXP_DEVCAP2_ATOMIC_ROUTE)) { - pcie_capability_clear_word(pdev, - PCI_EXP_DEVCTL2, - PCI_EXP_DEVCTL2_ATOMIC_REQ); - return; - } + goto disable; + + /* Check atomic routing support all the way to root complex */ + while (bridge->bus->parent) { + rc = pcie_capability_read_word(bridge, PCI_EXP_FLAGS, &flags); + if (rc || ((flags & PCI_EXP_FLAGS_VERS) < 2)) + goto disable; + + rc = pcie_capability_read_dword(bridge, PCI_EXP_DEVCAP2, &cap2); + if (rc) + goto disable; + + rc = pcie_capability_read_dword(bridge, PCI_EXP_DEVCTL2, &ctl2); + if (rc) + goto disable; + + if (!(cap2 & PCI_EXP_DEVCAP2_ATOMIC_ROUTE) || + (ctl2 & PCI_EXP_DEVCTL2_ATOMIC_EGRESS_BLOCK)) + goto disable; bridge = bridge->bus->parent->self; } - bridge = pdev->bus->self; - /* according to bridge capability */ - pcie_capability_read_dword(bridge, PCI_EXP_DEVCAP2, &val); - if (val & PCI_EXP_DEVCAP2_ATOMIC_COMP64) { - pcie_capability_set_word(pdev, PCI_EXP_DEVCTL2, - PCI_EXP_DEVCTL2_ATOMIC_REQ); - dev->atomic_cap = IB_ATOMIC_GLOB; - } else { - pcie_capability_clear_word(pdev, PCI_EXP_DEVCTL2, - PCI_EXP_DEVCTL2_ATOMIC_REQ); - } + rc = pcie_capability_read_word(bridge, PCI_EXP_FLAGS, &flags); + if (rc || ((flags & PCI_EXP_FLAGS_VERS) < 2)) + goto disable; + + rc = pcie_capability_read_dword(bridge, PCI_EXP_DEVCAP2, &cap2); + if (rc || !(cap2 & PCI_EXP_DEVCAP2_ATOMIC_COMP64)) + goto disable; + + /* Set atomic operations */ + pcie_capability_set_word(pdev, PCI_EXP_DEVCTL2, + PCI_EXP_DEVCTL2_ATOMIC_REQ); + dev->atomic_cap = IB_ATOMIC_GLOB; + + DP_DEBUG(dev, QEDR_MSG_INIT, "Atomic capability enabled\n"); + + return; + +disable: + pcie_capability_clear_word(pdev, PCI_EXP_DEVCTL2, + PCI_EXP_DEVCTL2_ATOMIC_REQ); + dev->atomic_cap = IB_ATOMIC_NONE; + + DP_DEBUG(dev, QEDR_MSG_INIT, "Atomic capability disabled\n"); + } static const struct qed_rdma_ops *qed_ops; @@ -423,14 +439,21 @@ static irqreturn_t qedr_irq_handler(int irq, void *handle) cq->arm_flags = 0; - if (cq->ibcq.comp_handler) + if (!cq->destroyed && cq->ibcq.comp_handler) (*cq->ibcq.comp_handler) (&cq->ibcq, cq->ibcq.cq_context); + /* The CQ's CNQ notification counter is checked before + * destroying the CQ in a busy-wait loop that waits for all of + * the CQ's CNQ interrupts to be processed. It is increased + * here, only after the completion handler, to ensure that the + * the handler is not running when the CQ is destroyed. + */ + cq->cnq_notif++; + sw_comp_cons = qed_chain_get_cons_idx(&cnq->pbl); cnq->n_comp++; - } qed_ops->rdma_cnq_prod_update(cnq->dev->rdma_ctx, cnq->index, @@ -587,9 +610,8 @@ void qedr_affiliated_event(void *context, u8 e_code, void *fw_handle) #define EVENT_TYPE_CQ 1 #define EVENT_TYPE_QP 2 struct qedr_dev *dev = (struct qedr_dev *)context; - union event_ring_data *data = fw_handle; - u64 roce_handle64 = ((u64)data->roce_handle.hi << 32) + - data->roce_handle.lo; + struct regpair *async_handle = (struct regpair *)fw_handle; + u64 roce_handle64 = ((u64) async_handle->hi << 32) + async_handle->lo; u8 event_type = EVENT_TYPE_NOT_DEFINED; struct ib_event event; struct ib_cq *ibcq; @@ -864,9 +886,9 @@ static void qedr_mac_address_change(struct qedr_dev *dev) memcpy(&sgid->raw[8], guid, sizeof(guid)); /* Update LL2 */ - rc = dev->ops->roce_ll2_set_mac_filter(dev->cdev, - dev->gsi_ll2_mac_address, - dev->ndev->dev_addr); + rc = dev->ops->ll2_set_mac_filter(dev->cdev, + dev->gsi_ll2_mac_address, + dev->ndev->dev_addr); ether_addr_copy(dev->gsi_ll2_mac_address, dev->ndev->dev_addr); @@ -880,7 +902,7 @@ static void qedr_mac_address_change(struct qedr_dev *dev) * initialization done before RoCE driver notifies * event to stack. */ -static void qedr_notify(struct qedr_dev *dev, enum qede_roce_event event) +static void qedr_notify(struct qedr_dev *dev, enum qede_rdma_event event) { switch (event) { case QEDE_UP: @@ -909,12 +931,12 @@ static struct qedr_driver qedr_drv = { static int __init qedr_init_module(void) { - return qede_roce_register_driver(&qedr_drv); + return qede_rdma_register_driver(&qedr_drv); } static void __exit qedr_exit_module(void) { - qede_roce_unregister_driver(&qedr_drv); + qede_rdma_unregister_driver(&qedr_drv); } module_init(qedr_init_module); diff --git a/drivers/infiniband/hw/qedr/qedr.h b/drivers/infiniband/hw/qedr/qedr.h index bb32e4792ec9..ab7784bfdac6 100644 --- a/drivers/infiniband/hw/qedr/qedr.h +++ b/drivers/infiniband/hw/qedr/qedr.h @@ -36,9 +36,10 @@ #include <rdma/ib_addr.h> #include <linux/qed/qed_if.h> #include <linux/qed/qed_chain.h> -#include <linux/qed/qed_roce_if.h> -#include <linux/qed/qede_roce.h> -#include "qedr_hsi.h" +#include <linux/qed/qed_rdma_if.h> +#include <linux/qed/qede_rdma.h> +#include <linux/qed/roce_common.h> +#include "qedr_hsi_rdma.h" #define QEDR_MODULE_VERSION "8.10.10.0" #define QEDR_NODE_DESC "QLogic 579xx RoCE HCA" @@ -57,7 +58,10 @@ #define QEDR_MSG_QP " QP" #define QEDR_MSG_GSI " GSI" -#define QEDR_CQ_MAGIC_NUMBER (0x11223344) +#define QEDR_CQ_MAGIC_NUMBER (0x11223344) + +#define FW_PAGE_SIZE (RDMA_RING_PAGE_SIZE) +#define FW_PAGE_SHIFT (12) struct qedr_dev; @@ -149,6 +153,8 @@ struct qedr_dev { u32 dp_module; u8 dp_level; u8 num_hwfns; + u8 gsi_ll2_handle; + uint wq_multiplier; u8 gsi_ll2_mac_address[ETH_ALEN]; int gsi_qp_created; @@ -271,6 +277,8 @@ struct qedr_cq { u32 cq_cons; struct qedr_userq q; + u8 destroyed; + u16 cnq_notif; }; struct qedr_pd { @@ -389,7 +397,7 @@ struct qedr_qp { struct qedr_ah { struct ib_ah ibah; - struct ib_ah_attr attr; + struct rdma_ah_attr attr; }; enum qedr_mr_type { @@ -428,7 +436,8 @@ struct qedr_mr { RDMA_CQE_RESPONDER_IMM_FLG_SHIFT) #define QEDR_RESP_RDMA (RDMA_CQE_RESPONDER_RDMA_FLG_MASK << \ RDMA_CQE_RESPONDER_RDMA_FLG_SHIFT) -#define QEDR_RESP_RDMA_IMM (QEDR_RESP_IMM | QEDR_RESP_RDMA) +#define QEDR_RESP_INV (RDMA_CQE_RESPONDER_INV_FLG_MASK << \ + RDMA_CQE_RESPONDER_INV_FLG_SHIFT) static inline void qedr_inc_sw_cons(struct qedr_qp_hwq_info *info) { @@ -442,19 +451,24 @@ static inline void qedr_inc_sw_prod(struct qedr_qp_hwq_info *info) } static inline int qedr_get_dmac(struct qedr_dev *dev, - struct ib_ah_attr *ah_attr, u8 *mac_addr) + struct rdma_ah_attr *ah_attr, u8 *mac_addr) { union ib_gid zero_sgid = { { 0 } }; struct in6_addr in6; + const struct ib_global_route *grh = rdma_ah_read_grh(ah_attr); + u8 *dmac; - if (!memcmp(&ah_attr->grh.dgid, &zero_sgid, sizeof(union ib_gid))) { + if (!memcmp(&grh->dgid, &zero_sgid, sizeof(union ib_gid))) { DP_ERR(dev, "Local port GID not supported\n"); eth_zero_addr(mac_addr); return -EINVAL; } - memcpy(&in6, ah_attr->grh.dgid.raw, sizeof(in6)); - ether_addr_copy(mac_addr, ah_attr->dmac); + memcpy(&in6, grh->dgid.raw, sizeof(in6)); + dmac = rdma_ah_retrieve_dmac(ah_attr); + if (!dmac) + return -EINVAL; + ether_addr_copy(mac_addr, dmac); return 0; } diff --git a/drivers/infiniband/hw/qedr/qedr_cm.c b/drivers/infiniband/hw/qedr/qedr_cm.c index a9a8d8745d2e..4689e802b332 100644 --- a/drivers/infiniband/hw/qedr/qedr_cm.c +++ b/drivers/infiniband/hw/qedr/qedr_cm.c @@ -43,14 +43,11 @@ #include <rdma/ib_addr.h> #include <rdma/ib_cache.h> -#include "qedr_hsi.h" #include <linux/qed/qed_if.h> -#include <linux/qed/qed_roce_if.h> +#include <linux/qed/qed_rdma_if.h> #include "qedr.h" -#include "qedr_hsi.h" #include "verbs.h" #include <rdma/qedr-abi.h> -#include "qedr_hsi.h" #include "qedr_cm.h" void qedr_inc_sw_gsi_cons(struct qedr_qp_hwq_info *info) @@ -67,9 +64,14 @@ void qedr_store_gsi_qp_cq(struct qedr_dev *dev, struct qedr_qp *qp, dev->gsi_qp = qp; } -void qedr_ll2_tx_cb(void *_qdev, struct qed_roce_ll2_packet *pkt) +void qedr_ll2_complete_tx_packet(void *cxt, + u8 connection_handle, + void *cookie, + dma_addr_t first_frag_addr, + bool b_last_fragment, bool b_last_packet) { - struct qedr_dev *dev = (struct qedr_dev *)_qdev; + struct qedr_dev *dev = (struct qedr_dev *)cxt; + struct qed_roce_ll2_packet *pkt = cookie; struct qedr_cq *cq = dev->gsi_sqcq; struct qedr_qp *qp = dev->gsi_qp; unsigned long flags; @@ -91,20 +93,26 @@ void qedr_ll2_tx_cb(void *_qdev, struct qed_roce_ll2_packet *pkt) (*cq->ibcq.comp_handler) (&cq->ibcq, cq->ibcq.cq_context); } -void qedr_ll2_rx_cb(void *_dev, struct qed_roce_ll2_packet *pkt, - struct qed_roce_ll2_rx_params *params) +void qedr_ll2_complete_rx_packet(void *cxt, + struct qed_ll2_comp_rx_data *data) { - struct qedr_dev *dev = (struct qedr_dev *)_dev; + struct qedr_dev *dev = (struct qedr_dev *)cxt; struct qedr_cq *cq = dev->gsi_rqcq; struct qedr_qp *qp = dev->gsi_qp; unsigned long flags; spin_lock_irqsave(&qp->q_lock, flags); - qp->rqe_wr_id[qp->rq.gsi_cons].rc = params->rc; - qp->rqe_wr_id[qp->rq.gsi_cons].vlan_id = params->vlan_id; - qp->rqe_wr_id[qp->rq.gsi_cons].sg_list[0].length = pkt->payload[0].len; - ether_addr_copy(qp->rqe_wr_id[qp->rq.gsi_cons].smac, params->smac); + qp->rqe_wr_id[qp->rq.gsi_cons].rc = data->u.data_length_error ? + -EINVAL : 0; + qp->rqe_wr_id[qp->rq.gsi_cons].vlan_id = data->vlan; + /* note: length stands for data length i.e. GRH is excluded */ + qp->rqe_wr_id[qp->rq.gsi_cons].sg_list[0].length = + data->length.data_length; + *((u32 *)&qp->rqe_wr_id[qp->rq.gsi_cons].smac[0]) = + ntohl(data->opaque_data_0); + *((u16 *)&qp->rqe_wr_id[qp->rq.gsi_cons].smac[4]) = + ntohs((u16)data->opaque_data_1); qedr_inc_sw_gsi_cons(&qp->rq); @@ -114,6 +122,14 @@ void qedr_ll2_rx_cb(void *_dev, struct qed_roce_ll2_packet *pkt, (*cq->ibcq.comp_handler) (&cq->ibcq, cq->ibcq.cq_context); } +void qedr_ll2_release_rx_packet(void *cxt, + u8 connection_handle, + void *cookie, + dma_addr_t rx_buf_addr, bool b_last_packet) +{ + /* Do nothing... */ +} + static void qedr_destroy_gsi_cq(struct qedr_dev *dev, struct ib_qp_init_attr *attrs) { @@ -162,27 +178,159 @@ static inline int qedr_check_gsi_qp_attrs(struct qedr_dev *dev, return 0; } +static int qedr_ll2_post_tx(struct qedr_dev *dev, + struct qed_roce_ll2_packet *pkt) +{ + enum qed_ll2_roce_flavor_type roce_flavor; + struct qed_ll2_tx_pkt_info ll2_tx_pkt; + int rc; + int i; + + memset(&ll2_tx_pkt, 0, sizeof(ll2_tx_pkt)); + + roce_flavor = (pkt->roce_mode == ROCE_V1) ? + QED_LL2_ROCE : QED_LL2_RROCE; + + if (pkt->roce_mode == ROCE_V2_IPV4) + ll2_tx_pkt.enable_ip_cksum = 1; + + ll2_tx_pkt.num_of_bds = 1 /* hdr */ + pkt->n_seg; + ll2_tx_pkt.vlan = 0; + ll2_tx_pkt.tx_dest = pkt->tx_dest; + ll2_tx_pkt.qed_roce_flavor = roce_flavor; + ll2_tx_pkt.first_frag = pkt->header.baddr; + ll2_tx_pkt.first_frag_len = pkt->header.len; + ll2_tx_pkt.cookie = pkt; + + /* tx header */ + rc = dev->ops->ll2_prepare_tx_packet(dev->rdma_ctx, + dev->gsi_ll2_handle, + &ll2_tx_pkt, 1); + if (rc) { + /* TX failed while posting header - release resources */ + dma_free_coherent(&dev->pdev->dev, pkt->header.len, + pkt->header.vaddr, pkt->header.baddr); + kfree(pkt); + + DP_ERR(dev, "roce ll2 tx: header failed (rc=%d)\n", rc); + return rc; + } + + /* tx payload */ + for (i = 0; i < pkt->n_seg; i++) { + rc = dev->ops->ll2_set_fragment_of_tx_packet( + dev->rdma_ctx, + dev->gsi_ll2_handle, + pkt->payload[i].baddr, + pkt->payload[i].len); + + if (rc) { + /* if failed not much to do here, partial packet has + * been posted we can't free memory, will need to wait + * for completion + */ + DP_ERR(dev, "ll2 tx: payload failed (rc=%d)\n", rc); + return rc; + } + } + + return 0; +} + +int qedr_ll2_stop(struct qedr_dev *dev) +{ + int rc; + + if (dev->gsi_ll2_handle == QED_LL2_UNUSED_HANDLE) + return 0; + + /* remove LL2 MAC address filter */ + rc = dev->ops->ll2_set_mac_filter(dev->cdev, + dev->gsi_ll2_mac_address, NULL); + + rc = dev->ops->ll2_terminate_connection(dev->rdma_ctx, + dev->gsi_ll2_handle); + if (rc) + DP_ERR(dev, "Failed to terminate LL2 connection (rc=%d)\n", rc); + + dev->ops->ll2_release_connection(dev->rdma_ctx, dev->gsi_ll2_handle); + + dev->gsi_ll2_handle = QED_LL2_UNUSED_HANDLE; + + return rc; +} + +int qedr_ll2_start(struct qedr_dev *dev, + struct ib_qp_init_attr *attrs, struct qedr_qp *qp) +{ + struct qed_ll2_acquire_data data; + struct qed_ll2_cbs cbs; + int rc; + + /* configure and start LL2 */ + cbs.rx_comp_cb = qedr_ll2_complete_rx_packet; + cbs.tx_comp_cb = qedr_ll2_complete_tx_packet; + cbs.rx_release_cb = qedr_ll2_release_rx_packet; + cbs.tx_release_cb = qedr_ll2_complete_tx_packet; + cbs.cookie = dev; + + memset(&data, 0, sizeof(data)); + data.input.conn_type = QED_LL2_TYPE_ROCE; + data.input.mtu = dev->ndev->mtu; + data.input.rx_num_desc = attrs->cap.max_recv_wr; + data.input.rx_drop_ttl0_flg = true; + data.input.rx_vlan_removal_en = false; + data.input.tx_num_desc = attrs->cap.max_send_wr; + data.input.tx_tc = 0; + data.input.tx_dest = QED_LL2_TX_DEST_NW; + data.input.ai_err_packet_too_big = QED_LL2_DROP_PACKET; + data.input.ai_err_no_buf = QED_LL2_DROP_PACKET; + data.input.gsi_enable = 1; + data.p_connection_handle = &dev->gsi_ll2_handle; + data.cbs = &cbs; + + rc = dev->ops->ll2_acquire_connection(dev->rdma_ctx, &data); + if (rc) { + DP_ERR(dev, + "ll2 start: failed to acquire LL2 connection (rc=%d)\n", + rc); + return rc; + } + + rc = dev->ops->ll2_establish_connection(dev->rdma_ctx, + dev->gsi_ll2_handle); + if (rc) { + DP_ERR(dev, + "ll2 start: failed to establish LL2 connection (rc=%d)\n", + rc); + goto err1; + } + + rc = dev->ops->ll2_set_mac_filter(dev->cdev, NULL, dev->ndev->dev_addr); + if (rc) + goto err2; + + return 0; + +err2: + dev->ops->ll2_terminate_connection(dev->rdma_ctx, dev->gsi_ll2_handle); +err1: + dev->ops->ll2_release_connection(dev->rdma_ctx, dev->gsi_ll2_handle); + + return rc; +} + struct ib_qp *qedr_create_gsi_qp(struct qedr_dev *dev, struct ib_qp_init_attr *attrs, struct qedr_qp *qp) { - struct qed_roce_ll2_params ll2_params; int rc; rc = qedr_check_gsi_qp_attrs(dev, attrs); if (rc) return ERR_PTR(rc); - /* configure and start LL2 */ - memset(&ll2_params, 0, sizeof(ll2_params)); - ll2_params.max_tx_buffers = attrs->cap.max_send_wr; - ll2_params.max_rx_buffers = attrs->cap.max_recv_wr; - ll2_params.cbs.tx_cb = qedr_ll2_tx_cb; - ll2_params.cbs.rx_cb = qedr_ll2_rx_cb; - ll2_params.cb_cookie = (void *)dev; - ll2_params.mtu = dev->ndev->mtu; - ether_addr_copy(ll2_params.mac_address, dev->ndev->dev_addr); - rc = dev->ops->roce_ll2_start(dev->cdev, &ll2_params); + rc = qedr_ll2_start(dev, attrs, qp); if (rc) { DP_ERR(dev, "create gsi qp: failed on ll2 start. rc=%d\n", rc); return ERR_PTR(rc); @@ -217,7 +365,7 @@ struct ib_qp *qedr_create_gsi_qp(struct qedr_dev *dev, err: kfree(qp->rqe_wr_id); - rc = dev->ops->roce_ll2_stop(dev->cdev); + rc = qedr_ll2_stop(dev); if (rc) DP_ERR(dev, "create gsi qp: failed destroy on create\n"); @@ -226,15 +374,7 @@ err: int qedr_destroy_gsi_qp(struct qedr_dev *dev) { - int rc; - - rc = dev->ops->roce_ll2_stop(dev->cdev); - if (rc) - DP_ERR(dev, "destroy gsi qp: failed (rc=%d)\n", rc); - else - DP_DEBUG(dev, QEDR_MSG_GSI, "destroy gsi qp: success\n"); - - return rc; + return qedr_ll2_stop(dev); } #define QEDR_MAX_UD_HEADER_SIZE (100) @@ -246,8 +386,8 @@ static inline int qedr_gsi_build_header(struct qedr_dev *dev, int *roce_mode) { bool has_vlan = false, has_grh_ipv6 = true; - struct ib_ah_attr *ah_attr = &get_qedr_ah(ud_wr(swr)->ah)->attr; - struct ib_global_route *grh = &ah_attr->grh; + struct rdma_ah_attr *ah_attr = &get_qedr_ah(ud_wr(swr)->ah)->attr; + const struct ib_global_route *grh = rdma_ah_read_grh(ah_attr); union ib_gid sgid; int send_size = 0; u16 vlan_id = 0; @@ -263,31 +403,34 @@ static inline int qedr_gsi_build_header(struct qedr_dev *dev, for (i = 0; i < swr->num_sge; ++i) send_size += swr->sg_list[i].length; - rc = ib_get_cached_gid(qp->ibqp.device, ah_attr->port_num, + rc = ib_get_cached_gid(qp->ibqp.device, rdma_ah_get_port_num(ah_attr), grh->sgid_index, &sgid, &sgid_attr); if (rc) { DP_ERR(dev, "gsi post send: failed to get cached GID (port=%d, ix=%d)\n", - ah_attr->port_num, grh->sgid_index); + rdma_ah_get_port_num(ah_attr), + grh->sgid_index); return rc; } - vlan_id = rdma_vlan_dev_vlan_id(sgid_attr.ndev); - if (vlan_id < VLAN_CFI_MASK) - has_vlan = true; - if (sgid_attr.ndev) + if (sgid_attr.ndev) { + vlan_id = rdma_vlan_dev_vlan_id(sgid_attr.ndev); + if (vlan_id < VLAN_CFI_MASK) + has_vlan = true; + dev_put(sgid_attr.ndev); + } if (!memcmp(&sgid, &zgid, sizeof(sgid))) { DP_ERR(dev, "gsi post send: GID not found GID index %d\n", - ah_attr->grh.sgid_index); + grh->sgid_index); return -ENOENT; } has_udp = (sgid_attr.gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP); if (!has_udp) { /* RoCE v1 */ - ether_type = ETH_P_ROCE; + ether_type = ETH_P_IBOE; *roce_mode = ROCE_V1; } else if (ipv6_addr_v4mapped((struct in6_addr *)&sgid)) { /* RoCE v2 IPv4 */ @@ -310,7 +453,7 @@ static inline int qedr_gsi_build_header(struct qedr_dev *dev, } /* ENET + VLAN headers */ - ether_addr_copy(udh->eth.dmac_h, ah_attr->dmac); + ether_addr_copy(udh->eth.dmac_h, ah_attr->roce.dmac); ether_addr_copy(udh->eth.smac_h, dev->ndev->dev_addr); if (has_vlan) { udh->eth.type = htons(ETH_P_8021Q); @@ -344,13 +487,13 @@ static inline int qedr_gsi_build_header(struct qedr_dev *dev, u32 ipv4_addr; udh->ip4.protocol = IPPROTO_UDP; - udh->ip4.tos = htonl(ah_attr->grh.flow_label); + udh->ip4.tos = htonl(grh->flow_label); udh->ip4.frag_off = htons(IP_DF); - udh->ip4.ttl = ah_attr->grh.hop_limit; + udh->ip4.ttl = grh->hop_limit; ipv4_addr = qedr_get_ipv4_from_gid(sgid.raw); udh->ip4.saddr = ipv4_addr; - ipv4_addr = qedr_get_ipv4_from_gid(ah_attr->grh.dgid.raw); + ipv4_addr = qedr_get_ipv4_from_gid(grh->dgid.raw); udh->ip4.daddr = ipv4_addr; /* note: checksum is calculated by the device */ } @@ -421,7 +564,6 @@ int qedr_gsi_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, { struct qed_roce_ll2_packet *pkt = NULL; struct qedr_qp *qp = get_qedr_qp(ibqp); - struct qed_roce_ll2_tx_params params; struct qedr_dev *dev = qp->dev; unsigned long flags; int rc; @@ -449,8 +591,6 @@ int qedr_gsi_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, goto err; } - memset(¶ms, 0, sizeof(params)); - spin_lock_irqsave(&qp->q_lock, flags); rc = qedr_gsi_build_packet(dev, qp, wr, &pkt); @@ -459,7 +599,8 @@ int qedr_gsi_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, goto err; } - rc = dev->ops->roce_ll2_tx(dev->cdev, pkt, ¶ms); + rc = qedr_ll2_post_tx(dev, pkt); + if (!rc) { qp->wqe_wr_id[qp->sq.prod].wr_id = wr->wr_id; qedr_inc_sw_prod(&qp->sq); @@ -467,17 +608,6 @@ int qedr_gsi_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, "gsi post send: opcode=%d, in_irq=%ld, irqs_disabled=%d, wr_id=%llx\n", wr->opcode, in_irq(), irqs_disabled(), wr->wr_id); } else { - if (rc == QED_ROCE_TX_HEAD_FAILURE) { - /* TX failed while posting header - release resources */ - dma_free_coherent(&dev->pdev->dev, pkt->header.len, - pkt->header.vaddr, pkt->header.baddr); - kfree(pkt); - } else if (rc == QED_ROCE_TX_FRAG_FAILURE) { - /* NTD since TX failed while posting a fragment. We will - * release the resources on TX callback - */ - } - DP_ERR(dev, "gsi post send: failed to transmit (rc=%d)\n", rc); rc = -EAGAIN; *bad_wr = wr; @@ -504,10 +634,8 @@ int qedr_gsi_post_recv(struct ib_qp *ibqp, struct ib_recv_wr *wr, { struct qedr_dev *dev = get_qedr_dev(ibqp->device); struct qedr_qp *qp = get_qedr_qp(ibqp); - struct qed_roce_ll2_buffer buf; unsigned long flags; - int status = 0; - int rc; + int rc = 0; if ((qp->state != QED_ROCE_QP_STATE_RTR) && (qp->state != QED_ROCE_QP_STATE_RTS)) { @@ -518,8 +646,6 @@ int qedr_gsi_post_recv(struct ib_qp *ibqp, struct ib_recv_wr *wr, return -EINVAL; } - memset(&buf, 0, sizeof(buf)); - spin_lock_irqsave(&qp->q_lock, flags); while (wr) { @@ -530,10 +656,12 @@ int qedr_gsi_post_recv(struct ib_qp *ibqp, struct ib_recv_wr *wr, goto err; } - buf.baddr = wr->sg_list[0].addr; - buf.len = wr->sg_list[0].length; - - rc = dev->ops->roce_ll2_post_rx_buffer(dev->cdev, &buf, 0, 1); + rc = dev->ops->ll2_post_rx_buffer(dev->rdma_ctx, + dev->gsi_ll2_handle, + wr->sg_list[0].addr, + wr->sg_list[0].length, + 0 /* cookie */, + 1 /* notify_fw */); if (rc) { DP_ERR(dev, "gsi post recv: failed to post rx buffer (rc=%d)\n", @@ -553,7 +681,7 @@ int qedr_gsi_post_recv(struct ib_qp *ibqp, struct ib_recv_wr *wr, spin_unlock_irqrestore(&qp->q_lock, flags); - return status; + return rc; err: spin_unlock_irqrestore(&qp->q_lock, flags); *bad_wr = wr; diff --git a/drivers/infiniband/hw/qedr/qedr_cm.h b/drivers/infiniband/hw/qedr/qedr_cm.h index 9ba6e15cd93f..a55916323ea9 100644 --- a/drivers/infiniband/hw/qedr/qedr_cm.h +++ b/drivers/infiniband/hw/qedr/qedr_cm.h @@ -37,10 +37,9 @@ #define QEDR_GSI_MAX_RECV_SGE (1) /* LL2 FW limitation */ -#define ETH_P_ROCE (0x8915) #define QEDR_ROCE_V2_UDP_SPORT (0000) -static inline u32 qedr_get_ipv4_from_gid(u8 *gid) +static inline u32 qedr_get_ipv4_from_gid(const u8 *gid) { return *(u32 *)(void *)&gid[12]; } diff --git a/drivers/infiniband/hw/qedr/verbs.c b/drivers/infiniband/hw/qedr/verbs.c index c7d6c9a783bd..2ae71b8f1ba8 100644 --- a/drivers/infiniband/hw/qedr/verbs.c +++ b/drivers/infiniband/hw/qedr/verbs.c @@ -43,7 +43,8 @@ #include <rdma/ib_addr.h> #include <rdma/ib_cache.h> -#include "qedr_hsi.h" +#include <linux/qed/common_hsi.h> +#include "qedr_hsi_rdma.h" #include <linux/qed/qed_if.h> #include "qedr.h" #include "verbs.h" @@ -52,6 +53,14 @@ #define DB_ADDR_SHIFT(addr) ((addr) << DB_PWM_ADDR_OFFSET_SHIFT) +static inline int qedr_ib_copy_to_udata(struct ib_udata *udata, void *src, + size_t len) +{ + size_t min_len = min_t(size_t, len, udata->outlen); + + return ib_copy_to_udata(udata, src, min_len); +} + int qedr_query_pkey(struct ib_device *ibdev, u8 port, u16 index, u16 *pkey) { if (index > QEDR_ROCE_PKEY_TABLE_LEN) @@ -238,8 +247,8 @@ int qedr_query_port(struct ib_device *ibdev, u8 port, struct ib_port_attr *attr) } rdma_port = dev->ops->rdma_query_port(dev->rdma_ctx); - memset(attr, 0, sizeof(*attr)); + /* *attr being zeroed by the caller, avoid zeroing it here */ if (rdma_port->port_state == QED_RDMA_PORT_UP) { attr->state = IB_PORT_ACTIVE; attr->phys_state = 5; @@ -377,7 +386,7 @@ struct ib_ucontext *qedr_alloc_ucontext(struct ib_device *ibdev, uresp.sges_per_srq_wr = dev->attr.max_srq_sge; uresp.max_cqes = QEDR_MAX_CQES; - rc = ib_copy_to_udata(udata, &uresp, sizeof(uresp)); + rc = qedr_ib_copy_to_udata(udata, &uresp, sizeof(uresp)); if (rc) goto err; @@ -498,7 +507,7 @@ struct ib_pd *qedr_alloc_pd(struct ib_device *ibdev, uresp.pd_id = pd_id; - rc = ib_copy_to_udata(udata, &uresp, sizeof(uresp)); + rc = qedr_ib_copy_to_udata(udata, &uresp, sizeof(uresp)); if (rc) { DP_ERR(dev, "copy error pd_id=0x%x.\n", pd_id); dev->ops->rdma_dealloc_pd(dev->rdma_ctx, pd_id); @@ -652,14 +661,15 @@ static int qedr_prepare_pbl_tbl(struct qedr_dev *dev, static void qedr_populate_pbls(struct qedr_dev *dev, struct ib_umem *umem, struct qedr_pbl *pbl, - struct qedr_pbl_info *pbl_info) + struct qedr_pbl_info *pbl_info, u32 pg_shift) { int shift, pg_cnt, pages, pbe_cnt, total_num_pbes = 0; + u32 fw_pg_cnt, fw_pg_per_umem_pg; struct qedr_pbl *pbl_tbl; struct scatterlist *sg; struct regpair *pbe; + u64 pg_addr; int entry; - u32 addr; if (!pbl_info->num_pbes) return; @@ -680,31 +690,37 @@ static void qedr_populate_pbls(struct qedr_dev *dev, struct ib_umem *umem, pbe_cnt = 0; - shift = ilog2(umem->page_size); + shift = umem->page_shift; + + fw_pg_per_umem_pg = BIT(umem->page_shift - pg_shift); for_each_sg(umem->sg_head.sgl, sg, umem->nmap, entry) { pages = sg_dma_len(sg) >> shift; + pg_addr = sg_dma_address(sg); for (pg_cnt = 0; pg_cnt < pages; pg_cnt++) { - /* store the page address in pbe */ - pbe->lo = cpu_to_le32(sg_dma_address(sg) + - umem->page_size * pg_cnt); - addr = upper_32_bits(sg_dma_address(sg) + - umem->page_size * pg_cnt); - pbe->hi = cpu_to_le32(addr); - pbe_cnt++; - total_num_pbes++; - pbe++; - - if (total_num_pbes == pbl_info->num_pbes) - return; - - /* If the given pbl is full storing the pbes, - * move to next pbl. - */ - if (pbe_cnt == (pbl_info->pbl_size / sizeof(u64))) { - pbl_tbl++; - pbe = (struct regpair *)pbl_tbl->va; - pbe_cnt = 0; + for (fw_pg_cnt = 0; fw_pg_cnt < fw_pg_per_umem_pg;) { + pbe->lo = cpu_to_le32(pg_addr); + pbe->hi = cpu_to_le32(upper_32_bits(pg_addr)); + + pg_addr += BIT(pg_shift); + pbe_cnt++; + total_num_pbes++; + pbe++; + + if (total_num_pbes == pbl_info->num_pbes) + return; + + /* If the given pbl is full storing the pbes, + * move to next pbl. + */ + if (pbe_cnt == + (pbl_info->pbl_size / sizeof(u64))) { + pbl_tbl++; + pbe = (struct regpair *)pbl_tbl->va; + pbe_cnt = 0; + } + + fw_pg_cnt++; } } } @@ -721,7 +737,7 @@ static int qedr_copy_cq_uresp(struct qedr_dev *dev, uresp.db_offset = DB_ADDR_SHIFT(DQ_PWM_OFFSET_UCM_RDMA_CQ_CONS_32BIT); uresp.icid = cq->icid; - rc = ib_copy_to_udata(udata, &uresp, sizeof(uresp)); + rc = qedr_ib_copy_to_udata(udata, &uresp, sizeof(uresp)); if (rc) DP_ERR(dev, "copy error cqid=0x%x.\n", cq->icid); @@ -753,7 +769,7 @@ static inline int qedr_init_user_queue(struct ib_ucontext *ib_ctx, u64 buf_addr, size_t buf_len, int access, int dmasync) { - int page_cnt; + u32 fw_pages; int rc; q->buf_addr = buf_addr; @@ -765,16 +781,21 @@ static inline int qedr_init_user_queue(struct ib_ucontext *ib_ctx, return PTR_ERR(q->umem); } - page_cnt = ib_umem_page_count(q->umem); - rc = qedr_prepare_pbl_tbl(dev, &q->pbl_info, page_cnt, 0); + fw_pages = ib_umem_page_count(q->umem) << + (q->umem->page_shift - FW_PAGE_SHIFT); + + rc = qedr_prepare_pbl_tbl(dev, &q->pbl_info, fw_pages, 0); if (rc) goto err0; q->pbl_tbl = qedr_alloc_pbl_tbl(dev, &q->pbl_info, GFP_KERNEL); - if (IS_ERR_OR_NULL(q->pbl_tbl)) + if (IS_ERR(q->pbl_tbl)) { + rc = PTR_ERR(q->pbl_tbl); goto err0; + } - qedr_populate_pbls(dev, q->umem, q->pbl_tbl, &q->pbl_info); + qedr_populate_pbls(dev, q->umem, q->pbl_tbl, &q->pbl_info, + FW_PAGE_SHIFT); return 0; @@ -819,6 +840,17 @@ int qedr_arm_cq(struct ib_cq *ibcq, enum ib_cq_notify_flags flags) { struct qedr_cq *cq = get_qedr_cq(ibcq); unsigned long sflags; + struct qedr_dev *dev; + + dev = get_qedr_dev(ibcq->device); + + if (cq->destroyed) { + DP_ERR(dev, + "warning: arm was invoked after destroy for cq %p (icid=%d)\n", + cq, cq->icid); + return -EINVAL; + } + if (cq->cq_type == QEDR_CQ_TYPE_GSI) return 0; @@ -911,7 +943,7 @@ struct ib_cq *qedr_create_cq(struct ib_device *ibdev, QED_CHAIN_CNT_TYPE_U32, chain_entries, sizeof(union rdma_cqe), - &cq->pbl); + &cq->pbl, NULL); if (rc) goto err1; @@ -984,35 +1016,82 @@ int qedr_resize_cq(struct ib_cq *ibcq, int new_cnt, struct ib_udata *udata) return 0; } +#define QEDR_DESTROY_CQ_MAX_ITERATIONS (10) +#define QEDR_DESTROY_CQ_ITER_DURATION (10) + int qedr_destroy_cq(struct ib_cq *ibcq) { struct qedr_dev *dev = get_qedr_dev(ibcq->device); struct qed_rdma_destroy_cq_out_params oparams; struct qed_rdma_destroy_cq_in_params iparams; struct qedr_cq *cq = get_qedr_cq(ibcq); + int iter; + int rc; + + DP_DEBUG(dev, QEDR_MSG_CQ, "destroy cq %p (icid=%d)\n", cq, cq->icid); - DP_DEBUG(dev, QEDR_MSG_CQ, "destroy cq: cq_id %d", cq->icid); + cq->destroyed = 1; /* GSIs CQs are handled by driver, so they don't exist in the FW */ - if (cq->cq_type != QEDR_CQ_TYPE_GSI) { - int rc; + if (cq->cq_type == QEDR_CQ_TYPE_GSI) + goto done; - iparams.icid = cq->icid; - rc = dev->ops->rdma_destroy_cq(dev->rdma_ctx, &iparams, - &oparams); - if (rc) - return rc; - dev->ops->common->chain_free(dev->cdev, &cq->pbl); - } + iparams.icid = cq->icid; + rc = dev->ops->rdma_destroy_cq(dev->rdma_ctx, &iparams, &oparams); + if (rc) + return rc; + + dev->ops->common->chain_free(dev->cdev, &cq->pbl); if (ibcq->uobject && ibcq->uobject->context) { qedr_free_pbl(dev, &cq->q.pbl_info, cq->q.pbl_tbl); ib_umem_release(cq->q.umem); } + /* We don't want the IRQ handler to handle a non-existing CQ so we + * wait until all CNQ interrupts, if any, are received. This will always + * happen and will always happen very fast. If not, then a serious error + * has occured. That is why we can use a long delay. + * We spin for a short time so we don’t lose time on context switching + * in case all the completions are handled in that span. Otherwise + * we sleep for a while and check again. Since the CNQ may be + * associated with (only) the current CPU we use msleep to allow the + * current CPU to be freed. + * The CNQ notification is increased in qedr_irq_handler(). + */ + iter = QEDR_DESTROY_CQ_MAX_ITERATIONS; + while (oparams.num_cq_notif != READ_ONCE(cq->cnq_notif) && iter) { + udelay(QEDR_DESTROY_CQ_ITER_DURATION); + iter--; + } + + iter = QEDR_DESTROY_CQ_MAX_ITERATIONS; + while (oparams.num_cq_notif != READ_ONCE(cq->cnq_notif) && iter) { + msleep(QEDR_DESTROY_CQ_ITER_DURATION); + iter--; + } + + if (oparams.num_cq_notif != cq->cnq_notif) + goto err; + + /* Note that we don't need to have explicit code to wait for the + * completion of the event handler because it is invoked from the EQ. + * Since the destroy CQ ramrod has also been received on the EQ we can + * be certain that there's no event handler in process. + */ +done: + cq->sig = ~cq->sig; + kfree(cq); return 0; + +err: + DP_ERR(dev, + "CQ %p (icid=%d) not freed, expecting %d ints but got %d ints\n", + cq, cq->icid, oparams.num_cq_notif, cq->cnq_notif); + + return -EINVAL; } static inline int get_gid_info_from_table(struct ib_qp *ibqp, @@ -1023,13 +1102,15 @@ static inline int get_gid_info_from_table(struct ib_qp *ibqp, { enum rdma_network_type nw_type; struct ib_gid_attr gid_attr; + const struct ib_global_route *grh = rdma_ah_read_grh(&attr->ah_attr); union ib_gid gid; u32 ipv4_addr; int rc = 0; int i; - rc = ib_get_cached_gid(ibqp->device, attr->ah_attr.port_num, - attr->ah_attr.grh.sgid_index, &gid, &gid_attr); + rc = ib_get_cached_gid(ibqp->device, + rdma_ah_get_port_num(&attr->ah_attr), + grh->sgid_index, &gid, &gid_attr); if (rc) return rc; @@ -1046,7 +1127,7 @@ static inline int get_gid_info_from_table(struct ib_qp *ibqp, memcpy(&qp_params->sgid.bytes[0], &gid.raw[0], sizeof(qp_params->sgid)); memcpy(&qp_params->dgid.bytes[0], - &attr->ah_attr.grh.dgid, + &grh->dgid, sizeof(qp_params->dgid)); qp_params->roce_mode = ROCE_V2_IPV6; SET_FIELD(qp_params->modify_flags, @@ -1056,7 +1137,7 @@ static inline int get_gid_info_from_table(struct ib_qp *ibqp, memcpy(&qp_params->sgid.bytes[0], &gid.raw[0], sizeof(qp_params->sgid)); memcpy(&qp_params->dgid.bytes[0], - &attr->ah_attr.grh.dgid, + &grh->dgid, sizeof(qp_params->dgid)); qp_params->roce_mode = ROCE_V1; break; @@ -1066,7 +1147,7 @@ static inline int get_gid_info_from_table(struct ib_qp *ibqp, ipv4_addr = qedr_get_ipv4_from_gid(gid.raw); qp_params->sgid.ipv4_addr = ipv4_addr; ipv4_addr = - qedr_get_ipv4_from_gid(attr->ah_attr.grh.dgid.raw); + qedr_get_ipv4_from_gid(grh->dgid.raw); qp_params->dgid.ipv4_addr = ipv4_addr; SET_FIELD(qp_params->modify_flags, QED_ROCE_MODIFY_QP_VALID_ROCE_MODE, 1); @@ -1086,30 +1167,6 @@ static inline int get_gid_info_from_table(struct ib_qp *ibqp, return 0; } -static void qedr_cleanup_user_sq(struct qedr_dev *dev, struct qedr_qp *qp) -{ - qedr_free_pbl(dev, &qp->usq.pbl_info, qp->usq.pbl_tbl); - ib_umem_release(qp->usq.umem); -} - -static void qedr_cleanup_user_rq(struct qedr_dev *dev, struct qedr_qp *qp) -{ - qedr_free_pbl(dev, &qp->urq.pbl_info, qp->urq.pbl_tbl); - ib_umem_release(qp->urq.umem); -} - -static void qedr_cleanup_kernel_sq(struct qedr_dev *dev, struct qedr_qp *qp) -{ - dev->ops->common->chain_free(dev->cdev, &qp->sq.pbl); - kfree(qp->wqe_wr_id); -} - -static void qedr_cleanup_kernel_rq(struct qedr_dev *dev, struct qedr_qp *qp) -{ - dev->ops->common->chain_free(dev->cdev, &qp->rq.pbl); - kfree(qp->rqe_wr_id); -} - static int qedr_check_qp_attrs(struct ib_pd *ibpd, struct qedr_dev *dev, struct ib_qp_init_attr *attrs) { @@ -1189,7 +1246,7 @@ static int qedr_copy_qp_uresp(struct qedr_dev *dev, uresp.atomic_supported = dev->atomic_cap != IB_ATOMIC_NONE; uresp.qp_id = qp->qp_id; - rc = ib_copy_to_udata(udata, &uresp, sizeof(uresp)); + rc = qedr_ib_copy_to_udata(udata, &uresp, sizeof(uresp)); if (rc) DP_ERR(dev, "create qp: failed a copy to user space with qp icid=0x%x.\n", @@ -1198,15 +1255,13 @@ static int qedr_copy_qp_uresp(struct qedr_dev *dev, return rc; } -static void qedr_set_qp_init_params(struct qedr_dev *dev, - struct qedr_qp *qp, - struct qedr_pd *pd, - struct ib_qp_init_attr *attrs) +static void qedr_set_common_qp_params(struct qedr_dev *dev, + struct qedr_qp *qp, + struct qedr_pd *pd, + struct ib_qp_init_attr *attrs) { - qp->pd = pd; - spin_lock_init(&qp->q_lock); - + qp->pd = pd; qp->qp_type = attrs->qp_type; qp->max_inline_data = attrs->cap.max_inline_data; qp->sq.max_sges = attrs->cap.max_send_sge; @@ -1215,272 +1270,278 @@ static void qedr_set_qp_init_params(struct qedr_dev *dev, qp->sq_cq = get_qedr_cq(attrs->send_cq); qp->rq_cq = get_qedr_cq(attrs->recv_cq); qp->dev = dev; + qp->rq.max_sges = attrs->cap.max_recv_sge; DP_DEBUG(dev, QEDR_MSG_QP, + "RQ params:\trq_max_sges = %d, rq_cq_id = %d\n", + qp->rq.max_sges, qp->rq_cq->icid); + DP_DEBUG(dev, QEDR_MSG_QP, "QP params:\tpd = %d, qp_type = %d, max_inline_data = %d, state = %d, signaled = %d, use_srq=%d\n", pd->pd_id, qp->qp_type, qp->max_inline_data, qp->state, qp->signaled, (attrs->srq) ? 1 : 0); DP_DEBUG(dev, QEDR_MSG_QP, "SQ params:\tsq_max_sges = %d, sq_cq_id = %d\n", qp->sq.max_sges, qp->sq_cq->icid); - qp->rq.max_sges = attrs->cap.max_recv_sge; - DP_DEBUG(dev, QEDR_MSG_QP, - "RQ params:\trq_max_sges = %d, rq_cq_id = %d\n", - qp->rq.max_sges, qp->rq_cq->icid); } -static inline void -qedr_init_qp_user_params(struct qed_rdma_create_qp_in_params *params, - struct qedr_create_qp_ureq *ureq) -{ - /* QP handle to be written in CQE */ - params->qp_handle_lo = ureq->qp_handle_lo; - params->qp_handle_hi = ureq->qp_handle_hi; -} - -static inline void -qedr_init_qp_kernel_doorbell_sq(struct qedr_dev *dev, struct qedr_qp *qp) +static void qedr_set_roce_db_info(struct qedr_dev *dev, struct qedr_qp *qp) { qp->sq.db = dev->db_addr + DB_ADDR_SHIFT(DQ_PWM_OFFSET_XCM_RDMA_SQ_PROD); qp->sq.db_data.data.icid = qp->icid + 1; + qp->rq.db = dev->db_addr + + DB_ADDR_SHIFT(DQ_PWM_OFFSET_TCM_ROCE_RQ_PROD); + qp->rq.db_data.data.icid = qp->icid; } static inline void -qedr_init_qp_kernel_doorbell_rq(struct qedr_dev *dev, struct qedr_qp *qp) +qedr_init_common_qp_in_params(struct qedr_dev *dev, + struct qedr_pd *pd, + struct qedr_qp *qp, + struct ib_qp_init_attr *attrs, + bool fmr_and_reserved_lkey, + struct qed_rdma_create_qp_in_params *params) { - qp->rq.db = dev->db_addr + - DB_ADDR_SHIFT(DQ_PWM_OFFSET_TCM_ROCE_RQ_PROD); - qp->rq.db_data.data.icid = qp->icid; + /* QP handle to be written in an async event */ + params->qp_handle_async_lo = lower_32_bits((uintptr_t) qp); + params->qp_handle_async_hi = upper_32_bits((uintptr_t) qp); + + params->signal_all = (attrs->sq_sig_type == IB_SIGNAL_ALL_WR); + params->fmr_and_reserved_lkey = fmr_and_reserved_lkey; + params->pd = pd->pd_id; + params->dpi = pd->uctx ? pd->uctx->dpi : dev->dpi; + params->sq_cq_id = get_qedr_cq(attrs->send_cq)->icid; + params->stats_queue = 0; + params->rq_cq_id = get_qedr_cq(attrs->recv_cq)->icid; + params->srq_id = 0; + params->use_srq = false; } -static inline int -qedr_init_qp_kernel_params_rq(struct qedr_dev *dev, - struct qedr_qp *qp, struct ib_qp_init_attr *attrs) +static inline void qedr_qp_user_print(struct qedr_dev *dev, struct qedr_qp *qp) { - /* Allocate driver internal RQ array */ - qp->rqe_wr_id = kcalloc(qp->rq.max_wr, sizeof(*qp->rqe_wr_id), - GFP_KERNEL); - if (!qp->rqe_wr_id) - return -ENOMEM; + DP_DEBUG(dev, QEDR_MSG_QP, "create qp: successfully created user QP. " + "qp=%p. " + "sq_addr=0x%llx, " + "sq_len=%zd, " + "rq_addr=0x%llx, " + "rq_len=%zd" + "\n", + qp, + qp->usq.buf_addr, + qp->usq.buf_len, qp->urq.buf_addr, qp->urq.buf_len); +} - DP_DEBUG(dev, QEDR_MSG_QP, "RQ max_wr set to %d.\n", qp->rq.max_wr); +static void qedr_cleanup_user(struct qedr_dev *dev, struct qedr_qp *qp) +{ + if (qp->usq.umem) + ib_umem_release(qp->usq.umem); + qp->usq.umem = NULL; - return 0; + if (qp->urq.umem) + ib_umem_release(qp->urq.umem); + qp->urq.umem = NULL; } -static inline int -qedr_init_qp_kernel_params_sq(struct qedr_dev *dev, - struct qedr_qp *qp, - struct ib_qp_init_attr *attrs, - struct qed_rdma_create_qp_in_params *params) +static int qedr_create_user_qp(struct qedr_dev *dev, + struct qedr_qp *qp, + struct ib_pd *ibpd, + struct ib_udata *udata, + struct ib_qp_init_attr *attrs) { - u32 temp_max_wr; + struct qed_rdma_create_qp_in_params in_params; + struct qed_rdma_create_qp_out_params out_params; + struct qedr_pd *pd = get_qedr_pd(ibpd); + struct ib_ucontext *ib_ctx = NULL; + struct qedr_ucontext *ctx = NULL; + struct qedr_create_qp_ureq ureq; + int rc = -EINVAL; - /* Allocate driver internal SQ array */ - temp_max_wr = attrs->cap.max_send_wr * dev->wq_multiplier; - temp_max_wr = min_t(u32, temp_max_wr, dev->attr.max_sqe); + ib_ctx = ibpd->uobject->context; + ctx = get_qedr_ucontext(ib_ctx); - /* temp_max_wr < attr->max_sqe < u16 so the casting is safe */ - qp->sq.max_wr = (u16)temp_max_wr; - qp->wqe_wr_id = kcalloc(qp->sq.max_wr, sizeof(*qp->wqe_wr_id), - GFP_KERNEL); - if (!qp->wqe_wr_id) - return -ENOMEM; + memset(&ureq, 0, sizeof(ureq)); + rc = ib_copy_from_udata(&ureq, udata, sizeof(ureq)); + if (rc) { + DP_ERR(dev, "Problem copying data from user space\n"); + return rc; + } + + /* SQ - read access only (0), dma sync not required (0) */ + rc = qedr_init_user_queue(ib_ctx, dev, &qp->usq, ureq.sq_addr, + ureq.sq_len, 0, 0); + if (rc) + return rc; - DP_DEBUG(dev, QEDR_MSG_QP, "SQ max_wr set to %d.\n", qp->sq.max_wr); + /* RQ - read access only (0), dma sync not required (0) */ + rc = qedr_init_user_queue(ib_ctx, dev, &qp->urq, ureq.rq_addr, + ureq.rq_len, 0, 0); - /* QP handle to be written in CQE */ - params->qp_handle_lo = lower_32_bits((uintptr_t)qp); - params->qp_handle_hi = upper_32_bits((uintptr_t)qp); + if (rc) + return rc; + + memset(&in_params, 0, sizeof(in_params)); + qedr_init_common_qp_in_params(dev, pd, qp, attrs, false, &in_params); + in_params.qp_handle_lo = ureq.qp_handle_lo; + in_params.qp_handle_hi = ureq.qp_handle_hi; + in_params.sq_num_pages = qp->usq.pbl_info.num_pbes; + in_params.sq_pbl_ptr = qp->usq.pbl_tbl->pa; + in_params.rq_num_pages = qp->urq.pbl_info.num_pbes; + in_params.rq_pbl_ptr = qp->urq.pbl_tbl->pa; + + qp->qed_qp = dev->ops->rdma_create_qp(dev->rdma_ctx, + &in_params, &out_params); + + if (!qp->qed_qp) { + rc = -ENOMEM; + goto err1; + } + + qp->qp_id = out_params.qp_id; + qp->icid = out_params.icid; + + rc = qedr_copy_qp_uresp(dev, qp, udata); + if (rc) + goto err; + + qedr_qp_user_print(dev, qp); return 0; +err: + rc = dev->ops->rdma_destroy_qp(dev->rdma_ctx, qp->qed_qp); + if (rc) + DP_ERR(dev, "create qp: fatal fault. rc=%d", rc); + +err1: + qedr_cleanup_user(dev, qp); + return rc; } -static inline int qedr_init_qp_kernel_sq(struct qedr_dev *dev, - struct qedr_qp *qp, - struct ib_qp_init_attr *attrs) +static int +qedr_roce_create_kernel_qp(struct qedr_dev *dev, + struct qedr_qp *qp, + struct qed_rdma_create_qp_in_params *in_params, + u32 n_sq_elems, u32 n_rq_elems) { - u32 n_sq_elems, n_sq_entries; + struct qed_rdma_create_qp_out_params out_params; int rc; - /* A single work request may take up to QEDR_MAX_SQ_WQE_SIZE elements in - * the ring. The ring should allow at least a single WR, even if the - * user requested none, due to allocation issues. - */ - n_sq_entries = attrs->cap.max_send_wr; - n_sq_entries = min_t(u32, n_sq_entries, dev->attr.max_sqe); - n_sq_entries = max_t(u32, n_sq_entries, 1); - n_sq_elems = n_sq_entries * QEDR_MAX_SQE_ELEMENTS_PER_SQE; rc = dev->ops->common->chain_alloc(dev->cdev, QED_CHAIN_USE_TO_PRODUCE, QED_CHAIN_MODE_PBL, QED_CHAIN_CNT_TYPE_U32, n_sq_elems, QEDR_SQE_ELEMENT_SIZE, - &qp->sq.pbl); - if (rc) { - DP_ERR(dev, "failed to allocate QP %p SQ\n", qp); + &qp->sq.pbl, NULL); + + if (rc) return rc; - } - DP_DEBUG(dev, QEDR_MSG_SQ, - "SQ Pbl base addr = %llx max_send_wr=%d max_wr=%d capacity=%d, rc=%d\n", - qed_chain_get_pbl_phys(&qp->sq.pbl), attrs->cap.max_send_wr, - n_sq_entries, qed_chain_get_capacity(&qp->sq.pbl), rc); - return 0; -} + in_params->sq_num_pages = qed_chain_get_page_cnt(&qp->sq.pbl); + in_params->sq_pbl_ptr = qed_chain_get_pbl_phys(&qp->sq.pbl); -static inline int qedr_init_qp_kernel_rq(struct qedr_dev *dev, - struct qedr_qp *qp, - struct ib_qp_init_attr *attrs) -{ - u32 n_rq_elems, n_rq_entries; - int rc; - - /* A single work request may take up to QEDR_MAX_RQ_WQE_SIZE elements in - * the ring. There ring should allow at least a single WR, even if the - * user requested none, due to allocation issues. - */ - n_rq_entries = max_t(u32, attrs->cap.max_recv_wr, 1); - n_rq_elems = n_rq_entries * QEDR_MAX_RQE_ELEMENTS_PER_RQE; rc = dev->ops->common->chain_alloc(dev->cdev, QED_CHAIN_USE_TO_CONSUME_PRODUCE, QED_CHAIN_MODE_PBL, QED_CHAIN_CNT_TYPE_U32, n_rq_elems, QEDR_RQE_ELEMENT_SIZE, - &qp->rq.pbl); - - if (rc) { - DP_ERR(dev, "failed to allocate memory for QP %p RQ\n", qp); - return -ENOMEM; - } + &qp->rq.pbl, NULL); + if (rc) + return rc; - DP_DEBUG(dev, QEDR_MSG_RQ, - "RQ Pbl base addr = %llx max_recv_wr=%d max_wr=%d capacity=%d, rc=%d\n", - qed_chain_get_pbl_phys(&qp->rq.pbl), attrs->cap.max_recv_wr, - n_rq_entries, qed_chain_get_capacity(&qp->rq.pbl), rc); + in_params->rq_num_pages = qed_chain_get_page_cnt(&qp->rq.pbl); + in_params->rq_pbl_ptr = qed_chain_get_pbl_phys(&qp->rq.pbl); - /* n_rq_entries < u16 so the casting is safe */ - qp->rq.max_wr = (u16)n_rq_entries; + qp->qed_qp = dev->ops->rdma_create_qp(dev->rdma_ctx, + in_params, &out_params); - return 0; -} + if (!qp->qed_qp) + return -EINVAL; -static inline void -qedr_init_qp_in_params_sq(struct qedr_dev *dev, - struct qedr_pd *pd, - struct qedr_qp *qp, - struct ib_qp_init_attr *attrs, - struct ib_udata *udata, - struct qed_rdma_create_qp_in_params *params) -{ - /* QP handle to be written in an async event */ - params->qp_handle_async_lo = lower_32_bits((uintptr_t)qp); - params->qp_handle_async_hi = upper_32_bits((uintptr_t)qp); + qp->qp_id = out_params.qp_id; + qp->icid = out_params.icid; - params->signal_all = (attrs->sq_sig_type == IB_SIGNAL_ALL_WR); - params->fmr_and_reserved_lkey = !udata; - params->pd = pd->pd_id; - params->dpi = pd->uctx ? pd->uctx->dpi : dev->dpi; - params->sq_cq_id = get_qedr_cq(attrs->send_cq)->icid; - params->max_sq_sges = 0; - params->stats_queue = 0; + qedr_set_roce_db_info(dev, qp); - if (udata) { - params->sq_num_pages = qp->usq.pbl_info.num_pbes; - params->sq_pbl_ptr = qp->usq.pbl_tbl->pa; - } else { - params->sq_num_pages = qed_chain_get_page_cnt(&qp->sq.pbl); - params->sq_pbl_ptr = qed_chain_get_pbl_phys(&qp->sq.pbl); - } + return 0; } -static inline void -qedr_init_qp_in_params_rq(struct qedr_qp *qp, - struct ib_qp_init_attr *attrs, - struct ib_udata *udata, - struct qed_rdma_create_qp_in_params *params) +static void qedr_cleanup_kernel(struct qedr_dev *dev, struct qedr_qp *qp) { - params->rq_cq_id = get_qedr_cq(attrs->recv_cq)->icid; - params->srq_id = 0; - params->use_srq = false; + dev->ops->common->chain_free(dev->cdev, &qp->sq.pbl); + kfree(qp->wqe_wr_id); - if (udata) { - params->rq_num_pages = qp->urq.pbl_info.num_pbes; - params->rq_pbl_ptr = qp->urq.pbl_tbl->pa; - } else { - params->rq_num_pages = qed_chain_get_page_cnt(&qp->rq.pbl); - params->rq_pbl_ptr = qed_chain_get_pbl_phys(&qp->rq.pbl); - } + dev->ops->common->chain_free(dev->cdev, &qp->rq.pbl); + kfree(qp->rqe_wr_id); } -static inline void qedr_qp_user_print(struct qedr_dev *dev, struct qedr_qp *qp) +static int qedr_create_kernel_qp(struct qedr_dev *dev, + struct qedr_qp *qp, + struct ib_pd *ibpd, + struct ib_qp_init_attr *attrs) { - DP_DEBUG(dev, QEDR_MSG_QP, - "create qp: successfully created user QP. qp=%p, sq_addr=0x%llx, sq_len=%zd, rq_addr=0x%llx, rq_len=%zd\n", - qp, qp->usq.buf_addr, qp->usq.buf_len, qp->urq.buf_addr, - qp->urq.buf_len); -} + struct qed_rdma_create_qp_in_params in_params; + struct qedr_pd *pd = get_qedr_pd(ibpd); + int rc = -EINVAL; + u32 n_rq_elems; + u32 n_sq_elems; + u32 n_sq_entries; -static inline int qedr_init_user_qp(struct ib_ucontext *ib_ctx, - struct qedr_dev *dev, - struct qedr_qp *qp, - struct qedr_create_qp_ureq *ureq) -{ - int rc; + memset(&in_params, 0, sizeof(in_params)); - /* SQ - read access only (0), dma sync not required (0) */ - rc = qedr_init_user_queue(ib_ctx, dev, &qp->usq, ureq->sq_addr, - ureq->sq_len, 0, 0); - if (rc) - return rc; + /* A single work request may take up to QEDR_MAX_SQ_WQE_SIZE elements in + * the ring. The ring should allow at least a single WR, even if the + * user requested none, due to allocation issues. + * We should add an extra WR since the prod and cons indices of + * wqe_wr_id are managed in such a way that the WQ is considered full + * when (prod+1)%max_wr==cons. We currently don't do that because we + * double the number of entries due an iSER issue that pushes far more + * WRs than indicated. If we decline its ib_post_send() then we get + * error prints in the dmesg we'd like to avoid. + */ + qp->sq.max_wr = min_t(u32, attrs->cap.max_send_wr * dev->wq_multiplier, + dev->attr.max_sqe); - /* RQ - read access only (0), dma sync not required (0) */ - rc = qedr_init_user_queue(ib_ctx, dev, &qp->urq, ureq->rq_addr, - ureq->rq_len, 0, 0); + qp->wqe_wr_id = kzalloc(qp->sq.max_wr * sizeof(*qp->wqe_wr_id), + GFP_KERNEL); + if (!qp->wqe_wr_id) { + DP_ERR(dev, "create qp: failed SQ shadow memory allocation\n"); + return -ENOMEM; + } - if (rc) - qedr_cleanup_user_sq(dev, qp); - return rc; -} + /* QP handle to be written in CQE */ + in_params.qp_handle_lo = lower_32_bits((uintptr_t) qp); + in_params.qp_handle_hi = upper_32_bits((uintptr_t) qp); -static inline int -qedr_init_kernel_qp(struct qedr_dev *dev, - struct qedr_qp *qp, - struct ib_qp_init_attr *attrs, - struct qed_rdma_create_qp_in_params *params) -{ - int rc; + /* A single work request may take up to QEDR_MAX_RQ_WQE_SIZE elements in + * the ring. There ring should allow at least a single WR, even if the + * user requested none, due to allocation issues. + */ + qp->rq.max_wr = (u16) max_t(u32, attrs->cap.max_recv_wr, 1); - rc = qedr_init_qp_kernel_sq(dev, qp, attrs); - if (rc) { - DP_ERR(dev, "failed to init kernel QP %p SQ\n", qp); - return rc; + /* Allocate driver internal RQ array */ + qp->rqe_wr_id = kzalloc(qp->rq.max_wr * sizeof(*qp->rqe_wr_id), + GFP_KERNEL); + if (!qp->rqe_wr_id) { + DP_ERR(dev, + "create qp: failed RQ shadow memory allocation\n"); + kfree(qp->wqe_wr_id); + return -ENOMEM; } - rc = qedr_init_qp_kernel_params_sq(dev, qp, attrs, params); - if (rc) { - dev->ops->common->chain_free(dev->cdev, &qp->sq.pbl); - DP_ERR(dev, "failed to init kernel QP %p SQ params\n", qp); - return rc; - } + qedr_init_common_qp_in_params(dev, pd, qp, attrs, true, &in_params); - rc = qedr_init_qp_kernel_rq(dev, qp, attrs); - if (rc) { - qedr_cleanup_kernel_sq(dev, qp); - DP_ERR(dev, "failed to init kernel QP %p RQ\n", qp); - return rc; - } + n_sq_entries = attrs->cap.max_send_wr; + n_sq_entries = min_t(u32, n_sq_entries, dev->attr.max_sqe); + n_sq_entries = max_t(u32, n_sq_entries, 1); + n_sq_elems = n_sq_entries * QEDR_MAX_SQE_ELEMENTS_PER_SQE; - rc = qedr_init_qp_kernel_params_rq(dev, qp, attrs); - if (rc) { - DP_ERR(dev, "failed to init kernel QP %p RQ params\n", qp); - qedr_cleanup_kernel_sq(dev, qp); - dev->ops->common->chain_free(dev->cdev, &qp->rq.pbl); - return rc; - } + n_rq_elems = qp->rq.max_wr * QEDR_MAX_RQE_ELEMENTS_PER_RQE; + + rc = qedr_roce_create_kernel_qp(dev, qp, &in_params, + n_sq_elems, n_rq_elems); + if (rc) + qedr_cleanup_kernel(dev, qp); return rc; } @@ -1490,12 +1551,7 @@ struct ib_qp *qedr_create_qp(struct ib_pd *ibpd, struct ib_udata *udata) { struct qedr_dev *dev = get_qedr_dev(ibpd->device); - struct qed_rdma_create_qp_out_params out_params; - struct qed_rdma_create_qp_in_params in_params; struct qedr_pd *pd = get_qedr_pd(ibpd); - struct ib_ucontext *ib_ctx = NULL; - struct qedr_ucontext *ctx = NULL; - struct qedr_create_qp_ureq ureq; struct qedr_qp *qp; struct ib_qp *ibqp; int rc = 0; @@ -1510,101 +1566,42 @@ struct ib_qp *qedr_create_qp(struct ib_pd *ibpd, if (attrs->srq) return ERR_PTR(-EINVAL); - qp = kzalloc(sizeof(*qp), GFP_KERNEL); - if (!qp) - return ERR_PTR(-ENOMEM); - DP_DEBUG(dev, QEDR_MSG_QP, - "create qp: sq_cq=%p, sq_icid=%d, rq_cq=%p, rq_icid=%d\n", + "create qp: called from %s, event_handler=%p, eepd=%p sq_cq=%p, sq_icid=%d, rq_cq=%p, rq_icid=%d\n", + udata ? "user library" : "kernel", attrs->event_handler, pd, get_qedr_cq(attrs->send_cq), get_qedr_cq(attrs->send_cq)->icid, get_qedr_cq(attrs->recv_cq), get_qedr_cq(attrs->recv_cq)->icid); - qedr_set_qp_init_params(dev, qp, pd, attrs); + qp = kzalloc(sizeof(*qp), GFP_KERNEL); + if (!qp) { + DP_ERR(dev, "create qp: failed allocating memory\n"); + return ERR_PTR(-ENOMEM); + } + + qedr_set_common_qp_params(dev, qp, pd, attrs); if (attrs->qp_type == IB_QPT_GSI) { - if (udata) { - DP_ERR(dev, - "create qp: unexpected udata when creating GSI QP\n"); - goto err0; - } ibqp = qedr_create_gsi_qp(dev, attrs, qp); if (IS_ERR(ibqp)) kfree(qp); return ibqp; } - memset(&in_params, 0, sizeof(in_params)); - - if (udata) { - if (!(udata && ibpd->uobject && ibpd->uobject->context)) - goto err0; - - ib_ctx = ibpd->uobject->context; - ctx = get_qedr_ucontext(ib_ctx); - - memset(&ureq, 0, sizeof(ureq)); - if (ib_copy_from_udata(&ureq, udata, sizeof(ureq))) { - DP_ERR(dev, - "create qp: problem copying data from user space\n"); - goto err0; - } - - rc = qedr_init_user_qp(ib_ctx, dev, qp, &ureq); - if (rc) - goto err0; - - qedr_init_qp_user_params(&in_params, &ureq); - } else { - rc = qedr_init_kernel_qp(dev, qp, attrs, &in_params); - if (rc) - goto err0; - } - - qedr_init_qp_in_params_sq(dev, pd, qp, attrs, udata, &in_params); - qedr_init_qp_in_params_rq(qp, attrs, udata, &in_params); - - qp->qed_qp = dev->ops->rdma_create_qp(dev->rdma_ctx, - &in_params, &out_params); + if (udata) + rc = qedr_create_user_qp(dev, qp, ibpd, udata, attrs); + else + rc = qedr_create_kernel_qp(dev, qp, ibpd, attrs); - if (!qp->qed_qp) - goto err1; + if (rc) + goto err; - qp->qp_id = out_params.qp_id; - qp->icid = out_params.icid; qp->ibqp.qp_num = qp->qp_id; - if (udata) { - rc = qedr_copy_qp_uresp(dev, qp, udata); - if (rc) - goto err2; - - qedr_qp_user_print(dev, qp); - } else { - qedr_init_qp_kernel_doorbell_sq(dev, qp); - qedr_init_qp_kernel_doorbell_rq(dev, qp); - } - - DP_DEBUG(dev, QEDR_MSG_QP, "created %s space QP %p\n", - udata ? "user" : "kernel", qp); - return &qp->ibqp; -err2: - rc = dev->ops->rdma_destroy_qp(dev->rdma_ctx, qp->qed_qp); - if (rc) - DP_ERR(dev, "create qp: fatal fault. rc=%d", rc); -err1: - if (udata) { - qedr_cleanup_user_sq(dev, qp); - qedr_cleanup_user_rq(dev, qp); - } else { - qedr_cleanup_kernel_sq(dev, qp); - qedr_cleanup_kernel_rq(dev, qp); - } - -err0: +err: kfree(qp); return ERR_PTR(-EFAULT); @@ -1772,6 +1769,7 @@ int qedr_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, struct qedr_qp *qp = get_qedr_qp(ibqp); struct qed_rdma_modify_qp_in_params qp_params = { 0 }; struct qedr_dev *dev = get_qedr_dev(&qp->dev->ibdev); + const struct ib_global_route *grh = rdma_ah_read_grh(&attr->ah_attr); enum ib_qp_state old_qp_state, new_qp_state; int rc = 0; @@ -1854,17 +1852,17 @@ int qedr_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, SET_FIELD(qp_params.modify_flags, QED_ROCE_MODIFY_QP_VALID_ADDRESS_VECTOR, 1); - qp_params.traffic_class_tos = attr->ah_attr.grh.traffic_class; - qp_params.flow_label = attr->ah_attr.grh.flow_label; - qp_params.hop_limit_ttl = attr->ah_attr.grh.hop_limit; + qp_params.traffic_class_tos = grh->traffic_class; + qp_params.flow_label = grh->flow_label; + qp_params.hop_limit_ttl = grh->hop_limit; - qp->sgid_idx = attr->ah_attr.grh.sgid_index; + qp->sgid_idx = grh->sgid_index; rc = get_gid_info_from_table(ibqp, attr, attr_mask, &qp_params); if (rc) { DP_ERR(dev, "modify qp: problems with GID index %d (rc=%d)\n", - attr->ah_attr.grh.sgid_index, rc); + grh->sgid_index, rc); return rc; } @@ -2049,25 +2047,21 @@ int qedr_query_qp(struct ib_qp *ibqp, qp_attr->cap.max_inline_data = ROCE_REQ_MAX_INLINE_DATA_SIZE; qp_init_attr->cap = qp_attr->cap; - memcpy(&qp_attr->ah_attr.grh.dgid.raw[0], ¶ms.dgid.bytes[0], - sizeof(qp_attr->ah_attr.grh.dgid.raw)); - - qp_attr->ah_attr.grh.flow_label = params.flow_label; - qp_attr->ah_attr.grh.sgid_index = qp->sgid_idx; - qp_attr->ah_attr.grh.hop_limit = params.hop_limit_ttl; - qp_attr->ah_attr.grh.traffic_class = params.traffic_class_tos; - - qp_attr->ah_attr.ah_flags = IB_AH_GRH; - qp_attr->ah_attr.port_num = 1; - qp_attr->ah_attr.sl = 0; + qp_attr->ah_attr.type = RDMA_AH_ATTR_TYPE_ROCE; + rdma_ah_set_grh(&qp_attr->ah_attr, NULL, + params.flow_label, qp->sgid_idx, + params.hop_limit_ttl, params.traffic_class_tos); + rdma_ah_set_dgid_raw(&qp_attr->ah_attr, ¶ms.dgid.bytes[0]); + rdma_ah_set_port_num(&qp_attr->ah_attr, 1); + rdma_ah_set_sl(&qp_attr->ah_attr, 0); qp_attr->timeout = params.timeout; qp_attr->rnr_retry = params.rnr_retry; qp_attr->retry_cnt = params.retry_cnt; qp_attr->min_rnr_timer = params.min_rnr_nak_timer; qp_attr->pkey_index = params.pkey_index; qp_attr->port_num = 1; - qp_attr->ah_attr.src_path_bits = 0; - qp_attr->ah_attr.static_rate = 0; + rdma_ah_set_path_bits(&qp_attr->ah_attr, 0); + rdma_ah_set_static_rate(&qp_attr->ah_attr, 0); qp_attr->alt_pkey_index = 0; qp_attr->alt_port_num = 0; qp_attr->alt_timeout = 0; @@ -2085,6 +2079,24 @@ err: return rc; } +int qedr_free_qp_resources(struct qedr_dev *dev, struct qedr_qp *qp) +{ + int rc = 0; + + if (qp->qp_type != IB_QPT_GSI) { + rc = dev->ops->rdma_destroy_qp(dev->rdma_ctx, qp->qed_qp); + if (rc) + return rc; + } + + if (qp->ibqp.uobject && qp->ibqp.uobject->context) + qedr_cleanup_user(dev, qp); + else + qedr_cleanup_kernel(dev, qp); + + return 0; +} + int qedr_destroy_qp(struct ib_qp *ibqp) { struct qedr_qp *qp = get_qedr_qp(ibqp); @@ -2107,28 +2119,17 @@ int qedr_destroy_qp(struct ib_qp *ibqp) qedr_modify_qp(ibqp, &attr, attr_mask, NULL); } - if (qp->qp_type != IB_QPT_GSI) { - rc = dev->ops->rdma_destroy_qp(dev->rdma_ctx, qp->qed_qp); - if (rc) - return rc; - } else { + if (qp->qp_type == IB_QPT_GSI) qedr_destroy_gsi_qp(dev); - } - if (ibqp->uobject && ibqp->uobject->context) { - qedr_cleanup_user_sq(dev, qp); - qedr_cleanup_user_rq(dev, qp); - } else { - qedr_cleanup_kernel_sq(dev, qp); - qedr_cleanup_kernel_rq(dev, qp); - } + qedr_free_qp_resources(dev, qp); kfree(qp); return rc; } -struct ib_ah *qedr_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr, +struct ib_ah *qedr_create_ah(struct ib_pd *ibpd, struct rdma_ah_attr *attr, struct ib_udata *udata) { struct qedr_ah *ah; @@ -2182,8 +2183,8 @@ static int init_mr_info(struct qedr_dev *dev, struct mr_info *info, goto done; info->pbl_table = qedr_alloc_pbl_tbl(dev, &info->pbl_info, GFP_KERNEL); - if (!info->pbl_table) { - rc = -ENOMEM; + if (IS_ERR(info->pbl_table)) { + rc = PTR_ERR(info->pbl_table); goto done; } @@ -2194,7 +2195,7 @@ static int init_mr_info(struct qedr_dev *dev, struct mr_info *info, * list and allocating another one */ tmp = qedr_alloc_pbl_tbl(dev, &info->pbl_info, GFP_KERNEL); - if (!tmp) { + if (IS_ERR(tmp)) { DP_DEBUG(dev, QEDR_MSG_MR, "Extra PBL is not allocated\n"); goto done; } @@ -2243,7 +2244,7 @@ struct ib_mr *qedr_reg_user_mr(struct ib_pd *ibpd, u64 start, u64 len, goto err1; qedr_populate_pbls(dev, mr->umem, mr->info.pbl_table, - &mr->info.pbl_info); + &mr->info.pbl_info, mr->umem->page_shift); rc = dev->ops->rdma_alloc_tid(dev->rdma_ctx, &mr->hw_mr.itid); if (rc) { @@ -2264,7 +2265,7 @@ struct ib_mr *qedr_reg_user_mr(struct ib_pd *ibpd, u64 start, u64 len, mr->hw_mr.pbl_ptr = mr->info.pbl_table[0].pa; mr->hw_mr.pbl_two_level = mr->info.pbl_info.two_layered; mr->hw_mr.pbl_page_size_log = ilog2(mr->info.pbl_info.pbl_size); - mr->hw_mr.page_size_log = ilog2(mr->umem->page_size); + mr->hw_mr.page_size_log = mr->umem->page_shift; mr->hw_mr.fbo = ib_umem_offset(mr->umem); mr->hw_mr.length = len; mr->hw_mr.vaddr = usr_addr; @@ -2699,6 +2700,8 @@ static int qedr_prepare_reg(struct qedr_qp *qp, fwqe1->addr.lo = lower_32_bits(mr->ibmr.iova); fwqe1->l_key = wr->key; + fwqe2->access_ctrl = 0; + SET_FIELD2(fwqe2->access_ctrl, RDMA_SQ_FMR_WQE_2ND_REMOTE_READ, !!(wr->access & IB_ACCESS_REMOTE_READ)); SET_FIELD2(fwqe2->access_ctrl, RDMA_SQ_FMR_WQE_2ND_REMOTE_WRITE, @@ -3224,6 +3227,10 @@ static int process_req(struct qedr_dev *dev, struct qedr_qp *qp, case IB_WC_REG_MR: qp->wqe_wr_id[qp->sq.cons].mr->info.completed++; break; + case IB_WC_RDMA_READ: + case IB_WC_SEND: + wc->byte_len = qp->wqe_wr_id[qp->sq.cons].bytes_len; + break; default: break; } @@ -3345,57 +3352,81 @@ static int qedr_poll_cq_req(struct qedr_dev *dev, return cnt; } -static void __process_resp_one(struct qedr_dev *dev, struct qedr_qp *qp, - struct qedr_cq *cq, struct ib_wc *wc, - struct rdma_cqe_responder *resp, u64 wr_id) +static inline int qedr_cqe_resp_status_to_ib(u8 status) { - enum ib_wc_status wc_status = IB_WC_SUCCESS; - u8 flags; - - wc->opcode = IB_WC_RECV; - wc->wc_flags = 0; - - switch (resp->status) { + switch (status) { case RDMA_CQE_RESP_STS_LOCAL_ACCESS_ERR: - wc_status = IB_WC_LOC_ACCESS_ERR; - break; + return IB_WC_LOC_ACCESS_ERR; case RDMA_CQE_RESP_STS_LOCAL_LENGTH_ERR: - wc_status = IB_WC_LOC_LEN_ERR; - break; + return IB_WC_LOC_LEN_ERR; case RDMA_CQE_RESP_STS_LOCAL_QP_OPERATION_ERR: - wc_status = IB_WC_LOC_QP_OP_ERR; - break; + return IB_WC_LOC_QP_OP_ERR; case RDMA_CQE_RESP_STS_LOCAL_PROTECTION_ERR: - wc_status = IB_WC_LOC_PROT_ERR; - break; + return IB_WC_LOC_PROT_ERR; case RDMA_CQE_RESP_STS_MEMORY_MGT_OPERATION_ERR: - wc_status = IB_WC_MW_BIND_ERR; - break; + return IB_WC_MW_BIND_ERR; case RDMA_CQE_RESP_STS_REMOTE_INVALID_REQUEST_ERR: - wc_status = IB_WC_REM_INV_RD_REQ_ERR; - break; + return IB_WC_REM_INV_RD_REQ_ERR; case RDMA_CQE_RESP_STS_OK: - wc_status = IB_WC_SUCCESS; - wc->byte_len = le32_to_cpu(resp->length); + return IB_WC_SUCCESS; + default: + return IB_WC_GENERAL_ERR; + } +} + +static inline int qedr_set_ok_cqe_resp_wc(struct rdma_cqe_responder *resp, + struct ib_wc *wc) +{ + wc->status = IB_WC_SUCCESS; + wc->byte_len = le32_to_cpu(resp->length); - flags = resp->flags & QEDR_RESP_RDMA_IMM; + if (resp->flags & QEDR_RESP_IMM) { + wc->ex.imm_data = le32_to_cpu(resp->imm_data_or_inv_r_Key); + wc->wc_flags |= IB_WC_WITH_IMM; - if (flags == QEDR_RESP_RDMA_IMM) + if (resp->flags & QEDR_RESP_RDMA) wc->opcode = IB_WC_RECV_RDMA_WITH_IMM; - if (flags == QEDR_RESP_RDMA_IMM || flags == QEDR_RESP_IMM) { - wc->ex.imm_data = - le32_to_cpu(resp->imm_data_or_inv_r_Key); - wc->wc_flags |= IB_WC_WITH_IMM; - } - break; - default: - wc->status = IB_WC_GENERAL_ERR; - DP_ERR(dev, "Invalid CQE status detected\n"); + if (resp->flags & QEDR_RESP_INV) + return -EINVAL; + + } else if (resp->flags & QEDR_RESP_INV) { + wc->ex.imm_data = le32_to_cpu(resp->imm_data_or_inv_r_Key); + wc->wc_flags |= IB_WC_WITH_INVALIDATE; + + if (resp->flags & QEDR_RESP_RDMA) + return -EINVAL; + + } else if (resp->flags & QEDR_RESP_RDMA) { + return -EINVAL; } - /* fill WC */ - wc->status = wc_status; + return 0; +} + +static void __process_resp_one(struct qedr_dev *dev, struct qedr_qp *qp, + struct qedr_cq *cq, struct ib_wc *wc, + struct rdma_cqe_responder *resp, u64 wr_id) +{ + /* Must fill fields before qedr_set_ok_cqe_resp_wc() */ + wc->opcode = IB_WC_RECV; + wc->wc_flags = 0; + + if (likely(resp->status == RDMA_CQE_RESP_STS_OK)) { + if (qedr_set_ok_cqe_resp_wc(resp, wc)) + DP_ERR(dev, + "CQ %p (icid=%d) has invalid CQE responder flags=0x%x\n", + cq, cq->icid, resp->flags); + + } else { + wc->status = qedr_cqe_resp_status_to_ib(resp->status); + if (wc->status == IB_WC_GENERAL_ERR) + DP_ERR(dev, + "CQ %p (icid=%d) contains an invalid CQE status %d\n", + cq, cq->icid, resp->status); + } + + /* Fill the rest of the WC */ wc->vendor_err = 0; wc->src_qp = qp->id; wc->qp = &qp->ibqp; @@ -3490,6 +3521,13 @@ int qedr_poll_cq(struct ib_cq *ibcq, int num_entries, struct ib_wc *wc) int update = 0; int done = 0; + if (cq->destroyed) { + DP_ERR(dev, + "warning: poll was invoked after destroy for cq %p (icid=%d)\n", + cq, cq->icid); + return 0; + } + if (cq->cq_type == QEDR_CQ_TYPE_GSI) return qedr_gsi_poll_cq(ibcq, num_entries, wc); @@ -3569,14 +3607,15 @@ int qedr_port_immutable(struct ib_device *ibdev, u8 port_num, struct ib_port_attr attr; int err; - err = qedr_query_port(ibdev, port_num, &attr); + immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE | + RDMA_CORE_PORT_IBA_ROCE_UDP_ENCAP; + + err = ib_query_port(ibdev, port_num, &attr); if (err) return err; immutable->pkey_tbl_len = attr.pkey_tbl_len; immutable->gid_tbl_len = attr.gid_tbl_len; - immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE | - RDMA_CORE_PORT_IBA_ROCE_UDP_ENCAP; immutable->max_mad_size = IB_MGMT_MAD_SIZE; return 0; diff --git a/drivers/infiniband/hw/qedr/verbs.h b/drivers/infiniband/hw/qedr/verbs.h index 070677ca4d19..0f8ab49d5a1a 100644 --- a/drivers/infiniband/hw/qedr/verbs.h +++ b/drivers/infiniband/hw/qedr/verbs.h @@ -70,7 +70,7 @@ int qedr_query_qp(struct ib_qp *, struct ib_qp_attr *qp_attr, int qp_attr_mask, struct ib_qp_init_attr *); int qedr_destroy_qp(struct ib_qp *ibqp); -struct ib_ah *qedr_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr, +struct ib_ah *qedr_create_ah(struct ib_pd *ibpd, struct rdma_ah_attr *attr, struct ib_udata *udata); int qedr_destroy_ah(struct ib_ah *ibah); diff --git a/drivers/infiniband/hw/qib/qib_common.h b/drivers/infiniband/hw/qib/qib_common.h index 1d6e63eb1146..a4a1f56ce824 100644 --- a/drivers/infiniband/hw/qib/qib_common.h +++ b/drivers/infiniband/hw/qib/qib_common.h @@ -742,11 +742,7 @@ struct qib_tid_session_member { #define SIZE_OF_CRC 1 #define QIB_DEFAULT_P_KEY 0xFFFF -#define QIB_AETH_CREDIT_SHIFT 24 -#define QIB_AETH_CREDIT_MASK 0x1F -#define QIB_AETH_CREDIT_INVAL 0x1F #define QIB_PSN_MASK 0xFFFFFF -#define QIB_MSN_MASK 0xFFFFFF #define QIB_EAGER_TID_ID QLOGIC_IB_I_TID_MASK #define QIB_MULTICAST_QPN 0xFFFFFF diff --git a/drivers/infiniband/hw/qib/qib_dma.c b/drivers/infiniband/hw/qib/qib_dma.c deleted file mode 100644 index 59fe092b4b0f..000000000000 --- a/drivers/infiniband/hw/qib/qib_dma.c +++ /dev/null @@ -1,169 +0,0 @@ -/* - * Copyright (c) 2006, 2009, 2010 QLogic, Corporation. All rights reserved. - * - * 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/types.h> -#include <linux/scatterlist.h> - -#include "qib_verbs.h" - -#define BAD_DMA_ADDRESS ((u64) 0) - -/* - * The following functions implement driver specific replacements - * for the ib_dma_*() functions. - * - * These functions return kernel virtual addresses instead of - * device bus addresses since the driver uses the CPU to copy - * data instead of using hardware DMA. - */ - -static int qib_mapping_error(struct ib_device *dev, u64 dma_addr) -{ - return dma_addr == BAD_DMA_ADDRESS; -} - -static u64 qib_dma_map_single(struct ib_device *dev, void *cpu_addr, - size_t size, enum dma_data_direction direction) -{ - BUG_ON(!valid_dma_direction(direction)); - return (u64) cpu_addr; -} - -static void qib_dma_unmap_single(struct ib_device *dev, u64 addr, size_t size, - enum dma_data_direction direction) -{ - BUG_ON(!valid_dma_direction(direction)); -} - -static u64 qib_dma_map_page(struct ib_device *dev, struct page *page, - unsigned long offset, size_t size, - enum dma_data_direction direction) -{ - u64 addr; - - BUG_ON(!valid_dma_direction(direction)); - - if (offset + size > PAGE_SIZE) { - addr = BAD_DMA_ADDRESS; - goto done; - } - - addr = (u64) page_address(page); - if (addr) - addr += offset; - /* TODO: handle highmem pages */ - -done: - return addr; -} - -static void qib_dma_unmap_page(struct ib_device *dev, u64 addr, size_t size, - enum dma_data_direction direction) -{ - BUG_ON(!valid_dma_direction(direction)); -} - -static int qib_map_sg(struct ib_device *dev, struct scatterlist *sgl, - int nents, enum dma_data_direction direction) -{ - struct scatterlist *sg; - u64 addr; - int i; - int ret = nents; - - BUG_ON(!valid_dma_direction(direction)); - - for_each_sg(sgl, sg, nents, i) { - addr = (u64) page_address(sg_page(sg)); - /* TODO: handle highmem pages */ - if (!addr) { - ret = 0; - break; - } - sg->dma_address = addr + sg->offset; -#ifdef CONFIG_NEED_SG_DMA_LENGTH - sg->dma_length = sg->length; -#endif - } - return ret; -} - -static void qib_unmap_sg(struct ib_device *dev, - struct scatterlist *sg, int nents, - enum dma_data_direction direction) -{ - BUG_ON(!valid_dma_direction(direction)); -} - -static void qib_sync_single_for_cpu(struct ib_device *dev, u64 addr, - size_t size, enum dma_data_direction dir) -{ -} - -static void qib_sync_single_for_device(struct ib_device *dev, u64 addr, - size_t size, - enum dma_data_direction dir) -{ -} - -static void *qib_dma_alloc_coherent(struct ib_device *dev, size_t size, - u64 *dma_handle, gfp_t flag) -{ - struct page *p; - void *addr = NULL; - - p = alloc_pages(flag, get_order(size)); - if (p) - addr = page_address(p); - if (dma_handle) - *dma_handle = (u64) addr; - return addr; -} - -static void qib_dma_free_coherent(struct ib_device *dev, size_t size, - void *cpu_addr, u64 dma_handle) -{ - free_pages((unsigned long) cpu_addr, get_order(size)); -} - -struct ib_dma_mapping_ops qib_dma_mapping_ops = { - .mapping_error = qib_mapping_error, - .map_single = qib_dma_map_single, - .unmap_single = qib_dma_unmap_single, - .map_page = qib_dma_map_page, - .unmap_page = qib_dma_unmap_page, - .map_sg = qib_map_sg, - .unmap_sg = qib_unmap_sg, - .sync_single_for_cpu = qib_sync_single_for_cpu, - .sync_single_for_device = qib_sync_single_for_device, - .alloc_coherent = qib_dma_alloc_coherent, - .free_coherent = qib_dma_free_coherent -}; diff --git a/drivers/infiniband/hw/qib/qib_file_ops.c b/drivers/infiniband/hw/qib/qib_file_ops.c index 2d1eacf1dfed..9396c1807cc3 100644 --- a/drivers/infiniband/hw/qib/qib_file_ops.c +++ b/drivers/infiniband/hw/qib/qib_file_ops.c @@ -893,7 +893,7 @@ bail: /* * qib_file_vma_fault - handle a VMA page fault. */ -static int qib_file_vma_fault(struct vm_area_struct *vma, struct vm_fault *vmf) +static int qib_file_vma_fault(struct vm_fault *vmf) { struct page *page; diff --git a/drivers/infiniband/hw/qib/qib_fs.c b/drivers/infiniband/hw/qib/qib_fs.c index f1e66efea98a..1d940a2885c9 100644 --- a/drivers/infiniband/hw/qib/qib_fs.c +++ b/drivers/infiniband/hw/qib/qib_fs.c @@ -512,7 +512,7 @@ static int qibfs_fill_super(struct super_block *sb, void *data, int silent) unsigned long flags; int ret; - static struct tree_descr files[] = { + static const struct tree_descr files[] = { [2] = {"driver_stats", &driver_ops[0], S_IRUGO}, [3] = {"driver_stats_names", &driver_ops[1], S_IRUGO}, {""}, diff --git a/drivers/infiniband/hw/qib/qib_iba6120.c b/drivers/infiniband/hw/qib/qib_iba6120.c index 92399d3ffd15..e423b71e6ea0 100644 --- a/drivers/infiniband/hw/qib/qib_iba6120.c +++ b/drivers/infiniband/hw/qib/qib_iba6120.c @@ -707,7 +707,7 @@ static void qib_6120_clear_freeze(struct qib_devdata *dd) /* disable error interrupts, to avoid confusion */ qib_write_kreg(dd, kr_errmask, 0ULL); - /* also disable interrupts; errormask is sometimes overwriten */ + /* also disable interrupts; errormask is sometimes overwritten */ qib_6120_set_intr_state(dd, 0); qib_cancel_sends(dd->pport); @@ -3295,13 +3295,11 @@ static int init_6120_variables(struct qib_devdata *dd) dd->rhdrhead_intr_off = 1ULL << 32; /* setup the stats timer; the add_timer is done at end of init */ - init_timer(&dd->stats_timer); - dd->stats_timer.function = qib_get_6120_faststats; - dd->stats_timer.data = (unsigned long) dd; + setup_timer(&dd->stats_timer, qib_get_6120_faststats, + (unsigned long)dd); - init_timer(&dd->cspec->pma_timer); - dd->cspec->pma_timer.function = pma_6120_timer; - dd->cspec->pma_timer.data = (unsigned long) ppd; + setup_timer(&dd->cspec->pma_timer, pma_6120_timer, + (unsigned long)ppd); dd->ureg_align = qib_read_kreg32(dd, kr_palign); diff --git a/drivers/infiniband/hw/qib/qib_iba7220.c b/drivers/infiniband/hw/qib/qib_iba7220.c index e55e31a69195..c3679c48e61c 100644 --- a/drivers/infiniband/hw/qib/qib_iba7220.c +++ b/drivers/infiniband/hw/qib/qib_iba7220.c @@ -1259,7 +1259,7 @@ static void qib_7220_clear_freeze(struct qib_devdata *dd) /* disable error interrupts, to avoid confusion */ qib_write_kreg(dd, kr_errmask, 0ULL); - /* also disable interrupts; errormask is sometimes overwriten */ + /* also disable interrupts; errormask is sometimes overwritten */ qib_7220_set_intr_state(dd, 0); qib_cancel_sends(dd->pport); @@ -4074,9 +4074,8 @@ static int qib_init_7220_variables(struct qib_devdata *dd) if (!qib_mini_init) qib_write_kreg(dd, kr_rcvbthqp, QIB_KD_QP); - init_timer(&ppd->cpspec->chase_timer); - ppd->cpspec->chase_timer.function = reenable_7220_chase; - ppd->cpspec->chase_timer.data = (unsigned long)ppd; + setup_timer(&ppd->cpspec->chase_timer, reenable_7220_chase, + (unsigned long)ppd); qib_num_cfg_vls = 1; /* if any 7220's, only one VL */ diff --git a/drivers/infiniband/hw/qib/qib_iba7322.c b/drivers/infiniband/hw/qib/qib_iba7322.c index c4a3616062f1..bb2439fff8fa 100644 --- a/drivers/infiniband/hw/qib/qib_iba7322.c +++ b/drivers/infiniband/hw/qib/qib_iba7322.c @@ -2053,7 +2053,7 @@ static void qib_7322_clear_freeze(struct qib_devdata *dd) qib_write_kreg_port(dd->pport + pidx, krp_errmask, 0ULL); - /* also disable interrupts; errormask is sometimes overwriten */ + /* also disable interrupts; errormask is sometimes overwritten */ qib_7322_set_intr_state(dd, 0); /* clear the freeze, and be sure chip saw it */ @@ -2893,7 +2893,6 @@ static void qib_setup_7322_cleanup(struct qib_devdata *dd) dd->cspec->gpio_mask &= ~mask; qib_write_kreg(dd, kr_gpio_mask, dd->cspec->gpio_mask); spin_unlock_irqrestore(&dd->cspec->gpio_lock, flags); - qib_qsfp_deinit(&dd->pport[i].cpspec->qsfp_data); } } } @@ -6612,9 +6611,8 @@ static int qib_init_7322_variables(struct qib_devdata *dd) if (!qib_mini_init) write_7322_init_portregs(ppd); - init_timer(&cp->chase_timer); - cp->chase_timer.function = reenable_chase; - cp->chase_timer.data = (unsigned long)ppd; + setup_timer(&cp->chase_timer, reenable_chase, + (unsigned long)ppd); ppd++; } @@ -6640,9 +6638,8 @@ static int qib_init_7322_variables(struct qib_devdata *dd) (u64) rcv_int_count << IBA7322_HDRHEAD_PKTINT_SHIFT; /* setup the stats timer; the add_timer is done at end of init */ - init_timer(&dd->stats_timer); - dd->stats_timer.function = qib_get_7322_faststats; - dd->stats_timer.data = (unsigned long) dd; + setup_timer(&dd->stats_timer, qib_get_7322_faststats, + (unsigned long)dd); dd->ureg_align = 0x10000; /* 64KB alignment */ @@ -7069,7 +7066,7 @@ static void qib_7322_txchk_change(struct qib_devdata *dd, u32 start, unsigned long flags; while (wait) { - unsigned long shadow; + unsigned long shadow = 0; int cstart, previ = -1; /* diff --git a/drivers/infiniband/hw/qib/qib_init.c b/drivers/infiniband/hw/qib/qib_init.c index b50240b1d5a4..6c16ba1107ba 100644 --- a/drivers/infiniband/hw/qib/qib_init.c +++ b/drivers/infiniband/hw/qib/qib_init.c @@ -233,9 +233,8 @@ int qib_init_pportdata(struct qib_pportdata *ppd, struct qib_devdata *dd, spin_lock_init(&ppd->cc_shadow_lock); init_waitqueue_head(&ppd->state_wait); - init_timer(&ppd->symerr_clear_timer); - ppd->symerr_clear_timer.function = qib_clear_symerror_on_linkup; - ppd->symerr_clear_timer.data = (unsigned long)ppd; + setup_timer(&ppd->symerr_clear_timer, qib_clear_symerror_on_linkup, + (unsigned long)ppd); ppd->qib_wq = NULL; ppd->ibport_data.pmastats = @@ -429,9 +428,8 @@ static int loadtime_init(struct qib_devdata *dd) qib_get_eeprom_info(dd); /* setup time (don't start yet) to verify we got interrupt */ - init_timer(&dd->intrchk_timer); - dd->intrchk_timer.function = verify_interrupt; - dd->intrchk_timer.data = (unsigned long) dd; + setup_timer(&dd->intrchk_timer, verify_interrupt, + (unsigned long)dd); done: return ret; } @@ -755,9 +753,8 @@ done: continue; if (dd->flags & QIB_HAS_SEND_DMA) ret = qib_setup_sdma(ppd); - init_timer(&ppd->hol_timer); - ppd->hol_timer.function = qib_hol_event; - ppd->hol_timer.data = (unsigned long)ppd; + setup_timer(&ppd->hol_timer, qib_hol_event, + (unsigned long)ppd); ppd->hol_state = QIB_HOL_UP; } diff --git a/drivers/infiniband/hw/qib/qib_keys.c b/drivers/infiniband/hw/qib/qib_keys.c index 2c3c93572c17..8fdf79f8d4e4 100644 --- a/drivers/infiniband/hw/qib/qib_keys.c +++ b/drivers/infiniband/hw/qib/qib_keys.c @@ -158,10 +158,7 @@ int qib_rkey_ok(struct rvt_qp *qp, struct rvt_sge *sge, unsigned n, m; size_t off; - /* - * We use RKEY == zero for kernel virtual addresses - * (see qib_get_dma_mr and qib_dma.c). - */ + /* We use RKEY == zero for kernel virtual addresses */ rcu_read_lock(); if (rkey == 0) { struct rvt_pd *pd = ibpd_to_rvtpd(qp->ibqp.pd); diff --git a/drivers/infiniband/hw/qib/qib_mad.c b/drivers/infiniband/hw/qib/qib_mad.c index d2ac29861af5..da295e0392ed 100644 --- a/drivers/infiniband/hw/qib/qib_mad.c +++ b/drivers/infiniband/hw/qib/qib_mad.c @@ -717,9 +717,10 @@ static int subn_set_portinfo(struct ib_smp *smp, struct ib_device *ibdev, spin_lock_irqsave(&ibp->rvp.lock, flags); if (ibp->rvp.sm_ah) { if (smlid != ibp->rvp.sm_lid) - ibp->rvp.sm_ah->attr.dlid = smlid; + rdma_ah_set_dlid(&ibp->rvp.sm_ah->attr, + smlid); if (msl != ibp->rvp.sm_sl) - ibp->rvp.sm_ah->attr.sl = msl; + rdma_ah_set_sl(&ibp->rvp.sm_ah->attr, msl); } spin_unlock_irqrestore(&ibp->rvp.lock, flags); if (smlid != ibp->rvp.sm_lid) @@ -2500,5 +2501,5 @@ void qib_notify_free_mad_agent(struct rvt_dev_info *rdi, int port_idx) del_timer_sync(&dd->pport[port_idx].cong_stats.timer); if (dd->pport[port_idx].ibport_data.smi_ah) - ib_destroy_ah(&dd->pport[port_idx].ibport_data.smi_ah->ibah); + rdma_destroy_ah(&dd->pport[port_idx].ibport_data.smi_ah->ibah); } diff --git a/drivers/infiniband/hw/qib/qib_pcie.c b/drivers/infiniband/hw/qib/qib_pcie.c index 6abe1c621aa4..c379b8342a09 100644 --- a/drivers/infiniband/hw/qib/qib_pcie.c +++ b/drivers/infiniband/hw/qib/qib_pcie.c @@ -682,13 +682,6 @@ qib_pci_slot_reset(struct pci_dev *pdev) return PCI_ERS_RESULT_CAN_RECOVER; } -static pci_ers_result_t -qib_pci_link_reset(struct pci_dev *pdev) -{ - qib_devinfo(pdev, "QIB link_reset function called, ignored\n"); - return PCI_ERS_RESULT_CAN_RECOVER; -} - static void qib_pci_resume(struct pci_dev *pdev) { @@ -707,7 +700,6 @@ qib_pci_resume(struct pci_dev *pdev) const struct pci_error_handlers qib_pci_err_handler = { .error_detected = qib_pci_error_detected, .mmio_enabled = qib_pci_mmio_enabled, - .link_reset = qib_pci_link_reset, .slot_reset = qib_pci_slot_reset, .resume = qib_pci_resume, }; diff --git a/drivers/infiniband/hw/qib/qib_qp.c b/drivers/infiniband/hw/qib/qib_qp.c index 99d31efe4c2f..a343e3b5d4cb 100644 --- a/drivers/infiniband/hw/qib/qib_qp.c +++ b/drivers/infiniband/hw/qib/qib_qp.c @@ -61,43 +61,6 @@ static inline unsigned find_next_offset(struct rvt_qpn_table *qpt, return off; } -/* - * Convert the AETH credit code into the number of credits. - */ -static u32 credit_table[31] = { - 0, /* 0 */ - 1, /* 1 */ - 2, /* 2 */ - 3, /* 3 */ - 4, /* 4 */ - 6, /* 5 */ - 8, /* 6 */ - 12, /* 7 */ - 16, /* 8 */ - 24, /* 9 */ - 32, /* A */ - 48, /* B */ - 64, /* C */ - 96, /* D */ - 128, /* E */ - 192, /* F */ - 256, /* 10 */ - 384, /* 11 */ - 512, /* 12 */ - 768, /* 13 */ - 1024, /* 14 */ - 1536, /* 15 */ - 2048, /* 16 */ - 3072, /* 17 */ - 4096, /* 18 */ - 6144, /* 19 */ - 8192, /* 1A */ - 12288, /* 1B */ - 16384, /* 1C */ - 24576, /* 1D */ - 32768 /* 1E */ -}; - const struct rvt_operation_params qib_post_parms[RVT_OPERATION_MAX] = { [IB_WR_RDMA_WRITE] = { .length = sizeof(struct ib_rdma_wr), @@ -141,10 +104,9 @@ const struct rvt_operation_params qib_post_parms[RVT_OPERATION_MAX] = { }; -static void get_map_page(struct rvt_qpn_table *qpt, struct rvt_qpn_map *map, - gfp_t gfp) +static void get_map_page(struct rvt_qpn_table *qpt, struct rvt_qpn_map *map) { - unsigned long page = get_zeroed_page(gfp); + unsigned long page = get_zeroed_page(GFP_KERNEL); /* * Free the page if someone raced with us installing it. @@ -163,7 +125,7 @@ static void get_map_page(struct rvt_qpn_table *qpt, struct rvt_qpn_map *map, * zero/one for QP type IB_QPT_SMI/IB_QPT_GSI. */ int qib_alloc_qpn(struct rvt_dev_info *rdi, struct rvt_qpn_table *qpt, - enum ib_qp_type type, u8 port, gfp_t gfp) + enum ib_qp_type type, u8 port) { u32 i, offset, max_scan, qpn; struct rvt_qpn_map *map; @@ -197,7 +159,7 @@ int qib_alloc_qpn(struct rvt_dev_info *rdi, struct rvt_qpn_table *qpt, max_scan = qpt->nmaps - !offset; for (i = 0;;) { if (unlikely(!map->page)) { - get_map_page(qpt, map, gfp); + get_map_page(qpt, map); if (unlikely(!map->page)) break; } @@ -354,76 +316,16 @@ u32 qib_mtu_from_qp(struct rvt_dev_info *rdi, struct rvt_qp *qp, u32 pmtu) return ib_mtu_enum_to_int(pmtu); } -/** - * qib_compute_aeth - compute the AETH (syndrome + MSN) - * @qp: the queue pair to compute the AETH for - * - * Returns the AETH. - */ -__be32 qib_compute_aeth(struct rvt_qp *qp) -{ - u32 aeth = qp->r_msn & QIB_MSN_MASK; - - if (qp->ibqp.srq) { - /* - * Shared receive queues don't generate credits. - * Set the credit field to the invalid value. - */ - aeth |= QIB_AETH_CREDIT_INVAL << QIB_AETH_CREDIT_SHIFT; - } else { - u32 min, max, x; - u32 credits; - struct rvt_rwq *wq = qp->r_rq.wq; - u32 head; - u32 tail; - - /* sanity check pointers before trusting them */ - head = wq->head; - if (head >= qp->r_rq.size) - head = 0; - tail = wq->tail; - if (tail >= qp->r_rq.size) - tail = 0; - /* - * Compute the number of credits available (RWQEs). - * XXX Not holding the r_rq.lock here so there is a small - * chance that the pair of reads are not atomic. - */ - credits = head - tail; - if ((int)credits < 0) - credits += qp->r_rq.size; - /* - * Binary search the credit table to find the code to - * use. - */ - min = 0; - max = 31; - for (;;) { - x = (min + max) / 2; - if (credit_table[x] == credits) - break; - if (credit_table[x] > credits) - max = x; - else if (min == x) - break; - else - min = x; - } - aeth |= x << QIB_AETH_CREDIT_SHIFT; - } - return cpu_to_be32(aeth); -} - -void *qib_qp_priv_alloc(struct rvt_dev_info *rdi, struct rvt_qp *qp, gfp_t gfp) +void *qib_qp_priv_alloc(struct rvt_dev_info *rdi, struct rvt_qp *qp) { struct qib_qp_priv *priv; - priv = kzalloc(sizeof(*priv), gfp); + priv = kzalloc(sizeof(*priv), GFP_KERNEL); if (!priv) return ERR_PTR(-ENOMEM); priv->owner = qp; - priv->s_hdr = kzalloc(sizeof(*priv->s_hdr), gfp); + priv->s_hdr = kzalloc(sizeof(*priv->s_hdr), GFP_KERNEL); if (!priv->s_hdr) { kfree(priv); return ERR_PTR(-ENOMEM); @@ -448,7 +350,6 @@ void qib_stop_send_queue(struct rvt_qp *qp) struct qib_qp_priv *priv = qp->priv; cancel_work_sync(&priv->s_work); - del_timer_sync(&qp->s_timer); } void qib_quiesce_qp(struct rvt_qp *qp) @@ -474,43 +375,6 @@ void qib_flush_qp_waiters(struct rvt_qp *qp) } /** - * qib_get_credit - flush the send work queue of a QP - * @qp: the qp who's send work queue to flush - * @aeth: the Acknowledge Extended Transport Header - * - * The QP s_lock should be held. - */ -void qib_get_credit(struct rvt_qp *qp, u32 aeth) -{ - u32 credit = (aeth >> QIB_AETH_CREDIT_SHIFT) & QIB_AETH_CREDIT_MASK; - - /* - * If the credit is invalid, we can send - * as many packets as we like. Otherwise, we have to - * honor the credit field. - */ - if (credit == QIB_AETH_CREDIT_INVAL) { - if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT)) { - qp->s_flags |= RVT_S_UNLIMITED_CREDIT; - if (qp->s_flags & RVT_S_WAIT_SSN_CREDIT) { - qp->s_flags &= ~RVT_S_WAIT_SSN_CREDIT; - qib_schedule_send(qp); - } - } - } else if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT)) { - /* Compute new LSN (i.e., MSN + credit) */ - credit = (aeth + credit_table[credit]) & QIB_MSN_MASK; - if (qib_cmp24(credit, qp->s_lsn) > 0) { - qp->s_lsn = credit; - if (qp->s_flags & RVT_S_WAIT_SSN_CREDIT) { - qp->s_flags &= ~RVT_S_WAIT_SSN_CREDIT; - qib_schedule_send(qp); - } - } - } -} - -/** * qib_check_send_wqe - validate wr/wqe * @qp - The qp * @wqe - The built wqe @@ -624,7 +488,7 @@ void qib_qp_iter_print(struct seq_file *s, struct qib_qp_iter *iter) qp->s_last, qp->s_acked, qp->s_cur, qp->s_tail, qp->s_head, qp->s_size, qp->remote_qpn, - qp->remote_ah_attr.dlid); + rdma_ah_get_dlid(&qp->remote_ah_attr)); } #endif diff --git a/drivers/infiniband/hw/qib/qib_qsfp.c b/drivers/infiniband/hw/qib/qib_qsfp.c index 4c7c3c84a741..295d40a83bb6 100644 --- a/drivers/infiniband/hw/qib/qib_qsfp.c +++ b/drivers/infiniband/hw/qib/qib_qsfp.c @@ -485,16 +485,6 @@ void qib_qsfp_init(struct qib_qsfp_data *qd, dd->f_gpio_mod(dd, mask, mask, mask); } -void qib_qsfp_deinit(struct qib_qsfp_data *qd) -{ - /* - * There is nothing to do here for now. our work is scheduled - * with queue_work(), and flush_workqueue() from remove_one - * will block until all work setup with queue_work() - * completes. - */ -} - int qib_qsfp_dump(struct qib_pportdata *ppd, char *buf, int len) { struct qib_qsfp_cache cd; diff --git a/drivers/infiniband/hw/qib/qib_qsfp.h b/drivers/infiniband/hw/qib/qib_qsfp.h index 91908f533a2b..ad8dbd6ac0cf 100644 --- a/drivers/infiniband/hw/qib/qib_qsfp.h +++ b/drivers/infiniband/hw/qib/qib_qsfp.h @@ -186,4 +186,3 @@ extern int qib_refresh_qsfp_cache(struct qib_pportdata *ppd, extern int qib_qsfp_mod_present(struct qib_pportdata *ppd); extern void qib_qsfp_init(struct qib_qsfp_data *qd, void (*fevent)(struct work_struct *)); -extern void qib_qsfp_deinit(struct qib_qsfp_data *qd); diff --git a/drivers/infiniband/hw/qib/qib_rc.c b/drivers/infiniband/hw/qib/qib_rc.c index 031433cb7206..4ddbcac5eabe 100644 --- a/drivers/infiniband/hw/qib/qib_rc.c +++ b/drivers/infiniband/hw/qib/qib_rc.c @@ -38,7 +38,6 @@ /* cut down ridiculously long IB macro names */ #define OP(x) IB_OPCODE_RC_##x -static void rc_timeout(unsigned long arg); static u32 restart_sge(struct rvt_sge_state *ss, struct rvt_swqe *wqe, u32 psn, u32 pmtu) @@ -50,19 +49,10 @@ static u32 restart_sge(struct rvt_sge_state *ss, struct rvt_swqe *wqe, ss->sg_list = wqe->sg_list + 1; ss->num_sge = wqe->wr.num_sge; ss->total_len = wqe->length; - qib_skip_sge(ss, len, 0); + rvt_skip_sge(ss, len, false); return wqe->length - len; } -static void start_timer(struct rvt_qp *qp) -{ - qp->s_flags |= RVT_S_TIMER; - qp->s_timer.function = rc_timeout; - /* 4.096 usec. * (1 << qp->timeout) */ - qp->s_timer.expires = jiffies + qp->timeout_jiffies; - add_timer(&qp->s_timer); -} - /** * qib_make_rc_ack - construct a response packet (ACK, NAK, or RDMA read) * @dev: the device for this QP @@ -144,7 +134,7 @@ static int qib_make_rc_ack(struct qib_ibdev *dev, struct rvt_qp *qp, qp->s_ack_state = OP(RDMA_READ_RESPONSE_ONLY); e->sent = 1; } - ohdr->u.aeth = qib_compute_aeth(qp); + ohdr->u.aeth = rvt_compute_aeth(qp); hwords++; qp->s_ack_rdma_psn = e->psn; bth2 = qp->s_ack_rdma_psn++ & QIB_PSN_MASK; @@ -153,7 +143,7 @@ static int qib_make_rc_ack(struct qib_ibdev *dev, struct rvt_qp *qp, qp->s_cur_sge = NULL; len = 0; qp->s_ack_state = OP(ATOMIC_ACKNOWLEDGE); - ohdr->u.at.aeth = qib_compute_aeth(qp); + ohdr->u.at.aeth = rvt_compute_aeth(qp); ib_u64_put(e->atomic_data, &ohdr->u.at.atomic_ack_eth); hwords += sizeof(ohdr->u.at) / sizeof(u32); bth2 = e->psn & QIB_PSN_MASK; @@ -174,7 +164,7 @@ static int qib_make_rc_ack(struct qib_ibdev *dev, struct rvt_qp *qp, if (len > pmtu) len = pmtu; else { - ohdr->u.aeth = qib_compute_aeth(qp); + ohdr->u.aeth = rvt_compute_aeth(qp); hwords++; qp->s_ack_state = OP(RDMA_READ_RESPONSE_LAST); e = &qp->s_ack_queue[qp->s_tail_ack_queue]; @@ -197,11 +187,11 @@ normal: qp->s_cur_sge = NULL; if (qp->s_nak_state) ohdr->u.aeth = - cpu_to_be32((qp->r_msn & QIB_MSN_MASK) | + cpu_to_be32((qp->r_msn & IB_MSN_MASK) | (qp->s_nak_state << - QIB_AETH_CREDIT_SHIFT)); + IB_AETH_CREDIT_SHIFT)); else - ohdr->u.aeth = qib_compute_aeth(qp); + ohdr->u.aeth = rvt_compute_aeth(qp); hwords++; len = 0; bth0 = OP(ACKNOWLEDGE) << 24; @@ -244,7 +234,7 @@ int qib_make_rc_req(struct rvt_qp *qp, unsigned long *flags) int delta; ohdr = &priv->s_hdr->u.oth; - if (qp->remote_ah_attr.ah_flags & IB_AH_GRH) + if (rdma_ah_get_ah_flags(&qp->remote_ah_attr) & IB_AH_GRH) ohdr = &priv->s_hdr->u.l.oth; /* Sending responses has higher priority over sending requests. */ @@ -257,7 +247,7 @@ int qib_make_rc_req(struct rvt_qp *qp, unsigned long *flags) goto bail; /* We are in the error state, flush the work request. */ smp_read_barrier_depends(); /* see post_one_send() */ - if (qp->s_last == ACCESS_ONCE(qp->s_head)) + if (qp->s_last == READ_ONCE(qp->s_head)) goto bail; /* If DMAs are in progress, we can't flush immediately. */ if (atomic_read(&priv->s_dma_busy)) { @@ -303,7 +293,8 @@ int qib_make_rc_req(struct rvt_qp *qp, unsigned long *flags) newreq = 0; if (qp->s_cur == qp->s_tail) { /* Check if send work queue is empty. */ - if (qp->s_tail == qp->s_head) + smp_read_barrier_depends(); /* see post_one_send() */ + if (qp->s_tail == READ_ONCE(qp->s_head)) goto bail; /* * If a fence is requested, wait for previous @@ -330,7 +321,7 @@ int qib_make_rc_req(struct rvt_qp *qp, unsigned long *flags) case IB_WR_SEND_WITH_IMM: /* If no credit, return. */ if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT) && - qib_cmp24(wqe->ssn, qp->s_lsn + 1) > 0) { + rvt_cmp_msn(wqe->ssn, qp->s_lsn + 1) > 0) { qp->s_flags |= RVT_S_WAIT_SSN_CREDIT; goto bail; } @@ -361,7 +352,7 @@ int qib_make_rc_req(struct rvt_qp *qp, unsigned long *flags) case IB_WR_RDMA_WRITE_WITH_IMM: /* If no credit, return. */ if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT) && - qib_cmp24(wqe->ssn, qp->s_lsn + 1) > 0) { + rvt_cmp_msn(wqe->ssn, qp->s_lsn + 1) > 0) { qp->s_flags |= RVT_S_WAIT_SSN_CREDIT; goto bail; } @@ -646,9 +637,11 @@ void qib_send_rc_ack(struct rvt_qp *qp) lrh0 = QIB_LRH_BTH; /* header size in 32-bit words LRH+BTH+AETH = (8+12+4)/4. */ hwords = 6; - if (unlikely(qp->remote_ah_attr.ah_flags & IB_AH_GRH)) { + if (unlikely(rdma_ah_get_ah_flags(&qp->remote_ah_attr) & + IB_AH_GRH)) { hwords += qib_make_grh(ibp, &hdr.u.l.grh, - &qp->remote_ah_attr.grh, hwords, 0); + rdma_ah_read_grh(&qp->remote_ah_attr), + hwords, 0); ohdr = &hdr.u.l.oth; lrh0 = QIB_LRH_GRH; } @@ -657,17 +650,18 @@ void qib_send_rc_ack(struct rvt_qp *qp) if (qp->s_mig_state == IB_MIG_MIGRATED) bth0 |= IB_BTH_MIG_REQ; if (qp->r_nak_state) - ohdr->u.aeth = cpu_to_be32((qp->r_msn & QIB_MSN_MASK) | + ohdr->u.aeth = cpu_to_be32((qp->r_msn & IB_MSN_MASK) | (qp->r_nak_state << - QIB_AETH_CREDIT_SHIFT)); + IB_AETH_CREDIT_SHIFT)); else - ohdr->u.aeth = qib_compute_aeth(qp); - lrh0 |= ibp->sl_to_vl[qp->remote_ah_attr.sl] << 12 | - qp->remote_ah_attr.sl << 4; + ohdr->u.aeth = rvt_compute_aeth(qp); + lrh0 |= ibp->sl_to_vl[rdma_ah_get_sl(&qp->remote_ah_attr)] << 12 | + rdma_ah_get_sl(&qp->remote_ah_attr) << 4; hdr.lrh[0] = cpu_to_be16(lrh0); - hdr.lrh[1] = cpu_to_be16(qp->remote_ah_attr.dlid); + hdr.lrh[1] = cpu_to_be16(rdma_ah_get_dlid(&qp->remote_ah_attr)); hdr.lrh[2] = cpu_to_be16(hwords + SIZE_OF_CRC); - hdr.lrh[3] = cpu_to_be16(ppd->lid | qp->remote_ah_attr.src_path_bits); + hdr.lrh[3] = cpu_to_be16(ppd->lid | + rdma_ah_get_path_bits(&qp->remote_ah_attr)); ohdr->bth[0] = cpu_to_be32(bth0); ohdr->bth[1] = cpu_to_be32(qp->remote_qpn); ohdr->bth[2] = cpu_to_be32(qp->r_ack_psn & QIB_PSN_MASK); @@ -836,7 +830,7 @@ done: * Back up requester to resend the last un-ACKed request. * The QP r_lock and s_lock should be held and interrupts disabled. */ -static void qib_restart_rc(struct rvt_qp *qp, u32 psn, int wait) +void qib_restart_rc(struct rvt_qp *qp, u32 psn, int wait) { struct rvt_swqe *wqe = rvt_get_swqe_ptr(qp, qp->s_acked); struct qib_ibport *ibp; @@ -869,46 +863,6 @@ static void qib_restart_rc(struct rvt_qp *qp, u32 psn, int wait) } /* - * This is called from s_timer for missing responses. - */ -static void rc_timeout(unsigned long arg) -{ - struct rvt_qp *qp = (struct rvt_qp *)arg; - struct qib_ibport *ibp; - unsigned long flags; - - spin_lock_irqsave(&qp->r_lock, flags); - spin_lock(&qp->s_lock); - if (qp->s_flags & RVT_S_TIMER) { - ibp = to_iport(qp->ibqp.device, qp->port_num); - ibp->rvp.n_rc_timeouts++; - qp->s_flags &= ~RVT_S_TIMER; - del_timer(&qp->s_timer); - qib_restart_rc(qp, qp->s_last_psn + 1, 1); - qib_schedule_send(qp); - } - spin_unlock(&qp->s_lock); - spin_unlock_irqrestore(&qp->r_lock, flags); -} - -/* - * This is called from s_timer for RNR timeouts. - */ -void qib_rc_rnr_retry(unsigned long arg) -{ - struct rvt_qp *qp = (struct rvt_qp *)arg; - unsigned long flags; - - spin_lock_irqsave(&qp->s_lock, flags); - if (qp->s_flags & RVT_S_WAIT_RNR) { - qp->s_flags &= ~RVT_S_WAIT_RNR; - del_timer(&qp->s_timer); - qib_schedule_send(qp); - } - spin_unlock_irqrestore(&qp->s_lock, flags); -} - -/* * Set qp->s_sending_psn to the next PSN after the given one. * This would be psn+1 except when RDMA reads are present. */ @@ -944,7 +898,7 @@ void qib_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr) u32 opcode; u32 psn; - if (!(ib_rvt_state_ops[qp->state] & RVT_PROCESS_OR_FLUSH_SEND)) + if (!(ib_rvt_state_ops[qp->state] & RVT_SEND_OR_FLUSH_OR_RECV_OK)) return; /* Find out where the BTH is */ @@ -971,7 +925,7 @@ void qib_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr) if ((psn & IB_BTH_REQ_ACK) && qp->s_acked != qp->s_tail && !(qp->s_flags & (RVT_S_TIMER | RVT_S_WAIT_RNR | RVT_S_WAIT_PSN)) && (ib_rvt_state_ops[qp->state] & RVT_PROCESS_RECV_OK)) - start_timer(qp); + rvt_add_retry_timer(qp); while (qp->s_last != qp->s_acked) { u32 s_last; @@ -987,7 +941,10 @@ void qib_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr) /* see post_send() */ barrier(); rvt_put_swqe(wqe); - rvt_qp_swqe_complete(qp, wqe, IB_WC_SUCCESS); + rvt_qp_swqe_complete(qp, + wqe, + ib_qib_wc_opcode[wqe->wr.opcode], + IB_WC_SUCCESS); } /* * If we were waiting for sends to complete before resending, @@ -1032,7 +989,10 @@ static struct rvt_swqe *do_rc_completion(struct rvt_qp *qp, qp->s_last = s_last; /* see post_send() */ barrier(); - rvt_qp_swqe_complete(qp, wqe, IB_WC_SUCCESS); + rvt_qp_swqe_complete(qp, + wqe, + ib_qib_wc_opcode[wqe->wr.opcode], + IB_WC_SUCCESS); } else this_cpu_inc(*ibp->rvp.rc_delayed_comp); @@ -1084,12 +1044,6 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, u32 ack_psn; int diff; - /* Remove QP from retry timer */ - if (qp->s_flags & (RVT_S_TIMER | RVT_S_WAIT_RNR)) { - qp->s_flags &= ~(RVT_S_TIMER | RVT_S_WAIT_RNR); - del_timer(&qp->s_timer); - } - /* * Note that NAKs implicitly ACK outstanding SEND and RDMA write * requests and implicitly NAK RDMA read and atomic requests issued @@ -1097,7 +1051,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, * request but will include an ACK'ed request(s). */ ack_psn = psn; - if (aeth >> 29) + if (aeth >> IB_AETH_NAK_SHIFT) ack_psn--; wqe = rvt_get_swqe_ptr(qp, qp->s_acked); ibp = to_iport(qp->ibqp.device, qp->port_num); @@ -1177,7 +1131,7 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, break; } - switch (aeth >> 29) { + switch (aeth >> IB_AETH_NAK_SHIFT) { case 0: /* ACK */ this_cpu_inc(*ibp->rvp.rc_acks); if (qp->s_acked != qp->s_tail) { @@ -1185,27 +1139,30 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, * We are expecting more ACKs so * reset the retransmit timer. */ - start_timer(qp); + rvt_mod_retry_timer(qp); /* * We can stop resending the earlier packets and * continue with the next packet the receiver wants. */ if (qib_cmp24(qp->s_psn, psn) <= 0) reset_psn(qp, psn + 1); - } else if (qib_cmp24(qp->s_psn, psn) <= 0) { - qp->s_state = OP(SEND_LAST); - qp->s_psn = psn + 1; + } else { + /* No more acks - kill all timers */ + rvt_stop_rc_timers(qp); + if (qib_cmp24(qp->s_psn, psn) <= 0) { + qp->s_state = OP(SEND_LAST); + qp->s_psn = psn + 1; + } } if (qp->s_flags & RVT_S_WAIT_ACK) { qp->s_flags &= ~RVT_S_WAIT_ACK; qib_schedule_send(qp); } - qib_get_credit(qp, aeth); + rvt_get_credit(qp, aeth); qp->s_rnr_retry = qp->s_rnr_retry_cnt; qp->s_retry = qp->s_retry_cnt; update_last_psn(qp, psn); - ret = 1; - goto bail; + return 1; case 1: /* RNR NAK */ ibp->rvp.n_rnr_naks++; @@ -1228,21 +1185,17 @@ static int do_rc_ack(struct rvt_qp *qp, u32 aeth, u32 psn, int opcode, reset_psn(qp, psn); qp->s_flags &= ~(RVT_S_WAIT_SSN_CREDIT | RVT_S_WAIT_ACK); - qp->s_flags |= RVT_S_WAIT_RNR; - qp->s_timer.function = qib_rc_rnr_retry; - qp->s_timer.expires = jiffies + usecs_to_jiffies( - ib_qib_rnr_table[(aeth >> QIB_AETH_CREDIT_SHIFT) & - QIB_AETH_CREDIT_MASK]); - add_timer(&qp->s_timer); - goto bail; + rvt_stop_rc_timers(qp); + rvt_add_rnr_timer(qp, aeth); + return 0; case 3: /* NAK */ if (qp->s_acked == qp->s_tail) goto bail; /* The last valid PSN is the previous PSN. */ update_last_psn(qp, psn - 1); - switch ((aeth >> QIB_AETH_CREDIT_SHIFT) & - QIB_AETH_CREDIT_MASK) { + switch ((aeth >> IB_AETH_CREDIT_SHIFT) & + IB_AETH_CREDIT_MASK) { case 0: /* PSN sequence error */ ibp->rvp.n_seq_naks++; /* @@ -1290,6 +1243,7 @@ reserved: } bail: + rvt_stop_rc_timers(qp); return ret; } @@ -1303,10 +1257,7 @@ static void rdma_seq_err(struct rvt_qp *qp, struct qib_ibport *ibp, u32 psn, struct rvt_swqe *wqe; /* Remove QP from retry timer */ - if (qp->s_flags & (RVT_S_TIMER | RVT_S_WAIT_RNR)) { - qp->s_flags &= ~(RVT_S_TIMER | RVT_S_WAIT_RNR); - del_timer(&qp->s_timer); - } + rvt_stop_rc_timers(qp); wqe = rvt_get_swqe_ptr(qp, qp->s_acked); @@ -1390,7 +1341,7 @@ static void qib_rc_rcv_resp(struct qib_ibport *ibp, /* Ignore invalid responses. */ smp_read_barrier_depends(); /* see post_one_send */ - if (qib_cmp24(psn, ACCESS_ONCE(qp->s_next_psn)) >= 0) + if (qib_cmp24(psn, READ_ONCE(qp->s_next_psn)) >= 0) goto ack_done; /* Ignore duplicate responses. */ @@ -1399,8 +1350,8 @@ static void qib_rc_rcv_resp(struct qib_ibport *ibp, /* Update credits for "ghost" ACKs */ if (diff == 0 && opcode == OP(ACKNOWLEDGE)) { aeth = be32_to_cpu(ohdr->u.aeth); - if ((aeth >> 29) == 0) - qib_get_credit(qp, aeth); + if ((aeth >> IB_AETH_NAK_SHIFT) == 0) + rvt_get_credit(qp, aeth); } goto ack_done; } @@ -1461,8 +1412,7 @@ read_middle: * We got a response so update the timeout. * 4.096 usec. * (1 << qp->timeout) */ - qp->s_flags |= RVT_S_TIMER; - mod_timer(&qp->s_timer, jiffies + qp->timeout_jiffies); + rvt_mod_retry_timer(qp); if (qp->s_flags & RVT_S_WAIT_ACK) { qp->s_flags &= ~RVT_S_WAIT_ACK; qib_schedule_send(qp); @@ -1764,25 +1714,6 @@ send_ack: return 0; } -void qib_rc_error(struct rvt_qp *qp, enum ib_wc_status err) -{ - unsigned long flags; - int lastwqe; - - spin_lock_irqsave(&qp->s_lock, flags); - lastwqe = rvt_error_qp(qp, err); - spin_unlock_irqrestore(&qp->s_lock, flags); - - if (lastwqe) { - struct ib_event ev; - - ev.device = qp->ibqp.device; - ev.element.qp = &qp->ibqp; - ev.event = IB_EVENT_QP_LAST_WQE_REACHED; - qp->ibqp.event_handler(&ev, qp->ibqp.qp_context); - } -} - static inline void qib_update_ack_queue(struct rvt_qp *qp, unsigned n) { unsigned next; @@ -1894,17 +1825,8 @@ void qib_rc_rcv(struct qib_ctxtdata *rcd, struct ib_header *hdr, break; } - if (qp->state == IB_QPS_RTR && !(qp->r_flags & RVT_R_COMM_EST)) { - qp->r_flags |= RVT_R_COMM_EST; - if (qp->ibqp.event_handler) { - struct ib_event ev; - - ev.device = qp->ibqp.device; - ev.element.qp = &qp->ibqp; - ev.event = IB_EVENT_COMM_EST; - qp->ibqp.event_handler(&ev, qp->ibqp.qp_context); - } - } + if (qp->state == IB_QPS_RTR && !(qp->r_flags & RVT_R_COMM_EST)) + rvt_comm_est(qp); /* OK, process the packet. */ switch (opcode) { @@ -1985,8 +1907,8 @@ send_last: wc.opcode = IB_WC_RECV; wc.qp = &qp->ibqp; wc.src_qp = qp->remote_qpn; - wc.slid = qp->remote_ah_attr.dlid; - wc.sl = qp->remote_ah_attr.sl; + wc.slid = rdma_ah_get_dlid(&qp->remote_ah_attr); + wc.sl = rdma_ah_get_sl(&qp->remote_ah_attr); /* zero fields that are N/A */ wc.vendor_err = 0; wc.pkey_index = 0; @@ -2034,8 +1956,10 @@ send_last: ret = qib_get_rwqe(qp, 1); if (ret < 0) goto nack_op_err; - if (!ret) + if (!ret) { + rvt_put_ss(&qp->r_sge); goto rnr_nak; + } wc.ex.imm_data = ohdr->u.rc.imm_data; hdrsize += 4; wc.wc_flags = IB_WC_WITH_IMM; @@ -2196,7 +2120,7 @@ rnr_nak: return; nack_op_err: - qib_rc_error(qp, IB_WC_LOC_QP_OP_ERR); + rvt_rc_error(qp, IB_WC_LOC_QP_OP_ERR); qp->r_nak_state = IB_NAK_REMOTE_OPERATIONAL_ERROR; qp->r_ack_psn = qp->r_psn; /* Queue NAK for later */ @@ -2210,7 +2134,7 @@ nack_op_err: nack_inv_unlck: spin_unlock_irqrestore(&qp->s_lock, flags); nack_inv: - qib_rc_error(qp, IB_WC_LOC_QP_OP_ERR); + rvt_rc_error(qp, IB_WC_LOC_QP_OP_ERR); qp->r_nak_state = IB_NAK_INVALID_REQUEST; qp->r_ack_psn = qp->r_psn; /* Queue NAK for later */ @@ -2224,7 +2148,7 @@ nack_inv: nack_acc_unlck: spin_unlock_irqrestore(&qp->s_lock, flags); nack_acc: - qib_rc_error(qp, IB_WC_LOC_PROT_ERR); + rvt_rc_error(qp, IB_WC_LOC_PROT_ERR); qp->r_nak_state = IB_NAK_REMOTE_ACCESS_ERROR; qp->r_ack_psn = qp->r_psn; send_ack: diff --git a/drivers/infiniband/hw/qib/qib_ruc.c b/drivers/infiniband/hw/qib/qib_ruc.c index e54a2feeeb10..bd09de7c6e56 100644 --- a/drivers/infiniband/hw/qib/qib_ruc.c +++ b/drivers/infiniband/hw/qib/qib_ruc.c @@ -38,44 +38,6 @@ #include "qib_mad.h" /* - * Convert the AETH RNR timeout code into the number of microseconds. - */ -const u32 ib_qib_rnr_table[32] = { - 655360, /* 00: 655.36 */ - 10, /* 01: .01 */ - 20, /* 02 .02 */ - 30, /* 03: .03 */ - 40, /* 04: .04 */ - 60, /* 05: .06 */ - 80, /* 06: .08 */ - 120, /* 07: .12 */ - 160, /* 08: .16 */ - 240, /* 09: .24 */ - 320, /* 0A: .32 */ - 480, /* 0B: .48 */ - 640, /* 0C: .64 */ - 960, /* 0D: .96 */ - 1280, /* 0E: 1.28 */ - 1920, /* 0F: 1.92 */ - 2560, /* 10: 2.56 */ - 3840, /* 11: 3.84 */ - 5120, /* 12: 5.12 */ - 7680, /* 13: 7.68 */ - 10240, /* 14: 10.24 */ - 15360, /* 15: 15.36 */ - 20480, /* 16: 20.48 */ - 30720, /* 17: 30.72 */ - 40960, /* 18: 40.96 */ - 61440, /* 19: 61.44 */ - 81920, /* 1A: 81.92 */ - 122880, /* 1B: 122.88 */ - 163840, /* 1C: 163.84 */ - 245760, /* 1D: 245.76 */ - 327680, /* 1E: 327.68 */ - 491520 /* 1F: 491.52 */ -}; - -/* * Validate a RWQE and fill in the SGE state. * Return 1 if OK. */ @@ -233,7 +195,7 @@ void qib_migrate_qp(struct rvt_qp *qp) qp->s_mig_state = IB_MIG_MIGRATED; qp->remote_ah_attr = qp->alt_ah_attr; - qp->port_num = qp->alt_ah_attr.port_num; + qp->port_num = rdma_ah_get_port_num(&qp->alt_ah_attr); qp->s_pkey_index = qp->s_alt_pkey_index; ev.device = qp->ibqp.device; @@ -273,18 +235,23 @@ int qib_ruc_check_hdr(struct qib_ibport *ibp, struct ib_header *hdr, if (qp->s_mig_state == IB_MIG_ARMED && (bth0 & IB_BTH_MIG_REQ)) { if (!has_grh) { - if (qp->alt_ah_attr.ah_flags & IB_AH_GRH) + if (rdma_ah_get_ah_flags(&qp->alt_ah_attr) & + IB_AH_GRH) goto err; } else { - if (!(qp->alt_ah_attr.ah_flags & IB_AH_GRH)) + const struct ib_global_route *grh; + + if (!(rdma_ah_get_ah_flags(&qp->alt_ah_attr) & + IB_AH_GRH)) goto err; - guid = get_sguid(ibp, qp->alt_ah_attr.grh.sgid_index); + grh = rdma_ah_read_grh(&qp->alt_ah_attr); + guid = get_sguid(ibp, grh->sgid_index); if (!gid_ok(&hdr->u.l.grh.dgid, ibp->rvp.gid_prefix, guid)) goto err; if (!gid_ok(&hdr->u.l.grh.sgid, - qp->alt_ah_attr.grh.dgid.global.subnet_prefix, - qp->alt_ah_attr.grh.dgid.global.interface_id)) + grh->dgid.global.subnet_prefix, + grh->dgid.global.interface_id)) goto err; } if (!qib_pkey_ok((u16)bth0, @@ -297,27 +264,33 @@ int qib_ruc_check_hdr(struct qib_ibport *ibp, struct ib_header *hdr, goto err; } /* Validate the SLID. See Ch. 9.6.1.5 and 17.2.8 */ - if (be16_to_cpu(hdr->lrh[3]) != qp->alt_ah_attr.dlid || - ppd_from_ibp(ibp)->port != qp->alt_ah_attr.port_num) + if ((be16_to_cpu(hdr->lrh[3]) != + rdma_ah_get_dlid(&qp->alt_ah_attr)) || + ppd_from_ibp(ibp)->port != + rdma_ah_get_port_num(&qp->alt_ah_attr)) goto err; spin_lock_irqsave(&qp->s_lock, flags); qib_migrate_qp(qp); spin_unlock_irqrestore(&qp->s_lock, flags); } else { if (!has_grh) { - if (qp->remote_ah_attr.ah_flags & IB_AH_GRH) + if (rdma_ah_get_ah_flags(&qp->remote_ah_attr) & + IB_AH_GRH) goto err; } else { - if (!(qp->remote_ah_attr.ah_flags & IB_AH_GRH)) + const struct ib_global_route *grh; + + if (!(rdma_ah_get_ah_flags(&qp->remote_ah_attr) & + IB_AH_GRH)) goto err; - guid = get_sguid(ibp, - qp->remote_ah_attr.grh.sgid_index); + grh = rdma_ah_read_grh(&qp->remote_ah_attr); + guid = get_sguid(ibp, grh->sgid_index); if (!gid_ok(&hdr->u.l.grh.dgid, ibp->rvp.gid_prefix, guid)) goto err; if (!gid_ok(&hdr->u.l.grh.sgid, - qp->remote_ah_attr.grh.dgid.global.subnet_prefix, - qp->remote_ah_attr.grh.dgid.global.interface_id)) + grh->dgid.global.subnet_prefix, + grh->dgid.global.interface_id)) goto err; } if (!qib_pkey_ok((u16)bth0, @@ -330,7 +303,8 @@ int qib_ruc_check_hdr(struct qib_ibport *ibp, struct ib_header *hdr, goto err; } /* Validate the SLID. See Ch. 9.6.1.5 */ - if (be16_to_cpu(hdr->lrh[3]) != qp->remote_ah_attr.dlid || + if (be16_to_cpu(hdr->lrh[3]) != + rdma_ah_get_dlid(&qp->remote_ah_attr) || ppd_from_ibp(ibp)->port != qp->port_num) goto err; if (qp->s_mig_state == IB_MIG_REARM && @@ -566,8 +540,8 @@ again: wc.byte_len = wqe->length; wc.qp = &qp->ibqp; wc.src_qp = qp->remote_qpn; - wc.slid = qp->remote_ah_attr.dlid; - wc.sl = qp->remote_ah_attr.sl; + wc.slid = rdma_ah_get_dlid(&qp->remote_ah_attr); + wc.sl = rdma_ah_get_sl(&qp->remote_ah_attr); wc.port_num = 1; /* Signal completion event if the solicited bit is set. */ rvt_cq_enter(ibcq_to_rvtcq(qp->ibqp.recv_cq), &wc, @@ -599,11 +573,8 @@ rnr_nak: spin_lock_irqsave(&sqp->s_lock, flags); if (!(ib_rvt_state_ops[sqp->state] & RVT_PROCESS_RECV_OK)) goto clr_busy; - sqp->s_flags |= RVT_S_WAIT_RNR; - sqp->s_timer.function = qib_rc_rnr_retry; - sqp->s_timer.expires = jiffies + - usecs_to_jiffies(ib_qib_rnr_table[qp->r_min_rnr_timer]); - add_timer(&sqp->s_timer); + rvt_add_rnr_timer(sqp, qp->r_min_rnr_timer << + IB_AETH_CREDIT_SHIFT); goto clr_busy; op_err: @@ -621,7 +592,7 @@ acc_err: wc.status = IB_WC_LOC_PROT_ERR; err: /* responder goes to error state */ - qib_rc_error(qp, wc.status); + rvt_rc_error(qp, wc.status); serr: spin_lock_irqsave(&sqp->s_lock, flags); @@ -660,7 +631,7 @@ done: * Return the size of the header in 32 bit words. */ u32 qib_make_grh(struct qib_ibport *ibp, struct ib_grh *hdr, - struct ib_global_route *grh, u32 hwords, u32 nwords) + const struct ib_global_route *grh, u32 hwords, u32 nwords) { hdr->version_tclass_flow = cpu_to_be32((IB_GRH_VERSION << IB_GRH_VERSION_SHIFT) | @@ -693,20 +664,23 @@ void qib_make_ruc_header(struct rvt_qp *qp, struct ib_other_headers *ohdr, extra_bytes = -qp->s_cur_size & 3; nwords = (qp->s_cur_size + extra_bytes) >> 2; lrh0 = QIB_LRH_BTH; - if (unlikely(qp->remote_ah_attr.ah_flags & IB_AH_GRH)) { - qp->s_hdrwords += qib_make_grh(ibp, &priv->s_hdr->u.l.grh, - &qp->remote_ah_attr.grh, - qp->s_hdrwords, nwords); + if (unlikely(rdma_ah_get_ah_flags(&qp->remote_ah_attr) & IB_AH_GRH)) { + qp->s_hdrwords += + qib_make_grh(ibp, &priv->s_hdr->u.l.grh, + rdma_ah_read_grh(&qp->remote_ah_attr), + qp->s_hdrwords, nwords); lrh0 = QIB_LRH_GRH; } - lrh0 |= ibp->sl_to_vl[qp->remote_ah_attr.sl] << 12 | - qp->remote_ah_attr.sl << 4; + lrh0 |= ibp->sl_to_vl[rdma_ah_get_sl(&qp->remote_ah_attr)] << 12 | + rdma_ah_get_sl(&qp->remote_ah_attr) << 4; priv->s_hdr->lrh[0] = cpu_to_be16(lrh0); - priv->s_hdr->lrh[1] = cpu_to_be16(qp->remote_ah_attr.dlid); + priv->s_hdr->lrh[1] = + cpu_to_be16(rdma_ah_get_dlid(&qp->remote_ah_attr)); priv->s_hdr->lrh[2] = cpu_to_be16(qp->s_hdrwords + nwords + SIZE_OF_CRC); - priv->s_hdr->lrh[3] = cpu_to_be16(ppd_from_ibp(ibp)->lid | - qp->remote_ah_attr.src_path_bits); + priv->s_hdr->lrh[3] = + cpu_to_be16(ppd_from_ibp(ibp)->lid | + rdma_ah_get_path_bits(&qp->remote_ah_attr)); bth0 |= qib_get_pkey(ibp, qp->s_pkey_index); bth0 |= extra_bytes << 20; if (qp->s_mig_state == IB_MIG_MIGRATED) @@ -744,7 +718,8 @@ void qib_do_send(struct rvt_qp *qp) if ((qp->ibqp.qp_type == IB_QPT_RC || qp->ibqp.qp_type == IB_QPT_UC) && - (qp->remote_ah_attr.dlid & ~((1 << ppd->lmc) - 1)) == ppd->lid) { + (rdma_ah_get_dlid(&qp->remote_ah_attr) & + ~((1 << ppd->lmc) - 1)) == ppd->lid) { qib_ruc_loopback(qp); return; } @@ -810,7 +785,10 @@ void qib_send_complete(struct rvt_qp *qp, struct rvt_swqe *wqe, qp->ibqp.qp_type == IB_QPT_GSI) atomic_dec(&ibah_to_rvtah(wqe->ud_wr.ah)->refcount); - rvt_qp_swqe_complete(qp, wqe, status); + rvt_qp_swqe_complete(qp, + wqe, + ib_qib_wc_opcode[wqe->wr.opcode], + status); if (qp->s_acked == old_last) qp->s_acked = last; diff --git a/drivers/infiniband/hw/qib/qib_uc.c b/drivers/infiniband/hw/qib/qib_uc.c index 5b2d483451ad..498e2202e72c 100644 --- a/drivers/infiniband/hw/qib/qib_uc.c +++ b/drivers/infiniband/hw/qib/qib_uc.c @@ -74,7 +74,7 @@ int qib_make_uc_req(struct rvt_qp *qp, unsigned long *flags) } ohdr = &priv->s_hdr->u.oth; - if (qp->remote_ah_attr.ah_flags & IB_AH_GRH) + if (rdma_ah_get_ah_flags(&qp->remote_ah_attr) & IB_AH_GRH) ohdr = &priv->s_hdr->u.l.oth; /* header size in 32-bit words LRH+BTH = (8+12)/4. */ @@ -325,17 +325,8 @@ inv: goto inv; } - if (qp->state == IB_QPS_RTR && !(qp->r_flags & RVT_R_COMM_EST)) { - qp->r_flags |= RVT_R_COMM_EST; - if (qp->ibqp.event_handler) { - struct ib_event ev; - - ev.device = qp->ibqp.device; - ev.element.qp = &qp->ibqp; - ev.event = IB_EVENT_COMM_EST; - qp->ibqp.event_handler(&ev, qp->ibqp.qp_context); - } - } + if (qp->state == IB_QPS_RTR && !(qp->r_flags & RVT_R_COMM_EST)) + rvt_comm_est(qp); /* OK, process the packet. */ switch (opcode) { @@ -403,8 +394,8 @@ last_imm: wc.status = IB_WC_SUCCESS; wc.qp = &qp->ibqp; wc.src_qp = qp->remote_qpn; - wc.slid = qp->remote_ah_attr.dlid; - wc.sl = qp->remote_ah_attr.sl; + wc.slid = rdma_ah_get_dlid(&qp->remote_ah_attr); + wc.sl = rdma_ah_get_sl(&qp->remote_ah_attr); /* zero fields that are N/A */ wc.vendor_err = 0; wc.pkey_index = 0; @@ -527,7 +518,7 @@ drop: return; op_err: - qib_rc_error(qp, IB_WC_LOC_QP_OP_ERR); + rvt_rc_error(qp, IB_WC_LOC_QP_OP_ERR); return; } diff --git a/drivers/infiniband/hw/qib/qib_ud.c b/drivers/infiniband/hw/qib/qib_ud.c index f45cad1198b0..341a123ee95c 100644 --- a/drivers/infiniband/hw/qib/qib_ud.c +++ b/drivers/infiniband/hw/qib/qib_ud.c @@ -54,7 +54,7 @@ static void qib_ud_loopback(struct rvt_qp *sqp, struct rvt_swqe *swqe) struct qib_devdata *dd = ppd->dd; struct rvt_dev_info *rdi = &dd->verbs_dev.rdi; struct rvt_qp *qp; - struct ib_ah_attr *ah_attr; + struct rdma_ah_attr *ah_attr; unsigned long flags; struct rvt_sge_state ssge; struct rvt_sge *sge; @@ -92,13 +92,13 @@ static void qib_ud_loopback(struct rvt_qp *sqp, struct rvt_swqe *swqe) pkey1 = qib_get_pkey(ibp, sqp->s_pkey_index); pkey2 = qib_get_pkey(ibp, qp->s_pkey_index); if (unlikely(!qib_pkey_ok(pkey1, pkey2))) { - lid = ppd->lid | (ah_attr->src_path_bits & + lid = ppd->lid | (rdma_ah_get_path_bits(ah_attr) & ((1 << ppd->lmc) - 1)); qib_bad_pqkey(ibp, IB_NOTICE_TRAP_BAD_PKEY, pkey1, - ah_attr->sl, + rdma_ah_get_sl(ah_attr), sqp->ibqp.qp_num, qp->ibqp.qp_num, cpu_to_be16(lid), - cpu_to_be16(ah_attr->dlid)); + cpu_to_be16(rdma_ah_get_dlid(ah_attr))); goto drop; } } @@ -116,13 +116,13 @@ static void qib_ud_loopback(struct rvt_qp *sqp, struct rvt_swqe *swqe) if (unlikely(qkey != qp->qkey)) { u16 lid; - lid = ppd->lid | (ah_attr->src_path_bits & + lid = ppd->lid | (rdma_ah_get_path_bits(ah_attr) & ((1 << ppd->lmc) - 1)); qib_bad_pqkey(ibp, IB_NOTICE_TRAP_BAD_QKEY, qkey, - ah_attr->sl, + rdma_ah_get_sl(ah_attr), sqp->ibqp.qp_num, qp->ibqp.qp_num, cpu_to_be16(lid), - cpu_to_be16(ah_attr->dlid)); + cpu_to_be16(rdma_ah_get_dlid(ah_attr))); goto drop; } } @@ -152,7 +152,7 @@ static void qib_ud_loopback(struct rvt_qp *sqp, struct rvt_swqe *swqe) ret = qib_get_rwqe(qp, 0); if (ret < 0) { - qib_rc_error(qp, IB_WC_LOC_QP_OP_ERR); + rvt_rc_error(qp, IB_WC_LOC_QP_OP_ERR); goto bail_unlock; } if (!ret) { @@ -168,16 +168,16 @@ static void qib_ud_loopback(struct rvt_qp *sqp, struct rvt_swqe *swqe) goto bail_unlock; } - if (ah_attr->ah_flags & IB_AH_GRH) { + if (rdma_ah_get_ah_flags(ah_attr) & IB_AH_GRH) { struct ib_grh grh; - struct ib_global_route grd = ah_attr->grh; + const struct ib_global_route *grd = rdma_ah_read_grh(ah_attr); - qib_make_grh(ibp, &grh, &grd, 0, 0); + qib_make_grh(ibp, &grh, grd, 0, 0); qib_copy_sge(&qp->r_sge, &grh, sizeof(grh), 1); wc.wc_flags |= IB_WC_GRH; } else - qib_skip_sge(&qp->r_sge, sizeof(struct ib_grh), 1); + rvt_skip_sge(&qp->r_sge, sizeof(struct ib_grh), true); ssge.sg_list = swqe->sg_list + 1; ssge.sge = *swqe->sg_list; ssge.num_sge = swqe->wr.num_sge; @@ -220,9 +220,10 @@ static void qib_ud_loopback(struct rvt_qp *sqp, struct rvt_swqe *swqe) wc.src_qp = sqp->ibqp.qp_num; wc.pkey_index = qp->ibqp.qp_type == IB_QPT_GSI ? swqe->ud_wr.pkey_index : 0; - wc.slid = ppd->lid | (ah_attr->src_path_bits & ((1 << ppd->lmc) - 1)); - wc.sl = ah_attr->sl; - wc.dlid_path_bits = ah_attr->dlid & ((1 << ppd->lmc) - 1); + wc.slid = ppd->lid | (rdma_ah_get_path_bits(ah_attr) & + ((1 << ppd->lmc) - 1)); + wc.sl = rdma_ah_get_sl(ah_attr); + wc.dlid_path_bits = rdma_ah_get_dlid(ah_attr) & ((1 << ppd->lmc) - 1); wc.port_num = qp->port_num; /* Signal completion event if the solicited bit is set. */ rvt_cq_enter(ibcq_to_rvtcq(qp->ibqp.recv_cq), &wc, @@ -246,7 +247,7 @@ int qib_make_ud_req(struct rvt_qp *qp, unsigned long *flags) { struct qib_qp_priv *priv = qp->priv; struct ib_other_headers *ohdr; - struct ib_ah_attr *ah_attr; + struct rdma_ah_attr *ah_attr; struct qib_pportdata *ppd; struct qib_ibport *ibp; struct rvt_swqe *wqe; @@ -289,14 +290,15 @@ int qib_make_ud_req(struct rvt_qp *qp, unsigned long *flags) ibp = to_iport(qp->ibqp.device, qp->port_num); ppd = ppd_from_ibp(ibp); ah_attr = &ibah_to_rvtah(wqe->ud_wr.ah)->attr; - if (ah_attr->dlid >= be16_to_cpu(IB_MULTICAST_LID_BASE)) { - if (ah_attr->dlid != be16_to_cpu(IB_LID_PERMISSIVE)) + if (rdma_ah_get_dlid(ah_attr) >= be16_to_cpu(IB_MULTICAST_LID_BASE)) { + if (rdma_ah_get_dlid(ah_attr) != + be16_to_cpu(IB_LID_PERMISSIVE)) this_cpu_inc(ibp->pmastats->n_multicast_xmit); else this_cpu_inc(ibp->pmastats->n_unicast_xmit); } else { this_cpu_inc(ibp->pmastats->n_unicast_xmit); - lid = ah_attr->dlid & ~((1 << ppd->lmc) - 1); + lid = rdma_ah_get_dlid(ah_attr) & ~((1 << ppd->lmc) - 1); if (unlikely(lid == ppd->lid)) { unsigned long tflags = *flags; /* @@ -328,17 +330,17 @@ int qib_make_ud_req(struct rvt_qp *qp, unsigned long *flags) qp->s_hdrwords = 7; qp->s_cur_size = wqe->length; qp->s_cur_sge = &qp->s_sge; - qp->s_srate = ah_attr->static_rate; + qp->s_srate = rdma_ah_get_static_rate(ah_attr); qp->s_wqe = wqe; qp->s_sge.sge = wqe->sg_list[0]; qp->s_sge.sg_list = wqe->sg_list + 1; qp->s_sge.num_sge = wqe->wr.num_sge; qp->s_sge.total_len = wqe->length; - if (ah_attr->ah_flags & IB_AH_GRH) { + if (rdma_ah_get_ah_flags(ah_attr) & IB_AH_GRH) { /* Header size in 32-bit words. */ qp->s_hdrwords += qib_make_grh(ibp, &priv->s_hdr->u.l.grh, - &ah_attr->grh, + rdma_ah_read_grh(ah_attr), qp->s_hdrwords, nwords); lrh0 = QIB_LRH_GRH; ohdr = &priv->s_hdr->u.l.oth; @@ -357,18 +359,20 @@ int qib_make_ud_req(struct rvt_qp *qp, unsigned long *flags) bth0 = IB_OPCODE_UD_SEND_ONLY_WITH_IMMEDIATE << 24; } else bth0 = IB_OPCODE_UD_SEND_ONLY << 24; - lrh0 |= ah_attr->sl << 4; + lrh0 |= rdma_ah_get_sl(ah_attr) << 4; if (qp->ibqp.qp_type == IB_QPT_SMI) lrh0 |= 0xF000; /* Set VL (see ch. 13.5.3.1) */ else - lrh0 |= ibp->sl_to_vl[ah_attr->sl] << 12; + lrh0 |= ibp->sl_to_vl[rdma_ah_get_sl(ah_attr)] << 12; priv->s_hdr->lrh[0] = cpu_to_be16(lrh0); - priv->s_hdr->lrh[1] = cpu_to_be16(ah_attr->dlid); /* DEST LID */ + priv->s_hdr->lrh[1] = + cpu_to_be16(rdma_ah_get_dlid(ah_attr)); /* DEST LID */ priv->s_hdr->lrh[2] = cpu_to_be16(qp->s_hdrwords + nwords + SIZE_OF_CRC); lid = ppd->lid; if (lid) { - lid |= ah_attr->src_path_bits & ((1 << ppd->lmc) - 1); + lid |= rdma_ah_get_path_bits(ah_attr) & + ((1 << ppd->lmc) - 1); priv->s_hdr->lrh[3] = cpu_to_be16(lid); } else priv->s_hdr->lrh[3] = IB_LID_PERMISSIVE; @@ -382,8 +386,9 @@ int qib_make_ud_req(struct rvt_qp *qp, unsigned long *flags) /* * Use the multicast QP if the destination LID is a multicast LID. */ - ohdr->bth[1] = ah_attr->dlid >= be16_to_cpu(IB_MULTICAST_LID_BASE) && - ah_attr->dlid != be16_to_cpu(IB_LID_PERMISSIVE) ? + ohdr->bth[1] = rdma_ah_get_dlid(ah_attr) >= + be16_to_cpu(IB_MULTICAST_LID_BASE) && + rdma_ah_get_dlid(ah_attr) != be16_to_cpu(IB_LID_PERMISSIVE) ? cpu_to_be32(QIB_MULTICAST_QPN) : cpu_to_be32(wqe->ud_wr.remote_qpn); ohdr->bth[2] = cpu_to_be32(wqe->psn & QIB_PSN_MASK); @@ -548,7 +553,7 @@ void qib_ud_rcv(struct qib_ibport *ibp, struct ib_header *hdr, ret = qib_get_rwqe(qp, 0); if (ret < 0) { - qib_rc_error(qp, IB_WC_LOC_QP_OP_ERR); + rvt_rc_error(qp, IB_WC_LOC_QP_OP_ERR); return; } if (!ret) { @@ -567,7 +572,7 @@ void qib_ud_rcv(struct qib_ibport *ibp, struct ib_header *hdr, sizeof(struct ib_grh), 1); wc.wc_flags |= IB_WC_GRH; } else - qib_skip_sge(&qp->r_sge, sizeof(struct ib_grh), 1); + rvt_skip_sge(&qp->r_sge, sizeof(struct ib_grh), true); qib_copy_sge(&qp->r_sge, data, wc.byte_len - sizeof(struct ib_grh), 1); rvt_put_ss(&qp->r_sge); if (!test_and_clear_bit(RVT_R_WRID_VALID, &qp->r_aflags)) diff --git a/drivers/infiniband/hw/qib/qib_user_pages.c b/drivers/infiniband/hw/qib/qib_user_pages.c index 75f08624ac05..ce83ba9a12ef 100644 --- a/drivers/infiniband/hw/qib/qib_user_pages.c +++ b/drivers/infiniband/hw/qib/qib_user_pages.c @@ -32,6 +32,7 @@ */ #include <linux/mm.h> +#include <linux/sched/signal.h> #include <linux/device.h> #include "qib.h" diff --git a/drivers/infiniband/hw/qib/qib_user_sdma.c b/drivers/infiniband/hw/qib/qib_user_sdma.c index 3e0677c51276..926f3c8eba69 100644 --- a/drivers/infiniband/hw/qib/qib_user_sdma.c +++ b/drivers/infiniband/hw/qib/qib_user_sdma.c @@ -144,8 +144,8 @@ qib_user_sdma_rb_search(struct rb_root *root, pid_t pid) struct rb_node *node = root->rb_node; while (node) { - sdma_rb_node = container_of(node, - struct qib_user_sdma_rb_node, node); + sdma_rb_node = rb_entry(node, struct qib_user_sdma_rb_node, + node); if (pid < sdma_rb_node->pid) node = node->rb_left; else if (pid > sdma_rb_node->pid) @@ -164,7 +164,7 @@ qib_user_sdma_rb_insert(struct rb_root *root, struct qib_user_sdma_rb_node *new) struct qib_user_sdma_rb_node *got; while (*node) { - got = container_of(*node, struct qib_user_sdma_rb_node, node); + got = rb_entry(*node, struct qib_user_sdma_rb_node, node); parent = *node; if (new->pid < got->pid) node = &((*node)->rb_left); diff --git a/drivers/infiniband/hw/qib/qib_verbs.c b/drivers/infiniband/hw/qib/qib_verbs.c index 4b54c0ddd08a..ac42dce7e281 100644 --- a/drivers/infiniband/hw/qib/qib_verbs.c +++ b/drivers/infiniband/hw/qib/qib_verbs.c @@ -114,6 +114,19 @@ module_param_named(disable_sma, ib_qib_disable_sma, uint, S_IWUSR | S_IRUGO); MODULE_PARM_DESC(disable_sma, "Disable the SMA"); /* + * Translate ib_wr_opcode into ib_wc_opcode. + */ +const enum ib_wc_opcode ib_qib_wc_opcode[] = { + [IB_WR_RDMA_WRITE] = IB_WC_RDMA_WRITE, + [IB_WR_RDMA_WRITE_WITH_IMM] = IB_WC_RDMA_WRITE, + [IB_WR_SEND] = IB_WC_SEND, + [IB_WR_SEND_WITH_IMM] = IB_WC_SEND, + [IB_WR_RDMA_READ] = IB_WC_RDMA_READ, + [IB_WR_ATOMIC_CMP_AND_SWP] = IB_WC_COMP_SWAP, + [IB_WR_ATOMIC_FETCH_AND_ADD] = IB_WC_FETCH_ADD +}; + +/* * System image GUID. */ __be64 ib_qib_sys_image_guid; @@ -129,78 +142,16 @@ void qib_copy_sge(struct rvt_sge_state *ss, void *data, u32 length, int release) struct rvt_sge *sge = &ss->sge; while (length) { - u32 len = sge->length; + u32 len = rvt_get_sge_length(sge, length); - if (len > length) - len = length; - if (len > sge->sge_length) - len = sge->sge_length; - BUG_ON(len == 0); + WARN_ON_ONCE(len == 0); memcpy(sge->vaddr, data, len); - sge->vaddr += len; - sge->length -= len; - sge->sge_length -= len; - if (sge->sge_length == 0) { - if (release) - rvt_put_mr(sge->mr); - if (--ss->num_sge) - *sge = *ss->sg_list++; - } else if (sge->length == 0 && sge->mr->lkey) { - if (++sge->n >= RVT_SEGSZ) { - if (++sge->m >= sge->mr->mapsz) - break; - sge->n = 0; - } - sge->vaddr = - sge->mr->map[sge->m]->segs[sge->n].vaddr; - sge->length = - sge->mr->map[sge->m]->segs[sge->n].length; - } + rvt_update_sge(ss, len, release); data += len; length -= len; } } -/** - * qib_skip_sge - skip over SGE memory - XXX almost dup of prev func - * @ss: the SGE state - * @length: the number of bytes to skip - */ -void qib_skip_sge(struct rvt_sge_state *ss, u32 length, int release) -{ - struct rvt_sge *sge = &ss->sge; - - while (length) { - u32 len = sge->length; - - if (len > length) - len = length; - if (len > sge->sge_length) - len = sge->sge_length; - BUG_ON(len == 0); - sge->vaddr += len; - sge->length -= len; - sge->sge_length -= len; - if (sge->sge_length == 0) { - if (release) - rvt_put_mr(sge->mr); - if (--ss->num_sge) - *sge = *ss->sg_list++; - } else if (sge->length == 0 && sge->mr->lkey) { - if (++sge->n >= RVT_SEGSZ) { - if (++sge->m >= sge->mr->mapsz) - break; - sge->n = 0; - } - sge->vaddr = - sge->mr->map[sge->m]->segs[sge->n].vaddr; - sge->length = - sge->mr->map[sge->m]->segs[sge->n].length; - } - length -= len; - } -} - /* * Count the number of DMA descriptors needed to send length bytes of data. * Don't modify the qib_sge_state to get the count. @@ -405,7 +356,7 @@ void qib_ib_rcv(struct qib_ctxtdata *rcd, void *rhdr, void *data, u32 tlen) if (lnh != QIB_LRH_GRH) goto drop; - mcast = rvt_mcast_find(&ibp->rvp, &hdr->u.l.grh.dgid); + mcast = rvt_mcast_find(&ibp->rvp, &hdr->u.l.grh.dgid, lid); if (mcast == NULL) goto drop; this_cpu_inc(ibp->pmastats->n_multicast_rcv); @@ -468,27 +419,6 @@ static void mem_timer(unsigned long data) } } -static void update_sge(struct rvt_sge_state *ss, u32 length) -{ - struct rvt_sge *sge = &ss->sge; - - sge->vaddr += length; - sge->length -= length; - sge->sge_length -= length; - if (sge->sge_length == 0) { - if (--ss->num_sge) - *sge = *ss->sg_list++; - } else if (sge->length == 0 && sge->mr->lkey) { - if (++sge->n >= RVT_SEGSZ) { - if (++sge->m >= sge->mr->mapsz) - return; - sge->n = 0; - } - sge->vaddr = sge->mr->map[sge->m]->segs[sge->n].vaddr; - sge->length = sge->mr->map[sge->m]->segs[sge->n].length; - } -} - #ifdef __LITTLE_ENDIAN static inline u32 get_upper_bits(u32 data, u32 shift) { @@ -646,11 +576,11 @@ static void copy_io(u32 __iomem *piobuf, struct rvt_sge_state *ss, data = clear_upper_bytes(v, extra, 0); } } - update_sge(ss, len); + rvt_update_sge(ss, len, false); length -= len; } /* Update address before sending packet. */ - update_sge(ss, length); + rvt_update_sge(ss, length, false); if (flush_wc) { /* must flush early everything before trigger word */ qib_flush_wc(); @@ -1069,7 +999,7 @@ static int qib_verbs_send_pio(struct rvt_qp *qp, struct ib_header *ibhdr, u32 *addr = (u32 *) ss->sge.vaddr; /* Update address before sending packet. */ - update_sge(ss, len); + rvt_update_sge(ss, len, false); if (flush_wc) { qib_pio_copy(piobuf, addr, dwords - 1); /* must flush early everything before trigger word */ @@ -1303,6 +1233,7 @@ static int qib_query_port(struct rvt_dev_info *rdi, u8 port_num, enum ib_mtu mtu; u16 lid = ppd->lid; + /* props being zeroed by the caller, avoid zeroing it here */ props->lid = lid ? lid : be16_to_cpu(IB_LID_PERMISSIVE); props->lmc = ppd->lmc; props->state = dd->f_iblink_state(ppd->lastibcstat); @@ -1405,16 +1336,16 @@ static int qib_get_guid_be(struct rvt_dev_info *rdi, struct rvt_ibport *rvp, return 0; } -int qib_check_ah(struct ib_device *ibdev, struct ib_ah_attr *ah_attr) +int qib_check_ah(struct ib_device *ibdev, struct rdma_ah_attr *ah_attr) { - if (ah_attr->sl > 15) + if (rdma_ah_get_sl(ah_attr) > 15) return -EINVAL; return 0; } static void qib_notify_new_ah(struct ib_device *ibdev, - struct ib_ah_attr *ah_attr, + struct rdma_ah_attr *ah_attr, struct rvt_ah *ah) { struct qib_ibport *ibp; @@ -1425,25 +1356,29 @@ static void qib_notify_new_ah(struct ib_device *ibdev, * done being setup. We can however modify things which we need to set. */ - ibp = to_iport(ibdev, ah_attr->port_num); + ibp = to_iport(ibdev, rdma_ah_get_port_num(ah_attr)); ppd = ppd_from_ibp(ibp); - ah->vl = ibp->sl_to_vl[ah->attr.sl]; + ah->vl = ibp->sl_to_vl[rdma_ah_get_sl(&ah->attr)]; ah->log_pmtu = ilog2(ppd->ibmtu); } struct ib_ah *qib_create_qp0_ah(struct qib_ibport *ibp, u16 dlid) { - struct ib_ah_attr attr; + struct rdma_ah_attr attr; struct ib_ah *ah = ERR_PTR(-EINVAL); struct rvt_qp *qp0; + struct qib_pportdata *ppd = ppd_from_ibp(ibp); + struct qib_devdata *dd = dd_from_ppd(ppd); + u8 port_num = ppd->port; memset(&attr, 0, sizeof(attr)); - attr.dlid = dlid; - attr.port_num = ppd_from_ibp(ibp)->port; + attr.type = rdma_ah_find_type(&dd->verbs_dev.rdi.ibdev, port_num); + rdma_ah_set_dlid(&attr, dlid); + rdma_ah_set_port_num(&attr, port_num); rcu_read_lock(); qp0 = rcu_dereference(ibp->rvp.qp[0]); if (qp0) - ah = ib_create_ah(qp0->ibqp.pd, &attr); + ah = rdma_create_ah(qp0->ibqp.pd, &attr); rcu_read_unlock(); return ah; } @@ -1632,7 +1567,7 @@ int qib_register_ib_device(struct qib_devdata *dd) ibdev->owner = THIS_MODULE; ibdev->node_guid = ppd->guid; ibdev->phys_port_cnt = dd->num_pports; - ibdev->dma_device = &dd->pcidev->dev; + ibdev->dev.parent = &dd->pcidev->dev; ibdev->modify_device = qib_modify_device; ibdev->process_mad = qib_process_mad; @@ -1659,6 +1594,7 @@ int qib_register_ib_device(struct qib_devdata *dd) dd->verbs_dev.rdi.driver_f.stop_send_queue = qib_stop_send_queue; dd->verbs_dev.rdi.driver_f.flush_qp_waiters = qib_flush_qp_waiters; dd->verbs_dev.rdi.driver_f.notify_error_qp = qib_notify_error_qp; + dd->verbs_dev.rdi.driver_f.notify_restart_rc = qib_restart_rc; dd->verbs_dev.rdi.driver_f.mtu_to_path_mtu = qib_mtu_to_path_mtu; dd->verbs_dev.rdi.driver_f.mtu_from_qp = qib_mtu_from_qp; dd->verbs_dev.rdi.driver_f.get_pmtu_from_attr = qib_get_pmtu_from_attr; diff --git a/drivers/infiniband/hw/qib/qib_verbs.h b/drivers/infiniband/hw/qib/qib_verbs.h index 94fd30fdedac..a52fc67b40d7 100644 --- a/drivers/infiniband/hw/qib/qib_verbs.h +++ b/drivers/infiniband/hw/qib/qib_verbs.h @@ -270,18 +270,16 @@ int qib_snapshot_counters(struct qib_pportdata *ppd, u64 *swords, int qib_get_counters(struct qib_pportdata *ppd, struct qib_verbs_counters *cntrs); -__be32 qib_compute_aeth(struct rvt_qp *qp); - /* * Functions provided by qib driver for rdmavt to use */ unsigned qib_free_all_qps(struct rvt_dev_info *rdi); -void *qib_qp_priv_alloc(struct rvt_dev_info *rdi, struct rvt_qp *qp, gfp_t gfp); +void *qib_qp_priv_alloc(struct rvt_dev_info *rdi, struct rvt_qp *qp); void qib_qp_priv_free(struct rvt_dev_info *rdi, struct rvt_qp *qp); void qib_notify_qp_reset(struct rvt_qp *qp); int qib_alloc_qpn(struct rvt_dev_info *rdi, struct rvt_qpn_table *qpt, - enum ib_qp_type type, u8 port, gfp_t gfp); - + enum ib_qp_type type, u8 port); +void qib_restart_rc(struct rvt_qp *qp, u32 psn, int wait); #ifdef CONFIG_DEBUG_FS struct qib_qp_iter; @@ -294,8 +292,6 @@ void qib_qp_iter_print(struct seq_file *s, struct qib_qp_iter *iter); #endif -void qib_get_credit(struct rvt_qp *qp, u32 aeth); - unsigned qib_pkt_delay(u32 plen, u8 snd_mult, u8 rcv_mult); void qib_verbs_sdma_desc_avail(struct qib_pportdata *ppd, unsigned avail); @@ -308,15 +304,13 @@ int qib_verbs_send(struct rvt_qp *qp, struct ib_header *hdr, void qib_copy_sge(struct rvt_sge_state *ss, void *data, u32 length, int release); -void qib_skip_sge(struct rvt_sge_state *ss, u32 length, int release); - void qib_uc_rcv(struct qib_ibport *ibp, struct ib_header *hdr, int has_grh, void *data, u32 tlen, struct rvt_qp *qp); void qib_rc_rcv(struct qib_ctxtdata *rcd, struct ib_header *hdr, int has_grh, void *data, u32 tlen, struct rvt_qp *qp); -int qib_check_ah(struct ib_device *ibdev, struct ib_ah_attr *ah_attr); +int qib_check_ah(struct ib_device *ibdev, struct rdma_ah_attr *ah_attr); int qib_check_send_wqe(struct rvt_qp *qp, struct rvt_swqe *wqe); @@ -326,8 +320,6 @@ void qib_rc_rnr_retry(unsigned long arg); void qib_rc_send_complete(struct rvt_qp *qp, struct ib_header *hdr); -void qib_rc_error(struct rvt_qp *qp, enum ib_wc_status err); - int qib_post_ud_send(struct rvt_qp *qp, struct ib_send_wr *wr); void qib_ud_rcv(struct qib_ibport *ibp, struct ib_header *hdr, @@ -343,7 +335,7 @@ int qib_ruc_check_hdr(struct qib_ibport *ibp, struct ib_header *hdr, int has_grh, struct rvt_qp *qp, u32 bth0); u32 qib_make_grh(struct qib_ibport *ibp, struct ib_grh *hdr, - struct ib_global_route *grh, u32 hwords, u32 nwords); + const struct ib_global_route *grh, u32 hwords, u32 nwords); void qib_make_ruc_header(struct rvt_qp *qp, struct ib_other_headers *ohdr, u32 bth0, u32 bth2); diff --git a/drivers/infiniband/hw/usnic/usnic_common_pkt_hdr.h b/drivers/infiniband/hw/usnic/usnic_common_pkt_hdr.h index 596e0ed49a8e..bf7d197a9f42 100644 --- a/drivers/infiniband/hw/usnic/usnic_common_pkt_hdr.h +++ b/drivers/infiniband/hw/usnic/usnic_common_pkt_hdr.h @@ -34,7 +34,6 @@ #ifndef USNIC_CMN_PKT_HDR_H #define USNIC_CMN_PKT_HDR_H -#define USNIC_ROCE_ETHERTYPE (0x8915) #define USNIC_ROCE_GRH_VER (8) #define USNIC_PROTO_VER (1) #define USNIC_ROCE_GRH_VER_SHIFT (4) diff --git a/drivers/infiniband/hw/usnic/usnic_common_util.h b/drivers/infiniband/hw/usnic/usnic_common_util.h index b54986de5f0c..ddd81294fa46 100644 --- a/drivers/infiniband/hw/usnic/usnic_common_util.h +++ b/drivers/infiniband/hw/usnic/usnic_common_util.h @@ -34,21 +34,7 @@ #ifndef USNIC_CMN_UTIL_H #define USNIC_CMN_UTIL_H -static inline void -usnic_mac_to_gid(const char *const mac, char *raw_gid) -{ - raw_gid[0] = 0xfe; - raw_gid[1] = 0x80; - memset(&raw_gid[2], 0, 6); - raw_gid[8] = mac[0]^2; - raw_gid[9] = mac[1]; - raw_gid[10] = mac[2]; - raw_gid[11] = 0xff; - raw_gid[12] = 0xfe; - raw_gid[13] = mac[3]; - raw_gid[14] = mac[4]; - raw_gid[15] = mac[5]; -} +#include <net/addrconf.h> static inline void usnic_mac_ip_to_gid(const char *const mac, const __be32 inaddr, char *raw_gid) @@ -57,27 +43,7 @@ usnic_mac_ip_to_gid(const char *const mac, const __be32 inaddr, char *raw_gid) raw_gid[1] = 0x80; memset(&raw_gid[2], 0, 2); memcpy(&raw_gid[4], &inaddr, 4); - raw_gid[8] = mac[0]^2; - raw_gid[9] = mac[1]; - raw_gid[10] = mac[2]; - raw_gid[11] = 0xff; - raw_gid[12] = 0xfe; - raw_gid[13] = mac[3]; - raw_gid[14] = mac[4]; - raw_gid[15] = mac[5]; -} - -static inline void -usnic_write_gid_if_id_from_mac(char *mac, char *raw_gid) -{ - raw_gid[8] = mac[0]^2; - raw_gid[9] = mac[1]; - raw_gid[10] = mac[2]; - raw_gid[11] = 0xff; - raw_gid[12] = 0xfe; - raw_gid[13] = mac[3]; - raw_gid[14] = mac[4]; - raw_gid[15] = mac[5]; + addrconf_addr_eui48(&raw_gid[8], mac); } #endif /* USNIC_COMMON_UTIL_H */ diff --git a/drivers/infiniband/hw/usnic/usnic_fwd.h b/drivers/infiniband/hw/usnic/usnic_fwd.h index 3a8add9ddf46..b2ac22be0731 100644 --- a/drivers/infiniband/hw/usnic/usnic_fwd.h +++ b/drivers/infiniband/hw/usnic/usnic_fwd.h @@ -36,6 +36,7 @@ #include <linux/if.h> #include <linux/netdevice.h> +#include <linux/if_ether.h> #include <linux/pci.h> #include <linux/in.h> @@ -97,7 +98,7 @@ static inline void usnic_fwd_init_usnic_filter(struct filter *filter, uint32_t usnic_id) { filter->type = FILTER_USNIC_ID; - filter->u.usnic.ethtype = USNIC_ROCE_ETHERTYPE; + filter->u.usnic.ethtype = ETH_P_IBOE; filter->u.usnic.flags = FILTER_FIELD_USNIC_ETHTYPE | FILTER_FIELD_USNIC_ID | FILTER_FIELD_USNIC_PROTO; diff --git a/drivers/infiniband/hw/usnic/usnic_ib_main.c b/drivers/infiniband/hw/usnic/usnic_ib_main.c index 0a89a955550b..c0c1e8b027b1 100644 --- a/drivers/infiniband/hw/usnic/usnic_ib_main.c +++ b/drivers/infiniband/hw/usnic/usnic_ib_main.c @@ -321,7 +321,9 @@ static int usnic_port_immutable(struct ib_device *ibdev, u8 port_num, struct ib_port_attr attr; int err; - err = usnic_ib_query_port(ibdev, port_num, &attr); + immutable->core_cap_flags = RDMA_CORE_PORT_USNIC; + + err = ib_query_port(ibdev, port_num, &attr); if (err) return err; @@ -380,7 +382,7 @@ static void *usnic_ib_device_add(struct pci_dev *dev) us_ibdev->ib_dev.node_type = RDMA_NODE_USNIC_UDP; us_ibdev->ib_dev.phys_port_cnt = USNIC_IB_PORT_CNT; us_ibdev->ib_dev.num_comp_vectors = USNIC_IB_NUM_COMP_VECTORS; - us_ibdev->ib_dev.dma_device = &dev->dev; + us_ibdev->ib_dev.dev.parent = &dev->dev; us_ibdev->ib_dev.uverbs_abi_ver = USNIC_UVERBS_ABI_VERSION; strlcpy(us_ibdev->ib_dev.name, "usnic_%d", IB_DEVICE_NAME_MAX); diff --git a/drivers/infiniband/hw/usnic/usnic_ib_sysfs.c b/drivers/infiniband/hw/usnic/usnic_ib_sysfs.c index 80ef3f8998c8..32956f9f5715 100644 --- a/drivers/infiniband/hw/usnic/usnic_ib_sysfs.c +++ b/drivers/infiniband/hw/usnic/usnic_ib_sysfs.c @@ -44,6 +44,7 @@ #include "usnic_vnic.h" #include "usnic_ib_verbs.h" #include "usnic_log.h" +#include "usnic_ib_sysfs.h" static ssize_t usnic_ib_show_board(struct device *device, struct device_attribute *attr, @@ -80,7 +81,7 @@ usnic_ib_show_config(struct device *device, struct device_attribute *attr, left = PAGE_SIZE; mutex_lock(&us_ibdev->usdev_lock); - if (atomic_read(&us_ibdev->vf_cnt.refcount) > 0) { + if (kref_read(&us_ibdev->vf_cnt) > 0) { char *busname; /* @@ -99,7 +100,7 @@ usnic_ib_show_config(struct device *device, struct device_attribute *attr, PCI_FUNC(us_ibdev->pdev->devfn), netdev_name(us_ibdev->netdev), us_ibdev->ufdev->mac, - atomic_read(&us_ibdev->vf_cnt.refcount)); + kref_read(&us_ibdev->vf_cnt)); UPDATE_PTR_LEFT(n, ptr, left); for (res_type = USNIC_VNIC_RES_TYPE_EOL; @@ -147,7 +148,7 @@ usnic_ib_show_max_vf(struct device *device, struct device_attribute *attr, us_ibdev = container_of(device, struct usnic_ib_dev, ib_dev.dev); return scnprintf(buf, PAGE_SIZE, "%u\n", - atomic_read(&us_ibdev->vf_cnt.refcount)); + kref_read(&us_ibdev->vf_cnt)); } static ssize_t diff --git a/drivers/infiniband/hw/usnic/usnic_ib_verbs.c b/drivers/infiniband/hw/usnic/usnic_ib_verbs.c index 74819a7951e2..4996984885c2 100644 --- a/drivers/infiniband/hw/usnic/usnic_ib_verbs.c +++ b/drivers/infiniband/hw/usnic/usnic_ib_verbs.c @@ -46,6 +46,7 @@ #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 @@ -151,7 +152,7 @@ find_free_vf_and_create_qp_grp(struct usnic_ib_dev *us_ibdev, struct usnic_vnic *vnic; struct usnic_ib_qp_grp *qp_grp; struct device *dev, **dev_list; - int i, found = 0; + int i; BUG_ON(!mutex_is_locked(&us_ibdev->usdev_lock)); @@ -173,8 +174,13 @@ find_free_vf_and_create_qp_grp(struct usnic_ib_dev *us_ibdev, us_ibdev->ib_dev.name, pci_name(usnic_vnic_get_pdev( vnic))); - found = 1; - break; + qp_grp = usnic_ib_qp_grp_create(us_ibdev->ufdev, + vf, pd, + res_spec, + trans_spec); + + spin_unlock(&vf->lock); + goto qp_grp_check; } spin_unlock(&vf->lock); @@ -182,34 +188,30 @@ find_free_vf_and_create_qp_grp(struct usnic_ib_dev *us_ibdev, usnic_uiom_free_dev_list(dev_list); } - if (!found) { - /* Try to find resources on an unused vf */ - list_for_each_entry(vf, &us_ibdev->vf_dev_list, link) { - spin_lock(&vf->lock); - vnic = vf->vnic; - if (vf->qp_grp_ref_cnt == 0 && - usnic_vnic_check_room(vnic, res_spec) == 0) { - found = 1; - break; - } + /* Try to find resources on an unused vf */ + list_for_each_entry(vf, &us_ibdev->vf_dev_list, link) { + spin_lock(&vf->lock); + vnic = vf->vnic; + if (vf->qp_grp_ref_cnt == 0 && + usnic_vnic_check_room(vnic, res_spec) == 0) { + qp_grp = usnic_ib_qp_grp_create(us_ibdev->ufdev, vf, + pd, res_spec, + trans_spec); + spin_unlock(&vf->lock); + goto qp_grp_check; } + spin_unlock(&vf->lock); } - if (!found) { - usnic_info("No free qp grp found on %s\n", - us_ibdev->ib_dev.name); - return ERR_PTR(-ENOMEM); - } + usnic_info("No free qp grp found on %s\n", us_ibdev->ib_dev.name); + return ERR_PTR(-ENOMEM); - qp_grp = usnic_ib_qp_grp_create(us_ibdev->ufdev, vf, pd, res_spec, - trans_spec); - spin_unlock(&vf->lock); +qp_grp_check: if (IS_ERR_OR_NULL(qp_grp)) { usnic_err("Failed to allocate qp_grp\n"); return ERR_PTR(qp_grp ? PTR_ERR(qp_grp) : -ENOMEM); } - return qp_grp; } @@ -291,11 +293,11 @@ int usnic_ib_query_device(struct ib_device *ibdev, qp_per_vf = max(us_ibdev->vf_res_cnt[USNIC_VNIC_RES_TYPE_WQ], us_ibdev->vf_res_cnt[USNIC_VNIC_RES_TYPE_RQ]); props->max_qp = qp_per_vf * - atomic_read(&us_ibdev->vf_cnt.refcount); + kref_read(&us_ibdev->vf_cnt); props->device_cap_flags = IB_DEVICE_PORT_ACTIVE_EVENT | IB_DEVICE_SYS_IMAGE_GUID | IB_DEVICE_BLOCK_MULTICAST_LOOPBACK; props->max_cq = us_ibdev->vf_res_cnt[USNIC_VNIC_RES_TYPE_CQ] * - atomic_read(&us_ibdev->vf_cnt.refcount); + kref_read(&us_ibdev->vf_cnt); props->max_pd = USNIC_UIOM_MAX_PD_CNT; props->max_mr = USNIC_UIOM_MAX_MR_CNT; props->local_ca_ack_delay = 0; @@ -330,7 +332,7 @@ int usnic_ib_query_port(struct ib_device *ibdev, u8 port, mutex_lock(&us_ibdev->usdev_lock); __ethtool_get_link_ksettings(us_ibdev->netdev, &cmd); - memset(props, 0, sizeof(*props)); + /* props being zeroed by the caller, avoid zeroing it here */ props->lid = 0; props->lmc = 1; @@ -738,7 +740,7 @@ int usnic_ib_mmap(struct ib_ucontext *context, /* In ib callbacks section - Start of stub funcs */ struct ib_ah *usnic_ib_create_ah(struct ib_pd *pd, - struct ib_ah_attr *ah_attr, + struct rdma_ah_attr *ah_attr, struct ib_udata *udata) { diff --git a/drivers/infiniband/hw/usnic/usnic_ib_verbs.h b/drivers/infiniband/hw/usnic/usnic_ib_verbs.h index 0ed8e072329e..172e43b6fa95 100644 --- a/drivers/infiniband/hw/usnic/usnic_ib_verbs.h +++ b/drivers/infiniband/hw/usnic/usnic_ib_verbs.h @@ -75,7 +75,7 @@ int usnic_ib_dealloc_ucontext(struct ib_ucontext *ibcontext); int usnic_ib_mmap(struct ib_ucontext *context, struct vm_area_struct *vma); struct ib_ah *usnic_ib_create_ah(struct ib_pd *pd, - struct ib_ah_attr *ah_attr, + struct rdma_ah_attr *ah_attr, struct ib_udata *udata); int usnic_ib_destroy_ah(struct ib_ah *ah); diff --git a/drivers/infiniband/hw/usnic/usnic_uiom.c b/drivers/infiniband/hw/usnic/usnic_uiom.c index 1ccee6ea5bc3..c49db7c33979 100644 --- a/drivers/infiniband/hw/usnic/usnic_uiom.c +++ b/drivers/infiniband/hw/usnic/usnic_uiom.c @@ -34,7 +34,8 @@ #include <linux/mm.h> #include <linux/dma-mapping.h> -#include <linux/sched.h> +#include <linux/sched/signal.h> +#include <linux/sched/mm.h> #include <linux/hugetlb.h> #include <linux/iommu.h> #include <linux/workqueue.h> diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma.h b/drivers/infiniband/hw/vmw_pvrdma/pvrdma.h index 71e1d55d69d6..8e2f0a11690f 100644 --- a/drivers/infiniband/hw/vmw_pvrdma/pvrdma.h +++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma.h @@ -69,6 +69,9 @@ */ #define PCI_DEVICE_ID_VMWARE_PVRDMA 0x0820 +#define PVRDMA_NUM_RING_PAGES 4 +#define PVRDMA_QP_NUM_HEADER_PAGES 1 + struct pvrdma_dev; struct pvrdma_page_dir { @@ -196,13 +199,7 @@ struct pvrdma_dev { spinlock_t cmd_lock; /* Command lock. */ struct semaphore cmd_sema; struct completion cmd_done; - struct { - enum pvrdma_intr_type type; /* Intr type */ - struct msix_entry msix_entry[PVRDMA_MAX_INTERRUPTS]; - irq_handler_t handler[PVRDMA_MAX_INTERRUPTS]; - u8 enabled[PVRDMA_MAX_INTERRUPTS]; - u8 size; - } intr; + unsigned int nr_vectors; /* RDMA-related device information. */ union ib_gid *sgid_tbl; @@ -443,10 +440,10 @@ void pvrdma_global_route_to_ib(struct ib_global_route *dst, const struct pvrdma_global_route *src); void ib_global_route_to_pvrdma(struct pvrdma_global_route *dst, const struct ib_global_route *src); -void pvrdma_ah_attr_to_ib(struct ib_ah_attr *dst, - const struct pvrdma_ah_attr *src); -void ib_ah_attr_to_pvrdma(struct pvrdma_ah_attr *dst, - const struct ib_ah_attr *src); +void pvrdma_ah_attr_to_rdma(struct rdma_ah_attr *dst, + const struct pvrdma_ah_attr *src); +void rdma_ah_attr_to_pvrdma(struct pvrdma_ah_attr *dst, + const struct rdma_ah_attr *src); int pvrdma_uar_table_init(struct pvrdma_dev *dev); void pvrdma_uar_table_cleanup(struct pvrdma_dev *dev); diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_cq.c b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_cq.c index e429ca5b16aa..69bda611d313 100644 --- a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_cq.c +++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_cq.c @@ -373,7 +373,7 @@ retry: wc->sl = cqe->sl; wc->dlid_path_bits = cqe->dlid_path_bits; wc->port_num = cqe->port_num; - wc->vendor_err = 0; + wc->vendor_err = cqe->vendor_err; /* Update shared ring state */ pvrdma_idx_ring_inc(&cq->ring_state->rx.cons_head, cq->ibcq.cqe); diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_dev_api.h b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_dev_api.h index c06768635d65..09078ccfaec7 100644 --- a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_dev_api.h +++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_dev_api.h @@ -132,7 +132,7 @@ enum pvrdma_pci_resource { enum pvrdma_device_ctl { PVRDMA_DEVICE_CTL_ACTIVATE, /* Activate device. */ - PVRDMA_DEVICE_CTL_QUIESCE, /* Quiesce device. */ + PVRDMA_DEVICE_CTL_UNQUIESCE, /* Unquiesce device. */ PVRDMA_DEVICE_CTL_RESET, /* Reset device. */ }; @@ -149,12 +149,6 @@ enum pvrdma_intr_cause { PVRDMA_INTR_CAUSE_CQ = (1 << PVRDMA_INTR_VECTOR_CQ), }; -enum pvrdma_intr_type { - PVRDMA_INTR_TYPE_INTX, /* Legacy. */ - PVRDMA_INTR_TYPE_MSI, /* MSI. */ - PVRDMA_INTR_TYPE_MSIX, /* MSI-X. */ -}; - enum pvrdma_gos_bits { PVRDMA_GOS_BITS_UNK, /* Unknown. */ PVRDMA_GOS_BITS_32, /* 32-bit. */ diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_main.c b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_main.c index bd8fbd3d2032..34ebc7615411 100644 --- a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_main.c +++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_main.c @@ -56,7 +56,7 @@ #include "pvrdma.h" #define DRV_NAME "vmw_pvrdma" -#define DRV_VERSION "1.0.0.0-k" +#define DRV_VERSION "1.0.1.0-k" static DEFINE_MUTEX(pvrdma_device_list_lock); static LIST_HEAD(pvrdma_device_list); @@ -132,13 +132,14 @@ static int pvrdma_port_immutable(struct ib_device *ibdev, u8 port_num, struct ib_port_attr attr; int err; - err = pvrdma_query_port(ibdev, port_num, &attr); + immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE; + + err = ib_query_port(ibdev, port_num, &attr); if (err) return err; immutable->pkey_tbl_len = attr.pkey_tbl_len; immutable->gid_tbl_len = attr.gid_tbl_len; - immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE; immutable->max_mad_size = IB_MGMT_MAD_SIZE; return 0; } @@ -172,7 +173,7 @@ static int pvrdma_register_device(struct pvrdma_dev *dev) dev->flags = 0; dev->ib_dev.owner = THIS_MODULE; dev->ib_dev.num_comp_vectors = 1; - dev->ib_dev.dma_device = &dev->pdev->dev; + dev->ib_dev.dev.parent = &dev->pdev->dev; dev->ib_dev.uverbs_abi_ver = PVRDMA_UVERBS_ABI_VERSION; dev->ib_dev.uverbs_cmd_mask = (1ull << IB_USER_VERBS_CMD_GET_CONTEXT) | @@ -282,7 +283,7 @@ static irqreturn_t pvrdma_intr0_handler(int irq, void *dev_id) dev_dbg(&dev->pdev->dev, "interrupt 0 (response) handler\n"); - if (dev->intr.type != PVRDMA_INTR_TYPE_MSIX) { + if (!dev->pdev->msix_enabled) { /* Legacy intr */ icr = pvrdma_read_reg(dev, PVRDMA_REG_ICR); if (icr == 0) @@ -489,31 +490,13 @@ static irqreturn_t pvrdma_intrx_handler(int irq, void *dev_id) return IRQ_HANDLED; } -static void pvrdma_disable_msi_all(struct pvrdma_dev *dev) -{ - if (dev->intr.type == PVRDMA_INTR_TYPE_MSIX) - pci_disable_msix(dev->pdev); - else if (dev->intr.type == PVRDMA_INTR_TYPE_MSI) - pci_disable_msi(dev->pdev); -} - static void pvrdma_free_irq(struct pvrdma_dev *dev) { int i; dev_dbg(&dev->pdev->dev, "freeing interrupts\n"); - - if (dev->intr.type == PVRDMA_INTR_TYPE_MSIX) { - for (i = 0; i < dev->intr.size; i++) { - if (dev->intr.enabled[i]) { - free_irq(dev->intr.msix_entry[i].vector, dev); - dev->intr.enabled[i] = 0; - } - } - } else if (dev->intr.type == PVRDMA_INTR_TYPE_INTX || - dev->intr.type == PVRDMA_INTR_TYPE_MSI) { - free_irq(dev->pdev->irq, dev); - } + for (i = 0; i < dev->nr_vectors; i++) + free_irq(pci_irq_vector(dev->pdev, i), dev); } static void pvrdma_enable_intrs(struct pvrdma_dev *dev) @@ -528,126 +511,48 @@ static void pvrdma_disable_intrs(struct pvrdma_dev *dev) pvrdma_write_reg(dev, PVRDMA_REG_IMR, ~0); } -static int pvrdma_enable_msix(struct pci_dev *pdev, struct pvrdma_dev *dev) -{ - int i; - int ret; - - for (i = 0; i < PVRDMA_MAX_INTERRUPTS; i++) { - dev->intr.msix_entry[i].entry = i; - dev->intr.msix_entry[i].vector = i; - - switch (i) { - case 0: - /* CMD ring handler */ - dev->intr.handler[i] = pvrdma_intr0_handler; - break; - case 1: - /* Async event ring handler */ - dev->intr.handler[i] = pvrdma_intr1_handler; - break; - default: - /* Completion queue handler */ - dev->intr.handler[i] = pvrdma_intrx_handler; - break; - } - } - - ret = pci_enable_msix(pdev, dev->intr.msix_entry, - PVRDMA_MAX_INTERRUPTS); - if (!ret) { - dev->intr.type = PVRDMA_INTR_TYPE_MSIX; - dev->intr.size = PVRDMA_MAX_INTERRUPTS; - } else if (ret > 0) { - ret = pci_enable_msix(pdev, dev->intr.msix_entry, ret); - if (!ret) { - dev->intr.type = PVRDMA_INTR_TYPE_MSIX; - dev->intr.size = ret; - } else { - dev->intr.size = 0; - } - } - - dev_dbg(&pdev->dev, "using interrupt type %d, size %d\n", - dev->intr.type, dev->intr.size); - - return ret; -} - static int pvrdma_alloc_intrs(struct pvrdma_dev *dev) { - int ret = 0; - int i; + struct pci_dev *pdev = dev->pdev; + int ret = 0, i; - if (pci_find_capability(dev->pdev, PCI_CAP_ID_MSIX) && - pvrdma_enable_msix(dev->pdev, dev)) { - /* Try MSI */ - ret = pci_enable_msi(dev->pdev); - if (!ret) { - dev->intr.type = PVRDMA_INTR_TYPE_MSI; - } else { - /* Legacy INTR */ - dev->intr.type = PVRDMA_INTR_TYPE_INTX; - } + ret = pci_alloc_irq_vectors(pdev, 1, PVRDMA_MAX_INTERRUPTS, + PCI_IRQ_MSIX); + if (ret < 0) { + ret = pci_alloc_irq_vectors(pdev, 1, 1, + PCI_IRQ_MSI | PCI_IRQ_LEGACY); + if (ret < 0) + return ret; } + dev->nr_vectors = ret; - /* Request First IRQ */ - switch (dev->intr.type) { - case PVRDMA_INTR_TYPE_INTX: - case PVRDMA_INTR_TYPE_MSI: - ret = request_irq(dev->pdev->irq, pvrdma_intr0_handler, - IRQF_SHARED, DRV_NAME, dev); - if (ret) { - dev_err(&dev->pdev->dev, - "failed to request interrupt\n"); - goto disable_msi; - } - break; - case PVRDMA_INTR_TYPE_MSIX: - ret = request_irq(dev->intr.msix_entry[0].vector, - pvrdma_intr0_handler, 0, DRV_NAME, dev); - if (ret) { - dev_err(&dev->pdev->dev, - "failed to request interrupt 0\n"); - goto disable_msi; - } - dev->intr.enabled[0] = 1; - break; - default: - /* Not reached */ - break; + ret = request_irq(pci_irq_vector(dev->pdev, 0), pvrdma_intr0_handler, + pdev->msix_enabled ? 0 : IRQF_SHARED, DRV_NAME, dev); + if (ret) { + dev_err(&dev->pdev->dev, + "failed to request interrupt 0\n"); + goto out_free_vectors; } - /* For MSIX: request intr for each vector */ - if (dev->intr.size > 1) { - ret = request_irq(dev->intr.msix_entry[1].vector, - pvrdma_intr1_handler, 0, DRV_NAME, dev); + for (i = 1; i < dev->nr_vectors; i++) { + ret = request_irq(pci_irq_vector(dev->pdev, i), + i == 1 ? pvrdma_intr1_handler : + pvrdma_intrx_handler, + 0, DRV_NAME, dev); if (ret) { dev_err(&dev->pdev->dev, - "failed to request interrupt 1\n"); - goto free_irq; - } - dev->intr.enabled[1] = 1; - - for (i = 2; i < dev->intr.size; i++) { - ret = request_irq(dev->intr.msix_entry[i].vector, - pvrdma_intrx_handler, 0, - DRV_NAME, dev); - if (ret) { - dev_err(&dev->pdev->dev, - "failed to request interrupt %d\n", i); - goto free_irq; - } - dev->intr.enabled[i] = 1; + "failed to request interrupt %d\n", i); + goto free_irqs; } } return 0; -free_irq: - pvrdma_free_irq(dev); -disable_msi: - pvrdma_disable_msi_all(dev); +free_irqs: + while (--i >= 0) + free_irq(pci_irq_vector(dev->pdev, i), dev); +out_free_vectors: + pci_free_irq_vectors(pdev); return ret; } @@ -755,7 +660,16 @@ static void pvrdma_netdevice_event_handle(struct pvrdma_dev *dev, pvrdma_dispatch_event(dev, 1, IB_EVENT_PORT_ERR); break; case NETDEV_UP: - pvrdma_dispatch_event(dev, 1, IB_EVENT_PORT_ACTIVE); + pvrdma_write_reg(dev, PVRDMA_REG_CTL, + PVRDMA_DEVICE_CTL_UNQUIESCE); + + mb(); + + if (pvrdma_read_reg(dev, PVRDMA_REG_ERR)) + dev_err(&dev->pdev->dev, + "failed to activate device during link up\n"); + else + pvrdma_dispatch_event(dev, 1, IB_EVENT_PORT_ACTIVE); break; default: dev_dbg(&dev->pdev->dev, "ignore netdevice event %ld on %s\n", @@ -953,7 +867,7 @@ static int pvrdma_pci_probe(struct pci_dev *pdev, dev->dsr->resp_slot_dma = (u64)slot_dma; /* Async event ring */ - dev->dsr->async_ring_pages.num_pages = 4; + dev->dsr->async_ring_pages.num_pages = PVRDMA_NUM_RING_PAGES; ret = pvrdma_page_dir_init(dev, &dev->async_pdir, dev->dsr->async_ring_pages.num_pages, true); if (ret) @@ -962,7 +876,7 @@ static int pvrdma_pci_probe(struct pci_dev *pdev, dev->dsr->async_ring_pages.pdir_dma = dev->async_pdir.dir_dma; /* CQ notification ring */ - dev->dsr->cq_ring_pages.num_pages = 4; + dev->dsr->cq_ring_pages.num_pages = PVRDMA_NUM_RING_PAGES; ret = pvrdma_page_dir_init(dev, &dev->cq_pdir, dev->dsr->cq_ring_pages.num_pages, true); if (ret) @@ -1091,7 +1005,7 @@ err_free_uar_table: pvrdma_uar_table_cleanup(dev); err_free_intrs: pvrdma_free_irq(dev); - pvrdma_disable_msi_all(dev); + pci_free_irq_vectors(pdev); err_free_cq_ring: pvrdma_page_dir_cleanup(dev, &dev->cq_pdir); err_free_async_ring: @@ -1141,7 +1055,7 @@ static void pvrdma_pci_remove(struct pci_dev *pdev) pvrdma_disable_intrs(dev); pvrdma_free_irq(dev); - pvrdma_disable_msi_all(dev); + pci_free_irq_vectors(pdev); /* Deactivate pvrdma device */ pvrdma_write_reg(dev, PVRDMA_REG_CTL, PVRDMA_DEVICE_CTL_RESET); diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_misc.c b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_misc.c index 948b5ccd2a70..ec6a4ca1eeb7 100644 --- a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_misc.c +++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_misc.c @@ -194,7 +194,7 @@ int pvrdma_page_dir_insert_umem(struct pvrdma_page_dir *pdir, len = sg_dma_len(sg) >> PAGE_SHIFT; for (j = 0; j < len; j++) { dma_addr_t addr = sg_dma_address(sg) + - umem->page_size * j; + (j << umem->page_shift); ret = pvrdma_page_dir_insert_dma(pdir, i, addr); if (ret) @@ -277,28 +277,29 @@ void ib_global_route_to_pvrdma(struct pvrdma_global_route *dst, dst->traffic_class = src->traffic_class; } -void pvrdma_ah_attr_to_ib(struct ib_ah_attr *dst, - const struct pvrdma_ah_attr *src) +void pvrdma_ah_attr_to_rdma(struct rdma_ah_attr *dst, + const struct pvrdma_ah_attr *src) { - pvrdma_global_route_to_ib(&dst->grh, &src->grh); - dst->dlid = src->dlid; - dst->sl = src->sl; - dst->src_path_bits = src->src_path_bits; - dst->static_rate = src->static_rate; - dst->ah_flags = src->ah_flags; - dst->port_num = src->port_num; - memcpy(&dst->dmac, &src->dmac, sizeof(dst->dmac)); + dst->type = RDMA_AH_ATTR_TYPE_ROCE; + pvrdma_global_route_to_ib(rdma_ah_retrieve_grh(dst), &src->grh); + rdma_ah_set_dlid(dst, src->dlid); + rdma_ah_set_sl(dst, src->sl); + rdma_ah_set_path_bits(dst, src->src_path_bits); + rdma_ah_set_static_rate(dst, src->static_rate); + rdma_ah_set_ah_flags(dst, src->ah_flags); + rdma_ah_set_port_num(dst, src->port_num); + memcpy(dst->roce.dmac, &src->dmac, ETH_ALEN); } -void ib_ah_attr_to_pvrdma(struct pvrdma_ah_attr *dst, - const struct ib_ah_attr *src) +void rdma_ah_attr_to_pvrdma(struct pvrdma_ah_attr *dst, + const struct rdma_ah_attr *src) { - ib_global_route_to_pvrdma(&dst->grh, &src->grh); - dst->dlid = src->dlid; - dst->sl = src->sl; - dst->src_path_bits = src->src_path_bits; - dst->static_rate = src->static_rate; - dst->ah_flags = src->ah_flags; - dst->port_num = src->port_num; - memcpy(&dst->dmac, &src->dmac, sizeof(dst->dmac)); + ib_global_route_to_pvrdma(&dst->grh, rdma_ah_read_grh(src)); + dst->dlid = rdma_ah_get_dlid(src); + dst->sl = rdma_ah_get_sl(src); + dst->src_path_bits = rdma_ah_get_path_bits(src); + dst->static_rate = rdma_ah_get_static_rate(src); + dst->ah_flags = rdma_ah_get_ah_flags(src); + dst->port_num = rdma_ah_get_port_num(src); + memcpy(&dst->dmac, src->roce.dmac, sizeof(dst->dmac)); } diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_qp.c b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_qp.c index c8c01e558125..ed34d5a581fa 100644 --- a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_qp.c +++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_qp.c @@ -151,7 +151,7 @@ static int pvrdma_set_rq_size(struct pvrdma_dev *dev, } static int pvrdma_set_sq_size(struct pvrdma_dev *dev, struct ib_qp_cap *req_cap, - enum ib_qp_type type, struct pvrdma_qp *qp) + struct pvrdma_qp *qp) { if (req_cap->max_send_wr > dev->dsr->caps.max_qp_wr || req_cap->max_send_sge > dev->dsr->caps.max_sge) { @@ -170,8 +170,9 @@ static int pvrdma_set_sq_size(struct pvrdma_dev *dev, struct ib_qp_cap *req_cap, sizeof(struct pvrdma_sge) * qp->sq.max_sg); /* Note: one extra page for the header. */ - qp->npages_send = 1 + (qp->sq.wqe_cnt * qp->sq.wqe_size + - PAGE_SIZE - 1) / PAGE_SIZE; + qp->npages_send = PVRDMA_QP_NUM_HEADER_PAGES + + (qp->sq.wqe_cnt * qp->sq.wqe_size + PAGE_SIZE - 1) / + PAGE_SIZE; return 0; } @@ -276,8 +277,7 @@ struct ib_qp *pvrdma_create_qp(struct ib_pd *pd, qp->is_kernel = true; ret = pvrdma_set_sq_size(to_vdev(pd->device), - &init_attr->cap, - init_attr->qp_type, qp); + &init_attr->cap, qp); if (ret) goto err_qp; @@ -289,7 +289,7 @@ struct ib_qp *pvrdma_create_qp(struct ib_pd *pd, qp->npages = qp->npages_send + qp->npages_recv; /* Skip header page. */ - qp->sq.offset = PAGE_SIZE; + qp->sq.offset = PVRDMA_QP_NUM_HEADER_PAGES * PAGE_SIZE; /* Recv queue pages are after send pages. */ qp->rq.offset = qp->npages_send * PAGE_SIZE; @@ -342,7 +342,7 @@ struct ib_qp *pvrdma_create_qp(struct ib_pd *pd, cmd->qp_type = ib_qp_type_to_pvrdma(init_attr->qp_type); cmd->access_flags = IB_ACCESS_LOCAL_WRITE; cmd->total_chunks = qp->npages; - cmd->send_chunks = qp->npages_send - 1; + cmd->send_chunks = qp->npages_send - PVRDMA_QP_NUM_HEADER_PAGES; cmd->pdir_dma = qp->pdir.dir_dma; dev_dbg(&dev->pdev->dev, "create queuepair with %d, %d, %d, %d\n", @@ -533,8 +533,8 @@ int pvrdma_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, cmd->attrs.alt_port_num = attr->alt_port_num; cmd->attrs.alt_timeout = attr->alt_timeout; ib_qp_cap_to_pvrdma(&cmd->attrs.cap, &attr->cap); - ib_ah_attr_to_pvrdma(&cmd->attrs.ah_attr, &attr->ah_attr); - ib_ah_attr_to_pvrdma(&cmd->attrs.alt_ah_attr, &attr->alt_ah_attr); + rdma_ah_attr_to_pvrdma(&cmd->attrs.ah_attr, &attr->ah_attr); + rdma_ah_attr_to_pvrdma(&cmd->attrs.alt_ah_attr, &attr->alt_ah_attr); ret = pvrdma_cmd_post(dev, &req, &rsp, PVRDMA_CMD_MODIFY_QP_RESP); if (ret < 0) { @@ -555,13 +555,13 @@ out: return ret; } -static inline void *get_sq_wqe(struct pvrdma_qp *qp, int n) +static inline void *get_sq_wqe(struct pvrdma_qp *qp, unsigned int n) { return pvrdma_page_dir_get_ptr(&qp->pdir, qp->sq.offset + n * qp->sq.wqe_size); } -static inline void *get_rq_wqe(struct pvrdma_qp *qp, int n) +static inline void *get_rq_wqe(struct pvrdma_qp *qp, unsigned int n) { return pvrdma_page_dir_get_ptr(&qp->pdir, qp->rq.offset + n * qp->rq.wqe_size); @@ -599,9 +599,7 @@ int pvrdma_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, unsigned long flags; struct pvrdma_sq_wqe_hdr *wqe_hdr; struct pvrdma_sge *sge; - int i, index; - int nreq; - int ret; + int i, ret; /* * In states lower than RTS, we can fail immediately. In other states, @@ -614,9 +612,8 @@ int pvrdma_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, spin_lock_irqsave(&qp->sq.lock, flags); - index = pvrdma_idx(&qp->sq.ring->prod_tail, qp->sq.wqe_cnt); - for (nreq = 0; wr; nreq++, wr = wr->next) { - unsigned int tail; + while (wr) { + unsigned int tail = 0; if (unlikely(!pvrdma_idx_ring_has_space( qp->sq.ring, qp->sq.wqe_cnt, &tail))) { @@ -681,7 +678,7 @@ int pvrdma_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, } } - wqe_hdr = (struct pvrdma_sq_wqe_hdr *)get_sq_wqe(qp, index); + wqe_hdr = (struct pvrdma_sq_wqe_hdr *)get_sq_wqe(qp, tail); memset(wqe_hdr, 0, sizeof(*wqe_hdr)); wqe_hdr->wr_id = wr->wr_id; wqe_hdr->num_sge = wr->num_sge; @@ -772,12 +769,11 @@ int pvrdma_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, /* Make sure wqe is written before index update */ smp_wmb(); - index++; - if (unlikely(index >= qp->sq.wqe_cnt)) - index = 0; /* Update shared sq ring */ pvrdma_idx_ring_inc(&qp->sq.ring->prod_tail, qp->sq.wqe_cnt); + + wr = wr->next; } ret = 0; @@ -807,7 +803,6 @@ int pvrdma_post_recv(struct ib_qp *ibqp, struct ib_recv_wr *wr, struct pvrdma_qp *qp = to_vqp(ibqp); struct pvrdma_rq_wqe_hdr *wqe_hdr; struct pvrdma_sge *sge; - int index, nreq; int ret = 0; int i; @@ -822,9 +817,8 @@ int pvrdma_post_recv(struct ib_qp *ibqp, struct ib_recv_wr *wr, spin_lock_irqsave(&qp->rq.lock, flags); - index = pvrdma_idx(&qp->rq.ring->prod_tail, qp->rq.wqe_cnt); - for (nreq = 0; wr; nreq++, wr = wr->next) { - unsigned int tail; + while (wr) { + unsigned int tail = 0; if (unlikely(wr->num_sge > qp->rq.max_sg || wr->num_sge < 0)) { @@ -844,7 +838,7 @@ int pvrdma_post_recv(struct ib_qp *ibqp, struct ib_recv_wr *wr, goto out; } - wqe_hdr = (struct pvrdma_rq_wqe_hdr *)get_rq_wqe(qp, index); + wqe_hdr = (struct pvrdma_rq_wqe_hdr *)get_rq_wqe(qp, tail); wqe_hdr->wr_id = wr->wr_id; wqe_hdr->num_sge = wr->num_sge; wqe_hdr->total_len = 0; @@ -860,12 +854,11 @@ int pvrdma_post_recv(struct ib_qp *ibqp, struct ib_recv_wr *wr, /* Make sure wqe is written before index update */ smp_wmb(); - index++; - if (unlikely(index >= qp->rq.wqe_cnt)) - index = 0; /* Update shared rq ring */ pvrdma_idx_ring_inc(&qp->rq.ring->prod_tail, qp->rq.wqe_cnt); + + wr = wr->next; } spin_unlock_irqrestore(&qp->rq.lock, flags); @@ -945,8 +938,8 @@ int pvrdma_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, attr->alt_port_num = resp->attrs.alt_port_num; attr->alt_timeout = resp->attrs.alt_timeout; pvrdma_qp_cap_to_ib(&attr->cap, &resp->attrs.cap); - pvrdma_ah_attr_to_ib(&attr->ah_attr, &resp->attrs.ah_attr); - pvrdma_ah_attr_to_ib(&attr->alt_ah_attr, &resp->attrs.alt_ah_attr); + pvrdma_ah_attr_to_rdma(&attr->ah_attr, &resp->attrs.ah_attr); + pvrdma_ah_attr_to_rdma(&attr->alt_ah_attr, &resp->attrs.alt_ah_attr); qp->state = attr->qp_state; diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.c b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.c index c2aa52638dcb..28517042011d 100644 --- a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.c +++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.c @@ -135,7 +135,7 @@ int pvrdma_query_port(struct ib_device *ibdev, u8 port, return err; } - memset(props, 0, sizeof(*props)); + /* props being zeroed by the caller, avoid zeroing it here */ props->state = pvrdma_port_state_to_ib(resp->attrs.state); props->max_mtu = pvrdma_mtu_to_ib(resp->attrs.max_mtu); @@ -275,7 +275,7 @@ int pvrdma_modify_port(struct ib_device *ibdev, u8 port, int mask, } mutex_lock(&vdev->port_mutex); - ret = pvrdma_query_port(ibdev, port, &attr); + ret = ib_query_port(ibdev, port, &attr); if (ret) goto out; @@ -520,20 +520,20 @@ int pvrdma_dealloc_pd(struct ib_pd *pd) * * @return: the ib_ah pointer on success, otherwise errno. */ -struct ib_ah *pvrdma_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr, +struct ib_ah *pvrdma_create_ah(struct ib_pd *pd, struct rdma_ah_attr *ah_attr, struct ib_udata *udata) { struct pvrdma_dev *dev = to_vdev(pd->device); struct pvrdma_ah *ah; - enum rdma_link_layer ll; + const struct ib_global_route *grh; + u8 port_num = rdma_ah_get_port_num(ah_attr); - if (!(ah_attr->ah_flags & IB_AH_GRH)) + if (!(rdma_ah_get_ah_flags(ah_attr) & IB_AH_GRH)) return ERR_PTR(-EINVAL); - ll = rdma_port_get_link_layer(pd->device, ah_attr->port_num); - - if (ll != IB_LINK_LAYER_ETHERNET || - rdma_is_multicast_addr((struct in6_addr *)ah_attr->grh.dgid.raw)) + grh = rdma_ah_read_grh(ah_attr); + if ((ah_attr->type != RDMA_AH_ATTR_TYPE_ROCE) || + rdma_is_multicast_addr((struct in6_addr *)grh->dgid.raw)) return ERR_PTR(-EINVAL); if (!atomic_add_unless(&dev->num_ahs, 1, dev->dsr->caps.max_ah)) @@ -545,15 +545,15 @@ struct ib_ah *pvrdma_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr, return ERR_PTR(-ENOMEM); } - ah->av.port_pd = to_vpd(pd)->pd_handle | (ah_attr->port_num << 24); - ah->av.src_path_bits = ah_attr->src_path_bits; + ah->av.port_pd = to_vpd(pd)->pd_handle | (port_num << 24); + ah->av.src_path_bits = rdma_ah_get_path_bits(ah_attr); ah->av.src_path_bits |= 0x80; - ah->av.gid_index = ah_attr->grh.sgid_index; - ah->av.hop_limit = ah_attr->grh.hop_limit; - ah->av.sl_tclass_flowlabel = (ah_attr->grh.traffic_class << 20) | - ah_attr->grh.flow_label; - memcpy(ah->av.dgid, ah_attr->grh.dgid.raw, 16); - memcpy(ah->av.dmac, ah_attr->dmac, 6); + ah->av.gid_index = grh->sgid_index; + ah->av.hop_limit = grh->hop_limit; + ah->av.sl_tclass_flowlabel = (grh->traffic_class << 20) | + grh->flow_label; + memcpy(ah->av.dgid, grh->dgid.raw, 16); + memcpy(ah->av.dmac, ah_attr->roce.dmac, ETH_ALEN); ah->ibah.device = pd->device; ah->ibah.pd = pd; diff --git a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.h b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.h index bfbe96b56255..002a9b066e70 100644 --- a/drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.h +++ b/drivers/infiniband/hw/vmw_pvrdma/pvrdma_verbs.h @@ -417,7 +417,7 @@ int pvrdma_resize_cq(struct ib_cq *ibcq, int entries, int pvrdma_destroy_cq(struct ib_cq *cq); int pvrdma_poll_cq(struct ib_cq *ibcq, int num_entries, struct ib_wc *wc); int pvrdma_req_notify_cq(struct ib_cq *cq, enum ib_cq_notify_flags flags); -struct ib_ah *pvrdma_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr, +struct ib_ah *pvrdma_create_ah(struct ib_pd *pd, struct rdma_ah_attr *ah_attr, struct ib_udata *udata); int pvrdma_destroy_ah(struct ib_ah *ah); struct ib_qp *pvrdma_create_qp(struct ib_pd *pd, diff --git a/drivers/infiniband/sw/rdmavt/Kconfig b/drivers/infiniband/sw/rdmavt/Kconfig index 1da8d01a6855..fdd001ce13d8 100644 --- a/drivers/infiniband/sw/rdmavt/Kconfig +++ b/drivers/infiniband/sw/rdmavt/Kconfig @@ -1,5 +1,6 @@ config INFINIBAND_RDMAVT tristate "RDMA verbs transport library" depends on 64BIT + select DMA_VIRT_OPS ---help--- This is a common software verbs provider for RDMA networks. diff --git a/drivers/infiniband/sw/rdmavt/Makefile b/drivers/infiniband/sw/rdmavt/Makefile index ccaa7992ac97..78b276a90401 100644 --- a/drivers/infiniband/sw/rdmavt/Makefile +++ b/drivers/infiniband/sw/rdmavt/Makefile @@ -7,7 +7,7 @@ # obj-$(CONFIG_INFINIBAND_RDMAVT) += rdmavt.o -rdmavt-y := vt.o ah.o cq.o dma.o mad.o mcast.o mmap.o mr.o pd.o qp.o srq.o \ - trace.o +rdmavt-y := vt.o ah.o cq.o mad.o mcast.o mmap.o mr.o pd.o qp.o \ + rc.o srq.o trace.o CFLAGS_trace.o = -I$(src) diff --git a/drivers/infiniband/sw/rdmavt/ah.c b/drivers/infiniband/sw/rdmavt/ah.c index 16c446142c2a..a96d4aa80ae8 100644 --- a/drivers/infiniband/sw/rdmavt/ah.c +++ b/drivers/infiniband/sw/rdmavt/ah.c @@ -60,32 +60,35 @@ * Return: 0 on success */ int rvt_check_ah(struct ib_device *ibdev, - struct ib_ah_attr *ah_attr) + struct rdma_ah_attr *ah_attr) { int err; + int port_num = rdma_ah_get_port_num(ah_attr); struct ib_port_attr port_attr; struct rvt_dev_info *rdi = ib_to_rvt(ibdev); - enum rdma_link_layer link = rdma_port_get_link_layer(ibdev, - ah_attr->port_num); + enum rdma_link_layer link = rdma_port_get_link_layer(ibdev, port_num); + u32 dlid = rdma_ah_get_dlid(ah_attr); + u8 ah_flags = rdma_ah_get_ah_flags(ah_attr); + u8 static_rate = rdma_ah_get_static_rate(ah_attr); - err = ib_query_port(ibdev, ah_attr->port_num, &port_attr); + err = ib_query_port(ibdev, port_num, &port_attr); if (err) return -EINVAL; - if (ah_attr->port_num < 1 || - ah_attr->port_num > ibdev->phys_port_cnt) + if (port_num < 1 || + port_num > ibdev->phys_port_cnt) return -EINVAL; - if (ah_attr->static_rate != IB_RATE_PORT_CURRENT && - ib_rate_to_mbps(ah_attr->static_rate) < 0) + if (static_rate != IB_RATE_PORT_CURRENT && + ib_rate_to_mbps(static_rate) < 0) return -EINVAL; - if ((ah_attr->ah_flags & IB_AH_GRH) && - ah_attr->grh.sgid_index >= port_attr.gid_tbl_len) + if ((ah_flags & IB_AH_GRH) && + rdma_ah_read_grh(ah_attr)->sgid_index >= port_attr.gid_tbl_len) return -EINVAL; if (link != IB_LINK_LAYER_ETHERNET) { - if (ah_attr->dlid == 0) + if (dlid == 0) return -EINVAL; - if (ah_attr->dlid >= be16_to_cpu(IB_MULTICAST_LID_BASE) && - ah_attr->dlid != be16_to_cpu(IB_LID_PERMISSIVE) && - !(ah_attr->ah_flags & IB_AH_GRH)) + if (dlid >= be16_to_cpu(IB_MULTICAST_LID_BASE) && + dlid != be16_to_cpu(IB_LID_PERMISSIVE) && + !(ah_flags & IB_AH_GRH)) return -EINVAL; } if (rdi->driver_f.check_ah) @@ -104,7 +107,7 @@ EXPORT_SYMBOL(rvt_check_ah); * Return: newly allocated ah */ struct ib_ah *rvt_create_ah(struct ib_pd *pd, - struct ib_ah_attr *ah_attr) + struct rdma_ah_attr *ah_attr) { struct rvt_ah *ah; struct rvt_dev_info *dev = ib_to_rvt(pd->device); @@ -119,7 +122,7 @@ struct ib_ah *rvt_create_ah(struct ib_pd *pd, spin_lock_irqsave(&dev->n_ahs_lock, flags); if (dev->n_ahs_allocated == dev->dparms.props.max_ah) { - spin_unlock(&dev->n_ahs_lock); + spin_unlock_irqrestore(&dev->n_ahs_lock, flags); kfree(ah); return ERR_PTR(-ENOMEM); } @@ -167,7 +170,7 @@ int rvt_destroy_ah(struct ib_ah *ibah) * * Return: 0 on success */ -int rvt_modify_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr) +int rvt_modify_ah(struct ib_ah *ibah, struct rdma_ah_attr *ah_attr) { struct rvt_ah *ah = ibah_to_rvtah(ibah); @@ -186,7 +189,7 @@ int rvt_modify_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr) * * Return: always 0 */ -int rvt_query_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr) +int rvt_query_ah(struct ib_ah *ibah, struct rdma_ah_attr *ah_attr) { struct rvt_ah *ah = ibah_to_rvtah(ibah); diff --git a/drivers/infiniband/sw/rdmavt/ah.h b/drivers/infiniband/sw/rdmavt/ah.h index e9c36be87d79..16105af99189 100644 --- a/drivers/infiniband/sw/rdmavt/ah.h +++ b/drivers/infiniband/sw/rdmavt/ah.h @@ -51,9 +51,9 @@ #include <rdma/rdma_vt.h> struct ib_ah *rvt_create_ah(struct ib_pd *pd, - struct ib_ah_attr *ah_attr); + struct rdma_ah_attr *ah_attr); int rvt_destroy_ah(struct ib_ah *ibah); -int rvt_modify_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr); -int rvt_query_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr); +int rvt_modify_ah(struct ib_ah *ibah, struct rdma_ah_attr *ah_attr); +int rvt_query_ah(struct ib_ah *ibah, struct rdma_ah_attr *ah_attr); #endif /* DEF_RVTAH_H */ diff --git a/drivers/infiniband/sw/rdmavt/cq.c b/drivers/infiniband/sw/rdmavt/cq.c index 7aa7a4e312f1..0ae2ff8cf81e 100644 --- a/drivers/infiniband/sw/rdmavt/cq.c +++ b/drivers/infiniband/sw/rdmavt/cq.c @@ -50,6 +50,7 @@ #include <linux/kthread.h> #include "cq.h" #include "vt.h" +#include "trace.h" /** * rvt_cq_enter - add a new entry to the completion queue @@ -93,6 +94,7 @@ void rvt_cq_enter(struct rvt_cq *cq, struct ib_wc *entry, bool solicited) } return; } + trace_rvt_cq_enter(cq, entry, head); if (cq->ip) { wc->uqueue[head].wr_id = entry->wr_id; wc->uqueue[head].status = entry->status; @@ -482,6 +484,7 @@ int rvt_poll_cq(struct ib_cq *ibcq, int num_entries, struct ib_wc *entry) if (tail == wc->head) break; /* The kernel doesn't need a RMB since it has the lock. */ + trace_rvt_cq_poll(cq, &wc->kqueue[tail], npolled); *entry = wc->kqueue[tail]; if (tail >= cq->ibcq.cqe) tail = 0; diff --git a/drivers/infiniband/sw/rdmavt/dma.c b/drivers/infiniband/sw/rdmavt/dma.c deleted file mode 100644 index f2cefb0d9180..000000000000 --- a/drivers/infiniband/sw/rdmavt/dma.c +++ /dev/null @@ -1,198 +0,0 @@ -/* - * Copyright(c) 2016 Intel Corporation. - * - * This file is provided under a dual BSD/GPLv2 license. When using or - * redistributing this file, you may do so under either license. - * - * GPL LICENSE SUMMARY - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of version 2 of the GNU General Public License as - * published by the Free Software Foundation. - * - * This program is distributed in the hope that it will be useful, but - * WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * General Public License for more details. - * - * BSD LICENSE - * - * 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. - * - Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. - * - */ -#include <linux/types.h> -#include <linux/scatterlist.h> -#include <rdma/ib_verbs.h> - -#include "dma.h" - -#define BAD_DMA_ADDRESS ((u64)0) - -/* - * The following functions implement driver specific replacements - * for the ib_dma_*() functions. - * - * These functions return kernel virtual addresses instead of - * device bus addresses since the driver uses the CPU to copy - * data instead of using hardware DMA. - */ - -static int rvt_mapping_error(struct ib_device *dev, u64 dma_addr) -{ - return dma_addr == BAD_DMA_ADDRESS; -} - -static u64 rvt_dma_map_single(struct ib_device *dev, void *cpu_addr, - size_t size, enum dma_data_direction direction) -{ - if (WARN_ON(!valid_dma_direction(direction))) - return BAD_DMA_ADDRESS; - - return (u64)cpu_addr; -} - -static void rvt_dma_unmap_single(struct ib_device *dev, u64 addr, size_t size, - enum dma_data_direction direction) -{ - /* This is a stub, nothing to be done here */ -} - -static u64 rvt_dma_map_page(struct ib_device *dev, struct page *page, - unsigned long offset, size_t size, - enum dma_data_direction direction) -{ - u64 addr; - - if (WARN_ON(!valid_dma_direction(direction))) - return BAD_DMA_ADDRESS; - - addr = (u64)page_address(page); - if (addr) - addr += offset; - - return addr; -} - -static void rvt_dma_unmap_page(struct ib_device *dev, u64 addr, size_t size, - enum dma_data_direction direction) -{ - /* This is a stub, nothing to be done here */ -} - -static int rvt_map_sg(struct ib_device *dev, struct scatterlist *sgl, - int nents, enum dma_data_direction direction) -{ - struct scatterlist *sg; - u64 addr; - int i; - int ret = nents; - - if (WARN_ON(!valid_dma_direction(direction))) - return 0; - - for_each_sg(sgl, sg, nents, i) { - addr = (u64)page_address(sg_page(sg)); - if (!addr) { - ret = 0; - break; - } - sg->dma_address = addr + sg->offset; -#ifdef CONFIG_NEED_SG_DMA_LENGTH - sg->dma_length = sg->length; -#endif - } - return ret; -} - -static void rvt_unmap_sg(struct ib_device *dev, - struct scatterlist *sg, int nents, - enum dma_data_direction direction) -{ - /* This is a stub, nothing to be done here */ -} - -static int rvt_map_sg_attrs(struct ib_device *dev, struct scatterlist *sgl, - int nents, enum dma_data_direction direction, - unsigned long attrs) -{ - return rvt_map_sg(dev, sgl, nents, direction); -} - -static void rvt_unmap_sg_attrs(struct ib_device *dev, - struct scatterlist *sg, int nents, - enum dma_data_direction direction, - unsigned long attrs) -{ - return rvt_unmap_sg(dev, sg, nents, direction); -} - -static void rvt_sync_single_for_cpu(struct ib_device *dev, u64 addr, - size_t size, enum dma_data_direction dir) -{ -} - -static void rvt_sync_single_for_device(struct ib_device *dev, u64 addr, - size_t size, - enum dma_data_direction dir) -{ -} - -static void *rvt_dma_alloc_coherent(struct ib_device *dev, size_t size, - u64 *dma_handle, gfp_t flag) -{ - struct page *p; - void *addr = NULL; - - p = alloc_pages(flag, get_order(size)); - if (p) - addr = page_address(p); - if (dma_handle) - *dma_handle = (u64)addr; - return addr; -} - -static void rvt_dma_free_coherent(struct ib_device *dev, size_t size, - void *cpu_addr, u64 dma_handle) -{ - free_pages((unsigned long)cpu_addr, get_order(size)); -} - -struct ib_dma_mapping_ops rvt_default_dma_mapping_ops = { - .mapping_error = rvt_mapping_error, - .map_single = rvt_dma_map_single, - .unmap_single = rvt_dma_unmap_single, - .map_page = rvt_dma_map_page, - .unmap_page = rvt_dma_unmap_page, - .map_sg = rvt_map_sg, - .unmap_sg = rvt_unmap_sg, - .map_sg_attrs = rvt_map_sg_attrs, - .unmap_sg_attrs = rvt_unmap_sg_attrs, - .sync_single_for_cpu = rvt_sync_single_for_cpu, - .sync_single_for_device = rvt_sync_single_for_device, - .alloc_coherent = rvt_dma_alloc_coherent, - .free_coherent = rvt_dma_free_coherent -}; diff --git a/drivers/infiniband/sw/rdmavt/mad.c b/drivers/infiniband/sw/rdmavt/mad.c index f6e99778d7ca..d6981dc04adb 100644 --- a/drivers/infiniband/sw/rdmavt/mad.c +++ b/drivers/infiniband/sw/rdmavt/mad.c @@ -74,9 +74,9 @@ int rvt_process_mad(struct ib_device *ibdev, int mad_flags, u8 port_num, u16 *out_mad_pkey_index) { /* - * MAD processing is quite different between hfi1 and qib. Therfore this - * is expected to be provided by the driver. Other drivers in the future - * may chose to implement this but it should not be made into a + * MAD processing is quite different between hfi1 and qib. Therefore + * this is expected to be provided by the driver. Other drivers in the + * future may choose to implement this but it should not be made into a * requirement. */ if (ibport_num_to_idx(ibdev, port_num) < 0) @@ -160,7 +160,7 @@ void rvt_free_mad_agents(struct rvt_dev_info *rdi) ib_unregister_mad_agent(agent); } if (rvp->sm_ah) { - ib_destroy_ah(&rvp->sm_ah->ibah); + rdma_destroy_ah(&rvp->sm_ah->ibah); rvp->sm_ah = NULL; } diff --git a/drivers/infiniband/sw/rdmavt/mcast.c b/drivers/infiniband/sw/rdmavt/mcast.c index 05c8c2afb0e3..1f12b69a0d07 100644 --- a/drivers/infiniband/sw/rdmavt/mcast.c +++ b/drivers/infiniband/sw/rdmavt/mcast.c @@ -100,10 +100,11 @@ static void rvt_mcast_qp_free(struct rvt_mcast_qp *mqp) /** * mcast_alloc - allocate the multicast GID structure * @mgid: the multicast GID + * @lid: the muilticast LID (host order) * * A list of QPs will be attached to this structure. */ -static struct rvt_mcast *rvt_mcast_alloc(union ib_gid *mgid) +static struct rvt_mcast *rvt_mcast_alloc(union ib_gid *mgid, u16 lid) { struct rvt_mcast *mcast; @@ -111,7 +112,9 @@ static struct rvt_mcast *rvt_mcast_alloc(union ib_gid *mgid) if (!mcast) goto bail; - mcast->mgid = *mgid; + mcast->mcast_addr.mgid = *mgid; + mcast->mcast_addr.lid = lid; + INIT_LIST_HEAD(&mcast->qp_list); init_waitqueue_head(&mcast->wait); atomic_set(&mcast->refcount, 0); @@ -131,15 +134,19 @@ static void rvt_mcast_free(struct rvt_mcast *mcast) } /** - * rvt_mcast_find - search the global table for the given multicast GID + * rvt_mcast_find - search the global table for the given multicast GID/LID + * NOTE: It is valid to have 1 MLID with multiple MGIDs. It is not valid + * to have 1 MGID with multiple MLIDs. * @ibp: the IB port structure * @mgid: the multicast GID to search for + * @lid: the multicast LID portion of the multicast address (host order) * * The caller is responsible for decrementing the reference count if found. * * Return: NULL if not found. */ -struct rvt_mcast *rvt_mcast_find(struct rvt_ibport *ibp, union ib_gid *mgid) +struct rvt_mcast *rvt_mcast_find(struct rvt_ibport *ibp, union ib_gid *mgid, + u16 lid) { struct rb_node *n; unsigned long flags; @@ -153,15 +160,18 @@ struct rvt_mcast *rvt_mcast_find(struct rvt_ibport *ibp, union ib_gid *mgid) mcast = rb_entry(n, struct rvt_mcast, rb_node); - ret = memcmp(mgid->raw, mcast->mgid.raw, - sizeof(union ib_gid)); + ret = memcmp(mgid->raw, mcast->mcast_addr.mgid.raw, + sizeof(*mgid)); if (ret < 0) { n = n->rb_left; } else if (ret > 0) { n = n->rb_right; } else { - atomic_inc(&mcast->refcount); - found = mcast; + /* MGID/MLID must match */ + if (mcast->mcast_addr.lid == lid) { + atomic_inc(&mcast->refcount); + found = mcast; + } break; } } @@ -177,7 +187,8 @@ EXPORT_SYMBOL(rvt_mcast_find); * * Return: zero if both were added. Return EEXIST if the GID was already in * the table but the QP was added. Return ESRCH if the QP was already - * attached and neither structure was added. + * attached and neither structure was added. Return EINVAL if the MGID was + * found, but the MLID did NOT match. */ static int rvt_mcast_add(struct rvt_dev_info *rdi, struct rvt_ibport *ibp, struct rvt_mcast *mcast, struct rvt_mcast_qp *mqp) @@ -195,8 +206,9 @@ static int rvt_mcast_add(struct rvt_dev_info *rdi, struct rvt_ibport *ibp, pn = *n; tmcast = rb_entry(pn, struct rvt_mcast, rb_node); - ret = memcmp(mcast->mgid.raw, tmcast->mgid.raw, - sizeof(union ib_gid)); + ret = memcmp(mcast->mcast_addr.mgid.raw, + tmcast->mcast_addr.mgid.raw, + sizeof(mcast->mcast_addr.mgid)); if (ret < 0) { n = &pn->rb_left; continue; @@ -206,6 +218,11 @@ static int rvt_mcast_add(struct rvt_dev_info *rdi, struct rvt_ibport *ibp, continue; } + if (tmcast->mcast_addr.lid != mcast->mcast_addr.lid) { + ret = EINVAL; + goto bail; + } + /* Search the QP list to see if this is already there. */ list_for_each_entry_rcu(p, &tmcast->qp_list, list) { if (p->qp == mqp->qp) { @@ -276,7 +293,7 @@ int rvt_attach_mcast(struct ib_qp *ibqp, union ib_gid *gid, u16 lid) * Allocate data structures since its better to do this outside of * spin locks and it will most likely be needed. */ - mcast = rvt_mcast_alloc(gid); + mcast = rvt_mcast_alloc(gid, lid); if (!mcast) return -ENOMEM; @@ -296,6 +313,10 @@ int rvt_attach_mcast(struct ib_qp *ibqp, union ib_gid *gid, u16 lid) /* Exceeded the maximum number of mcast groups. */ ret = -ENOMEM; goto bail_mqp; + case EINVAL: + /* Invalid MGID/MLID pair */ + ret = -EINVAL; + goto bail_mqp; default: break; } @@ -344,14 +365,20 @@ int rvt_detach_mcast(struct ib_qp *ibqp, union ib_gid *gid, u16 lid) } mcast = rb_entry(n, struct rvt_mcast, rb_node); - ret = memcmp(gid->raw, mcast->mgid.raw, - sizeof(union ib_gid)); - if (ret < 0) + ret = memcmp(gid->raw, mcast->mcast_addr.mgid.raw, + sizeof(*gid)); + if (ret < 0) { n = n->rb_left; - else if (ret > 0) + } else if (ret > 0) { n = n->rb_right; - else + } else { + /* MGID/MLID must match */ + if (mcast->mcast_addr.lid != lid) { + spin_unlock_irq(&ibp->lock); + return -EINVAL; + } break; + } } /* Search the QP list. */ diff --git a/drivers/infiniband/sw/rdmavt/mmap.c b/drivers/infiniband/sw/rdmavt/mmap.c index e202b8142759..6b712eecbd37 100644 --- a/drivers/infiniband/sw/rdmavt/mmap.c +++ b/drivers/infiniband/sw/rdmavt/mmap.c @@ -170,9 +170,9 @@ struct rvt_mmap_info *rvt_create_mmap_info(struct rvt_dev_info *rdi, spin_lock_irq(&rdi->mmap_offset_lock); if (rdi->mmap_offset == 0) - rdi->mmap_offset = PAGE_SIZE; + rdi->mmap_offset = ALIGN(PAGE_SIZE, SHMLBA); ip->offset = rdi->mmap_offset; - rdi->mmap_offset += size; + rdi->mmap_offset += ALIGN(size, SHMLBA); spin_unlock_irq(&rdi->mmap_offset_lock); INIT_LIST_HEAD(&ip->pending_mmaps); diff --git a/drivers/infiniband/sw/rdmavt/mr.c b/drivers/infiniband/sw/rdmavt/mr.c index 52fd15276ee6..aa5f9ea318e4 100644 --- a/drivers/infiniband/sw/rdmavt/mr.c +++ b/drivers/infiniband/sw/rdmavt/mr.c @@ -120,10 +120,19 @@ static void rvt_deinit_mregion(struct rvt_mregion *mr) mr->mapsz = 0; while (i) kfree(mr->map[--i]); + percpu_ref_exit(&mr->refcount); +} + +static void __rvt_mregion_complete(struct percpu_ref *ref) +{ + struct rvt_mregion *mr = container_of(ref, struct rvt_mregion, + refcount); + + complete(&mr->comp); } static int rvt_init_mregion(struct rvt_mregion *mr, struct ib_pd *pd, - int count) + int count, unsigned int percpu_flags) { int m, i = 0; struct rvt_dev_info *dev = ib_to_rvt(pd->device); @@ -133,19 +142,23 @@ static int rvt_init_mregion(struct rvt_mregion *mr, struct ib_pd *pd, for (; i < m; i++) { mr->map[i] = kzalloc_node(sizeof(*mr->map[0]), GFP_KERNEL, dev->dparms.node); - if (!mr->map[i]) { - rvt_deinit_mregion(mr); - return -ENOMEM; - } + if (!mr->map[i]) + goto bail; mr->mapsz++; } init_completion(&mr->comp); /* count returning the ptr to user */ - atomic_set(&mr->refcount, 1); + if (percpu_ref_init(&mr->refcount, &__rvt_mregion_complete, + percpu_flags, GFP_KERNEL)) + goto bail; + atomic_set(&mr->lkey_invalid, 0); mr->pd = pd; mr->max_segs = count; return 0; +bail: + rvt_deinit_mregion(mr); + return -ENOMEM; } /** @@ -178,10 +191,10 @@ static int rvt_alloc_lkey(struct rvt_mregion *mr, int dma_region) tmr = rcu_access_pointer(dev->dma_mr); if (!tmr) { - rcu_assign_pointer(dev->dma_mr, mr); mr->lkey_published = 1; - } else { - rvt_put_mr(mr); + /* Insure published written first */ + rcu_assign_pointer(dev->dma_mr, mr); + rvt_get_mr(mr); } goto success; } @@ -212,8 +225,9 @@ static int rvt_alloc_lkey(struct rvt_mregion *mr, int dma_region) mr->lkey |= 1 << 8; rkt->gen++; } - rcu_assign_pointer(rkt->table[r], mr); mr->lkey_published = 1; + /* Insure published written first */ + rcu_assign_pointer(rkt->table[r], mr); success: spin_unlock_irqrestore(&rkt->lock, flags); out: @@ -239,22 +253,26 @@ static void rvt_free_lkey(struct rvt_mregion *mr) int freed = 0; spin_lock_irqsave(&rkt->lock, flags); - if (!mr->lkey_published) - goto out; - if (lkey == 0) { - RCU_INIT_POINTER(dev->dma_mr, NULL); + if (!lkey) { + if (mr->lkey_published) { + mr->lkey_published = 0; + /* insure published is written before pointer */ + rcu_assign_pointer(dev->dma_mr, NULL); + rvt_put_mr(mr); + } } else { + if (!mr->lkey_published) + goto out; r = lkey >> (32 - dev->dparms.lkey_table_size); - RCU_INIT_POINTER(rkt->table[r], NULL); + mr->lkey_published = 0; + /* insure published is written before pointer */ + rcu_assign_pointer(rkt->table[r], NULL); } - mr->lkey_published = 0; freed++; out: spin_unlock_irqrestore(&rkt->lock, flags); - if (freed) { - synchronize_rcu(); - rvt_put_mr(mr); - } + if (freed) + percpu_ref_kill(&mr->refcount); } static struct rvt_mr *__rvt_alloc_mr(int count, struct ib_pd *pd) @@ -269,7 +287,7 @@ static struct rvt_mr *__rvt_alloc_mr(int count, struct ib_pd *pd) if (!mr) goto bail; - rval = rvt_init_mregion(&mr->mr, pd, count); + rval = rvt_init_mregion(&mr->mr, pd, count, 0); if (rval) goto bail; /* @@ -294,8 +312,8 @@ bail: static void __rvt_free_mr(struct rvt_mr *mr) { - rvt_deinit_mregion(&mr->mr); rvt_free_lkey(&mr->mr); + rvt_deinit_mregion(&mr->mr); kfree(mr); } @@ -305,8 +323,8 @@ static void __rvt_free_mr(struct rvt_mr *mr) * @acc: access flags * * Return: the memory region on success, otherwise returns an errno. - * Note that all DMA addresses should be created via the - * struct ib_dma_mapping_ops functions (see dma.c). + * Note that all DMA addresses should be created via the functions in + * struct dma_virt_ops. */ struct ib_mr *rvt_get_dma_mr(struct ib_pd *pd, int acc) { @@ -323,7 +341,7 @@ struct ib_mr *rvt_get_dma_mr(struct ib_pd *pd, int acc) goto bail; } - rval = rvt_init_mregion(&mr->mr, pd, 0); + rval = rvt_init_mregion(&mr->mr, pd, 0, 0); if (rval) { ret = ERR_PTR(rval); goto bail; @@ -390,8 +408,7 @@ struct ib_mr *rvt_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, mr->mr.access_flags = mr_access_flags; mr->umem = umem; - if (is_power_of_2(umem->page_size)) - mr->mr.page_shift = ilog2(umem->page_size); + mr->mr.page_shift = umem->page_shift; m = 0; n = 0; for_each_sg(umem->sg_head.sgl, sg, umem->nmap, entry) { @@ -403,8 +420,9 @@ struct ib_mr *rvt_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, goto bail_inval; } mr->mr.map[m]->segs[n].vaddr = vaddr; - mr->mr.map[m]->segs[n].length = umem->page_size; - trace_rvt_mr_user_seg(&mr->mr, m, n, vaddr, umem->page_size); + mr->mr.map[m]->segs[n].length = BIT(umem->page_shift); + trace_rvt_mr_user_seg(&mr->mr, m, n, vaddr, + BIT(umem->page_shift)); n++; if (n == RVT_SEGSZ) { m++; @@ -445,8 +463,8 @@ int rvt_dereg_mr(struct ib_mr *ibmr) timeout = wait_for_completion_timeout(&mr->mr.comp, 5 * HZ); if (!timeout) { rvt_pr_err(rdi, - "rvt_dereg_mr timeout mr %p pd %p refcount %u\n", - mr, mr->mr.pd, atomic_read(&mr->mr.refcount)); + "rvt_dereg_mr timeout mr %p pd %p\n", + mr, mr->mr.pd); rvt_get_mr(&mr->mr); ret = -EBUSY; goto out; @@ -623,7 +641,8 @@ struct ib_fmr *rvt_alloc_fmr(struct ib_pd *pd, int mr_access_flags, if (!fmr) goto bail; - rval = rvt_init_mregion(&fmr->mr, pd, fmr_attr->max_pages); + rval = rvt_init_mregion(&fmr->mr, pd, fmr_attr->max_pages, + PERCPU_REF_INIT_ATOMIC); if (rval) goto bail; @@ -674,11 +693,12 @@ int rvt_map_phys_fmr(struct ib_fmr *ibfmr, u64 *page_list, struct rvt_fmr *fmr = to_ifmr(ibfmr); struct rvt_lkey_table *rkt; unsigned long flags; - int m, n, i; + int m, n; + unsigned long i; u32 ps; struct rvt_dev_info *rdi = ib_to_rvt(ibfmr->device); - i = atomic_read(&fmr->mr.refcount); + i = atomic_long_read(&fmr->mr.refcount.count); if (i > 2) return -EBUSY; @@ -782,7 +802,7 @@ int rvt_lkey_ok(struct rvt_lkey_table *rkt, struct rvt_pd *pd, /* * We use LKEY == zero for kernel virtual addresses - * (see rvt_get_dma_mr and dma.c). + * (see rvt_get_dma_mr() and dma_virt_ops). */ rcu_read_lock(); if (sge->lkey == 0) { @@ -805,16 +825,21 @@ int rvt_lkey_ok(struct rvt_lkey_table *rkt, struct rvt_pd *pd, goto ok; } mr = rcu_dereference(rkt->table[sge->lkey >> rkt->shift]); - if (unlikely(!mr || atomic_read(&mr->lkey_invalid) || - mr->lkey != sge->lkey || mr->pd != &pd->ibpd)) + if (!mr) goto bail; + rvt_get_mr(mr); + if (!READ_ONCE(mr->lkey_published)) + goto bail_unref; + + if (unlikely(atomic_read(&mr->lkey_invalid) || + mr->lkey != sge->lkey || mr->pd != &pd->ibpd)) + goto bail_unref; off = sge->addr - mr->user_base; if (unlikely(sge->addr < mr->user_base || off + sge->length > mr->length || (mr->access_flags & acc) != acc)) - goto bail; - rvt_get_mr(mr); + goto bail_unref; rcu_read_unlock(); off += mr->offset; @@ -850,6 +875,8 @@ int rvt_lkey_ok(struct rvt_lkey_table *rkt, struct rvt_pd *pd, isge->n = n; ok: return 1; +bail_unref: + rvt_put_mr(mr); bail: rcu_read_unlock(); return 0; @@ -880,7 +907,7 @@ int rvt_rkey_ok(struct rvt_qp *qp, struct rvt_sge *sge, /* * We use RKEY == zero for kernel virtual addresses - * (see rvt_get_dma_mr and dma.c). + * (see rvt_get_dma_mr() and dma_virt_ops). */ rcu_read_lock(); if (rkey == 0) { @@ -905,15 +932,20 @@ int rvt_rkey_ok(struct rvt_qp *qp, struct rvt_sge *sge, } mr = rcu_dereference(rkt->table[rkey >> rkt->shift]); - if (unlikely(!mr || atomic_read(&mr->lkey_invalid) || - mr->lkey != rkey || qp->ibqp.pd != mr->pd)) + if (!mr) goto bail; + rvt_get_mr(mr); + /* insure mr read is before test */ + if (!READ_ONCE(mr->lkey_published)) + goto bail_unref; + if (unlikely(atomic_read(&mr->lkey_invalid) || + mr->lkey != rkey || qp->ibqp.pd != mr->pd)) + goto bail_unref; off = vaddr - mr->iova; if (unlikely(vaddr < mr->iova || off + len > mr->length || (mr->access_flags & acc) == 0)) - goto bail; - rvt_get_mr(mr); + goto bail_unref; rcu_read_unlock(); off += mr->offset; @@ -949,6 +981,8 @@ int rvt_rkey_ok(struct rvt_qp *qp, struct rvt_sge *sge, sge->n = n; ok: return 1; +bail_unref: + rvt_put_mr(mr); bail: rcu_read_unlock(); return 0; diff --git a/drivers/infiniband/sw/rdmavt/pd.c b/drivers/infiniband/sw/rdmavt/pd.c index d1292f324c67..8a89afff3363 100644 --- a/drivers/infiniband/sw/rdmavt/pd.c +++ b/drivers/infiniband/sw/rdmavt/pd.c @@ -90,7 +90,7 @@ struct ib_pd *rvt_alloc_pd(struct ib_device *ibdev, spin_unlock(&dev->n_pds_lock); /* ib_alloc_pd() will initialize pd->ibpd. */ - pd->user = udata ? 1 : 0; + pd->user = !!udata; ret = &pd->ibpd; diff --git a/drivers/infiniband/sw/rdmavt/qp.c b/drivers/infiniband/sw/rdmavt/qp.c index 2a13ac660f2b..8876ee7bc326 100644 --- a/drivers/infiniband/sw/rdmavt/qp.c +++ b/drivers/infiniband/sw/rdmavt/qp.c @@ -1,5 +1,5 @@ /* - * Copyright(c) 2016 Intel Corporation. + * Copyright(c) 2016, 2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -51,10 +51,51 @@ #include <linux/vmalloc.h> #include <linux/slab.h> #include <rdma/ib_verbs.h> +#include <rdma/ib_hdrs.h> #include "qp.h" #include "vt.h" #include "trace.h" +static void rvt_rc_timeout(unsigned long arg); + +/* + * Convert the AETH RNR timeout code into the number of microseconds. + */ +static const u32 ib_rvt_rnr_table[32] = { + 655360, /* 00: 655.36 */ + 10, /* 01: .01 */ + 20, /* 02 .02 */ + 30, /* 03: .03 */ + 40, /* 04: .04 */ + 60, /* 05: .06 */ + 80, /* 06: .08 */ + 120, /* 07: .12 */ + 160, /* 08: .16 */ + 240, /* 09: .24 */ + 320, /* 0A: .32 */ + 480, /* 0B: .48 */ + 640, /* 0C: .64 */ + 960, /* 0D: .96 */ + 1280, /* 0E: 1.28 */ + 1920, /* 0F: 1.92 */ + 2560, /* 10: 2.56 */ + 3840, /* 11: 3.84 */ + 5120, /* 12: 5.12 */ + 7680, /* 13: 7.68 */ + 10240, /* 14: 10.24 */ + 15360, /* 15: 15.36 */ + 20480, /* 16: 20.48 */ + 30720, /* 17: 30.72 */ + 40960, /* 18: 40.96 */ + 61440, /* 19: 61.44 */ + 81920, /* 1A: 81.92 */ + 122880, /* 1B: 122.88 */ + 163840, /* 1C: 163.84 */ + 245760, /* 1D: 245.76 */ + 327680, /* 1E: 327.68 */ + 491520 /* 1F: 491.52 */ +}; + /* * Note that it is OK to post send work requests in the SQE and ERR * states; rvt_do_send() will process them and generate error @@ -76,28 +117,10 @@ const int ib_rvt_state_ops[IB_QPS_ERR + 1] = { }; EXPORT_SYMBOL(ib_rvt_state_ops); -/* - * Translate ib_wr_opcode into ib_wc_opcode. - */ -const enum ib_wc_opcode ib_rvt_wc_opcode[] = { - [IB_WR_RDMA_WRITE] = IB_WC_RDMA_WRITE, - [IB_WR_RDMA_WRITE_WITH_IMM] = IB_WC_RDMA_WRITE, - [IB_WR_SEND] = IB_WC_SEND, - [IB_WR_SEND_WITH_IMM] = IB_WC_SEND, - [IB_WR_RDMA_READ] = IB_WC_RDMA_READ, - [IB_WR_ATOMIC_CMP_AND_SWP] = IB_WC_COMP_SWAP, - [IB_WR_ATOMIC_FETCH_AND_ADD] = IB_WC_FETCH_ADD, - [IB_WR_SEND_WITH_INV] = IB_WC_SEND, - [IB_WR_LOCAL_INV] = IB_WC_LOCAL_INV, - [IB_WR_REG_MR] = IB_WC_REG_MR -}; -EXPORT_SYMBOL(ib_rvt_wc_opcode); - static void get_map_page(struct rvt_qpn_table *qpt, - struct rvt_qpn_map *map, - gfp_t gfp) + struct rvt_qpn_map *map) { - unsigned long page = get_zeroed_page(gfp); + unsigned long page = get_zeroed_page(GFP_KERNEL); /* * Free the page if someone raced with us installing it. @@ -149,7 +172,7 @@ static int init_qpn_table(struct rvt_dev_info *rdi, struct rvt_qpn_table *qpt) rdi->dparms.qpn_res_start, rdi->dparms.qpn_res_end); for (i = rdi->dparms.qpn_res_start; i <= rdi->dparms.qpn_res_end; i++) { if (!map->page) { - get_map_page(qpt, map, GFP_KERNEL); + get_map_page(qpt, map); if (!map->page) { ret = -ENOMEM; break; @@ -200,7 +223,8 @@ int rvt_driver_qp_init(struct rvt_dev_info *rdi) if (!rdi->driver_f.free_all_qps || !rdi->driver_f.qp_priv_alloc || !rdi->driver_f.qp_priv_free || - !rdi->driver_f.notify_qp_reset) + !rdi->driver_f.notify_qp_reset || + !rdi->driver_f.notify_restart_rc) return -EINVAL; /* allocate parent object */ @@ -317,14 +341,14 @@ static inline unsigned mk_qpn(struct rvt_qpn_table *qpt, * Return: The queue pair number */ static int alloc_qpn(struct rvt_dev_info *rdi, struct rvt_qpn_table *qpt, - enum ib_qp_type type, u8 port_num, gfp_t gfp) + enum ib_qp_type type, u8 port_num) { u32 i, offset, max_scan, qpn; struct rvt_qpn_map *map; u32 ret; if (rdi->driver_f.alloc_qpn) - return rdi->driver_f.alloc_qpn(rdi, qpt, type, port_num, gfp); + return rdi->driver_f.alloc_qpn(rdi, qpt, type, port_num); if (type == IB_QPT_SMI || type == IB_QPT_GSI) { unsigned n; @@ -349,7 +373,7 @@ static int alloc_qpn(struct rvt_dev_info *rdi, struct rvt_qpn_table *qpt, max_scan = qpt->nmaps - !offset; for (i = 0;;) { if (unlikely(!map->page)) { - get_map_page(qpt, map, gfp); + get_map_page(qpt, map); if (unlikely(!map->page)) break; } @@ -587,6 +611,7 @@ static void rvt_reset_qp(struct rvt_dev_info *rdi, struct rvt_qp *qp, /* Let drivers flush their waitlist */ rdi->driver_f.flush_qp_waiters(qp); + rvt_stop_rc_timers(qp); qp->s_flags &= ~(RVT_S_TIMER | RVT_S_ANY_WAIT); spin_unlock(&qp->s_lock); spin_unlock(&qp->s_hlock); @@ -594,7 +619,7 @@ static void rvt_reset_qp(struct rvt_dev_info *rdi, struct rvt_qp *qp, /* Stop the send queue and the retry timer */ rdi->driver_f.stop_send_queue(qp); - + rvt_del_timers_sync(qp); /* Wait for things to stop */ rdi->driver_f.quiesce_qp(qp); @@ -646,7 +671,6 @@ struct ib_qp *rvt_create_qp(struct ib_pd *ibpd, struct ib_qp *ret = ERR_PTR(-ENOMEM); struct rvt_dev_info *rdi = ib_to_rvt(ibpd->device); void *priv = NULL; - gfp_t gfp; size_t sqsize; if (!rdi) @@ -654,18 +678,9 @@ struct ib_qp *rvt_create_qp(struct ib_pd *ibpd, if (init_attr->cap.max_send_sge > rdi->dparms.props.max_sge || init_attr->cap.max_send_wr > rdi->dparms.props.max_qp_wr || - init_attr->create_flags & ~(IB_QP_CREATE_USE_GFP_NOIO)) + init_attr->create_flags) return ERR_PTR(-EINVAL); - /* GFP_NOIO is applicable to RC QP's only */ - - if (init_attr->create_flags & IB_QP_CREATE_USE_GFP_NOIO && - init_attr->qp_type != IB_QPT_RC) - return ERR_PTR(-EINVAL); - - gfp = init_attr->create_flags & IB_QP_CREATE_USE_GFP_NOIO ? - GFP_NOIO : GFP_KERNEL; - /* Check receive queue parameters if no SRQ is specified. */ if (!init_attr->srq) { if (init_attr->cap.max_recv_sge > rdi->dparms.props.max_sge || @@ -693,14 +708,7 @@ struct ib_qp *rvt_create_qp(struct ib_pd *ibpd, sz = sizeof(struct rvt_sge) * init_attr->cap.max_send_sge + sizeof(struct rvt_swqe); - if (gfp == GFP_NOIO) - swq = __vmalloc( - sqsize * sz, - gfp | __GFP_ZERO, PAGE_KERNEL); - else - swq = vzalloc_node( - sqsize * sz, - rdi->dparms.node); + swq = vzalloc_node(sqsize * sz, rdi->dparms.node); if (!swq) return ERR_PTR(-ENOMEM); @@ -715,7 +723,8 @@ struct ib_qp *rvt_create_qp(struct ib_pd *ibpd, } else if (init_attr->cap.max_recv_sge > 1) sg_list_sz = sizeof(*qp->r_sg_list) * (init_attr->cap.max_recv_sge - 1); - qp = kzalloc_node(sz + sg_list_sz, gfp, rdi->dparms.node); + qp = kzalloc_node(sz + sg_list_sz, GFP_KERNEL, + rdi->dparms.node); if (!qp) goto bail_swq; @@ -725,17 +734,22 @@ struct ib_qp *rvt_create_qp(struct ib_pd *ibpd, kzalloc_node( sizeof(*qp->s_ack_queue) * rvt_max_atomic(rdi), - gfp, + GFP_KERNEL, rdi->dparms.node); if (!qp->s_ack_queue) goto bail_qp; } + /* initialize timers needed for rc qp */ + setup_timer(&qp->s_timer, rvt_rc_timeout, (unsigned long)qp); + hrtimer_init(&qp->s_rnr_timer, CLOCK_MONOTONIC, + HRTIMER_MODE_REL); + qp->s_rnr_timer.function = rvt_rc_rnr_retry; /* * Driver needs to set up it's private QP structure and do any * initialization that is needed. */ - priv = rdi->driver_f.qp_priv_alloc(rdi, qp, gfp); + priv = rdi->driver_f.qp_priv_alloc(rdi, qp); if (IS_ERR(priv)) { ret = priv; goto bail_qp; @@ -755,11 +769,6 @@ struct ib_qp *rvt_create_qp(struct ib_pd *ibpd, qp->r_rq.wq = vmalloc_user( sizeof(struct rvt_rwq) + qp->r_rq.size * sz); - else if (gfp == GFP_NOIO) - qp->r_rq.wq = __vmalloc( - sizeof(struct rvt_rwq) + - qp->r_rq.size * sz, - gfp | __GFP_ZERO, PAGE_KERNEL); else qp->r_rq.wq = vzalloc_node( sizeof(struct rvt_rwq) + @@ -793,7 +802,7 @@ struct ib_qp *rvt_create_qp(struct ib_pd *ibpd, err = alloc_qpn(rdi, &rdi->qp_dev->qpn_table, init_attr->qp_type, - init_attr->port_num, gfp); + init_attr->port_num); if (err < 0) { ret = ERR_PTR(err); goto bail_rq_wq; @@ -1073,14 +1082,15 @@ int rvt_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, goto inval; if (attr_mask & IB_QP_AV) { - if (attr->ah_attr.dlid >= be16_to_cpu(IB_MULTICAST_LID_BASE)) + if (rdma_ah_get_dlid(&attr->ah_attr) >= + be16_to_cpu(IB_MULTICAST_LID_BASE)) goto inval; if (rvt_check_ah(qp->ibqp.device, &attr->ah_attr)) goto inval; } if (attr_mask & IB_QP_ALT_PATH) { - if (attr->alt_ah_attr.dlid >= + if (rdma_ah_get_dlid(&attr->alt_ah_attr) >= be16_to_cpu(IB_MULTICAST_LID_BASE)) goto inval; if (rvt_check_ah(qp->ibqp.device, &attr->alt_ah_attr)) @@ -1209,7 +1219,7 @@ int rvt_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, if (attr_mask & IB_QP_AV) { qp->remote_ah_attr = attr->ah_attr; - qp->s_srate = attr->ah_attr.static_rate; + qp->s_srate = rdma_ah_get_static_rate(&attr->ah_attr); qp->srate_mbps = ib_rate_to_mbps(qp->s_srate); } @@ -1222,7 +1232,7 @@ int rvt_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, qp->s_mig_state = attr->path_mig_state; if (mig) { qp->remote_ah_attr = qp->alt_ah_attr; - qp->port_num = qp->alt_ah_attr.port_num; + qp->port_num = rdma_ah_get_port_num(&qp->alt_ah_attr); qp->s_pkey_index = qp->s_alt_pkey_index; } } @@ -1248,9 +1258,7 @@ int rvt_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, if (attr_mask & IB_QP_TIMEOUT) { qp->timeout = attr->timeout; - qp->timeout_jiffies = - usecs_to_jiffies((4096UL * (1UL << qp->timeout)) / - 1000UL); + qp->timeout_jiffies = rvt_timeout_to_jiffies(qp->timeout); } if (attr_mask & IB_QP_QKEY) @@ -1393,7 +1401,8 @@ int rvt_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, attr->timeout = qp->timeout; attr->retry_cnt = qp->s_retry_cnt; attr->rnr_retry = qp->s_rnr_retry_cnt; - attr->alt_port_num = qp->alt_ah_attr.port_num; + attr->alt_port_num = + rdma_ah_get_port_num(&qp->alt_ah_attr); attr->alt_timeout = qp->alt_timeout; init_attr->event_handler = qp->ibqp.event_handler; @@ -1741,11 +1750,14 @@ static int rvt_post_one_wr(struct rvt_qp *qp, 0); qp->s_next_psn = wqe->lpsn + 1; } - trace_rvt_post_one_wr(qp, wqe); - if (unlikely(reserved_op)) + if (unlikely(reserved_op)) { + wqe->wr.send_flags |= RVT_SEND_RESERVE_USED; rvt_qp_wqe_reserve(qp, wqe); - else + } else { + wqe->wr.send_flags &= ~RVT_SEND_RESERVE_USED; qp->s_avail--; + } + trace_rvt_post_one_wr(qp, wqe); smp_wmb(); /* see request builders */ qp->s_head = next; @@ -1868,3 +1880,188 @@ int rvt_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr, } return 0; } + +/** + * qp_comm_est - handle trap with QP established + * @qp: the QP + */ +void rvt_comm_est(struct rvt_qp *qp) +{ + qp->r_flags |= RVT_R_COMM_EST; + if (qp->ibqp.event_handler) { + struct ib_event ev; + + ev.device = qp->ibqp.device; + ev.element.qp = &qp->ibqp; + ev.event = IB_EVENT_COMM_EST; + qp->ibqp.event_handler(&ev, qp->ibqp.qp_context); + } +} +EXPORT_SYMBOL(rvt_comm_est); + +void rvt_rc_error(struct rvt_qp *qp, enum ib_wc_status err) +{ + unsigned long flags; + int lastwqe; + + spin_lock_irqsave(&qp->s_lock, flags); + lastwqe = rvt_error_qp(qp, err); + spin_unlock_irqrestore(&qp->s_lock, flags); + + if (lastwqe) { + struct ib_event ev; + + ev.device = qp->ibqp.device; + ev.element.qp = &qp->ibqp; + ev.event = IB_EVENT_QP_LAST_WQE_REACHED; + qp->ibqp.event_handler(&ev, qp->ibqp.qp_context); + } +} +EXPORT_SYMBOL(rvt_rc_error); + +/* + * rvt_rnr_tbl_to_usec - return index into ib_rvt_rnr_table + * @index - the index + * return usec from an index into ib_rvt_rnr_table + */ +unsigned long rvt_rnr_tbl_to_usec(u32 index) +{ + return ib_rvt_rnr_table[(index & IB_AETH_CREDIT_MASK)]; +} +EXPORT_SYMBOL(rvt_rnr_tbl_to_usec); + +static inline unsigned long rvt_aeth_to_usec(u32 aeth) +{ + return ib_rvt_rnr_table[(aeth >> IB_AETH_CREDIT_SHIFT) & + IB_AETH_CREDIT_MASK]; +} + +/* + * rvt_add_retry_timer - add/start a retry timer + * @qp - the QP + * add a retry timer on the QP + */ +void rvt_add_retry_timer(struct rvt_qp *qp) +{ + struct ib_qp *ibqp = &qp->ibqp; + struct rvt_dev_info *rdi = ib_to_rvt(ibqp->device); + + lockdep_assert_held(&qp->s_lock); + qp->s_flags |= RVT_S_TIMER; + /* 4.096 usec. * (1 << qp->timeout) */ + qp->s_timer.expires = jiffies + qp->timeout_jiffies + + rdi->busy_jiffies; + add_timer(&qp->s_timer); +} +EXPORT_SYMBOL(rvt_add_retry_timer); + +/** + * rvt_add_rnr_timer - add/start an rnr timer + * @qp - the QP + * @aeth - aeth of RNR timeout, simulated aeth for loopback + * add an rnr timer on the QP + */ +void rvt_add_rnr_timer(struct rvt_qp *qp, u32 aeth) +{ + u32 to; + + lockdep_assert_held(&qp->s_lock); + qp->s_flags |= RVT_S_WAIT_RNR; + to = rvt_aeth_to_usec(aeth); + hrtimer_start(&qp->s_rnr_timer, + ns_to_ktime(1000 * to), HRTIMER_MODE_REL); +} +EXPORT_SYMBOL(rvt_add_rnr_timer); + +/** + * rvt_stop_rc_timers - stop all timers + * @qp - the QP + * stop any pending timers + */ +void rvt_stop_rc_timers(struct rvt_qp *qp) +{ + lockdep_assert_held(&qp->s_lock); + /* Remove QP from all timers */ + if (qp->s_flags & (RVT_S_TIMER | RVT_S_WAIT_RNR)) { + qp->s_flags &= ~(RVT_S_TIMER | RVT_S_WAIT_RNR); + del_timer(&qp->s_timer); + hrtimer_try_to_cancel(&qp->s_rnr_timer); + } +} +EXPORT_SYMBOL(rvt_stop_rc_timers); + +/** + * rvt_stop_rnr_timer - stop an rnr timer + * @qp - the QP + * + * stop an rnr timer and return if the timer + * had been pending. + */ +static int 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); + } + return rval; +} + +/** + * rvt_del_timers_sync - wait for any timeout routines to exit + * @qp - the QP + */ +void rvt_del_timers_sync(struct rvt_qp *qp) +{ + del_timer_sync(&qp->s_timer); + hrtimer_cancel(&qp->s_rnr_timer); +} +EXPORT_SYMBOL(rvt_del_timers_sync); + +/** + * This is called from s_timer for missing responses. + */ +static void rvt_rc_timeout(unsigned long arg) +{ + struct rvt_qp *qp = (struct rvt_qp *)arg; + struct rvt_dev_info *rdi = ib_to_rvt(qp->ibqp.device); + unsigned long flags; + + spin_lock_irqsave(&qp->r_lock, flags); + spin_lock(&qp->s_lock); + if (qp->s_flags & RVT_S_TIMER) { + struct rvt_ibport *rvp = rdi->ports[qp->port_num - 1]; + + qp->s_flags &= ~RVT_S_TIMER; + rvp->n_rc_timeouts++; + del_timer(&qp->s_timer); + trace_rvt_rc_timeout(qp, qp->s_last_psn + 1); + if (rdi->driver_f.notify_restart_rc) + rdi->driver_f.notify_restart_rc(qp, + qp->s_last_psn + 1, + 1); + rdi->driver_f.schedule_send(qp); + } + spin_unlock(&qp->s_lock); + spin_unlock_irqrestore(&qp->r_lock, flags); +} + +/* + * This is called from s_timer for RNR timeouts. + */ +enum hrtimer_restart rvt_rc_rnr_retry(struct hrtimer *t) +{ + struct rvt_qp *qp = container_of(t, struct rvt_qp, s_rnr_timer); + struct rvt_dev_info *rdi = ib_to_rvt(qp->ibqp.device); + unsigned long flags; + + spin_lock_irqsave(&qp->s_lock, flags); + rvt_stop_rnr_timer(qp); + rdi->driver_f.schedule_send(qp); + spin_unlock_irqrestore(&qp->s_lock, flags); + return HRTIMER_NORESTART; +} +EXPORT_SYMBOL(rvt_rc_rnr_retry); diff --git a/drivers/infiniband/sw/rdmavt/rc.c b/drivers/infiniband/sw/rdmavt/rc.c new file mode 100644 index 000000000000..6131cc558bdb --- /dev/null +++ b/drivers/infiniband/sw/rdmavt/rc.c @@ -0,0 +1,189 @@ +/* + * Copyright(c) 2016 Intel Corporation. + * + * This file is provided under a dual BSD/GPLv2 license. When using or + * redistributing this file, you may do so under either license. + * + * GPL LICENSE SUMMARY + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * BSD LICENSE + * + * 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. + * - Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + */ + +#include <rdma/rdma_vt.h> +#include <rdma/ib_hdrs.h> + +/* + * Convert the AETH credit code into the number of credits. + */ +static const u16 credit_table[31] = { + 0, /* 0 */ + 1, /* 1 */ + 2, /* 2 */ + 3, /* 3 */ + 4, /* 4 */ + 6, /* 5 */ + 8, /* 6 */ + 12, /* 7 */ + 16, /* 8 */ + 24, /* 9 */ + 32, /* A */ + 48, /* B */ + 64, /* C */ + 96, /* D */ + 128, /* E */ + 192, /* F */ + 256, /* 10 */ + 384, /* 11 */ + 512, /* 12 */ + 768, /* 13 */ + 1024, /* 14 */ + 1536, /* 15 */ + 2048, /* 16 */ + 3072, /* 17 */ + 4096, /* 18 */ + 6144, /* 19 */ + 8192, /* 1A */ + 12288, /* 1B */ + 16384, /* 1C */ + 24576, /* 1D */ + 32768 /* 1E */ +}; + +/** + * rvt_compute_aeth - compute the AETH (syndrome + MSN) + * @qp: the queue pair to compute the AETH for + * + * Returns the AETH. + */ +__be32 rvt_compute_aeth(struct rvt_qp *qp) +{ + u32 aeth = qp->r_msn & IB_MSN_MASK; + + if (qp->ibqp.srq) { + /* + * Shared receive queues don't generate credits. + * Set the credit field to the invalid value. + */ + aeth |= IB_AETH_CREDIT_INVAL << IB_AETH_CREDIT_SHIFT; + } else { + u32 min, max, x; + u32 credits; + struct rvt_rwq *wq = qp->r_rq.wq; + u32 head; + u32 tail; + + /* sanity check pointers before trusting them */ + head = wq->head; + if (head >= qp->r_rq.size) + head = 0; + tail = wq->tail; + if (tail >= qp->r_rq.size) + tail = 0; + /* + * Compute the number of credits available (RWQEs). + * There is a small chance that the pair of reads are + * not atomic, which is OK, since the fuzziness is + * resolved as further ACKs go out. + */ + credits = head - tail; + if ((int)credits < 0) + credits += qp->r_rq.size; + /* + * Binary search the credit table to find the code to + * use. + */ + min = 0; + max = 31; + for (;;) { + x = (min + max) / 2; + if (credit_table[x] == credits) + break; + if (credit_table[x] > credits) { + max = x; + } else { + if (min == x) + break; + min = x; + } + } + aeth |= x << IB_AETH_CREDIT_SHIFT; + } + return cpu_to_be32(aeth); +} +EXPORT_SYMBOL(rvt_compute_aeth); + +/** + * rvt_get_credit - flush the send work queue of a QP + * @qp: the qp who's send work queue to flush + * @aeth: the Acknowledge Extended Transport Header + * + * The QP s_lock should be held. + */ +void rvt_get_credit(struct rvt_qp *qp, u32 aeth) +{ + struct rvt_dev_info *rdi = ib_to_rvt(qp->ibqp.device); + u32 credit = (aeth >> IB_AETH_CREDIT_SHIFT) & IB_AETH_CREDIT_MASK; + + lockdep_assert_held(&qp->s_lock); + /* + * If the credit is invalid, we can send + * as many packets as we like. Otherwise, we have to + * honor the credit field. + */ + if (credit == IB_AETH_CREDIT_INVAL) { + if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT)) { + qp->s_flags |= RVT_S_UNLIMITED_CREDIT; + if (qp->s_flags & RVT_S_WAIT_SSN_CREDIT) { + qp->s_flags &= ~RVT_S_WAIT_SSN_CREDIT; + rdi->driver_f.schedule_send(qp); + } + } + } else if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT)) { + /* Compute new LSN (i.e., MSN + credit) */ + credit = (aeth + credit_table[credit]) & IB_MSN_MASK; + if (rvt_cmp_msn(credit, qp->s_lsn) > 0) { + qp->s_lsn = credit; + if (qp->s_flags & RVT_S_WAIT_SSN_CREDIT) { + qp->s_flags &= ~RVT_S_WAIT_SSN_CREDIT; + rdi->driver_f.schedule_send(qp); + } + } + } +} +EXPORT_SYMBOL(rvt_get_credit); diff --git a/drivers/infiniband/sw/rdmavt/trace.h b/drivers/infiniband/sw/rdmavt/trace.h index e2d23acb6a7d..bb4b1e710f22 100644 --- a/drivers/infiniband/sw/rdmavt/trace.h +++ b/drivers/infiniband/sw/rdmavt/trace.h @@ -1,5 +1,5 @@ /* - * Copyright(c) 2016 Intel Corporation. + * Copyright(c) 2016, 2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -52,3 +52,5 @@ #include "trace_qp.h" #include "trace_tx.h" #include "trace_mr.h" +#include "trace_cq.h" +#include "trace_rc.h" diff --git a/drivers/infiniband/sw/rdmavt/trace_cq.h b/drivers/infiniband/sw/rdmavt/trace_cq.h new file mode 100644 index 000000000000..a315850aa9bb --- /dev/null +++ b/drivers/infiniband/sw/rdmavt/trace_cq.h @@ -0,0 +1,127 @@ +/* + * Copyright(c) 2016 Intel Corporation. + * + * This file is provided under a dual BSD/GPLv2 license. When using or + * redistributing this file, you may do so under either license. + * + * GPL LICENSE SUMMARY + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * BSD LICENSE + * + * 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. + * - Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + */ +#if !defined(__RVT_TRACE_CQ_H) || defined(TRACE_HEADER_MULTI_READ) +#define __RVT_TRACE_CQ_H + +#include <linux/tracepoint.h> +#include <linux/trace_seq.h> + +#include <rdma/ib_verbs.h> +#include <rdma/rdmavt_cq.h> + +#undef TRACE_SYSTEM +#define TRACE_SYSTEM rvt_cq + +#define wc_opcode_name(opcode) { IB_WC_##opcode, #opcode } +#define show_wc_opcode(opcode) \ +__print_symbolic(opcode, \ + wc_opcode_name(SEND), \ + wc_opcode_name(RDMA_WRITE), \ + wc_opcode_name(RDMA_READ), \ + wc_opcode_name(COMP_SWAP), \ + wc_opcode_name(FETCH_ADD), \ + wc_opcode_name(LSO), \ + wc_opcode_name(LOCAL_INV), \ + wc_opcode_name(REG_MR), \ + wc_opcode_name(MASKED_COMP_SWAP), \ + wc_opcode_name(RECV), \ + wc_opcode_name(RECV_RDMA_WITH_IMM)) + +#define CQ_PRN \ +"[%s] idx %u wr_id %llx status %u opcode %u,%s length %u qpn %x" + +DECLARE_EVENT_CLASS( + rvt_cq_entry_template, + TP_PROTO(struct rvt_cq *cq, struct ib_wc *wc, u32 idx), + TP_ARGS(cq, wc, idx), + TP_STRUCT__entry( + RDI_DEV_ENTRY(cq->rdi) + __field(u64, wr_id) + __field(u32, status) + __field(u32, opcode) + __field(u32, qpn) + __field(u32, length) + __field(u32, idx) + ), + TP_fast_assign( + RDI_DEV_ASSIGN(cq->rdi) + __entry->wr_id = wc->wr_id; + __entry->status = wc->status; + __entry->opcode = wc->opcode; + __entry->length = wc->byte_len; + __entry->qpn = wc->qp->qp_num; + __entry->idx = idx; + ), + TP_printk( + CQ_PRN, + __get_str(dev), + __entry->idx, + __entry->wr_id, + __entry->status, + __entry->opcode, show_wc_opcode(__entry->opcode), + __entry->length, + __entry->qpn + ) +); + +DEFINE_EVENT( + rvt_cq_entry_template, rvt_cq_enter, + TP_PROTO(struct rvt_cq *cq, struct ib_wc *wc, u32 idx), + TP_ARGS(cq, wc, idx)); + +DEFINE_EVENT( + rvt_cq_entry_template, rvt_cq_poll, + TP_PROTO(struct rvt_cq *cq, struct ib_wc *wc, u32 idx), + TP_ARGS(cq, wc, idx)); + +#endif /* __RVT_TRACE_CQ_H */ + +#undef TRACE_INCLUDE_PATH +#undef TRACE_INCLUDE_FILE +#define TRACE_INCLUDE_PATH . +#define TRACE_INCLUDE_FILE trace_cq +#include <trace/define_trace.h> diff --git a/drivers/infiniband/sw/rdmavt/dma.h b/drivers/infiniband/sw/rdmavt/trace_rc.h index 979f07e09195..995276933a55 100644 --- a/drivers/infiniband/sw/rdmavt/dma.h +++ b/drivers/infiniband/sw/rdmavt/trace_rc.h @@ -1,8 +1,5 @@ -#ifndef DEF_RDMAVTDMA_H -#define DEF_RDMAVTDMA_H - /* - * Copyright(c) 2016 Intel Corporation. + * Copyright(c) 2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -47,7 +44,66 @@ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. * */ +#if !defined(__RVT_TRACE_RC_H) || defined(TRACE_HEADER_MULTI_READ) +#define __RVT_TRACE_RC_H + +#include <linux/tracepoint.h> +#include <linux/trace_seq.h> + +#include <rdma/ib_verbs.h> +#include <rdma/rdma_vt.h> + +#undef TRACE_SYSTEM +#define TRACE_SYSTEM rvt_rc + +DECLARE_EVENT_CLASS(rvt_rc_template, + TP_PROTO(struct rvt_qp *qp, u32 psn), + TP_ARGS(qp, psn), + TP_STRUCT__entry( + RDI_DEV_ENTRY(ib_to_rvt(qp->ibqp.device)) + __field(u32, qpn) + __field(u32, s_flags) + __field(u32, psn) + __field(u32, s_psn) + __field(u32, s_next_psn) + __field(u32, s_sending_psn) + __field(u32, s_sending_hpsn) + __field(u32, r_psn) + ), + TP_fast_assign( + RDI_DEV_ASSIGN(ib_to_rvt(qp->ibqp.device)) + __entry->qpn = qp->ibqp.qp_num; + __entry->s_flags = qp->s_flags; + __entry->psn = psn; + __entry->s_psn = qp->s_psn; + __entry->s_next_psn = qp->s_next_psn; + __entry->s_sending_psn = qp->s_sending_psn; + __entry->s_sending_hpsn = qp->s_sending_hpsn; + __entry->r_psn = qp->r_psn; + ), + TP_printk( + "[%s] qpn 0x%x s_flags 0x%x psn 0x%x s_psn 0x%x s_next_psn 0x%x s_sending_psn 0x%x sending_hpsn 0x%x r_psn 0x%x", + __get_str(dev), + __entry->qpn, + __entry->s_flags, + __entry->psn, + __entry->s_psn, + __entry->s_next_psn, + __entry->s_sending_psn, + __entry->s_sending_hpsn, + __entry->r_psn + ) +); + +DEFINE_EVENT(rvt_rc_template, rvt_rc_timeout, + TP_PROTO(struct rvt_qp *qp, u32 psn), + TP_ARGS(qp, psn) +); -extern struct ib_dma_mapping_ops rvt_default_dma_mapping_ops; +#endif /* __RVT_TRACE_RC_H */ -#endif /* DEF_RDMAVTDMA_H */ +#undef TRACE_INCLUDE_PATH +#undef TRACE_INCLUDE_FILE +#define TRACE_INCLUDE_PATH . +#define TRACE_INCLUDE_FILE trace_rc +#include <trace/define_trace.h> diff --git a/drivers/infiniband/sw/rdmavt/trace_tx.h b/drivers/infiniband/sw/rdmavt/trace_tx.h index 0e03173662d8..a613a2223751 100644 --- a/drivers/infiniband/sw/rdmavt/trace_tx.h +++ b/drivers/infiniband/sw/rdmavt/trace_tx.h @@ -71,10 +71,20 @@ __print_symbolic(opcode, \ wr_opcode_name(RDMA_READ_WITH_INV), \ wr_opcode_name(LOCAL_INV), \ wr_opcode_name(MASKED_ATOMIC_CMP_AND_SWP), \ - wr_opcode_name(MASKED_ATOMIC_FETCH_AND_ADD)) + wr_opcode_name(MASKED_ATOMIC_FETCH_AND_ADD), \ + wr_opcode_name(RESERVED1), \ + wr_opcode_name(RESERVED2), \ + wr_opcode_name(RESERVED3), \ + wr_opcode_name(RESERVED4), \ + wr_opcode_name(RESERVED5), \ + wr_opcode_name(RESERVED6), \ + wr_opcode_name(RESERVED7), \ + wr_opcode_name(RESERVED8), \ + wr_opcode_name(RESERVED9), \ + wr_opcode_name(RESERVED10)) #define POS_PRN \ -"[%s] wr_id %llx qpn %x psn 0x%x lpsn 0x%x length %u opcode 0x%.2x,%s size %u avail %u head %u last %u" +"[%s] wqe %p wr_id %llx send_flags %x qpn %x qpt %u psn %x lpsn %x ssn %x length %u opcode 0x%.2x,%s size %u avail %u head %u last %u pid %u num_sge %u" TRACE_EVENT( rvt_post_one_wr, @@ -83,7 +93,9 @@ TRACE_EVENT( TP_STRUCT__entry( RDI_DEV_ENTRY(ib_to_rvt(qp->ibqp.device)) __field(u64, wr_id) + __field(struct rvt_swqe *, wqe) __field(u32, qpn) + __field(u32, qpt) __field(u32, psn) __field(u32, lpsn) __field(u32, length) @@ -92,11 +104,17 @@ TRACE_EVENT( __field(u32, avail) __field(u32, head) __field(u32, last) + __field(u32, ssn) + __field(int, send_flags) + __field(pid_t, pid) + __field(int, num_sge) ), TP_fast_assign( RDI_DEV_ASSIGN(ib_to_rvt(qp->ibqp.device)) + __entry->wqe = wqe; __entry->wr_id = wqe->wr.wr_id; __entry->qpn = qp->ibqp.qp_num; + __entry->qpt = qp->ibqp.qp_type; __entry->psn = wqe->psn; __entry->lpsn = wqe->lpsn; __entry->length = wqe->length; @@ -105,20 +123,30 @@ TRACE_EVENT( __entry->avail = qp->s_avail; __entry->head = qp->s_head; __entry->last = qp->s_last; + __entry->pid = qp->pid; + __entry->ssn = wqe->ssn; + __entry->send_flags = wqe->wr.send_flags; + __entry->num_sge = wqe->wr.num_sge; ), TP_printk( POS_PRN, __get_str(dev), + __entry->wqe, __entry->wr_id, + __entry->send_flags, __entry->qpn, + __entry->qpt, __entry->psn, __entry->lpsn, + __entry->ssn, __entry->length, __entry->opcode, show_wr_opcode(__entry->opcode), __entry->size, __entry->avail, __entry->head, - __entry->last + __entry->last, + __entry->pid, + __entry->num_sge ) ); diff --git a/drivers/infiniband/sw/rdmavt/vt.c b/drivers/infiniband/sw/rdmavt/vt.c index d430c2f7cec4..0d7c6bb551d9 100644 --- a/drivers/infiniband/sw/rdmavt/vt.c +++ b/drivers/infiniband/sw/rdmavt/vt.c @@ -47,6 +47,7 @@ #include <linux/module.h> #include <linux/kernel.h> +#include <linux/dma-mapping.h> #include "vt.h" #include "trace.h" @@ -165,7 +166,7 @@ static int rvt_query_port(struct ib_device *ibdev, u8 port_num, return -EINVAL; rvp = rdi->ports[port_index]; - memset(props, 0, sizeof(*props)); + /* props being zeroed by the caller, avoid zeroing it here */ props->sm_lid = rvp->sm_lid; props->sm_sl = rvp->sm_sl; props->port_cap_flags = rvp->port_cap_flags; @@ -326,13 +327,14 @@ static int rvt_get_port_immutable(struct ib_device *ibdev, u8 port_num, if (port_index < 0) return -EINVAL; - err = rvt_query_port(ibdev, port_num, &attr); + immutable->core_cap_flags = rdi->dparms.core_cap_flags; + + err = ib_query_port(ibdev, port_num, &attr); if (err) return err; immutable->pkey_tbl_len = attr.pkey_tbl_len; immutable->gid_tbl_len = attr.gid_tbl_len; - immutable->core_cap_flags = rdi->dparms.core_cap_flags; immutable->max_mad_size = rdi->dparms.max_mad_size; return 0; @@ -777,8 +779,7 @@ int rvt_register_device(struct rvt_dev_info *rdi) } /* DMA Operations */ - rdi->ibdev.dma_ops = - rdi->ibdev.dma_ops ? : &rvt_default_dma_mapping_ops; + rdi->ibdev.dev.dma_ops = rdi->ibdev.dev.dma_ops ? : &dma_virt_ops; /* Protection Domain */ spin_lock_init(&rdi->n_pds_lock); diff --git a/drivers/infiniband/sw/rdmavt/vt.h b/drivers/infiniband/sw/rdmavt/vt.h index 6b01eaa4461b..f363505312be 100644 --- a/drivers/infiniband/sw/rdmavt/vt.h +++ b/drivers/infiniband/sw/rdmavt/vt.h @@ -50,7 +50,6 @@ #include <rdma/rdma_vt.h> #include <linux/pci.h> -#include "dma.h" #include "pd.h" #include "qp.h" #include "ah.h" diff --git a/drivers/infiniband/sw/rxe/Kconfig b/drivers/infiniband/sw/rxe/Kconfig index 1e4e628fe7b0..320bffc980d8 100644 --- a/drivers/infiniband/sw/rxe/Kconfig +++ b/drivers/infiniband/sw/rxe/Kconfig @@ -2,6 +2,8 @@ config RDMA_RXE tristate "Software RDMA over Ethernet (RoCE) driver" depends on INET && PCI && INFINIBAND depends on NET_UDP_TUNNEL + depends on CRYPTO_CRC32 + select DMA_VIRT_OPS ---help--- This driver implements the InfiniBand RDMA transport over the Linux network stack. It enables a system with a @@ -21,4 +23,4 @@ config RDMA_RXE To configure and work with soft-RoCE driver please use the following wiki page under "configure Soft-RoCE (RXE)" section: - https://github.com/SoftRoCE/rxe-dev/wiki/rxe-dev:-Home + https://github.com/linux-rdma/rdma-core/blob/master/Documentation/rxe.md diff --git a/drivers/infiniband/sw/rxe/Makefile b/drivers/infiniband/sw/rxe/Makefile index 3b3fb9d1c470..3f12beb7076f 100644 --- a/drivers/infiniband/sw/rxe/Makefile +++ b/drivers/infiniband/sw/rxe/Makefile @@ -14,11 +14,11 @@ rdma_rxe-y := \ rxe_qp.o \ rxe_cq.o \ rxe_mr.o \ - rxe_dma.o \ rxe_opcode.o \ rxe_mmap.o \ rxe_icrc.o \ rxe_mcast.o \ rxe_task.o \ rxe_net.o \ - rxe_sysfs.o + rxe_sysfs.o \ + rxe_hw_counters.o diff --git a/drivers/infiniband/sw/rxe/rxe.c b/drivers/infiniband/sw/rxe/rxe.c index ab6c3c25d7ff..c21c913f911a 100644 --- a/drivers/infiniband/sw/rxe/rxe.c +++ b/drivers/infiniband/sw/rxe/rxe.c @@ -31,6 +31,7 @@ * SOFTWARE. */ +#include <net/addrconf.h> #include "rxe.h" #include "rxe_loc.h" @@ -64,6 +65,8 @@ static void rxe_cleanup(struct rxe_dev *rxe) rxe_pool_cleanup(&rxe->mc_elem_pool); rxe_cleanup_ports(rxe); + + crypto_free_shash(rxe->tfm); } /* called when all references have been dropped */ @@ -178,7 +181,8 @@ static int rxe_init_ports(struct rxe_dev *rxe) return -ENOMEM; port->pkey_tbl[0] = 0xffff; - port->port_guid = rxe->ifc_ops->port_guid(rxe); + addrconf_addr_eui48((unsigned char *)&port->port_guid, + rxe->ndev->dev_addr); spin_lock_init(&port->port_lock); diff --git a/drivers/infiniband/sw/rxe/rxe.h b/drivers/infiniband/sw/rxe/rxe.h index a696af81e4a5..1ac5b8551a4d 100644 --- a/drivers/infiniband/sw/rxe/rxe.h +++ b/drivers/infiniband/sw/rxe/rxe.h @@ -50,6 +50,7 @@ #include <rdma/ib_umem.h> #include <rdma/ib_cache.h> #include <rdma/ib_addr.h> +#include <crypto/hash.h> #include "rxe_net.h" #include "rxe_opcode.h" @@ -64,6 +65,28 @@ #define RXE_ROCE_V2_SPORT (0xc000) +static inline u32 rxe_crc32(struct rxe_dev *rxe, + u32 crc, void *next, size_t len) +{ + u32 retval; + int err; + + SHASH_DESC_ON_STACK(shash, rxe->tfm); + + shash->tfm = rxe->tfm; + shash->flags = 0; + *(u32 *)shash_desc_ctx(shash) = crc; + err = crypto_shash_update(shash, next, len); + if (unlikely(err)) { + pr_warn_ratelimited("failed crc calculation, err: %d\n", err); + return crc32_le(crc, next, len); + } + + retval = *(u32 *)shash_desc_ctx(shash); + barrier_data(shash_desc_ctx(shash)); + return retval; +} + int rxe_set_mtu(struct rxe_dev *rxe, unsigned int dev_mtu); int rxe_add(struct rxe_dev *rxe, unsigned int mtu); diff --git a/drivers/infiniband/sw/rxe/rxe_av.c b/drivers/infiniband/sw/rxe/rxe_av.c index 604f6fee96bd..5bddf469361b 100644 --- a/drivers/infiniband/sw/rxe/rxe_av.c +++ b/drivers/infiniband/sw/rxe/rxe_av.c @@ -34,21 +34,22 @@ #include "rxe.h" #include "rxe_loc.h" -int rxe_av_chk_attr(struct rxe_dev *rxe, struct ib_ah_attr *attr) +int rxe_av_chk_attr(struct rxe_dev *rxe, struct rdma_ah_attr *attr) { struct rxe_port *port; - if (attr->port_num != 1) { - pr_info("invalid port_num = %d\n", attr->port_num); + if (rdma_ah_get_port_num(attr) != 1) { + pr_info("invalid port_num = %d\n", rdma_ah_get_port_num(attr)); return -EINVAL; } port = &rxe->port; - if (attr->ah_flags & IB_AH_GRH) { - if (attr->grh.sgid_index > port->attr.gid_tbl_len) { - pr_info("invalid sgid index = %d\n", - attr->grh.sgid_index); + if (rdma_ah_get_ah_flags(attr) & IB_AH_GRH) { + u8 sgid_index = rdma_ah_read_grh(attr)->sgid_index; + + if (sgid_index > port->attr.gid_tbl_len) { + pr_info("invalid sgid index = %d\n", sgid_index); return -EINVAL; } } @@ -57,30 +58,33 @@ int rxe_av_chk_attr(struct rxe_dev *rxe, struct ib_ah_attr *attr) } int rxe_av_from_attr(struct rxe_dev *rxe, u8 port_num, - struct rxe_av *av, struct ib_ah_attr *attr) + struct rxe_av *av, struct rdma_ah_attr *attr) { memset(av, 0, sizeof(*av)); - memcpy(&av->grh, &attr->grh, sizeof(attr->grh)); + memcpy(&av->grh, rdma_ah_read_grh(attr), + sizeof(*rdma_ah_read_grh(attr))); av->port_num = port_num; return 0; } int rxe_av_to_attr(struct rxe_dev *rxe, struct rxe_av *av, - struct ib_ah_attr *attr) + struct rdma_ah_attr *attr) { - memcpy(&attr->grh, &av->grh, sizeof(av->grh)); - attr->port_num = av->port_num; + attr->type = RDMA_AH_ATTR_TYPE_ROCE; + memcpy(rdma_ah_retrieve_grh(attr), &av->grh, sizeof(av->grh)); + rdma_ah_set_ah_flags(attr, IB_AH_GRH); + rdma_ah_set_port_num(attr, av->port_num); return 0; } int rxe_av_fill_ip_info(struct rxe_dev *rxe, struct rxe_av *av, - struct ib_ah_attr *attr, + struct rdma_ah_attr *attr, struct ib_gid_attr *sgid_attr, union ib_gid *sgid) { rdma_gid2ip(&av->sgid_addr._sockaddr, sgid); - rdma_gid2ip(&av->dgid_addr._sockaddr, &attr->grh.dgid); + rdma_gid2ip(&av->dgid_addr._sockaddr, &rdma_ah_read_grh(attr)->dgid); av->network_type = ib_gid_to_network_type(sgid_attr->gid_type, sgid); return 0; diff --git a/drivers/infiniband/sw/rxe/rxe_comp.c b/drivers/infiniband/sw/rxe/rxe_comp.c index d369f24425f9..9eb12c2e3c74 100644 --- a/drivers/infiniband/sw/rxe/rxe_comp.c +++ b/drivers/infiniband/sw/rxe/rxe_comp.c @@ -154,6 +154,8 @@ void rxe_comp_queue_pkt(struct rxe_dev *rxe, struct rxe_qp *qp, skb_queue_tail(&qp->resp_pkts, skb); must_sched = skb_queue_len(&qp->resp_pkts) > 1; + if (must_sched != 0) + rxe_counter_inc(rxe, RXE_CNT_COMPLETER_SCHED); rxe_run_task(&qp->comp.task, must_sched); } @@ -236,6 +238,7 @@ static inline enum comp_state check_ack(struct rxe_qp *qp, { unsigned int mask = pkt->mask; u8 syn; + struct rxe_dev *rxe = to_rdev(qp->ibqp.device); /* Check the sequence only */ switch (qp->comp.opcode) { @@ -254,7 +257,7 @@ static inline enum comp_state check_ack(struct rxe_qp *qp, } break; default: - WARN_ON(1); + WARN_ON_ONCE(1); } /* Check operation validity. */ @@ -298,6 +301,7 @@ static inline enum comp_state check_ack(struct rxe_qp *qp, return COMPST_WRITE_SEND; case AETH_RNR_NAK: + rxe_counter_inc(rxe, RXE_CNT_RCV_RNR); return COMPST_RNR_RETRY; case AETH_NAK: @@ -307,6 +311,8 @@ static inline enum comp_state check_ack(struct rxe_qp *qp, * before */ if (psn_compare(pkt->psn, qp->comp.psn) > 0) { + rxe_counter_inc(rxe, + RXE_CNT_RCV_SEQ_ERR); qp->comp.psn = pkt->psn; if (qp->req.wait_psn) { qp->req.wait_psn = 0; @@ -412,13 +418,21 @@ static void make_send_cqe(struct rxe_qp *qp, struct rxe_send_wqe *wqe, } } +/* + * IBA Spec. Section 10.7.3.1 SIGNALED COMPLETIONS + * ---------8<---------8<------------- + * ...Note that if a completion error occurs, a Work Completion + * will always be generated, even if the signaling + * indicator requests an Unsignaled Completion. + * ---------8<---------8<------------- + */ static void do_complete(struct rxe_qp *qp, struct rxe_send_wqe *wqe) { struct rxe_cqe cqe; if ((qp->sq_sig_type == IB_SIGNAL_ALL_WR) || (wqe->wr.send_flags & IB_SEND_SIGNALED) || - (qp->req.state == QP_STATE_ERROR)) { + wqe->status != IB_WC_SUCCESS) { make_send_cqe(qp, wqe, &cqe); advance_consumer(qp->sq.queue); rxe_cq_post(qp->scq, &cqe, 0); @@ -503,57 +517,41 @@ static inline enum comp_state complete_wqe(struct rxe_qp *qp, return COMPST_GET_WQE; } -int rxe_completer(void *arg) +static void rxe_drain_resp_pkts(struct rxe_qp *qp, bool notify) { - struct rxe_qp *qp = (struct rxe_qp *)arg; - struct rxe_send_wqe *wqe = wqe; - struct sk_buff *skb = NULL; - struct rxe_pkt_info *pkt = NULL; - enum comp_state state; - - rxe_add_ref(qp); - - if (!qp->valid) { - while ((skb = skb_dequeue(&qp->resp_pkts))) { - rxe_drop_ref(qp); - kfree_skb(skb); - } - skb = NULL; - pkt = NULL; - - while (queue_head(qp->sq.queue)) - advance_consumer(qp->sq.queue); + struct sk_buff *skb; + struct rxe_send_wqe *wqe; - goto exit; + while ((skb = skb_dequeue(&qp->resp_pkts))) { + rxe_drop_ref(qp); + kfree_skb(skb); } - if (qp->req.state == QP_STATE_ERROR) { - while ((skb = skb_dequeue(&qp->resp_pkts))) { - rxe_drop_ref(qp); - kfree_skb(skb); - } - skb = NULL; - pkt = NULL; - - while ((wqe = queue_head(qp->sq.queue))) { + while ((wqe = queue_head(qp->sq.queue))) { + if (notify) { wqe->status = IB_WC_WR_FLUSH_ERR; do_complete(qp, wqe); + } else { + advance_consumer(qp->sq.queue); } - - goto exit; } +} - if (qp->req.state == QP_STATE_RESET) { - while ((skb = skb_dequeue(&qp->resp_pkts))) { - rxe_drop_ref(qp); - kfree_skb(skb); - } - skb = NULL; - pkt = NULL; +int rxe_completer(void *arg) +{ + struct rxe_qp *qp = (struct rxe_qp *)arg; + struct rxe_dev *rxe = to_rdev(qp->ibqp.device); + struct rxe_send_wqe *wqe = wqe; + struct sk_buff *skb = NULL; + struct rxe_pkt_info *pkt = NULL; + enum comp_state state; - while (queue_head(qp->sq.queue)) - advance_consumer(qp->sq.queue); + rxe_add_ref(qp); + if (!qp->valid || qp->req.state == QP_STATE_ERROR || + qp->req.state == QP_STATE_RESET) { + rxe_drain_resp_pkts(qp, qp->valid && + qp->req.state == QP_STATE_ERROR); goto exit; } @@ -639,6 +637,7 @@ int rxe_completer(void *arg) if (pkt) { rxe_drop_ref(pkt->qp); kfree_skb(skb); + skb = NULL; } goto done; @@ -662,6 +661,7 @@ int rxe_completer(void *arg) qp->qp_timeout_jiffies) mod_timer(&qp->retrans_timer, jiffies + qp->qp_timeout_jiffies); + WARN_ON_ONCE(skb); goto exit; case COMPST_ERROR_RETRY: @@ -674,8 +674,10 @@ int rxe_completer(void *arg) */ /* there is nothing to retry in this case */ - if (!wqe || (wqe->state == wqe_state_posted)) + if (!wqe || (wqe->state == wqe_state_posted)) { + WARN_ON_ONCE(skb); goto exit; + } if (qp->comp.retry_cnt > 0) { if (qp->comp.retry_cnt != 7) @@ -688,8 +690,10 @@ int rxe_completer(void *arg) if (psn_compare(qp->req.psn, qp->comp.psn) > 0) { /* tell the requester to retry the - * send send queue next time around + * send queue next time around */ + rxe_counter_inc(rxe, + RXE_CNT_COMP_RETRY); qp->req.need_retry = 1; rxe_run_task(&qp->req.task, 1); } @@ -697,11 +701,14 @@ int rxe_completer(void *arg) if (pkt) { rxe_drop_ref(pkt->qp); kfree_skb(skb); + skb = NULL; } + WARN_ON_ONCE(skb); goto exit; } else { + rxe_counter_inc(rxe, RXE_CNT_RETRY_EXCEEDED); wqe->status = IB_WC_RETRY_EXC_ERR; state = COMPST_ERROR; } @@ -718,22 +725,30 @@ int rxe_completer(void *arg) mod_timer(&qp->rnr_nak_timer, jiffies + rnrnak_jiffies(aeth_syn(pkt) & ~AETH_TYPE_MASK)); + rxe_drop_ref(pkt->qp); + kfree_skb(skb); + skb = NULL; goto exit; } else { + rxe_counter_inc(rxe, + RXE_CNT_RNR_RETRY_EXCEEDED); wqe->status = IB_WC_RNR_RETRY_EXC_ERR; state = COMPST_ERROR; } break; case COMPST_ERROR: + WARN_ON_ONCE(wqe->status == IB_WC_SUCCESS); do_complete(qp, wqe); rxe_qp_error(qp); if (pkt) { rxe_drop_ref(pkt->qp); kfree_skb(skb); + skb = NULL; } + WARN_ON_ONCE(skb); goto exit; } } @@ -742,6 +757,7 @@ exit: /* we come here if we are done with processing and want the task to * exit from the loop calling us */ + WARN_ON_ONCE(skb); rxe_drop_ref(qp); return -EAGAIN; @@ -749,6 +765,7 @@ done: /* we come here if we have processed a packet we want the task to call * us again to see if there is anything else to do */ + WARN_ON_ONCE(skb); rxe_drop_ref(qp); return 0; } diff --git a/drivers/infiniband/sw/rxe/rxe_cq.c b/drivers/infiniband/sw/rxe/rxe_cq.c index e5e6a5e7dee9..49fe42c23f4d 100644 --- a/drivers/infiniband/sw/rxe/rxe_cq.c +++ b/drivers/infiniband/sw/rxe/rxe_cq.c @@ -156,9 +156,9 @@ int rxe_cq_post(struct rxe_cq *cq, struct rxe_cqe *cqe, int solicited) return 0; } -void rxe_cq_cleanup(void *arg) +void rxe_cq_cleanup(struct rxe_pool_entry *arg) { - struct rxe_cq *cq = arg; + struct rxe_cq *cq = container_of(arg, typeof(*cq), pelem); if (cq->queue) rxe_queue_cleanup(cq->queue); diff --git a/drivers/infiniband/sw/rxe/rxe_dma.c b/drivers/infiniband/sw/rxe/rxe_dma.c deleted file mode 100644 index a0f8af5851ae..000000000000 --- a/drivers/infiniband/sw/rxe/rxe_dma.c +++ /dev/null @@ -1,183 +0,0 @@ -/* - * Copyright (c) 2016 Mellanox Technologies Ltd. All rights reserved. - * Copyright (c) 2015 System Fabric Works, Inc. All rights reserved. - * - * 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 "rxe.h" -#include "rxe_loc.h" - -#define DMA_BAD_ADDER ((u64)0) - -static int rxe_mapping_error(struct ib_device *dev, u64 dma_addr) -{ - return dma_addr == DMA_BAD_ADDER; -} - -static u64 rxe_dma_map_single(struct ib_device *dev, - void *cpu_addr, size_t size, - enum dma_data_direction direction) -{ - WARN_ON(!valid_dma_direction(direction)); - return (uintptr_t)cpu_addr; -} - -static void rxe_dma_unmap_single(struct ib_device *dev, - u64 addr, size_t size, - enum dma_data_direction direction) -{ - WARN_ON(!valid_dma_direction(direction)); -} - -static u64 rxe_dma_map_page(struct ib_device *dev, - struct page *page, - unsigned long offset, - size_t size, enum dma_data_direction direction) -{ - u64 addr; - - WARN_ON(!valid_dma_direction(direction)); - - if (offset + size > PAGE_SIZE) { - addr = DMA_BAD_ADDER; - goto done; - } - - addr = (uintptr_t)page_address(page); - if (addr) - addr += offset; - -done: - return addr; -} - -static void rxe_dma_unmap_page(struct ib_device *dev, - u64 addr, size_t size, - enum dma_data_direction direction) -{ - WARN_ON(!valid_dma_direction(direction)); -} - -static int rxe_map_sg(struct ib_device *dev, struct scatterlist *sgl, - int nents, enum dma_data_direction direction) -{ - struct scatterlist *sg; - u64 addr; - int i; - int ret = nents; - - WARN_ON(!valid_dma_direction(direction)); - - for_each_sg(sgl, sg, nents, i) { - addr = (uintptr_t)page_address(sg_page(sg)); - if (!addr) { - ret = 0; - break; - } - sg->dma_address = addr + sg->offset; -#ifdef CONFIG_NEED_SG_DMA_LENGTH - sg->dma_length = sg->length; -#endif - } - - return ret; -} - -static void rxe_unmap_sg(struct ib_device *dev, - struct scatterlist *sg, int nents, - enum dma_data_direction direction) -{ - WARN_ON(!valid_dma_direction(direction)); -} - -static int rxe_map_sg_attrs(struct ib_device *dev, struct scatterlist *sgl, - int nents, enum dma_data_direction direction, - unsigned long attrs) -{ - return rxe_map_sg(dev, sgl, nents, direction); -} - -static void rxe_unmap_sg_attrs(struct ib_device *dev, - struct scatterlist *sg, int nents, - enum dma_data_direction direction, - unsigned long attrs) -{ - rxe_unmap_sg(dev, sg, nents, direction); -} - -static void rxe_sync_single_for_cpu(struct ib_device *dev, - u64 addr, - size_t size, enum dma_data_direction dir) -{ -} - -static void rxe_sync_single_for_device(struct ib_device *dev, - u64 addr, - size_t size, enum dma_data_direction dir) -{ -} - -static void *rxe_dma_alloc_coherent(struct ib_device *dev, size_t size, - u64 *dma_handle, gfp_t flag) -{ - struct page *p; - void *addr = NULL; - - p = alloc_pages(flag, get_order(size)); - if (p) - addr = page_address(p); - - if (dma_handle) - *dma_handle = (uintptr_t)addr; - - return addr; -} - -static void rxe_dma_free_coherent(struct ib_device *dev, size_t size, - void *cpu_addr, u64 dma_handle) -{ - free_pages((unsigned long)cpu_addr, get_order(size)); -} - -struct ib_dma_mapping_ops rxe_dma_mapping_ops = { - .mapping_error = rxe_mapping_error, - .map_single = rxe_dma_map_single, - .unmap_single = rxe_dma_unmap_single, - .map_page = rxe_dma_map_page, - .unmap_page = rxe_dma_unmap_page, - .map_sg = rxe_map_sg, - .unmap_sg = rxe_unmap_sg, - .map_sg_attrs = rxe_map_sg_attrs, - .unmap_sg_attrs = rxe_unmap_sg_attrs, - .sync_single_for_cpu = rxe_sync_single_for_cpu, - .sync_single_for_device = rxe_sync_single_for_device, - .alloc_coherent = rxe_dma_alloc_coherent, - .free_coherent = rxe_dma_free_coherent -}; diff --git a/drivers/infiniband/sw/rxe/rxe_hdr.h b/drivers/infiniband/sw/rxe/rxe_hdr.h index d57b5e956ceb..6cb18406f5b8 100644 --- a/drivers/infiniband/sw/rxe/rxe_hdr.h +++ b/drivers/infiniband/sw/rxe/rxe_hdr.h @@ -53,8 +53,16 @@ struct rxe_pkt_info { }; /* Macros should be used only for received skb */ -#define SKB_TO_PKT(skb) ((struct rxe_pkt_info *)(skb)->cb) -#define PKT_TO_SKB(pkt) container_of((void *)(pkt), struct sk_buff, cb) +static inline struct rxe_pkt_info *SKB_TO_PKT(struct sk_buff *skb) +{ + BUILD_BUG_ON(sizeof(struct rxe_pkt_info) > sizeof(skb->cb)); + return (void *)skb->cb; +} + +static inline struct sk_buff *PKT_TO_SKB(struct rxe_pkt_info *pkt) +{ + return container_of((void *)pkt, struct sk_buff, cb); +} /* * IBA header types and methods diff --git a/drivers/infiniband/sw/rxe/rxe_hw_counters.c b/drivers/infiniband/sw/rxe/rxe_hw_counters.c new file mode 100644 index 000000000000..7ef90aad7dfd --- /dev/null +++ b/drivers/infiniband/sw/rxe/rxe_hw_counters.c @@ -0,0 +1,78 @@ +/* + * Copyright (c) 2017 Mellanox Technologies Ltd. All rights reserved. + * + * 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 "rxe.h" +#include "rxe_hw_counters.h" + +const char * const rxe_counter_name[] = { + [RXE_CNT_SENT_PKTS] = "sent_pkts", + [RXE_CNT_RCVD_PKTS] = "rcvd_pkts", + [RXE_CNT_DUP_REQ] = "duplicate_request", + [RXE_CNT_OUT_OF_SEQ_REQ] = "out_of_sequence", + [RXE_CNT_RCV_RNR] = "rcvd_rnr_err", + [RXE_CNT_SND_RNR] = "send_rnr_err", + [RXE_CNT_RCV_SEQ_ERR] = "rcvd_seq_err", + [RXE_CNT_COMPLETER_SCHED] = "ack_deffered", + [RXE_CNT_RETRY_EXCEEDED] = "retry_exceeded_err", + [RXE_CNT_RNR_RETRY_EXCEEDED] = "retry_rnr_exceeded_err", + [RXE_CNT_COMP_RETRY] = "completer_retry_err", + [RXE_CNT_SEND_ERR] = "send_err", +}; + +int rxe_ib_get_hw_stats(struct ib_device *ibdev, + struct rdma_hw_stats *stats, + u8 port, int index) +{ + struct rxe_dev *dev = to_rdev(ibdev); + unsigned int cnt; + + if (!port || !stats) + return -EINVAL; + + for (cnt = 0; cnt < ARRAY_SIZE(rxe_counter_name); cnt++) + stats->value[cnt] = dev->stats_counters[cnt]; + + return ARRAY_SIZE(rxe_counter_name); +} + +struct rdma_hw_stats *rxe_ib_alloc_hw_stats(struct ib_device *ibdev, + u8 port_num) +{ + BUILD_BUG_ON(ARRAY_SIZE(rxe_counter_name) != RXE_NUM_OF_COUNTERS); + /* We support only per port stats */ + if (!port_num) + return NULL; + + return rdma_alloc_hw_stats_struct(rxe_counter_name, + ARRAY_SIZE(rxe_counter_name), + RDMA_HW_STATS_DEFAULT_LIFESPAN); +} diff --git a/drivers/infiniband/sw/rxe/rxe_hw_counters.h b/drivers/infiniband/sw/rxe/rxe_hw_counters.h new file mode 100644 index 000000000000..f44df1b76742 --- /dev/null +++ b/drivers/infiniband/sw/rxe/rxe_hw_counters.h @@ -0,0 +1,61 @@ +/* + * Copyright (c) 2017 Mellanox Technologies Ltd. All rights reserved. + * + * 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 RXE_HW_COUNTERS_H +#define RXE_HW_COUNTERS_H + +/* + * when adding counters to enum also add + * them to rxe_counter_name[] vector. + */ +enum rxe_counters { + RXE_CNT_SENT_PKTS, + RXE_CNT_RCVD_PKTS, + RXE_CNT_DUP_REQ, + RXE_CNT_OUT_OF_SEQ_REQ, + RXE_CNT_RCV_RNR, + RXE_CNT_SND_RNR, + RXE_CNT_RCV_SEQ_ERR, + RXE_CNT_COMPLETER_SCHED, + RXE_CNT_RETRY_EXCEEDED, + RXE_CNT_RNR_RETRY_EXCEEDED, + RXE_CNT_COMP_RETRY, + RXE_CNT_SEND_ERR, + RXE_NUM_OF_COUNTERS +}; + +struct rdma_hw_stats *rxe_ib_alloc_hw_stats(struct ib_device *ibdev, + u8 port_num); +int rxe_ib_get_hw_stats(struct ib_device *ibdev, + struct rdma_hw_stats *stats, + u8 port, int index); +#endif /* RXE_HW_COUNTERS_H */ diff --git a/drivers/infiniband/sw/rxe/rxe_icrc.c b/drivers/infiniband/sw/rxe/rxe_icrc.c index 413b56b23a06..39e0be31aab1 100644 --- a/drivers/infiniband/sw/rxe/rxe_icrc.c +++ b/drivers/infiniband/sw/rxe/rxe_icrc.c @@ -87,10 +87,10 @@ u32 rxe_icrc_hdr(struct rxe_pkt_info *pkt, struct sk_buff *skb) bth->qpn |= cpu_to_be32(~BTH_QPN_MASK); length = hdr_size + RXE_BTH_BYTES; - crc = crc32_le(crc, pshdr, length); + crc = rxe_crc32(pkt->rxe, crc, pshdr, length); /* And finish to compute the CRC on the remainder of the headers. */ - crc = crc32_le(crc, pkt->hdr + RXE_BTH_BYTES, - rxe_opcode[pkt->opcode].length - RXE_BTH_BYTES); + crc = rxe_crc32(pkt->rxe, crc, pkt->hdr + RXE_BTH_BYTES, + rxe_opcode[pkt->opcode].length - RXE_BTH_BYTES); return crc; } diff --git a/drivers/infiniband/sw/rxe/rxe_loc.h b/drivers/infiniband/sw/rxe/rxe_loc.h index efe4c6a35442..d6299edf9a5b 100644 --- a/drivers/infiniband/sw/rxe/rxe_loc.h +++ b/drivers/infiniband/sw/rxe/rxe_loc.h @@ -36,17 +36,17 @@ /* rxe_av.c */ -int rxe_av_chk_attr(struct rxe_dev *rxe, struct ib_ah_attr *attr); +int rxe_av_chk_attr(struct rxe_dev *rxe, struct rdma_ah_attr *attr); int rxe_av_from_attr(struct rxe_dev *rxe, u8 port_num, - struct rxe_av *av, struct ib_ah_attr *attr); + struct rxe_av *av, struct rdma_ah_attr *attr); int rxe_av_to_attr(struct rxe_dev *rxe, struct rxe_av *av, - struct ib_ah_attr *attr); + struct rdma_ah_attr *attr); int rxe_av_fill_ip_info(struct rxe_dev *rxe, struct rxe_av *av, - struct ib_ah_attr *attr, + struct rdma_ah_attr *attr, struct ib_gid_attr *sgid_attr, union ib_gid *sgid); @@ -64,7 +64,7 @@ int rxe_cq_resize_queue(struct rxe_cq *cq, int new_cqe, struct ib_udata *udata); int rxe_cq_post(struct rxe_cq *cq, struct rxe_cqe *cqe, int solicited); -void rxe_cq_cleanup(void *arg); +void rxe_cq_cleanup(struct rxe_pool_entry *arg); /* rxe_mcast.c */ int rxe_mcast_get_grp(struct rxe_dev *rxe, union ib_gid *mgid, @@ -78,7 +78,7 @@ int rxe_mcast_drop_grp_elem(struct rxe_dev *rxe, struct rxe_qp *qp, void rxe_drop_all_mcast_groups(struct rxe_qp *qp); -void rxe_mc_cleanup(void *arg); +void rxe_mc_cleanup(struct rxe_pool_entry *arg); /* rxe_mmap.c */ struct rxe_mmap_info { @@ -137,10 +137,24 @@ int mem_check_range(struct rxe_mem *mem, u64 iova, size_t length); int rxe_mem_map_pages(struct rxe_dev *rxe, struct rxe_mem *mem, u64 *page, int num_pages, u64 iova); -void rxe_mem_cleanup(void *arg); +void rxe_mem_cleanup(struct rxe_pool_entry *arg); int advance_dma_data(struct rxe_dma_info *dma, unsigned int length); +/* rxe_net.c */ +int rxe_loopback(struct sk_buff *skb); +int rxe_send(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, + struct sk_buff *skb); +struct sk_buff *rxe_init_packet(struct rxe_dev *rxe, struct rxe_av *av, + int paylen, struct rxe_pkt_info *pkt); +int rxe_prepare(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, + struct sk_buff *skb, u32 *crc); +enum rdma_link_layer rxe_link_layer(struct rxe_dev *rxe, unsigned int port_num); +const char *rxe_parent_name(struct rxe_dev *rxe, unsigned int port_num); +struct device *rxe_dma_device(struct rxe_dev *rxe); +int rxe_mcast_add(struct rxe_dev *rxe, union ib_gid *mgid); +int rxe_mcast_delete(struct rxe_dev *rxe, union ib_gid *mgid); + /* rxe_qp.c */ int rxe_qp_chk_init(struct rxe_dev *rxe, struct ib_qp_init_attr *init); @@ -162,7 +176,7 @@ void rxe_qp_error(struct rxe_qp *qp); void rxe_qp_destroy(struct rxe_qp *qp); -void rxe_qp_cleanup(void *arg); +void rxe_qp_cleanup(struct rxe_pool_entry *arg); static inline int qp_num(struct rxe_qp *qp) { @@ -221,10 +235,9 @@ int rxe_srq_from_attr(struct rxe_dev *rxe, struct rxe_srq *srq, struct ib_srq_attr *attr, enum ib_srq_attr_mask mask, struct ib_udata *udata); -extern struct ib_dma_mapping_ops rxe_dma_mapping_ops; - void rxe_release(struct kref *kref); +void rxe_drain_req_pkts(struct rxe_qp *qp, bool notify); int rxe_completer(void *arg); int rxe_requester(void *arg); int rxe_responder(void *arg); @@ -256,13 +269,14 @@ static inline int rxe_xmit_packet(struct rxe_dev *rxe, struct rxe_qp *qp, if (pkt->mask & RXE_LOOPBACK_MASK) { memcpy(SKB_TO_PKT(skb), pkt, sizeof(*pkt)); - err = rxe->ifc_ops->loopback(skb); + err = rxe_loopback(skb); } else { - err = rxe->ifc_ops->send(rxe, pkt, skb); + err = rxe_send(rxe, pkt, skb); } if (err) { rxe->xmit_errors++; + rxe_counter_inc(rxe, RXE_CNT_SEND_ERR); return err; } @@ -272,6 +286,7 @@ static inline int rxe_xmit_packet(struct rxe_dev *rxe, struct rxe_qp *qp, rxe_run_task(&qp->comp.task, 1); } + rxe_counter_inc(rxe, RXE_CNT_SENT_PKTS); goto done; drop: diff --git a/drivers/infiniband/sw/rxe/rxe_mcast.c b/drivers/infiniband/sw/rxe/rxe_mcast.c index fa95544ca7e0..522a7942c56c 100644 --- a/drivers/infiniband/sw/rxe/rxe_mcast.c +++ b/drivers/infiniband/sw/rxe/rxe_mcast.c @@ -61,7 +61,7 @@ int rxe_mcast_get_grp(struct rxe_dev *rxe, union ib_gid *mgid, rxe_add_key(grp, mgid); - err = rxe->ifc_ops->mcast_add(rxe, mgid); + err = rxe_mcast_add(rxe, mgid); if (err) goto err2; @@ -180,11 +180,11 @@ void rxe_drop_all_mcast_groups(struct rxe_qp *qp) } } -void rxe_mc_cleanup(void *arg) +void rxe_mc_cleanup(struct rxe_pool_entry *arg) { - struct rxe_mc_grp *grp = arg; + struct rxe_mc_grp *grp = container_of(arg, typeof(*grp), pelem); struct rxe_dev *rxe = grp->rxe; rxe_drop_key(grp); - rxe->ifc_ops->mcast_delete(rxe, &grp->mgid); + rxe_mcast_delete(rxe, &grp->mgid); } diff --git a/drivers/infiniband/sw/rxe/rxe_mmap.c b/drivers/infiniband/sw/rxe/rxe_mmap.c index c572a4c09359..bd812e00988e 100644 --- a/drivers/infiniband/sw/rxe/rxe_mmap.c +++ b/drivers/infiniband/sw/rxe/rxe_mmap.c @@ -156,10 +156,10 @@ struct rxe_mmap_info *rxe_create_mmap_info(struct rxe_dev *rxe, spin_lock_bh(&rxe->mmap_offset_lock); if (rxe->mmap_offset == 0) - rxe->mmap_offset = PAGE_SIZE; + rxe->mmap_offset = ALIGN(PAGE_SIZE, SHMLBA); ip->info.offset = rxe->mmap_offset; - rxe->mmap_offset += size; + rxe->mmap_offset += ALIGN(size, SHMLBA); spin_unlock_bh(&rxe->mmap_offset_lock); diff --git a/drivers/infiniband/sw/rxe/rxe_mr.c b/drivers/infiniband/sw/rxe/rxe_mr.c index d0faca294006..e37cc89987e1 100644 --- a/drivers/infiniband/sw/rxe/rxe_mr.c +++ b/drivers/infiniband/sw/rxe/rxe_mr.c @@ -59,9 +59,11 @@ int mem_check_range(struct rxe_mem *mem, u64 iova, size_t length) case RXE_MEM_TYPE_MR: case RXE_MEM_TYPE_FMR: - return ((iova < mem->iova) || - ((iova + length) > (mem->iova + mem->length))) ? - -EFAULT : 0; + if (iova < mem->iova || + length > mem->length || + iova > mem->iova + mem->length - length) + return -EFAULT; + return 0; default: return -EFAULT; @@ -89,9 +91,9 @@ static void rxe_mem_init(int access, struct rxe_mem *mem) mem->map_shift = ilog2(RXE_BUF_PER_MAP); } -void rxe_mem_cleanup(void *arg) +void rxe_mem_cleanup(struct rxe_pool_entry *arg) { - struct rxe_mem *mem = arg; + struct rxe_mem *mem = container_of(arg, typeof(*mem), pelem); int i; if (mem->umem) @@ -123,7 +125,7 @@ static int rxe_mem_alloc(struct rxe_dev *rxe, struct rxe_mem *mem, int num_buf) goto err2; } - WARN_ON(!is_power_of_2(RXE_BUF_PER_MAP)); + BUILD_BUG_ON(!is_power_of_2(RXE_BUF_PER_MAP)); mem->map_shift = ilog2(RXE_BUF_PER_MAP); mem->map_mask = RXE_BUF_PER_MAP - 1; @@ -189,10 +191,8 @@ int rxe_mem_init_user(struct rxe_dev *rxe, struct rxe_pd *pd, u64 start, goto err1; } - WARN_ON(!is_power_of_2(umem->page_size)); - - mem->page_shift = ilog2(umem->page_size); - mem->page_mask = umem->page_size - 1; + mem->page_shift = umem->page_shift; + mem->page_mask = BIT(umem->page_shift) - 1; num_buf = 0; map = mem->map; @@ -208,7 +208,7 @@ int rxe_mem_init_user(struct rxe_dev *rxe, struct rxe_pd *pd, u64 start, } buf->addr = (uintptr_t)vaddr; - buf->size = umem->page_size; + buf->size = BIT(umem->page_shift); num_buf++; buf++; @@ -368,14 +368,15 @@ int rxe_mem_copy(struct rxe_mem *mem, u64 iova, void *addr, int length, ((void *)(uintptr_t)iova) : addr; if (crcp) - *crcp = crc32_le(*crcp, src, length); + *crcp = rxe_crc32(to_rdev(mem->pd->ibpd.device), + *crcp, src, length); memcpy(dest, src, length); return 0; } - WARN_ON(!mem->map); + WARN_ON_ONCE(!mem->map); err = mem_check_range(mem, iova, length); if (err) { @@ -401,7 +402,8 @@ int rxe_mem_copy(struct rxe_mem *mem, u64 iova, void *addr, int length, bytes = length; if (crcp) - crc = crc32_le(crc, src, bytes); + crc = rxe_crc32(to_rdev(mem->pd->ibpd.device), + crc, src, bytes); memcpy(dest, src, bytes); diff --git a/drivers/infiniband/sw/rxe/rxe_net.c b/drivers/infiniband/sw/rxe/rxe_net.c index 4abdeb359fb4..08f3f90d2912 100644 --- a/drivers/infiniband/sw/rxe/rxe_net.c +++ b/drivers/infiniband/sw/rxe/rxe_net.c @@ -84,47 +84,19 @@ struct rxe_dev *get_rxe_by_name(const char *name) struct rxe_recv_sockets recv_sockets; -static __be64 rxe_mac_to_eui64(struct net_device *ndev) -{ - unsigned char *mac_addr = ndev->dev_addr; - __be64 eui64; - unsigned char *dst = (unsigned char *)&eui64; - - dst[0] = mac_addr[0] ^ 2; - dst[1] = mac_addr[1]; - dst[2] = mac_addr[2]; - dst[3] = 0xff; - dst[4] = 0xfe; - dst[5] = mac_addr[3]; - dst[6] = mac_addr[4]; - dst[7] = mac_addr[5]; - - return eui64; -} - -static __be64 node_guid(struct rxe_dev *rxe) -{ - return rxe_mac_to_eui64(rxe->ndev); -} - -static __be64 port_guid(struct rxe_dev *rxe) -{ - return rxe_mac_to_eui64(rxe->ndev); -} - -static struct device *dma_device(struct rxe_dev *rxe) +struct device *rxe_dma_device(struct rxe_dev *rxe) { struct net_device *ndev; ndev = rxe->ndev; - if (ndev->priv_flags & IFF_802_1Q_VLAN) + if (is_vlan_dev(ndev)) ndev = vlan_dev_real_dev(ndev); return ndev->dev.parent; } -static int mcast_add(struct rxe_dev *rxe, union ib_gid *mgid) +int rxe_mcast_add(struct rxe_dev *rxe, union ib_gid *mgid) { int err; unsigned char ll_addr[ETH_ALEN]; @@ -135,7 +107,7 @@ static int mcast_add(struct rxe_dev *rxe, union ib_gid *mgid) return err; } -static int mcast_delete(struct rxe_dev *rxe, union ib_gid *mgid) +int rxe_mcast_delete(struct rxe_dev *rxe, union ib_gid *mgid) { int err; unsigned char ll_addr[ETH_ALEN]; @@ -210,6 +182,39 @@ static struct dst_entry *rxe_find_route6(struct net_device *ndev, #endif +static struct dst_entry *rxe_find_route(struct rxe_dev *rxe, + struct rxe_qp *qp, + struct rxe_av *av) +{ + struct dst_entry *dst = NULL; + + if (qp_type(qp) == IB_QPT_RC) + dst = sk_dst_get(qp->sk->sk); + + if (!dst || !(dst->obsolete && dst->ops->check(dst, 0))) { + if (dst) + dst_release(dst); + + if (av->network_type == RDMA_NETWORK_IPV4) { + struct in_addr *saddr; + struct in_addr *daddr; + + saddr = &av->sgid_addr._sockaddr_in.sin_addr; + daddr = &av->dgid_addr._sockaddr_in.sin_addr; + dst = rxe_find_route4(rxe->ndev, saddr, daddr); + } else if (av->network_type == RDMA_NETWORK_IPV6) { + struct in6_addr *saddr6; + struct in6_addr *daddr6; + + saddr6 = &av->sgid_addr._sockaddr_in6.sin6_addr; + daddr6 = &av->dgid_addr._sockaddr_in6.sin6_addr; + dst = rxe_find_route6(rxe->ndev, saddr6, daddr6); + } + } + + return dst; +} + static int rxe_udp_encap_recv(struct sock *sk, struct sk_buff *skb) { struct udphdr *udph; @@ -243,8 +248,8 @@ static struct socket *rxe_setup_udp_tunnel(struct net *net, __be16 port, { int err; struct socket *sock; - struct udp_port_cfg udp_cfg = {0}; - struct udp_tunnel_sock_cfg tnl_cfg = {0}; + struct udp_port_cfg udp_cfg = { }; + struct udp_tunnel_sock_cfg tnl_cfg = { }; if (ipv6) { udp_cfg.family = AF_INET6; @@ -301,7 +306,7 @@ static void prepare_ipv4_hdr(struct dst_entry *dst, struct sk_buff *skb, skb_scrub_packet(skb, xnet); skb_clear_hash(skb); - skb_dst_set(skb, dst); + skb_dst_set(skb, dst_clone(dst)); memset(IPCB(skb), 0, sizeof(*IPCB(skb))); skb_push(skb, sizeof(struct iphdr)); @@ -349,13 +354,14 @@ static void prepare_ipv6_hdr(struct dst_entry *dst, struct sk_buff *skb, static int prepare4(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, struct sk_buff *skb, struct rxe_av *av) { + struct rxe_qp *qp = pkt->qp; struct dst_entry *dst; bool xnet = false; __be16 df = htons(IP_DF); struct in_addr *saddr = &av->sgid_addr._sockaddr_in.sin_addr; struct in_addr *daddr = &av->dgid_addr._sockaddr_in.sin_addr; - dst = rxe_find_route4(rxe->ndev, saddr, daddr); + dst = rxe_find_route(rxe, qp, av); if (!dst) { pr_err("Host not reachable\n"); return -EHOSTUNREACH; @@ -369,17 +375,24 @@ static int prepare4(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, prepare_ipv4_hdr(dst, skb, saddr->s_addr, daddr->s_addr, IPPROTO_UDP, av->grh.traffic_class, av->grh.hop_limit, df, xnet); + + if (qp_type(qp) == IB_QPT_RC) + sk_dst_set(qp->sk->sk, dst); + else + dst_release(dst); + return 0; } static int prepare6(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, struct sk_buff *skb, struct rxe_av *av) { - struct dst_entry *dst; + struct rxe_qp *qp = pkt->qp; + struct dst_entry *dst = NULL; struct in6_addr *saddr = &av->sgid_addr._sockaddr_in6.sin6_addr; struct in6_addr *daddr = &av->dgid_addr._sockaddr_in6.sin6_addr; - dst = rxe_find_route6(rxe->ndev, saddr, daddr); + dst = rxe_find_route(rxe, qp, av); if (!dst) { pr_err("Host not reachable\n"); return -EHOSTUNREACH; @@ -394,11 +407,17 @@ static int prepare6(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, prepare_ipv6_hdr(dst, skb, saddr, daddr, IPPROTO_UDP, av->grh.traffic_class, av->grh.hop_limit); + + if (qp_type(qp) == IB_QPT_RC) + sk_dst_set(qp->sk->sk, dst); + else + dst_release(dst); + return 0; } -static int prepare(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, - struct sk_buff *skb, u32 *crc) +int rxe_prepare(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, + struct sk_buff *skb, u32 *crc) { int err = 0; struct rxe_av *av = rxe_get_av(pkt); @@ -422,10 +441,11 @@ static void rxe_skb_tx_dtor(struct sk_buff *skb) if (unlikely(qp->need_req_skb && skb_out < RXE_INFLIGHT_SKBS_PER_QP_LOW)) rxe_run_task(&qp->req.task, 1); + + rxe_drop_ref(qp); } -static int send(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, - struct sk_buff *skb) +int rxe_send(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, struct sk_buff *skb) { struct sk_buff *nskb; struct rxe_av *av; @@ -455,13 +475,14 @@ static int send(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, return -EAGAIN; } + rxe_add_ref(pkt->qp); atomic_inc(&pkt->qp->skb_out); kfree_skb(skb); return 0; } -static int loopback(struct sk_buff *skb) +int rxe_loopback(struct sk_buff *skb) { return rxe_rcv(skb); } @@ -471,8 +492,8 @@ static inline int addr_same(struct rxe_dev *rxe, struct rxe_av *av) return rxe->port.port_guid == av->grh.dgid.global.interface_id; } -static struct sk_buff *init_packet(struct rxe_dev *rxe, struct rxe_av *av, - int paylen, struct rxe_pkt_info *pkt) +struct sk_buff *rxe_init_packet(struct rxe_dev *rxe, struct rxe_av *av, + int paylen, struct rxe_pkt_info *pkt) { unsigned int hdr_len; struct sk_buff *skb; @@ -511,31 +532,16 @@ static struct sk_buff *init_packet(struct rxe_dev *rxe, struct rxe_av *av, * this is required by rxe_cfg to match rxe devices in * /sys/class/infiniband up with their underlying ethernet devices */ -static char *parent_name(struct rxe_dev *rxe, unsigned int port_num) +const char *rxe_parent_name(struct rxe_dev *rxe, unsigned int port_num) { return rxe->ndev->name; } -static enum rdma_link_layer link_layer(struct rxe_dev *rxe, - unsigned int port_num) +enum rdma_link_layer rxe_link_layer(struct rxe_dev *rxe, unsigned int port_num) { return IB_LINK_LAYER_ETHERNET; } -static struct rxe_ifc_ops ifc_ops = { - .node_guid = node_guid, - .port_guid = port_guid, - .dma_device = dma_device, - .mcast_add = mcast_add, - .mcast_delete = mcast_delete, - .prepare = prepare, - .send = send, - .loopback = loopback, - .init_packet = init_packet, - .parent_name = parent_name, - .link_layer = link_layer, -}; - struct rxe_dev *rxe_net_add(struct net_device *ndev) { int err; @@ -545,7 +551,6 @@ struct rxe_dev *rxe_net_add(struct net_device *ndev) if (!rxe) return NULL; - rxe->ifc_ops = &ifc_ops; rxe->ndev = ndev; err = rxe_add(rxe, ndev->mtu); @@ -658,7 +663,7 @@ struct notifier_block rxe_net_notifier = { .notifier_call = rxe_notify, }; -int rxe_net_ipv4_init(void) +static int rxe_net_ipv4_init(void) { recv_sockets.sk4 = rxe_setup_udp_tunnel(&init_net, htons(ROCE_V2_UDP_DPORT), false); @@ -671,7 +676,7 @@ int rxe_net_ipv4_init(void) return 0; } -int rxe_net_ipv6_init(void) +static int rxe_net_ipv6_init(void) { #if IS_ENABLED(CONFIG_IPV6) diff --git a/drivers/infiniband/sw/rxe/rxe_param.h b/drivers/infiniband/sw/rxe/rxe_param.h index 13ed2cc6eaa2..1b596fbbe251 100644 --- a/drivers/infiniband/sw/rxe/rxe_param.h +++ b/drivers/infiniband/sw/rxe/rxe_param.h @@ -114,7 +114,6 @@ enum rxe_device_param { RXE_MAX_UCONTEXT = 512, RXE_NUM_PORT = 1, - RXE_NUM_COMP_VECTORS = 1, RXE_MIN_QP_INDEX = 16, RXE_MAX_QP_INDEX = 0x00020000, diff --git a/drivers/infiniband/sw/rxe/rxe_pool.c b/drivers/infiniband/sw/rxe/rxe_pool.c index d723947a8542..75d11ee635ec 100644 --- a/drivers/infiniband/sw/rxe/rxe_pool.c +++ b/drivers/infiniband/sw/rxe/rxe_pool.c @@ -102,7 +102,7 @@ struct rxe_type_info rxe_type_info[RXE_NUM_TYPES] = { }, }; -static inline char *pool_name(struct rxe_pool *pool) +static inline const char *pool_name(struct rxe_pool *pool) { return rxe_type_info[pool->type].name; } @@ -112,13 +112,6 @@ static inline struct kmem_cache *pool_cache(struct rxe_pool *pool) return rxe_type_info[pool->type].cache; } -static inline enum rxe_elem_type rxe_type(void *arg) -{ - struct rxe_pool_entry *elem = arg; - - return elem->pool->type; -} - int rxe_cache_init(void) { int err; @@ -273,6 +266,7 @@ static u32 alloc_index(struct rxe_pool *pool) if (index >= range) index = find_first_zero_bit(pool->table, range); + WARN_ON_ONCE(index >= range); set_bit(index, pool->table); pool->last = index; return index + pool->min_index; @@ -461,7 +455,7 @@ void *rxe_pool_get_index(struct rxe_pool *pool, u32 index) out: spin_unlock_irqrestore(&pool->pool_lock, flags); - return node ? (void *)elem : NULL; + return node ? elem : NULL; } void *rxe_pool_get_key(struct rxe_pool *pool, void *key) @@ -497,5 +491,5 @@ void *rxe_pool_get_key(struct rxe_pool *pool, void *key) out: spin_unlock_irqrestore(&pool->pool_lock, flags); - return node ? ((void *)elem) : NULL; + return node ? elem : NULL; } diff --git a/drivers/infiniband/sw/rxe/rxe_pool.h b/drivers/infiniband/sw/rxe/rxe_pool.h index 4d04830adcae..47df28e43acf 100644 --- a/drivers/infiniband/sw/rxe/rxe_pool.h +++ b/drivers/infiniband/sw/rxe/rxe_pool.h @@ -57,10 +57,12 @@ enum rxe_elem_type { RXE_NUM_TYPES, /* keep me last */ }; +struct rxe_pool_entry; + struct rxe_type_info { - char *name; + const char *name; size_t size; - void (*cleanup)(void *obj); + void (*cleanup)(struct rxe_pool_entry *obj); enum rxe_pool_flags flags; u32 max_index; u32 min_index; @@ -91,7 +93,7 @@ struct rxe_pool { spinlock_t pool_lock; /* pool spinlock */ size_t elem_size; struct kref ref_cnt; - void (*cleanup)(void *obj); + void (*cleanup)(struct rxe_pool_entry *obj); enum rxe_pool_state state; enum rxe_pool_flags flags; enum rxe_elem_type type; diff --git a/drivers/infiniband/sw/rxe/rxe_qp.c b/drivers/infiniband/sw/rxe/rxe_qp.c index 44b2108253bd..80ccc7c7c341 100644 --- a/drivers/infiniband/sw/rxe/rxe_qp.c +++ b/drivers/infiniband/sw/rxe/rxe_qp.c @@ -273,15 +273,11 @@ static int rxe_qp_init_req(struct rxe_dev *rxe, struct rxe_qp *qp, rxe_init_task(rxe, &qp->comp.task, qp, rxe_completer, "comp"); - init_timer(&qp->rnr_nak_timer); - qp->rnr_nak_timer.function = rnr_nak_timer; - qp->rnr_nak_timer.data = (unsigned long)qp; - - init_timer(&qp->retrans_timer); - qp->retrans_timer.function = retransmit_timer; - qp->retrans_timer.data = (unsigned long)qp; qp->qp_timeout_jiffies = 0; /* Can't be set for UD/UC in modify_qp */ - + if (init->qp_type == IB_QPT_RC) { + setup_timer(&qp->rnr_nak_timer, rnr_nak_timer, (unsigned long)qp); + setup_timer(&qp->retrans_timer, retransmit_timer, (unsigned long)qp); + } return 0; } @@ -635,8 +631,8 @@ int rxe_qp_from_attr(struct rxe_qp *qp, struct ib_qp_attr *attr, int mask, if (mask & IB_QP_AV) { ib_get_cached_gid(&rxe->ib_dev, 1, - attr->ah_attr.grh.sgid_index, &sgid, - &sgid_attr); + rdma_ah_read_grh(&attr->ah_attr)->sgid_index, + &sgid, &sgid_attr); rxe_av_from_attr(rxe, attr->port_num, &qp->pri_av, &attr->ah_attr); rxe_av_fill_ip_info(rxe, &qp->pri_av, &attr->ah_attr, @@ -646,9 +642,11 @@ int rxe_qp_from_attr(struct rxe_qp *qp, struct ib_qp_attr *attr, int mask, } if (mask & IB_QP_ALT_PATH) { - ib_get_cached_gid(&rxe->ib_dev, 1, - attr->alt_ah_attr.grh.sgid_index, &sgid, - &sgid_attr); + u8 sgid_index = + rdma_ah_read_grh(&attr->alt_ah_attr)->sgid_index; + + ib_get_cached_gid(&rxe->ib_dev, 1, sgid_index, + &sgid, &sgid_attr); rxe_av_from_attr(rxe, attr->alt_port_num, &qp->alt_av, &attr->alt_ah_attr); @@ -809,8 +807,10 @@ void rxe_qp_destroy(struct rxe_qp *qp) qp->qp_timeout_jiffies = 0; rxe_cleanup_task(&qp->resp.task); - del_timer_sync(&qp->retrans_timer); - del_timer_sync(&qp->rnr_nak_timer); + if (qp_type(qp) == IB_QPT_RC) { + del_timer_sync(&qp->retrans_timer); + del_timer_sync(&qp->rnr_nak_timer); + } rxe_cleanup_task(&qp->req.task); rxe_cleanup_task(&qp->comp.task); @@ -824,9 +824,9 @@ void rxe_qp_destroy(struct rxe_qp *qp) } /* called when the last reference to the qp is dropped */ -void rxe_qp_cleanup(void *arg) +void rxe_qp_cleanup(struct rxe_pool_entry *arg) { - struct rxe_qp *qp = arg; + struct rxe_qp *qp = container_of(arg, typeof(*qp), pelem); rxe_drop_all_mcast_groups(qp); @@ -851,6 +851,14 @@ void rxe_qp_cleanup(void *arg) qp->resp.mr = NULL; } + if (qp_type(qp) == IB_QPT_RC) { + struct dst_entry *dst = NULL; + + dst = sk_dst_get(qp->sk->sk); + if (dst) + dst_release(dst); + } + free_rd_atomic_resources(qp); kernel_sock_shutdown(qp->sk, SHUT_RDWR); diff --git a/drivers/infiniband/sw/rxe/rxe_recv.c b/drivers/infiniband/sw/rxe/rxe_recv.c index 252b4d637d45..fb8c83e055e1 100644 --- a/drivers/infiniband/sw/rxe/rxe_recv.c +++ b/drivers/infiniband/sw/rxe/rxe_recv.c @@ -387,9 +387,9 @@ int rxe_rcv(struct sk_buff *skb) pack_icrc = be32_to_cpu(*icrcp); calc_icrc = rxe_icrc_hdr(pkt, skb); - calc_icrc = crc32_le(calc_icrc, (u8 *)payload_addr(pkt), - payload_size(pkt)); - calc_icrc = cpu_to_be32(~calc_icrc); + calc_icrc = rxe_crc32(rxe, calc_icrc, (u8 *)payload_addr(pkt), + payload_size(pkt)); + calc_icrc = (__force u32)cpu_to_be32(~calc_icrc); if (unlikely(calc_icrc != pack_icrc)) { if (skb->protocol == htons(ETH_P_IPV6)) pr_warn_ratelimited("bad ICRC from %pI6c\n", @@ -403,6 +403,8 @@ int rxe_rcv(struct sk_buff *skb) goto drop; } + rxe_counter_inc(rxe, RXE_CNT_RCVD_PKTS); + if (unlikely(bth_qpn(pkt) == IB_MULTICAST_QPN)) rxe_rcv_mcast_pkt(rxe, skb); else @@ -417,4 +419,3 @@ drop: kfree_skb(skb); return 0; } -EXPORT_SYMBOL(rxe_rcv); diff --git a/drivers/infiniband/sw/rxe/rxe_req.c b/drivers/infiniband/sw/rxe/rxe_req.c index 73d4a97603a1..7ee465d1a1e1 100644 --- a/drivers/infiniband/sw/rxe/rxe_req.c +++ b/drivers/infiniband/sw/rxe/rxe_req.c @@ -32,6 +32,7 @@ */ #include <linux/skbuff.h> +#include <crypto/hash.h> #include "rxe.h" #include "rxe_loc.h" @@ -361,19 +362,14 @@ static inline int check_init_depth(struct rxe_qp *qp, struct rxe_send_wqe *wqe) return -EAGAIN; } -static inline int get_mtu(struct rxe_qp *qp, struct rxe_send_wqe *wqe) +static inline int get_mtu(struct rxe_qp *qp) { struct rxe_dev *rxe = to_rdev(qp->ibqp.device); - struct rxe_port *port; - struct rxe_av *av; if ((qp_type(qp) == IB_QPT_RC) || (qp_type(qp) == IB_QPT_UC)) return qp->mtu; - av = &wqe->av; - port = &rxe->port; - - return port->mtu_cap; + return rxe->port.mtu_cap; } static struct sk_buff *init_req_packet(struct rxe_qp *qp, @@ -409,7 +405,7 @@ static struct sk_buff *init_req_packet(struct rxe_qp *qp, /* init skb */ av = rxe_get_av(pkt); - skb = rxe->ifc_ops->init_packet(rxe, av, paylen, pkt); + skb = rxe_init_packet(rxe, av, paylen, pkt); if (unlikely(!skb)) return NULL; @@ -480,7 +476,7 @@ static int fill_packet(struct rxe_qp *qp, struct rxe_send_wqe *wqe, u32 *p; int err; - err = rxe->ifc_ops->prepare(rxe, pkt, skb, &crc); + err = rxe_prepare(rxe, pkt, skb, &crc); if (err) return err; @@ -488,8 +484,7 @@ static int fill_packet(struct rxe_qp *qp, struct rxe_send_wqe *wqe, if (wqe->wr.send_flags & IB_SEND_INLINE) { u8 *tmp = &wqe->dma.inline_data[wqe->dma.sge_offset]; - crc = crc32_le(crc, tmp, paylen); - + crc = rxe_crc32(rxe, crc, tmp, paylen); memcpy(payload_addr(pkt), tmp, paylen); wqe->dma.resid -= paylen; @@ -599,9 +594,14 @@ int rxe_requester(void *arg) rxe_add_ref(qp); next_wqe: - if (unlikely(!qp->valid || qp->req.state == QP_STATE_ERROR)) + if (unlikely(!qp->valid)) goto exit; + if (unlikely(qp->req.state == QP_STATE_ERROR)) { + rxe_drain_req_pkts(qp, true); + goto exit; + } + if (unlikely(qp->req.state == QP_STATE_RESET)) { qp->req.wqe_index = consumer_index(qp->sq.queue); qp->req.opcode = -1; @@ -635,6 +635,7 @@ next_wqe: goto exit; } rmr->state = RXE_MEM_STATE_FREE; + rxe_drop_ref(rmr); wqe->state = wqe_state_done; wqe->status = IB_WC_SUCCESS; } else if (wqe->wr.opcode == IB_WR_REG_MR) { @@ -679,7 +680,7 @@ next_wqe: goto exit; } - mtu = get_mtu(qp, wqe); + mtu = get_mtu(qp); payload = (mask & RXE_WRITE_OR_SEND) ? wqe->dma.resid : 0; if (payload > mtu) { if (qp_type(qp) == IB_QPT_UD) { @@ -728,11 +729,11 @@ next_wqe: ret = rxe_xmit_packet(to_rdev(qp->ibqp.device), qp, &pkt, skb); if (ret) { qp->need_req_skb = 1; - kfree_skb(skb); rollback_state(wqe, qp, &rollback_wqe, rollback_psn); if (ret == -EAGAIN) { + kfree_skb(skb); rxe_run_task(&qp->req.task, 1); goto exit; } @@ -748,17 +749,8 @@ err: kfree_skb(skb); wqe->status = IB_WC_LOC_PROT_ERR; wqe->state = wqe_state_error; - - /* - * IBA Spec. Section 10.7.3.1 SIGNALED COMPLETIONS - * ---------8<---------8<------------- - * ...Note that if a completion error occurs, a Work Completion - * will always be generated, even if the signaling - * indicator requests an Unsignaled Completion. - * ---------8<---------8<------------- - */ - wqe->wr.send_flags |= IB_SEND_SIGNALED; __rxe_do_task(&qp->comp.task); + exit: rxe_drop_ref(qp); return -EAGAIN; diff --git a/drivers/infiniband/sw/rxe/rxe_resp.c b/drivers/infiniband/sw/rxe/rxe_resp.c index 3435efff8799..a958ee918a49 100644 --- a/drivers/infiniband/sw/rxe/rxe_resp.c +++ b/drivers/infiniband/sw/rxe/rxe_resp.c @@ -149,6 +149,7 @@ static enum resp_states check_psn(struct rxe_qp *qp, struct rxe_pkt_info *pkt) { int diff = psn_compare(pkt->psn, qp->resp.psn); + struct rxe_dev *rxe = to_rdev(qp->ibqp.device); switch (qp_type(qp)) { case IB_QPT_RC: @@ -157,9 +158,11 @@ static enum resp_states check_psn(struct rxe_qp *qp, return RESPST_CLEANUP; qp->resp.sent_psn_nak = 1; + rxe_counter_inc(rxe, RXE_CNT_OUT_OF_SEQ_REQ); return RESPST_ERR_PSN_OUT_OF_SEQ; } else if (diff < 0) { + rxe_counter_inc(rxe, RXE_CNT_DUP_REQ); return RESPST_DUPLICATE_REQUEST; } @@ -307,7 +310,7 @@ static enum resp_states check_op_valid(struct rxe_qp *qp, break; default: - WARN_ON(1); + WARN_ON_ONCE(1); break; } @@ -418,7 +421,7 @@ static enum resp_states check_length(struct rxe_qp *qp, static enum resp_states check_rkey(struct rxe_qp *qp, struct rxe_pkt_info *pkt) { - struct rxe_mem *mem; + struct rxe_mem *mem = NULL; u64 va; u32 rkey; u32 resid; @@ -459,50 +462,48 @@ static enum resp_states check_rkey(struct rxe_qp *qp, mem = lookup_mem(qp->pd, access, rkey, lookup_remote); if (!mem) { state = RESPST_ERR_RKEY_VIOLATION; - goto err1; + goto err; } if (unlikely(mem->state == RXE_MEM_STATE_FREE)) { state = RESPST_ERR_RKEY_VIOLATION; - goto err1; + goto err; } if (mem_check_range(mem, va, resid)) { state = RESPST_ERR_RKEY_VIOLATION; - goto err2; + goto err; } if (pkt->mask & RXE_WRITE_MASK) { if (resid > mtu) { if (pktlen != mtu || bth_pad(pkt)) { state = RESPST_ERR_LENGTH; - goto err2; + goto err; } - - resid = mtu; } else { if (pktlen != resid) { state = RESPST_ERR_LENGTH; - goto err2; + goto err; } if ((bth_pad(pkt) != (0x3 & (-resid)))) { /* This case may not be exactly that * but nothing else fits. */ state = RESPST_ERR_LENGTH; - goto err2; + goto err; } } } - WARN_ON(qp->resp.mr); + WARN_ON_ONCE(qp->resp.mr); qp->resp.mr = mem; return RESPST_EXECUTE; -err2: - rxe_drop_ref(mem); -err1: +err: + if (mem) + rxe_drop_ref(mem); return state; } @@ -608,7 +609,7 @@ static struct sk_buff *prepare_ack_packet(struct rxe_qp *qp, pad = (-payload) & 0x3; paylen = rxe_opcode[opcode].length + payload + pad + RXE_ICRC_SIZE; - skb = rxe->ifc_ops->init_packet(rxe, &qp->pri_av, paylen, ack); + skb = rxe_init_packet(rxe, &qp->pri_av, paylen, ack); if (!skb) return NULL; @@ -637,7 +638,7 @@ static struct sk_buff *prepare_ack_packet(struct rxe_qp *qp, if (ack->mask & RXE_ATMACK_MASK) atmack_set_orig(ack, qp->resp.atomic_orig); - err = rxe->ifc_ops->prepare(rxe, ack, skb, &crc); + err = rxe_prepare(rxe, ack, skb, &crc); if (err) { kfree_skb(skb); return NULL; @@ -808,12 +809,10 @@ static enum resp_states execute(struct rxe_qp *qp, struct rxe_pkt_info *pkt) err = process_atomic(qp, pkt); if (err) return err; - } else + } else { /* Unreachable */ - WARN_ON(1); - - /* We successfully processed this new request. */ - qp->resp.msn++; + WARN_ON_ONCE(1); + } /* next expected psn, read handles this separately */ qp->resp.psn = (pkt->psn + 1) & BTH_PSN_MASK; @@ -821,9 +820,11 @@ static enum resp_states execute(struct rxe_qp *qp, struct rxe_pkt_info *pkt) qp->resp.opcode = pkt->opcode; qp->resp.status = IB_WC_SUCCESS; - if (pkt->mask & RXE_COMP_MASK) + if (pkt->mask & RXE_COMP_MASK) { + /* We successfully processed this new request. */ + qp->resp.msn++; return RESPST_COMPLETE; - else if (qp_type(qp) == IB_QPT_RC) + } else if (qp_type(qp) == IB_QPT_RC) return RESPST_ACKNOWLEDGE; else return RESPST_CLEANUP; @@ -906,6 +907,7 @@ static enum resp_states do_complete(struct rxe_qp *qp, return RESPST_ERROR; } rmr->state = RXE_MEM_STATE_FREE; + rxe_drop_ref(rmr); } wc->qp = &qp->ibqp; @@ -993,7 +995,9 @@ static int send_atomic_ack(struct rxe_qp *qp, struct rxe_pkt_info *pkt, free_rd_atomic_resource(qp, res); rxe_advance_resp_resource(qp); - memcpy(SKB_TO_PKT(skb), &ack_pkt, sizeof(skb->cb)); + memcpy(SKB_TO_PKT(skb), &ack_pkt, sizeof(ack_pkt)); + memset((unsigned char *)SKB_TO_PKT(skb) + sizeof(ack_pkt), 0, + sizeof(skb->cb) - sizeof(ack_pkt)); res->type = RXE_ATOMIC_MASK; res->atomic.skb = skb; @@ -1206,9 +1210,26 @@ static enum resp_states do_class_d1e_error(struct rxe_qp *qp) } } +void rxe_drain_req_pkts(struct rxe_qp *qp, bool notify) +{ + struct sk_buff *skb; + + while ((skb = skb_dequeue(&qp->req_pkts))) { + rxe_drop_ref(qp); + kfree_skb(skb); + } + + if (notify) + return; + + while (!qp->srq && qp->rq.queue && queue_head(qp->rq.queue)) + advance_consumer(qp->rq.queue); +} + int rxe_responder(void *arg) { struct rxe_qp *qp = (struct rxe_qp *)arg; + struct rxe_dev *rxe = to_rdev(qp->ibqp.device); enum resp_states state; struct rxe_pkt_info *pkt = NULL; int ret = 0; @@ -1297,6 +1318,7 @@ int rxe_responder(void *arg) break; case RESPST_ERR_RNR: if (qp_type(qp) == IB_QPT_RC) { + rxe_counter_inc(rxe, RXE_CNT_SND_RNR); /* RC - class B */ send_ack(qp, pkt, AETH_RNR_NAK | (~AETH_TYPE_MASK & @@ -1373,21 +1395,10 @@ int rxe_responder(void *arg) goto exit; - case RESPST_RESET: { - struct sk_buff *skb; - - while ((skb = skb_dequeue(&qp->req_pkts))) { - rxe_drop_ref(qp); - kfree_skb(skb); - } - - while (!qp->srq && qp->rq.queue && - queue_head(qp->rq.queue)) - advance_consumer(qp->rq.queue); - + case RESPST_RESET: + rxe_drain_req_pkts(qp, false); qp->resp.wqe = NULL; goto exit; - } case RESPST_ERROR: qp->resp.goto_error = 0; @@ -1396,7 +1407,7 @@ int rxe_responder(void *arg) goto exit; default: - WARN_ON(1); + WARN_ON_ONCE(1); } } diff --git a/drivers/infiniband/sw/rxe/rxe_verbs.c b/drivers/infiniband/sw/rxe/rxe_verbs.c index beb7021ff18a..af90a7d42b96 100644 --- a/drivers/infiniband/sw/rxe/rxe_verbs.c +++ b/drivers/infiniband/sw/rxe/rxe_verbs.c @@ -31,9 +31,12 @@ * SOFTWARE. */ +#include <linux/dma-mapping.h> +#include <net/addrconf.h> #include "rxe.h" #include "rxe_loc.h" #include "rxe_queue.h" +#include "rxe_hw_counters.h" static int rxe_query_device(struct ib_device *dev, struct ib_device_attr *attr, @@ -86,6 +89,7 @@ static int rxe_query_port(struct ib_device *dev, port = &rxe->port; + /* *attr being zeroed by the caller, avoid zeroing it here */ *attr = port->attr; mutex_lock(&rxe->usdev_lock); @@ -168,7 +172,7 @@ static int rxe_query_pkey(struct ib_device *device, struct rxe_port *port; if (unlikely(port_num != 1)) { - dev_warn(device->dma_device, "invalid port_num = %d\n", + dev_warn(device->dev.parent, "invalid port_num = %d\n", port_num); goto err1; } @@ -176,7 +180,7 @@ static int rxe_query_pkey(struct ib_device *device, port = &rxe->port; if (unlikely(index >= port->attr.pkey_tbl_len)) { - dev_warn(device->dma_device, "invalid index = %d\n", + dev_warn(device->dev.parent, "invalid index = %d\n", index); goto err1; } @@ -234,7 +238,7 @@ static enum rdma_link_layer rxe_get_link_layer(struct ib_device *dev, { struct rxe_dev *rxe = to_rdev(dev); - return rxe->ifc_ops->link_layer(rxe, port_num); + return rxe_link_layer(rxe, port_num); } static struct ib_ucontext *rxe_alloc_ucontext(struct ib_device *dev, @@ -261,13 +265,14 @@ static int rxe_port_immutable(struct ib_device *dev, u8 port_num, int err; struct ib_port_attr attr; - err = rxe_query_port(dev, port_num, &attr); + immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE_UDP_ENCAP; + + err = ib_query_port(dev, port_num, &attr); if (err) return err; immutable->pkey_tbl_len = attr.pkey_tbl_len; immutable->gid_tbl_len = attr.gid_tbl_len; - immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE_UDP_ENCAP; immutable->max_mad_size = IB_MGMT_MAD_SIZE; return 0; @@ -292,22 +297,22 @@ static int rxe_dealloc_pd(struct ib_pd *ibpd) return 0; } -static int rxe_init_av(struct rxe_dev *rxe, struct ib_ah_attr *attr, +static int rxe_init_av(struct rxe_dev *rxe, struct rdma_ah_attr *attr, struct rxe_av *av) { int err; union ib_gid sgid; struct ib_gid_attr sgid_attr; - err = ib_get_cached_gid(&rxe->ib_dev, attr->port_num, - attr->grh.sgid_index, &sgid, + err = ib_get_cached_gid(&rxe->ib_dev, rdma_ah_get_port_num(attr), + rdma_ah_read_grh(attr)->sgid_index, &sgid, &sgid_attr); if (err) { pr_err("Failed to query sgid. err = %d\n", err); return err; } - err = rxe_av_from_attr(rxe, attr->port_num, av, attr); + err = rxe_av_from_attr(rxe, rdma_ah_get_port_num(attr), av, attr); if (!err) err = rxe_av_fill_ip_info(rxe, av, attr, &sgid_attr, &sgid); @@ -316,7 +321,8 @@ static int rxe_init_av(struct rxe_dev *rxe, struct ib_ah_attr *attr, return err; } -static struct ib_ah *rxe_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr, +static struct ib_ah *rxe_create_ah(struct ib_pd *ibpd, + struct rdma_ah_attr *attr, struct ib_udata *udata) { @@ -351,7 +357,7 @@ err1: return ERR_PTR(err); } -static int rxe_modify_ah(struct ib_ah *ibah, struct ib_ah_attr *attr) +static int rxe_modify_ah(struct ib_ah *ibah, struct rdma_ah_attr *attr) { int err; struct rxe_dev *rxe = to_rdev(ibah->device); @@ -368,11 +374,13 @@ static int rxe_modify_ah(struct ib_ah *ibah, struct ib_ah_attr *attr) return 0; } -static int rxe_query_ah(struct ib_ah *ibah, struct ib_ah_attr *attr) +static int rxe_query_ah(struct ib_ah *ibah, struct rdma_ah_attr *attr) { struct rxe_dev *rxe = to_rdev(ibah->device); struct rxe_ah *ah = to_rah(ibah); + memset(attr, 0, sizeof(*attr)); + attr->type = ibah->type; rxe_av_to_attr(rxe, &ah->av, attr); return 0; } @@ -732,13 +740,8 @@ static int init_send_wqe(struct rxe_qp *qp, struct ib_send_wr *ibwr, sge = ibwr->sg_list; for (i = 0; i < num_sge; i++, sge++) { - if (qp->is_user && copy_from_user(p, (__user void *) - (uintptr_t)sge->addr, sge->length)) - return -EFAULT; - - else if (!qp->is_user) - memcpy(p, (void *)(uintptr_t)sge->addr, - sge->length); + memcpy(p, (void *)(uintptr_t)sge->addr, + sge->length); p += sge->length; } @@ -911,6 +914,9 @@ static int rxe_post_recv(struct ib_qp *ibqp, struct ib_recv_wr *wr, spin_unlock_irqrestore(&rq->producer_lock, flags); + if (qp->resp.state == QP_STATE_ERROR) + rxe_run_task(&qp->resp.task, 1); + err1: return err; } @@ -1209,10 +1215,8 @@ static ssize_t rxe_show_parent(struct device *device, { struct rxe_dev *rxe = container_of(device, struct rxe_dev, ib_dev.dev); - char *name; - name = rxe->ifc_ops->parent_name(rxe, 1); - return snprintf(buf, 16, "%s\n", name); + return snprintf(buf, 16, "%s\n", rxe_parent_name(rxe, 1)); } static DEVICE_ATTR(parent, S_IRUGO, rxe_show_parent, NULL); @@ -1233,11 +1237,14 @@ int rxe_register_device(struct rxe_dev *rxe) dev->owner = THIS_MODULE; dev->node_type = RDMA_NODE_IB_CA; dev->phys_port_cnt = 1; - dev->num_comp_vectors = RXE_NUM_COMP_VECTORS; - dev->dma_device = rxe->ifc_ops->dma_device(rxe); + dev->num_comp_vectors = num_possible_cpus(); + dev->dev.parent = rxe_dma_device(rxe); dev->local_dma_lkey = 0; - dev->node_guid = rxe->ifc_ops->node_guid(rxe); - dev->dma_ops = &rxe_dma_mapping_ops; + addrconf_addr_eui48((unsigned char *)&dev->node_guid, + rxe->ndev->dev_addr); + dev->dev.dma_ops = &dma_virt_ops; + dma_coerce_mask_and_coherent(&dev->dev, + dma_get_required_mask(dev->dev.parent)); dev->uverbs_abi_ver = RXE_UVERBS_ABI_VERSION; dev->uverbs_cmd_mask = BIT_ULL(IB_USER_VERBS_CMD_GET_CONTEXT) @@ -1317,6 +1324,15 @@ int rxe_register_device(struct rxe_dev *rxe) dev->map_mr_sg = rxe_map_mr_sg; dev->attach_mcast = rxe_attach_mcast; dev->detach_mcast = rxe_detach_mcast; + dev->get_hw_stats = rxe_ib_get_hw_stats; + dev->alloc_hw_stats = rxe_ib_alloc_hw_stats; + + rxe->tfm = crypto_alloc_shash("crc32", 0, 0); + if (IS_ERR(rxe->tfm)) { + pr_err("failed to allocate crc algorithm err:%ld\n", + PTR_ERR(rxe->tfm)); + return PTR_ERR(rxe->tfm); + } err = ib_register_device(dev, NULL); if (err) { @@ -1338,6 +1354,8 @@ int rxe_register_device(struct rxe_dev *rxe) err2: ib_unregister_device(dev); err1: + crypto_free_shash(rxe->tfm); + return err; } diff --git a/drivers/infiniband/sw/rxe/rxe_verbs.h b/drivers/infiniband/sw/rxe/rxe_verbs.h index cac1d52a08f0..5a180fbe40d9 100644 --- a/drivers/infiniband/sw/rxe/rxe_verbs.h +++ b/drivers/infiniband/sw/rxe/rxe_verbs.h @@ -38,6 +38,7 @@ #include <rdma/rdma_user_rxe.h> #include "rxe_pool.h" #include "rxe_task.h" +#include "rxe_hw_counters.h" static inline int pkey_match(u16 key1, u16 key2) { @@ -372,26 +373,6 @@ struct rxe_port { u32 qp_gsi_index; }; -/* callbacks from rdma_rxe to network interface layer */ -struct rxe_ifc_ops { - void (*release)(struct rxe_dev *rxe); - __be64 (*node_guid)(struct rxe_dev *rxe); - __be64 (*port_guid)(struct rxe_dev *rxe); - struct device *(*dma_device)(struct rxe_dev *rxe); - int (*mcast_add)(struct rxe_dev *rxe, union ib_gid *mgid); - int (*mcast_delete)(struct rxe_dev *rxe, union ib_gid *mgid); - int (*prepare)(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, - struct sk_buff *skb, u32 *crc); - int (*send)(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, - struct sk_buff *skb); - int (*loopback)(struct sk_buff *skb); - struct sk_buff *(*init_packet)(struct rxe_dev *rxe, struct rxe_av *av, - int paylen, struct rxe_pkt_info *pkt); - char *(*parent_name)(struct rxe_dev *rxe, unsigned int port_num); - enum rdma_link_layer (*link_layer)(struct rxe_dev *rxe, - unsigned int port_num); -}; - struct rxe_dev { struct ib_device ib_dev; struct ib_device_attr attr; @@ -400,8 +381,6 @@ struct rxe_dev { struct kref ref_cnt; struct mutex usdev_lock; - struct rxe_ifc_ops *ifc_ops; - struct net_device *ndev; int xmit_errors; @@ -423,10 +402,18 @@ struct rxe_dev { spinlock_t mmap_offset_lock; /* guard mmap_offset */ int mmap_offset; + u64 stats_counters[RXE_NUM_OF_COUNTERS]; + struct rxe_port port; struct list_head list; + struct crypto_shash *tfm; }; +static inline void rxe_counter_inc(struct rxe_dev *rxe, enum rxe_counters cnt) +{ + rxe->stats_counters[cnt]++; +} + static inline struct rxe_dev *to_rdev(struct ib_device *dev) { return dev ? container_of(dev, struct rxe_dev, ib_dev) : NULL; @@ -475,6 +462,6 @@ static inline struct rxe_mem *to_rmw(struct ib_mw *mw) int rxe_register_device(struct rxe_dev *rxe); int rxe_unregister_device(struct rxe_dev *rxe); -void rxe_mc_cleanup(void *arg); +void rxe_mc_cleanup(struct rxe_pool_entry *arg); #endif /* RXE_VERBS_H */ diff --git a/drivers/infiniband/ulp/Makefile b/drivers/infiniband/ulp/Makefile index f3c7dcf03098..c28af1823a2d 100644 --- a/drivers/infiniband/ulp/Makefile +++ b/drivers/infiniband/ulp/Makefile @@ -3,3 +3,4 @@ obj-$(CONFIG_INFINIBAND_SRP) += srp/ obj-$(CONFIG_INFINIBAND_SRPT) += srpt/ obj-$(CONFIG_INFINIBAND_ISER) += iser/ obj-$(CONFIG_INFINIBAND_ISERT) += isert/ +obj-$(CONFIG_INFINIBAND_OPA_VNIC) += opa_vnic/ diff --git a/drivers/infiniband/ulp/ipoib/ipoib.h b/drivers/infiniband/ulp/ipoib/ipoib.h index da12717a3eb7..ff50a7bd66d8 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib.h +++ b/drivers/infiniband/ulp/ipoib/ipoib.h @@ -52,7 +52,6 @@ #include <rdma/ib_pack.h> #include <rdma/ib_sa.h> #include <linux/sched.h> - /* constants */ enum ipoib_flush_level { @@ -153,6 +152,13 @@ static inline void skb_add_pseudo_hdr(struct sk_buff *skb) skb_pull(skb, IPOIB_HARD_LEN); } +static inline struct ipoib_dev_priv *ipoib_priv(const struct net_device *dev) +{ + struct rdma_netdev *rn = netdev_priv(dev); + + return rn->clnt_priv; +} + /* Used for all multicast joins (broadcast, IPv4 mcast and IPv6 mcast) */ struct ipoib_mcast { struct ib_sa_mcmember_rec mcmember; @@ -404,6 +410,7 @@ struct ipoib_dev_priv { struct timer_list poll_timer; unsigned max_send_sge; bool sm_fullmember_sendonly_support; + const struct net_device_ops *rn_ops; }; struct ipoib_ah { @@ -416,7 +423,7 @@ struct ipoib_ah { struct ipoib_path { struct net_device *dev; - struct ib_sa_path_rec pathrec; + struct sa_path_rec pathrec; struct ipoib_ah *ah; struct sk_buff_head queue; @@ -472,7 +479,7 @@ void ipoib_ib_completion(struct ib_cq *cq, void *dev_ptr); void ipoib_send_comp_handler(struct ib_cq *cq, void *dev_ptr); struct ipoib_ah *ipoib_create_ah(struct net_device *dev, - struct ib_pd *pd, struct ib_ah_attr *attr); + struct ib_pd *pd, struct rdma_ah_attr *attr); void ipoib_free_ah(struct kref *kref); static inline void ipoib_put_ah(struct ipoib_ah *ah) { @@ -482,27 +489,28 @@ int ipoib_open(struct net_device *dev); int ipoib_add_pkey_attr(struct net_device *dev); int ipoib_add_umcast_attr(struct net_device *dev); -void ipoib_send(struct net_device *dev, struct sk_buff *skb, - struct ipoib_ah *address, u32 qpn); +int ipoib_send(struct net_device *dev, struct sk_buff *skb, + struct ib_ah *address, u32 dqpn); void ipoib_reap_ah(struct work_struct *work); struct ipoib_path *__path_find(struct net_device *dev, void *gid); void ipoib_mark_paths_invalid(struct net_device *dev); void ipoib_flush_paths(struct net_device *dev); -int ipoib_check_sm_sendonly_fullmember_support(struct ipoib_dev_priv *priv); -struct ipoib_dev_priv *ipoib_intf_alloc(const char *format); - -int ipoib_ib_dev_init(struct net_device *dev, struct ib_device *ca, int port); +struct ipoib_dev_priv *ipoib_intf_alloc(struct ib_device *hca, u8 port, + const char *format); +void ipoib_ib_tx_timer_func(unsigned long ctx); void ipoib_ib_dev_flush_light(struct work_struct *work); void ipoib_ib_dev_flush_normal(struct work_struct *work); void ipoib_ib_dev_flush_heavy(struct work_struct *work); void ipoib_pkey_event(struct work_struct *work); void ipoib_ib_dev_cleanup(struct net_device *dev); +int ipoib_ib_dev_open_default(struct net_device *dev); int ipoib_ib_dev_open(struct net_device *dev); -int ipoib_ib_dev_up(struct net_device *dev); -int ipoib_ib_dev_down(struct net_device *dev); int ipoib_ib_dev_stop(struct net_device *dev); +void ipoib_ib_dev_up(struct net_device *dev); +void ipoib_ib_dev_down(struct net_device *dev); +int ipoib_ib_dev_stop_default(struct net_device *dev); void ipoib_pkey_dev_check_presence(struct net_device *dev); int ipoib_dev_init(struct net_device *dev, struct ib_device *ca, int port); @@ -513,7 +521,7 @@ void ipoib_mcast_carrier_on_task(struct work_struct *work); void ipoib_mcast_send(struct net_device *dev, u8 *daddr, struct sk_buff *skb); void ipoib_mcast_restart_task(struct work_struct *work); -int ipoib_mcast_start_thread(struct net_device *dev); +void ipoib_mcast_start_thread(struct net_device *dev); int ipoib_mcast_stop_thread(struct net_device *dev); void ipoib_mcast_dev_down(struct net_device *dev); @@ -562,8 +570,10 @@ void ipoib_path_iter_read(struct ipoib_path_iter *iter, struct ipoib_path *path); #endif -int ipoib_mcast_attach(struct net_device *dev, u16 mlid, - union ib_gid *mgid, int set_qkey); +int ipoib_mcast_attach(struct net_device *dev, struct ib_device *hca, + union ib_gid *mgid, u16 mlid, int set_qkey, u32 qkey); +int ipoib_mcast_detach(struct net_device *dev, struct ib_device *hca, + union ib_gid *mgid, u16 mlid); void ipoib_mcast_remove_list(struct list_head *remove_list); void ipoib_check_and_add_mcast_sendonly(struct ipoib_dev_priv *priv, u8 *mgid, struct list_head *remove_list); @@ -587,13 +597,13 @@ void __exit ipoib_netlink_fini(void); void ipoib_set_umcast(struct net_device *ndev, int umcast_val); int ipoib_set_mode(struct net_device *dev, const char *buf); -void ipoib_setup(struct net_device *dev); +void ipoib_setup_common(struct net_device *dev); void ipoib_pkey_open(struct ipoib_dev_priv *priv); void ipoib_drain_cq(struct net_device *dev); void ipoib_set_ethtool_ops(struct net_device *dev); -int ipoib_set_dev_features(struct ipoib_dev_priv *priv, struct ib_device *hca); +void ipoib_set_dev_features(struct ipoib_dev_priv *priv, struct ib_device *hca); #define IPOIB_FLAGS_RC 0x80 #define IPOIB_FLAGS_UC 0x40 @@ -607,14 +617,14 @@ extern int ipoib_max_conn_qp; static inline int ipoib_cm_admin_enabled(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); return IPOIB_CM_SUPPORTED(dev->dev_addr) && test_bit(IPOIB_FLAG_ADMIN_CM, &priv->flags); } static inline int ipoib_cm_enabled(struct net_device *dev, u8 *hwaddr) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); return IPOIB_CM_SUPPORTED(hwaddr) && test_bit(IPOIB_FLAG_ADMIN_CM, &priv->flags); } @@ -637,13 +647,13 @@ static inline void ipoib_cm_set(struct ipoib_neigh *neigh, struct ipoib_cm_tx *t static inline int ipoib_cm_has_srq(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); return !!priv->cm.srq; } static inline unsigned int ipoib_cm_max_mtu(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); return priv->cm.max_cm_mtu; } diff --git a/drivers/infiniband/ulp/ipoib/ipoib_cm.c b/drivers/infiniband/ulp/ipoib/ipoib_cm.c index 096c4f6fbd65..f87d104837dc 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_cm.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_cm.c @@ -38,6 +38,8 @@ #include <linux/slab.h> #include <linux/vmalloc.h> #include <linux/moduleparam.h> +#include <linux/sched/signal.h> +#include <linux/sched/mm.h> #include "ipoib.h" @@ -91,7 +93,7 @@ static void ipoib_cm_dma_unmap_rx(struct ipoib_dev_priv *priv, int frags, static int ipoib_cm_post_receive_srq(struct net_device *dev, int id) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ib_recv_wr *bad_wr; int i, ret; @@ -117,7 +119,7 @@ static int ipoib_cm_post_receive_nonsrq(struct net_device *dev, struct ib_recv_wr *wr, struct ib_sge *sge, int id) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ib_recv_wr *bad_wr; int i, ret; @@ -144,7 +146,7 @@ static struct sk_buff *ipoib_cm_alloc_rx_skb(struct net_device *dev, u64 mapping[IPOIB_CM_RX_SG], gfp_t gfp) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct sk_buff *skb; int i; @@ -195,7 +197,7 @@ partial_error: static void ipoib_cm_free_rx_ring(struct net_device *dev, struct ipoib_cm_rx_buf *rx_ring) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); int i; for (i = 0; i < ipoib_recvq_size; ++i) @@ -234,7 +236,7 @@ static void ipoib_cm_start_rx_drain(struct ipoib_dev_priv *priv) static void ipoib_cm_rx_event_handler(struct ib_event *event, void *ctx) { struct ipoib_cm_rx *p = ctx; - struct ipoib_dev_priv *priv = netdev_priv(p->dev); + struct ipoib_dev_priv *priv = ipoib_priv(p->dev); unsigned long flags; if (event->event != IB_EVENT_QP_LAST_WQE_REACHED) @@ -250,7 +252,7 @@ static void ipoib_cm_rx_event_handler(struct ib_event *event, void *ctx) static struct ib_qp *ipoib_cm_create_rx_qp(struct net_device *dev, struct ipoib_cm_rx *p) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ib_qp_init_attr attr = { .event_handler = ipoib_cm_rx_event_handler, .send_cq = priv->recv_cq, /* For drain WR */ @@ -275,7 +277,7 @@ static int ipoib_cm_modify_rx_qp(struct net_device *dev, struct ib_cm_id *cm_id, struct ib_qp *qp, unsigned psn) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ib_qp_attr qp_attr; int qp_attr_mask, ret; @@ -330,7 +332,7 @@ static void ipoib_cm_init_rx_wr(struct net_device *dev, struct ib_recv_wr *wr, struct ib_sge *sge) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); int i; for (i = 0; i < priv->cm.num_frags; ++i) @@ -348,7 +350,7 @@ static void ipoib_cm_init_rx_wr(struct net_device *dev, static int ipoib_cm_nonsrq_init_rx(struct net_device *dev, struct ib_cm_id *cm_id, struct ipoib_cm_rx *rx) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct { struct ib_recv_wr wr; struct ib_sge sge[IPOIB_CM_RX_SG]; @@ -363,7 +365,7 @@ static int ipoib_cm_nonsrq_init_rx(struct net_device *dev, struct ib_cm_id *cm_i t = kmalloc(sizeof *t, GFP_KERNEL); if (!t) { ret = -ENOMEM; - goto err_free; + goto err_free_1; } ipoib_cm_init_rx_wr(dev, &t->wr, t->sge); @@ -410,6 +412,8 @@ err_count: err_free: kfree(t); + +err_free_1: ipoib_cm_free_rx_ring(dev, rx->rx_ring); return ret; @@ -419,7 +423,7 @@ static int ipoib_cm_send_rep(struct net_device *dev, struct ib_cm_id *cm_id, struct ib_qp *qp, struct ib_cm_req_event_param *req, unsigned psn) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ipoib_cm_data data = {}; struct ib_cm_rep_param rep = {}; @@ -439,7 +443,7 @@ static int ipoib_cm_send_rep(struct net_device *dev, struct ib_cm_id *cm_id, static int ipoib_cm_req_handler(struct ib_cm_id *cm_id, struct ib_cm_event *event) { struct net_device *dev = cm_id->context; - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ipoib_cm_rx *p; unsigned psn; int ret; @@ -512,7 +516,7 @@ static int ipoib_cm_rx_handler(struct ib_cm_id *cm_id, /* Fall through */ case IB_CM_REJ_RECEIVED: p = cm_id->context; - priv = netdev_priv(p->dev); + priv = ipoib_priv(p->dev); if (ib_modify_qp(p->qp, &ipoib_cm_err_attr, IB_QP_STATE)) ipoib_warn(priv, "unable to move qp to error state\n"); /* Fall through */ @@ -556,7 +560,7 @@ static void skb_put_frags(struct sk_buff *skb, unsigned int hdr_space, void ipoib_cm_handle_rx_wc(struct net_device *dev, struct ib_wc *wc) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ipoib_cm_rx_buf *rx_ring; unsigned int wr_id = wc->wr_id & ~(IPOIB_OP_CM | IPOIB_OP_RECV); struct sk_buff *skb, *newskb; @@ -705,7 +709,7 @@ static inline int post_send(struct ipoib_dev_priv *priv, void ipoib_cm_send(struct net_device *dev, struct sk_buff *skb, struct ipoib_cm_tx *tx) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ipoib_tx_buf *tx_req; int rc; unsigned usable_sge = tx->max_send_sge - !!skb_headlen(skb); @@ -783,7 +787,7 @@ void ipoib_cm_send(struct net_device *dev, struct sk_buff *skb, struct ipoib_cm_ void ipoib_cm_handle_tx_wc(struct net_device *dev, struct ib_wc *wc) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ipoib_cm_tx *tx = wc->qp->qp_context; unsigned int wr_id = wc->wr_id & ~IPOIB_OP_CM; struct ipoib_tx_buf *tx_req; @@ -820,9 +824,12 @@ void ipoib_cm_handle_tx_wc(struct net_device *dev, struct ib_wc *wc) wc->status != IB_WC_WR_FLUSH_ERR) { struct ipoib_neigh *neigh; - ipoib_dbg(priv, "failed cm send event " - "(status=%d, wrid=%d vend_err %x)\n", - wc->status, wr_id, wc->vendor_err); + if (wc->status != IB_WC_RNR_RETRY_EXC_ERR) + ipoib_warn(priv, "failed cm send event (status=%d, wrid=%d vend_err %x)\n", + wc->status, wr_id, wc->vendor_err); + else + ipoib_dbg(priv, "failed cm send event (status=%d, wrid=%d vend_err %x)\n", + wc->status, wr_id, wc->vendor_err); spin_lock_irqsave(&priv->lock, flags); neigh = tx->neigh; @@ -849,7 +856,7 @@ void ipoib_cm_handle_tx_wc(struct net_device *dev, struct ib_wc *wc) int ipoib_cm_dev_open(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); int ret; if (!IPOIB_CM_SUPPORTED(dev->dev_addr)) @@ -881,7 +888,7 @@ err_cm: static void ipoib_cm_free_rx_reap_list(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ipoib_cm_rx *rx, *n; LIST_HEAD(list); @@ -904,7 +911,7 @@ static void ipoib_cm_free_rx_reap_list(struct net_device *dev) void ipoib_cm_dev_stop(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ipoib_cm_rx *p; unsigned long begin; int ret; @@ -948,7 +955,7 @@ void ipoib_cm_dev_stop(struct net_device *dev) break; } spin_unlock_irq(&priv->lock); - msleep(1); + usleep_range(1000, 2000); ipoib_drain_cq(dev); spin_lock_irq(&priv->lock); } @@ -963,7 +970,7 @@ void ipoib_cm_dev_stop(struct net_device *dev) static int ipoib_cm_rep_handler(struct ib_cm_id *cm_id, struct ib_cm_event *event) { struct ipoib_cm_tx *p = cm_id->context; - struct ipoib_dev_priv *priv = netdev_priv(p->dev); + struct ipoib_dev_priv *priv = ipoib_priv(p->dev); struct ipoib_cm_data *data = event->private_data; struct sk_buff_head skqueue; struct ib_qp_attr qp_attr; @@ -1015,9 +1022,10 @@ static int ipoib_cm_rep_handler(struct ib_cm_id *cm_id, struct ib_cm_event *even while ((skb = __skb_dequeue(&skqueue))) { skb->dev = p->dev; - if (dev_queue_xmit(skb)) - ipoib_warn(priv, "dev_queue_xmit failed " - "to requeue packet\n"); + ret = dev_queue_xmit(skb); + if (ret) + ipoib_warn(priv, "%s:dev_queue_xmit failed to re-queue packet, ret:%d\n", + __func__, ret); } ret = ib_send_cm_rtu(cm_id, NULL, 0); @@ -1030,7 +1038,7 @@ static int ipoib_cm_rep_handler(struct ib_cm_id *cm_id, struct ib_cm_event *even static struct ib_qp *ipoib_cm_create_tx_qp(struct net_device *dev, struct ipoib_cm_tx *tx) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ib_qp_init_attr attr = { .send_cq = priv->recv_cq, .recv_cq = priv->recv_cq, @@ -1040,9 +1048,8 @@ static struct ib_qp *ipoib_cm_create_tx_qp(struct net_device *dev, struct ipoib_ .sq_sig_type = IB_SIGNAL_ALL_WR, .qp_type = IB_QPT_RC, .qp_context = tx, - .create_flags = IB_QP_CREATE_USE_GFP_NOIO + .create_flags = 0 }; - struct ib_qp *tx_qp; if (dev->features & NETIF_F_SG) @@ -1050,10 +1057,6 @@ static struct ib_qp *ipoib_cm_create_tx_qp(struct net_device *dev, struct ipoib_ min_t(u32, priv->ca->attrs.max_sge, MAX_SKB_FRAGS + 1); tx_qp = ib_create_qp(priv->pd, &attr); - if (PTR_ERR(tx_qp) == -EINVAL) { - attr.create_flags &= ~IB_QP_CREATE_USE_GFP_NOIO; - tx_qp = ib_create_qp(priv->pd, &attr); - } tx->max_send_sge = attr.cap.max_send_sge; return tx_qp; } @@ -1061,9 +1064,9 @@ static struct ib_qp *ipoib_cm_create_tx_qp(struct net_device *dev, struct ipoib_ static int ipoib_cm_send_req(struct net_device *dev, struct ib_cm_id *id, struct ib_qp *qp, u32 qpn, - struct ib_sa_path_rec *pathrec) + struct sa_path_rec *pathrec) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ipoib_cm_data data = {}; struct ib_cm_req_param req = {}; @@ -1098,7 +1101,7 @@ static int ipoib_cm_send_req(struct net_device *dev, static int ipoib_cm_modify_tx_init(struct net_device *dev, struct ib_cm_id *cm_id, struct ib_qp *qp) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ib_qp_attr qp_attr; int qp_attr_mask, ret; ret = ib_find_pkey(priv->ca, priv->port, priv->pkey, &qp_attr.pkey_index); @@ -1121,13 +1124,14 @@ static int ipoib_cm_modify_tx_init(struct net_device *dev, } static int ipoib_cm_tx_init(struct ipoib_cm_tx *p, u32 qpn, - struct ib_sa_path_rec *pathrec) + struct sa_path_rec *pathrec) { - struct ipoib_dev_priv *priv = netdev_priv(p->dev); + struct ipoib_dev_priv *priv = ipoib_priv(p->dev); + unsigned int noio_flag; int ret; - p->tx_ring = __vmalloc(ipoib_sendq_size * sizeof *p->tx_ring, - GFP_NOIO, PAGE_KERNEL); + noio_flag = memalloc_noio_save(); + p->tx_ring = vzalloc(ipoib_sendq_size * sizeof(*p->tx_ring)); if (!p->tx_ring) { ret = -ENOMEM; goto err_tx; @@ -1135,9 +1139,10 @@ static int ipoib_cm_tx_init(struct ipoib_cm_tx *p, u32 qpn, memset(p->tx_ring, 0, ipoib_sendq_size * sizeof *p->tx_ring); p->qp = ipoib_cm_create_tx_qp(p->dev, p); + memalloc_noio_restore(noio_flag); if (IS_ERR(p->qp)) { ret = PTR_ERR(p->qp); - ipoib_warn(priv, "failed to allocate tx qp: %d\n", ret); + ipoib_warn(priv, "failed to create tx qp: %d\n", ret); goto err_qp; } @@ -1151,13 +1156,13 @@ static int ipoib_cm_tx_init(struct ipoib_cm_tx *p, u32 qpn, ret = ipoib_cm_modify_tx_init(p->dev, p->id, p->qp); if (ret) { ipoib_warn(priv, "failed to modify tx qp to rtr: %d\n", ret); - goto err_modify; + goto err_modify_send; } ret = ipoib_cm_send_req(p->dev, p->id, p->qp, qpn, pathrec); if (ret) { ipoib_warn(priv, "failed to send cm req: %d\n", ret); - goto err_send_cm; + goto err_modify_send; } ipoib_dbg(priv, "Request connection 0x%x for gid %pI6 qpn 0x%x\n", @@ -1165,8 +1170,7 @@ static int ipoib_cm_tx_init(struct ipoib_cm_tx *p, u32 qpn, return 0; -err_send_cm: -err_modify: +err_modify_send: ib_destroy_cm_id(p->id); err_id: p->id = NULL; @@ -1180,7 +1184,7 @@ err_tx: static void ipoib_cm_tx_destroy(struct ipoib_cm_tx *p) { - struct ipoib_dev_priv *priv = netdev_priv(p->dev); + struct ipoib_dev_priv *priv = ipoib_priv(p->dev); struct ipoib_tx_buf *tx_req; unsigned long begin; @@ -1200,7 +1204,7 @@ static void ipoib_cm_tx_destroy(struct ipoib_cm_tx *p) goto timeout; } - msleep(1); + usleep_range(1000, 2000); } } @@ -1230,7 +1234,7 @@ static int ipoib_cm_tx_handler(struct ib_cm_id *cm_id, struct ib_cm_event *event) { struct ipoib_cm_tx *tx = cm_id->context; - struct ipoib_dev_priv *priv = netdev_priv(tx->dev); + struct ipoib_dev_priv *priv = ipoib_priv(tx->dev); struct net_device *dev = priv->dev; struct ipoib_neigh *neigh; unsigned long flags; @@ -1281,7 +1285,7 @@ static int ipoib_cm_tx_handler(struct ib_cm_id *cm_id, struct ipoib_cm_tx *ipoib_cm_create_tx(struct net_device *dev, struct ipoib_path *path, struct ipoib_neigh *neigh) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ipoib_cm_tx *tx; tx = kzalloc(sizeof *tx, GFP_ATOMIC); @@ -1300,7 +1304,7 @@ struct ipoib_cm_tx *ipoib_cm_create_tx(struct net_device *dev, struct ipoib_path void ipoib_cm_destroy_tx(struct ipoib_cm_tx *tx) { - struct ipoib_dev_priv *priv = netdev_priv(tx->dev); + struct ipoib_dev_priv *priv = ipoib_priv(tx->dev); unsigned long flags; if (test_and_clear_bit(IPOIB_FLAG_INITIALIZED, &tx->flags)) { spin_lock_irqsave(&priv->lock, flags); @@ -1326,7 +1330,7 @@ static void ipoib_cm_tx_start(struct work_struct *work) struct ipoib_path *path; int ret; - struct ib_sa_path_rec pathrec; + struct sa_path_rec pathrec; u32 qpn; netif_tx_lock_bh(dev); @@ -1388,7 +1392,7 @@ static void ipoib_cm_tx_reap(struct work_struct *work) while (!list_empty(&priv->cm.reap_list)) { p = list_entry(priv->cm.reap_list.next, typeof(*p), list); - list_del(&p->list); + list_del_init(&p->list); spin_unlock_irqrestore(&priv->lock, flags); netif_tx_unlock_bh(dev); ipoib_cm_tx_destroy(p); @@ -1435,7 +1439,7 @@ static void ipoib_cm_skb_reap(struct work_struct *work) void ipoib_cm_skb_too_long(struct net_device *dev, struct sk_buff *skb, unsigned int mtu) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); int e = skb_queue_empty(&priv->cm.skb_queue); if (skb_dst(skb)) @@ -1484,7 +1488,8 @@ static void ipoib_cm_stale_task(struct work_struct *work) static ssize_t show_mode(struct device *d, struct device_attribute *attr, char *buf) { - struct ipoib_dev_priv *priv = netdev_priv(to_net_dev(d)); + struct net_device *dev = to_net_dev(d); + struct ipoib_dev_priv *priv = ipoib_priv(dev); if (test_bit(IPOIB_FLAG_ADMIN_CM, &priv->flags)) return sprintf(buf, "connected\n"); @@ -1497,7 +1502,7 @@ static ssize_t set_mode(struct device *d, struct device_attribute *attr, { struct net_device *dev = to_net_dev(d); int ret; - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); if (test_bit(IPOIB_FLAG_GOING_DOWN, &priv->flags)) return -EPERM; @@ -1507,12 +1512,14 @@ static ssize_t set_mode(struct device *d, struct device_attribute *attr, ret = ipoib_set_mode(dev, buf); - rtnl_unlock(); - - if (!ret) - return count; + /* The assumption is that the function ipoib_set_mode returned + * with the rtnl held by it, if not the value -EBUSY returned, + * then no need to rtnl_unlock + */ + if (ret != -EBUSY) + rtnl_unlock(); - return ret; + return (!ret || ret == -EBUSY) ? count : ret; } static DEVICE_ATTR(mode, S_IWUSR | S_IRUGO, show_mode, set_mode); @@ -1524,7 +1531,7 @@ int ipoib_cm_add_mode_attr(struct net_device *dev) static void ipoib_cm_create_srq(struct net_device *dev, int max_sge) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ib_srq_init_attr srq_init_attr = { .srq_type = IB_SRQT_BASIC, .attr = { @@ -1553,7 +1560,7 @@ static void ipoib_cm_create_srq(struct net_device *dev, int max_sge) int ipoib_cm_dev_init(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); int max_srq_sge, i; INIT_LIST_HEAD(&priv->cm.passive_ids); @@ -1614,7 +1621,7 @@ int ipoib_cm_dev_init(struct net_device *dev) void ipoib_cm_dev_cleanup(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); int ret; if (!priv->cm.srq) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_ethtool.c b/drivers/infiniband/ulp/ipoib/ipoib_ethtool.c index 7b6d40ff1acf..7871379342f4 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_ethtool.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_ethtool.c @@ -60,12 +60,12 @@ static const struct ipoib_stats ipoib_gstrings_stats[] = { static void ipoib_get_drvinfo(struct net_device *netdev, struct ethtool_drvinfo *drvinfo) { - struct ipoib_dev_priv *priv = netdev_priv(netdev); + struct ipoib_dev_priv *priv = ipoib_priv(netdev); ib_get_device_fw_str(priv->ca, drvinfo->fw_version, sizeof(drvinfo->fw_version)); - strlcpy(drvinfo->bus_info, dev_name(priv->ca->dma_device), + strlcpy(drvinfo->bus_info, dev_name(priv->ca->dev.parent), sizeof(drvinfo->bus_info)); strlcpy(drvinfo->version, ipoib_driver_version, @@ -77,7 +77,7 @@ static void ipoib_get_drvinfo(struct net_device *netdev, static int ipoib_get_coalesce(struct net_device *dev, struct ethtool_coalesce *coal) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); coal->rx_coalesce_usecs = priv->ethtool.coalesce_usecs; coal->rx_max_coalesced_frames = priv->ethtool.max_coalesced_frames; @@ -88,7 +88,7 @@ static int ipoib_get_coalesce(struct net_device *dev, static int ipoib_set_coalesce(struct net_device *dev, struct ethtool_coalesce *coal) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); int ret; /* @@ -155,7 +155,66 @@ static int ipoib_get_sset_count(struct net_device __always_unused *dev, return -EOPNOTSUPP; } +/* Return lane speed in unit of 1e6 bit/sec */ +static inline int ib_speed_enum_to_int(int speed) +{ + switch (speed) { + case IB_SPEED_SDR: + return SPEED_2500; + case IB_SPEED_DDR: + return SPEED_5000; + case IB_SPEED_QDR: + case IB_SPEED_FDR10: + return SPEED_10000; + case IB_SPEED_FDR: + return SPEED_14000; + case IB_SPEED_EDR: + return SPEED_25000; + } + + return SPEED_UNKNOWN; +} + +static int ipoib_get_link_ksettings(struct net_device *netdev, + struct ethtool_link_ksettings *cmd) +{ + struct ipoib_dev_priv *priv = ipoib_priv(netdev); + struct ib_port_attr attr; + int ret, speed, width; + + if (!netif_carrier_ok(netdev)) { + cmd->base.speed = SPEED_UNKNOWN; + cmd->base.duplex = DUPLEX_UNKNOWN; + return 0; + } + + ret = ib_query_port(priv->ca, priv->port, &attr); + if (ret < 0) + return -EINVAL; + + speed = ib_speed_enum_to_int(attr.active_speed); + width = ib_width_enum_to_int(attr.active_width); + + if (speed < 0 || width < 0) + return -EINVAL; + + /* Except the following are set, the other members of + * the struct ethtool_link_settings are initialized to + * zero in the function __ethtool_get_link_ksettings. + */ + cmd->base.speed = speed * width; + cmd->base.duplex = DUPLEX_FULL; + + cmd->base.phy_address = 0xFF; + + cmd->base.autoneg = AUTONEG_ENABLE; + cmd->base.port = PORT_OTHER; + + return 0; +} + static const struct ethtool_ops ipoib_ethtool_ops = { + .get_link_ksettings = ipoib_get_link_ksettings, .get_drvinfo = ipoib_get_drvinfo, .get_coalesce = ipoib_get_coalesce, .set_coalesce = ipoib_set_coalesce, diff --git a/drivers/infiniband/ulp/ipoib/ipoib_fs.c b/drivers/infiniband/ulp/ipoib/ipoib_fs.c index 6bd5740e2691..11f74cbe6660 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_fs.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_fs.c @@ -210,16 +210,16 @@ static int ipoib_path_seq_show(struct seq_file *file, void *iter_ptr) seq_printf(file, "GID: %s\n" " complete: %6s\n", - gid_buf, path.pathrec.dlid ? "yes" : "no"); + gid_buf, sa_path_get_dlid(&path.pathrec) ? "yes" : "no"); - if (path.pathrec.dlid) { + if (sa_path_get_dlid(&path.pathrec)) { rate = ib_rate_to_mbps(path.pathrec.rate); seq_printf(file, " DLID: 0x%04x\n" " SL: %12d\n" " rate: %8d.%d Gb/sec\n", - be16_to_cpu(path.pathrec.dlid), + be32_to_cpu(sa_path_get_dlid(&path.pathrec)), path.pathrec.sl, rate / 1000, rate % 1000); } @@ -261,7 +261,7 @@ static const struct file_operations ipoib_path_fops = { void ipoib_create_debug_files(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); char name[IFNAMSIZ + sizeof "_path"]; snprintf(name, sizeof name, "%s_mcg", dev->name); @@ -279,10 +279,13 @@ void ipoib_create_debug_files(struct net_device *dev) void ipoib_delete_debug_files(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); + WARN_ONCE(!priv->mcg_dentry, "null mcg debug file\n"); + WARN_ONCE(!priv->path_dentry, "null path debug file\n"); debugfs_remove(priv->mcg_dentry); debugfs_remove(priv->path_dentry); + priv->mcg_dentry = priv->path_dentry = NULL; } int ipoib_register_debugfs(void) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_ib.c b/drivers/infiniband/ulp/ipoib/ipoib_ib.c index 5038f9d2d753..57a9655e844d 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_ib.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_ib.c @@ -52,7 +52,7 @@ MODULE_PARM_DESC(data_debug_level, #endif struct ipoib_ah *ipoib_create_ah(struct net_device *dev, - struct ib_pd *pd, struct ib_ah_attr *attr) + struct ib_pd *pd, struct rdma_ah_attr *attr) { struct ipoib_ah *ah; struct ib_ah *vah; @@ -65,13 +65,13 @@ struct ipoib_ah *ipoib_create_ah(struct net_device *dev, ah->last_send = 0; kref_init(&ah->ref); - vah = ib_create_ah(pd, attr); + vah = rdma_create_ah(pd, attr); if (IS_ERR(vah)) { kfree(ah); ah = (struct ipoib_ah *)vah; } else { ah->ah = vah; - ipoib_dbg(netdev_priv(dev), "Created ah %p\n", ah->ah); + ipoib_dbg(ipoib_priv(dev), "Created ah %p\n", ah->ah); } return ah; @@ -80,7 +80,7 @@ struct ipoib_ah *ipoib_create_ah(struct net_device *dev, void ipoib_free_ah(struct kref *kref) { struct ipoib_ah *ah = container_of(kref, struct ipoib_ah, ref); - struct ipoib_dev_priv *priv = netdev_priv(ah->dev); + struct ipoib_dev_priv *priv = ipoib_priv(ah->dev); unsigned long flags; @@ -99,7 +99,7 @@ static void ipoib_ud_dma_unmap_rx(struct ipoib_dev_priv *priv, static int ipoib_ib_post_receive(struct net_device *dev, int id) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ib_recv_wr *bad_wr; int ret; @@ -121,7 +121,7 @@ static int ipoib_ib_post_receive(struct net_device *dev, int id) static struct sk_buff *ipoib_alloc_rx_skb(struct net_device *dev, int id) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct sk_buff *skb; int buf_size; u64 *mapping; @@ -153,7 +153,7 @@ error: static int ipoib_ib_post_receives(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); int i; for (i = 0; i < ipoib_recvq_size; ++i) { @@ -172,7 +172,7 @@ static int ipoib_ib_post_receives(struct net_device *dev) static void ipoib_ib_handle_rx_wc(struct net_device *dev, struct ib_wc *wc) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); unsigned int wr_id = wc->wr_id & ~IPOIB_OP_RECV; struct sk_buff *skb; u64 mapping[IPOIB_UD_RX_SG]; @@ -381,7 +381,7 @@ free_res: static void ipoib_ib_handle_tx_wc(struct net_device *dev, struct ib_wc *wc) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); unsigned int wr_id = wc->wr_id; struct ipoib_tx_buf *tx_req; @@ -485,14 +485,14 @@ poll_more: void ipoib_ib_completion(struct ib_cq *cq, void *dev_ptr) { struct net_device *dev = dev_ptr; - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); napi_schedule(&priv->napi); } static void drain_tx_cq(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); netif_tx_lock(dev); while (poll_tx(priv)) @@ -506,14 +506,14 @@ static void drain_tx_cq(struct net_device *dev) void ipoib_send_comp_handler(struct ib_cq *cq, void *dev_ptr) { - struct ipoib_dev_priv *priv = netdev_priv(dev_ptr); + struct ipoib_dev_priv *priv = ipoib_priv(dev_ptr); mod_timer(&priv->poll_timer, jiffies); } static inline int post_send(struct ipoib_dev_priv *priv, unsigned int wr_id, - struct ib_ah *address, u32 qpn, + struct ib_ah *address, u32 dqpn, struct ipoib_tx_buf *tx_req, void *head, int hlen) { @@ -523,7 +523,7 @@ static inline int post_send(struct ipoib_dev_priv *priv, ipoib_build_sge(priv, tx_req); priv->tx_wr.wr.wr_id = wr_id; - priv->tx_wr.remote_qpn = qpn; + priv->tx_wr.remote_qpn = dqpn; priv->tx_wr.ah = address; if (head) { @@ -537,10 +537,10 @@ static inline int post_send(struct ipoib_dev_priv *priv, return ib_post_send(priv->qp, &priv->tx_wr.wr, &bad_wr); } -void ipoib_send(struct net_device *dev, struct sk_buff *skb, - struct ipoib_ah *address, u32 qpn) +int ipoib_send(struct net_device *dev, struct sk_buff *skb, + struct ib_ah *address, u32 dqpn) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ipoib_tx_buf *tx_req; int hlen, rc; void *phead; @@ -554,7 +554,7 @@ void ipoib_send(struct net_device *dev, struct sk_buff *skb, ++dev->stats.tx_dropped; ++dev->stats.tx_errors; dev_kfree_skb_any(skb); - return; + return -1; } } else { if (unlikely(skb->len > priv->mcast_mtu + IPOIB_ENCAP_LEN)) { @@ -563,7 +563,7 @@ void ipoib_send(struct net_device *dev, struct sk_buff *skb, ++dev->stats.tx_dropped; ++dev->stats.tx_errors; ipoib_cm_skb_too_long(dev, skb, priv->mcast_mtu); - return; + return -1; } phead = NULL; hlen = 0; @@ -574,7 +574,7 @@ void ipoib_send(struct net_device *dev, struct sk_buff *skb, ++dev->stats.tx_dropped; ++dev->stats.tx_errors; dev_kfree_skb_any(skb); - return; + return -1; } /* Does skb_linearize return ok without reducing nr_frags? */ if (skb_shinfo(skb)->nr_frags > usable_sge) { @@ -582,12 +582,13 @@ void ipoib_send(struct net_device *dev, struct sk_buff *skb, ++dev->stats.tx_dropped; ++dev->stats.tx_errors; dev_kfree_skb_any(skb); - return; + return -1; } } - ipoib_dbg_data(priv, "sending packet, length=%d address=%p qpn=0x%06x\n", - skb->len, address, qpn); + ipoib_dbg_data(priv, + "sending packet, length=%d address=%p dqpn=0x%06x\n", + skb->len, address, dqpn); /* * We put the skb into the tx_ring _before_ we call post_send() @@ -601,7 +602,7 @@ void ipoib_send(struct net_device *dev, struct sk_buff *skb, if (unlikely(ipoib_dma_map_tx(priv->ca, tx_req))) { ++dev->stats.tx_errors; dev_kfree_skb_any(skb); - return; + return -1; } if (skb->ip_summed == CHECKSUM_PARTIAL) @@ -620,7 +621,7 @@ void ipoib_send(struct net_device *dev, struct sk_buff *skb, skb_dst_drop(skb); rc = post_send(priv, priv->tx_head & (ipoib_sendq_size - 1), - address->ah, qpn, tx_req, phead, hlen); + address, dqpn, tx_req, phead, hlen); if (unlikely(rc)) { ipoib_warn(priv, "post_send failed, error %d\n", rc); ++dev->stats.tx_errors; @@ -629,21 +630,24 @@ void ipoib_send(struct net_device *dev, struct sk_buff *skb, dev_kfree_skb_any(skb); if (netif_queue_stopped(dev)) netif_wake_queue(dev); + rc = 0; } else { netif_trans_update(dev); - address->last_send = priv->tx_head; + rc = priv->tx_head; ++priv->tx_head; } if (unlikely(priv->tx_outstanding > MAX_SEND_CQE)) while (poll_tx(priv)) ; /* nothing */ + + return rc; } static void __ipoib_reap_ah(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ipoib_ah *ah, *tah; LIST_HEAD(remove_list); unsigned long flags; @@ -654,7 +658,7 @@ static void __ipoib_reap_ah(struct net_device *dev) list_for_each_entry_safe(ah, tah, &priv->dead_ahs, list) if ((int) priv->tx_tail - (int) ah->last_send >= 0) { list_del(&ah->list); - ib_destroy_ah(ah->ah); + rdma_destroy_ah(ah->ah); kfree(ah); } @@ -677,7 +681,7 @@ void ipoib_reap_ah(struct work_struct *work) static void ipoib_flush_ah(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); cancel_delayed_work(&priv->ah_reap_task); flush_workqueue(priv->wq); @@ -686,30 +690,124 @@ static void ipoib_flush_ah(struct net_device *dev) static void ipoib_stop_ah(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); set_bit(IPOIB_STOP_REAPER, &priv->flags); ipoib_flush_ah(dev); } -static void ipoib_ib_tx_timer_func(unsigned long ctx) +static int recvs_pending(struct net_device *dev) { - drain_tx_cq((struct net_device *)ctx); + struct ipoib_dev_priv *priv = ipoib_priv(dev); + int pending = 0; + int i; + + for (i = 0; i < ipoib_recvq_size; ++i) + if (priv->rx_ring[i].skb) + ++pending; + + return pending; } -int ipoib_ib_dev_open(struct net_device *dev) +int ipoib_ib_dev_stop_default(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); - int ret; + struct ipoib_dev_priv *priv = ipoib_priv(dev); + struct ib_qp_attr qp_attr; + unsigned long begin; + struct ipoib_tx_buf *tx_req; + int i; - ipoib_pkey_dev_check_presence(dev); + if (test_bit(IPOIB_FLAG_INITIALIZED, &priv->flags)) + napi_disable(&priv->napi); - if (!test_bit(IPOIB_PKEY_ASSIGNED, &priv->flags)) { - ipoib_warn(priv, "P_Key 0x%04x is %s\n", priv->pkey, - (!(priv->pkey & 0x7fff) ? "Invalid" : "not found")); - return -1; + ipoib_cm_dev_stop(dev); + + /* + * Move our QP to the error state and then reinitialize in + * when all work requests have completed or have been flushed. + */ + qp_attr.qp_state = IB_QPS_ERR; + if (ib_modify_qp(priv->qp, &qp_attr, IB_QP_STATE)) + ipoib_warn(priv, "Failed to modify QP to ERROR state\n"); + + /* Wait for all sends and receives to complete */ + begin = jiffies; + + while (priv->tx_head != priv->tx_tail || recvs_pending(dev)) { + if (time_after(jiffies, begin + 5 * HZ)) { + ipoib_warn(priv, + "timing out; %d sends %d receives not completed\n", + priv->tx_head - priv->tx_tail, + recvs_pending(dev)); + + /* + * assume the HW is wedged and just free up + * all our pending work requests. + */ + while ((int)priv->tx_tail - (int)priv->tx_head < 0) { + tx_req = &priv->tx_ring[priv->tx_tail & + (ipoib_sendq_size - 1)]; + ipoib_dma_unmap_tx(priv, tx_req); + dev_kfree_skb_any(tx_req->skb); + ++priv->tx_tail; + --priv->tx_outstanding; + } + + for (i = 0; i < ipoib_recvq_size; ++i) { + struct ipoib_rx_buf *rx_req; + + rx_req = &priv->rx_ring[i]; + if (!rx_req->skb) + continue; + ipoib_ud_dma_unmap_rx(priv, + priv->rx_ring[i].mapping); + dev_kfree_skb_any(rx_req->skb); + rx_req->skb = NULL; + } + + goto timeout; + } + + ipoib_drain_cq(dev); + + usleep_range(1000, 2000); } + ipoib_dbg(priv, "All sends and receives done.\n"); + +timeout: + del_timer_sync(&priv->poll_timer); + qp_attr.qp_state = IB_QPS_RESET; + if (ib_modify_qp(priv->qp, &qp_attr, IB_QP_STATE)) + ipoib_warn(priv, "Failed to modify QP to RESET state\n"); + + ib_req_notify_cq(priv->recv_cq, IB_CQ_NEXT_COMP); + + return 0; +} + +int ipoib_ib_dev_stop(struct net_device *dev) +{ + struct ipoib_dev_priv *priv = ipoib_priv(dev); + + priv->rn_ops->ndo_stop(dev); + + clear_bit(IPOIB_FLAG_INITIALIZED, &priv->flags); + ipoib_flush_ah(dev); + + return 0; +} + +void ipoib_ib_tx_timer_func(unsigned long ctx) +{ + drain_tx_cq((struct net_device *)ctx); +} + +int ipoib_ib_dev_open_default(struct net_device *dev) +{ + struct ipoib_dev_priv *priv = ipoib_priv(dev); + int ret; + ret = ipoib_init_qp(dev); if (ret) { ipoib_warn(priv, "ipoib_init_qp returned %d\n", ret); @@ -719,33 +817,59 @@ int ipoib_ib_dev_open(struct net_device *dev) ret = ipoib_ib_post_receives(dev); if (ret) { ipoib_warn(priv, "ipoib_ib_post_receives returned %d\n", ret); - goto dev_stop; + goto out; } ret = ipoib_cm_dev_open(dev); if (ret) { ipoib_warn(priv, "ipoib_cm_dev_open returned %d\n", ret); - goto dev_stop; + goto out; + } + + if (!test_bit(IPOIB_FLAG_INITIALIZED, &priv->flags)) + napi_enable(&priv->napi); + + return 0; +out: + return -1; +} + +int ipoib_ib_dev_open(struct net_device *dev) +{ + struct ipoib_dev_priv *priv = ipoib_priv(dev); + + ipoib_pkey_dev_check_presence(dev); + + if (!test_bit(IPOIB_PKEY_ASSIGNED, &priv->flags)) { + ipoib_warn(priv, "P_Key 0x%04x is %s\n", priv->pkey, + (!(priv->pkey & 0x7fff) ? "Invalid" : "not found")); + return -1; } clear_bit(IPOIB_STOP_REAPER, &priv->flags); queue_delayed_work(priv->wq, &priv->ah_reap_task, round_jiffies_relative(HZ)); - if (!test_and_set_bit(IPOIB_FLAG_INITIALIZED, &priv->flags)) - napi_enable(&priv->napi); + if (priv->rn_ops->ndo_open(dev)) { + pr_warn("%s: Failed to open dev\n", dev->name); + goto dev_stop; + } + + set_bit(IPOIB_FLAG_INITIALIZED, &priv->flags); return 0; + dev_stop: - if (!test_and_set_bit(IPOIB_FLAG_INITIALIZED, &priv->flags)) - napi_enable(&priv->napi); + set_bit(IPOIB_STOP_REAPER, &priv->flags); + cancel_delayed_work(&priv->ah_reap_task); + set_bit(IPOIB_FLAG_INITIALIZED, &priv->flags); ipoib_ib_dev_stop(dev); return -1; } void ipoib_pkey_dev_check_presence(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); if (!(priv->pkey & 0x7fff) || ib_find_pkey(priv->ca, priv->port, priv->pkey, @@ -755,25 +879,25 @@ void ipoib_pkey_dev_check_presence(struct net_device *dev) set_bit(IPOIB_PKEY_ASSIGNED, &priv->flags); } -int ipoib_ib_dev_up(struct net_device *dev) +void ipoib_ib_dev_up(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); ipoib_pkey_dev_check_presence(dev); if (!test_bit(IPOIB_PKEY_ASSIGNED, &priv->flags)) { ipoib_dbg(priv, "PKEY is not assigned.\n"); - return 0; + return; } set_bit(IPOIB_FLAG_OPER_UP, &priv->flags); - return ipoib_mcast_start_thread(dev); + ipoib_mcast_start_thread(dev); } -int ipoib_ib_dev_down(struct net_device *dev) +void ipoib_ib_dev_down(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); ipoib_dbg(priv, "downing ib_dev\n"); @@ -784,26 +908,11 @@ int ipoib_ib_dev_down(struct net_device *dev) ipoib_mcast_dev_flush(dev); ipoib_flush_paths(dev); - - return 0; -} - -static int recvs_pending(struct net_device *dev) -{ - struct ipoib_dev_priv *priv = netdev_priv(dev); - int pending = 0; - int i; - - for (i = 0; i < ipoib_recvq_size; ++i) - if (priv->rx_ring[i].skb) - ++pending; - - return pending; } void ipoib_drain_cq(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); int i, n; /* @@ -840,109 +949,6 @@ void ipoib_drain_cq(struct net_device *dev) local_bh_enable(); } -int ipoib_ib_dev_stop(struct net_device *dev) -{ - struct ipoib_dev_priv *priv = netdev_priv(dev); - struct ib_qp_attr qp_attr; - unsigned long begin; - struct ipoib_tx_buf *tx_req; - int i; - - if (test_and_clear_bit(IPOIB_FLAG_INITIALIZED, &priv->flags)) - napi_disable(&priv->napi); - - ipoib_cm_dev_stop(dev); - - /* - * Move our QP to the error state and then reinitialize in - * when all work requests have completed or have been flushed. - */ - qp_attr.qp_state = IB_QPS_ERR; - if (ib_modify_qp(priv->qp, &qp_attr, IB_QP_STATE)) - ipoib_warn(priv, "Failed to modify QP to ERROR state\n"); - - /* Wait for all sends and receives to complete */ - begin = jiffies; - - while (priv->tx_head != priv->tx_tail || recvs_pending(dev)) { - if (time_after(jiffies, begin + 5 * HZ)) { - ipoib_warn(priv, "timing out; %d sends %d receives not completed\n", - priv->tx_head - priv->tx_tail, recvs_pending(dev)); - - /* - * assume the HW is wedged and just free up - * all our pending work requests. - */ - while ((int) priv->tx_tail - (int) priv->tx_head < 0) { - tx_req = &priv->tx_ring[priv->tx_tail & - (ipoib_sendq_size - 1)]; - ipoib_dma_unmap_tx(priv, tx_req); - dev_kfree_skb_any(tx_req->skb); - ++priv->tx_tail; - --priv->tx_outstanding; - } - - for (i = 0; i < ipoib_recvq_size; ++i) { - struct ipoib_rx_buf *rx_req; - - rx_req = &priv->rx_ring[i]; - if (!rx_req->skb) - continue; - ipoib_ud_dma_unmap_rx(priv, - priv->rx_ring[i].mapping); - dev_kfree_skb_any(rx_req->skb); - rx_req->skb = NULL; - } - - goto timeout; - } - - ipoib_drain_cq(dev); - - msleep(1); - } - - ipoib_dbg(priv, "All sends and receives done.\n"); - -timeout: - del_timer_sync(&priv->poll_timer); - qp_attr.qp_state = IB_QPS_RESET; - if (ib_modify_qp(priv->qp, &qp_attr, IB_QP_STATE)) - ipoib_warn(priv, "Failed to modify QP to RESET state\n"); - - ipoib_flush_ah(dev); - - ib_req_notify_cq(priv->recv_cq, IB_CQ_NEXT_COMP); - - return 0; -} - -int ipoib_ib_dev_init(struct net_device *dev, struct ib_device *ca, int port) -{ - struct ipoib_dev_priv *priv = netdev_priv(dev); - - priv->ca = ca; - priv->port = port; - priv->qp = NULL; - - if (ipoib_transport_dev_init(dev, ca)) { - printk(KERN_WARNING "%s: ipoib_transport_dev_init failed\n", ca->name); - return -ENODEV; - } - - setup_timer(&priv->poll_timer, ipoib_ib_tx_timer_func, - (unsigned long) dev); - - if (dev->flags & IFF_UP) { - if (ipoib_ib_dev_open(dev)) { - ipoib_transport_dev_cleanup(dev); - return -ENODEV; - } - } - - return 0; -} - /* * Takes whatever value which is in pkey index 0 and updates priv->pkey * returns 0 if the pkey value was changed. @@ -971,6 +977,19 @@ static inline int update_parent_pkey(struct ipoib_dev_priv *priv) */ priv->dev->broadcast[8] = priv->pkey >> 8; priv->dev->broadcast[9] = priv->pkey & 0xff; + + /* + * Update the broadcast address in the priv->broadcast object, + * in case it already exists, otherwise no one will do that. + */ + if (priv->broadcast) { + spin_lock_irq(&priv->lock); + memcpy(priv->broadcast->mcmember.mgid.raw, + priv->dev->broadcast + 4, + sizeof(union ib_gid)); + spin_unlock_irq(&priv->lock); + } + return 0; } @@ -1220,7 +1239,7 @@ void ipoib_ib_dev_flush_heavy(struct work_struct *work) void ipoib_ib_dev_cleanup(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); ipoib_dbg(priv, "cleaning up ib_dev\n"); /* @@ -1240,7 +1259,13 @@ void ipoib_ib_dev_cleanup(struct net_device *dev) */ ipoib_stop_ah(dev); - ipoib_transport_dev_cleanup(dev); -} + clear_bit(IPOIB_PKEY_ASSIGNED, &priv->flags); + priv->rn_ops->ndo_uninit(dev); + + if (priv->pd) { + ib_dealloc_pd(priv->pd); + priv->pd = NULL; + } +} diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c index 3ce0765a05ab..4ce315c92b48 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_main.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c @@ -108,9 +108,36 @@ static struct ib_client ipoib_client = { .get_net_dev_by_params = ipoib_get_net_dev_by_params, }; +#ifdef CONFIG_INFINIBAND_IPOIB_DEBUG +static int ipoib_netdev_event(struct notifier_block *this, + unsigned long event, void *ptr) +{ + struct netdev_notifier_info *ni = ptr; + struct net_device *dev = ni->dev; + + if (dev->netdev_ops->ndo_open != ipoib_open) + return NOTIFY_DONE; + + switch (event) { + case NETDEV_REGISTER: + ipoib_create_debug_files(dev); + break; + case NETDEV_CHANGENAME: + ipoib_delete_debug_files(dev); + ipoib_create_debug_files(dev); + break; + case NETDEV_UNREGISTER: + ipoib_delete_debug_files(dev); + break; + } + + return NOTIFY_DONE; +} +#endif + int ipoib_open(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); ipoib_dbg(priv, "bringing up interface\n"); @@ -126,8 +153,7 @@ int ipoib_open(struct net_device *dev) goto err_disable; } - if (ipoib_ib_dev_up(dev)) - goto err_stop; + ipoib_ib_dev_up(dev); if (!test_bit(IPOIB_FLAG_SUBINTERFACE, &priv->flags)) { struct ipoib_dev_priv *cpriv; @@ -150,9 +176,6 @@ int ipoib_open(struct net_device *dev) return 0; -err_stop: - ipoib_ib_dev_stop(dev); - err_disable: clear_bit(IPOIB_FLAG_ADMIN_UP, &priv->flags); @@ -161,7 +184,7 @@ err_disable: static int ipoib_stop(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); ipoib_dbg(priv, "stopping interface\n"); @@ -199,7 +222,7 @@ static void ipoib_uninit(struct net_device *dev) static netdev_features_t ipoib_fix_features(struct net_device *dev, netdev_features_t features) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); if (test_bit(IPOIB_FLAG_ADMIN_CM, &priv->flags)) features &= ~(NETIF_F_IP_CSUM | NETIF_F_TSO); @@ -209,7 +232,8 @@ static netdev_features_t ipoib_fix_features(struct net_device *dev, netdev_featu static int ipoib_change_mtu(struct net_device *dev, int new_mtu) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); + int ret = 0; /* dev->mtu > 2K ==> connected mode */ if (ipoib_cm_admin_enabled(dev)) { @@ -229,9 +253,38 @@ static int ipoib_change_mtu(struct net_device *dev, int new_mtu) priv->admin_mtu = new_mtu; - dev->mtu = min(priv->mcast_mtu, priv->admin_mtu); + if (priv->mcast_mtu < priv->admin_mtu) + ipoib_dbg(priv, "MTU must be smaller than the underlying " + "link layer MTU - 4 (%u)\n", priv->mcast_mtu); - return 0; + new_mtu = min(priv->mcast_mtu, priv->admin_mtu); + + if (priv->rn_ops->ndo_change_mtu) { + bool carrier_status = netif_carrier_ok(dev); + + netif_carrier_off(dev); + + /* notify lower level on the real mtu */ + ret = priv->rn_ops->ndo_change_mtu(dev, new_mtu); + + if (carrier_status) + netif_carrier_on(dev); + } else { + dev->mtu = new_mtu; + } + + return ret; +} + +static void ipoib_get_stats(struct net_device *dev, + struct rtnl_link_stats64 *stats) +{ + struct ipoib_dev_priv *priv = ipoib_priv(dev); + + if (priv->rn_ops->ndo_get_stats64) + priv->rn_ops->ndo_get_stats64(dev, stats); + else + netdev_stats_to_stats64(stats, &dev->stats); } /* Called with an RCU read lock taken */ @@ -468,7 +521,14 @@ static struct net_device *ipoib_get_net_dev_by_params( int ipoib_set_mode(struct net_device *dev, const char *buf) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); + + if ((test_bit(IPOIB_FLAG_ADMIN_CM, &priv->flags) && + !strcmp(buf, "connected\n")) || + (!test_bit(IPOIB_FLAG_ADMIN_CM, &priv->flags) && + !strcmp(buf, "datagram\n"))) { + return 0; + } /* flush paths if we switch modes so that connections are restarted */ if (IPOIB_CM_SUPPORTED(dev->dev_addr) && !strcmp(buf, "connected\n")) { @@ -481,8 +541,7 @@ int ipoib_set_mode(struct net_device *dev, const char *buf) priv->tx_wr.wr.send_flags &= ~IB_SEND_IP_CSUM; ipoib_flush_paths(dev); - rtnl_lock(); - return 0; + return (!rtnl_trylock()) ? -EBUSY : 0; } if (!strcmp(buf, "datagram\n")) { @@ -491,8 +550,7 @@ int ipoib_set_mode(struct net_device *dev, const char *buf) dev_set_mtu(dev, min(priv->mcast_mtu, dev->mtu)); rtnl_unlock(); ipoib_flush_paths(dev); - rtnl_lock(); - return 0; + return (!rtnl_trylock()) ? -EBUSY : 0; } return -EINVAL; @@ -500,7 +558,7 @@ int ipoib_set_mode(struct net_device *dev, const char *buf) struct ipoib_path *__path_find(struct net_device *dev, void *gid) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct rb_node *n = priv->path_tree.rb_node; struct ipoib_path *path; int ret; @@ -524,7 +582,7 @@ struct ipoib_path *__path_find(struct net_device *dev, void *gid) static int __path_add(struct net_device *dev, struct ipoib_path *path) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct rb_node **n = &priv->path_tree.rb_node; struct rb_node *pn = NULL; struct ipoib_path *tpath; @@ -559,7 +617,7 @@ static void path_free(struct net_device *dev, struct ipoib_path *path) while ((skb = __skb_dequeue(&path->queue))) dev_kfree_skb_irq(skb); - ipoib_dbg(netdev_priv(dev), "path_free\n"); + ipoib_dbg(ipoib_priv(dev), "path_free\n"); /* remove all neigh connected to this path */ ipoib_del_neighs_by_gid(dev, path->pathrec.dgid.raw); @@ -593,7 +651,7 @@ struct ipoib_path_iter *ipoib_path_iter_init(struct net_device *dev) int ipoib_path_iter_next(struct ipoib_path_iter *iter) { - struct ipoib_dev_priv *priv = netdev_priv(iter->dev); + struct ipoib_dev_priv *priv = ipoib_priv(iter->dev); struct rb_node *n; struct ipoib_path *path; int ret = 1; @@ -630,95 +688,32 @@ void ipoib_path_iter_read(struct ipoib_path_iter *iter, void ipoib_mark_paths_invalid(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ipoib_path *path, *tp; spin_lock_irq(&priv->lock); list_for_each_entry_safe(path, tp, &priv->path_list, list) { - ipoib_dbg(priv, "mark path LID 0x%04x GID %pI6 invalid\n", - be16_to_cpu(path->pathrec.dlid), - path->pathrec.dgid.raw); + ipoib_dbg(priv, "mark path LID 0x%08x GID %pI6 invalid\n", + be32_to_cpu(sa_path_get_dlid(&path->pathrec)), + path->pathrec.dgid.raw); path->valid = 0; } spin_unlock_irq(&priv->lock); } -struct classport_info_context { - struct ipoib_dev_priv *priv; - struct completion done; - struct ib_sa_query *sa_query; -}; - -static void classport_info_query_cb(int status, struct ib_class_port_info *rec, - void *context) +static void push_pseudo_header(struct sk_buff *skb, const char *daddr) { - struct classport_info_context *cb_ctx = context; - struct ipoib_dev_priv *priv; - - WARN_ON(!context); - - priv = cb_ctx->priv; - - if (status || !rec) { - pr_debug("device: %s failed query classport_info status: %d\n", - priv->dev->name, status); - /* keeps the default, will try next mcast_restart */ - priv->sm_fullmember_sendonly_support = false; - goto out; - } - - if (ib_get_cpi_capmask2(rec) & - IB_SA_CAP_MASK2_SENDONLY_FULL_MEM_SUPPORT) { - pr_debug("device: %s enabled fullmember-sendonly for sendonly MCG\n", - priv->dev->name); - priv->sm_fullmember_sendonly_support = true; - } else { - pr_debug("device: %s disabled fullmember-sendonly for sendonly MCG\n", - priv->dev->name); - priv->sm_fullmember_sendonly_support = false; - } - -out: - complete(&cb_ctx->done); -} - -int ipoib_check_sm_sendonly_fullmember_support(struct ipoib_dev_priv *priv) -{ - struct classport_info_context *callback_context; - int ret; - - callback_context = kmalloc(sizeof(*callback_context), GFP_KERNEL); - if (!callback_context) - return -ENOMEM; - - callback_context->priv = priv; - init_completion(&callback_context->done); - - ret = ib_sa_classport_info_rec_query(&ipoib_sa_client, - priv->ca, priv->port, 3000, - GFP_KERNEL, - classport_info_query_cb, - callback_context, - &callback_context->sa_query); - if (ret < 0) { - pr_info("%s failed to send ib_sa_classport_info query, ret: %d\n", - priv->dev->name, ret); - kfree(callback_context); - return ret; - } - - /* waiting for the callback to finish before returnning */ - wait_for_completion(&callback_context->done); - kfree(callback_context); + struct ipoib_pseudo_header *phdr; - return ret; + phdr = skb_push(skb, sizeof(*phdr)); + memcpy(phdr->hwaddr, daddr, INFINIBAND_ALEN); } void ipoib_flush_paths(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ipoib_path *path, *tp; LIST_HEAD(remove_list); unsigned long flags; @@ -747,12 +742,12 @@ void ipoib_flush_paths(struct net_device *dev) } static void path_rec_completion(int status, - struct ib_sa_path_rec *pathrec, + struct sa_path_rec *pathrec, void *path_ptr) { struct ipoib_path *path = path_ptr; struct net_device *dev = path->dev; - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ipoib_ah *ah = NULL; struct ipoib_ah *old_ah = NULL; struct ipoib_neigh *neigh, *tn; @@ -762,7 +757,8 @@ static void path_rec_completion(int status, if (!status) ipoib_dbg(priv, "PathRec LID 0x%04x for GID %pI6\n", - be16_to_cpu(pathrec->dlid), pathrec->dgid.raw); + be32_to_cpu(sa_path_get_dlid(pathrec)), + pathrec->dgid.raw); else ipoib_dbg(priv, "PathRec status %d for GID %pI6\n", status, path->pathrec.dgid.raw); @@ -770,7 +766,7 @@ static void path_rec_completion(int status, skb_queue_head_init(&skqueue); if (!status) { - struct ib_ah_attr av; + struct rdma_ah_attr av; if (!ib_init_ah_from_path(priv->ca, priv->port, pathrec, &av)) ah = ipoib_create_ah(dev, priv->pd, &av); @@ -785,7 +781,8 @@ static void path_rec_completion(int status, path->ah = ah; ipoib_dbg(priv, "created address handle %p for LID 0x%04x, SL %d\n", - ah, be16_to_cpu(pathrec->dlid), pathrec->sl); + ah, be32_to_cpu(sa_path_get_dlid(pathrec)), + pathrec->sl); while ((skb = __skb_dequeue(&path->queue))) __skb_queue_tail(&skqueue, skb); @@ -834,16 +831,18 @@ static void path_rec_completion(int status, ipoib_put_ah(old_ah); while ((skb = __skb_dequeue(&skqueue))) { + int ret; skb->dev = dev; - if (dev_queue_xmit(skb)) - ipoib_warn(priv, "dev_queue_xmit failed " - "to requeue packet\n"); + ret = dev_queue_xmit(skb); + if (ret) + ipoib_warn(priv, "%s: dev_queue_xmit failed to re-queue packet, ret:%d\n", + __func__, ret); } } static struct ipoib_path *path_rec_create(struct net_device *dev, void *gid) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ipoib_path *path; if (!priv->broadcast) @@ -859,6 +858,10 @@ static struct ipoib_path *path_rec_create(struct net_device *dev, void *gid) 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); @@ -871,7 +874,7 @@ static struct ipoib_path *path_rec_create(struct net_device *dev, void *gid) static int path_rec_start(struct net_device *dev, struct ipoib_path *path) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); ipoib_dbg(priv, "Start path record lookup for %pI6\n", path->pathrec.dgid.raw); @@ -902,7 +905,8 @@ static int path_rec_start(struct net_device *dev, static void neigh_add_path(struct sk_buff *skb, u8 *daddr, struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); + struct rdma_netdev *rn = netdev_priv(dev); struct ipoib_path *path; struct ipoib_neigh *neigh; unsigned long flags; @@ -940,8 +944,7 @@ static void neigh_add_path(struct sk_buff *skb, u8 *daddr, } if (skb_queue_len(&neigh->queue) < IPOIB_MAX_PATH_REC_QUEUE) { - /* put pseudoheader back on for next time */ - skb_push(skb, IPOIB_PSEUDO_LEN); + push_pseudo_header(skb, neigh->daddr); __skb_queue_tail(&neigh->queue, skb); } else { ipoib_warn(priv, "queue length limit %d. Packet drop.\n", @@ -950,7 +953,8 @@ static void neigh_add_path(struct sk_buff *skb, u8 *daddr, } } else { spin_unlock_irqrestore(&priv->lock, flags); - ipoib_send(dev, skb, path->ah, IPOIB_QPN(daddr)); + path->ah->last_send = rn->send(dev, skb, path->ah->ah, + IPOIB_QPN(daddr)); ipoib_neigh_put(neigh); return; } @@ -959,10 +963,12 @@ static void neigh_add_path(struct sk_buff *skb, u8 *daddr, if (!path->query && path_rec_start(dev, path)) goto err_path; - if (skb_queue_len(&neigh->queue) < IPOIB_MAX_PATH_REC_QUEUE) + if (skb_queue_len(&neigh->queue) < IPOIB_MAX_PATH_REC_QUEUE) { + push_pseudo_header(skb, neigh->daddr); __skb_queue_tail(&neigh->queue, skb); - else + } else { goto err_drop; + } } spin_unlock_irqrestore(&priv->lock, flags); @@ -982,7 +988,8 @@ err_drop: static void unicast_arp_send(struct sk_buff *skb, struct net_device *dev, struct ipoib_pseudo_header *phdr) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); + struct rdma_netdev *rn = netdev_priv(dev); struct ipoib_path *path; unsigned long flags; @@ -998,8 +1005,7 @@ static void unicast_arp_send(struct sk_buff *skb, struct net_device *dev, } if (path) { if (skb_queue_len(&path->queue) < IPOIB_MAX_PATH_REC_QUEUE) { - /* put pseudoheader back on for next time */ - skb_push(skb, IPOIB_PSEUDO_LEN); + push_pseudo_header(skb, phdr->hwaddr); __skb_queue_tail(&path->queue, skb); } else { ++dev->stats.tx_dropped; @@ -1023,16 +1029,16 @@ static void unicast_arp_send(struct sk_buff *skb, struct net_device *dev, } if (path->ah) { - ipoib_dbg(priv, "Send unicast ARP to %04x\n", - be16_to_cpu(path->pathrec.dlid)); + ipoib_dbg(priv, "Send unicast ARP to %08x\n", + be32_to_cpu(sa_path_get_dlid(&path->pathrec))); spin_unlock_irqrestore(&priv->lock, flags); - ipoib_send(dev, skb, path->ah, IPOIB_QPN(phdr->hwaddr)); + path->ah->last_send = rn->send(dev, skb, path->ah->ah, + IPOIB_QPN(phdr->hwaddr)); return; } else if ((path->query || !path_rec_start(dev, path)) && skb_queue_len(&path->queue) < IPOIB_MAX_PATH_REC_QUEUE) { - /* put pseudoheader back on for next time */ - skb_push(skb, IPOIB_PSEUDO_LEN); + push_pseudo_header(skb, phdr->hwaddr); __skb_queue_tail(&path->queue, skb); } else { ++dev->stats.tx_dropped; @@ -1044,7 +1050,8 @@ static void unicast_arp_send(struct sk_buff *skb, struct net_device *dev, static int ipoib_start_xmit(struct sk_buff *skb, struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); + struct rdma_netdev *rn = netdev_priv(dev); struct ipoib_neigh *neigh; struct ipoib_pseudo_header *phdr; struct ipoib_header *header; @@ -1108,13 +1115,13 @@ send_using_neigh: goto unref; } } else if (neigh->ah) { - ipoib_send(dev, skb, neigh->ah, IPOIB_QPN(phdr->hwaddr)); + neigh->ah->last_send = rn->send(dev, skb, neigh->ah->ah, + IPOIB_QPN(phdr->hwaddr)); goto unref; } if (skb_queue_len(&neigh->queue) < IPOIB_MAX_PATH_REC_QUEUE) { - /* put pseudoheader back on for next time */ - skb_push(skb, sizeof(*phdr)); + push_pseudo_header(skb, phdr->hwaddr); spin_lock_irqsave(&priv->lock, flags); __skb_queue_tail(&neigh->queue, skb); spin_unlock_irqrestore(&priv->lock, flags); @@ -1131,7 +1138,7 @@ unref: static void ipoib_timeout(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); ipoib_warn(priv, "transmit timeout: latency %d msecs\n", jiffies_to_msecs(jiffies - dev_trans_start(dev))); @@ -1146,10 +1153,9 @@ static int ipoib_hard_header(struct sk_buff *skb, unsigned short type, const void *daddr, const void *saddr, unsigned len) { - struct ipoib_pseudo_header *phdr; struct ipoib_header *header; - header = (struct ipoib_header *) skb_push(skb, sizeof *header); + header = skb_push(skb, sizeof *header); header->proto = htons(type); header->reserved = 0; @@ -1159,15 +1165,14 @@ static int ipoib_hard_header(struct sk_buff *skb, * destination address into skb hard header so we can figure out where * to send the packet later. */ - phdr = (struct ipoib_pseudo_header *) skb_push(skb, sizeof(*phdr)); - memcpy(phdr->hwaddr, daddr, INFINIBAND_ALEN); + push_pseudo_header(skb, daddr); return IPOIB_HARD_LEN; } static void ipoib_set_mcast_list(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); if (!test_bit(IPOIB_FLAG_OPER_UP, &priv->flags)) { ipoib_dbg(priv, "IPOIB_FLAG_OPER_UP not set"); @@ -1179,7 +1184,7 @@ static void ipoib_set_mcast_list(struct net_device *dev) static int ipoib_get_iflink(const struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); /* parent interface */ if (!test_bit(IPOIB_FLAG_SUBINTERFACE, &priv->flags)) @@ -1207,7 +1212,7 @@ static u32 ipoib_addr_hash(struct ipoib_neigh_hash *htbl, u8 *daddr) struct ipoib_neigh *ipoib_neigh_get(struct net_device *dev, u8 *daddr) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ipoib_neigh_table *ntbl = &priv->ntbl; struct ipoib_neigh_hash *htbl; struct ipoib_neigh *neigh = NULL; @@ -1286,7 +1291,7 @@ static void __ipoib_reap_neigh(struct ipoib_dev_priv *priv) rcu_dereference_protected(neigh->hnext, lockdep_is_held(&priv->lock))); /* remove from path/mc list */ - list_del(&neigh->list); + list_del_init(&neigh->list); call_rcu(&neigh->rcu, ipoib_neigh_reclaim); } else { np = &neigh->hnext; @@ -1336,7 +1341,7 @@ static struct ipoib_neigh *ipoib_neigh_ctor(u8 *daddr, struct ipoib_neigh *ipoib_neigh_alloc(u8 *daddr, struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ipoib_neigh_table *ntbl = &priv->ntbl; struct ipoib_neigh_hash *htbl; struct ipoib_neigh *neigh; @@ -1393,7 +1398,7 @@ void ipoib_neigh_dtor(struct ipoib_neigh *neigh) { /* neigh reference count was dropprd to zero */ struct net_device *dev = neigh->dev; - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct sk_buff *skb; if (neigh->ah) ipoib_put_ah(neigh->ah); @@ -1403,7 +1408,7 @@ void ipoib_neigh_dtor(struct ipoib_neigh *neigh) } if (ipoib_cm_get(neigh)) ipoib_cm_destroy_tx(ipoib_cm_get(neigh)); - ipoib_dbg(netdev_priv(dev), + ipoib_dbg(ipoib_priv(dev), "neigh free for %06x %pI6\n", IPOIB_QPN(neigh->daddr), neigh->daddr + 4); @@ -1425,7 +1430,7 @@ static void ipoib_neigh_reclaim(struct rcu_head *rp) void ipoib_neigh_free(struct ipoib_neigh *neigh) { struct net_device *dev = neigh->dev; - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ipoib_neigh_table *ntbl = &priv->ntbl; struct ipoib_neigh_hash *htbl; struct ipoib_neigh __rcu **np; @@ -1450,7 +1455,7 @@ void ipoib_neigh_free(struct ipoib_neigh *neigh) rcu_dereference_protected(neigh->hnext, lockdep_is_held(&priv->lock))); /* remove from parent list */ - list_del(&neigh->list); + list_del_init(&neigh->list); call_rcu(&neigh->rcu, ipoib_neigh_reclaim); return; } else { @@ -1508,7 +1513,7 @@ static void neigh_hash_free_rcu(struct rcu_head *head) void ipoib_del_neighs_by_gid(struct net_device *dev, u8 *gid) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ipoib_neigh_table *ntbl = &priv->ntbl; struct ipoib_neigh_hash *htbl; unsigned long flags; @@ -1535,7 +1540,7 @@ void ipoib_del_neighs_by_gid(struct net_device *dev, u8 *gid) rcu_dereference_protected(neigh->hnext, lockdep_is_held(&priv->lock))); /* remove from parent list */ - list_del(&neigh->list); + list_del_init(&neigh->list); call_rcu(&neigh->rcu, ipoib_neigh_reclaim); } else { np = &neigh->hnext; @@ -1577,7 +1582,7 @@ static void ipoib_flush_neighs(struct ipoib_dev_priv *priv) rcu_dereference_protected(neigh->hnext, lockdep_is_held(&priv->lock))); /* remove from path/mc list */ - list_del(&neigh->list); + list_del_init(&neigh->list); call_rcu(&neigh->rcu, ipoib_neigh_reclaim); } } @@ -1594,7 +1599,7 @@ out_unlock: static void ipoib_neigh_hash_uninit(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); int stopped; ipoib_dbg(priv, "ipoib_neigh_hash_uninit\n"); @@ -1611,10 +1616,28 @@ static void ipoib_neigh_hash_uninit(struct net_device *dev) wait_for_completion(&priv->ntbl.deleted); } +static void ipoib_dev_uninit_default(struct net_device *dev) +{ + struct ipoib_dev_priv *priv = ipoib_priv(dev); + + ipoib_transport_dev_cleanup(dev); -int ipoib_dev_init(struct net_device *dev, struct ib_device *ca, int port) + netif_napi_del(&priv->napi); + + ipoib_cm_dev_cleanup(dev); + + kfree(priv->rx_ring); + vfree(priv->tx_ring); + + priv->rx_ring = NULL; + priv->tx_ring = NULL; +} + +static int ipoib_dev_init_default(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); + + netif_napi_add(dev, &priv->napi, ipoib_poll, NAPI_POLL_WEIGHT); /* Allocate RX/TX "rings" to hold queued skbs */ priv->rx_ring = kzalloc(ipoib_recvq_size * sizeof *priv->rx_ring, @@ -1625,46 +1648,112 @@ int ipoib_dev_init(struct net_device *dev, struct ib_device *ca, int port) 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", - ca->name, ipoib_sendq_size); + priv->ca->name, ipoib_sendq_size); goto out_rx_ring_cleanup; } /* priv->tx_head, tx_tail & tx_outstanding are already 0 */ - if (ipoib_ib_dev_init(dev, ca, port)) + if (ipoib_transport_dev_init(dev, priv->ca)) { + pr_warn("%s: ipoib_transport_dev_init failed\n", + priv->ca->name); goto out_tx_ring_cleanup; + } + + /* after qp created set dev address */ + priv->dev->dev_addr[1] = (priv->qp->qp_num >> 16) & 0xff; + priv->dev->dev_addr[2] = (priv->qp->qp_num >> 8) & 0xff; + priv->dev->dev_addr[3] = (priv->qp->qp_num) & 0xff; + + setup_timer(&priv->poll_timer, ipoib_ib_tx_timer_func, + (unsigned long)dev); + + return 0; + +out_tx_ring_cleanup: + vfree(priv->tx_ring); + +out_rx_ring_cleanup: + kfree(priv->rx_ring); + +out: + netif_napi_del(&priv->napi); + return -ENOMEM; +} + +int ipoib_dev_init(struct net_device *dev, struct ib_device *ca, int port) +{ + struct ipoib_dev_priv *priv = ipoib_priv(dev); + int ret = -ENOMEM; + + priv->ca = ca; + priv->port = port; + priv->qp = NULL; /* - * Must be after ipoib_ib_dev_init so we can allocate a per - * device wq there and use it here + * the various IPoIB tasks assume they will never race against + * themselves, so always use a single thread workqueue */ - if (ipoib_neigh_hash_init(priv) < 0) + priv->wq = alloc_ordered_workqueue("ipoib_wq", WQ_MEM_RECLAIM); + if (!priv->wq) { + pr_warn("%s: failed to allocate device WQ\n", dev->name); + goto out; + } + + /* create pd, which used both for control and datapath*/ + priv->pd = ib_alloc_pd(priv->ca, 0); + if (IS_ERR(priv->pd)) { + pr_warn("%s: failed to allocate PD\n", ca->name); + goto clean_wq; + } + + ret = priv->rn_ops->ndo_init(dev); + if (ret) { + pr_warn("%s failed to init HW resource\n", dev->name); + goto out_free_pd; + } + + if (ipoib_neigh_hash_init(priv) < 0) { + pr_warn("%s failed to init neigh hash\n", dev->name); goto out_dev_uninit; + } + + if (dev->flags & IFF_UP) { + if (ipoib_ib_dev_open(dev)) { + pr_warn("%s failed to open device\n", dev->name); + ret = -ENODEV; + goto out_dev_uninit; + } + } return 0; out_dev_uninit: ipoib_ib_dev_cleanup(dev); -out_tx_ring_cleanup: - vfree(priv->tx_ring); +out_free_pd: + if (priv->pd) { + ib_dealloc_pd(priv->pd); + priv->pd = NULL; + } -out_rx_ring_cleanup: - kfree(priv->rx_ring); +clean_wq: + if (priv->wq) { + destroy_workqueue(priv->wq); + priv->wq = NULL; + } out: - return -ENOMEM; + return ret; } void ipoib_dev_cleanup(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev), *cpriv, *tcpriv; + struct ipoib_dev_priv *priv = ipoib_priv(dev), *cpriv, *tcpriv; LIST_HEAD(head); ASSERT_RTNL(); - ipoib_delete_debug_files(dev); - /* Delete any child interfaces first */ list_for_each_entry_safe(cpriv, tcpriv, &priv->child_intfs, list) { /* Stop GC on child */ @@ -1674,24 +1763,21 @@ void ipoib_dev_cleanup(struct net_device *dev) } unregister_netdevice_many(&head); - /* - * Must be before ipoib_ib_dev_cleanup or we delete an in use - * work queue - */ ipoib_neigh_hash_uninit(dev); ipoib_ib_dev_cleanup(dev); - kfree(priv->rx_ring); - vfree(priv->tx_ring); - - priv->rx_ring = NULL; - priv->tx_ring = NULL; + /* no more works over the priv->wq */ + if (priv->wq) { + flush_workqueue(priv->wq); + destroy_workqueue(priv->wq); + priv->wq = NULL; + } } static int ipoib_set_vf_link_state(struct net_device *dev, int vf, int link_state) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); return ib_set_vf_link_state(priv->ca, vf, priv->port, link_state); } @@ -1699,7 +1785,7 @@ static int ipoib_set_vf_link_state(struct net_device *dev, int vf, int link_stat static int ipoib_get_vf_config(struct net_device *dev, int vf, struct ifla_vf_info *ivf) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); int err; err = ib_get_vf_config(priv->ca, vf, priv->port, ivf); @@ -1713,7 +1799,7 @@ static int ipoib_get_vf_config(struct net_device *dev, int vf, static int ipoib_set_vf_guid(struct net_device *dev, int vf, u64 guid, int type) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); if (type != IFLA_VF_IB_NODE_GUID && type != IFLA_VF_IB_PORT_GUID) return -EINVAL; @@ -1724,7 +1810,7 @@ static int ipoib_set_vf_guid(struct net_device *dev, int vf, u64 guid, int type) static int ipoib_get_vf_stats(struct net_device *dev, int vf, struct ifla_vf_stats *vf_stats) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); return ib_get_vf_stats(priv->ca, vf, priv->port, vf_stats); } @@ -1748,6 +1834,7 @@ static const struct net_device_ops ipoib_netdev_ops_pf = { .ndo_get_vf_stats = ipoib_get_vf_stats, .ndo_set_vf_guid = ipoib_set_vf_guid, .ndo_set_mac_address = ipoib_set_mac, + .ndo_get_stats64 = ipoib_get_stats, }; static const struct net_device_ops ipoib_netdev_ops_vf = { @@ -1762,21 +1849,12 @@ static const struct net_device_ops ipoib_netdev_ops_vf = { .ndo_get_iflink = ipoib_get_iflink, }; -void ipoib_setup(struct net_device *dev) +void ipoib_setup_common(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); - - if (priv->hca_caps & IB_DEVICE_VIRTUAL_FUNCTION) - dev->netdev_ops = &ipoib_netdev_ops_vf; - else - dev->netdev_ops = &ipoib_netdev_ops_pf; - dev->header_ops = &ipoib_header_ops; ipoib_set_ethtool_ops(dev); - netif_napi_add(dev, &priv->napi, ipoib_poll, NAPI_POLL_WEIGHT); - dev->watchdog_timeo = HZ; dev->flags |= IFF_BROADCAST | IFF_MULTICAST; @@ -1790,11 +1868,14 @@ void ipoib_setup(struct net_device *dev) netif_keep_dst(dev); memcpy(dev->broadcast, ipv4_bcast_addr, INFINIBAND_ALEN); +} - priv->dev = dev; +static void ipoib_build_priv(struct net_device *dev) +{ + struct ipoib_dev_priv *priv = ipoib_priv(dev); + priv->dev = dev; spin_lock_init(&priv->lock); - init_rwsem(&priv->vlan_rwsem); INIT_LIST_HEAD(&priv->path_list); @@ -1812,22 +1893,100 @@ void ipoib_setup(struct net_device *dev) INIT_DELAYED_WORK(&priv->neigh_reap_task, ipoib_reap_neigh); } -struct ipoib_dev_priv *ipoib_intf_alloc(const char *name) +static const struct net_device_ops ipoib_netdev_default_pf = { + .ndo_init = ipoib_dev_init_default, + .ndo_uninit = ipoib_dev_uninit_default, + .ndo_open = ipoib_ib_dev_open_default, + .ndo_stop = ipoib_ib_dev_stop_default, +}; + +static struct net_device +*ipoib_create_netdev_default(struct ib_device *hca, + const char *name, + unsigned char name_assign_type, + void (*setup)(struct net_device *)) { struct net_device *dev; + struct rdma_netdev *rn; - dev = alloc_netdev((int)sizeof(struct ipoib_dev_priv), name, - NET_NAME_UNKNOWN, ipoib_setup); + dev = alloc_netdev((int)sizeof(struct rdma_netdev), + name, + name_assign_type, setup); if (!dev) return NULL; - return netdev_priv(dev); + rn = netdev_priv(dev); + + rn->send = ipoib_send; + rn->attach_mcast = ipoib_mcast_attach; + rn->detach_mcast = ipoib_mcast_detach; + rn->free_rdma_netdev = free_netdev; + rn->hca = hca; + + dev->netdev_ops = &ipoib_netdev_default_pf; + + return dev; +} + +static struct net_device *ipoib_get_netdev(struct ib_device *hca, u8 port, + const char *name) +{ + struct net_device *dev; + + if (hca->alloc_rdma_netdev) { + dev = hca->alloc_rdma_netdev(hca, port, + RDMA_NETDEV_IPOIB, name, + NET_NAME_UNKNOWN, + ipoib_setup_common); + if (IS_ERR_OR_NULL(dev) && PTR_ERR(dev) != -EOPNOTSUPP) + return NULL; + } + + if (!hca->alloc_rdma_netdev || PTR_ERR(dev) == -EOPNOTSUPP) + dev = ipoib_create_netdev_default(hca, name, NET_NAME_UNKNOWN, + ipoib_setup_common); + + return dev; +} + +struct ipoib_dev_priv *ipoib_intf_alloc(struct ib_device *hca, u8 port, + const char *name) +{ + struct net_device *dev; + struct ipoib_dev_priv *priv; + struct rdma_netdev *rn; + + priv = kzalloc(sizeof(*priv), GFP_KERNEL); + if (!priv) + return NULL; + + dev = ipoib_get_netdev(hca, port, name); + if (!dev) + goto free_priv; + + priv->rn_ops = dev->netdev_ops; + + /* fixme : should be after the query_cap */ + if (priv->hca_caps & IB_DEVICE_VIRTUAL_FUNCTION) + dev->netdev_ops = &ipoib_netdev_ops_vf; + else + dev->netdev_ops = &ipoib_netdev_ops_pf; + + rn = netdev_priv(dev); + rn->clnt_priv = priv; + ipoib_build_priv(dev); + + return priv; +free_priv: + kfree(priv); + return NULL; } static ssize_t show_pkey(struct device *dev, struct device_attribute *attr, char *buf) { - struct ipoib_dev_priv *priv = netdev_priv(to_net_dev(dev)); + struct net_device *ndev = to_net_dev(dev); + struct ipoib_dev_priv *priv = ipoib_priv(ndev); return sprintf(buf, "0x%04x\n", priv->pkey); } @@ -1836,14 +1995,15 @@ static DEVICE_ATTR(pkey, S_IRUGO, show_pkey, NULL); static ssize_t show_umcast(struct device *dev, struct device_attribute *attr, char *buf) { - struct ipoib_dev_priv *priv = netdev_priv(to_net_dev(dev)); + struct net_device *ndev = to_net_dev(dev); + struct ipoib_dev_priv *priv = ipoib_priv(ndev); return sprintf(buf, "%d\n", test_bit(IPOIB_FLAG_UMCAST, &priv->flags)); } void ipoib_set_umcast(struct net_device *ndev, int umcast_val) { - struct ipoib_dev_priv *priv = netdev_priv(ndev); + struct ipoib_dev_priv *priv = ipoib_priv(ndev); if (umcast_val > 0) { set_bit(IPOIB_FLAG_UMCAST, &priv->flags); @@ -1916,7 +2076,7 @@ static int ipoib_check_lladdr(struct net_device *dev, static int ipoib_set_mac(struct net_device *dev, void *addr) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct sockaddr_storage *ss = addr; int ret; @@ -1984,20 +2144,18 @@ int ipoib_add_pkey_attr(struct net_device *dev) return device_create_file(&dev->dev, &dev_attr_pkey); } -int ipoib_set_dev_features(struct ipoib_dev_priv *priv, struct ib_device *hca) +void ipoib_set_dev_features(struct ipoib_dev_priv *priv, struct ib_device *hca) { priv->hca_caps = hca->attrs.device_cap_flags; if (priv->hca_caps & IB_DEVICE_UD_IP_CSUM) { - priv->dev->hw_features = NETIF_F_IP_CSUM | NETIF_F_RXCSUM; + priv->dev->hw_features |= NETIF_F_IP_CSUM | NETIF_F_RXCSUM; if (priv->hca_caps & IB_DEVICE_UD_TSO) priv->dev->hw_features |= NETIF_F_TSO; priv->dev->features |= priv->dev->hw_features; } - - return 0; } static struct net_device *ipoib_add_port(const char *format, @@ -2007,11 +2165,11 @@ static struct net_device *ipoib_add_port(const char *format, struct ib_port_attr attr; int result = -ENOMEM; - priv = ipoib_intf_alloc(format); + priv = ipoib_intf_alloc(hca, port, format); if (!priv) goto alloc_mem_failed; - SET_NETDEV_DEV(priv->dev, hca->dma_device); + SET_NETDEV_DEV(priv->dev, hca->dev.parent); priv->dev->dev_id = port - 1; result = ib_query_port(hca, port, &attr); @@ -2037,9 +2195,7 @@ static struct net_device *ipoib_add_port(const char *format, goto device_init_failed; } - result = ipoib_set_dev_features(priv, hca); - if (result) - goto device_init_failed; + ipoib_set_dev_features(priv, hca); /* * Set the full membership bit, so that we join the right @@ -2083,8 +2239,7 @@ static struct net_device *ipoib_add_port(const char *format, goto register_failed; } - ipoib_create_debug_files(priv->dev); - + result = -ENOMEM; if (ipoib_cm_add_mode_attr(priv->dev)) goto sysfs_failed; if (ipoib_add_pkey_attr(priv->dev)) @@ -2099,7 +2254,6 @@ static struct net_device *ipoib_add_port(const char *format, return priv->dev; sysfs_failed: - ipoib_delete_debug_files(priv->dev); unregister_netdev(priv->dev); register_failed: @@ -2115,6 +2269,7 @@ event_failed: device_init_failed: free_netdev(priv->dev); + kfree(priv); alloc_mem_failed: return ERR_PTR(result); @@ -2139,7 +2294,7 @@ static void ipoib_add_one(struct ib_device *device) continue; dev = ipoib_add_port("ib%d", device, p); if (!IS_ERR(dev)) { - priv = netdev_priv(dev); + priv = ipoib_priv(dev); list_add_tail(&priv->list, dev_list); count++; } @@ -2155,13 +2310,15 @@ static void ipoib_add_one(struct ib_device *device) static void ipoib_remove_one(struct ib_device *device, void *client_data) { - struct ipoib_dev_priv *priv, *tmp; + struct ipoib_dev_priv *priv, *tmp, *cpriv, *tcpriv; struct list_head *dev_list = client_data; if (!dev_list) return; list_for_each_entry_safe(priv, tmp, dev_list, list) { + struct rdma_netdev *rn = netdev_priv(priv->dev); + ib_unregister_event_handler(&priv->event_handler); flush_workqueue(ipoib_workqueue); @@ -2178,12 +2335,23 @@ static void ipoib_remove_one(struct ib_device *device, void *client_data) flush_workqueue(priv->wq); unregister_netdev(priv->dev); - free_netdev(priv->dev); + rn->free_rdma_netdev(priv->dev); + + list_for_each_entry_safe(cpriv, tcpriv, &priv->child_intfs, list) + kfree(cpriv); + + kfree(priv); } kfree(dev_list); } +#ifdef CONFIG_INFINIBAND_IPOIB_DEBUG +static struct notifier_block ipoib_netdev_notifier = { + .notifier_call = ipoib_netdev_event, +}; +#endif + static int __init ipoib_init_module(void) { int ret; @@ -2236,6 +2404,9 @@ static int __init ipoib_init_module(void) if (ret) goto err_client; +#ifdef CONFIG_INFINIBAND_IPOIB_DEBUG + register_netdevice_notifier(&ipoib_netdev_notifier); +#endif return 0; err_client: @@ -2253,6 +2424,9 @@ err_fs: static void __exit ipoib_cleanup_module(void) { +#ifdef CONFIG_INFINIBAND_IPOIB_DEBUG + unregister_netdevice_notifier(&ipoib_netdev_notifier); +#endif ipoib_netlink_fini(); ib_unregister_client(&ipoib_client); ib_sa_unregister_client(&ipoib_sa_client); diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c index fddff403d5d2..057f58e6afca 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c @@ -114,7 +114,7 @@ static void ipoib_mcast_free(struct ipoib_mcast *mcast) struct net_device *dev = mcast->dev; int tx_dropped = 0; - ipoib_dbg_mcast(netdev_priv(dev), "deleting multicast group %pI6\n", + ipoib_dbg_mcast(ipoib_priv(dev), "deleting multicast group %pI6\n", mcast->mcmember.mgid.raw); /* remove all neigh connected to this mcast */ @@ -158,7 +158,7 @@ static struct ipoib_mcast *ipoib_mcast_alloc(struct net_device *dev, static struct ipoib_mcast *__ipoib_mcast_find(struct net_device *dev, void *mgid) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct rb_node *n = priv->multicast_tree.rb_node; while (n) { @@ -182,7 +182,7 @@ static struct ipoib_mcast *__ipoib_mcast_find(struct net_device *dev, void *mgid static int __ipoib_mcast_add(struct net_device *dev, struct ipoib_mcast *mcast) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct rb_node **n = &priv->multicast_tree.rb_node, *pn = NULL; while (*n) { @@ -212,8 +212,10 @@ static int ipoib_mcast_join_finish(struct ipoib_mcast *mcast, struct ib_sa_mcmember_rec *mcmember) { struct net_device *dev = mcast->dev; - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); + struct rdma_netdev *rn = netdev_priv(dev); struct ipoib_ah *ah; + struct rdma_ah_attr av; int ret; int set_qkey = 0; @@ -260,8 +262,9 @@ static int ipoib_mcast_join_finish(struct ipoib_mcast *mcast, return 0; } - ret = ipoib_mcast_attach(dev, be16_to_cpu(mcast->mcmember.mlid), - &mcast->mcmember.mgid, set_qkey); + ret = rn->attach_mcast(dev, priv->ca, &mcast->mcmember.mgid, + be16_to_cpu(mcast->mcmember.mlid), + set_qkey, priv->qkey); if (ret < 0) { ipoib_warn(priv, "couldn't attach QP to multicast group %pI6\n", mcast->mcmember.mgid.raw); @@ -271,40 +274,34 @@ static int ipoib_mcast_join_finish(struct ipoib_mcast *mcast, } } - { - struct ib_ah_attr av = { - .dlid = be16_to_cpu(mcast->mcmember.mlid), - .port_num = priv->port, - .sl = mcast->mcmember.sl, - .ah_flags = IB_AH_GRH, - .static_rate = mcast->mcmember.rate, - .grh = { - .flow_label = be32_to_cpu(mcast->mcmember.flow_label), - .hop_limit = mcast->mcmember.hop_limit, - .sgid_index = 0, - .traffic_class = mcast->mcmember.traffic_class - } - }; - av.grh.dgid = mcast->mcmember.mgid; - - ah = ipoib_create_ah(dev, priv->pd, &av); - if (IS_ERR(ah)) { - ipoib_warn(priv, "ib_address_create failed %ld\n", - -PTR_ERR(ah)); - /* use original error */ - return PTR_ERR(ah); - } else { - spin_lock_irq(&priv->lock); - mcast->ah = ah; - spin_unlock_irq(&priv->lock); - - ipoib_dbg_mcast(priv, "MGID %pI6 AV %p, LID 0x%04x, SL %d\n", - mcast->mcmember.mgid.raw, - mcast->ah->ah, - be16_to_cpu(mcast->mcmember.mlid), - mcast->mcmember.sl); - } + memset(&av, 0, sizeof(av)); + av.type = rdma_ah_find_type(priv->ca, priv->port); + rdma_ah_set_dlid(&av, be16_to_cpu(mcast->mcmember.mlid)), + rdma_ah_set_port_num(&av, priv->port); + rdma_ah_set_sl(&av, mcast->mcmember.sl); + rdma_ah_set_static_rate(&av, mcast->mcmember.rate); + + rdma_ah_set_grh(&av, &mcast->mcmember.mgid, + be32_to_cpu(mcast->mcmember.flow_label), + 0, mcast->mcmember.hop_limit, + mcast->mcmember.traffic_class); + + ah = ipoib_create_ah(dev, priv->pd, &av); + if (IS_ERR(ah)) { + ipoib_warn(priv, "ib_address_create failed %ld\n", + -PTR_ERR(ah)); + /* use original error */ + return PTR_ERR(ah); } + spin_lock_irq(&priv->lock); + mcast->ah = ah; + spin_unlock_irq(&priv->lock); + + ipoib_dbg_mcast(priv, "MGID %pI6 AV %p, LID 0x%04x, SL %d\n", + mcast->mcmember.mgid.raw, + mcast->ah->ah, + be16_to_cpu(mcast->mcmember.mlid), + mcast->mcmember.sl); /* actually send any queued packets */ netif_tx_lock_bh(dev); @@ -314,9 +311,11 @@ static int ipoib_mcast_join_finish(struct ipoib_mcast *mcast, netif_tx_unlock_bh(dev); skb->dev = dev; - if (dev_queue_xmit(skb)) - ipoib_warn(priv, "dev_queue_xmit failed to requeue packet\n"); + ret = dev_queue_xmit(skb); + if (ret) + ipoib_warn(priv, "%s:dev_queue_xmit failed to re-queue packet, ret:%d\n", + __func__, ret); netif_tx_lock_bh(dev); } netif_tx_unlock_bh(dev); @@ -329,7 +328,6 @@ void ipoib_mcast_carrier_on_task(struct work_struct *work) struct ipoib_dev_priv *priv = container_of(work, struct ipoib_dev_priv, carrier_on_task); struct ib_port_attr attr; - int ret; if (ib_query_port(priv->ca, priv->port, &attr) || attr.state != IB_PORT_ACTIVE) { @@ -342,11 +340,9 @@ void ipoib_mcast_carrier_on_task(struct work_struct *work) * because the broadcast group must always be joined first and is always * re-joined if the SM changes substantially. */ - ret = ipoib_check_sm_sendonly_fullmember_support(priv); - if (ret < 0) - pr_debug("%s failed query sm support for sendonly-fullmember (ret: %d)\n", - priv->dev->name, ret); - + priv->sm_fullmember_sendonly_support = + ib_sa_sendonly_fullmem_support(&ipoib_sa_client, + priv->ca, priv->port); /* * Take rtnl_lock to avoid racing with ipoib_stop() and * turning the carrier back on while a device is being @@ -373,7 +369,7 @@ static int ipoib_mcast_join_complete(int status, { struct ipoib_mcast *mcast = multicast->context; struct net_device *dev = mcast->dev; - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); ipoib_dbg_mcast(priv, "%sjoin completion for %pI6 (status %d)\n", test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags) ? @@ -475,7 +471,7 @@ out_locked: */ static int ipoib_mcast_join(struct net_device *dev, struct ipoib_mcast *mcast) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ib_sa_multicast *multicast; struct ib_sa_mcmember_rec rec = { .join_state = 1 @@ -487,6 +483,9 @@ static int ipoib_mcast_join(struct net_device *dev, struct ipoib_mcast *mcast) !test_bit(IPOIB_FLAG_OPER_UP, &priv->flags)) return -EINVAL; + init_completion(&mcast->done); + set_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags); + ipoib_dbg_mcast(priv, "joining MGID %pI6\n", mcast->mcmember.mgid.raw); rec.mgid = mcast->mcmember.mgid; @@ -645,8 +644,6 @@ void ipoib_mcast_join_task(struct work_struct *work) if (mcast->backoff == 1 || time_after_eq(jiffies, mcast->delay_until)) { /* Found the next unjoined group */ - init_completion(&mcast->done); - set_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags); if (ipoib_mcast_join(dev, mcast)) { spin_unlock_irq(&priv->lock); return; @@ -666,17 +663,15 @@ out: queue_delayed_work(priv->wq, &priv->mcast_task, delay_until - jiffies); } - if (mcast) { - init_completion(&mcast->done); - set_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags); + if (mcast) ipoib_mcast_join(dev, mcast); - } + spin_unlock_irq(&priv->lock); } -int ipoib_mcast_start_thread(struct net_device *dev) +void ipoib_mcast_start_thread(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); unsigned long flags; ipoib_dbg_mcast(priv, "starting multicast thread\n"); @@ -684,13 +679,11 @@ int ipoib_mcast_start_thread(struct net_device *dev) spin_lock_irqsave(&priv->lock, flags); __ipoib_mcast_schedule_join_thread(priv, NULL, 0); spin_unlock_irqrestore(&priv->lock, flags); - - return 0; } int ipoib_mcast_stop_thread(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); unsigned long flags; ipoib_dbg_mcast(priv, "stopping multicast thread\n"); @@ -706,7 +699,8 @@ int ipoib_mcast_stop_thread(struct net_device *dev) static int ipoib_mcast_leave(struct net_device *dev, struct ipoib_mcast *mcast) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); + struct rdma_netdev *rn = netdev_priv(dev); int ret = 0; if (test_and_clear_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags)) @@ -720,8 +714,8 @@ static int ipoib_mcast_leave(struct net_device *dev, struct ipoib_mcast *mcast) mcast->mcmember.mgid.raw); /* Remove ourselves from the multicast group */ - ret = ib_detach_mcast(priv->qp, &mcast->mcmember.mgid, - be16_to_cpu(mcast->mcmember.mlid)); + ret = rn->detach_mcast(dev, priv->ca, &mcast->mcmember.mgid, + be16_to_cpu(mcast->mcmember.mlid)); if (ret) ipoib_warn(priv, "ib_detach_mcast failed (result = %d)\n", ret); } else if (!test_bit(IPOIB_MCAST_FLAG_SENDONLY, &mcast->flags)) @@ -762,7 +756,8 @@ void ipoib_mcast_remove_list(struct list_head *remove_list) void ipoib_mcast_send(struct net_device *dev, u8 *daddr, struct sk_buff *skb) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); + struct rdma_netdev *rn = netdev_priv(dev); struct ipoib_mcast *mcast; unsigned long flags; void *mgid = daddr + 4; @@ -825,7 +820,8 @@ void ipoib_mcast_send(struct net_device *dev, u8 *daddr, struct sk_buff *skb) } } spin_unlock_irqrestore(&priv->lock, flags); - ipoib_send(dev, skb, mcast->ah, IB_MULTICAST_QPN); + mcast->ah->last_send = rn->send(dev, skb, mcast->ah->ah, + IB_MULTICAST_QPN); if (neigh) ipoib_neigh_put(neigh); return; @@ -837,7 +833,7 @@ unlock: void ipoib_mcast_dev_flush(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); LIST_HEAD(remove_list); struct ipoib_mcast *mcast, *tmcast; unsigned long flags; @@ -1029,7 +1025,7 @@ struct ipoib_mcast_iter *ipoib_mcast_iter_init(struct net_device *dev) int ipoib_mcast_iter_next(struct ipoib_mcast_iter *iter) { - struct ipoib_dev_priv *priv = netdev_priv(iter->dev); + struct ipoib_dev_priv *priv = ipoib_priv(iter->dev); struct rb_node *n; struct ipoib_mcast *mcast; int ret = 1; diff --git a/drivers/infiniband/ulp/ipoib/ipoib_netlink.c b/drivers/infiniband/ulp/ipoib/ipoib_netlink.c index cdc7df4fdb8a..3e44087935ae 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_netlink.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_netlink.c @@ -44,7 +44,7 @@ static const struct nla_policy ipoib_policy[IFLA_IPOIB_MAX + 1] = { static int ipoib_fill_info(struct sk_buff *skb, const struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); u16 val; if (nla_put_u16(skb, IFLA_IPOIB_PKEY, priv->pkey)) @@ -64,8 +64,9 @@ nla_put_failure: return -EMSGSIZE; } -static int ipoib_changelink(struct net_device *dev, - struct nlattr *tb[], struct nlattr *data[]) +static int ipoib_changelink(struct net_device *dev, struct nlattr *tb[], + struct nlattr *data[], + struct netlink_ext_ack *extack) { u16 mode, umcast; int ret = 0; @@ -93,7 +94,8 @@ out_err: } static int ipoib_new_child_link(struct net *src_net, struct net_device *dev, - struct nlattr *tb[], struct nlattr *data[]) + struct nlattr *tb[], struct nlattr *data[], + struct netlink_ext_ack *extack) { struct net_device *pdev; struct ipoib_dev_priv *ppriv; @@ -107,7 +109,7 @@ static int ipoib_new_child_link(struct net *src_net, struct net_device *dev, if (!pdev || pdev->type != ARPHRD_INFINIBAND) return -ENODEV; - ppriv = netdev_priv(pdev); + ppriv = ipoib_priv(pdev); if (test_bit(IPOIB_FLAG_SUBINTERFACE, &ppriv->flags)) { ipoib_warn(ppriv, "child creation disallowed for child devices\n"); @@ -129,10 +131,11 @@ static int ipoib_new_child_link(struct net *src_net, struct net_device *dev, */ child_pkey |= 0x8000; - err = __ipoib_vlan_add(ppriv, netdev_priv(dev), child_pkey, IPOIB_RTNL_CHILD); + err = __ipoib_vlan_add(ppriv, ipoib_priv(dev), + child_pkey, IPOIB_RTNL_CHILD); if (!err && data) - err = ipoib_changelink(dev, tb, data); + err = ipoib_changelink(dev, tb, data, extack); return err; } @@ -140,8 +143,8 @@ static void ipoib_unregister_child_dev(struct net_device *dev, struct list_head { struct ipoib_dev_priv *priv, *ppriv; - priv = netdev_priv(dev); - ppriv = netdev_priv(priv->parent); + priv = ipoib_priv(dev); + ppriv = ipoib_priv(priv->parent); down_write(&ppriv->vlan_rwsem); unregister_netdevice_queue(dev, head); @@ -161,7 +164,7 @@ static struct rtnl_link_ops ipoib_link_ops __read_mostly = { .maxtype = IFLA_IPOIB_MAX, .policy = ipoib_policy, .priv_size = sizeof(struct ipoib_dev_priv), - .setup = ipoib_setup, + .setup = ipoib_setup_common, .newlink = ipoib_new_child_link, .changelink = ipoib_changelink, .dellink = ipoib_unregister_child_dev, diff --git a/drivers/infiniband/ulp/ipoib/ipoib_verbs.c b/drivers/infiniband/ulp/ipoib/ipoib_verbs.c index 189dcd1709d2..bb64baf25309 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_verbs.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_verbs.c @@ -35,9 +35,10 @@ #include "ipoib.h" -int ipoib_mcast_attach(struct net_device *dev, u16 mlid, union ib_gid *mgid, int set_qkey) +int ipoib_mcast_attach(struct net_device *dev, struct ib_device *hca, + union ib_gid *mgid, u16 mlid, int set_qkey, u32 qkey) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ib_qp_attr *qp_attr = NULL; int ret; u16 pkey_index; @@ -56,7 +57,7 @@ int ipoib_mcast_attach(struct net_device *dev, u16 mlid, union ib_gid *mgid, int goto out; /* set correct QKey for QP */ - qp_attr->qkey = priv->qkey; + qp_attr->qkey = qkey; ret = ib_modify_qp(priv->qp, qp_attr, IB_QP_QKEY); if (ret) { ipoib_warn(priv, "failed to modify QP, ret = %d\n", ret); @@ -74,9 +75,20 @@ out: return ret; } +int ipoib_mcast_detach(struct net_device *dev, struct ib_device *hca, + union ib_gid *mgid, u16 mlid) +{ + struct ipoib_dev_priv *priv = ipoib_priv(dev); + int ret; + + ret = ib_detach_mcast(priv->qp, mgid, mlid); + + return ret; +} + int ipoib_init_qp(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); int ret; struct ib_qp_attr qp_attr; int attr_mask; @@ -130,7 +142,7 @@ out_fail: int ipoib_transport_dev_init(struct net_device *dev, struct ib_device *ca) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); struct ib_qp_init_attr init_attr = { .cap = { .max_send_wr = ipoib_sendq_size, @@ -147,22 +159,6 @@ int ipoib_transport_dev_init(struct net_device *dev, struct ib_device *ca) int ret, size; int i; - priv->pd = ib_alloc_pd(priv->ca, 0); - if (IS_ERR(priv->pd)) { - printk(KERN_WARNING "%s: failed to allocate PD\n", ca->name); - return -ENODEV; - } - - /* - * the various IPoIB tasks assume they will never race against - * themselves, so always use a single thread workqueue - */ - priv->wq = alloc_ordered_workqueue("ipoib_wq", WQ_MEM_RECLAIM); - if (!priv->wq) { - printk(KERN_WARNING "ipoib: failed to allocate device WQ\n"); - goto out_free_pd; - } - size = ipoib_recvq_size + 1; ret = ipoib_cm_dev_init(dev); if (!ret) { @@ -173,7 +169,7 @@ int ipoib_transport_dev_init(struct net_device *dev, struct ib_device *ca) size += ipoib_recvq_size * ipoib_max_conn_qp; } else if (ret != -ENOSYS) - goto out_free_wq; + return -ENODEV; cq_attr.cqe = size; priv->recv_cq = ib_create_cq(priv->ca, ipoib_ib_completion, NULL, @@ -212,10 +208,6 @@ int ipoib_transport_dev_init(struct net_device *dev, struct ib_device *ca) goto out_free_send_cq; } - priv->dev->dev_addr[1] = (priv->qp->qp_num >> 16) & 0xff; - priv->dev->dev_addr[2] = (priv->qp->qp_num >> 8) & 0xff; - priv->dev->dev_addr[3] = (priv->qp->qp_num ) & 0xff; - for (i = 0; i < MAX_SKB_FRAGS + 1; ++i) priv->tx_sge[i].lkey = priv->pd->local_dma_lkey; @@ -247,26 +239,18 @@ out_free_recv_cq: out_cm_dev_cleanup: ipoib_cm_dev_cleanup(dev); -out_free_wq: - destroy_workqueue(priv->wq); - priv->wq = NULL; - -out_free_pd: - ib_dealloc_pd(priv->pd); - return -ENODEV; } void ipoib_transport_dev_cleanup(struct net_device *dev) { - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); if (priv->qp) { if (ib_destroy_qp(priv->qp)) ipoib_warn(priv, "ib_qp_destroy failed\n"); priv->qp = NULL; - clear_bit(IPOIB_PKEY_ASSIGNED, &priv->flags); } if (ib_destroy_cq(priv->send_cq)) @@ -274,16 +258,6 @@ void ipoib_transport_dev_cleanup(struct net_device *dev) if (ib_destroy_cq(priv->recv_cq)) ipoib_warn(priv, "ib_cq_destroy (recv) failed\n"); - - ipoib_cm_dev_cleanup(dev); - - if (priv->wq) { - flush_workqueue(priv->wq); - destroy_workqueue(priv->wq); - priv->wq = NULL; - } - - ib_dealloc_pd(priv->pd); } void ipoib_event(struct ib_event_handler *handler, diff --git a/drivers/infiniband/ulp/ipoib/ipoib_vlan.c b/drivers/infiniband/ulp/ipoib/ipoib_vlan.c index fd811115af49..081b33deff1b 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_vlan.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_vlan.c @@ -31,6 +31,7 @@ */ #include <linux/module.h> +#include <linux/sched/signal.h> #include <linux/init.h> #include <linux/seq_file.h> @@ -43,7 +44,7 @@ static ssize_t show_parent(struct device *d, struct device_attribute *attr, char *buf) { struct net_device *dev = to_net_dev(d); - struct ipoib_dev_priv *priv = netdev_priv(dev); + struct ipoib_dev_priv *priv = ipoib_priv(dev); return sprintf(buf, "%s\n", priv->parent->name); } @@ -61,9 +62,7 @@ int __ipoib_vlan_add(struct ipoib_dev_priv *ppriv, struct ipoib_dev_priv *priv, priv->parent = ppriv->dev; set_bit(IPOIB_FLAG_SUBINTERFACE, &priv->flags); - result = ipoib_set_dev_features(priv, ppriv->ca); - if (result) - goto err; + ipoib_set_dev_features(priv, ppriv->ca); priv->pkey = pkey; @@ -87,8 +86,6 @@ int __ipoib_vlan_add(struct ipoib_dev_priv *ppriv, struct ipoib_dev_priv *priv, goto register_failed; } - ipoib_create_debug_files(priv->dev); - /* RTNL childs don't need proprietary sysfs entries */ if (type == IPOIB_LEGACY_CHILD) { if (ipoib_cm_add_mode_attr(priv->dev)) @@ -109,7 +106,6 @@ int __ipoib_vlan_add(struct ipoib_dev_priv *ppriv, struct ipoib_dev_priv *priv, sysfs_failed: result = -ENOMEM; - ipoib_delete_debug_files(priv->dev); unregister_netdevice(priv->dev); register_failed: @@ -129,20 +125,21 @@ int ipoib_vlan_add(struct net_device *pdev, unsigned short pkey) if (!capable(CAP_NET_ADMIN)) return -EPERM; - ppriv = netdev_priv(pdev); + ppriv = ipoib_priv(pdev); if (test_bit(IPOIB_FLAG_GOING_DOWN, &ppriv->flags)) return -EPERM; snprintf(intf_name, sizeof intf_name, "%s.%04x", ppriv->dev->name, pkey); - priv = ipoib_intf_alloc(intf_name); - if (!priv) - return -ENOMEM; if (!rtnl_trylock()) return restart_syscall(); + priv = ipoib_intf_alloc(ppriv->ca, ppriv->port, intf_name); + if (!priv) + return -ENOMEM; + down_write(&ppriv->vlan_rwsem); /* @@ -168,11 +165,13 @@ int ipoib_vlan_add(struct net_device *pdev, unsigned short pkey) out: up_write(&ppriv->vlan_rwsem); - if (result) - free_netdev(priv->dev); - rtnl_unlock(); + if (result) { + free_netdev(priv->dev); + kfree(priv); + } + return result; } @@ -184,7 +183,7 @@ int ipoib_vlan_delete(struct net_device *pdev, unsigned short pkey) if (!capable(CAP_NET_ADMIN)) return -EPERM; - ppriv = netdev_priv(pdev); + ppriv = ipoib_priv(pdev); if (test_bit(IPOIB_FLAG_GOING_DOWN, &ppriv->flags)) return -EPERM; @@ -196,7 +195,6 @@ int ipoib_vlan_delete(struct net_device *pdev, unsigned short pkey) list_for_each_entry_safe(priv, tpriv, &ppriv->child_intfs, list) { if (priv->pkey == pkey && priv->child_type == IPOIB_LEGACY_CHILD) { - unregister_netdevice(priv->dev); list_del(&priv->list); dev = priv->dev; break; @@ -204,10 +202,16 @@ int ipoib_vlan_delete(struct net_device *pdev, unsigned short pkey) } up_write(&ppriv->vlan_rwsem); + if (dev) { + ipoib_dbg(ppriv, "delete child vlan %s\n", dev->name); + unregister_netdevice(dev); + } + rtnl_unlock(); if (dev) { free_netdev(dev); + kfree(priv); return 0; } diff --git a/drivers/infiniband/ulp/iser/iscsi_iser.c b/drivers/infiniband/ulp/iser/iscsi_iser.c index e71af717e71b..37b33d708c2d 100644 --- a/drivers/infiniband/ulp/iser/iscsi_iser.c +++ b/drivers/infiniband/ulp/iser/iscsi_iser.c @@ -83,6 +83,7 @@ static struct scsi_host_template iscsi_iser_sht; static struct iscsi_transport iscsi_iser_transport; static struct scsi_transport_template *iscsi_iser_scsi_transport; static struct workqueue_struct *release_wq; +static DEFINE_MUTEX(unbind_iser_conn_mutex); struct iser_global ig; int iser_debug_level = 0; @@ -550,12 +551,14 @@ iscsi_iser_conn_stop(struct iscsi_cls_conn *cls_conn, int flag) */ if (iser_conn) { mutex_lock(&iser_conn->state_mutex); + mutex_lock(&unbind_iser_conn_mutex); iser_conn_terminate(iser_conn); iscsi_conn_stop(cls_conn, flag); /* unbind */ iser_conn->iscsi_conn = NULL; conn->dd_data = NULL; + mutex_unlock(&unbind_iser_conn_mutex); complete(&iser_conn->stop_completion); mutex_unlock(&iser_conn->state_mutex); @@ -652,7 +655,7 @@ iscsi_iser_session_create(struct iscsi_endpoint *ep, } if (iscsi_host_add(shost, - ib_conn->device->ib_device->dma_device)) { + ib_conn->device->ib_device->dev.parent)) { mutex_unlock(&iser_conn->state_mutex); goto free_host; } @@ -977,13 +980,21 @@ static int iscsi_iser_slave_alloc(struct scsi_device *sdev) struct iser_conn *iser_conn; struct ib_device *ib_dev; + mutex_lock(&unbind_iser_conn_mutex); + session = starget_to_session(scsi_target(sdev))->dd_data; iser_conn = session->leadconn->dd_data; + if (!iser_conn) { + mutex_unlock(&unbind_iser_conn_mutex); + return -ENOTCONN; + } ib_dev = iser_conn->ib_conn.device->ib_device; if (!(ib_dev->attrs.device_cap_flags & IB_DEVICE_SG_GAPS_REG)) blk_queue_virt_boundary(sdev->request_queue, ~MASK_4K); + mutex_unlock(&unbind_iser_conn_mutex); + return 0; } @@ -994,6 +1005,7 @@ static struct scsi_host_template iscsi_iser_sht = { .change_queue_depth = scsi_change_queue_depth, .sg_tablesize = ISCSI_ISER_DEF_SG_TABLESIZE, .cmd_per_lun = ISER_DEF_CMD_PER_LUN, + .eh_timed_out = iscsi_eh_cmd_timed_out, .eh_abort_handler = iscsi_eh_abort, .eh_device_reset_handler= iscsi_eh_device_reset, .eh_target_reset_handler = iscsi_eh_recover_target, diff --git a/drivers/infiniband/ulp/iser/iscsi_iser.h b/drivers/infiniband/ulp/iser/iscsi_iser.h index 9d0b22ad58c1..c1ae4aeae2f9 100644 --- a/drivers/infiniband/ulp/iser/iscsi_iser.h +++ b/drivers/infiniband/ulp/iser/iscsi_iser.h @@ -430,6 +430,7 @@ struct iser_fr_desc { struct list_head list; struct iser_reg_resources rsc; struct iser_pi_context *pi_ctx; + struct list_head all_list; }; /** @@ -443,6 +444,7 @@ struct iser_fr_pool { struct list_head list; spinlock_t lock; int size; + struct list_head all_list; }; /** diff --git a/drivers/infiniband/ulp/iser/iser_initiator.c b/drivers/infiniband/ulp/iser/iser_initiator.c index 81ae2e30dd12..2a07692007bd 100644 --- a/drivers/infiniband/ulp/iser/iser_initiator.c +++ b/drivers/infiniband/ulp/iser/iser_initiator.c @@ -137,8 +137,10 @@ iser_prepare_write_cmd(struct iscsi_task *task, if (unsol_sz < edtl) { hdr->flags |= ISER_WSV; - hdr->write_stag = cpu_to_be32(mem_reg->rkey); - hdr->write_va = cpu_to_be64(mem_reg->sge.addr + unsol_sz); + if (buf_out->data_len > imm_sz) { + hdr->write_stag = cpu_to_be32(mem_reg->rkey); + hdr->write_va = cpu_to_be64(mem_reg->sge.addr + unsol_sz); + } iser_dbg("Cmd itt:%d, WRITE tags, RKEY:%#.4X " "VA:%#llX + unsol:%d\n", @@ -612,7 +614,7 @@ iser_check_remote_inv(struct iser_conn *iser_conn, iser_conn, rkey); if (unlikely(!iser_conn->snd_w_inv)) { - iser_err("conn %p: unexepected remote invalidation, " + iser_err("conn %p: unexpected remote invalidation, " "terminating connection\n", iser_conn); return -EPROTO; } diff --git a/drivers/infiniband/ulp/iser/iser_verbs.c b/drivers/infiniband/ulp/iser/iser_verbs.c index 6a9d1cb548ee..26a004e97ae0 100644 --- a/drivers/infiniband/ulp/iser/iser_verbs.c +++ b/drivers/infiniband/ulp/iser/iser_verbs.c @@ -362,6 +362,7 @@ int iser_alloc_fastreg_pool(struct ib_conn *ib_conn, int i, ret; INIT_LIST_HEAD(&fr_pool->list); + INIT_LIST_HEAD(&fr_pool->all_list); spin_lock_init(&fr_pool->lock); fr_pool->size = 0; for (i = 0; i < cmds_max; i++) { @@ -373,6 +374,7 @@ int iser_alloc_fastreg_pool(struct ib_conn *ib_conn, } list_add_tail(&desc->list, &fr_pool->list); + list_add_tail(&desc->all_list, &fr_pool->all_list); fr_pool->size++; } @@ -392,13 +394,13 @@ void iser_free_fastreg_pool(struct ib_conn *ib_conn) struct iser_fr_desc *desc, *tmp; int i = 0; - if (list_empty(&fr_pool->list)) + if (list_empty(&fr_pool->all_list)) return; iser_info("freeing conn %p fr pool\n", ib_conn); - list_for_each_entry_safe(desc, tmp, &fr_pool->list, list) { - list_del(&desc->list); + list_for_each_entry_safe(desc, tmp, &fr_pool->all_list, all_list) { + list_del(&desc->all_list); iser_free_reg_res(&desc->rsc); if (desc->pi_ctx) iser_free_pi_ctx(desc->pi_ctx); @@ -597,7 +599,9 @@ static void iser_free_ib_conn_res(struct iser_conn *iser_conn, iser_conn, ib_conn->cma_id, ib_conn->qp); if (ib_conn->qp != NULL) { + mutex_lock(&ig.connlist_mutex); ib_conn->comp->active_qps--; + mutex_unlock(&ig.connlist_mutex); rdma_destroy_qp(ib_conn->cma_id); ib_conn->qp = NULL; } @@ -704,8 +708,14 @@ iser_calc_scsi_params(struct iser_conn *iser_conn, unsigned short sg_tablesize, sup_sg_tablesize; sg_tablesize = DIV_ROUND_UP(max_sectors * 512, SIZE_4K); - sup_sg_tablesize = min_t(unsigned, ISCSI_ISER_MAX_SG_TABLESIZE, - device->ib_device->attrs.max_fast_reg_page_list_len); + if (device->ib_device->attrs.device_cap_flags & + IB_DEVICE_MEM_MGT_EXTENSIONS) + sup_sg_tablesize = + min_t( + uint, ISCSI_ISER_MAX_SG_TABLESIZE, + device->ib_device->attrs.max_fast_reg_page_list_len); + else + sup_sg_tablesize = ISCSI_ISER_MAX_SG_TABLESIZE; iser_conn->scsi_sg_tablesize = min(sg_tablesize, sup_sg_tablesize); } diff --git a/drivers/infiniband/ulp/isert/ib_isert.c b/drivers/infiniband/ulp/isert/ib_isert.c index 314e95516068..0e662656ef42 100644 --- a/drivers/infiniband/ulp/isert/ib_isert.c +++ b/drivers/infiniband/ulp/isert/ib_isert.c @@ -728,7 +728,7 @@ isert_disconnected_handler(struct rdma_cm_id *cma_id, iscsit_cause_connection_reinstatement(isert_conn->conn, 0); break; default: - isert_warn("conn %p teminating in state %d\n", + isert_warn("conn %p terminating in state %d\n", isert_conn, isert_conn->state); } mutex_unlock(&isert_conn->mutex); @@ -817,6 +817,7 @@ isert_post_recvm(struct isert_conn *isert_conn, u32 count) rx_wr->sg_list = &rx_desc->rx_sg; rx_wr->num_sge = 1; rx_wr->next = rx_wr + 1; + rx_desc->in_use = false; } rx_wr--; rx_wr->next = NULL; /* mark end of work requests list */ @@ -835,6 +836,15 @@ isert_post_recv(struct isert_conn *isert_conn, struct iser_rx_desc *rx_desc) struct ib_recv_wr *rx_wr_failed, rx_wr; int ret; + if (!rx_desc->in_use) { + /* + * if the descriptor is not in-use we already reposted it + * for recv, so just silently return + */ + return 0; + } + + rx_desc->in_use = false; rx_wr.wr_cqe = &rx_desc->rx_cqe; rx_wr.sg_list = &rx_desc->rx_sg; rx_wr.num_sge = 1; @@ -1397,6 +1407,8 @@ isert_recv_done(struct ib_cq *cq, struct ib_wc *wc) return; } + rx_desc->in_use = true; + ib_dma_sync_single_for_cpu(ib_dev, rx_desc->dma_addr, ISER_RX_PAYLOAD_SIZE, DMA_FROM_DEVICE); @@ -1440,7 +1452,7 @@ static void isert_login_recv_done(struct ib_cq *cq, struct ib_wc *wc) { struct isert_conn *isert_conn = wc->qp->qp_context; - struct ib_device *ib_dev = isert_conn->cm_id->device; + struct ib_device *ib_dev = isert_conn->device->ib_device; if (unlikely(wc->status != IB_WC_SUCCESS)) { isert_print_wc(wc, "login recv"); @@ -1659,10 +1671,23 @@ isert_rdma_write_done(struct ib_cq *cq, struct ib_wc *wc) ret = isert_check_pi_status(cmd, isert_cmd->rw.sig->sig_mr); isert_rdma_rw_ctx_destroy(isert_cmd, isert_conn); - if (ret) - transport_send_check_condition_and_sense(cmd, cmd->pi_err, 0); - else - isert_put_response(isert_conn->conn, isert_cmd->iscsi_cmd); + if (ret) { + /* + * transport_generic_request_failure() expects to have + * plus two references to handle queue-full, so re-add + * one here as target-core will have already dropped + * it after the first isert_put_datain() callback. + */ + kref_get(&cmd->cmd_kref); + transport_generic_request_failure(cmd, cmd->pi_err); + } else { + /* + * XXX: isert_put_response() failure is not retried. + */ + ret = isert_put_response(isert_conn->conn, isert_cmd->iscsi_cmd); + if (ret) + pr_warn_ratelimited("isert_put_response() ret: %d\n", ret); + } } static void @@ -1699,13 +1724,15 @@ isert_rdma_read_done(struct ib_cq *cq, struct ib_wc *wc) cmd->i_state = ISTATE_RECEIVED_LAST_DATAOUT; spin_unlock_bh(&cmd->istate_lock); - if (ret) { - target_put_sess_cmd(se_cmd); - transport_send_check_condition_and_sense(se_cmd, - se_cmd->pi_err, 0); - } else { + /* + * transport_generic_request_failure() will drop the extra + * se_cmd->cmd_kref reference after T10-PI error, and handle + * any non-zero ->queue_status() callback error retries. + */ + if (ret) + transport_generic_request_failure(se_cmd, se_cmd->pi_err); + else target_execute_cmd(se_cmd); - } } static void @@ -2171,26 +2198,28 @@ isert_put_datain(struct iscsi_conn *conn, struct iscsi_cmd *cmd) chain_wr = &isert_cmd->tx_desc.send_wr; } - isert_rdma_rw_ctx_post(isert_cmd, isert_conn, cqe, chain_wr); - isert_dbg("Cmd: %p posted RDMA_WRITE for iSER Data READ\n", isert_cmd); - return 1; + rc = isert_rdma_rw_ctx_post(isert_cmd, isert_conn, cqe, chain_wr); + isert_dbg("Cmd: %p posted RDMA_WRITE for iSER Data READ rc: %d\n", + isert_cmd, rc); + return rc; } static int isert_get_dataout(struct iscsi_conn *conn, struct iscsi_cmd *cmd, bool recovery) { struct isert_cmd *isert_cmd = iscsit_priv_cmd(cmd); + int ret; isert_dbg("Cmd: %p RDMA_READ data_length: %u write_data_done: %u\n", isert_cmd, cmd->se_cmd.data_length, cmd->write_data_done); isert_cmd->tx_desc.tx_cqe.done = isert_rdma_read_done; - isert_rdma_rw_ctx_post(isert_cmd, conn->context, - &isert_cmd->tx_desc.tx_cqe, NULL); + ret = isert_rdma_rw_ctx_post(isert_cmd, conn->context, + &isert_cmd->tx_desc.tx_cqe, NULL); - isert_dbg("Cmd: %p posted RDMA_READ memory for ISER Data WRITE\n", - isert_cmd); - return 0; + isert_dbg("Cmd: %p posted RDMA_READ memory for ISER Data WRITE rc: %d\n", + isert_cmd, ret); + return ret; } static int diff --git a/drivers/infiniband/ulp/isert/ib_isert.h b/drivers/infiniband/ulp/isert/ib_isert.h index c02ada57d7f5..87d994de8c91 100644 --- a/drivers/infiniband/ulp/isert/ib_isert.h +++ b/drivers/infiniband/ulp/isert/ib_isert.h @@ -60,7 +60,7 @@ #define ISER_RX_PAD_SIZE (ISCSI_DEF_MAX_RECV_SEG_LEN + 4096 - \ (ISER_RX_PAYLOAD_SIZE + sizeof(u64) + sizeof(struct ib_sge) + \ - sizeof(struct ib_cqe))) + sizeof(struct ib_cqe) + sizeof(bool))) #define ISCSI_ISER_SG_TABLESIZE 256 @@ -85,6 +85,7 @@ struct iser_rx_desc { u64 dma_addr; struct ib_sge rx_sg; struct ib_cqe rx_cqe; + bool in_use; char pad[ISER_RX_PAD_SIZE]; } __packed; diff --git a/drivers/infiniband/ulp/opa_vnic/Kconfig b/drivers/infiniband/ulp/opa_vnic/Kconfig new file mode 100644 index 000000000000..48132ab5e6b9 --- /dev/null +++ b/drivers/infiniband/ulp/opa_vnic/Kconfig @@ -0,0 +1,8 @@ +config INFINIBAND_OPA_VNIC + tristate "Intel OPA VNIC support" + depends on X86_64 && INFINIBAND + ---help--- + This is Omni-Path (OPA) Virtual Network Interface Controller (VNIC) + driver for Ethernet over Omni-Path feature. It implements the HW + independent VNIC functionality. It interfaces with Linux stack for + data path and IB MAD for the control path. diff --git a/drivers/infiniband/ulp/opa_vnic/Makefile b/drivers/infiniband/ulp/opa_vnic/Makefile new file mode 100644 index 000000000000..8061b287cfe4 --- /dev/null +++ b/drivers/infiniband/ulp/opa_vnic/Makefile @@ -0,0 +1,7 @@ +# Makefile - Intel Omni-Path Virtual Network Controller driver +# Copyright(c) 2017, Intel Corporation. +# +obj-$(CONFIG_INFINIBAND_OPA_VNIC) += opa_vnic.o + +opa_vnic-y := opa_vnic_netdev.o opa_vnic_encap.o opa_vnic_ethtool.o \ + opa_vnic_vema.o opa_vnic_vema_iface.o diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.c b/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.c new file mode 100644 index 000000000000..afa938bd26d6 --- /dev/null +++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.c @@ -0,0 +1,475 @@ +/* + * Copyright(c) 2017 Intel Corporation. + * + * This file is provided under a dual BSD/GPLv2 license. When using or + * redistributing this file, you may do so under either license. + * + * GPL LICENSE SUMMARY + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * BSD LICENSE + * + * 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. + * - Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + */ + +/* + * This file contains OPA VNIC encapsulation/decapsulation function. + */ + +#include <linux/if_ether.h> +#include <linux/if_vlan.h> + +#include "opa_vnic_internal.h" + +/* OPA 16B Header fields */ +#define OPA_16B_LID_MASK 0xFFFFFull +#define OPA_16B_SLID_HIGH_SHFT 8 +#define OPA_16B_SLID_MASK 0xF00ull +#define OPA_16B_DLID_MASK 0xF000ull +#define OPA_16B_DLID_HIGH_SHFT 12 +#define OPA_16B_LEN_SHFT 20 +#define OPA_16B_SC_SHFT 20 +#define OPA_16B_RC_SHFT 25 +#define OPA_16B_PKEY_SHFT 16 + +#define OPA_VNIC_L4_HDR_SHFT 16 + +/* L2+L4 hdr len is 20 bytes (5 quad words) */ +#define OPA_VNIC_HDR_QW_LEN 5 + +static inline void opa_vnic_make_header(u8 *hdr, u32 slid, u32 dlid, u16 len, + u16 pkey, u16 entropy, u8 sc, u8 rc, + u8 l4_type, u16 l4_hdr) +{ + /* h[1]: LT=1, 16B L2=10 */ + u32 h[OPA_VNIC_HDR_QW_LEN] = {0, 0xc0000000, 0, 0, 0}; + + h[2] = l4_type; + h[3] = entropy; + h[4] = l4_hdr << OPA_VNIC_L4_HDR_SHFT; + + /* Extract and set 4 upper bits and 20 lower bits of the lids */ + h[0] |= (slid & OPA_16B_LID_MASK); + h[2] |= ((slid >> (20 - OPA_16B_SLID_HIGH_SHFT)) & OPA_16B_SLID_MASK); + + h[1] |= (dlid & OPA_16B_LID_MASK); + h[2] |= ((dlid >> (20 - OPA_16B_DLID_HIGH_SHFT)) & OPA_16B_DLID_MASK); + + h[0] |= (len << OPA_16B_LEN_SHFT); + h[1] |= (rc << OPA_16B_RC_SHFT); + h[1] |= (sc << OPA_16B_SC_SHFT); + h[2] |= ((u32)pkey << OPA_16B_PKEY_SHFT); + + memcpy(hdr, h, OPA_VNIC_HDR_LEN); +} + +/* + * Using a simple hash table for mac table implementation with the last octet + * of mac address as a key. + */ +static void opa_vnic_free_mac_tbl(struct hlist_head *mactbl) +{ + struct opa_vnic_mac_tbl_node *node; + struct hlist_node *tmp; + int bkt; + + if (!mactbl) + return; + + vnic_hash_for_each_safe(mactbl, bkt, tmp, node, hlist) { + hash_del(&node->hlist); + kfree(node); + } + kfree(mactbl); +} + +static struct hlist_head *opa_vnic_alloc_mac_tbl(void) +{ + u32 size = sizeof(struct hlist_head) * OPA_VNIC_MAC_TBL_SIZE; + struct hlist_head *mactbl; + + mactbl = kzalloc(size, GFP_KERNEL); + if (!mactbl) + return ERR_PTR(-ENOMEM); + + vnic_hash_init(mactbl); + return mactbl; +} + +/* opa_vnic_release_mac_tbl - empty and free the mac table */ +void opa_vnic_release_mac_tbl(struct opa_vnic_adapter *adapter) +{ + struct hlist_head *mactbl; + + mutex_lock(&adapter->mactbl_lock); + mactbl = rcu_access_pointer(adapter->mactbl); + rcu_assign_pointer(adapter->mactbl, NULL); + synchronize_rcu(); + opa_vnic_free_mac_tbl(mactbl); + mutex_unlock(&adapter->mactbl_lock); +} + +/* + * opa_vnic_query_mac_tbl - query the mac table for a section + * + * This function implements query of specific function of the mac table. + * The function also expects the requested range to be valid. + */ +void opa_vnic_query_mac_tbl(struct opa_vnic_adapter *adapter, + struct opa_veswport_mactable *tbl) +{ + struct opa_vnic_mac_tbl_node *node; + struct hlist_head *mactbl; + int bkt; + u16 loffset, lnum_entries; + + rcu_read_lock(); + mactbl = rcu_dereference(adapter->mactbl); + if (!mactbl) + goto get_mac_done; + + loffset = be16_to_cpu(tbl->offset); + lnum_entries = be16_to_cpu(tbl->num_entries); + + vnic_hash_for_each(mactbl, bkt, node, hlist) { + struct __opa_vnic_mactable_entry *nentry = &node->entry; + struct opa_veswport_mactable_entry *entry; + + if ((node->index < loffset) || + (node->index >= (loffset + lnum_entries))) + continue; + + /* populate entry in the tbl corresponding to the index */ + entry = &tbl->tbl_entries[node->index - loffset]; + memcpy(entry->mac_addr, nentry->mac_addr, + ARRAY_SIZE(entry->mac_addr)); + memcpy(entry->mac_addr_mask, nentry->mac_addr_mask, + ARRAY_SIZE(entry->mac_addr_mask)); + entry->dlid_sd = cpu_to_be32(nentry->dlid_sd); + } + tbl->mac_tbl_digest = cpu_to_be32(adapter->info.vport.mac_tbl_digest); +get_mac_done: + rcu_read_unlock(); +} + +/* + * opa_vnic_update_mac_tbl - update mac table section + * + * This function updates the specified section of the mac table. + * The procedure includes following steps. + * - Allocate a new mac (hash) table. + * - Add the specified entries to the new table. + * (except the ones that are requested to be deleted). + * - Add all the other entries from the old mac table. + * - If there is a failure, free the new table and return. + * - Switch to the new table. + * - Free the old table and return. + * + * The function also expects the requested range to be valid. + */ +int opa_vnic_update_mac_tbl(struct opa_vnic_adapter *adapter, + struct opa_veswport_mactable *tbl) +{ + struct opa_vnic_mac_tbl_node *node, *new_node; + struct hlist_head *new_mactbl, *old_mactbl; + int i, bkt, rc = 0; + u8 key; + u16 loffset, lnum_entries; + + mutex_lock(&adapter->mactbl_lock); + /* allocate new mac table */ + new_mactbl = opa_vnic_alloc_mac_tbl(); + if (IS_ERR(new_mactbl)) { + mutex_unlock(&adapter->mactbl_lock); + return PTR_ERR(new_mactbl); + } + + loffset = be16_to_cpu(tbl->offset); + lnum_entries = be16_to_cpu(tbl->num_entries); + + /* add updated entries to the new mac table */ + for (i = 0; i < lnum_entries; i++) { + struct __opa_vnic_mactable_entry *nentry; + struct opa_veswport_mactable_entry *entry = + &tbl->tbl_entries[i]; + u8 *mac_addr = entry->mac_addr; + u8 empty_mac[ETH_ALEN] = { 0 }; + + v_dbg("new mac entry %4d: %02x:%02x:%02x:%02x:%02x:%02x %x\n", + loffset + i, mac_addr[0], mac_addr[1], mac_addr[2], + mac_addr[3], mac_addr[4], mac_addr[5], + entry->dlid_sd); + + /* if the entry is being removed, do not add it */ + if (!memcmp(mac_addr, empty_mac, ARRAY_SIZE(empty_mac))) + continue; + + node = kzalloc(sizeof(*node), GFP_KERNEL); + if (!node) { + rc = -ENOMEM; + goto updt_done; + } + + node->index = loffset + i; + nentry = &node->entry; + memcpy(nentry->mac_addr, entry->mac_addr, + ARRAY_SIZE(nentry->mac_addr)); + memcpy(nentry->mac_addr_mask, entry->mac_addr_mask, + ARRAY_SIZE(nentry->mac_addr_mask)); + nentry->dlid_sd = be32_to_cpu(entry->dlid_sd); + key = node->entry.mac_addr[OPA_VNIC_MAC_HASH_IDX]; + vnic_hash_add(new_mactbl, &node->hlist, key); + } + + /* add other entries from current mac table to new mac table */ + old_mactbl = rcu_access_pointer(adapter->mactbl); + if (!old_mactbl) + goto switch_tbl; + + vnic_hash_for_each(old_mactbl, bkt, node, hlist) { + if ((node->index >= loffset) && + (node->index < (loffset + lnum_entries))) + continue; + + new_node = kzalloc(sizeof(*new_node), GFP_KERNEL); + if (!new_node) { + rc = -ENOMEM; + goto updt_done; + } + + new_node->index = node->index; + memcpy(&new_node->entry, &node->entry, sizeof(node->entry)); + key = new_node->entry.mac_addr[OPA_VNIC_MAC_HASH_IDX]; + vnic_hash_add(new_mactbl, &new_node->hlist, key); + } + +switch_tbl: + /* switch to new table */ + rcu_assign_pointer(adapter->mactbl, new_mactbl); + synchronize_rcu(); + + adapter->info.vport.mac_tbl_digest = be32_to_cpu(tbl->mac_tbl_digest); +updt_done: + /* upon failure, free the new table; otherwise, free the old table */ + if (rc) + opa_vnic_free_mac_tbl(new_mactbl); + else + opa_vnic_free_mac_tbl(old_mactbl); + + mutex_unlock(&adapter->mactbl_lock); + return rc; +} + +/* opa_vnic_chk_mac_tbl - check mac table for dlid */ +static uint32_t opa_vnic_chk_mac_tbl(struct opa_vnic_adapter *adapter, + struct ethhdr *mac_hdr) +{ + struct opa_vnic_mac_tbl_node *node; + struct hlist_head *mactbl; + u32 dlid = 0; + u8 key; + + rcu_read_lock(); + mactbl = rcu_dereference(adapter->mactbl); + if (unlikely(!mactbl)) + goto chk_done; + + key = mac_hdr->h_dest[OPA_VNIC_MAC_HASH_IDX]; + vnic_hash_for_each_possible(mactbl, node, hlist, key) { + struct __opa_vnic_mactable_entry *entry = &node->entry; + + /* if related to source mac, skip */ + if (unlikely(OPA_VNIC_DLID_SD_IS_SRC_MAC(entry->dlid_sd))) + continue; + + if (!memcmp(node->entry.mac_addr, mac_hdr->h_dest, + ARRAY_SIZE(node->entry.mac_addr))) { + /* mac address found */ + dlid = OPA_VNIC_DLID_SD_GET_DLID(node->entry.dlid_sd); + break; + } + } + +chk_done: + rcu_read_unlock(); + return dlid; +} + +/* opa_vnic_get_dlid - find and return the DLID */ +static uint32_t opa_vnic_get_dlid(struct opa_vnic_adapter *adapter, + struct sk_buff *skb, u8 def_port) +{ + struct __opa_veswport_info *info = &adapter->info; + struct ethhdr *mac_hdr = (struct ethhdr *)skb_mac_header(skb); + u32 dlid; + + dlid = opa_vnic_chk_mac_tbl(adapter, mac_hdr); + if (dlid) + return dlid; + + if (is_multicast_ether_addr(mac_hdr->h_dest)) { + dlid = info->vesw.u_mcast_dlid; + } else { + if (is_local_ether_addr(mac_hdr->h_dest)) { + dlid = ((uint32_t)mac_hdr->h_dest[5] << 16) | + ((uint32_t)mac_hdr->h_dest[4] << 8) | + mac_hdr->h_dest[3]; + if (unlikely(!dlid)) + v_warn("Null dlid in MAC address\n"); + } else if (def_port != OPA_VNIC_INVALID_PORT) { + dlid = info->vesw.u_ucast_dlid[def_port]; + } + } + + return dlid; +} + +/* opa_vnic_get_sc - return the service class */ +static u8 opa_vnic_get_sc(struct __opa_veswport_info *info, + struct sk_buff *skb) +{ + struct ethhdr *mac_hdr = (struct ethhdr *)skb_mac_header(skb); + u16 vlan_tci; + u8 sc; + + if (!__vlan_get_tag(skb, &vlan_tci)) { + u8 pcp = OPA_VNIC_VLAN_PCP(vlan_tci); + + if (is_multicast_ether_addr(mac_hdr->h_dest)) + sc = info->vport.pcp_to_sc_mc[pcp]; + else + sc = info->vport.pcp_to_sc_uc[pcp]; + } else { + if (is_multicast_ether_addr(mac_hdr->h_dest)) + sc = info->vport.non_vlan_sc_mc; + else + sc = info->vport.non_vlan_sc_uc; + } + + return sc; +} + +u8 opa_vnic_get_vl(struct opa_vnic_adapter *adapter, struct sk_buff *skb) +{ + struct ethhdr *mac_hdr = (struct ethhdr *)skb_mac_header(skb); + struct __opa_veswport_info *info = &adapter->info; + u8 vl; + + if (skb_vlan_tag_present(skb)) { + u8 pcp = skb_vlan_tag_get(skb) >> VLAN_PRIO_SHIFT; + + if (is_multicast_ether_addr(mac_hdr->h_dest)) + vl = info->vport.pcp_to_vl_mc[pcp]; + else + vl = info->vport.pcp_to_vl_uc[pcp]; + } else { + if (is_multicast_ether_addr(mac_hdr->h_dest)) + vl = info->vport.non_vlan_vl_mc; + else + vl = info->vport.non_vlan_vl_uc; + } + + return vl; +} + +/* opa_vnic_calc_entropy - calculate the packet entropy */ +u8 opa_vnic_calc_entropy(struct opa_vnic_adapter *adapter, struct sk_buff *skb) +{ + u16 hash16; + + /* + * Get flow based 16-bit hash and then XOR the upper and lower bytes + * to get the entropy. + * __skb_tx_hash limits qcount to 16 bits. Hence, get 15-bit hash. + */ + hash16 = __skb_tx_hash(adapter->netdev, skb, BIT(15)); + return (u8)((hash16 >> 8) ^ (hash16 & 0xff)); +} + +/* opa_vnic_get_def_port - get default port based on entropy */ +static inline u8 opa_vnic_get_def_port(struct opa_vnic_adapter *adapter, + u8 entropy) +{ + u8 flow_id; + + /* Add the upper and lower 4-bits of entropy to get the flow id */ + flow_id = ((entropy & 0xf) + (entropy >> 4)); + return adapter->flow_tbl[flow_id & (OPA_VNIC_FLOW_TBL_SIZE - 1)]; +} + +/* Calculate packet length including OPA header, crc and padding */ +static inline int opa_vnic_wire_length(struct sk_buff *skb) +{ + u32 pad_len; + + /* padding for 8 bytes size alignment */ + pad_len = -(skb->len + OPA_VNIC_ICRC_TAIL_LEN) & 0x7; + pad_len += OPA_VNIC_ICRC_TAIL_LEN; + + return (skb->len + pad_len) >> 3; +} + +/* opa_vnic_encap_skb - encapsulate skb packet with OPA header and meta data */ +void opa_vnic_encap_skb(struct opa_vnic_adapter *adapter, struct sk_buff *skb) +{ + struct __opa_veswport_info *info = &adapter->info; + struct opa_vnic_skb_mdata *mdata; + u8 def_port, sc, entropy, *hdr; + u16 len, l4_hdr; + u32 dlid; + + hdr = skb_push(skb, OPA_VNIC_HDR_LEN); + + entropy = opa_vnic_calc_entropy(adapter, skb); + def_port = opa_vnic_get_def_port(adapter, entropy); + len = opa_vnic_wire_length(skb); + dlid = opa_vnic_get_dlid(adapter, skb, def_port); + sc = opa_vnic_get_sc(info, skb); + l4_hdr = info->vesw.vesw_id; + + mdata = skb_push(skb, sizeof(*mdata)); + mdata->vl = opa_vnic_get_vl(adapter, skb); + mdata->entropy = entropy; + mdata->flags = 0; + if (unlikely(!dlid)) { + mdata->flags = OPA_VNIC_SKB_MDATA_ENCAP_ERR; + return; + } + + opa_vnic_make_header(hdr, info->vport.encap_slid, dlid, len, + info->vesw.pkey, entropy, sc, 0, + OPA_VNIC_L4_ETHR, l4_hdr); +} diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.h b/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.h new file mode 100644 index 000000000000..4c434b9dd84c --- /dev/null +++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.h @@ -0,0 +1,489 @@ +#ifndef _OPA_VNIC_ENCAP_H +#define _OPA_VNIC_ENCAP_H +/* + * Copyright(c) 2017 Intel Corporation. + * + * This file is provided under a dual BSD/GPLv2 license. When using or + * redistributing this file, you may do so under either license. + * + * GPL LICENSE SUMMARY + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * BSD LICENSE + * + * 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. + * - Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + */ + +/* + * This file contains all OPA VNIC declaration required for encapsulation + * and decapsulation of Ethernet packets + */ + +#include <linux/types.h> +#include <rdma/ib_mad.h> + +/* EMA class version */ +#define OPA_EMA_CLASS_VERSION 0x80 + +/* + * Define the Intel vendor management class for OPA + * ETHERNET MANAGEMENT + */ +#define OPA_MGMT_CLASS_INTEL_EMA 0x34 + +/* EM attribute IDs */ +#define OPA_EM_ATTR_CLASS_PORT_INFO 0x0001 +#define OPA_EM_ATTR_VESWPORT_INFO 0x0011 +#define OPA_EM_ATTR_VESWPORT_MAC_ENTRIES 0x0012 +#define OPA_EM_ATTR_IFACE_UCAST_MACS 0x0013 +#define OPA_EM_ATTR_IFACE_MCAST_MACS 0x0014 +#define OPA_EM_ATTR_DELETE_VESW 0x0015 +#define OPA_EM_ATTR_VESWPORT_SUMMARY_COUNTERS 0x0020 +#define OPA_EM_ATTR_VESWPORT_ERROR_COUNTERS 0x0022 + +/* VNIC configured and operational state values */ +#define OPA_VNIC_STATE_DROP_ALL 0x1 +#define OPA_VNIC_STATE_FORWARDING 0x3 + +#define OPA_VESW_MAX_NUM_DEF_PORT 16 +#define OPA_VNIC_MAX_NUM_PCP 8 + +#define OPA_VNIC_EMA_DATA (OPA_MGMT_MAD_SIZE - IB_MGMT_VENDOR_HDR) + +/* Defines for vendor specific notice(trap) attributes */ +#define OPA_INTEL_EMA_NOTICE_TYPE_INFO 0x04 + +/* INTEL OUI */ +#define INTEL_OUI_1 0x00 +#define INTEL_OUI_2 0x06 +#define INTEL_OUI_3 0x6a + +/* Trap opcodes sent from VNIC */ +#define OPA_VESWPORT_TRAP_IFACE_UCAST_MAC_CHANGE 0x1 +#define OPA_VESWPORT_TRAP_IFACE_MCAST_MAC_CHANGE 0x2 +#define OPA_VESWPORT_TRAP_ETH_LINK_STATUS_CHANGE 0x3 + +#define OPA_VNIC_DLID_SD_IS_SRC_MAC(dlid_sd) (!!((dlid_sd) & 0x20)) +#define OPA_VNIC_DLID_SD_GET_DLID(dlid_sd) ((dlid_sd) >> 8) + +/* VNIC Ethernet link status */ +#define OPA_VNIC_ETH_LINK_UP 1 +#define OPA_VNIC_ETH_LINK_DOWN 2 + +/** + * struct opa_vesw_info - OPA vnic switch information + * @fabric_id: 10-bit fabric id + * @vesw_id: 12-bit virtual ethernet switch id + * @def_port_mask: bitmask of default ports + * @pkey: partition key + * @u_mcast_dlid: unknown multicast dlid + * @u_ucast_dlid: array of unknown unicast dlids + * @eth_mtu: MTUs for each vlan PCP + * @eth_mtu_non_vlan: MTU for non vlan packets + */ +struct opa_vesw_info { + __be16 fabric_id; + __be16 vesw_id; + + u8 rsvd0[6]; + __be16 def_port_mask; + + u8 rsvd1[2]; + __be16 pkey; + + u8 rsvd2[4]; + __be32 u_mcast_dlid; + __be32 u_ucast_dlid[OPA_VESW_MAX_NUM_DEF_PORT]; + + u8 rsvd3[44]; + __be16 eth_mtu[OPA_VNIC_MAX_NUM_PCP]; + __be16 eth_mtu_non_vlan; + u8 rsvd4[2]; +} __packed; + +/** + * struct opa_per_veswport_info - OPA vnic per port information + * @port_num: port number + * @eth_link_status: current ethernet link state + * @base_mac_addr: base mac address + * @config_state: configured port state + * @oper_state: operational port state + * @max_mac_tbl_ent: max number of mac table entries + * @max_smac_ent: max smac entries in mac table + * @mac_tbl_digest: mac table digest + * @encap_slid: base slid for the port + * @pcp_to_sc_uc: sc by pcp index for unicast ethernet packets + * @pcp_to_vl_uc: vl by pcp index for unicast ethernet packets + * @pcp_to_sc_mc: sc by pcp index for multicast ethernet packets + * @pcp_to_vl_mc: vl by pcp index for multicast ethernet packets + * @non_vlan_sc_uc: sc for non-vlan unicast ethernet packets + * @non_vlan_vl_uc: vl for non-vlan unicast ethernet packets + * @non_vlan_sc_mc: sc for non-vlan multicast ethernet packets + * @non_vlan_vl_mc: vl for non-vlan multicast ethernet packets + * @uc_macs_gen_count: generation count for unicast macs list + * @mc_macs_gen_count: generation count for multicast macs list + */ +struct opa_per_veswport_info { + __be32 port_num; + + u8 eth_link_status; + u8 rsvd0[3]; + + u8 base_mac_addr[ETH_ALEN]; + u8 config_state; + u8 oper_state; + + __be16 max_mac_tbl_ent; + __be16 max_smac_ent; + __be32 mac_tbl_digest; + u8 rsvd1[4]; + + __be32 encap_slid; + + u8 pcp_to_sc_uc[OPA_VNIC_MAX_NUM_PCP]; + u8 pcp_to_vl_uc[OPA_VNIC_MAX_NUM_PCP]; + u8 pcp_to_sc_mc[OPA_VNIC_MAX_NUM_PCP]; + u8 pcp_to_vl_mc[OPA_VNIC_MAX_NUM_PCP]; + + u8 non_vlan_sc_uc; + u8 non_vlan_vl_uc; + u8 non_vlan_sc_mc; + u8 non_vlan_vl_mc; + + u8 rsvd2[48]; + + __be16 uc_macs_gen_count; + __be16 mc_macs_gen_count; + + u8 rsvd3[8]; +} __packed; + +/** + * struct opa_veswport_info - OPA vnic port information + * @vesw: OPA vnic switch information + * @vport: OPA vnic per port information + * + * On host, each of the virtual ethernet ports belongs + * to a different virtual ethernet switches. + */ +struct opa_veswport_info { + struct opa_vesw_info vesw; + struct opa_per_veswport_info vport; +}; + +/** + * struct opa_veswport_mactable_entry - single entry in the forwarding table + * @mac_addr: MAC address + * @mac_addr_mask: MAC address bit mask + * @dlid_sd: Matching DLID and side data + * + * On the host each virtual ethernet port will have + * a forwarding table. These tables are used to + * map a MAC to a LID and other data. For more + * details see struct opa_veswport_mactable_entries. + * This is the structure of a single mactable entry + */ +struct opa_veswport_mactable_entry { + u8 mac_addr[ETH_ALEN]; + u8 mac_addr_mask[ETH_ALEN]; + __be32 dlid_sd; +} __packed; + +/** + * struct opa_veswport_mactable - Forwarding table array + * @offset: mac table starting offset + * @num_entries: Number of entries to get or set + * @mac_tbl_digest: mac table digest + * @tbl_entries[]: Array of table entries + * + * The EM sends down this structure in a MAD indicating + * the starting offset in the forwarding table that this + * entry is to be loaded into and the number of entries + * that that this MAD instance contains + * The mac_tbl_digest has been added to this MAD structure. It will be set by + * the EM and it will be used by the EM to check if there are any + * discrepancies with this value and the value + * maintained by the EM in the case of VNIC port being deleted or unloaded + * A new instantiation of a VNIC will always have a value of zero. + * This value is stored as part of the vnic adapter structure and will be + * accessed by the GET and SET routines for both the mactable entries and the + * veswport info. + */ +struct opa_veswport_mactable { + __be16 offset; + __be16 num_entries; + __be32 mac_tbl_digest; + struct opa_veswport_mactable_entry tbl_entries[0]; +} __packed; + +/** + * struct opa_veswport_summary_counters - summary counters + * @vp_instance: vport instance on the OPA port + * @vesw_id: virtual ethernet switch id + * @veswport_num: virtual ethernet switch port number + * @tx_errors: transmit errors + * @rx_errors: receive errors + * @tx_packets: transmit packets + * @rx_packets: receive packets + * @tx_bytes: transmit bytes + * @rx_bytes: receive bytes + * @tx_unicast: unicast packets transmitted + * @tx_mcastbcast: multicast/broadcast packets transmitted + * @tx_untagged: non-vlan packets transmitted + * @tx_vlan: vlan packets transmitted + * @tx_64_size: transmit packet length is 64 bytes + * @tx_65_127: transmit packet length is >=65 and < 127 bytes + * @tx_128_255: transmit packet length is >=128 and < 255 bytes + * @tx_256_511: transmit packet length is >=256 and < 511 bytes + * @tx_512_1023: transmit packet length is >=512 and < 1023 bytes + * @tx_1024_1518: transmit packet length is >=1024 and < 1518 bytes + * @tx_1519_max: transmit packet length >= 1519 bytes + * @rx_unicast: unicast packets received + * @rx_mcastbcast: multicast/broadcast packets received + * @rx_untagged: non-vlan packets received + * @rx_vlan: vlan packets received + * @rx_64_size: received packet length is 64 bytes + * @rx_65_127: received packet length is >=65 and < 127 bytes + * @rx_128_255: received packet length is >=128 and < 255 bytes + * @rx_256_511: received packet length is >=256 and < 511 bytes + * @rx_512_1023: received packet length is >=512 and < 1023 bytes + * @rx_1024_1518: received packet length is >=1024 and < 1518 bytes + * @rx_1519_max: received packet length >= 1519 bytes + * + * All the above are counters of corresponding conditions. + */ +struct opa_veswport_summary_counters { + __be16 vp_instance; + __be16 vesw_id; + __be32 veswport_num; + + __be64 tx_errors; + __be64 rx_errors; + __be64 tx_packets; + __be64 rx_packets; + __be64 tx_bytes; + __be64 rx_bytes; + + __be64 tx_unicast; + __be64 tx_mcastbcast; + + __be64 tx_untagged; + __be64 tx_vlan; + + __be64 tx_64_size; + __be64 tx_65_127; + __be64 tx_128_255; + __be64 tx_256_511; + __be64 tx_512_1023; + __be64 tx_1024_1518; + __be64 tx_1519_max; + + __be64 rx_unicast; + __be64 rx_mcastbcast; + + __be64 rx_untagged; + __be64 rx_vlan; + + __be64 rx_64_size; + __be64 rx_65_127; + __be64 rx_128_255; + __be64 rx_256_511; + __be64 rx_512_1023; + __be64 rx_1024_1518; + __be64 rx_1519_max; + + __be64 reserved[16]; +} __packed; + +/** + * struct opa_veswport_error_counters - error counters + * @vp_instance: vport instance on the OPA port + * @vesw_id: virtual ethernet switch id + * @veswport_num: virtual ethernet switch port number + * @tx_errors: transmit errors + * @rx_errors: receive errors + * @tx_smac_filt: smac filter errors + * @tx_dlid_zero: transmit packets with invalid dlid + * @tx_logic: other transmit errors + * @tx_drop_state: packet tansmission in non-forward port state + * @rx_bad_veswid: received packet with invalid vesw id + * @rx_runt: received ethernet packet with length < 64 bytes + * @rx_oversize: received ethernet packet with length > MTU size + * @rx_eth_down: received packets when interface is down + * @rx_drop_state: received packets in non-forwarding port state + * @rx_logic: other receive errors + * + * All the above are counters of corresponding erorr conditions. + */ +struct opa_veswport_error_counters { + __be16 vp_instance; + __be16 vesw_id; + __be32 veswport_num; + + __be64 tx_errors; + __be64 rx_errors; + + __be64 rsvd0; + __be64 tx_smac_filt; + __be64 rsvd1; + __be64 rsvd2; + __be64 rsvd3; + __be64 tx_dlid_zero; + __be64 rsvd4; + __be64 tx_logic; + __be64 rsvd5; + __be64 tx_drop_state; + + __be64 rx_bad_veswid; + __be64 rsvd6; + __be64 rx_runt; + __be64 rx_oversize; + __be64 rsvd7; + __be64 rx_eth_down; + __be64 rx_drop_state; + __be64 rx_logic; + __be64 rsvd8; + + __be64 rsvd9[16]; +} __packed; + +/** + * struct opa_veswport_trap - Trap message sent to EM by VNIC + * @fabric_id: 10 bit fabric id + * @veswid: 12 bit virtual ethernet switch id + * @veswportnum: logical port number on the Virtual switch + * @opaportnum: physical port num (redundant on host) + * @veswportindex: switch port index on opa port 0 based + * @opcode: operation + * @reserved: 32 bit for alignment + * + * The VNIC will send trap messages to the Ethernet manager to + * inform it about changes to the VNIC config, behaviour etc. + * This is the format of the trap payload. + */ +struct opa_veswport_trap { + __be16 fabric_id; + __be16 veswid; + __be32 veswportnum; + __be16 opaportnum; + u8 veswportindex; + u8 opcode; + __be32 reserved; +} __packed; + +/** + * struct opa_vnic_iface_macs_entry - single entry in the mac list + * @mac_addr: MAC address + */ +struct opa_vnic_iface_mac_entry { + u8 mac_addr[ETH_ALEN]; +}; + +/** + * struct opa_veswport_iface_macs - Msg to set globally administered MAC + * @start_idx: position of first entry (0 based) + * @num_macs_in_msg: number of MACs in this message + * @tot_macs_in_lst: The total number of MACs the agent has + * @gen_count: gen_count to indicate change + * @entry: The mac list entry + * + * Same attribute IDS and attribute modifiers as in locally administered + * addresses used to set globally administered addresses + */ +struct opa_veswport_iface_macs { + __be16 start_idx; + __be16 num_macs_in_msg; + __be16 tot_macs_in_lst; + __be16 gen_count; + struct opa_vnic_iface_mac_entry entry[0]; +} __packed; + +/** + * struct opa_vnic_vema_mad - Generic VEMA MAD + * @mad_hdr: Generic MAD header + * @rmpp_hdr: RMPP header for vendor specific MADs + * @oui: Unique org identifier + * @data: MAD data + */ +struct opa_vnic_vema_mad { + struct ib_mad_hdr mad_hdr; + struct ib_rmpp_hdr rmpp_hdr; + u8 reserved; + u8 oui[3]; + u8 data[OPA_VNIC_EMA_DATA]; +}; + +/** + * struct opa_vnic_notice_attr - Generic Notice MAD + * @gen_type: Generic/Specific bit and type of notice + * @oui_1: Vendor ID byte 1 + * @oui_2: Vendor ID byte 2 + * @oui_3: Vendor ID byte 3 + * @trap_num: Trap number + * @toggle_count: Notice toggle bit and count value + * @issuer_lid: Trap issuer's lid + * @issuer_gid: Issuer GID (only if Report method) + * @raw_data: Trap message body + */ +struct opa_vnic_notice_attr { + u8 gen_type; + u8 oui_1; + u8 oui_2; + u8 oui_3; + __be16 trap_num; + __be16 toggle_count; + __be32 issuer_lid; + __be32 reserved; + u8 issuer_gid[16]; + u8 raw_data[64]; +} __packed; + +/** + * struct opa_vnic_vema_mad_trap - Generic VEMA MAD Trap + * @mad_hdr: Generic MAD header + * @rmpp_hdr: RMPP header for vendor specific MADs + * @oui: Unique org identifier + * @notice: Notice structure + */ +struct opa_vnic_vema_mad_trap { + struct ib_mad_hdr mad_hdr; + struct ib_rmpp_hdr rmpp_hdr; + u8 reserved; + u8 oui[3]; + struct opa_vnic_notice_attr notice; +}; + +#endif /* _OPA_VNIC_ENCAP_H */ diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_ethtool.c b/drivers/infiniband/ulp/opa_vnic/opa_vnic_ethtool.c new file mode 100644 index 000000000000..62390e9e0023 --- /dev/null +++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_ethtool.c @@ -0,0 +1,187 @@ +/* + * Copyright(c) 2017 Intel Corporation. + * + * This file is provided under a dual BSD/GPLv2 license. When using or + * redistributing this file, you may do so under either license. + * + * GPL LICENSE SUMMARY + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * BSD LICENSE + * + * 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. + * - Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + */ + +/* + * This file contains OPA VNIC ethtool functions + */ + +#include <linux/ethtool.h> + +#include "opa_vnic_internal.h" + +enum {NETDEV_STATS, VNIC_STATS}; + +struct vnic_stats { + char stat_string[ETH_GSTRING_LEN]; + struct { + int sizeof_stat; + int stat_offset; + }; +}; + +#define VNIC_STAT(m) { FIELD_SIZEOF(struct opa_vnic_stats, m), \ + offsetof(struct opa_vnic_stats, m) } + +static struct vnic_stats vnic_gstrings_stats[] = { + /* NETDEV stats */ + {"rx_packets", VNIC_STAT(netstats.rx_packets)}, + {"tx_packets", VNIC_STAT(netstats.tx_packets)}, + {"rx_bytes", VNIC_STAT(netstats.rx_bytes)}, + {"tx_bytes", VNIC_STAT(netstats.tx_bytes)}, + {"rx_errors", VNIC_STAT(netstats.rx_errors)}, + {"tx_errors", VNIC_STAT(netstats.tx_errors)}, + {"rx_dropped", VNIC_STAT(netstats.rx_dropped)}, + {"tx_dropped", VNIC_STAT(netstats.tx_dropped)}, + + /* SUMMARY counters */ + {"tx_unicast", VNIC_STAT(tx_grp.unicast)}, + {"tx_mcastbcast", VNIC_STAT(tx_grp.mcastbcast)}, + {"tx_untagged", VNIC_STAT(tx_grp.untagged)}, + {"tx_vlan", VNIC_STAT(tx_grp.vlan)}, + + {"tx_64_size", VNIC_STAT(tx_grp.s_64)}, + {"tx_65_127", VNIC_STAT(tx_grp.s_65_127)}, + {"tx_128_255", VNIC_STAT(tx_grp.s_128_255)}, + {"tx_256_511", VNIC_STAT(tx_grp.s_256_511)}, + {"tx_512_1023", VNIC_STAT(tx_grp.s_512_1023)}, + {"tx_1024_1518", VNIC_STAT(tx_grp.s_1024_1518)}, + {"tx_1519_max", VNIC_STAT(tx_grp.s_1519_max)}, + + {"rx_unicast", VNIC_STAT(rx_grp.unicast)}, + {"rx_mcastbcast", VNIC_STAT(rx_grp.mcastbcast)}, + {"rx_untagged", VNIC_STAT(rx_grp.untagged)}, + {"rx_vlan", VNIC_STAT(rx_grp.vlan)}, + + {"rx_64_size", VNIC_STAT(rx_grp.s_64)}, + {"rx_65_127", VNIC_STAT(rx_grp.s_65_127)}, + {"rx_128_255", VNIC_STAT(rx_grp.s_128_255)}, + {"rx_256_511", VNIC_STAT(rx_grp.s_256_511)}, + {"rx_512_1023", VNIC_STAT(rx_grp.s_512_1023)}, + {"rx_1024_1518", VNIC_STAT(rx_grp.s_1024_1518)}, + {"rx_1519_max", VNIC_STAT(rx_grp.s_1519_max)}, + + /* ERROR counters */ + {"rx_fifo_errors", VNIC_STAT(netstats.rx_fifo_errors)}, + {"rx_length_errors", VNIC_STAT(netstats.rx_length_errors)}, + + {"tx_fifo_errors", VNIC_STAT(netstats.tx_fifo_errors)}, + {"tx_carrier_errors", VNIC_STAT(netstats.tx_carrier_errors)}, + + {"tx_dlid_zero", VNIC_STAT(tx_dlid_zero)}, + {"tx_drop_state", VNIC_STAT(tx_drop_state)}, + {"rx_drop_state", VNIC_STAT(rx_drop_state)}, + {"rx_oversize", VNIC_STAT(rx_oversize)}, + {"rx_runt", VNIC_STAT(rx_runt)}, +}; + +#define VNIC_STATS_LEN ARRAY_SIZE(vnic_gstrings_stats) + +/* vnic_get_drvinfo - get driver info */ +static void vnic_get_drvinfo(struct net_device *netdev, + struct ethtool_drvinfo *drvinfo) +{ + strlcpy(drvinfo->driver, opa_vnic_driver_name, sizeof(drvinfo->driver)); + strlcpy(drvinfo->version, opa_vnic_driver_version, + sizeof(drvinfo->version)); + strlcpy(drvinfo->bus_info, dev_name(netdev->dev.parent), + sizeof(drvinfo->bus_info)); +} + +/* vnic_get_sset_count - get string set count */ +static int vnic_get_sset_count(struct net_device *netdev, int sset) +{ + return (sset == ETH_SS_STATS) ? VNIC_STATS_LEN : -EOPNOTSUPP; +} + +/* vnic_get_ethtool_stats - get statistics */ +static void vnic_get_ethtool_stats(struct net_device *netdev, + struct ethtool_stats *stats, u64 *data) +{ + struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev); + struct opa_vnic_stats vstats; + int i; + + memset(&vstats, 0, sizeof(vstats)); + spin_lock(&adapter->stats_lock); + adapter->rn_ops->ndo_get_stats64(netdev, &vstats.netstats); + spin_unlock(&adapter->stats_lock); + for (i = 0; i < VNIC_STATS_LEN; i++) { + char *p = (char *)&vstats + vnic_gstrings_stats[i].stat_offset; + + data[i] = (vnic_gstrings_stats[i].sizeof_stat == + sizeof(u64)) ? *(u64 *)p : *(u32 *)p; + } +} + +/* vnic_get_strings - get strings */ +static void vnic_get_strings(struct net_device *netdev, u32 stringset, u8 *data) +{ + int i; + + if (stringset != ETH_SS_STATS) + return; + + for (i = 0; i < VNIC_STATS_LEN; i++) + memcpy(data + i * ETH_GSTRING_LEN, + vnic_gstrings_stats[i].stat_string, + ETH_GSTRING_LEN); +} + +/* ethtool ops */ +static const struct ethtool_ops opa_vnic_ethtool_ops = { + .get_drvinfo = vnic_get_drvinfo, + .get_link = ethtool_op_get_link, + .get_strings = vnic_get_strings, + .get_sset_count = vnic_get_sset_count, + .get_ethtool_stats = vnic_get_ethtool_stats, +}; + +/* opa_vnic_set_ethtool_ops - set ethtool ops */ +void opa_vnic_set_ethtool_ops(struct net_device *netdev) +{ + netdev->ethtool_ops = &opa_vnic_ethtool_ops; +} diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_internal.h b/drivers/infiniband/ulp/opa_vnic/opa_vnic_internal.h new file mode 100644 index 000000000000..ca29e6d5aedc --- /dev/null +++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_internal.h @@ -0,0 +1,329 @@ +#ifndef _OPA_VNIC_INTERNAL_H +#define _OPA_VNIC_INTERNAL_H +/* + * Copyright(c) 2017 Intel Corporation. + * + * This file is provided under a dual BSD/GPLv2 license. When using or + * redistributing this file, you may do so under either license. + * + * GPL LICENSE SUMMARY + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * BSD LICENSE + * + * 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. + * - Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + */ + +/* + * This file contains OPA VNIC driver internal declarations + */ + +#include <linux/bitops.h> +#include <linux/etherdevice.h> +#include <linux/hashtable.h> +#include <linux/sizes.h> +#include <rdma/opa_vnic.h> + +#include "opa_vnic_encap.h" + +#define OPA_VNIC_VLAN_PCP(vlan_tci) \ + (((vlan_tci) & VLAN_PRIO_MASK) >> VLAN_PRIO_SHIFT) + +/* Flow to default port redirection table size */ +#define OPA_VNIC_FLOW_TBL_SIZE 32 + +/* Invalid port number */ +#define OPA_VNIC_INVALID_PORT 0xff + +struct opa_vnic_adapter; + +/** + * struct __opa_vesw_info - OPA vnic virtual switch info + * + * Same as opa_vesw_info without bitwise attribute. + */ +struct __opa_vesw_info { + u16 fabric_id; + u16 vesw_id; + + u8 rsvd0[6]; + u16 def_port_mask; + + u8 rsvd1[2]; + u16 pkey; + + u8 rsvd2[4]; + u32 u_mcast_dlid; + u32 u_ucast_dlid[OPA_VESW_MAX_NUM_DEF_PORT]; + + u8 rsvd3[44]; + u16 eth_mtu[OPA_VNIC_MAX_NUM_PCP]; + u16 eth_mtu_non_vlan; + u8 rsvd4[2]; +} __packed; + +/** + * struct __opa_per_veswport_info - OPA vnic per port info + * + * Same as opa_per_veswport_info without bitwise attribute. + */ +struct __opa_per_veswport_info { + u32 port_num; + + u8 eth_link_status; + u8 rsvd0[3]; + + u8 base_mac_addr[ETH_ALEN]; + u8 config_state; + u8 oper_state; + + u16 max_mac_tbl_ent; + u16 max_smac_ent; + u32 mac_tbl_digest; + u8 rsvd1[4]; + + u32 encap_slid; + + u8 pcp_to_sc_uc[OPA_VNIC_MAX_NUM_PCP]; + u8 pcp_to_vl_uc[OPA_VNIC_MAX_NUM_PCP]; + u8 pcp_to_sc_mc[OPA_VNIC_MAX_NUM_PCP]; + u8 pcp_to_vl_mc[OPA_VNIC_MAX_NUM_PCP]; + + u8 non_vlan_sc_uc; + u8 non_vlan_vl_uc; + u8 non_vlan_sc_mc; + u8 non_vlan_vl_mc; + + u8 rsvd2[48]; + + u16 uc_macs_gen_count; + u16 mc_macs_gen_count; + + u8 rsvd3[8]; +} __packed; + +/** + * struct __opa_veswport_info - OPA vnic port info + * + * Same as opa_veswport_info without bitwise attribute. + */ +struct __opa_veswport_info { + struct __opa_vesw_info vesw; + struct __opa_per_veswport_info vport; +}; + +/** + * struct __opa_veswport_trap - OPA vnic trap info + * + * Same as opa_veswport_trap without bitwise attribute. + */ +struct __opa_veswport_trap { + u16 fabric_id; + u16 veswid; + u32 veswportnum; + u16 opaportnum; + u8 veswportindex; + u8 opcode; + u32 reserved; +} __packed; + +/** + * struct opa_vnic_ctrl_port - OPA virtual NIC control port + * @ibdev: pointer to ib device + * @ops: opa vnic control operations + * @num_ports: number of opa ports + */ +struct opa_vnic_ctrl_port { + struct ib_device *ibdev; + struct opa_vnic_ctrl_ops *ops; + u8 num_ports; +}; + +/** + * struct opa_vnic_adapter - OPA VNIC netdev private data structure + * @netdev: pointer to associated netdev + * @ibdev: ib device + * @cport: pointer to opa vnic control port + * @rn_ops: rdma netdev's net_device_ops + * @port_num: OPA port number + * @vport_num: vesw port number + * @lock: adapter lock + * @info: virtual ethernet switch port information + * @vema_mac_addr: mac address configured by vema + * @umac_hash: unicast maclist hash + * @mmac_hash: multicast maclist hash + * @mactbl: hash table of MAC entries + * @mactbl_lock: mac table lock + * @stats_lock: statistics lock + * @flow_tbl: flow to default port redirection table + * @trap_timeout: trap timeout + * @trap_count: no. of traps allowed within timeout period + */ +struct opa_vnic_adapter { + struct net_device *netdev; + struct ib_device *ibdev; + struct opa_vnic_ctrl_port *cport; + const struct net_device_ops *rn_ops; + + u8 port_num; + u8 vport_num; + + /* Lock used around concurrent updates to netdev */ + struct mutex lock; + + struct __opa_veswport_info info; + u8 vema_mac_addr[ETH_ALEN]; + u32 umac_hash; + u32 mmac_hash; + struct hlist_head __rcu *mactbl; + + /* Lock used to protect updates to mac table */ + struct mutex mactbl_lock; + + /* Lock used to protect access to vnic counters */ + spinlock_t stats_lock; + + u8 flow_tbl[OPA_VNIC_FLOW_TBL_SIZE]; + + unsigned long trap_timeout; + u8 trap_count; +}; + +/* Same as opa_veswport_mactable_entry, but without bitwise attribute */ +struct __opa_vnic_mactable_entry { + u8 mac_addr[ETH_ALEN]; + u8 mac_addr_mask[ETH_ALEN]; + u32 dlid_sd; +} __packed; + +/** + * struct opa_vnic_mac_tbl_node - OPA VNIC mac table node + * @hlist: hash list handle + * @index: index of entry in the mac table + * @entry: entry in the table + */ +struct opa_vnic_mac_tbl_node { + struct hlist_node hlist; + u16 index; + struct __opa_vnic_mactable_entry entry; +}; + +#define v_dbg(format, arg...) \ + netdev_dbg(adapter->netdev, format, ## arg) +#define v_err(format, arg...) \ + netdev_err(adapter->netdev, format, ## arg) +#define v_info(format, arg...) \ + netdev_info(adapter->netdev, format, ## arg) +#define v_warn(format, arg...) \ + netdev_warn(adapter->netdev, format, ## arg) + +#define c_err(format, arg...) \ + dev_err(&cport->ibdev->dev, format, ## arg) +#define c_info(format, arg...) \ + dev_info(&cport->ibdev->dev, format, ## arg) +#define c_dbg(format, arg...) \ + dev_dbg(&cport->ibdev->dev, format, ## arg) + +/* The maximum allowed entries in the mac table */ +#define OPA_VNIC_MAC_TBL_MAX_ENTRIES 2048 +/* Limit of smac entries in mac table */ +#define OPA_VNIC_MAX_SMAC_LIMIT 256 + +/* The last octet of the MAC address is used as the key to the hash table */ +#define OPA_VNIC_MAC_HASH_IDX 5 + +/* The VNIC MAC hash table is of size 2^8 */ +#define OPA_VNIC_MAC_TBL_HASH_BITS 8 +#define OPA_VNIC_MAC_TBL_SIZE BIT(OPA_VNIC_MAC_TBL_HASH_BITS) + +/* VNIC HASH MACROS */ +#define vnic_hash_init(hashtable) __hash_init(hashtable, OPA_VNIC_MAC_TBL_SIZE) + +#define vnic_hash_add(hashtable, node, key) \ + hlist_add_head(node, \ + &hashtable[hash_min(key, ilog2(OPA_VNIC_MAC_TBL_SIZE))]) + +#define vnic_hash_for_each_safe(name, bkt, tmp, obj, member) \ + for ((bkt) = 0, obj = NULL; \ + !obj && (bkt) < OPA_VNIC_MAC_TBL_SIZE; (bkt)++) \ + hlist_for_each_entry_safe(obj, tmp, &name[bkt], member) + +#define vnic_hash_for_each_possible(name, obj, member, key) \ + hlist_for_each_entry(obj, \ + &name[hash_min(key, ilog2(OPA_VNIC_MAC_TBL_SIZE))], member) + +#define vnic_hash_for_each(name, bkt, obj, member) \ + for ((bkt) = 0, obj = NULL; \ + !obj && (bkt) < OPA_VNIC_MAC_TBL_SIZE; (bkt)++) \ + hlist_for_each_entry(obj, &name[bkt], member) + +extern char opa_vnic_driver_name[]; +extern const char opa_vnic_driver_version[]; + +struct opa_vnic_adapter *opa_vnic_add_netdev(struct ib_device *ibdev, + u8 port_num, u8 vport_num); +void opa_vnic_rem_netdev(struct opa_vnic_adapter *adapter); +void opa_vnic_encap_skb(struct opa_vnic_adapter *adapter, struct sk_buff *skb); +u8 opa_vnic_get_vl(struct opa_vnic_adapter *adapter, struct sk_buff *skb); +u8 opa_vnic_calc_entropy(struct opa_vnic_adapter *adapter, struct sk_buff *skb); +void opa_vnic_process_vema_config(struct opa_vnic_adapter *adapter); +void opa_vnic_release_mac_tbl(struct opa_vnic_adapter *adapter); +void opa_vnic_query_mac_tbl(struct opa_vnic_adapter *adapter, + struct opa_veswport_mactable *tbl); +int opa_vnic_update_mac_tbl(struct opa_vnic_adapter *adapter, + struct opa_veswport_mactable *tbl); +void opa_vnic_query_ucast_macs(struct opa_vnic_adapter *adapter, + struct opa_veswport_iface_macs *macs); +void opa_vnic_query_mcast_macs(struct opa_vnic_adapter *adapter, + struct opa_veswport_iface_macs *macs); +void opa_vnic_get_summary_counters(struct opa_vnic_adapter *adapter, + struct opa_veswport_summary_counters *cntrs); +void opa_vnic_get_error_counters(struct opa_vnic_adapter *adapter, + struct opa_veswport_error_counters *cntrs); +void opa_vnic_get_vesw_info(struct opa_vnic_adapter *adapter, + struct opa_vesw_info *info); +void opa_vnic_set_vesw_info(struct opa_vnic_adapter *adapter, + struct opa_vesw_info *info); +void opa_vnic_get_per_veswport_info(struct opa_vnic_adapter *adapter, + struct opa_per_veswport_info *info); +void opa_vnic_set_per_veswport_info(struct opa_vnic_adapter *adapter, + struct opa_per_veswport_info *info); +void opa_vnic_vema_report_event(struct opa_vnic_adapter *adapter, u8 event); +void opa_vnic_set_ethtool_ops(struct net_device *netdev); +void opa_vnic_vema_send_trap(struct opa_vnic_adapter *adapter, + struct __opa_veswport_trap *data, u32 lid); + +#endif /* _OPA_VNIC_INTERNAL_H */ diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_netdev.c b/drivers/infiniband/ulp/opa_vnic/opa_vnic_netdev.c new file mode 100644 index 000000000000..1a3c25364b64 --- /dev/null +++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_netdev.c @@ -0,0 +1,387 @@ +/* + * Copyright(c) 2017 Intel Corporation. + * + * This file is provided under a dual BSD/GPLv2 license. When using or + * redistributing this file, you may do so under either license. + * + * GPL LICENSE SUMMARY + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * BSD LICENSE + * + * 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. + * - Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + */ + +/* + * This file contains OPA Virtual Network Interface Controller (VNIC) driver + * netdev functionality. + */ + +#include <linux/module.h> +#include <linux/if_vlan.h> +#include <linux/crc32.h> + +#include "opa_vnic_internal.h" + +#define OPA_TX_TIMEOUT_MS 1000 + +#define OPA_VNIC_SKB_HEADROOM \ + ALIGN((OPA_VNIC_HDR_LEN + OPA_VNIC_SKB_MDATA_LEN), 8) + +/* This function is overloaded for opa_vnic specific implementation */ +static void opa_vnic_get_stats64(struct net_device *netdev, + struct rtnl_link_stats64 *stats) +{ + struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev); + struct opa_vnic_stats vstats; + + memset(&vstats, 0, sizeof(vstats)); + spin_lock(&adapter->stats_lock); + adapter->rn_ops->ndo_get_stats64(netdev, &vstats.netstats); + spin_unlock(&adapter->stats_lock); + memcpy(stats, &vstats.netstats, sizeof(*stats)); +} + +/* opa_netdev_start_xmit - transmit function */ +static netdev_tx_t opa_netdev_start_xmit(struct sk_buff *skb, + struct net_device *netdev) +{ + struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev); + + v_dbg("xmit: queue %d skb len %d\n", skb->queue_mapping, skb->len); + /* pad to ensure mininum ethernet packet length */ + if (unlikely(skb->len < ETH_ZLEN)) { + if (skb_padto(skb, ETH_ZLEN)) + return NETDEV_TX_OK; + + skb_put(skb, ETH_ZLEN - skb->len); + } + + opa_vnic_encap_skb(adapter, skb); + return adapter->rn_ops->ndo_start_xmit(skb, netdev); +} + +static u16 opa_vnic_select_queue(struct net_device *netdev, struct sk_buff *skb, + void *accel_priv, + select_queue_fallback_t fallback) +{ + struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev); + struct opa_vnic_skb_mdata *mdata; + int rc; + + /* pass entropy and vl as metadata in skb */ + mdata = skb_push(skb, sizeof(*mdata)); + mdata->entropy = opa_vnic_calc_entropy(adapter, skb); + mdata->vl = opa_vnic_get_vl(adapter, skb); + rc = adapter->rn_ops->ndo_select_queue(netdev, skb, + accel_priv, fallback); + skb_pull(skb, sizeof(*mdata)); + return rc; +} + +/* opa_vnic_process_vema_config - process vema configuration updates */ +void opa_vnic_process_vema_config(struct opa_vnic_adapter *adapter) +{ + struct __opa_veswport_info *info = &adapter->info; + struct rdma_netdev *rn = netdev_priv(adapter->netdev); + u8 port_num[OPA_VESW_MAX_NUM_DEF_PORT] = { 0 }; + struct net_device *netdev = adapter->netdev; + u8 i, port_count = 0; + u16 port_mask; + + /* If the base_mac_addr is changed, update the interface mac address */ + if (memcmp(info->vport.base_mac_addr, adapter->vema_mac_addr, + ARRAY_SIZE(info->vport.base_mac_addr))) { + struct sockaddr saddr; + + memcpy(saddr.sa_data, info->vport.base_mac_addr, + ARRAY_SIZE(info->vport.base_mac_addr)); + mutex_lock(&adapter->lock); + eth_mac_addr(netdev, &saddr); + memcpy(adapter->vema_mac_addr, + info->vport.base_mac_addr, ETH_ALEN); + mutex_unlock(&adapter->lock); + } + + rn->set_id(netdev, info->vesw.vesw_id); + + /* Handle MTU limit change */ + rtnl_lock(); + netdev->max_mtu = max_t(unsigned int, info->vesw.eth_mtu_non_vlan, + netdev->min_mtu); + if (netdev->mtu > netdev->max_mtu) + dev_set_mtu(netdev, netdev->max_mtu); + rtnl_unlock(); + + /* Update flow to default port redirection table */ + port_mask = info->vesw.def_port_mask; + for (i = 0; i < OPA_VESW_MAX_NUM_DEF_PORT; i++) { + if (port_mask & 1) + port_num[port_count++] = i; + port_mask >>= 1; + } + + /* + * Build the flow table. Flow table is required when destination LID + * is not available. Up to OPA_VNIC_FLOW_TBL_SIZE flows supported. + * Each flow need a default port number to get its dlid from the + * u_ucast_dlid array. + */ + for (i = 0; i < OPA_VNIC_FLOW_TBL_SIZE; i++) + adapter->flow_tbl[i] = port_count ? port_num[i % port_count] : + OPA_VNIC_INVALID_PORT; + + /* Operational state can only be DROP_ALL or FORWARDING */ + if (info->vport.config_state == OPA_VNIC_STATE_FORWARDING) { + info->vport.oper_state = OPA_VNIC_STATE_FORWARDING; + netif_dormant_off(netdev); + } else { + info->vport.oper_state = OPA_VNIC_STATE_DROP_ALL; + netif_dormant_on(netdev); + } +} + +/* + * Set the power on default values in adapter's vema interface structure. + */ +static inline void opa_vnic_set_pod_values(struct opa_vnic_adapter *adapter) +{ + adapter->info.vport.max_mac_tbl_ent = OPA_VNIC_MAC_TBL_MAX_ENTRIES; + adapter->info.vport.max_smac_ent = OPA_VNIC_MAX_SMAC_LIMIT; + adapter->info.vport.config_state = OPA_VNIC_STATE_DROP_ALL; + adapter->info.vport.eth_link_status = OPA_VNIC_ETH_LINK_DOWN; +} + +/* opa_vnic_set_mac_addr - change mac address */ +static int opa_vnic_set_mac_addr(struct net_device *netdev, void *addr) +{ + struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev); + struct sockaddr *sa = addr; + int rc; + + if (!memcmp(netdev->dev_addr, sa->sa_data, ETH_ALEN)) + return 0; + + mutex_lock(&adapter->lock); + rc = eth_mac_addr(netdev, addr); + mutex_unlock(&adapter->lock); + if (rc) + return rc; + + adapter->info.vport.uc_macs_gen_count++; + opa_vnic_vema_report_event(adapter, + OPA_VESWPORT_TRAP_IFACE_UCAST_MAC_CHANGE); + return 0; +} + +/* + * opa_vnic_mac_send_event - post event on possible mac list exchange + * Send trap when digest from uc/mc mac list differs from previous run. + * Digest is evaluated similar to how cksum does. + */ +static void opa_vnic_mac_send_event(struct net_device *netdev, u8 event) +{ + struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev); + struct netdev_hw_addr *ha; + struct netdev_hw_addr_list *hw_list; + u32 *ref_crc; + u32 l, crc = 0; + + switch (event) { + case OPA_VESWPORT_TRAP_IFACE_UCAST_MAC_CHANGE: + hw_list = &netdev->uc; + adapter->info.vport.uc_macs_gen_count++; + ref_crc = &adapter->umac_hash; + break; + case OPA_VESWPORT_TRAP_IFACE_MCAST_MAC_CHANGE: + hw_list = &netdev->mc; + adapter->info.vport.mc_macs_gen_count++; + ref_crc = &adapter->mmac_hash; + break; + default: + return; + } + netdev_hw_addr_list_for_each(ha, hw_list) { + crc = crc32_le(crc, ha->addr, ETH_ALEN); + } + l = netdev_hw_addr_list_count(hw_list) * ETH_ALEN; + crc = ~crc32_le(crc, (void *)&l, sizeof(l)); + + if (crc != *ref_crc) { + *ref_crc = crc; + opa_vnic_vema_report_event(adapter, event); + } +} + +/* opa_vnic_set_rx_mode - handle uc/mc mac list change */ +static void opa_vnic_set_rx_mode(struct net_device *netdev) +{ + opa_vnic_mac_send_event(netdev, + OPA_VESWPORT_TRAP_IFACE_UCAST_MAC_CHANGE); + + opa_vnic_mac_send_event(netdev, + OPA_VESWPORT_TRAP_IFACE_MCAST_MAC_CHANGE); +} + +/* opa_netdev_open - activate network interface */ +static int opa_netdev_open(struct net_device *netdev) +{ + struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev); + int rc; + + rc = adapter->rn_ops->ndo_open(adapter->netdev); + if (rc) { + v_dbg("open failed %d\n", rc); + return rc; + } + + /* Update eth link status and send trap */ + adapter->info.vport.eth_link_status = OPA_VNIC_ETH_LINK_UP; + opa_vnic_vema_report_event(adapter, + OPA_VESWPORT_TRAP_ETH_LINK_STATUS_CHANGE); + return 0; +} + +/* opa_netdev_close - disable network interface */ +static int opa_netdev_close(struct net_device *netdev) +{ + struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev); + int rc; + + rc = adapter->rn_ops->ndo_stop(adapter->netdev); + if (rc) { + v_dbg("close failed %d\n", rc); + return rc; + } + + /* Update eth link status and send trap */ + adapter->info.vport.eth_link_status = OPA_VNIC_ETH_LINK_DOWN; + opa_vnic_vema_report_event(adapter, + OPA_VESWPORT_TRAP_ETH_LINK_STATUS_CHANGE); + return 0; +} + +/* netdev ops */ +static const struct net_device_ops opa_netdev_ops = { + .ndo_open = opa_netdev_open, + .ndo_stop = opa_netdev_close, + .ndo_start_xmit = opa_netdev_start_xmit, + .ndo_get_stats64 = opa_vnic_get_stats64, + .ndo_set_rx_mode = opa_vnic_set_rx_mode, + .ndo_select_queue = opa_vnic_select_queue, + .ndo_set_mac_address = opa_vnic_set_mac_addr, +}; + +/* opa_vnic_add_netdev - create vnic netdev interface */ +struct opa_vnic_adapter *opa_vnic_add_netdev(struct ib_device *ibdev, + u8 port_num, u8 vport_num) +{ + struct opa_vnic_adapter *adapter; + struct net_device *netdev; + struct rdma_netdev *rn; + int rc; + + netdev = ibdev->alloc_rdma_netdev(ibdev, port_num, + RDMA_NETDEV_OPA_VNIC, + "veth%d", NET_NAME_UNKNOWN, + ether_setup); + if (!netdev) + return ERR_PTR(-ENOMEM); + else if (IS_ERR(netdev)) + return ERR_CAST(netdev); + + rn = netdev_priv(netdev); + adapter = kzalloc(sizeof(*adapter), GFP_KERNEL); + if (!adapter) { + rc = -ENOMEM; + goto adapter_err; + } + + rn->clnt_priv = adapter; + rn->hca = ibdev; + rn->port_num = port_num; + adapter->netdev = netdev; + adapter->ibdev = ibdev; + adapter->port_num = port_num; + adapter->vport_num = vport_num; + adapter->rn_ops = netdev->netdev_ops; + + netdev->netdev_ops = &opa_netdev_ops; + netdev->priv_flags |= IFF_LIVE_ADDR_CHANGE; + netdev->hard_header_len += OPA_VNIC_SKB_HEADROOM; + mutex_init(&adapter->lock); + mutex_init(&adapter->mactbl_lock); + spin_lock_init(&adapter->stats_lock); + + SET_NETDEV_DEV(netdev, ibdev->dev.parent); + + opa_vnic_set_ethtool_ops(netdev); + + opa_vnic_set_pod_values(adapter); + + rc = register_netdev(netdev); + if (rc) + goto netdev_err; + + netif_carrier_off(netdev); + netif_dormant_on(netdev); + v_info("initialized\n"); + + return adapter; +netdev_err: + mutex_destroy(&adapter->lock); + mutex_destroy(&adapter->mactbl_lock); + kfree(adapter); +adapter_err: + rn->free_rdma_netdev(netdev); + + return ERR_PTR(rc); +} + +/* opa_vnic_rem_netdev - remove vnic netdev interface */ +void opa_vnic_rem_netdev(struct opa_vnic_adapter *adapter) +{ + struct net_device *netdev = adapter->netdev; + struct rdma_netdev *rn = netdev_priv(netdev); + + v_info("removing\n"); + unregister_netdev(netdev); + opa_vnic_release_mac_tbl(adapter); + mutex_destroy(&adapter->lock); + mutex_destroy(&adapter->mactbl_lock); + kfree(adapter); + rn->free_rdma_netdev(netdev); +} diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_vema.c b/drivers/infiniband/ulp/opa_vnic/opa_vnic_vema.c new file mode 100644 index 000000000000..cf768dd78d1b --- /dev/null +++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_vema.c @@ -0,0 +1,1056 @@ +/* + * Copyright(c) 2017 Intel Corporation. + * + * This file is provided under a dual BSD/GPLv2 license. When using or + * redistributing this file, you may do so under either license. + * + * GPL LICENSE SUMMARY + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * BSD LICENSE + * + * 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. + * - Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + */ + +/* + * This file contains OPA Virtual Network Interface Controller (VNIC) + * Ethernet Management Agent (EMA) driver + */ + +#include <linux/module.h> +#include <rdma/ib_addr.h> +#include <rdma/ib_smi.h> + +#include "opa_vnic_internal.h" + +#define DRV_VERSION "1.0" +char opa_vnic_driver_name[] = "opa_vnic"; +const char opa_vnic_driver_version[] = DRV_VERSION; + +/* + * The trap service level is kept in bits 3 to 7 in the trap_sl_rsvd + * field in the class port info MAD. + */ +#define GET_TRAP_SL_FROM_CLASS_PORT_INFO(x) (((x) >> 3) & 0x1f) + +/* Cap trap bursts to a reasonable limit good for normal cases */ +#define OPA_VNIC_TRAP_BURST_LIMIT 4 + +/* + * VNIC trap limit timeout. + * Inverse of cap2_mask response time out (1.0737 secs) = 0.9 + * secs approx IB spec 13.4.6.2.1 PortInfoSubnetTimeout and + * 13.4.9 Traps. + */ +#define OPA_VNIC_TRAP_TIMEOUT ((4096 * (1UL << 18)) / 1000) + +#define OPA_VNIC_UNSUP_ATTR \ + cpu_to_be16(IB_MGMT_MAD_STATUS_UNSUPPORTED_METHOD_ATTRIB) + +#define OPA_VNIC_INVAL_ATTR \ + cpu_to_be16(IB_MGMT_MAD_STATUS_INVALID_ATTRIB_VALUE) + +#define OPA_VNIC_CLASS_CAP_TRAP 0x1 + +/* Maximum number of VNIC ports supported */ +#define OPA_VNIC_MAX_NUM_VPORT 255 + +/** + * struct opa_vnic_vema_port -- VNIC VEMA port details + * @cport: pointer to port + * @mad_agent: pointer to mad agent for port + * @class_port_info: Class port info information. + * @tid: Transaction id + * @port_num: OPA port number + * @vport_idr: vnic ports idr + * @event_handler: ib event handler + * @lock: adapter interface lock + */ +struct opa_vnic_vema_port { + struct opa_vnic_ctrl_port *cport; + struct ib_mad_agent *mad_agent; + struct opa_class_port_info class_port_info; + u64 tid; + u8 port_num; + struct idr vport_idr; + struct ib_event_handler event_handler; + + /* Lock to query/update network adapter */ + struct mutex lock; +}; + +static void opa_vnic_vema_add_one(struct ib_device *device); +static void opa_vnic_vema_rem_one(struct ib_device *device, + void *client_data); + +static struct ib_client opa_vnic_client = { + .name = opa_vnic_driver_name, + .add = opa_vnic_vema_add_one, + .remove = opa_vnic_vema_rem_one, +}; + +/** + * vema_get_vport_num -- Get the vnic from the mad + * @recvd_mad: Received mad + * + * Return: returns value of the vnic port number + */ +static inline u8 vema_get_vport_num(struct opa_vnic_vema_mad *recvd_mad) +{ + return be32_to_cpu(recvd_mad->mad_hdr.attr_mod) & 0xff; +} + +/** + * vema_get_vport_adapter -- Get vnic port adapter from recvd mad + * @recvd_mad: received mad + * @port: ptr to port struct on which MAD was recvd + * + * Return: vnic adapter + */ +static inline struct opa_vnic_adapter * +vema_get_vport_adapter(struct opa_vnic_vema_mad *recvd_mad, + struct opa_vnic_vema_port *port) +{ + u8 vport_num = vema_get_vport_num(recvd_mad); + + return idr_find(&port->vport_idr, vport_num); +} + +/** + * vema_mac_tbl_req_ok -- Check if mac request has correct values + * @mac_tbl: mac table + * + * This function checks for the validity of the offset and number of + * entries required. + * + * Return: true if offset and num_entries are valid + */ +static inline bool vema_mac_tbl_req_ok(struct opa_veswport_mactable *mac_tbl) +{ + u16 offset, num_entries; + u16 req_entries = ((OPA_VNIC_EMA_DATA - sizeof(*mac_tbl)) / + sizeof(mac_tbl->tbl_entries[0])); + + offset = be16_to_cpu(mac_tbl->offset); + num_entries = be16_to_cpu(mac_tbl->num_entries); + + return ((num_entries <= req_entries) && + (offset + num_entries <= OPA_VNIC_MAC_TBL_MAX_ENTRIES)); +} + +/* + * Return the power on default values in the port info structure + * in big endian format as required by MAD. + */ +static inline void vema_get_pod_values(struct opa_veswport_info *port_info) +{ + memset(port_info, 0, sizeof(*port_info)); + port_info->vport.max_mac_tbl_ent = + cpu_to_be16(OPA_VNIC_MAC_TBL_MAX_ENTRIES); + port_info->vport.max_smac_ent = + cpu_to_be16(OPA_VNIC_MAX_SMAC_LIMIT); + port_info->vport.oper_state = OPA_VNIC_STATE_DROP_ALL; + port_info->vport.config_state = OPA_VNIC_STATE_DROP_ALL; +} + +/** + * vema_add_vport -- Add a new vnic port + * @port: ptr to opa_vnic_vema_port struct + * @vport_num: vnic port number (to be added) + * + * Return a pointer to the vnic adapter structure + */ +static struct opa_vnic_adapter *vema_add_vport(struct opa_vnic_vema_port *port, + u8 vport_num) +{ + struct opa_vnic_ctrl_port *cport = port->cport; + struct opa_vnic_adapter *adapter; + + adapter = opa_vnic_add_netdev(cport->ibdev, port->port_num, vport_num); + if (!IS_ERR(adapter)) { + int rc; + + adapter->cport = cport; + rc = idr_alloc(&port->vport_idr, adapter, vport_num, + vport_num + 1, GFP_NOWAIT); + if (rc < 0) { + opa_vnic_rem_netdev(adapter); + adapter = ERR_PTR(rc); + } + } + + return adapter; +} + +/** + * vema_get_class_port_info -- Get class info for port + * @port: Port on whic MAD was received + * @recvd_mad: pointer to the received mad + * @rsp_mad: pointer to respose mad + * + * This function copies the latest class port info value set for the + * port and stores it for generating traps + */ +static void vema_get_class_port_info(struct opa_vnic_vema_port *port, + struct opa_vnic_vema_mad *recvd_mad, + struct opa_vnic_vema_mad *rsp_mad) +{ + struct opa_class_port_info *port_info; + + port_info = (struct opa_class_port_info *)rsp_mad->data; + memcpy(port_info, &port->class_port_info, sizeof(*port_info)); + port_info->base_version = OPA_MGMT_BASE_VERSION, + port_info->class_version = OPA_EMA_CLASS_VERSION; + + /* + * Set capability mask bit indicating agent generates traps, + * and set the maximum number of VNIC ports supported. + */ + port_info->cap_mask = cpu_to_be16((OPA_VNIC_CLASS_CAP_TRAP | + (OPA_VNIC_MAX_NUM_VPORT << 8))); + + /* + * Since a get routine is always sent by the EM first we + * set the expected response time to + * 4.096 usec * 2^18 == 1.0737 sec here. + */ + port_info->cap_mask2_resp_time = cpu_to_be32(18); +} + +/** + * vema_set_class_port_info -- Get class info for port + * @port: Port on whic MAD was received + * @recvd_mad: pointer to the received mad + * @rsp_mad: pointer to respose mad + * + * This function updates the port class info for the specific vnic + * and sets up the response mad data + */ +static void vema_set_class_port_info(struct opa_vnic_vema_port *port, + struct opa_vnic_vema_mad *recvd_mad, + struct opa_vnic_vema_mad *rsp_mad) +{ + memcpy(&port->class_port_info, recvd_mad->data, + sizeof(port->class_port_info)); + + vema_get_class_port_info(port, recvd_mad, rsp_mad); +} + +/** + * vema_get_veswport_info -- Get veswport info + * @port: source port on which MAD was received + * @recvd_mad: pointer to the received mad + * @rsp_mad: pointer to respose mad + */ +static void vema_get_veswport_info(struct opa_vnic_vema_port *port, + struct opa_vnic_vema_mad *recvd_mad, + struct opa_vnic_vema_mad *rsp_mad) +{ + struct opa_veswport_info *port_info = + (struct opa_veswport_info *)rsp_mad->data; + struct opa_vnic_adapter *adapter; + + adapter = vema_get_vport_adapter(recvd_mad, port); + if (adapter) { + memset(port_info, 0, sizeof(*port_info)); + opa_vnic_get_vesw_info(adapter, &port_info->vesw); + opa_vnic_get_per_veswport_info(adapter, + &port_info->vport); + } else { + vema_get_pod_values(port_info); + } +} + +/** + * vema_set_veswport_info -- Set veswport info + * @port: source port on which MAD was received + * @recvd_mad: pointer to the received mad + * @rsp_mad: pointer to respose mad + * + * This function gets the port class infor for vnic + */ +static void vema_set_veswport_info(struct opa_vnic_vema_port *port, + struct opa_vnic_vema_mad *recvd_mad, + struct opa_vnic_vema_mad *rsp_mad) +{ + struct opa_vnic_ctrl_port *cport = port->cport; + struct opa_veswport_info *port_info; + struct opa_vnic_adapter *adapter; + u8 vport_num; + + vport_num = vema_get_vport_num(recvd_mad); + + adapter = vema_get_vport_adapter(recvd_mad, port); + if (!adapter) { + adapter = vema_add_vport(port, vport_num); + if (IS_ERR(adapter)) { + c_err("failed to add vport %d: %ld\n", + vport_num, PTR_ERR(adapter)); + goto err_exit; + } + } + + port_info = (struct opa_veswport_info *)recvd_mad->data; + opa_vnic_set_vesw_info(adapter, &port_info->vesw); + opa_vnic_set_per_veswport_info(adapter, &port_info->vport); + + /* Process the new config settings */ + opa_vnic_process_vema_config(adapter); + + vema_get_veswport_info(port, recvd_mad, rsp_mad); + return; + +err_exit: + rsp_mad->mad_hdr.status = OPA_VNIC_INVAL_ATTR; +} + +/** + * vema_get_mac_entries -- Get MAC entries in VNIC MAC table + * @port: source port on which MAD was received + * @recvd_mad: pointer to the received mad + * @rsp_mad: pointer to respose mad + * + * This function gets the MAC entries that are programmed into + * the VNIC MAC forwarding table. It checks for the validity of + * the index into the MAC table and the number of entries that + * are to be retrieved. + */ +static void vema_get_mac_entries(struct opa_vnic_vema_port *port, + struct opa_vnic_vema_mad *recvd_mad, + struct opa_vnic_vema_mad *rsp_mad) +{ + struct opa_veswport_mactable *mac_tbl_in, *mac_tbl_out; + struct opa_vnic_adapter *adapter; + + adapter = vema_get_vport_adapter(recvd_mad, port); + if (!adapter) { + rsp_mad->mad_hdr.status = OPA_VNIC_INVAL_ATTR; + return; + } + + mac_tbl_in = (struct opa_veswport_mactable *)recvd_mad->data; + mac_tbl_out = (struct opa_veswport_mactable *)rsp_mad->data; + + if (vema_mac_tbl_req_ok(mac_tbl_in)) { + mac_tbl_out->offset = mac_tbl_in->offset; + mac_tbl_out->num_entries = mac_tbl_in->num_entries; + opa_vnic_query_mac_tbl(adapter, mac_tbl_out); + } else { + rsp_mad->mad_hdr.status = OPA_VNIC_INVAL_ATTR; + } +} + +/** + * vema_set_mac_entries -- Set MAC entries in VNIC MAC table + * @port: source port on which MAD was received + * @recvd_mad: pointer to the received mad + * @rsp_mad: pointer to respose mad + * + * This function sets the MAC entries in the VNIC forwarding table + * It checks for the validity of the index and the number of forwarding + * table entries to be programmed. + */ +static void vema_set_mac_entries(struct opa_vnic_vema_port *port, + struct opa_vnic_vema_mad *recvd_mad, + struct opa_vnic_vema_mad *rsp_mad) +{ + struct opa_veswport_mactable *mac_tbl; + struct opa_vnic_adapter *adapter; + + adapter = vema_get_vport_adapter(recvd_mad, port); + if (!adapter) { + rsp_mad->mad_hdr.status = OPA_VNIC_INVAL_ATTR; + return; + } + + mac_tbl = (struct opa_veswport_mactable *)recvd_mad->data; + if (vema_mac_tbl_req_ok(mac_tbl)) { + if (opa_vnic_update_mac_tbl(adapter, mac_tbl)) + rsp_mad->mad_hdr.status = OPA_VNIC_UNSUP_ATTR; + } else { + rsp_mad->mad_hdr.status = OPA_VNIC_UNSUP_ATTR; + } + vema_get_mac_entries(port, recvd_mad, rsp_mad); +} + +/** + * vema_set_delete_vesw -- Reset VESW info to POD values + * @port: source port on which MAD was received + * @recvd_mad: pointer to the received mad + * @rsp_mad: pointer to respose mad + * + * This function clears all the fields of veswport info for the requested vesw + * and sets them back to the power-on default values. It does not delete the + * vesw. + */ +static void vema_set_delete_vesw(struct opa_vnic_vema_port *port, + struct opa_vnic_vema_mad *recvd_mad, + struct opa_vnic_vema_mad *rsp_mad) +{ + struct opa_veswport_info *port_info = + (struct opa_veswport_info *)rsp_mad->data; + struct opa_vnic_adapter *adapter; + + adapter = vema_get_vport_adapter(recvd_mad, port); + if (!adapter) { + rsp_mad->mad_hdr.status = OPA_VNIC_INVAL_ATTR; + return; + } + + vema_get_pod_values(port_info); + opa_vnic_set_vesw_info(adapter, &port_info->vesw); + opa_vnic_set_per_veswport_info(adapter, &port_info->vport); + + /* Process the new config settings */ + opa_vnic_process_vema_config(adapter); + + opa_vnic_release_mac_tbl(adapter); + + vema_get_veswport_info(port, recvd_mad, rsp_mad); +} + +/** + * vema_get_mac_list -- Get the unicast/multicast macs. + * @port: source port on which MAD was received + * @recvd_mad: Received mad contains fields to set vnic parameters + * @rsp_mad: Response mad to be built + * @attr_id: Attribute ID indicating multicast or unicast mac list + */ +static void vema_get_mac_list(struct opa_vnic_vema_port *port, + struct opa_vnic_vema_mad *recvd_mad, + struct opa_vnic_vema_mad *rsp_mad, + u16 attr_id) +{ + struct opa_veswport_iface_macs *macs_in, *macs_out; + int max_entries = (OPA_VNIC_EMA_DATA - sizeof(*macs_out)) / ETH_ALEN; + struct opa_vnic_adapter *adapter; + + adapter = vema_get_vport_adapter(recvd_mad, port); + if (!adapter) { + rsp_mad->mad_hdr.status = OPA_VNIC_INVAL_ATTR; + return; + } + + macs_in = (struct opa_veswport_iface_macs *)recvd_mad->data; + macs_out = (struct opa_veswport_iface_macs *)rsp_mad->data; + + macs_out->start_idx = macs_in->start_idx; + if (macs_in->num_macs_in_msg) + macs_out->num_macs_in_msg = macs_in->num_macs_in_msg; + else + macs_out->num_macs_in_msg = cpu_to_be16(max_entries); + + if (attr_id == OPA_EM_ATTR_IFACE_MCAST_MACS) + opa_vnic_query_mcast_macs(adapter, macs_out); + else + opa_vnic_query_ucast_macs(adapter, macs_out); +} + +/** + * vema_get_summary_counters -- Gets summary counters. + * @port: source port on which MAD was received + * @recvd_mad: Received mad contains fields to set vnic parameters + * @rsp_mad: Response mad to be built + */ +static void vema_get_summary_counters(struct opa_vnic_vema_port *port, + struct opa_vnic_vema_mad *recvd_mad, + struct opa_vnic_vema_mad *rsp_mad) +{ + struct opa_veswport_summary_counters *cntrs; + struct opa_vnic_adapter *adapter; + + adapter = vema_get_vport_adapter(recvd_mad, port); + if (adapter) { + cntrs = (struct opa_veswport_summary_counters *)rsp_mad->data; + opa_vnic_get_summary_counters(adapter, cntrs); + } else { + rsp_mad->mad_hdr.status = OPA_VNIC_INVAL_ATTR; + } +} + +/** + * vema_get_error_counters -- Gets summary counters. + * @port: source port on which MAD was received + * @recvd_mad: Received mad contains fields to set vnic parameters + * @rsp_mad: Response mad to be built + */ +static void vema_get_error_counters(struct opa_vnic_vema_port *port, + struct opa_vnic_vema_mad *recvd_mad, + struct opa_vnic_vema_mad *rsp_mad) +{ + struct opa_veswport_error_counters *cntrs; + struct opa_vnic_adapter *adapter; + + adapter = vema_get_vport_adapter(recvd_mad, port); + if (adapter) { + cntrs = (struct opa_veswport_error_counters *)rsp_mad->data; + opa_vnic_get_error_counters(adapter, cntrs); + } else { + rsp_mad->mad_hdr.status = OPA_VNIC_INVAL_ATTR; + } +} + +/** + * vema_get -- Process received get MAD + * @port: source port on which MAD was received + * @recvd_mad: Received mad + * @rsp_mad: Response mad to be built + */ +static void vema_get(struct opa_vnic_vema_port *port, + struct opa_vnic_vema_mad *recvd_mad, + struct opa_vnic_vema_mad *rsp_mad) +{ + u16 attr_id = be16_to_cpu(recvd_mad->mad_hdr.attr_id); + + switch (attr_id) { + case OPA_EM_ATTR_CLASS_PORT_INFO: + vema_get_class_port_info(port, recvd_mad, rsp_mad); + break; + case OPA_EM_ATTR_VESWPORT_INFO: + vema_get_veswport_info(port, recvd_mad, rsp_mad); + break; + case OPA_EM_ATTR_VESWPORT_MAC_ENTRIES: + vema_get_mac_entries(port, recvd_mad, rsp_mad); + break; + case OPA_EM_ATTR_IFACE_UCAST_MACS: + /* fall through */ + case OPA_EM_ATTR_IFACE_MCAST_MACS: + vema_get_mac_list(port, recvd_mad, rsp_mad, attr_id); + break; + case OPA_EM_ATTR_VESWPORT_SUMMARY_COUNTERS: + vema_get_summary_counters(port, recvd_mad, rsp_mad); + break; + case OPA_EM_ATTR_VESWPORT_ERROR_COUNTERS: + vema_get_error_counters(port, recvd_mad, rsp_mad); + break; + default: + rsp_mad->mad_hdr.status = OPA_VNIC_UNSUP_ATTR; + break; + } +} + +/** + * vema_set -- Process received set MAD + * @port: source port on which MAD was received + * @recvd_mad: Received mad contains fields to set vnic parameters + * @rsp_mad: Response mad to be built + */ +static void vema_set(struct opa_vnic_vema_port *port, + struct opa_vnic_vema_mad *recvd_mad, + struct opa_vnic_vema_mad *rsp_mad) +{ + u16 attr_id = be16_to_cpu(recvd_mad->mad_hdr.attr_id); + + switch (attr_id) { + case OPA_EM_ATTR_CLASS_PORT_INFO: + vema_set_class_port_info(port, recvd_mad, rsp_mad); + break; + case OPA_EM_ATTR_VESWPORT_INFO: + vema_set_veswport_info(port, recvd_mad, rsp_mad); + break; + case OPA_EM_ATTR_VESWPORT_MAC_ENTRIES: + vema_set_mac_entries(port, recvd_mad, rsp_mad); + break; + case OPA_EM_ATTR_DELETE_VESW: + vema_set_delete_vesw(port, recvd_mad, rsp_mad); + break; + default: + rsp_mad->mad_hdr.status = OPA_VNIC_UNSUP_ATTR; + break; + } +} + +/** + * vema_send -- Send handler for VEMA MAD agent + * @mad_agent: pointer to the mad agent + * @mad_wc: pointer to mad send work completion information + * + * Free all the data structures associated with the sent MAD + */ +static void vema_send(struct ib_mad_agent *mad_agent, + struct ib_mad_send_wc *mad_wc) +{ + rdma_destroy_ah(mad_wc->send_buf->ah); + ib_free_send_mad(mad_wc->send_buf); +} + +/** + * vema_recv -- Recv handler for VEMA MAD agent + * @mad_agent: pointer to the mad agent + * @send_buf: Send buffer if found, else NULL + * @mad_wc: pointer to mad send work completion information + * + * Handle only set and get methods and respond to other methods + * as unsupported. Allocate response buffer and address handle + * for the response MAD. + */ +static void vema_recv(struct ib_mad_agent *mad_agent, + struct ib_mad_send_buf *send_buf, + struct ib_mad_recv_wc *mad_wc) +{ + struct opa_vnic_vema_port *port; + struct ib_ah *ah; + struct ib_mad_send_buf *rsp; + struct opa_vnic_vema_mad *vema_mad; + + if (!mad_wc || !mad_wc->recv_buf.mad) + return; + + port = mad_agent->context; + ah = ib_create_ah_from_wc(mad_agent->qp->pd, mad_wc->wc, + mad_wc->recv_buf.grh, mad_agent->port_num); + if (IS_ERR(ah)) + goto free_recv_mad; + + rsp = ib_create_send_mad(mad_agent, mad_wc->wc->src_qp, + mad_wc->wc->pkey_index, 0, + IB_MGMT_VENDOR_HDR, OPA_VNIC_EMA_DATA, + GFP_KERNEL, OPA_MGMT_BASE_VERSION); + if (IS_ERR(rsp)) + goto err_rsp; + + rsp->ah = ah; + vema_mad = rsp->mad; + memcpy(vema_mad, mad_wc->recv_buf.mad, IB_MGMT_VENDOR_HDR); + vema_mad->mad_hdr.method = IB_MGMT_METHOD_GET_RESP; + vema_mad->mad_hdr.status = 0; + + /* Lock ensures network adapter is not removed */ + mutex_lock(&port->lock); + + switch (mad_wc->recv_buf.mad->mad_hdr.method) { + case IB_MGMT_METHOD_GET: + vema_get(port, (struct opa_vnic_vema_mad *)mad_wc->recv_buf.mad, + vema_mad); + break; + case IB_MGMT_METHOD_SET: + vema_set(port, (struct opa_vnic_vema_mad *)mad_wc->recv_buf.mad, + vema_mad); + break; + default: + vema_mad->mad_hdr.status = OPA_VNIC_UNSUP_ATTR; + break; + } + mutex_unlock(&port->lock); + + if (!ib_post_send_mad(rsp, NULL)) { + /* + * with post send successful ah and send mad + * will be destroyed in send handler + */ + goto free_recv_mad; + } + + ib_free_send_mad(rsp); + +err_rsp: + rdma_destroy_ah(ah); +free_recv_mad: + ib_free_recv_mad(mad_wc); +} + +/** + * vema_get_port -- Gets the opa_vnic_vema_port + * @cport: pointer to control dev + * @port_num: Port number + * + * This function loops through the ports and returns + * the opa_vnic_vema port structure that is associated + * with the OPA port number + * + * Return: ptr to requested opa_vnic_vema_port strucure + * if success, NULL if not + */ +static struct opa_vnic_vema_port * +vema_get_port(struct opa_vnic_ctrl_port *cport, u8 port_num) +{ + struct opa_vnic_vema_port *port = (void *)cport + sizeof(*cport); + + if (port_num > cport->num_ports) + return NULL; + + return port + (port_num - 1); +} + +/** + * opa_vnic_vema_send_trap -- This function sends a trap to the EM + * @cport: pointer to vnic control port + * @data: pointer to trap data filled by calling function + * @lid: issuers lid (encap_slid from vesw_port_info) + * + * This function is called from the VNIC driver to send a trap if there + * is somethng the EM should be notified about. These events currently + * are + * 1) UNICAST INTERFACE MACADDRESS changes + * 2) MULTICAST INTERFACE MACADDRESS changes + * 3) ETHERNET LINK STATUS changes + * While allocating the send mad the remote site qpn used is 1 + * as this is the well known QP. + * + */ +void opa_vnic_vema_send_trap(struct opa_vnic_adapter *adapter, + struct __opa_veswport_trap *data, u32 lid) +{ + struct opa_vnic_ctrl_port *cport = adapter->cport; + struct ib_mad_send_buf *send_buf; + struct opa_vnic_vema_port *port; + struct ib_device *ibp; + struct opa_vnic_vema_mad_trap *trap_mad; + struct opa_class_port_info *class; + struct rdma_ah_attr ah_attr; + struct ib_ah *ah; + struct opa_veswport_trap *trap; + u32 trap_lid; + u16 pkey_idx; + + if (!cport) + goto err_exit; + ibp = cport->ibdev; + port = vema_get_port(cport, data->opaportnum); + if (!port || !port->mad_agent) + goto err_exit; + + if (time_before(jiffies, adapter->trap_timeout)) { + if (adapter->trap_count == OPA_VNIC_TRAP_BURST_LIMIT) { + v_warn("Trap rate exceeded\n"); + goto err_exit; + } else { + adapter->trap_count++; + } + } else { + adapter->trap_count = 0; + } + + class = &port->class_port_info; + /* Set up address handle */ + memset(&ah_attr, 0, sizeof(ah_attr)); + ah_attr.type = rdma_ah_find_type(ibp, port->port_num); + rdma_ah_set_sl(&ah_attr, + GET_TRAP_SL_FROM_CLASS_PORT_INFO(class->trap_sl_rsvd)); + rdma_ah_set_port_num(&ah_attr, port->port_num); + trap_lid = be32_to_cpu(class->trap_lid); + /* + * check for trap lid validity, must not be zero + * The trap sink could change after we fashion the MAD but since traps + * are not guaranteed we won't use a lock as anyway the change will take + * place even with locking. + */ + if (!trap_lid) { + c_err("%s: Invalid dlid\n", __func__); + goto err_exit; + } + + rdma_ah_set_dlid(&ah_attr, trap_lid); + ah = rdma_create_ah(port->mad_agent->qp->pd, &ah_attr); + if (IS_ERR(ah)) { + c_err("%s:Couldn't create new AH = %p\n", __func__, ah); + c_err("%s:dlid = %d, sl = %d, port = %d\n", __func__, + rdma_ah_get_dlid(&ah_attr), rdma_ah_get_sl(&ah_attr), + rdma_ah_get_port_num(&ah_attr)); + goto err_exit; + } + + if (ib_find_pkey(ibp, data->opaportnum, IB_DEFAULT_PKEY_FULL, + &pkey_idx) < 0) { + c_err("%s:full key not found, defaulting to partial\n", + __func__); + if (ib_find_pkey(ibp, data->opaportnum, IB_DEFAULT_PKEY_PARTIAL, + &pkey_idx) < 0) + pkey_idx = 1; + } + + send_buf = ib_create_send_mad(port->mad_agent, 1, pkey_idx, 0, + IB_MGMT_VENDOR_HDR, IB_MGMT_MAD_DATA, + GFP_ATOMIC, OPA_MGMT_BASE_VERSION); + if (IS_ERR(send_buf)) { + c_err("%s:Couldn't allocate send buf\n", __func__); + goto err_sndbuf; + } + + send_buf->ah = ah; + + /* Set up common MAD hdr */ + trap_mad = send_buf->mad; + trap_mad->mad_hdr.base_version = OPA_MGMT_BASE_VERSION; + trap_mad->mad_hdr.mgmt_class = OPA_MGMT_CLASS_INTEL_EMA; + trap_mad->mad_hdr.class_version = OPA_EMA_CLASS_VERSION; + trap_mad->mad_hdr.method = IB_MGMT_METHOD_TRAP; + port->tid++; + trap_mad->mad_hdr.tid = cpu_to_be64(port->tid); + trap_mad->mad_hdr.attr_id = IB_SMP_ATTR_NOTICE; + + /* Set up vendor OUI */ + trap_mad->oui[0] = INTEL_OUI_1; + trap_mad->oui[1] = INTEL_OUI_2; + trap_mad->oui[2] = INTEL_OUI_3; + + /* Setup notice attribute portion */ + trap_mad->notice.gen_type = OPA_INTEL_EMA_NOTICE_TYPE_INFO << 1; + trap_mad->notice.oui_1 = INTEL_OUI_1; + trap_mad->notice.oui_2 = INTEL_OUI_2; + trap_mad->notice.oui_3 = INTEL_OUI_3; + trap_mad->notice.issuer_lid = cpu_to_be32(lid); + + /* copy the actual trap data */ + trap = (struct opa_veswport_trap *)trap_mad->notice.raw_data; + trap->fabric_id = cpu_to_be16(data->fabric_id); + trap->veswid = cpu_to_be16(data->veswid); + trap->veswportnum = cpu_to_be32(data->veswportnum); + trap->opaportnum = cpu_to_be16(data->opaportnum); + trap->veswportindex = data->veswportindex; + trap->opcode = data->opcode; + + /* If successful send set up rate limit timeout else bail */ + if (ib_post_send_mad(send_buf, NULL)) { + ib_free_send_mad(send_buf); + } else { + if (adapter->trap_count) + return; + adapter->trap_timeout = jiffies + + usecs_to_jiffies(OPA_VNIC_TRAP_TIMEOUT); + return; + } + +err_sndbuf: + rdma_destroy_ah(ah); +err_exit: + v_err("Aborting trap\n"); +} + +static int vema_rem_vport(int id, void *p, void *data) +{ + struct opa_vnic_adapter *adapter = p; + + opa_vnic_rem_netdev(adapter); + return 0; +} + +static int vema_enable_vport(int id, void *p, void *data) +{ + struct opa_vnic_adapter *adapter = p; + + netif_carrier_on(adapter->netdev); + return 0; +} + +static int vema_disable_vport(int id, void *p, void *data) +{ + struct opa_vnic_adapter *adapter = p; + + netif_carrier_off(adapter->netdev); + return 0; +} + +static void opa_vnic_event(struct ib_event_handler *handler, + struct ib_event *record) +{ + struct opa_vnic_vema_port *port = + container_of(handler, struct opa_vnic_vema_port, event_handler); + struct opa_vnic_ctrl_port *cport = port->cport; + + if (record->element.port_num != port->port_num) + return; + + c_dbg("OPA_VNIC received event %d on device %s port %d\n", + record->event, record->device->name, record->element.port_num); + + if (record->event == IB_EVENT_PORT_ERR) + idr_for_each(&port->vport_idr, vema_disable_vport, NULL); + if (record->event == IB_EVENT_PORT_ACTIVE) + idr_for_each(&port->vport_idr, vema_enable_vport, NULL); +} + +/** + * vema_unregister -- Unregisters agent + * @cport: pointer to control port + * + * This deletes the registration by VEMA for MADs + */ +static void vema_unregister(struct opa_vnic_ctrl_port *cport) +{ + int i; + + for (i = 1; i <= cport->num_ports; i++) { + struct opa_vnic_vema_port *port = vema_get_port(cport, i); + + if (!port->mad_agent) + continue; + + /* Lock ensures no MAD is being processed */ + mutex_lock(&port->lock); + idr_for_each(&port->vport_idr, vema_rem_vport, NULL); + mutex_unlock(&port->lock); + + ib_unregister_mad_agent(port->mad_agent); + port->mad_agent = NULL; + mutex_destroy(&port->lock); + idr_destroy(&port->vport_idr); + ib_unregister_event_handler(&port->event_handler); + } +} + +/** + * vema_register -- Registers agent + * @cport: pointer to control port + * + * This function registers the handlers for the VEMA MADs + * + * Return: returns 0 on success. non zero otherwise + */ +static int vema_register(struct opa_vnic_ctrl_port *cport) +{ + struct ib_mad_reg_req reg_req = { + .mgmt_class = OPA_MGMT_CLASS_INTEL_EMA, + .mgmt_class_version = OPA_MGMT_BASE_VERSION, + .oui = { INTEL_OUI_1, INTEL_OUI_2, INTEL_OUI_3 } + }; + int i; + + set_bit(IB_MGMT_METHOD_GET, reg_req.method_mask); + set_bit(IB_MGMT_METHOD_SET, reg_req.method_mask); + + /* register ib event handler and mad agent for each port on dev */ + for (i = 1; i <= cport->num_ports; i++) { + struct opa_vnic_vema_port *port = vema_get_port(cport, i); + int ret; + + port->cport = cport; + port->port_num = i; + + INIT_IB_EVENT_HANDLER(&port->event_handler, + cport->ibdev, opa_vnic_event); + ret = ib_register_event_handler(&port->event_handler); + if (ret) { + c_err("port %d: event handler register failed\n", i); + vema_unregister(cport); + return ret; + } + + idr_init(&port->vport_idr); + mutex_init(&port->lock); + port->mad_agent = ib_register_mad_agent(cport->ibdev, i, + IB_QPT_GSI, ®_req, + IB_MGMT_RMPP_VERSION, + vema_send, vema_recv, + port, 0); + if (IS_ERR(port->mad_agent)) { + ret = PTR_ERR(port->mad_agent); + port->mad_agent = NULL; + mutex_destroy(&port->lock); + idr_destroy(&port->vport_idr); + vema_unregister(cport); + return ret; + } + } + + return 0; +} + +/** + * opa_vnic_vema_add_one -- Handle new ib device + * @device: ib device pointer + * + * Allocate the vnic control port and initialize it. + */ +static void opa_vnic_vema_add_one(struct ib_device *device) +{ + struct opa_vnic_ctrl_port *cport; + int rc, size = sizeof(*cport); + + if (!rdma_cap_opa_vnic(device)) + return; + + size += device->phys_port_cnt * sizeof(struct opa_vnic_vema_port); + cport = kzalloc(size, GFP_KERNEL); + if (!cport) + return; + + cport->num_ports = device->phys_port_cnt; + cport->ibdev = device; + + /* Initialize opa vnic management agent (vema) */ + rc = vema_register(cport); + if (!rc) + c_info("VNIC client initialized\n"); + + ib_set_client_data(device, &opa_vnic_client, cport); +} + +/** + * opa_vnic_vema_rem_one -- Handle ib device removal + * @device: ib device pointer + * @client_data: ib client data + * + * Uninitialize and free the vnic control port. + */ +static void opa_vnic_vema_rem_one(struct ib_device *device, + void *client_data) +{ + struct opa_vnic_ctrl_port *cport = client_data; + + if (!cport) + return; + + c_info("removing VNIC client\n"); + vema_unregister(cport); + kfree(cport); +} + +static int __init opa_vnic_init(void) +{ + int rc; + + pr_info("OPA Virtual Network Driver - v%s\n", + opa_vnic_driver_version); + + rc = ib_register_client(&opa_vnic_client); + if (rc) + pr_err("VNIC driver register failed %d\n", rc); + + return rc; +} +module_init(opa_vnic_init); + +static void opa_vnic_deinit(void) +{ + ib_unregister_client(&opa_vnic_client); +} +module_exit(opa_vnic_deinit); + +MODULE_LICENSE("Dual BSD/GPL"); +MODULE_AUTHOR("Intel Corporation"); +MODULE_DESCRIPTION("Intel OPA Virtual Network driver"); +MODULE_VERSION(DRV_VERSION); diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_vema_iface.c b/drivers/infiniband/ulp/opa_vnic/opa_vnic_vema_iface.c new file mode 100644 index 000000000000..c2733964379c --- /dev/null +++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_vema_iface.c @@ -0,0 +1,390 @@ +/* + * Copyright(c) 2017 Intel Corporation. + * + * This file is provided under a dual BSD/GPLv2 license. When using or + * redistributing this file, you may do so under either license. + * + * GPL LICENSE SUMMARY + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * BSD LICENSE + * + * 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. + * - Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + */ + +/* + * This file contains OPA VNIC EMA Interface functions. + */ + +#include "opa_vnic_internal.h" + +/** + * opa_vnic_vema_report_event - sent trap to report the specified event + * @adapter: vnic port adapter + * @event: event to be reported + * + * This function calls vema api to sent a trap for the given event. + */ +void opa_vnic_vema_report_event(struct opa_vnic_adapter *adapter, u8 event) +{ + struct __opa_veswport_info *info = &adapter->info; + struct __opa_veswport_trap trap_data; + + trap_data.fabric_id = info->vesw.fabric_id; + trap_data.veswid = info->vesw.vesw_id; + trap_data.veswportnum = info->vport.port_num; + trap_data.opaportnum = adapter->port_num; + trap_data.veswportindex = adapter->vport_num; + trap_data.opcode = event; + + opa_vnic_vema_send_trap(adapter, &trap_data, info->vport.encap_slid); +} + +/** + * opa_vnic_get_error_counters - get summary counters + * @adapter: vnic port adapter + * @cntrs: pointer to destination summary counters structure + * + * This function populates the summary counters that is maintained by the + * given adapter to destination address provided. + */ +void opa_vnic_get_summary_counters(struct opa_vnic_adapter *adapter, + struct opa_veswport_summary_counters *cntrs) +{ + struct opa_vnic_stats vstats; + __be64 *dst; + u64 *src; + + memset(&vstats, 0, sizeof(vstats)); + spin_lock(&adapter->stats_lock); + adapter->rn_ops->ndo_get_stats64(adapter->netdev, &vstats.netstats); + spin_unlock(&adapter->stats_lock); + + cntrs->vp_instance = cpu_to_be16(adapter->vport_num); + cntrs->vesw_id = cpu_to_be16(adapter->info.vesw.vesw_id); + cntrs->veswport_num = cpu_to_be32(adapter->port_num); + + cntrs->tx_errors = cpu_to_be64(vstats.netstats.tx_errors); + cntrs->rx_errors = cpu_to_be64(vstats.netstats.rx_errors); + cntrs->tx_packets = cpu_to_be64(vstats.netstats.tx_packets); + cntrs->rx_packets = cpu_to_be64(vstats.netstats.rx_packets); + cntrs->tx_bytes = cpu_to_be64(vstats.netstats.tx_bytes); + cntrs->rx_bytes = cpu_to_be64(vstats.netstats.rx_bytes); + + /* + * This loop depends on layout of + * opa_veswport_summary_counters opa_vnic_stats structures. + */ + for (dst = &cntrs->tx_unicast, src = &vstats.tx_grp.unicast; + dst < &cntrs->reserved[0]; dst++, src++) { + *dst = cpu_to_be64(*src); + } +} + +/** + * opa_vnic_get_error_counters - get error counters + * @adapter: vnic port adapter + * @cntrs: pointer to destination error counters structure + * + * This function populates the error counters that is maintained by the + * given adapter to destination address provided. + */ +void opa_vnic_get_error_counters(struct opa_vnic_adapter *adapter, + struct opa_veswport_error_counters *cntrs) +{ + struct opa_vnic_stats vstats; + + memset(&vstats, 0, sizeof(vstats)); + spin_lock(&adapter->stats_lock); + adapter->rn_ops->ndo_get_stats64(adapter->netdev, &vstats.netstats); + spin_unlock(&adapter->stats_lock); + + cntrs->vp_instance = cpu_to_be16(adapter->vport_num); + cntrs->vesw_id = cpu_to_be16(adapter->info.vesw.vesw_id); + cntrs->veswport_num = cpu_to_be32(adapter->port_num); + + cntrs->tx_errors = cpu_to_be64(vstats.netstats.tx_errors); + cntrs->rx_errors = cpu_to_be64(vstats.netstats.rx_errors); + cntrs->tx_dlid_zero = cpu_to_be64(vstats.tx_dlid_zero); + cntrs->tx_drop_state = cpu_to_be64(vstats.tx_drop_state); + cntrs->tx_logic = cpu_to_be64(vstats.netstats.tx_fifo_errors + + vstats.netstats.tx_carrier_errors); + + cntrs->rx_bad_veswid = cpu_to_be64(vstats.netstats.rx_nohandler); + cntrs->rx_runt = cpu_to_be64(vstats.rx_runt); + cntrs->rx_oversize = cpu_to_be64(vstats.rx_oversize); + cntrs->rx_drop_state = cpu_to_be64(vstats.rx_drop_state); + cntrs->rx_logic = cpu_to_be64(vstats.netstats.rx_fifo_errors); +} + +/** + * opa_vnic_get_vesw_info -- Get the vesw information + * @adapter: vnic port adapter + * @info: pointer to destination vesw info structure + * + * This function copies the vesw info that is maintained by the + * given adapter to destination address provided. + */ +void opa_vnic_get_vesw_info(struct opa_vnic_adapter *adapter, + struct opa_vesw_info *info) +{ + struct __opa_vesw_info *src = &adapter->info.vesw; + int i; + + info->fabric_id = cpu_to_be16(src->fabric_id); + info->vesw_id = cpu_to_be16(src->vesw_id); + memcpy(info->rsvd0, src->rsvd0, ARRAY_SIZE(src->rsvd0)); + info->def_port_mask = cpu_to_be16(src->def_port_mask); + memcpy(info->rsvd1, src->rsvd1, ARRAY_SIZE(src->rsvd1)); + info->pkey = cpu_to_be16(src->pkey); + + memcpy(info->rsvd2, src->rsvd2, ARRAY_SIZE(src->rsvd2)); + info->u_mcast_dlid = cpu_to_be32(src->u_mcast_dlid); + for (i = 0; i < OPA_VESW_MAX_NUM_DEF_PORT; i++) + info->u_ucast_dlid[i] = cpu_to_be32(src->u_ucast_dlid[i]); + + memcpy(info->rsvd3, src->rsvd3, ARRAY_SIZE(src->rsvd3)); + for (i = 0; i < OPA_VNIC_MAX_NUM_PCP; i++) + info->eth_mtu[i] = cpu_to_be16(src->eth_mtu[i]); + + info->eth_mtu_non_vlan = cpu_to_be16(src->eth_mtu_non_vlan); + memcpy(info->rsvd4, src->rsvd4, ARRAY_SIZE(src->rsvd4)); +} + +/** + * opa_vnic_set_vesw_info -- Set the vesw information + * @adapter: vnic port adapter + * @info: pointer to vesw info structure + * + * This function updates the vesw info that is maintained by the + * given adapter with vesw info provided. Reserved fields are stored + * and returned back to EM as is. + */ +void opa_vnic_set_vesw_info(struct opa_vnic_adapter *adapter, + struct opa_vesw_info *info) +{ + struct __opa_vesw_info *dst = &adapter->info.vesw; + int i; + + dst->fabric_id = be16_to_cpu(info->fabric_id); + dst->vesw_id = be16_to_cpu(info->vesw_id); + memcpy(dst->rsvd0, info->rsvd0, ARRAY_SIZE(info->rsvd0)); + dst->def_port_mask = be16_to_cpu(info->def_port_mask); + memcpy(dst->rsvd1, info->rsvd1, ARRAY_SIZE(info->rsvd1)); + dst->pkey = be16_to_cpu(info->pkey); + + memcpy(dst->rsvd2, info->rsvd2, ARRAY_SIZE(info->rsvd2)); + dst->u_mcast_dlid = be32_to_cpu(info->u_mcast_dlid); + for (i = 0; i < OPA_VESW_MAX_NUM_DEF_PORT; i++) + dst->u_ucast_dlid[i] = be32_to_cpu(info->u_ucast_dlid[i]); + + memcpy(dst->rsvd3, info->rsvd3, ARRAY_SIZE(info->rsvd3)); + for (i = 0; i < OPA_VNIC_MAX_NUM_PCP; i++) + dst->eth_mtu[i] = be16_to_cpu(info->eth_mtu[i]); + + dst->eth_mtu_non_vlan = be16_to_cpu(info->eth_mtu_non_vlan); + memcpy(dst->rsvd4, info->rsvd4, ARRAY_SIZE(info->rsvd4)); +} + +/** + * opa_vnic_get_per_veswport_info -- Get the vesw per port information + * @adapter: vnic port adapter + * @info: pointer to destination vport info structure + * + * This function copies the vesw per port info that is maintained by the + * given adapter to destination address provided. + * Note that the read only fields are not copied. + */ +void opa_vnic_get_per_veswport_info(struct opa_vnic_adapter *adapter, + struct opa_per_veswport_info *info) +{ + struct __opa_per_veswport_info *src = &adapter->info.vport; + + info->port_num = cpu_to_be32(src->port_num); + info->eth_link_status = src->eth_link_status; + memcpy(info->rsvd0, src->rsvd0, ARRAY_SIZE(src->rsvd0)); + + memcpy(info->base_mac_addr, src->base_mac_addr, + ARRAY_SIZE(info->base_mac_addr)); + info->config_state = src->config_state; + info->oper_state = src->oper_state; + info->max_mac_tbl_ent = cpu_to_be16(src->max_mac_tbl_ent); + info->max_smac_ent = cpu_to_be16(src->max_smac_ent); + info->mac_tbl_digest = cpu_to_be32(src->mac_tbl_digest); + memcpy(info->rsvd1, src->rsvd1, ARRAY_SIZE(src->rsvd1)); + + info->encap_slid = cpu_to_be32(src->encap_slid); + memcpy(info->pcp_to_sc_uc, src->pcp_to_sc_uc, + ARRAY_SIZE(info->pcp_to_sc_uc)); + memcpy(info->pcp_to_vl_uc, src->pcp_to_vl_uc, + ARRAY_SIZE(info->pcp_to_vl_uc)); + memcpy(info->pcp_to_sc_mc, src->pcp_to_sc_mc, + ARRAY_SIZE(info->pcp_to_sc_mc)); + memcpy(info->pcp_to_vl_mc, src->pcp_to_vl_mc, + ARRAY_SIZE(info->pcp_to_vl_mc)); + info->non_vlan_sc_uc = src->non_vlan_sc_uc; + info->non_vlan_vl_uc = src->non_vlan_vl_uc; + info->non_vlan_sc_mc = src->non_vlan_sc_mc; + info->non_vlan_vl_mc = src->non_vlan_vl_mc; + memcpy(info->rsvd2, src->rsvd2, ARRAY_SIZE(src->rsvd2)); + + info->uc_macs_gen_count = cpu_to_be16(src->uc_macs_gen_count); + info->mc_macs_gen_count = cpu_to_be16(src->mc_macs_gen_count); + memcpy(info->rsvd3, src->rsvd3, ARRAY_SIZE(src->rsvd3)); +} + +/** + * opa_vnic_set_per_veswport_info -- Set vesw per port information + * @adapter: vnic port adapter + * @info: pointer to vport info structure + * + * This function updates the vesw per port info that is maintained by the + * given adapter with vesw per port info provided. Reserved fields are + * stored and returned back to EM as is. + */ +void opa_vnic_set_per_veswport_info(struct opa_vnic_adapter *adapter, + struct opa_per_veswport_info *info) +{ + struct __opa_per_veswport_info *dst = &adapter->info.vport; + + dst->port_num = be32_to_cpu(info->port_num); + memcpy(dst->rsvd0, info->rsvd0, ARRAY_SIZE(info->rsvd0)); + + memcpy(dst->base_mac_addr, info->base_mac_addr, + ARRAY_SIZE(dst->base_mac_addr)); + dst->config_state = info->config_state; + memcpy(dst->rsvd1, info->rsvd1, ARRAY_SIZE(info->rsvd1)); + + dst->encap_slid = be32_to_cpu(info->encap_slid); + memcpy(dst->pcp_to_sc_uc, info->pcp_to_sc_uc, + ARRAY_SIZE(dst->pcp_to_sc_uc)); + memcpy(dst->pcp_to_vl_uc, info->pcp_to_vl_uc, + ARRAY_SIZE(dst->pcp_to_vl_uc)); + memcpy(dst->pcp_to_sc_mc, info->pcp_to_sc_mc, + ARRAY_SIZE(dst->pcp_to_sc_mc)); + memcpy(dst->pcp_to_vl_mc, info->pcp_to_vl_mc, + ARRAY_SIZE(dst->pcp_to_vl_mc)); + dst->non_vlan_sc_uc = info->non_vlan_sc_uc; + dst->non_vlan_vl_uc = info->non_vlan_vl_uc; + dst->non_vlan_sc_mc = info->non_vlan_sc_mc; + dst->non_vlan_vl_mc = info->non_vlan_vl_mc; + memcpy(dst->rsvd2, info->rsvd2, ARRAY_SIZE(info->rsvd2)); + memcpy(dst->rsvd3, info->rsvd3, ARRAY_SIZE(info->rsvd3)); +} + +/** + * opa_vnic_query_mcast_macs - query multicast mac list + * @adapter: vnic port adapter + * @macs: pointer mac list + * + * This function populates the provided mac list with the configured + * multicast addresses in the adapter. + */ +void opa_vnic_query_mcast_macs(struct opa_vnic_adapter *adapter, + struct opa_veswport_iface_macs *macs) +{ + u16 start_idx, num_macs, idx = 0, count = 0; + struct netdev_hw_addr *ha; + + start_idx = be16_to_cpu(macs->start_idx); + num_macs = be16_to_cpu(macs->num_macs_in_msg); + netdev_for_each_mc_addr(ha, adapter->netdev) { + struct opa_vnic_iface_mac_entry *entry = &macs->entry[count]; + + if (start_idx > idx++) + continue; + else if (num_macs == count) + break; + memcpy(entry, ha->addr, sizeof(*entry)); + count++; + } + + macs->tot_macs_in_lst = cpu_to_be16(netdev_mc_count(adapter->netdev)); + macs->num_macs_in_msg = cpu_to_be16(count); + macs->gen_count = cpu_to_be16(adapter->info.vport.mc_macs_gen_count); +} + +/** + * opa_vnic_query_ucast_macs - query unicast mac list + * @adapter: vnic port adapter + * @macs: pointer mac list + * + * This function populates the provided mac list with the configured + * unicast addresses in the adapter. + */ +void opa_vnic_query_ucast_macs(struct opa_vnic_adapter *adapter, + struct opa_veswport_iface_macs *macs) +{ + u16 start_idx, tot_macs, num_macs, idx = 0, count = 0; + struct netdev_hw_addr *ha; + + start_idx = be16_to_cpu(macs->start_idx); + num_macs = be16_to_cpu(macs->num_macs_in_msg); + /* loop through dev_addrs list first */ + for_each_dev_addr(adapter->netdev, ha) { + struct opa_vnic_iface_mac_entry *entry = &macs->entry[count]; + + /* Do not include EM specified MAC address */ + if (!memcmp(adapter->info.vport.base_mac_addr, ha->addr, + ARRAY_SIZE(adapter->info.vport.base_mac_addr))) + continue; + + if (start_idx > idx++) + continue; + else if (num_macs == count) + break; + memcpy(entry, ha->addr, sizeof(*entry)); + count++; + } + + /* loop through uc list */ + netdev_for_each_uc_addr(ha, adapter->netdev) { + struct opa_vnic_iface_mac_entry *entry = &macs->entry[count]; + + if (start_idx > idx++) + continue; + else if (num_macs == count) + break; + memcpy(entry, ha->addr, sizeof(*entry)); + count++; + } + + tot_macs = netdev_hw_addr_list_count(&adapter->netdev->dev_addrs) + + netdev_uc_count(adapter->netdev); + macs->tot_macs_in_lst = cpu_to_be16(tot_macs); + macs->num_macs_in_msg = cpu_to_be16(count); + macs->gen_count = cpu_to_be16(adapter->info.vport.uc_macs_gen_count); +} diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c index 79bf48477ddb..2354c742caa1 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.c +++ b/drivers/infiniband/ulp/srp/ib_srp.c @@ -40,6 +40,7 @@ #include <linux/parser.h> #include <linux/random.h> #include <linux/jiffies.h> +#include <linux/lockdep.h> #include <rdma/ib_cache.h> #include <linux/atomic.h> @@ -311,6 +312,11 @@ static int srp_new_cm_id(struct srp_rdma_ch *ch) if (ch->cm_id) ib_destroy_cm_id(ch->cm_id); ch->cm_id = new_cm_id; + if (rdma_cap_opa_ah(target->srp_host->srp_dev->dev, + target->srp_host->port)) + ch->path.rec_type = SA_PATH_REC_TYPE_OPA; + else + ch->path.rec_type = SA_PATH_REC_TYPE_IB; ch->path.sgid = target->sgid; ch->path.dgid = target->orig_dgid; ch->path.pkey = target->pkey; @@ -371,7 +377,6 @@ static struct srp_fr_pool *srp_create_fr_pool(struct ib_device *device, struct srp_fr_desc *d; struct ib_mr *mr; int i, ret = -EINVAL; - enum ib_mr_type mr_type; if (pool_size <= 0) goto err; @@ -385,13 +390,9 @@ static struct srp_fr_pool *srp_create_fr_pool(struct ib_device *device, spin_lock_init(&pool->lock); INIT_LIST_HEAD(&pool->free_list); - if (device->attrs.device_cap_flags & IB_DEVICE_SG_GAPS_REG) - mr_type = IB_MR_TYPE_SG_GAPS; - else - mr_type = IB_MR_TYPE_MEM_REG; - for (i = 0, d = &pool->desc[0]; i < pool->size; i++, d++) { - mr = ib_alloc_mr(pd, mr_type, max_page_list_len); + mr = ib_alloc_mr(pd, IB_MR_TYPE_MEM_REG, + max_page_list_len); if (IS_ERR(mr)) { ret = PTR_ERR(mr); if (ret == -ENOMEM) @@ -470,9 +471,13 @@ static struct srp_fr_pool *srp_alloc_fr_pool(struct srp_target_port *target) * completion handler can access the queue pair while it is * being destroyed. */ -static void srp_destroy_qp(struct ib_qp *qp) +static void srp_destroy_qp(struct srp_rdma_ch *ch, struct ib_qp *qp) { - ib_drain_rq(qp); + spin_lock_irq(&ch->lock); + ib_process_cq_direct(ch->send_cq, -1); + spin_unlock_irq(&ch->lock); + + ib_drain_qp(qp); ib_destroy_qp(qp); } @@ -546,7 +551,7 @@ static int srp_create_ch_ib(struct srp_rdma_ch *ch) } if (ch->qp) - srp_destroy_qp(ch->qp); + srp_destroy_qp(ch, ch->qp); if (ch->recv_cq) ib_free_cq(ch->recv_cq); if (ch->send_cq) @@ -570,7 +575,7 @@ static int srp_create_ch_ib(struct srp_rdma_ch *ch) return 0; err_qp: - srp_destroy_qp(qp); + ib_destroy_qp(qp); err_send_cq: ib_free_cq(send_cq); @@ -613,7 +618,7 @@ static void srp_free_ch_ib(struct srp_target_port *target, ib_destroy_fmr_pool(ch->fmr_pool); } - srp_destroy_qp(ch->qp); + srp_destroy_qp(ch, ch->qp); ib_free_cq(ch->send_cq); ib_free_cq(ch->recv_cq); @@ -643,7 +648,7 @@ static void srp_free_ch_ib(struct srp_target_port *target, } static void srp_path_rec_completion(int status, - struct ib_sa_path_rec *pathrec, + struct sa_path_rec *pathrec, void *ch_ptr) { struct srp_rdma_ch *ch = ch_ptr; @@ -1804,6 +1809,8 @@ static struct srp_iu *__srp_get_tx_iu(struct srp_rdma_ch *ch, s32 rsv = (iu_type == SRP_IU_TSK_MGMT) ? 0 : SRP_TSK_MGMT_SQ_SIZE; struct srp_iu *iu; + lockdep_assert_held(&ch->lock); + ib_process_cq_direct(ch->send_cq, -1); if (list_empty(&ch->free_tx)) @@ -1824,6 +1831,11 @@ static struct srp_iu *__srp_get_tx_iu(struct srp_rdma_ch *ch, return iu; } +/* + * Note: if this function is called from inside ib_drain_sq() then it will + * be called without ch->lock being held. If ib_drain_sq() dequeues a WQE + * with status IB_WC_SUCCESS then that's a bug. + */ static void srp_send_done(struct ib_cq *cq, struct ib_wc *wc) { struct srp_iu *iu = container_of(wc->wr_cqe, struct srp_iu, cqe); @@ -1834,6 +1846,8 @@ static void srp_send_done(struct ib_cq *cq, struct ib_wc *wc) return; } + lockdep_assert_held(&ch->lock); + list_add(&iu->list, &ch->free_tx); } @@ -1889,17 +1903,24 @@ static void srp_process_rsp(struct srp_rdma_ch *ch, struct srp_rsp *rsp) if (unlikely(rsp->tag & SRP_TAG_TSK_MGMT)) { spin_lock_irqsave(&ch->lock, flags); ch->req_lim += be32_to_cpu(rsp->req_lim_delta); + if (rsp->tag == ch->tsk_mgmt_tag) { + ch->tsk_mgmt_status = -1; + if (be32_to_cpu(rsp->resp_data_len) >= 4) + ch->tsk_mgmt_status = rsp->data[3]; + complete(&ch->tsk_mgmt_done); + } else { + shost_printk(KERN_ERR, target->scsi_host, + "Received tsk mgmt response too late for tag %#llx\n", + rsp->tag); + } spin_unlock_irqrestore(&ch->lock, flags); - - ch->tsk_mgmt_status = -1; - if (be32_to_cpu(rsp->resp_data_len) >= 4) - ch->tsk_mgmt_status = rsp->data[3]; - complete(&ch->tsk_mgmt_done); } else { scmnd = scsi_host_find_tag(target->scsi_host, rsp->tag); - if (scmnd) { + if (scmnd && scmnd->host_scribble) { req = (void *)scmnd->host_scribble; scmnd = srp_claim_req(ch, req, NULL, scmnd); + } else { + scmnd = NULL; } if (!scmnd) { shost_printk(KERN_ERR, target->scsi_host, @@ -2383,12 +2404,12 @@ static void srp_cm_rej_handler(struct ib_cm_id *cm_id, switch (event->param.rej_rcvd.reason) { case IB_CM_REJ_PORT_CM_REDIRECT: cpi = event->param.rej_rcvd.ari; - ch->path.dlid = cpi->redirect_lid; + sa_path_set_dlid(&ch->path, htonl(ntohs(cpi->redirect_lid))); ch->path.pkey = cpi->redirect_pkey; cm_id->remote_cm_qpn = be32_to_cpu(cpi->redirect_qp) & 0x00ffffff; memcpy(ch->path.dgid.raw, cpi->redirect_gid, 16); - ch->status = ch->path.dlid ? + ch->status = sa_path_get_dlid(&ch->path) ? SRP_DLID_REDIRECT : SRP_PORT_REDIRECT; break; @@ -2531,19 +2552,18 @@ srp_change_queue_depth(struct scsi_device *sdev, int qdepth) } static int srp_send_tsk_mgmt(struct srp_rdma_ch *ch, u64 req_tag, u64 lun, - u8 func) + u8 func, u8 *status) { struct srp_target_port *target = ch->target; struct srp_rport *rport = target->rport; struct ib_device *dev = target->srp_host->srp_dev->dev; struct srp_iu *iu; struct srp_tsk_mgmt *tsk_mgmt; + int res; if (!ch->connected || target->qp_in_error) return -1; - init_completion(&ch->tsk_mgmt_done); - /* * Lock the rport mutex to avoid that srp_create_ch_ib() is * invoked while a task management function is being sent. @@ -2566,10 +2586,16 @@ static int srp_send_tsk_mgmt(struct srp_rdma_ch *ch, u64 req_tag, u64 lun, tsk_mgmt->opcode = SRP_TSK_MGMT; int_to_scsilun(lun, &tsk_mgmt->lun); - tsk_mgmt->tag = req_tag | SRP_TAG_TSK_MGMT; tsk_mgmt->tsk_mgmt_func = func; tsk_mgmt->task_tag = req_tag; + spin_lock_irq(&ch->lock); + ch->tsk_mgmt_tag = (ch->tsk_mgmt_tag + 1) | SRP_TAG_TSK_MGMT; + tsk_mgmt->tag = ch->tsk_mgmt_tag; + spin_unlock_irq(&ch->lock); + + init_completion(&ch->tsk_mgmt_done); + ib_dma_sync_single_for_device(dev, iu->dma, sizeof *tsk_mgmt, DMA_TO_DEVICE); if (srp_post_send(ch, iu, sizeof(*tsk_mgmt))) { @@ -2578,13 +2604,15 @@ static int srp_send_tsk_mgmt(struct srp_rdma_ch *ch, u64 req_tag, u64 lun, return -1; } + res = wait_for_completion_timeout(&ch->tsk_mgmt_done, + msecs_to_jiffies(SRP_ABORT_TIMEOUT_MS)); + if (res > 0 && status) + *status = ch->tsk_mgmt_status; mutex_unlock(&rport->mutex); - if (!wait_for_completion_timeout(&ch->tsk_mgmt_done, - msecs_to_jiffies(SRP_ABORT_TIMEOUT_MS))) - return -1; + WARN_ON_ONCE(res < 0); - return 0; + return res > 0 ? 0 : -1; } static int srp_abort(struct scsi_cmnd *scmnd) @@ -2610,7 +2638,7 @@ static int srp_abort(struct scsi_cmnd *scmnd) shost_printk(KERN_ERR, target->scsi_host, "Sending SRP abort for tag %#x\n", tag); if (srp_send_tsk_mgmt(ch, tag, scmnd->device->lun, - SRP_TSK_ABORT_TASK) == 0) + SRP_TSK_ABORT_TASK, NULL) == 0) ret = SUCCESS; else if (target->rport->state == SRP_RPORT_LOST) ret = FAST_IO_FAIL; @@ -2628,14 +2656,15 @@ static int srp_reset_device(struct scsi_cmnd *scmnd) struct srp_target_port *target = host_to_target(scmnd->device->host); struct srp_rdma_ch *ch; int i; + u8 status; shost_printk(KERN_ERR, target->scsi_host, "SRP reset_device called\n"); ch = &target->ch[0]; if (srp_send_tsk_mgmt(ch, SRP_TAG_NO_REQ, scmnd->device->lun, - SRP_TSK_LUN_RESET)) + SRP_TSK_LUN_RESET, &status)) return FAILED; - if (ch->tsk_mgmt_status) + if (status) return FAILED; for (i = 0; i < target->ch_count; i++) { @@ -2664,9 +2693,8 @@ static int srp_slave_alloc(struct scsi_device *sdev) struct Scsi_Host *shost = sdev->host; struct srp_target_port *target = host_to_target(shost); struct srp_device *srp_dev = target->srp_host->srp_dev; - struct ib_device *ibdev = srp_dev->dev; - if (!(ibdev->attrs.device_cap_flags & IB_DEVICE_SG_GAPS_REG)) + if (true) blk_queue_virt_boundary(sdev->request_queue, ~srp_dev->mr_page_mask); @@ -2869,6 +2897,7 @@ static struct scsi_host_template srp_template = { .info = srp_target_info, .queuecommand = srp_queuecommand, .change_queue_depth = srp_change_queue_depth, + .eh_timed_out = srp_timed_out, .eh_abort_handler = srp_abort, .eh_device_reset_handler = srp_reset_device, .eh_host_reset_handler = srp_reset_host, @@ -2909,7 +2938,7 @@ static int srp_add_target(struct srp_host *host, struct srp_target_port *target) sprintf(target->target_name, "SRP.T10:%016llX", be64_to_cpu(target->id_ext)); - if (scsi_add_host(target->scsi_host, host->srp_dev->dev->dma_device)) + if (scsi_add_host(target->scsi_host, host->srp_dev->dev->dev.parent)) return -ENODEV; memcpy(ids.port_id, &target->id_ext, 8); @@ -3421,11 +3450,12 @@ static ssize_t srp_create_target(struct device *dev, ret = srp_connect_ch(ch, multich); if (ret) { shost_printk(KERN_ERR, target->scsi_host, - PFX "Connection %d/%d failed\n", + PFX "Connection %d/%d to %pI6 failed\n", ch_start + cpu_idx, - target->ch_count); + target->ch_count, + ch->target->orig_dgid.raw); if (node_idx == 0 && cpu_idx == 0) { - goto err_disconnect; + goto free_ch; } else { srp_free_ch_ib(target, ch); srp_free_req_data(target, ch); @@ -3472,6 +3502,7 @@ put: err_disconnect: srp_disconnect_target(target); +free_ch: for (i = 0; i < target->ch_count; i++) { ch = &target->ch[i]; srp_free_ch_ib(target, ch); @@ -3520,7 +3551,7 @@ static struct srp_host *srp_add_port(struct srp_device *device, u8 port) host->port = port; host->dev.class = &srp_class; - host->dev.parent = device->dev->dma_device; + host->dev.parent = device->dev->dev.parent; dev_set_name(&host->dev, "srp-%s-%d", device->dev->name, port); if (device_register(&host->dev)) diff --git a/drivers/infiniband/ulp/srp/ib_srp.h b/drivers/infiniband/ulp/srp/ib_srp.h index 21c69695f9d4..ab9077b81d5a 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.h +++ b/drivers/infiniband/ulp/srp/ib_srp.h @@ -152,7 +152,7 @@ struct srp_rdma_ch { struct completion done; int status; - struct ib_sa_path_rec path; + struct sa_path_rec path; struct ib_sa_query *path_query; int path_query_id; @@ -163,6 +163,7 @@ struct srp_rdma_ch { int max_ti_iu_len; int comp_vector; + u64 tsk_mgmt_tag; struct completion tsk_mgmt_done; u8 tsk_mgmt_status; bool connected; diff --git a/drivers/infiniband/ulp/srpt/ib_srpt.c b/drivers/infiniband/ulp/srpt/ib_srpt.c index d21ba9d857c3..402275be0931 100644 --- a/drivers/infiniband/ulp/srpt/ib_srpt.c +++ b/drivers/infiniband/ulp/srpt/ib_srpt.c @@ -417,7 +417,7 @@ static void srpt_mgmt_method_get(struct srpt_port *sp, struct ib_mad *rq_mad, static void srpt_mad_send_handler(struct ib_mad_agent *mad_agent, struct ib_mad_send_wc *mad_wc) { - ib_destroy_ah(mad_wc->send_buf->ah); + rdma_destroy_ah(mad_wc->send_buf->ah); ib_free_send_mad(mad_wc->send_buf); } @@ -481,7 +481,7 @@ static void srpt_mad_recv_handler(struct ib_mad_agent *mad_agent, ib_free_send_mad(rsp); err_rsp: - ib_destroy_ah(ah); + rdma_destroy_ah(ah); err: ib_free_recv_mad(mad_wc); } @@ -500,6 +500,7 @@ static int srpt_refresh_port(struct srpt_port *sport) struct ib_mad_reg_req reg_req; struct ib_port_modify port_modify; struct ib_port_attr port_attr; + __be16 *guid; int ret; memset(&port_modify, 0, sizeof(port_modify)); @@ -522,10 +523,17 @@ static int srpt_refresh_port(struct srpt_port *sport) if (ret) goto err_query_port; + sport->port_guid_wwn.priv = sport; + guid = (__be16 *)&sport->gid.global.interface_id; snprintf(sport->port_guid, sizeof(sport->port_guid), - "0x%016llx%016llx", - be64_to_cpu(sport->gid.global.subnet_prefix), - be64_to_cpu(sport->gid.global.interface_id)); + "%04x:%04x:%04x:%04x", + be16_to_cpu(guid[0]), be16_to_cpu(guid[1]), + be16_to_cpu(guid[2]), be16_to_cpu(guid[3])); + sport->port_gid_wwn.priv = sport; + snprintf(sport->port_gid, sizeof(sport->port_gid), + "0x%016llx%016llx", + be64_to_cpu(sport->gid.global.subnet_prefix), + be64_to_cpu(sport->gid.global.interface_id)); if (!sport->mad_agent) { memset(®_req, 0, sizeof(reg_req)); @@ -1149,8 +1157,8 @@ static int srpt_abort_cmd(struct srpt_send_ioctx *ioctx) } spin_unlock_irqrestore(&ioctx->spinlock, flags); - pr_debug("Aborting cmd with state %d and tag %lld\n", state, - ioctx->cmd.tag); + pr_debug("Aborting cmd with state %d -> %d and tag %lld\n", state, + ioctx->state, ioctx->cmd.tag); switch (state) { case SRPT_STATE_NEW: @@ -1838,6 +1846,7 @@ static int srpt_cm_req_recv(struct ib_cm_id *cm_id, struct srp_login_rej *rej; struct ib_cm_rep_param *rep_param; struct srpt_rdma_ch *ch, *tmp_ch; + __be16 *guid; u32 it_iu_len; int i, ret = 0; @@ -1983,26 +1992,30 @@ static int srpt_cm_req_recv(struct ib_cm_id *cm_id, goto destroy_ib; } - /* - * Use the initator port identifier as the session name, when - * checking against se_node_acl->initiatorname[] this can be - * with or without preceeding '0x'. - */ + guid = (__be16 *)¶m->primary_path->sgid.global.interface_id; + snprintf(ch->ini_guid, sizeof(ch->ini_guid), "%04x:%04x:%04x:%04x", + be16_to_cpu(guid[0]), be16_to_cpu(guid[1]), + be16_to_cpu(guid[2]), be16_to_cpu(guid[3])); snprintf(ch->sess_name, sizeof(ch->sess_name), "0x%016llx%016llx", be64_to_cpu(*(__be64 *)ch->i_port_id), be64_to_cpu(*(__be64 *)(ch->i_port_id + 8))); pr_debug("registering session %s\n", ch->sess_name); - ch->sess = target_alloc_session(&sport->port_tpg_1, 0, 0, + if (sport->port_guid_tpg.se_tpg_wwn) + ch->sess = target_alloc_session(&sport->port_guid_tpg, 0, 0, + TARGET_PROT_NORMAL, + ch->ini_guid, ch, NULL); + if (sport->port_gid_tpg.se_tpg_wwn && IS_ERR_OR_NULL(ch->sess)) + ch->sess = target_alloc_session(&sport->port_gid_tpg, 0, 0, TARGET_PROT_NORMAL, ch->sess_name, ch, NULL); /* Retry without leading "0x" */ - if (IS_ERR(ch->sess)) - ch->sess = target_alloc_session(&sport->port_tpg_1, 0, 0, + if (sport->port_gid_tpg.se_tpg_wwn && IS_ERR_OR_NULL(ch->sess)) + ch->sess = target_alloc_session(&sport->port_gid_tpg, 0, 0, TARGET_PROT_NORMAL, ch->sess_name + 2, ch, NULL); - if (IS_ERR(ch->sess)) { + if (IS_ERR_OR_NULL(ch->sess)) { pr_info("Rejected login because no ACL has been configured yet for initiator %s.\n", ch->sess_name); rej->reason = cpu_to_be32((PTR_ERR(ch->sess) == -ENOMEM) ? @@ -2289,12 +2302,8 @@ static void srpt_queue_response(struct se_cmd *cmd) } spin_unlock_irqrestore(&ioctx->spinlock, flags); - if (unlikely(transport_check_aborted_status(&ioctx->cmd, false) - || WARN_ON_ONCE(state == SRPT_STATE_CMD_RSP_SENT))) { - atomic_inc(&ch->req_lim_delta); - srpt_abort_cmd(ioctx); + if (unlikely(WARN_ON_ONCE(state == SRPT_STATE_CMD_RSP_SENT))) return; - } /* For read commands, transfer the data to the initiator. */ if (ioctx->cmd.data_direction == DMA_FROM_DEVICE && @@ -2420,7 +2429,7 @@ static int srpt_release_sdev(struct srpt_device *sdev) return 0; } -static struct srpt_port *__srpt_lookup_port(const char *name) +static struct se_wwn *__srpt_lookup_wwn(const char *name) { struct ib_device *dev; struct srpt_device *sdev; @@ -2435,23 +2444,25 @@ static struct srpt_port *__srpt_lookup_port(const char *name) for (i = 0; i < dev->phys_port_cnt; i++) { sport = &sdev->port[i]; - if (!strcmp(sport->port_guid, name)) - return sport; + if (strcmp(sport->port_guid, name) == 0) + return &sport->port_guid_wwn; + if (strcmp(sport->port_gid, name) == 0) + return &sport->port_gid_wwn; } } return NULL; } -static struct srpt_port *srpt_lookup_port(const char *name) +static struct se_wwn *srpt_lookup_wwn(const char *name) { - struct srpt_port *sport; + struct se_wwn *wwn; spin_lock(&srpt_dev_lock); - sport = __srpt_lookup_port(name); + wwn = __srpt_lookup_wwn(name); spin_unlock(&srpt_dev_lock); - return sport; + return wwn; } /** @@ -2464,8 +2475,7 @@ static void srpt_add_one(struct ib_device *device) struct ib_srq_init_attr srq_attr; int i; - pr_debug("device = %p, device->dma_ops = %p\n", device, - device->dma_ops); + pr_debug("device = %p\n", device); sdev = kzalloc(sizeof(*sdev), GFP_KERNEL); if (!sdev) @@ -2643,11 +2653,19 @@ static char *srpt_get_fabric_name(void) return "srpt"; } +static struct srpt_port *srpt_tpg_to_sport(struct se_portal_group *tpg) +{ + return tpg->se_tpg_wwn->priv; +} + static char *srpt_get_fabric_wwn(struct se_portal_group *tpg) { - struct srpt_port *sport = container_of(tpg, struct srpt_port, port_tpg_1); + struct srpt_port *sport = srpt_tpg_to_sport(tpg); - return sport->port_guid; + WARN_ON_ONCE(tpg != &sport->port_guid_tpg && + tpg != &sport->port_gid_tpg); + return tpg == &sport->port_guid_tpg ? sport->port_guid : + sport->port_gid; } static u16 srpt_get_tag(struct se_portal_group *tpg) @@ -2667,7 +2685,8 @@ static void srpt_release_cmd(struct se_cmd *se_cmd) struct srpt_rdma_ch *ch = ioctx->ch; unsigned long flags; - WARN_ON(ioctx->state != SRPT_STATE_DONE); + WARN_ON_ONCE(ioctx->state != SRPT_STATE_DONE && + !(ioctx->cmd.transport_state & CMD_T_ABORTED)); if (ioctx->n_rw_ctx) { srpt_free_rw_ctxs(ch, ioctx); @@ -2737,6 +2756,19 @@ static int srpt_get_tcm_cmd_state(struct se_cmd *se_cmd) return srpt_get_cmd_state(ioctx); } +static int srpt_parse_guid(u64 *guid, const char *name) +{ + u16 w[4]; + int ret = -EINVAL; + + if (sscanf(name, "%hx:%hx:%hx:%hx", &w[0], &w[1], &w[2], &w[3]) != 4) + goto out; + *guid = get_unaligned_be64(w); + ret = 0; +out: + return ret; +} + /** * srpt_parse_i_port_id() - Parse an initiator port ID. * @name: ASCII representation of a 128-bit initiator port ID. @@ -2772,20 +2804,23 @@ out: */ static int srpt_init_nodeacl(struct se_node_acl *se_nacl, const char *name) { + u64 guid; u8 i_port_id[16]; + int ret; - if (srpt_parse_i_port_id(i_port_id, name) < 0) { + ret = srpt_parse_guid(&guid, name); + if (ret < 0) + ret = srpt_parse_i_port_id(i_port_id, name); + if (ret < 0) pr_err("invalid initiator port ID %s\n", name); - return -EINVAL; - } - return 0; + return ret; } static ssize_t srpt_tpg_attrib_srp_max_rdma_size_show(struct config_item *item, char *page) { struct se_portal_group *se_tpg = attrib_to_tpg(item); - struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + struct srpt_port *sport = srpt_tpg_to_sport(se_tpg); return sprintf(page, "%u\n", sport->port_attrib.srp_max_rdma_size); } @@ -2794,7 +2829,7 @@ static ssize_t srpt_tpg_attrib_srp_max_rdma_size_store(struct config_item *item, const char *page, size_t count) { struct se_portal_group *se_tpg = attrib_to_tpg(item); - struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + struct srpt_port *sport = srpt_tpg_to_sport(se_tpg); unsigned long val; int ret; @@ -2822,7 +2857,7 @@ static ssize_t srpt_tpg_attrib_srp_max_rsp_size_show(struct config_item *item, char *page) { struct se_portal_group *se_tpg = attrib_to_tpg(item); - struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + struct srpt_port *sport = srpt_tpg_to_sport(se_tpg); return sprintf(page, "%u\n", sport->port_attrib.srp_max_rsp_size); } @@ -2831,7 +2866,7 @@ static ssize_t srpt_tpg_attrib_srp_max_rsp_size_store(struct config_item *item, const char *page, size_t count) { struct se_portal_group *se_tpg = attrib_to_tpg(item); - struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + struct srpt_port *sport = srpt_tpg_to_sport(se_tpg); unsigned long val; int ret; @@ -2859,7 +2894,7 @@ static ssize_t srpt_tpg_attrib_srp_sq_size_show(struct config_item *item, char *page) { struct se_portal_group *se_tpg = attrib_to_tpg(item); - struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + struct srpt_port *sport = srpt_tpg_to_sport(se_tpg); return sprintf(page, "%u\n", sport->port_attrib.srp_sq_size); } @@ -2868,7 +2903,7 @@ static ssize_t srpt_tpg_attrib_srp_sq_size_store(struct config_item *item, const char *page, size_t count) { struct se_portal_group *se_tpg = attrib_to_tpg(item); - struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + struct srpt_port *sport = srpt_tpg_to_sport(se_tpg); unsigned long val; int ret; @@ -2906,7 +2941,7 @@ static struct configfs_attribute *srpt_tpg_attrib_attrs[] = { static ssize_t srpt_tpg_enable_show(struct config_item *item, char *page) { struct se_portal_group *se_tpg = to_tpg(item); - struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + struct srpt_port *sport = srpt_tpg_to_sport(se_tpg); return snprintf(page, PAGE_SIZE, "%d\n", (sport->enabled) ? 1: 0); } @@ -2915,7 +2950,7 @@ static ssize_t srpt_tpg_enable_store(struct config_item *item, const char *page, size_t count) { struct se_portal_group *se_tpg = to_tpg(item); - struct srpt_port *sport = container_of(se_tpg, struct srpt_port, port_tpg_1); + struct srpt_port *sport = srpt_tpg_to_sport(se_tpg); struct srpt_device *sdev = sport->sdev; struct srpt_rdma_ch *ch; unsigned long tmp; @@ -2967,15 +3002,19 @@ static struct se_portal_group *srpt_make_tpg(struct se_wwn *wwn, struct config_group *group, const char *name) { - struct srpt_port *sport = container_of(wwn, struct srpt_port, port_wwn); + struct srpt_port *sport = wwn->priv; + static struct se_portal_group *tpg; int res; - /* Initialize sport->port_wwn and sport->port_tpg_1 */ - res = core_tpg_register(&sport->port_wwn, &sport->port_tpg_1, SCSI_PROTOCOL_SRP); + WARN_ON_ONCE(wwn != &sport->port_guid_wwn && + wwn != &sport->port_gid_wwn); + tpg = wwn == &sport->port_guid_wwn ? &sport->port_guid_tpg : + &sport->port_gid_tpg; + res = core_tpg_register(wwn, tpg, SCSI_PROTOCOL_SRP); if (res) return ERR_PTR(res); - return &sport->port_tpg_1; + return tpg; } /** @@ -2984,11 +3023,10 @@ static struct se_portal_group *srpt_make_tpg(struct se_wwn *wwn, */ static void srpt_drop_tpg(struct se_portal_group *tpg) { - struct srpt_port *sport = container_of(tpg, - struct srpt_port, port_tpg_1); + struct srpt_port *sport = srpt_tpg_to_sport(tpg); sport->enabled = false; - core_tpg_deregister(&sport->port_tpg_1); + core_tpg_deregister(tpg); } /** @@ -2999,19 +3037,7 @@ static struct se_wwn *srpt_make_tport(struct target_fabric_configfs *tf, struct config_group *group, const char *name) { - struct srpt_port *sport; - int ret; - - sport = srpt_lookup_port(name); - pr_debug("make_tport(%s)\n", name); - ret = -EINVAL; - if (!sport) - goto err; - - return &sport->port_wwn; - -err: - return ERR_PTR(ret); + return srpt_lookup_wwn(name) ? : ERR_PTR(-EINVAL); } /** @@ -3020,9 +3046,6 @@ err: */ static void srpt_drop_tport(struct se_wwn *wwn) { - struct srpt_port *sport = container_of(wwn, struct srpt_port, port_wwn); - - pr_debug("drop_tport(%s\n", config_item_name(&sport->port_wwn.wwn_group.cg_item)); } static ssize_t srpt_wwn_version_show(struct config_item *item, char *buf) diff --git a/drivers/infiniband/ulp/srpt/ib_srpt.h b/drivers/infiniband/ulp/srpt/ib_srpt.h index 581878782854..cc1183851af5 100644 --- a/drivers/infiniband/ulp/srpt/ib_srpt.h +++ b/drivers/infiniband/ulp/srpt/ib_srpt.h @@ -258,6 +258,7 @@ enum rdma_ch_state { * against concurrent modification by the cm_id spinlock. * @sess: Session information associated with this SRP channel. * @sess_name: Session name. + * @ini_guid: Initiator port GUID. * @release_work: Allows scheduling of srpt_release_channel(). * @release_done: Enables waiting for srpt_release_channel() completion. */ @@ -284,6 +285,7 @@ struct srpt_rdma_ch { struct list_head cmd_wait_list; struct se_session *sess; u8 sess_name[36]; + u8 ini_guid[24]; struct work_struct release_work; struct completion *release_done; }; @@ -306,28 +308,34 @@ struct srpt_port_attrib { * @mad_agent: per-port management datagram processing information. * @enabled: Whether or not this target port is enabled. * @port_guid: ASCII representation of Port GUID + * @port_gid: ASCII representation of Port GID * @port: one-based port number. * @sm_lid: cached value of the port's sm_lid. * @lid: cached value of the port's lid. * @gid: cached value of the port's gid. * @port_acl_lock spinlock for port_acl_list: * @work: work structure for refreshing the aforementioned cached values. - * @port_tpg_1 Target portal group = 1 data. - * @port_wwn: Target core WWN data. + * @port_guid_tpg: TPG associated with target port GUID. + * @port_guid_wwn: WWN associated with target port GUID. + * @port_gid_tpg: TPG associated with target port GID. + * @port_gid_wwn: WWN associated with target port GID. * @port_acl_list: Head of the list with all node ACLs for this port. */ struct srpt_port { struct srpt_device *sdev; struct ib_mad_agent *mad_agent; bool enabled; - u8 port_guid[64]; + u8 port_guid[24]; + u8 port_gid[64]; u8 port; u16 sm_lid; u16 lid; union ib_gid gid; struct work_struct work; - struct se_portal_group port_tpg_1; - struct se_wwn port_wwn; + struct se_portal_group port_guid_tpg; + struct se_wwn port_guid_wwn; + struct se_portal_group port_gid_tpg; + struct se_wwn port_gid_wwn; struct srpt_port_attrib port_attrib; }; |
