diff options
Diffstat (limited to 'net/smc')
| -rw-r--r-- | net/smc/af_smc.c | 210 | ||||
| -rw-r--r-- | net/smc/smc.h | 20 | ||||
| -rw-r--r-- | net/smc/smc_core.c | 5 | ||||
| -rw-r--r-- | net/smc/smc_diag.c | 2 | ||||
| -rw-r--r-- | net/smc/smc_pnet.c | 50 | ||||
| -rw-r--r-- | net/smc/smc_pnet.h | 2 |
6 files changed, 231 insertions, 58 deletions
diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c index 961854e56736..284befa90967 100644 --- a/net/smc/af_smc.c +++ b/net/smc/af_smc.c @@ -183,7 +183,7 @@ static int smc_release(struct socket *sock) { struct sock *sk = sock->sk; struct smc_sock *smc; - int rc = 0; + int old_state, rc = 0; if (!sk) goto out; @@ -191,8 +191,10 @@ static int smc_release(struct socket *sock) sock_hold(sk); /* sock_put below */ smc = smc_sk(sk); + old_state = sk->sk_state; + /* cleanup for a dangling non-blocking connect */ - if (smc->connect_nonblock && sk->sk_state == SMC_INIT) + if (smc->connect_nonblock && old_state == SMC_INIT) tcp_abort(smc->clcsock->sk, ECONNABORTED); if (cancel_work_sync(&smc->connect_work)) @@ -206,6 +208,10 @@ static int smc_release(struct socket *sock) else lock_sock(sk); + if (old_state == SMC_INIT && sk->sk_state == SMC_ACTIVE && + !smc->use_fallback) + smc_close_active_abort(smc); + rc = __smc_release(smc); /* detach socket */ @@ -566,12 +572,118 @@ static void smc_stat_fallback(struct smc_sock *smc) mutex_unlock(&net->smc.mutex_fback_rsn); } -static void smc_switch_to_fallback(struct smc_sock *smc, int reason_code) +/* must be called under rcu read lock */ +static void smc_fback_wakeup_waitqueue(struct smc_sock *smc, void *key) { - wait_queue_head_t *smc_wait = sk_sleep(&smc->sk); - wait_queue_head_t *clc_wait = sk_sleep(smc->clcsock->sk); - unsigned long flags; + struct socket_wq *wq; + __poll_t flags; + + wq = rcu_dereference(smc->sk.sk_wq); + if (!skwq_has_sleeper(wq)) + return; + /* wake up smc sk->sk_wq */ + if (!key) { + /* sk_state_change */ + wake_up_interruptible_all(&wq->wait); + } else { + flags = key_to_poll(key); + if (flags & (EPOLLIN | EPOLLOUT)) + /* sk_data_ready or sk_write_space */ + wake_up_interruptible_sync_poll(&wq->wait, flags); + else if (flags & EPOLLERR) + /* sk_error_report */ + wake_up_interruptible_poll(&wq->wait, flags); + } +} + +static int smc_fback_mark_woken(wait_queue_entry_t *wait, + unsigned int mode, int sync, void *key) +{ + struct smc_mark_woken *mark = + container_of(wait, struct smc_mark_woken, wait_entry); + + mark->woken = true; + mark->key = key; + return 0; +} + +static void smc_fback_forward_wakeup(struct smc_sock *smc, struct sock *clcsk, + void (*clcsock_callback)(struct sock *sk)) +{ + struct smc_mark_woken mark = { .woken = false }; + struct socket_wq *wq; + + init_waitqueue_func_entry(&mark.wait_entry, + smc_fback_mark_woken); + rcu_read_lock(); + wq = rcu_dereference(clcsk->sk_wq); + if (!wq) + goto out; + add_wait_queue(sk_sleep(clcsk), &mark.wait_entry); + clcsock_callback(clcsk); + remove_wait_queue(sk_sleep(clcsk), &mark.wait_entry); + + if (mark.woken) + smc_fback_wakeup_waitqueue(smc, mark.key); +out: + rcu_read_unlock(); +} + +static void smc_fback_state_change(struct sock *clcsk) +{ + struct smc_sock *smc = + smc_clcsock_user_data(clcsk); + + if (!smc) + return; + smc_fback_forward_wakeup(smc, clcsk, smc->clcsk_state_change); +} + +static void smc_fback_data_ready(struct sock *clcsk) +{ + struct smc_sock *smc = + smc_clcsock_user_data(clcsk); + + if (!smc) + return; + smc_fback_forward_wakeup(smc, clcsk, smc->clcsk_data_ready); +} + +static void smc_fback_write_space(struct sock *clcsk) +{ + struct smc_sock *smc = + smc_clcsock_user_data(clcsk); + + if (!smc) + return; + smc_fback_forward_wakeup(smc, clcsk, smc->clcsk_write_space); +} + +static void smc_fback_error_report(struct sock *clcsk) +{ + struct smc_sock *smc = + smc_clcsock_user_data(clcsk); + + if (!smc) + return; + smc_fback_forward_wakeup(smc, clcsk, smc->clcsk_error_report); +} + +static int smc_switch_to_fallback(struct smc_sock *smc, int reason_code) +{ + struct sock *clcsk; + int rc = 0; + + mutex_lock(&smc->clcsock_release_lock); + if (!smc->clcsock) { + rc = -EBADF; + goto out; + } + clcsk = smc->clcsock->sk; + + if (smc->use_fallback) + goto out; smc->use_fallback = true; smc->fallback_rsn = reason_code; smc_stat_fallback(smc); @@ -582,22 +694,41 @@ static void smc_switch_to_fallback(struct smc_sock *smc, int reason_code) smc->clcsock->wq.fasync_list = smc->sk.sk_socket->wq.fasync_list; - /* There may be some entries remaining in - * smc socket->wq, which should be removed - * to clcsocket->wq during the fallback. + /* There might be some wait entries remaining + * in smc sk->sk_wq and they should be woken up + * as clcsock's wait queue is woken up. */ - spin_lock_irqsave(&smc_wait->lock, flags); - spin_lock_nested(&clc_wait->lock, SINGLE_DEPTH_NESTING); - list_splice_init(&smc_wait->head, &clc_wait->head); - spin_unlock(&clc_wait->lock); - spin_unlock_irqrestore(&smc_wait->lock, flags); + smc->clcsk_state_change = clcsk->sk_state_change; + smc->clcsk_data_ready = clcsk->sk_data_ready; + smc->clcsk_write_space = clcsk->sk_write_space; + smc->clcsk_error_report = clcsk->sk_error_report; + + clcsk->sk_state_change = smc_fback_state_change; + clcsk->sk_data_ready = smc_fback_data_ready; + clcsk->sk_write_space = smc_fback_write_space; + clcsk->sk_error_report = smc_fback_error_report; + + smc->clcsock->sk->sk_user_data = + (void *)((uintptr_t)smc | SK_USER_DATA_NOCOPY); } +out: + mutex_unlock(&smc->clcsock_release_lock); + return rc; } /* fall back during connect */ static int smc_connect_fallback(struct smc_sock *smc, int reason_code) { - smc_switch_to_fallback(smc, reason_code); + struct net *net = sock_net(&smc->sk); + int rc = 0; + + rc = smc_switch_to_fallback(smc, reason_code); + if (rc) { /* fallback fails */ + this_cpu_inc(net->smc.smc_stats->clnt_hshake_err_cnt); + if (smc->sk.sk_state == SMC_INIT) + sock_put(&smc->sk); /* passive closing */ + return rc; + } smc_copy_sock_settings_to_clc(smc); smc->connect_nonblock = 0; if (smc->sk.sk_state == SMC_INIT) @@ -1518,11 +1649,12 @@ static void smc_listen_decline(struct smc_sock *new_smc, int reason_code, { /* RDMA setup failed, switch back to TCP */ smc_conn_abort(new_smc, local_first); - if (reason_code < 0) { /* error, no fallback possible */ + if (reason_code < 0 || + smc_switch_to_fallback(new_smc, reason_code)) { + /* error, no fallback possible */ smc_listen_out_err(new_smc); return; } - smc_switch_to_fallback(new_smc, reason_code); if (reason_code && reason_code != SMC_CLC_DECL_PEERDECL) { if (smc_clc_send_decline(new_smc, reason_code, version) < 0) { smc_listen_out_err(new_smc); @@ -1964,8 +2096,11 @@ static void smc_listen_work(struct work_struct *work) /* check if peer is smc capable */ if (!tcp_sk(newclcsock->sk)->syn_smc) { - smc_switch_to_fallback(new_smc, SMC_CLC_DECL_PEERNOSMC); - smc_listen_out_connected(new_smc); + rc = smc_switch_to_fallback(new_smc, SMC_CLC_DECL_PEERNOSMC); + if (rc) + smc_listen_out_err(new_smc); + else + smc_listen_out_connected(new_smc); return; } @@ -2094,10 +2229,9 @@ out: static void smc_clcsock_data_ready(struct sock *listen_clcsock) { - struct smc_sock *lsmc; + struct smc_sock *lsmc = + smc_clcsock_user_data(listen_clcsock); - lsmc = (struct smc_sock *) - ((uintptr_t)listen_clcsock->sk_user_data & ~SK_USER_DATA_NOCOPY); if (!lsmc) return; lsmc->clcsk_data_ready(listen_clcsock); @@ -2254,7 +2388,9 @@ static int smc_sendmsg(struct socket *sock, struct msghdr *msg, size_t len) if (msg->msg_flags & MSG_FASTOPEN) { if (sk->sk_state == SMC_INIT && !smc->connect_nonblock) { - smc_switch_to_fallback(smc, SMC_CLC_DECL_OPTUNSUPP); + rc = smc_switch_to_fallback(smc, SMC_CLC_DECL_OPTUNSUPP); + if (rc) + goto out; } else { rc = -EINVAL; goto out; @@ -2447,6 +2583,11 @@ static int smc_setsockopt(struct socket *sock, int level, int optname, /* generic setsockopts reaching us here always apply to the * CLC socket */ + mutex_lock(&smc->clcsock_release_lock); + if (!smc->clcsock) { + mutex_unlock(&smc->clcsock_release_lock); + return -EBADF; + } if (unlikely(!smc->clcsock->ops->setsockopt)) rc = -EOPNOTSUPP; else @@ -2456,6 +2597,7 @@ static int smc_setsockopt(struct socket *sock, int level, int optname, sk->sk_err = smc->clcsock->sk->sk_err; sk_error_report(sk); } + mutex_unlock(&smc->clcsock_release_lock); if (optlen < sizeof(int)) return -EINVAL; @@ -2472,7 +2614,7 @@ static int smc_setsockopt(struct socket *sock, int level, int optname, case TCP_FASTOPEN_NO_COOKIE: /* option not supported by SMC */ if (sk->sk_state == SMC_INIT && !smc->connect_nonblock) { - smc_switch_to_fallback(smc, SMC_CLC_DECL_OPTUNSUPP); + rc = smc_switch_to_fallback(smc, SMC_CLC_DECL_OPTUNSUPP); } else { rc = -EINVAL; } @@ -2515,13 +2657,23 @@ static int smc_getsockopt(struct socket *sock, int level, int optname, char __user *optval, int __user *optlen) { struct smc_sock *smc; + int rc; smc = smc_sk(sock->sk); + mutex_lock(&smc->clcsock_release_lock); + if (!smc->clcsock) { + mutex_unlock(&smc->clcsock_release_lock); + return -EBADF; + } /* socket options apply to the CLC socket */ - if (unlikely(!smc->clcsock->ops->getsockopt)) + if (unlikely(!smc->clcsock->ops->getsockopt)) { + mutex_unlock(&smc->clcsock_release_lock); return -EOPNOTSUPP; - return smc->clcsock->ops->getsockopt(smc->clcsock, level, optname, - optval, optlen); + } + rc = smc->clcsock->ops->getsockopt(smc->clcsock, level, optname, + optval, optlen); + mutex_unlock(&smc->clcsock_release_lock); + return rc; } static int smc_ioctl(struct socket *sock, unsigned int cmd, @@ -2935,12 +3087,14 @@ static int __init smc_init(void) rc = tcp_register_ulp(&smc_ulp_ops); if (rc) { pr_err("%s: tcp_ulp_register fails with %d\n", __func__, rc); - goto out_sock; + goto out_ib; } static_branch_enable(&tcp_have_smc); return 0; +out_ib: + smc_ib_unregister_client(); out_sock: sock_unregister(PF_SMC); out_proto6: diff --git a/net/smc/smc.h b/net/smc/smc.h index 3d0b8e300deb..37b2001a0255 100644 --- a/net/smc/smc.h +++ b/net/smc/smc.h @@ -139,6 +139,12 @@ enum smc_urg_state { SMC_URG_READ = 3, /* data was already read */ }; +struct smc_mark_woken { + bool woken; + void *key; + wait_queue_entry_t wait_entry; +}; + struct smc_connection { struct rb_node alert_node; struct smc_link_group *lgr; /* link group of connection */ @@ -228,8 +234,14 @@ struct smc_connection { struct smc_sock { /* smc sock container */ struct sock sk; struct socket *clcsock; /* internal tcp socket */ + void (*clcsk_state_change)(struct sock *sk); + /* original stat_change fct. */ void (*clcsk_data_ready)(struct sock *sk); - /* original data_ready fct. **/ + /* original data_ready fct. */ + void (*clcsk_write_space)(struct sock *sk); + /* original write_space fct. */ + void (*clcsk_error_report)(struct sock *sk); + /* original error_report fct. */ struct smc_connection conn; /* smc connection */ struct smc_sock *listen_smc; /* listen parent */ struct work_struct connect_work; /* handle non-blocking connect*/ @@ -264,6 +276,12 @@ static inline struct smc_sock *smc_sk(const struct sock *sk) return (struct smc_sock *)sk; } +static inline struct smc_sock *smc_clcsock_user_data(struct sock *clcsk) +{ + return (struct smc_sock *) + ((uintptr_t)clcsk->sk_user_data & ~SK_USER_DATA_NOCOPY); +} + extern struct workqueue_struct *smc_hs_wq; /* wq for handshake work */ extern struct workqueue_struct *smc_close_wq; /* wq for close work */ diff --git a/net/smc/smc_core.c b/net/smc/smc_core.c index 29525d03b253..be7d704976ff 100644 --- a/net/smc/smc_core.c +++ b/net/smc/smc_core.c @@ -1161,8 +1161,8 @@ void smc_conn_free(struct smc_connection *conn) cancel_work_sync(&conn->abort_work); } if (!list_empty(&lgr->list)) { - smc_lgr_unregister_conn(conn); smc_buf_unuse(conn, lgr); /* allow buffer reuse */ + smc_lgr_unregister_conn(conn); } if (!lgr->conns_num) @@ -1864,7 +1864,8 @@ int smc_conn_create(struct smc_sock *smc, struct smc_init_info *ini) (ini->smcd_version == SMC_V2 || lgr->vlan_id == ini->vlan_id) && (role == SMC_CLNT || ini->is_smcd || - lgr->conns_num < SMC_RMBS_PER_LGR_MAX)) { + (lgr->conns_num < SMC_RMBS_PER_LGR_MAX && + !bitmap_full(lgr->rtokens_used_mask, SMC_RMBS_PER_LGR_MAX)))) { /* link group found */ ini->first_contact_local = 0; conn->lgr = lgr; diff --git a/net/smc/smc_diag.c b/net/smc/smc_diag.c index b8898c787d23..1fca2f90a9c7 100644 --- a/net/smc/smc_diag.c +++ b/net/smc/smc_diag.c @@ -146,13 +146,11 @@ static int __smc_diag_dump(struct sock *sk, struct sk_buff *skb, (req->diag_ext & (1 << (SMC_DIAG_LGRINFO - 1))) && !list_empty(&smc->conn.lgr->list)) { struct smc_link *link = smc->conn.lnk; - struct net *net = read_pnet(&link->smcibdev->ibdev->coredev.rdma_net); struct smc_diag_lgrinfo linfo = { .role = smc->conn.lgr->role, .lnk[0].ibport = link->ibport, .lnk[0].link_id = link->link_id, - .lnk[0].net_cookie = net->net_cookie, }; memcpy(linfo.lnk[0].ibname, diff --git a/net/smc/smc_pnet.c b/net/smc/smc_pnet.c index 291f1484a1b7..29f0a559d884 100644 --- a/net/smc/smc_pnet.c +++ b/net/smc/smc_pnet.c @@ -113,7 +113,7 @@ static int smc_pnet_remove_by_pnetid(struct net *net, char *pnet_name) pnettable = &sn->pnettable; /* remove table entry */ - write_lock(&pnettable->lock); + mutex_lock(&pnettable->lock); list_for_each_entry_safe(pnetelem, tmp_pe, &pnettable->pnetlist, list) { if (!pnet_name || @@ -131,7 +131,7 @@ static int smc_pnet_remove_by_pnetid(struct net *net, char *pnet_name) rc = 0; } } - write_unlock(&pnettable->lock); + mutex_unlock(&pnettable->lock); /* if this is not the initial namespace, stop here */ if (net != &init_net) @@ -192,7 +192,7 @@ static int smc_pnet_add_by_ndev(struct net_device *ndev) sn = net_generic(net, smc_net_id); pnettable = &sn->pnettable; - write_lock(&pnettable->lock); + mutex_lock(&pnettable->lock); list_for_each_entry_safe(pnetelem, tmp_pe, &pnettable->pnetlist, list) { if (pnetelem->type == SMC_PNET_ETH && !pnetelem->ndev && !strncmp(pnetelem->eth_name, ndev->name, IFNAMSIZ)) { @@ -206,7 +206,7 @@ static int smc_pnet_add_by_ndev(struct net_device *ndev) break; } } - write_unlock(&pnettable->lock); + mutex_unlock(&pnettable->lock); return rc; } @@ -224,7 +224,7 @@ static int smc_pnet_remove_by_ndev(struct net_device *ndev) sn = net_generic(net, smc_net_id); pnettable = &sn->pnettable; - write_lock(&pnettable->lock); + mutex_lock(&pnettable->lock); list_for_each_entry_safe(pnetelem, tmp_pe, &pnettable->pnetlist, list) { if (pnetelem->type == SMC_PNET_ETH && pnetelem->ndev == ndev) { dev_put_track(pnetelem->ndev, &pnetelem->dev_tracker); @@ -237,7 +237,7 @@ static int smc_pnet_remove_by_ndev(struct net_device *ndev) break; } } - write_unlock(&pnettable->lock); + mutex_unlock(&pnettable->lock); return rc; } @@ -368,12 +368,9 @@ static int smc_pnet_add_eth(struct smc_pnettable *pnettable, struct net *net, new_pe->type = SMC_PNET_ETH; memcpy(new_pe->pnet_name, pnet_name, SMC_MAX_PNETID_LEN); strncpy(new_pe->eth_name, eth_name, IFNAMSIZ); - new_pe->ndev = ndev; - if (ndev) - netdev_tracker_alloc(ndev, &new_pe->dev_tracker, GFP_KERNEL); rc = -EEXIST; new_netdev = true; - write_lock(&pnettable->lock); + mutex_lock(&pnettable->lock); list_for_each_entry(tmp_pe, &pnettable->pnetlist, list) { if (tmp_pe->type == SMC_PNET_ETH && !strncmp(tmp_pe->eth_name, eth_name, IFNAMSIZ)) { @@ -382,10 +379,15 @@ static int smc_pnet_add_eth(struct smc_pnettable *pnettable, struct net *net, } } if (new_netdev) { + if (ndev) { + new_pe->ndev = ndev; + netdev_tracker_alloc(ndev, &new_pe->dev_tracker, + GFP_ATOMIC); + } list_add_tail(&new_pe->list, &pnettable->pnetlist); - write_unlock(&pnettable->lock); + mutex_unlock(&pnettable->lock); } else { - write_unlock(&pnettable->lock); + mutex_unlock(&pnettable->lock); kfree(new_pe); goto out_put; } @@ -446,7 +448,7 @@ static int smc_pnet_add_ib(struct smc_pnettable *pnettable, char *ib_name, new_pe->ib_port = ib_port; new_ibdev = true; - write_lock(&pnettable->lock); + mutex_lock(&pnettable->lock); list_for_each_entry(tmp_pe, &pnettable->pnetlist, list) { if (tmp_pe->type == SMC_PNET_IB && !strncmp(tmp_pe->ib_name, ib_name, IB_DEVICE_NAME_MAX)) { @@ -456,9 +458,9 @@ static int smc_pnet_add_ib(struct smc_pnettable *pnettable, char *ib_name, } if (new_ibdev) { list_add_tail(&new_pe->list, &pnettable->pnetlist); - write_unlock(&pnettable->lock); + mutex_unlock(&pnettable->lock); } else { - write_unlock(&pnettable->lock); + mutex_unlock(&pnettable->lock); kfree(new_pe); } return (new_ibdev) ? 0 : -EEXIST; @@ -603,7 +605,7 @@ static int _smc_pnet_dump(struct net *net, struct sk_buff *skb, u32 portid, pnettable = &sn->pnettable; /* dump pnettable entries */ - read_lock(&pnettable->lock); + mutex_lock(&pnettable->lock); list_for_each_entry(pnetelem, &pnettable->pnetlist, list) { if (pnetid && !smc_pnet_match(pnetelem->pnet_name, pnetid)) continue; @@ -618,7 +620,7 @@ static int _smc_pnet_dump(struct net *net, struct sk_buff *skb, u32 portid, break; } } - read_unlock(&pnettable->lock); + mutex_unlock(&pnettable->lock); return idx; } @@ -862,7 +864,7 @@ int smc_pnet_net_init(struct net *net) struct smc_pnetids_ndev *pnetids_ndev = &sn->pnetids_ndev; INIT_LIST_HEAD(&pnettable->pnetlist); - rwlock_init(&pnettable->lock); + mutex_init(&pnettable->lock); INIT_LIST_HEAD(&pnetids_ndev->list); rwlock_init(&pnetids_ndev->lock); @@ -942,7 +944,7 @@ static int smc_pnet_find_ndev_pnetid_by_table(struct net_device *ndev, sn = net_generic(net, smc_net_id); pnettable = &sn->pnettable; - read_lock(&pnettable->lock); + mutex_lock(&pnettable->lock); list_for_each_entry(pnetelem, &pnettable->pnetlist, list) { if (pnetelem->type == SMC_PNET_ETH && ndev == pnetelem->ndev) { /* get pnetid of netdev device */ @@ -951,7 +953,7 @@ static int smc_pnet_find_ndev_pnetid_by_table(struct net_device *ndev, break; } } - read_unlock(&pnettable->lock); + mutex_unlock(&pnettable->lock); return rc; } @@ -1154,7 +1156,7 @@ int smc_pnetid_by_table_ib(struct smc_ib_device *smcibdev, u8 ib_port) sn = net_generic(&init_net, smc_net_id); pnettable = &sn->pnettable; - read_lock(&pnettable->lock); + mutex_lock(&pnettable->lock); list_for_each_entry(tmp_pe, &pnettable->pnetlist, list) { if (tmp_pe->type == SMC_PNET_IB && !strncmp(tmp_pe->ib_name, ib_name, IB_DEVICE_NAME_MAX) && @@ -1164,7 +1166,7 @@ int smc_pnetid_by_table_ib(struct smc_ib_device *smcibdev, u8 ib_port) break; } } - read_unlock(&pnettable->lock); + mutex_unlock(&pnettable->lock); return rc; } @@ -1183,7 +1185,7 @@ int smc_pnetid_by_table_smcd(struct smcd_dev *smcddev) sn = net_generic(&init_net, smc_net_id); pnettable = &sn->pnettable; - read_lock(&pnettable->lock); + mutex_lock(&pnettable->lock); list_for_each_entry(tmp_pe, &pnettable->pnetlist, list) { if (tmp_pe->type == SMC_PNET_IB && !strncmp(tmp_pe->ib_name, ib_name, IB_DEVICE_NAME_MAX)) { @@ -1192,7 +1194,7 @@ int smc_pnetid_by_table_smcd(struct smcd_dev *smcddev) break; } } - read_unlock(&pnettable->lock); + mutex_unlock(&pnettable->lock); return rc; } diff --git a/net/smc/smc_pnet.h b/net/smc/smc_pnet.h index 14039272f7e4..80a88eea4949 100644 --- a/net/smc/smc_pnet.h +++ b/net/smc/smc_pnet.h @@ -29,7 +29,7 @@ struct smc_link_group; * @pnetlist: List of PNETIDs */ struct smc_pnettable { - rwlock_t lock; + struct mutex lock; struct list_head pnetlist; }; |