diff options
Diffstat (limited to 'net/l2tp/l2tp_core.c')
-rw-r--r-- | net/l2tp/l2tp_core.c | 827 |
1 files changed, 517 insertions, 310 deletions
diff --git a/net/l2tp/l2tp_core.c b/net/l2tp/l2tp_core.c index 8d21ff25f160..3eec23ac5ab1 100644 --- a/net/l2tp/l2tp_core.c +++ b/net/l2tp/l2tp_core.c @@ -39,7 +39,6 @@ #include <linux/ip.h> #include <linux/udp.h> #include <linux/l2tp.h> -#include <linux/hash.h> #include <linux/sort.h> #include <linux/file.h> #include <linux/nsproxy.h> @@ -61,7 +60,6 @@ #include <linux/atomic.h> #include "l2tp_core.h" -#include "trace.h" #define CREATE_TRACE_POINTS #include "trace.h" @@ -88,6 +86,11 @@ /* Default trace flags */ #define L2TP_DEFAULT_DEBUG_FLAGS 0 +#define L2TP_DEPTH_NESTING 2 +#if L2TP_DEPTH_NESTING == SINGLE_DEPTH_NESTING +#error "L2TP requires its own lockdep subclass" +#endif + /* Private data stored for received packets in the skb. */ struct l2tp_skb_cb { @@ -107,99 +110,106 @@ struct l2tp_net { /* Lock for write access to l2tp_tunnel_idr */ spinlock_t l2tp_tunnel_idr_lock; struct idr l2tp_tunnel_idr; - struct hlist_head l2tp_session_hlist[L2TP_HASH_SIZE_2]; - /* Lock for write access to l2tp_session_hlist */ - spinlock_t l2tp_session_hlist_lock; + /* Lock for write access to l2tp_v[23]_session_idr/htable */ + spinlock_t l2tp_session_idr_lock; + struct idr l2tp_v2_session_idr; + struct idr l2tp_v3_session_idr; + struct hlist_head l2tp_v3_session_htable[16]; }; -#if IS_ENABLED(CONFIG_IPV6) -static bool l2tp_sk_is_v6(struct sock *sk) +static u32 l2tp_v2_session_key(u16 tunnel_id, u16 session_id) { - return sk->sk_family == PF_INET6 && - !ipv6_addr_v4mapped(&sk->sk_v6_daddr); + return ((u32)tunnel_id) << 16 | session_id; } -#endif -static inline struct l2tp_net *l2tp_pernet(const struct net *net) +static unsigned long l2tp_v3_session_hashkey(struct sock *sk, u32 session_id) { - return net_generic(net, l2tp_net_id); + return ((unsigned long)sk) + session_id; } -/* Session hash global list for L2TPv3. - * The session_id SHOULD be random according to RFC3931, but several - * L2TP implementations use incrementing session_ids. So we do a real - * hash on the session_id, rather than a simple bitmask. - */ -static inline struct hlist_head * -l2tp_session_id_hash_2(struct l2tp_net *pn, u32 session_id) +#if IS_ENABLED(CONFIG_IPV6) +static bool l2tp_sk_is_v6(struct sock *sk) { - return &pn->l2tp_session_hlist[hash_32(session_id, L2TP_HASH_BITS_2)]; + return sk->sk_family == PF_INET6 && + !ipv6_addr_v4mapped(&sk->sk_v6_daddr); } +#endif -/* Session hash list. - * The session_id SHOULD be random according to RFC2661, but several - * L2TP implementations (Cisco and Microsoft) use incrementing - * session_ids. So we do a real hash on the session_id, rather than a - * simple bitmask. - */ -static inline struct hlist_head * -l2tp_session_id_hash(struct l2tp_tunnel *tunnel, u32 session_id) +static struct l2tp_net *l2tp_pernet(const struct net *net) { - return &tunnel->session_hlist[hash_32(session_id, L2TP_HASH_BITS)]; + return net_generic(net, l2tp_net_id); } static void l2tp_tunnel_free(struct l2tp_tunnel *tunnel) { + struct sock *sk = tunnel->sock; + trace_free_tunnel(tunnel); - sock_put(tunnel->sock); - /* the tunnel is freed in the socket destructor */ + + if (sk) { + /* Disable udp encapsulation */ + switch (tunnel->encap) { + case L2TP_ENCAPTYPE_UDP: + /* No longer an encapsulation socket. See net/ipv4/udp.c */ + WRITE_ONCE(udp_sk(sk)->encap_type, 0); + udp_sk(sk)->encap_rcv = NULL; + udp_sk(sk)->encap_destroy = NULL; + break; + case L2TP_ENCAPTYPE_IP: + break; + } + + tunnel->sock = NULL; + sock_put(sk); + } + + kfree_rcu(tunnel, rcu); } static void l2tp_session_free(struct l2tp_session *session) { trace_free_session(session); if (session->tunnel) - l2tp_tunnel_dec_refcount(session->tunnel); - kfree(session); + l2tp_tunnel_put(session->tunnel); + kfree_rcu(session, rcu); } -struct l2tp_tunnel *l2tp_sk_to_tunnel(struct sock *sk) +struct l2tp_tunnel *l2tp_sk_to_tunnel(const struct sock *sk) { - struct l2tp_tunnel *tunnel = sk->sk_user_data; + const struct net *net = sock_net(sk); + unsigned long tunnel_id, tmp; + struct l2tp_tunnel *tunnel; + struct l2tp_net *pn; - if (tunnel) - if (WARN_ON(tunnel->magic != L2TP_TUNNEL_MAGIC)) - return NULL; + rcu_read_lock_bh(); + pn = l2tp_pernet(net); + idr_for_each_entry_ul(&pn->l2tp_tunnel_idr, tunnel, tmp, tunnel_id) { + if (tunnel && + tunnel->sock == sk && + refcount_inc_not_zero(&tunnel->ref_count)) { + rcu_read_unlock_bh(); + return tunnel; + } + } + rcu_read_unlock_bh(); - return tunnel; + return NULL; } EXPORT_SYMBOL_GPL(l2tp_sk_to_tunnel); -void l2tp_tunnel_inc_refcount(struct l2tp_tunnel *tunnel) -{ - refcount_inc(&tunnel->ref_count); -} -EXPORT_SYMBOL_GPL(l2tp_tunnel_inc_refcount); - -void l2tp_tunnel_dec_refcount(struct l2tp_tunnel *tunnel) +void l2tp_tunnel_put(struct l2tp_tunnel *tunnel) { if (refcount_dec_and_test(&tunnel->ref_count)) l2tp_tunnel_free(tunnel); } -EXPORT_SYMBOL_GPL(l2tp_tunnel_dec_refcount); +EXPORT_SYMBOL_GPL(l2tp_tunnel_put); -void l2tp_session_inc_refcount(struct l2tp_session *session) -{ - refcount_inc(&session->ref_count); -} -EXPORT_SYMBOL_GPL(l2tp_session_inc_refcount); - -void l2tp_session_dec_refcount(struct l2tp_session *session) +void l2tp_session_put(struct l2tp_session *session) { if (refcount_dec_and_test(&session->ref_count)) l2tp_session_free(session); } -EXPORT_SYMBOL_GPL(l2tp_session_dec_refcount); +EXPORT_SYMBOL_GPL(l2tp_session_put); /* Lookup a tunnel. A new reference is held on the returned tunnel. */ struct l2tp_tunnel *l2tp_tunnel_get(const struct net *net, u32 tunnel_id) @@ -219,92 +229,202 @@ struct l2tp_tunnel *l2tp_tunnel_get(const struct net *net, u32 tunnel_id) } EXPORT_SYMBOL_GPL(l2tp_tunnel_get); -struct l2tp_tunnel *l2tp_tunnel_get_nth(const struct net *net, int nth) +struct l2tp_tunnel *l2tp_tunnel_get_next(const struct net *net, unsigned long *key) { struct l2tp_net *pn = l2tp_pernet(net); - unsigned long tunnel_id, tmp; - struct l2tp_tunnel *tunnel; - int count = 0; + struct l2tp_tunnel *tunnel = NULL; rcu_read_lock_bh(); - idr_for_each_entry_ul(&pn->l2tp_tunnel_idr, tunnel, tmp, tunnel_id) { - if (tunnel && ++count > nth && - refcount_inc_not_zero(&tunnel->ref_count)) { +again: + tunnel = idr_get_next_ul(&pn->l2tp_tunnel_idr, key); + if (tunnel) { + if (refcount_inc_not_zero(&tunnel->ref_count)) { rcu_read_unlock_bh(); return tunnel; } + (*key)++; + goto again; } rcu_read_unlock_bh(); return NULL; } -EXPORT_SYMBOL_GPL(l2tp_tunnel_get_nth); +EXPORT_SYMBOL_GPL(l2tp_tunnel_get_next); -struct l2tp_session *l2tp_tunnel_get_session(struct l2tp_tunnel *tunnel, - u32 session_id) +struct l2tp_session *l2tp_v3_session_get(const struct net *net, struct sock *sk, u32 session_id) { - struct hlist_head *session_list; + const struct l2tp_net *pn = l2tp_pernet(net); struct l2tp_session *session; - session_list = l2tp_session_id_hash(tunnel, session_id); - rcu_read_lock_bh(); - hlist_for_each_entry_rcu(session, session_list, hlist) - if (session->session_id == session_id) { - l2tp_session_inc_refcount(session); - rcu_read_unlock_bh(); + session = idr_find(&pn->l2tp_v3_session_idr, session_id); + if (session && !hash_hashed(&session->hlist) && + refcount_inc_not_zero(&session->ref_count)) { + rcu_read_unlock_bh(); + return session; + } - return session; + /* If we get here and session is non-NULL, the session_id + * collides with one in another tunnel. If sk is non-NULL, + * find the session matching sk. + */ + if (session && sk) { + unsigned long key = l2tp_v3_session_hashkey(sk, session->session_id); + + hash_for_each_possible_rcu(pn->l2tp_v3_session_htable, session, + hlist, key) { + /* session->tunnel may be NULL if another thread is in + * l2tp_session_register and has added an item to + * l2tp_v3_session_htable but hasn't yet added the + * session to its tunnel's session_list. + */ + struct l2tp_tunnel *tunnel = READ_ONCE(session->tunnel); + + if (session->session_id == session_id && + tunnel && tunnel->sock == sk && + refcount_inc_not_zero(&session->ref_count)) { + rcu_read_unlock_bh(); + return session; + } } + } rcu_read_unlock_bh(); return NULL; } -EXPORT_SYMBOL_GPL(l2tp_tunnel_get_session); +EXPORT_SYMBOL_GPL(l2tp_v3_session_get); -struct l2tp_session *l2tp_session_get(const struct net *net, u32 session_id) +struct l2tp_session *l2tp_v2_session_get(const struct net *net, u16 tunnel_id, u16 session_id) { - struct hlist_head *session_list; + u32 session_key = l2tp_v2_session_key(tunnel_id, session_id); + const struct l2tp_net *pn = l2tp_pernet(net); struct l2tp_session *session; - session_list = l2tp_session_id_hash_2(l2tp_pernet(net), session_id); + rcu_read_lock_bh(); + session = idr_find(&pn->l2tp_v2_session_idr, session_key); + if (session && refcount_inc_not_zero(&session->ref_count)) { + rcu_read_unlock_bh(); + return session; + } + rcu_read_unlock_bh(); + + return NULL; +} +EXPORT_SYMBOL_GPL(l2tp_v2_session_get); + +struct l2tp_session *l2tp_session_get(const struct net *net, struct sock *sk, int pver, + u32 tunnel_id, u32 session_id) +{ + if (pver == L2TP_HDR_VER_2) + return l2tp_v2_session_get(net, tunnel_id, session_id); + else + return l2tp_v3_session_get(net, sk, session_id); +} +EXPORT_SYMBOL_GPL(l2tp_session_get); + +static struct l2tp_session *l2tp_v2_session_get_next(const struct net *net, + u16 tid, + unsigned long *key) +{ + struct l2tp_net *pn = l2tp_pernet(net); + struct l2tp_session *session = NULL; + + /* Start searching within the range of the tid */ + if (*key == 0) + *key = l2tp_v2_session_key(tid, 0); rcu_read_lock_bh(); - hlist_for_each_entry_rcu(session, session_list, global_hlist) - if (session->session_id == session_id) { - l2tp_session_inc_refcount(session); - rcu_read_unlock_bh(); +again: + session = idr_get_next_ul(&pn->l2tp_v2_session_idr, key); + if (session) { + struct l2tp_tunnel *tunnel = READ_ONCE(session->tunnel); + /* ignore sessions with id 0 as they are internal for pppol2tp */ + if (session->session_id == 0) { + (*key)++; + goto again; + } + + if (tunnel->tunnel_id == tid && + refcount_inc_not_zero(&session->ref_count)) { + rcu_read_unlock_bh(); return session; } + + (*key)++; + if (tunnel->tunnel_id == tid) + goto again; + } rcu_read_unlock_bh(); return NULL; } -EXPORT_SYMBOL_GPL(l2tp_session_get); -struct l2tp_session *l2tp_session_get_nth(struct l2tp_tunnel *tunnel, int nth) +static struct l2tp_session *l2tp_v3_session_get_next(const struct net *net, + u32 tid, struct sock *sk, + unsigned long *key) { - int hash; - struct l2tp_session *session; - int count = 0; + struct l2tp_net *pn = l2tp_pernet(net); + struct l2tp_session *session = NULL; rcu_read_lock_bh(); - for (hash = 0; hash < L2TP_HASH_SIZE; hash++) { - hlist_for_each_entry_rcu(session, &tunnel->session_hlist[hash], hlist) { - if (++count > nth) { - l2tp_session_inc_refcount(session); +again: + session = idr_get_next_ul(&pn->l2tp_v3_session_idr, key); + if (session && !hash_hashed(&session->hlist)) { + struct l2tp_tunnel *tunnel = READ_ONCE(session->tunnel); + + if (tunnel && tunnel->tunnel_id == tid && + refcount_inc_not_zero(&session->ref_count)) { + rcu_read_unlock_bh(); + return session; + } + + (*key)++; + goto again; + } + + /* If we get here and session is non-NULL, the IDR entry may be one + * where the session_id collides with one in another tunnel. Check + * session_htable for a match. There can only be one session of a given + * ID per tunnel so we can return as soon as a match is found. + */ + if (session && hash_hashed(&session->hlist)) { + unsigned long hkey = l2tp_v3_session_hashkey(sk, session->session_id); + u32 sid = session->session_id; + + hash_for_each_possible_rcu(pn->l2tp_v3_session_htable, session, + hlist, hkey) { + struct l2tp_tunnel *tunnel = READ_ONCE(session->tunnel); + + if (session->session_id == sid && + tunnel && tunnel->tunnel_id == tid && + refcount_inc_not_zero(&session->ref_count)) { rcu_read_unlock_bh(); return session; } } + + /* If no match found, the colliding session ID isn't in our + * tunnel so try the next session ID. + */ + (*key)++; + goto again; } rcu_read_unlock_bh(); return NULL; } -EXPORT_SYMBOL_GPL(l2tp_session_get_nth); + +struct l2tp_session *l2tp_session_get_next(const struct net *net, struct sock *sk, int pver, + u32 tunnel_id, unsigned long *key) +{ + if (pver == L2TP_HDR_VER_2) + return l2tp_v2_session_get_next(net, tunnel_id, key); + else + return l2tp_v3_session_get_next(net, tunnel_id, sk, key); +} +EXPORT_SYMBOL_GPL(l2tp_session_get_next); /* Lookup a session by interface name. * This is very inefficient but is only used by management interfaces. @@ -313,86 +433,190 @@ struct l2tp_session *l2tp_session_get_by_ifname(const struct net *net, const char *ifname) { struct l2tp_net *pn = l2tp_pernet(net); - int hash; + unsigned long tunnel_id, tmp; struct l2tp_session *session; + struct l2tp_tunnel *tunnel; rcu_read_lock_bh(); - for (hash = 0; hash < L2TP_HASH_SIZE_2; hash++) { - hlist_for_each_entry_rcu(session, &pn->l2tp_session_hlist[hash], global_hlist) { - if (!strcmp(session->ifname, ifname)) { - l2tp_session_inc_refcount(session); - rcu_read_unlock_bh(); - - return session; + idr_for_each_entry_ul(&pn->l2tp_tunnel_idr, tunnel, tmp, tunnel_id) { + if (tunnel) { + list_for_each_entry_rcu(session, &tunnel->session_list, list) { + if (!strcmp(session->ifname, ifname)) { + refcount_inc(&session->ref_count); + rcu_read_unlock_bh(); + + return session; + } } } } - rcu_read_unlock_bh(); return NULL; } EXPORT_SYMBOL_GPL(l2tp_session_get_by_ifname); +static void l2tp_session_coll_list_add(struct l2tp_session_coll_list *clist, + struct l2tp_session *session) +{ + refcount_inc(&session->ref_count); + WARN_ON_ONCE(session->coll_list); + session->coll_list = clist; + spin_lock(&clist->lock); + list_add(&session->clist, &clist->list); + spin_unlock(&clist->lock); +} + +static int l2tp_session_collision_add(struct l2tp_net *pn, + struct l2tp_session *session1, + struct l2tp_session *session2) +{ + struct l2tp_session_coll_list *clist; + + lockdep_assert_held(&pn->l2tp_session_idr_lock); + + if (!session2) + return -EEXIST; + + /* If existing session is in IP-encap tunnel, refuse new session */ + if (session2->tunnel->encap == L2TP_ENCAPTYPE_IP) + return -EEXIST; + + clist = session2->coll_list; + if (!clist) { + /* First collision. Allocate list to manage the collided sessions + * and add the existing session to the list. + */ + clist = kmalloc(sizeof(*clist), GFP_ATOMIC); + if (!clist) + return -ENOMEM; + + spin_lock_init(&clist->lock); + INIT_LIST_HEAD(&clist->list); + refcount_set(&clist->ref_count, 1); + l2tp_session_coll_list_add(clist, session2); + } + + /* If existing session isn't already in the session hlist, add it. */ + if (!hash_hashed(&session2->hlist)) + hash_add_rcu(pn->l2tp_v3_session_htable, &session2->hlist, + session2->hlist_key); + + /* Add new session to the hlist and collision list */ + hash_add_rcu(pn->l2tp_v3_session_htable, &session1->hlist, + session1->hlist_key); + refcount_inc(&clist->ref_count); + l2tp_session_coll_list_add(clist, session1); + + return 0; +} + +static void l2tp_session_collision_del(struct l2tp_net *pn, + struct l2tp_session *session) +{ + struct l2tp_session_coll_list *clist = session->coll_list; + unsigned long session_key = session->session_id; + struct l2tp_session *session2; + + lockdep_assert_held(&pn->l2tp_session_idr_lock); + + hash_del_rcu(&session->hlist); + + if (clist) { + /* Remove session from its collision list. If there + * are other sessions with the same ID, replace this + * session's IDR entry with that session, otherwise + * remove the IDR entry. If this is the last session, + * the collision list data is freed. + */ + spin_lock(&clist->lock); + list_del_init(&session->clist); + session2 = list_first_entry_or_null(&clist->list, struct l2tp_session, clist); + if (session2) { + void *old = idr_replace(&pn->l2tp_v3_session_idr, session2, session_key); + + WARN_ON_ONCE(IS_ERR_VALUE(old)); + } else { + void *removed = idr_remove(&pn->l2tp_v3_session_idr, session_key); + + WARN_ON_ONCE(removed != session); + } + session->coll_list = NULL; + spin_unlock(&clist->lock); + if (refcount_dec_and_test(&clist->ref_count)) + kfree(clist); + l2tp_session_put(session); + } +} + int l2tp_session_register(struct l2tp_session *session, struct l2tp_tunnel *tunnel) { - struct l2tp_session *session_walk; - struct hlist_head *g_head; - struct hlist_head *head; - struct l2tp_net *pn; + struct l2tp_net *pn = l2tp_pernet(tunnel->l2tp_net); + struct l2tp_session *other_session = NULL; + void *old = NULL; + u32 session_key; int err; - head = l2tp_session_id_hash(tunnel, session->session_id); + spin_lock_bh(&tunnel->list_lock); + spin_lock_bh(&pn->l2tp_session_idr_lock); - spin_lock_bh(&tunnel->hlist_lock); if (!tunnel->acpt_newsess) { err = -ENODEV; - goto err_tlock; + goto out; } - hlist_for_each_entry(session_walk, head, hlist) - if (session_walk->session_id == session->session_id) { - err = -EEXIST; - goto err_tlock; - } - if (tunnel->version == L2TP_HDR_VER_3) { - pn = l2tp_pernet(tunnel->l2tp_net); - g_head = l2tp_session_id_hash_2(pn, session->session_id); - - spin_lock_bh(&pn->l2tp_session_hlist_lock); - + session_key = session->session_id; + err = idr_alloc_u32(&pn->l2tp_v3_session_idr, NULL, + &session_key, session_key, GFP_ATOMIC); /* IP encap expects session IDs to be globally unique, while - * UDP encap doesn't. + * UDP encap doesn't. This isn't per the RFC, which says that + * sessions are identified only by the session ID, but is to + * support existing userspace which depends on it. */ - hlist_for_each_entry(session_walk, g_head, global_hlist) - if (session_walk->session_id == session->session_id && - (session_walk->tunnel->encap == L2TP_ENCAPTYPE_IP || - tunnel->encap == L2TP_ENCAPTYPE_IP)) { - err = -EEXIST; - goto err_tlock_pnlock; - } - - l2tp_tunnel_inc_refcount(tunnel); - hlist_add_head_rcu(&session->global_hlist, g_head); - - spin_unlock_bh(&pn->l2tp_session_hlist_lock); + if (err == -ENOSPC && tunnel->encap == L2TP_ENCAPTYPE_UDP) { + other_session = idr_find(&pn->l2tp_v3_session_idr, + session_key); + err = l2tp_session_collision_add(pn, session, + other_session); + } } else { - l2tp_tunnel_inc_refcount(tunnel); + session_key = l2tp_v2_session_key(tunnel->tunnel_id, + session->session_id); + err = idr_alloc_u32(&pn->l2tp_v2_session_idr, NULL, + &session_key, session_key, GFP_ATOMIC); + } + + if (err) { + if (err == -ENOSPC) + err = -EEXIST; + goto out; } - hlist_add_head_rcu(&session->hlist, head); - spin_unlock_bh(&tunnel->hlist_lock); + refcount_inc(&tunnel->ref_count); + WRITE_ONCE(session->tunnel, tunnel); + list_add_rcu(&session->list, &tunnel->session_list); - trace_register_session(session); + /* this makes session available to lockless getters */ + if (tunnel->version == L2TP_HDR_VER_3) { + if (!other_session) + old = idr_replace(&pn->l2tp_v3_session_idr, session, session_key); + } else { + old = idr_replace(&pn->l2tp_v2_session_idr, session, session_key); + } - return 0; + /* old should be NULL, unless something removed or modified + * the IDR entry after our idr_alloc_32 above (which shouldn't + * happen). + */ + WARN_ON_ONCE(old); +out: + spin_unlock_bh(&pn->l2tp_session_idr_lock); + spin_unlock_bh(&tunnel->list_lock); -err_tlock_pnlock: - spin_unlock_bh(&pn->l2tp_session_hlist_lock); -err_tlock: - spin_unlock_bh(&tunnel->hlist_lock); + if (!err) + trace_register_session(session); return err; } @@ -693,7 +917,8 @@ void l2tp_recv_common(struct l2tp_session *session, struct sk_buff *skb, if (!session->lns_mode && !session->send_seq) { trace_session_seqnum_lns_enable(session); session->send_seq = 1; - l2tp_session_set_header_len(session, tunnel->version); + l2tp_session_set_header_len(session, tunnel->version, + tunnel->encap); } } else { /* No sequence numbers. @@ -714,7 +939,8 @@ void l2tp_recv_common(struct l2tp_session *session, struct sk_buff *skb, if (!session->lns_mode && session->send_seq) { trace_session_seqnum_lns_disable(session); session->send_seq = 0; - l2tp_session_set_header_len(session, tunnel->version); + l2tp_session_set_header_len(session, tunnel->version, + tunnel->encap); } else if (session->send_seq) { pr_debug_ratelimited("%s: recv data has no seq numbers when required. Discarding.\n", session->name); @@ -785,18 +1011,14 @@ static void l2tp_session_queue_purge(struct l2tp_session *session) } } -/* Internal UDP receive frame. Do the real work of receiving an L2TP data frame - * here. The skb is not on a list when we get here. - * Returns 0 if the packet was a data packet and was successfully passed on. - * Returns 1 if the packet was not a good data packet and could not be - * forwarded. All such packets are passed up to userspace to deal with. - */ -static int l2tp_udp_recv_core(struct l2tp_tunnel *tunnel, struct sk_buff *skb) +/* UDP encapsulation receive handler. See net/ipv4/udp.c for details. */ +int l2tp_udp_encap_recv(struct sock *sk, struct sk_buff *skb) { struct l2tp_session *session = NULL; + struct l2tp_tunnel *tunnel = NULL; + struct net *net = sock_net(sk); unsigned char *ptr, *optr; u16 hdrflags; - u32 tunnel_id, session_id; u16 version; int length; @@ -806,11 +1028,8 @@ static int l2tp_udp_recv_core(struct l2tp_tunnel *tunnel, struct sk_buff *skb) __skb_pull(skb, sizeof(struct udphdr)); /* Short packet? */ - if (!pskb_may_pull(skb, L2TP_HDR_SIZE_MAX)) { - pr_debug_ratelimited("%s: recv short packet (len=%d)\n", - tunnel->name, skb->len); - goto invalid; - } + if (!pskb_may_pull(skb, L2TP_HDR_SIZE_MAX)) + goto pass; /* Point to L2TP header */ optr = skb->data; @@ -819,13 +1038,8 @@ static int l2tp_udp_recv_core(struct l2tp_tunnel *tunnel, struct sk_buff *skb) /* Get L2TP header flags */ hdrflags = ntohs(*(__be16 *)ptr); - /* Check protocol version */ + /* Get protocol version */ version = hdrflags & L2TP_HDR_VER_MASK; - if (version != tunnel->version) { - pr_debug_ratelimited("%s: recv protocol version mismatch: got %d expected %d\n", - tunnel->name, version, tunnel->version); - goto invalid; - } /* Get length of L2TP packet */ length = skb->len; @@ -837,7 +1051,9 @@ static int l2tp_udp_recv_core(struct l2tp_tunnel *tunnel, struct sk_buff *skb) /* Skip flags */ ptr += 2; - if (tunnel->version == L2TP_HDR_VER_2) { + if (version == L2TP_HDR_VER_2) { + u16 tunnel_id, session_id; + /* If length is present, skip it */ if (hdrflags & L2TP_HDRFLAG_L) ptr += 2; @@ -847,33 +1063,40 @@ static int l2tp_udp_recv_core(struct l2tp_tunnel *tunnel, struct sk_buff *skb) ptr += 2; session_id = ntohs(*(__be16 *)ptr); ptr += 2; + + session = l2tp_v2_session_get(net, tunnel_id, session_id); } else { + u32 session_id; + ptr += 2; /* skip reserved bits */ - tunnel_id = tunnel->tunnel_id; session_id = ntohl(*(__be32 *)ptr); ptr += 4; + + session = l2tp_v3_session_get(net, sk, session_id); } - /* Find the session context */ - session = l2tp_tunnel_get_session(tunnel, session_id); if (!session || !session->recv_skb) { if (session) - l2tp_session_dec_refcount(session); + l2tp_session_put(session); /* Not found? Pass to userspace to deal with */ - pr_debug_ratelimited("%s: no session found (%u/%u). Passing up.\n", - tunnel->name, tunnel_id, session_id); goto pass; } - if (tunnel->version == L2TP_HDR_VER_3 && + tunnel = session->tunnel; + + /* Check protocol version */ + if (version != tunnel->version) + goto invalid; + + if (version == L2TP_HDR_VER_3 && l2tp_v3_ensure_opt_in_linear(session, skb, &ptr, &optr)) { - l2tp_session_dec_refcount(session); + l2tp_session_put(session); goto invalid; } l2tp_recv_common(session, skb, ptr, optr, hdrflags, length); - l2tp_session_dec_refcount(session); + l2tp_session_put(session); return 0; @@ -886,38 +1109,25 @@ pass: return 1; } +EXPORT_SYMBOL_GPL(l2tp_udp_encap_recv); -/* UDP encapsulation receive handler. See net/ipv4/udp.c. - * Return codes: - * 0 : success. - * <0: error - * >0: skb should be passed up to userspace as UDP. - */ -int l2tp_udp_encap_recv(struct sock *sk, struct sk_buff *skb) +/* UDP encapsulation receive error handler. See net/ipv4/udp.c for details. */ +static void l2tp_udp_encap_err_recv(struct sock *sk, struct sk_buff *skb, int err, + __be16 port, u32 info, u8 *payload) { - struct l2tp_tunnel *tunnel; - - /* Note that this is called from the encap_rcv hook inside an - * RCU-protected region, but without the socket being locked. - * Hence we use rcu_dereference_sk_user_data to access the - * tunnel data structure rather the usual l2tp_sk_to_tunnel - * accessor function. - */ - tunnel = rcu_dereference_sk_user_data(sk); - if (!tunnel) - goto pass_up; - if (WARN_ON(tunnel->magic != L2TP_TUNNEL_MAGIC)) - goto pass_up; - - if (l2tp_udp_recv_core(tunnel, skb)) - goto pass_up; + sk->sk_err = err; + sk_error_report(sk); - return 0; - -pass_up: - return 1; + if (ip_hdr(skb)->version == IPVERSION) { + if (inet_test_bit(RECVERR, sk)) + return ip_icmp_error(sk, skb, err, port, info, payload); +#if IS_ENABLED(CONFIG_IPV6) + } else { + if (inet6_test_bit(RECVERR6, sk)) + return ipv6_icmp_error(sk, skb, err, port, info, payload); +#endif + } } -EXPORT_SYMBOL_GPL(l2tp_udp_encap_recv); /************************************************************************ * Transmit handling @@ -1041,7 +1251,13 @@ static int l2tp_xmit_core(struct l2tp_session *session, struct sk_buff *skb, uns IPCB(skb)->flags &= ~(IPSKB_XFRM_TUNNEL_SIZE | IPSKB_XFRM_TRANSFORMED | IPSKB_REROUTED); nf_reset_ct(skb); - bh_lock_sock_nested(sk); + /* L2TP uses its own lockdep subclass to avoid lockdep splats caused by + * nested socket calls on the same lockdep socket class. This can + * happen when data from a user socket is routed over l2tp, which uses + * another userspace socket. + */ + spin_lock_nested(&sk->sk_lock.slock, L2TP_DEPTH_NESTING); + if (sock_owned_by_user(sk)) { kfree_skb(skb); ret = NET_XMIT_DROP; @@ -1093,7 +1309,7 @@ static int l2tp_xmit_core(struct l2tp_session *session, struct sk_buff *skb, uns ret = l2tp_xmit_queue(tunnel, skb, &inet->cork.fl); out_unlock: - bh_unlock_sock(sk); + spin_unlock(&sk->sk_lock.slock); return ret; } @@ -1124,66 +1340,38 @@ EXPORT_SYMBOL_GPL(l2tp_xmit_skb); * Tinnel and session create/destroy. *****************************************************************************/ -/* Tunnel socket destruct hook. - * The tunnel context is deleted only when all session sockets have been - * closed. - */ -static void l2tp_tunnel_destruct(struct sock *sk) -{ - struct l2tp_tunnel *tunnel = l2tp_sk_to_tunnel(sk); - - if (!tunnel) - goto end; - - /* Disable udp encapsulation */ - switch (tunnel->encap) { - case L2TP_ENCAPTYPE_UDP: - /* No longer an encapsulation socket. See net/ipv4/udp.c */ - WRITE_ONCE(udp_sk(sk)->encap_type, 0); - udp_sk(sk)->encap_rcv = NULL; - udp_sk(sk)->encap_destroy = NULL; - break; - case L2TP_ENCAPTYPE_IP: - break; - } - - /* Remove hooks into tunnel socket */ - write_lock_bh(&sk->sk_callback_lock); - sk->sk_destruct = tunnel->old_sk_destruct; - sk->sk_user_data = NULL; - write_unlock_bh(&sk->sk_callback_lock); - - /* Call the original destructor */ - if (sk->sk_destruct) - (*sk->sk_destruct)(sk); - - kfree_rcu(tunnel, rcu); -end: - return; -} - -/* Remove an l2tp session from l2tp_core's hash lists. */ +/* Remove an l2tp session from l2tp_core's lists. */ static void l2tp_session_unhash(struct l2tp_session *session) { struct l2tp_tunnel *tunnel = session->tunnel; - /* Remove the session from core hashes */ if (tunnel) { - /* Remove from the per-tunnel hash */ - spin_lock_bh(&tunnel->hlist_lock); - hlist_del_init_rcu(&session->hlist); - spin_unlock_bh(&tunnel->hlist_lock); - - /* For L2TPv3 we have a per-net hash: remove from there, too */ - if (tunnel->version != L2TP_HDR_VER_2) { - struct l2tp_net *pn = l2tp_pernet(tunnel->l2tp_net); - - spin_lock_bh(&pn->l2tp_session_hlist_lock); - hlist_del_init_rcu(&session->global_hlist); - spin_unlock_bh(&pn->l2tp_session_hlist_lock); + struct l2tp_net *pn = l2tp_pernet(tunnel->l2tp_net); + struct l2tp_session *removed = session; + + spin_lock_bh(&tunnel->list_lock); + spin_lock_bh(&pn->l2tp_session_idr_lock); + + /* Remove from the per-tunnel list */ + list_del_init(&session->list); + + /* Remove from per-net IDR */ + if (tunnel->version == L2TP_HDR_VER_3) { + if (hash_hashed(&session->hlist)) + l2tp_session_collision_del(pn, session); + else + removed = idr_remove(&pn->l2tp_v3_session_idr, + session->session_id); + } else { + u32 session_key = l2tp_v2_session_key(tunnel->tunnel_id, + session->session_id); + removed = idr_remove(&pn->l2tp_v2_session_idr, + session_key); } + WARN_ON_ONCE(removed && removed != session); - synchronize_rcu(); + spin_unlock_bh(&pn->l2tp_session_idr_lock); + spin_unlock_bh(&tunnel->list_lock); } } @@ -1192,37 +1380,24 @@ static void l2tp_session_unhash(struct l2tp_session *session) static void l2tp_tunnel_closeall(struct l2tp_tunnel *tunnel) { struct l2tp_session *session; - int hash; - spin_lock_bh(&tunnel->hlist_lock); + spin_lock_bh(&tunnel->list_lock); tunnel->acpt_newsess = false; - for (hash = 0; hash < L2TP_HASH_SIZE; hash++) { -again: - hlist_for_each_entry_rcu(session, &tunnel->session_hlist[hash], hlist) { - hlist_del_init_rcu(&session->hlist); - - spin_unlock_bh(&tunnel->hlist_lock); - l2tp_session_delete(session); - spin_lock_bh(&tunnel->hlist_lock); - - /* Now restart from the beginning of this hash - * chain. We always remove a session from the - * list so we are guaranteed to make forward - * progress. - */ - goto again; - } - } - spin_unlock_bh(&tunnel->hlist_lock); + list_for_each_entry(session, &tunnel->session_list, list) + l2tp_session_delete(session); + spin_unlock_bh(&tunnel->list_lock); } /* Tunnel socket destroy hook for UDP encapsulation */ static void l2tp_udp_encap_destroy(struct sock *sk) { - struct l2tp_tunnel *tunnel = l2tp_sk_to_tunnel(sk); + struct l2tp_tunnel *tunnel; - if (tunnel) + tunnel = l2tp_sk_to_tunnel(sk); + if (tunnel) { l2tp_tunnel_delete(tunnel); + l2tp_tunnel_put(tunnel); + } } static void l2tp_tunnel_remove(struct net *net, struct l2tp_tunnel *tunnel) @@ -1256,10 +1431,10 @@ static void l2tp_tunnel_del_work(struct work_struct *work) l2tp_tunnel_remove(tunnel->l2tp_net, tunnel); /* drop initial ref */ - l2tp_tunnel_dec_refcount(tunnel); + l2tp_tunnel_put(tunnel); /* drop workqueue ref */ - l2tp_tunnel_dec_refcount(tunnel); + l2tp_tunnel_put(tunnel); } /* Create a socket for the tunnel, if one isn't set up by @@ -1405,10 +1580,10 @@ int l2tp_tunnel_create(int fd, int version, u32 tunnel_id, u32 peer_tunnel_id, tunnel->tunnel_id = tunnel_id; tunnel->peer_tunnel_id = peer_tunnel_id; - tunnel->magic = L2TP_TUNNEL_MAGIC; sprintf(&tunnel->name[0], "tunl %u", tunnel_id); - spin_lock_init(&tunnel->hlist_lock); + spin_lock_init(&tunnel->list_lock); tunnel->acpt_newsess = true; + INIT_LIST_HEAD(&tunnel->session_list); tunnel->encap = encap; @@ -1418,8 +1593,6 @@ int l2tp_tunnel_create(int fd, int version, u32 tunnel_id, u32 peer_tunnel_id, /* Init delete workqueue struct */ INIT_WORK(&tunnel->del_work, l2tp_tunnel_del_work); - INIT_LIST_HEAD(&tunnel->list); - err = 0; err: if (tunnelp) @@ -1432,6 +1605,8 @@ EXPORT_SYMBOL_GPL(l2tp_tunnel_create); static int l2tp_validate_socket(const struct sock *sk, const struct net *net, enum l2tp_encap_type encap) { + struct l2tp_tunnel *tunnel; + if (!net_eq(sock_net(sk), net)) return -EINVAL; @@ -1445,9 +1620,15 @@ static int l2tp_validate_socket(const struct sock *sk, const struct net *net, (encap == L2TP_ENCAPTYPE_IP && sk->sk_protocol != IPPROTO_L2TP)) return -EPROTONOSUPPORT; - if (sk->sk_user_data) + if (encap == L2TP_ENCAPTYPE_UDP && sk->sk_user_data) return -EBUSY; + tunnel = l2tp_sk_to_tunnel(sk); + if (tunnel) { + l2tp_tunnel_put(tunnel); + return -EBUSY; + } + return 0; } @@ -1485,22 +1666,19 @@ int l2tp_tunnel_register(struct l2tp_tunnel *tunnel, struct net *net, ret = l2tp_validate_socket(sk, net, tunnel->encap); if (ret < 0) goto err_inval_sock; - rcu_assign_sk_user_data(sk, tunnel); write_unlock_bh(&sk->sk_callback_lock); if (tunnel->encap == L2TP_ENCAPTYPE_UDP) { struct udp_tunnel_sock_cfg udp_cfg = { - .sk_user_data = tunnel, .encap_type = UDP_ENCAP_L2TPINUDP, .encap_rcv = l2tp_udp_encap_recv, + .encap_err_rcv = l2tp_udp_encap_err_recv, .encap_destroy = l2tp_udp_encap_destroy, }; setup_udp_tunnel_sock(net, sock, &udp_cfg); } - tunnel->old_sk_destruct = sk->sk_destruct; - sk->sk_destruct = &l2tp_tunnel_destruct; sk->sk_allocation = GFP_ATOMIC; release_sock(sk); @@ -1539,7 +1717,7 @@ void l2tp_tunnel_delete(struct l2tp_tunnel *tunnel) { if (!test_and_set_bit(0, &tunnel->dead)) { trace_delete_tunnel(tunnel); - l2tp_tunnel_inc_refcount(tunnel); + refcount_inc(&tunnel->ref_count); queue_work(l2tp_wq, &tunnel->del_work); } } @@ -1547,23 +1725,37 @@ EXPORT_SYMBOL_GPL(l2tp_tunnel_delete); void l2tp_session_delete(struct l2tp_session *session) { - if (test_and_set_bit(0, &session->dead)) - return; + if (!test_and_set_bit(0, &session->dead)) { + trace_delete_session(session); + refcount_inc(&session->ref_count); + queue_work(l2tp_wq, &session->del_work); + } +} +EXPORT_SYMBOL_GPL(l2tp_session_delete); + +/* Workqueue session deletion function */ +static void l2tp_session_del_work(struct work_struct *work) +{ + struct l2tp_session *session = container_of(work, struct l2tp_session, + del_work); - trace_delete_session(session); l2tp_session_unhash(session); l2tp_session_queue_purge(session); if (session->session_close) (*session->session_close)(session); - l2tp_session_dec_refcount(session); + /* drop initial ref */ + l2tp_session_put(session); + + /* drop workqueue ref */ + l2tp_session_put(session); } -EXPORT_SYMBOL_GPL(l2tp_session_delete); /* We come here whenever a session's send_seq, cookie_len or * l2specific_type parameters are set. */ -void l2tp_session_set_header_len(struct l2tp_session *session, int version) +void l2tp_session_set_header_len(struct l2tp_session *session, int version, + enum l2tp_encap_type encap) { if (version == L2TP_HDR_VER_2) { session->hdr_len = 6; @@ -1572,7 +1764,7 @@ void l2tp_session_set_header_len(struct l2tp_session *session, int version) } else { session->hdr_len = 4 + session->cookie_len; session->hdr_len += l2tp_get_l2specific_len(session); - if (session->tunnel->encap == L2TP_ENCAPTYPE_UDP) + if (encap == L2TP_ENCAPTYPE_UDP) session->hdr_len += 4; } } @@ -1586,7 +1778,6 @@ struct l2tp_session *l2tp_session_create(int priv_size, struct l2tp_tunnel *tunn session = kzalloc(sizeof(*session) + priv_size, GFP_KERNEL); if (session) { session->magic = L2TP_SESSION_MAGIC; - session->tunnel = tunnel; session->session_id = session_id; session->peer_session_id = peer_session_id; @@ -1606,8 +1797,11 @@ struct l2tp_session *l2tp_session_create(int priv_size, struct l2tp_tunnel *tunn skb_queue_head_init(&session->reorder_q); + session->hlist_key = l2tp_v3_session_hashkey(tunnel->sock, session->session_id); INIT_HLIST_NODE(&session->hlist); - INIT_HLIST_NODE(&session->global_hlist); + INIT_LIST_HEAD(&session->clist); + INIT_LIST_HEAD(&session->list); + INIT_WORK(&session->del_work, l2tp_session_del_work); if (cfg) { session->pwtype = cfg->pw_type; @@ -1622,7 +1816,7 @@ struct l2tp_session *l2tp_session_create(int priv_size, struct l2tp_tunnel *tunn memcpy(&session->peer_cookie[0], &cfg->peer_cookie[0], cfg->peer_cookie_len); } - l2tp_session_set_header_len(session, tunnel->version); + l2tp_session_set_header_len(session, tunnel->version, tunnel->encap); refcount_set(&session->ref_count, 1); @@ -1640,25 +1834,22 @@ EXPORT_SYMBOL_GPL(l2tp_session_create); static __net_init int l2tp_init_net(struct net *net) { struct l2tp_net *pn = net_generic(net, l2tp_net_id); - int hash; idr_init(&pn->l2tp_tunnel_idr); spin_lock_init(&pn->l2tp_tunnel_idr_lock); - for (hash = 0; hash < L2TP_HASH_SIZE_2; hash++) - INIT_HLIST_HEAD(&pn->l2tp_session_hlist[hash]); - - spin_lock_init(&pn->l2tp_session_hlist_lock); + idr_init(&pn->l2tp_v2_session_idr); + idr_init(&pn->l2tp_v3_session_idr); + spin_lock_init(&pn->l2tp_session_idr_lock); return 0; } -static __net_exit void l2tp_exit_net(struct net *net) +static __net_exit void l2tp_pre_exit_net(struct net *net) { struct l2tp_net *pn = l2tp_pernet(net); struct l2tp_tunnel *tunnel = NULL; unsigned long tunnel_id, tmp; - int hash; rcu_read_lock_bh(); idr_for_each_entry_ul(&pn->l2tp_tunnel_idr, tunnel, tmp, tunnel_id) { @@ -1667,18 +1858,34 @@ static __net_exit void l2tp_exit_net(struct net *net) } rcu_read_unlock_bh(); - if (l2tp_wq) - flush_workqueue(l2tp_wq); - rcu_barrier(); + if (l2tp_wq) { + /* Run all TUNNEL_DELETE work items just queued. */ + __flush_workqueue(l2tp_wq); + + /* Each TUNNEL_DELETE work item will queue a SESSION_DELETE + * work item for each session in the tunnel. Flush the + * workqueue again to process these. + */ + __flush_workqueue(l2tp_wq); + } +} + +static __net_exit void l2tp_exit_net(struct net *net) +{ + struct l2tp_net *pn = l2tp_pernet(net); - for (hash = 0; hash < L2TP_HASH_SIZE_2; hash++) - WARN_ON_ONCE(!hlist_empty(&pn->l2tp_session_hlist[hash])); + WARN_ON_ONCE(!idr_is_empty(&pn->l2tp_v2_session_idr)); + idr_destroy(&pn->l2tp_v2_session_idr); + WARN_ON_ONCE(!idr_is_empty(&pn->l2tp_v3_session_idr)); + idr_destroy(&pn->l2tp_v3_session_idr); + WARN_ON_ONCE(!idr_is_empty(&pn->l2tp_tunnel_idr)); idr_destroy(&pn->l2tp_tunnel_idr); } static struct pernet_operations l2tp_net_ops = { .init = l2tp_init_net, .exit = l2tp_exit_net, + .pre_exit = l2tp_pre_exit_net, .id = &l2tp_net_id, .size = sizeof(struct l2tp_net), }; |