From b1a78b9b98862cda167b643690e43662ea060625 Mon Sep 17 00:00:00 2001 From: Xin Long Date: Sat, 28 Jan 2023 10:58:39 -0500 Subject: net: add support for ipv4 big tcp Similar to Eric's IPv6 BIG TCP, this patch is to enable IPv4 BIG TCP. Firstly, allow sk->sk_gso_max_size to be set to a value greater than GSO_LEGACY_MAX_SIZE by not trimming gso_max_size in sk_trim_gso_size() for IPv4 TCP sockets. Then on TX path, set IP header tot_len to 0 when skb->len > IP_MAX_MTU in __ip_local_out() to allow to send BIG TCP packets, and this implies that skb->len is the length of a IPv4 packet; On RX path, use skb->len as the length of the IPv4 packet when the IP header tot_len is 0 and skb->len > IP_MAX_MTU in ip_rcv_core(). As the API iph_set_totlen() and skb_ip_totlen() are used in __ip_local_out() and ip_rcv_core(), we only need to update these APIs. Also in GRO receive, add the check for ETH_P_IP/IPPROTO_TCP, and allows the merged packet size >= GRO_LEGACY_MAX_SIZE in skb_gro_receive(). In GRO complete, set IP header tot_len to 0 when the merged packet size greater than IP_MAX_MTU in iph_set_totlen() so that it can be processed on RX path. Note that by checking skb_is_gso_tcp() in API iph_totlen(), it makes this implementation safe to use iph->len == 0 indicates IPv4 BIG TCP packets. Signed-off-by: Xin Long Reviewed-by: David Ahern Reviewed-by: Eric Dumazet Signed-off-by: Jakub Kicinski --- net/ipv4/af_inet.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'net/ipv4/af_inet.c') diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c index 6c0ec2789943..2f992a323b95 100644 --- a/net/ipv4/af_inet.c +++ b/net/ipv4/af_inet.c @@ -1485,6 +1485,7 @@ struct sk_buff *inet_gro_receive(struct list_head *head, struct sk_buff *skb) if (unlikely(ip_fast_csum((u8 *)iph, 5))) goto out; + NAPI_GRO_CB(skb)->proto = proto; id = ntohl(*(__be32 *)&iph->id); flush = (u16)((ntohl(*(__be32 *)iph) ^ skb_gro_len(skb)) | (id & ~IP_DF)); id >>= 16; @@ -1618,9 +1619,9 @@ int inet_recv_error(struct sock *sk, struct msghdr *msg, int len, int *addr_len) int inet_gro_complete(struct sk_buff *skb, int nhoff) { - __be16 newlen = htons(skb->len - nhoff); struct iphdr *iph = (struct iphdr *)(skb->data + nhoff); const struct net_offload *ops; + __be16 totlen = iph->tot_len; int proto = iph->protocol; int err = -ENOSYS; @@ -1629,8 +1630,8 @@ int inet_gro_complete(struct sk_buff *skb, int nhoff) skb_set_inner_network_header(skb, nhoff); } - csum_replace2(&iph->check, iph->tot_len, newlen); - iph->tot_len = newlen; + iph_set_totlen(iph, skb->len - nhoff); + csum_replace2(&iph->check, totlen, iph->tot_len); ops = rcu_dereference(inet_offloads[proto]); if (WARN_ON(!ops || !ops->callbacks.gro_complete)) -- cgit v1.2.3 From fe33311c3e371855c4f4c0ab8a5fce5b9a9fdafd Mon Sep 17 00:00:00 2001 From: Jason Xing Date: Tue, 14 Feb 2023 12:14:10 +0800 Subject: net: no longer support SOCK_REFCNT_DEBUG feature Commit e48c414ee61f ("[INET]: Generalise the TCP sock ID lookup routines") commented out the definition of SOCK_REFCNT_DEBUG in 2005 and later another commit 463c84b97f24 ("[NET]: Introduce inet_connection_sock") removed it. Since we could track all of them through bpf and kprobe related tools and the feature could print loads of information which might not be that helpful even under a little bit pressure, the whole feature which has been inactive for many years is no longer supported. Link: https://lore.kernel.org/lkml/20230211065153.54116-1-kerneljasonxing@gmail.com/ Suggested-by: Kuniyuki Iwashima Signed-off-by: Jason Xing Reviewed-by: Kuniyuki Iwashima Acked-by: Wenjia Zhang Reviewed-by: Eric Dumazet Acked-by: Matthieu Baerts Signed-off-by: David S. Miller --- include/net/sock.h | 28 ---------------------------- net/core/sock.c | 13 ------------- net/ipv4/af_inet.c | 3 --- net/ipv4/inet_connection_sock.c | 2 -- net/ipv4/inet_timewait_sock.c | 3 --- net/ipv6/af_inet6.c | 10 ---------- net/ipv6/ipv6_sockglue.c | 12 ------------ net/mptcp/protocol.c | 1 - net/packet/af_packet.c | 4 ---- net/sctp/ipv6.c | 2 -- net/sctp/protocol.c | 2 -- net/smc/af_smc.c | 3 --- net/xdp/xsk.c | 4 ---- 13 files changed, 87 deletions(-) (limited to 'net/ipv4/af_inet.c') diff --git a/include/net/sock.h b/include/net/sock.h index 937e842dc930..2cb258fde072 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1349,9 +1349,6 @@ struct proto { char name[32]; struct list_head node; -#ifdef SOCK_REFCNT_DEBUG - atomic_t socks; -#endif int (*diag_destroy)(struct sock *sk, int err); } __randomize_layout; @@ -1359,31 +1356,6 @@ int proto_register(struct proto *prot, int alloc_slab); void proto_unregister(struct proto *prot); int sock_load_diag_module(int family, int protocol); -#ifdef SOCK_REFCNT_DEBUG -static inline void sk_refcnt_debug_inc(struct sock *sk) -{ - atomic_inc(&sk->sk_prot->socks); -} - -static inline void sk_refcnt_debug_dec(struct sock *sk) -{ - atomic_dec(&sk->sk_prot->socks); - printk(KERN_DEBUG "%s socket %p released, %d are still alive\n", - sk->sk_prot->name, sk, atomic_read(&sk->sk_prot->socks)); -} - -static inline void sk_refcnt_debug_release(const struct sock *sk) -{ - if (refcount_read(&sk->sk_refcnt) != 1) - printk(KERN_DEBUG "Destruction of the %s socket %p delayed, refcnt=%d\n", - sk->sk_prot->name, sk, refcount_read(&sk->sk_refcnt)); -} -#else /* SOCK_REFCNT_DEBUG */ -#define sk_refcnt_debug_inc(sk) do { } while (0) -#define sk_refcnt_debug_dec(sk) do { } while (0) -#define sk_refcnt_debug_release(sk) do { } while (0) -#endif /* SOCK_REFCNT_DEBUG */ - INDIRECT_CALLABLE_DECLARE(bool tcp_stream_memory_free(const struct sock *sk, int wake)); static inline int sk_forward_alloc_get(const struct sock *sk) diff --git a/net/core/sock.c b/net/core/sock.c index afbb02984d5f..341c565dbc26 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -2340,17 +2340,6 @@ struct sock *sk_clone_lock(const struct sock *sk, const gfp_t priority) smp_wmb(); refcount_set(&newsk->sk_refcnt, 2); - /* Increment the counter in the same struct proto as the master - * sock (sk_refcnt_debug_inc uses newsk->sk_prot->socks, that - * is the same as sk->sk_prot->socks, as this field was copied - * with memcpy). - * - * This _changes_ the previous behaviour, where - * tcp_create_openreq_child always was incrementing the - * equivalent to tcp_prot->socks (inet_sock_nr), so this have - * to be taken into account in all callers. -acme - */ - sk_refcnt_debug_inc(newsk); sk_set_socket(newsk, NULL); sk_tx_queue_clear(newsk); RCU_INIT_POINTER(newsk->sk_wq, NULL); @@ -3710,8 +3699,6 @@ void sk_common_release(struct sock *sk) xfrm_sk_free_policy(sk); - sk_refcnt_debug_release(sk); - sock_put(sk); } EXPORT_SYMBOL(sk_common_release); diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c index 2c778b013cb0..8db6747f892f 100644 --- a/net/ipv4/af_inet.c +++ b/net/ipv4/af_inet.c @@ -156,7 +156,6 @@ void inet_sock_destruct(struct sock *sk) kfree(rcu_dereference_protected(inet->inet_opt, 1)); dst_release(rcu_dereference_protected(sk->sk_dst_cache, 1)); dst_release(rcu_dereference_protected(sk->sk_rx_dst, 1)); - sk_refcnt_debug_dec(sk); } EXPORT_SYMBOL(inet_sock_destruct); @@ -357,8 +356,6 @@ lookup_protocol: inet->mc_list = NULL; inet->rcv_tos = 0; - sk_refcnt_debug_inc(sk); - if (inet->inet_num) { /* It assumes that any protocol which allows * the user to assign a number at socket diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c index 7d206a10ad14..eedcf4146d29 100644 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -1199,8 +1199,6 @@ void inet_csk_destroy_sock(struct sock *sk) xfrm_sk_free_policy(sk); - sk_refcnt_debug_release(sk); - this_cpu_dec(*sk->sk_prot->orphan_count); sock_put(sk); diff --git a/net/ipv4/inet_timewait_sock.c b/net/ipv4/inet_timewait_sock.c index beed32fff484..40052414c7c7 100644 --- a/net/ipv4/inet_timewait_sock.c +++ b/net/ipv4/inet_timewait_sock.c @@ -77,9 +77,6 @@ void inet_twsk_free(struct inet_timewait_sock *tw) { struct module *owner = tw->tw_prot->owner; twsk_destructor((struct sock *)tw); -#ifdef SOCK_REFCNT_DEBUG - pr_debug("%s timewait_sock %p released\n", tw->tw_prot->name, tw); -#endif kmem_cache_free(tw->tw_prot->twsk_prot->twsk_slab, tw); module_put(owner); } diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c index 847934763868..38689bedfce7 100644 --- a/net/ipv6/af_inet6.c +++ b/net/ipv6/af_inet6.c @@ -239,16 +239,6 @@ lookup_protocol: inet->pmtudisc = IP_PMTUDISC_DONT; else inet->pmtudisc = IP_PMTUDISC_WANT; - /* - * Increment only the relevant sk_prot->socks debug field, this changes - * the previous behaviour of incrementing both the equivalent to - * answer->prot->socks (inet6_sock_nr) and inet_sock_nr. - * - * This allows better debug granularity as we'll know exactly how many - * UDPv6, TCPv6, etc socks were allocated, not the sum of all IPv6 - * transport protocol socks. -acme - */ - sk_refcnt_debug_inc(sk); if (inet->inet_num) { /* It assumes that any protocol which allows diff --git a/net/ipv6/ipv6_sockglue.c b/net/ipv6/ipv6_sockglue.c index 9ce51680290b..2917dd8d198c 100644 --- a/net/ipv6/ipv6_sockglue.c +++ b/net/ipv6/ipv6_sockglue.c @@ -464,13 +464,6 @@ int do_ipv6_setsockopt(struct sock *sk, int level, int optname, __ipv6_sock_mc_close(sk); __ipv6_sock_ac_close(sk); - /* - * Sock is moving from IPv6 to IPv4 (sk_prot), so - * remove it from the refcnt debug socks count in the - * original family... - */ - sk_refcnt_debug_dec(sk); - if (sk->sk_protocol == IPPROTO_TCP) { struct inet_connection_sock *icsk = inet_csk(sk); @@ -507,11 +500,6 @@ int do_ipv6_setsockopt(struct sock *sk, int level, int optname, inet6_cleanup_sock(sk); - /* - * ... and add it to the refcnt debug socks count - * in the new family. -acme - */ - sk_refcnt_debug_inc(sk); module_put(THIS_MODULE); retv = 0; break; diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index c9817aa0f413..3ad9c46202fc 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -2875,7 +2875,6 @@ static void __mptcp_destroy_sock(struct sock *sk) sk_stream_kill_queues(sk); xfrm_sk_free_policy(sk); - sk_refcnt_debug_release(sk); sock_put(sk); } diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index 8ffb19c643ab..d4e76e2ae153 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -1335,8 +1335,6 @@ static void packet_sock_destruct(struct sock *sk) pr_err("Attempt to release alive packet socket: %p\n", sk); return; } - - sk_refcnt_debug_dec(sk); } static bool fanout_flow_is_huge(struct packet_sock *po, struct sk_buff *skb) @@ -3174,7 +3172,6 @@ static int packet_release(struct socket *sock) skb_queue_purge(&sk->sk_receive_queue); packet_free_pending(po); - sk_refcnt_debug_release(sk); sock_put(sk); return 0; @@ -3364,7 +3361,6 @@ static int packet_create(struct net *net, struct socket *sock, int protocol, packet_cached_dev_reset(po); sk->sk_destruct = packet_sock_destruct; - sk_refcnt_debug_inc(sk); /* * Attach a protocol block diff --git a/net/sctp/ipv6.c b/net/sctp/ipv6.c index 097bd60ce964..62b436a2c8fe 100644 --- a/net/sctp/ipv6.c +++ b/net/sctp/ipv6.c @@ -807,8 +807,6 @@ static struct sock *sctp_v6_create_accept_sk(struct sock *sk, newsk->sk_v6_rcv_saddr = sk->sk_v6_rcv_saddr; - sk_refcnt_debug_inc(newsk); - if (newsk->sk_prot->init(newsk)) { sk_common_release(newsk); newsk = NULL; diff --git a/net/sctp/protocol.c b/net/sctp/protocol.c index 909a89a1cff4..c365df24ad33 100644 --- a/net/sctp/protocol.c +++ b/net/sctp/protocol.c @@ -601,8 +601,6 @@ static struct sock *sctp_v4_create_accept_sk(struct sock *sk, newinet->inet_daddr = asoc->peer.primary_addr.v4.sin_addr.s_addr; - sk_refcnt_debug_inc(newsk); - if (newsk->sk_prot->init(newsk)) { sk_common_release(newsk); newsk = NULL; diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c index b163266e581a..d7a7420e81ec 100644 --- a/net/smc/af_smc.c +++ b/net/smc/af_smc.c @@ -360,8 +360,6 @@ static void smc_destruct(struct sock *sk) return; if (!sock_flag(sk, SOCK_DEAD)) return; - - sk_refcnt_debug_dec(sk); } static struct sock *smc_sock_alloc(struct net *net, struct socket *sock, @@ -390,7 +388,6 @@ static struct sock *smc_sock_alloc(struct net *net, struct socket *sock, spin_lock_init(&smc->accept_q_lock); spin_lock_init(&smc->conn.send_lock); sk->sk_prot->hash(sk); - sk_refcnt_debug_inc(sk); mutex_init(&smc->clcsock_release_lock); smc_init_saved_callbacks(smc); diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c index 9f0561b67c12..a245c1b4a21b 100644 --- a/net/xdp/xsk.c +++ b/net/xdp/xsk.c @@ -845,7 +845,6 @@ static int xsk_release(struct socket *sock) sock_orphan(sk); sock->sk = NULL; - sk_refcnt_debug_release(sk); sock_put(sk); return 0; @@ -1396,8 +1395,6 @@ static void xsk_destruct(struct sock *sk) if (!xp_put_pool(xs->pool)) xdp_put_umem(xs->umem, !xs->pool); - - sk_refcnt_debug_dec(sk); } static int xsk_create(struct net *net, struct socket *sock, int protocol, @@ -1427,7 +1424,6 @@ static int xsk_create(struct net *net, struct socket *sock, int protocol, sk->sk_family = PF_XDP; sk->sk_destruct = xsk_destruct; - sk_refcnt_debug_inc(sk); sock_set_flag(sk, SOCK_RCU_FREE); -- cgit v1.2.3