From 71cea17ed39fdf1c0634f530ddc6a2c2fc601c2b Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 20 May 2013 06:52:26 +0000 Subject: tcp: md5: remove spinlock usage in fast path TCP md5 code uses per cpu variables but protects access to them with a shared spinlock, which is a contention point. [ tcp_md5sig_pool_lock is locked twice per incoming packet ] Makes things much simpler, by allocating crypto structures once, first time a socket needs md5 keys, and not deallocating them as they are really small. Next step would be to allow crypto allocations being done in a NUMA aware way. Signed-off-by: Eric Dumazet Cc: Herbert Xu Signed-off-by: David S. Miller --- net/ipv4/tcp_ipv4.c | 10 +--------- 1 file changed, 1 insertion(+), 9 deletions(-) (limited to 'net/ipv4/tcp_ipv4.c') diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 719652305a2..d20ede0c959 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -1026,7 +1026,7 @@ int tcp_md5_do_add(struct sock *sk, const union tcp_md5_addr *addr, key = sock_kmalloc(sk, sizeof(*key), gfp); if (!key) return -ENOMEM; - if (hlist_empty(&md5sig->head) && !tcp_alloc_md5sig_pool(sk)) { + if (!tcp_alloc_md5sig_pool()) { sock_kfree_s(sk, key, sizeof(*key)); return -ENOMEM; } @@ -1044,9 +1044,7 @@ EXPORT_SYMBOL(tcp_md5_do_add); int tcp_md5_do_del(struct sock *sk, const union tcp_md5_addr *addr, int family) { - struct tcp_sock *tp = tcp_sk(sk); struct tcp_md5sig_key *key; - struct tcp_md5sig_info *md5sig; key = tcp_md5_do_lookup(sk, (union tcp_md5_addr *)&addr, AF_INET); if (!key) @@ -1054,10 +1052,6 @@ int tcp_md5_do_del(struct sock *sk, const union tcp_md5_addr *addr, int family) hlist_del_rcu(&key->node); atomic_sub(sizeof(*key), &sk->sk_omem_alloc); kfree_rcu(key, rcu); - md5sig = rcu_dereference_protected(tp->md5sig_info, - sock_owned_by_user(sk)); - if (hlist_empty(&md5sig->head)) - tcp_free_md5sig_pool(); return 0; } EXPORT_SYMBOL(tcp_md5_do_del); @@ -1071,8 +1065,6 @@ static void tcp_clear_md5_list(struct sock *sk) md5sig = rcu_dereference_protected(tp->md5sig_info, 1); - if (!hlist_empty(&md5sig->head)) - tcp_free_md5sig_pool(); hlist_for_each_entry_safe(key, n, &md5sig->head, node) { hlist_del_rcu(&key->node); atomic_sub(sizeof(*key), &sk->sk_omem_alloc); -- cgit v1.2.3-18-g5258 From 28850dc7c71da9d0c0e39246e9ff6913f41f8d0a Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Fri, 7 Jun 2013 05:11:46 +0000 Subject: net: tcp: move GRO/GSO functions to tcp_offload Would be good to make things explicit and move those functions to a new file called tcp_offload.c, thus make this similar to tcpv6_offload.c. While moving all related functions into tcp_offload.c, we can also make some of them static, since they are only used there. Also, add an explicit registration function. Suggested-by: Eric Dumazet Signed-off-by: Daniel Borkmann Acked-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/tcp_ipv4.c | 66 +---------------------------------------------------- 1 file changed, 1 insertion(+), 65 deletions(-) (limited to 'net/ipv4/tcp_ipv4.c') diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index d20ede0c959..289039b4d8d 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -545,8 +545,7 @@ out: sock_put(sk); } -static void __tcp_v4_send_check(struct sk_buff *skb, - __be32 saddr, __be32 daddr) +void __tcp_v4_send_check(struct sk_buff *skb, __be32 saddr, __be32 daddr) { struct tcphdr *th = tcp_hdr(skb); @@ -571,23 +570,6 @@ void tcp_v4_send_check(struct sock *sk, struct sk_buff *skb) } EXPORT_SYMBOL(tcp_v4_send_check); -int tcp_v4_gso_send_check(struct sk_buff *skb) -{ - const struct iphdr *iph; - struct tcphdr *th; - - if (!pskb_may_pull(skb, sizeof(*th))) - return -EINVAL; - - iph = ip_hdr(skb); - th = tcp_hdr(skb); - - th->check = 0; - skb->ip_summed = CHECKSUM_PARTIAL; - __tcp_v4_send_check(skb, iph->saddr, iph->daddr); - return 0; -} - /* * This routine will send an RST to the other tcp. * @@ -2795,52 +2777,6 @@ void tcp4_proc_exit(void) } #endif /* CONFIG_PROC_FS */ -struct sk_buff **tcp4_gro_receive(struct sk_buff **head, struct sk_buff *skb) -{ - const struct iphdr *iph = skb_gro_network_header(skb); - __wsum wsum; - __sum16 sum; - - switch (skb->ip_summed) { - case CHECKSUM_COMPLETE: - if (!tcp_v4_check(skb_gro_len(skb), iph->saddr, iph->daddr, - skb->csum)) { - skb->ip_summed = CHECKSUM_UNNECESSARY; - break; - } -flush: - NAPI_GRO_CB(skb)->flush = 1; - return NULL; - - case CHECKSUM_NONE: - wsum = csum_tcpudp_nofold(iph->saddr, iph->daddr, - skb_gro_len(skb), IPPROTO_TCP, 0); - sum = csum_fold(skb_checksum(skb, - skb_gro_offset(skb), - skb_gro_len(skb), - wsum)); - if (sum) - goto flush; - - skb->ip_summed = CHECKSUM_UNNECESSARY; - break; - } - - return tcp_gro_receive(head, skb); -} - -int tcp4_gro_complete(struct sk_buff *skb) -{ - const struct iphdr *iph = ip_hdr(skb); - struct tcphdr *th = tcp_hdr(skb); - - th->check = ~tcp_v4_check(skb->len - skb_transport_offset(skb), - iph->saddr, iph->daddr, 0); - skb_shinfo(skb)->gso_type = SKB_GSO_TCPV4; - - return tcp_gro_complete(skb); -} - struct proto tcp_prot = { .name = "TCP", .owner = THIS_MODULE, -- cgit v1.2.3-18-g5258 From d30e383bb856f614ddb5bbbb5a7d3f86240e41ec Mon Sep 17 00:00:00 2001 From: Eliezer Tamir Date: Mon, 10 Jun 2013 11:40:10 +0300 Subject: tcp: add low latency socket poll support. Adds low latency socket poll support for TCP. In tcp_v[46]_rcv() add a call to sk_mark_ll() to copy the napi_id from the skb to the sk. In tcp_recvmsg(), when there is no data in the socket we busy-poll. This is a good example of how to add busy-poll support to more protocols. Signed-off-by: Alexander Duyck Signed-off-by: Jesse Brandeburg Signed-off-by: Eliezer Tamir Acked-by: Eric Dumazet Tested-by: Willem de Bruijn Signed-off-by: David S. Miller --- net/ipv4/tcp_ipv4.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'net/ipv4/tcp_ipv4.c') diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 289039b4d8d..1063bb83e34 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -75,6 +75,7 @@ #include #include #include +#include #include #include @@ -1993,6 +1994,7 @@ process: if (sk_filter(sk, skb)) goto discard_and_relse; + sk_mark_ll(sk, skb); skb->dev = NULL; bh_lock_sock_nested(sk); -- cgit v1.2.3-18-g5258 From 076bb0c82a44fbe46fe2c8527a5b5b64b69f679d Mon Sep 17 00:00:00 2001 From: Eliezer Tamir Date: Wed, 10 Jul 2013 17:13:17 +0300 Subject: net: rename include/net/ll_poll.h to include/net/busy_poll.h Rename the file and correct all the places where it is included. Signed-off-by: Eliezer Tamir Signed-off-by: David S. Miller --- net/ipv4/tcp_ipv4.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'net/ipv4/tcp_ipv4.c') diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 35675e46aff..3a261b41a00 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -75,7 +75,7 @@ #include #include #include -#include +#include #include #include -- cgit v1.2.3-18-g5258 From 8b80cda536ea9bceec0364e897868a30ee13b992 Mon Sep 17 00:00:00 2001 From: Eliezer Tamir Date: Wed, 10 Jul 2013 17:13:26 +0300 Subject: net: rename ll methods to busy-poll Rename ndo_ll_poll to ndo_busy_poll. Rename sk_mark_ll to sk_mark_napi_id. Rename skb_mark_ll to skb_mark_napi_id. Correct all useres of these functions. Update comments and defines in include/net/busy_poll.h Signed-off-by: Eliezer Tamir Signed-off-by: David S. Miller --- net/ipv4/tcp_ipv4.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'net/ipv4/tcp_ipv4.c') diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 3a261b41a00..b299da5ff49 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -1994,7 +1994,7 @@ process: if (sk_filter(sk, skb)) goto discard_and_relse; - sk_mark_ll(sk, skb); + sk_mark_napi_id(sk, skb); skb->dev = NULL; bh_lock_sock_nested(sk); -- cgit v1.2.3-18-g5258