Commit 58d607d3 authored by Eric Dumazet's avatar Eric Dumazet Committed by David S. Miller
Browse files

tcp: provide skb->hash to synack packets 



In commit b73c3d0e ("net: Save TX flow hash in sock and set in skbuf
on xmit"), Tom provided a l4 hash to most outgoing TCP packets.

We'd like to provide one as well for SYNACK packets, so that all packets
of a given flow share same txhash, to later enable bonding driver to
also use skb->hash to perform slave selection.

Note that a SYNACK retransmit shuffles the tx hash, as Tom did
in commit 265f94ff ("net: Recompute sk_txhash on negative routing
advice") for established sockets.

This has nice effect making TCP flows resilient to some kind of black
holes, even at connection establish phase.

Signed-off-by: default avatarEric Dumazet <edumazet@google.com>
Cc: Tom Herbert <tom@herbertland.com>
Cc: Mahesh Bandewar <maheshb@google.com>
Acked-by: default avatarTom Herbert <tom@herbertland.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
parent bbe83731

include/linux/tcp.h

+1 −0
Original line number Original line Diff line number Diff line
@@ -113,6 +113,7 @@ struct tcp_request_sock { 
	struct inet_request_sock 	req;

	struct inet_request_sock 	req;

	const struct tcp_request_sock_ops *af_specific;

	const struct tcp_request_sock_ops *af_specific;

	bool				tfo_listener;

	bool				tfo_listener;

	u32				txhash;

	u32				rcv_isn;

	u32				rcv_isn;

	u32				snt_isn;

	u32				snt_isn;

	u32				snt_synack; /* synack sent time */

	u32				snt_synack; /* synack sent time */

include/net/sock.h

+8 −4
Original line number Original line Diff line number Diff line
@@ -1654,12 +1654,16 @@ static inline void sock_graft(struct sock *sk, struct socket *parent) 
kuid_t sock_i_uid(struct sock *sk);

kuid_t sock_i_uid(struct sock *sk);

unsigned long sock_i_ino(struct sock *sk);

unsigned long sock_i_ino(struct sock *sk);





static inline void sk_set_txhash(struct sock *sk)

static inline u32 net_tx_rndhash(void)

{

{

	sk->sk_txhash = prandom_u32();

	u32 v = prandom_u32();



	return v ?: 1;

}





	if (unlikely(!sk->sk_txhash))

static inline void sk_set_txhash(struct sock *sk)

		sk->sk_txhash = 1;

{

	sk->sk_txhash = net_tx_rndhash();

}

}





static inline void sk_rethink_txhash(struct sock *sk)

static inline void sk_rethink_txhash(struct sock *sk)

net/ipv4/tcp_input.c

+1 −0
Original line number Original line Diff line number Diff line
@@ -6228,6 +6228,7 @@ int tcp_conn_request(struct request_sock_ops *rsk_ops, 
	}

	}





	tcp_rsk(req)->snt_isn = isn;

	tcp_rsk(req)->snt_isn = isn;

	tcp_rsk(req)->txhash = net_tx_rndhash();

	tcp_openreq_init_rwin(req, sk, dst);

	tcp_openreq_init_rwin(req, sk, dst);

	fastopen = !want_cookie &&

	fastopen = !want_cookie &&

		   tcp_try_fastopen(sk, skb, req, &foc, dst);

		   tcp_try_fastopen(sk, skb, req, &foc, dst);

net/ipv4/tcp_ipv4.c

+1 −1
Original line number Original line Diff line number Diff line
@@ -1276,8 +1276,8 @@ struct sock *tcp_v4_syn_recv_sock(struct sock *sk, struct sk_buff *skb, 
	newinet->mc_index     = inet_iif(skb);

	newinet->mc_index     = inet_iif(skb);

	newinet->mc_ttl	      = ip_hdr(skb)->ttl;

	newinet->mc_ttl	      = ip_hdr(skb)->ttl;

	newinet->rcv_tos      = ip_hdr(skb)->tos;

	newinet->rcv_tos      = ip_hdr(skb)->tos;

	newsk->sk_txhash      = tcp_rsk(req)->txhash;

	inet_csk(newsk)->icsk_ext_hdr_len = 0;

	inet_csk(newsk)->icsk_ext_hdr_len = 0;

	sk_set_txhash(newsk);

	if (inet_opt)

	if (inet_opt)

		inet_csk(newsk)->icsk_ext_hdr_len = inet_opt->opt.optlen;

		inet_csk(newsk)->icsk_ext_hdr_len = inet_opt->opt.optlen;

	newinet->inet_id = newtp->write_seq ^ jiffies;

	newinet->inet_id = newtp->write_seq ^ jiffies;

net/ipv4/tcp_output.c

+2 −0
Original line number Original line Diff line number Diff line
@@ -2987,6 +2987,7 @@ struct sk_buff *tcp_make_synack(struct sock *sk, struct dst_entry *dst, 
	rcu_read_lock();

	rcu_read_lock();

	md5 = tcp_rsk(req)->af_specific->req_md5_lookup(sk, req_to_sk(req));

	md5 = tcp_rsk(req)->af_specific->req_md5_lookup(sk, req_to_sk(req));

#endif

#endif

	skb_set_hash(skb, tcp_rsk(req)->txhash, PKT_HASH_TYPE_L4);

	tcp_header_size = tcp_synack_options(sk, req, mss, skb, &opts, md5,

	tcp_header_size = tcp_synack_options(sk, req, mss, skb, &opts, md5,

					     foc) + sizeof(*th);

					     foc) + sizeof(*th);
@@ -3505,6 +3506,7 @@ int tcp_rtx_synack(struct sock *sk, struct request_sock *req) 
	struct flowi fl;

	struct flowi fl;

	int res;

	int res;





	tcp_rsk(req)->txhash = net_tx_rndhash();

	res = af_ops->send_synack(sk, NULL, &fl, req, 0, NULL);

	res = af_ops->send_synack(sk, NULL, &fl, req, 0, NULL);

	if (!res) {

	if (!res) {

		TCP_INC_STATS_BH(sock_net(sk), TCP_MIB_RETRANSSEGS);

		TCP_INC_STATS_BH(sock_net(sk), TCP_MIB_RETRANSSEGS);

CRA Git | Maintained and supported by SUSTech CRA and CCSE