diff options
| author | Eric Dumazet <edumazet@google.com> | 2016-04-14 07:05:39 +0200 |
|---|---|---|
| committer | David S. Miller <davem@davemloft.net> | 2016-04-15 22:45:44 +0200 |
| commit | b3d051477cf94e9d71d6acadb8a90de15237b9c1 (patch) | |
| tree | 59009bc698f472b31b15059972e6904cd9272d32 /include/net/tcp.h | |
| parent | qlge: Replace create_singlethread_workqueue with alloc_ordered_workqueue (diff) | |
| download | linux-b3d051477cf94e9d71d6acadb8a90de15237b9c1.tar.xz linux-b3d051477cf94e9d71d6acadb8a90de15237b9c1.zip | |
tcp: do not mess with listener sk_wmem_alloc
When removing sk_refcnt manipulation on synflood, I missed that
using skb_set_owner_w() was racy, if sk->sk_wmem_alloc had already
transitioned to 0.
We should hold sk_refcnt instead, but this is a big deal under attack.
(Doing so increase performance from 3.2 Mpps to 3.8 Mpps only)
In this patch, I chose to not attach a socket to syncookies skb.
Performance is now 5 Mpps instead of 3.2 Mpps.
Following patch will remove last known false sharing in
tcp_rcv_state_process()
Fixes: 3b24d854cb35 ("tcp/dccp: do not touch listener sk_refcnt under synflood")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'include/net/tcp.h')
| -rw-r--r-- | include/net/tcp.h | 9 |
1 files changed, 7 insertions, 2 deletions
diff --git a/include/net/tcp.h b/include/net/tcp.h index 74d3ed5eb219..fd40f8c64d5f 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -452,10 +452,15 @@ struct sock *tcp_v4_syn_recv_sock(const struct sock *sk, struct sk_buff *skb, int tcp_v4_do_rcv(struct sock *sk, struct sk_buff *skb); int tcp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len); int tcp_connect(struct sock *sk); +enum tcp_synack_type { + TCP_SYNACK_NORMAL, + TCP_SYNACK_FASTOPEN, + TCP_SYNACK_COOKIE, +}; struct sk_buff *tcp_make_synack(const struct sock *sk, struct dst_entry *dst, struct request_sock *req, struct tcp_fastopen_cookie *foc, - bool attach_req); + enum tcp_synack_type synack_type); int tcp_disconnect(struct sock *sk, int flags); void tcp_finish_connect(struct sock *sk, struct sk_buff *skb); @@ -1728,7 +1733,7 @@ struct tcp_request_sock_ops { int (*send_synack)(const struct sock *sk, struct dst_entry *dst, struct flowi *fl, struct request_sock *req, struct tcp_fastopen_cookie *foc, - bool attach_req); + enum tcp_synack_type synack_type); }; #ifdef CONFIG_SYN_COOKIES |