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: 3b24d854cb
("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>
This commit is contained in:

committed by
David S. Miller

parent
ac18dd9e84
commit
b3d051477c
@@ -6327,7 +6327,7 @@ int tcp_conn_request(struct request_sock_ops *rsk_ops,
|
||||
}
|
||||
if (fastopen_sk) {
|
||||
af_ops->send_synack(fastopen_sk, dst, &fl, req,
|
||||
&foc, false);
|
||||
&foc, TCP_SYNACK_FASTOPEN);
|
||||
/* Add the child socket directly into the accept queue */
|
||||
inet_csk_reqsk_queue_add(sk, req, fastopen_sk);
|
||||
sk->sk_data_ready(sk);
|
||||
@@ -6337,8 +6337,9 @@ int tcp_conn_request(struct request_sock_ops *rsk_ops,
|
||||
tcp_rsk(req)->tfo_listener = false;
|
||||
if (!want_cookie)
|
||||
inet_csk_reqsk_queue_hash_add(sk, req, TCP_TIMEOUT_INIT);
|
||||
af_ops->send_synack(sk, dst, &fl, req,
|
||||
&foc, !want_cookie);
|
||||
af_ops->send_synack(sk, dst, &fl, req, &foc,
|
||||
!want_cookie ? TCP_SYNACK_NORMAL :
|
||||
TCP_SYNACK_COOKIE);
|
||||
if (want_cookie) {
|
||||
reqsk_free(req);
|
||||
return 0;
|
||||
|
Reference in New Issue
Block a user