在 2024/6/21 03:14, Kuniyuki Iwashima 写道:
From: luoxuanqiang <luoxuanqiang@xxxxxxxxxx> Date: Thu, 20 Jun 2024 10:38:36 +0800在 2024/6/20 03:53, Kuniyuki Iwashima 写道:From: luoxuanqiang <luoxuanqiang@xxxxxxxxxx> Date: Wed, 19 Jun 2024 14:54:15 +0800在 2024/6/18 01:59, Kuniyuki Iwashima 写道:From: luoxuanqiang <luoxuanqiang@xxxxxxxxxx> Date: Mon, 17 Jun 2024 15:56:40 +0800When bonding is configured in BOND_MODE_BROADCAST mode, if two identical SYN packets are received at the same time and processed on different CPUs, it can potentially create the same sk (sock) but two different reqsk (request_sock) in tcp_conn_request(). These two different reqsk will respond with two SYNACK packets, and since the generation of the seq (ISN) incorporates a timestamp, the final two SYNACK packets will have different seq values. The consequence is that when the Client receives and replies with an ACK to the earlier SYNACK packet, we will reset(RST) it. ======================================================================== This behavior is consistently reproducible in my local setup, which comprises: | NETA1 ------ NETB1 | PC_A --- bond --- | | --- bond --- PC_B | NETA2 ------ NETB2 | - PC_A is the Server and has two network cards, NETA1 and NETA2. I have bonded these two cards using BOND_MODE_BROADCAST mode and configured them to be handled by different CPU. - PC_B is the Client, also equipped with two network cards, NETB1 and NETB2, which are also bonded and configured in BOND_MODE_BROADCAST mode. If the client attempts a TCP connection to the server, it might encounter a failure. Capturing packets from the server side reveals: 10.10.10.10.45182 > localhost: Flags [S], seq 320236027, 10.10.10.10.45182 > localhost: Flags [S], seq 320236027, localhost > 10.10.10.10.45182: Flags [S.], seq 2967855116, localhost > 10.10.10.10.45182: Flags [S.], seq 2967855123, <== 10.10.10.10.45182 > localhost: Flags [.], ack 4294967290, 10.10.10.10.45182 > localhost: Flags [.], ack 4294967290, localhost > 10.10.10.10.45182: Flags [R], seq 2967855117, <== localhost > 10.10.10.10.45182: Flags [R], seq 2967855117, Two SYNACKs with different seq numbers are sent by localhost, resulting in an anomaly. ======================================================================== The attempted solution is as follows: In the tcp_conn_request(), while inserting reqsk into the ehash table, it also checks if an entry already exists. If found, it avoids reinsertion and releases it. Simultaneously, In the reqsk_queue_hash_req(), the start of the req->rsk_timer is adjusted to be after successful insertion. Signed-off-by: luoxuanqiang <luoxuanqiang@xxxxxxxxxx> --- include/net/inet_connection_sock.h | 4 ++-- net/dccp/ipv4.c | 2 +- net/dccp/ipv6.c | 2 +- net/ipv4/inet_connection_sock.c | 19 +++++++++++++------ net/ipv4/tcp_input.c | 9 ++++++++- 5 files changed, 25 insertions(+), 11 deletions(-) diff --git a/include/net/inet_connection_sock.h b/include/net/inet_connection_sock.h index 7d6b1254c92d..8ebab6220dbc 100644 --- a/include/net/inet_connection_sock.h +++ b/include/net/inet_connection_sock.h @@ -263,8 +263,8 @@ struct dst_entry *inet_csk_route_child_sock(const struct sock *sk, struct sock *inet_csk_reqsk_queue_add(struct sock *sk, struct request_sock *req, struct sock *child); -void inet_csk_reqsk_queue_hash_add(struct sock *sk, struct request_sock *req, - unsigned long timeout); +bool inet_csk_reqsk_queue_hash_add(struct sock *sk, struct request_sock *req, + unsigned long timeout, bool *found_dup_sk); struct sock *inet_csk_complete_hashdance(struct sock *sk, struct sock *child, struct request_sock *req, bool own_req); diff --git a/net/dccp/ipv4.c b/net/dccp/ipv4.c index ff41bd6f99c3..13aafdeb9205 100644 --- a/net/dccp/ipv4.c +++ b/net/dccp/ipv4.c @@ -657,7 +657,7 @@ int dccp_v4_conn_request(struct sock *sk, struct sk_buff *skb) if (dccp_v4_send_response(sk, req)) goto drop_and_free;- inet_csk_reqsk_queue_hash_add(sk, req, DCCP_TIMEOUT_INIT);+ inet_csk_reqsk_queue_hash_add(sk, req, DCCP_TIMEOUT_INIT, NULL); reqsk_put(req); return 0;diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.cindex 85f4b8fdbe5e..493cdb12ce2b 100644 --- a/net/dccp/ipv6.c +++ b/net/dccp/ipv6.c @@ -400,7 +400,7 @@ static int dccp_v6_conn_request(struct sock *sk, struct sk_buff *skb) if (dccp_v6_send_response(sk, req)) goto drop_and_free;- inet_csk_reqsk_queue_hash_add(sk, req, DCCP_TIMEOUT_INIT);+ inet_csk_reqsk_queue_hash_add(sk, req, DCCP_TIMEOUT_INIT, NULL); reqsk_put(req); return 0;diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.cindex d81f74ce0f02..2fa9b33ae26a 100644 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -1122,25 +1122,32 @@ static void reqsk_timer_handler(struct timer_list *t) inet_csk_reqsk_queue_drop_and_put(oreq->rsk_listener, oreq); }-static void reqsk_queue_hash_req(struct request_sock *req,- unsigned long timeout) +static bool reqsk_queue_hash_req(struct request_sock *req, + unsigned long timeout, bool *found_dup_sk) {Given any changes here in reqsk_queue_hash_req() conflicts with 4.19 (oldest stable) and DCCP does not check found_dup_sk, you can define found_dup_sk here, then you need not touch DCCP at all.Apologies for not fully understanding your advice. If we cannot modify the content of reqsk_queue_hash_req() and should avoid touching the DCCP part, it seems the issue requires reworking some interfaces. Specifically: The call flow to add reqsk to ehash is as follows: tcp_conn_request() dccp_v4(6)_conn_request() -> inet_csk_reqsk_queue_hash_add() -> reqsk_queue_hash_req() -> inet_ehash_insert() tcp_conn_request() needs to call the same interface inet_csk_reqsk_queue_hash_add() as dccp_v4(6)_conn_request(), but the critical section for installation check and insertion into ehash is within inet_ehash_insert(). If reqsk_queue_hash_req() should not be modified, then we need to rewrite the interfaces to distinguish them. I don't see how redefining found_dup_sk alone can resolve this conflict point.I just said we cannot avoid conflict so suggested avoiding found_dup_sk in inet_csk_reqsk_queue_hash_add(). But I finally ended up modifying DCCP because we return before setting refcnt. ---8<--- diff --git a/net/dccp/ipv4.c b/net/dccp/ipv4.c index ff41bd6f99c3..b2a8aed35eb0 100644 --- a/net/dccp/ipv4.c +++ b/net/dccp/ipv4.c @@ -657,8 +657,11 @@ int dccp_v4_conn_request(struct sock *sk, struct sk_buff *skb) if (dccp_v4_send_response(sk, req)) goto drop_and_free;- inet_csk_reqsk_queue_hash_add(sk, req, DCCP_TIMEOUT_INIT);- reqsk_put(req); + if (unlikely(inet_csk_reqsk_queue_hash_add(sk, req, DCCP_TIMEOUT_INIT))) + reqsk_free(req); + else + reqsk_put(req); + return 0;drop_and_free:diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c index d81f74ce0f02..7dd6892b10b9 100644 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -1122,25 +1122,33 @@ static void reqsk_timer_handler(struct timer_list *t) inet_csk_reqsk_queue_drop_and_put(oreq->rsk_listener, oreq); }-static void reqsk_queue_hash_req(struct request_sock *req,+static bool reqsk_queue_hash_req(struct request_sock *req, unsigned long timeout) { + bool found_dup_sk; + + if (!inet_ehash_insert(req_to_sk(req), NULL, &found_dup_sk)) + return false; + timer_setup(&req->rsk_timer, reqsk_timer_handler, TIMER_PINNED); mod_timer(&req->rsk_timer, jiffies + timeout);- inet_ehash_insert(req_to_sk(req), NULL, NULL);/* before letting lookups find us, make sure all req fields * are committed to memory and refcnt initialized. */ smp_wmb(); refcount_set(&req->rsk_refcnt, 2 + 1); + return true; }-void inet_csk_reqsk_queue_hash_add(struct sock *sk, struct request_sock *req,+bool inet_csk_reqsk_queue_hash_add(struct sock *sk, struct request_sock *req, unsigned long timeout) { - reqsk_queue_hash_req(req, timeout); + if (!reqsk_queue_hash_req(req, timeout)) + return false; + inet_csk_reqsk_queue_added(sk); + return true; } EXPORT_SYMBOL_GPL(inet_csk_reqsk_queue_hash_add);---8<---Thank you for your patient explanation. I understand your point and will send out the V4 version, looking forward to your review. Also, I'd like to confirm a detail with you. For the DCCP part, is it sufficient to simply call reqsk_free() for the return value, or should we use goto drop_and_free? The different return values here will determine whether a reset is sent, and I lack a comprehensive understanding of DCCP. so could you please help me confirm this from a higher-level perspective?I think we need not send RST because in this case there's another establishing request already in ehash and we should not abort it as we don't do so in tcp_conn_request(). Thanks!
Thank you for confirming. Best wishes! :)
diff --git a/net/dccp/ipv4.c b/net/dccp/ipv4.c index ff41bd6f99c3..5926159a6f20 100644 --- a/net/dccp/ipv4.c +++ b/net/dccp/ipv4.c @@ -657,8 +657,11 @@ int dccp_v4_conn_request(struct sock *sk, struct sk_buff *skb) if (dccp_v4_send_response(sk, req)) goto drop_and_free;- inet_csk_reqsk_queue_hash_add(sk, req, DCCP_TIMEOUT_INIT);- reqsk_put(req); + if (unlikely(!inet_csk_reqsk_queue_hash_add(sk, req, DCCP_TIMEOUT_INIT))) + reqsk_free(req); // or goto drop_and_free: <============ + else + reqsk_put(req); + return 0;drop_and_free:reqsk_free(req); drop: __DCCP_INC_STATS(DCCP_MIB_ATTEMPTFAILS); return -1; }
