Merge branch 'mptcp-refactor-inet_accept-and-mib-updates'

Mat Martineau says:

====================
mptcp: Refactor inet_accept() and MIB updates

Patches 1 and 2 refactor inet_accept() to provide a new __inet_accept()
that's usable with locked sockets, and then make use of that helper to
simplify mptcp_stream_accept().

Patches 3 and 4 add some new MIBS related to MPTCP address advertisement
and update related selftest scripts.

Patch 5 modifies the selftests to ensure MIBS are only printed once when
a test case fails.
====================

Link: https://lore.kernel.org/r/20230516-send-net-next-20220516-v1-0-e91822b7b6e0@kernel.orgSigned-off-by: Jakub Kicinski <kuba@kernel.org>

include/net/inet_common.h

@@ -31,6 +31,8 @@ int inet_dgram_connect(struct socket *sock, struct sockaddr *uaddr,
 		       int addr_len, int flags);
 int inet_accept(struct socket *sock, struct socket *newsock, int flags,
 		bool kern);
+void __inet_accept(struct socket *sock, struct socket *newsock,
+		   struct sock *newsk);
 int inet_send_prepare(struct sock *sk);
 int inet_sendmsg(struct socket *sock, struct msghdr *msg, size_t size);
 ssize_t inet_sendpage(struct socket *sock, struct page *page, int offset,

net/ipv4/af_inet.c

@@ -730,6 +730,20 @@ int inet_stream_connect(struct socket *sock, struct sockaddr *uaddr,
 }
 EXPORT_SYMBOL(inet_stream_connect);
 
+void __inet_accept(struct socket *sock, struct socket *newsock, struct sock *newsk)
+{
+	sock_rps_record_flow(newsk);
+	WARN_ON(!((1 << newsk->sk_state) &
+		  (TCPF_ESTABLISHED | TCPF_SYN_RECV |
+		  TCPF_CLOSE_WAIT | TCPF_CLOSE)));
+
+	if (test_bit(SOCK_SUPPORT_ZC, &sock->flags))
+		set_bit(SOCK_SUPPORT_ZC, &newsock->flags);
+	sock_graft(newsk, newsock);
+
+	newsock->state = SS_CONNECTED;
+}
+
 /*
  *	Accept a pending connection. The TCP layer now gives BSD semantics.
  */
@@ -743,24 +757,12 @@ int inet_accept(struct socket *sock, struct socket *newsock, int flags,
 	/* IPV6_ADDRFORM can change sk->sk_prot under us. */
 	sk2 = READ_ONCE(sk1->sk_prot)->accept(sk1, flags, &err, kern);
 	if (!sk2)
-		goto do_err;
+		return err;
 
 	lock_sock(sk2);
-
-	sock_rps_record_flow(sk2);
-	WARN_ON(!((1 << sk2->sk_state) &
-		  (TCPF_ESTABLISHED | TCPF_SYN_RECV |
-		  TCPF_CLOSE_WAIT | TCPF_CLOSE)));
-
-	if (test_bit(SOCK_SUPPORT_ZC, &sock->flags))
-		set_bit(SOCK_SUPPORT_ZC, &newsock->flags);
-	sock_graft(sk2, newsock);
-
-	newsock->state = SS_CONNECTED;
-	err = 0;
+	__inet_accept(sock, newsock, sk2);
 	release_sock(sk2);
-do_err:
-	return err;
+	return 0;
 }
 EXPORT_SYMBOL(inet_accept);
 

net/mptcp/mib.c

@@ -34,7 +34,11 @@ static const struct snmp_mib mptcp_snmp_list[] = {
 	SNMP_MIB_ITEM("NoDSSInWindow", MPTCP_MIB_NODSSWINDOW),
 	SNMP_MIB_ITEM("DuplicateData", MPTCP_MIB_DUPDATA),
 	SNMP_MIB_ITEM("AddAddr", MPTCP_MIB_ADDADDR),
+	SNMP_MIB_ITEM("AddAddrTx", MPTCP_MIB_ADDADDRTX),
+	SNMP_MIB_ITEM("AddAddrTxDrop", MPTCP_MIB_ADDADDRTXDROP),
 	SNMP_MIB_ITEM("EchoAdd", MPTCP_MIB_ECHOADD),
+	SNMP_MIB_ITEM("EchoAddTx", MPTCP_MIB_ECHOADDTX),
+	SNMP_MIB_ITEM("EchoAddTxDrop", MPTCP_MIB_ECHOADDTXDROP),
 	SNMP_MIB_ITEM("PortAdd", MPTCP_MIB_PORTADD),
 	SNMP_MIB_ITEM("AddAddrDrop", MPTCP_MIB_ADDADDRDROP),
 	SNMP_MIB_ITEM("MPJoinPortSynRx", MPTCP_MIB_JOINPORTSYNRX),
@@ -44,6 +48,8 @@ static const struct snmp_mib mptcp_snmp_list[] = {
 	SNMP_MIB_ITEM("MismatchPortAckRx", MPTCP_MIB_MISMATCHPORTACKRX),
 	SNMP_MIB_ITEM("RmAddr", MPTCP_MIB_RMADDR),
 	SNMP_MIB_ITEM("RmAddrDrop", MPTCP_MIB_RMADDRDROP),
+	SNMP_MIB_ITEM("RmAddrTx", MPTCP_MIB_RMADDRTX),
+	SNMP_MIB_ITEM("RmAddrTxDrop", MPTCP_MIB_RMADDRTXDROP),
 	SNMP_MIB_ITEM("RmSubflow", MPTCP_MIB_RMSUBFLOW),
 	SNMP_MIB_ITEM("MPPrioTx", MPTCP_MIB_MPPRIOTX),
 	SNMP_MIB_ITEM("MPPrioRx", MPTCP_MIB_MPPRIORX),

net/mptcp/mib.h

@@ -27,7 +27,15 @@ enum linux_mptcp_mib_field {
 	MPTCP_MIB_NODSSWINDOW,		/* Segments not in MPTCP windows */
 	MPTCP_MIB_DUPDATA,		/* Segments discarded due to duplicate DSS */
 	MPTCP_MIB_ADDADDR,		/* Received ADD_ADDR with echo-flag=0 */
+	MPTCP_MIB_ADDADDRTX,		/* Sent ADD_ADDR with echo-flag=0 */
+	MPTCP_MIB_ADDADDRTXDROP,	/* ADD_ADDR with echo-flag=0 not send due to
+					 * resource exhaustion
+					 */
 	MPTCP_MIB_ECHOADD,		/* Received ADD_ADDR with echo-flag=1 */
+	MPTCP_MIB_ECHOADDTX,		/* Send ADD_ADDR with echo-flag=1 */
+	MPTCP_MIB_ECHOADDTXDROP,	/* ADD_ADDR with echo-flag=1 not send due
+					 * to resource exhaustion
+					 */
 	MPTCP_MIB_PORTADD,		/* Received ADD_ADDR with a port-number */
 	MPTCP_MIB_ADDADDRDROP,		/* Dropped incoming ADD_ADDR */
 	MPTCP_MIB_JOINPORTSYNRX,	/* Received a SYN MP_JOIN with a different port-number */
@@ -37,6 +45,8 @@ enum linux_mptcp_mib_field {
 	MPTCP_MIB_MISMATCHPORTACKRX,	/* Received an ACK MP_JOIN with a mismatched port-number */
 	MPTCP_MIB_RMADDR,		/* Received RM_ADDR */
 	MPTCP_MIB_RMADDRDROP,		/* Dropped incoming RM_ADDR */
+	MPTCP_MIB_RMADDRTX,		/* Sent RM_ADDR */
+	MPTCP_MIB_RMADDRTXDROP,		/* RM_ADDR not sent due to resource exhaustion */
 	MPTCP_MIB_RMSUBFLOW,		/* Remove a subflow */
 	MPTCP_MIB_MPPRIOTX,		/* Transmit a MP_PRIO */
 	MPTCP_MIB_MPPRIORX,		/* Received a MP_PRIO */
@@ -63,6 +73,14 @@ struct mptcp_mib {
 	unsigned long mibs[LINUX_MIB_MPTCP_MAX];
 };
 
+static inline void MPTCP_ADD_STATS(struct net *net,
+				   enum linux_mptcp_mib_field field,
+				   int val)
+{
+	if (likely(net->mib.mptcp_statistics))
+		SNMP_ADD_STATS(net->mib.mptcp_statistics, field, val);
+}
+
 static inline void MPTCP_INC_STATS(struct net *net,
 				   enum linux_mptcp_mib_field field)
 {

net/mptcp/options.c

@@ -687,9 +687,12 @@ static bool mptcp_established_options_add_addr(struct sock *sk, struct sk_buff *
 	}
 	opts->suboptions |= OPTION_MPTCP_ADD_ADDR;
 	if (!echo) {
+		MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_ADDADDRTX);
 		opts->ahmac = add_addr_generate_hmac(msk->local_key,
 						     msk->remote_key,
 						     &opts->addr);
+	} else {
+		MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_ECHOADDTX);
 	}
 	pr_debug("addr_id=%d, ahmac=%llu, echo=%d, port=%d",
 		 opts->addr.id, opts->ahmac, echo, ntohs(opts->addr.port));
@@ -723,7 +726,7 @@ static bool mptcp_established_options_rm_addr(struct sock *sk,
 
 	for (i = 0; i < opts->rm_list.nr; i++)
 		pr_debug("rm_list_ids[%d]=%d", i, opts->rm_list.ids[i]);
-
+	MPTCP_ADD_STATS(sock_net(sk), MPTCP_MIB_RMADDRTX, opts->rm_list.nr);
 	return true;
 }
 

net/mptcp/pm.c

@@ -26,7 +26,8 @@ int mptcp_pm_announce_addr(struct mptcp_sock *msk,
 
 	if (add_addr &
 	    (echo ? BIT(MPTCP_ADD_ADDR_ECHO) : BIT(MPTCP_ADD_ADDR_SIGNAL))) {
-		pr_warn("addr_signal error, add_addr=%d, echo=%d", add_addr, echo);
+		MPTCP_INC_STATS(sock_net((struct sock *)msk),
+				echo ? MPTCP_MIB_ECHOADDTXDROP : MPTCP_MIB_ADDADDRTXDROP);
 		return -EINVAL;
 	}
 
@@ -48,7 +49,8 @@ int mptcp_pm_remove_addr(struct mptcp_sock *msk, const struct mptcp_rm_list *rm_
 	pr_debug("msk=%p, rm_list_nr=%d", msk, rm_list->nr);
 
 	if (rm_addr) {
-		pr_warn("addr_signal error, rm_addr=%d", rm_addr);
+		MPTCP_ADD_STATS(sock_net((struct sock *)msk),
+				MPTCP_MIB_RMADDRTXDROP, rm_list->nr);
 		return -EINVAL;
 	}
 

net/mptcp/protocol.c

@@ -3747,6 +3747,7 @@ static int mptcp_stream_accept(struct socket *sock, struct socket *newsock,
 {
 	struct mptcp_sock *msk = mptcp_sk(sock->sk);
 	struct socket *ssock;
+	struct sock *newsk;
 	int err;
 
 	pr_debug("msk=%p", msk);
@@ -3758,17 +3759,20 @@ static int mptcp_stream_accept(struct socket *sock, struct socket *newsock,
 	if (!ssock)
 		return -EINVAL;
 
-	err = ssock->ops->accept(sock, newsock, flags, kern);
-	if (err == 0 && !mptcp_is_tcpsk(newsock->sk)) {
-		struct mptcp_sock *msk = mptcp_sk(newsock->sk);
+	newsk = mptcp_accept(sock->sk, flags, &err, kern);
+	if (!newsk)
+		return err;
+
+	lock_sock(newsk);
+
+	__inet_accept(sock, newsock, newsk);
+	if (!mptcp_is_tcpsk(newsock->sk)) {
+		struct mptcp_sock *msk = mptcp_sk(newsk);
 		struct mptcp_subflow_context *subflow;
-		struct sock *newsk = newsock->sk;
 
 		set_bit(SOCK_CUSTOM_SOCKOPT, &newsock->flags);
 		msk->in_accept_queue = 0;
 
-		lock_sock(newsk);
-
 		/* set ssk->sk_socket of accept()ed flows to mptcp socket.
 		 * This is needed so NOSPACE flag can be set from tcp stack.
 		 */
@@ -3789,11 +3793,10 @@ static int mptcp_stream_accept(struct socket *sock, struct socket *newsock,
 			if (unlikely(list_empty(&msk->conn_list)))
 				inet_sk_state_store(newsk, TCP_CLOSE);
 		}
-
-		release_sock(newsk);
 	}
+	release_sock(newsk);
 
-	return err;
+	return 0;
 }
 
 static __poll_t mptcp_check_writeable(struct mptcp_sock *msk)