Merge branch 'mptcp-misc-fixes-involving-fallback-to-tcp'

Matthieu Baerts says:

====================
mptcp: misc. fixes involving fallback to TCP

- Patch 1: better handle DSS corruptions from a bugged peer: reducing
  warnings, doing a fallback or a reset depending on the subflow state.
  For >= v5.7.

- Patch 2: fix DSS corruption due to large pmtu xmit, where MPTCP was
  not taken into account. For >= v5.6.

- Patch 3: fallback when MPTCP opts are dropped after the first data
  packet, instead of resetting the connection. For >= v5.6.

- Patch 4: restrict the removal of a subflow to other closing states, a
  better fix, for a recent one. For >= v5.10.
====================

Link: https://patch.msgid.link/20241008-net-mptcp-fallback-fixes-v1-0-c6fb8e93e551@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

6 files changed, 32 insertions, 10 deletions

diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index 4fd746bd4d54..68804fd01daf 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -2342,10 +2342,7 @@ static bool tcp_can_coalesce_send_queue_head(struct sock *sk, int len)
 		if (len <= skb->len)
 			break;
 
-		if (unlikely(TCP_SKB_CB(skb)->eor) ||
-		    tcp_has_tx_tstamp(skb) ||
-		    !skb_pure_zcopy_same(skb, next) ||
-		    skb_frags_readable(skb) != skb_frags_readable(next))
+		if (tcp_has_tx_tstamp(skb) || !tcp_skb_can_collapse(skb, next))
 			return false;
 
 		len -= skb->len;
diff --git a/net/mptcp/mib.c b/net/mptcp/mib.c
index 38c2efc82b94..ad88bd3c58df 100644
--- a/net/mptcp/mib.c
+++ b/net/mptcp/mib.c
@@ -32,6 +32,8 @@ static const struct snmp_mib mptcp_snmp_list[] = {
 	SNMP_MIB_ITEM("MPJoinSynTxBindErr", MPTCP_MIB_JOINSYNTXBINDERR),
 	SNMP_MIB_ITEM("MPJoinSynTxConnectErr", MPTCP_MIB_JOINSYNTXCONNECTERR),
 	SNMP_MIB_ITEM("DSSNotMatching", MPTCP_MIB_DSSNOMATCH),
+	SNMP_MIB_ITEM("DSSCorruptionFallback", MPTCP_MIB_DSSCORRUPTIONFALLBACK),
+	SNMP_MIB_ITEM("DSSCorruptionReset", MPTCP_MIB_DSSCORRUPTIONRESET),
 	SNMP_MIB_ITEM("InfiniteMapTx", MPTCP_MIB_INFINITEMAPTX),
 	SNMP_MIB_ITEM("InfiniteMapRx", MPTCP_MIB_INFINITEMAPRX),
 	SNMP_MIB_ITEM("DSSNoMatchTCP", MPTCP_MIB_DSSTCPMISMATCH),
diff --git a/net/mptcp/mib.h b/net/mptcp/mib.h
index c8ffe18a8722..3206cdda8bb1 100644
--- a/net/mptcp/mib.h
+++ b/net/mptcp/mib.h
@@ -27,6 +27,8 @@ enum linux_mptcp_mib_field {
 	MPTCP_MIB_JOINSYNTXBINDERR,	/* Not able to bind() the address when sending a SYN + MP_JOIN */
 	MPTCP_MIB_JOINSYNTXCONNECTERR,	/* Not able to connect() when sending a SYN + MP_JOIN */
 	MPTCP_MIB_DSSNOMATCH,		/* Received a new mapping that did not match the previous one */
+	MPTCP_MIB_DSSCORRUPTIONFALLBACK,/* DSS corruption detected, fallback */
+	MPTCP_MIB_DSSCORRUPTIONRESET,	/* DSS corruption detected, MPJ subflow reset */
 	MPTCP_MIB_INFINITEMAPTX,	/* Sent an infinite mapping */
 	MPTCP_MIB_INFINITEMAPRX,	/* Received an infinite mapping */
 	MPTCP_MIB_DSSTCPMISMATCH,	/* DSS-mapping did not map with TCP's sequence numbers */
diff --git a/net/mptcp/pm_netlink.c b/net/mptcp/pm_netlink.c
index 64fe0e7d87d7..f6f0a38a0750 100644
--- a/net/mptcp/pm_netlink.c
+++ b/net/mptcp/pm_netlink.c
@@ -860,7 +860,8 @@ static void mptcp_pm_nl_rm_addr_or_subflow(struct mptcp_sock *msk,
 			int how = RCV_SHUTDOWN | SEND_SHUTDOWN;
 			u8 id = subflow_get_local_id(subflow);
 
-			if (inet_sk_state_load(ssk) == TCP_CLOSE)
+			if ((1 << inet_sk_state_load(ssk)) &
+			    (TCPF_FIN_WAIT1 | TCPF_FIN_WAIT2 | TCPF_CLOSING | TCPF_CLOSE))
 				continue;
 			if (rm_type == MPTCP_MIB_RMADDR && remote_id != rm_id)
 				continue;
diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c
index c2317919fc14..6d0e201c3eb2 100644
--- a/net/mptcp/protocol.c
+++ b/net/mptcp/protocol.c
@@ -620,6 +620,18 @@ static bool mptcp_check_data_fin(struct sock *sk)
 	return ret;
 }
 
+static void mptcp_dss_corruption(struct mptcp_sock *msk, struct sock *ssk)
+{
+	if (READ_ONCE(msk->allow_infinite_fallback)) {
+		MPTCP_INC_STATS(sock_net(ssk),
+				MPTCP_MIB_DSSCORRUPTIONFALLBACK);
+		mptcp_do_fallback(ssk);
+	} else {
+		MPTCP_INC_STATS(sock_net(ssk), MPTCP_MIB_DSSCORRUPTIONRESET);
+		mptcp_subflow_reset(ssk);
+	}
+}
+
 static bool __mptcp_move_skbs_from_subflow(struct mptcp_sock *msk,
 					   struct sock *ssk,
 					   unsigned int *bytes)
@@ -692,10 +704,16 @@ static bool __mptcp_move_skbs_from_subflow(struct mptcp_sock *msk,
 				moved += len;
 			seq += len;
 
-			if (WARN_ON_ONCE(map_remaining < len))
-				break;
+			if (unlikely(map_remaining < len)) {
+				DEBUG_NET_WARN_ON_ONCE(1);
+				mptcp_dss_corruption(msk, ssk);
+			}
 		} else {
-			WARN_ON_ONCE(!fin);
+			if (unlikely(!fin)) {
+				DEBUG_NET_WARN_ON_ONCE(1);
+				mptcp_dss_corruption(msk, ssk);
+			}
+
 			sk_eat_skb(ssk, skb);
 			done = true;
 		}
diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c
index 1040b3b9696b..25dde81bcb75 100644
--- a/net/mptcp/subflow.c
+++ b/net/mptcp/subflow.c
@@ -975,8 +975,10 @@ static bool skb_is_fully_mapped(struct sock *ssk, struct sk_buff *skb)
 	unsigned int skb_consumed;
 
 	skb_consumed = tcp_sk(ssk)->copied_seq - TCP_SKB_CB(skb)->seq;
-	if (WARN_ON_ONCE(skb_consumed >= skb->len))
+	if (unlikely(skb_consumed >= skb->len)) {
+		DEBUG_NET_WARN_ON_ONCE(1);
 		return true;
+	}
 
 	return skb->len - skb_consumed <= subflow->map_data_len -
 					  mptcp_subflow_get_map_offset(subflow);
@@ -1280,7 +1282,7 @@ static bool subflow_can_fallback(struct mptcp_subflow_context *subflow)
 	else if (READ_ONCE(msk->csum_enabled))
 		return !subflow->valid_csum_seen;
 	else
-		return !subflow->fully_established;
+		return READ_ONCE(msk->allow_infinite_fallback);
 }
 
 static void mptcp_subflow_fail(struct mptcp_sock *msk, struct sock *ssk)