tcp: propagate MPTCP skb extensions on xmit splits

When the TCP stack splits a packet on the write queue, the tail
half currently lose the associated skb extensions, and will not
carry the DSM on the wire.

The above does not cause functional problems and is allowed by
the RFC, but interact badly with GRO and RX coalescing, as possible
candidates for aggregation will carry different TCP options.

This change tries to improve the MPTCP behavior, propagating the
skb extensions on split.

Additionally, we must prevent the MPTCP stack from updating the
mapping after the split occur: that will both violate the RFC and
fool the reader.

Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

diff --git a/include/net/mptcp.h b/include/net/mptcp.h
index 753ba7e755d6d1427634110ee975ab61474f3420..6e706d838e4ebd593f5f896dae16d553de670ef0 100644 (file)
--- a/include/net/mptcp.h
+++ b/include/net/mptcp.h
@@ -29,7 +29,8 @@ struct mptcp_ext {
                        use_ack:1,
                        ack64:1,
                        mpc_map:1,
-                       __unused:2;
+                       frozen:1,
+                       __unused:1;
        /* one byte hole */
 };
 
@@ -106,6 +107,19 @@ static inline void mptcp_skb_ext_move(struct sk_buff *to,
        from->active_extensions = 0;
 }
 
+static inline void mptcp_skb_ext_copy(struct sk_buff *to,
+                                     struct sk_buff *from)
+{
+       struct mptcp_ext *from_ext;
+
+       from_ext = skb_ext_find(from, SKB_EXT_MPTCP);
+       if (!from_ext)
+               return;
+
+       from_ext->frozen = 1;
+       skb_ext_copy(to, from);
+}
+
 static inline bool mptcp_ext_matches(const struct mptcp_ext *to_ext,
                                     const struct mptcp_ext *from_ext)
 {
@@ -193,6 +207,11 @@ static inline void mptcp_skb_ext_move(struct sk_buff *to,
 {
 }
 
+static inline void mptcp_skb_ext_copy(struct sk_buff *to,
+                                     struct sk_buff *from)
+{
+}
+
 static inline bool mptcp_skb_can_collapse(const struct sk_buff *to,
                                          const struct sk_buff *from)
 {

diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index bf48cd73e96787a14b6f9af8beddb1067a7cb8dc..9abf5a0358d54869aa91a1e1a3125c1072af7305 100644 (file)
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -1569,6 +1569,7 @@ int tcp_fragment(struct sock *sk, enum tcp_queue tcp_queue,
        if (!buff)
                return -ENOMEM; /* We'll just try again later. */
        skb_copy_decrypted(buff, skb);
+       mptcp_skb_ext_copy(buff, skb);
 
        sk_wmem_queued_add(sk, buff->truesize);
        sk_mem_charge(sk, buff->truesize);
@@ -2123,6 +2124,7 @@ static int tso_fragment(struct sock *sk, struct sk_buff *skb, unsigned int len,
        if (unlikely(!buff))
                return -ENOMEM;
        skb_copy_decrypted(buff, skb);
+       mptcp_skb_ext_copy(buff, skb);
 
        sk_wmem_queued_add(sk, buff->truesize);
        sk_mem_charge(sk, buff->truesize);
@@ -2393,6 +2395,7 @@ static int tcp_mtu_probe(struct sock *sk)
 
        skb = tcp_send_head(sk);
        skb_copy_decrypted(nskb, skb);
+       mptcp_skb_ext_copy(nskb, skb);
 
        TCP_SKB_CB(nskb)->seq = TCP_SKB_CB(skb)->seq;
        TCP_SKB_CB(nskb)->end_seq = TCP_SKB_CB(skb)->seq + probe_size;

diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c
index f5bacfc55006758cc18294ad48c4bc3788a45ca7..25d12183d1ca5e15a97ac80f5c4069be7109f36d 100644 (file)
--- a/net/mptcp/protocol.c
+++ b/net/mptcp/protocol.c
@@ -771,8 +771,11 @@ static bool mptcp_skb_can_collapse_to(u64 write_seq,
        if (!tcp_skb_can_collapse_to(skb))
                return false;
 
-       /* can collapse only if MPTCP level sequence is in order */
-       return mpext && mpext->data_seq + mpext->data_len == write_seq;
+       /* can collapse only if MPTCP level sequence is in order and this
+        * mapping has not been xmitted yet
+        */
+       return mpext && mpext->data_seq + mpext->data_len == write_seq &&
+              !mpext->frozen;
 }
 
 static bool mptcp_frag_can_collapse_to(const struct mptcp_sock *msk,