diff mbox series

[net-next,v2,3/7] tcp: propagate MPTCP skb extensions on xmit splits

Message ID 20201103190509.27416-4-mathew.j.martineau@linux.intel.com
State Accepted
Delegated to: David Miller
Headers show
Series mptcp: Miscellaneous MPTCP fixes | expand

Checks

Context Check Description
jkicinski/cover_letter success Link
jkicinski/fixes_present success Link
jkicinski/patch_count success Link
jkicinski/tree_selection success Clearly marked for net-next
jkicinski/subject_prefix success Link
jkicinski/source_inline success Was 0 now: 0
jkicinski/verify_signedoff success Link
jkicinski/module_param success Was 0 now: 0
jkicinski/build_32bit success Errors and warnings before: 1527 this patch: 1527
jkicinski/kdoc success Errors and warnings before: 0 this patch: 0
jkicinski/verify_fixes success Link
jkicinski/checkpatch success total: 0 errors, 0 warnings, 0 checks, 73 lines checked
jkicinski/build_allmodconfig_warn success Errors and warnings before: 1513 this patch: 1513
jkicinski/header_inline success Link
jkicinski/stable success Stable not CCed

Commit Message

Mat Martineau Nov. 3, 2020, 7:05 p.m. UTC 
From: Paolo Abeni <pabeni@redhat.com>

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>
---
 include/net/mptcp.h   | 21 ++++++++++++++++++++-
 net/ipv4/tcp_output.c |  3 +++
 net/mptcp/protocol.c  |  7 +++++--
 3 files changed, 28 insertions(+), 3 deletions(-)
diff mbox series

Patch

diff --git a/include/net/mptcp.h b/include/net/mptcp.h
index 753ba7e755d6..6e706d838e4e 100644
--- 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 bf48cd73e967..9abf5a0358d5 100644
--- 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 f5bacfc55006..25d12183d1ca 100644
--- 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,