summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorChristoph Paasch <cpaasch@openai.com>2025-08-15 19:28:19 +0200
committerJakub Kicinski <kuba@kernel.org>2025-08-18 17:39:58 -0700
commitccab044697980c6c01ab51f43f48f13b8a3e5c33 (patch)
treedafc3d3bc4ba75583935b0ef260322cfe81c927c
parent84967deee9d9870b15bc4c3acb50f1d401807902 (diff)
mptcp: drop skb if MPTCP skb extension allocation fails
When skb_ext_add(skb, SKB_EXT_MPTCP) fails in mptcp_incoming_options(), we used to return true, letting the segment proceed through the TCP receive path without a DSS mapping. Such segments can leave inconsistent mapping state and trigger a mid-stream fallback to TCP, which in testing collapsed (by artificially forcing failures in skb_ext_add) throughput to zero. Return false instead so the TCP input path drops the skb (see tcp_data_queue() and step-7 processing). This is the safer choice under memory pressure: it preserves MPTCP correctness and provides backpressure to the sender. Control packets remain unaffected: ACK updates and DATA_FIN handling happen before attempting the extension allocation, and tcp_reset() continues to ignore the return value. With this change, MPTCP continues to work at high throughput if we artificially inject failures into skb_ext_add. Fixes: 6787b7e350d3 ("mptcp: avoid processing packet if a subflow reset") Cc: stable@vger.kernel.org Signed-off-by: Christoph Paasch <cpaasch@openai.com> Reviewed-by: Matthieu Baerts (NGI0) <matttbe@kernel.org> Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org> Link: https://patch.msgid.link/20250815-net-mptcp-misc-fixes-6-17-rc2-v1-1-521fe9957892@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>
-rw-r--r--net/mptcp/options.c6
1 files changed, 4 insertions, 2 deletions
diff --git a/net/mptcp/options.c b/net/mptcp/options.c
index 70c0ab0ecf90..2a8ea28442b2 100644
--- a/net/mptcp/options.c
+++ b/net/mptcp/options.c
@@ -1118,7 +1118,9 @@ static bool add_addr_hmac_valid(struct mptcp_sock *msk,
return hmac == mp_opt->ahmac;
}
-/* Return false if a subflow has been reset, else return true */
+/* Return false in case of error (or subflow has been reset),
+ * else return true.
+ */
bool mptcp_incoming_options(struct sock *sk, struct sk_buff *skb)
{
struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(sk);
@@ -1222,7 +1224,7 @@ bool mptcp_incoming_options(struct sock *sk, struct sk_buff *skb)
mpext = skb_ext_add(skb, SKB_EXT_MPTCP);
if (!mpext)
- return true;
+ return false;
memset(mpext, 0, sizeof(*mpext));