From nobody Mon Jan 26 01:46:27 2026 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-1.web.codeaurora.org [10.30.226.201]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 7C19DA41 for ; Tue, 13 Jan 2026 06:13:55 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=10.30.226.201 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1768284835; cv=none; b=Aj/KCIvf+S77JliTYtJEiOQ/DbXOupKCvC4sBGs1wLuug4ODDhTj6E8I2SLOWiYzTqsDjD5i0XRvyS+NnufBidzp3qS/pT4SFtssB8yxGhVysRp1NpJJ00b3Qhx7cvm8OlRqLaQjF1Re8HHrlc1EcTq5rP4G3+dJHFEAReA0EZ4= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1768284835; c=relaxed/simple; bh=J9JjOTy4iHK94lhHXIPXji1D1kdYy/srKBioKY2BVjg=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=FfBxTHCY/jNGeFuBHBj5qylTZnfLrqESk8dBb+OB7sXpKveJaNwW3D0C53XSvs0gb6Vhf1BsVlzTGQRxBsueyMptbz4QJZWUUZXpZVR5VlBx8apQLdD3aAZr7KXfoqionUzmzG31ieJXRPCdsGoTz1HsDcmwNJZSqp6/YVAUhh0= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=SvdtSdBM; arc=none smtp.client-ip=10.30.226.201 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="SvdtSdBM" Received: by smtp.kernel.org (Postfix) with ESMTPSA id DC95EC116C6; Tue, 13 Jan 2026 06:13:53 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1768284835; bh=J9JjOTy4iHK94lhHXIPXji1D1kdYy/srKBioKY2BVjg=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=SvdtSdBM91GateGBWyZc1tTrKN/MncPoLQkJnZnqs5cW9EVRBzHwWkwf1walGzsQY C17/bm/MmtWhGSe1FQjhz8FD6W0sOy7NjjVwSJPV9W0i+cGgO9jzI6/tib1JWVZnC0 wLC8pnDiitFk2T57jQU3x3NGse6lrQgqJrgG3JOZxZREmMsd+HfIJi9LEyXG8xNOJs e6KKEtfWfHl1oWhGvR9VClTDJj5LKd+dfEkd+nXGBC9He9q0HtZC+xPZrf9m6gjQFo 1M+nVSonXrYKRr2F89UA4BLbMkexSZLApFmOFRBe+W3D+RjODwkwxBeaaZNpxsY5BF 1UMXEDeLEJWWQ== From: Geliang Tang To: mptcp@lists.linux.dev Cc: Geliang Tang , Gang Yan Subject: [RFC mptcp-next v7 6/9] mptcp: implement tls_mptcp_ops Date: Tue, 13 Jan 2026 14:13:32 +0800 Message-ID: <7ebe2388bbacb63ac91be1c7a9e6836c0554b292.1768284047.git.tanggeliang@kylinos.cn> X-Mailer: git-send-email 2.51.0 In-Reply-To: References: Precedence: bulk X-Mailing-List: mptcp@lists.linux.dev List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" From: Geliang Tang This patch implements the MPTCP-specific struct tls_prot_ops, named 'tls_mptcp_ops'. Note that there is a slight difference between mptcp_inq() and mptcp_inq_hint(), it does not return 1 when the socket is closed or shut down; instead, it returns 0. Otherwise, it would break the condition "inq < 1" in tls_strp_read_sock(). A direct call to mptcp_read_sock() could lead to a deadlock, as 'read_sock' interface of TLS might be invoked from within a softirq context. In such a scenario, lock_sock_fast(), which is called by mptcp_rcv_space_adjust() or mptcp_cleanup_rbuf(), would cause the deadlocks. To resolve it, use in_softirq() to determine whether to call mptcp_read_sock() or mptcp_read_sock_noack(). Passing an MPTCP socket to tcp_sock_rate_check_app_limited() can trigger a crash. Here, an MPTCP version of check_app_limited() is implemented, which calls tcp_sock_rate_check_app_limited() for each subflow. MPTCP TLS_HW mode is not yet implemented, returning EOPNOTSUPP here. Co-developed-by: Gang Yan Signed-off-by: Gang Yan Signed-off-by: Geliang Tang --- include/net/mptcp.h | 2 + include/net/tcp.h | 1 + net/ipv4/tcp_rate.c | 9 +++- net/mptcp/protocol.c | 116 ++++++++++++++++++++++++++++++++++++++++--- net/tls/tls_main.c | 6 +++ 5 files changed, 126 insertions(+), 8 deletions(-) diff --git a/include/net/mptcp.h b/include/net/mptcp.h index 4cf59e83c1c5..02564eceeb7e 100644 --- a/include/net/mptcp.h +++ b/include/net/mptcp.h @@ -132,6 +132,8 @@ struct mptcp_pm_ops { void (*release)(struct mptcp_sock *msk); } ____cacheline_aligned_in_smp; =20 +extern struct tls_prot_ops tls_mptcp_ops; + #ifdef CONFIG_MPTCP void mptcp_init(void); =20 diff --git a/include/net/tcp.h b/include/net/tcp.h index 1ff682763ed3..4b2b9daada49 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -1372,6 +1372,7 @@ void tcp_rate_skb_delivered(struct sock *sk, struct s= k_buff *skb, struct rate_sample *rs); void tcp_rate_gen(struct sock *sk, u32 delivered, u32 lost, bool is_sack_reneg, struct rate_sample *rs); +void tcp_sock_rate_check_app_limited(struct tcp_sock *tp); void tcp_rate_check_app_limited(struct sock *sk); =20 static inline bool tcp_skb_sent_after(u64 t1, u64 t2, u32 seq1, u32 seq2) diff --git a/net/ipv4/tcp_rate.c b/net/ipv4/tcp_rate.c index a8f6d9d06f2e..93bf22ae58c4 100644 --- a/net/ipv4/tcp_rate.c +++ b/net/ipv4/tcp_rate.c @@ -191,9 +191,9 @@ void tcp_rate_gen(struct sock *sk, u32 delivered, u32 l= ost, } =20 /* If a gap is detected between sends, mark the socket application-limited= . */ -void tcp_rate_check_app_limited(struct sock *sk) +void tcp_sock_rate_check_app_limited(struct tcp_sock *tp) { - struct tcp_sock *tp =3D tcp_sk(sk); + struct sock *sk =3D (struct sock *)tp; =20 if (/* We have less than one packet to send. */ tp->write_seq - tp->snd_nxt < tp->mss_cache && @@ -206,4 +206,9 @@ void tcp_rate_check_app_limited(struct sock *sk) tp->app_limited =3D (tp->delivered + tcp_packets_in_flight(tp)) ? : 1; } + +void tcp_rate_check_app_limited(struct sock *sk) +{ + tcp_sock_rate_check_app_limited(tcp_sk(sk)); +} EXPORT_SYMBOL_GPL(tcp_rate_check_app_limited); diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index e5b59310bd27..7d95c0401672 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -24,11 +24,12 @@ #include #include #include +#include #include #include "protocol.h" #include "mib.h" =20 -static unsigned int mptcp_inq_hint(const struct sock *sk); +static unsigned int mptcp_inq_hint(struct sock *sk); =20 #define CREATE_TRACE_POINTS #include @@ -1884,7 +1885,7 @@ static void mptcp_rps_record_subflows(const struct mp= tcp_sock *msk) } } =20 -static int mptcp_sendmsg(struct sock *sk, struct msghdr *msg, size_t len) +static int mptcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_= t len) { struct mptcp_sock *msk =3D mptcp_sk(sk); struct page_frag *pfrag; @@ -1895,8 +1896,6 @@ static int mptcp_sendmsg(struct sock *sk, struct msgh= dr *msg, size_t len) /* silently ignore everything else */ msg->msg_flags &=3D MSG_MORE | MSG_DONTWAIT | MSG_NOSIGNAL | MSG_FASTOPEN; =20 - lock_sock(sk); - mptcp_rps_record_subflows(msk); =20 if (unlikely(inet_test_bit(DEFER_CONNECT, sk) || @@ -2004,7 +2003,6 @@ static int mptcp_sendmsg(struct sock *sk, struct msgh= dr *msg, size_t len) __mptcp_push_pending(sk, msg->msg_flags); =20 out: - release_sock(sk); return copied; =20 do_error: @@ -2015,6 +2013,17 @@ static int mptcp_sendmsg(struct sock *sk, struct msg= hdr *msg, size_t len) goto out; } =20 +static int mptcp_sendmsg(struct sock *sk, struct msghdr *msg, size_t len) +{ + int ret; + + lock_sock(sk); + ret =3D mptcp_sendmsg_locked(sk, msg, len); + release_sock(sk); + + return ret; +} + static void mptcp_rcv_space_adjust(struct mptcp_sock *msk, int copied); =20 static void mptcp_eat_recv_skb(struct sock *sk, struct sk_buff *skb) @@ -2242,7 +2251,7 @@ static bool mptcp_move_skbs(struct sock *sk) return enqueued; } =20 -static unsigned int mptcp_inq_hint(const struct sock *sk) +static int mptcp_inq(struct sock *sk) { const struct mptcp_sock *msk =3D mptcp_sk(sk); const struct sk_buff *skb; @@ -2257,6 +2266,16 @@ static unsigned int mptcp_inq_hint(const struct sock= *sk) return (unsigned int)hint_val; } =20 + return 0; +} + +static unsigned int mptcp_inq_hint(struct sock *sk) +{ + unsigned int inq =3D mptcp_inq(sk); + + if (inq) + return inq; + if (sk->sk_state =3D=3D TCP_CLOSE || (sk->sk_shutdown & RCV_SHUTDOWN)) return 1; =20 @@ -4678,3 +4697,88 @@ int __init mptcp_proto_v6_init(void) return err; } #endif + +static struct sk_buff *mptcp_recv_skb_tls(struct sock *sk, u32 seq, u32 *o= ff) +{ + return mptcp_recv_skb(sk, off); +} + +static void mptcp_read_done(struct sock *sk, size_t len) +{ + struct mptcp_sock *msk =3D mptcp_sk(sk); + struct sk_buff *skb; + size_t left; + u32 offset; + + msk_owned_by_me(msk); + + if (sk->sk_state =3D=3D TCP_LISTEN) + return; + + left =3D len; + while (left && (skb =3D mptcp_recv_skb(sk, &offset)) !=3D NULL) { + int used; + + used =3D min_t(size_t, skb->len - offset, left); + msk->bytes_consumed +=3D used; + MPTCP_SKB_CB(skb)->offset +=3D used; + MPTCP_SKB_CB(skb)->map_seq +=3D used; + left -=3D used; + + if (skb->len > offset + used) + break; + + mptcp_eat_recv_skb(sk, skb); + } + + mptcp_rcv_space_adjust(msk, len - left); + + /* Clean up data we have read: This will do ACK frames. */ + if (left !=3D len) + mptcp_cleanup_rbuf(msk, len - left); +} + +static u32 mptcp_get_skb_seq(struct sk_buff *skb) +{ + return MPTCP_SKB_CB(skb)->map_seq; +} + +static int mptcp_read_sock_tls(struct sock *sk, read_descriptor_t *desc, + sk_read_actor_t recv_actor) +{ + return __mptcp_read_sock(sk, desc, recv_actor, in_softirq()); +} + +static bool mptcp_epollin_ready_tls(const struct sock *sk, int target) +{ + return mptcp_epollin_ready(sk); +} + +static void mptcp_check_app_limited(struct sock *sk) +{ + struct mptcp_sock *msk =3D mptcp_sk(sk); + struct mptcp_subflow_context *subflow; + + mptcp_for_each_subflow(msk, subflow) { + struct sock *ssk =3D mptcp_subflow_tcp_sock(subflow); + bool slow; + + slow =3D lock_sock_fast(ssk); + tcp_sock_rate_check_app_limited(tcp_sk(ssk)); + unlock_sock_fast(ssk, slow); + } +} + +struct tls_prot_ops tls_mptcp_ops =3D { + .protocol =3D IPPROTO_MPTCP, + .inq =3D mptcp_inq, + .sendmsg_locked =3D mptcp_sendmsg_locked, + .recv_skb =3D mptcp_recv_skb_tls, + .read_done =3D mptcp_read_done, + .get_skb_seq =3D mptcp_get_skb_seq, + .read_sock =3D mptcp_read_sock_tls, + .poll =3D mptcp_poll, + .epollin_ready =3D mptcp_epollin_ready_tls, + .check_app_limited =3D mptcp_check_app_limited, +}; +EXPORT_SYMBOL(tls_mptcp_ops); diff --git a/net/tls/tls_main.c b/net/tls/tls_main.c index e8bb745029e4..75e01f2dde26 100644 --- a/net/tls/tls_main.c +++ b/net/tls/tls_main.c @@ -772,6 +772,9 @@ static int do_tls_setsockopt_conf(struct sock *sk, sock= ptr_t optval, tls_sw_strparser_arm(sk, ctx); } =20 + if (conf =3D=3D TLS_HW && sk->sk_protocol =3D=3D IPPROTO_MPTCP) + return -EOPNOTSUPP; + if (tx) ctx->tx_conf =3D conf; else @@ -1332,6 +1335,9 @@ static int __init tls_register(void) tcp_register_ulp(&tcp_tls_ulp_ops); =20 tls_register_prot_ops(&tls_tcp_ops); +#ifdef CONFIG_MPTCP + tls_register_prot_ops(&tls_mptcp_ops); +#endif =20 return 0; err_strp: --=20 2.51.0