From nobody Sat Sep 26 12:28:38 2026 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-alma10-1.taild15c8.ts.net [100.103.45.18]) (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 BB29D35C690 for ; Sun, 23 Aug 2026 01:02:17 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=100.103.45.18 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1787446947; cv=none; b=Cl3SNlzx6U120rZ5xD9FCjj5aTyh98yo5uW0Uj5Y21A6Zaw5LxrTEbQVVr+COOaRjVzEe85cPn7vUlVRAnya6DSOFP+a5tqhukdU79pdFNH4GDRCKWvlZUPqHsmGeNHF3CmbdJEmAA1Acse0VtWWLoBc+1gYDW/MihohvFLDkug= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1787446947; c=relaxed/simple; bh=fwNhWA7/HhhWlyx6Eqpb2J7lCcvZ1hmAHxC2ESYEDK4=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=aUvkbOCndbj5fZl5/Sp61jzVj6iPOMzcW+OhVq5oBGjNYI4oJKm/PXJVcBPaYGhNu+6QZmLbLSO/uclt+N4zrp1EfvCXpiPRkTDaJvZP6scW5Llzk+OvTvB3lwtikaTl2MTffoq97alq9FLOvIDR3FjpMVMnttR36rINBiIt4xk= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=DCPUgU52; arc=none smtp.client-ip=100.103.45.18 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="DCPUgU52" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 0B3971F00A3A; Sun, 23 Aug 2026 01:02:10 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1787446932; bh=EbtbHCsIuO1lhJC8yuDXVRCqhrT67dAhhQnq/7Fqubg=; h=From:To:Cc:Subject:Date:In-Reply-To:References; b=DCPUgU52uhGvtB8Dom5d2acw0EHBSNc4cKLV0dYJ6XAGgABcIpI2tsc9HM7tT2Hrw zza02M5PdDZLPyCxc/ZfatWKuODMbJjoqdbhSSlgGS1pQYze28OquhPjIJH6MsXKz3 vL3HnnyugFKMRbIp2EiMq65tY0ELTK6VzxRwCJYlZDz6lIaTr2gMOwfNCJMiLVQiE4 a2zqaVpC8w68C7XAL4+Uj9ayaHu2yr8CH/CHXy3bsulWxgJc2J6qMzSzzANXKUZnLR S1GW8bNkHCG6OKjIjrPLd067YlVkt4cDPRKbjZQVpvvzZwb1RwfYSRNFwlR6X0p/T9 nyIYFpA3GvKiQ== From: Geliang Tang To: mptcp@lists.linux.dev Cc: Paolo Abeni , Geliang Tang Subject: [PATCH mptcp-next v6 1/7] mptcp: drop the mptcp_ooo_try_coalesce() helper Date: Sun, 23 Aug 2026 09:01:31 +0800 Message-ID: <32ab30e519aca2c4bf4b18777462b71b376899e3.1787446274.git.tanggeliang@kylinos.cn> X-Mailer: git-send-email 2.53.0 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: Paolo Abeni It's used to save an additional comparison for in-order skbs, but is also a barrier to remove CB offset. Remove the helper, let __mptcp_try_coalesce() always perform the sequence check and remove duplicate checks from the callers. Co-developed-by: Geliang Tang Signed-off-by: Geliang Tang Signed-off-by: Paolo Abeni --- net/mptcp/protocol.c | 21 ++++++--------------- 1 file changed, 6 insertions(+), 15 deletions(-) diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index f22d64ab1c53..7421b44934f0 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -167,7 +167,8 @@ static bool __mptcp_try_coalesce(struct sock *sk, struc= t sk_buff *to, { int limit =3D READ_ONCE(sk->sk_rcvbuf); =20 - if (unlikely(MPTCP_SKB_CB(to)->cant_coalesce) || + if (MPTCP_SKB_CB(from)->map_seq !=3D MPTCP_SKB_CB(to)->end_seq || + unlikely(MPTCP_SKB_CB(to)->cant_coalesce) || MPTCP_SKB_CB(from)->offset || ((to->len + from->len) > (limit >> 3)) || !skb_try_coalesce(to, from, fragstolen, delta)) @@ -200,15 +201,6 @@ static bool mptcp_try_coalesce(struct sock *sk, struct= sk_buff *to, return true; } =20 -static bool mptcp_ooo_try_coalesce(struct mptcp_sock *msk, struct sk_buff = *to, - struct sk_buff *from) -{ - if (MPTCP_SKB_CB(from)->map_seq !=3D MPTCP_SKB_CB(to)->end_seq) - return false; - - return mptcp_try_coalesce((struct sock *)msk, to, from); -} - /* "inspired" by tcp_rcvbuf_grow(), main difference: * - mptcp does not maintain a msk-level window clamp * - returns true when the receive buffer is actually updated @@ -348,7 +340,7 @@ static void mptcp_data_queue_ofo(struct mptcp_sock *msk= , struct sk_buff *skb) /* with 2 subflows, adding at end of ooo queue is quite likely * Use of ooo_last_skb avoids the O(Log(N)) rbtree lookup. */ - if (mptcp_ooo_try_coalesce(msk, msk->ooo_last_skb, skb)) { + if (mptcp_try_coalesce(sk, msk->ooo_last_skb, skb)) { MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_OFOMERGE); MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_OFOQUEUETAIL); return; @@ -394,7 +386,7 @@ static void mptcp_data_queue_ofo(struct mptcp_sock *msk= , struct sk_buff *skb) MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_DUPDATA); goto merge_right; } - } else if (mptcp_ooo_try_coalesce(msk, skb1, skb)) { + } else if (mptcp_try_coalesce(sk, skb1, skb)) { MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_OFOMERGE); return; } @@ -777,8 +769,7 @@ static void __mptcp_add_backlog(struct sock *sk, if (!list_empty(&msk->backlog_list)) tail =3D list_last_entry(&msk->backlog_list, struct sk_buff, list); =20 - if (tail && MPTCP_SKB_CB(skb)->map_seq =3D=3D MPTCP_SKB_CB(tail)->end_seq= && - ssk =3D=3D tail->sk && + if (tail && ssk =3D=3D tail->sk && __mptcp_try_coalesce(sk, tail, skb, &fragstolen, &delta)) { skb->truesize -=3D delta; kfree_skb_partial(skb, fragstolen); @@ -902,7 +893,7 @@ static bool __mptcp_ofo_queue(struct mptcp_sock *msk) =20 end_seq =3D MPTCP_SKB_CB(skb)->end_seq; tail =3D skb_peek_tail(&sk->sk_receive_queue); - if (!tail || !mptcp_ooo_try_coalesce(msk, tail, skb)) { + if (!tail || !mptcp_try_coalesce(sk, tail, skb)) { int delta =3D msk->ack_seq - MPTCP_SKB_CB(skb)->map_seq; =20 /* skip overlapping data, if any */ --=20 2.53.0 From nobody Sat Sep 26 12:28:38 2026 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-alma10-1.taild15c8.ts.net [100.103.45.18]) (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 B806E3515FA for ; Sun, 23 Aug 2026 01:02:17 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=100.103.45.18 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1787446945; cv=none; b=lM7okksMYTZ4E0IIbFPPTHg54hZQrps8FW7OE54luizyGonMGOyCX3/EgW1nZsGahYmKIbG3G15v4N+b5aPpPrYXqnq7eZ+USINrn5GaFUgiMT7E+1apjqHlC9HsLaIbw9/25u9JykaZt6hs0Zu5+HmGBDXET814HiKNAQuO1fc= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1787446945; c=relaxed/simple; bh=htG3MmuUPTONUv+1571RfQX29p2Ug1TjS2fBKhYSx08=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=IXMpqN0KYLd4tViTC74G2gncSUeurA+GkIEeE2xYmkrmW8enwY6WJSafQM9xAS16NTAWVX2x0WcsNAvaK9f6AVJqpOXaxo9oG9l3A3iOpf71Uy+f7U74qMlYFyLGDS7ezyrkhEYZhHOLEBgJEYvFbmcEs4Wlw5GMDZ6/+QAoZgo= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=UiqyLPb/; arc=none smtp.client-ip=100.103.45.18 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="UiqyLPb/" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 8ECCA1F00A3D; Sun, 23 Aug 2026 01:02:12 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1787446933; bh=6AW5Nf03rOKATQ+qztjjl2akUJu+gS3WASFy4PAsDZE=; h=From:To:Cc:Subject:Date:In-Reply-To:References; b=UiqyLPb/zB2V8lU6VQqYeRViSXC/LOXDaW8DVDeaQQtgnT9u6bqx8uCUgmSdh4NZ9 PU5B3vqScixhF9bxfdZpEHJG5P0uRVm1TgsKj075VPffYEho1M+FuiXU8TPsErvV11 E52at8YV0LnzpeK9wW6aIt4Ew53EQAuxevs4wBFxnbx2lxPFdwsN+8S0TLeLaghg9q pT/WbkFEZ25UTpcPhKJwd7kAgGV0O/OvY8t4/1f3eIcLENIPBOv06Lk6+ywMCuIZ4y Nbqadt3dGODxoohY0aXoWcrQn1SRRNj8D5EhqzlA4Xwe4Qj94u7OaaWvBiyMFBX8Fe GiYk8KdmqE01g== From: Geliang Tang To: mptcp@lists.linux.dev Cc: Paolo Abeni , Geliang Tang Subject: [PATCH mptcp-next v6 2/7] mptcp: drop the cant_coalesce CB field Date: Sun, 23 Aug 2026 09:01:32 +0800 Message-ID: X-Mailer: git-send-email 2.53.0 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: Paolo Abeni Such field is used to ensure in-sequence processing in case of fastopen. Instead let's perform synchronization of the fastopen skb sequence when the IASN becomes available with the 3rd ack. When the `cant_coalesce` field has been introduced, commit f03afb3aeb9d ("mptcp: drop __mptcp_fastopen_gen_msk_ackseq()") noted that updating the already queued skb for passive fastopen socket at 3rd ack time would be difficult and race prone. The main point is that such update don't need to be synchronously performed at 3rd ack time, but is sufficient to perform it before the next segment is introduced into the msk. To such extent, add an explicit test in __mptcp_move_skb(). Performance wise this trades a conditional in the fast path - in __mptcp_try_coalesce() - with a similar one in __mptcp_move_skb() and a couple more in slow paths. After this change the user-space will always observe consistent sequence numbers in the receive queue, even in the TFO dummy mapping case. There is still a potential race in mptcp_inq_hint() that will be addressed by a later patch in the series. Co-developed-by: Geliang Tang Signed-off-by: Geliang Tang Signed-off-by: Paolo Abeni --- net/mptcp/fastopen.c | 2 +- net/mptcp/protocol.c | 35 +++++++++++++++++++++++++++++++++-- net/mptcp/protocol.h | 4 +++- net/mptcp/subflow.c | 10 ++++++++++ 4 files changed, 47 insertions(+), 4 deletions(-) diff --git a/net/mptcp/fastopen.c b/net/mptcp/fastopen.c index f717750906ff..d6895c2200cc 100644 --- a/net/mptcp/fastopen.c +++ b/net/mptcp/fastopen.c @@ -49,11 +49,11 @@ void mptcp_fastopen_subflow_synack_set_params(struct mp= tcp_subflow_context *subf MPTCP_SKB_CB(skb)->end_seq =3D 0; MPTCP_SKB_CB(skb)->offset =3D 0; MPTCP_SKB_CB(skb)->has_rxtstamp =3D has_rxtstamp; - MPTCP_SKB_CB(skb)->cant_coalesce =3D 1; =20 mptcp_data_lock(sk); DEBUG_NET_WARN_ON_ONCE(sock_owned_by_user_nocheck(sk)); =20 + mptcp_sk(sk)->rcvd_dummy_seq =3D true; mptcp_borrow_fwdmem(sk, skb); skb_set_owner_r(skb, sk); __skb_queue_tail(&sk->sk_receive_queue, skb); diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index 7421b44934f0..0a90bed02b41 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -168,7 +168,6 @@ static bool __mptcp_try_coalesce(struct sock *sk, struc= t sk_buff *to, int limit =3D READ_ONCE(sk->sk_rcvbuf); =20 if (MPTCP_SKB_CB(from)->map_seq !=3D MPTCP_SKB_CB(to)->end_seq || - unlikely(MPTCP_SKB_CB(to)->cant_coalesce) || MPTCP_SKB_CB(from)->offset || ((to->len + from->len) > (limit >> 3)) || !skb_try_coalesce(to, from, fragstolen, delta)) @@ -430,7 +429,6 @@ static void mptcp_init_skb(struct sock *ssk, struct sk_= buff *skb, int offset, MPTCP_SKB_CB(skb)->end_seq =3D MPTCP_SKB_CB(skb)->map_seq + copy_len; MPTCP_SKB_CB(skb)->offset =3D offset; MPTCP_SKB_CB(skb)->has_rxtstamp =3D has_rxtstamp; - MPTCP_SKB_CB(skb)->cant_coalesce =3D 0; =20 __skb_unlink(skb, &ssk->sk_receive_queue); =20 @@ -438,6 +436,26 @@ static void mptcp_init_skb(struct sock *ssk, struct sk= _buff *skb, int offset, skb_dst_drop(skb); } =20 +void __mptcp_sync_rcv_sequence(struct sock *sk) +{ + struct mptcp_sock *msk =3D mptcp_sk(sk); + struct sk_buff *skb; + u32 offset; + + if (likely(!msk->rcvd_dummy_seq)) + return; + + /* User space can have already received the TFO skb. */ + msk->rcvd_dummy_seq =3D false; + skb =3D skb_peek_tail(&sk->sk_receive_queue); + if (!skb) + return; + + offset =3D MPTCP_SKB_CB(skb)->offset; + MPTCP_SKB_CB(skb)->map_seq =3D msk->ack_seq - skb->len + offset; + MPTCP_SKB_CB(skb)->end_seq =3D msk->ack_seq; +} + static bool __mptcp_move_skb(struct sock *sk, struct sk_buff *skb) { u64 copy_len =3D MPTCP_SKB_CB(skb)->end_seq - MPTCP_SKB_CB(skb)->map_seq; @@ -446,6 +464,12 @@ static bool __mptcp_move_skb(struct sock *sk, struct s= k_buff *skb) =20 mptcp_borrow_fwdmem(sk, skb); =20 + /* Be sure to sync the eventual fastopen dummy mapping before any other + * skb lands into the msk. + */ + if (unlikely(msk->rcvd_dummy_seq)) + __mptcp_sync_rcv_sequence(sk); + if (MPTCP_SKB_CB(skb)->map_seq =3D=3D msk->ack_seq) { /* in sequence */ insert: @@ -3895,6 +3919,13 @@ static void mptcp_release_cb(struct sock *sk) __mptcp_error_report(sk); if (__test_and_clear_bit(MPTCP_SYNC_SNDBUF, &msk->cb_flags)) __mptcp_sync_sndbuf(sk); + if (test_and_clear_bit(MPTCP_SYNC_SEQ, &msk->cb_flags)) { + /* Ensure we see the updated ack_seq after seeing + * the flag + */ + smp_rmb(); + __mptcp_sync_rcv_sequence(sk); + } } } =20 diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h index 3d250e8204d5..0b20cacf56a8 100644 --- a/net/mptcp/protocol.h +++ b/net/mptcp/protocol.h @@ -126,13 +126,13 @@ #define MPTCP_FLUSH_JOIN_LIST 5 #define MPTCP_SYNC_STATE 6 #define MPTCP_SYNC_SNDBUF 7 +#define MPTCP_SYNC_SEQ 8 =20 struct mptcp_skb_cb { u64 map_seq; u64 end_seq; u32 offset; u8 has_rxtstamp; - u8 cant_coalesce; }; =20 #define MPTCP_SKB_CB(__skb) ((struct mptcp_skb_cb *)&((__skb)->cb[0])) @@ -313,6 +313,7 @@ struct mptcp_sock { u32 token; unsigned long flags; unsigned long cb_flags; + bool rcvd_dummy_seq; bool recovery; /* closing subflow write queue reinjected */ bool can_ack; bool fully_established; @@ -1172,6 +1173,7 @@ void mptcp_event_pm_listener(const struct sock *ssk, enum mptcp_event_type event); bool mptcp_userspace_pm_active(const struct mptcp_sock *msk); =20 +void __mptcp_sync_rcv_sequence(struct sock *sk); void mptcp_fastopen_subflow_synack_set_params(struct mptcp_subflow_context= *subflow, struct request_sock *req); int mptcp_pm_genl_fill_addr(struct sk_buff *msg, diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c index 2d7ccb01d234..ed3a48cf9c97 100644 --- a/net/mptcp/subflow.c +++ b/net/mptcp/subflow.c @@ -476,6 +476,8 @@ static void subflow_set_remote_key(struct mptcp_sock *m= sk, struct mptcp_subflow_context *subflow, const struct mptcp_options_received *mp_opt) { + struct sock *sk =3D (struct sock *)msk; + /* active MPC subflow will reach here multiple times: * at subflow_finish_connect() time and at 4th ack time */ @@ -494,6 +496,14 @@ static void subflow_set_remote_key(struct mptcp_sock *= msk, WRITE_ONCE(msk->ack_seq, subflow->iasn); WRITE_ONCE(msk->can_ack, true); atomic64_set(&msk->rcv_wnd_sent, subflow->iasn); + + if (!sock_owned_by_user(sk)) { + __mptcp_sync_rcv_sequence(sk); + } else { + /* Ensure ack_seq is visible before setting the flag */ + smp_wmb(); + set_bit(MPTCP_SYNC_SEQ, &msk->cb_flags); + } } =20 static void mptcp_propagate_state(struct sock *sk, struct sock *ssk, --=20 2.53.0 From nobody Sat Sep 26 12:28:38 2026 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-alma10-1.taild15c8.ts.net [100.103.45.18]) (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 CBF6333F8C4 for ; Sun, 23 Aug 2026 01:02:18 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=100.103.45.18 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1787446948; cv=none; b=is+GhIeZ5Gqs0Aus3WyNok4G7gGvUT1wAEUMkVzk98iy/ZCXggIfmcIcySctca+h7tAWMVbacWgF7Z6s2Q165sIC6El4b7o0XqaVRafmOXCgdVCL0lrtLOhfAQ32CYW8WRKlkzxEBJ3rXbxvxHH/xt5ALr92FpVGCFZdvkrmyyM= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1787446948; c=relaxed/simple; bh=ub0mrAcRrSey2Y28qqjMuAYnNcZizkKPZaVhyZEQfd8=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=PDiPFGwH4cLkML5iXt+5w2eZUcFN3xeiVnRwYkH/7rj2qnbZYuw+WllGkO6CpLOFpoix/QkcEIrdT9OIJFxabafBfBe4Uzl4oI6+0VoWKGLc+3a6ZHPdMnaHDXce23ggPxYMfOxk+8y95AppgZpM/n/iP+AIyaDBHlIqbIRYYrs= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=MQ0hK/B9; arc=none smtp.client-ip=100.103.45.18 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="MQ0hK/B9" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 1C4641F00A3E; Sun, 23 Aug 2026 01:02:13 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1787446935; bh=WQgNNHCVcGANzmc20X7ZMvgIikPUxxnLbmcqJU84bQ0=; h=From:To:Cc:Subject:Date:In-Reply-To:References; b=MQ0hK/B96IVAe7enPBjdouKyCVg404qsjxQzqgPemoyTEihQk9CiwZKNlxIVdyE5B FeALWCTVCVSCf3SjmnW+zxZ9R8PTvUo9Cc+R4SKlXngPhTHvmtQepu64+SvJ3DOsuv +GCe8dG2KvnjEI9et5oc2lSHIMzsDzIZMJI1HtHwP0Yfm1wNSAxatRhKAmwpLjArBX W1M1qHLClfjsy/kt/I9VGz/KQIfmZ2ft6xgFi3YJNeEnb8eMuCPbx8sT+iGHNNUPwy JjWn+AfdrPJhEPLNcll1grmaMGBJUjplRe2EdsZ5PUh18EqrwvczAEibdk8Ta8Ilhm 0J0Vy6Qxw9YgA== From: Geliang Tang To: mptcp@lists.linux.dev Cc: Paolo Abeni , Geliang Tang Subject: [PATCH mptcp-next v6 3/7] mptcp: remove CB offset field Date: Sun, 23 Aug 2026 09:01:33 +0800 Message-ID: <95f5d91022661e9783da7a08c369659b3a14f802.1787446274.git.tanggeliang@kylinos.cn> X-Mailer: git-send-email 2.53.0 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: Paolo Abeni Instead, use a new msk-level field to track the bytes already consumed inside each skb, carrying the amount of bytes already copied to user-space, alike what TCP is already doing. The newly introduce `copied_seq` field is always accessed under the msk socket lock, delegating the synchronization with IASN to the msk release CB, when the socket is owned by the user-space at remote key reception time. Such synchronization preserves any partial progress (copy) made on the TFO packet. Note that the explicit synchronization in __mptcp_move_skb() is needed to ensure that the TFO skb in the receive queue got its map_seq synched before the next skb lands into the receive queue when spooling the backlog at mptcp_release_cb() time, as the release CB synchronization will happen later. Prior to this patch, the TFO skb dummy mapping was always ignored, now it affects the `copied_seq` initial update: be sure to extends the sign correctly of such mapping initialization time. Overall this simplify a bit the __mptcp_recvmsg_mskq(), mptcp_inq_hint() and the __mptcp_move_skb() code and will also make possible the next patch. Initialize MPTCP sequence space to 0 in mptcp_propagate_state() when mp_opt is NULL, ensuring SKB's map_seq starts from 0 to match msk->copied_seq and prevent offset calculation underflow in fallback mode. Co-developed-by: Geliang Tang Signed-off-by: Geliang Tang Signed-off-by: Paolo Abeni --- net/mptcp/fastopen.c | 15 ++++-- net/mptcp/protocol.c | 111 +++++++++++++++++++++---------------------- net/mptcp/protocol.h | 8 +++- net/mptcp/subflow.c | 9 ++++ 4 files changed, 80 insertions(+), 63 deletions(-) diff --git a/net/mptcp/fastopen.c b/net/mptcp/fastopen.c index d6895c2200cc..421a50a85547 100644 --- a/net/mptcp/fastopen.c +++ b/net/mptcp/fastopen.c @@ -9,6 +9,7 @@ void mptcp_fastopen_subflow_synack_set_params(struct mptcp_subflow_context= *subflow, struct request_sock *req) { + struct mptcp_sock *msk; struct sock *sk, *ssk; struct sk_buff *skb; struct tcp_sock *tp; @@ -44,20 +45,24 @@ void mptcp_fastopen_subflow_synack_set_params(struct mp= tcp_subflow_context *subf subflow->ssn_offset +=3D skb->len; has_rxtstamp =3D TCP_SKB_CB(skb)->has_rxtstamp; =20 - /* Only the sequence delta is relevant */ - MPTCP_SKB_CB(skb)->map_seq =3D -skb->len; + /* The TFO segment data sits before the IASN; before receiving + * the remote key, IASN is assumed being 0. + */ + MPTCP_SKB_CB(skb)->map_seq =3D -(u64)skb->len; MPTCP_SKB_CB(skb)->end_seq =3D 0; - MPTCP_SKB_CB(skb)->offset =3D 0; MPTCP_SKB_CB(skb)->has_rxtstamp =3D has_rxtstamp; =20 mptcp_data_lock(sk); DEBUG_NET_WARN_ON_ONCE(sock_owned_by_user_nocheck(sk)); =20 - mptcp_sk(sk)->rcvd_dummy_seq =3D true; + msk =3D mptcp_sk(sk); + msk->rcvd_dummy_seq =3D true; + msk->copied_seq =3D MPTCP_SKB_CB(skb)->map_seq; + msk->tfo_skb_len =3D skb->len; mptcp_borrow_fwdmem(sk, skb); skb_set_owner_r(skb, sk); __skb_queue_tail(&sk->sk_receive_queue, skb); - mptcp_sk(sk)->bytes_received +=3D skb->len; + msk->bytes_received +=3D skb->len; =20 sk->sk_data_ready(sk); =20 diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index 0a90bed02b41..0700cdab7f2e 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -29,7 +29,7 @@ #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 @@ -168,7 +168,6 @@ static bool __mptcp_try_coalesce(struct sock *sk, struc= t sk_buff *to, int limit =3D READ_ONCE(sk->sk_rcvbuf); =20 if (MPTCP_SKB_CB(from)->map_seq !=3D MPTCP_SKB_CB(to)->end_seq || - MPTCP_SKB_CB(from)->offset || ((to->len + from->len) > (limit >> 3)) || !skb_try_coalesce(to, from, fragstolen, delta)) return false; @@ -415,8 +414,7 @@ static void mptcp_data_queue_ofo(struct mptcp_sock *msk= , struct sk_buff *skb) skb_set_owner_r(skb, sk); } =20 -static void mptcp_init_skb(struct sock *ssk, struct sk_buff *skb, int offs= et, - int copy_len) +static void mptcp_init_skb(struct sock *ssk, struct sk_buff *skb, int offs= et) { struct mptcp_subflow_context *subflow =3D mptcp_subflow_ctx(ssk); bool has_rxtstamp =3D TCP_SKB_CB(skb)->has_rxtstamp; @@ -425,9 +423,9 @@ static void mptcp_init_skb(struct sock *ssk, struct sk_= buff *skb, int offset, * mptcp_subflow_get_mapped_dsn() is based on the current tp->copied_seq * value */ - MPTCP_SKB_CB(skb)->map_seq =3D mptcp_subflow_get_mapped_dsn(subflow); - MPTCP_SKB_CB(skb)->end_seq =3D MPTCP_SKB_CB(skb)->map_seq + copy_len; - MPTCP_SKB_CB(skb)->offset =3D offset; + MPTCP_SKB_CB(skb)->map_seq =3D mptcp_subflow_get_mapped_dsn(subflow) - + offset; + MPTCP_SKB_CB(skb)->end_seq =3D MPTCP_SKB_CB(skb)->map_seq + skb->len; MPTCP_SKB_CB(skb)->has_rxtstamp =3D has_rxtstamp; =20 __skb_unlink(skb, &ssk->sk_receive_queue); @@ -440,7 +438,6 @@ void __mptcp_sync_rcv_sequence(struct sock *sk) { struct mptcp_sock *msk =3D mptcp_sk(sk); struct sk_buff *skb; - u32 offset; =20 if (likely(!msk->rcvd_dummy_seq)) return; @@ -451,9 +448,8 @@ void __mptcp_sync_rcv_sequence(struct sock *sk) if (!skb) return; =20 - offset =3D MPTCP_SKB_CB(skb)->offset; - MPTCP_SKB_CB(skb)->map_seq =3D msk->ack_seq - skb->len + offset; - MPTCP_SKB_CB(skb)->end_seq =3D msk->ack_seq; + MPTCP_SKB_CB(skb)->map_seq =3D mptcp_iasn(msk) - skb->len; + MPTCP_SKB_CB(skb)->end_seq =3D MPTCP_SKB_CB(skb)->map_seq + skb->len; } =20 static bool __mptcp_move_skb(struct sock *sk, struct sk_buff *skb) @@ -464,6 +460,12 @@ static bool __mptcp_move_skb(struct sock *sk, struct s= k_buff *skb) =20 mptcp_borrow_fwdmem(sk, skb); =20 + if (test_and_clear_bit(MPTCP_SYNC_SEQ, &msk->cb_flags)) { + /* Ensure we see the updated ack_seq after seeing the flag */ + smp_rmb(); + msk->copied_seq +=3D mptcp_iasn(msk); + } + /* Be sure to sync the eventual fastopen dummy mapping before any other * skb lands into the msk. */ @@ -496,10 +498,6 @@ static bool __mptcp_move_skb(struct sock *sk, struct s= k_buff *skb) /* Partial packet */ if (after64(MPTCP_SKB_CB(skb)->end_seq, msk->ack_seq)) { copy_len =3D MPTCP_SKB_CB(skb)->end_seq - msk->ack_seq; - MPTCP_SKB_CB(skb)->offset +=3D msk->ack_seq - - MPTCP_SKB_CB(skb)->map_seq; - MPTCP_SKB_CB(skb)->map_seq +=3D msk->ack_seq - - MPTCP_SKB_CB(skb)->map_seq; goto insert; } =20 @@ -857,7 +855,7 @@ static bool __mptcp_move_skbs_from_subflow(struct mptcp= _sock *msk, if (offset < skb->len) { size_t len =3D skb->len - offset; =20 - mptcp_init_skb(ssk, skb, offset, len); + mptcp_init_skb(ssk, skb, offset); =20 if (own_msk) { mptcp_subflow_lend_fwdmem(subflow, skb); @@ -924,8 +922,6 @@ static bool __mptcp_ofo_queue(struct mptcp_sock *msk) pr_debug("uncoalesced seq=3D%llx ack seq=3D%llx delta=3D%d\n", MPTCP_SKB_CB(skb)->map_seq, msk->ack_seq, delta); - MPTCP_SKB_CB(skb)->offset +=3D delta; - MPTCP_SKB_CB(skb)->map_seq +=3D delta; __skb_queue_tail(&sk->sk_receive_queue, skb); } msk->bytes_received +=3D end_seq - msk->ack_seq; @@ -2199,33 +2195,24 @@ static void mptcp_eat_recv_skb(struct sock *sk, str= uct sk_buff *skb) } =20 static int __mptcp_recvmsg_mskq(struct sock *sk, struct msghdr *msg, - size_t len, int flags, int copied_total, + size_t len, int flags, u64 *seq, struct scm_timestamping_internal *tss, int *cmsg_flags, struct sk_buff **last) { struct mptcp_sock *msk =3D mptcp_sk(sk); struct sk_buff *skb, *tmp; - int total_data_len =3D 0; int copied =3D 0; =20 skb_queue_walk_safe(&sk->sk_receive_queue, skb, tmp) { - u32 delta, offset =3D MPTCP_SKB_CB(skb)->offset; + u64 offset =3D *seq - MPTCP_SKB_CB(skb)->map_seq; u32 data_len =3D skb->len - offset; u32 count; int err; =20 - if (flags & MSG_PEEK) { - /* skip already peeked skbs */ - if (total_data_len + data_len <=3D copied_total) { - total_data_len +=3D data_len; - *last =3D skb; - continue; - } - - /* skip the already peeked data in the current skb */ - delta =3D copied_total - total_data_len; - offset +=3D delta; - data_len -=3D delta; + /* Skip the already peeked data. */ + if (offset >=3D skb->len) { + *last =3D skb; + continue; } =20 count =3D min_t(size_t, len - copied, data_len); @@ -2244,14 +2231,12 @@ static int __mptcp_recvmsg_mskq(struct sock *sk, st= ruct msghdr *msg, } =20 copied +=3D count; + *seq +=3D count; =20 if (!(flags & MSG_PEEK)) { msk->bytes_consumed +=3D count; - if (count < data_len) { - MPTCP_SKB_CB(skb)->offset +=3D count; - MPTCP_SKB_CB(skb)->map_seq +=3D count; + if (count < data_len) break; - } =20 mptcp_eat_recv_skb(sk, skb); } else { @@ -2404,25 +2389,23 @@ static bool mptcp_move_skbs(struct sock *sk) return enqueued; } =20 -static unsigned int mptcp_inq_hint(const struct sock *sk) +static unsigned int mptcp_inq_hint(struct sock *sk) { const struct mptcp_sock *msk =3D mptcp_sk(sk); - const struct sk_buff *skb; - - skb =3D skb_peek(&sk->sk_receive_queue); - if (skb) { - u64 hint_val =3D READ_ONCE(msk->ack_seq) - MPTCP_SKB_CB(skb)->map_seq; + u64 hint_val; =20 - if (hint_val >=3D INT_MAX) - return INT_MAX; - - return (unsigned int)hint_val; - } + /* Avoid races vs ack_seq updates. */ + mptcp_data_lock(sk); + hint_val =3D msk->ack_seq - msk->copied_seq; + mptcp_data_unlock(sk); + if (hint_val >=3D INT_MAX) + return INT_MAX; =20 - if (sk->sk_state =3D=3D TCP_CLOSE || (sk->sk_shutdown & RCV_SHUTDOWN)) + if (!hint_val && + (sk->sk_state =3D=3D TCP_CLOSE || (sk->sk_shutdown & RCV_SHUTDOWN))) return 1; =20 - return 0; + return (unsigned int)hint_val; } =20 static int mptcp_recvmsg(struct sock *sk, struct msghdr *msg, size_t len, @@ -2431,6 +2414,7 @@ static int mptcp_recvmsg(struct sock *sk, struct msgh= dr *msg, size_t len, struct mptcp_sock *msk =3D mptcp_sk(sk); struct scm_timestamping_internal tss; int copied =3D 0, cmsg_flags =3D 0; + u64 peek_seq, *seq; int target; long timeo; =20 @@ -2449,6 +2433,11 @@ static int mptcp_recvmsg(struct sock *sk, struct msg= hdr *msg, size_t len, =20 len =3D min_t(size_t, len, INT_MAX); target =3D sock_rcvlowat(sk, flags & MSG_WAITALL, len); + seq =3D &msk->copied_seq; + if (flags & MSG_PEEK) { + peek_seq =3D msk->copied_seq; + seq =3D &peek_seq; + } =20 if (unlikely(msk->recvmsg_inq)) cmsg_flags =3D MPTCP_CMSG_INQ; @@ -2458,7 +2447,7 @@ static int mptcp_recvmsg(struct sock *sk, struct msgh= dr *msg, size_t len, int err, bytes_read; =20 bytes_read =3D __mptcp_recvmsg_mskq(sk, msg, len - copied, flags, - copied, &tss, &cmsg_flags, + seq, &tss, &cmsg_flags, &last); if (unlikely(bytes_read < 0)) { if (!copied) @@ -2468,8 +2457,11 @@ static int mptcp_recvmsg(struct sock *sk, struct msg= hdr *msg, size_t len, =20 copied +=3D bytes_read; =20 - if (!list_empty(&msk->backlog_list) && mptcp_move_skbs(sk)) + if (!list_empty(&msk->backlog_list) && mptcp_move_skbs(sk)) { + if (flags & MSG_PEEK) + peek_seq =3D msk->copied_seq + copied; continue; + } =20 /* only the MPTCP socket status is relevant here. The exit * conditions mirror closely tcp_recvmsg() @@ -2513,6 +2505,10 @@ static int mptcp_recvmsg(struct sock *sk, struct msg= hdr *msg, size_t len, err =3D copied ? : err; goto out_err; } + + /* Recompute peek offset after eventual seq resync. */ + if (flags & MSG_PEEK) + peek_seq =3D msk->copied_seq + copied; } =20 mptcp_cleanup_rbuf(msk, copied); @@ -3690,11 +3686,13 @@ static int mptcp_disconnect(struct sock *sk, int fl= ags) msk->bytes_retrans =3D 0; msk->rcvspace_init =3D 0; msk->fastclosing =3D 0; + msk->tfo_skb_len =3D 0; mptcp_init_rtt_est(msk); =20 /* for fallback's sake */ WRITE_ONCE(msk->ack_seq, 0); atomic64_set(&msk->rcv_wnd_sent, 0); + msk->copied_seq =3D 0; =20 WRITE_ONCE(sk->sk_shutdown, 0); sk_error_report(sk); @@ -3924,6 +3922,7 @@ static void mptcp_release_cb(struct sock *sk) * the flag */ smp_rmb(); + msk->copied_seq +=3D mptcp_iasn(msk); __mptcp_sync_rcv_sequence(sk); } } @@ -4588,7 +4587,7 @@ static struct sk_buff *mptcp_recv_skb(struct sock *sk= , u32 *off) mptcp_move_skbs(sk); =20 while ((skb =3D skb_peek(&sk->sk_receive_queue)) !=3D NULL) { - offset =3D MPTCP_SKB_CB(skb)->offset; + offset =3D msk->copied_seq - MPTCP_SKB_CB(skb)->map_seq; if (offset < skb->len) { *off =3D offset; return skb; @@ -4630,11 +4629,9 @@ static int __mptcp_read_sock(struct sock *sk, read_d= escriptor_t *desc, copied +=3D count; =20 msk->bytes_consumed +=3D count; - if (count < data_len) { - MPTCP_SKB_CB(skb)->offset +=3D count; - MPTCP_SKB_CB(skb)->map_seq +=3D count; + msk->copied_seq +=3D count; + if (count < data_len) break; - } =20 mptcp_eat_recv_skb(sk, skb); if (!desc->count) diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h index 0b20cacf56a8..72a357b11b0f 100644 --- a/net/mptcp/protocol.h +++ b/net/mptcp/protocol.h @@ -131,7 +131,6 @@ struct mptcp_skb_cb { u64 map_seq; u64 end_seq; - u32 offset; u8 has_rxtstamp; }; =20 @@ -292,6 +291,7 @@ struct mptcp_sock { u64 bytes_sent; u64 snd_nxt; u64 bytes_received; + u64 copied_seq; u64 ack_seq; atomic64_t rcv_wnd_sent; u64 rcv_data_fin_seq; @@ -311,6 +311,7 @@ struct mptcp_sock { u32 last_ack_recv; unsigned long timer_ival; u32 token; + u32 tfo_skb_len; unsigned long flags; unsigned long cb_flags; bool rcvd_dummy_seq; @@ -865,6 +866,11 @@ struct sock *mptcp_subflow_get_retrans(struct mptcp_so= ck *msk); int mptcp_sched_get_send(struct mptcp_sock *msk); int mptcp_sched_get_retrans(struct mptcp_sock *msk); =20 +static inline u64 mptcp_iasn(const struct mptcp_sock *msk) +{ + return msk->ack_seq - msk->bytes_received + msk->tfo_skb_len; +} + static inline u64 mptcp_data_avail(const struct mptcp_sock *msk) { return READ_ONCE(msk->bytes_received) - READ_ONCE(msk->bytes_consumed); diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c index ed3a48cf9c97..6bffa02c6962 100644 --- a/net/mptcp/subflow.c +++ b/net/mptcp/subflow.c @@ -498,6 +498,8 @@ static void subflow_set_remote_key(struct mptcp_sock *m= sk, atomic64_set(&msk->rcv_wnd_sent, subflow->iasn); =20 if (!sock_owned_by_user(sk)) { + /* User space could have already read partially the TFO skb */ + msk->copied_seq +=3D subflow->iasn; __mptcp_sync_rcv_sequence(sk); } else { /* Ensure ack_seq is visible before setting the flag */ @@ -520,6 +522,13 @@ static void mptcp_propagate_state(struct sock *sk, str= uct sock *ssk, WRITE_ONCE(msk->snd_una, subflow->idsn + 1); WRITE_ONCE(msk->wnd_end, subflow->idsn + 1 + tcp_sk(ssk)->snd_wnd); subflow_set_remote_key(msk, subflow, mp_opt); + } else { + /* Fallback: initialize sequence space to 0 (no remote key) */ + subflow->map_seq =3D 0; + /* ensure mptcp_subflow_get_map_offset() returns 0 */ + subflow->map_subflow_seq =3D subflow->ssn_offset + + tcp_sk(ssk)->copied_seq; + WRITE_ONCE(msk->ack_seq, 0); } =20 if (!sock_owned_by_user(sk)) { --=20 2.53.0 From nobody Sat Sep 26 12:28:38 2026 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-alma10-1.taild15c8.ts.net [100.103.45.18]) (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 D64DB35975 for ; Sun, 23 Aug 2026 01:02:19 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=100.103.45.18 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1787446949; cv=none; b=SHktvTVQ1SErCGNl16PYWU5idcBtqCrdGDLl28AQvvRF6znBa+HKmjfs6lnP63YqwjeTvqDRI/so6RfgZOo7GLx9fKZryfBFQIArJheVkqanPzBQsxZeSUNqG9rfDUFGL+Je/B9kiUK31UEYvxjzhs6duX48gB4IkTiT4DZqToc= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1787446949; c=relaxed/simple; bh=TcDuRGtqZ5sGxfR2iWHDuCDXcqb0+s1v5XCgdnsAivk=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=HcxPHC3degmywG5UnJEDPbmisT4/pUpgUKj06u+t8yCotKf99uGuMxPnDAIfDGIW9WJt96aB+x0c4eKdR7+TGTOlvfm6IKHzfoU4UT19XH/BsQQciH55iSwgysIeOm0zeYA8mlWmu28U/iQcCl/1KMt4wGzWXO7WkYLTwj3vEMQ= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=OW43o8MH; arc=none smtp.client-ip=100.103.45.18 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="OW43o8MH" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 8ABCA1F00A3F; Sun, 23 Aug 2026 01:02:15 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1787446936; bh=hCJlxJa3tYjGV393+8hxPb+hErAvfPARK7NAgUbK/+k=; h=From:To:Cc:Subject:Date:In-Reply-To:References; b=OW43o8MHKATeI6iy+DX2gCPfW2YgP/cTPR0jl4V8Jr+gfpkN/ACJ0RBN/2xW/83R7 vOkMoFJbwLmemXZ06wThHiSEmoUZkvyuEcrHNACEnR4VuGzc1sYf5bIb8TJUSgtuLE 28fOVrSA5gont35UpUu3xPbpoEIUQlovOS5u9d7LHMR6D0ezy0K3wGbFipZsZVchba YBjm4VjcXE/nu6xi5AnWnyKavptOjcuL6X33SgintzaBcmy5Ub3nm999tUBuIyjlvL QWEHosw9drWsKppgsS3JJbln5tmfGMb15Rbv3rpY0M3vW7XuRCem6sNYtS8rWkM04u sDs6dHUf2cKbA== From: Geliang Tang To: mptcp@lists.linux.dev Cc: Paolo Abeni , Geliang Tang Subject: [PATCH mptcp-next v6 4/7] mptcp: sync mptcp skb cb layout with tcp one Date: Sun, 23 Aug 2026 09:01:34 +0800 Message-ID: X-Mailer: git-send-email 2.53.0 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: Paolo Abeni The MPTCP protocol uses a significantly different CB layout WRT TCP, as it includes different information and use 64 bits for the sequence numbers. As the msk-level rcvbuf buffer size is limited by the core socket code the INT_MAX; after validating the incoming skb vs the current receive window, we can safely use 32 bits for MPTCP-level sequence number. This allow updating the MPTCP CB layout so that fields with a corresponding TCP-level data use the same area inside the CB itself. Add build time check to ensure the latter invariant. Co-developed-by: Geliang Tang Signed-off-by: Geliang Tang Signed-off-by: Paolo Abeni --- net/mptcp/fastopen.c | 6 ++-- net/mptcp/protocol.c | 85 +++++++++++++++++++++++++++----------------- net/mptcp/protocol.h | 7 ++-- 3 files changed, 61 insertions(+), 37 deletions(-) diff --git a/net/mptcp/fastopen.c b/net/mptcp/fastopen.c index 421a50a85547..0d339dd454f9 100644 --- a/net/mptcp/fastopen.c +++ b/net/mptcp/fastopen.c @@ -48,8 +48,10 @@ void mptcp_fastopen_subflow_synack_set_params(struct mpt= cp_subflow_context *subf /* The TFO segment data sits before the IASN; before receiving * the remote key, IASN is assumed being 0. */ - MPTCP_SKB_CB(skb)->map_seq =3D -(u64)skb->len; + MPTCP_SKB_CB(skb)->map_seq64 =3D -(u64)skb->len; + MPTCP_SKB_CB(skb)->map_seq =3D MPTCP_SKB_CB(skb)->map_seq64; MPTCP_SKB_CB(skb)->end_seq =3D 0; + MPTCP_SKB_CB(skb)->flags =3D 0; MPTCP_SKB_CB(skb)->has_rxtstamp =3D has_rxtstamp; =20 mptcp_data_lock(sk); @@ -57,7 +59,7 @@ void mptcp_fastopen_subflow_synack_set_params(struct mptc= p_subflow_context *subf =20 msk =3D mptcp_sk(sk); msk->rcvd_dummy_seq =3D true; - msk->copied_seq =3D MPTCP_SKB_CB(skb)->map_seq; + msk->copied_seq =3D MPTCP_SKB_CB(skb)->map_seq64; msk->tfo_skb_len =3D skb->len; mptcp_borrow_fwdmem(sk, skb); skb_set_owner_r(skb, sk); diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index 0700cdab7f2e..2e0e8e85b4d0 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -172,7 +172,7 @@ static bool __mptcp_try_coalesce(struct sock *sk, struc= t sk_buff *to, !skb_try_coalesce(to, from, fragstolen, delta)) return false; =20 - pr_debug("colesced seq %llx into %llx new len %d new end seq %llx\n", + pr_debug("colesced seq %x into %x new len %d new end seq %x\n", MPTCP_SKB_CB(from)->map_seq, MPTCP_SKB_CB(to)->map_seq, to->len, MPTCP_SKB_CB(from)->end_seq); MPTCP_SKB_CB(to)->end_seq =3D MPTCP_SKB_CB(from)->end_seq; @@ -254,8 +254,8 @@ static void mptcp_prune_ofo_queue(struct sock *sk, struct sk_buff *skb =3D rb_to_skb(node); =20 /* Stop pruning if the incoming skb would land in OoO tail. */ - if (after64(MPTCP_SKB_CB(in_skb)->map_seq, - MPTCP_SKB_CB(skb)->map_seq)) + if (after(MPTCP_SKB_CB(in_skb)->map_seq, + MPTCP_SKB_CB(skb)->map_seq)) break; =20 pruned =3D true; @@ -301,15 +301,19 @@ static void mptcp_data_queue_ofo(struct mptcp_sock *m= sk, struct sk_buff *skb) { struct sock *sk =3D (struct sock *)msk; struct rb_node **p, *parent; - u64 seq, end_seq, max_seq; + u64 end_seq, max_seq; struct sk_buff *skb1; + u32 seq; =20 seq =3D MPTCP_SKB_CB(skb)->map_seq; - end_seq =3D MPTCP_SKB_CB(skb)->end_seq; + end_seq =3D MPTCP_SKB_CB(skb)->map_seq64 + skb->len; max_seq =3D atomic64_read(&msk->rcv_wnd_sent); =20 - pr_debug("msk=3D%p seq=3D%llx limit=3D%llx empty=3D%d\n", msk, seq, max_s= eq, + pr_debug("msk=3D%p seq=3D%x limit=3D%llx empty=3D%d\n", msk, seq, max_seq, RB_EMPTY_ROOT(&msk->out_of_order_queue)); + /* Use the full sequence space to perform the admission checks, to + * protect vs possible wrap-arounds. + */ if (after64(end_seq, max_seq)) { /* out of window */ mptcp_drop(sk, skb); @@ -345,7 +349,7 @@ static void mptcp_data_queue_ofo(struct mptcp_sock *msk= , struct sk_buff *skb) } =20 /* Can avoid an rbtree lookup if we are adding skb after ooo_last_skb */ - if (!before64(seq, MPTCP_SKB_CB(msk->ooo_last_skb)->end_seq)) { + if (!before(seq, MPTCP_SKB_CB(msk->ooo_last_skb)->end_seq)) { MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_OFOQUEUETAIL); parent =3D &msk->ooo_last_skb->rbnode; p =3D &parent->rb_right; @@ -357,18 +361,18 @@ static void mptcp_data_queue_ofo(struct mptcp_sock *m= sk, struct sk_buff *skb) while (*p) { parent =3D *p; skb1 =3D rb_to_skb(parent); - if (before64(seq, MPTCP_SKB_CB(skb1)->map_seq)) { + if (before(seq, MPTCP_SKB_CB(skb1)->map_seq)) { p =3D &parent->rb_left; continue; } - if (before64(seq, MPTCP_SKB_CB(skb1)->end_seq)) { - if (!after64(end_seq, MPTCP_SKB_CB(skb1)->end_seq)) { + if (before(seq, MPTCP_SKB_CB(skb1)->end_seq)) { + if (!after(end_seq, MPTCP_SKB_CB(skb1)->end_seq)) { /* All the bits are present. Drop. */ mptcp_drop(sk, skb); MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_DUPDATA); return; } - if (after64(seq, MPTCP_SKB_CB(skb1)->map_seq)) { + if (after(seq, MPTCP_SKB_CB(skb1)->map_seq)) { /* partial overlap: * | skb | * | skb1 | @@ -399,7 +403,7 @@ static void mptcp_data_queue_ofo(struct mptcp_sock *msk= , struct sk_buff *skb) merge_right: /* Remove other segments covered by skb. */ while ((skb1 =3D skb_rb_next(skb)) !=3D NULL) { - if (before64(end_seq, MPTCP_SKB_CB(skb1)->end_seq)) + if (before((u32)end_seq, MPTCP_SKB_CB(skb1)->end_seq)) break; rb_erase(&skb1->rbnode, &msk->out_of_order_queue); mptcp_drop(sk, skb1); @@ -421,11 +425,13 @@ static void mptcp_init_skb(struct sock *ssk, struct s= k_buff *skb, int offset) =20 /* the skb map_seq accounts for the skb offset: * mptcp_subflow_get_mapped_dsn() is based on the current tp->copied_seq - * value + * value; note that end seq number is only available in 32bits format. */ - MPTCP_SKB_CB(skb)->map_seq =3D mptcp_subflow_get_mapped_dsn(subflow) - - offset; + MPTCP_SKB_CB(skb)->map_seq64 =3D mptcp_subflow_get_mapped_dsn(subflow) - + offset; + MPTCP_SKB_CB(skb)->map_seq =3D (u32)MPTCP_SKB_CB(skb)->map_seq64; MPTCP_SKB_CB(skb)->end_seq =3D MPTCP_SKB_CB(skb)->map_seq + skb->len; + MPTCP_SKB_CB(skb)->flags =3D 0; MPTCP_SKB_CB(skb)->has_rxtstamp =3D has_rxtstamp; =20 __skb_unlink(skb, &ssk->sk_receive_queue); @@ -448,13 +454,14 @@ void __mptcp_sync_rcv_sequence(struct sock *sk) if (!skb) return; =20 - MPTCP_SKB_CB(skb)->map_seq =3D mptcp_iasn(msk) - skb->len; + MPTCP_SKB_CB(skb)->map_seq64 =3D mptcp_iasn(msk) - skb->len; + MPTCP_SKB_CB(skb)->map_seq =3D (u32)MPTCP_SKB_CB(skb)->map_seq64; MPTCP_SKB_CB(skb)->end_seq =3D MPTCP_SKB_CB(skb)->map_seq + skb->len; } =20 static bool __mptcp_move_skb(struct sock *sk, struct sk_buff *skb) { - u64 copy_len =3D MPTCP_SKB_CB(skb)->end_seq - MPTCP_SKB_CB(skb)->map_seq; + u32 copy_len =3D MPTCP_SKB_CB(skb)->end_seq - MPTCP_SKB_CB(skb)->map_seq; struct mptcp_sock *msk =3D mptcp_sk(sk); struct sk_buff *tail; =20 @@ -472,7 +479,7 @@ static bool __mptcp_move_skb(struct sock *sk, struct sk= _buff *skb) if (unlikely(msk->rcvd_dummy_seq)) __mptcp_sync_rcv_sequence(sk); =20 - if (MPTCP_SKB_CB(skb)->map_seq =3D=3D msk->ack_seq) { + if (MPTCP_SKB_CB(skb)->map_seq64 =3D=3D msk->ack_seq) { /* in sequence */ insert: if (!mptcp_try_rmem_schedule(sk, skb)) { @@ -490,14 +497,14 @@ static bool __mptcp_move_skb(struct sock *sk, struct = sk_buff *skb) skb_set_owner_r(skb, sk); __skb_queue_tail(&sk->sk_receive_queue, skb); return true; - } else if (after64(MPTCP_SKB_CB(skb)->map_seq, msk->ack_seq)) { + } else if (after64(MPTCP_SKB_CB(skb)->map_seq64, msk->ack_seq)) { mptcp_data_queue_ofo(msk, skb); return false; } =20 /* Partial packet */ - if (after64(MPTCP_SKB_CB(skb)->end_seq, msk->ack_seq)) { - copy_len =3D MPTCP_SKB_CB(skb)->end_seq - msk->ack_seq; + if (after64(MPTCP_SKB_CB(skb)->map_seq64 + skb->len, msk->ack_seq)) { + copy_len =3D MPTCP_SKB_CB(skb)->end_seq - (u32)msk->ack_seq; goto insert; } =20 @@ -892,40 +899,40 @@ static bool __mptcp_ofo_queue(struct mptcp_sock *msk) { struct sock *sk =3D (struct sock *)msk; struct sk_buff *skb, *tail; + u32 seq_delta, ack_seq; bool moved =3D false; struct rb_node *p; - u64 end_seq; =20 p =3D rb_first(&msk->out_of_order_queue); pr_debug("msk=3D%p empty=3D%d\n", msk, RB_EMPTY_ROOT(&msk->out_of_order_q= ueue)); while (p) { + ack_seq =3D msk->ack_seq; skb =3D rb_to_skb(p); - if (after64(MPTCP_SKB_CB(skb)->map_seq, msk->ack_seq)) + if (after(MPTCP_SKB_CB(skb)->map_seq, ack_seq)) break; =20 p =3D rb_next(p); rb_erase(&skb->rbnode, &msk->out_of_order_queue); =20 - if (unlikely(!after64(MPTCP_SKB_CB(skb)->end_seq, - msk->ack_seq))) { + if (unlikely(!after(MPTCP_SKB_CB(skb)->end_seq, ack_seq))) { mptcp_drop(sk, skb); MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_DUPDATA); continue; } =20 - end_seq =3D MPTCP_SKB_CB(skb)->end_seq; + seq_delta =3D MPTCP_SKB_CB(skb)->end_seq - ack_seq; tail =3D skb_peek_tail(&sk->sk_receive_queue); if (!tail || !mptcp_try_coalesce(sk, tail, skb)) { - int delta =3D msk->ack_seq - MPTCP_SKB_CB(skb)->map_seq; + int delta =3D ack_seq - MPTCP_SKB_CB(skb)->map_seq; =20 /* skip overlapping data, if any */ - pr_debug("uncoalesced seq=3D%llx ack seq=3D%llx delta=3D%d\n", - MPTCP_SKB_CB(skb)->map_seq, msk->ack_seq, + pr_debug("uncoalesced seq=3D%x ack seq=3D%x delta=3D%d\n", + MPTCP_SKB_CB(skb)->map_seq, ack_seq, delta); __skb_queue_tail(&sk->sk_receive_queue, skb); } - msk->bytes_received +=3D end_seq - msk->ack_seq; - WRITE_ONCE(msk->ack_seq, end_seq); + msk->bytes_received +=3D seq_delta; + WRITE_ONCE(msk->ack_seq, msk->ack_seq + seq_delta); moved =3D true; } return moved; @@ -2204,7 +2211,7 @@ static int __mptcp_recvmsg_mskq(struct sock *sk, stru= ct msghdr *msg, int copied =3D 0; =20 skb_queue_walk_safe(&sk->sk_receive_queue, skb, tmp) { - u64 offset =3D *seq - MPTCP_SKB_CB(skb)->map_seq; + u32 offset =3D (u32)(*seq) - MPTCP_SKB_CB(skb)->map_seq; u32 data_len =3D skb->len - offset; u32 count; int err; @@ -4587,7 +4594,7 @@ static struct sk_buff *mptcp_recv_skb(struct sock *sk= , u32 *off) mptcp_move_skbs(sk); =20 while ((skb =3D skb_peek(&sk->sk_receive_queue)) !=3D NULL) { - offset =3D msk->copied_seq - MPTCP_SKB_CB(skb)->map_seq; + offset =3D (u32)msk->copied_seq - MPTCP_SKB_CB(skb)->map_seq; if (offset < skb->len) { *off =3D offset; return skb; @@ -4838,11 +4845,23 @@ static int mptcp_napi_poll(struct napi_struct *napi= , int budget) return work_done; } =20 +#define CHK_CB_FIELD(mptcp_field, tcp_field) \ + ({ \ + BUILD_BUG_ON(offsetof(struct mptcp_skb_cb, mptcp_field) !=3D \ + offsetof(struct tcp_skb_cb, tcp_field)); \ + BUILD_BUG_ON(offsetofend(struct mptcp_skb_cb, mptcp_field) !=3D \ + offsetofend(struct tcp_skb_cb, tcp_field)); \ + }) + void __init mptcp_proto_init(void) { struct mptcp_delegated_action *delegated; int cpu; =20 + CHK_CB_FIELD(map_seq, seq); + CHK_CB_FIELD(end_seq, end_seq); + CHK_CB_FIELD(flags, tcp_flags); + mptcp_prot.h.hashinfo =3D tcp_prot.h.hashinfo; =20 if (percpu_counter_init(&mptcp_sockets_allocated, 0, GFP_KERNEL)) diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h index 72a357b11b0f..3a677a88ee4c 100644 --- a/net/mptcp/protocol.h +++ b/net/mptcp/protocol.h @@ -129,9 +129,12 @@ #define MPTCP_SYNC_SEQ 8 =20 struct mptcp_skb_cb { - u64 map_seq; - u64 end_seq; + u32 map_seq; + u32 end_seq; + u32 unused; + u16 flags; u8 has_rxtstamp; + u64 map_seq64; }; =20 #define MPTCP_SKB_CB(__skb) ((struct mptcp_skb_cb *)&((__skb)->cb[0])) --=20 2.53.0 From nobody Sat Sep 26 12:28:38 2026 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-alma10-1.taild15c8.ts.net [100.103.45.18]) (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 B291E335066 for ; Sun, 23 Aug 2026 01:02:23 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=100.103.45.18 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1787446953; cv=none; b=fMNs3HfQyM94ZWzLChit4Skr7M7A3o2wmhpmk7olSTVGLUEGlxEoauAfnzEis378iXBfu/Q76tYXXQOgaT1ljuZ3yG/3MAIMC5ou99W+D+sjXqtwOy0+Ox3zv8aPlVVUMbT2Y/F3g33pNhOQSZvB2XxUm2IIay0sKGE1dY6XCJU= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1787446953; c=relaxed/simple; bh=FzZpPXJM/REiCK60jX3+6L5IR6pAe9N6tQs4rrcpDh0=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=C502b12ERnbWBTI/RpDl48mF3Bsvj5AhgdLtax+cY/BNQv8Zm+343KJEjMumY3hXVG+OgZmyRPpHNbV3fglYdox9sWNQQTAzbk4o/yracVmaPvR5SpzYSJxTAqOdVvq8UymcWtmhEcNOAe0kmolJtBmEhjpF2slsWuW5kh8hyAM= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=F6bZvwkA; arc=none smtp.client-ip=100.103.45.18 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="F6bZvwkA" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 2DD491F00AC4; Sun, 23 Aug 2026 01:02:16 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1787446938; bh=dhe3mQCFRcDvmP6ACUABzvGaPjcIMbJPYWgER9Qdn5k=; h=From:To:Cc:Subject:Date:In-Reply-To:References; b=F6bZvwkAbv3SpJGdzFOHp9uR6b+1K7PgHXMyqQFtIfz8qD6ZfBSaGUrpmEtkEYrpF nyU5proX1pYnCXVoKKNOCA1aExLfm377TBmNaoAkPpaeK4IvMoVQntsG59xc0iGTLW Gd5fH7aNUDF7m9AkftTUChKOj4gxhn670ffPlTre1zLsp6x+fwoGMHmnAeoBpceQsB VtfgG4w89F+f/i+to6QkH+0OuTXBjU4SkZ5GxLaWX7VVwYPU4nFNSFU3KXkIdUk8gk uuvAsIW/Z+lO7OWm+nDDwto8rdlxZTUJISpNHIRI6CMMPZIujYxEdxLp7HehZBuitM eWPrkCRXSISxQ== From: Geliang Tang To: mptcp@lists.linux.dev Cc: Paolo Abeni , Geliang Tang Subject: [PATCH mptcp-next v6 5/7] mptcp: defer read_sock cleanup to mptcp_worker Date: Sun, 23 Aug 2026 09:01:35 +0800 Message-ID: <9c6520cbada2fb6554ee360c0017d32bab84fb9b.1787446274.git.tanggeliang@kylinos.cn> X-Mailer: git-send-email 2.53.0 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: Paolo Abeni When MPTCP carries TLS, the data path runs under mptcp_data_lock(). Reaching sk->sk_data_ready(sk) synchronously ends up at tls_strp_check_rcv() -> mptcp_recv_skb() -> mptcp_move_skbs(), which calls mptcp_data_lock() on the same sk and recurses on sk_lock.slock. The TLS path is not the only constraint: before the mptcp_recv_skb() calls, the TLS code would also reach __mptcp_read_sock(), which calls mptcp_rcv_space_adjust() and mptcp_cleanup_rbuf(). Both require holding the msk socket lock in process context, while the mptcp/TLS caller is in BH scope. Fix this by deferring sk->sk_data_ready(sk) to mptcp_worker() via a new MPTCP_WORK_READ_COMPLETE bit, reusing the existing mptcp_schedule_work()/ mptcp_cancel_work() infrastructure. The wakeup bit is consumed after the SOCK_DEAD && TCP_CLOSE destroy branch, so a socket that reaches the destroy path drops the pending wakeup rather than running it post-free. Co-developed-by: Geliang Tang Signed-off-by: Geliang Tang Signed-off-by: Paolo Abeni --- net/mptcp/protocol.c | 25 +++++++++++++++++++------ net/mptcp/protocol.h | 2 ++ 2 files changed, 21 insertions(+), 6 deletions(-) diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index 2e0e8e85b4d0..3e3461e5804e 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -3136,6 +3136,15 @@ static void mptcp_backlog_purge(struct sock *sk) sk_mem_reclaim(sk); } =20 +static void mptcp_read_complete(struct sock *sk) +{ + struct mptcp_sock *msk =3D mptcp_sk(sk); + + mptcp_rcv_space_adjust(msk, msk->read_copied); + mptcp_cleanup_rbuf(msk, msk->read_copied); + msk->read_copied =3D 0; +} + static void mptcp_do_fastclose(struct sock *sk) { struct mptcp_subflow_context *subflow, *tmp; @@ -3212,6 +3221,9 @@ static void mptcp_worker(struct work_struct *work) if (test_and_clear_bit(MPTCP_WORK_RTX, &msk->flags)) __mptcp_retrans(sk); =20 + if (test_and_clear_bit(MPTCP_WORK_READ_COMPLETE, &msk->flags)) + mptcp_read_complete(sk); + fail_tout =3D msk->first ? READ_ONCE(mptcp_subflow_ctx(msk->first)->fail_= tout) : 0; if (fail_tout && time_after(jiffies, fail_tout)) mptcp_mp_fail_no_response(msk); @@ -4590,9 +4602,6 @@ static struct sk_buff *mptcp_recv_skb(struct sock *sk= , u32 *off) struct sk_buff *skb; u32 offset; =20 - if (!list_empty(&msk->backlog_list)) - mptcp_move_skbs(sk); - while ((skb =3D skb_peek(&sk->sk_receive_queue)) !=3D NULL) { offset =3D (u32)msk->copied_seq - MPTCP_SKB_CB(skb)->map_seq; if (offset < skb->len) { @@ -4607,6 +4616,7 @@ static struct sk_buff *mptcp_recv_skb(struct sock *sk= , u32 *off) /* * Note: * - It is assumed that the socket was locked by the caller. + * - Can be invoked in BH scope. */ static int __mptcp_read_sock(struct sock *sk, read_descriptor_t *desc, sk_read_actor_t recv_actor, bool noack) @@ -4648,11 +4658,14 @@ static int __mptcp_read_sock(struct sock *sk, read_= descriptor_t *desc, if (noack) goto out; =20 - mptcp_rcv_space_adjust(msk, copied); - + /* The backlog flushing is only needed when some data is actually + * moved and will take place in the workers's release callback. + */ if (copied > 0) { mptcp_recv_skb(sk, &offset); - mptcp_cleanup_rbuf(msk, copied); + msk->read_copied +=3D copied; + set_bit(MPTCP_WORK_READ_COMPLETE, &msk->flags); + mptcp_schedule_work(sk); } out: return copied; diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h index 3a677a88ee4c..f08803593d66 100644 --- a/net/mptcp/protocol.h +++ b/net/mptcp/protocol.h @@ -117,6 +117,7 @@ #define MPTCP_FALLBACK_DONE 2 #define MPTCP_WORK_CLOSE_SUBFLOW 3 #define MPTCP_RTX_DISABLED 4 +#define MPTCP_WORK_READ_COMPLETE 5 =20 /* MPTCP socket release cb flags */ #define MPTCP_PUSH_PENDING 1 @@ -312,6 +313,7 @@ struct mptcp_sock { u32 last_data_sent; u32 last_data_recv; u32 last_ack_recv; + int read_copied; unsigned long timer_ival; u32 token; u32 tfo_skb_len; --=20 2.53.0 From nobody Sat Sep 26 12:28:38 2026 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-alma10-1.taild15c8.ts.net [100.103.45.18]) (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 5583E35F8C5 for ; Sun, 23 Aug 2026 01:02:23 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=100.103.45.18 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1787446956; cv=none; b=DIjcAp/CBbgAU+Ocl7z3WPeCVuZU7sZNAZLorlsYW+/ItJNWQ6e4YDPxyWWXZNr6n9JHloxBe8LQfaTHOn+h8fGqNKYOeKwU7tUnMYzs9izHGmhjEEXbmrMhAjsHVNdR4rMMLzi01JYh4raK/dvxZ13jTE+Rx2HpOAp6zl71twk= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1787446956; c=relaxed/simple; bh=NbfCdi3r/GOiL5ulNdIZ+cgQbCOawzaVKvvVkNd/Trk=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=uDE3kq/+d2iJOZaIYAm0fAR3vsP7E57UB2iBEoHsI/hpkmzsvtwrDoOHlEs34qnSsW7UupmPOcAFkqhiqLcUFbocUXkJK/N41m2FBwfgV5fWDrKbLaLIXrzo8ihKkQDfpx6t+WHNFzBzH2KWFmW/UfGgZ/9njfkkBRZ84UN6mns= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=KOtHUhHt; arc=none smtp.client-ip=100.103.45.18 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="KOtHUhHt" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 98BD31F00ACA; Sun, 23 Aug 2026 01:02:18 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1787446939; bh=AQIQ9AtDZ1q565UJ8CKzsH601OoqxmLBA65HcXTIQ1A=; h=From:To:Cc:Subject:Date:In-Reply-To:References; b=KOtHUhHt7XnpidTC35reC/Oir1o7aCJMgSZSn72FVkCoQMOcq7TH0LKYVuGlMYSW6 EBj61HqfFzKtZ3nOrcLeYtxZgvRCd/yWV1R3XDzJTFAhOcOVptBfnFtus6DKkqsKfh HfKlUV+DnlqwZyWejhjirzY/xxceFmXKoWS6YSQjOwewJp2zi8mJvgXu4AXjgDo3MX Dy3PU8dVgQa3D4QhP1wmN8HV0qyc8hKAVVhWoteQPfecR/eIit4hPlKMPSoWRIbFYA I5nWPiNxMV5US73zpPuiE9L3eWm0iHsoFdxOS5rzNIr0Wf1tCWNyMCVpifhgFZdcVD 8Zdf6/ePE204Q== From: Geliang Tang To: mptcp@lists.linux.dev Cc: Geliang Tang Subject: [PATCH mptcp-next v6 6/7] mptcp: track app-limited state in mptcp_sendmsg Date: Sun, 23 Aug 2026 09:01:36 +0800 Message-ID: X-Mailer: git-send-email 2.53.0 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 The application-limited accounting in TCP is updated by tcp_rate_check_app_limited(), which currently takes a struct sock * and internally calls tcp_sk(). MPTCP needs to apply the same accounting to each subflow individually - every subflow is an independent TCP socket with its own tp->app_limited / delivered state - so wrapping the call as a struct sock * -> tcp_sk() helper is awkward at the call site. Split the existing function: keep the logic as tcp_sock_rate_check_app_limited(struct tcp_sock *tp), and turn tcp_rate_check_app_limited(struct sock *) into a thin wrapper so the exported API is unchanged for other TCP users. Then add mptcp_sock_rate_check_app_limited() that walks every subflow of the mptcp_sock and runs tcp_sock_rate_check_app_limited() under each subflow's socket lock. Invoke it from mptcp_sendmsg() right after the send-side setup, so the delivery-rate app_limited state stays in sync with what the application actually has to send across each subflow. With this in place, TCP_INFO.tcpi_delivery_rate_app_limited is reported correctly for MPTCP connections instead of being left at 0. Signed-off-by: Geliang Tang --- include/net/tcp.h | 1 + net/ipv4/tcp.c | 9 +++++++-- net/mptcp/protocol.c | 18 ++++++++++++++++++ 3 files changed, 26 insertions(+), 2 deletions(-) diff --git a/include/net/tcp.h b/include/net/tcp.h index 2c5b889530b5..9d436816411b 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -849,6 +849,7 @@ static inline int tcp_bound_to_half_wnd(struct tcp_sock= *tp, int pktsize) =20 /* tcp.c */ void tcp_get_info(struct sock *, struct tcp_info *); +void tcp_sock_rate_check_app_limited(struct tcp_sock *tp); void tcp_rate_check_app_limited(struct sock *sk); =20 /* Read 'sendfile()'-style from a TCP socket */ diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 1452c40e22d4..f0f843fd1897 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -1096,9 +1096,9 @@ int tcp_sendmsg_fastopen(struct sock *sk, struct msgh= dr *msg, int *copied, } =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 && @@ -1111,6 +1111,11 @@ 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); =20 int tcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_t size) diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index 3e3461e5804e..f39c62f956c4 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -2048,6 +2048,21 @@ static void mptcp_rps_record_subflows(const struct m= ptcp_sock *msk) } } =20 +static void mptcp_sock_rate_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_nested(ssk); + tcp_sock_rate_check_app_limited(tcp_sk(ssk)); + unlock_sock_fast(ssk, slow); + } +} + static int mptcp_sendmsg(struct sock *sk, struct msghdr *msg, size_t len) { struct mptcp_sock *msk =3D mptcp_sk(sk); @@ -2078,6 +2093,9 @@ static int mptcp_sendmsg(struct sock *sk, struct msgh= dr *msg, size_t len) =20 timeo =3D sock_sndtimeo(sk, msg->msg_flags & MSG_DONTWAIT); =20 + /* is sending application-limited? */ + mptcp_sock_rate_check_app_limited(sk); + if ((1 << sk->sk_state) & ~(TCPF_ESTABLISHED | TCPF_CLOSE_WAIT)) { ret =3D sk_stream_wait_connect(sk, &timeo); if (ret) --=20 2.53.0 From nobody Sat Sep 26 12:28:38 2026 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-alma10-1.taild15c8.ts.net [100.103.45.18]) (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 AEFB223ABBF for ; Sun, 23 Aug 2026 01:02:29 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=100.103.45.18 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1787446963; cv=none; b=J4cEZK8SZThFg0V09gcVJfFIfT2SLrqBO4/WdxMx1/YZgTNyO5PsYzW65RL/+Okc3NcV+pMfL64DRfwNOifapN8M1bty8eACVi/u/gFOwYOcTL97KDp+hAFvLMegu8eTkkR5nGr+qXoWbYVipLkncGNYtyAr5NMN20M+Sd1oREw= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1787446963; c=relaxed/simple; bh=JxT3qeiSHuBS12KtGnTPLFRzVxKd0fvCXwqrRkI3nnM=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=szO5yKVgk6OlbZtcG9rVZlGbJ0rx3PcZ80bRDiSLPVVGkkj82Wpc2Hvv1GkWL4AdsZvGtq7CLsv3eztRjr9SoFP4tn5kGJVMAbEWOaLSttR7IN7eYxbUsNvBV0iGX+XtQAJ0YZ7VUpAquWLkBMiGuekgVbPpDpzUsI9AyIMLCyI= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=RU01Y5VN; arc=none smtp.client-ip=100.103.45.18 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="RU01Y5VN" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 0EE711F000E9; Sun, 23 Aug 2026 01:02:19 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1787446940; bh=xNNDwMcimuhvgxiiYj/oGzahS6NB1VtBfCB7+LyvfjA=; h=From:To:Cc:Subject:Date:In-Reply-To:References; b=RU01Y5VNjX5S4XwsGprD+ZxbpNKz/qMmEl6feq1sk3UfLc5jDUxiOWU1jYeGwoUWP PKXsadoRL7GSZsrMghJlVWU9UpmVbkeAaEoT4y9C+gZx70WVaDqgejh0mjL8jlw3gF r2B+fP0TDRA3UYSFFyFI+eZcXmP8s2pV+IrIcJGM8LCQm/ZGehEZ59kRttMuimbgs3 oRMmgESGCBHYcKpXHFSCUvKov//kH44VLCr2CXibPy8AbcofJAucWqavf86b7G/BdU OgDDzHs8Zw2+zFCuKovOYz6DQJmh9/tg7dLSdUpiVkq4UnKKuPOovmfmoSNdVOBxNt +YDMExNzHo9mg== From: Geliang Tang To: mptcp@lists.linux.dev Cc: Geliang Tang Subject: [PATCH mptcp-next v6 7/7] selftests: mptcp: sockopt: check app_limited Date: Sun, 23 Aug 2026 09:01:37 +0800 Message-ID: <39b2a50221fdd8aea820d7b5ad9359cc40dff108.1787446274.git.tanggeliang@kylinos.cn> X-Mailer: git-send-email 2.53.0 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 connect_one_server() in mptcp_sockopt exchanges only a few packets between the client and server, then closes the socket. After such a small transfer the application has nothing further to send, so the connection is, by definition, application-limited. Extend the TCP_INFO readback at the end of the function to assert s.tcp_info.tcpi_delivery_rate_app_limited =3D=3D 1. Without the preceding commit, mptcp_sendmsg() never updates the per-subflow app-limited state, and this field stays at 0 - the assertion would fail. With it in place, the value is forced to 1, turning this into a regression guard for the subflow-side application-limited accounting. Signed-off-by: Geliang Tang --- tools/testing/selftests/net/mptcp/mptcp_sockopt.c | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/testing/selftests/net/mptcp/mptcp_sockopt.c b/tools/test= ing/selftests/net/mptcp/mptcp_sockopt.c index d68515b7903b..8d712bdb4325 100644 --- a/tools/testing/selftests/net/mptcp/mptcp_sockopt.c +++ b/tools/testing/selftests/net/mptcp/mptcp_sockopt.c @@ -640,6 +640,7 @@ static void connect_one_server(int fd, int pipefd) total +=3D 1; /* sequence advances due to FIN */ =20 assert(s.mptcpi_rcv_delta =3D=3D (uint64_t)total); + assert(s.tcp_info.tcpi_delivery_rate_app_limited =3D=3D 1); close(fd); } =20 --=20 2.53.0