From nobody Fri Jul 24 21:29:32 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 DB66B439351; Fri, 24 Jul 2026 14:02:38 +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=1784901760; cv=none; b=Wid44g6LaZS6GP7sTwwVlUOhftHGcJiRWc4nBGCTHqIGLsTtKW73DTkPGGa4C7D6LW/bqxvPAY+vi61PlGLNGRxTWIQBbODk+oUSUYEKfuOzUBinQV7o8y2X2bqXzlk9iBFeo+DzabrAHmdg+XHRd/rSZshgIsk3ymxUXYRrIRM= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1784901760; c=relaxed/simple; bh=++AX/TdTvWwgODLnwbzLSGJMOH0kealvdqqMPU1BGdY=; h=From:Date:Subject:MIME-Version:Content-Type:Message-Id:References: In-Reply-To:To:Cc; b=iTJKha3cICHqjEOIUXSDKz1v1kNa/kV0JIK50NP0U9AZHsyhW+w0lU+H2ZzwZSmGc0OU7tfmjvhJrV1ZgF3PYDVmU3aEIdegZy1LqB6r9XOHbv+Nhtd0iCsJqgmyBbVsNfNN7+/AxtOaQs/MoGKsHn6YyUkA0Mj98sbvV6N339k= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=b8rAVKSH; 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="b8rAVKSH" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 748651F00A3D; Fri, 24 Jul 2026 14:02:36 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1784901758; bh=xB/FsxzRoiYwJDzNNomLKMj1scIXpQggJ8ZrpM9x+Gs=; h=From:Date:Subject:References:In-Reply-To:To:Cc; b=b8rAVKSHocqGnURnCL2skSbgnMfBVfXso+rh+xSghJfgMPkI8qedD69sg1KoiiwwV NcFEbdmDhBamdeh48KBMsCg0Dlkx30b0tRaAOa9ztZXivA9/ZMK+dsXXIyJaVW7Qe4 uLyRtvS86vBv67UgxMbg8FZswFSaUfT+4ywxX0OzwIPuSWHpRCBJxxhyldrrJHaoTE /Ji+0/PjgWuP++2c4urWNSQ3X0jeqDDZlsFnZtsiJ0FtI4a1HOKDvNgFoVb1HkbcII uHOQwCp0thtAPlRAUTV76JqM4HICIH/UakEGqg+LvJlx592BPdGW9+1cp4YpY/v2mz L4ZnPD0lQbKxA== From: "Matthieu Baerts (NGI0)" Date: Fri, 24 Jul 2026 16:02:13 +0200 Subject: [PATCH net-next 1/5] mptcp: move the retrans loop to a separate helper Precedence: bulk X-Mailing-List: mptcp@lists.linux.dev List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Message-Id: <20260724-net-next-mptcp-oooq-pruning-v1-1-5dd4dec63a54@kernel.org> References: <20260724-net-next-mptcp-oooq-pruning-v1-0-5dd4dec63a54@kernel.org> In-Reply-To: <20260724-net-next-mptcp-oooq-pruning-v1-0-5dd4dec63a54@kernel.org> To: Mat Martineau , Geliang Tang , "David S. Miller" , Eric Dumazet , Jakub Kicinski , Paolo Abeni , Simon Horman Cc: netdev@vger.kernel.org, mptcp@lists.linux.dev, linux-kernel@vger.kernel.org, "Matthieu Baerts (NGI0)" , Gang Yan X-Mailer: b4 0.15.2 X-Developer-Signature: v=1; a=openpgp-sha256; l=3211; i=matttbe@kernel.org; h=from:subject:message-id; bh=roDpodaahk2BWLOIYD/W304v8cqk1FmemTEUD9iezUs=; b=owGbwMvMwCVWo/Th0Gd3rumMp9WSGLKSC8quFP348ct6me+M875Hvd3mf9reUrsqq5Kz+dyHs 4m/Y/91dpSyMIhxMciKKbJIt0Xmz3xexVvi5WcBM4eVCWQIAxenAEzkggYjw6WvSqZKr3dvs2Ge l35ndmXwrqwJcu4nOHe6LEo+Wb9NO5Phf2jPfu3LR+ISbzZVx9m8W3T25nole4O5KXu8/ac0mhW e5gAA X-Developer-Key: i=matttbe@kernel.org; a=openpgp; fpr=E8CB85F76877057A6E27F77AF6B7824F4269A073 From: Paolo Abeni This is a cleanup in order to make the next patch simpler. No functional change intended. Tested-by: Gang Yan Tested-by: Geliang Tang Acked-by: Geliang Tang Signed-off-by: Paolo Abeni Signed-off-by: Matthieu Baerts (NGI0) --- net/mptcp/protocol.c | 74 ++++++++++++++++++++++++++++++------------------= ---- 1 file changed, 43 insertions(+), 31 deletions(-) diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index ca644ec53eed..290d14e2fa5b 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -2785,41 +2785,14 @@ static void mptcp_check_fastclose(struct mptcp_sock= *msk) sk_error_report(sk); } =20 -static void __mptcp_retrans(struct sock *sk) +/* Retransmit the specified data fragment on all the selected subflows. */ +static int __mptcp_push_retrans(struct sock *sk, struct mptcp_data_frag *d= frag) { struct mptcp_sendmsg_info info =3D { .data_lock_held =3D true, }; struct mptcp_sock *msk =3D mptcp_sk(sk); struct mptcp_subflow_context *subflow; - struct mptcp_data_frag *dfrag; struct sock *ssk; - int ret, err; - u16 len =3D 0; - - mptcp_clean_una_wakeup(sk); - - /* first check ssk: need to kick "stale" logic */ - err =3D mptcp_sched_get_retrans(msk); - dfrag =3D mptcp_rtx_head(sk); - if (!dfrag) { - if (mptcp_data_fin_enabled(msk)) { - struct inet_connection_sock *icsk =3D inet_csk(sk); - - WRITE_ONCE(icsk->icsk_retransmits, - icsk->icsk_retransmits + 1); - mptcp_set_datafin_timeout(sk); - mptcp_send_ack(msk); - - goto reset_timer; - } - - if (!mptcp_send_head(sk)) - goto clear_scheduled; - - goto reset_timer; - } - - if (err) - goto reset_timer; + int ret, len =3D 0; =20 mptcp_for_each_subflow(msk, subflow) { if (READ_ONCE(subflow->scheduled)) { @@ -2847,7 +2820,7 @@ static void __mptcp_retrans(struct sock *sk) !msk->allow_subflows) { spin_unlock_bh(&msk->fallback_lock); release_sock(ssk); - goto clear_scheduled; + return -1; } =20 while (info.sent < info.limit) { @@ -2870,6 +2843,45 @@ static void __mptcp_retrans(struct sock *sk) release_sock(ssk); } } + return len; +} + +static void __mptcp_retrans(struct sock *sk) +{ + struct mptcp_sock *msk =3D mptcp_sk(sk); + struct mptcp_subflow_context *subflow; + struct mptcp_data_frag *dfrag; + int err, len; + + mptcp_clean_una_wakeup(sk); + + /* first check ssk: need to kick "stale" logic */ + err =3D mptcp_sched_get_retrans(msk); + dfrag =3D mptcp_rtx_head(sk); + if (!dfrag) { + if (mptcp_data_fin_enabled(msk)) { + struct inet_connection_sock *icsk =3D inet_csk(sk); + + WRITE_ONCE(icsk->icsk_retransmits, + icsk->icsk_retransmits + 1); + mptcp_set_datafin_timeout(sk); + mptcp_send_ack(msk); + + goto reset_timer; + } + + if (!mptcp_send_head(sk)) + goto clear_scheduled; + + goto reset_timer; + } + + if (err) + goto reset_timer; + + len =3D __mptcp_push_retrans(sk, dfrag); + if (len < 0) + goto clear_scheduled; =20 msk->bytes_retrans +=3D len; dfrag->already_sent =3D max(dfrag->already_sent, len); --=20 2.53.0 From nobody Fri Jul 24 21:29:32 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 6D3B133D4E8; Fri, 24 Jul 2026 14:02:40 +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=1784901764; cv=none; b=CxiQawQE5x/6JBZcJZjQO31c7keIX23MxCc2tG9JtrVttaJ4SHRhoDeN1zBc8A0Lh3Q/NDpp2CDqW/eAJtDbLn7lykfJm/4kck72PVK2yB+aMpRdNAspwg51ue9uyFS6Mc15fdI0jBzyUxdtYf0b8iBei0uiZPBJ5FQ0nwNL18Q= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1784901764; c=relaxed/simple; bh=+RqkXn1fraP6cVhzylmca3p6JDvhLR0jInwTN075Ip4=; h=From:Date:Subject:MIME-Version:Content-Type:Message-Id:References: In-Reply-To:To:Cc; b=GbU9Z2WnlfRJi3DVcGoX15WsCBYTePcncqP14JSRF1t7Uyv4Tgf3PAzlukkgfa2KejySOGP1kjvhOFd3IRj9m18ROZWHbtlpeOXKBXQJYIDfJGZWaU8AhlGd995nLmYAd+mFgQ+6yuVq+dYCAUTON5ohJH3+L79e40fpTgojpGs= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=gSSQ3nen; 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="gSSQ3nen" Received: by smtp.kernel.org (Postfix) with ESMTPSA id D1DD41F000E9; Fri, 24 Jul 2026 14:02:38 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1784901760; bh=5+GFzuFgZQNxbrz5hLonyNttgqfCDeTaSqr0K2b4YtA=; h=From:Date:Subject:References:In-Reply-To:To:Cc; b=gSSQ3nennMCw/SD7ZrKW6VtiWg+tRTcbcWKr0xDOs1U5nyymvLNnZw4ha7+Fw/q4R Z5OCz5QStfnCEQDN9opeD+r0rGze5DnMBMQsYvq2XJrtOr72yzM18TlJaKm6reRxyB jQLz87QQmXnQlQ4Z91JBqrAxqRszYkTS8GsN/0wsREznnkf/IMDNgN45ns2hiOFDKY xEDotfsVFql4uywJSlOQbxx2y7vRWSQNK8Fg2OfBNpeksmuKNpYMoGiXy5Qktbdj0G sPOUAn8uXfJiAPBRswVuaPmi62mOW+Hh5GguhPK3NUUWbf9xrC/yNpMC8UcLgj5n96 K4gxLAxDGzDig== From: "Matthieu Baerts (NGI0)" Date: Fri, 24 Jul 2026 16:02:14 +0200 Subject: [PATCH net-next 2/5] mptcp: let the retrans scheduler do its job Precedence: bulk X-Mailing-List: mptcp@lists.linux.dev List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Message-Id: <20260724-net-next-mptcp-oooq-pruning-v1-2-5dd4dec63a54@kernel.org> References: <20260724-net-next-mptcp-oooq-pruning-v1-0-5dd4dec63a54@kernel.org> In-Reply-To: <20260724-net-next-mptcp-oooq-pruning-v1-0-5dd4dec63a54@kernel.org> To: Mat Martineau , Geliang Tang , "David S. Miller" , Eric Dumazet , Jakub Kicinski , Paolo Abeni , Simon Horman Cc: netdev@vger.kernel.org, mptcp@lists.linux.dev, linux-kernel@vger.kernel.org, "Matthieu Baerts (NGI0)" , Gang Yan X-Mailer: b4 0.15.2 X-Developer-Signature: v=1; a=openpgp-sha256; l=6309; i=matttbe@kernel.org; h=from:subject:message-id; bh=kPuyETNsuPEY+YBx+aLERj0Xbk9Jfu2rb6IkdWMvqvM=; b=kA0DAAoWfCLwwvNHCpcByyZiAGpjcHbILWWRYYtGuVhUu4biMo4n4Tp1s8ERLhXtfTdkCLi6A Yh1BAAWCgAdFiEEG4ZZb5nneg10Sk44fCLwwvNHCpcFAmpjcHYACgkQfCLwwvNHCpc8PgEAjXaW 0QQavdkpMGOcqsYcVHDIKZ9mvduOidAzlE0P4GUBAJ5qZ0+nTIB426TssJk1krSIRxSe1bFTwl8 /SugDuHUJ X-Developer-Key: i=matttbe@kernel.org; a=openpgp; fpr=E8CB85F76877057A6E27F77AF6B7824F4269A073 From: Paolo Abeni Currently the MPTCP core enforces that when MPTCP-level retrans timer fires, at most a single dfrag is retransmitted. In some corner-cases, it may be necessary to retransmit multiple dfrags, and the MPTCP socket will need to wait multiple retrans timeout to accomplish that. Remove the mentioned constraint, allowing to transmit multiple dfrags per retrans period, as long as the scheduler keeps selecting subflows for retransmissions and pending data is available in the rtx queue. The default scheduler will transmit a dfrag per available subflow. Tested-by: Gang Yan Tested-by: Geliang Tang Acked-by: Geliang Tang Signed-off-by: Paolo Abeni Signed-off-by: Matthieu Baerts (NGI0) --- net/mptcp/protocol.c | 119 +++++++++++++++++++++++++++++++++++++----------= ---- 1 file changed, 87 insertions(+), 32 deletions(-) diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index 290d14e2fa5b..72b1fa3ca71c 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -1136,13 +1136,6 @@ static void __mptcp_clean_una_wakeup(struct sock *sk) mptcp_write_space(sk); } =20 -static void mptcp_clean_una_wakeup(struct sock *sk) -{ - mptcp_data_lock(sk); - __mptcp_clean_una_wakeup(sk); - mptcp_data_unlock(sk); -} - static void mptcp_enter_memory_pressure(struct sock *sk) { struct mptcp_subflow_context *subflow; @@ -2785,8 +2778,12 @@ static void mptcp_check_fastclose(struct mptcp_sock = *msk) sk_error_report(sk); } =20 -/* Retransmit the specified data fragment on all the selected subflows. */ -static int __mptcp_push_retrans(struct sock *sk, struct mptcp_data_frag *d= frag) +/* + * Retransmit the specified data fragment on all the selected subflows, + * starting from the specified sequence + */ +static int __mptcp_push_retrans(struct sock *sk, struct mptcp_data_frag *d= frag, + u64 sent_seq) { struct mptcp_sendmsg_info info =3D { .data_lock_held =3D true, }; struct mptcp_sock *msk =3D mptcp_sk(sk); @@ -2796,6 +2793,7 @@ static int __mptcp_push_retrans(struct sock *sk, stru= ct mptcp_data_frag *dfrag) =20 mptcp_for_each_subflow(msk, subflow) { if (READ_ONCE(subflow->scheduled)) { + u16 offset =3D sent_seq - dfrag->data_seq; u16 copied =3D 0; =20 mptcp_subflow_set_scheduled(subflow, false); @@ -2805,7 +2803,7 @@ static int __mptcp_push_retrans(struct sock *sk, stru= ct mptcp_data_frag *dfrag) lock_sock(ssk); =20 /* limit retransmission to the bytes already sent on some subflows */ - info.sent =3D 0; + info.sent =3D offset; info.limit =3D READ_ONCE(msk->csum_enabled) ? dfrag->data_len : dfrag->already_sent; =20 @@ -2851,14 +2849,88 @@ static void __mptcp_retrans(struct sock *sk) struct mptcp_sock *msk =3D mptcp_sk(sk); struct mptcp_subflow_context *subflow; struct mptcp_data_frag *dfrag; + bool need_retrans; + u64 retrans_seq; int err, len; =20 - mptcp_clean_una_wakeup(sk); - - /* first check ssk: need to kick "stale" logic */ - err =3D mptcp_sched_get_retrans(msk); + mptcp_data_lock(sk); + __mptcp_clean_una_wakeup(sk); + retrans_seq =3D msk->snd_una; dfrag =3D mptcp_rtx_head(sk); - if (!dfrag) { + need_retrans =3D !!dfrag; + mptcp_data_unlock(sk); + if (!dfrag) + goto check_data_fin; + + for (;;) { + bool already_retrans; + u64 sent_seq; + + /* The default scheduler will kick "stale" logic, that in + * turn can process incoming acks and clean the RTX queue; + * ensure that the current dfrag will still be around + * afterwards. + */ + get_page(dfrag->page); + err =3D mptcp_sched_get_retrans(msk); + if (err) { + put_page(dfrag->page); + break; + } + + /* Incoming acks can have moved retrans sequence after + * the current dfrag, if so try to start again from RTX head. + */ + mptcp_data_lock(sk); + already_retrans =3D !before64(msk->snd_una, dfrag->data_seq + + dfrag->already_sent); + put_page(dfrag->page); + if (already_retrans) { + __mptcp_clean_una_wakeup(sk); + retrans_seq =3D msk->snd_una; + dfrag =3D mptcp_rtx_head(sk); + need_retrans =3D !!dfrag; + } else if (after64(msk->snd_una, retrans_seq)) { + retrans_seq =3D msk->snd_una; + } + mptcp_data_unlock(sk); + + /* `already_sent` can be 0 for `dfrag` belonging to the RTX + * queue due to __mptcp_retransmit_pending_data(). + */ + if (!dfrag || !dfrag->already_sent) + break; + + /* Can fail only in case of fallback. */ + len =3D __mptcp_push_retrans(sk, dfrag, retrans_seq); + if (len < 0) + goto clear_scheduled; + + retrans_seq +=3D len; + msk->bytes_retrans +=3D len; + dfrag->already_sent =3D max_t(u16, dfrag->already_sent, + retrans_seq - dfrag->data_seq); + + /* With csum enabled retransmission can send new data. */ + sent_seq =3D dfrag->already_sent + dfrag->data_seq; + if (after64(sent_seq, msk->snd_nxt)) + WRITE_ONCE(msk->snd_nxt, sent_seq); + + /* Attempt the next fragment only if the current one is + * completely retransmitted. + */ + if (before64(retrans_seq, dfrag->data_seq + dfrag->data_len)) + break; + + dfrag =3D list_is_last(&dfrag->list, &msk->rtx_queue) ? + NULL : list_next_entry(dfrag, list); + if (!dfrag) + break; + } + + /* Attempt data-fin retransmission only when the RTX queue is empty. */ + if (!need_retrans) { +check_data_fin: if (mptcp_data_fin_enabled(msk)) { struct inet_connection_sock *icsk =3D inet_csk(sk); =20 @@ -2866,30 +2938,13 @@ static void __mptcp_retrans(struct sock *sk) icsk->icsk_retransmits + 1); mptcp_set_datafin_timeout(sk); mptcp_send_ack(msk); - goto reset_timer; } =20 if (!mptcp_send_head(sk)) goto clear_scheduled; - - goto reset_timer; } =20 - if (err) - goto reset_timer; - - len =3D __mptcp_push_retrans(sk, dfrag); - if (len < 0) - goto clear_scheduled; - - msk->bytes_retrans +=3D len; - dfrag->already_sent =3D max(dfrag->already_sent, len); - - /* With csum enabled retransmission can send new data. */ - if (after64(dfrag->already_sent + dfrag->data_seq, msk->snd_nxt)) - WRITE_ONCE(msk->snd_nxt, dfrag->already_sent + dfrag->data_seq); - reset_timer: mptcp_check_and_set_pending(sk); =20 --=20 2.53.0 From nobody Fri Jul 24 21:29:32 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 A76D6439906; Fri, 24 Jul 2026 14:02:44 +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=1784901767; cv=none; b=Csa1JQqQpUETab8fPj4DP/y7MZzlDYjRzNtMs8oPpFhxaM9rdFvP26dAkjZDveHzmXh1GHYjzXnVTBCt6lKdCZdpE7nsAdb5LwF7CND0BsxX8H+76gqstme5mz1hOL0zTz4HvVgQq0UbfazLukKqb2yO0Uvwh+IYHGMkNiqr1s0= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1784901767; c=relaxed/simple; bh=tRCs/Ywn6YThfo2FendTPgyTtyAcAJvsU0yUabYM+8k=; h=From:Date:Subject:MIME-Version:Content-Type:Message-Id:References: In-Reply-To:To:Cc; b=t45L54QtNAkSOOpjMiG8RPk8g1BPooYJu1VA1nafPObis97ATVdlH24CfBDE1IxSE3KH9HP/VU1t/nAbIkc4a7hX2k3SJ3lCMWv3sC1HmZGejmLL30m7gIGo/lNfbWlYBXsSGvmKy0vfGn2/FgeUmsD+izwtRubR5Vg7gk5VAGA= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=Q7wReyni; 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="Q7wReyni" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 3AEAE1F00A3A; Fri, 24 Jul 2026 14:02:41 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1784901763; bh=3L2urecN0rup7TreSmxzOfhOmTGRwimuPlF5G6bq1K8=; h=From:Date:Subject:References:In-Reply-To:To:Cc; b=Q7wReyniWU63uNctpmvnqXOYur5FW0JGM5HgkTQeY0ch29NYl9xhEfL+ootlpD/V4 BPjPV2DptUAYTe/iqt+52D4k6E6BgLpe+qqC8wh/ZWJRGXKD9Q/hASVSXTAD36gcN+ Y2Y+HnzJLHdWRAeZ+2Ju24Mq6xExp1rMrOOYuvqApWNTS3+M36w/7xk5qAnDIU9CpM aQPr6PtF7ZLvwGiVxpmUNz+K4z2aPfEq/u21uBpXIT5WQoksGXWOFsMcs+P+Ujhqv3 OJHpjpAVG63AThWKk6it5IU1O5IWQQcheS75JUCFUFTZA0IHlS6M7ZPmCK3CAlq2K3 hxTjQ/dW+Djaw== From: "Matthieu Baerts (NGI0)" Date: Fri, 24 Jul 2026 16:02:15 +0200 Subject: [PATCH net-next 3/5] mptcp: explicitly drop over memory limits Precedence: bulk X-Mailing-List: mptcp@lists.linux.dev List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Message-Id: <20260724-net-next-mptcp-oooq-pruning-v1-3-5dd4dec63a54@kernel.org> References: <20260724-net-next-mptcp-oooq-pruning-v1-0-5dd4dec63a54@kernel.org> In-Reply-To: <20260724-net-next-mptcp-oooq-pruning-v1-0-5dd4dec63a54@kernel.org> To: Mat Martineau , Geliang Tang , "David S. Miller" , Eric Dumazet , Jakub Kicinski , Paolo Abeni , Simon Horman Cc: netdev@vger.kernel.org, mptcp@lists.linux.dev, linux-kernel@vger.kernel.org, "Matthieu Baerts (NGI0)" X-Mailer: b4 0.15.2 X-Developer-Signature: v=1; a=openpgp-sha256; l=6495; i=matttbe@kernel.org; h=from:subject:message-id; bh=FdLKg8Yu88qWDJDxfs1mL1EwkqxvxVMEzBy1/9NUA/o=; b=owGbwMvMwCVWo/Th0Gd3rumMp9WSGLKSC8pz/c4/Xrq0w6U9Y56zysU11q5KiU73WCN3aP/Xc VsRWDCro5SFQYyLQVZMkUW6LTJ/5vMq3hIvPwuYOaxMIEMYuDgFYCJzyhkZzt92bT9QYZaT/q7L +uC0O8uFHk8129hwSpJNtLH8ajHPZYb/AXld7dEb7jSVvOk5qi9g/8+s+NFFGeOICrVja5znxNt zAwA= X-Developer-Key: i=matttbe@kernel.org; a=openpgp; fpr=E8CB85F76877057A6E27F77AF6B7824F4269A073 From: Paolo Abeni Currently the enforcement of the rcvbuf constraint is implemented when moving the skbs into the msk receive or OoO queue, keeping the incoming skbs in the subflow queue when over limits. Under significant memory pressure the above can cause permanent data transfer stalls, as the skb needed to make forward progress can be stuck in a subflow queue. Over memory limits, drop the incoming skb, relying on MPTCP-level retransmissions. Note that fallback socket must perform the limit before the skb reaches the subflow-level queue, as dropping an in-sequence already acked skb would break the stream. This is not a complete fix for the stall issue, as the drop strategy needs refinements that will come in the next patches. Signed-off-by: Paolo Abeni Reviewed-by: Matthieu Baerts (NGI0) Signed-off-by: Matthieu Baerts (NGI0) --- net/mptcp/mib.c | 2 ++ net/mptcp/mib.h | 2 ++ net/mptcp/options.c | 28 +++++++++++++++++++++++++--- net/mptcp/protocol.c | 31 +++++++++++++++++++++++-------- 4 files changed, 52 insertions(+), 11 deletions(-) diff --git a/net/mptcp/mib.c b/net/mptcp/mib.c index f23fda0c55a7..ef65e2df709f 100644 --- a/net/mptcp/mib.c +++ b/net/mptcp/mib.c @@ -85,6 +85,8 @@ static const struct snmp_mib mptcp_snmp_list[] =3D { SNMP_MIB_ITEM("SimultConnectFallback", MPTCP_MIB_SIMULTCONNFALLBACK), SNMP_MIB_ITEM("FallbackFailed", MPTCP_MIB_FALLBACKFAILED), SNMP_MIB_ITEM("WinProbe", MPTCP_MIB_WINPROBE), + SNMP_MIB_ITEM("BacklogDrop", MPTCP_MIB_BACKLOGDROP), + SNMP_MIB_ITEM("RcvPruned", MPTCP_MIB_RCVPRUNED), }; =20 /* mptcp_mib_alloc - allocate percpu mib counters diff --git a/net/mptcp/mib.h b/net/mptcp/mib.h index 812218b5ed2b..c84eb853d499 100644 --- a/net/mptcp/mib.h +++ b/net/mptcp/mib.h @@ -88,6 +88,8 @@ enum linux_mptcp_mib_field { MPTCP_MIB_SIMULTCONNFALLBACK, /* Simultaneous connect */ MPTCP_MIB_FALLBACKFAILED, /* Can't fallback due to msk status */ MPTCP_MIB_WINPROBE, /* MPTCP-level zero window probe */ + MPTCP_MIB_BACKLOGDROP, /* Backlog over memory limit */ + MPTCP_MIB_RCVPRUNED, /* Dropped due to memory constrains */ __MPTCP_MIB_MAX }; =20 diff --git a/net/mptcp/options.c b/net/mptcp/options.c index c664023d37ba..7b954ef78672 100644 --- a/net/mptcp/options.c +++ b/net/mptcp/options.c @@ -1127,8 +1127,30 @@ static bool add_addr_hmac_valid(struct mptcp_sock *m= sk, return hmac =3D=3D mp_opt->ahmac; } =20 -/* Return false in case of error (or subflow has been reset), - * else return true. +static bool mptcp_over_limit(struct sock *sk, struct sock *ssk, + const struct sk_buff *skb) +{ + struct mptcp_sock *msk =3D mptcp_sk(sk); + u64 mem =3D sk_rmem_alloc_get(sk); + + mem +=3D READ_ONCE(msk->backlog_len); + if (likely(mem <=3D READ_ONCE(sk->sk_rcvbuf))) + return false; + + /* Avoid silently dropping pure acks, fin or zero win probes. */ + if (TCP_SKB_CB(skb)->seq =3D=3D TCP_SKB_CB(skb)->end_seq || + TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN || + !after(TCP_SKB_CB(skb)->end_seq, tcp_sk(ssk)->rcv_nxt)) + return false; + + /* Dropped due to memory constraints, schedule an ack. */ + inet_csk(ssk)->icsk_ack.pending |=3D ICSK_ACK_NOMEM | ICSK_ACK_NOW; + inet_csk_schedule_ack(ssk); + return true; +} + +/* Return false when the caller must drop the packet, i.e. in case of erro= r, + * subflow has been reset, or over memory limits. */ bool mptcp_incoming_options(struct sock *sk, struct sk_buff *skb) { @@ -1154,7 +1176,7 @@ bool mptcp_incoming_options(struct sock *sk, struct s= k_buff *skb) =20 __mptcp_data_acked(subflow->conn); mptcp_data_unlock(subflow->conn); - return true; + return !mptcp_over_limit(subflow->conn, sk, skb); } =20 mptcp_get_options(skb, &mp_opt); diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index 72b1fa3ca71c..5f7d8340a3d9 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -381,6 +381,16 @@ static bool __mptcp_move_skb(struct sock *sk, struct s= k_buff *skb) =20 mptcp_borrow_fwdmem(sk, skb); =20 + /* Can't drop packets for fallback socket this late, or the stream + * will break. + */ + if (unlikely(sk_rmem_alloc_get(sk) > READ_ONCE(sk->sk_rcvbuf)) && + !__mptcp_check_fallback(msk)) { + MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_RCVPRUNED); + mptcp_drop(sk, skb); + return false; + } + if (MPTCP_SKB_CB(skb)->map_seq =3D=3D msk->ack_seq) { /* in sequence */ msk->bytes_received +=3D copy_len; @@ -675,6 +685,7 @@ static void __mptcp_add_backlog(struct sock *sk, struct sk_buff *tail =3D NULL; struct sock *ssk =3D skb->sk; bool fragstolen; + u64 limit; int delta; =20 if (unlikely(sk->sk_state =3D=3D TCP_CLOSE)) { @@ -682,6 +693,16 @@ static void __mptcp_add_backlog(struct sock *sk, return; } =20 + /* Similar additional allowance as plain TCP. */ + limit =3D READ_ONCE(sk->sk_rcvbuf); + limit +=3D (limit >> 1) + 64 * 1024; + limit =3D min_t(u64, limit, UINT_MAX); + if (msk->backlog_len > limit && !__mptcp_check_fallback(msk)) { + __MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_BACKLOGDROP); + kfree_skb_reason(skb, SKB_DROP_REASON_SOCKET_BACKLOG); + return; + } + /* Try to coalesce with the last skb in our backlog */ if (!list_empty(&msk->backlog_list)) tail =3D list_last_entry(&msk->backlog_list, struct sk_buff, list); @@ -753,7 +774,7 @@ static bool __mptcp_move_skbs_from_subflow(struct mptcp= _sock *msk, =20 mptcp_init_skb(ssk, skb, offset, len); =20 - if (own_msk && sk_rmem_alloc_get(sk) < sk->sk_rcvbuf) { + if (own_msk) { mptcp_subflow_lend_fwdmem(subflow, skb); ret |=3D __mptcp_move_skb(sk, skb); } else { @@ -2204,10 +2225,6 @@ static bool __mptcp_move_skbs(struct sock *sk, struc= t list_head *skbs, u32 *delt =20 *delta =3D 0; while (1) { - /* If the msk recvbuf is full stop, don't drop */ - if (sk_rmem_alloc_get(sk) > sk->sk_rcvbuf) - break; - prefetch(skb->next); list_del(&skb->list); *delta +=3D skb->truesize; @@ -2235,9 +2252,7 @@ static bool mptcp_can_spool_backlog(struct sock *sk, = struct list_head *skbs) DEBUG_NET_WARN_ON_ONCE(msk->backlog_unaccounted && sk->sk_socket && mem_cgroup_from_sk(sk)); =20 - /* Don't spool the backlog if the rcvbuf is full. */ - if (list_empty(&msk->backlog_list) || - sk_rmem_alloc_get(sk) > sk->sk_rcvbuf) + if (list_empty(&msk->backlog_list)) return false; =20 INIT_LIST_HEAD(skbs); --=20 2.53.0 From nobody Fri Jul 24 21:29:32 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 7D33243230E; Fri, 24 Jul 2026 14:02:45 +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=1784901769; cv=none; b=k/ol7djJMi4F7RgIdW4/gpDw65iwqzh7C3Rqnaf1eFmBoZYUYbVWnWMz2CujSUVMiV9HK6H5/geJOkN6VuLwUwOHwU/El1hg4OBR4trdUmC9gv1MuozLig5Fdn1JpYuviSgz3hocHs1XKfhEIWVNTbgZEqwwoZemroqz6IovUA0= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1784901769; c=relaxed/simple; bh=FGREC4W6Mh/CHWXFlthiI7UvopuXq/RVGFIwYGmO6W4=; h=From:Date:Subject:MIME-Version:Content-Type:Message-Id:References: In-Reply-To:To:Cc; b=WFdgHrDnq1m7IdGyOQ/c3ToTKl/Y3osBqpyyI51hIgsd69UL3XpHgxN0KGYB70W+qUzBE5kALcqVRFELhOClrk91K16ziQsurY6qf9E1K2s3WoPcSw+p53tdXb/vjmMTCf8sABDEDDF45sYedLfZAd/v3UeCx5ZcsnAMK/TvCcU= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=UcpZEdyh; 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="UcpZEdyh" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 725EB1F00A3D; Fri, 24 Jul 2026 14:02:43 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1784901765; bh=36lAnMkjB07VIYCCBLdzVx8RoT1sIGgW6HHKCR5eWMg=; h=From:Date:Subject:References:In-Reply-To:To:Cc; b=UcpZEdyhat91TWRDwkFKUzDUIIAdLPG5fJH6W3uXjKWjvgc3HdQmk0cQu+gvR6+HO N2c/fbL1wo7wGy5v4DBmKUrd82LgiliEi9vYOZOWpqOeW3PkWvZdd5QgI/rQYmSzIc 0rdt7sgBuy5uX1NU2q5yvFNrUmCHGqVUFnHsi9D4apGmDtbJVpaC/mL4p6v+7+z4U+ Mr11amWEc99fcqdwRGwZ++Z1G+3biN+krwslszSUwytQnwKI/73o5uP0rcAOs4BChC 4JeTCNx3nQDoBf/4vThIEFGibxEPX37wrt3YmcefZULykCs4iaYJ2/z6VN9d3M2grl igosbsBDwo9dA== From: "Matthieu Baerts (NGI0)" Date: Fri, 24 Jul 2026 16:02:16 +0200 Subject: [PATCH net-next 4/5] mptcp: enforce hard limit on backlog flushing Precedence: bulk X-Mailing-List: mptcp@lists.linux.dev List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Message-Id: <20260724-net-next-mptcp-oooq-pruning-v1-4-5dd4dec63a54@kernel.org> References: <20260724-net-next-mptcp-oooq-pruning-v1-0-5dd4dec63a54@kernel.org> In-Reply-To: <20260724-net-next-mptcp-oooq-pruning-v1-0-5dd4dec63a54@kernel.org> To: Mat Martineau , Geliang Tang , "David S. Miller" , Eric Dumazet , Jakub Kicinski , Paolo Abeni , Simon Horman Cc: netdev@vger.kernel.org, mptcp@lists.linux.dev, linux-kernel@vger.kernel.org, "Matthieu Baerts (NGI0)" X-Mailer: b4 0.15.2 X-Developer-Signature: v=1; a=openpgp-sha256; l=2868; i=matttbe@kernel.org; h=from:subject:message-id; bh=8oehzmC/Ja20Y/2KJD9sivd0mO9h5X8ufkiQiwQH0Fc=; b=owGbwMvMwCVWo/Th0Gd3rumMp9WSGLKSCyosFGPsFjp0+fAtvFGjs4mNedMBoR1Lu2dO3ri2N uC5ZLNxRykLgxgXg6yYIot0W2T+zOdVvCVefhYwc1iZQIYwcHEKwESOvGZkWNL3XuPrP+0QmXVf Vm755Nm6Yd7jvjYTM0fPUuPLpTdWazEy/AmJbJz+9HTUqZf+XHw3Vq3facpgP/1myKQc5tOzGLi jGAE= X-Developer-Key: i=matttbe@kernel.org; a=openpgp; fpr=E8CB85F76877057A6E27F77AF6B7824F4269A073 From: Paolo Abeni Currently a wild producer could keep the backlog flushing operation spinning for an unbound time. Since the previous patch, the amount of data present in the backlog is hard-limited. Move the backlog len update at the end of the flush loop to prevent it spinning forever. Also, no need to splice back the remaining skbs list into the backlog, as such list is always empty after each backlog processing loop. Signed-off-by: Paolo Abeni Reviewed-by: Matthieu Baerts (NGI0) Signed-off-by: Matthieu Baerts (NGI0) --- 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 5f7d8340a3d9..7f257419c5ea 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -2223,7 +2223,6 @@ static bool __mptcp_move_skbs(struct sock *sk, struct= list_head *skbs, u32 *delt struct mptcp_sock *msk =3D mptcp_sk(sk); bool moved =3D false; =20 - *delta =3D 0; while (1) { prefetch(skb->next); list_del(&skb->list); @@ -2260,20 +2259,12 @@ static bool mptcp_can_spool_backlog(struct sock *sk= , struct list_head *skbs) return true; } =20 -static void mptcp_backlog_spooled(struct sock *sk, u32 moved, - struct list_head *skbs) -{ - struct mptcp_sock *msk =3D mptcp_sk(sk); - - WRITE_ONCE(msk->backlog_len, msk->backlog_len - moved); - list_splice(skbs, &msk->backlog_list); -} - static bool mptcp_move_skbs(struct sock *sk) { + struct mptcp_sock *msk =3D mptcp_sk(sk); struct list_head skbs; bool enqueued =3D false; - u32 moved; + u32 moved =3D 0; =20 mptcp_data_lock(sk); while (mptcp_can_spool_backlog(sk, &skbs)) { @@ -2281,8 +2272,8 @@ static bool mptcp_move_skbs(struct sock *sk) enqueued |=3D __mptcp_move_skbs(sk, &skbs, &moved); =20 mptcp_data_lock(sk); - mptcp_backlog_spooled(sk, moved, &skbs); } + WRITE_ONCE(msk->backlog_len, msk->backlog_len - moved); mptcp_data_unlock(sk); =20 if (enqueued && mptcp_epollin_ready(sk)) @@ -3752,12 +3743,12 @@ static void mptcp_release_cb(struct sock *sk) __must_hold(&sk->sk_lock.slock) { struct mptcp_sock *msk =3D mptcp_sk(sk); + u32 moved =3D 0; =20 for (;;) { unsigned long flags =3D (msk->cb_flags & MPTCP_FLAGS_PROCESS_CTX_NEED); struct list_head join_list, skbs; bool spool_bl; - u32 moved; =20 spool_bl =3D mptcp_can_spool_backlog(sk, &skbs); if (!flags && !spool_bl) @@ -3790,9 +3781,9 @@ static void mptcp_release_cb(struct sock *sk) =20 cond_resched(); spin_lock_bh(&sk->sk_lock.slock); - if (spool_bl) - mptcp_backlog_spooled(sk, moved, &skbs); } + if (moved) + WRITE_ONCE(msk->backlog_len, msk->backlog_len - moved); =20 if (__test_and_clear_bit(MPTCP_CLEAN_UNA, &msk->cb_flags)) __mptcp_clean_una_wakeup(sk); --=20 2.53.0 From nobody Fri Jul 24 21:29:32 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 9BF0F43A81B; Fri, 24 Jul 2026 14:02:48 +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=1784901770; cv=none; b=VRS5VY6BPJ7Muoa1PpZQEEFsyooKDLPDtOVQ5GmPk3mSmjXOK3Q6U5UlO0Pf65qGmxhtl8jalg081Fa/f8U2ZJqz3zqJpm9l1I0Axj4An3m52aVB3UTC7dGFK8QPah1jKouLuTu6xvVHmQ06JWcQkwf1zP2FdKlutgvboFo5mkM= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1784901770; c=relaxed/simple; bh=S3MdV3UAR+ADbXfkUpujD9LpujITJkIrZfRUogdwTOI=; h=From:Date:Subject:MIME-Version:Content-Type:Message-Id:References: In-Reply-To:To:Cc; b=rG2FlA3VWKYywyOvxeV4DhhgCiNBrj15WmOAOBmZad4sYigbyt6FbmOxV61nvscg+ZqH0QawJIQkzjEDsiKowc4I7Dr9ukHXqxra5RgUpPt7TAZ+HSo8lPfn9PKYgfvT5vgPgqMGV4R9PD2fuAIJ9NElLkPiDBfmrS0MwfKGYww= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=gYZp48Gp; 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="gYZp48Gp" Received: by smtp.kernel.org (Postfix) with ESMTPSA id A9BA21F000E9; Fri, 24 Jul 2026 14:02:45 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1784901767; bh=eiIkZX9UKn62Wt3UHZGy8oEPX9InDuKj5kOafRz6D3M=; h=From:Date:Subject:References:In-Reply-To:To:Cc; b=gYZp48Gpn8wxKY98OEwH4fJVZL/ZImbDl5uOCR75i3jG1+w6tpJjsKv9omFE/qsfg qlVXjsCaNpgE0t5Xd4bojwezYpSRi63sVzMR1PpjKeiOKwk8+q7HButPdUUtiA7gTC heLbLjBUV1btBtC08rVvq1P3olYLxdQxDBCxU4jp+UDCe8/CURCRxI5E1PW5qvxEGF XNuyvtI/cPgJxoqqMwcR6cEGxDtAil/xufYr25hqOuPIMsOMk/Rg5e/Na9TQ1RGp/+ 6awyAluITZNaWBu9PcOn55eSIYCcD+DOMO3ENQh6HG6MeU2Zt8Bhj1Sl0z9upjcyOL WyeIpLtgeJnvA== From: "Matthieu Baerts (NGI0)" Date: Fri, 24 Jul 2026 16:02:17 +0200 Subject: [PATCH net-next 5/5] mptcp: implemented OoO queue pruning Precedence: bulk X-Mailing-List: mptcp@lists.linux.dev List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Message-Id: <20260724-net-next-mptcp-oooq-pruning-v1-5-5dd4dec63a54@kernel.org> References: <20260724-net-next-mptcp-oooq-pruning-v1-0-5dd4dec63a54@kernel.org> In-Reply-To: <20260724-net-next-mptcp-oooq-pruning-v1-0-5dd4dec63a54@kernel.org> To: Mat Martineau , Geliang Tang , "David S. Miller" , Eric Dumazet , Jakub Kicinski , Paolo Abeni , Simon Horman Cc: netdev@vger.kernel.org, mptcp@lists.linux.dev, linux-kernel@vger.kernel.org, "Matthieu Baerts (NGI0)" , Gang Yan X-Mailer: b4 0.15.2 X-Developer-Signature: v=1; a=openpgp-sha256; l=3583; i=matttbe@kernel.org; h=from:subject:message-id; bh=0TP0HmzvNCrhNepV2D684uiABQ8/iw2HmKqfNTiVn14=; b=owGbwMvMwCVWo/Th0Gd3rumMp9WSGLKSCyrLXn2vsjnL+X8Vx8zbfV+DjLU4WhbdMzs3bX0kY 0/l02OpHaUsDGJcDLJiiizSbZH5M59X8ZZ4+VnAzGFlAhnCwMUpABPRn8/wz1ak6GH+cv6mFh0H y6Bshheu4WzVNTHaIc3OfNsm7pWdychwo+XsVcMN/mdVOCrmRrebG8vMdfS7uSPs5LzQdWopp52 YAQ== X-Developer-Key: i=matttbe@kernel.org; a=openpgp; fpr=E8CB85F76877057A6E27F77AF6B7824F4269A073 From: Paolo Abeni When moving incoming skbs in the msk receive queue and the latter is above limits, prune it as needed quite alike what TCP is doing at the subflow level. The main difference relies in the stop condition: since MPTCP does not perform collapsing, it's better off dropping the bare minimum to fit the (newer) incoming packet. Signed-off-by: Paolo Abeni Tested-by: Gang Yan Reviewed-by: Matthieu Baerts (NGI0) Signed-off-by: Matthieu Baerts (NGI0) --- net/mptcp/mib.c | 1 + net/mptcp/mib.h | 1 + net/mptcp/protocol.c | 46 +++++++++++++++++++++++++++++++++++++++++++++- 3 files changed, 47 insertions(+), 1 deletion(-) diff --git a/net/mptcp/mib.c b/net/mptcp/mib.c index ef65e2df709f..d9bd4f4afcc0 100644 --- a/net/mptcp/mib.c +++ b/net/mptcp/mib.c @@ -87,6 +87,7 @@ static const struct snmp_mib mptcp_snmp_list[] =3D { SNMP_MIB_ITEM("WinProbe", MPTCP_MIB_WINPROBE), SNMP_MIB_ITEM("BacklogDrop", MPTCP_MIB_BACKLOGDROP), SNMP_MIB_ITEM("RcvPruned", MPTCP_MIB_RCVPRUNED), + SNMP_MIB_ITEM("OfoPruned", MPTCP_MIB_OFO_PRUNED), }; =20 /* mptcp_mib_alloc - allocate percpu mib counters diff --git a/net/mptcp/mib.h b/net/mptcp/mib.h index c84eb853d499..18f35f7e0a2d 100644 --- a/net/mptcp/mib.h +++ b/net/mptcp/mib.h @@ -90,6 +90,7 @@ enum linux_mptcp_mib_field { MPTCP_MIB_WINPROBE, /* MPTCP-level zero window probe */ MPTCP_MIB_BACKLOGDROP, /* Backlog over memory limit */ MPTCP_MIB_RCVPRUNED, /* Dropped due to memory constrains */ + MPTCP_MIB_OFO_PRUNED, /* MPTCP-level OoO queue pruned */ __MPTCP_MIB_MAX }; =20 diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index 7f257419c5ea..5cf44eab76f1 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -373,6 +373,49 @@ static void mptcp_init_skb(struct sock *ssk, struct sk= _buff *skb, int offset, skb_dst_drop(skb); } =20 +/* "Inspired" from the TCP version; main difference: stop as soon as the M= PTCP + * socket is under memory limit. + */ +static bool mptcp_prune_ofo_queue(struct sock *sk, u64 seq) +{ + struct mptcp_sock *msk =3D mptcp_sk(sk); + struct rb_node *node, *prev; + bool pruned =3D false; + u64 mem; + + if (RB_EMPTY_ROOT(&msk->out_of_order_queue)) + goto out; + + node =3D &msk->ooo_last_skb->rbnode; + + do { + struct sk_buff *skb =3D rb_to_skb(node); + + /* Stop pruning if the incoming skb would land in OoO tail. */ + if (after64(seq, MPTCP_SKB_CB(skb)->map_seq)) + break; + + pruned =3D true; + prev =3D rb_prev(node); + rb_erase(node, &msk->out_of_order_queue); + mptcp_drop(sk, skb); + msk->ooo_last_skb =3D rb_to_skb(prev); + + mem =3D (unsigned int)sk_rmem_alloc_get(sk); + if (mem <=3D sk->sk_rcvbuf) + break; + + node =3D prev; + } while (node); + + if (pruned) + MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_OFO_PRUNED); + +out: + mem =3D (unsigned int)sk_rmem_alloc_get(sk); + return mem <=3D sk->sk_rcvbuf; +} + 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; @@ -385,7 +428,8 @@ static bool __mptcp_move_skb(struct sock *sk, struct sk= _buff *skb) * will break. */ if (unlikely(sk_rmem_alloc_get(sk) > READ_ONCE(sk->sk_rcvbuf)) && - !__mptcp_check_fallback(msk)) { + !__mptcp_check_fallback(msk) && + !mptcp_prune_ofo_queue(sk, MPTCP_SKB_CB(skb)->map_seq)) { MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_RCVPRUNED); mptcp_drop(sk, skb); return false; --=20 2.53.0