From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mga04.intel.com (mga04.intel.com [192.55.52.120]) (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 6D37E70 for ; Wed, 7 Jul 2021 23:07:43 +0000 (UTC) X-IronPort-AV: E=McAfee;i="6200,9189,10037"; a="207581022" X-IronPort-AV: E=Sophos;i="5.84,222,1620716400"; d="scan'208";a="207581022" Received: from fmsmga005.fm.intel.com ([10.253.24.32]) by fmsmga104.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 07 Jul 2021 16:07:41 -0700 X-IronPort-AV: E=Sophos;i="5.84,222,1620716400"; d="scan'208";a="648094543" Received: from wilesamy-mobl.amr.corp.intel.com ([10.212.141.74]) by fmsmga005-auth.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 07 Jul 2021 16:07:41 -0700 Date: Wed, 7 Jul 2021 16:07:40 -0700 (PDT) From: Mat Martineau To: Geliang Tang cc: mptcp@lists.linux.dev Subject: Re: [MPTCP][PATCH v3 mptcp-next 1/8] mptcp: MP_FAIL suboption sending In-Reply-To: <3cc57b70ec147b4c77610737241aeadf4150f323.1624854005.git.geliangtang@gmail.com> Message-ID: References: <3cc57b70ec147b4c77610737241aeadf4150f323.1624854005.git.geliangtang@gmail.com> Precedence: bulk X-Mailing-List: mptcp@lists.linux.dev List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII; format=flowed On Mon, 28 Jun 2021, Geliang Tang wrote: > This patch added the MP_FAIL suboption sending support. > > Add a new flag named send_mp_fail in struct mptcp_subflow_context. If > this flag is set, send out MP_FAIL suboption. > > Add a new member fail_seq in struct mptcp_out_options to save the data > sequence number to put into the MP_FAIL suboption. > > Signed-off-by: Geliang Tang > --- > include/net/mptcp.h | 1 + > net/mptcp/options.c | 54 ++++++++++++++++++++++++++++++++++++++++++-- > net/mptcp/protocol.h | 4 ++++ > 3 files changed, 57 insertions(+), 2 deletions(-) > > diff --git a/include/net/mptcp.h b/include/net/mptcp.h > index cb580b06152f..f48d3b5a3fd4 100644 > --- a/include/net/mptcp.h > +++ b/include/net/mptcp.h > @@ -72,6 +72,7 @@ struct mptcp_out_options { > u32 nonce; > u64 thmac; > u32 token; > + u64 fail_seq; > u8 hmac[20]; > struct mptcp_ext ext_copy; > #endif > diff --git a/net/mptcp/options.c b/net/mptcp/options.c > index b5850afea343..b78defe1aed9 100644 > --- a/net/mptcp/options.c > +++ b/net/mptcp/options.c > @@ -771,6 +771,28 @@ static noinline void mptcp_established_options_rst(struct sock *sk, struct sk_bu > opts->reset_reason = subflow->reset_reason; > } > > +static bool mptcp_established_options_mp_fail(struct sock *sk, > + unsigned int *size, > + unsigned int remaining, > + struct mptcp_out_options *opts) > +{ > + struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(sk); > + > + if (!subflow->send_mp_fail) > + return false; > + > + if (remaining < TCPOLEN_MPTCP_FAIL) > + return false; > + > + *size = TCPOLEN_MPTCP_FAIL; > + opts->suboptions |= OPTION_MPTCP_FAIL; > + opts->fail_seq = subflow->fail_seq; > + > + pr_debug("MP_FAIL fail_seq=%llu", opts->fail_seq); > + > + return true; > +} > + > bool mptcp_established_options(struct sock *sk, struct sk_buff *skb, > unsigned int *size, unsigned int remaining, > struct mptcp_out_options *opts) > @@ -787,15 +809,29 @@ bool mptcp_established_options(struct sock *sk, struct sk_buff *skb, > return false; > > if (unlikely(skb && TCP_SKB_CB(skb)->tcp_flags & TCPHDR_RST)) { > - mptcp_established_options_rst(sk, skb, size, remaining, opts); > + if (mptcp_established_options_mp_fail(sk, &opt_size, remaining, opts)) { > + *size += opt_size; > + remaining -= opt_size; > + } > + mptcp_established_options_rst(sk, skb, &opt_size, remaining, opts); > + *size += opt_size; > + remaining -= opt_size; If there isn't enough space for MP_TCPRST, opt_size can still have the value set by mptcp_established_options_mp_fail() and the previous two lines would add or subtract incorrectly. mptcp_established_options_rst() could return bool to make the size adjustments conditional like the new mp_fail lines above. - Mat > return true; > } > > snd_data_fin = mptcp_data_fin_enabled(msk); > if (mptcp_established_options_mp(sk, skb, snd_data_fin, &opt_size, remaining, opts)) > ret = true; > - else if (mptcp_established_options_dss(sk, skb, snd_data_fin, &opt_size, remaining, opts)) > + else if (mptcp_established_options_dss(sk, skb, snd_data_fin, &opt_size, remaining, opts)) { > ret = true; > + if (opts->ext_copy.use_ack) { > + if (mptcp_established_options_mp_fail(sk, &opt_size, remaining, opts)) { > + *size += opt_size; > + remaining -= opt_size; > + ret = true; > + } > + } > + } > > /* we reserved enough space for the above options, and exceeding the > * TCP option space would be fatal > @@ -1334,6 +1370,20 @@ void mptcp_write_options(__be32 *ptr, const struct tcp_sock *tp, > opts->backup, TCPOPT_NOP); > } > > + if (OPTION_MPTCP_FAIL & opts->suboptions) { > + const struct sock *ssk = (const struct sock *)tp; > + struct mptcp_subflow_context *subflow; > + > + subflow = mptcp_subflow_ctx(ssk); > + subflow->send_mp_fail = 0; > + > + *ptr++ = mptcp_option(MPTCPOPT_MP_FAIL, > + TCPOLEN_MPTCP_FAIL, > + 0, 0); > + put_unaligned_be64(opts->fail_seq, ptr); > + ptr += 2; > + } > + > if (OPTION_MPTCP_MPJ_SYN & opts->suboptions) { > *ptr++ = mptcp_option(MPTCPOPT_MP_JOIN, > TCPOLEN_MPTCP_MPJ_SYN, > diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h > index 426ed80fe72f..007af5e4ba3d 100644 > --- a/net/mptcp/protocol.h > +++ b/net/mptcp/protocol.h > @@ -26,6 +26,7 @@ > #define OPTION_MPTCP_FASTCLOSE BIT(8) > #define OPTION_MPTCP_PRIO BIT(9) > #define OPTION_MPTCP_RST BIT(10) > +#define OPTION_MPTCP_FAIL BIT(11) > > /* MPTCP option subtypes */ > #define MPTCPOPT_MP_CAPABLE 0 > @@ -67,6 +68,7 @@ > #define TCPOLEN_MPTCP_PRIO_ALIGN 4 > #define TCPOLEN_MPTCP_FASTCLOSE 12 > #define TCPOLEN_MPTCP_RST 4 > +#define TCPOLEN_MPTCP_FAIL 12 > > #define TCPOLEN_MPTCP_MPC_ACK_DATA_CSUM (TCPOLEN_MPTCP_DSS_CHECKSUM + TCPOLEN_MPTCP_MPC_ACK_DATA) > > @@ -417,6 +419,7 @@ struct mptcp_subflow_context { > mpc_map : 1, > backup : 1, > send_mp_prio : 1, > + send_mp_fail : 1, > rx_eof : 1, > can_ack : 1, /* only after processing the remote a key */ > disposable : 1; /* ctx can be free at ulp release time */ > @@ -431,6 +434,7 @@ struct mptcp_subflow_context { > u8 reset_seen:1; > u8 reset_transient:1; > u8 reset_reason:4; > + u64 fail_seq; > > long delegated_status; > struct list_head delegated_node; /* link into delegated_action, protected by local BH */ > -- > 2.31.1 > > > -- Mat Martineau Intel