From: Dmytro Shytyi <dmytro@shytyi.net>
To: Paolo Abeni <pabeni@redhat.com>, mptcp@lists.linux.dev
Subject: Re: [RFC PATCH mptcp-next v7 07/11] mptfo variables for msk, options. Fix loop retrans
Date: Tue, 20 Sep 2022 15:40:36 +0200 [thread overview]
Message-ID: <dcedb781-507b-700d-68b2-71dfb80b6fa2@shytyi.net> (raw)
In-Reply-To: <be427fe3457391b7b50af213c255d8fd1f3703c7.camel@redhat.com>
This is refactored in v8.
Best regards,
Dmytro
On 9/19/2022 1:11 PM, Paolo Abeni wrote:
> On Sun, 2022-09-18 at 00:28 +0200, Dmytro Shytyi wrote:
>> Introduce mptfo variables for msk and options.
>> Also fixe the infinite retransmissions in the end of second session.
>> Suggestion @palbeni (SEP 1) during the meting to 'look at ack'
>>
>> Signed-off-by: Dmytro Shytyi <dmytro@shytyi.net>
>> ---
>> net/mptcp/fastopen.c | 16 ++++++++++++++++
>> net/mptcp/options.c | 3 +++
>> net/mptcp/protocol.h | 6 +++++-
>> 3 files changed, 24 insertions(+), 1 deletion(-)
>>
>> diff --git a/net/mptcp/fastopen.c b/net/mptcp/fastopen.c
>> index 436e773d798a..149a4b1d3dac 100644
>> --- a/net/mptcp/fastopen.c
>> +++ b/net/mptcp/fastopen.c
>> @@ -91,3 +91,19 @@ int mptcp_setsockopt_sol_tcp_fastopen(struct mptcp_sock *msk, sockptr_t optval,
>>
>> return ret;
>> }
>> +
>> +void mptcp_treat_hshake_ack_fastopen(struct mptcp_sock *msk, struct mptcp_subflow_context *subflow,
>> + struct mptcp_options_received mp_opt)
>> +{
>> + u64 ack_seq;
>> +
>> + if (mp_opt.suboptions & OPTIONS_MPTCP_MPC && mp_opt.is_mptfo && msk->is_mptfo) {
>> + msk->can_ack = true;
>> + msk->remote_key = mp_opt.sndr_key;
>> + mptcp_crypto_key_sha(msk->remote_key, NULL, &ack_seq);
>> + ack_seq++;
>> + WRITE_ONCE(msk->ack_seq, ack_seq);
>> + pr_debug("ack_seq=%llu sndr_key=%llu", msk->ack_seq, mp_opt.sndr_key);
>> + atomic64_set(&msk->rcv_wnd_sent, ack_seq);
> It's not clear to me why the above is needed. According to the RFC we
> should not do additional MPTCP-level key management for TFO' sake ?!?
>
>> + }
>> +}
>> diff --git a/net/mptcp/options.c b/net/mptcp/options.c
>> index 30d289044e71..185b069e57f4 100644
>> --- a/net/mptcp/options.c
>> +++ b/net/mptcp/options.c
>> @@ -91,6 +91,7 @@ static void mptcp_parse_option(const struct sk_buff *skb,
>> ptr += 8;
>> }
>> if (opsize >= TCPOLEN_MPTCP_MPC_ACK) {
>> + mp_opt->is_mptfo = 1;
> This looks incorrect. Plain MPTCP MP_CAPABLE handshake will be marked
> as TFO regardless of the actual TCP option exchange.
>
> It looks like the only already available hook to catch the 'incoming
> fast open' info would be implementing custom mptcp subflow {v4,v6}
> send_synack and checking the 'foc' pointer.
>
> Otherwise you can add TCPOPT_FASTOPEN parsing capability to
> mptcp_parse_option(), but that would be more duplicate code, I think it
> would be better to avoid that if possible.
>
>> mp_opt->rcvr_key = get_unaligned_be64(ptr);
>> ptr += 8;
>> }
>> @@ -1124,6 +1125,8 @@ bool mptcp_incoming_options(struct sock *sk, struct sk_buff *skb)
>> return sk->sk_state != TCP_CLOSE;
>>
>> if (unlikely(mp_opt.suboptions != OPTION_MPTCP_DSS)) {
>> + mptcp_treat_hshake_ack_fastopen(msk, subflow, mp_opt);
>> +
>> if ((mp_opt.suboptions & OPTION_MPTCP_FASTCLOSE) &&
>> msk->local_key == mp_opt.rcvr_key) {
>> WRITE_ONCE(msk->rcv_fastclose, true);
>> diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h
>> index 8caaeeedb9da..b90279c734ae 100644
>> --- a/net/mptcp/protocol.h
>> +++ b/net/mptcp/protocol.h
>> @@ -155,7 +155,8 @@ struct mptcp_options_received {
>> echo:1,
>> backup:1,
>> deny_join_id0:1,
>> - __unused:2;
>> + is_mptfo:1,
>> + __unused:1;
>> u8 join_id;
>> u64 thmac;
>> u8 hmac[MPTCPOPT_HMAC_LEN];
>> @@ -282,6 +283,7 @@ struct mptcp_sock {
>> bool use_64bit_ack; /* Set when we received a 64-bit DSN */
>> bool csum_enabled;
>> bool allow_infinite_fallback;
>> + bool is_mptfo;
>> u8 mpc_endpoint_id;
>> u8 recvmsg_inq:1,
>> cork:1,
>> @@ -845,6 +847,8 @@ int mptcp_sendmsg_fastopen(struct sock *sk, struct msghdr *msg,
>> size_t *copied);
>> int mptcp_setsockopt_sol_tcp_fastopen(struct mptcp_sock *msk, sockptr_t optval,
>> unsigned int optlen);
>> +void mptcp_treat_hshake_ack_fastopen(struct mptcp_sock *msk, struct mptcp_subflow_context *subflow,
>> + struct mptcp_options_received mp_opt);
>> // Fast Open Mechanism functions end
>>
>> static inline bool mptcp_pm_should_add_signal(struct mptcp_sock *msk)
next prev parent reply other threads:[~2022-09-20 13:40 UTC|newest]
Thread overview: 36+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-09-17 22:28 [RFC PATCH mptcp-next v7 00/11] mptcp: Fast Open Mechanism Dmytro Shytyi
2022-09-17 22:28 ` [RFC PATCH mptcp-next v7 01/11] Add separate fastopen.c file Dmytro Shytyi
2022-09-19 10:21 ` Paolo Abeni
2022-09-20 13:29 ` Dmytro Shytyi
2022-09-17 22:28 ` [RFC PATCH mptcp-next v7 02/11] add mptcp_stream_connect to protocol.h Dmytro Shytyi
2022-09-19 10:19 ` Paolo Abeni
2022-09-20 13:30 ` Dmytro Shytyi
2022-09-17 22:28 ` [RFC PATCH mptcp-next v7 03/11] Initiator: MSG_FASTOPEN sendto(). request cookie Dmytro Shytyi
2022-09-19 10:35 ` Paolo Abeni
2022-09-19 10:44 ` Paolo Abeni
2022-09-19 11:22 ` Matthieu Baerts
2022-09-20 13:32 ` Dmytro Shytyi
2022-09-17 22:28 ` [RFC PATCH mptcp-next v7 04/11] rfree(), rmem_uncharge() prototypes to protocol.h Dmytro Shytyi
2022-09-17 22:28 ` [RFC PATCH mptcp-next v7 05/11] Initiator: add locks() to mptcp_sendmsg_fastopen Dmytro Shytyi
2022-09-19 10:46 ` Paolo Abeni
2022-09-20 13:33 ` Dmytro Shytyi
2022-09-17 22:28 ` [RFC PATCH mptcp-next v7 06/11] add mptcp_setsockopt_fastopen Dmytro Shytyi
2022-09-19 10:48 ` Paolo Abeni
2022-09-20 13:33 ` Dmytro Shytyi
2022-09-20 9:37 ` Matthieu Baerts
2022-09-20 13:34 ` Dmytro Shytyi
2022-09-17 22:28 ` [RFC PATCH mptcp-next v7 07/11] mptfo variables for msk, options. Fix loop retrans Dmytro Shytyi
2022-09-19 11:11 ` Paolo Abeni
2022-09-19 11:26 ` Paolo Abeni
2022-09-20 16:32 ` Christoph Paasch
2022-09-20 13:40 ` Dmytro Shytyi [this message]
2022-09-17 22:28 ` [RFC PATCH mptcp-next v7 08/11] Fix unxpctd val of subflow->map_seq(dscrd packet) Dmytro Shytyi
2022-09-19 11:27 ` Paolo Abeni
2022-09-19 11:31 ` Matthieu Baerts
2022-09-19 14:02 ` Paolo Abeni
2022-09-17 22:28 ` [RFC PATCH mptcp-next v7 09/11] Listener: Add received skb to msk Dmytro Shytyi
2022-09-17 22:28 ` [RFC PATCH mptcp-next v7 10/11] mptcp_fastopen_add_skb() helpers (skb to msk) Dmytro Shytyi
2022-09-19 14:06 ` Paolo Abeni
2022-09-20 13:36 ` Dmytro Shytyi
2022-09-17 22:28 ` [RFC PATCH mptcp-next v7 11/11] selftests: mptfo initiator/listener Dmytro Shytyi
2022-09-19 14:11 ` Paolo Abeni
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=dcedb781-507b-700d-68b2-71dfb80b6fa2@shytyi.net \
--to=dmytro@shytyi.net \
--cc=mptcp@lists.linux.dev \
--cc=pabeni@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).