From: Jianguo Wu <wujianguo106@163.com>
To: Mat Martineau <mathew.j.martineau@linux.intel.com>
Cc: mptcp@lists.linux.dev, pabeni@redhat.com, fw@strlen.de
Subject: Re: [PATCH v5 3/4] mptcp: fix syncookie process if mptcp can not_accept new subflow
Date: Mon, 21 Jun 2021 14:24:01 +0800 [thread overview]
Message-ID: <cebfba8a-7576-3b9a-2017-66ab276d1f06@163.com> (raw)
In-Reply-To: <3cdb43ff-a2c2-6599-4e8-3cb569e1a26a@linux.intel.com>
Hi Mat,
On 2021/6/19 7:19, Mat Martineau wrote:
> On Wed, 16 Jun 2021, wujianguo106@163.com wrote:
>
>> From: Jianguo Wu <wujianguo@chinatelecom.cn>
>>
>> Lots of "TCP: tcp_fin: Impossible, sk->sk_state=7" in client side
>> when doing stress testing.
>
> Hi Jianguo -
>
> Like patch 1, do the existing syncookie selftests trigger this sometimes?
>
> Or are there selftests that could be added to this series, or packetdrill tests, to do such syncookie stress tests?
>
This is triggered by the same test in patch 1,
I will try selftests too, or add selftest if necessary.
> Thanks,
>
> Mat
>
>
>>
>> There are at least two cases may trigger this warning:
>> 1.mptcp is in syncookie, and server recv MP_JOIN SYN request,
>> in subflow_check_req(), the mptcp_can_accept_new_subflow()
>> return false, so subflow_init_req_cookie_join_save() isn't
>> called, i.e. not store the data present in the MP_JOIN syn
>> request and the random nonce in hash table - join_entries[],
>> but still send synack. When recv 3rd-ack,
>> mptcp_token_join_cookie_init_state() will return false, and
>> 3rd-ack is dropped, then if mptcp conn is closed by client,
>> client will send a DATA_FIN and a MPTCP FIN, the DATA_FIN
>> doesn't have MP_CAPABLE or MP_JOIN,
>> so mptcp_subflow_init_cookie_req() will return 0, and pass
>> the cookie check, MP_JOIN request is fallback to normal TCP.
>> Server will send a TCP FIN if closed, in client side,
>> when process TCP FIN, it will do reset, the code path is:
>> tcp_data_queue()->mptcp_incoming_options()
>> ->check_fully_established()->mptcp_subflow_reset().
>> mptcp_subflow_reset() will set sock state to TCP_CLOSE,
>> so tcp_fin will hit TCP_CLOSE, and print the warning.
>> 2.mptcp is in syncookie, and server recv 3rd-ack, in
>> mptcp_subflow_init_cookie_req(), mptcp_can_accept_new_subflow()
>> return false, and subflow_req->mp_join is not set to 1,
>> so in subflow_syn_recv_sock() will not reset the MP_JOIN
>> subflow, but fallback to normal TCP, and then the same thing
>> happens when server will send a TCP FIN if closed.
>>
>> For case1, subflow_check_req() return -EPERM,
>> then tcp_conn_request() will drop MP_JOIN SYN.
>>
>> For case2, let subflow_syn_recv_sock() call mptcp_can_accept_new_subflow(),
>> and do fatal fallback, send reset.
>>
>> Fixes: 9466a1ccebbe("mptcp: enable JOIN requests even if cookies are in use")
>> Signed-off-by: Jianguo Wu <wujianguo@chinatelecom.cn>
>> ---
>> net/mptcp/subflow.c | 6 +++---
>> 1 file changed, 3 insertions(+), 3 deletions(-)
>>
>> diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c
>> index 75ed530706c0..6d98e19a20aa 100644
>> --- a/net/mptcp/subflow.c
>> +++ b/net/mptcp/subflow.c
>> @@ -224,6 +224,8 @@ static int subflow_check_req(struct request_sock *req,
>> if (unlikely(req->syncookie)) {
>> if (mptcp_can_accept_new_subflow(subflow_req->msk))
>> subflow_init_req_cookie_join_save(subflow_req, skb);
>> + else
>> + return -EPERM;
>> }
>>
>> pr_debug("token=%u, remote_nonce=%u msk=%p", subflow_req->token,
>> @@ -263,9 +265,7 @@ int mptcp_subflow_init_cookie_req(struct request_sock *req,
>> if (!mptcp_token_join_cookie_init_state(subflow_req, skb))
>> return -EINVAL;
>>
>> - if (mptcp_can_accept_new_subflow(subflow_req->msk))
>> - subflow_req->mp_join = 1;
>> -
>> + subflow_req->mp_join = 1;
>> subflow_req->ssn_offset = TCP_SKB_CB(skb)->seq - 1;
>> }
>>
>> --
>> 1.8.3.1
>>
>>
>>
>
> --
> Mat Martineau
> Intel
next prev parent reply other threads:[~2021-06-21 6:24 UTC|newest]
Thread overview: 24+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-06-16 10:49 [PATCH v5 0/4] Fix some mptcp syncookie process bugs wujianguo106
2021-06-16 10:49 ` [PATCH v5 1/4] mptcp: fix warning in __skb_flow_dissect() when do syn cookie for subflow join wujianguo106
2021-06-18 22:40 ` Mat Martineau
2021-06-21 6:14 ` Jianguo Wu
2021-06-21 10:09 ` Jianguo Wu
2021-06-22 23:38 ` Mat Martineau
2021-06-16 10:49 ` [PATCH v5 2/4] mptcp: remove redundant req destruct in subflow_check_req() wujianguo106
2021-06-16 10:49 ` [PATCH v5 3/4] mptcp: fix syncookie process if mptcp can not_accept new subflow wujianguo106
2021-06-18 23:19 ` Mat Martineau
2021-06-21 6:24 ` Jianguo Wu [this message]
2021-06-24 2:08 ` Jianguo Wu
2021-06-24 22:36 ` Mat Martineau
2021-06-16 10:49 ` [PATCH v5 4/4] mptcp: avoid processing packet if a subflow reset wujianguo106
2021-06-19 0:19 ` Mat Martineau
2021-06-21 6:35 ` Jianguo Wu
2021-06-23 0:00 ` Mat Martineau
2021-06-23 9:48 ` Paolo Abeni
2021-06-24 1:57 ` Jianguo Wu
2021-06-24 10:02 ` Paolo Abeni
2021-06-24 22:38 ` Mat Martineau
2021-06-25 0:51 ` Jianguo Wu
2021-06-19 0:26 ` [PATCH v5 0/4] Fix some mptcp syncookie process bugs Mat Martineau
2021-06-21 6:39 ` Jianguo Wu
2021-06-23 0:07 ` Mat Martineau
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=cebfba8a-7576-3b9a-2017-66ab276d1f06@163.com \
--to=wujianguo106@163.com \
--cc=fw@strlen.de \
--cc=mathew.j.martineau@linux.intel.com \
--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).