From: sashiko-bot@kernel.org
To: "Geliang Tang" <geliang@kernel.org>, mptcp@lists.linux.dev
Subject: Re: [PATCH mptcp-next v28 09/11] mptcp: drop data after DATA_FIN before skb init
Date: Mon, 16 Nov 2026 07:37:37 +0000 [thread overview]
Message-ID: <sashiko-outbox-216443@kernel.org> (raw)
In-Reply-To: <0b6321dafceaf638b198d43b6918ed87a58cf5f1.1794810809.git.tanggeliang@kylinos.cn>
Thank you for your contribution! Sashiko AI review found 4 potential
issues to consider:
- [Critical] mptcp: unadvanced tp->copied_seq when dropping data after
DATA_FIN
- [High] mptcp: premature data drop on local RCV_SHUTDOWN in
mptcp_seq_after_data_fin()
- [High] mptcp: post-DATA_FIN OOO, backlog, and straddling skbs bypass
filtering
- [High] mptcp: rcv_data_fin clearing in mptcp_check_data_fin() allows
DATA_FIN bypass
--
commit 16698b96108faa15e512d4467eb72e9ebc40cd5c
Author: Geliang Tang <geliang@kernel.org>
mptcp: drop data after DATA_FIN before skb init
This commit adds mptcp_seq_after_data_fin() and calls it in
__mptcp_move_skbs_from_subflow() before mptcp_init_skb() to drop incoming
subflow data at or after the DATA_FIN sequence number.
> diff --git a/net/mptcp/options.c b/net/mptcp/options.c
> index 71579238a951c..425a7fe94e0c4 100644
> --- a/net/mptcp/options.c
> +++ b/net/mptcp/options.c
> @@ -1177,6 +1177,7 @@ bool mptcp_update_rcv_data_fin(struct mptcp_sock *msk, u64 data_fin_seq, bool us
>
> WRITE_ONCE(msk->rcv_data_fin_seq,
> mptcp_expand_seq(READ_ONCE(msk->ack_seq), data_fin_seq, use_64bit));
> + smp_wmb(); /* ensure rcv_data_fin_seq is visible before rcv_data_fin */
> WRITE_ONCE(msk->rcv_data_fin, 1);
>
> return true;
[Severity: High]
Can data at or after the initial DATA_FIN bypass mptcp_seq_after_data_fin()
once mptcp_check_data_fin() clears msk->rcv_data_fin?
net/mptcp/protocol.c:mptcp_check_data_fin() {
...
if (mptcp_pending_data_fin(sk, &rcv_data_fin_seq)) {
WRITE_ONCE(msk->ack_seq, msk->ack_seq + 1);
WRITE_ONCE(msk->rcv_data_fin, 0);
WRITE_ONCE(sk->sk_shutdown, sk->sk_shutdown | RCV_SHUTDOWN);
...
}
mptcp_update_rcv_data_fin() only checks READ_ONCE(msk->rcv_data_fin), so
after mptcp_check_data_fin() clears msk->rcv_data_fin to 0, a peer sending
a second DATA_FIN with a higher data_fin_seq overwrites
msk->rcv_data_fin_seq.
In addition, mptcp_check_data_fin() clears msk->rcv_data_fin before setting
RCV_SHUTDOWN in sk->sk_shutdown without mptcp_data_lock(sk) or a write
barrier, so a concurrent mptcp_seq_after_data_fin() can observe both as 0:
CPU0 CPU1
mptcp_check_data_fin()
WRITE_ONCE(msk->ack_seq, ...)
WRITE_ONCE(msk->rcv_data_fin, 0)
__mptcp_move_skbs_from_subflow()
mptcp_seq_after_data_fin()
READ_ONCE(msk->rcv_data_fin) == 0
(sk->sk_shutdown & RCV_SHUTDOWN) == 0
return false <-- bypasses check
WRITE_ONCE(sk->sk_shutdown,
... | RCV_SHUTDOWN)
In either case, data at or after the original DATA_FIN is queued into
sk->sk_receive_queue after msk->ack_seq has advanced past msk->copied_seq,
triggering offset wraparound in __mptcp_recvmsg_mskq() and wrapping
mptcp_data_avail() to U64_MAX.
> diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c
> index ddc98497f0eb0..78ca580c6d9c5 100644
> --- a/net/mptcp/protocol.c
> +++ b/net/mptcp/protocol.c
> @@ -811,6 +811,25 @@ static void __mptcp_add_backlog(struct sock *sk,
> msk->backlog_unaccounted += delta;
> }
>
> +static bool mptcp_seq_after_data_fin(struct mptcp_sock *msk,
> + struct sock *ssk, u32 len)
> +{
> + struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(ssk);
> + struct sock *sk = (struct sock *)msk;
> +
> + if (READ_ONCE(msk->rcv_data_fin) || (sk->sk_shutdown & RCV_SHUTDOWN)) {
> + u64 seq = mptcp_subflow_get_mapped_dsn(subflow);
> + u64 fin_seq;
> +
> + smp_rmb(); /* ensure we see the updated rcv_data_fin_seq */
> + fin_seq = READ_ONCE(msk->rcv_data_fin_seq);
[Severity: High]
Does checking (sk->sk_shutdown & RCV_SHUTDOWN) drop valid in-sequence peer
data when RCV_SHUTDOWN was set locally before any DATA_FIN arrived?
__mptcp_close() and inet_shutdown() both set RCV_SHUTDOWN on
sk->sk_shutdown while msk->rcv_data_fin_seq is still 0 (or holds a stale
value from a previous connection after mptcp_disconnect()). When
fin_seq == 0, !before64(seq, 0) is (__s64)seq >= 0, which is true for half
of all random initial 64-bit DSNs. Dropping the peer's pre-DATA_FIN data
prevents msk->ack_seq from reaching the peer's DATA_FIN sequence number, so
the connection hangs in TCP_FIN_WAIT1 or TCP_FIN_WAIT2 until the close
timer expires.
> + if (!before64(seq, fin_seq) ||
> + !before64(seq + len, fin_seq + 1))
> + return true;
> + }
> + return false;
> +}
[Severity: High]
What happens when post-DATA_FIN segments are already queued in
msk->out_of_order_queue or msk->backlog_list before DATA_FIN arrives, or
when a single skb straddles fin_seq?
Segments that arrived before the packet carrying DATA_FIN have already
passed __mptcp_move_skbs_from_subflow(). When DATA_FIN later arrives,
__mptcp_ofo_queue() and __mptcp_move_skbs() move those post-DATA_FIN
segments into sk->sk_receive_queue without checking rcv_data_fin_seq,
causing the same offset underflow in __mptcp_recvmsg_mskq(),
mptcp_data_avail() wraparound to U64_MAX, and skipped DATA_FIN
acknowledgment.
Conversely, if a single segment straddles fin_seq (seq < fin_seq and
seq + len > fin_seq), !before64(seq + len, fin_seq + 1) drops the entire
skb including the valid [seq, fin_seq) prefix, preventing msk->ack_seq from
reaching fin_seq.
> +
> static bool __mptcp_move_skbs_from_subflow(struct mptcp_sock *msk,
> struct sock *ssk, bool own_msk)
> {
> @@ -826,7 +845,8 @@ static bool __mptcp_move_skbs_from_subflow(struct mptcp_sock *msk,
> u32 map_remaining, offset;
> u32 seq = tp->copied_seq;
> struct sk_buff *skb;
> - bool fin;
> + bool fin, data_fin;
> + size_t len;
>
> /* try to move as much data as available */
> map_remaining = subflow->map_data_len -
> @@ -850,9 +870,9 @@ static bool __mptcp_move_skbs_from_subflow(struct mptcp_sock *msk,
> if (fin)
> seq++;
>
> - if (offset < skb->len) {
> - size_t len = skb->len - offset;
> -
> + len = skb->len - offset;
> + data_fin = mptcp_seq_after_data_fin(msk, ssk, len);
> + if (len > 0 && !data_fin) {
> mptcp_init_skb(ssk, skb, offset, len);
>
> if (own_msk) {
> @@ -870,7 +890,7 @@ static bool __mptcp_move_skbs_from_subflow(struct mptcp_sock *msk,
> } else {
> sk_eat_skb(ssk, skb);
>
> - if (unlikely(!fin)) {
> + if (len == 0 && unlikely(!fin)) {
> DEBUG_NET_WARN_ON_ONCE(1);
> mptcp_dss_corruption(msk, ssk);
> }
[Severity: Critical]
When len > 0 and data_fin is true, does the else branch eat skb without
advancing seq by len before writing tp->copied_seq?
Leaving tp->copied_seq behind by len prevents mptcp_subflow_data_available()
from clearing subflow->map_valid and causes offset = tp->copied_seq -
TCP_SKB_CB(skb)->seq to underflow u32 on the next segment on that subflow.
In subflow_check_data_avail() -> mptcp_subflow_discard_data(), that
underflow triggers WARN_ON_ONCE(offset > skb->len) and returns without
advancing tp->copied_seq or freeing the skb, causing the caller's loop to
spin forever in softirq context:
net/mptcp/subflow.c:subflow_check_data_avail() {
...
old_ack = READ_ONCE(msk->ack_seq);
ack_seq = mptcp_subflow_get_mapped_dsn(subflow);
if (unlikely(before64(ack_seq, old_ack))) {
mptcp_subflow_discard_data(ssk, skb, old_ack - ack_seq);
continue;
}
...
}
Replacing "if (offset < skb->len)" with an unconditional
"len = skb->len - offset" also wraps len in 32-bit unsigned arithmetic if
offset ever exceeds skb->len.
--
Sashiko AI review · https://sashiko.dev/#/patchset/cover.1794810809.git.tanggeliang@kylinos.cn?part=9
next prev parent reply other threads:[~2026-11-16 7:37 UTC|newest]
Thread overview: 24+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-11-16 6:38 [PATCH mptcp-next v28 00/11] Reduce the differences between TCP and MPTCP for TLS usage Geliang Tang
2026-11-16 6:38 ` [PATCH mptcp-net v28 01/11] mptcp: fix TFO SIOCINQ overflow in mptcp_inq_hint Geliang Tang
2026-11-17 22:41 ` netdev-bot+sashiko
2026-11-16 6:38 ` [PATCH mptcp-next v28 02/11] selftests: mptcp: connect: check TFO SIOCINQ overflow Geliang Tang
2026-11-16 6:57 ` sashiko-bot
2026-11-17 22:41 ` netdev-bot+sashiko
2026-11-16 6:38 ` [PATCH mptcp-net v28 03/11] mptcp: fix fallback dummy mapping anchor Geliang Tang
2026-11-16 6:38 ` [PATCH mptcp-next v28 04/11] mptcp: drop the mptcp_ooo_try_coalesce() helper Geliang Tang
2026-11-16 6:38 ` [PATCH mptcp-next v28 05/11] mptcp: drop the cant_coalesce CB field Geliang Tang
2026-11-16 6:38 ` [PATCH mptcp-next v28 06/11] mptcp: add flags field to mptcp_skb_cb Geliang Tang
2026-11-17 22:41 ` netdev-bot+sashiko
2026-11-16 6:39 ` [PATCH mptcp-next v28 07/11] mptcp: encapsulate CB DSN field access with helpers Geliang Tang
2026-11-17 22:41 ` netdev-bot+sashiko
2026-11-16 6:39 ` [PATCH mptcp-next v28 08/11] mptcp: remove CB offset field Geliang Tang
2026-11-17 22:41 ` netdev-bot+sashiko
2026-11-16 6:39 ` [PATCH mptcp-next v28 09/11] mptcp: drop data after DATA_FIN before skb init Geliang Tang
2026-11-16 7:37 ` sashiko-bot [this message]
2026-11-17 22:41 ` netdev-bot+sashiko
2026-11-16 6:39 ` [PATCH mptcp-next v28 10/11] mptcp: sync mptcp skb cb layout with tcp one Geliang Tang
2026-11-16 8:10 ` sashiko-bot
2026-11-17 22:41 ` netdev-bot+sashiko
2026-11-16 6:39 ` [PATCH mptcp-next v28 11/11] mptcp: defer read_sock cleanup to mptcp_worker Geliang Tang
2026-11-17 22:41 ` netdev-bot+sashiko
2026-11-16 8:33 ` [PATCH mptcp-next v28 00/11] Reduce the differences between TCP and MPTCP for TLS usage MPTCP CI
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=sashiko-outbox-216443@kernel.org \
--to=sashiko-bot@kernel.org \
--cc=geliang@kernel.org \
--cc=mptcp@lists.linux.dev \
--cc=sashiko-reviews@lists.linux.dev \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.