Re: [PATCH net 2/7] mptcp: push pending data after a sendmsg copy fault
From: netdev-bot+sashiko
Date: Tue Nov 17 2026 - 17:30:02 EST
Thank you for your contribution! Sashiko AI review found more severe
issue(s) in other patches of this series. This mail is only an FYI, in case
you want to address the nit picks below when you respin.
Low priority / nit picks for your consideration:
- [Low] mptcp: Fixes trailer in commit message cites 69800e516e96 instead
of d9ca1de8c0cd
--
> diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c
> --- a/net/mptcp/protocol.c
> +++ b/net/mptcp/protocol.c
> @@ -2130,8 +2130,10 @@ static int mptcp_sendmsg(struct sock *sk, struct msghdr *msg, size_t len)
> return copied;
>
> do_error:
> - if (copied)
> + if (copied) {
> + __mptcp_push_pending(sk, msg->msg_flags);
> goto out;
> + }
>
> copied = sk_stream_error(sk, msg->msg_flags, ret);
> goto out;
>
[Severity: Low]
The Fixes tag looks off to me. Before 69800e516e96 ("mptcp: propagate
fastclose error"), a copy fault in mptcp_sendmsg() already set
ret = -EFAULT and jumped straight to out. The out label sat after the
if (copied) __mptcp_push_pending() block and returned copied ? : ret.
So with copied > 0, the already queued dfrag was left on rtx_queue
without a push even back then. 69800e516e96 only introduced do_error
to pass ret through sk_stream_error() when copied == 0. It kept the
existing if (copied) goto out; behaviour.
As far as I can tell, skipping the push on a partial copy error goes
back to d9ca1de8c0cd ("mptcp: move page frag allocation in
mptcp_sendmsg()"). That commit made mptcp_sendmsg() queue everything
first and only push after the loop.
With the current tag, stable tooling would miss kernels that have
d9ca1de8c0cd but not 69800e516e96.
Could you double check, and if you agree, point the Fixes tag at
d9ca1de8c0cd instead?
--
Sashiko AI review · https://netdev-ai.bots.linux.dev/sashiko/#/patchset/20261116-net-mptcp-misc-fixes-7-4-rc4-v1-0-bf7656de653e%40kernel.org