Hi Jakub, Paolo, Stephen,
On 23/02/2024 5:14 pm, Matthieu Baerts (NGI0) wrote:
> From: Paolo Abeni <[email protected]>
>
> Such value should be inherited from the first subflow, but
> passive sockets always used 'rsk_rcv_wnd'.
>
> Fixes: 6f8a612a33e4 ("mptcp: keep track of advertised windows right edge")
> Cc: [email protected]
> Signed-off-by: Paolo Abeni <[email protected]>
> Reviewed-by: Mat Martineau <[email protected]>
> Signed-off-by: Matthieu Baerts (NGI0) <[email protected]>
> ---
> net/mptcp/protocol.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c
> index 442fa7d9b57a..2c8f931c6d5b 100644
> --- a/net/mptcp/protocol.c
> +++ b/net/mptcp/protocol.c
> @@ -3211,7 +3211,7 @@ struct sock *mptcp_sk_clone_init(const struct sock *sk,
> msk->write_seq = subflow_req->idsn + 1;
> msk->snd_nxt = msk->write_seq;
> msk->snd_una = msk->write_seq;
> - msk->wnd_end = msk->snd_nxt + req->rsk_rcv_wnd;
> + msk->wnd_end = msk->snd_nxt + tcp_sk(ssk)->snd_wnd;
Please note that this patch will conflict with the following commit from
net-next:
3f83d8a77eee ("mptcp: fix more tx path fields initialization")
That's because this commit modifies the same line as the one modified
here. We cannot avoid a conflict here. To fix it, please use
'WRITE_ONCE()' with the new line from -net:
WRITE_ONCE(msk->wnd_end, msk->snd_nxt + tcp_sk(ssk)->snd_wnd);
A 3-way diff has been attached to this email.
> msk->setsockopt_seq = mptcp_sk(sk)->setsockopt_seq;
> mptcp_init_sched(msk, mptcp_sk(sk)->sched);
>
>
Cheers,
Matt
--
Sponsored by the NGI0 Core fund.