[PATCH net-next] mptcp: adjust tcp rcvspace after moving skbs from ssk to sk queue

[PATCH net-next] mptcp: adjust tcp rcvspace after moving skbs from ssk to sk queue

[Florian Westphal]

TCP does tcp rcvbuf tuning when copying packets to userspace, e.g. in
tcp_read_sock().  In case of mptcp, that function is only rarely used
(when discarding retransmitted duplicate data).

Instead, skbs are moved from the tcp rx queue to the mptcp socket rx
queue.
Adjust subflow rcvbuf when we do so, its the last spot where we can
adjust the ssk rcvbuf -- later we only have the mptcp-level socket.

This greatly improves performance on mptcp bulk transfers.

Signed-off-by: Florian Westphal <fw@strlen.de>
---
 net/mptcp/protocol.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c
index ba9d3d5c625f..dbb86cbb9e77 100644
--- a/net/mptcp/protocol.c
+++ b/net/mptcp/protocol.c
@@ -248,6 +248,9 @@ static bool __mptcp_move_skbs_from_subflow(struct mptcp_sock *msk,
 
 	*bytes = moved;
 
+	if (moved)
+		tcp_rcv_space_adjust(ssk);
+
 	return done;
 }
 
@@ -1263,6 +1266,7 @@ static int mptcp_init_sock(struct sock *sk)
 		return ret;
 
 	sk_sockets_allocated_inc(sk);
+	sk->sk_rcvbuf = sock_net(sk)->ipv4.sysctl_tcp_rmem[1];
 	sk->sk_sndbuf = sock_net(sk)->ipv4.sysctl_tcp_wmem[2];
 
 	return 0;
-- 
2.26.2

Re: [PATCH net-next] mptcp: adjust tcp rcvspace after moving skbs from ssk to sk queue

[Paolo Abeni]

Hi,

On Fri, 2020-05-22 at 14:43 +0200, Florian Westphal wrote:
> TCP does tcp rcvbuf tuning when copying packets to userspace, e.g. in
> tcp_read_sock().  In case of mptcp, that function is only rarely used
> (when discarding retransmitted duplicate data).
> 
> Instead, skbs are moved from the tcp rx queue to the mptcp socket rx
> queue.
> Adjust subflow rcvbuf when we do so, its the last spot where we can
> adjust the ssk rcvbuf -- later we only have the mptcp-level socket.
> 
> This greatly improves performance on mptcp bulk transfers.
> 
> Signed-off-by: Florian Westphal <fw@strlen.de>
> ---
>  net/mptcp/protocol.c | 4 ++++
>  1 file changed, 4 insertions(+)
> 
> diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c
> index ba9d3d5c625f..dbb86cbb9e77 100644
> --- a/net/mptcp/protocol.c
> +++ b/net/mptcp/protocol.c
> @@ -248,6 +248,9 @@ static bool __mptcp_move_skbs_from_subflow(struct mptcp_sock *msk,
>  
>  	*bytes = moved;
>  
> +	if (moved)
> +		tcp_rcv_space_adjust(ssk);
> +
>  	return done;
>  }

It looks like this way ssk rcvbuf will grow up to tcp_rmem[2] even if
there is no user-space reader - assuming the link is fast enough.

Don't we need to somehow cap that? e.g. moving mptcp rcvbuf update in
mptcp_revmsg()?

Thanks,

Paolo

Re: [PATCH net-next] mptcp: adjust tcp rcvspace after moving skbs from ssk to sk queue

[Florian Westphal]

Paolo Abeni <pabeni@redhat.com> wrote:
> It looks like this way ssk rcvbuf will grow up to tcp_rmem[2] even if
> there is no user-space reader - assuming the link is fast enough.
> 
> Don't we need to somehow cap that? e.g. moving mptcp rcvbuf update in
> mptcp_revmsg()?

From irc discussion it looks like you already have a patch for that.

I marked this patch as "Changes Requested" and do not plan to resend
it for the time being.