[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20180222.141647.1474414618530735347.davem@davemloft.net>
Date: Thu, 22 Feb 2018 14:16:47 -0500 (EST)
From: David Miller <davem@...emloft.net>
To: eric.dumazet@...il.com
Cc: oleksandr@...alenko.name, edumazet@...gle.com,
netdev@...r.kernel.org, soheil@...gle.com, ncardwell@...gle.com
Subject: Re: [PATCH net] tcp_bbr: better deal with suboptimal GSO
From: Eric Dumazet <eric.dumazet@...il.com>
Date: Wed, 21 Feb 2018 06:43:03 -0800
> From: Eric Dumazet <edumazet@...gle.com>
>
> BBR uses tcp_tso_autosize() in an attempt to probe what would be the
> burst sizes and to adjust cwnd in bbr_target_cwnd() with following
> gold formula :
>
> /* Allow enough full-sized skbs in flight to utilize end systems. */
> cwnd += 3 * bbr->tso_segs_goal;
>
> But GSO can be lacking or be constrained to very small
> units (ip link set dev ... gso_max_segs 2)
>
> What we really want is to have enough packets in flight so that both
> GSO and GRO are efficient.
>
> So in the case GSO is off or downgraded, we still want to have the same
> number of packets in flight as if GSO/TSO was fully operational, so
> that GRO can hopefully be working efficiently.
>
> To fix this issue, we make tcp_tso_autosize() unaware of
> sk->sk_gso_max_segs
>
> Only tcp_tso_segs() has to enforce the gso_max_segs limit.
. ..
> Fixes: 0f8782ea1497 ("tcp_bbr: add BBR congestion control")
> Signed-off-by: Eric Dumazet <edumazet@...gle.com>
> Reported-by: Oleksandr Natalenko <oleksandr@...alenko.name>
Applied and queued up for -stable, thanks Eric.
Powered by blists - more mailing lists