From: Wei Wang <[email protected]> In tcp_select_initial_window(), we only set rcv_wnd to tcp_default_init_rwnd() if current mss > (1 << wscale). Otherwise, rcv_wnd is kept at the full receive space of the socket which is a value way larger than tcp_default_init_rwnd(). With larger initial rcv_wnd value, receive buffer autotuning logic takes longer to kick in and increase the receive buffer.
In a TCP throughput test where receiver has rmem[2] set to 125MB (wscale is 11), we see the connection gets recvbuf limited at the beginning of the connection and gets less throughput overall. Signed-off-by: Wei Wang <[email protected]> Acked-by: Eric Dumazet <[email protected]> Acked-by: Soheil Hassas Yeganeh <[email protected]> Acked-by: Yuchung Cheng <[email protected]> --- net/ipv4/tcp_output.c | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index 95feffb6d53f..d07c0dcc99aa 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -229,11 +229,9 @@ void tcp_select_initial_window(const struct sock *sk, int __space, __u32 mss, } } - if (mss > (1 << *rcv_wscale)) { - if (!init_rcv_wnd) /* Use default unless specified otherwise */ - init_rcv_wnd = tcp_default_init_rwnd(mss); - *rcv_wnd = min(*rcv_wnd, init_rcv_wnd * mss); - } + if (!init_rcv_wnd) /* Use default unless specified otherwise */ + init_rcv_wnd = tcp_default_init_rwnd(mss); + *rcv_wnd = min(*rcv_wnd, init_rcv_wnd * mss); /* Set the clamp no higher than max representable value */ (*window_clamp) = min_t(__u32, U16_MAX << (*rcv_wscale), *window_clamp); -- 2.17.0.484.g0c8726318c-goog
