Thread (3 messages) flat view 3 messages, 2 authors, 2021-03-03

Re: [PATCH net-next v2] tcp: make TCP Fast Open retransmission ignore Tx status

From: Eric Dumazet <edumazet@google.com>
Date: 2021-03-03 04:23:53

On Tue, Mar 2, 2021 at 7:00 PM Jakub Kicinski [off-list ref] wrote:
On Tue,  2 Mar 2021 09:52:59 -0800 Jakub Kicinski wrote:
quoted
When receiver does not accept TCP Fast Open it will only ack
the SYN, and not the data. We detect this and immediately queue
the data for (re)transmission in tcp_rcv_fastopen_synack().

In DC networks with very low RTT and without RFS the SYN-ACK
may arrive before NIC driver reported Tx completion on
the original SYN. In which case skb_still_in_host_queue()
returns true and sender will need to wait for the retransmission
timer to fire milliseconds later.

Work around this issue by passing negative segment count to
__tcp_retransmit_skb() as suggested by Eric.

The condition triggers more often when Tx coalescing is configured
higher than Rx coalescing on the underlying NIC, but it does happen
even with relatively moderate and even settings (e.g. 33us).

Note that DC machines usually run configured to always accept
TCP FastOpen data so the problem may not be very common.

Suggested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Neil Spring <redacted>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
.. and now I realized net-next is closed. I'll keep an eye on patchwork
and resend as needed, sorry.
Ah, just open it ;)
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help