* [PATCH net-next] tcp: Make SYN ACK RTO tunable by BPF programs with TFO
@ 2022-08-15 20:29 Jie Meng
2022-08-16 13:36 ` Neal Cardwell
` (2 more replies)
0 siblings, 3 replies; 6+ messages in thread
From: Jie Meng @ 2022-08-15 20:29 UTC (permalink / raw)
To: netdev; +Cc: kafai, kuba, edumazet, bpf, Jie Meng
Instead of the hardcoded TCP_TIMEOUT_INIT, this diff calls tcp_timeout_init
to initiate req->timeout like the non TFO SYN ACK case.
Tested using the following packetdrill script, on a host with a BPF
program that sets the initial connect timeout to 10ms.
`../../common/defaults.sh`
// Initialize connection
0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_TCP, TCP_FASTOPEN, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
+0 < S 0:0(0) win 32792 <mss 1000,sackOK,FO TFO_COOKIE>
+0 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK>
+.01 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK>
+.02 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK>
+.04 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK>
+.01 < . 1:1(0) ack 1 win 32792
+0 accept(3, ..., ...) = 4
Signed-off-by: Jie Meng <jmeng@fb.com>
---
net/ipv4/tcp_fastopen.c | 3 ++-
net/ipv4/tcp_timer.c | 2 +-
2 files changed, 3 insertions(+), 2 deletions(-)
diff --git a/net/ipv4/tcp_fastopen.c b/net/ipv4/tcp_fastopen.c
index 825b216d11f5..45cc7f1ca296 100644
--- a/net/ipv4/tcp_fastopen.c
+++ b/net/ipv4/tcp_fastopen.c
@@ -272,8 +272,9 @@ static struct sock *tcp_fastopen_create_child(struct sock *sk,
* The request socket is not added to the ehash
* because it's been added to the accept queue directly.
*/
+ req->timeout = tcp_timeout_init(child);
inet_csk_reset_xmit_timer(child, ICSK_TIME_RETRANS,
- TCP_TIMEOUT_INIT, TCP_RTO_MAX);
+ req->timeout, TCP_RTO_MAX);
refcount_set(&req->rsk_refcnt, 2);
diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c
index b4dfb82d6ecb..cb79127f45c3 100644
--- a/net/ipv4/tcp_timer.c
+++ b/net/ipv4/tcp_timer.c
@@ -428,7 +428,7 @@ static void tcp_fastopen_synack_timer(struct sock *sk, struct request_sock *req)
if (!tp->retrans_stamp)
tp->retrans_stamp = tcp_time_stamp(tp);
inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS,
- TCP_TIMEOUT_INIT << req->num_timeout, TCP_RTO_MAX);
+ req->timeout << req->num_timeout, TCP_RTO_MAX);
}
--
2.30.2
^ permalink raw reply related [flat|nested] 6+ messages in thread* Re: [PATCH net-next] tcp: Make SYN ACK RTO tunable by BPF programs with TFO
2022-08-15 20:29 [PATCH net-next] tcp: Make SYN ACK RTO tunable by BPF programs with TFO Jie Meng
@ 2022-08-16 13:36 ` Neal Cardwell
2022-08-16 16:11 ` Yuchung Cheng
2022-08-16 16:59 ` Martin KaFai Lau
2022-08-17 9:30 ` patchwork-bot+netdevbpf
2 siblings, 1 reply; 6+ messages in thread
From: Neal Cardwell @ 2022-08-16 13:36 UTC (permalink / raw)
To: Jie Meng; +Cc: netdev, kafai, kuba, edumazet, bpf, Yuchung Cheng, Wei Wang
On Mon, Aug 15, 2022 at 8:30 PM Jie Meng <jmeng@fb.com> wrote:
>
> Instead of the hardcoded TCP_TIMEOUT_INIT, this diff calls tcp_timeout_init
> to initiate req->timeout like the non TFO SYN ACK case.
>
> Tested using the following packetdrill script, on a host with a BPF
> program that sets the initial connect timeout to 10ms.
>
> `../../common/defaults.sh`
>
> // Initialize connection
> 0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
> +0 setsockopt(3, SOL_TCP, TCP_FASTOPEN, [1], 4) = 0
> +0 bind(3, ..., ...) = 0
> +0 listen(3, 1) = 0
>
> +0 < S 0:0(0) win 32792 <mss 1000,sackOK,FO TFO_COOKIE>
> +0 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK>
> +.01 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK>
> +.02 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK>
> +.04 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK>
> +.01 < . 1:1(0) ack 1 win 32792
>
> +0 accept(3, ..., ...) = 4
>
> Signed-off-by: Jie Meng <jmeng@fb.com>
> ---
> net/ipv4/tcp_fastopen.c | 3 ++-
> net/ipv4/tcp_timer.c | 2 +-
> 2 files changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/net/ipv4/tcp_fastopen.c b/net/ipv4/tcp_fastopen.c
> index 825b216d11f5..45cc7f1ca296 100644
> --- a/net/ipv4/tcp_fastopen.c
> +++ b/net/ipv4/tcp_fastopen.c
> @@ -272,8 +272,9 @@ static struct sock *tcp_fastopen_create_child(struct sock *sk,
> * The request socket is not added to the ehash
> * because it's been added to the accept queue directly.
> */
> + req->timeout = tcp_timeout_init(child);
> inet_csk_reset_xmit_timer(child, ICSK_TIME_RETRANS,
> - TCP_TIMEOUT_INIT, TCP_RTO_MAX);
> + req->timeout, TCP_RTO_MAX);
>
> refcount_set(&req->rsk_refcnt, 2);
>
> diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c
> index b4dfb82d6ecb..cb79127f45c3 100644
> --- a/net/ipv4/tcp_timer.c
> +++ b/net/ipv4/tcp_timer.c
> @@ -428,7 +428,7 @@ static void tcp_fastopen_synack_timer(struct sock *sk, struct request_sock *req)
> if (!tp->retrans_stamp)
> tp->retrans_stamp = tcp_time_stamp(tp);
> inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS,
> - TCP_TIMEOUT_INIT << req->num_timeout, TCP_RTO_MAX);
> + req->timeout << req->num_timeout, TCP_RTO_MAX);
> }
>
>
> --
Looks good to me. Thanks for the feature!
Acked-by: Neal Cardwell <ncardwell@google.com>
neal
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH net-next] tcp: Make SYN ACK RTO tunable by BPF programs with TFO
2022-08-16 13:36 ` Neal Cardwell
@ 2022-08-16 16:11 ` Yuchung Cheng
0 siblings, 0 replies; 6+ messages in thread
From: Yuchung Cheng @ 2022-08-16 16:11 UTC (permalink / raw)
To: Neal Cardwell; +Cc: Jie Meng, netdev, kafai, kuba, edumazet, bpf, Wei Wang
On Tue, Aug 16, 2022 at 6:37 AM Neal Cardwell <ncardwell@google.com> wrote:
>
> On Mon, Aug 15, 2022 at 8:30 PM Jie Meng <jmeng@fb.com> wrote:
> >
> > Instead of the hardcoded TCP_TIMEOUT_INIT, this diff calls tcp_timeout_init
> > to initiate req->timeout like the non TFO SYN ACK case.
> >
> > Tested using the following packetdrill script, on a host with a BPF
> > program that sets the initial connect timeout to 10ms.
> >
> > `../../common/defaults.sh`
> >
> > // Initialize connection
> > 0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
> > +0 setsockopt(3, SOL_TCP, TCP_FASTOPEN, [1], 4) = 0
> > +0 bind(3, ..., ...) = 0
> > +0 listen(3, 1) = 0
> >
> > +0 < S 0:0(0) win 32792 <mss 1000,sackOK,FO TFO_COOKIE>
> > +0 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK>
> > +.01 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK>
> > +.02 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK>
> > +.04 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK>
> > +.01 < . 1:1(0) ack 1 win 32792
> >
> > +0 accept(3, ..., ...) = 4
> >
> > Signed-off-by: Jie Meng <jmeng@fb.com>
> > ---
> > net/ipv4/tcp_fastopen.c | 3 ++-
> > net/ipv4/tcp_timer.c | 2 +-
> > 2 files changed, 3 insertions(+), 2 deletions(-)
> >
> > diff --git a/net/ipv4/tcp_fastopen.c b/net/ipv4/tcp_fastopen.c
> > index 825b216d11f5..45cc7f1ca296 100644
> > --- a/net/ipv4/tcp_fastopen.c
> > +++ b/net/ipv4/tcp_fastopen.c
> > @@ -272,8 +272,9 @@ static struct sock *tcp_fastopen_create_child(struct sock *sk,
> > * The request socket is not added to the ehash
> > * because it's been added to the accept queue directly.
> > */
> > + req->timeout = tcp_timeout_init(child);
> > inet_csk_reset_xmit_timer(child, ICSK_TIME_RETRANS,
> > - TCP_TIMEOUT_INIT, TCP_RTO_MAX);
> > + req->timeout, TCP_RTO_MAX);
> >
> > refcount_set(&req->rsk_refcnt, 2);
> >
> > diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c
> > index b4dfb82d6ecb..cb79127f45c3 100644
> > --- a/net/ipv4/tcp_timer.c
> > +++ b/net/ipv4/tcp_timer.c
> > @@ -428,7 +428,7 @@ static void tcp_fastopen_synack_timer(struct sock *sk, struct request_sock *req)
> > if (!tp->retrans_stamp)
> > tp->retrans_stamp = tcp_time_stamp(tp);
> > inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS,
> > - TCP_TIMEOUT_INIT << req->num_timeout, TCP_RTO_MAX);
> > + req->timeout << req->num_timeout, TCP_RTO_MAX);
> > }
> >
> >
> > --
>
> Looks good to me. Thanks for the feature!
>
> Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Would be great to have a companion patch on SYN timeout as well.
>
> neal
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH net-next] tcp: Make SYN ACK RTO tunable by BPF programs with TFO
2022-08-15 20:29 [PATCH net-next] tcp: Make SYN ACK RTO tunable by BPF programs with TFO Jie Meng
2022-08-16 13:36 ` Neal Cardwell
@ 2022-08-16 16:59 ` Martin KaFai Lau
2022-08-17 9:30 ` patchwork-bot+netdevbpf
2 siblings, 0 replies; 6+ messages in thread
From: Martin KaFai Lau @ 2022-08-16 16:59 UTC (permalink / raw)
To: Jie Meng; +Cc: netdev, kuba, edumazet, bpf
On Mon, Aug 15, 2022 at 01:29:00PM -0700, Jie Meng wrote:
> Instead of the hardcoded TCP_TIMEOUT_INIT, this diff calls tcp_timeout_init
> to initiate req->timeout like the non TFO SYN ACK case.
>
> Tested using the following packetdrill script, on a host with a BPF
> program that sets the initial connect timeout to 10ms.
>
> `../../common/defaults.sh`
>
> // Initialize connection
> 0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
> +0 setsockopt(3, SOL_TCP, TCP_FASTOPEN, [1], 4) = 0
> +0 bind(3, ..., ...) = 0
> +0 listen(3, 1) = 0
>
> +0 < S 0:0(0) win 32792 <mss 1000,sackOK,FO TFO_COOKIE>
> +0 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK>
> +.01 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK>
> +.02 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK>
> +.04 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK>
> +.01 < . 1:1(0) ack 1 win 32792
>
> +0 accept(3, ..., ...) = 4
Acked-by: Martin KaFai Lau <kafai@fb.com>
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH net-next] tcp: Make SYN ACK RTO tunable by BPF programs with TFO
2022-08-15 20:29 [PATCH net-next] tcp: Make SYN ACK RTO tunable by BPF programs with TFO Jie Meng
2022-08-16 13:36 ` Neal Cardwell
2022-08-16 16:59 ` Martin KaFai Lau
@ 2022-08-17 9:30 ` patchwork-bot+netdevbpf
2 siblings, 0 replies; 6+ messages in thread
From: patchwork-bot+netdevbpf @ 2022-08-17 9:30 UTC (permalink / raw)
To: Jie Meng; +Cc: netdev, kafai, kuba, edumazet, bpf
Hello:
This patch was applied to netdev/net-next.git (master)
by David S. Miller <davem@davemloft.net>:
On Mon, 15 Aug 2022 13:29:00 -0700 you wrote:
> Instead of the hardcoded TCP_TIMEOUT_INIT, this diff calls tcp_timeout_init
> to initiate req->timeout like the non TFO SYN ACK case.
>
> Tested using the following packetdrill script, on a host with a BPF
> program that sets the initial connect timeout to 10ms.
>
> `../../common/defaults.sh`
>
> [...]
Here is the summary with links:
- [net-next] tcp: Make SYN ACK RTO tunable by BPF programs with TFO
https://git.kernel.org/netdev/net-next/c/8ea731d4c2ce
You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html
^ permalink raw reply [flat|nested] 6+ messages in thread
[parent not found: <20220806000635.472853-1-jmeng@fb.com>]
* Re: [PATCH net-next] tcp: Make SYN ACK RTO tunable by BPF programs with TFO
[not found] <20220806000635.472853-1-jmeng@fb.com>
@ 2022-08-10 22:30 ` Martin KaFai Lau
0 siblings, 0 replies; 6+ messages in thread
From: Martin KaFai Lau @ 2022-08-10 22:30 UTC (permalink / raw)
To: Jie Meng; +Cc: netdev, bpf
On Fri, Aug 05, 2022 at 05:06:35PM -0700, Jie Meng wrote:
> Instead of the hardcoded TCP_TIMEOUT_INIT, this diff calls tcp_timeout_init
> to initiate req->timeout like the non TFO SYN ACK case.
>
> Tested using the following packetdrill script, on a host with a BPF
> program that sets the initial connect timeout to 10ms.
Please also cc the bpf mailing list.
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2022-08-17 9:30 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-08-15 20:29 [PATCH net-next] tcp: Make SYN ACK RTO tunable by BPF programs with TFO Jie Meng
2022-08-16 13:36 ` Neal Cardwell
2022-08-16 16:11 ` Yuchung Cheng
2022-08-16 16:59 ` Martin KaFai Lau
2022-08-17 9:30 ` patchwork-bot+netdevbpf
[not found] <20220806000635.472853-1-jmeng@fb.com>
2022-08-10 22:30 ` Martin KaFai Lau
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).