Skip to contentSkip to conversation

tcp: make TCP_USER_TIMEOUT accurate for zero window probes

PATCH In review Enke Chen 4 participants 9 messages Jan 22, 2021 – Jan 24, 2021

Subsystems (3)
lore
Share & tools
lore

Enke Chen

author
#0 lore raw

PATCH 1/1 tcp: make TCP_USER_TIMEOUT accurate for zero window probes

From: Enke Chen <[email protected]>

The TCP_USER_TIMEOUT is checked by the 0-window probe timer. As the timer has backoff with a max interval of about two minutes, the actual timeout for TCP_USER_TIMEOUT can be off by up to two minutes.

In this patch the TCP_USER_TIMEOUT is made more accurate by taking it into account when computing the timer value for the 0-window probes.

This patch is similar to the one that made TCP_USER_TIMEOUT accurate for RTOs in commit b701a99e431d ("tcp: Add tcp_clamp_rto_to_user_timeout() helper to improve accuracy").

---
 include/net/tcp.h     |  1 +
 net/ipv4/tcp_input.c  |  4 ++--
 net/ipv4/tcp_output.c |  2 ++
 net/ipv4/tcp_timer.c  | 18 ++++++++++++++++++
 4 files changed, 23 insertions(+), 2 deletions(-)
include/net/tcp.h +1 -0
index 78d13c88720f..ca7e2c6cc663 100644
@@ -630,6 +630,7 @@ static inline void tcp_clear_xmit_timers(struct sock *sk)
 
 unsigned int tcp_sync_mss(struct sock *sk, u32 pmtu);
 unsigned int tcp_current_mss(struct sock *sk);
+u32 tcp_clamp_probe0_to_user_timeout(const struct sock *sk, u32 when);
 
 /* Bound MSS / TSO packet size with the half of the window */
 static inline int tcp_bound_to_half_wnd(struct tcp_sock *tp, int pktsize)
net/ipv4/tcp_input.c +2 -2
index bafcab75f425..4923cdbea95a 100644
@@ -3392,8 +3392,8 @@ static void tcp_ack_probe(struct sock *sk)
 	} else {
 		unsigned long when = tcp_probe0_when(sk, TCP_RTO_MAX);
 
-		tcp_reset_xmit_timer(sk, ICSK_TIME_PROBE0,
-				     when, TCP_RTO_MAX);
+		when = tcp_clamp_probe0_to_user_timeout(sk, when);
+		tcp_reset_xmit_timer(sk, ICSK_TIME_PROBE0, when, TCP_RTO_MAX);
 	}
 }
 
net/ipv4/tcp_output.c +2 -0
index ab458697881e..8478cf749821 100644
@@ -4099,6 +4099,8 @@ void tcp_send_probe0(struct sock *sk)
 		 */
 		timeout = TCP_RESOURCE_PROBE_INTERVAL;
 	}
+
+	timeout = tcp_clamp_probe0_to_user_timeout(sk, timeout);
 	tcp_reset_xmit_timer(sk, ICSK_TIME_PROBE0, timeout, TCP_RTO_MAX);
 }
 
net/ipv4/tcp_timer.c +18 -0
index 454732ecc8f3..90722e30ad90 100644
@@ -40,6 +40,24 @@ static u32 tcp_clamp_rto_to_user_timeout(const struct sock *sk)
 	return min_t(u32, icsk->icsk_rto, msecs_to_jiffies(remaining));
 }
 
+u32 tcp_clamp_probe0_to_user_timeout(const struct sock *sk, u32 when)
+{
+	struct inet_connection_sock *icsk = inet_csk(sk);
+	u32 remaining;
+	s32 elapsed;
+
+	if (!icsk->icsk_user_timeout || !icsk->icsk_probes_tstamp)
+		return when;
+
+	elapsed = tcp_jiffies32 - icsk->icsk_probes_tstamp;
+	if (unlikely(elapsed < 0))
+		elapsed = 0;
+	remaining = msecs_to_jiffies(icsk->icsk_user_timeout) - elapsed;
+	remaining = max_t(u32, remaining, TCP_TIMEOUT_MIN);
+
+	return min_t(u32, remaining, when);
+}
+
 /**
  *  tcp_write_err() - close socket and save error info
  *  @sk:  The socket the error has appeared on.
Signature · 1 line
2.29.2
#1 lore raw
Quoted 10 lines · Enke Chen

From: Enke Chen <[email protected]>

The TCP_USER_TIMEOUT is checked by the 0-window probe timer. As the timer has backoff with a max interval of about two minutes, the actual timeout for TCP_USER_TIMEOUT can be off by up to two minutes.

In this patch the TCP_USER_TIMEOUT is made more accurate by taking it into account when computing the timer value for the 0-window probes.

Quoted 7 lines

This patch is similar to the one that made TCP_USER_TIMEOUT accurate for RTOs in commit b701a99e431d ("tcp: Add tcp_clamp_rto_to_user_timeout() helper to improve accuracy").

---

SGTM, thanks !

maintainer
#2 lore raw
In reply to Enke Chen
Quoted 9 lines · On Fri, 22 Jan 2021 11:13:06 -0800 Enke Chen wrote:

From: Enke Chen <[email protected]>

The TCP_USER_TIMEOUT is checked by the 0-window probe timer. As the timer has backoff with a max interval of about two minutes, the actual timeout for TCP_USER_TIMEOUT can be off by up to two minutes.

In this patch the TCP_USER_TIMEOUT is made more accurate by taking it into account when computing the timer value for the 0-window probes.

Quoted 6 lines

This patch is similar to the one that made TCP_USER_TIMEOUT accurate for RTOs in commit b701a99e431d ("tcp: Add tcp_clamp_rto_to_user_timeout() helper to improve accuracy").

This is targeting net, any guidance on Fixes / backporting?

author
#3 lore raw

Hi, Jakub:

In terms of backporting, this patch should go together with:

9d9b1ee0b2d1 tcp: fix TCP_USER_TIMEOUT with zero window

Thanks. -- Enke

Quoted 18 lines · Jakub Kicinski
Quoted 9 lines · On Fri, 22 Jan 2021 11:13:06 -0800 Enke Chen wrote:

From: Enke Chen <[email protected]>

The TCP_USER_TIMEOUT is checked by the 0-window probe timer. As the timer has backoff with a max interval of about two minutes, the actual timeout for TCP_USER_TIMEOUT can be off by up to two minutes.

In this patch the TCP_USER_TIMEOUT is made more accurate by taking it into account when computing the timer value for the 0-window probes.

Quoted 6 lines

This patch is similar to the one that made TCP_USER_TIMEOUT accurate for RTOs in commit b701a99e431d ("tcp: Add tcp_clamp_rto_to_user_timeout() helper to improve accuracy").

This is targeting net, any guidance on Fixes / backporting?

maintainer
#4 lore raw
Quoted 5 lines · On Fri, 22 Jan 2021 18:28:23 -0800 Enke Chen wrote:

Hi, Jakub:

In terms of backporting, this patch should go together with:

9d9b1ee0b2d1 tcp: fix TCP_USER_TIMEOUT with zero window

As in it:

Fixes 9d9b1ee0b2d1 tcp: fix TCP_USER_TIMEOUT with zero window

or does it further fix the same issue, so:

Fixes 9721e709fa68 ("tcp: simplify window probe aborting on USER_TIMEOUT")

?

author
#5 lore raw

Hi, Jakub:

Quoted 9 lines · Jakub Kicinski
Quoted 5 lines · On Fri, 22 Jan 2021 18:28:23 -0800 Enke Chen wrote:

Hi, Jakub:

In terms of backporting, this patch should go together with:

9d9b1ee0b2d1 tcp: fix TCP_USER_TIMEOUT with zero window

As in it:

Quoted 7 lines
Fixes 9d9b1ee0b2d1 tcp: fix TCP_USER_TIMEOUT with zero window

or does it further fix the same issue, so:

Fixes 9721e709fa68 ("tcp: simplify window probe aborting on USER_TIMEOUT")

?

Let me clarify:

1) 9d9b1ee0b2d1 tcp: fix TCP_USER_TIMEOUT with zero window

fixes the bug and makes it work.

2) The current patch makes the TCP_USER_TIMEOUT accurate for 0-window probes. It's independent.

With 1) and 2), the known issues with TCP_USER_TIMEOUT for 0-window probes would be resolved.

Thanks. -- Enke

maintainer
#6 lore raw
Quoted 21 lines · Enke Chen

Hi, Jakub:

Quoted 9 lines · Jakub Kicinski
Quoted 5 lines · On Fri, 22 Jan 2021 18:28:23 -0800 Enke Chen wrote:

Hi, Jakub:

In terms of backporting, this patch should go together with:

9d9b1ee0b2d1 tcp: fix TCP_USER_TIMEOUT with zero window

As in it:

Quoted 7 lines
Fixes 9d9b1ee0b2d1 tcp: fix TCP_USER_TIMEOUT with zero window

or does it further fix the same issue, so:

Fixes 9721e709fa68 ("tcp: simplify window probe aborting on USER_TIMEOUT")

?

Quoted 8 lines

Let me clarify:

1) 9d9b1ee0b2d1 tcp: fix TCP_USER_TIMEOUT with zero window

fixes the bug and makes it work.

2) The current patch makes the TCP_USER_TIMEOUT accurate for 0-window probes. It's independent.

Patch (2) ("tcp: make TCP_USER_TIMEOUT accurate for zero window probes") is indeed conceptually independent of (1) but its implementation depends on the icsk_probes_tstamp field defined in (1), so AFAICT (2) cannot be backported further back than (1).

Patch (1) fixes a bug in 5.1:

Fixes 9721e709fa68 ("tcp: simplify window probe aborting on USER_TIMEOUT")

So probably (1) and (2) should be backported as a pair, and only back as far as 5.1. (That covers 2 LTS kernels, 5.4 and 5.10, so hopefully that is good enough.)

neal

author
#7 lore raw

Hi, Neal:

What you described is more accurate, and is correct.

Thanks. -- Enke

Quoted 44 lines · Neal Cardwell
Quoted 21 lines · Enke Chen

Hi, Jakub:

Quoted 9 lines · Jakub Kicinski
Quoted 5 lines · On Fri, 22 Jan 2021 18:28:23 -0800 Enke Chen wrote:

Hi, Jakub:

In terms of backporting, this patch should go together with:

9d9b1ee0b2d1 tcp: fix TCP_USER_TIMEOUT with zero window

As in it:

Quoted 7 lines
Fixes 9d9b1ee0b2d1 tcp: fix TCP_USER_TIMEOUT with zero window

or does it further fix the same issue, so:

Fixes 9721e709fa68 ("tcp: simplify window probe aborting on USER_TIMEOUT")

?

Quoted 8 lines

Let me clarify:

1) 9d9b1ee0b2d1 tcp: fix TCP_USER_TIMEOUT with zero window

fixes the bug and makes it work.

2) The current patch makes the TCP_USER_TIMEOUT accurate for 0-window probes. It's independent.

Patch (2) ("tcp: make TCP_USER_TIMEOUT accurate for zero window probes") is indeed conceptually independent of (1) but its implementation depends on the icsk_probes_tstamp field defined in (1), so AFAICT (2) cannot be backported further back than (1).

Patch (1) fixes a bug in 5.1:

Fixes 9721e709fa68 ("tcp: simplify window probe aborting on USER_TIMEOUT")

So probably (1) and (2) should be backported as a pair, and only back as far as 5.1. (That covers 2 LTS kernels, 5.4 and 5.10, so hopefully that is good enough.)

neal

maintainer
#8 lore raw
Quoted 32 lines · On Sat, 23 Jan 2021 16:56:43 -0800 Enke Chen wrote:
Quoted 31 lines · Neal Cardwell
Quoted 16 lines · Enke Chen
Quoted 7 lines · Jakub Kicinski
Quoted 3 lines · On Fri, 22 Jan 2021 18:28:23 -0800 Enke Chen wrote:

In terms of backporting, this patch should go together with:

9d9b1ee0b2d1 tcp: fix TCP_USER_TIMEOUT with zero window

As in it:

Quoted 7 lines
Fixes 9d9b1ee0b2d1 tcp: fix TCP_USER_TIMEOUT with zero window

or does it further fix the same issue, so:

Fixes 9721e709fa68 ("tcp: simplify window probe aborting on USER_TIMEOUT")

?

Quoted 8 lines

Let me clarify:

1) 9d9b1ee0b2d1 tcp: fix TCP_USER_TIMEOUT with zero window

fixes the bug and makes it work.

2) The current patch makes the TCP_USER_TIMEOUT accurate for 0-window probes. It's independent.

Patch (2) ("tcp: make TCP_USER_TIMEOUT accurate for zero window probes") is indeed conceptually independent of (1) but its implementation depends on the icsk_probes_tstamp field defined in (1), so AFAICT (2) cannot be backported further back than (1).

Quoted 8 lines
Quoted 6 lines

Patch (1) fixes a bug in 5.1:

Fixes 9721e709fa68 ("tcp: simplify window probe aborting on USER_TIMEOUT")

So probably (1) and (2) should be backported as a pair, and only back as far as 5.1. (That covers 2 LTS kernels, 5.4 and 5.10, so hopefully that is good enough.)

What you described is more accurate, and is correct.

That makes it clear.

I added a Fixes tag, reworded the message slightly and applied, thanks!