2019-05-19 20:08:55 +08:00
|
|
|
// SPDX-License-Identifier: GPL-2.0-only
|
2005-04-17 06:20:36 +08:00
|
|
|
/*
|
|
|
|
|
* INET An implementation of the TCP/IP protocol suite for the LINUX
|
|
|
|
|
* operating system. INET is implemented using the BSD Socket
|
|
|
|
|
* interface as the means of communication with the user level.
|
|
|
|
|
*
|
|
|
|
|
* Implementation of the Transmission Control Protocol(TCP).
|
|
|
|
|
*
|
2005-05-06 07:16:16 +08:00
|
|
|
* Authors: Ross Biro
|
2005-04-17 06:20:36 +08:00
|
|
|
* Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
|
|
|
|
|
* Mark Evans, <evansmp@uhura.aston.ac.uk>
|
|
|
|
|
* Corey Minyard <wf-rch!minyard@relay.EU.net>
|
|
|
|
|
* Florian La Roche, <flla@stud.uni-sb.de>
|
|
|
|
|
* Charles Hedrick, <hedrick@klinzhai.rutgers.edu>
|
|
|
|
|
* Linus Torvalds, <torvalds@cs.helsinki.fi>
|
|
|
|
|
* Alan Cox, <gw4pts@gw4pts.ampr.org>
|
|
|
|
|
* Matthew Dillon, <dillon@apollo.west.oic.com>
|
|
|
|
|
* Arnt Gulbrandsen, <agulbra@nvg.unit.no>
|
|
|
|
|
* Jorge Cwik, <jorge@laser.satlink.net>
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* Changes: Pedro Roque : Retransmit queue handled by TCP.
|
|
|
|
|
* : Fragmentation on mtu decrease
|
|
|
|
|
* : Segment collapse on retransmit
|
|
|
|
|
* : AF independence
|
|
|
|
|
*
|
|
|
|
|
* Linus Torvalds : send_delayed_ack
|
|
|
|
|
* David S. Miller : Charge memory using the right skb
|
|
|
|
|
* during syn/ack processing.
|
|
|
|
|
* David S. Miller : Output engine completely rewritten.
|
|
|
|
|
* Andrea Arcangeli: SYNACK carry ts_recent in tsecr.
|
|
|
|
|
* Cacophonix Gaul : draft-minshall-nagle-01
|
|
|
|
|
* J Hadi Salim : ECN support
|
|
|
|
|
*
|
|
|
|
|
*/
|
|
|
|
|
|
2012-05-15 22:11:54 +08:00
|
|
|
#define pr_fmt(fmt) "TCP: " fmt
|
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
#include <net/tcp.h>
|
2020-01-22 08:56:16 +08:00
|
|
|
#include <net/mptcp.h>
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
|
#include <linux/compiler.h>
|
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 16:04:11 +08:00
|
|
|
#include <linux/gfp.h>
|
2005-04-17 06:20:36 +08:00
|
|
|
#include <linux/module.h>
|
2017-10-25 17:01:45 +08:00
|
|
|
#include <linux/static_key.h>
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2017-10-14 04:03:16 +08:00
|
|
|
#include <trace/events/tcp.h>
|
2006-06-14 13:33:04 +08:00
|
|
|
|
2018-09-21 23:51:49 +08:00
|
|
|
/* Refresh clocks of a TCP socket,
|
|
|
|
|
* ensuring monotically increasing values.
|
|
|
|
|
*/
|
|
|
|
|
void tcp_mstamp_refresh(struct tcp_sock *tp)
|
|
|
|
|
{
|
|
|
|
|
u64 val = tcp_clock_ns();
|
|
|
|
|
|
2019-03-22 21:26:29 +08:00
|
|
|
tp->tcp_clock_cache = val;
|
|
|
|
|
tp->tcp_mstamp = div_u64(val, NSEC_PER_USEC);
|
2018-09-21 23:51:49 +08:00
|
|
|
}
|
|
|
|
|
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
static bool tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle,
|
|
|
|
|
int push_one, gfp_t gfp);
|
2009-12-03 02:14:19 +08:00
|
|
|
|
2009-07-22 07:00:40 +08:00
|
|
|
/* Account for new data that has been sent to the network. */
|
2017-10-06 13:21:27 +08:00
|
|
|
static void tcp_event_new_data_sent(struct sock *sk, struct sk_buff *skb)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
struct inet_connection_sock *icsk = inet_csk(sk);
|
[TCP]: Sed magic converts func(sk, tp, ...) -> func(sk, ...)
This is (mostly) automated change using magic:
sed -e '/struct sock \*sk/ N' -e '/struct sock \*sk/ N'
-e '/struct sock \*sk/ N' -e '/struct sock \*sk/ N'
-e 's|struct sock \*sk,[\n\t ]*struct tcp_sock \*tp\([^{]*\n{\n\)|
struct sock \*sk\1\tstruct tcp_sock *tp = tcp_sk(sk);\n|g'
-e 's|struct sock \*sk, struct tcp_sock \*tp|
struct sock \*sk|g' -e 's|sk, tp\([^-]\)|sk\1|g'
Fixed four unused variable (tp) warnings that were introduced.
In addition, manually added newlines after local variables and
tweaked function arguments positioning.
$ gcc --version
gcc (GCC) 4.1.1 20060525 (Red Hat 4.1.1-1)
...
$ codiff -fV built-in.o.old built-in.o.new
net/ipv4/route.c:
rt_cache_flush | +14
1 function changed, 14 bytes added
net/ipv4/tcp.c:
tcp_setsockopt | -5
tcp_sendpage | -25
tcp_sendmsg | -16
3 functions changed, 46 bytes removed
net/ipv4/tcp_input.c:
tcp_try_undo_recovery | +3
tcp_try_undo_dsack | +2
tcp_mark_head_lost | -12
tcp_ack | -15
tcp_event_data_recv | -32
tcp_rcv_state_process | -10
tcp_rcv_established | +1
7 functions changed, 6 bytes added, 69 bytes removed, diff: -63
net/ipv4/tcp_output.c:
update_send_head | -9
tcp_transmit_skb | +19
tcp_cwnd_validate | +1
tcp_write_wakeup | -17
__tcp_push_pending_frames | -25
tcp_push_one | -8
tcp_send_fin | -4
7 functions changed, 20 bytes added, 63 bytes removed, diff: -43
built-in.o.new:
18 functions changed, 40 bytes added, 178 bytes removed, diff: -138
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-21 13:18:02 +08:00
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
2007-12-31 20:43:57 +08:00
|
|
|
unsigned int prior_packets = tp->packets_out;
|
[TCP]: Sed magic converts func(sk, tp, ...) -> func(sk, ...)
This is (mostly) automated change using magic:
sed -e '/struct sock \*sk/ N' -e '/struct sock \*sk/ N'
-e '/struct sock \*sk/ N' -e '/struct sock \*sk/ N'
-e 's|struct sock \*sk,[\n\t ]*struct tcp_sock \*tp\([^{]*\n{\n\)|
struct sock \*sk\1\tstruct tcp_sock *tp = tcp_sk(sk);\n|g'
-e 's|struct sock \*sk, struct tcp_sock \*tp|
struct sock \*sk|g' -e 's|sk, tp\([^-]\)|sk\1|g'
Fixed four unused variable (tp) warnings that were introduced.
In addition, manually added newlines after local variables and
tweaked function arguments positioning.
$ gcc --version
gcc (GCC) 4.1.1 20060525 (Red Hat 4.1.1-1)
...
$ codiff -fV built-in.o.old built-in.o.new
net/ipv4/route.c:
rt_cache_flush | +14
1 function changed, 14 bytes added
net/ipv4/tcp.c:
tcp_setsockopt | -5
tcp_sendpage | -25
tcp_sendmsg | -16
3 functions changed, 46 bytes removed
net/ipv4/tcp_input.c:
tcp_try_undo_recovery | +3
tcp_try_undo_dsack | +2
tcp_mark_head_lost | -12
tcp_ack | -15
tcp_event_data_recv | -32
tcp_rcv_state_process | -10
tcp_rcv_established | +1
7 functions changed, 6 bytes added, 69 bytes removed, diff: -63
net/ipv4/tcp_output.c:
update_send_head | -9
tcp_transmit_skb | +19
tcp_cwnd_validate | +1
tcp_write_wakeup | -17
__tcp_push_pending_frames | -25
tcp_push_one | -8
tcp_send_fin | -4
7 functions changed, 20 bytes added, 63 bytes removed, diff: -43
built-in.o.new:
18 functions changed, 40 bytes added, 178 bytes removed, diff: -138
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-21 13:18:02 +08:00
|
|
|
|
2019-10-11 11:17:42 +08:00
|
|
|
WRITE_ONCE(tp->snd_nxt, TCP_SKB_CB(skb)->end_seq);
|
2007-11-26 20:17:38 +08:00
|
|
|
|
2017-10-06 13:21:27 +08:00
|
|
|
__skb_unlink(skb, &sk->sk_write_queue);
|
|
|
|
|
tcp_rbtree_insert(&sk->tcp_rtx_queue, skb);
|
|
|
|
|
|
2019-12-27 16:52:37 +08:00
|
|
|
if (tp->highest_sack == NULL)
|
|
|
|
|
tp->highest_sack = skb;
|
|
|
|
|
|
2007-12-31 20:43:57 +08:00
|
|
|
tp->packets_out += tcp_skb_pcount(skb);
|
2017-01-13 14:11:39 +08:00
|
|
|
if (!prior_packets || icsk->icsk_pending == ICSK_TIME_LOSS_PROBE)
|
2012-05-02 21:30:04 +08:00
|
|
|
tcp_rearm_rto(sk);
|
2014-03-04 04:31:36 +08:00
|
|
|
|
2014-03-07 04:03:17 +08:00
|
|
|
NET_ADD_STATS(sock_net(sk), LINUX_MIB_TCPORIGDATASENT,
|
|
|
|
|
tcp_skb_pcount(skb));
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
|
2017-02-18 01:04:55 +08:00
|
|
|
/* SND.NXT, if window was not shrunk or the amount of shrunk was less than one
|
|
|
|
|
* window scaling factor due to loss of precision.
|
2005-04-17 06:20:36 +08:00
|
|
|
* If window has been shrunk, what should we make? It is not clear at all.
|
|
|
|
|
* Using SND.UNA we will fail to open window, SND.NXT is out of window. :-(
|
|
|
|
|
* Anything in between SND.UNA...SND.UNA+SND.WND also can be already
|
|
|
|
|
* invalid. OK, let's make this for now:
|
|
|
|
|
*/
|
2011-10-21 17:22:42 +08:00
|
|
|
static inline __u32 tcp_acceptable_seq(const struct sock *sk)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
2011-10-21 17:22:42 +08:00
|
|
|
const struct tcp_sock *tp = tcp_sk(sk);
|
[TCP]: Sed magic converts func(sk, tp, ...) -> func(sk, ...)
This is (mostly) automated change using magic:
sed -e '/struct sock \*sk/ N' -e '/struct sock \*sk/ N'
-e '/struct sock \*sk/ N' -e '/struct sock \*sk/ N'
-e 's|struct sock \*sk,[\n\t ]*struct tcp_sock \*tp\([^{]*\n{\n\)|
struct sock \*sk\1\tstruct tcp_sock *tp = tcp_sk(sk);\n|g'
-e 's|struct sock \*sk, struct tcp_sock \*tp|
struct sock \*sk|g' -e 's|sk, tp\([^-]\)|sk\1|g'
Fixed four unused variable (tp) warnings that were introduced.
In addition, manually added newlines after local variables and
tweaked function arguments positioning.
$ gcc --version
gcc (GCC) 4.1.1 20060525 (Red Hat 4.1.1-1)
...
$ codiff -fV built-in.o.old built-in.o.new
net/ipv4/route.c:
rt_cache_flush | +14
1 function changed, 14 bytes added
net/ipv4/tcp.c:
tcp_setsockopt | -5
tcp_sendpage | -25
tcp_sendmsg | -16
3 functions changed, 46 bytes removed
net/ipv4/tcp_input.c:
tcp_try_undo_recovery | +3
tcp_try_undo_dsack | +2
tcp_mark_head_lost | -12
tcp_ack | -15
tcp_event_data_recv | -32
tcp_rcv_state_process | -10
tcp_rcv_established | +1
7 functions changed, 6 bytes added, 69 bytes removed, diff: -63
net/ipv4/tcp_output.c:
update_send_head | -9
tcp_transmit_skb | +19
tcp_cwnd_validate | +1
tcp_write_wakeup | -17
__tcp_push_pending_frames | -25
tcp_push_one | -8
tcp_send_fin | -4
7 functions changed, 20 bytes added, 63 bytes removed, diff: -43
built-in.o.new:
18 functions changed, 40 bytes added, 178 bytes removed, diff: -138
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-21 13:18:02 +08:00
|
|
|
|
2017-02-18 01:04:55 +08:00
|
|
|
if (!before(tcp_wnd_end(tp), tp->snd_nxt) ||
|
|
|
|
|
(tp->rx_opt.wscale_ok &&
|
|
|
|
|
((tp->snd_nxt - tcp_wnd_end(tp)) < (1 << tp->rx_opt.rcv_wscale))))
|
2005-04-17 06:20:36 +08:00
|
|
|
return tp->snd_nxt;
|
|
|
|
|
else
|
2007-12-31 20:48:41 +08:00
|
|
|
return tcp_wnd_end(tp);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Calculate mss to advertise in SYN segment.
|
|
|
|
|
* RFC1122, RFC1063, draft-ietf-tcpimpl-pmtud-01 state that:
|
|
|
|
|
*
|
|
|
|
|
* 1. It is independent of path mtu.
|
|
|
|
|
* 2. Ideally, it is maximal possible segment size i.e. 65535-40.
|
|
|
|
|
* 3. For IPv4 it is reasonable to calculate it from maximal MTU of
|
|
|
|
|
* attached devices, because some buggy hosts are confused by
|
|
|
|
|
* large MSS.
|
|
|
|
|
* 4. We do not make 3, we advertise MSS, calculated from first
|
|
|
|
|
* hop device mtu, but allow to raise it to ip_rt_min_advmss.
|
|
|
|
|
* This may be overridden via information stored in routing table.
|
|
|
|
|
* 5. Value 65535 for MSS is valid in IPv6 and means "as large as possible,
|
|
|
|
|
* probably even Jumbo".
|
|
|
|
|
*/
|
|
|
|
|
static __u16 tcp_advertise_mss(struct sock *sk)
|
|
|
|
|
{
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
2011-10-21 17:22:42 +08:00
|
|
|
const struct dst_entry *dst = __sk_dst_get(sk);
|
2005-04-17 06:20:36 +08:00
|
|
|
int mss = tp->advmss;
|
|
|
|
|
|
2010-12-14 04:52:14 +08:00
|
|
|
if (dst) {
|
|
|
|
|
unsigned int metric = dst_metric_advmss(dst);
|
|
|
|
|
|
|
|
|
|
if (metric < mss) {
|
|
|
|
|
mss = metric;
|
|
|
|
|
tp->advmss = mss;
|
|
|
|
|
}
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return (__u16)mss;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* RFC2861. Reset CWND after idle period longer RTO to "restart window".
|
tcp: fix slow start after idle vs TSO/GSO
slow start after idle might reduce cwnd, but we perform this
after first packet was cooked and sent.
With TSO/GSO, it means that we might send a full TSO packet
even if cwnd should have been reduced to IW10.
Moving the SSAI check in skb_entail() makes sense, because
we slightly reduce number of times this check is done,
especially for large send() and TCP Small queue callbacks from
softirq context.
As Neal pointed out, we also need to perform the check
if/when receive window opens.
Tested:
Following packetdrill test demonstrates the problem
// Test of slow start after idle
`sysctl -q net.ipv4.tcp_slow_start_after_idle=1`
0.000 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
+0 < S 0:0(0) win 65535 <mss 1000,sackOK,nop,nop,nop,wscale 7>
+0 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 6>
+.100 < . 1:1(0) ack 1 win 511
+0 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_SOCKET, SO_SNDBUF, [200000], 4) = 0
+0 write(4, ..., 26000) = 26000
+0 > . 1:5001(5000) ack 1
+0 > . 5001:10001(5000) ack 1
+0 %{ assert tcpi_snd_cwnd == 10 }%
+.100 < . 1:1(0) ack 10001 win 511
+0 %{ assert tcpi_snd_cwnd == 20, tcpi_snd_cwnd }%
+0 > . 10001:20001(10000) ack 1
+0 > P. 20001:26001(6000) ack 1
+.100 < . 1:1(0) ack 26001 win 511
+0 %{ assert tcpi_snd_cwnd == 36, tcpi_snd_cwnd }%
+4 write(4, ..., 20000) = 20000
// If slow start after idle works properly, we should send 5 MSS here (cwnd/2)
+0 > . 26001:31001(5000) ack 1
+0 %{ assert tcpi_snd_cwnd == 10, tcpi_snd_cwnd }%
+0 > . 31001:36001(5000) ack 1
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-08-22 03:30:00 +08:00
|
|
|
* This is the first part of cwnd validation mechanism.
|
|
|
|
|
*/
|
|
|
|
|
void tcp_cwnd_restart(struct sock *sk, s32 delta)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
2005-08-10 11:10:42 +08:00
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
tcp: fix slow start after idle vs TSO/GSO
slow start after idle might reduce cwnd, but we perform this
after first packet was cooked and sent.
With TSO/GSO, it means that we might send a full TSO packet
even if cwnd should have been reduced to IW10.
Moving the SSAI check in skb_entail() makes sense, because
we slightly reduce number of times this check is done,
especially for large send() and TCP Small queue callbacks from
softirq context.
As Neal pointed out, we also need to perform the check
if/when receive window opens.
Tested:
Following packetdrill test demonstrates the problem
// Test of slow start after idle
`sysctl -q net.ipv4.tcp_slow_start_after_idle=1`
0.000 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
+0 < S 0:0(0) win 65535 <mss 1000,sackOK,nop,nop,nop,wscale 7>
+0 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 6>
+.100 < . 1:1(0) ack 1 win 511
+0 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_SOCKET, SO_SNDBUF, [200000], 4) = 0
+0 write(4, ..., 26000) = 26000
+0 > . 1:5001(5000) ack 1
+0 > . 5001:10001(5000) ack 1
+0 %{ assert tcpi_snd_cwnd == 10 }%
+.100 < . 1:1(0) ack 10001 win 511
+0 %{ assert tcpi_snd_cwnd == 20, tcpi_snd_cwnd }%
+0 > . 10001:20001(10000) ack 1
+0 > P. 20001:26001(6000) ack 1
+.100 < . 1:1(0) ack 26001 win 511
+0 %{ assert tcpi_snd_cwnd == 36, tcpi_snd_cwnd }%
+4 write(4, ..., 20000) = 20000
// If slow start after idle works properly, we should send 5 MSS here (cwnd/2)
+0 > . 26001:31001(5000) ack 1
+0 %{ assert tcpi_snd_cwnd == 10, tcpi_snd_cwnd }%
+0 > . 31001:36001(5000) ack 1
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-08-22 03:30:00 +08:00
|
|
|
u32 restart_cwnd = tcp_init_cwnd(tp, __sk_dst_get(sk));
|
2005-04-17 06:20:36 +08:00
|
|
|
u32 cwnd = tp->snd_cwnd;
|
|
|
|
|
|
2005-08-10 15:03:31 +08:00
|
|
|
tcp_ca_event(sk, CA_EVENT_CWND_RESTART);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2005-08-10 15:03:31 +08:00
|
|
|
tp->snd_ssthresh = tcp_current_ssthresh(sk);
|
2005-04-17 06:20:36 +08:00
|
|
|
restart_cwnd = min(restart_cwnd, cwnd);
|
|
|
|
|
|
2005-08-10 11:10:42 +08:00
|
|
|
while ((delta -= inet_csk(sk)->icsk_rto) > 0 && cwnd > restart_cwnd)
|
2005-04-17 06:20:36 +08:00
|
|
|
cwnd >>= 1;
|
|
|
|
|
tp->snd_cwnd = max(cwnd, restart_cwnd);
|
2017-05-17 05:00:04 +08:00
|
|
|
tp->snd_cwnd_stamp = tcp_jiffies32;
|
2005-04-17 06:20:36 +08:00
|
|
|
tp->snd_cwnd_used = 0;
|
|
|
|
|
}
|
|
|
|
|
|
2009-07-22 07:00:40 +08:00
|
|
|
/* Congestion state accounting after a packet has been sent. */
|
2006-01-04 08:03:49 +08:00
|
|
|
static void tcp_event_data_sent(struct tcp_sock *tp,
|
2011-10-21 17:22:42 +08:00
|
|
|
struct sock *sk)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
2005-08-10 11:10:42 +08:00
|
|
|
struct inet_connection_sock *icsk = inet_csk(sk);
|
2017-05-17 05:00:03 +08:00
|
|
|
const u32 now = tcp_jiffies32;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2015-09-10 12:54:37 +08:00
|
|
|
if (tcp_packets_in_flight(tp) == 0)
|
|
|
|
|
tcp_ca_event(sk, CA_EVENT_TX_START);
|
|
|
|
|
|
2019-01-26 02:53:20 +08:00
|
|
|
/* If this is the first data packet sent in response to the
|
|
|
|
|
* previous received data,
|
|
|
|
|
* and it is a reply for ato after last received packet,
|
|
|
|
|
* increase pingpong count.
|
2005-04-17 06:20:36 +08:00
|
|
|
*/
|
2019-01-26 02:53:20 +08:00
|
|
|
if (before(tp->lsndtime, icsk->icsk_ack.lrcvtime) &&
|
|
|
|
|
(u32)(now - icsk->icsk_ack.lrcvtime) < icsk->icsk_ack.ato)
|
|
|
|
|
inet_csk_inc_pingpong_cnt(sk);
|
|
|
|
|
|
|
|
|
|
tp->lsndtime = now;
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
|
2009-07-22 07:00:40 +08:00
|
|
|
/* Account for an ACK we sent. */
|
2018-07-19 04:56:35 +08:00
|
|
|
static inline void tcp_event_ack_sent(struct sock *sk, unsigned int pkts,
|
|
|
|
|
u32 rcv_nxt)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
tcp: add SACK compression
When TCP receives an out-of-order packet, it immediately sends
a SACK packet, generating network load but also forcing the
receiver to send 1-MSS pathological packets, increasing its
RTX queue length/depth, and thus processing time.
Wifi networks suffer from this aggressive behavior, but generally
speaking, all these SACK packets add fuel to the fire when networks
are under congestion.
This patch adds a high resolution timer and tp->compressed_ack counter.
Instead of sending a SACK, we program this timer with a small delay,
based on RTT and capped to 1 ms :
delay = min ( 5 % of RTT, 1 ms)
If subsequent SACKs need to be sent while the timer has not yet
expired, we simply increment tp->compressed_ack.
When timer expires, a SACK is sent with the latest information.
Whenever an ACK is sent (if data is sent, or if in-order
data is received) timer is canceled.
Note that tcp_sack_new_ofo_skb() is able to force a SACK to be sent
if the sack blocks need to be shuffled, even if the timer has not
expired.
A new SNMP counter is added in the following patch.
Two other patches add sysctls to allow changing the 1,000,000 and 44
values that this commit hard-coded.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Acked-by: Toke Høiland-Jørgensen <toke@toke.dk>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-05-18 05:47:26 +08:00
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
|
2020-05-01 01:35:41 +08:00
|
|
|
if (unlikely(tp->compressed_ack)) {
|
2018-05-18 05:47:27 +08:00
|
|
|
NET_ADD_STATS(sock_net(sk), LINUX_MIB_TCPACKCOMPRESSED,
|
2020-05-01 01:35:41 +08:00
|
|
|
tp->compressed_ack);
|
|
|
|
|
tp->compressed_ack = 0;
|
tcp: add SACK compression
When TCP receives an out-of-order packet, it immediately sends
a SACK packet, generating network load but also forcing the
receiver to send 1-MSS pathological packets, increasing its
RTX queue length/depth, and thus processing time.
Wifi networks suffer from this aggressive behavior, but generally
speaking, all these SACK packets add fuel to the fire when networks
are under congestion.
This patch adds a high resolution timer and tp->compressed_ack counter.
Instead of sending a SACK, we program this timer with a small delay,
based on RTT and capped to 1 ms :
delay = min ( 5 % of RTT, 1 ms)
If subsequent SACKs need to be sent while the timer has not yet
expired, we simply increment tp->compressed_ack.
When timer expires, a SACK is sent with the latest information.
Whenever an ACK is sent (if data is sent, or if in-order
data is received) timer is canceled.
Note that tcp_sack_new_ofo_skb() is able to force a SACK to be sent
if the sack blocks need to be shuffled, even if the timer has not
expired.
A new SNMP counter is added in the following patch.
Two other patches add sysctls to allow changing the 1,000,000 and 44
values that this commit hard-coded.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Acked-by: Toke Høiland-Jørgensen <toke@toke.dk>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-05-18 05:47:26 +08:00
|
|
|
if (hrtimer_try_to_cancel(&tp->compressed_ack_timer) == 1)
|
|
|
|
|
__sock_put(sk);
|
|
|
|
|
}
|
2018-07-19 04:56:35 +08:00
|
|
|
|
|
|
|
|
if (unlikely(rcv_nxt != tp->rcv_nxt))
|
|
|
|
|
return; /* Special ACK sent by DCTCP to reflect ECN */
|
2005-08-10 11:10:42 +08:00
|
|
|
tcp_dec_quickack_mode(sk, pkts);
|
|
|
|
|
inet_csk_clear_xmit_timer(sk, ICSK_TIME_DACK);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Determine a window scaling and initial window to offer.
|
|
|
|
|
* Based on the assumption that the given amount of space
|
|
|
|
|
* will be offered. Store the results in the tp structure.
|
|
|
|
|
* NOTE: for smooth operation initial space offering should
|
|
|
|
|
* be a multiple of mss if possible. We assume here that mss >= 1.
|
|
|
|
|
* This MUST be enforced by all callers.
|
|
|
|
|
*/
|
2017-10-27 22:47:24 +08:00
|
|
|
void tcp_select_initial_window(const struct sock *sk, int __space, __u32 mss,
|
2005-04-17 06:20:36 +08:00
|
|
|
__u32 *rcv_wnd, __u32 *window_clamp,
|
2009-12-15 19:15:28 +08:00
|
|
|
int wscale_ok, __u8 *rcv_wscale,
|
|
|
|
|
__u32 init_rcv_wnd)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
|
|
|
|
unsigned int space = (__space < 0 ? 0 : __space);
|
|
|
|
|
|
|
|
|
|
/* If no clamp set the clamp to the max possible scaled window */
|
|
|
|
|
if (*window_clamp == 0)
|
2017-04-04 21:09:48 +08:00
|
|
|
(*window_clamp) = (U16_MAX << TCP_MAX_WSCALE);
|
2005-04-17 06:20:36 +08:00
|
|
|
space = min(*window_clamp, space);
|
|
|
|
|
|
|
|
|
|
/* Quantize space offering to a multiple of mss if possible. */
|
|
|
|
|
if (space > mss)
|
2017-04-04 21:09:48 +08:00
|
|
|
space = rounddown(space, mss);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
|
/* NOTE: offering an initial window larger than 32767
|
2006-03-21 14:40:29 +08:00
|
|
|
* will break some buggy TCP stacks. If the admin tells us
|
|
|
|
|
* it is likely we could be speaking with such a buggy stack
|
|
|
|
|
* we will truncate our initial window offering to 32K-1
|
|
|
|
|
* unless the remote has sent us a window scaling option,
|
|
|
|
|
* which we interpret as a sign the remote TCP is not
|
|
|
|
|
* misinterpreting the window field as a signed quantity.
|
2005-04-17 06:20:36 +08:00
|
|
|
*/
|
2017-10-27 22:47:24 +08:00
|
|
|
if (sock_net(sk)->ipv4.sysctl_tcp_workaround_signed_windows)
|
2006-03-21 14:40:29 +08:00
|
|
|
(*rcv_wnd) = min(space, MAX_TCP_WINDOW);
|
|
|
|
|
else
|
2018-09-28 02:21:19 +08:00
|
|
|
(*rcv_wnd) = min_t(u32, space, U16_MAX);
|
|
|
|
|
|
|
|
|
|
if (init_rcv_wnd)
|
|
|
|
|
*rcv_wnd = min(*rcv_wnd, init_rcv_wnd * mss);
|
2006-03-21 14:40:29 +08:00
|
|
|
|
2018-11-29 23:56:20 +08:00
|
|
|
*rcv_wscale = 0;
|
2005-04-17 06:20:36 +08:00
|
|
|
if (wscale_ok) {
|
2017-04-04 21:09:48 +08:00
|
|
|
/* Set window scaling on max possible window */
|
2017-11-07 16:29:28 +08:00
|
|
|
space = max_t(u32, space, sock_net(sk)->ipv4.sysctl_tcp_rmem[2]);
|
2016-07-29 21:34:02 +08:00
|
|
|
space = max_t(u32, space, sysctl_rmem_max);
|
2006-08-22 15:06:11 +08:00
|
|
|
space = min_t(u32, space, *window_clamp);
|
2018-11-29 23:56:20 +08:00
|
|
|
*rcv_wscale = clamp_t(int, ilog2(space) - 15,
|
|
|
|
|
0, TCP_MAX_WSCALE);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
/* Set the clamp no higher than max representable value */
|
2017-04-04 21:09:48 +08:00
|
|
|
(*window_clamp) = min_t(__u32, U16_MAX << (*rcv_wscale), *window_clamp);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
2010-07-10 05:22:10 +08:00
|
|
|
EXPORT_SYMBOL(tcp_select_initial_window);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
|
/* Chose a new window to advertise, update state in tcp_sock for the
|
|
|
|
|
* socket, and return result with RFC1323 scaling applied. The return
|
|
|
|
|
* value can be stuffed directly into th->window for an outgoing
|
|
|
|
|
* frame.
|
|
|
|
|
*/
|
2006-01-04 08:03:49 +08:00
|
|
|
static u16 tcp_select_window(struct sock *sk)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
2014-02-25 21:34:32 +08:00
|
|
|
u32 old_win = tp->rcv_wnd;
|
2005-04-17 06:20:36 +08:00
|
|
|
u32 cur_win = tcp_receive_window(tp);
|
|
|
|
|
u32 new_win = __tcp_select_window(sk);
|
|
|
|
|
|
|
|
|
|
/* Never shrink the offered window */
|
2007-03-09 12:45:19 +08:00
|
|
|
if (new_win < cur_win) {
|
2005-04-17 06:20:36 +08:00
|
|
|
/* Danger Will Robinson!
|
|
|
|
|
* Don't update rcv_wup/rcv_wnd here or else
|
|
|
|
|
* we will not be able to advertise a zero
|
|
|
|
|
* window in time. --DaveM
|
|
|
|
|
*
|
|
|
|
|
* Relax Will Robinson.
|
|
|
|
|
*/
|
2014-02-25 21:34:32 +08:00
|
|
|
if (new_win == 0)
|
|
|
|
|
NET_INC_STATS(sock_net(sk),
|
|
|
|
|
LINUX_MIB_TCPWANTZEROWINDOWADV);
|
[TCP]: Fix shrinking windows with window scaling
When selecting a new window, tcp_select_window() tries not to shrink
the offered window by using the maximum of the remaining offered window
size and the newly calculated window size. The newly calculated window
size is always a multiple of the window scaling factor, the remaining
window size however might not be since it depends on rcv_wup/rcv_nxt.
This means we're effectively shrinking the window when scaling it down.
The dump below shows the problem (scaling factor 2^7):
- Window size of 557 (71296) is advertised, up to 3111907257:
IP 172.2.2.3.33000 > 172.2.2.2.33000: . ack 3111835961 win 557 <...>
- New window size of 514 (65792) is advertised, up to 3111907217, 40 bytes
below the last end:
IP 172.2.2.3.33000 > 172.2.2.2.33000: . 3113575668:3113577116(1448) ack 3111841425 win 514 <...>
The number 40 results from downscaling the remaining window:
3111907257 - 3111841425 = 65832
65832 / 2^7 = 514
65832 % 2^7 = 40
If the sender uses up the entire window before it is shrunk, this can have
chaotic effects on the connection. When sending ACKs, tcp_acceptable_seq()
will notice that the window has been shrunk since tcp_wnd_end() is before
tp->snd_nxt, which makes it choose tcp_wnd_end() as sequence number.
This will fail the receivers checks in tcp_sequence() however since it
is before it's tp->rcv_wup, making it respond with a dupack.
If both sides are in this condition, this leads to a constant flood of
ACKs until the connection times out.
Make sure the window is never shrunk by aligning the remaining window to
the window scaling factor.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-21 07:11:27 +08:00
|
|
|
new_win = ALIGN(cur_win, 1 << tp->rx_opt.rcv_wscale);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
tp->rcv_wnd = new_win;
|
|
|
|
|
tp->rcv_wup = tp->rcv_nxt;
|
|
|
|
|
|
|
|
|
|
/* Make sure we do not exceed the maximum possible
|
|
|
|
|
* scaled window.
|
|
|
|
|
*/
|
2017-10-27 22:47:24 +08:00
|
|
|
if (!tp->rx_opt.rcv_wscale &&
|
|
|
|
|
sock_net(sk)->ipv4.sysctl_tcp_workaround_signed_windows)
|
2005-04-17 06:20:36 +08:00
|
|
|
new_win = min(new_win, MAX_TCP_WINDOW);
|
|
|
|
|
else
|
|
|
|
|
new_win = min(new_win, (65535U << tp->rx_opt.rcv_wscale));
|
|
|
|
|
|
|
|
|
|
/* RFC1323 scaling applied */
|
|
|
|
|
new_win >>= tp->rx_opt.rcv_wscale;
|
|
|
|
|
|
2017-08-31 01:24:58 +08:00
|
|
|
/* If we advertise zero window, disable fast path. */
|
2014-02-25 21:34:32 +08:00
|
|
|
if (new_win == 0) {
|
2017-08-31 01:24:58 +08:00
|
|
|
tp->pred_flags = 0;
|
2014-02-25 21:34:32 +08:00
|
|
|
if (old_win)
|
|
|
|
|
NET_INC_STATS(sock_net(sk),
|
|
|
|
|
LINUX_MIB_TCPTOZEROWINDOWADV);
|
|
|
|
|
} else if (old_win == 0) {
|
|
|
|
|
NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPFROMZEROWINDOWADV);
|
|
|
|
|
}
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
|
return new_win;
|
|
|
|
|
}
|
|
|
|
|
|
2009-07-22 07:00:40 +08:00
|
|
|
/* Packet ECN state for a SYN-ACK */
|
2014-09-29 19:08:30 +08:00
|
|
|
static void tcp_ecn_send_synack(struct sock *sk, struct sk_buff *skb)
|
2007-05-27 17:04:16 +08:00
|
|
|
{
|
2014-09-27 04:37:33 +08:00
|
|
|
const struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
|
2011-09-28 01:25:05 +08:00
|
|
|
TCP_SKB_CB(skb)->tcp_flags &= ~TCPHDR_CWR;
|
2008-01-01 06:57:14 +08:00
|
|
|
if (!(tp->ecn_flags & TCP_ECN_OK))
|
2011-09-28 01:25:05 +08:00
|
|
|
TCP_SKB_CB(skb)->tcp_flags &= ~TCPHDR_ECE;
|
2017-07-01 11:02:49 +08:00
|
|
|
else if (tcp_ca_needs_ecn(sk) ||
|
|
|
|
|
tcp_bpf_ca_needs_ecn(sk))
|
2014-09-27 04:37:33 +08:00
|
|
|
INET_ECN_xmit(sk);
|
2007-05-27 17:04:16 +08:00
|
|
|
}
|
|
|
|
|
|
2009-07-22 07:00:40 +08:00
|
|
|
/* Packet ECN state for a SYN. */
|
2014-09-29 19:08:30 +08:00
|
|
|
static void tcp_ecn_send_syn(struct sock *sk, struct sk_buff *skb)
|
2007-05-27 17:04:16 +08:00
|
|
|
{
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
2017-07-01 11:02:49 +08:00
|
|
|
bool bpf_needs_ecn = tcp_bpf_ca_needs_ecn(sk);
|
net: allow setting ecn via routing table
This patch allows to set ECN on a per-route basis in case the sysctl
tcp_ecn is not set to 1. In other words, when ECN is set for specific
routes, it provides a tcp_ecn=1 behaviour for that route while the rest
of the stack acts according to the global settings.
One can use 'ip route change dev $dev $net features ecn' to toggle this.
Having a more fine-grained per-route setting can be beneficial for various
reasons, for example, 1) within data centers, or 2) local ISPs may deploy
ECN support for their own video/streaming services [1], etc.
There was a recent measurement study/paper [2] which scanned the Alexa's
publicly available top million websites list from a vantage point in US,
Europe and Asia:
Half of the Alexa list will now happily use ECN (tcp_ecn=2, most likely
blamed to commit 255cac91c3 ("tcp: extend ECN sysctl to allow server-side
only ECN") ;)); the break in connectivity on-path was found is about
1 in 10,000 cases. Timeouts rather than receiving back RSTs were much
more common in the negotiation phase (and mostly seen in the Alexa
middle band, ranks around 50k-150k): from 12-thousand hosts on which
there _may_ be ECN-linked connection failures, only 79 failed with RST
when _not_ failing with RST when ECN is not requested.
It's unclear though, how much equipment in the wild actually marks CE
when buffers start to fill up.
We thought about a fallback to non-ECN for retransmitted SYNs as another
global option (which could perhaps one day be made default), but as Eric
points out, there's much more work needed to detect broken middleboxes.
Two examples Eric mentioned are buggy firewalls that accept only a single
SYN per flow, and middleboxes that successfully let an ECN flow establish,
but later mark CE for all packets (so cwnd converges to 1).
[1] http://www.ietf.org/proceedings/89/slides/slides-89-tsvarea-1.pdf, p.15
[2] http://ecn.ethz.ch/
Joint work with Daniel Borkmann.
Reference: http://thread.gmane.org/gmane.linux.network/335797
Suggested-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Daniel Borkmann <dborkman@redhat.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-11-04 00:35:03 +08:00
|
|
|
bool use_ecn = sock_net(sk)->ipv4.sysctl_tcp_ecn == 1 ||
|
2017-07-01 11:02:49 +08:00
|
|
|
tcp_ca_needs_ecn(sk) || bpf_needs_ecn;
|
net: allow setting ecn via routing table
This patch allows to set ECN on a per-route basis in case the sysctl
tcp_ecn is not set to 1. In other words, when ECN is set for specific
routes, it provides a tcp_ecn=1 behaviour for that route while the rest
of the stack acts according to the global settings.
One can use 'ip route change dev $dev $net features ecn' to toggle this.
Having a more fine-grained per-route setting can be beneficial for various
reasons, for example, 1) within data centers, or 2) local ISPs may deploy
ECN support for their own video/streaming services [1], etc.
There was a recent measurement study/paper [2] which scanned the Alexa's
publicly available top million websites list from a vantage point in US,
Europe and Asia:
Half of the Alexa list will now happily use ECN (tcp_ecn=2, most likely
blamed to commit 255cac91c3 ("tcp: extend ECN sysctl to allow server-side
only ECN") ;)); the break in connectivity on-path was found is about
1 in 10,000 cases. Timeouts rather than receiving back RSTs were much
more common in the negotiation phase (and mostly seen in the Alexa
middle band, ranks around 50k-150k): from 12-thousand hosts on which
there _may_ be ECN-linked connection failures, only 79 failed with RST
when _not_ failing with RST when ECN is not requested.
It's unclear though, how much equipment in the wild actually marks CE
when buffers start to fill up.
We thought about a fallback to non-ECN for retransmitted SYNs as another
global option (which could perhaps one day be made default), but as Eric
points out, there's much more work needed to detect broken middleboxes.
Two examples Eric mentioned are buggy firewalls that accept only a single
SYN per flow, and middleboxes that successfully let an ECN flow establish,
but later mark CE for all packets (so cwnd converges to 1).
[1] http://www.ietf.org/proceedings/89/slides/slides-89-tsvarea-1.pdf, p.15
[2] http://ecn.ethz.ch/
Joint work with Daniel Borkmann.
Reference: http://thread.gmane.org/gmane.linux.network/335797
Suggested-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Daniel Borkmann <dborkman@redhat.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-11-04 00:35:03 +08:00
|
|
|
|
|
|
|
|
if (!use_ecn) {
|
|
|
|
|
const struct dst_entry *dst = __sk_dst_get(sk);
|
|
|
|
|
|
|
|
|
|
if (dst && dst_feature(dst, RTAX_FEATURE_ECN))
|
|
|
|
|
use_ecn = true;
|
|
|
|
|
}
|
2007-05-27 17:04:16 +08:00
|
|
|
|
|
|
|
|
tp->ecn_flags = 0;
|
net: allow setting ecn via routing table
This patch allows to set ECN on a per-route basis in case the sysctl
tcp_ecn is not set to 1. In other words, when ECN is set for specific
routes, it provides a tcp_ecn=1 behaviour for that route while the rest
of the stack acts according to the global settings.
One can use 'ip route change dev $dev $net features ecn' to toggle this.
Having a more fine-grained per-route setting can be beneficial for various
reasons, for example, 1) within data centers, or 2) local ISPs may deploy
ECN support for their own video/streaming services [1], etc.
There was a recent measurement study/paper [2] which scanned the Alexa's
publicly available top million websites list from a vantage point in US,
Europe and Asia:
Half of the Alexa list will now happily use ECN (tcp_ecn=2, most likely
blamed to commit 255cac91c3 ("tcp: extend ECN sysctl to allow server-side
only ECN") ;)); the break in connectivity on-path was found is about
1 in 10,000 cases. Timeouts rather than receiving back RSTs were much
more common in the negotiation phase (and mostly seen in the Alexa
middle band, ranks around 50k-150k): from 12-thousand hosts on which
there _may_ be ECN-linked connection failures, only 79 failed with RST
when _not_ failing with RST when ECN is not requested.
It's unclear though, how much equipment in the wild actually marks CE
when buffers start to fill up.
We thought about a fallback to non-ECN for retransmitted SYNs as another
global option (which could perhaps one day be made default), but as Eric
points out, there's much more work needed to detect broken middleboxes.
Two examples Eric mentioned are buggy firewalls that accept only a single
SYN per flow, and middleboxes that successfully let an ECN flow establish,
but later mark CE for all packets (so cwnd converges to 1).
[1] http://www.ietf.org/proceedings/89/slides/slides-89-tsvarea-1.pdf, p.15
[2] http://ecn.ethz.ch/
Joint work with Daniel Borkmann.
Reference: http://thread.gmane.org/gmane.linux.network/335797
Suggested-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Daniel Borkmann <dborkman@redhat.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-11-04 00:35:03 +08:00
|
|
|
|
|
|
|
|
if (use_ecn) {
|
2011-09-28 01:25:05 +08:00
|
|
|
TCP_SKB_CB(skb)->tcp_flags |= TCPHDR_ECE | TCPHDR_CWR;
|
2007-05-27 17:04:16 +08:00
|
|
|
tp->ecn_flags = TCP_ECN_OK;
|
2017-07-01 11:02:49 +08:00
|
|
|
if (tcp_ca_needs_ecn(sk) || bpf_needs_ecn)
|
2014-09-27 04:37:33 +08:00
|
|
|
INET_ECN_xmit(sk);
|
2007-05-27 17:04:16 +08:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
tcp: add rfc3168, section 6.1.1.1. fallback
This work as a follow-up of commit f7b3bec6f516 ("net: allow setting ecn
via routing table") and adds RFC3168 section 6.1.1.1. fallback for outgoing
ECN connections. In other words, this work adds a retry with a non-ECN
setup SYN packet, as suggested from the RFC on the first timeout:
[...] A host that receives no reply to an ECN-setup SYN within the
normal SYN retransmission timeout interval MAY resend the SYN and
any subsequent SYN retransmissions with CWR and ECE cleared. [...]
Schematic client-side view when assuming the server is in tcp_ecn=2 mode,
that is, Linux default since 2009 via commit 255cac91c3c9 ("tcp: extend
ECN sysctl to allow server-side only ECN"):
1) Normal ECN-capable path:
SYN ECE CWR ----->
<----- SYN ACK ECE
ACK ----->
2) Path with broken middlebox, when client has fallback:
SYN ECE CWR ----X crappy middlebox drops packet
(timeout, rtx)
SYN ----->
<----- SYN ACK
ACK ----->
In case we would not have the fallback implemented, the middlebox drop
point would basically end up as:
SYN ECE CWR ----X crappy middlebox drops packet
(timeout, rtx)
SYN ECE CWR ----X crappy middlebox drops packet
(timeout, rtx)
SYN ECE CWR ----X crappy middlebox drops packet
(timeout, rtx)
In any case, it's rather a smaller percentage of sites where there would
occur such additional setup latency: it was found in end of 2014 that ~56%
of IPv4 and 65% of IPv6 servers of Alexa 1 million list would negotiate
ECN (aka tcp_ecn=2 default), 0.42% of these webservers will fail to connect
when trying to negotiate with ECN (tcp_ecn=1) due to timeouts, which the
fallback would mitigate with a slight latency trade-off. Recent related
paper on this topic:
Brian Trammell, Mirja Kühlewind, Damiano Boppart, Iain Learmonth,
Gorry Fairhurst, and Richard Scheffenegger:
"Enabling Internet-Wide Deployment of Explicit Congestion Notification."
Proc. PAM 2015, New York.
http://ecn.ethz.ch/ecn-pam15.pdf
Thus, when net.ipv4.tcp_ecn=1 is being set, the patch will perform RFC3168,
section 6.1.1.1. fallback on timeout. For users explicitly not wanting this
which can be in DC use case, we add a net.ipv4.tcp_ecn_fallback knob that
allows for disabling the fallback.
tp->ecn_flags are not being cleared in tcp_ecn_clear_syn() on output, but
rather we let tcp_ecn_rcv_synack() take that over on input path in case a
SYN ACK ECE was delayed. Thus a spurious SYN retransmission will not prevent
ECN being negotiated eventually in that case.
Reference: https://www.ietf.org/proceedings/92/slides/slides-92-iccrg-1.pdf
Reference: https://www.ietf.org/proceedings/89/slides/slides-89-tsvarea-1.pdf
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Mirja Kühlewind <mirja.kuehlewind@tik.ee.ethz.ch>
Signed-off-by: Brian Trammell <trammell@tik.ee.ethz.ch>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Dave That <dave.taht@gmail.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-05-20 03:04:22 +08:00
|
|
|
static void tcp_ecn_clear_syn(struct sock *sk, struct sk_buff *skb)
|
|
|
|
|
{
|
|
|
|
|
if (sock_net(sk)->ipv4.sysctl_tcp_ecn_fallback)
|
|
|
|
|
/* tp->ecn_flags are cleared at a later point in time when
|
|
|
|
|
* SYN ACK is ultimatively being received.
|
|
|
|
|
*/
|
|
|
|
|
TCP_SKB_CB(skb)->tcp_flags &= ~(TCPHDR_ECE | TCPHDR_CWR);
|
|
|
|
|
}
|
|
|
|
|
|
2014-09-29 19:08:30 +08:00
|
|
|
static void
|
2015-09-25 22:39:18 +08:00
|
|
|
tcp_ecn_make_synack(const struct request_sock *req, struct tcphdr *th)
|
2007-05-27 17:04:16 +08:00
|
|
|
{
|
2015-09-25 22:39:18 +08:00
|
|
|
if (inet_rsk(req)->ecn_ok)
|
2007-05-27 17:04:16 +08:00
|
|
|
th->ece = 1;
|
|
|
|
|
}
|
|
|
|
|
|
2009-07-22 07:00:40 +08:00
|
|
|
/* Set up ECN state for a packet on a ESTABLISHED socket that is about to
|
|
|
|
|
* be sent.
|
|
|
|
|
*/
|
2014-09-29 19:08:30 +08:00
|
|
|
static void tcp_ecn_send(struct sock *sk, struct sk_buff *skb,
|
2016-05-14 00:16:40 +08:00
|
|
|
struct tcphdr *th, int tcp_header_len)
|
2007-05-27 17:04:16 +08:00
|
|
|
{
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
|
|
|
|
|
if (tp->ecn_flags & TCP_ECN_OK) {
|
|
|
|
|
/* Not-retransmitted data segment: set ECT and inject CWR. */
|
|
|
|
|
if (skb->len != tcp_header_len &&
|
|
|
|
|
!before(TCP_SKB_CB(skb)->seq, tp->snd_nxt)) {
|
|
|
|
|
INET_ECN_xmit(sk);
|
2008-01-01 06:57:14 +08:00
|
|
|
if (tp->ecn_flags & TCP_ECN_QUEUE_CWR) {
|
2007-05-27 17:04:16 +08:00
|
|
|
tp->ecn_flags &= ~TCP_ECN_QUEUE_CWR;
|
2016-05-14 00:16:40 +08:00
|
|
|
th->cwr = 1;
|
2007-05-27 17:04:16 +08:00
|
|
|
skb_shinfo(skb)->gso_type |= SKB_GSO_TCP_ECN;
|
|
|
|
|
}
|
2014-09-27 04:37:33 +08:00
|
|
|
} else if (!tcp_ca_needs_ecn(sk)) {
|
2007-05-27 17:04:16 +08:00
|
|
|
/* ACK or retransmitted segment: clear ECT|CE */
|
|
|
|
|
INET_ECN_dontxmit(sk);
|
|
|
|
|
}
|
|
|
|
|
if (tp->ecn_flags & TCP_ECN_DEMAND_CWR)
|
2016-05-14 00:16:40 +08:00
|
|
|
th->ece = 1;
|
2007-05-27 17:04:16 +08:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2008-01-04 12:39:01 +08:00
|
|
|
/* Constructs common control bits of non-data skb. If SYN/FIN is present,
|
|
|
|
|
* auto increment end seqno.
|
|
|
|
|
*/
|
|
|
|
|
static void tcp_init_nondata_skb(struct sk_buff *skb, u32 seq, u8 flags)
|
|
|
|
|
{
|
2010-04-09 02:32:30 +08:00
|
|
|
skb->ip_summed = CHECKSUM_PARTIAL;
|
2008-01-04 12:39:01 +08:00
|
|
|
|
2011-09-28 01:25:05 +08:00
|
|
|
TCP_SKB_CB(skb)->tcp_flags = flags;
|
2008-01-04 12:39:01 +08:00
|
|
|
TCP_SKB_CB(skb)->sacked = 0;
|
|
|
|
|
|
2014-09-24 19:11:22 +08:00
|
|
|
tcp_skb_pcount_set(skb, 1);
|
2008-01-04 12:39:01 +08:00
|
|
|
|
|
|
|
|
TCP_SKB_CB(skb)->seq = seq;
|
2010-06-12 22:01:43 +08:00
|
|
|
if (flags & (TCPHDR_SYN | TCPHDR_FIN))
|
2008-01-04 12:39:01 +08:00
|
|
|
seq++;
|
|
|
|
|
TCP_SKB_CB(skb)->end_seq = seq;
|
|
|
|
|
}
|
|
|
|
|
|
2012-05-17 07:15:34 +08:00
|
|
|
static inline bool tcp_urg_mode(const struct tcp_sock *tp)
|
2008-10-08 05:43:06 +08:00
|
|
|
{
|
|
|
|
|
return tp->snd_una != tp->snd_up;
|
|
|
|
|
}
|
|
|
|
|
|
2008-07-19 15:04:31 +08:00
|
|
|
#define OPTION_SACK_ADVERTISE (1 << 0)
|
|
|
|
|
#define OPTION_TS (1 << 1)
|
|
|
|
|
#define OPTION_MD5 (1 << 2)
|
2009-10-01 14:41:59 +08:00
|
|
|
#define OPTION_WSCALE (1 << 3)
|
2012-07-19 14:43:05 +08:00
|
|
|
#define OPTION_FAST_OPEN_COOKIE (1 << 8)
|
2017-10-25 17:01:45 +08:00
|
|
|
#define OPTION_SMC (1 << 9)
|
2020-01-22 08:56:16 +08:00
|
|
|
#define OPTION_MPTCP (1 << 10)
|
2017-10-25 17:01:45 +08:00
|
|
|
|
|
|
|
|
static void smc_options_write(__be32 *ptr, u16 *options)
|
|
|
|
|
{
|
|
|
|
|
#if IS_ENABLED(CONFIG_SMC)
|
|
|
|
|
if (static_branch_unlikely(&tcp_have_smc)) {
|
|
|
|
|
if (unlikely(OPTION_SMC & *options)) {
|
|
|
|
|
*ptr++ = htonl((TCPOPT_NOP << 24) |
|
|
|
|
|
(TCPOPT_NOP << 16) |
|
|
|
|
|
(TCPOPT_EXP << 8) |
|
|
|
|
|
(TCPOLEN_EXP_SMC_BASE));
|
|
|
|
|
*ptr++ = htonl(TCPOPT_SMC_MAGIC);
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
}
|
2008-07-19 15:04:31 +08:00
|
|
|
|
|
|
|
|
struct tcp_out_options {
|
2012-07-19 14:43:05 +08:00
|
|
|
u16 options; /* bit field of OPTION_* */
|
|
|
|
|
u16 mss; /* 0 to disable */
|
2008-07-19 15:04:31 +08:00
|
|
|
u8 ws; /* window scale, 0 to disable */
|
|
|
|
|
u8 num_sack_blocks; /* number of SACK blocks to include */
|
2009-12-03 02:23:05 +08:00
|
|
|
u8 hash_size; /* bytes in hash_location */
|
2020-08-21 03:00:52 +08:00
|
|
|
u8 bpf_opt_len; /* length of BPF hdr option */
|
2009-12-03 02:23:05 +08:00
|
|
|
__u8 *hash_location; /* temporary pointer, overloaded */
|
2012-07-19 14:43:05 +08:00
|
|
|
__u32 tsval, tsecr; /* need to include OPTION_TS */
|
|
|
|
|
struct tcp_fastopen_cookie *fastopen_cookie; /* Fast open cookie */
|
2020-01-22 08:56:16 +08:00
|
|
|
struct mptcp_out_options mptcp;
|
2008-07-19 15:04:31 +08:00
|
|
|
};
|
|
|
|
|
|
2020-11-20 03:46:02 +08:00
|
|
|
static void mptcp_options_write(__be32 *ptr, const struct tcp_sock *tp,
|
|
|
|
|
struct tcp_out_options *opts)
|
2020-01-22 08:56:16 +08:00
|
|
|
{
|
|
|
|
|
#if IS_ENABLED(CONFIG_MPTCP)
|
|
|
|
|
if (unlikely(OPTION_MPTCP & opts->options))
|
2020-11-20 03:46:02 +08:00
|
|
|
mptcp_write_options(ptr, tp, &opts->mptcp);
|
2020-01-22 08:56:16 +08:00
|
|
|
#endif
|
|
|
|
|
}
|
|
|
|
|
|
2020-08-21 03:00:52 +08:00
|
|
|
#ifdef CONFIG_CGROUP_BPF
|
bpf: tcp: Allow bpf prog to write and parse TCP header option
[ Note: The TCP changes here is mainly to implement the bpf
pieces into the bpf_skops_*() functions introduced
in the earlier patches. ]
The earlier effort in BPF-TCP-CC allows the TCP Congestion Control
algorithm to be written in BPF. It opens up opportunities to allow
a faster turnaround time in testing/releasing new congestion control
ideas to production environment.
The same flexibility can be extended to writing TCP header option.
It is not uncommon that people want to test new TCP header option
to improve the TCP performance. Another use case is for data-center
that has a more controlled environment and has more flexibility in
putting header options for internal only use.
For example, we want to test the idea in putting maximum delay
ACK in TCP header option which is similar to a draft RFC proposal [1].
This patch introduces the necessary BPF API and use them in the
TCP stack to allow BPF_PROG_TYPE_SOCK_OPS program to parse
and write TCP header options. It currently supports most of
the TCP packet except RST.
Supported TCP header option:
───────────────────────────
This patch allows the bpf-prog to write any option kind.
Different bpf-progs can write its own option by calling the new helper
bpf_store_hdr_opt(). The helper will ensure there is no duplicated
option in the header.
By allowing bpf-prog to write any option kind, this gives a lot of
flexibility to the bpf-prog. Different bpf-prog can write its
own option kind. It could also allow the bpf-prog to support a
recently standardized option on an older kernel.
Sockops Callback Flags:
──────────────────────
The bpf program will only be called to parse/write tcp header option
if the following newly added callback flags are enabled
in tp->bpf_sock_ops_cb_flags:
BPF_SOCK_OPS_PARSE_UNKNOWN_HDR_OPT_CB_FLAG
BPF_SOCK_OPS_PARSE_ALL_HDR_OPT_CB_FLAG
BPF_SOCK_OPS_WRITE_HDR_OPT_CB_FLAG
A few words on the PARSE CB flags. When the above PARSE CB flags are
turned on, the bpf-prog will be called on packets received
at a sk that has at least reached the ESTABLISHED state.
The parsing of the SYN-SYNACK-ACK will be discussed in the
"3 Way HandShake" section.
The default is off for all of the above new CB flags, i.e. the bpf prog
will not be called to parse or write bpf hdr option. There are
details comment on these new cb flags in the UAPI bpf.h.
sock_ops->skb_data and bpf_load_hdr_opt()
─────────────────────────────────────────
sock_ops->skb_data and sock_ops->skb_data_end covers the whole
TCP header and its options. They are read only.
The new bpf_load_hdr_opt() helps to read a particular option "kind"
from the skb_data.
Please refer to the comment in UAPI bpf.h. It has details
on what skb_data contains under different sock_ops->op.
3 Way HandShake
───────────────
The bpf-prog can learn if it is sending SYN or SYNACK by reading the
sock_ops->skb_tcp_flags.
* Passive side
When writing SYNACK (i.e. sock_ops->op == BPF_SOCK_OPS_WRITE_HDR_OPT_CB),
the received SYN skb will be available to the bpf prog. The bpf prog can
use the SYN skb (which may carry the header option sent from the remote bpf
prog) to decide what bpf header option should be written to the outgoing
SYNACK skb. The SYN packet can be obtained by getsockopt(TCP_BPF_SYN*).
More on this later. Also, the bpf prog can learn if it is in syncookie
mode (by checking sock_ops->args[0] == BPF_WRITE_HDR_TCP_SYNACK_COOKIE).
The bpf prog can store the received SYN pkt by using the existing
bpf_setsockopt(TCP_SAVE_SYN). The example in a later patch does it.
[ Note that the fullsock here is a listen sk, bpf_sk_storage
is not very useful here since the listen sk will be shared
by many concurrent connection requests.
Extending bpf_sk_storage support to request_sock will add weight
to the minisock and it is not necessary better than storing the
whole ~100 bytes SYN pkt. ]
When the connection is established, the bpf prog will be called
in the existing PASSIVE_ESTABLISHED_CB callback. At that time,
the bpf prog can get the header option from the saved syn and
then apply the needed operation to the newly established socket.
The later patch will use the max delay ack specified in the SYN
header and set the RTO of this newly established connection
as an example.
The received ACK (that concludes the 3WHS) will also be available to
the bpf prog during PASSIVE_ESTABLISHED_CB through the sock_ops->skb_data.
It could be useful in syncookie scenario. More on this later.
There is an existing getsockopt "TCP_SAVED_SYN" to return the whole
saved syn pkt which includes the IP[46] header and the TCP header.
A few "TCP_BPF_SYN*" getsockopt has been added to allow specifying where to
start getting from, e.g. starting from TCP header, or from IP[46] header.
The new getsockopt(TCP_BPF_SYN*) will also know where it can get
the SYN's packet from:
- (a) the just received syn (available when the bpf prog is writing SYNACK)
and it is the only way to get SYN during syncookie mode.
or
- (b) the saved syn (available in PASSIVE_ESTABLISHED_CB and also other
existing CB).
The bpf prog does not need to know where the SYN pkt is coming from.
The getsockopt(TCP_BPF_SYN*) will hide this details.
Similarly, a flags "BPF_LOAD_HDR_OPT_TCP_SYN" is also added to
bpf_load_hdr_opt() to read a particular header option from the SYN packet.
* Fastopen
Fastopen should work the same as the regular non fastopen case.
This is a test in a later patch.
* Syncookie
For syncookie, the later example patch asks the active
side's bpf prog to resend the header options in ACK. The server
can use bpf_load_hdr_opt() to look at the options in this
received ACK during PASSIVE_ESTABLISHED_CB.
* Active side
The bpf prog will get a chance to write the bpf header option
in the SYN packet during WRITE_HDR_OPT_CB. The received SYNACK
pkt will also be available to the bpf prog during the existing
ACTIVE_ESTABLISHED_CB callback through the sock_ops->skb_data
and bpf_load_hdr_opt().
* Turn off header CB flags after 3WHS
If the bpf prog does not need to write/parse header options
beyond the 3WHS, the bpf prog can clear the bpf_sock_ops_cb_flags
to avoid being called for header options.
Or the bpf-prog can select to leave the UNKNOWN_HDR_OPT_CB_FLAG on
so that the kernel will only call it when there is option that
the kernel cannot handle.
[1]: draft-wang-tcpm-low-latency-opt-00
https://tools.ietf.org/html/draft-wang-tcpm-low-latency-opt-00
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200820190104.2885895-1-kafai@fb.com
2020-08-21 03:01:04 +08:00
|
|
|
static int bpf_skops_write_hdr_opt_arg0(struct sk_buff *skb,
|
|
|
|
|
enum tcp_synack_type synack_type)
|
|
|
|
|
{
|
|
|
|
|
if (unlikely(!skb))
|
|
|
|
|
return BPF_WRITE_HDR_TCP_CURRENT_MSS;
|
|
|
|
|
|
|
|
|
|
if (unlikely(synack_type == TCP_SYNACK_COOKIE))
|
|
|
|
|
return BPF_WRITE_HDR_TCP_SYNACK_COOKIE;
|
|
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
2020-08-21 03:00:52 +08:00
|
|
|
/* req, syn_skb and synack_type are used when writing synack */
|
|
|
|
|
static void bpf_skops_hdr_opt_len(struct sock *sk, struct sk_buff *skb,
|
|
|
|
|
struct request_sock *req,
|
|
|
|
|
struct sk_buff *syn_skb,
|
|
|
|
|
enum tcp_synack_type synack_type,
|
|
|
|
|
struct tcp_out_options *opts,
|
|
|
|
|
unsigned int *remaining)
|
|
|
|
|
{
|
bpf: tcp: Allow bpf prog to write and parse TCP header option
[ Note: The TCP changes here is mainly to implement the bpf
pieces into the bpf_skops_*() functions introduced
in the earlier patches. ]
The earlier effort in BPF-TCP-CC allows the TCP Congestion Control
algorithm to be written in BPF. It opens up opportunities to allow
a faster turnaround time in testing/releasing new congestion control
ideas to production environment.
The same flexibility can be extended to writing TCP header option.
It is not uncommon that people want to test new TCP header option
to improve the TCP performance. Another use case is for data-center
that has a more controlled environment and has more flexibility in
putting header options for internal only use.
For example, we want to test the idea in putting maximum delay
ACK in TCP header option which is similar to a draft RFC proposal [1].
This patch introduces the necessary BPF API and use them in the
TCP stack to allow BPF_PROG_TYPE_SOCK_OPS program to parse
and write TCP header options. It currently supports most of
the TCP packet except RST.
Supported TCP header option:
───────────────────────────
This patch allows the bpf-prog to write any option kind.
Different bpf-progs can write its own option by calling the new helper
bpf_store_hdr_opt(). The helper will ensure there is no duplicated
option in the header.
By allowing bpf-prog to write any option kind, this gives a lot of
flexibility to the bpf-prog. Different bpf-prog can write its
own option kind. It could also allow the bpf-prog to support a
recently standardized option on an older kernel.
Sockops Callback Flags:
──────────────────────
The bpf program will only be called to parse/write tcp header option
if the following newly added callback flags are enabled
in tp->bpf_sock_ops_cb_flags:
BPF_SOCK_OPS_PARSE_UNKNOWN_HDR_OPT_CB_FLAG
BPF_SOCK_OPS_PARSE_ALL_HDR_OPT_CB_FLAG
BPF_SOCK_OPS_WRITE_HDR_OPT_CB_FLAG
A few words on the PARSE CB flags. When the above PARSE CB flags are
turned on, the bpf-prog will be called on packets received
at a sk that has at least reached the ESTABLISHED state.
The parsing of the SYN-SYNACK-ACK will be discussed in the
"3 Way HandShake" section.
The default is off for all of the above new CB flags, i.e. the bpf prog
will not be called to parse or write bpf hdr option. There are
details comment on these new cb flags in the UAPI bpf.h.
sock_ops->skb_data and bpf_load_hdr_opt()
─────────────────────────────────────────
sock_ops->skb_data and sock_ops->skb_data_end covers the whole
TCP header and its options. They are read only.
The new bpf_load_hdr_opt() helps to read a particular option "kind"
from the skb_data.
Please refer to the comment in UAPI bpf.h. It has details
on what skb_data contains under different sock_ops->op.
3 Way HandShake
───────────────
The bpf-prog can learn if it is sending SYN or SYNACK by reading the
sock_ops->skb_tcp_flags.
* Passive side
When writing SYNACK (i.e. sock_ops->op == BPF_SOCK_OPS_WRITE_HDR_OPT_CB),
the received SYN skb will be available to the bpf prog. The bpf prog can
use the SYN skb (which may carry the header option sent from the remote bpf
prog) to decide what bpf header option should be written to the outgoing
SYNACK skb. The SYN packet can be obtained by getsockopt(TCP_BPF_SYN*).
More on this later. Also, the bpf prog can learn if it is in syncookie
mode (by checking sock_ops->args[0] == BPF_WRITE_HDR_TCP_SYNACK_COOKIE).
The bpf prog can store the received SYN pkt by using the existing
bpf_setsockopt(TCP_SAVE_SYN). The example in a later patch does it.
[ Note that the fullsock here is a listen sk, bpf_sk_storage
is not very useful here since the listen sk will be shared
by many concurrent connection requests.
Extending bpf_sk_storage support to request_sock will add weight
to the minisock and it is not necessary better than storing the
whole ~100 bytes SYN pkt. ]
When the connection is established, the bpf prog will be called
in the existing PASSIVE_ESTABLISHED_CB callback. At that time,
the bpf prog can get the header option from the saved syn and
then apply the needed operation to the newly established socket.
The later patch will use the max delay ack specified in the SYN
header and set the RTO of this newly established connection
as an example.
The received ACK (that concludes the 3WHS) will also be available to
the bpf prog during PASSIVE_ESTABLISHED_CB through the sock_ops->skb_data.
It could be useful in syncookie scenario. More on this later.
There is an existing getsockopt "TCP_SAVED_SYN" to return the whole
saved syn pkt which includes the IP[46] header and the TCP header.
A few "TCP_BPF_SYN*" getsockopt has been added to allow specifying where to
start getting from, e.g. starting from TCP header, or from IP[46] header.
The new getsockopt(TCP_BPF_SYN*) will also know where it can get
the SYN's packet from:
- (a) the just received syn (available when the bpf prog is writing SYNACK)
and it is the only way to get SYN during syncookie mode.
or
- (b) the saved syn (available in PASSIVE_ESTABLISHED_CB and also other
existing CB).
The bpf prog does not need to know where the SYN pkt is coming from.
The getsockopt(TCP_BPF_SYN*) will hide this details.
Similarly, a flags "BPF_LOAD_HDR_OPT_TCP_SYN" is also added to
bpf_load_hdr_opt() to read a particular header option from the SYN packet.
* Fastopen
Fastopen should work the same as the regular non fastopen case.
This is a test in a later patch.
* Syncookie
For syncookie, the later example patch asks the active
side's bpf prog to resend the header options in ACK. The server
can use bpf_load_hdr_opt() to look at the options in this
received ACK during PASSIVE_ESTABLISHED_CB.
* Active side
The bpf prog will get a chance to write the bpf header option
in the SYN packet during WRITE_HDR_OPT_CB. The received SYNACK
pkt will also be available to the bpf prog during the existing
ACTIVE_ESTABLISHED_CB callback through the sock_ops->skb_data
and bpf_load_hdr_opt().
* Turn off header CB flags after 3WHS
If the bpf prog does not need to write/parse header options
beyond the 3WHS, the bpf prog can clear the bpf_sock_ops_cb_flags
to avoid being called for header options.
Or the bpf-prog can select to leave the UNKNOWN_HDR_OPT_CB_FLAG on
so that the kernel will only call it when there is option that
the kernel cannot handle.
[1]: draft-wang-tcpm-low-latency-opt-00
https://tools.ietf.org/html/draft-wang-tcpm-low-latency-opt-00
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200820190104.2885895-1-kafai@fb.com
2020-08-21 03:01:04 +08:00
|
|
|
struct bpf_sock_ops_kern sock_ops;
|
|
|
|
|
int err;
|
|
|
|
|
|
2020-08-21 03:00:52 +08:00
|
|
|
if (likely(!BPF_SOCK_OPS_TEST_FLAG(tcp_sk(sk),
|
|
|
|
|
BPF_SOCK_OPS_WRITE_HDR_OPT_CB_FLAG)) ||
|
|
|
|
|
!*remaining)
|
|
|
|
|
return;
|
|
|
|
|
|
bpf: tcp: Allow bpf prog to write and parse TCP header option
[ Note: The TCP changes here is mainly to implement the bpf
pieces into the bpf_skops_*() functions introduced
in the earlier patches. ]
The earlier effort in BPF-TCP-CC allows the TCP Congestion Control
algorithm to be written in BPF. It opens up opportunities to allow
a faster turnaround time in testing/releasing new congestion control
ideas to production environment.
The same flexibility can be extended to writing TCP header option.
It is not uncommon that people want to test new TCP header option
to improve the TCP performance. Another use case is for data-center
that has a more controlled environment and has more flexibility in
putting header options for internal only use.
For example, we want to test the idea in putting maximum delay
ACK in TCP header option which is similar to a draft RFC proposal [1].
This patch introduces the necessary BPF API and use them in the
TCP stack to allow BPF_PROG_TYPE_SOCK_OPS program to parse
and write TCP header options. It currently supports most of
the TCP packet except RST.
Supported TCP header option:
───────────────────────────
This patch allows the bpf-prog to write any option kind.
Different bpf-progs can write its own option by calling the new helper
bpf_store_hdr_opt(). The helper will ensure there is no duplicated
option in the header.
By allowing bpf-prog to write any option kind, this gives a lot of
flexibility to the bpf-prog. Different bpf-prog can write its
own option kind. It could also allow the bpf-prog to support a
recently standardized option on an older kernel.
Sockops Callback Flags:
──────────────────────
The bpf program will only be called to parse/write tcp header option
if the following newly added callback flags are enabled
in tp->bpf_sock_ops_cb_flags:
BPF_SOCK_OPS_PARSE_UNKNOWN_HDR_OPT_CB_FLAG
BPF_SOCK_OPS_PARSE_ALL_HDR_OPT_CB_FLAG
BPF_SOCK_OPS_WRITE_HDR_OPT_CB_FLAG
A few words on the PARSE CB flags. When the above PARSE CB flags are
turned on, the bpf-prog will be called on packets received
at a sk that has at least reached the ESTABLISHED state.
The parsing of the SYN-SYNACK-ACK will be discussed in the
"3 Way HandShake" section.
The default is off for all of the above new CB flags, i.e. the bpf prog
will not be called to parse or write bpf hdr option. There are
details comment on these new cb flags in the UAPI bpf.h.
sock_ops->skb_data and bpf_load_hdr_opt()
─────────────────────────────────────────
sock_ops->skb_data and sock_ops->skb_data_end covers the whole
TCP header and its options. They are read only.
The new bpf_load_hdr_opt() helps to read a particular option "kind"
from the skb_data.
Please refer to the comment in UAPI bpf.h. It has details
on what skb_data contains under different sock_ops->op.
3 Way HandShake
───────────────
The bpf-prog can learn if it is sending SYN or SYNACK by reading the
sock_ops->skb_tcp_flags.
* Passive side
When writing SYNACK (i.e. sock_ops->op == BPF_SOCK_OPS_WRITE_HDR_OPT_CB),
the received SYN skb will be available to the bpf prog. The bpf prog can
use the SYN skb (which may carry the header option sent from the remote bpf
prog) to decide what bpf header option should be written to the outgoing
SYNACK skb. The SYN packet can be obtained by getsockopt(TCP_BPF_SYN*).
More on this later. Also, the bpf prog can learn if it is in syncookie
mode (by checking sock_ops->args[0] == BPF_WRITE_HDR_TCP_SYNACK_COOKIE).
The bpf prog can store the received SYN pkt by using the existing
bpf_setsockopt(TCP_SAVE_SYN). The example in a later patch does it.
[ Note that the fullsock here is a listen sk, bpf_sk_storage
is not very useful here since the listen sk will be shared
by many concurrent connection requests.
Extending bpf_sk_storage support to request_sock will add weight
to the minisock and it is not necessary better than storing the
whole ~100 bytes SYN pkt. ]
When the connection is established, the bpf prog will be called
in the existing PASSIVE_ESTABLISHED_CB callback. At that time,
the bpf prog can get the header option from the saved syn and
then apply the needed operation to the newly established socket.
The later patch will use the max delay ack specified in the SYN
header and set the RTO of this newly established connection
as an example.
The received ACK (that concludes the 3WHS) will also be available to
the bpf prog during PASSIVE_ESTABLISHED_CB through the sock_ops->skb_data.
It could be useful in syncookie scenario. More on this later.
There is an existing getsockopt "TCP_SAVED_SYN" to return the whole
saved syn pkt which includes the IP[46] header and the TCP header.
A few "TCP_BPF_SYN*" getsockopt has been added to allow specifying where to
start getting from, e.g. starting from TCP header, or from IP[46] header.
The new getsockopt(TCP_BPF_SYN*) will also know where it can get
the SYN's packet from:
- (a) the just received syn (available when the bpf prog is writing SYNACK)
and it is the only way to get SYN during syncookie mode.
or
- (b) the saved syn (available in PASSIVE_ESTABLISHED_CB and also other
existing CB).
The bpf prog does not need to know where the SYN pkt is coming from.
The getsockopt(TCP_BPF_SYN*) will hide this details.
Similarly, a flags "BPF_LOAD_HDR_OPT_TCP_SYN" is also added to
bpf_load_hdr_opt() to read a particular header option from the SYN packet.
* Fastopen
Fastopen should work the same as the regular non fastopen case.
This is a test in a later patch.
* Syncookie
For syncookie, the later example patch asks the active
side's bpf prog to resend the header options in ACK. The server
can use bpf_load_hdr_opt() to look at the options in this
received ACK during PASSIVE_ESTABLISHED_CB.
* Active side
The bpf prog will get a chance to write the bpf header option
in the SYN packet during WRITE_HDR_OPT_CB. The received SYNACK
pkt will also be available to the bpf prog during the existing
ACTIVE_ESTABLISHED_CB callback through the sock_ops->skb_data
and bpf_load_hdr_opt().
* Turn off header CB flags after 3WHS
If the bpf prog does not need to write/parse header options
beyond the 3WHS, the bpf prog can clear the bpf_sock_ops_cb_flags
to avoid being called for header options.
Or the bpf-prog can select to leave the UNKNOWN_HDR_OPT_CB_FLAG on
so that the kernel will only call it when there is option that
the kernel cannot handle.
[1]: draft-wang-tcpm-low-latency-opt-00
https://tools.ietf.org/html/draft-wang-tcpm-low-latency-opt-00
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200820190104.2885895-1-kafai@fb.com
2020-08-21 03:01:04 +08:00
|
|
|
/* *remaining has already been aligned to 4 bytes, so *remaining >= 4 */
|
|
|
|
|
|
|
|
|
|
/* init sock_ops */
|
|
|
|
|
memset(&sock_ops, 0, offsetof(struct bpf_sock_ops_kern, temp));
|
|
|
|
|
|
|
|
|
|
sock_ops.op = BPF_SOCK_OPS_HDR_OPT_LEN_CB;
|
|
|
|
|
|
|
|
|
|
if (req) {
|
|
|
|
|
/* The listen "sk" cannot be passed here because
|
|
|
|
|
* it is not locked. It would not make too much
|
|
|
|
|
* sense to do bpf_setsockopt(listen_sk) based
|
|
|
|
|
* on individual connection request also.
|
|
|
|
|
*
|
|
|
|
|
* Thus, "req" is passed here and the cgroup-bpf-progs
|
|
|
|
|
* of the listen "sk" will be run.
|
|
|
|
|
*
|
|
|
|
|
* "req" is also used here for fastopen even the "sk" here is
|
|
|
|
|
* a fullsock "child" sk. It is to keep the behavior
|
|
|
|
|
* consistent between fastopen and non-fastopen on
|
|
|
|
|
* the bpf programming side.
|
|
|
|
|
*/
|
|
|
|
|
sock_ops.sk = (struct sock *)req;
|
|
|
|
|
sock_ops.syn_skb = syn_skb;
|
|
|
|
|
} else {
|
|
|
|
|
sock_owned_by_me(sk);
|
|
|
|
|
|
|
|
|
|
sock_ops.is_fullsock = 1;
|
|
|
|
|
sock_ops.sk = sk;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
sock_ops.args[0] = bpf_skops_write_hdr_opt_arg0(skb, synack_type);
|
|
|
|
|
sock_ops.remaining_opt_len = *remaining;
|
|
|
|
|
/* tcp_current_mss() does not pass a skb */
|
|
|
|
|
if (skb)
|
|
|
|
|
bpf_skops_init_skb(&sock_ops, skb, 0);
|
|
|
|
|
|
|
|
|
|
err = BPF_CGROUP_RUN_PROG_SOCK_OPS_SK(&sock_ops, sk);
|
|
|
|
|
|
|
|
|
|
if (err || sock_ops.remaining_opt_len == *remaining)
|
|
|
|
|
return;
|
|
|
|
|
|
|
|
|
|
opts->bpf_opt_len = *remaining - sock_ops.remaining_opt_len;
|
|
|
|
|
/* round up to 4 bytes */
|
|
|
|
|
opts->bpf_opt_len = (opts->bpf_opt_len + 3) & ~3;
|
|
|
|
|
|
|
|
|
|
*remaining -= opts->bpf_opt_len;
|
2020-08-21 03:00:52 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static void bpf_skops_write_hdr_opt(struct sock *sk, struct sk_buff *skb,
|
|
|
|
|
struct request_sock *req,
|
|
|
|
|
struct sk_buff *syn_skb,
|
|
|
|
|
enum tcp_synack_type synack_type,
|
|
|
|
|
struct tcp_out_options *opts)
|
|
|
|
|
{
|
bpf: tcp: Allow bpf prog to write and parse TCP header option
[ Note: The TCP changes here is mainly to implement the bpf
pieces into the bpf_skops_*() functions introduced
in the earlier patches. ]
The earlier effort in BPF-TCP-CC allows the TCP Congestion Control
algorithm to be written in BPF. It opens up opportunities to allow
a faster turnaround time in testing/releasing new congestion control
ideas to production environment.
The same flexibility can be extended to writing TCP header option.
It is not uncommon that people want to test new TCP header option
to improve the TCP performance. Another use case is for data-center
that has a more controlled environment and has more flexibility in
putting header options for internal only use.
For example, we want to test the idea in putting maximum delay
ACK in TCP header option which is similar to a draft RFC proposal [1].
This patch introduces the necessary BPF API and use them in the
TCP stack to allow BPF_PROG_TYPE_SOCK_OPS program to parse
and write TCP header options. It currently supports most of
the TCP packet except RST.
Supported TCP header option:
───────────────────────────
This patch allows the bpf-prog to write any option kind.
Different bpf-progs can write its own option by calling the new helper
bpf_store_hdr_opt(). The helper will ensure there is no duplicated
option in the header.
By allowing bpf-prog to write any option kind, this gives a lot of
flexibility to the bpf-prog. Different bpf-prog can write its
own option kind. It could also allow the bpf-prog to support a
recently standardized option on an older kernel.
Sockops Callback Flags:
──────────────────────
The bpf program will only be called to parse/write tcp header option
if the following newly added callback flags are enabled
in tp->bpf_sock_ops_cb_flags:
BPF_SOCK_OPS_PARSE_UNKNOWN_HDR_OPT_CB_FLAG
BPF_SOCK_OPS_PARSE_ALL_HDR_OPT_CB_FLAG
BPF_SOCK_OPS_WRITE_HDR_OPT_CB_FLAG
A few words on the PARSE CB flags. When the above PARSE CB flags are
turned on, the bpf-prog will be called on packets received
at a sk that has at least reached the ESTABLISHED state.
The parsing of the SYN-SYNACK-ACK will be discussed in the
"3 Way HandShake" section.
The default is off for all of the above new CB flags, i.e. the bpf prog
will not be called to parse or write bpf hdr option. There are
details comment on these new cb flags in the UAPI bpf.h.
sock_ops->skb_data and bpf_load_hdr_opt()
─────────────────────────────────────────
sock_ops->skb_data and sock_ops->skb_data_end covers the whole
TCP header and its options. They are read only.
The new bpf_load_hdr_opt() helps to read a particular option "kind"
from the skb_data.
Please refer to the comment in UAPI bpf.h. It has details
on what skb_data contains under different sock_ops->op.
3 Way HandShake
───────────────
The bpf-prog can learn if it is sending SYN or SYNACK by reading the
sock_ops->skb_tcp_flags.
* Passive side
When writing SYNACK (i.e. sock_ops->op == BPF_SOCK_OPS_WRITE_HDR_OPT_CB),
the received SYN skb will be available to the bpf prog. The bpf prog can
use the SYN skb (which may carry the header option sent from the remote bpf
prog) to decide what bpf header option should be written to the outgoing
SYNACK skb. The SYN packet can be obtained by getsockopt(TCP_BPF_SYN*).
More on this later. Also, the bpf prog can learn if it is in syncookie
mode (by checking sock_ops->args[0] == BPF_WRITE_HDR_TCP_SYNACK_COOKIE).
The bpf prog can store the received SYN pkt by using the existing
bpf_setsockopt(TCP_SAVE_SYN). The example in a later patch does it.
[ Note that the fullsock here is a listen sk, bpf_sk_storage
is not very useful here since the listen sk will be shared
by many concurrent connection requests.
Extending bpf_sk_storage support to request_sock will add weight
to the minisock and it is not necessary better than storing the
whole ~100 bytes SYN pkt. ]
When the connection is established, the bpf prog will be called
in the existing PASSIVE_ESTABLISHED_CB callback. At that time,
the bpf prog can get the header option from the saved syn and
then apply the needed operation to the newly established socket.
The later patch will use the max delay ack specified in the SYN
header and set the RTO of this newly established connection
as an example.
The received ACK (that concludes the 3WHS) will also be available to
the bpf prog during PASSIVE_ESTABLISHED_CB through the sock_ops->skb_data.
It could be useful in syncookie scenario. More on this later.
There is an existing getsockopt "TCP_SAVED_SYN" to return the whole
saved syn pkt which includes the IP[46] header and the TCP header.
A few "TCP_BPF_SYN*" getsockopt has been added to allow specifying where to
start getting from, e.g. starting from TCP header, or from IP[46] header.
The new getsockopt(TCP_BPF_SYN*) will also know where it can get
the SYN's packet from:
- (a) the just received syn (available when the bpf prog is writing SYNACK)
and it is the only way to get SYN during syncookie mode.
or
- (b) the saved syn (available in PASSIVE_ESTABLISHED_CB and also other
existing CB).
The bpf prog does not need to know where the SYN pkt is coming from.
The getsockopt(TCP_BPF_SYN*) will hide this details.
Similarly, a flags "BPF_LOAD_HDR_OPT_TCP_SYN" is also added to
bpf_load_hdr_opt() to read a particular header option from the SYN packet.
* Fastopen
Fastopen should work the same as the regular non fastopen case.
This is a test in a later patch.
* Syncookie
For syncookie, the later example patch asks the active
side's bpf prog to resend the header options in ACK. The server
can use bpf_load_hdr_opt() to look at the options in this
received ACK during PASSIVE_ESTABLISHED_CB.
* Active side
The bpf prog will get a chance to write the bpf header option
in the SYN packet during WRITE_HDR_OPT_CB. The received SYNACK
pkt will also be available to the bpf prog during the existing
ACTIVE_ESTABLISHED_CB callback through the sock_ops->skb_data
and bpf_load_hdr_opt().
* Turn off header CB flags after 3WHS
If the bpf prog does not need to write/parse header options
beyond the 3WHS, the bpf prog can clear the bpf_sock_ops_cb_flags
to avoid being called for header options.
Or the bpf-prog can select to leave the UNKNOWN_HDR_OPT_CB_FLAG on
so that the kernel will only call it when there is option that
the kernel cannot handle.
[1]: draft-wang-tcpm-low-latency-opt-00
https://tools.ietf.org/html/draft-wang-tcpm-low-latency-opt-00
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200820190104.2885895-1-kafai@fb.com
2020-08-21 03:01:04 +08:00
|
|
|
u8 first_opt_off, nr_written, max_opt_len = opts->bpf_opt_len;
|
|
|
|
|
struct bpf_sock_ops_kern sock_ops;
|
|
|
|
|
int err;
|
|
|
|
|
|
|
|
|
|
if (likely(!max_opt_len))
|
2020-08-21 03:00:52 +08:00
|
|
|
return;
|
|
|
|
|
|
bpf: tcp: Allow bpf prog to write and parse TCP header option
[ Note: The TCP changes here is mainly to implement the bpf
pieces into the bpf_skops_*() functions introduced
in the earlier patches. ]
The earlier effort in BPF-TCP-CC allows the TCP Congestion Control
algorithm to be written in BPF. It opens up opportunities to allow
a faster turnaround time in testing/releasing new congestion control
ideas to production environment.
The same flexibility can be extended to writing TCP header option.
It is not uncommon that people want to test new TCP header option
to improve the TCP performance. Another use case is for data-center
that has a more controlled environment and has more flexibility in
putting header options for internal only use.
For example, we want to test the idea in putting maximum delay
ACK in TCP header option which is similar to a draft RFC proposal [1].
This patch introduces the necessary BPF API and use them in the
TCP stack to allow BPF_PROG_TYPE_SOCK_OPS program to parse
and write TCP header options. It currently supports most of
the TCP packet except RST.
Supported TCP header option:
───────────────────────────
This patch allows the bpf-prog to write any option kind.
Different bpf-progs can write its own option by calling the new helper
bpf_store_hdr_opt(). The helper will ensure there is no duplicated
option in the header.
By allowing bpf-prog to write any option kind, this gives a lot of
flexibility to the bpf-prog. Different bpf-prog can write its
own option kind. It could also allow the bpf-prog to support a
recently standardized option on an older kernel.
Sockops Callback Flags:
──────────────────────
The bpf program will only be called to parse/write tcp header option
if the following newly added callback flags are enabled
in tp->bpf_sock_ops_cb_flags:
BPF_SOCK_OPS_PARSE_UNKNOWN_HDR_OPT_CB_FLAG
BPF_SOCK_OPS_PARSE_ALL_HDR_OPT_CB_FLAG
BPF_SOCK_OPS_WRITE_HDR_OPT_CB_FLAG
A few words on the PARSE CB flags. When the above PARSE CB flags are
turned on, the bpf-prog will be called on packets received
at a sk that has at least reached the ESTABLISHED state.
The parsing of the SYN-SYNACK-ACK will be discussed in the
"3 Way HandShake" section.
The default is off for all of the above new CB flags, i.e. the bpf prog
will not be called to parse or write bpf hdr option. There are
details comment on these new cb flags in the UAPI bpf.h.
sock_ops->skb_data and bpf_load_hdr_opt()
─────────────────────────────────────────
sock_ops->skb_data and sock_ops->skb_data_end covers the whole
TCP header and its options. They are read only.
The new bpf_load_hdr_opt() helps to read a particular option "kind"
from the skb_data.
Please refer to the comment in UAPI bpf.h. It has details
on what skb_data contains under different sock_ops->op.
3 Way HandShake
───────────────
The bpf-prog can learn if it is sending SYN or SYNACK by reading the
sock_ops->skb_tcp_flags.
* Passive side
When writing SYNACK (i.e. sock_ops->op == BPF_SOCK_OPS_WRITE_HDR_OPT_CB),
the received SYN skb will be available to the bpf prog. The bpf prog can
use the SYN skb (which may carry the header option sent from the remote bpf
prog) to decide what bpf header option should be written to the outgoing
SYNACK skb. The SYN packet can be obtained by getsockopt(TCP_BPF_SYN*).
More on this later. Also, the bpf prog can learn if it is in syncookie
mode (by checking sock_ops->args[0] == BPF_WRITE_HDR_TCP_SYNACK_COOKIE).
The bpf prog can store the received SYN pkt by using the existing
bpf_setsockopt(TCP_SAVE_SYN). The example in a later patch does it.
[ Note that the fullsock here is a listen sk, bpf_sk_storage
is not very useful here since the listen sk will be shared
by many concurrent connection requests.
Extending bpf_sk_storage support to request_sock will add weight
to the minisock and it is not necessary better than storing the
whole ~100 bytes SYN pkt. ]
When the connection is established, the bpf prog will be called
in the existing PASSIVE_ESTABLISHED_CB callback. At that time,
the bpf prog can get the header option from the saved syn and
then apply the needed operation to the newly established socket.
The later patch will use the max delay ack specified in the SYN
header and set the RTO of this newly established connection
as an example.
The received ACK (that concludes the 3WHS) will also be available to
the bpf prog during PASSIVE_ESTABLISHED_CB through the sock_ops->skb_data.
It could be useful in syncookie scenario. More on this later.
There is an existing getsockopt "TCP_SAVED_SYN" to return the whole
saved syn pkt which includes the IP[46] header and the TCP header.
A few "TCP_BPF_SYN*" getsockopt has been added to allow specifying where to
start getting from, e.g. starting from TCP header, or from IP[46] header.
The new getsockopt(TCP_BPF_SYN*) will also know where it can get
the SYN's packet from:
- (a) the just received syn (available when the bpf prog is writing SYNACK)
and it is the only way to get SYN during syncookie mode.
or
- (b) the saved syn (available in PASSIVE_ESTABLISHED_CB and also other
existing CB).
The bpf prog does not need to know where the SYN pkt is coming from.
The getsockopt(TCP_BPF_SYN*) will hide this details.
Similarly, a flags "BPF_LOAD_HDR_OPT_TCP_SYN" is also added to
bpf_load_hdr_opt() to read a particular header option from the SYN packet.
* Fastopen
Fastopen should work the same as the regular non fastopen case.
This is a test in a later patch.
* Syncookie
For syncookie, the later example patch asks the active
side's bpf prog to resend the header options in ACK. The server
can use bpf_load_hdr_opt() to look at the options in this
received ACK during PASSIVE_ESTABLISHED_CB.
* Active side
The bpf prog will get a chance to write the bpf header option
in the SYN packet during WRITE_HDR_OPT_CB. The received SYNACK
pkt will also be available to the bpf prog during the existing
ACTIVE_ESTABLISHED_CB callback through the sock_ops->skb_data
and bpf_load_hdr_opt().
* Turn off header CB flags after 3WHS
If the bpf prog does not need to write/parse header options
beyond the 3WHS, the bpf prog can clear the bpf_sock_ops_cb_flags
to avoid being called for header options.
Or the bpf-prog can select to leave the UNKNOWN_HDR_OPT_CB_FLAG on
so that the kernel will only call it when there is option that
the kernel cannot handle.
[1]: draft-wang-tcpm-low-latency-opt-00
https://tools.ietf.org/html/draft-wang-tcpm-low-latency-opt-00
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200820190104.2885895-1-kafai@fb.com
2020-08-21 03:01:04 +08:00
|
|
|
memset(&sock_ops, 0, offsetof(struct bpf_sock_ops_kern, temp));
|
|
|
|
|
|
|
|
|
|
sock_ops.op = BPF_SOCK_OPS_WRITE_HDR_OPT_CB;
|
|
|
|
|
|
|
|
|
|
if (req) {
|
|
|
|
|
sock_ops.sk = (struct sock *)req;
|
|
|
|
|
sock_ops.syn_skb = syn_skb;
|
|
|
|
|
} else {
|
|
|
|
|
sock_owned_by_me(sk);
|
|
|
|
|
|
|
|
|
|
sock_ops.is_fullsock = 1;
|
|
|
|
|
sock_ops.sk = sk;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
sock_ops.args[0] = bpf_skops_write_hdr_opt_arg0(skb, synack_type);
|
|
|
|
|
sock_ops.remaining_opt_len = max_opt_len;
|
|
|
|
|
first_opt_off = tcp_hdrlen(skb) - max_opt_len;
|
|
|
|
|
bpf_skops_init_skb(&sock_ops, skb, first_opt_off);
|
|
|
|
|
|
|
|
|
|
err = BPF_CGROUP_RUN_PROG_SOCK_OPS_SK(&sock_ops, sk);
|
|
|
|
|
|
|
|
|
|
if (err)
|
|
|
|
|
nr_written = 0;
|
|
|
|
|
else
|
|
|
|
|
nr_written = max_opt_len - sock_ops.remaining_opt_len;
|
|
|
|
|
|
|
|
|
|
if (nr_written < max_opt_len)
|
|
|
|
|
memset(skb->data + first_opt_off + nr_written, TCPOPT_NOP,
|
|
|
|
|
max_opt_len - nr_written);
|
2020-08-21 03:00:52 +08:00
|
|
|
}
|
|
|
|
|
#else
|
|
|
|
|
static void bpf_skops_hdr_opt_len(struct sock *sk, struct sk_buff *skb,
|
|
|
|
|
struct request_sock *req,
|
|
|
|
|
struct sk_buff *syn_skb,
|
|
|
|
|
enum tcp_synack_type synack_type,
|
|
|
|
|
struct tcp_out_options *opts,
|
|
|
|
|
unsigned int *remaining)
|
|
|
|
|
{
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static void bpf_skops_write_hdr_opt(struct sock *sk, struct sk_buff *skb,
|
|
|
|
|
struct request_sock *req,
|
|
|
|
|
struct sk_buff *syn_skb,
|
|
|
|
|
enum tcp_synack_type synack_type,
|
|
|
|
|
struct tcp_out_options *opts)
|
|
|
|
|
{
|
|
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
|
2009-07-22 07:00:40 +08:00
|
|
|
/* Write previously computed TCP options to the packet.
|
|
|
|
|
*
|
|
|
|
|
* Beware: Something in the Internet is very sensitive to the ordering of
|
2008-10-24 05:06:35 +08:00
|
|
|
* TCP options, we learned this through the hard way, so be careful here.
|
|
|
|
|
* Luckily we can at least blame others for their non-compliance but from
|
2013-12-09 04:15:44 +08:00
|
|
|
* inter-operability perspective it seems that we're somewhat stuck with
|
2008-10-24 05:06:35 +08:00
|
|
|
* the ordering which we have been using if we want to keep working with
|
|
|
|
|
* those broken things (not that it currently hurts anybody as there isn't
|
|
|
|
|
* particular reason why the ordering would need to be changed).
|
|
|
|
|
*
|
|
|
|
|
* At least SACK_PERM as the first option is known to lead to a disaster
|
|
|
|
|
* (but it may well be that other scenarios fail similarly).
|
|
|
|
|
*/
|
2008-07-19 15:04:31 +08:00
|
|
|
static void tcp_options_write(__be32 *ptr, struct tcp_sock *tp,
|
2009-12-03 02:23:05 +08:00
|
|
|
struct tcp_out_options *opts)
|
|
|
|
|
{
|
2012-07-19 14:43:05 +08:00
|
|
|
u16 options = opts->options; /* mungable copy */
|
2009-12-03 02:23:05 +08:00
|
|
|
|
|
|
|
|
if (unlikely(OPTION_MD5 & options)) {
|
2013-03-17 16:23:34 +08:00
|
|
|
*ptr++ = htonl((TCPOPT_NOP << 24) | (TCPOPT_NOP << 16) |
|
|
|
|
|
(TCPOPT_MD5SIG << 8) | TCPOLEN_MD5SIG);
|
2009-12-03 02:23:05 +08:00
|
|
|
/* overload cookie hash location */
|
|
|
|
|
opts->hash_location = (__u8 *)ptr;
|
2008-07-19 15:04:31 +08:00
|
|
|
ptr += 4;
|
2006-01-04 08:03:49 +08:00
|
|
|
}
|
2008-07-19 15:04:31 +08:00
|
|
|
|
2008-10-24 05:06:35 +08:00
|
|
|
if (unlikely(opts->mss)) {
|
|
|
|
|
*ptr++ = htonl((TCPOPT_MSS << 24) |
|
|
|
|
|
(TCPOLEN_MSS << 16) |
|
|
|
|
|
opts->mss);
|
|
|
|
|
}
|
|
|
|
|
|
2009-12-03 02:23:05 +08:00
|
|
|
if (likely(OPTION_TS & options)) {
|
|
|
|
|
if (unlikely(OPTION_SACK_ADVERTISE & options)) {
|
2008-07-19 15:04:31 +08:00
|
|
|
*ptr++ = htonl((TCPOPT_SACK_PERM << 24) |
|
|
|
|
|
(TCPOLEN_SACK_PERM << 16) |
|
|
|
|
|
(TCPOPT_TIMESTAMP << 8) |
|
|
|
|
|
TCPOLEN_TIMESTAMP);
|
2009-12-03 02:23:05 +08:00
|
|
|
options &= ~OPTION_SACK_ADVERTISE;
|
2008-07-19 15:04:31 +08:00
|
|
|
} else {
|
|
|
|
|
*ptr++ = htonl((TCPOPT_NOP << 24) |
|
|
|
|
|
(TCPOPT_NOP << 16) |
|
|
|
|
|
(TCPOPT_TIMESTAMP << 8) |
|
|
|
|
|
TCPOLEN_TIMESTAMP);
|
|
|
|
|
}
|
|
|
|
|
*ptr++ = htonl(opts->tsval);
|
|
|
|
|
*ptr++ = htonl(opts->tsecr);
|
|
|
|
|
}
|
|
|
|
|
|
2009-12-03 02:23:05 +08:00
|
|
|
if (unlikely(OPTION_SACK_ADVERTISE & options)) {
|
2008-07-19 15:04:31 +08:00
|
|
|
*ptr++ = htonl((TCPOPT_NOP << 24) |
|
|
|
|
|
(TCPOPT_NOP << 16) |
|
|
|
|
|
(TCPOPT_SACK_PERM << 8) |
|
|
|
|
|
TCPOLEN_SACK_PERM);
|
|
|
|
|
}
|
|
|
|
|
|
2009-12-03 02:23:05 +08:00
|
|
|
if (unlikely(OPTION_WSCALE & options)) {
|
2008-07-19 15:04:31 +08:00
|
|
|
*ptr++ = htonl((TCPOPT_NOP << 24) |
|
|
|
|
|
(TCPOPT_WINDOW << 16) |
|
|
|
|
|
(TCPOLEN_WINDOW << 8) |
|
|
|
|
|
opts->ws);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (unlikely(opts->num_sack_blocks)) {
|
|
|
|
|
struct tcp_sack_block *sp = tp->rx_opt.dsack ?
|
|
|
|
|
tp->duplicate_sack : tp->selective_acks;
|
2006-01-04 08:03:49 +08:00
|
|
|
int this_sack;
|
|
|
|
|
|
|
|
|
|
*ptr++ = htonl((TCPOPT_NOP << 24) |
|
|
|
|
|
(TCPOPT_NOP << 16) |
|
|
|
|
|
(TCPOPT_SACK << 8) |
|
2008-07-19 15:04:31 +08:00
|
|
|
(TCPOLEN_SACK_BASE + (opts->num_sack_blocks *
|
2006-01-04 08:03:49 +08:00
|
|
|
TCPOLEN_SACK_PERBLOCK)));
|
2007-03-09 12:45:19 +08:00
|
|
|
|
2008-07-19 15:04:31 +08:00
|
|
|
for (this_sack = 0; this_sack < opts->num_sack_blocks;
|
|
|
|
|
++this_sack) {
|
2006-01-04 08:03:49 +08:00
|
|
|
*ptr++ = htonl(sp[this_sack].start_seq);
|
|
|
|
|
*ptr++ = htonl(sp[this_sack].end_seq);
|
|
|
|
|
}
|
2007-03-09 12:45:19 +08:00
|
|
|
|
2009-03-14 22:23:01 +08:00
|
|
|
tp->rx_opt.dsack = 0;
|
2006-01-04 08:03:49 +08:00
|
|
|
}
|
2012-07-19 14:43:05 +08:00
|
|
|
|
|
|
|
|
if (unlikely(OPTION_FAST_OPEN_COOKIE & options)) {
|
|
|
|
|
struct tcp_fastopen_cookie *foc = opts->fastopen_cookie;
|
2015-04-07 05:37:26 +08:00
|
|
|
u8 *p = (u8 *)ptr;
|
|
|
|
|
u32 len; /* Fast Open option length */
|
|
|
|
|
|
|
|
|
|
if (foc->exp) {
|
|
|
|
|
len = TCPOLEN_EXP_FASTOPEN_BASE + foc->len;
|
|
|
|
|
*ptr = htonl((TCPOPT_EXP << 24) | (len << 16) |
|
|
|
|
|
TCPOPT_FASTOPEN_MAGIC);
|
|
|
|
|
p += TCPOLEN_EXP_FASTOPEN_BASE;
|
|
|
|
|
} else {
|
|
|
|
|
len = TCPOLEN_FASTOPEN_BASE + foc->len;
|
|
|
|
|
*p++ = TCPOPT_FASTOPEN;
|
|
|
|
|
*p++ = len;
|
|
|
|
|
}
|
2012-07-19 14:43:05 +08:00
|
|
|
|
2015-04-07 05:37:26 +08:00
|
|
|
memcpy(p, foc->val, foc->len);
|
|
|
|
|
if ((len & 3) == 2) {
|
|
|
|
|
p[foc->len] = TCPOPT_NOP;
|
|
|
|
|
p[foc->len + 1] = TCPOPT_NOP;
|
2012-07-19 14:43:05 +08:00
|
|
|
}
|
2015-04-07 05:37:26 +08:00
|
|
|
ptr += (len + 3) >> 2;
|
2012-07-19 14:43:05 +08:00
|
|
|
}
|
2017-10-25 17:01:45 +08:00
|
|
|
|
|
|
|
|
smc_options_write(ptr, &options);
|
2020-01-22 08:56:16 +08:00
|
|
|
|
2020-11-20 03:46:02 +08:00
|
|
|
mptcp_options_write(ptr, tp, opts);
|
2017-10-25 17:01:45 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static void smc_set_option(const struct tcp_sock *tp,
|
|
|
|
|
struct tcp_out_options *opts,
|
|
|
|
|
unsigned int *remaining)
|
|
|
|
|
{
|
|
|
|
|
#if IS_ENABLED(CONFIG_SMC)
|
|
|
|
|
if (static_branch_unlikely(&tcp_have_smc)) {
|
|
|
|
|
if (tp->syn_smc) {
|
|
|
|
|
if (*remaining >= TCPOLEN_EXP_SMC_BASE_ALIGNED) {
|
|
|
|
|
opts->options |= OPTION_SMC;
|
|
|
|
|
*remaining -= TCPOLEN_EXP_SMC_BASE_ALIGNED;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static void smc_set_option_cond(const struct tcp_sock *tp,
|
|
|
|
|
const struct inet_request_sock *ireq,
|
|
|
|
|
struct tcp_out_options *opts,
|
|
|
|
|
unsigned int *remaining)
|
|
|
|
|
{
|
|
|
|
|
#if IS_ENABLED(CONFIG_SMC)
|
|
|
|
|
if (static_branch_unlikely(&tcp_have_smc)) {
|
|
|
|
|
if (tp->syn_smc && ireq->smc_ok) {
|
|
|
|
|
if (*remaining >= TCPOLEN_EXP_SMC_BASE_ALIGNED) {
|
|
|
|
|
opts->options |= OPTION_SMC;
|
|
|
|
|
*remaining -= TCPOLEN_EXP_SMC_BASE_ALIGNED;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
#endif
|
2008-07-19 15:04:31 +08:00
|
|
|
}
|
|
|
|
|
|
2020-01-22 08:56:18 +08:00
|
|
|
static void mptcp_set_option_cond(const struct request_sock *req,
|
|
|
|
|
struct tcp_out_options *opts,
|
|
|
|
|
unsigned int *remaining)
|
|
|
|
|
{
|
|
|
|
|
if (rsk_is_mptcp(req)) {
|
|
|
|
|
unsigned int size;
|
|
|
|
|
|
|
|
|
|
if (mptcp_synack_options(req, &size, &opts->mptcp)) {
|
|
|
|
|
if (*remaining >= size) {
|
|
|
|
|
opts->options |= OPTION_MPTCP;
|
|
|
|
|
*remaining -= size;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2009-07-22 07:00:40 +08:00
|
|
|
/* Compute TCP options for SYN packets. This is not the final
|
|
|
|
|
* network wire format yet.
|
|
|
|
|
*/
|
2012-04-15 13:58:06 +08:00
|
|
|
static unsigned int tcp_syn_options(struct sock *sk, struct sk_buff *skb,
|
2008-07-19 15:04:31 +08:00
|
|
|
struct tcp_out_options *opts,
|
2011-10-21 17:22:42 +08:00
|
|
|
struct tcp_md5sig_key **md5)
|
|
|
|
|
{
|
2008-07-19 15:04:31 +08:00
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
2012-04-15 13:58:06 +08:00
|
|
|
unsigned int remaining = MAX_TCP_OPTION_SPACE;
|
2012-07-19 14:43:07 +08:00
|
|
|
struct tcp_fastopen_request *fastopen = tp->fastopen_req;
|
2008-07-19 15:04:31 +08:00
|
|
|
|
2018-04-24 05:46:25 +08:00
|
|
|
*md5 = NULL;
|
2006-11-15 11:07:45 +08:00
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
2019-02-27 01:49:11 +08:00
|
|
|
if (static_branch_unlikely(&tcp_md5_needed) &&
|
2018-11-28 07:03:21 +08:00
|
|
|
rcu_access_pointer(tp->md5sig_info)) {
|
2018-04-24 05:46:25 +08:00
|
|
|
*md5 = tp->af_specific->md5_lookup(sk, sk);
|
|
|
|
|
if (*md5) {
|
|
|
|
|
opts->options |= OPTION_MD5;
|
|
|
|
|
remaining -= TCPOLEN_MD5SIG_ALIGNED;
|
|
|
|
|
}
|
2006-11-15 11:07:45 +08:00
|
|
|
}
|
|
|
|
|
#endif
|
2008-07-19 15:04:31 +08:00
|
|
|
|
|
|
|
|
/* We always get an MSS option. The option bytes which will be seen in
|
|
|
|
|
* normal data packets should timestamps be used, must be in the MSS
|
|
|
|
|
* advertised. But we subtract them from tp->mss_cache so that
|
|
|
|
|
* calculations in tcp_sendmsg are simpler etc. So account for this
|
|
|
|
|
* fact here if necessary. If we don't do this correctly, as a
|
|
|
|
|
* receiver we won't recognize data packets as being full sized when we
|
|
|
|
|
* should, and thus we won't abide by the delayed ACK rules correctly.
|
|
|
|
|
* SACKs don't matter, we never delay an ACK when we have any of those
|
|
|
|
|
* going out. */
|
|
|
|
|
opts->mss = tcp_advertise_mss(sk);
|
2009-12-03 02:23:05 +08:00
|
|
|
remaining -= TCPOLEN_MSS_ALIGNED;
|
2008-07-19 15:04:31 +08:00
|
|
|
|
2017-06-08 01:34:39 +08:00
|
|
|
if (likely(sock_net(sk)->ipv4.sysctl_tcp_timestamps && !*md5)) {
|
2008-07-19 15:04:31 +08:00
|
|
|
opts->options |= OPTION_TS;
|
2014-09-06 06:33:33 +08:00
|
|
|
opts->tsval = tcp_skb_timestamp(skb) + tp->tsoffset;
|
2008-07-19 15:04:31 +08:00
|
|
|
opts->tsecr = tp->rx_opt.ts_recent;
|
2009-12-03 02:23:05 +08:00
|
|
|
remaining -= TCPOLEN_TSTAMP_ALIGNED;
|
2008-07-19 15:04:31 +08:00
|
|
|
}
|
2017-06-08 01:34:38 +08:00
|
|
|
if (likely(sock_net(sk)->ipv4.sysctl_tcp_window_scaling)) {
|
2008-07-19 15:04:31 +08:00
|
|
|
opts->ws = tp->rx_opt.rcv_wscale;
|
2009-10-01 14:41:59 +08:00
|
|
|
opts->options |= OPTION_WSCALE;
|
2009-12-03 02:23:05 +08:00
|
|
|
remaining -= TCPOLEN_WSCALE_ALIGNED;
|
2008-07-19 15:04:31 +08:00
|
|
|
}
|
2017-06-08 01:34:37 +08:00
|
|
|
if (likely(sock_net(sk)->ipv4.sysctl_tcp_sack)) {
|
2008-07-19 15:04:31 +08:00
|
|
|
opts->options |= OPTION_SACK_ADVERTISE;
|
2008-07-22 09:45:34 +08:00
|
|
|
if (unlikely(!(OPTION_TS & opts->options)))
|
2009-12-03 02:23:05 +08:00
|
|
|
remaining -= TCPOLEN_SACKPERM_ALIGNED;
|
2008-07-19 15:04:31 +08:00
|
|
|
}
|
|
|
|
|
|
2012-07-19 14:43:07 +08:00
|
|
|
if (fastopen && fastopen->cookie.len >= 0) {
|
2015-04-07 05:37:27 +08:00
|
|
|
u32 need = fastopen->cookie.len;
|
|
|
|
|
|
|
|
|
|
need += fastopen->cookie.exp ? TCPOLEN_EXP_FASTOPEN_BASE :
|
|
|
|
|
TCPOLEN_FASTOPEN_BASE;
|
2012-07-19 14:43:07 +08:00
|
|
|
need = (need + 3) & ~3U; /* Align to 32 bits */
|
|
|
|
|
if (remaining >= need) {
|
|
|
|
|
opts->options |= OPTION_FAST_OPEN_COOKIE;
|
|
|
|
|
opts->fastopen_cookie = &fastopen->cookie;
|
|
|
|
|
remaining -= need;
|
|
|
|
|
tp->syn_fastopen = 1;
|
2015-04-07 05:37:27 +08:00
|
|
|
tp->syn_fastopen_exp = fastopen->cookie.exp ? 1 : 0;
|
2012-07-19 14:43:07 +08:00
|
|
|
}
|
|
|
|
|
}
|
2009-12-03 02:23:05 +08:00
|
|
|
|
2017-10-25 17:01:45 +08:00
|
|
|
smc_set_option(tp, opts, &remaining);
|
|
|
|
|
|
2020-01-22 08:56:18 +08:00
|
|
|
if (sk_is_mptcp(sk)) {
|
|
|
|
|
unsigned int size;
|
|
|
|
|
|
2020-01-22 08:56:31 +08:00
|
|
|
if (mptcp_syn_options(sk, skb, &size, &opts->mptcp)) {
|
2020-01-22 08:56:18 +08:00
|
|
|
opts->options |= OPTION_MPTCP;
|
|
|
|
|
remaining -= size;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2020-08-21 03:00:52 +08:00
|
|
|
bpf_skops_hdr_opt_len(sk, skb, NULL, NULL, 0, opts, &remaining);
|
|
|
|
|
|
2009-12-03 02:23:05 +08:00
|
|
|
return MAX_TCP_OPTION_SPACE - remaining;
|
2006-01-04 08:03:49 +08:00
|
|
|
}
|
|
|
|
|
|
2009-07-22 07:00:40 +08:00
|
|
|
/* Set up TCP options for SYN-ACKs. */
|
2017-10-25 17:01:45 +08:00
|
|
|
static unsigned int tcp_synack_options(const struct sock *sk,
|
|
|
|
|
struct request_sock *req,
|
2015-09-25 22:39:17 +08:00
|
|
|
unsigned int mss, struct sk_buff *skb,
|
|
|
|
|
struct tcp_out_options *opts,
|
|
|
|
|
const struct tcp_md5sig_key *md5,
|
tcp: md5: do not send silly options in SYNCOOKIES
Whenever cookie_init_timestamp() has been used to encode
ECN,SACK,WSCALE options, we can not remove the TS option in the SYNACK.
Otherwise, tcp_synack_options() will still advertize options like WSCALE
that we can not deduce later when receiving the packet from the client
to complete 3WHS.
Note that modern linux TCP stacks wont use MD5+TS+SACK in a SYN packet,
but we can not know for sure that all TCP stacks have the same logic.
Before the fix a tcpdump would exhibit this wrong exchange :
10:12:15.464591 IP C > S: Flags [S], seq 4202415601, win 65535, options [nop,nop,md5 valid,mss 1400,sackOK,TS val 456965269 ecr 0,nop,wscale 8], length 0
10:12:15.464602 IP S > C: Flags [S.], seq 253516766, ack 4202415602, win 65535, options [nop,nop,md5 valid,mss 1400,nop,nop,sackOK,nop,wscale 8], length 0
10:12:15.464611 IP C > S: Flags [.], ack 1, win 256, options [nop,nop,md5 valid], length 0
10:12:15.464678 IP C > S: Flags [P.], seq 1:13, ack 1, win 256, options [nop,nop,md5 valid], length 12
10:12:15.464685 IP S > C: Flags [.], ack 13, win 65535, options [nop,nop,md5 valid], length 0
After this patch the exchange looks saner :
11:59:59.882990 IP C > S: Flags [S], seq 517075944, win 65535, options [nop,nop,md5 valid,mss 1400,sackOK,TS val 1751508483 ecr 0,nop,wscale 8], length 0
11:59:59.883002 IP S > C: Flags [S.], seq 1902939253, ack 517075945, win 65535, options [nop,nop,md5 valid,mss 1400,sackOK,TS val 1751508479 ecr 1751508483,nop,wscale 8], length 0
11:59:59.883012 IP C > S: Flags [.], ack 1, win 256, options [nop,nop,md5 valid,nop,nop,TS val 1751508483 ecr 1751508479], length 0
11:59:59.883114 IP C > S: Flags [P.], seq 1:13, ack 1, win 256, options [nop,nop,md5 valid,nop,nop,TS val 1751508483 ecr 1751508479], length 12
11:59:59.883122 IP S > C: Flags [.], ack 13, win 256, options [nop,nop,md5 valid,nop,nop,TS val 1751508483 ecr 1751508483], length 0
11:59:59.883152 IP S > C: Flags [P.], seq 1:13, ack 13, win 256, options [nop,nop,md5 valid,nop,nop,TS val 1751508484 ecr 1751508483], length 12
11:59:59.883170 IP C > S: Flags [.], ack 13, win 256, options [nop,nop,md5 valid,nop,nop,TS val 1751508484 ecr 1751508484], length 0
Of course, no SACK block will ever be added later, but nothing should break.
Technically, we could remove the 4 nops included in MD5+TS options,
but again some stacks could break seeing not conventional alignment.
Fixes: 4957faade11b ("TCPCT part 1g: Responder Cookie => Initiator")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Florian Westphal <fw@strlen.de>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2020-07-02 03:41:23 +08:00
|
|
|
struct tcp_fastopen_cookie *foc,
|
2020-08-21 03:00:52 +08:00
|
|
|
enum tcp_synack_type synack_type,
|
|
|
|
|
struct sk_buff *syn_skb)
|
2009-12-03 02:25:27 +08:00
|
|
|
{
|
2008-07-19 15:04:31 +08:00
|
|
|
struct inet_request_sock *ireq = inet_rsk(req);
|
2012-04-15 13:58:06 +08:00
|
|
|
unsigned int remaining = MAX_TCP_OPTION_SPACE;
|
2008-07-19 15:04:31 +08:00
|
|
|
|
2006-11-15 11:07:45 +08:00
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
2015-03-25 06:58:52 +08:00
|
|
|
if (md5) {
|
2008-07-19 15:04:31 +08:00
|
|
|
opts->options |= OPTION_MD5;
|
2009-12-03 02:25:27 +08:00
|
|
|
remaining -= TCPOLEN_MD5SIG_ALIGNED;
|
|
|
|
|
|
|
|
|
|
/* We can't fit any SACK blocks in a packet with MD5 + TS
|
|
|
|
|
* options. There was discussion about disabling SACK
|
|
|
|
|
* rather than TS in order to fit in better with old,
|
|
|
|
|
* buggy kernels, but that was deemed to be unnecessary.
|
|
|
|
|
*/
|
tcp: md5: do not send silly options in SYNCOOKIES
Whenever cookie_init_timestamp() has been used to encode
ECN,SACK,WSCALE options, we can not remove the TS option in the SYNACK.
Otherwise, tcp_synack_options() will still advertize options like WSCALE
that we can not deduce later when receiving the packet from the client
to complete 3WHS.
Note that modern linux TCP stacks wont use MD5+TS+SACK in a SYN packet,
but we can not know for sure that all TCP stacks have the same logic.
Before the fix a tcpdump would exhibit this wrong exchange :
10:12:15.464591 IP C > S: Flags [S], seq 4202415601, win 65535, options [nop,nop,md5 valid,mss 1400,sackOK,TS val 456965269 ecr 0,nop,wscale 8], length 0
10:12:15.464602 IP S > C: Flags [S.], seq 253516766, ack 4202415602, win 65535, options [nop,nop,md5 valid,mss 1400,nop,nop,sackOK,nop,wscale 8], length 0
10:12:15.464611 IP C > S: Flags [.], ack 1, win 256, options [nop,nop,md5 valid], length 0
10:12:15.464678 IP C > S: Flags [P.], seq 1:13, ack 1, win 256, options [nop,nop,md5 valid], length 12
10:12:15.464685 IP S > C: Flags [.], ack 13, win 65535, options [nop,nop,md5 valid], length 0
After this patch the exchange looks saner :
11:59:59.882990 IP C > S: Flags [S], seq 517075944, win 65535, options [nop,nop,md5 valid,mss 1400,sackOK,TS val 1751508483 ecr 0,nop,wscale 8], length 0
11:59:59.883002 IP S > C: Flags [S.], seq 1902939253, ack 517075945, win 65535, options [nop,nop,md5 valid,mss 1400,sackOK,TS val 1751508479 ecr 1751508483,nop,wscale 8], length 0
11:59:59.883012 IP C > S: Flags [.], ack 1, win 256, options [nop,nop,md5 valid,nop,nop,TS val 1751508483 ecr 1751508479], length 0
11:59:59.883114 IP C > S: Flags [P.], seq 1:13, ack 1, win 256, options [nop,nop,md5 valid,nop,nop,TS val 1751508483 ecr 1751508479], length 12
11:59:59.883122 IP S > C: Flags [.], ack 13, win 256, options [nop,nop,md5 valid,nop,nop,TS val 1751508483 ecr 1751508483], length 0
11:59:59.883152 IP S > C: Flags [P.], seq 1:13, ack 13, win 256, options [nop,nop,md5 valid,nop,nop,TS val 1751508484 ecr 1751508483], length 12
11:59:59.883170 IP C > S: Flags [.], ack 13, win 256, options [nop,nop,md5 valid,nop,nop,TS val 1751508484 ecr 1751508484], length 0
Of course, no SACK block will ever be added later, but nothing should break.
Technically, we could remove the 4 nops included in MD5+TS options,
but again some stacks could break seeing not conventional alignment.
Fixes: 4957faade11b ("TCPCT part 1g: Responder Cookie => Initiator")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Florian Westphal <fw@strlen.de>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2020-07-02 03:41:23 +08:00
|
|
|
if (synack_type != TCP_SYNACK_COOKIE)
|
|
|
|
|
ireq->tstamp_ok &= !ireq->sack_ok;
|
2006-11-15 11:07:45 +08:00
|
|
|
}
|
|
|
|
|
#endif
|
2008-07-19 15:04:31 +08:00
|
|
|
|
2009-12-03 02:25:27 +08:00
|
|
|
/* We always send an MSS option. */
|
2008-07-19 15:04:31 +08:00
|
|
|
opts->mss = mss;
|
2009-12-03 02:25:27 +08:00
|
|
|
remaining -= TCPOLEN_MSS_ALIGNED;
|
2008-07-19 15:04:31 +08:00
|
|
|
|
|
|
|
|
if (likely(ireq->wscale_ok)) {
|
|
|
|
|
opts->ws = ireq->rcv_wscale;
|
2009-10-01 14:41:59 +08:00
|
|
|
opts->options |= OPTION_WSCALE;
|
2009-12-03 02:25:27 +08:00
|
|
|
remaining -= TCPOLEN_WSCALE_ALIGNED;
|
2008-07-19 15:04:31 +08:00
|
|
|
}
|
2010-05-18 13:35:36 +08:00
|
|
|
if (likely(ireq->tstamp_ok)) {
|
2008-07-19 15:04:31 +08:00
|
|
|
opts->options |= OPTION_TS;
|
2016-12-01 18:32:06 +08:00
|
|
|
opts->tsval = tcp_skb_timestamp(skb) + tcp_rsk(req)->ts_off;
|
2008-07-19 15:04:31 +08:00
|
|
|
opts->tsecr = req->ts_recent;
|
2009-12-03 02:25:27 +08:00
|
|
|
remaining -= TCPOLEN_TSTAMP_ALIGNED;
|
2008-07-19 15:04:31 +08:00
|
|
|
}
|
|
|
|
|
if (likely(ireq->sack_ok)) {
|
|
|
|
|
opts->options |= OPTION_SACK_ADVERTISE;
|
2010-05-18 13:35:36 +08:00
|
|
|
if (unlikely(!ireq->tstamp_ok))
|
2009-12-03 02:25:27 +08:00
|
|
|
remaining -= TCPOLEN_SACKPERM_ALIGNED;
|
2008-07-19 15:04:31 +08:00
|
|
|
}
|
2015-04-07 05:37:26 +08:00
|
|
|
if (foc != NULL && foc->len >= 0) {
|
|
|
|
|
u32 need = foc->len;
|
|
|
|
|
|
|
|
|
|
need += foc->exp ? TCPOLEN_EXP_FASTOPEN_BASE :
|
|
|
|
|
TCPOLEN_FASTOPEN_BASE;
|
2012-08-31 20:29:12 +08:00
|
|
|
need = (need + 3) & ~3U; /* Align to 32 bits */
|
|
|
|
|
if (remaining >= need) {
|
|
|
|
|
opts->options |= OPTION_FAST_OPEN_COOKIE;
|
|
|
|
|
opts->fastopen_cookie = foc;
|
|
|
|
|
remaining -= need;
|
|
|
|
|
}
|
|
|
|
|
}
|
2013-03-17 16:23:34 +08:00
|
|
|
|
2020-01-22 08:56:18 +08:00
|
|
|
mptcp_set_option_cond(req, opts, &remaining);
|
|
|
|
|
|
2017-10-25 17:01:45 +08:00
|
|
|
smc_set_option_cond(tcp_sk(sk), ireq, opts, &remaining);
|
|
|
|
|
|
2020-08-21 03:00:52 +08:00
|
|
|
bpf_skops_hdr_opt_len((struct sock *)sk, skb, req, syn_skb,
|
|
|
|
|
synack_type, opts, &remaining);
|
|
|
|
|
|
2009-12-03 02:25:27 +08:00
|
|
|
return MAX_TCP_OPTION_SPACE - remaining;
|
2008-07-19 15:04:31 +08:00
|
|
|
}
|
|
|
|
|
|
2009-07-22 07:00:40 +08:00
|
|
|
/* Compute TCP options for ESTABLISHED sockets. This is not the
|
|
|
|
|
* final wire format yet.
|
|
|
|
|
*/
|
2012-04-15 13:58:06 +08:00
|
|
|
static unsigned int tcp_established_options(struct sock *sk, struct sk_buff *skb,
|
2008-07-19 15:04:31 +08:00
|
|
|
struct tcp_out_options *opts,
|
2011-10-21 17:22:42 +08:00
|
|
|
struct tcp_md5sig_key **md5)
|
|
|
|
|
{
|
2008-07-19 15:04:31 +08:00
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
2012-04-15 13:58:06 +08:00
|
|
|
unsigned int size = 0;
|
2009-02-28 12:44:38 +08:00
|
|
|
unsigned int eff_sacks;
|
2008-07-19 15:04:31 +08:00
|
|
|
|
2013-10-01 04:29:11 +08:00
|
|
|
opts->options = 0;
|
|
|
|
|
|
2018-04-24 05:46:25 +08:00
|
|
|
*md5 = NULL;
|
2008-07-19 15:04:31 +08:00
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
2019-02-27 01:49:11 +08:00
|
|
|
if (static_branch_unlikely(&tcp_md5_needed) &&
|
2018-11-28 07:03:21 +08:00
|
|
|
rcu_access_pointer(tp->md5sig_info)) {
|
2018-04-24 05:46:25 +08:00
|
|
|
*md5 = tp->af_specific->md5_lookup(sk, sk);
|
|
|
|
|
if (*md5) {
|
|
|
|
|
opts->options |= OPTION_MD5;
|
|
|
|
|
size += TCPOLEN_MD5SIG_ALIGNED;
|
|
|
|
|
}
|
2008-07-19 15:04:31 +08:00
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
if (likely(tp->rx_opt.tstamp_ok)) {
|
|
|
|
|
opts->options |= OPTION_TS;
|
2014-09-06 06:33:33 +08:00
|
|
|
opts->tsval = skb ? tcp_skb_timestamp(skb) + tp->tsoffset : 0;
|
2008-07-19 15:04:31 +08:00
|
|
|
opts->tsecr = tp->rx_opt.ts_recent;
|
|
|
|
|
size += TCPOLEN_TSTAMP_ALIGNED;
|
|
|
|
|
}
|
|
|
|
|
|
2020-01-22 08:56:18 +08:00
|
|
|
/* MPTCP options have precedence over SACK for the limited TCP
|
|
|
|
|
* option space because a MPTCP connection would be forced to
|
|
|
|
|
* fall back to regular TCP if a required multipath option is
|
|
|
|
|
* missing. SACK still gets a chance to use whatever space is
|
|
|
|
|
* left.
|
|
|
|
|
*/
|
|
|
|
|
if (sk_is_mptcp(sk)) {
|
|
|
|
|
unsigned int remaining = MAX_TCP_OPTION_SPACE - size;
|
|
|
|
|
unsigned int opt_size = 0;
|
|
|
|
|
|
|
|
|
|
if (mptcp_established_options(sk, skb, &opt_size, remaining,
|
|
|
|
|
&opts->mptcp)) {
|
|
|
|
|
opts->options |= OPTION_MPTCP;
|
|
|
|
|
size += opt_size;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2009-02-28 12:44:38 +08:00
|
|
|
eff_sacks = tp->rx_opt.num_sacks + tp->rx_opt.dsack;
|
|
|
|
|
if (unlikely(eff_sacks)) {
|
2012-04-15 13:58:06 +08:00
|
|
|
const unsigned int remaining = MAX_TCP_OPTION_SPACE - size;
|
2020-01-09 23:59:22 +08:00
|
|
|
if (unlikely(remaining < TCPOLEN_SACK_BASE_ALIGNED +
|
|
|
|
|
TCPOLEN_SACK_PERBLOCK))
|
|
|
|
|
return size;
|
|
|
|
|
|
2008-07-19 15:04:31 +08:00
|
|
|
opts->num_sack_blocks =
|
2012-04-15 13:58:06 +08:00
|
|
|
min_t(unsigned int, eff_sacks,
|
2008-07-19 15:04:31 +08:00
|
|
|
(remaining - TCPOLEN_SACK_BASE_ALIGNED) /
|
|
|
|
|
TCPOLEN_SACK_PERBLOCK);
|
2020-01-09 23:59:22 +08:00
|
|
|
|
|
|
|
|
size += TCPOLEN_SACK_BASE_ALIGNED +
|
|
|
|
|
opts->num_sack_blocks * TCPOLEN_SACK_PERBLOCK;
|
2008-07-19 15:04:31 +08:00
|
|
|
}
|
|
|
|
|
|
2020-08-21 03:00:52 +08:00
|
|
|
if (unlikely(BPF_SOCK_OPS_TEST_FLAG(tp,
|
|
|
|
|
BPF_SOCK_OPS_WRITE_HDR_OPT_CB_FLAG))) {
|
|
|
|
|
unsigned int remaining = MAX_TCP_OPTION_SPACE - size;
|
|
|
|
|
|
|
|
|
|
bpf_skops_hdr_opt_len(sk, skb, NULL, NULL, 0, opts, &remaining);
|
|
|
|
|
|
|
|
|
|
size = MAX_TCP_OPTION_SPACE - remaining;
|
|
|
|
|
}
|
|
|
|
|
|
2008-07-19 15:04:31 +08:00
|
|
|
return size;
|
2006-01-04 08:03:49 +08:00
|
|
|
}
|
2005-04-17 06:20:36 +08:00
|
|
|
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
|
|
|
|
|
/* TCP SMALL QUEUES (TSQ)
|
|
|
|
|
*
|
|
|
|
|
* TSQ goal is to keep small amount of skbs per tcp flow in tx queues (qdisc+dev)
|
|
|
|
|
* to reduce RTT and bufferbloat.
|
|
|
|
|
* We do this using a special skb destructor (tcp_wfree).
|
|
|
|
|
*
|
|
|
|
|
* Its important tcp_wfree() can be replaced by sock_wfree() in the event skb
|
|
|
|
|
* needs to be reallocated in a driver.
|
2013-12-09 04:15:44 +08:00
|
|
|
* The invariant being skb->truesize subtracted from sk->sk_wmem_alloc
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
*
|
|
|
|
|
* Since transmit from skb destructor is forbidden, we use a tasklet
|
|
|
|
|
* to process all sockets that eventually need to send more skbs.
|
|
|
|
|
* We use one tasklet per cpu, with its own queue of sockets.
|
|
|
|
|
*/
|
|
|
|
|
struct tsq_tasklet {
|
|
|
|
|
struct tasklet_struct tasklet;
|
|
|
|
|
struct list_head head; /* queue of tcp sockets */
|
|
|
|
|
};
|
|
|
|
|
static DEFINE_PER_CPU(struct tsq_tasklet, tsq_tasklet);
|
|
|
|
|
|
2018-05-11 05:59:43 +08:00
|
|
|
static void tcp_tsq_write(struct sock *sk)
|
2012-07-20 13:45:50 +08:00
|
|
|
{
|
|
|
|
|
if ((1 << sk->sk_state) &
|
|
|
|
|
(TCPF_ESTABLISHED | TCPF_FIN_WAIT1 | TCPF_CLOSING |
|
2016-09-21 13:45:58 +08:00
|
|
|
TCPF_CLOSE_WAIT | TCPF_LAST_ACK)) {
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
|
|
|
|
|
if (tp->lost_out > tp->retrans_out &&
|
2017-10-22 12:13:16 +08:00
|
|
|
tp->snd_cwnd > tcp_packets_in_flight(tp)) {
|
|
|
|
|
tcp_mstamp_refresh(tp);
|
2016-09-21 13:45:58 +08:00
|
|
|
tcp_xmit_retransmit_queue(sk);
|
2017-10-22 12:13:16 +08:00
|
|
|
}
|
2016-09-21 13:45:58 +08:00
|
|
|
|
|
|
|
|
tcp_write_xmit(sk, tcp_current_mss(sk), tp->nonagle,
|
2014-02-10 10:40:11 +08:00
|
|
|
0, GFP_ATOMIC);
|
2016-09-21 13:45:58 +08:00
|
|
|
}
|
2012-07-20 13:45:50 +08:00
|
|
|
}
|
2018-05-11 05:59:43 +08:00
|
|
|
|
|
|
|
|
static void tcp_tsq_handler(struct sock *sk)
|
|
|
|
|
{
|
|
|
|
|
bh_lock_sock(sk);
|
|
|
|
|
if (!sock_owned_by_user(sk))
|
|
|
|
|
tcp_tsq_write(sk);
|
|
|
|
|
else if (!test_and_set_bit(TCP_TSQ_DEFERRED, &sk->sk_tsq_flags))
|
|
|
|
|
sock_hold(sk);
|
|
|
|
|
bh_unlock_sock(sk);
|
|
|
|
|
}
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
/*
|
2013-12-09 04:15:44 +08:00
|
|
|
* One tasklet per cpu tries to send more skbs.
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
* We run in tasklet context but need to disable irqs when
|
2013-12-09 04:15:44 +08:00
|
|
|
* transferring tsq->head because tcp_wfree() might
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
* interrupt us (non NAPI drivers)
|
|
|
|
|
*/
|
2020-11-03 17:18:17 +08:00
|
|
|
static void tcp_tasklet_func(struct tasklet_struct *t)
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
{
|
2020-11-03 17:18:17 +08:00
|
|
|
struct tsq_tasklet *tsq = from_tasklet(tsq, t, tasklet);
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
LIST_HEAD(list);
|
|
|
|
|
unsigned long flags;
|
|
|
|
|
struct list_head *q, *n;
|
|
|
|
|
struct tcp_sock *tp;
|
|
|
|
|
struct sock *sk;
|
|
|
|
|
|
|
|
|
|
local_irq_save(flags);
|
|
|
|
|
list_splice_init(&tsq->head, &list);
|
|
|
|
|
local_irq_restore(flags);
|
|
|
|
|
|
|
|
|
|
list_for_each_safe(q, n, &list) {
|
|
|
|
|
tp = list_entry(q, struct tcp_sock, tsq_node);
|
|
|
|
|
list_del(&tp->tsq_node);
|
|
|
|
|
|
|
|
|
|
sk = (struct sock *)tp;
|
2016-12-21 21:42:43 +08:00
|
|
|
smp_mb__before_atomic();
|
2016-12-04 03:14:57 +08:00
|
|
|
clear_bit(TSQ_QUEUED, &sk->sk_tsq_flags);
|
|
|
|
|
|
2018-05-11 05:59:43 +08:00
|
|
|
tcp_tsq_handler(sk);
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
sk_free(sk);
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2016-12-04 03:14:50 +08:00
|
|
|
#define TCP_DEFERRED_ALL (TCPF_TSQ_DEFERRED | \
|
|
|
|
|
TCPF_WRITE_TIMER_DEFERRED | \
|
|
|
|
|
TCPF_DELACK_TIMER_DEFERRED | \
|
|
|
|
|
TCPF_MTU_REDUCED_DEFERRED)
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
/**
|
|
|
|
|
* tcp_release_cb - tcp release_sock() callback
|
|
|
|
|
* @sk: socket
|
|
|
|
|
*
|
|
|
|
|
* called from release_sock() to perform protocol dependent
|
|
|
|
|
* actions before socket release.
|
|
|
|
|
*/
|
|
|
|
|
void tcp_release_cb(struct sock *sk)
|
|
|
|
|
{
|
2012-07-20 13:45:50 +08:00
|
|
|
unsigned long flags, nflags;
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
|
2012-07-20 13:45:50 +08:00
|
|
|
/* perform an atomic operation only if at least one flag is set */
|
|
|
|
|
do {
|
2016-12-04 03:14:57 +08:00
|
|
|
flags = sk->sk_tsq_flags;
|
2012-07-20 13:45:50 +08:00
|
|
|
if (!(flags & TCP_DEFERRED_ALL))
|
|
|
|
|
return;
|
|
|
|
|
nflags = flags & ~TCP_DEFERRED_ALL;
|
2016-12-04 03:14:57 +08:00
|
|
|
} while (cmpxchg(&sk->sk_tsq_flags, flags, nflags) != flags);
|
2012-07-20 13:45:50 +08:00
|
|
|
|
2018-05-11 05:59:43 +08:00
|
|
|
if (flags & TCPF_TSQ_DEFERRED) {
|
|
|
|
|
tcp_tsq_write(sk);
|
|
|
|
|
__sock_put(sk);
|
|
|
|
|
}
|
tcp: tcp_release_cb() should release socket ownership
Lars Persson reported following deadlock :
-000 |M:0x0:0x802B6AF8(asm) <-- arch_spin_lock
-001 |tcp_v4_rcv(skb = 0x8BD527A0) <-- sk = 0x8BE6B2A0
-002 |ip_local_deliver_finish(skb = 0x8BD527A0)
-003 |__netif_receive_skb_core(skb = 0x8BD527A0, ?)
-004 |netif_receive_skb(skb = 0x8BD527A0)
-005 |elk_poll(napi = 0x8C770500, budget = 64)
-006 |net_rx_action(?)
-007 |__do_softirq()
-008 |do_softirq()
-009 |local_bh_enable()
-010 |tcp_rcv_established(sk = 0x8BE6B2A0, skb = 0x87D3A9E0, th = 0x814EBE14, ?)
-011 |tcp_v4_do_rcv(sk = 0x8BE6B2A0, skb = 0x87D3A9E0)
-012 |tcp_delack_timer_handler(sk = 0x8BE6B2A0)
-013 |tcp_release_cb(sk = 0x8BE6B2A0)
-014 |release_sock(sk = 0x8BE6B2A0)
-015 |tcp_sendmsg(?, sk = 0x8BE6B2A0, ?, ?)
-016 |sock_sendmsg(sock = 0x8518C4C0, msg = 0x87D8DAA8, size = 4096)
-017 |kernel_sendmsg(?, ?, ?, ?, size = 4096)
-018 |smb_send_kvec()
-019 |smb_send_rqst(server = 0x87C4D400, rqst = 0x87D8DBA0)
-020 |cifs_call_async()
-021 |cifs_async_writev(wdata = 0x87FD6580)
-022 |cifs_writepages(mapping = 0x852096E4, wbc = 0x87D8DC88)
-023 |__writeback_single_inode(inode = 0x852095D0, wbc = 0x87D8DC88)
-024 |writeback_sb_inodes(sb = 0x87D6D800, wb = 0x87E4A9C0, work = 0x87D8DD88)
-025 |__writeback_inodes_wb(wb = 0x87E4A9C0, work = 0x87D8DD88)
-026 |wb_writeback(wb = 0x87E4A9C0, work = 0x87D8DD88)
-027 |wb_do_writeback(wb = 0x87E4A9C0, force_wait = 0)
-028 |bdi_writeback_workfn(work = 0x87E4A9CC)
-029 |process_one_work(worker = 0x8B045880, work = 0x87E4A9CC)
-030 |worker_thread(__worker = 0x8B045880)
-031 |kthread(_create = 0x87CADD90)
-032 |ret_from_kernel_thread(asm)
Bug occurs because __tcp_checksum_complete_user() enables BH, assuming
it is running from softirq context.
Lars trace involved a NIC without RX checksum support but other points
are problematic as well, like the prequeue stuff.
Problem is triggered by a timer, that found socket being owned by user.
tcp_release_cb() should call tcp_write_timer_handler() or
tcp_delack_timer_handler() in the appropriate context :
BH disabled and socket lock held, but 'owned' field cleared,
as if they were running from timer handlers.
Fixes: 6f458dfb4092 ("tcp: improve latencies of timer triggered events")
Reported-by: Lars Persson <lars.persson@axis.com>
Tested-by: Lars Persson <lars.persson@axis.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-03-11 00:50:11 +08:00
|
|
|
/* Here begins the tricky part :
|
|
|
|
|
* We are called from release_sock() with :
|
|
|
|
|
* 1) BH disabled
|
|
|
|
|
* 2) sk_lock.slock spinlock held
|
|
|
|
|
* 3) socket owned by us (sk->sk_lock.owned == 1)
|
|
|
|
|
*
|
|
|
|
|
* But following code is meant to be called from BH handlers,
|
|
|
|
|
* so we should keep BH disabled, but early release socket ownership
|
|
|
|
|
*/
|
|
|
|
|
sock_release_ownership(sk);
|
|
|
|
|
|
2016-12-04 03:14:50 +08:00
|
|
|
if (flags & TCPF_WRITE_TIMER_DEFERRED) {
|
2012-07-20 13:45:50 +08:00
|
|
|
tcp_write_timer_handler(sk);
|
tcp: fix possible socket refcount problem
Commit 6f458dfb40 (tcp: improve latencies of timer triggered events)
added bug leading to following trace :
[ 2866.131281] IPv4: Attempt to release TCP socket in state 1 ffff880019ec0000
[ 2866.131726]
[ 2866.132188] =========================
[ 2866.132281] [ BUG: held lock freed! ]
[ 2866.132281] 3.6.0-rc1+ #622 Not tainted
[ 2866.132281] -------------------------
[ 2866.132281] kworker/0:1/652 is freeing memory ffff880019ec0000-ffff880019ec0a1f, with a lock still held there!
[ 2866.132281] (sk_lock-AF_INET-RPC){+.+...}, at: [<ffffffff81903619>] tcp_sendmsg+0x29/0xcc6
[ 2866.132281] 4 locks held by kworker/0:1/652:
[ 2866.132281] #0: (rpciod){.+.+.+}, at: [<ffffffff81083567>] process_one_work+0x1de/0x47f
[ 2866.132281] #1: ((&task->u.tk_work)){+.+.+.}, at: [<ffffffff81083567>] process_one_work+0x1de/0x47f
[ 2866.132281] #2: (sk_lock-AF_INET-RPC){+.+...}, at: [<ffffffff81903619>] tcp_sendmsg+0x29/0xcc6
[ 2866.132281] #3: (&icsk->icsk_retransmit_timer){+.-...}, at: [<ffffffff81078017>] run_timer_softirq+0x1ad/0x35f
[ 2866.132281]
[ 2866.132281] stack backtrace:
[ 2866.132281] Pid: 652, comm: kworker/0:1 Not tainted 3.6.0-rc1+ #622
[ 2866.132281] Call Trace:
[ 2866.132281] <IRQ> [<ffffffff810bc527>] debug_check_no_locks_freed+0x112/0x159
[ 2866.132281] [<ffffffff818a0839>] ? __sk_free+0xfd/0x114
[ 2866.132281] [<ffffffff811549fa>] kmem_cache_free+0x6b/0x13a
[ 2866.132281] [<ffffffff818a0839>] __sk_free+0xfd/0x114
[ 2866.132281] [<ffffffff818a08c0>] sk_free+0x1c/0x1e
[ 2866.132281] [<ffffffff81911e1c>] tcp_write_timer+0x51/0x56
[ 2866.132281] [<ffffffff81078082>] run_timer_softirq+0x218/0x35f
[ 2866.132281] [<ffffffff81078017>] ? run_timer_softirq+0x1ad/0x35f
[ 2866.132281] [<ffffffff810f5831>] ? rb_commit+0x58/0x85
[ 2866.132281] [<ffffffff81911dcb>] ? tcp_write_timer_handler+0x148/0x148
[ 2866.132281] [<ffffffff81070bd6>] __do_softirq+0xcb/0x1f9
[ 2866.132281] [<ffffffff81a0a00c>] ? _raw_spin_unlock+0x29/0x2e
[ 2866.132281] [<ffffffff81a1227c>] call_softirq+0x1c/0x30
[ 2866.132281] [<ffffffff81039f38>] do_softirq+0x4a/0xa6
[ 2866.132281] [<ffffffff81070f2b>] irq_exit+0x51/0xad
[ 2866.132281] [<ffffffff81a129cd>] do_IRQ+0x9d/0xb4
[ 2866.132281] [<ffffffff81a0a3ef>] common_interrupt+0x6f/0x6f
[ 2866.132281] <EOI> [<ffffffff8109d006>] ? sched_clock_cpu+0x58/0xd1
[ 2866.132281] [<ffffffff81a0a172>] ? _raw_spin_unlock_irqrestore+0x4c/0x56
[ 2866.132281] [<ffffffff81078692>] mod_timer+0x178/0x1a9
[ 2866.132281] [<ffffffff818a00aa>] sk_reset_timer+0x19/0x26
[ 2866.132281] [<ffffffff8190b2cc>] tcp_rearm_rto+0x99/0xa4
[ 2866.132281] [<ffffffff8190dfba>] tcp_event_new_data_sent+0x6e/0x70
[ 2866.132281] [<ffffffff8190f7ea>] tcp_write_xmit+0x7de/0x8e4
[ 2866.132281] [<ffffffff818a565d>] ? __alloc_skb+0xa0/0x1a1
[ 2866.132281] [<ffffffff8190f952>] __tcp_push_pending_frames+0x2e/0x8a
[ 2866.132281] [<ffffffff81904122>] tcp_sendmsg+0xb32/0xcc6
[ 2866.132281] [<ffffffff819229c2>] inet_sendmsg+0xaa/0xd5
[ 2866.132281] [<ffffffff81922918>] ? inet_autobind+0x5f/0x5f
[ 2866.132281] [<ffffffff810ee7f1>] ? trace_clock_local+0x9/0xb
[ 2866.132281] [<ffffffff8189adab>] sock_sendmsg+0xa3/0xc4
[ 2866.132281] [<ffffffff810f5de6>] ? rb_reserve_next_event+0x26f/0x2d5
[ 2866.132281] [<ffffffff8103e6a9>] ? native_sched_clock+0x29/0x6f
[ 2866.132281] [<ffffffff8103e6f8>] ? sched_clock+0x9/0xd
[ 2866.132281] [<ffffffff810ee7f1>] ? trace_clock_local+0x9/0xb
[ 2866.132281] [<ffffffff8189ae03>] kernel_sendmsg+0x37/0x43
[ 2866.132281] [<ffffffff8199ce49>] xs_send_kvec+0x77/0x80
[ 2866.132281] [<ffffffff8199cec1>] xs_sendpages+0x6f/0x1a0
[ 2866.132281] [<ffffffff8107826d>] ? try_to_del_timer_sync+0x55/0x61
[ 2866.132281] [<ffffffff8199d0d2>] xs_tcp_send_request+0x55/0xf1
[ 2866.132281] [<ffffffff8199bb90>] xprt_transmit+0x89/0x1db
[ 2866.132281] [<ffffffff81999bcd>] ? call_connect+0x3c/0x3c
[ 2866.132281] [<ffffffff81999d92>] call_transmit+0x1c5/0x20e
[ 2866.132281] [<ffffffff819a0d55>] __rpc_execute+0x6f/0x225
[ 2866.132281] [<ffffffff81999bcd>] ? call_connect+0x3c/0x3c
[ 2866.132281] [<ffffffff819a0f33>] rpc_async_schedule+0x28/0x34
[ 2866.132281] [<ffffffff810835d6>] process_one_work+0x24d/0x47f
[ 2866.132281] [<ffffffff81083567>] ? process_one_work+0x1de/0x47f
[ 2866.132281] [<ffffffff819a0f0b>] ? __rpc_execute+0x225/0x225
[ 2866.132281] [<ffffffff81083a6d>] worker_thread+0x236/0x317
[ 2866.132281] [<ffffffff81083837>] ? process_scheduled_works+0x2f/0x2f
[ 2866.132281] [<ffffffff8108b7b8>] kthread+0x9a/0xa2
[ 2866.132281] [<ffffffff81a12184>] kernel_thread_helper+0x4/0x10
[ 2866.132281] [<ffffffff81a0a4b0>] ? retint_restore_args+0x13/0x13
[ 2866.132281] [<ffffffff8108b71e>] ? __init_kthread_worker+0x5a/0x5a
[ 2866.132281] [<ffffffff81a12180>] ? gs_change+0x13/0x13
[ 2866.308506] IPv4: Attempt to release TCP socket in state 1 ffff880019ec0000
[ 2866.309689] =============================================================================
[ 2866.310254] BUG TCP (Not tainted): Object already free
[ 2866.310254] -----------------------------------------------------------------------------
[ 2866.310254]
The bug comes from the fact that timer set in sk_reset_timer() can run
before we actually do the sock_hold(). socket refcount reaches zero and
we free the socket too soon.
timer handler is not allowed to reduce socket refcnt if socket is owned
by the user, or we need to change sk_reset_timer() implementation.
We should take a reference on the socket in case TCP_DELACK_TIMER_DEFERRED
or TCP_DELACK_TIMER_DEFERRED bit are set in tsq_flags
Also fix a typo in tcp_delack_timer(), where TCP_WRITE_TIMER_DEFERRED
was used instead of TCP_DELACK_TIMER_DEFERRED.
For consistency, use same socket refcount change for TCP_MTU_REDUCED_DEFERRED,
even if not fired from a timer.
Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Tested-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-08-20 08:22:46 +08:00
|
|
|
__sock_put(sk);
|
|
|
|
|
}
|
2016-12-04 03:14:50 +08:00
|
|
|
if (flags & TCPF_DELACK_TIMER_DEFERRED) {
|
2012-07-20 13:45:50 +08:00
|
|
|
tcp_delack_timer_handler(sk);
|
tcp: fix possible socket refcount problem
Commit 6f458dfb40 (tcp: improve latencies of timer triggered events)
added bug leading to following trace :
[ 2866.131281] IPv4: Attempt to release TCP socket in state 1 ffff880019ec0000
[ 2866.131726]
[ 2866.132188] =========================
[ 2866.132281] [ BUG: held lock freed! ]
[ 2866.132281] 3.6.0-rc1+ #622 Not tainted
[ 2866.132281] -------------------------
[ 2866.132281] kworker/0:1/652 is freeing memory ffff880019ec0000-ffff880019ec0a1f, with a lock still held there!
[ 2866.132281] (sk_lock-AF_INET-RPC){+.+...}, at: [<ffffffff81903619>] tcp_sendmsg+0x29/0xcc6
[ 2866.132281] 4 locks held by kworker/0:1/652:
[ 2866.132281] #0: (rpciod){.+.+.+}, at: [<ffffffff81083567>] process_one_work+0x1de/0x47f
[ 2866.132281] #1: ((&task->u.tk_work)){+.+.+.}, at: [<ffffffff81083567>] process_one_work+0x1de/0x47f
[ 2866.132281] #2: (sk_lock-AF_INET-RPC){+.+...}, at: [<ffffffff81903619>] tcp_sendmsg+0x29/0xcc6
[ 2866.132281] #3: (&icsk->icsk_retransmit_timer){+.-...}, at: [<ffffffff81078017>] run_timer_softirq+0x1ad/0x35f
[ 2866.132281]
[ 2866.132281] stack backtrace:
[ 2866.132281] Pid: 652, comm: kworker/0:1 Not tainted 3.6.0-rc1+ #622
[ 2866.132281] Call Trace:
[ 2866.132281] <IRQ> [<ffffffff810bc527>] debug_check_no_locks_freed+0x112/0x159
[ 2866.132281] [<ffffffff818a0839>] ? __sk_free+0xfd/0x114
[ 2866.132281] [<ffffffff811549fa>] kmem_cache_free+0x6b/0x13a
[ 2866.132281] [<ffffffff818a0839>] __sk_free+0xfd/0x114
[ 2866.132281] [<ffffffff818a08c0>] sk_free+0x1c/0x1e
[ 2866.132281] [<ffffffff81911e1c>] tcp_write_timer+0x51/0x56
[ 2866.132281] [<ffffffff81078082>] run_timer_softirq+0x218/0x35f
[ 2866.132281] [<ffffffff81078017>] ? run_timer_softirq+0x1ad/0x35f
[ 2866.132281] [<ffffffff810f5831>] ? rb_commit+0x58/0x85
[ 2866.132281] [<ffffffff81911dcb>] ? tcp_write_timer_handler+0x148/0x148
[ 2866.132281] [<ffffffff81070bd6>] __do_softirq+0xcb/0x1f9
[ 2866.132281] [<ffffffff81a0a00c>] ? _raw_spin_unlock+0x29/0x2e
[ 2866.132281] [<ffffffff81a1227c>] call_softirq+0x1c/0x30
[ 2866.132281] [<ffffffff81039f38>] do_softirq+0x4a/0xa6
[ 2866.132281] [<ffffffff81070f2b>] irq_exit+0x51/0xad
[ 2866.132281] [<ffffffff81a129cd>] do_IRQ+0x9d/0xb4
[ 2866.132281] [<ffffffff81a0a3ef>] common_interrupt+0x6f/0x6f
[ 2866.132281] <EOI> [<ffffffff8109d006>] ? sched_clock_cpu+0x58/0xd1
[ 2866.132281] [<ffffffff81a0a172>] ? _raw_spin_unlock_irqrestore+0x4c/0x56
[ 2866.132281] [<ffffffff81078692>] mod_timer+0x178/0x1a9
[ 2866.132281] [<ffffffff818a00aa>] sk_reset_timer+0x19/0x26
[ 2866.132281] [<ffffffff8190b2cc>] tcp_rearm_rto+0x99/0xa4
[ 2866.132281] [<ffffffff8190dfba>] tcp_event_new_data_sent+0x6e/0x70
[ 2866.132281] [<ffffffff8190f7ea>] tcp_write_xmit+0x7de/0x8e4
[ 2866.132281] [<ffffffff818a565d>] ? __alloc_skb+0xa0/0x1a1
[ 2866.132281] [<ffffffff8190f952>] __tcp_push_pending_frames+0x2e/0x8a
[ 2866.132281] [<ffffffff81904122>] tcp_sendmsg+0xb32/0xcc6
[ 2866.132281] [<ffffffff819229c2>] inet_sendmsg+0xaa/0xd5
[ 2866.132281] [<ffffffff81922918>] ? inet_autobind+0x5f/0x5f
[ 2866.132281] [<ffffffff810ee7f1>] ? trace_clock_local+0x9/0xb
[ 2866.132281] [<ffffffff8189adab>] sock_sendmsg+0xa3/0xc4
[ 2866.132281] [<ffffffff810f5de6>] ? rb_reserve_next_event+0x26f/0x2d5
[ 2866.132281] [<ffffffff8103e6a9>] ? native_sched_clock+0x29/0x6f
[ 2866.132281] [<ffffffff8103e6f8>] ? sched_clock+0x9/0xd
[ 2866.132281] [<ffffffff810ee7f1>] ? trace_clock_local+0x9/0xb
[ 2866.132281] [<ffffffff8189ae03>] kernel_sendmsg+0x37/0x43
[ 2866.132281] [<ffffffff8199ce49>] xs_send_kvec+0x77/0x80
[ 2866.132281] [<ffffffff8199cec1>] xs_sendpages+0x6f/0x1a0
[ 2866.132281] [<ffffffff8107826d>] ? try_to_del_timer_sync+0x55/0x61
[ 2866.132281] [<ffffffff8199d0d2>] xs_tcp_send_request+0x55/0xf1
[ 2866.132281] [<ffffffff8199bb90>] xprt_transmit+0x89/0x1db
[ 2866.132281] [<ffffffff81999bcd>] ? call_connect+0x3c/0x3c
[ 2866.132281] [<ffffffff81999d92>] call_transmit+0x1c5/0x20e
[ 2866.132281] [<ffffffff819a0d55>] __rpc_execute+0x6f/0x225
[ 2866.132281] [<ffffffff81999bcd>] ? call_connect+0x3c/0x3c
[ 2866.132281] [<ffffffff819a0f33>] rpc_async_schedule+0x28/0x34
[ 2866.132281] [<ffffffff810835d6>] process_one_work+0x24d/0x47f
[ 2866.132281] [<ffffffff81083567>] ? process_one_work+0x1de/0x47f
[ 2866.132281] [<ffffffff819a0f0b>] ? __rpc_execute+0x225/0x225
[ 2866.132281] [<ffffffff81083a6d>] worker_thread+0x236/0x317
[ 2866.132281] [<ffffffff81083837>] ? process_scheduled_works+0x2f/0x2f
[ 2866.132281] [<ffffffff8108b7b8>] kthread+0x9a/0xa2
[ 2866.132281] [<ffffffff81a12184>] kernel_thread_helper+0x4/0x10
[ 2866.132281] [<ffffffff81a0a4b0>] ? retint_restore_args+0x13/0x13
[ 2866.132281] [<ffffffff8108b71e>] ? __init_kthread_worker+0x5a/0x5a
[ 2866.132281] [<ffffffff81a12180>] ? gs_change+0x13/0x13
[ 2866.308506] IPv4: Attempt to release TCP socket in state 1 ffff880019ec0000
[ 2866.309689] =============================================================================
[ 2866.310254] BUG TCP (Not tainted): Object already free
[ 2866.310254] -----------------------------------------------------------------------------
[ 2866.310254]
The bug comes from the fact that timer set in sk_reset_timer() can run
before we actually do the sock_hold(). socket refcount reaches zero and
we free the socket too soon.
timer handler is not allowed to reduce socket refcnt if socket is owned
by the user, or we need to change sk_reset_timer() implementation.
We should take a reference on the socket in case TCP_DELACK_TIMER_DEFERRED
or TCP_DELACK_TIMER_DEFERRED bit are set in tsq_flags
Also fix a typo in tcp_delack_timer(), where TCP_WRITE_TIMER_DEFERRED
was used instead of TCP_DELACK_TIMER_DEFERRED.
For consistency, use same socket refcount change for TCP_MTU_REDUCED_DEFERRED,
even if not fired from a timer.
Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Tested-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-08-20 08:22:46 +08:00
|
|
|
__sock_put(sk);
|
|
|
|
|
}
|
2016-12-04 03:14:50 +08:00
|
|
|
if (flags & TCPF_MTU_REDUCED_DEFERRED) {
|
2014-08-15 00:40:05 +08:00
|
|
|
inet_csk(sk)->icsk_af_ops->mtu_reduced(sk);
|
tcp: fix possible socket refcount problem
Commit 6f458dfb40 (tcp: improve latencies of timer triggered events)
added bug leading to following trace :
[ 2866.131281] IPv4: Attempt to release TCP socket in state 1 ffff880019ec0000
[ 2866.131726]
[ 2866.132188] =========================
[ 2866.132281] [ BUG: held lock freed! ]
[ 2866.132281] 3.6.0-rc1+ #622 Not tainted
[ 2866.132281] -------------------------
[ 2866.132281] kworker/0:1/652 is freeing memory ffff880019ec0000-ffff880019ec0a1f, with a lock still held there!
[ 2866.132281] (sk_lock-AF_INET-RPC){+.+...}, at: [<ffffffff81903619>] tcp_sendmsg+0x29/0xcc6
[ 2866.132281] 4 locks held by kworker/0:1/652:
[ 2866.132281] #0: (rpciod){.+.+.+}, at: [<ffffffff81083567>] process_one_work+0x1de/0x47f
[ 2866.132281] #1: ((&task->u.tk_work)){+.+.+.}, at: [<ffffffff81083567>] process_one_work+0x1de/0x47f
[ 2866.132281] #2: (sk_lock-AF_INET-RPC){+.+...}, at: [<ffffffff81903619>] tcp_sendmsg+0x29/0xcc6
[ 2866.132281] #3: (&icsk->icsk_retransmit_timer){+.-...}, at: [<ffffffff81078017>] run_timer_softirq+0x1ad/0x35f
[ 2866.132281]
[ 2866.132281] stack backtrace:
[ 2866.132281] Pid: 652, comm: kworker/0:1 Not tainted 3.6.0-rc1+ #622
[ 2866.132281] Call Trace:
[ 2866.132281] <IRQ> [<ffffffff810bc527>] debug_check_no_locks_freed+0x112/0x159
[ 2866.132281] [<ffffffff818a0839>] ? __sk_free+0xfd/0x114
[ 2866.132281] [<ffffffff811549fa>] kmem_cache_free+0x6b/0x13a
[ 2866.132281] [<ffffffff818a0839>] __sk_free+0xfd/0x114
[ 2866.132281] [<ffffffff818a08c0>] sk_free+0x1c/0x1e
[ 2866.132281] [<ffffffff81911e1c>] tcp_write_timer+0x51/0x56
[ 2866.132281] [<ffffffff81078082>] run_timer_softirq+0x218/0x35f
[ 2866.132281] [<ffffffff81078017>] ? run_timer_softirq+0x1ad/0x35f
[ 2866.132281] [<ffffffff810f5831>] ? rb_commit+0x58/0x85
[ 2866.132281] [<ffffffff81911dcb>] ? tcp_write_timer_handler+0x148/0x148
[ 2866.132281] [<ffffffff81070bd6>] __do_softirq+0xcb/0x1f9
[ 2866.132281] [<ffffffff81a0a00c>] ? _raw_spin_unlock+0x29/0x2e
[ 2866.132281] [<ffffffff81a1227c>] call_softirq+0x1c/0x30
[ 2866.132281] [<ffffffff81039f38>] do_softirq+0x4a/0xa6
[ 2866.132281] [<ffffffff81070f2b>] irq_exit+0x51/0xad
[ 2866.132281] [<ffffffff81a129cd>] do_IRQ+0x9d/0xb4
[ 2866.132281] [<ffffffff81a0a3ef>] common_interrupt+0x6f/0x6f
[ 2866.132281] <EOI> [<ffffffff8109d006>] ? sched_clock_cpu+0x58/0xd1
[ 2866.132281] [<ffffffff81a0a172>] ? _raw_spin_unlock_irqrestore+0x4c/0x56
[ 2866.132281] [<ffffffff81078692>] mod_timer+0x178/0x1a9
[ 2866.132281] [<ffffffff818a00aa>] sk_reset_timer+0x19/0x26
[ 2866.132281] [<ffffffff8190b2cc>] tcp_rearm_rto+0x99/0xa4
[ 2866.132281] [<ffffffff8190dfba>] tcp_event_new_data_sent+0x6e/0x70
[ 2866.132281] [<ffffffff8190f7ea>] tcp_write_xmit+0x7de/0x8e4
[ 2866.132281] [<ffffffff818a565d>] ? __alloc_skb+0xa0/0x1a1
[ 2866.132281] [<ffffffff8190f952>] __tcp_push_pending_frames+0x2e/0x8a
[ 2866.132281] [<ffffffff81904122>] tcp_sendmsg+0xb32/0xcc6
[ 2866.132281] [<ffffffff819229c2>] inet_sendmsg+0xaa/0xd5
[ 2866.132281] [<ffffffff81922918>] ? inet_autobind+0x5f/0x5f
[ 2866.132281] [<ffffffff810ee7f1>] ? trace_clock_local+0x9/0xb
[ 2866.132281] [<ffffffff8189adab>] sock_sendmsg+0xa3/0xc4
[ 2866.132281] [<ffffffff810f5de6>] ? rb_reserve_next_event+0x26f/0x2d5
[ 2866.132281] [<ffffffff8103e6a9>] ? native_sched_clock+0x29/0x6f
[ 2866.132281] [<ffffffff8103e6f8>] ? sched_clock+0x9/0xd
[ 2866.132281] [<ffffffff810ee7f1>] ? trace_clock_local+0x9/0xb
[ 2866.132281] [<ffffffff8189ae03>] kernel_sendmsg+0x37/0x43
[ 2866.132281] [<ffffffff8199ce49>] xs_send_kvec+0x77/0x80
[ 2866.132281] [<ffffffff8199cec1>] xs_sendpages+0x6f/0x1a0
[ 2866.132281] [<ffffffff8107826d>] ? try_to_del_timer_sync+0x55/0x61
[ 2866.132281] [<ffffffff8199d0d2>] xs_tcp_send_request+0x55/0xf1
[ 2866.132281] [<ffffffff8199bb90>] xprt_transmit+0x89/0x1db
[ 2866.132281] [<ffffffff81999bcd>] ? call_connect+0x3c/0x3c
[ 2866.132281] [<ffffffff81999d92>] call_transmit+0x1c5/0x20e
[ 2866.132281] [<ffffffff819a0d55>] __rpc_execute+0x6f/0x225
[ 2866.132281] [<ffffffff81999bcd>] ? call_connect+0x3c/0x3c
[ 2866.132281] [<ffffffff819a0f33>] rpc_async_schedule+0x28/0x34
[ 2866.132281] [<ffffffff810835d6>] process_one_work+0x24d/0x47f
[ 2866.132281] [<ffffffff81083567>] ? process_one_work+0x1de/0x47f
[ 2866.132281] [<ffffffff819a0f0b>] ? __rpc_execute+0x225/0x225
[ 2866.132281] [<ffffffff81083a6d>] worker_thread+0x236/0x317
[ 2866.132281] [<ffffffff81083837>] ? process_scheduled_works+0x2f/0x2f
[ 2866.132281] [<ffffffff8108b7b8>] kthread+0x9a/0xa2
[ 2866.132281] [<ffffffff81a12184>] kernel_thread_helper+0x4/0x10
[ 2866.132281] [<ffffffff81a0a4b0>] ? retint_restore_args+0x13/0x13
[ 2866.132281] [<ffffffff8108b71e>] ? __init_kthread_worker+0x5a/0x5a
[ 2866.132281] [<ffffffff81a12180>] ? gs_change+0x13/0x13
[ 2866.308506] IPv4: Attempt to release TCP socket in state 1 ffff880019ec0000
[ 2866.309689] =============================================================================
[ 2866.310254] BUG TCP (Not tainted): Object already free
[ 2866.310254] -----------------------------------------------------------------------------
[ 2866.310254]
The bug comes from the fact that timer set in sk_reset_timer() can run
before we actually do the sock_hold(). socket refcount reaches zero and
we free the socket too soon.
timer handler is not allowed to reduce socket refcnt if socket is owned
by the user, or we need to change sk_reset_timer() implementation.
We should take a reference on the socket in case TCP_DELACK_TIMER_DEFERRED
or TCP_DELACK_TIMER_DEFERRED bit are set in tsq_flags
Also fix a typo in tcp_delack_timer(), where TCP_WRITE_TIMER_DEFERRED
was used instead of TCP_DELACK_TIMER_DEFERRED.
For consistency, use same socket refcount change for TCP_MTU_REDUCED_DEFERRED,
even if not fired from a timer.
Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Tested-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-08-20 08:22:46 +08:00
|
|
|
__sock_put(sk);
|
|
|
|
|
}
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
}
|
|
|
|
|
EXPORT_SYMBOL(tcp_release_cb);
|
|
|
|
|
|
|
|
|
|
void __init tcp_tasklet_init(void)
|
|
|
|
|
{
|
|
|
|
|
int i;
|
|
|
|
|
|
|
|
|
|
for_each_possible_cpu(i) {
|
|
|
|
|
struct tsq_tasklet *tsq = &per_cpu(tsq_tasklet, i);
|
|
|
|
|
|
|
|
|
|
INIT_LIST_HEAD(&tsq->head);
|
2020-11-03 17:18:17 +08:00
|
|
|
tasklet_setup(&tsq->tasklet, tcp_tasklet_func);
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* Write buffer destructor automatically called from kfree_skb.
|
2013-12-09 04:15:44 +08:00
|
|
|
* We can't xmit new skbs from this context, as we might already
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
* hold qdisc lock.
|
|
|
|
|
*/
|
2013-04-12 19:31:52 +08:00
|
|
|
void tcp_wfree(struct sk_buff *skb)
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
{
|
|
|
|
|
struct sock *sk = skb->sk;
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
2016-12-04 03:14:51 +08:00
|
|
|
unsigned long flags, nval, oval;
|
2014-10-13 21:27:47 +08:00
|
|
|
|
|
|
|
|
/* Keep one reference on sk_wmem_alloc.
|
|
|
|
|
* Will be released by sk_free() from here or tcp_tasklet_func()
|
|
|
|
|
*/
|
2017-06-30 18:08:00 +08:00
|
|
|
WARN_ON(refcount_sub_and_test(skb->truesize - 1, &sk->sk_wmem_alloc));
|
2014-10-13 21:27:47 +08:00
|
|
|
|
|
|
|
|
/* If this softirq is serviced by ksoftirqd, we are likely under stress.
|
|
|
|
|
* Wait until our queues (qdisc + devices) are drained.
|
|
|
|
|
* This gives :
|
|
|
|
|
* - less callbacks to tcp_write_xmit(), reducing stress (batches)
|
|
|
|
|
* - chance for incoming ACK (processed by another cpu maybe)
|
|
|
|
|
* to migrate this flow (skb->ooo_okay will be eventually set)
|
|
|
|
|
*/
|
2017-06-30 18:08:00 +08:00
|
|
|
if (refcount_read(&sk->sk_wmem_alloc) >= SKB_TRUESIZE(1) && this_cpu_ksoftirqd() == current)
|
2014-10-13 21:27:47 +08:00
|
|
|
goto out;
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
|
2016-12-04 03:14:57 +08:00
|
|
|
for (oval = READ_ONCE(sk->sk_tsq_flags);; oval = nval) {
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
struct tsq_tasklet *tsq;
|
2016-12-04 03:14:53 +08:00
|
|
|
bool empty;
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
|
2016-12-04 03:14:51 +08:00
|
|
|
if (!(oval & TSQF_THROTTLED) || (oval & TSQF_QUEUED))
|
|
|
|
|
goto out;
|
|
|
|
|
|
2018-05-11 05:59:43 +08:00
|
|
|
nval = (oval & ~TSQF_THROTTLED) | TSQF_QUEUED;
|
2016-12-04 03:14:57 +08:00
|
|
|
nval = cmpxchg(&sk->sk_tsq_flags, oval, nval);
|
2016-12-04 03:14:51 +08:00
|
|
|
if (nval != oval)
|
|
|
|
|
continue;
|
|
|
|
|
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
/* queue this socket to tasklet queue */
|
|
|
|
|
local_irq_save(flags);
|
2014-08-18 01:30:35 +08:00
|
|
|
tsq = this_cpu_ptr(&tsq_tasklet);
|
2016-12-04 03:14:53 +08:00
|
|
|
empty = list_empty(&tsq->head);
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
list_add(&tp->tsq_node, &tsq->head);
|
2016-12-04 03:14:53 +08:00
|
|
|
if (empty)
|
|
|
|
|
tasklet_schedule(&tsq->tasklet);
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
local_irq_restore(flags);
|
2014-10-13 21:27:47 +08:00
|
|
|
return;
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
}
|
2014-10-13 21:27:47 +08:00
|
|
|
out:
|
|
|
|
|
sk_free(sk);
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
}
|
|
|
|
|
|
2018-05-11 05:59:43 +08:00
|
|
|
/* Note: Called under soft irq.
|
|
|
|
|
* We can call TCP stack right away, unless socket is owned by user.
|
2017-05-16 19:24:36 +08:00
|
|
|
*/
|
|
|
|
|
enum hrtimer_restart tcp_pace_kick(struct hrtimer *timer)
|
|
|
|
|
{
|
|
|
|
|
struct tcp_sock *tp = container_of(timer, struct tcp_sock, pacing_timer);
|
|
|
|
|
struct sock *sk = (struct sock *)tp;
|
|
|
|
|
|
2018-05-11 05:59:43 +08:00
|
|
|
tcp_tsq_handler(sk);
|
|
|
|
|
sock_put(sk);
|
2017-05-16 19:24:36 +08:00
|
|
|
|
|
|
|
|
return HRTIMER_NORESTART;
|
|
|
|
|
}
|
|
|
|
|
|
2018-10-16 00:37:54 +08:00
|
|
|
static void tcp_update_skb_after_send(struct sock *sk, struct sk_buff *skb,
|
|
|
|
|
u64 prior_wstamp)
|
2017-10-05 03:59:58 +08:00
|
|
|
{
|
tcp: switch tcp and sch_fq to new earliest departure time model
TCP keeps track of tcp_wstamp_ns by itself, meaning sch_fq
no longer has to do it.
Thanks to this model, TCP can get more accurate RTT samples,
since pacing no longer inflates them.
This has the nice effect of removing some delays caused by FQ
quantum mechanism, causing inflated max/P99 latencies.
Also we might relax TCP Small Queue tight limits in the future,
since this new model allow TCP to build bigger batches, since
sch_fq (or a device with earliest departure time offload) ensure
these packets will be delivered on time.
Note that other protocols are not converted (they will probably
never be) so sch_fq has still support for SO_MAX_PACING_RATE
Tested:
Test showing FQ pacing quantum artifact for low-rate flows,
adding unexpected throttles for RPC flows, inflating max and P99 latencies.
The parameters chosen here are to show what happens typically when
a TCP flow has a reduced pacing rate (this can be caused by a reduced
cwin after few losses, or/and rtt above few ms)
MIBS="MIN_LATENCY,MEAN_LATENCY,MAX_LATENCY,P99_LATENCY,STDDEV_LATENCY"
Before :
$ netperf -H 10.246.7.133 -t TCP_RR -Cc -T6,6 -- -q 2000000 -r 100,100 -o $MIBS
MIGRATED TCP REQUEST/RESPONSE TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.246.7.133 () port 0 AF_INET : first burst 0 : cpu bind
Minimum Latency Microseconds,Mean Latency Microseconds,Maximum Latency Microseconds,99th Percentile Latency Microseconds,Stddev Latency Microseconds
19,82.78,5279,3825,482.02
After :
$ netperf -H 10.246.7.133 -t TCP_RR -Cc -T6,6 -- -q 2000000 -r 100,100 -o $MIBS
MIGRATED TCP REQUEST/RESPONSE TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.246.7.133 () port 0 AF_INET : first burst 0 : cpu bind
Minimum Latency Microseconds,Mean Latency Microseconds,Maximum Latency Microseconds,99th Percentile Latency Microseconds,Stddev Latency Microseconds
20,49.94,128,63,3.18
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-09-21 23:51:52 +08:00
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
|
|
|
|
|
if (sk->sk_pacing_status != SK_PACING_NONE) {
|
net: extend sk_pacing_rate to unsigned long
sk_pacing_rate has beed introduced as a u32 field in 2013,
effectively limiting per flow pacing to 34Gbit.
We believe it is time to allow TCP to pace high speed flows
on 64bit hosts, as we now can reach 100Gbit on one TCP flow.
This patch adds no cost for 32bit kernels.
The tcpi_pacing_rate and tcpi_max_pacing_rate were already
exported as 64bit, so iproute2/ss command require no changes.
Unfortunately the SO_MAX_PACING_RATE socket option will stay
32bit and we will need to add a new option to let applications
control high pacing rates.
State Recv-Q Send-Q Local Address:Port Peer Address:Port
ESTAB 0 1787144 10.246.9.76:49992 10.246.9.77:36741
timer:(on,003ms,0) ino:91863 sk:2 <->
skmem:(r0,rb540000,t66440,tb2363904,f605944,w1822984,o0,bl0,d0)
ts sack bbr wscale:8,8 rto:201 rtt:0.057/0.006 mss:1448
rcvmss:536 advmss:1448
cwnd:138 ssthresh:178 bytes_acked:256699822585 segs_out:177279177
segs_in:3916318 data_segs_out:177279175
bbr:(bw:31276.8Mbps,mrtt:0,pacing_gain:1.25,cwnd_gain:2)
send 28045.5Mbps lastrcv:73333
pacing_rate 38705.0Mbps delivery_rate 22997.6Mbps
busy:73333ms unacked:135 retrans:0/157 rcv_space:14480
notsent:2085120 minrtt:0.013
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-10-16 00:37:53 +08:00
|
|
|
unsigned long rate = sk->sk_pacing_rate;
|
tcp: switch tcp and sch_fq to new earliest departure time model
TCP keeps track of tcp_wstamp_ns by itself, meaning sch_fq
no longer has to do it.
Thanks to this model, TCP can get more accurate RTT samples,
since pacing no longer inflates them.
This has the nice effect of removing some delays caused by FQ
quantum mechanism, causing inflated max/P99 latencies.
Also we might relax TCP Small Queue tight limits in the future,
since this new model allow TCP to build bigger batches, since
sch_fq (or a device with earliest departure time offload) ensure
these packets will be delivered on time.
Note that other protocols are not converted (they will probably
never be) so sch_fq has still support for SO_MAX_PACING_RATE
Tested:
Test showing FQ pacing quantum artifact for low-rate flows,
adding unexpected throttles for RPC flows, inflating max and P99 latencies.
The parameters chosen here are to show what happens typically when
a TCP flow has a reduced pacing rate (this can be caused by a reduced
cwin after few losses, or/and rtt above few ms)
MIBS="MIN_LATENCY,MEAN_LATENCY,MAX_LATENCY,P99_LATENCY,STDDEV_LATENCY"
Before :
$ netperf -H 10.246.7.133 -t TCP_RR -Cc -T6,6 -- -q 2000000 -r 100,100 -o $MIBS
MIGRATED TCP REQUEST/RESPONSE TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.246.7.133 () port 0 AF_INET : first burst 0 : cpu bind
Minimum Latency Microseconds,Mean Latency Microseconds,Maximum Latency Microseconds,99th Percentile Latency Microseconds,Stddev Latency Microseconds
19,82.78,5279,3825,482.02
After :
$ netperf -H 10.246.7.133 -t TCP_RR -Cc -T6,6 -- -q 2000000 -r 100,100 -o $MIBS
MIGRATED TCP REQUEST/RESPONSE TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.246.7.133 () port 0 AF_INET : first burst 0 : cpu bind
Minimum Latency Microseconds,Mean Latency Microseconds,Maximum Latency Microseconds,99th Percentile Latency Microseconds,Stddev Latency Microseconds
20,49.94,128,63,3.18
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-09-21 23:51:52 +08:00
|
|
|
|
|
|
|
|
/* Original sch_fq does not pace first 10 MSS
|
|
|
|
|
* Note that tp->data_segs_out overflows after 2^32 packets,
|
|
|
|
|
* this is a minor annoyance.
|
|
|
|
|
*/
|
net: extend sk_pacing_rate to unsigned long
sk_pacing_rate has beed introduced as a u32 field in 2013,
effectively limiting per flow pacing to 34Gbit.
We believe it is time to allow TCP to pace high speed flows
on 64bit hosts, as we now can reach 100Gbit on one TCP flow.
This patch adds no cost for 32bit kernels.
The tcpi_pacing_rate and tcpi_max_pacing_rate were already
exported as 64bit, so iproute2/ss command require no changes.
Unfortunately the SO_MAX_PACING_RATE socket option will stay
32bit and we will need to add a new option to let applications
control high pacing rates.
State Recv-Q Send-Q Local Address:Port Peer Address:Port
ESTAB 0 1787144 10.246.9.76:49992 10.246.9.77:36741
timer:(on,003ms,0) ino:91863 sk:2 <->
skmem:(r0,rb540000,t66440,tb2363904,f605944,w1822984,o0,bl0,d0)
ts sack bbr wscale:8,8 rto:201 rtt:0.057/0.006 mss:1448
rcvmss:536 advmss:1448
cwnd:138 ssthresh:178 bytes_acked:256699822585 segs_out:177279177
segs_in:3916318 data_segs_out:177279175
bbr:(bw:31276.8Mbps,mrtt:0,pacing_gain:1.25,cwnd_gain:2)
send 28045.5Mbps lastrcv:73333
pacing_rate 38705.0Mbps delivery_rate 22997.6Mbps
busy:73333ms unacked:135 retrans:0/157 rcv_space:14480
notsent:2085120 minrtt:0.013
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-10-16 00:37:53 +08:00
|
|
|
if (rate != ~0UL && rate && tp->data_segs_out >= 10) {
|
2018-10-16 00:37:54 +08:00
|
|
|
u64 len_ns = div64_ul((u64)skb->len * NSEC_PER_SEC, rate);
|
|
|
|
|
u64 credit = tp->tcp_wstamp_ns - prior_wstamp;
|
|
|
|
|
|
|
|
|
|
/* take into account OS jitter */
|
|
|
|
|
len_ns -= min_t(u64, len_ns / 2, credit);
|
|
|
|
|
tp->tcp_wstamp_ns += len_ns;
|
tcp: switch tcp and sch_fq to new earliest departure time model
TCP keeps track of tcp_wstamp_ns by itself, meaning sch_fq
no longer has to do it.
Thanks to this model, TCP can get more accurate RTT samples,
since pacing no longer inflates them.
This has the nice effect of removing some delays caused by FQ
quantum mechanism, causing inflated max/P99 latencies.
Also we might relax TCP Small Queue tight limits in the future,
since this new model allow TCP to build bigger batches, since
sch_fq (or a device with earliest departure time offload) ensure
these packets will be delivered on time.
Note that other protocols are not converted (they will probably
never be) so sch_fq has still support for SO_MAX_PACING_RATE
Tested:
Test showing FQ pacing quantum artifact for low-rate flows,
adding unexpected throttles for RPC flows, inflating max and P99 latencies.
The parameters chosen here are to show what happens typically when
a TCP flow has a reduced pacing rate (this can be caused by a reduced
cwin after few losses, or/and rtt above few ms)
MIBS="MIN_LATENCY,MEAN_LATENCY,MAX_LATENCY,P99_LATENCY,STDDEV_LATENCY"
Before :
$ netperf -H 10.246.7.133 -t TCP_RR -Cc -T6,6 -- -q 2000000 -r 100,100 -o $MIBS
MIGRATED TCP REQUEST/RESPONSE TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.246.7.133 () port 0 AF_INET : first burst 0 : cpu bind
Minimum Latency Microseconds,Mean Latency Microseconds,Maximum Latency Microseconds,99th Percentile Latency Microseconds,Stddev Latency Microseconds
19,82.78,5279,3825,482.02
After :
$ netperf -H 10.246.7.133 -t TCP_RR -Cc -T6,6 -- -q 2000000 -r 100,100 -o $MIBS
MIGRATED TCP REQUEST/RESPONSE TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 10.246.7.133 () port 0 AF_INET : first burst 0 : cpu bind
Minimum Latency Microseconds,Mean Latency Microseconds,Maximum Latency Microseconds,99th Percentile Latency Microseconds,Stddev Latency Microseconds
20,49.94,128,63,3.18
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-09-21 23:51:52 +08:00
|
|
|
}
|
|
|
|
|
}
|
2017-10-05 03:59:58 +08:00
|
|
|
list_move_tail(&skb->tcp_tsorted_anchor, &tp->tsorted_sent_queue);
|
|
|
|
|
}
|
|
|
|
|
|
2020-06-20 03:12:34 +08:00
|
|
|
INDIRECT_CALLABLE_DECLARE(int ip_queue_xmit(struct sock *sk, struct sk_buff *skb, struct flowi *fl));
|
|
|
|
|
INDIRECT_CALLABLE_DECLARE(int inet6_csk_xmit(struct sock *sk, struct sk_buff *skb, struct flowi *fl));
|
2020-06-20 03:12:35 +08:00
|
|
|
INDIRECT_CALLABLE_DECLARE(void tcp_v4_send_check(struct sock *sk, struct sk_buff *skb));
|
2020-06-20 03:12:34 +08:00
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/* This routine actually transmits TCP packets queued in by
|
|
|
|
|
* tcp_do_sendmsg(). This is used by both the initial
|
|
|
|
|
* transmission and possible later retransmissions.
|
|
|
|
|
* All SKB's seen here are completely headerless. It is our
|
|
|
|
|
* job to build the TCP header, and pass the packet down to
|
|
|
|
|
* IP so it can do the same plus pass the packet off to the
|
|
|
|
|
* device.
|
|
|
|
|
*
|
|
|
|
|
* We are working here with either a clone of the original
|
|
|
|
|
* SKB, or a fresh unique copy made by the retransmit engine.
|
|
|
|
|
*/
|
2018-07-19 04:56:34 +08:00
|
|
|
static int __tcp_transmit_skb(struct sock *sk, struct sk_buff *skb,
|
|
|
|
|
int clone_it, gfp_t gfp_mask, u32 rcv_nxt)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
2005-12-07 08:24:52 +08:00
|
|
|
const struct inet_connection_sock *icsk = inet_csk(sk);
|
|
|
|
|
struct inet_sock *inet;
|
|
|
|
|
struct tcp_sock *tp;
|
|
|
|
|
struct tcp_skb_cb *tcb;
|
2008-07-19 15:04:31 +08:00
|
|
|
struct tcp_out_options opts;
|
2012-04-15 13:58:06 +08:00
|
|
|
unsigned int tcp_options_size, tcp_header_size;
|
tcp: update skb->skb_mstamp more carefully
liujian reported a problem in TCP_USER_TIMEOUT processing with a patch
in tcp_probe_timer() :
https://www.spinics.net/lists/netdev/msg454496.html
After investigations, the root cause of the problem is that we update
skb->skb_mstamp of skbs in write queue, even if the attempt to send a
clone or copy of it failed. One reason being a routing problem.
This patch prevents this, solving liujian issue.
It also removes a potential RTT miscalculation, since
__tcp_retransmit_skb() is not OR-ing TCP_SKB_CB(skb)->sacked with
TCPCB_EVER_RETRANS if a failure happens, but skb->skb_mstamp has
been changed.
A future ACK would then lead to a very small RTT sample and min_rtt
would then be lowered to this too small value.
Tested:
# cat user_timeout.pkt
--local_ip=192.168.102.64
0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
+0 `ifconfig tun0 192.168.102.64/16; ip ro add 192.0.2.1 dev tun0`
+0 < S 0:0(0) win 0 <mss 1460>
+0 > S. 0:0(0) ack 1 <mss 1460>
+.1 < . 1:1(0) ack 1 win 65530
+0 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_USER_TIMEOUT, [3000], 4) = 0
+0 write(4, ..., 24) = 24
+0 > P. 1:25(24) ack 1 win 29200
+.1 < . 1:1(0) ack 25 win 65530
//change the ipaddress
+1 `ifconfig tun0 192.168.0.10/16`
+1 write(4, ..., 24) = 24
+1 write(4, ..., 24) = 24
+1 write(4, ..., 24) = 24
+1 write(4, ..., 24) = 24
+0 `ifconfig tun0 192.168.102.64/16`
+0 < . 1:2(1) ack 25 win 65530
+0 `ifconfig tun0 192.168.0.10/16`
+3 write(4, ..., 24) = -1
# ./packetdrill user_timeout.pkt
Signed-off-by: Eric Dumazet <edumazet@googl.com>
Reported-by: liujian <liujian56@huawei.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-09-14 11:30:39 +08:00
|
|
|
struct sk_buff *oskb = NULL;
|
2006-11-15 11:07:45 +08:00
|
|
|
struct tcp_md5sig_key *md5;
|
2005-12-07 08:24:52 +08:00
|
|
|
struct tcphdr *th;
|
2018-10-16 00:37:54 +08:00
|
|
|
u64 prior_wstamp;
|
2005-12-07 08:24:52 +08:00
|
|
|
int err;
|
|
|
|
|
|
|
|
|
|
BUG_ON(!skb || !tcp_skb_pcount(skb));
|
2016-06-09 12:16:44 +08:00
|
|
|
tp = tcp_sk(sk);
|
2019-01-17 07:05:29 +08:00
|
|
|
prior_wstamp = tp->tcp_wstamp_ns;
|
|
|
|
|
tp->tcp_wstamp_ns = max(tp->tcp_wstamp_ns, tp->tcp_clock_cache);
|
|
|
|
|
skb->skb_mstamp_ns = tp->tcp_wstamp_ns;
|
2013-10-10 23:43:00 +08:00
|
|
|
if (clone_it) {
|
2016-06-09 12:16:44 +08:00
|
|
|
TCP_SKB_CB(skb)->tx.in_flight = TCP_SKB_CB(skb)->end_seq
|
|
|
|
|
- tp->snd_una;
|
tcp: update skb->skb_mstamp more carefully
liujian reported a problem in TCP_USER_TIMEOUT processing with a patch
in tcp_probe_timer() :
https://www.spinics.net/lists/netdev/msg454496.html
After investigations, the root cause of the problem is that we update
skb->skb_mstamp of skbs in write queue, even if the attempt to send a
clone or copy of it failed. One reason being a routing problem.
This patch prevents this, solving liujian issue.
It also removes a potential RTT miscalculation, since
__tcp_retransmit_skb() is not OR-ing TCP_SKB_CB(skb)->sacked with
TCPCB_EVER_RETRANS if a failure happens, but skb->skb_mstamp has
been changed.
A future ACK would then lead to a very small RTT sample and min_rtt
would then be lowered to this too small value.
Tested:
# cat user_timeout.pkt
--local_ip=192.168.102.64
0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
+0 `ifconfig tun0 192.168.102.64/16; ip ro add 192.0.2.1 dev tun0`
+0 < S 0:0(0) win 0 <mss 1460>
+0 > S. 0:0(0) ack 1 <mss 1460>
+.1 < . 1:1(0) ack 1 win 65530
+0 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_USER_TIMEOUT, [3000], 4) = 0
+0 write(4, ..., 24) = 24
+0 > P. 1:25(24) ack 1 win 29200
+.1 < . 1:1(0) ack 25 win 65530
//change the ipaddress
+1 `ifconfig tun0 192.168.0.10/16`
+1 write(4, ..., 24) = 24
+1 write(4, ..., 24) = 24
+1 write(4, ..., 24) = 24
+1 write(4, ..., 24) = 24
+0 `ifconfig tun0 192.168.102.64/16`
+0 < . 1:2(1) ack 25 win 65530
+0 `ifconfig tun0 192.168.0.10/16`
+3 write(4, ..., 24) = -1
# ./packetdrill user_timeout.pkt
Signed-off-by: Eric Dumazet <edumazet@googl.com>
Reported-by: liujian <liujian56@huawei.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-09-14 11:30:39 +08:00
|
|
|
oskb = skb;
|
2017-10-05 03:59:58 +08:00
|
|
|
|
|
|
|
|
tcp_skb_tsorted_save(oskb) {
|
|
|
|
|
if (unlikely(skb_cloned(oskb)))
|
|
|
|
|
skb = pskb_copy(oskb, gfp_mask);
|
|
|
|
|
else
|
|
|
|
|
skb = skb_clone(oskb, gfp_mask);
|
|
|
|
|
} tcp_skb_tsorted_restore(oskb);
|
|
|
|
|
|
2005-12-07 08:24:52 +08:00
|
|
|
if (unlikely(!skb))
|
|
|
|
|
return -ENOBUFS;
|
2020-03-20 03:49:55 +08:00
|
|
|
/* retransmit skbs might have a non zero value in skb->dev
|
|
|
|
|
* because skb->dev is aliased with skb->rbnode.rb_left
|
|
|
|
|
*/
|
|
|
|
|
skb->dev = NULL;
|
2005-12-07 08:24:52 +08:00
|
|
|
}
|
2018-10-16 00:37:52 +08:00
|
|
|
|
2005-12-07 08:24:52 +08:00
|
|
|
inet = inet_sk(sk);
|
|
|
|
|
tcb = TCP_SKB_CB(skb);
|
2008-07-19 15:04:31 +08:00
|
|
|
memset(&opts, 0, sizeof(opts));
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2019-09-11 05:49:28 +08:00
|
|
|
if (unlikely(tcb->tcp_flags & TCPHDR_SYN)) {
|
2008-07-19 15:04:31 +08:00
|
|
|
tcp_options_size = tcp_syn_options(sk, skb, &opts, &md5);
|
2019-09-11 05:49:28 +08:00
|
|
|
} else {
|
2008-07-19 15:04:31 +08:00
|
|
|
tcp_options_size = tcp_established_options(sk, skb, &opts,
|
|
|
|
|
&md5);
|
2019-09-11 05:49:28 +08:00
|
|
|
/* Force a PSH flag on all (GSO) packets to expedite GRO flush
|
|
|
|
|
* at receiver : This slightly improve GRO performance.
|
|
|
|
|
* Note that we do not force the PSH flag for non GSO packets,
|
|
|
|
|
* because they might be sent under high congestion events,
|
|
|
|
|
* and in this case it is better to delay the delivery of 1-MSS
|
|
|
|
|
* packets and thus the corresponding ACK packet that would
|
|
|
|
|
* release the following packet.
|
|
|
|
|
*/
|
|
|
|
|
if (tcp_skb_pcount(skb) > 1)
|
|
|
|
|
tcb->tcp_flags |= TCPHDR_PSH;
|
|
|
|
|
}
|
2008-07-19 15:04:31 +08:00
|
|
|
tcp_header_size = tcp_options_size + sizeof(struct tcphdr);
|
2007-02-09 22:24:47 +08:00
|
|
|
|
2013-05-23 15:44:20 +08:00
|
|
|
/* if no packet is in qdisc/device queue, then allow XPS to select
|
2014-10-11 09:06:35 +08:00
|
|
|
* another queue. We can be called from tcp_tsq_handler()
|
2018-05-11 05:59:43 +08:00
|
|
|
* which holds one reference to sk.
|
2014-10-11 09:06:35 +08:00
|
|
|
*
|
|
|
|
|
* TODO: Ideally, in-flight pure ACK packets should not matter here.
|
|
|
|
|
* One way to get this would be to set skb->truesize = 2 on them.
|
2013-05-23 15:44:20 +08:00
|
|
|
*/
|
2014-10-11 09:06:35 +08:00
|
|
|
skb->ooo_okay = sk_wmem_alloc_get(sk) < SKB_TRUESIZE(1);
|
2005-12-07 08:24:52 +08:00
|
|
|
|
2017-02-03 12:40:08 +08:00
|
|
|
/* If we had to use memory reserve to allocate this skb,
|
|
|
|
|
* this might cause drops if packet is looped back :
|
|
|
|
|
* Other socket might not have SOCK_MEMALLOC.
|
|
|
|
|
* Packets not looped back do not care about pfmemalloc.
|
|
|
|
|
*/
|
|
|
|
|
skb->pfmemalloc = 0;
|
|
|
|
|
|
2007-04-11 12:04:22 +08:00
|
|
|
skb_push(skb, tcp_header_size);
|
|
|
|
|
skb_reset_transport_header(skb);
|
tcp: TCP Small Queues
This introduce TSQ (TCP Small Queues)
TSQ goal is to reduce number of TCP packets in xmit queues (qdisc &
device queues), to reduce RTT and cwnd bias, part of the bufferbloat
problem.
sk->sk_wmem_alloc not allowed to grow above a given limit,
allowing no more than ~128KB [1] per tcp socket in qdisc/dev layers at a
given time.
TSO packets are sized/capped to half the limit, so that we have two
TSO packets in flight, allowing better bandwidth use.
As a side effect, setting the limit to 40000 automatically reduces the
standard gso max limit (65536) to 40000/2 : It can help to reduce
latencies of high prio packets, having smaller TSO packets.
This means we divert sock_wfree() to a tcp_wfree() handler, to
queue/send following frames when skb_orphan() [2] is called for the
already queued skbs.
Results on my dev machines (tg3/ixgbe nics) are really impressive,
using standard pfifo_fast, and with or without TSO/GSO.
Without reduction of nominal bandwidth, we have reduction of buffering
per bulk sender :
< 1ms on Gbit (instead of 50ms with TSO)
< 8ms on 100Mbit (instead of 132 ms)
I no longer have 4 MBytes backlogged in qdisc by a single netperf
session, and both side socket autotuning no longer use 4 Mbytes.
As skb destructor cannot restart xmit itself ( as qdisc lock might be
taken at this point ), we delegate the work to a tasklet. We use one
tasklest per cpu for performance reasons.
If tasklet finds a socket owned by the user, it sets TSQ_OWNED flag.
This flag is tested in a new protocol method called from release_sock(),
to eventually send new segments.
[1] New /proc/sys/net/ipv4/tcp_limit_output_bytes tunable
[2] skb_orphan() is usually called at TX completion time,
but some drivers call it in their start_xmit() handler.
These drivers should at least use BQL, or else a single TCP
session can still fill the whole NIC TX ring, since TSQ will
have no effect.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Dave Taht <dave.taht@bufferbloat.net>
Cc: Tom Herbert <therbert@google.com>
Cc: Matt Mathis <mattmathis@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-07-11 13:50:31 +08:00
|
|
|
|
|
|
|
|
skb_orphan(skb);
|
|
|
|
|
skb->sk = sk;
|
2016-05-03 01:56:27 +08:00
|
|
|
skb->destructor = skb_is_tcp_pure_ack(skb) ? __sock_wfree : tcp_wfree;
|
2014-07-02 12:32:17 +08:00
|
|
|
skb_set_hash_from_sk(skb, sk);
|
2017-06-30 18:08:00 +08:00
|
|
|
refcount_add(skb->truesize, &sk->sk_wmem_alloc);
|
2005-12-07 08:24:52 +08:00
|
|
|
|
2017-02-07 05:14:14 +08:00
|
|
|
skb_set_dst_pending_confirm(skb, sk->sk_dst_pending_confirm);
|
|
|
|
|
|
2005-12-07 08:24:52 +08:00
|
|
|
/* Build TCP header and checksum it. */
|
2016-05-14 00:16:40 +08:00
|
|
|
th = (struct tcphdr *)skb->data;
|
2009-10-15 14:30:45 +08:00
|
|
|
th->source = inet->inet_sport;
|
|
|
|
|
th->dest = inet->inet_dport;
|
2005-12-07 08:24:52 +08:00
|
|
|
th->seq = htonl(tcb->seq);
|
2018-07-19 04:56:34 +08:00
|
|
|
th->ack_seq = htonl(rcv_nxt);
|
2006-09-28 09:38:52 +08:00
|
|
|
*(((__be16 *)th) + 6) = htons(((tcp_header_size >> 2) << 12) |
|
2011-09-28 01:25:05 +08:00
|
|
|
tcb->tcp_flags);
|
2005-12-07 08:24:52 +08:00
|
|
|
|
|
|
|
|
th->check = 0;
|
|
|
|
|
th->urg_ptr = 0;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2008-10-08 05:43:06 +08:00
|
|
|
/* The urg_mode check is necessary during a below snd_una win probe */
|
2009-02-22 15:52:29 +08:00
|
|
|
if (unlikely(tcp_urg_mode(tp) && before(tcb->seq, tp->snd_up))) {
|
|
|
|
|
if (before(tp->snd_up, tcb->seq + 0x10000)) {
|
|
|
|
|
th->urg_ptr = htons(tp->snd_up - tcb->seq);
|
|
|
|
|
th->urg = 1;
|
|
|
|
|
} else if (after(tcb->seq + 0xFFFF, tp->snd_nxt)) {
|
2010-04-21 10:06:52 +08:00
|
|
|
th->urg_ptr = htons(0xFFFF);
|
2009-02-22 15:52:29 +08:00
|
|
|
th->urg = 1;
|
|
|
|
|
}
|
2005-12-07 08:24:52 +08:00
|
|
|
}
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2015-06-12 00:15:16 +08:00
|
|
|
skb_shinfo(skb)->gso_type = sk->sk_gso_type;
|
2016-05-14 00:16:40 +08:00
|
|
|
if (likely(!(tcb->tcp_flags & TCPHDR_SYN))) {
|
|
|
|
|
th->window = htons(tcp_select_window(sk));
|
|
|
|
|
tcp_ecn_send(sk, skb, th, tcp_header_size);
|
|
|
|
|
} else {
|
|
|
|
|
/* RFC1323: The window in SYN & SYN/ACK segments
|
|
|
|
|
* is never scaled.
|
|
|
|
|
*/
|
|
|
|
|
th->window = htons(min(tp->rcv_wnd, 65535U));
|
|
|
|
|
}
|
2020-11-20 03:46:02 +08:00
|
|
|
|
|
|
|
|
tcp_options_write((__be32 *)(th + 1), tp, &opts);
|
|
|
|
|
|
2006-11-15 11:07:45 +08:00
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
|
|
|
|
/* Calculate the MD5 hash, as we have all we need now */
|
|
|
|
|
if (md5) {
|
2010-05-16 15:36:33 +08:00
|
|
|
sk_nocaps_add(sk, NETIF_F_GSO_MASK);
|
2009-12-03 02:23:05 +08:00
|
|
|
tp->af_specific->calc_md5_hash(opts.hash_location,
|
2015-03-25 06:58:55 +08:00
|
|
|
md5, sk, skb);
|
2006-11-15 11:07:45 +08:00
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
|
2020-08-21 03:00:52 +08:00
|
|
|
/* BPF prog is the last one writing header option */
|
|
|
|
|
bpf_skops_write_hdr_opt(sk, skb, NULL, NULL, 0, &opts);
|
|
|
|
|
|
2020-06-20 03:12:35 +08:00
|
|
|
INDIRECT_CALL_INET(icsk->icsk_af_ops->send_check,
|
|
|
|
|
tcp_v6_send_check, tcp_v4_send_check,
|
|
|
|
|
sk, skb);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2011-09-28 01:25:05 +08:00
|
|
|
if (likely(tcb->tcp_flags & TCPHDR_ACK))
|
2018-07-19 04:56:35 +08:00
|
|
|
tcp_event_ack_sent(sk, tcp_skb_pcount(skb), rcv_nxt);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2016-03-15 01:52:15 +08:00
|
|
|
if (skb->len != tcp_header_size) {
|
2011-10-21 17:22:42 +08:00
|
|
|
tcp_event_data_sent(tp, sk);
|
2016-03-15 01:52:15 +08:00
|
|
|
tp->data_segs_out += tcp_skb_pcount(skb);
|
2018-08-01 08:46:21 +08:00
|
|
|
tp->bytes_sent += skb->len - tcp_header_size;
|
2016-03-15 01:52:15 +08:00
|
|
|
}
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2006-08-08 12:04:15 +08:00
|
|
|
if (after(tcb->end_seq, tp->snd_nxt) || tcb->seq == tcb->end_seq)
|
2010-04-22 15:00:24 +08:00
|
|
|
TCP_ADD_STATS(sock_net(sk), TCP_MIB_OUTSEGS,
|
|
|
|
|
tcp_skb_pcount(skb));
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2015-05-21 07:35:41 +08:00
|
|
|
tp->segs_out += tcp_skb_pcount(skb);
|
2015-06-12 00:15:18 +08:00
|
|
|
/* OK, its time to fill skb_shinfo(skb)->gso_{segs|size} */
|
2014-09-24 19:11:22 +08:00
|
|
|
skb_shinfo(skb)->gso_segs = tcp_skb_pcount(skb);
|
2015-06-12 00:15:18 +08:00
|
|
|
skb_shinfo(skb)->gso_size = tcp_skb_mss(skb);
|
2014-09-24 19:11:22 +08:00
|
|
|
|
2018-09-21 23:51:50 +08:00
|
|
|
/* Leave earliest departure time in skb->tstamp (skb->skb_mstamp_ns) */
|
2014-09-28 00:50:57 +08:00
|
|
|
|
|
|
|
|
/* Cleanup our debris for IP stacks */
|
|
|
|
|
memset(skb->cb, 0, max(sizeof(struct inet_skb_parm),
|
|
|
|
|
sizeof(struct inet6_skb_parm)));
|
|
|
|
|
|
tcp: add optional per socket transmit delay
Adding delays to TCP flows is crucial for studying behavior
of TCP stacks, including congestion control modules.
Linux offers netem module, but it has unpractical constraints :
- Need root access to change qdisc
- Hard to setup on egress if combined with non trivial qdisc like FQ
- Single delay for all flows.
EDT (Earliest Departure Time) adoption in TCP stack allows us
to enable a per socket delay at a very small cost.
Networking tools can now establish thousands of flows, each of them
with a different delay, simulating real world conditions.
This requires FQ packet scheduler or a EDT-enabled NIC.
This patchs adds TCP_TX_DELAY socket option, to set a delay in
usec units.
unsigned int tx_delay = 10000; /* 10 msec */
setsockopt(fd, SOL_TCP, TCP_TX_DELAY, &tx_delay, sizeof(tx_delay));
Note that FQ packet scheduler limits might need some tweaking :
man tc-fq
PARAMETERS
limit
Hard limit on the real queue size. When this limit is
reached, new packets are dropped. If the value is lowered,
packets are dropped so that the new limit is met. Default
is 10000 packets.
flow_limit
Hard limit on the maximum number of packets queued per
flow. Default value is 100.
Use of TCP_TX_DELAY option will increase number of skbs in FQ qdisc,
so packets would be dropped if any of the previous limit is hit.
Use of a jump label makes this support runtime-free, for hosts
never using the option.
Also note that TSQ (TCP Small Queues) limits are slightly changed
with this patch : we need to account that skbs artificially delayed
wont stop us providind more skbs to feed the pipe (netem uses
skb_orphan_partial() for this purpose, but FQ can not use this trick)
Because of that, using big delays might very well trigger
old bugs in TSO auto defer logic and/or sndbuf limited detection.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2019-06-13 02:57:25 +08:00
|
|
|
tcp_add_tx_delay(skb, tp);
|
|
|
|
|
|
2020-06-20 03:12:34 +08:00
|
|
|
err = INDIRECT_CALL_INET(icsk->icsk_af_ops->queue_xmit,
|
|
|
|
|
inet6_csk_xmit, ip_queue_xmit,
|
|
|
|
|
sk, skb, &inet->cork.fl);
|
2014-09-06 06:33:33 +08:00
|
|
|
|
tcp: update skb->skb_mstamp more carefully
liujian reported a problem in TCP_USER_TIMEOUT processing with a patch
in tcp_probe_timer() :
https://www.spinics.net/lists/netdev/msg454496.html
After investigations, the root cause of the problem is that we update
skb->skb_mstamp of skbs in write queue, even if the attempt to send a
clone or copy of it failed. One reason being a routing problem.
This patch prevents this, solving liujian issue.
It also removes a potential RTT miscalculation, since
__tcp_retransmit_skb() is not OR-ing TCP_SKB_CB(skb)->sacked with
TCPCB_EVER_RETRANS if a failure happens, but skb->skb_mstamp has
been changed.
A future ACK would then lead to a very small RTT sample and min_rtt
would then be lowered to this too small value.
Tested:
# cat user_timeout.pkt
--local_ip=192.168.102.64
0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
+0 `ifconfig tun0 192.168.102.64/16; ip ro add 192.0.2.1 dev tun0`
+0 < S 0:0(0) win 0 <mss 1460>
+0 > S. 0:0(0) ack 1 <mss 1460>
+.1 < . 1:1(0) ack 1 win 65530
+0 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_USER_TIMEOUT, [3000], 4) = 0
+0 write(4, ..., 24) = 24
+0 > P. 1:25(24) ack 1 win 29200
+.1 < . 1:1(0) ack 25 win 65530
//change the ipaddress
+1 `ifconfig tun0 192.168.0.10/16`
+1 write(4, ..., 24) = 24
+1 write(4, ..., 24) = 24
+1 write(4, ..., 24) = 24
+1 write(4, ..., 24) = 24
+0 `ifconfig tun0 192.168.102.64/16`
+0 < . 1:2(1) ack 25 win 65530
+0 `ifconfig tun0 192.168.0.10/16`
+3 write(4, ..., 24) = -1
# ./packetdrill user_timeout.pkt
Signed-off-by: Eric Dumazet <edumazet@googl.com>
Reported-by: liujian <liujian56@huawei.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-09-14 11:30:39 +08:00
|
|
|
if (unlikely(err > 0)) {
|
|
|
|
|
tcp_enter_cwr(sk);
|
|
|
|
|
err = net_xmit_eval(err);
|
|
|
|
|
}
|
2017-09-16 07:47:42 +08:00
|
|
|
if (!err && oskb) {
|
2018-10-16 00:37:54 +08:00
|
|
|
tcp_update_skb_after_send(sk, oskb, prior_wstamp);
|
2017-09-16 07:47:42 +08:00
|
|
|
tcp_rate_skb_sent(sk, oskb);
|
|
|
|
|
}
|
tcp: update skb->skb_mstamp more carefully
liujian reported a problem in TCP_USER_TIMEOUT processing with a patch
in tcp_probe_timer() :
https://www.spinics.net/lists/netdev/msg454496.html
After investigations, the root cause of the problem is that we update
skb->skb_mstamp of skbs in write queue, even if the attempt to send a
clone or copy of it failed. One reason being a routing problem.
This patch prevents this, solving liujian issue.
It also removes a potential RTT miscalculation, since
__tcp_retransmit_skb() is not OR-ing TCP_SKB_CB(skb)->sacked with
TCPCB_EVER_RETRANS if a failure happens, but skb->skb_mstamp has
been changed.
A future ACK would then lead to a very small RTT sample and min_rtt
would then be lowered to this too small value.
Tested:
# cat user_timeout.pkt
--local_ip=192.168.102.64
0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
+0 `ifconfig tun0 192.168.102.64/16; ip ro add 192.0.2.1 dev tun0`
+0 < S 0:0(0) win 0 <mss 1460>
+0 > S. 0:0(0) ack 1 <mss 1460>
+.1 < . 1:1(0) ack 1 win 65530
+0 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_USER_TIMEOUT, [3000], 4) = 0
+0 write(4, ..., 24) = 24
+0 > P. 1:25(24) ack 1 win 29200
+.1 < . 1:1(0) ack 25 win 65530
//change the ipaddress
+1 `ifconfig tun0 192.168.0.10/16`
+1 write(4, ..., 24) = 24
+1 write(4, ..., 24) = 24
+1 write(4, ..., 24) = 24
+1 write(4, ..., 24) = 24
+0 `ifconfig tun0 192.168.102.64/16`
+0 < . 1:2(1) ack 25 win 65530
+0 `ifconfig tun0 192.168.0.10/16`
+3 write(4, ..., 24) = -1
# ./packetdrill user_timeout.pkt
Signed-off-by: Eric Dumazet <edumazet@googl.com>
Reported-by: liujian <liujian56@huawei.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-09-14 11:30:39 +08:00
|
|
|
return err;
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
|
2018-07-19 04:56:34 +08:00
|
|
|
static int tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, int clone_it,
|
|
|
|
|
gfp_t gfp_mask)
|
|
|
|
|
{
|
|
|
|
|
return __tcp_transmit_skb(sk, skb, clone_it, gfp_mask,
|
|
|
|
|
tcp_sk(sk)->rcv_nxt);
|
|
|
|
|
}
|
|
|
|
|
|
2009-07-22 07:00:40 +08:00
|
|
|
/* This routine just queues the buffer for sending.
|
2005-04-17 06:20:36 +08:00
|
|
|
*
|
|
|
|
|
* NOTE: probe0 timer is not checked, do not forget tcp_push_pending_frames,
|
|
|
|
|
* otherwise socket can stall.
|
|
|
|
|
*/
|
|
|
|
|
static void tcp_queue_skb(struct sock *sk, struct sk_buff *skb)
|
|
|
|
|
{
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
|
|
|
|
|
/* Advance write_seq and place onto the write_queue. */
|
2019-10-11 11:17:41 +08:00
|
|
|
WRITE_ONCE(tp->write_seq, TCP_SKB_CB(skb)->end_seq);
|
2014-09-23 07:29:32 +08:00
|
|
|
__skb_header_release(skb);
|
2007-03-08 04:12:44 +08:00
|
|
|
tcp_add_write_queue_tail(sk, skb);
|
2019-10-11 11:17:46 +08:00
|
|
|
sk_wmem_queued_add(sk, skb->truesize);
|
2007-12-31 16:11:19 +08:00
|
|
|
sk_mem_charge(sk, skb->truesize);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
|
2009-07-22 07:00:40 +08:00
|
|
|
/* Initialize TSO segments for a packet. */
|
2015-06-12 00:15:17 +08:00
|
|
|
static void tcp_set_skb_tso_segs(struct sk_buff *skb, unsigned int mss_now)
|
2005-07-06 06:18:03 +08:00
|
|
|
{
|
2018-02-20 03:56:51 +08:00
|
|
|
if (skb->len <= mss_now) {
|
2005-07-06 06:18:03 +08:00
|
|
|
/* Avoid the costly divide in the normal
|
|
|
|
|
* non-TSO case.
|
|
|
|
|
*/
|
2014-09-24 19:11:22 +08:00
|
|
|
tcp_skb_pcount_set(skb, 1);
|
2015-06-12 00:15:18 +08:00
|
|
|
TCP_SKB_CB(skb)->tcp_gso_size = 0;
|
2005-07-06 06:18:03 +08:00
|
|
|
} else {
|
2014-09-24 19:11:22 +08:00
|
|
|
tcp_skb_pcount_set(skb, DIV_ROUND_UP(skb->len, mss_now));
|
2015-06-12 00:15:18 +08:00
|
|
|
TCP_SKB_CB(skb)->tcp_gso_size = mss_now;
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2009-04-02 07:15:17 +08:00
|
|
|
/* Pcount in the middle of the write queue got changed, we need to do various
|
|
|
|
|
* tweaks to fix counters
|
|
|
|
|
*/
|
2011-10-21 17:22:42 +08:00
|
|
|
static void tcp_adjust_pcount(struct sock *sk, const struct sk_buff *skb, int decr)
|
2009-04-02 07:15:17 +08:00
|
|
|
{
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
|
|
|
|
|
tp->packets_out -= decr;
|
|
|
|
|
|
|
|
|
|
if (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_ACKED)
|
|
|
|
|
tp->sacked_out -= decr;
|
|
|
|
|
if (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_RETRANS)
|
|
|
|
|
tp->retrans_out -= decr;
|
|
|
|
|
if (TCP_SKB_CB(skb)->sacked & TCPCB_LOST)
|
|
|
|
|
tp->lost_out -= decr;
|
|
|
|
|
|
|
|
|
|
/* Reno case is special. Sigh... */
|
|
|
|
|
if (tcp_is_reno(tp) && decr > 0)
|
|
|
|
|
tp->sacked_out -= min_t(u32, tp->sacked_out, decr);
|
|
|
|
|
|
|
|
|
|
if (tp->lost_skb_hint &&
|
|
|
|
|
before(TCP_SKB_CB(skb)->seq, TCP_SKB_CB(tp->lost_skb_hint)->seq) &&
|
2017-11-09 05:01:26 +08:00
|
|
|
(TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_ACKED))
|
2009-04-02 07:15:17 +08:00
|
|
|
tp->lost_cnt_hint -= decr;
|
|
|
|
|
|
|
|
|
|
tcp_verify_left_out(tp);
|
|
|
|
|
}
|
|
|
|
|
|
2016-04-28 11:39:01 +08:00
|
|
|
static bool tcp_has_tx_tstamp(const struct sk_buff *skb)
|
|
|
|
|
{
|
|
|
|
|
return TCP_SKB_CB(skb)->txstamp_ack ||
|
|
|
|
|
(skb_shinfo(skb)->tx_flags & SKBTX_ANY_TSTAMP);
|
|
|
|
|
}
|
|
|
|
|
|
2014-08-13 03:08:12 +08:00
|
|
|
static void tcp_fragment_tstamp(struct sk_buff *skb, struct sk_buff *skb2)
|
|
|
|
|
{
|
|
|
|
|
struct skb_shared_info *shinfo = skb_shinfo(skb);
|
|
|
|
|
|
2016-04-28 11:39:01 +08:00
|
|
|
if (unlikely(tcp_has_tx_tstamp(skb)) &&
|
2014-08-13 03:08:12 +08:00
|
|
|
!before(shinfo->tskey, TCP_SKB_CB(skb2)->seq)) {
|
|
|
|
|
struct skb_shared_info *shinfo2 = skb_shinfo(skb2);
|
|
|
|
|
u8 tsflags = shinfo->tx_flags & SKBTX_ANY_TSTAMP;
|
|
|
|
|
|
|
|
|
|
shinfo->tx_flags &= ~tsflags;
|
|
|
|
|
shinfo2->tx_flags |= tsflags;
|
|
|
|
|
swap(shinfo->tskey, shinfo2->tskey);
|
tcp: Carry txstamp_ack in tcp_fragment_tstamp
When a tcp skb is sliced into two smaller skbs (e.g. in
tcp_fragment() and tso_fragment()), it does not carry
the txstamp_ack bit to the newly created skb if it is needed.
The end result is a timestamping event (SCM_TSTAMP_ACK) will
be missing from the sk->sk_error_queue.
This patch carries this bit to the new skb2
in tcp_fragment_tstamp().
BPF Output Before:
~~~~~~
<No output due to missing SCM_TSTAMP_ACK timestamp>
BPF Output After:
~~~~~~
<...>-2050 [000] d.s. 100.928763: : ee_data:14599
Packetdrill Script:
~~~~~~
+0 `sysctl -q -w net.ipv4.tcp_min_tso_segs=10`
+0 `sysctl -q -w net.ipv4.tcp_no_metrics_save=1`
+0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
0.100 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 7>
0.100 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 7>
0.200 < . 1:1(0) ack 1 win 257
0.200 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_NODELAY, [1], 4) = 0
+0 setsockopt(4, SOL_SOCKET, 37, [2688], 4) = 0
0.200 write(4, ..., 14600) = 14600
+0 setsockopt(4, SOL_SOCKET, 37, [2176], 4) = 0
0.200 > . 1:7301(7300) ack 1
0.200 > P. 7301:14601(7300) ack 1
0.300 < . 1:1(0) ack 14601 win 257
0.300 close(4) = 0
0.300 > F. 14601:14601(0) ack 1
0.400 < F. 1:1(0) ack 16062 win 257
0.400 > . 14602:14602(0) ack 2
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Soheil Hassas Yeganeh <soheil@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Tested-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-20 13:50:47 +08:00
|
|
|
TCP_SKB_CB(skb2)->txstamp_ack = TCP_SKB_CB(skb)->txstamp_ack;
|
|
|
|
|
TCP_SKB_CB(skb)->txstamp_ack = 0;
|
2014-08-13 03:08:12 +08:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
tcp: Handle eor bit when fragmenting a skb
When fragmenting a skb, the next_skb should carry
the eor from prev_skb. The eor of prev_skb should
also be reset.
Packetdrill script for testing:
~~~~~~
+0 `sysctl -q -w net.ipv4.tcp_min_tso_segs=10`
+0 `sysctl -q -w net.ipv4.tcp_no_metrics_save=1`
+0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
0.100 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 7>
0.100 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 7>
0.200 < . 1:1(0) ack 1 win 257
0.200 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_NODELAY, [1], 4) = 0
0.200 sendto(4, ..., 15330, MSG_EOR, ..., ...) = 15330
0.200 sendto(4, ..., 730, 0, ..., ...) = 730
0.200 > . 1:7301(7300) ack 1
0.200 > . 7301:14601(7300) ack 1
0.300 < . 1:1(0) ack 14601 win 257
0.300 > P. 14601:15331(730) ack 1
0.300 > P. 15331:16061(730) ack 1
0.400 < . 1:1(0) ack 16061 win 257
0.400 close(4) = 0
0.400 > F. 16061:16061(0) ack 1
0.400 < F. 1:1(0) ack 16062 win 257
0.400 > . 16062:16062(0) ack 2
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Soheil Hassas Yeganeh <soheil@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-26 05:44:50 +08:00
|
|
|
static void tcp_skb_fragment_eor(struct sk_buff *skb, struct sk_buff *skb2)
|
|
|
|
|
{
|
|
|
|
|
TCP_SKB_CB(skb2)->eor = TCP_SKB_CB(skb)->eor;
|
|
|
|
|
TCP_SKB_CB(skb)->eor = 0;
|
|
|
|
|
}
|
|
|
|
|
|
2017-10-06 13:21:27 +08:00
|
|
|
/* Insert buff after skb on the write or rtx queue of sk. */
|
|
|
|
|
static void tcp_insert_write_queue_after(struct sk_buff *skb,
|
|
|
|
|
struct sk_buff *buff,
|
|
|
|
|
struct sock *sk,
|
|
|
|
|
enum tcp_queue tcp_queue)
|
|
|
|
|
{
|
|
|
|
|
if (tcp_queue == TCP_FRAG_IN_WRITE_QUEUE)
|
|
|
|
|
__skb_queue_after(&sk->sk_write_queue, skb, buff);
|
|
|
|
|
else
|
|
|
|
|
tcp_rbtree_insert(&sk->tcp_rtx_queue, buff);
|
|
|
|
|
}
|
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/* Function to create two new TCP segments. Shrinks the given segment
|
|
|
|
|
* to the specified size and appends a new segment with the rest of the
|
2007-02-09 22:24:47 +08:00
|
|
|
* packet to the list. This won't be called frequently, I hope.
|
2005-04-17 06:20:36 +08:00
|
|
|
* Remember, these are still headerless SKBs at this point.
|
|
|
|
|
*/
|
2017-10-06 13:21:27 +08:00
|
|
|
int tcp_fragment(struct sock *sk, enum tcp_queue tcp_queue,
|
|
|
|
|
struct sk_buff *skb, u32 len,
|
2014-06-06 22:32:37 +08:00
|
|
|
unsigned int mss_now, gfp_t gfp)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
struct sk_buff *buff;
|
2005-09-02 13:47:01 +08:00
|
|
|
int nsize, old_factor;
|
2019-07-20 02:52:33 +08:00
|
|
|
long limit;
|
2006-04-20 12:35:00 +08:00
|
|
|
int nlen;
|
2009-02-28 12:44:42 +08:00
|
|
|
u8 flags;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2011-04-02 12:47:41 +08:00
|
|
|
if (WARN_ON(len > skb->len))
|
|
|
|
|
return -EINVAL;
|
2005-11-11 09:14:59 +08:00
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
nsize = skb_headlen(skb) - len;
|
|
|
|
|
if (nsize < 0)
|
|
|
|
|
nsize = 0;
|
|
|
|
|
|
2019-07-20 02:52:33 +08:00
|
|
|
/* tcp_sendmsg() can overshoot sk_wmem_queued by one full size skb.
|
|
|
|
|
* We need some allowance to not penalize applications setting small
|
|
|
|
|
* SO_SNDBUF values.
|
|
|
|
|
* Also allow first and last skb in retransmit queue to be split.
|
|
|
|
|
*/
|
|
|
|
|
limit = sk->sk_sndbuf + 2 * SKB_TRUESIZE(GSO_MAX_SIZE);
|
|
|
|
|
if (unlikely((sk->sk_wmem_queued >> 1) > limit &&
|
|
|
|
|
tcp_queue != TCP_FRAG_IN_WRITE_QUEUE &&
|
|
|
|
|
skb != tcp_rtx_queue_head(sk) &&
|
|
|
|
|
skb != tcp_rtx_queue_tail(sk))) {
|
2019-05-18 20:12:05 +08:00
|
|
|
NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPWQUEUETOOBIG);
|
|
|
|
|
return -ENOMEM;
|
|
|
|
|
}
|
|
|
|
|
|
2014-06-06 22:32:37 +08:00
|
|
|
if (skb_unclone(skb, gfp))
|
2005-04-17 06:20:36 +08:00
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
|
|
/* Get a new skb... force flag on. */
|
2015-05-20 04:26:55 +08:00
|
|
|
buff = sk_stream_alloc_skb(sk, nsize, gfp, true);
|
2015-04-03 16:17:26 +08:00
|
|
|
if (!buff)
|
2005-04-17 06:20:36 +08:00
|
|
|
return -ENOMEM; /* We'll just try again later. */
|
2019-08-08 08:03:59 +08:00
|
|
|
skb_copy_decrypted(buff, skb);
|
2020-11-04 03:05:05 +08:00
|
|
|
mptcp_skb_ext_copy(buff, skb);
|
2006-04-19 04:24:14 +08:00
|
|
|
|
2019-10-11 11:17:46 +08:00
|
|
|
sk_wmem_queued_add(sk, buff->truesize);
|
2007-12-31 16:11:19 +08:00
|
|
|
sk_mem_charge(sk, buff->truesize);
|
2006-04-20 12:35:00 +08:00
|
|
|
nlen = skb->len - len - nsize;
|
|
|
|
|
buff->truesize += nlen;
|
|
|
|
|
skb->truesize -= nlen;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
|
/* Correct the sequence numbers. */
|
|
|
|
|
TCP_SKB_CB(buff)->seq = TCP_SKB_CB(skb)->seq + len;
|
|
|
|
|
TCP_SKB_CB(buff)->end_seq = TCP_SKB_CB(skb)->end_seq;
|
|
|
|
|
TCP_SKB_CB(skb)->end_seq = TCP_SKB_CB(buff)->seq;
|
|
|
|
|
|
|
|
|
|
/* PSH and FIN should only be set in the second packet. */
|
2011-09-28 01:25:05 +08:00
|
|
|
flags = TCP_SKB_CB(skb)->tcp_flags;
|
|
|
|
|
TCP_SKB_CB(skb)->tcp_flags = flags & ~(TCPHDR_FIN | TCPHDR_PSH);
|
|
|
|
|
TCP_SKB_CB(buff)->tcp_flags = flags;
|
2005-09-20 09:18:38 +08:00
|
|
|
TCP_SKB_CB(buff)->sacked = TCP_SKB_CB(skb)->sacked;
|
tcp: Handle eor bit when fragmenting a skb
When fragmenting a skb, the next_skb should carry
the eor from prev_skb. The eor of prev_skb should
also be reset.
Packetdrill script for testing:
~~~~~~
+0 `sysctl -q -w net.ipv4.tcp_min_tso_segs=10`
+0 `sysctl -q -w net.ipv4.tcp_no_metrics_save=1`
+0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
0.100 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 7>
0.100 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 7>
0.200 < . 1:1(0) ack 1 win 257
0.200 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_NODELAY, [1], 4) = 0
0.200 sendto(4, ..., 15330, MSG_EOR, ..., ...) = 15330
0.200 sendto(4, ..., 730, 0, ..., ...) = 730
0.200 > . 1:7301(7300) ack 1
0.200 > . 7301:14601(7300) ack 1
0.300 < . 1:1(0) ack 14601 win 257
0.300 > P. 14601:15331(730) ack 1
0.300 > P. 15331:16061(730) ack 1
0.400 < . 1:1(0) ack 16061 win 257
0.400 close(4) = 0
0.400 > F. 16061:16061(0) ack 1
0.400 < F. 1:1(0) ack 16062 win 257
0.400 > . 16062:16062(0) ack 2
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Soheil Hassas Yeganeh <soheil@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-26 05:44:50 +08:00
|
|
|
tcp_skb_fragment_eor(skb, buff);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2018-02-20 03:56:52 +08:00
|
|
|
skb_split(skb, buff, len);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2018-02-20 03:56:52 +08:00
|
|
|
buff->ip_summed = CHECKSUM_PARTIAL;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2005-08-15 08:24:31 +08:00
|
|
|
buff->tstamp = skb->tstamp;
|
2014-08-13 03:08:12 +08:00
|
|
|
tcp_fragment_tstamp(skb, buff);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2005-09-02 13:47:01 +08:00
|
|
|
old_factor = tcp_skb_pcount(skb);
|
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/* Fix up tso_factor for both original and new SKB. */
|
2015-06-12 00:15:17 +08:00
|
|
|
tcp_set_skb_tso_segs(skb, mss_now);
|
|
|
|
|
tcp_set_skb_tso_segs(buff, mss_now);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
tcp: track data delivery rate for a TCP connection
This patch generates data delivery rate (throughput) samples on a
per-ACK basis. These rate samples can be used by congestion control
modules, and specifically will be used by TCP BBR in later patches in
this series.
Key state:
tp->delivered: Tracks the total number of data packets (original or not)
delivered so far. This is an already-existing field.
tp->delivered_mstamp: the last time tp->delivered was updated.
Algorithm:
A rate sample is calculated as (d1 - d0)/(t1 - t0) on a per-ACK basis:
d1: the current tp->delivered after processing the ACK
t1: the current time after processing the ACK
d0: the prior tp->delivered when the acked skb was transmitted
t0: the prior tp->delivered_mstamp when the acked skb was transmitted
When an skb is transmitted, we snapshot d0 and t0 in its control
block in tcp_rate_skb_sent().
When an ACK arrives, it may SACK and ACK some skbs. For each SACKed
or ACKed skb, tcp_rate_skb_delivered() updates the rate_sample struct
to reflect the latest (d0, t0).
Finally, tcp_rate_gen() generates a rate sample by storing
(d1 - d0) in rs->delivered and (t1 - t0) in rs->interval_us.
One caveat: if an skb was sent with no packets in flight, then
tp->delivered_mstamp may be either invalid (if the connection is
starting) or outdated (if the connection was idle). In that case,
we'll re-stamp tp->delivered_mstamp.
At first glance it seems t0 should always be the time when an skb was
transmitted, but actually this could over-estimate the rate due to
phase mismatch between transmit and ACK events. To track the delivery
rate, we ensure that if packets are in flight then t0 and and t1 are
times at which packets were marked delivered.
If the initial and final RTTs are different then one may be corrupted
by some sort of noise. The noise we see most often is sending gaps
caused by delayed, compressed, or stretched acks. This either affects
both RTTs equally or artificially reduces the final RTT. We approach
this by recording the info we need to compute the initial RTT
(duration of the "send phase" of the window) when we recorded the
associated inflight. Then, for a filter to avoid bandwidth
overestimates, we generalize the per-sample bandwidth computation
from:
bw = delivered / ack_phase_rtt
to the following:
bw = delivered / max(send_phase_rtt, ack_phase_rtt)
In large-scale experiments, this filtering approach incorporating
send_phase_rtt is effective at avoiding bandwidth overestimates due to
ACK compression or stretched ACKs.
Signed-off-by: Van Jacobson <vanj@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-09-20 11:39:14 +08:00
|
|
|
/* Update delivered info for the new segment */
|
|
|
|
|
TCP_SKB_CB(buff)->tx = TCP_SKB_CB(skb)->tx;
|
|
|
|
|
|
2005-09-02 13:47:01 +08:00
|
|
|
/* If this packet has been sent out already, we must
|
|
|
|
|
* adjust the various packet counters.
|
|
|
|
|
*/
|
2005-09-09 06:10:52 +08:00
|
|
|
if (!before(tp->snd_nxt, TCP_SKB_CB(buff)->end_seq)) {
|
2005-09-02 13:47:01 +08:00
|
|
|
int diff = old_factor - tcp_skb_pcount(skb) -
|
|
|
|
|
tcp_skb_pcount(buff);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2009-04-02 07:15:17 +08:00
|
|
|
if (diff)
|
|
|
|
|
tcp_adjust_pcount(sk, skb, diff);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Link BUFF into the send queue. */
|
2014-09-23 07:29:32 +08:00
|
|
|
__skb_header_release(buff);
|
2017-10-06 13:21:27 +08:00
|
|
|
tcp_insert_write_queue_after(skb, buff, sk, tcp_queue);
|
2017-11-03 09:10:03 +08:00
|
|
|
if (tcp_queue == TCP_FRAG_IN_RTX_QUEUE)
|
|
|
|
|
list_add(&buff->tcp_tsorted_anchor, &skb->tcp_tsorted_anchor);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
2017-06-02 05:18:36 +08:00
|
|
|
/* This is similar to __pskb_pull_tail(). The difference is that pulled
|
|
|
|
|
* data is not copied, but immediately discarded.
|
2005-04-17 06:20:36 +08:00
|
|
|
*/
|
2017-04-27 08:15:40 +08:00
|
|
|
static int __pskb_trim_head(struct sk_buff *skb, int len)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
2013-12-06 14:31:30 +08:00
|
|
|
struct skb_shared_info *shinfo;
|
2005-04-17 06:20:36 +08:00
|
|
|
int i, k, eat;
|
|
|
|
|
|
2011-12-04 16:51:08 +08:00
|
|
|
eat = min_t(int, len, skb_headlen(skb));
|
|
|
|
|
if (eat) {
|
|
|
|
|
__skb_pull(skb, eat);
|
|
|
|
|
len -= eat;
|
|
|
|
|
if (!len)
|
2017-04-27 08:15:40 +08:00
|
|
|
return 0;
|
2011-12-04 16:51:08 +08:00
|
|
|
}
|
2005-04-17 06:20:36 +08:00
|
|
|
eat = len;
|
|
|
|
|
k = 0;
|
2013-12-06 14:31:30 +08:00
|
|
|
shinfo = skb_shinfo(skb);
|
|
|
|
|
for (i = 0; i < shinfo->nr_frags; i++) {
|
|
|
|
|
int size = skb_frag_size(&shinfo->frags[i]);
|
2011-10-19 05:00:24 +08:00
|
|
|
|
|
|
|
|
if (size <= eat) {
|
2011-08-23 07:44:59 +08:00
|
|
|
skb_frag_unref(skb, i);
|
2011-10-19 05:00:24 +08:00
|
|
|
eat -= size;
|
2005-04-17 06:20:36 +08:00
|
|
|
} else {
|
2013-12-06 14:31:30 +08:00
|
|
|
shinfo->frags[k] = shinfo->frags[i];
|
2005-04-17 06:20:36 +08:00
|
|
|
if (eat) {
|
2019-07-30 22:40:33 +08:00
|
|
|
skb_frag_off_add(&shinfo->frags[k], eat);
|
2013-12-06 14:31:30 +08:00
|
|
|
skb_frag_size_sub(&shinfo->frags[k], eat);
|
2005-04-17 06:20:36 +08:00
|
|
|
eat = 0;
|
|
|
|
|
}
|
|
|
|
|
k++;
|
|
|
|
|
}
|
|
|
|
|
}
|
2013-12-06 14:31:30 +08:00
|
|
|
shinfo->nr_frags = k;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
|
skb->data_len -= len;
|
|
|
|
|
skb->len = skb->data_len;
|
2017-04-27 08:15:40 +08:00
|
|
|
return len;
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
|
2009-07-22 07:00:40 +08:00
|
|
|
/* Remove acked data from a packet in the transmit queue. */
|
2005-04-17 06:20:36 +08:00
|
|
|
int tcp_trim_head(struct sock *sk, struct sk_buff *skb, u32 len)
|
|
|
|
|
{
|
2017-04-27 08:15:40 +08:00
|
|
|
u32 delta_truesize;
|
|
|
|
|
|
2013-02-14 17:44:49 +08:00
|
|
|
if (skb_unclone(skb, GFP_ATOMIC))
|
2005-04-17 06:20:36 +08:00
|
|
|
return -ENOMEM;
|
|
|
|
|
|
2017-04-27 08:15:40 +08:00
|
|
|
delta_truesize = __pskb_trim_head(skb, len);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
|
TCP_SKB_CB(skb)->seq += len;
|
2006-08-30 07:44:56 +08:00
|
|
|
skb->ip_summed = CHECKSUM_PARTIAL;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2017-04-27 08:15:40 +08:00
|
|
|
if (delta_truesize) {
|
|
|
|
|
skb->truesize -= delta_truesize;
|
2019-10-11 11:17:46 +08:00
|
|
|
sk_wmem_queued_add(sk, -delta_truesize);
|
2017-04-27 08:15:40 +08:00
|
|
|
sk_mem_uncharge(sk, delta_truesize);
|
|
|
|
|
}
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2012-01-29 01:29:46 +08:00
|
|
|
/* Any change of skb->len requires recalculation of tso factor. */
|
2005-04-17 06:20:36 +08:00
|
|
|
if (tcp_skb_pcount(skb) > 1)
|
2015-06-12 00:15:17 +08:00
|
|
|
tcp_set_skb_tso_segs(skb, tcp_skb_mss(skb));
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
2013-02-22 16:59:06 +08:00
|
|
|
/* Calculate MSS not accounting any TCP options. */
|
|
|
|
|
static inline int __tcp_mtu_to_mss(struct sock *sk, int pmtu)
|
2006-03-21 09:53:41 +08:00
|
|
|
{
|
2011-10-21 17:22:42 +08:00
|
|
|
const struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
const struct inet_connection_sock *icsk = inet_csk(sk);
|
2006-03-21 09:53:41 +08:00
|
|
|
int mss_now;
|
|
|
|
|
|
|
|
|
|
/* Calculate base mss without TCP options:
|
|
|
|
|
It is MMS_S - sizeof(tcphdr) of rfc1122
|
|
|
|
|
*/
|
|
|
|
|
mss_now = pmtu - icsk->icsk_af_ops->net_header_len - sizeof(struct tcphdr);
|
|
|
|
|
|
ipv6: RTAX_FEATURE_ALLFRAG causes inefficient TCP segment sizing
Quoting Tore Anderson from :
https://bugzilla.kernel.org/show_bug.cgi?id=42572
When RTAX_FEATURE_ALLFRAG is set on a route, the effective TCP segment
size does not take into account the size of the IPv6 Fragmentation
header that needs to be included in outbound packets, causing every
transmitted TCP segment to be fragmented across two IPv6 packets, the
latter of which will only contain 8 bytes of actual payload.
RTAX_FEATURE_ALLFRAG is typically set on a route in response to
receving a ICMPv6 Packet Too Big message indicating a Path MTU of less
than 1280 bytes. 1280 bytes is the minimum IPv6 MTU, however ICMPv6
PTBs with MTU < 1280 are still valid, in particular when an IPv6
packet is sent to an IPv4 destination through a stateless translator.
Any ICMPv4 Need To Fragment packets originated from the IPv4 part of
the path will be translated to ICMPv6 PTB which may then indicate an
MTU of less than 1280.
The Linux kernel refuses to reduce the effective MTU to anything below
1280 bytes, instead it sets it to exactly 1280 bytes, and
RTAX_FEATURE_ALLFRAG is also set. However, the TCP segment size appears
to be set to 1240 bytes (1280 Path MTU - 40 bytes of IPv6 header),
instead of 1232 (additionally taking into account the 8 bytes required
by the IPv6 Fragmentation extension header).
This in turn results in rather inefficient transmission, as every
transmitted TCP segment now is split in two fragments containing
1232+8 bytes of payload.
After this patch, all the outgoing packets that includes a
Fragmentation header all are "atomic" or "non-fragmented" fragments,
i.e., they both have Offset=0 and More Fragments=0.
With help from David S. Miller
Reported-by: Tore Anderson <tore@fud.no>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Maciej Żenczykowski <maze@google.com>
Cc: Tom Herbert <therbert@google.com>
Tested-by: Tore Anderson <tore@fud.no>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-04-24 15:37:38 +08:00
|
|
|
/* IPv6 adds a frag_hdr in case RTAX_FEATURE_ALLFRAG is set */
|
|
|
|
|
if (icsk->icsk_af_ops->net_frag_header_len) {
|
|
|
|
|
const struct dst_entry *dst = __sk_dst_get(sk);
|
|
|
|
|
|
|
|
|
|
if (dst && dst_allfrag(dst))
|
|
|
|
|
mss_now -= icsk->icsk_af_ops->net_frag_header_len;
|
|
|
|
|
}
|
|
|
|
|
|
2006-03-21 09:53:41 +08:00
|
|
|
/* Clamp it (mss_clamp does not include tcp options) */
|
|
|
|
|
if (mss_now > tp->rx_opt.mss_clamp)
|
|
|
|
|
mss_now = tp->rx_opt.mss_clamp;
|
|
|
|
|
|
|
|
|
|
/* Now subtract optional transport overhead */
|
|
|
|
|
mss_now -= icsk->icsk_ext_hdr_len;
|
|
|
|
|
|
|
|
|
|
/* Then reserve room for full set of TCP options and 8 bytes of data */
|
2019-06-07 00:15:31 +08:00
|
|
|
mss_now = max(mss_now, sock_net(sk)->ipv4.sysctl_tcp_min_snd_mss);
|
2006-03-21 09:53:41 +08:00
|
|
|
return mss_now;
|
|
|
|
|
}
|
|
|
|
|
|
2013-02-22 16:59:06 +08:00
|
|
|
/* Calculate MSS. Not accounting for SACKs here. */
|
|
|
|
|
int tcp_mtu_to_mss(struct sock *sk, int pmtu)
|
|
|
|
|
{
|
|
|
|
|
/* Subtract TCP options size, not including SACKs */
|
|
|
|
|
return __tcp_mtu_to_mss(sk, pmtu) -
|
|
|
|
|
(tcp_sk(sk)->tcp_header_len - sizeof(struct tcphdr));
|
|
|
|
|
}
|
|
|
|
|
|
2006-03-21 09:53:41 +08:00
|
|
|
/* Inverse of above */
|
ipv6: RTAX_FEATURE_ALLFRAG causes inefficient TCP segment sizing
Quoting Tore Anderson from :
https://bugzilla.kernel.org/show_bug.cgi?id=42572
When RTAX_FEATURE_ALLFRAG is set on a route, the effective TCP segment
size does not take into account the size of the IPv6 Fragmentation
header that needs to be included in outbound packets, causing every
transmitted TCP segment to be fragmented across two IPv6 packets, the
latter of which will only contain 8 bytes of actual payload.
RTAX_FEATURE_ALLFRAG is typically set on a route in response to
receving a ICMPv6 Packet Too Big message indicating a Path MTU of less
than 1280 bytes. 1280 bytes is the minimum IPv6 MTU, however ICMPv6
PTBs with MTU < 1280 are still valid, in particular when an IPv6
packet is sent to an IPv4 destination through a stateless translator.
Any ICMPv4 Need To Fragment packets originated from the IPv4 part of
the path will be translated to ICMPv6 PTB which may then indicate an
MTU of less than 1280.
The Linux kernel refuses to reduce the effective MTU to anything below
1280 bytes, instead it sets it to exactly 1280 bytes, and
RTAX_FEATURE_ALLFRAG is also set. However, the TCP segment size appears
to be set to 1240 bytes (1280 Path MTU - 40 bytes of IPv6 header),
instead of 1232 (additionally taking into account the 8 bytes required
by the IPv6 Fragmentation extension header).
This in turn results in rather inefficient transmission, as every
transmitted TCP segment now is split in two fragments containing
1232+8 bytes of payload.
After this patch, all the outgoing packets that includes a
Fragmentation header all are "atomic" or "non-fragmented" fragments,
i.e., they both have Offset=0 and More Fragments=0.
With help from David S. Miller
Reported-by: Tore Anderson <tore@fud.no>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Maciej Żenczykowski <maze@google.com>
Cc: Tom Herbert <therbert@google.com>
Tested-by: Tore Anderson <tore@fud.no>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-04-24 15:37:38 +08:00
|
|
|
int tcp_mss_to_mtu(struct sock *sk, int mss)
|
2006-03-21 09:53:41 +08:00
|
|
|
{
|
2011-10-21 17:22:42 +08:00
|
|
|
const struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
const struct inet_connection_sock *icsk = inet_csk(sk);
|
2006-03-21 09:53:41 +08:00
|
|
|
int mtu;
|
|
|
|
|
|
|
|
|
|
mtu = mss +
|
|
|
|
|
tp->tcp_header_len +
|
|
|
|
|
icsk->icsk_ext_hdr_len +
|
|
|
|
|
icsk->icsk_af_ops->net_header_len;
|
|
|
|
|
|
ipv6: RTAX_FEATURE_ALLFRAG causes inefficient TCP segment sizing
Quoting Tore Anderson from :
https://bugzilla.kernel.org/show_bug.cgi?id=42572
When RTAX_FEATURE_ALLFRAG is set on a route, the effective TCP segment
size does not take into account the size of the IPv6 Fragmentation
header that needs to be included in outbound packets, causing every
transmitted TCP segment to be fragmented across two IPv6 packets, the
latter of which will only contain 8 bytes of actual payload.
RTAX_FEATURE_ALLFRAG is typically set on a route in response to
receving a ICMPv6 Packet Too Big message indicating a Path MTU of less
than 1280 bytes. 1280 bytes is the minimum IPv6 MTU, however ICMPv6
PTBs with MTU < 1280 are still valid, in particular when an IPv6
packet is sent to an IPv4 destination through a stateless translator.
Any ICMPv4 Need To Fragment packets originated from the IPv4 part of
the path will be translated to ICMPv6 PTB which may then indicate an
MTU of less than 1280.
The Linux kernel refuses to reduce the effective MTU to anything below
1280 bytes, instead it sets it to exactly 1280 bytes, and
RTAX_FEATURE_ALLFRAG is also set. However, the TCP segment size appears
to be set to 1240 bytes (1280 Path MTU - 40 bytes of IPv6 header),
instead of 1232 (additionally taking into account the 8 bytes required
by the IPv6 Fragmentation extension header).
This in turn results in rather inefficient transmission, as every
transmitted TCP segment now is split in two fragments containing
1232+8 bytes of payload.
After this patch, all the outgoing packets that includes a
Fragmentation header all are "atomic" or "non-fragmented" fragments,
i.e., they both have Offset=0 and More Fragments=0.
With help from David S. Miller
Reported-by: Tore Anderson <tore@fud.no>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Maciej Żenczykowski <maze@google.com>
Cc: Tom Herbert <therbert@google.com>
Tested-by: Tore Anderson <tore@fud.no>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-04-24 15:37:38 +08:00
|
|
|
/* IPv6 adds a frag_hdr in case RTAX_FEATURE_ALLFRAG is set */
|
|
|
|
|
if (icsk->icsk_af_ops->net_frag_header_len) {
|
|
|
|
|
const struct dst_entry *dst = __sk_dst_get(sk);
|
|
|
|
|
|
|
|
|
|
if (dst && dst_allfrag(dst))
|
|
|
|
|
mtu += icsk->icsk_af_ops->net_frag_header_len;
|
|
|
|
|
}
|
2006-03-21 09:53:41 +08:00
|
|
|
return mtu;
|
|
|
|
|
}
|
2016-09-20 11:39:19 +08:00
|
|
|
EXPORT_SYMBOL(tcp_mss_to_mtu);
|
2006-03-21 09:53:41 +08:00
|
|
|
|
2009-07-22 07:00:40 +08:00
|
|
|
/* MTU probing init per socket */
|
2006-03-21 09:53:41 +08:00
|
|
|
void tcp_mtup_init(struct sock *sk)
|
|
|
|
|
{
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
struct inet_connection_sock *icsk = inet_csk(sk);
|
2015-02-10 09:53:16 +08:00
|
|
|
struct net *net = sock_net(sk);
|
2006-03-21 09:53:41 +08:00
|
|
|
|
2015-02-10 09:53:16 +08:00
|
|
|
icsk->icsk_mtup.enabled = net->ipv4.sysctl_tcp_mtu_probing > 1;
|
2006-03-21 09:53:41 +08:00
|
|
|
icsk->icsk_mtup.search_high = tp->rx_opt.mss_clamp + sizeof(struct tcphdr) +
|
2007-02-09 22:24:47 +08:00
|
|
|
icsk->icsk_af_ops->net_header_len;
|
2015-02-10 09:53:16 +08:00
|
|
|
icsk->icsk_mtup.search_low = tcp_mss_to_mtu(sk, net->ipv4.sysctl_tcp_base_mss);
|
2006-03-21 09:53:41 +08:00
|
|
|
icsk->icsk_mtup.probe_size = 0;
|
2015-03-06 11:18:24 +08:00
|
|
|
if (icsk->icsk_mtup.enabled)
|
2017-05-17 05:00:08 +08:00
|
|
|
icsk->icsk_mtup.probe_timestamp = tcp_jiffies32;
|
2006-03-21 09:53:41 +08:00
|
|
|
}
|
2010-07-10 05:22:10 +08:00
|
|
|
EXPORT_SYMBOL(tcp_mtup_init);
|
2006-03-21 09:53:41 +08:00
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/* This function synchronize snd mss to current pmtu/exthdr set.
|
|
|
|
|
|
|
|
|
|
tp->rx_opt.user_mss is mss set by user by TCP_MAXSEG. It does NOT counts
|
|
|
|
|
for TCP options, but includes only bare TCP header.
|
|
|
|
|
|
|
|
|
|
tp->rx_opt.mss_clamp is mss negotiated at connection setup.
|
2005-11-11 09:13:47 +08:00
|
|
|
It is minimum of user_mss and mss received with SYN.
|
2005-04-17 06:20:36 +08:00
|
|
|
It also does not include TCP options.
|
|
|
|
|
|
2005-12-14 15:26:10 +08:00
|
|
|
inet_csk(sk)->icsk_pmtu_cookie is last pmtu, seen by this function.
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
|
tp->mss_cache is current effective sending mss, including
|
|
|
|
|
all tcp options except for SACKs. It is evaluated,
|
|
|
|
|
taking into account current pmtu, but never exceeds
|
|
|
|
|
tp->rx_opt.mss_clamp.
|
|
|
|
|
|
|
|
|
|
NOTE1. rfc1122 clearly states that advertised MSS
|
|
|
|
|
DOES NOT include either tcp or ip options.
|
|
|
|
|
|
2005-12-14 15:26:10 +08:00
|
|
|
NOTE2. inet_csk(sk)->icsk_pmtu_cookie and tp->mss_cache
|
|
|
|
|
are READ ONLY outside this function. --ANK (980731)
|
2005-04-17 06:20:36 +08:00
|
|
|
*/
|
|
|
|
|
unsigned int tcp_sync_mss(struct sock *sk, u32 pmtu)
|
|
|
|
|
{
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
2005-12-14 15:26:10 +08:00
|
|
|
struct inet_connection_sock *icsk = inet_csk(sk);
|
2006-03-21 09:53:41 +08:00
|
|
|
int mss_now;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2006-03-21 09:53:41 +08:00
|
|
|
if (icsk->icsk_mtup.search_high > pmtu)
|
|
|
|
|
icsk->icsk_mtup.search_high = pmtu;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2006-03-21 09:53:41 +08:00
|
|
|
mss_now = tcp_mtu_to_mss(sk, pmtu);
|
2008-01-01 06:57:40 +08:00
|
|
|
mss_now = tcp_bound_to_half_wnd(tp, mss_now);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
|
/* And store cached results */
|
2005-12-14 15:26:10 +08:00
|
|
|
icsk->icsk_pmtu_cookie = pmtu;
|
2006-03-21 09:53:41 +08:00
|
|
|
if (icsk->icsk_mtup.enabled)
|
|
|
|
|
mss_now = min(mss_now, tcp_mtu_to_mss(sk, icsk->icsk_mtup.search_low));
|
2005-07-06 06:24:38 +08:00
|
|
|
tp->mss_cache = mss_now;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
|
return mss_now;
|
|
|
|
|
}
|
2010-07-10 05:22:10 +08:00
|
|
|
EXPORT_SYMBOL(tcp_sync_mss);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
|
/* Compute the current effective MSS, taking SACKs and IP options,
|
|
|
|
|
* and even PMTU discovery events into account.
|
|
|
|
|
*/
|
2009-03-14 22:23:05 +08:00
|
|
|
unsigned int tcp_current_mss(struct sock *sk)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
2011-10-21 17:22:42 +08:00
|
|
|
const struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
const struct dst_entry *dst = __sk_dst_get(sk);
|
2005-07-06 06:24:38 +08:00
|
|
|
u32 mss_now;
|
2012-04-15 13:58:06 +08:00
|
|
|
unsigned int header_len;
|
2008-07-19 15:04:31 +08:00
|
|
|
struct tcp_out_options opts;
|
|
|
|
|
struct tcp_md5sig_key *md5;
|
2005-07-06 06:24:38 +08:00
|
|
|
|
|
|
|
|
mss_now = tp->mss_cache;
|
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
if (dst) {
|
|
|
|
|
u32 mtu = dst_mtu(dst);
|
2005-12-14 15:26:10 +08:00
|
|
|
if (mtu != inet_csk(sk)->icsk_pmtu_cookie)
|
2005-04-17 06:20:36 +08:00
|
|
|
mss_now = tcp_sync_mss(sk, mtu);
|
|
|
|
|
}
|
|
|
|
|
|
2008-07-19 15:04:31 +08:00
|
|
|
header_len = tcp_established_options(sk, NULL, &opts, &md5) +
|
|
|
|
|
sizeof(struct tcphdr);
|
|
|
|
|
/* The mss_cache is sized based on tp->tcp_header_len, which assumes
|
|
|
|
|
* some common options. If this is an odd packet (because we have SACK
|
|
|
|
|
* blocks etc) then our calculated header_len will be different, and
|
|
|
|
|
* we have to adjust mss_now correspondingly */
|
|
|
|
|
if (header_len != tp->tcp_header_len) {
|
|
|
|
|
int delta = (int) header_len - tp->tcp_header_len;
|
|
|
|
|
mss_now -= delta;
|
|
|
|
|
}
|
2006-11-15 11:07:45 +08:00
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
return mss_now;
|
|
|
|
|
}
|
|
|
|
|
|
2014-04-18 12:27:46 +08:00
|
|
|
/* RFC2861, slow part. Adjust cwnd, after it was not full during one rto.
|
|
|
|
|
* As additional protections, we do not touch cwnd in retransmission phases,
|
|
|
|
|
* and if application hit its sndbuf limit recently.
|
|
|
|
|
*/
|
|
|
|
|
static void tcp_cwnd_application_limited(struct sock *sk)
|
|
|
|
|
{
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
|
|
|
|
|
if (inet_csk(sk)->icsk_ca_state == TCP_CA_Open &&
|
|
|
|
|
sk->sk_socket && !test_bit(SOCK_NOSPACE, &sk->sk_socket->flags)) {
|
|
|
|
|
/* Limited by application or receiver window. */
|
|
|
|
|
u32 init_win = tcp_init_cwnd(tp, __sk_dst_get(sk));
|
|
|
|
|
u32 win_used = max(tp->snd_cwnd_used, init_win);
|
|
|
|
|
if (win_used < tp->snd_cwnd) {
|
|
|
|
|
tp->snd_ssthresh = tcp_current_ssthresh(sk);
|
|
|
|
|
tp->snd_cwnd = (tp->snd_cwnd + win_used) >> 1;
|
|
|
|
|
}
|
|
|
|
|
tp->snd_cwnd_used = 0;
|
|
|
|
|
}
|
2017-05-17 05:00:04 +08:00
|
|
|
tp->snd_cwnd_stamp = tcp_jiffies32;
|
2014-04-18 12:27:46 +08:00
|
|
|
}
|
|
|
|
|
|
2014-05-22 22:41:08 +08:00
|
|
|
static void tcp_cwnd_validate(struct sock *sk, bool is_cwnd_limited)
|
2005-07-06 06:18:51 +08:00
|
|
|
{
|
2017-05-06 03:53:23 +08:00
|
|
|
const struct tcp_congestion_ops *ca_ops = inet_csk(sk)->icsk_ca_ops;
|
[TCP]: Sed magic converts func(sk, tp, ...) -> func(sk, ...)
This is (mostly) automated change using magic:
sed -e '/struct sock \*sk/ N' -e '/struct sock \*sk/ N'
-e '/struct sock \*sk/ N' -e '/struct sock \*sk/ N'
-e 's|struct sock \*sk,[\n\t ]*struct tcp_sock \*tp\([^{]*\n{\n\)|
struct sock \*sk\1\tstruct tcp_sock *tp = tcp_sk(sk);\n|g'
-e 's|struct sock \*sk, struct tcp_sock \*tp|
struct sock \*sk|g' -e 's|sk, tp\([^-]\)|sk\1|g'
Fixed four unused variable (tp) warnings that were introduced.
In addition, manually added newlines after local variables and
tweaked function arguments positioning.
$ gcc --version
gcc (GCC) 4.1.1 20060525 (Red Hat 4.1.1-1)
...
$ codiff -fV built-in.o.old built-in.o.new
net/ipv4/route.c:
rt_cache_flush | +14
1 function changed, 14 bytes added
net/ipv4/tcp.c:
tcp_setsockopt | -5
tcp_sendpage | -25
tcp_sendmsg | -16
3 functions changed, 46 bytes removed
net/ipv4/tcp_input.c:
tcp_try_undo_recovery | +3
tcp_try_undo_dsack | +2
tcp_mark_head_lost | -12
tcp_ack | -15
tcp_event_data_recv | -32
tcp_rcv_state_process | -10
tcp_rcv_established | +1
7 functions changed, 6 bytes added, 69 bytes removed, diff: -63
net/ipv4/tcp_output.c:
update_send_head | -9
tcp_transmit_skb | +19
tcp_cwnd_validate | +1
tcp_write_wakeup | -17
__tcp_push_pending_frames | -25
tcp_push_one | -8
tcp_send_fin | -4
7 functions changed, 20 bytes added, 63 bytes removed, diff: -43
built-in.o.new:
18 functions changed, 40 bytes added, 178 bytes removed, diff: -138
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-21 13:18:02 +08:00
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
2005-07-06 06:18:51 +08:00
|
|
|
|
2014-05-22 22:41:08 +08:00
|
|
|
/* Track the maximum number of outstanding packets in each
|
|
|
|
|
* window, and remember whether we were cwnd-limited then.
|
|
|
|
|
*/
|
|
|
|
|
if (!before(tp->snd_una, tp->max_packets_seq) ||
|
tcp: fix cwnd-limited bug for TSO deferral where we send nothing
When cwnd is not a multiple of the TSO skb size of N*MSS, we can get
into persistent scenarios where we have the following sequence:
(1) ACK for full-sized skb of N*MSS arrives
-> tcp_write_xmit() transmit full-sized skb with N*MSS
-> move pacing release time forward
-> exit tcp_write_xmit() because pacing time is in the future
(2) TSQ callback or TCP internal pacing timer fires
-> try to transmit next skb, but TSO deferral finds remainder of
available cwnd is not big enough to trigger an immediate send
now, so we defer sending until the next ACK.
(3) repeat...
So we can get into a case where we never mark ourselves as
cwnd-limited for many seconds at a time, even with
bulk/infinite-backlog senders, because:
o In case (1) above, every time in tcp_write_xmit() we have enough
cwnd to send a full-sized skb, we are not fully using the cwnd
(because cwnd is not a multiple of the TSO skb size). So every time we
send data, we are not cwnd limited, and so in the cwnd-limited
tracking code in tcp_cwnd_validate() we mark ourselves as not
cwnd-limited.
o In case (2) above, every time in tcp_write_xmit() that we try to
transmit the "remainder" of the cwnd but defer, we set the local
variable is_cwnd_limited to true, but we do not send any packets, so
sent_pkts is zero, so we don't call the cwnd-limited logic to update
tp->is_cwnd_limited.
Fixes: ca8a22634381 ("tcp: make cwnd-limited checks measurement-based, and gentler")
Reported-by: Ingemar Johansson <ingemar.s.johansson@ericsson.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20201209035759.1225145-1-ncardwell.kernel@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-09 11:57:59 +08:00
|
|
|
tp->packets_out > tp->max_packets_out ||
|
|
|
|
|
is_cwnd_limited) {
|
2014-05-22 22:41:08 +08:00
|
|
|
tp->max_packets_out = tp->packets_out;
|
|
|
|
|
tp->max_packets_seq = tp->snd_nxt;
|
|
|
|
|
tp->is_cwnd_limited = is_cwnd_limited;
|
|
|
|
|
}
|
tcp: fix cwnd limited checking to improve congestion control
Yuchung discovered tcp_is_cwnd_limited() was returning false in
slow start phase even if the application filled the socket write queue.
All congestion modules take into account tcp_is_cwnd_limited()
before increasing cwnd, so this behavior limits slow start from
probing the bandwidth at full speed.
The problem is that even if write queue is full (aka we are _not_
application limited), cwnd can be under utilized if TSO should auto
defer or TCP Small queues decided to hold packets.
So the in_flight can be kept to smaller value, and we can get to the
point tcp_is_cwnd_limited() returns false.
With TCP Small Queues and FQ/pacing, this issue is more visible.
We fix this by having tcp_cwnd_validate(), which is supposed to track
such things, take into account unsent_segs, the number of segs that we
are not sending at the moment due to TSO or TSQ, but intend to send
real soon. Then when we are cwnd-limited, remember this fact while we
are processing the window of ACKs that comes back.
For example, suppose we have a brand new connection with cwnd=10; we
are in slow start, and we send a flight of 9 packets. By the time we
have received ACKs for all 9 packets we want our cwnd to be 18.
We implement this by setting tp->lsnd_pending to 9, and
considering ourselves to be cwnd-limited while cwnd is less than
twice tp->lsnd_pending (2*9 -> 18).
This makes tcp_is_cwnd_limited() more understandable, by removing
the GSO/TSO kludge, that tried to work around the issue.
Note the in_flight parameter can be removed in a followup cleanup
patch.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-05-01 02:58:13 +08:00
|
|
|
|
2014-05-03 12:18:05 +08:00
|
|
|
if (tcp_is_cwnd_limited(sk)) {
|
2005-07-06 06:18:51 +08:00
|
|
|
/* Network is feed fully. */
|
|
|
|
|
tp->snd_cwnd_used = 0;
|
2017-05-17 05:00:04 +08:00
|
|
|
tp->snd_cwnd_stamp = tcp_jiffies32;
|
2005-07-06 06:18:51 +08:00
|
|
|
} else {
|
|
|
|
|
/* Network starves. */
|
|
|
|
|
if (tp->packets_out > tp->snd_cwnd_used)
|
|
|
|
|
tp->snd_cwnd_used = tp->packets_out;
|
|
|
|
|
|
2017-10-27 12:54:59 +08:00
|
|
|
if (sock_net(sk)->ipv4.sysctl_tcp_slow_start_after_idle &&
|
2017-05-17 05:00:04 +08:00
|
|
|
(s32)(tcp_jiffies32 - tp->snd_cwnd_stamp) >= inet_csk(sk)->icsk_rto &&
|
2017-05-06 03:53:23 +08:00
|
|
|
!ca_ops->cong_control)
|
2005-07-06 06:18:51 +08:00
|
|
|
tcp_cwnd_application_limited(sk);
|
2016-11-28 15:07:16 +08:00
|
|
|
|
|
|
|
|
/* The following conditions together indicate the starvation
|
|
|
|
|
* is caused by insufficient sender buffer:
|
|
|
|
|
* 1) just sent some data (see tcp_write_xmit)
|
|
|
|
|
* 2) not cwnd limited (this else condition)
|
2017-10-06 13:21:27 +08:00
|
|
|
* 3) no more data to send (tcp_write_queue_empty())
|
2016-11-28 15:07:16 +08:00
|
|
|
* 4) application is hitting buffer limit (SOCK_NOSPACE)
|
|
|
|
|
*/
|
2017-10-06 13:21:27 +08:00
|
|
|
if (tcp_write_queue_empty(sk) && sk->sk_socket &&
|
2016-11-28 15:07:16 +08:00
|
|
|
test_bit(SOCK_NOSPACE, &sk->sk_socket->flags) &&
|
|
|
|
|
(1 << sk->sk_state) & (TCPF_ESTABLISHED | TCPF_CLOSE_WAIT))
|
|
|
|
|
tcp_chrono_start(sk, TCP_CHRONO_SNDBUF_LIMITED);
|
2005-07-06 06:18:51 +08:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
tcp: refine TSO splits
While investigating performance problems on small RPC workloads,
I noticed linux TCP stack was always splitting the last TSO skb
into two parts (skbs). One being a multiple of MSS, and a small one
with the Push flag. This split is done even if TCP_NODELAY is set,
or if no small packet is in flight.
Example with request/response of 4K/4K
IP A > B: . ack 68432 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: . 65537:68433(2896) ack 69632 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: P 68433:69633(1200) ack 69632 win 2783 <nop,nop,timestamp 6524593 6525001>
IP B > A: . ack 68433 win 2768 <nop,nop,timestamp 6525001 6524593>
IP B > A: . 69632:72528(2896) ack 69633 win 2768 <nop,nop,timestamp 6525001 6524593>
IP B > A: P 72528:73728(1200) ack 69633 win 2768 <nop,nop,timestamp 6525001 6524593>
IP A > B: . ack 72528 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: . 69633:72529(2896) ack 73728 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: P 72529:73729(1200) ack 73728 win 2783 <nop,nop,timestamp 6524593 6525001>
We can avoid this split by including the Nagle tests at the right place.
Note : If some NIC had trouble sending TSO packets with a partial
last segment, we would have hit the problem in GRO/forwarding workload already.
tcp_minshall_update() is moved to tcp_output.c and is updated as we might
feed a TSO packet with a partial last segment.
This patch tremendously improves performance, as the traffic now looks
like :
IP A > B: . ack 98304 win 2783 <nop,nop,timestamp 6834277 6834685>
IP A > B: P 94209:98305(4096) ack 98304 win 2783 <nop,nop,timestamp 6834277 6834685>
IP B > A: . ack 98305 win 2768 <nop,nop,timestamp 6834686 6834277>
IP B > A: P 98304:102400(4096) ack 98305 win 2768 <nop,nop,timestamp 6834686 6834277>
IP A > B: . ack 102400 win 2783 <nop,nop,timestamp 6834279 6834686>
IP A > B: P 98305:102401(4096) ack 102400 win 2783 <nop,nop,timestamp 6834279 6834686>
IP B > A: . ack 102401 win 2768 <nop,nop,timestamp 6834687 6834279>
IP B > A: P 102400:106496(4096) ack 102401 win 2768 <nop,nop,timestamp 6834687 6834279>
IP A > B: . ack 106496 win 2783 <nop,nop,timestamp 6834280 6834687>
IP A > B: P 102401:106497(4096) ack 106496 win 2783 <nop,nop,timestamp 6834280 6834687>
IP B > A: . ack 106497 win 2768 <nop,nop,timestamp 6834688 6834280>
IP B > A: P 106496:110592(4096) ack 106497 win 2768 <nop,nop,timestamp 6834688 6834280>
Before :
lpq83:~# nstat >/dev/null;perf stat ./super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K
280774
Performance counter stats for './super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K':
205719.049006 task-clock # 9.278 CPUs utilized
8,449,968 context-switches # 0.041 M/sec
1,935,997 CPU-migrations # 0.009 M/sec
160,541 page-faults # 0.780 K/sec
548,478,722,290 cycles # 2.666 GHz [83.20%]
455,240,670,857 stalled-cycles-frontend # 83.00% frontend cycles idle [83.48%]
272,881,454,275 stalled-cycles-backend # 49.75% backend cycles idle [66.73%]
166,091,460,030 instructions # 0.30 insns per cycle
# 2.74 stalled cycles per insn [83.39%]
29,150,229,399 branches # 141.699 M/sec [83.30%]
1,943,814,026 branch-misses # 6.67% of all branches [83.32%]
22.173517844 seconds time elapsed
lpq83:~# nstat | egrep "IpOutRequests|IpExtOutOctets"
IpOutRequests 16851063 0.0
IpExtOutOctets 23878580777 0.0
After patch :
lpq83:~# nstat >/dev/null;perf stat ./super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K
280877
Performance counter stats for './super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K':
107496.071918 task-clock # 4.847 CPUs utilized
5,635,458 context-switches # 0.052 M/sec
1,374,707 CPU-migrations # 0.013 M/sec
160,920 page-faults # 0.001 M/sec
281,500,010,924 cycles # 2.619 GHz [83.28%]
228,865,069,307 stalled-cycles-frontend # 81.30% frontend cycles idle [83.38%]
142,462,742,658 stalled-cycles-backend # 50.61% backend cycles idle [66.81%]
95,227,712,566 instructions # 0.34 insns per cycle
# 2.40 stalled cycles per insn [83.43%]
16,209,868,171 branches # 150.795 M/sec [83.20%]
874,252,952 branch-misses # 5.39% of all branches [83.37%]
22.175821286 seconds time elapsed
lpq83:~# nstat | egrep "IpOutRequests|IpExtOutOctets"
IpOutRequests 11239428 0.0
IpExtOutOctets 23595191035 0.0
Indeed, the occupancy of tx skbs (IpExtOutOctets/IpOutRequests) is higher :
2099 instead of 1417, thus helping GRO to be more efficient when using FQ packet
scheduler.
Many thanks to Neal for review and ideas.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Cc: Van Jacobson <vanj@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Tested-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-14 05:51:23 +08:00
|
|
|
/* Minshall's variant of the Nagle send check. */
|
|
|
|
|
static bool tcp_minshall_check(const struct tcp_sock *tp)
|
|
|
|
|
{
|
|
|
|
|
return after(tp->snd_sml, tp->snd_una) &&
|
|
|
|
|
!after(tp->snd_sml, tp->snd_nxt);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Update snd_sml if this skb is under mss
|
|
|
|
|
* Note that a TSO packet might end with a sub-mss segment
|
|
|
|
|
* The test is really :
|
|
|
|
|
* if ((skb->len % mss) != 0)
|
|
|
|
|
* tp->snd_sml = TCP_SKB_CB(skb)->end_seq;
|
|
|
|
|
* But we can avoid doing the divide again given we already have
|
|
|
|
|
* skb_pcount = skb->len / mss_now
|
2007-12-25 13:33:45 +08:00
|
|
|
*/
|
tcp: refine TSO splits
While investigating performance problems on small RPC workloads,
I noticed linux TCP stack was always splitting the last TSO skb
into two parts (skbs). One being a multiple of MSS, and a small one
with the Push flag. This split is done even if TCP_NODELAY is set,
or if no small packet is in flight.
Example with request/response of 4K/4K
IP A > B: . ack 68432 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: . 65537:68433(2896) ack 69632 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: P 68433:69633(1200) ack 69632 win 2783 <nop,nop,timestamp 6524593 6525001>
IP B > A: . ack 68433 win 2768 <nop,nop,timestamp 6525001 6524593>
IP B > A: . 69632:72528(2896) ack 69633 win 2768 <nop,nop,timestamp 6525001 6524593>
IP B > A: P 72528:73728(1200) ack 69633 win 2768 <nop,nop,timestamp 6525001 6524593>
IP A > B: . ack 72528 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: . 69633:72529(2896) ack 73728 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: P 72529:73729(1200) ack 73728 win 2783 <nop,nop,timestamp 6524593 6525001>
We can avoid this split by including the Nagle tests at the right place.
Note : If some NIC had trouble sending TSO packets with a partial
last segment, we would have hit the problem in GRO/forwarding workload already.
tcp_minshall_update() is moved to tcp_output.c and is updated as we might
feed a TSO packet with a partial last segment.
This patch tremendously improves performance, as the traffic now looks
like :
IP A > B: . ack 98304 win 2783 <nop,nop,timestamp 6834277 6834685>
IP A > B: P 94209:98305(4096) ack 98304 win 2783 <nop,nop,timestamp 6834277 6834685>
IP B > A: . ack 98305 win 2768 <nop,nop,timestamp 6834686 6834277>
IP B > A: P 98304:102400(4096) ack 98305 win 2768 <nop,nop,timestamp 6834686 6834277>
IP A > B: . ack 102400 win 2783 <nop,nop,timestamp 6834279 6834686>
IP A > B: P 98305:102401(4096) ack 102400 win 2783 <nop,nop,timestamp 6834279 6834686>
IP B > A: . ack 102401 win 2768 <nop,nop,timestamp 6834687 6834279>
IP B > A: P 102400:106496(4096) ack 102401 win 2768 <nop,nop,timestamp 6834687 6834279>
IP A > B: . ack 106496 win 2783 <nop,nop,timestamp 6834280 6834687>
IP A > B: P 102401:106497(4096) ack 106496 win 2783 <nop,nop,timestamp 6834280 6834687>
IP B > A: . ack 106497 win 2768 <nop,nop,timestamp 6834688 6834280>
IP B > A: P 106496:110592(4096) ack 106497 win 2768 <nop,nop,timestamp 6834688 6834280>
Before :
lpq83:~# nstat >/dev/null;perf stat ./super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K
280774
Performance counter stats for './super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K':
205719.049006 task-clock # 9.278 CPUs utilized
8,449,968 context-switches # 0.041 M/sec
1,935,997 CPU-migrations # 0.009 M/sec
160,541 page-faults # 0.780 K/sec
548,478,722,290 cycles # 2.666 GHz [83.20%]
455,240,670,857 stalled-cycles-frontend # 83.00% frontend cycles idle [83.48%]
272,881,454,275 stalled-cycles-backend # 49.75% backend cycles idle [66.73%]
166,091,460,030 instructions # 0.30 insns per cycle
# 2.74 stalled cycles per insn [83.39%]
29,150,229,399 branches # 141.699 M/sec [83.30%]
1,943,814,026 branch-misses # 6.67% of all branches [83.32%]
22.173517844 seconds time elapsed
lpq83:~# nstat | egrep "IpOutRequests|IpExtOutOctets"
IpOutRequests 16851063 0.0
IpExtOutOctets 23878580777 0.0
After patch :
lpq83:~# nstat >/dev/null;perf stat ./super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K
280877
Performance counter stats for './super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K':
107496.071918 task-clock # 4.847 CPUs utilized
5,635,458 context-switches # 0.052 M/sec
1,374,707 CPU-migrations # 0.013 M/sec
160,920 page-faults # 0.001 M/sec
281,500,010,924 cycles # 2.619 GHz [83.28%]
228,865,069,307 stalled-cycles-frontend # 81.30% frontend cycles idle [83.38%]
142,462,742,658 stalled-cycles-backend # 50.61% backend cycles idle [66.81%]
95,227,712,566 instructions # 0.34 insns per cycle
# 2.40 stalled cycles per insn [83.43%]
16,209,868,171 branches # 150.795 M/sec [83.20%]
874,252,952 branch-misses # 5.39% of all branches [83.37%]
22.175821286 seconds time elapsed
lpq83:~# nstat | egrep "IpOutRequests|IpExtOutOctets"
IpOutRequests 11239428 0.0
IpExtOutOctets 23595191035 0.0
Indeed, the occupancy of tx skbs (IpExtOutOctets/IpOutRequests) is higher :
2099 instead of 1417, thus helping GRO to be more efficient when using FQ packet
scheduler.
Many thanks to Neal for review and ideas.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Cc: Van Jacobson <vanj@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Tested-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-14 05:51:23 +08:00
|
|
|
static void tcp_minshall_update(struct tcp_sock *tp, unsigned int mss_now,
|
|
|
|
|
const struct sk_buff *skb)
|
|
|
|
|
{
|
|
|
|
|
if (skb->len < tcp_skb_pcount(skb) * mss_now)
|
|
|
|
|
tp->snd_sml = TCP_SKB_CB(skb)->end_seq;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Return false, if packet can be sent now without violation Nagle's rules:
|
|
|
|
|
* 1. It is full sized. (provided by caller in %partial bool)
|
|
|
|
|
* 2. Or it contains FIN. (already checked by caller)
|
|
|
|
|
* 3. Or TCP_CORK is not set, and TCP_NODELAY is set.
|
|
|
|
|
* 4. Or TCP_CORK is not set, and all sent packets are ACKed.
|
|
|
|
|
* With Minshall's modification: all sent small packets are ACKed.
|
|
|
|
|
*/
|
|
|
|
|
static bool tcp_nagle_check(bool partial, const struct tcp_sock *tp,
|
2014-03-24 14:49:34 +08:00
|
|
|
int nonagle)
|
tcp: refine TSO splits
While investigating performance problems on small RPC workloads,
I noticed linux TCP stack was always splitting the last TSO skb
into two parts (skbs). One being a multiple of MSS, and a small one
with the Push flag. This split is done even if TCP_NODELAY is set,
or if no small packet is in flight.
Example with request/response of 4K/4K
IP A > B: . ack 68432 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: . 65537:68433(2896) ack 69632 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: P 68433:69633(1200) ack 69632 win 2783 <nop,nop,timestamp 6524593 6525001>
IP B > A: . ack 68433 win 2768 <nop,nop,timestamp 6525001 6524593>
IP B > A: . 69632:72528(2896) ack 69633 win 2768 <nop,nop,timestamp 6525001 6524593>
IP B > A: P 72528:73728(1200) ack 69633 win 2768 <nop,nop,timestamp 6525001 6524593>
IP A > B: . ack 72528 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: . 69633:72529(2896) ack 73728 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: P 72529:73729(1200) ack 73728 win 2783 <nop,nop,timestamp 6524593 6525001>
We can avoid this split by including the Nagle tests at the right place.
Note : If some NIC had trouble sending TSO packets with a partial
last segment, we would have hit the problem in GRO/forwarding workload already.
tcp_minshall_update() is moved to tcp_output.c and is updated as we might
feed a TSO packet with a partial last segment.
This patch tremendously improves performance, as the traffic now looks
like :
IP A > B: . ack 98304 win 2783 <nop,nop,timestamp 6834277 6834685>
IP A > B: P 94209:98305(4096) ack 98304 win 2783 <nop,nop,timestamp 6834277 6834685>
IP B > A: . ack 98305 win 2768 <nop,nop,timestamp 6834686 6834277>
IP B > A: P 98304:102400(4096) ack 98305 win 2768 <nop,nop,timestamp 6834686 6834277>
IP A > B: . ack 102400 win 2783 <nop,nop,timestamp 6834279 6834686>
IP A > B: P 98305:102401(4096) ack 102400 win 2783 <nop,nop,timestamp 6834279 6834686>
IP B > A: . ack 102401 win 2768 <nop,nop,timestamp 6834687 6834279>
IP B > A: P 102400:106496(4096) ack 102401 win 2768 <nop,nop,timestamp 6834687 6834279>
IP A > B: . ack 106496 win 2783 <nop,nop,timestamp 6834280 6834687>
IP A > B: P 102401:106497(4096) ack 106496 win 2783 <nop,nop,timestamp 6834280 6834687>
IP B > A: . ack 106497 win 2768 <nop,nop,timestamp 6834688 6834280>
IP B > A: P 106496:110592(4096) ack 106497 win 2768 <nop,nop,timestamp 6834688 6834280>
Before :
lpq83:~# nstat >/dev/null;perf stat ./super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K
280774
Performance counter stats for './super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K':
205719.049006 task-clock # 9.278 CPUs utilized
8,449,968 context-switches # 0.041 M/sec
1,935,997 CPU-migrations # 0.009 M/sec
160,541 page-faults # 0.780 K/sec
548,478,722,290 cycles # 2.666 GHz [83.20%]
455,240,670,857 stalled-cycles-frontend # 83.00% frontend cycles idle [83.48%]
272,881,454,275 stalled-cycles-backend # 49.75% backend cycles idle [66.73%]
166,091,460,030 instructions # 0.30 insns per cycle
# 2.74 stalled cycles per insn [83.39%]
29,150,229,399 branches # 141.699 M/sec [83.30%]
1,943,814,026 branch-misses # 6.67% of all branches [83.32%]
22.173517844 seconds time elapsed
lpq83:~# nstat | egrep "IpOutRequests|IpExtOutOctets"
IpOutRequests 16851063 0.0
IpExtOutOctets 23878580777 0.0
After patch :
lpq83:~# nstat >/dev/null;perf stat ./super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K
280877
Performance counter stats for './super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K':
107496.071918 task-clock # 4.847 CPUs utilized
5,635,458 context-switches # 0.052 M/sec
1,374,707 CPU-migrations # 0.013 M/sec
160,920 page-faults # 0.001 M/sec
281,500,010,924 cycles # 2.619 GHz [83.28%]
228,865,069,307 stalled-cycles-frontend # 81.30% frontend cycles idle [83.38%]
142,462,742,658 stalled-cycles-backend # 50.61% backend cycles idle [66.81%]
95,227,712,566 instructions # 0.34 insns per cycle
# 2.40 stalled cycles per insn [83.43%]
16,209,868,171 branches # 150.795 M/sec [83.20%]
874,252,952 branch-misses # 5.39% of all branches [83.37%]
22.175821286 seconds time elapsed
lpq83:~# nstat | egrep "IpOutRequests|IpExtOutOctets"
IpOutRequests 11239428 0.0
IpExtOutOctets 23595191035 0.0
Indeed, the occupancy of tx skbs (IpExtOutOctets/IpOutRequests) is higher :
2099 instead of 1417, thus helping GRO to be more efficient when using FQ packet
scheduler.
Many thanks to Neal for review and ideas.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Cc: Van Jacobson <vanj@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Tested-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-14 05:51:23 +08:00
|
|
|
{
|
|
|
|
|
return partial &&
|
|
|
|
|
((nonagle & TCP_NAGLE_CORK) ||
|
|
|
|
|
(!nonagle && tp->packets_out && tcp_minshall_check(tp)));
|
|
|
|
|
}
|
tcp: refine TSO autosizing
Commit 95bd09eb2750 ("tcp: TSO packets automatic sizing") tried to
control TSO size, but did this at the wrong place (sendmsg() time)
At sendmsg() time, we might have a pessimistic view of flow rate,
and we end up building very small skbs (with 2 MSS per skb).
This is bad because :
- It sends small TSO packets even in Slow Start where rate quickly
increases.
- It tends to make socket write queue very big, increasing tcp_ack()
processing time, but also increasing memory needs, not necessarily
accounted for, as fast clones overhead is currently ignored.
- Lower GRO efficiency and more ACK packets.
Servers with a lot of small lived connections suffer from this.
Lets instead fill skbs as much as possible (64KB of payload), but split
them at xmit time, when we have a precise idea of the flow rate.
skb split is actually quite efficient.
Patch looks bigger than necessary, because TCP Small Queue decision now
has to take place after the eventual split.
As Neal suggested, introduce a new tcp_tso_autosize() helper, so that
tcp_tso_should_defer() can be synchronized on same goal.
Rename tp->xmit_size_goal_segs to tp->gso_segs, as this variable
contains number of mss that we can put in GSO packet, and is not
related to the autosizing goal anymore.
Tested:
40 ms rtt link
nstat >/dev/null
netperf -H remote -l -2000000 -- -s 1000000
nstat | egrep "IpInReceives|IpOutRequests|TcpOutSegs|IpExtOutOctets"
Before patch :
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/s
87380 2000000 2000000 0.36 44.22
IpInReceives 600 0.0
IpOutRequests 599 0.0
TcpOutSegs 1397 0.0
IpExtOutOctets 2033249 0.0
After patch :
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
87380 2000000 2000000 0.36 44.27
IpInReceives 221 0.0
IpOutRequests 232 0.0
TcpOutSegs 1397 0.0
IpExtOutOctets 2013953 0.0
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-12-08 04:22:18 +08:00
|
|
|
|
|
|
|
|
/* Return how many segs we'd like on a TSO packet,
|
|
|
|
|
* to send one TSO packet per ms
|
|
|
|
|
*/
|
2018-03-01 06:40:46 +08:00
|
|
|
static u32 tcp_tso_autosize(const struct sock *sk, unsigned int mss_now,
|
|
|
|
|
int min_tso_segs)
|
tcp: refine TSO autosizing
Commit 95bd09eb2750 ("tcp: TSO packets automatic sizing") tried to
control TSO size, but did this at the wrong place (sendmsg() time)
At sendmsg() time, we might have a pessimistic view of flow rate,
and we end up building very small skbs (with 2 MSS per skb).
This is bad because :
- It sends small TSO packets even in Slow Start where rate quickly
increases.
- It tends to make socket write queue very big, increasing tcp_ack()
processing time, but also increasing memory needs, not necessarily
accounted for, as fast clones overhead is currently ignored.
- Lower GRO efficiency and more ACK packets.
Servers with a lot of small lived connections suffer from this.
Lets instead fill skbs as much as possible (64KB of payload), but split
them at xmit time, when we have a precise idea of the flow rate.
skb split is actually quite efficient.
Patch looks bigger than necessary, because TCP Small Queue decision now
has to take place after the eventual split.
As Neal suggested, introduce a new tcp_tso_autosize() helper, so that
tcp_tso_should_defer() can be synchronized on same goal.
Rename tp->xmit_size_goal_segs to tp->gso_segs, as this variable
contains number of mss that we can put in GSO packet, and is not
related to the autosizing goal anymore.
Tested:
40 ms rtt link
nstat >/dev/null
netperf -H remote -l -2000000 -- -s 1000000
nstat | egrep "IpInReceives|IpOutRequests|TcpOutSegs|IpExtOutOctets"
Before patch :
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/s
87380 2000000 2000000 0.36 44.22
IpInReceives 600 0.0
IpOutRequests 599 0.0
TcpOutSegs 1397 0.0
IpExtOutOctets 2033249 0.0
After patch :
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
87380 2000000 2000000 0.36 44.27
IpInReceives 221 0.0
IpOutRequests 232 0.0
TcpOutSegs 1397 0.0
IpExtOutOctets 2013953 0.0
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-12-08 04:22:18 +08:00
|
|
|
{
|
|
|
|
|
u32 bytes, segs;
|
|
|
|
|
|
net: extend sk_pacing_rate to unsigned long
sk_pacing_rate has beed introduced as a u32 field in 2013,
effectively limiting per flow pacing to 34Gbit.
We believe it is time to allow TCP to pace high speed flows
on 64bit hosts, as we now can reach 100Gbit on one TCP flow.
This patch adds no cost for 32bit kernels.
The tcpi_pacing_rate and tcpi_max_pacing_rate were already
exported as 64bit, so iproute2/ss command require no changes.
Unfortunately the SO_MAX_PACING_RATE socket option will stay
32bit and we will need to add a new option to let applications
control high pacing rates.
State Recv-Q Send-Q Local Address:Port Peer Address:Port
ESTAB 0 1787144 10.246.9.76:49992 10.246.9.77:36741
timer:(on,003ms,0) ino:91863 sk:2 <->
skmem:(r0,rb540000,t66440,tb2363904,f605944,w1822984,o0,bl0,d0)
ts sack bbr wscale:8,8 rto:201 rtt:0.057/0.006 mss:1448
rcvmss:536 advmss:1448
cwnd:138 ssthresh:178 bytes_acked:256699822585 segs_out:177279177
segs_in:3916318 data_segs_out:177279175
bbr:(bw:31276.8Mbps,mrtt:0,pacing_gain:1.25,cwnd_gain:2)
send 28045.5Mbps lastrcv:73333
pacing_rate 38705.0Mbps delivery_rate 22997.6Mbps
busy:73333ms unacked:135 retrans:0/157 rcv_space:14480
notsent:2085120 minrtt:0.013
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-10-16 00:37:53 +08:00
|
|
|
bytes = min_t(unsigned long,
|
2019-12-17 10:51:03 +08:00
|
|
|
sk->sk_pacing_rate >> READ_ONCE(sk->sk_pacing_shift),
|
net: extend sk_pacing_rate to unsigned long
sk_pacing_rate has beed introduced as a u32 field in 2013,
effectively limiting per flow pacing to 34Gbit.
We believe it is time to allow TCP to pace high speed flows
on 64bit hosts, as we now can reach 100Gbit on one TCP flow.
This patch adds no cost for 32bit kernels.
The tcpi_pacing_rate and tcpi_max_pacing_rate were already
exported as 64bit, so iproute2/ss command require no changes.
Unfortunately the SO_MAX_PACING_RATE socket option will stay
32bit and we will need to add a new option to let applications
control high pacing rates.
State Recv-Q Send-Q Local Address:Port Peer Address:Port
ESTAB 0 1787144 10.246.9.76:49992 10.246.9.77:36741
timer:(on,003ms,0) ino:91863 sk:2 <->
skmem:(r0,rb540000,t66440,tb2363904,f605944,w1822984,o0,bl0,d0)
ts sack bbr wscale:8,8 rto:201 rtt:0.057/0.006 mss:1448
rcvmss:536 advmss:1448
cwnd:138 ssthresh:178 bytes_acked:256699822585 segs_out:177279177
segs_in:3916318 data_segs_out:177279175
bbr:(bw:31276.8Mbps,mrtt:0,pacing_gain:1.25,cwnd_gain:2)
send 28045.5Mbps lastrcv:73333
pacing_rate 38705.0Mbps delivery_rate 22997.6Mbps
busy:73333ms unacked:135 retrans:0/157 rcv_space:14480
notsent:2085120 minrtt:0.013
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-10-16 00:37:53 +08:00
|
|
|
sk->sk_gso_max_size - 1 - MAX_TCP_HEADER);
|
tcp: refine TSO autosizing
Commit 95bd09eb2750 ("tcp: TSO packets automatic sizing") tried to
control TSO size, but did this at the wrong place (sendmsg() time)
At sendmsg() time, we might have a pessimistic view of flow rate,
and we end up building very small skbs (with 2 MSS per skb).
This is bad because :
- It sends small TSO packets even in Slow Start where rate quickly
increases.
- It tends to make socket write queue very big, increasing tcp_ack()
processing time, but also increasing memory needs, not necessarily
accounted for, as fast clones overhead is currently ignored.
- Lower GRO efficiency and more ACK packets.
Servers with a lot of small lived connections suffer from this.
Lets instead fill skbs as much as possible (64KB of payload), but split
them at xmit time, when we have a precise idea of the flow rate.
skb split is actually quite efficient.
Patch looks bigger than necessary, because TCP Small Queue decision now
has to take place after the eventual split.
As Neal suggested, introduce a new tcp_tso_autosize() helper, so that
tcp_tso_should_defer() can be synchronized on same goal.
Rename tp->xmit_size_goal_segs to tp->gso_segs, as this variable
contains number of mss that we can put in GSO packet, and is not
related to the autosizing goal anymore.
Tested:
40 ms rtt link
nstat >/dev/null
netperf -H remote -l -2000000 -- -s 1000000
nstat | egrep "IpInReceives|IpOutRequests|TcpOutSegs|IpExtOutOctets"
Before patch :
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/s
87380 2000000 2000000 0.36 44.22
IpInReceives 600 0.0
IpOutRequests 599 0.0
TcpOutSegs 1397 0.0
IpExtOutOctets 2033249 0.0
After patch :
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
87380 2000000 2000000 0.36 44.27
IpInReceives 221 0.0
IpOutRequests 232 0.0
TcpOutSegs 1397 0.0
IpExtOutOctets 2013953 0.0
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-12-08 04:22:18 +08:00
|
|
|
|
|
|
|
|
/* Goal is to send at least one packet per ms,
|
|
|
|
|
* not one big TSO packet every 100 ms.
|
|
|
|
|
* This preserves ACK clocking and is consistent
|
|
|
|
|
* with tcp_tso_should_defer() heuristic.
|
|
|
|
|
*/
|
2016-09-20 11:39:18 +08:00
|
|
|
segs = max_t(u32, bytes / mss_now, min_tso_segs);
|
tcp: refine TSO autosizing
Commit 95bd09eb2750 ("tcp: TSO packets automatic sizing") tried to
control TSO size, but did this at the wrong place (sendmsg() time)
At sendmsg() time, we might have a pessimistic view of flow rate,
and we end up building very small skbs (with 2 MSS per skb).
This is bad because :
- It sends small TSO packets even in Slow Start where rate quickly
increases.
- It tends to make socket write queue very big, increasing tcp_ack()
processing time, but also increasing memory needs, not necessarily
accounted for, as fast clones overhead is currently ignored.
- Lower GRO efficiency and more ACK packets.
Servers with a lot of small lived connections suffer from this.
Lets instead fill skbs as much as possible (64KB of payload), but split
them at xmit time, when we have a precise idea of the flow rate.
skb split is actually quite efficient.
Patch looks bigger than necessary, because TCP Small Queue decision now
has to take place after the eventual split.
As Neal suggested, introduce a new tcp_tso_autosize() helper, so that
tcp_tso_should_defer() can be synchronized on same goal.
Rename tp->xmit_size_goal_segs to tp->gso_segs, as this variable
contains number of mss that we can put in GSO packet, and is not
related to the autosizing goal anymore.
Tested:
40 ms rtt link
nstat >/dev/null
netperf -H remote -l -2000000 -- -s 1000000
nstat | egrep "IpInReceives|IpOutRequests|TcpOutSegs|IpExtOutOctets"
Before patch :
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/s
87380 2000000 2000000 0.36 44.22
IpInReceives 600 0.0
IpOutRequests 599 0.0
TcpOutSegs 1397 0.0
IpExtOutOctets 2033249 0.0
After patch :
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
87380 2000000 2000000 0.36 44.27
IpInReceives 221 0.0
IpOutRequests 232 0.0
TcpOutSegs 1397 0.0
IpExtOutOctets 2013953 0.0
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-12-08 04:22:18 +08:00
|
|
|
|
2018-02-21 22:43:03 +08:00
|
|
|
return segs;
|
tcp: refine TSO autosizing
Commit 95bd09eb2750 ("tcp: TSO packets automatic sizing") tried to
control TSO size, but did this at the wrong place (sendmsg() time)
At sendmsg() time, we might have a pessimistic view of flow rate,
and we end up building very small skbs (with 2 MSS per skb).
This is bad because :
- It sends small TSO packets even in Slow Start where rate quickly
increases.
- It tends to make socket write queue very big, increasing tcp_ack()
processing time, but also increasing memory needs, not necessarily
accounted for, as fast clones overhead is currently ignored.
- Lower GRO efficiency and more ACK packets.
Servers with a lot of small lived connections suffer from this.
Lets instead fill skbs as much as possible (64KB of payload), but split
them at xmit time, when we have a precise idea of the flow rate.
skb split is actually quite efficient.
Patch looks bigger than necessary, because TCP Small Queue decision now
has to take place after the eventual split.
As Neal suggested, introduce a new tcp_tso_autosize() helper, so that
tcp_tso_should_defer() can be synchronized on same goal.
Rename tp->xmit_size_goal_segs to tp->gso_segs, as this variable
contains number of mss that we can put in GSO packet, and is not
related to the autosizing goal anymore.
Tested:
40 ms rtt link
nstat >/dev/null
netperf -H remote -l -2000000 -- -s 1000000
nstat | egrep "IpInReceives|IpOutRequests|TcpOutSegs|IpExtOutOctets"
Before patch :
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/s
87380 2000000 2000000 0.36 44.22
IpInReceives 600 0.0
IpOutRequests 599 0.0
TcpOutSegs 1397 0.0
IpExtOutOctets 2033249 0.0
After patch :
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
87380 2000000 2000000 0.36 44.27
IpInReceives 221 0.0
IpOutRequests 232 0.0
TcpOutSegs 1397 0.0
IpExtOutOctets 2013953 0.0
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-12-08 04:22:18 +08:00
|
|
|
}
|
|
|
|
|
|
2016-09-20 11:39:17 +08:00
|
|
|
/* Return the number of segments we want in the skb we are transmitting.
|
|
|
|
|
* See if congestion control module wants to decide; otherwise, autosize.
|
|
|
|
|
*/
|
|
|
|
|
static u32 tcp_tso_segs(struct sock *sk, unsigned int mss_now)
|
|
|
|
|
{
|
|
|
|
|
const struct tcp_congestion_ops *ca_ops = inet_csk(sk)->icsk_ca_ops;
|
2018-03-01 06:40:46 +08:00
|
|
|
u32 min_tso, tso_segs;
|
2016-09-20 11:39:17 +08:00
|
|
|
|
2018-03-01 06:40:46 +08:00
|
|
|
min_tso = ca_ops->min_tso_segs ?
|
|
|
|
|
ca_ops->min_tso_segs(sk) :
|
|
|
|
|
sock_net(sk)->ipv4.sysctl_tcp_min_tso_segs;
|
|
|
|
|
|
|
|
|
|
tso_segs = tcp_tso_autosize(sk, mss_now, min_tso);
|
2018-02-21 22:43:03 +08:00
|
|
|
return min_t(u32, tso_segs, sk->sk_gso_max_segs);
|
2016-09-20 11:39:17 +08:00
|
|
|
}
|
|
|
|
|
|
tcp: refine TSO splits
While investigating performance problems on small RPC workloads,
I noticed linux TCP stack was always splitting the last TSO skb
into two parts (skbs). One being a multiple of MSS, and a small one
with the Push flag. This split is done even if TCP_NODELAY is set,
or if no small packet is in flight.
Example with request/response of 4K/4K
IP A > B: . ack 68432 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: . 65537:68433(2896) ack 69632 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: P 68433:69633(1200) ack 69632 win 2783 <nop,nop,timestamp 6524593 6525001>
IP B > A: . ack 68433 win 2768 <nop,nop,timestamp 6525001 6524593>
IP B > A: . 69632:72528(2896) ack 69633 win 2768 <nop,nop,timestamp 6525001 6524593>
IP B > A: P 72528:73728(1200) ack 69633 win 2768 <nop,nop,timestamp 6525001 6524593>
IP A > B: . ack 72528 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: . 69633:72529(2896) ack 73728 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: P 72529:73729(1200) ack 73728 win 2783 <nop,nop,timestamp 6524593 6525001>
We can avoid this split by including the Nagle tests at the right place.
Note : If some NIC had trouble sending TSO packets with a partial
last segment, we would have hit the problem in GRO/forwarding workload already.
tcp_minshall_update() is moved to tcp_output.c and is updated as we might
feed a TSO packet with a partial last segment.
This patch tremendously improves performance, as the traffic now looks
like :
IP A > B: . ack 98304 win 2783 <nop,nop,timestamp 6834277 6834685>
IP A > B: P 94209:98305(4096) ack 98304 win 2783 <nop,nop,timestamp 6834277 6834685>
IP B > A: . ack 98305 win 2768 <nop,nop,timestamp 6834686 6834277>
IP B > A: P 98304:102400(4096) ack 98305 win 2768 <nop,nop,timestamp 6834686 6834277>
IP A > B: . ack 102400 win 2783 <nop,nop,timestamp 6834279 6834686>
IP A > B: P 98305:102401(4096) ack 102400 win 2783 <nop,nop,timestamp 6834279 6834686>
IP B > A: . ack 102401 win 2768 <nop,nop,timestamp 6834687 6834279>
IP B > A: P 102400:106496(4096) ack 102401 win 2768 <nop,nop,timestamp 6834687 6834279>
IP A > B: . ack 106496 win 2783 <nop,nop,timestamp 6834280 6834687>
IP A > B: P 102401:106497(4096) ack 106496 win 2783 <nop,nop,timestamp 6834280 6834687>
IP B > A: . ack 106497 win 2768 <nop,nop,timestamp 6834688 6834280>
IP B > A: P 106496:110592(4096) ack 106497 win 2768 <nop,nop,timestamp 6834688 6834280>
Before :
lpq83:~# nstat >/dev/null;perf stat ./super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K
280774
Performance counter stats for './super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K':
205719.049006 task-clock # 9.278 CPUs utilized
8,449,968 context-switches # 0.041 M/sec
1,935,997 CPU-migrations # 0.009 M/sec
160,541 page-faults # 0.780 K/sec
548,478,722,290 cycles # 2.666 GHz [83.20%]
455,240,670,857 stalled-cycles-frontend # 83.00% frontend cycles idle [83.48%]
272,881,454,275 stalled-cycles-backend # 49.75% backend cycles idle [66.73%]
166,091,460,030 instructions # 0.30 insns per cycle
# 2.74 stalled cycles per insn [83.39%]
29,150,229,399 branches # 141.699 M/sec [83.30%]
1,943,814,026 branch-misses # 6.67% of all branches [83.32%]
22.173517844 seconds time elapsed
lpq83:~# nstat | egrep "IpOutRequests|IpExtOutOctets"
IpOutRequests 16851063 0.0
IpExtOutOctets 23878580777 0.0
After patch :
lpq83:~# nstat >/dev/null;perf stat ./super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K
280877
Performance counter stats for './super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K':
107496.071918 task-clock # 4.847 CPUs utilized
5,635,458 context-switches # 0.052 M/sec
1,374,707 CPU-migrations # 0.013 M/sec
160,920 page-faults # 0.001 M/sec
281,500,010,924 cycles # 2.619 GHz [83.28%]
228,865,069,307 stalled-cycles-frontend # 81.30% frontend cycles idle [83.38%]
142,462,742,658 stalled-cycles-backend # 50.61% backend cycles idle [66.81%]
95,227,712,566 instructions # 0.34 insns per cycle
# 2.40 stalled cycles per insn [83.43%]
16,209,868,171 branches # 150.795 M/sec [83.20%]
874,252,952 branch-misses # 5.39% of all branches [83.37%]
22.175821286 seconds time elapsed
lpq83:~# nstat | egrep "IpOutRequests|IpExtOutOctets"
IpOutRequests 11239428 0.0
IpExtOutOctets 23595191035 0.0
Indeed, the occupancy of tx skbs (IpExtOutOctets/IpOutRequests) is higher :
2099 instead of 1417, thus helping GRO to be more efficient when using FQ packet
scheduler.
Many thanks to Neal for review and ideas.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Cc: Van Jacobson <vanj@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Tested-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-14 05:51:23 +08:00
|
|
|
/* Returns the portion of skb which can be sent right away */
|
|
|
|
|
static unsigned int tcp_mss_split_point(const struct sock *sk,
|
|
|
|
|
const struct sk_buff *skb,
|
|
|
|
|
unsigned int mss_now,
|
|
|
|
|
unsigned int max_segs,
|
|
|
|
|
int nonagle)
|
2005-07-06 06:24:38 +08:00
|
|
|
{
|
2011-10-21 17:22:42 +08:00
|
|
|
const struct tcp_sock *tp = tcp_sk(sk);
|
tcp: refine TSO splits
While investigating performance problems on small RPC workloads,
I noticed linux TCP stack was always splitting the last TSO skb
into two parts (skbs). One being a multiple of MSS, and a small one
with the Push flag. This split is done even if TCP_NODELAY is set,
or if no small packet is in flight.
Example with request/response of 4K/4K
IP A > B: . ack 68432 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: . 65537:68433(2896) ack 69632 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: P 68433:69633(1200) ack 69632 win 2783 <nop,nop,timestamp 6524593 6525001>
IP B > A: . ack 68433 win 2768 <nop,nop,timestamp 6525001 6524593>
IP B > A: . 69632:72528(2896) ack 69633 win 2768 <nop,nop,timestamp 6525001 6524593>
IP B > A: P 72528:73728(1200) ack 69633 win 2768 <nop,nop,timestamp 6525001 6524593>
IP A > B: . ack 72528 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: . 69633:72529(2896) ack 73728 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: P 72529:73729(1200) ack 73728 win 2783 <nop,nop,timestamp 6524593 6525001>
We can avoid this split by including the Nagle tests at the right place.
Note : If some NIC had trouble sending TSO packets with a partial
last segment, we would have hit the problem in GRO/forwarding workload already.
tcp_minshall_update() is moved to tcp_output.c and is updated as we might
feed a TSO packet with a partial last segment.
This patch tremendously improves performance, as the traffic now looks
like :
IP A > B: . ack 98304 win 2783 <nop,nop,timestamp 6834277 6834685>
IP A > B: P 94209:98305(4096) ack 98304 win 2783 <nop,nop,timestamp 6834277 6834685>
IP B > A: . ack 98305 win 2768 <nop,nop,timestamp 6834686 6834277>
IP B > A: P 98304:102400(4096) ack 98305 win 2768 <nop,nop,timestamp 6834686 6834277>
IP A > B: . ack 102400 win 2783 <nop,nop,timestamp 6834279 6834686>
IP A > B: P 98305:102401(4096) ack 102400 win 2783 <nop,nop,timestamp 6834279 6834686>
IP B > A: . ack 102401 win 2768 <nop,nop,timestamp 6834687 6834279>
IP B > A: P 102400:106496(4096) ack 102401 win 2768 <nop,nop,timestamp 6834687 6834279>
IP A > B: . ack 106496 win 2783 <nop,nop,timestamp 6834280 6834687>
IP A > B: P 102401:106497(4096) ack 106496 win 2783 <nop,nop,timestamp 6834280 6834687>
IP B > A: . ack 106497 win 2768 <nop,nop,timestamp 6834688 6834280>
IP B > A: P 106496:110592(4096) ack 106497 win 2768 <nop,nop,timestamp 6834688 6834280>
Before :
lpq83:~# nstat >/dev/null;perf stat ./super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K
280774
Performance counter stats for './super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K':
205719.049006 task-clock # 9.278 CPUs utilized
8,449,968 context-switches # 0.041 M/sec
1,935,997 CPU-migrations # 0.009 M/sec
160,541 page-faults # 0.780 K/sec
548,478,722,290 cycles # 2.666 GHz [83.20%]
455,240,670,857 stalled-cycles-frontend # 83.00% frontend cycles idle [83.48%]
272,881,454,275 stalled-cycles-backend # 49.75% backend cycles idle [66.73%]
166,091,460,030 instructions # 0.30 insns per cycle
# 2.74 stalled cycles per insn [83.39%]
29,150,229,399 branches # 141.699 M/sec [83.30%]
1,943,814,026 branch-misses # 6.67% of all branches [83.32%]
22.173517844 seconds time elapsed
lpq83:~# nstat | egrep "IpOutRequests|IpExtOutOctets"
IpOutRequests 16851063 0.0
IpExtOutOctets 23878580777 0.0
After patch :
lpq83:~# nstat >/dev/null;perf stat ./super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K
280877
Performance counter stats for './super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K':
107496.071918 task-clock # 4.847 CPUs utilized
5,635,458 context-switches # 0.052 M/sec
1,374,707 CPU-migrations # 0.013 M/sec
160,920 page-faults # 0.001 M/sec
281,500,010,924 cycles # 2.619 GHz [83.28%]
228,865,069,307 stalled-cycles-frontend # 81.30% frontend cycles idle [83.38%]
142,462,742,658 stalled-cycles-backend # 50.61% backend cycles idle [66.81%]
95,227,712,566 instructions # 0.34 insns per cycle
# 2.40 stalled cycles per insn [83.43%]
16,209,868,171 branches # 150.795 M/sec [83.20%]
874,252,952 branch-misses # 5.39% of all branches [83.37%]
22.175821286 seconds time elapsed
lpq83:~# nstat | egrep "IpOutRequests|IpExtOutOctets"
IpOutRequests 11239428 0.0
IpExtOutOctets 23595191035 0.0
Indeed, the occupancy of tx skbs (IpExtOutOctets/IpOutRequests) is higher :
2099 instead of 1417, thus helping GRO to be more efficient when using FQ packet
scheduler.
Many thanks to Neal for review and ideas.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Cc: Van Jacobson <vanj@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Tested-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-14 05:51:23 +08:00
|
|
|
u32 partial, needed, window, max_len;
|
2005-07-06 06:24:38 +08:00
|
|
|
|
2007-12-31 20:48:41 +08:00
|
|
|
window = tcp_wnd_end(tp) - TCP_SKB_CB(skb)->seq;
|
2012-07-31 00:11:42 +08:00
|
|
|
max_len = mss_now * max_segs;
|
2007-12-25 13:33:45 +08:00
|
|
|
|
2012-07-31 00:11:42 +08:00
|
|
|
if (likely(max_len <= window && skb != tcp_write_queue_tail(sk)))
|
|
|
|
|
return max_len;
|
2007-12-25 13:33:45 +08:00
|
|
|
|
2008-03-12 08:55:27 +08:00
|
|
|
needed = min(skb->len, window);
|
|
|
|
|
|
2012-07-31 00:11:42 +08:00
|
|
|
if (max_len <= needed)
|
|
|
|
|
return max_len;
|
2007-12-25 13:33:45 +08:00
|
|
|
|
tcp: refine TSO splits
While investigating performance problems on small RPC workloads,
I noticed linux TCP stack was always splitting the last TSO skb
into two parts (skbs). One being a multiple of MSS, and a small one
with the Push flag. This split is done even if TCP_NODELAY is set,
or if no small packet is in flight.
Example with request/response of 4K/4K
IP A > B: . ack 68432 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: . 65537:68433(2896) ack 69632 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: P 68433:69633(1200) ack 69632 win 2783 <nop,nop,timestamp 6524593 6525001>
IP B > A: . ack 68433 win 2768 <nop,nop,timestamp 6525001 6524593>
IP B > A: . 69632:72528(2896) ack 69633 win 2768 <nop,nop,timestamp 6525001 6524593>
IP B > A: P 72528:73728(1200) ack 69633 win 2768 <nop,nop,timestamp 6525001 6524593>
IP A > B: . ack 72528 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: . 69633:72529(2896) ack 73728 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: P 72529:73729(1200) ack 73728 win 2783 <nop,nop,timestamp 6524593 6525001>
We can avoid this split by including the Nagle tests at the right place.
Note : If some NIC had trouble sending TSO packets with a partial
last segment, we would have hit the problem in GRO/forwarding workload already.
tcp_minshall_update() is moved to tcp_output.c and is updated as we might
feed a TSO packet with a partial last segment.
This patch tremendously improves performance, as the traffic now looks
like :
IP A > B: . ack 98304 win 2783 <nop,nop,timestamp 6834277 6834685>
IP A > B: P 94209:98305(4096) ack 98304 win 2783 <nop,nop,timestamp 6834277 6834685>
IP B > A: . ack 98305 win 2768 <nop,nop,timestamp 6834686 6834277>
IP B > A: P 98304:102400(4096) ack 98305 win 2768 <nop,nop,timestamp 6834686 6834277>
IP A > B: . ack 102400 win 2783 <nop,nop,timestamp 6834279 6834686>
IP A > B: P 98305:102401(4096) ack 102400 win 2783 <nop,nop,timestamp 6834279 6834686>
IP B > A: . ack 102401 win 2768 <nop,nop,timestamp 6834687 6834279>
IP B > A: P 102400:106496(4096) ack 102401 win 2768 <nop,nop,timestamp 6834687 6834279>
IP A > B: . ack 106496 win 2783 <nop,nop,timestamp 6834280 6834687>
IP A > B: P 102401:106497(4096) ack 106496 win 2783 <nop,nop,timestamp 6834280 6834687>
IP B > A: . ack 106497 win 2768 <nop,nop,timestamp 6834688 6834280>
IP B > A: P 106496:110592(4096) ack 106497 win 2768 <nop,nop,timestamp 6834688 6834280>
Before :
lpq83:~# nstat >/dev/null;perf stat ./super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K
280774
Performance counter stats for './super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K':
205719.049006 task-clock # 9.278 CPUs utilized
8,449,968 context-switches # 0.041 M/sec
1,935,997 CPU-migrations # 0.009 M/sec
160,541 page-faults # 0.780 K/sec
548,478,722,290 cycles # 2.666 GHz [83.20%]
455,240,670,857 stalled-cycles-frontend # 83.00% frontend cycles idle [83.48%]
272,881,454,275 stalled-cycles-backend # 49.75% backend cycles idle [66.73%]
166,091,460,030 instructions # 0.30 insns per cycle
# 2.74 stalled cycles per insn [83.39%]
29,150,229,399 branches # 141.699 M/sec [83.30%]
1,943,814,026 branch-misses # 6.67% of all branches [83.32%]
22.173517844 seconds time elapsed
lpq83:~# nstat | egrep "IpOutRequests|IpExtOutOctets"
IpOutRequests 16851063 0.0
IpExtOutOctets 23878580777 0.0
After patch :
lpq83:~# nstat >/dev/null;perf stat ./super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K
280877
Performance counter stats for './super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K':
107496.071918 task-clock # 4.847 CPUs utilized
5,635,458 context-switches # 0.052 M/sec
1,374,707 CPU-migrations # 0.013 M/sec
160,920 page-faults # 0.001 M/sec
281,500,010,924 cycles # 2.619 GHz [83.28%]
228,865,069,307 stalled-cycles-frontend # 81.30% frontend cycles idle [83.38%]
142,462,742,658 stalled-cycles-backend # 50.61% backend cycles idle [66.81%]
95,227,712,566 instructions # 0.34 insns per cycle
# 2.40 stalled cycles per insn [83.43%]
16,209,868,171 branches # 150.795 M/sec [83.20%]
874,252,952 branch-misses # 5.39% of all branches [83.37%]
22.175821286 seconds time elapsed
lpq83:~# nstat | egrep "IpOutRequests|IpExtOutOctets"
IpOutRequests 11239428 0.0
IpExtOutOctets 23595191035 0.0
Indeed, the occupancy of tx skbs (IpExtOutOctets/IpOutRequests) is higher :
2099 instead of 1417, thus helping GRO to be more efficient when using FQ packet
scheduler.
Many thanks to Neal for review and ideas.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Cc: Van Jacobson <vanj@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Tested-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-14 05:51:23 +08:00
|
|
|
partial = needed % mss_now;
|
|
|
|
|
/* If last segment is not a full MSS, check if Nagle rules allow us
|
|
|
|
|
* to include this last segment in this skb.
|
|
|
|
|
* Otherwise, we'll split the skb at last MSS boundary
|
|
|
|
|
*/
|
2014-03-24 14:49:34 +08:00
|
|
|
if (tcp_nagle_check(partial != 0, tp, nonagle))
|
tcp: refine TSO splits
While investigating performance problems on small RPC workloads,
I noticed linux TCP stack was always splitting the last TSO skb
into two parts (skbs). One being a multiple of MSS, and a small one
with the Push flag. This split is done even if TCP_NODELAY is set,
or if no small packet is in flight.
Example with request/response of 4K/4K
IP A > B: . ack 68432 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: . 65537:68433(2896) ack 69632 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: P 68433:69633(1200) ack 69632 win 2783 <nop,nop,timestamp 6524593 6525001>
IP B > A: . ack 68433 win 2768 <nop,nop,timestamp 6525001 6524593>
IP B > A: . 69632:72528(2896) ack 69633 win 2768 <nop,nop,timestamp 6525001 6524593>
IP B > A: P 72528:73728(1200) ack 69633 win 2768 <nop,nop,timestamp 6525001 6524593>
IP A > B: . ack 72528 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: . 69633:72529(2896) ack 73728 win 2783 <nop,nop,timestamp 6524593 6525001>
IP A > B: P 72529:73729(1200) ack 73728 win 2783 <nop,nop,timestamp 6524593 6525001>
We can avoid this split by including the Nagle tests at the right place.
Note : If some NIC had trouble sending TSO packets with a partial
last segment, we would have hit the problem in GRO/forwarding workload already.
tcp_minshall_update() is moved to tcp_output.c and is updated as we might
feed a TSO packet with a partial last segment.
This patch tremendously improves performance, as the traffic now looks
like :
IP A > B: . ack 98304 win 2783 <nop,nop,timestamp 6834277 6834685>
IP A > B: P 94209:98305(4096) ack 98304 win 2783 <nop,nop,timestamp 6834277 6834685>
IP B > A: . ack 98305 win 2768 <nop,nop,timestamp 6834686 6834277>
IP B > A: P 98304:102400(4096) ack 98305 win 2768 <nop,nop,timestamp 6834686 6834277>
IP A > B: . ack 102400 win 2783 <nop,nop,timestamp 6834279 6834686>
IP A > B: P 98305:102401(4096) ack 102400 win 2783 <nop,nop,timestamp 6834279 6834686>
IP B > A: . ack 102401 win 2768 <nop,nop,timestamp 6834687 6834279>
IP B > A: P 102400:106496(4096) ack 102401 win 2768 <nop,nop,timestamp 6834687 6834279>
IP A > B: . ack 106496 win 2783 <nop,nop,timestamp 6834280 6834687>
IP A > B: P 102401:106497(4096) ack 106496 win 2783 <nop,nop,timestamp 6834280 6834687>
IP B > A: . ack 106497 win 2768 <nop,nop,timestamp 6834688 6834280>
IP B > A: P 106496:110592(4096) ack 106497 win 2768 <nop,nop,timestamp 6834688 6834280>
Before :
lpq83:~# nstat >/dev/null;perf stat ./super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K
280774
Performance counter stats for './super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K':
205719.049006 task-clock # 9.278 CPUs utilized
8,449,968 context-switches # 0.041 M/sec
1,935,997 CPU-migrations # 0.009 M/sec
160,541 page-faults # 0.780 K/sec
548,478,722,290 cycles # 2.666 GHz [83.20%]
455,240,670,857 stalled-cycles-frontend # 83.00% frontend cycles idle [83.48%]
272,881,454,275 stalled-cycles-backend # 49.75% backend cycles idle [66.73%]
166,091,460,030 instructions # 0.30 insns per cycle
# 2.74 stalled cycles per insn [83.39%]
29,150,229,399 branches # 141.699 M/sec [83.30%]
1,943,814,026 branch-misses # 6.67% of all branches [83.32%]
22.173517844 seconds time elapsed
lpq83:~# nstat | egrep "IpOutRequests|IpExtOutOctets"
IpOutRequests 16851063 0.0
IpExtOutOctets 23878580777 0.0
After patch :
lpq83:~# nstat >/dev/null;perf stat ./super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K
280877
Performance counter stats for './super_netperf 200 -t TCP_RR -H lpq84 -l 20 -- -r 4K,4K':
107496.071918 task-clock # 4.847 CPUs utilized
5,635,458 context-switches # 0.052 M/sec
1,374,707 CPU-migrations # 0.013 M/sec
160,920 page-faults # 0.001 M/sec
281,500,010,924 cycles # 2.619 GHz [83.28%]
228,865,069,307 stalled-cycles-frontend # 81.30% frontend cycles idle [83.38%]
142,462,742,658 stalled-cycles-backend # 50.61% backend cycles idle [66.81%]
95,227,712,566 instructions # 0.34 insns per cycle
# 2.40 stalled cycles per insn [83.43%]
16,209,868,171 branches # 150.795 M/sec [83.20%]
874,252,952 branch-misses # 5.39% of all branches [83.37%]
22.175821286 seconds time elapsed
lpq83:~# nstat | egrep "IpOutRequests|IpExtOutOctets"
IpOutRequests 11239428 0.0
IpExtOutOctets 23595191035 0.0
Indeed, the occupancy of tx skbs (IpExtOutOctets/IpOutRequests) is higher :
2099 instead of 1417, thus helping GRO to be more efficient when using FQ packet
scheduler.
Many thanks to Neal for review and ideas.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Nandita Dukkipati <nanditad@google.com>
Cc: Van Jacobson <vanj@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Tested-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-14 05:51:23 +08:00
|
|
|
return needed - partial;
|
|
|
|
|
|
|
|
|
|
return needed;
|
2005-07-06 06:24:38 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Can at least one segment of SKB be sent right now, according to the
|
|
|
|
|
* congestion window rules? If so, return how many segments are allowed.
|
|
|
|
|
*/
|
2011-10-21 17:22:42 +08:00
|
|
|
static inline unsigned int tcp_cwnd_test(const struct tcp_sock *tp,
|
|
|
|
|
const struct sk_buff *skb)
|
2005-07-06 06:24:38 +08:00
|
|
|
{
|
tcp: limit GSO packets to half cwnd
In DC world, GSO packets initially cooked by tcp_sendmsg() are usually
big, as sk_pacing_rate is high.
When network is congested, cwnd can be smaller than the GSO packets
found in socket write queue. tcp_write_xmit() splits GSO packets
using the available cwnd, and we end up sending a single GSO packet,
consuming all available cwnd.
With GRO aggregation on the receiver, we might handle a single GRO
packet, sending back a single ACK.
1) This single ACK might be lost
TLP or RTO are forced to attempt a retransmit.
2) This ACK releases a full cwnd, sender sends another big GSO packet,
in a ping pong mode.
This behavior does not fill the pipes in the best way, because of
scheduling artifacts.
Make sure we always have at least two GSO packets in flight.
This allows us to safely increase GRO efficiency without risking
spurious retransmits.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-11-14 01:45:22 +08:00
|
|
|
u32 in_flight, cwnd, halfcwnd;
|
2005-07-06 06:24:38 +08:00
|
|
|
|
|
|
|
|
/* Don't be strict about the congestion window for the final FIN. */
|
2011-09-28 01:25:05 +08:00
|
|
|
if ((TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN) &&
|
|
|
|
|
tcp_skb_pcount(skb) == 1)
|
2005-07-06 06:24:38 +08:00
|
|
|
return 1;
|
|
|
|
|
|
|
|
|
|
in_flight = tcp_packets_in_flight(tp);
|
|
|
|
|
cwnd = tp->snd_cwnd;
|
tcp: limit GSO packets to half cwnd
In DC world, GSO packets initially cooked by tcp_sendmsg() are usually
big, as sk_pacing_rate is high.
When network is congested, cwnd can be smaller than the GSO packets
found in socket write queue. tcp_write_xmit() splits GSO packets
using the available cwnd, and we end up sending a single GSO packet,
consuming all available cwnd.
With GRO aggregation on the receiver, we might handle a single GRO
packet, sending back a single ACK.
1) This single ACK might be lost
TLP or RTO are forced to attempt a retransmit.
2) This ACK releases a full cwnd, sender sends another big GSO packet,
in a ping pong mode.
This behavior does not fill the pipes in the best way, because of
scheduling artifacts.
Make sure we always have at least two GSO packets in flight.
This allows us to safely increase GRO efficiency without risking
spurious retransmits.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-11-14 01:45:22 +08:00
|
|
|
if (in_flight >= cwnd)
|
|
|
|
|
return 0;
|
2005-07-06 06:24:38 +08:00
|
|
|
|
tcp: limit GSO packets to half cwnd
In DC world, GSO packets initially cooked by tcp_sendmsg() are usually
big, as sk_pacing_rate is high.
When network is congested, cwnd can be smaller than the GSO packets
found in socket write queue. tcp_write_xmit() splits GSO packets
using the available cwnd, and we end up sending a single GSO packet,
consuming all available cwnd.
With GRO aggregation on the receiver, we might handle a single GRO
packet, sending back a single ACK.
1) This single ACK might be lost
TLP or RTO are forced to attempt a retransmit.
2) This ACK releases a full cwnd, sender sends another big GSO packet,
in a ping pong mode.
This behavior does not fill the pipes in the best way, because of
scheduling artifacts.
Make sure we always have at least two GSO packets in flight.
This allows us to safely increase GRO efficiency without risking
spurious retransmits.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-11-14 01:45:22 +08:00
|
|
|
/* For better scheduling, ensure we have at least
|
|
|
|
|
* 2 GSO packets in flight.
|
|
|
|
|
*/
|
|
|
|
|
halfcwnd = max(cwnd >> 1, 1U);
|
|
|
|
|
return min(halfcwnd, cwnd - in_flight);
|
2005-07-06 06:24:38 +08:00
|
|
|
}
|
|
|
|
|
|
tree-wide: fix comment/printk typos
"gadget", "through", "command", "maintain", "maintain", "controller", "address",
"between", "initiali[zs]e", "instead", "function", "select", "already",
"equal", "access", "management", "hierarchy", "registration", "interest",
"relative", "memory", "offset", "already",
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2010-11-02 03:38:34 +08:00
|
|
|
/* Initialize TSO state of a skb.
|
2009-07-22 07:00:40 +08:00
|
|
|
* This must be invoked the first time we consider transmitting
|
2005-07-06 06:24:38 +08:00
|
|
|
* SKB onto the wire.
|
|
|
|
|
*/
|
2015-06-12 00:15:17 +08:00
|
|
|
static int tcp_init_tso_segs(struct sk_buff *skb, unsigned int mss_now)
|
2005-07-06 06:24:38 +08:00
|
|
|
{
|
|
|
|
|
int tso_segs = tcp_skb_pcount(skb);
|
|
|
|
|
|
2008-12-04 13:24:48 +08:00
|
|
|
if (!tso_segs || (tso_segs > 1 && tcp_skb_mss(skb) != mss_now)) {
|
2015-06-12 00:15:17 +08:00
|
|
|
tcp_set_skb_tso_segs(skb, mss_now);
|
2005-07-06 06:24:38 +08:00
|
|
|
tso_segs = tcp_skb_pcount(skb);
|
|
|
|
|
}
|
|
|
|
|
return tso_segs;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
2012-05-17 07:15:34 +08:00
|
|
|
/* Return true if the Nagle test allows this packet to be
|
2005-07-06 06:24:38 +08:00
|
|
|
* sent now.
|
|
|
|
|
*/
|
2012-05-17 07:15:34 +08:00
|
|
|
static inline bool tcp_nagle_test(const struct tcp_sock *tp, const struct sk_buff *skb,
|
|
|
|
|
unsigned int cur_mss, int nonagle)
|
2005-07-06 06:24:38 +08:00
|
|
|
{
|
|
|
|
|
/* Nagle rule does not apply to frames, which sit in the middle of the
|
|
|
|
|
* write_queue (they have no chances to get new data).
|
|
|
|
|
*
|
|
|
|
|
* This is implemented in the callers, where they modify the 'nonagle'
|
|
|
|
|
* argument based upon the location of SKB in the send queue.
|
|
|
|
|
*/
|
|
|
|
|
if (nonagle & TCP_NAGLE_PUSH)
|
2012-05-17 07:15:34 +08:00
|
|
|
return true;
|
2005-07-06 06:24:38 +08:00
|
|
|
|
2013-03-20 21:32:58 +08:00
|
|
|
/* Don't use the nagle rule for urgent data (or for the final FIN). */
|
|
|
|
|
if (tcp_urg_mode(tp) || (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN))
|
2012-05-17 07:15:34 +08:00
|
|
|
return true;
|
2005-07-06 06:24:38 +08:00
|
|
|
|
2014-03-24 14:49:34 +08:00
|
|
|
if (!tcp_nagle_check(skb->len < cur_mss, tp, nonagle))
|
2012-05-17 07:15:34 +08:00
|
|
|
return true;
|
2005-07-06 06:24:38 +08:00
|
|
|
|
2012-05-17 07:15:34 +08:00
|
|
|
return false;
|
2005-07-06 06:24:38 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Does at least the first segment of SKB fit into the send window? */
|
2012-05-17 07:15:34 +08:00
|
|
|
static bool tcp_snd_wnd_test(const struct tcp_sock *tp,
|
|
|
|
|
const struct sk_buff *skb,
|
|
|
|
|
unsigned int cur_mss)
|
2005-07-06 06:24:38 +08:00
|
|
|
{
|
|
|
|
|
u32 end_seq = TCP_SKB_CB(skb)->end_seq;
|
|
|
|
|
|
|
|
|
|
if (skb->len > cur_mss)
|
|
|
|
|
end_seq = TCP_SKB_CB(skb)->seq + cur_mss;
|
|
|
|
|
|
2007-12-31 20:48:41 +08:00
|
|
|
return !after(end_seq, tcp_wnd_end(tp));
|
2005-07-06 06:24:38 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Trim TSO SKB to LEN bytes, put the remaining data into a new packet
|
|
|
|
|
* which is put after SKB on the list. It is very much like
|
|
|
|
|
* tcp_fragment() except that it may make several kinds of assumptions
|
|
|
|
|
* in order to speed up the splitting operation. In particular, we
|
|
|
|
|
* know that all the data is in scatter-gather pages, and that the
|
|
|
|
|
* packet has never been sent out before (and thus is not cloned).
|
|
|
|
|
*/
|
2019-02-27 01:49:13 +08:00
|
|
|
static int tso_fragment(struct sock *sk, struct sk_buff *skb, unsigned int len,
|
2010-06-24 09:00:22 +08:00
|
|
|
unsigned int mss_now, gfp_t gfp)
|
2005-07-06 06:24:38 +08:00
|
|
|
{
|
|
|
|
|
int nlen = skb->len - len;
|
2019-02-27 01:49:13 +08:00
|
|
|
struct sk_buff *buff;
|
2009-02-28 12:44:42 +08:00
|
|
|
u8 flags;
|
2005-07-06 06:24:38 +08:00
|
|
|
|
|
|
|
|
/* All of a TSO frame must be composed of paged data. */
|
2005-08-17 11:43:40 +08:00
|
|
|
if (skb->len != skb->data_len)
|
2019-02-27 01:49:13 +08:00
|
|
|
return tcp_fragment(sk, TCP_FRAG_IN_WRITE_QUEUE,
|
|
|
|
|
skb, len, mss_now, gfp);
|
2005-07-06 06:24:38 +08:00
|
|
|
|
2015-05-20 04:26:55 +08:00
|
|
|
buff = sk_stream_alloc_skb(sk, 0, gfp, true);
|
2015-04-03 16:17:26 +08:00
|
|
|
if (unlikely(!buff))
|
2005-07-06 06:24:38 +08:00
|
|
|
return -ENOMEM;
|
2019-08-08 08:03:59 +08:00
|
|
|
skb_copy_decrypted(buff, skb);
|
2020-11-04 03:05:05 +08:00
|
|
|
mptcp_skb_ext_copy(buff, skb);
|
2005-07-06 06:24:38 +08:00
|
|
|
|
2019-10-11 11:17:46 +08:00
|
|
|
sk_wmem_queued_add(sk, buff->truesize);
|
2007-12-31 16:11:19 +08:00
|
|
|
sk_mem_charge(sk, buff->truesize);
|
2006-04-20 12:35:00 +08:00
|
|
|
buff->truesize += nlen;
|
2005-07-06 06:24:38 +08:00
|
|
|
skb->truesize -= nlen;
|
|
|
|
|
|
|
|
|
|
/* Correct the sequence numbers. */
|
|
|
|
|
TCP_SKB_CB(buff)->seq = TCP_SKB_CB(skb)->seq + len;
|
|
|
|
|
TCP_SKB_CB(buff)->end_seq = TCP_SKB_CB(skb)->end_seq;
|
|
|
|
|
TCP_SKB_CB(skb)->end_seq = TCP_SKB_CB(buff)->seq;
|
|
|
|
|
|
|
|
|
|
/* PSH and FIN should only be set in the second packet. */
|
2011-09-28 01:25:05 +08:00
|
|
|
flags = TCP_SKB_CB(skb)->tcp_flags;
|
|
|
|
|
TCP_SKB_CB(skb)->tcp_flags = flags & ~(TCPHDR_FIN | TCPHDR_PSH);
|
|
|
|
|
TCP_SKB_CB(buff)->tcp_flags = flags;
|
2005-07-06 06:24:38 +08:00
|
|
|
|
|
|
|
|
/* This packet was never sent out yet, so no SACK bits. */
|
|
|
|
|
TCP_SKB_CB(buff)->sacked = 0;
|
|
|
|
|
|
tcp: Handle eor bit when fragmenting a skb
When fragmenting a skb, the next_skb should carry
the eor from prev_skb. The eor of prev_skb should
also be reset.
Packetdrill script for testing:
~~~~~~
+0 `sysctl -q -w net.ipv4.tcp_min_tso_segs=10`
+0 `sysctl -q -w net.ipv4.tcp_no_metrics_save=1`
+0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
0.100 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 7>
0.100 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 7>
0.200 < . 1:1(0) ack 1 win 257
0.200 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_NODELAY, [1], 4) = 0
0.200 sendto(4, ..., 15330, MSG_EOR, ..., ...) = 15330
0.200 sendto(4, ..., 730, 0, ..., ...) = 730
0.200 > . 1:7301(7300) ack 1
0.200 > . 7301:14601(7300) ack 1
0.300 < . 1:1(0) ack 14601 win 257
0.300 > P. 14601:15331(730) ack 1
0.300 > P. 15331:16061(730) ack 1
0.400 < . 1:1(0) ack 16061 win 257
0.400 close(4) = 0
0.400 > F. 16061:16061(0) ack 1
0.400 < F. 1:1(0) ack 16062 win 257
0.400 > . 16062:16062(0) ack 2
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Soheil Hassas Yeganeh <soheil@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-26 05:44:50 +08:00
|
|
|
tcp_skb_fragment_eor(skb, buff);
|
|
|
|
|
|
2018-02-20 03:56:52 +08:00
|
|
|
buff->ip_summed = CHECKSUM_PARTIAL;
|
2005-07-06 06:24:38 +08:00
|
|
|
skb_split(skb, buff, len);
|
2014-08-13 03:08:12 +08:00
|
|
|
tcp_fragment_tstamp(skb, buff);
|
2005-07-06 06:24:38 +08:00
|
|
|
|
|
|
|
|
/* Fix up tso_factor for both original and new SKB. */
|
2015-06-12 00:15:17 +08:00
|
|
|
tcp_set_skb_tso_segs(skb, mss_now);
|
|
|
|
|
tcp_set_skb_tso_segs(buff, mss_now);
|
2005-07-06 06:24:38 +08:00
|
|
|
|
|
|
|
|
/* Link BUFF into the send queue. */
|
2014-09-23 07:29:32 +08:00
|
|
|
__skb_header_release(buff);
|
2019-02-27 01:49:13 +08:00
|
|
|
tcp_insert_write_queue_after(skb, buff, sk, TCP_FRAG_IN_WRITE_QUEUE);
|
2005-07-06 06:24:38 +08:00
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Try to defer sending, if possible, in order to minimize the amount
|
|
|
|
|
* of TSO splitting we do. View it as a kind of TSO Nagle test.
|
|
|
|
|
*
|
|
|
|
|
* This algorithm is from John Heffner.
|
|
|
|
|
*/
|
2014-05-22 22:41:08 +08:00
|
|
|
static bool tcp_tso_should_defer(struct sock *sk, struct sk_buff *skb,
|
2018-12-07 01:58:24 +08:00
|
|
|
bool *is_cwnd_limited,
|
|
|
|
|
bool *is_rwnd_limited,
|
|
|
|
|
u32 max_segs)
|
2005-07-06 06:24:38 +08:00
|
|
|
{
|
2005-08-10 15:03:31 +08:00
|
|
|
const struct inet_connection_sock *icsk = inet_csk(sk);
|
2018-11-11 22:41:30 +08:00
|
|
|
u32 send_win, cong_win, limit, in_flight;
|
tcp: tso: restore IW10 after TSO autosizing
With sysctl_tcp_min_tso_segs being 4, it is very possible
that tcp_tso_should_defer() decides not sending last 2 MSS
of initial window of 10 packets. This also applies if
autosizing decides to send X MSS per GSO packet, and cwnd
is not a multiple of X.
This patch implements an heuristic based on age of first
skb in write queue : If it was sent very recently (less than half srtt),
we can predict that no ACK packet will come in less than half rtt,
so deferring might cause an under utilization of our window.
This is visible on initial send (IW10) on web servers,
but more generally on some RPC, as the last part of the message
might need an extra RTT to get delivered.
Tested:
Ran following packetdrill test
// A simple server-side test that sends exactly an initial window (IW10)
// worth of packets.
`sysctl -e -q net.ipv4.tcp_min_tso_segs=4`
0.000 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
+.1 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 7>
+0 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 6>
+.1 < . 1:1(0) ack 1 win 257
+0 accept(3, ..., ...) = 4
+0 write(4, ..., 14600) = 14600
+0 > . 1:5841(5840) ack 1 win 457
+0 > . 5841:11681(5840) ack 1 win 457
// Following packet should be sent right now.
+0 > P. 11681:14601(2920) ack 1 win 457
+.1 < . 1:1(0) ack 14601 win 257
+0 close(4) = 0
+0 > F. 14601:14601(0) ack 1
+.1 < F. 1:1(0) ack 14602 win 257
+0 > . 14602:14602(0) ack 2
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-02-27 06:10:19 +08:00
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
struct sk_buff *head;
|
2010-12-07 20:03:55 +08:00
|
|
|
int win_divisor;
|
2018-11-11 22:41:30 +08:00
|
|
|
s64 delta;
|
2005-07-06 06:24:38 +08:00
|
|
|
|
2015-07-26 15:45:24 +08:00
|
|
|
if (icsk->icsk_ca_state >= TCP_CA_Recovery)
|
2006-10-19 11:36:48 +08:00
|
|
|
goto send_now;
|
|
|
|
|
|
2015-02-27 06:10:18 +08:00
|
|
|
/* Avoid bursty behavior by allowing defer
|
2018-11-11 22:41:31 +08:00
|
|
|
* only if the last write was recent (1 ms).
|
|
|
|
|
* Note that tp->tcp_wstamp_ns can be in the future if we have
|
|
|
|
|
* packets waiting in a qdisc or device for EDT delivery.
|
2015-02-27 06:10:18 +08:00
|
|
|
*/
|
2018-11-11 22:41:31 +08:00
|
|
|
delta = tp->tcp_clock_cache - tp->tcp_wstamp_ns - NSEC_PER_MSEC;
|
|
|
|
|
if (delta > 0)
|
2006-10-19 11:36:48 +08:00
|
|
|
goto send_now;
|
2005-07-06 06:43:58 +08:00
|
|
|
|
2005-07-06 06:24:38 +08:00
|
|
|
in_flight = tcp_packets_in_flight(tp);
|
|
|
|
|
|
2018-01-05 07:38:05 +08:00
|
|
|
BUG_ON(tcp_skb_pcount(skb) <= 1);
|
|
|
|
|
BUG_ON(tp->snd_cwnd <= in_flight);
|
2005-07-06 06:24:38 +08:00
|
|
|
|
2007-12-31 20:48:41 +08:00
|
|
|
send_win = tcp_wnd_end(tp) - TCP_SKB_CB(skb)->seq;
|
2005-07-06 06:24:38 +08:00
|
|
|
|
|
|
|
|
/* From in_flight test above, we know that cwnd > in_flight. */
|
|
|
|
|
cong_win = (tp->snd_cwnd - in_flight) * tp->mss_cache;
|
|
|
|
|
|
|
|
|
|
limit = min(send_win, cong_win);
|
|
|
|
|
|
2006-03-12 10:51:49 +08:00
|
|
|
/* If a full-sized TSO skb can be sent, do it. */
|
tcp: refine TSO autosizing
Commit 95bd09eb2750 ("tcp: TSO packets automatic sizing") tried to
control TSO size, but did this at the wrong place (sendmsg() time)
At sendmsg() time, we might have a pessimistic view of flow rate,
and we end up building very small skbs (with 2 MSS per skb).
This is bad because :
- It sends small TSO packets even in Slow Start where rate quickly
increases.
- It tends to make socket write queue very big, increasing tcp_ack()
processing time, but also increasing memory needs, not necessarily
accounted for, as fast clones overhead is currently ignored.
- Lower GRO efficiency and more ACK packets.
Servers with a lot of small lived connections suffer from this.
Lets instead fill skbs as much as possible (64KB of payload), but split
them at xmit time, when we have a precise idea of the flow rate.
skb split is actually quite efficient.
Patch looks bigger than necessary, because TCP Small Queue decision now
has to take place after the eventual split.
As Neal suggested, introduce a new tcp_tso_autosize() helper, so that
tcp_tso_should_defer() can be synchronized on same goal.
Rename tp->xmit_size_goal_segs to tp->gso_segs, as this variable
contains number of mss that we can put in GSO packet, and is not
related to the autosizing goal anymore.
Tested:
40 ms rtt link
nstat >/dev/null
netperf -H remote -l -2000000 -- -s 1000000
nstat | egrep "IpInReceives|IpOutRequests|TcpOutSegs|IpExtOutOctets"
Before patch :
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/s
87380 2000000 2000000 0.36 44.22
IpInReceives 600 0.0
IpOutRequests 599 0.0
TcpOutSegs 1397 0.0
IpExtOutOctets 2033249 0.0
After patch :
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
87380 2000000 2000000 0.36 44.27
IpInReceives 221 0.0
IpOutRequests 232 0.0
TcpOutSegs 1397 0.0
IpExtOutOctets 2013953 0.0
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-12-08 04:22:18 +08:00
|
|
|
if (limit >= max_segs * tp->mss_cache)
|
2006-10-19 11:36:48 +08:00
|
|
|
goto send_now;
|
2006-03-12 10:51:49 +08:00
|
|
|
|
2009-02-28 12:44:29 +08:00
|
|
|
/* Middle in queue won't get any more data, full sendable already? */
|
|
|
|
|
if ((skb != tcp_write_queue_tail(sk)) && (limit >= skb->len))
|
|
|
|
|
goto send_now;
|
|
|
|
|
|
Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next
Pull networking updates from David Miller:
"Highlights:
1) Maintain the TCP retransmit queue using an rbtree, with 1GB
windows at 100Gb this really has become necessary. From Eric
Dumazet.
2) Multi-program support for cgroup+bpf, from Alexei Starovoitov.
3) Perform broadcast flooding in hardware in mv88e6xxx, from Andrew
Lunn.
4) Add meter action support to openvswitch, from Andy Zhou.
5) Add a data meta pointer for BPF accessible packets, from Daniel
Borkmann.
6) Namespace-ify almost all TCP sysctl knobs, from Eric Dumazet.
7) Turn on Broadcom Tags in b53 driver, from Florian Fainelli.
8) More work to move the RTNL mutex down, from Florian Westphal.
9) Add 'bpftool' utility, to help with bpf program introspection.
From Jakub Kicinski.
10) Add new 'cpumap' type for XDP_REDIRECT action, from Jesper
Dangaard Brouer.
11) Support 'blocks' of transformations in the packet scheduler which
can span multiple network devices, from Jiri Pirko.
12) TC flower offload support in cxgb4, from Kumar Sanghvi.
13) Priority based stream scheduler for SCTP, from Marcelo Ricardo
Leitner.
14) Thunderbolt networking driver, from Amir Levy and Mika Westerberg.
15) Add RED qdisc offloadability, and use it in mlxsw driver. From
Nogah Frankel.
16) eBPF based device controller for cgroup v2, from Roman Gushchin.
17) Add some fundamental tracepoints for TCP, from Song Liu.
18) Remove garbage collection from ipv6 route layer, this is a
significant accomplishment. From Wei Wang.
19) Add multicast route offload support to mlxsw, from Yotam Gigi"
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next: (2177 commits)
tcp: highest_sack fix
geneve: fix fill_info when link down
bpf: fix lockdep splat
net: cdc_ncm: GetNtbFormat endian fix
openvswitch: meter: fix NULL pointer dereference in ovs_meter_cmd_reply_start
netem: remove unnecessary 64 bit modulus
netem: use 64 bit divide by rate
tcp: Namespace-ify sysctl_tcp_default_congestion_control
net: Protect iterations over net::fib_notifier_ops in fib_seq_sum()
ipv6: set all.accept_dad to 0 by default
uapi: fix linux/tls.h userspace compilation error
usbnet: ipheth: prevent TX queue timeouts when device not ready
vhost_net: conditionally enable tx polling
uapi: fix linux/rxrpc.h userspace compilation errors
net: stmmac: fix LPI transitioning for dwmac4
atm: horizon: Fix irq release error
net-sysfs: trigger netlink notification on ifalias change via sysfs
openvswitch: Using kfree_rcu() to simplify the code
openvswitch: Make local function ovs_nsh_key_attr_size() static
openvswitch: Fix return value check in ovs_meter_cmd_features()
...
2017-11-16 03:56:19 +08:00
|
|
|
win_divisor = READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_tso_win_divisor);
|
2010-12-07 20:03:55 +08:00
|
|
|
if (win_divisor) {
|
2005-07-06 06:24:38 +08:00
|
|
|
u32 chunk = min(tp->snd_wnd, tp->snd_cwnd * tp->mss_cache);
|
|
|
|
|
|
|
|
|
|
/* If at least some fraction of a window is available,
|
|
|
|
|
* just use it.
|
|
|
|
|
*/
|
2010-12-07 20:03:55 +08:00
|
|
|
chunk /= win_divisor;
|
2005-07-06 06:24:38 +08:00
|
|
|
if (limit >= chunk)
|
2006-10-19 11:36:48 +08:00
|
|
|
goto send_now;
|
2005-07-06 06:24:38 +08:00
|
|
|
} else {
|
|
|
|
|
/* Different approach, try not to defer past a single
|
|
|
|
|
* ACK. Receiver should ACK every other full sized
|
|
|
|
|
* frame, so if we have space for more than 3 frames
|
|
|
|
|
* then send now.
|
|
|
|
|
*/
|
2011-11-22 01:15:14 +08:00
|
|
|
if (limit > tcp_max_tso_deferred_mss(tp) * tp->mss_cache)
|
2006-10-19 11:36:48 +08:00
|
|
|
goto send_now;
|
2005-07-06 06:24:38 +08:00
|
|
|
}
|
|
|
|
|
|
2017-10-06 13:21:27 +08:00
|
|
|
/* TODO : use tsorted_sent_queue ? */
|
|
|
|
|
head = tcp_rtx_queue_head(sk);
|
|
|
|
|
if (!head)
|
|
|
|
|
goto send_now;
|
2018-11-11 22:41:30 +08:00
|
|
|
delta = tp->tcp_clock_cache - head->tstamp;
|
tcp: tso: restore IW10 after TSO autosizing
With sysctl_tcp_min_tso_segs being 4, it is very possible
that tcp_tso_should_defer() decides not sending last 2 MSS
of initial window of 10 packets. This also applies if
autosizing decides to send X MSS per GSO packet, and cwnd
is not a multiple of X.
This patch implements an heuristic based on age of first
skb in write queue : If it was sent very recently (less than half srtt),
we can predict that no ACK packet will come in less than half rtt,
so deferring might cause an under utilization of our window.
This is visible on initial send (IW10) on web servers,
but more generally on some RPC, as the last part of the message
might need an extra RTT to get delivered.
Tested:
Ran following packetdrill test
// A simple server-side test that sends exactly an initial window (IW10)
// worth of packets.
`sysctl -e -q net.ipv4.tcp_min_tso_segs=4`
0.000 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
+.1 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 7>
+0 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 6>
+.1 < . 1:1(0) ack 1 win 257
+0 accept(3, ..., ...) = 4
+0 write(4, ..., 14600) = 14600
+0 > . 1:5841(5840) ack 1 win 457
+0 > . 5841:11681(5840) ack 1 win 457
// Following packet should be sent right now.
+0 > P. 11681:14601(2920) ack 1 win 457
+.1 < . 1:1(0) ack 14601 win 257
+0 close(4) = 0
+0 > F. 14601:14601(0) ack 1
+.1 < F. 1:1(0) ack 14602 win 257
+0 > . 14602:14602(0) ack 2
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-02-27 06:10:19 +08:00
|
|
|
/* If next ACK is likely to come too late (half srtt), do not defer */
|
2018-11-11 22:41:30 +08:00
|
|
|
if ((s64)(delta - (u64)NSEC_PER_USEC * (tp->srtt_us >> 4)) < 0)
|
tcp: tso: restore IW10 after TSO autosizing
With sysctl_tcp_min_tso_segs being 4, it is very possible
that tcp_tso_should_defer() decides not sending last 2 MSS
of initial window of 10 packets. This also applies if
autosizing decides to send X MSS per GSO packet, and cwnd
is not a multiple of X.
This patch implements an heuristic based on age of first
skb in write queue : If it was sent very recently (less than half srtt),
we can predict that no ACK packet will come in less than half rtt,
so deferring might cause an under utilization of our window.
This is visible on initial send (IW10) on web servers,
but more generally on some RPC, as the last part of the message
might need an extra RTT to get delivered.
Tested:
Ran following packetdrill test
// A simple server-side test that sends exactly an initial window (IW10)
// worth of packets.
`sysctl -e -q net.ipv4.tcp_min_tso_segs=4`
0.000 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
+.1 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 7>
+0 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 6>
+.1 < . 1:1(0) ack 1 win 257
+0 accept(3, ..., ...) = 4
+0 write(4, ..., 14600) = 14600
+0 > . 1:5841(5840) ack 1 win 457
+0 > . 5841:11681(5840) ack 1 win 457
// Following packet should be sent right now.
+0 > P. 11681:14601(2920) ack 1 win 457
+.1 < . 1:1(0) ack 14601 win 257
+0 close(4) = 0
+0 > F. 14601:14601(0) ack 1
+.1 < F. 1:1(0) ack 14602 win 257
+0 > . 14602:14602(0) ack 2
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-02-27 06:10:19 +08:00
|
|
|
goto send_now;
|
|
|
|
|
|
2018-12-07 01:58:24 +08:00
|
|
|
/* Ok, it looks like it is advisable to defer.
|
|
|
|
|
* Three cases are tracked :
|
|
|
|
|
* 1) We are cwnd-limited
|
|
|
|
|
* 2) We are rwnd-limited
|
|
|
|
|
* 3) We are application limited.
|
|
|
|
|
*/
|
|
|
|
|
if (cong_win < send_win) {
|
|
|
|
|
if (cong_win <= skb->len) {
|
|
|
|
|
*is_cwnd_limited = true;
|
|
|
|
|
return true;
|
|
|
|
|
}
|
|
|
|
|
} else {
|
|
|
|
|
if (send_win <= skb->len) {
|
|
|
|
|
*is_rwnd_limited = true;
|
|
|
|
|
return true;
|
|
|
|
|
}
|
|
|
|
|
}
|
2006-10-19 11:36:48 +08:00
|
|
|
|
2018-12-07 01:58:24 +08:00
|
|
|
/* If this packet won't get more data, do not wait. */
|
2018-12-10 22:10:02 +08:00
|
|
|
if ((TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN) ||
|
|
|
|
|
TCP_SKB_CB(skb)->eor)
|
2018-12-07 01:58:24 +08:00
|
|
|
goto send_now;
|
2014-05-22 22:41:08 +08:00
|
|
|
|
2012-05-17 07:15:34 +08:00
|
|
|
return true;
|
2006-10-19 11:36:48 +08:00
|
|
|
|
|
|
|
|
send_now:
|
2012-05-17 07:15:34 +08:00
|
|
|
return false;
|
2005-07-06 06:24:38 +08:00
|
|
|
}
|
|
|
|
|
|
2015-03-06 11:18:24 +08:00
|
|
|
static inline void tcp_mtu_check_reprobe(struct sock *sk)
|
|
|
|
|
{
|
|
|
|
|
struct inet_connection_sock *icsk = inet_csk(sk);
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
struct net *net = sock_net(sk);
|
|
|
|
|
u32 interval;
|
|
|
|
|
s32 delta;
|
|
|
|
|
|
|
|
|
|
interval = net->ipv4.sysctl_tcp_probe_interval;
|
2017-05-17 05:00:08 +08:00
|
|
|
delta = tcp_jiffies32 - icsk->icsk_mtup.probe_timestamp;
|
2015-03-06 11:18:24 +08:00
|
|
|
if (unlikely(delta >= interval * HZ)) {
|
|
|
|
|
int mss = tcp_current_mss(sk);
|
|
|
|
|
|
|
|
|
|
/* Update current search range */
|
|
|
|
|
icsk->icsk_mtup.probe_size = 0;
|
|
|
|
|
icsk->icsk_mtup.search_high = tp->rx_opt.mss_clamp +
|
|
|
|
|
sizeof(struct tcphdr) +
|
|
|
|
|
icsk->icsk_af_ops->net_header_len;
|
|
|
|
|
icsk->icsk_mtup.search_low = tcp_mss_to_mtu(sk, mss);
|
|
|
|
|
|
|
|
|
|
/* Update probe time stamp */
|
2017-05-17 05:00:08 +08:00
|
|
|
icsk->icsk_mtup.probe_timestamp = tcp_jiffies32;
|
2015-03-06 11:18:24 +08:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2018-02-12 18:57:04 +08:00
|
|
|
static bool tcp_can_coalesce_send_queue_head(struct sock *sk, int len)
|
|
|
|
|
{
|
|
|
|
|
struct sk_buff *skb, *next;
|
|
|
|
|
|
|
|
|
|
skb = tcp_send_head(sk);
|
|
|
|
|
tcp_for_write_queue_from_safe(skb, next, sk) {
|
|
|
|
|
if (len <= skb->len)
|
|
|
|
|
break;
|
|
|
|
|
|
2019-08-28 03:09:33 +08:00
|
|
|
if (unlikely(TCP_SKB_CB(skb)->eor) || tcp_has_tx_tstamp(skb))
|
2018-02-12 18:57:04 +08:00
|
|
|
return false;
|
|
|
|
|
|
|
|
|
|
len -= skb->len;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return true;
|
|
|
|
|
}
|
|
|
|
|
|
2006-03-21 09:53:41 +08:00
|
|
|
/* Create a new MTU probe if we are ready.
|
2009-07-22 07:00:40 +08:00
|
|
|
* MTU probe is regularly attempting to increase the path MTU by
|
|
|
|
|
* deliberately sending larger packets. This discovers routing
|
|
|
|
|
* changes resulting in larger path MTUs.
|
|
|
|
|
*
|
2006-03-21 09:53:41 +08:00
|
|
|
* Returns 0 if we should wait to probe (no cwnd available),
|
|
|
|
|
* 1 if a probe was sent,
|
2008-01-01 06:57:14 +08:00
|
|
|
* -1 otherwise
|
|
|
|
|
*/
|
2006-03-21 09:53:41 +08:00
|
|
|
static int tcp_mtu_probe(struct sock *sk)
|
|
|
|
|
{
|
|
|
|
|
struct inet_connection_sock *icsk = inet_csk(sk);
|
2016-12-04 03:14:55 +08:00
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
2006-03-21 09:53:41 +08:00
|
|
|
struct sk_buff *skb, *nskb, *next;
|
2015-03-06 11:18:23 +08:00
|
|
|
struct net *net = sock_net(sk);
|
2006-03-21 09:53:41 +08:00
|
|
|
int probe_size;
|
2007-11-23 19:08:16 +08:00
|
|
|
int size_needed;
|
2016-12-04 03:14:55 +08:00
|
|
|
int copy, len;
|
2006-03-21 09:53:41 +08:00
|
|
|
int mss_now;
|
2015-03-06 11:18:23 +08:00
|
|
|
int interval;
|
2006-03-21 09:53:41 +08:00
|
|
|
|
|
|
|
|
/* Not currently probing/verifying,
|
|
|
|
|
* not in recovery,
|
|
|
|
|
* have enough cwnd, and
|
2016-12-04 03:14:55 +08:00
|
|
|
* not SACKing (the variable headers throw things off)
|
|
|
|
|
*/
|
|
|
|
|
if (likely(!icsk->icsk_mtup.enabled ||
|
|
|
|
|
icsk->icsk_mtup.probe_size ||
|
|
|
|
|
inet_csk(sk)->icsk_ca_state != TCP_CA_Open ||
|
|
|
|
|
tp->snd_cwnd < 11 ||
|
|
|
|
|
tp->rx_opt.num_sacks || tp->rx_opt.dsack))
|
2006-03-21 09:53:41 +08:00
|
|
|
return -1;
|
|
|
|
|
|
2015-03-06 11:18:23 +08:00
|
|
|
/* Use binary search for probe_size between tcp_mss_base,
|
|
|
|
|
* and current mss_clamp. if (search_high - search_low)
|
|
|
|
|
* smaller than a threshold, backoff from probing.
|
|
|
|
|
*/
|
2009-03-14 22:23:05 +08:00
|
|
|
mss_now = tcp_current_mss(sk);
|
2015-03-06 11:18:23 +08:00
|
|
|
probe_size = tcp_mtu_to_mss(sk, (icsk->icsk_mtup.search_high +
|
|
|
|
|
icsk->icsk_mtup.search_low) >> 1);
|
2007-11-23 19:08:16 +08:00
|
|
|
size_needed = probe_size + (tp->reordering + 1) * tp->mss_cache;
|
2015-03-06 11:18:23 +08:00
|
|
|
interval = icsk->icsk_mtup.search_high - icsk->icsk_mtup.search_low;
|
2015-03-06 11:18:24 +08:00
|
|
|
/* When misfortune happens, we are reprobing actively,
|
|
|
|
|
* and then reprobe timer has expired. We stick with current
|
|
|
|
|
* probing process by not resetting search range to its orignal.
|
|
|
|
|
*/
|
2015-03-06 11:18:23 +08:00
|
|
|
if (probe_size > tcp_mtu_to_mss(sk, icsk->icsk_mtup.search_high) ||
|
2015-03-06 11:18:24 +08:00
|
|
|
interval < net->ipv4.sysctl_tcp_probe_threshold) {
|
|
|
|
|
/* Check whether enough time has elaplased for
|
|
|
|
|
* another round of probing.
|
|
|
|
|
*/
|
|
|
|
|
tcp_mtu_check_reprobe(sk);
|
2006-03-21 09:53:41 +08:00
|
|
|
return -1;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Have enough data in the send queue to probe? */
|
2007-11-23 19:10:56 +08:00
|
|
|
if (tp->write_seq - tp->snd_nxt < size_needed)
|
2006-03-21 09:53:41 +08:00
|
|
|
return -1;
|
|
|
|
|
|
2007-11-23 19:08:16 +08:00
|
|
|
if (tp->snd_wnd < size_needed)
|
|
|
|
|
return -1;
|
2007-12-31 20:48:41 +08:00
|
|
|
if (after(tp->snd_nxt + size_needed, tcp_wnd_end(tp)))
|
2007-11-23 19:08:16 +08:00
|
|
|
return 0;
|
2006-03-21 09:53:41 +08:00
|
|
|
|
2007-12-02 06:48:01 +08:00
|
|
|
/* Do we need to wait to drain cwnd? With none in flight, don't stall */
|
|
|
|
|
if (tcp_packets_in_flight(tp) + 2 > tp->snd_cwnd) {
|
|
|
|
|
if (!tcp_packets_in_flight(tp))
|
2006-03-21 09:53:41 +08:00
|
|
|
return -1;
|
|
|
|
|
else
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
2018-02-12 18:57:04 +08:00
|
|
|
if (!tcp_can_coalesce_send_queue_head(sk, probe_size))
|
|
|
|
|
return -1;
|
|
|
|
|
|
2006-03-21 09:53:41 +08:00
|
|
|
/* We're allowed to probe. Build it now. */
|
2015-05-20 04:26:55 +08:00
|
|
|
nskb = sk_stream_alloc_skb(sk, probe_size, GFP_ATOMIC, false);
|
2015-04-03 16:17:26 +08:00
|
|
|
if (!nskb)
|
2006-03-21 09:53:41 +08:00
|
|
|
return -1;
|
2019-10-11 11:17:46 +08:00
|
|
|
sk_wmem_queued_add(sk, nskb->truesize);
|
2007-12-31 16:11:19 +08:00
|
|
|
sk_mem_charge(sk, nskb->truesize);
|
2006-03-21 09:53:41 +08:00
|
|
|
|
2007-03-08 04:12:44 +08:00
|
|
|
skb = tcp_send_head(sk);
|
2019-08-08 08:03:59 +08:00
|
|
|
skb_copy_decrypted(nskb, skb);
|
2020-11-04 03:05:05 +08:00
|
|
|
mptcp_skb_ext_copy(nskb, skb);
|
2006-03-21 09:53:41 +08:00
|
|
|
|
|
|
|
|
TCP_SKB_CB(nskb)->seq = TCP_SKB_CB(skb)->seq;
|
|
|
|
|
TCP_SKB_CB(nskb)->end_seq = TCP_SKB_CB(skb)->seq + probe_size;
|
2011-09-28 01:25:05 +08:00
|
|
|
TCP_SKB_CB(nskb)->tcp_flags = TCPHDR_ACK;
|
2006-03-21 09:53:41 +08:00
|
|
|
TCP_SKB_CB(nskb)->sacked = 0;
|
|
|
|
|
nskb->csum = 0;
|
2018-02-20 03:56:52 +08:00
|
|
|
nskb->ip_summed = CHECKSUM_PARTIAL;
|
2006-03-21 09:53:41 +08:00
|
|
|
|
2007-12-02 06:48:00 +08:00
|
|
|
tcp_insert_write_queue_before(nskb, skb, sk);
|
2017-10-31 14:08:20 +08:00
|
|
|
tcp_highest_sack_replace(sk, skb, nskb);
|
2007-12-02 06:48:00 +08:00
|
|
|
|
2006-03-21 09:53:41 +08:00
|
|
|
len = 0;
|
2007-12-02 06:48:02 +08:00
|
|
|
tcp_for_write_queue_from_safe(skb, next, sk) {
|
2006-03-21 09:53:41 +08:00
|
|
|
copy = min_t(int, skb->len, probe_size - len);
|
2018-02-20 03:56:52 +08:00
|
|
|
skb_copy_bits(skb, 0, skb_put(nskb, copy), copy);
|
2006-03-21 09:53:41 +08:00
|
|
|
|
|
|
|
|
if (skb->len <= copy) {
|
|
|
|
|
/* We've eaten all the data from this skb.
|
|
|
|
|
* Throw it away. */
|
2011-09-28 01:25:05 +08:00
|
|
|
TCP_SKB_CB(nskb)->tcp_flags |= TCP_SKB_CB(skb)->tcp_flags;
|
2018-02-12 18:57:04 +08:00
|
|
|
/* If this is the last SKB we copy and eor is set
|
|
|
|
|
* we need to propagate it to the new skb.
|
|
|
|
|
*/
|
|
|
|
|
TCP_SKB_CB(nskb)->eor = TCP_SKB_CB(skb)->eor;
|
2019-08-28 03:09:33 +08:00
|
|
|
tcp_skb_collapse_tstamp(nskb, skb);
|
2007-03-08 04:12:44 +08:00
|
|
|
tcp_unlink_write_queue(skb, sk);
|
2007-12-31 16:11:19 +08:00
|
|
|
sk_wmem_free_skb(sk, skb);
|
2006-03-21 09:53:41 +08:00
|
|
|
} else {
|
2011-09-28 01:25:05 +08:00
|
|
|
TCP_SKB_CB(nskb)->tcp_flags |= TCP_SKB_CB(skb)->tcp_flags &
|
2010-06-12 22:01:43 +08:00
|
|
|
~(TCPHDR_FIN|TCPHDR_PSH);
|
2006-03-21 09:53:41 +08:00
|
|
|
if (!skb_shinfo(skb)->nr_frags) {
|
|
|
|
|
skb_pull(skb, copy);
|
|
|
|
|
} else {
|
|
|
|
|
__pskb_trim_head(skb, copy);
|
2015-06-12 00:15:17 +08:00
|
|
|
tcp_set_skb_tso_segs(skb, mss_now);
|
2006-03-21 09:53:41 +08:00
|
|
|
}
|
|
|
|
|
TCP_SKB_CB(skb)->seq += copy;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
len += copy;
|
2007-12-02 06:48:02 +08:00
|
|
|
|
|
|
|
|
if (len >= probe_size)
|
|
|
|
|
break;
|
2006-03-21 09:53:41 +08:00
|
|
|
}
|
2015-06-12 00:15:17 +08:00
|
|
|
tcp_init_tso_segs(nskb, nskb->len);
|
2006-03-21 09:53:41 +08:00
|
|
|
|
|
|
|
|
/* We're ready to send. If this fails, the probe will
|
2014-09-06 06:33:33 +08:00
|
|
|
* be resegmented into mss-sized pieces by tcp_write_xmit().
|
|
|
|
|
*/
|
2006-03-21 09:53:41 +08:00
|
|
|
if (!tcp_transmit_skb(sk, nskb, 1, GFP_ATOMIC)) {
|
|
|
|
|
/* Decrement cwnd here because we are sending
|
2008-01-01 06:57:14 +08:00
|
|
|
* effectively two packets. */
|
2006-03-21 09:53:41 +08:00
|
|
|
tp->snd_cwnd--;
|
2007-12-31 20:43:57 +08:00
|
|
|
tcp_event_new_data_sent(sk, nskb);
|
2006-03-21 09:53:41 +08:00
|
|
|
|
|
|
|
|
icsk->icsk_mtup.probe_size = tcp_mss_to_mtu(sk, nskb->len);
|
2006-03-21 13:32:58 +08:00
|
|
|
tp->mtu_probe.probe_seq_start = TCP_SKB_CB(nskb)->seq;
|
|
|
|
|
tp->mtu_probe.probe_seq_end = TCP_SKB_CB(nskb)->end_seq;
|
2006-03-21 09:53:41 +08:00
|
|
|
|
|
|
|
|
return 1;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return -1;
|
|
|
|
|
}
|
|
|
|
|
|
2018-10-16 00:37:56 +08:00
|
|
|
static bool tcp_pacing_check(struct sock *sk)
|
2017-05-16 19:24:36 +08:00
|
|
|
{
|
2018-10-16 00:37:56 +08:00
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
|
|
|
|
|
if (!tcp_needs_internal_pacing(sk))
|
|
|
|
|
return false;
|
|
|
|
|
|
|
|
|
|
if (tp->tcp_wstamp_ns <= tp->tcp_clock_cache)
|
|
|
|
|
return false;
|
|
|
|
|
|
|
|
|
|
if (!hrtimer_is_queued(&tp->pacing_timer)) {
|
|
|
|
|
hrtimer_start(&tp->pacing_timer,
|
|
|
|
|
ns_to_ktime(tp->tcp_wstamp_ns),
|
|
|
|
|
HRTIMER_MODE_ABS_PINNED_SOFT);
|
|
|
|
|
sock_hold(sk);
|
|
|
|
|
}
|
|
|
|
|
return true;
|
2017-05-16 19:24:36 +08:00
|
|
|
}
|
|
|
|
|
|
2016-09-21 13:45:58 +08:00
|
|
|
/* TCP Small Queues :
|
|
|
|
|
* Control number of packets in qdisc/devices to two packets / or ~1 ms.
|
|
|
|
|
* (These limits are doubled for retransmits)
|
|
|
|
|
* This allows for :
|
|
|
|
|
* - better RTT estimation and ACK scheduling
|
|
|
|
|
* - faster recovery
|
|
|
|
|
* - high rates
|
|
|
|
|
* Alas, some drivers / subsystems require a fair amount
|
|
|
|
|
* of queued bytes to ensure line rate.
|
|
|
|
|
* One example is wifi aggregation (802.11 AMPDU)
|
|
|
|
|
*/
|
|
|
|
|
static bool tcp_small_queue_check(struct sock *sk, const struct sk_buff *skb,
|
|
|
|
|
unsigned int factor)
|
|
|
|
|
{
|
net: extend sk_pacing_rate to unsigned long
sk_pacing_rate has beed introduced as a u32 field in 2013,
effectively limiting per flow pacing to 34Gbit.
We believe it is time to allow TCP to pace high speed flows
on 64bit hosts, as we now can reach 100Gbit on one TCP flow.
This patch adds no cost for 32bit kernels.
The tcpi_pacing_rate and tcpi_max_pacing_rate were already
exported as 64bit, so iproute2/ss command require no changes.
Unfortunately the SO_MAX_PACING_RATE socket option will stay
32bit and we will need to add a new option to let applications
control high pacing rates.
State Recv-Q Send-Q Local Address:Port Peer Address:Port
ESTAB 0 1787144 10.246.9.76:49992 10.246.9.77:36741
timer:(on,003ms,0) ino:91863 sk:2 <->
skmem:(r0,rb540000,t66440,tb2363904,f605944,w1822984,o0,bl0,d0)
ts sack bbr wscale:8,8 rto:201 rtt:0.057/0.006 mss:1448
rcvmss:536 advmss:1448
cwnd:138 ssthresh:178 bytes_acked:256699822585 segs_out:177279177
segs_in:3916318 data_segs_out:177279175
bbr:(bw:31276.8Mbps,mrtt:0,pacing_gain:1.25,cwnd_gain:2)
send 28045.5Mbps lastrcv:73333
pacing_rate 38705.0Mbps delivery_rate 22997.6Mbps
busy:73333ms unacked:135 retrans:0/157 rcv_space:14480
notsent:2085120 minrtt:0.013
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-10-16 00:37:53 +08:00
|
|
|
unsigned long limit;
|
2016-09-21 13:45:58 +08:00
|
|
|
|
net: extend sk_pacing_rate to unsigned long
sk_pacing_rate has beed introduced as a u32 field in 2013,
effectively limiting per flow pacing to 34Gbit.
We believe it is time to allow TCP to pace high speed flows
on 64bit hosts, as we now can reach 100Gbit on one TCP flow.
This patch adds no cost for 32bit kernels.
The tcpi_pacing_rate and tcpi_max_pacing_rate were already
exported as 64bit, so iproute2/ss command require no changes.
Unfortunately the SO_MAX_PACING_RATE socket option will stay
32bit and we will need to add a new option to let applications
control high pacing rates.
State Recv-Q Send-Q Local Address:Port Peer Address:Port
ESTAB 0 1787144 10.246.9.76:49992 10.246.9.77:36741
timer:(on,003ms,0) ino:91863 sk:2 <->
skmem:(r0,rb540000,t66440,tb2363904,f605944,w1822984,o0,bl0,d0)
ts sack bbr wscale:8,8 rto:201 rtt:0.057/0.006 mss:1448
rcvmss:536 advmss:1448
cwnd:138 ssthresh:178 bytes_acked:256699822585 segs_out:177279177
segs_in:3916318 data_segs_out:177279175
bbr:(bw:31276.8Mbps,mrtt:0,pacing_gain:1.25,cwnd_gain:2)
send 28045.5Mbps lastrcv:73333
pacing_rate 38705.0Mbps delivery_rate 22997.6Mbps
busy:73333ms unacked:135 retrans:0/157 rcv_space:14480
notsent:2085120 minrtt:0.013
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-10-16 00:37:53 +08:00
|
|
|
limit = max_t(unsigned long,
|
|
|
|
|
2 * skb->truesize,
|
2019-12-17 10:51:03 +08:00
|
|
|
sk->sk_pacing_rate >> READ_ONCE(sk->sk_pacing_shift));
|
2018-11-11 23:34:28 +08:00
|
|
|
if (sk->sk_pacing_status == SK_PACING_NONE)
|
|
|
|
|
limit = min_t(unsigned long, limit,
|
|
|
|
|
sock_net(sk)->ipv4.sysctl_tcp_limit_output_bytes);
|
2016-09-21 13:45:58 +08:00
|
|
|
limit <<= factor;
|
|
|
|
|
|
tcp: add optional per socket transmit delay
Adding delays to TCP flows is crucial for studying behavior
of TCP stacks, including congestion control modules.
Linux offers netem module, but it has unpractical constraints :
- Need root access to change qdisc
- Hard to setup on egress if combined with non trivial qdisc like FQ
- Single delay for all flows.
EDT (Earliest Departure Time) adoption in TCP stack allows us
to enable a per socket delay at a very small cost.
Networking tools can now establish thousands of flows, each of them
with a different delay, simulating real world conditions.
This requires FQ packet scheduler or a EDT-enabled NIC.
This patchs adds TCP_TX_DELAY socket option, to set a delay in
usec units.
unsigned int tx_delay = 10000; /* 10 msec */
setsockopt(fd, SOL_TCP, TCP_TX_DELAY, &tx_delay, sizeof(tx_delay));
Note that FQ packet scheduler limits might need some tweaking :
man tc-fq
PARAMETERS
limit
Hard limit on the real queue size. When this limit is
reached, new packets are dropped. If the value is lowered,
packets are dropped so that the new limit is met. Default
is 10000 packets.
flow_limit
Hard limit on the maximum number of packets queued per
flow. Default value is 100.
Use of TCP_TX_DELAY option will increase number of skbs in FQ qdisc,
so packets would be dropped if any of the previous limit is hit.
Use of a jump label makes this support runtime-free, for hosts
never using the option.
Also note that TSQ (TCP Small Queues) limits are slightly changed
with this patch : we need to account that skbs artificially delayed
wont stop us providind more skbs to feed the pipe (netem uses
skb_orphan_partial() for this purpose, but FQ can not use this trick)
Because of that, using big delays might very well trigger
old bugs in TSO auto defer logic and/or sndbuf limited detection.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2019-06-13 02:57:25 +08:00
|
|
|
if (static_branch_unlikely(&tcp_tx_delay_enabled) &&
|
|
|
|
|
tcp_sk(sk)->tcp_tx_delay) {
|
|
|
|
|
u64 extra_bytes = (u64)sk->sk_pacing_rate * tcp_sk(sk)->tcp_tx_delay;
|
|
|
|
|
|
|
|
|
|
/* TSQ is based on skb truesize sum (sk_wmem_alloc), so we
|
|
|
|
|
* approximate our needs assuming an ~100% skb->truesize overhead.
|
|
|
|
|
* USEC_PER_SEC is approximated by 2^20.
|
|
|
|
|
* do_div(extra_bytes, USEC_PER_SEC/2) is replaced by a right shift.
|
|
|
|
|
*/
|
|
|
|
|
extra_bytes >>= (20 - 1);
|
|
|
|
|
limit += extra_bytes;
|
|
|
|
|
}
|
2017-06-30 18:08:00 +08:00
|
|
|
if (refcount_read(&sk->sk_wmem_alloc) > limit) {
|
2017-10-06 13:21:27 +08:00
|
|
|
/* Always send skb if rtx queue is empty.
|
2016-12-04 03:14:54 +08:00
|
|
|
* No need to wait for TX completion to call us back,
|
|
|
|
|
* after softirq/tasklet schedule.
|
|
|
|
|
* This helps when TX completions are delayed too much.
|
|
|
|
|
*/
|
2017-10-06 13:21:27 +08:00
|
|
|
if (tcp_rtx_queue_empty(sk))
|
2016-12-04 03:14:54 +08:00
|
|
|
return false;
|
|
|
|
|
|
2016-12-04 03:14:57 +08:00
|
|
|
set_bit(TSQ_THROTTLED, &sk->sk_tsq_flags);
|
2016-09-21 13:45:58 +08:00
|
|
|
/* It is possible TX completion already happened
|
|
|
|
|
* before we set TSQ_THROTTLED, so we must
|
|
|
|
|
* test again the condition.
|
|
|
|
|
*/
|
|
|
|
|
smp_mb__after_atomic();
|
2017-06-30 18:08:00 +08:00
|
|
|
if (refcount_read(&sk->sk_wmem_alloc) > limit)
|
2016-09-21 13:45:58 +08:00
|
|
|
return true;
|
|
|
|
|
}
|
|
|
|
|
return false;
|
|
|
|
|
}
|
|
|
|
|
|
2016-11-28 15:07:13 +08:00
|
|
|
static void tcp_chrono_set(struct tcp_sock *tp, const enum tcp_chrono new)
|
|
|
|
|
{
|
2017-05-17 05:00:09 +08:00
|
|
|
const u32 now = tcp_jiffies32;
|
2017-07-28 22:41:37 +08:00
|
|
|
enum tcp_chrono old = tp->chrono_type;
|
2016-11-28 15:07:13 +08:00
|
|
|
|
2017-07-28 22:41:37 +08:00
|
|
|
if (old > TCP_CHRONO_UNSPEC)
|
|
|
|
|
tp->chrono_stat[old - 1] += now - tp->chrono_start;
|
2016-11-28 15:07:13 +08:00
|
|
|
tp->chrono_start = now;
|
|
|
|
|
tp->chrono_type = new;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void tcp_chrono_start(struct sock *sk, const enum tcp_chrono type)
|
|
|
|
|
{
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
|
|
|
|
|
/* If there are multiple conditions worthy of tracking in a
|
2016-11-28 15:07:14 +08:00
|
|
|
* chronograph then the highest priority enum takes precedence
|
|
|
|
|
* over the other conditions. So that if something "more interesting"
|
2016-11-28 15:07:13 +08:00
|
|
|
* starts happening, stop the previous chrono and start a new one.
|
|
|
|
|
*/
|
|
|
|
|
if (type > tp->chrono_type)
|
|
|
|
|
tcp_chrono_set(tp, type);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void tcp_chrono_stop(struct sock *sk, const enum tcp_chrono type)
|
|
|
|
|
{
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
|
2016-11-28 15:07:14 +08:00
|
|
|
|
|
|
|
|
/* There are multiple conditions worthy of tracking in a
|
|
|
|
|
* chronograph, so that the highest priority enum takes
|
|
|
|
|
* precedence over the other conditions (see tcp_chrono_start).
|
|
|
|
|
* If a condition stops, we only stop chrono tracking if
|
|
|
|
|
* it's the "most interesting" or current chrono we are
|
|
|
|
|
* tracking and starts busy chrono if we have pending data.
|
|
|
|
|
*/
|
2017-10-06 13:21:27 +08:00
|
|
|
if (tcp_rtx_and_write_queues_empty(sk))
|
2016-11-28 15:07:14 +08:00
|
|
|
tcp_chrono_set(tp, TCP_CHRONO_UNSPEC);
|
|
|
|
|
else if (type == tp->chrono_type)
|
|
|
|
|
tcp_chrono_set(tp, TCP_CHRONO_BUSY);
|
2016-11-28 15:07:13 +08:00
|
|
|
}
|
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/* This routine writes packets to the network. It advances the
|
|
|
|
|
* send_head. This happens as incoming acks open up the remote
|
|
|
|
|
* window for us.
|
|
|
|
|
*
|
2008-12-04 13:24:48 +08:00
|
|
|
* LARGESEND note: !tcp_urg_mode is overkill, only frames between
|
|
|
|
|
* snd_up-64k-mss .. snd_up cannot be large. However, taking into
|
|
|
|
|
* account rare use of URG, this is not a big flaw.
|
|
|
|
|
*
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
* Send at most one packet when push_one > 0. Temporarily ignore
|
|
|
|
|
* cwnd limit to force at most one packet out when push_one == 2.
|
|
|
|
|
|
2012-05-17 07:15:34 +08:00
|
|
|
* Returns true, if no segments are in flight and we have queued segments,
|
|
|
|
|
* but cannot send anything now because of SWS or another problem.
|
2005-04-17 06:20:36 +08:00
|
|
|
*/
|
2012-05-17 07:15:34 +08:00
|
|
|
static bool tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle,
|
|
|
|
|
int push_one, gfp_t gfp)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
2005-07-06 06:19:06 +08:00
|
|
|
struct sk_buff *skb;
|
2005-07-06 06:24:38 +08:00
|
|
|
unsigned int tso_segs, sent_pkts;
|
|
|
|
|
int cwnd_quota;
|
2006-03-21 09:53:41 +08:00
|
|
|
int result;
|
2016-11-28 15:07:15 +08:00
|
|
|
bool is_cwnd_limited = false, is_rwnd_limited = false;
|
tcp: refine TSO autosizing
Commit 95bd09eb2750 ("tcp: TSO packets automatic sizing") tried to
control TSO size, but did this at the wrong place (sendmsg() time)
At sendmsg() time, we might have a pessimistic view of flow rate,
and we end up building very small skbs (with 2 MSS per skb).
This is bad because :
- It sends small TSO packets even in Slow Start where rate quickly
increases.
- It tends to make socket write queue very big, increasing tcp_ack()
processing time, but also increasing memory needs, not necessarily
accounted for, as fast clones overhead is currently ignored.
- Lower GRO efficiency and more ACK packets.
Servers with a lot of small lived connections suffer from this.
Lets instead fill skbs as much as possible (64KB of payload), but split
them at xmit time, when we have a precise idea of the flow rate.
skb split is actually quite efficient.
Patch looks bigger than necessary, because TCP Small Queue decision now
has to take place after the eventual split.
As Neal suggested, introduce a new tcp_tso_autosize() helper, so that
tcp_tso_should_defer() can be synchronized on same goal.
Rename tp->xmit_size_goal_segs to tp->gso_segs, as this variable
contains number of mss that we can put in GSO packet, and is not
related to the autosizing goal anymore.
Tested:
40 ms rtt link
nstat >/dev/null
netperf -H remote -l -2000000 -- -s 1000000
nstat | egrep "IpInReceives|IpOutRequests|TcpOutSegs|IpExtOutOctets"
Before patch :
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/s
87380 2000000 2000000 0.36 44.22
IpInReceives 600 0.0
IpOutRequests 599 0.0
TcpOutSegs 1397 0.0
IpExtOutOctets 2033249 0.0
After patch :
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
87380 2000000 2000000 0.36 44.27
IpInReceives 221 0.0
IpOutRequests 232 0.0
TcpOutSegs 1397 0.0
IpExtOutOctets 2013953 0.0
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-12-08 04:22:18 +08:00
|
|
|
u32 max_segs;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2005-07-06 06:19:06 +08:00
|
|
|
sent_pkts = 0;
|
2006-03-21 09:53:41 +08:00
|
|
|
|
2017-10-27 12:21:40 +08:00
|
|
|
tcp_mstamp_refresh(tp);
|
2008-12-06 14:48:55 +08:00
|
|
|
if (!push_one) {
|
|
|
|
|
/* Do MTU probing. */
|
|
|
|
|
result = tcp_mtu_probe(sk);
|
|
|
|
|
if (!result) {
|
2012-05-17 07:15:34 +08:00
|
|
|
return false;
|
2008-12-06 14:48:55 +08:00
|
|
|
} else if (result > 0) {
|
|
|
|
|
sent_pkts = 1;
|
|
|
|
|
}
|
2006-03-21 09:53:41 +08:00
|
|
|
}
|
|
|
|
|
|
2016-09-20 11:39:17 +08:00
|
|
|
max_segs = tcp_tso_segs(sk, mss_now);
|
2007-03-08 04:12:44 +08:00
|
|
|
while ((skb = tcp_send_head(sk))) {
|
2005-08-17 11:43:40 +08:00
|
|
|
unsigned int limit;
|
|
|
|
|
|
2018-10-19 00:12:19 +08:00
|
|
|
if (unlikely(tp->repair) && tp->repair_queue == TCP_SEND_QUEUE) {
|
|
|
|
|
/* "skb_mstamp_ns" is used as a start point for the retransmit timer */
|
|
|
|
|
skb->skb_mstamp_ns = tp->tcp_wstamp_ns = tp->tcp_clock_cache;
|
|
|
|
|
list_move_tail(&skb->tcp_tsorted_anchor, &tp->tsorted_sent_queue);
|
2019-02-24 07:51:51 +08:00
|
|
|
tcp_init_tso_segs(skb, mss_now);
|
2018-10-19 00:12:19 +08:00
|
|
|
goto repair; /* Skip network transmission */
|
|
|
|
|
}
|
|
|
|
|
|
2017-05-16 19:24:36 +08:00
|
|
|
if (tcp_pacing_check(sk))
|
|
|
|
|
break;
|
|
|
|
|
|
2015-06-12 00:15:17 +08:00
|
|
|
tso_segs = tcp_init_tso_segs(skb, mss_now);
|
2005-07-06 06:24:38 +08:00
|
|
|
BUG_ON(!tso_segs);
|
2005-07-06 06:20:09 +08:00
|
|
|
|
2005-08-05 10:52:02 +08:00
|
|
|
cwnd_quota = tcp_cwnd_test(tp, skb);
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
if (!cwnd_quota) {
|
|
|
|
|
if (push_one == 2)
|
|
|
|
|
/* Force out a loss probe pkt. */
|
|
|
|
|
cwnd_quota = 1;
|
|
|
|
|
else
|
|
|
|
|
break;
|
|
|
|
|
}
|
2005-08-05 10:52:02 +08:00
|
|
|
|
2016-11-28 15:07:15 +08:00
|
|
|
if (unlikely(!tcp_snd_wnd_test(tp, skb, mss_now))) {
|
|
|
|
|
is_rwnd_limited = true;
|
2005-08-05 10:52:02 +08:00
|
|
|
break;
|
2016-11-28 15:07:15 +08:00
|
|
|
}
|
2005-08-05 10:52:02 +08:00
|
|
|
|
2015-05-26 23:55:28 +08:00
|
|
|
if (tso_segs == 1) {
|
2005-07-06 06:24:38 +08:00
|
|
|
if (unlikely(!tcp_nagle_test(tp, skb, mss_now,
|
|
|
|
|
(tcp_skb_is_last(sk, skb) ?
|
|
|
|
|
nonagle : TCP_NAGLE_PUSH))))
|
|
|
|
|
break;
|
|
|
|
|
} else {
|
2014-05-22 22:41:08 +08:00
|
|
|
if (!push_one &&
|
tcp: refine TSO autosizing
Commit 95bd09eb2750 ("tcp: TSO packets automatic sizing") tried to
control TSO size, but did this at the wrong place (sendmsg() time)
At sendmsg() time, we might have a pessimistic view of flow rate,
and we end up building very small skbs (with 2 MSS per skb).
This is bad because :
- It sends small TSO packets even in Slow Start where rate quickly
increases.
- It tends to make socket write queue very big, increasing tcp_ack()
processing time, but also increasing memory needs, not necessarily
accounted for, as fast clones overhead is currently ignored.
- Lower GRO efficiency and more ACK packets.
Servers with a lot of small lived connections suffer from this.
Lets instead fill skbs as much as possible (64KB of payload), but split
them at xmit time, when we have a precise idea of the flow rate.
skb split is actually quite efficient.
Patch looks bigger than necessary, because TCP Small Queue decision now
has to take place after the eventual split.
As Neal suggested, introduce a new tcp_tso_autosize() helper, so that
tcp_tso_should_defer() can be synchronized on same goal.
Rename tp->xmit_size_goal_segs to tp->gso_segs, as this variable
contains number of mss that we can put in GSO packet, and is not
related to the autosizing goal anymore.
Tested:
40 ms rtt link
nstat >/dev/null
netperf -H remote -l -2000000 -- -s 1000000
nstat | egrep "IpInReceives|IpOutRequests|TcpOutSegs|IpExtOutOctets"
Before patch :
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/s
87380 2000000 2000000 0.36 44.22
IpInReceives 600 0.0
IpOutRequests 599 0.0
TcpOutSegs 1397 0.0
IpExtOutOctets 2033249 0.0
After patch :
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
87380 2000000 2000000 0.36 44.27
IpInReceives 221 0.0
IpOutRequests 232 0.0
TcpOutSegs 1397 0.0
IpExtOutOctets 2013953 0.0
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-12-08 04:22:18 +08:00
|
|
|
tcp_tso_should_defer(sk, skb, &is_cwnd_limited,
|
2018-12-07 01:58:24 +08:00
|
|
|
&is_rwnd_limited, max_segs))
|
2005-07-06 06:24:38 +08:00
|
|
|
break;
|
|
|
|
|
}
|
2005-07-06 06:20:09 +08:00
|
|
|
|
tcp: refine TSO autosizing
Commit 95bd09eb2750 ("tcp: TSO packets automatic sizing") tried to
control TSO size, but did this at the wrong place (sendmsg() time)
At sendmsg() time, we might have a pessimistic view of flow rate,
and we end up building very small skbs (with 2 MSS per skb).
This is bad because :
- It sends small TSO packets even in Slow Start where rate quickly
increases.
- It tends to make socket write queue very big, increasing tcp_ack()
processing time, but also increasing memory needs, not necessarily
accounted for, as fast clones overhead is currently ignored.
- Lower GRO efficiency and more ACK packets.
Servers with a lot of small lived connections suffer from this.
Lets instead fill skbs as much as possible (64KB of payload), but split
them at xmit time, when we have a precise idea of the flow rate.
skb split is actually quite efficient.
Patch looks bigger than necessary, because TCP Small Queue decision now
has to take place after the eventual split.
As Neal suggested, introduce a new tcp_tso_autosize() helper, so that
tcp_tso_should_defer() can be synchronized on same goal.
Rename tp->xmit_size_goal_segs to tp->gso_segs, as this variable
contains number of mss that we can put in GSO packet, and is not
related to the autosizing goal anymore.
Tested:
40 ms rtt link
nstat >/dev/null
netperf -H remote -l -2000000 -- -s 1000000
nstat | egrep "IpInReceives|IpOutRequests|TcpOutSegs|IpExtOutOctets"
Before patch :
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/s
87380 2000000 2000000 0.36 44.22
IpInReceives 600 0.0
IpOutRequests 599 0.0
TcpOutSegs 1397 0.0
IpExtOutOctets 2033249 0.0
After patch :
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
87380 2000000 2000000 0.36 44.27
IpInReceives 221 0.0
IpOutRequests 232 0.0
TcpOutSegs 1397 0.0
IpExtOutOctets 2013953 0.0
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-12-08 04:22:18 +08:00
|
|
|
limit = mss_now;
|
2015-05-26 23:55:28 +08:00
|
|
|
if (tso_segs > 1 && !tcp_urg_mode(tp))
|
tcp: refine TSO autosizing
Commit 95bd09eb2750 ("tcp: TSO packets automatic sizing") tried to
control TSO size, but did this at the wrong place (sendmsg() time)
At sendmsg() time, we might have a pessimistic view of flow rate,
and we end up building very small skbs (with 2 MSS per skb).
This is bad because :
- It sends small TSO packets even in Slow Start where rate quickly
increases.
- It tends to make socket write queue very big, increasing tcp_ack()
processing time, but also increasing memory needs, not necessarily
accounted for, as fast clones overhead is currently ignored.
- Lower GRO efficiency and more ACK packets.
Servers with a lot of small lived connections suffer from this.
Lets instead fill skbs as much as possible (64KB of payload), but split
them at xmit time, when we have a precise idea of the flow rate.
skb split is actually quite efficient.
Patch looks bigger than necessary, because TCP Small Queue decision now
has to take place after the eventual split.
As Neal suggested, introduce a new tcp_tso_autosize() helper, so that
tcp_tso_should_defer() can be synchronized on same goal.
Rename tp->xmit_size_goal_segs to tp->gso_segs, as this variable
contains number of mss that we can put in GSO packet, and is not
related to the autosizing goal anymore.
Tested:
40 ms rtt link
nstat >/dev/null
netperf -H remote -l -2000000 -- -s 1000000
nstat | egrep "IpInReceives|IpOutRequests|TcpOutSegs|IpExtOutOctets"
Before patch :
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/s
87380 2000000 2000000 0.36 44.22
IpInReceives 600 0.0
IpOutRequests 599 0.0
TcpOutSegs 1397 0.0
IpExtOutOctets 2033249 0.0
After patch :
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
87380 2000000 2000000 0.36 44.27
IpInReceives 221 0.0
IpOutRequests 232 0.0
TcpOutSegs 1397 0.0
IpExtOutOctets 2013953 0.0
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-12-08 04:22:18 +08:00
|
|
|
limit = tcp_mss_split_point(sk, skb, mss_now,
|
|
|
|
|
min_t(unsigned int,
|
|
|
|
|
cwnd_quota,
|
|
|
|
|
max_segs),
|
|
|
|
|
nonagle);
|
|
|
|
|
|
|
|
|
|
if (skb->len > limit &&
|
2019-02-27 01:49:13 +08:00
|
|
|
unlikely(tso_fragment(sk, skb, limit, mss_now, gfp)))
|
tcp: refine TSO autosizing
Commit 95bd09eb2750 ("tcp: TSO packets automatic sizing") tried to
control TSO size, but did this at the wrong place (sendmsg() time)
At sendmsg() time, we might have a pessimistic view of flow rate,
and we end up building very small skbs (with 2 MSS per skb).
This is bad because :
- It sends small TSO packets even in Slow Start where rate quickly
increases.
- It tends to make socket write queue very big, increasing tcp_ack()
processing time, but also increasing memory needs, not necessarily
accounted for, as fast clones overhead is currently ignored.
- Lower GRO efficiency and more ACK packets.
Servers with a lot of small lived connections suffer from this.
Lets instead fill skbs as much as possible (64KB of payload), but split
them at xmit time, when we have a precise idea of the flow rate.
skb split is actually quite efficient.
Patch looks bigger than necessary, because TCP Small Queue decision now
has to take place after the eventual split.
As Neal suggested, introduce a new tcp_tso_autosize() helper, so that
tcp_tso_should_defer() can be synchronized on same goal.
Rename tp->xmit_size_goal_segs to tp->gso_segs, as this variable
contains number of mss that we can put in GSO packet, and is not
related to the autosizing goal anymore.
Tested:
40 ms rtt link
nstat >/dev/null
netperf -H remote -l -2000000 -- -s 1000000
nstat | egrep "IpInReceives|IpOutRequests|TcpOutSegs|IpExtOutOctets"
Before patch :
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/s
87380 2000000 2000000 0.36 44.22
IpInReceives 600 0.0
IpOutRequests 599 0.0
TcpOutSegs 1397 0.0
IpExtOutOctets 2033249 0.0
After patch :
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
87380 2000000 2000000 0.36 44.27
IpInReceives 221 0.0
IpOutRequests 232 0.0
TcpOutSegs 1397 0.0
IpExtOutOctets 2013953 0.0
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-12-08 04:22:18 +08:00
|
|
|
break;
|
|
|
|
|
|
2016-09-21 13:45:58 +08:00
|
|
|
if (tcp_small_queue_check(sk, skb, 0))
|
|
|
|
|
break;
|
tcp: TSQ can use a dynamic limit
When TCP Small Queues was added, we used a sysctl to limit amount of
packets queues on Qdisc/device queues for a given TCP flow.
Problem is this limit is either too big for low rates, or too small
for high rates.
Now TCP stack has rate estimation in sk->sk_pacing_rate, and TSO
auto sizing, it can better control number of packets in Qdisc/device
queues.
New limit is two packets or at least 1 to 2 ms worth of packets.
Low rates flows benefit from this patch by having even smaller
number of packets in queues, allowing for faster recovery,
better RTT estimations.
High rates flows benefit from this patch by allowing more than 2 packets
in flight as we had reports this was a limiting factor to reach line
rate. [ In particular if TX completion is delayed because of coalescing
parameters ]
Example for a single flow on 10Gbp link controlled by FQ/pacing
14 packets in flight instead of 2
$ tc -s -d qd
qdisc fq 8001: dev eth0 root refcnt 32 limit 10000p flow_limit 100p
buckets 1024 quantum 3028 initial_quantum 15140
Sent 1168459366606 bytes 771822841 pkt (dropped 0, overlimits 0
requeues 6822476)
rate 9346Mbit 771713pps backlog 953820b 14p requeues 6822476
2047 flow, 2046 inactive, 1 throttled, delay 15673 ns
2372 gc, 0 highprio, 0 retrans, 9739249 throttled, 0 flows_plimit
Note that sk_pacing_rate is currently set to twice the actual rate, but
this might be refined in the future when a flow is in congestion
avoidance.
Additional change : skb->destructor should be set to tcp_wfree().
A future patch (for linux 3.13+) might remove tcp_limit_output_bytes
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Wei Liu <wei.liu2@citrix.com>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-09-27 18:28:54 +08:00
|
|
|
|
2019-12-13 04:55:29 +08:00
|
|
|
/* Argh, we hit an empty skb(), presumably a thread
|
|
|
|
|
* is sleeping in sendmsg()/sk_stream_wait_memory().
|
|
|
|
|
* We do not want to send a pure-ack packet and have
|
|
|
|
|
* a strange looking rtx queue with empty packet(s).
|
|
|
|
|
*/
|
|
|
|
|
if (TCP_SKB_CB(skb)->end_seq == TCP_SKB_CB(skb)->seq)
|
|
|
|
|
break;
|
|
|
|
|
|
2008-12-06 14:48:55 +08:00
|
|
|
if (unlikely(tcp_transmit_skb(sk, skb, 1, gfp)))
|
2005-07-06 06:19:06 +08:00
|
|
|
break;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2012-11-15 12:03:17 +08:00
|
|
|
repair:
|
2005-07-06 06:19:06 +08:00
|
|
|
/* Advance the send_head. This one is sent out.
|
|
|
|
|
* This call will increment packets_out.
|
|
|
|
|
*/
|
2007-12-31 20:43:57 +08:00
|
|
|
tcp_event_new_data_sent(sk, skb);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2005-07-06 06:19:06 +08:00
|
|
|
tcp_minshall_update(tp, mss_now, skb);
|
2011-08-22 04:21:57 +08:00
|
|
|
sent_pkts += tcp_skb_pcount(skb);
|
2008-12-06 14:48:55 +08:00
|
|
|
|
|
|
|
|
if (push_one)
|
|
|
|
|
break;
|
2005-07-06 06:19:06 +08:00
|
|
|
}
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2016-11-28 15:07:15 +08:00
|
|
|
if (is_rwnd_limited)
|
|
|
|
|
tcp_chrono_start(sk, TCP_CHRONO_RWND_LIMITED);
|
|
|
|
|
else
|
|
|
|
|
tcp_chrono_stop(sk, TCP_CHRONO_RWND_LIMITED);
|
|
|
|
|
|
tcp: fix cwnd-limited bug for TSO deferral where we send nothing
When cwnd is not a multiple of the TSO skb size of N*MSS, we can get
into persistent scenarios where we have the following sequence:
(1) ACK for full-sized skb of N*MSS arrives
-> tcp_write_xmit() transmit full-sized skb with N*MSS
-> move pacing release time forward
-> exit tcp_write_xmit() because pacing time is in the future
(2) TSQ callback or TCP internal pacing timer fires
-> try to transmit next skb, but TSO deferral finds remainder of
available cwnd is not big enough to trigger an immediate send
now, so we defer sending until the next ACK.
(3) repeat...
So we can get into a case where we never mark ourselves as
cwnd-limited for many seconds at a time, even with
bulk/infinite-backlog senders, because:
o In case (1) above, every time in tcp_write_xmit() we have enough
cwnd to send a full-sized skb, we are not fully using the cwnd
(because cwnd is not a multiple of the TSO skb size). So every time we
send data, we are not cwnd limited, and so in the cwnd-limited
tracking code in tcp_cwnd_validate() we mark ourselves as not
cwnd-limited.
o In case (2) above, every time in tcp_write_xmit() that we try to
transmit the "remainder" of the cwnd but defer, we set the local
variable is_cwnd_limited to true, but we do not send any packets, so
sent_pkts is zero, so we don't call the cwnd-limited logic to update
tp->is_cwnd_limited.
Fixes: ca8a22634381 ("tcp: make cwnd-limited checks measurement-based, and gentler")
Reported-by: Ingemar Johansson <ingemar.s.johansson@ericsson.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20201209035759.1225145-1-ncardwell.kernel@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2020-12-09 11:57:59 +08:00
|
|
|
is_cwnd_limited |= (tcp_packets_in_flight(tp) >= tp->snd_cwnd);
|
|
|
|
|
if (likely(sent_pkts || is_cwnd_limited))
|
|
|
|
|
tcp_cwnd_validate(sk, is_cwnd_limited);
|
|
|
|
|
|
2005-07-06 06:20:09 +08:00
|
|
|
if (likely(sent_pkts)) {
|
2012-09-03 01:38:04 +08:00
|
|
|
if (tcp_in_cwnd_reduction(sk))
|
|
|
|
|
tp->prr_out += sent_pkts;
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
|
|
|
|
|
/* Send one loss probe per tail loss episode. */
|
|
|
|
|
if (push_one != 2)
|
tcp: when scheduling TLP, time of RTO should account for current ACK
Fix the TLP scheduling logic so that when scheduling a TLP probe, we
ensure that the estimated time at which an RTO would fire accounts for
the fact that ACKs indicating forward progress should push back RTO
times.
After the following fix:
df92c8394e6e ("tcp: fix xmit timer to only be reset if data ACKed/SACKed")
we had an unintentional behavior change in the following kind of
scenario: suppose the RTT variance has been very low recently. Then
suppose we send out a flight of N packets and our RTT is 100ms:
t=0: send a flight of N packets
t=100ms: receive an ACK for N-1 packets
The response before df92c8394e6e that was:
-> schedule a TLP for now + RTO_interval
The response after df92c8394e6e is:
-> schedule a TLP for t=0 + RTO_interval
Since RTO_interval = srtt + RTT_variance, this means that we have
scheduled a TLP timer at a point in the future that only accounts for
RTT_variance. If the RTT_variance term is small, this means that the
timer fires soon.
Before df92c8394e6e this would not happen, because in that code, when
we receive an ACK for a prefix of flight, we did:
1) Near the top of tcp_ack(), switch from TLP timer to RTO
at write_queue_head->paket_tx_time + RTO_interval:
if (icsk->icsk_pending == ICSK_TIME_LOSS_PROBE)
tcp_rearm_rto(sk);
2) In tcp_clean_rtx_queue(), update the RTO to now + RTO_interval:
if (flag & FLAG_ACKED) {
tcp_rearm_rto(sk);
3) In tcp_ack() after tcp_fastretrans_alert() switch from RTO
to TLP at now + RTO_interval:
if (icsk->icsk_pending == ICSK_TIME_RETRANS)
tcp_schedule_loss_probe(sk);
In df92c8394e6e we removed that 3-phase dance, and instead directly
set the TLP timer once: we set the TLP timer in cases like this to
write_queue_head->packet_tx_time + RTO_interval. So if the RTT
variance is small, then this means that this is setting the TLP timer
to fire quite soon. This means if the ACK for the tail of the flight
takes longer than an RTT to arrive (often due to delayed ACKs), then
the TLP timer fires too quickly.
Fixes: df92c8394e6e ("tcp: fix xmit timer to only be reset if data ACKed/SACKed")
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-11-18 10:06:14 +08:00
|
|
|
tcp_schedule_loss_probe(sk, false);
|
2012-05-17 07:15:34 +08:00
|
|
|
return false;
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
2017-10-06 13:21:27 +08:00
|
|
|
return !tp->packets_out && !tcp_write_queue_empty(sk);
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
}
|
|
|
|
|
|
tcp: when scheduling TLP, time of RTO should account for current ACK
Fix the TLP scheduling logic so that when scheduling a TLP probe, we
ensure that the estimated time at which an RTO would fire accounts for
the fact that ACKs indicating forward progress should push back RTO
times.
After the following fix:
df92c8394e6e ("tcp: fix xmit timer to only be reset if data ACKed/SACKed")
we had an unintentional behavior change in the following kind of
scenario: suppose the RTT variance has been very low recently. Then
suppose we send out a flight of N packets and our RTT is 100ms:
t=0: send a flight of N packets
t=100ms: receive an ACK for N-1 packets
The response before df92c8394e6e that was:
-> schedule a TLP for now + RTO_interval
The response after df92c8394e6e is:
-> schedule a TLP for t=0 + RTO_interval
Since RTO_interval = srtt + RTT_variance, this means that we have
scheduled a TLP timer at a point in the future that only accounts for
RTT_variance. If the RTT_variance term is small, this means that the
timer fires soon.
Before df92c8394e6e this would not happen, because in that code, when
we receive an ACK for a prefix of flight, we did:
1) Near the top of tcp_ack(), switch from TLP timer to RTO
at write_queue_head->paket_tx_time + RTO_interval:
if (icsk->icsk_pending == ICSK_TIME_LOSS_PROBE)
tcp_rearm_rto(sk);
2) In tcp_clean_rtx_queue(), update the RTO to now + RTO_interval:
if (flag & FLAG_ACKED) {
tcp_rearm_rto(sk);
3) In tcp_ack() after tcp_fastretrans_alert() switch from RTO
to TLP at now + RTO_interval:
if (icsk->icsk_pending == ICSK_TIME_RETRANS)
tcp_schedule_loss_probe(sk);
In df92c8394e6e we removed that 3-phase dance, and instead directly
set the TLP timer once: we set the TLP timer in cases like this to
write_queue_head->packet_tx_time + RTO_interval. So if the RTT
variance is small, then this means that this is setting the TLP timer
to fire quite soon. This means if the ACK for the tail of the flight
takes longer than an RTT to arrive (often due to delayed ACKs), then
the TLP timer fires too quickly.
Fixes: df92c8394e6e ("tcp: fix xmit timer to only be reset if data ACKed/SACKed")
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-11-18 10:06:14 +08:00
|
|
|
bool tcp_schedule_loss_probe(struct sock *sk, bool advancing_rto)
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
{
|
|
|
|
|
struct inet_connection_sock *icsk = inet_csk(sk);
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
2017-08-03 21:19:53 +08:00
|
|
|
u32 timeout, rto_delta_us;
|
2017-10-27 12:54:56 +08:00
|
|
|
int early_retrans;
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
|
|
|
|
|
/* Don't do any loss probe on a Fast Open connection before 3WHS
|
|
|
|
|
* finishes.
|
|
|
|
|
*/
|
2019-10-11 11:17:38 +08:00
|
|
|
if (rcu_access_pointer(tp->fastopen_rsk))
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
return false;
|
|
|
|
|
|
2017-10-27 12:54:56 +08:00
|
|
|
early_retrans = sock_net(sk)->ipv4.sysctl_tcp_early_retrans;
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
/* Schedule a loss probe in 2*RTT for SACK capable connections
|
tcp: allow TLP in ECN CWR
This patch enables tail loss probe in cwnd reduction (CWR) state
to detect potential losses. Prior to this patch, since the sender
uses PRR to determine the cwnd in CWR state, the combination of
CWR+PRR plus tcp_tso_should_defer() could cause unnecessary stalls
upon losses: PRR makes cwnd so gentle that tcp_tso_should_defer()
defers sending wait for more ACKs. The ACKs may not come due to
packet losses.
Disallowing TLP when there is unused cwnd had the primary effect
of disallowing TLP when there is TSO deferral, Nagle deferral,
or we hit the rwin limit. Because basically every application
write() or incoming ACK will cause us to run tcp_write_xmit()
to see if we can send more, and then if we sent something we call
tcp_schedule_loss_probe() to see if we should schedule a TLP. At
that point, there are a few common reasons why some cwnd budget
could still be unused:
(a) rwin limit
(b) nagle check
(c) TSO deferral
(d) TSQ
For (d), after the next packet tx completion the TSQ mechanism
will allow us to send more packets, so we don't really need a
TLP (in practice it shouldn't matter whether we schedule one
or not). But for (a), (b), (c) the sender won't send any more
packets until it gets another ACK. But if the whole flight was
lost, or all the ACKs were lost, then we won't get any more ACKs,
and ideally we should schedule and send a TLP to get more feedback.
In particular for a long time we have wanted some kind of timer for
TSO deferral, and at least this would give us some kind of timer
Reported-by: Steve Ibanez <sibanez@stanford.edu>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Reviewed-by: Nandita Dukkipati <nanditad@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-12-12 07:42:53 +08:00
|
|
|
* not in loss recovery, that are either limited by cwnd or application.
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
*/
|
2017-10-27 12:54:56 +08:00
|
|
|
if ((early_retrans != 3 && early_retrans != 4) ||
|
2017-01-13 14:11:39 +08:00
|
|
|
!tp->packets_out || !tcp_is_sack(tp) ||
|
tcp: allow TLP in ECN CWR
This patch enables tail loss probe in cwnd reduction (CWR) state
to detect potential losses. Prior to this patch, since the sender
uses PRR to determine the cwnd in CWR state, the combination of
CWR+PRR plus tcp_tso_should_defer() could cause unnecessary stalls
upon losses: PRR makes cwnd so gentle that tcp_tso_should_defer()
defers sending wait for more ACKs. The ACKs may not come due to
packet losses.
Disallowing TLP when there is unused cwnd had the primary effect
of disallowing TLP when there is TSO deferral, Nagle deferral,
or we hit the rwin limit. Because basically every application
write() or incoming ACK will cause us to run tcp_write_xmit()
to see if we can send more, and then if we sent something we call
tcp_schedule_loss_probe() to see if we should schedule a TLP. At
that point, there are a few common reasons why some cwnd budget
could still be unused:
(a) rwin limit
(b) nagle check
(c) TSO deferral
(d) TSQ
For (d), after the next packet tx completion the TSQ mechanism
will allow us to send more packets, so we don't really need a
TLP (in practice it shouldn't matter whether we schedule one
or not). But for (a), (b), (c) the sender won't send any more
packets until it gets another ACK. But if the whole flight was
lost, or all the ACKs were lost, then we won't get any more ACKs,
and ideally we should schedule and send a TLP to get more feedback.
In particular for a long time we have wanted some kind of timer for
TSO deferral, and at least this would give us some kind of timer
Reported-by: Steve Ibanez <sibanez@stanford.edu>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Reviewed-by: Nandita Dukkipati <nanditad@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-12-12 07:42:53 +08:00
|
|
|
(icsk->icsk_ca_state != TCP_CA_Open &&
|
|
|
|
|
icsk->icsk_ca_state != TCP_CA_CWR))
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
return false;
|
|
|
|
|
|
2017-07-20 06:41:26 +08:00
|
|
|
/* Probe timeout is 2*rtt. Add minimum RTO to account
|
2015-09-19 02:40:33 +08:00
|
|
|
* for delayed ack when there's one outstanding packet. If no RTT
|
|
|
|
|
* sample is available then probe after TCP_TIMEOUT_INIT.
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
*/
|
2017-07-20 06:41:26 +08:00
|
|
|
if (tp->srtt_us) {
|
|
|
|
|
timeout = usecs_to_jiffies(tp->srtt_us >> 2);
|
|
|
|
|
if (tp->packets_out == 1)
|
|
|
|
|
timeout += TCP_RTO_MIN;
|
|
|
|
|
else
|
|
|
|
|
timeout += TCP_TIMEOUT_MIN;
|
|
|
|
|
} else {
|
|
|
|
|
timeout = TCP_TIMEOUT_INIT;
|
|
|
|
|
}
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
|
2017-08-03 21:19:53 +08:00
|
|
|
/* If the RTO formula yields an earlier time, then use that time. */
|
tcp: when scheduling TLP, time of RTO should account for current ACK
Fix the TLP scheduling logic so that when scheduling a TLP probe, we
ensure that the estimated time at which an RTO would fire accounts for
the fact that ACKs indicating forward progress should push back RTO
times.
After the following fix:
df92c8394e6e ("tcp: fix xmit timer to only be reset if data ACKed/SACKed")
we had an unintentional behavior change in the following kind of
scenario: suppose the RTT variance has been very low recently. Then
suppose we send out a flight of N packets and our RTT is 100ms:
t=0: send a flight of N packets
t=100ms: receive an ACK for N-1 packets
The response before df92c8394e6e that was:
-> schedule a TLP for now + RTO_interval
The response after df92c8394e6e is:
-> schedule a TLP for t=0 + RTO_interval
Since RTO_interval = srtt + RTT_variance, this means that we have
scheduled a TLP timer at a point in the future that only accounts for
RTT_variance. If the RTT_variance term is small, this means that the
timer fires soon.
Before df92c8394e6e this would not happen, because in that code, when
we receive an ACK for a prefix of flight, we did:
1) Near the top of tcp_ack(), switch from TLP timer to RTO
at write_queue_head->paket_tx_time + RTO_interval:
if (icsk->icsk_pending == ICSK_TIME_LOSS_PROBE)
tcp_rearm_rto(sk);
2) In tcp_clean_rtx_queue(), update the RTO to now + RTO_interval:
if (flag & FLAG_ACKED) {
tcp_rearm_rto(sk);
3) In tcp_ack() after tcp_fastretrans_alert() switch from RTO
to TLP at now + RTO_interval:
if (icsk->icsk_pending == ICSK_TIME_RETRANS)
tcp_schedule_loss_probe(sk);
In df92c8394e6e we removed that 3-phase dance, and instead directly
set the TLP timer once: we set the TLP timer in cases like this to
write_queue_head->packet_tx_time + RTO_interval. So if the RTT
variance is small, then this means that this is setting the TLP timer
to fire quite soon. This means if the ACK for the tail of the flight
takes longer than an RTT to arrive (often due to delayed ACKs), then
the TLP timer fires too quickly.
Fixes: df92c8394e6e ("tcp: fix xmit timer to only be reset if data ACKed/SACKed")
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-11-18 10:06:14 +08:00
|
|
|
rto_delta_us = advancing_rto ?
|
|
|
|
|
jiffies_to_usecs(inet_csk(sk)->icsk_rto) :
|
|
|
|
|
tcp_rto_delta_us(sk); /* How far in future is RTO? */
|
2017-08-03 21:19:53 +08:00
|
|
|
if (rto_delta_us > 0)
|
|
|
|
|
timeout = min_t(u32, timeout, usecs_to_jiffies(rto_delta_us));
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
|
2020-05-05 02:27:49 +08:00
|
|
|
tcp_reset_xmit_timer(sk, ICSK_TIME_LOSS_PROBE, timeout, TCP_RTO_MAX);
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
return true;
|
|
|
|
|
}
|
|
|
|
|
|
2014-04-21 08:58:17 +08:00
|
|
|
/* Thanks to skb fast clones, we can detect if a prior transmit of
|
|
|
|
|
* a packet is still in a qdisc or driver queue.
|
|
|
|
|
* In this case, there is very little point doing a retransmit !
|
|
|
|
|
*/
|
|
|
|
|
static bool skb_still_in_host_queue(const struct sock *sk,
|
|
|
|
|
const struct sk_buff *skb)
|
|
|
|
|
{
|
2014-10-31 01:32:34 +08:00
|
|
|
if (unlikely(skb_fclone_busy(sk, skb))) {
|
2016-04-30 05:16:47 +08:00
|
|
|
NET_INC_STATS(sock_net(sk),
|
|
|
|
|
LINUX_MIB_TCPSPURIOUS_RTX_HOSTQUEUES);
|
2014-04-21 08:58:17 +08:00
|
|
|
return true;
|
|
|
|
|
}
|
|
|
|
|
return false;
|
|
|
|
|
}
|
|
|
|
|
|
2015-08-13 02:18:19 +08:00
|
|
|
/* When probe timeout (PTO) fires, try send a new segment if possible, else
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
* retransmit the last segment.
|
|
|
|
|
*/
|
|
|
|
|
void tcp_send_loss_probe(struct sock *sk)
|
|
|
|
|
{
|
2013-03-11 18:00:44 +08:00
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
struct sk_buff *skb;
|
|
|
|
|
int pcount;
|
|
|
|
|
int mss = tcp_current_mss(sk);
|
|
|
|
|
|
2020-07-24 03:00:06 +08:00
|
|
|
/* At most one outstanding TLP */
|
|
|
|
|
if (tp->tlp_high_seq)
|
|
|
|
|
goto rearm_timer;
|
|
|
|
|
|
|
|
|
|
tp->tlp_retrans = 0;
|
2015-08-13 02:18:19 +08:00
|
|
|
skb = tcp_send_head(sk);
|
2017-10-06 13:21:27 +08:00
|
|
|
if (skb && tcp_snd_wnd_test(tp, skb, mss)) {
|
|
|
|
|
pcount = tp->packets_out;
|
|
|
|
|
tcp_write_xmit(sk, mss, TCP_NAGLE_OFF, 2, GFP_ATOMIC);
|
|
|
|
|
if (tp->packets_out > pcount)
|
|
|
|
|
goto probe_sent;
|
|
|
|
|
goto rearm_timer;
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
}
|
2017-10-06 13:21:27 +08:00
|
|
|
skb = skb_rb_last(&sk->tcp_rtx_queue);
|
2018-12-06 06:38:38 +08:00
|
|
|
if (unlikely(!skb)) {
|
|
|
|
|
WARN_ONCE(tp->packets_out,
|
|
|
|
|
"invalid inflight: %u state %u cwnd %u mss %d\n",
|
|
|
|
|
tp->packets_out, sk->sk_state, tp->snd_cwnd, mss);
|
|
|
|
|
inet_csk(sk)->icsk_pending = 0;
|
|
|
|
|
return;
|
|
|
|
|
}
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
|
2014-04-21 08:58:17 +08:00
|
|
|
if (skb_still_in_host_queue(sk, skb))
|
|
|
|
|
goto rearm_timer;
|
|
|
|
|
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
pcount = tcp_skb_pcount(skb);
|
|
|
|
|
if (WARN_ON(!pcount))
|
|
|
|
|
goto rearm_timer;
|
|
|
|
|
|
|
|
|
|
if ((pcount > 1) && (skb->len > (pcount - 1) * mss)) {
|
2017-10-06 13:21:27 +08:00
|
|
|
if (unlikely(tcp_fragment(sk, TCP_FRAG_IN_RTX_QUEUE, skb,
|
|
|
|
|
(pcount - 1) * mss, mss,
|
2014-06-06 22:32:37 +08:00
|
|
|
GFP_ATOMIC)))
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
goto rearm_timer;
|
2017-10-06 13:21:27 +08:00
|
|
|
skb = skb_rb_next(skb);
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (WARN_ON(!skb || !tcp_skb_pcount(skb)))
|
|
|
|
|
goto rearm_timer;
|
|
|
|
|
|
tcp-tso: do not split TSO packets at retransmit time
Linux TCP stack painfully segments all TSO/GSO packets before retransmits.
This was fine back in the days when TSO/GSO were emerging, with their
bugs, but we believe the dark age is over.
Keeping big packets in write queues, but also in stack traversal
has a lot of benefits.
- Less memory overhead, because write queues have less skbs
- Less cpu overhead at ACK processing.
- Better SACK processing, as lot of studies mentioned how
awful linux was at this ;)
- Less cpu overhead to send the rtx packets
(IP stack traversal, netfilter traversal, drivers...)
- Better latencies in presence of losses.
- Smaller spikes in fq like packet schedulers, as retransmits
are not constrained by TCP Small Queues.
1 % packet losses are common today, and at 100Gbit speeds, this
translates to ~80,000 losses per second.
Losses are often correlated, and we see many retransmit events
leading to 1-MSS train of packets, at the time hosts are already
under stress.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-22 01:55:23 +08:00
|
|
|
if (__tcp_retransmit_skb(sk, skb, 1))
|
2015-08-13 02:18:19 +08:00
|
|
|
goto rearm_timer;
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
|
2020-07-24 03:00:06 +08:00
|
|
|
tp->tlp_retrans = 1;
|
|
|
|
|
|
|
|
|
|
probe_sent:
|
2013-03-11 18:00:44 +08:00
|
|
|
/* Record snd_nxt for loss detection. */
|
2015-08-13 02:18:19 +08:00
|
|
|
tp->tlp_high_seq = tp->snd_nxt;
|
2013-03-11 18:00:44 +08:00
|
|
|
|
2016-04-30 05:16:47 +08:00
|
|
|
NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPLOSSPROBES);
|
2015-08-13 02:18:19 +08:00
|
|
|
/* Reset s.t. tcp_rearm_rto will restart timer from now */
|
|
|
|
|
inet_csk(sk)->icsk_pending = 0;
|
tcp: Tail loss probe (TLP)
This patch series implement the Tail loss probe (TLP) algorithm described
in http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01. The
first patch implements the basic algorithm.
TLP's goal is to reduce tail latency of short transactions. It achieves
this by converting retransmission timeouts (RTOs) occuring due
to tail losses (losses at end of transactions) into fast recovery.
TLP transmits one packet in two round-trips when a connection is in
Open state and isn't receiving any ACKs. The transmitted packet, aka
loss probe, can be either new or a retransmission. When there is tail
loss, the ACK from a loss probe triggers FACK/early-retransmit based
fast recovery, thus avoiding a costly RTO. In the absence of loss,
there is no change in the connection state.
PTO stands for probe timeout. It is a timer event indicating
that an ACK is overdue and triggers a loss probe packet. The PTO value
is set to max(2*SRTT, 10ms) and is adjusted to account for delayed
ACK timer when there is only one oustanding packet.
TLP Algorithm
On transmission of new data in Open state:
-> packets_out > 1: schedule PTO in max(2*SRTT, 10ms).
-> packets_out == 1: schedule PTO in max(2*RTT, 1.5*RTT + 200ms)
-> PTO = min(PTO, RTO)
Conditions for scheduling PTO:
-> Connection is in Open state.
-> Connection is either cwnd limited or no new data to send.
-> Number of probes per tail loss episode is limited to one.
-> Connection is SACK enabled.
When PTO fires:
new_segment_exists:
-> transmit new segment.
-> packets_out++. cwnd remains same.
no_new_packet:
-> retransmit the last segment.
Its ACK triggers FACK or early retransmit based recovery.
ACK path:
-> rearm RTO at start of ACK processing.
-> reschedule PTO if need be.
In addition, the patch includes a small variation to the Early Retransmit
(ER) algorithm, such that ER and TLP together can in principle recover any
N-degree of tail loss through fast recovery. TLP is controlled by the same
sysctl as ER, tcp_early_retrans sysctl.
tcp_early_retrans==0; disables TLP and ER.
==1; enables RFC5827 ER.
==2; delayed ER.
==3; TLP and delayed ER. [DEFAULT]
==4; TLP only.
The TLP patch series have been extensively tested on Google Web servers.
It is most effective for short Web trasactions, where it reduced RTOs by 15%
and improved HTTP response time (average by 6%, 99th percentile by 10%).
The transmitted probes account for <0.5% of the overall transmissions.
Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-11 18:00:43 +08:00
|
|
|
rearm_timer:
|
2015-08-13 02:18:18 +08:00
|
|
|
tcp_rearm_rto(sk);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
|
2005-07-06 06:18:51 +08:00
|
|
|
/* Push out any pending frames which were held back due to
|
|
|
|
|
* TCP_CORK or attempt at coalescing tiny packets.
|
|
|
|
|
* The socket must be locked by the caller.
|
|
|
|
|
*/
|
[TCP]: Sed magic converts func(sk, tp, ...) -> func(sk, ...)
This is (mostly) automated change using magic:
sed -e '/struct sock \*sk/ N' -e '/struct sock \*sk/ N'
-e '/struct sock \*sk/ N' -e '/struct sock \*sk/ N'
-e 's|struct sock \*sk,[\n\t ]*struct tcp_sock \*tp\([^{]*\n{\n\)|
struct sock \*sk\1\tstruct tcp_sock *tp = tcp_sk(sk);\n|g'
-e 's|struct sock \*sk, struct tcp_sock \*tp|
struct sock \*sk|g' -e 's|sk, tp\([^-]\)|sk\1|g'
Fixed four unused variable (tp) warnings that were introduced.
In addition, manually added newlines after local variables and
tweaked function arguments positioning.
$ gcc --version
gcc (GCC) 4.1.1 20060525 (Red Hat 4.1.1-1)
...
$ codiff -fV built-in.o.old built-in.o.new
net/ipv4/route.c:
rt_cache_flush | +14
1 function changed, 14 bytes added
net/ipv4/tcp.c:
tcp_setsockopt | -5
tcp_sendpage | -25
tcp_sendmsg | -16
3 functions changed, 46 bytes removed
net/ipv4/tcp_input.c:
tcp_try_undo_recovery | +3
tcp_try_undo_dsack | +2
tcp_mark_head_lost | -12
tcp_ack | -15
tcp_event_data_recv | -32
tcp_rcv_state_process | -10
tcp_rcv_established | +1
7 functions changed, 6 bytes added, 69 bytes removed, diff: -63
net/ipv4/tcp_output.c:
update_send_head | -9
tcp_transmit_skb | +19
tcp_cwnd_validate | +1
tcp_write_wakeup | -17
__tcp_push_pending_frames | -25
tcp_push_one | -8
tcp_send_fin | -4
7 functions changed, 20 bytes added, 63 bytes removed, diff: -43
built-in.o.new:
18 functions changed, 40 bytes added, 178 bytes removed, diff: -138
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
2007-04-21 13:18:02 +08:00
|
|
|
void __tcp_push_pending_frames(struct sock *sk, unsigned int cur_mss,
|
|
|
|
|
int nonagle)
|
2005-07-06 06:18:51 +08:00
|
|
|
{
|
2008-12-06 14:43:56 +08:00
|
|
|
/* If we are closed, the bytes will have to remain here.
|
|
|
|
|
* In time closedown will finish, we empty the write queue and
|
|
|
|
|
* all will be happy.
|
|
|
|
|
*/
|
|
|
|
|
if (unlikely(sk->sk_state == TCP_CLOSE))
|
|
|
|
|
return;
|
|
|
|
|
|
2012-08-01 07:44:14 +08:00
|
|
|
if (tcp_write_xmit(sk, cur_mss, nonagle, 0,
|
2015-12-01 00:57:28 +08:00
|
|
|
sk_gfp_mask(sk, GFP_ATOMIC)))
|
2008-12-06 14:43:56 +08:00
|
|
|
tcp_check_probe_timer(sk);
|
2005-07-06 06:18:51 +08:00
|
|
|
}
|
|
|
|
|
|
2005-07-06 06:24:38 +08:00
|
|
|
/* Send _single_ skb sitting at the send head. This function requires
|
|
|
|
|
* true push pending frames to setup probe timer etc.
|
|
|
|
|
*/
|
|
|
|
|
void tcp_push_one(struct sock *sk, unsigned int mss_now)
|
|
|
|
|
{
|
2007-03-08 04:12:44 +08:00
|
|
|
struct sk_buff *skb = tcp_send_head(sk);
|
2005-07-06 06:24:38 +08:00
|
|
|
|
|
|
|
|
BUG_ON(!skb || skb->len < mss_now);
|
|
|
|
|
|
2008-12-06 14:48:55 +08:00
|
|
|
tcp_write_xmit(sk, mss_now, TCP_NAGLE_PUSH, 1, sk->sk_allocation);
|
2005-07-06 06:24:38 +08:00
|
|
|
}
|
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/* This function returns the amount that we can raise the
|
|
|
|
|
* usable window based on the following constraints
|
2007-02-09 22:24:47 +08:00
|
|
|
*
|
2005-04-17 06:20:36 +08:00
|
|
|
* 1. The window can never be shrunk once it is offered (RFC 793)
|
|
|
|
|
* 2. We limit memory per socket
|
|
|
|
|
*
|
|
|
|
|
* RFC 1122:
|
|
|
|
|
* "the suggested [SWS] avoidance algorithm for the receiver is to keep
|
|
|
|
|
* RECV.NEXT + RCV.WIN fixed until:
|
|
|
|
|
* RCV.BUFF - RCV.USER - RCV.WINDOW >= min(1/2 RCV.BUFF, MSS)"
|
|
|
|
|
*
|
|
|
|
|
* i.e. don't raise the right edge of the window until you can raise
|
|
|
|
|
* it at least MSS bytes.
|
|
|
|
|
*
|
|
|
|
|
* Unfortunately, the recommended algorithm breaks header prediction,
|
|
|
|
|
* since header prediction assumes th->window stays fixed.
|
|
|
|
|
*
|
|
|
|
|
* Strictly speaking, keeping th->window fixed violates the receiver
|
|
|
|
|
* side SWS prevention criteria. The problem is that under this rule
|
|
|
|
|
* a stream of single byte packets will cause the right side of the
|
|
|
|
|
* window to always advance by a single byte.
|
2007-02-09 22:24:47 +08:00
|
|
|
*
|
2005-04-17 06:20:36 +08:00
|
|
|
* Of course, if the sender implements sender side SWS prevention
|
|
|
|
|
* then this will not be a problem.
|
2007-02-09 22:24:47 +08:00
|
|
|
*
|
2005-04-17 06:20:36 +08:00
|
|
|
* BSD seems to make the following compromise:
|
2007-02-09 22:24:47 +08:00
|
|
|
*
|
2005-04-17 06:20:36 +08:00
|
|
|
* If the free space is less than the 1/4 of the maximum
|
|
|
|
|
* space available and the free space is less than 1/2 mss,
|
|
|
|
|
* then set the window to 0.
|
|
|
|
|
* [ Actually, bsd uses MSS and 1/4 of maximal _window_ ]
|
|
|
|
|
* Otherwise, just prevent the window from shrinking
|
|
|
|
|
* and from being larger than the largest representable value.
|
|
|
|
|
*
|
|
|
|
|
* This prevents incremental opening of the window in the regime
|
|
|
|
|
* where TCP is limited by the speed of the reader side taking
|
|
|
|
|
* data out of the TCP receive queue. It does nothing about
|
|
|
|
|
* those cases where the window is constrained on the sender side
|
|
|
|
|
* because the pipeline is full.
|
|
|
|
|
*
|
|
|
|
|
* BSD also seems to "accidentally" limit itself to windows that are a
|
|
|
|
|
* multiple of MSS, at least until the free space gets quite small.
|
|
|
|
|
* This would appear to be a side effect of the mbuf implementation.
|
|
|
|
|
* Combining these two algorithms results in the observed behavior
|
|
|
|
|
* of having a fixed window size at almost all times.
|
|
|
|
|
*
|
|
|
|
|
* Below we obtain similar behavior by forcing the offered window to
|
|
|
|
|
* a multiple of the mss when it is feasible to do so.
|
|
|
|
|
*
|
|
|
|
|
* Note, we don't "adjust" for TIMESTAMP or SACK option bytes.
|
|
|
|
|
* Regular options like TIMESTAMP are taken into account.
|
|
|
|
|
*/
|
|
|
|
|
u32 __tcp_select_window(struct sock *sk)
|
|
|
|
|
{
|
2005-08-10 11:10:42 +08:00
|
|
|
struct inet_connection_sock *icsk = inet_csk(sk);
|
2005-04-17 06:20:36 +08:00
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
2005-11-11 09:13:47 +08:00
|
|
|
/* MSS for the peer's data. Previous versions used mss_clamp
|
2005-04-17 06:20:36 +08:00
|
|
|
* here. I don't know if the value based on our guesses
|
|
|
|
|
* of peer's MSS is better for the performance. It's more correct
|
|
|
|
|
* but may be worse for the performance because of rcv_mss
|
|
|
|
|
* fluctuations. --SAW 1998/11/1
|
|
|
|
|
*/
|
2005-08-10 11:10:42 +08:00
|
|
|
int mss = icsk->icsk_ack.rcv_mss;
|
2005-04-17 06:20:36 +08:00
|
|
|
int free_space = tcp_space(sk);
|
tcp: use zero-window when free_space is low
Currently the kernel tries to announce a zero window when free_space
is below the current receiver mss estimate.
When a sender is transmitting small packets and reader consumes data
slowly (or not at all), receiver might be unable to shrink the receive
win because
a) we cannot withdraw already-commited receive window, and,
b) we have to round the current rwin up to a multiple of the wscale
factor, else we would shrink the current window.
This causes the receive buffer to fill up until the rmem limit is hit.
When this happens, we start dropping packets.
Moreover, tcp_clamp_window may continue to grow sk_rcvbuf towards rmem[2]
even if socket is not being read from.
As we cannot avoid the "current_win is rounded up to multiple of mss"
issue [we would violate a) above] at least try to prevent the receive buf
growth towards tcp_rmem[2] limit by attempting to move to zero-window
announcement when free_space becomes less than 1/16 of the current
allowed receive buffer maximum. If tcp_rmem[2] is large, this will
increase our chances to get a zero-window announcement out in time.
Reproducer:
On server:
$ nc -l -p 12345
<suspend it: CTRL-Z>
Client:
#!/usr/bin/env python
import socket
import time
sock = socket.socket()
sock.setsockopt(socket.IPPROTO_TCP, socket.TCP_NODELAY, 1)
sock.connect(("192.168.4.1", 12345));
while True:
sock.send('A' * 23)
time.sleep(0.005)
socket buffer on server-side will grow until tcp_rmem[2] is hit,
at which point the client rexmits data until -EDTIMEOUT:
tcp_data_queue invokes tcp_try_rmem_schedule which will call
tcp_prune_queue which calls tcp_clamp_window(). And that function will
grow sk->sk_rcvbuf up until it eventually hits tcp_rmem[2].
Thanks to Eric Dumazet for running regression tests.
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Tested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-19 19:51:10 +08:00
|
|
|
int allowed_space = tcp_full_space(sk);
|
2020-04-24 18:31:50 +08:00
|
|
|
int full_space, window;
|
|
|
|
|
|
|
|
|
|
if (sk_is_mptcp(sk))
|
|
|
|
|
mptcp_space(sk, &free_space, &allowed_space);
|
|
|
|
|
|
|
|
|
|
full_space = min_t(int, tp->window_clamp, allowed_space);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2017-02-02 00:33:53 +08:00
|
|
|
if (unlikely(mss > full_space)) {
|
2007-02-09 22:24:47 +08:00
|
|
|
mss = full_space;
|
2017-02-02 00:33:53 +08:00
|
|
|
if (mss <= 0)
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
2007-12-21 13:48:32 +08:00
|
|
|
if (free_space < (full_space >> 1)) {
|
2005-08-10 11:10:42 +08:00
|
|
|
icsk->icsk_ack.quick = 0;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2015-05-16 03:39:27 +08:00
|
|
|
if (tcp_under_memory_pressure(sk))
|
2008-01-01 06:57:14 +08:00
|
|
|
tp->rcv_ssthresh = min(tp->rcv_ssthresh,
|
|
|
|
|
4U * tp->advmss);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
tcp: use zero-window when free_space is low
Currently the kernel tries to announce a zero window when free_space
is below the current receiver mss estimate.
When a sender is transmitting small packets and reader consumes data
slowly (or not at all), receiver might be unable to shrink the receive
win because
a) we cannot withdraw already-commited receive window, and,
b) we have to round the current rwin up to a multiple of the wscale
factor, else we would shrink the current window.
This causes the receive buffer to fill up until the rmem limit is hit.
When this happens, we start dropping packets.
Moreover, tcp_clamp_window may continue to grow sk_rcvbuf towards rmem[2]
even if socket is not being read from.
As we cannot avoid the "current_win is rounded up to multiple of mss"
issue [we would violate a) above] at least try to prevent the receive buf
growth towards tcp_rmem[2] limit by attempting to move to zero-window
announcement when free_space becomes less than 1/16 of the current
allowed receive buffer maximum. If tcp_rmem[2] is large, this will
increase our chances to get a zero-window announcement out in time.
Reproducer:
On server:
$ nc -l -p 12345
<suspend it: CTRL-Z>
Client:
#!/usr/bin/env python
import socket
import time
sock = socket.socket()
sock.setsockopt(socket.IPPROTO_TCP, socket.TCP_NODELAY, 1)
sock.connect(("192.168.4.1", 12345));
while True:
sock.send('A' * 23)
time.sleep(0.005)
socket buffer on server-side will grow until tcp_rmem[2] is hit,
at which point the client rexmits data until -EDTIMEOUT:
tcp_data_queue invokes tcp_try_rmem_schedule which will call
tcp_prune_queue which calls tcp_clamp_window(). And that function will
grow sk->sk_rcvbuf up until it eventually hits tcp_rmem[2].
Thanks to Eric Dumazet for running regression tests.
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Tested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-19 19:51:10 +08:00
|
|
|
/* free_space might become our new window, make sure we don't
|
|
|
|
|
* increase it due to wscale.
|
|
|
|
|
*/
|
|
|
|
|
free_space = round_down(free_space, 1 << tp->rx_opt.rcv_wscale);
|
|
|
|
|
|
|
|
|
|
/* if free space is less than mss estimate, or is below 1/16th
|
|
|
|
|
* of the maximum allowed, try to move to zero-window, else
|
|
|
|
|
* tcp_clamp_window() will grow rcv buf up to tcp_rmem[2], and
|
|
|
|
|
* new incoming data is dropped due to memory limits.
|
|
|
|
|
* With large window, mss test triggers way too late in order
|
|
|
|
|
* to announce zero window in time before rmem limit kicks in.
|
|
|
|
|
*/
|
|
|
|
|
if (free_space < (allowed_space >> 4) || free_space < mss)
|
2005-04-17 06:20:36 +08:00
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (free_space > tp->rcv_ssthresh)
|
|
|
|
|
free_space = tp->rcv_ssthresh;
|
|
|
|
|
|
|
|
|
|
/* Don't do rounding if we are using window scaling, since the
|
|
|
|
|
* scaled window will not line up with the MSS boundary anyway.
|
|
|
|
|
*/
|
|
|
|
|
if (tp->rx_opt.rcv_wscale) {
|
|
|
|
|
window = free_space;
|
|
|
|
|
|
|
|
|
|
/* Advertise enough space so that it won't get scaled away.
|
|
|
|
|
* Import case: prevent zero window announcement if
|
|
|
|
|
* 1<<rcv_wscale > mss.
|
|
|
|
|
*/
|
2017-03-30 06:49:19 +08:00
|
|
|
window = ALIGN(window, (1 << tp->rx_opt.rcv_wscale));
|
2005-04-17 06:20:36 +08:00
|
|
|
} else {
|
2017-03-30 06:49:19 +08:00
|
|
|
window = tp->rcv_wnd;
|
2005-04-17 06:20:36 +08:00
|
|
|
/* Get the largest window that is a nice multiple of mss.
|
|
|
|
|
* Window clamp already applied above.
|
|
|
|
|
* If our current window offering is within 1 mss of the
|
|
|
|
|
* free space we just keep it. This prevents the divide
|
|
|
|
|
* and multiply from happening most of the time.
|
|
|
|
|
* We also don't do any window rounding when the free space
|
|
|
|
|
* is too small.
|
|
|
|
|
*/
|
|
|
|
|
if (window <= free_space - mss || window > free_space)
|
2017-03-30 06:49:19 +08:00
|
|
|
window = rounddown(free_space, mss);
|
2007-04-03 04:56:32 +08:00
|
|
|
else if (mss == full_space &&
|
2007-12-21 13:48:32 +08:00
|
|
|
free_space > window + (full_space >> 1))
|
2007-04-03 04:56:32 +08:00
|
|
|
window = free_space;
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return window;
|
|
|
|
|
}
|
|
|
|
|
|
tcp: Merge tx_flags and tskey in tcp_shifted_skb
After receiving sacks, tcp_shifted_skb() will collapse
skbs if possible. tx_flags and tskey also have to be
merged.
This patch reuses the tcp_skb_collapse_tstamp() to handle
them.
BPF Output Before:
~~~~~
<no-output-due-to-missing-tstamp-event>
BPF Output After:
~~~~~
<...>-2024 [007] d.s. 88.644374: : ee_data:14599
Packetdrill Script:
~~~~~
+0 `sysctl -q -w net.ipv4.tcp_min_tso_segs=10`
+0 `sysctl -q -w net.ipv4.tcp_no_metrics_save=1`
+0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
0.100 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 7>
0.100 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 7>
0.200 < . 1:1(0) ack 1 win 257
0.200 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_NODELAY, [1], 4) = 0
0.200 write(4, ..., 1460) = 1460
+0 setsockopt(4, SOL_SOCKET, 37, [2688], 4) = 0
0.200 write(4, ..., 13140) = 13140
0.200 > P. 1:1461(1460) ack 1
0.200 > . 1461:8761(7300) ack 1
0.200 > P. 8761:14601(5840) ack 1
0.300 < . 1:1(0) ack 1 win 257 <sack 1461:14601,nop,nop>
0.300 > P. 1:1461(1460) ack 1
0.400 < . 1:1(0) ack 14601 win 257
0.400 close(4) = 0
0.400 > F. 14601:14601(0) ack 1
0.500 < F. 1:1(0) ack 14602 win 257
0.500 > . 14602:14602(0) ack 2
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Soheil Hassas Yeganeh <soheil@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Tested-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-20 13:39:29 +08:00
|
|
|
void tcp_skb_collapse_tstamp(struct sk_buff *skb,
|
|
|
|
|
const struct sk_buff *next_skb)
|
tcp: Merge tx_flags and tskey in tcp_collapse_retrans
If two skbs are merged/collapsed during retransmission, the current
logic does not merge the tx_flags and tskey. The end result is
the SCM_TSTAMP_ACK timestamp could be missing for a packet.
The patch:
1. Merge the tx_flags
2. Overwrite the prev_skb's tskey with the next_skb's tskey
BPF Output Before:
~~~~~~
<no-output-due-to-missing-tstamp-event>
BPF Output After:
~~~~~~
packetdrill-2092 [001] d.s. 453.998486: : ee_data:1459
Packetdrill Script:
~~~~~~
+0 `sysctl -q -w net.ipv4.tcp_min_tso_segs=10`
+0 `sysctl -q -w net.ipv4.tcp_no_metrics_save=1`
+0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
0.100 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 7>
0.100 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 7>
0.200 < . 1:1(0) ack 1 win 257
0.200 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_NODELAY, [1], 4) = 0
0.200 write(4, ..., 730) = 730
+0 setsockopt(4, SOL_SOCKET, 37, [2688], 4) = 0
0.200 write(4, ..., 730) = 730
+0 setsockopt(4, SOL_SOCKET, 37, [2176], 4) = 0
0.200 write(4, ..., 11680) = 11680
+0 setsockopt(4, SOL_SOCKET, 37, [2688], 4) = 0
0.200 > P. 1:731(730) ack 1
0.200 > P. 731:1461(730) ack 1
0.200 > . 1461:8761(7300) ack 1
0.200 > P. 8761:13141(4380) ack 1
0.300 < . 1:1(0) ack 1 win 257 <sack 1461:2921,nop,nop>
0.300 < . 1:1(0) ack 1 win 257 <sack 1461:4381,nop,nop>
0.300 < . 1:1(0) ack 1 win 257 <sack 1461:5841,nop,nop>
0.300 > P. 1:1461(1460) ack 1
0.400 < . 1:1(0) ack 13141 win 257
0.400 close(4) = 0
0.400 > F. 13141:13141(0) ack 1
0.500 < F. 1:1(0) ack 13142 win 257
0.500 > . 13142:13142(0) ack 2
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Soheil Hassas Yeganeh <soheil@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Tested-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-20 13:39:28 +08:00
|
|
|
{
|
2016-04-28 11:39:01 +08:00
|
|
|
if (unlikely(tcp_has_tx_tstamp(next_skb))) {
|
|
|
|
|
const struct skb_shared_info *next_shinfo =
|
|
|
|
|
skb_shinfo(next_skb);
|
tcp: Merge tx_flags and tskey in tcp_collapse_retrans
If two skbs are merged/collapsed during retransmission, the current
logic does not merge the tx_flags and tskey. The end result is
the SCM_TSTAMP_ACK timestamp could be missing for a packet.
The patch:
1. Merge the tx_flags
2. Overwrite the prev_skb's tskey with the next_skb's tskey
BPF Output Before:
~~~~~~
<no-output-due-to-missing-tstamp-event>
BPF Output After:
~~~~~~
packetdrill-2092 [001] d.s. 453.998486: : ee_data:1459
Packetdrill Script:
~~~~~~
+0 `sysctl -q -w net.ipv4.tcp_min_tso_segs=10`
+0 `sysctl -q -w net.ipv4.tcp_no_metrics_save=1`
+0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
0.100 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 7>
0.100 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 7>
0.200 < . 1:1(0) ack 1 win 257
0.200 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_NODELAY, [1], 4) = 0
0.200 write(4, ..., 730) = 730
+0 setsockopt(4, SOL_SOCKET, 37, [2688], 4) = 0
0.200 write(4, ..., 730) = 730
+0 setsockopt(4, SOL_SOCKET, 37, [2176], 4) = 0
0.200 write(4, ..., 11680) = 11680
+0 setsockopt(4, SOL_SOCKET, 37, [2688], 4) = 0
0.200 > P. 1:731(730) ack 1
0.200 > P. 731:1461(730) ack 1
0.200 > . 1461:8761(7300) ack 1
0.200 > P. 8761:13141(4380) ack 1
0.300 < . 1:1(0) ack 1 win 257 <sack 1461:2921,nop,nop>
0.300 < . 1:1(0) ack 1 win 257 <sack 1461:4381,nop,nop>
0.300 < . 1:1(0) ack 1 win 257 <sack 1461:5841,nop,nop>
0.300 > P. 1:1461(1460) ack 1
0.400 < . 1:1(0) ack 13141 win 257
0.400 close(4) = 0
0.400 > F. 13141:13141(0) ack 1
0.500 < F. 1:1(0) ack 13142 win 257
0.500 > . 13142:13142(0) ack 2
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Soheil Hassas Yeganeh <soheil@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Tested-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-20 13:39:28 +08:00
|
|
|
struct skb_shared_info *shinfo = skb_shinfo(skb);
|
|
|
|
|
|
2016-04-28 11:39:01 +08:00
|
|
|
shinfo->tx_flags |= next_shinfo->tx_flags & SKBTX_ANY_TSTAMP;
|
tcp: Merge tx_flags and tskey in tcp_collapse_retrans
If two skbs are merged/collapsed during retransmission, the current
logic does not merge the tx_flags and tskey. The end result is
the SCM_TSTAMP_ACK timestamp could be missing for a packet.
The patch:
1. Merge the tx_flags
2. Overwrite the prev_skb's tskey with the next_skb's tskey
BPF Output Before:
~~~~~~
<no-output-due-to-missing-tstamp-event>
BPF Output After:
~~~~~~
packetdrill-2092 [001] d.s. 453.998486: : ee_data:1459
Packetdrill Script:
~~~~~~
+0 `sysctl -q -w net.ipv4.tcp_min_tso_segs=10`
+0 `sysctl -q -w net.ipv4.tcp_no_metrics_save=1`
+0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
0.100 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 7>
0.100 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 7>
0.200 < . 1:1(0) ack 1 win 257
0.200 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_NODELAY, [1], 4) = 0
0.200 write(4, ..., 730) = 730
+0 setsockopt(4, SOL_SOCKET, 37, [2688], 4) = 0
0.200 write(4, ..., 730) = 730
+0 setsockopt(4, SOL_SOCKET, 37, [2176], 4) = 0
0.200 write(4, ..., 11680) = 11680
+0 setsockopt(4, SOL_SOCKET, 37, [2688], 4) = 0
0.200 > P. 1:731(730) ack 1
0.200 > P. 731:1461(730) ack 1
0.200 > . 1461:8761(7300) ack 1
0.200 > P. 8761:13141(4380) ack 1
0.300 < . 1:1(0) ack 1 win 257 <sack 1461:2921,nop,nop>
0.300 < . 1:1(0) ack 1 win 257 <sack 1461:4381,nop,nop>
0.300 < . 1:1(0) ack 1 win 257 <sack 1461:5841,nop,nop>
0.300 > P. 1:1461(1460) ack 1
0.400 < . 1:1(0) ack 13141 win 257
0.400 close(4) = 0
0.400 > F. 13141:13141(0) ack 1
0.500 < F. 1:1(0) ack 13142 win 257
0.500 > . 13142:13142(0) ack 2
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Soheil Hassas Yeganeh <soheil@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Tested-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-20 13:39:28 +08:00
|
|
|
shinfo->tskey = next_shinfo->tskey;
|
tcp: Merge txstamp_ack in tcp_skb_collapse_tstamp
When collapsing skbs, txstamp_ack also needs to be merged.
Retrans Collapse Test:
~~~~~~
0.200 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_NODELAY, [1], 4) = 0
0.200 write(4, ..., 730) = 730
+0 setsockopt(4, SOL_SOCKET, 37, [2688], 4) = 0
0.200 write(4, ..., 730) = 730
+0 setsockopt(4, SOL_SOCKET, 37, [2176], 4) = 0
0.200 write(4, ..., 11680) = 11680
0.200 > P. 1:731(730) ack 1
0.200 > P. 731:1461(730) ack 1
0.200 > . 1461:8761(7300) ack 1
0.200 > P. 8761:13141(4380) ack 1
0.300 < . 1:1(0) ack 1 win 257 <sack 1461:2921,nop,nop>
0.300 < . 1:1(0) ack 1 win 257 <sack 1461:4381,nop,nop>
0.300 < . 1:1(0) ack 1 win 257 <sack 1461:5841,nop,nop>
0.300 > P. 1:1461(1460) ack 1
0.400 < . 1:1(0) ack 13141 win 257
BPF Output Before:
~~~~~
<No output due to missing SCM_TSTAMP_ACK timestamp>
BPF Output After:
~~~~~
<...>-2027 [007] d.s. 79.765921: : ee_data:1459
Sacks Collapse Test:
~~~~~
0.200 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_NODELAY, [1], 4) = 0
0.200 write(4, ..., 1460) = 1460
+0 setsockopt(4, SOL_SOCKET, 37, [2688], 4) = 0
0.200 write(4, ..., 13140) = 13140
+0 setsockopt(4, SOL_SOCKET, 37, [2176], 4) = 0
0.200 > P. 1:1461(1460) ack 1
0.200 > . 1461:8761(7300) ack 1
0.200 > P. 8761:14601(5840) ack 1
0.300 < . 1:1(0) ack 1 win 257 <sack 1461:14601,nop,nop>
0.300 > P. 1:1461(1460) ack 1
0.400 < . 1:1(0) ack 14601 win 257
BPF Output Before:
~~~~~
<No output due to missing SCM_TSTAMP_ACK timestamp>
BPF Output After:
~~~~~
<...>-2049 [007] d.s. 89.185538: : ee_data:14599
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Soheil Hassas Yeganeh <soheil@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Tested-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-20 13:50:48 +08:00
|
|
|
TCP_SKB_CB(skb)->txstamp_ack |=
|
|
|
|
|
TCP_SKB_CB(next_skb)->txstamp_ack;
|
tcp: Merge tx_flags and tskey in tcp_collapse_retrans
If two skbs are merged/collapsed during retransmission, the current
logic does not merge the tx_flags and tskey. The end result is
the SCM_TSTAMP_ACK timestamp could be missing for a packet.
The patch:
1. Merge the tx_flags
2. Overwrite the prev_skb's tskey with the next_skb's tskey
BPF Output Before:
~~~~~~
<no-output-due-to-missing-tstamp-event>
BPF Output After:
~~~~~~
packetdrill-2092 [001] d.s. 453.998486: : ee_data:1459
Packetdrill Script:
~~~~~~
+0 `sysctl -q -w net.ipv4.tcp_min_tso_segs=10`
+0 `sysctl -q -w net.ipv4.tcp_no_metrics_save=1`
+0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
0.100 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 7>
0.100 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 7>
0.200 < . 1:1(0) ack 1 win 257
0.200 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_NODELAY, [1], 4) = 0
0.200 write(4, ..., 730) = 730
+0 setsockopt(4, SOL_SOCKET, 37, [2688], 4) = 0
0.200 write(4, ..., 730) = 730
+0 setsockopt(4, SOL_SOCKET, 37, [2176], 4) = 0
0.200 write(4, ..., 11680) = 11680
+0 setsockopt(4, SOL_SOCKET, 37, [2688], 4) = 0
0.200 > P. 1:731(730) ack 1
0.200 > P. 731:1461(730) ack 1
0.200 > . 1461:8761(7300) ack 1
0.200 > P. 8761:13141(4380) ack 1
0.300 < . 1:1(0) ack 1 win 257 <sack 1461:2921,nop,nop>
0.300 < . 1:1(0) ack 1 win 257 <sack 1461:4381,nop,nop>
0.300 < . 1:1(0) ack 1 win 257 <sack 1461:5841,nop,nop>
0.300 > P. 1:1461(1460) ack 1
0.400 < . 1:1(0) ack 13141 win 257
0.400 close(4) = 0
0.400 > F. 13141:13141(0) ack 1
0.500 < F. 1:1(0) ack 13142 win 257
0.500 > . 13142:13142(0) ack 2
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Soheil Hassas Yeganeh <soheil@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Tested-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-20 13:39:28 +08:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2008-11-25 13:03:43 +08:00
|
|
|
/* Collapses two adjacent SKB's during retransmission. */
|
tcp: enhance tcp_collapse_retrans() with skb_shift()
In commit 2331ccc5b323 ("tcp: enhance tcp collapsing"),
we made a first step allowing copying right skb to left skb head.
Since all skbs in socket write queue are headless (but possibly the very
first one), this strategy often does not work.
This patch extends tcp_collapse_retrans() to perform frag shifting,
thanks to skb_shift() helper.
This helper needs to not BUG on non headless skbs, as callers are ok
with that.
Tested:
Following packetdrill test now passes :
0.000 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
+0 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 8>
+0 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
+.100 < . 1:1(0) ack 1 win 257
+0 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_NODELAY, [1], 4) = 0
+0 write(4, ..., 200) = 200
+0 > P. 1:201(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 201:401(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 401:601(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 601:801(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 801:1001(200) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1001:1101(100) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1101:1201(100) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1201:1301(100) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1301:1401(100) ack 1
+.099 < . 1:1(0) ack 201 win 257
+.001 < . 1:1(0) ack 201 win 257 <nop,nop,sack 1001:1401>
+0 > P. 201:1001(800) ack 1
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-11-16 04:51:50 +08:00
|
|
|
static bool tcp_collapse_retrans(struct sock *sk, struct sk_buff *skb)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
2017-10-06 13:21:27 +08:00
|
|
|
struct sk_buff *next_skb = skb_rb_next(skb);
|
2018-08-01 09:59:56 +08:00
|
|
|
int next_skb_size;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2007-12-31 20:51:11 +08:00
|
|
|
next_skb_size = next_skb->len;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2007-12-31 20:51:11 +08:00
|
|
|
BUG_ON(tcp_skb_pcount(skb) != 1 || tcp_skb_pcount(next_skb) != 1);
|
2007-09-26 13:44:14 +08:00
|
|
|
|
tcp: enhance tcp_collapse_retrans() with skb_shift()
In commit 2331ccc5b323 ("tcp: enhance tcp collapsing"),
we made a first step allowing copying right skb to left skb head.
Since all skbs in socket write queue are headless (but possibly the very
first one), this strategy often does not work.
This patch extends tcp_collapse_retrans() to perform frag shifting,
thanks to skb_shift() helper.
This helper needs to not BUG on non headless skbs, as callers are ok
with that.
Tested:
Following packetdrill test now passes :
0.000 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
+0 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 8>
+0 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
+.100 < . 1:1(0) ack 1 win 257
+0 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_NODELAY, [1], 4) = 0
+0 write(4, ..., 200) = 200
+0 > P. 1:201(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 201:401(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 401:601(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 601:801(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 801:1001(200) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1001:1101(100) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1101:1201(100) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1201:1301(100) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1301:1401(100) ack 1
+.099 < . 1:1(0) ack 201 win 257
+.001 < . 1:1(0) ack 201 win 257 <nop,nop,sack 1001:1401>
+0 > P. 201:1001(800) ack 1
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-11-16 04:51:50 +08:00
|
|
|
if (next_skb_size) {
|
|
|
|
|
if (next_skb_size <= skb_availroom(skb))
|
|
|
|
|
skb_copy_bits(next_skb, 0, skb_put(skb, next_skb_size),
|
|
|
|
|
next_skb_size);
|
2019-05-18 08:17:22 +08:00
|
|
|
else if (!tcp_skb_shift(skb, next_skb, 1, next_skb_size))
|
tcp: enhance tcp_collapse_retrans() with skb_shift()
In commit 2331ccc5b323 ("tcp: enhance tcp collapsing"),
we made a first step allowing copying right skb to left skb head.
Since all skbs in socket write queue are headless (but possibly the very
first one), this strategy often does not work.
This patch extends tcp_collapse_retrans() to perform frag shifting,
thanks to skb_shift() helper.
This helper needs to not BUG on non headless skbs, as callers are ok
with that.
Tested:
Following packetdrill test now passes :
0.000 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
+0 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 8>
+0 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
+.100 < . 1:1(0) ack 1 win 257
+0 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_NODELAY, [1], 4) = 0
+0 write(4, ..., 200) = 200
+0 > P. 1:201(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 201:401(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 401:601(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 601:801(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 801:1001(200) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1001:1101(100) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1101:1201(100) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1201:1301(100) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1301:1401(100) ack 1
+.099 < . 1:1(0) ack 201 win 257
+.001 < . 1:1(0) ack 201 win 257 <nop,nop,sack 1001:1401>
+0 > P. 201:1001(800) ack 1
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-11-16 04:51:50 +08:00
|
|
|
return false;
|
|
|
|
|
}
|
2017-10-31 14:08:20 +08:00
|
|
|
tcp_highest_sack_replace(sk, next_skb, skb);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2007-12-31 20:51:11 +08:00
|
|
|
/* Update sequence range on original skb. */
|
|
|
|
|
TCP_SKB_CB(skb)->end_seq = TCP_SKB_CB(next_skb)->end_seq;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2009-02-28 12:44:35 +08:00
|
|
|
/* Merge over control information. This moves PSH/FIN etc. over */
|
2011-09-28 01:25:05 +08:00
|
|
|
TCP_SKB_CB(skb)->tcp_flags |= TCP_SKB_CB(next_skb)->tcp_flags;
|
2007-12-31 20:51:11 +08:00
|
|
|
|
|
|
|
|
/* All done, get rid of second SKB and account for it so
|
|
|
|
|
* packet counting does not break.
|
|
|
|
|
*/
|
|
|
|
|
TCP_SKB_CB(skb)->sacked |= TCP_SKB_CB(next_skb)->sacked & TCPCB_EVER_RETRANS;
|
tcp: Handle eor bit when coalescing skb
This patch:
1. Prevent next_skb from coalescing to the prev_skb if
TCP_SKB_CB(prev_skb)->eor is set
2. Update the TCP_SKB_CB(prev_skb)->eor if coalescing is
allowed
Packetdrill script for testing:
~~~~~~
+0 `sysctl -q -w net.ipv4.tcp_min_tso_segs=10`
+0 `sysctl -q -w net.ipv4.tcp_no_metrics_save=1`
+0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
0.100 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 7>
0.100 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 7>
0.200 < . 1:1(0) ack 1 win 257
0.200 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_NODELAY, [1], 4) = 0
0.200 sendto(4, ..., 730, MSG_EOR, ..., ...) = 730
0.200 sendto(4, ..., 730, MSG_EOR, ..., ...) = 730
0.200 write(4, ..., 11680) = 11680
0.200 > P. 1:731(730) ack 1
0.200 > P. 731:1461(730) ack 1
0.200 > . 1461:8761(7300) ack 1
0.200 > P. 8761:13141(4380) ack 1
0.300 < . 1:1(0) ack 1 win 257 <sack 1461:13141,nop,nop>
0.300 > P. 1:731(730) ack 1
0.300 > P. 731:1461(730) ack 1
0.400 < . 1:1(0) ack 13141 win 257
0.400 close(4) = 0
0.400 > F. 13141:13141(0) ack 1
0.500 < F. 1:1(0) ack 13142 win 257
0.500 > . 13142:13142(0) ack 2
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Soheil Hassas Yeganeh <soheil@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-26 05:44:49 +08:00
|
|
|
TCP_SKB_CB(skb)->eor = TCP_SKB_CB(next_skb)->eor;
|
2007-12-31 20:51:11 +08:00
|
|
|
|
|
|
|
|
/* changed transmit queue under us so clear hints */
|
2008-09-21 12:25:15 +08:00
|
|
|
tcp_clear_retrans_hints_partial(tp);
|
|
|
|
|
if (next_skb == tp->retransmit_skb_hint)
|
|
|
|
|
tp->retransmit_skb_hint = skb;
|
2007-12-31 20:51:11 +08:00
|
|
|
|
2009-04-02 07:15:17 +08:00
|
|
|
tcp_adjust_pcount(sk, next_skb, tcp_skb_pcount(next_skb));
|
|
|
|
|
|
tcp: Merge tx_flags and tskey in tcp_collapse_retrans
If two skbs are merged/collapsed during retransmission, the current
logic does not merge the tx_flags and tskey. The end result is
the SCM_TSTAMP_ACK timestamp could be missing for a packet.
The patch:
1. Merge the tx_flags
2. Overwrite the prev_skb's tskey with the next_skb's tskey
BPF Output Before:
~~~~~~
<no-output-due-to-missing-tstamp-event>
BPF Output After:
~~~~~~
packetdrill-2092 [001] d.s. 453.998486: : ee_data:1459
Packetdrill Script:
~~~~~~
+0 `sysctl -q -w net.ipv4.tcp_min_tso_segs=10`
+0 `sysctl -q -w net.ipv4.tcp_no_metrics_save=1`
+0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
0.100 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 7>
0.100 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 7>
0.200 < . 1:1(0) ack 1 win 257
0.200 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_NODELAY, [1], 4) = 0
0.200 write(4, ..., 730) = 730
+0 setsockopt(4, SOL_SOCKET, 37, [2688], 4) = 0
0.200 write(4, ..., 730) = 730
+0 setsockopt(4, SOL_SOCKET, 37, [2176], 4) = 0
0.200 write(4, ..., 11680) = 11680
+0 setsockopt(4, SOL_SOCKET, 37, [2688], 4) = 0
0.200 > P. 1:731(730) ack 1
0.200 > P. 731:1461(730) ack 1
0.200 > . 1461:8761(7300) ack 1
0.200 > P. 8761:13141(4380) ack 1
0.300 < . 1:1(0) ack 1 win 257 <sack 1461:2921,nop,nop>
0.300 < . 1:1(0) ack 1 win 257 <sack 1461:4381,nop,nop>
0.300 < . 1:1(0) ack 1 win 257 <sack 1461:5841,nop,nop>
0.300 > P. 1:1461(1460) ack 1
0.400 < . 1:1(0) ack 13141 win 257
0.400 close(4) = 0
0.400 > F. 13141:13141(0) ack 1
0.500 < F. 1:1(0) ack 13142 win 257
0.500 > . 13142:13142(0) ack 2
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Soheil Hassas Yeganeh <soheil@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Tested-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-20 13:39:28 +08:00
|
|
|
tcp_skb_collapse_tstamp(skb, next_skb);
|
|
|
|
|
|
2017-10-06 13:21:27 +08:00
|
|
|
tcp_rtx_queue_unlink_and_free(next_skb, sk);
|
tcp: enhance tcp_collapse_retrans() with skb_shift()
In commit 2331ccc5b323 ("tcp: enhance tcp collapsing"),
we made a first step allowing copying right skb to left skb head.
Since all skbs in socket write queue are headless (but possibly the very
first one), this strategy often does not work.
This patch extends tcp_collapse_retrans() to perform frag shifting,
thanks to skb_shift() helper.
This helper needs to not BUG on non headless skbs, as callers are ok
with that.
Tested:
Following packetdrill test now passes :
0.000 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
+0 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 8>
+0 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
+.100 < . 1:1(0) ack 1 win 257
+0 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_NODELAY, [1], 4) = 0
+0 write(4, ..., 200) = 200
+0 > P. 1:201(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 201:401(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 401:601(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 601:801(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 801:1001(200) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1001:1101(100) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1101:1201(100) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1201:1301(100) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1301:1401(100) ack 1
+.099 < . 1:1(0) ack 201 win 257
+.001 < . 1:1(0) ack 201 win 257 <nop,nop,sack 1001:1401>
+0 > P. 201:1001(800) ack 1
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-11-16 04:51:50 +08:00
|
|
|
return true;
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
|
2009-07-22 07:00:40 +08:00
|
|
|
/* Check if coalescing SKBs is legal. */
|
2012-05-17 07:15:34 +08:00
|
|
|
static bool tcp_can_collapse(const struct sock *sk, const struct sk_buff *skb)
|
2008-11-25 13:03:43 +08:00
|
|
|
{
|
|
|
|
|
if (tcp_skb_pcount(skb) > 1)
|
2012-05-17 07:15:34 +08:00
|
|
|
return false;
|
2008-11-25 13:03:43 +08:00
|
|
|
if (skb_cloned(skb))
|
2012-05-17 07:15:34 +08:00
|
|
|
return false;
|
tcp: enhance tcp collapsing
As Ilya Lesokhin suggested, we can collapse two skbs at retransmit
time even if the skb at the right has fragments.
We simply have to use more generic skb_copy_bits() instead of
skb_copy_from_linear_data() in tcp_collapse_retrans()
Also need to guard this skb_copy_bits() in case there is nothing to
copy, otherwise skb_put() could panic if left skb has frags.
Tested:
Used following packetdrill test
// Establish a connection.
0.000 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
+0 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 8>
+0 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
+.100 < . 1:1(0) ack 1 win 257
+0 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_NODELAY, [1], 4) = 0
+0 write(4, ..., 200) = 200
+0 > P. 1:201(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 201:401(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 401:601(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 601:801(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 801:1001(200) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1001:1101(100) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1101:1201(100) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1201:1301(100) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1301:1401(100) ack 1
+.100 < . 1:1(0) ack 1 win 257 <nop,nop,sack 1001:1401>
// Check that TCP collapse works :
+0 > P. 1:1001(1000) ack 1
Reported-by: Ilya Lesokhin <ilyal@mellanox.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-11-02 01:53:42 +08:00
|
|
|
/* Some heuristics for collapsing over SACK'd could be invented */
|
2008-11-25 13:03:43 +08:00
|
|
|
if (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_ACKED)
|
2012-05-17 07:15:34 +08:00
|
|
|
return false;
|
2008-11-25 13:03:43 +08:00
|
|
|
|
2012-05-17 07:15:34 +08:00
|
|
|
return true;
|
2008-11-25 13:03:43 +08:00
|
|
|
}
|
|
|
|
|
|
2009-07-22 07:00:40 +08:00
|
|
|
/* Collapse packets in the retransmit queue to make to create
|
|
|
|
|
* less packets on the wire. This is only done on retransmission.
|
|
|
|
|
*/
|
2008-11-25 13:03:43 +08:00
|
|
|
static void tcp_retrans_try_collapse(struct sock *sk, struct sk_buff *to,
|
|
|
|
|
int space)
|
|
|
|
|
{
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
struct sk_buff *skb = to, *tmp;
|
2012-05-17 07:15:34 +08:00
|
|
|
bool first = true;
|
2008-11-25 13:03:43 +08:00
|
|
|
|
2017-10-27 12:55:00 +08:00
|
|
|
if (!sock_net(sk)->ipv4.sysctl_tcp_retrans_collapse)
|
2008-11-25 13:03:43 +08:00
|
|
|
return;
|
2011-09-28 01:25:05 +08:00
|
|
|
if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN)
|
2008-11-25 13:03:43 +08:00
|
|
|
return;
|
|
|
|
|
|
2017-10-06 13:21:27 +08:00
|
|
|
skb_rbtree_walk_from_safe(skb, tmp) {
|
2008-11-25 13:03:43 +08:00
|
|
|
if (!tcp_can_collapse(sk, skb))
|
|
|
|
|
break;
|
|
|
|
|
|
2020-01-09 23:59:20 +08:00
|
|
|
if (!tcp_skb_can_collapse(to, skb))
|
tcp: Handle eor bit when coalescing skb
This patch:
1. Prevent next_skb from coalescing to the prev_skb if
TCP_SKB_CB(prev_skb)->eor is set
2. Update the TCP_SKB_CB(prev_skb)->eor if coalescing is
allowed
Packetdrill script for testing:
~~~~~~
+0 `sysctl -q -w net.ipv4.tcp_min_tso_segs=10`
+0 `sysctl -q -w net.ipv4.tcp_no_metrics_save=1`
+0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
0.100 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 7>
0.100 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 7>
0.200 < . 1:1(0) ack 1 win 257
0.200 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_NODELAY, [1], 4) = 0
0.200 sendto(4, ..., 730, MSG_EOR, ..., ...) = 730
0.200 sendto(4, ..., 730, MSG_EOR, ..., ...) = 730
0.200 write(4, ..., 11680) = 11680
0.200 > P. 1:731(730) ack 1
0.200 > P. 731:1461(730) ack 1
0.200 > . 1461:8761(7300) ack 1
0.200 > P. 8761:13141(4380) ack 1
0.300 < . 1:1(0) ack 1 win 257 <sack 1461:13141,nop,nop>
0.300 > P. 1:731(730) ack 1
0.300 > P. 731:1461(730) ack 1
0.400 < . 1:1(0) ack 13141 win 257
0.400 close(4) = 0
0.400 > F. 13141:13141(0) ack 1
0.500 < F. 1:1(0) ack 13142 win 257
0.500 > . 13142:13142(0) ack 2
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Soheil Hassas Yeganeh <soheil@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-26 05:44:49 +08:00
|
|
|
break;
|
|
|
|
|
|
2008-11-25 13:03:43 +08:00
|
|
|
space -= skb->len;
|
|
|
|
|
|
|
|
|
|
if (first) {
|
2012-05-17 07:15:34 +08:00
|
|
|
first = false;
|
2008-11-25 13:03:43 +08:00
|
|
|
continue;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (space < 0)
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
if (after(TCP_SKB_CB(skb)->end_seq, tcp_wnd_end(tp)))
|
|
|
|
|
break;
|
|
|
|
|
|
tcp: enhance tcp_collapse_retrans() with skb_shift()
In commit 2331ccc5b323 ("tcp: enhance tcp collapsing"),
we made a first step allowing copying right skb to left skb head.
Since all skbs in socket write queue are headless (but possibly the very
first one), this strategy often does not work.
This patch extends tcp_collapse_retrans() to perform frag shifting,
thanks to skb_shift() helper.
This helper needs to not BUG on non headless skbs, as callers are ok
with that.
Tested:
Following packetdrill test now passes :
0.000 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0
+0 < S 0:0(0) win 32792 <mss 1460,sackOK,nop,nop,nop,wscale 8>
+0 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 8>
+.100 < . 1:1(0) ack 1 win 257
+0 accept(3, ..., ...) = 4
+0 setsockopt(4, SOL_TCP, TCP_NODELAY, [1], 4) = 0
+0 write(4, ..., 200) = 200
+0 > P. 1:201(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 201:401(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 401:601(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 601:801(200) ack 1
+.001 write(4, ..., 200) = 200
+0 > P. 801:1001(200) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1001:1101(100) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1101:1201(100) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1201:1301(100) ack 1
+.001 write(4, ..., 100) = 100
+0 > P. 1301:1401(100) ack 1
+.099 < . 1:1(0) ack 201 win 257
+.001 < . 1:1(0) ack 201 win 257 <nop,nop,sack 1001:1401>
+0 > P. 201:1001(800) ack 1
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-11-16 04:51:50 +08:00
|
|
|
if (!tcp_collapse_retrans(sk, to))
|
|
|
|
|
break;
|
2008-11-25 13:03:43 +08:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/* This retransmits one SKB. Policy decisions and retransmit queue
|
|
|
|
|
* state updates are done by the caller. Returns non-zero if an
|
|
|
|
|
* error occurred which prevented the send.
|
|
|
|
|
*/
|
tcp-tso: do not split TSO packets at retransmit time
Linux TCP stack painfully segments all TSO/GSO packets before retransmits.
This was fine back in the days when TSO/GSO were emerging, with their
bugs, but we believe the dark age is over.
Keeping big packets in write queues, but also in stack traversal
has a lot of benefits.
- Less memory overhead, because write queues have less skbs
- Less cpu overhead at ACK processing.
- Better SACK processing, as lot of studies mentioned how
awful linux was at this ;)
- Less cpu overhead to send the rtx packets
(IP stack traversal, netfilter traversal, drivers...)
- Better latencies in presence of losses.
- Smaller spikes in fq like packet schedulers, as retransmits
are not constrained by TCP Small Queues.
1 % packet losses are common today, and at 100Gbit speeds, this
translates to ~80,000 losses per second.
Losses are often correlated, and we see many retransmit events
leading to 1-MSS train of packets, at the time hosts are already
under stress.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-22 01:55:23 +08:00
|
|
|
int __tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb, int segs)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
2006-03-21 09:53:41 +08:00
|
|
|
struct inet_connection_sock *icsk = inet_csk(sk);
|
tcp-tso: do not split TSO packets at retransmit time
Linux TCP stack painfully segments all TSO/GSO packets before retransmits.
This was fine back in the days when TSO/GSO were emerging, with their
bugs, but we believe the dark age is over.
Keeping big packets in write queues, but also in stack traversal
has a lot of benefits.
- Less memory overhead, because write queues have less skbs
- Less cpu overhead at ACK processing.
- Better SACK processing, as lot of studies mentioned how
awful linux was at this ;)
- Less cpu overhead to send the rtx packets
(IP stack traversal, netfilter traversal, drivers...)
- Better latencies in presence of losses.
- Smaller spikes in fq like packet schedulers, as retransmits
are not constrained by TCP Small Queues.
1 % packet losses are common today, and at 100Gbit speeds, this
translates to ~80,000 losses per second.
Losses are often correlated, and we see many retransmit events
leading to 1-MSS train of packets, at the time hosts are already
under stress.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-22 01:55:23 +08:00
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
2008-05-22 07:42:20 +08:00
|
|
|
unsigned int cur_mss;
|
tcp-tso: do not split TSO packets at retransmit time
Linux TCP stack painfully segments all TSO/GSO packets before retransmits.
This was fine back in the days when TSO/GSO were emerging, with their
bugs, but we believe the dark age is over.
Keeping big packets in write queues, but also in stack traversal
has a lot of benefits.
- Less memory overhead, because write queues have less skbs
- Less cpu overhead at ACK processing.
- Better SACK processing, as lot of studies mentioned how
awful linux was at this ;)
- Less cpu overhead to send the rtx packets
(IP stack traversal, netfilter traversal, drivers...)
- Better latencies in presence of losses.
- Smaller spikes in fq like packet schedulers, as retransmits
are not constrained by TCP Small Queues.
1 % packet losses are common today, and at 100Gbit speeds, this
translates to ~80,000 losses per second.
Losses are often correlated, and we see many retransmit events
leading to 1-MSS train of packets, at the time hosts are already
under stress.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-22 01:55:23 +08:00
|
|
|
int diff, len, err;
|
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
|
tcp-tso: do not split TSO packets at retransmit time
Linux TCP stack painfully segments all TSO/GSO packets before retransmits.
This was fine back in the days when TSO/GSO were emerging, with their
bugs, but we believe the dark age is over.
Keeping big packets in write queues, but also in stack traversal
has a lot of benefits.
- Less memory overhead, because write queues have less skbs
- Less cpu overhead at ACK processing.
- Better SACK processing, as lot of studies mentioned how
awful linux was at this ;)
- Less cpu overhead to send the rtx packets
(IP stack traversal, netfilter traversal, drivers...)
- Better latencies in presence of losses.
- Smaller spikes in fq like packet schedulers, as retransmits
are not constrained by TCP Small Queues.
1 % packet losses are common today, and at 100Gbit speeds, this
translates to ~80,000 losses per second.
Losses are often correlated, and we see many retransmit events
leading to 1-MSS train of packets, at the time hosts are already
under stress.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-22 01:55:23 +08:00
|
|
|
/* Inconclusive MTU probe */
|
|
|
|
|
if (icsk->icsk_mtup.probe_size)
|
2006-03-21 09:53:41 +08:00
|
|
|
icsk->icsk_mtup.probe_size = 0;
|
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/* Do not sent more than we queued. 1/4 is reserved for possible
|
2005-11-11 09:13:47 +08:00
|
|
|
* copying overhead: fragmentation, tunneling, mangling etc.
|
2005-04-17 06:20:36 +08:00
|
|
|
*/
|
2017-06-30 18:08:00 +08:00
|
|
|
if (refcount_read(&sk->sk_wmem_alloc) >
|
2016-09-15 23:12:33 +08:00
|
|
|
min_t(u32, sk->sk_wmem_queued + (sk->sk_wmem_queued >> 2),
|
|
|
|
|
sk->sk_sndbuf))
|
2005-04-17 06:20:36 +08:00
|
|
|
return -EAGAIN;
|
|
|
|
|
|
2014-04-21 08:58:17 +08:00
|
|
|
if (skb_still_in_host_queue(sk, skb))
|
|
|
|
|
return -EBUSY;
|
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
if (before(TCP_SKB_CB(skb)->seq, tp->snd_una)) {
|
2018-05-15 12:14:26 +08:00
|
|
|
if (unlikely(before(TCP_SKB_CB(skb)->end_seq, tp->snd_una))) {
|
|
|
|
|
WARN_ON_ONCE(1);
|
|
|
|
|
return -EINVAL;
|
|
|
|
|
}
|
2005-04-17 06:20:36 +08:00
|
|
|
if (tcp_trim_head(sk, skb, tp->snd_una - TCP_SKB_CB(skb)->seq))
|
|
|
|
|
return -ENOMEM;
|
|
|
|
|
}
|
|
|
|
|
|
2008-05-22 07:42:20 +08:00
|
|
|
if (inet_csk(sk)->icsk_af_ops->rebuild_header(sk))
|
|
|
|
|
return -EHOSTUNREACH; /* Routing failure or similar. */
|
|
|
|
|
|
2009-03-14 22:23:05 +08:00
|
|
|
cur_mss = tcp_current_mss(sk);
|
2008-05-22 07:42:20 +08:00
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/* If receiver has shrunk his window, and skb is out of
|
|
|
|
|
* new window, do not retransmit it. The exception is the
|
|
|
|
|
* case, when window is shrunk to zero. In this case
|
|
|
|
|
* our retransmit serves as a zero window probe.
|
|
|
|
|
*/
|
2009-11-24 02:41:23 +08:00
|
|
|
if (!before(TCP_SKB_CB(skb)->seq, tcp_wnd_end(tp)) &&
|
|
|
|
|
TCP_SKB_CB(skb)->seq != tp->snd_una)
|
2005-04-17 06:20:36 +08:00
|
|
|
return -EAGAIN;
|
|
|
|
|
|
tcp-tso: do not split TSO packets at retransmit time
Linux TCP stack painfully segments all TSO/GSO packets before retransmits.
This was fine back in the days when TSO/GSO were emerging, with their
bugs, but we believe the dark age is over.
Keeping big packets in write queues, but also in stack traversal
has a lot of benefits.
- Less memory overhead, because write queues have less skbs
- Less cpu overhead at ACK processing.
- Better SACK processing, as lot of studies mentioned how
awful linux was at this ;)
- Less cpu overhead to send the rtx packets
(IP stack traversal, netfilter traversal, drivers...)
- Better latencies in presence of losses.
- Smaller spikes in fq like packet schedulers, as retransmits
are not constrained by TCP Small Queues.
1 % packet losses are common today, and at 100Gbit speeds, this
translates to ~80,000 losses per second.
Losses are often correlated, and we see many retransmit events
leading to 1-MSS train of packets, at the time hosts are already
under stress.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-22 01:55:23 +08:00
|
|
|
len = cur_mss * segs;
|
|
|
|
|
if (skb->len > len) {
|
2017-10-06 13:21:27 +08:00
|
|
|
if (tcp_fragment(sk, TCP_FRAG_IN_RTX_QUEUE, skb, len,
|
|
|
|
|
cur_mss, GFP_ATOMIC))
|
2005-04-17 06:20:36 +08:00
|
|
|
return -ENOMEM; /* We'll try again later. */
|
2009-02-28 12:44:31 +08:00
|
|
|
} else {
|
tcp-tso: do not split TSO packets at retransmit time
Linux TCP stack painfully segments all TSO/GSO packets before retransmits.
This was fine back in the days when TSO/GSO were emerging, with their
bugs, but we believe the dark age is over.
Keeping big packets in write queues, but also in stack traversal
has a lot of benefits.
- Less memory overhead, because write queues have less skbs
- Less cpu overhead at ACK processing.
- Better SACK processing, as lot of studies mentioned how
awful linux was at this ;)
- Less cpu overhead to send the rtx packets
(IP stack traversal, netfilter traversal, drivers...)
- Better latencies in presence of losses.
- Smaller spikes in fq like packet schedulers, as retransmits
are not constrained by TCP Small Queues.
1 % packet losses are common today, and at 100Gbit speeds, this
translates to ~80,000 losses per second.
Losses are often correlated, and we see many retransmit events
leading to 1-MSS train of packets, at the time hosts are already
under stress.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-22 01:55:23 +08:00
|
|
|
if (skb_unclone(skb, GFP_ATOMIC))
|
|
|
|
|
return -ENOMEM;
|
2009-04-02 07:18:20 +08:00
|
|
|
|
tcp-tso: do not split TSO packets at retransmit time
Linux TCP stack painfully segments all TSO/GSO packets before retransmits.
This was fine back in the days when TSO/GSO were emerging, with their
bugs, but we believe the dark age is over.
Keeping big packets in write queues, but also in stack traversal
has a lot of benefits.
- Less memory overhead, because write queues have less skbs
- Less cpu overhead at ACK processing.
- Better SACK processing, as lot of studies mentioned how
awful linux was at this ;)
- Less cpu overhead to send the rtx packets
(IP stack traversal, netfilter traversal, drivers...)
- Better latencies in presence of losses.
- Smaller spikes in fq like packet schedulers, as retransmits
are not constrained by TCP Small Queues.
1 % packet losses are common today, and at 100Gbit speeds, this
translates to ~80,000 losses per second.
Losses are often correlated, and we see many retransmit events
leading to 1-MSS train of packets, at the time hosts are already
under stress.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-22 01:55:23 +08:00
|
|
|
diff = tcp_skb_pcount(skb);
|
|
|
|
|
tcp_set_skb_tso_segs(skb, cur_mss);
|
|
|
|
|
diff -= tcp_skb_pcount(skb);
|
|
|
|
|
if (diff)
|
|
|
|
|
tcp_adjust_pcount(sk, skb, diff);
|
|
|
|
|
if (skb->len < cur_mss)
|
|
|
|
|
tcp_retrans_try_collapse(sk, skb, cur_mss);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
|
tcp: add rfc3168, section 6.1.1.1. fallback
This work as a follow-up of commit f7b3bec6f516 ("net: allow setting ecn
via routing table") and adds RFC3168 section 6.1.1.1. fallback for outgoing
ECN connections. In other words, this work adds a retry with a non-ECN
setup SYN packet, as suggested from the RFC on the first timeout:
[...] A host that receives no reply to an ECN-setup SYN within the
normal SYN retransmission timeout interval MAY resend the SYN and
any subsequent SYN retransmissions with CWR and ECE cleared. [...]
Schematic client-side view when assuming the server is in tcp_ecn=2 mode,
that is, Linux default since 2009 via commit 255cac91c3c9 ("tcp: extend
ECN sysctl to allow server-side only ECN"):
1) Normal ECN-capable path:
SYN ECE CWR ----->
<----- SYN ACK ECE
ACK ----->
2) Path with broken middlebox, when client has fallback:
SYN ECE CWR ----X crappy middlebox drops packet
(timeout, rtx)
SYN ----->
<----- SYN ACK
ACK ----->
In case we would not have the fallback implemented, the middlebox drop
point would basically end up as:
SYN ECE CWR ----X crappy middlebox drops packet
(timeout, rtx)
SYN ECE CWR ----X crappy middlebox drops packet
(timeout, rtx)
SYN ECE CWR ----X crappy middlebox drops packet
(timeout, rtx)
In any case, it's rather a smaller percentage of sites where there would
occur such additional setup latency: it was found in end of 2014 that ~56%
of IPv4 and 65% of IPv6 servers of Alexa 1 million list would negotiate
ECN (aka tcp_ecn=2 default), 0.42% of these webservers will fail to connect
when trying to negotiate with ECN (tcp_ecn=1) due to timeouts, which the
fallback would mitigate with a slight latency trade-off. Recent related
paper on this topic:
Brian Trammell, Mirja Kühlewind, Damiano Boppart, Iain Learmonth,
Gorry Fairhurst, and Richard Scheffenegger:
"Enabling Internet-Wide Deployment of Explicit Congestion Notification."
Proc. PAM 2015, New York.
http://ecn.ethz.ch/ecn-pam15.pdf
Thus, when net.ipv4.tcp_ecn=1 is being set, the patch will perform RFC3168,
section 6.1.1.1. fallback on timeout. For users explicitly not wanting this
which can be in DC use case, we add a net.ipv4.tcp_ecn_fallback knob that
allows for disabling the fallback.
tp->ecn_flags are not being cleared in tcp_ecn_clear_syn() on output, but
rather we let tcp_ecn_rcv_synack() take that over on input path in case a
SYN ACK ECE was delayed. Thus a spurious SYN retransmission will not prevent
ECN being negotiated eventually in that case.
Reference: https://www.ietf.org/proceedings/92/slides/slides-92-iccrg-1.pdf
Reference: https://www.ietf.org/proceedings/89/slides/slides-89-tsvarea-1.pdf
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Mirja Kühlewind <mirja.kuehlewind@tik.ee.ethz.ch>
Signed-off-by: Brian Trammell <trammell@tik.ee.ethz.ch>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Dave That <dave.taht@gmail.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-05-20 03:04:22 +08:00
|
|
|
/* RFC3168, section 6.1.1.1. ECN fallback */
|
|
|
|
|
if ((TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN_ECN) == TCPHDR_SYN_ECN)
|
|
|
|
|
tcp_ecn_clear_syn(sk, skb);
|
|
|
|
|
|
2017-01-28 08:24:39 +08:00
|
|
|
/* Update global and local TCP statistics. */
|
|
|
|
|
segs = tcp_skb_pcount(skb);
|
|
|
|
|
TCP_ADD_STATS(sock_net(sk), TCP_MIB_RETRANSSEGS, segs);
|
|
|
|
|
if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN)
|
|
|
|
|
__NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPSYNRETRANS);
|
|
|
|
|
tp->total_retrans += segs;
|
2018-08-01 08:46:22 +08:00
|
|
|
tp->bytes_retrans += skb->len;
|
2017-01-28 08:24:39 +08:00
|
|
|
|
2013-04-11 18:57:18 +08:00
|
|
|
/* make sure skb->data is aligned on arches that require it
|
|
|
|
|
* and check if ack-trimming & collapsing extended the headroom
|
|
|
|
|
* beyond what csum_start can cover.
|
|
|
|
|
*/
|
|
|
|
|
if (unlikely((NET_IP_ALIGN && ((unsigned long)skb->data & 3)) ||
|
|
|
|
|
skb_headroom(skb) >= 0xFFFF)) {
|
2016-05-10 11:55:16 +08:00
|
|
|
struct sk_buff *nskb;
|
|
|
|
|
|
2017-10-05 03:59:58 +08:00
|
|
|
tcp_skb_tsorted_save(skb) {
|
|
|
|
|
nskb = __pskb_copy(skb, MAX_TCP_HEADER, GFP_ATOMIC);
|
2020-03-20 23:52:02 +08:00
|
|
|
if (nskb) {
|
|
|
|
|
nskb->dev = NULL;
|
|
|
|
|
err = tcp_transmit_skb(sk, nskb, 0, GFP_ATOMIC);
|
|
|
|
|
} else {
|
|
|
|
|
err = -ENOBUFS;
|
|
|
|
|
}
|
2017-10-05 03:59:58 +08:00
|
|
|
} tcp_skb_tsorted_restore(skb);
|
|
|
|
|
|
2017-10-25 07:44:42 +08:00
|
|
|
if (!err) {
|
2018-10-16 00:37:54 +08:00
|
|
|
tcp_update_skb_after_send(sk, skb, tp->tcp_wstamp_ns);
|
2017-10-25 07:44:42 +08:00
|
|
|
tcp_rate_skb_sent(sk, skb);
|
|
|
|
|
}
|
2011-12-04 05:39:53 +08:00
|
|
|
} else {
|
2014-03-01 08:42:26 +08:00
|
|
|
err = tcp_transmit_skb(sk, skb, 1, GFP_ATOMIC);
|
2011-12-04 05:39:53 +08:00
|
|
|
}
|
2014-03-01 08:42:26 +08:00
|
|
|
|
2019-01-17 07:05:29 +08:00
|
|
|
/* To avoid taking spuriously low RTT samples based on a timestamp
|
|
|
|
|
* for a transmit that never happened, always mark EVER_RETRANS
|
|
|
|
|
*/
|
|
|
|
|
TCP_SKB_CB(skb)->sacked |= TCPCB_EVER_RETRANS;
|
|
|
|
|
|
2018-01-26 08:14:14 +08:00
|
|
|
if (BPF_SOCK_OPS_TEST_FLAG(tp, BPF_SOCK_OPS_RETRANS_CB_FLAG))
|
|
|
|
|
tcp_call_bpf_3arg(sk, BPF_SOCK_OPS_RETRANS_CB,
|
|
|
|
|
TCP_SKB_CB(skb)->seq, segs, err);
|
|
|
|
|
|
2014-04-29 13:00:29 +08:00
|
|
|
if (likely(!err)) {
|
2017-10-14 04:03:16 +08:00
|
|
|
trace_tcp_retransmit_skb(sk, skb);
|
2017-01-28 08:24:39 +08:00
|
|
|
} else if (err != -EBUSY) {
|
2018-11-29 08:06:44 +08:00
|
|
|
NET_ADD_STATS(sock_net(sk), LINUX_MIB_TCPRETRANSFAIL, segs);
|
2014-04-29 13:00:29 +08:00
|
|
|
}
|
2014-03-01 08:42:26 +08:00
|
|
|
return err;
|
2012-12-06 16:45:32 +08:00
|
|
|
}
|
|
|
|
|
|
tcp-tso: do not split TSO packets at retransmit time
Linux TCP stack painfully segments all TSO/GSO packets before retransmits.
This was fine back in the days when TSO/GSO were emerging, with their
bugs, but we believe the dark age is over.
Keeping big packets in write queues, but also in stack traversal
has a lot of benefits.
- Less memory overhead, because write queues have less skbs
- Less cpu overhead at ACK processing.
- Better SACK processing, as lot of studies mentioned how
awful linux was at this ;)
- Less cpu overhead to send the rtx packets
(IP stack traversal, netfilter traversal, drivers...)
- Better latencies in presence of losses.
- Smaller spikes in fq like packet schedulers, as retransmits
are not constrained by TCP Small Queues.
1 % packet losses are common today, and at 100Gbit speeds, this
translates to ~80,000 losses per second.
Losses are often correlated, and we see many retransmit events
leading to 1-MSS train of packets, at the time hosts are already
under stress.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-22 01:55:23 +08:00
|
|
|
int tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb, int segs)
|
2012-12-06 16:45:32 +08:00
|
|
|
{
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
tcp-tso: do not split TSO packets at retransmit time
Linux TCP stack painfully segments all TSO/GSO packets before retransmits.
This was fine back in the days when TSO/GSO were emerging, with their
bugs, but we believe the dark age is over.
Keeping big packets in write queues, but also in stack traversal
has a lot of benefits.
- Less memory overhead, because write queues have less skbs
- Less cpu overhead at ACK processing.
- Better SACK processing, as lot of studies mentioned how
awful linux was at this ;)
- Less cpu overhead to send the rtx packets
(IP stack traversal, netfilter traversal, drivers...)
- Better latencies in presence of losses.
- Smaller spikes in fq like packet schedulers, as retransmits
are not constrained by TCP Small Queues.
1 % packet losses are common today, and at 100Gbit speeds, this
translates to ~80,000 losses per second.
Losses are often correlated, and we see many retransmit events
leading to 1-MSS train of packets, at the time hosts are already
under stress.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-22 01:55:23 +08:00
|
|
|
int err = __tcp_retransmit_skb(sk, skb, segs);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
|
if (err == 0) {
|
|
|
|
|
#if FASTRETRANS_DEBUG > 0
|
2008-01-01 06:57:14 +08:00
|
|
|
if (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_RETRANS) {
|
2012-05-14 05:56:26 +08:00
|
|
|
net_dbg_ratelimited("retrans_out leaked\n");
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
TCP_SKB_CB(skb)->sacked |= TCPCB_RETRANS;
|
|
|
|
|
tp->retrans_out += tcp_skb_pcount(skb);
|
|
|
|
|
}
|
2014-07-03 03:07:16 +08:00
|
|
|
|
2019-01-17 07:05:30 +08:00
|
|
|
/* Save stamp of the first (attempted) retransmit. */
|
|
|
|
|
if (!tp->retrans_stamp)
|
|
|
|
|
tp->retrans_stamp = tcp_skb_timestamp(skb);
|
|
|
|
|
|
2014-07-03 03:07:16 +08:00
|
|
|
if (tp->undo_retrans < 0)
|
|
|
|
|
tp->undo_retrans = 0;
|
|
|
|
|
tp->undo_retrans += tcp_skb_pcount(skb);
|
2005-04-17 06:20:36 +08:00
|
|
|
return err;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* This gets called after a retransmit timeout, and the initially
|
|
|
|
|
* retransmitted data is acknowledged. It tries to continue
|
|
|
|
|
* resending the rest of the retransmit queue, until either
|
|
|
|
|
* we've sent it all or the congestion window limit is reached.
|
|
|
|
|
*/
|
|
|
|
|
void tcp_xmit_retransmit_queue(struct sock *sk)
|
|
|
|
|
{
|
2005-08-10 15:03:31 +08:00
|
|
|
const struct inet_connection_sock *icsk = inet_csk(sk);
|
2017-10-17 10:38:35 +08:00
|
|
|
struct sk_buff *skb, *rtx_head, *hole = NULL;
|
2005-04-17 06:20:36 +08:00
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
2020-05-05 02:27:50 +08:00
|
|
|
bool rearm_timer = false;
|
2017-01-13 14:11:38 +08:00
|
|
|
u32 max_segs;
|
2008-09-21 12:22:59 +08:00
|
|
|
int mib_idx;
|
2005-11-11 09:14:59 +08:00
|
|
|
|
2010-07-19 09:16:18 +08:00
|
|
|
if (!tp->packets_out)
|
|
|
|
|
return;
|
|
|
|
|
|
2017-10-17 10:38:35 +08:00
|
|
|
rtx_head = tcp_rtx_queue_head(sk);
|
|
|
|
|
skb = tp->retransmit_skb_hint ?: rtx_head;
|
2016-09-20 11:39:17 +08:00
|
|
|
max_segs = tcp_tso_segs(sk, tcp_current_mss(sk));
|
2017-10-06 13:21:27 +08:00
|
|
|
skb_rbtree_walk_from(skb) {
|
2016-08-17 22:48:36 +08:00
|
|
|
__u8 sacked;
|
tcp-tso: do not split TSO packets at retransmit time
Linux TCP stack painfully segments all TSO/GSO packets before retransmits.
This was fine back in the days when TSO/GSO were emerging, with their
bugs, but we believe the dark age is over.
Keeping big packets in write queues, but also in stack traversal
has a lot of benefits.
- Less memory overhead, because write queues have less skbs
- Less cpu overhead at ACK processing.
- Better SACK processing, as lot of studies mentioned how
awful linux was at this ;)
- Less cpu overhead to send the rtx packets
(IP stack traversal, netfilter traversal, drivers...)
- Better latencies in presence of losses.
- Smaller spikes in fq like packet schedulers, as retransmits
are not constrained by TCP Small Queues.
1 % packet losses are common today, and at 100Gbit speeds, this
translates to ~80,000 losses per second.
Losses are often correlated, and we see many retransmit events
leading to 1-MSS train of packets, at the time hosts are already
under stress.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-22 01:55:23 +08:00
|
|
|
int segs;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2017-05-16 19:24:36 +08:00
|
|
|
if (tcp_pacing_check(sk))
|
|
|
|
|
break;
|
|
|
|
|
|
2008-09-21 12:23:49 +08:00
|
|
|
/* we could do better than to assign each time */
|
2015-04-03 16:17:26 +08:00
|
|
|
if (!hole)
|
2008-09-21 12:24:21 +08:00
|
|
|
tp->retransmit_skb_hint = skb;
|
2008-09-21 12:23:49 +08:00
|
|
|
|
tcp-tso: do not split TSO packets at retransmit time
Linux TCP stack painfully segments all TSO/GSO packets before retransmits.
This was fine back in the days when TSO/GSO were emerging, with their
bugs, but we believe the dark age is over.
Keeping big packets in write queues, but also in stack traversal
has a lot of benefits.
- Less memory overhead, because write queues have less skbs
- Less cpu overhead at ACK processing.
- Better SACK processing, as lot of studies mentioned how
awful linux was at this ;)
- Less cpu overhead to send the rtx packets
(IP stack traversal, netfilter traversal, drivers...)
- Better latencies in presence of losses.
- Smaller spikes in fq like packet schedulers, as retransmits
are not constrained by TCP Small Queues.
1 % packet losses are common today, and at 100Gbit speeds, this
translates to ~80,000 losses per second.
Losses are often correlated, and we see many retransmit events
leading to 1-MSS train of packets, at the time hosts are already
under stress.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-22 01:55:23 +08:00
|
|
|
segs = tp->snd_cwnd - tcp_packets_in_flight(tp);
|
|
|
|
|
if (segs <= 0)
|
2020-05-05 02:27:50 +08:00
|
|
|
break;
|
2016-08-17 22:48:36 +08:00
|
|
|
sacked = TCP_SKB_CB(skb)->sacked;
|
2016-06-27 23:38:50 +08:00
|
|
|
/* In case tcp_shift_skb_data() have aggregated large skbs,
|
|
|
|
|
* we need to make sure not sending too bigs TSO packets
|
|
|
|
|
*/
|
|
|
|
|
segs = min_t(int, segs, max_segs);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2017-01-13 14:11:38 +08:00
|
|
|
if (tp->retrans_out >= tp->lost_out) {
|
|
|
|
|
break;
|
2008-09-21 12:24:21 +08:00
|
|
|
} else if (!(sacked & TCPCB_LOST)) {
|
2015-04-03 16:17:26 +08:00
|
|
|
if (!hole && !(sacked & (TCPCB_SACKED_RETRANS|TCPCB_SACKED_ACKED)))
|
2008-09-21 12:24:21 +08:00
|
|
|
hole = skb;
|
|
|
|
|
continue;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2008-09-21 12:24:21 +08:00
|
|
|
} else {
|
|
|
|
|
if (icsk->icsk_ca_state != TCP_CA_Loss)
|
|
|
|
|
mib_idx = LINUX_MIB_TCPFASTRETRANS;
|
|
|
|
|
else
|
|
|
|
|
mib_idx = LINUX_MIB_TCPSLOWSTARTRETRANS;
|
|
|
|
|
}
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2008-09-21 12:24:21 +08:00
|
|
|
if (sacked & (TCPCB_SACKED_ACKED|TCPCB_SACKED_RETRANS))
|
2005-04-17 06:20:36 +08:00
|
|
|
continue;
|
|
|
|
|
|
2016-09-21 13:45:58 +08:00
|
|
|
if (tcp_small_queue_check(sk, skb, 1))
|
2020-05-05 02:27:50 +08:00
|
|
|
break;
|
2016-09-21 13:45:58 +08:00
|
|
|
|
tcp-tso: do not split TSO packets at retransmit time
Linux TCP stack painfully segments all TSO/GSO packets before retransmits.
This was fine back in the days when TSO/GSO were emerging, with their
bugs, but we believe the dark age is over.
Keeping big packets in write queues, but also in stack traversal
has a lot of benefits.
- Less memory overhead, because write queues have less skbs
- Less cpu overhead at ACK processing.
- Better SACK processing, as lot of studies mentioned how
awful linux was at this ;)
- Less cpu overhead to send the rtx packets
(IP stack traversal, netfilter traversal, drivers...)
- Better latencies in presence of losses.
- Smaller spikes in fq like packet schedulers, as retransmits
are not constrained by TCP Small Queues.
1 % packet losses are common today, and at 100Gbit speeds, this
translates to ~80,000 losses per second.
Losses are often correlated, and we see many retransmit events
leading to 1-MSS train of packets, at the time hosts are already
under stress.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-04-22 01:55:23 +08:00
|
|
|
if (tcp_retransmit_skb(sk, skb, segs))
|
2020-05-05 02:27:50 +08:00
|
|
|
break;
|
2013-07-13 02:33:04 +08:00
|
|
|
|
2016-09-22 07:16:14 +08:00
|
|
|
NET_ADD_STATS(sock_net(sk), mib_idx, tcp_skb_pcount(skb));
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2012-09-03 01:38:04 +08:00
|
|
|
if (tcp_in_cwnd_reduction(sk))
|
2011-08-22 04:21:57 +08:00
|
|
|
tp->prr_out += tcp_skb_pcount(skb);
|
|
|
|
|
|
2017-10-06 13:21:27 +08:00
|
|
|
if (skb == rtx_head &&
|
2017-01-13 14:11:33 +08:00
|
|
|
icsk->icsk_pending != ICSK_TIME_REO_TIMEOUT)
|
2020-05-05 02:27:50 +08:00
|
|
|
rearm_timer = true;
|
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
2020-05-05 02:27:50 +08:00
|
|
|
if (rearm_timer)
|
|
|
|
|
tcp_reset_xmit_timer(sk, ICSK_TIME_RETRANS,
|
|
|
|
|
inet_csk(sk)->icsk_rto,
|
|
|
|
|
TCP_RTO_MAX);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
|
2015-04-22 09:32:24 +08:00
|
|
|
/* We allow to exceed memory limits for FIN packets to expedite
|
|
|
|
|
* connection tear down and (memory) recovery.
|
2015-04-24 01:42:39 +08:00
|
|
|
* Otherwise tcp_send_fin() could be tempted to either delay FIN
|
|
|
|
|
* or even be forced to close flow without any FIN.
|
2015-05-16 03:39:26 +08:00
|
|
|
* In general, we want to allow one skb per socket to avoid hangs
|
|
|
|
|
* with edge trigger epoll()
|
2015-04-22 09:32:24 +08:00
|
|
|
*/
|
2015-05-16 03:39:26 +08:00
|
|
|
void sk_forced_mem_schedule(struct sock *sk, int size)
|
2015-04-22 09:32:24 +08:00
|
|
|
{
|
net: tcp_memcontrol: sanitize tcp memory accounting callbacks
There won't be a tcp control soft limit, so integrating the memcg code
into the global skmem limiting scheme complicates things unnecessarily.
Replace this with simple and clear charge and uncharge calls--hidden
behind a jump label--to account skb memory.
Note that this is not purely aesthetic: as a result of shoehorning the
per-memcg code into the same memory accounting functions that handle the
global level, the old code would compare the per-memcg consumption
against the smaller of the per-memcg limit and the global limit. This
allowed the total consumption of multiple sockets to exceed the global
limit, as long as the individual sockets stayed within bounds. After
this change, the code will always compare the per-memcg consumption to
the per-memcg limit, and the global consumption to the global limit, and
thus close this loophole.
Without a soft limit, the per-memcg memory pressure state in sockets is
generally questionable. However, we did it until now, so we continue to
enter it when the hard limit is hit, and packets are dropped, to let
other sockets in the cgroup know that they shouldn't grow their transmit
windows, either. However, keep it simple in the new callback model and
leave memory pressure lazily when the next packet is accepted (as
opposed to doing it synchroneously when packets are processed). When
packets are dropped, network performance will already be in the toilet,
so that should be a reasonable trade-off.
As described above, consumption is now checked on the per-memcg level
and the global level separately. Likewise, memory pressure states are
maintained on both the per-memcg level and the global level, and a
socket is considered under pressure when either level asserts as much.
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: Vladimir Davydov <vdavydov@virtuozzo.com>
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-01-15 07:21:14 +08:00
|
|
|
int amt;
|
2015-04-22 09:32:24 +08:00
|
|
|
|
|
|
|
|
if (size <= sk->sk_forward_alloc)
|
|
|
|
|
return;
|
|
|
|
|
amt = sk_mem_pages(size);
|
|
|
|
|
sk->sk_forward_alloc += amt * SK_MEM_QUANTUM;
|
net: tcp_memcontrol: sanitize tcp memory accounting callbacks
There won't be a tcp control soft limit, so integrating the memcg code
into the global skmem limiting scheme complicates things unnecessarily.
Replace this with simple and clear charge and uncharge calls--hidden
behind a jump label--to account skb memory.
Note that this is not purely aesthetic: as a result of shoehorning the
per-memcg code into the same memory accounting functions that handle the
global level, the old code would compare the per-memcg consumption
against the smaller of the per-memcg limit and the global limit. This
allowed the total consumption of multiple sockets to exceed the global
limit, as long as the individual sockets stayed within bounds. After
this change, the code will always compare the per-memcg consumption to
the per-memcg limit, and the global consumption to the global limit, and
thus close this loophole.
Without a soft limit, the per-memcg memory pressure state in sockets is
generally questionable. However, we did it until now, so we continue to
enter it when the hard limit is hit, and packets are dropped, to let
other sockets in the cgroup know that they shouldn't grow their transmit
windows, either. However, keep it simple in the new callback model and
leave memory pressure lazily when the next packet is accepted (as
opposed to doing it synchroneously when packets are processed). When
packets are dropped, network performance will already be in the toilet,
so that should be a reasonable trade-off.
As described above, consumption is now checked on the per-memcg level
and the global level separately. Likewise, memory pressure states are
maintained on both the per-memcg level and the global level, and a
socket is considered under pressure when either level asserts as much.
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: Vladimir Davydov <vdavydov@virtuozzo.com>
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-01-15 07:21:14 +08:00
|
|
|
sk_memory_allocated_add(sk, amt);
|
|
|
|
|
|
2016-01-15 07:21:17 +08:00
|
|
|
if (mem_cgroup_sockets_enabled && sk->sk_memcg)
|
|
|
|
|
mem_cgroup_charge_skmem(sk->sk_memcg, amt);
|
2015-04-22 09:32:24 +08:00
|
|
|
}
|
|
|
|
|
|
2015-04-24 01:42:39 +08:00
|
|
|
/* Send a FIN. The caller locks the socket for us.
|
|
|
|
|
* We should try to send a FIN packet really hard, but eventually give up.
|
2005-04-17 06:20:36 +08:00
|
|
|
*/
|
|
|
|
|
void tcp_send_fin(struct sock *sk)
|
|
|
|
|
{
|
2019-12-13 04:55:30 +08:00
|
|
|
struct sk_buff *skb, *tskb, *tail = tcp_write_queue_tail(sk);
|
2007-02-09 22:24:47 +08:00
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
|
2015-04-24 01:42:39 +08:00
|
|
|
/* Optimization, tack on the FIN if we have one skb in write queue and
|
|
|
|
|
* this skb was not yet sent, or we are under memory pressure.
|
|
|
|
|
* Note: in the latter case, FIN packet will be sent after a timeout,
|
|
|
|
|
* as TCP stack thinks it has already been transmitted.
|
2005-04-17 06:20:36 +08:00
|
|
|
*/
|
2019-12-13 04:55:30 +08:00
|
|
|
tskb = tail;
|
2017-10-06 13:21:27 +08:00
|
|
|
if (!tskb && tcp_under_memory_pressure(sk))
|
|
|
|
|
tskb = skb_rb_last(&sk->tcp_rtx_queue);
|
|
|
|
|
|
|
|
|
|
if (tskb) {
|
2015-04-24 01:42:39 +08:00
|
|
|
TCP_SKB_CB(tskb)->tcp_flags |= TCPHDR_FIN;
|
|
|
|
|
TCP_SKB_CB(tskb)->end_seq++;
|
2005-04-17 06:20:36 +08:00
|
|
|
tp->write_seq++;
|
2019-12-13 04:55:30 +08:00
|
|
|
if (!tail) {
|
2015-04-24 01:42:39 +08:00
|
|
|
/* This means tskb was already sent.
|
|
|
|
|
* Pretend we included the FIN on previous transmit.
|
|
|
|
|
* We need to set tp->snd_nxt to the value it would have
|
|
|
|
|
* if FIN had been sent. This is because retransmit path
|
|
|
|
|
* does not change tp->snd_nxt.
|
|
|
|
|
*/
|
2019-10-11 11:17:42 +08:00
|
|
|
WRITE_ONCE(tp->snd_nxt, tp->snd_nxt + 1);
|
2015-04-24 01:42:39 +08:00
|
|
|
return;
|
|
|
|
|
}
|
2005-04-17 06:20:36 +08:00
|
|
|
} else {
|
2015-04-24 01:42:39 +08:00
|
|
|
skb = alloc_skb_fclone(MAX_TCP_HEADER, sk->sk_allocation);
|
2019-04-04 22:46:03 +08:00
|
|
|
if (unlikely(!skb))
|
2015-04-24 01:42:39 +08:00
|
|
|
return;
|
2019-04-04 22:46:03 +08:00
|
|
|
|
2017-10-05 03:59:58 +08:00
|
|
|
INIT_LIST_HEAD(&skb->tcp_tsorted_anchor);
|
2015-04-22 09:32:24 +08:00
|
|
|
skb_reserve(skb, MAX_TCP_HEADER);
|
2015-05-16 03:39:26 +08:00
|
|
|
sk_forced_mem_schedule(sk, skb->truesize);
|
2005-04-17 06:20:36 +08:00
|
|
|
/* FIN eats a sequence byte, write_seq advanced by tcp_queue_skb(). */
|
2008-01-04 12:39:01 +08:00
|
|
|
tcp_init_nondata_skb(skb, tp->write_seq,
|
2010-06-12 22:01:43 +08:00
|
|
|
TCPHDR_ACK | TCPHDR_FIN);
|
2005-04-17 06:20:36 +08:00
|
|
|
tcp_queue_skb(sk, skb);
|
|
|
|
|
}
|
2015-04-24 01:42:39 +08:00
|
|
|
__tcp_push_pending_frames(sk, tcp_current_mss(sk), TCP_NAGLE_OFF);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* We get here when a process closes a file descriptor (either due to
|
|
|
|
|
* an explicit close() or as a byproduct of exit()'ing) and there
|
|
|
|
|
* was unread data in the receive queue. This behavior is recommended
|
2007-04-29 12:21:46 +08:00
|
|
|
* by RFC 2525, section 2.17. -DaveM
|
2005-04-17 06:20:36 +08:00
|
|
|
*/
|
2005-10-07 14:46:04 +08:00
|
|
|
void tcp_send_active_reset(struct sock *sk, gfp_t priority)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
|
|
|
|
struct sk_buff *skb;
|
|
|
|
|
|
2017-04-06 23:05:49 +08:00
|
|
|
TCP_INC_STATS(sock_net(sk), TCP_MIB_OUTRSTS);
|
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/* NOTE: No TCP options attached and we never retransmit this. */
|
|
|
|
|
skb = alloc_skb(MAX_TCP_HEADER, priority);
|
|
|
|
|
if (!skb) {
|
2008-07-17 11:30:14 +08:00
|
|
|
NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPABORTFAILED);
|
2005-04-17 06:20:36 +08:00
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Reserve space for headers and prepare control bits. */
|
|
|
|
|
skb_reserve(skb, MAX_TCP_HEADER);
|
2008-01-04 12:39:01 +08:00
|
|
|
tcp_init_nondata_skb(skb, tcp_acceptable_seq(sk),
|
2010-06-12 22:01:43 +08:00
|
|
|
TCPHDR_ACK | TCPHDR_RST);
|
2017-05-17 05:00:14 +08:00
|
|
|
tcp_mstamp_refresh(tcp_sk(sk));
|
2005-04-17 06:20:36 +08:00
|
|
|
/* Send it off. */
|
2005-12-07 08:24:52 +08:00
|
|
|
if (tcp_transmit_skb(sk, skb, 0, priority))
|
2008-07-17 11:30:14 +08:00
|
|
|
NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPABORTFAILED);
|
2017-10-24 00:20:24 +08:00
|
|
|
|
|
|
|
|
/* skb of trace_tcp_send_reset() keeps the skb that caused RST,
|
|
|
|
|
* skb here is different to the troublesome skb, so use NULL
|
|
|
|
|
*/
|
|
|
|
|
trace_tcp_send_reset(sk, NULL);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
|
2009-07-22 07:00:40 +08:00
|
|
|
/* Send a crossed SYN-ACK during socket establishment.
|
|
|
|
|
* WARNING: This routine must only be called when we have already sent
|
2005-04-17 06:20:36 +08:00
|
|
|
* a SYN packet that crossed the incoming SYN that caused this routine
|
|
|
|
|
* to get called. If this assumption fails then the initial rcv_wnd
|
|
|
|
|
* and rcv_wscale values will not be correct.
|
|
|
|
|
*/
|
|
|
|
|
int tcp_send_synack(struct sock *sk)
|
|
|
|
|
{
|
2008-01-01 06:57:14 +08:00
|
|
|
struct sk_buff *skb;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2017-10-06 13:21:27 +08:00
|
|
|
skb = tcp_rtx_queue_head(sk);
|
2015-04-03 16:17:26 +08:00
|
|
|
if (!skb || !(TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN)) {
|
2017-10-06 13:21:27 +08:00
|
|
|
pr_err("%s: wrong queue state\n", __func__);
|
2005-04-17 06:20:36 +08:00
|
|
|
return -EFAULT;
|
|
|
|
|
}
|
2011-09-28 01:25:05 +08:00
|
|
|
if (!(TCP_SKB_CB(skb)->tcp_flags & TCPHDR_ACK)) {
|
2005-04-17 06:20:36 +08:00
|
|
|
if (skb_cloned(skb)) {
|
2017-10-05 03:59:58 +08:00
|
|
|
struct sk_buff *nskb;
|
|
|
|
|
|
|
|
|
|
tcp_skb_tsorted_save(skb) {
|
|
|
|
|
nskb = skb_copy(skb, GFP_ATOMIC);
|
|
|
|
|
} tcp_skb_tsorted_restore(skb);
|
2015-04-03 16:17:26 +08:00
|
|
|
if (!nskb)
|
2005-04-17 06:20:36 +08:00
|
|
|
return -ENOMEM;
|
2017-10-05 03:59:58 +08:00
|
|
|
INIT_LIST_HEAD(&nskb->tcp_tsorted_anchor);
|
2020-01-23 13:03:00 +08:00
|
|
|
tcp_highest_sack_replace(sk, skb, nskb);
|
2017-10-06 13:21:27 +08:00
|
|
|
tcp_rtx_queue_unlink_and_free(skb, sk);
|
2014-09-23 07:29:32 +08:00
|
|
|
__skb_header_release(nskb);
|
2017-10-06 13:21:27 +08:00
|
|
|
tcp_rbtree_insert(&sk->tcp_rtx_queue, nskb);
|
2019-10-11 11:17:46 +08:00
|
|
|
sk_wmem_queued_add(sk, nskb->truesize);
|
2007-12-31 16:11:19 +08:00
|
|
|
sk_mem_charge(sk, nskb->truesize);
|
2005-04-17 06:20:36 +08:00
|
|
|
skb = nskb;
|
|
|
|
|
}
|
|
|
|
|
|
2011-09-28 01:25:05 +08:00
|
|
|
TCP_SKB_CB(skb)->tcp_flags |= TCPHDR_ACK;
|
2014-09-29 19:08:30 +08:00
|
|
|
tcp_ecn_send_synack(sk, skb);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
2005-12-07 08:24:52 +08:00
|
|
|
return tcp_transmit_skb(sk, skb, 1, GFP_ATOMIC);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
|
2012-06-04 04:33:21 +08:00
|
|
|
/**
|
2020-08-21 03:00:52 +08:00
|
|
|
* tcp_make_synack - Allocate one skb and build a SYNACK packet.
|
|
|
|
|
* @sk: listener socket
|
|
|
|
|
* @dst: dst entry attached to the SYNACK. It is consumed and caller
|
|
|
|
|
* should not use it again.
|
|
|
|
|
* @req: request_sock pointer
|
|
|
|
|
* @foc: cookie for tcp fast open
|
|
|
|
|
* @synack_type: Type of synack to prepare
|
|
|
|
|
* @syn_skb: SYN packet just received. It could be NULL for rtx case.
|
2012-06-04 04:33:21 +08:00
|
|
|
*/
|
2015-09-25 22:39:19 +08:00
|
|
|
struct sk_buff *tcp_make_synack(const struct sock *sk, struct dst_entry *dst,
|
2009-12-03 02:07:39 +08:00
|
|
|
struct request_sock *req,
|
2015-10-03 02:43:35 +08:00
|
|
|
struct tcp_fastopen_cookie *foc,
|
2020-08-21 03:00:52 +08:00
|
|
|
enum tcp_synack_type synack_type,
|
|
|
|
|
struct sk_buff *syn_skb)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
[NET] Generalise TCP's struct open_request minisock infrastructure
Kept this first changeset minimal, without changing existing names to
ease peer review.
Basicaly tcp_openreq_alloc now receives the or_calltable, that in turn
has two new members:
->slab, that replaces tcp_openreq_cachep
->obj_size, to inform the size of the openreq descendant for
a specific protocol
The protocol specific fields in struct open_request were moved to a
class hierarchy, with the things that are common to all connection
oriented PF_INET protocols in struct inet_request_sock, the TCP ones
in tcp_request_sock, that is an inet_request_sock, that is an
open_request.
I.e. this uses the same approach used for the struct sock class
hierarchy, with sk_prot indicating if the protocol wants to use the
open_request infrastructure by filling in sk_prot->rsk_prot with an
or_calltable.
Results? Performance is improved and TCP v4 now uses only 64 bytes per
open request minisock, down from 96 without this patch :-)
Next changeset will rename some of the structs, fields and functions
mentioned above, struct or_calltable is way unclear, better name it
struct request_sock_ops, s/struct open_request/struct request_sock/g,
etc.
Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-19 13:46:52 +08:00
|
|
|
struct inet_request_sock *ireq = inet_rsk(req);
|
2015-09-25 22:39:19 +08:00
|
|
|
const struct tcp_sock *tp = tcp_sk(sk);
|
2015-03-25 06:58:52 +08:00
|
|
|
struct tcp_md5sig_key *md5 = NULL;
|
2015-09-25 22:39:19 +08:00
|
|
|
struct tcp_out_options opts;
|
|
|
|
|
struct sk_buff *skb;
|
2009-12-03 02:23:05 +08:00
|
|
|
int tcp_header_size;
|
2015-09-25 22:39:19 +08:00
|
|
|
struct tcphdr *th;
|
2008-09-21 15:21:51 +08:00
|
|
|
int mss;
|
tcp: add optional per socket transmit delay
Adding delays to TCP flows is crucial for studying behavior
of TCP stacks, including congestion control modules.
Linux offers netem module, but it has unpractical constraints :
- Need root access to change qdisc
- Hard to setup on egress if combined with non trivial qdisc like FQ
- Single delay for all flows.
EDT (Earliest Departure Time) adoption in TCP stack allows us
to enable a per socket delay at a very small cost.
Networking tools can now establish thousands of flows, each of them
with a different delay, simulating real world conditions.
This requires FQ packet scheduler or a EDT-enabled NIC.
This patchs adds TCP_TX_DELAY socket option, to set a delay in
usec units.
unsigned int tx_delay = 10000; /* 10 msec */
setsockopt(fd, SOL_TCP, TCP_TX_DELAY, &tx_delay, sizeof(tx_delay));
Note that FQ packet scheduler limits might need some tweaking :
man tc-fq
PARAMETERS
limit
Hard limit on the real queue size. When this limit is
reached, new packets are dropped. If the value is lowered,
packets are dropped so that the new limit is met. Default
is 10000 packets.
flow_limit
Hard limit on the maximum number of packets queued per
flow. Default value is 100.
Use of TCP_TX_DELAY option will increase number of skbs in FQ qdisc,
so packets would be dropped if any of the previous limit is hit.
Use of a jump label makes this support runtime-free, for hosts
never using the option.
Also note that TSQ (TCP Small Queues) limits are slightly changed
with this patch : we need to account that skbs artificially delayed
wont stop us providind more skbs to feed the pipe (netem uses
skb_orphan_partial() for this purpose, but FQ can not use this trick)
Because of that, using big delays might very well trigger
old bugs in TSO auto defer logic and/or sndbuf limited detection.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2019-06-13 02:57:25 +08:00
|
|
|
u64 now;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2015-10-03 02:43:35 +08:00
|
|
|
skb = alloc_skb(MAX_TCP_HEADER, GFP_ATOMIC);
|
2012-06-04 04:33:21 +08:00
|
|
|
if (unlikely(!skb)) {
|
|
|
|
|
dst_release(dst);
|
2005-04-17 06:20:36 +08:00
|
|
|
return NULL;
|
2012-06-04 04:33:21 +08:00
|
|
|
}
|
2005-04-17 06:20:36 +08:00
|
|
|
/* Reserve space for headers. */
|
|
|
|
|
skb_reserve(skb, MAX_TCP_HEADER);
|
|
|
|
|
|
2016-04-14 13:05:39 +08:00
|
|
|
switch (synack_type) {
|
|
|
|
|
case TCP_SYNACK_NORMAL:
|
2015-11-02 07:36:55 +08:00
|
|
|
skb_set_owner_w(skb, req_to_sk(req));
|
2016-04-14 13:05:39 +08:00
|
|
|
break;
|
|
|
|
|
case TCP_SYNACK_COOKIE:
|
|
|
|
|
/* Under synflood, we do not attach skb to a socket,
|
|
|
|
|
* to avoid false sharing.
|
|
|
|
|
*/
|
|
|
|
|
break;
|
|
|
|
|
case TCP_SYNACK_FASTOPEN:
|
2015-10-03 02:43:35 +08:00
|
|
|
/* sk is a const pointer, because we want to express multiple
|
|
|
|
|
* cpu might call us concurrently.
|
|
|
|
|
* sk->sk_wmem_alloc in an atomic, we can promote to rw.
|
|
|
|
|
*/
|
|
|
|
|
skb_set_owner_w(skb, (struct sock *)sk);
|
2016-04-14 13:05:39 +08:00
|
|
|
break;
|
2015-10-03 02:43:35 +08:00
|
|
|
}
|
2012-06-04 04:33:21 +08:00
|
|
|
skb_dst_set(skb, dst);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2017-02-03 00:04:56 +08:00
|
|
|
mss = tcp_mss_clamp(tp, dst_metric_advmss(dst));
|
2008-09-21 15:21:51 +08:00
|
|
|
|
2008-07-19 15:04:31 +08:00
|
|
|
memset(&opts, 0, sizeof(opts));
|
tcp: add optional per socket transmit delay
Adding delays to TCP flows is crucial for studying behavior
of TCP stacks, including congestion control modules.
Linux offers netem module, but it has unpractical constraints :
- Need root access to change qdisc
- Hard to setup on egress if combined with non trivial qdisc like FQ
- Single delay for all flows.
EDT (Earliest Departure Time) adoption in TCP stack allows us
to enable a per socket delay at a very small cost.
Networking tools can now establish thousands of flows, each of them
with a different delay, simulating real world conditions.
This requires FQ packet scheduler or a EDT-enabled NIC.
This patchs adds TCP_TX_DELAY socket option, to set a delay in
usec units.
unsigned int tx_delay = 10000; /* 10 msec */
setsockopt(fd, SOL_TCP, TCP_TX_DELAY, &tx_delay, sizeof(tx_delay));
Note that FQ packet scheduler limits might need some tweaking :
man tc-fq
PARAMETERS
limit
Hard limit on the real queue size. When this limit is
reached, new packets are dropped. If the value is lowered,
packets are dropped so that the new limit is met. Default
is 10000 packets.
flow_limit
Hard limit on the maximum number of packets queued per
flow. Default value is 100.
Use of TCP_TX_DELAY option will increase number of skbs in FQ qdisc,
so packets would be dropped if any of the previous limit is hit.
Use of a jump label makes this support runtime-free, for hosts
never using the option.
Also note that TSQ (TCP Small Queues) limits are slightly changed
with this patch : we need to account that skbs artificially delayed
wont stop us providind more skbs to feed the pipe (netem uses
skb_orphan_partial() for this purpose, but FQ can not use this trick)
Because of that, using big delays might very well trigger
old bugs in TSO auto defer logic and/or sndbuf limited detection.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2019-06-13 02:57:25 +08:00
|
|
|
now = tcp_clock_ns();
|
2008-10-27 14:10:12 +08:00
|
|
|
#ifdef CONFIG_SYN_COOKIES
|
2020-07-31 03:25:50 +08:00
|
|
|
if (unlikely(synack_type == TCP_SYNACK_COOKIE && ireq->tstamp_ok))
|
2019-11-08 03:51:18 +08:00
|
|
|
skb->skb_mstamp_ns = cookie_init_timestamp(req, now);
|
2008-10-27 14:10:12 +08:00
|
|
|
else
|
|
|
|
|
#endif
|
2019-04-30 06:46:15 +08:00
|
|
|
{
|
tcp: add optional per socket transmit delay
Adding delays to TCP flows is crucial for studying behavior
of TCP stacks, including congestion control modules.
Linux offers netem module, but it has unpractical constraints :
- Need root access to change qdisc
- Hard to setup on egress if combined with non trivial qdisc like FQ
- Single delay for all flows.
EDT (Earliest Departure Time) adoption in TCP stack allows us
to enable a per socket delay at a very small cost.
Networking tools can now establish thousands of flows, each of them
with a different delay, simulating real world conditions.
This requires FQ packet scheduler or a EDT-enabled NIC.
This patchs adds TCP_TX_DELAY socket option, to set a delay in
usec units.
unsigned int tx_delay = 10000; /* 10 msec */
setsockopt(fd, SOL_TCP, TCP_TX_DELAY, &tx_delay, sizeof(tx_delay));
Note that FQ packet scheduler limits might need some tweaking :
man tc-fq
PARAMETERS
limit
Hard limit on the real queue size. When this limit is
reached, new packets are dropped. If the value is lowered,
packets are dropped so that the new limit is met. Default
is 10000 packets.
flow_limit
Hard limit on the maximum number of packets queued per
flow. Default value is 100.
Use of TCP_TX_DELAY option will increase number of skbs in FQ qdisc,
so packets would be dropped if any of the previous limit is hit.
Use of a jump label makes this support runtime-free, for hosts
never using the option.
Also note that TSQ (TCP Small Queues) limits are slightly changed
with this patch : we need to account that skbs artificially delayed
wont stop us providind more skbs to feed the pipe (netem uses
skb_orphan_partial() for this purpose, but FQ can not use this trick)
Because of that, using big delays might very well trigger
old bugs in TSO auto defer logic and/or sndbuf limited detection.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2019-06-13 02:57:25 +08:00
|
|
|
skb->skb_mstamp_ns = now;
|
2019-04-30 06:46:15 +08:00
|
|
|
if (!tcp_rsk(req)->snt_synack) /* Timestamp first SYNACK */
|
|
|
|
|
tcp_rsk(req)->snt_synack = tcp_skb_timestamp_us(skb);
|
|
|
|
|
}
|
2015-03-25 06:58:52 +08:00
|
|
|
|
|
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
|
|
|
|
rcu_read_lock();
|
2015-03-25 06:58:56 +08:00
|
|
|
md5 = tcp_rsk(req)->af_specific->req_md5_lookup(sk, req_to_sk(req));
|
2015-03-25 06:58:52 +08:00
|
|
|
#endif
|
2015-09-16 06:24:20 +08:00
|
|
|
skb_set_hash(skb, tcp_rsk(req)->txhash, PKT_HASH_TYPE_L4);
|
2020-08-21 03:00:52 +08:00
|
|
|
/* bpf program will be interested in the tcp_flags */
|
|
|
|
|
TCP_SKB_CB(skb)->tcp_flags = TCPHDR_SYN | TCPHDR_ACK;
|
2017-10-25 17:01:45 +08:00
|
|
|
tcp_header_size = tcp_synack_options(sk, req, mss, skb, &opts, md5,
|
2020-08-21 03:00:52 +08:00
|
|
|
foc, synack_type,
|
|
|
|
|
syn_skb) + sizeof(*th);
|
2006-11-15 11:07:45 +08:00
|
|
|
|
2007-04-11 12:04:22 +08:00
|
|
|
skb_push(skb, tcp_header_size);
|
|
|
|
|
skb_reset_transport_header(skb);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2016-05-14 00:16:40 +08:00
|
|
|
th = (struct tcphdr *)skb->data;
|
2005-04-17 06:20:36 +08:00
|
|
|
memset(th, 0, sizeof(struct tcphdr));
|
|
|
|
|
th->syn = 1;
|
|
|
|
|
th->ack = 1;
|
2015-09-25 22:39:18 +08:00
|
|
|
tcp_ecn_make_synack(req, th);
|
2013-10-10 15:04:37 +08:00
|
|
|
th->source = htons(ireq->ir_num);
|
2013-10-10 06:21:29 +08:00
|
|
|
th->dest = ireq->ir_rmt_port;
|
2017-07-03 21:51:50 +08:00
|
|
|
skb->mark = ireq->ir_mark;
|
2017-11-03 03:30:25 +08:00
|
|
|
skb->ip_summed = CHECKSUM_PARTIAL;
|
|
|
|
|
th->seq = htonl(tcp_rsk(req)->snt_isn);
|
2012-08-31 20:29:12 +08:00
|
|
|
/* XXX data is queued and acked as is. No buffer/window check */
|
|
|
|
|
th->ack_seq = htonl(tcp_rsk(req)->rcv_nxt);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
|
/* RFC1323: The window in SYN & SYN/ACK segments is never scaled. */
|
2015-10-09 10:33:23 +08:00
|
|
|
th->window = htons(min(req->rsk_rcv_wnd, 65535U));
|
2015-09-25 22:39:19 +08:00
|
|
|
tcp_options_write((__be32 *)(th + 1), NULL, &opts);
|
2005-04-17 06:20:36 +08:00
|
|
|
th->doff = (tcp_header_size >> 2);
|
2016-04-28 07:44:32 +08:00
|
|
|
__TCP_INC_STATS(sock_net(sk), TCP_MIB_OUTSEGS);
|
2006-11-15 11:07:45 +08:00
|
|
|
|
|
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
|
|
|
|
/* Okay, we have all we need - do the md5 hash if needed */
|
2015-03-25 06:58:52 +08:00
|
|
|
if (md5)
|
2009-12-03 02:23:05 +08:00
|
|
|
tcp_rsk(req)->af_specific->calc_md5_hash(opts.hash_location,
|
2015-03-25 06:58:55 +08:00
|
|
|
md5, req_to_sk(req), skb);
|
2015-03-25 06:58:52 +08:00
|
|
|
rcu_read_unlock();
|
2006-11-15 11:07:45 +08:00
|
|
|
#endif
|
|
|
|
|
|
2020-08-21 03:00:52 +08:00
|
|
|
bpf_skops_write_hdr_opt((struct sock *)sk, skb, req, syn_skb,
|
|
|
|
|
synack_type, &opts);
|
|
|
|
|
|
tcp: add optional per socket transmit delay
Adding delays to TCP flows is crucial for studying behavior
of TCP stacks, including congestion control modules.
Linux offers netem module, but it has unpractical constraints :
- Need root access to change qdisc
- Hard to setup on egress if combined with non trivial qdisc like FQ
- Single delay for all flows.
EDT (Earliest Departure Time) adoption in TCP stack allows us
to enable a per socket delay at a very small cost.
Networking tools can now establish thousands of flows, each of them
with a different delay, simulating real world conditions.
This requires FQ packet scheduler or a EDT-enabled NIC.
This patchs adds TCP_TX_DELAY socket option, to set a delay in
usec units.
unsigned int tx_delay = 10000; /* 10 msec */
setsockopt(fd, SOL_TCP, TCP_TX_DELAY, &tx_delay, sizeof(tx_delay));
Note that FQ packet scheduler limits might need some tweaking :
man tc-fq
PARAMETERS
limit
Hard limit on the real queue size. When this limit is
reached, new packets are dropped. If the value is lowered,
packets are dropped so that the new limit is met. Default
is 10000 packets.
flow_limit
Hard limit on the maximum number of packets queued per
flow. Default value is 100.
Use of TCP_TX_DELAY option will increase number of skbs in FQ qdisc,
so packets would be dropped if any of the previous limit is hit.
Use of a jump label makes this support runtime-free, for hosts
never using the option.
Also note that TSQ (TCP Small Queues) limits are slightly changed
with this patch : we need to account that skbs artificially delayed
wont stop us providind more skbs to feed the pipe (netem uses
skb_orphan_partial() for this purpose, but FQ can not use this trick)
Because of that, using big delays might very well trigger
old bugs in TSO auto defer logic and/or sndbuf limited detection.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2019-06-13 02:57:25 +08:00
|
|
|
skb->skb_mstamp_ns = now;
|
|
|
|
|
tcp_add_tx_delay(skb, tp);
|
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
return skb;
|
|
|
|
|
}
|
2010-07-10 05:22:10 +08:00
|
|
|
EXPORT_SYMBOL(tcp_make_synack);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
net: tcp: add per route congestion control
This work adds the possibility to define a per route/destination
congestion control algorithm. Generally, this opens up the possibility
for a machine with different links to enforce specific congestion
control algorithms with optimal strategies for each of them based
on their network characteristics, even transparently for a single
application listening on all links.
For our specific use case, this additionally facilitates deployment
of DCTCP, for example, applications can easily serve internal
traffic/dsts in DCTCP and external one with CUBIC. Other scenarios
would also allow for utilizing e.g. long living, low priority
background flows for certain destinations/routes while still being
able for normal traffic to utilize the default congestion control
algorithm. We also thought about a per netns setting (where different
defaults are possible), but given its actually a link specific
property, we argue that a per route/destination setting is the most
natural and flexible.
The administrator can utilize this through ip-route(8) by appending
"congctl [lock] <name>", where <name> denotes the name of a
congestion control algorithm and the optional lock parameter allows
to enforce the given algorithm so that applications in user space
would not be allowed to overwrite that algorithm for that destination.
The dst metric lookups are being done when a dst entry is already
available in order to avoid a costly lookup and still before the
algorithms are being initialized, thus overhead is very low when the
feature is not being used. While the client side would need to drop
the current reference on the module, on server side this can actually
even be avoided as we just got a flat-copied socket clone.
Joint work with Florian Westphal.
Suggested-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Daniel Borkmann <dborkman@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-01-06 06:57:48 +08:00
|
|
|
static void tcp_ca_dst_init(struct sock *sk, const struct dst_entry *dst)
|
|
|
|
|
{
|
|
|
|
|
struct inet_connection_sock *icsk = inet_csk(sk);
|
|
|
|
|
const struct tcp_congestion_ops *ca;
|
|
|
|
|
u32 ca_key = dst_metric(dst, RTAX_CC_ALGO);
|
|
|
|
|
|
|
|
|
|
if (ca_key == TCP_CA_UNSPEC)
|
|
|
|
|
return;
|
|
|
|
|
|
|
|
|
|
rcu_read_lock();
|
|
|
|
|
ca = tcp_ca_find_key(ca_key);
|
2020-01-09 08:35:08 +08:00
|
|
|
if (likely(ca && bpf_try_module_get(ca, ca->owner))) {
|
|
|
|
|
bpf_module_put(icsk->icsk_ca_ops, icsk->icsk_ca_ops->owner);
|
net: tcp: add per route congestion control
This work adds the possibility to define a per route/destination
congestion control algorithm. Generally, this opens up the possibility
for a machine with different links to enforce specific congestion
control algorithms with optimal strategies for each of them based
on their network characteristics, even transparently for a single
application listening on all links.
For our specific use case, this additionally facilitates deployment
of DCTCP, for example, applications can easily serve internal
traffic/dsts in DCTCP and external one with CUBIC. Other scenarios
would also allow for utilizing e.g. long living, low priority
background flows for certain destinations/routes while still being
able for normal traffic to utilize the default congestion control
algorithm. We also thought about a per netns setting (where different
defaults are possible), but given its actually a link specific
property, we argue that a per route/destination setting is the most
natural and flexible.
The administrator can utilize this through ip-route(8) by appending
"congctl [lock] <name>", where <name> denotes the name of a
congestion control algorithm and the optional lock parameter allows
to enforce the given algorithm so that applications in user space
would not be allowed to overwrite that algorithm for that destination.
The dst metric lookups are being done when a dst entry is already
available in order to avoid a costly lookup and still before the
algorithms are being initialized, thus overhead is very low when the
feature is not being used. While the client side would need to drop
the current reference on the module, on server side this can actually
even be avoided as we just got a flat-copied socket clone.
Joint work with Florian Westphal.
Suggested-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Daniel Borkmann <dborkman@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-01-06 06:57:48 +08:00
|
|
|
icsk->icsk_ca_dst_locked = tcp_ca_dst_locked(dst);
|
|
|
|
|
icsk->icsk_ca_ops = ca;
|
|
|
|
|
}
|
|
|
|
|
rcu_read_unlock();
|
|
|
|
|
}
|
|
|
|
|
|
2009-07-22 07:00:40 +08:00
|
|
|
/* Do all connect socket setups that can be done AF independent. */
|
2013-12-30 03:39:51 +08:00
|
|
|
static void tcp_connect_init(struct sock *sk)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
2011-10-21 17:22:42 +08:00
|
|
|
const struct dst_entry *dst = __sk_dst_get(sk);
|
2005-04-17 06:20:36 +08:00
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
__u8 rcv_wscale;
|
2017-07-01 11:02:44 +08:00
|
|
|
u32 rcv_wnd;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
|
/* We'll fix this up when we get a response from the other end.
|
|
|
|
|
* See tcp_input.c:tcp_rcv_state_process case TCP_SYN_SENT.
|
|
|
|
|
*/
|
2017-06-08 01:34:39 +08:00
|
|
|
tp->tcp_header_len = sizeof(struct tcphdr);
|
|
|
|
|
if (sock_net(sk)->ipv4.sysctl_tcp_timestamps)
|
|
|
|
|
tp->tcp_header_len += TCPOLEN_TSTAMP_ALIGNED;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2006-11-15 11:07:45 +08:00
|
|
|
#ifdef CONFIG_TCP_MD5SIG
|
2015-04-03 16:17:27 +08:00
|
|
|
if (tp->af_specific->md5_lookup(sk, sk))
|
2006-11-15 11:07:45 +08:00
|
|
|
tp->tcp_header_len += TCPOLEN_MD5SIG_ALIGNED;
|
|
|
|
|
#endif
|
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/* If user gave his TCP_MAXSEG, record it to clamp */
|
|
|
|
|
if (tp->rx_opt.user_mss)
|
|
|
|
|
tp->rx_opt.mss_clamp = tp->rx_opt.user_mss;
|
|
|
|
|
tp->max_window = 0;
|
2006-03-21 09:53:41 +08:00
|
|
|
tcp_mtup_init(sk);
|
2005-04-17 06:20:36 +08:00
|
|
|
tcp_sync_mss(sk, dst_mtu(dst));
|
|
|
|
|
|
net: tcp: add per route congestion control
This work adds the possibility to define a per route/destination
congestion control algorithm. Generally, this opens up the possibility
for a machine with different links to enforce specific congestion
control algorithms with optimal strategies for each of them based
on their network characteristics, even transparently for a single
application listening on all links.
For our specific use case, this additionally facilitates deployment
of DCTCP, for example, applications can easily serve internal
traffic/dsts in DCTCP and external one with CUBIC. Other scenarios
would also allow for utilizing e.g. long living, low priority
background flows for certain destinations/routes while still being
able for normal traffic to utilize the default congestion control
algorithm. We also thought about a per netns setting (where different
defaults are possible), but given its actually a link specific
property, we argue that a per route/destination setting is the most
natural and flexible.
The administrator can utilize this through ip-route(8) by appending
"congctl [lock] <name>", where <name> denotes the name of a
congestion control algorithm and the optional lock parameter allows
to enforce the given algorithm so that applications in user space
would not be allowed to overwrite that algorithm for that destination.
The dst metric lookups are being done when a dst entry is already
available in order to avoid a costly lookup and still before the
algorithms are being initialized, thus overhead is very low when the
feature is not being used. While the client side would need to drop
the current reference on the module, on server side this can actually
even be avoided as we just got a flat-copied socket clone.
Joint work with Florian Westphal.
Suggested-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Daniel Borkmann <dborkman@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-01-06 06:57:48 +08:00
|
|
|
tcp_ca_dst_init(sk, dst);
|
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
if (!tp->window_clamp)
|
|
|
|
|
tp->window_clamp = dst_metric(dst, RTAX_WINDOW);
|
2017-02-03 00:04:56 +08:00
|
|
|
tp->advmss = tcp_mss_clamp(tp, dst_metric_advmss(dst));
|
2008-09-21 15:21:51 +08:00
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
tcp_initialize_rcv_mss(sk);
|
|
|
|
|
|
tcp: allow effective reduction of TCP's rcv-buffer via setsockopt
Via setsockopt it is possible to reduce the socket RX buffer
(SO_RCVBUF). TCP method to select the initial window and window scaling
option in tcp_select_initial_window() currently misbehaves and do not
consider a reduced RX socket buffer via setsockopt.
Even though the server's RX buffer is reduced via setsockopt() to 256
byte (Initial Window 384 byte => 256 * 2 - (256 * 2 / 4)) the window
scale option is still 7:
192.168.1.38.40676 > 78.47.222.210.5001: Flags [S], seq 2577214362, win 5840, options [mss 1460,sackOK,TS val 338417 ecr 0,nop,wscale 0], length 0
78.47.222.210.5001 > 192.168.1.38.40676: Flags [S.], seq 1570631029, ack 2577214363, win 384, options [mss 1452,sackOK,TS val 2435248895 ecr 338417,nop,wscale 7], length 0
192.168.1.38.40676 > 78.47.222.210.5001: Flags [.], ack 1, win 5840, options [nop,nop,TS val 338421 ecr 2435248895], length 0
Within tcp_select_initial_window() the original space argument - a
representation of the rx buffer size - is expanded during
tcp_select_initial_window(). Only sysctl_tcp_rmem[2], sysctl_rmem_max
and window_clamp are considered to calculate the initial window.
This patch adjust the window_clamp argument if the user explicitly
reduce the receive buffer.
Signed-off-by: Hagen Paul Pfeifer <hagen@jauu.net>
Cc: David S. Miller <davem@davemloft.net>
Cc: Patrick McHardy <kaber@trash.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-08-19 14:33:05 +08:00
|
|
|
/* limit the window selection if the user enforce a smaller rx buffer */
|
|
|
|
|
if (sk->sk_userlocks & SOCK_RCVBUF_LOCK &&
|
|
|
|
|
(tp->window_clamp > tcp_full_space(sk) || tp->window_clamp == 0))
|
|
|
|
|
tp->window_clamp = tcp_full_space(sk);
|
|
|
|
|
|
2017-07-01 11:02:44 +08:00
|
|
|
rcv_wnd = tcp_rwnd_init_bpf(sk);
|
|
|
|
|
if (rcv_wnd == 0)
|
|
|
|
|
rcv_wnd = dst_metric(dst, RTAX_INITRWND);
|
|
|
|
|
|
2017-10-27 22:47:24 +08:00
|
|
|
tcp_select_initial_window(sk, tcp_full_space(sk),
|
2005-04-17 06:20:36 +08:00
|
|
|
tp->advmss - (tp->rx_opt.ts_recent_stamp ? tp->tcp_header_len - sizeof(struct tcphdr) : 0),
|
|
|
|
|
&tp->rcv_wnd,
|
|
|
|
|
&tp->window_clamp,
|
2017-06-08 01:34:38 +08:00
|
|
|
sock_net(sk)->ipv4.sysctl_tcp_window_scaling,
|
2009-12-15 19:15:28 +08:00
|
|
|
&rcv_wscale,
|
2017-07-01 11:02:44 +08:00
|
|
|
rcv_wnd);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
|
tp->rx_opt.rcv_wscale = rcv_wscale;
|
|
|
|
|
tp->rcv_ssthresh = tp->rcv_wnd;
|
|
|
|
|
|
|
|
|
|
sk->sk_err = 0;
|
|
|
|
|
sock_reset_flag(sk, SOCK_DONE);
|
|
|
|
|
tp->snd_wnd = 0;
|
2009-03-03 14:42:02 +08:00
|
|
|
tcp_init_wl(tp, 0);
|
2018-05-15 12:14:26 +08:00
|
|
|
tcp_write_queue_purge(sk);
|
2005-04-17 06:20:36 +08:00
|
|
|
tp->snd_una = tp->write_seq;
|
|
|
|
|
tp->snd_sml = tp->write_seq;
|
2008-10-08 05:43:06 +08:00
|
|
|
tp->snd_up = tp->write_seq;
|
2019-10-11 11:17:42 +08:00
|
|
|
WRITE_ONCE(tp->snd_nxt, tp->write_seq);
|
2012-04-19 11:40:39 +08:00
|
|
|
|
|
|
|
|
if (likely(!tp->repair))
|
|
|
|
|
tp->rcv_nxt = 0;
|
2013-08-27 16:20:40 +08:00
|
|
|
else
|
2017-05-17 05:00:07 +08:00
|
|
|
tp->rcv_tstamp = tcp_jiffies32;
|
2012-04-19 11:40:39 +08:00
|
|
|
tp->rcv_wup = tp->rcv_nxt;
|
2019-10-11 11:17:40 +08:00
|
|
|
WRITE_ONCE(tp->copied_seq, tp->rcv_nxt);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2017-07-01 11:02:42 +08:00
|
|
|
inet_csk(sk)->icsk_rto = tcp_timeout_init(sk);
|
2005-08-10 11:10:42 +08:00
|
|
|
inet_csk(sk)->icsk_retransmits = 0;
|
2005-04-17 06:20:36 +08:00
|
|
|
tcp_clear_retrans(tp);
|
|
|
|
|
}
|
|
|
|
|
|
2012-07-19 14:43:07 +08:00
|
|
|
static void tcp_connect_queue_skb(struct sock *sk, struct sk_buff *skb)
|
|
|
|
|
{
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
struct tcp_skb_cb *tcb = TCP_SKB_CB(skb);
|
|
|
|
|
|
|
|
|
|
tcb->end_seq += skb->len;
|
2014-09-23 07:29:32 +08:00
|
|
|
__skb_header_release(skb);
|
2019-10-11 11:17:46 +08:00
|
|
|
sk_wmem_queued_add(sk, skb->truesize);
|
2012-07-19 14:43:07 +08:00
|
|
|
sk_mem_charge(sk, skb->truesize);
|
2019-10-11 11:17:41 +08:00
|
|
|
WRITE_ONCE(tp->write_seq, tcb->end_seq);
|
2012-07-19 14:43:07 +08:00
|
|
|
tp->packets_out += tcp_skb_pcount(skb);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Build and send a SYN with data and (cached) Fast Open cookie. However,
|
|
|
|
|
* queue a data-only packet after the regular SYN, such that regular SYNs
|
|
|
|
|
* are retransmitted on timeouts. Also if the remote SYN-ACK acknowledges
|
|
|
|
|
* only the SYN sequence, the data are retransmitted in the first ACK.
|
|
|
|
|
* If cookie is not cached or other error occurs, falls back to send a
|
|
|
|
|
* regular SYN with Fast Open cookie request option.
|
|
|
|
|
*/
|
|
|
|
|
static int tcp_send_syn_data(struct sock *sk, struct sk_buff *syn)
|
|
|
|
|
{
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
struct tcp_fastopen_request *fo = tp->fastopen_req;
|
2017-01-24 02:59:20 +08:00
|
|
|
int space, err = 0;
|
2014-11-18 15:06:20 +08:00
|
|
|
struct sk_buff *syn_data;
|
2012-07-19 14:43:10 +08:00
|
|
|
|
2012-07-19 14:43:11 +08:00
|
|
|
tp->rx_opt.mss_clamp = tp->advmss; /* If MSS is not cached */
|
2017-01-24 02:59:20 +08:00
|
|
|
if (!tcp_fastopen_cookie_check(sk, &tp->rx_opt.mss_clamp, &fo->cookie))
|
2012-07-19 14:43:07 +08:00
|
|
|
goto fallback;
|
|
|
|
|
|
|
|
|
|
/* MSS for SYN-data is based on cached MSS and bounded by PMTU and
|
|
|
|
|
* user-MSS. Reserve maximum option space for middleboxes that add
|
|
|
|
|
* private TCP options. The cost is reduced data space in SYN :(
|
|
|
|
|
*/
|
2017-02-03 00:04:56 +08:00
|
|
|
tp->rx_opt.mss_clamp = tcp_mss_clamp(tp, tp->rx_opt.mss_clamp);
|
|
|
|
|
|
2013-02-22 16:59:06 +08:00
|
|
|
space = __tcp_mtu_to_mss(sk, inet_csk(sk)->icsk_pmtu_cookie) -
|
2012-07-19 14:43:07 +08:00
|
|
|
MAX_TCP_OPTION_SPACE;
|
|
|
|
|
|
2014-02-21 02:09:18 +08:00
|
|
|
space = min_t(size_t, space, fo->size);
|
|
|
|
|
|
|
|
|
|
/* limit to order-0 allocations */
|
|
|
|
|
space = min_t(size_t, space, SKB_MAX_HEAD(MAX_TCP_HEADER));
|
|
|
|
|
|
2015-05-20 04:26:55 +08:00
|
|
|
syn_data = sk_stream_alloc_skb(sk, space, sk->sk_allocation, false);
|
2014-11-18 15:06:20 +08:00
|
|
|
if (!syn_data)
|
2012-07-19 14:43:07 +08:00
|
|
|
goto fallback;
|
2014-11-18 15:06:20 +08:00
|
|
|
syn_data->ip_summed = CHECKSUM_PARTIAL;
|
|
|
|
|
memcpy(syn_data->cb, syn->cb, sizeof(syn->cb));
|
2015-12-17 05:53:10 +08:00
|
|
|
if (space) {
|
|
|
|
|
int copied = copy_from_iter(skb_put(syn_data, space), space,
|
|
|
|
|
&fo->data->msg_iter);
|
|
|
|
|
if (unlikely(!copied)) {
|
2017-10-19 05:20:30 +08:00
|
|
|
tcp_skb_tsorted_anchor_cleanup(syn_data);
|
2015-12-17 05:53:10 +08:00
|
|
|
kfree_skb(syn_data);
|
|
|
|
|
goto fallback;
|
|
|
|
|
}
|
|
|
|
|
if (copied != space) {
|
|
|
|
|
skb_trim(syn_data, copied);
|
|
|
|
|
space = copied;
|
|
|
|
|
}
|
2019-01-26 00:17:23 +08:00
|
|
|
skb_zcopy_set(syn_data, fo->uarg, NULL);
|
2014-11-29 02:40:20 +08:00
|
|
|
}
|
2014-11-18 15:06:20 +08:00
|
|
|
/* No more data pending in inet_wait_for_connect() */
|
|
|
|
|
if (space == fo->size)
|
|
|
|
|
fo->data = NULL;
|
|
|
|
|
fo->copied = space;
|
2012-07-19 14:43:07 +08:00
|
|
|
|
2014-11-18 15:06:20 +08:00
|
|
|
tcp_connect_queue_skb(sk, syn_data);
|
2016-11-28 15:07:14 +08:00
|
|
|
if (syn_data->len)
|
|
|
|
|
tcp_chrono_start(sk, TCP_CHRONO_BUSY);
|
2012-07-19 14:43:07 +08:00
|
|
|
|
2014-11-18 15:06:20 +08:00
|
|
|
err = tcp_transmit_skb(sk, syn_data, 1, sk->sk_allocation);
|
2012-07-19 14:43:07 +08:00
|
|
|
|
2018-09-21 23:51:50 +08:00
|
|
|
syn->skb_mstamp_ns = syn_data->skb_mstamp_ns;
|
2014-03-10 08:36:02 +08:00
|
|
|
|
2014-11-18 15:06:20 +08:00
|
|
|
/* Now full SYN+DATA was cloned and sent (or not),
|
|
|
|
|
* remove the SYN from the original skb (syn_data)
|
|
|
|
|
* we keep in write queue in case of a retransmit, as we
|
|
|
|
|
* also have the SYN packet (with no data) in the same queue.
|
|
|
|
|
*/
|
|
|
|
|
TCP_SKB_CB(syn_data)->seq++;
|
|
|
|
|
TCP_SKB_CB(syn_data)->tcp_flags = TCPHDR_ACK | TCPHDR_PSH;
|
|
|
|
|
if (!err) {
|
2012-07-19 14:43:11 +08:00
|
|
|
tp->syn_data = (fo->copied > 0);
|
2017-10-06 13:21:27 +08:00
|
|
|
tcp_rbtree_insert(&sk->tcp_rtx_queue, syn_data);
|
2014-03-04 04:31:36 +08:00
|
|
|
NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPORIGDATASENT);
|
2012-07-19 14:43:07 +08:00
|
|
|
goto done;
|
|
|
|
|
}
|
|
|
|
|
|
2017-10-06 13:21:27 +08:00
|
|
|
/* data was not sent, put it in write_queue */
|
|
|
|
|
__skb_queue_tail(&sk->sk_write_queue, syn_data);
|
tcp: fastopen: fix on syn-data transmit failure
Our recent change exposed a bug in TCP Fastopen Client that syzkaller
found right away [1]
When we prepare skb with SYN+DATA, we attempt to transmit it,
and we update socket state as if the transmit was a success.
In socket RTX queue we have two skbs, one with the SYN alone,
and a second one containing the DATA.
When (malicious) ACK comes in, we now complain that second one had no
skb_mstamp.
The proper fix is to make sure that if the transmit failed, we do not
pretend we sent the DATA skb, and make it our send_head.
When 3WHS completes, we can now send the DATA right away, without having
to wait for a timeout.
[1]
WARNING: CPU: 0 PID: 100189 at net/ipv4/tcp_input.c:3117 tcp_clean_rtx_queue+0x2057/0x2ab0 net/ipv4/tcp_input.c:3117()
WARN_ON_ONCE(last_ackt == 0);
Modules linked in:
CPU: 0 PID: 100189 Comm: syz-executor1 Not tainted
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
0000000000000000 ffff8800b35cb1d8 ffffffff81cad00d 0000000000000000
ffffffff828a4347 ffff88009f86c080 ffffffff8316eb20 0000000000000d7f
ffff8800b35cb220 ffffffff812c33c2 ffff8800baad2440 00000009d46575c0
Call Trace:
[<ffffffff81cad00d>] __dump_stack
[<ffffffff81cad00d>] dump_stack+0xc1/0x124
[<ffffffff812c33c2>] warn_slowpath_common+0xe2/0x150
[<ffffffff812c361e>] warn_slowpath_null+0x2e/0x40
[<ffffffff828a4347>] tcp_clean_rtx_queue+0x2057/0x2ab0 n
[<ffffffff828ae6fd>] tcp_ack+0x151d/0x3930
[<ffffffff828baa09>] tcp_rcv_state_process+0x1c69/0x4fd0
[<ffffffff828efb7f>] tcp_v4_do_rcv+0x54f/0x7c0
[<ffffffff8258aacb>] sk_backlog_rcv
[<ffffffff8258aacb>] __release_sock+0x12b/0x3a0
[<ffffffff8258ad9e>] release_sock+0x5e/0x1c0
[<ffffffff8294a785>] inet_wait_for_connect
[<ffffffff8294a785>] __inet_stream_connect+0x545/0xc50
[<ffffffff82886f08>] tcp_sendmsg_fastopen
[<ffffffff82886f08>] tcp_sendmsg+0x2298/0x35a0
[<ffffffff82952515>] inet_sendmsg+0xe5/0x520
[<ffffffff8257152f>] sock_sendmsg_nosec
[<ffffffff8257152f>] sock_sendmsg+0xcf/0x110
Fixes: 8c72c65b426b ("tcp: update skb->skb_mstamp more carefully")
Fixes: 783237e8daf1 ("net-tcp: Fast Open client - sending SYN-data")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Dmitry Vyukov <dvyukov@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-09-20 01:05:57 +08:00
|
|
|
tp->packets_out -= tcp_skb_pcount(syn_data);
|
|
|
|
|
|
2012-07-19 14:43:07 +08:00
|
|
|
fallback:
|
|
|
|
|
/* Send a regular SYN with Fast Open cookie request option */
|
|
|
|
|
if (fo->cookie.len > 0)
|
|
|
|
|
fo->cookie.len = 0;
|
|
|
|
|
err = tcp_transmit_skb(sk, syn, 1, sk->sk_allocation);
|
|
|
|
|
if (err)
|
|
|
|
|
tp->syn_fastopen = 0;
|
|
|
|
|
done:
|
|
|
|
|
fo->cookie.len = -1; /* Exclude Fast Open option for SYN retries */
|
|
|
|
|
return err;
|
|
|
|
|
}
|
|
|
|
|
|
2009-07-22 07:00:40 +08:00
|
|
|
/* Build a SYN and send it off. */
|
2005-04-17 06:20:36 +08:00
|
|
|
int tcp_connect(struct sock *sk)
|
|
|
|
|
{
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
struct sk_buff *buff;
|
2010-11-16 19:52:49 +08:00
|
|
|
int err;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2018-01-26 08:14:09 +08:00
|
|
|
tcp_call_bpf(sk, BPF_SOCK_OPS_TCP_CONNECT_CB, 0, NULL);
|
tcp: fastopen: tcp_connect() must refresh the route
With new TCP_FASTOPEN_CONNECT socket option, there is a possibility
to call tcp_connect() while socket sk_dst_cache is either NULL
or invalid.
+0 socket(..., SOCK_STREAM, IPPROTO_TCP) = 4
+0 fcntl(4, F_SETFL, O_RDWR|O_NONBLOCK) = 0
+0 setsockopt(4, SOL_TCP, TCP_FASTOPEN_CONNECT, [1], 4) = 0
+0 connect(4, ..., ...) = 0
<< sk->sk_dst_cache becomes obsolete, or even set to NULL >>
+1 sendto(4, ..., 1000, MSG_FASTOPEN, ..., ...) = 1000
We need to refresh the route otherwise bad things can happen,
especially when syzkaller is running on the host :/
Fixes: 19f6d3f3c8422 ("net/tcp-fastopen: Add new API support")
Reported-by: Dmitry Vyukov <dvyukov@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Wei Wang <weiwan@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Wei Wang <weiwan@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-08-08 16:41:58 +08:00
|
|
|
|
|
|
|
|
if (inet_csk(sk)->icsk_af_ops->rebuild_header(sk))
|
|
|
|
|
return -EHOSTUNREACH; /* Routing failure or similar. */
|
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
tcp_connect_init(sk);
|
|
|
|
|
|
2012-11-22 09:13:58 +08:00
|
|
|
if (unlikely(tp->repair)) {
|
|
|
|
|
tcp_finish_connect(sk, NULL);
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
2015-05-20 04:26:55 +08:00
|
|
|
buff = sk_stream_alloc_skb(sk, 0, sk->sk_allocation, true);
|
2014-11-18 15:06:20 +08:00
|
|
|
if (unlikely(!buff))
|
2005-04-17 06:20:36 +08:00
|
|
|
return -ENOBUFS;
|
|
|
|
|
|
2010-06-12 22:01:43 +08:00
|
|
|
tcp_init_nondata_skb(buff, tp->write_seq++, TCPHDR_SYN);
|
2017-05-17 05:00:14 +08:00
|
|
|
tcp_mstamp_refresh(tp);
|
|
|
|
|
tp->retrans_stamp = tcp_time_stamp(tp);
|
2012-07-19 14:43:07 +08:00
|
|
|
tcp_connect_queue_skb(sk, buff);
|
2014-09-29 19:08:30 +08:00
|
|
|
tcp_ecn_send_syn(sk, buff);
|
2017-10-06 13:21:27 +08:00
|
|
|
tcp_rbtree_insert(&sk->tcp_rtx_queue, buff);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2012-07-19 14:43:07 +08:00
|
|
|
/* Send off SYN; include data in Fast Open. */
|
|
|
|
|
err = tp->fastopen_req ? tcp_send_syn_data(sk, buff) :
|
|
|
|
|
tcp_transmit_skb(sk, buff, 1, sk->sk_allocation);
|
2010-11-16 19:52:49 +08:00
|
|
|
if (err == -ECONNREFUSED)
|
|
|
|
|
return err;
|
2006-08-08 12:04:15 +08:00
|
|
|
|
|
|
|
|
/* We change tp->snd_nxt after the tcp_transmit_skb() call
|
|
|
|
|
* in order to make this packet get counted in tcpOutSegs.
|
|
|
|
|
*/
|
2019-10-11 11:17:42 +08:00
|
|
|
WRITE_ONCE(tp->snd_nxt, tp->write_seq);
|
2006-08-08 12:04:15 +08:00
|
|
|
tp->pushed_seq = tp->write_seq;
|
tcp: fastopen: fix on syn-data transmit failure
Our recent change exposed a bug in TCP Fastopen Client that syzkaller
found right away [1]
When we prepare skb with SYN+DATA, we attempt to transmit it,
and we update socket state as if the transmit was a success.
In socket RTX queue we have two skbs, one with the SYN alone,
and a second one containing the DATA.
When (malicious) ACK comes in, we now complain that second one had no
skb_mstamp.
The proper fix is to make sure that if the transmit failed, we do not
pretend we sent the DATA skb, and make it our send_head.
When 3WHS completes, we can now send the DATA right away, without having
to wait for a timeout.
[1]
WARNING: CPU: 0 PID: 100189 at net/ipv4/tcp_input.c:3117 tcp_clean_rtx_queue+0x2057/0x2ab0 net/ipv4/tcp_input.c:3117()
WARN_ON_ONCE(last_ackt == 0);
Modules linked in:
CPU: 0 PID: 100189 Comm: syz-executor1 Not tainted
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
0000000000000000 ffff8800b35cb1d8 ffffffff81cad00d 0000000000000000
ffffffff828a4347 ffff88009f86c080 ffffffff8316eb20 0000000000000d7f
ffff8800b35cb220 ffffffff812c33c2 ffff8800baad2440 00000009d46575c0
Call Trace:
[<ffffffff81cad00d>] __dump_stack
[<ffffffff81cad00d>] dump_stack+0xc1/0x124
[<ffffffff812c33c2>] warn_slowpath_common+0xe2/0x150
[<ffffffff812c361e>] warn_slowpath_null+0x2e/0x40
[<ffffffff828a4347>] tcp_clean_rtx_queue+0x2057/0x2ab0 n
[<ffffffff828ae6fd>] tcp_ack+0x151d/0x3930
[<ffffffff828baa09>] tcp_rcv_state_process+0x1c69/0x4fd0
[<ffffffff828efb7f>] tcp_v4_do_rcv+0x54f/0x7c0
[<ffffffff8258aacb>] sk_backlog_rcv
[<ffffffff8258aacb>] __release_sock+0x12b/0x3a0
[<ffffffff8258ad9e>] release_sock+0x5e/0x1c0
[<ffffffff8294a785>] inet_wait_for_connect
[<ffffffff8294a785>] __inet_stream_connect+0x545/0xc50
[<ffffffff82886f08>] tcp_sendmsg_fastopen
[<ffffffff82886f08>] tcp_sendmsg+0x2298/0x35a0
[<ffffffff82952515>] inet_sendmsg+0xe5/0x520
[<ffffffff8257152f>] sock_sendmsg_nosec
[<ffffffff8257152f>] sock_sendmsg+0xcf/0x110
Fixes: 8c72c65b426b ("tcp: update skb->skb_mstamp more carefully")
Fixes: 783237e8daf1 ("net-tcp: Fast Open client - sending SYN-data")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Dmitry Vyukov <dvyukov@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-09-20 01:05:57 +08:00
|
|
|
buff = tcp_send_head(sk);
|
|
|
|
|
if (unlikely(buff)) {
|
2019-10-11 11:17:42 +08:00
|
|
|
WRITE_ONCE(tp->snd_nxt, TCP_SKB_CB(buff)->seq);
|
tcp: fastopen: fix on syn-data transmit failure
Our recent change exposed a bug in TCP Fastopen Client that syzkaller
found right away [1]
When we prepare skb with SYN+DATA, we attempt to transmit it,
and we update socket state as if the transmit was a success.
In socket RTX queue we have two skbs, one with the SYN alone,
and a second one containing the DATA.
When (malicious) ACK comes in, we now complain that second one had no
skb_mstamp.
The proper fix is to make sure that if the transmit failed, we do not
pretend we sent the DATA skb, and make it our send_head.
When 3WHS completes, we can now send the DATA right away, without having
to wait for a timeout.
[1]
WARNING: CPU: 0 PID: 100189 at net/ipv4/tcp_input.c:3117 tcp_clean_rtx_queue+0x2057/0x2ab0 net/ipv4/tcp_input.c:3117()
WARN_ON_ONCE(last_ackt == 0);
Modules linked in:
CPU: 0 PID: 100189 Comm: syz-executor1 Not tainted
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
0000000000000000 ffff8800b35cb1d8 ffffffff81cad00d 0000000000000000
ffffffff828a4347 ffff88009f86c080 ffffffff8316eb20 0000000000000d7f
ffff8800b35cb220 ffffffff812c33c2 ffff8800baad2440 00000009d46575c0
Call Trace:
[<ffffffff81cad00d>] __dump_stack
[<ffffffff81cad00d>] dump_stack+0xc1/0x124
[<ffffffff812c33c2>] warn_slowpath_common+0xe2/0x150
[<ffffffff812c361e>] warn_slowpath_null+0x2e/0x40
[<ffffffff828a4347>] tcp_clean_rtx_queue+0x2057/0x2ab0 n
[<ffffffff828ae6fd>] tcp_ack+0x151d/0x3930
[<ffffffff828baa09>] tcp_rcv_state_process+0x1c69/0x4fd0
[<ffffffff828efb7f>] tcp_v4_do_rcv+0x54f/0x7c0
[<ffffffff8258aacb>] sk_backlog_rcv
[<ffffffff8258aacb>] __release_sock+0x12b/0x3a0
[<ffffffff8258ad9e>] release_sock+0x5e/0x1c0
[<ffffffff8294a785>] inet_wait_for_connect
[<ffffffff8294a785>] __inet_stream_connect+0x545/0xc50
[<ffffffff82886f08>] tcp_sendmsg_fastopen
[<ffffffff82886f08>] tcp_sendmsg+0x2298/0x35a0
[<ffffffff82952515>] inet_sendmsg+0xe5/0x520
[<ffffffff8257152f>] sock_sendmsg_nosec
[<ffffffff8257152f>] sock_sendmsg+0xcf/0x110
Fixes: 8c72c65b426b ("tcp: update skb->skb_mstamp more carefully")
Fixes: 783237e8daf1 ("net-tcp: Fast Open client - sending SYN-data")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Dmitry Vyukov <dvyukov@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-09-20 01:05:57 +08:00
|
|
|
tp->pushed_seq = TCP_SKB_CB(buff)->seq;
|
|
|
|
|
}
|
2008-07-17 11:22:04 +08:00
|
|
|
TCP_INC_STATS(sock_net(sk), TCP_MIB_ACTIVEOPENS);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
|
/* Timer for repeating the SYN until an answer. */
|
2005-08-10 11:11:08 +08:00
|
|
|
inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS,
|
|
|
|
|
inet_csk(sk)->icsk_rto, TCP_RTO_MAX);
|
2005-04-17 06:20:36 +08:00
|
|
|
return 0;
|
|
|
|
|
}
|
2010-07-10 05:22:10 +08:00
|
|
|
EXPORT_SYMBOL(tcp_connect);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
|
/* Send out a delayed ack, the caller does the policy checking
|
|
|
|
|
* to see if we should even be here. See tcp_input.c:tcp_ack_snd_check()
|
|
|
|
|
* for details.
|
|
|
|
|
*/
|
|
|
|
|
void tcp_send_delayed_ack(struct sock *sk)
|
|
|
|
|
{
|
2005-08-10 11:10:42 +08:00
|
|
|
struct inet_connection_sock *icsk = inet_csk(sk);
|
|
|
|
|
int ato = icsk->icsk_ack.ato;
|
2005-04-17 06:20:36 +08:00
|
|
|
unsigned long timeout;
|
|
|
|
|
|
|
|
|
|
if (ato > TCP_DELACK_MIN) {
|
2005-08-10 11:10:42 +08:00
|
|
|
const struct tcp_sock *tp = tcp_sk(sk);
|
2008-01-01 06:57:14 +08:00
|
|
|
int max_ato = HZ / 2;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2019-01-26 02:53:19 +08:00
|
|
|
if (inet_csk_in_pingpong_mode(sk) ||
|
2008-01-01 06:57:14 +08:00
|
|
|
(icsk->icsk_ack.pending & ICSK_ACK_PUSHED))
|
2005-04-17 06:20:36 +08:00
|
|
|
max_ato = TCP_DELACK_MAX;
|
|
|
|
|
|
|
|
|
|
/* Slow path, intersegment interval is "high". */
|
|
|
|
|
|
|
|
|
|
/* If some rtt estimate is known, use it to bound delayed ack.
|
2005-08-10 11:10:42 +08:00
|
|
|
* Do not use inet_csk(sk)->icsk_rto here, use results of rtt measurements
|
2005-04-17 06:20:36 +08:00
|
|
|
* directly.
|
|
|
|
|
*/
|
2014-02-27 06:02:48 +08:00
|
|
|
if (tp->srtt_us) {
|
|
|
|
|
int rtt = max_t(int, usecs_to_jiffies(tp->srtt_us >> 3),
|
|
|
|
|
TCP_DELACK_MIN);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
|
|
|
|
if (rtt < max_ato)
|
|
|
|
|
max_ato = rtt;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
ato = min(ato, max_ato);
|
|
|
|
|
}
|
|
|
|
|
|
2020-08-21 03:00:21 +08:00
|
|
|
ato = min_t(u32, ato, inet_csk(sk)->icsk_delack_max);
|
|
|
|
|
|
2005-04-17 06:20:36 +08:00
|
|
|
/* Stay within the limit we were given */
|
|
|
|
|
timeout = jiffies + ato;
|
|
|
|
|
|
|
|
|
|
/* Use new timeout only if there wasn't a older one earlier. */
|
2005-08-10 11:10:42 +08:00
|
|
|
if (icsk->icsk_ack.pending & ICSK_ACK_TIMER) {
|
2020-09-30 20:54:56 +08:00
|
|
|
/* If delack timer is about to expire, send ACK now. */
|
|
|
|
|
if (time_before_eq(icsk->icsk_ack.timeout, jiffies + (ato >> 2))) {
|
2005-04-17 06:20:36 +08:00
|
|
|
tcp_send_ack(sk);
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
2005-08-10 11:10:42 +08:00
|
|
|
if (!time_before(timeout, icsk->icsk_ack.timeout))
|
|
|
|
|
timeout = icsk->icsk_ack.timeout;
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
2005-08-10 11:10:42 +08:00
|
|
|
icsk->icsk_ack.pending |= ICSK_ACK_SCHED | ICSK_ACK_TIMER;
|
|
|
|
|
icsk->icsk_ack.timeout = timeout;
|
|
|
|
|
sk_reset_timer(sk, &icsk->icsk_delack_timer, timeout);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* This routine sends an ack and also updates the window. */
|
2018-07-19 04:56:34 +08:00
|
|
|
void __tcp_send_ack(struct sock *sk, u32 rcv_nxt)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
2007-12-31 20:51:11 +08:00
|
|
|
struct sk_buff *buff;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2007-12-31 20:51:11 +08:00
|
|
|
/* If we have been reset, we may not send again. */
|
|
|
|
|
if (sk->sk_state == TCP_CLOSE)
|
|
|
|
|
return;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2007-12-31 20:51:11 +08:00
|
|
|
/* We are not putting this on the write queue, so
|
|
|
|
|
* tcp_transmit_skb() will set the ownership to this
|
|
|
|
|
* sock.
|
|
|
|
|
*/
|
2015-12-01 00:57:28 +08:00
|
|
|
buff = alloc_skb(MAX_TCP_HEADER,
|
|
|
|
|
sk_gfp_mask(sk, GFP_ATOMIC | __GFP_NOWARN));
|
|
|
|
|
if (unlikely(!buff)) {
|
2020-09-30 20:54:57 +08:00
|
|
|
struct inet_connection_sock *icsk = inet_csk(sk);
|
|
|
|
|
unsigned long delay;
|
|
|
|
|
|
|
|
|
|
delay = TCP_DELACK_MAX << icsk->icsk_ack.retry;
|
|
|
|
|
if (delay < TCP_RTO_MAX)
|
|
|
|
|
icsk->icsk_ack.retry++;
|
2007-12-31 20:51:11 +08:00
|
|
|
inet_csk_schedule_ack(sk);
|
2020-09-30 20:54:57 +08:00
|
|
|
icsk->icsk_ack.ato = TCP_ATO_MIN;
|
|
|
|
|
inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK, delay, TCP_RTO_MAX);
|
2007-12-31 20:51:11 +08:00
|
|
|
return;
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
2007-12-31 20:51:11 +08:00
|
|
|
|
|
|
|
|
/* Reserve space for headers and prepare control bits. */
|
|
|
|
|
skb_reserve(buff, MAX_TCP_HEADER);
|
2010-06-12 22:01:43 +08:00
|
|
|
tcp_init_nondata_skb(buff, tcp_acceptable_seq(sk), TCPHDR_ACK);
|
2007-12-31 20:51:11 +08:00
|
|
|
|
tcp: do not pace pure ack packets
When we added pacing to TCP, we decided to let sch_fq take care
of actual pacing.
All TCP had to do was to compute sk->pacing_rate using simple formula:
sk->pacing_rate = 2 * cwnd * mss / rtt
It works well for senders (bulk flows), but not very well for receivers
or even RPC :
cwnd on the receiver can be less than 10, rtt can be around 100ms, so we
can end up pacing ACK packets, slowing down the sender.
Really, only the sender should pace, according to its own logic.
Instead of adding a new bit in skb, or call yet another flow
dissection, we tweak skb->truesize to a small value (2), and
we instruct sch_fq to use new helper and not pace pure ack.
Note this also helps TCP small queue, as ack packets present
in qdisc/NIC do not prevent sending a data packet (RPC workload)
This helps to reduce tx completion overhead, ack packets can use regular
sock_wfree() instead of tcp_wfree() which is a bit more expensive.
This has no impact in the case packets are sent to loopback interface,
as we do not coalesce ack packets (were we would detect skb->truesize
lie)
In case netem (with a delay) is used, skb_orphan_partial() also sets
skb->truesize to 1.
This patch is a combination of two patches we used for about one year at
Google.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2015-02-04 10:31:53 +08:00
|
|
|
/* We do not want pure acks influencing TCP Small Queues or fq/pacing
|
|
|
|
|
* too much.
|
|
|
|
|
* SKB_TRUESIZE(max(1 .. 66, MAX_TCP_HEADER)) is unfortunately ~784
|
|
|
|
|
*/
|
|
|
|
|
skb_set_tcp_pure_ack(buff);
|
|
|
|
|
|
2007-12-31 20:51:11 +08:00
|
|
|
/* Send it off, this clears delayed acks for us. */
|
2018-07-19 04:56:34 +08:00
|
|
|
__tcp_transmit_skb(sk, buff, 0, (__force gfp_t)0, rcv_nxt);
|
|
|
|
|
}
|
2018-07-19 04:56:35 +08:00
|
|
|
EXPORT_SYMBOL_GPL(__tcp_send_ack);
|
2018-07-19 04:56:34 +08:00
|
|
|
|
|
|
|
|
void tcp_send_ack(struct sock *sk)
|
|
|
|
|
{
|
|
|
|
|
__tcp_send_ack(sk, tcp_sk(sk)->rcv_nxt);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* This routine sends a packet with an out of date sequence
|
|
|
|
|
* number. It assumes the other end will try to ack it.
|
|
|
|
|
*
|
|
|
|
|
* Question: what should we make while urgent mode?
|
|
|
|
|
* 4.4BSD forces sending single byte of data. We cannot send
|
|
|
|
|
* out of window data, because we have SND.NXT==SND.MAX...
|
|
|
|
|
*
|
|
|
|
|
* Current solution: to send TWO zero-length segments in urgent mode:
|
|
|
|
|
* one is with SEG.SEQ=SND.UNA to deliver urgent pointer, another is
|
|
|
|
|
* out-of-date with SND.UNA-1 to probe window.
|
|
|
|
|
*/
|
2015-05-07 05:26:25 +08:00
|
|
|
static int tcp_xmit_probe_skb(struct sock *sk, int urgent, int mib)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
|
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
struct sk_buff *skb;
|
|
|
|
|
|
|
|
|
|
/* We don't queue it, tcp_transmit_skb() sets ownership. */
|
2015-12-01 00:57:28 +08:00
|
|
|
skb = alloc_skb(MAX_TCP_HEADER,
|
|
|
|
|
sk_gfp_mask(sk, GFP_ATOMIC | __GFP_NOWARN));
|
2015-04-03 16:17:26 +08:00
|
|
|
if (!skb)
|
2005-04-17 06:20:36 +08:00
|
|
|
return -1;
|
|
|
|
|
|
|
|
|
|
/* Reserve space for headers and set control bits. */
|
|
|
|
|
skb_reserve(skb, MAX_TCP_HEADER);
|
|
|
|
|
/* Use a previous sequence. This should cause the other
|
|
|
|
|
* end to send an ack. Don't queue or clone SKB, just
|
|
|
|
|
* send it.
|
|
|
|
|
*/
|
2010-06-12 22:01:43 +08:00
|
|
|
tcp_init_nondata_skb(skb, tp->snd_una - !urgent, TCPHDR_ACK);
|
2015-10-20 04:51:34 +08:00
|
|
|
NET_INC_STATS(sock_net(sk), mib);
|
2015-12-01 00:57:28 +08:00
|
|
|
return tcp_transmit_skb(sk, skb, 0, (__force gfp_t)0);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
|
2017-05-17 05:00:00 +08:00
|
|
|
/* Called from setsockopt( ... TCP_REPAIR ) */
|
2012-04-19 11:40:39 +08:00
|
|
|
void tcp_send_window_probe(struct sock *sk)
|
|
|
|
|
{
|
|
|
|
|
if (sk->sk_state == TCP_ESTABLISHED) {
|
|
|
|
|
tcp_sk(sk)->snd_wl1 = tcp_sk(sk)->rcv_nxt - 1;
|
2017-05-17 05:00:14 +08:00
|
|
|
tcp_mstamp_refresh(tcp_sk(sk));
|
2015-05-07 05:26:25 +08:00
|
|
|
tcp_xmit_probe_skb(sk, 0, LINUX_MIB_TCPWINPROBE);
|
2012-04-19 11:40:39 +08:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2009-07-22 07:00:40 +08:00
|
|
|
/* Initiate keepalive or window probe from timer. */
|
2015-05-07 05:26:25 +08:00
|
|
|
int tcp_write_wakeup(struct sock *sk, int mib)
|
2005-04-17 06:20:36 +08:00
|
|
|
{
|
2007-12-31 20:51:11 +08:00
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
|
|
|
|
struct sk_buff *skb;
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2007-12-31 20:51:11 +08:00
|
|
|
if (sk->sk_state == TCP_CLOSE)
|
|
|
|
|
return -1;
|
|
|
|
|
|
2015-04-03 16:17:27 +08:00
|
|
|
skb = tcp_send_head(sk);
|
|
|
|
|
if (skb && before(TCP_SKB_CB(skb)->seq, tcp_wnd_end(tp))) {
|
2007-12-31 20:51:11 +08:00
|
|
|
int err;
|
2009-03-14 22:23:05 +08:00
|
|
|
unsigned int mss = tcp_current_mss(sk);
|
2007-12-31 20:51:11 +08:00
|
|
|
unsigned int seg_size = tcp_wnd_end(tp) - TCP_SKB_CB(skb)->seq;
|
|
|
|
|
|
|
|
|
|
if (before(tp->pushed_seq, TCP_SKB_CB(skb)->end_seq))
|
|
|
|
|
tp->pushed_seq = TCP_SKB_CB(skb)->end_seq;
|
|
|
|
|
|
|
|
|
|
/* We are probing the opening of a window
|
|
|
|
|
* but the window size is != 0
|
|
|
|
|
* must have been a result SWS avoidance ( sender )
|
|
|
|
|
*/
|
|
|
|
|
if (seg_size < TCP_SKB_CB(skb)->end_seq - TCP_SKB_CB(skb)->seq ||
|
|
|
|
|
skb->len > mss) {
|
|
|
|
|
seg_size = min(seg_size, mss);
|
2011-09-28 01:25:05 +08:00
|
|
|
TCP_SKB_CB(skb)->tcp_flags |= TCPHDR_PSH;
|
2017-10-06 13:21:27 +08:00
|
|
|
if (tcp_fragment(sk, TCP_FRAG_IN_WRITE_QUEUE,
|
|
|
|
|
skb, seg_size, mss, GFP_ATOMIC))
|
2007-12-31 20:51:11 +08:00
|
|
|
return -1;
|
|
|
|
|
} else if (!tcp_skb_pcount(skb))
|
2015-06-12 00:15:17 +08:00
|
|
|
tcp_set_skb_tso_segs(skb, mss);
|
2007-12-31 20:51:11 +08:00
|
|
|
|
2011-09-28 01:25:05 +08:00
|
|
|
TCP_SKB_CB(skb)->tcp_flags |= TCPHDR_PSH;
|
2007-12-31 20:51:11 +08:00
|
|
|
err = tcp_transmit_skb(sk, skb, 1, GFP_ATOMIC);
|
|
|
|
|
if (!err)
|
|
|
|
|
tcp_event_new_data_sent(sk, skb);
|
|
|
|
|
return err;
|
|
|
|
|
} else {
|
2008-10-08 05:43:06 +08:00
|
|
|
if (between(tp->snd_up, tp->snd_una + 1, tp->snd_una + 0xFFFF))
|
2015-05-07 05:26:25 +08:00
|
|
|
tcp_xmit_probe_skb(sk, 1, mib);
|
|
|
|
|
return tcp_xmit_probe_skb(sk, 0, mib);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* A window probe timeout has occurred. If window is not closed send
|
|
|
|
|
* a partial packet else a zero probe.
|
|
|
|
|
*/
|
|
|
|
|
void tcp_send_probe0(struct sock *sk)
|
|
|
|
|
{
|
2005-08-10 11:10:42 +08:00
|
|
|
struct inet_connection_sock *icsk = inet_csk(sk);
|
2005-04-17 06:20:36 +08:00
|
|
|
struct tcp_sock *tp = tcp_sk(sk);
|
2016-02-03 15:46:54 +08:00
|
|
|
struct net *net = sock_net(sk);
|
2019-01-17 07:05:35 +08:00
|
|
|
unsigned long timeout;
|
2005-04-17 06:20:36 +08:00
|
|
|
int err;
|
|
|
|
|
|
2015-05-07 05:26:25 +08:00
|
|
|
err = tcp_write_wakeup(sk, LINUX_MIB_TCPWINPROBE);
|
2005-04-17 06:20:36 +08:00
|
|
|
|
2017-10-06 13:21:27 +08:00
|
|
|
if (tp->packets_out || tcp_write_queue_empty(sk)) {
|
2005-04-17 06:20:36 +08:00
|
|
|
/* Cancel probe timer, if it is not required. */
|
2005-08-10 15:03:31 +08:00
|
|
|
icsk->icsk_probes_out = 0;
|
2005-08-10 11:10:42 +08:00
|
|
|
icsk->icsk_backoff = 0;
|
2005-04-17 06:20:36 +08:00
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
2019-01-17 07:05:35 +08:00
|
|
|
icsk->icsk_probes_out++;
|
2005-04-17 06:20:36 +08:00
|
|
|
if (err <= 0) {
|
2016-02-03 15:46:54 +08:00
|
|
|
if (icsk->icsk_backoff < net->ipv4.sysctl_tcp_retries2)
|
2005-08-10 11:10:42 +08:00
|
|
|
icsk->icsk_backoff++;
|
2019-01-17 07:05:35 +08:00
|
|
|
timeout = tcp_probe0_when(sk, TCP_RTO_MAX);
|
2005-04-17 06:20:36 +08:00
|
|
|
} else {
|
|
|
|
|
/* If packet was not sent due to local congestion,
|
2019-01-17 07:05:35 +08:00
|
|
|
* Let senders fight for local resources conservatively.
|
2005-04-17 06:20:36 +08:00
|
|
|
*/
|
2019-01-17 07:05:35 +08:00
|
|
|
timeout = TCP_RESOURCE_PROBE_INTERVAL;
|
|
|
|
|
}
|
2020-05-05 02:27:49 +08:00
|
|
|
tcp_reset_xmit_timer(sk, ICSK_TIME_PROBE0, timeout, TCP_RTO_MAX);
|
2005-04-17 06:20:36 +08:00
|
|
|
}
|
2014-06-25 22:09:59 +08:00
|
|
|
|
2015-09-25 22:39:23 +08:00
|
|
|
int tcp_rtx_synack(const struct sock *sk, struct request_sock *req)
|
2014-06-25 22:09:59 +08:00
|
|
|
{
|
|
|
|
|
const struct tcp_request_sock_ops *af_ops = tcp_rsk(req)->af_specific;
|
|
|
|
|
struct flowi fl;
|
|
|
|
|
int res;
|
|
|
|
|
|
2015-09-16 06:24:20 +08:00
|
|
|
tcp_rsk(req)->txhash = net_tx_rndhash();
|
2020-08-21 03:00:52 +08:00
|
|
|
res = af_ops->send_synack(sk, NULL, &fl, req, NULL, TCP_SYNACK_NORMAL,
|
|
|
|
|
NULL);
|
2014-06-25 22:09:59 +08:00
|
|
|
if (!res) {
|
2016-04-28 07:44:32 +08:00
|
|
|
__TCP_INC_STATS(sock_net(sk), TCP_MIB_RETRANSSEGS);
|
2016-04-28 07:44:39 +08:00
|
|
|
__NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPSYNRETRANS);
|
2016-09-22 07:16:15 +08:00
|
|
|
if (unlikely(tcp_passive_fastopen(sk)))
|
|
|
|
|
tcp_sk(sk)->total_retrans++;
|
2017-10-31 05:41:35 +08:00
|
|
|
trace_tcp_retransmit_synack(sk, req);
|
2014-06-25 22:09:59 +08:00
|
|
|
}
|
|
|
|
|
return res;
|
|
|
|
|
}
|
|
|
|
|
EXPORT_SYMBOL(tcp_rtx_synack);
|