Both TLP and Fast Open call __tcp_retransmit_skb() instead of
tcp_retransmit_skb() to avoid changing tp->retrans_out.

This has the side effect of missing SNMP counters increments as well
as tcp_info tcpi_total_retrans updates.

Fix this by moving the stats increments of into __tcp_retransmit_skb()
Signed-off-by: NEric Dumazet <edumazet@google.com>
Acked-by: NNandita Dukkipati <nanditad@google.com>
Acked-by: NNeal Cardwell <ncardwell@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

ip_queue_xmit() assumes the skb it has to transmit is attached to an
inet socket. Commit 31c70d59 ("l2tp: keep original skb ownership")
changed l2tp to not change skb ownership and thus broke this assumption.

One fix is to add a new 'struct sock *sk' parameter to ip_queue_xmit(),
so that we do not assume skb->sk points to the socket used by l2tp
tunnel.

Fixes: 31c70d59 ("l2tp: keep original skb ownership")
Reported-by: NZhan Jianyu <nasa4836@gmail.com>
Tested-by: NZhan Jianyu <nasa4836@gmail.com>
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

There is no need to allocate 15 bytes in excess for a SYNACK packet,
as it contains no data, only headers.

SYNACK are always generated in softirq context, and contain a single
segment, we can use TCP_INC_STATS_BH()
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

After commit d4589926 (tcp: refine TSO splits), tcp_nagle_check() does
not use parameter mss_now anymore.
Signed-off-by: NWeiping Pan <panweiping3@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

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: 6f458dfb ("tcp: improve latencies of timer triggered events")
Reported-by: NLars Persson <lars.persson@axis.com>
Tested-by: NLars Persson <lars.persson@axis.com>
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

All skb in socket write queue should be properly timestamped.

In case of FastOpen, we special case the SYN+DATA 'message' as we
queue in socket wrote queue the two fallback skbs:

1) SYN message by itself.
2) DATA segment by itself.

We should make sure these skbs have proper timestamps.

Add a WARN_ON_ONCE() to eventually catch future violations.

Fixes: 740b0f18 ("tcp: switch rtt estimations to usec resolution")
Signed-off-by: NEric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: NNeal Cardwell <ncardwell@google.com>
Acked-by: NYuchung Cheng <ycheng@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Usage of skb->tstamp should remain private to TCP stack
(only set on packets on write queue, not on cloned ones)

Otherwise, packets given to loopback interface with a non null tstamp
can confuse netif_rx() / net_timestamp_check()

Other possibility would be to clear tstamp in loopback_xmit(),
as done in skb_scrub_packet()

Fixes: 740b0f18 ("tcp: switch rtt estimations to usec resolution")
Signed-off-by: NEric Dumazet <edumazet@google.com>
Reported-by: NWillem de Bruijn <willemb@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Can be invoked from non-BH context.

Based upon a patch by Eric Dumazet.

Fixes: f19c29e3 ("tcp: snmp stats for Fast Open, SYN rtx, and data pkts")
Reported-by: NSergey Senozhatsky <sergey.senozhatsky@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add the following snmp stats:

TCPFastOpenActiveFail: Fast Open attempts (SYN/data) failed beacuse
the remote does not accept it or the attempts timed out.

TCPSynRetrans: number of SYN and SYN/ACK retransmits to break down
retransmissions into SYN, fast-retransmits, timeout retransmits, etc.

TCPOrigDataSent: number of outgoing packets with original data (excluding
retransmission but including data-in-SYN). This counter is different from
TcpOutSegs because TcpOutSegs also tracks pure ACKs. TCPOrigDataSent is
more useful to track the TCP retransmission rate.

Change TCPFastOpenActive to track only successful Fast Opens to be symmetric to
TCPFastOpenPassive.
Signed-off-by: NYuchung Cheng <ycheng@google.com>
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NNandita Dukkipati <nanditad@google.com>
Signed-off-by: NLawrence Brakmo <brakmo@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

RTT may be bogus with tall loss probe (TLP) when a packet
is retransmitted and latter (s)acked without TCPCB_SACKED_RETRANS flag.

For example, TLP calls __tcp_retransmit_skb() instead of
tcp_retransmit_skb(). The skb timestamps are updated but the sacked
flag is not marked with TCPCB_SACKED_RETRANS. As a result we'll
get bogus RTT in tcp_clean_rtx_queue() or in tcp_sacktag_one() on
spurious retransmission.

The fix is to apply the sticky flag TCP_EVER_RETRANS to enforce Karn's
check on RTT sampling. However this will disable F-RTO if timeout occurs
after TLP, by resetting undo_marker in tcp_enter_loss(). We relax this
check to only if any pending retransmists are still in-flight.
Signed-off-by: NYuchung Cheng <ycheng@google.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Acked-by: NNeal Cardwell <ncardwell@google.com>
Acked-by: NNandita Dukkipati <nanditad@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Upcoming congestion controls for TCP require usec resolution for RTT
estimations. Millisecond resolution is simply not enough these days.

FQ/pacing in DC environments also require this change for finer control
and removal of bimodal behavior due to the current hack in
tcp_update_pacing_rate() for 'small rtt'

TCP_CONG_RTT_STAMP is no longer needed.

As Julian Anastasov pointed out, we need to keep user compatibility :
tcp_metrics used to export RTT and RTTVAR in msec resolution,
so we added RTT_US and RTTVAR_US. An iproute2 patch is needed
to use the new attributes if provided by the kernel.

In this example ss command displays a srtt of 32 usecs (10Gbit link)

lpk51:~# ./ss -i dst lpk52
Netid  State      Recv-Q Send-Q   Local Address:Port       Peer
Address:Port
tcp    ESTAB      0      1         10.246.11.51:42959
10.246.11.52:64614
         cubic wscale:6,6 rto:201 rtt:0.032/0.001 ato:40 mss:1448
cwnd:10 send
3620.0Mbps pacing_rate 7240.0Mbps unacked:1 rcv_rtt:993 rcv_space:29559

Updated iproute2 ip command displays :

lpk51:~# ./ip tcp_metrics | grep 10.246.11.52
10.246.11.52 age 561.914sec cwnd 10 rtt 274us rttvar 213us source
10.246.11.51

Old binary displays :

lpk51:~# ip tcp_metrics | grep 10.246.11.52
10.246.11.52 age 561.914sec cwnd 10 rtt 250us rttvar 125us source
10.246.11.51

With help from Julian Anastasov, Stephen Hemminger and Yuchung Cheng
Signed-off-by: NEric Dumazet <edumazet@google.com>
Acked-by: NNeal Cardwell <ncardwell@google.com>
Cc: Stephen Hemminger <stephen@networkplumber.org>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Larry Brakmo <brakmo@google.com>
Cc: Julian Anastasov <ja@ssi.bg>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Three counters are added:
- one to track when we went from non-zero to zero window
- one to track the reverse
- one counter incremented when we want to announce zero window,
  but can't because we would shrink current window.
Suggested-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NFlorian Westphal <fw@strlen.de>
Acked-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

While LINUX_MIB_TCPSPURIOUS_RTX_HOSTQUEUES can only be incremented
in tcp_transmit_skb() from softirq (incoming message or timer
activation), it is better to use NET_INC_STATS() instead of
NET_INC_STATS_BH() as tcp_transmit_skb() can be called from process
context.

This will avoid copy/paste confusion when/if we want to add
other SNMP counters in tcp_transmit_skb()
Signed-off-by: NEric Dumazet <edumazet@google.com>
Cc: Hannes Frederic Sowa <hannes@stressinduktion.org>
Cc: Florian Westphal <fw@strlen.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch fixes two bugs in fastopen :

1) The tcp_sendmsg(...,  @size) argument was ignored.

   Code was relying on user not fooling the kernel with iovec mismatches

2) When MTU is about 64KB, tcp_send_syn_data() attempts order-5
allocations, which are likely to fail when memory gets fragmented.

Fixes: 783237e8 ("net-tcp: Fast Open client - sending SYN-data")
Signed-off-by: NEric Dumazet <edumazet@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: NYuchung Cheng <ycheng@google.com>
Tested-by: NYuchung Cheng <ycheng@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

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: NEric Dumazet <edumazet@google.com>
Tested-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NFlorian Westphal <fw@strlen.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

One of my pet coding style peeves is the practice of
adding extra return; at the end of function.
Kill several instances of this in network code.

I suppose some coccinelle wizardy could do this automatically.
Signed-off-by: NStephen Hemminger <stephen@networkplumber.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit 46d3ceab ("tcp: TCP Small Queues") introduced a possible
regression for applications using TCP_NODELAY.

If TCP session is throttled because of tsq, we should consult
tp->nonagle when TX completion is done and allow us to send additional
segment, especially if this segment is not a full MSS.
Otherwise this segment is sent after an RTO.

[edumazet] : Cooked the changelog, added another fix about testing
sk_wmem_alloc twice because TX completion can happen right before
setting TSQ_THROTTLED bit.

This problem is particularly visible with recent auto corking,
but might also be triggered with low tcp_limit_output_bytes
values or NIC drivers delaying TX completion by hundred of usec,
and very low rtt.

Thomas Glanzmann for example reported an iscsi regression, caused
by tcp auto corking making this bug quite visible.

Fixes: 46d3ceab ("tcp: TCP Small Queues")
Signed-off-by: NJohn Ogness <john.ogness@linutronix.de>
Signed-off-by: NEric Dumazet <edumazet@google.com>
Reported-by: NThomas Glanzmann <thomas@glanzmann.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

TCP pacing depends on an accurate srtt estimation.

Current srtt estimation is using jiffie resolution,
and has an artificial offset of at least 1 ms, which can produce
slowdowns when FQ/pacing is used, especially in DC world,
where typical rtt is below 1 ms.

We are planning a switch to usec resolution for linux-3.15,
but in the meantime, this patch removes the 1 ms offset.

All we need is to have tp->srtt minimal value of 1 to differentiate
the case of srtt being initialized or not, not 8.

The problematic behavior was observed on a 40Gbit testbed,
where 32 concurrent netperf were reaching 12Gbps of aggregate
speed, instead of line speed.

This patch also has the effect of reporting more accurate srtt and send
rates to iproute2 ss command as in :

$ ss -i dst cca2
Netid  State      Recv-Q Send-Q          Local Address:Port
Peer Address:Port
tcp    ESTAB      0      0                10.244.129.1:56984
10.244.129.2:12865
	 cubic wscale:6,6 rto:200 rtt:0.25/0.25 ato:40 mss:1448 cwnd:10 send
463.4Mbps rcv_rtt:1 rcv_space:29200
tcp    ESTAB      0      390960           10.244.129.1:60247
10.244.129.2:50204
	 cubic wscale:6,6 rto:200 rtt:0.875/0.75 mss:1448 cwnd:73 ssthresh:51
send 966.4Mbps unacked:73 retrans:0/121 rcv_space:29200
Reported-by: NVytautas Valancius <valas@google.com>
Signed-off-by: NEric Dumazet <edumazet@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Acked-by: NNeal Cardwell <ncardwell@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The following are only used in one file:
  tcp_connect_init
  tcp_set_rto
Signed-off-by: NStephen Hemminger <stephen@networkplumber.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

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: NEric 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: NNeal Cardwell <ncardwell@google.com>
Tested-by: NNeal Cardwell <ncardwell@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Various spelling fixes in networking stack
Signed-off-by: NStephen Hemminger <stephen@networkplumber.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Compiler doesn't know skb_shinfo(skb) pointer is usually constant.

By using a temporary variable, we help generating smaller code.

For example, tcp_init_nondata_skb() is inlined after this patch.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

snd_nxt must be updated synchronously with sk_send_head.  Otherwise
tp->packets_out may be updated incorrectly, what may bring a kernel panic.

Here is a kernel panic from my host.
[  103.043194] BUG: unable to handle kernel NULL pointer dereference at 0000000000000048
[  103.044025] IP: [<ffffffff815aaaaf>] tcp_rearm_rto+0xcf/0x150
...
[  146.301158] Call Trace:
[  146.301158]  [<ffffffff815ab7f0>] tcp_ack+0xcc0/0x12c0

Before this panic a tcp socket was restored. This socket had sent and
unsent data in the write queue. Sent data was restored in repair mode,
then the socket was switched from reapair mode and unsent data was
restored. After that the socket was switched back into repair mode.

In that moment we had a socket where write queue looks like this:
snd_una    snd_nxt   write_seq
   |_________|________|
             |
	  sk_send_head

After a second switching from repair mode the state of socket was
changed:

snd_una          snd_nxt, write_seq
   |_________ ________|
             |
	  sk_send_head

This state is inconsistent, because snd_nxt and sk_send_head are not
synchronized.

Bellow you can find a call trace, how packets_out can be incremented
twice for one skb, if snd_nxt and sk_send_head are not synchronized.
In this case packets_out will be always positive, even when
sk_write_queue is empty.

tcp_write_wakeup
	skb = tcp_send_head(sk);
	tcp_fragment
		if (!before(tp->snd_nxt, TCP_SKB_CB(buff)->end_seq))
			tcp_adjust_pcount(sk, skb, diff);
	tcp_event_new_data_sent
		tp->packets_out += tcp_skb_pcount(skb);

I think update of snd_nxt isn't required, when a socket is switched from
repair mode.  Because it's initialized in tcp_connect_init. Then when a
write queue is restored, snd_nxt is incremented in tcp_event_new_data_sent,
so it's always is in consistent state.

I have checked, that the bug is not reproduced with this patch and
all tests about restoring tcp connections work fine.

Cc: Pavel Emelyanov <xemul@parallels.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Alexey Kuznetsov <kuznet@ms2.inr.ac.ru>
Cc: James Morris <jmorris@namei.org>
Cc: Hideaki YOSHIFUJI <yoshfuji@linux-ipv6.org>
Cc: Patrick McHardy <kaber@trash.net>
Signed-off-by: NAndrey Vagin <avagin@openvz.org>
Acked-by: NPavel Emelyanov <xemul@parallels.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

After commit c9eeec26 ("tcp: TSQ can use a dynamic limit"), several
users reported throughput regressions, notably on mvneta and wifi
adapters.

802.11 AMPDU requires a fair amount of queueing to be effective.

This patch partially reverts the change done in tcp_write_xmit()
so that the minimal amount is sysctl_tcp_limit_output_bytes.

It also remove the use of this sysctl while building skb stored
in write queue, as TSO autosizing does the right thing anyway.

Users with well behaving NICS and correct qdisc (like sch_fq),
can then lower the default sysctl_tcp_limit_output_bytes value from
128KB to 8KB.

This new usage of sysctl_tcp_limit_output_bytes permits each driver
authors to check how their driver performs when/if the value is set
to a minimum of 4KB.

Normally, line rate for a single TCP flow should be possible,
but some drivers rely on timers to perform TX completion and
too long TX completion delays prevent reaching full throughput.

Fixes: c9eeec26 ("tcp: TSQ can use a dynamic limit")
Signed-off-by: NEric Dumazet <edumazet@google.com>
Reported-by: NSujith Manoharan <sujith@msujith.org>
Reported-by: NArnaud Ebalard <arno@natisbad.org>
Tested-by: NSujith Manoharan <sujith@msujith.org>
Cc: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Remove the specialized code in __tcp_retransmit_skb() that tries to trim
any ACKed payload preceding a FIN before we retransmit (this was added
in 1999 in v2.2.3pre3). This trimming code was made unreachable by the
more general code added above it that uses tcp_trim_head() to trim any
ACKed payload, with or without a FIN (this was added in "[NET]: Add
segmentation offload support to TCP." in 2002 circa v2.5.33).
Signed-off-by: NNeal Cardwell <ncardwell@google.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Acked-by: NYuchung Cheng <ycheng@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

sk_can_gso() should only be used as a hint in tcp_sendmsg() to build GSO
packets in the first place. (As a performance hint)

Once we have GSO packets in write queue, we can not decide they are no
longer GSO only because flow now uses a route which doesn't handle
TSO/GSO.

Core networking stack handles the case very well for us, all we need
is keeping track of packet counts in MSS terms, regardless of
segmentation done later (in GSO or hardware)

Right now, if  tcp_fragment() splits a GSO packet in two parts,
@left and @right, and route changed through a non GSO device,
both @left and @right have pcount set to 1, which is wrong,
and leads to incorrect packet_count tracking.

This problem was added in commit d5ac99a6 ("[TCP]: skb pcount with MTU
discovery")
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NNeal Cardwell <ncardwell@google.com>
Signed-off-by: NYuchung Cheng <ycheng@google.com>
Reported-by: NMaciej Żenczykowski <maze@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

TCP stack should make sure it owns skbs before mangling them.

We had various crashes using bnx2x, and it turned out gso_size
was cleared right before bnx2x driver was populating TC descriptor
of the _previous_ packet send. TCP stack can sometime retransmit
packets that are still in Qdisc.

Of course we could make bnx2x driver more robust (using
ACCESS_ONCE(shinfo->gso_size) for example), but the bug is TCP stack.

We have identified two points where skb_unclone() was needed.

This patch adds a WARN_ON_ONCE() to warn us if we missed another
fix of this kind.

Kudos to Neal for finding the root cause of this bug. Its visible
using small MSS.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NNeal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

1) We need to take a timestamp only for skb that should be cloned.

Other skbs are not in write queue and no rtt estimation is done on them.

2) the unlikely() hint is wrong for receivers (they send pure ACK)
Signed-off-by: NEric Dumazet <edumazet@google.com>
Cc: MF Nowlan <fitz@cs.yale.edu>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Acked-By: NYuchung Cheng <ycheng@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In commit 634fb979 ("inet: includes a sock_common in request_sock")
I forgot that the two ports in sock_common do not have same byte order :

skc_dport is __be16 (network order), but skc_num is __u16 (host order)

So sparse complains because ir_loc_port (mapped into skc_num) is
considered as __u16 while it should be __be16

Let rename ir_loc_port to ireq->ir_num (analogy with inet->inet_num),
and perform appropriate htons/ntohs conversions.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Reported-by: NWu Fengguang <fengguang.wu@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

TCP listener refactoring, part 5 :

We want to be able to insert request sockets (SYN_RECV) into main
ehash table instead of the per listener hash table to allow RCU
lookups and remove listener lock contention.

This patch includes the needed struct sock_common in front
of struct request_sock

This means there is no more inet6_request_sock IPv6 specific
structure.

Following inet_request_sock fields were renamed as they became
macros to reference fields from struct sock_common.
Prefix ir_ was chosen to avoid name collisions.

loc_port   -> ir_loc_port
loc_addr   -> ir_loc_addr
rmt_addr   -> ir_rmt_addr
rmt_port   -> ir_rmt_port
iif        -> ir_iif
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

tcp_established_options assumes opts->options is 0 before calling,
as it read modify writes it.

For the tcp_current_mss() case the opts structure is not zeroed,
so this can be done with uninitialized values.

This is ok, because ->options is not read in this path.
But it's still better to avoid the operation on the uninitialized
field. This shuts up a static code analyzer, and presumably
may help the optimizer.

Cc: netdev@vger.kernel.org
Signed-off-by: NAndi Kleen <ak@linux.intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

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: NEric 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: NNeal Cardwell <ncardwell@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In commit 90ba9b19 (tcp: tcp_make_synack() can use alloc_skb()), Eric changed
the call to sock_wmalloc in tcp_make_synack to alloc_skb.  In doing so,
the netfilter owner match lost its ability to block the SYNACK packet on
outbound listening sockets.  Revert the change, restoring the owner match
functionality.

This closes netfilter bugzilla #847.
Signed-off-by: NPhil Oester <kernel@linuxace.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

After hearing many people over past years complaining against TSO being
bursty or even buggy, we are proud to present automatic sizing of TSO
packets.

One part of the problem is that tcp_tso_should_defer() uses an heuristic
relying on upcoming ACKS instead of a timer, but more generally, having
big TSO packets makes little sense for low rates, as it tends to create
micro bursts on the network, and general consensus is to reduce the
buffering amount.

This patch introduces a per socket sk_pacing_rate, that approximates
the current sending rate, and allows us to size the TSO packets so
that we try to send one packet every ms.

This field could be set by other transports.

Patch has no impact for high speed flows, where having large TSO packets
makes sense to reach line rate.

For other flows, this helps better packet scheduling and ACK clocking.

This patch increases performance of TCP flows in lossy environments.

A new sysctl (tcp_min_tso_segs) is added, to specify the
minimal size of a TSO packet (default being 2).

A follow-up patch will provide a new packet scheduler (FQ), using
sk_pacing_rate as an input to perform optional per flow pacing.

This explains why we chose to set sk_pacing_rate to twice the current
rate, allowing 'slow start' ramp up.

sk_pacing_rate = 2 * cwnd * mss / srtt

v2: Neal Cardwell reported a suspect deferring of last two segments on
initial write of 10 MSS, I had to change tcp_tso_should_defer() to take
into account tp->xmit_size_goal_segs
Signed-off-by: NEric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Van Jacobson <vanj@google.com>
Cc: Tom Herbert <therbert@google.com>
Acked-by: NYuchung Cheng <ycheng@google.com>
Acked-by: NNeal Cardwell <ncardwell@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

u32 rcv_tstamp;     /* timestamp of last received ACK */

Its value used in tcp_retransmit_timer, which closes socket
if the last ack was received more then TCP_RTO_MAX ago.

Currently rcv_tstamp is initialized to zero and if tcp_retransmit_timer
is called before receiving a first ack, the connection is closed.

This patch initializes rcv_tstamp to a timestamp, when a socket was
restored.

Cc: Pavel Emelyanov <xemul@parallels.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Alexey Kuznetsov <kuznet@ms2.inr.ac.ru>
Cc: James Morris <jmorris@namei.org>
Cc: Hideaki YOSHIFUJI <yoshfuji@linux-ipv6.org>
Cc: Patrick McHardy <kaber@trash.net>
Reported-by: NCyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: NAndrey Vagin <avagin@openvz.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Idea of this patch is to add optional limitation of number of
unsent bytes in TCP sockets, to reduce usage of kernel memory.

TCP receiver might announce a big window, and TCP sender autotuning
might allow a large amount of bytes in write queue, but this has little
performance impact if a large part of this buffering is wasted :

Write queue needs to be large only to deal with large BDP, not
necessarily to cope with scheduling delays (incoming ACKS make room
for the application to queue more bytes)

For most workloads, using a value of 128 KB or less is OK to give
applications enough time to react to POLLOUT events in time
(or being awaken in a blocking sendmsg())

This patch adds two ways to set the limit :

1) Per socket option TCP_NOTSENT_LOWAT

2) A sysctl (/proc/sys/net/ipv4/tcp_notsent_lowat) for sockets
not using TCP_NOTSENT_LOWAT socket option (or setting a zero value)
Default value being UINT_MAX (0xFFFFFFFF), meaning this has no effect.

This changes poll()/select()/epoll() to report POLLOUT
only if number of unsent bytes is below tp->nosent_lowat

Note this might increase number of sendmsg()/sendfile() calls
when using non blocking sockets,
and increase number of context switches for blocking sockets.

Note this is not related to SO_SNDLOWAT (as SO_SNDLOWAT is
defined as :
 Specify the minimum number of bytes in the buffer until
 the socket layer will pass the data to the protocol)

Tested:

netperf sessions, and watching /proc/net/protocols "memory" column for TCP

With 200 concurrent netperf -t TCP_STREAM sessions, amount of kernel memory
used by TCP buffers shrinks by ~55 % (20567 pages instead of 45458)

lpq83:~# echo -1 >/proc/sys/net/ipv4/tcp_notsent_lowat
lpq83:~# (super_netperf 200 -t TCP_STREAM -H remote -l 90 &); sleep 60 ; grep TCP /proc/net/protocols
TCPv6     1880      2   45458   no     208   yes  ipv6        y  y  y  y  y  y  y  y  y  y  y  y  y  n  y  y  y  y  y
TCP       1696    508   45458   no     208   yes  kernel      y  y  y  y  y  y  y  y  y  y  y  y  y  n  y  y  y  y  y

lpq83:~# echo 131072 >/proc/sys/net/ipv4/tcp_notsent_lowat
lpq83:~# (super_netperf 200 -t TCP_STREAM -H remote -l 90 &); sleep 60 ; grep TCP /proc/net/protocols
TCPv6     1880      2   20567   no     208   yes  ipv6        y  y  y  y  y  y  y  y  y  y  y  y  y  n  y  y  y  y  y
TCP       1696    508   20567   no     208   yes  kernel      y  y  y  y  y  y  y  y  y  y  y  y  y  n  y  y  y  y  y

Using 128KB has no bad effect on the throughput or cpu usage
of a single flow, although there is an increase of context switches.

A bonus is that we hold socket lock for a shorter amount
of time and should improve latencies of ACK processing.

lpq83:~# echo -1 >/proc/sys/net/ipv4/tcp_notsent_lowat
lpq83:~# perf stat -e context-switches ./netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3
OMNI Send TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 7.7.7.84 () port 0 AF_INET : +/-2.500% @ 99% conf.
Local       Remote      Local  Elapsed Throughput Throughput  Local Local  Remote Remote Local   Remote  Service
Send Socket Recv Socket Send   Time               Units       CPU   CPU    CPU    CPU    Service Service Demand
Size        Size        Size   (sec)                          Util  Util   Util   Util   Demand  Demand  Units
Final       Final                                             %     Method %      Method
1651584     6291456     16384  20.00   17447.90   10^6bits/s  3.13  S      -1.00  U      0.353   -1.000  usec/KB

 Performance counter stats for './netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3':

           412,514 context-switches

     200.034645535 seconds time elapsed

lpq83:~# echo 131072 >/proc/sys/net/ipv4/tcp_notsent_lowat
lpq83:~# perf stat -e context-switches ./netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3
OMNI Send TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 7.7.7.84 () port 0 AF_INET : +/-2.500% @ 99% conf.
Local       Remote      Local  Elapsed Throughput Throughput  Local Local  Remote Remote Local   Remote  Service
Send Socket Recv Socket Send   Time               Units       CPU   CPU    CPU    CPU    Service Service Demand
Size        Size        Size   (sec)                          Util  Util   Util   Util   Demand  Demand  Units
Final       Final                                             %     Method %      Method
1593240     6291456     16384  20.00   17321.16   10^6bits/s  3.35  S      -1.00  U      0.381   -1.000  usec/KB

 Performance counter stats for './netperf -H 7.7.7.84 -t omni -l 20 -c -i10,3':

         2,675,818 context-switches

     200.029651391 seconds time elapsed
Signed-off-by: NEric Dumazet <edumazet@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Acked-By: NYuchung Cheng <ycheng@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Change snmp RETRANSFAILS stat to include timeout retransmit failures
in addition to other loss recoveries.
Signed-off-by: NYuchung Cheng <ycheng@google.com>
Acked-by: NNeal Cardwell <ncardwell@google.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In previous discussions, I tried to find some reasonable heuristics
for delayed ACK, however this seems not possible, according to Eric:

	"ACKS might also be delayed because of bidirectional
	traffic, and is more controlled by the application
	response time. TCP stack can not easily estimate it."

	"ACK can be incredibly useful to recover from losses in
	a short time.

	The vast majority of TCP sessions are small lived, and we
	send one ACK per received segment anyway at beginning or
	retransmits to let the sender smoothly increase its cwnd,
	so an auto-tuning facility wont help them that much."

and according to David:

	"ACKs are the only information we have to detect loss.

	And, for the same reasons that TCP VEGAS is fundamentally
	broken, we cannot measure the pipe or some other
	receiver-side-visible piece of information to determine
	when it's "safe" to stretch ACK.

	And even if it's "safe", we should not do it so that losses are
	accurately detected and we don't spuriously retransmit.

	The only way to know when the bandwidth increases is to
	"test" it, by sending more and more packets until drops happen.
	That's why all successful congestion control algorithms must
	operate on explicited tested pieces of information.

	Similarly, it's not really possible to universally know if
	it's safe to stretch ACK or not."

It still makes sense to enable or disable quick ack mode like
what TCP_QUICK_ACK does.

Similar to TCP_QUICK_ACK option, but for people who can't
modify the source code and still wants to control
TCP delayed ACK behavior. As David suggested, this should belong
to per-path scope, since different pathes may want different
behaviors.

Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Rick Jones <rick.jones2@hp.com>
Cc: Stephen Hemminger <stephen@networkplumber.org>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Thomas Graf <tgraf@suug.ch>
CC: David Laight <David.Laight@ACULAB.COM>
Signed-off-by: NCong Wang <amwang@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NWeiping Pan <wpan@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Linux sends new unset data during disorder and recovery state if all
(suspected) lost packets have been retransmitted ( RFC5681, section
3.2 step 1 & 2, RFC3517 section 4, NexSeg() Rule 2).  One requirement
is to keep the receive window about twice the estimated sender's
congestion window (tcp_rcv_space_adjust()), assuming the fast
retransmits repair the losses in the next round trip.

But currently it's not the case on the first round trip in either
normal or Fast Open connection, beucase the initial receive window
is identical to (expected) sender's initial congestion window. The
fix is to double it.
Signed-off-by: NYuchung Cheng <ycheng@google.com>
Acked-by: NNeal Cardwell <ncardwell@google.com>
Acked-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>