tcp: Push bit failure to set in fastpath

[rrs]

Summary

Recently changes were made to the tcp stack to use a macro/function
to set tcp flags. In the process the PUSH bit setting in the fastpath of
rack was broken. This fixes that as well as cleans up a warning that
is occurring when you don't have INVARIANT on (inp used in KASSERT).

We can use the tcp test suite to find this bug the test plan shows the script
that fails due to the missing push bit

Test Plan

All rights reserved.

Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions
are met:
1. Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
2. Redistributions in binary form must reproduce the above copyright
notice, this list of conditions and the following disclaimer in the
documentation and/or other materials provided with the distribution.

THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
SUCH DAMAGE.
//

--ip_version=ipv4
--tolerance_usecs=50000

0.00 sysctl -w net.inet.tcp.hostcache.purgenow=1
+0.00 sysctl -w net.inet.tcp.syncookies_only=0
+0.00 sysctl -w net.inet.tcp.syncookies=1
+0.00 sysctl -w net.inet.tcp.rfc1323=0
+0.00 sysctl -w net.inet.tcp.sack.enable=0
+0.00 sysctl -w net.inet.tcp.ecn.enable=2
+0.00 sysctl -w net.inet.tcp.recvspace=65536

+0.00 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0.00 fcntl(3, F_GETFL) = 0x02 (flags O_RDWR)
+0.00 fcntl(3, F_SETFL, O_RDWR | O_NONBLOCK) = 0
+0.00 setsockopt(3, IPPROTO_TCP, TCP_NOOPT, [1], 4) = 0
+0.00 connect(3, ..., ...) = -1 EINPROGRESS (Operation now in progress)
+0.00 > S 0:0(0) win 65535
+0.05 < S 0:0(0) win 65535 <sackOK,TS val 100 ecr 0>
+0.00 > S. 0:0(0) ack 1 win 65535
+0.00 < . 1:1(0) ack 1 win 65535
+0.00 %{ assert tcpi_state == TCPI_ESTABLISHED }%
+0.00 %{ assert tcpi_options == 0}%
+0.05 write(3, ..., 5360) = 5360
+0.00 > . 1:537(536) ack 1 win 65535
+0.00 > . 537:1073(536) ack 1 win 65535
+0.00 > . 1073:1609(536) ack 1 win 65535
+0.00 > . 1609:2145(536) ack 1 win 65535
+0.00 > . 2145:2681(536) ack 1 win 65535
+0.00 > . 2681:3217(536) ack 1 win 65535
+0.00 > . 3217:3753(536) ack 1 win 65535
+0.00 > . 3753:4289(536) ack 1 win 65535
+0.00 > . 4289:4825(536) ack 1 win 65535
+0.00 > P. 4825:5361(536) ack 1 win 65535

// Validate DupThresh loss recovery works
+0.01 < . 1:1(0) ack 537 win 65535
+0.01 < . 1:1(0) ack 537 win 65535
+0.01 < . 1:1(0) ack 537 win 65535
+0.01 < . 1:1(0) ack 537 win 65535
+0.00 > . 537:1073(536) ack 1 win 65535

// Verify that no SACK block is sent out
+0.00 < . 10:20(10) ack 5361 win 65535
+0.00 > . 5361:5361(0) ack 1 win 65535

+0.01 close(3) = 0

[rscheff]

Sorry for the oversight.

sys/netinet/tcp_stacks/rack.c

16224

This should have been '''flags |= TH_PUSH''' here, since

16241

down here, tcp_ecn_output_established() will set some of the TCP header flags.

16252

Thus, the flags have to be transferred to the header only after all else.

[rrs]

sys/netinet/tcp_stacks/rack.c

16224

I am fine with that, I will move the line back down.. I was
just following what your earlier code changes had done... I
believe before your changes it *was* a flags |= TH_PUSH :)

[rrs]

changes back to what it was before i.e. a |= and moved the flags
set back down.

[rscheff]

Possibly - probably missed this over my iterations. Thanks!

[tuexen]

sys/netinet/tcp_stacks/rack.c

6262–6263

Why not use:

struct inpcb *inp = tp->t_inpcb;

This reduces one #ifdef...

6264

Please keep an empty line between the declarations and the code.