tcp_output: don't try to send SACK option larger than txsegsize.
fix a panic like "panic: m_copydata: off 0, len -7".

tcp_output: remove duplicated code and tweak indent. no functional changes.

back out revision 1.144 calculating txsegsizep since it unmasks
other bugs. See PR kern/34674

consider sb_lowat when limiting the transmit length to keep acks on the wire

Import of TCP ECN algorithm for congestion control.
Both available for IPv4 and IPv6.
Basic implementation test results are available at
http://netbsd-soc.sourceforge.net/projects/ecn/testresults.htm

Import of TCP ECN algorithm for congestion control.
Both available for IPv4 and IPv6.
Basic implementation test results are available at
http://netbsd-soc.sourceforge.net/projects/ecn/testresults.html.

Work sponsored by the Google Summer of Code project 2006.
Special thanks to Kentaro Kurahone, Allen Briggs and Matt Thomas for their
help, comments and support during the project.

show more ...

Slight simplification of hdr len calculation in tcp_segsize().
No functional change.

Remove leading __ from __(const|inline|signed|volatile) -- it is obsolete.

merge ktrace-lwp.

wrap INET-only code by #if defined(INET).

ipv6 tx checksum offloading. reviewed by Jason Thorpe.

Implement PMTU checks from:

http://www.gont.com.ar/drafts/icmp-attacks-against-tcp.html

1. Don't act on ICMP-need-frag immediately if adhoc checks on the
advertised MTU fail. The MTU update

Implement PMTU checks from:

http://www.gont.com.ar/drafts/icmp-attacks-against-tcp.html

1. Don't act on ICMP-need-frag immediately if adhoc checks on the
advertised MTU fail. The MTU update is delayed until a TCP retransmit
happens.
2. Ignore ICMP Source Quench messages meant for TCP connections.

From OpenBSD.

show more ...

typo in comment

- add const
- remove bogus casts
- avoid nested variables

tcp_output: account FIN when building sack option.

tcp_output: don't try to send more data than we have. PR/30160.

tcp_output: clear TH_FIN where appropriate. related to PR/30160.

add a function to handle M_CSUM_TSOv4 by software.

fix problems related to loopback interface checksum omission. PR/29971.

- for ipv4, defer decision to ip layer as h/w checksum offloading does
so that it can check the actual interface the packet

fix problems related to loopback interface checksum omission. PR/29971.

- for ipv4, defer decision to ip layer as h/w checksum offloading does
so that it can check the actual interface the packet is going to.
- for ipv6, disable it.
(maybe will be revisited when it implements h/w checksum offloading.)

ok'ed by Jason Thorpe.

show more ...

tcp_output: lock reass queue when building sack.

simplify data receiver side sack processing.
- introduce t_segqlen, the number of segments in segq/timeq.
the name is from freebsd.
- rather than maintaining a copy of sack blocks (rcv_sack_block[]

simplify data receiver side sack processing.
- introduce t_segqlen, the number of segments in segq/timeq.
the name is from freebsd.
- rather than maintaining a copy of sack blocks (rcv_sack_block[]),
build it directly from the segment list when needed.

show more ...

- use full sized segments unless we actually have SACKs to send.
- avoid TSO duplicate D-SACK.
- send SACKs regardless of TF_ACKNOW.
- don't clear rcv_sack_num when transmitting.

discussed on tech-n

- use full sized segments unless we actually have SACKs to send.
- avoid TSO duplicate D-SACK.
- send SACKs regardless of TF_ACKNOW.
- don't clear rcv_sack_num when transmitting.

discussed on tech-net@.

show more ...

don't try to use TSO to transmit a single segment.
- there's no benefit.
- rtl8169 seems to be stuck with it.

For AF_INET, always set m->m_pkthdr.csum_data. Don't or TSOv4, just set it.

tcp_sack_option: the max number of sack blocks in a packet is 4, not 3.

Add a /*CONSTCOND*/ to last.