Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22364, verified by myself.

Make the fast-search stuff an option. There are still reports on
problem with it.

Fix previous bug. Thanks to Enami for spotting the (obvious) error, and
to other people with much help with bug reports etc.
While fixing, change some of the code I added last time to make it
cleaner

Fix previous bug. Thanks to Enami for spotting the (obvious) error, and
to other people with much help with bug reports etc.
While fixing, change some of the code I added last time to make it
cleaner and simpler.

show more ...

Disable the code I checked in yesterday; reports that samba (!) are crashing
machines with it. Will do some more tests.

Back out the lwp/ktrace changes. They contained a lot of colateral damage,
and need to be examined and discussed more.

Add code to remember where in the send queue of mbufs the last packet was
sent from. This change avoid a linear search through all mbufs when using
large TCP windows, and therefore permit high-speed

Add code to remember where in the send queue of mbufs the last packet was
sent from. This change avoid a linear search through all mbufs when using
large TCP windows, and therefore permit high-speed connections on long
distances.

Tested on a 1 Gigabit connection between Luleå and San Francisco, a distance
of about 15000km. With TCP windows of just over 20 Mbytes it could keep up
with 950Mbit/s.

After discussions with Matt Thomas and Jason Thorpe.

show more ...

no need for ip_v recovery in output path too
(tcp_template includes ip_v setting)

Allow TCP connections to hosts on a local network to use a larger
slow start initial window. Default this larger initial window to
4 packets, allowing it to be adjusted with net.inet.tcp.init_win_lo

Allow TCP connections to hosts on a local network to use a larger
slow start initial window. Default this larger initial window to
4 packets, allowing it to be adjusted with net.inet.tcp.init_win_local.

show more ...

Add MBUFTRACE kernel option.
Do a little mbuf rework while here. Change all uses of MGET*(*, M_WAIT, *)
to m_get*(M_WAIT, *). These are not performance critical and making them
call m_get saves con

Add MBUFTRACE kernel option.
Do a little mbuf rework while here. Change all uses of MGET*(*, M_WAIT, *)
to m_get*(M_WAIT, *). These are not performance critical and making them
call m_get saves considerable space. Add m_clget analogue of MCLGET and
make corresponding change for M_WAIT uses.
Modify netinet, gem, fxp, tulip, nfs to support MBUFTRACE.
Begin to change netstat to use sysctl.

show more ...

Fix a genuine uninitialised variable warning.

cleanup ipsec.h dependency. commented by perry, sync w/kame

In the txsegsize bounding code, it is not necessary to adjust for the options
length.

Never send more than half a socket buffer of data. This insures that
we can always keep 2 packets on the wire, no matter what SO_SNDBUF is,
and therefore ACKs will never be delayed unless we run out

Never send more than half a socket buffer of data. This insures that
we can always keep 2 packets on the wire, no matter what SO_SNDBUF is,
and therefore ACKs will never be delayed unless we run out of data to
transmit. The problem is quite easy to tickle when the MTU of the
outgoing interface is larger than the socket buffer size (e.g. loopback).

Fix from Charles Hannum.

show more ...

avoid swapping endian of ip_len and ip_off on mbuf, to meet with M_LEADINGSPACE
optimization made last year. should solve PR 17867 and 10195.

IP_HDRINCL behavior of raw ip socket is kept unchanged.

avoid swapping endian of ip_len and ip_off on mbuf, to meet with M_LEADINGSPACE
optimization made last year. should solve PR 17867 and 10195.

IP_HDRINCL behavior of raw ip socket is kept unchanged. we may want to
provide IP_HDRINCL variant that does not swap endian.

show more ...

Disable TCP Congestion Window Monitoring by default; there are
performance problems in the face of tinygrams.

whitespace

attach nd_ifinfo structure into if_afdata.
split IPv6 link MTU (advertised by RA) from real link MTU.
sync with kame

path MTU discovery blackhole detection.
PR 12790 (sorry for not committing it for a long time)

* Instrument tcp_build_datapkt().
* Remove the code that allocates a cluster if the packet would
fit in one; it totally defeats doing references to M_EXT mbufs
in the socket buffer. This drastic

* Instrument tcp_build_datapkt().
* Remove the code that allocates a cluster if the packet would
fit in one; it totally defeats doing references to M_EXT mbufs
in the socket buffer. This drastically reduces the number of
data copies in the tcp_output() path for applications which use
large writes. Kudos to Matt Thomas for pointing me in the right
direction.

show more ...

In tcp_segsize(), move a label so that option length is considered
when using the default TCP MSS as well. From Matt Thomas.

place NRL copyright notice itself, not a reference to it.

Fix TCP segment size computation. From Rick Byersm, PR kern/14799.

add RCSIDs

Use callouts for TCP timers, rather than traversing the list of
all open TCP connections in tcp_slowtimo() (which is called 2x
per second). It's fairly rare for TCP timers to actually fire,
so savin

Use callouts for TCP timers, rather than traversing the list of
all open TCP connections in tcp_slowtimo() (which is called 2x
per second). It's fairly rare for TCP timers to actually fire,
so saving this list traversal is good, especially if you want
to scale to thousands of open connections.

show more ...

Change the way receive idle time and round trip time are measured.
Instead of incrementing t_idle and t_rtt in tcp_slowtimo(), we now
take a timstamp (via tcp_now) and use subtraction to compute the

Change the way receive idle time and round trip time are measured.
Instead of incrementing t_idle and t_rtt in tcp_slowtimo(), we now
take a timstamp (via tcp_now) and use subtraction to compute the
delta when we actually need it (using unsigned arithmetic so that
tcp_now wrapping is handled correctly).

Based on similar changes in FreeBSD.

show more ...