TCA_HTB_DIRECT_QLEN attribute is supported since linux-3.10

HTB classes use an internal pfifo queue, which limit was not reported
by tc, and value inherited from device tx_queue_len at setup time.

With this patch, tc displays the value and can change it.
Signed-off-by: Eric Dumazet <edumazet@google.com>

Since linux-3.11, rate estimator can provide TCA_STATS_RATE_EST64
when rate (bytes per second) is above 2^32 (~34 Mbits)

Change tc to use this attribute for high rates.
Signed-off-by: Eric Dumazet <edumazet@google.com>

This augments bridge(8) manual page with `bridge mdb' and `bridge
monitor mdb' commands which have been added recently.
Signed-off-by: Petr Písař <ppisar@redhat.com>

This iproute2 tc patch is connected to the kernel
 - commit 8a8e3d84b17 (net_sched: restore "linklayer atm" handling)

The rate table calculated by tc, have gotten replaced in the kernel
and is no-longer used for lookups.

This happened in kernel release v3.8 caused by kernel
 - commit 56b765b79 ("htb: improved accuracy at high rates").
This change unfortunately caused breakage of tc overhead and
linklayer parameters.

 Kernel overhead handling got fixed in kernel v3.10 by
 - commit 01cb71d2d47 (net_sched: restore "overhead xxx" handling)

 Kernel linklayer handling got fixed in kernel v3.11 by
 - commit 8a8e3d84b17 (net_sched: restore "linklayer atm" handling)

The linklayer fix introduced a struct change, that allow the linklayer
attribute to be transferred between tc and kernel. This patch make use
of this linklayer attribute.

The linklayer setting is transfer to the kernel.  And linklayer
setting received from the kernel is printed with a prefixed
"linklayer" when listing current configuration.  The default
TC_LINKLAYER_ETHERNET is only printed in detailed output mode.
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>

Since commit a05f6511, ip batch mode is broken when using 'netns exec' cmd.

When WIFEXITED() returns true, it means that the child exited normally, hence
we must not call exit() but just returns the status. If we call exit, the next
commands in the file file are not executed.
If WIFEXITED() returns false, we can call exit() because it means that the
child failed.

This patch partially reverts commit a05f6511.
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>

Last minute addition to pkt_sched.h

The git-commit dc8867d0, that added support for displaying the
extra-flags of a state, introduced a potential segfault.
Trying to show a state without the extra-flag attribute and show_stats
enabled, would cause the NULL pointer in tb[XFRMA_SA_EXTRA_FLAGS] to be
dereferenced.
Signed-off-by: Thomas Egerer <thomas.egerer@secunet.com>

The mode information is contained in IFLA_MACVLAN_MODE instead
of IFLA_VLAN_ID (both evaluating to "1" in their enums).
Signed-off-by: Lutz Jaenicke <ljaenicke@innominate.com>

In SYNOPSIS section:

 - Add 'reordering'
 - Add missing '[' before 'quickack'
Signed-off-by: Richard Godbee <richard@godbee.net>

Signed-off-by: Richard Godbee <richard@godbee.net>

Fix two spacing problems around square brackets in usage text.
Signed-off-by: Richard Godbee <richard@godbee.net>

Useful to be able to monitor arp and IPv6 nd seperately.
Default is both.

Pfifo_fast needs no options. So don't force it to have parsing code.

Prefix labelling is currently only activated when monitoring "all"
objects.  However, the output can still be confusing when monitoring
more than 1 object, so add an option to always print prefix labels.
Signed-off-by: Martin Schwenke <martin@meltin.net>

Minor errors found by codespell

Split it naturally

Add support for the bridge fdb replace command to replace an
existing entry in the vxlan device driver forwarding data base.
The entry is identified with its unicast mac address and its
corresponding remote destination information is updated.

This is useful for virtual machine migration and replaces the
bridge fdb del and bridge fdb add commands.

It follows the same interface as ip neigh replace commands.
Signed-off-by: Thomas Richter <tmricht@linux.vnet.ibm.com>

This change adds the interface group to the output of "ip link show".

It also makes "ip link" print _all_ devices if no group filter is specified;
previously, only interfaces of the default group (0) were shown.
Signed-off-by: Stefan Tomanek <stefan.tomanek@wertarbyte.de>

-Werror just doesn't work because it changes too much
between compiler versions.

It's now possible to have IPv4 and IPv6 over IPv4 tunnels with the module sit.
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>

Because only IPv4 was supported, the size was static. Now, IPv6 also supports
peer address.
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>

This patch allows setting VXLAN destination to unicast address.
It allows that VXLAN can be used as peer-to-peer tunnel without
multicast.

v6: change back to the v3 except for using new attribute because
    replacing command-line parameters breaks existing scripts,
    based by Cong Wang's comments.

v5: rebase on the latest.

v4: replace "group" with "remote" based by David Stevens's comments.

v3: move a new attribute REMOTE into the last of an enum list
    based by Stephen Hemminger's comments.
    fix the usage to show explicitly that both "remote" and "group"
    cannot be specified, based by Ben Hutchings's comments.

v2: use a new argument "remote" instead of "group" based by
    Stephen Hemminger's comments.
Signed-off-by: Atzm Watanabe <atzm@stratosphere.co.jp>

Sanitized headers from upstream

Change the code for tc and ip so that batch mode is handled
the same.

The netns code was using EXIT_SUCCESS/EXIT_FAILURE but the rest of the ip
code used -1 explictly, so change to follow convention. Also, certain types
of errors like fork failure should abort a batch operation, rather than just
returning an error.

These do nothing on an 80-column display. They were clearly somebody's
boilerplate way of setting up hanging indents, but the syntax lines
are way too short to require them. And since most were argumentless
they would have been no-ops on any sized display.

The ip link command line help and the ip-link.8.in
man page are outdated in regards to the vxlan support.
The patch updates both the command line help for the
ip command and its man page.
Signed-off-by: Thomas Richter <tmricht@linux.vnet.ibm.com>

The bridge fdb command line help and the bridge.8
man page are outdated in regards to the vxlan support.
The patch updates both the command line help for the
bridge command and its man page.
Signed-off-by: Thomas Richter <tmricht@linux.vnet.ibm.com>

execvp() does not return when the command succeed, hence all commands in the
batch file after the line 'ip netns exec' are not executed.

Let's fork before calling execvp() if batch mode is used..

Example:
$ cat test.batch
netns add netns1
netns exec netns1 ip l
netns
$ ip -b test.batch
1: lo: <LOOPBACK> mtu 65536 qdisc noop state DOWN mode DEFAULT
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: sit0: <NOARP> mtu 1480 qdisc noop state DOWN mode DEFAULT
    link/sit 0.0.0.0 brd 0.0.0.0

All command after 'netns exec' are never executed.

With the patch:
$ ip -b test.batch
1: lo: <LOOPBACK> mtu 65536 qdisc noop state DOWN mode DEFAULT
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: sit0: <NOARP> mtu 1480 qdisc noop state DOWN mode DEFAULT
    link/sit 0.0.0.0 brd 0.0.0.0
netns1

Now, existing netns are displayed.
Signed-off-by: JunweiZhang <junwei.zhang@6wind.com>
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>

Without patch, I got:

	# ./ip/ip tunnel show
	ip_vti0: ioctl 89f4 failed: Invalid argument
	ip_vti0: ip/ip  remote any  local any  ttl inherit  nopmtudisc key 0

this is due to VTI_ISVTI has the same numeric value with SIT_ISATAP,
but only sit tunnel has SIOCGETPRL, therefore it should check for SIT
tunnel first.

Cc: Stephen Hemminger <stephen@networkplumber.org>
Signed-off-by: Cong Wang <amwang@redhat.com>

The command synopsis is regularized and part of it split off into an
OPTIONS section.  This allows the page to lift to XML-DocBook.

An invalid \p escape was removed.

This page was written by someone who didn't understand the use of
definite and indefinite articles in English, nor its punctuation rules.
I've fixed these mistakes, and some glitches in punctuation and
capitalization.

This includes x32, and, per Linus' decree, any future arch with longs
shorter than 64 bits.
Signed-off-by: Adam Borowski <kilobyte@angband.pl>

Allow ss -i to display more TCP informations :

unacked:N   Number of un-acked packets
retrans:X/Y   X: number of outstanding retransmit packets
              Y: total number of retransmits for the session
lost:N       Number of lost packets (tcpi_lost)
sacked:N     Number of sacked packets (tcpi_sacked)
facked:N     Number of facked packets (tcpi_facked)
reordering:N Reordering level (if different of 3)

Example :

$ ss -emoi dst 10.7.7.83
tcp   ESTAB      0      1154056   10.7.7.84:54127    10.7.7.83:34342
timer:(on,200ms,0) ino:57003 sk:ffff88063c51d0c0 <->
	 skmem:(r0,rb89280,t0,tb2097152,f726504,w1436184,o0,bl0) ts sack cubic
wscale:7,6 rto:310 rtt:107.375/1 mss:1448 cwnd:568 ssthresh:108 send
61.3Mbps unacked:568 retrans:0/21 reordering:127 rcv_space:29200
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>

Enclosed patch fixes inappropriate uses of the .SS macro.  Fuller explanation
in the change comment.

There are other problems in these pages that block lifting to
XML-DocBook, most notably in the command synopses.  They will take
some creativity to fix.  I'm working on it

>From 75745adba4b45b87577b61a2daa886dd444f44da Mon Sep 17 00:00:00 2001
From: "Eric S. Raymond" <esr@thyrsus.com>
Date: Fri, 21 Jun 2013 15:27:38 -0400
Subject: [PATCH] Abolish presentation-level misuse of the .SS macro.

This change fixes most (but not all) fatal errors in attempts to lift
the iproute2 manual pages to XML-DocBook.  Where .SS is still used it
is a real subsection header, not just a way to outdent and bold text.
Presentation-level instances are turned into .TP calls and tables.