iproute4mac 0.6.0

iproute for Mac

This is a macOS network wrapper to imitate GNU/Linux iproute2 suite, inspired by the iproute2mac project.

Working staff

Command ip:

objects	implemented	supported	note
address	yes	yes
address	yes	yes
addrlabel	no	?	IPv6 protocol address label
maddress	no	?
route	yes	yes
rule	no	?	(e.g. source based routing with FreeBSD…)
neighbor	yes	yes	using ARP for IPv4 and NDP for IPv6
ntable	no	?
ntbl	no	?
link	yes	yes
l2tp	no	?
fou	no	?	IP-IP tunnel over UDP?
ila	no	?	IPv6 Identifier Locator Addressing
macsec	no	no
tunnel	no	yes	IP-IP only
tuntap	no	?	Tunnelblick third party tuntaposx?
token	no	?	May be related to non-numeric IPv6 mask?
tcpmetrics	no	?
monitor	no	no
xfrm	no	no
mroute	no	?	See Max OS: no multicast route for 127.0.0.1
mrule	no	?
netns	no	no
netconf	no	yes
vrf	no	?	Virtual Routing and Forwarding
sr	no	?	IPv6 Segment Routing management
nexthop	no	?
mptcp	no	no	Multipath TCP
ioam	no	?	IPv6 In-situ OAM (IOAM)
help	yes	yes
stats	no	?

Command bridge:

objects	implemented	supported	note
link	yes	yes	show only
fdb	yes	yes	show only

Examples:

• ip address [ list | show ]

• ip address { add | change | replace | delete }

• ip link [ list | show ]

• ip link { add | set | change | delete }

• ip route [ list | show ]

• ip route { add | change | replace | delete }

• ip route get

• ip neigh [ list | show ]

• ip neigh flush

• bridge link [ list | show ]

• bridge fdb [ list | show ]

Installation

Homebrew

The preferred method of installation is Homebrew.

In order to use this tap, you can install directly the package:

brew install signal-09/tap/iproute4mac

Or subscribe the tap repository and then install the package:

brew tap signal-09/tap
brew install iproute4mac

PyPI

Create a Virtual Environment and upgrade pip module:

python3 -m venv venv
source venv/bin/activate
python3 -m pip install -U pip

Then install iproute4mac:

python3 -m pip install iproute4mac

Usage

iproute4mac try to imitate as much as possible the look & feel of the GNU/Linux iproute2 suite, but macOS (Darwin) is a BSD derivative, and some aspects cannot be replicated at all.

ip address show: look at protocol addresses

Implemented syntax:

ip address [ show [ dev IFNAME ] [ master DEVICE | nomaster ] [ type TYPE ] [up] ]

TYPE := { bond | bond_slave | bridge | bridge_slave | feth⁵ | vlan }

Shows IPv4 and IPv6 addresses assigned to all network interfaces. The ‘show’ subcommand can be omitted:

ip address show

Same as above except that only IPv6 addresses assigned to active network interfaces are shown:

ip -6 address show up

Shows IPv4 and IPv6 addresses assigned to network interface en0 in pretty printed JSON:

ip -j -p address show dev en0

Shows IPv4 only addresses assigned to networks member of bridge0:

ip -4 address show master bridge0

Shows IP addresses belonging to private C class network 192.168.0.0/24:

ip address show to 192.168.0.0/24

Note:

1. qdisc (queuing discipline) is part of the Linux Traffic Control subsystem (TC) managed via the tc utility. Even if this information is not reported, similar results in traffic control and shaping can be achieved using dnctl and pfctl.

2. address lifetime for IPv6 addresses (-L flag of ifconfig) is not provided anymore in Sonoma (macOS 14); for IPv4, addresses valid and prederred lifetime, is supposed forever (0xffffffff = 4.294.967.295 = 32 bit).

3. sysctl net.inet6.ip6.temppltime specifies the “preferred lifetime” for privacy addresses, in seconds, and defaults to 86400 (one day).

4. sysctl net.inet6.ip6.tempvltime specifies the “valid lifetime” for privacy addresses, in second, and defaults to 604800 (one week).

5. veth can be replaced by feth in macOS

ip address add: add new protocol address

ip address change: change protocol address

ip address replace: change or add protocol address

Implemented syntax:

ip address {add|change|replace} IFADDR dev IFNAME [ LIFETIME ] [ CONFFLAG-LIST ]

Note:

1. {change|replace} option “really” change address properties (e.g. broadcast) while Linux simply ignore them.

ip address delete: delete protocol address

ip link show: display device attributes

Implemented syntax:

ip link show [ DEVICE ] [ up ] [ master DEVICE | nomaster ] [ type ETYPE ]

ETYPE := [ TYPE | bridge_slave | bond_slave ]

TYPE := [ bridge | bond ]

Shows the state of all network interfaces on the system:

ip link show

Shows the bridge devices:

ip link show type bridge

Shows the vlan devices:

ip link show type vlan

Shows devices enslaved by bridge0:

ip link show master bridge0

Note:

1. txqlen (the transmit queue length) is not configurable on specific interface; a system default value is managed via sysctl net.link.generic.system.sndq_maxlen (or net.link.generic.system.rcvq_maxlen).

ip link add: add virtual link

Implemented syntax:

ip link add [ link DEV ] [ name ] NAME [ address LLADDR ] [ mtu MTU ] type TYPE [ ARGS ]

Create a VLAN with TAG 100 linked to en1:

ip link add link en1 name vlan100 type vlan id 100

Create a new bridge interface (auto numbering1):

ip link add type bridge

Create a new bridge with a specified name:

ip link add bridge20 type bridge

Create a new static bond (vs lacp) interface:

ip link add bond1 type bond mode active-backup

Note:

1. macOS ifconfig print the created interface name to the standard output

ip link delete: delete virtual link

Implemented syntax:

ip link delete { DEVICE | dev DEVICE } type TYPE [ ARGS ]

Delete any kind of virtual interface:

ip link del vlan100

ip link set (or change): change device attributes

Implemented syntax:

ip route show: list routes

Implemented syntax:

ip route [ show [ SELECTOR ] ]

SELECTOR := [ proto RTPROTO ] [ type TYPE ] [ scope SCOPE ]

TYPE := { unicast | broadcast | multicast | blackhole }

SCOPE := [ host | link | global ]

RTPROTO := [ kernel | static ]

List routes using a specific gateway:

ip route show via 192.168.0.1

List IPv6 routes using a specific network interface:

ip -6 route show dev en1

List routes for multicast:

ip route show type multicast

List availabe routes to reach specific network:

ip route show to match 192.168.1.0/24

List IPv4 and IPv6 routes2

ip route show table all

Note:

1. iif is not honored (is treated like dev and oif).

2. Route tables are not implemented in macOS (Darwin), but “table all” will result in show IPv4 + IPv6 routes

ip route add: add new route

ip route delete: delete route

ip route change: change route

ip route replace: change or add new one

Implemented syntax:

ip route { add | delete | change | replace } ROUTE

ROUTE := [ TYPE ] PREFIX [ tos TOS ] [ proto RTPROTO ] [ scope SCOPE ]

TYPE := { unicast | broadcast | multicast | blackhole }

SCOPE := [ host | link | global ]

RTPROTO := [ kernel | boot | static ]

Add direct routing on a specific network interface:

ip route add 192.168.22.0/24 dev en1

Change interface for a given route:

ip route change 192.168.22.0/24 dev en0

Replace default gateway:

ip route replace default via 192.168.0.254

Delete route:

ip route del 192.168.22.0/24

ip route get: get a single route

Implemented syntax:

ip route get ADDRESS

Shows the route to reach Google DNS 8.8.8.8:

ip route get 8.8.8.8

ip neigh show: list neighbour entries

Note:

1. NOARP and PERMANENT states are not catched

ip neigh flush: flush neighbour entries

Same syntax of ip neigh show

Contributing

Every contribute is welcome!

Fork the repository

Fork button

Clone the fork

git clone https://github.com/YOUR-USERNAME/iproute4mac

Create a branch

Before making changes to the project, you should create a new branch and check it out (see “GitHub flow”).

git branch BRANCH-NAME
git checkout BRANCH-NAME

Create a developer environment

python3 -m venv venv
source venv/bin/activate

Then install requiered packages:

python3 -m pip install -U pip
python3 -m pip install pre-commit pytest pytest-console-scripts
pre-commit install

Coding style

Ruff is used to enforce coding style. You can checkout the compliance with the following command:

pre-commit run --all-files [--show-diff-on-failure]

Commit your work

Create as few commit as possible to make diff checking easier. In case of modification of already pushed commit, amend it if possible:

git add -A
git commit --amend
git push --force

In case of multiple and not organic commits, “Squash and merge” policy will be applied.