Date: Mon, 13 Nov 2023 14:42:51 +0200
From: Nikolay Aleksandrov <razor@...ckwall.org>
To: Daniel Borkmann <daniel@...earbox.net>, stephen@...workplumber.org
Cc: martin.lau@...nel.org, netdev@...r.kernel.org
Subject: Re: [PATCH iproute2] ip, link: Add support for netkit

On 11/13/23 05:23, Daniel Borkmann wrote:
> Add base support for creating/dumping netkit devices.
> 
> Minimal example usage:
> 
>    # ip link add type netkit
>    # ip -d a
>    [...]
>    7: nk0@nk1: <BROADCAST,MULTICAST,NOARP,M-DOWN> mtu 1500 qdisc noop state DOWN group default qlen 1000
>      link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff promiscuity 0 allmulti 0 minmtu 68 maxmtu 65535
>      netkit mode l3 type peer policy forward numtxqueues 1 numrxqueues 1 [...]
>    8: nk1@nk0: <BROADCAST,MULTICAST,NOARP,M-DOWN> mtu 1500 qdisc noop state DOWN group default qlen 1000
>      link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff promiscuity 0 allmulti 0 minmtu 68 maxmtu 65535
>      netkit mode l3 type primary policy forward numtxqueues 1 numrxqueues 1 [...]
> 
> Example usage with netns (for BPF examples, see BPF selftests linked below):
> 
>    # ip netns add blue
>    # ip link add nk0 type netkit peer nk1 netns blue
>    # ip link set up nk0
>    # ip addr add 10.0.0.1/24 dev nk0
>    # ip -n blue link set up nk1
>    # ip -n blue addr add 10.0.0.2/24 dev nk1
>    # ping -c1 10.0.0.2
>    PING 10.0.0.2 (10.0.0.2) 56(84) bytes of data.
>    64 bytes from 10.0.0.2: icmp_seq=1 ttl=64 time=0.021 ms
> 
> Example usage with L2 mode and peer blackholing when no BPF is attached:
> 
>    # ip link add foo type netkit mode l2 forward peer blackhole bar
>    # ip -d a
>    [...]
>    13: bar@foo: <BROADCAST,MULTICAST,M-DOWN> mtu 1500 qdisc noop state DOWN group default qlen 1000
>       link/ether 5e:5b:81:17:02:27 brd ff:ff:ff:ff:ff:ff promiscuity 0 allmulti 0 minmtu 68 maxmtu 65535
>       netkit mode l2 type peer policy blackhole numtxqueues 1 numrxqueues 1 [...]
>    14: foo@bar: <BROADCAST,MULTICAST,M-DOWN> mtu 1500 qdisc noop state DOWN group default qlen 1000
>       link/ether de:01:a5:88:9e:99 brd ff:ff:ff:ff:ff:ff promiscuity 0 allmulti 0 minmtu 68 maxmtu 65535
>       netkit mode l2 type primary policy forward numtxqueues 1 numrxqueues 1 [...]
> 
> Signed-off-by: Daniel Borkmann <daniel@...earbox.net>
> Link: https://git.kernel.org/torvalds/c/35dfaad7188c
> Link: https://git.kernel.org/torvalds/c/05c31b4ab205
> Link: https://git.kernel.org/torvalds/c/ace15f91e569
> ---
>   (Targeted for iproute2 v6.7.0.)
> 

>   ip/Makefile              |   2 +-
>   ip/iplink.c              |   4 +-
>   ip/iplink_netkit.c       | 160 +++++++++++++++++++++++++++++++++++++++
>   man/man8/ip-address.8.in |   3 +-
>   man/man8/ip-link.8.in    |  44 +++++++++++
>   5 files changed, 209 insertions(+), 4 deletions(-)
>   create mode 100644 ip/iplink_netkit.c
> 
> diff --git a/ip/Makefile b/ip/Makefile
> index 8fd9e295..3535ba78 100644
> --- a/ip/Makefile
> +++ b/ip/Makefile
> @@ -13,7 +13,7 @@ IPOBJ=ip.o ipaddress.o ipaddrlabel.o iproute.o iprule.o ipnetns.o \
>       ipvrf.o iplink_xstats.o ipseg6.o iplink_netdevsim.o iplink_rmnet.o \
>       ipnexthop.o ipmptcp.o iplink_bareudp.o iplink_wwan.o ipioam6.o \
>       iplink_amt.o iplink_batadv.o iplink_gtp.o iplink_virt_wifi.o \
> -    ipstats.o
> +    iplink_netkit.o ipstats.o
>   
>   RTMONOBJ=rtmon.o
>   
> diff --git a/ip/iplink.c b/ip/iplink.c
> index 9a548dd3..6989cc4d 100644
> --- a/ip/iplink.c
> +++ b/ip/iplink.c
> @@ -46,8 +46,8 @@ void iplink_types_usage(void)
>   		"          dsa | dummy | erspan | geneve | gre | gretap | gtp | ifb |\n"
>   		"          ip6erspan | ip6gre | ip6gretap | ip6tnl |\n"
>   		"          ipip | ipoib | ipvlan | ipvtap |\n"
> -		"          macsec | macvlan | macvtap |\n"
> -		"          netdevsim | nlmon | rmnet | sit | team | team_slave |\n"
> +		"          macsec | macvlan | macvtap | netdevsim |\n"
> +		"          netkit | nlmon | rmnet | sit | team | team_slave |\n"
>   		"          vcan | veth | vlan | vrf | vti | vxcan | vxlan | wwan |\n"
>   		"          xfrm | virt_wifi }\n");
>   }
> diff --git a/ip/iplink_netkit.c b/ip/iplink_netkit.c
> new file mode 100644
> index 00000000..c539777a
> --- /dev/null
> +++ b/ip/iplink_netkit.c
> @@ -0,0 +1,160 @@
> +/* SPDX-License-Identifier: GPL-2.0-or-later */
> +/*
> + * iplink_netkit.c netkit device management
> + *
> + * Authors:        Daniel Borkmann <daniel@...earbox.net>
> + */
> +#include <stdio.h>
> +#include <stdlib.h>
> +#include <string.h>
> +#include <sys/socket.h>
> +#include <linux/if_link.h>
> +
> +#include "rt_names.h"
> +#include "utils.h"
> +#include "ip_common.h"
> +
> +static void explain(struct link_util *lu, FILE *f)
> +{
> +	fprintf(f,
> +		"Usage: ... %s [ mode MODE ] [ POLICY ] [ peer [ POLICY <options> ] ]\n"
> +		"\n"
> +		"MODE: l3 | l2\n"
> +		"POLICY: forward | blackhole\n"
> +		"(first values are the defaults if nothing is specified)\n"
> +		"\n"
> +		"To get <options> type 'ip link add help'.\n",
> +		lu->id);
> +}
> +
> +static bool seen_mode, seen_peer;
> +static struct rtattr *data;
> +
> +static int netkit_parse_opt(struct link_util *lu, int argc, char **argv,
> +			    struct nlmsghdr *n)
> +{
> +	__u32 ifi_flags, ifi_change, ifi_index;
> +	struct ifinfomsg *ifm, *peer_ifm;
> +	int err;
> +
> +	ifm = NLMSG_DATA(n);
> +	ifi_flags = ifm->ifi_flags;
> +	ifi_change = ifm->ifi_change;
> +	ifi_index = ifm->ifi_index;
> +	ifm->ifi_flags = 0;
> +	ifm->ifi_change = 0;
> +	ifm->ifi_index = 0;
> +	while (argc > 0) {
> +		if (matches(*argv, "mode") == 0) {

matches() should not be used anymore, only strcmp

> +			__u32 mode = 0;
> +
> +			NEXT_ARG();
> +			if (seen_mode)
> +				duparg("mode", *argv);
> +			seen_mode = true;
> +
> +			if (strcmp(*argv, "l3") == 0)
> +				mode = NETKIT_L3;
> +			else if (strcmp(*argv, "l2") == 0)
> +				mode = NETKIT_L2;

minor nit: curly braces for all cases

> +			else {
> +				fprintf(stderr, "Error: argument of \"mode\" must be either \"l3\" or \"l2\"\n");
> +				return -1;
> +			}
> +			addattr32(n, 1024, IFLA_NETKIT_MODE, mode);
> +		} else if (matches(*argv, "forward") == 0 ||
> +			   matches(*argv, "blackhole") == 0) {

same about matches and strcmp

> +			int attr_name = seen_peer ?
> +					IFLA_NETKIT_PEER_POLICY :
> +					IFLA_NETKIT_POLICY;
> +			__u32 policy = 0;
> +
> +			if (strcmp(*argv, "forward") == 0)
> +				policy = NETKIT_PASS;
> +			else if (strcmp(*argv, "blackhole") == 0)
> +				policy = NETKIT_DROP;
> +			else {
> +				fprintf(stderr, "Error: policy must be either \"forward\" or \"blackhole\"\n");
> +				return -1;
> +			}

same minor nit about the braces

> +			addattr32(n, 1024, attr_name, policy);
> +		} else if (matches(*argv, "peer") == 0) {

matches -> strcmp

> +			if (seen_peer)
> +				duparg("peer", *(argv + 1));
> +			seen_peer = true;
> +		} else {
> +			char *type = NULL;
> +
> +			if (seen_peer) {
> +				data = addattr_nest(n, 1024, IFLA_NETKIT_PEER_INFO);
> +				n->nlmsg_len += sizeof(struct ifinfomsg);
> +				err = iplink_parse(argc, argv, (struct iplink_req *)n, &type);
> +				if (err < 0)
> +					return err;
> +				if (type)
> +					duparg("type", argv[err]);
> +				goto out_ok;
> +			}
> +			fprintf(stderr, "%s: unknown option \"%s\"?\n",
> +				lu->id, *argv);
> +			explain(lu, stderr);
> +			return -1;
> +		}
> +		argc--;
> +		argv++;
> +	}
> +out_ok:
> +	if (data) {
> +		peer_ifm = RTA_DATA(data);
> +		peer_ifm->ifi_index = ifm->ifi_index;
> +		peer_ifm->ifi_flags = ifm->ifi_flags;
> +		peer_ifm->ifi_change = ifm->ifi_change;
> +		addattr_nest_end(n, data);
> +	}
> +	ifm->ifi_flags = ifi_flags;
> +	ifm->ifi_change = ifi_change;
> +	ifm->ifi_index = ifi_index;
> +	return 0;
> +}
> +
> +static void netkit_print_opt(struct link_util *lu, FILE *f, struct rtattr *tb[])
> +{
> +	if (!tb)
> +		return;
> +	if (tb[IFLA_NETKIT_MODE]) {
> +		__u32 mode = rta_getattr_u32(tb[IFLA_NETKIT_MODE]);
> +		const char *mode_str =
> +			mode == NETKIT_L2 ? "l2" :
> +			mode == NETKIT_L3 ? "l3" : "unknown";
> +
> +		print_string(PRINT_ANY, "mode", "mode %s ", mode_str);
> +	}
> +	if (tb[IFLA_NETKIT_PRIMARY]) {
> +		__u8 primary = rta_getattr_u8(tb[IFLA_NETKIT_PRIMARY]);
> +		const char *type_str = primary ? "primary" : "peer";
> +
> +		print_string(PRINT_ANY, "type", "type %s ", type_str);
> +	}
> +	if (tb[IFLA_NETKIT_POLICY]) {
> +		__u32 policy = rta_getattr_u32(tb[IFLA_NETKIT_POLICY]);
> +		const char *policy_str =
> +			policy == NETKIT_PASS ? "forward" :
> +			policy == NETKIT_DROP ? "blackhole" : "unknown";
> +
> +		print_string(PRINT_ANY, "policy", "policy %s ", policy_str);
> +	}
> +}
> +
> +static void netkit_print_help(struct link_util *lu,
> +			      int argc, char **argv, FILE *f)
> +{
> +	explain(lu, f);
> +}
> +
> +struct link_util netkit_link_util = {
> +	.id		= "netkit",
> +	.maxattr	= IFLA_NETKIT_MAX,
> +	.parse_opt	= netkit_parse_opt,
> +	.print_opt	= netkit_print_opt,
> +	.print_help	= netkit_print_help,
> +};
> diff --git a/man/man8/ip-address.8.in b/man/man8/ip-address.8.in
> index b9a476a5..9d34a6a1 100644
> --- a/man/man8/ip-address.8.in
> +++ b/man/man8/ip-address.8.in
> @@ -146,7 +146,8 @@ ip-address \- protocol address management
>   .BR ipvlan " |"
>   .BR lowpan " |"
>   .BR geneve " |"
> -.BR macsec " ]"
> +.BR macsec " |"
> +.BR netkit " ]"
>   
>   .SH "DESCRIPTION"
>   The
> diff --git a/man/man8/ip-link.8.in b/man/man8/ip-link.8.in
> index e82b2dbb..ca49b008 100644
> --- a/man/man8/ip-link.8.in
> +++ b/man/man8/ip-link.8.in
> @@ -247,6 +247,7 @@ ip-link \- network device configuration
>   .BR macvlan  " | "
>   .BR macvtap  " | "
>   .BR netdevsim " |"
> +.BR netkit " |"
>   .BR nlmon " |"
>   .BR rmnet " |"
>   .BR sit " |"
> @@ -384,6 +385,9 @@ Link types:
>   .BR netdevsim
>   - Interface for netdev API tests
>   .sp
> +.BR netkit
> +- BPF-programmable network device
> +.sp
>   .BR nlmon
>   - Netlink monitoring device
>   .sp
> @@ -848,6 +852,46 @@ tunnel.
>   
>   .in -8
>   
> +.TP
> +netkit Type Support
> +For a link of type
> +.I netkit
> +the following additional arguments are supported:
> +
> +.BI "ip link add " DEVICE
> +.BR type " netkit "
> +[
> +.BI mode " MODE "
> +] [
> +.I "POLICY "
> +] [
> +.BR peer
> +[
> +.I "POLICY "
> +] [
> +.I "NAME "
> +] ]
> +
> +.in +8
> +
> +.sp
> +.BI mode " MODE"
> +- specifies the operation mode of the netkit device with "l3" and "l2"
> +as possible values. Default option is "l3".
> +
> +.sp
> +.I "POLICY"
> +- specifies the default device policy when no BPF programs are attached
> +with "forward" and "blackhole" as possible values. Default option is
> +"forward". Specifying policy before the peer option refers to the primary
> +device, after the peer option refers to the peer device.
> +
> +.sp
> +.I "NAME"
> +- specifies the device name of the peer device.
> +
> +.in -8
> +
>   .TP
>   IPIP, SIT Type Support
>   For a link of type

Powered by blists - more mailing lists