Re: [PATCH net-next v4] ipv6: add `force_forwarding` sysctl to enable per-interface forwarding

[Kuniyuki Iwashima <kuniyu@google.com>]

From: Gabriel Goller <g.goller@proxmox.com>
Date: Thu,  3 Jul 2025 18:01:53 +0200
> diff --git a/Documentation/networking/ip-sysctl.rst b/Documentation/networking/ip-sysctl.rst
> index 0f1251cce314..ec7fa1e890f1 100644
> --- a/Documentation/networking/ip-sysctl.rst
> +++ b/Documentation/networking/ip-sysctl.rst
> @@ -2292,6 +2292,11 @@ conf/all/forwarding - BOOLEAN
>  proxy_ndp - BOOLEAN
>  	Do proxy ndp.
>  
> +force_forwarding - BOOLEAN
> +	Enable forwarding on this interface only -- regardless of the setting on
> +	``conf/all/forwarding``. When setting ``conf.all.forwarding`` to 0,
> +	the ``force_forwarding`` flag will be reset on all interfaces.

Please update conf/all/forwarding too as it will be stale after
this patch.


> +
>  fwmark_reflect - BOOLEAN
>  	Controls the fwmark of kernel-generated IPv6 reply packets that are not
>  	associated with a socket for example, TCP RSTs or ICMPv6 echo replies).
> diff --git a/include/linux/ipv6.h b/include/linux/ipv6.h
> index 5aeeed22f35b..5380107e466c 100644
> --- a/include/linux/ipv6.h
> +++ b/include/linux/ipv6.h
> @@ -19,6 +19,7 @@ struct ipv6_devconf {
>  	__s32		forwarding;
>  	__s32		disable_policy;
>  	__s32		proxy_ndp;
> +	__s32		force_forwarding;

nit: place force_forwarding just after forwarding.


>  	__cacheline_group_end(ipv6_devconf_read_txrx);
>  
>  	__s32		accept_ra;
> @@ -857,6 +859,15 @@ static void addrconf_forward_change(struct net *net, __s32 newf)
>  		idev = __in6_dev_get_rtnl_net(dev);
>  		if (idev) {
>  			int changed = (!idev->cnf.forwarding) ^ (!newf);
> +			/*
> +			 * With the introduction of force_forwarding, we need to be backwards
> +			 * compatible, so that means we need to set the force_forwarding flag

Strictly backward compatibility is not related I think because
the per iface conf is disabled by default, and this is a new
behaviour.  Maybe simply say

/* Disabling all.forwarding sets 0 to force_forwarding for all interfaces */

> +			 * on every interface to 0 if net.ipv6.conf.all.forwarding is set to 0.
> +			 * This allows the global forwarding flag to disable forwarding for
> +			 * all interfaces.
> +			 */
> +			if (newf == 0)
> +				WRITE_ONCE(idev->cnf.force_forwarding, newf);
>  
>  			WRITE_ONCE(idev->cnf.forwarding, newf);
>  			if (changed)
> @@ -5719,6 +5730,7 @@ static void ipv6_store_devconf(const struct ipv6_devconf *cnf,
>  	array[DEVCONF_ACCEPT_UNTRACKED_NA] =
>  		READ_ONCE(cnf->accept_untracked_na);
>  	array[DEVCONF_ACCEPT_RA_MIN_LFT] = READ_ONCE(cnf->accept_ra_min_lft);
> +	array[DEVCONF_FORCE_FORWARDING] = READ_ONCE(cnf->force_forwarding);
>  }
>  
[...]
>  static inline size_t inet6_ifla6_size(void)
> @@ -6747,6 +6759,78 @@ static int addrconf_sysctl_disable_policy(const struct ctl_table *ctl, int write
>  	return ret;
>  }
>  
> +static void addrconf_force_forward_change(struct net *net, __s32 newf)
> +{
> +	ASSERT_RTNL();

__in6_dev_get_rtnl_net() has the same check so this is not needed.


> +	struct net_device *dev;
> +	struct inet6_dev *idev;
> +
> +	for_each_netdev(net, dev) {
> +		idev = __in6_dev_get_rtnl_net(dev);
> +		if (idev) {
> +			int changed = (!idev->cnf.force_forwarding) ^ (!newf);
> +
> +			WRITE_ONCE(idev->cnf.force_forwarding, newf);
> +			if (changed) {
> +				inet6_netconf_notify_devconf(dev_net(dev), RTM_NEWNETCONF,
> +							     NETCONFA_FORCE_FORWARDING,
> +							     dev->ifindex, &idev->cnf);
> +			}
> +		}
> +	}
> +}
> +
> +static int addrconf_sysctl_force_forwarding(const struct ctl_table *ctl, int write,
> +					    void *buffer, size_t *lenp, loff_t *ppos)
> +{
> +	struct inet6_dev *idev = ctl->extra1;
> +	struct net *net = ctl->extra2;
> +	int *valp = ctl->data;
> +	loff_t pos = *ppos;

nit: Please keep the reverse xmas tree order.
https://www.kernel.org/doc/html/latest/process/maintainer-netdev.html#local-variable-ordering-reverse-xmas-tree-rcs

> +	int new_val = *valp;
> +	int old_val = *valp;
> +	int ret;
> +
> +	struct ctl_table tmp_ctl = *ctl;

same here.

> +
> +	tmp_ctl.extra1 = SYSCTL_ZERO;
> +	tmp_ctl.extra2 = SYSCTL_ONE;

As you are copying *ctl, please specify this in addrconf_sysctl[].


> +	tmp_ctl.data = &new_val;
> +
> +	ret = proc_douintvec_minmax(&tmp_ctl, write, buffer, lenp, ppos);
> +
> +	if (write && old_val != new_val) {
> +		if (!rtnl_net_trylock(net))
> +			return restart_syscall();
> +
> +		if (valp == &net->ipv6.devconf_dflt->force_forwarding) {
> +			inet6_netconf_notify_devconf(net, RTM_NEWNETCONF,
> +						     NETCONFA_FORCE_FORWARDING,
> +						     NETCONFA_IFINDEX_DEFAULT,
> +						     net->ipv6.devconf_dflt);
> +		} else if (valp == &net->ipv6.devconf_all->force_forwarding) {
> +			inet6_netconf_notify_devconf(net, RTM_NEWNETCONF,
> +						     NETCONFA_FORCE_FORWARDING,
> +						     NETCONFA_IFINDEX_ALL,
> +						     net->ipv6.devconf_all);
> +
> +			addrconf_force_forward_change(net, new_val);
> +		} else {
> +			inet6_netconf_notify_devconf(net, RTM_NEWNETCONF,
> +						     NETCONFA_FORCE_FORWARDING,
> +						     idev->dev->ifindex,
> +						     &idev->cnf);
> +		}
> +		rtnl_net_unlock(net);
> +	}
> +
> +	if (write)
> +		WRITE_ONCE(*valp, new_val);
> +	if (ret)
> +		*ppos = pos;
> +	return ret;
> +}
> +
>  static int minus_one = -1;
>  static const int two_five_five = 255;
>  static u32 ioam6_if_id_max = U16_MAX;
> @@ -7217,6 +7301,13 @@ static const struct ctl_table addrconf_sysctl[] = {
>  		.extra1		= SYSCTL_ZERO,
>  		.extra2		= SYSCTL_TWO,
>  	},
> +	{
> +		.procname	= "force_forwarding",
> +		.data		= &ipv6_devconf.force_forwarding,
> +		.maxlen		= sizeof(int),
> +		.mode		= 0644,
> +		.proc_handler	= addrconf_sysctl_force_forwarding,

Here for extra{1,2}.


> +	},
>  };
>  
>  static int __addrconf_sysctl_register(struct net *net, char *dev_name,
> diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
> index 7bd29a9ff0db..440b9efced72 100644
> --- a/net/ipv6/ip6_output.c
> +++ b/net/ipv6/ip6_output.c
> @@ -509,7 +509,8 @@ int ip6_forward(struct sk_buff *skb)
>  	u32 mtu;
>  
>  	idev = __in6_dev_get_safely(dev_get_by_index_rcu(net, IP6CB(skb)->iif));
> -	if (READ_ONCE(net->ipv6.devconf_all->forwarding) == 0)
> +	if (idev && !READ_ONCE(idev->cnf.force_forwarding) &&
> +	    !READ_ONCE(net->ipv6.devconf_all->forwarding))

Now this ignores devconf_all when !idev whose dev was not
found or has not had a valid mtu.

if (!READ_ONCE(net->ipv6.devconf_all->forwarding) &&
    (!idev || !READ_ONCE(idev->cnf.force_forwarding)))