Re: [PATCH v3] net: netfilter: Fix rpfilter dropping vrf packets by mistake

[linmiaohe]

On 2019/5/13 17:42, Pablo Neira Ayuso wrote:
> On Thu, Apr 25, 2019 at 09:43:53PM +0800, linmiaohe wrote:
>> From: Miaohe Lin <linmiaohe@xxxxxxxxxx>
>>
>> When firewalld is enabled with ipv4/ipv6 rpfilter, vrf
>> ipv4/ipv6 packets will be dropped because in device is
>> vrf but out device is an enslaved device. So failed with
>> the check of the rpfilter.
>>
>> Signed-off-by: Miaohe Lin <linmiaohe@xxxxxxxxxx>
>> ---
>>  net/ipv4/netfilter/ipt_rpfilter.c  |  1 +
>>  net/ipv6/netfilter/ip6t_rpfilter.c | 10 +++++++++-
>>  2 files changed, 10 insertions(+), 1 deletion(-)
>>
> 
> Suggestion: Could you just call l3mdev_master_ifindex_rcu() when
> invoking rpfilter_lookup_reverse6() ?
> 
> diff --git a/net/ipv6/netfilter/ip6t_rpfilter.c b/net/ipv6/netfilter/ip6t_rpfilter.c
> index c3c6b09acdc4..ce64ff5d6fb6 100644
> --- a/net/ipv6/netfilter/ip6t_rpfilter.c
> +++ b/net/ipv6/netfilter/ip6t_rpfilter.c
> @@ -101,7 +101,8 @@ static bool rpfilter_mt(const struct sk_buff *skb,
> struct xt_action_param *par)
>         if (unlikely(saddrtype == IPV6_ADDR_ANY))
>                 return true ^ invert; /* not routable: forward path will drop it */
>  
> -       return rpfilter_lookup_reverse6(xt_net(par), skb, xt_in(par),
> +       return rpfilter_lookup_reverse6(xt_net(par), skb,
> +                                       l3mdev_master_ifindex_rcu(xt_in(par)),
>                                         info->flags) ^ invert;
>  }
> 
> .
>     rpfilter_lookup_reverse6 requests struct net_device *dev as third argument, so
what you really mean is this ?
 diff --git a/net/ipv6/netfilter/ip6t_rpfilter.c b/net/ipv6/netfilter/ip6t_rpfilter.c
 index c3c6b09acdc4..ce64ff5d6fb6 100644
 --- a/net/ipv6/netfilter/ip6t_rpfilter.c
 +++ b/net/ipv6/netfilter/ip6t_rpfilter.c
 @@ -101,7 +101,8 @@ static bool rpfilter_mt(const struct sk_buff *skb,
 struct xt_action_param *par)
         if (unlikely(saddrtype == IPV6_ADDR_ANY))
                 return true ^ invert; /* not routable: forward path will drop it */

 -       return rpfilter_lookup_reverse6(xt_net(par), skb, xt_in(par),
 +       return rpfilter_lookup_reverse6(xt_net(par), skb,
 +                                       l3mdev_master_dev_rcu(xt_in(par)) ? : xt_in(par),
                                         info->flags) ^ invert;
  }
    I'am sorry but I tested this. It doesn't work. When flags with XT_RPFILTER_LOOSE set,
we need set fl6.flowi6_oif to complete fib lookup in an l3mdev domain. And we need
enslaved network device to compute rpfilter rather than l3 master device.
    Many thanks for your suggestion.
    Best regards.