[PATCH] net: route: Fix vrf dst_entry ref count false increasing

[PATCH] net: route: Fix vrf dst_entry ref count false increasing

[linmiaohe]

From: Suanming.Mou <mousuanming@huawei.com>

When config ip in default vrf same as the ip in specified
vrf, fib_lookup will return the route from table local
even if the in device is an enslaved l3mdev. Then the
dst_entry will hold the vrf device rather than loopback
device in local_input of function ip_route_input_slow.
So vrf dst_entry is false increased by route from table
local.

Here is reproduce step:
1.enslave enp4s0 to vrf2, and config ip address:
ip link add vrf2 type vrf table 1
ip link set vrf2 up
ip link set enp4s0 master vrf2
ip addr ad 125.1.1.1/16 dev enp4s0

2.config same ip in default vrf:
ip addr ad 125.1.1.1/16 dev enp6s0

3.config peer and ping:
ip vrf exec vrf2 ping 125.1.1.2 -c 3

4.del vrf2 link:
ip link del vrf2

And "unregister_netdevice: waiting for vrf2 to become free.
Usage count = 1" will occur.

Signed-off-by: Suanming.Mou <mousuanming@huawei.com>
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
---
 net/core/fib_rules.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/net/core/fib_rules.c b/net/core/fib_rules.c
index ffbb827723a2..1a2c11ed1585 100644
--- a/net/core/fib_rules.c
+++ b/net/core/fib_rules.c
@@ -263,6 +263,11 @@ static int fib_rule_match(struct fib_rule *rule, struct fib_rules_ops *ops,
 	if (rule->tun_id && (rule->tun_id != fl->flowi_tun_key.tun_id))
 		goto out;

+	if (!rule->l3mdev &&
+	    (netif_index_is_l3_master(rule->fr_net, fl->flowi_iif) ||
+	     netif_index_is_l3_master(rule->fr_net, fl->flowi_oif)))
+		goto out;
+
 	if (rule->l3mdev && !l3mdev_fib_rule_match(rule->fr_net, fl, arg))
 		goto out;

-- 
2.21.GIT

Re: [PATCH] net: route: Fix vrf dst_entry ref count false increasing

[linmiaohe]

On 2019/5/4 16:03, linmiaohe wrote:
> From: Suanming.Mou <mousuanming@huawei.com>
> 
> When config ip in default vrf same as the ip in specified
> vrf, fib_lookup will return the route from table local
> even if the in device is an enslaved l3mdev. Then the
> dst_entry will hold the vrf device rather than loopback
> device in local_input of function ip_route_input_slow.
> So vrf dst_entry is false increased by route from table
> local.
> 
> Here is reproduce step:
> 1.enslave enp4s0 to vrf2, and config ip address:
> ip link add vrf2 type vrf table 1
> ip link set vrf2 up
> ip link set enp4s0 master vrf2
> ip addr ad 125.1.1.1/16 dev enp4s0
> 
> 2.config same ip in default vrf:
> ip addr ad 125.1.1.1/16 dev enp6s0
> 
> 3.config peer and ping:
> ip vrf exec vrf2 ping 125.1.1.2 -c 3
> 
> 4.del vrf2 link:
> ip link del vrf2
> 
> And "unregister_netdevice: waiting for vrf2 to become free.
> Usage count = 1" will occur.
> 
> Signed-off-by: Suanming.Mou <mousuanming@huawei.com>
> Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
> ---
>  net/core/fib_rules.c | 5 +++++
>  1 file changed, 5 insertions(+)
> 
> diff --git a/net/core/fib_rules.c b/net/core/fib_rules.c
> index ffbb827723a2..1a2c11ed1585 100644
> --- a/net/core/fib_rules.c
> +++ b/net/core/fib_rules.c
> @@ -263,6 +263,11 @@ static int fib_rule_match(struct fib_rule *rule, struct fib_rules_ops *ops,
>  	if (rule->tun_id && (rule->tun_id != fl->flowi_tun_key.tun_id))
>  		goto out;
> 
> +	if (!rule->l3mdev &&
> +	    (netif_index_is_l3_master(rule->fr_net, fl->flowi_iif) ||
> +	     netif_index_is_l3_master(rule->fr_net, fl->flowi_oif)))
> +		goto out;
> +
>  	if (rule->l3mdev && !l3mdev_fib_rule_match(rule->fr_net, fl, arg))
>  		goto out;
> 

I'am sorry, but I think this fix looks bad because this patch make vrf
working with other ip rule impossible. I will send anothor patch to fix
this. Thanks.