Le 20 nov. 2018 à 15:57, David Ahern <d...@cumulusnetworks.com> a écrit :
> 
> On 11/20/18 7:23 AM, Alexis Bauvin wrote:
>> Creating a VXLAN device with is underlay in the non-default VRF makes
>> egress route lookup fail or incorrect since it will resolve in the
>> default VRF, and ingress fail because the socket listens in the default
>> VRF.
>> 
>> This patch binds the underlying UDP tunnel socket to the l3mdev of the
>> lower device of the VXLAN device. This will listen in the proper VRF and
>> output traffic from said l3mdev, matching l3mdev routing rules and
>> looking up the correct routing table.
>> 
>> When the VXLAN device does not have a lower device, or the lower device
>> is in the default VRF, the socket will not be bound to any interface,
>> keeping the previous behaviour.
>> 
>> The underlay l3mdev is deduced from the VXLAN lower device
>> (IFLA_VXLAN_LINK).
>> 
>> The l3mdev_master_upper_ifindex_by_index function has been added to
>> l3mdev. Its goal is to fetch the effective l3mdev of an interface which
>> is not a direct slave of said l3mdev. It handles the following example,
>> properly resolving the l3mdev of eth0 to vrf-blue:
>> 
>> +----------+                         +---------+
>> |          |                         |         |
>> | vrf-blue |                         | vrf-red |
>> |          |                         |         |
>> +----+-----+                         +----+----+
>>     |                                    |
>>     |                                    |
>> +----+-----+                         +----+----+
>> |          |                         |         |
>> | br-blue  |                         | br-red  |
>> |          |                         |         |
>> +----+-----+                         +---+-+---+
>>     |                                   | |
>>     |                             +-----+ +-----+
>>     |                             |             |
>> +----+-----+                +------+----+   +----+----+
>> |          |  lower device  |           |   |         |
>> |   eth0   | <- - - - - - - | vxlan-red |   | tap-red | (... more taps)
>> |          |                |           |   |         |
>> +----------+                +-----------+   +---------+
> 
> same here. Very helpful diagram.
> 
> 
>> diff --git a/drivers/net/vxlan.c b/drivers/net/vxlan.c
>> index 27bd586b94b0..a3de08122269 100644
>> --- a/drivers/net/vxlan.c
>> +++ b/drivers/net/vxlan.c
> 
> The vxlan changes look ok to me. It would be good for someone who know
> that code better than I do to review it.
> 
> Move the following l3mdev changes to a separate patch - introduce infra
> changes separate from their use:

Sure, will do in next version. Thanks!

>> diff --git a/include/net/l3mdev.h b/include/net/l3mdev.h
>> index 3832099289c5..78fa0ac4613c 100644
>> --- a/include/net/l3mdev.h
>> +++ b/include/net/l3mdev.h
>> @@ -101,6 +101,17 @@ struct net_device *l3mdev_master_dev_rcu(const struct 
>> net_device *_dev)
>>      return master;
>> }
>> 
>> +int l3mdev_master_upper_ifindex_by_index_rcu(struct net *net, int ifindex);
>> +static inline
>> +int l3mdev_master_upper_ifindex_by_index(struct net *net, int ifindex)
>> +{
>> +    rcu_read_lock();
>> +    ifindex = l3mdev_master_upper_ifindex_by_index_rcu(net, ifindex);
>> +    rcu_read_unlock();
>> +
>> +    return ifindex;
>> +}
>> +
>> u32 l3mdev_fib_table_rcu(const struct net_device *dev);
>> u32 l3mdev_fib_table_by_index(struct net *net, int ifindex);
>> static inline u32 l3mdev_fib_table(const struct net_device *dev)
>> @@ -207,6 +218,17 @@ static inline int l3mdev_master_ifindex_by_index(struct 
>> net *net, int ifindex)
>>      return 0;
>> }
>> 
>> +static inline
>> +int l3mdev_master_upper_ifindex_by_index_rcu(struct net *net, int ifindex)
>> +{
>> +    return 0;
>> +}
>> +static inline
>> +int l3mdev_master_upper_ifindex_by_index(struct net *net, int ifindex)
>> +{
>> +    return 0;
>> +}
>> +
>> static inline
>> struct net_device *l3mdev_master_dev_rcu(const struct net_device *dev)
>> {
>> diff --git a/net/l3mdev/l3mdev.c b/net/l3mdev/l3mdev.c
>> index 8da86ceca33d..309dee76724e 100644
>> --- a/net/l3mdev/l3mdev.c
>> +++ b/net/l3mdev/l3mdev.c
>> @@ -46,6 +46,24 @@ int l3mdev_master_ifindex_rcu(const struct net_device 
>> *dev)
>> }
>> EXPORT_SYMBOL_GPL(l3mdev_master_ifindex_rcu);
>> 
>> +/**
>> + *  l3mdev_master_upper_ifindex_by_index - get index of upper l3 master
>> + *                                         device
>> + *  @net: network namespace for device index lookup
>> + *  @ifindex: targeted interface
>> + */
>> +int l3mdev_master_upper_ifindex_by_index_rcu(struct net *net, int ifindex)
>> +{
>> +    struct net_device *dev;
>> +
>> +    dev = dev_get_by_index_rcu(net, ifindex);
>> +    while (dev && !netif_is_l3_master(dev))
>> +            dev = netdev_master_upper_dev_get(dev);
>> +
>> +    return dev ? dev->ifindex : 0;
>> +}
>> +EXPORT_SYMBOL_GPL(l3mdev_master_upper_ifindex_by_index_rcu);
>> +
>> /**
>>  *   l3mdev_fib_table - get FIB table id associated with an L3
>>  *                             master interface

Reply via email to