Re: [RFC 3/9] net: dsa: mv88e6xxx: add support for VTU ops

[Florian Fainelli]

On 02/06/15 23:53, Scott Feldman wrote:
> On Tue, Jun 2, 2015 at 3:31 PM, nolan <nolan@xxxxxxxxxxxxxxxxxxx> wrote:
>> On 06/02/2015 12:44 AM, Scott Feldman wrote:
>>>
>>> That brings up an interesting point about having multiple bridges with
>>> the same vlan configured.  I struggled with that problem with rocker
>>> also and I don't have an answer other than "don't do that".  Or,
>>> better put, if you have multiple bridge on the same vlan, just use one
>>> bridge for that vlan.  Otherwise, I don't know how at the device level
>>> to partition the vlan between the bridges.  Maybe that's what Vivien
>>> is facing also?  I can see how this works for software-only bridges,
>>> because they should be isolated from each other and independent.  But
>>> when offloading to a device which sees VLAN XXX global across the
>>> entire switch, I don't see how we can preserve the bridge boundaries.
>>
>>
>> Scott,
>>
>> I'm confused by this.  I think you're saying this config is problematic:
>>
>> br0: eth0.100, eth1.100
>> br1: eth2.100, eth3.100
>>
>>
>> But this works fine today.
> 
> Ya, this is going to work because br0 and br1 are bridging untagged
> traffic, but br0 and br1 have different internal VLAN ids for untagged
> traffic, so there is no crosstalk between bridges.
> 
> (I'm assuming since you used the ethX.100 format, you've vconfig
> created a vlan interface on ethX and added the vlan interface to brY).
> The vlan interface eats the vlan tag and the bridge sees untagged
> traffic.
> 
>> Could you clarify the issue you're referring to?
> 
> I'm talking about bridging tagged traffic.  E.g.:
> 
> ip link add name br0 type bridge
> ip link add name br1 type bridge
> 
> ip link set dev sw1p1 master br0
> ip link set dev sw1p2 master br0
> ip link set dev sw1p3 master br1
> ip link set dev sw1p4 master br1
> 
> bridge vlan add vid 100 dev sw1p1
> bridge vlan add vid 100 dev sw1p2
> bridge vlan add vid 100 dev sw1p3
> bridge vlan add vid 100 dev sw1p4
> 
> Now the ports are trunking vlan 100 and the bridge/device see tagged
> traffic.  If the device used floods vlan 100 pkt to all ports in vlan
> 100, it'll flood to a port outside the bridge.  Oops!  For the device
> I'm using (rocker w/ OF-DPA) the bridging table matches on vlan ID and
> mac_dst.  There is no prevision to isolate vlans per bridge.
> 
> How do you solve the above case with your hardware?

For switches that support double-tagging, I suppose you could utilize
the fact that packets need to be normalized to include an internal outer
tag as well (for both tagged and untagged packets), and utilize that to
make sure there is no cross-talk. That way, you can have the same inner
VLAN tags on both of your bridges but the internal outer tag can be
different.

There might be a better solution, like having proper partitioning based
on whatever the switch is capable of?
-- 
Florian
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/