On Thu, Sep 10, 2020 at 12:05:17PM -0700, Florian Fainelli wrote: > On 9/10/2020 12:01 PM, Vladimir Oltean wrote: > > On Thu, Sep 10, 2020 at 11:42:02AM -0700, Florian Fainelli wrote: > > > On 9/10/2020 8:07 AM, Vladimir Oltean wrote: > > > Yes, doing what you suggest would make perfect sense for a DSA master that > > > is capable of VLAN filtering, I did encounter that problem with e1000 and > > > the dsa-loop.c mockup driver while working on a mock-up 802.1Q data path. > > > > Yes, I have another patch where I add those VLANs from tag_8021q.c which > > I did not show here. > > > > But if the DSA switch that uses tag_8021q is cascaded to another one, > > that's of little use if the upper switch does not propagate that > > configuration to its own upstream. > > Yes, that would not work. As soon as you have a bridge spanning any of those > switches, does not the problem go away by virtue of the switch port forcing > the DSA master/upstream to be in promiscuous mode?
Well, yes, bridged it works but standalone it doesn't. A bit strange if you ask me. -Vladimir
