On Mon, Mar 31, 2014 at 4:05 PM, Damian Kaczkowski
<damian.kaczkowski+open...@gmail.com> wrote:
> On 31 March 2014 14:05, Saverio Proto <ziopr...@gmail.com> wrote:
>> Hello,
>> I cleaned up and rebased on current trunk the patch found on the
>> following ticket:
>> https://dev.openwrt.org/ticket/12181
>> I ask for help for proper testing.
> Hello there.
> Have a question. What about those switch chips:
> - ar8316,
> - ar8227,
> - ar7240.
> Are they also capable of doing tagged+untagged vlan on the same port at the
> same time? Do they benefit from this pach? Could anyone with a deeper
> knowledge comment on this topic please? Many thanks in advance.

It's a bit more complicated.

ar8216, ar8316:
- do not support mixed traffic at all. Either everything is tagged or
nothing is tagged on a port.

ar8227/8, ar8236
- support having the PVID untagged and everything else tagged through
a special port-tag-mode, so here only one vlan untagged at most, and
it always matches the PVID.

ar8327/8, ar8337/8.
- support per untagging/tagging on a per vlan per port basis on
egress. So here more than one vlan could be untagged on egress, but
the PVID still sets the incoming VID.

No idea about ar7240, ar933x, ar934x, ar935x, qca955x, although I
suspect ar7240 will fall into the same category as ar8216, and ar933x
and newer into the ar8227/8 category, maybe the "enterprise" class
ones even like ar8327.

openwrt-devel mailing list

Reply via email to