On Wed, Dec 25, 2019 at 9:10 PM Mike C. <[email protected]> wrote:

> >
> > This behavior is going to depend on the switch chip embedded in the SoC
> > onTomas's device. The Linux kernel or its bridging behavior won't be
> > involved
> > until the traffic leaves the switch.
> >
>
> Ah, now I know why I didn't understand this nor agree with it.
>
> The embedded image of the Asus WL 500G wireless router that I tried to
> share but was rejected shows the 4 LAN ports hardwired together as a
> multi-port bridge all in VLAN 0.
>

That isn't a "bridge" in the linux sense. It is a switch chip. The switch
chip (e.g. the mt7530 in the mt7621 SoC[1] in my DLink router) is
configured from Linux, but it operates autonomously. All Linux sees are the
packets coming from the switch port the CPU is connected to. In cases where
WAN and LAN ports are both connected to the switch, those packets are VLAN
tagged so the CPU can tell them apart. Unless the switch chip has logic to
support some kind of loop avoidance (the three examples I cited don't),
they'll loop.

[1] https://patchwork.kernel.org/patch/8437611/
_______________________________________________
PLUG mailing list
[email protected]
http://lists.pdxlinux.org/mailman/listinfo/plug

Reply via email to