Re: [RFC PATCH 0/4] net: dsa: link aggregation support

[Vladimir Oltean <olteanv@gmail.com>]

Hi Tobias,

On Tue, Oct 27, 2020 at 11:51:13AM +0100, Tobias Waldekranz wrote:
> I would really appreciate feedback on the following:
> 
> All LAG configuration is cached in `struct dsa_lag`s. I realize that
> the standard M.O. of DSA is to read back information from hardware
> when required. With LAGs this becomes very tricky though. For example,
> the change of a link state on one switch will require re-balancing of
> LAG hash buckets on another one, which in turn depends on the total
> number of active links in the LAG. Do you agree that this is
> motivated?

I don't really have an issue with that.

> The LAG driver ops all receive the LAG netdev as an argument when this
> information is already available through the port's lag pointer. This
> was done to match the way that the bridge netdev is passed to all VLAN
> ops even though it is in the port's bridge_dev. Is there a reason for
> this or should I just remove it from the LAG ops?

Maybe because on "leave", the bridge/LAG net device pointer inside
struct dsa_port is first set to NULL, then the DSA notifier is called?

> At least on mv88e6xxx, the exact source port is not available when
> packets are received on the CPU. The way I see it, there are two ways
> around that problem:
> 
> - Inject the packet directly on the LAG device (what this series
>   does). Feels right because it matches all that we actually know; the
>   packet came in on the LAG. It does complicate dsa_switch_rcv
>   somewhat as we can no longer assume that skb->dev is a DSA port.
> 
> - Inject the packet on "the designated port", i.e. some port in the
>   LAG. This lets us keep the current Rx path untouched. The problem is
>   that (a) the port would have to be dynamically updated to match the
>   expectations of the LAG driver (team/bond) as links are
>   enabled/disabled and (b) we would be presenting a lie because
>   packets would appear to ingress on netdevs that they might not in
>   fact have been physically received on.

Since ocelot/felix does not have this restriction, and supports
individual port addressing even under a LAG, you can imagine I am not
very happy to see the RX data path punishing everyone else that is not
mv88e6xxx.

> (mv88e6xxx) What is the policy regarding the use of DSA vs. EDSA?  It
> seems like all chips capable of doing EDSA are using that, except for
> the Peridot.

I have no documentation whatsoever for mv88e6xxx, but just wondering,
what is the benefit brought by EDSA here vs DSA? Does DSA have the
same restriction when the ports are in a LAG?