Re: [PATCH v2 net-next 02/12] tsnep: deny tc-taprio changes to per-tc max SDU

[Vladimir Oltean]

On Mon, Sep 26, 2022 at 01:40:25PM -0700, Jakub Kicinski wrote:
> On Fri, 23 Sep 2022 19:33:00 +0300 Vladimir Oltean wrote:
> > Since the driver does not act upon the max_sdu argument, deny any other
> > values except the default all-zeroes, which means that all traffic
> > classes should use the same MTU as the port itself.
> 
> Don't all the driver patches make you wanna turn this into an opt-in?

Presumably you're thinking of a way through which the caller of
ndo_setup_tc(TC_SETUP_QDISC_TAPRIO, struct tc_taprio_qopt_offload *)
knows whether the driver took the new max_sdu field into consideration,
and not just accepted it blindly?

I'm not exactly up to date with all the techniques which can achieve
that without changes in drivers, and I haven't noticed other qdisc
offloads doing it either... but this would be a great trick to learn for
sure. Do you have any idea?

> What are the chances we'll catch all drivers missing the validation 
> in review?

Not that slim I think, they are all identifiable if you search for
TC_SETUP_QDISC_TAPRIO.