Re: [RFC PATCH v2 net-next 10/15] net: dsa: Pass ndo_setup_tc slave callback to drivers

Hi Vladimir,

On Mon, Sep 02, 2019 at 11:49:30AM +0300, Vladimir Oltean wrote:

I did something similar in v1 with a .port_setup_taprio in "[RFC PATCH
net-next 3/6] net: dsa: Pass tc-taprio offload to drivers".

Okay, didn't see that one.

Would this address Ilias's comment about DSA not really needing to
have this level of awareness into the qdisc offload type? Rightfully I
can agree that the added-value of making a .port_set_schedule and
.port_del_schedule in DSA compared to simply passing the ndo_setup_tc
is not that great.

I wanted to avoid that drivers have to the same kind of work, and it put
it therefore into the core part. However, I agree that the added-value
is not that high for TAPRIO.

By the way, thanks for the iproute2 patch for parsing 64-bit base time
on ARM 32, saved me a bit of debugging time :)

No problem :). That cost me a bit of time.

Regards,
-Vladimir

Thanks,
Kurt