Re: [RFC PATCH net] net: taprio: Validate offload support using NETIF_F_HW_TC in hw_features

["Karumanchi, Vineeth" <vineeth@amd.com>]

Hi Eric,

On 7/29/2025 7:06 PM, Eric Dumazet wrote:
<...>
>>>> Note: Some drivers already set NETIF_F_HW_TC alongside .ndo_setup_tc.
>>>> Follow-up patches will be submitted to update remaining drivers if this
>>>> approach is accepted.
>>> Hi Vineeth
>>>
>>> Could you give more details ? "Some IP versions of a driver" and "Some
>>> drivers" are rather vague.
>> At present, I’m only familiar with the GEM IP, which supports TSN Qbv in
>> its later versions. The GEM implementations found in Zynq and ZynqMP
>> devices do not support TSN Qbv, whereas the updated versions integrated
>> into Versal devices do offer TSN Qbv support.
> Is this an out-of-tree driver ? I do not find macb_taprio_setup_replace()
> 
> I think most drivers should return -EOPNOTSUPP in this case.
> 


These are the patches of taprio implementation in macb.

https://lore.kernel.org/netdev/20250722154111.1871292-4-vineeth.karumanchi@amd.com/

https://lore.kernel.org/netdev/20250722154111.1871292-6-vineeth.karumanchi@amd.com/
Here’s a clearer and more polished version of your message:

  I initially considered adding the check in macb driver, but since it's 
a generic validation, I believe the ideal place for it would be within 
the TC framework.

But, I'm okay with adding the check in macb driver.

Thanks
-- 
🙏 vineeth