> -----Original Message-----
> From: [email protected] <[email protected]>
> Sent: Wednesday, September 3, 2025 9:58 AM
> To: [email protected]
> Cc: [email protected]; [email protected];
> [email protected]; [email protected]; Keller, Jacob E
> <[email protected]>; Loktionov, Aleksandr
> <[email protected]>; Nguyen, Anthony L
> <[email protected]>; Kitszel, Przemyslaw
> <[email protected]>; Mohammad Heib <[email protected]>
> Subject: [PATCH net-next,2/2] i40e: support generic devlink param
> "max_mac_per_vf"
>
> From: Mohammad Heib <[email protected]>
>
> Add support for the new generic devlink runtime parameter
> "max_mac_per_vf", which controls the maximum number of MAC addresses a
> trusted VF can use.
Good day Mohammad,
Thanks for working on this and for the clear explanation in the commit message.
I have a couple of questions and thoughts:
1) Scope of the parameter
The name max_mac_per_vf is a bit ambiguous. From the description,
it seems to apply only to trusted VFs, but the name does not make that
obvious.
Would it make sense to either:
- Make the name reflect that (e.g., max_mac_per_trusted_vf), or
- Introduce two separate parameters for trusted and untrusted VFs if
both cases need to be handled differently?
2)Problem statement
It would help to better understand the underlying problem this parameter is
solving.
Is the goal to enforce a global cap for all VFs, or to provide operators
with a way
to fine-tune per-VF limits? From my perspective, the most important part is
clearly stating the problem and the use case.
3)Granularity
If the intent is to give operators flexibility, a single global parameter
might not be enough.
For example, limiting the number of MAC filters per specific VF (or having
different limits for trusted vs. untrusted)
could be a real-world requirement. This patch doesn't seem to address that
scenario.
Could you share more details about the use case and whether per-VF granularity
was considered?
Thanks again for the work on this. Looking forward to your thoughts.
Best regards,
Aleksandr
>
> By default (value 0), the driver enforces its internally calculated
> per-VF MAC filter limit. A non-zero value acts as a strict cap,
> overriding the internal calculation.
>
> Please note that the configured value is only a theoretical maximum
> and a hardware limits may still apply.
>
> - Previous discussion about this change:
> https://lore.kernel.org/netdev/20250805134042.2604897-1-
> [email protected]
> https://lore.kernel.org/netdev/20250823094952.182181-1-
> [email protected]
>
> Signed-off-by: Mohammad Heib <[email protected]>
> ---
...
> --
> 2.50.1
