[PATCH v7 1/3] dts: allow mbuf_fast_free to be set with testpmd shell

Luca Vizzarro luca.vizzarro at arm.com
Tue Oct 21 15:28:51 CEST 2025


Hi Andrew,

Apologies for not spotting this before. Everything else is fine just as
before but:

On Thu, Oct 16, 2025 at 09:24:19AM +0000, Andrew Bailey wrote:
> diff --git a/dts/api/testpmd/__init__.py b/dts/api/testpmd/__init__.py
> index a060ab5639..39696ddde2 100644
> --- a/dts/api/testpmd/__init__.py
> +++ b/dts/api/testpmd/__init__.py
> @@ -1292,3 +1292,63 @@ def get_capabilities_physical_function(
>              supported_capabilities.add(NicCapability.PHYSICAL_FUNCTION)
>          else:
>              unsupported_capabilities.add(NicCapability.PHYSICAL_FUNCTION)
> +
> +    @_requires_stopped_ports
> +    def set_port_mbuf_fast_free(
> +        self,
> +        on: bool,
> +        port_id: int,

Out of consistency with the actual testpmd call and with all the other
methods, the port_id should come first. The arguments generally reflect
the other in which they are called in testpmd to make it "easier" for
someone who is familiar with it to pick up.

> +        /,
> +        verify: bool = True,
> +    ) -> None:
> +        """Sets the mbuf_fast_free configuration for the Tx offload of a given port.
> +
> +        Args:
> +            on: If :data:`True` mbuf_fast_free will be enabled, disable it otherwise.
> +            port_id: The ID of the port to configure mbuf_fast_free on.

Don't forget to swap it here as well.

> +            verify: If :data:`True` the output of the command will be scanned in an attempt to
> +                verify that the mbuf_fast_free was set successfully.
> +
> +        Raises:
> +            InteractiveCommandExecutionError: If mbuf_fast_free could not be set successfully.
> +        """
<snip>
> +
> +    @_requires_stopped_ports
> +    def set_queue_mbuf_fast_free(
> +        self,
> +        on: bool,
> +        port_id: int,

same here.

> +        /,
> +        queue_id: int = 0,
> +        verify: bool = True,
> +    ) -> None:

and the docstring here too.

Best,
Luca


More information about the dev mailing list