[dpdk-dev] [PATCH 3/3] net/virtio: reject unsupported Rx multi queue modes

[Andrew Rybchenko]

On 10/10/19 10:42 AM, Thomas Monjalon wrote:
> 09/10/2019 13:24, Andrew Rybchenko:
>> On 10/9/19 1:41 PM, Tiwei Bie wrote:
>>> On Wed, Oct 09, 2019 at 12:00:28PM +0300, Andrew Rybchenko wrote:
>>>> And one more thought...
>>>> It looks like if more than one Rx queue is configured, mq_mode must
>>>> be ETH_MQ_RX_RSS and must not be ETH_MQ_RX_NONE.
>>> Some apps in DPDK will set mq_mode to ETH_MQ_RX_NONE while
>>> enabling multiple queue pairs, e.g.:
>>>
>>> https://github.com/DPDK/dpdk/blob/5b5bb56532fd5dea5d6260c0a0e405c0e32da2a0/app/test/test_link_bonding.c#L137
>>> https://github.com/DPDK/dpdk/blob/5b5bb56532fd5dea5d6260c0a0e405c0e32da2a0/app/test/test_link_bonding.c#L181-L182
>>> https://github.com/DPDK/dpdk/blob/5b5bb56532fd5dea5d6260c0a0e405c0e32da2a0/app/test/test_link_bonding.c#L3938-L3948
>>>
>>> Testpmd will also set mq_mode to ETH_MQ_RX_NONE when multiple
>>> Rx queues are enabled but rss_hf is empty:
>>>
>>> https://github.com/DPDK/dpdk/blob/5b5bb56532fd5dea5d6260c0a0e405c0e32da2a0/app/test-pmd/testpmd.c#L2935-L2938
>>> https://github.com/DPDK/dpdk/blob/5b5bb56532fd5dea5d6260c0a0e405c0e32da2a0/app/test-pmd/testpmd.c#L2945-L2948
>>> (the flow_type_rss_offloads reported by virtio-PMD is zero)
>>>
>>> My understanding is that, setting mq_mode to ETH_MQ_RX_NONE means
>>> no method is enforced on how to route packets to MQs.
>> I'm not sure. It is definitely a place to be improved in
>> ethdev documentation. Thomas, Ferruh, what do you think?
>> Is it really a definition of ETH_MQ_RX_NONE?
> I think it means everything go to queue 0.

I understand it this way as well.

> The comment says no DCB, RSS or VMDQ.
> It looks like the "NONE" value has been abused for some custom steering.
> We have two options:
> 	- document NONE as a possible case of custom steering
> 	- add a new CUSTOM value

I'd prefer to say that ETH_MQ_RX_RSS with rss_hf equal to 0 means
unspecified/unknown steering. If application just want to spread
traffic across many Rx queues, it is natural choice to say that
it want RSS, but do not care about spreading algorithm etc.
It allows driver use recommended defaults if rss_hf is controllable,
or just spread in virtio case.

Yes, it means that testpmd should be fixed for this specific case
to be able to enable RSS with rss_hf equal to 0.