Re: [PATCH net-next v2 1/2] selftests: drv-net: devmem: fix 5-tuple flow steering

On Fri, Aug 7, 2026 at 2:54 PM Jakub Kicinski <[email protected]> wrote:
>
> On Thu, 6 Aug 2026 13:59:53 -0700 Mina Almasry wrote:
> > Do you mean forking all the test cases so that we have
> > check_rx_3_tuple and check_rx_5_tuple, etc? Won't that be annoying in
> > the future if every test case needs to be duplicated twice? And
> > potentially again if there is another deviation in driver config
> > support specifics?
>
> No no, I just meant a trivial test that installs the rule and maybe
> sends a packet to check via getsockopt(NAPI_ID) that it arrived on
> the right queue.
>
> You can probably extend ntuple.py instead. I was thinking yesterday
> that it's nice to have all zcrx stuff in one place. But in reality
> we already have a bunch of scripts testing zcrx so ntuple.py is
> probably better.

Ack, will do. I assume you mean do that in a separate series. I'll
submit another series that extends ntuple.py with 2 tests, one for
3-tuple and one for 5-tuple flow steering.

For this series, I realized, yes, patch 2 is not needed indeed if I do
this on my test setup:

sudo sysctl -w net.ipv6.conf.eth1.accept_dad=0
sudo sysctl -w net.ipv4.ip_nonlocal_bind=1

so I'm going to drop patch 2. But I still need a bit from this patch.
Without setting the flow_steer= arg correctly on check_rx_hds that
test case specifically fails on 5-tuple drivers. Let me know if I
understood your feedback correctly.


-- 
Thanks,
Mina