Re: [PATCH v4 net-next 13/15] selftests: xsk: list the hardware test cases

[netdev-bot+sashiko@kernel.org]

Thank you for your contribution! Sashiko AI review found more severe
issue(s) in other patches of this series. This mail is only an FYI, in case
you want to address the nit picks below when you respin.

Low priority / nit picks for your consideration:

- [Low] The new comment above the test tables in
  tools/testing/selftests/net/lib/xsk/test_xsk.h (lines 307-312) names…

--

selftests: xsk: list the hardware test cases

This splits tests[] in test_xsk.h into hw_rx_tx_tests[], hw_rx_tests[],
hw_tx_tests[] and generic_tests[], grouped in test_groups[], and test IDs
now follow the order of those arrays. It also adds a -L option to
xskxceiver, which lists each hardware test once per role it covers.

> diff --git a/tools/testing/selftests/net/lib/xsk/test_xsk.h b/tools/testing/selftests/net/lib/xsk/test_xsk.h
> index 15534f4ec8f4b..a59b34b156d79 100644
> --- a/tools/testing/selftests/net/lib/xsk/test_xsk.h
> +++ b/tools/testing/selftests/net/lib/xsk/test_xsk.h
> @@ -304,49 +304,73 @@ int testapp_xdp_shared_umem(struct test_spec *test);
>  int worker_testapp_validate_rx(struct test_spec *test);
>  int worker_testapp_validate_tx(struct test_spec *test);
>  
> -static const struct test_spec tests[] = {
> +/* Tests by the role of the endpoint that they check. A hardware run, see
> + * xskxceiver -L, gives that role to its zero-copy endpoint. It does not do
> + * the generic tests: BIDIRECTIONAL swaps the directions, which
> + * drivers/net/hw/xsk.py sets up one at a time, and the others skip in
> + * zero-copy mode.
> + */

[Severity: Low]
This isn't a bug, but the comment above hw_rx_tx_tests[] points to
drivers/net/hw/xsk.py, and that file is not in the tree at this commit.

The only xsk script under tools/testing/selftests/drivers/net/hw/ here is
xsk_reconfig.py. Nothing reads the hw_rx/hw_tx fields of struct test_group
or the -L output yet. The in-tree runner, net/test_xsk.py, still uses -l.

The next patch in the series, "selftests: drv-net: test AF_XDP zero-copy
with an SKB peer", adds drivers/net/hw/xsk.py. Its _get_cases() runs
xskxceiver -L and parses the "ID: NAME rx|tx" lines. So the reference
is correct once the whole series is applied.

Could the comment be added in the patch that adds xsk.py instead, so it
is accurate at every point in the series?

> +static const struct test_spec hw_rx_tx_tests[] = {
>  	{.name = "SEND_RECEIVE", .test_func = testapp_send_receive},
>  	{.name = "SEND_RECEIVE_2K_FRAME", .test_func = testapp_send_receive_2k_frame},
>  	{.name = "SEND_RECEIVE_SINGLE_PKT", .test_func = testapp_single_pkt},

[ ... ]

-- 
Sashiko AI review · https://netdev-ai.bots.linux.dev/sashiko/#/patchset/20261016174338.787135-1-maciej.fijalkowski%40intel.com