Re: [PATCH net-next v2 7/8] selftests: tls: skip the zero_len tests when TLS is unavailable

[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] Backport problem.

--

selftests: tls: skip the zero_len tests when TLS is unavailable

This patch makes TEST_F(zero_len, test) skip when the fixture has recorded
notls. Without the check, all eight variants fail on a kernel built without
CONFIG_TLS, because the raw records go over a plain TCP socket.

> diff --git a/tools/testing/selftests/net/tls.c b/tools/testing/selftests/net/tls.c
> index 419c6cc0cc5e1..e6876b8caac89 100644
> --- a/tools/testing/selftests/net/tls.c
> +++ b/tools/testing/selftests/net/tls.c
> @@ -2663,6 +2663,9 @@ TEST_F(zero_len, test)
>  	int rec_off;
>  	int i;
>  
> +	if (self->notls)
> +		SKIP(return, "no TLS support");
> +
>  	zero_len_send_recs(_metadata, self->fd, variant->recs);
                ^^^^

[Severity: Low]
This isn't a bug, but will this apply cleanly when it is backported on its
own?

The Fixes: tag points at a61a3e961baf ("selftests: tls: add tests for
zero-length records"), so stable tooling will probably try to pick this up.
However, the trailing context line above, the zero_len_send_recs() call,
comes from the previous patch in this series, "selftests: tls: add peek and
splice coverage for zero-length records". That patch has no Fixes: tag.

A tree that has a61a3e961baf but not that refactor still has the open-coded
send loop at this spot in TEST_F(zero_len, test):

	for (i = 0; i < 4 && variant->recs[i]; i++)
		EXPECT_EQ(send(self->fd, variant->recs[i]->cipher_data,
			       variant->recs[i]->cipher_len, 0),
			  variant->recs[i]->cipher_len);

So the hunk context won't match there, and someone would have to rebase the
fix by hand. Until then, stable kernels without TLS would keep reporting the
false zero_len failures.

The missing notls check has been there since a61a3e961baf. Could this patch
go first in the series, written against the original send loop, so it can
be backported without the refactor?

-- 
Sashiko AI review · https://netdev-ai.bots.linux.dev/sashiko/#/patchset/20261001-tls-follow-on-v2-0-2dd1947bb642%40kernel.org