Re: [PATCH] selftests: rtnetlink: add checks for ifconfig and iproute2

[Hangbin Liu <liuhangbin@gmail.com>]

On Thu, Aug 21, 2025 at 10:45:19AM +0200, Alessandro wrote:
> On Thu, 21 Aug 2025 at 10:25, Hangbin Liu <liuhangbin@gmail.com> wrote:
> >
> > On Thu, Aug 21, 2025 at 09:43:11AM +0200, Alessandro Ratti wrote:
> > > On systems where `ifconfig` is not available (e.g., modern Debian), the
> > > `kci_test_promote_secondaries` test fails. Wrap the call in a check.
> > >
> > > Additionally, `do_test_address_proto` fails on iproute2 versions that
> > > lack support for `proto` in `ip address` commands. Add a minimal feature
> > > check and skip the test with a proper message if unsupported.
> > >
> > > These changes allow the tests to run and report SKIP instead of FAIL on
> > > platforms with older tools.
> > >
> > > Signed-off-by: Alessandro Ratti <alessandro@0x65c.net>
> > > ---
> > >  tools/testing/selftests/net/rtnetlink.sh | 10 +++++++++-
> > >  1 file changed, 9 insertions(+), 1 deletion(-)
> > >
> > > diff --git a/tools/testing/selftests/net/rtnetlink.sh b/tools/testing/selftests/net/rtnetlink.sh
> > > index d6c00efeb664..9bff620ef595 100755
> > > --- a/tools/testing/selftests/net/rtnetlink.sh
> > > +++ b/tools/testing/selftests/net/rtnetlink.sh
> > > @@ -330,7 +330,9 @@ kci_test_promote_secondaries()
> > >       for i in $(seq 2 254);do
> > >               IP="10.23.11.$i"
> > >               ip -f inet addr add $IP/16 brd + dev "$devdummy"
> > > -             ifconfig "$devdummy" $IP netmask 255.255.0.0
> > > +             if command -v ifconfig >/dev/null 2>&1; then
> > > +                     ifconfig "$devdummy" $IP netmask 255.255.0.0
> > > +             fi
> >
> > Maybe just skip the promote_secondaries test if ifconfig is not available?
> >
> 
> Thank you for your review and comment.
> 
> My takeaway here is that the test works because the IP addresses are set on the
> $devdummy by the previous ip(8) command, and ifconfig seems a bit redundant.

No, please check the git log to see why we use ifconfig here.

Thanks
Hangbin

> 
> Also, considering we are testing netlink, I was baffled to see ifconfig there
> that, if I'm not mistaken, uses ioctl(); but I might be missing
> something obvious
> here, considering I'm looking at these tests for the first time, so bear with
> me :)
> 
> If it's better to skip the test altogether when ifconfig is missing, I'll
> submit another patch to do so.
> 
> Thank you
> 
> Best regards,
> Alessandro