Re: [PATCH net-next v6 7/7] bonding: Selftest and documentation for the arp_ip_target parameter.

[Jakub Kicinski <kuba@kernel.org>]

On Fri, 8 Aug 2025 17:32:08 +0000 David Wilder wrote:
> > iproute2 is built from source a month ago, with some pending patches,
> > but not yours. Presumably building from source without your patches
> > should give similar effect (IIRC the patches I applied related
> > to MC routing)
> >  
> > > Can I access the logs from the CI run?  
> >
> > Yes
> >
> > https://netdev.bots.linux.dev/contest.html?pw-n=0&branch=net-next-2025-07-19--00-00 
> >  
> > > Is there a way I can debug in you CI environment?  
> >
> > Not at this point, unfortunately.
> >  
> > > Can I submit debug patches?  
> >
> > https://github.com/linux-netdev/nipa/wiki/How-to-run-netdev-selftests-CI-style   
> 
> I am able to run my tests in this environment, but sadly I still cant reproduce the failure.
> Can you tell me more about the environment you are using for CI?  What version of GCC is used?
> What distro and release is the environment build from?

Not sure the compiler and distro matters here:

the bond options tests prints a bunch of:

# RTNETLINK answers: Message too long
# Cannot send link get request: Message too long
# RTNETLINK answers: Message too long
# Cannot send link get request: Message too long
# Not enough information: "dev" argument is required.
# RTNETLINK answers: Message too long

> I have made some changes to the function that is failing In an attempt to fix the problem.
> If the problem continues with my next version would it be possible to save the build artifacts
> used in the test (vmlinux)?

Right, let's see if the problem persists. I updated iproute2 recently
too, so maybe that will help.