netdev - Re: [PATCH net-next v13 6/7] bonding: Update for extended arp_ip

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20251016124908.759bbb63@kernel.org>
Date: Thu, 16 Oct 2025 12:49:08 -0700
From: Jakub Kicinski <kuba@...nel.org>
To: Paolo Abeni <pabeni@...hat.com>
Cc: David Wilder <wilder@...ibm.com>, netdev@...r.kernel.org,
 jv@...sburgh.net, pradeep@...ibm.com, i.maximets@....org,
 amorenoz@...hat.com, haliu@...hat.com, stephen@...workplumber.org,
 horms@...nel.org, andrew+netdev@...n.ch, edumazet@...gle.com
Subject: Re: [PATCH net-next v13 6/7] bonding: Update for extended
 arp_ip_target format.

On Thu, 16 Oct 2025 13:50:52 +0200 Paolo Abeni wrote:
> > +		if (nla_put(skb, i, size, &data))
> > +			goto nla_put_failure;
> >  	}
> >  
> >  	if (targets_added)  
> 
> I guess you should update bond_get_size() accordingly???
> 
> Also changing the binary layout of an existing NL type does not feel
> safe. @Jakub: is that something we can safely allow?

In general extending attributes is fine, but going from a scalar 
to a struct is questionable. YNL for example will not allow it.

I haven't looked at the series more closely until now.

Why are there multiple vlan tags per target?

Is this configuration really something we should support in the kernel?
IDK how much we should push "OvS-compatibility" into other parts of the
stack. If user knows that they have to apply this funny configuration
on the bond maybe they should just arp from user space?