netdev - Re: [PATCH net] net: dsa: fix bonding with ARP monitoring by updating trans

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] [day] [month] [year] [list]

Date:   Mon, 25 Jul 2022 21:53:19 +0000
From:   Vladimir Oltean <vladimir.oltean@....com>
To:     Jonathan Toppins <jtoppins@...hat.com>
CC:     Jakub Kicinski <kuba@...nel.org>,
        "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
        "David S. Miller" <davem@...emloft.net>,
        Eric Dumazet <edumazet@...gle.com>,
        Paolo Abeni <pabeni@...hat.com>, Andrew Lunn <andrew@...n.ch>,
        Vivien Didelot <vivien.didelot@...il.com>,
        Florian Fainelli <f.fainelli@...il.com>,
        Jay Vosburgh <j.vosburgh@...il.com>,
        Veaceslav Falico <vfalico@...il.com>,
        Hangbin Liu <liuhangbin@...il.com>,
        Brian Hutchinson <b.hutchman@...il.com>
Subject: Re: [PATCH net] net: dsa: fix bonding with ARP monitoring by updating
 trans_start manually

On Mon, Jul 25, 2022 at 05:39:33PM -0400, Jonathan Toppins wrote:
> > +	/* Stacked network interfaces usually have NETIF_F_LLTX so
> > +	 * netdev_start_xmit() -> txq_trans_update() fails to do anything,
> > +	 * because they don't lock the TX queue. However, layers such as the
> > +	 * bonding arp monitor may still use dev_trans_start() on slave
> > +	 * interfaces such as vlan, macvlan, DSA (or potentially stacked
> > +	 * combinations of the above). In this case, walk the adjacency lists
> > +	 * all the way down, hoping that the lower-most device won't have
> > +	 * NETIF_F_LLTX.
> > +	 */
> 
> I am not sure this assumption holds for virtual devices like veth unless the
> virtual interfaces are updated to modify trans_start, see
> e66e257a5d83 ("veth: Add updating of trans_start")
> 
> And not all the virtual interfaces update trans_start iirc.

Agree this hack won't give callers of dev_trans_start() something to chew
in all cases when "dev" is a virtual interface, that's why I said 'stacked'
and not 'virtual'.