netdev - Re: [PATCH net-next] net/hsr: add interlink to fill

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]

Message-ID: <6d67143b-533d-450b-abb1-2db0857f563f@suse.de>
Date: Sat, 18 Oct 2025 12:10:59 +0200
From: Fernando Fernandez Mancera <fmancera@...e.de>
To: Jakub Kicinski <kuba@...nel.org>
Cc: Jan Vaclav <jvaclav@...hat.com>, "David S. Miller" <davem@...emloft.net>,
 Eric Dumazet <edumazet@...gle.com>, Simon Horman <horms@...nel.org>,
 Paolo Abeni <pabeni@...hat.com>, netdev@...r.kernel.org
Subject: Re: [PATCH net-next] net/hsr: add interlink to fill_info output



On 10/18/25 12:17 AM, Jakub Kicinski wrote:
> On Fri, 17 Oct 2025 12:17:21 +0200 Fernando Fernandez Mancera wrote:
>> On 10/17/25 12:57 AM, Jakub Kicinski wrote:
>>> On Wed, 15 Oct 2025 12:10:02 +0200 Jan Vaclav wrote:
>>>> Currently, it is possible to configure the interlink port, but no
>>>> way to read it back from userspace.
>>>>
>>>> Add it to the output of hsr_fill_info(), so it can be read from
>>>> userspace, for example:
>>>>
>>>> $ ip -d link show hsr0 12: hsr0: <BROADCAST,MULTICAST> mtu ... ...
>>>> hsr slave1 veth0 slave2 veth1 interlink veth2 ...
>>>
>>> Not entirely cleat at a glance how this driver deals with the slaves
>>> or interlink being in a different netns, but I guess that's a pre-
>>> existing problem..
>>>    
>>
>> FTR, I just did a quick round of testing and it handles it correctly.
>> When moving a port to a different netns it notifies NETDEV_UNREGISTER -
>> net/hsr/hsr_main.c handles the notification removing the port from the
>> list. If the port list is empty, removes the hsr link.
>>
>> All good or at least as I would expect.
> 
> Did you try to make the slave/interlink be in a different namespace
> when HSR interface is created? It should be possible with the right
> combination of NEWLINK attributes, and that's the config I was
> particularly concerned about..
> 

Ugh, right. It is completely messed up indeed. I am handling this in a 
patch series. Thanks you Jakub.