linux-kernel - Re: [PATCH net-next v3] net: phy: Fix suspicious rcu

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 for Android: free password hash cracker in your pocket

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

Message-ID: <20250123150537.5288d894@kmaincent-XPS-13-7390>
Date: Thu, 23 Jan 2025 15:05:37 +0100
From: Kory Maincent <kory.maincent@...tlin.com>
To: Claudiu Beznea <claudiu.beznea@...on.dev>
Cc: Paul Barker <paul.barker.ct@...renesas.com>, Jakub Kicinski
 <kuba@...nel.org>, "David S. Miller" <davem@...emloft.net>,
 netdev@...r.kernel.org, linux-kernel@...r.kernel.org, Claudiu Beznea
 <claudiu.beznea.uj@...renesas.com>, thomas.petazzoni@...tlin.com, Andrew
 Lunn <andrew@...n.ch>, Heiner Kallweit <hkallweit1@...il.com>, Russell King
 <linux@...linux.org.uk>, Eric Dumazet <edumazet@...gle.com>, Paolo Abeni
 <pabeni@...hat.com>, Niklas Söderlund
 <niklas.soderlund@...natech.se>, Sergey Shtylyov <s.shtylyov@....ru>
Subject: Re: [PATCH net-next v3] net: phy: Fix suspicious rcu_dereference
 usage

On Thu, 23 Jan 2025 13:25:57 +0200
Claudiu Beznea <claudiu.beznea@...on.dev> wrote:

> >> ravb_ptp_stop() modifies a couple of device registers and calls
> >> ptp_clock_unregister(). I don't see anything to suggest that this
> >> requires the rtnl lock to be held, unless I am missing something.  
> > 
> > What happens if two ptp_clock_unregister() with the same ptp_clock pointer
> > are called simultaneously? From ravb_suspend and ravb_set_ringparam for
> > example. It may cause some errors.  
> 
> Can this happen? I see ethtool_ops::set_ringparam() references only in
> ethtool or ioctl files:
> 
> net/ethtool/ioctl.c:2066
> net/ethtool/ioctl.c:2081
> net/ethtool/rings.c:212
> net/ethtool/rings.c:304
> 
> At the time the suspend/resume APIs are called the user space threads are
> frozen.

Maybe, I don't know the suspend path, and what the state of user space threads
at that time. This was an example but Wake on Lan setup could also have some
issue. IMHO I think it is more precautions to have it under rtnl lock.

Regards,
-- 
Köry Maincent, Bootlin
Embedded Linux and kernel engineering
https://bootlin.com