Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.37 as permitted sender) client-ip=23.128.96.37;
Date:   Thu, 23 Nov 2023 14:35:02 +0100
From:   Maxime Chevallier <maxime.chevallier@bootlin.com>
To:     "Russell King (Oracle)" <linux@armlinux.org.uk>
Cc:     Andrew Lunn <andrew@lunn.ch>, davem@davemloft.net,
        netdev@vger.kernel.org, linux-kernel@vger.kernel.org,
        thomas.petazzoni@bootlin.com, Jakub Kicinski <kuba@kernel.org>,
        Eric Dumazet <edumazet@google.com>,
        Paolo Abeni <pabeni@redhat.com>,
        linux-arm-kernel@lists.infradead.org,
        Christophe Leroy <christophe.leroy@csgroup.eu>,
        Herve Codina <herve.codina@bootlin.com>,
        Florian Fainelli <f.fainelli@gmail.com>,
        Heiner Kallweit <hkallweit1@gmail.com>,
        Vladimir Oltean <vladimir.oltean@nxp.com>,
        =?UTF-8?B?S8O2cnk=?= Maincent <kory.maincent@bootlin.com>,
        Jesse Brandeburg <jesse.brandeburg@intel.com>
Subject: Re: [RFC PATCH net-next v2 04/10] net: sfp: Add helper to return
 the SFP bus name
Message-ID: <20231123143502.3a9a9047@device.home>
In-Reply-To: <ZVyEe0zH8Zo1NLFO@shell.armlinux.org.uk>
References: <20231117162323.626979-1-maxime.chevallier@bootlin.com>
        <20231117162323.626979-5-maxime.chevallier@bootlin.com>
        <00d26b50-56f1-4eac-a37f-36cf321bd46a@lunn.ch>
        <ZVyEe0zH8Zo1NLFO@shell.armlinux.org.uk>
Organization: Bootlin
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Precedence: bulk

Hi Andrew, Russell,

On Tue, 21 Nov 2023 10:20:43 +0000
"Russell King (Oracle)" <linux@armlinux.org.uk> wrote:

> On Tue, Nov 21, 2023 at 02:00:58AM +0100, Andrew Lunn wrote:
> > > +const char *sfp_get_name(struct sfp_bus *bus)
> > > +{
> > > +	if (bus->sfp_dev)
> > > +		return dev_name(bus->sfp_dev);
> > > +
> > > +	return NULL;
> > > +}  
> > 
> > Locking? Do you assume rtnl? Does this function need to take rtnl?  
> 
> Yes, rtnl needs to be held to safely access bus->sfp_dev, and that
> either needs to happen in this function, or be documented as being
> requried (and ASSERT_RTNL() added here.)
> 
> The reason is that sfp_dev is the SFP socket device which can be
> unbound via sfp_unregister_socket(), which will set bus->sfp_dev to
> NULL. This could race with the above.
> 

That's right, I'll add an assert and document it, thanks for spotting
this.

Maxime