[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20170406142555.GD17713@lunn.ch>
Date: Thu, 6 Apr 2017 16:25:55 +0200
From: Andrew Lunn <andrew@...n.ch>
To: Florian Fainelli <f.fainelli@...il.com>
Cc: Juergen Borleis <jbe@...gutronix.de>, netdev@...r.kernel.org,
linux-kernel@...r.kernel.org, kernel@...gutronix.de,
vivien.didelot@...oirfairelinux.com, davem@...emloft.net
Subject: Re: [PATCH 4/4] net: dsa: LAN9303: add MDIO managed mode support
On Thu, Apr 06, 2017 at 06:53:10AM -0700, Florian Fainelli wrote:
>
>
> On 04/05/2017 12:32 PM, Andrew Lunn wrote:
> >> +static const struct of_device_id lan9303_mdio_of_match[] = {
> >> + { .compatible = "smsc,lan9303" },
> >> + { /* sentinel */ },
> >> +};
> >> +MODULE_DEVICE_TABLE(of, lan9303_mdio_of_match);
> >
> > If i'm reading this right, i think you have the same compatible string
> > for both the i2c and the mdio driver. Does that work? I've no idea.
>
> This should not be a problem, actually, this does seem like the right
> thing to do. The bus topology would solve the Linux device creation such
> that you would probe either via mdio or i2c (or both) but through the
> appropriate drivers' probe functions.
Hi Florian
I was more thinking about what happens when there is a call out to
userspace to find a kernel module to load to driver this hardware. Is
the compatible string used for this? Is the subsystem string also used
when finding the correct kernel module?
Andrew
Powered by blists - more mailing lists