Re: [PATCH v9 1/8] drivers: phy: add generic PHY framework

[Greg KH <gregkh@linuxfoundation.org>]

On Wed, Jul 17, 2013 at 03:02:59PM +0530, Kishon Vijay Abraham I wrote:
> Hi,
> 
> On Wednesday 17 July 2013 11:59 AM, Greg KH wrote:
> > On Wed, Jun 26, 2013 at 05:17:29PM +0530, Kishon Vijay Abraham I wrote:
> >> +menuconfig GENERIC_PHY
> >> +	tristate "PHY Subsystem"
> >> +	help
> >> +	  Generic PHY support.
> >> +
> >> +	  This framework is designed to provide a generic interface for PHY
> >> +	  devices present in the kernel. This layer will have the generic
> >> +	  API by which phy drivers can create PHY using the phy framework and
> >> +	  phy users can obtain reference to the PHY.
> > 
> > Shouldn't this be something that other drivers select?  How will anyone
> > know if they need this or not?
> 
> All the PHY drivers should go here. So only if *GENERIC_PHY* is enabled those
> PHY drivers can be selected like in [1].
> The PHY consumer driver should ideally use *depends on* in their Kconfig.
> 
> [1] ->
> http://archive.arm.linux.org.uk/lurker/message/20130628.134308.4a8f7668.ca.html

No, switch it around the other way.  How would I even _know_ that I need
to enable the generic PHY subsystem in the first place?  How can I
determine that I need this for my hardware?  You need to give hints to
someone who doesn't even know what a PHY is, otherwise they will always
disable it and move on.

> >> --- /dev/null
> >> +++ b/drivers/phy/phy-core.c
> >> @@ -0,0 +1,544 @@
> >> +/*
> >> + * phy-core.c  --  Generic Phy framework.
> >> + *
> >> + * Copyright (C) 2013 Texas Instruments
> >> + *
> >> + * Author: Kishon Vijay Abraham I <kishon@ti.com>
> >> + *
> >> + * This program is free software; you can redistribute  it and/or modify it
> >> + * under  the terms of  the GNU General  Public License as published by the
> >> + * Free Software Foundation;  either version 2 of the  License, or (at your
> >> + * option) any later version.
> > 
> > You really mean "any later version" (I have to ask)?
> 
> That was copied from somewhere :-s

So, is that what you really mean to have for this code?

> >> + * This program is distributed in the hope that it will be useful,
> >> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> >> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> >> + * GNU General Public License for more details.
> >> + *
> >> + * You should have received a copy of the GNU General Public License
> >> + * along with this program.  If not, see <http://www.gnu.org/licenses/>.
> > 
> > Are these two paragraphs needed?  This isn't a "program", and they got a
> > copy of the GPL already with the kernel.
> 
> hmm.. I can remove that.
> > 
> >> +static struct class *phy_class;
> > 
> > Why do you need a class?
> 
> Wanted to group all the PHY drivers to be used by different subsystems
> (SATA/USB/PCIE/HDMI/VIDEO) into a single entity. There were some comments in my
> initial version [3] on using a bus_type instead of class but then it was
> decided to go with class itself.
> 
> [3] -> http://lkml.indiana.edu/hypermail/linux/kernel/1302.2/01389.html

Ok, but what does the class usage get you?

> > When modifying/adding new sysfs stuff, you need a Documentation/ABI/
> > entry as well.
> 
> I'm not actually adding any new sysfs entry other than what a *class_create*
> must have created. Do I need to add one for that?

If you are not creating anything in sysfs at all, why use the driver
model?  (hint, I think you need to do something here to justify it...)

thanks,

greg k-h