[PATCH 1/3] usb: host: st-hcd: Add USB HCD support for STi SoCs

[stern@rowland.harvard.edu (Alan Stern)]

On Mon, 14 Jul 2014, Lee Jones wrote:

> On Fri, 11 Jul 2014, Alan Stern wrote:
> > On Fri, 11 Jul 2014, Peter Griffin wrote:
> > > On Thu, 10 Jul 2014, Alan Stern wrote:
> > > 
> > > > On Thu, 10 Jul 2014, Peter Griffin wrote:
> > > > 
> > > > > This driver adds support for the USB HCD present in STi
> > > > > SoC's from STMicroelectronics. It has been tested on the
> > > > > stih416-b2020 board.
> > > > 
> > > > This driver file, along with the Kconfig changes, belongs in the
> > > > arch/platform-specific source directory.  Not in drivers/usb/host/.  
> > > > It is, after all, a platform driver that registers two platform
> > > > devices.
> > > 
> > > Why do think that?
> > 
> > Because it is true.  One can easily see that st-hcd.c is a platform
> > driver: It contains a module_platform_driver() line and a struct
> > platform_driver definition near the end.  And it registers a platform
> > device by calling platform_device_add() in st_hcd_device_create(),
> > which is called twice by st_hcd_probe().
> 
> You are correct that this is indeed a platform driver and in the 'old
> days' these would have been stuffed into the ARM sub-architecture
> directories.  However, these became far too bloated and created too
> much churn which angered Linus.  A great deal has changed since his
> ARM rant [1].  All drivers (platform or otherwise) are now to live in
> 'drivers', which makes a great deal of sense really, doesn't it?
> 
> Did you grep kernel wide for module_platform_driver()?
> 
> $ git grep module_platform_driver -- arch/ | wc -l
> 12
> 
> $ git grep module_platform_driver -- drivers/ | wc -l
> 1138
> 
> Most platform drivers have already been moved.

Okay, I grant the point.  Objections withdrawn.  Fix up the spelling 
error and the other stuff in the Kconfig help text, and this will be 
acceptable.

Alan Stern