Re: [PATCH] Input: drivers/joystick: use parallel port device model

On Fri, Jul 31, 2015 at 01:43:06PM -0700, Dmitry Torokhov wrote:

On Fri, Jul 31, 2015 at 01:36:18PM -0700, Greg KH wrote:

quoted

On Fri, Jul 31, 2015 at 11:04:26PM +0530, Sudip Mukherjee wrote:

quoted

On Fri, Jul 31, 2015 at 09:54:27AM -0700, Dmitry Torokhov wrote:

quoted

On Fri, Jul 31, 2015 at 03:45:25PM +0530, Sudip Mukherjee wrote:

quoted

On Thu, Jul 30, 2015 at 09:53:23AM -0700, Dmitry Torokhov wrote:

quoted

Hi Sudip,

Hi Dmitry,

quoted

On Thu, Jul 30, 2015 at 07:36:34PM +0530, Sudip Mukherjee wrote:

<snip>

quoted

quoted

quoted

No, unfortunately I do not.

Since neither of us can test the change what is the benefit of doing the
conversion? What will be gained by doing it? Are there plans for parport
subsystem to remove the old style initialization?

Yes, that is the plan. Well, if you are not comfortable with introducing
attach and detach functions then this can be done in another way where
there will be very minimum change in the code. But I will prefer to have
attach and detach then it can take advantage of the hotplug feature.
Adding Greg in To: list for his comments.

Converting to the "new" api is the end goal here, no need to keep the
old one around anymore.

OK, then I guess we can do the conversion right (dropping db9_base
module-global) and see if anyone screams at us.

I am working on it now to remove db9_base. But in the detach callback we
will get struct parport * and from parport to get a pardevice we need to
get it from port->physport->devices. Since it is having PARPORT_FLAG_EXCL
it is ok, but should we really depend on and work with the internal data
structures of the parport rather than working with the exported api?

regards
sudip