Re: [PATCH 1/1] HID: add have_special_driver hid module parameter

Hi Nikolai

On Fri, Mar 30, 2012 at 5:22 PM, Nikolai Kondrashov [off-list ref] wrote:

Hi Jiri,

On Fri, Mar 30, 2012 at 4:49 PM, Jiri Kosina [off-list ref] wrote:

quoted

On Mon, 20 Feb 2012, Nikolai Kondrashov wrote:

quoted

Add "have_special_driver" hid module parameter - a list of additional HID
devices which have specialized driver on HID bus. Needed to support out-of-tree
HID drivers.

I find this a little bit hackish :) How about for example the userspace
transport drivers for Bluetooth-LE that are currently in the works?

I like the idea and the implementation seems good, but I can't see how it
is related to the task of having out-of-tree HID drivers, sorry.

quoted

I'd very much prefer just using sysfs (bind/unbind) way of doing things if
possible for this purpose. What do you think?

To my shame I must admit I didn't know or forgot of this mechanism. Although
I did some research asking kernel developers at work about possibility to
rebind a different driver. They either forgot about it at the time too, or I
didn't ask properly.

I think, I can craft an udev rule which will detach a tablet from the
generic HID driver and attach it to my driver through sysfs.

Please forget about this patch for now, I will resume bothering you in case
it's not possible or not appropriate :)

Rebinding works perfectly well for me. I use it all the time when
testing HID drivers. Furthermore, there is a dynid-implementation in
the kernel so when hacking on something like this I'd recommend using
this. It would also remove the hard limit that you currently have.

Thanks!

Sincerely,
Nick

Regards
David