Re: [opensuse-kernel] Re: Fwd: OBS HEAD kernel-desktop

Re: [opensuse-kernel] Re: Fwd: OBS HEAD kernel-desktop

[Daniel FAIVRE]

Jiri :  CONFIG_HID_*=m is THE solution !!! Please commit it as soon as 
possible !!! ;-)

> On Wed, 14 Oct 2009, Greg KH wrote:
> > > There's a problem with builtin hid-ntrig in kernel-desktop : the
> > > hid-ntrig is
> > > NOT working (stylus button do NOT work) : it would be a great idea NOT
> > > to embed hid-ntrig in the kernel, and to include it only as a module,
> > > until developments evolutes enough to get a correct basic support.
> > >
> > > Actually, there's no way to get it work except recompiling the whole
> > > kernel without builtin hid-ntrig ...
> >
> > So just pick one hid device as a module?  Sure, I don't object to doing
> > that if it fixes problems for people.
> >
> > What specific config option should be changed?
>
> Apparently we are having CONFIG_HID_*=y for -desktop flavour. I think it
> might be good idea to change it back to =m, to be in sync with other
> flavours, vanilla defaults, etc.

Yes !!! That's the best idea : as a workaround, i recompiled a kernel with 
 CONFIG_HID_*=m

>
> Daniel, could you please report the problems you are having with hid-ntrig
> upstream? That would mean sending the description of the misbehavior of
> the driver to
>
> 	Rafi Rubin <rafi@seas.upenn.edu>
> 	Stephane Chatty <chatty@enac.fr>
> 	Jiri Kosina <jkosina@suse.cz>
> 	linux-input@vger.kernel.org
>
> Thanks,

n-trig is known NOT to work with opensuse 11.1, due to missing support in 
2.6.27 kernel.
A workaround could be installing a new kernel from HEAD. Unfortunately, with 
2.6.31 kernel from openSuSE HEAD repository, the stylus button do NOT work.
(wacdump report stylus button click as a left mouseclick)
Some peoples suggest patches for kernels > 2.6.30
Rafi suggested a patch here : http://ofb.net/~rafi/latitude_xt.html
Will it fix the stylus button issue ? Nevertheless, a python script available 
on several Ubuntu forums show that the button is working properly, BUT remain 
unavailable for linuxwacom (which need to be patched too just to enable 1b96 
vendor and 0x01 and 0x02 devices). 

I will test patched hid-ntrig.ko tomorrow, with a recompiled kernel, and 
different patches.

Nevertheless, I'm interested to contribute both for solving these issue and 
for improving opensuse/linux tabletPC computing.

Under linux, we still lack some cool features available on windows, and that 
is a serious disability to promote our favourite OS on tabletPCs :
- right-click on delayed touch / stylus pressure, with visual feedback (may be 
possible with improved linuxwacom / xorg configuration, except for visual 
feedback ? I saw some xorg.conf with things like "longtouch" ... for evtouch 
driver ;-/ )
- multitouch (many things to do, from low-level drivers to KDE integration !)

And more difficult:
- handwriting recognition (cellwriter is far under the windows tool)
- speech recognition (embedded in windows since Vista !!!). Some engines 
exists, but not much anything else ...

One last thing, about hp-tx2 : does somebody get the three buttons under the 
fingerprint reader works and mapped to relevant actions under kde ???

Bests regards,
-- 
Daniel FAIVRE

Re: [opensuse-kernel] Re: Fwd: OBS HEAD kernel-desktop

[Rafi Rubin]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Daniel FAIVRE wrote:
> Jiri :  CONFIG_HID_*=m is THE solution !!! Please commit it as soon as 
> possible !!! ;-)
> 
>> On Wed, 14 Oct 2009, Greg KH wrote:
>>>> There's a problem with builtin hid-ntrig in kernel-desktop : the
>>>> hid-ntrig is
>>>> NOT working (stylus button do NOT work) : it would be a great idea NOT
>>>> to embed hid-ntrig in the kernel, and to include it only as a module,
>>>> until developments evolutes enough to get a correct basic support.
>>>>
>>>> Actually, there's no way to get it work except recompiling the whole
>>>> kernel without builtin hid-ntrig ...
>>> So just pick one hid device as a module?  Sure, I don't object to doing
>>> that if it fixes problems for people.
>>>
>>> What specific config option should be changed?
>> Apparently we are having CONFIG_HID_*=y for -desktop flavour. I think it
>> might be good idea to change it back to =m, to be in sync with other
>> flavours, vanilla defaults, etc.
> 
> Yes !!! That's the best idea : as a workaround, i recompiled a kernel with 
>  CONFIG_HID_*=m
> 
>> Daniel, could you please report the problems you are having with hid-ntrig
>> upstream? That would mean sending the description of the misbehavior of
>> the driver to
>>
>> 	Rafi Rubin <rafi@seas.upenn.edu>
>> 	Stephane Chatty <chatty@enac.fr>
>> 	Jiri Kosina <jkosina@suse.cz>
>> 	linux-input@vger.kernel.org
>>
>> Thanks,
> 
> n-trig is known NOT to work with opensuse 11.1, due to missing support in 
> 2.6.27 kernel.
> A workaround could be installing a new kernel from HEAD. Unfortunately, with 
> 2.6.31 kernel from openSuSE HEAD repository, the stylus button do NOT work.
> (wacdump report stylus button click as a left mouseclick)
> Some peoples suggest patches for kernels > 2.6.30
> Rafi suggested a patch here : http://ofb.net/~rafi/latitude_xt.html
> Will it fix the stylus button issue ? Nevertheless, a python script available 
> on several Ubuntu forums show that the button is working properly, BUT remain 
> unavailable for linuxwacom (which need to be patched too just to enable 1b96 
> vendor and 0x01 and 0x02 devices). 

The break in the stylus buttons was the result of minor mistake added with the conversion to a multitouch driver.  Jiri already
passed my most recent patches up to Linus, and I can confirm the fix is in HEAD.  I think its been in there since before 2.6.32-rc1.

My apologies for missing the bug with in the previous version, I really should have caught it immediately.  The error really was
just masking out the stylus button events during initialization, so quite easy to fix.

My website has no further patches for the kernel driver for the tablet at this time.

I did send in the patch to the linuxwacom project, but as of yet it has not been accepted.  The patch adds two things:
the resolution and adds the ntrig id to a predicate which otherwise rejects all non-wacom devices.  The resolution is not strictly
necessary since the run-time configuration provides sufficient options.  As for the manufacturer limiting, that was added only after
the initial ntrig patches to the kernel.  Given the circumstances it seems likely that the specific blocking is intentional.  While
I understand why they might prefer to limit the support of that driver, it works very well with my digitizer and I expect to
continue using it and supporting the id patch.  Thank you Ping for such a nice driver, it has been quite helpful.


Multitouch on the ntrig is another question.  It might just be a matter of initialization with the current firmware.  Stephane has
used multitouch with older firmwares (without extra init), but with those versions he was not able to use the stylus.

> I will test patched hid-ntrig.ko tomorrow, with a recompiled kernel, and 
> different patches.

You might want to just grab the hid-ntrig.c file from HEAD:
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob_plain;f=drivers/hid/hid-ntrig.c;hb=HEAD

> Nevertheless, I'm interested to contribute both for solving these issue and 
> for improving opensuse/linux tabletPC computing.
> 
> Under linux, we still lack some cool features available on windows, and that 
> is a serious disability to promote our favourite OS on tabletPCs :
> - right-click on delayed touch / stylus pressure, with visual feedback (may be 
> possible with improved linuxwacom / xorg configuration, except for visual 
> feedback ? I saw some xorg.conf with things like "longtouch" ... for evtouch 
> driver ;-/ )
> - multitouch (many things to do, from low-level drivers to KDE integration !)

I'd also like to see something that shows feedback without interfering.  I suspect the xtest extension might be a good place to
start looking.


Rafi
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAkrWvvgACgkQwuRiAT9o608Q8QCgjBaCulVWzvpWrId05I0U7FyZ
Fq8AoPQPlygv/LhRYTuhe4ROnL94ErXs
=1oMY
-----END PGP SIGNATURE-----

N-trig support

[Daniel FAIVRE]

Hi,

After our last mails about n-trig support, I have :

- reported a bug for  OBS HEAD kernel-desktop (some opensuse build service 
kernels do not include drivers as modules)
https://bugzilla.novell.com/show_bug.cgi?id=548445

- reported a bug for outdated n--trig.c , suggesting is replacement by the one 
from git as suggested by Rafi.
https://bugzilla.novell.com/show_bug.cgi?id=548463

- tested the last linuxwacom-dev from cvs, patched to support n-trig :
the touch support is far better with the dev version : a touch tap is a click 
far more easily than with the last release. Touch beacame very smooth, as 
much as comfortable as under windows !
Last commit on linuxwacom was "Support Tablet PCs 0xE2 and 0xE3 as well as 
touch only tablet 0x9F" : THE NTRIG PATCH IS STILL MISSING !!!

Multitouch support :
In wcmUSB.c, we can see that support for multitouch is now included, waiting 
an event BTN_TOOL_TRIPLETAP.
A new option for xorg.conf, "Gesture", is used. 
Actually, this could not work for n-trig because hid-ntrig.ko does NOT send 
any BTN_TOOL_TRIPLETAP event, but it may be something to investigate ?

Screen buttons support :
Among patches for linuxwacom on sourceforge, one is used to detect and convert 
events from these three buttons.

X crashes : I encounter some X crashes with wacom_drv.so :
The system log reports :
wacom: v1.49-pc-1:USB Wacom Graphire and Wacom Intuos tablet driver
[Xorg[2694]: segfault at 6a0000000f ip 00007fded576e5c5 sp 00007fff73602780 
error 4 in wacom_drv.so [7fded578000+18000]

Wacom X drivers compiled from last release or from cvs have the same issue. 
That happens always when quitting an X session, and freeze display and all 
input devices, including keyboard. That also sometimes happens at boottime 
when starting X.
When X crash after booting, logs always reports things like :
0 is not supported by linuxwacom or 
1F is not supported by linuxwacom ...
When closing the X session, it seems to be related to :
X(deleteInputDeviceRequest+0x3b) [0x47713b]

In both cases, this issue seems related to wrong device vendor/product used in 
linuxwacom. Any idea to fix that ?

Best regards,
-- 
Daniel FAIVRE

Re: N-trig support

[Rafi Rubin]

[-- Attachment #1: Type: text/plain, Size: 3986 bytes --]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Daniel FAIVRE wrote:
>
> - reported a bug for outdated n--trig.c , suggesting is replacement by the one
> from git as suggested by Rafi.
> https://bugzilla.novell.com/show_bug.cgi?id=548463
>
> - tested the last linuxwacom-dev from cvs, patched to support n-trig :
> the touch support is far better with the dev version : a touch tap is a click
> far more easily than with the last release. Touch beacame very smooth, as
> much as comfortable as under windows !
> Last commit on linuxwacom was "Support Tablet PCs 0xE2 and 0xE3 as well as
> touch only tablet 0x9F" : THE NTRIG PATCH IS STILL MISSING !!!

Yes, the tapping behavior improved, but after a personal request I took another look and have an experimental version which is
actually quite a bit better (for tapping).

I have not posted this yet because it still has a couple of odd behaviors and I haven't had time to sort them out.  First, sometimes
buttons get stuck down (not sure if its only the eraser or other stylus buttons).  Secondly, I'm seeing warping (pointer jumps from
a distant location).

I have not planned to spend time on it for at least another week, but if you find the tap improvement is important and you're trying
to prepare for a release, I will try to get this fixed this week.

> Multitouch support :
> In wcmUSB.c, we can see that support for multitouch is now included, waiting
> an event BTN_TOOL_TRIPLETAP.
> A new option for xorg.conf, "Gesture", is used.
> Actually, this could not work for n-trig because hid-ntrig.ko does NOT send
> any BTN_TOOL_TRIPLETAP event, but it may be something to investigate ?

As for multitouch I have not gone as considering the wacom support.  I have never seen events multitouch events come from my
digitizer to the kernel in linux.  Stephane did have that working, but only with an earlier version of the firmware.  I'm not sure
what he used for his demos, but it wasn't the wacom driver.

Our current working hypothesis is that we just need to pass the digitizer some commands during initialization.  I'd be happy for
help if anyone has the desire and the time to work on that.


> Screen buttons support :
> Among patches for linuxwacom on sourceforge, one is used to detect and convert
> events from these three buttons.

At least for the xt (4 buttons) and xt2 (3 buttons).  The screen buttons just send relatively normal key events, just with funny
codes.  I personally just map them at boot with setkeycodes:
	setkeycodes e01f 72     e01b 75 e023 77 e01e 80 e058 120 e018 121 e01a 122 e017 123
(the other four are arrow buttons on the side of the screen of the xt).

> X crashes : I encounter some X crashes with wacom_drv.so :
> The system log reports :
> wacom: v1.49-pc-1:USB Wacom Graphire and Wacom Intuos tablet driver
> [Xorg[2694]: segfault at 6a0000000f ip 00007fded576e5c5 sp 00007fff73602780
> error 4 in wacom_drv.so [7fded578000+18000]
>
> Wacom X drivers compiled from last release or from cvs have the same issue.
> That happens always when quitting an X session, and freeze display and all
> input devices, including keyboard. That also sometimes happens at boottime
> when starting X.
> When X crash after booting, logs always reports things like :
> 0 is not supported by linuxwacom or
> 1F is not supported by linuxwacom ...
> When closing the X session, it seems to be related to :
> X(deleteInputDeviceRequest+0x3b) [0x47713b]
>
> In both cases, this issue seems related to wrong device vendor/product used in
> linuxwacom. Any idea to fix that ?

I'm using an update from cvs from October 5.  I have not seen any crashes related to the wacom driver, nor have I seen any such
crashes in recent memory (at least a year).

Rafi
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAkre3w4ACgkQwuRiAT9o60/N0gCfazn/u2wIr1DoF3/2gHnv0PDu
fOUAn0nHfVhburgRKScDZStr68+q/GC+
=8XAC
-----END PGP SIGNATURE-----

[-- Attachment #2: hid-ntrig.c --]
[-- Type: text/x-csrc, Size: 8457 bytes --]

/*
 *  HID driver for N-Trig touchscreens
 *
 *  Copyright (c) 2008 Rafi Rubin
 *  Copyright (c) 2009 Stephane Chatty
 *
 */

/*
 * 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.
 */

#include <linux/device.h>
#include <linux/hid.h>
#include <linux/module.h>

#include "hid-ids.h"

#define NTRIG_DUPLICATE_USAGES	0x001

#define nt_map_key_clear(c)	hid_map_usage_clear(hi, usage, bit, max, \
					EV_KEY, (c))

struct ntrig_data {
	__s32 x, y, id, w, h;
	char reading_a_point, found_contact_id;
	char pen_active;
	char finger_active;
	char inverted;
	__s32 tool;
	char inrange;
};

/*
 * this driver is aimed at two firmware versions in circulation:
 *  - dual pen/finger single touch
 *  - finger multitouch, pen not working
 */

static int ntrig_input_mapping(struct hid_device *hdev, struct hid_input *hi,
		struct hid_field *field, struct hid_usage *usage,
		unsigned long **bit, int *max)
{
	switch (usage->hid & HID_USAGE_PAGE) {

	case HID_UP_GENDESK:
		switch (usage->hid) {
		case HID_GD_X:
			hid_map_usage(hi, usage, bit, max,
					EV_ABS, ABS_MT_POSITION_X);
			input_set_abs_params(hi->input, ABS_X,
					field->logical_minimum,
					field->logical_maximum, 0, 0);
			return 1;
		case HID_GD_Y:
			hid_map_usage(hi, usage, bit, max,
					EV_ABS, ABS_MT_POSITION_Y);
			input_set_abs_params(hi->input, ABS_Y,
					field->logical_minimum,
					field->logical_maximum, 0, 0);
			return 1;
		}
		return 0;

	case HID_UP_DIGITIZER:
		switch (usage->hid) {
		/* we do not want to map these for now */
		case HID_DG_CONTACTID: /* value is useless */
		case HID_DG_INPUTMODE:
		case HID_DG_DEVICEINDEX:
		case HID_DG_CONTACTCOUNT:
		case HID_DG_CONTACTMAX:
			return 0;

		/* original mapping by Rafi Rubin */
		case HID_DG_CONFIDENCE:
			nt_map_key_clear(BTN_TOOL_DOUBLETAP);
			return 1;

		/* width/height mapped on TouchMajor/TouchMinor/Orientation */
		case HID_DG_WIDTH:
			hid_map_usage(hi, usage, bit, max,
					EV_ABS, ABS_MT_TOUCH_MAJOR);
			return 1;
		case HID_DG_HEIGHT:
			hid_map_usage(hi, usage, bit, max,
					EV_ABS, ABS_MT_TOUCH_MINOR);
			input_set_abs_params(hi->input, ABS_MT_ORIENTATION,
					0, 1, 0, 0);
			return 1;
		}
		return 0;

	case 0xff000000:
		/* we do not want to map these: no input-oriented meaning */
		return -1;
	}

	return 0;
}

static int ntrig_input_mapped(struct hid_device *hdev, struct hid_input *hi,
		struct hid_field *field, struct hid_usage *usage,
		unsigned long **bit, int *max)
{
	if (usage->type == EV_KEY || usage->type == EV_REL
			|| usage->type == EV_ABS)
		clear_bit(usage->code, *bit);

	return 0;
}

/*
 * this function is called upon all reports
 * so that we can filter contact point information,
 * decide whether we are in multi or single touch mode
 * and call input_mt_sync after each point if necessary
 */
static int ntrig_event (struct hid_device *hid, struct hid_field *field,
		                        struct hid_usage *usage, __s32 value)
{
	struct input_dev *input = field->hidinput->input;
	struct ntrig_data *nd = hid_get_drvdata(hid);

        if (hid->claimed & HID_CLAIMED_INPUT) {
		switch (usage->hid) {

		case HID_DG_INRANGE:
			input_sync(input);
			if (nd->tool!= field->application)
			{
				if (field->application == HID_DG_PEN)
				{
					input_report_key(input, BTN_TOOL_DOUBLETAP, 0);
					input_report_key(input, nd->inverted ? BTN_TOOL_RUBBER : 
							BTN_TOOL_PEN, (value != 0));
				}
				else if(field->application == HID_DG_TOUCHSCREEN)
				{
					input_report_key(input, BTN_TOOL_RUBBER, 0);
					input_report_key(input, BTN_TOOL_PEN, 0);
					input_report_key(input, BTN_TOOL_DOUBLETAP, (value != 0));
				}
				else
				{
				}
				input_sync(input);
				nd->tool=field->application;
			}
			else
			{
				if((value != 0) && (nd->inrange == 0))
				{
					if (field->application == HID_DG_PEN)
					{
						input_report_key(input, nd->inverted ? BTN_TOOL_RUBBER : 
								BTN_TOOL_PEN, 1);
					}
					//HID_DG_TOUCHSCREEN
					else if(field->application == HID_DG_TOUCHSCREEN)
					{
						input_report_key(input, BTN_TOOL_DOUBLETAP, 1);
					}
				}
				input_sync(input);
			}
			nd->inrange= (value != 0);
			break;

		case HID_DG_INVERT:
			if (field->application == HID_DG_PEN && nd->inverted != value)
			{
				input_report_key(input,
						nd->inverted ? BTN_TOOL_RUBBER : BTN_TOOL_PEN
						, 0);
				input_report_key(input,
						value ? BTN_TOOL_RUBBER : BTN_TOOL_PEN
						, nd->inrange);
				input_sync(input);
			}
			nd->inverted = value;
			return 0;

		case HID_GD_X:
			nd->x = value;
			nd->reading_a_point = 1;
			break;
		case HID_GD_Y:
			nd->y = value;
			break;
		case HID_DG_CONTACTID:
			nd->id = value;
			/* we receive this only when in multitouch mode */
			nd->found_contact_id = 1;
			break;
		case HID_DG_WIDTH:
			nd->w = value;
			break;
		case HID_DG_HEIGHT:
			nd->h = value;
			break;
		case HID_DG_CONFIDENCE:
				input_event(input, EV_ABS, ABS_X, nd->x);
				input_event(input, EV_ABS, ABS_Y, nd->y);
				input_sync(input);
			return 0;
		case HID_DG_TIPPRESSURE:
			/*
			 * when in single touch mode, this is the last
			 * report received in a pen event. We want
			 * to emit a normal (X, Y) position
			 */
			if (! nd->found_contact_id) {
				input_event(input, EV_ABS, ABS_X, nd->x);
				input_event(input, EV_ABS, ABS_Y, nd->y);
				input_event(input, EV_ABS, ABS_PRESSURE, value);
				input_sync(input);
			}
			break;
		case 0xff000002:
			/*
			 * we receive this when the device is in multitouch
			 * mode. The first of the three values tagged with
			 * this usage tells if the contact point is real
			 * or a placeholder
			 */
			if (!nd->reading_a_point || value != 1)
				break;
			/* emit a normal (X, Y) for the first point only */
			if (nd->id == 0) {
				input_event(input, EV_ABS, ABS_X, nd->x);
				input_event(input, EV_ABS, ABS_Y, nd->y);
			}
			input_event(input, EV_ABS, ABS_MT_POSITION_X, nd->x);
			input_event(input, EV_ABS, ABS_MT_POSITION_Y, nd->y);
			if (nd->w > nd->h) {
				input_event(input, EV_ABS,
						ABS_MT_ORIENTATION, 1);
				input_event(input, EV_ABS,
						ABS_MT_TOUCH_MAJOR, nd->w);
				input_event(input, EV_ABS,
						ABS_MT_TOUCH_MINOR, nd->h);
			} else {
				input_event(input, EV_ABS,
						ABS_MT_ORIENTATION, 0);
				input_event(input, EV_ABS,
						ABS_MT_TOUCH_MAJOR, nd->h);
				input_event(input, EV_ABS,
						ABS_MT_TOUCH_MINOR, nd->w);
			}
			input_mt_sync(field->hidinput->input);
			nd->reading_a_point = 0;
			nd->found_contact_id = 0;
			break;

		default:
			/* fallback to the generic hidinput handling */
			return 0;
		}
	}

	/* we have handled the hidinput part, now remains hiddev */
        if (hid->claimed & HID_CLAIMED_HIDDEV && hid->hiddev_hid_event)
                hid->hiddev_hid_event(hid, field, usage, value);

	return 1;
}

static int ntrig_probe(struct hid_device *hdev, const struct hid_device_id *id)
{
	int ret;
	struct ntrig_data *nd;

	nd = kmalloc(sizeof(struct ntrig_data), GFP_KERNEL);
	nd->inrange=0;
	nd->tool=0;
	if (!nd) {
		dev_err(&hdev->dev, "cannot allocate N-Trig data\n");
		return -ENOMEM;
	}
	nd->reading_a_point = 0;
	nd->found_contact_id = 0;
	hid_set_drvdata(hdev, nd);

	ret = hid_parse(hdev);
	if (!ret)
		ret = hid_hw_start(hdev, HID_CONNECT_DEFAULT);

	if (ret)
		kfree (nd);

	return ret;
}

static void ntrig_remove(struct hid_device *hdev)
{
	hid_hw_stop(hdev);
	kfree(hid_get_drvdata(hdev));
}

static const struct hid_device_id ntrig_devices[] = {
	{ HID_USB_DEVICE(USB_VENDOR_ID_NTRIG, USB_DEVICE_ID_NTRIG_TOUCH_SCREEN),
		.driver_data = NTRIG_DUPLICATE_USAGES },
	{ }
};
MODULE_DEVICE_TABLE(hid, ntrig_devices);

static const struct hid_usage_id ntrig_grabbed_usages[] = {
	{ HID_ANY_ID, HID_ANY_ID, HID_ANY_ID },
	{ HID_ANY_ID - 1, HID_ANY_ID - 1, HID_ANY_ID - 1}
};

static struct hid_driver ntrig_driver = {
	.name = "ntrig",
	.id_table = ntrig_devices,
	.probe = ntrig_probe,
	.remove = ntrig_remove,
	.input_mapping = ntrig_input_mapping,
	.input_mapped = ntrig_input_mapped,
	.usage_table = ntrig_grabbed_usages,
	.event = ntrig_event,
};

static int __init ntrig_init(void)
{
	return hid_register_driver(&ntrig_driver);
}

static void __exit ntrig_exit(void)
{
	hid_unregister_driver(&ntrig_driver);
}

module_init(ntrig_init);
module_exit(ntrig_exit);
MODULE_LICENSE("GPL");