Input event for blocking touchpad

Input event for blocking touchpad

[Thadeu Lima de Souza Cascardo]

[-- Attachment #1: Type: text/plain, Size: 478 bytes --]

Hello,

I have a notebook with an ACPI system that notifies about the pressing
of the key for blocking touchpad. However, it does not notify it by
itself. I was considering sending the event to userspace and let it does
the blocking.

Since other keys are also notified by this same ACPI object, it's even
more natural that I use an input device to send this to userspace.

The problem is: there seems to be no event to represent this. Any
suggestions?

Best regards,
Cascardo.

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 836 bytes --]

Re: Input event for blocking touchpad

[Dmitry Torokhov]

On Fri, Nov 26, 2010 at 07:10:57PM -0200, Thadeu Lima de Souza Cascardo wrote:
> Hello,
> 
> I have a notebook with an ACPI system that notifies about the pressing
> of the key for blocking touchpad. However, it does not notify it by
> itself. I was considering sending the event to userspace and let it does
> the blocking.
> 
> Since other keys are also notified by this same ACPI object, it's even
> more natural that I use an input device to send this to userspace.
> 
> The problem is: there seems to be no event to represent this. Any
> suggestions?
> 

I have the patch below from Bastien that I will be queueing shortly.

Thanks.

-- 
Dmitry


Input: add keycodes for touchpad on/off keys

From: Bastien Nocera <hadess@hadess.net>

Some laptops will have a "touchpad toggle" soft button, which expects
user-space to turn off the touchpad themselves, some other devices will
do this in hardware, but send key events telling us that the touchpad
has been turned off/on.

KEY_TOUCHPAD_ON/KEY_TOUCHPAD_OFF will be used by user-space to show a
popup with the status of the touchpad.

Signed-off-by: Bastien Nocera <hadess@hadess.net>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
---

 include/linux/input.h |    4 ++++
 1 files changed, 4 insertions(+), 0 deletions(-)


diff --git a/include/linux/input.h b/include/linux/input.h
index 6c3c986..a8af21d 100644
--- a/include/linux/input.h
+++ b/include/linux/input.h
@@ -643,6 +643,10 @@ struct input_keymap_entry {
 #define KEY_CAMERA_FOCUS	0x210
 #define KEY_WPS_BUTTON		0x211	/* WiFi Protected Setup key */
 
+#define KEY_TOUCHPAD_TOGGLE	0x212	/* Request switch touchpad on or off */
+#define KEY_TOUCHPAD_ON		0x213
+#define KEY_TOUCHPAD_OFF	0x214
+
 #define BTN_TRIGGER_HAPPY		0x2c0
 #define BTN_TRIGGER_HAPPY1		0x2c0
 #define BTN_TRIGGER_HAPPY2		0x2c1

Re: Input event for blocking touchpad

[Bastien Nocera]

On Sun, 2010-11-28 at 00:07 -0800, Dmitry Torokhov wrote:
> On Fri, Nov 26, 2010 at 07:10:57PM -0200, Thadeu Lima de Souza Cascardo wrote:
> > Hello,
> > 
> > I have a notebook with an ACPI system that notifies about the pressing
> > of the key for blocking touchpad. However, it does not notify it by
> > itself. I was considering sending the event to userspace and let it does
> > the blocking.
> > 
> > Since other keys are also notified by this same ACPI object, it's even
> > more natural that I use an input device to send this to userspace.
> > 
> > The problem is: there seems to be no event to represent this. Any
> > suggestions?
> > 
> 
> I have the patch below from Bastien that I will be queueing shortly.

Note that KEY_TOUCHPAD_{ON,OFF} are for when:
1) Your touchpad disables in hardware
2) You can know for certain the status of the device

You'll need to use KEY_TOUCHPAD_TOGGLE in your patch (and make changes
to udev's keyboard remap scripts to map KEY_TOUCHPAD_TOGGLE to F21 until
X can get access to those keycodes.

Cheers

Re: Input event for blocking touchpad

[Zhang Rui]

On Sat, 2010-11-27 at 05:10 +0800, Thadeu Lima de Souza Cascardo wrote:
> Hello,
> 
> I have a notebook with an ACPI system that notifies about the pressing
> of the key for blocking touchpad. However, it does not notify it by
> itself. I was considering sending the event to userspace and let it does
> the blocking.
> 
how do you know the event is sent via ACPI?

please try "cat /proc/interrupts > interrupt-1;
grep . /sys/firmware/acpi/interrupts/* > gpe-1; sleep 5;
cat /proc/interrupts > interrupt-2;
grep . /sys/firmware/acpi/interrupts/* > gpe-2", and press the block key
during the 5 seconds sleep.
Then please attach the 4 files here to see if there are ACPI interrupts
generated by the key pressing.

thanks,
rui
> Since other keys are also notified by this same ACPI object, it's even
> more natural that I use an input device to send this to userspace.
> 
> The problem is: there seems to be no event to represent this. Any
> suggestions?
> 
> Best regards,
> Cascardo.

Re: Input event for blocking touchpad

[Thadeu Lima de Souza Cascardo]

[-- Attachment #1: Type: text/plain, Size: 2001 bytes --]

On Tue, Nov 30, 2010 at 09:34:13AM +0800, Zhang Rui wrote:
> On Sat, 2010-11-27 at 05:10 +0800, Thadeu Lima de Souza Cascardo wrote:
> > Hello,
> > 
> > I have a notebook with an ACPI system that notifies about the pressing
> > of the key for blocking touchpad. However, it does not notify it by
> > itself. I was considering sending the event to userspace and let it does
> > the blocking.
> > 
> how do you know the event is sent via ACPI?
> 
> please try "cat /proc/interrupts > interrupt-1;
> grep . /sys/firmware/acpi/interrupts/* > gpe-1; sleep 5;
> cat /proc/interrupts > interrupt-2;
> grep . /sys/firmware/acpi/interrupts/* > gpe-2", and press the block key
> during the 5 seconds sleep.
> Then please attach the 4 files here to see if there are ACPI interrupts
> generated by the key pressing.
> 
> thanks,
> rui

I know because I've already done some investigation and have even
written a driver for it. It's WMI to be more precise. After calling a
WMI method, it does start notification through the PNP0C14 object, using
the notify_id of a GUID. Then, I can get the key calling a query block
for a third GUID. I get values for all the keys whose functions were
missing.

I am about to send the driver soon. It requires the fix for WMI that
I've sent yesterday to work at all, since the GUIDs I use were being
wrongly considerated as duplicates.

Thanks for the tip, though. I'll be sure to use it to lookup the source
for key presses next time. In fact, I did end up looking at *gpe* at
/sys/ while doing this investigation. Now, I'm sure that's one of the
first places to look up while searching for ACPI notifications.

Thanks Rui.

Regards,
Cascardo.

> > Since other keys are also notified by this same ACPI object, it's even
> > more natural that I use an input device to send this to userspace.
> > 
> > The problem is: there seems to be no event to represent this. Any
> > suggestions?
> > 
> > Best regards,
> > Cascardo.
> 
> 

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 836 bytes --]